United States Industrial Environmental Research
Environmental Protection Laboratory
Agency Research Triangle Park NC 27711
EPA-600/7-78-113
June 1978
&ER&
Procedures Manual
for Fabric Filter
Evaluation
Interagency
Energy/Environment
R&D Program Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the
effort funded under the 17-agency Federal Energy/Environment Research and
Development Program. These studies relate to EPA's mission to protect the public
health and welfare from adverse effects of pollutants associated with energy sys-
tems. The goal of the Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology. Investigations include analy-
ses of the transport of energy-related pollutants and their health and ecological
effects; assessments of, and development of, control technologies for energy
systems; and integrated assessments of a wide range of energy-related environ-
mental issues.
EPA REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved
for publication. Approval does not signify that the contents necessarily reflect
the views and policies of the Government, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/7-78-113
June 1978
Procedures Manual
for
Fabric Filter Evaluation
by
Kenneth M. Gushing and Wallace B. Smith
Southern Research Institute
2000 Ninth Avenue, South
Birmingham, Alabama 35205
Contract No. 68-02-2131
W.A. 21104
Program Element No. EHE624
EPA Project Officer: D. Bruce Harris
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
TABLE OF CONTENTS
Page
Abstract x
Acknowledgements xi
1. INTRODUCTION 1
1.1. FABRIC FILTER INSTALLATIONS 2
1.1.1. Particle Filtering Mechanisms 2
1.1.2. Factors Affecting Filter Performance . . 9
1.1.3. Filter Fabrics 19
1.1.4. Types of Fabric Filters 24
1.2. PARTICULATE SAMPLING FOR FABRIC FILTER EVALUA-
TION 28
1.2.1. General Considerations 28
1.2.2. Particulate Mass Measurements 31
1.2.3. Particle Sizing Techniques 32
2. TECHNICAL DISCUSSION 33
2.1. MECHANICAL CHARACTERIZATION OF A FABRIC FILTER
INSTALLATION 33
2.1.1. Mechanical Design and Operating Data . . 34
4.L.2. The Fabric Filter Bags 35
2.1.3. Filter Fabrics 37
2.1.4. Dust Removal Systems 39
2.1.5. Baghouse Operation-General Maintenance
Considerations 39
2.2. MASS EMISSION MEASUREMENTS 46
2.2.1. General Discussion 46
2.2.2. EPA-Type Particulate Sampling Train
(Method 5) 47
2.2.3. ASTM-Type Particulate Sampling Train . . 49
2.2.4. ASME-Type Particulate Sampling Train . . 49
2.2.5. General Sampling Procedures 51
2.3. PARTICLE SIZE MEASUREMENT TECHNIQUES 54
2.3.1. General Discussion 54
2.3.2. Inertial Particle Sizing Devices .... 56
2.3.3. Optical Measurement Techniques 73
2.3.4. Ultrafine Particle Sizing Techniques . . 75
11
-------
TABLE OF CONTENTS (Cont'd.)
Page?
2.4. PROCESS EFFLUENT GAS ANALYSIS
2.4.1. General Discussion
2.4.2. Qualitative Gas Analysis
2.4.3. Quantitative Gas Analysis
3. DEVELOPMENT OF TEST PLANS FOR FABRIC FILTER EVALUATIONS . 87
3.1. OBJECTIVES OF CONTROL DEVICE TESTS ......... 87
3.2. TYPE AND NUMBER OF TESTS REQUIRED ......... 88
3.2.1. Fabric Filter Level A Evaluation ...... 93
3.2.1.1. Plant Operating Data ....... 93
3.2.1.2. Baghouse-Fabric Filter Design
Data .............. 94
3.2.1.3. Flue Gas Characteristics, Bag-
house AP, Maintenance Data ... 94
3.2.2. Fabric Filter Level B Evaluation ...... 98
3.2.2.1. Quantitative Gas Analysis .... 98
3.2.2.2. Inlet and Outlet Mass Concentra-
tion Measurements Total Mass
Collection Efficiency ...... 99
3.2.3. Fabric Filter Level C Evaluation ...... 99
3.3. GENERAL PROBLEMS AND CONSIDERATIONS . ....... 102
APPENDICES
Appendix A - AEROSOL FUNDAMENTALS, NOMENCLATURE, AND DEFINI-
TIONS ......... . ........... 107
Appendix B - PARTICULATE MASS CONCENTRATION MEASUREMENTS. . . 129
Appendix C - CASCADE IMP ACTOR SAMPLING TECHNIQUES ...... 188
Appendix D - SIZE DISTRIBUTIONS OF SUBMICRON AEROSOL PARTI-
CLES ..................... 250
Appendix E - SUMMARY OF SOURCE PERFORMANCE METHODS ...... 302
Appendix F - FEDERAL STATIONARY SOURCE PERFORMANCE STANDARDS. 420
Bibliography ......................... 426
111
-------
LIST OF FIGURES
Figure No. Page
1 Typical air flow—fabric filter configura-
tions 3
2 Baghouse air flow and collection diagram .... 4
3 Typical simple fabric filter baghouse design . . 5
4 Typical pulse-jet baghouse with screw conveyor
dust removal system 6
5 Types of dust filtration 8
6 Resistance changes during fly ash filtration
for well-used fabrics 10
7 Calculated effluent concentration vs. time for
fly ash and ambient air dust based on measure-
ments with an optical particle counter 13
8 Effluent concentration vs. filter fabric
loading 14
9 Effluent concentration vs. filter fabric load-
ing and particle size. . 15
10 Penetration as a function of particle diameter
and coal-fired boiler load. . 16
11 Average outlet mass concentration (with one
standard deviation limits) as a function of air-
to-cloth ratio. 17
12 Penetration (with one standard deviation limits)
as a function of air-to-cloth ratio 18
13 Typical filter cloth weaves 21
14 Basic fabric weaves used for woven filter bags . 22
15 Residual dust left after several periods of
shaking of a fabric filter bag 25
16 Reverse air flexing to clean dust collector bags
by repressuring 26
-------
LIST OF FIGURES (Cont'd.)
Figure No. Pa9e
17 Typical reverse—pulse baghouse during
cleaning cycle
18 Cleaning methods for filter bags 29
19 Types of fabric filter systems depending on
cleaning method 30
20 EPA Method 5 Particulate Sampling Train. ... 48
21 ASTM Type Particulate Sampling Train 50
22 Typical gas velocity distribution at the inlet
to a baghouse 53
23 Operation principle and typical performance
for a cascade impactor 58
24 Andersen Mark III Stack Sampler 60
25 Modified Brink Model BMS-11 Cascade Impactor . 61
26 MRI Model 1502 Inertial Cascade Impactor ... 62
27 Sierra Model 226 Cascade Impactor 63
28 University of Washington Mark III Source Test
Cascade Impactor 64
29 Schematic of the Source Assessment Sampling
System 67
30 Three Stage Series Cyclone System 69
31 Five Stage Series Cyclone System 70
32 Laboratory Calibration for the Five Stage
Series Cyclone System 71
33 Operating principle for an optical particle
counter 74
v
-------
LIST OF FIGURES (Cont'd.)
Figure No. Page
34a Parallel plate diffusion battery 77
34b Parallel plate diffusion battery penetration
curves for monodisperse aerosols (12 channels,
0.1 x 10 x 48 cm) 77
35 Screen type diffusion battery 78
36 The electric mobility analyzer principle. ... 80
37 Flow schematic and electronic block diagram of
the Electrical Aerosol Analyser 81
38 Schematic diagram of apparatus for the collec-
tion of S03 by the condensation method .... 85
Al Examples of frequency or particle size dis-
tributions 109
A2 A single particle size distribution presented
in four ways 113
A3 Size distributions plotted on log probability
paper 114
Bl Thermocouple junction 131
B2 Standard pitot tube 134
B3 S-type pitot tube 135
B4 Probe and nozzle assembly 139
B5 Set-up for calibration of dry gas meter and
orifice meter 143
B6 Data sheet for calibrating dry gas meter and
orifice meter 144
B7 Minimum number of traverse points per sample
obtained from "Distances to Disturbances",
upstream and downstream 152
B8 Examples of equal area sample points 154
B9 Percent water vapor in air at saturation. . . . 159
BIO Percent water vapor with wet and dry bulb . . . 160
VI
-------
LIST OF FIGURES (Cont'd.)
Figure No. Pa9§.
Bll Isokinetic flow patterns 163
B12 Stack Sampling Nomograph (side 1) 166
B13 Stack Sampling Nomograph (side 2) 167
Cl Presurvey sampling with a cascade impactor . . 189
C2 Typical sample train with a heated impactor . . 193
C3 Nomograph for selecting nozzles for isokinetic
sampling 202
C4 Nomograph for sampling time selection (50 mg
sample) 207
C5 Hypothetical particle size distribution at a
fabric filter outlet determined from Andersen
impactor data 245
C6 Hypothetical particle size distribution at a
fabric filter outlet determined from Andersen
impactor data 246
C7 Hypothetical particle size distribution at a
fabric filter inlet determined from Brink
impactor data 247
C8 Hypothetical fabric filter fractional efficien-
cy based on the data presented in Figures C5
and C7 248
Dl Probe losses due to settling and diffusion
for spherical particles having a density of
2.5 gm/cm3 under conditions of laminar flow. . 252
D2 Diffusional adsorption apparatus for removal
of H20 from sample aerosol 255
D3 Sample Extraction-Dilution System (SEDS).... 258
D4 Sample Extraction Diluter, cut-away view. . . . 260
D5 Calculated dilution versus true dilution for
the Southern Research Institute Ultrafine
Particle Diluter, 0.092 ym particles 261
VII
-------
LTST OF FIGURES (Cont'd.)
Figure No. Page
D6 Calculated dilution versus true dilution for
the Southern Research Institute Ultrafine
Particle Diluter, 0.15 um particles 262
D7 Parallel Plate Diffusion Battery 265
D8 Screen type diffusion battery 268
D9 Theoretical parallel plate diffusion battery
penetration curves 285
D10 Schematic diagram of the electrical aerosol ana-
lyzer 286
Dll Hypothetical inlet size distribution at a fabric
filter on a coal-fired boiler 300
VI 11
-------
LIST OF TABLES
Table No.
I Summary of physical and chemical properties
of industrial filter fibers
II Type and frequency of fabric filter problems
reported
Ill Commercial cascade impactor sampling system ...
IV Particulate control device tasks '1
V Three levels of effort for fabric filter evalua-
tion
92
VI Maintenance Report - Type and Frequency of
Problems 96
Bl Duct traversing length factors 153
B2 Example - Particulate source test data -
ESP Inlet 180
Cl Impactor decision making 191
C2 Sampling information required 212
C3 Sample Calculation - Input Data and Results . . . 232
C4 Program flow 249
Dl Diffusion battery sample calculation data - CNC . 275
D2 Program 1 - Particle Diffusivity (Df) 279
D3 Program 2 - Particle Penetration (p) 282
D4 EAA (Model 3030) Data Reduction Form 291
D5 EAA Current Readings (I, in picoamps and Dilution
Factors) 294
D6 EAA (Model 3030) Data Reduction Form 297
D7 EAA (Model 3030) Data Reduction Form 298
IX
-------
ABSTRACT
The purpose of this procedures manual is to describe methods
to be used in experimentally characterizing the performance of
fabric filters for pollution control. A detailed description
of the mechanical characteristics of fabric filters is presented.
Procedures are described for measuring the particle size distri-
bution, the mass concentration of particulate matter, and the
concentration of major gaseous components of the flue gas-particle
mixture. A concise discussion and outline is presented which
describes the development of a test plan for the evaluation of
a fabric filter installation. By following this outline useful
tests may be performed which range in complexity from qualitative
and relatively inexpensive to rather elaborate research programs.
-------
ACKNOWLEDGEMENTS
Members of the Environmental Engineering Division of
Southern Research Institute who contributed to the prepara-
tion of this report include Mr. W. R. Dickson, Research Chemist
and Miss Annette Duncan, Assistant Physicist.
We also appreciate the assistance and guidance of our
Project Officer, Mr. D. Bruce Harris.
XI
-------
1. INTRODUCTION
Many different types of measurements must be made in order
to accurately evaluate the performance of a fabric filter (bag-
house)* installed to remove suspended particulate matter from
an industrial process stream. Among the required measurements
are determinations of the compositions of the gas and suspended
particles and of the particles' concentration and size distribu-
tion. Also, the baghouse geometry and operating parameters must
be recorded for proper interpretation of the measurements.
This document provides information and guidelines for use
in planning and conducting tests to obtain the necessary data.
A brief description of fabric filters and various evalua-
tion methods is provided in the remainder of this section. In
Section 2 the methods of measuring the fabric filter operating
parameters and the technical background and procedures for flue
gas and particulate characterization are discussed. Section 3
describes the logic and procedures to be used in developing a
test plan for the evaluation of a fabric filter installation.
The Appendices contain detailed information on the test methods,
as well as a listing of the Federal Stationary Source Performance
Standards and Federal Stationary Source Testing Reference Methods,
*In this manual the terms fabric filter and baghouse are used
interchangeably throughout the text.
-------
1.1 FABRIC FILTER INSTALLATIONS
A fabric filter consists of a porous, flexible layer of tex-
tile material through which a dusty gas is passed to separate
particles from the gas stream. Schematic representations of
typical bag-air flow configurations are shown in Figures 1 and
2. As particles accumulate on the fabric, resistance to gas flow
increases. Deposits are removed periodically by vigorous cleaning
of the cloth to maintain the pressure drop across the filter
within practical operating limits.
A baghouse is a large metal structure divided into two func-
tional areas which holds the fabric filter bags. The gas inlet
section may be part of the baghouse structure proper or it may
be designed as a gas distribution manifold. The function of this
dirty air plenum is to distribute evenly the gas to the fabric
filter bags.
The gas outlet section is designed for the recombination
of clean air exiting each of the fabric filter elements. The
medium separating the clean and dirty gas areas is the fabric
filter formed either as tubular bags or flat sheets. The hop-
pers that collect the particulate matter periodically dislodged
from the fabric filters are usually designed as part of the dirty
air plenum. Two typical baghouse designs are depicted in Fig-
ures 3 and 4.
1.1.1 Particle Filtering Mechanisms
Three distinct periods of fabric filter dust filtration can
be identified (Koscianowski and Koscianowski, 1976). The start
up phase is initiated with the unused fabric and is completed
with the first cleaning of the filters. The transition phase
begins when the filters are cleaned for the first time, and is
completed when the remaining fabric dust load remains stabilized
-------
0
ENVELOPE OR FRAME TYPE
UP, DOWN, OR THROUGH FLOW
CYLINDRICAL TYPES
OUTSIDE
FILTERING
INSIDE
FILTERING
NORMAL
(UPWARD)
FLOW
u
X
1
/
1
'•I
,-k
\
DOWN
FLOW
1 _
"X ~"N
X
\ I
\
1
I
1
\
/
r~3
U
"**^\
/
V
_I
(TUBE)
Figure 1. Typical air flow—fabric filter configurations.
-------
Figure 2. Baghouse air flow and collection diagram.
-------
CLEAN AIR
OUTLET
DIRTY AIR
INLET
CLEAN AIR
SIDE
FILTER
BAGS
CELL PLATE
Figure 3. Typical simple fabric filter baghouse design.
-------
Figure 4. Typical pulse-jet baghouse with screw conveyor dust
removal system.
-------
after each cleaning cycle. The final phase is characterized by
the fabric filters being filled with particulate matter and having
a steady value of pressure drop and dust load immediately after
each cleaning cycle. Figure 5 illustrates these three types of
fabric filtering.
During the start-up phase of filtering the collection mech-
anisms of impaction, interception, and diffusion on and within
the filter fibers operate (Dorman, 1966 and Billings and Wilder,
1970). Fabric materials woven from staple yarns (cotton, wool,
etc.) capture particles by single projecting fibers within the
air flow field, such as the interyarn spaces or pores. Smooth
surface yarns (fiberglass and synthetic materials) have deep
tunnel-like pores with few of these projecting fibers. Dust depo-
sition occurs within these pores. Dust collected by the fabric
filter builds up initially near these interyarn spaces. As it
builds, a dust layer or cake is formed through which most of the
flow must pass. This dust layer filtration continues until the
pressure drop reaches a point where cleaning is initiated.
During the transition phase of filtration the residual dust
load within the fabric accumulates. After each cleaning, the
fabric is never returned to its unused condition since dust in
at least a monolayer remains on the fibers. After each succes-
sive cleaning cycle, more dust is left in the fabric until this
residual dust load remains constant. The length of time for this
to occur is a function of several variables which include weave
pattern, pore-size, air-to-cloth ratio (air flow divided by filter
area), frequency of cleaning, method of cleaning and particle
size distribution.
After the transition phase of filter life is complete, the
final phase begins. At the beginning of each filtering cycle,
the fabric pores are coated with dust. This residual dust plays
an important part in the rapid formation of a new dust layer or
-------
REGENERATION
REGENERATION
DUST
FILTRATION
START UP PHASE
DUST
FILTRATION
TRANSITION PHASE
DUST
FILTRATION
FINAL PHASE
STABILIZED VALUE
OF PRESSURE DROP
• INCREASED PRESSURE DROP
AFTER REGENERATION
TIME
Figure 5. Types of dust filtration.
-------
filter cake. Filtration begins with the capture of single par-
ticles by particles in the residual dust layer. These mechanisms
are impaction, interception, and diffusion. These new particles
act as obstructions and aid in the capture of still more parti-
cles. A series of particle aggregate chains project into the
gas stream. Further accumulation occurs until the chains become
long enough to bridge most of the interyarn spaces and the filter
cake then is formed. During formation of the filter cake the
pressure drop increases non-linearly with time- After the cake
has been formed the pressure drop increases linearly with time
(Billings and Wilder, 1970). Figure 6 shows this type of behavior
for four different fabrics. This final phase filtering descrip-
tion applies to shake or reverse-air cleaning systems, but not
to pulse-jet situations.
There is no comprehensive theory to describe the actual
interrelationship or operation of these physical processes, how-
ever, Dennis, Cass, and Langley (1975) in their recent model
studies of fabric filter collection assume that the flow and
collection by the dust layer cake are like those in nuclepore
and membrane filters with low porosity- Their model also makes
the assumption that some of the interyarn pores always stay open
during the cleaning and filtering cycles.
1.1.2 Factors Affecting Filter Performance
Several factors can influence fabric filter performance.
These include fabric structure, air-to-cloth ratio, maximum pres-
sure drop before cleaning, method of cleaning, cleaning frequency,
intensity of cleaning, and flue gas temperature and humidity.
These factors can be grouped into two areas. One is those af-
fecting fabric filter design and the other is those factors which
relate to the system as installed.
During the design phase questions of type of fabric, method
of cleaning, air-to-cloth ratio, operating temperature and hu-
midity should be considered. Once installed, the cleaning rate,
-------
01
o
LU
DC
A NAPPED COTTON
D UNMAPPED COTTON
• CROWFOOT DACRON
O PLAIN WEAVE DACRON
Figure 6. Resistance changes during fly ash filtration for well-used fabrics.
Inlet loading-3.5 gr/ft3, filter velocity-3 ft/min, 30 min filter
cycle. After Billings and Wilder(1970).
10
-------
duration, intensity of cleaning and other operating variables
are adjusted to keep the system at peak efficiency.
The collection efficiency can also be affected by several
other occurrences. Bags can be blinded (that is, clogged) during
any filtering phase. Temperature excursions through acid dew
points may cause condensation with resulting damage to the bag-
house structure and adverse affects on the bag life and perform-
ance. Also, air preheater failures at utility boilers can cause
temperature excursions above safe limits.
Fabric filter performance is not affected by some of the
variables that are crucial for the success of scrubbers and elec-
trostatic precipitators. Woven fabric filters are relatively
insensitive to inlet dust loadings. Also, dust resistivity is
of little consequence since electrostatic attraction is not the
main method of removal. Particle wettability, necessary for good
scrubber performance, is not important in fabric filters.
The insensitivity of fabric filters to inlet dust loading
is well known (Billings and Wilder, 1970). A laboratory study
(Dennis and Wilder, 1975) determined that the average mass emis-
sion and its size distribution may be almost independent of the
loading and distribution of the inlet dust for a specified dust/
collector combination. The authors concluded that there was no
simple relationship between typical inlet and outlet concentra-
tions for most filter systems. This insensitivity of outlet
loadings is understandable if a majority of the penetration is
due to seepage. If this is the case, then the majority of emis-
sions consist of previously deposited dust in the fabric.
The pressure drop across a baghouse is a factor in its per-
formance. An operating pressure of 3-4 inches of water is typi-
cal for baghouses installed on utility boilers. Some industrial
boilers, however, operate in excess of ten inches of water pres-
11
-------
sure drop. When filtration begins the pressure drop increases
non-linearly until a cake is formed, after which it increases
linearly with time. This behavior is illustrated in Figure 6.
Figure 7 shows that directly after a cleaning cycle, the
filter efficiency drops to levels similar to those found in most
precipitators (Dennis and Wilder, 1975). Penetration can either
increase or decrease depending on the type of filter material,
construction, and dust properties. Figures 8 and 9 illustrate
two cases where penetration leveled off and remained fairly con-
stant for seasoned fiberglass bags (Dennis, Cass, and Langley,
1975) .
Fabric filter penetration increases rapidly for increasing
air-to-cloth ratio or face velocity. Figures 10, 11, and 12 il-
lustrate this performance characteristic for a utility boiler
baghouse (Ensor, Hooper, and Scheck, 1976). Figure 10 shows bag-
house penetration as a function of particle size and air-to-cloth
ratio. Here the penetration of particles larger than 1.0 microm-
eter diameter increases with an increase in air-to-cloth ratio.
This may be due to an increase in seepage of larger agglomerated
particles. Such an increase has been measured during recent bag-
house performance tests (Cass and Bradway, 1975). Figure 11 shows
typical outlet mass concentrations as a function of boiler load.
At lower air-to-cloth ratios, particles with larger diameters
show a reduced penetration as shown in Figure 12.
The presence of pinhole inclusions in filter dust cakes can
cause a degradation in performance. Pinholes are observed through
collected dust surfaces, and occur for both woven and filtered
fabrics (Holland and Rothwell, 1977). It has been thought that
the increase in penetration in the 2 ym to 5 ym range of particle
sizes is a direct result of pinhole leaks that allow coarser par-
ticles to pass through (Dennis, Cass, and Langley, 1975).
12
-------
1000 —
A = ROOM AIR FILTRATION
<10'4 GRAINS/FT3
B
-------
up
b
x
CO
o
h-
Z
111
o
o
o
UJ
_J
O
D
2
—i r
MEASUREMENTS BY CONDENSATION
NUCLEI COUNTER
99.12% AVERAGE WEIGHT COLLECTION EFFICIENCY
0.1
0.2
FABRIC LOADING, W. Ib/ft2
3726-024
Figure 8. Effluent concentration vs. filter fabric loading. Sunbury, PN
coal-fired boiler fly ash. After Dennis, Cass, and Langley(1975).
14
-------
101
up
b
r~
X
CO
f
\ITRATION, no./n
«k
%
LU
O
^f
0
O
LU
_J
O
H
OC
°- 10'1
n
x I I
^ MEASUREMENTS BY
BAUSCH & LOMB
SINGLE PARTICLE COUNTER
•
Q
A ••••*••• «
Q 4 SYMBOL SIZE, /jm
O • >0.3
0 A >0.5
"""* ^ ^1,0 "^
O >2.0
^ >3.0
x x ^
A
A
A AA A X Ax A0
— x x _
O
O
O
1 1
0.1 0.2
FABRIC LOADING, W, Ib/ft2
3726-025
Figure 9. Effluent concentration vs. filter fabric loading and particle size.
Used Nucla, Colorado baghouse fabric and fly ash. After Dennis,
Cass. and Langley(1975).
15
-------
CO
O
LLJ
in
O
I
C3
<
CO
I
C3
O
DC
I
I-
<
cc
•z.
LU
Q.
t-
o
—I—
SYMBOL
D
O
LOAD AIR-TO-CLOTH RATIO
MW ft3/min/ft2
6 1.87
11 2.47
12 2.74
0.1
1.0
PARTICLE DIAMETER, specific gravity = 2.0 g/cm3 , /mi
10.0
3726-028
Figure 10. Penetration as a function of particle diameter and coal-fired
boiler load. After Ensor, Hooper, and Scheck(1976).
16
-------
CO
16
15
14
13
12
111
110
I 9
<
F 8
u
o
u
V)
1
u
o
SYMBOL
O
D
A
LOAD
MW
6
11
12
I
INLET GRAIN
LOADING, gr/ft3
0.939
3.19
1.35
I
0 1.0 2.0
AIR-TO-CLOTH RATIO, ft/min
Figure 11. Average outlet mass concentration (with one standard
deviation limits) as a function of air-to-cloth ratio.
After Ensor(1976).
3.0
17
-------
UJ
Q
i
<
i
D
o
CC
X
l_
^^
z
o
<
OC
h-
UJ
z
UJ
o.
^
^
CO
OC
a.
2
^
X
I
^2
UJ
z
5
cc
Ul
111
a
14
13
12
11
10
9
8
7
6
5
4
3
2
1
n
I I I I
PRESSURE
DROP
BETWEEN
~~ LOAD CLEANING CLEANING
SYMBOL MW FREQUENCY IN H2O
O 6 NONE 3.0
_ D 11 HOURLY 3-4.5
A 12 HOURLY-CONTINUOUS 3-45
—
_
—
^_
—
—
— -j- _
— T*— " ~~^^
—
•
—
—
"""
f H
/I
/
/
/
/
/
/
/ j
-
r /
/
/
1
—
~ DESIGN SIX COMPARTMENTS ~~
1 1 1 1
1
99.86
99.87
99.88
99.89
99.90
99.91
99.92
99.93
99.94
99.95
99.96
99.97
99.98
99.99
ss
U
Ul
U
LL.
UJ
0.5
1.0 1.5 2.0
AIR-TO-CLOTH RATIO
BETWEEN COMPARTMENT CLEANING, acfm/ft2
2.5
3.0
Figure 12. Penetration (with one standard deviation limits) as a
function of air-to-cloth ratio. After Ensor(1976).
18
-------
There is some agreement as a result of these findings that
emissions from fabric filters are the result of indirect fault
processes such as seepage or pinhole leaks, or both, in contrast
to failure to collect these particles initially. If this is the
case then the following penetration characteristics can be ex-
plained (Oglesby, 1977).
a. Greater penetration as a result of higher face velocity can
be understood to result from an increased incidence of pin-
holes at high velocity, as well as from increased pressure
drop which may lead to increased seepage.
b. Constant penetration as the dust cake thickens can be under-
stood as a result of the inability of the pinholes to seal
after they have been formed. The high air flow rate through
these pinholes may tend to keep them open, unless large par-
ticles are present.
c. Continuously decreasing penetration is explainable if, as
the dust cake builds up, the fabric and dust do not inter-
act to create pinholes.
d. Because these pinholes have no fractionating capability,
constant penetration for all particle sizes is explained if
particles pass through these holes.
e. Considerable penetration due to seepage may explain the in-
sensitivity of outlet emission rates to inlet concentrations
of particles.
1.1.3 Filter Fabrics
Two basic types of fabric cloth are used in fabric filters.
They are woven cloth and felted cloth.
19
-------
Woven fabric filters in conventional baghouses usually have
air-to-cloth ratios of 1:1 to 5:1. Woven fabric permeability
can be varied, thus changing the operating air-to-cloth ratio.
Woven fabric permeability, or porosity, is varied by using dif-
ferent yarns, fabric count, cloth weights, and weave patterns.
The three basic forms of yarn are monofilament, multifilament,
and spun-staple. The basic weaves used for fabric filters are
plain, twill, and sateen. See Figures 13 and 14. The more com-
mon woven fabric bags are made from cotton, wool, Dacron, Nylon,
Orion, Nomex, polypropylene, Teflon, and fiberglass. Dimensional
stability is an important factor in filter fabric. Cotton and
wool must be preshrunk and synthetics are usually given a cor-
responding treatment called "heat-setting." Fiberglass-type
fabric bags are treated with silicone, mixtures of silicone and
colloidal graphite, or Teflon lubricants to provide protection
against flexing of the fabrics and abrasion. These lubricants
are effective up to about 290°C.
Felted fabrics serve as filter media and are used in reverse-
air and pulse-jet baghouses with air-to-cloth ratios of 6:1 to
16:1. Felted bags are more expensive than woven bags. Wool is
the only fiber which will produce a true felt. However, synthetic
materials can be needled to function as a felt fiber. Of the
materials mentioned above cotton and fiberglass do not yet make
felted fabrics.
Fabric filter bags have an average life of 18 to 36 months.
Several causes of bag failure include blinding, melting, caking,
tearing, burning, decomposition, abrasion, chemical attack, and
aging. The circumstances that can lead to bag failure vary widely.
Some occur during normal operation; others are due to misapplica-
tion or improper operation.
Table I presents a summary of the physical and chemical pro-
perties of current leading filter fibers.
20
-------
PLAIN
TWILL 2/1
TWILL 2/2
TTl
in
BUI
WARP
DIRECTION
TWILL 3/1 TWILL 3/2
SATEEN
Figure 13. Typical filter cloth weaves.
21
-------
PLAIN
TWILL
SATEEN
Figure 14. Basic fabric weaves used for woven filter bags.
(Courtesy of West Point Pepperell, Industrial Fabrics Division)
22
-------
TABLE I.
SUMMARY OF PHYSICAL AND CHEMICAL PROPERTIES OF INDUSTRIAL FILTER FIBERS
MAX RECOMMENDED
SPECIFIC OPERATING RESISTANCE TO RESISTANCE TO RESISTANCE TO RESISTANCE
Creslan
Orion
Nylon
.c 1 1 tron
Dynel
Nomex
Nylon
Dacron
Polypro-
pylene
Teflon
Fiber-
crlass
:otton
Wool
GRAVITY
1.18
1.14
1.38
1.38
.90
2.10
2.54
1.50
1.32
PEMPERATURE
135
135
93
132
71
232
148
107
260
288
107
93
C ABRASION
G
G
E
VG
F
E
E
E
F
P
G
G
MINERAL ACIDS
G
G
F
VG
VG
F
G
E
E
E
P
F
ORGANIC ACIDS
G
G
F
VG
VG
E
G
E
E
G
G
F
TO ALKALIES
F
F
E
G
VG
G
G
E
E
G
G
P
E - Excellent
^: - Very Good
G - Good
F - Fair
P - Poor
23
-------
1.1.4 Types of Fabric Filters
Baghouses are characterized according to the method used
to remove the participate matter from the bags. In shaker-type
units, the filter bags are hung from a structural framework.
The structure is supported so that it will oscillate freely when
driven by an electric motor. At set intervals, a damper is used
to isolate a baghouse compartment such that there is no gas flow
through it. The bags are then shaken for a preset period of time.
The collected dust is dislodged from the bags and falls into a
hopper from which it is subsequently removed. A shaker-type bag-
house is depicted in Figure 3. Figure 15 shows the amount of
dust remaining on a bag after several shakes during a cleaning
cycle.
Reverse-flow baghouses are equipped with a secondary fan
which forces air through the bags in an isolated compartment in
the direction opposite to filtraton. This action collapses the
bag and fractures the dust layer. When the filter bags are rein-
flated by being brought back on line, the broken dust layer is
dislodged into the hopper. If the main process fan is located
downstream of the baghouse, the reduced pressure in the structure
may eliminate the need for an auxiliary fan. Sometimes shaking
and reverse-flow cleaning mechanisms are combined in the same
baghouse unit. This reverse-flow cleaning action is illustrated
in Figure 16.
Reverse-pulse fabric filters use a short pulse of compressed
air directed from the top to the bottom of each bag. This burst,
usually less than 100 milliseconds, aspirates secondary air as
it passes through a nozzle or venturi. The resulting air mass
expands the bag and expells the collected dust. An example of
this cleaning mechanism is diagramed in Figure 17.
24
-------
100
1st THROUGH 7th SHAKE, EACH 5 SEC
8th THROUGH 11th SHAKE, EACH 2 MIN
12th SHAKE, 5 MIN
13th SHAKE, 10 MIN
1
TERMINAL AP: 3.0 IN. H,O
AVG FILTER VELOCITY: 0.5 FT/MIN
INLET DUST CONC.: 7.5 GRAINS/CU FT
I
100
150 200 250
DUST MASS, grains/sq ft
300
350
400
Figure 15. Residual dust left after several periods of shaking of a fabric filter bag.
The dust mass varies along the length of the bag in a manner charac-
teristic of the cleaning technique. After Stephan and Walsh(1960).
25
-------
V
V
,1,
s.
1
X
'EXHAUST
REPRESSURING
VALVE
SIDE VIEW
FILTERING
SIDE VIEW
COLLAPSING
V.;:;.y SIDE VIEW
CLEANING
Figure 16. Reverse air flexing to clean dust collector bags by repressuring.
26
-------
SOLENOID
VALVES
TO
EXHAUSTER
DUST
LADEN
AIR
COMPRESSED AIR
BLOW PIPE
INDUCED FLOW
FILTER CLOTH
BAG RETAINER
MATERIAL DISCHARGE
Figure 17. Typical reverse-pulse baghouse during cleaning cycle.
27
-------
Shaker and reverse-flow baghouses collect dust on the inside
of the bags, while reverse-pulse units collect dust on the outside
of the filters. Several hybrid combinations of these three clean-
ing methods have been developed. They include reverse-flow (shake
assist), reverse-flow (vibrator assist), pulse, and pulse (off-
line) . See Figures 18 and 19.
1.2 PARTICULATE SAMPLING FOR FABRIC FILTER EVALUATION
1.2.1 General Considerations
Measurements of particle size and concentration are usually
made at both the fabric filter inlet and outlet to obtain an ac-
curate characterization of the baghouse performance.
Gas velocities may vary from 15 m/sec to 30 m/sec, in ducts
leading toward or away from the baghouse, to as low as 1-5 m/sec
in transforms immediately upstream or downstream of the baghouse.
If samples are taken near the baghouse proper, then low gas
velocities pose problems because of the difficulty of sampling
isokinetically at such locations. Large diameter nozzles are
required to sample at low gas velocities when normal sampling
train flow rates are used. Conversely, high gas velocities can
create situations where nozzles of very small diameter (one or
two millimeters) are necessary.
Particulate concentrations at the sampling point may range
from 10 g/m3 to 0.001 g/m3 depending upon the type of source and
the collection device efficiency. Sampling times to collect ac-
curate data vary approximately inversely with the dust concentra-
tion. At baghouse outlets it is not uncommon to sample for twelve
hours or longer to collect a sufficient amount of particulate
matter. In fact, the particulate concentration can be too low
for some instruments to detect or measure. Sampling times can
28
-------
VIBRATOR
FLUSING
AIR IN
REVERSE
FLOW
CRUDE GAS
PULSED
FLUSING
AIR
u
Jil
.COMPRESSED—^ |
AIR
.SUPPORT.
CAGE
—
Figure 18. Cleaning methods for filter bags, (a to c: mechanical cleaning by
shaking [rapping mechanism or vibrators], d to h: pneumatic
cleaning) After Bate!(1973).
29
-------
PRESSURE
TYPE
SUCTION
TYPE
INTERMITTENT
CLEANING
-Q-X--X
X
INTERMITTENT
REVERSE FLOW
CLEANING (CON
TINOUS IF
COMPARTMENTED)
X-0
CONTINOUS COM-
PARTMENTED RE-
VERSE FLOW
CLEANING
CONTINUOUS PULSE
OR JET CLEANING
KEY:
FABRIC FILTER
PRIMARY FAN
AUXILIARY BLOWER OR COMPRESSOR
Figure 19. Types of fabric filter systems depending on cleaning method.
After Billings and Wilder (1970).
30
-------
also be discontinuous if studies are being conducted to isolate
the effects of baghouse compartment cleaning.
At some sources, condensation can occur within the pollution
control system or in the stack, and particles may grow larger
and change composition, or be created, by this mechanism. These
occurrences will cause changes in the observed particle concen-
tration and size distribution.
In many instances the duct dimensions at the sampling loca-
tion can pose problems, especially if sampling from the top of
the duct is required. The duct may be up to seven meters in depth.
A hoist must be constructed in some cases to handle probes re-
quired to obtain traverse samples near the bottom of such a duct.
1.2.2 Particulate Mass Measurements
Measurements of the particulate mass concentration are made
by pumping the dust laden gas through a system containing a filter
and a means of measuring the volume of the gas sampled. The total
material collected on all surfaces within the system is recovered
and weighed. This weight, normalized to a unit of gas volume,
is the suspended particulate concentration. The samples are col-
lected using a prescribed traversing procedure, which, in effect,
yields an approximate integration of the average mass emission
rate past a cross-section of the duct or stack. The velocity
distribution of the gas is also measured as part of the test.
The gas flow rate and velocity distribution and the particulate
mass concentration are used to calculate the mass flow rate or
emission rate at the point of interest. Measurements of the mass
flow rate are made at both the inlet and outlet to determine the
fabric filter's collection efficiency. Section 2.2 contains a
summary of the methods for conducting mass measurements. Appendix
B describes in detail the EPA Method 5 particulate mass measure-
ment procedures.
31
-------
1.2.3 Particle Sizing Techniques
Measurements of particle size distributions in industrial
flue gas streams are made for several reasons. The aerosol must
be characterized as completely as possible in order to assess
the potential for adverse health or environmental effects; emis-
sion measurements can be useful as a process monitor; and the
aerosol particle size distribution must be known in order to com-
pletely quantify the behavior of a control device. Also, particle
size measurements on uncontrolled sources are sometimes useful
in fabric filter design.
In recent years the emphasis in pollution control has been
placed on "fine" particles which are defined by the EPA as those
particles having aerodynamic diameters smaller than three microm-
eters. Fine particles are more difficult to control than large
particles, and because they are respirable, may constitute a
greater hazard to health.
Several techniques must be applied if information on par-
ticle size over a wide range of diameters is required, or if real
time data is desired. As a general rule, most particle sizing
techniques yield accurate information over approximately a factor
of ten range in particle diameter. The particle size range with
which this manual is concerned is 0.01 to 10 micrometers diameter.
Therefore, several techniques must be employed.
Section 2.3 contains a summary of the instruments that are
available for particle sizing, as well as a discussion of their
applicability to specific tests. Appendices C and D contain
detailed descriptions of the procedures for measurement of par-
ticle size distributions in flue gases.
32
-------
2. TECHNICAL DISCUSSION
2.1 MECHANICAL CHARACTERIZATION OF A FABRIC FILTER INSTALLATION
In order to perform a comprehensive analysis of the perform-
ance of a baghouse it is necessary to obtain information on the
design and operating parameters of the complete installation.
These data are generally used, with the test results, to assure
compliance with regulations, or to make comparisons with other
fabric filter evaluations or with performance predictions of theo-
retical or empirical models. Plant or process operating data
should also be obtained as part of each test program, and corre-
lated with the control device performance. Plant data are rou-
tinely recorded by the plant personnel and arrangements can usually
be made to obtain copies of these records or to have a member
of the test crew record the data during a test.
In the remainder of Section 2.1, the nomenclature that is
used to described fabric filters is defined and a number of param-
eters that should ideally be measured or noted during a baghouse
evaluation are listed. Some of the data listed are essential
to a meaningful evaluation, while some may be difficult or im-
practical to obtain and must be sacrificed. Also, some instal-
lations may have individual peculiarities requiring that addi-
tional data be taken. Although this manual is intended as a
comprehensive guide, there is no substitute for intelligent and
experienced judgement in the discrimination of essential from
nonessential data.
33
-------
2.1.1 Mechanical Design and Operating Data
The objective of a baghouse design is to combine the compo-
nent parts into an effective arrangement that results in an opti-
mum collection efficiency at the least operating expense. Ex-
perience, efficiency requirements, and economics generally dictate
the best arrangement. Fabric filter technology can be improved
through the study and comparison of many of the items listed below
and in the following sections.
For completeness, determinations of the following items per-
taining to the physical layout and mechanical operation of the
baghouse should be made during the test. This data may be ob-
tained by observation, from plant personnel, or from manufacturers'
literature.
Number of baghouses — A baghouse is a metal box-like struc-
ture with inlet and outlet gas manifolds and which contains the
filtering elements.
Dimensions of the baghouse — The physical height, length,
and width of each baghouse should be noted.
Inlet and outlet manifolds — The physical dimensions of
the gas inlet and outlet manifolds should be noted; whether they
are round or rectangular; whether they have any bends that might
cause poor particulate concentration distribution; whether there
are any flow distribution devices or baffles in the ducts leading
to the baghouse proper. Any or all of these items can affect
the observed inlet and outlet dust concentration measurements.
Sampling ports are usually located in these inlet and outlet
ducts. If this is the case, their size, shape, number, and loca-
tion should be noted.
34
-------
Number 'of baghouse compartments — Each large, nonreverse
pulse cleaned baghouse is divided into several independent com-
partments. During a complete cleaning cycle, each compartment is
taken off line in turn and the bags are cleaned. When one com-
partment is returned on line, another compartment is taken off
line for cleaning.
Total gas volume, temperature, and moisture content — The
average total gas volume through the baghouse, its temperature,
and normal moisture content can usually be obtained from plant
data. During a normal evaluation procedure, however, these data
are obtained as part of the mass concentration (EPA Method 5)
measurements.
System air movers — It is usually worth noting the amount
of fan capacity and also whether the system is induced draft or
forced draft. The amount of air flow through the baghouse will
be affected by the buildup of dust on the bags and their clean-
ing cycle. The fan capacity should be capable of handling the
maximum filter pressure drop anticipated during normal operation.
Compartment access — The accessibility to the baghouse com-
partments should be determined. This will indicate how readily
old bags can be replaced with a minimum of down time.
2.1.2 The Fabric Filter Bags
The entire baghouse is designed around the filtering element
or bag; thus, detailed information on their design, structure,
and application is required.
Total number of bags — The total number of bags in each
baghouse should be noted.
Number of bags per compartment — The number of bags in each
baghouse compartment should be noted.
35
-------
Bag shape and size — The bag shape as well as the length
and diameter should be determined.
Bag construction — Several items with respect to the con-
struction of the bags should be noted. These include seams,
closures, seals, clamps, cuffs, reinforcements (cage, rings, or
frame), attachments and thimbles. This information can usually
be obtained from plant personnel or manufacturers' literature.
•
Bag placement and spacing — The arrangement of the bags
in each compartment should be determined as well as their spacing,
bag to bag and bag to wall. These parameters affect bag wear,
air movement, and ease of bag replacement.
Type of bag cleaning mechanism — The type of mechanical
action by which the dust is removed from the bags should be noted.
Currently available types include (1) off-line, gravity; (2) off-
line, collapse; (3) off-line, collapse, reverse-air; (4) off-line,
shake or vibratory; (5) reverse-air; (6) pulse-jet; and (7) re-
verse-air, shake assist.
Cycle for cleaning bags — The sequence of compartment bag
cleaning should be noted, including how often each compartment
is taken off-line for cleaning, the duration of each cleaning
cycle, and the amount of each cycle used for shaking, reverse
air, pulse, etc.
Pressure drop to initiate cleaning — Generally bag cleaning
commences when the pressure drop across a compartment reaches
a predetermined value. Once this pressure drop is attained, the
compartment cleaning cycle commences automatically, and continues
until the pressure drop either has been lowered to a preset level
or a preset length of time has elapsed. Generally each manufac-
turer has a particular type of cleaning system developed that
is somewhat unique from any other.
36
-------
2.1.3 Filter Fabrics
The filtering element in the baghouse installation is the
fabric. A wide variety of commercial synthetic and natural fibers
have been developed to cope with many different types of indus-
trial situations.
Fiber material — The type of industrial process will gen-
erally determine the fiber material that is used in the fabric
filters. Many synthetic materials have recently been marketed
which will withstand temperatures above 100°C. A list of the more
commonly used fibers can be found in Section 1.1.2.
Fiber and fabric characteristics — There are many properties
of the filter fabric that can be specified; however, they are
not all essential to a complete evaluation. Some of the inform-
ation also may be very difficult to find, except through the
manufacturer's literature. For completeness, a list of these proper
ties is presented below:
Surface properties
Fiber size, length/thickness, texture
Specific gravity
Strain, recovery, creep
Flexibility, elasticity
Stiffness
Hardness, deformation
Modulus, compliance
Moisture regain
Friction, cohesiveness
Electrical properties, dielectric constant
Surface asperities
Physical resistance to the environment
Chemical resistance to the environment
Age effects, degradation
37
-------
Bacteriological effects
Polymer additives, treatments
Yarn twist, lay
Staple, filament configuration
Number of fibers/yarn
Thread used to sew
Weave style
Crimp, geometric properties
Porosity
Yarn count, warp, fill
Yarn tension
Interyarn spacing
Available free fiber
Fabric treatments — To aid in a longer bag life most commer-
cial filter bags are treated with special lubricants, such as
graphite and Teflon, to cut down on fiber to fiber and bag to
bag abrasion. Also most commercial fibers are treated or "heat-
set" to reduce stretching or shrinking at elevated temperatures.
Fabric thickness and weight — The weight of the fabric
should be noted. It is measured commonly in ounces per square
yard. This will give an indication of the durability of the product.
Air-to-cloth ratio or air flow permeability — The air-to-
cloth ratio gives the volume of air per minute passing through
one square foot of cloth at a predetermined pressure drop. This
is also the filtering velocity. A knowledge of this quantity
is important in determining the final system operating pressure
drop, as well as the pressure drop at which cleaning is initiated.
Generally the pressure drop across each baghouse compartment is
monitored. Figures 10 and 11 show experimental data indicating
the relationship between air-to-cloth ratio and fabric filtering
efficiency at a coal-fired boiler.
38
-------
2.1.4 Dust Removal Systems
It is important during any fabric filter evaluation to de-
termine that the dust removal system is working properly and to
specifications. If the system is not working properly, discus-
sions should take place with plant personnel to see whether these
problems can be fixed in a reasonable time span. If this is not
the case, then testing should be delayed.
Hoppers are used to collect and store dry particulate which
is removed from the bags. If the hoppers are allowed to over-
flow, the collected dust may reach the level of the bags and cause
blinding, chafing, or chemical attack.
Several types of systems exist for removal of dusts accumu-
lated in hoppers. These include container removal, dry vacuum,
wet vacuum, screw conveyors, and scrape bottom systems.
As much information about the dust removal system should
be reported as is practical.
2.1.5 Baghouse Operation-General Maintenance Considerations
Although noting general maintenance procedures is not neces-
sarily a part of an evaluation procedure, the status of baghouse
operation at the time of testing will depend on the previous
operation and maintenance schedule. Thus a brief description
of these procedures is in order.
One way to minimize operating and maintenance costs while pro-
longing the life of the fabric filters is to keep the air flow and
temperature into the filters as low as possible, limited only by the
danger of reaching the dewpoint. The day-to-day use of a filter sys-
tem requires frequent observation and occasional adjustments in
order to determine and adhere to the best overall compromise between
performance, bag life, power, costs, etc.
39
-------
Use of instruments — A single manometer across each bag-
house compartment generally provides all the data necessary on
the operation of that section of the unit. It indicates the per-
meability of the cloth, how heavy the dust deposit is before
cleaning- how complete the cleaning is, and whether the fabric
is starting to plug or blind. It indicates what surges in veloc-
ity the dust deposit is under going and whether there is any flow
through the cloth during the cleaning cycle. It is advisable
to have a list of normal differential pressures through one filtra-
tion cycle, as a means to quick detection of later trouble. A
high differential pressure may mean:
• an increase of air flow
• the beginning of blinding of the cloth
• hoppers so full as to block off the bags
• condensation in the cloth
• cleaning mechanism inoperative
while a low pressure differential may mean
• the fan is slowing down, perhaps due to belt slip-
page or fan motor problems
• broken or undamped bags
• plugged inlet ducting or valves closed
• leakage between sections of the filterhouse.
Other monitoring instruments can be nearly as valuable, since
transients frequently occur in pressure, flow rate, temperature,
humidity.
Flow variations — At processes requiring multiple or vari-
able dust pick-up points, there will be variations in flow rate
and filtering velocity. All the branches of the inlet ducting
may be open or some may be shut off, depending on plant activity.
Also, while one collector is down for its cleaning cycle, another
may be down for bag changes or still another for inspection.
These system changes affect the flow through each filter. Too
40
-------
much flow through too few bags amounts to an overload or too high
a filter velocity, leading to insufficient filtering, plugging
of the cloth, loosening or bursting of the bags. Too low a flow
is a frequent cause of condensation.
Cleaning cycle — As the cloth ages, adjustments in the
cleaning cycle either in the amount or intensity of cleaning or
in the length of the cleaning cycle are generally necessary. To
prolong the life of the fabric, as little cleaning as necessary
is advisable; but enough cleaning is required to keep the dif-
ferential pressure at an economic level. There is a point of
optimization, although this may be difficult to locate in prac-
tice. Operating at the point of minimum cleaning is indicated
by a gradual build-up of differential pressure, perhaps over a
period of a few days. Before the pressure gets too high, a slight
increase in cleaning action should reverse the trend in pressure.
After a few cycles or a few hours the pressure should reach a
sufficiently low level to reduce the cleaning action. If this
gradual fluctuation in differential pressure is not observed,
it may mean that the fabric is being overcleaned. Changes in
fabric condition or process will undoubtedly cause the cleaning
requirements to shift from week to week.
Types and frequency of fabric filter problems — A survey
in the late 1960's obtained information on specific problems en-
countered with fabric filter installations. Table II lists 23
types of operational problems encountered by the installations
surveyed (Billings and Wilder, 1970). They have been grouped
into five causality categories as follows:
41
-------
1. Fabric-dust interactions 9 types of problems 60 problems
reported
2. Filter element difficul- 2 " 6 "
ties
3. Filter element-hardware 5 " 12 "
interactions
4. Collector design prob- 5 ' 27 "
lems
5. System design problems 2 " 7 "
23 types of problems 112 problems
reported
42
-------
TABLE II
TYPE AND FREQUENCY OF FABRIC FILTER PROBLEMS REPORTED
1. Fabric - dust deposit interactions Frequency
a. interstitial deposit related 8
abrasion, wear
b. flexture wear failure 10
c. seeping 4
d. blinding 14
e. burning, heat 6
f. holes, pinholes, shot holes 6
g. hydroscopicity 4
h. condensation 5
i. deposited dust hardens, cake 3
tears, cracks bag
Subtotal 60
Fabrication failures not particularly related to dust inter-
action, mechanical
a. seams, sewing 2
b. tears at top 4
Subtotal 6
Design or maintenance failures related to tensioning, sup-
ports, rings, collars, or cleaning device interactions
a. chafe on housing or other bags 3
b. tensioning, bags too loose 1
c. cage, wire, ring abrasion, wear 5
(also dust related), support
mechanism interact
d. cleaning carriage bag wear 1
e. seals around cloth-metal collars 2
Subtotal 12
43
-------
TABLE II (Continued)
Collector design problems, internal, external mechanisms,
incl. auxiliaries, (ex. pipes, hoods)
a. unable to enter collector to service or 4
maintain during operation
b. hole detection problems or performance 3
effluent monitor
c. hopper dust sticking, holdup, screw 6
conveying plug
d. internal mechanism wear 4
e. external mechanism wear, timer, shaker, 10
fan, bearings, doors, seals, wall
failure
Subtotal 27
Dust collecting system design problems, external to col-
lector (incl. pipes, hoods)
a. piping, elbows, .abrasion 4
b. hood inlet control poor 3
Subtotal 7
Total: 112
44
-------
45
-------
2.2 MASS EMISSION MEASUREMENTS
2.2.1 General Discussion
As part of most fabric filter evaluation studies, the parti-
culate collection efficiency is determined experimentally by
making measurements of the mass concentration and gas flow rate
at the baghouse inlet and outlet.
A number of test procedures have been developed for perform-
ing mass emission measurements on process streams. All of the
"standard" sampling trains are similar and basically are composed
of a nozzle, a probe, a filter, one or more devices for monitor-
ing gas flow, and a pump. Generally pitot tubes and thermocouple
assemblies are also used to measure the gas velocity and tempera-
ture. The nozzle is streamlined to minimize flow disturbances
and the diameter is chosen for isokinetic sampling. The probe
must be rigid so that the nozzle can be positioned accurately
at the selected sampling points.
Filters of various compositions and geometries are used,
although glass fiber is most common. Any filter chosen must be
an efficient collector of submicron particles. A gas meter and
sometimes a calibrated orifice are used to measure the gas flow
rate and total volume sampled. Condensers are used to measure
the water vapor content, as well as for pump and meter protec-
tion. In some instances impinger bubblers and liquid traps are
used and the contents analyzed after sampling for various vola-
tile elements.
Many emissions contain substances which condense at tempera-
tures well above ambient to form solid or liquid particles. Care
must be taken that the temperature in all parts of the system
upstream from the filter be kept at temperatures high enough to
prevent condensation. Also, there is sometimes considerable depo-
46
-------
sition on the surfaces of the nozzles, probes, etc., and these
must be thoroughly cleaned as part of normal sampling and analysis
Three systems which are commonly used to measure mass emis-
sions are described in the following paragraphs.
2.2.2 EPA-Type Particulate Sampling Train (Method 5)
Official performance testing of control devices on stationary
sources in the United States must be conducted with the "EPA
Method 5 Sampling Train" (Federal Register, 1977) illustrated
in Figure 21. A heated sample probe is used to transport the
particulate sample to a glass fiber filter which is maintained
at 120 ± 14°C. A reverse type pitot tube is attached to the probe
to insure that isokinetic sampling conditions are maintained
during a traverse of the duct or stack. According to the EPA
method the glass fiber filter must have a penetration value equal
to or less than that of MSA 1106 BH (approximately 0.05% for a
standard 0.3 ym DOP aerosol penetration test at 5 cm/sec face
velocity). Gases, vapors, and any particles that penetrate the
heated filter enter a series of impingers or condensers that are
immersed in an ice bath. They trap the uncombined water that
is present in the gas stream so that the moisture content can
be determined; they also prevent the hot, humid gases from en-
tering the gas metering system and pump.
After the traverse, the filters are dried and the probe is
washed to remove and collect particulate matter from the probe
walls for subsequent analysis. The EPA Method 5 requires that
acetone be used for the probe wash. This creates a problem when
the long metal probes are hot from sampling, so distilled water,
with the Administrator's approval, is often used instead of ace-
tone. The probe wash liquid is collected and evaporated to dry-
ness so that the amount of particulate matter removed from the
probe can be weighed. This weight, and the weight of the parti-
47
-------
HEATED PROBE
IMPINGER TRAIN OPTIONAL:
MAY BE REPLACED BY AN
EQUIVALENT CONDENSER
*-L—rlXW-tX}
CHECK
VALVE
MANOMETER DRY TEST METER AIR TIGHT PUMP
0700-14.16
Figure 20. EPA Method 5 Paniculate Samp/ing Train.
48
-------
culate matter on the filter, and the measured gas flow are used
to determine the mass emission rate of the source. Approved sys-
tems of this type are currently available in low (230-472 cm3/sec)
(0.5-1 ft3/min) and high (up to 170 LPS) (6 ft3/min) flow rate
configurations. A modification to the EPA particulate sampling
procedure (Method 17) has been promulgated to allow the use of
in-stack filters when testing Kraft pulp mills (see Appendix E).
2.2.3 ASTM-Type Particulate Sampling Train
The American Society of Testing and Materials (ASTM, 1977)
has described a particulate sampling train which is illustrated
in Figure 22. The main difference between this method and the
EPA Method 5 is the use of an in-stack particulate filter. With
this arrangement, a thimble-shaped filter is used to sample high
mass concentrations, and a conventional, disk-shaped, filter is
used for low mass concentrations. It is important to heat the
filter holder to insure that the filter temperature is maintained
above the dew point temperature if condensible vapors are present
in the gas stream. The advantage of this system is that the
particles are trapped before they enter the probe and a probe
wash is not required. Also, external heating of the filter is
often unnecessary. A condenser and gas cooler are still required
between the probe and the gas metering system. The pitot tube,
pump, and other parts of the system are similar to the EPA Method
5 Sampling Train. The thimble-filter system is often used in
engineering tests to evaluate the performance of a control device.
2.2.4 ASME-Type Particulate Sampling Train
The American Society of Mechanical Engineers (ASME, 1957)
has described in its Power Test Codes the use of a sampling train
to measure particulate emissions from industrial sources. To
meet the ASME specifications the particulate sampling train must
have the following parts:
49
-------
SAMPLING
NOZZLE
GLASS FIBER THIMBLE FILTER
HOLDER AND PROBE(HEATED)
REVERSE-TYPE
PITOTTUBE
CHECK
VALVE
DRY TEST METER
AIR-TIGHT PUMP
0700-14.17
Figure 21. ASTM Type Paniculate Sampling Train.
50
-------
A tube or nozzle for insertion into the gas stream and
through which the sample is drawn.
A filter (thimble, flat disk, or bag type) for removing the
particulate. For the purpose of the Power Test Code, 99-0%
efficiency by weight is satisfactory.
A means of checking the equality of the velocity of the gas
entering the nozzle and the velocity of the gas in the flue
at the point of sampling.
A method by which the quantity of gas sampled is determined.
A pump for drawing the gas stream through the nozzle, filter.
and metering device along with the necessary tubing. It
is important that the temperature of the gas be above the
dew point until after it has passed the filter.
This sampling procedure is not very restrictive. Both the
EPA and ASTM particulate trains comply with the ASME requirements,
2.2.5 General Sampling Procedures
Because the EPA Method 5 is required for most particulate
compliance testing, it is the only one discussed in this section
and in Appendix B, where the details of performing a mass emis-
sions test are described.
The general sampling procedure outlined herein is presented
in the EPA Test Method 5 "Determination of Particulate Emissions
from Stationary Sources." (See Appendix E.) Before sampling,
however, it is necessary to determine the number of sampling
points appropriate for the particular duct or stack under con-
sideration. EPA Test Method 1 (Federal Register, 1977) "Sample
and Velocity Traverses for Stationary Sources" (see Appendix E)
51
-------
describes the computations to determine the number of sampling
points for both the velocity traverse and mass sampling traverse.
The number of points will depend on the size and shape of the
duct. If the velocity traverse indicates that the velocity pro-
file in the duct is very unstable, the number of sample points
should be increased to obtain a more accurate integrated mass
emission rate (see Figure 23).
The use of the S-Type pitot tube, and its calibration for
measuring the stack gas velocity and flow rate, is described in
the EPA Test Method 2 (Federal Register, 1977) "Determination
of Stack Gas Velocity and Volumetric Flow Rate (Type S Pitot
Tube)." (See Appendix E.) The S-Type pitot tube is used because
it is less susceptible to clogging in high dust loading environ-
ments. It is also advantageous to perform a temperature traverse
of the duct during the velocity traverse. This can be easily
accomplished by attaching a thermocouple temperature sensor near
the end of the pitot. During actual mass sampling both the tem-
perature and gas velocity are monitored to allow isokinetic sampl-
ing at each traverse point.
Mass sampling at fabric filter installations can pose prob-
lems because of the large difference in dust concentration be-
tween the inlet and outlet ports. Baghouse inlet mass loadings
are usually high, while outlet loadings can be up to a factor
of 103 smaller. Extended sampling times are required in many
situations at baghouse outlets. Problems in obtaining average
inlet and outlet mass loadings can also occur at sources which
have cyclical operations. However, the outlet mass concentra-
tions at several baghouses remain fairly stable even if the inlet
concentration is fluctuating. The person preparing for a sampling
procedure must take all of these variables into account when de-
signing a test plan.
52
-------
15
o
O
>
e/j
ol_L
I
I
I I I
8 10 12
TRAVERSE POINTS
14
16
18 20
0700-11.5
Figure 22. Typical gas velocity distribution at the inlet to a baghouse.
53
-------
2.3 PARTICLE SIZE MEASUREMENT TECHNIQUES
2.3.1 General Discussion
Any detailed experimental program designed to evaluate fab-
ric filters must include measurements of the particle size distri-
butions at the inlet and outlet. These size distributions can
then be used to calculate the baghouse collection efficiency
versus particle size, or "fractional efficiency curve".
Although most of the mass emitted from a particular pollu-
tion source may consist of large particles, in general, the largest
number of particles is in the fine particle range. Thus, high
mass collection efficiency does not always imply high number col-
lection efficiency nor does it insure that a particular opacity
standard will be met. Fine particles also contribute more to
visible light scattering and opacity and present a greater health
hazard than do the larger particles.
An ideal particle size measurement device would be located
in situ and give a real time readout of particle size distribu-
tions and particle number concentration over the size range from
0.01 ym to 10 pm diameter. At the present time, however, par-
ticle size distribution measurements are made using several in-
struments which operate over limited size ranges and do not yield
instantaneous data.
Particle sizing methods may involve instruments which are
operated in-stack, or out of stack where the samples are taken
using probes. For in-stack sampling, the sample aerosol flow
rate is usually adjusted to maintain near isokinetic sampling
conditions in order to avoid concentration errors which result
from under or oversampling large particles (dia. > 3 pm) which
have too high an inertia to follow the gas flow streams in the
54
-------
vicinity of the sampling nozzle. Since many particulate sizing
devices have size fractionation points that are flow rate depen-
dent, the necessity for isokinetic sampling in the case of large
particles can result in undesirable compromises in obtaining data -
either in the number of points sampled or in the validity or
precision of the data for large particles.
In general, particulate concentrations within a duct or flue
are stratified to some degree with strong gradients often found
for larger particles and in some cases for small particles. Such
concentration gradients, which can be due to inertial effects,
gravitational settling, leaking compartments, etc., imply that
multipoint (traverse) sampling must be used. Even the careful
use of multipoint traverse techniques will not guarantee that
representative data are obtained. The location of the sampling
points during process changes or variations in baghouse operation
can lead to significant scatter in the data.
Choices of particulate measurement devices or methods for
individual applications are dependent on the availability of suit-
able techniques which permit the required temporal and/or spatial
resolution or integration. In many instances the properties of
the particulate are subject to large changes in not only size
distribution and concentration, but also in chemical composition
(for example, emissions from the open hearth steel making pro-
cess) . Different methods or sampling devices are generally re-
quired to obtain data for long term process averages as opposed
to the isolation of certain portions of the process in order to
determine the cause of a particular type of emission.
Interferences exist which can affect most sampling methods.
Two commonly occurring problems are the condensation of vapor
phase components from the gas stream and reactions of gas, liquid,
or solid phase materials with various portions of the sampling
systems. An example of the latter is the formation of sulfates
55
-------
in appreciable (several milligram) quantities on several of the
commonly used glass fiber filter media by reactions involving
SO and trace constituents of the filter media. Sulfuric acid
A
condensation in cascade impactors and in the probes used for ex-
tractive sampling is an example of the former.
If extractive sampling is used and the sample is conveyed
through lengthy probes and transport lines, as is the case with
several particulate sizing methods, special attention must be
given toward recognition, minimization, and compensation for
losses by various mechanisms in the transport lines. The degree
of such losses can be quite large for certain particle sizes.
In this section, individual particulate sampling systems
and procedures are discussed. These are categorized according
to the physical mechanism that is used to obtain the data: in-
ertial, optical, electrical, or diffusional.
2.3.2 Inertial Particle Sizing Devices
Two devices which fall into the inertial sizing category
are impactors and cyclones. In both of these devices, the aerosol
stream is constrained to follow a path of such curvature that
the particles tend to move radially outward toward a collection
surface because of their inertia. Subsequent analysis of the
particle size distribution may be made by gravimetric means,
quantitative chemical analysis, or microscopic inspection.
Particle size distribution measurements related to fabric
filter evaluations have largely been made using cascade impactors,
which are effective in the size range from 0.3 to 20 urn diameter;
although, in some cases, hybrid cyclone-impactor units, or cy-
clones have also been used. The particle size distributions are
normally calculated from experimental data by relating the mass
collected on various stages to the theoretical or calibrated size
cutpoints associated with those stage geometries.
56
-------
Cascade impactors — Because of its compact arrangement and
mechanical simplicity, the cascade impactor has gained wide ac-
ceptance as a practical means of making particle size measurements
in flue gases. In most cases, the impactors can be inserted
directly into the duct or flue, eliminating many condensation
and sample loss problems which occur when probes are used for
extractive sampling.
Figure 24 is a schematic which illustrates the principle
of particle collection which is common to all cascade impactors.
The sample aerosol is constrained to pass through a slit or cir-
cular hole to form a jet which is directed toward an impaction
surface. Particles which have lower momentum will follow the
air stream to lower stages where the jet velocities are progres-
sively higher. For each stage there is a characteristic particle
size which theoretically has a 50% probability of striking the
collection surface. This particle size, or D50/ is called the
effective cut size for that stage. The number of holes or jets
on any one stage ranges from one to several hundred depending
on the desired jet velocity and total volumetric flow rate. The
number of jet stages in an impactor ranges from one to about
twenty for various impactor geometries reported in the literature.
Most commercially available impactors have between five and ten
stages.
Parameters which determine the collection efficiency for
a particular geometry are the gas viscosity, the particle density,
the jet diameter or width, the jet-to-plate spacing, and the
velocity of the air jet.
Most modern impactor designs are based on the semi-empirical
theory of Ranz and Wong (1952). More comprehensive theories have
been developed by Davies and Aylward (1951) and by Marple (1970).
In practice, deviations from ideal behavior in actual impactors
dictate that they be calibrated experimentally.
57
-------
JET
TRAJECTORY OF
LARGE PARTICLES
TRAJECTORY OF
SMALL PARTICLES
GAS STREAMLINES
IMPACTION SURFACE
o
o
U-
UJ
O
UJ
o
o
A. TYPICAL IMPACTOR JET AND COLLECTION PLATE
100
50
U50
PARTICLE DIAMETER
B. GENERALIZED STAGE COLLECTION EFFICIENCY CURVE
3630-059
Figure 23. Operation principle and typical performance for a
cascade impactor.
58
-------
A large number of experimental studies have been published
on cascade impactor design and performance in the laboratory en-
vironment. Most of these have been reviewed in the dissertations
of Marple (1970) and Rao (1975). Recently, Gushing et al (1976)
have presented calibration data on several commercially available
cascade impactors. Figures 25, 26, 27, 28, and 29 show schematics
of the commercial impactor designs which are commonly used in
source testing. Table III gives a listing of the manufacturers,
and some operational information for stack sampling. The details
of cascade impactor applications are discussed in Appendix C.
It is usually impractical to use the same impactor at the
inlet and outlet of a fabric filter when making fractional ef-
ficiency measurements because of the large difference in parti-
culate loading. For example, if a sampling time of thirty minutes
is adequate at the inlet, for the same impactor operating condi-
tions and the same amount of sample collected, approximately 3000
minutes sampling time would be required at the outlet (a collec-
tion efficiency of 99% is assumed). Although impactor flow rates
can be varied, they cannot be adjusted enough to compensate for
this difference in particulate loading without creating other
problems. Extremely high sampling rates result in particle bounce
and in scouring of impacted particles from the lower stages of
the impactor where the jet velocities become excessively high.
Short sampling times may result in atypical samples being obtained
as a result of momentary fluctuations in the particle concentra-
tion or size distribution within the duct. Normally, a low flow
rate impactor is used at the inlet and a high flow rate impactor
at the outlet. The impactors are then operated at their respec-
tive optimum flow rates, and the sampling times are dictated by
the time required to collect weighable samples on each stage
without overloading any single stage.
59
-------
JET STAGE (9 TOTAL)
SPACERS
GLASS FIBER
COLLECTION
SUBSTRATE
NOZZLE
INLET
BACKUP
FILTER -
PLATE
HOLDER
CORE
0700-14.3
Figure 24. Andersen Mark III Stack Sampler.
60
-------
NOZZLE
PRECQLLECTION
CYCLONE
JET STAGE
(7 TOTAL)
COLLECTION
PLATE
SPRING
0700 14.1
Figure 25. Modified Brink Model BMS-11 Cascade Impactor.
61
-------
NOZZLE
INLET JET STAGE NO
STAGE NO. 2
STAGE NO. 3
STAGE NO. 4
STAGE NO. 5
STAGE NO. 6
STAGE NO.7
FILTER
IMPACTOR BASE
363O-046
Figure 26. MRI Model 1502 Inertial Cascade Impactor.
62
-------
NOZZLE
INLET CONE
STAGE 0
STAGE 1
STAGE 2
STAGE 3
STAGE 4
STAGE 5
FILTER
SUPPORT
IMPACTOR
BASE
3630-053
Figure 27. Sierra Model 226 Cascade Impactor.
63
-------
JET STAGE O-RING
COLLECTION PLATE
INLET
/ \
FILTER HOLDER
COLLECTION
PLATE (7 TOTAL)
JET STAGE
(7 TOTAL)
O7OO-14.2
Figure 28. University of Washington Mark III Source Test
Cascade Impactor.
64
-------
TABLE III
COMMERCIAL CASCADE IMPACTOR SAMPLING SYSTEMS
01
Name
Andersen Stack Sampler
(Precollection Cyclone
Avail. )
Univ. of Washington
Mark III Source Test
Cascade Impactor
(Precollection Cyclone
Avail. )
Brink Cascade Impactor
(Precollection Cyclone
Avail. )
Nominal Flow rate
(cm3/sec)
236
236
14.2
Substrates
Glass Fiber (Available from
manufacturer)
Stainless Steel Inserts,
Glass Fiber, Grease
Glass Fiber, Aluminum,
Grease
Manufacturer
Andersen 2000, Inc.
P.O. Box 20769
Atlanta, GA 30320
Pollution Control
System Corp.
321 Evergreen Bldg.
Renton, WA 98055
Monsanto EnviroChem
Systems, Inc.
St. Louis, MO 63166
Sierra Source Cascade
Impactor - Model 226
(Precollection Cyclone
Avail.)
MRI Inertial Cascade
Impactor
118
236
Glass Fiber (Available
from manufacturer)
Stainless Steel, Alumi-
num, Mylar, Teflon.
Optional: Gold, Silver,
Nickel
Sierra Instruments, Inc.
P.O. Box 909
Village Square
Carmel Valley, CA 93924
Meteorology Research, Inc.
Box 637
Altadena, CA 91001
-------
Series cyclone particulate sampling techniques — Prototype
series cyclone sampling systems have been developed for industrial
source sampling (Chang, 1974) . In general, series cyclones are
easy to use, trouble-free, and efficient collectors of large
quantities of size segregated particulate. Their main drawbacks
atT- tnat their size limits the number of size segregated samples
duiing each test that can be obtained as compared to most com-
mercial impactors and an accurate theory of operation has not
been developed. However, the ability to collect large quantities
of sized material for analysis makes these devices irreplaceable
for some applications.
The Source Assessment Sampling System (SASS) incorporates
t'uee cyclones and a back up filter (Blake, 1978). A schematic
ct this system is illustrated in Figure 30. It is operated at
a t low rate of 3065 cm3/sec (6.5 ft3/mi.n) with approximate cyclone
cut points of 10, 3, and 1 micrometer aerodynamic diameter and
a gas temperature of 205°C. Besides obtaining information on
the particle size distribution, this system collects gram quanti-
ties of particles for later chemical or biological analyses.
The SASS train is large and requires extractive sampling through
a heated probe. The cyclones are mounted in an oven to keep the
air stream at stack temperature or above the dew point until the
particles are collected. This system is supplied with nozzles,
a probe-pi tot-tube-thermocouple assembly, cyclones, back up filter,
oven, a gas conditioning chamber, and a flow metering device and
pump adapted from the Aerotherm High Volume Sampling System.
Cyclone calibration details are furnished along with equations
for calculating approximate cyclone cutpoints for operating con-
ditions other than those measured during the calibration procedure.
It is mandatory that the gas velocity and temperature through
the cyclones be maintained at a constant setting while sampling,
because the cyclone cut points are dependent upon the gas flow
rate and temperature. This usually means that periods of non-
66
-------
HEATER
CONTROLLER
STACK T.C.
<=5
~H
^
X
i—f
S-T>
SS PROBE
— t-
rPE PITOT)
/^~/'
fi=£\
/0-4\
^^__ r*/^M\/cr*Tir»M nwcui rlLICn
• UUIMV tC 1 IUIM UVbIM /
1 71 GAS COOLER
1 f /
1 ' 1 in I 1 i 1 i ' ./
j P| ' "r- J [ ^M [ '/^ ^^ _ N^
^ l_ °*> C\i
y i i_j i_|_i ^. jl-^
XAD-2 I TEMPERATURE
CARTRIDGE \V7/ T-C. _<
, X
s*
iMP/pnr^i FB
CONDENSATE JRACE ELEMENT-
COLLECTOR COLLECTOR
DRY GAS METER
^S|_ ORIFICE METER
w*=-
CENTRALIZ
AND PRES
FO TFMPFRATIIRF ^S
SURE READOUT -T.
1MPINGER
T.C.
CONTROL MODULE t^T"
10 CFM VACUUM PUMP
3630-054
Figure 29. Schematic of the Source Assessment Sampling System.
-------
isokinetic sampling may occur. Depending on magnitude of fluctua-
tions in the velocity of the sampled stream, this may or may not
introduce significant errors in the sizing process.
Southern Research Institute, under EPA sponsorship, has de-
signed and built a prototype three-stage series cyclone for in-
stack use. A sketch of this system is shown in Figure 31. It
is designed to operate at 472 cm3/sec (1 ft3/min). The calibrated
cut points for these cyclones are 3.0, 1.6, and 0.6 micrometer
aerodynamic at 21°C. A 47 mm Gelman filter holder is used as
a back up filter after the last cyclone. This series cyclone
system was designed for in-stack use and requires a six inch sampl-
ing port. A sampling system similar to that for a high flow rate
impactor is usually adequate, although a more powerful pump may
be required under some sampling situations. As with the SASS
train, a constant flow rate through the cyclone system is required
to maintain stable cyclone cut points.
Figure 32 shows a second generation EPA/Southern Research
series cyclone system under development which contains five cy-
clones and a back up filter. It is a compact system and will
fit through 4 inch diameter ports. The initial prototype was
made of anodized aluminum with stainless steel connecting hard-
ware. A second prototype, for in-stack evaluation, is made of
titanium. The development of this cyclone system is summarized
in a report by Smith and Wilson (1978) .
Figure 33 shows laboratory calibrations of the five cyclones
in the prototype system. The cut points, at the test conditions
are 0.32, 0.6, 1.3, 2.6, and 7.5 ym. A continuing research pro-
gram includes studies to investigate the dependence of the cyclone
cut points upon the sample flow rate and temperature so that the
behavior of the cyclones at stack conditions can be predicted
more accurately.
68
-------
TO PUMP
BACK-UP FILTER.
-El
CYCLONE 2 •>
^
-CYCLONE 3
NOZZLE
CYCLONE 1
3630-055
Figure 30. Three Stage Series Cyclone System.
69
-------
CYCLONE 1
CYCLONE 4
CYCLONE 5
CYCLONE 2
OUTLET
Figure 31. Five Stage Series Cyclone System.
CYCLONE 3
INLET NOZZLE
3630-056
70
-------
100
90
80
89
o 70
1 «>
LL
t 50
| t
o
LU
:j so
o
° 20
10
0.2 0.3 0.40.50.60.8 1.0 2 3 4 5 6 8 10 20
PARTICLE DIAMETER, micrometers
• FIRST STAGE CYCLONE
• SECOND STAGE CYCLONE
£ THIRD STAGE CYCLONE
V FOURTH STAGE CYCLONE
O FIFTH STAGE CYCLONE
3630-057
Figure 32. Laboratory Calibration for the Five Stage Series
Cyclone System. (472 cm^/sec, particle
density— 1.0 gm/cm3)
71
-------
Small cyclone systems appear to be practical alternatives
to cascade impactors as instruments for measuring particle size
distributions in process streams. Cyclones offer several advan-
tages:
Large, size segregated samples are obtained.
There are no substrates to interfere with analyses.
They are convenient and reliable to operate.
They allow long sampling times under high mass loading con-
ditions for a better process emission average.
They may be operated at a wide range of flow rates without
particle bounce or reentrainment.
On the other hand, there are some negative aspects of cyclone
systems which require further investigation:
Unduly long sampling times may be required to obtain large
samples at relatively clean sources.
The existing theories do not accurately predict cyclone per-
formance.
Cyclone systems are bulkier than impactors and may require
larger ports for in-stack use.
As discussed above, cyclones are now used on an experimental
basis by the EPA and EPA contractors. If current research pro-
grams are successful in developing a better understanding of
cyclone behavior, they may play a very important role in control
device evaluation.
72
-------
2.3.3. Optical Measurement Techniques
The basic operating principle for one type of optical parti-
cle counter is illustrated in Figure 34. Light is scattered by
individual particles as they pass through a small viewing volume,
the intensity of the scattered light being measured by a photo-
detector. The sizes of the particles determine the amplitude
of the scattered light pulses, and the rate at which the pulses
occur is related to the particle concentration. Thus, a counter
of this type gives both size and concentration information. The
simultaneous presence of more than one particle in the viewing
volume is interpreted by the counter as a larger single particle.
To avoid errors arising from this effect, dilution to about 300
particles/cm3 is generally necessary. Errors in counting rate
also occur as a result of electronics deadtime and from statis-
tical effects resulting from the presence of high concentrations
of sub-countable (D < 0.3 ym) particles in the sample gas stream
(Whitby and Liu, 1967) . The intensity of the scattered light
depends upon the viewing angle, particle index of refraction,
particle optical absorptivity, and shape, in addition to the
particle size. The schematic in Figure 34 shows a system which
utilizes "integrated near forward" scattering. Different viewing
angles might be chosen to optimize some aspect of the counter
performance. For example, near forward scattering minimizes the
affect of variations in the indicated particle size with index
of refraction, but for this geometry, there is a severe loss of
resolution for particle diameters near 1 ym. Right angle, or
90° scattering smooths out the response curve, but the intensity
is more dependent on the particle index of refraction.
73
-------
SENSOR
CHAMBER
PHOTOMULTIPLIER
INLET
CONE ,?
S
VIEW VOLUME
MIRROR
LAMP
CALIBRATOR
3630-060
Figure 33. Operating principle for an optical particle counter. Courtesy
of Climet Instruments Company.
74
-------
Available geometries are:
Bausch & Lomb 40-1 Near Forward Scattering
Royco 220 Right Angle
Royco 245, 225 Near Forward
Climet CI-201, 208 Integrated Near Forward
Optical particle counters have not been used extensively
in stack sampling because they cannot be applied directly to the
effluent gas stream. The sample must be extracted, cooled, and
diluted; a procedure which requires great care to avoid intro-
ducing serious errors into calculations of the particle size
distribution. The main advantage of optical counters is the capa-
bility of observing emission fluctuations in real time. After
extraction, the useful particle size range is approximately 0.3
to 1.5 vim.
2.3.4 Ultrafine Particle Sizing Techniques
There are two physical properties of ultrafine particles
(diameter < 0.5 ym) which are size dependent and which can be
predicted with sufficient accuracy under controlled conditions
to be used to measure particle size. These are the particle dif-
fusivity and electrical mobility. Although ultrafine particle
size distribution measurements are still in a developmental stage,
instruments are available which can be used for this purpose,
and some field measurements have been made. A practical limita-
tion on the lower size limit for this type of measurement is the
loss of particles by diffusion in the sampling lines and instru-
mentation. These losses are excessive for particle sizes below
about 0.01 ym where the samples are extracted from a duct and
diluted to concentrations within the capability of the sensing
devices.
75
-------
Diffusional sizing — Diffusion batteries may consist of
a number of long, narrow, parallel channels, a cluster of small
bore tubes, or a series of screens. Figure 35a shows a typical
parallel channel diffusion battery, and Figure 35b shows the
aerosol penetration characteristics of this geometry at two flow
rates. The parallel plate geometry is convenient because of ease
of fabrication and the availability of suitable materials, and
also because sedimentation can be ignored if the slots are ver-
tical, while additional information can be gained through settl-
ing, if the slots are horizontal.
Breslin et al (1971) and Sinclair (1972) report success with
more compact, tube-type and screen-type arrangements in laboratory
studies and a commercial version of Sinclair's geometry is avail-
able.* Although the screen-type diffusion battery must be cali-
brated empirically, it offers convenience in cleaning and opera-
tion, and compact size. Figure 36 shows Sinclair's geometry.
This battery is 21 cm long, approximately 4 cm in diameter, and
weighs 0.9 kg.
Variations in the length and number of channels (tubes, or
screens) and in the aerosol flow rate are used as means of mea-
suring the number of particles in a selected size range. As the
aerosol moves in streamline flow through the channels, the par-
ticles diffuse to the walls at a predictable rate, depending on
the particle size and the diffusion battery geometry. It is
assumed that every particle which reaches the battery wall will
adhere, therefore, only a fraction of the influent particles will
appear at the effluent of a battery. It is only necessary to
measure the total number concentration of particles with a con-
densation nuclei counter at the inlet and outlet to the diffusion
battery under a number of conditions in order to calculate the
particle size distribution.
*TSI Incorporated, 500 Cardigan Road, St. Paul, MN 55165.
76
-------
0700-14.11
Figure 34a. Parallel plate diffusion battery.
0.01
PARTICLE DIAMETER,
0700-14.12
Figure 34b. Parallel plate diffusion battery penetration curves for
monodisperse aerosols (12 channels, 0.1 x 10 x 48 cm).
77
-------
SAMPLING
PORT (TYP)
SECTION CONTAINING
SCREENS (TYP)
3630-045
Figure 35. Screen type diffusion battery. The battery is 21 cm
long, 4 cm in diameter, and contains 55, 635 mesh
stainless steel screens.
78
-------
Diffusional measurements are less dependent upon the aerosol
parameters than the other techniques discussed and perhaps are
on a more firm basis from a theoretical standpoint.
Disadvantages of the diffusional technique are the bulk of
the parallel plate diffusional batteries, although advanced tech-
nology may alleviate this problem; the long time required to mea-
sure a size distribution; and problems with sample conditioning when
condensible vapors are present.
Electrical particle counters — Several aerosol spectrometers,
or mobility analyzers, have been demonstrated that employ the dia-
meter-mobility relationship and to classify particles according
to their size, and Figure 36 illustrates the principle on which
these devices operate. Particles are charged under conditions
of homogeneous electric field and ion concentration, and then
passed into the spectrometer. Clean air flows down the length of
the device and a transverse electric field is applied. From a
knowledge of the system geometry and operating conditions, the
mobility is derived for any position of deposition on the grounded
electrode. The particle diameter is then readily calculated from
a knowledge of the electric charge and mobility.
Difficulties with mobility analyzers are associated primarily
with charging the particles to a known value with a minimum of
loss by precipitation and obtaining accurate analyses of the quan-
tity of particles in each size range. The latter may be done gra-
vimetrically, optically, or electrically.
The concept described above has been used by Liu, Whitby, and
Pui (1974) at the University of Minnesota, to develop a series of
Electrical Aerosol Analyzers (EAA). A commercial version of the
U. of Minnesota devices is now marketed by Thermosysterns, Inc. as
the Model 3030 (Figure 37).* The EAA is designed to measure the
*TSI Incorporated, 500 Cardigan Road, St. Paul, MN 55165.
79
-------
CHARGED PARTICLES
HV
CLEAN AIR
LAMINAR FLOW
te
*
A
k
\
SMALLER PARTICLES OF
HIGH ELECTRICAL MOBILITY
LARGER PARTICLES OF
LOW ELECTRICAL MOBILITY
3630-252
Figure 36. The electric mobility analyzer principle.
80
-------
CONTROL MODULE
ANALYZER OUTPUT SIGNAL
DATA READ COMMAND
CYCLE START COMMAND • -
CYCLE RESET COMMAND
AEROSOL FLOWMETER READOUT
CHARJER CURRENT READOUT
- --CHARGER VOLTAGE READOUT
AUTOMATIC HIGH VOLTAGE CONTROL AND READOUT
ELECTROMETER (ANALYZER CURRENT) READOUT
TOTAL FLOWMETER READOUT
00
-» EXTERNAL
-ft DATA
—'ACQUISITION'
—' SYSTEM
TO VACUUM PUMP
3630-043
Figure 37. Flow schematic and electronic block diagram of the Electrical
Aerosol Analyser. After Liu, Whitby, and Pui(1973).
-------
size distribution of particles in the range from 0.0032 to 1.0
diameter. The concentration range for best operation is 1 to
1000 ug/m3, thus dilution is required for most industrial gas
aerosols.
The EAA has the distinct advantage of very rapid data acqui-
sition compared to diffusion batteries and condensation nuclei
counters (two minutes as opposed to two hours for a single size
distribution analysis.)
Disadvantages of this type of measurement system are dif-
ficulties in predicting the particle charge, and the fraction
of the particles bearing a charge, with sufficient accuracy, and
the requirement for sample dilution when making particle size
distribution measurements in flue gases.
The details of sampling with optical particle counters, dif-
fusion batteries with condensation nuclei counters, and EAA using
extractive sampling and dilution techniques are presented in Ap-
pendix D.
2.4 PROCESS EFFLUENT GAS ANALYSIS
2.4.1 General Discussion
In evaluating the performance of a fabric filter it is ad-
vantageous to know something of the constituents of the effluent
gas stream, excluding the particulate matter. Information on
gas composition can indicate the potential for future problems
with the life of the fabric filter, the integrity of the baghouse
structure, or changes in the industrial process. Existence of
humidity and temperature near dew points can lead to blinding,
caking, and cracking of filter bags, if temperature excusions
are likely for the particular process. High SO content can cause
A
potential formation of HzSO^ mists which can ruin bags and bag-
houses, as well as mask ultrafine particle measurements data.
S02 can also react with certain glass filter substrate media in
sampling trains which cause weight changes that mask actual mass
collection gains. Also, if chemical agents are being injected
82
-------
to modify the industrial process, or if removal of polluting gases
(such as S02 by Nahcolite injection) is occur ing, then measure-
ments of gas composition are necessary.
A gas analysis generally concerns the amount of N2, 02, CO,
C02, and H2O in the effluent. These can usually be determined
using an ORSAT apparatus or one of the EPA Stationary Source Test
Methods developed for gas analysis. Studying effluent modifica-
tion by injection of chemical agents has also resulted in the
need for measuring the amounts of other gases such as NH3, S02,
SO,, NO , etc. These measurement methods are briefly discussed
J\
in the following sections.
2.4.2 Qualitative Gas Analysis
Interviewing key plant personnel or obtaining the proper
records should be sufficient to determine the qualitative nature
of the process gas stream. This data should include the average
fractional amounts of C02, CO, 02, N2, and H20 in the process
effluent gas stream during normal operation. Depending on the
industrial process under consideration, there may also be measur-
able amounts of NH3, S02, S03, NO , HF, sulfuric acid mist, or
other volatile substances.
Other qualitative information which should be gathered in-
cludes the average gas temperature at the baghouse inlet and out-
let, and the average actual and standard volumetric flow rates
through the structure.
2.4.3 Quantitative Gas Analysis
Flue gas constituents normally specified for analysis are
N2, 02, CO, and C02, and H20. In addition to these analyses,
S02 and SO3 concentrations are usually measured and sometimes
NH3, NOX, HF, or other vapor concentrations are determined.
83
-------
Oxygen, CO, and C02 concentrations are measured with a com-
mercial Orsat-type apparatus. Two Orsat-type analyzers are used
to determine the oxygen content of the gas entering and leaving
the baghouse simultaneously. Comparisons of the inlet and outlet
oxygen concentration provides a check for leakage of gas into
or out of the baghouse. Although in principle leakage can be
determined from an examination of inlet and outlet gas velocity
profiles, in practice, flow disturbances at available sampling
locations often severely limit the accuracy with which flow de-
termination can be made. Therefore, the simultaneous inlet and
outlet oxygen determinations may be a more sensitive indicator
of the physical integrity of the baghouse casing.
The Environmental Protection Agency has developed several
Stationary Source Test Methods for the determination of various
flue gas components. These methods, of course, are not the only
way by which the quantity of the gases can be measured. There
are many other acceptable analytical methods developed by dif-
ferent testing societies such as the ASME and ASTM. EPA Test
Method Number 4 (Federal Register, 1977) describes a procedure
to determine the H2O content of the flue gas. EPA Test Method
Number 6 (Federal Register, 1977) can be used to determine the
S02 content of the flue gas. EPA Test Method Number 7 (Federal
Register, 1977) explains a method for measuring the nitrogen oxide
in the flue gas. The amount of sulfuric acid mist and S02 con-
tent can be determined using the EPA Test Method Number 8 (Federal
Register, 1977) (see Appendix E).
Figure 39 illustrates a system not described in the Federal
Register, which has been found to be accurate and convenient to
use for measurements of S02 and SO3 concentrations. This is the
Controlled Condensation System developed for the EPA by TRW Systems
Group (Maddalone and Garner, 1977) to accurately determine SO2
and SO3 concentrations in process gas streams. This procedure
84
-------
FLUE-GAS
SAMPLE
1
7
WALL OF FLUE
HEATED
'SAMPLING PROBE
I
SO3 CONDENSER
SO2 ABSORBER
(PEROXIDE-WATER
SOLUTION)
VENT
DRIERITE OR
COLD TRAP
3630-092
Figure 38. Schematic diagram of apparatus for the collection of SOg
by the condensation method.
85
-------
is applicable in high or low mass loading environments with tem-
peratures up to 300°C and S02 concentrations up to 600 ppm.
The Controlled Condensation System is based on the separa-
tion of SO 3 as E2SO^ from S02 by cooling the gas stream below
the dewpoint of H2SO,,, but above the H2O dewpoint. Cooling
is accomplished by a water-jacketed coil where the H-jSO^ is col-
lected. Particulate matter is collected by a quartz filter mat
inserted in the line prior to the condensation coil. The parti-
culate filter system is maintained at a temperature of 288°C to
insure that none of the H-jSO^ will condense on the filter mat
or filter holder.
The Controlled Condensation Coil (CCC) is a modified Graham
Condenser. The water jacket is maintained at 60°C. This is ade-
quate to reduce the flue gas below the dewpoint of the HjSO^.
Following the CCC are two impingers for removing the SO2 and H20.
The SO2 scrubber is a bubbler filled with a 3% solution of H2O2
in water. The water vapor is removed by a silica gel filled
impinger. A vacuum pump with a capacity of 472 cm3/sec is recom-
mended. The total volume of gas sampled is measured with a dry
gas meter.
The sampling rate/time is normally 135 cm3/sec for one hour.
An indication of the proper amount of sample comes from watching
the condenser coil. When the H2S0lf fog has crept about one-half
to two-thirds of the way along the coil, the sampling can cease.
After sampling has been completed, the probe and coil are rinsed
with deionized H20 and the recovered solution is analysed in the
lab. The amount of I^SO^ in the condensation coil and probe can
be determined by a sulfate or H+ titration. Because of its sim-
plicity and sensitivity, the H+ titration is preferred. This
recommended acid/base titration uses Brom phenol Blue as the in-
dicator, since the endpoint of the NaOH and HgSO^ titration falls
near the pH range (3-4.6) of the Brom phenol Blue color change.
Complete details can be found in the reference above.
86
-------
3. DEVELOPMENT OF TEST PLANS FOR FABRIC FILTER EVALUATIONS
3.1 OBJECTIVES OF CONTROL DEVICE TESTS
Several reasons exist for performing fabric filter control
device evaluations. These reasons may range from a verification
of compliance with emissions requirements, to programs related
strictly to research.
The majority of stationary air pollution sources need some
type of control device to satisfy the national, state, or local
air pollution regulations that limit the allowable emissions.
In order to determine whether the plant is in compliance with
these regulations, tests are performed to measure the amount of
air pollutant emissions from the control device in question.
This is one type of control device evaluation and it is usually
the simplest and least expensive.
Another reason for performing tests on a fabric filter con-
trol device is to optimize the performance of the installation.
These tests might be requested by the owners of the plant where
the control device is installed, or by the control device manu-
facturer. Usually tests of both the inlet and outlet particulate
mass concentration are made resulting in a measure of the parti-
culate collection efficiency. In some instances the fractional
efficiency (efficiency as a function of particle size) is desired
and measurements of the particle size distributions of the inlet
and outlet dusts are necessary-
If a particular fabric filter installation is performing
poorly due to poor maintenance, or poor design, etc., then tests
might be required in order to obtain data to be used in designing
additional or replacement fabric filter units.
87
-------
To obtain data for purely research purposes is a fourth
reason for performing a fabric filter control device evaluation.
In each test the data may be used to confirm existing theories
of fabric filter operation or to develop new theories for model-
ling and predicting baghouse performance. Research tests may
involve total systems studies on the source/control device com-
bination. These tests are usually the most complicated and ex-
pensive Because of the amount of data that is desired.
3.2 TYPE AND NUMBER OF TESTS REQUIRED
As mentioned in the previous section, the type and number
of tests that are performed during a fabric filter evaluation
depend on the reason for the tests and the amount of funding avail-
able.
In most all cases the standard compliance test involves a
determination of the particulate mass concentration at the control
device outlet. Depending on the type of source, some measurements
of the gaseous emissions may also be required. The minimum number
of tests to be performed during a compliance test is usually set
by Federal or State regulations.
In order to study the performance of a fabric filter, mea-
surements of both the inlet and outlet mass concentration are
performed. These data are required for calculations of the parti-
culate collection efficiency. If the collection efficiency is
to be related to particle size, particle size measurements must
be performed at the inlet and outlet. If the source is stable,
fewer tests will be required than if the plant process is cyclic
or variable over an indeterminant time period. If a fabric filter
appears to be performing poorly, then other tests might be neces-
sary depending on the type of problem encountered. At a fabric
filter installation, the problem might be torn bags in one bag-
house compartment. This might require a special test strategy
to isolate this compartment.
88
-------
Data that are required for fabric filter design are the parti-
culate mass concentration; the particle size distribution; the
physical and chemical proeprties of the dust to be collected;
and the effluent gas temperature, pressure, and composition.
A fairly extensive testing program is necessary in order to obtain
these data. Tests should be performed during all normal process
cycles and with all types of expected feedstock to insure that
the fabric filter will not be designed undersize.
If testing is to be performed on a fabric filter for research
purposes only, then the tests that are made are dependent on the
information which is desired as well as the amount of funding.
As is true of all experimental type programs, the more data that
are obtained, the more reliable will be the conclusions based
on those data. Usually control device research programs are de-
signed to gather as much information as practical for the money
available. Generally, research studies concern the particulate
mass concentrations at the inlet and outlet, the inlet and outlet
particle size distributions, gas analysis, the dust properties,
the control device operation parameters, plant process data, pre-
vious control device maintenance data, and the economics of the
particular control device. Of course, the type of fabric filter
and plant will determine the specific tests which are conducted.
In some instances the type of tests which are conducted de-
pend on cooperation from the plant personnel. They may or may
not be willing to alter the feedstock or change the settings on
the particle collector controls, for example. Usually these prob-
lems are worked out as the test plan is developed.
In summary, fabric filter testing is not a routine operation
that has had all the problems worked out or specific procedures
developed. Each plant-control device combination is unique and
should be treated as such. Certain problems exist at one instal-
lation which might not be encountered at any other control device
89
-------
installation. The number of specific tests which should be per-
formed will depend on the type of control device, the stability
of the source, the length of time allowed for testing, and the
available funding. It is usually advisable to perform as many
tests as practical to insure that adequate data exist, because
later it may be found that some tests must be disqualified.
Table IV indicates some of the considerations and problems
that must be dealt with in developing a test plan for general
control device evaluations. Although this table is designed to
serve as a planning outline, the relative importance of the facets
of the plan, or considerations that are not listed, can only be
established from a good understanding of the plant-control de-
vice system and the objectives of the test.
For the purposes of developing specific baghouse test plans,
three basic levels of source-fabric filter data can be identified.
Table V depicts these categories in a flow chart where the pro-
gression is from minimum data to maximum data, or from relatively
inexpensive and qualitative to very expensive and quantitative
programs. In general, a basic test plan can be developed around
this flow chart. Depending on the interest in different types
of information, as well as the amount of funding available, the
actual test plan would be modelled after a Level A, B, or C fabric
filter evaluation.
In the following sections each of the components of these
test plans are briefly discussed. Also, a list of important items
to be noted under each category is presented.
90
-------
TABLE IV
PARTICULATE CONTROL DEVICE TASKS
Assure Compliance
with EPA
Objective of Tests Regulation
Tests Required
Gas Temperature
Control Device Data
Plant Process Data
Technical Considerations
(Decisions/Problems)
Adequate Space,
Condensible Vapors/
Volatile Particles
Mass Concentration/
Process/Emission
Variations
Select Particle Sizing
Methods
Select Gas Analysis Methods
Filter Mass Stability
Sample Preservation
n
n
o
x
3
X
p
o
p
c
n
Optimize Performance
of Control Device
0.
n
y
n
X
r
r
p
p
D
p
Obtain Design
Data for
Control Device
i
i
|
I
I
I
I
x
Y
p
p
p
c
1 0
i n
>(U
rj
o
p
Obtain Data
for Modeling
Studies
i n
i n
1 0
i n
1 0
y
i n
ccp nn|,.
X
X
rj
i n
x
P
n
r
r
i n
i n
n
0
P
Systems Studies
Process and
Control Device
i n
i n
\ 0
i n
1 0
i n
i n*
Y
Y
Y
V
o
x
i n
x
Y
P
p
n
p
r
i n
i n
n
o
p
n
Key: 0 Outlet
I Inlet
X Required
D Decision based on specific site or test objectives
C Must be considered
* vs. Particle Diameter
91
-------
TABLE V
THREE LEVELS OF EFFORT FOR FABRIC FILTER EVALUATION
LEVEL
: B i
ENGINEERING AND OPERATING DATA
Baghouse Design, Operating and Maintenance Data
Qualitative Gas Composition
Compartment Pressure Drop Data
Opacity Monitor Check For Leaks
Operating Air-to-Cloth Ratio
Two men on site for two days. Two man weeks for
analysis and reporting.
PARTICULATE MASS CONCENTRATION AND FABRIC FILTER
EFFICIENCY
Pretest Site Survey, Port Installation
Inlet and Outlet Mass Sampling
Quantitative Gas Analysis
Five men on site for five days.
for analysis and reporting.
One man month
PARTICLE SIZING AND FABRIC FILTER FRACTIONAL
EFFICIENCY
Inlet and Outlet Particle Size Distribution
Ten to fifteen men on site for one or two
weeks
Six to twenty man months for analysis, modelling,
and reporting.
92
-------
3.2.1 Fabric Filter Level A Evaluation
A Level A Fabric Filter Evaluation includes baghouse design,
operating, and maintenance data, baghouse compartment pressure
drop data, qualitative flue gas analysis, operating air-to-cloth
ratios, and opacity monitor information for each baghouse com-
partment.
3.2.1.1 Plant Operating Data
Generally the main inputs in plant operating data are the
raw input materials, the output materials, and the waste efflu-
ent, as well as the conditions of plant operation. Sometimes
the material collected by the fabric filter is the prime plant
product or is a by-product which is recycled or sold. An impor-
tant aspect of plant operation data is the variability of the
source. The plant may have periods of high or low output, or
it may be cyclic over a predetermined period. Also the raw ma-
terials may vary in composition over this cycle. The following
list encompasses most areas which should be noted. This data
is normally available from plant personnel.
a. Raw input materials
b. Output products
c. Waste effluent materials
d. Qualitative waste gas composition
e. Qualitative waste particulate effluent
f. Normal plant operating cycles
g. Normal plant outage periods
h. Plant economics: unit input energy/unit output product
i. Plant operating peculiarities
93
-------
3.2.1.2 Baghouse-Fabric Filter Design Data
All aspects of baghouse-fabric filter design should be noted,
if available. This includes structural design and filter design
data. This information will usually be available from plant
personnel and manufacturers literature.
a. Type of baghouse - number of baghouses, manufacturer
b. Dimensions of baghouse
c. Inlet and outlet manifolds
d. Number of baghouse compartments
e. Fans and capacity
f. Compartment accessibility
g. Number of bags
h. Number of bags per compartment
i. Bag shape and size
j. Bag construction
k. Bag placement and spacing
1. Bag cleaning mechanism
m. Pressure drop to initiate cleaning
n. Bag cleaning cycles
o. Type of fabric filter material
p. Fiber and fabric characteristics
q. Fabric treatments
r. Fabric thickness and weight
s. Air-to-cloth ratio
t. Dust removal systems
Refer to Sections 1.1, 2.1, and 2.2.
3.2.1.3 Flue Gas Characteristics. Baghouse AP, Maintenance Data
Under this section actual quantitative data concerning bag-
house operation is noted. This includes gas flow rates, tempera-
ture, pressure drops during filtering and cleaning cycles across
94
-------
each baghouse compartment, and any maintenance data that would
be useful in evaluating the performance of the baghouse. Main-
tenance data might include bag failure data, bag replacement
scheduling, start-up difficulties, etc. A sample check list for
noting potential problems is presented in Table VI.
a. Effluent gas volumetric flow rates
b. Qualitative gas composition (H20, 02, N2, C02, CO)
c. Qualitative evidence of S02, S03, H2S04 and other gas
forms depending on plant process
d. Gas temperatures
e. Pressure drop cycles across each compartment
f. Bag failures, opacity monitor data for each compartment
g. Bag replacement schedule
h. Any start-up difficulties
i. Current bag conditions
j. General maintenance status
Refer to Sections 2.1.2 and 2.1.5.
95
-------
TABLE VI
MAINTENANCE REPORT - TYPE AND FREQUENCY OF PROBLEMS
Problem Type Frequency
1. Fabric - dust deposit interactions
a. intestitial deposit related, abrasion, wear
b. flexture wear failures
c. seeping
d. blinding
e. burning, heat
f. holes, pinholes, shot holes
g. hygroscopicity
h. condensation
i. hardened cake tears, cracked bags
j. others
2. Fabrication failures, mechanical bag failures
a. bag seams, sewing failures
b. bag top tears
c. others
3. Design or maintenance failures related to
tensioning, supports, rings, collars, or
cleaning device interactions.
a. chafing on housing or other bags
b. tensioning, bags too loose
c. cage, wire, ring abrasion wear, support
mechanism interaction
d. cleaning carriage bag wear
e. seals around cloth-metal collars
f. others
96
-------
TABLE VI (Continued)
Problem Type Frequency
4. Collector design problems; internal, external
mechanisms
a. unable to enter collector to service or
maintain during operation
b. hole detection problems
c. hopper dust sticking, holdup, screw
conveyor plugging.
d. internal mechanism wear
e. external mechanism wear, timer, shaker,
fan, bearings, doors, seals, wall
failures, etc.
f. others
5. Dust collecting system design problems, external
to collectors (inc. piping, hoods)
a. piping, elbows, abrasion
b. hood inlet control poor
c. others
97
-------
3.2.2. Fabric Filter Level B Evaluation
A Level B Fabric Filter Evaluation includes all elements
of a Level A Evaluation as well as Quantitative Gas Analysis and
Inlet and Outlet Mass Concentration Measurements. These concen-
tration measurements will allow a calculation of the Total Mass
Collection Efficiency.
3.2.2.1 Quantitative Gas Analysis
Quantitative gas analysis involves actual measurements on
the effluent gases to determine the components of the gas mixture.
This usually includes at the minimum determinations of the water
vapor content and amounts of N2, O2, C02, and CO. Depending on
the type of plant, determinations of the S02 and S03 content will
be necessary. At some installations the flue gas or filter bags
are treated with Nahcolite or other S02 adsorbing materials to
remove this gas. In these cases measurements of the S02 content
upstream and downstream of the baghouse might be required. Also
depending on the particular industrial process, there may be other
specific gas components which might affect baghouse operation.
a. ORSAT Measurement of 02, N2, CO, and C02
b. EPA Reference Methods for H20, S02, HaSO.*, NO
c. Alternate methods for determining S02 and S03
d. Other gases peculiar to particular site
e. Inlet and outlet S02, if S02 removal chemicals being
injected.
Refer to Section 2.5.3.
x
98
-------
3.2.2.2 Inlet and Outlet Mass Concentration Measurements
Total Mass Collection Efficiency
In order to determine the overall mass collection efficiency,
the mass concentration of the effluent particulate at the inlet
and outlet of the baghouse must be determined. Usually this will
involve either the EPA Reference Method Number 5 or the ASTM method
as described under Section 2. In most cases simultaneous inlet
and outlet sampling will be necessary, as well as a traversing
capability. Also sampling times at the inlet and outlet will
be very different due to the usual excellent efficiency of bag-
house units.
a. Sampling train type
b. Flue depth
c. Traversing capability
d. Velocity/Temperature Traverse
e. Probe heating required
f. Filter integrity
g. Isokinetic sampling
h. Sampling times
i. Number of samples
j. Process variations
k. Integrated or time averaged sample
Refer to Appendix B and Appendix E.
3.2.3 Fabric Filter Level C Evaluation
A Level C Fabric Filter Evaluation includes the data gathered
under a Level B Evaluation and Inlet and Outlet Particle Size
Distribution Measurements. This data will allow determination
of the control device fractional efficiency.
99
-------
To measure the collection efficiency of a baghouse as a
function of particle size or the fractional efficiency, determina-
tions of the inlet and outlet particle size distributions must
be made. There are several types of instruments which can be
used for this purpose. These include cascade impactors, series
cyclones, optical particle counters, diffusion batteries/condensa-
tion nuclei counters, and electrical aerosol analyzers. In gen-
eral process variations will have an effect on sampling times
for impactors and cyclones at baghouse inlets. Recent experi-
mental data show that outlet emissions from seasoned baghouses
are fairly stable even if the inlet concentration is fluctuating.
Thus inlet sampling devices may have to sample during many parts
of a cyclic operation to obtain a representative inlet size dis-
tribution.
Several sizing systems are listed below along with several
aspects of operation to be considered in developing a test plan.
Cascade Impactors
Size Range 0.3-10.0 micrometers
a. High Loading Low Loading
Brink Andersen
MR I
Sierra
U. of W.
b. Isokinetic Sampling
c. Jet Velocity Limits
d. Nozzle Selection
e. Precutter Selection
f. Loading Limits
Refer to Appendix C and Section 2.3.2.
100
-------
Series Cyclones
Size Range 0.3-10 micrometers
a. Accuracy/Resolution
b. Sampling Times
c. Back Up Filters
d. Filter Integrity
e. Filter Loading
f. Constant Flow Rate
g. Traversing
h. Isokinetic Sampling
i. Extractive Sampling
j. Probe Losses
Refer to Section 2.3.2.
Optical Particle Counters
Size Range 0.3-10 micrometers
Refer to Section 2.3.3.
Diffusional/Electrical Aerosol Analyzer
Size Range 0.01-0.3 micrometers
a. Extractive Sampling
b. Sample Conditioning
c. Dilution of Sample
d. Gas Composition
e. Condensible Vapors
f. Heated Lines
g. Probe Losses
h. Real Time Monitoring
i. Duct Pressures
Refer to Appendix D and Section 2.3.4.
101
-------
3.3 GENERAL PROBLEMS AND CONSIDERATIONS
It is rare that a fabric filter evaluation program does not
encounter several problems in performing the required tests.
These problems can cover a wide range of circumstances and affect
the ability to complete the test program successfully. Although
it is impossible to anticipate every contingency, careful planning
can reduce the likelihood of complete failure of the test program.
A discussion of the more commonly encountered problems and situa-
tions is presented below.
Plant Location
Plant location will generally not be a problem unless it
is a long distance to a city where acceptable accommodations and
supplies are available, or if the nearest airport is not convenient
for shipping equipment or for transportation of personnel. Also,
depending on the time of year, the local weather can force post-
ponement of testing, unusual working hours, or require the con-
struction of special shelters for test crew members required to
work out of doors.
Sampling Location and Accessibility
In most new plants the requirements of compliance test-
ing at the control device outlet (ports, platforms, power,
etc.) have been taken into account in designing the facility.
This is frequently not true of older plants. At many sites,
the equipment must be hand carried or hoisted to the sampling
location. Stack testing can be difficult without a properly
designed platform. Sometimes platforms and scaffolding must
be erected to allow direct access to the sampling location.
It is recommended that a pre-test site survey be conducted
to determine if any platforms or shelter must be built
102
-------
prior to actual testing. This is also a good time to inspect
the entire plant and establish contact with the plant employee
who will be responsible for liaison with the plant managers.
Type of Ports
Almost all sampling of flue gases and dusts requires some
type of port. Before the tests begin it is advisable to know
the location, type, number, and size of the ports that are avail-
able (inlet, outlet, stack). For some types of test equipment,
the ports may be too small and require replacement with larger
diameter ports. The number of ports will also determine the
flexibility that one has in planning for traverses of the duct
to obtain representative samples. There may also be some diffi-
culties with the type of ports that are installed, whether flanged
or threaded internally or externally. Other problems commonly
encountered with sampling ports are the length of the port nipple,
rusting of port caps onto the nipples, and caking of dust inside
the ports which must be chiseled away before sampling can begin,
etc.
Dust Size
The duct size will generally determine whether traversing
is feasible. Traversing twenty foot deep ducts is not a simple
matter, especially if the probes must be heat traced. Special
hoists sometimes must be erected. Small circular ducts usually
cannot be effectively traversed, and in some cases, instruments
that are normally operated in situ, must be operated in an oven
with special sampling probes for extracting the samples.
Flue Gas Velocity and Nozzle Sizes
Depending on the location of the sampling ports, the flue
gas velocity can sometimes be very high or very low. Isokinetic
sampling is highly desirable when sampling dusts. Gas velocities
103
-------
are usually lowest in the transforms immediately upstream or down-
stream from control devices. In these instances the nozzle sizes
required for isokinetic sampling may be larger than standard sizes.
On the other hand high gas velocities can require impractically
small nozzles, especially when sampling less than 14 liters per
minute. If the concern is with particles smaller than about five
micrometers diameter, errors from non-isokinetic sampling are
less significant.
Power Requirements
Depending on the amount of equipment operating at one time,
the accessible power outlets at most sampling locations may or
may not be adequate. In many instances long extension cords are
necessary, and in some cases a transformer is needed to change
the available power to 110 volts. Before testing, the power re-
quirements should be calculated and plant personnel contacted
if it appears that additional power outlets will be required.
Gas Temperature and Dew Point
Under some circumstances the gas temperature can cause diffi-
culties. Too high a temperature can cause galling, metal fatigue,
collection substrate problems, and poor vacuum sealing for in.
situ sampling equipment. Probes and other sampling equipment
may have to be insulated or heat traced to prevent premature
cooling of the gases. Low gas temperatures can be especially
troublesome when the slightest temperature drop can cause excur-
sion through dew points causing condensation within the probe
or on collection filters. H2SOlt condensation, or chemical re-
action can mask particulate weight gains on glass fiber collection
substrates. Usually heating to 17°C above the gas dew point is
recommended to avoid condensation. At some plant temperature
fluctuations can occur as a result of process variations or the
104
-------
amount of excess air in boiler operations. A knowledge of this
type of activity is desirable before testing.
Corrosive Gases
At some locations, particle growth, such as H2SOi, mists,
can mask the true concentration of fine particulate matter. This
can only be alleviated by keeping the gas temperature in the
sampling train sufficiently high or by dilution with clean dry
air. Reevaporation of H2SO., mists requires very high tempera-
tures, and this problem can usually more easily be avoided than
corrected.
Volatile Components
In planning an effective sampling protocol, it is necessary
to consider whether or not volatile components make up a signi-
ficant part of the emissions. Smelting processes are a notable
example of sources where much of the mass emissions consist of
compounds that exist in vapor form at flue gas temperatures, but
condense to form solid particles upon cooling in the atmosphere.
For process streams such as these, the nature or quantity of the
sample is dependent on the temperature of the sampling train.
It is usually advantageous to design a special train with several
stages kept at progressively lower temperatures, in order to fully
understand the nature of the emissions. Sulfuric acid is a good
example of a volatile pollutant that can present control and
sampling problems.
Long and Short Sampling Times
In general, control devices are very efficient particle col-
lectors. At the inlet, high dust concentrations may necessitate
undesirably short sampling durations. Extremely short (less than
five minutes) sampling durations do not allow an adequate period
105
-------
of integration over the plant process cycles unless the cycles
are very stable and long. On the other hand, low dust concentra-
tions at control device outlets sometimes require sampling time
as long as 12 hours, hampering the study of emissions from each
part of a process cycle. It is helpful to use low flow rate sampl-
ing devices at control device inlets and high flow rate devices
at outlets in order to obtain reasonable sampling times.
Laboratory Space
Usually arrangements can be made to obtain the use of a por-
tion of the chemical laboratory normally found at most industrial
plants. In some cases, however, the location of this laboratory
may not be convenient to the sampling site. As part of the pre-
test site survey, a decision should be made as to whether some
type of temporary, mobile lab or trailer would be more convenient
than the plant laboratory space.
Process Cycles and Feedstock Variations
In many plants, such as iron and steel mills and smelting
operations, the effluent gas and dust characteristics vary dramat-
ically over a single process cycle. If the test objective is
to obtain a good average of the emissions, the sampling time is
quite flexible. However, if the test objective is to isolate
emissions from a particular part or from each part of an average
cycle, the sampling time must be short, or the tests interrupted
periodically and run only during the part of interest. At some
plants the supply of fuel or feedstock can change. Normally a
plant will maintain logs of the important process parameters,
and this information should be obtained and correlated with the
test data. This can avoid costly repetition of test procedures
or invalidation of the test data.
106
-------
APPENDIX A
AEROSOL FUNDAMENTALS, NOMENCLATURE, AND DEFINITIONS
A.I GENERAL DISCUSSION
The particulate matter suspended in industrial gas streams
may be in the form of nearly perfect spheres, regular crystal-
line forms other than spheres, irregular or random shapes, or
as agglomerates made up from combinations of these. It is pos-
sible to discuss particle size in terms of the volume, surface
area, projected area, projected perimeter, linear dimensions,
light scattering properties, or in terms of drag forces in a
liquid or gas (mobility). Particle sizing work is frequently
done on a statistical basis where large numbers of particles,
rather than individuals, are sampled. For this reason the par-
ticles are normally assumed to be spherical. This convention
also makes transformation from one basis to another more conven-
ient.
Experimental measurements of particle size normally cannot
be made with a single instrument if the size range of interest
extends over much more than a decimal order of magnitude. Pres-
entations of size distributions covering broad ranges of sizes
then must include data points which may have been obtained using
different physical mechanisms. Normally the data points are
converted by calculation to the same basis and put into tabular
form or fitted with a histogram or smooth curve to represent the
particle size distribution. Frequently used bases for particle
size distributions are the relative number, volume, surface area,
or mass of particles within a size range. The size range might
be specified in terms of aerodynamic, Stokes, or equivalent PSL
107
-------
diameter. There is no standard equation for statistical distri-
butions which can be universally applied to describe the results
given by experimental particle size measurements. However, the
log-normal distribution function has been found to be a fair ap-
proximation for some sources of particulate matter and has several
features which make it convenient to use. For industrial sources
the best procedure is to plot the distribution in different size
ranges separately, rather than trying to characterize the entire
distribution by two or three parameters. The ready availability
of inexpensive programmable calculators which can be used to con-
vert from one basis to another compensates greatly for the lack
of an analytical expression for the size distribution.
A.2 PARTICLE SIZE DISTRIBUTIONS
Figure Al shows plots of generalized unimodal particle size
distributions which will be used to graphically illustrate the
terms which are commonly used to characterize an aerosol. Oc-
casionally size distribution plots exhibit more than one peak.
A size distribution with two peaks would be called bimodal. Such
distributions can frequently be shown to be equivalent to the
sum of two or more distributions of the types shown in Figure
Al. If a distribution is symmetric or bell shaped when plotted
along a linear abscissa, it is called a "normal" distribution
(Figure Ale). A distribution that is symmetric or bell shaped
when plotted on a-logarithmic abscissa is called "log-normal"
(Figure Aid) .
Interpretation of the frequency or relative frequency shown
as f in Figure Al is very subtle. One is tempted to interpret
this as the amount of particulate of a given size. This inter-
pretation is erroneous, however, and would require that an in-
finite number of particles be present. The most useful convention
is to define f in such a way that the area bounded by the curve
108
-------
a. Distribution Skewed Left
b. Distribution Skewed Right
<
LU
5
z"
Q
LU
u
D
O
c. Normal Distribution
LOG a
Q
UJ
U
cc
I-
LOG og
68.27%
LOG D
d. Log Normal Distribution
3630-091
Figure A1. Examples of frequency or particle size distributions. D is
the particle diameter.
109
-------
(f) and vertical lines intersecting the abscissa at any two
diameters is equal to the amount of particulate matter in the
size range indicated by the diameters selected. f then is equal
to the relative amount of particulate matter in a narrow size
range about a given diameter.
The median divides the area under the frequency curve in
half. For example, the mass median diameter (HMD) of a particle
size distribution is the size at which 50% of the mass consists
of particles of larger diameter, and 50% of the mass consists
of particles having smaller diameters. Similar definitions apply
for the number median diameter (NMD) and the surface median diam-
eter (SMD).
The term "mean" is used to denote the arithmetic mean of
the distribution. In a particle size distribution the mass mean
diameter is the diameter of a particle which has the average mass
for the entire particle distribution. Again, similar definitions
hold for the surface and number mean diameters.
The mode represents the diameter which occurs most commonly
in a particle size distribution. The mode is seldom used as a
descriptive term in aerosol physics.
The geometric mean diameter is the diameter of a particle
which has the logarithmic mean for the size distribution. This
can be expressed mathematically as:
log Dj + log D2 + . . . . log D
log Dg = _ S (Ala)
110
-------
or as
Dg = (DlD2D! . . . . DN) V» (Alb)
The standard deviation (a) and relative standard deviation
(a) are measures of the dispersion (spread, or polydispersity)
of a set of numbers. The relative standard deviation is the stan-
dard deviation of a distribution divided by the mean, where a
and the mean are calculated on the same basis; i.e., number, mass,
or surface area. A monodisperse aerosol has a standard deviation
and relative standard deviation of zero. For many purposes the
standard deviation is preferred because it has the same dimensions
(units) as the set of interest. In the case of a normal distri-
bution, 68.27% of the events fall within one standard deviation
of the mean, 95.45% within two standard deviations, and 99.73%
within three standard deviations.
A.2.1 Cumulative and Differential Graphs
Field measurements of particle size usually yield a set of
discrete data points which must be manipulated or transformed
to some extent before interpretation. The resultant particle
size distribution may be shown as tables, histograms, or graphs.
Graphical presentations are the conventional and most convenient
format and these can be of several forms.
Cumulative size distributions — Cumulative mass size dis-
tributions are formed by summing all the mass containing particles
less than a certain diameter and plotting this mass versus that
diameter. The ordinate is specifically equal to
111
-------
where M is the amount of mass contained in the size interval
between D- and D. ,.
The abscissa would be equal to D.. Cumulative plots can be made
for surface area and number of particles per unit volume in the
same manner. Examples of cumulative mass and number graphs are
shown in Figures A2b and A2a, respectively, for the effluent from
a coal fired power boiler. Although cumulative plots obscure
some information, the median diameter and total mass per unit
volume can be obtained readily from the curve. Because both the
ordinate and abscissa extend over several orders of magnitude,
logarithmic axes are normally used for both.
A second form of cumulative plot which is frequently used
is the cumulative percent of mass, number, or surface area con-
tained in particles having diameter smaller than a given size.
In this case the ordinate would be, on a mass basis:
j
Z M
Cumulative percent of mass less than size = t=l x 100%. (A2)
N
E M
t=l c
The abscissa would be log D.. Special log-probability paper
is used for these graphs, and for log-normal distributions the
data set would lie along a straight line. For such distributions
the median diameter and geometric standard deviation can be easily
obtained graphically. Figures A3a and A3b show cumulative percent
graphs for the size distribution shown in Figure A2a and a log-
normal size distribution.
112
-------
1014
UjCO
0 Q
t <
e/> 5
HI 2
_l —
<-> 2
P <
CC X
LL 111
O O
1010
0.01 0.1 1.0
PARTICLE DIAMETER, ;um
a. Cumulative No. Graph
10
cc
HI
-------
re
a. Cumulative Percent Graph
o 99.99
5?
Q
LU
Q
I
LU
O
cc
<
a.
99
3 "
50
10
1
0.01
.02
0.1
b. Cumulative Percent Graph
(Log Normal Distribution)
MMD
I
I
1.0
10.0 .02
0.1
1.0
10.0
PARTICLE DIAMETER,
3630-087
Figure A3. Size distributions plotted on log probability paper.
114
-------
Differential size distributions — Differential particle
size distribution curves are obtained from cumulative plots by
taking the average slope over a small size range as the ordinate
and the geometric mean diameter of the range as the abscissa.
If the cumulative plot were made on logarithmic paper the frequency
(slope) would be, taking finite differences:
AM = Mj " Mj-l (A3)
A (log D) log D- - log D. j^ '
and the abscissa would be D = \/D .D. •, where the size range of
interest is bounded by D. and D.,. M. and M._, correspond to
the cumulative masses below these sizes. Differential number
and surface area distributions can be obtained from cumulative
graphs in precisely the same way- Differential graphs show visually
the size range where the particles are concentrated with respect
to the parameter of interest. The area under the curve in any
size range is equal to the amount of mass (number, or surface
area) consisting of particles in that range, and the total area
under the curve corresponds to the entire mass (number, or surface
area) of particulate in a unit volume. Again, because of the
extent in particle size and the emphasis on the fine particle
fraction, these plots are normally made on logarithmic scales.
Figures A2c and A2d are examples of differential graphs of par-
ticle size distributions.
Log-normal size distributions — The formation of aerosols
by different means frequently result in particle size distribu-
tions which obey the log-normal law. For log-normal particle
size distributions the geometric mean and median diameters co-
incide.
The normal distribution law is, on a mass basis:
115
-------
dM =
exp
(D-D V
V m)
2a;
(A4)
The log-normal distribution law is derived from this equation
by the transformation D "* log D
f =
dM
d(log D) log a
exp
log D-log D
log af
gm
(A5)
where a , the geometric standard deviation, is obtained by using
the transformation D "* log D in equation A4. This distribution
is symmetric when plotted along a logarithmic abscissa and has
the feature that 68.3% of the distribution lies within one geo-
metric standard deviation of the geometric mean on such a plot.
Mathematically, this implies that log a = log Dg4 14 -log D
or log D-log DIS gg where DQ4 14 is the diameter below which
84.14% of the distribution is found, etc. This can be simplified
to yield:
D
84
D,
(A6)
a -
or
(A7)
(A8)
When plotted on log-probability paper, the log-normal distri-
bution is a straight line on any basis and is determined comple-
tely by the knowledge of Dg and a This is illustrated in Figure
A3b. Another important feature is the relatively simple relation-
ships among log-normal distributions of different bases. if D ,
gm
116
-------
D , D , and D „ are the geometric mean diameter of the mass,
gs gvs gN
surface area, volume-surface, and number distribution, then;
log Dgs = log Dgm - 4.6 Iog2ag, (A9)
log Dgvs = log Dgm ~ 1°151 Iogag' and
log DgN = log Dgm - 6.9 Iog2ag. (All)
The geometric standard deviation remains the same for all bases.
More examples of particle size distribution graphs are given
in the data reduction sections of Appendices C and D. The fol-
lowing section in this Appendix lists useful definitions, equa-
tions, and nomenclature for aerosol sampling.
A. 3 NOMENCLATURE AND DEFINITIONS
A. 3.1 Definition of Particle Diameter
Aerodynamic diameter, D, — The aerodynamic diameter of a
£\
particle is the diameter of a sphere of unit density which has
the same settling velocity in the gas as the particle of interest.
Aerodynamic impact ion diameter, D^T — The D,, of a particle
AJ. A J.
is an indication of the way that a particle behaves in an inertial
impactor or in a control device where inertial impaction is the
primary mechanism for collection. If the particle Stokes diameter
is known, Dg, the DAI is equal to:
DAI = Ds >/pC~ ' (A12)
117
-------
where p is the particle density, gm/cm3, and
C is the slip correction factor.
Stokes diameter, D — If the density of a particle is known,
s
the Stokes diameter may be used to describe particle size. This
is the diameter of a sphere having the same density which behaves
aerodynamically as the particle of interest. For spherical par-
ticles, the Stokes number is equal to the actual dimensions of
the particle.
An average density for the particles can be obtained from
volume-weight data using a helium pycnometer if large enough
samples are available. The validity of size information based
on an average density depends upon the uniformity of the density
from particle to particle, particularly with respect to size.
Visual inspection of some size-classified samples from flue gases
sometimes shows a variation in color with size which would seem
to indicate compositional inhomogeneities.
Equivalent polystyrene latex (PSL) diameter — The intensity
of light scattered by a particle at any given angle is dependent
upon the particle size, shape, and index of refraction. It is
impractical to measure each of these parameters and the theory
for irregularly shaped particles is not well developed. Sizes
based on light scattering by single particles are therefore usu-
ally estimated by comparison of the intensity of scattered light
from the particle with the intensities due to a series of cali-
bration spheres of very precisely known size. Most commonly
these are PSL spheres.* Spinning disc and vibrating orifice
aerosol generators can be used to generate monodisperse calibra-
tion aerosols of different physical properties. Because most
*Available from The Dow Chemical Company, P.O. Box 68511,
Indianapolis, Indiana 46268.
118
-------
manufacturers of optical particle sizing instruments use PSL
spheres to calibrate their instruments, it is convenient to define
an equivalent PSL diameter as the diameter of a PSL sphere which
gives the same response with a particular optical instrument as
the particle of interest.
Equivalent volume diameter — Certain instruments, such as
the Coulter Counter, have, as the measured size parameter, the
volumes of the individual particles. Size distributions from
such techniques are given in terms of spheres having the same
volume as the particles of interest.
A. 3. 2 Mean Free Path of Gas
The mean free path of a gas, which is the average distance
that molecules travel between collisions, is an important param-
eter in determining the aerodynamic behavior of particles. For
practical purposes, the mean free path is given with sufficient
accuracy by the following equation
X . - 11 - .
1.01xl06P \ 3 MM
where y is the viscosity of the gas, poise,
P is the pressure of the gas, atm,
T is the temperature, °Kelvin, and
MM is the mean molecular weight.
A. 3. 3 Slip Correction Factor
Stokes law can be applied to submicron particles if a cor
rection factor, C, is used.
119
-------
c =
2A h.23 + 0.41 exp P^J , (A14)
where X is the mean free path of the gas, Mm, and
D is the particle diameter, ym.
The constants in equation A14 were determined empirically
for air at standard temperature and pressure, and are thus only
approximate for stack conditions. If the exponential term is
neglected, equation A14 is referred to as the Cunningham correc-
tion factor.
A. 3. 4 Viscosity of Gas
Viscosity in fluid motion is the analog of friction in the
motion of solids. It is a measure of tangential forces between
layers of fluid in relative motion and the resulting dissipation
of mechanical energy. In aerosol physics, the study of viscosity
is concerned with the force exerted on or the energy dissipated
due to a solid surface in a (relatively) moving gas stream. Gas
viscosity is defined as the force per unit area a moving gas
exerts on a surface divided by the net transverse velocity grad-
ient of the gas at the surface. The unit of viscosity in the
International System of Units is the pascal-second and in the
cgs system is the poise [0.1 pascal-second or g/ (cm-sec)]. In
air and most stack gases, viscosity increases with increasing
temperature.
In order to find the viscosity of the flue gas, y, the vis-
cosity of the pure gas components of the flue gas must first be
found. Viscosity is a function of temperature, and the tempera-
ture difference in different flue gases can be quite significant.
The following equations (derived from curves fitted to viscosity
data from the Handbook of Chemistry and Physics, Chemical Rubber
120
-------
Company Publisher, 54 Edition, 1973-1974, pp. F52-55), are used
to find the viscosities of C02(Ui), C0(y2), N2(y3), 02(Mi,), and
H20(y5) .
M! = 138.494 + 0.499T - 0.267 x 10~3T2 + 0.972 x 10~7T3
y2 = 165.763 + 0.442T - 0.213 x 10~3T2
y3 = 167.086 + 0.417T - 0.139 x 10~3T2
UU = 190.187 + 0.558T - 0.336 x 10~3T2 + 0.139 x 10~6T3
y5 = 87.800 + 0.374T + 0.238 x lO'^T2
where T is the temperature of the flue gas in degrees Celsius.
The units of y are 10~6 g/cm-sec. Next, these values of yj through
ys are used in a general viscosity equation for a mixture of any
number of components (See "A Viscosity Equation for Gas Mixtures"
by C. R. Wilke, Journal of Chemical Physics, Volume 8, Number
4, April 1950, page 517) to find the viscosity of the flue gas:
n
y = £
(A15)
where 4>. • is given by the equation:
(A16)
(4 A/5") i +
[i
and M = molecular weight of a component in the mixture,
X = mole fraction of a component in the mixture,
y = viscosity, g/cm-sec; yi, ^2, etc., refer to the pure
components at the temperature and pressure of mix-
ture, y is the viscosity of the mixture, and
-------
A. 3. 5 Particle-Gas Interactions
Particle relaxation time — For the purposes of this docu-
ment, the particle relaxation time, T, may be defined as the time
required for a particle to accelerate from some initial velocity
to the velocity of the carrier gas.
2P r2C
T = —E - (A17)
where p is the particle density, gm/cm3 ,
r is the particle radius, cm,
C is the slip correction factor, and
U is the gas viscosity, poise.
Particle stopping distance — The particle stopping distance,
I, is the distance travelled by a particle as it accelerates from
zero velocity to the velocity of the gas stream.
I = TV (A18)
where T is the relaxation time (sec) , and
V is the velocity of the gas stream (cm/sec) .
Stokes number — The Stokes number, Stk, is the ratio of
the particle stopping distance to some characteristic dimension
of the sampling system. For example, if the stopping distance
for particles of a given diameter is much smaller than the radius
of a sampling nozzle, (Stk « 1) the particles will be sampled
accurately in spite of flow disturbances due to the nozzle design
or sampling velocity. If the particle stopping distance is com-
parable in magnitude to the nozzle diameter, however, the par-
ticles may cross flow streamlines and either enter or miss the
nozzle in quantities which are not proportional to the particle
122
-------
concentration in the duct. Thus, for Stk on the, order of 0.1
or greater, isokinetic sampling is required. In impaction theory,
the characteristic dimension of the system is the radius or half
width of the jet, R.
2p r2CV
Thus Stk =
9yR
where V0 is the particle velocity, cm/sec,
3
p is the particle density, g/cm,
r is the radius of the particle, cm,
C is the slip correction factor,
M is the gas viscosity, poise, and
R is the radius or half width of the jet, cm.
Particle mobility — The ratio of the velocity of a particle
to the force causing steady motion is called the mobility, b.
b =
where y is the gas viscosity, poise,
D is the particle diameter, cm, and
C is the slip correction factor.
A. 3. 6 Cascade Impactor Terminology
Blank — A blank usually refers to a controlled cascade im-
pactor test run in which the particles are removed by a prefilter
If the measured impactor stage weights are found tochange signifi
cantly and consistently, the normal runs should be corrected for
this background.
Bounce — Bounce in this document refers to inadequate re-
tention of particles in cascade impactors which strike the impac-
tion surface. If the particle does not adhere, it is said to
bounce.
123
-------
Cascade impactor — An instrument consisting of a series
of impaction stages of increasing efficiency with which particles
can be segregated into relatively narrow intervals of aerodynamic
diameters.
Control — A control run is a technique which is used to
confirm that variables are isolated. The control is made up to
be as similar as possible to an actual run, but it is not run
through the test situation. The control is then examined as would
an actual test run. If the experimental variable changes signifi-
cantly, the experiment is not properly set up.
Cut-point — The cut-point of an impactor stage is the par-
ticle diameter for which all particles of equal or greater diam-
eter are captured and all particles with smaller diameters are
not captured. No real impactor actually has a sharp step function
cut-point, but the theoretically defined Dso of a stage is often
called its cut-point.
D5 o — The D50 of an impactor stage is the particle diameter
at which the device is 50 percent efficient. Fifty percent of
the particles of that diameter are captured and 50% are passed
to the next stage.
n = i-*-uij i* M R
50 A /- /> V
where Stk = Stokes number, determined by calibration for 50%
collection efficiency, dimensionless,
jj = gas viscosity, poise,
R = impactor jet radius (for slot impactors, the slot
half width), cm,
Vj = gas velocity through impactor jet, cm/sec,
C = slip correction factor, dimensionless,
p = particle density, g/cm3.
124
-------
D50(AI), aerodynamic impaction diameter, is found by setting
C and p = 1.0.
D50(A), the aerodynamic diameter, is found by setting p = 1.0,
and
D50(S), Stokes diameter, is found by setting p = the actual
particle density.
Extractive sampling — Sampling of a particulate laden pro-
cess effluent stream by means of a probe inserted inside the
process stream duct to allow transport of the gas to some type
of sampling instrument located outside the process stream duct.
Filter — A mat of fibers or a porous membrane used to col-
lect airborne particles.
Grease — In impactor terminology, grease is a substance
which is placed on an impactor stage or substrate to serve as
a particle adhesive.
Impaction — The process in which the inertia of particles
in an air stream that is deflected about an obstacle causes them
to strike the obstacle.
Inertial impaction parameter, f — The in^rtial impaction
parameter, ¥, is similar to the Stokes number, however, the charac-
teristic dimension of the system is the diameter or width of the
jet, not the radius or half width. Thus
Y = 2CV0ppr2/9yD
125
-------
where V0 is the particle velocity, cm/sec,
p is the particle density, g/cm ,
r is the radius of the particle, cm,
C is the slip correction factor,
y is the gas viscosity, poise, and
D is the diameter or width of the jet, cm.
In situ sampling ~ Placement of a sampling device directly
into a process gas stream in order to sample the particles or
gas directly-
Isokinetic sampling — The condition in which ambient air
flow has the same speed and direction as air flowing into a sampl-
ing inlet. This prevents sample bias.
Preconditioning — Unwanted weight changes of impactor glass
fiber collection substrates may be reduced by placing a large
number of substrates inside the duct which is to be sampled, and
pumping or passing filtered flue gas through them for several
hours. Such a procedure is referred to as "preconditioning" the
substrates.
Precutter or precollector — A collection device, often a
cyclone, which is put ahead of the impactor in order to reduce
the first stage loading. This is necessary because in some streams
the high loading of large particulate would overload the first
stage before an acceptable sample had been gathered on the last
stages.
Probe — A probe is any pipe used for the transport of pro-
cess effluent gas from the interior of the process ducting to
a sampling instrument. Usually probes are insulated and heat
traced and have some type of nozzle attachment at the end to be
inserted in the gas stream for isokinetic sampling. In the case
of in situ sampling, the probe is used to connect the sampling
126
-------
instrument inside the duct to accessory equipment outside the
duct. If there is no accessory equipment, the probe is used as
a handle for inserting, transversing, securing, and removing the
sampling instrument.
Rebound — Return of particles to an air stream when they
fail to adhere after striking a collecting surface.
Re-entrainment — Return of particles to an air stream some
time after their deposition on a collecting surface.
Stage — A stage of an impactor is usually considered to
be the accelerating jet (or plate containing multiple jets) and
the surface on which the accelerated particles impact.
Substrate — The removable, often disposable, surface on
which impacted particles are collected. Substrates are charac-
teristically light and can be weighed on a microbalance.
Wall loss — The deposition of particles on sampler surfaces
other than those designed to collect particles. They should be
collected if possible and assigned to the proper position where
they would have been collected.
A.3.7 Temperature and Pressure Standards
Laboratory standard conditions have generally been recognized
for many years as 0°C and 760 mm Hg. Recently the US EPA has
set standard conditions for all stationary source testing to be
20°C and 760 mm Hg. Engineering standards have been defined for
some time and are 70°F and 29.92 in. Hg. In order to avoid con-
fusion, the designation "normal", (N), is used to denote engineer-
ing standard conditions in metric units (21°C, 760 mm Hg).
127
-------
When denoting measures of gas volume, the letter "d", or
"D", (for "dry") is sometimes included to signify that the volume
measured contains no water vapor. In stationary source testing,
the letter "a", or "A", (for "actual") signifies the volume of
the gas at the actual stack conditions, for example, the volume
the gas would have at 200°C, 740 mm Hg, and 10% H2O.
Examples of stationary source testing nomenclature:
s.d.c.f. (or DSCF or SDCF)—standard dry cubic feet—a gas volume
measured at 20°C, 760 mm Hg, and 0% H20.
a.c.f. (or ACF)—actual cubic feet—a gas volume measured at
conditions other than standard, usually given in the text.
ACM (or Am3)—actual cubic meters
ACCM (or Acm3)—actual cubic centimeters
DNCM (or DNm )—dry normal cubic meters—a gas volume measured
at 21°C, 760 mm Hg, and 0% H20.
128
-------
APPENDIX B
PARTICULATE MASS CONCENTRATION MEASUREMENTS
This appendix contains the details for conducting particulate
mass concentration measurements using the EPA Reference Method
5 procedure (See Appendix E). The information is presented in
four sections. Section Bl describes the equipment used in con-
ducting these tests. Section B2 is concerned with preliminary
procedures prior to sampling. Section B3 details the actual mass
sampling procedures. Section B4 deals with data analysis after
testing has been concluded. The sampling system described is
depicted in Figure 21 of Section 2.2.2.
B.I GENERAL SAMPLING EQUIPMENT
B.I.I Temperature Measurement
Several temperature measurements are required in conducting
a test for particulate mass loading, including the temperature
of the stack gas, the particulate filter, and the cooled sample
stream. The relative errors encountered in temperature measure-
ments are usually small since absolute temperatures are used in
all gas law calculations. In source testing, dial thermometers
and thermocouples are usually used for making temperature mea-
surements.
Two common scales are used in temperature measurement, the
Celcius scale and the absolute Kelvin scale. Conversions for
these are shown below for °C to °K and °F to °C.
129
-------
°K = °C + 273°
°C = 5/9(°F - 32)
Mercury bulb thermometers — The mercury bulb thermometer
operates by measurement of mercury expansion with temperature
increase. This expansion is linear over the range of the tem-
perature scale. Glass mercury thermometers break easily and a
risk is involved in using these in source sampling.
Dial thermometers — Two types of dial thermometers are avail-
able for use in sampling. One is bimetallic and the other is
a gas bulb thermometer. The bimetallic thermometer contains a
strip of two different metals bonded together. Because of the
different thermal expansion coefficients of the two metals, the
bonded strip will deform with temperature, and depending on the
configuration of the strip, this deformation will be transferred
to a dial movement which contains a temperature scale.
Gas bulb thermometers rely on the expansion of an inert gas
with temperature. This expansion is sensed as a change in pres-
sure. The dial temperature scale is actually a pressure scale.
Gas bulb thermometers are used for lower temperature ranges.
Thermocouples — Thermocouples are the most popular device
for measuring high temperatures. These consist of two dissimilar
wires welded together at one junction (See Figure Bl). These
two wires are joined to a third wire held at a reference tempera-
ture. The difference between the temperature in question and
the reference junction temperature causes an electromotive force
in the system which can be sensed by a potentiometer. Several
metal pair types are available. Generally Chromel/Alumel is the
most common choice due to resistance to oxidation. This pair
is useful from -184°C to 1260°C. Other common pairs are Copper/
Constantan (-184°C to 350°C), Iron/Constantan (-158°C to 1010°C),
and Platinum/Platinum 10% Rh (0°C to 1538°C).
130
-------
INDICATOR
METAL A
METAL B
'"I
| METAL C
I
LEADS
(LOW RESISTANCE AND
AS SHORT AS PRACTICAL)
METAL C
REFERENCE
JUNCTION
Figure B1. Thermocouple junction.
3630-067
131
-------
B.I.2 Pressure Measurement
Pressure is defined as a force per unit area. Most pressure
measurements are made with local atmospheric pressure as refer-
ence. The pressure above atmospheric is considered positive,
and that below negative. The absolute pressure at a point is
the atmsopheric plus the pressure differential.
An easy way to measure a low pressure is to balance a column
of liquid against the pressure. The magnitude of the pressure
can be calculated based on the measured height of the liquid
column. Devices which do this are called manometers. In source
sampling, manometers are often used for the determination of the
stack gas velocity and the sample train flow rate. For small
pressure differentials, the manometer is often inclined to in-
crease the sensitivity. The inclined manometer is used to measure
the stack velocity pressure and sample stream orifice pressure
differential. It is advantageous to use manometers which have
some means of protection against accidental blow out.
Mechanical pressure gauges are also available to measure
low differential pressures commonly encountered, such as velocity
pressure and orifice meter heads in sampling systems. The Magne-
helic gauge manufactured by F. W. Dwyer Mfg. Co. in Michigan City,
Indiana is an example of such an instrument. Inclined manometers,
however, are generally more reliable and easier to use. They
are also easier to repair.
B.I.3 Velocity Measurement
The measurement of velocity in a duct utilizes one of the
simplest devices in most sampling systems, the pitot tube. These
devices are necessary because it is impossible to determine the
total volumetric flow through large ducts. Only by measuring
the velocity at many points and knowing the area of the duct can
132
-------
an accurate determination of duct volumetric flow be made. The
pitot tube will not directly measure the average duct velocity
but measure only the instantaneous velocity at the point at which
it is located.
Several configurations are possible for pitot tubes. One,
the Prandtl or standard type is shown in Figure B2. The static
pressure is measured at point W. The velocity pressure is mea-
sured at point P. The velocity, V , then is given by
s
Vs -
where V is the gas velocity, cm/sec,
S
P is the measured velocity pressure, mm Hg,
P is the measured static pressure, mm Hg,
C is the pitot coefficient, dimensionless,
2.666 x 10
3 (VPw)
ps
p is the density of the stack gas, gm/cm
S
3
The pitot tube described here is usually called a standard pitot
tube? the C value for this configuration is approximately 0.99.
One configuration which does not resemble the standard pitot
is the reversed or Stausscheibe (S-type) pitot tube. The S-type
pitot tube is used primarily for stack testing because of one
major advantage: it will not rapidly clog in gases with heavy
dust concentrations. An example of one of these is shown in Fig-
ure 33- The S-type does not give the same velocity pressure as
the standard pitot tube. The observed delta P is larger for a
given velocity because the rear part of the tube faces downstream.
This P is a wake pressure which is lower than the static pres-
£}
sure. When used with a water manometer, the S-type pitot tube
equation becomes
133
-------
MANOMETER
3630-069
Figure B2. Standard pitot tube.
134
-------
GAS FLOW
MANOMETER
3630-068
Figure B3. S-type pitot tube.
135
-------
= 101.55 C
where V is the gas velocity, cm/sec,
s
AP is the pressure differential (P -P_) , mm Hg,
P t>
p is the density of the stack gas, gm/cm3,
o
C is the pitot tube correction factor.
P
For the limits of 0.025 to 25.4 cm of water velocity pressure,
C for a standard (Prandtl) pitot tube usually takes on values
of 0.98 to 1.00. The C of an S-type pitot tube usually is be-
tween 0.83 and 0.87. Each must be calibrated before a test, pre-
ferably in a gas stream in which the gas properties and velocity
are similar to those of the test conditions.
The calibration of a pitot tube requires a gas stream of
constant and known velocity. Thus, a wind tunnel facility should
be available. However, the S-type pitot tube can be calibrated
against the standard pitot tube. This is the procedure desig-
nated in EPA Reference Method 2 (Federal Register, 1977).
The following equation applies in this case:
'ptest Cpstd\APtestj
where C = Coefficient of the S-type pitot tube,
ptest
C = Coefficient of the standard pitot tube,
F
std
= Pressure differential measured by the S-type pitot
tube, mm of Hg, and
= Pressure differential measured by the standard
std
pi, tot tube, mm of Hg.
136
-------
To calibrate the S-type pitot tube, the velocity pressure
is measured at the same point with both the S-type and standard
pitot tubes. Both pitot tubes must be properly aligned in the
flow field. The appropriate values are inserted in the above
equation and the coefficient for the S-type pitot tube is cal-
culated. If C is not known, then a value of 0.99 should be
used. Pstd
The coefficients for the S-type pitot tube should be deter-
mined first with one leg, then with the other leg pointed down-
stream. If the computed coefficients differ by more than 0.01,
the pitot tube should not be used without proper labeling.
The determination of the average stack gas velocity, V ,
s
is one of the greatest sources of error in stack sampling. There-
fore, it is recommended that the pitot tube be recalibrated on
a regular basis.
At extremely low or high velocities the pitot method is in-
accurate and unreliable. There are several other mechanical and
electronic methods which are available, i.e., hot wire anemom-
eters, rotating vane anemometers and certain fluidic devices.
B.I.4 Nozzles
The nozzle is considered the initial sampling system bound-
ary. It removes a portion of the effluent from the duct and
delivers it to the sampling probe. The nozzle has several re-
strictions in its use:
1. It should not disturb the duct gas stream flow.
2. It should not alter the particulate being sampled.
3. It should not add to the sample being collected.
4. It should be of a size allowing easy access.
137
-------
For particulate sampling, the nozzle should disturb the gas
flow as little as possible or the sample will not be representa-
tive. Any nozzle will disturb the flow, but a thin wall, sharp
edged nozzle disturbs it the least. In the case of particulates,
any bends in the nozzle will cause impingement of larger particles.
The nozzle must then be cleaned carefully and any material found
within it must be added to the total collected particulate.
B. 1.5 Probes
The probe is the sampling interface between the gas stream
in the duct and the external sampling train (See Figure B4).
It is exposed internally and externally to the flue gas at the
nozzle end and ambient air at the exit end. The probe should
not alter the sample in any way. It must be able to support it-
self. It must be easy to clean and it must not add to the sample.
Ideally the sample should be delivered to the sample train at
the stack temperature. In most instances this requires that the
probe be heated to maintain the sample at stack conditions and
prevent condensation. Structurally the probe must support itself,
the nozzle, the pitot tube, and sometimes a thermocouple. This
strength is obtained by use of a metal sheath around the probe.
The requirement of probe cleaning conflicts with the structural
needs. The surface of a glass tube is much more easily cleaned
than a metal tube, but it is more fragile. In most cases glass
probes over 2 meters in length are impractical. If a metal tube
is used, some material may become trapped in the rough surface
of the probe.
Glass has another advantage which should be considered.
It is for all practical purposes chemically inert. This is not
true of stainless steel especially if there are acid gases present
in the gas stream which is being sampled.
138
-------
GLASS
BALL
JOINTV
GLASS
PROBE
HIGH
TEMPERATURE
TAPE
PROBE
SHEATH
O-RING
FRONT
FERRULE
NOZZLE
RUBBER
STOPPER
POWEi
CORD
UNION
REAR
FERRULE
REAR
FERRULE
3630-070
Figure B4. Probe and nozzle assembly.
139
-------
In the case of extremely high temperatures, the only prac-
tical choice is to use a water-cooled, high quality, stainless
steel probe.
B.I.6 Gaseous Sample Collectors
There are four main types of gas sample collection devices.
One type is the cold trap which condenses vapors in the sample
flow stream. Another type is that which contains a solid adsor-
bent. This removes the gas from the stream by surface adsorption.
The third type is the grab sample container. The fourth uses
the principle of gas absorption by a liquid.
Gas absorbers are generally called impingers or bubblers.
Their efficiency depends on the diffusivity of the gases, the
retention time in the devices, the bubble size, and the gas solu-
bility. There are four general types of impingers; the midget
impinger, the fritted glass bubbler, the modified impinger tip
bubbler, and the most commonly used type, the Greenburg-Smith
impinger.
B.I.7 Particulate Sample Collectors
Filtration is the basic method for particulate collection.
There are three major types of filters available today: the flat
glass fiber filter, the ceramic Alundum filter, and the glass
fiber bag.
Alundum thimbles are subject to variations in their particle
retention efficiency as they are used because of changes in their
porosity as a particulate cake forms on them. Therefore, new
filters should not be used when testing relatively clean gas
streams. In addition, when the thimble is used in relatively
clean gas streams, only small amounts of particulate may be col-
lected and weighing accuracy suffers. There are sometimes prob-
140
-------
lems with unknown penetration characteristics of the glass fiber
bags and flat filters. In many sampling trains a small cyclone
pre-collector is used to remove larger particulate and allow
longer sampling times.
B.I.8 Sample Flow Rate Meters
The sharp edged orifice meter is a simple and accurate method
to measure instantaneous volumetric flow rate. In source sampl-
ing it is used in conjunction with a total volume gas meter.
As the gas passes through the orifice restriction, a pressure
drop is created. The following equation is used for determining
the flow rate through an orifice.
T Ap
where Q = Gas flow rate, cm3/sec,
K = Proportionality factor determined by calibration,
T = Upstream gas temperature, °K,
AP = Orifice meter pressure drop, mm H20,
P = Upstream absolute pressure, mm Hg, and
M = Molecular weight of gas, gm/cm-mole.
For a given orifice, K must be determined by calibration. Km
is a function of Reynolds number and thus will not be a constant
over the entire range of flow rates. However, for a small range
of Q , such as for most sampling cases, K is a constant. Thus,
it is important to calibrate the orifice for the range of flow
rates anticipated. Generally, commercial sampling trains have
orifices with delta P of 0-25 cm H20 over the useful flow rate
range. For calibration, see Section B.I. 9.
141
-------
B.I.9 Total Sample Volume
The total volume of gas sampled must be determined in most
sampling trains. This provides the volume necessary to calculate
the particulate concentration. Dry gas meters with capacities
from 0.094-10.8 liters/sec (0.2 to 150 ft3/min) are generally
used. For sampling, the smallest dial face division should be
5 cm3(0.01 ft3) because the meter movement is not smooth over
one revolution. The dry gas meter is calibrated using a wet test
meter. In a wet test meter the gas displaces water in a chamber
and causes the rotor to revolve. It should be noted that the
gas leaving the wet test meter is saturated with water vapor.
The setup for calibrating a dry gas meter and sampling orifice
is shown in Figure B5. The pertinent information should be entered
on a form similar to that shown in Figure B6. The following
equation is used to determine how well the dry gas meter performs
as compared with the wet test meter. If the ratio defined by
the equation is less than 0.99 or greater than 1.01, then the
dry gas meter should be readjusted and recalibrated.
v pu
Y = : W b
a (V"
where y is the ratio of accuracy of the wet test meter to the
dry gas meter,
Vw is the gas volume passing through the wet test meter,
cm3,
P^ is the absolute barometric pressure, mm Hg,
Td is the average temperature of the gas in the dry
gas meter, °K
TW is the temperature of the gas in the wet test meter,
AM is the orifice meter pressure drop, mm Hg,
Vd is the volume of gas passing through the dry gas
meter, cm3, and
6 is the time, seconds, for sampling the gas volume.
V
142
-------
WET TEST
METER
FLOW
INCLINED
MANOMETER
(Am)
3630-086
Figure B5. Set-up for calibration of dry gas meter and orifice meter.
143
-------
DRY GAS METER NO.
DATE
ORIFICE METER NO.
WET TEST METER NO.
BAROMETRIC PRESSURE Pb =
. mm Hg CALIBRATED BY .
ORIFICE
SETTING
AM
mm Hg
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
WET TEST
METER
VOLUME
Vw 3
cm
DRY TEST
METER
VOLUME
Vd 3
cmj
TEMPERATURE
WET TEST
tw
°K
DRY TEST
fdl
°K
fdo
°K
td
°K
TIME
0
sec
AVERAGE ^
7
km
363O-083
Figure B6. Data sheet for calibrating dry gas meter and orifice meter.
144
-------
The data in Figure B6 can be used to calibrate the orifice meter.
In making these calculations the wet test meter is used for the
flow rate, Q , and
V
o = " w
Qm ~0 *
B.I.10 Gas Conditioning
Often the gas sample must be conditioned or treated before
or while it is passed through the sampling train components.
This is done either to preserve the sample or to prevent damage
to the sampling train. Typical gas conditioning operations in-
clude condensing, drying, heating, and dilution.
Condensing
Condensers are used to remove water and other vapors from
the gas sample. They work on the principle that the partial pres-
sure of water vapor decreases with a decrease in sample tempera-
ture. For example, as indicated in stream tables, the partial
pressure of water vapor at 0°C is only 0.15% of the partial pres-
sure at 150°C. Thus, the use of an ice bath type condenser is
an effective way to remove water vapor from a gas sample. This
provides the moisture content which must be determined in order
to calculate the molecular weight of the stack gas.
The ice bath condenser is also used in source sampling to
protect other components from damage. The deposition of water
vapor and water soluble constituents in such components as the
dry gas meter and pump can cause severe damage.
The ice bath condenser usually consists of several wet and
dry impingers connected in series, but it may be as simple as
a piece of coiled tubing submerged in an ice water bath. A mea-
145
-------
sured initial amount of water is put into the impinger-type con-
denser to assist in the condensation process. When the sampling
train is operated a known amount of sample gas is passed through
the system. By observing the pressure and temperature operating
conditions, the amount of water condensed, and the amount of gas
passing through the system, the moisture content of the gas stream
can be calculated.
Drying
In source sampling trains, gas drying is used to accomplish
the same objectives as condensing. The drying operation is achieved
using special chemicals which have a great affinity for water
vapor. One such chemical is silica gel. The silica gel strongly
adsorbs water and hence its change in weight can be used to cal-
culate the moisture content of the gas stream. Indicating silica
gel, which is granular and has a bright blue color, can be obtained
commercially. As it becomes saturated with water vapor, its color
changes to a light pink. If this method is to be used to deter-
mine the moisture content of a gas sample, care must be taken
to insure that all particulate matter is removed first and there
is no other major constituent in the gas stream which may also
be adsorbed by the silica gel. The silica gel releases (desorbs)
the adsorbed water vapor upon heating to 177°C and can be reused.
Often the condenser and drying tube are used in series to
increase water vapor collection efficiency and obtain a high
capacity for the water removed. Large mesh silica gel (6-16)
is used with a filter support backing to prevent the possibility
of entrainment of small particles which might damage other com-
ponents.
146
-------
Heating
The heating operation is used solely to preserve the gas
sample prior to passing it through the particulate collector.
This is an effective method of preventing condensation of water
vapor and high molecular weight substances. Therefore, it is
common practice to heat sampling probes and particulate collec-
tors such as filters to prevent deposition by condensation. If
such a condensation process were allowed to occur, it would cause
loss of gaseous constituents from the gas sample. In addition
to causing a sampling error these materials could be deposited
in inaccessible areas of the sampling system and lead to later
malfunctions.
Ideally it is a good policy to try to maintain stack gas
temperatures throughout the sampling train preceding the filter.
However, high temperatures favor chemical reactions such as oxi-
dation of hydrocarbons in a gas stream containing appreciable
amounts of oxygen. Low temperatures, as mentioned, are conductive
to condensation of water vapor and high molecular weight hydro-
carbons. Hence a compromise is required and most probes and
heated filter boxes operate at about 120°C. Heat sensitive sampl-
ing train components will not be affected by this temperature.
Water vapor will not condense and some of the safety problems
involved with the handling of hot equipment will be alleviated.
Particulate compliance tests require that a gas stream tempera-
ture no higher than 120°C±14°C be maintained prior to particulate
filtration. Any condensables which are taken out by the probe
and filter under this condition are considered to be part of the
particulate catch. A recent ruling however, allows the use of
temperature up to 165°C±14°C when testing at fossil fuel utility
boilers.
147
-------
Dilution
Addition of a dry gas can be an effective method for pre-
venting condensation. When this gas is added the sample is di-
luted. Condensation is prevented because the dry gas is capable
of supporting a part of the water vapor from the gas sample even
though the temperature of the mixture is reduced. The dry gas
must be added in such a manner that the original sample constitu-
ents are not altered. This could be a problem with respect to
particulate matter because the dilution (mixing) process could
cause such events as particle agglomeration, deposition, and
condensation.
B.I.11 Pumps
The purpose of a pump is to pull the sampled gas through
the sampling train components. The detail of the particular type
of pump required will depend on several criteria. The pump must
provide adequate flow and pressure characteristics and be durable
and portable. The pump must be able to overcome the pressure
drop of the other sampling train components and thereby provide
the desired flow rate. It must be able to provide a wide range
of flow rates as required by isokinetic sampling conditions.
Often the head loss across the filter increases through the sampl-
ing tests. This puts an added burden on the pump which must still
be able to maintain the required sampling rate at the nozzle tip
within the stack.
The pump must be leakless when it is located ahead of the
gas meter in the sampling train. If it isn't, then the metered
volume will be greater than the sampled volume and hence the mea-
sured particulate concentration will be less than the true parti-
culate concentration. The EPA Method 5 sampling train falls in
this category. In many of the other sampling trains, the pump
148
-------
is located after the gas meters and therefore no error is involved
if a leak exists in the pump.
The pump must be durable in that it is exposed to corrosive
environment of the sample gas. During most source tests it is
in constant operation and should be a long life component. The
design should enable this component to be maintained easily; the
key components should be accessible and replaceable with a minimum
amount of time.
The need for portability becomes readily apparent when per-
forming source tests, and consequently a small, lightweight pump
is desirable.
There are several types of pumps suitable for source sampling
trains. All are of positive displacement types which are capable
of producing relatively high vacuums (^686 mm of Hg below atmos-
phere pressure) and operate with a direct linear correspondence
between the flow rate and inlet pressure. In commercial source
sampling equipment, reciprocating diaphragm and rotary vane pumps
are commonly used.
The diaphragm pump operates on the moving diaphragm principle.
Gas is drawn into the chamber on a suction stroke and pushed out
on the discharge stroke. On the suction stroke a suction valve
is open, allowing gas to flow in. On the discharge stroke the
section valve closes and a discharge valve opens allowing the
gas to flow out. This intermittent operation can cause some flow
fluctuation (pulsation) in the sampling train. However, this
problem can be somewhat reduced by running two such pumps in paral-
lel or by a specifically designed surge chamber in the flow line.
The diaphragm in these pumps is made out of metal, rubber or
plastic.
149
-------
The rotary vane pump is one rotor in a casing, which is
machined eccentrically in relation to the shaft. The rotor con-
tains a series of movable vanes which seal against the pump casing,
The vanes are free to slide in and out of the slots as the rotor
turns. If the pump must be leakless, then only the fiber vane
type pump with an oiler should be used. The oiler may have to
be modified so that no ambient air leaks into the system through
the oil bowl.
B.I.12 Flow Control
Flow regulation for most sampling trains is accomplished
by using a throttling valve preceding the pump. This valve varies
the vacuum the pump must work against and thereby changes the
flow rate. A more sophisticated arrangement uses two valves.
One precedes the pump and provides a coarse control while a second
one is installed in a recycle (by-pass) loop to protect the pump
and provide a fine control. This latter arrangement is used in
the EPA Method 5 particulate sampling train. This double valve
arrangement is also easier on the pump, allowing longer pump life.
The major requirements of the flow control valve are: (a)
it allows sensitive flow rate adjustment to meet proportional
sampling (isokinetic) conditions and (b) it does not allow any
leakage. Both these requirements depend upon the valve construc-
tion. The leakage problem poses the same potential error as was
discussed for the pumps. Good quality needle valves are required
in most source sampling applications.
B.2 FABRIC FILTER SAMPLING-PRELIMINARY PROCEDURES
B.2.1 Introduction
Before a sample is taken, several preliminary tests must
be made to determine some of the characteristics of the sampling
150
-------
location and the gas stream. The results of these preliminary
tests are the basis for determining suitability of the sampling
location, the nozzle size, number of sampling points, sampling
time, and nomograph settings.
B.2.2 Physical Sampling Location Characteristics
The sampling location should be at a position where the gas
flow is sufficiently uniform that an accurate sample can be ob-
tained. Eight to ten duct diameters downstream and two duct
diameters upstream from any disturbance such as bends, duct in-
lets, duct outlets, or changes in diameter should give suffici-
ently uniform flow. If the flow at the sampling location is very
uniform, a minimum number of twelve sample points per sample may
be used. If the upstream and downstream diameter requirement
is not met, the flow at the sampling location is likely to be
very non-uniform and require an increase in the number of sampling
points. In such a case, Figure B7 is used as a guide to determine
the number of sample points necessary to define the velocity
profile adequately. Figure B7 is used by reading the number of
sampling points corresponding to both the number of downstream
(A) and upstream (B) diameters and selecting the greater number
of sampling points. A quick pitot tube survey should indicate
whether a sufficient number of points has been chosen to define
the velocity profile adequately. After the number of sample
points is selected, the cross section of the duct is divided into
a number of equal areas as shown in Figure B8. In the case of
round stacks, the sample point is located such that half of the
area increment represented by that point is radially on each side
of the sample point. The location of sampling points is deter-
mined as shown in Table Bl. The area increments must be small
enough to insure that the flow at the sampling point in each area
is representative of the flow in the area; however, the total
number of area increments must be limited enough so that all the
points may be sampled within a reasonable period of time.
151
-------
50
0.5
DUCT DIAMETERS UPSTREAM FROM DISTURBANCE* (DISTANCE A)
1.0 1-5 2.0
2.5
{2 40
O
a.
ai
oo
cc
LLJ
> 30
<
cc
i-
cc
111 __
CQ 20
£ 10
T
T
^
T
A
i
i
1
1
B
J
L
PI
V
7
DISTURBANCE
_
MEASUREMENT
SITE
DISTURBANCE
^
*FROM POINT OF ANY TYPE OF DISTURBANCE (BEND, EXPANSION,
CONTRACTION, ECT. )
I
I
I
I
I
34567 89
DUCT DIAMETERS DOWNSTREAM FROM FLOW DISTURBANCE* (DISTANCE B)
10
Figure B7. Minimum number of traverse points per sample obtained
from "Distances to Disturbances", upstream and downstream.
152
-------
TABLE B.I
Duct Traversing Length Factors
LENGTH FACTORS, K_
L
(Fraction of stk. diam. from inside wall to traverse pt.)
Ul
U)
Traverse
Point
Number
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
2468
.146 .067 .044 .033
.854 .250 .147 .105
.750 .295 .194
.933 .705 .323
.853 .677
.956 .806
.895
.967
10
.025
.082
.146
.226
.342
.658
.774
.854
.918
.975
12
.021
.067
. 118
.177
.250
.355
.645
.750
.823
.882
.933
.979
14
.018
.057
.099
.146
.201
.269
.366
.634
.731
.799
.854
.901
.943
.982
A DIAMETER
16
.016
.049
.085
.125
.169
.220
.283
.375
.625
.717
.780
.831
.875
.9.15
.951
.984
18
.014
.044
.075
.109
.146
. 188
.236
.296
.382
.618
.704
.764
.812
.854
.891
.925
.956
.986
20
.013
.039
.067
.097
.129
.165
.204
.250
.306
.388
.612
.694
.750
.796
.835
.871
.903
.933
.961
.987
22
.011
.035
.060
.087
.116
.146
.180
.218
.261
.315
.393
.607
.685
.739
.782
.820
.854
.884
.913
.940
.965
.989
24
.011
.032
.055
.079
.105
.132
.161
.194
.230
.272
.323
.398
.602
.677
.728
.770
.806
.839
.868
.895
.92.1
.945
.968
.989
-------
CROSS SECTION OF CIRCULAR DUCT
DIVIDED INTO 12 EQUAL AREAS,
SHOWING LOCATION OF TRAVERSE
POINTS.
CROSS SECTION OF RECTANGULAR DUCT
DIVIDED INTO 12 EQUAL AREAS, WITH
TRAVERSE POINTS AT CENTROID OF
EACH AREA.
3630-093
Figure B8. Examples of equal area sample points.
154
-------
In rectangular ducts or stacks, the cross section is divided
into a number of equal area rectangles. The sample is taken at
the centroid of each rectangular area. These areas should be
laid off such that the ratio of the length to the width of the
elemental areas is between one and two.
The minimum number of sample points is twelve, and the same
criteria as for round ducts is used to insure that the velocity
profile is adequately defined. The equivalent diameter of a
square duct is approximated by 2LW/L+W where L and W are the duct
cross-sectional dimensions.
In most cases the dimensions of the ducts obtained from con-
struction drawings are accurate; however, the inside dimensions
should still be measured if feasible, particularly in the case
of horizontal ducts on the bottoms of which dust deposits of con-
siderable thickness are often found. The pitot tube may be used
to make this measurement but the ends should be protected to pre-
vent material from the back wall from clogging the ends. Another
critical measurement is the length of the port extension. With
these measurements, the pitot tube is marked at the points to
be sampled. If the outermost points are less than one inch from
the walls, they should be located at one inch and noted on the
sampling form. The required length of the probe (pitot tube)
for each of the points may be marked with hose clamps, tape, or
other suitable material compatible with flue gas conditions.
Stack pressure is determined with a leveled and zeroed manom-
eter. The pitot tubes are aligned perpendicular to the flow
stream in the stack and one of the two pitot lines is disconnected
from the console. If the stack pressure is positive gauge pressure,
the manometer will show positive deflection with the one pitot
line connected to the positive side of the manometer. If the
155
-------
stack pressure is negative gauge pressure, the manometer will
show positive deflection with the one pitot tube line connected
to the negative side of the manometer. The stack gauge pressure
(P ) in millimeters of Hg is obtained by adding the stack to
ambient differential pressure, APD, to the ambient pressure (PAMB)•
ps = APD + PAMB
The preliminary flue gas temperature is obtained by a suitable
means such as a stem thermometer placed in the sampling port or
a thermocouple with an appropriate readout device.
B.2.3 Velocity Determination
Before velocity measurements are taken, the inclined manom-
eter must be leveled and zeroed and must remain level during sampl-
ing. The openings of the pitot tubes should be shielded from
any wind currents but not be completely closed off when the manom-
eter is zeroed. Correct connection of the pitot tube lines may
be checked by blowing gently on the upstream pitot tube opening
and noting the response on the manometer. The probe is then in-
serted, the pitot reading noted and the pitot tube lines are
switched both on the console and on the probe. If the manometer
reading is the same as that prior to switching the pitot tube
lines it is reasonably certain that there is no significant leak
in the lines. If a leak is detected, it must be eliminated before
any readings are taken.
The pitot tube lines must not be pinched or the tube stopped
up during the traverse. If fluctuations in the manometer are
noted, pieces of cotton or glass wool may be placed in the pitot
tube lines to dampen the fluctuation but should not be packed
too tightly. Since the pitot tube measures pressure differences,
there is no actual air flow through the lines. If one line is
156
-------
completely plugged, however, the results will not be accurate.
In some instances condensation of water within the pitot tube
lines can cause difficulties and erroneous readings.
Once the traverse has been completed and the pressure and
temperature readings have been recorded the velocity may be de-
termined. Velocity may be calculated as follows:
(T ) (4P)
(V.) - (422.67) C
'
avg s s
where (v "\ is the stack gas velocity, cm/sec,
v s/avg
T_ is the average stack temperature, °K,
S
Ap is average stack gas velocity head, mm Hg,
M is the molecular weight of the stack gas, wet
S
basis, gm/gm-mole,
P is the absolute stack pressure, mm Hg, and
S
C is the pitot tube correction factor.
B.2.4 Stack Moisture Content
The stack moisture content is an important factor in stack
sampling. Nozzle size selection and sampling rate are both de-
pendent on the moisture content. A condenser method or a deter-
mination (based on the dry bulb temperature with knowledge that
saturated conditions exist in the stack) are two ways of deter-
mining moisture content. A wet bulb-dry bulb technique requires
less equipment but must be limited to non-acid gas streams with
moisture contents of less than 15% and dew points less than 52°C.
The condenser method works well for most gas streams and is rela-
tively easy to perform.
157
-------
B.2.4.1 Condenser Method
Several condenser techniques can be used to determine stack
moisture content. One such technique uses a Greenburg-Smith
impinger approximately half full of water followed by a straight
impinger approximately half full of silica gel. A measured volume
of stack gas, usually 10 cubic feet, is drawn through the impingers
at a moderate flow rate. The total change in weight of the im-
pingers is the weight of the moisture caught. The impingers
should be in an ice bath while the stack gas is drawn.
With the impinger volume increase in milliliters, V , the
we
stack gas volume in cm3 (V ) corrected by the dry gas meter cor-
rection factor, the absolute average meter temperature in °K (T ),
and the meter pressure in mm of mercury (P ), the moisture frac-
tion (B ) is calculated as follows:
V_ (1243.34 cm3/ml
wo
V (1243.34 cm3/ml(H,0)) + V
we z m
B.2.4.2 Saturation Method
If water droplets are present in the stack and the stack
gas temperature is below 100°C, the gas stream may be assumed
to be saturated. The moisture content is read from the satura-
tion curve on the psychrometric chart at the stack gas tempera-
ture, Figure B9.
B.2.4.3 Wet Bulb-Dry Bulb Method
In this method two thermometers are placed in the gas stream,
One is dry and the other has a wet sock over the bulb. The tem-
peratures are read after they stabilize. If the stack or duct
pressure is near atmospheric pressure, the percent moisture may
then be found from the psychrometric chart, Figure BIO.
158
-------
100,
_l
o
CO
U
CC
cc
o
Q.
cc
HI
80
60
40
20
I
I
I
I
60 80 100 120 140 160
STACK GAS TEMPERATURE, °F
180
200
220
3630-071
Figure B9. Percent water vapor in air at saturation.
159
-------
i—i—i—i rrn
30 40 50 60
70
80 90 100 110 120 130 140
DRY BULB TEMPERATURE, °F
150 160 170 180
3630-072
Figure BIO. Percent water vapor with wet and dry bulb.
160
-------
The percent water vapor by volume is found directly on the
ordinate axis. Inputs are the dry bulb temperature on the ab-
scissa, the wet bulb temperature, and the sloping lines which
terminate at the saturated vapor line.
When obtaining wet bulb-dry bulb readings with a sling psy-
chrometer, the plane of the thermometers should be perpendicular
to the flow of gas. If it is parallel to the flow, the dry bulb
should be upstream of the wet bulb. The gas velocity past the
wet bulb should be from 3.7 to 9.3 meters per second. Sufficient
time must be allowed for the wet bulb temperature to stabilize
or inaccurate results will be obtained.
B.2.5 Molecular Weight of the Stack Gas
The most common method of determining the composition of
combustion effluents is the Orsat apparatus. Although flue gases
vary in composition, they normally contain C02, CO, 02, H20, and
N2. The Orsat analysis determines the quantities of these com-
ponents (except H20) present by successive removal using suitable
absorbents and measurement of the volume changes of the original
sample. The Orsat analysis, as it is normally used, measures
the percentage of C02, 02, and CO in the sample. The difference
is largely N2. By changing the absorbents other components may
also be measured.
After the Orsat analysis has been performed, the molecular
weight of the stack gas may be determined by the following equa-
tion:
M =
s
B
wo
18.0 + B
DG
proportion by vol.
of component on
dry gas basis
X
Msc of
component
161
-------
where Mc is the molecular weight of the stack gas,
s
BDG is dry gas fraction of the stack gas,
M is the molecular weight of each component of the
Ow
stack gas, and
B is the moisture fraction.
wo
B.3 FABRIC FILTER SAMPLING FOR PARTICULATE AND GASES
B.3.1 Isokinetic Sampling
To obtain a representative particulate sample, the sample
must be collected at a rate as nearly isokinetic as possible,
i.e., the kinetic energy of the gas stream in the stack is equal
to the kinetic energy of the gas stream through the sampling noz-
zle. Since the composition of the two gas streams is the same,
this energy balance simplifies to: the velocity in the stack
is equal to the velocity through the nozzle. If a particulate
sample is not pulled isokinetically, inaccurate results may be
obtained.
Whenever an object is placed in a moving gas stream, some
disturbance of the flow patterns will occur. The purpose of
isokinetic sampling is to minimize any disturbance caused by the
sampling nozzle. A sample collected isokinetically through a
sharp-edged nozzle, should create very little disturbance. Fig-
ure Bll illustrates this point.
Large (heavy) particles tend to travel in a straight line
and are not greatly affected by flow disturbances, whereas small
(light) particles tend to follow the flow lines. In a gas stream
with a homogeneous distribution of large and small particles,
over-isokinetic sampling will give a low particulate mass rate
(PMR) because fewer large particles will be caught than are rep-
resentative of the flow stream from which the gas was withdrawn.
On the other hand, under-isokinetic sampling will give a high
162
-------
II
I I.I
Y\\
\\
1. 11
I I I i I I I
ISOKINETIC
OVER ISOKINETIC
UNDER ISOKINETIC
3630-064
Figure B11. Isokinetic flow patterns.
163
-------
PMR due to a greater than representative number of large particles
that will be caught.
The velocity of a gas stream in a stack generally varies
from point to point; therefore, the flow rate of velocity through
the sampling nozzle must be adjusted to maintain isokinetic con-
ditions at each sampling point. In the sampling train, deter-
mination of the nozzle volume and the flow rate through the nozzle
are based on dry gas volume and flow rate measured at approxi-
mately ambient temperatures. For this reason, the flow rate
through the orifice meter which corresponds to the desired flow
rate through the sampling nozzle must be determined. The stack
velocity as measured by the pressure drop (AP) across the pitot
tube and the velocity through the nozzle as measured by the pres-
sure drop (AH) across an orifice meter at the end of the sampling
train must be equal in order to maintain isokinetic flow. To
speed up this calculation, two nomographs were developed by the
old National Air Pollution Control Administration (NAPCA). Through
the use of these nomographs, the proper size sampling nozzle may
be selected and the flow adjustments required to maintain iso-
kinetic sampling conditions may be determined. An alternate
method employs the use of one of the NAPCA nomographs or a cal-
culator for adjusting flow rates along with a mathematical method
of nozzle selection. These two procedures give equivalent results.
B.3.2 Sampling for Effluent Gases
When effluent gases and particulates are to be sampled simul-
taneously the sampling must be performed isokinetically. It is
advisable however, to choose a nozzle which will give a low volu-
metric flow rate to optimize gas absorption efficiency in the
impinger train; 142 cm3/sec (0.3 ft3/min) Or less is desirable.
If only gases are to be sampled and if they are well mixed, iso-
kinetic sampling is not necessary.
164
-------
B.3.3 Nozzle Selection and Nomograph Setting
Isokinetic sampling involves maintaining the flow rate through
the sampling nozzle such that the velocity in the nozzle equals
the velocity in the stack at the sampling point. Obviously, the
flow rate through almost any size nozzle could be adjusted such
that this velocity requirement is met. There are certain physical
limitations, however, placed on nozzle size by the sampling equip-
ment: pump capacity, port diameter, filter efficiency, and the
critical flow through the Greenburg-Smith impingers. Another
limiting factor involves the reliability of the sample. Small
nozzles can yield less representative samples when large particles
are present. As the ratio of nozzle tip area to stack cross-sec-
tional area decreases, the chance of sampling at a point where
the flow is not representative of the flow in the stack area that
the point represents increases. The nozzle should also be at
least a little larger than the largest particles that might be
encountered in the stack. Some guidelines for nozzle size selec-
tion are given in the next section.
B.3.4 EPA Reference Method 5 Procedure
Isokinetic sampling, the condition of equal velocities, im-
plies a mathematical relationship between the two pressure drops,
AP and AH. The pressure drop measured by the pitot tube, AP,
indicates the stack velocity and the desired velocity through
the sampling nozzle. The pressure drop across the orifice plate,
AH, represents the flow rate of dry gas through the dry gas meter.
This relationship has been incorporated into two stack sampling
nomographs, Figures B12 and B13. To operate the nomographs, a
factor which is a composite of the test constants is obtained
from one nomograph and is used to set up the second nomograph.
This factor will be called "C". As the pressure drop across the
pitot tube (AP) changes from point to point, an updated desired
value for the orifice pressure drop (AH) is found from the second
165
-------
AH
.3.0
.2.0
-1.5
•1.0
REF 1
\
% H2O
T °F
'rrr r
150 —
-100-=rr
50-
"S
\
\
0-
-50-
rREF 2
2.0
10-
I 1.0
— as"
0.6~
0.5
EXAMPLES AH = 2.7 in.
H20
AH = 2.7 IN. H2O
Tm = 0°F
%H2O = 30
Ps/Pm= 1.1
FIND C
20-
30:
40
• 0.9
• 0.8
50-
DRAW LINE FROM AH TO Tm(°F) TO OBTAIN POINT A ON REF. 1.
DRAW LINE FROM POINT A TO %H20 AND READ B ON REF 2.
DRAW LINE FROM POINT B TO Ps/Pm, AND OBTAIN ANSWER OF
0.85 FOR C.
3630-061
Figure B12. Stack Sampling Nomograph (side 1).
166
-------
ORIFICE READING
AH
10 1
9 —
8 —
— •;
7 — i
-^
6— j
5— i
4—=
3— £
2 ^
.
i ^^^«
0.9
0.8
0.7—=
_2
0.6—1
0.5-4
0.4—=
;
—
0.3-^
J
-
0.2— -
-
—
™
-
-
1 ~"
0.1
Ref.
— Ref.
2.0
—
1-5 Q
I CORRECTION
FACTOR
1 0
— 0.9
— 0.8
0.7
-
0.6
"
0.5
2500
2000
1500
TOOO
800
600
500
400
300
200
100
0
^^^^^
5
— n
o
~ £,
1-
LLJ
OC
D
~~ (—
zr oc
ZT < §
1— LU
JT i
tS) r—
^~
z__
B~~
1 —
1-
=—
SLIDING
SCALE
V
vv /
^\^/
CUT ALONG LINES
0.001 — i
K FA
AH = IN. H2o
CTOR PITOT READING -
AP -^
0.002— E
_z
0.003—!
0.004— 1
0.005—=
PROBE 0.006-^
TIP DIAMETER 0.008-=
D 0.01 —
C= DIMENSIONLESS
TS °F
K = DIMENSIONLESS
D = IN.
AP= IN. H2O
c— 1.0 :
~
E— 0.9 -E
E 002—^
— 0.8 • :
E— 0.7 0.03—5
— i
0.04 —=
=—0.6 :
^^_ ;
E 0.06—=
=—0.5 ~±
- 0.08 —
=—0.4 -
_ ~
^ —
~ 0.2 -\
H _=
^—0.3 0.3—:
I 0.4—:
~ 0.5— \
~ 0.6 ^f
0.2 0.8 -E
— ~~~
— ™T
_
"^
2—=
3 — =
4—=
r
0.1 5— |
6~
8~E
10 —
3630-084
Figure B13. Stack Sampling Nomograph (side 2).
167
-------
nomograph. The flow rate may then be adjusted to give this de-
sired pressure drop across the orifice which establishes isoki-
netic conditions.
In using the first nomograph, the following parameters are
required to determine the "C" factor which will be carried over
to the second nomograph:
1. Pressure drop (AH) across the orifice plate when 0.75 cubic
feet per minute of dry gas is flowing and the pressure and
temperature are 760 mm mercury and 21.1°C respectively is
determined. Note that the calibration factor of the orifice
must be determined in the lab.
2. Percent H20 in the stack gas is determined prior to sampling.
3. Expected temperature at the orifice must be estimated from
experience. It is usually at least ten degrees Fahrenheit
higher than ambient temperature and is the same as the out-
let temperature from the dry gas meter.
4. The ratio of absolute stack pressure to atmospheric pressure
is also required. This ratio is approximately equal to one.
Proceed to find "C" as follows:
A. Obtain a point on the "reference one" line by connecting
the orifice pressure drop, AH, (for 0.75 cubic feet per
minute) to the temperature of the orifice plate (T ).
B. Draw a line from this point on the "reference one" line to
the percent H20 to obtain a point on the "reference two"
line.
168
-------
C. Connect this point to the pressure ratio point; this line
crosses the "C" scale and gives the value required to set
up the second nomograph.
After the initial velocity traverse, and prior to sampling, set
up the second nomograph as follows:
A. Set the vertical sliding scale so that the "C" factor men-
tioned above is at the reference point.
B. From the velocity traverse, calculate the average AP. A
line from this point to the stack temperature intersects
the probe tip (nozzle) diameter scale.
C. From the nozzles on hand, select one near this size. Re-
construct the line through the stack temperature and use
the selected probe tip diameter to give a reference point
on the A? scale.
D,
A line from the reference point on the AP scale to the per-
manent reference point on the AH scale locates the pivot
point on the K factor line. Lock the pivot bolt that carries
the clean plastic rule at this point. The nomograph is now
set up for isokinetic sampling. Note that stack tempera-
ture is assumed to remain constant.
For each point of the sampling traverse the pressure drop
across the pitot tube, AP, is used as input to the nomograph.
The output is the desired pressure drop across the orifice, AH,
which is required to maintain isokinetic flow. If the tempera-
ture of the stack changes appreciably during sampling, (±5% on
an absolute temperature basis) the pivot point must be reset.
Once the pivot point on the K factor line has been set, it is
suggested that the maximum and minimum AP's from the velocity
169
-------
traverse be used to determine the range of AH'S. This range
should lie between 0.3 and 6 inches of H20. If it does not, a
different choice of nozzle tip diameter should be considered.
B.3.5 General Sampling Procedures
B.3.5.1 Preparing the Glassware
The glassware must be prepared and placed in the sampling
box. All of the glassware should be clean and dry. The glass
liner in the probe should be washed and the nozzle attached to
the probe. In the sampling box, the sample first passes through
a glass cyclone which has a flask attached to the bottom to catch
large particles that the cyclone separates from the gas stream.
The sample then goes through a fritted glass filter holder. The
filter holder is 6.4 cm to 10 cm in diameter. A glass fiber
filter is placed over the fritted surface. The filter should
be preweighed to ±0.0001 grams. From the filter the sample goes
into the condenser section. The type and number of impingers
to be used in this section is dictated by the type of sample to
be taken. The different impinger solutions and their uses are
listed in the following paragraphs. Gummed labels can be used
to label the impingers and should be placed near the top to pre-
vent water from being absorbed from the ice slurry. The impingers
should be weighed to ±0,1 gram before and after sampling. A very
light coating of silicone lubricant should also be placed on the
ball joints connecting the glassware to insure a vacuum-tight
seal. The impinger solution volume is normally 200 ml.
B.3.5.2 Impinger Trains
The following tables give some recommended impinger (IMP)
trains.
170
-------
Preparation of Impinger Solutions
80% Isopropanol: 160 ml Technical Grade isopropanol +
40 ml deionized (DI) water.
6% H202: 40 ml Reagent Grade 30% Hydrogen Peroxide + 160
ml DI water.
0.1 N NaOH: 20 ml 1 N NaOH + 180 ml DI water.
1 N NaOH: 40 grams Reagent Grade NaOH (pellet form)
dissolved in DI water made up to 1.0 liter volume.
Alkaline arsenite solution (0.500 N NaAs02 in 2.5 N NaOH):
Dissolve 100 grams Reagent Grade NaOH (pellet form) and
32.5 grams of NaAs02 in DI water and dilute to 1.000 liter
with DI water.
0.1 N HzSO,,: 10 ml 2 N E2SO^ diluted with DI water to
200 ml. Prepare 2 N H2SO^ by adding 1 volume of Reagent
Grade concentrated HjSO^ (18 M) to 17 volumes of DI water.
0.1 M zinc acetate: 22 grams Zn(C2H302)2. 2H20 in 1.0
liter DI water.
I. Particulate Only
Application: Hot mix plants, ore sintering processes,
gypsum manufacturing, etc.
IMP. POLLUTANT IMPINGER TIP
NO. CONTENTS OR COMPONENT CONFIGURATION
1 DI Water Particulate Straight
2 DI Water Particulate Greenburg-Smith (G-S)
3 Dry Water Straight
4 Silica Gel Water Straight
II. Particulate + Sulfur Dioxide
Application: Non-ferrous smelters, Portland cement
kilns, coal-fired boilers, fluid catalytic cracking
units, sulfuric acid plants, boiler recovery stacks
(kraft paper mills).
IMP POLLUTANT IMPINGER TIP
NO. CONTENTS OR COMPONENT CONFIGURATION
1 80% Particulate, Straight
isopropanol HzSO,,, S03
171
-------
IMP.
NO.
2
3
4
5
CONTENTS
6% H202
6% H202
Dry
Silica Gel
POLLUTANT
OR COMPONENT
S02
S02
Water
Water
IMPINGER TIP
CONFIGURATION
G-S
G-S
Straight
Straight
NOTES: 1. The impinger train must be purged with two
cubic feet of ambient air at the end of the
run to sweep S02 out of impinger #1 into im-
pingers 2 and 3.
2. The first impinger not only is effective in
trapping sulfuric acid but will also trap
submicron particulate that passes through
the filter.
III. Particulate + Chlorine and/or Chlorides
Application: Pulp bleaching (effluent from), Magnesium
plants (drying of magnesium chloride).
IMP
NO.
1
2*
4
5
CONTENTS
DI Water
Alkaline
Arsenite
Solution
Alkaline
Arsenite
Solution
Dry
Silica Gel
POLLUTANT
OR COMPONENT
HC1, Cl~
Particulate
HC1, C12
Particulate
HC1, C12
Water
Water
IMPINGER TIP
CONFIGURATION
G-S
G-S
G-S
Straight
Straight
*NOTES: It is important to measure the volume of this
solution as accurately as possible since the
analysis will be based on the molar quantity of
arsenite remaining.
IV. Particulate in the Presence of Hydrogen Fluoride
Application: Alumina reduction plants and phosphate
fertilizer manufacturing (acidulation process).
172
-------
IMP
NO.
V-
3
4
NOTE:
CONTENTS
Water
Water
Dry
Silica Gel
POLLUTANT
OR COMPONENT
Particulates,
inorganic
fluoride
particulate
Particulates,
inorganic
fluoride
particulate
Water
Water
IMPINGER TIP
CONFIGURATION
G-S
G-S
Straight
Straight
Use filter bypass instead of fritted filter.
Particulate + Ammonia
Application: Effluent from manufacturing, prilling
and drying of ammonium nitrate or ammonium phosphate
fertilizer.
IMP
NO.
1
2
3
4
5
NOTE:
CONTENTS
Water
0.1 N HjjSO^
0.1 N E2SO,,
Dry
Silica Gel
POLLUTANT
OR COMPONENT
Particulate
NH3
NH3
Water
Water
IMPINGER TIP
CONFIGURATION
G-S
Straight
Straight
Straight
Straight
Water is used in the first impinger in order that
particulate may be measured. The contents of this
impinger will be saturated with ammonia. Partial
recovery of ammonia from the first impinger may
be accomplished by purging the impinger train with
approximately 2 ft3 of ambient air.
VI. Particulate + Hydrogen Sulfide
A. Application: Carbon black plants (furnace effluent),
lime kilns (kraft paper mills).
IMP
NO.
1
2
3
4
CONTENTS
0.1 M zinc
acetate
0.1 M zinc
acetate
Dry
Silica Gel
POLLUTANT
OR COMPONENT
H2S
H2S
Water
Water
IMPINGER TIP
CONFIGURATION
G-S
G-S
Straight
Straight
173
-------
IMP
NO.
1
2
3
4
5
CONTENTS
Water
0.1 M zinc
acetate
0.1 M zinc
acetate
Dry
Silica Gel
POLLUTANT
OR COMPONENT
Particulate
S03, H2SO^
H2S
H2S
Water
Water
B. Application: kraft pulp mills, boiler recovery
stacks.
IMPINGER TIP
CONFIGURATION
Straight
G-S
G-S
Straight
Straight
NOTE: With a conventional sampling train H2S and S02
cannot be determined simultaneously.
B.3.5.3 Checking the Sampling Train for Leaks (Vacuum Check)
A vacuum-tight system is necessary to prevent any dilution
air from being pulled into the sampling line. After the probe
and sample box are connected and suspended from the monorail or
other support, the umbilical cord containing the sample line,
pitot tube lines and thermocouple leads may be connected. The
inlet side of the cyclone is then sealed for the vacuum check.
The pump is then started with the coarse valve closed and bypass
valve open. As the coarse valve is slowly opened the vacuum will
begin to increase. The bypass valve is slowly closed until the
vacuum reaches 381 mm gauge. At this vacuum the flow rate through
the dry gas meter should not exceed 9 cm3/sec (0.02 ft3/min).
If a leak is present all connections should be checked to elimi-
nate the leak and the above procedure repeated. The seal on the
cyclone must be removed slowly before the pump is turned off to
prevent liquid backup in the impingers. After the above vacuum
check is performed the probe liner is connected to the cyclone
and a vacuum check on the total system may be performed. The
pump is started again and the coarse valve is slowly opened.
After the flow starts, the nozzle tip is sealed and vacuum should
start to build in the system. When the sample line vacuum reaches
174
-------
381 mm the coarse valve should be closed. The vacuum should hold
steady if there are no leaks. If it holds for about fifteen
seconds, the nozzle tip is opened and the vacuum should drop.
Then the pump is turned off with care not to back up liquid in
the impingers. Under no circumstances should the pump be turned
off while the nozzle tip is sealed. The power line is then plug-
ged in and the heater turned on. The probe heater, if used,
should be connected and the probe heater turned on. The pitot
tube manometer may then be zeroed by shielding the end of the
pitot tube from any wind or disturbance. The ends should not
be plugged, however. The orifice manometer should also be zeroed.
The manometer must be level at this time and throughout the sampl-
ing period. The pitot tube should also be checked to be sure
the lines are hooked up properly. The man on the platform can
blow lightly on the upstream pitot tube while the console operator
checks the manometer displacement. Before sampling begins, the
console operator should be sure the manometers are leveled and
zeroed, the temperature indicator is working properly, and the
nomograph or calculator is properly set up.
The console operator should have decided on how long to
sample at each point. The number of sample points is determined
by generally accepted rules, e.g., velocity traverse, but the
sampling time is based on knowledge of the plant operations and
the approximate particulate loading which may be obtained from
plant personnel on a pretest. The sample collected on the filter
should be large enough to weigh accurately and be representative
of the conditions in the stack; however, care must be exercised
to prevent the filter from clogging, or in wet stacks, the im-
pingers from filling up. It is preferable to sample for at least
an hour. If the process is cyclic and portions of the cycle give
high particulate mass rates (PMR) which are not upset conditions,
these portions of the cycle must be included in the sample. The
sample size should be at least 850 liters at standard conditions.
If conditions permit, some adjustments in nozzle size and sample
175
-------
flow rate can often be made to satisfy time requirements. It
is usually best to sample each point for no less than three min-
utes as this allows time for adjusting flows and recording data.
B.3.5.4 Sampling
When the console operator is ready and the probe and sample
box have heated sufficiently, the initial dry gas meter reading
is recorded and the sampling probe is pushed carefully into the
duct to the point nearest the back wall. This allows the probe
to cool in hot stacks as it comes out, shortening the time re-
quired for cooling after the sample is taken. This also allows
the use of the stack heat to help heat the probe. The nozzle
must not hit the back wall or the inside of the port where de-
posited material might contaminate the sample. If this procedure
is followed, the last point sampled will be the point nearest
the port. As soon as the probe is positioned the operator should
record the time and start the run by turning on the pump, opening
the coarse valve and adjusting the bypass valve until the desired
flow rate calculated from the nomograph or K factor is obtained.
The data should then be recorded. The probe crew should be noti-
fied 15 to 30 seconds before the probe is to be moved and the
signal to move to the next point is given approximately 5 seconds
before that time. When the probe is repositioned, the operator
should read the new Ap, use the nomograph or calculator, adjust
the flow rate, and record the data required on sampling field
data sheets. The opening at the port should be plugged to prevent
dilution or abnormal distortion of the flow patterns in the stack.
After the last point has been sampled, the operator turns off
the pump and records the meter reading. If more than one port
is to be sampled, the sampling box is transferred to the next
port and the above procedure repeated. The probe is then removed
carefully from the port so the open end of the nozzle does not
hit the port. The probe should also be kept horizontal and the
176
-------
nozzle plugged as soon as possible after it is removed from the
stack to prevent loss of sample. In some cases condensate will
collect in the probe and if the probe is tipped, some of the
sample might be lost. The particulate trapped in the probe can
represent a significant portion of the total sample. Normally
three sample runs are taken. Each run normally consists of two
traverses. Each run is considered as a separate sample, and the
calculations are performed for each sample. The results of the
runs are then averaged.
Before making a second run the percent isokinetic of the
first run should be checked. The nomograph or calculated K factor
should be changed and an additional sample taken if the percent
isokinetic varies by more than ±10% from 100%. (See Data Reduc-
tion Section.)
B.3.5.5 Sample Handling
At the end of each sample run the electrical power is dis-
connected and the hot side of the sample box is opened to allow
the glassware to start cooling. The pitot tube lines to the probe
are disconnected and the probe is removed from the box and the
ends plugged as soon as possible. The impingers may now be re-
moved and weighed. After the impingers are weighed, the liquid
in the impingers is placed in clean sample bottles to be taken
to the lab. Some particulate or condensable compounds will oc-
casionally get past the filter and be collected in the liquid.
The sample catch from the filter and the probe washings should
be collected and stored separately from the impinger catch. The
probe and nozzle are washed carefully and the washings collected
in a clean sample bottle to be taken to the lab, evaporated and
then weighed. The cyclone and flask and the connecting glass-
ware are washed and the washings added to the probe washings.
The weight of any water caught in the cyclone must be determined
and added to the impinger weight gain. The filter is removed
177
-------
from the holder, folded with the participate side in and placed
in an envelope to be taken to the lab and weighed. The filter
holder is washed and the washings added to the probe wash. Usu-
ally distilled water is used to wash the glassware, but in some
instances, acetone may be used. Precautions must be taken to
eliminate the possibility of tampering with, accidental destruc-
tion of, and/or physical and chemical action on the samples.
To reduce the possibility of invalidating the results, all
components of the sample should be carefully removed from the
sampling train and placed in sealed, nonreactive, numbered con-
tainers. The samples are then delivered to the laboratory for
analysis. It is recommended that this be done on the same day
that the samples are taken. If this is impractical, all samples
should be placed in a carrying case (preferably locked) in which
they are protected from breakage, contamination, and loss.
Each container should have a unique identification to in-
sure positive identification and to preclude the possibility of
interchange. The identification of the container should be re-
corded on the analysis data sheet so it will be associated with
the sample throughout the test and analysis.
The samples should be handled only by persons associated
in some way with the task. A good general rule to follow is "the
fewer hands the better", even though a properly sealed sample
may pass through a number of hands without affecting its inte-
grity.
B.4 MASS CONCENTRATION DATA REDUCTION
After performing particulate mass measurements on the inlet
and outlet of a fabric filter, the data must be reduced to ob-
tain particulate emissions concentrations and other pertinent
parameters.
178
-------
To illustrate these calculations, an example of data col-
lected on a single test will be presented along with the appro-
priate calculations to obtain the necessary results.
From the particulate source test data given in Table B2 we
find:
Orifice AP „„ = 3.69 mm Hg
ci V ^
T = 23.6°C = 296. 9 °K
mavg
B.4.1 Volume of Gas Sampled
T .,(P. +AH
std\ bar
mc mstd m Tm Pstd
294 (749 + 3.69)
= 2.8 2g? 7go
=2.75 DNM3
where V is the volume of gas sampled through the dry gas
meter at standard conditions (21°C, 760 mm Hg),
DNM3 .
V is the volume of gas samples through the dry gas
meter at meter conditions, m3.
Tstd is 21°C, 294°K.
Tm is the average dry gas meter temperature, 23°C,
297°K.
PL is the uncorrected barometric pressure at the
outlet of the orifice meter, 749 mm Hg.
AHave is the average pressure drop across the orifice
meter, 3.69 mm Hg.
Pstd is the abs°lute pressure at standard conditions,
760 mm Hg.
179
-------
TABLE B2
EXAMPLE - PARTICULATE SOURCE TEST DATA - BAGHOUSE INLET
General
Firm - CM & S Public Service
Source - Coal Boiler #1
Sampling Location - Baghouse Inlet
Personnel - K.M.C., J.D.M., & W.B.S.
Date - 3/18/77
Test No.-5
Witnessed by - D.B.H.
Equipment
Gas Meter No.-3
Sample Box No.-8
Probe No.-3 Length-305 cm Cp-.87
Nozzle Dia. - 0.89 cm
Filter No. Silica Gel No.
212 87
Probe Wash Sample No. - KC10
Conditions
Orifice AH - 3.74 mm Hg
Assumed %H20 - 8% Tg - 149°C
Assumed AP...--1.5 mm Hg
AVG P./P -1.0
Assumed TM~26.7&C mA - 9.29 m2
Bar Pressure - 749 mm Hg
C-Factor - .860
Probe Heater Set - 135°C
Filter Oven Set - 121°C
Traverse Time
Pt. (sec)
1
M 2
00
f*t
0 3
4
5
6
7
8
9
10
11
12
300
300
300
300
300
300
300
300
300
300
300
300
Dry Gas
Meter
(xlO-3m3]
53.5
244.4
415.7
677.4
929.5
1088.3
1311.5
1613.4
1928.6
2137.9
2331.3
2560.1
Pitot AP
1 mm Hg
.99
1.08
1.18
0.45
1.53
1.62
1.77
1.76
1.55
1.18
] .12
1.08
Meter
Orifice AH Inlet
mm Hg °C
4.52
4.89
4.86
4.30
3.31
3.16
2.97
2.80
3.29
3.10
3.38
3.74
23
24
24
24
25
25
25
25
26
27
28
28
Temp.
Outlet
°C
20
20
21
22
22
22
22
22
22
23
23
24
Stack
Temp.
°C
146
146
150
150
145
147
151
151
153
150
147
145
Pump
Vacuum
mm Hg
152.4
152.4
152.4
177.8
177.8
152.4
152.4
152.4
127.0
152.4
177.8
177.8
Stack
Pressure Test End - 12:00 noon
-7.5 mti Hg Test Start-ll:00 a.m.
Impinger Outlet
Max Temperature
10°C
Condensate
Collected
#1 100 ml
#2 50 ml
#3 10 ml
AH =3.69 mmHg
CO:
11.0%
11.5%
12.0%
Silica
Charge
Filter
Oz
6.0%
5.5%
5.8%
Gel Mass
- +46 grams
and Probe
Wash Catch - 150 mg
Totals
3600
2800
Averages
1.28
3.69
25.3
21.9
148.4
-------
B.4.2 Volume of H,0 Vapor in Stack Gas
Water condensed in impingers = 160 ml, water absorbed on
silica gel = 46 grains.
- v _ ?nfi ii°0 (6.2383x10") (294) _
"" 18-° 76°
where V is the total volume of water in the sampled gas
WC
at standard conditions (21°C, 760 mm Hg), m3
V1 is the total volume of liquid H20 collected in
c impingers and on silica gel, 206 ml.
PH n is the density of water at standard conditions,
H2U
1.00 g/ml.
R is the ideal gas constant, 6.2383 x 101* mm Hg
cm3/gm mole-°K.
Tstd is the standard temperature, 21°C, 294°K.
P . , is the standard pressure, 760 mm Hg.
MH 0 is the molecular weight of water, gm/gm-mole.
B.4.3 Moisture Content of Stack Gas
Bwo = V = 2.75'f0.28 = °'0924 or 9'24%
mstd wc
where B is the mole fraction or proportion by volume of
WO
H20 vapor in the stack gas.
1 S
wc
V is 0.28 m3 H20 vapor at standard conditions.
V is 2.75 m3 of dry sampled gas at standard conditions.
mstd
181
-------
B.4.4 Molecular Weight of Stack Gas
From Orsat analysis per EPA Method 3:
CO2 ave = 11.5%
02 ave = 5.8% by volume (Dry basis)
CO ave = 0%
From moisture analysis per EPA Method 5:
H20 avg = 9.24%
From EPA Method 3:
% N2 (dry basis) = 100% - (% CO2 + O2 + % CO) = 100% - (17.3%)
= 82.7%
Mw = (1-Bwo)(BC02MC02 + B02M02 + BCOMCO + BN2MN2)+ BwoMH20
= (1-.0924) [(0.115) (44.0) + (0.058) (32.0) + (0.0) (28.0) +
(0.827) (28.0) ]• + (0.0924)(18.0) = 28.96 gm/gm-mole
where M is the molecular weight of the stack gas,
gm/gm-mole, wet basis.
B. is the mole fraction of the component gas.
M. is the molecular weight of the component gas.
B is the mole fraction of the water vapor in the
VVCJ
stack gas.
B.4.5 Excess Combustion Air (per EPA Method 3)
(%Q2) - 0.5(%CO)
0.264(%N2) - (%02) + 0.5(%CO)
(5.8) - 0.5(0)
0.264(82.7) - (5.8) - 0.5(0)
182
-------
where %EA is the percent excess combustion air.
%02 is the percent 02 by volume on a dry basis.
%N2 is the percent N2 by volume on a dry basis.
%CO is the percent CO by volume on a dry basis.
0.264 is the ratio of 02 to N2 in air by volume.
B.4.6 Particulate Emissions Concentration
M
C1 = ^— = ^VS = 54.55 mg/DNM;
mstd
where C' is the concentration of particulate matter in the
S
stack gas, mg/DNM3.
Mn is the total particulate mass collected on filter
media, 150.0 mg.
V is the volume of stack gas sampled (volume through
c t- r\
dry gas meter), DNM3.
B.4.7 Stack Gas Volumetric Flow Rate (per EPA Method 2)
h
V \ = K C (v£p~)
s/ ava P P V '
avg
avg
T
s avg
P_M
s w
where (v \ is the average stack gas velocity, m/sec.
\ / ava
avg
K is
p " sec Igm-mole-
C is the pitot tube coefficient.
C
(T ) is the average stack gas temperature, °K.
avg
\,)avg is tne average of the square roots of velocity
heads of stack gas, in H20, determined ac-
cording to EPA Methods 1 and 2.
183
-------
p is the absolute stack gas pressure, mm Hg,
s
M is the molecular weight of the stack gas,
w
gm/gm-mole (wet basis).
From Table B2:
avg
= 1.13
C = 0.860
P
(Ts)avg = 421.7'
Ps = Pbar + (Ps) = 749 +
= 748.25 mm Hg
M =28.96 gm/gm-mole
= (128.83)(0.860)(1.13)
avg
421.7
(748.25)(28.96)
/V \ =17.47 m/sec
\ s'avg
Then calculating the average stack gas volumetric flow rate:
= 3600
(>-
B V
T . ,P
std s
^avg Pstd
where Q is the average volumetric flow rate at dry standard
S
conditions, DNM3/h
BWQ is the mole fraction of water vapor in the stack gas,
dimensionless, and
A is the cross-sectional area of the stack at the
o
sample point, 9.293m (from the Particulate Test
Data Sheet).
184
-------
Then,
Q = (3.600 x 103) (.9076) (17.47) (9.29)
S
= 3.64 x 10s DNM3/h
Therefore, the particulate inlet mass emission rate can be
calculated by PMR = 5 c
S S S
where PMR is the average particulate inlet mass emission rate,
S
(mg/hr calculated by the concentration method).
PMR0 = 3.64 x 10s ^--54.55 mg/DNM3
s nr
= 1.986 x 107 mg/hr of particulate.
B.4.8 Average Isokinetic Ratio
V
avg (V )
v s/avg
where ^va •"•s tne ave^age percent isokinetic.
( avg
v« )-,,,« is tne average gas velocity into the nozzle
n / aVy
entrance.
\ s)a is the average gas velocity of the stack gases.
Writing the equations for the nozzle velocity independent
of the velocity pressure measured in the stack.
V T P
me s m
T P M,
Iv \ = m s d
V nJav9 An 6
185
-------
where V is the average nozzle velocity.
n avg
V is the volume of gas through the dry gas meter,
m i
DNM (dry) .
T is the average absolute stack gas temperature °K,
s
P is the average absolute stack pressure, mm Hg.
S
P is the average absolute dry gas meter pressure,
mm Hg
T is the average absolute meter temperature, °K
M, is the average mole fraction of the dry stack
gas, (1-BWQ).
A is the area of the nozzle, m2.
9 is the total sampling time, sec.
Then,
(2.8) 421.7 752.69 1
_ 294 748.21 .9076 ,Q 00 .
Vn ^n = (6.22 x 10V) (3600 sec) = 19'88 m/S6C
V avg ,
'avg ' -TTTvg ' 17717 ' *•" «
To conform with the federal performance standards, the
average isokinetic rate Iavg, must be >0.90 and <1.10. In
this example, since lavg does not fall within these limits,
this test would be discarded.
186
-------
B.4.9 Mass Collection Efficiency Calculation
This example has shown calculations for a test at a bag-
house inlet. Similar calculations are made on data taken at
baghouse outlets. After determining the inlet and outlet
mass loading concentrations, the fabric filter efficiency
can be calculated from
PMRT -, .. - rt ,., t
%EFF = ln Outlet x 10Q%
j , .
Inlet
= Particulate inlet mass emission rate.
PMROutlet = Particulate outlet mass emission rate,
187
-------
APPENDIX C
CASCADE IMPACTOR SAMPLING TECHNIQUES
This appendix contains detailed guidelines for the opera-
tion of cascade impactors to measure the particle size distri-
butions at the inlet and outlet of fabric filters. These instruc-
tions are taken from "PROCEDURES FOR CASCADE IMPACTOR CALIBRATION
AND OPERATION IN PROCESS STREAMS", EPA-600/2-77-004, by D. B.
Harris. Minor modifications have been made to make them specifi-
cally applicable to fabric filter evaluations, and a more detailed
description of data reduction techniques and data presentation
formats has been added.
C.I THE PRESURVEY
The key to performing a successful fractional efficiency
evaluation is thorough planning based on a complete pretest site
survey. The survey should provide adequate information at as
low a cost as possible. Some sites will require more informa-
tion and some less. As far as is possible, the information noted
during the presurvey should be measured rather than obtained from
plant records or personnel.
As the presurvey is generally conducted by one or two men
"traveling lightly," the apparatus used during the presurvey
should be as light and compact as possible. A presurvey sample
train is shown in Figure Cl. This system can be built into a
single, suitcase-size package, and serves well as a presurvey
sample train. The impactor which is to be used during the main
test program should normally be used during the presurvey. This
188
-------
0-2 IN. H2O
DIFFERENTIAL PRESSURE
INDICATORS
0-10 IN. Hg
NEEDLE
VALVE
3-WAY
VALVE
MOISTURE
TRAP
COOLING
COIL
STATIC-IMPACT
PI TOT-TUBE
STACK ATMOSPHERE
STATIC PRESSURE
™tbbUHt IMPAC1
COLLAPSIBLE
PITOT-TUBE
j
(
L ^
TEMPERATURE
W/CONTROLLER
0-500° F
OUT
IMPACTOR
HEATING
TAPE
Figure C1. Presurvey sampling with a cascade impactor.
i!
OUTPUT
TO HEATING
TAPE
TEMPERATURE
SENSOR
SIGNAL
3630-082
189
-------
is because the suitability of substrates and adhesives must be
checked out. These problems are discussed more fully in later
sections.
In general, the presurvey work should be done using the tech-
niques described in this manual. Less precision is required,
but the accuracy must be high enough to provide useful informa-
tion in designing the test program. The decisions which must
be made are summarized in Table Cl.
C.2 EQUIPMENT SELECTION
C.2.1 Impactor Selection
The selection of the proper impactor for a particular test
situation is primarily dependent upon the mass loading of the
gas stream and its effect on sampling time. There are three major
criteria to be met to match an impactor to a particulate stream.
1. The sampling period must be long enough to provide a
reasonable averaging of any short term transient in the
gas stream.
2. The loading on a given impactor stage must be low enough
to prevent re-entrainment.
3. The sampling rate through the impactor must be low enough
to prevent scouring of impacted particles by high gas
velocities.
For these reasons, an impactor with a comparatively low
sample rate must be used in a gas stream with a high mass loading.
The low sample rate allows a longer sampling time, although in
some situations it will still be undesirably short. Conversely,
190
-------
Item
TABLE Cl. IMPACTOR DECISION MAKING
Information Required
Criteria
Impactor
Sampling rate
Nozzle
Pre-cutter
Sampling time
Collection substrates
Number of sample points
Loading and size estimate
Loading and gas velocity
Gas velocity
Size and loading
Loading and flow rate
Temperature and gas
composition
Velocity distribution
and duct configuration
a. If particle concentration below 5.0 pro is less
than 0.46 gm/am3 (0.2 grain/acf), use high flow
rate impactor.
b. If particle concentration below 5.0 \m is greater
than 0.46 gm/am3 (0.2 grain/acf), use low flow
rate impactor.
a. Fixed, near isokinetic
b. Limit so last jet velocity does not exceed:
60 m/sec greased
35 m/sec without grease
a. Near isokinetic, ±10%
b. Sharp edged; min 1.4 mm ID
If pre-cutter loading is comparable to first stage
loading—use pre-cutter.
a. Per Figure C4
b. No stage loading greater than 10 mg
a. Use metallic foil or fiber substrates whenever
possible
b. Use adhesive coatings whenever possible
a. At least two points per station
b. At least two samples per point
-------
in a low mass loading situation such as a control device outlet,
a high sample rate device must be used if a significant amount
of sample is to be gathered in a reasonable amount of time.
A cascade impactor can normally yield useful information
over a range of sample rates differing by a factor of 2 or 3.
As high efficiency control devices cause the outlet mass loading
to differ from the inlet by a factor of 103, the same impactor
can seldom be used on both inlet and outlet. Both high and low
flow rate impactors are usually required to determine the frac-
tional efficiency of fabric filter installations. Some commer-
cial impactors are constructed such that their stage and nozzle
configurations can be altered, and they can serve as either high
or low sample rate impactors. Others are fixed with respect to
sample rate.
C.2.2 Sample Trains
Figure C2 is a flow diagram of a typical impactor sampling
train. As shown, it is desirable to have the impactor inside
the stack with a straight nozzle. The various parts of the sample
train are discussed below.
A sampling probe leading to an impactor outside of the duct
should be used only if absolutely necessary. The probe should
be as short as possible and contain the fewest possible bends.
It is recommended that a pre-cutter cyclone be mounted at the
probe inlet to remove particles larger than approximately ten
micrometers and thus reduce line losses.
Heating system — The criteria for heating are given in Sec-
tion C.4.3. If heating is required, the entire impactor must
be either wrapped in a heating tape or put in a custom-fitted
heating mantle. The temperature control should be based on the
192
-------
HEATING
JACKET
TEMPERATURE
CONTROLLER
INPUT T
OUTPUT
TO JACKET
HEAT
EXCHANGER
ICE BATH
BLEED
CALIBRATED
ORIFICE
MM Ml
MANOMETER
MANOMETER
VENT
u
VACUUM
DRY GAS PUMP
METER
LEGEND
© - PRESSURE MEASUREMENT POINT
m- TEMPERATURE MEASUREMENT POINT
Figure C2. Typical sample train with a heated impactor.
3630-081
193
-------
temperature at the outlet end of the impactor. Often the tem-
perature is measured between the last stage and back-up filter.
The impactor temperature can be controlled either manually or
automatically- An automatic controller has usually been found
to be worth the money by releasing the operator for other tasks.
Flue gas conditioning — It is usually necessary to cool
and dry the flue gas before it reaches the flow measuring sec-
tion. Condensation in the orifice would distort the measurement;
also, it is useful to protect the equipment from the condensate,
which, in S02-containing gases, is likely to be sulfuric acid.
The type of condensers shown are usually satisfactory. Packed
bed drying columns are commercially available. The heat exchange
coil is used to bring the gas temperature to essentially ambient
so that there will not be significant temperature gradient across
the flow measuring devices.
Flow measurements — At least two flow measuring devices
are used in series. Normally, a calibrated orifice is used in
conjunction with a dry gas meter, as shown. At very low sample
rates the dry gas meter may be inaccurate. The commonly used
diaphragm-type positive displacement gas meter becomes increas-
ingly inaccurate at flow rates less than five percent of rated
capacity. For a typical stack sampling gas meter this would be
approximately 23.3 cm3/sec (0.05 cfm). Another calibrated orifice
or a rotameter should then be used as the second flow meter.
Vacuum pumps — The vacuum pump should usually be placed
at the end of the sample train. This is because vacuum pumps
tend to leak and all of the flow measurements must be made up-
stream of any leak. The flow rate can be controlled by using
an inlet side air bleed or with a recirculating bypass.
194
-------
Pressure measurements — Most of the pressure measurements
are made with manometers, but calibrated differential pressure
meters are equally acceptable. The in-stack pressure needed is
the static pressure, which is not exactly the downstream pressure
of an S-type pitot tube. A true static pressure measurement
should be made. It is not necessary that this be part of the
impactor train, but it can be.
The pressure at the downstream end of the impactor, between
the last stage and the final filter, must be known. It can be
measured, but this is often inconvenient. If a flow rate pres-
sure drop calibration is available for the impactor (without final
filter), it is normally acceptable to calculate the pressure drop.
Correction must be made for pressure and temperature differences
between the calibration conditions and the actual conditions.
The pressure at the inlet to the metering devices must be
known. In the system shown in Figure C2, the pressure is metered
ahead of the calibrated orifice and the orifice pressure drop
is used to calculate the pressure going into the dry gas meter.
The dry gas meter pressure should be measured if there is a reason
to think the procedure above was not adequate.
Temperature measurements — It is necessary to know the tem-
perature at all points where the pressure is measured. Any con-
venient device of known accuracy can be used to make the measure-
ments. The measurement in-stack can easily be made at the probe
end with a thermocouple. The temperature at the downstream end
of the impactor is made directly behind the final filter and is
used to control the heating tape if one is used. If the heat
exchanger in the train brings the gas temperature to about am-
bient, only one temperature reading will be necessary at the flow
meters. This is usually most conveniently done at the dry gas
meter, as taps are available on the meter.
195
-------
C.2.3 Balance Requirements
For the accurate weighing of the collected material a balance
with a sensitivity of at least 0.05 mg is required. This is
especially true for the lower stages of the low sample rate im-
pactors where collection of 0.3 mg or less is not uncommon. The
balance must also be insensitive to vibration if it is to be used
in the field. It is also desirable to have a balance with a large
weighing chamber. These capabilities are available in several
electrobalances marketed in the United States.
More information on selected sampling train components can
be found in Appendix B, Section B.I.
C.3 IMPACTOR SUBSTRATES
C.3.1 Collection Substrates
For reasons which have been discussed, very accurate deter-
minations of impactor stage catch weights are necessary. Impactor
stages are generally too heavy for the tare capacity of field-
usable precision balances. Thus, a substrate which can be weighed
on the balance is used. Generally, these substrates are made
of metal foil or glass fiber.
Glass fiber substrates — Glass fiber substrates are used
on some commercial impactors. In addition to providing a light-
weight impaction surface, glass fiber mats greatly reduce re-en-
trainment due to particle bounce. They are superior to greased
metal substrates in very high temperature applications where the
greases tend to evaporate. Care must be taken when using glass
fiber substrates in streams containing sulfur dioxide, however.
Recent experimentation (Felix, et al, 1977) has shown that glass
fiber materials often exhibit anomalous weight gains due to sul-
fate uptake on the substrates. Apparently, sulfur dioxide in
a gas stream can react with basic sites on most glass fiber ma-
terials and form sulfates.
196
-------
There are two approaches to alleviate this problem. Sub-
strates which will gain weight from sulfate uptake can be pre-
conditoned in the flue gas before weighing. Two to six hours
of exposure to the flue gas will suffice where mass loadings are
high and sample times are short. In the situations where sample
times are long and the collected amount of particulate matter
is small, it may be necessary to condition the substrates for
as long as twenty-four hours to eliminate significant sulfate
uptake and weight gains. Repeated weighings to check weight
gains are necessary to confirm that the substrates can be used.
Another approach is to use a fibrous substrate which shows little
weight gain in a sulfur dioxide stream, if one can be found.
It should be noted that the particle retention characteristics
of different fiber materials vary, and the impactor calibration
could change significantly if the substrate is changed. In the
report on substrate experiments (Felix, et al, 1977), a laboratory
preconditioning procedure is described in detail.
Greased substrates — "Grease" must often be used on metal
foil substrates to improve their particle retention character-
istics. This is particularly important with hard, bouncy parti-
cles. Impactor stage velocities of 60-65 m/sec have been used
on greased substrates with good results, while particle bounce
can become a problem at about half of that rate on ungreased
substrates.
Finding a suitable grease can be difficult. The grease
should not flow at operating temperature, and it must be essenti-
ally nonvolatile. Gas chromatographic materials such as poly-
ethylene glycol 600 have exhibited more consistent characteristics
than materials such as stopcock grease. Another class of materials
which may be suitable are high vacuum greases; Apiezon L and H
in particular have performed well at temperatures up to 120°C.
The greased substrates must be tested as blanks in filtered pro-
cess gas before they are used in the test program.
197
-------
The greases are normally applied as suspensions or solutions
of 10-20 percent grease in toluene or benzene. The mixture is
placed on the substrate with a brush or eyedropper, baked at 204°C
for 1 to 2 hours, and then desiccated for 12 to 24 hours prior
to weighing. It is important to avoid an excess of grease. The
desiccated, greased substrate should be tacky, but not slippery,
with a film thickness about equal to the diameter of the particles
which are to be captured.
Horizontal operation of the impactors with greased substrates
is not recommended due to possible flow of the grease. Care must
also be taken to ensure that grease is not blown off the substrates
(which tends to occur at jet velocities greater than 60 m/sec).
To some degree, grease blow-off can be avoided by using a light
coating of grease on the last stages. This is normally satis-
factory from an adhesive standpoint, as the last stages usually
have the lightest loading along with the highest jet velocity.
Inspection of the stage catches is the best way to check on this
problem.
C.3.2 Back Up Filters
Back up filters are used on all impactors to collect the ma-
terial that passes the last impaction stage. Binderless glass
fiber filter material is normally used for this purpose in all
the impactors, although the exact configuration varies.
Glass fiber back up filters have the same problems as do
glass fiber substrates. Their use in process gases containing
sulfur dioxide is suspect, and blanks must be run to check out
the problems. Pure Teflon filters may alleviate this problem
if they can be used.
198
-------
C.4 PREPARATION AND SAMPLING
C.4.1 Substrate Preparation
It is assumed that the substrates have been properly prepared
and that the necessary quality assurance steps have been taken.
The substrates should be carefully weighed and kept in a desic-
cator until they are to be placed in the impactor.
C.4.2 Impactor Orientation
Whenever possible, the impactor should be oriented vertically
to minimize gravitational effects such as flow of grease or fall-
off of collected particles. Sampling situations requiring hori-
zontal placement will occur, and extra care must be taken on such
occasions not to bump the impactor against the port during entry
or removal.
C.4.3 Heating the Impactor
All condensable vapors must be in a gaseous state until they
exit from the impactor, unless a condensate is the prime aerosol
being measured. In gas streams above 177°C, auxiliary heating
should not be required. Below 177°C the exit temperature of the
impactor should be maintained at least 10°C above the process
temperature if condensable vapors are present. A thermocouple
feedback temperature controller has proven useful.
When condensable vapors are present, it is sometimes neces-
sary to heat the impactor probe to prevent any condensate formed
in the probe from entering the impactor and contaminating the
substrates. Water vapor is the primary problem. The probe tem-
perature should be maintained above the vapor's dewpoint.
199
-------
Whether the impactor is being heated in the duct or externally
with heater tape, an allowance of 45 minutes warmup time is recom-
mended as a minimum to ensure that the impactor has been heated
to duct or operating temperature. Thermocouple monitoring of
the impactor temperature and gas temperature is recommended.
C.4.4 Probes
Sampling probes leading to an impactor outside of the duct
should be used only if there is no other way. They should be as
short as possible and contain the fewest possible bends. It is
recommended that a precutter be mounted at the duct end of the
probe to remove the large (>10 ym) particles and thus reduce line
losses.
C.4.5 Nozzle and Sampling Rate Selection
It, is preferable to use as large a nozzle diameter as pos-
sible to minimize sampling errors resulting from nozzle inlet
geometry. When very small nozzles have been used with the Brink
impactor, there have been some cases in which large amounts of
material were retained in the nozzle or the nozzle was completely
blocked. It is recommended that the inlet nozzle not be smaller
than 1.4 mm, and some types of particulate material may require
a larger minimum nozzle size. In some instances bent nozzles
are necessary due to port location and gas direction, but these
should be avoided. Problems occur in cleaning bent nozzles, and
it is difficult to determine the size interval in which the de-
posited material originated. If they cannot be avoided, bends
should be as smooth as possible and of minimum angle in order
to minimize the losses in the fine particle region.
For hard, "bouncy", particulate, the sampling rate must be
such that the last stage velocity does not exceed 60 m/sec for
greased collection surfaces or 35 m/sec for ungreased plates if
200
-------
no suitable substrate can be found to limit particle bounce. The
flow rates above should not be considered the final word on nozzle
velocity. Particle bounce has been observed at nozzle velocities
as low as 10 m/sec. Some particulate materials are "sticky" and
will adhere at well above the maximum velocity for hard particles.
The exposed substrate should be visually examined for evidence
of re-entrainment and the rates adjusted accordingly.
It is apparent that sample rate and nozzle size are closely
coupled. The requirements for isokinetic or near-isokinetic
nozzle flow sometimes impose a compromise on nozzle selection.
The general order of priorities when choosing the sample rate
is nozzle diameter (at least 1.4 mm), last stage jet velocity,
and flow rate required for isokinetic sampling. Selection of
nozzle diameter and impactor flow rate combinations for achieving
near-isokinetic sampling conditions can be made from Figure C3.
If a choice must be made between undersized and oversized
nozzles, undersized nozzles will usually result in lower sampling
errors than will oversize.
C.4.6 Use of Precollector Cyclones
In many instances the percentage (by weight) of material
with sizes larger than the first impaction stage cut point is
quite high. In such cases a precollector cyclone is necessary
to prevent the upper impactor stages from overloading. A pre-
cutter should always be used for the first test. If the weight
of material obtained by the precutter is greater than or equal
to that on the first stage, the precollector should be used in
all subsequent runs. Cyclones can be obtained from the impactor
manufacturer or can be shop made. The use of two first stages
in series has also been suggested and appears to be a valid ap-
proach; however, no data are available.
201
-------
l/min acfm
0.283-0.01
30.5
GAS VELOCITY
4 6 8 100
3050
3630-080
Figure C3. Nomograph for selecting nozzles for isokinetic samp/ing.
202
-------
C.4.7 Impactor Flow Rate Measurement
The flow rate through an impactor must be accurately mea-
sured in order to set the isokinetic sampling rate and to deter-
mine the correct impactor stage cut points. Unfortunately, it
is usually very inconvenient, and sometimes impossible, to mea-
sure the impactor flow rate at the conditions present in the im-
pactor. The gas is normally drier, cooler, and at a lower pres-
sure by the time the flow rate is measured; and the flow must
be corrected to impactor conditions. The use of calibrated ori-
fices and dry gas meters is discussed below.
Units — The equations presented are valid only if the units
of the various terms are consistent. For instance, the pressure
drop terms could be in units of mm H20 or cm H20 or something
else, but all pressure drop terms must have the same units. The
same is true for the other properties. Note that pressure and
temperature are both absolute measurements.
Orifice meters — The gas flow rate through a particulate
orifice meter is related to the pressure drop across that orifice
by an equation of the form:
* „ AP
Q = C — (Cl)
where Q = volumetric flow rate at upstream conditions,
AP = pressure drop across orifice,
p = density of gas at upstream conditions, and
C = dimensional constant, (length) 5 (mass) (time) ~2 (force) -1
Solving for the constant, C, in equation (Cl) , one obtains:
r
c
_ QP
~ ~
203
-------
As C is a constant at all conditions, its value can be ob-
tained at a convenient set of conditions with a known flow rate
and used later to calculate the actual sampling flow rate. Equa-
tion (C2) can be rewritten:
Q 2 P
c = _S £ (C2a)
c AP
c
The subscript "c" indicates that these parameters were de-
termined during a calibration. Density and flow rate are at
upstream conditions.
Substituting equation (C2a) into equation (Cl) yields an
equation suitable for obtaining flow rates from a calibrated
orifice:
Q.
2 =
AP
c
AP
m
Pm
(C3)
The subscript "m" denotes the parameters of the gas as it
is being "measured". All are at conditions immediately upstream
of the orifice.
For use with impactors, the measured flow rate, Q , must
be converted to a flow rate at stack conditions, Q . Assuming
s
that the stack gas was dried as well as altered in temperature
and pressure, the stack flow rate is related to the measured flow
rate by:
Qs (* - Bwo) 5* ' Qm £ 1C4)
s m
204
-------
where B = water removed from flue gas, expressed as a volu-
metric fraction,
P = absolute pressure, and
T = absolute temperature.
The subscript "s" refers to stack conditions.
At the usual conditions of relatively high temperature and
low pressure which occur during stack sampling, the flue gas
behaves very much like an ideal gas. The density of an ideal
gas can be approximated as:
P(MW) ,
p = " (C5)
where R = the universal gas constant, and
MW = the molecular weight of the gas.
Equations (C3) , (C4) , and C5) can be combined and rearranged
into a form which gives the pressure drop which must exist across
the calibrated orifice, P , to obtain the required impactor flow
rate, Q .
Apm ' 4Pc !K (wU'lsVII^H vfl (C6)
where (MW),,, = molecular weight of the stack gas at the orifice,
normally the dry molecular weight,
(MW) = molecular weight of the calibration gas.
c
Dry gas meter — The dry gas meter, like the orifice, can
only directly measure the flow rate of the gas which passes through
it. This measured flow rate can be converted to the flow rate
through the impactor (which is at stack conditions) using equation
(C4a):
205
-------
- B } (C4a)
_ leimll-rtll
m
C.4.8 Sampling Time
The length of the sampling time is dictated by mass loading
and the particle size distribution. An estimate for initial tests
can be obtained from Figure C4. Two conflicting criteria compli-
cate the choice of the sampling time. It is desirable from the
standpoint of minimizing weighing errors to collect several milli-
grams on each stage. However, most size distributions are such
that the upper stages are overloaded and are re-entraining par-
ticles by the time the lower stages reach a few milligrams. A
rule of thumb is that no stage should be loaded above 10 mg, but
the determining factor is whether or not re-entrainment occurs.
As is discussed later, a comparison of the relative distribution
determined by a long run with that from a shorter (about half
as long) run can be used to check on re-entrainment due to stage
overloading.
C.4.9 Readying the Impactor
As equipment is not always cleaned as well as it should be,
the impactor should be inspected prior to use. The nozzles must
be clean, gaskets in good shape, and the interior clean. Nozzles
can be cleaned with fine wire if necessary.
After inspection, the impactor should be carefully loaded
with the preweighed stage substrates and assembled. Teflon thread
sealant tape or antiseize compound should be applied to the threads,
especially when high temperatures (>215°C) are encountered. The
thread sealant tape generally works better and causes fewer prob-
lems but probably cannot be successfully used at temperatures
above 290°C.
206
-------
FLUE GAS MASS LOADING
I I I I MM
I II Mill
I I I IIIII-
READ DOWN FROM MASS LOADING TO IMPACTOR
SAMPLE RATE. READ LEFT TO TIME REQUIRED
TO COLLECT A 50 MG SAMPLE AT THAT SAMPLE —
RATE. _
\SAMPLE '
= LOAD NG x RATE x TIME
1.0 0.6 0.2 0.1 0.04 0.02 0.01
28.3 2.83 0.283
IMPACTOR SAMPLE RATE
3630-079
Figure C4. Nomograph for sampling time selection (50 mg sample).
207
-------
If supplemental heating is required, a heating device and
temperature monitor need to be added. A thermocouple mounted
in the gas flow immediately after the impactor is best for control-
ling impactor cut points.
The supplemental heat can be supplied with either a heating
mantle which has been made to fit the impactor or by using heating
tapes. If the tapes are to be used, a heating tape of sufficient
wattage is wrapped around the impactor. Glass fiber tape works
well for holding the heating tape. Insulation such as asbestos
tape is then wound around the impactor. Glass fiber tape again
is used to hold the asbestos in place and also acts as additional
insulation. The impactor can now be mounted on the appropriate
probe, taken to the sampling position, and installed in the sampl-
ing system.
C.4.10 Pre-Sample Checks
Impactors are prone to leak, and they must be checked at
operating temperature for leaks. This can be done in several
ways. The nozzle can be plugged and the impactor pressure-tested
or vacuum-tested. Because impactors are basically a series of
orifices, they should have a constant flow to pressure drop re-
lationship. Checking the pressure drop on various flows of fil-
tered air will point out deviations from normal operations--both
leaks (external or internal) and plugged jets.
C.4.11 Taking the Sample
The impactor should be preheated for at least 45 minutes
before sampling. If supplemental heat is being used, the impactor
should be brought up to temperature outside the duct and then
allowed some time to equilibrate after insertion. The nozzle
should not point into the flow field during this phase. Without
208
-------
supplemental heat, the whole warm-up is conducted within the duct,
again with the nozzle pointed away from the flow field.
A predetermined flow rate must be maintained to ensure stable
cut points. Any attempt to modulate flow to provide isokinetic
sampling will destroy the utility of the data by changing the
cut points of the individual stages. Rapid establishment of the
correct flow rate is especially important for the short sampling
times typically found at the inlets to control devices. Capping
the nozzle during preheat in the flue is also desirable.
C.4.12 Number of Sample Points
As the velocity and particulate distributions in industrial
ductwork are unlikely to be ideal, a large number of samples are
often required for accurate particulate measurements. A velocity
traverse should be run to check on the velocity distribution.
At least two points within a duct should be sampled in each mea-
surement plane, and at least two samples taken at each of these
points. These are the minimum sampling efforts and are appro-
priate only for locations with well developed flow profiles in
the absence of significant concentration stratification. If the
flow profile at the inlet or outlet is uncertain due to duct con-
figuration and/or the mass loading is not uniform, the number
of samples may need to be increased for reliable results.
C.5 SAMPLE RETRIEVAL AND WEIGHING
C.5.1 Impactor Clean-up
The careful disassembly of the impactor and removal of the
collected particulate are essential to the success of the test
program. The crucial points are to make sure that the collected
material stays where it originally impacted and to remove all
209
-------
the particulate. After the sampling run, the impactor should
be carefully removed from the duct without jarring it, removed
from the probe, and allowed to cool. Disassembly can be difficult
in some cases, particularly if the impactor was used at elevated
temperatures.
Typically, not all of the particulate which collects in an
impactor collects on the substrates. Some accumulates on the
interior surfaces, especially in the nozzle. By convention, all
of the particles collected upstream of a given impaction stage
are assigned to that stage.
The collection of these "misdirected" particles is often
troublesome. If the particles are hard and dry, they can be
brushed off into the weighing container. A No. 7 Portrait brush
or its equivalent is suggested, and care must be taken to prevent
brush hairs from contaminating the sample. If the particles are
sticky or wet, some type of washdown procedure should be used.
The solvent must be considerably more volatile then the parti-
culate matter.
C.5.2 Drying and Weighing
All of the samples must be dried to constant weight, with
2 hour checks used to establish the uniformity of the weights.
Hard, nonvolatile particles are often dried in a convection oven
to 100°C, desiccated until cooled to room temperature, weighed,
then check weighed. Volatile materials will require some other
technique using low temperature. Whatever the technique used,
constant weight of the sample with further drying is the criteria
to be met.
210
-------
C.5.3 Data Logging
Permanent records should be kept of all pertinent information.
It is generally necessary to keep records in three places—in
the lab with the balance using a bound notebook, and using either
looseleaf data forms of a bound notebook at both the inlet and
outlet of the control device. Table C2 presents a fairly complete
listing of the information required concerning an impactor run-
In addition, records of the weighing of the catches must be kept.
Notes should be taken on any abnormalities which occur and on
the apparent condition of the stage catches.
C.6 QUALITY ASSURANCE
The field use of cascade impactors is a difficult task.
The accuracy required is more appropriate for a laboratory program
than for a field test. There are many places in the operational
sequence where errors can occur in spite of a conscientious effort
to do a good job. Quality assurance attempts to discover inac-
curacies before they are propagated throughout the test program.
The techniques presented in this section are not the only ways
to ensure quality data. However, they have been used successfully
in field testing with impactors.
C.6.1 Impactor Techniques
Glass fiber substrates — As has been discussed previously,
glass fiber substrates are not without problems. Two potentially
serious problems are S02 uptake on the substrate and mechanical
or manual abrasion of the filter mat.
211
-------
TABLE C2. SAMPLING INFORMATION REQUIRED
Date
Time
Run Code Number
Impactor Type and Identification Number
Operator
Port Number/Sampling Location
Ambient Temperature
Ambient Pressure
Impactor In-Stack or Out-of-Stack
Impactor Orientation
Number of Traverse Points
Stack Pressure
Stack Temperature
Nozzle Diameter/Type
Probe Depth, if used
Stack Pitot Tube Delta P/Stack Gas Velocity
Desired Impactor Flow Rate for Isokinetic Sampling
Metering Orifice Identification Number
Metering Orifice Delta P
Impactor Temperature
Scalping Cyclone in Use? Identification
Prefilter Identification
Postfilter Identification
Substrate Set Identification
Pressure Drop Across Impactor
Test Start/End Time: Duration of Test
Gas Meter Start/End Readings? Gas Meter Volume
Agreement Between Meter and Orifice
Volume of Condensible H20 in Flue Gas
Gas Meter Temperature
212
-------
The problem of S02 uptake on the substrate is discussed by
Felix et al (1977). The two approaches available are to use a
substrate which does not change weight in the flue gas or to pre-
condition the substrate in filtered flue gas prior to weighing.
Using a new glass fiber material, which does not react with the
S02, may alter the particle retention characteristics of the im-
pactor and change the impactor's calibration. This must be checked
and the data reported. The use of preconditioned filter mats
requires that the glass fiber substrates be preconditioned long
enough to reduce the weight change during the expected duration
of the impactor runs to 10 percent or less of the minimum stage
weight. At the present time, this S02 reaction phenomenon is not
well understood, and only rough guidelines are available. For
some common glass fiber materials tested, the saturation times
were on the order of 2 to 6 hours at the temperatures tested.
The applicability of the method chosen to overcome this sub-
strate problem must be tested during the presurvey and periodi-
cally during the test runs by running blanks.
Glass fiber substrates must be handled carefully to prevent
damage and possible loss of fibers. Loose surface fibers should
be removed by shaking prior to initial weighing. After weighing,
every precaution must be taken to prevent the loss of any part
of the substrate. One approach which will quantify the problem
of substrate abrasion is to prepare a substrate set, load the
impactor, then disassemble and reweigh.
Greased metal substrates — The problems which occur with
the use of greased substrates are usually related to the properties
of the grease. A grease which has been applied too heavily or
has a low viscosity at operating temperature can be physically
blown off the impactor stage. The grease could also react chemi-
cally with the flue gas or be excessively volatile at the operating
213
-------
temperature. Again, these phenomenon must be checked during the
presurvey and periodically during the test program.
Re-entrainment — Re-entrainment is the phenomenon of an
impacted particle being blown off the stage on which it was col-
lected downstream. This can be caused by excessive jet velocities
or by overloaded stages. The effect of re-entrainment can oe
serious, because only a few large particles on a small particle
stage will considerably affect the size distribution.
One way to spot re-entrainment is to very carefully examine
the stage catches. If, for example, a low velocity through the
jets resulted in a well-defined pile of particulate and a high
velocity sample gave a diffuse deposit, re-entrainment should
be suspected at the high sampling rate. Microscopic examination
of the lower stages and final filter for large particles (which
should have been collected upstream) is another way to check for
re-entrainment.
Re-entrainment due to stage overloading can be detected by
running two otherwise identical tests for two different test dura-
tions. If the two size distributions are not the same, overload-
ing should be suspected at the higher stage loadings.
Impactor leaks — Two types of leaks can occur with impac-
tors—internal or external. A flow rate versus pressure drop
check of a pressure test will pick up most leaks. An internal
leak, where part of the airstream is bypassing the proper flow
path, will give results similar to re-entrainment. Leak checks
must be made at operating temperature.
General procedure — A general procedure for impactor use,
concentrating on quality assurance, has been outlined below:
214
-------
1. Prepare Impactor
a. Wash impactor. Use ultrasonic cleaner if available.
b. Visually check cleanliness. Jets must be clear,
sidewalls clean. Must be done in good lighting.
c. Obtain preweighed substrates and assemble impactor.
2. Sampling
a. Assemble impactor train and heat to operating tem-
perature.
b. Leak check the impactor.
c. Sample with impactor.
d. Disassemble impactor, examine stage catches and
impactor walls. Note any anomalies.
3. Substrate and Re-entrainment Checks
a. Check during presurvey-
b. Check substrates if flue gas composition changes
significantly.
C.6.2 Weighing Techniques
Precision and calibration — The manufacturer's directions
should be followed when operating the balance. The balance should
be calibrated at least once a day. The repeatability of measure-
ments should be checked by repeatedly weighing a substrate and
a test weight.
Technique — The assembly and disassembly of an impactor
should have no effect on the substrate weights. This should be
checked by weighing up a set of substrates, assembling them in
an impactor, then disassembling and reweighing. Any weight losses
from this process should be within the repeatability of the balance
215
-------
(approximately 0.02 milligram for an electrobalance). Dry weight
checks are made by desiccating the substrate, weighing, then
desiccating again and reweighing. When the agreement is within
the repeatability of the balance, dry weight has been achieved.
C.6.3 General Notes
Spare parts — The well equipped sampling team will travel
with an adequate supply of spares. Improvisation due to an equip-
ment failure can lead to poor quality data.
Flow meters — At least two flow meters should be used in
series. If they do not agree, the problem should be investigated.
Pumps — Typically, vacuum pumps in sampling trains leak.
For this reason, the flow meters should be upstream of the pump.
C.6.4 Data Analysis
Final filter data — The fine particulate information ob-
tained from the final filter can sometimes be misleading. It
is assumed for analysis that a stage captures everything larger
than its D50/ and captures nothing smaller. A real stage misses
some large particles. Under some conditions (including but not
limited to re-entrainment), large particles will penetrate to
the final filter. In this case the size distribution will be
skewed towards the small particles. Microscopic examination of
the final filter may provide an indication of this problem. If
it occurs, the best choice in data analysis is probably to ignore
the final filter on runs where this phenomenon was encountered.
Cumulative size data analysis — If either a probe or a pre-
cutter cyclone is used with an impactor, the resultant probe losses
and precutter catches must be included in cumulative size analysis.
Failure to do so will lead to an incorrect cumulative distribution.
216
-------
Inspection of data — After the data have been collected,
they should be examined for any inconsistencies and outliers
should be rejected before the final averaging is done.
C.7 USE OF COMMERCIAL IMPACTORS
C.7.1 Brink BMS-11 Cascade Impactor
Additional information on the following impactors can be
found in Section 2.3.2 under Table I.
The Brink impactor is a five-stage, low sample rate, cascade
impactor, suitable for measurements in high mass loading situa-
tions. The Brink uses a single round jet on each of its stages.
Sampling rate — The usual sampling rates for the Brink
are in the range of 9-33 cm3/sec (0.02 to 0.07 acfm). The sampl-
ing rate must be low enough to prevent re-entrainment of particles
from the lower stages. With hard, bouncy particulate, the last
stage nozzle velocity must be less than 30-35 m/sec with ungreased
collection plates and less than 60 m/sec with greased collection
plates.
Collection substrates and adhesives — The Brink impactor
collection stage is too heavy to use without some type of sub-
strate insert. Foil cups are commonly preformed and fitted into
the collection cups of the Brink stages. If grease is to be used,
the top stages require about 5 or 6 drops of solution while the
bottom stages normally require only about one drop in the center
of the cup. Glass fiber substrates cut to fit the collection
cups have also been found satisfactory in many situations.
Back up filter — The Brink back up filter is normally made
of binderless glass fiber filter material. Two 2.5 cm diameter
disks of filter material are placed under the spring in the last
217
-------
stage of the impactor. The filter is protected by a Teflon O-
ring and the second filter disk acts as a support.
Precutter cyclone — A precutter cyclone for the Brink is
not commercially available.
Sampling train — The Brink uses the usual type of sampling
train. Orifices on the order of 0.77, 1.52, and 2.29 mm in diam-
eter allow full coverage of its range of sampling rates at rea-
sonable pressure drops.
Brink clean-up — Careful disassembly of a Brink impactor
is necessary for obtaining good stage weights. If a precollector
cyclone has been used, all material from the nozzle to the outlet
of the cyclone is included with the cyclone catch. All of this
material should be brushed onto a small, tared, 2.5 x 2.5 cm alu-
minum foil square to be saved for weighing. Cleaning the nozzle
is also important, especially if it is a small bore nozzle. All
material between the cyclone outlet and the second stage nozzle
is included with material collected on the first collection sub-
strate. All appropriate walls should be brushed off, as well
as around the underside of the nozzle, where as much as 30 percent
of the sample has been found.
C.I.2 Amdersen Mark III Stack Sampler^
The Andersen impactor is a relatively high sample rate impac-
tor. The normal sample rate is about 236 cm /sec (0.5 acfm).
The Andersen is a multiple jet, round hole impactor.
Sampling rate — The nominal Andersen sampling rate is given
above. As with other cascade impactors, the flow rate must be
low enough to prevent re-entrainment of impacted dust.
218
-------
Collection substrates and adhesives — Andersen substrates
are obtained precut from the manufacturer. The substrates are
glass fiber and of two types—one cut for the odd numbered stages,
one for the even. As discussed earlier, normal Andersen substrates
have a tendency to absorb S02 on basic sites in the substrate
and therefore gain weight.
The Andersen requires careful assembly, as overtightening
will cause the substrates to stick to the metal separator rings.
Back up filter — The Andersen uses a 63.5 mm diameter disk
placed above the final F-stage. (This F-stage is an option not
normally included with the standard stack head.) The filter
should be cut from binderless glass fiber filter material, such
as Reeve-Angel 934AH filter paper.
Precutter cyclone — A precutter cyclone for the Andersen
is available from the manufacturer. It is necessary to have a
6-inch diameter or larger sampling port when using the precutter
cyclone with its nozzle.
Andersen sampling train — The Andersen requires the usual
type of sample train. The pumping and metering systems of the
commercial EPA Method 5 mass sampling trains are appropriately
sized for use with the Andersen.
Care should be exercised never to allow a gas flow reversal
to occur through the impactor. Material could be blown off the
collection substrate onto the underside of the jet plate or the
collection substrates could be disturbed. A check valve or main-
tenance of a very low flow while removing the impactor from the
duct avoids this problem.
219
-------
Andersen clean-up — Cleaning an Andersen impactor is dif-
ficult. Foils should be cut to hold the substrates, and each
foil and substrate weighed together before and after the run.
For disassembly, the foil to hold the stage 1 substrate should
be laid out. Next the nozzle and entrance cone should be brushed
out and onto the foil. Then the material on stage 0 should be
brushed onto the foil. The stage 1 filter substrate material
should then be placed on the foil and, finally, the top of the
stage 1 plate, 0-ring, and cross piece should be brushed off.
Depending on how tightly the impactor was assembled, some filter
material may stick to the 0-ring edge contacting the substrate.
This should be carefully brushed onto the appropriate foil. This
process is continued through the lower stages. Finally, the after
filter is carefully removed.
C.7.3 University of Washington Mark III (Pilat) Impactor
The Mark III impactor is a seven-stage, high flow rate de-
vice with generally the same characteristics as the Andersen.
The Mark III is a round hole, multiple jet impactor.
Sample rate — The Mark III sampling rate is on the order
of 236 cm3/sec (0.5 acfm). The flow rate must be low enough to
keep scouring of impacted particles to a minimum.
Collection substrates and adhesives — The Mark III has often
been used with supplementary foil (aluminum or stainless steel)
substrates. These substrates require the use of grease for easily
re-entrained particles. Enough of the grease solution is placed
evenly on the substrate to adequately cover the area under the
jets. The normal cautions on the use of greased substrates apply
as discussed in the text.
220
-------
Precutter cyclone — A BCURA (British Coal Utilization Re-
search Association) designed precutter cyclone is available from
the manufacturer.
Mark III sampling train -- As the Mark III is a high flow
rate device, its sampling train is similar to that of the Ander-
sen.
Mark III clean-up — Mark III impactor clean-up is similar
to that for the Brink. Some problems have been noted with 0-rings
sticking rather tenaciously and care must be exercised not to
dislodge the sample while trying to separate the stages.
C.7.4 Meteorology Research, Inc. (MRI) Model 1502 Cascade Impactor
The MRI impactor is a high flow rate sampler. The body of
the instrument is constructed from quick-disconnect rings which
allow flexibility in configuration of the impactor and a posi-
tive gas seal between stages. The impactor uses multiple round
jets in its stages.
Sampling rate — The sampling rate is nominally 236 cm3/sec
(0.5 acfm) in the seven-stage configuration. Higher flow rates
have been used by removing the last stages.
Collection substrates and adhesives — The MRI collection
disc is a self-supporting foil (316 stainless steel) which is
functionally similar to the collection cup or tray and inserts
used in other impactors. The collection discs are mass produced
and normally are used only once and discarded.
Grease applied as described earlier is recommended for most
applications.
221
-------
Back-up filter — The MRI impactor has a built-in filter
holder for 47 nun diameter filters. Normally, binderless glass
filters are used. Filter losses can be prevented by placing tared
Teflon washers on both sides of the filter during the test.
MRI sampling train — The MRI sampling train is similar to
that of the Andersen.
MRI clean-up — The clean up of the MRI impactor is similar
to the Brink. The device is clamped in a vice and all of the
sections and nozzles are loosened with wrenches. The wall losses
are carefully brushed onto the appropriate collection disc. Care
is taken not to brush contamination from the threads into the
sample. A tared foil dish is used to hold the back up filter.
Any worn 0-rings should be replaced and the whole unit carefully
cleaned before the next test.
C.7.5 Sierra Model 226 Source Cascade Impactor
The Sierra impactor is a six-stage, high sample rate cascade
impactor. The Sierra instrument uses a radial-slot design.
Sampling rate — The Sierra impactor has a nominal sampling
rate 236 cm3/sec (0.5 acfm). The flow rate must be low enough
to prevent re-entrainment of particles. Laboratory research
(Gushing, 1976) has shown that a flow rate of 118 cm3/sec (0.25
acfm) gives better stage collection efficiency characteristics
for this impactor.
Collection substrates and adhesives — Substrates for the
Sierra are obtained precut from the manufacturer. These are glass
fiber substrates and should be checked for weight gain. Stain-
less steel substrates are also available and these should normally
be coated with grease as described earlier.
222
-------
Back-up filter — The back-up filter uses a 47 mm glass fiber
filter mat. It is supported by a screen from below.
Precutter cyclone — A precutter cyclone is available from
the manufacturer.
Sampling train — The sampling train for the Sierra is similar
to that of the Andersen.
Clean-up — Clean-up of the Sierra is fairly similar to
Andersen clean-up. Care should be taken to be sure the glass
fiber substrates are removed intact.
C.8 DATA ANALYSIS
C.8.1 General Discussion
The majority of impactor data reduction is done using a
technique referred to as the "D50" method. In this method it
is assumed that all of the particles caught by an impactor stage
consist of particles having diameters equal to or greater than
the D50 of that stage, but less than the diameter of the stage
above. The mass caught by the first stage or cyclone is considered
to be equal to or greater than the D5o of that stage, but smaller
than the largest particle diameter present in the aerosol. The
largest diameter may be determined, approximately by microscopic
examination of the stage catch, or it may be assigned some reason-
able arbitrary value, say 100 ym.
Particle size distributions may be presented on a differ-
ential or a cumulative basis. When using the D50 method, either
type of presentation may be easily employed.
The size parameter reported can be aerodynamic diameter,
aerodynamic impaction diameter, or Stokes diameter. In all cases,
223
-------
the particles are assumed to be spherical. The method of report-
ing diameters depends to a large extent upon the ultimate use
of the size distribution information. For this reason it is
suggested that the data be reported in three parallel sets: one
set based on aerodynamic impaction diameter, one based on aerody-
namic diameter, and one based on the Stokes diameter.
C.8.2 Calculation of Impactor Stage DSO'S
The reduction of field data obtained with a cascade impactor
can sometimes be troublesome and time consuming because of the
computations involved. The equations below are based on the
motion of particles in the Stokes regime for which the Reynolds
number is less than 1000. Although this is not always true for
impactor s, the equations are often a good approximation. The
basic equation that defines the theoretical impaction behavior
of a given stage of a cascade impactor is
P d 2V.C
where ty. is the inertial impaction parameter,
p is the particle density, gm/cm3 ,
V. is the velocity of the gas (and particles) in the im
pactor jet (cm/sec) ,
y is the gas viscosity, poise,
D. is the diameter (width for slots) of the jet, cm,
C is the slip correction factor, and
d is the particle diameter, cm.
If the value of \|> for 50 percent collection, ty50, can be
determined, equation C7 can be inverted to give the stage D50
for a wide range of test conditions. Historically, the experi
mental values reported by Ranz and Wong (1952) have been used.
These are:
224
-------
For round jets, ^50 = 0.145, and
For rectangular jets, i|>50 = 0.44.
Subsequent studies, however, have shown that there is no
universal value for i\>so and the actual value must be determined
by calibration for each impactor design. Several papers and re-
ports are available which tabulate stage constants of different
impactors, and outline procedures for impactor calibration, Harris
(1977) and Gushing (1976).
From equation C7,
(C8)
As equation (C8) is written, with the actual particle density
and the calculated slip correction factor, it defines the Stokes
diameter. If the particles are treated as if their density, p ,
was 1.0, equation (C8) defines the aerodynamic diameter. If the
slip correction factor is also assumed to be equal to 1.0, the
aerodynamic impaction diameter is defined by equation (C8) . See
Section A. 3. 6.
Since C, the slip correction factor, contains D, this equa-
tion must be solved by iteration where D50 and C are calculated
alternately-
Equation C8 may be written more conveniently in terms of
the test parameters: For round jet impactors
CPPQSPS
yD.
~ J J i
(C9)
225
-------
where X. is the number of jets on the stage,
P. is the absolute pressure downstream of the jet(s),
mm Hg,
p is the absolute pressure in the stack, mm Hg, and
s 3
Q is the sample flow rate, cm /sec.
S
and for rectangular jet impactors
18ip50 yW 2L P
D J J _ _ (CIO)
"
where W. is the jet width, cm, and
L. is the total jet length, cm.
One approach that can be used to simplify the computations
is to develop curves for the impactor stage cut points at one
set of conditions—e.g., air at standard conditions and a par-
ticle density of 1.0. Then a suitable correction factor can be
applied to these curves for the actual sampling conditions. Un-
fortunately, further steps are involved in making the correction
factor simple enough to be of value. Therefore, the use of this
type of approach suffers from some restrictions.
All of the assumptions and calculations involved in going
from equations C9 to CIO to the calibration curve can be quite
awkward, particularly in cases where different types of sources
are being sampled. The best and easiest approach is to write
or obtain a computer program based on the rigorous equations given
initially. Such a program can calculate impactor stage cut points,
compute concentrations of particles in each size range, determine
the baghouse efficiency, and plot graphs.
226
-------
A sophisticated computer data reduction program is available
from the EPA (Johnson, et. al., 1978) and also less powerful pro-
grams are available for the Hewlett Packard HP-65 (Ragland, 1976)
and HP-25 (Ragland, 1977) programmable calculators.
C.8.3 Cumulative Particle Size Distributions
Impactor data may be presented on a cumulative basis by sum-
ming the mass on all the collection stages and back up filter,
and plotting the fraction of the mass below a given size versus
size. This is frequently done on special log-probability paper.
Semi-log paper may be preferable for distributions that are not
log-normal.
Cumulative distributions are very easy to understand and
present the data with clarity. Cumulative distributions, however,
have a couple of disadvantages when compared to differential dis-
tributions. An error in stage weight will be propagated through-
out a cumulative analysis, but will be isolated by the differen-
tial approach. Also the differential method does not involve
the use of total mass concentration or total size distribution
from diameters of zero to infinity, and so is useful in comparing
instruments with overlapping but different size fractionation
ranges and different stage cut points while cumulative analysis
is not.
When cumulative plots are used, the abscissa is normally
the logarithm of the particle diameter and the ordinate is the
weight percent smaller than this size. The value of the ordinate
at a given (Dso)k would be
227
-------
k-1
Weight percent smaller than (D50)k = -^ x 100%, (Cll)
1=0
where i = o corresponds to the filter,
i = k corresponds to the stage under study, and
i = K corresponds to the coarsest jet or cyclone.
This equation requires that the stages be counted from the
final filter up. There is no (D50) , as the "o" corresponds to
the filter. (D50)-, is the cut point of the last stage, which
collects mass, AM-,.
C.8.4 Differential Size Distributions
Differential particle size distributions are used to plot
the relative concentration versus particle diameter. The area
under the frequency curve, between two designated diameters, is
equal to the mass of particles in that size range. Differential
curves may be obtained directly from the reduced impactor data,
or by differentiation of curves fitted to the cumulative particle
size distributions.
Many cascade impactors are designed so that the relationship
between successive stage D50's is logarithmic. For this reason,
and to minimize graphical scaling problems, the differential par-
ticle size distributions are plotted on log-log or semi-log paper
with AM/A(logD) as the ordinate and geometric mean of A(logD)
as the abscissa. The mass on stage "n" is designated by M and
is, in approximation, the particulate mass with diameters between
228
-------
(D50)n and (D50) .. The A(logD) associated with AMR is
l°g(Dso)n+1 - log(D50) . Using these approximations, the de
rivative term (ordinate) associated with stage "n" is:
n _ Mass on Stage "n"
log(D50)n+1 - log(D50)n
and the abscissa, D , is
[<»..>„+! * <»..)„]"
-------
UCL90 = Average + -^=r = Average + C.I.90 (C14)
VN
LCL90 = Average - t9°a = Average - C.I.go (CIS)
where Average = The average of the N values for AM/AlogD or
AN/AlogD at a particular particle size.
tgo = The Student's t distribution for 90% confidence
limits with N sets of data.
a = The Standard Deviation of the N sets of data.
N = The number of data values in the average.
The average cumulative mass and cumulative per cent graphs
are obtained from AM/AlogD data by piecewise integration, after
discarding outliers. The confidence interval of the cumulative
graph, at a particular size, is equal to the square root of the
sum of the squares of the confidence intervals of each size incre-
ment of the differential graph less than or equal to that size.
Clearly, the average values are more reliable and the con-
fidence limits are small when the number of data points is large.
A determination of fabric filter fractional penetration is
made by taking the ratio of the average AM/AlogD outlet values
to the average AM/AlogD inlet values at a series of particle
sizes. The confidence limits of the calculated penetration are
given by:
UCL90 = Average Penetration + CIP90 = P + CIP90 (C16)
and
LCL90 = Average Penetration - CIP90 = P - CIP90 (C17)
230
-------
where:
CIP
9 0
(C.I.go, Outlet)2 + p2 (C.I.9n, Inlet)
(Inlet Average) (Inlet Average)
(CIS)
and
p _ Outlet Average
Inlet Average
C.8.6 Cascade Impactor Data Reduction—Sample Calculation
This section contains a detailed description of the calcula-
tions that are required to derive particle size distributions
and a fabric filter fractional efficiency curve from raw impactor
data. The specific example given is for a single hypothetical
test performed with an Andersen impactor. The calculation pro-
cedures outlined here can be used with any impactor, however.
Normally the results of all tests made under the same process
and baghouse operating conditions are grouped and averaged, and
confidence limits calculated as described in Section C.8.5.
This discussion is based on Table C3 which was generated
by a computer program (Johnson, et. al., 1978). All of the cal-
culations, however, can be done on programmable calculators.
In the example shown, the data is reduced using a particle density
of 1.35 gm/cm3; thus the diameters reported are Stokes diameters.
For aerodynamic, or aerodynamic impaction diameters, p or p and
C , are set to unity, respectively- (p is particle density and
C is the slip correction factor.)
Information obtained from the data log sheets for each test
is printed at the top of Table C3. The maximum particle diameter
is measured by examining the particles collected on the first
stage (or first cyclone) with an optical microscope. Gas analysis
231
-------
TABLE C3
SAMPLE CALCULATION - INPUT DATA AND RESULTS
to
(jO
to
Hypothetical Andersen
Impactoc Flowrate = 0.500 ACFM
Impactor Pressure Drop =1.5 In. of Hg
Assumed Particle Density = 1.35 gm/cu.cm.
Gas Composition (Percent) C02
Calc. Mass Loading = 8.0711E-03 gr/acf
Impactor Stage
Stage Index Number
D50 (Micrometers)
Mass (Milligrams)
MG/DSCM/STAGE
Cum. Percent Of Mass Smaller Than
(MG/ACM) Smaller Than D50
(GR/ACF) Smaller Than D50
(GR/DSCF) Smaller Than D50
Mean Dia. (Micrometers)
Impactor Temperature = 400.0 F = 204.4 C
Stack Temperature = 400.0 F = 204.4 C
Stack Pressure = 26.50 In. of Hg
0.95 CO = 0.00 N2 = 76.53
Sampling Duration = 20.00 Min
Max. Particle Diameter = 100.0 Micrometers
02 = 20.53 H20 =1.00
1.4948E -02 gr/rfscf
1. .8470E+01 mg/acm
3.4207E+01 mg/dscm
Cum
Cum
Cum
Geo
DM/DLOGD (MG/DSCM)
DN/DLOGD (NO. PARTICLES/DSCM)
SI
1
10.74
0.72
4.71E+00
D50 86.24
1.59E+01
6.96E-03
1.29E-02
3.28E+01
4.86E+00
1.95E+05
S2
2
9.95
0.50
2.62E+00
78.59
1.45E+01
6.34E-03
1.17E-02
1.03E+01
7.94E+01
1.02E+08
S3
3
6.36
0.53
3.47E+00
68.46
1..26E+01
5.53E-03
1..02E-02
7.96E+00
1.79E+01
5.01E+07
S4
4
4.19
0.09
5.89E-01
66.74
1.23E+01
5.39E-03
9.98E-03
5.17E+00
3.25E+00
3.33E+07
35
•}
2.22
0.38
2.49E+00
59.47
1.10E+01
4.80E-03
8.89E-03
3.05E+00
8.99E+00
4.48E+08
86
6
1 .29
1.43
9.35E+00
32.13
5.93E+00
2.59E-03
4.80E-03
1.69E+00
3.99E4-01
1.16E-HO
S7
7
0.69
1 .25
8.18E+00
8.23
1.52E+00
6.64E-04
1..23E-03
9.43E-01
2.98E+01
5.03E+10
S3
8
0.33
0.04
2.62E+01
7.46
1.38E+00
6.02E-04
1.12E-03
4.74E-01
8.09E-01
1.08E+10
FILTER
9
0.39
2.55E+00
2.31E-01
8.47E+00
9.74E+11
Normal or standard conditions are 21°C and 760 mm Hg
-------
samples are taken at the same time the impactor is run. The mass
loading is calculated from the total mass of the particles col-
lected by the impactor, and listed in four different systems of
units after the heading CALC. MASS LOADING. The symbols are
defined as:
GR/ACF - grains per actual cubic foot of gas at stack conditions
of temperature, pressure, and water content.
GR/DSCF - grains per dry standard cubic foot of gas at engineer-
ing standard conditions of the gas. Engineering stan-
dard conditions are defined as 0% water content, 70°F,
and 29.92 inches of Hg.
MG/ACM - milligrams per actual cubic meter of gas at stack con-
ditions of temperature, pressure, and water content.
MG/DSCM - milligrams per dry standard cubic meter of gas at en-
gineering standard conditions of the gas. Engineering
standard conditions are defined as 0% water content,
21°C, and 760 mm of Hg (Torr).
The conditions at which the impactor was run determine stage
D50 cut points. These are calculated by iterative solution of
the following equations:
'so
-1
=
14.1 yD 3\bso P X.
°i i Si 1
2 X.
i
D50 x iO'"
1.23 + 0.41
D
5 0
A .
(C19)
(C20)
where D50i
ci
stage (i) cut point (cm),
gas viscosity (poise)
stage jet diameter (cm),
233
-------
P = local pressure at stage jet (mm Hg),
si
p = particle density (gm/cm3),
Q = impactor flow rate (cm3/sec),
P0 = ambient pressure at impactor inlet (mm Hg),
C. = Slip Correction Factor,
X. = gas mean free path (cm),
x. = number of holes per stage (i), and
0 = inertial impaction parameter of stage (i) .
To find the pressure P at each impactor stage, the follow-
si
ing equation is used:
P = P - (F.\ (DP), (C21)
0
- (F.\
where P0 is the ambient pressure at the impactor inlet,
F. is the fraction of the total impactor pressure drop at
each stage, and
DP is the total pressure drop across the impactor.
The total pressure drop DP across the impactor is given by
the following equation.
DP = K Q2 p MM, (C22)
where K = Empirically determined constant for each impactor,
Q = Flow rate through impactor (cm3/sec) ,
p = Gas density (gm/cm ) , and
MM = Mean Molecular Weight of gas (gm/gm-mole) .
To calculate the gas mean free path, X., for each impactor
stage, the following equation is used:
234
-------
;.3i x 107 T
^ = ^-M •%/ K
i 1.013 x 106 P x f 3 MM
Si
where y is the gas viscosity (poise) ,
P is the pressure at each impactor stage (atm),
si
TR is the gas temperature at the impactor stage (°K), and
MM is the average flue gas molecular weight.
To find the viscosity of the flue gas, y, the viscosity of
the pure gas components of the flue gas must first be found.
Viscosity is a function of temperature, and the temperature dif-
ference in different flue gases can be quite significant. The
following equations (derived from curves fitted to viscosity data
from the Handbook of Chemistry and Physics, Chemical Rubber Comp-
any Publisher, 54 Edition, 1973-1974, pp. F52-55), are used to
find the viscosities of C02(\i1) , C0(y2), N2(M3), 02(y^) and
H20(y5) .
MX = 138.494 + 0.499T - 0.267 x 10~3T2 + 0.972 x 10~7T3
M2 = 165.763 + 0.442T - 0.213 x 10"3T2
U3 = 167.086 + 0.417T + 0.417 - 0.139 x 10~3T2
M,, = 190.187 + 0.558T - 0.336 x 10~V + 0.139 x lO'V
ys = 87.800 + 0.374T + 0.238 x lO'^T2
where T is the temperature of the flue gas in degrees Celsius.
The units of y are 10~6 g/cm-sec. Next, these values of yx through
ys are used in a general viscosity equation for a mixture of any
number of components (See "A Viscosity Equation for Gas Mixtures"
by C. R. Wilke, Journal of Chemical Physics, Volume 8, Number
4, April 1950, page 517) used to find the viscosity of the flue
gas:
235
-------
n
- E
±
ED=n -i
• + JI 5 Yd
where <|>. . is given by the equation
(C25)
13 V 2
[l + (M./M.
and M = molecular weight of. a component in the mixture,
X = mole fraction of a component in the mixture,
y = viscosity, gm/cm-sec; y: , y2/ etc. refer to the pure
components at the temperature and pressure of the
mixture, y is the viscosity of the mixture, and
4> = dimensionless constant defined above.
Below these data the information pertinent to each stage
is summarized in columnar form in order of decreasing particle
size from left to right. Thus SI is the first stage, S8 is the
last stage, and FILTER is the back-up filter. If a cyclone was
used, then to the left of SI a column labelled CYC would appear
and information relevant to the cyclone would be listed in this
column. Beneath each impactor stage number is listed the cor-
responding stage index numbers, which also serve as identification
for the stages. Directly beneath these listings is the stage
cut point calculated from equations C19 and C20 for the actual
test conditions. It is labelled DSO and is given in micrometer
units. The stage weights are likewise listed for the respective
stages, labelled MASS and are in milligram units.
The mass loadings per unit volume of gas sampled indicated
by the stage weights are labelled MG/DSCM/STAGE and are written
236
-------
in milligrams per dry standard cubic meter. The /STAGE indicates
that it is not cumulative. It is calculated for particular stage
j by the formula
MASS.
MG/DSCM/STAGE.. = SAMPLING DURATION (minutes)
35.31 cubic feet/cubic meter Absolute Stack Temperature
x FLOWRATE (ACFM) x Absolute Standard Temperature
Absolute Standard Pressure _ 1 _ (C26)
x Absolute Stack Pressure x (1 - Fraction of H20)
where absolute means the temperature and pressure are in absolute
units-degrees Rankin or degrees Kelvin for temperature, and at-
mosphere, inches or millimeters of mercury for pressure. For
SI,
Mr/nqrM/cTArp _ -72 m9 , 35.31 cubic feet/cubic meter
MG/DSCM/STAGE - x
(400 + 460)°R 29.92 in. fig _ 1 __ . 71 /ncnM
x (70 + 460) °R X 26.50 in. Hg x (1.0 - 0.01) ~ 4'/J- mq/DbCM
The subscripts indicate stage index numbers.
The percent of the mass of particles with diameters smaller
than the corresponding D50 is called the CUMULATIVE PERCENT OF
MASS SMALLER THAN D5 0 . It is the cumulative mass at stage j di-
vided by the total mass collected on all the stages, and converted
to a percentage:
V MASS.
CDM «J = TO Mass * 10°
-------
For example, for S6, the cumulative percent is given by
CUM % - MASS7 + MASS, + MASS,
CUM %6 Total Mass
- 1.25 mg + 0.04 mg + 0.39 mg = 32
-------
rriM ,Mr-/n™n MASS, + MASSK + MASS7 + MASSfl + MASS9
CUM. (MG/ACM) „ = 2Q minutes
35.31 cubic feet/cubic meter
X - 0.500 ACFM -
For S8, the mass of the particulate collected on the filter is
again used,
~Tn. ,..,-,/»™ x MASS, 35.31 cubic feet/cubic meter
CUM. (MG/ACM) = 2Q minu*es x 0.500 ACFM
0.39 mg 35.31 cubic feet/cubic meter
) minutes
= 1.38 mg/ACM
20 minutes x 0.500 ACFM
The cumulative mass loading of particles smaller in diameter
than the corresponding D5 „ in grains per actual cubic foot (CUM.
(GR/ACF) SMALLER THAN D50) for a particular stage j is given by
the formula
CUM. (MG/ACM) .
2.28B 1000 mg/gram
foot y/ ^
for S7,
_ 1.52 mg/ACM
CUM.(GR/ACF) =
= 6.64 x 10 * grains/ACF
The cumulative mass loading of particles smaller in diameter
than the corresponding D50 in grains per dry standard cubic foot
(CUM. (GR/DSCF) SMALLER THAN D50) is calculated to show what the
above cumulative would be for one cubic foot of dry gas at 70°F
and at a pressure of 29.92 inches of mercury. For a particular
stage j.
239
-------
Absolute Stack Temperature
CUM. (GR/DSCF) .. - CUM. (GR/ACF) j x Absolute standard Temperature
Absolute Standard Pressure x
x Absolute Stack Pressure (1-Fraction of H20)
(C31)
where absolute means the temperature and pressure are in absolute
units-degrees Rankin or degrees Kelvin for temperature, and atmo-
spheres, inches or millimeters of mercury for pressure. For SI,
_, (400 + 460)°R
CUM.(GR/DSCF)! = 6.96 x 10 3 gr/ACF x (?Q + 46Q)OR
29.92 in. Hg 1 , 9Q v in-2 nr/nqrp
x 26.50 in. Hg X (1.00-0.01) ~ l'29 X 10 9C/DSCF
The particle size distribution may be presented on a dif-
ferential basis which is the slope of the cumulative curve.
Differential size distributions may be derived two ways:
1. Curves may be fitted to the graphs of cumulative mass
vs. particle size, and then the differential curves (slope) of
each test would be calculated by taking finite differences along
the ordinate and abscissa of the fitted curve.
2. The finite differences may be taken equal to the differ-
ences in D50's from stage to stage (abscissa) and the particulate
mass on each stage (ordinate). This technique was used to cal-
culate the differential size distribution data in Table C3, and
is described in detail in the following paragraphs.
240
-------
If we define the terms:
AM. = MG/DSCM/STAGE . and
(AlogDjj = log! 0(D5 „.._!) - log10(D ..) , then
MG/DSCM/STAGE.
C32)
- log
10
_
(D50..\
Because the computer printer does not contain Greek letters, the
computer printout sheet reads DM/DLOGD instead of AM/AlogD. For
S6,
/ AM \ _ _ 9.35 mg/DSCM - -ao 7
' 6 ~ log10(2.22) - Iog10(1.29) ' 39'7
Note that AM/AlogD has the dimensions of the numerator since the
denominator is dimensionless. In the calculation for SI, a maxi-
mum particle diameter is used. For this example, MAX. PARTICLE
DIAMETER = 100.0 micrometers.
/ AM \ _ _ 4.71 mg/DSCM _ mn/nqrM
, - log, 0 (100) - log10(10.74)- 4'86 ^9/DSCM
For the filter stage, the D50 is arbitrarily chosen to be
one-half of the D50 for stage eight (S8) . For this example, it
is chosen to be 0.33 micrometers/2 = 0.165 micrometers. Thus,
/ AM \ = _ 2.55 mg/DSCM = 8 47
(AlogDJ g log10(0.33) - Iog10 (0.165) 8'47
The geometric mean diameter in micrometers GEO. MEAN DIA.
(MICROMETERS) for a particular stage j is given by the formula
GEO. MEAN DIA.. = ^D50^ x DSO^.J^ (C33)
241
-------
For S8,
GEO. MEAN DIA. 8 = ^0.33 x 0.69 micrometers
= 0.477 micrometers
As in the ALOGD calculation, we again use the maximum particle
diameter for the stage one calculation and one-half the D50 for
stage eight for the filter stage calculation.
For SI,
GEO. MEAN DIA.l = V10-74 x 100.0 micrometers
= 32.8 micrometers
For the filter,
GEO. MEAN DIA. 9 = V0.165 x 0.33 micrometers
= 0.23 micrometers
A differential number distribution (for comparison with ultra-
fine data) can also be derived. Since AM. = MG/DSCM/STAGE . is
the mass per unit volume for stage j, then we can define AN. as
AN. = NUMBER OF PARTICLES/DSCM/STAGE . or the number of particles
per unit volume for stage j. Now AM. and AN. are related by the
equation M = N. x m , where m is the average mass of the par-
ticles collected on one stage. Dividing both sides of the equa-
tion by m x AlogD yields
(AM/AlogD)
- 1 _
mp AlogD .
Now mp = ppvp where pp is the assumed particle density and V
is the average volume of one particle on one stage. To obtain
mp in roiHigram units when p is in grams per cubic centimeter
242
-------
and V is in cubic micrometers, certain conversion factors must
be used. The complete formula, using the correct conversion fac-
tors and the expression (4/3) (IT) (d/2)3 for V where d is the
geometric mean diameter in micrometers, is:
/103 mg\/4TT\/dV/ 10"12 cm3 \
mp pp y 1 gm Ms 1\2 I \i cubic micrometer)
= 5.23599 x 10~10 p d3.
P
Therefore,
(AM/AlogD).
AN \ = :
5.23599 x 10~10p c
P
where AM/AlogD is in units of mg/DSCM, p is in gm/cm3, d is in
microns, and AN/AlogD is in number of particles/DSCM. For S3,
/ AN \ 17.9 mq/DSCM
\AlogDJ 3 = (5.23599 x 10~10) x (1.35 gm/cc) x (7.96 microns)3
= 5.02 x 107 particles/DSCM.
8.47 mq/DSCM
For the filter stage,
/ AN \ = 3
\AlogD/9 (5.23599 x 10 10) x (1.35 gm/cc) x (0.231 microns)3
= 9.72 x 1011 particles/DSCM
The test data are usually classified according to sampling lo-
cation (outlet or inlet), sampling time (day, week, etc.) and
combustion chamber or pollution control device conditions (high
or low sulfur coal for coal plants, normal or below normal fuel
consumption, etc.). When classified, all of the data taken in
a single classification are usually averaged and plotted on ap-
propriate graph paper.
243
-------
Curves are drawn through the discrete points. At selected
particle sizes values of AM/AlogD and AN/AlogD are chosen from
these curves for averaging.
Figure C5 shows a typical average AM/AlogD outlet particle
size distribution curve, from Andersen impactor data, and Fig-
ure C6 shows the same data presented as AN/AlogD.
Figure C7 shows similar data for a hypothetical inlet test
when a Brink impactor was used. The size distributions of Fig-
ures C5 or C7 were used to calculate the penetration/efficiency
curve of Figure C8. Confidence limits were calculated for the
average data as described in Section C.8.
A computer program (Johnson, 1978) has been written which
does all of the calculations necessary to generate plotted graphs
of the particle size distributions and fractional efficiency curves
from the raw field data. Table C4 shows a flow diagram for this
program. Copies of the report describing this computer program
are available from the EPA or NTIS.
244
-------
102
o
w
O
•a
o
o
-------
ERROR BARS INDICATE 90%
CONFIDENCE INTERVAL *
105
10° 101 102
GEOMETRIC MEAN DIAMETER, micrometers
3630-098
Figure C6. Hypothetical particle size distribution at a fabric
filter outlet determined from Andersen impactor
data.
246
-------
104
103
u
V)
Q
o>
C3
O
102
ERROR BARS INDICATE 90% CONFIDENCE INTERVAL
I
10° 101
GEOMETRIC MEAN DIAMETER, micrometers
102
3630-097
Figure C7. Hypothetical particle size distribution at a fabric
filter inlet determined from Brink impactor data.
247
-------
ERROR BARS INDICATE 90% CONFIDENCE INTERVAL
U.UI
0.1
0.2
1
2
§ 5
5 10
cc
uj 20
z
LU
Q.
* 4°
60
80
90
95
98
99
—
—
—
^_
™~
—
' —
—
—
—
I I I I
C
h * * *
, i
III I
III I
$
5
I I
$ 5
I 1
| I
—
—
—
___
••^
—
—
—
—
__
VfM U
99.9
99.8
99
98
95
90
80
60
40
20
10
5
2
o
2
UJ
O
LL
U.
LU
0.1
0.2
0.4 0.6 0.81.0
2.
4. 6. 8.10.0
20.
40. 60. 80.100.
GEOMETRIC MEAN DIAMETER, micrometers
3630-096
Figure C8. Hypothetical fabric filter fractional efficiency based
on the data presented in Figures C5 and C7.
248
-------
TABLE C4
PROGRAM FLOW
For all inlet data;
I. Impactor Program (MPPROG)
Takes testing conditions, stage weights, and impactor constants
to produce stage D50's, cumulative and cumulative % mass concen-
trations < D50, geometric mean diameters, and mass and number
size distributions.
II. Fitting Program (SPLINl)
Uses modified spline technique to fit cumulative mass loading
points for each plot. Stores fitting coefficients and boundary
points on file.
III. Graphing Program (GRAPH)
Produces individual run graphs with points based on stage weights
and impactor D50's. Also superimposes plot based on fitted data,
if desired. Graphs include cum. mass loading, cum. % mass load-
ing, and mass and number size distributions.
IV. Statistical Program (STATIS)
Recalls cum. mass loading fitting coefficients to produce avg.
cum. mass loading, avg. % cum. mass loading, avg. mass size dis-
tribution, and avg. number size distribution plots each with 90%
confidence bars.
Repeat programs I-IV for outlet data.
V. Efficiency Program (PENTRA)
Recalls avg. mass size distribution values along with 90% con-
fidence limits for inlet and outlet to plot percent penetration
and efficiency with 90% confidence bars.
249
-------
APPENDIX D
SIZE DISTRIBUTIONS OF SUBMICRON AEROSOL PARTICLES
If it is desirable to measure the fabric filter collection
efficiency for ultrafine particles, measurements of the particle
size distribution must be made at the inlet and outlet. This
appendix describes procedures for making measurements of the par-
ticle size and concentration from 0.01 to 2 ym diameter.
D.I SYSTEM FOR EXTRACTIVE SAMPLING
When possible, in-stack sampling is preferred because it
eliminates many condensation and sample loss problems which occur
when probes are used for extractive sampling. Unfortunately,
existing submicron sizing techniques and instrumentation are de-
signed for a laboratory environment and cannot be used in-stack.
Particulate concentrations are usually extremely high in
industrial flues and vary by orders of magnitude from one indus-
trial process to another and from the inlet side of a control
device to the outlet of the same device. Temperature, pressure,
moisture content, and the physical properties of the particulate
also vary widely from one industrial process to another. Because
of this complexity and the limited useful concentration range
for particle sizing techniques, extensive sample dilution and
conditioning is required to obtain information on submicron par-
ticles in an industrial gas stream.
250
-------
D.I.I Line Losses
When extracting the sample, attention should be given to
line losses which can be a problem for particles smaller than
0.01 vim or greater than 1.5-2.0 ym diameter. Since the objective
is to measure the concentration of ultrafine particles, there
is little interest in measuring particles larger than about 2.0
Vim. Thus, losses due to impaction and settling are not signifi-
cant, and isokinetic sampling is unnecessary. Diffusional and
electrostatic line losses are of concern, however. For example,
at a sampling rate of 1 ilpm, a sample line will remove 0.005 ym
particles by diffusion at a rate of about 8% of the instantaneous
concentration for every foot of sample line, independent of the
radius of the sample line. This problem can never be eliminated
for the case of out-of-stack sampling, but it can be minimized
by (1) using short probes and as high a flow rate as is practical,
or (2) by using high, nonturbulent, flow rates through a probe
and connecting lines to a conditioning-dilution system in which
a sample is split off at the required rate. This second technique
allows one to use long probes and change sampling points without
disconnecting and reconnecting all sampling lines. Diffusional
line losses for non-turbulent flow can be estimated from the equa-
tions for diffusion to the walls in a circular geometry (Figure
Dl). A similar discussion of diffusional line losses has been
given by Ensor and Jackson (1975).
D.I.2 Condensation of Gases
Another problem of concern is condensation. Elements which
are at a gaseous state at stack temperature (SOs/HaSO^ in parti-
cular) can drop below their dew point and form high concentra-
tions of very small particles resulting in anomalously high read-
ings. In the case of SOa in the presence of H2O, a sulfuric acid
fume can be formed if temperatures fall below the acid dew point
251
-------
100
DIA. FLOW RATE
0.64 cm 0.1 LPM
0.64 cm 0.5 LPM
0.64 cm 0.5 LPM
1.6 cm 10 LPM
LENGTH
A 183 cm
B 305 cm
C 183 cm
D 305 cm
0.01
0.1 1.0
PARTICLE DIAMETER, /im
3630-035
Figure D1. Probe losses due to settling and diffusion for spherical particles
having a density of 2.5 gm/cm^ under conditions of laminar flow.
252
-------
(temperature, pressure, and concentration sensitive). Once
this fume has been formed, very high temperatures are required
to re-evaporate the droplets. For this reason temperatures above
the dew point must be maintained throughout the system until the
gaseous SO3/H2SOi, can be removed or diluted. Two techniques ap-
pear to be useful for doing this: (1) diffusion to an absorber
reagent and (2) dilution of the SOa/HaSCK while hot, to levels
at which the mist will not be formed (Sem, 1975). The main prob-
lem with the second technique, however, is that a further dilution
is required to bring the aerosol to a temperature within the
operating range of the instrument. As a result, the minimum total
dilution, which is the product of hot stage and cold stage dilu-
tions, may be excessive for low particulate concentration levels
such as those found at the outlet of some gas cleaning devices.
Hence the concentration of ultrafine particles even at the mini-
mum dilution may be below the minimum detection level of the
sizing instrument.
At low levels of S03, such as those at power plants burning
low sulfur coal, copper has been found to be a successful SOa/HaSCK
absorber reagent through the formation of CuSO.,. This reagent
has the particular advantage that the reaction product is water
soluble, and the absorber can easily be rejuvenated. Activated
charcoal is also a very effective absorber of SO , even at re-
X
lative high concentrations.
A special effort must be made to detect any particles generating
interferences and to eliminate these if possible by conditioning
the extracted sample. Condensation may be observed by periodi-
cally checking the linearity of the dilution system. When the
dilution system is adjusted to produce a many-fold change in dilu-
tion, the indicated concentration should reflect an equal change
in measured concentration.
253
-------
Figure D2 shows a diffusional absorber/dryer for the removal
of water vapor from the sample stream. Figure D3 shows a sample
extraction system which includes diffusional absorbers for high
temperature use.
D.I.3 Temperatures
Consideration must also be given to reducing the gas tempera-
ture to a level at which the instruments were designed to operate.
This is normally done by using a large volume of cool, dry dilu-
tion air but in some cases can be done by simply pulling the
sample through an ice bath condenser. It should be remembered,
however, that unless condensible gases have previously been re-
moved, it is possible to form a condensation fume when using a
simple condenser.
D.I.4 Electrostatic Losses
Electrostatic line losses can also be a problem. Due to
charges present on the particles, large static electric fields
can be established which result in particle deposition and non-
representative sampling. Line losses may be more severe if the
aerosol particles are charged. Electric fields may exist which
result in particle deposition and nonrepresentative sampling.
This problem is most severe if the sample lines are made of in-
sulating materials. Particle losses may also be a problem in
the diffusion batteries where the theory assumes that the par-
ticles are electrically neutral and no consideration is given
to unknown electrical forces. Also, it is assumed in the appli-
cation of the electrical aerosol analyzer that the sample aerosol
particles initially bear no charge. Precharged particles could
acquire charges different from tnose of calibration, causing them
to exhibit different mobility vs^ size characteristics. It is
desirable to neutralize the particle charge prior to entering
254
-------
GLASS CYLINDER
SAMPLE AEROSOL
DRIERITE
100 MESH STAINLESS
SCREEN
3630-05
Figure D2. Diffusional adsorption apparatus for removal of f-/20 from
sample aerosol.
255
-------
the probe nozzle, and charge neutralization to Boltzmann equili-
brium can be accomplished by exposure to an ion field created
by radioactive materials, however, no suitable radioactive source
has been developed for in-stack applications. The approach gener-
ally taken has been to use radioactive materials such as P0210
in the diluter to neutralize the particulate after it has been
cooled and before it goes to the sizing instruments. The extent
of electrostatic interferences, however, has not been well quanti-
fied.
D.I.5 Humidity
High humidities can alter the charging characteristics of
electrical mobility analyzers and can cause water to condense
on particles and change their size (similar to the controlled
process occurring in a condensation nuclei counter). Humidity
problems can be eliminated by the use of (1) dry dilution air
or (2) diffusional dryers.
D.I.6 Dilution
In-stack concentrations generally exceed instrument concen-
tration limits and must be reduced to levels at which the instru-
ment functions properly. Large changes in the aerosol size distri-
bution due to coagulation are also of concern, particularly in
a diffusional configuration involving long residence times, since
the loss rate due to coagulation for a given size particle rises
rapidly with increasing concentration.
Most of the dilution systems used to date involve the mea-
surement of two different flow rates. Several methods are avail-
able for doing this: (1) Rotameters, (2) Orifices, (3) Venturis,
and (4) Mass Flowmeters. Spink (1957) has summarized the use
of the first three and Parry and Meyer (1974) have described the
256
-------
use of several mass flowmeters in an automated system. Rotameters
are convenient for measuring cool dilution air flow rates but
could cause sample losses when measuring sample gas flows. They
are also sensitive to temperature, pressure, and gas composition.
Flowmetering orifices are useful for measuring flow rates of hot
gases but significant sample losses can occur for particles larger
than about 2 ym. Orifices also require, as do Venturis, that
pressure drops, etc. be monitored in order to calculate the flow
rate. Venturis have the advantage that less turbulence occurs
in the meter and hence size dependent losses should be much less
than those for orifices. Venturis have the disadvantage that
the pressure tap must be taken at the Vena Contracta making them
difficult to construct for low flow rates. Mass flowmeters are
ideal but may require line restrictions in order to throttle the
flow rate.
D.I.7 Sample Extraction-Dilution Systems (SEDS)
Figure D3 is a block diagram of a sample extraction-dilu-
tion system (SEDS) developed by Southern Research Institute under
EPA Contract No. 68-02-2114. In this system a 233 cm3/sec sample
flow is removed from the process exhaust stream and is pulled
through a rigid probe, a flexible connector hose, and a cyclone,
into a "T" where the flow splits. The excess flow is dumped and
the desired sample flow goes into the diluter via a calibrated
orifice and an optional bank of sulfur oxide absorbers. The cyc-
lone, orifice, and sulfur oxide absorber bank are housed in a
heated box so that all components of the system except the diluter
can be maintained at the stack temperature (up to 200°C) to pre-
vent condensation. Pressure taps for the cyclone and the orifice
allow continuous monitoring of the cyclone flow rate and the
orifice flow rate by reading the pressure drop across the respec-
tive component.
257
-------
TIME
AVERAGING
CHAMBER
to
Ul
CO
BLEED DILUTION DEVICE
CHARGE NEUTRALIZES
""-"
PROCESS EXHAUST LINE
CHARGE NEUTRALIZER
CYCLONE
ORIFICE WITH BALL AND SOCKET
JOINTS FOR QUICK RELEASE
SOX ABSORBERS (OPTIONAL)
HEATED INSULATED BOX
RECIRCULATED CLEAN, DRY, DILUTION AIR
o
FILTER BLEED NO. 2
MANOMETER
COOLING COIL
3630-036
PRESSURE
BALANCING
LINE
BLEED NO. 1
Figure D3. Sample Extraction • Dilution System (SEDS)
-------
Currently, charge neutralization is done in the cone of the
diluter by two 500 yC Polonium-210 strips, mounted as shown.
The sample gas enters the dilution chamber at the apex of a per-
forated cone into which clean, dry air is pumped through the per-
forations, creating a highly turbulent mixing zone. Calibration
data for the diluter, Figure D4, is shown in Figures D5 and D6.
At a downstream point, after adequate mixing has occurred, the
diluted sample is extracted and conveyed to the sizing instrument.
This diluted sample passes through a diffusional dryer where any
remaining moisture is removed. The major drying action is accom-
plished by using dilution air which has been passed through an
ice bath condenser.
Just prior to entering the sizing instrument, the sample
passes through a plenum to damp out short term concentration
changes. The exhaust from each instrument is returned to the
diluter to reduce pressure drops across the sizing devices. If
further drying of this recycled gas is necessary, absorption
driers are placed in the instrument exhaust lines.
Changing the sample air flow and the dilution air flow allows
one to change the dilution ratio. Sample air and dilution air
flowmeters are controlled by two bleed valves on the dilution
air pump, one upstream of the pump (#1) and one downstream (#2).
Manipulation of these valves changes the internal pressure of
the diluter which, in turn, sets the sampling rate. As the pres-
sure in the diluter is reduced, the sample flow rate is increased.
In practice, the operation of these valves changes the dilution
air flow only about 10% for a many-fold change in sample flow.
McCain, et al (1974), has reported some problems with an
earlier prototype conditioning system; growth of particles when
high concentrations of S03 are present, pluggage of orifices,
single point sampling limitations, and diffusional losses in
259
-------
SAMPLE
IN
DILUTION
IN
SAMPLE
OUT
3630-O37
Figure D4. Sample Extraction Diluter, cut-away view.
260
-------
10,000
1,000
CC
g
o
o
I-
D
_J
5
iu
100
10
O BASE AT POINT 1 COLLISON
D BASE AT POINT 1 SPRAYER
OBASE AT POINT 2 - COLLISON
O BASE AT POINT 2 - SPRAYER
SAMPLE ORIFICE DESIGNATIONS
.082 .059 . .042 ..029. .021K
.014K
10
100 1,000
CALCULATED DILUTION FACTOR
10,000
3630-038
Figure D5. Calculated dilution versus true dilution for the Southern
Research Institute Ultra fine Particle Diluter, 0.092 \im
particles.
261
-------
10,000
1,000 —
cc
o
UJ
D
CC
O BASE AT POINT 1
D BASE AT POINT 1
O BASE AT POINT 2
O BASE AT POINT 2
COLLISON
SPRAYER
COLLISON
SPRAYER
SAMPLE ORIFICE DESIGNATIONS
100 —
100 1,000
CALCULATED DILUTION FACTOR
10,000
3630-039
Figure D6. Calculated dilution versus true dilution for the Southern
Research Institute Ultrafine Particle Diluter, 0.15
particles.
262
-------
sampling lines. The configuration described above allows the
use of optional SO absorber chambers, rapid replacement for
A
plugged orifices, quick positive determination of partial plug-
gage, full traverse sampling capability, and decreased sample
line losses. The hot box configuration also increases the total
time available for data acquisition by decreasing the time needed
to change orifices or dilution ratios (2 minutes compared to about
30 minutes).
Different sampling and dilution systems have been developed
and reported by Ensor and Jackson (1975), Bradway and Cass (1975) ,
and Schmidt et al (1976) .
It is clear that the most difficult problem in making size
distribution measurements of submicron particles is extracting
a representative sample from the duct and conditioning it for
compatibility with the sizing instruments.
A variety of instruments are available to characterize the
conditioned sample, and these are described in the remainder of
this appendix.
D.2 SUBMICRON PARTICLE SIZING TECHNIQUES BASED ON PARTICLE
DIFFUSIVITY
D.2.1 Diffusion Batteries
Fuchs (1954) has reviewed diffusional sizing work up until
1956, while Sinclair (1972, 1975), Sinclair and Countess (1975),
Breslin et al (1971), Twomey (1963), Thomas (1955), and Sansone
and Weyel (1971) have reported more recent work, both theoretical
and experimental.
263
-------
Diffusion batteries may consist of a number of long, narrow,
parallel channels, a cluster of small bore tubes, or a series
of screens. Variations in the length and number of channels
(tubes, or screens) and in the aerosol flow rate are used as a
means of measuring the number of particles in a selected size
range. As the aerosol moves in streamline flow through the chan-
nels, the particles diffuse to the walls at a predictable rate
depending on the particle size and the diffusion battery geometry.
It is assumed that every particle which reaches the battery wall
will adhere; therefore, only a fraction of the influent particles
will appear as the effluent of a battery. It is only necessary
to measure the total number concentration of particles at the
inlet and outlet to the diffusion battery under a number of con-
ditions in order to calculate the particle size distribution.
When the Stokes diameter is used to describe particle size,
the penetration of diffusion batteries is virtually independent
of physical properties of the individual aerosol particles.
Parallel plate geometry — The parallel plate geometry is
convenient because of ease of fabrication and the availability
of suitable materials, and also because sedimentation can be
ignored if the slots are vertical, while additional information
can be gained through settling if the slots are horizontal. See
Figure D7. Disadvantages of the parallel plate diffusion bat-
teries are (1) the bulk of the diffusional batteries, and (2)
the long transport time required to measure a size distribution.
The mathematical expression for the penetration of a rec-
tangular slot or parallel plate diffusion battery by a monodis-
perse aerosol was given in series form by Twomey (1963).
264
-------
CHANNEL DIMENSIONS
MULTI CHANNEL BATTERY
3630-040
Figure D7. Parallel Plate Diffusion Battery.
265
-------
By varying the number of diffusion batteries in series and
the flow rate, it is possible to measure penetrations under a
variety of conditions. Using a set of diffusion batteries such
as those developed at Southern Research (four-98 channel diffusion
batteries, and a 13 channel diffusion battery) and measuring the
penetration at three different flow rates, yields fifteen data
points from which the particle size distribution (0.01-0.2 ym
diameter) can be reconstructed.
When calculating the fraction (n/n0) of the aerosol which
penetrates a series of diffusion batteries, the transport time
through the diffusion batteries must be taken into account. This
transport time is about 3%-5 minutes for a 98 channel, parallel
plate diffusion battery. Thus, with four of these diffusion bat-
teries in series, n, at a particular time, would be related to
n0 at a point 16-20 minutes earlier in time on a chart recording.
Making a complete measurement of a particle size distribution
requires 2-4 hours and diffusional measurements are the most
useful on stable sources where the distribution is constant in
time when using parallel plate diffusion batteries. The transport
time is determined by the open air volume of the diffusion battery
and flow rate. If the same geometry constant can be obtained
with smaller open air volumes, the transport time can be reduced.
This can be accomplished by decreasing the plate spacing, but
one very quickly approaches problems with material flatness errors,
or equipment survivability when such materials as precision ground
graphite sheets are used. These problems seem to have been elimi-
nated with the screen geometry, which is discussed below.
Screen geometry — Breslin e_t al (1971) and Sinclair (1972)
report success with more compact, tube-type and screen-type ar-
rangements in laboratory studies.
266
-------
Although the screen-type diffusion battery must be calibrated
empirically, it offers convenience in cleaning and operation,
and compact size. Figure D8 shows Sinclair's geometry. This
battery is 21 cm long, approximately 4 cm in diameter, and weighs
0.9 kg.
This system is commercially available from TSI Incorporated,
St. Paul, MN 55113 as the Model 3040 Diffusion Battery. Because
of the small internal volume of the battery, the time necessary
to obtain a test on one battery is reduced by about a factor of
ten as compared to the parallel plate batteries. This diffusion
battery system allows data to be collected on process streams
where the particle concentration is somewhat unstable (^15 minutes
per cycle).
A disadvantage of the small volume is the sensitivity to
surges in flow rate caused by the commercially available CNC
counter.
D.2.2 Particle Concentration Indicators-Condensation Nuclei
Counters (CNC)
Condensation nuclei counters function on the principle that
particles act as nuclei for the condensation of water or other
condensable vapors in a supersaturated environment. This process
is used to detect and count particles in the 0.002 to 0.3 micron
range (often referred to as condensation or Aitken nuclei). In
condensation nuclei detectors, a sample is withdrawn from the
gas stream, humidified, and brought to a supersaturated condition
by reducing the pressure. In this supersaturated condition, con-
densation will be initiated on all particles larger than a certain
critical size and will continue as long as the sample is super-
saturated. This condensation process forms a homogeneous aerosol,
predominantly composed of the condensed vapor containing one drop
267
-------
10
SAMPLING
PORT (TYP)
SECTION CONTAINING
SCREENS (TYP)
3630-045
Figure D8. Screen type diffusion battery. The battery is 21 cm
long, 4 cm in diameter, and contains 55 635 mesh
stainless steel screens.
268
-------
for each original particle whose size was greater than the criti-
cal size appropriate to the degree of supersaturation obtained;
a greater degree of supersaturation is used to initiate growth
on smaller particles. The number of particles that are formed
is estimated from the light scattering properties of the final
aerosol.
Because of the nature of this process, measurements of very
high concentrations can be in error as a result of a lack of cor-
respondence between particle concentration and scattering or at-
tenuation of light. Additional errors can result from depletion
of the vapor available for condensation. Certain condensation
nuclei measuring techniques can also obtain information on the
size distribution of the nuclei; that is, variations in the degree
of supersaturation will provide size discrimination by changing
the critical size for which condensation will occur. However,
MacLauren and Junge (1971) have predicted that the critical size
for initiating condensation is also affected by the volume frac-
tion of water soluble material contained in the original aerosol
particle, so the critical size will be uncertain unless the solu-
bility of the aerosol particles is known. At very high degrees
of supersaturation (about 400%), solubility effects are only minor
and essentially all particles in the original aerosol with diam-
eters larger than 0.002 ym will initiate the condensation process.
A continuous flow CN counter has been described by Sinclair
and Hoopes (1975) and an absolute calibration of a CN counter
has been done by Liu and Pui (1974). The theory and principles
of operation of CN counters has been described by Haberl (1970)
and Haberl and Fusco (1970).
269
-------
D.2.3 Using Diffusion Batteries With Condensation Nuclei Counters
and a Sample Extraction-Dilution System To Measure Concen-
trations of Submicron Aerosols In Industrial Flue Gases
Before taking the equipment into the field, a preliminary
examination of the sampling site should be made. There must be
ample space for the diffusion batteries (D.B.) and the condensa-
tion nuclei counter (CNC). The sample lines to and from the D.B.'s
should be as short as is practical. The diffusion batteries
should be placed out of direct sunlight to reduce thermal inter-
ferences .
Four models of CMC's are presently used. Two automatic
models are the General Electric (GE)* and the Environment One
Model Rich 100 (E-l).** Small manual particle detectors are also
commercially available from Gardner Associates*** and Environment
One.** The valving system in the GE is mechanical, and pressure
differentials across the valves are permissible. However, the
E-l has pneumatic valves and a pressure difference of greater
than 2 inches of water across the valves will lock them either
open or shut. Soderholm (1976) has described some modifications
to the E-l Rich 100 that replace the pneumatic valves with sole-
noid valves and smooth out pulsations in the sample flow. A re-
turn line is normally used on the GE but may not be connected
to the diluter in some circumstances. The GE creates substantial
pulsation in the sample lines which is intolerable for diffusional
analysis, and an antipulsation device consisting of two metal
cylinders connected by a small orifice may be used as a pneumatic
R-C network to damp the oscillations to an acceptable level.
* General Electric - Ordnance Systems, Electronics Systems
Division, Pittsfield, MA 01201.
** Environment-One Corporation, Schenectady NY 12301.
*** Gardner Associates, Schenectady, NY 12301.
270
-------
Once all sample and return lines have been connected, the
equipment is turned on and allowed to warm up. The flows are
then adjusted to the proper rates. If the GE model is used, the
vacuum gage on the front panel should read 8 inches of Hg as
recommended in the operation manual. This gives a nominal flow
rate of 6 fcpm and a sample supersaturation after expansion of
approximately 400%. The E-l may be adjusted to any desired flow
rate between about 0.6 fcpm and 4.2 fi-pm.
The water supplies for the humidifiers are filled with a
mixture of distilled water and a wetting agent. About 0.5% photo-
flow (ethylene glycol) is used in the water. This reduces the
surface tension of the water and allows the wick to wet better
and more quickly.
Before any data can be taken the system should be leak
checked. This is done by connecting all the D.B.'s in series,
clamping one end off and pumping several inches of mercury vacuum.
If the vacuum holds, data can be taken. If there is a leak it
can be isolated by repeating the process with successively fewer
D.B.'s until the leaky one is found. The exact location of the
leak can be determined by putting a concentrated condensation
nuclei source (e.g., a burning string) close to various parts
of the D.B. When the leak is found, the CNC panel meter will
rapidly rise.
There are two methods of data collection: graphing the CNC
output on a strip chart recorder or directly writing down the
meter reading. In general, a combination of the two is used to
insure that no faulty connection exists between the CNC and the
chart recorder or that the chart recorder is malfunctioning.
Since with either the GE or the E-l CNC the largest flow
possible is 6 fcpm, a certain minimum amount of time is required
to pull the sample through the parallel plate D.B.'s. If gra-
271
-------
phical techniques are used a characteristic output will be ob-
served. The indicated concentration is zero while the clean air
already in the D.B.'s is being exhausted. Then the output rises
to a peak and stabilizes. It is at this point that meaningful
data is being taken. However, if data is taken by meter readings
alone this characteristic response is very difficult to follow.
Adequate time must be allowed for transport through the D.B.'s
before each reading is taken. These transport times have been
calculated for 6 &pm for two types of parallel plate D.B.'s used:
Thirty seconds for the 13 channel parallel plate and five minutes
for each 98 channel parallel plate D.B. used. Quite frequently
the in-stack concentrations are unsteady and data is normally
recorded for several minutes after the system has had time to
stabilize as a double check on the validity of the data and to
provide some time averaging of the data on fluctuating sources.
With the screen type diffusion batteries, transport times of about
1/10 of the times given above can be used.
If the GE CNC is used, diffusional sizing cut points are
selected by changing the number of D.B.'s, since the flow rate
is constant. However, with the E-l, the cut sizes may be adjusted
by either changing the number of D.B.'s or the flow rate. The
flow rate is sometimes held constant because source fluctuations
introduce so much uncertainty that attempts to achieve high resolu-
tion are futile.
In order to obtain a set of data, the CNC is first connected
directly to the diluter to obtain total concentration. Data are
then taken by pulling the sample next through a single battery,
two batteries, three batteries, etc., until all the available
permutations of geometry and flow rate have been used.
If only one set of equipment is available, after data are
collected at the inlet the equipment is carried to the outlet,
set up, leak checked, and the above procedure is repeated.
272
-------
D.2.4 Data Reduction Techniques and a Sample Calculation
Fuchs et al (1962) presented a technique for calculating
the particle size distribution from raw data, assuming that the
size distribution was log normal. A technique suggested by Sin-
clair (1972) does not include this restriction. In Sinclair's
method a nomograph is prepared using the penetration for each
diffusion battery geometry and flow rate and a large number of
monodisperse particle sizes. Comparing this nomograph with ex-
perimental penetrations, one calculates the particle size dis-
tribution using a "graphical stripping" process. However, it
is usually more convenient to use a "D5o" technique like that
used for the reduction of cascade impactor data.
Data for a given test condition are averaged as follows to
yield a representation of the source at that condition. All in-
strument readings are converted to indicated concentrations by
means of individual instrument calibration curves. These values
are then corrected for dilution to obtain flue gas concentration.
Since the CNC calibration curve is non-linear, multiplying by
the dilution factor before converting to concentration will yield
erroneous results. Next, process averaging is accomplished by
taking appropriately weighted averages of the data obtained through
the various process cycles. These weighted averages are found
for each D.B. arrangement, including no D.B.'s, and test condi-
tion, both inlet and outlet.
For the sample calculation it is assumed that five parallel
plate diffusion batteries were used, in four configurations, and
at three flow rates to obtain data at twelve D50 sizes. These
five diffusion batteries consisted of one 13 channel (Type A)
and four 98 channel (Type B) units. The four sampling configura-
tions were (1) one Type A, (2) one Type B, (3) two Type B in series,
and (4) four Type B in series. These diffusion batteries are
273
-------
similar to the one depicted in Figure D7. The three flow rates
were 1, 6, and 10 liters per minute. The aerosol was sampled
from a Sample Extraction and Dilution System and the total number
of particles entering and exiting the diffusion batteries was
determined using a condensation nuclei counter (CNC). The sample
data for this experiment is shown in Table Dl.
To calculate the particle size for 50% penetration (the D50)
through a diffusion battery configuration, the following equations
must be used.
The penetration of a rectangular plate diffusion battery
is given by
-21.43lDfY
0.015e-62'317DfY + 0.0068e-- (Dl)
where Df = Particle diffusivity, cm /sec,
Y = Diffusion battery flow rate-geometry constant,
sec/cm ,
m = Number of identical batteries in series,
n0 = Diffusion battery inlet concentration, #/cm3 , and
n = Diffusion battery outlet concentration, #/cm3 .
The particle diffusivity, Df, is given by
Df = kTB, (D2)
where k = Boltzmann's Constant, gm cm2/sec2 °K,
T = Absolute Temperature, °K, and
B = Particle Mechanical Mobility, sec/gm.
274
-------
TABLE Dl
DIFFUSION BATTERY SAMPLE CALCULATION DATA
CNC
Configura-
tion
1 Type A
1 Type B
2 Type B
4 Type B
1 Type A
1 Type B
2 Type B
4 Type B
1 Type A
1 Type B
2 Type B
4 Type B
Flow
Rate
U/min)
6
6
6
6
1
1
1
1
10
10
10
10
Calc.
D5o
(lam)
0.02
0.06
0.098
0.17
0.028
0.07
0.12
0.20
0.015
0.045
0.08
0.14
Reading
(#/cm3)
53000
43000
45000
11000
84000
33600
29000
5500
64000
62000
66500
20500
Dilution
Factor
1000
500
200
200
500
500
200
200
1000
500
200
200
Actual
Concentration
(#/cm3 x 106)
53.0
21.5
9.0
2.2
42.0
16.8
5.8
1.1
64.0
31.0
13.3
4.1
275
-------
The particle mechanical mobility is given by
B = [l + 2.49 (X/d) + 0.84 (X/d)e~ (0' 44) (d/X)J /Supd (D3)
where X = gas mean free path, cm,
d = particle diameter, cm, and
\i = gas viscosity, gm/cm-sec.
The gas mean free path is given by
X = 3.109 x lO"1^ T/M
-------
The two diffusion batteries under consideration in this
example have the following dimensions.
Type A Type B
L 45.72 cm 45.72 cm
h 10.15 cm 11.46 cm
W 0.10 cm 0.10 cm
N 13 98
Thus, YA = 2.358 x 103 sec/cm2 and YB = 2.006 x 10" sec/cm2 for
Q = 6 £/min.
The manipulation of these equations allows one to calculate
the penetration of the diffusion batteries at different flow rate
and particle-size combinations. After plotting the penetration
versus particle size, the Dso's can be determined.
To aid in the calculation of the particle diffusivity and
penetration of a particular diffusion battery arrangement, two
programs have been written for the Hewlett-Packard HP-65 Program-
mable Calculator. If a programmable calculator is not available,
the calculations can be done manually using the equations given
above. The calculator programs and their applications are de-
scribed below:
Program 1 is used to calculate the viscosity (y) and mean
free path (X) of standard air. These values are then used to
calculate the diffusivity (Df) for a monodisperse aerosol having
diameter d. Given values for the flow rate-geometry configuration
(Y) and the diffusivity (Df), Program 2 is used to calculate the
theoretical penetration (n/n0).
277
-------
Program 1; Diffusivity (Program Steps are Listed in Table D2.)
Over the temperature range from 0-350°C, the viscosity of
dry standard air is very nearly linear. For a given temperature
p can be found, in CGS units, from:
y =[0.494 (T - 294°K) + 18l]x 10 6 poise
Knowing the viscosity, the mean free path is given by
X = 5.77^VT x 10~2 cm for standard air (P is the absolute
pressure, atm.) .
From a knowledge of y and X for the carrier gas, the dif-
fusivity (Df.) of a monodisperse aerosol having diameter (d.)
is given by the following equation:
Df
. = (l.
46 x
Ud.
1 + 2.49f^-
^-(0.44) (d./X)
i
for T in °K, y in poise, X and d. in cm.
User Instructions. Enter the program shown in Table D2.
To calculate y and X for Standard Air having 250°K < T <
600°K;
1. Load storage registers with the following variables
Temperature T,
Absolute Pressure, P,
"Hg,
STO 2
STO 3
2. Start program "y, \"
"B1
278
-------
TABLE D2
PROGRAM 1 - PARTICLE DIFFUSIVITY (Df)
HP-65 CALCULATOR PROGRAM FOR A PARALLEL PLATE DIFFUSION BATTERY
CODE
LBL
A
RCL 1
RCL 5
-r
•
4
3
5
CHS
X
f1
LN
RCL 5
RCL 1
-7-
STO 8
X
•
8
4
X
1
+
RCL 8
2
•
4
9
2
X
+
RCL 2
X
1
4
KEYS
23
11
34 01
34 05
81
83
04
03
05
42
71
32
07
34 05
34 01
81
33 08
71
83
08
04
71
01
61
34 08
02
83
04
09
02
71
61
3402
71
01
04
CODE
6
EEX
CHS
1
9
X
RCL 4
-j-
RCL 1
-f-
STO 7
RTN
LBL
B
•
4
9
5
RCL 2
2
9
4
—
X
1
8
2
+
EEX
CHS
6
X
STO 4
RCL 3
-r
RCL2
KEYS
06
43
42
01
09
71
3404
81
34 01
81
33 07
24
23
12
83
04
09
05
34 02
02
09
04
51
71
01
08
02
61
43
42
06
71
33 04
34 03
81
34 02
CODE
f
V
X
•
0
5
7
7
X
STO 5
RTN
KEYS
31
09
71
83
00
05
07
07
71
33 05
24
R-| d
R2
RS
T
P
R4
RB
M
X
Rg (BLANK)
R7 Dfj
R8 (WORK)
Rg (BLANK)
279
-------
3. Output:
Values of y are stored in the correct storage register for
retrieval during calculation of Df
To display y (units of poise) RCL 4
To display X (units of cm) RCL 5
To calculate Df. for a monodisperse aerosol
4. Load Storage registers with the following variables
Temperature, T,
Viscosity, y ,
Mean Free Path, X ,
Diameter of the
monodisperse
aerosol, d. ,
°K, STO 2
poise, STO 4
cm, STO 5
cm, STO 1
Performed
in Steps
1 and 2
5. Start program "Diff" "A"
6. Output:
Displayed value is "Df" in cgs units (also stored in
register 7).
Test Problem
Find y, X, and Dfi for T = 75.2°F, P = 14.39 PSIA, and
d = 0.023 microns (ym) .
1. Convert T, P, and d. to the proper units then store in
registers 2, 3, and 1, respectively.
75.2°F = 24°C = 297°K, store in register 2
280
-------
14.39 PSIA x 2.036 1"l°9 = 29.30 in. Hg, store in register 3
Jr & .LA
0.023 pm = 0.023 x 1Q-1* cm = 2.3 x 10~6 cm, store in
register 1
2. Start Program B. X is displayed (X = 6.23 x 10~6 cm), RCL 4
to display (vi = 1.83 x 10"1* poise).
3. Start Program A. Df. is displayed (Df. = 9.95 x 10~5cm2/sec).
Program 2; Theoretical Penetration (Program Steps are Listed
in Table D3.)
When the diffusivity (Df.) of a monodisperse aerosol of diam-
eter d. is known, and the flow rate geometry configuration (Y.)
for the diffusional apparatus (diffusion battery) are known, frac-
tional penetration as a function of size (d.) is given by
-1.8852Df.Y. -21.43lDf.Y.
p. . = 0.9104e 1 D + 0.0531e 1 D
!/ D
-62.317Df.Y. -124.537Df.Y.
+ 0.015 1 -1 + 0.068e i D
If several diffusional configurations are used in series
where the aerosol is allowed to remix before going through the
next configuration, the final penetration is the product of the
penetration for each, thus
p. = II p. .
*i FifD
If m diffusion batteries having the same Y. value are used
in series, then
m
281
-------
TABLE D3
PROGRAM 2 - PARTICLE PENETRATION (p)
HP-65 CALCULATOR PROGRAM FOR A PARALLEL PLATE DIFFUSION BATTERY
CODE
LBL
A
RCL 7
RCL 6
X
STD 8
1
8
8
5
2
CHS
X
f1
LN
•
9
1
0
4
X
RCL 8
2
1
4
3
1
CHS
X
f1
LN
•
0
5
KEYS
24
11
34 07
3406
71
33 08
01
83
08
08
05
02
42
71
32
07
83
09
01
00
04
71
34 08
02
01
83
04
03
01
42
71
32
07
83
00
05
CODE
3
1
X
+
RCL 8
6
2
•
3
1
7
CHS
X
f1
LN
•
0
1
5
3
X
+
RCL 8
1
2
4
-
5
3
7
CHS
X
f1
LN
0
KEYS
03
01
71
61
34 08
06
02
83
03
01
07
42
71
32
07
83
00
01
05
03
71
61
3408
01
02
04
83
05
03
07
42
71
32
07
83
00
CODE
0
6
8
X
RCL
9
9
yx
RTN
KEYS
00
06
08
71
61
34
09
35
05
24
RI
R2
(BLANK)
(BLANK)
RS p
R4
(BLANK)
R5 (BLANK)
Re
Yi
R? Dfi
RS
Rg
(WORK)
m
282
-------
To calculate p. . for m identical batteries in series,
enter program in Table D3.
1. Load storage registers with the collowing variables
Diffusivity, Df
Flow rate-geometry Y.
configuration
Number of identical m,
batteries in series
cgsf units,
cgs units,
integer,
STO 7
STO 6
STO 9
2. Start Program "A"
Displayed value is the fractional penetration (p. •) of
1 f J
this monodisperse aerosol of size d^^ through m identical
diffusion batteries in series.
Test Problems
1. Load data
Y. = 2.36 x 103 in cgs units
f.^ = 5.39 x 10~6 in cgs units
m = 1
STO 6
STO 7
STO 9
2. Calculate p
p = 9.38 x 101 = 93.8%
Therefore, the penetration of the diffusion battery Type A at
a flow rate of 6 liters/minute is 93.8% for 0.023 ym particles
283
-------
After calculating the penetration, a graph similar to Figure
D9 will be obtained. The Dso's can be determined from this graph
as indicated. The experimental data for this sample calculation
are shown in Figure Dll.
A cumulative particle size distribution is plotted using
the corrected concentration from the last column in Table Dl as
the ordinate and the Dso's for each configuration as the abscissa.
Differential number graphs are obtained by differentiation of
the cumulative curve (finite differences) as described in Sec-
tions C.8 and C.9.
D.3 SUBMICRON PARTICLE SIZING TECHNIQUES USING THE ELECTRICAL
MOBILITY PRINCIPLE
D.3.1 Electrical Mobility
If a particle charge is known, measurements of the electrical
mobility can be used to determine the particle size. This concept
has been used by Liu, Whitby, and Pui (1974) at the University
of Minnesota to develop a series of Electrical Aerosol Analyzers
(EAA). A schematic of this device is shown in Figure DIG. The
EAA has the distinct advantage of very rapid data acquisition
compared to parallel plate diffusion batteries used with condensa-
tion nuclei counters (two minutes as opposed to two hours for
a single size distribution analysis).
D.3.2 Thermosystems Model 3030 Electrical Aerosol Size Analyzer*
The EAA is designed to size particles in the range of 0.0032
micrometer diameter to 1.0 micrometer diameter. It can size solids
Thermosystems, Inc., St. Paul, MN 55113.
284
-------
0.01
0.02 0.03 0.04 0.05 0.1
PARTICLE DIAMETER, //m
0.2
0.3 0.4 0.5
3630-042
Figure D 9. Theoretical parallel plate diffusion battery penetration
curves.
285
-------
CONTROL MODULE
ANALYZER OUTPUT SIGNAL - - -
DATA READ COMMAND - - -
CYCLE START COMMAND - - • -
CYCLE RESET COMMAND - •
• AEROSOL FLOWMETER READOUT
CHARJED CURRENT READOUT
CHARGER VOLTAGE READOUT
AUTOMATIC HIGH VOLTAGE CONTROL AND READOUT
ELECTROMETER (ANALYZER CURRENT I READOUT
TOTAL FLOWMETER READOUT
NJ
00
FORCES OH MBTICLE
,-CLECTAOSTATIC FOKCC
d-ACROOYNAMIC DAAO
-» EXTERNAL
-» DATA
^ACQUISITION
1 SYSTEM
— TO VACUUM PUMP
3630-043
Figure D10. Schematic diagram of the electrical aerosol analyser.
After Liu and Pui(1975).
-------
and non-volatile liquids. The instrument's concentration range
of 1 to 1000 yg/m3 requires that a sample extraction-dilution
system (as described earlier) be used.
The EAA is operated in the following manner. As a vacuum
pump draws the aerosol through the analyzer (See Figure D10),
a positive corona (generated at a high voltage wire within the
charging section) gives the sample a positive electrical charge.
The charged aerosol flows from the charger to the analyzer
section as an annular cylinder of aerosol surrounding a cone of
clean air. A metal rod, to which a variable, negative voltage
can be applied, passes axially through the center of the analyzer
tube. Particles smaller than a certain size (with highest elec-
trical mobility) are drawn to the collecting rod when the voltage
corresponding to that size is on the rod. Larger particles pass
through the analyzer tube and are collected by a filter. The
electrical charges on these particles drain off through an electro-
meter, giving a measure of current.
A step increase in rod voltage will cause particles of a
larger size to be collected by the rod with a resulting decrease
in electrometer current. This decrease in current can be related
to the increased number of collected particles by the current
sensitivities (AN/AI) as determined by Liu and Pui (1975). Liu's
work determined empirical values for the current sensitivities
for the case when the ion electrical mobility had a value of 1.4
cm3/volt-sec. Marlow, Reist, and Dwiggins (1976) have done work
that indicates that the presence of certain trace elements (e.g.,
0.01 ppm SO2) in the instrument's sheath air can significantly
affect data obtained with the EAA. They suggested several pos-
sible explanations, one of which was that any S02, present in
the charger sheath air, can contribute substantially to the cur-
rent carrying ion species, and thus, significantly alter the
287
-------
effective ion mobility for particle charging. A total of eleven
voltage steps divide the 0.0032 to 1.0 micron size range of the
instrument into ten equal logarithmic size intervals. Different
size intervals can be programmed via an optional plug-in memory
card.
The electrical aerosol analyzer can be operated either auto-
matically or manually. In the automatic mode, the analyzer steps
through the entire size range. For size and concentration moni-
toring over an extended period of time, the analyzer may be inter-
mittently triggered by an external timer. The standard readout
consists of a digital display within the control circuit module,
although a chart recorder output is available. It is almost
always advantageous to use a strip chart recorder to record the
data. This allows the operator to identify a stable reading
superimposed on source variations and gives a permanent record
of the raw data.
D.3.3 Using the TSI Model 3030 Electrical Aerosol Analyzer
With a Sample Extraction-Dilution System To Measure
Concentrations of Submicron Aerosols in Industrial
Flue Gases
Once the equipment is set up as shown in Figure D3, the flows
are adjusted through the sample orifice and the dilution air
orifice, to obtain the desired dilution factor. The EAA is placed
in a manual mode and the current readings for each channel are
recorded with a strip chart recorder. Manual control allows run
times of from two to five minutes in each of the nine channels.
This allows one to average out rapid source fluctuations. At
the beginning of each day, the internal calibration points and
flows through the EAA are checked, as described in the instrument
manual. In order to assure correct residence time through the
charger section the operator must measure the temperature and
288
-------
pressure of the aerosol in order to calculate the mass flows that
yield actual flows of 4 &pm and 50 5,pro.
D.3.4 Data Reduction Techniques and a Sample Calculation
It is assumed that a Thermo-Systems Inc. Model 3030 Elec-
trical Aerosol Size Analyzer (EAA) with a 0.0032 ym to 0.360 ym
range at the normal operating conditions has been used to deter-
mine concentration vs size information in the ultrafine size range
for the effluent of a fabric filter. The EAA sampled the gas
stream after the sample was extracted with a Sample Extraction
and Dilution System as described in Section D.I.7.
The EAA was placed in a manual scan mode and the current
readings for each channel were recorded with a strip chart re-
corder. Manual control allowed run times of up to one minute
for each of the nine channels. This allowed the averaging of
rapid source fluctuations.
The theory of operation and basic equations for the EAA have
been given by Sem (1975). The EAA was initially marketed in a
configuration having an N. product of 7 x 106 (ions/cm)(sec) and
a set of preprogrammed collection rod voltages aimed at obtaining
eleven size cuts between 0.0032 and 1.00 ym. Calibration by Liu
and Pui (1975) showed that for an N. product of 7 x 106 [and an
assumed ion electrical mobility of 1.4 cm3/(volt-sec)] these pre-
programmed voltages resulted in eleven size cuts between 0.032
and 0.36 ym. Liu's work determined the correct size cuts and
associated current sensitivities for the preprogrammed voltages.
It is these values that are shown in Table D4 and used in the
example. Liu's work also determined an N. value and set of col-
lection rod voltages that obtained the originally desired set
of size cuts (from 0.0032 to 1.00 ym) . This N. = 1 x 107 con-
figuration and the set of voltages used by Liu are now the
289
-------
standard configuration for the Thermo System Inc. EAA. The me-
chanics of the data reduction for the Nfc = 1 x 107 configuration
and the N = 7 x 106 configuration (shown in the example herein)
are identical. A computerized data reduction program for the
N = 1 x 107 configuration (Liu and Kapodia, 1977) is now com-
mercially available from Thermo Systems Inc. This program at-
tempts to correct for the multiple charging effects noted by Liu
by assuming a log normal distribution (unimodel or bimodel) and
using Liu's calibration matrix to arrive at a best fit to the
measured set of I values.
Table D4 is essentially self-explanatory- The heading "D ,
ym" (Column 3) is the particle diameter in micrometers. A value
of 0.100 ym means that particles equal to or smaller than this
diameter are collected in the analyzer tube while larger particles
penetrate to the current collecting filter where an electrometer
measures the total current carried by the unprecipitated particles.
This current represents the charges on all particles larger than
0.100 ym. This measured current is the basic output of the Model
3030.
The fourth column (D . f jam) is the geometric mean diameter
of the particles represented by the current difference of two
successive steps (Channel No.*s). For example, the difference
in current for the 0.100 ym cut-off and the current for the 0.0178
Vim cut-off is the total current collected from particles between
these sizes, or rather for a mean diameter of 0.0133 ym. The
current differences are entered in Column 8 headed "I, pA" (pico-
Amps).
The fifth column gives the revised calibration factor (based
on the calibration by Liu and Pui (1975)) for each of the eight
size bands. These factors are in units of particles per cm per
picoAmpere. Multiplying this size specific current sensitivity,
AN/AI, (Column 5) by the current difference, AI, (Column 8) gives
290
-------
TABLE D4
EAA (Model 3030) Data Reduction Form
Concentration, Cumulative Concentration, and AN /ALogD From
Scan No. For DF =
Nt = 7x10°
1
Channel
No.
3
to 4
10
H
5
6
7
8
9
10
11
2
Collector
Voltage
196
593
1220
2183
3515
5387
7152
8642
9647
3
D , ym
0.0100
0.0178
0.026
0.036
0.070
0.120
0.185
0.260
0.360
4
Dpi, pm
0.0133
0.0215
0.0306
0.0502
0.0917
0.149
0.219
0.306
5
AN/AI
4.76xl05
2. 33x10 5
1.47xl05
8.33x10"
4.26x10"
2.47x10"
1.56x10"
1. 10x10"
6 7 8 9 10 11 12
AlogD I/pA AI,pA AN AN IN AN /AlogD
P S 5 S
0.250
0.165
0.141
0.289
0.234
0.188
0.148
0.141
-------
the total number of particles, AN, (Column 9) in units of par-
ticles per cm3, within this size band (Column 4) for the diluted
aerosol. To correct for dilution and find in-stack concentrations,
multiply Column 9 by the dilution factor (DF) and enter the re-
sult, AN , in Column 10. Columns 6 and 12 are used for AN /ALogD
information calculated from the number distribution in Column
10. Column 11 is used for cumulative concentrations, corrected
for dilution. All concentration have been corrected to engineer-
ing standard (normal) conditions by the dilution factor. Engineer-
ing standard or normal conditions are defined as dry gas at 21°C
and 760 mm Hg pressure.
The basic data from the EAA is cumulative current for each
of nine channels (Column 7). One must then take the differences
of the current readings for successive channels (Column 8) in
order to find AN, etc. The Al values are multiplied by a series
of constants (AN/AI., DF.) to arrive at AN (concentration in
1 D s
stack corrected to dry, standard conditions). While a single
scan should be made at a constant dilution, different scans may
be made at different dilutions. To simplify the arithmetic for
each test condition, the product a. = Al. . x DF. is formed and
all such inlet (outlet) products for the same size band are
averaged.
SUMMARY OF THE CALCULATION FORMAT
STEP 1
A. Calculate the average instrument reading (I) for each channel
as obtained from the strip chart recording of channel current
vs. time.
B. Calculate all dilution factors (DF.
292
-------
STEP 2
Calculate current differences (AI. .) from adjacent channels
11J
and average the a. products (a. = AI. . x DF.) for the same size
1 1 !r3 D
band for all scans taken for the same test conditions. Calculate
90% confidence intervals for each a.. Note: the i subscript
denotes size and the j subscript denotes dilution setting.
STEP 3
Using a. and Table D6 calculate "number concentration" (AN ),
1 5
"average cumulative concentration of all particles having diam-
eter greater than the indicated size" Z(AN ), and "AN /ALogD"
s s
for each size band for each test condition.
STEP 4
Plot "Cumulative Concentration vs. Size" for each test con-
dition.
STEP 5
Plot AN /ALogD (with upper and lower 90% confidence limits)
S
vs. size for each test condition.
SAMPLE CALCULATION FOLLOWING THE CALCULATION FORMAT
Table D5 contains hypothetical test data for the following
sample calculation.
STEP 1
A. Calculate the average instrument reading (I) for each channel
as obtained from the strip chart recording of channel current
vs. time. Each complete size scan (Table D5) consists of nine
293
-------
TABLE D5
EAA Current Readings (I, in picoamps and Dilution Factors)
For This Sample Calculation: Hypothetical Inlet Data
Nt = 7xlO?
SCAN
1
2
3
4
5
6
7
8
9
10
Time
l:30p
1:32
1:34
1:36
1:38
1:40
1:45
1:47
1:49
1:51
CH 3
2.869
2.835
2.841
2.859
2.866
2.866
6.477
6.580
6.377
6.390
CH 4
2.734
2.711
2.709
2.722
2.740
2.736
6.188
6.288
6.087
6.094
CH 5
2.519
2.495
2.500
2.522
2.530
2.531
5.716
5.818
5.620
5.614
CH 6
2.227
2.205
2.200
2.235
2.251
2.238
5.056
5.153
4.960
4.956
CH 7
1.362
1.344
1.340
1.368
1.381
1.378
3.111
3.233
3.021
3.006
CH 8
.682
.669
.655
.676
.714
.698
1.575
1.613
1.526
1.467
CH 9
.242
.220
.218
.226
.279
.255
.565
.510
.537
.492
CH 10
.102
.075
.081
.096
.137
.115
.243
.195
.227
.187
CH 11
.020
- .010
.001
.010
.052
.033
.053
.010 _
.032
.005
Dilution Factoi
255
255
255
255
255
255
113
113
113
H3 , .. .
-------
instrument readings (I, Column 7 of Table D4). These instrument
readings are the average current outputs as taken from the strip
chart recordings, for each of the nine channels. Run times were
manually controlled and varied from two to ten minutes per chan-
nels 3, 4, 5, ..., 11. Table D5 gives the instrument readings
used as data for the sample calculation (10 scans, 90 average
current readings).
B. Calculate all dilution factors (DF.; corrected to engineer-
ing standard (normal) conditions: 70°F (20°C) and 29.92 inches
of mercury pressure (760 mm Hg)) .
STEP 2
Calculate current differences (AI- •) from adjacent chan-
nels and average the a- products for the same size band for all
scans taken at the same test condition. Calculate 90% confidence
intervals for each a.^. (Refer to Section C.8.5).
ai = AIi,j x DFj
where i denotes the size band and j denotes the dilution value.
For channels 3-4 we have:
Scan #1: a3_4 j^ = (0.135) (255) pA
#2: o^-/! = (0.124) (255) pA
#3: ctQ ' = (0.132) (255) pA
19: a3_4 2 = (0.290)(113) pA
#10: a3_4'2 = (0.296)(113) pA
thus, a_ = 33.179 pA; n = 10 and CI = 0.579,
295
-------
In a similar manner we can find a4_5» «5_6' • • -' aio-ll*
Thus, the mean, with upper and lower 90% confidence limits for
S3_4 is given by:
a = (33.179 ± 0.579) pA
or
a3_4 = (33.2 ± 0.6) pA
STEP 3
Using a- and Table D6 calculate "number concentration" (ANg) ,
"average cumulative concentration . . ." (ZANg), and "ANg/ALogD"
for each size band for each test condition.
Table D7 shows these calculations for the sample data of
Table D5. Column 7 is a as shown in Step 2. Column 8 is the
product of columns 7 and 5. Column 9 is the summation of 8 for
all sizes "equal to or greater than the indicated size." Column
10 is column 5 times column 7 divided by column 6.
STEP 4
Plot cumulative concentration vs. size for each test con-
ditions. For the sample data set of Table D5 this would be the
concentrations in Table D7 column 9 plotted against the sizes
in column 4. No errors bars are used.
STEP 5
Plot ANg/ALogD with upper and lower 90% confidence limits
for each test condition.
296
-------
TABLE D6
EAA (Model 3030) Data Reduction Form
Concentration, Cumulative Concentration, and AN /ALogD
From Average a For Condition s
Nt = 7xlOb
1
Channel
No.
10
3
4
5
6
7
8
9
10
11
2
Collector
Voltage
196
593
1220
2183
3515
5387
7152
8642
9647
3
D , ym
0.0100
0.0178
0.026
0.036
0.070
0.120
0.185
0.260
0.360
4
Dpi, ym
0.0133
0.0215
0.0306
0.0502
0.0917
0.149
0.219
0.306
5
AN/A I
4. 76x10 5
2.33xl05
1.47xl05
8.33x10"
4.26x10"
2.47x10"
1.56x10"
1.10x10"
6789 10
AlogD a AN ZAN AN /AlogD
p s s s
0.250
0.165
0.141
0.289
0.234
0.188
0.148
0.141
-------
TABLE D7
EAA (Model 3030) Data Reduction Form
Concentration, Cumulative Concentration, and AN /ALogD
From Average Al For Condition Inlet
(Sample Calculation)
Nt = 7xl06
10
Channel
No.
NJ
vo
00
3
4
5
6
7
8
9
10
11
Collector
Voltage
196
593
1220
2183
3515
5387
7152
8642
9647
D , ym
0.0100
0.0178
0.026
0.036
0.070
0.120
0.185
0.260
0.360
Dpi, ym
0.0133
0.0215
0.0306
0.0502
0.0917
0.149
0.219
0.306
AN/AI
4. 76x10 5
2.33xlOs
1.47xlOs
8.33x10"
4.26x10"
2.47x10"
1.56x10"
1.10x10"
A log D a
0.250
0.165
0.141
0.289
0.234
0.183
0.148
0.141
ANg
xlO6
33.2+. 6
53.3+. 7
74.3+. 8
219.8+.8
174+2
114+2
35.4+. 6
21.2+.3
15.
12.
10.
18.
7.
2.
•
S+.3
4+.2
~
9+.1
3+.1
41+. 09
82+. 05
552+. 009
233+. 003
EAN
s
xlO6
68. 4
52.6
40.2
29.3
11.0
3. 61
.785
.233
AN
s
xlO
63.
75.
77.
63.
31.
15.
3.
1.
/ALogD
6
2+1.1
3+1.0
5+.S
4+.2
~
7+.4
0+.3
73+. 06
~
65+. 02
-------
For the sample data set of Table D5 this would be the con-
centrations in Table D7 , column 10 plotted against the sizes in
column 4. The upper error bar is the value plus the 90% confi-
dence interval. The lower error bar is the value minus the 90%
confidence interval. For a3_4 in Table D7 we would have <*3_4
= 33.2 ± 0.6
thus:
AM /AT rv 33.2 x 4.76 x 10 0.6 x 4.76 x 10
ANs/ALogD -- Q
= (63.2 ± 1.1) x 106
The data shown in column 10, Table D7 is graphically displayed
in Figure Dll.
D.4 OPTICAL PARTICLE COUNTERS
Optical particle counters are a useful adjunct to the ultra-
fine sizing system, giving additional data in the region where
impactors, diffusional and electrical methods have the worst
resolution.
A number of commercial optical particle counters are avail-
able which will perform adequately in the field environment associ-
ated with precipitator evaluation. These instruments can be ob-
tained with several optical sensor configurations and may usually
be specified to perform in a threshold mode, where all particles
larger than a certain size are counted, or in a window mode where
a narrow range of particle sizes is indicated. In either instance,
the smallest particle which will be detected is approximately
0.3 pm diameter. In this respect, most optical counters size
particles in the same range as cascade impactors and cyclones.
299
-------
102
(0
o
X
CO
5
2
Q
d
2
O
t-
<
cc
LU
o
2
O
O
CC
111
CO
2
LU
O
HYPOTHETICAL DATA
• PARALLEL PLATE DIFFUSION BATTERIES
O ELECTRICAL AEROSOL ANALYSER
9
10°
«
O
10
-1
10'2
10
1-1
10°
PARTICLE DIAMETER, micrometers
10'
3630-044
Figure D11. Hypothetical inlet size distribution at a fabric filter
on a coal-fired boiler. Sample data for parallel plate
diffusion batteries and electrical aerosol analyser are
shown.
300
-------
The calibration of optical particle counters is usually done
with monodisperse plastic spheres which have a refractive index
of 1.59. Any deviations from this index of refraction, or spheri-
city of the particles will cause the indicated sizes to be in
error. If high accuracy in the optical particle counter data
is important to the test, it is possible to do calibrations in
the field, using the test aerosol. McCain (1974) has used a sedi-
mentation technique which employs parallel plate diffusion bat-
teries as sedimentation chambers to correlate aerodynamic par-
ticle diameter with the indicated, or equivalent PSL, diameter.
Marple and Rubow (1976) reported the development of special im-
paction devices which may be used for this same purpose.
In the majority of full sampling situations, optical par-
ticle counters must be used with some type of Sample Extraction
and Dilution System as described in Section D.I. The sample is
thus cooled and diluted to less than 300 particles per cm3. Also
it is usually necessary to place the return line back into the
dilution system to minimize pressure gradients across the counter
system.
Most commercially available systems come with real-time
analog outputs and digital outputs with one to 10 minute accumu-
lation or integration times. Analog signals are particularly
useful for monitoring source variations and rapping emissions.
Data accumulated by optical particle counters are given as
particles per unit volume and, after multiplication by the proper
dilution factor, may be plotted on a cumulative or differential
number basis. See paragraph C.8 for a discussion of data plot-
ting. Figure 34 in Section 2.3.4. shows data taken using an
optical-diffusional impactor system.
301
-------
APPENDIX E
SUMMARY OF SOURCE PERFORMANCE REFERENCE METHODS
To evaluate the performance standards for new stationary
sources, the EPA has promulgated reference methods which specify
the manner in which certain tests must be performed. These Ref-
erence Methods can be found in the Code of Federal Regulations
under Title 40 - Protection of Environment; Chapter 1 - Environ-
mental Protection Agency; Subchapter C - Air Programs; Part 60 -
Standards of Performance for New Stationary Sources; Appendix
A - Reference Methods.
In the first section of this appendix these reference methods
are summarized. In the latter part of this appendix these ref-
erence methods, as presented in the Code of Federal Regulations,
are reproduced verbatim. They have been corrected and amended
through April 1, 1978.
302
-------
Title 4O—Protection of Environment
CHAPTER I—ENVIRONMENTAL
PROTECTION AGENCY
[PRL 754-5]
PART 60—STANDARDS OF PERFORM-
ANCE FOR NEW STATIONARY SOURCES
Revision to Reference Methods 1-8
AGENCY: Environmental Protection
Agency.
ACTION: Final Rule.
SUMMARY: This rule revises Reference
Methods 1 through 8, the detailed re-
quirements used to measure emissions
from affected facilities to determine
whether they are in compliance with a
standard of performance. The methods
were' originally promulgated December
23. 1971, and since that time several re-
visions became apparent which would
clarify, correct and improve the meth-
ods. These revisions make the methods
easier to use, and improve their accuracy
and reliability.
EFFECTIVE DATE:-September 19, 1977.
ADDRESSES: Copies of the comment
letters are'available for public inspection
and copying at the U.S. Environmental
Protection Agency, Public Information
Reference Unit (EPA Library), Room
2922, 401 M Street, S.W.. Washington,
D.C. 20460. A summary of the comments
and EPA's responses may be obtained
upon written request from the EPA Pub-
lic Information Center (PM-215). 401
M Street, S.W., Washington, D.C. 20460
(specify "Public Comment Summary:
Revisions to Reference Methods 1-8 In
Appendix A of Standards of Performance
for New Stationary Sources").
FOR FURTHEK INFORMATION CON-
TACT:
Don R. Goodwin, Emission Standards
and Engineering Division, Environ-
mental Protection Agency, Research
Triangle Park, North Carolina 27711,
telephone No. 919-541-5271.
SUPPLEMENTARY INFORMATION:
The amendments were proposed on June
8.1976 (40 FR 23060). A total of 55 com-
ment letters were receivedN during the
comment period—34 from Industry. 15
from governmental agencies, and 6 from
other interested parties. They contained
numerous suggestions which were incor-
porated in the final revisions.
Changes common' to all eight of the
reference methods are: (1) the clarifica-
tion of procedures and equipment spec-
ifications resulting from the comments,
(2) the addition of guidelines for al-
ternative procedures and equipment to
make prior approval of the Administra-
tor unnecessary and (3) the addition of
an introduction to each reference meth-
od discussing the general use of the
method and delineating the procedure
for using alternative methods and equip-
ment.
Specific changes to the methods are:
METHOD 1
1. The provision for the use of more
than two traverse diameters, when spec-
ified by ttxe Administrator, has been
deleted. If one traverse diameter Is in a
plane containing the greatest expected
concentration variation, the intended
purpose of the deleted paragraph will be
fulfilled.
2. Based on recent data from Fluldyne
(Particulate Sampling Strategies for
Large Power Plants Including Nonuni-
form Flow. EPA-600/2-76-170, June
1976) and Entropy Environmentalists
(Determination of the Optimum Number
of Traverse Points: An Analysis of
Method 1 Criteria (draft), Contract No.
68-01-3172), the number of traverse
points for velocity measurements has
been reduced and the 2:1 length to width
ratio requirement for cross-sectional lay-
out of rectangular ducts has been re-
placed by a "balanced matrix" scheme.
3. Guidelines for sampling in stacks
containing cyclonic flow and stacks
smaller than about 0.31 meter in diam-
eter or 0.071 m* hi cross-sectional area
will be published at a later date.
4. Clarification has been made as to
when a check for cyclonic flow is neces-
sary; also, the suggested procedure for
determination of unacceptable Sow con-
ditions has been revised.
METHOD 2
1. The calibration of certain pitot tubes
has been made optional; Appropriate con-
struction and application guidelines have
been Included.
2. A detailed calibration procedure for
temperature gauges has been Included.
3. A leak check procedure for pitot
lines has been Included.
METHOD 3
1. The applicability of the method has
303
-------
been confined to fossil-fuel combustion
processes and to other processes where It
has been determined that components
other than d, CO2. CO,'and N3 are not
present in concentrations sufficient to
affect the final results.
2. Based on recent research informa-
tion (Participate Sampling Strategies for
Large Power Plants Including Nonuni-
form Flow, EPA-600/2-7'i-170, June
1976), the requirement for proportional
sampling has been dropped and replaced
with the requirement for constant rate
sampling. Proportional and constant rate
sampling have been found to give essen-
tially the same result.
3. The "three consecutive" require-
ment has been replaced by "iny three"
for the determination of molecular
weight, CO, and O2.
4. The equation for excess s,ir has been
revised to account for the presence of CO.
5. A clearer distinction has t*sen made
between molecular weight determination
and emission rate correction factor
determination.
8. Single point, integrated sampling
has been included.
METHOD 4
1. The sampling time of 1 hour has
been changed to a total sampling time
which will span the length of time the
pollutant emission rate Js b°ing deter-
mined or such time as specified in an
applicable subpart of the standards.
2. The requirement for proportional
sampling has been dropped and replaced
with the requirement fnr constant rate
sampUncr.
3. The leak cheek before the test run
has bsen made optional; the leak check
after the run remains mandatory.
METHOD 5
1. The following alternatives have
been included in the method:
a. The use of metal probe liners.
b. The use of other ropterials of con-
struction for filter holders and probe
liner parts.
c. The use of polyethylene wash bot-
tles and sample storag3 containers.
d. The use of r'.-siccants other than
silica gel or calcium sulfate, when
appropriate.
e. The ujse of stopw'< iprease other
than silicone grease, when appropriate.
f. The drying of filters and probe-filter
catches at elevated temperf-tures, when
appropriate.
g. The combining of the filter and
probe washes into one container.
2. The leak check prior to a test run
has been made optional. The post-test
leak check remains mandatory. A meth-
od for correcting sample volume for ex-
cessive leakaee rates has been included.
3. Detailed leak check and calibration
procedures for .the meter.'.ng system have
been included.
METHOD R
1. Possible interfering agents of the
method have be°n delineated.
2. The options of: (a) using a Method
8 impinger rytem, ir (b) determining
SOi pimultaceously nH.tai pprticulate
matter, have been included in the
method.
3. Based on recent research data, the
requirement for proportional sampling
has been dropped and replaced with the
requirement for constant rate sampling.
4. Tests have shown that isopropanol
obtained from commercial sources oc-
casionally has peroxide impurities that
will cause erroneously low SO. measure-
ments. Therefore, a test for detecting
peroxides in isopropanol has been in-
cluded in the method.
5. The leak check before the test run
has been made ootional; the leak check
after the run remains mandatory.
6. A detailed calibration procedure for
the metering system has been included
in the method.
METHOD 7
1. For variable wave length spectro-
photometers, a scanning procedure for
determining the point of maximum ab-
sorbancc has been incorporated as an
option.
METHOD 8
1. Known interfering compounds have
been listed to avoid misapplication of
the Ihethod.
2. The determination of filterable
particulate matter (including acid mist)
simultaneously with SO3 and SO., has
been allowed whore applicable.
3. Since occasionally some commer-
cially available quantities of isopropanol
have peroxide impurities that wffl cause
erroneously high sitlfurle acid mist meas-
urements, a test for peroxides in Isopro-
panol hss been included in the method.
4. The gravimetric technique for mois-
ture content (rather than volumetric)
has been specified because a mixture of
isopropyl alcohol and water will have a
volume less than the sura of the volumes
of its content.
5. A closer correspondence has been
made between similar parts of Methods
8 and 5.
MISCELLANEOUS
Several commenters questioned the
meaning of the term "subjert to the ap-
proval of the Administrator' in relation
to using alternate test methods and pro-
cedures. As defined in § P0.2 of subpart
A, the "Administrator" includes any au-
thorized representative of the Adminis-
trator of the Environmental Protection
Agency. Authorized representatives are
EPA officials in EPA Regional Offices or
State, local, and regional governmental
officials who have been delegated the re-
sponsibility of enforcing regulations un-
der 40 CFR 60. These officials m consulta-
tion with other staff members familiar
with technical aspects of source testing
will render decisions regarding accept-
aWe alternate test procedures.
In accordance with section 117 of the
Act, publication of these methods was
preceded by consultation with appropri-
ate advisory co-oimitt^cs, Independent
experts, and Federal departments and
agencies.
(Sees. HI, 114 and SOI (a) or the Clean All
Act, BKC. 4(8.) e>f Pi'b. L. No. 91-6O4, 84 Stet
304
-------
1683; sec. *Ca) of Pub. L. No. 91-6O4, 84 Stat.
1687; sec. 2 of Pub. L. No. 90-148. 81 Stat. 504
|42 U.S.C.' 18670-6, 1857C-9, 1857g(a) ].)
NOTE.—The Environmental Protection
Agency has determine^ that this document
does not contain a major proposal requiring
preparation of an Economic Impact Analysis
under Executive Orders 11821 and 11949 and
OMB Circular A-107.
Dated: August 10, 1977.
DOUGLAS M. COSTLE,
Administrator.
Part 60 of Chapter I of Title 40 of ttie
Code of Federal Regulations is amended
by revising Methods 1 through 8 of Ap-
pendix A—Reference Methods as
follows:
APPENDIX A—JltycitENCE METHODS
The reference methods in this appendix are referred to
in § 60.8 (Performance Tests) and J 60.11 (Compliance
With Standards and Maintenance Requirements) of 40
CFB Part 60, Subparf A (General Provisions). Specific
uses of these reference methods are described m the
standards of performance contained in the subparts,
beginning with Subpart D.
Within each standard of performance, a section titled
"Test Methods and Procedures" is provided to (1)
identify the test methods applicable to the facility
subject to the respective standard and (2) identify any
special instructions or conditions to be followed when
applying a method to the respective facility. Such in-
structions (for example, establish'sampling rates vol-
umes, or temperatures) are to be used either in addition
to, or as a substitute for procedures in a reference method.
Similarly, for sources subject to emission monitoring
requirements, specific instructions pertaining to any use
of a referenre method are provided in the snhpart or in
Appendix B.
Inclusion of methods In this appendix is not intended
as an endorsement or denial of their applicability to
wrarces tliat are not subject to standards of performance.
The methods are ppten tially applicable to other sources;
however, applicability should be confirmed by careful
ana appropriate evaluation ol the conditions prevalent
at such soiirrcs.
The approach followed in the formulation of the ref-
erence methods Involves spivifieations for equipment,
procedures, and performinii-o. lit concept, a performance
^pecitii at ion approach wuuld be preferable in all methods
because iliis allows the greatest flexibility to the user.
In practice, however, this approach is impractical in most
t as^s berjiu^o performance specilit-ations cannot lie
establislu-d. Most of mo methods described herein,
tlvrefore, involve specihe equipment Ppefifk-ations and
pwduivs and only :i fi w mcihuds ill tl.is appendix rely
tin pcrfoi'iiiiinu'
-------
METHOD 1
Sample and Velocity Traverses for Stationary Sources;
Procedure for selecting stack sampling site and selecting and
locating minimum number of traverse points.
Federal Register
Vol. No. Date Page
36 247 12/23/71 24882
38 99 5/23/73 13562
41 111 6/08/76 23061
42 160 8/18/77 41755
Regulation Promulgated
Regulation Amended
Proposed Revised Regulation
Revised Regulation Promulgated
METHOD 2
Determination of Stack Gas and Volumetric Flowrate;
Procedure for determining stack gas velocity from gas density and
velocity head using Type S pitot tube. Volumetric flowrate is
calculated from gas velocity and stack cross-sectional area.
Federal Register
Vol. No. Date Page
36 247 12/23/71 24886
41 111 6/08/76 23063
42 160 8/18/77 41758
Regulation Promulgated
Proposed Revised Regulation
Revised Regulation Promulgated
306
-------
METHOD 3
Gas Analysis for Carbon Dioxide/ Excess Air, and Dry Molecular
Weight;
C02/ CO, and 02 are determined by absorbing in appropriate reagents
(Orsat analysis).
Federal Register
Vol. No. Date Page
36 247 12/23/71 24886
41 111 6/08/76 23069
42 160 8/18/77 41768
Regulation Promulgated
Proposed Revised Regulation
Revised Regulation Promulgated
METHOD 4
Determination of Moisture in Stack Gases;
Stack gas moisture content is determined by condensation and
measuring condensed water volumetrically.
Federal Register
Vol. No. Date Page
36 247 12/23/71 24887
41 111 6/08/76 23072
42 160 8/18/77 41771
Regulation Promulgated
Proposed Revised Regulation
Revised Regulation Promulgated
METHOD 5
Determination of Particulate Emissions from Stationary Sources;
Stack gas is sampled isokinetically, particulate matter filtered
and weighed after removal of uncombined water.
307
-------
Regulation Promulgated
Regulation Amended
Proposed Revised Regulation
Revised Regulation Promulgated
Federal Register
Vol. No. Date Page
36 247 12/23/71 24888
38 99 5/23/73 13563
41 111 6/08/76 23076
42 160 8/18/77 41776
METHOD 6
Determination of Sulfur Dioxide Emissions from Stationary Sources:
S02 is separated from any SO3 and HaSO^ mist present using aqueous
isopropyl alcohol. The S02 is passed through dilute hydrogen per-
oxide which is titrated by the barium-thoris method to determine
sulfate formation.
Federal Register
Vol. No. Date Page
36 247 12/23/71 24890
41 111 6/08/76 23083
42 160 8/18/77 41782
Regulation Promulgated
Proposed Revised Regulation
Revised Regulation Promulgated
METHOD 7
Determination of Nitrogen Oxide Emissions from Stationary Sources;
A grab sample is drawn into an evacuated flask containing actified
hydrogen peroxide which converts NO and N02 to HN03. Analysis is
made by the colorimetrie PDS method.
Federal Register
Vol. No. Date Page
Regulation Promulgated 36 247 12/23/71 24891
Proposed Revised Regulation 41 111 6/08/76 23085
Revised Regulation Promulgated 42 160 8/18/77 41784
308
-------
METHOD 8
Determination of Sulfuric Acid Mist and Sulfur Dioxide Emissions
from Stationary Sources;
S03 and H2SOi, mist are absorbed in aqueous isopropyl alcohol and
S02 passes through to be collected in hydrogen peroxide solution.
Each solution is titrated by the barium-thoris method to determine
sulfate formation.
Federal Register
Vol. No. Date Page
36 247 12/23/71 24893
41 111 6/08/76 23087
42 160 8/18/77 41786
Regulation Promulgated
Proposed Revised Regulation
Revised Regulation Promulgated
METHOD 9
Visual Determination of the Opacity of Emissions from Stationary
Sources:
Opacity of emissions from a stationary source is determined visually
by qualified observers.
Federal Register
Vol. No. Date Page
36 247 12/23/71 24895
39 177 9/11/74 32857
39 219 11/12/74 39874
Regulation Promulgated
Proposed Revised Regulation
Revised Regulation Promulgated
METHOD 10
Determination of Carbon Monoxide Emissions from Stationary Sources:
Integrated or continuous gas samples from stacks are analysed for
CO content using NDIR analysis.
Federal Register
Vol. No. Date Page
Regulation Promulgated 39 47 3/08/74 9319
Regulation Amended 39 75 4/07/74 13776
309
-------
METHOD 11
Determination of Hydrogen Sulfide Emissions from Stationary Sources^
H2S is collected in alkaline Cd(OH)2 to form CdS, which is dissolved
in HCL and reacted with a known amount of iodine. Amount of iodine
consumed is measure of H2S present in gas.
Federal Register
Vol. No. Date Page
Regulation Promulgated 39 47 3/08/74 9321
Regulation Amended 39 75 4/17/74 13776
Proposed Revised Regulation 42 99 5/23/77 26222
METHOD 12
Determination of Sulfur Dioxide Emissions from Stationary Sources
by Continuous Monitors;
No specific continuous monitor required; results must show acceptable
relationship to those determined by Method 6 or 8.
(RESERVED)
METHOD 13A
Determination of Total Fluoride Emissions from Stationary Sources -
SPADNS Zirconium Lake Method;
Gaseous and particulate fluorides are withdrawn isokinetically and
concentration determined by SPADNS method.
Federal Register
Vol. No. Date Page
Regulation Promulgated 40 152 8/06/76 33157
Regulation Amended 41 230 11/29/76 52299
310
-------
METHOD 13B
Determination of Total Fluoride Emissions from Stationary Sources -
Specific Ion Electrode Method;
Gaseous and particulate fluorides are withdrawn isokinetically and
concentration determined by a specific ion electrode.
Federal Register
Vol. No. Date Page
Regulation Promulgated 40 152 8/06/75 33163
Regulation Amended 41 230 11/29/76 52299
METHOD 14
Determination of Fluoride Emissions from Pot Room Roof Monitors
of Primary Aluminum Plants;
A permanent sampling manifold is constructed to isokinetically
withdraw sample gas from the roof monitor. The gas sample is
brought to ground level and sampled using Method 13A or 13B for
determination of fluoride concentration.
Federal Register
Vol. No. Date Page
Regulation Promulgated 41 17 1/26/76 3829
METHOD 15
Determination of EzS, COS, and CS2 Emissions from Stationary Sources;
A gas sample is extracted and diluted with clean dry air. An aliquot
of the diluted sample is then analysed by gas chromatographic
separation (GC) and flame photometric detection (FPD).
Federal Register
Vol. No. Date Page
Proposed Regulation 41 193 10/04/76 43870
Regulation Promulgated 43 51 3/15/78 10870
311
-------
METHOD 16
Semicontinuous Determination of Sulfur Emissions from Stationary
Sources;
A gas sample is extracted and diluted with clean dry air. An
aliquot of the diluted sample is then analysed for gas sulfur.
Two GC/FPD analysers are used for resolution of both high and
low molecular weight sulfur compounds.
Federal Register
Vol.
Proposed Regulation 41
Regulation Promultaged 43
METHOD 17
Determination of Particulate Emissions
No. Date Page
187 9/24/76 42017
37 2/23/78 7575
from Stationary Sources
(In-Stack Filtration Method) :
Particulate matter is withdrawn isokinetically and collected on a
filter maintained at stack temperature. Particulate matter mass
is determined gravimetrically after removal of uncombined H20.
Federal Register
Vol. No. Date Page
Proposed Regulation 41 187 9/24/76 42020
Regulation Promulgated 43 37 2/23/78 7585
312
-------
METHOD 1
SAMPLE AND VELOCITY TRAVERSES FOR STATIONARY SOURCES
1 1'rinciftlc and Applicability
1.1 Principle. To aid in the representative measure-
ment of pollutant emissions and/or total volumetric! flow
rate from £ stationary source, a measurement site where
the effluent stream is flowing in a known direction is
selected, and the cross-section of the slack is divided Into
a number of equal areas. A traverse point is then located
within each of these equal areas.
1.2 Applicability. This method is applicable to flow-
ing gas streams in ducts, stacks, and flues. The method
vannot he used when: (1) flow is cyclonic or swirling (see
Section 2.4), (2) a stack is smaller than about 0.30 meter
(12 in.) in diameter, or 0.071 m> (113 in.!) in cross-sec-
tional area, or (3) the measurement site is less than two
stack or duct diameters downstream or less than a ball
diameter upstream from a flow disturbance.
The rev ujiatriam from any flow disturbance such as
a bend, expansion, or contraction in the stack, or from a
visible flame. If necessary, an alternative location may
be selected, at a position at least two stack or duct di-
ameters downstream and a half diameter upstream from
eny flow rusturb&ucc. For a rectangular cross section,
an efjUivuIt!ni diameter (7).) shall be calculated from (he
following equation, to determine the upstrt-am and
dowiis;r>Hi>! distances:
_ZL\V
'~L+W
whereT=length and H'=width.
2.2 Determining the Number of Traverse Points.
2.2.1 Particulate Traverses. When the eight- and
two-diameter criterion can be met, the zninimumnumber
of traverse points shall be: (1) twelve, for circular or
rectangular stacks with diameters (or equivalent di-
ameters) greater than 0.61 meter (24 in.); (2) eight, for
circular stacks with diameters between 0.30 and 0.61
meter (12-24 in.); (3) nine, for rectangular stacks with
equivalent diameters between 0.30 and 0.61 meter (12-24
in.).
When the eight- and two-diameter criterion cannot be
met, the minimum number of traverse points is deter-
mined from Figure 1-1. Before referring to the figure,
however, determine the distances from the chosen meas-
urement site to the nearest upstreatn and downstream
disturbances, and divide each distance by the stack
diameter or equivalent diameter, to determine the
distance in terms of the number of duct diameters. Then.
determine from Figure 1-1 the K^Mmyi™ number of
traverse points that corresponds: (1) to the number of
duet diameters upstream; and (2) to the number of
diameters downstream. Select the higher of the two
minimum numbers of traverse points, or a greater value,
so that for circular stacks the number is a multiple of 4,
and for rectangular stacks, the number Is one of those
shown in Table 1-1.
TtnLE 1-1. Cr la
essentially parallel to the stack walls. However.
cyclonic now may exist 0) after such devices as cyclones
and Inertia! denusters following venturi scrubbers, or
P) In stacks having tangential Inlets or other duct con-
figurations which tend to Induce swirling; in these
instances, the presence or absence of cyclonic now at
the sampling location must be determined. The following
techniques are acceptable for this determination.
1
1
1 1__.
' . I
1
1
Figure 1-4. Example showing rectangular stack cross
section divided into 12 equal areas, with a traverse
point at centroid of each area.
313
-------
U)
50
£40
30
0.5
O
oe
I 20
i 10
DUCT DIAMETERS UPSTREAM FROM FLOW DISTURBANCE (DISTANCE A)
1.0 1.5 2.0
I
I
I
2.5
* FROM POINT OF ANY TYPE OF
DISTURBANCE {BEND, EXPANSION, CONTRACTION, ETC.)
I
I
T
\
A
-
1
-
3
I
"
1
^
'DISTURBANCE
MEASUREMENT
r- SITE
DISTURBANCE
I
1
3456789
DUCT DIAMETERS DOWNSTREAM FROM FLOW DISTURBANCE (DISTANCE B)
Figure 1-1. Minimum number of traverse points for particulate traverses.
10
-------
U)
50
0.5
DUCT DIAMETERS UPSTREAM FROM FLOW DISTURBANCE (DISTANCE A)
1.0 1.5 2.0
t/i
O
a.
40
30
20
5 10
I
I
I
I
2.5
I
\
T
A
~
1
i
\
i
••^H
4s
'DISTURBANCE
MEASUREMENT
£-' SITE
DISTURBANCE
I
1
1
I
I
3456789
DUCT DIAMETERS DOWNSTREAM FROM FLOW DISTURBANCE (DISTANCE R)
10
Figure 1-2. Minimum number of traverse points for velocity (nonparticulate) traverses.
-------
Level and eero the manometer. Connect a Type 8
pilot tnbe to the manometer. Position the Type 8 pilot
tube at each traverse point, in succession, so that the
planes of the lace openings of the pilot tube are perpendic-
ular to the stack cross-sectional plane: when the Type 8
pilot tnbe is in this position, it is at "0° reference." Note
the differential pressure (Ap) reading at each traverse
point. II a null (zero) pilot reading is obtained M. V
reference at a given traverse point, an acceptable flow
and record the value of the rotation angle (or) to the
nearest degree. After the null technique has been applied
at each travrse point, calculate the average of the abso-
lute values of
-------
Table 1-2. 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
T2J
13
14
15
16
37
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
?3.3
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
317
-------
METHOD 2
DETERMINATION OF STACK GAS AND VOLUMETRIC FLOWRATE
I. Prindplt and ApfilaMlUj
1.1 Principle- The average gas velocity in a stack is
determined from the gas density and from measurement
o[ the average velocity head with a Type S (Stausscheibe
or reverse type) pilot tube.
1.2 Applicability. This method is applicable for
measurement of the average velocity ol a gas stream and
for Quantifying gas flow.
This procedure is not applicable at measurement sites
which fail to meet the criteria of Method 1, Section 2.1.
Also, the method cannot be used for direct measurement
in cyclonic or swirling gas streams; Section 2.4 of Method
1 shows how to determine cyclonic or swirling flow con-
ditions. When unacceptable conditions exist, alternative
procedures, subject to the approval of the Administrator,
U.S. Environmental Protection Agency, must be em-
ployed to make accurate flow rate determinations:
examples of such alternative procedures are: (1) to install
straightening vanes; (2) to calculate the total volumetric
flow rate stoichiomotrically, or (3) to move to another
measurement site at which the flow is acceptable.
2. Ajifaratm
Specifications for'the apparatus are given below. Any
other apparatus that has been demonstrated (subject to
approval of the Administrator) to be capable of meeting
the specifications will be considered acceptable.
2.1 Type 8 Pltot Tube. The Type 8 pilot tabe
igure 3-1) shall be made of metal tubing , B standard
pitot tube is used as a reference. The standard pitot
tube shall, preferably, have a known coefficient, obtained
either (1) directly from the National Bureau of Stand-
ards, Route 270, Quince Orchard Road, Qaithersburg,
Maryland, or (2) by calibration against another standard
pitot tube with an N US-traceable coefficient. Alter-
natively, a standard pitot tube designed according to
the criteria given in 2.7.1 through 2.7.5 below and illus-
trated In Figure 2-4 (see also Citations 7. 8, and 17 In
318
-------
1.90-2.54 cm*
(0.75 -1.0 in.)
LSJ.
"• *"—* CHEb!9S3X^OI^9^
I I 7.62 cm (3 in.)*
TEMPERATURE SENSOR
LEAK-FREE
CONNECTIONS
MANOMETER
•SUGGESTED (INTERFERENCE FREE)
PITOT TUBE • THERMOCOUPLE SPACING
Figure 2-1. Type S pitot tube manometer assembly.
319
-------
TRANSVERSE
TUBE AXIS
\
FACE
OPENING
PLANES
(a)
1
4
LONGITUDINAL
TUBE AXIS
' Dt
A SI DEPLANE
1
" y / * „
A _y PA
* " R "X.
* B ,-,. \ PR
- I
' p, (
NOTE:
1.050t
-------
TRANSVERSE
TUBE AXIS "
LONGITUDINAL
TUBE AXIS—
(e)
-t
(f)
(9)
Figure 2-3. Types of face-opening misalignment that can result from field use or im-
proper construction of Type S pitot tubes. These will not affect the baseline value
of.Cp(s) so long as 01 and 02 < 10°, 01 and 02 < 5°. z < 0.32 cm (1/8 in.) and w <
0.08 cm (1/32 in.) (citation 11 in Section 6).
321
-------
&
1
"SS»
T
CURVED OR
MITERED JUNCTION
HEMISPHERICAL
TIP
STATIC
HOLES
Figure 2-4. • Standard pitot tube design specifications.
322
-------
Section 6) may be used. Pilot tubes Unsigned arrnrdini;
to these specifications will bave baseline eocllieicnts oi
ibout0.90±0.01.
2.7.1 Heniisijlicrlciil (shown la Figure 2-4), eUlu.soiila!,
IT conical tip.
2.7.2 A mijilmum of six diameters straight, run (bos.'d
ipon D, tlie external diameter of the luhe) be.tv.ve.ii tlio
;ip and the static pressure holes.
2.7.3 A minimum of eight diameters slruinlit rim
between the static pressure holes and the ci'iite.rlim: of
the external tube, following the 90 degree bend.
2.7.4 Static pressure holes of equal size. (approximately
0.1 1)}, et|ii;illy spaced in a piezometer ring ennliKuration.
.2.7.0 Ninety degree bend, with curvud ,r mitered
jiinrlioi1..
2.8 DilTerent-iul Pressure Gauge for Typo 8 Pilot
Tube Calibration. An inclined manometer or equivalent
is used. If the single-velocity calibration technujue is
employed (see Section 4.1.2.3), the calibration differen-
tial pressure gauge shall be readable to the nearest 0.13
mm HjO (0.005 in. IltO). For multivolocily calibrations,
the gauge shall be readable to the nearest 0.13 mm IIzO
(0.005 in HiO) for Ap values between 1.3 and 25 mm HzO
(0.05 and 1.0 in. HjO), and to the nearest 1.3 mm HiO
(0.05 in. HiO) for Ap values above 25 mm HjO (1.0 In.
HiO). A special, more sensitive gauge will be required
to read Ap values below 1.3 mm H2O [0.05 In. HiO]
(see Citation 18 In Section 6).
3. Procedure
3.1 Set up the apparatus as shown in Figure 2-1.
Capillary tubing at surge tanks installed between the
manometer and pltot tube may be used to dampen Ap
fluctuations. It is recommended, but not required, that
a pretest leak-check be conducted, as follows: (1) blow
through the pitot Impact opening until at least 7.6 cm
(3 in.) HiO velocity pressure registers on the manometer;
then, close off the impact opening. The pressure shall
remain stable for at least 15 seconds; (2) do the same for
the static pressure side, except using suction to obtain
the minimum of 7.6 cm (3 in.) HiO. Other leak-cheek
procedures, subject to the approval of the Administrator,
may be used. :
3.2 Level and zero the manometer. Because the ma
nometer level and zero may drift due to vibrations and
temperature changes, make periodic, checks during the
traverse. Record all necessary data as shown in the
example data sheet (Figure 2-5).
3.3 Measure the velocity head and temperature at the
traverse points specified by Method 1. Ensure that the
proper differential pressure gauge is being used for the
range of Ap values encountered (see Section 2.2). If it is
necessary to change to a more sensitive gauge, do so, and
remeasure the Ap and temperature readings at each tra-
verse point. Conduct a post-test leak-check (mandatory),
as described in Section 3.1 above, to validate the traverse
run.
3.4 Measure the static pressure in the stack. One
reading is usually adequate.
3.6 Determine the atmospheric pressure.
3.6 Determine the stack gas dry molecular weight.
For combustion processes or processes that emit essen-
lially COi, Oi, CO, and Nj, use Method 3. For processes
emitting essentially air, an analysis need not be con-
ducted; use a dry molecular weight of 29.0. For other
processes, other methods, subject to the approval of the
Administrator, must be used.
3.7 Obtain the moisture content from Reference
Method 4 (or equivalent) or from Method 5.
:).8 Determine the cross-sectional area of the stack
or duct at the sampling location. Whenever possible,
physically measure tho stack dimensions rather than
using blueprints.
4 Calibration
4.1 Type 8 Pitot Tube. Before its initial use, care-
hilly examine the Type 8 pitot tube in top, side, and
end views to verify that the face openings of the tube
are aligned within the specifications illustrated in Figure
2-2 or 2-3. The pltot tube shall not be used if it fails to
meet these alignment specifications.
After verifying the face opening alignment, measure
and record the following dimensions of the pltot tube:
(a) the external tubing diameter (dimension Ui, Figure
2-2b); and (b) the base-to-opening plane distances
(dimensions Pi and PB, Figure 2-2b). If Di is between
0.48 and 0.95 cm (Mo and % in.) and if PA and Pa are
equal and between 1.05 and l.SOK,, there are two possible
options: (1) the pltot tube may be calibrated according
to the procedure outlined in Sections 4.1.2 through
4.1.5 below, or (2) a baseline (Isolated tube) coefficient
value of 0.84 may be assigned to the pitot tube. Note,
however, that if the pitot tube is part of an assembly,
calibration may still be required, despite knowledge
of the baseline coefficient value (see Section 4.1.1).
If Di, PA, and PB are outside the specified limits, the
pitot tube must be calibrated as outlined In 4.1.2 through
4.1.5 below.
4.1.1 Type S Pltot Tube Assemblies. During sample
and velocity traverses, the isolated Type 8 pitot tube is
not always used; In many instances, the pitot tube is
used in combination with other source-sampling compon-
ents (thermocouple, sampling probe, nozzle) as part of
an "assembly." The presence of other sampling compo-
nents can sometimes affect the baseline value of the Type
8 pitot tube coefficient (Citation 9 in Section 6); therefore
an assigned (or otherwise known) baseline coefficient
value may or may not be valid for a given assembly The
baseline and assembly coefficient values will ho identical
only when the relative placement of the components in
the assembly is such that aerodynamic inlerferunco
effects are eliminated. Figures 2-6 through '1- b illustrate
interference-free component arrangements for Typo S
pilot tubes having external tubing diamri.Ts between
0.48 and 0.06 cm (Me and 9
-------
PLANT
DATE
.RUN NO.
STACK DIAMETER OR DIMENSIONS, m(in.)
BAROMETRIC PRESSURE, mm Hg (in. Hg)—
CROSS SECTIONAL AREA. m2(ftZ}
OPERATORS
PITOTTUBEI.D.NO.
AVG. COEFFICIENT. Cp = .
LAST DATE CALIBRATED.
SCHEMATIC OF STACK
CROSS SECTION
Traverse
Pt.No.
Vel.Hd..4i
mm (inj HzO
Stack Temperature
ts.0C(°F)
Average
Ts.0K{°R)
P9
mm Hg (in.Hg)
>fAr
Figure 2-5. Velocity traverse data.
324
-------
TYPE SPITOT TUBE
I
x £ 1.90 cm (3/4 in.) FOR On -1.3 cm (1/2 in.)
SAMPLING NOZZLE
A. BOTTOM VIEW; SNOWING MINIMUM PITOT NOZZLE SEPARATION.
SAMPLING
PROBE
SAMPLING
NOZZLE
STATIC PRESSURE
OPENING PLANE
IMPACT PRESSURE
OPENING PLANE
1—J
TYPES
PITOT TUBE
NOZZLE ENTRY
PLANE
•. SIDE VIEW; TO PREVENT PITOT TUBE
FROM INTERFERING WITH GAS FLOW
STREAMLINES APPROACHING THE
NOZZLE. THE IMPACT PRESSURE
OPENING PLANE OF THE PITOT TUBE
SHALL BE EVEN WITH OR ABOVE THE
NOZZLE ENTRY PLANE.
Figure 2-6. ^Proper pitot tube • sampling nozzle configuration to prevent
aerodynamic interference; buttonhook- type nozzle; centers of nozzle
and pitot opening aligned; Dt between 0.48 and 0.95 cm (3/16 and
3/8 in.).
325
-------
THERMOCOUPLE
W>7.i2o*
-rr
-fr
Z>i.Mcm(3/4in.)
THERMOCOUPLE
Z>S.Mcm
<2i«4
TYPE SPITOT TUBE
r
-U-
TYPE SPITOT TUBE
SAMPLE PROBE
_ SAMPLE PROBE
Figure 2-7. Proper thermocouple placement to prevent interference;
Dt between 0.48 and 0.95 cm (3/16 and 3/8 in.).
£
•^fft^r>>-
•&Mti&-
Dt
TYPE SPITOT TUBE
T~
I n;i!i in
SAMPLE PROBE
Y >7.62 em (3 in J
Figure 2-8. Minimum pitot-sample probe separatfon needed to prevent interference;
Dt between 0.48 and 0.95 cm (3/16 and 3/8 in.).
326
-------
PITOTTUBE IDENTIFICATION NUMBER:
CALIBRATED BYr.
.DATE:.
RUN NO.
1
2
3
"A" SIDE CALIBRATION
Apstd
cmHaO
(in.H20)
AP($)
cmH20
(in. H20)
Cp (SIDE A)
Cp(s)
DEVIATION
Cp(,)-Cp(A)
RUNVfr.
1
2
3
"B" SIDE CALIBRATION
Apstd
CfflH20
(in.H20)
Ap($)
emHaO
(in. H20)
Cp (SIDE B)
Cp(s)
DEVIATION
Cp(,)-Cp(B)
AVERAGE DEVIATION = 0 (A OR B)
S|Cp(s)-Cp(AORB)]
ilOE A)-Cp (SIDE B) (-4-MUST BE <0.01
Figure 2-9. Pitot tube calibration data.
•MUSTBE<0.01
327
-------
4.1.3.6 Read Ap. and enter its value in the data table.
Kemove the Type S pitot tube from lie duct and dis-
connect it from the manometer.
4.1.3.7 Repeat steps 4.1.3.3 through 4.1.3.6 above until
three pairs of Ap readings have been obtained.
4.1.3.8 Repeat steps 4.1.3.3 tlirongh 4,1.3.7 above for
the B side of the Type S pitot tube.
4.1.3.9 Perform calculations, as described in Section
4.1.4 below.
4.1.4 Calculation--;.
4.1.4.1 For each of the six pairs of Ap readings (1.6..
three from side A and three from side B) obtained in
Section 4.1.3 above, calculate the value of the Type 8
pilot tube ooellicieut as follows:
Equation 2-2
C,(.)=Typ« B pilot tube coefficient
= Standard pitot tab* coefficient; an OJ» il the
coefficient is unknown and the tube la designed
according to the criteria of Sections 2.7.1 to
2.7 Ji of this method.
= Velocity head measured by the standard pitot
tube, em HiO (in. HjO)
Ap.= Velocity head measured by the Type B pitot
tube, em HiO On. HiO)
4.1.42 Calculate C, (doe A), the mean A-sMe coef-
ficient, and
-------
EXTERNAL
SHEATH
ESTIMATED
SHEATH
BLOCKAGE
ED r ixw i
JE = [DUCT ARE AJ
x 100
Figure 2-10. Projected-area m.odels 'For typical pitot tube assemblies.
329
-------
for the English system.
3/j=Molecufar weight of stack gas, dry basis (sea
Section 3.6) g/B-mole (Ib/lb-mole).
M.=Molecular weight of stack gas, wet basis, g/g-
moje (IbAb-mole).
=Mi (1—B.J-t-18.0 B— Equation 2-5
Pb«f=Barometric pressure at measurement site, mm
Hg (in. Hg).
P,=Stack static pressure, mm Hg (in. Hg).
P,=Absolnto stack gas pressure, mm Hg (in. Hg).
=Ptm,+P, Equation 2-6
Pud=Standard absolute pressure, 760 mm Hg (29-92
in. Hg).
Q,d=Dry volumetric stack gas flow rate corrected to
standard conditions, dscm/hr (dscf/hr).
«.=Slack temperature, °C (*F).
T.=Absolute stack temperature, °K (°R).
=273+t. for metric
=460-H. for English
Equation 2-7
Equation 2-8
r«d=8tandard absolute temperature, 293 °K (528* R)
r.=Average stack gas velocity, m/sec (ft/sec).
Ap=Velocity head of stack gas, mm UiO (in. HiO).
3,600= Con version factor, sec/hr.
18.0=Molecalar weight of water, g/g-mole (Ib-lb-
molo).
5.2 Average stack gas velocity.
PM.
Equation 2-9
5.3 Average slack gas dry volumetric flow rate.
Equation 2-10
6. Bibliography
1. Mark, L. 8. Mechanical Engineers' Handbook. New
York. McGraw-Hill Book Co., Ine. 1951.
2. Perry. J. H. Chemical Engineers' Handbook. New
York. McGraw-Hill Book Co., Inc. I960.
3. Sliigeliara, H. T.. W. F. Todd, and W. 8. Smith.
Significance of Errors in Stack Sampling Measurements.
U.S. Environmental Protection Agency, Research
Triangle Park, N.C . (Presented at the Annual Meeting of
the Air Pollution Control Association, St. Louis, Mo.,
Juno 14-19, 1970.)
4. Standard Method for Sampling Stacks for Paniculate
Matter. In: 1971 Book of A8TM Standards, Fart 23.
Philadelphia, Pa. 1971. ASTM Designation D-2928-71.
S. Vcnnard, J. K. Elementary Fluid Mechanics. New
York. John Wiley and Sons. Inc. 1947.
6. Fluid Meiers—Their Theory and Application.
American Socirl y of Mechanical Engineers, New York,
N.Y. 1»».
7. AS1IRAK Hiindbook of Fundamentals. 1072. p. 208.
8. Annual Book of ASTM Standards, Part 26.1974. p.
OIK.
•J. Vollaro. R. F. Guidelines for Typo S Pilot Tube
Calibration. U.S. Environmental Protection Agency.
Research Tianglc Park, N.C. (Presented at 1st Annual
Meeting, Source Evaluation Society, Dayton, Ohio,
September 18,1. Velocity Calibration of EPA
Type Source Sampling Probe. United Technologies
Corporation, Pratt and Whitney Aircraft Division,
East Hartford, Conn. 1975.
16. Vollaro, R. F. Recommended Procedure for Sample
Traverses in Ducts Smaller than 12 Inches in Diameter.
U.S. Environmental Protection Agency, Emission
Measurement Branch, Research Triangle Park, N.C.
November 1976.
17. Ower, E. and R. C. Pankhurst. The Measurement
of Air Flow, 4th Ed., London, Pergamon Press. 1966.
18. Vollaro, R. F. A survey of Commercially Available
Instrumentation for the Measurement of Low-Range
Gas Velocities. U.S. Environmental Protection Agency.
Emission Measurement Branch, Research Triangle
Park, N.C. November 1976. (Unpublished Paper)
19. Onyp, A. W.. C. C. St. Pierre, D. 8. Smith, D.
Motion, and J. Sterner. An Experimental Investigation
of the Effect of Pitot Tube-Sampling Probe Configura-
tions on the Magnitude of the 8 Type Pitot Tube Co-
efficient for Commercially Available Source Sampling
Probes. Prepared by the University of Windsor for the
Ministry of the Environment, Toronto, Canada. Feb-
ruary 1975.
330
-------
METHOD 3
GAS ANALYSIS FOR CARBON DIOXIDE, EXCESS AIR, AND DRY MOLECULAR
WEIGHT
1. Principle and Applicability
1.1 Principle. A gas sample is oil raclcd from a slack,
by one of the following methods: (1) single-point, grab
sampling; (2) single-point, integrated sampling; or (3)
mnltl-point, Integrated sampling. The gas sample Is
analyzed for perc«nt carbon dioxide (COj), percent oxy-
gen (Os), and, If necessary, percent carbon monoxide
(CO). If a dry molecular weight determination is to be
made, either an Orsat or a Fyrite ' analyzer may be used
for the analysis; for excess air or emission rate correction
factor determination, an Orsat analyzer must bo used.
1.2 Applicability. This method is applicable for de-
termining CO; and Oz concentrations, excess air, and
dry molecular weight of a sample from a gas stream of a
fossil-fuel combustion process. The method may also be
applicable to other processes where it has been determined
that compounds other than CO>, Oi, CO, and nitrogen
(Ni) are not present in concentrations sufficient to
affect the results.
Other methods, as well as modifications to the proce-
dure described herein, are also applicable for some or ail
of the above determinations. Examples of specific meth-
ods and modifications include: (1) a multi-point sami)-
ling method using an Orsat analyzer to analyze indi-
vidual grab samples obtained at each point; (2) a method
using CO2 or O? and stoichiometric calculations to deter-
mine dry molecular weight and excess air; (3) assigning a
value of 30.0 for dry molecular weight, in lieu of actual
measurements, for processes burning natural gas, coal, or
oil. These methods and modifications may be used, but
are subject to the approval of the Administrator.
2. Apparatus
As an alternative to the sampling apparatus nnd sys-
tems described herein, other sampling systems (e.g.,
liquid displacement) may be used provided such systems
are capable of obtaining a representative sample and
maintaining a constant sampling rate, and are otherwise
capable of yielding acceptable results. Use of such
systems is subject to the approval of the Administrator.
2.1 Qrab Sampling (Figure 3-1).
2.1.1 Probe. The probe should be made of stainless
steel or borosilicate glass tubing and should be equipped
with an in-stack or out-stack filter to remove paniculate
matter (a plug of glass wool is satisfactory for this pur-
pose). Any other material inert to Oi, COi, CO, and Nj
and resistant to temperature at sampling conditions may
be used for the probe; examples of such material are
aluminum, copper, quartz glass and Teflon.
2.1.2 Pump. A one-way squeeze bulb, or equivalent,
is used to transport the gas sample to the analyzer.
2.2 Integrated Sampling (Figure 3-2).
2.2.1 Probe. A probe such as that described in Section
2.1.1 is suitable.
2.2.2 Condenser. An air-cooled or water-cooled eon-
denser, or ether condenser that will not remove Oj,
COi, CO, and Nj, may be used to remove excess moisture
which would interfere with the operation of the pump
and flow meter.
2.2.3 Valve. A needle valve is used (o adjust sample
fan flow rate.
2.2.4 Pump. A leak-free, diaphragm-type pump, or
equivalent, is used to transport sample gas to the flexible
bag. Install a small surge tank between the pump and
r:ite meter to eliminate the pulsation effect of the dia-
phragm pump on the rotametcr.
2.2.8 Rate Meter. The rotameter, or equivalent rate
nwter, used should be capable of measuring now rate
to within ±2 percent of the selected flow rat«. A flow
rate range of 500 to 1000 cm'/min is suggested.
2.2.6 Flexible Bag. Any leak-free plastic (eg,, Tedlar,
Mylar, Teflon) or plastic-coated aluminum (e.g., alnmi-
nized Mylar) bag, or equivalent, having a capacity
consistent with the selected flow rale and time length
of the test run, may be used. A capacity in the range of
55 to 90 liters is suggested.
To leak-check the bag, connect it to a water manometer
•nd pressurize the bag to 6 to 10 cm Ii?O (2 to 4 in. H«O).
Allow to stand for 10 minutes. Any displacement in the
water manometer indicates a leak. An alternative leak-
cheek method is to pressurize the bag to fi to 10 em HzO
(2 to 4 in. HiO) and allow to stand overnight. A deflated
i Mention of trade names or specific products does not
constitute endorsement by the Environmental Protec-
tion Agency.
tag indicates a leak.
2.2.7 Pressure Gauge. A water-filled U-tube manom-
eter, or equivalent, of about 28 cm (12 in.) is used for
the flexible bag leak-check.
2.2.8 Vacuum Gauge. A mercury manometer, or
equivalent, of at least 760 mm Hg (30 in. Hg) is used for
the sampling train leak-check.
2.3 Analysis. For Orsat and Fyrite analyzer main-
tenance and operation procedures, follow the instructions
recommended by the manufacturer, unless otherwise
specified herein.
2.3.1 Dry Molecular Weight Determination. An Orsat
analyzer or Fyrite type combustion gas analyzer may be
used.
2.3.2 Emission Bate Correction Factor or Excess Air
Determination. An Orsat analyzer must be used. For
low COi (less than 4.0 percent) or high O> (greater than
15.0 percent) concentrations, the measuring burette of
the Orsat must have at least 0.1 percent subdivisions.
3. Dry Molecular Weight Determination
Any of the three sampling and analytical procedures
described below may be used for determining the dry
molecular weight.
3.1 Single-Point, Grab Sampling and Analytical
Procedure.
3.1.1 The sampling point in the duct shall either be
at the centroid of the cross section or at a point no closer
to the walls than 1.00m (3.3 ft), unless otherwise specified
by the Administrator.
3.1.2 Bet up the equipment as shown in Figure 3-1,
making sure all connections ahead of the analyzer are
tight and leak-free. If an Orsat analyzer la used, it is
recommended that the analyzer be leaked-checked by
following the procedure In Section 5; however, the leak-
check is optional.
3.1.3 Place the probe in the stack, with the tip of the
probe positioned at the sampling point; purge the sampl-
ing line. Draw a sample into the analyzer and imme-
diately analyze it for percent CO) and percent Oi. Deter-
mine the percentage of the gas that is Ni and CO by
subtracting the sum of the percent COi and percent Oi
from 100 percent. Calculate the dry molecular weight as
indicated in Section 6.3.
3.1.4 Repeat the sampling, analysis, and calculation
procedures, until the dry molecular weights of any three
grab samples differ from their mean by no more than
0.3 g/g-mole (0.3 lb/lb-mole). Average these three molec-
ular weights, and report the results to the nearest
0.1 g/g-mole (Ib/lb-mole).
3.2 Single-Point, Integrated Sampling and Analytical
Procedure.
3.2.1 The sampling point in the duct shall be located
as specified in Section 3.1.1.
3.2.2 Leak-check (optional) the flexible bag as in
Reel ion. 2.2.6. Bet up the equipment as shown in Figure
:j-2. Just prior to sampling, leak-check (optional) the
train by placing a vacuum gauge at the condenser inlet,
pulling a vacuum of at least 250 mm Hg (10 in. Hg),
plugging the outlet at the quick disconnect, and then
i urning oft" the pump. The vacuum should remain stable
(oral least 0.5 minute. Evacuate the flexible bag. Connect
the probe and place it in the stack, with the tip of the
probe positioned at the sampling point; purge the sampl-
ing line. Next, connect the bag and make sure that all
connections are tight and leak free.
3.2.3 Sampl« at a constant rate. The sampling run
should be simultaneous with, and for the same total
length of lime as, the pollutant emission rate determina-
tion. Collection of at least 30 liters (1.00 ft') of sample gas
js recommended; however, smaller volumes may be
collected, if desired.
3 2.4 Obtain one integrated flue gas sample during
*•». h pollutant emission rate determination. Within s
hours after the sample is taken, analyze it for percent
COi and percent O> using either an Orsat analyzer or a
Fyrite-type combustion gas analyzer. If an Orsat ana-
lyzer is used, It is recommended that the Orsat leak-
••beck described hi Section 5 be performed before tins
determination; however, the check Is optional.'Deter-
mine the percentage of the gas that is N j and CO by sub-
tracting the sum of the percent CO: and percent Oi
331
-------
PROBE
FLEXIBLE TUBING
\
FILTER (GLASS WOOL)
TO ANALYZER
SQUEEZE BULB
Figure 3-1. Grab-sampling train.
RATE METER
AIR-COOLED
CONDENSER
PROBE
\
^H
FILTER
(GLASS WOOL)
QUICK DISCONNECT
n
RIGID CONTAINER
Figure 3-2. Integrated gas-sampling train.
332
-------
from 100 percent. Calculate the dry molecular weight as
indicated in Section 6.3.
1A5 Repeat the analysis and calculation procedures
until the individual dry molecular weights for any three
analyses differ from their mean by no more than 0 3
g/g-mole (0.3 Ib/lb-mole). Average these three molecular
weights, and report the results to the nearest 0.1 g/g-mole
(0.1 Ib/lb-mole).
3.3 Multi-Point, Integrated Sampling mid Analytical
Procedure.
8.3.1 Unless otherwise specified by the Adminis-
trator, a minimum of eight traverse points shall be used
for circular stacks having diameters less then 0.61 m
(24 In.), a minimum of nine shall be used for rectangular
stacks having equivalent diameters less than 0.61 m
(24 In.), and a minimum of twelve traverse points shall
be used for all other cases. The traverse points shall be
located according to Method 1. The use of fewer points
is subject to approval of the Administrator.
3.3.2 Follow the procedures outlined in Sections 3.3.2
through 3.2.5, except for the following: traverse all sam-
pling points and sample at each point for an equal length
of time. Itccord sampling dai a as shown in Figure 3-3.
4. Emission Hate Cuntctiiin Factor or 1'jcta Air Dii'.i-
mmalioii
NOTE.—A Fyrite-type combustion gas analyzer is not.
acceptable for excess air or emission rate correction factor
determination, unless approved by the Administrator.
If both percent COi and percent Oj are measured, the
analytical results of any of the three procedures given
below may also be used for calculating the dry molecular
weight.
Each of the three procedures below shall be u<=<-J only
when specified in an applicable subpart of the standards.
The use of these procedures for other purposes must i,ave
specific prior approval of the Administrator.
4.1 Single-Point, Grab Sampling and Anal>(i'-al
Procedure.
4.1.1 The sampling point in the duct shall eith.-r no
at the centroid of the cross-sei-tion or at a point no closer
to the walls than 1.00m (3.3 ft), unless otherwise siKxini-d
by the Administrator.
4.1.2 Set up the equipment as shown in Figure 31,
making sure all connections ahead of the analyzer are
tight and leak-free. Leak-check the Orsat analyzer ac-
cording to the procedure dea-ribed m Scctiou 5. This
leak-check is mandatory.
4.1.3 Place the probe in the stack, with the tip of the
probe positioned at the sampling point; purge the sam-
pling line. Draw a sample into the analyier. For emission
rate correction factor determination, immediately ana-
lyze the sample, as outlined in Sections 4.1.4 and 4.1.5,
for percent CO] or percent Oi. If excess ah- is desired,
proceed as follows: (1) immediately analyze the sample,
as in Sections 4.1.4 and 4.1.5, for percent COi Of, and
CO; (2) determine the percentage of the gas that Is Ni
by subtracting the sum of the percent COi, percent Oi,
and percent CO from 100 percent; and (3) calculate
percent excess air as outlined in Section 6.2.
4.1.4 To ensure complete absorption of the COi, Oi,
or II applicable, CO, mate repeated passes through each
absorbing solution until two consecutive readings are
the same. Several passes (three or four) should be made
between readings. (If constant readings cannot be
obtained after three consecutive readings, replace the
absorbing solution.)
4.1.6 After the analysis is completed, leak-check
(mandatory) the Orsat analyzer once again, as described
in Section 6. For the results of the analysis to be valid,
the Orsat analyzer must pass this leak test before and
after the analysis. NOTE.—Since this single-point, grab
sampling and analytical procedure is normally conducted
in conjunction with a single-point, grab sampling and
analytical procedure for a pollutant, only one analysis
is ordinarily conducted. Therefore, great care must be
taken to obtain a valid sample and analysis. Although
In most cases only COi or Ot is required, it is recom-
mended that both CO; and O> be measured, and that
Citation 5 in the Bibliography be used to validate the
analytical data.
4.2 Single-Point, Integrated Sampling and Analytical
Procedure..
4.2.1 The sampling point in the duct shall be located
as specified in Section 4.1.1.
4.2.2 Leak-check (mandatory) the flexible bag as in
Mcclion 2.2.6. Set up the equipment as shown in Figure
3-2. Just prior to sampling, leak-check (mandatory) the
train by placing a vacuum gauge at the condenser inlet,
pulling a vacuum of at least 260 rum Hg (10 in. Hg),
plugging the outlet at the quick disconnect, and then
turning off the pump. The vacuum shall remain stable
for at least 0.5 minute. Evacuate the flexible bag. Con-
nect the probe and place it in the stack, with the Up of the
probe positioned at the sampling point; purge the sam-
pling line. Next, connect the bag and make sure that
all connections are tight and leak free.
4.2.3 Sample at a constant rate, or as specified by the
Administrator. The sampling run must be simultaneous
with, and for the same total length of time of, the pollut-
ant emission rate determination. Collect at least 30
liters (1.00 ft») of sample gas. Smaller volnmi s may be
collected, subject to approval of the Administrator.
4.2.4 Obtain one integrated fine gas sample during
each pollutant emission rate determination. For cii,iss{ou
rate correction factor determination, analyze the s.uni)i.;
within 4 hours after it is taken for percent CO j or percent
Oi (as outlined in Sections 4.2.5 through 427^ The
Orsat analyzer must be leak-checked (see Suction o)
before the analysis. If excess air is desired, proceed as
follows: (1) within 4 hours after the sample is taktii.
analyze it (as in Sections 4.2. j through 4.2 7) (or percent
COt, Of, and CO; (2) determine the puci•niaiv of din
gas that is Njby subiracting the sum of the pi-nn-nt Cot,
percent Of, and percent CO (rom 100 peri-i-m: .3) cal-
culate percent excess air, as outlined in Sei-tion G.'i
4.2.5 To ensure complete alisorption of the <"Oi, Ot,
or if applicable, CO, make repealed passes U.pji.gh each
absorbing solution until two consecutive readings ate the
same. Several passes (three or four) should be marie be-
tween readings. (If constant readings cannot lir-olrtalned
after three consecutive readings, n:]>iace the abfoiuing
solution.)
4.2.6 Repeat the analysis until the following .nl tical procedure
until the results of any three analyses differ by im iuo/e
than (a) 0.3 percent by volume when Ot Is less than 15.0
percent or (b) 0.2 percent by volume when Ot is greater
than 15.0 percent. Average the three acceptable values of
percent Ot and report the results to the nearest 0.1
IKTCellt.
4.2.6.3 For percent CO, repeat the analytical proce-
dure until the results of any three analyses differ by no
more than 0.3 percent. Average the three acceptable
values of percent CO and report the results to the nearest
0.1 percent.
4.2.7 After the analysis is completed, leak-check
(mandatory) the Orsat analyzer once again, as described
in Sections. Forthe results of the analysis to be valid, the
Orsat analyzer must pass this leak test before and after
the analysis. Note: Although in most instances only COt
or Ot is required, it is recommended that both COt and
Oi be measured, and that Citation 5 in tbc Bibliography
be used to validate the analytical data.
4.3 Multi-Point, Integrated Sampling and Analytical
Procedure.
4.3.1 Both the minimum number of sampling points
and the sampling point location shall be as specified in
Section 3.3.1 of this method. The use of fewer points than
specified is Mbject to the approval of the Administrator.
4.3.2 Follow the procedures outlined in Sections 4.2.2
through 4.2.7, except for the following: Traverse all
sampling points and sample at each point for an equal
length of time. Record sampling data as shown in Figure
3-3.
6. Leak-Check Procedure for Orsat Analy:era
Moving an Orsat analyzer frequently causes it to leak.
Therefore, an Orsat analyzer should be thoroughly leak-
checked on site before the flue gas sample is introduced
into it. The procedure for leak-checking an Orsat analyzer
is:
5.1.1 Bring trie liquid level in each pipette up to the
reference mark on the capillary tubing and then close the
pipette stopcock.
5.1.2 Raise the leveling bulb sufficiently to bring the
confining liquid meniscus onto the graduated portion of
the burette and then close the manifold stopcock.
5.1.3 Record the meniscus position.
5.1.4 Observe the meniscus in the burette and the
liquid level In the pipette for movement over the next 4
minutes.
5.1.5 For the Orsat analyzer to pass the leak-check,
two conditions must be met.
5.1.5.1 The liquid level in each pipette must not fall
below the bottom of the capillary tubing during this
4-minuteinterval.
5.1.5.2 The meniscus in the burette must not change
by more than 0.2 ml during this 4-minute Interval.
5.1.6 If the analyzer falls the leak-check procedure, all
rubber connections and stopcocks should be checked
until the cause of the leak is identified. Leaking stopcocks
must be disassembled, cleaned, and regreased. Leaking
rubber connections must be replaced. After the analyzer
is reassembled, the leak-check procedure must be
repeated.
6. CalculaHom
6.1 Nomenclature.
Mj-Dry molecular weight, g/g-mole (Ib/lb-mole).
%EA=Percent excess air.
%COt=Percent COf by volume (dry basis).
%Of=Percent Of by volume (dry basis).
%CO=Percent CO by volume (dry basis).
%Nt=Percent Nt by volume (dry basis).
0.264= Ratio of Oi to Ni In air, v/v.
0.280=Molecular weight of Nt or CO, divided by 100.
0.320=Molecular weight of Ot divided by 100.
0.440=Molecular weight of CO: divided by 100.
6.2 Percent Excess Air. Calculate the percent excess
air (it applicable), by substituting the appropriate
values of percent Ot, CO,and N: (obtained from Section
4.1.3 or 4.2.4) into Equation 3-1.
%02-0.5%CO
0.264 %N2(%0:
0.5 %CO "I
b02-0.5%CO>J
-0.5%CO>J100
Equation 3-1
333
-------
NOTE.—The equation above assumes that ambient
air Is used as the source of O: and that the fuel does not
contain appreciable amounts of Ni (as do coke oven or
blast furnace gases). For those cases when appreciable
amounts of Ni are present (coal, oil, and natural gas
do not contain appreciable amounts of N>) or when
oxygen enrichment is used, alternate methods, subject
to approval of the Administrator, are required.
6.3 Dry Molecular Weight. Use Equation 8-2 to
calculate the dry molecular weight of the stack gas
Afd=0.140(^COj)+0.320(%.0:)+0.280(%Ni+%CO)
Equation 3-2
NOTE.—The above equation does not consider argon
la air (about 0.9 percent, molecular weight of 37.7).
A negative error of about 0.4 percent is introduced.
The tester may opt to include Mgon in the analysis using
procedures subject to approval of the Administrator.
7.
1. Altshuller, A. P. Storage of Oases and Vapors in
Plastic Bags. International Journal of Air and Water
Pollution. 6:75-81.1963.
2. Conner, William D. and J. 9. Nader. Air Sampling
Plastic Bags. Journal of the American Industrial Hy-
giene Association. «S:291-297. 1964.
3. Burrell Manual for Oas Analysts, Seventh edition.
Burrell Corporation, 2223 Fifth Avenue, Pittsburgh,
Pa. 15219.1951.
4. Mitchell, W. J. and M. R. Mldgett. Field Reliability
of the Orsat Analyzer. Journal of Air Pollution Control
Association 26:491-495. May 1976.
5. Shigehara, R. T., R. M. Neullcht, and W. B. Smith.
Validating Orsat Analysis Data from Fossil Fuel-Fired
Units. Stack Sampling News. 4(2):21-26. August, 1976.
TIME
TRAVERSE
PT.
AVERAGE
Q
1pm
% DEV.a
'%DEV=
(MUSTBE<10%)
Figure 3=3. Sampling rate data.
334
-------
METHOD 4
DETERMINATION OF MOISTURE IN STACK GASES
1. Principle anil Applicability
1.1 Principle. A gas sample is extracted at a consult
rate from the source; moisture is removed from the sam-
ple stream and determined either volurnctricully or
gravimelrically.
1.2 Applicability. This met hod is applicable for
determining the moisture content ol slack gas.
Two procedures are given. The first is a reference
method, for accurate determinations of moisture content
(such as arc needed to calculate emission data). The
second is an approximation melhoil, which provides
estimates of percent moisture 10 aid in suiting isokinclic
sampling rates prior to a pollutant emission measure-
ment run. The approximation method described herein
is only a suggested approach; alternative means for
approximating the moisture content, e.g., drying tubes,
wet bulb-dry bulb techniques, condo.nstit ion techniques,
stoiohionietric calculations, pie.vious experience, etc.,
are also acceptable.
The reference method is often conducted simultane-
ously with a pollutant emission measurement run; when
it is, calculation of percent isokinelic, pollutant emission
rate, etc., for the run shall bo based upon the results ol
tlie reference method or its equivalent; these calculations
shall not be based upon the results of the approximation
method, unless the approximation method is shown, lo
the satisfaction of the Administrator, U.S. Environmen-
tal Protection Agency, to be capable of yielding results
within 1 percent HzO of the reference method.
NOTE.—The reference method may yield questionable
results when applied to saturated gas streams or to
streams that contain water droplets. Therefore, when
these conditions exist or are suspected, a second deter-
mination of the moisture content shall be made simul-
taneously with the reference method, as follows: Assume
that the gas stream is saturated. Attach a temperature
sensor [capable of measuring to ±1° C (2° F)| to the
reference method probe. Measure the stack gas tempera-
ture at each traverse point (sec*. Section 2.2.1) during (ho
reference method traverse; calculate the average stack
gas temperature. Next, determine the moisture percent-
age, either by: (1) using a psychrometric chart and
making appropriate corrections if stack pressure is
different from that of the chart, or (2) using saturation
vapor pressure tables. In cases where the psychrometric
chart or the saturation vapor pressure tables are not
applicable (based on evaluation of the process), alternate
methods, subject to the approval of the Administrator,
shall be used.
2. Reference Method
The procedure described in Method 5 for determining
moisture content is acceptable as a reference method.
2.1 Apparatus. A schematic of the sampling train
used in this reference method is shown in Figure 4-1.
All components shall be maintained and calibrated
according to the procedure outlined in Method 5.
2.1.1 Probe. The probe is constructed ol stainless
rteel or glass tubing, sufficiently heated to prevent
water condensation, and is equipped with a filter, either
In-stack (e.g., a plug of glass wool inserted into the end
of the probe) or heated out-stack (e.g., as described in
Method 5), to remove paniculate matter.
When stack conditions permit, other nwtals or plastic
tubing may be used for the probe, subject to the approval
nl the Administrator.
2.1.Z Condenser. The condenser consists of four
hnpingws connected in series with ground glass, leak-
free fittings or any similarly leak-free non-contaminating
fittings. The first, third, and fourth impingers shall be
of the Oreenburg-Sinith design, modified by replacing
the tip with a 1.3 centimeter (% inch) ID glass tube
extending to about 1.3 cm (K in.) from the bottom of
the flask. The second impinger shall be of the Greenburg-
Bmilh design with the standard tip. Modifications (e.g.,
using flexible connections between the impingers, using
materials other than glass, or using flexible vacuum lines
lo connect the filter holder to the condenser) may be
used, subject to the approval of the Administrator.
The first two impingers shall contain known volumes
of water, the third shall be empty, and the fourth shall
contain a known weight of 6- to Id-mesh Indicating typo
nlica gel, or equivalent desiccant. If the silica gel has
been previously used, dry at 175° C (350P V) for 2 hours.
New silica gel may be used as received. A thermometer,
capable of measuring temperature to within 1° C (2° F),
shall be placed at the outlet ol the fourth impinger, for
monitoring purposes.
Alternatively, any system may be used (subject to
the approval of the Administrator) that cools tho sample
pas stream and allows measurement of both the water
that has been condensed and the moisture leaving the
condenser, each to within 1 ml or 1 g. Acceptable means
are to measure the condensed water, either gravi-
»»etrically or volumetrically, and to measure the mois-
ture leaving the condenser by: (1) monitoring, the
temperature and pressure at the exit of the condenser
»ud usine Dnltnn's law of partial pressures, or (2) passing
the sample gas stream through a tared silica gel (or
equivalent desiccant) trap, with exit gases kept below
20° C (68° F), and determining the weight gain.
If means other than silica gel are used to determine the
amount of moisture leaving the condenser, it is recom-
mended that silica gel (or equivalent) still be used be-
tween tlie condenser system and pump, to prevent
moisture condensation in the pump and metering
devices and to avoid the need to make corrections for
moisture in the metered volume.
2.1.3 Cooling System. An ice bath container and
crushed ice (or equivalent) are used to aid in condensing
moisture.
2.1.4 Metering System. This system includes a vac-
uum gauge, leak-free pump, thermometers capable of
measuring temperature to within 3° C (6.4° F), dry gas
meter capable of measuring volume to within 2 percent,
and related equipment as shown in Figure 4-1. Other
metering systems, capable of maintaining a constant
sampling rate and determining sample gas volume, may
bo used, subject to the approval of the Administrator.
2.1.5 Barometer. Mercury, aneroid, or other barom-
eter capable of measuring atmospheric pressure to within
2.6 mm Hg (0.1 in. Hg) may be used. In many cases, the
barometric reading may be obtained from a nearby
national weather service station, in which case the sta-
tion value (which is the absolute barometric pressure)
shall be requested and an adjustment for elevation
differences between the weather station and the sam-
pling point shall be applied at a rate of minus 2.5mm Hg
(0.1 in. Hg) per 30 m (100 It) elevation increase or vice
versa for elevation decrease.
2.1.6 Graduated Cylinder and/or Balance. These
items are used to measure condensed water and moisture
caught in the silica gel to within 1 ml or 0.8 g. Graduated
cylinders shall have subdivisions no greater than 2 ml.
Most laboratory balances are capable of weighing to the
nearest 0.5 g or less. These balances are suitable for
use here.
2.2 Procedure. The following procedure is written for
a condenser system (such as the impinger system de-
scribed in Section 2.1.2) incorporating volumetric analy-
sis to measure the condensed moisture, and silica gel and
gravimetric analysis to measure the moisture leaving the
coudenser.
2.2.1 Unless otherwise specified by the Administrator,
a minimum of eight traverse points shall be used for
circular stacks having diameters less than 0.61 m (24 in.),
a minimum of nine points shall be used for rectangular
stacks having equivalent diameters \tss than 0.61 m
(24 in.), and a minimum of twelve travcrs points shall
be used in all other cases. The traverse points shall he
located according to Method 1. The use of fewer points
is subject to the approval of the Administrator. Select a
suitable probe and probe length such that all travt-r.se
points can be sampled. Consider sampling from opposite
sides of the stack (four total sampling port.1-:) for large
stacks, to permit use of shorter probe length!!. Mark the
probe with heat resistant tape or by some oth.-r method
to denote the proper distance into the stack or duet for
each sampling point. Place known volumes ol water in
the first two impingers. Weigh and record tho weigh I ol
the silica gel to the nearest 0.5 K, and tran-fer the silica
gel to the fourth impinger; alternatively, the- silica pel
may first be transferred to the impinger, and the wi .gin
of the silica gel plus impinger recorded.
2.2.2 Select a total sampling time such that a i/.:n,-
mum total gas volume of O.UU Sum (21 set; will !><• col-
lected, at a rate no greater than 0.021 m', mm (0.75 elm).
When both moisture content and pollutant emission iatc
335
-------
FILTER
(EITHER IN STACK
OR OUT OF STACK)
WALL
CONDENSER-ICE BATH SYSTEM INCLUDING
SILICA GEL TUBE—-j
Figure 4-1. Moisture sampling train-reference method
336
-------
PLANT
OCATION
OPERATOR
DATE
RUN NO
AMBIENT TEMPERATURE-
BAROMETRIC PRESSURE.
PROBE LENGTH rnlft)
SCHEMATIC OF STACK CROSS SECTION
TRAVERSE POINT
NUMBER
TOTAL
SAMPLING
TIME
(9). mu.
AVERAGE
STACK
TEMPERATURE
•C (°F)
PRESSURE
DIFFERENTIAL
ACROSS
ORIFICE METER
(AH).
•m(»J HjO
METER
READING
GAS SAMPLE
VOLUME
m3(ft3)
AV«
1.3 (f|3)
CAS SAMPLE TEMPERATURE
AT DRV GAS METER
INLET
ffmin),»C(aF)
A*
Aug..
OUTLET
(Tn.oull.°C(°F)
A*.
TEMPERATURE
OF GAS
LEAVING
CONDENSER OR
LAST IMPINGER.
°C <°F)
Figure 4-2. Field moisture determination-reference method.
337
-------
are to be determined, the moisture deit.Tinination shall
be simultaneous with, and for the same tulal length ul
time as, the pollutant emission rate, run, unless ollu-rv, ise
specified in an applicable subpart of the standards.
2,2.3 Set up the sampling tram as shown in Figure
4-1. Turn on the probe heater and (if applicable', the
fllter heating system to temperatures of about 120° C
(243° F), to prevent water condensation ahead ol the
condenser; allow time for the temperatures to stabilise.
Place crushed ice In the ice bath container. It is recom-
mended, but not required, that a leak check be done, u
follows: Disconnect the probe from tbe first impingor or
(If applicable) from the fllterholder. Plug the Inlet to the
lirst impinger (or filter holder) and puli a 380 mm (15 in.)
Jig vacuum; a lower vacuum may be used, provided that
il is not exceeded during the test. A leakage rate in
excess of -1 percent of the average sampling rate or 0.00057
niVmin (0.02 c(m), whichever is less, is unacceptable.
Following the; cak check, reconnect the probe to the
. iinpling train.
2.2.-1 During the sampling nm, maintain a sampling
i ate within 10 percent of constant rate, or as specified by
i!ie Administrator. For each run, record the data re-
iinired on the example data sheet shown in Figure 4-2.
lit- sim1 to record the dry gas meter reading at the begin-
ning and end of each sampling time inurement and when-
ever sampling is halted. Take other appropriate readings
at each sample point, at least once during each time
increment.
2.2.5 To begin sampling, position the probe tip at the
first traverse point. Immediately start the pump and
adjust the flow to the desired rate. Traverse the cross
section, sampling at each traverse point for an equal
length of time. Add more ice and, if necessary, salt to
maintain a temperature of less than 20° C (68° F) at the
silica gel outlet.
2.2.6 After collecting the sample, disconnect the probe
from the filter holder (or from the first impinger) and con-
duct a leak check (mandatory) as described in Section
2.2.3. Record the leak rate. If the leakage rate exceeds the
allowable rate, the tester shall either reject the test re-
sults or shall correct the sample volume as in Section 6 3
of Method 5. Next, measure ihc volume of the moisture
condensed to the nearest ml. Determine the increase in
weight of the silica gel (or silica gel pins impinger) to the
nearest 0.5 g. Record this information (see example data
sheet, Figure 4-3) and calculate the inoisluio pcrcuntaeo
as described in 2.3 below. '
2.3 Calculations. Carry out the following calculations.
retaining at least one extra decimal figure beyond that of
the acquired data. Round off flames after final calcula-
tion.
2.3.2 \ olumo ol water vapor condensed
FINAL
INITIAL
DIFFERENCE
IMPINGER
VOLUME,
ml '
SILICA GEL
WEIGHT.
a
Figure 4 3. Analytical data - reference method.
2.3.1 Nomenclature.
B,tt—Proportion of \\ater vapor, hy volume, in
the gas stream.
Mw = Molecular weight of water. 18.0 g/g-mole
(18.01b/lb-mole).
>*OT=Absolute pressure (for this method, same
as barometric pressure) at the dry gas meter,
mm Hg (in. He,).
P.Ij=Standard absolute pressure, TOO mm Hg
(29.92 in. Hg).
R- Ideal gas constant, 0.06236 (mm Hg) (m!)/
(g-iuole) (°K) for metric units and 21.85 (hi.
Hg) (ft»)/(lb-mole) (°R) for English units.
7'm-Absolute temperature, at meter, °K (°H).
3'iij=Standard absolute temperature, 293° K
(528° R).
V« = Dry gas volume measured by dry gas meter,
dcm (dcf).
AK,.=Incrcniontal dry gas volume measured by
dry gas meter at each traverse point, dcm
(dcf).
Vn(»/d) = Dry gas volume measured hy the dry gas
meter, corrected to standard conditions,
dscm (dscf).
V..(«j)=Volume of water vapor condensed corrected
to standard conditions, scm (scf).
Vi»t(,id) =Volume of water vapor collected in silica
gel corrected to standard conditions, scm
(scf).
Vv=Final volume of condenser water, ml.
Vi = TjiJtial volume, if any, of condenser water
ml.
IP; = Final weight of silica gel or silica gel plus
impinger, g.
W,p=Initial weight of silica gel or silica gel pluc
impiuger, g.
K=Dry gas meter calibration factor.
P»=Density of waier, 0.9982 g/nil .0.002201
Ib/ml).
Kqiiiition 4 1
where:
A'i=0.00)333 ni'/ml for metric units
=0.04707 ftJ/ml for English units
3.3.3 Volume of water vapor collected in silica p 3 of M.-iiio
'2 3 5 Moisture Content.
n,^ ..Jl-.sc",>-1-1' - ;>•:>
Virc (SM) + 1 .. ~i -nlA) r I ai (.1.1)
NUTK. — hi saturated or moistm* droplet-laden gftf
blreams, (wo calculations of the moisture content of the
stack gas shall be made, one using a value based upon
the saturated conditions (see Section 1 2), and anothei
based upon the results of the impingur analysis. The
lower of these two values of if. . shall be considered cor-
rect.
230 Vciification of constant Campling rate. For each
time liK'i-pinent, di'tennine the AVm. C'alculale the
average. If the value for any time inorcinrnt dill'ers from
the average by more than 10 perceni. rejri 1 Die results
and repeat the nm. •
3. Approximation Mithod
The approximation method descnhcd belov is PMJ.
sented only as a suggested incihod (see Section I 21
3.1 Apparatus.
3.1.1 Probe. Stainless steel or glass tubing, sniliciently
heated to prevent water condensation and equipped
with a filter (either in-stack or heated out-stack) to re-
move participate matter. A plug of glass wool, inserted
inta*he end ol the probe, is a satisfactory filter
3.1.2 Impingers. Two midget impingers, each with
30 ml capacity, or equivalent.
3.1.3 Ice TCath. Container and ice, to aid in condens-
ing moisture in impingers.
3.1.4 Drying Tube. Tube packed with imv or re-
generated 0- to 16-mesh indicatinji-lype silica gel (or
equivalent ilcsiccant), to dry the sample gas and to pro-
tect the metor and pump.
3.1.5 VaJve. Needle valve, lo regulate the sample gas
flow rate.
3.1.6 1'iiinp. Leak-free, diaplnaiim type, or equiva-
lent, to pull the gas sample through the train.
3.1.7 Volume meter. Dry gas meter, sufficiently ac-
curate to measure, the sample volume within 2%, and
calibrated over Uie range of flow rates and conditions
actually encountered during sampling.
3.1.8 Hate Meter. Rotameter, to nleasmv Hit How
range from 0 to 3 1 pm (0 to 0.11 cfm ) .
3.1.9 Graduated Cylinder. 25 ml.
3.1.10 Barometer. Mercury, aneroid, or olhei barom-
otor, as described in Section 2.1.5 above.
3.1.11 Vacuum Gauge. At least 760 mm Tig (30 in.
Hg) gauge, to be used for the sampling leak chock
3.2 Procedure.
3.2.1 Place exactly 6 ml distilled water in each 1111-
pingcr. Assemble the apparatus without the probe as
shown in Figure 4-4. Leak check the train by placing a
vacuum gauge at the inlet to the first impingcr and
drawing a vacuum of at least 250 mm Hg (10 in. Hg),
plugging the outlet of the rotaraeter, and then turning
off the pump. The vacuum shall remain constant for at
east one minute. Carefully release the vacuum gauge
Ibcfore unplugging the rotameter end
3.2.2 Connect tbe probe, insert it into the stack, and
sample at a constant rate of 2 1pm (0.071 cfm). Continue
sampling until the dry gas meter registers about 30
liters (1.1 ft') or until visible liquid droplets are carried
over from the first Impinger to the second. Record
temperature, pressure, and dry gas meter readings as
required by Figure 4^5.
3.2.3 After collecting the sample, combine the con-
tents of the two impingers and measure the volume to the
nearest 0.6 ml.
338
-------
HEATED PROBE
SILICA GEL TUBE
RATE METER,
VALVE
MIDGET IMPIIMGERS
PUMP
Figure 4-4. Moisture-sampling train - approximation method.
LOCATION.
TEST
COMMENTS
DATE
OPERATOR
BAROMETRIC PRESSURE
CLOCK TIME
GAS VOLUME THROUGH
METER, (Vm),
m3 (ft3)
RATE METER SETTING
m^/min. (ft^/min.)
METER TEMPERATURE,
°C(°F)
Figure 4-5. Field moisture determination - approximation method.
339
-------
3.3 Calculations. The calculation method presented is
designed to estimate the moisture in the stack gas;
therefore, other data, which are only necessary for ac-
curate moisture determinations, are not collected. Tho
following equations adequately estimate the moisture
content, for the purpose of determining isokinetic sam-
pling rate settings.
3.3.1 Nomenclature.
B»»=Approsimate proportion, by volume, of
water vapor in the gas stream leaving the
second impinger, 0.025.
£„.=Water vapor in the gas stream, proportion by
volume.
Af.=Moleeular weight of water, 18.0 gig-mole
(18.0 Ib/lb-mole)
P.,=Absolute pressure (for this method, same as
barometric pressure) at the dry gas meter.
P,/d=Standard absolute pressure, 760 mm Hg
(29.92 in. Hg).
J?=IdeaJ gas constant, 0.06238 (mm Eg) (m«)/
(g-mole) (°K) for metric units and 21.85
(in. Hg) (ft«)/lb-mole) (°R) for English
units.
T.=Absolute temperature at meter, °K (°R)
T.,i=Standard absolute temperature, 293° K
(628° R)
Vf= Final volume of impinger contents, ml.
Vi=InitiaI volume of impinger contents, ml.
V*=Dry gas volume measured by dry gas meter,
dcm (dcf).
V.(,u)=Dry gas volume measured by dry gas meter,
corrected to standard conditions, dscm
(dscf).
V«.(,ij)=Volnme of water vapor condensed, corrected
to standard conditions, scm (scf).
«.=Density of water, 0.9982 g/ml (0.002201 Ib'ml).
3.3.2 Volume of water vapor collected.
Equation 4-5
vhere:
Ki=0.001333 m'/ml lor metric units
=0.04707 ftl/ml for English unite.
3.3.3 Oas volume.
=K,
v P
"m* »
Equation 4-6
£i=0.3868 °K/mm Hg for metric units
=17.84 °R/in. Hg for English units
3.3.4 Approximate moisture content.
V
JO __ r <"
|_ D
"T *»»
v,c
Equation 4-7
4. Calibration
4.1 For the reference metliod, calibrate equipment as
specified in the following sections of Method 6: Section 5.3
(metering system); Section 5.5 (temperature gauges);
and Section 5.7 (barometer). The recommended leak
check of the metering system (Section 5.6 of Method 5)
also applies to the reference method. For the approxima-
tion method, use the procedures outlined in Section 5.1.1
of Method 6 to calibrate the metering system, and the
procedure of Method 5, Section 5.7 to calibrate the
barometer.
5. Bibliography
1. Air Pollution Engineering Manual (Second Edition).
Danielson, J. A. (ed.). U.S. Environmental Protection
Agency, Office of Air Quality Planning and Standards.
Research Triangle Park, N.C. Publication No. AP-40.
1973.
2. Devorkin, Howard, et al. Air Pollution Source Test-
ing Manual. Air Pollution Control District. Los Angeles,
Calif. November, 1963.
3. Methods for Determination of Velocity, Volume,
Dust and Mist Content of Gases. Western Precipitation
Division of Joy Manufacturing Co., I/os Angeles, Calif.
Bulletin WP-50.1968.
340
-------
METHOD 5
DETERMINATION OF PARTICULATE EMISSIONS FROM STATIONARY SOURCES
1. Principle and Applicability
1.1 Principle. Paniculate matter is withdrawn iso-
kinetically from the source and collected on a glass
fiber filter maintained at a temperature in the range ot
120±M« C (248±2S° F) or such other temperature as
specified by an applicable subpart of the standards or
approved by the Administrator, U.S. Environmental
Protection Agency, for a particular application. The
particulate mass, which includes any material that
condenses at or above the nitration temperature, is
determined gravimetrically after removal of uncombined
water.
1.2 Applicability. This method is applicable for the
determination of particulate emissions from stationary
sources.
2. Apparatus
2.1 Sampling Train. A schematic of the sampling
train used in this method is shown in Figure 5-1. Com-
plete construction details are given in APTD-0581
(Citation 2 in Section 7); commercial models of this
train are also available. For changes from APTD-0581
and tot allowable modifications of the train shown in
Figure 9-1, see the following subsections.
The operating and maintenance procedures for the
sampling train are described in APTD-0578 (Citation 3
in Section 7). Since correct usage is important in obtain-
ing valid results, all users should read APTD-0576 and
adopt the operating and maintenance procedures out-
lined in it, unless otherwise specified herein. The sam-
pling train consists of the following components:
2.1.1 Probe Nozzle. Stainless steel (316) or glass with
sharp, tapered leading "edge. The angle of taper shall
be <30° and the taper shall be on the outside to preserve
a constant internal diameter. The proble nozzle shall be
of the button-hook or elbow design, unless otherwise
specified by the Administrator. If made of-stainless
steel, the nozzle shall be constructed from seamless tub-
ing; other materials of construction may be used, subject
to the approval of the Administrator.
A range of nozzle sizes suitable for isoklnetlc sampling
should be available, e.g., 0.32 to 1.27 cm (H to J» in.)—
or larger if higher volume sampling trains are used-
inside diameter (ID) nozzles in increments of 0.16 cm
(Hi in.). Each nozzle shall be calibrated according to
the procedures outlined in Section 5.
2.1.2 Probe Liner. Borosilicate or quartz glass tubing
with a heating system capable of maintaining a gas tem-
perature at the exit end during sampling of 120±14° C
(248±25° F), or such other temperature as specified by
an applicable subpart of the standards or approved by
tbe Administrator for a particular application. (The
tester may opt to operate the equipment at a temperature
lower than that specified.) Since the actual temperature
at the outlet of the probe is not usually monitored during
sampling, probes constructed according to APTD-0581
and utilizing the calibration curves of APTD-0576 (or
calibrated according to the procedure outlined in
APTD-0576) will be considered acceptable.
Either borosiliccte or quartz glass probe liners may be
nsed for stack temperatures up to about 480° C ;900° F):
quartz liners shall be used for temperatures between 480
and 900° C (900 and 1,650° F). Both types of liners may
be used at higher temperatures than specified for short
periods of time, subject to the approval of the Adminis-
trator. The softening temperature for borosih'cate is
820° C (1,508*F), and far quartz it is l,50t° C (2,732° F).
Whenever practical, every effort should be made to use
borosih'cate ox quartz glass probe liners. Alternatively,
metal liners (e.g., 316 stainless steel. Incoloy 825,2 or ether
corrosion resistant metals) made of seamless tubing may
be used, subject to the approval of the Administrator.
2.1.3 Pilot Tub*. Type S, as described in Section 2.1
of Method 2, or other device approved by the Adminis-
trator. The pitot tube shall be attached to the probe (as
shown in Figure 5-1) to allow constant monitoring of the
stack gas velocity Tbe impact (high pressure) opening
plane of the pitot tube shall be even with or above the
nozzle entry plane (see Method 2, Figure 3-6b) during
' Mention of trade names or specific products does not
constitute endorsement by the Environmental Protec-
tion Agency.
sampling. The Type S pitot tube assembly shall have a
known coefficient, determined as outlined In Section 4 of
Method 2.
2.1.4 Differential Pressure Gauge. Inclined manom-
eter or equivalent dev^o itwo), as escribed in Section
2.2 of Method '2. One manometer shall be used .or velocity
head (Ap) readings, and the other, for orifice differential
pressure readings.
2.1.5 Filter Holder. Borosilicate glass, with a glass
frit filter support and a silicoue rubber gasket. Other
materials of construction (e.g., stainless steel, Teflon,
Viton) may be used, subject to approval of the Ad-
ministrator. The holder design shall provide a positive
seal against leakage irom the outside or around the filter.
The holder shall be attached immediately at the outlet
of the probe (or cyclone, if used).
2.1.6 Filter Heating System. Any heating system
capable of maintaining a temperature around the filter
holder during sampling o. 120±14° C (24S±2V F), or
such other temperature as specified by an applicable
subpart of the standards or approved by the Adminis-
trator for a particular application. Alternatively, the
tester may opt to operate the equipment at a temperature
lower than that specified. A temperature gauge capable
of measuring temperature to within 3" C (5.4° F) shall
be Installed so that the temperature around the filter
holder can be regulated and monitored during sampling.
Heating systems other than the one shown in APTD-
0581 may be used.
2.1.7 Condenser. The fallowing system shall be used
to determine the stack gas moisture coutent: Four
impingers connected In series with leak-lree ground
glass fittings or any similar leak-free non-contaminating
fittings. The first, third, and fourth Impingers shall be
ol the Greenburg-Smith design, modified by replacing
the tip with 1.3 cm (>$ in.) ID glass Lube extending to
about 1.3 cm (M in.) from the bottom of the flask. The
second impinger shall be of the Grecnburg-Smith design
with the standard tip. Modifications (e.g., using flexible
connections between the impingers, using materials
other than glass, or using flexible vacuum lines to connect
the filter holder to the condenser) may be used, subject
to the approval of tbe Administrator. Tbe first and
second tmpingers shall contain known quantities of
water (Section 4.1.3), the third shall be empty, and the
fourth shall contain a known weight of silica gel, or
equivalent desiccant. A thermometer, capable of measur-
ing temperature to within 1° C (2° Q shall be placed
at the outlet of the fourth Impinger for monitoring
purposes.
Alternatively, any system that cools the sample gas
stream and allows measurement of the water condensed
and moisture leaving the condenser, each to within
1 ml or 1 g may be used, subject to the approval of the
Administrator. Acceptable means are to measure the
condensed water either gravimetrically or volumetrically
and to measure the moisture leaving the condenser by:
(1) monitoring the temperature and pressure at the
exit of the condenser and using Dalton's law of partial
pressures; or (2) passing the sample gas stream through
a tared silica gel (or equivalent desiccant) trap with
exit gases kept below 20° C (68° F) and determining
the weight gain.
If means other than silica gel are used to determine
the amount of moisture leaving the condenser, it is
recommended that silica gel (or equivalent) still be
used between the condenser system and pump to prevent
moisture condensation in the pump and metering devices
and to avoid the need to make corrections, for moisture iu
the metered volume.
NOTE.—If a determination of the particulate matter
collected In the Impingers is desired in addition to mois-
ture content, the impinger system described above shall
be used, without modification. Individua, States or
control agencies requiring this information shall be
contacted as to the sample recovery and analysis ol the
impinger contents.
2.1.8 Metering System. Vacuum gauge, leak-free
pump, thermometers capable of measuring temperature
to within 3° C (5.4° F), dry gas meter capable ol measuring
volume to within 2 percent, and related equipment, as
shown in Figure 5-1. Other metering systems capable of
maintaining sampling rates within 10 percent of iso-
kinetic and ol determining sample volumes to within 2
341
-------
TEMPERATURE SENSOR
PROBE
TEMPERATURE
SENSOR
KWPIWGER TRAIN OPTIONAL, MAY BE REPLACED
BY AN EQUIVALENT CONDENSER
HEATED AREA THERMOMETER
THERMOMETER
REVERSE-TYPE
PITOTTUBE
IMPINGERS ICE BATH
BY-PASS VALVE
PITOT MANOMETER
ORIFICE
THERMOMETERS
DRY GAS METER
AIR-TIGHT
PUMP
CHECK
VALVE
VACUUM
LINE
Fjgure 5-1, Particulate-sampling train.
342
-------
percent may be used, subject to Iho approval oflho
Administrator. When the meteriug system is used in
conjunction with a pilot tube, tile system shall enable
checks ol isokinctic rales.
Sampling trains utilizing metering systems designed for
higher flow rates than that described in APTD-05S1 or
APTD-057G may be used provided that the specifica-
tions Oi this method arc met.
2.1.9 Barometer. Mercury, aneroid, or other barometer
capable or measuring atmospheric pressure to within
2.5 mm Hg (0.1 in. Hg). In many cases, the barometric
reading may be obtained from a nearby national weather
service station, in which case the station value (which is
the absolute barometric pressure) shall be requested and
:u\ adjustment for elevation differences between the
w> ather station and sampling point shall be applied at a
rate of minus 2.5 mm Hg (0.1 in. Hg) per 30 m (100 ft)
elevation increase or vice versa for elevation decrease.
2.1.10 Gas Density Determination Equipment.
Temperature sensor and pressure gauge, as described
in Sections 2.3 and 2.4 of Method 2, and gas analyzer,
if necessary, as described in Method 3. The temperature
sensor shall, preferably, be permanently attached to
the pilot tube or sampling probe in a fixed configuration,
such that the tip of the sensor extends beyond theleading
edge of the probe sheath and does not touch any metal.
Alternatively, the sensor may be attached just prior
to use ill the field. Note, however, that if the temperature
sensor is attached in the field, the sensor must be placed
in- an interference-free arrangement with respect to the
Type S pitot tube openings (see Method 2, Figure 2-7).
As a second alternative, if a di/Terence of not more than
1 percent in the average velocity measurement is to be
introduced, the temperature gauge need not be attached
to the probe or pilot tube. (This alternative is subject
to the approval of the Administrator.)
2.2 Sample Recovery. The following items are
needed;
2.2.1 Probe-Liner and Probe-Nozzle Brushes. Nylon
bristle brushes with stainless steel wire handles. The
probe brush shall have extensions (at least as long as
the probe) of stainless steel, Nylon, Teflon, or similarly
inert material. The brushes shall be properly sized and
shaped to brush out the probe liner and nozzle.
2.2.2 Wash Bottles—Two. Glass wash bottles are
recommended; polyethylene wash bottles may be used
at the option of the tester. It is recommended that acetone
not be stored in polyethylene bottles for longer than a
month.
2.2.3 Glass Sample Storage Containers. Chemically
resistant, borosilicate glass bottles, for acetone washes,
SOO ml or 1000 ml. Screw cap liners shall either be rubber-
backed Teflon or shall be constructed so as to be leak-free
and resistant to chemical attack by acetone. (Narrow
mouth glass bottles have been found to be less prone to
leakage.) Alternatively, polyethylene bottles may be
used.
2.2.4 Petri Dishes. For filter samples, glass or polA-
wthylene, unless otherwise specified by the Admin-
istrator.
2.2.5 Graduated Cylinder and/or Balance. To meas-
ure condensed water to within 1 ml or 1 g. Graduated
cylinders shall have subdivisions no greater than 2 ml.
Most laboratory balances are capable of weighing to the
nearest 0.5 g or less. Any of these balances is suitable for
use here and in Section 2.3.4.
2.2.6 Plastic Storage Containers. Air-tight containers
to store silica gel.
2.2.7 Funnel and Rubber Policeman. To aid in
transfer of silica gel to container; not necessary if silica
gel is weighed in the field.
2.2.8 Funnel. Glass or polyethlene, to aid in sample
recovery.
2.3 Analysis. For analysis, the following equipment is
needed.
2 3.1 Glass Weighing Dishes.
2.3.2 Desiccator.
2.3.3 Analytical Balance. To measure to within 0.1
ing.
2.3.4 Balance. To measure to within 0:5 g.
2.3.5 Beakers. 250 ml.
2.3.6 Hygrometer. To measure the relative humidity
of the laboratory environment.
2.3.7 Temperature Gauge. To measure the tempera-
ture of the laboratory environment.
3. Reagents
3.1 Sampling. The reagents used in sampling are as
follows:
3.1.1 Filters. Glass fiber filters, without organic
binder, exhibiting at least 99.95 percent efficiency (<0.05
percent penetration) on 0.3-micron dioctyl phthalate
smoke particles. The filter efficiency test shall be con-
ducted in accordance with ASTM standard method D
2980-71. Test data from the supplier's quality control
program are sufficient for this purpose.
3.1.2. Silica Gel. Indicating type, 6 to 1ft mesh. If
previously used, dry at 175° C (350* F) for 2 hours. New
silica gel may bo used as received. Alternatively, other
types of desiccants (equivalent or better) may be used,
subject to the approval of the Administrator.
3.1.3 Water. When analysis of the material caught in
the impingers is required, distilled water shall be used.
Run blanks prior to field use to eliminate a high blank
on teat samples.
3.1.4 Crushed Ice.
3.1.5 Stopcock Grease. Acetone-insoluble, heat-stable
sUicone grease. This is not necessary if screw-on con-
nectors with Teflon sleeves, or similar, are used. Alterna-
tively, other types of stopcock grease may be used, sub-
3.2 Sample Recovery. Acetone—reagent grade, <0.001
percent residue, iu glass bottles—is required. Acetone
from metal containers generally has a bigh residue blank
and should not be used. Sometimes, suppliers transfer
acetone to glass bottles from metal containers; thus,
acetone blanks shall be run prior to field use and only
acetone with low blank values (<0.001 percent) shall be
used. In no case shall a blank value of greater than 0.001
percent of the weight of acetone used be subtracted from
the cample weight.
3.3 Analysis. Two reagents are required for lhc analy-
sis:
3.3.1 Acetone. Same as 3.2.
3.3.2 • Desiccant. Anhydrous calcium sulfate, indicat-
ing type. Alternatively, other types of desiccants may be
used, subject to the approval of the Administrator.
4. Procedure
4.1 Sampling. The complexity of this method is such
that, in order to obtain reliable results, testers should be
trained and experienced with the tesl procedures.
4.1.1 Pretest Preparation. All the components shall
be maintained and calibrated according to .0 procedure
described in APTD-0.r>7G, unless 'olhenvKe, specified
herein.
Weigh several 2flOto 3001> portions of sillc-agcljn air-tight
containers to the nearest 6.5 g. Record the total weight of
the silica gel plus container, on eai-li container. As an
aJirniiitive, the silica gel need not be preweighed, but
may be weighed directly in its impinger or sampling
holder jusl prior to train assembly.
Check filters visually against light for irregularities anil
flaws or pinhole leaks. Label filters of t lie proper diameter
on the. back side near the edge using numbering machine
ink. As an alternative, label the shipping containers
(glass or plastic petri dishes) and keep the liliors in these
containers at all times except during sampling and
weighing.
Desiccate the fillers at 20±5.6° C (68±10° F) and
ambient pressure for at least 24 hours aiid weigh at in-
tervals of at least 6 hours to a constant weight, i.e.,
<0.5 mg change from previous weighing; record results
to the nearest 0.1 mg. During each weighing the filter
must not be exposed to the laboratory atmosphere for a
period greater than 2 minutes and a relative humidity
above 50 percent. Alternatively (unless otherwise speci-
fied by the Administrator), the filters may be oven
dried al 105° C (220° F) for 2 to 3 hours, desiccated for 2
hours, and weighed. Procedures other than those de-
scribed, which account for relative humidity effects, may
be used, subject to the approval of the Administrator.
4.1.2 Preliminary Determinations. Select the sam-
pling site and the minimum number of sampling points
according to Method 1 or as specified by the Administra-
tor. Determine the stack pressure, temperature, and the
range of velocity heads using Method 2; it is recommended
that a leak-check of the pitot lines (see Method 2, Sec-
tion 3.1) be performed. Determine tlic moisture content
using Approximation Method 4 or its alternatives for
the purpose of making isokinetic sampling rate settings.
Determine the stack gas dry molecular weight, as des-
cribed in Method 2, Section 3.6; if integrated Method 3
sampling is used for molecular weight determination, Ihe
integrated bag sample shall be taken simultaneously
with, and for the same total length of lime as, the par-
ticulate sample run.
Select a noMle sixe based on. the range of velocity heads,
such that it is not necessary to change the nozzle size in
order to maintain isokinetic sampling rates. During the
run, do not change .the nozzle size. Ensure that the
proper differential pressure gauge is chosen for the range
of velocity heads encountered (see Section 2.2 of Method
2).
Select a suitable probe liner and probe length such that
all traverse points can be-sampled. For large stacks,
consider sampling from opposite sides of the stack to
reduce the length of probes.
Select a total sampling time greater than or equal to
the minimum total sampling time specified in the test
procedures for the specific industry such that (1) the
sampling time per point is not less than 2 rain (or some
greater time interval as specified by the Administrator),
and (2) the sample volume taken (corrected to standard
conditions) will exceed the required minimum total gas
sample volume. The latter is based on an approximate
average sampling rate.
It is recommended that the number of minutes sam-
pled at each point be an integer or an integer plus one-
naif minute, in order to avoid timekeeping errors.
- In some circumstances, e.g., batch cycles, it may be
necessary to sample for shorter times at the traverse
points and to obtain smaller gas sample volumes. In
these cases, the Administrator's approval must first
be obtained.
'4.1.3 Preparation of-Collection Train. During prep-
aration and assembly of the sampling train, keep all
openings where contamination can occur covered until
just prior to assembly or until sampling is about to begin.
Place 100 ml of water in each of the first two impingers,
leave the third impinger empty, and transfer approxi-
mately 200 to 300 g of preweighed silica gel from its
container to the fourth impinger. More silica gel may be
used, but care should be taken to ensure that it is not
entrained and carried out from the impinger during
sampling. Place the container in a clean place for later
use in the sample recovery. Alternatively, the weight of
tbe silica gel plus impinger may be determined to the
nearest 0.5 g and recorded.
Using a tweezer or clean disposable surgical gloves,
place a labeled (identified) and weighed filter in the
filter holder. Be sure that the filter is properly centered
343
-------
and the gasket properly placed so as to prevent the
sample gas stream from circumventing the alter. Check
the filter tor tears after assembly is complet«d.
When glass liners are used, install the selected nozzle
using a Viton A 0-ring when stack temperatures are
less than 260° C (500° F) snd an asbestos string gasket
when temperatures are higher, See APTD-0676 lor
details. Other connecting systems niJi.p either 311; Mum
less steel or Tellon ferrules may be used. When metal
liners are used, install the nozzle as above or by a leak-
free direct mechanical connection. Mark the probe with
heat resistant tape or by some other method to denote
the proper distance into the stack or duel for each .sam-
pling point.
Set up the train as in Figure 5-1, using (if neeess;u>)
a very light coat of silicone grease on nil ground glass
Joints, greasing only the outer portion (sec APTD-IKM
to avoid possibility of contamination by the sili'-une
grease. Subject to the approval of the Administrator, ;i
glass cyclone may be used between the probe and hluT
holder when Hie total parliculale catch is cxp.-i tc7t> and APTD-0581 may bu
helpful. Start the pump with bypass valve fully open
and coarse adjust valve completely closed. Partially
open the coarse adjust valve and slowly close tho bypass
valve until the desired vacuum is reached. Do not reverse
direction of bypass valve; this will cause water to back
up into the filter holder. If the desired vacuum is ex-
ceeded, either leak-check at. this higher vacuum or cud
tho leak check as shown below and start over.
When the leak-cheek is completed, llrst slowly remove
the plug from tho inlet to the probe, filler holder, or
cyclone (if applicable) and immediately turn off the
vaccum pump. This prevents the waler in the impingers
from being forced backward into the niter holder nnd
silica gel from being entrained backward into the third
impinger.
4.1.4.2 Leak-Checks During Sample, Run. If, during
the sampling run, a. component (e.g., iilter assembly
or impinger) change becomes necessary, a leak-check
shall be conducted immediately before the change is
made. The leak-check shall be done according to the
procedure outlined in Section 4.1.4.1 above, except that
It shall be done at. a vacuum equal to or greater than the
maximum value recorded up to that point in the test.
If the leakage rate is found to be no greater than 0.00057
m'/min (0.02 cfm) or 4 percent of the average sampling
rate (whichever is less), the results are acceptable, and
no correction will need to be applied to the total volume
of dry gas metered; if, however, a higher leakage rate
is obtained, the tester shall either record tho leakage
rate and plan to correct the sample volume as shown hi
Section ti.3 of this method, or shall void the sampling
run.
Immediately after component changes, leak-checks
are optional; if such leak-checks are done, the procedure
outlined in Section 4.1.4.1 above shall bo used.
4.1.4.3 Post-test Leak-Cheek. A leak-check is manda-
tory at the conclusion of each sampling run. The leak-
check shall be done in accordance with the procedures
outlined in Section 4.1.4.1, except that it shall be con-
ducted at a vacuum equal to or greater than the maxi-
mum value reached during the sampling run. If the
leakage rate is found to be no greater than 0.000.17 in'.'mln
(0.02 cfm) or 4 percent of the average sampling rale
(whichever is less), the results are acceptable, and uo
correction need be applied to the total volume of dry gas
motered. If, however, a higher leakage rate is obtained,
the tester shall either record the leakage rale and correct
the sample volume as shown in Section 6.3 of t his method,
or shall void the sampling run.
4.1.5 Particular Train Operation. During Ihe
sampling run, maintain an isokinetic sampling rat«
(within 10 percent of true isokinetic unless otherwise
specified by the Administrator) and a temperature
around the filter of 120±14° C (248±25° F), or such other
temperature as specified by an applicable subpart of the
standards or approved by the Administrator.
For each run, record the data required on a data sheet
such as the one shown in Figure 5-2. Be sure to record tho
initial dry gas meter reading. Record tho dry gas meter
readings at the beginning and end of each sampling time
Increment, wben changes in flow rates are maofi, Defore
and after each leak check, and when sampling is halted.'
Take other readings required by Figure 5-2 at least once
at each sample point during each time increment and
adilitional readings when significant changes (20 percent
variation in velocity bead readings) necessitate addi-
tional adjustments in flow rate. Level and tero the
manometer. Because the manometer level and zero may
drift due to vibrations and temperature changes, make
periodic checks during Ihe traverse.
Clean th« portholes prior to the test run to mlnlmlza
the chance of sampling deposited malarial. To begin
sampling, remove the nozzle cap, verily that the filter
and probe heating systems are up to temperature, and
that the pltot tube and probe are properly positioned.
Position the nozzle at the first traverse point with the tip
pointing directly into the gas stream. Immediately start
the pump and adjust the flow to isokinotic conditions.
Nomographs are available, which aid in the rapid adjust-
ment of the Isokinetic sampling rate without excessive
computations. These nomographs are designed for use
when the Type 8 pilot tube coefficient is 0.85±0,02, and
the stack gas equivalent density (dry molecular weight)
is equal to 29±4. APTD-0576 details Iho procedure for
using the nomographs. If C, and Mt are outside the
above stated ranges do not use. the nomographs unless
appropriate steps (see Citation 7 in Section 7t are luken
to compensate for the deviations.
When the stack is under significant negative pressure
(height of impinger stem), take care to close the coarse
adjust valve before inserting the probe into the stack to
prevent water from backing into the filter holder. If
necessary, the pump may be turned on with the coarse
adjust valve closed.
When the probe is in position, block off the openings
around the probe and porthole to prevent unrepre-
senlalive dilution of the gas stream.
Traverse the slack cross-section, as required by Method
1 or as specified by the Administrator, being careful not
to bump tho probe nozzle into the stack walls when
sampling near the walls or when removing or inserting
the probe through the portholes; this minimizes the
chance of extracting deposited material.
During the test run, make periodic adjustments to
keep the temperature around the filter holder at the
proper level; add more ice and, if necessary, salt to
maintain a temperature of less than 20° C (68° Fl at the
condenser/silica gel outlet. Also, periodically check
the level and zero of the manometer.
If the pressure drop across the filter becomes too high,
making isokinetic sampling difficult to maintain, the
liltcr may be replaced in the midst of a sample run. It
is recommended that another complete filter assembly
he used rather than attempting to change the filter itself.
Before a new filter assembly is installed, conduct a leak-
check (see Section 4.1.4.2). The total particulate weight
shall include the summation of all filter assembly catches.
A single train shall be used for the entire sample run,
except in cases where simultaneous sampling is required
in two or more separate ducts or at two or more different
locations within the same duct, or, in cases, where Equip-
ment failure necessitates a change of trains. In all other
si tuations, the use of two or more trains will be subject to
the approval of thn Artminictrator.
Note that when two or more trains are used, separate
analyses of the front-half and (if applicable) impingei
catches from each train shall be performed, unless identi-
cal nozzle sizes were used on all trains, in which case, the
front-half catches from the individual trains may be
combined (as may the impinger catches) and one analysis
of front-half catch and one analysis of impingor catch
may be performed. Consult with the Administrator for
details concerning the calculation of results when two or
more trains are used.
At the end of the sample run, turn off the coarse adjust
valve, remove the probe and nozzle from the stack, turn
off the pump, record the final dry gas meter reading, and
conduct a post-test leak-check, as outlined in Section
4,1.4.3. Also, leak-check the pilot lines as described in
Method 2, Section 3.1; the lines must pass this leak-check,
in order to validate the velocity head dala.
4.1.6 Calculation of Percent Isokinetic. Calculate
percent isokinetic (see Calculations, Section 6) to deter-
mine whether Ihe run was valid or another test run
should be made. If there wag difficulty in mainlaining
isokinelic rates due to source conditions, consult with
the Administrator for possible variance on the isokinetic
rates.
4.2 Sample Eecovery. Proper cleanup procedure
begins as soon as the probe is removed from the stack at
the end of the sampling period. Allow the probe to cool.
When the probe can be safely handled, wipe off all
external particulate matter near Ihe Up of the probe
nozzle and place a cap over il to prevenl losing or gaining
particulate matter. Do not cap off the probe tip tightly
while the sampling train is cooling down as this would
create a vacuum in the filter holder, thus drawing water
from the impingers into the filter holder.
Before moving the sample train lo Ihe cleanup site,
remove Ihe probe from Ihe sample train, wipe off Ihe
silicone grease, and cap the open outlet of the probe. Bo
careful not to lose any condensato that might be present;
Wipe off the silicone grease from the filter diet where the
probe was fastened and cap it. Kemove the umbilical
cord from the last impiuger and cap the impinger. If a
flexible line is used between the first impinger or con-
denser and the filter holder, disconnect the line at the
344
-------
PLANT
LOCATION
OPERATOR,
DATE
RUN NO
SAMPLE BOX NO..
METER BOX NO. _
METER AH@
C FACTOR
AMBIENT TEMPERATURE.
BAROMETRIC PRESSURE.
ASSUMED MOISTURE, K_
PROBE LENGTH, nv (ft)
PITOT TUBE COEFFICIENT, Cp.
SCHEMATIC OF STACK CROSS SECTION
NOZZLE IDENTIFICATION N0.__
AVERAGE CALIBRATED NOZZLE DIAMETER, em (in.).
PROBE HEATER SETTING
LEAK RATE, m3/min.(cfm)
PROBE LINER MATERIAL
STATIC PRESSURE, mm Hg (in. Hgt
FILTER NO
TRAVERSE POINT
. NUMBER
TOTAL
SAMPLING
TIME
(0). min.
AVERAGE
VACUUM
mmHg
(in. Hg}
STACK
TEMPERATURE
(TS)
•C {«F1
VELOCITY
HEAD
(APsJ.
mmfln.H^O
PRESSURE
DIFFERENTIAL
ACROSS
ORIFICE
METER
mnU20
(in. HaO)
GAS SAMPLE
VOLUME
m3 (ft3)
GAS SAMPLE TEMPERATURE
AT DRY GAS METER
INLET
"C (°F)
Avg.
OUTLET
"C ("Fl
Avg.
Avg.
FILTER HOLDER
TEMPERATURE.
•C ("F)
TEMPERATURE
° OF GAS '
LEAVING
CONDENSER OR
LAST IMPINGER.
CCI°FI
Figure 5-2. Paniculate field data.
345
-------
Plant.
Date
Run No..
Filter No
Amount liquid lost during transport
Acetone blank volume, ml
Acetone wash volume, ml.
Acetone blank concentration, mg/mg (equation 54).
Acetone wash blank, mg (equation 5-5}
CONTAINER
NUMBER
1
2
TOTAL
WEIGHT OF PARTICULATE COLLECTED,
mg
FINAL WEIGHT
I^x^H
TARE WEIGHT
72X^T
Less acetone blank
Weight of parti cu I ate matter
WEIGHT GAIN
.
VOLUME OF LIQUID
WATER COLLECTED
JMPINGER
VOLUME,
ml.
FINAL I
INITIAL
LIQUID COLLECTED
TOTAL VOLUME COLLECTED
SILICA GEL
WEIGHT,
9
g«| ml
* CONVERT WEIGHT OF WATER TO VOLUME BY DIVIDING TOTAL WEIGHT
INCREASE BY DENSITY OF WATER (1g/ml)
INCREASE' 9 - VOLUME WATER, ml
1 g/ml
Figure 5-3. Analytical data.
346
-------
filter holder and let any condensed water or liquid
drain into the impingers or condenser. After wiping oft
the silicone grease, cap off the filler holder outlet and
impinger inlet. Either ground-glass stoppers, plastic.
caps, or serum caps may ha used lo close these openings.
Transfer the probe and filter-impinger assembly to the
cleanup area. This area should be clean and proleete«l
from the wind so that the chances of contaminating or
losing the sample will be minimized.
Save a portion of the acetone, used for i-ltanup a.i :i
blank. Take 200 nil of this acetone directly from the wa-.li
bottle being used and place It in a gl.iss sample citiit.iiij-lr
labeled "acetone Ijluitk."
Inspect the train prior to and timing • li.-.a~-.M'i,il»ly an-l
note any abnormal conditions. Tu-al the ^impli-^ a*
follows:
Container No. 1. Carefully remove the filler from th->
filter holder and place it in its identified petii dish con-
tainer. Use a pair of tweezers and/or clean disposal-^
surgical gloves to handle the Alter. If it is nccf vary to
fold the iilter. do so such that the parliculutb cake it,
inside the fold. Carefully transfer to the petri dish anj
particulato matter and/or filter fibers which adhere to
the filler holder gasket, by using a dry nylon bristle
brush and/or a sharp-edged blade. Seal the container.
Container No. t. Taking care to see that dust on the
outside of the probe or other exterior surfaces does not
get Into the sample, quantitatively recover particulate
matter or any condensate from the probe nozzle, probe
fitting, probe liner, and front half of the tiller holder by
washing these components with acetone and placing the
wash in a glass container. Distilled water may be wed
instead of acetone when approved by the Administrator
and shall be used when specified by tbe Administrator;
in these cases, save a water blank and follow the Admin-
istrator's directions on analysis. Perform tbe acetone
rinses as follows:
('areuilly remove the probe no?.nle and clean the inside
-in face !>y rinsing; with acetone from a wash.boll!'1 and
brushing with a nylon bristle brush, Brush until tlie
neetone rinse shows no visible panicles, after which
make a final nnse of the inside surface with acetone.
Brush and rinse the inside pans of the Swagelnk
fitting with aeelone in a similar way until no visible
panicles remain. *
HinM- the probe li'n-r willi arelone by tilting and
(n!atni3 the prulje while Sfiuiniiifr arHone into its upper
t-nrt so thnt fill in.-itle surfaces will be. wetted willi r.cf-
lune. Let the aci lone drain from the lower end into the
.-ample container. A funnel (glass or polyethylene) may
be used to aid in transferring Ht|iiM washes to the cou-
l.-liner. Follow the acetone rinse with a probe brush.
Hold thu probe in an inclined position, squirt acetone
inlo the upper fnd as the probe l.ru.-li is being pushed
with a twisting action through the probe; hold a sample
container underneath the lover enrl of the probe, and
f-ateh any acetone and paniculate matter which is
brushed from the probe. Him the brush through the
probe three times or more until no visible partu-ulate
matter is carried out with the acetone or until none
remains in the probe liner on visual inspection. With
stainless steel or other mnal probes, run the brash
through in the above pi-scribed manner at least sis
limes since metal prubi-s have small crevices in which
paniculate matte-- can br entrapped. Kinse the brush
with acetone, and quaniiiMiM-ly collect these, washings
m the sample container. After (he brushing, make a
linal acetone rinse of the probe as described above.
it is recommended that two people be used to clean
the probe to minimize sample losses. Between sampling
runs, keep brushes clean and protected from contamina-
tion.
After ensuring that all joints have been wiped clean
of sillcone grease, clean the inside of the front half of t-lie
Iilter holder by rubbing the surfaces with a nylon bristle
brush and rinsing with acetone. Rinse each surface
three times or more if needed to remove visible particu-
lale. Make a final rinse of the brush and filter holder.
Carefully rinse out the glass cyclone, also (if applicable).
After all acetone washings and participate matter have
been collected in the sample container, tighten the lid
on the sample container so that acetone will not leak
out wben it is shipped to the laboratory. Mark the
height of the fluid level to determine whether or not
leakage occurred during transport. Label the container
to clearly identify its contents.
Container No. 3. Note the color of the indicating silica
gel to determine if it has been completely spent and make
a notation of its condition. Transfer the silica gel from
the fourth impinger to its original container and seal.
A funnel may make it easier to pour t lie silica gel wit hoi it
spilling. A rubber policeman may Be used as an aid in
removing the silica gel from the impinger. It is not
necessary to remove the email amount of dust particles
that may adhere to the impinger wall and are difficult
to remove. Bjnce the gain in weight is to be used for
moisture calculations, dp not use any water or other
liquids to transfer the silica gel. If a balance is available
in the field, follow the procedure for container No. 3
in Section 4.3.
Impinger Water. Treat the impingers as follows; Make
a notation of any color or film in the liquid catch. Measure
the liquid which is in the first three impingers to within
*1 ml by using a graduated cylinder or by weighing it
to within ±0.5 g by using a balance (if one is available).
Record the volume or weight of liquid present. This
information is required to calculate the moisture content
of tbe effluent gas.
Discard tbe liquid after measuring and rerouting the
volume or weight, unless analysis of the impinger catch
is required (see Note, Section 2.1.7).
If a different type of condenser is used, measure the
amount of moisture condensed cither volumetrically or
gravimetric-ally.
Whenever possible, containers should be shipped in
such a way that they remain upright at all times.
4.3 Analysis. Record the data required on a sheet
mch as the one shown in Figure 3-3. Handle each sampl-
container as follows:
Container No. 1. Leave the contents in the shipping
container or transfer the filter and any loose particulate
from the sample container to a tared glass weighing dish.
Desiccate for 24 hours in a desiccator containing anhy-
drous calcium sulfate. Weigh to a constant weight and
report the results to the nearest 0.1 mg. For purposes of
this Section. 4.3, the term "constant weight" means a
difference of no more than 0.5 mg or 1 percent of total
weight less tare weight, whichever is greater, between
two consecutive weighings, with no less than 6 hours of
desiccation time between weighings.
Alternatively, the sample may be oven dried at 105° C
(220° F) for 2 to 3 hours, cooled in the desiccator, and
weighed to a constant weight, unless otherwise specified
by the Administrator. The tester may also opt to oven
dry the sample at 105 ° C (220 ° F) for 2 to 3 hours, weigh
the sample, and use this weight as a final weight.
Container No. f. Note the level of liquid in the container
and confirm on the analysis sheet whether or not leakagft
occurred during transport. If a noticeable amount of
leakage has occurred, either void the sample or use
methods, subject to the approval of the Administrator,
to correct the final results. Measure the liquid in this
container either volumetrically to ±1 ml or gravi-
inetrically to ±0.~> g. Transfer the contents to a tared
2"iO-ml beaker and evaporate to dryness at ambient
temperature and pressure. Desiccate for 24 hours and
weigh to a constant weight. Report the results to the
nearest 0.1 mg.
Container Aro. S. Weigh the spent silica gel for silica gel
plus impinger) lo the nearest 0.5 g using a balance. This
step may be conducted in the field.
1Aatone Blank" Container. Measure acetone in this
container either volumetrically or gravimetrically.
Transfer the acetone to a tared 250-ml beaker and evap-
orate to dryness at ambient temperature and pressure.
Desiccate for 24 hours and weigh to a conlsant weight.
Report the results to the nearest 0.1 mg.
NOTE.—At the option of the tester, the contents of
Container No. 2 as well as the acetone blank container
may be evaporated at temperatures higher than ambi-
ent. If evaporation is done at an elevated temperature,
the temperature must bo below the boiling point of the
solvent; also, to preve,nt "bumping," the evaporation
process must be closely supervised, and the contents of
the beakor must be swirled occasionally to maintain an
even temperature. Use extreme care, as acetone is highly
flammable and has a low Hash point.
5. Calibration
Maintain a laboratory log of all calibrations.
5.1 Probe Nozzle. Probe nozzles shall be calibrated
before their initial use in the field. Using a micrometer,
measure the inside diameter of the nozzle to the nearest
0.025 mm (0.001 in.). Make three separate measurements
using different diameters each time, and obtain the aver-
age of the measurements. The difference between the high
and low numbers shall not exceed 0.1 mm (0.004 in.).
When nozzles become nicked, dented, or corroded, they
shall be reshaped, sharpened, and recalibrated before
use. Each nozzle shall be permanently and uniquely
identified.
5.2 Pilot Tube. The Type S pitot tube assembly shall
be calibrated according to the procedure outliued in
Section 4 of Method 2.
5.3 Metering System. Before its initial use in the field,
the metering system shall be calibrated according to the
procedure outlined in AFTD-057C. Instead of physically
adjusting the dry gas meter dial readings to correspond
to the wet test meter readings, calibration factors may be
used to mathematically correct the gas meter dial readings
to the proper values. Before calibrating the metering sys-
tem, it is suggested that a leak-check be conducted.
For metering systems having diaphragm pumps, the
normal leak-check procedure will not detect leakages
within the pump. For these cases the following leak-
check procedure is suggested: make a 10-minute calibra-
tion run at 0.00057 m a/min (0.02 cfm); at the end of the
run, take the difference of the measured wet test meter
and dry gas meter volumes; divide the difference by 10,
to got the leak rate. The leak rate should not exceed
0.00057 m Vmin (0.02 cfm).
After each field use, the calibration of the metering
system shall be checked by performing three calibration
runs at a single, inteririediate orifice setting (based on
tho previous field test), with the vacuum set at the
maximum value reached during the test series. To
adjust the vacuum, insert a valve between the wet test
meter and the inlet of the metering system. Calculate
the average value of the calibration factor. If the calibra-
tion has changed by more than 5 percent, recalibrate
the meter over the full range of orifice settings, as out-
lined in APTD-0576.
Alternative procedures, e.g., using the orifice meter
coefficients, may be used, subject to the approval of the
Administrator.
NOTE.—If t he dry gas meter coefficient values obtained
before and after a test scries differ by more than 5 percent,
the test series shall either be voided, or calculations for
the test series shall bo performed using whichever meter
coefficient value (I.e., before or after) gives the lower
value of total sample volume.
347
-------
5.4 Probe Heater Calibiulion. The- probe heating
system shall be calibrated before Its initial uso in the
field according to the prow-dun- outlined in Al'TI>-OJ76.
Probe.s constructed accordini! to APTD-OSS1 need not
be calibrated if tbe calibration ciirvi-s in APTD-057G
are used.
5.5 Temperature Gauges. Use. the procedure in
Section 4.3 of Method 2 to calibrate in-stack temperature
gauges. Dial thermometers, such as are used for the dry
gas meter and condenser outlet, filial! be calibrated
against mercury-in-glass thermometers.
5.6 Leak Cheek of Metering System Shown in Finuro
5-1. That portion of the sampling train from the pump
to the orifice meter should ba leak checked prior to Initial
use and after each shipment. Leakage after the pump will
result In less volume being recorded than is p.etually
sampled. The following procedure is suggested (see
Figure 5-4): Close the main valve on tbo meter box.
Insert a one-hole rubber stopper with rubber tubing
attached into the orifice exhaust pipe. Disconnect aud
vent the low side of the orifice manometer. Close off tho
low side orifice tap. Pressurize the system to 13 to 18 era
(6 to 7 in.) water e,olumn by blowing into the rubber
tubing. Pinch off tbe tubing and observe the manometer
for one minut». A Ions <-! pressure on tbe manometer
indicates a leak In tbe meter bo^; leaks, if present, must
be corrected.
5.7 Barometer. Calibrate acjinsl a mercury barom-
eter.
6. Calculation!
Carry out calculations, retaining at least one extra
decimal figure beyond t.^*t of the Required data. Hound
oft figures after the fine! calculation. Other-forms of th"
equations may be used as long as Ihey give equivalent
results.
8.1 Nomenclature
A, = Cross-sectional area of nozzle, in- (ft!).
Bf, =Water vunor in the gns stream, proportion
by volume.
C, =Acntono blank residue concentrations, mg/g.
c, =Coiwutratioa of particalate matter in stack
gas, a— basis corrected to standard condi-
tions, g'Mscm fe/dscf).
/ =Percent of isokinetio sampling.
Z. = Maximum acceptable leakage rate (or either a
pret.e.u leak check or for a leak check follow-
ing v. component change; enual to 0.00057
ia»/m'n (O.fH cfro) or 4 percent of the average
sampling rate, whichever is less.
Li = IndivM"al leakage ratfl observed during the
lev* che<*k conducted prior to the "fit"
component change ((=>!, 2, 3 .... ?i),
m3/mjr (cfm).
Lt — T.*eakfwe rate ob?rryed during the post-test
le^k rheck, m8/uiin (cfra).
m« =^otnl amount of paniculate matter collected,
mir.
.Vw =Moleeulir wniirM of wat«r, )S.O g/g-mole
(IS.OlhAh-molej.
m, =Mass of residue of acetone after evaporation,
me.
JPb« = Barometric pressure at the sampling site,
mm Jig (in. HR).
P, "A bsolnte s.tsu* gas pressure, rttm Hg (in. Hg).
Ptu =8tf**'iri\ pbsolutfl pressure 760 mm Hg
(-W.P2 in. H«r).
R =Ideal gas constant, 0.0823/i nun Hg-ms/°K-g-
mole fSl.CS in. Hg-ftyR-lb-roole).
Tm =.Vbsolute fi.verage dry gas meter tempwatura
fp-i PiRnre fr-2), °K (°K).
T, =Ahsolut<> averags stack gas temperature (see
Figure 5--2),1>K(0R).
T.a ^Standard absolute temperature, 293" K
. fKS° R).
V* =Volurrie of acetone blank, ml.
V. a =Volurrte of acetone used in wash, ml.
Vi,=TotM volume of liquid collected in impingers
end silica eel (see Figure 5-S), ml.
V«=1rolm;ie of gas simple as sieasured by dr7 gas
meter, dcm (dcf).
V«(,id)=VoiuTrie of pns sample measured by the dry
R^s meter, corrected to standard conditions,
d'CTO (doC.f).
V'»(«fj)='s'>lunie of water vapor In tbe gas sample,
corrected to standard conditions, scm (set).
V,=Stack gas velocity, calculated by Method 2,
F.quation 2-9, using data obtained from
Method 5, m/sec (ft/sec).
Wt—Weight of residue iii acetone wash, mg.
K=Dry gas meter calibration factor.
AH= Average pressure differential across the orifice
meter (see Fjjnire 6-2), mm HiO (in. HiO).
P«=Donsity of acetorw, mg/ml (s«« label on
bottlB).
p.=rinnsity of wfttei, 0.9932 g/ml (0.002201
\b/ml).
9=Tot,3l sampling time, m'o,
fii = f>amnliug time interval, from the beginning
of a run until the first component change,
jnin.
0<=8prapling time interval, between two suc-
cessive component changes, rieglnning with
the interval between the first and second
changes, min.
0p = SamD)ing time interval, from the final (7i*M
component change until the. end of the
samrilino; run. min
13.G=Specific gravity of mercury
60=Sec/min.
100= Conversion to percent
6.2 Average dry gas meter temperature and average
orifice pressure drop. See data sheet (Figure 5-2).
6.3 Dry Gas Volume. Correct the sample volume
measured by the dry gas meter to standard conditions
(20* C, 760 mm Hg or 68° F, 20.92 in. Hg) by using
Equation 5-1.
Equation 5-1
vhare:
mi=0.3SM °K'mm Hg for metric units
-17.64'R/in. Hg (or English units
NOTE.—Equation £-1 can be used at written unless
the leakage rate observed during any of the mandatory
lt*k checks (I.e., the port-test V-ik cheek or tnk checks
conducted prior to component changes) excmds £.. II
Jk, or /« exceeds JS., Equation 5-1 murt be modified as
follows;
(») Cam I. No component changes made during
mmpllng run. In this case, replace K. in Equation 5-1
Trtth the expression;
(b) Om IT. One or more component changes made
Curing tlie campling run. In this case, replace K. in
Rqoatiort 5-1 by the expression:
Vm-(Ll-L,)e1
i-2
and nibstituU only for those leakage rates (Li or L,)
vhich
Equation 5-2
M Volume of water vapor.
vhere:
JKi=fl.001333 m'/ml for metric units
=t).(H707 .i'/ml for English units.
tJi Moisture Content.
rt ' v (ltd)
Equation 5-3
K.—Tn »tamf«d or w»l«r droplet-laden gas
rtrewron, tro calculntions ot th« moisture content of the
stack gaa shall be made, one from the impinger analysis
(Equation S-3), and a second from the assumption ol
saturated conditions. The lower of the two values ol
B». shall lie considered correct. Tbe procedure tor deter-
mining'the moisture content based upon assumption ot
saturated conditions is given In the Note of Section 1.2
at Method 4. For the purposes of this method, the average
stixjt gas temperature from Figure 6-2 may be used to
roafce this determination, provided that the accuracy ol
the in-fltack temperature sensor is ± 1° C (2° F).
6.6 Acetone Blank Concentration.
6.7 Acetone \Vash Hlanl..
E»;iiaiimi5-4
Equation 5-5
6.8 Total Particular Weight. Determine the total
partlcntote catch from the sum of the weights obtained
from containers 1 and 2 less the acetone blank (see Figure
6-3). NOTE.— Refer to Section 4.1.5 to wsist to calculation
of results Involving two or more filter assemblies or two
or more sampling trains.
8.9 Farticulate Concentration.
e.= (0.001 g/mg) (
6.10 Conveision Factors:
Vm ,.,.„)
Equation 6-t
From
To
Multiply by
&
gr/ft'
lb/ft>
gym"
348
-------
RUBBER
TUBING
RUBBER
STOPPER
ORIFICE
VACUUM
GAUGE
BLOW INTO TUBING
UNTIL MANOMETER
READS 5 TO 7 INCHES
WATER COLUMN
ORIFICE
MANOMETER
AIR-TIGHT
PUMP
Figure 5-4. Leak check of meter box.
349
-------
0.11 Isoklnetic Variation.
6.11.1 Calculation From Raw Data.
loor.t/r.v,.
,) ( Pb.,+ Ag/13.6) ]
Equation 5-7
vbere:
Jfi=0.0034">4 mm Hg-m>;ml-°K for metric units.
-0.002669 in. Hg-ft'/ml-°K tor English unite.
4.11.2 Calculation From IntermediauValues.
___
P.V.A,e(l-B,r.)
vbere:
K i=4.320 for metric units
0.00460 for English units.
Equation 3-8
.
6.12 Acceptable Results. If 90 percent
-------
METHOD 6
DETERMINATION OF SULFUR DIOXIDE EMISSIONS FROM STATIONARY SOURCES
I. Principle and Applicability
1.1 Principle. A gas sample is extracted from tKe
sampling point in the stack. Tbe suifuric acid mist
^including suliur trioxide) and the sulfur dioxide are
separated. The sulfur dioxide traction is measured by
tbe barium-thorin Utratlon method.
1.2 Applicability. This method is applicable for the
determination of sulfur dioxide emissions from stationary
sources. Tbe minimum detectable limit of the method
has been determined to be 3.4 milligrams unpl of SOi'm8
(2.12X10-' Ib/ft'). Although no upper limit has been
established, tests have shown that concentrations as
liigh as 80,000 mg/m» of SOj can be collected efficiently
in two midget impingers, each containing 15 milliliters
of 3 percent hydrogen peroxide, at a rat« of 1.0 1pm for
20 minutes. Based on theoretical calculations, the upper
concentration limit in a 20-liter sample is about 93,300
nig/in'.
Possible interferents are free ammonia, water-soluble
cations, and fluorides. The cations and fluorides are
removed by glass wool niters and an isopropanol bubbler,
and hence do not affect the SOi analysis. When samples
are being taken from a gas stream with high concentra-
tions of very fine metallic fumes (such as in inlets to
control devices), a high-efficiency glass fiber filter must
be used in place of the glass wool plug (i.e., the one in
the probe) to remove the cation interferents.
Free ammonia interferes by reacting with SOi to form
partioulate sulfite and by reacting with the indicator.
If free ammonia is present (this can be determined by
knowledge of the process and noticing white particular*
matter In the probe and isopropauol bubbler"i, alterna-
tive methods, subject to the approval of tbe Administra-
tor, U.S. Environmental PrutHtion Agency, ar«
required.
2. Apparatus
2.1 Sampling. The sampling train is shown in Figure
6-1, and component part* are discussed below. The
taster has tbe option of substituting sampling equip-
ment described in Method 8 in place of the midget im-
pinger equipment of Method 6. However, tbe Method 8
train must be modified to include a heated filter between
the probe and isopropanol impinger, and tbe operation
of toe sampling train and sample analysis must be at
tbe flow rates and solution volumes denned in Method 8.
The tester also has the option of determining SOi
simultaneously with participate matter and moisture
determinations by (1) replacing tbe water in a Method 5
implnger system with 3 percent perioxide solution, or
(2) by replacing the Method 5 water impinger system
with a Method 8 isopropanol-nlter-peroxide system. The
analysis for SO) must be consistent with the procedure
In Method 8.
2.1.1 Probe. Borosilicate glass, or stainless steel (other
materials of construction may be used, subject to the
approval of the Administrator), approximately 6-mjn
inside diameter, with a beating system to prevent water
condensation and a filter (either in-slack or heated out-
stack) to remove partlcnlato matter, including suifuric
acid mist. A plug of glass wool is a satisfactory filter.
2.1.2 Bubbler and Impingers. One midget bubbler,
with medium-coarse glass frit and borosQicate or quartz
glass wool packed In top (see Figure 6-1) to prevent
snlfuric acid mist carryover, and three 30-ml midget
Impingers. Tbe bubbler and midget impingers must be
connected In series with leak-free glass connectors. Sill-
cone grease may be used, if necessary, to prevent leakage.
At the option of the tester, a midget Impinger may be
used in place of the midget bubbler.
Other collection absorbers and flow rates may be used,
hut are subject to the approval of the Administrator.
Also, collection efficiency must be shown to be at least
99 percent for each test run and must be documented in
the report. If the efficiency is found to be acceptable after
a series of three tests, further documentation is not
required. To conduct the efficiency test, an extra ab-
sorber must be added and analyzed separately. This
extra absorber must not contain more than 1 percent of
the total SO«.
2.1.3 Glass WooL Borosilicate or quart*.
2.1.4 Stopcock Grease. Acetone-insoluble, heat-
stable silicone grease may be used, if necessary.
2.1.5 Temperature Gauge. Dial thermometer, or
equivalent, to measure temperature of gas leaving im-
pinger train to within 1° C (2*F.)
2.1.6 Drying Tube. Tube packed with 6- to 16-mesh
Indicating type silica gel, or equivalent, to dry tbe gas
sample and to protect the meter and pump. If the sillac
gel has been used previously, dry at 175° C (350* F) for
2 hours. New silica gel may be used as received. Alterna-
tively, other types of dedccants (equivalent or better)
may be used, subject to approval of the Administrator.
2.1.7 Value. Needle value, to regulate sample gas flow
rate.
2.1.8 Pump. Leak-free diaphragm pump, or equiv-
alent, to pull gas through tbe train. Install a small tank
between the pump aad rate meter to eliminate the
pulsation effect of the diaphragm pump on the rotameter.
2.1.9 Rate Meter. Rotameter, or equivalent, capable
of measuring flow rate to within 2 percent of the selected
flow rate of about 1000 cc/min.
2.1.10 Volume Meter. Dry gas meter, sufficiently
accurate to measure, the sample volume within 2 percent,
calibrated at the selected flow rate and conditions
actually encountered during sampling, and equipped
with a temperature gauge (dial thermometer, or equiv-
alent) capable of measuring temperature to within
3°C (5.48F ).
2.1.11 Barometer. Mercury, ameroid, or other barom-
eter capable of measuring atmospheric pressure to within
2.8 mm Hg (0.1 in. Hg). In many cases, the barometric
reading may be obtained from a nearby national weather
service station, in which case the station value (which
is the absolute barometric pressure) shall be requested
and an adjustment for elevation differences between
the weather station and sampling point shall be applied
at a rate of minus 2.5 mm Hg (0.1 ID. Hg) per 30m (100ft)
elevation increase or vice versa for elevation decrease.
2.1.12 Vacuum Gauge. At least 760 mm Hg (30 in.
Hg) gauge, to be used for leak check of the sampling
train.
2.2 Sample Recovery.
2.2.1 Wash bottles. Polyethylene or glass, 500 ml,
two.
2.2.2 Storage Bottles. Polyethylene, 100 ml, to store
impinger samples (one per sample).
2.3 Analysis.
2.3.1 Pipettes. Volumetric type, 5-ml, 20-ml (one per
sample), and 25-ml sizes.
2.3.2 volumetric Flasks. 100-ml size (one per sample)
and 100-ml sue.
2.3.3 Burettes. 6- and 50-ml sizes.
2.3.4 Erlenmeyer Flasks. 250 ml-aize (one for each
sample, blank, and standard).
2.3.5 Dropping Bottle. 125-ml site, to add indicator.
2.3.6 Graduated Cylinder. 100-ml size.
2.3.7 Spectrophotometer. To measure absorbance at
3S2 nanometers.
3. Btayent*
Unless otherwise Indicated, all reagents must conform
to the specifications established by the Committee on
Analytical Reagents of the American Chemical Society.
Where such specifications are not available, use the best
available grade.
3.1 Sampling.
3.1.1 WaterTbeionized, distilled to conform to A8TM
specification D1U3-74, Type 3. At the option of the
analyst, the KMnO< test for oxidizable organic mattei
may be omitted when high concentrations of organic
matter are not expected to be present.
3.1.2 Isopropanol, 80 percent. Mix 80 ml of isopropanol
with 20 ml of deionized, distilled water. Check each lot ol
Isopropanol for peroxide impurities as follows: shake 10
ml of isopropanol with 10 ml of freshly prepared 10
percent potassium iodide solution. Prepare a blank by
similarly treating 10 ml of distilled water. After 1 minute,
read the absorbance at 362 nanometers on a spectro-
photometer. If absorbance exceeds 0.1, reject alcohol fo*
use.
Peroxides may be removed from isopropanol by redis-
tilling or by passage through a column of activated
alumina; however, reagent grade isopropanol with
suitably low peroxide levels may be obtained from com-
mercial sources. Rejection of contaminated lots may,
therefore, be a more efficient procedure.
3.1.3 Hydrogen Peroxide, 3 Percent. Dilute 30 percent
hydrogen peroxide 1:9 (v/v) with deionized, distilled
351
-------
water (30 ml is needed per sample). Prepare fresh dally
314 Potassium Iodide Solution, 10 Percent. Dissolve
10.6 grams KI in deionised, distilled water and dilute to
100 ml. Prepare when needed.
3.2 Sample Recovery.
3 2 1 Water. Drionked, distilled, as in 3.1.1.
322 Isopropanol,SOPercent.MixSOmlofisopropanol
with 20 ml of deionized, distilled water.
3 3 Analysis.
3.3.1 Water. Deionized, distilled, as in 3.1.1.
3.3.2 Isopropanol, 100 percent.
333 Thorin Indicator. l-(o-arsonophenylazo)-2-
naphthol-3,Misulfonic acid, disodium salt, or equiva-
lent. Dissolve 0.20 g in 100 ml of deionized, distilled
water.
334 Barium Perohlorate Solution, 0.0100 N. Dis-
solve 1.95 g of bariumperchlorate trlhydrate [Ba(ClOOr
SHjO] in 200 ml distilled water and dilute to 1 liter with
sopropanol. Alternatively, 1.22 g of [BaClr2H!O] may
be used instead of the perchlorate. Standardize as in
Section 5.5.
3.3.5 SuMuric Acid Standard, 0.0100 N. Purchase or
standardize to *0.0002 N against 0.0100 N NaOH which
has previously been standardized against potassium
add phthalate (primary standard grade).
4. Procedure.
4.1 Sampling.
4.1.1 Preparation of collection train. Measure 19 ml of
80 percent Isopropanol into the midget bubbler and 15
ml of 3 percent hydrogen peroxide Into each of the first
two midget Impingers. Leave the final midget impinger
dry. Assemble the train as shown In Figure 6-1. Adjust
probe heater to a temperature sufficient to prevent water
condensation. Place crushed ice and water around the
impingers.
4.1.2 Leak-check procedure. A leak check prior to the
sampling run is optional; however, a leak check after the
sampling run is mandatory. The leak-check procedure Is
as follows:
With the probe disconnected, place a vacuum gauge at
the inlet to the bubbler and pull a vacuum of 250 mm
(10 in.) Hg; plug or pinch off the outlet of the now meter,
and then turn oft the pump. The vacuum shall remain
stable for at least 30 seconds. Carefully release the
vacuum gauge before releasing the flow meter end to
prevent bark flow of the Impinger fluid.
Other leak-check procedures may be used, subject to
the approval of the Administrator, U.S. Environmental
Protection Agency. The procedure used in Method 5 is
lot suitable for diaphragm pumps.
4.1.3 Sample collection. Record the initial dry gas
.. .
meter reading and barometric pressure. To begin sam-
pling, position the tip of the probe at the sampling point,
connect the probe to the bubbler, and start the pump.
Adjust the sample flow to a constant rate of ap-
proximately 1.0 Uter/mln as Indicated by the rotameter.
Maintain this constant rate (*10 percent) during the
entire sampling run. Take readings (dry gas meter,
temperatures at dry gas meter and at impinger outlet
and rate meter) at least every 5 minutes. Ada more ice
during the run to keep tile temperature of the gases
leaving the last impinger at 20° C (68° F) or less. At the
conclusion of each run, turn off the pump, remove probe
from the stack, and record the final readings. Conduct a
leak check as in Section 4.1.2. (This leak check Is manda-
tory.) If a leak Is found, void the test run. Drain the ice
bath, and purge the remaining port of the train by draw-
ing clean ambient air through the system for 15 minutes
at the sampling rate.
Clean ambient air can be provided by passing air
through a charcoal filter or through an extra midget
impinger with 15 ml of 3 percent HaOi. The tester may
opt to simply use ambient air, without purification.
4.2 Sample Recovery. Disconnect the Impingers after
purging. Discard the contents of the midget bubbler. Four
the contents of the midget Impingers Into a leak-free
polyethylene bottle for shipment. Rinse the three midget
impingers and the connecting tubes with deionized,
distilled water, and add the washings to the same storage
container. Mark the fluid level. Seal and Identify the
sample container.
4.3 Sample Analysis. Note level of liquid in container,
and confirm whether any sample was lost during ship-
ment; note this on analytical data sheet. If a noticeable
amount of leakage has occurred, either void the sample
or use methods, subject to the approval of the Adminis-
trator, to correct the final results.
Transfer the contents of the storage container to a
100-ml volumetric flask and dilute to exactly 100 ml
with, deionized, distilled water. Pipette a 20-ml aliquot of
this solution into a 250-ml Erienmeyer flask, add 80 ml
of 100 percent Isopropanol and two to four drops of thorin
indicator, and titrate to a pink endpoint using 0.0100 N
barium perchlorate. Repeat and average the tltration
volumes. Run a blank with each series of samples. Repli-
cate tltrations must agree within 1 percent or 0.2 ml,
whichever is larger.
(NOTE.—Protect the 0.0100 N barium perchlorate
solution from evaporation at all times.)
5. Calibration
S.I Metering System.
5.1.1 Initial Calibration. Before its initial use in the
field, first leak check the metering system (drying tube,
needle valve, pump, rotameter, ana dry gas meter) as
follows: place a vacuum gauge at the inlet to the drying
tube and pull a vacuum of 250 mm (10 in.) Hg; plug or
pinch off the outlet or the flow meter, and then turn off
the pump. The vacuum shall remain stable for at least
30 seconds. Carefully release the vacuum gauge before
releasing the flow meter end.
Next, calibrate the metering system (at the sampling
flow rat« specified by the method) as follows: connect
an appropriately sized wet test meter (e-g., 1 liter per
revolution) to the inlet of the drying tube. Make three
Independent calibration runs, using at least five revolu-
tions of the dry gas meter per run. Calculate the calibra-
tion factor, Y (wet test meter calibration volume divided
by the dry gas meter volume, both volumes adjusted to
the same reference temperature and pressure), for each
run, and average the results. If any r value deviates by
more than 2 percent from the average, the metering
system is unacceptable for usa. Otherwise, use the aver-
age as the calibration factor for subsequent test runs.
5.1.2 Post-Test Calibration Check. After each field
test series, conduct a calibration check as in Section 5.1.1
above, except for the following variations: (a) the leak
check Is not to be conducted, (b) three, or more revolu-
tions of the dry gas meter may be used, and (c) only two
independent runs need be made. If the calibration factor
does not deviate by more than 5 percent from the Initial
calibration factor (determined in Section 5.1.1), then the
dry gas meter volumes obtained during the test series
are acceptable. If the calibration factor deviates by more
than 5 percent, recalibrate the metering system as in
Section 5.1.1, and for the calculations, use the calibration
factor (initial or recalibration) that yields the lower gas
volume for each test run.
5.2 Thermometers. Calibrate against merenry-in-
glass thermometers.
5.3 Rotameter. The rotameter need not be calibrated
but should be cleaned and maintained according to the
manufacturer's Instruction.
5.4 Barometer. Calibrate against a mercury barom-
eter.
5.5 Barium Perchlorate Solution. Standardize the
barium perchlorate solution against 25 ml of standard
sulfurie acid toVhlch 100 ml of 100 percent isopropanol
has been added.
6. Calculation*
Carry out calculations, retaining at least one extra
decimal figure beyond that of the acquired data. Round
off figures after final calculation.
6.1 Nomenclature.
C«> = Concentration of sulfur dioxide, dry basis
corrected to standard conditions, mg/dscm
(Ib/dscf).
N=Normality of barium perchlorate tltrant,
mllliequivalents/ml.
Pb.r=Barometric pressure at the exit orifice of the
dry gas meter, mm Hg (in. Hg).
P.td=Standard absolute pressure, 760 mm Hg
(29.92 In. Hg).
rm=Average dry gas meter absolute temperature,
T,id=Standard absolute temperature, 293° K
(528° R).
V0= Volume of sample aliquot titrated, ml.
V»=Dry gas volume as measured by the dry gas
meter, dcm (dcf).
V* (>td)=Dry gas volume measured by the dry gas
meter, corrected to standard conditions,
dscm (dm*).
F,oiD=Total volume of solution in which the sulfur
dioxide sample Is contained, 100 ml.
Vi= Volume of barium perchlorate tltrant used
for the sample, ml (average of replicate
titrations).
Vu=Volume of barium perchlorate titrant used
for the blank, ml.
F= Dry gas meter calibration factor.
32.03=Equivalent weight of sulfur dioxide.
6.2 Dry sample gas volume, corrected to standard
conditions.
Vm(.ld) =
Equation 6-1
where:
Jfi=0.3858 °K/mm HB for metric units.
=•17.64 °R/ln. Hg for English units-
6.3 Sulfur dioxide concentration.
n(atd)
Equation 6-2
where:
Jfi=32.03 mp/meq. for metric units.
=7.061X10-* Ib/meq. for English units.
7. Biblloaraphv
1. Atmospheric Emissions from Sulfuric Acid Manu-
facturing Processes. U.S. DHEW, PHS, Division of Air
Pollution. Public Health Service Publication No.
999-AP-13. Cincinnati, Ohio. 1965.
2. Corbett, P. F. The Determination of SOi and SOi
In Flue Oases. Journal of the Institute of Fuel, tv 237-
243 1961
3! Matty, R. E. and E. K. Dlehl. Measuring Flue-Gas
SOj and SOi. Power. 101:94-97. November 1957.
352
-------
4. Patton. W. Jr. and J. A. Brink, Jr. New Equipment
and Techniques for Sampling Chemical Process Gases.
I. Air Pollution Control Association, li: US. 1903.
5. Rom, J.J. Maintenance. Calibration, and Operation
of Isokinetie Source-Sampling Equipment. Office of
Air Programs, Environmental Protection Agency.
Research Triangle Park, N.C. APTD-0576. March 1972.
6. Hamil, H. F. and D. E. Camann. Collaborative
Study of Method for the Determination of Sulfur Dioxide
Emissions from Stationary Sources (Fossil-Fuel Fired
Steam Generators). Environmental Protection Agency,
Research Triangle Park, N.C. EPA-650/4-74-024.
December 1073.
7. Annual Book of ASTM Standards. Part 31; Water,
Atmospheric Analysis. American Society for Testing
and Materials. Philadelphia, Pa. 1974. pp. 40-42.
8. Knoll, J. E. and M. R. Midgett. The Application ol
EPA Method 6 to High Sulfur Dioxide Concentrations.
Environmental Protection Agency. Research Triangle
Park, N.C. EPA-600/4-76-038. July IOTA.
THERMOMETER
PROBE (END PACKED
WITH QUARTZ OR
PYREX WOOL)
ICE BATH
THERMOMETER
SILICA GEL
DRYING TUBE
PUMP
Figure 6-1. SOg sampling train.
SURGE TANK
353
-------
METHOD 7
DETERMINATION OF NITROGEN OXIDE EMISSIONS FROM STATIONARY SOURCES
1. Prtntipfc and AppUabUilt
1.1 Principle. A grab sample is collected in an evacu-
ated flask containing a dilate salfnric acid-hydrogen
peroxide absorbing solution, and the nitrogen oxides,
except nitrons oxide, are measured colorimeterically
using the phenoldisulfonic acid (PD8) procedure.
1.2 Applicability. This method is applicable to the
measurement of nitrogen oxides emitted from stationary
sources. The range of the method has been determined
to be 2 to 400 milligrams NO, (as NO>) per dry standard
cubic meter, without having to dilute the sample.
Z.Apparotiu
2.1 Sampling (see Figure 7-1). Other grab sampling
systems or equipment, capable of measuring sample
volume to within ±2.0 percent and collecting a sufficient
sample volume to allow analytical reprodncibuity to
within ±5 percent, will be considered acceptable alter-
natives, subject to approval of the Administrator, U.S.
Environmental Protection Agency. The following
equipment is used in sampling:
2.1.1 Probe. Borosilicate glass tubing, sufficiently
heated to prevent water condensation and equipped
with an in-stack or out-stack filter to remove particulate
matter (a plug of glass wool is satisfactory for this
purpose). Stainless steel or Teflon »tubing may also be
used for the probe. Heating is not necessary if the probe
remains dry during the purging period.
2.1.2 Collection Flask. Two-liter borosilicate, round
bottom flask, with short neck and 24/40 standard taper
opening, protected against implosion or breakage.
2.1.3 Flask Valve. T-bore stopcock connected to a
24/40 standard taper Joint.
2.1.4 Temperature Gauge. Dial-type thermometer, or
other temperature gauge, capable of measuring 1? C
(2° F) intervals Irom -5 to SO4 C (25 to 125° F).
2.1.5 Vacuum Line. Tubing capable of withstanding
a vacuum of 75 mm Hg (3 in. Hg) absolute pressure, with
"T" connection and T-bore stopcock.
2.1.G Vacuum Gange. U-tube manometer. 1 meter
(36 in.), with 1-mm (0.1-in.) divisions, or other gauge
capable of measuring pressure to within ±2.5 mm Hg
(0.10 in. Hg).
2.1.7 Pump. Capable of evacuating the collection
flask to a pressure equal to or less than 75 mm Hg (3 in.
Hg) absolute.
2.1.8 Squeeze Bulb. One-way.
2.1.9 Volumetric Pipette. 25 ml.
2.1.10 Stopcock an4 Ground Joint Grease. A high-
vacuum, high-temperature chlorofluorocarbon grease is
required. Halocatbon 25-53 has beenfound to be effective.
2.1.11 Barometer. Mercury, aneroid, or other barom-
eter capable of measuring atmospheric pressure to within
2.5 mm Hg (0.1 in. Hg). In many cases, the barometric
reading may be obtained from a nearby national weather
service station, in which case the station value (which is
the absolute barometric pressure) shall be requested and
an adjustment for elevation differences between the
weather station and sampling point shall be applied at a
rate of minus '<• .1 mm HE (0.1 in. Hg) per 30 m (100 ft)
elevation increase or vice versa for elevation decrease.
2.2 Sample Rc-i overy The following equipment is
required for sample recovery
2.2.1 Graduated Cylinder 50 ml with 1-ml divisions.
2.2 2 Storage Containers. Leak-free polyethylene
bottles.
2.2.3 Wash Bottle. Polyethylene or glass.
2.2.4 Glass Stirring Rod.
2.2.5 Test Paper for Indicating pH. To cover the pH
range of 7 to 14
2.3 Analysis. For the analysis, the following equip-
ment is needed: *
2.3.1 Volumetric Pipettes. Two 1 ml, two 2 ml, one
3 ml, one 4 ml, two 10 ml. and one 25 ml for each sample
and standard.
2.3.2 Porcelain Evaporating Dishes. 175- to 250-ml
capacity with lip lor pouring, one for each sample and
Mention of trade names or specific products does not
constitute endorsement by the Environmental Pro-
tection Agency.
each standard. The Coors No. 45006 (shallow-form, 195
ml) has been found to be satisfactory. Alternatively,
polymethyl pentene beakers (Nalge No. 1203,150ml), or
glass beakers (150 ml) may be used. When glass beakers
are used, etching of the beakers may cause solid matter
to be present in the.-analytical steo. the solids should bo
removed by nitration (see Section 4.3).
2.3.3 Steam Bath. Low-temperature ovens or thermo-
statically controlled hot plates kept below 70° C (160° F)
are acceptable alternatives.
2.3.4 Dropping Pipette or Dropper. Three required.
2.3.5 Polyethylene Policeman. One for each sample
and each standard.
2.3.6 Graduated Cylinder. 100ml with 1-ml divisions.
2.3.7 Volumetric Flasks. 50 ml (one for each sample),
100ml (one for each sample and each standard, and one
for the working standard KNOi solution), and 1000 ml
(one).
2.3.8 Spectrophotometer. To measure absorbance at
410 run.
2.3.9 Graduated Pipette. 10 ml with 0.1-ml divisions.
2.3.10 Test Paper for Indicating pH. To cover the
pH range of 7 to 14.
2.3.11 Analytical Balance. To measure to within 0.1
mg.
3. Rcagenls
Unless otherwise indicated, it is intended that all
reagents conform to the specifications established by the
Committee on Analytical Reagents of the American
Chemical Society, where such specifications are avail
able; otherwise, use the best available grade.
3.1 Sampling. To prepare the absorbing solution,
cautiously add 2.8 ml concentrated HiSOi to 1 liter ol
dcionized, distilled water. Mix well and add 6 ml of 3
percent hydrogen peroxide, freshly prepared from 30
percent hydrogen peroxide solution. The absorbing
solution should be used within 1 week of its preparation.
Do not expose to extreme heat or direct sunlight.
3.2 Sample Recovery. Two reagents are required for
sample recovery:
3.2.1 Sodium Hydroxide (IN). Dissolve 40 g NaOH
in deionized, distilled water and dilute to 1 liter.
3.2.2 Water Deionized, distilled to conform te A8TM
specification D1193-74, Type 3. At the option of the
analyst, the KMNO. test for oxidizable organic matter
may be omitted when high concentrations of organic
matter are not expected to De present.
3.3 Analysis. For the analysis, the following reagents
are required:
3.3.1 Fuming Sulfuric Acid. 15 to 18 percent by weight
free sulfur trioxide. HANDLE WITH CAUTION.
3.3.2 Phenol. White solid.
3.3.3 Sulfuric Acid. Concentrated, 95 percent mini-
mum assay. HANDLE WITH CAUTION.
3.3.4 Potassium Nitrate. Dried at 105 to 110° C (220
to 230° F) for a minimum of 2 hours Just prior to prepara-
tion of standard solution.
3.3.5 Standard KNOi Solution. Dissolve exactly
2.198 g of dried potassium nitrate (KNOi) in deionized,
distilled water and dilute to 1 liter with deionized,
distilled water in a 1,000-ml volumetric flask.
3.3.6 Working Standard KNOj Solution. Dilute 10
ml of the standard solution to 100 ml with deionized
distilled water. One milliliter of the working standard
solution is equivalent to 100 jjg nitrogen dioxide (NO>).
3.3.7 Water. Deionized, distilled as in Section 3.2.2.
3.3.8 Phenoldisulfonic Acid Solution. Dissolve 25 g
of pure white phenol in 150 ml concentrated sulfuric
acid on a steam bath. Cool, add 75 ml fuming sulfuric
acid, and heat at 100° C (212° F) for 2 hours. Store in
a dark, stoppered bottle.
4. Procedure*
4.1 Sampling.
4.1.1 Pipette 25 ml of absorbing solution into a sample
flask, retaining a sufficient quantity for use in preparing
the calibration standards. Insert the flask valve stopper
into the flask with the valve in the "purge" position.
Assemble the sampling train as shown in Figure 7-1
and place the probe at the sampling point. Make sure
that all fittings are tight and leak-free, and that all
ground glass Joints have been properly greased with a
high-vacuum, high-temperature chlorofluorocarbon-
based stopcock grease. Turn the flask valve_and the
354
-------
pump valve to their "evacuate" positions, evacuate
the flask to 75 mm Hg (3 in. Hg) absolute pressure, or
less. Evacuation to a pressure approaching the vapor
pressure of water at the existing temperature is desirable.
Turn the pump valve to its "vent" position and turn
oft toe pump. Check for leakage by observing the ma-
nometer for any pressure fluctuation. (Any variation
greater than 10 mm Hg (0.4 in. Hg) over a period of
1 minute is not acceptable, and the flask is not to be
used until the leakage problem is corrected. Pressure
in the flask is not to exceed 76 mm Hg (3 in. Hg) absolute
at the time sampling Is commenced.) Record the volume
of the flask and valve (V/), the flask temperature (T,),
and the barometric pressure. Turn the flask valve
counterclockwise to its "purge" position and do the
same with the pump valve. Purge the probe and the
vacuum tube using the squeeze bulb. If condensation
occurs in the probe and the flask valve area, beat the
probe and purge until the condensation disappears.
Next, turn the pump valve to its "vent" position. Turn
the flask valve clockwise to its "evacuate" position and
record the difference in the mercury levels in the manom-
eter. The absolute internal pressure in the flask (Pi)
is equal to the barometric pressure less the manometer
reading. Immediately turn tbe flask valve to the "sam-
ple" position and permit tbe gas to enter the flask until
pressures In the flask and sample line (I.e., duct, stack)
are equal. This will usually require about 15 seconds;
a longer period Indicates a "plug" in the probe, which
must be corrected before sampling is continued. After
collecting tbe sample, turn the flask valve to its "purge"
position and disconnect the flask from the sampling
train. Shake the flask for at least 5 minutes.
4.1.2 If the gas being sampled contains insufficient
oxygen for tbe conversion of NO to NO: (e.g., an ap-
plicable subpart of the standard may require taking a
sample of a calibration gas mixture of NO in Nt). then
oxygen shall be introduced into the flask to permit this
conversion. Oxygen may be Introduced into the flask
by one of three methods; (1) Before evacuating the
sampling flask, flush with pure cylinder oxygen, then
evacuate flask to 75 mm Hg (3 in. Hg) absolute pressure
or less; or (2) inject oxygen into the flask after sampling;
or (3) terminate sampling with a minimum of 50 mm
Hg (2 in. Hg) vacuum remaining in the flask, record
this final pressure, and then vent the flask to the at-
mosphere until the flask pressure is almost equal to
atmospheric pressure.
4.2 Sample Recovery. Let the flask set for a minimum
of 16 hours and then shake the contents for 2 minutes.
Connect the flask to a mercury filled U-tube manometer.
Open the valve from the flask to the manometer and
record the flask temperature (T/), the barometric
pressure, and tbe difference between the mercury levels
n the manometer. The absolute internal pressure in
the flask (Pi) is the barometric pressure less the man-
ometer reading. Transfer the contents of the flask to a
leak-free polyethylene bottle. Rinse the flask twice
with 5-ml portions of deionized, distilled water and add
the rinse water to the bottle. Adjust the pH to between
9 and 12 by adding sodium hydroxide (1 N), dropwise
(about 25 to 35 drops). Check the pH by dipping a
stirring rod Into the solution and then touching the rod
to the pH test paper. Remove as little material as possible
during this step. Mark the height of the liquid level so
that the container can be checked for leakage after
transport. Label the container to clearly identify its
contents. Seal the container for shipping.
4.3 Analysis. Note tbe level of the liquid in container
and confirm whether or not any sample was lost during
shipment; note this on the analytical data sheet. If a
noticeable amount of leakage has occurred, either void
the sample or use methods, subject to the approval of
the Administrator, to correct the final results. Immedi-
ately prior to analysis, transfer the contents of tbe
shipping container to a 50-ml volumetric flask, and
rinse the container twice with 5-ml portions of deionized,
distilled water. Add the rinse water to the flask and
dilute to the mark with deionized, distilled water; mix
thoroughly. Pipette a 25-ml aliquot into the procelain
evaporating dish. Return any unused portion of the
sample to tbe polyethylene storage bottle. Evaporate
the 25-ml aliquot to dryness on a steam bath and allow
to cool. Add 2 ml phenoldisulfonic acid solution to the
dried residue and triturate thoroughly with a poylethyl-
ene policeman. Make sure the solution contacts all the
residue. Add 1 ml deionized, distilled water and four
drops of concentrated sulfuric acid. Heat the solution
on a steam bath for 3 minutes with occasional stirring.
Allow the solution to cool, add 20 ml deionized, distilled
water, mix well by stirring, and add concentrated am-
monium hydroxide, dropwise, with constant stirring,
until the pH is 10 (as determined by pH paper). If the
sample contains solids, these must be removed by
nitration (centrifugation is an acceptable alternative,
subject to the approval of the Administrator), as follows:
filter through Whatman No. 41 filter paper into a 100-ml
volumetric flask; rinse the evaporating dish with three
5-ml portions of deionized, distilled water; filter these
three rinses. Wash the filter with at least three 15-ml
portions of deionized, distilled water. Add the filter
washings to the contents of the volumetric flask and
dilute to the mark with deionized, distilled water. If
solids are absent, the solution can be transferred directly
to the 100-ml volumetric flask and diluted to the mark
with deionized, distilled water. Mix the contents of the
flask thoroughly, and measure the absorbance at the
optimum wavelength used for the standards (Section
5.2.1), using the blank solution as a zero reference. Dilute
the sample and the blank with equal volumes of deion-
ized, distilled water if the absorhance exceeds A,, the
absorbance of the 400 = Volflme of flask and valve, ml.
1',, = Volume of absorbing solution, 26 ml.
2=60/26, the aliquot factor. (If other than a 25-ml
aliquot was used for analysis, the correspond-
ing factor must tie substituted).
6.2 Sample volume, dry basis, corrected to standard
conditions.
'tc—~n \y f 'a) l-Trr ~m \
r*.td L 1 1 liJ
where:
, = 0.3858
= 17.64 r
°K
mm Hg
°R
Equation 7-2
for metric units
in. Hg
for English units
355
-------
6.3 Total 11% NOi per sample.
Equation 7-3
NOTK.— If other than a 25-ml aliquot is used for analy-
sis, the factor 2 must be replaced by a corresponding
factor.
6.4 Sample concentration, dry basis, Corrected to
standard conditions.
where:
Equation 7-4
for metric units
=6.243X10-' ~ for English units
1. Standard Methods of Chemical Analysis. 6th ed.
329^330 V™* Nostrand Co- Inc 196a- Vo>- L
2. Standard Method of Test for Oxides of Nitrogen in
Gaseous Combustion Products (Phenoldisnllbnic Add
Procedure). In: 1968 Book of ASTM Standards, Part 26.
* a' 1968' AS™ Desien'>t»™ D-1608-«0,
3. Jacob, M. B. The Chemical Analysis of Air Pollut-
ants. New York. Interscience Publishers, Inc 1960
vol. 10, p. 351-356.
4. Beatty, B. L., L. B. Berger, and H. H. Bchrenk.
Determination of Oxides of Nitrogen by the Pheooldisul-
fonic Acid Method. Bureau of Mines, U.S. Dept. of
Interior. R. I. 3687. February 1943.
5. Bamil, H. F. and D. E. Camann. Collaborative
Study of Method for the Determination of Nitrogen
Oxide Emissions from Stationary Sources (Fossil Fuel-
Fired Steam Generators). Southwest Research Institute
report for Environmental Protection Agency . Research
Triangle Park, N.C. October 5, 1973.
6. Hamil H. F. and R. E. Thomas. Collaborative
Study of Method for the Drtermination of Nitrogen
Oxide Emissions from Stationary Sources (Nitric Acid
Plants). Southwest Research Institute report for En-
vironmental Protection Agency. Research Triangle
Park, N.C. May 8, 1974. ^^
PROBE
FLASK VAC
as_
FILTER
GROUND-GLASS SOCKET.
§ NO. 12/5
f
110mm
3-WAY STOPCOCKr
T-BORE. $ PYREX.
2tnjn BORE. 8-mm OD
GROUl
STANDARD TAPER.
§ SLEEVE NO. 24/40
SQUEEZE BULB
VALVE
PUMP
FLASK SHIELD_ ,\
THERMOMETER
Figure 7-1. Sampling train, flask valve, and flask.
FOAM ENCASEMENT
GROUND-GLASS
SOCKET. § NO. 12/S
PYREX
BOILING FLASK -
2-LITER. ROUND-BOTTOM. SHORT NECK.
WITH I SLEEVE NO. 24/40
356
-------
METHOD 8
DETERMINATION OF SULFURIC ACID MIST AND SULFUR DIOXIDE EMISSIONS
FROM STATIONARY SOURCES
1. Principle and Applicability
1.1 Principle. A gas sample is extracted isokinetically
from the stack. The sulfuric acid mist (including salfur
trioiide) and the sulfur dioxide are separated, and both
fractions are measured separately by the barium-therm
titrarjon method.
1.2 Applicability. This method is applicable for the
determination of sulfuric acid mist (including sulfur
trioxide, and in the absence of other paniculate matter)
and sulfur dioxide emissions from stationary sources.
Collaborative tests have shown that the minimum
detectable limits of the method are 0.05 milligrams/cubic
meter (0.03X10-' pounds/cubic foot) for sulfur trioiide
and 1.2 nig/m' (0.74 10-' lb/ft») for sulfur dioxide. No
upper limits have been established. Based on theoretical
calculations for 200 mUliliters of 3 percent hydrogen
peroxide solution, the upper concentration limit for
sulfur dioxide in a 1.0 m3 (35.3 ft1) gas sample is about
12,500 rag/mi (7.TX10-* lb/ft'). The upper limit can be
extended by increasing the quantity of peroxide solution
in the impingers.
Possible interfering agents of this method are fluorides,
free ammonia, and dimethyl aniline. If any of these
interfering agents are present (this can bo determined by
knowledge of the process), alternative methods, subject
to the approval of the Administrator, are required.
Filterable participate matter may be determined along
with SO] and SOi (subject to the approval of the Ad-
ministrator) ; however, the procedure used for paniculate
matter must be consistent with the specifications and
procedures given in Method 5.
2. Apparatus
2.1 Sampling. A schematic of the sampling train
used in this method Is shown in Figure 8-1; ft is similar
to the Method 5 train except that the filter position Is
different and the filter holder does not have to be heated.
Commercial models of this train are available. For those
who desire to build their own, however, complete con-
struction details are described in APTD-O581. Changes
from the APTD-0581 document and allowable modi-
fications to Figure 8-1 are discussed in the following
subsections.
The operating and maintenance procedures for the
sampling train are described in APTI)-0376. Since correct
usage Is important in obtaining valid results, all users
should read the APTD-0576 document, and adopt the
operating and maintenance procedures outlined in it,
unless otherwise specified herein. Further details and
guidelines on operation and maintenance aru given in
Method 5 and should be read and followed whenever
they are applicable.
2.1.1 Probe Nozzle. Same as Method 5, Section y.l.l.
2.1.2 Probe Liner. Borosilicate or quartz glass, with a
heating system to prevent visible condensation during
sampling. Do not use metal probe liners.
2.1.3 Pilot Tube. Same as Method 5. Section 2.1.3.
2.1.4 Differential Pressure Qange. 8»me as Method 5,
Section 2.1.4.
2.1.5 Filter Holder. BorosUkato glass, with a glass
frit filter support and a sUlcone rubber gasket. Other
gasket materials, e.g., Teflon or Vlton, may be used sub-
ject to the approval of the Administrator. The holder
design shall provide a positive seal against leakage from
the outside or around the filter. The filter holder shall
be placed between the first and second Impingers. Note:
Do not heat the filter holder.
2.1.6 Impingers-rFour, as shown in Figure 8-1. The
first and third shall be of the Oreenburg-Smlth design
with standard tips. The second and fourth shall be of
the Oreenburg-Smltb design, modified by replacing the
Insert with an approximately 13 millimeter (0.5 in.) ID
glass tube, having an unconstricted tip located 13 mm
(0.5 in.) from the bottom of the flask. Similar collection
systems, which have been approved by the Adminis-
trator, may be used.
2.1.7 Metering System. Same as Method 5, Section
2.1.8 Barometer. Same as Method 5. Section 2.1.9.
2.1.9 Qas Density Determination Equipment. Same
as Method 5, Section 2.1.10.
2.1.10 Temperature Gauge. Thermometer, or equiva-
lent, to measure the temperature of the gas leaving the
impinger train to within 1° C (2° F).
2.2 Sample Recovery.
2.2.1 Wash Bottles. Polyethylene or glass, 500 mL
(two).
2.2.2 Graduated Cylinders. 250 ml. 1 liter. (Volu-
metric flasks may also be used.)
2.2.3 Storage Bottles. Leak-free polyethylene bottles,
1000 ml size (two for each munpiing run).
2JS.4 Trip Balance. SOOgmm capacity, to measure to
±0.5 £ (necessary only If a moisture content analysis is
to be done).
2.3 Analysis.
2.3.1 Pipettes. Volumetric 25 ml, 100ml.
2.3.2 Burette. 50 ml.
2.3.3 Brienmeyer Flask. 250 ml. (one for each sample
blank and standard).
2A4 Graduated Cylinder. 100ml.
2.3.5 Trip Balance. 500 g capacity, to measure to
±0.5 g.
2.3.6 Dropping Bottle. To add Indicator solution,
125-ml site.
Unless otherwise Indicated, all reagents are to conform
to the specifications established by the Committee on
Analytical Reagents of the American Chemical Society,
where such specifications an available. Otherwise, use
the best available grade.
3.1 Sampling.
3.1.1 Filters. Same as Method 5, Section 3.1.1.
3.1.2 Silica Gel. Same as Method 5, Section 3.1.2.
3.1.3 Water. Delonieed, distilled to conform to ASTM
specification D1193-74, Type 3. At the option of the
analyst, the KMnO4 test for oxidkable organic matter
may be omitted when high concentrations of organic
matter are not expected to be present.
3.1.4 Isopropanol. 80 Percent. Mix 800 ml of isopro-
nanol with 200 ml of delonbed, distilled water.
Non.— Experience has shown that only A.C.8. grade
Isopropanol Is satisfactory. Tests have shown that
isopropanol obtained from commercial sources occa-
cadonally has peroxide Impurities that will cause er-
roneously high sulfnrio add mist measurement. Use
the following test for detecting peroxides In each lot of
isopropanol: Shake 10 ml of the Isopropanol with 10 ml
of freshly prepared 10 percent potassium iodide solution.
Prepare a blank by similarly treating 10 ml of distilled
water. After 1 minute, read the absorbance on a spectro-
pbotometer at 352 nanometers. If the absorbance exceeds
0.1, the isopropanol shall not be used. Peroxides may be
removed from isopropanol by redistilling, or by passage
through a column of activated alumina. However, re-
agen tirade isopropanol with suitably low peroxide levels
is readily available from commercial sources; therefore,
rejection of contaminated lots may be more efficient
than following the peroxide removal procedure.
3.1.5 Hydrogen Peroxide, 3 Percent. Dilute 100 ml
of 30 percent hydrogen peroxide to 1 liter with delonlted,
distilled water. Prepare fresh daily.
3.1.0 Crushed Ice.
3.2 Sample Recovery.
3.2.1 Water. Same as 3.1.3.
3.2.2 Isopropanol, 80 Percent. Same as 3.1.4.
3.3 Analysis.
3.3.1 Water. Same as 3.1.3.
3.3.2 Isopropanol, 100 Percent.
3.3.3 Thorin Indicator. l-(o-arsonophenylazo)-2-naph-
thol-3, 6-dfsulfonlc acid, dlsodlum salt, or equivalent.
Dissolve 0.20 g in 100 ml of deionlzed, distilled water.
3.3.4 Barium Perchlorate (0.0100 Normal). Dissolve
1. 95 g of barium perchlorata trlhydrate(Ba(C10i)i-3HiO)
in 200 ml deionized, distilled water, and dilute to 1 liter
with Isopropanol; 1.22 g of barium chloride dihydrate
(BaClj-2HiO) may be used Instead of the barium per-
chlorate. Standardize with sulfuric acid as in Section 5.2.
This solution must be protected against evaporation at
all times.
3 3.5 Sulfuric Acid Standard (0.0100 N). Purchase or
standardize to ±0.0002 N against. 0.0100 N NaOH that
has previously been standardized against primary
standard potassium add phthalate.
357
-------
"fe
7
REVERSE TYPE
PITOT TUBE
TEMPERATURE SENSOR
PROBE
PITOT TUBE
TEMPERATURE SENSOR
FILTER HOLDER
VACUUM
LINE
VACUUM
GAUGE
MAIN VALVE
DRY TEST METER
Figure 8-1. Su If uric acid mist sampling train.
358
-------
4. Procedure
4.1 Sampling.
4.1.1 Pretest Preparation. Follow the procedure out-
lined In Method 5, Section 4.1.1; filters should be in-
spected, hot need not be desiccated, weighed, or identi-
fied. If the effluent gas can be considered dry, I.e., mois-
ture free,, the silica gel need not be weighed.
4.12 Preliminary Determinations. Follow the pro-
cedure outlined in Method 5, Section 4.1.2.
4.1.S Preparation of Collection Train. Follow the pro-
cedure outlined in Method 5, Section 4.1.3 (except for
the second paragraph and other obviously inapplicable
parts) and use Figure 8-1 instead of Figure 5-1. Replace
the second paragraph with: Place 100 ml of 80 percent
isopropanol in the first impinger, 100 ml of 3 percent
hydrogen peroxide in both the second and third im-
pingers; retain a portion of each reagent for use as a
blank solution. Place about 200 g of silica gel in the fourth
impinger.
NOTI.—If moisture content is to be determined by
impinger analysis, weigh each of the first three impingers
(plus absorbing solution) to the nearest 0.5 g and record
these welghts/The weight of the silica gel (or silica gel
plus container) must abo be determined to the nearest
0.5 g and recorded.
4.1.4 Pretest Leak-Check Procedure. Follow the
basic procedure outlined In Method 5, Section 4.1.4.1,
noting that the probe heater shall be adjusted to the
Tnirjimnin temperature required to prevent condensa-
tion, and abo that verbage such as, ' * * plugging the
inlet to the filter holder * * *," shall be replaced by,
"• * * plugging the inlet to the first Impinger • • *."
The pratestleak-check is optional.
4.1.5 Train Operation. Follow the basic procedures
outlined in Method 5, Section 4.1.5, in conjunction with
the following special instructions. Data shall be recorded
on a sheet similar to the one hi Figure 8-2. The sampling
rate shall not exceed 0.030 m'/min (l.Ocfm) during the
run. Periodically during the test, observe the connecting
line between the probe and first impinger for signs of
condensation. If it does occur, adjust the probe heater
setting upward to the minimum temperature required
to prevent condensation. If component changes become
necessary during a run, a leak-check shall be done im-
mediately before each change, according to the procedure
outlined in Section 4.1.4.2 of Method 5 (with appropriate
modifications, as mentioned in Section 4.1.4 of this
method); record all leak rates. If the leakage rate(s)
exceed the specified rate, the tester shall either void the
run or shall plan to correct the sample volume as out-
lined in Section 6.3 of Method 5. Immediately after com-
ponent changes, leak-checks are optional. If these
leak-checks are done, the procedure outlined in Section
4.1.4.1 qf Method 5 (with appropriate modifications)
shall be used.
After turning off the pump and recording the final
readings at the conclusion of each run, remove the probe
from the stack. Conduct a post-test (mandatory) leak-
check as in Section 4.1.4.3 of Method 5 (with appropriate
modification) and record the leak rate. If the post-test
leakage rate exceeds the specified acceptable rate, the
tester shall either correct the sample volume, as outlined
in Section 6.3 of Method 5, or shall void the run.
Drain the ice bath and, with the probe disconnected,
purge the remaining part of the train, by drawing clean
ambient air through the system for 15 minutes at the
average flow rate used for sampling.
NOTE.—Clean ambient air can be provided by passing
air through a charcoal filter. At the option of the tester,
ambient ah- (without cleaning) may be used. ,
4.1.6 Calculation of Percent Isokinetic. Follow the
procedure outlined in Method 5, Section 4.1.6.
4.2 Sample Recovery.'
4.2.1 Container No. 1. If a moisture content analysis
is to be done, weigh the first impinger plus contents to
the nearest 0.5 g and record this weight.
Transfer the contents of the first impinger to a 250-ml
graduated cylinder. Rinse the probe, first impinger, all
connecting glassware before the filter, and the front half
of the filter holder with 80 percent isopropanol. Add the
rinse solution to the cylinder. Dilute to 250 ml with 80
percent isopropanol. Add the filter to the solution, mix,
and transfer to the storage container. Protect the solution
against evaporation. Mark the level of liquid on bet
container and identify the sample container.
4.2.2 Container No."2. If a moisture content analysis
is to be done, weigh the second and third Impingers
(plus contents) to the nearest 0.5 g and record these
weights. Abo, weigh the spent silica gel (or silica gel
plus impinger) to the nearest 0.5 g.
Transfer the solutions from the second and third
Impingers to a 1000-ml graduated cylinder. Rinse all
connecting glassware (including back half of filter holder)
between the filter and silica geTimpinger with deionfzed,
distilled water, and add this rinse water to the cylinder.
Dilute to a volume of 1000 ml with deionized distilled
water. Transfer the solution to a storage container Mark
the level of liquid on the container. Seal and identify the
sample container.
4.3 Analysis.
Note the level of liquid in containers 1 and 2, and con-
firm whether or not any sample was lost during ship-
ment; note this on the analytical data sheet. If a notice-
able amount of leakage has occurred, either void the
sample or use methods, subject to the approval of the
Administrator, to correct the final results.
4.3.1 Container No. 1. Shake the container holding
the Isopropanol solution and the filter. If the filter
breaks up, allow the fragments to settle for a few minutes
before removing a sample. Pipette a 100-ml aliquot of
this solution into a 250-ml Erlenmeyer flask, add 2 to 4
drops of thorin indicator, and titrate to a pink endpoint
using 0.0100 N barium perchlorate. Repeat the titration
with a second aliquot of sample and average the titration
Tallies. Replicate Utrations must agree within 1 percent
or 03 ml, whichever Is greater.
4*3 Container No. 1. Thoroughly mix the solution
in the container holding the content* of the second and
third impingen. Pipette a 10-ml aliquot of sample Into a
260-ml Krtenmeypr flask. Add ml of Isopropanol, 2 to
4 drops of thoriu Indicator, and titrate to a pink endpoint
using 0.0100 N barium perchlonie. Repeal the titration
with a second aliquot of sample sad average the titration
value*. Replicate tttratlons must agree wrtiun 1 percent
or 03 ml, whichever b greater.
5. OaUbnOon
5.1 Calibrate equipment using the procedures speci-
fied In the following sections of"Method fi: Section &3
(metering system); Section &fi (temperature gauges);
Section 5.7 (barometer). Note that the recommended
leak-check of the metering system, described hi Section
&fl of Method «, also applies to this method.
6.2 BtandardUe the barium perchlorate solution with
25 ml of standard snlfuric sddVto which 100 ml of 100
percent isopropanol has been added.
6, CttUvlattont
Note.—Carry out calculations retaining at least one
extra decimal figure beyond that of the acquired date
Bound off figures after final calculatiqn.
64 Nomenclature.
•d.-Cross-aectional ana of noule, m» (fP).
B«-Water vapor in the gas stream, proportion
by volume.
CHiSO,=>8ullnrIc acid (Including SO,) concentration,
g/dsom (Ib/dscf).
CTSOt'oSulfnr HhpAA* concentration, g/dsem (lb/
dsd),
/»Percent of Isokinetic sampling.
^= Normality of barium perchlorate titrant, g
equivalents/liter.
Pbar—Barometric pressure at the ;»mpHng site,
mm Hg On. Hg).
P.-Absorate stack gas pressure, mm Hg (in.
Pstd-Btandard absolute presBure, TOO mm Hg
(29.92 in. Hg).
T«—Average absolute dry gas meter temperature
(see Figure 8-2), • K(° K).
T.=* Average absolute stack gas temperature (see
Figure 8-Z), ° K C" B).
Tstd-Btandard absolute temperature, 293° K
F.=Votame of sample aliquot titrated, 100 ml
for HtSOi audio ml for SOi.
Vi ,=Total volume of liquid collected In implngen
and silica gel, ml.
V_=Volume of gas sample as measured by dry
gas meter, dcm (del).
V,(std)=Volume of gas sample measured by the dry
gas meter corrected to standard conditions,
dscm (dscf).
r,>= Average stack gas velocity, calculated by
Method 2. Equation 2-9. using data obtained
from Method 8, m/sec (ft/sec).
Fsoln=Total volume of solution in which the
snlfurio acid or sulfur dioxide sample is
contained, 250 ml or 1,000 ml, respectively.
V,=Volnme of barium perchlorate titrant used
for the sample, ml.
Vi»=Volume of barium perchlorate titrant used
for the blank, ml;
K= Dry gas meter calibration factor.
AH=Average pressure drop across orifice meter,
mm (in.) HiO.
e=Total sampling time, min.
13.6=-8peoiflo gravity of mercury.
ou=sec/min.
100=Converaion to percent.
6J Average dry gas meter temperature and avenge
orifice pressure drop. See data sheet (Figure 8-2).
6J Dry Qas Volume. Correct the sample volume
measured by the dry gas meter to standard conditions
(30° C and TOO mm Hg or 68* F and 29.92 in. Hg) by using
Equation 8-1.
=
•*«
Equation 8-1
359
-------
PLANT.
LOCATION
OPERATOR
DATE
RUN NO
SAMPLE BOX NO..
METER BOX NO. _
METER A Hp
C FACTOR
PITOT TUBE COEFFICIENT, Cp.
STATIC PRESSURE, mm H| (m. H|)
AMBIENT TEMPERATURE
BAROMETRIC PRESSURE
ASSUMED MOISTURE, X
PROBE LENGTH,™ (ft)
SCHEMATIC OF STACK CROSS SECTION
NOZZLE IDENTIFICATION NO
AVERAGE CALIBRATED NOZZLE DIAMETER, em(in.).
PROBE HEATER SETTING
LEAK RATE,m3/min,(dm)
PROBE LINER MATERIAL
FILTER NO.
TRAVERSE POINT
NUMBER
TOTAL
SAMPLING
TIME
(ei.ini*.
AVERAGE
VACUUM
nmH|
(hi. H|)
STACK
TEMPERATURE
(Ts).
°C(»F)
VELOCITY
HEAD
-------
where:
JTi<*o.385g "K/nun Hg for metric units.
-17.64«R/ta. HK for English units.
NOTK.—If the leak rate observed during any manda-
tory leak-checks exceeds the specified acceptable rate,
the tester shall either correct the value of Vm In Equation
8-1 (as described In Section 6.3 of Method 5), or shall
Invalidate the test run.
0.4 Volume of Water Vapor and Moisture Content.
Calculate the volume of water vapor using Equation
5-2 of Method 5: the weight of water collected in the
impingerB and silica gel can be directly converted to
mfflillters (the specific gravity of water Is 1 g/ml). Cal-
culate the moisture content of the stack gas. using Equa-
tion 6~a of Method 6. The "Note" In Section 6.5 of Method
5 also applies to this method. Note that If the effluent gas
stream can be considered dry, the volume of water vapor
and moisture content need not be calculated.
6.6 Sulfurle acid mist (Including SOi) concentration.
ff(Vt~Vu,)
Equation 8-2
where:
£Ti=0.04S04 g/milllequivalent for metric units.
=1.081X10-" Ib/meq lor English units.
8.6 Sulfur dioxide concentration.
N(Vt-Vlt)
Equation 8-3
where:
£i=0.n3203 i/men for metric units.
=7.«51X10-«lD/ineq for English units.
6.7 laokinetlc Variation.
6.7.1 Calculation from raw data.
, 100 T.(Kt Vle+ (VJT*) P
600V. P. A
-f- Ag/13.6)]
Equation 8-1
where:
JTi=-0.0034M nun Hg-m'/ml-°K for metric units.
=0.002676 in. He-ft*/ml-°B for English units.
6.7.2 Calculation from Intermediate values.
r r.V,(.M)P.,,|100
_ K"
" §
^ (,ul)
Equation 8-5
where:
Jfi=4.320 tor metric units.
-0.09450 for English units.
6.8 Acceptable Results. If 90 percent and SO>
in Flue Oases. Journal of the Institute of Fuel. l£237-243.
1SS1.
3. Martin, Robert M. Construction Details of Isoktnetic
Source Sampling Equipment. Environmental Protection
Agency. Research Triangle Park, N.C. Air Pollution
Control Office Publication No. APTD-OS81. April, 1971.
4. Fatten, W. F. and 3. A. Brink, Jr. New Equipment
and Techniques for Sampling Chemical Process Gases.
Journal of Air Pollution Control Association. 13:IW. 1963.
5. Bom, J. J. Maintenance, Calibration, and Operation
of Isoklnetic Source-Sampling Equipment. Office of
Air Programs, Environmental Protection Agency.
Research Triangle Park, N.C. APTD-0576. March, 1972.
6. Hamil, H. F. and D. E. Camann. Collaborative
Study of Method tor Determination of Sulfur Dioxide
Emissions from Stationary Sources (Fossil Fuel-Fired
Steam Generators). Environmental Protection Agency.
Research Triangle Park, N.C. EPA-650/4-74-024.
December, 1973.
7. Annual Book of A8TM Standards. Part 31; Water,
Atmospheric Analysis, pp. 40-42. American Society
for Testing and Materials. Philadelphia, Pa. 1974.
361
-------
METHOD 9
VISUAL DETERMINATION OF THE OPACITY OF EMISSIONS FROM
STATIONARY SOURCES
Many stationary sources discharge visible
emissions Into the atmosphere: these emis-
sions are usually in the shape of a p!ume.
Tills method Involves the determination of
plume opacity by qualified observers. The
method includes procedures for the training
and certification of observers, and procedures
to be used In the field for determination of
plume opacity. The appearance of a plume as
viewed by an observer depends upon a num-
ber of variables, some of which may be con-
trollable and some of which may not be
controllable in the field. Variables which can
be controlled to an extent to which they 110
longer exert a significant influence upoa
plume appearance Include: Angle of the ob-
server with respect to the plume; angle of the
observer with respect to the sun; point of
observation of attached and detached steam
plume; and angle of the observer with re-
spect to a plume emitted from a rectangular
stack with a large length to width ratio. The
method Includes specific criteria applicable
to these variables.
Other variables which may not be control-
lable In the field are luminescence and color
contrast between the plume and the back-
ground against which the plume Is viewed.
These variables exert an Influence upon the
appearance of a plume as viewed by an ob-
server, and can affect the ability of the ob-
server to accurately assign opacity values
to the observed plume. Studies of the theory
of plume opacity and field studies have dem-
onstrated that a plume Is most visible and
presents the greatest apparent opacity when
viewed against a contrasting background. It
follows from this, and Is confirmed by field
trials, that the opacity of a plume, viewed
under conditions where a contrasting back-
ground Is present can be assigned with the
greatest degree of accuracy. However, the po-
tential for a positive error Is also the greatest
when a plume Is viewed under such contrast-
Ing conditions. Under conditions presenting
a less contrasting background, the apparent
opacity of a plume Is less and approaches
zero as the color and luminescence contrast
decrease toward zero. As a result, significant
negative bias and negative errors can be
made when a plume Is viewed under less
contrasting conditions. A negative bias de-
creases rather than Increases the possibility
that a plant operator will be cited for a vio-
lation of opacity standards due to observer
error
Studies have been undertaken to determine
the magnitude of positive errors which can
be made by qualified observers while read-
Ing plumes under contrasting conditions and
using the procedures set forth In this
method. The results of these studies (field
trials) which involve a. total of 769 sets of
25 readings each are as follows:
1 For a set, positive error=average opacity
determined by observers' 25 observations —
average opacity determined from transmls-
someter's 25 recordings.
(1) For black plumes (133 sets at a smoke
generator), 100 percent of the sets were
read with a positive error1 of less than 7.5
percent opacity; 99 percent were read with
a positive error of less than 5 percent opacity.
(2) For white plumes (170 sets at a smoke
generator, 168 sets at a coal-fired power plant,
298 sets at a sulfurlc acid plant), 99 percent
of the sets were read with a positive error of
less than 7.5 percent opacity; 95 percent were
read with a positive error of less than 5 per-
cent opacity.
The positive observational error associated
with an average of twenty-five readings Is
therefore established. The accuracy of the
method must he taken Into account when
determining possible violations of appli-
cable opacity standards.
1. Principle and applicability.
1.1 Principle. The opacity of emissions
from stationary sources Is determined vis-
ually by a qualified observer.
1.2 Applicability. This method Is appli-
cable for the determination of the opacity
of emissions from stationary sources pur-
suant to 5 60.11 (b) and for qualifying ob-
servers for visually determining opacity of
emissions.
2. Procedures. The observer qualified In
accordance with paragraph 3 of this method
shall use the following procedures for vis-
ually determining the opacity of emissions:
2.1 Position. The qualified observer shall
stand at a distance sufficient to provide a
clear view of the emissions with the sun
oriented In the 140° sector to his back. Con-
sistent with maintaining the above require-
ment, the observer shall, as much as possible,
make his observations from a position such
that his line of vision Is approximately
perpendicular to the plume direction, and
when observing opacity of emissions from
rectangular outlets (e.g. roof monitors, open
baghouses, nonclrcular stacks), approxi-
mately perpendicular to the longer axis of
the outlet. The observer's line of sight should
not Include more than one plume at a time
when multiple stacks are Involved, and In
any case the observer should make his ob-
servations with his line of sight perpendicu-
lar to the longer axis of such a set of multi-
ple stacks (e.g. stub stacks on baghouses).
2.2 Field records. The observer shall re-
cord the name of the plant, emission loca-
tion, type facility, observer's name and
affiliation, and the date on a field data sheet
(Figure 9—1). The time, estimated distance
to the emission location, approximate wind
direction, estimated wind speed, description
of the sky condition (presence and color of
clouds), and plume background are recorded
on a field data sheet at the time opacity read-
tags are Initiated and completed.
2.3 Observations. Opacity observations
shall be made at the point of greatest opacity
in that portion of the plume where con-
densed water vapor Is not present. The ob-
server shall not look continuously at the
plume, but Instead shall observe the plume
momentarily at 15-second Intervals.
362
-------
2.3.1 Attached steam plumes. When con-
densed water vapor Is present within the
plume as It emerges from the emission out-
let, opiclty observations shall be made be-
yond the point In the plume at which con-
densed water vapor Is no longer visible. The
observer shall record the approximate dis-
tance from the emission outlet to the point
In the plume at which the observations are
made.
2.3.2 Detached steam plume. When water
vapor In the plume condenses and becomes
visible at a distinct distance from the emis-
sion outlet, the opacity of emissions should
be evaluated at the emission outlet prior to
the condensation of water vapor and the for-
mation of the steam plume.
2.4 Recording observations. Opacity ob-
servations shall be recorded to the nearest 5
percent at 15-second Intervals on an ob-
servational record sheet. (See Figure 9-2 for
an example.) A minimum of 24 observations
shall be recorded. Each momentary observa-
tion recorded shall be deemed to represent
the average opacity of emissions for a 15-
seconti period.
2.5 Data Reduction. Opacity shall be de-
termined as an average of 24 consecutive
observations recorded at 15-second intervals.
Divide the observations recorded on the rec-
ord sheet Into sets of 24 consecutive obser-
vations. A set is composed of any 24 con-
secutive observations. Sets need not be con-
secutive in time and in no case shall two
sets overlap. For each set of 24 observations,
calculate the average by summing the opacity
of the 24 observations and dividing this sum
by 24. If an applicable standard specifies an
averaging time requiring more than 24 ob-
servations, calculate the average for all ob-
servations made during the specified time
period. Record the average opacity on a record
sheet. (See Figure 9-1 for an example.)
3. Qualifications and testing.
3.1 Certification requirements. To receive
certification as a qualified observer, a can-
didate must be tested and demonstrate the
ability to assign opacity readings in 5 percent
Increments to 25 different black plumes and
25 different white plumes, with an error
not to exceed 15 percent opacity on any one
reading and an average error not to exceed
7.5 percent opacity in each category Candi-
dates shall be tested according to the pro-
cedures described in paragraph 3.2. Smoke
generators used pv.rsuant to paragraph 3.2
shall be equipped with a smoke meter which
meets the requirements of paragraph 3.3.
The certification shall be valid for a period
of 6 months, at which time the qualification
procedure must be repeated by any observer
in order to retain certification.
3 2 Certification procedure. The certifica-
tion test consists of showing the candidate a
complete run of 50 plumes—25 black plumes
and 25 white plumes—generated by a smoke
generator. Plumes within each set of 25 black
and 25 white runs shall be presented In ran-
dom order. The candidate assigns an opacity
value to each plume and records his obser-
vation on a suitable form. At the completion
of each run of 50 readings, the score of the
candidate is determined. If a candidate falls
to qualify, the complete run of 50 readings
must be repeated In any retest. The smoke
test may be administered as part of a smoke
school or training program, and may be pre-
ceded by training or familiarization runs of
the smoke generator during which candidates
are shown black and white plumes of known
opacity.
3.3 Smoke generator specifications. Anv
smoke generator used for the purposes of
paragraph 3.2 shall be equipped with a smnke
meter installed to measure opacity across
the diameter of the smoke generator stack.
The smoke meter output shall display In-
siack opacity based upon a pathlength equal
to the stack exit diameter, on a full 0 to 100
percent chart recorder scale. The smnke
meter optical design c.nd performance shall
meet the specifications shown In Table 9-1.
The smoke me-ter shall be calibrated as pre-
scribed in paragraph 3.3.1 prior to the con-
duct of each smoke reading test. At the
completion of each test, the zero and span
drift shall be checked and if the drift ex-
ceeds =tl percent opacity, the condition shall
be corrected prior to conducting any subse-
quent test runs. The smoke meter shall be
demonstrated, at the time of Installation, to
meet the specifications listed in Table 9-1.
This demonstration shall be repeated fol-
lowing any subsequent repair or replacement
of the photocell or associated electronic cir-
cuitry including the chart recorder or output
meter, or every 6 months, whichever occurs
first.
TABLE 9-1 SMOKE METER DESIGN AND
PERFORMANCE SPECIFICATIONS
Parameter: Specification
a. Light source Incandescent lamp
operated at nominal
rated voltage.
b. Spectral response Photopic (daylight
of photocell. spectral response of
the human eye—
reference 4.3).
c. Angle of view 15° maximum total
angle.
d. Angle of projec- 15° maximum total
tlon. angle.
e. Calibration error- ±3% opacity, maxi-
mum.
1. Zero and span ±1 ""-> opacity, 30
drift. minutes.
g. Response time f.b sucoads.
3.3.1 Calibration. The smoke meter is
calibrated after allowing a nilnimum of 30
rri.iut'>s warmup by altc-rnafe'.y producing
simulated opacity of 0 percent and 100 per-
cent. When stable response; at 0 nercen: or
100 percent, is noted, the smoke meter is ad-
Justed to produce an output of 0 percent or
100 percent, as appropriate. This calibration
shall be repeated until stable 0 percent and
100 percent readings are produced without
adjustment. Simulated 0 percent and 100
percent opacity vo.lues may be produced by
alternately switching the power to the liaht
source on and off v.'hile the smoke generator
i= not producing smoVe.
3.3.2 Smoke rneter evaluation The smoke
meter design and perfr.rmanre are to be
evx'.xiated as follows:
3.3.2.1 Light source. Verify from manu-
facturer's data and from voltage measure-
ments made at the lamp, as installed, that
the lamp is operated within ±5 percent of
the nominal rated voltage.
3.3.2.2 Spectral response of photocell.
Verify from manufacturer's data that the
photocell has a photoplc response; i.e., the
spectral sensitivity of the cell shall closely
approximate the standard spactral-luminos-
Itv curve for photopic vision which is refer-
enced in (b) of Table 9-1.
363
-------
3.3.2.3 Angle of view. Check construction
geometry to ensure that the total angle of
view of the smoke plume, as seen by the
photocell, does not exceed 15' The total
angle of view may be calculated from: 0=2
tan-' d/2L, where fl;=total angle of view;
d = the sum of the photocell diameter+the
diameter of the limiting aperture: and
L=the distance from the photocell to the
limiting aperture. The limiting aperture is
the point in the path between the photocell
and the smoke plume wliere the angle of
vlevr ;s most restricted. In smoke generator
Bmoke meters this Is normally an orifice
plate.
3.3.2.4 Angle of projection. Check con-
struction geometry to ensure that the total
angle of projection of the lamp on the
smoke plume does not exceed 15°. The total
angle of projection may be calculated from:
0=2 tan-' d/2L, where «= total angle of pro-
jection; d= the sum of the length of the
lamp filament + the diameter of the limiting
aperture; and L=: the distance from the lamp
to the limiting aperture.
3.3.2.5 Calibration error. Using neutral-
density filters of known opacity, check the
error between the actxial response and the
theoretical linear response of the smoke
meter. This check is accomplished by first
calibrating the emoke meter according to
3.3.1 and then Inserting a series of three
neutral-density filters of nominal opacity of
20, 50, and 75 percent In the smoke meter
pathlength. Filters callbarted within ±2 per-
cent shall be used. Care should be taken
when inserting the niters to prevent stray
light from affecting the meter. Make a total
of five nonconsecutlve readings for each
filter. The """clnnirn error on any one read-
Ing shall be 3 percent opacity.
3.3.2-6 Zero and span drift. Determine
the zero and span drift by calibrating and
operating the smoke generator In a normal
manner over a 1-hour period. The drift la
measured by checking the zero and spaa at
the end of this period.
3.3.2.7 Response time. Determine the re-
sponse time by producng the series of five
simulated 0 percent and 100 percent opacity
values and observing the time required to
reach stable response. Opacity values of 0
percent and 100 percent may be simulated
by alternately switching the power to the
light source off and on while the smoke
generator Is not operating.
4. References.
4.1 Air Pollution Control District Rules
and Regulations, Los Angeles County Air
Pollution Control District. Regulation IV.
Prohibitions, Rule 60.
42 Welsburd, Melvln I., Field Operations
and Enforcement Manual for Air. UJS. Envi-
ronmental Protection Agency, Research Tri-
angle Park. N.C.. APTD-1100. August 1972.
pp. 4.1-4.36.
4.3 Condon, E. U., and Odlshaw, H., Hand-
book of Physics. McGraw-Hill Co.. N.Y.. N.Y..
1058. Table 3.1. p. 6-52.
FIGURE 9-1
RECORD OF VISUAL DETEUl I NATION OF OPACITY
PAGE of
CO",?A.'IY
LOCATION
TEST !!UMDER_
D-'.TE
TYPE FACILITY_
CC'JTROL DEVICE
ur.; or on<:p,y.'Uio:!_
OBSERVER CEP.TIFICATiOil OATE_
OBSERVER AFFILIATION
POINT OF EMISSIONS
HEIGHT OF DISCHARGE POINT
CLOCK TIME
OCSERVER LOCATION
Distance to Discharge
Direction from Discharge
Height of Observation Point
BACKGROUND DESCRIPTION
HEATHER CONDITIONS
Hind Direction
Wind Speed
Arbient Terperature
SKY CONDITIONS (clear.
overcast, X clouds, etc.)
PLl'IE DESCRIPTION
Color
Distance Visible
CIliLi! IHIOPr.'iTlOM
Initial
Final
SUGARY OF AVERAGE OPACITY
Set
Number
Tlnp
Start—End
Opacity
Sjm
•vfragc
Readings ranged from
to
opacity
The source was/was not 1n conpliance with
the tire evaluation was made.
at
364
-------
FIGURE 9-2 OBSERVATION RECORD
PAGE
OF
COMPANY
LOCATION
TEST NUMBER
DATE
OBSERVER
TYPE FACILITY
POINT OF EMISSIONS
Hr.
Min.
0
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
26
27
28
29
-fr-
Se
rrr
conrls
30
*-b
STEAM PLUME
(check if applicable)
Attached
Detached
COMMENTS
365
-------
METHOD 10
DETERMINATION OF CARBON MONOXIDE EMISSIONS FROM STATIONARY SOURCES
1. Principle and Applicability
1.1 Principle. An integrated or continuous
gas sample Is extracted from a sampling point
and analyzed for carbon monoxide (CO) con-
tent using a Luft-type nondisperslve infra-
red analyser (NDIR) cr equivalent.
1 2 Applicability. This method is appli-
cable for the determination of carbon mon-
oxide emissions from stationary sources only
when specified by the test procedures for
determining compliance with new source
performance standards. The test procedure
will indicate whether a Continuous or an
Integrated sample is to be used.
2. Range and sensitivity.
2.1 Range. 0 to 1,000 ppm.
2.2 Sensitivity. Minimum detectable con-
centration is 20 ppm for a 0 to 1.000 ppm
span.
3. Interferences. Any substance having a
strong absorption of Infrared energy will
Interfere to some extent. For example, dis-
crimination ratios for water (HnO) and car-
bon dioxide (CO,) are 3-5 percent H,O per
7 ppm CO and 10 percent CO2 per 10 ppm
CO, respectively, for devices measuring in the
1,500 to 3,000 ppm range. For devices meas-
uring in the 0 to 100 ppm range. Interference
ratios can be as high as 3.5 percent H2O per
25 ppm CO and 10 percent CO, per 50 ppm
CO. The use of silica gel and ascarite traps
will alleviate the major Interference prob-
lems. The measured gas volume must be
corrected if these traps are used.
4 Precision and accuracy.
4 1 Precision. The precision of most NDIR
analyzers is approximately ±2 percent of
span.
4.2 Accuracy. The accuracy of most NDIR
analyzers is approximately ±5 percent of
span after calibration.
5. Apparatus.
5.1 Continuous sample (Figure 10-1).
5.1.1 Probe. Stainless steel or sheathed
Pyrex' glass, equipped with a Biter to remove
particulate matter.
5.1.2 Air-cooled condenser or equivalent.
To remove any excess moisture.
52 Integrated sample (Figure 10-2).
5.2.1 Probe. Stainless steel or sheathed
Pyrex glass, equipped with a filter to remove
particulate matter.
5.2.2 Air-cooled condenser or equivalent.
To remove any excess moisture.
5.2.3 Valve. Needle valve, or equivalent, to
to adjust flow rate.
5.2.4 Pump. Leak-free diaphragm type, or
equivalent, to transport gas.
5.2.5 Rate meter Rotameter, or equivalent,
to measure a flow range from 0 to 1.0 liter
per mln. (0.035 cfm)
5.2.6 Flexible bag. Tedlar, or equivalent,
with a capacity of 60 to 90 liters (2 to 3 ft»).
Leak-test the bag In the laboratory before
using by evacuating bag with a pump fol-
lowed by a dry gas meter. When evacuation
Is complete, there should be no flow through
the meter.
5.2.7 Pilot tube. Type S, or equivalent, at-
tached to the probe so that the sampling
rate can be regulated proportional to the
stack gas velocity when velocity Is varying
with the time or a sample traverse is cou-
ducted.
5.3 Analysis (Figure 10-3).
5.3.1 Carbon monoxide analyzer. Nondisper-
sive infrared spectrometer, or equivalent.
This instrument should be demonstrated,
preferably by the manufacturer, to meet or
exceed manufacturer's specifications and
those described in this method.
5.3.2 Drying tube. To contain approxi-
mately 200 g of silica gel.
5.3.3 Calibration gas. Refer to paragraph
5.1-
5.3.4 Filter. As recommended by NDIR
manufacturer.
AIR-COOLED CONDENSER
FILTER (GLASS WOOL!
5.3.5 COj removal tube. To contain approxi-
mately 500 g of ascarite.
5.3.6 Ice water bath. For ascarite and silica
gel tubes.
5.3.7 Valve. Needle valve, or equivalent, to
adjust flow rate
5.3.8 Rate meter. Rotameter or equivalent
to measure gas flow rate of 0 to 1.0 liter per
min. (0.035 cfm) through NDIR.
5.3.9 Recorder (optional). To provide per-
manent record of NDIR readings.
6. Reagents.
1 Mention of trade names or specific prod-
ucts does not constitute endorsement by the
Environmental Protection Agency.
366
-------
COMPANY
LOCATION
TEST NUMBER
DATE
FIGURE 9-2 OBSERVATION RECORD
(Continued)
OBSERVER
PAGE
OF
TYPE FACILITY __
POINT OF EMISSIONS
Hr. 1 Mir..
30
31
32
33
L 34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
Seconds
0
15
30
4b
STEAM PLUME
(check if applicable)
Attached
Detached
COMMENTS
367
-------
6.1 Calibration gases. Known concentration
of CO in nitrogen (N2) for instrument span,
prepurifled grade of N. for zero, and two addi-
tional concentrations corresponding approxi-
mately to CO percent and 30 percent span. The
span concen' ition shall not exceed 1.5 times
the applica .<; source performance standard.
The call oration gases shall be certified by
the manufacturer to be within ±2 percent
of the specified concentration.
F.gure 10-3 A . c* v".- p~nl.
6.2 Silica gel. Indicating type, 6 to 16 mesh.
dried at 175° C (347° P) for 2 hours.
G.3 Ascarite. Commercially available.
7. Procedure.
7.1 Sampling.
7.1.1 Continuous sampling. Set up the
equipment as shown in Figure 10-1 making
sure all connections are leak free. Place the
probe in the stack at a sampling point and
purge the sampling line. Connect the ana-
lyzer and begin drawing sample into the
analyzer. Allow 5 minutes for the system
to stabilize, then record the analyzer read-
Ing as required by the test procedure. (See
11 7.2 and 8). CO2 content of the gas may be
determined by using the Method 3 inte-
grated sample procedure (36 FB 24886), or
by weighing the ascarlte CO, removal tube
and computing CO concentration from the
gas volume sampled and the weight gain
of the tube.
7.1.2 Integrated sampling. Evacuate the
flexible bag. Set up the oonipmeiit a'5 shown
in Figure 10-2 wit i the bag disconnected.
Place the probe i:i the stack and purge the
sampling line. Connect the bag, making sure
that all connections are leuk free. Sample at
a rate proportional to the stack velocity.
CO., c outeiit of the yr.s may be determined
by uKii:g the Method 3 integrated sample
procedures (36 FR 24886), or by weighing
the as'_urite CO., removal tube and comput-
ing CO., concentration from the gas volume
sampled and the weight gain of the tube.
7.2 CO Analysis. Assemble the apparatus as
shown In Figure 10-3, calibrate the instru-
ment, and perform other required operations
as described in paragraph 8. Purge analyzer
with N2 prior to introduction of each sample.
Direct the sample stream through the instru-
ment for the test period, recording the read-
ings. Check the zero and span again after the
test to assure that any drift or malfunction
is detected. Record the sample data on Table
10-1.
8. Calibration. Assemble the apparatus ac-
cording to Figure 10-3. Generally an instru-
ment requires a warm-up period before sta-
bility is obtained. Follow the manufacturer's
instructions for specific procedure. Allow a
minimum time of one hour for warm-up.
During this time check the sample condi-
tioning apparatus, i.e., filter, condenser, dry-
Ing tube, and CO2 removal tube, to ensure
that each component is in good operating
condition. Zero and calibrate the Instrument
according to the manufacturer's procedures
using, respectively, nitrogen and the calibra-
tion gases.
9. Calculation—Concentration of carbon monoxide. Calculate the concentration of carbon
monoxide in the stack using equation 10-1.
equation 10-1
where:
Ccost.ck = concentration of CO in stack, ppm by volume 'dry basis).
(VoNUIP — concentration of CO measured by NDIR annlyzer, ppm by volume (dry
basis).
/• c 0 =r-\i nirr>e fraction of CO2 in sample, i.e., pm^nt COj from Great analysis
di\idcd by 100.
10.3 MSA LIRA Infrared Gas and Liquid
Analyzer Instruction Book, Mine Safety
Appliances Co., Technical Products Di-
vision, Pittsburgh, Pa.
10.4 Models 215A, 315A, and 415A Infrared
Analyzers, Beckman Instruments, Inc.,
Beckman Instructions 1635-B, Fuller-
ton, Calif., October 1967.
10.5 Continuous CO Monitoring System,
Model A5611, Intertech Corp., Princeton,
N.J
10.6 UNOR Infrared Gas Analyzers, Bendlx
Corp., Ronceverte, West Virginia.
10 Biblf. iphy
10.1 Mc^iroy, Frank. The Intertech NDIR-CO
Analyzer, Presented at llth Methods
Coi-.eronci -i Air Pollution, University
of California, Berkeley, Calif., April 1,
;r:7i>
i0.2 Jacobs, M. B., et al.. Continuous Deter-
mination of Carbon Monoxide and Hy-
drocarbons in Air by a Modified Infra-
red Analyzer, J. Air Pollution Control
Association, 9(2) :110-114, August 1959.
TABLE 10-1.—Field data
Location Comments:
Test
Date
Operator
Clock time
Rotameter setting, liters per minute
(cubic feet per minute)
368
-------
ADDENDA
A. Performance Specifications for NDIR Carbon Monoxide Analyzers.
Range (minimum) 0-1000ppm.
Output (minimum) 0-10mV.
Minimum detectable sensitivity 20 ppm.
Rise time. 90 percent (maximum) 30 seconds.
:Fall time. 90 percent (maximum) 30 seconds.
Zero drift (maximum) 10% in 8 hours.
Span drift (maximum) 10% lii 8 hours.
Precision (minimum) ± 2% of full scale.
Noise (maximum) ± 1 % of full scale.
Linearity (maximum deviation) 2 % of full scale.
Interference rejection ratio CO2—1000 to 1. H2O—500 to 1.
B. Definitions of Performance Specifica-
tions.
Range—The minimum and maximum
measurement limits.
Output—Electrical signal which is propor-
tional to the measurement; Intended for con-
nection to readout or data processing devices.
Usually expressed as millivolts or mllliamps
full scale at a given Impedance.
Full scale—The maximum measuring limit
for a given range.
Minimum detectable sensitivity—The
smallest amount of Input concentration that
can be detected as the concentration ap-
proaches zero.
Accuracy—The degree of agreement be-
tween a measured value and the true value;
usually expressed as ± percent of full scale.
Time to 90 percent response—The time In-
terval from a step change In the Input con-
centration at the Instrument Inlet to a read-
Ing of 90 percent of the ultimate recorded
concentration.
Rise Time (90 percent)—The Interval be-
tween Initial response time and time to 90
percent response after a step Increase In the
Inlet concentration.
Fall Time (90 percent)—The Interval be-
tween Initial response time and time to 90
percent response after a step decrease In the
Inlet concentration.
Zero Drift—The change In Instrument out-
put over a stated time period, usually 24
hours, of unadjusted continuous operation
when the Input concentration Is zero; usually
expressed as percent full scale.
Span Drift—The change In Instrument out-
put over a stated time period, usually 24
hours, of unadjusted continuous operation
when the Input concentration Is a stated
i upscale value; usually expressed as percent
I full scale.
Precision—The degree of agreement be-
tween repeated measurements of the same
concentration, expressed as the average de-
viation of the single results from the mean.
Noise—Spontaneous deviations from a
mean output not caused by Input concen-
tration changes.
Linearity—The maximum deviation be-
tween an actual Instrument reading and the
reading predicted by a straight line drawn
between upper and lower calibration points.
369
-------
METHOD 11
DETERMINATION OF HYDROGEN SULFIDE EMISSIONS FROM STATIONARY SOURCES
1. Principle and applicability.
1.1 Principle. Hydrogen sulflde (H,S) la
collected from the source In a series of midget
Implngers and reacted with alkaline cad-
mium hydroxide [Cd(OH)2] to form cad-
mium sulflde (CdS). The precipitated CdS
Is then dissolved In hydrochloric acid and
absorbed in a known volume of iodine solu-
tion. The iodine consumed is a measure of
the HjS content of the gaa. An Implnger con.
talnlng hydrogen peroxide is Included to re-
move SOj as an interfering species.
1.2 Applicability. This method is applica-
ble for the determination of hydrogen sul-
flde emissions from stationary sources only
when specified by the test procedures for
determining compliance with the new source
performance standards.
2. Apparatus.
2.1 Sampling train.
2.1.1 Sampling line—6- to 7-mm (14-inch)
Teflon > tubing to connect sampling train to
sampling valve, with provisions for heating
to prevent condensation. A pressure reduc-
ing valve prior to the Teflon sampling line
may be required depending on sampling
stream pressure.
2.1.2 Impingers—Five midget Implngers,
each with 30-ml capacity, or equivalent.
2.1.3 Ice bath container—1o maintain ab-
sorbing solution at a constant temperature.
2.1.4 Silica gel drying tube—To protect
pump and dry gas meter.
2.1.5 Needle valve, or equivalent—Stainless
steel or other corrosion resistant material, to
adjust gas flow rate.
2.1.6 Pump—Leak free, diaphragm type, or
equivalent, to transport gas. (Not required
If sampling stream under positive pressure.)
2.1.7 Dry gas meter—Sufficiently accurate
to measure sample volume to within 1 per-
cent.
2.1.8 Rate meter—Rotameter, or equivalent,
to measure a flow rate of 0 to 3 liters per
minute (0.1 ft'/mln).
2.1.9 Graduated cylinder—25 ml.
2.1.10 Barometer—To measure atmospheric
pressure within ±2.5 mm (0.1 In.) Hg.
2.2 Sample Recovery.
2.2.1 Sample container—500-ml glass-stop-
pered iodine flask.
2.2.2 Pipette—50-mI volumetric type.
2.2.3 Beakers—250 ml.
2.2.4 Wash, bottle—Glass.
2.3 Analysis.
2.3.1 Flask—500-ml glass-stoppered iodine
flask.
2.3.2 Burette—One 50 ml.
~ 2.3 2 Flask—125-ml conical.
3. Reagents.
3.1 Sampling.
3 1.1 Absorbing solution—Cadmium hy-
droxide (Cd(OH)2)—Mix 4.3 g cadmium sul-
fate hydrate (3 CdSOt.8HaC>) and 0.3 g of
sodium hydroxide (NaOH) In 1 liter of dis-
tilled water (H,O). Mix well.
1 Mention of trade names or specific prod-
ucts does not constitute endorsement by the
Environmental Protection Agency.
Note: The cadmium hydroxide formed in
this mixture will precipitate as a white sus-
pension. Therefore, this solution must be
thoroughly mixed before using to ensure an
even distribution of the cadmium hydroxide.
3.1.2 Hydrogen peroxide, 3 percent—Dilute
30 percent hydrogen peroxide to 3 percent
as needed. Prepare fresh dally;
3.2 Sample recovery.
3.2.1 Hydrochloric acid solution (HCl), 10
percent by iciight—mix 230 ml of concen-
trated HCl (specific gravity 1.19) and 770 ml
of distilled H..O.
3.2.2 Iodine solution, 0.1 N—Dissolve 24 g
potassium iodide (KI) in 30 mi of distilled
H..O in a 1-liter graduated cylinder. Weigh
12.7 g of resublimed iodine (I,) into a weigh-
ing bottle and add to the potassium iodide
solution. Shake the mixture until the Iodine
is completely dissolved. Slowly dilute the so-
lution to 1 liter with distilled H..O, with
swirling, p'llter the solution, if cloudy, and
store in a brown glabs-stoppered bottle.
3.2.3 Standard iodine solution, 0.01 N—Di-
lute 100 ml of the 0.1 N Iodine solution in a
volumetric flask to 1 liter with distilled
water.
Standardize dally as follows: Pipette 25 ml
of the 0.01 N iodine solution into a 125-ml
conical flask. Titrate with standard 0.01 N
thlosulfate solution (see paragraph 3.3.2) un-
til the solution is a light yellow. Add a few
drops of the starch solution and continue
titrating until the blue color Just disap-
pears. Prom the results of this titration, cal-
culate the exact normality of the iodine
solut'.on (see paragraph 5.1).
3.2.-1 Distilled, deionized water.
3.3 Analysis.
3.3.1 Sodium thiosulfate solution, standard
0.1 N—For each liter of solution, dissolve
24.8 g of sodium thlosulfate (NA.S.,0., • 5H..O)
In distilled water arid add 0.01 g o'f anhydrous
sodium carbonate (Na^COJ and 0.4 ml of
chloroform (CHC13) to" stabilize. Mix thor-
oughly by shaking or by aerating with r.itro-
gtn for approximately 15 minutes, and store
in a glass-stoppered glass bottle.
Standardize frequently as follows: Weigh
into a 500-ml volumetric flask about 2 g of
potassium dlchromate (K:Cr,,O7) weighed
to the nearest milligram and dilute to the
500-ml mark with distilled H:O. Use dl-
chromate which has been crystallized from
distilled water and oven-dried at 182 °C to
199°C (360°F to 390°F). Dissolve approxi-
mately 3 g of potassium iodide (KI) In 50 ml
of distilled water in a glass-stoppered, 500-ml
conical flask, then add 5 ml of 20-percent
hydrochloric acid solution. Pipette 50 ml ot
the dichromate solution Into this mixture.
Gently swirl the solution once and allow It
to stand in the dark for 5 minutes. Dilute
the solution with 100 to 200 ml of distilled
water, washing down the sides of the flask
with part of the water. Swirl the solution
slowly'and titrate with the thlosulfate solu-
tion until the solution Is light yellow. Add
4 ml of starch solution and continue with a
370
-------
Slow titratlon with the thlosulfate untU the
bright blue color has disappeared and only
the pale green color of the chromic Ion re-
mains. Prom this iltratlon, calculate the ex-
act normality of the sodium thiosulfate solu-
tion (see paragraph 5.2).
3.3.2 Sodium thiosulfate solution, standard
0.01 N—Pipette 100 ml of the standard 0.1 N
thiosulfate solution into a volumetric flask
and dilute to oiie liter with distilled water.
3.3.3 Starch indicator solution—Suspend
10 g of soluble starch in 100 ml of distilled
water and add 15 g of potassium hydroxide
pelltts. Stir until dissolved, dilute with 900
ml of distilled water, aud let stand 1 hour.
Neutralize the alkali with concentrated hy-
drochloric acid, using au indicator paper
similar to Alkacid test ribbon, then add 2 ml
of glacial acetic acid as a preservative.
Test for decomposition by titrating 4 ml of
etarch solution In 200 ml of distilled water
with 0.01 N iodine solution. If more than 4
drops of the 0.01 N iodine solution are re-
quired to obtain the blue color, make up a
fresh starch solution.
4. Procedure.
4.1 Sampling.
4.1.1 Assemble the sampling train as shown
In Figure 11-1, connecting the five midget
impingers in series. Place 15 ml of 3 percent
hydrogen peroxide In the first impinger. Place
15 ml of the absorbing solution in each of
the next three Impingers, leaving the fifth
dry. Place crushed ice around the Impingers.
Add more Ice during the run to keep the
temperature of the gases leaving the last
Impinger at about 20°C (70°P), or less.
4.1.2 Purge the connecting line between
the sampling valve and the first Impinger.
Connect the sample line to the train. Record
the initial reading on the dry gas meter aa
shown in Table 11-1.
TABLE 11-1.—Field data
Location Comments:
Test
Date
Operator
Barometric pressure
Clock
time
Gaa volume
throueh
meter (V"«),
liters (cubic
feet)
Rotanieter
setting, Lpm
(cubic feet
per minute)
Meter
temperature.
0 C (° F)
4.1.3 Open the flow control valve and ad-
lust tb«j sampling rate to 1.13 liters per
minute (0.04 cfm). Read the meter temper-
ature and record on Table 11-1.
4.1.4 Continue sampling a minimum of 10
minutes. If the yellow color of cadmium sul-
nde is visible in the third impinger, analysis
should confirm that the applicable standard
.has been exceeded. At the end of the sample
time, close the flow control valve and read
ithe final meter volume and temperature.
. 4.1.5 Disconnect the impinger train from
the sampling line. Furge the train with clean
lambient air for 15 minutes to ensure that all
IL.S is removed from the hydrogen peroxide.
Cap the open ends and move to the sample
'clean-up area.
| 4.2 Sample recovery.
' 4.2.1 Pipette 50 ml of O.OL N iodine solution
, Into a 250-ml beaker. Add 50 ml of 10 percent
HC1 to the solution. Mix well.
4.2.2 Discard the contents of the hydrogen
peroxide impinger. Carefully transfer the con-
tents of the remaining four impingers to a
500-ml iodine flask.
I 4.2.3 Rinse the four absorbing Impingers
'and connecting glassware with three portions
:of the acidified iodine solution. Use the en-
tire 100 ml of acidified iodine for this pur-
pose. Immediately after pouring the acidified
iodine into an impinger, stopper it r, nd shake
for a. few moments before transferring the
rinse to the iodine flask. Do not transfer any
rinse portion from one impinger to another;
transfer It directly to the iodine flask. Once
acidified iodine solution has been poured into
any glassware containing cadmium sulflde
sample, the container must be tightly stop-
pered at all times except when adding more
solution, and this must be done as quickly
and carefully as possible. After adding any
acidified iodine solution to the iodine flask,
allow a few minutes for absorption of the H2S
into the iodine before adding any further
rinses.
4.2.4 Follow this rinse with two more rinses
using distilled water. Add the distilled water
rinses to the iodine flask. Stopper the flask
and shake well. Allow about 30 minutes for
absorption of the H,S into the Iodine, then
complete the analysis titratlon.
Caution: Keep the iodine flask stoppered
except when adding sample or tltrant.
4.2.5 Prepare a blank in an iodir.e fiask
using 45 ml of the absorbing solution, 50 ml
of 0.01 N Iodine solution, and 50 ml of 10
percent HC1. Stopper the flask, shake well
and analyze with the samples.
4.3 Analysis.
Note: This analysis titratlon should be
conducted at the sampling location in order
to prevent loss of Iodine from the sample.
Titration should never be made in direct
sunlight.
4.3.1 Titrate the solution In the flask with
0.01 N sodium thiosulfate solution until the
solution is light yellow. Add 4 ml of the
starch Indicator solution and continue
titrating until the blue color Just disappears.
4.3.2 Titrate the blanks In the same man-
ner aa the samples.
5. Calculations.
5.1 Normality of the standard iodine solution.
NTVr
equation lt-1
371
-------
where :
.V7=normality of iodine, g-eq/liler.
V,= volume of iodine used, nil.
AV= normality of sodium thiosulfate, g-eq/liter.
VT= volume of sodium thiosulfate used, ml.
5.2 Normality of the standard Ihiosuljale solution.
W
Vr equation 11 -2
where:
W— weight of A'jCVjO? used, g.
Vr= volume of NaiSiO, used, ml.
AV= normality of standard thiosulfate solution, g-eq/litcr.
2.04 = con version factor
= (0 eq /j/mole gjCr20;) (1,000 ml/1) _
"" (294.2 g #2CV207/mole) (10 aliquot factor).
5.3 Dry gas volume. Correct the sample volume measured by the dry gas nu-tc: to
standard conditions [21°C(70°F)] and 760 mm (29.92 inches) llg] by using equation 11-3.
V = V (T"a} (Pl'"
''-.id r"\Tm J \P.<
equation 11-3
where:
Fm>lil = volume at standard conditions of gas sample through the dry gas meter,
standard liters (scf).
Vm = volume of gas sample through the dry gas meter (meter conditions), liters
(cu. ft.).
T.td = absolute temperature at standard conditions, 294°K (530°R).
7V. = average dry gas meter temperature, °K (°R).
P,..r= barometric pressure at the orifice meter, mm Hg (in. Hg).
P.td = absolute pressure at standard conditions, 760 mm Hg (29.92 in. Hg).
5.4 Concentration of H^S. — Calculate the concentration of H2S in the gas stream at
standard conditions using equation 11-4:
wheie (metric units):
Cn2e — concentration of H2S at standard conditions, mg/dscm
.K = couversion factor=17.0X103
= (34.07 g/mole H2S)( 1,000 l/ni3)(l,000 mg/g)
~ (1,000 ml/l)(2H2S eq/inole)
V/ = volume of standaid iodine solution, ml.
JVj = nornuility of standard iodine solution, g-eq/liter.
VT= volume of standard sodium thiosulfate solution, ml.
A'r= normality of standard sodium thiosulfate solution, g-ecj/llter.
Vm>iU = dry gas volume at standard conditions, liters.
where (English units):
17.0(15.43 gr/g)
(1,000 l/m«)
m=rSC'-
6. References.
1^ °rf Hydr°geP Sulfide, Ammoniacal Cadmium Chloride Method,
/2-54- T°: Manual on Disposal of Refinery Wastes, Vol. V: Sampling
, C 19M Peculate Matter, American Petroleum InstHute!
in 6N»tT,r^ar Ve M^h°di fnr D?,terniination of Hydrogen Sulfide and Mercaptan Sulfur
cation No 2265l65a5 Process^s Association, Tulsa, Oklahoma, NGPA Publi-
catin No 2265l65i9G5
372
-------
METHOD 13A
DETERMINATION OF TOTAL FLUORIDE EMISSIONS FROM STATIONARY SOURCES
SPADNS ZIRCONIUM LAKE METHOD
1. Principle and Applicability.
1.1 Principle. Gaseous and partlculate
fluorides are withdrawn isokinetically from.
the soxu-ce using a sampling train. The fluo-
rides are collected in the Impinger water and
on the filter of the sampling train. The
weight of total fluorides in the train Is de-
termined by the SPADNS Zirconium Lake
colorimetric method.
1.2 Applicability. This method Is applica-
ble for the determination of fluoride emis-
sions from stationary sources only when
specified by the test procedures for deter-
mining compliance with new source per-
formance standards. Pluorocarbons, such as
Freons, are not quantitatively collected or
measured by this procedure.
2. Range and Sensitivity.
The SPADNS Zirconium Lake analytical
method covers the range from 0—1.4 /tg/ml
fluoride. Sensitivity has not been determined.
3. Interferences.
During the laboratory analysis, aluminum
In excess of 300 mg/liter and silicon dioxide
in excess of 300 /ig/liter will prevent com-
plete recovery of fluoride. Chloride will distill
over and interfere with the SPADNS Zirconi-
um Lake color reaction. If chloride Ion Is
present, use of Specific Ion Electrode (Method
13B) Is recommended; otherwise a chloride
determination Is required and 5 nag of silver
sulfate (see section 7.3.6) must be added for
each nig of chloride to prevent chloride In-
terference. If sulfurlc acid is carried over In
the distillation, It will cause a positive Inter-
ference. To avoid sulfurlc acid carryover. It
Is important to stop distillation at 175°C.
4. Precision, Accuracy and Stability.
4.1 Analysis. A relative standard devia-
tion of 3 percent was obtained from twenty
replicate Intralaboratory determinations on
stack emission samples with a concentration
range of 39 to 360 mg/1. A phosphate rock
standard which was analyzed by this pro-
cedure contained a certified value of 3.84
percent. The average of five determinations
was 3.88 percent fluoride.
4.2 Stability. The color obtained when
the sample and colorimetric reagent are
mixed Is stable for approximately two hours.
After formation of the color, the absorbances
of the sample and standard solutions should
be measured at the same temperature. A 3°C
temperature difference between sample and
standard solutions will produce an error of
approximately 0.005 mg F/llter.
6. Apparatus.
5.1 Sample train. See Figure 13A-1; It Is
similar to the Method 5 train except for the
luterchangeability of the position of the ni-
ter. Commercial models of this train are
available. However, if one desires to build his
own. complete construction details are de-
scribed in APTD-0581; for changes from the
APTD-05B1 document and for allowable
modifications to Figure 13A-1, see the fol-
lowing subsections.
The operating and maintenance procedures
for the sampling train are described In
APTD-0576. Since correct usage is important
in obtaining valid results, all users should
read the APTD-0576 document and adopt
the operating and maintenance procedures
outlined in it, unless otherwise specified
herein.
5.1.1 Probe nozzle—Stainless steel (316)
with sharp, tapered leading edge. The angle
of taper shall be S30° and the taper shall
be on the outside to preserve a constant
internal diameter. The probe nozzle shall be
of the button-hook or elbow design, unless
otherwise specified by the Administrator. The
wall thickness of the nozzle shall be less than
or equal to that of 20 gauge tubing, i.e.,
0.165 cm (0.065 in.) and the distance from
the tip of the nozzle to the first bend or
point of disturbance shall be at least two
times the outside nozzle diameter. The nozzle
' shall be constructed from seamless stainless
steel tubing. Other configurations and con-
struction material may be used with approval
from the Administrator.
A range of sizes suitable for Isoklnetlc
sampling should be available, e.g., 0.32 cm
C/8 in.) up to 1.27 cm (>/4 In.) (or larger if
higher volume sampling trains are used) in-
side diameter (ID) nozzles in increments of
0.16 cm (i/IO in.). Each nozzle shall be cali-
brated according to the procedures outlined
in the calibration section.
5.1.2 Probe liner—Boroslllcate glass or
stainless steel (316). When the filter is lo-
cated immediately after the probe, a probe
neating system may be used to prevent filter
plugging resulting from moisture condensa-
tion. The temperature In the probe shall not
exceed 120 ± 14°C (248 ± 25°F).
5.1.3 Pltot tube—Type S, or other device
approved by the Administrator, attached to
probe to allow constant monitoring of the
stack gas velocity. The face openings of the
pitot tube and the probe nozzle shall be
adjacent and parallel to each other, not
necessarily on the same plane, during sam-
pling. The free space between the nozzle and
pitot tube shall be at least 1.9 cm (0.75 in.).
The free space shall be set based on a 1.3 cm
(0.5 in.) ID nozzle, which is the largest size
nozzle used.
The pitot tube must also meet the criteria
specified in Method 2 and be calibrated ac-
cording to the procedure in the calibration
section of that method.
5.1.4 Differential pressure gauge—In-
clined manometer capable of measuring ve-
locity head to within 10% of the minimum
measured value. Below a differential pressure
of 1.3 mm (0.05 in.) water gauge, micro-
manometers with sensitivities of 0.013 mm
(0.0005 In.) should be used. However, micro-
manometers are not easily adaptable to field
conditions and are not easy to use with pul-
sating flow. Thus, other methods or devices
acceptable to the Administrator may be
used when conditions warrant.
5.1.5 Filter holder—Borosilicate glass with
a glass frit filter support and a silicone rub-
ber gasket. Other materials of construction
may be used with approval from the Ad-
ministrator, e.g.. It probe liner is stainless
steel, then filter holder may be stainless steel.
The holder design shall provide a positive
seal against leakage from the outside or
around the filter.
373
-------
liM'r'T TEMPERATURE
(Y ' ^S SENSOR .
J/PROBE
1.9cm (0.75 in.) ^ ^^=
PITOTTUBE
./
V«| STACK WALL
PROBE L/
g4.-.--------==-^
V
OPTIONAL
FILTER HOLDER
LOCATION
THERMOMETER CHEC((
FILTER HOLDER ^7i VALVE
/
REVERSETYPE
PITOTTUBE
r x
PITOT MANOMETER
X
ORIFICE MANOMETER
Fiyurc 13A-1. Fluoride sampling (rain.
CONNECTING TUBE
12-mm ID
f24 40 V
THERMOMETER TIFMUST EXTEND BELOW
THE LIQUID LEVEL
WITH S10/30
124/40
$24/40
CONDENSER
HEATING
MANTLE
250ml
VOLUMETRIC
FLASK
Figure 13A-2. Fluoride Distillation Apparatus
374
-------
5.1.6 Filter heating system—When mois-
ture condensation Is a problem, any heating
system capable of maintaining a temperature
around the filter holder during sampling of
no greater than 120±14°C (248±25"F).
A temperature gauge capable of measuring
temperature to within 3°C (5.4°F) "shall be
installed so that when the niter heater is
used, the temperature around the filter
holder can be regulated and monitored dur-
ing sampling. Heating systems other than
the one shown in APTD-0581 may be used.
5.1.7 Impingers—Four impingers con-
nected as shown in Figure 13A-1 with ground
glass (or equivalent), vacuum tight fittings.
The first, third, and fourth impingers are
of the Greenburg-Smith design, modified by
replacing the tip with a I1/! cm (% In.)
inside diameter glass tube extending to 114
cm (V2 in.) from the bottom of the flask.
The second impinger Is of the Greenburg-
Smith design with the standard tip.
5.1.8 Metering system—Vacuum gauge,
leak-free pump, thermometers capable of
measuring temperature to within 3°C
(~5°F), dry gas meter with 2% accuracy at
the required sampling rate, and related
equipment, or equivalent, as required to
maintain an Isokinetlc sampling rate and
to determine sample volume. When the
metering system Is used in conjunction with
a pltot tube, the system shall enable checks
of Isokinetlc rates.
5.19 Barometer—Mercury, aneroid, or
other barometers capable of measuring at-
mospheric pressure to within 2.5 mm Hg
(0.1 In. Hg). In many cases, the barometric
reading may be obtained from a nearby
weather bureau station, in which case the
station value shall be requested and an ad-
justment for elevation differences shall be
applied at a rate of minus 2.5 mm Hg (0.1
in. Hg) per 30 m"(100 ft) elevation Increase.
5.2 Sample recovery.
5.2.1 Probe liner and probe nozzle
brushes—Nylon bristles with stainless steel
wire handles. The probe brush shall have
extensions, at least as long as the probe, of
stainless steel, teflon, or similarly Inert mate-
rial. Both brushes shall be properly sized and
shaped to brush out the probe liner and
nozzle.
5.2.2 Glass wash bottles—Two.
5.2.3 Sample storage containers—Wide
mouth, high density polyethylene bottles,
1 liter.
5.2.4 Plastic storage containers—Air tight
containers of sufficient volume to store silica
gel.
5.2.5 Graduated cylinder—250 ml.
5.2.6 Funnel and rubber policeman—to
aid in transfer of silica gel to container; not
necessary if silica gel is weighed In the field.
5.3 Analysis.
5.3.1 Distillation apparatus—Glass distil-
lation apparatus assembled as shown in Fig-
ure 13A-2.
5.3.2 Hot plate—Capable of heating to
500° C.
5.3.3 Electric muffle furnace—Capable of
heating'to 600° C.
5.3.4 Crucibles—Nickel, 75 to 100 ml ca-
pacity.
5.3.5 Beaker, 1500 ml.
5.3.6 Volumetric flask—50 ml.
5.3.7 Erlenmeyer flask or plastic bottle—
500 ml.
5.3.8 Constant temperature bath—Capa-
ble of maintaining a constant temperature of
±1.0" C In the range of room temperature.
5.3.9 Balance—300 g capacity to measure
to ±0.5 g.
5.3.10 Spectrophotometer — Instrument
capable of measuring absorbance at 570 nm
and providing at least a 1 cm light path.
5.3.11 Spectrophotometer cells—1 cm.
6. Reagents
6.1 Sampling.
6.1.1 Filters—Whatman No. 1 filters, or
equivalent, sized to fit filter holder.
6.1.2 Silica gel—Indicating type, 6-16
mesh. If previously used, dry at 175° C
.(350° P) for 2 hours. New silica gel may be
•used as received.
6.1.3 Water—Distilled.
6.1.4 Crushed Ice.
6.1.5 Stopcock grease—Acetone Insoluble.
heat stable slllcone grease. This Is not neces-
sary if screw-on connectors with teflon
sleeves, or similar, are used.
6.2 Sample recovery.
6.2.1 Water—Distilled from same con-
tamer as 6.1.3.
6.3 Analysis.
6.3.1 Calcium oxide (CaO)—Certified
grade containing 0.005 percent fluoride or
less.
6.3.2 Phenolphthaleln Indicator—0.1 per-
cent in 1:1 ethanol-water mixture.
6.3.3 Silver sulfate (Ag,SO,)—ACS re-
agent grade, or equivalent.
6.3.4 Sodium hydroxide (NaOH)—Pellets,
ACS reagent grade, or equivalent.
6.3.5 SuHuric acid (H..SO,)—Concen-
trated, ACS reagent grade, or equivalent.
6.3.6 Filters—Whatman No. 541, or equiv-
alent.
6.3.7 Hydrochloric acid (HC1)—Concen-
trated, ACS reagent grade, or equivalent.
6.3.8 Water—Distilled, from same con-
tainer as 6.1.3.
6.3.9 Sodium fluoride—Standard solution.
Dissolve 0.2210 g of sodium fluoride in 1
liter of distilled water. Dilute 100 ml of this
solution to 1 liter with distilled water. One
milliliter of the solution contains 0.01 mg
of fluoride.
6.3.10 SPADNS solution—[4,5dihydroxy-
3- (p-sulfophenylazo) -2,7-naphthalene dl-
sulfonic acid trisodium salt). Dissolve 0.960
±.010 g of SPADNS reagent in 500 ml dis-
tilled water. This solution is stable for at
least one month, if stored in a well-sealed
bottle protected from sunlight.
6.3.11 Reference solution—Add 10 ml of
SPADNS solution (6.3.10) to 100 ml distilled
water and acidify with a solution prepared by
diluting 7 ml of concentrated HC1 to 10 ml
with distilled water. This solution is used to
set the Spectrophotometer zero point and
should be prepared daily.
6.3.12 SPADNS Mixed Reagent—Dissolve
0.135 ±0.005 g of zirconyl chloride octahy-
drate (ZrOCl2.8H.O), in 25 ml distilled water.
Add 350 ml of concentrated HC1 and dilute to
500 ml with distilled water. Mix equal vol-
umes of this solution and SPADNS solution
to form a single reagent. This reagent is
stable for at least two months.
7. Procedure.
NOTE: The fusion and distillation steps of
this procedure will not be required, If It can
be shown to the satisfaction of the Adminis-
trator that the samples contain only water-
soluble fluorides.
7.1 Sampling. The sampling shall be con-
ducted by competent personnel experienced
with this test procedure.
7.1.1 Pretest preparation. All train com-
ponents shall be maintained and calibrated
according to the procedure described in
APTD-0576, unless otherwise specified herein.
Weigh approximately 200-300 g of silica gel
in air tight containers to the nearest 0.5 g.
Record the total weight, both silica gel and
container, on the container. More silica gal
may be used but care should be taken during
sampling that It Is not entrained and carried
out from the Impinger. As an alternative, the
silica gel may be weighed directly in the Im-
pinger or Its sampling holder Just prior to
the train assembly.
375
-------
7.1.2 Preliminary determinations, select
the sampling site and the minimum number
of sampling points according to Method 1 or
as specified by the Administrator. Determine
the stack pressure, temperature, and the
range of velocity heads using Method 2 and
moisture content using Approximation Meth-
od 4 or Its alternatives for the purpose of
making Isoklnetic sampling rate calculations.
Estimates may be used. However, final results
will be based on actual measurements made
during the test.
Select a nozzle size based on the range of
velocity heads such that It Is not necessary
to change the nozzle size in order to main-
tain isokinetic sampling rates. During the
run, do not change the nozzle size. Ensure
that the differential pressure gauge is capable
of measuring the minimum velocity head
value to within 10%, or as specified by the
Administrator.
Select a suitable probe liner and probe
length such that all traverse points can be
sampled. Consider sampling from opposite
sides for large stacks to reduce the length of
probes.
Select a total sampling time greater than
or equal to the minimum total sampling time
specified In the test procedures for the spe-
cific Industry such that the sampling time
per point is not less than 2 min. or select
some greater time Interval as specified by the
Administrator, and such that the sample
volume that will be taken will exceed the re-
quired minimum total gas sample volume
specified In the test procedures for the spe-
cific industry. The latter is based on an ap-
proximate average sampling rate. Note also
that the minimum total sample volume is
corrected to standard conditions.
It is recommended that a half-integral or
Integral number of minutes be sampled at
each point in order to avoid timekeeping
errors.
In some circumstances, e.g. batch cycles, it
may be necessary to sumple for shorter times
at the traverse points and to obtain smaller
gas sample volumes. In these cases, the Ad-
ministrator's approval must first be obtained.
7.1.3 Preparation of collection train. Dur-
ing preparation and assembly of the sam-
pling train, keep all openings where contami-
nation can occur covered until Just prior to
assembly or until sampling is about to begin.
Place 100 ml of water In each of the first
two implngers, leave the third impinger
empty, and place approximately 200-300 g
or more, If necessary, of preweighed silica
gel in the fourth Impinger. Record the weight
of the silica gel and container on the data
sheet. Place the empty container In a clean
place for later use in the sample recovery.
Place a filter In the filter holder. Be sure
that the filter Is properly centered and the
gasket properly placed so as to not allow the
sample gas stream to circumvent the filter.
Check filter for tears after assembly Is com-
pleted.
When glass liners are used, install selected
nozzle using a Viton A O-rlng; the Viton A
0-ring is installed as a seal where the nozzle
Is connected to a glass liner. See APTD-0576
for details. When metal liners are used, In-
stall the nozzle as above or by a leak free
direct mechanical connection. Mark the
probe with heat resistant tape or by some
other method to denote the proper distance
Into the stack or duct for each sampling
point.
Unless otherwise specified by the Admin-
istrator, attach a temperature probe to the
metal sheath of the sampling probe so that
the sensor extends beyond the probe tip and
does not touch any metal. Its position should
be about 1.9 to 2.54 cm (0.75 to 1 In.) from
jthe pitot tube and probe nozzle to avoid
interference with the gas flow.
Assemble the train as shown in Figure
13A-1 with the lilter between the third and
fourth impingcrs. Alternatively, the filter
may be placed between the probe and the
first impinger. A filter heating system may
be used to prevent moisture condensation,
but the temperature around the filter holder
shall not exceed 120±14'C (248±25°F).
((Note: Whatman No. 1 filter decomposes at
150'C (300•F)).) Record filter location on
the data sheet.
Place crushed ice around the implngers.
7.14 Leak check procedure—After the
sampling train has been assembled, turn on
and set (if applicable) the probe and filter
heating system (s) to reach a temperature
sufficient to avoid condensation In the probe.
Allow time for the temperature to stabilize.
Leak check the train at the sampling site by
plugging the nozzle and pulling a 380 mm Hg
(15 in. Hg) vacuum. A leakage rate In ex-
cess of 4% of the average sampling rate or
0.00057 m-'/min. (0.02 cfm), whichever Is less,
is unacceptable.
The following leak check instructions for
the sampling train described In APTD-0576
and APTD-0581 muy be helpful. Start the
pump with by-pass valve fully open and
coarse adjust valve completely closed. Par-
tially open the coarse adjust valve and slowly
close the by-pass valve until 380 mm Hg (15
In. Hg) vacuum is reached. Do not reverse
direction of by-pass valve. This will cause
water to back up into the filter holder. If
380 mm Hg (15 In. Hg) is exceeded, either
leak check at this higher vacuum or end the
leak check as described below and start over.
When the leak check is completed, first
slowly remove the plug from the inlet to the
probe or filter holder and Immediately turn
off the vacuum pump. This prevents the
water in the Implngers from being forced
backward Into the filter holder (if placed
before the Implngers) and silica gel from
being entrained backward Into the third
impinger.
Leak checks sho.ll be conducted as described
whenever the train is disengaged, e.g. for
silica gel or filter changes during the test,
prior to each test run, and at the completion
of each test run. If leaks are found to be In
excess of the acceptable rate, the test will be
considered Invalid. To reduce lost time due
to leakage occurrences, it is recommended
that leak checks be conducted between port
changes.
7.1.5 Particulate train operation—During
the sampling run, an isokinetic sampling rate
within 10%, or as specified by the Adminis-
trator, of true isokinetic shall be maintained.
For each run, record the data required on
the example data sheet shown in Figure 13A-
3. Be sure to recoicl tbe initial dry gas meter
reading. Record tr.e dry sjas meter readings at
the beginning and end of each sampling Ume
increment, v/nen changes hi flow rates are
made, and when sampling is halted. Take
other data point readings tit least once at
each sample point during each time Incre-
ment and additional readings when signifi-
cant changes (20% variation In velocity head
readings) necessitate additional adjustments
In flow rate. Be sure to level and zero the
manometer.
Clean the portholes prior to the test run to
minimize chance of sampling deposited
material. To begin sampling, remove the
nozzle cap, verify (if applicable) that the
probe heater Is working and filter heater K
up to temperature, and that the pltot tube
and probe are properly positioned. Position
the noz?,le at the first traverse point with the
tip pointing directly into the gas stream. Im-
mediately start the pump and adjust the
flow to Isokinetic conditions. Nomographs are
iftvallablft for sampling trains using type S
pltot tubes with 0.85±0.02 coefficients (Cn),
376
-------
'ana when sampling in nir or a stack gas with
equivalent density (molecular weight. M.I,
equal to 29 !_-4), which aid in the rapid ad-
justment of the isokinetlc sampling rate
without excessive computations. APTD-0576
details the procedure for using these nomo-
graphs. If CP and M.T are outside the above
stated ranges, do not use the nomograph
unless appropirate steps are taken to com-
pensate for the deviations.
When the stack is under significant nega-
tive pressure (height of impinger stem), take
cpre to close the coarse adjust valve before
inserting the probe into the stack to a-old
water backing into the filter holder. Jf neces-
sary, the pump may he turned on with the
coarse adjxist valve closed.
When the probe is In position, block off
the openings around the probe and porthole
to prevent unrepresentative dilution of the
gas stream.
Traverse? the stack cross section, as required
by Method 1 or as specified by the Adminis-
trator, being careful not to bump the probe
nozzle into the stack walls when sampling
near the walls cr when removing or inserting
the probe through the portholes to mi.'i'niize
chance: of extracting deposited material.
During the trst run, make periodic adjust-
ments to keep the probe and (if applicable)
filter temperatures at their proper values. Acid
more ice and, if necessary, salt to the ice
bath, to maintain a temperature of less than
20°C (68°P) at the Impinger/sillca gel outlet.
to avoid excessive moisture losses. Also, pe-
riodically check the level and zero of the
manometer.
If the pressure drop across the filter be-
comes high enough to make isokinetic sam-
pling difficult to maintain, the filter may be
replaced in the midst of a sample run. It is
recommended that another complete filter
assembly be used rather than attempting to
change the filter Itself. After the new filter or
filter assembly is installed conduct a leak
check. The final emission results shall be
based on the summation of all filter catches.
A single train shall be used for the entire
sample run, except for filter and silica gel
changes. However, If approved by the Admin-
istrator, two or more trains may be used for
a single test run when there are two or more
ducts or sampling ports. The final emission
results shall be based on the total of all
sampling train catches.
At the end of the sample run, turn off the
pump, remove the probe and nozzle from
the stack, and record the final dry gas meter
reading. Perform a leak check.' Calculate
percent Isokinetlc (see calculation section)
to determine whether another test run
should be made. If there Is difficulty In main-
taining Isokinetlc rates due to source con-
ditions, consult with the Administrator for
possible variance on the Isokinetlc rates.
7.2 Sample recovery. Proper cleanup pro-
cedure begins as soon as the probe Is re-
moved from the stack at the end of the
sampling period.
When the probe can be safely handled,
wipe off all external participate matter near
the tip of the probe nozzle and place a cap
over It to keep from losing part of the
sample. Do not cap off the probe Up tightly
while the sampling train is cooling down, as
this would create a vacuum In the filter
holder, thus drawing water from the 1m-
pingers Into the filter.
Before moving the sample train to the
cleanup site, remove the probe from the
sample train, wipe off the slllcone grease, and
cap the open outlet of the probe. Be careful
not to lose any condensate, If present. Wipe
off the slllcone grease from the filter inlet
where the probe was fastened and cap It.
i With acceptability of the test run to be
based on the same criterion as in 7.1.4.
Remove the umbilical cord from the last
Implnger and cap the Implnger. After wip-
ing off the sllicone grease, cap off the filter
holder outlet and Implnger Inlet. Ground
glass stoppers, plastic caps, or serum caps
may be used to close these openings.
Transfer the probe and fllter-impinger as-
sembly to the cleanup area. This area should
be clean and protected from the wind so that
the chances of contaminating or losing the
sample will be minimized.
Inspect the train prior to and during dis-
assembly and note any abnormal conditions.
Using a graduated cylinder, measure and re-
cord the volume of the water In the first
three Impingers, to the nearest ml; any con-
densate in the probe should be Included in
this determination. Treat the samples as
follows:
7.2.1 Container No. 1. Transfer the 1m-
pinger water from the graduated cylinder to
this container. Add the filter to this i.on-
talner. Wash all sample exposed surfaces,
Including the probe tip, probe, first three
Impingers, Implnger connectors, filter holder,
and graduated cylinder thoroughly with dis-
tilled water. Wash each component three
separate times with water and clean the
probe and nozzle with brushes. A maximum
wash of 5CO ml is used, and the washings are
added to the sample container which must
be made of polyethylene.
7.2.2 Container No. 2. Transfer the silica
gel from the fourth implnger to this con-
tainer and seal.
7.3 Analysis. Treat the contents of each
sample container as described below.
7.3.1 Container No. 1.
7.3.1.1 Filter this container's contents, In-
cluding the Whatman No. 1 filter, through
Whatman No. 541 filter paper, or equivalent
into a 1500 ml beaker. Note: If filtrate volume
i exceeds 900 ml make filtrate basic with
NaOH to phenolphthaleln and evaporate to
less than 900 ml.
7.3.1.2 Place the Whatman No. 541 filter
containing the insoluble matter (Including
the Whatman No. 1 filter) in a nickel cruci-
ble, add a few ml of water and macerate the
filter with a glass rod.
Add 100 mg CaO to the crucible and mix
the contents thoroughly to form a slurry.
Add n couple of drops of phenolphthaleln
Indicator. The indicator will turn red in a
basic medium. The slurry should remain
basic during the evaporation of the water
or fluoride Ion will be lost. If the Indicator
turns colorless during the evaporation, an
acidic condition is indicated. If this happens
add CaO until the color turns red again.
Place the crucible in a hood under infra-
red lamps or on a hot plate at low heat. Evap-
orate the water completely.
After evaporation of the water, place the
crucible on a hot plate under a hood and
slowly increase the temperature until the
paper chars. It may take several hours for
complete charring of the filter to occur.
Place the crucible in a cold muffle furnace
and gradually (to prevent smoking) Increase
the temperature to 600 °C, and maintain un-
til the contents are reduced to an ash. Re-
move the crucible from the furnace and allow
it to cool.
7.3.1.3 Add approximately 4 g of crushed
NaOH to the crucible and mix. Return the
crucible to the muffle furnace, and fuse the
sample for 10 minutes at 600°C.
Remove the sample from the furnace and
cool to ambient temperature. Using several
rinsings of warm distilled water transfer the
contents of the crucible to the beaker con-
taining the nitrate from container No. 1
(7.3.1). To assure complete sample removal.
377
-------
rinse finally with two 20 ml portions of 25
percent (v/v) sulfurlc acid and carefully add
to the beaker. Mix well and transfer to one-
liter volumetric flask. Dilute to volume with
distilled water and mix thoroughly. Allow
any uiidlssolved solids to settle.
7.3.2 Container No. 2. Weigh the spent
silica gel and report to the nearest 0.5 g.
7.3.3 Adjustment of acid/water ratio In
distillation flask—(Utilize a protective shield
when carrying out this procedure.) Place 400
ml of distilled water In the distilling flask
and add 200 ml of concentrated H^O,. Cau-
tion: Observe standard precautions when
mixing the HaSO4 by slowly adding the acid
to the flask with constant swirling. Add some
soft glass beads and several small pieces of
broken glass tubing and assemble the ap-
paratus as shown in Figure 13A-2. Heat the
flask until It reaches a temperature of 175'C
to adjust the acid/water ratio for subsequent
distillations. Discard the distillate.
7.3.4 Distillation—Cool the contents of
the distillation flask to below 80°C. Pipette
an aliquot of sample containing less than 0.6
mg F directly Into the distilling flask and add
distilled water to make a total volume of 220
ml added to the distilling flask. [For an es-
timate of what size aliquot does not exceed
0.6 mg F, select an aliquot of the solution
and treat as described in Section 7.3.6. This
will give an approximation of the fluoride
content, but only an approximation since
Interfering Ions have not been removed by
the distillation step.]
Place a 250 ml volumetric flask at the con-
denser exit. Now begin distillation and grad-
ually Increase the heat and collect all the
distillation up to 175°C. Caution: Heating
the solution above 175°C will cause sulfuric
acid to distill over.
The acid in the distilling flask can be used
until there Is carryover of Interferences or
poor fluoride recovery. An occasional check of
fluoride recovery with standard solutions Is
advised. The acid should be changed when-
ever there is less than 90 percent recovery
or blank values are higher than 0.1 Mg/ml.
Note: If the sample contains chloride, add
5 mg AgjSO, to the flask for every mg of
chloride. Gradually increase the heat and
collect all the distillate up to 175° C. Do not
exceed 176 °C.
7.3.5 Determination of Concentration—
Bring the distillate In the 250 ml volumetric
flask to the mark with distilled water and
mix thoroughly. Pipette a suitable aliquot
from the distillate (containing 10 ^g to 40
mg fluoride) and dilute to 50 ml with dis-
tilled water. Add 10 ml of SPADNS Mixed Rea-
gent (see Section 6.3.12) and mix thoroughly.
After mixing, place the sample in a con-
stant temperature bath containing the stand-
ard solution for thirty minutes before read-
Ing the absorbance with the spectropho-
tometer.
• Set the spectrophotometer to zero absorb-
ance at 570 nm with reference solution
(6.3.11), and check the spectrophotometer
calibration with the standard solution. De-
termine the absorbance of the samples and
determine the concentration from the cali-
bration curve. If the concentration does not
fall within the range of the calibration curve,
repeat the procedure using a different size
aliquot.
8. Calibration.
Maintain a laboratory log of all calibrations.
8.1 Sampling Train.
8.1.1 Probe nozzle—Using a micrometer,
measure the Inside diameter of the nozzle
to the nearest 0.025 mm (0.001 in.). Make
3 separate measurements using different
diameters each time and obtain the average
of the measurements. The difference between
the high and low numbers shall not exceed
O.lmm (0.004 In.).
When nozzles become nicked, dented, or
corroded, they shall be reshaped, sharpened,
and recalibrated before use.
Each nozzle shall be permanently and
uniquely identified.
8.1.2 Pltot tube—The pltot tube shall be
calibrated according to the procedure out-
lined in Method 2.
8.1.3 Dry gas meter and orifice meter.
Both meters shall be calibrated according to
the procedure outlined in APTD-0576. When
diaphragm pumps with by-pass valves are
used, check for proper metering system de-
sign by calibrating the dry gas meter at an
additional flow rate of 0.0057 mVmln. (0.2
cfm) with the by-pass valve fully opened
and then with It fully closed. If there Is more
than ±2 percent difference In flow rates
when compared to the fully closed position
of the by-pass valve, the system Is not de-
signed properly and must be corrected.
8.1.4 Probe heater calibration—The probe
heating system shall be calibrated according
to the procedure contained in APTD-0576.
Probes constructed according to APTD-0581
need not be calibrated If the calibration
curves In APTD-0576 are used.
8.1.5 Temperature gauges—Calibrate dial
and liquid filled bulb thermometers against
mercury-ln-glass thermometers. Thermo-
couples need not be calibrated. For other
devices, check with the Administrator.
8.2 Analytical Apparatus. Spectrophotom-
eter. Prepare the blank standard by adding
10 ml of SPADNS mixed reagent to 50 ml of
distilled water. Accurately prepare a series
of standards from the standard fluoride solu-
tion (see Section 6.3.9) by diluting 2, 4, 6,
8, 10, 12, and 14 ml volumes to 100 ml with
distilled water. Pipette 50 ml from each solu-
tion and transfer to a 100 ml beaker. Then
add 10 ml of SPADNS mixed reagent to each.
These standards will contain 0, 10, 20, 30,
4"0', 50, 60, and 70 Ag of fluoride (0—1.4 Ag/ml)
respectively.
After mixing, place the reference standards
and reference solution In a constant tem-
perature bath for thirty minutes before read-
Ing the absorbance with the spectrophotom-
eter. All samples should be adjusted to this
same temperature before analyzing. Since
a 3°C temperature difference between samples
and standards will produce an error of ap-
proximately 0.005 mg F/liter, care mvst be
taken to sea that samples and standards are
at nearly Identical temperatures when ab-
sorbances are recorded.
With the spectrophotometer at 570 nm,
use the reference solution (see section 6.3.11)
to set the absorbance to zero.
Determine the absorbance of the stand-
ards. Prepare a calibration curve by plotting
11% F/50 ml versus absorbance on linear graph
paper. A standard curve should be prepared
initially and thereafter whenever the
SPADNS mixed reagent Is newly made. Also,
a calibration standard should be run with
each set of samples and if It differs from the
calibration curve by ±2 percent, a new
standard curve should be prepared.
9. Calculations.
Carry out calculations, retaining at least
one extra decimal figure beyond that of the
acquired data. Round off figures after final
calculation.
9.1 Nomenclature.
At — Aliquot of distillate taken for color
development, ml.
An—Cross sectional area of nozzle, m" (ft5).
Ai = Aliquot of total sample added to still,
ml.
B»i=Water vaoor In the gas stream, propor-
tion by voluma.
378
-------
'C, = Concentration of fluoride In stack gas.
mg/m', corrected to standard conditions
of 20- C. 760 mm Hg (68' P. 29.92 In. Hg)
on dry basis.
Fi=Total weight of fluoride In sample, mg.
jiSF.F'r: Concentration from the calibration
curve, Mg.
7=Percent of Isoklnetlc sampling.
ntn=Total amount of particulate matter
collected, mg.
Jf« —Molecular weight of water, 18 g/g-mole
(18 Ib/lb-mole).
ma = Mass of residue of acetone after evap-
oration, mg.
Jf>i,ar = Barometric pressure at the sampling
site, mm Hg (In. Hg).
P. = Absolute stack gas pressure, mm Hg (In.
Hg).
P»i
(in. Hi-O).
pa = Density of acetone, nig/ml (see label on
bottle).
Plr=- Density of water, 1 g/ml (O.C0220 lb.'
ml).
0 = Total sampling time, min.
13.6 = Specific gravity of mercury.
60 = Sec/min.
100 = Conversion to percent.
9.2 Average dry gas meter temperature
and average orifice pressure drop. See data
sheet (fig. 13A-3).
9.3 Dry gas volume. Correct the sample
volume measured by the dry gas meter to
standard conditions [20° C, 760 mm KS (68s
P, 29.92 Inches Hg) ] by using equation
13A-1.
Vm(,td) = Vm —=—
rPba +M.
\ I*-"
I jrj
[_ 1 ltd
= KVm
Tm
where:
K=0.3855 °K/mm Hg for metric units.
= 17.65 "R/in. Hg for English units.
9.4 Volume of water vapor.
F»(i((0 = T
where:
K=0.00134 mVml for metric units.
= 0.0472 ftvml for English unite.
9.5 Moisture content.
Ic -.7-
r,"
r,td
equation 13A-1
equation 13A-2
r u(afrf)
equation 13A-3
If the liquid droplets are present in the
gas stream assume the stream to be saturated
and use a psychrometrlc chart to obtain an
approximation of the moisture percentage.
9.6 Concentration.
9.6.1 Calculate the amount of fluoride In
the sample according to Equation 13A-4,
equation 13A-4
•where:
K = lo-~ me/eg.
9.6.2 Concentration of fluoride in stack
gas. Determine the concentration of fluoride
In the stack gas according to Equation 13A-5.
equation 13A-5
where:
#=35.31 ftVm».
379
-------
9.7 Isokinetlc variation.
9.7.1 Calculations from raw data.
100 T. \KVu-\-(VJTm) (Ptar+AH/13.C)]
00 8t), P./ln
equation 13A-6
where:
K = 0.00346 mm Hg-mVml-°K for metric
units.
=0.00267 In. Hg-ftVml-°R for English
units.
9.7.2 Calculations from Intermediate val-
ues.
r
100
(l-Bat)
T.V,
P.v.A.6
where:
K = 4.323 for metric units.
=0.0944 for English units.
9.8 Acceptable results. The following
range sets the limit on acceptable isokinetlc
sampling results:
If 90 percent
-------
Reference Method ISA Is amended
as follows:
(a) In section 3., the phrase "300
/ig/liter" is corrected to read "300 rag/
liter" and the parenthetical phrase "(see
section 7.3.8)" is corrected to read "(see
section 7.3.4)"-
(b) Section 5.1.5 is revised to read as
follows:
5.1.5 Filter bolder—If located between the
probe and first Implnger. borosUlcate glass
with a 20 mesh stainless steel screen niter
support and a silicons rubber gasket; neither
a glass frit filter support nor a sintered metal
filter support may be used if the filter Is In
front of the Implngers. If located between
the third and fourth Implngers, borosUlcate
glass with a glass frit filter support and a
slllcone rubber gasket. Other materials of
construction may be used with approval from
the Administrator, e.g.. If probe liner Is stain-
less steel, then filter holder may be stainless
steel. The holder design shall provide a posi-
tive seal against leakage from the outside or
around the filter.
(c) Section 7.1.3 is amended by re-
vising the first two sentences of the sixth
paragraph to read as follows:
7.1.3 Preparation of collection train. • • •
Assemble the train as shown In Figure
13A-1 with the filter between the third and
fourth Implngers. Alternatively, the filter
may be placed between the probe and first
Implnger If a 20 mesh stainless steel screen
Is used for the filter support. • • •
• • • • •
(d) In section 7.3.4, the reference in
the first paragraph to "section 7.3.6" is
corrected to read "section 7.3.5".
381
-------
METHOD 13B
DETERMINATION OF TOTAL FLUORIDE EMISSIONS FROM STATIONARY SOURCES
SPECIFIC ION ELECTRODE METHOD
1. Principle and. Applicability.
1 1 Principle. Gaseous and paniculate flu-
orides are withdrawn IsoklneticaUy from the
.soiree using a sampling train. The fluorides
are collected in the impliiger water and on
the filter of the sampling train. The weight
of total fluorides In the train is determined
by the specific Ion electrode method.
1 2 Applicability. This method Is ap-
plicable for the determination of fluoride
ch'.lssicns from stationary sources only when
specified by the test procedures for deter-
mining compliance with new source per-
formance standards. Fluorocarbons such as
Freons, are not quantitatively collected or
measured by this procedure.
2. Range and Sensitivity.
The fluoride specific Ion electrode analyti-
cal method covers the range of 0.02-2,000 /ig
P ml; however, measurements of less than
0.1 Mg F/ml require extra care. Sensitivity has
net been determined.
3. Interferences.
Daring the laboratory analysis, aluminum
in excess of 300 mg/liter and silicon dioxide
in excess of 300 /ig/liter will prevent complete
recovery of fluoride.
4. Precision, Accuracy and Stability.
Tr.e accuracy-of-ftuorlde electrode measure-
ments has been reported by various re-
searrhers to be In the range of 1-5 percent In
a concentration range of 0.04 to 80 mg/1. A
change in the temperature of the sample will
change the electrode ruspoase; a change of.
1°C will produce a 1.5 percent relative error
in the measurement. Lack of stability in the
electrometer usod to measxire EMF can intro-
duce error. An error of 1 millivolt In the EMF
measurement produces a relative error of 4
percent regardless of the absolute concen-
tration being measured.
5. Apparatus.
5.1 Sample train. See Figure 13A-1
(Method 13A); it is similar to the Method 5
train except for the interclmngeability of
the position of the filter. Commercial models
of this train are available. However, If one
desires to build his own, complete construc-
tion details are described in APTD-0531; foi;
changes from the APTD-0581 document "and
for allowable modifications to Figure 13A-1,
see the following subsections.
The operating and maintenance procedures
for the sampling train are described in
APTD-0576. Since correct usarre Is impor-
tant in obtaining valid results, all users
should read the APTD-0576 document nnd
adopt the operating and maintenance pro-
cedures outlined In it, unless otherwise spec-
ified herein.
5.1.1 Probe nozzle—Stainless steel (316)
with sharp, tapered leading edge. The angle,
of taper shall be S30° and the taper shall be
on the outside to preserve a constant Inter-!
nal diameter. The probe nozzle shall be of'
the button-hook or elbow design, unless!
otherwise specified by the Administrator.
The wall thickness of the nozzle shall be
less than or equal to that of 20 gauge tub-
Ing, I.e., 0.165 cm (0.065 in.) and the distance
from the tip of the nozzle to the trst bend
or point of disturbance shall be at least two
times the outside nozzle diameter. The noz-
zle shall be constructed from seamless stain-
less steel tubing. Other configurations and
construction material may be used with ap-
proval from the Administrator.
A range of sizes suitable for Isokinetlc
sampling should be available, e.g., 0.32 cm
("/a in.) up to 1.27 cm (\'2 in.) (or larger If
'higher volume sampling trains are used)
Inside diameter (ID) nozzles in Increments
of 0.16 cm (i/io in.). Each nozzle shall be
calibrated according to the procedures out-
lined in the calibration section.
5.1.2 Probs liner—Borosllicate glass or
stainless steel (316). When the filter Is lo-
cated Immediately after the probe, a probe
heating system may be used to prevent filter
plugging resulting from moisture conden-
sation. The temperature In the probe shall
not exceed 120±14CC (24B±25°F).
5.1.3 Pilot tube—Type S, or other device
approved by the Administrator, attached to
probe to allow constant monitoring of the
stack gas velocity. The face openings of the
pitot tube and the probe nozzle shall be ad-
jacent and parallel to each other, not neces-
isarily on the same plane, during sampling.
.The free space between the nozzle and pitot
tube shall be at least 1.9 cm (0.75 in.). The
'free space shall be set based on a 1.3 cm
(0.5 In.) ID nozzle, which Is the largest size
nozzle used.
The pitot tube must also meet the criteria
specified In Method 2 and be calibrated ac-
cording to the procedure In the calibration
section of that method.
5.1.4 Differential pressure gauge—In-
clined manometer capable of measuring
velocity head to within 10 percent of the
minimum measured value. Below a differen-
tial pressure of 1.3 mm (0.05 In.) water
gauge, mlcromanometers with sensitivities
of 0.013 mm (0.0005 In.) should be used.
However, mlcromanometers are not easily
adaptable to field conditions and are not
easy to use with pulsating flow. Thus, other
methods or devices acceptable to the Ad-
ministrator may be used when conditions
warrant.
5.1.5 Filter holder—Borosllicate glass
with a glass frit filter support and a slllcone
rubber gasket. Other materials of construc-
tion may be used with approval from the
Administrator, e.g. If probe liner Is stain-
less steel, then filter holder may be stainless
steel. The holder design shall provide a posi-
tive seal against leakage from the outside
or around the filter.
5.1.6 Filter heating system—When mois-
ture condensation Is a problem, any heating
system capable of maintaining a temperature
around the filter holder during sampling of
no greater than 120±14°C (248 ±25°F). A
temperature gauge capable of measuring tem-
perature to within 3°C (5.4°F) shall be In-
stalled so that when the filter heater Is used,
the temperature around the filter holder can
be regulated and monitored during sampling.
382
-------
Healing systems other than the one shown
In APTD-0581 may be used.
5.1.7 Impingers—Four Impingers con-
nected as shown In Figure 13A-1 with ground
glass (or equivalent), vacuum tight fittings.
The first, third, and fourth impingers are of
the Greenburg-Smlth design, modified by re-
placing the tip with a 1% cm (>,£ in.) insido
diameter glass tube extending to 1V4 c'n (V4
in.) from the bottom of the flask. The second
impinger is of the Greenburg-Smith design
with the standard tip.
5.1.8 Metering system—Vacuum gauge.
leak-free pump, thermometers capable of
measuring temperature to within 3"C
(~5°F), dry gas meter with 2 percent ac-
curacy at the required sampling rate, and
relr.ted equipment, or equivalent, as required
to maintain an isoklnetlc sampling rate and
to determine sample volume. When the
metering system is used In conjunction with
a pltot tube, the system shall enable checks
of isokinetic rates.
5.1.9 Barometer—Mercury, anercld, or
other barometers capable of measuring at-
mospheric pressure to within 2.5 mm Hg (0.1
in Hg). In many cases, the barometric read-
ing may be obtained from a nearby weather
bureau station, in which case the station
value shall be requested and an adjustment
for elevation differences shall be applied at a
rate of minus 2.5 mm Hg (0.1 In. Hg) per 30
m (100 ft) elevation Increase.
6.2 Sample recovery.
5.2.1 Probe liner and probe nozzle
brushes—Nylon bristles with stainless steel
wire handles. The probe brush shall have
extensions, at least as long as the probe, of
stainless steel, teflon, or similarly inert mate-
rial. Both brushes shall be properly sized and
shaped to brush out the probe liner and noz-
zle.
5.2.2 Glass wash bottles—Two.
5.2.3 Sample storage containers—Wide
mouth, high density polyethylene bottles, 1
liter.
5.2.4 Plastic storage containers—Air tight
containers of sufficient volume to store silica
gel.
5.2.5 Graduated cylinder—250 ml.
5.2.6 Funnel and rubber policeman—To
aid In transfer of silica gel to container; not
necessary If silica gel is weighed In the field.
5.3 Analysis.
5.3.1 Distillation apparatus—Glass distil-
lation apparatus assembled as shown in Fig-
ure 13A-2 (Method ISA).
5.3.2 Hot plate—Capable of heating to
500"C.
5.3.3 Electric muffle furnace—Capable of
heating to 600°C.
5.3.4 Crucibles—Nickel, 75 to 100 ml
capacity.
5.3.5 Beaker—1500 ml.
5.3.6 Volumetric flask—50 ml.
5.3.7 Erlenmeyer flask or plastic bottle—
500 ml.
5.3.8 Constant temperature bath—Cap-
able of maintaining a constant temperature
of ±1.0°C In the range of room temperature.
5.3.9 Trip balance—300 g capacity to
measure to ±0.5 g.
5.3.10 Fluoride ion activity sensing elec-
trode.
5.3.11 Reference electrode—^Single Junc-
tion; sleeve type. (A combination-type elec-
trode having the references electrode and
the fluoride-Ion sensing electrode built Into
one unit may also be used).
5.3.12 Electrometer—A pH meter with
millivolt scale capable of ±0.1 mv resolu-
tion, or a specific ion meter made specifically
for specific ion use.
5.3.13 Magnetic stlrrer and TFE fluoro-
carbon coated stripping bars.
6. Reagents.
6.1 Sampling.
6.1.1 Filters—Whatman No. 1 filters, or
equivalent, sized to fit filter holder.
6.1.2 Silica gel—Indicating type, 6-16
mesh. If previously used, dry at 175°C
(350°F) for 2 hours. New silica gel may be
used as received.
6.1.3 Water—Distilled.
6.1.4 Crushed ice.
6.1.5 Stopcock grease—Acetone Insoluble,
heat stable silicone grease. This is not neces-
sary If screw-on connectors with teflon
sleeves, or similar, are used.
6.2 Sample recovery.
6.2.1 Water—Distilled from same con-
tainer as 6.1.3.
6.3 Analysis.
6.3.1 Calcium oxide (CaO)—Certified
grade containing 0.005 percent fluoride or
less.
6.3.2 Phenolphthalein Indicator—0.1 per-
cent in 1:1 ethanol water mixture.
6.3.3 Sodium hydroxide (NaOH)—Pel-
lets, ACS reagent grade or equivalent.
6.3.4 Sulfurlc acid (HjSOJ—Concen-
trated. ACS reagent grade or equivalent.
6.3.5 Filters—Whatman No. 541, or
equivalent.
6.3.6 Water—Distilled, from same con-
tainer as 6.1.3.
6.3.7 Total Ionic Strength Adjustment1
Buffer (TISAB)—Place approximately 500
ml of distilled water In a 1-llter beaker. Add
57 ml glacial acetic acid, 58 g sodium chlo-
ride, and 4 g CDTA (Cyclohexylene dinitrilo
tetraacetlc acid). Stir to dissolve. Place the
beaker In a water bath to cool It. Slowly
add 5 M NaOH to the solution, measuring
the pH continuously with a calibrated pH/
reference electrode pair, until the pH Is 5.3.
Cool to room temperature. Pour into a 1-llter
flask and dilute to volume with distilled
water. Commercially prepared TISAB buffer
may be substituted for the above.
6.3.8 Fluoride Standard Solution—0.1 M
fluoride reference solution. Add 4.20 grams of
reagent grade sodium fluoride (NaF) to a 1-
llter volumetric flask and add enough dis-
tilled water to dissolve. Dilute to volume
with distilled water.
7. Procedure.
NOTE: The fusion and distillation steps of
this procedure will not be required, If It can
be shown to the satisfaction of the Admin-
istrator that the samples contain only water-
soluble fluorides.
7.1 Sampling. The sampling shall be con-
ducted by competent personnel experienced
with this test procedure.
7.1.1 Pretest preparation. All train com-
ponents shall be maintained and calibrated
according to the procedure described in
APTD-0576, unless otherwise specified
herein.
Weigh approximately 200-300 g of silica gel
In air tight containers to the nearest 0.5 g.
Record the total weight, both silica gel and
container, on the container. More silica gel
may be used but care should be taken during
sampling that It Is not entrained and carried
out from the impinger. As an alternative, the
silica gel may be weighed directly in the Im-
pinger or Its sampling holder just prior to
the train assembly.
7.1.2 Preliminary determinations. Select
the sampling site and the minimum number
of sampling points according to Method 1 or
as specified by the Administrator. Determine
the stack pressure, temperature, and the
range of velocity heads using Method 2 and
moisture content using Approximation
Method 4 or Its alternatives for the purpose
of making Isoklnetlc sampling rate calcula-
tions. Estimates may be used. However, final
383
-------
results will be based on actual measure-
ments made during the test.
Select a nozzle size based on the range of
velocity heads such that It Is not necessary
to change the nozzle size In order to maintain
Isolclnetlc sampling rates. During the run, do
not change the nozzle size. Ensure that the
differential pressure gauge Is capable of
measuring the minimum velocity head value
to within 10 percent, or as specified by the
Administrator.
Select a suitable probe liner and probe
length such that all traverse points can be
sampled. Consider sampling from opposite
sides for large stacks to reduce the length of
probes.
Select a total sampling time greater than
or equal to the minimum total sampling
time specified In the test procedures for the
specific Industry such that the sampling time
per point Is not less than 2 mln. or select
some greater time Interval as specified by
the Administrator, and such that the sample
volume that will be taken will exceed the re-
quired minimum total gas sample rolume
specified In the test procedures for the spe-
cific Industry. The latter Is based on an ap-
proximate average sampling rate. Note also
that the minimum total sample volume Is
corrected to standard conditions.
It Is recommended that a half-Integral or
Integral number of minutes be sampled at
each point In order to avoid timekeeping
errors.
In some circumstances, e.g. batch cycles, It
may be necessary to sample for shorter times
at the traverse points and to obtain smaller
gas sample volumes. In these cases, the Ad-
ministrator's approval must first be obtained.
7.1.3 Preparation of collection train. Dur-
ing preparation and assembly of the sampling
train, keep all openings where contamination
can occur covered until Just.prlor to assembly
or until sampling Is about to begin.
Place 100 ml of water In each of the first
two impiugers, leave the third Implnger
empty, and place approximately 200-300 g or
more, if necessary, of prewelghed silica gel In
the fourth Implnger. Record the weight of
the silica gel and container on the data sheet.
Place the empty container in a clean place
for later use in the sample recovery.
Place a filter In the filter holder. Be sure
that the filter Is properly centered and the
gasket properly placed so as to not allow the
sample gas stream to circumvent the filter.
Check filter for tears after assembly is com-
pleted.
When glass liners are used, install selected
nozzle using a Vlton A O-rlng; the Viton A
O-ring is installed as a seal where the nozzle
Is connected to a glass liner. See APTD-0576
for details. When metal liners are used. In-
stall the nozzle as above or by a leak free
direct mechanical connection. Mark the probe
with heat resistant tape or by some other
method to denote the proper distance into
the stack or duct for each sampling point.
Unless otherwise specified by the Admin-
istrator, attach a temperature probe to the
metal sheath of the sampling probe so that
the sensor extends beyond the probe tip and
does not touch any metal. Its position should
be about 1.9 to 2.54 cm (0.75 to 1 In.) from
the pitot tube and probe nozzle to avoid in-
terference with the gas flow.
Assemble the train as shown In Figure
13A-1 (Method 13A) with the filter between
the third and fourth impingers. Alterna-
tively, the filter may be placed between the
probe and first Implnger. A filter heating sys-
tem may be used to prevent moisture con-
densation, but the temperature around the
filter holder shall not exceed 1200±14°C
(243-25-F). [(Note: Whatman No. 1 filter
decomposes at 150"C (300°P)).] Record
filter location on the data sheet.
Place crushed ice around the Impingers.
7.1.4 Leak check procedure—After the
sampling train has been assembled, turn on
and set (If applicable) the probe and filter
heating system(s) to reach a temperature
sufficient to avoid condensation in the probe.
Allow time for the temperature to stabilize.
Leak check the train at the sampling site by
plugging the nozzle and pulling a 380 mm
Hg (15 in. Hg) vacuum. A leakage rate in ex-
cess of 4% of the average sampling rate of
0.0057 mVmln. (0.02 cfm), whichever Is less,
is unacceptable.
The following leak check Instruction for
the sampling train described in APTD-0576
and APTD-0581 may be helpful. Start the
pump with by-pass valve fully open and
coarse adjust valve completely closed. Par-
tially open the coarse adjust valve and slow-
ly close the by-pass valve until 380 mm Hg
(15 In. Hg) vacuum is reached. Do not re-
verse direction of by-pass valve. This will
cause water to back up Into the filter holder.
If 380 mm Hg (15 In. Hg) Is exceeded, either
leak check at this higher vacuum or end the
leak check as described below and start over.
When the leak check is completed, first
slowly remove the plug from the Inlet to the
probe or filter holder and immediately turn
off the vacuum pump. This prevents the
water In the Impingers from being forced
backward into the filter holder (if placed
before the impingers) and silica gel from
being entrained backward into the third
impinger.
Leak checks shall be conducted as de-
scribed whenever the train is disengaged, e.g.
for silica g«l or filter changes during the test,
prior to each test run, and at the completion
of each test run. If leaks are found to be in
excess of the acceptable rate, the test will be
considered invalid. To reduce lost time due to
leakage occurrences, it Is recommended that
leak checks be conducted between port
changes.
7.1.5 Particulate train operation—During
the sampling run, an Isokinetic sampling
rate within 10%, or as specified by the Ad-
ministrator, of true isokinetic shall be main-
tained.
For each run, record the data required on
the example data sheet shown in Figure
13A-3 (Method 13A). Be sure to record the
initial dry gas meter reading. Record the
dry gas meter readings at the beginning and
end of each sampling time increment, when
changes in flow rates are made, and when
sampling is halted. Take other data point
readings at least once at each sample point
during each time Increment and additional
readings when significant changes (20%
variation In velocity head readings) neces-
sitate additional adjustments in flow rate. Be
sure to level and zero the manometer.
Clean the portholes prior to the test run
to minimize chance of sampling deposited
material. To begin sampling, remove the
nozzle cap, verify (if applicable) that the
probe heater Is working and filter heater Is
up to temperature, and that the pitot tub*
and probe are properly positioned. Position
the nozzle at the first traverse point with
the tip pointing directly into the gas stream.
Immediately start the pump and adjust the
flow to isokinetic conditions. Nomographs are
available for sampling trains using type S
pitot tubes with 0.85±0.02 coefficients (Ci>),
and when sampling in air or a stack gas with
equivalent density (molecular weight, Md,
equal to 29±4), which aid in the rapid ad-
justment of the Isokinetic sampling rate
without excessive computations. APTD-0576
details the procedure for using these nomo-
graphs. If Cp and Md are outside the above
stated ranges, do not use the nomograph un-
384
-------
less appropriate steps are taken to compen-
sate for the deviations.
When the stack Is under significant neg-
ative pressure (height of Implnger stem),
take care to close the coarse adjust valve
before Inserting the probe into the stack to
avoid water backing Into the filter holder. If
necessary, the pump may be turned on with
the coarse adjust valve closed.
When the probe is in position, block off
the openings around the nrobe and porthole
to prevent unrepresentative dilution of the
gas stream.
Traverse the stack cross section, as re-
quired by Method 1 or as specified by the Ad-
ministrator, being careful not to bump the
probe nozzle into the stack walls when
sampling near the walls or when removing-
or Inserting the probe through the port-
holes to minimize chance of extracting de-
posited material.
During the test run, make periodic adjust-
ments to keep the probe and (if applicable)
alter temperatures at their proper values.
Add more Ice and. If necessary, salt to the
Ice bath, to maintain a temperature of ICES
than 20"C (68°F) at the Impinger/sllica gel
outlet, to avoid excessive moisture losses.
Also, periodically check the level and zero
of the manometer.
If the pressure drop across the filter be-
comes high enough to make Isoklnetic sam-
pling difficult to maintain, the filter may be
replaced In the midst of a sample run. It Is
recommended that another complete filter as-
sembly be used rather than attempting to
change the filter itself. After the new filter
or filter assembly Is Installed, conduct a
leak check. The final emission results shall
be based on the summation of all filter
catches.
A single train shall be used for the entire
sample run, except for filter and silica gel
changes. However, if approved by the Admin-
istrator, two or more trains may be used for
a single test run when there are two or more
ducts or sampling ports. The final emission
results shall be based on the total of all
sampling train catches.
At the end of the sample run, turn off the
pump, remove the probe and nozzle from
the stack, and record the final dry gas meter
reading. Perform a leak check.1 Calculate
percent Isokinetic (see calculation section) to
determine whether another test run should
be made. If there Is difficulty In maintaining
isoklnetic rates due to source conditions, con-
sult with the Administrator for possible
variance on the Isokinetic rates.
7.2 Sample recovery. Proper cleanup pro-
cedure begins as soon as the probe is re-
moved from the stack at the end of the
sampling period.
When the probe can be safely handled,
wipe off all external partlculate matter near
the Up of the probe nozzle and place a cap
over It to keep from losing part of the sam-
ple. Do not cap off the probe tip tightiy
while the sampling train is cooling down,
as this would create a vacuum in the filter
holder, thus drawing water from the im-
plngers into the filter.
Before moving the cample train to the
cleanup site, remove the probe from the
sample train, wipe off the sillcone grease,
and cap the open outlet of the probe. Be
careful not to lose any condensate, if pre->-
ent. Wipe off the sllicoue grease from the,
filter inlet where the probe was fastened:
and cap it. Remove the umbilical cord from
the last implnger and cap the irapinger. After
wiping off the silicone grease, cap oil the
niter holder outlet and Impinger inict.
1 With acceptability, of the test run to be
based on the same criterion as In 7.1.4.
Ground glaso stoppers, plastic caps, rr ser.i.-i
caps mav be used to close these opening."-.
Tran-f-.-r the probe and filter-impinger as-
sembly to the cleanup area. This area -.iiou:.1
be clean anu protected from the wire! i.i Hi- t
the chances of contaminating or losing the
sample will be minimized.
Inspect the train prior to and during cii .-
assembly and note any abnormal conditions.
Using a graduated cylinder, measure and re-
cord the volume of the water In the fir-t
three impingers, to the nearest ml: a:iy con-
densate in the probe should be included in
this determination. Treat the sampler as
follows:
7.2.1 Container No. 1. Transfer the Im-
pinger water from the graduated cylinder
to this container. Add the filter to this
container. Wash all sample exposed sur-
faces, Including the probe tip, probe, first
three impingers, implnger connectors, filter
holder, and graduated cylinder thoroughly
with distilled water. Wash each component
three separate times with water and clean
the probe and nozzle with brushes. A max-
imum wash of 500 ml is used, and the wash-
Ings are added to the sample container
which must be made of polyethylene.
7.2.2 Container No. 2. Transfer the silica
gel from the fourth implnger to this con-
tainer and seal.
7.3 Analysis. Treat the contents of each
sample container as described below.
7.3.1 Container No. 1.
7.3.1.1 Filter this container's contents, In-
cluding the Whatman No 1 filter, through
Whatman No. 541 filter paper, or equivalent
Into a, 1500 ml beaker. NOTE: If nitrate vol-
ume exceeds 900 ml make filtrate basic with
NaOH to phenclphthalein and evaporate to
less than 900 ml.
7.3.1.2 Place the Whatman No. 541 filter
containing the insoluble matter (Including
the Whatman No. 1 filter) to a nickel cru-
cib'e, add a few ml of water and macerate
the filter with a glass rod.
Add 100 rag CaO to the crucible and mix
the contents thoroughly to form a slurry. Add
a couple of drops of phenolphthaleln Indi-
cator. The indicator will turn red in a basic
medium. The slurry should remain basic
during the evaporation of the water or
fluoride Ion will be lost. If the Indicator
turns colorless during the evaporation, an
acidic condition is Indicated. If this happens
add CaO until the color turns red again.
Place the crucible In » hood under In-
frared lamps or on a hot plate at low heat.
Evaporate the water completely.
After evaporation of the water, place the
crucible on a hot plate under a, hood and
slowly increase the temperature until the
paper chars. It may take several hours for
complete charring of the filter to occur.
Place the crucible In a cold muffle furnace
and gradually (to prevent smoking) Increase
the temperature to 600°C, and maintain until
the contents are reduced to an ash. Remove
the crucible from the furnace and allow it to
cor.l.
7.3.1.3 Add approximately 4 g of crushed
NaOH to the crucible and mix. Return the
crucible to the muffle furnace, and fuse the
sample for 10 minutes at 600°C.
Remove the sample from the furnace and
cool to ambient temperature. Using several
rinsings of warm distilled water transfer
the contents of the crucible to the beaker
containing the filtrate from container No.
1 (7.3.1). To assure complete sample re-
moval, rinse finally with two 20 ml portions
of 25 percent (v/v) snlfurlc acid and care-
fully add to the beaker. Mix well and trans-
fer to a one-liter volumetric flask. Dilute
385
-------
to volume with distilled water and mix
thoroughly. Allow any undlssolved solids to
settle.
7.3.2 Container No. 2. Weigh the spent
silica gel and report to the nearest 0.5 g.
7.3.3 Adjustment of acid/water ratio In
distillation flask—(Utilize a protective shield
when carrying out this procedure). Place 400
ml of distilled water In the distilling flask
and add 200 ml of concentrated H..SO,. Cau-
tion: Observe standard precautions when
mixing the H,SO4 by slowly adding the acid
to the flask with constant swirling. Add some
soft glass beads and several small pieces of
broken glass tubing and assemble the ap-
paratus as shown In Figure 13A-2. Heat the
flask until it reaches a temperature of 175°C
to adjust the acid/water ratio for subsequent
distillations. Discard the distillate.
7.3.4 Distillation—Cool Uie contents of
tho distillation flask to below 80'C.- Pipette
an aliquot of sample containing less
than 0.6 mg F directly into the distilling
flask and add distilled water to make a total
volume of 220 ml added to the distilling
flask. [For an estimate of what size aliquot
does not exceed 0.6 mg F, select an aliquot
of the solution and treat as described In
Section 7.3.6. This will give an approxima-
tion of the fluoride content, but only an ap-
proximation since Interfering ions have not
been removed by the distillation step.]
Place a 250 ml volumetric flask at the con-
denser exit. Now begin distillation and
gradually Increase the heat and collect all the
distillate up to 175°C. Caution: Heating the
solution above 175°C will cause sulfuric acid
to distill over.
The acid in the distilling flask can be
used until there is carryover of interferences
or poor fluoride recovery. An occasional
check of fluoride recovery with standard
solutions is advised. The acid should
be changed whenever there is less than 90
percent recovery or blank values are higher
than 0.1 ^g/ml.
7.3.5 Determination of concentration—
Bring the distillate in the 250 ml volumetric
flask to the mark with distilled water and
mix thoroughly. Pipette a 25 ml aliquot from
the distillate. Add an equal volume of TISAB
and mix. The sample should be at the
same temperature as the calibration stand-
ards when measurements are made. If
ambient lab temperature fluctuates more
than ±2°C from the temperature at which
the calibration standards were measured,
condition samples and standards in a con-
stant temperature bath measurement. Stir
the sample with a magnetic stirrer during
measurement to minimize electrode response
time. If the stirrer generates enough heat to
'change solution temperature, place a piece
of insulating material such as cork
between the stirrer and the beaker. Dilute
samples (below 10-* M fluoride ion content)
should be held in polyethylene or poly-
propylene beakers during measurement.
Insert the fluoride and reference electrodes
Into the solution. When a steady millivolt
reading is obtained, record It. This may take
several minutes. Determine concentration
from the calibration curve. Between elec-
trode measurements, soak the fluoride sens-
ing electrode in distilled water for 30 seconds
and then remove and blot dry.
8. Calibration.
Maintain a laboratory log of all
calibrations.
8.1 Sampling Train.
8.1.1 Probe nozzle—Using * micrometer,
measure the inside diameter of the nozzle
to the nearest 0.025 mm (0.001 in.). Make
3 separate measurements using different
diameters each time and obtain the average
of the measurements. The difference between
the high and low numbers shall not exceed
0.1 mm (0.004 In.).
When nozzles become nicked, dented, or
'corroded, they shall be reshaped, sharpened,
and recalibrated before use.
Each nozzle shall be permanently and
uniquely identified.
8.1.2 Pltot tube—The pltot tube shall be
calibrated according to the procedure out-
lined in Method 2.
8.1.3 Dry gas meter and orifice meter.
Both meters shall be calibrated according to
the procedure outlined in APTD-0576. When
diaphragm pumps with by-pass valves are
used, check for proper metering system
design by calibrating the dry gas meter at an
additional flow rate of 0.0057 mVmln. (0.2
cfm) with the by-pass valve fully opened
and then with it fully closed. If there is
more than -±2 percent difference in flow
rates when compared to the fully closed posi-
tion of the by-pass valve, the system is not
designed properly and must be corrected.
8.1.4 Probe heater calibration—The probe
heating system shall be calibrated according
to the procedure contained In APTD-0576.
Probes constructed according to APTD-0581
need not be calibrated if the calibration
curves in APTD-0576 are used.
8.1.5 Temperature gauges—Calibrate dial
and liquid filled bulb thermometers against
mercury-in-glass thermometers. Thermo-
couples need not be calibrated. For other
devices, check with the Administrator.
8.2 Analytical Apparatus.
8.2.1 Fluoride Electrode—Prepare fluoride
standardizing solutions by serial dilution of
the 0.1 M fluoride standard solution. Pipette
10 ml of 0.1 M NaF into a 100 ml volumetric
flask and make up to the mark with distilled
water for a 10-* M standard solution. Use 10
ml of 10-2 M solution to make a 10-3 M solu-
tion in the same manner. Repeat for 10-4 and
10-5 M solutions.
Pipette 50 ml of each standard Into a sep-
arate beaker. Add 50 ml of TISAB to each
beaker. Place the electrode in the most dilute
standard solution. When a steady millivolt
reading is obtained, plot the value on the
linear axis of semi-log graph paper versus
concentration on the log axis. Plot the
nominal value for concentration of the
standard on the log axis, e.g., when 60 ml of
10-2 M standard Is diluted with 50 ml TISAB.
the concentration is still designated "10-2 M".
Between measurements soak the fluoride
sensing electrode in distilled water for 30
seconds, and then remove and blot dry.
Analyze the standards going from dilute to
concentrated standards. A straight-line cali-
bration curve will be obtained, with nominal
concentrations of 10-5, 10-', 10-", 10-', 10->
fluoride molarity on the log axis plotted
versus electrode potential (In millivolts) on
the linear scale.
Calibrate the fluoride electrode dally, and
check It hourly. Prepare fresh fluoride stand-
ardizing solutions dally of 10-= M or less.
Store fluoride standardizing solutions in
polyethylene or polypropylene containers.
(Note: Certain specific Ion meters have been
designed specifically for fluoride electrode
use and give a direct readout of fluoride Ion
concentration. These meters may be used In
lieu of calibration curves for fluoride meas-
urements over narrow concentration ranges.
Calibrate the meter according to manufac-
turer's Instructions.)
9. Calculations.
Carry out calculations, retaining at least
one extra decimal figure beyond that of the
386
-------
acquired data. Round off figures after final
calculation.
9.1 Nomenclature.
A*=Cross sectional area of nozzle, m1 (ft2).
At = Aliquot of total sample added to still,
ml.
B*i = Water vapor In the gas stream, propor-
tion by volume.
C. = Concentration of fluoride In stack gas,
mg/m3. corrected to standard conditions
of 20° C. 760 mm Hg (68° F, 29.92 In. Hg)
on dry basis.
F» = Total weight of fluoride In sample, mg.
/ = Percent of Isoklnetic sampling.
M = Concentration of fluoride from calibra-
tion curve, molarity.
77in=Total amount of particulate matter
collected, mg.
M »> = Molecular weight of water, 18 g/g-mole
(18 Ib/lb-mole).
ma—Mass of residue of acetone after evap-
oration, mg.
Pb«r = Barometric pressure at the sampling
site, mm Hg (In. Hg).
Pi = Absolute stack gas pressure, mm Hg (In.
Hg).
P. td = Standard absolute pressure, 760 mm
Hg (29.92 in. Hg).
R = Ideal gas constant. 0.06236 mm Hg-mV
°K-g-mole (21.83 in. Hg-ftV°R-lb-mole).
Tm = Absolute average dry gas meter tem-
perature (see fig. 13A-3), °K (°R).
Tt = Absolute average stack gas temperature
(see fig. 13A-3), °K (°R).
T. m=standard absolute temperature, 293°
K (528° R).
Va=Volume of acetone blank, ml.
Va» = Volume of acetone used in wash, ml.
Vd = Volume of distillate collected, ml.
Vic = Total volume of liquid collected in 1m-
pingers and silica gel, ml. Volume of water
in silica gel equals silica gel weight In-
crease In grams times 1 ml/gram. Volume
of liquid collected in impinger equals final
volume minus initial volume.
Vm =Volume of gas sample as measured by
dry gas meter, dcm (dcf).
Vm(.u) = Volume of gas sample measured by
the dry gas meter corrected to standard
conditions, dscm (dscf).
V«did>=:Volume of water vapor In the gas
sample corrected to standard conditions,
scm (scf).
Vr = Total volume of sample, ml.
v, = Stack gas velocity, calculated by Method
2, Equation 2-7 using data obtained from
Method 5, m/sec (ft/sec).
Wa = Weight of residue In acetone wash, mg.
&H=Average pressure differential across the
orifice (see fig. 13A-3), meter, mm H-O
(In. H=O).
p0=Denslty of acetone, mg/ml (see label on
bottle).
pu, = Denslty of water, 1 g/ml (0.00220 lb/
ml).
0—Total sampling time, mln.
13.6 —Spe'cific gravity of mercury.
60 = Sec/mln.
100 —Conversion to percent.
9.2 Average dry gas meter temperature
and average orifice pressure drop. See data
sheet (Figure 13A-3 of Method 13A).
9.3 Dry gas volume. Use Section 9.3 of
Method 13A.
9.4 Volume of Water Vapor. Use Section
9.4 of Method 13A.
9.5 Moisture Content. Use Section 9.5 of
Method 13A.
9.6 Concentration
9.6.1 Calculate the amount of fluoride In
the sample according to equation 13B-1.
Vi
Fr~K-(V,i) (AT)
A,
where:
K = 19 mg/ml.
9.6.2 Concentration of fluoride In stack
gas. Use Section 9.6.2 of Method 13A.
9.7 Isoklnetic variation. Use Section 9.7
of Method 13A.
9.8 Acceptable results. Use Section 9.8 of
Method 13A.
10. References.
Bellack, Ervin, "Simplified Fluoride Distil-
lation Method," Journal of the American
Water Works Association #50: 530-6 (1958).
MacLeod, Kathryn E., and Howard L. Crist,
"Comparison of the SPADNS—Zirconium
Lake and Specific Ion Electrode Methods of
Fluoride Determination In Stack Emission
Samples," Analytical Chemistry 45: 1272-1273
(1973).
Martin, Robert M. "Construction Details of
Isoklnetic Source Sampling Equipment,"
Environmental Protection Agency, Air Pol-
lution Control Office Publication No. APTD-
0581.
1973 Annual Book of ASTM Standards, Part
23, Designation: D 1179-72.
Pom, Jerome J., "Maintenance, Calibration,
and Operation of Isoklnetic Source Sampling
Equipment," Environmental Protection
Agency, Air Pollution Control Office Publica-
tion No. APTD-0576.
Standard Methods for the Examination of
Water and Waste Water, published Jointly by
American Public Health Association, Ameri-
can Water Works Association and Water Pol-
lution Control Federation, 13th Edition
(1971).
387
-------
Reference Method 13B is amended
as follows:
(a) In the third line of section 3, the
phrase "300,ug/liter" is corrected to read
"300 mg/liter".
Section 5.1.5 is revised to read as
follows:
5.1.6 Filter holder-r&f located between tha
probe and first Implnger. boroslUcate glass
with a 20 mesh stainless steel screen filter
support and a slllcone rubber gasket; neither
a glass frit filter support nor a sintered metal
filter support may be used If the filter is in
front of the Implngers. If located between
the third and fourth Implngers, borosllicate
glass with a glass frit filter support and a
sillcone rubber gasket. Other materials of
construction may be used with approval from
the Administrator, e.g., if probe liner is stain-
less steel, then filter holder may be stainless
steel. The holder design shall provide a posi-
tive seal against leakage from the outside or
around the filter.
(c) Section 7.1.3 Is amended by revis-
ing the first two sentences of the sixth
paragraph to read as follows:
7.1.3 Preparation of collection train. • • •
Assemble the train as shown In Figure
13A-1 (Method 13A) with the filter between
the third and fourth Implngers. Alterna-
tively, the filter may be placed between the
probe the first Implnger if a 20 mesh stain-
less steel screen Is used for the filter sup-
port. • • •
• • • • •
(d) In section 7.3.4, the reference In
the first paragraph to "section 7.3.6" is
corrected to read "section 7.3.5".
388
-------
METHOD 14
DETERMINATION OF FLUORIDE EMISSIONS FROM POT ROOM ROOF MONITORS
OF PRIMARY ALUMINUM PLANTS
1. Principle and applicability.
1.1 Principle. Gaseous and partlculate
fluoride roof monitor emissions are drawn
.Into a permanent sampling manifold through
several large nozzles. The sample Is trans-
ported from the sampling manifold to ground
level through a duct. The gas In the duct Is
sampled using Method 13A or 13B—DETER-
MINATION OP TOTAL FLUORIDE EMIS-
SIONS PROM STATIONARY SOURCES. Ef-
fluent velocity and volumetric flow rate are
determined with anemometers permanently
located In the roof monitor.
1.2 Applicability. This method is applica-
ble for the determination of fluoride emis-
sions from stationary sources only when
specified by the test procedures for deter-
mining compliance with new source perform-
ance standards.
2. Apparatus.
2.1 Velocity measurement apparatus.
2.1.1 Anemometers. Vane or propeller
anemometers with a velocity measuring
threshold as low as 15 meters/minute and a
range up to at least 600 meters/minute. Each
anemometer shall generate an electrical sig-
nal which can be calibrated to the velocity
measured by the anemometer. Anemometers
shall be able to withstand dusty and corro-
sive atmospheres.
One anemometer shall be Installed for
every 85 meters of roof monitor length. If
the roof monitor length divided by 85 meters
is not a whole number, round the fraction
to the nearest whole number to determine
the number of anemometers needed. Use one
anemometer for any roof monitor less than
85 meters long. Permanently mount the
anemometers at the center of each equal
length along the roof monitor. One anemom-
eter shall be Installed In the same section
of the roof monitor that contains the sam-
pling manifold (see section 2.2.1). Make a
velocity traverse of the width of the roof
monitor where an anemometer is to be placed.
This traverse may be made with any suit-
able low velocity measuring device, and shall
be made during normal process operating
conditions. Install the anemometer at a point
of average velocity along this traverse.
2.1.2 Recorders. Recorders equipped with
signal transducers for converting the electri-
cal signal from each anemometer to a con-
tinuous recording of air flow velocity, or to
an Integrated measure of volumetric flow.
For the purpose of recording velocity, "con-
tinuous" shall mean one readout per 15-
mlnute or shorter time Interval. A constant
amount of time shall elapse between read-
ings. Volumetric flow rate may be determined
by an electrical count of anemometer revo-
lutions. The recorders or counters shall per-
mit Identification of the velocities or flow
rate measured by each Individual anemom-
eter.
2.2 Roof monitor air sampling system.
2.2.1 Sampling ductwork. The manifold
system and connecting duct shall be per-
manently installed to draw an air sample
from the roof monitor to ground level. A
SAV.PIE EXTRACTION
DUCT
ISeml.D.
ROOF MONITOR
txHAUST
STACK
1C DUCT Dlft'
MINIMUM
MINIMUM
SAMPLE PGHTSIN
VERTICAL DiJCT
SECTION AS SHOWN
Y
f'< PCT ROOM
'-1
EXHAUST BLOWER
Figuie 14 1. Hnof Vomtc' S
-------
'samnle-exposed surfaces within the nozzles.
ma: ifold and sample duct of 316 stainless
steel. Aluminum may be used if a new duct-
work system Is conditioned with fluoride-
laden roof monitor air for a period of six
\veeks prior to initial testing. Other materials
cf construction may be used if it is demon-
strated through comparative testing that
there Is no loss of flucrldes in the system. All
connections in the ductwork shall be leak
free.
Locate two sample ports in a vertical sec-
tion of the duct between the roof monitor
and exhaust fan. The sample ports shall be at
least 10 duct diameters downstream and
two diameters upstream from any flow dis-
turbance such as a bend or contraction. The
two sample ports shall be situated 90" apart.
One of the sample ports shall be situated so
that the duct can be traversed in the plane
of the nearest upstream duct bend.
2.2.2 Exhaust fan. An industrial fan or
blower to be attached t$ the sample duct
at ground level. (See Figure 14—1.) This ex-
haust fan shall have » maximum capacity
such that a large enough volume of air can
be pulled through the ductwork to main-
tain an isokinetlc sampling rate In all the
sample nozzles for all flow rates normally en-
countered in the roof monitor.
The exhaust fan volumetric flow rate shall
be adjustable so that the roof monitor air
can be drawn Isokinetically into the sample
nozzles. This control of flow may be achieved
by a damper on the inlet to the exhauster or
by any other workable method.
2.3 Temperature measurement apparatus.
2.3.1 Thermocouple. Installed In the roof
monitor near the sample duct.
2.3.2 Signal transducer. Transducer to
change the thermocouple voltage output to
a temperature readout.
2.3.3 Thermocouple wire. To reach from
roof monitor to signal transducer and
recorder.
2.3.4 Sampling train. Use the train de-
scribed in Methods 13A and 13B-—Determi-
nation of total fluoride emissions from sta-
tionary sources.
3. Reagents.
3.1 Sampling and analysis. Use reagents
described in Method ISA or 13B—Determi-
nation of total fluoride emissions from sta-
tionary sources.
4. Calibration.
4.1 Propeller anemometer. Calibrate the
anemometers so that their electrical signal
outnut corresponds to the velocity or volu-
metric flow they are measuring. Calibrate
according to manufacturer's Instructions.
4.2 Manifold intake nozzles. Adjust the ex-
haust fan to draw a volumetric flow rate
(refer to Equation 14-1) such that the en-
trance velocity into each manifold nozzle
approximates the average effluent velocity In
the roof monitor. Measure the velocity of the
air entering earn nozzle by inserting an S
type pltot tube into a 2.5 cm or less diameter
hole (sse Figure 14-2) located In the mani-
fold between each blast gate (or valve) and
nozrJs The pitot tube tip shall be extended
in.*o the center of the manifold. Take care
to insure that there is no leakage around the
pitot probe which could affect the Indicated
velocity in the manifold leg. If the velocity
of air being drawn Into each nozzle Is not
the same, open or close each blast gate (or
valve) until the velocity In each nozzle Is the
same. Fasten each blast gate (or valve) so
that It will remain In this position and close
the pltot port holes. This calibration shall be
performed whan the manifold system Is In-
stalled. (Note: It la recommended that this
calibration be repeated at least once a year.)
5. Procedure.
5.1 Roof monitor velocity determination.
5.1.1 Velocity value for setting isokinetic
flow. During the 24 hours preceding a test
run, determine the velocity Indicated by the
propeller anemometer In the section of root
monitor containing the sampling manifold.
Velocity readings shall be taken every 15
minutes or at shorter equal time Intervals.
Calculate the average velocity for the 24-hour
period.
5.1.2 Velocity determination during a test
run. During the actual test run, record the
velocity or volume readings of each propeller
anemometer in the roof monitor. Velocity
readings shall be taken for each anemometer
every 15 minutes or at shorter equal time
Intervals (or continuously).
5.2 Temperature recording. Record the
temperature of the roof monitor every two
hours during the test run.
5.3 Sampling.
5.3.1 Preliminary air flow in duct. During
the 24 hours preceding the test, turn ou the
exhaust fan and draw roof monitor air
through the manifold duct to condition the
ductwork. Adjust the fan to draw a volu-
metric flow through the duct such that the
velocity of gas entering the manifold nozzles
approximates the average velocity of the air
leaving the roof monitor.
5.3.2 Isokinetic sample rate adjustment.
Adjust the fan so that the volumetric flow
rate In the duct Is such that air enters Into
the manifold sample nozzles at a velocity
equal to the 24-hour average velocity deter-
mined under 5.1.1. Equation 14-1 gives the
correct stream velocity which Is needed In the
duct at the sample ports In order for sample
gas to be drawn Isokinetically Into the mani-
fold nozzles. Perform a pltot traverse of the
duct at the sample ports to determine If the
correct average velocity In the duct has been
achieved. Perform the pltot determination
according to Method 2. Make this determina-
tion before the start of a test run. The fan
setting need not be changed during the run.
8
(Vm)
1 minute
60 sec
Eq. 14-1
where:
V*=desired velocity In duct at sample
ports, meter/sec.
Dn=dlameter of a roof monitor manifold
nozzle, meters.
Da=dlameter of duct at sample port,
meters.
Vm=average velocity of the air stream In
the roof monitor, meters/minute, as
determined under section 5.1,1.
5.3.3 Sample train operation. Sample the
duct using the standard fluoride train and
methods described In Methods 13A and 13B—
Determination of total fluoride emissions
from stationary sources. Select sample trav-
erse points according to Method 1. If a se-
lected sampling point is less than one Inch
from the stack wall, adjust the location of
that point to one inch away from the wall.
5.3.4 Each test run shall last eight hours
or more. It a question exists concerning the
representativeness of an eight-hour test, a
longer test period up to 24 hours may be se-
lected. Conduct each run during a period
when all normal operations are performed
underneath the sampling manifold, i.e. tap-
ping, anode changes, maintenance, and either
normal duties. All pots In the potroom shall
be operated In a normal manner during the
test period.
5.3.5 Sample recovery. Same as Method
13A or 13B—Determination of total fluoride
emissions from stationary sources.
5.4 Analysis. Same as Method 13A or 13B—
Determination of total fluoride emissions
from stationary sources.
390
-------
6. Calculations.
6.1 Isokinetic sampling test. Calculate the
mean velocity measured during each sam-
pling run by the anemometer In the section
of the roof monitor containing the sampling
manifold. If the mean velocity recorded dur-
ing a particular test run does not fall within
±20 percent of the mean velocity established
according to 5.3.2. repeat the run.
6.2 Average velocity of roof monitor gases
Calculate the average roof monitor velocltj
using all the velocity or volumetric flow read-
Ings from section 5.1.2.
6.3 Roof monitor temperature. Calculate
the mean value of the temperatures recorded
in section 5.2.
6.4 Concentration of fluorides in roof moni-
tor air in mg F/m*. This Is given by Equation
13A-5 in Method 13A—Determination of
total fluoride emissions from stationary
sources.
6.5 Average volumetric flow from roof Is
given by Equation 14-2.
™, (A) (Md) Pm (294°K)
.-. Eq.14-2
* (Tm + 273°) (760mmHg)
where:
Q«=average volumetric flow from roof
monitor at standard conditions on
a dry basis, mVmln.
ji=roo£ monitor open area, m*. •
Vm i = average velocity of air in the rool
monitor, meters/minute from sec-
tion 6.2.
Pm=atmospherlc pressure, mm Hg.
Tm=roof monitor temperature, "C, from
section 6.3.
Mj = mole fraction of dry gas, which is
^ „„ 100—100 (B».)
given by Md—
Bn«=ls the proportion by volume of water
vapor In the gas stream, from
Equation 13A-3. Method 13A—De-
termination of total fluoride emis-
sions from stationary sources.
[36 FB 24877, Dec. 23, 1971, as amended at
38 FB 13562, May 23, 1973; 39 FB 9319,
Mar. 8, 1974; 39 FB 13776, Apr. 17, 1974; 39
FB 20794, June 14, 1974; 39 FB 39874. Nov.
12, 1974; 40 FB 33157, Aug. 6, 1975; 41 FB
3828. Jan. 26, 1976]
391
-------
METHOD 15
DETERMINATION OF H2S, COS, and CS2 EMISSIONS FROM
STATIONARY SOURCES
The
INTRODUCTION
method describc-d below uses the
principle of pas chromatographic separation
and flame photometric detection (FPD).
Sinoe there are many rvstems cr sets ct op-
erating conditions that represent usable
methods ol determining sulfur emissions, all
systems which employ this principle, but
differ only in details of touipment and oper-
ation. may be used a.< alternative methods,
provided that the criteria set below are met.
1. Principle and applicability
1.1 Principle. A gas sample is extracted
from the emission source and diluted \vilh
c'.ean dry p.ir. An aliquot of the diluted
sample- is then analyzed for hydrogen xul-
fide <.H,S). carbony'. suHide (COS), and
carbon disulfide (CS,) by gas chromatogra-
phic (GCj separation and flame photomet-
ric detection !FTD).
1.2 Applicability. This method is applica-
ble for determination of the above sulfur
compounds from nil gas control units of
sulfu'- recovery plants.
2. Range and sensitivity
2.1 Range. Coupled v. i'h a gas chrcmto-
graphic sy.-tem utilising a 1-milliliter sample
size, the maximum limit of the FPD lor
ench scifur compound is approximately 10
pprn. It may be necessary to dilute gas sam-
ples from .sulfur recov"ry plams hundred-
fold (99:1) rosu'tiiiis in an upper limit of
abu'.it 1000 ppm frr each compound.
2.'1 The n'inimum d'.'li-ctabie concentra-
tion of the FPD is :,l,--o de::>«ridf;nt on sample
size and would be aoout 0.5 ppm for a 1 ml
sample.
3. Interferences
3.1 Moisture Condensation. Moisture con-
densation in .'he .--ami1'1* drliverv system, the
analylu-aJ foH.'inn. or the FPD burner block
ci>.'.\ c:\ii.if t,».-.;\ 5 or ni.er'erf.ices. This p<>-
ler.i-.i1 is •'l.niir.-.iN --; In lieatmu the sample
line, and b> a lull" .-..,1 . the sample wim
dry dilution ::ir lo If ....... it.-. dfw point b* low
the ui'iTAtin.'. U';i|..'i,iin.v o: the GC/1'T'O
anal> iti'ul s.\steiH pne.r to nnalvsis.
3.2 C.ir'oon M.nin.Mur rd Carbon Dioxide.
CO and CO; have .sub-.Ianlial desen.-.iti.:inif
of feels on (he flrime photometric detector
even after D:l il:!>.ition. (Acceptable systems
must demonstrate that they have eliminat-
ed this iniei lerence bv sonir procedure such
as eludir.K CO and CO. beiore any of the
sulfur compuundi to be measured.) Compli-
ance with tl'i:-. iciiuirpmont can be demon-
strated by submitting c'iro"i.-itoerams of
c-iilibr.i'ion mises with and \utliout CO, in
the ciiliitnt Pii-s. The CO-, irvel should be ap-
proximately 10 percent fc.- the case with
CO, prevonr. The two c!iromntoi:raphs
should show agreement within the precision
limits of section 4.1.
3.3 Elemental Sulfur. The condensation of
sulfur vapor in the sampling line can lead to
eventual coating and even blockage of the
sample line. This problem can be eliminated
along with the moisture problem by heating
the sample line.
4. Precision
4.1 Calibration Precision. A series of three
consecutive injections of the same calibra-
tion gas. at any dilution, shall produce re-
sults \\l-.icb do not vary by more than ±13
percent from the mean of the three injec-
tions.
4.2 Calibration Drift. The calibration drift
determined from the mean of three injec-
tions made at the beginning and end of any
8-hour period shall not exceed ±5 percent.
S. Apparatus
5.1.1 Probe. The probe must be made of
inert material such as stainless steel 01
glass. It should be designed to incorporate a
litter and to allow calibration gas to enter
the probe at or near the sample entry point.
Any portion of the probe not exposed to the
stack gas must be heated to prevent mois-
ture condensation.
5.1.2 The sample line must be made of
Teflon,'no greater inan 1.3 cm C/4 in) inside
diameter. All parts from the probe to the di-
lution system must be thermostatically
hewtfcd to 120° C.
5.1.3 Sample Pump. The sample pump
shall be a leakless Teflon coated diaphragm
Lype or equivalent. If the purnp is upstream
of the dilution system, the pump head must
be heated to 120° C.
5.2 Dilution System. The dilution system
must be constructed such that all sample
contacts are made of ineit material (e.g.
stainless steel or Teflon). It must be heated
to 120° C and be capable of approximately a
9:1 dilution of the sample.
5.3 Gas Chromatograph. The gas chroma-
tograph must have at least the following
components:.
5.3.1 O\vn. Capable of maintaining the
separation column at the proper operating
temperature ±r C.
5.3.2 Temperature Gauge. To monitor
column oven, detector, and exhaust tem-
perature ±r c.
5.3.a Flow System. Gas metering system to
measure sample, fuel, combustion gas, and
carrier gas flow.-,.
'Miri'ion of trade names or specific prod-
ucts c-'es not
-------
5.3.4 Fh'ine Photometric Detector.
E.3.4.1 Electrometer. Capable of full scale
amplification of linear ranges of 10"" to 10"'
amperrs full scale.
5.3.4.2 Power Supply. Capable of deliver-
ing up to 750 volts.
5.3.-J.3 Recorder. Compatible with the
output voltage range of the electrometer.
5.4 Gas ChromatoRraph Columns. The
column system must be demonstrated to be
capable of resolving three major reduced
sulfur compounds: H,S. COo, and CS,.
To demonstrate that adequate resolution
lias been achieved the tester must submit a
chromatogniph of a calibration gas contain-
ing all three reduced sulfur compounds in
the concentration range of the applicable
standard. Adequate resolution will be de-
fined as base line separation of adjacent
peaks when the amplifier attenuation is set
so that the smaller peak is at least 50 per-
cent of full scale. Base line separation is de-
fined as a return to zero ±5 percent in the
Interval between peaks. Systems not meet-
ing this criteria may be considered alternate
methods subject to the approval of the Ad-
ministrator.
5.5.1 Calibration System. The calibration
system must contain the following compo-
nents.
5.5.2 Flow System. To measure air flow
over permeation tubes at ±2 percent. Each
flowmeter shall be calibrated after a com-.
plete test series with a \vet test meter. If the
flow measuring device differs from the wet
test meter by 5 percent, the completed test
shall be discarded. Alternatively, the tester
may elect to use the flow data that would
yield the lowest flow measurement. Calibra-
tion with a wet test meter before a test is
optional.
5.5.3 Constant Temperature Bath. Device
capable of maintaining the permeation
tubes at the calibration temperature within
±1.1'C.
5.5.4 Temperature Gauge. Thermometer
or equivalent to monitor bath temperature
within ±1'C.
6. Reagents
6.1 Fuel. Hydrogen (H,) prepurified grade '
or better.
6.2 Combustion Gas. Oxygen (Oa) or air,
research purity or better.
6.3 Carrier Gas. Prepurified grade or
better.
6.4 Diluent. Air containing less than 0.5
ppm total sulfur compounds and less than
10 ppm each of moisture and total hydro-
carbons.
6.5 Calibration Gases. Permeation tubes,
one each of H2S, COS, and CS,, gravimetri-
caily calibrated and certified at some conve-
nient operating temperature. These tubes
consist of hermetically sealed FEP Teflon
tubing in which a liquified gaseous sub-
stance is enclosed. The enclosed gas perme-
ates through the tubing wall at a constant
rate. When the temperature is constant,
calibration Eases covering a wide range of
known concentrations can be generated by
varying and accurately measuring the flow
rale of diluent gas passing over the tubes.
These calibration p.ases are used to calibrate
the GC/FPD system and the dilution
system.
7. Pretest Procedures
The following procedures are optional but
would be helpful in preventing any problem
which might occur later and invalidate the
entire te.st.
7.1 After the complete me?surement
svstem has br-'u set up at the si'o and
deemeJ to be operational, the follow ing pro-
cedures shoulit be computed before sam-
pling is initiated.
7.1.1 Leak Test. Appropriate leak test pro-
cedures should be cviplojecl to \erify the in-
tegrity of all components, sample lines, niH
connections. The following U-ak test proce-
dure Is suggested: For components upstream
of the sample pump, attach the probe end
of the sample line to a manometer or
vacuum gauge, start the pump and pull
greater than 50 mm (2 in.) Hi? vacuum, clcse
off the pump outlet, and then stop the
pump and ascertain that there is no leak for
1 minute. For components after the pump,
apply a slight positive pressure and check
for leaks by applying a liquid (detergent in
v.'ater. for example) at each joint. Bubbling
indicates the presence cf a leak.
7.1.2 System Performance. Since the com-
plete system is calibrated following each
test, the precise calibration of each compo-
nent is not critical. However, these compo-
nents should be verified to be operating
properly. This verification can be performed
by observing the response of flowtneters or
of the GC output to changes in flow rates or
calibration gas concentrations and ascer-
taining the response to be within predicted
limits. If any component or the complete
system fails to respond in a normal and pre-
dictable manner, the source of the discrep-
ancy should be identifed and corrected
before proceeding.
8. Calibration
Prior to any sampling run, calibrate the
system using the following procedures. (If
more than one run is performed during any
24-hour period, a calibration need not be
performed prior to the second and any sub-
sequent runs. The calibration must, howev-
er, be verified as prescribed in section 10,
after the last run made within the 24-hour
period.)
8.1 General Considerations. This section
outlines steps to be followed for use of the
GC/FPD and the dilution system. The pro-
cedure does not include detailed instruc-
tions because the operation of these systems
is complex, and it requires an understanding
of the individual system being used. Each
system should include a written operating
manual describing in detail the operating
procedures associated with each component
in the measurement system. In addition, the
operator shuld be familiar with the operat-
ing principles of the components; particular-
ly the GC/FPD. The citations in the Bib-
liography at the end of this method are rec-
ommended for review for this purpose.
3.2 Calibration Procedure. Insert the per-
meation tubes into the tube chamber. Check
the bath temperature to assure agreement
with the calibration temperature of the
tubes within ±U.1°C. Allow 24 hours for the
tubes to equilibrate. Alternatively equilibra-
tion may be verified by injecting samples of
calibration gas at 1-hour intervals. The per-
meation tubes can be assumed to have
reached equilibiium when consecutive
hourly samples agree within the precision
limits of section 4.1.
Vary the amount of air flowing over the
tubes to produce the desired concentrations
for calibrating the analytical and dilution
systems. The air flow act ass the tubes must
at all timet exceed the flow requirement of
the analytical sy.s'cms. The concentration in
parts per million generated by a bubo con-
393
-------
taming a specific permeant can be calculat-
ed as foKov.'s:
Equation 15-1
where:
C--Cor.c< ntration of permeant produced
in p^:n.
P,^ ITI mention rate of the tube in fig/
mln.
M= Molecular weight of the permeant: g/
K-mole.
L=Flo\v rate. 1/mln. of air over permeant
@ 20'C. 760 mm Hg.
K=Gas constant at 20°C and 760 mm
Hg =24 04 I/ E mole.
8.3 Calibration of analysis system. Gener-
ate a series of three or more known concen-
trations spanning the linear range of the
FPD (approximately 0.05 to 1.0 ppm) for
each of the four major sulfur compounds.
Bypassing the dilution system, inject these
standards in to the GC/FPD analyzers and
monitor the responses. Three injects for
each concentration must yield the precision
described in section 4.1. Failure to attain
this precision is an indication of a problem
in the calibration or analytical system. Any
such problem most be identified and cor-
rected before proceeding.
8.4 Calibration Curves. Plot the GC/PPD
response in current (amperes) versus their
causative concentrations in ppm on log-log
coordinate graph paper lor each sulfur com-
pound. Alternatively, a least squares equa-
tion may be generated from the calibration
data.
8.5 Calibration of Dilution System. Gener-
ate a know concentration of hydrogen sui-
fied using the permeation tube system.
Adjust the flow rate of diluent air for the
first dilution stage so that the desired level
of dilution is approximated. Inject the dilut-
ed calibration gas into the GC/FPD system
and monitor its response. Three injections
for each dilution must yield the precision
described in section 4.1. Failure to attain
this precision in this step is an indication 01
a problem in the dilution system. Any such
problem must be identified and corrected
before proceeding. Using the calibration
data for H,S (developed under 8.3) deter-
mine the diluted calibration gas concentra-
tion in ppm. Then calculate the dilution
factor as the ratio of the calibration ga^
concentration before dilution to the diluted
calibration gas concentration determined
under this paragraph. Repeat this proce-
dure for each stage of dilution required. Al-
ternatively, the GC/FPD system may te
calibrated by generating a series of three or
more concentrations of each sulfur com-
pound and diluting these samples before in-
jecting them into the GC/FPD system. Th^s
data wUl then serve as the calibration data.
for the unknown samples and a s-jparate de-
termination of the dilution factor will no:
be necessary. However, the precision re-
quirements of section 4.1 are still applicable.
S. Sampling and Analysis Procedure
9 1 Sampling. Insert the sampling probe
into the test port making certain that no di-
lution air enters the stack through the port.
Begin sampling and dilute the sample ap-
proximately 9:i using the dilution system.
Note that Ihe precise dilution factor is thai
which Is determined in paragraph 8.5. Con-
dition the entire system with sample for a.
minimum of 15 minutes prior to commenc-
ing analjsis.
9.2 Analysis. Aliquots of diluted sample
are injected into the GC/FPD analyzer for
analysis.
9.2.1 Sample Run. A sample run is com-
posed of 1G individual analyses (injects) per-
formed over n period of not less than 3
hours or more than 6 hours.
9.2.2 Observation for Clogging of Probe. If
reductions in sample concentrations are ob-
served during a sample run that cannot be
explained by process conditions, the sam-
pling must be interrupted to di-lermine i!
the sample probe is cloBced with pnr'.lmlale
matter. If the probe is loimd to be clorE'-d.
the test must be stopped and thr ri-uiiu up
to that point discarded. Tcstlnu may rrsunie
after cleaning the prube or replacing It with
a clean one. After each run. the sample
probe must be inspected and. if necessary.
dismantled and cleaned.
JO. Post-Test Procedures
10.1 Sample Line Loss. A known concen-
tration of hydrogen sulflde at the level of
the applicable standard. 3:20 percent, must
be introduced into the sampling system at
the opening of the probe In sufficient, quan-
tities to ensure that there is an excess of
sample which must be vented to thf atmo-
sphere. The sample must be transported
through the entire sampling system to the
measurement system in the normal manner.
The resulting measured concentration
should be compared to the known value to
determine the sampling system loss. A sam-
pling system loss of more than 20 percent Is
unacceptable. Sampling losses of 0-20 per-
cent must be corrected by dividing the re-
sulting sample concentration by the frac-
tion of recovery. The known gas sample may
be generated using permeation tubes. Alter-
natively, cylinders of hydrogen sulfide
mixed in air may be used provided they are
traceable to permeition tubes. The optional
pretest procedures provide a good guideline
for determining if there are leaks in the
sampling system.
10.2 Recalibration. After each run. or
after a series of runs made within a 24-hour
period, perform a partial recalibration using
the procedures -in section 8. Only HaS (or
other permeant) need be used to recalibrate
the GC/FPD analysis system (8.3) and the
dilution system (8.5).
10.3 Determination of Calibration Drift.
Compare the calibration curves obtained
prior to the runs, to the calibration curves
obtained under paragraph 10.1. The calibra-
tion drift should not exceed the limits set
forth in paragraph 4.2. If the drift exceeds
this limit, the intervening ri n or runs
should be considered not valid. The tester.
however, may instead have the option of
choosing the calibration data set which
would give the highest sample values.
11. Calculations
11.1 Determine the concentrations of each
reduced sulfur compound detected directly
from the calibration curves. Alternatively.
the concentrations may be calculated using
the equation for the least squares line.
11.2 Calculation of SO-. Equivalent. SO,
equivalent will be determined for each anal-
ysis made by summing the concentrations of
each reduced sulfur compound resolved
during the given analysis.
SO, equivalent = £(H2S. CCS. 2 CSa)d
Equation 15-2
394
-------
where:
SOi equivalent -The sum of the concen-
tration of each of the measured com-
pounds (COS, H,S. CS,) expressed as
sulfur dioxide in ppm.
HaS - Hydnvc-n sulfidr, ppm.
COS^Carbonyl sulfiile. ppm.
CS> -Carbon cUsulfidr. ppm.
d --Dilution favlor. dinviisiontess.
11.3 Average Sd equivalent will be deter-
mined as follows:
N
I S0 equiv.
u SO, equivalent » 1 • 1
_
H (1 - Bwol
Equation 15-3
where:
Average SO, equivalent. = Average SO,
equivalent in ppm, dry basis.
Average SO, equivalen:, = SO, in ppm as
determined by liquation 15-2.
N = Number of analyses performed.
Bwo = Fraction of volume of water vapor
In the gas stream as determined by
Method 4 — Determination of Moisture
in Stack Gases (36 PR 24887).
12. Example System
Described below is a system utilized by
EPA in gathering NSPS data. This system
does not now reflect all the latest develop-
ments in equipment and column technology,
but it does represent one system that has
been demonstrated to work.
li.l Apparatus.
12.1.1 Sample System.
12.1.1.1 Probe. Stainless steel tubing, 6.35
mm (*/t in.) outside diameter, packed with
glass wool.
12.1.1.2 Sample Line. Vis inch inside diam-
eter Teflon tubing heated to 120°C. This
temperature is controlled by a thermostatic
heater.
12.1.1.3 Sample Pump. Leakless Teflon
coated diaphragm type or equivalent. The
pump head is heated to 120' C by enclosing
it in the sample dilution box (12.2.4 below).
12.1.2 Dilution System. A schematic dia-
gram of the dynamic dilution system is
given in Figure 15-2. The dilution system is
constructed such that all sample contacts
are made of inert materials. The dilution
system which is heated to 120' C must be ca-
pable of a minimum of 9:1 dilution of
sample. Equipment used in the dilution
system is listed below:
12.1.2.1 Dilution Pump. Model A-150 Koh-
myhr Teflon positive displacement type.
nonadjustable 150 cc/min. ±2.0 percent, or
equivalent, per dilution stage. A 9:1 dilution
of sampie is accomplished by combining 150
cc of sample with 1350 cc of clean dry air as
shown in Pifrure 15-2.
12.1.2.2 Valves. Three-way Teflon solenoid
or manual type.
12.1.2.3 Tubing. Teflon tubing and fittings
are used throughout from the sample probe
to the GC/FPD to present an inert surface
for sample gas.
12.1.2.4 Box. Insulated box. heated and
maintained at 120'C. of sufficient dimen-
sions to house dilution apparatus.
12.1.2.5 Flov.-meters. Rotamc-ters or equiv-
alent to measure flow Irom 0 to 1500 ml/
min. ± I percent per dilution stage.
12.1.3.0 Cias Ch: omatoirraph.
12.1.3.1 Column— 1.83 m (6 ft.) length of
Teilon tubing, 2.16 mm (O.OC5 in.) inside di-
ameter. packed with deactivated silica gel.
or equivalent.
12.1.3.2 Sample Valve. Teflon six port gas
sampling v.Uvc. equipped with a 1 ml sample
loop, actuated by compressed air (figure 15-
1).
12.1.3.3 Oven. For containing sample
valve, stripper column anci separation
column. The oven should bf capable of
maintaining an elevated tempi-rature rang-
ing from ambient to ICO C, co;u,lant within
±1'C.
12.1.3.4 Temperature Monitor. Thermo-
couple pyrometer to measure column oven.
detector, and exhaust temperature » f C.
12.1.3.5 Flow System. Gas tnetcrliiR
system to measure sample f!ow, hydrogen
flow, oxygen flow and nitrogen carrier EOS
flow.
12.1.3.6 Detector. Flame photometric de-
tector.
12.1.3.7 Electrometer. Capable of full scale
amplification of linear ranges of 10"'to 10"'
amperes full scale.
12.1.3.8 Power Supply. Capable of deliver-
ing up to 750 volts.
12.1.3.9 Recorder Compatible with the
output voltage ran^e of the electrometer.
12.1.4 Calibration. Permeation tube
system (Figure 15-3).
12.1.4.1 Tube Chamber. Glass chamber of
sufficient dimensions to house perme-ation
tubes.
12.1.4.2 Mass Flowmeters. Two mass flow-
mrters in the range 0-3 1/mln. and 0-10 I/
min. to measure air flow over permeation
tubes at ±2 percent. These flowmeters shall
be cross-calibrated at the beginning of each
test. Using a convenient flow rate in the
measuring range of both flowmeters, set
and monitor the flow rate of gas over the
permeation tubes. Injection of calibrat.ion
gas generated at this flow rate as measured
by one flowrneter followed by injection of
calibration gas at the same flow rate ns mea-
sured by the other flowrneter should apree
within the specified precision limits. If ihey
do not, then there is a problem with the
mass flow measurement. Each mass flow-
meter shall be calibrated prior to ihe first
test with a wet test meter and thereafter at
least once each year.
12.1.4.3 Constant Temperature Bath. Ca-
pable of maintaining permeation •tubes at
certification temperature of 30'C within
±0.1° C.
12.2 Reagents.
12.2.1 Fuel. Hydrogen (H,) prepurified
grade or better.
12.2.2 Combustion Gas. Oxygen (O.) re-
search purity or better.
12.2.3 Carrier Gas. Nitrogen (N,) prepuri-
fied grade or better.
12.2.4 Diluent. Air containing less than 0.5
ppm total sulfur compounds and less than
10 ppm each of moisture and total hydro-
carbons, and filtered using MSA tillers
46727 and 79030. or equivalent. Removal of
sulfur compounds can be verified by inject-
ing dilution air only, described in section
8.3.
12.2.5 Compressed Air. 60 psig for GC
valve actuation.
12.2.6 Calibration Gases. Permeation
tubes gravimetrically calibrated and certi-
fied at 30.0' C.
12.3 Operating Parameters. The operating
parameters for the GC/FPD system are as
follows: nitrogen carrier eas flow rate of 100
cc/min. exhaust temperature of 110* C. de-
tector temperature 105' C. oven tempera-
ture of 40' C, hydrogen flow rate of 80 cc/
minute, oxygen flow rate of 20 cc/'mimite.
and sample flow rate of 80 cc/minutc.
395
-------
12.4 Analysis. The sample valve is -Actu-
ated for 1 ir.inute in which time an aliquot
of diluted sample is injected onto the .sepa-
ration column. The valve is then deactivated
for the remainder of analysis cycle in v hich
time the sample loop is refilled and the sep-
aration column continues to be foreflusl.ed.
The t-lut,on time for each compound will be
determined during calibration.
13. Bibliography
13.1 O'Kceffe. A. E. and G. C. Oilman.
"Primary Standards for Trace Gai Analy-
sis." Anal. Chcin. Jfl.760 (1966).
13.2 Stevens. H. K.. A. E. O'KeofJc. and
O. C. Ortmnn. "Absolute Calibration of a
Flame Photometric Detector to Voiatiie
Sulfur CompounrJs at Sub-Part-Per-M:llion
Leveis." Environmental Science and Tech-
nology 3:7 (July. 19SB).
13.3Mulick. J. D.. R. K. Stevens, and R.
Baamgardner. "An Analytical System De-
signed to Measure MultiDle Malodorous
Compounds Related to Kr*tt Mill Activi-
ties." Presented at -he 12th Conference on
Methods in Air Pollution and Industrial Hy-
giene Studies. University of Southern Cali-
fornia. Los Angeles. Calif April 8-8. 1971.
13.4 Devonald. R. H.. R. S. Serenlus. and
A. D. Mclntyre. "Evi'lur.lion of the flame
Photometric Detector for Analysis of Sulfur
Compounds." Pulp and Paper Magazine of
Canada. 73.3 (March, 1972).
13.5 Grlmley. K. W.. W. S. Smith, and
R. M. Martin. "The Use of a Dynamic Dilu-
tion System in the Conditioning of Stack
Gases for Automated Analysis by a Mobile
Sampling Van." Presented at the 63rd
Annual APCA Meeting in St. lrf>uis. Mo.
June 14-19. 1970.
13.6 General Reference. Standard Meth-
ods of Chemical Analysis Volume III A and
B Instrumental Methods. Sixth Edition.
Van Ncstrand Reinhold Co.
396
-------
METHOD 16
SEMICONTINUOUS DETERMINATION OF SULFUR EMISSIONS FROM
STATIONARY SOURCES
Introduction
The method described below uses the
principle of gas chromatographic separation
and flame photometric detection. Since
there are many systems or sets of operating
conditions that represent usable methods of
determining sulfur emissions, all systems
which employ this principle, but differ only
In details of equipment and operation, may
be used as alternative methods, provided
that the criteria set below are met.
1. Principle and Applicability.
1.1 Principle. A gas sample is extracted
from the emission source and diluted with
clean dry air. An aliquot of the diluted
sample is then analyzed for hydrogen sul-
fide (H,S), methyl mercaptan (MeSH). di-
methyl sulfide (DMS) and dimethyl disul-
fide (DMDS) by gas chromatographic (GO
separation and flame photometric detection
(FPD). These four compounds are known
collectively as total reduced sulfur fTRS).
1.2 Applicability. This metliod is applica-
ble for determination of TRS compounds
from recovery furnaces, lime kilns, and
smelt dissolving tanks at kraft pulp mills.
2. Range and Sensitivity.
2.1 Range. Coupled with a gas chromato-
graphic system utilizing a ten milliliter
sample size, the maximum limit of the FPD
for each sulfur compound is approximately
1 ppm. This limit is expanded by dilution of
the sample gas before analysis. Kraft mill
gas samples are normally diluted tenfold
(9:1). resulting in an upper limit of about 10
ppm for each compound.
For sources with emirsion levels between
10 and 100 ppm. the measuring range can be
best extended by reducing the sample size
to 1 milliliter.
2.2 Using the sample size, the mlnin-.um
detectable concentration is approximately
50 ppb.
3. interferences.
3.1 Moisture Condensation. Moisture
condensation in the sample delivery system,
the analytical column, or the FPD buiner
block can cause losses or interferences. This
potential is eliminated by heating the
sample line,'and by conditioning the sample
with dry dilution air to lower its dew point
below the operating temperature of the
GC/FPD analytical system prior to analysis.
3.2 Carbon Monoxide and Carbon Diox-
ide. CO and CO, have substantial desensitiz-
ing effect on the name photometric detec-
tor even after 9:1 dilution. Acceptable sys-
tems must demonstrate thai they have
eliminated this Interference by some proce-
dure such as cluting these compounds
before any of the compounds to be mea-
sured. Compliance with this requirement
can be demonstrated by submitting chroma-
tograms of calibration gases with and with-
out COt In the diluent gas. The CO, level
should be approximately 10 percent for the
case with CO, present. The two chromato-
graphs should show agreement within the
precision limits of Section 4.1.
3.3 Participate Matter. Participate
matter in gas samples can cause interfer-
ence by eventual clogging of the analytical
system. This interference must be eliminat-
ed by use of a probe filler.
3.4 Sulfur Dioxide. SO, is not a specific
Interferent but may be present in such large
amounts that it cannot be effectively sepa-
rated from other compounds of Interest.
The procedure must be designed to elimi-
nate this problem either by the choice of
separation columns or by removal of SO,
from the sample.
Compliance with this section can be dem-
onstrated by submitting chromatographs of
calibration gases with SO, present in the
same quantities expected from the emission
source to be tested. Acceptable systems
shall show baseline separation with the am-
plifier attenuation set so that the reduced
sulfur compound cf concern is at least 50
percent of full scale. Base line separation is
defined as a return to zero ± percent in the
interval between peaks.
4. Precision and Accuracy.
4.1 GC/PPD and Dilution System Cali-
bration Precision. A series of three consecu-
tive injections of the same calibration gas.
at any dilution, shall produce results which
do not vary by more than ±3 percent from
the mean of the three injections.
4.2 GC/PPD and Dilution System Cali-
bration Drift. The calibration drift deter-
mined from the mean of three injections
made at the beginning and end of any 8-
hour period shall not exceed ± percent.
4.3 System Calibration Accuracy. The
complete system musi, quantitatively trans-
port and analyze with an accuracy of 20 per-
cent. A correction factor is developed to
adjust calibration accuracy to 100 percent.
5. Apparatus (See Figure 16-1).
5.1.1 Probe. The probe must be made of
inert material such as stainless steel or
glass. It should be designed to incorporate a
filter and to allow calibration gas to enter
the probe at or near the sample entry point.
Any portion of the probe not exposed to the
stack gas must be heated to prevent mois-
ture condensation.
5.1.2 Sample Line. The sample line must
be made of Teflon.1 no greater than 1.3 cm
-------
5.2 Dilution System. The dilution system
must he constructed such that all sample
contacts are made of inert materials (e.g..
stainless steel or Teflon). It must be heated
to 120' C. and be capable of approximately a
9:1 dilution of the sample.
5.3 Gas Chromatograph. The gas chro-
matograph must have at least the following
components:
5.3.1 Oven. Capable of maintaining the
separation column at the proper operating
temperature ±1' C.
5.3.2 Temperature Gauge. To monitor
column oven, detector, and exhaust tem-
perature ±r C.
5.3.3 Flow System. Gas metering system
to measure sample, fuel, combustion cas.
and carrier gas flows.
5.3.4 Flame Photometric Detector.
5.3.4.1 Electrometer. Capable of full scale
amplification of linear ranges of !<)-• to 10-'
amperes full scale.
5.3.4.2 Power Supply. Capable of deliver-
ing up to 750 volts.
5.3.4.3 Recorder. Compatible with the
output voltage ranee of the electrometer.
5.4 Gas Chromatograph Columns. The
column system must be demonstrated to be
capble of resolving the four major reduced
sulfur compounds: H.S. MeSH. DMS. and
DMDS. It must also demonstrate freedom
from known interferences.
To demonstrate that adequate resolution
has been achieved, the tester must submit a
chromatograph of a calibration gas contain-
ing all four of the TRS compounds in the
concentration range of the applicable stan-
dard. Adequate resolution will be defined as
base line separation of adjacent peaks when
the amplifier attenuation is set so that the
smaller peak is at least 50 percent of full
scale. Base line separation is defined in Sec-
tion 3.4. Systems not meeting this criteria
may be considered alternate methods sub-
ject to the approval of the Administrator.
5.5. Calibration System. The calibration
system must contain the following compo-
nents.
5.5.1 Tube Chamber. Chamber of glass or
Teflon of sufficient dimensions to house
permeation tubes.
5.5.2 Flow System. To measure air flow
over permeation tubes at ±2 percent. Each
flowmeter shall be calibrated after a com-
plete test series with a wet test meter. If the
flow measuring device differs from the wet
test meter by 5 percent, the completed test
shall be discarded. Alternatively, the tester
may elect to use the flow data that would
yield the lowest flow measurement. Calibra-
tion with a wet test meter before a test is
optional.
5.5.3 Constant Temperature Bath, Device
capable of maintaining the permeation
tubes at the calibration temperature within
±0.1' C.
5.5.4 Temperature Gauge. Thermometer
or equivalent to monitor bath temperature
within ±1- C.
6. Reagents.
6.1 Fuel. Hydrogen (H,) prepurlfled
grade or better.
6.2 Combustion Gas. Oxygen (O.) or air,
research purity or better.
6.3 Carrier Gas. Prepurlfied grade or
better.
6.4 Diluent. Air containing less than 50
ppb total sulfur compounds and less than 10
ppm each of moisture and total hydrocar-
bons. This gas must be healed prior to
mixing with the sample to avoid water con-
densation at the point of contact.
6.5 Calibration Gases. Permeation tubes.
one each of H,S. MeSH, DMS. and DMDS.
awavlmetrically calibrated and certified at
some convenient operating temperature.
These tubes consist of hermetically sealed
FEP Teflon tubing in which a liquified gas-
eous substance is enclosed. The enclosed gas
permeates through the tubing wall at a con-
stant rate. When the temperature lr con-
stant, calibration pases Governing a wide
range of known concentrations can be gen-
erated by varying and accurately measuring
the flow rate of diluent gas passing over the
tubes. These calibration gases are used to
calibrate the GC/FPD system and the dilu-
tion system.
7. Pretext Procedures. The following proce-
dures are optional but would be helpful In
preventing any problem which might occui
later and Invalidate the entire test.
7.1 After the complete measurement
.system has been set up at the site and
di-emcd to be operational, the following- pro-
cr durcs should be completed before sam-
pling is initiated.
7.1.1 Leak Test. Appropriate leak test
procedures should be employed to verify the
integrity of all components, sample lines,
and connections. The following leak test
procedure is suggested: For components up-
stream of the sample pump, attach the
probe end of the sample line to a ma- no-
meier or vacuum gauge, start the pump and
pull greater than 50 mm (2 in.) Hg \acuum,
close off the pump outlet, and then stop the
pump and ascertain that there is no leak for
1 minute. For components af.er tlie pump.
apply a slight positive pressure and check
lor leaks by applying a liquid (detergent in
water, for example) at each joint. Bubbling
indicates the presence of a leak.
7.1.2 System Performance. Since the
complete system is calibrated following each
test, the precise calibration of eacii compo-
nent is not critical. However, these compo-
nents should bo verified to be operating
properly. This verification can be performed
by observing the response of flowmeters or
of the GC output to changes in flow rates or
calibration gas concentrations and ascer-
taining the response to be within predicted
limits. In any component, or if the complete
system fails to respond in a normal and pre-
dictable manner, the source of the discrep-
ancy should be identified and corrected
before proceeding.
8. Calibration. Prior to any sampling run,
calibrate the system using the following
procedures. (If more than one run is per-
formed during any 24-hour period, a calibra-
tion need not be performed prior to the
second and any subsequent runs. The cali-
bration must, however, be verified as pre-
scribed in Section 10, after the last run
made within the 24-hour period.)
8.1 General Considerations. This section
outlines steps to be followed for use of the
GC/FPD and the diludon system. The pro-
cedure do-js not include detailed Instruc-
tions because the operation of these systems
is complex, and it requires a understanding
of the individual system being used. Each
system should include a written operating
manual describing In detail the operating
procedures associated with each component
in the measurement system. In addition, the
operator should be familiar with the operat-
ing principles of the components; particular-
ly the GC/FPD. The citations in the Bib-
liography at the end of this method are rec-
ommended for review for this purpose.
8.2 Calibration Procedure. Insert the per-
meation tubes into the tube chamber.
398
-------
Check the bath temperature to assure
agreement with the calibration temperature
of the tubes within ±0.1' C. Allow 24 hours
for the tubes to equilibrate. Alternatively
equilibration may be verified by injecting
samples of calibration gas at 1-hour inter-
vals. The permeation tubes can be assumed
to have reached equilibrium when consecu-
tive hourly samples agree within the preci-
sion limits of Section 4.1.
Vary the amount of air flowing over the
tubes to produce the desired concentrations
for calibrating the analytical and dilution
systems. The air flow across the tubes must
at all times exceed the flow requirement of
the analytical systems. The concentration In
parts per million generated by a tube con-
taining a specific pcrmeant can be calculat-
ed as follows: p
c = KR[
Equation 16-1
where:
C= Concentration of permeant produced in
ppm.
Pr== Permeation rate of the tube in ppr/mln.
M= Molecular weight of the permeant (g/g-
mole).
L=Flow rate, 1/min. of air over permeant (5>
20' C, 760 mm Hg.
K = Gas constant at 20' C and 760 mm
Hg = 24.04 1/gmole.
8.3 Calibration of analysis system. Gen-
erate a series of three or more known con-
centrations spanning the linear range of the
IT?D (approximately 0.05 to 1.0 ppm) for
each of the four major sulfur compounds.
Bypassing the dilution system, inject these
standards into the GC/FPD analyzers and
monitor the responses. Three injects for
each concentration must yield the precision
described in Section 4.1. Failure to attain
this precision is an indication of a problem
in the C2libration or analytical system. Any
such problem innst be identified and cor-
rected before proceeding.
8.4 Calibration Curves. Plot the GC/FPD
response in current (amperes) versos their
causative concentrations in ppm on log-log
coordinate graph paper for each sulfur com-
pound. Alternatively, a least squares equa-
tion may be generated from the calibration
data.
8.5 Calibration of Dilution System. Gen-
erate a known concentration of hydrogen
sulfide using the permeation tube system.
Adjust the flow rate of diluent air for the
first dilution stage so that the desired level
of dilution is approximated. Inject the dilut-
ed calibration gas into the GC/FPD system
and monitor its response. Three injections
for each dilution must yield the precision
described in Section 4.1. Failure to attain
this precision in this step is an indication of
a problem in the dilution system. Any such
problem must be identified and corrected
before proceeding. Using the calibration
data for IIjS (developed under 8.3) deter-
oiine the diluted calibration gas concentra-
tion in pprn. Then calculate the dilution
factor as the ratio of the calibration gas
concentration before dilution to the diluted
calibration gas concentration determined
under this paragraph. Repeat this proce-
dure for each stage of dilution required. Al-
ternatively, the GC/PPD system may be
calibrated by generating a series of three or
more concentrations of each sulfur com-
pound and diluting these samples before in-
jecting them into the GC/FPD system. This
•data will then serve as the calibration data
for the unknown samples and a separate de-
termination of the dilution factor will not
be necessary. However, the precision re-
quirements of Section 4.1 are still applica-
ble.
9. Sampling and Analysis Procedure.
9.1 Sampling. Insert the sampling probe
into the test port making certain that no di-
lution air enters the stack through the port.
Begin sampling and dilute the sample ap-
proximtely 9:1 using the dilution system.
Note that the precise dilution factor is that
which is determined in paragraph 8.5. Con-
dition the entire system with sample for a
minimum of 15 minutes prior to commenc-
ing analysis.
9.2 Analysis. Aliquots of diluted sample
are injected Into the GC/FPD analyzer for
analysis.
9.2.1 Sample Run. A sample run is com-
posed of 16 individual analyses (injects) per-
formed over a period of not less than 3
hours or more than 6 hours.
9.2.2 Observation for Clogging of Probe.
If reductions in sample concrntrations are
observed during a sample run that cannot
be explained by process conditions, the sam-
pling must be interrupted to determine If
the sample probe is clogged with particulate
matter. If the probe Is found to be clogged,
the test must be st.opped and the results up
to that point discarded. Testing may resume
after cleaning the probe or replacing It with
a clean one. After each run, the sample
probe must be inspected and, if necessary,
dismantled and cleaned.
10. Post-Test Procedures.
10.1 Sample Line Loss. A known concen-
tration of hydrogen sulficie at the level of
the applicable standard, ±20 percent, must
be introduced into the sampling system at
the opening of the probe in sufficient quan-
tities to insure that there is an excess of
sample which must be vented to the atmo-
sphere. The sample must be transported
through the entire sampling system to the
measurement system in the normal manner.
The resulting measured concentration
should be compared to the known value to
determine the sampling system loss. A sam-
pling system loss of more than 20 percent is
unacceptable. Sampling losses of 0-20 per-
cent must be corrected for by dividing the
resulting sample concentration by the frac-
tion of recovery. The known gas sample may
be generated using permeation tubes. Alter-
natively, cylinders of hydrogen sulfide
mixed in air may be used provided they are
traceable to permeation tubes. The optional
pretest procedures provide a good guideline
for determining if there are leaks in the
sampling system.
10.2 Recalibration. After each nin. or
after a series of runs made within a 24-hour
period, perform a partial recalibration using
the procedures in Section 8. Only HaS (or
other penneant) need be used to recalibrate
the OC/FPD analysis system (8.3) and the
dilution system (8.5).
10.3 Determination of Calibration Drift.
Compare the calibration curves obtained
prior to the runs, to the calibration curves
obtained under paragraph 10.1. The calibra-
tion drift should not exceed the limits set
forth in paragraph 4.2. If the drift exceeds
this limit, the intervening run or runs
should be considered not valid. The tester.
however, may instead have the option of
choosing the calibration data set which
would give the highest sample values.
399
-------
11. Calculations.
111 Determine the concentrations of
each reduced sulfur compound detected di-
re; Uy from the calibration curves. Alterna-
tively, the concentrations may be calculated
using the equation for the least square line.
11.2 Calculation of TRS. Total reduced
sulfur will be determined for each aiiaylsis
made by summing the concentrations of
each reduced sulfur compound resolved
during a given analysis.
TRS=i (H.S. MeSH. DAIS. 2DMDS)d
Equation 16-2
where:
TRS=Total reduced sulfur in ppm. wet
basis.
HiS=Hydrogen sulfide, ppm.
MeSH --Methyl mercp.pt.an. ppm.
DMS.=-Dimethyl sulfide. ppm.
DMDS^Dimethyl disulfide. ppm.
d=Dilution factor, dimcnslonless.
11.3 Average TRS. The average TRS will
be determined as follows:
N
I TRS
Average TRS
.
)
wo
Average TRS = Average total reduced suflur
in ppm, dry basis.
TRS,-^ Total roduced sulfur in ppm as deter-
mined by Equation 16-2.
N=Number of samples.
B«=Praction of volume of water vapor in
the gas stream as determined by method
4—Determination of Moisture in Stack
Gases (36 FR 24S87).
11.4 Average concentration of Individual
reduced sulfur compounds.
N
I s.
1 = 1
Equation 16-3
where:
S,=Concentration of any reduced sulfur
compound from the ith sample injec-
tion, ppm.
C=Average concentration of any one of the
reduced sulfur compounds for the entire
run, ppm.
N=Number of injections in any run period.
12. Example System. Described below is a
system utilized by EPA in gathering NSPS
data. This system does not now reflect all
the latest developments in equipment and
column technology, but it does represent
one system that has been demonstrated to
work.
12.1 Apparatus.
12.1.1 Sampling System.
12.1.1.1 Probe. Figure 16-1 illustrates the
probe used in lime kilns and other sources
where significant amounts of particulate
matter are present, the probe is designed
with the deflector shield placed between the
sample and the gas inlet holes and the glass
wool plugs to reduce clogging of the filter
and possible adsorption of sample gas. The
exposed portion of the probe between the
sampling port and the sample line is heated
with heating tape.
12.1.1.2 Sample Line Vie inch inside diam-
eter TeHon tubing, heated to 120" C. This
temperature is controlled by a thermostatic
heater.
12.1.1.3 Sample Pump. Lcakless Teflon
coated diaphragm type or equivalent. The
pump head is heated to 120' C by enclosing
it in the sample dilution box (12.2.4 below).
12.1.2 Dilution System. A schematic dia-
gram of the dynamic dilution system is
given in Figure 16-2. The dilution system is
constructed such that all sample contacts
arc made of inert materials. The dilution
system which is he;ited to 120' C must be ca-
pable of a minimum of 9:1 dilution of
sample. Equipment used in the dilution
system is listed below:
12.1.2.1 Dilution Pump. Model A-150
Kohmyhr Teflon positive displacement
type, nonadjustable 150 cc/min. ±2.0 per-
cent, or equivalent, per dilution stage. A 9:1
dilution of sample is accomplished by com-
bining 150 cc of sample with 1.350 cc of
clean dry air as shown in Figure 16-2.
12.1.2.2 Valves. Three-way Teflon sole-
noid or manual type.
12.1.2.3 Tubing. Teflon tubing and fit-
tings are used throughout from the sample
probe to the GC/FPD to present an Inert
surface for sample gas.,
12.1.2.4 Box. Insulated box. heated and
maintained at 120° C, of sufficient dimen-
sions to house dilution apparatus.
12.1.2.5 Flownieters. Rotameters or
equivalent to measure flow from 0 to 1500
ml/min ±1 percent per dilution stage.
12.1.3 Gas Ciiromatograph Columns.
Two types of columns are used for separa-
tion of low and high molecular weight
sulfur compounds:
12.1.3.1 Low Molecular Weight Sulfur
Compounds Column (GC/FPD-1).
12.1.3.1 Separation Column. 11 in by 2.16
mm (36 ft by 0.085 in) inside diameter
Teflon tubing packed with 30/60 mesh
Teflon coated with 5 percent polyphenyl
ether and 0.05 percent orthophosphoric
acid, or equivalent (see Figure 16-3).
12.1.3.1.2 Stripper or Precolumn. 0.6 m
by 2.16 mm (2 ft by 0.085 in) inside diameter
Teflon tubing packed as in 5.3.1.
12.1.3.1.3 Sample Valve. Teflon 10-port
gas sampling valve, equipped with a 10 ml
sample loop, actuated by compressed air
(Figure 16-3).
12.1.3.1.4 Oven. For containing sample
valve, stripper column and separation
column. The oven should be capable of
maintaining an elevated temperature rang-
ing from ambient to 100° C, constant within
±1'C.
12.1.3.1.5 Temperature Monitor. Thermo-
couple pyrometer to measure column oven,
detector, and exhaust temperature ±1' C.
12.1.3.1.6 Flow System. Gas metering
system to measure sample flow, hydrogen
flow, and oxygen flow (and nitrogen carrier
gas flow).
12.1.3.1.7 Detector. Flame photometric
detector.
12.1.3.1.8 Electrometer. Capable of full
scale amplification of linear ranges of 10~*
to 10-' amperes full scale.
12.1.3.1.9 Power Supply. Capable of deli-
vering up to 750 volts.
12.1.3.1.10 Recorder. Compatible with
400
-------
the output voltage range of the electrom-
eter.
12.1.3.2 High Molecular Weight Com-
pounds Column (GC/PPD-11).
12.1.3.2.1. Separation Column. 3.05 m by
2.16 mm (10 ft by 0.0885 in) inside diameter
Teflon tubing packed with 30/60 mesh
Teflon coated with 10 percent Triton X-305,
or equivalent.
12.1.3.2.2 Sample Valve. Teflon 6-port gas
sampling valve equipped with a 10 ml
sample loop, actuated by compressed air
(Figure 16-3).
12.1.3.2.3 Other Components. All compo-
nents same as in 12.1.3.1.4 to 12.1.3.1.10.
12.1.4 Calibration. Permeation tube
system (figure 16-4).
12.1.4.1 Tube Chamber. Glass chamber
of sufficient dimensions to house perme-
ation tubes.
12.1.4.2 Mass Flowmeters. Two mass
flowmeters in the range 0-3 1/min. and 0-10
1/min. to measure air flow over permeation
tubes at ±2 percent. These flowmeters shall
be cross-calibrated at the beginning of each
test. Using a convenient flow rate in the
measuring range of both flowmeters. set
and monitor the flow rate of gas over the
permeation tubes. Injection of calibration
gas generated at this flow rate as measured
by one flcwmetcr followed by injection of
calibration gas at the same flow rate as mea-
sured by the other flowmeter should np.rre
within the specified precision limits. If they
do not. then there is a problem with the
mass flow measurement. Each mass flow-
meter shall be calibrated prior to the first
test with a wet test meter and thereafter, at
least once each year.
12.1.4.3 Constant Temperature Bath. Ca-
pable of maintaining permeation tubes at
certification temperature of 30' C. within
±o.r c.
12.2 Reagents
12.2.1 Fuel. Hydrogen (H,) prepurlfied
grade or better.
12.2.2. Combustion Gas. Oxygen (O.) re-
search purity or better.
12.2.3 Carrier Gas. Nitrogen (Ni) prepuri-
fied grade or better.
12.2.4 Diluent. Air containing less than
50 ppb total sulfur compounds and less than
10 ppm each of moisture and total hydro-
carbons, and filtered using MSA filters
46727 and 79030, or equivalent. Removal of
sulfur compounds can be verified by inject-
ing dilution air only, described in Section
8.3.
12.2.5 Compressed Air. 60 psig for GC
valve actuation.
12.2.6 Calibrated Gases. Permeation
tubes gravimetrically calibrated and certi-
fied at 30.0° C.
12.3 Operating Parameters.
12.3.1 Low-Molecular Weight Sulfur
Compounds. The operating parameters for
the GC/FPD system used for low molecular
weight compounds are as follows: nitrogen
carrier gas flow rate of 50 cc/min, exhaust
temperature of 110° C, detector temperature
of 105° C, oven temperature of 40* C, hydro-
gen How rate of 80 cc/min. oxygen flow rate
of 20 cc/min. and sample flow rate between
20 and 80 cc/min.
12.3.2 High-Molecular Weight Sulfur
Compounds. The operating parameters for
the GC/FPD system for high molecular
weight compounds are the same as in 12.3.1
except: oven temperature of 70° C, and ni-
trogen carrier gas flow of 100 cc/min.
12.4 Analysis Procedure.
12.4.1 Analysis. Aliquots of diluted
sample are injected simultaneously into
both GC/FPD analyzers for analysis. GC/
FPD-I Is used to measure the low-molecular
weight reduced sulfur compounds. The low
molecular weight compounds include hydro-
gen sulfide, methyl mercaptan. and di-
methyl sulfide. GC/FPD-II is used to re-
solve the high-molecular weight compound.
The high-molecular weight compound is di
methyl disulfide.
12.4.1.1 Analysis of Low-Molecular
Weight Sulfur Compounds. The sample
valve is actuated for 3 minutes in which
time an aliquot of diluted sample is injected
into the stripper column and analytic.il
column. The valve is then deactivated for
approximately 12 minutes in which time.
the analytical column continues to be fore-
flushed, the stripper column is backflushed.
and the sample loop is refilled. Monitor the
responses. The elution time for each com-
pound will be determined during calibra-
tion.
12.4.1.2 'Analysis of High-Molecular
Weight Sulfur Compounds. The procedure
is essentially the same as above except that
no stripper column is needed.
13. Bibliography.
13.1 O'Keeffe, A. E. and G. C. Ortman.
"Primary Standards for Trace Gas Analy-
sis." Analytical Chemical Journal. 38,760
(196G).
13.2 Stevens. R. K.. A. E. O'Keeffe, and
G. C. Ortman. "Absolute Calibration of a
Flame Photometric Detector to Volatile
Sulfur Compounds at Sub-Part-Per-Million
Levels." Environmental Science and Tech-
nology. 3:7 (July. 1969).
13.3 Mulick, J. D., R. K. Stevens, and R.
Baumgardner. "An Analytical System De-
signed to Measure Multiple Malodorous
Compounds Related to Kraft Mill Activi-
ties." Presented at the 12th Conference on
Methods in Air Pollution and Industrial Hy-
giene Studies. University of Southern Cali-
fornia. Los Angeles. CA. April 6-8. 1971.
13.4 Devonald. R. H.. R. S. Serenius. and
A. D. Mclntyre. "Evaluation of the Flame
Photometric Detector for Analysis of Sulfur
Compounds." Pulp and Paper Magazine of
Canada. 73,3 (March, 1972).
13.5 Grimley. K. W., W. S. Smith, end R.
M. Martin. "The Use of a Dynp.mlc Diinllon
System in the Conditioning of Stack Ga.«es
for Automated Analysis by a Mobile Sam-
pling Van." Presented at the 63rd Annual
APCA Meeting in St. Louis. Mo. June 14-19,
1970.
13.6 General Reference. Standard Meth-
ods of Chemical Analysis Volume III A and
B Instrumental Methods. Sixth Edition.
Van Nostrand Reinhold Co.
401
-------
Figure 16-1. Probe used for sample gas containing high particulate loadings.
-------
-X
PROBE
STACK
VJf. > '
TO GC/FPD ANALYZERS
10:1 102:1
A
FILTER
(GLASS WOOL)
FILTER
r~
HEATED
SAMPLE
LINE
o
U)
PuSiTi'v'E
DISPLACEMENT
- PUMP
PERMEATION
TUBE
CALIBRATION
GAS
fo
DIAPHRAGM
(HEATED)
^
d
DILUTION BOX HEATED
TO 100°C
DILUENT AIR
3 -WAY
.S. VALVE
' .*
13!
C
1
iOcc/
If
d r
mi L.
I
25 PS
»LcA
RYA
FLOWMETER
VENT
Figure 16- 2. Sampling and dilution apparatus.
-------
SAMPLING VALVE
GC/FPO-I
SAV.PLE
OR
CALIBRATION
GAS
STRiPPFR
JIMM
FLAME PHOTOMETRIC DETECTOR
EXHAUST
750V
POWER SUPPLY
SAMPLING VALVE FOR
GC/FPD-11
VACUUM—s;.
SAMPLED
OR
CALIBRATION
GAS
N2
•CARRIER
•^-TOGC/FPD-I!
Figure 16-3. Gas chrcmatographic-f lame photometric analyzers..
-------
TO INSTRUMENTS
AND
•DILUTION SYSTEM
CONSTANT
TEMPERATURE
BATH
PERMEATION
TUBE
Figure 16-4. Apparatus for field calibration.
405
-------
VENT
VENT
PROBE
>£>
O
O
P v,
So
13
<
u
SAMPLE
LINE
DILUTION
SYSTEM
VENT
GAS
CHROMATOGRAPH
Figure 16- 5. Determination of sample line loss.
-------
METHOD 17
DETERMINATION OF PARTICULATE EMISSIONS FROM STATIONARY SOURCES
(IN-STACK FILTRATION METHOD)
Introduction
Partlculnte matter Is not an absolute
quantity; rather, it is a function of tempera-
ture and pressure. Therefore, to prevent
variability in participate matter emission
regulations and/or associated test methods.
the temperature and pressure at which par-
ticulate n-.atlcr is to be measured must be
carefully defined. Of the two variables (i.e..
temperature and pressure), temperature has
the greater effect upon the amount of par-
ticulate matter in an effluent gas stream: in
most stationary source categories, the effect
of pressure appears to be negligible.
In method 5, 250* F is established as a
nominal reference temperature. Thus.
where Method 5 is specified in an applicable
subpart of the standards, particulate matter
is defined with respect to temperature. In
order to maintain a. collection temperature
of 250* F, Method 5 employs a heated glass
sample probe and a heated filter holder.
This equipment is somewhat cumbersome
and requires care in Its operation. There-
fore, where particulate matter concentra-
tions (over the normal rani;e of temperature
acvsociated with a specified source category)
are known to be independent of tempera-
ture, it is desirable to eliminate the glass
probe and heating systems, and sample at
stack temperature.
This method describes an in-stack sam-
pling system and sampling procedures for
use in such cases. It is intended to be used
only when specified by an applicable sub-
part of the standards, and only within the
applicable temperature limits (if specified).
or when otherwise approved by the Admin-
istrator.
1. Principle and Applicability.
1.1 Principle. Particulate matter is with-
drawn isokinetically from the source and
collected on a glass fiber filter maintained
at strck temperature. The particulate mass
is determined gravimetrically after removal
of uncombined water.
1 2 Applicability. This method applies to
the determination of particulate emissions
from stationary sources for determining
compliance with new source performance
standards, only when specifically provided
for in an applicable subpart of the stan-
dards. This method is not applicable to
stacks that contain liquid droplets or are
saturated with water vapor. In addition, this
method shall not be used as written If the
projected cross-sectional area of the probe
exttnsion-filter holder assembly covers
more than 5 percent of the stack cross-sec-
tional area (see Section 4.1.2).
2. Apparatus.
2.1 Sampling Train. A schematic of the
sampling train used-in this method is shown
in Figure 17-1. Construction details for
many, but not all, of the train components
are given in APTD-0581 (Citation 2 in Sec-
tion 7); for changes from the APTD-0581
document and for allowable modifications
to Figure 17-1. consult with the Administra-
tor.
The operating and maintenance proce-
dures for many of the sampling train com-
ponents are described in APTD-0576 (Cita-
tion 3 in Section 7). Since correct usage is
Important in obtaining valid results, all
users should read the APTD-0576 document
and adopt the operating and maintenance
procedures outlined in it. unless otherwise
specified herein. The sampling train con-
sists of the following components:
2.1.1 Probe Nozzle. Stainless steel (316)
or glass, with sharp, tapered leading edge.
The angle of taper shall be 030' and the
taper shall be on the outside to preserve a
constant internal diameter. The probe
nozzle shall be of the button-hook or elbow
design, unless otherwise specified by the Ad-
ministrator. If made of stainless steel, the
nozzle shall be constructed from seamless
tubing. Other materials of construction may
be used subject to the approval of the Ad-
ministrator.
A range of sizes suitable for isokinetic
sampling should be available, e.g., 0.32 to
1.27 cm (Vi to V4 in)—or larger if higher
volume sampling trains are used—inside di-
ameter (ID) nozzles in increments of 0.16 cm
(Vie in). Each nozzle shall be calibrated ac-
cording to the procedures outlined In Sec-
tion 5.1.
2.1.2 Filter Holder. The in-stack filter
holder shall be constructed of borosilicate
or quartz glass, or stainless steel; it a gasket
is used, it shall be made of silicone rubber.
Teflon, or stainless steel. Other holder and
gasket materials may be used subject to the
approval of the Administrator. The filter
holder shall be designed to provide a posi-
tive seal against leakage from the outside or
around the filter.
2.1.3 Probe Extension. Any suitable rigid
probe extension may be used after the filter
holder.
2.1.4 Pitot Tube. Type S. as described in
Section 2.1 of Method 2. or other device ap-
proved by the Administrator; the pitot tube
shall be attached to the probe extension to
allow constant monitoring of the stack gas
velocity (see Figure 17-1). The impact (high
pressure) opening plane of the pitot tube
shall .be even with or above the nozzle entry
plane during sampling (see Method 2,
Figure 2-6b). It is recommended: (1) that
the pitot tube have a known baseline coeffi-
cient, determined as outlined in Section 4 of
Method 2: and (2) that this known coeffi-
cient be preserved by placing the pitot tube
407
-------
n an interference-free arrangement with re-
-pe;.-! to the sampling nozzle, filter holder.
End temperature sensor (see Figure 17-1).
No'e that the 1.9 cm (0.75 in) free-space be-
tween the nozile and pito', tube shown in
figure 17-1. is based on s. 1.3 cm (0.5 in) ID
nozzle. If the sar-'ylins train is designed for
sampling at high'T flow rates than that de-
scribed in AFTD-0581. thus necessitating
the use of larger sized nozzles, the free-
sparp shail be 1.9 cm 10.15 in) with the larg-
est sized no.v.le in place.
Source-sampling assemblies that do not
meet the minimum spacing requirements of
Figure 17-1 (or the equivalent of these re-
quirements, e.g . Figure 2-7 of Method D
may be use 1, however, the pilot tube coeffi-
cionts of such assemblies shall be deter-
mi:i -d bv calibration, using methods subjoct
to the approvr.l of the Administrator.
2.1.5 Differential Pressure Gauge. In-
clined manometer or equivalent device
. as described in Section 2.2 of Method
2 One manometer shall be used for velocity
head readings, and the other, for ori-
fice differential pressure readings.
2 1 6 Condenser. It is recommended that
i!\e jpipini' i -.!>..! em described in Method 5
t' usej to dc''-riniiie the moisture foment
of Hie st.irk ga.s Alternatively, any system
tl>at allows n\ -.. urement of both the 'vatci
condensed and tlir moisture leaving the con-
denser, each to within 1 ml or 1 f. may bf
used The muis'ure leaving the connrnscr
<•»•> bi mea-urid citlior by: (1) monitorinf
tl-i- temperature and pressure at the exit ol
ll.c condenser und using Dalton's law ol
partial pres-snn-.-.. or (2) passing the sample
gas stream throurh a silica gel trap with
CM' gases kept below 20' C (68* F) and de-
termining the weu-hl gain.
Tiexible tubing may be used between the
prooe extension and condenser. If meanr
other than suicn gel are used to determine
the amount of moistore leaving the con
denser, it ib recommended that silica gel still
be used between the condenser system and
pump 10 prevent moisture condensation in
the pump and metering devices and to avoid
the np?d to make corrections for moisture
in the metered volume.
2.1.7 Metering System. Vacuum gauge,
leak-free pump, thermometers capable of
measuring temperature to within 3' C (5.4*
F). dry gas meter capable of measuring
volume to within 2 percent, and related
equipment, as shown in Figure 17-1. Other
metering systems capable of maintaining
sampling rates within 10 percent of isokine-
tic and of determining sample volumes to
within 2 percent may be used, subject to the
approval of the Administrator When the
metering system is used in conjunction with
a pilot tube, the system shall enable checks
of isokmetic rates.
Sampling trains utilizing metering sys-
Ums aesigned for higher flow rates than
thai described in APTD-0581 or APTD-0576
may be u.-ed provided that the specifica-
tions of this method are met.
218 Barometer. Mercury, aneroid, or
otaer barometer capable of measuring at-
mospheric pressure to within 2.5 mm Hg
O.I :n Hg). In many cases, the barometric
reading may be obi-ained from a nearby na-
tional weather service station, in which case
tne station value (which is the absolute
barometric pressure) shall be requested and
MI adiustment for elevation differences be-
lA-pn tr-- weather station and sampling
po.nt --ha.l he applied at a rate of minus 2.5
mm Ug (01 in. Kg) per 30 m (100 ft) eleva-
tion increase or vice versa for elevation de-
crease.
2.1.9 Gas Density Determination Equip-
ment. Temperature sensor and pressure
gauge, as described in Sections 2.3 and 2.4 of
Method 2. and gas analyzer, if necessary, as
described in Method 3.
The temperature sensor shall be attached
to either the pilot tube or to the probe ex-
tension, in a fixed configuration. If the tem-
perature sensor is attached in the field; the
sensor shall be placed in an interterence-
free arrangement with respect to the Type
S pilot tube openings (as shown in Figure
17-1 or in Figure 2-7 of Method 2). Alterna-
tively, the temperature sensor need not be
attached to either the probe extension or
pilot tube during sampling, provided that a
difference of not more than 1 percent in the
average velocity measurement is introduced.
This alternative is subject to the approval
of the Administrator.
2.2 Sample Recovery.
2.2.1 Probe Nozzle Brush. Nylon bristle
brush with stainless steel wire handle. The
brush shall be properly sized and shaped to
brush out the probe nozzle. -
2.2.2 Wash Botllos—Two. Gloss wash.
bolllrs are recommended; polyethylene
wash bottles rtiii.v bo usej at the option of
the te.stor. It is recommended that acetone
not be stored In polyethylene bottles for
loMKi't than a month.
2.2.3 Glass Sample Storage Containers.
Chemically resistant. boro.silicate glass bot-
tles, for acetone washes. 5UO ml or 1000 ml.
Screw cap liners shall cither be rubber-
bucked Teflon or shall be constructed so as
to be leak-free and resistant to chemical
attack by acetone. (Narrow mouth glass bot-
tles have been found to be less prone to
leakage.) Alternatively, polyethylene bottles
mny be used.
2.2.4 Petrt Dishes. For filter samples;
glass or polyethylene, unless otherwise
specified by the Administrator.
2.2.5 Graduated Cylinder and/or Bal-
ance. To measure condensed water to within
1 ml or 1 g. Graduated cylinders shall have
subdivisions no greater than 2 ml. Most lab-
oratory balances are capable of weighing to
the nearest 0.0 g or less. Any of these bal-
ances is suitable for use here and in Section
2.3.4.
2.2.6 Plastic Storage Containers. Air
tight containers to store silica gel.
2.2.7 Funnel and Rubber Policeman. To
aid in transfer of silica gel to container; not
necessary if silica gel is weighed in the field.
2.2.8 Funnel. Glass or polyethylene, to
aid in sample recovery.
2.3 Analysis.
2.3.1 Glass Weighing Dishes.
2.3.2 Desiccator.
2.3.3 Analytical Balance. To measure to
within 0.1 mg.
2.3.4 Balance. To measure to within 0.5
mg.
2.3.5 Beakers. 250 ml.
2.3.6 Hygrometer. To measure the rela-
tive humidity of the laboratory environ-
ment.
2.3.7 Temperature Gauge. To measure
the temperature of the laboratory environ-
ment.
3. Reagents.
3.1 Sampling.
3.1.1 Filters. The in-stack filters shall be
glass mats or thimble fiber filters, without
organic binders, and shall exhibit at least
99.95 percent efficiency (00.05 percent pene-
tration) on 0.3 micron dioctyl phthalate
408
-------
smoke particles. The filter efficiency tests
shall be conducted In accordance with
ASTM standard method D 2986-71. Test
data, from the supplier's quality control pro-
gram are sufficient for this purpose.
3.1.2 Silica Gel. Indicating type. 6- to 16-
Exesh. If previously used, dry at 175' C (350'
F) for 2 hours. New silica gel may be used as
received. Alternatively, other types of desic-
canls (equivalent or better) may be used.
subject to the approval of the Administra-
tor.
3.1.3 Crushed Ice.
3.1.4 Stopcock Grease. Acetone-Insoluble.
heat-stable silicone grease. This is not nec-
essary if screw-on connectors with Teflon
sleeves, or similar, are used. Alternatively.
other types of stopcock grease may be used,
subject to the approval of the Administra-
tor.
3.2 Sample Recovery. Acetone, reagent
grade. 00.001 percent residue, in glass bot-
tles. Acetone from metal containers general-
ly has a high residue blank and should not
be used. Sometimes, suppliers transfer ac-
etone to glass bottles from metal containers.
Thus, acetone blanks shall be run prior to
field use and only acetone with low blank
values (00.001 percent) shall be used. In no
case shall a blank value of greater than
0.001 percent of the weight of acetone used
be subtracted from the sample weight.
3.3 Analysis.
3.3.1 Acetone. Same as 3.2.
3.3.2 Dcsiccant. Anhydrous calcium sul-
fate. indicating type. Alternatively, other
types of dcsiccants may be used, subject to
the approval of the Administrator.
4. Procedure.
4.1 Sampling. The complexity of this
method is such that, in order to obtain reli-
able results, testers should be trained and
experienced with the test procedures.
4.1.1 Pretest Preparation. All compo-
nents shall be maintained and calibrated ac-
cording to the procedure described in
APTD-0576. unless otherwise specified
herein.
Weigh several 200 to 300 g portions of
silica gel in air-tight containers to the near-
est 0.5 g. Record the total weight of the
silica gel plus container, on each container.
As an alternative, the silica gel need not be
preweighed, but may be weighed directly in
its impinger or sampling holder just prior to
train assembly.
Check filters visually against light for Ir-
regularities and flaws or plnhole leaks.
Label filters of the proper size on the back
side near the edpe using numbering ma-
chine ink. As an alternative, label the ship-
ping containers (glass or plastic petri dishes)
and keep the filters in these containers at
all times except during sampling and weigh-
ing.
Desiccate the filters at 20+5.6" C (68 + 10'
P) and ambient pressure for at least 24
hours and weigh at intervals of at least 6
hours to a constant weight, i.e.. 00.5 mg
change from previous weighing; record re-
sults to the nearest 0.1 mg. During each
weighing the filter must not be exposed to
the laboratory atmosphere for a period
greater than 2 minutes and a relative hu-
midity above 50 percent. Alternatively
(unless otherwise specified by the Adminis-
trator), the filters may be oven dried at 105°
C (220° P) for 2 to 3 hours, desiccated for 2
hours, and weighed. Procedures other than
those described, which account for relative
humidity effects, may be used, subject to
the approval of the Administrator.
4.1.2 Preliminary Determinations. Select
the sampling site and the minimum number
of sampling points according to Method 1 or
as specified by the Administrator. Make a
projccted-area model of the probe exten-
sion-filter holder assembly, with the pitot
tube face openings positioned alone the ccn-
terllne of the stark, as shown in Figure 17-2.
Calculate the estimated cross-section block-
age, as shown in -Figure 17-2. If the blockage
exceeds 5 percent of the duct cross sectional
area, the tester has the following options:
(Da suitable out-of-stack filtration method
may be used instead of in-slack filtration: or
(2) a special in-slack arrangement, in which
the sampling and velocity measurement
sites are separate, may be used; for details
concerning this approach, consult wilh the
Administrator (see also Citation 10 in Sec-
tion 7). Determine the stack pressure, tem-
perature, and the range of velocity heads
using Method 2; it is recommended that a
leak-check of the pitot lines (see Method 2,
Section 3.1) be performed. Determine the
moisture * content using Approximation
Method 4 or its alternatives for the purpose
of making isokinetic sampling rate settings.
Determine the stack gas dry molecular
weight, as described in Method 2, Section
3.6; if integrated Method 3 sampling is used
for molecular weight determination, the in-
tegrated bag sample shall be taken simulta-
neously with, and for the same total length
of time as. the particular sample run.
Select a nozzle size based on the range of
velocity heads, such that It is not necessary
to change the nozzle size In o'-der to main-
tain Isokinetic sampling rates. During the
run. do not chanpro the nozzle size. Ensure
that the proper differential pressure gauge
Is chosen for the range of velocity heads en-
countered (see Section 2.2 of Method 2).
Select a probe extension length such that
all traverse points can be sampled. For large
stacks, consider sampling from opposite
sides of the stack to reduce the length of
probes.
Select a total sampling time greater than
or equal to the minimum total sampling
time specified in the test procedures for the
specific industry such that (1) the sampling
time per point is not less than 2 minutes (or
some greater time interval if specified by
the Administrator), and (2) the sample
volume taken (corrected to standard condi-
tions) will exceed the required minimum
total gas sample volume. The latter is based
on an approximate average sampling rate.
It is recommended that the number of
minutes sampled at each point be an integer
or an integer plus one-half minute, in order
to avoid timekeeping errors.
In some circumstances, e.g., batch cycles,
it may be necessary to sample for shorter
times at the traverse points and to obtain
smaller gas sample volumes. In these cases,
the Administrator's approval must first be
obtained.
4.1.3 Preparation of Collection Train.
During preparation and assembly of the
sampling train, keep all openings where con-
tamination can occur covered until just
prior to assembly or until sampling is about
to begin.
If impingers are used to .condense stack
gas moisture, prepare them as follows: place
100 ml of water in each of the first two im-
pingers, leave the third impinger empty,
and transfer approximately 200 to 300 g of
preweighed silica gel from its container to
409
-------
the fourth impinger. More silica gel may be
used, but care should be taken to ensure
that it is net entrained and carried out from
the impinger during sampling. Place the
container in a clean place for later use in
the sample recovery. Alternatively, the
weight of the silica eel plus impinger may
be determined to the nearest 0.5 g and re-
corded. •'
If some means other than impingers Is
used to condense moisture, prepare the con-
denser (and, if appropriate, silica gel for
condenser outlet) for use.
Usinp a tweezer or clean disposable surgi-
cal gloves, place a labeled (identified) and
weighed filter in the filter holder. Be sure
that the filter is properly centered »nd the
Basnet properly placed so as not to allow the
sample.- gas slreiun to circumvent the filter.
Check filter for tears after assembly is com-
Dleied. Mark the probe extension with heat
resistant tape or by some other method to
denote the proper distance Into the stack or
duct for each sampling point.
Assemble the tram as in Figure 17-1, using
a very light coat of silicone grease on all
ground glass joints and greasing only the
outer portion (.see APTD-0576) to avoid pos-
sibility of contamination by the silicone
grease. Place crushed ice around the im-
pingers.
4.1.4 Leak Check Procedures.
4.1.4.1 Pretest Leak-Check. A pretest
leak check is recommended, but not re-
quired. If the tester opts to conduct the pre-
test leak-check, the following procedure
shall be used.
After the sampling train has been assem-
bled, plug the inlet to the probe nozzie with
a material th:U will be able to withstand the
stack temperature. Insert the lilter holder
into the stack and wait approximately 5
minutes (or longer, if necessary) to allow
the system to come to equilibrium with the
temperature of the stack cas stream. Turn
on the pump and draw a vacuum of at least
380 mm Hg (15 in. Hg); note that a lower
vacuum may be used, provided that it Is not
exceeded during the test. Detennine the
leakage rate. A leakage rate in excess of 4
percent of the average sampling rate or
0.00057 m'/min. (0.02 cfm), whichever is
less, is unacceptable.
The following leak-check instructions for
the sampling train described in APTD-0576
and APTD-0581 may be helpiul. Start the
pump with by-pass valve fully open and
coarse adjust valve completely closed. Par-
tially open the coarse adjust valve and
slowly c}ose the by-pass valve until the de-
sired vacuum is reached. Do not reverse di-
rection of b.v-pass valve. If the desired
vacuum is exceeded, either leak-check at
this higher vacuum or end the leak-check as
ihov.'n below and start over.
When the leak-check is completed, first
slowly remove the plug from the inlet to the
probe nozzl-- and immediately turn off the
vacuum pump. This prevents water from
being forced backward and keeps silica gel
flora being entrained backward.
4 1.4 2 Le^K-Chccks During Sample Run.
If, during the sampling run, a component
(e.g.. filte. assembly or impinger) change be-
comes neceb.sa.ry, a leak-check shall be con-
ducted immediately before the change is
made. The leak-check shall be done accord-
ing to the procedure outlined In Section
4.1.4.1 above, except that it shall be cione at
a vacuum equal to or greater than the maxi-
mum value recorded up to that point in the
U-.st. If the leakage rate is found to be no
greater than 0.00057 m'/min (0.02 cfm) or 4
percent of the average sampling; rate
(whichever Is less), the results are accept-
able, and no correction will need to he ap-
plied to the total volume of dry gas metered:
if. however, a higher leakage rate is ob-
tained, the tester shall either record the
leakage rate and plan to correct the sample
volume as shown in Section 6.3 of this
method, or shall void the sampling run.
Immediately after component changes,
leak-checks are optional; if such leak-checks
are done, the procedure outlined in Section
4.1.4.1 above shall be used.
4.1.4.3 Post-Test Leak-Check. A leak-
check is mandatory at the conclusion of
each sampling run. The leak-check shall be
done in accordance with the procedures out-
lined iii Section 4.1.4.1. except that it shall
be conducted at a vacuum equal to or great-
er than the maximum value reached during
the sampling run. If the leakage rate is
found to be no greater than C.00057 m'/min
(0.02 cfm) or 4 percent of the average sam-
pling rate (whichever is less), the results are
acceptable, and no correction need be ap-
plied to the total volume of dry gas metered.
If, however, a higher leakage rate is ob-
tained, the tester shall either record the
leakage rate and correct the sample volume
as shown in Section 6.3 of this method, or
shall void the sampling run.
4.1.5 Particulate Train Operation.
During the sampling run, maintain a sam-
pling rate such that sampling is within 10
percent of true isokinetic, unless otherwise
specified by the Administrator.
For each run, record the data required on
the example data sheet shown in Figure 17-
3. Be sure to record the initial dry gas meter
reading. Record the dry gas meter readings
at the beginning and end of each sampling
time increment, when changes in flow rates
are made, before and after each leak, check,
and when sampling is halted. Take other
readings required by Figure 17-3 at least
once at each sample point during each time
increment and additional readings when sig-
nificant changes (20 percent variation in ve-
locity head readinps) necessitate additional
adjustments in flow rate. Level and zero the
manometer. Because the manometer level
and zero may drift due to vibrations and
temperature changes, make periodic checks
during the traverse.
Clean the portholes prior to the test run
to minimize the chance of sampling the de-
posited material. To begin sampling, remove
the nozzle cap and verify that the pitot tube
and probe extension are properly posi-
tioned. Position the nozzle at the first tra-
verse point with the tip pointing directly
Into the gas stream. Immediately start the
pump and adjust the flow to isokinetic con-
ditions. Nomographs are available, which
aid in the rapid adjustment to the isokinetic
sampling rate without excessive computa-
tions. These nomographs are designed for
use when the Type S pitot tube coefficient
Is 0.85±0.02. and the stack gas equivalent
density (dry molecular weight) is equal to
29-4. APTD-0576 details the, procedure for
using the nomographs. If C, and M,, are out-
side the above stated ranges, do not use the
nomographs unless appropriate steps (see
Citation 7 in Section 7) are taken to com-
pensate for the deviations.
When the stack Is under significant nega-
tive pressure (height of impiruser stem),
take care to close the coarse aajust valve
before insertins the probe ex-iension a-isem-
bly into the stack to prevent water fro-n
410
-------
being forced backward, if necessary, the
pump may be turned on with the coarse
adjust valve closed.
When the probe is in position, block off
the openings around the probe and porthole
to prevent unrepresentative dilution of the
gas stream.
Traverse the stack cross section, as re-
quired by Method 1 or as specified by the
Administrator, being careful not to bump
the probe nozzle into the stack walls when
sampling near the walls or when removing
or inserting the probe extension through
the portholes, to minimize chance of ex-
tracting deposited material.
During the test run. take appropriate
steps (e.g., adding crushed ice to the im-
pinger ice bath) to maintain a temperature
of less than 20° C (68' P) at the condenser
outlet; this will prevent excessive moisture
losses. Also, periodically check the level and
zero of the manometer.
If the pressure drop across the filter be-
comes too high, making isokinetic sampling
difficult to maintain, the filter may be re-
placed in the midst of a sample run. It is
recommended that another complete filter
holder assembly be used rather than at-
tempting to change the filter itself. Before a
new filter holder is installed, conduct a leak
check, as outlined in Section 4.1.4.2. The
total particulate weight shall include the
summation of all filter assembly catches.
A single train shall be used for the entire
sample run, except in cases where simulta-
neous sampling is required in two or more
separate duels or at two or more different
locations within the same duct, or, in cases
where equipment failure necessitates a
change of trains. In all other situations, the
use of two or more trains will be subject to
the approval of the Administrator. Note
that when two or more trains are used, a
separate analysis of the collected particu-
late from each train shall be performed,
unless identical nozzle sizes were used on all
trains, in which case the particulate catches
from the individual trains r.iay be combined
and a single analysis performed.
At the end of the sample run. turn off the
pump, remove ths probe extension assembly
from the stack, and record the final dry gas
meter reading. Perform a leak-check, ns out-
lined In Section 4.1.4.3. Also, Irak-check the
pilot lines as described in Section 3.1 ol
Method 2; the lines must pass this leak-
check. In order to validate the velocity head
data.
4.1.6 Calculation of Percent Isokinetic.
Calculate percent isokinellc (see Serlion
6.11) lo delrrmine whether another lest run
should be made. If there is difficulty In
maintaining Isokinellc rates due to source
conditions, consult with the Administrator
for possible variance on the Isokinetic rates.
4.2 Sample Recovery. Proper cleanup
procedure begins as soon as the probe ex-
tension assembly Is removed from the stack
at the end of the sampling period. Allow the
assembly to cool.
When the assembly can be safely handled,
wipe off all external particulate matter near
the tip of the probe nozzle and place a cap
over it to prevent losing or gaining particu-
late matter. Do not cap off the probe tip
tightly while the sampling train is cooling
down as this would create a vacuum in the
filter holder, forcing condenser water back-
ward.
Before moving the sample train to the
cleanup site, disconnect the filter holder-
probe nozzle assembly from the probe ex-
tension; cap the open iiilet of the probe ex-
tension. Be careful not to lose any conden-
sat«, if present. Remove the umbilical cord
from the condenser outlet and cap the
outlet. If a flexible line is used between the
first fmpinger (or condenser) and the probe
extension, disconnect the line at the probe
extension and let any condensed water or
liquid drain into the implngers or condens-
er. Disconnect the probe extension from the
condenser, cap the probe extension outlet.
After v.-iping off the silicone grease, cap off
the condenser inlet. Ground glass stoppers,
plastic caps, or serum caps (whichever are
appropriate) may be used to close these
openings.
Transfer both the filter holder-probe
nozz'.e assembly and the condenser to the
cleanup area. This area should be clean and
protected from the wind so that the chances
of contaminating or losing the sample will
be minimized.
Save a portion of the acetone used for
cleanup as a blank. Take 200 ml of this ac-
etone directly from the wash bottle being
used and place it in a glass sample container
labeled "acetone blank."
Inspect the train prior to and during dis-
assembly and note any abnormal conditions.
Treat the samples as follows:
Container No. 1. Carefully remove the
filter from the filter holder and place it in
its identified petri dish container. Use a pair
of tweezers and/or clean disposable surgical
gloves to handle the filter. If it is necessary
to fold the filter, do so such that the partic-
ulate cake is inside the fold. Carefully trans-
fer to the petri dish any particulate matter
and/or filter fibers which adhere to the
filter holder gasket, by using a dry Nylon
bristle brush and/or a sharp-edged blade.
Seal the container.
Container No. 2. Taking care to see that
dust on the outside of the probe nozzle or
other exterior surfaces does not get into the
sample, quantitatively recover particulf.te
matter or any condi>nsate from the probe
nozzle, fitting, and front half of the filter
holder by washing these components with
acetone and placing the wash in a glass con-
tainer. Distilled water may be ustd instead
of acetone when approved by the Adminis-
trator and shall be used when specified by
the Administrator; in these cases, save a
water blank and follow Administrator's di-
rections on analysis. Perform the acetone
rinses as follows:
Carefully remove the probe nox/.le ana
clean the Inside surface by rinsing with ac-
etone from a wash bottle and brushlnp with
a Nylon bristle brush. Brush until arrtone
rinse shows no visible particles, after which
make a final rinse of the Inside surface with
acetone.
Brush and rinse with acetone the inside
parts of the fitting In a similar way until no
visible partk-les remain. A funnel (glass or
polyethylene) may be used to aid in trans-
ferring liquid washes to the container. Rinse
the brush with acetone and quantitatively
collect these washings in the sample con-
tainer. Between sampling runs, keep
brushes clean and protected from contami-
nation.
After ensuring that all joints are wiped
clean of silicone grease (if applicable), clean
the inside of the front half of the filter
holder by rubbing the surfaces with a Nylon
bristle brush and rinsing with acetone.
Rinse each surface three times or more if
needed to remove visible particulate. Make
final rinse of the brush and filter holder.
411
-------
After all acetone washings and particulaie
matter are collected in the sample contain-
er, tighten the lid on the sample container
so that acetone wi'l not leak out when it is
shipped to the laboratory. Mark the height
of the fluid level to determine whether or
not leakage occurred during transport.
Label the container to clearly identify its
contents.
Container No. 3. if silica gel is used in the
condenser system for mositure content de-
termination, note the color of the gel to de-
termine if it has been completely spent:
make a notation of its condition. Transfer
the silica gel back to its original container
and seal. A funnel may make it easier to
pour the silica gel without spilling, and a
rubber policeman may be used as an aid in
removing the silica gel. It is not necessary to
remove the small amount of dust particles
that may adhere to the walls and are diffi-
cult to remove. Since the gain in weignt is to
be used for moisture calculations, do not use
any water or other liquids to transfer the
silica gel. If a balance is available in the
field, follow the procedure for Container
No. 3 under "Analysis."
Condenser Water. Treat the condenser or
impinger .water as follows: make a notation
of any color or film in the liquid catch. Mea-
sure the liquid volume to within ±1 ml by
using a graduated cylinder or, if a balance is
available, determine the liquid weight to
within ±0.5 g. Record the total volume or
weight of liquid present. This iniormation is
required to calculate the moisture content
of the effluent gas. Discard the liquid after
measuring and recording the volume or
weight.
4.3 Analysis. Record the data required on
the example sheet shown in Figure 17-4.
Handle each sample container as follows:
Container No. 1. Jjeave the contents in the
shipping container or transfer the filter and
any loose particulate Irom the sample con-
tainer to a tared glass weighing dish. Dejic-
care for 24 hours in a desiccator containing
anhydrous calcium sulfale. \Veitfh to a con-
stant weipht and report thn results to the
nearest O.j mg. For purposes of this Section,
4.3. the term "constant weight" moans a dif-
Icrence o! no more than 0.5 mg or 1 percent
ol local At-ight less tare weight, whichever is
i! eater, between two consecutive weighings.
uith no less than 0 hours of desiccation
tune between weighings.
Alternatively the sample may be oven
dried at the average stack temperature or
105' C <220' F). -Ahichever is less! tor 2 to 3
hours, cooled m the desiccator, and weighed
to a constant Y\fiKht, unless otherwise speci-
fi« d bv the Administrator. The tester mny
ai:. cr:i to oirii ikv flu- sample at the aver-
age stack temperature or 105' C (220* F).
whichever is less, for 2 to 3 hours. wclRh the
-.iir.plr and use this weight as a fuiai
we-.phl.
Container No. 2. Note the level of liquid in
the container and confirm on the analysis
sheet whether or not leakage occurred
during transport. If a noticeable amount of
leakage has occurred, either void the sample
or use methods, subject to the approval of
the Administrator, to correct the final re-
sults. Measure the liquid in this container
either volumetrically to ±1 ml or gravimc-
trically to • «.5 g. Transfer the contents to a
tared 250-ml beaker and evaporate to dry-
ness at ambient temperature and pressure.
Desiccate for 24 hours and weigh to a con-
stant weight Report the rtsulcs to the near-
est 0.1 mg.
Container Ho. 3. This step may be con-
ducted in the field. Weigh the spent silica
gel (or silica gel plus impinger) to the near-
est 0.5 g using a balance.
"Acetone Blank" Container., Measure ac-
etone in this container either volumetrically
or gravimetrically. Transfer the acetone to a
tared 250-ml beaker and evaporate to dry-
ness at ambient temperature and pressure.
Desiccate for 24 hours and weigh to a con-
stant weight. Report the results to the near-
est 0.1 mg.
NOTE.—At the option of the tester, the
contents of Container No. 2 as well as the
acetone blank container may be evaporated
at temperatures higher than ambient. If
evaporation is done at an elevated tempera-
ture, the temperature must be below the
boiling point of the solvent; also, to prevent
"bumping." the evaporation process must be
closely supervised, and the contents of the
beaker must be swirled occasionally to
maintain an even temperature. Use extreme
care, as acetone is highly flammable and
has a low flash point.
5. Calibration. Maintain a laboratory log
of all calibrations.
5.1 Probe Nozzle. Probe nozzles shall be
calibrated before their initial use in the
field. Using a micrometer, measure the
inside diameter of the nozzle to the nearest
0.025 mm (0.001 In.). Make three separate
measurements using different diameters
each time, and obtain the average of the
measurements. The difference between the
hifrh and low numbers shall not exceed 0.1
mm (0.004 in.). When no/zlcs become
nicked, dented, or corroded, they shall be
reshaped, sharpened, and recalibrated
before use. Each nozzle shall be permanent-
ly and uniquely identified.
5.2 Pilot Tube. If the pilot tube is placed
in an interference-free arrangement with re-
spect to the other probe assembly compo-
nents, its baseline (isolated tube) coefficient
shall be determined as outlined in Section 4
of Method 2. If the probe assembly is not in-
terference-free, the pilot tube assembly co-
efficient shall be determined by calibration,
using methods subject to the approval of
the Administrator.
5.3 Metering System. Before its initial
use in the field, the metering system shall
be calibrated according to the procedure
outlined in APTD-0576. Instead of physical-
ly adjusting the dry gas meter dial readings
to correspond to the wet test meter read-
ings, calibration factors may be used to
mathematically correct the gas meter dial
readings to the proper values.
Before calibrating the metering system, it
is suggested that a leak-check be conducted.
For metering systems having diaphragm
pumps, the normal leak-check procedure
will not detect leakages within the pump.
For these cases the following leak-check
procedure is suggested: make a 10-minute
calibration run at 0.00057 m'/min (0.02
cfm); at the end of the run, take the differ-
ence of the measured wet test meter and
dry gas meter volumes; divide the difference
by 10, to get the leak rate. The leak rate
should not exceed 0.00057 m'/min (0.02
cfm).
After each field use. the calibration of the
metering system shall be checked by per-
forming three calibralion runs at a single.
intermediate orifice setting (based on the
previous field test), with the vacuum set at
412
-------
Ihc maximum value reached during the tost
scries. To adjust the vacuum, insert a valve
between the wet test meter and the Inlot of
the metering system. Calculate the average
value of the calibration factor. If the cali-
bration has changed by more than 5 per-
cent, recalibrate the meter over the full
range of orifice settings, as outlined in
APTD-0576.
Alternative procedures, e.g., using the ori-
fice meter coefficients, may be used, subject
to the approval of the Administrator.
NOTE.—If the dry gas meter coefficient
values obtained before and after a test
series differ by more than 5 percent, the
test series shall either be voided, or calcula-
tions for the test scries shall be performed
using whichever meter coefficient value
(i.e., before or after) gives the lower value of
total sample volume.
5.4 Temperature Gauges. Use the proce-
dure in Section 4.3 of Method 2 to calibrate
in-stack temperature gauges. Dial thermom-
eters, such as are used for the dry gas meter
and condenser outlet, shall be calibrated
against mercury-in-glass thermometers.
5.5 Leak Check of Metering System
Shown in Figure 17-1. That portion of the
sampling train from the pump to the orifice
meter should be leak checked prior to initial
use and after each shipment. Leakage after
the pump will result in less volume being re-
corded than is actually sampled. The follow-
ing procedure is suggested (see Figure 17-5).
Close the main valve on the meter box.
Insert a one-hole rubber stopper with
rubber tubing attached into the orifice ex-
haust pipe. Disconnect and vent the low side
of the orifice manometer. Close off the low
side orifice tap. Pressurize the system to 13
to 18 cm (5 to 7 in.) water column by blow-
ing into the rubber tubing. Pinch off the
tubing and observe the manometer for one
minute. A loss of pressure on the mano-
meter indicates a leak in the meter box;
leaks, if present, must be corrected.
. 5.6 Barometer. Calibrate against a mer-
cury barometer.
6. Calculations. Carry out calculations, re-
taining at least one extra decimal figure
beyond that of the acquired data. Round off
figures after the final calculation. Other
forms of the equations may be used as long
as they give equivalent results.
6.1 Nomenclature.
An=Cross-sectional area of nozzle, m" (ft1).
B,,=Water vapor in the gas stream, propor-
tion by volume.
C.=Acetone blank residue concentration,
mg/g.
c.=Concentration of particulatc matter in
stack gas, dry basis, corrected to stan-
dard conditions, g/dscm (g/dscf).
I=Percent of isokinetic sampling.
L, = Maximum acceptable leakage rate for
either a pretest leak check or for a leak
check following a component change;
equal to 0.00057 m'/rnin (0.02 cfm) or 4
percent of the average sampling rate,
whichever is less.
li!=lndlvidual leakage rate observed during
the leak check conducted prior to the
"I"1" component change (1 = 1. 2. 3 . . . n),
mVmin (cfm).
Lp=Leakage rate observed during the post-
test leak check, m'/min (cfm).
m.=Total amount of particulate matter col-
lected, mg.
Mw=Molecular weight of water, 18.0 g/g-
mole (18.0 Ib/lb-mole).
m.=Mass of residue- of acetone after evapo-
ration, mg.
Pi»i=Barometric pressure at the sampling
site, mm Hg (in. Hg).
P.=Absolute stack gas pressure, mm Hg (in.
Hg).
P,w=Standard absolute pressure, 760 mm
Hg (29.92 in. Hg).
R=Ideal gas constant, 0.06236 mm Hg-m'/
•K-g-mole (21.85 in. Hg-ftVR-lb-mole).
Tn=Absolute average dry gas meter tem-
perature (see Figure 17-3). "K CR).
T.=Absolute average stack gas temperature
(see Figure 17-3). °K CR).
T.M=Standard absolute temperature, 293'K
(528=R>.
V.=Volume of acetone blank, ml.
V..=Volume of acetone used in wash. ml.
Vu=Total volume of liquid collected in im-
pingers and silica gel (see Figure 17-4),
ml.
Vm=Volume of gas sample as measured by
dry gas meter, dcm (dcf).
Vnuid)=Vo!ume of gas sample measured by
the dry gas meter, corrected to standard
conditions, dscm (dscf).
V«(.,=Volume of water vapor in the gas
sample, corrected to standard condi-
tions, scm (scf).
v.=Stack gas velocity, calculated by Method
2. Equation 2-Q, using data obtained
from Method 17, m/sec (ft/sec).
W.=Weight of residue in acetone wash. mg.
Y=Dry gas meter calibration coefficient.
AH=Average pressure differential across
the orifice meter (see Figure 17-3). mm
H,O (in. HjO).
/>.=Density of acetone, mg/ml (see label on
bottle).
sw=Density of water. 0.9982 g/ml (0.002201
Ib/ml).
0=Total sampling time, min.
0, = Samp)inb' lime interval, from the begin-
ning of a run until the first component
change, min.
0,=Sampling time Interval, between two
successive component changes, begin-
ning with the interval between Uie first
and second changes, min.
P,=Sampllng time Interval, from the final
(nlh) component change until the end ol
the sampling run, min.
13.6 = Specific gravity of mercury.
60 = Sec/min.
100 = Con version to percent.
6.2 Average dry gas meter temperature
and average orifice pressure drop. See data
sheet (Figure 17-3).
6.3 Dry Gas Volume. Correct the sample
volume measured by the dry gas meter to
standard conditions (20' C. 760 mm Hg or
68" F. 29.92 in. Hg) by using Equation 17-1.
Vm(std) " V
. „
bar
AH
Kstd
Pbar + (AH/13.6)
Equation 17-1
where:
K,=0.3858" K/mm Hg for metric units;
17.64' R/in. Hg for English units.
NOTE.—Equation 17-1 can be used as writ-
ten unless the leakage rate observed during
any of the mandatory leak checks (i.e., the
413
-------
post-test leak cneck or leak checks conduct-
ed prior to component changes) exceeds L..
If 1+ or I, exceeds L.. Equation 17-i must be
modified as follows-
(a) Case I No component changes made
during sampling run. In this case, replace
Vm in Equation 17-1 with the expression:
(b) Case II. One or more component
changes made during the sampling run. In
this case, replace Vm in Equation 17-1 by the
expression:
91 '
Li ' La>
6.11 Isokinetic Variation.
6.11.1 Calculation from Raw Data.
100 Ts 1K3V1C » (V/T ) (Pbar *
A ~ " JCh a>, ft - it
60 ev P A
n
Equation 17-7
where:
K,=0.003454 mm Hg-m'/ml-'K for metric
units: 0.002669 in. Hg-ft'/ml-'R for Eng-
lish units.
6.11.2 Calculation from Intermediate
Values.
- La>
and substitute only for those leakage rates
(L, or L»> which exceed L..
6.4 Volume of water vapor.
Equation 17-2
where:
K,^ O.OOU33 m'/ml for metric units; 0.04707
It'/inl for English units.
6.5 Moir.ture Content.
B - "w(std)
ws Vm(std) + Vw(std)
Equation 17-3
6.6 Acetone Blank Concentration.
"' " Va "a
Equation 17-4
6.7 Acetone Wash Blank.
W.=C.V..p.
Equation 17-5
6.8 Total Particulate Weight. Determine
the total paniculate catch from the sum of
the weights obtained from containers 1 and
2 less the acetone blank (see Figure 17-4).
NOTE.—Refer to Section 4.1.5 to assist in
calculation of results involving two or more
filter assemblies or two or more sampling
trains.
6.9 Particulate Concentration.
c.=(0.001 g/mg) (m./V.taa))
Equation 17-6
6.10 Conversion Factors:
Prom
To
Multiply by
scf.— _ m'
g/ff gr/ft»
g/ff lb/ft'
K'ff g/m' —
_ 0.02832
15.43
2.205x10-'
35.31
Ts Vstd)Pstd
std
Ts Vm(std)
vs Aje (,-Cws)
Equation 17-6
where:
K«=4.320 for metric units; 0.09450 for Eng-
lish units.
• 6.12 Acceptable Results. If 90 percent
010110 percent, the results are acceptable. If
the results are low In comparison to the
standard and I is beyond the acceptable
range, or. if I is less than 90 percent, tile Ad-
ministrator may opt to accept the results.
Use Citation 4 in Section 7 to innke judg-
ments. Otherwise, reject the results and
repeat the test.
7. Bibliography.
1. Addendum to Specifications for Inciner-
ator Testing at Federal Facilities. PUS.
NCAPC. December 6. 1967.
2. Martin, Robert M.. Construction Details
of Isokinetic Source-Sampling Equipment.
Environmental Protection Agency. Re-
search Triangle Park, N.C. APTD-0581.
April. 1971.
3. Rom. Jerome J.. Maintenance. Calibra-
tion, and Operation of Isokinetic Source-
Sampling Equipment. Environmental Pro-
tection Agency. Research Triangle Park.
N.C. APTD-0576. March. 1972.
4. Smith, W. S., R. T. Shigehara. and W.
F. Toild. A Method of Interpreting Stack
Sampling Data. Paper Presented at the 63rd
Annual Meeting of the Air Pollution Con-
trol Association. St. Louts, Mo. June 14-19
1970.
5. Smith, W. S., et al.. Stack Gas Sampling
Improved and Simplified with New Equip-
ment. APCA Paper No. 67-119. 1967.
6. Specifications for Incinerator Testing at
Federal Facilities. PHS, NCAPC. 19G7.
7. Shigehara. R. T.. Adjustments in the
EPA Nomograph for Different Pitot Tube
Coefficients and Dry Molecular Weights.
Stack Sampling News 2:4-11. October, 1974.
8. Vollaro. R. F.. A Survey of Commercial-
ly Available Instrumentation for the Mea-
surement of Low-Range Gas Velocities. U.S.
Environmental Protection Agency, Emission
Measurement Branch. Research Triangle
Park.. N.C. November. 1976 (unpublished
paper).
9. Annual Book of ASTM Standards. Part
26. Garoous Fuels; Coal and Coke: At mo-
414
-------
TEMPERATURE
SENSOR
IN-STAOt
FILTER HOLOW
x«
y > 1.9 cm (0.75 in.)*
z>7.6 em (3 in.)*
TYPE-S
PITOTTUBE
TEMPERATURE
SENSOR
IMPINGER TRAIN OPTIONAL. MAY BL REPLACED
BY AN EQUIVALENT CONDEfcSER
SAMPLING
NOZZLE
IN-STACK
FILTER
HOLDER
REVERSE-TYPE
PITOTTUBt
V
STACK
WALL
PROBE
EXTENSIOM
>
THERMOMETER
. CHECK
/ VALVE
jf
^
PITOTMANOMETER
ORIFICE MANOMETER
1 SUGGESTED (INTERFERENCE-FREE) SPACINGS
=• =
If
VACUUM
LINE
3- 3 p
a TJ
IP?
t fS o
f* » o
COS
5T ^
AIRTIGHT
PUMP
2
DRY GAS METER
Figure 17-1. Particulate-Sampling Train, Equipped with In-Stack Filter.
-------
STACK
WALL
IN-STACK FILTER-
PROBE EXTENSION
ASSEMBLY
ESTIMATED
BLOCKAGE
DUCT AREA
AREA]
REA J
X 100
Figure 17-2. Projected-area model of cross-section blockage (approximate average for
a sample traverse) caused by an in-stack filter holder-probe extension assembly.
416
-------
PLANT
LOCATION.
OPERATOR.
DATE
RUN NO.
SAMPLE BOX NO..
METER BOX N0._
METERAH@
CFACTOR
PITOT TUBE COEFFICIENT. Cp.
BAROMETRIC PRESSURE.
ASSUMED MOISTURE %_
PROBE EXTENSION LENGTH, m(ft.).
NOZZLE IDENTIFICATION NO
AVERAGE CALIBRATED NOZZLE DIAMETER cm (in.!
FILTER NO
LEAK RATE, m3/min,(cfm)
STATIC PRESSURE, mm Hg (in. Hg).
SCHEMATIC OF STACK CROSS SECTION
TRAVERSE POINT
NUMBER
TOTAL
SAMPLING
TIME
(01. min.
AVERAGE
VACUUM
mm Hg
(in. Hg)
STACK
TEMPERATURE
Jty
°C (*F)
VELOCITY
HEAD
(A PS),
mm H20
(in. H20)
PRESSURE
DIFFERENTIAL
ACROSS
ORIFICE
METER,
mm H20
(in. H20)
GAS SAMPLE
VOLUME,
m3 (ft3)
GAS SAMPLE TEMPERATURE
AT DRY GAS METER
INLET.
°C(°F)
Avg
OUTLET.
°C(°F)
Avci
Avg
TEMPERATURE
OF GAS
LEAVING
CONDENSER OR
LASTIMPINGER,
°C(°F)
1
-------
Plant
Date_
Run No
Filter No.
Amount liquid lost during transport
Acetone blank volume, ml
Acetone wash volume, ml
Acetone black concentration, mg/mg (equation 174)
Acetone wash blank, mg (equation 17-5)
CONTAINER
NUMBER
1
2
TOTAL
WEIGHT OF PARTICULATE COLLECTED,
mg
FINAL WEIGHT
I^xCl
TARE WEIGHT
IT>~
-------
RUBBER
TUBING
RUBBER
STOPPER
ORIFICE
BY-PASS VALVE
VACUUM
GAUGE
IO
BLOW INTO TUBING
UNTIL MANOMETER
READS 5 TO 7 INCHES
WATER COLUMN
ORIFICE
MANOMETER
AIR-TIGHT
PUMP
Figure 17-5. Leak check of meter box.
-------
APPENDIX F
FEDERAL STATIONARY SOURCE PERFORMANCE STANDARDS
Since December 1971 the Environmental Protection Agency has
promulgated source performance standards for twenty industrial
air pollution sources. These regulations have been presented
in numerous issues of the Federal Register and can now be found
in the Code of Federal Regulations under Title 40 - Protection
of Environment; Chapter 1 - Environmental Protection Agency;
Subchapter C - Air Programs; Part 60 - Standards of Performance
For New Stationary Sources.
This appendix presents a summary of these regulations as
found in the Code of Federal Regulations. This summary is an
updated listing as found originally in the following reference:
L.S. Chaput. 1976. Federal Standards of Performance For
New Stationary Sources Of Air Pollution - A Summary
of Regulations. Journal of the Air Pollution Control
Association. Vol. 26, No. 11j 1055-1060.
420
-------
Standards of Performance - 40 CFR Part 60.
Source Allected
category facility
Subpart D:
Steam generators Coal fired boilers
(>250 million Btu/hr)
Promulgated
12/23/71 (36 FR 24876)
Revised
7/26/72 (37 FR 14877)
6/14/74 (39 FR 20790) oil fired boilers
1 /1 6/75 (40 FR 2803)
10/6/75 (40 FR 46250)
3/7/78(43FR9278)
Gas fired boilers
Subpart E:
Incinerators Incinerators
(>50 tons/day)
Pollutant
Participate
Opacity
SO2
NOX
(except lignite
and coal
refuse)
Particulate
Opacity
S02
NOX
Particulate
Opacity
NOX
Particulate
Emission level
0.10 lb/106 Btu
20%
1.2 lb/106 Btu
0.70 lb/106 Btu
0.10 lb/106 Btu
20% ; 40% 2 min/hr
0.80 lb/106 Btu
0.30 lb/106 Btu
0.10 lb/106 Btu
20%
0.20 lb/106 Btu
0.80 gr/dscf corrected
to 12% CO
Monitoring
requirement
No requirement
Continuous
Continuous
Continuous
No requirement
Continuous
Continuous
Continuous
No requirement
No requirement
Continuous
No requirement
Promulgated
12/23/71 (36 FR 24876)
Revised
6/14/74 (39 FR 20790)
Subpart F:
Portland cement plants
Promulgated
12/23/71 (36 FR 24876)
Revised
6/14/74 (39 FR 20790)
1 1/1 2/74 (39 FR 39874)
10/6/75 (40 FR 46250)
Subpart G:
Nitric acid plants
Kiln
Clinker cooler
Fugitive
Emission points
Process equipment
Particulate
Opacity
Particulate
Opacity
Opacity
Opacity
NOX
0.30 Ib/ton
20%
0.10 Ib/ton
10%
10%
10%
3.0 Ib/ton
No requirement
No requirement
No requirement
No requirement
No requirement
No requirement
Continuous
Promulgated
12/23/71 (36 FR 24876)
Revised
5/23/73 (38 FR 13562)
6/14/74 (39 FR 20790)
10/6/75 (40 FR 46250)
Subpart H:
Sulfuric acid plants
Promulgated
12/23/71 (36 FR 24876)
Revised
5/23/73 (38 FR 13562)
6/14/74 (39 FR 20790)
10/6/75 (40 FR 46250)
Process equipment
SO2
Acid mist
Opacity
4.0 Ib/ton
0.15 Ib/ton
10%
Continuous
No requirement
No requirement
421
-------
Source
category
Emliaion level
Monitoring
requirement
Subpart I:
Asphalt concrete plants
Promulgated
3/8/74 (39 FR 9308)
Reused
10/6/75 (40 FR 46250)
Dryers; screening and Particulate 0.04 gr/dscf
weighing systems; (90 mg/dscm)
storage, transfer, and Opacity 20%
loading systems; and
dust handling equipment
No requirement
No requirement
Subpart J:
Petroleum refineries
Promulgated
3/8/74 (39 FR 9308)
Catalytic cracker
Particulate 1.0lb/1000lb No requirement
Opacity 30% (3 min. exemption) Continuous
CO 0.05% Continuous
Rew'sed
10/6/75 (40 FR 46250)
3/15/78(43FR10868)
Fuel gas
combination
SO2
0.1 gr H2S/dscf
(230 mg/dscm)
Continuous
Subpart K:
Storage vessels for
petroleum liquids
Promulgated
3/8/74 (39 FR 9308)
Revised
4/17/74 (39 FR 13776)
6/14/74 (39 FR 20790)
Storage tanks
>40,000 gal. capacity
Hydrocarbons
For vapor pressure
78-570 mm Hg, equip
with floating roof,
vapor recovery system,
or equivalent; for
vapor pressure >570
mm Hg, equip with
vapor recovery system
or equivalent
No requirement
Subpart L:
Secondary lead
smelteis
Promulgated
3/8/74 (39 FR 9308)
Reviced
4/17/74 (39 FR 13776)
10/6/75 (40 FR 46250)
Reverberatory and
blast furnaces
Pot furnaces
Particulate
Opacity
Opacity
0.022 gr/dscf
(50 mg/dscm)
20%
10%
No requirement
No requirement
No requirement
Subpart M:
Secondary brass and
bronze plants
Promulgated
3/8/74 (39 FR 9308)
flew'sed
10/6/75 (40 FR 46250)
Reverberatory
furnace
Blast and
electric furnaces
Particulate
Opacity
Opacity
0.022 gr/dscf
(50 mg/dscm)
20%
10%
No requirement
No requirement
No requirement
Subpart N:
Iron and steel plants
Promulgated
3/8/74 (39 FR 9308)
Basic oxygen
process furnace
Particulate
0.022 gr/dscf
(50 mg/dscm)
No requirement
422
-------
Source
calopory
AHeelcd
facility
Pollutant
EmlMion Uv»l
Monitoring
requirement
Subpart O:
Sewage treatment
plants
Promulgated
3/8/74 (39 FR 9308)
Revised
4/17/74 (39 FR 13776)
5/3/74 (39 FR 15396)
10/6/75 (40 FR 46250)
Sludge incinerators
Paniculate
Opacity
1.30lb/ton
20%
Mass or volume
of sludge
No requirement
Subpart P:
Primary copper
smelters
Promulgated
1/15/76(41 FR2331)
flews ed
2/26/76 (41 FR 8346)
Dryer
Roaster, smelting
furnace,* copper
converter
'Reverberaiory furnaces
that process high-
impurity feed materials
are exempt from
SO2 standard
Particulate
Opacity
S02
Opacity
0.022 gr/dscf
(50 mg/dscm)
20%
0.065%
20%
No requirement
Continuous
Continuous
No requirement
Subpart Q:
Primary zinc
smelters
Promulgated
1/15/76(41 FR2331)
Subpart R.:
Primary lead
smelters
Promulgated
1/15/76(41 FR2331)
Subpart S:
Primary aluminum
reduction plants
Promulgated
1/26/76 (41 FR 3825)
Sintering machine
Roaster
Blast or reverberatory
furnace, sintering
macnine discharge end
Sintering machine,
electric smelting
furnace, converter
Potroom group
(a) Soderberg
plant
(b) Prebake
plant
Anode bake plants
Particulate
Opacity
SO2
Opacity
Particulate
Opacity
S02
Opacity
(a) Total
fluorides
Opacity
(b) Total
fluorides
Opacity
Total fluorides
Opacity
0.022 gr/dscf
(50 mg/dscm)
20%
0.065%
20%
0.022 gr/dscf
(50 mg/dscm)
20%
0.065%
20%
2.0 Ib/ton
10%
1.9 Ib/ton
10%
0.1 Ib/ton
20%
No requirement
Continuous
Continuous
No requirement
No requirement
Continuous
Continuous
No requirement
No requirement
No requirement
No requirement
No requirement
No requirement
No requirement
423
-------
Source
category
Subpart T:
Phosphate fertilizer
plants
Promulgated
8/6/75 (40 FR 331 52)
Subpart U:
Subpart V:
Subpart W:
Subpart X:
Subpart Y:
Coal preparation
plants
Promulgated
1/15/76(41 FR2232)
^""'f1 Pollutant Emission level
facility
Wet process Total fluorides 0.02 Ib/ton
phosphoric acid
Superphosphoric acid Total fluorides 0.01 Ib/ton
Diammonium Total fluorides 0.06 Ib/ton
phosphate
Triple super- Total fluorides 0.2 Ib/ton
phosphate
Granular triple Total fluorides 5.0 xlO'4
superphosphate Ib/hr/ton
Thermal dryer Particulate 0.031 gr/dscf
(0.070 g/dscm)
Opacity 20%
Pneumatic coal Particulate 0.01 8 gr/dscf
cleaning equipment (0.040 g/dscm)
Opacity 10%
Processing and Conveying Opacity 20%
equipment, storage
systems, transfer and
loading systems
Monitoring
requirement
Total pressure
drop across
process
scrubbing
system
Total pressure
drop across
process
scrubbing
system
Total pressure
drop across
process
scrubbing
system
Total pressure
drop across
process
scrubbing
system
Total pressure
drop across
process
scrubbing
system
Temperature
Scrubber
pressure loss
Water pressure
No requirement
No requirement
No requirement
No requirement
Subpart Z:
Ferroalloy production
tacilities
Promulgated
5/4/76(41 FR 18497)
Revised
5/20/76 (41 FR 20659)
Electric submerged
arc furnaces
Particulate
0.99 Ib/Mw-hr
(0.45kg/Mw-hr)
("high silicon alloys")
0.51 Ib/Mw-hr
(0.23 kg/Mw-hr)
(chrome and
manganese alloys)
No visible emissions
may escape furnace
capture system
No requirement
Flowrate
monitoring
in hood
424
-------
Source
catogory
Affected
facility
Pollutant
Emission level
Monitoring
requirement
Ferroalloy production
facilities (cont.)
Opacity
CO
No visible emission
may escape tapping
system for >40% of
each tapping period
15%
20% volume basis
Flowrate
monitoring
in hood
Continuous
No requirement
Oust handling equipment
Opacity
10%
No requirement
Subpart AA:
Iron and steel
plants
Promulgated
9/23/75 (40 FR 43850)
Electric arc furnaces
Particulate
Opacity
(a) control
device
(b) shop roof
Bus. handling equipment
Opacity
0.0052 gr/dscf
(12 mg/dscm)
3%
0, except
20%—charging
40%—tapping
10%
No requirement
Continuous
Flowrate
monitoring
in capture hood
Pressure
monitoring
in DSE system
No requirement
Subpart BB:
Kraft pulp mills
Promulgated
2/23/78(43FR7572)
Straight and cross
recovery furnace
Straight kraft
recovery furnace
Particulate
Opacity
Total reduced
sulfur
0.044 gr/dscf corrected
to 8% oxygen
35%
5 ppm by volume on a
dry basis, corrected
to 8% oxygen
Continuous
Continuous
Cross recovery
furnace
Total reduced
sulfur
25 ppm by volume on a
dry basis, corrected
to 8% oxygen
Continuous
Smelt dissolving
tank
Lime kiln
Particulate
Total reduced
sulfur
Particulate
Total reduced
sulfur
0.1 g/kg black liquor solids
(dry weight)
0.0084 g/kg black liquor solids ' Continuous
(dry weight)
0.067 gr/dscf corrected to
10 % oxygen when gaseous
fossil fuel is burned
0.13 gr/dscf corrected to
10% oxygen when liquid
fossil fuel is burned
8 ppm by volume on a
dry basis, corrected to
10% oxygen
Continuous
Digester system,
brown stock
washer system,
multiple-effect
evaporator system,
black liquor
oxidation system,
condensate stripper
system
Total reduced
sulfur
425
5 ppm by volume on a
dry basis, corrected to
10% oxygen, except as
noted in CFR.
Continuous
-------
BIBLIOGRAPHY
Air Pollution Control Association (ed.). Proceedings of a Spe-
cialty Conference on: The User and Fabric Filtration Equip-
ment. October 14-16, 1973, Buffalo, New York. 176 pp.
Air Pollution Control Association (ed.). Proceedings of a Second
Specialty Conference on: The User and Fabric Filtration
Equipment. October 5-7, 1975, Buffalo, New York. 197 pp.
Allen, T. 1975. Particle Size Measurement. John Wiley and Sons,
New York, New York. 454 pp.
American Air Filter Company. Proceedings: 2nd International
Fabric Alternatives Forum, Denver, Colorado, July 27 and
28, 1977.
American Society For Testing and Materials. Standard Method For
Sampling Stacks For Particulate Matter. Designation D2928-
71, Annual Book of ASTM Standards, 1977.
American Society of Mechanical Engineers. Power Test Code-27:
Determining Dust Concentrations in a Gas Stream. ASME,
New York, April, 1957.
Batel, W. State of Development and Tendencies for the Filtering
Dust Collector. Staub Reinhaltung der Luft in English,
33 (1973), No. 9, September. pp. 352-362.
426
-------
Bibliography (Cont'd.)
Billings, C. E. (ed.). Proceedings: Fabric Filter Symposium
March 16-18, 1971, Charleston, South Carolina. Sponsored
by Division of Process Control Engineering, Air Pollution
Control Office, U.S. Environmental Protection Agency.
Billings, C. E., and J. Wilder. Handbook of Fabric Filter Tech-
nology. Vol. I, II, III, IV. Furnished to the National
Air Pollution Control Administration Under Contract Number
CPA-22-69-33 by GCA Corporation. December 1970.
Bird, A. N., Jr. 1973. Summary of Techniques Used for Measuring
Concentration and Size of Fine Particulate Emission. Memo-
randum Prepared for the Working Group for Stationary Source
Air Pollution Control Technology- Southern Research Insti-
tute, Birmingham, Alabama. 14 pp.
Blake, D. E. 1976. Symposium on Particulate Control in Energy
Processes. EPA-600/7-76-010. U.S. Environmental Protec-
tion Agency, Research Triangle Park, North Carolina 27711.
584 pp.
Blake, D. E. Source Assessment Sampling System: Design and
Development. EPA-600/7-78-018, February, 1978. U.S.
Environmental Protection Agency, Research Triangle Park,
NC. 221 pp.
Bradway, R. M., and R. W. Cass. Fractional Efficiency of a
Utility Boiler Baghouse - Nucla Generating Plant. EPA-600/2-
75-013a, U.S. Environmental Protection Agency, Washington,
D.C., August, 1975. 148 pp.
427
-------
Bibliography (Cont'd.)
Brenchley, D. L., G. D. Turley, and R. F. Yarmac. 1974. Indus-
trial Source Sampling. Ann Arbor Science Publishers, Inc.,
Ann Arbor, Michigan. 484 pp.
Breslin, A. J., S. F. Guggenheim, and A. C. George. Compact High
Efficiency Diffusion Batteries. Staub (in English), 31(8),
pp. 1-5 (1971).
Cass, R. W., and R. M. Bradway. Fractional Efficiency of a Utility
Boiler Baghouse: Sunbury Steam-Electric Station. EPA-600/2-
76-077a, U.S. Environmental Protection Agency, Washington,
B.C., March, 1976. 244 pp.
Chang, Hsing-Chi. A Parallel Multicyclone Size-Selective Par-
ticulate Sampling Train. Amer. Ind. Hyg. Assoc. J., Sep-
tember, 1974.
Cooper, Douglas W., and Vladimir Hampl. Fabric Filtration Per-
formance Model. In: Proc. Conf. Particulate Collection
Problems Converting Low Sulfur Coals. EPA-600/7-76-016,
U.S. Environmental Protection Agency, Research Triangle Park,
NC, 1976. pp. 149-185.
Gushing, K. M., G. Lacey, J. D. McCain, and W. B. Smith. Parti-
culate Sizing Techniques for Control Device Evaluation:
Cascade Impactor Calibrations. EPA-600/2-76-280, U.S.
Environmental Protection Agency, Research Triangle Park,
North Carolina, October, 1976, 94 pp, NTIS-PB-245184 ($5.75).
Davies, C. N. 1966. Aerosol Science. Academic Press, New York,
New York 10003. 468 pp.
428
-------
Bibliography (Cont'd.)
Davies, C. N. and M. Aylward. Proc. Phys. Soc., (64) :889, 1951.
Dennis, Richard, and J. Wilder. Fabric Filter Cleaning Studies.
EPA-650/2-75-009, U.S. Environmental Protection Agency,
Washington, D.C., January, 1975. 438 pp.
Dennis, R., R. W. Cass, and J. E. Langley. Development of Data
Base and Fabric Filtration Models for Design of Particulate
Control Systems, Field Related Tests. Contract No. 68-02-
1438, U.S. Environmental Protection Agency, Research Triangle
Park, NC, 1975. 119 pp.
Dorman, R. G. Chapter VIII. In: Aerosol Science, C. N. Davies,
ed. New York, Academic Press, 1966.
Draemel, D. C. Relationship Between Fabric Structure and Filtra-
tion Performance in Dust Filtration. EPA-R2-73-288, U. S.
Environmental Protection Agency, Research Triangle Park,
North Carolina, July, 1973. 83 pp.
Englund, H. M., and W. T. Berry (ed.). Control Technology: Par-
ticulates. APCA Reprint Series. Air Pollution Control
Association, Pittsburgh, PA, July, 1973. 80 pp.
Ensor, David S., R. G. Hooper, and R. W. Scheck. Determination
of the Fractional Efficiency, Opacity Characteristics,
Engineering and Economic Aspects of a Fabric Filter Operat-
ing on a Utility Boiler. Electric Power Research Institute
Report No. EPRI FP-297. November, 1976.
429
-------
Bibliography (Cont'd.)
Ensor, D. S. and B. S. Jackson. Evaluation of a Particulate
Scrubber On A Coal-Fired Utility Boiler. EPA-600/2-75-074,
1975. U.S. Environmental Protection Agency, Research Triangle
Park, NC 27711.
Federal Register. Standards of Performance For New Stationary
Sources: Revision to Reference Methods 1-8. Volume 42,
Number 160, Thursday, August 18, 1977, pp. 41753-41789-
Federal Register. Standards of Performance for New Stationary
Sources. Vol. 36, No. 247, December 23, 1971.
Felix, L. G., G. I. Clinard, G. E. Lacey, and J. D. McCain. In-
ertial Cascade Impactor Substrate Media For Flue Gas Sampling.
Interagency Energy-Environment Research and Development
Program Report, EPA-600/7-77-060, June, 1977.
Fuchs, N. A., I. B. Stechkina, and V.I. Starasselskii. On The
Determination Of Particle Size Distribution On Polydisperse
Aerosols By The Diffusion Method. Brit. J. Appl. Phys.,
13, 280-81 (1962).
Fuchs, N. A. The Mechanics of Aerosols. The MacMillan Co.
60 - 5th Avenue, NY (1954), pp. 204-12.
Haberl, J. B. and S. J. Fusco. Condensation Nuclei Counters:
Theory and Principles of Operation. General Electric Tech-
nical Information Series, No. 70-POD 12 (1970) .
Hall, R. R., and R. Dennis. Mobile Fabric Filter System Design
and Field Test Results. EPA-650/2-75-059, U.S. Environmental
Protection Agency, Washington, D.C., July, 1975. 136 pp.
430
-------
Bibliography (Cont'd.)
Hamersma, J. W., S. L. Reynolds, and R. F. Maddalone. Procedures
Manual for Level 1 Environmental Assessment. EPA-600/2-76-
160a, EPA, Research Triangle Park, North Carolina.
Harris, D. B. Procedures For Cascade Impactor Calibration And
Operation In Process Streams. U.S. Environmental Protection
Agency Report No. EPA-600/2-77-004, January, 1977.
Hewitt, G. W. The Charging of Small Particles for Electrostatic
Precipitation. AIEE Winter General Meeting, New York. Paper
No. 73-283, 1957.
Holland, C. R., and E. Rothwell. Model Studies of Fabric Dust
Filtration 1. Flow Characteristics of Dust Cakes Uniformly
Distributed on Filter Fabrics. Filtration Separation 14:30,
1977.
Huang, C. M., B. G. McKinney, G. A. Hollinden, and J. R. Crooks.
Review of Major Developments In Particulate Control Tech-
nologies in USA. Tennessee Valley Authority, Chattanooga,
Tennessee, October, 1976.
Johnson, J. W., G. I. Clinard, L. G.Felix, and J. D. McCain.
A Computer-Based Cascade Impactor Data Reduction System.
EPA-600/7-78-042, March, 1978. U.S. Environmental Protection
Agency, Research Triangle Park, N.C. 602 pp.
Jones, A. A. Theory and Application of Fabric Dust Filters.
Pit and Quarry, 9 (n.a.), 110-120, 1972.
431
-------
Bibliography (Cont'd.)
Koscianowski, J. R. and L. Koscianowski. Effect of Filtration
Parameters on Dust Cleaning Fabrics. EPA-600/2-76-074,
U.S. Environmental Protection Agency, Washington, D.C.,
March, 1976. 185 pp.
Koscianowski, L., J. Zareba, and J. Koscianowski. Study of Fabric
Filters For Industrial Dust Collecting Devices. National
Technical Information Service Report No. 258835-T, 1974.
Limbach, R. F. Nodular Deposits in Fabric Filters. Paper No.
77-51.6,, Presented at the 70th Annual Meeting of the Air
Pollution Control Association, Toronto, Canada. June 20-
24, 1977.
Liu, B. Y. H., and D. Y. H. Pui. A Sub-micron Aerosol Standard
And The Primary, Absolute Calibration Of The Condensation
Nuclei Counter. J. of Colloid and Interface Science, 47(1),
(1974) .
Liu, B. Y. H., and D. Y. H. Pui. On the Performance of the Elec-
trical Aerosol Analyzer. J. Aerosol Science, 6, pp. 249-
64 (1975).
Liu, B. Y. H. (ed.). 1976. Fine Particles: Aerosol Generation,
Measurement, Sampling, and Analysis. Academic Press. New
York, New York. 837 pp.
Liu, B. Y. H., K. T. Whitby, and D. Y. H. Pui. A Portable Elec-
trical Aerosol Analyzer for Size Distribution Measurements
of Submicron Aerosols. APCA J. 24(11):1067-72, 1974.
432
-------
Bibliography (Cont'd)
MacLauren, E., and C. Junge. Relationship of Cloud Nuclei-System
to Aerosol Size Distribution and Composition. J. of Atmos-
pheric Science, 4, (1971).
Maddalone, R. and N. Garner. Process Measurements Procedures -
Sulfuric Acid Emissions. TRW Systems Group Document 28055-
6004-RO-OO, February, 1977.
Marlow, W. H., P. C. Reist, and G. A. Dwiggins. Aspects of the
Performance of the Electrical Aerosol Analyses Under Non
Ideal Conditions, J. Aerosol Science, 7, 457 (1976).
Marple, V. A. and K. L. Rubow. Aerodynamic Particle Size Calibra-
tions of Optical Particle Counters. J. Aerosol Science,
7, pp. 425-438 (1976).
Marple, V. A. and K. Willeke. 1976. Impactor Design. Atmos-
pheric Environment. Vol 10: 891-896.
Marple, V. A. Doctoral Thesis, University of Minnesota, Minnea-
polis, MN, 1970.
McCain, J. D., K. M. Gushing, and A. N. Bird, Jr. 1973. Field
Measurements Of Particle Size Distribution With Inertial
Sizing Devices. EPA-650/2-73-035. U.S. Environmental Pro-
tection Agency, Washington, D.C. 20460. 52 pp. NTIS-PB
226292/$.75.
McCain, J. D., K. M. Gushing, and W. B. Smith. 1974. Methods
For Determining Particulate Mass and Size Properties: Labora-
tory and Field Measurements. Journal of the Air Pollution
Control Association, Vol. 24, No. 12: 1172-1176.
433
-------
Bibliography (Cont'd.)
McKee, D. E., R. E. Lucas, and J. E. Smith. A Review of Pulse
Jet Baghouse and Dual Mechanical Collector Performance on
Spreader Stokes Fired Boilers. Paper No. 77-26.4. Presented
at the 70th Annual Meeting of the Air Pollution Control Asso-
ciation, Toronto, Canada, June 20-24, 1977.
National Air Pollution Control Administration. Control Techniques
For Particulate Air Pollutants Publication Number AP-51,
January, 1969.
Oglesby, S., Jr. 1975. Survey Of Information On Fine Particle
Control. EPRI 259/Final Report. Southern Research Insti-
tute, Birmingham, Alabama. Prepared For Electric Power
Research Institute, Palo Alto, California. 247
Oglesby, S. A Review of Technology for Control of Fly Ash Emis-
sions from Coal in Electric Power Generation. Report to
Argonne National Laboratory. Southern Research Institute
Report No, SORI-EAS-77-243. July 1, 1977. 245 pp.
Parry, E. P. and R. A. Meyer. Gas Flow Measurements in Air Pol-
lution Monitoring. Presented at the 67th Annual Meeting
of the Air Pollution Control Association, Denver, CO (June
1974) .
Prill, F. W., L. V. Bing, and D. R. Noll. Fabric Filter Appli-
cations on Drum Mix Asphalt Plants. Paper Number 77-16-3.
70th Annual Meeting of the Air Pollution Control Association
June 20-24, 1977.
434
-------
Bibliography (Cont'd.)
Ragland, J. W., K. M. Gushing, J. D. McCain, and W. B. Smith.
HP-65 Programmable Pocket Calculator Applied To Air Pollu-
tion Measurement Studies: Stationary Sources. U.S. Environ-
mental Protection Agency Report No. EPA-600/2-77-002, October
1976.
Ragland, J. W., K. M. Gushing, J. D. McCain, and W. B. Smith.
HP-25 Programmable Pocket Calculator Applied To Air Pollu-
tion Measurement Studies: Stationary Sources. Interagency
Energy-Environment Research and Development Program Report,
EPA-600/7-77-058, June 1977.
Ragland, J. W., D. H. Pontius, and W. B. Smith. 1976. Design,
Construct, and Test A Field Usable Prototype System For
Sizing Particles Smaller Than 0.5 pm Diameter - Special
Summary Report (SORI-EAS-76-094) Southern Research Insti-
tute, Birmingham, Alabama. 65 pp.
Ramsey, G. H., et al. EPA Fabric Filtration Studies: 2. Per-
formance of Non-woven Polyester Filter Bags. EPA-600/2-76-
168b, U.S. Environmental Protection Agency, June, 1976.
37 pp.
Ranz, W. E., and J. B. Wong. Impaction of Dust and Smoke Par-
ticles. Industrial Engineering Chemistry. 44(6):137-138,
1952.
Rao, A. K. Doctoral Thesis, University of Minnesota, Minneapolis,
MN, 1975.
Reigel, S. A., and R. P. Bundy. Why the swing to baghouses?
Power (January, 1977), pp. 68-73.
435
-------
Bibliography (Cont'd.)
Sansone, E. B. and D. A. Weyel. A Note On The Penetration Of
A Circular Tube By An Aerosol With A Log-Normal Size Dis-
tribution. Aerosol Science, 2, pp. 413-15 (1971).
Schmidt, E. W., J. A. Greseke, and J. M. Allen. Size Distribu-
tions Of Fine Particulate Emissions From A Coal-Fired Power
Plant. Atmospheric Environment, 10, 1065-1069 (1976).
Schrag, M. P-, and L. J. Shannon. Evaluation of Electric Field
Fabric Filtration. EPA-600/2-76-041, U.S. Environmental
Protection Agency, Washington, D.C., February, 1976.
25 pp.
Seale, L. M. (ed.). Proceedings: Symposium on the Use of Fabric
Filters for the Control of Submicron Particulates, April
8-10, 1974, Boston, Massachusetts. EPA-650/2-74-043, U.S.
Environmental Protection Agency, Washington, D.C., May, 1974.
316 pp.
Sem, G. J., e_t al. 1971. State of the Art: 1971. Instrumenta-
tion for Measurement of Particulate Emissions From Combustion
Sources. Volume I - PB 202665, Volume II PB 202666, Volume
III - PB 223393, Volume IV - PB 231919. Thermosystems, Inc.,
St. Paul, MN. 605 pp.
Sem, G. J. Submicron Particle Size Measurement of Stack Emissions
Using The Electrical Mobility Technique. Presented at the
Electric Power Research Institute Workshop on Sampling,
Analysis, and Monitoring of Stack Emissions, October (1975).
Session 32 - Particulate Control (Filters). Proceedings presented
at the 70th Annual Meeting of the Air Pollution Control
Association, Toronto, Canada, June 20-24, 1977.
436
-------
Bibliography (Cont'd.)
Session 57 - Control Equipment For Particulate Emissions, III -
Filtration. Proceedings presented at the 68th Annual Meeting
of the Air Pollution Control Association, Boston, Massachu-
setts, June 15-20, 1975.
Sinclair, D., and G. S. Hoopes. A Continuous Flow Condensation
Nuclei Counter. Aerosol Science, 6, pp. 1-7 1975.
Sinclair, D. A Novel Form of Diffusion Battery. Am. Ind. Hygiene
Assoc. J-, 36(1), pp. 39-42 (1975).
Sinclair, D. A Portable Diffusion Battery. Am. Ind. Hygiene
Assoc. J., 33(11), pp. 729-35 (1972).
Sinclair, D., R. J. Countess, B. Y. H. Liu, and D. Y. H. Pui.
Experimental Verification of Diffusion Battery Theory.
Presented at the 68th Annual Meeting of the Air Pollution
Control Association, Boston, MA, 1975. Paper No. 75-41.4.
Smith, W. B., K. M. Gushing, and J. D. McCain. 1974. Particle
Sizing Techniques For Control Device Evaluation. EPA-650/2-
74-102. U.S. Environmental Protection Agency, Washington,
D.C. 127 pages. NTIS-PB 240670/$5.75.
Smith, W. B., and R. R. Wilson, Jr. Development and Laboratory
Evaluation of A Five-Stage Cyclone System. EPA-600/7-78-
008, January 1978. U.S. Environmental Protection Agency,
Research Triangle Park, NC. 66 pp.
Smith, W. B., K. M. Gushing, G. E. Lacey, and J. D. McCain. 1975.
Particle Sizing Techniques For Control Device Evaluation.
EPA-650/2-74-012a. U.S. Environmental Protection Agency,
Washington, D.C. 132 pp. NTIS-PB 245184/$5.75.
437
-------
Bibliography (Cont'd.)
Smith, F. H. 1974. The Effects Of Nozzle Design And Sampling
Techniques On Aerosol Measurements. EPA-650/2-74-076.
U.S. Environmental Protection Agency, Washington, D.C.
89 pp.
Soderholm, S. C. Modification of A Commercial Condensation Nuclei
Counter For Steady Flow. Atmospheric Environment, Vol. 10:
657-660, 1976.
Southern Research Institute. A Review of Technology for Control
of Fly Ash Emissions From Coal in Electric Power Generation.
Report Number SORI-EAS-77-243. Birmingham, AL, July, 1977.
Southern Research Institute. 1975. Proceedings Of The Workshop
On Sampling, Analysis, And Monitoring Of Stack Emissions.
EPRI Special Report 41. Prepared for the Electric Power
Research Institute, Palo Alto, CA. 346 pp.
Spink, L. K. and D. M. Considine. Process Instruments and Con-
trols Handbook. McGraw-Hill, 1957.
Standard Havens, Inc. Baghouses and Evaporative Coolers For Air
Pollution Control. Kansas City, MO, 1973. 64 pp.
Stephan, D. G. and G. W. Walsh. Residual Dust Profiles In Air
Filtrations. Ing. Eng. Chem., (52), 999 (1960).
Stern, A. C. 1968. Air Pollution. Academic Press, New York,
New York. 644 pp.
Strauss, W. Industrial Gas Cleaning. 2nd Edition. Pergamon
Press, Oxford, England, 1975. 621 pp.
438
-------
Bibliography (Cont'd.)
Thomas, J. W. The Diffusion Battery Method For Aerosol Particle
Size Determination. J. Colloid Science, 10, p. 246 (1955).
Twomey, S. The Determination of Aerosol Size Distribution From
Diffusional Decay Measurements. J. Franklin Inst., 275,
pp. 121-38 (1963).
U.S.-U.S.S.R. Working Group - Stationary Source Air Pollution
Control Technology. Proceedings of the Symposium On Control
of Fine Particulate Emissions From Industrial Sources. Janu-
ary 15-18, 1974. San Francisco, CA. Published by Southern
Research Institute, Birmingham, Alabama.
Whitby, K. T. and B. Y. H. Liu. Generation of Countable Pulses
by High Concentrations of Subcountable Sized Particles in
the Sensing Volume of Optical Counters. J. Colloid Inter-
face Science, 25: 537-46, 1967.
Yeh, H. C. A Fundamental Study of Aerosol Filtration by Fibrous
Filters. Ph.D. Thesis, University of Minnesota, March, 1972,
439
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
. REPORT NO
EPA-600/7-78-113
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
Procedures Manual for Fabric Filter Evaluation
5. REPORT DATE
June 1978
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Kenneth M. Gushing and Wallace B. Smith
8. PERFORMING ORGANIZATION REPORT NO.
SORI-EAS-78-355FFM
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Southern Research Institute
2000 Ninth Avenue, South
Birmingham, Alabama 35205
10. PROGRAM ELEMENT NO.
EHE624
11. CONTRACT/GRANT NO.
68-02-2131, WA 21104
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT A
Task Final; ""
OVERED
4. SPONSORING AGENCY CODE
EPA/600/13
5. SUPPLEMENTARY NOTES TERL_RTP project officer is D. Bruce Harris, Mail Drop 62,
919/541-2557.
6. ABSTRACT
The report describes methods to be used in experimentally characterizing
the performance of fabric filters for pollution control. It, gives a detailed description
of the mechanical characteristics of fabric filters. It describes procedures for
measuring particle size distribution, the mass concentration of particulate matter,
and the concentration of major gaseous components of t.he flue-gas/particle mix-
ture. It gives a concise discussion and outline, describing the development of a
test plan for evaluating a fabric filter installation. By following this outline, useful
tests may be performed, ranging in complexity .from qualitative and relatively
inexpensive to rather elaborate research programs.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
ENDED TERMS
c. COSATI Field/Group
I3B
Air Pollution
Dust Control
Gas Filters
Fabrics
valuation
Measurement
Air Pollution Control
Stationary Sources
Particulate
Fabric Filters
Size Distribution
Mass Concentration
13K
HE
14B
. D.STRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (This Report I
Unclassified
21. NO. OF PAGES
451
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
440
------- |