FINAL REPORT
IMPROVED CHEMICAL METHODS FOR SAMPLING AND
ANALYSIS OF GASEOUS POLLUTANTS FROM THE
COMBUSTION OF FOSSIL FUELS
VOLUME I
SULFUR OXIDES
Contract No. CPA 22-69-95
JUNE 1971
Prepared for
ENVIRONMENTAL PROTECTION AGENCY
Cincinnati, Ohio
-------
FINAL REPORT
IMPROVED CHEMICAL METHODS FOR SAMPLING AND
ANALYSIS OF GASEOUS POLLUTANTS FROM
THE COMBUSTION OF FOSSIL FUELS
VOLUME I
SULFUR OXIDES
Prepared under
Contract No. CPA 22-69-95
by
J. N. Driscoll
and
A. W. Berger
Maiden Research Corporation
359 Allston Street
Cambridge, Massachusetts
June 1971
Prepared for
Environmental Protection Agency
Cincinnati, Ohio
-------
When u.s. Government drawings, specifications, or other data
are used for any purpose other than a definitely related
Government procurement operation, the Government thereby in-
curs no responsibility nor any obligation whatsoever, and
the fact that the Government may have formulated, furnished,
or in any way supplied the said drawings, specifications, or
other data, is not to be regarded by implication or other-
wise, or in any manner licensing the holder or any other per-
son or corporation, or conveying any rights or permission to
manufacture, use, or sell any patented invention that may in
any way be related thereto.
References to named commercial products in this report are not
to be considered in any sense as an endorsement of the product
by the Government.
iii
WALDEN RESEARCH CORPORATION
-------
FOREWORD
The compilation of information contained in this publication was
performed pursuant to Contract No. CPA 22-69-95, Air Pollution Control
Office, Environmental Protection Agency.
The information was compiled by Walden Research Corporation and
their subcontractor, Arthur D. little, during the period June 12, 1969
to September 11, 1970.
Volume I of this report examines state-of-the-art manual methods
for sampling and analysis of sulfur oxides. The determination of sul-
fur dioxide (10-3000 ppm) and sulfur trioxide (5-300 ppm) is reviewed
for stationary fossil fuel combustion sources and for control equipment
which may be installed to reduce the emissions from such sources.
Volume II reviews methods for the determination of oxides of nitro-
gen in stationary fossil fuel combustion sources for the concentration
range 5-2000 ppm for nitric oxide (NO), nitrogen dioxide (N02)' and
total oxides of nitrogen (NO + N02 or NOx)'
Volume III of this report examines state-of-the-art manual methods
for sampling and analysis of carbon monoxide in stationary fossil fuel
combustion sources for the concentration range 10-1000 ppm.
The major pollutants of interest in this study are sulfur and nitro-
gen oxides. The report subsequently includes a literature search, re-
view of methodology and laboratory investigation on the most promising
methods. Carbon monoxide, the minor pollutant, was limited to a litera-
ture search and review of methodology.
This report has been reviewed and approved.
-------~--- ---'
R~~~
Robert l. larkin
Project Officer
Process Measurements Section
Division of Control Systems
Office of Air Programs
Environmental Protection Agency
v
WALDEN RESEARCH CORPORATION
-------
Section
1
2
3
4
TABLE OF CONTENTS
Tit1 e
Introduction
...... ..... .... ..... ....... ..... .....
1.1 Na~io~a! SOx Emissions and Method Application
Prl0r,tles ..................................
The Combustion Effluent Environment ..............
2. 1 I ntroducti on ................................
2 . 2 Gene ra 1 .....................................
2.3 Temperatures.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.4 Mo is tu re ....................................
2.5 Particulate Matter ..........................
2.6 Su1 fur Compounds............................
2.7 Other Gaseous Speci es .......................
2.8 Trans; ents ..................................
Chemistry of,Su1fur Oxides in Flue Gases .........
3.1 Introduction ................................
3.2 Properties of S02 and S03 ...................
3.2.1 Equilibrium Constants for Formation of
'502 and SO~ from the Elements ........
3.2.2 The Vapor res sure of.S93 :...........
3.2.3 The S03-H20-H2S04 Equl11Drlum ........
3.3 The Oxidation of Gaseous S02 ................
3.3.1 Homogeneous Oxidation ................
3.3.2 Heterogeneous Oxidation ..............
3.3.3 Sulfate Decomposition ................
3.4 Adsorption of S02 ...........................
3.5 Solution Chemistry of S02 ...................
3.5.1 The Solubility of S02 ................
3.5.2 Acid Dissociation Constants ..........
3.5.3 Redox Chemi stry ......................
3.6 Conclusions .................................
Manual Sampling Methods for Gaseous Pollutants ...
4.1 Introduction................................
4.2 Convnon Factors..............................
4.2.1 Pre-Test Planning ....................
4.2.2 Stack Measurements...................
vii
Page
1
5
5
5
7
10
10
11
11
12
13
13
13
13
15
15
15
15
19
19
21
21
21
25
25
26
28
28
28
28
29
WALDEN RESEARCH CORPORATION
-------
Section
TABLE OF CONTENTS (Cont.)
Ti tl e
4.2.3 Sampling Train Components
.............
4.2.3. 1 General. . . . . . . . . . . . . . . . . . . . . .
4.2.3.2 Evaluation of Specific Com-
ponents ......................
4.2.4 Representative Stack Sampling for Gases
4.2.5 Sampling Errors in Ducts..............
4.2.6 Variance of Concentration in Large
Du c ts .................................
4.3 Flue Gas Flow Measurement Methods ............
4.3.1 Introduction ..........................
4.3.2 Pi tot Tube Traverse...................
4.3.2.1
4.3.2.2
Velocity Measurements ........
T ra ve rs e .....................
4.3.2.2.1
Equal Area Division
of Ducts of Circu-
lar Cross Section.
Equal Area Division
of Ducts of Rectang-
ular Cross Section
Flow Variations
During Traverse ...
4.3.2.2.2
4.3.2.2.3
4.3.2.3 Calculations of Volumetric Flow
From Velocity Pressure .......
4.3.3 Low Velocity Flow Measurement .........
4.3.3.1
4.3.3.2
4.3.3.3
4.3.3.4
4.3.3.5
Micromanometer - Pi tot Tube ..
Electronic Micromanometers ...
Hot-Wire Anemometer ..........
Vane Anemometers.............
Plate Orifice Meter ..........
4.3.4 Comparison of Volumetric Flow Measuring
Methods. . . . . . . . . . . . . . . . . . . . . . . . .
4.3.5 Indirect Determinations of Volumet~i~"
Flow
......... .........................
4.3.5.1 Stoichiometric Method ........
4.3.5.2 Dilution Technique ...........
viii
Page
30
30
36
40
42
44
47
47
50
52
55
56
58
61
62
63
63
63
65
65
66
68
68
68
68
WALDEN RESEARCH CORPORATION
-------
Section
5
TABLE OF CONTENTS (Cant.)
Titl e
Sampling Methods for Sulfur Oxides
..................
5. 1 Introducti on ...................................
5.2 Probes. . . . . . . . .. . . . . . . . . . . . . . . . . . . .. .. .. . . . . . ..
5.2.1 Simple Probes ...........................
5.2.2 Temperature Controlled Probes ...........
5.2.2. 1 Hea ted .........................
5.2.2.2 Water-Cooled ...................
5.3 Collection Methods for S03 .....................
5.3.1 Introduction ............................
5.3.2 Absorption in a Liquid ..................
5.3.2.1
5.3.2.2
5.3.2.3
5.3.2.4
80% Isopropanol................
Boiling Water ..................
Aqueous Sodium Hydroxide .......
Distilled Water ................
5.3.3 Condensation Methods ....................
5.3.3.1 Controlled Condensation........
5.3.3.2 Uncontrolled Condensation ......
5.3.3.3 Dewpoint Method................
5.4 Collection Methods for S02 .....................
5.4.1 Introduction ............................
5.4.2 Aqueous Absorbents ......................
Page
71
71
71
72
74
74
74
74
74
78
78
87
87
89
89
91
94
94
95
95
96
Hydrogen Peroxi de .............. 96
Sodi urn Hydroxi de ............... 97
lodi ne ......................... 99
Sodium Tetrachloromercurate .... 101
Stannous Chloride.............. 103
Sodium Acetate ................. 104
Collection Efficiency of
Aqueous Scrubbers.............. 105
5.4.3 Solid Adsorbents ........................ 112
5.4.2.1
5.4.2.2
5.4.2.3
5.4.2.4
5.4.2.5
5.4.2.6
5.4.2.7
5.4.3.1 Silica Gel.....................
5.4.3.2 Adsorption on Molecular Sieve or
lon-Exchange Resins............
5.4.3.3 Reactive Solid Sorbents ........
5.4.4 Dilution Methods ........................
ix
112
116
118
125
WALDEN RESEARCH CORPORATION
-------
TABLE OF CONTENTS (Cont.)
Section
Ti tl e
Page
5.5 Comparison of Analytical
ft1etho ds ..............
126
5.5. 1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . 126
5.5.2 Analytical Methods for the Determination
of Sulfate Ion or Sulfuric Acid ........ 127
5.5.2.1
5.5.2.2
5.5.2.3
5.5.2.4
5.5.2.5
5.5.2.6
5.5.2.7
5.5.2.8
5.5.2.9
Titration with Barium Ion and
Selected Indicators........... 127
Titration with Barium Ion Using
Other Substituted Naphthalene
Disulfonic Acid Derivatives ... 128
Titration with NaOH ........... 130
Colorimetry Using Barium Chlor-
ani 1 ate ....................... 131
Turbidimetry.................. 132
Conductivity.................. 133
Precipitation Reactions ....... 134
Ti1:r_ations InvoJving Ethyl-
enediaminetetraacetic Acid .... 135
Summary and Conclusions ....... 136
5.5.3 Analytical Methods for the Measurement of
Sul fi te Ion Concentrati on .............. 137
5.5.3.1
5.5.3.2
5.5.3.3
5.5.3.4
5.5.3.5
Iodine Titration by Decoloriza-
tion of a Standard Iodine Solu-
t ion. . . . . . . . . . . . . . . . . . . . . . . . . .
137
Iodine Titration of NaHS03
Solution ...................... 137
West-Gaeke Colorimetric Det-
enni nati on of S02 ............. 138
Acid Titration - NaHS03 Solu-
tion .......................... 139
Conclusion .................... 139
5.5.4 Laboratory Investigation of Analytical
Methods for Determination of Sulfate
Ion. . . . . . . . . . . . . . . . . . . . . . . . . . .' . . . . . . . . .
5.5.4.1
5.5.4.2
140
Titration of Sulfate Solutions
wtth Barium Perchlorate ....... 140
Determination of Sulfate by the
Barium Chloranilate Method "" 145
---~-----
Statistics of-FTeld Sa-nlplingand Analysis...~...... 154
'6-.lIntroduction """"""'''''''''''''''''''''' 154
6
x
WALD'EN RESEARCH CORPORATION
-------
Secti on
7
Appendix 1
Appendix 2
Appendix 3
Appendix 4
Appendix 5
Appendix 6
TABLE OF CONTENTS (Cont.)
Title
6. 1 . 1 Accuracy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6. 1 .2 Preci s i on [[[
6.2 Precision Obtainable in Field Sampling of Sulfur
Ox; des.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..
6.3 Comparison of Sampling Errors and Analytical Er-
rors in the Determination of Sulfur Oxides .....
6.4 Comparison of Stoichiometric to Measured S02
Fi e 1 d Res u 1 ts [[[
6.5 Conclusions ....................................
Recornrnended Methods[[[
7.1 Coll.ection Methods for S03 .....................
7.2 Collection Methods for S02 .....................
7.3 Analytical Methods .............................
7.4 Simp 1 i fi ed Methods.............................
7.5 Correlation of Combustion Source Type with Se-
lection of Sampling and Analysis Techniques ....
Literature Ci ted [[[
Reference Data on the Combustion Effluent Environ-
n1en t [[[
-------
Figure No.
2-1
3-1
3-2
3-3
3-4
3-5
3-6
3-7
4-1
4-2
4-3
4-4
4-5
4-6
4-7
4-8
4-9
4-10
4-11
4-12
4-13
4-14
5-1
5-2
5-3
5-4
5-5
5-6
5-7
LIST OF ILLUSTRATIONS
Caption
Schematic of Stationary Combustion Systems ..........
Equilibrium Constants for Formation of Sulfur Oxides
Equilibrium Concentrations of S02 and S03 vs. Temp-
erature at Atmospheric Pressure .....................
The Vapor Pressure of S03 ...........................
The H2S04 = H20 + S03 Equilibrium ...................
The Equilibrium Constant for the Reaction (1) vs.
Temperature. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Fraction of Total SOx ... Oxidized to S03 ...........
Adsorption of S02 on Silica Gel.....................
Observed Variations of Flow in Ducts ................
Samp 1 i ng System Componen ts ..........................
Probability of Obtaining an Accuracy within 15% of
9 Point Analysis for 02 in a Large Duct .............
Schematic of Large Combustion Unit ..................
Normalized CO Traverse Data at Dust Collector of
Coal-Fired Pl~nts ...................................
A Standard Pi tot-Stati c Tube........................
Effect of Yaw on Pitot-Static Combinations ..........
S taubschei be Pi tot Tube.............................
Calibration Factors, K, for Staubscheibe Pitot Tube.
Tangential Method for Duct Division .................
Types of Asymmetric Velocity Distribution in Pipes ..
Traverse Plan for Rectangular Duct ..................
NACA Mi cromanometer .................................
Vane Anemometers............................."......
Sulfuric Acid Dewpoint as a Function of S03 Concen-
t ra t ion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Heated and Cool ed Probes............................
Co 11 ecti on Devi ces for S03 ..........................
Efficiency of S03 Collection........................
Collection Efficiency of Midget Impingers ...........
S03 Controlled Condensation Apparatus ...............
Aci d Deposi ti on Probe...............................
xiii
Page
6
14
16
17
18
20
20
23
31
32
43
45
46
51
54
55
55
57
59
61
64
66
73
75
80
86
88
92
95
WALDEN RESEARCH CORPORATION
-------
Fi gure No.
5-8
5-9
5-10
6-1
A4-1
A4-2
A4-3
A5-1
A5-2
A5-3
LIST OF ILLUSTRATIONS (Cont.)
Capti on
Scrubber Configurations .............................
Collection Efficiency of a Single Midget Impinger for
502 .................................................
Comparison of Detector Tube and Peroxide Collection
Ana 1 ys is for S02 .... 8. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Measured S02 (PHS-Shell Method), ppm ................
Probe Modul e ........................................
Co 11 ector Modul e ....................................
Control ft10dule ......................................
Simple Dilution System for High Concentrations of S02
and NO ..............................................
Di 1 ut ion Sys tern. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
503 De 1 ; ve ry Sys tern. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
xiv
Page
106
111
115
166
222
224
225
230
231
232
WALDEN RESEARCH CORPORATION
-------
Table No.
1-1
1-2
2-1
2-2
2-3
3-1
3-2
3-3
4-1
4-2
4-3
4-4
4-5
4-6
4-7
4-8
5-1
5-2
5-3
5-4
5-5
5-6
5-7
5-8
5-9
5-10
5-11
LIST OF TABLES
Ti t1 e
Page
Nationwide Emissions of S02' 1966 ................... 2
Nationwide Emissions of S02' 1967 ................... 3
Size Categories for Stationary Combustion Sources ... 6
Emission Parameters for Fossil Fuel Combustion ...... 8
50 Con t ro 1 Sys terns. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
x
Adsorption of S02 on Carbon Black ................... 22
501 ubi 1 i ty of S02 ................................... 24
Acid Dissociation Constants ......................... 25
Comparison of Satisfactory Gas Meters for SOx Sampling 38
Observed Coefficient of Variation for C02 Traverse at
Various Samp1 ing Locations.......................... 48
Observed Coefficient of Variation for C02 Traverse
for Various P1 ant Types.............................
Pi tot Static Factors at Low Air Speeds ..............
Location of Measuring Points for Log-Linear Method ..
Test Points for Rectangular Ducts ...................
Diameter Ratios for Vane Anemometers ................
Comparison of Volumetric Flow Measuring Methods .....
Typical Power Plant Sampling Points .................
Comparative Summary of Sampling Devices .............
Collection Efficiency of S03 in 80% IPA .............
Collection Efficiency of Midget Impingers for S03 ...
Collection Efficiency of Midget Bubblers for S03 ....
Collection Efficiency of Lamp Sulfur Absorber for S03
Comparison of S03 Analyses as a Function of Collection
Method. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
SO Collection Efficiency Using Controlled Condensa-
t i ~n ................................................ 93
Efficiencies of Absorbing Devices and Reagents for
502 ................................................. 1 07
49
53
58
60
67
69
72
77
79
82
84
85
Collection Efficiencies of Single Midget Impingers
fo r S02 ............................................. 1 09
Collection Efficiency of Midget Impingers for S02 in
Pe ro xi de ............................................ 11 0
xv
WALDEN RESEARCH CORPORATION
-------
ACKNOWLEDGMENTS
We are pleased to acknowledge the assi stance of Robert L.larkin,
the project leader for NAPCA. We thank Mr. larkin and James Dorsey for
sharing their experience with the project staff. We are indebted to
Matt Stengel and Jim Martin of Combustion Engineering and John F. Smith
of the Bureau of Mines for making available unpublished flue gas analy-
sis data. We thank the following for their active interest and coopera-
tion:
D. Barnhardt and J. Moore
H. Chapman
E. Diehl and R. Young
F. Gottlich
J. Hultz, H. lang,
W.OINeil
A. Plumley. W. Taylor
J. Perci val
R. Stevens, W. Smith
F. Gartrell
Babcock and Wilcox
Bethlehem Steel
Bituminous Coal Res.
Boston Edison
Bureau of Mines
Combustion Engineering
Esso Res. & Engineering
NAPCA-Durham
Tennessee Valley Authority
We thank Professors Paul Giever (Univ. of Michigan) and Melvin First
(Harvard Univ.) and Alfred Thompson of Riley Stoker Corp. for valuable
consultations. We also thank Mr. Cirvillo of Air Pollution Technical In-
formation Center for the literature search conducted at our request.
xvii
WALDEN RESEARCH CORPORATION
-------
PROJECT STAFF
Major technical contributions to this report have been made by the
.followihg:
W~lden Research Corporation
J. Becker
A. Berger
K. Brummer
A. Doyle
J. Driscoll
J. McCoy
P. Morgenstern
ArfhurD. Little, Inc.
J. Funkhouser
M. Scofield
C. Summers
A. Whittier
xix
WALDEN RESEARCH CORPORATION
-------
1.
INTRODUCTION
This is Volume I of three volumes of the Final Report on "Improved
Chemical Methods for Sampling and Analysis of Gaseous Pollutants from
the Combustion of Fossil Fuels (in Stationary Sources). II The three
volumes are divided by subject as follows:
I Sulfur Oxides
II Nitrogen Oxides
III Carbon Monoxide
This report was prepared under Contract No. CPA 22-69-95 covering
the period June 12, 1969 to September 11,1970. The principal objec-
tives of this program were: (1) to review the state-of-the-art methods
for manual (wet chemical) methods for the determination of emissions of
specified gaseous pollutants from (indirect-fired) stationary fossil
fuel sources, (2) to select and develop improved procedures for high pre-
cision emissions determinations, and (3) to select and develop simplified
procedures where high precision determinations were judged to be complex
(i.e., required a highly trained research and development group).
The determination of sulfur dioxide (10-3000 ppm) and sulfur trioxide
(5 to 10-300 ppm in the gas phase under stack conditions) is reviewed for
effluent (or stream) temperatures up to ca. 900°F for stationary fossil
fuel combustion sources and for control equipment which may be installed
for such sources. Laboratory studies of the most promising methods are
also included.
Since the practical utility of an analytical method is limited by the
fundamental problem of obtaining a representative sample, we have neces-
sarily included a discussion of the principles of flue sampling and the
limitations so introduced to the overall accuracy and precision of a
given determination.
In the following subsections, we consider method development prior-
ities based upon national SOx emissions, summarize the state-of-the-art
in manual determination of sulfur oxides, and describe laboratory studies
of the barium chloranilate and barium ion-thorin t~tration procedures.
1
WALDEN RESEARCH CORPORATION
-------
1.1
National SO Emissions and Method Application Priorities
x
A rough guide to method development priorities may be
from the relative contributions of various sources to national
sions (Table 1-1).
obtained
SO emis-
x
TABLE 1-1
*
NATIONWIDE EMISSIONS OF SOx' 1966 (521)
106 tons/year
Total from Fossil Fuel Combustion in Stationary
Sources
22. 1
By Fuel
Coal
Residual Fuel Oil
Distillate Fuel Oil
18.3
3.3
0.5
By Sector
Power Generationa
Industrial
Residential and Commercial Heating
14.0
5.6
2.5
aAbout 90% of the power generation contribution to SOx emission is de-
rived from coal-fired units and about 10% from oil (522).
Since SO emissions ~re primarily fuel dependent (518) rather
x
than equipment dependent, the data of Table 1-1 essentially integrate
the product of fuel use by sulfur content of the fuel. The above do not,
however, distinguish fossil fuel effluents mixed with process emissions.
A somewhat more detailed breakdown, which excludes (direct-fired)
process emissions, is available from work conducted at Walden (480), and
is given in Table 1-2.
As may be seen from the Boiler Category of Table 1-2, 90% of
total (non-process) emissions are attributable to watertube boilers,
for which the smallest size range if about 25,000 pounds of steam per
*Literature references are listed at the end of the report.
2
WALDEN RESEARCH CORPORATIO~
-------
TABLE 1-2
NATIONWIDE EMISSIONS OF S02' 1967
*
(Excluding Process Emissions) (480)
106 tons/year
Total from Fossil Fuel Combustion in Stationary
Sources
18.63
By Fue 1
Coal
Residual Fuel Oil
Distillate Fuel Oil
By Sector
Utilities
Industrial
Commerci al
Residential
15.60
2.76
.27
12.58
4.28
1. 53
.24
By Boiler Category (excluding residential)
Watertube ~500,000 pph
Watertube ~500,000 pph
Firetube
Cast Iron
9.29
6.50
.88
.72
*
Also excludes stationary turbines and reciprocating internal combustion
stationary engines.
3
WALDEN RESEARCH CORPORATION
-------
hour (or ca. 3 x 107 Btu/hr). The S02 emissions determination problem is
centered in these relatively large sources. Further examination of the
categories of Tables 1-1 and 1-2 reveals that the highest priority is
clearly to be placed upon analytical methods and equipment applicable to
large boilers, both as the largest sources, and in consequence of the
probable first application of control techniques to these sources.
It does not appear desirable to develop specially adapted high
precision methods applicable to the myriad and relatively unimportant
small sources at this time.
In the following section, we describe the effluent compositions
from fossil fuel fired stationary combustion sources. The emphasis is
upon those factors which influence the selection of sampling and analyti-
cal methods for sulfur oxides.
4
WALDEN RESEARCH CORPORATION
-------
THE COMBUSTION EFFLUENT ENVIRONMENT
2.
2. 1
Introduction
A prerequisite for the selection of manual chemical methods
for sampling and analysis of gaseous pollutants from the combustion of
fossil fuels is the description of the environments to be sampled. The
combustion products from all stationary sources such as steam generators,
process heaters, water heaters and air (space) heaters are to be con-
sidered, except for those cases where process emissions are combined
with the combustion products, e.g., a lime kiln or blast furnace. The
measurements will normally be required at stack conditions; however,
they may also include conditions at the inlet and outlet of sulfur ox-
ides control equipment.
2.2 General
All combustion equipment can be generalized in terms of the
schematic shown in Figure 2-1. Fuel is burned in a furnace to release
heat. Combustion products and ash are generated as undesirable by-
products. The combustion heat is transferred in a heat exchanger to
some convenient thermal fluid, usually water, steam or air. Domestic
heaters are familiar, simple examples of this generalized system. They
are designed to provide unattended service over long periods of time
to untrained users. Consequently, they are characterized by a low
level of sophistication in combustion control, heat transfer (thermal
efficiency) or air pollution control. On the other end ~f the sophisti-
cation spectrum are the new steam generating units at power plants.
While the generalized combustion system shown in Figure 2-1 is still
applicable, it is embellished considerably by combustion air pre-
heaters, economizers, high pressure heat exchangers and pollution con-
trol equipment. In addition, computerized systems are utilized to
monitor and control furnace and heat exchanger operation. The combus-
tion products may be modified by the use of additives, either in the
fuel, or added in the furnace or heat exchangers.
5
WALDEN RESEARCH CORPORATION
-------
Useful Heat
Fuel
Furnace
Heat
Exchanger
Coni> us ti on
Products
Air
Ash
Cool
Therma 1 Fl ui,d
Figure 2-1. Schematic of Stationary Combustion Systems
For this evaluation, we have divided combustion equipment into
three arbitrary size groups as shown in Table 2-1 in several approxi-
mately equivalent ways. There are relatively few suppliers of equip-
ment in the large size range (4 major sources) with 1968 annual sales
TABLE 2-1
SIZE CATEGORIES FOR STATIONARY COMBUSTION SOURCES
Large Intermediate Small
BTU Per Hour >5xl08 3xl05-5xl08 <3xl05
Boiler Horsepower ~15,000 10-15,000 <10
Pounds of Steam
Per Hour >500,000 350-500,000 <330
Megawatts >50 <50MW Not Applicable
of the order of 75 units (556). On the other hand, there are several
hundred suppliers of small units with 1967 annual shipments of approxi-
mately 1,500,000 units (557). Useful life is perhaps 15 and 30 years,
respectively, for the small and large units. Combining the large
number of installed units with the disparate designs, range of fuels
used, load factors, varying operating practices, etc., it is clear
6
WALDEN RESEARCH CORPORATioN
-------
that it is totally impractical to attempt to specify the environment
for individual cases. Rather, we have reviewed the range of environ-
ments reported in the literature for each of the three size categories
shown in Table 2-1 for oil, coal and gas. A summary of emission param-
eters is given in Table 2-2 in consistent units (#/106 BTU). Details
are given in Appendix 1. A summary of conditions before and after
control equipment for the most prominent SO control systems is given
x
in Table 2-3. From these data the various requirements for principal
factors such as required sensitivity, possible interferences, transients
and temperature regimes may be determined. The major factors are briefly
described below.
2.3 Temperatures
Sampling equipment must be designed to tolerate Inorma1" op-
eration as well as unusual temperature excursions. In general, large
units are designed for stack temperatures of 300-4000F. Gas-fired unit
emissions, as a result of their freedom from the effects of corrosive
sulfuric acid mist, run cooler, about 27SoF. In domestic units stack
temperatures are much higher, to 700-8000F. Inlet temperatures of the
Monsanto SO control process also run at this level. Intermediate com-
x
bustion units cover the range of temperatures between these extremes
( 400-7000 F) .
The implications of this temperature variation for sampling
are four-fol d:
1. Materials of construction must be adequate to tolerate
the highest temperature.
2. Materials must be chosen to eliminate high temperature
catalytic oxidation of S02 to 503 in the sample probe and lines.
3. Water and acid condensation must be taken into considera-
tion in design of the sampling train. The sample stream should be
kept hot insofar as possible, up to the collection equipment.
7
WALDEN RESEARCH CORPORATION
-------
TABLE 2-2
EMISSION PARAMETERS FOR FOSSIL FUEL COMBUSTIONa (pounds/l
La rge
GASb
Intermed.
Small
OIL c
Large Intermed. Small
(Resid.) (Resid./Dist.) (Dist.)
2000-
3000
Velocity
ft/mi n
2000-
3000
Moisture
18% @ 0% XS Ai r
----- 15% @ 20% XS Air ------
10% @ 0% XS Air
----- 9% @ 20% XS Air ------
Excess Air 15 15-75 75 20 20-75 75
%
Stack Temp. 275- 400- 750- 300- 400- 750-
of (of 400 750 900 400 750 900
00 Breeching)
CO Neg. 4xl0-4 4xl0-4 3xl0-4 0.01/0.01 .01
S02 4xl0-4 4xl0-4 4xl0-4 1.65 1.65/0.31 0.31
S03 0.03 0.016/3xl0-3 3xl0-3
NO
N02 0.35 0.2 0.1 0.68 0.47/0.51 .08
~ Parti cu- 0.014 0.016 0.017 0.07 0.15/0.10 0.06
~ late (9m
o
",
z Hydrocarbons Neg. Neg. Neg. 0.022 0.01/0.01 .02
;:u 8>
",
CJ)
~ Transients 10% of Capacity Per Min. 10% of Capacity Per Min.
()
:r Details in Appendix 1, Fi na 1 Report, Sdist. 0.39%, 142,000
() a. Vol. I c.
o Sresid. 1.6%, 152,000
;:u b. 11 00 BTU/SCF
~
;:u d. S = 2.5%. 8% ash. 13.(
~
~i5
~.<.:"~
-------
TABLE 2-3
SO CONTROL SYSTEMS (553)
x
BEFORE AF"
Developer Process Temp.oF Input 0 Out
Temp. F
CE Wet Scrub- 250-300 Normal 250-300 99%
bing (Do1- Double Load cu1
omite) Particulates mov
Monsanto CAT-OX 850 Follows 99% 250
+ Elect.
Precip.
\.0
Bumines Alkalized 625 Recommend 3 250 Fo1
Alumina 0.9 grift c10
ara
Rei n1 uft Char. 250-300 Recommend 3 250
Absorption 0.9 grift
We 11 man - A1ka1i- 300 Only as 250 +95
Lord Sulfite Req'd by cu1
Absorption Power Plant mov
~ Equipment Spe
>
,... Scr
c
",
2
:;tI TV A Dry Dol omi te Additive 2-3 times
",
en Absorpti on Rate of Particulate
~ (Intermit- 2 Times Load in Flue
:;tI
(') tent) Stoic hi- Gas Stream
:x:
(') ometric and Bottom
0 Requi re- Ash
:;tI
~ ment
:;tI
~
0
2
*
Design Efficiencies Assumed
-------
4. Hot probes and hot stacks make for uncomfortable working
conditions requiring asbestos gloves. Severe manipulative constraints
are imposed in handling of delicate apparatus, small connectors, etc.
which should, therefore, be avoided.
2.4 Moisture
Combustion effluents possess a high moisture content which
introduces a number of problems.
1. Uncontrolled condensation losses in sample lines may
losses of soluble species (S02' S03).
2. Uncontrolled condensation in the sampler may lead to
dilution of reagents.
1 ea d to
3. Uncontrolled condensation can confuse the determination
of water content of the stack gases. Determination of water is re-
quired whether results are reported on a "dry gas'. basis or at "stack
conditions".
4. Sample train components must not corrode in the high
moisture/acid environment.
2.5 Particulate Matter
The objective of the methods under discussion is to det-
ermine gaseous contaminants. However, particulate matter can be an
impediment and should be removed from the sample stream prior to the
collection of gaseous S02 and S03 for several reasons.
1.
Material adsorbed on particles will falsify the results.
2. Fly ash can catalyze the oxidation of S02 to S03 [even
in dilute solutions, (527)], thereby distorting the true conditions.
3. Acid mist should be quantitatively removed from a
s~mPle stream if one is to determine gaseous S03. Heated sampling
llnes are then used to prevent further condensation of gaseous S03.
10
WALDEN RESEARCH CORPORATIQ~
-------
The uncertainty involved in collecting 503 mist and not re-evaporating
it is great in a heated sampling system. Consequently, the preferred
method is to remove the acid mist in a prefilter heated well above the
dewpoint and to maintain such a temperature through sampling lines to
the collector.
Obviously, the determination of 503 associated with fly ash
or soot is ambiguous using this procedure and contributes to the con-
fusion in S03 values reported in the literature. Particulate matter
interference with S03 determination in the important large coal-fired
units may be a very serious problem because of the high particulate
loading.
Particulate matter from residual and distillate oil-firing
is substantially smaller. Although the sulfur content of the residual
oils can be significant, the analysis problem is less severe than that
for coal. Particulate emissions from gas fired units are very low, as
is sulfur content of the effluent gases.
2.6 Sulfur Compounds
Sulfur compound emissions are insignificant for most gas-
fired units but substantial for oil and coal. Sulfur compounds in the
combustion effluent from coal are ca 104 higher than for comparable
gas-fired units. The chemistry of the sulfur compounds in flue gases
is discussed in Section 3. Typical values and their ranges for the
major fuels are discussed in detail in Section 6.
2.7 Other Gaseous Species
Other gaseous species, both major (C02' 02) and minor (NOx'
CO, HC1, etc.) are of interest, here, specifically as possible inter-
ferences in methods for the determination of sulfur compounds. Since
these effects are specific, they are discussed in detail in Section 5
for each method considered.
11
WALDEN RESEARCH CORPORATION
-------
2.8 Transients
Description of a transient requires the ability to sample for
several short intervals over the duration of the transient. Transients
related to load changes can be expected to vary from 1 or 2 minutes to
approximately 30 minutes. For convenience, power plants will usually
go from no to full-load in about one-half hour and will make smaller
load changes (say, 100 + 70%) over about 15 minutes. However, new
boilers are designed for fast load changes with gas and oil-fired units
capable of 10% of full capacity per minute [10 minutes to full load
(555)]. Because of greater difficulty in fuel transfer, metering and
valving, coal-fired units can only be designed for about one half this
load change rate (20 minutes to full load).
Domestic oil-fired units generally do not modulate the fuel
flow but go from "off" to "on". Steady state combusti on based upon
soot emission takes about 10 minutes (512). This period is precisely
the average "on" time for a domestic oil furnace (512). Consequently.
domestic units are essentially always operating in a transient mode.
Manual methods for sulfur compounds described in the fol-
lowing are best suited for continuous sampling and are, therefore,
averaging techniques of limited time resolution. Although it is cer-
tainly feasible to increase sampling rates to increase time resolution,
say to the order of five minutes, transients are clearly best de-
scribed by fast response instrumental methods.
12
WALDEN RESEARCH CORPORATION
-------
CHEMISTRY OF SULFUR OXIDES IN FLUE GASES
3.
3.1
Introduction
Sulfur, contained as an impurity in fuel, is converted nearly
quantitatively. during the combustion process, into the two stable
oxides of sulfur, S02 and S03. About 98% of the effluent sulfur is
released as S02 and the remainder as S03. S03 is a local pollutant;
as it leaves the stack, it reacts with water vapor to form an acid
mist of dilute H2S04 droplets. This mist may fall near the plant,
causing corrosion of metal surfaces. Some of the S03 may leave the
stack as "acid smut". This material is fonned by S03 condensation on
cool spots inside the stack (below about 2800F) and adsorption onto
flakes of carbon, which are deposited simultaneously. Where growth
occurs on metal surfaces, corrosion produces ferric sulfate which is
incorporated into the carbon flake.
In the following we briefly summarize the properties of the
sulfur oxides most significant for flue gas sampling: equilibria,
vapor pressures, gas phase oxidation equilibria and kinetics, adsorp-
tion properties of S02' and physical chemistry of aqueous solutions
of S02. Finally, we discuss conclusions, bearing upon the analysis
program, which may be drawn from the data presented.
3.2 Properties of S02 and 503
3.2.1
Equilibrium Constants for Formation of 502 and 503
From the Elements
The equilibrium constants, Kp, for the formation of
S02 and S03 from sulfur and oxygen are shown as a function of temp-
erature in Figure 3-1 (488) where,
K =
PSO
2
PSO
2
Po .PS
2
, and
K =
PSO
3
PSO
3
(P )3/2 P
02 S
13
WALDEN RESEARCH CORPORATION
-------
100
o
90
80
70
60
50
Q.
~
C'1
0
r-
40
o
o
30
~4:j.Lfqs~?: .!::~~ ;~;:l~:f t~: ,;;~;~: ~~~ ~.. ;':'I-ii mtm'H~. ' . , ,fiti!;' ~:t: Jl~r 11~'"r[, ,tij
1+:+ .tl:'1 j , ri~ 'I .'1+ t ,- .~~-. .1. -,""'0> ~.... ~-+ -t.+>--.:-, i-I" .-.: -:- ~' ~'~- ,. LH- . .
j[1!TTljt:TtUT1~7~,':J:~i ::1~dl'LHh(S :!ii~ ~lt,~-:-, ~~;I;i tf~,.2:c.1 ,.- .,' ~ Ii" ", ,. ,,' if1, "~1fi}l{ " ,i
.I." '+, ,.1 ! T I.J "-.' ....t-r-./.-.+-ij-I-I-~-+!1..,-t ,'''- -'M r,...' +-+--t-t- ,..-t- ...--',-++-j tr--" - -r ----, ~r- + ,.,. .
'" '~r' "'''''''''''1--(:;'''' ""--~~n::+c ,:!:t; ~"i~r~j:;,,~dlili'~'H t:Jj:'-',"'''h}>- ,'d' ,IH+ ;, J +', t
Jlti ,t~; 'jl; d:-I ~-.~i ','.: ~~~~1..!..Jt~::~f;_;j:~-q t' :---- ..;-Ltt:I ,~~~!:~:~~ I-ttr::tF~tp:-rV;-+1-d-...~-rt1:i+ r+ . -: -ttrJ-ttl- -t+, +-tf -J.l
. ~-H '~+ ~~~: ~~!~ ,~. '. :_:.,~I~~I '~lT~~__~ h-~ -~'- :'~-r :~.:- ~+o-:~~~ ~ ~-~~: ~ + ~,:-~ :.:, ~.t~~-:-j t ~ ~ 1 t . 1- . .~..,+ '-f ~ : I. ~- t
Ii I"H",L"I~'Lf;!1:" 1.11 q::t,:iti::tb"r' 1, t "-i::t: m"~" :TE: 'H+;:cq i'IH'" "-'-H :011> It-- 1:t ".1- 8+ Ht ; t' ",
~-<-~. '~~ +fl4~:r.-;t~;rl-\I~'~ ~nt-~.rt!ti+i.~~T;t rt ++: +-r ~ ;..lrt.:i:~ ;'I-U-~-'h+ ,:.:.ji:.ti~r-,~p:tti-t J.."'; }if_+- ti:, 1 -.t+-~-..t~-~1--" - - il.!li.
Hi 11 r:-fW'J-rztiJ ;+;:.::t!J1JI.:I}~pjJltr;.;.i~r+--,+tI:ffit tj:H~ r'+i' H .;:tlerr.;:~-++ =B:i::trt,~-r'~'f":!t;==J: ++ t!
+t , ii !ntl) 'i L~I;njt' !i~:: ~ ~tl ;511~ "4'~"4-[l ~ 1-,., 141, tt "~!'J";+': 1:;.:': -" .,., ~tLt': ,,,,+'!TU .I+c :)4 :tf
j--< j." .t'I:J,---~-:-"1-'-t.J..- -,I-j- ."'T1J.~~---+~"t'1J..:t:til .--.....-~I-t-t+- "',,'.'+'," "".Llli-r.J.:,' ,~-T{:;:".,i" ,>r:Hi ,,"'ftt,,!1R:~,~'~ !I"j~F" }>! +: t JL i + ~,'!tli, u-' I-
[ll!! Ii 'It ,:" ;/'..: ;:~ 1I;-1;p: l:-!~: .L-tl.~:-tt--ll .l1_i:,~I~~T ,~-1-t~rlt~::.p~ tr-;~U I-:.ttt.- tL.. ", ,tt-. -,r;:j:I.ttitp ~ ~.
,'i il,"i 1,1 '" ,,': I'" Ji"' ';d '" ,., '''1 r::i" ' , i: t'" ,,', ,...; t'., 1.~" 1=\' '!'t' r!::r 'jJ:r' ., + t 'H tj-- t
~m.' 'til.~~~; H:I ~:~I i} 11?-lli',::iH: ~I!="1. Gf'j-t!, 1H" ,lLiW 8 :itiHf:! it11j :!1£ lit t ] "k.in .1 0" I",
j11111 Iitt ,li Jrtii:~j 1:;:Ltb: :>;ii~j c,;;j:!.~i :tIi j1'tj h -i'dr ;:f!'!:Hi.IJ~r: LH';:it 'UEr 'p; tl'; L, 't, 'j-tliW1tttt! "i,,:I:i','
, .,'~," 1:1", ,JI,H:ttl '1- "j, --I,lt_t:~ :.-TI -~t,tr;..L -.-1-,-+,+,UiJ -, - j-t ttt:-n~t ......-r1- - - 1-:r tt.t-it- -t!('t-t.tf:t-:t-t ':1:.f .
'lillil' ,I Y i IN it! H,,: !L; ~m H: 'I;:~", ;;d{t. :-4-:.t~i m J:i:;t~fLd~1~;},131 ...'jitHwtli ;gn'mt rt. 'f t :fiH., i, if't ttn ticl J 1,11 , tt1
,I ill: '~~'Trr1~'~1.;t\.~'I:::: ~rfr i~!; ;;;; }t~f. ~~tjt Imi ;:~ ti:#ff~tm ~!1rtW~ mi ~ittt*f 11 ~"~4!j;U1ii 'ilr] mmt,;+ ~'~i W:.
Jj ,::::->.;.. ,":bHt'-:I'~I~!~-':~:~-'~'.gJ+.-~ .~I--l-:,--:q,~.:-tt...._~~:-~~-:',. +1--0- ~:~_I":;'+~"L~-'~ ~I ---~ J-+- ~". t- 1-.L4+1+~ - - ~4,
I! I' ::JJ1' \J :, . I:; 11 r~t;- 'Lih;:: D ' ill::,: 1irl 'i,!~;tHt!E~ :::~ E1r I:~H;H JJ:H2,~t.t.lUH111~L t" Jb 'r; EtH!tTftln ,tlHtJ:ml if
IItr[ Fil'::nl:!\ IT'1:~if;:~': ;tt~~T~' nplU tid nH~W;,;HfJnffi T~jl H:tt~l-tr tdl( -ITi..t,. ~~ 1t~ttHfliNid1:' Hi.:
-~frl~"ii';:~':'I',"~c:';;::t~'lC;'i~~l:tl ;"~"~;:Bti',:~~q ,I' '.tt ":;";:,;r;~, ',',' .' ~~, U,,~~J.':: t;~" '.'
LI JIj~:, ~~~l::~ "~T"\ : :!.; :1.:.' 'JJ', ht,tlV., tj ill11'lJ.,f,' ,t':'.~,.JIH!J..L.'tltlli,lt..,,: IU ~j,Im:i:i+ffift,.t' i~lnL i:J!,l t ti, 1;ti. 1-4-1,' jJ~<~n~}:l hk-:
Ltll\TI I";I'II:~. :1':' 0;; -n;:i-r~:~I'-..:t:;;:!~~L...:tJ~ - -Ll1~~ :r7;- _.:-.:~_:~d.::-J-l;-o-.o-I-""-"~ y.. I -. . -, t-4 ~iT" I-L..j :+H ,.~ JOT 4 -,
'Iii 'rm:1i'~F rl ,""\;" :~,;"q0tt'''1i:r;:''It.i'i' ;.][..1."" r,.J._d ;:j,,:,tt :U::rl';t-.Litl1tll' '-i'I-!-i-I++i1 ~jJtj"::ttr't":J::!""'i1H '
!' ; ,~!' ::" ,-tttqi ,.l.T ;:~ !-,: ,:;: ;;~ ~:,:'; U'L- ,;.; ~~~1 ;:\i ;:: ~::i ~~; '~:.; i,::" ..:..' + ',. "..:' ';' :: I~':d~ :1-:': t,\:
tPW ,:, ;r~"; 1 :'&~~ I':. ',4';; ,-it'. ;;[) :1:1 \:8': .Jj1 'lJ''ttj, 1';.u :i~jj-HtH;H i' i. i~J ~inf}f fmll: HU .;
, ifnP 1[/: i en: > r~!l;1i'£::I' ;: ;Ir~ i+tf ; ii n~ ti~; d~mfitj~: 1I,~-tf lr,iffj;~~'1t' , ",4 j:;jli' f;'~;~:M:i ~~!, i1~
,1\ t:,: ijt:: , i,: i::' :!~~;:.~: ;;: :';; ;;~>"It 1 ~:l: :[!; 1"~ ;::~hf" ~ ~ ~.~ ;:J.L,...t " ,,','; ':~i' \t!J~ ~'
, Ii i i I Iii j i:: l'l :,' ft~ :.I..~ ,::: ; "j;~ J~+tJ:\ ~LJii; :::~h:::t rr:!tW1 fBl.ffffi,+jI:..;tp' ,1f)'td t;jt " . '~f11'4' ,t' ", +
Ht+t':'-I~ :7:-: ,.p-, ~:--: +~~'II~-o-.~~~::.~~~ ..+ - : ., ~;. .1'- . t- -'-+-1 ~ J--.-~ l-l. -d.->..- ... . j H- I
iU!;Jl:;;iJ..':il.:.'::I.ri:,!l::j:i'-<:,:~~;:tl;:h11~:Rti~;f;:I;!;;~r1~::-t~,;t~.g.:j;t!!.~,lli~ ft.t, :t!., " ' .' "£' iliitJ:t ttj: .'
HmTT2(r,'...l:..JI-;:'~' -1~1;~~.~- 'I'~:'~: t-:~.-t1':i-'."-JT,":.:~~_:~:'~;;:-l.;;-'t-t~;-d-"'-~~J+"-.: ~:~~~ . ":_' .......... .. - -,: '~,--, ffiT"
4 :1", :, :"'::'1,: ;:~1 i,"';'''' :, Ii: !:ttlUi ~r."" ~:.. ~;!~ ,:lliti m}~!l;.t!~t. tJ:j~F->:-- ql,,,: ' ,ti ~ j+ ,t..;t J" ,r1l ,'£-
i,'! 1!;-;r7;;:,~::I,':J:1-afL<:r Ii,"cl~~"'n :":jl:~~ir,=-~:+,td-i7""';:,.,:£ ~t1'l~W''''~ ;t, 'nj,.., '11 ,t'li'lEft iUti+
I: "I: I", "'.~':.;' ,:1,;1,;,. ,;:1:::, :I~; N: p.:,l ,,:i'Ji. :,: ::,,' l~lt1rtjliHit1 ;r:-!-i,:iiii1H i.j:.j:Htt,rJ:' It .J:itHt:J.", tB ~j:il
,; 1f1jj'~I"'-~'-;-+"':' j-<'I_~H-q~_l""~~I'\1~'~I"-I:--<"~:>':~~'.j.'~_.l.-~--i-- ,--'+-~+-~~"'~ of..-.. --:' --.~' "-m't.-n~ht j-.~t.U--;-g: j
, i Iii! ii'J'!!' ;;'11";:' iU: r;,; 11'1.'- ::,t '~l'.lfi~~J ~.' ':i4,"";~' , .i.. t.Lrhti, :-!t,i:! ,~j~nl.t: '~.h. +'. ' .\ , , ,,'
1- ~I' 'fT<~~~7tPr:-t:'~~i ~~;: -~:-~ r~~+~ ;~fu-fB-41 r!-:'- '~fiT-~;Ji;..:f:h:tt~-H~~~.~~thP ~+-:- -J, '. 111 -.. .: - -j
~ . i i i i ; ~ :h:. i fli !'! n:rjJt~r ;-~~~ ;::: H;; \:~~lD tT+ +~" -" ':; 1'1: 1tf.~ #Lmf jb"Pt-;: :;;;b i,tiP; + ' : 'r~1 , I! It! - , ":t I'
i ,:r i:;:I:::iii~fll:",tilf "t:bN; ': r :!.:;::- f-t'7!: 'ritii~:t'~ u..:,Jj":;:,m;i ,:t.: 't,tH:\:t~;+ ;':; 4. ""j. I; ~ " IH; :(:, "d',' ,:
Wi, H-!-:-!+; ~'. ., '~'~,I---T" ,-+" .~L ,. ,I U.' .;1:.1 Jr.. t +.' Tj:4 F-i ",i', ~, ...' ::t." ~L!i fi1.iip" 't.,;-;- ffi~, t' , tTb ". .li!:it, :ii I:L t' j. it
liii! iJlj/i:.;rt';;!tl: 1,'{1 n~n:,':~'::" ;W ;::1 ;1ii ,:t: 'tt":,,:t :J t:,; tl,~.~;tt.!.;:" j-j+i':~ '+":t ~'tH.+chl~ti" .: I:
- - ~ I:'~: ~ f: ,J,: j:, ! +:~-. ~~~~:: +.~' '~~ ~:~~ l:~-~-~: " 1': j ';-+~ -HI :t.-;;j!:1.+" R+ ,-' :It . -'1 'rr L-:fl--b:i-r :t~ ~ .H j:
If J i,::I,: r :J"0.<1',t; ,pc '\ r~il ;;11 :tJ.!rFtP' tJe' 'c: Hi; ,1:)' ;::~t.'trt"" ~~ '~+.!' ,I" 'iit t"l
~ ~-~:t'--~hl,.:t~~I.LCJ~~ii :i~;!::>::~: 1W ~;~~;i:i ~i; :~~' :bl:;;;' "i~ r::b_;11 ~.,: ~';, ,+> ."', ~I ~ rjj~ :~t ;1:b tW:1,~ . ,-
: ' ,J"" "i ~':l.',l -- '", D', ,,+ 41''';:' "+I ",.e, --en" ~i: r':~ ~,t~I "'r ~ ,d H' "n'" + " "';
i I:' i",l' ":--'+I.tt -t, =t-.~1' - Jtt~.. ;
,'r ;t't ""t:; ;1",' c, ,,,, ,:~'''I''.""" --1:tb.J-++ .:. .f'. ,T~. " '+<-,,' ~II,.: ,tt,,'!] hl,t', ~ttjj:! tf11,' j:t j 1ittl
I !. :jl. 11"i~t.f :--'-,J~~::,":I}:T; ::::~~:':..: I' t - -t-t !w.... I-t"7 ....:} 11-'----< c !.-o- -. -' ~--... . L r'; -' - p... - ..L ...
"1"';:':-: ::"'YI,~:1;,'!T:~::C:~:;1';i'=:::-Ic-c1'lt:.;j.t:-:'.1~4':;ttt:+j' ,. :..+~,.t<}l:!J:,t ',.' ,"jffi, ~'id~-1 ,tl' ,
l~rl:iliD "11 ~,'~ ~;~ ~:'"t:I~. -r-:tr:-~:-:~:'i--i-D:.F~ fit: .-::I:tit.J. -+=4-U:~~- ..... ,"" ~~~-f n-t"j - ,~.j -., ~fH ~ - -<- t
r-;tn r:~'I':; ~.-- :r:::r -:-11': ,~o-~-.oq:d ~~-~ -t..~ ...~J \ J -H J~;j ~tlt -rt1t Ti-i-.. +~ -'-tt:i.-F, t. :ttf tt:t'.l ~-r
~ll. .p.:.~ . '''r L~J "JL .. .:.--:~ht-':!~, --'-,'1" -1+' ''-'''',+ 1+tli!~l i;":;-i't1t;i 't' ffin I :H- L
.1:! .~Tf,i~~t'~t:j:'~;;J!;,~)i~~ ;t~L~~:~!.~,;.~.' ~ l~.,~ ~~jr~~ ~~i ~hl i# ~C;J~ ~Jj , , 1 ~;, I~L!' . ili thj:~~1
;;;j'rd1.,l,. T.1 "1,' :iH:-: '::.}1::." ~'~ 1:i1.t'1' -'1::+ ~:, :;::]"',"h+, j~r 'F.'-
" .L.....:..i ~:.:-/ _-:r~' ..., 'u, ."~ , , "+ =; ,=..~ '\:1-,= "t WALDEN RES
ii!~ .:!',I'", :1-:--:-f'I: :c:'-~"':; j"'~L""'" :::.:.~., ~~'::';'h ~~.. EARCH CORPORATION
2tttj~! ',~:f-l j ~.~r!:l¥~~~ ':~i; ~t ~ :~:.~' :~0.:;i,~~i, ~il:~ f~, ,-' , . ' .
~'K';;I"iT i iT, ',",' t ~'+ ~ ,1"'-q.t.:g.l:::C::;h~ ..:.T , "'"IT.+.:J',::: :::ri~:~" ,',n~i:r~r:~c ':ri~ "':'.i" Tlr., ,':'" """, 'i"~'."" " ,
!~~~ ~:, : I ~' 1 ,i .. ..-: ~'.~-' j' , J_~:~. ;~~ -..- . .~. " :c>: .~~-ub~ ~~~I--~ li'~1~0 ~..}~: ;~-it~"~! ~~~~\t: :~j-LtI-; ..lL~ ~~-:: -
,:;, "'. : 2 ~ :F 1 .~!:, :~"I' )"'.1:::; ,":'1:....,.,.: Jo~ t" '...:.i--, ';-" :.1. ' . :--",I;,~j ..~ 1:t,:::il'U ,",\-' +" . ,
.. ',' ,I " I +, I, ,,'!-:-' -+ -- . ~ '.' ,'':;'- I c' r':- '::E"C:'; ,.tr;,;:: . -,I ... ' 'i",t!" -- .~~., '
"fic, j':T"rl-. r~",. :c:,- '~C".. '"1':'+-'--:-" ::,Ai it":;-j~;:;:.rti4ltiltjtmt-
('0'; I, ':O':<".j ",1-"~' ..." I'.>::::':: ' j ; l!~\ "CLJU:~':' .
200 500 l~~~erature 1500° f-1,~
:1
:l
j
<-t
-"" -t t l-tt
t _1. ''''''
1. " +1-
~ t
:t:
. nJ
.+
ftW-tU '
-_I +t:.-
,
I-i
,
I
1
~
1
:b
+
11
1
t
20
10
o
-------
From Figure 3-1, it follows that S02 is the stable species at high temp-
erature and S03 is the stable species at low temperatures. The composi-
tion of the equilibrium mixture as a function of temperature is shown
in Figure 3-2.
3.2.2 The Vapor Pressure of S03
The vapor pressure of S03 between 00 and 1000C is shown
in Figure 3-3. Solid (a) S03 is the stable form below ca 6SoC. The
vapor pressures of the less stable solid phases (S,y) are essentially
indistinguishable from that of the liquid.
3.2.3 The S03-H20-H2S04 Equilibrium
The dewpoint of S03 varies with composition in the flue
gas, i.e., with the concentrations of S03 and H20; the usual dewpoint is
about 280oF. Observed dewpoints in a flue gas are discussed subsequently
(Section 5). The calculated concentrations of S03 and H20 in equilibrium
with pure sulfuric acid are shown in Figure 3-4.
3.3 The Oxidation of Gaseous S02
3.3.1
Homogeneous Oxidation
The oxidation of S02 to S03 has been measured in a
furnace (530). where it was found that, at 2-3% 02' the ratio of S03
to the total SO concentration is about 0.04 at 1080oC. This ratio,
x
D', is an experimental result and not necessarily the equilibrium value.
A comparison between D' and the corresponding equilibrium value, D, is
given below.
There are two principal mechanisms for the homogeneous
oxidation of S02:
S02 + 1/2 02 + S03
(1)
S02 + ° + S03
(2)
15
WALDEN RESEARCH CORPORATION
-------
100%
M 60
o
(./')
~
0
u
s:::: 40
o
u
~
20
Fi gure 3-2.
80
0%
o
400
1200 1600 2000 24000F~
800
--- -- _. - --
~/
Equi1 ibrium ~o~centration-Of?02 and S(f3 vS Temperature
at Atmospherlc Pressure (51) .
---- - ---
16
WALDEN RESEARCH CORPORATION
-------
Temperature, of
40 100 200 ~ no
: I
----- --- -- ,
) 1--
. ~
~
/
/
~I
,
"
. I
~
II
/
II'-
~J
;;
~ I
J
I I
~ J
~ /
I I
I
/ J
J' I
I I
I / .
,
/ 1 i qu ~d .
b( --- - so 1 i d(c ) i
j I
1.0
10
b.1
200
50 a 100
Temperature, C
150
Figure 3-3.
The Vapor Pressure of 503 (475).
17
.
10
E
.j..J
fa
..
OJ
s-
:3
IJ')
IJ')
OJ
s-
o..
s-
o
0..
fa
:>
WALDEN RESEARCH CORPORATION
-------
I ' ,I" ,1',1:I,rt:11 :1:'t::jT',;1;~:'i;-::-;i:~: J;~:;L;~~t;i£ltlt;:;t.:;~,I::::B~:ltffi!tmtrill.trt:1tJJ!jt~ im T 1[.1' :f
!-"1., - ~I .. -, 4~'L.:.":I:'1:::t1Jtl1t"H:b!...d!~_.rlhJ:1-tLri~ 1-0, 1 , ~I-J-t
~: ,,-, .~..1....- ~ -~" !f;~ ::;'~f+~tl rLj! Ii',
.,-",: ~- ~.1- ~-'" -~-~ ~ -~ ,~L
~, ~T!' :8 ill! ,di ',,1
f - ~t. 'U ~ ~ ~ '~
:~t: L , -t1' :.1f 11 :iIfEU H
,~ Figure 3-4. The H 504 = H20 + 503 Equilibrium ~.:m-;,'~:~fi\~; J~ltIS"
"', 'I' 2 , , ',- ,~ -~ ~H' If+t+''''.H+T'' !iT ;~, 11 W
r :" ~ , . -. -~ 1: 1. - . ...j. .,. -- -, rr -.-t 1- .1 tl -I, ~ hW-Tm ~ ~ t - I. t ~ttt -j 1"+1 1 t- - i 1.1 11ff ~
~'l::"I"::,:,,,I"t:, : ';lil;,I"rJ~:r.: "~I: 1I''-!'H~trtt:=F ~~~I ~1t:; r;~ I:Jhd ~tf~ tT;it..'t:t+~ttlMi ~ "Sn If:- ~,H t, 1
I., "I' ," I '-l" !d" -I~~~:ht -,' t "-I " H-;~t- tj, -m-~ ;-1 p;. d.l+1- , t-. - 1 i
" i~ ~ .;.;~.;:~, ~':, II' ' ;; ::: I : , ,: ;': I ' " I t~ : Ii , : T :'.: ! l;t I: U r : : I TjJ, ~!:! j: t! 1:;: T: t 'II tt-!1! ' t, t: i 1 i Ij ! i 11 ,,;.j , ' 1 !
:jl:r~+! i::;it;;~: ." :;:+~H ~11;:~j~ )++ 1,f;tj!J}i f!fi tHrf[~ jiiHHmtt :tH t!it ~[tt~~t' 't li It, IITJ : If!' tt ," " '
100 !i,' :::1 :":,", ',:"", '", ; i:I::" ;:;;; d, ,t,,,, ,. I," "'''''' , "': , . , "-,'. "" , " 'H' rr III :t '
trill; ,~ 'II' , ;::.:l=rili'H~ ~':~:::cI=S-.2-li~1l~1.fu~ l~F:mUTj~'lli:m~ fig#f~~fl ",' 'Ht::t'4m~tH t':t;r' 1:+ .1 ~, :B'
-4:11-':;: ~;;i~ "."~-':.~.lL;.':-: ;-Ltl1-1 '- H~' ~~~~ Ir.,~ '1 r ....;->- ~~.. ~~_. ~:~... +--~,-:' '-,.-~ +--~' ,.~:t1 t',-, - - l.j:1-t.o...j. ....~t.l- - ~ ! .'R
't' , m-: t:.:1 '~i" i;'.' ::r: :l~ J{ ~;-" ~:2- 'lf: E1+ m~ fi:ii I1J 1i~ ;;;:8:ill ft' E j~cf :1~~ -fu' '1" i' t+ ....It ' " " q1:t :' I- 'tt~
t l,t~;'~:",I""'...,.'1':'j::'t,,1..,"Iq. >-i 1'~J:.,.t- ',:".-':: -;';-,:" "":.2;,, Hit 'f ':It:. a~H1,;'t' !j:Jj
l' J ,tit "I" -- i' .." 1;" "'~f hi. . --',I ;2 '~-: J.j:o .::ffi'irrti ~':-i:]-&,:::tt. "..,i-! L ,1 ';I' lr;: r'~1i' . ~:It. ++ r1:1 :!H i 4 t + +Ii
'~.. ~ i1. .~- j-~
14~f-:-';I' ,;.;Ht1 ,~:~ " ::: ~ :~;: ::~' :.,~ ~;+.'" ,':t:'~:~tc.~t: ',~t;~. 18m' ,; '~,. ,h,' t, j1 ji" ,
JJ:~" ,IJ;'ll ,;411;" ~.., .." 1r:lI,:".:; t';.;: ':-:-t .,-" 4 ;::';'~1H;tq11 '.,'" d, "'" w ,," j' 'Ft,
11:; :;n 't!, :--+-,1 rj-,,:'i; :-;'-~ tjt-:.. :~~t ~-:~' ...::.+ I-~Ij ...-......; ~-+L ,.... ~+ H.;~; ..H-d+h I ,-I 1.+ ! - -..,- "'. - .
80 ,I:l!:;;;,:!:j; ~ ;:t4t: J hti': i:i1;h~if;i:Ei iliHi,:; tW ;Eli1 ~~lj jSf:tiif4t : ~ , r:+! ~ ' , :t r+} , L:J:: fl' , .
,.1 ~ ~+~t+-8'~~~ I ':f . .,.';';"'.:;~ '~:~.::t;J .~~'~~-H-- - -t-fJ.... ~j- -'-J-P-t-"- 1.~ittlct~1. l-j.. +- 4 -+4- ,+-, ~~:I.-..r.t -
'fn ijl: i;:~Llh!'I::; fI!:ibL,"::~:q;/i;:;f:~;F~ 1t~i:it~~;~+~';~~:~,t~;;~~t'-':,,:, tl: 'I"n't~~ ~'f~I~1tl4: "~'II'+
liltt\!'''::';:!!: ,,!,::,,!;'L'L';':":'::-:-'::""'t:..;H',':":"Htt':'~:,II t.''''',:i+Eh+~it " L.. +i,' b tl"'fJ';X ' ,'::
.~llL ~t!r+-t~f +H ,;:: :H,: >~,~h~~~f;; :~2,;:~H,r:,:~';~;1 ;~;~, ~~~l;;:;: ~;~trU ;-)+- 't~ l;',': I' '", ' 'mtt ~' ;jt1,'k, ' 'r' ,f1t:fJi / '
! Iii :,~! '",'i !;!i <;+:", ,::' :~::;, ,,:.:" -:~i '~,;: I',' ::, -:'=:: :<:; ! liir~'j..t, iftl '>:. ,tt t t", ,,' 1> H"',
70 Iltl!1 ';;'r;;:: ife" ':;I::t1 ;t,: ,,: .':: ;".: 'if fill ;W~jH1i!" i:,:U J1;( ~l:i :r:,:!;)~HiE~li ,j, ,,;,,~i;ffJ jr;tf~'nd 1:1,1]:1 ;:ftirt~1
tH'~ff-l :.:s :IL~,~ ::,; :.: .:.;.:' ;:;::, ":::.', ,1,,' ',: :'~ .:;:;rii1 '~':, ',~, :::~ " <4+ '-h' ,.;, I ' ,. '. ' ,
I ,!il! Ii i!ji l1TIlf:i i>: I/L~if Ji,i ::: ,::; ;~:: Hr.Bl~ ilf-it:: j:;,~;.; "Ui~jj llii""\ ,'h n~Ht:11rk H"iLRrdt +t,t-
~I,q ill}+-:, j " ":"'. ',:j~: ~~,...: ~':. .:~..t~1:- ,'-il> ,::.::':.:.~: ~- - -'~:J_-~-t-'''''++ ~ + t .+I-~ h t ,.tt~t:1~t, ~ -t,
l! mil il'I:,I: :":i:~~-:' ~1!l;Ln 61: !J:, :: I.,: l!~:- ~;:: 'flU dE ':::j'T: :::, ft1:1 UIH"t-il 11: :#1 tn . t ' IH~ ~di !fLi.r,T;:!1 ::G 11 ,'f
.lL ~'f-'lUL I;ti +;-:~';-: I+L:+ t-::P..~+++'~:::'..,.:+:+i+~':','i :, " ~ ::ii:~ :~' ffi ""' '0: ~:' '; iIlm ' Jit ilti}tli :h::.h 1
IT ',till I.. 'I' '1""'1' t'" II[ '1'":" "'I i'!' JI: I':,!; ~.:. .L~~:: ',11 j:ltS;-!-' t~l1t! j: HI" f1t1 t,-, ~t1t Hi'.! ,
60 ~ i-i+~1-+i1 ,-;1: I;:: "~I~:::I;l ::'::' ,~.:: :,~:: :~;";:: :':,:;::~ ':: '; '~, ';; ~ '~' , t ,;,; 'f;' ~;;: ;;i'"1 "","
I: In: :~II II ii, ,:: ~: I 11': :'3t~ :A::p nit!:.,: 1tt~~:i: : :nt~~lli1l;: if bf) !t: ~ tlif ~:trl+1::t. 1. ,111 !:H Ih4 h:- tf!l 110 i3 liL
i I 1\: +1: Tit;" i" 7'::~;: ~I!;J"": ;;:-:jT:, a.:~,i::; T' r:: ,;,:1:':': ::::,;:. I::: ~II;: It;' filL lit. I', !~! iJ'l ;t!:llfi 1::111 :1;'I'i,!iff ft!ti:
Iii :;j , r! ::1 ,\ I;': .q .I~j ;:. :.:~,::. :iir'~~;J ~,~~ .1t: 1;i!l- ,': ;'~: 1~; ~~.!--: ,:.' ~~ 1-1-1 f ~ i.. 11- 111. 1+1. :~~, 11 I !Hftt '
fTIT '" "'fI'+J""'''' "".." """., "'i",I"'I"t" "'"1j"'II.ji"t'llfl" 'j,t~~mt{'1 "",!, W""" !f~t'Jl'lJ 'i
11 I, Ii" I'" Ii 'I I, ,"'" ,I, , , '/'" "1' '''I ' 'I !'" 'it':"~ '" , "t::I,'" i : :+11 'I: I tj',' 'j tt ,:1 t', H'; 'Jrl! r 1;
I Jj'.' ':'r:::: I' :11:I:ij ;mll:: ::.' ~:; I;;' ,; ,I. iL ::1: ';i. :~:.;.tii:i i.S" 'f"Ll': 1." ,,"I ,j' '11.'1' ""'" 1 '
++, '1r' " 1,1,1-+"',''71,' ,iT, t;t:,: " ',1 ',:;',:: .It :1:' Hot,':::' j'.,1:'fr;;::t; ::::1:11; :;1;' : ",' '1'1: j;,' II" ,'., , , "'JI:: 11n tJj ;,'1: +,'1\, !,':
,1 ":II:-':!'i':I::ltq:,l:t~,:::,,I';:I,;:!t';-:t!:tG:lirr.:.:t';11,::"1 ,[iiY:::"i,:t"'l!ff' I, 'I I U :1+1- ~'Ii.t nll",t},
~ 50 II ""!:111'H8':~ 1:'~-'!if,t!:it!;:,;'!;:,l;i;r.':j:hltt:-!~t1::t1::.r~t-t,III'I,.t,,-,'HJ Pq:t 'I+t, ,', 11' It!" H;jj.I,IH,It" ':1
111 Jj:illl",:1 11:,1,1+;1'1"'" ",.I.." ,;.II.!1TI-.-;"~;!"'lrf,q'""".1'j+:, ,+ ,t.,':t!c't' , );:lI.'t1, ,Ltr!i,C,~J:tl-rrt:' t."'
,~ iln : rI :,'; :~:i7I~' ,;': i:,; ::;; +;Jr~~ iI,~ ::~; ':,::1 t;1 ~'):i ~;i~ ~jrr:j~ ~;:'1' :i:: :;: '1;1'~'li "~: ~:~11 m~ d~li ,jii 't:tJ1 1j,t '1 +11 H~,~'1j tltt ~~t"lil{ff,t4~, ffi
II III ,1:' , 1,:","<", :II ,,~, ',' "H ~--.., -+l-t+ .."' .. ,e. ~t', ,,' "i= ,,[lJ,.~ H4 Ht, ~H'o ,tJ: .. ,Hi j- .'il--HF ,,,,,.. It!
,iff, !;~: ::': :;:, :' 'Gdi fti! itfi t£t2~; ;;j:!iE WHW ;t~ ii.ii :!:c ~1~~ ~:;! Si,! Ti :iti i-tI11if tfl~lItt ' =:$ J{!tJtt!:t m .1;jiW ~~" !+
~t 4i} r7-~.+h- ;'!7:'h~f-i7:+!fi:1 ;:(1' :"1 ", I:f~' Te t' ,,' 'I';'; '1-' .;~: ,tT, '1-1 "1'''~ 'rt! t~l: ' , " ", ;'!~I ITs IiffF, :HI, !Hi
I:! l:i! ~.i;!\~~-ii :;': ;;: nn ttT; +~~~ }t~; :t~: L~!~ ~::!- t-_~ :-~!t 1~: :~:~11:~ it.- ~ -- ~~+ ~-~ t1t!- J:.. ~ - ~~ tt tt-H-HtJ.1 11~ L-4H+ ~1i 1 tttt
40 !fT;!~,:j:ji ;::' ,:1 ,J;l-~il~ :1,1 rl i,:r'-i~l! ,:1 ' ,ijHd::t ~i:~-i.i;';~:J:.r 'nt:i ~+ itt-tt ; HIt, -;/J,' t llil
il ~;I! [1:1 T!Jr ~~i~ iH.1 111:1 ;~~ r~~t 4~t....,. ~1 1 -t ~.:!4-r 1:;:, V lJ::r.iJ ~-,.,; .~1 t - - . - -t~-I J:tt
t' i'li :::1' :'1, ,:.[1"1. ;iji fE, j"j :,j' :'1: 1"Sitt1 . ° ;.:;i"Wtinti. It;! ff! l!:H tit - Lj' j+++;" 1rt; '1'"
ti j! ;;;j ..~> : I J" l'lt--t;''- -+-r~ .i::",~ -.HJ.. ~...u :1ti r..it.. ! .1- I. '-r '-it! 1,+ +H--!-- -t1i~:- .... -{
, ' I ,d I:!: p, "'t If, 'r;!' ".;.. .. ", ::jtt ~ J I-- ~c . , 'loti. "w '~,
,Jil1 ;," ii' ":T"I :01 u+lt!lii;;t ,""'311iS. l: "';'5 u~~:~, ':1 J .~'~ till t::l n . "' "'I-\-it:{U j ,
,,~f:!+11!-I~i,.l:;i+j"~~~-:-;;' ,+r~;:'''''::;'~:;'lj;..,;,L, 0'", L 1-" '~'-i-'-.'+H+ll~'fu~.,:;+ +'''" ,,,"
1nILllildll:'1:i::r:H:::,~~~~~'::i::"hfS+," ,;::11m:~;'j>n+j";ftr"f:!ti~:;tl::i~ir;: W, 'r:f S1f;,' [-L f
30 Uti' ;1, I ilii I ~IL: ;'.1 ~t;" ;jH~1: ::;1 G~ ~':tt ~ ~i.-':; ir;.j: W":ri! ~~ :l;,' ' I 't ~j;" - t+ ~fj" .:
:!ffi imlli[ ~:'l T': ;i'; ;'c~: i W ili;i;!i ~;Iili' ,,; : ~ ,,: ; '11 . . - j . ,,;;. - "' ,
t~ ~~'~;' ~,.~ ~: ~ i;~- i~; - ;;r ~;; -:-~~1 ~Ui.i ---~ :Ji .t:~~ ~~; ~~-~ -;' -.. t-: ~h+.
pI. '''I,j :',1;',1. ','~,' --it .,',',,'1,',"'" ';';1:. .'.. ,'-'i; ~'f",+, '+'1' rrJ I~'" :f~'~',~ t , ,.. t + ~rtt'4
llil Jill f'.I t----+.- ' ,-" i';~ '~~, ,-" ',,~~.r.- +- --+T' I..:;..L- Llt- 1 . -.. ..... '...4 L -
!!~!1!' n-i~I'~: IW :~:,:~~l~g r~~j )Jt.~ ~ ~:St~~:~~ :",~ti-~" !~t!!B t " - h!~,
20 "[',t'., ':(j"'"t1"':::;;::' '1=;:1'1 ":::';:;.i: '+~f:!E:+r;,k'~ct!, t: ,."t:Etr2i ;:tLif;q .1' Wr~r, 'I'';:''''~ 1
1:11r-~-~~~:!R-~:rJ...:~:~-i.1,+t ""'.'-:7~, '. --., ~r.i J".+,:-:-~~.i::~': t~:lftit1 .0- -1> i.--j 1
'~';l": ,I :I"ifl-": ~iUH:; ~~j at.,.;:S-.' j im:i!it'~'.:r.I:tt::;:.:c;.r' W** L, ",I'-'-<"f ~Hj,":,' t
':71~ t-' I, ;~-~'~ -i-H:'-,r,;+-H- ... ,..- ,'- +-t. ' Irn~u:-u:::t:L1-:-":-'-'1, ~~ ~~J j~j1
:::+:T: /1'" >: :t:I '. :~, ,'~;, ~. ;:'1'~ :r~t '-,;~ I, ~:::;rH~FY,~:::: :i~ :; ;~r~i'lf:!,i ~,' L, 1t 't'![ ~m lSti ~ I
:':'" '----t--=-I:"I'; f';---" 'lib' "~:t',:- " ~~'h+;" '- ,.,.'t:....l F.' "" ''-j-, --:, '." :.;, "', ,- . ''I, '''t '""
10 :::';::k; .1i, r ~ilir "F:1:t;:- ;.:-r: z:.: ,:..:... c':: 10th- ::t~ E;::"L '.: ~:;! :~~l :¥:= ~2 ~::HTIi ~1 '~: .~ 1 tt',: ;,~:,~1; c~ ~U,: ,if4P Ih.. -
I': "II ""Ii::' '1,:;' ',r 1.t;J L'!o+: :'''It=''' :'7:: g.'- ,;,;4 ",;1'.' :::; ~ ;i~: r;-:,tti1, - r
~-i "" 'cd, i \".. 'Jy+ p' ""Ift' ," "o_;;=:. -r:.. '""". .; ~~~::,:';-£' +~:I,
tIT": ': ,1:li ,': , "'j'" ,,"". --,+ Cj-"'T,'-'+:i-tdE-:t:.:,t:" I" \ ,:,:t
'Sj~~:~Oi T!400;~' F'f~~~l':Co~b~:r ;~~f rr~lO :1"i;~oILt:ij;~:f~
16 Temperature
I::
o
......
+-'
fO
.......
U
o
VI
VI
~
o
,
~
~I,;-f-r,r, 1
-~tt H
> ,--,
~ 18
-------
The thermodynamic (equilibrium) data for reaction (1) are shown in
Figures 3-S and 3-6 (S30).
The point marked IXI in Figure 3-6 corresponds to
the results previously quoted from reference (Sl). It is not possible
to extend these measurements over a wider range of temperature since
the reaction is frozen below about 10000C under the conditions in a
furnace.
It is not known whether reactions (1) or (2) repre-
sent the mechanism of 503 formation. Evidence from measurements made
in H25 flames that 502 is oxidized by ° atoms (reaction 2) (48S) is:
first, there is a close parallel between the ° atom concentration
and the rate of 503 formation and, second, there is no evidence for
a molecular equilibrium reaction.
3.3.2 Heterogeneous Oxidation
Many metal oxides catalyze the oxidation of gaseous
S02' Vadadium oxide is one of the most efficient and a significant
component of many residual oils. Nickel compounds, also present in
oil, decompose to nickel oxide, which also catalyze the oxidation of
502' Investigation of this catalytic process is difficult, because
the adsorption of 502 on the walls changes continuously, particularly
where carbon deposition is high (S30). However, the rate of oxida-
tion increases with temperature (484).
At 1000-SOOoC, V20S is thought to react chemically
with 502 (Sl), V20S ~ V204 + 503' Ferric oxide oxidizes 502 at 800°C
by the reaction (474) 3 Fe203 + S02 ~ 2 Fe304 + 503'
3.3.3 Sulfate Decomposition
Alkali metal sulfates and ferrous sulfate, present
in furnace deposits, decompose on heating to product 503' The dis-
sociation temperatures of calcium and magnesium sulfates at one at-
mosphere, are lS9SoC and 1150°C, respectively (476). At S20°C
19
WALDEN RESEARCH CORPORATION
-------
Figure 3-5.
Figure 3-6.
- -- ------ ----- - --- -
.,
-------------- ----.---
'" If"
.1
.J
'1
,Kf" P,;OJ~~
IPSO} ',Fb1t
I I
I '
I
. /"
"7l
.r
., I
4'
8:rJ,0j 1100
( '0"
,
, ,
/fDJ JDD MJIJ
1ro I oro T 01(/,
, ,.so,.,41'~ J \;,
I,D
UJ
The Equilibrium Constant for Reaction (l) vs Temperature
~/.O '.,."
.
~~~r-r ~ -1-\-
q6 1_-1-~\~-~ -+-ITI-1
: \,:~t=U? J~~~~:
:.: !j~~:\HjjfE'
JI' "'~v..0'!. 02 I
QJ I -\ \~.--1-!-1-,--
r\: \0..--+---;----i--
I '+, ",:.......: . ,
,,;.:~~:~:-~m .
I ~--.,.~
JO() @ 5a) 6(X) 700 lJ(X) 900 1I:w,iix;-'20.il]O()'c
Q9
Qi
0./
The Fraction of the Total 50 Which is Oxidized to 503
Calculated from the Equilibrfum Constant
20
WALDEN RESEARCH CORPORATION
-------
ferrous su1 fate is oxi di zed to ferri c sulfat.e whi ch decomposes at
6500C.
The decomposition of sulfates has recently been exten-
sively reveiwed by Thomas et a1 (493).
3.4 Adsorption of S02
Data on the adsorption of S02 on carbon black is given below
(Table 3-1). The adsorption isotherms on silica over the range 0-1000C
are shown in Figure 3-7. Adsorption on silica gel is almost entirely
capillary condensation, rather than direct surface adsorption, as the
isotherms fit the Freundlich equation (51), but at 300 and 4000C the
adsorption process is complicated by the reaction, 3 S02 ~ 2S03 + S.
Many other substances adsorb S02; generally much water is
adsorbed simultaneously. Salts of the alkali metals, alkaline earth
metals and the sulfides of nickel, zinc and iron have also been in-
vestigated (491). Many of these substances react with S02 rather than
adsorb it.
Potentially, carbon is the best sorbent for S02' first be-
cause its capacity is high and, second, because, in a flue gas, water
and oxygen are simultaneously adsorbed and at elevated temperatures
oxidize S02 to S03 and convert the product to dilute sulfuric acid.
The collection of (S02 + S03 + H2S04) would therefore be relatively
simple and free from much foreign adsorbed material. Carbon is the
closest approximation to a selective (non-reactive) sorbent for total
sulfur oxides in flue gases.
3.5 Solution Chemistry of S02
3.5. 1
The Solubility of S02
Solubility data for S02 in water and in dilute H2S04
are summarized in Table 3-2. The solubility of S02 in water is signifi-
cant but decreases with increasing concentration of (dilute) H2S04'
21
WALDEN RESEARCH CORPORATION
-------
~
r
C
JTI
Z
::II
JTI
CIJ
~
::II
o
::J:
o
o
::II
~
::II
~
(5
z
TABLE 3-1
ADSORPTION OF S02 ON CARBON BLACK (477)
Temperature °c 6.5 12.0 16.0 18.5 25.5 43.5 56.0 71.0 86.0
Adsorbed Vol. cc/gm
Carbon at One
Atmosphere 495 408 375 320 261 185 136 97 76
Pressure in mm Mg 827.5 759.5 649.5 551.5 404.5 322.5 258.5 194.5 126.6 78.5 47.0
Adsorbed Vol. cc/gm 402.9 379.4 320.8 278.9 235.6 195.4 171.9 140.6 101.5 60.4 21.4
Carbon
N
N
-------
o
100"C.
80
70
60 --'-
-J
~
u.
o
~ ~O
.....
11).
E
u
ill
~ 40
o
\oJ
en
cr
o
i 30
20
10
300
Figure 3-7.
Adsorption of S02 on Silica Gel (478)
23
WALDEN RESEARCH CORPORATION
-------
TABLE 3-2
SOLUBILITY OF S02
S02 Partial Solubility
Pressure Temp 9 S02/1000 9
(atm) (oC) Solvent Solvent Reference
1 62 Water 21.5 (482)
1 41 Water 48.1 ( 482 )
1 20 Water 118 (472)
1 0 Water 186 (473)
5.7x10-4 50 Water 0.5 (483)
1.7x10-4 11 Wa te r 0.2 (483)
1 62 10% H2S04 17. 1 (482)
30.3% H2S04 12.6 (482)
41 7.3% H2S04 36.5 (482)
18% H2S04
24
WALDEN RESEARCH CORPORATION
-------
In the presence of air and trace concentrations of
transition metal ions, the apparent solubility may be greatly increased
by oxidation to SO; (549,527). Manganese sulfate (0.0028%) increases
the apparent solubility of S02 in water by 600% by catalyzing its oxi-
dation to S03' However, the presence of phenol (9) cresol and xylenol
as impurities (0.005%) inhibit the oxidation of S02 in the presence
of Mn++ ions; catechol, resorcinol, and pyrogallol and other poly-
hydric phenols are claimed to stop the reaction completely (but see
subsequent discussion of analytical chemistry).
3.5.2 Acid Dissociation Constants
The dissociation constants of sulfuric, sulfurous and,
for comparison, carbonic acids are given in Table 3-3 (481).
TABLE 3-3
ACID DISSOCIATION CONSTANTS
H2S04
H2S03
H2C03
1. 7xl 0-2
3.5xlO-7
250C
180C
2nd Temp.
2xlO-2 180C
5xlO-6 250C
4 .4xl 0-11 250C
1st
Temp.
The acid strength of H2S03 is comparable to that of
HS04 and far greater than that of H2C03'
3.5.3 Redox Chemistry
The oxidation-reduction potentials for the sulfite-
sulfate couple vs H2 are (479),
Acid solution
-+ --
H2S03 +- S04
-- -+ --
S03 +- S04
EO = -0. 1 7 V
E~ = +0. 93 V
Alkaline solution
25
WALDEN RESEARCH CORPORATION
-------
The standard potentials for reaction with molecular
oxygen are (479):
Acid Solution
= +
H2S03 + 1/2 02 = S04 + 2H
EO = 1.06V
Alkaline Solution
SO; + 1/2 02 = S04
Junge and Ryan (549) have shown that the aqueous oxidation rate of S02
(4 ppm in air) decreased with decreasing pH and ceased at pH 2.2 (in
the presence of FeC12 catalyst). Berger et a1 (527) repeated these
experiments at 17 ppm S02 in the presence of fly ash and found that
oxidation ceased at pH ~ 3.
EO = 1.33V
In view of the large favorable potentials, it appears
unlikely that thermodynamic factors playa significant role in the
acid inhibition phenomenon.
3.6 Conclusions
The general properties of S02 and S03' discussed above, have
been selected for their bearing on some of the problems and sources of
error in flue gas analysis.
The solubility data, for example, show that S02 is absorbed in
water, but less so in dilute acid. Both S02 and 503 are obviously
soluble in alkali, but the oxidation of S02 to S03 cannot readily be
inhibited in alkali. In water or dilute acid, many organic phenol
compounds at least partially inhibit the oxidation of S02 to S03' al-
though they are less effective in the presence of ions of the trans-
ition metals.
Adsorption data is probably not a useful method for analysis,
since even the relatively high absorption on silica gel and carbon is
26
WALDEN RESEARCH CORPORATION.
-------
by capillary condensation and therefore not specific. The oxidative
adsorption on carbon may, however, be useful for total sulfur oxides.
The very high adsorption on carbon is
possible source of error in SO analysis where
x
controlled.
an indication of a
combustion is poorly
The low vapor pressures of S03 and H2S04' compared to the re-
maining constituents of the flue gas permits efficient condensation
of S03 as acid mist.
27
WALDEN RESEARCH CORPORATION
-------
4.
MANUAL SAMPLING METHODS FOR GASEOUS POLLUTANTS
4.1
Introduction
There is no universal method for sampling gaseous pollutants,
even where the number of pollutants is quite restricted (as it is
here). Changes in the effluent environment due to the variety of
source and fuel types, operating characteristics (steady-state vs.
transient), and control techniques establish the unreasonableness
of developing a single recommended standard method or even a single
simplified method for the several contaminants and the wide range of
concentrations encountered. Therefore, a number of manual methods
for specific situations must be considered. While different in de-
tail, they share common requirements for pre-test planning, sample
train components, and determination of velocity, moisture content,
temperature and pressure. The brief discussion below outlines
these common factors only in the detail necessary to demonstrate
the relationship between the various aspects of the program.
4.2 Common Factors
4.2.1
Pre-Test Planning
Analysis of the objectives of an emissions determina-
tion, the characteristics of the source and the available resources
are essential elements of the pre-test plan. Such factors as
steady-state vs. transient operation, lengths of cycles, peak
periods of emission, estimated gas composition and pollutant con-
centrations, gas temperatures, duct size, gas velocity and humidity
are examined to select an appropriate collection method, method of
analysis~ sampling rate, sampling interval, number of tests re-
quired for proper statistical analysis, as well as to consider
less obvious needs, such as special materials of construction
(temperature or corrosion effects, etc.). Such mundane factors as
availability of ladders, scaffolding, sampling ports, lighting (for
28
WALDEN RESEARCH CORPORATION
-------
night work), and coordination with operating personnel must be in-
cluded in the test plan.
Adequate planning increases the probability for
representativeness of the sample which is the real measure of success
in--"1:he tests to follow. The alternative to pre-test planning is the
too familiar trial-and-error method which leads to extended experi-
mental programs, overlooked key data and a generally more costly and
less representative result.
4.2.2 Stack Measurements
The initial step in the experimental procedure is to
establish the velocity, pressure, temperature, and water content of
the stack gases. Velocity measurements are required to permit integra-
tion over the cross-sectional area to compute the total volume of ef-
fluent. Time dependence of flow rate, temperature and composition
should also be observ~d over a complete operating cycle.
Rigorous methods for velocity measurements (including
sampling locations, number of samples per cross-section and their
position) are described by Bloomfield (525), ASME (502), Hawksley (537).
Industrial and power boilers and heaters are usually designed for
velocities of 2000-3000 ftjmin, which is a convenient range for pitot
tub1 use. A pitot traverse is generally employed. (See Section 4.3.2.)
Sma~l natural draft furnaces found in residential heating produce low
velocities (approx. 10 ftjsec) (539). Pitot tubes with conventional
manometers or other flow measurement schemes are necessary. (See
Section 4.3.3.)
There is much variability in the configurations of the
ducts containing the effluent. This may introduce difficulties in es-
tablishing the cross section of the duct (necessary for volume and
mass flow) and may lead to flow irregularities following bends or con-
strictions. One is generally advised to sample several diameters down-
stream and a few diameters upstream of any flow disturbing element. In
29
WALDEN RESEARCH CORPORATION
-------
practice it is rare to find an ideal location. The deviation from
spatial uniformity of flow in a duct is well illustrated in Figure
4-1 (537) which shows the velocity ranges encountered under some
test conditions. The necessity of a velocity traverse is obvious.
(A review of methods for determination of gas velocity in ducts is
presented in Section 4.2.6.)
The pitot sampling points are frequently used sub-
sequently as contaminant sampling points. Stack temperature, pres-
sure, and moisture content must also be measured to compute contami-
nation at stack and standard conditions. Temperature and moisture
content generally change in the sampling system and the volumes in-
dicated by the sample flow meter must be corrected back to the in-
itial conditions. The BSI (379) method is satisfactory for determi-
nation of the moisture content of the flue gas.
4.2.3 Sampling Train Components
4.2.3.1
Genera 1
The apparatus used for the manual sampling
of gases has been divided into modular components. A typical array
of sampling system components is shown in Figure 4-2 [after Bloomfield
(525)], which depicts, schematically, a duct, sampling nozzle, probe
(sometimes temperature-controlled), collection device, water condenser
and/or dessicant column, sample flow meter, (gas temperature, pressure
and water content at the meter are essential), and air mover. Analo-
gous sampling arrays are discussed by American Congress of Government
Industrial Hygienists (540), Arthur D. Little, Inc. (541), ASME (502),
Hawksley (537), IIAir Pollution Source Testing Manual" (1), ASTM (547),
Manufacturing Chemists Association (542). and Industrial Gas Cleaning
Institute (543).
For gas sampling the nozzle design is not
critical, although provision is frequently made for a simple preliminary
30
WAlDEN RESEARCH CORPORATION
-------
0,'
0.,
"0
1'0
1.1
\.1
GAS VELOCITY
"'OM IOILIR.
\
1'1
'.1 ()AJ 0.7 oe
I') 09 0,7 ().6
.
I
I
,
TO + LD. fM
M£RAG& GAl Yti.CKITY . H "/t£..
I') 1.0 ()AJ ()AJ
1
',4 "Z I'Z I'J
GAS VELOCITY
FROM IOILER
0.5 "0 0,3 0.4
,.3 "0 1'.3 0"
1'3 1'1 1'2 "4
1,2 0'" 1'1 1'2
GAS VELOCITY
AVIRAGE GAl VlLOCITY ... njae:
FAOhI
I.D. fAH
AVERAGE GAS VELOCITY. as fT/SEC
Fi gure 4-1.
Observed Variations of Flow in Ducts
(After Hawksley) (537)
31
WALDEN RESEARCH CORPORATiON
-------
~
(E) CYCLONE ~
(F) PARTICULATE OR (i) CONDENSER I
GAS ABSORPTION IF REQUIRED
(C) PROBE
TEMPfRATURE
NOZZLE-CYCLONE ~ CONTROL BATH (L) ROTAMETER
ASSEMBLY AND SERIES- WHEN REQUIRED T
CONNECTED FILTER
(G) FREEZE-OUT TRAP
tr. TlVATED CARBON
---.II '--- SILICA GEL, ALU-
~ MINA, ETC.
(H) ADSORPTION
WHERE REQUIRED, PLACE
IN HEATED ENCLOSUR~NTO~
PREVENT CONDENSATIO~
(0) FABRiC, PAPER, GLASS,
MEMBRANE, CERAMIC
OR METAL FILTER MEDIA
P, T, I pv
I
!
(
(A) PROBE
f
NULL OR iNTER-
CHANGEABLE OR
SINGLE SIZE NOZZLE
(B) PROBE
FILTER MFDIA IN
NOZZLE ASSEMBLY
DUCT
Figure 4-2.
@nll
(J) ORIFICE F'LOWMETER
(M) GAS METER
Sampling System Components
(After Bloomfield) (525)
32
If
(N) PUMP
WATER, STEAM OR
COMPRESSED AIR
+
+
-
(0) EJECTOR
WALDEN RESEARCH CORPORATION
-------
filter to remove combustion particulates which may interfere with the
analysis or be destructive to meters, pumps, etc.
Probes are of two general classes, simple
and temperature-controlled. Probes are frequently heated to prevent
loss of sample through condensation. The effectiveness of this tech-
nique in sampling for S02 and S03 is shown by Lisle and Sensenbaugh
(80). Alternatively, one can use a simple probe, but one must use
very careful handling techniques and quantitatively wash condensate
and precipitate from the probe and sampling lines. This material is
then combined with the material in the collector. This approach in-
creases the opportunity of error and is generally undesirable. Cooled
probes are used where one is sampling directly in the furnace at tem-
peratures above about 8000F.
Gas sample collecting devices may be divided
into two classes, those which concentrate a sample and those which do
not. The latter are sometimes called grab samplers and simply retain
a gas sample for subsequent analysis in the laboratory. Most of the
familiar methods, i.e., absorption, adsorption, condensation, or
freeze-out are concentrating mechanisms. Of these, by far the most
used is absorption.
In liquid absorption effectiveness is de-
termined by:
contact between gas and absorbent
rates of diffusion and mixing
solubility in or reactivity of the absorbent
volatility of contaminant
concentration of reactant in absorbing solution
volumes of gas sampled and absorbent
Clearly, the selection of a specific contaminant collection system must
be carefully made. Scrubbing efficiency (discussed below) cannot be
assumed to be complete and quantitative.
33
WALDEN RESEARCH CORPORATION
-------
A variety of adsorbents are available for gas
sampling, although the most used is active carbon. The adsorbed gas
is then removed by heat or eluted by a carrier gas. Solvent stripping
and vacuum removal may also be used. Other adsorbents in common use
are silica gel, lithium chloride, alumina and fuller1s earth.
Condensation or freezeout methods are also
employed. Condensation is commonly used to remove water from the
combustion gases prior to metering and may be used for specific con-
taminants such as S03' The temperatures required to collect a given
contaminant depend upon its properties.
Grab samples are taken in bottles or flasks
and several types of plastic bags or balloons. Plastic bags have
the advantage that there is no pressure change on filling or with-
drawal of the sample.
Some precautions which also apply to other
collecting devices, are most important in grab sampling, since samples
are sometimes retained for an appreciable time prior to analysis.
Specifically, these problems include chemical interactions in the
sample, photochemical decomposition, adsorption on container walls,
and leaks.
Indicator tubes are a special class of com-
mercially available grab samplers for specific gaseous contaminants.
A granular solid treated to produce a color reaction when exposed to
a specific contaminant is confined in a glass or plastic tube. A
"known volume" sample is then drawn through the tube (usually with
a squeeze bulb or piston pump) and the concentration of contaminant
is determined by the existence and length of the color stain in the
tube. Treated papers and paints based upon the same principle have
also been used.
34
WALDEN RESEARCH CORPORATION
~-"', '
-------
Sample metering is required for a complete sys-
tem. For a specific sampling system, one must select from the following
types of meters: constriction types (orifice, nozzle, venturi), dry gas,
wet test, rotameters, positive displacement (rotary. piston, gasometer),
turbine, laminar flow (capillary), deflection vane, vane anemometer, and
evaluated tanks (pressure differences compensated for temperature). For
a complete mechanical description, see Jacobs (507), Altieri (533) or the
AGA Manual (534).
A manual sampling method demands ruggedness and
portability and must be able to tolerate corrosive combustion products.
Based on these considerations and accuracy requirements, a number of meter-
ing techniques may quickly be eliminated. Final selection will thus be
made from the following, more restricted group: dry gas, orifice, nozzle
or venturi, capillary, rotameter, and evacuated tanks. All of these de-
vices must be carefully calibrated at their operating conditions. Pres-
sure, temperature, and water content at the meter must be known in order
to correct the combustion gas sample back to initial conditions.
Air movers must be able to overcome the pres-
sure drops in the sampling system at the required gas flow rate. The
following types of air movers have been used: turbine blower (high
volume, low pressure differential), mechanical vane pump (high pres-
sure differential, moderate volume), mechanical piston, diaphragm,
or positive-displacement gear pump (high pressure differential, high
volume), air, water, or steam ejectors (low pressure differential,
moderate volume, useful in hazardous atmospheres, and where electricity
is unavailable), pressure difference in the system (special cases only),
and evacuated tanks.
Flow control can be achieved in several al-
ternative ways, viz: a valve between the flow meter and the air
mover, a bleed or recycle valve at the air mover inlet, speed control
of the air mover motor (usually varying the input voltage to a uni-
versal motor). or a constant flow constriction (can be used as flow
35
WALDEN RESEARCH CORPORATION
-------
meter also; requires vacuum source able to maintain
orifice and cannot be used for an on-off control).
is presented in Section 4.2.3.2, following.
sonic flow through
Further ana lys is
Materials of Construction - The materials of
construction are often a major factor in selection of components. Cor-
rosive species of the environment may react with container walls and
destroy the sample as well as the equipment. For example, ASTM (01605)
cautions against the use of metal containers for H2S, S02 or oxides of
nitrogen. Most sampling devices are made of glass or plastic although
specific stainless steels can be used. Concern about proper selection
of materials of construction is reduced downstream of the collector,
although it is still necessary to insure reasonable equipment life.
4.2.3.2 Evaluation of Specific Components
Sample Metering Device - The accuracy of an
emission determination depends as much upon the accuracy of sample
gas volume measurements as upon the analysis. This implies that suit-
able sample gas metering devices must provide measurement with errors
of 1% or less. Within this accuracy constraint, the particular
measuring device will be selected in harmony with the pollutant col-
lection scheme and within the general engineering design requirements
of:
a.
b.
portability
instrumental and operational
rel i abil i ty and
ruggedness
simplicity
c.
d.
For manual sampling systems for SO pollutants
in flue gas streams, the critical orifice is the best engin~ering solu-
tion for sample gas measurement. When calibrated, critical orifices
meet the measurement accuracy requirements plus the design requirements
listed above. The collectors used in an SOx system are a condenser for
S03 and midget impingers for S02. In order to obtain high collection
36
WALDEN RESEARCH CORPORATION
-------
efficiency, this system will operate at flow rates of three liters
per minute or less. It is not necessary to vary the flow rate during
sampling; hence, discrete flow rate selection is satisfactory. Criti-
cal orifices are commercially available (Millipore Corporation, Bedford,
Mass.) in flow rates that include 0.5,1.0 and 3.0 l/min. These de-
vices are made of stainless steel, weigh 3 grams, and occupy a volume
of less than a cubic centimeter. It is obvious that these devices
are readily portable and instrumentally simple. They are mechanically
rugged; hence, no special handling techniques are required to prevent
damage during field use. Since the flow rate is set by the particular
orifice, operation simply requires starting the sampling pump and
measurement of the sampling time. Sample gas pressure and temperature
must of course be measured for density corrections, and the pressure
differential across the orifice noted to assure that critical flow
conditions through the orifice are maintained.
Other sample gas measuring devices which
may be satisfactorily used in SO source sampling are summarized in
x
Table 4-1. However, they are obviously not the simplest instrumental
or operational solutions.
Air Movers - For the SO system the selection
x
of air movers is limited to mechanical pumps which can maintain the
critical pressure drop across a 3 llmin critical orifice when con-
nected to the sampling system. A particularly convenient pump is an
eight pound, oil-less, motor-vacuum pump unit (Gast Model 1531).
Based on our experience, this unit is readily portable, reliable
and ragged. Similar units should also be satisfactory.
37
WALDEN RESEARCH CORPORATION
-------
~
o
IT!
Z
::0
IT!
CJ)
~
::0
(')
:r
(')
o
::0
~
::0
~
;.0
"2
TABLE 4-1
COMPARISON OF SATISFACTORY GAS METERS FOR SOx SAMPLING
TOTAL VOLUME GAS METERS
Typi ca 1
Best Instrumental Operational Field
Meter Accuracy Portabil ity Simpl icity Simplicity Rel i abil i ty Ruggedness
Wet Test :to. 5% Very Poor Fai r Good Good Poor
Meter
Dry Gas '\J:tl% Good Fair Good Good
Rema rks
w
ex>
Fair
(frequent
ca1. neces-
sary)
Good
Good
Ro ta ry
Gas
Meter
'\J:tl%
( 1 a rge
vo 1. )
Fai r
Good
Poor
Constant
Vol.
Evac.
Tank
'\J:tl%
Good
Fa; r
Good
Good
Good
Ma i n 1 y a 1 ab
instrument
Poor for flow
<3 llmin
Not commonly
a v a 11 ab 1 e fo r
small flows
Total sample
volume must
be smaller
than tank
volume
-------
TABLE 4-1 (Cont.)
GAS FLOW METE RS
Typical
Best Instrumental Operational Field
Meter Accuracy Portabi 1 i ty Simpl i ci ty Simp 1 i ci ty Reliability Ruggedness Remarks
Calibra- 11 to 2% Good Good Fair Good Good Sample volume
ted Ori- (with calculated as a
fice ca re- function of flow
Meter ful cali - and time
bration)
Capi 1- :tl to 2% Good Good Fai r Good Good (Same as Above)
lary (with
Flow care-
Meter ful cali-
eN bration)
1.0
Ven tu ri 11 to 2% Good Good Fai r Good Good (Same as Above)
Meter (with
care-
fu 1 ca 1 i -
bration)
Ro ta - 11 to 2% Good Good Fai r Good Fair (Same as Above)
meter (cali-
~ brated
with flue
r gas)
c
lot
:z
:;0 Mass '\112% Fai r Poor Good Informati on Fair Recorder output
I'T1
(/) Flow not avail- available
~
:;0 Meters able
("')
r
("')
0
:;0
~
:;0
~
0
:z
-------
4.2.4 Representative Stack Sampling for Gases
The classical problem of stack sampling for gases is
to associate the (average) samples* taken generally at different times
and positions, with either an average concentration, or with greater
difficulty, an average throughput. The precision and accuracy of the
determination are a function of the method(s), the experimental de-
sign and the variability of the source. In any sampling program, one
must specify the desired precision as determined by the purpose of the
work, before the ~gnitude (or even the feasibility) of the sampling
program can be established.
Before attempting high precision analysis of total
emissions, it is necessary to identify and distinguish variability in
the source from error and imprecision introduced by the methodology.
IITrueli va ri abi 1 ity wi 11 exi st in total flow and in component concentra-
tions, as functions of time and location in the duct. Each measure-
ment (velocity, duct size, sample flow rate, sampling time, mass of
contaminant, etc.) will have an associated error. As a result of the
finite number of replications there will be an associated imprecision
to the limiting means and, finally, since only a finite number of
locations are selected, there will be an associated imprecision in
mapping the flow field.
Flue gas flow measurement techniques are reviewed
in Section 4.2.6, following, which discusses velocity measurements,
duct traverses, calculation of volumetric flow and their associated
errors. Sampling locations in ducts and sampling errors are dis-
cussed in Section 4.2.7.
*
The.collec~ion.of a ga~ sample which accurately represents the flow at
a gl~en pOlnt lS ~ar slmple~ th~n the complexities of representative
p~rtlculate sampllng. Isoklnetlc sampling rates are not a requirement
Slnce t~e (~he~m~l and turbulent) eddy diffusivities are far too lar e
to perm~t slg~lflcant molecular fractionation in the very weak centri-
fugal flelds lmposed by any velocity mismatch attal.nabl . . 1.
(523). e ln samp lng
40
WAlDEN RESEARCH CORPORATiON
-------
Errors in manual sampling methods are discussed in
Section 5; the observed precision and accuracy of field determination
of sulfur oxide concentrations are discussed in Section 6.
"True" variability in contaminant concentration in
both location and time is best approached by rapid response instru-
mental methods. Gross variations in contaminant concentration
caused by poor mixing of infiltration air can be indirectly compen-
sated by profiling (and monitoring) with a continuous oxygen or
carbon dioxide analyzer and suitable selection or weighting of
sampling points. (See Section 4.2.5, following.)
In large coal-fired power plants, short term fluc-
tuations are far greater than longer term variations. Slag buildup
from boiler. tubes (soot-blowing) may be removed every few hours;
thus, the week to week variations in average gas temperature due
to slag buildup is less than the hour to hour fluctuations. Coal
supplies, delivered to large plants from a number of mines and dif-
ferent seams also contribute to source variation.
A summary of the short and long term variances in
ash and sulfur content of coal samples taken from the feed to a
large' power plant is given below (494):
% Ash
% Sulfur
Long-term variance (weekly)
Short-term variance (hourly)
0.31
0.72
0.12
1.05
Thus the hour-to-hour fluctuation in ash and sulfur content (and the
SOx content of the flue gas) is undoubtedly greater than the week-
to-week fluctuations.
It is thus clear that no single, a priori, experi-
mental plan can be drawn up which will optimize a given emissions
41
WALDEN RESEARCH CORPORATION
-------
output determination. The plan will be a function of the resources
allocated, the accuracy desired, and the nature of the source.
4.2.5 Sampling Errors in Ducts
Literature on sampling locations for obtaining repre-
sentative average concentrations of gases within ducts is scarce.
Although this subject is frequently discussed in the literature for
particulate sampling, the discussions are concerned with inertial
separation of particles from the gas stream. Since inertial separa-
tion of pollutant gases from carrier gases does not occur, the par-
ticulate sampling discussions are not generally applicable (i.e.,
isokinetic sampling is not necessary and stratification due to duct
bends, etc. is not present).
There is a widely held belief that gas stratification
is not present in ducts with turbulent gas flow (613). However, eddy
diffusion studies (571) show that a straight duct length of the order
of 102 duct diameters is required for good mixing of a highly strati-
fied gas. It follows that in many large scale combustion systems
where infiltration air is known to occur there is no direct location
where the gas is well mixed. Luxl (524) carried out a total of 792
Orsat oxygen traverses in 10 different ducts of fossil fuel combina-
tion sources. He found (Figure 4-3) that stratification is generally
present and that single point samples are usually non-representative
in 1 arge ducts.
ASME (502) specifies a multi-point sampling system
for Orsat analysis of flue gases and a single point sampling system
for S03 and S02' but there is no discussion of this inconsistency.
The multi-point locations specified without discussion are then also
selected by ASME for velocity traverse. It may be argued that velo-
city traverse schemes can give representative average velocities
in a duct where the velocity profile is not generally flat; hence,
42
WALDEN RESEARCH CORPORATI~H;
-------
10
......--
~ ~
- 2'l)bObt't'Y '01 3 ,,( b.,
/ ."0" (,lit., O' 0"".
-- --
j - -
I
- - - -
J 'rcbnh,I,1' f", It. 5,"",1. "ob~
lexQ!ed ." ,tit Cenlf'r O' the Duct
I --
-J
*
to
80
10
Figure 4-3.
Probability of
obtaining an
accuracy within
15% of 9 poi nt
analysis for 02
in a large
duct (524).
!6O
.¥
1:'10
I.:
:i
j.o
30
20
o
I
) . , 6
I't..."b., 01 Probrl UI,I,r.d
10
.
the same techniques should give representative values of emissions.
However, until a careful study is made the accuracy of this technique
cannot be defined. (Compare Section 4.2.6, following.)
The emission of material from a combustion source is
described by the general equation,
+ +
Ea = fA Ca v . n dA
where Ea is the emission of material (a), Ca is the concentration of
(a), v the flue gas velocity along the duct, A is the cross sectional
area of the duct, and n the unit vector normal to A. It follows from
Equation (1) that C and v are coupled if neither is constant across
a
the duct; hence, they should be measured together.
A traverse using a continuous oxygen or carbon dioxide
analyzer can quickly determine the extent of infiltration air strati-
fication. If the gases are not significantly stratified, the pol-
lutant gases may be assumed to be well mixed and the concentration
43
WALDEN RESEARCH CORPORATION
-------
determined from a single sampling point.
dent to sample at more than one point as
pollutant gases.
However, it is more pru-
a check on the mixing of the
For rigourous measurements a set of replicate samples
should be taken at traverse points and a statistical analysis performed
in order to establish the flow pattern and estimate of the residual
error. For subsequent measurements in this duct, fewer sampling points
can be used and the accuracy predicted (537).
Methods for the determination of C and V for irregular
a
distributions are discussed in Sections 4.2.6 and 4.3, following.
4.2.6 Variance of Concentration in Large Ducts
A major problem in the high precision determination
of pollutant emissions is the variation of species concentration which
may exist in a large duct as a result of air infiltration and poor
mixing (stratification). In the course of NAPCA's extensive studies
of coal-fired power plant effluents (548) many C02 concentration pro-
files were obtained by traverse of large ducts at different sampling
locations. The following discussion, based upon a random selection.
of this test data and, therefore, incomplete, outlines the magnitude
of the problem and the influence of both sampling location and equip-
ment type on the results obtained.
Typical sampling locations are illustrated in the
power plant schematic (Figure 4-4). The most common locations are
at the entrance and exit of the dust collection equipment. The
statistical treatment adopted was to calculate the mean CO con-
. 2
centratlon for each traverse plane, the standard deviation from the
mean, and the coefficient of variation (CV = 100 cr/mean). Two ex-
amples, illustrating relatively homogeneous and stratified flows
are given in Figure 4-5. The observed coefficient of variation of
44
WALDEN RESEARCH CORPORATION
-------
..J:::o
U"I
pulverizer bottoms
~
~
c
",
Z
::0
",
(/J
~
::0
o
:r
o
o
::0
C3
::0
~
o
z
infiltration air
1
bottom ash
@
[J
[l]
from GJ
economizer
air
pre-
heat
Combustion air
infiltration
air
w
collected ash
LS.
ppt
fly ash 5
S025S03r:1
t L? I
GJ
e
occasionally the E.S. precipitator may be upstream of the air-preheater
Plant input data
Flue gas sampling points
Figure 4-4.
Schematic Large Combustion Unit
Stack
-------
"HOMOGENEOUS"
1.03 1.00 0.98 1.00 1.01 r
1.03 1.01 0.98 1.00 1.01 3"3"
-*-
1.00 1.01 0.98 1.01 1.01
~ 221 ~
Ave C02 = 11.7%
CV = 1. 5%
II STRATI FlED"
.74 .92 1.0 1.0 .97 .95 T
1.01 1.02 1.01 1.00 .90
.82 418"
.75 1.05 .99 1.02 1.01 .90 ~
.82 1.00 .93 .93 1.01 .86
~ 14110" ~
Ave C02 = 12.6
CV = 9.3%
Random selection from six plants.
Figure 4-5.
Normalized C02 Traverse Data at Dust Cpllector of Coal-
Fired Power Plants.
'., .
- -- -
46
WALDEN RESEARCH CORPORATION
-------
the C02 concentration is given in Table 4-2 as a function of sampling
location. The relatively low value of CV at the outlet of the dust
collector, presumably the result of good mixing, suggests this to be
the location of choice for both simplicity and good precision. The
relatively high value of CV at the inlet to the dust collector may
be associated with air infiltration at the air preheater, a common
occurrence.
The most frequent sampling locations for emissions
determinations are the inlet and outlet of the dust collectors. The
coefficient of variation of the C02 concentration at these locations
is given in Table 4-3 as a function of equipment type. With the ex-
ception of Plant No.3, CV at the inlet is relatively large compared
to the outlet values.
Conclusions which may be drawn from these data are:
1. C02 (or 02) traverses are extremely valuable for
selection of sampling locations and determination of the number of
samples required for a high precision emissions determination. C02
traverse data may be utilized for normalization of 50 concentrations
x
made at the selected locations.
4.3 Flue Gas Flow Measurement Methods
4.3.1
Introduction
The emission of a material from a combustion source
is described by the general relationship
+ +
E =1 C v . ndA
a A a
(l)
where Ea is the emission of material (a). Ca is the concentration of
(a), A is the cross sectional area of the flue or duct in question,
v is the velocity and n is the unit vector normal to A.
47
WALDEN RESEARCH CORPORATION
-------
TABLE 4-2
OBSERVED COEFFICIENT OF VARIATION FOR C02 TRAVERSE
AT VARIOUS SAMPLING LOCATIONS
Sampling Position ro(F) No. of Traverse Points C02 (CvL %
* 2400 12 4.0
Furnace (1) >4.6
12 5.2
Economizer Inlet (2) 860 8 3.8 >3.8
8 3.8
Economizer Outlet (3) 360 24 7. 1 >6.2
(Dust Collector Inlet) 24 5.4
Outlet of Dust Collec- 350 18 3.2 >3.2
tor (4) 18 3.2
(2) Sampling should be conducted at the outlet of dust collectors in the
absence of other information.
(3) For simplified methods, single point sampling at the dust collector
outlet appear to be feasible (CV < 5%).
*
See Figure 4-4.
48
WALDEN RESEARCH CORPORATION
-------
TABLE 4-3
OBSERVED COEFFICIENT OF VARIATION FOR C02 TP~VERSE FOR VARIOUS COAL-FIRED PLANTS
Plant Dust Collection Sampling No. of
No. Type of Boiler Firing Equipment Location Traverse Points C02(CV}, %
1 Horizontally Opposed C I 24 9.3
o 12 2.3,1.4
2 Cyclone E I 24 4.6
o 24 3.2
3 Spreader Stoker C I 18 1.5
o 9 1.02
4 Corner C,E I 18 8.8
o 12 0.97
.p.
\0 5 Vertical C,E I 24 7. 1
o 12 3.2
~
C
"..
Z
::0
"..
C/!
~
::0
(")
::r
(")
o
::0
~
::0
~
o
z
C = cyclone
E = electrostatic precipitator
I = dust collector inlet
o = dust collector outlet
-------
If the concentration is not a function of position,
i.e., no stratification exists, then Equation (1) can be written as:
E = C . V
a a
(2)
where
-+- -+-
V = fA v . ndA
(3)
which is the volumetric flow.
This discussion is limited to determining the volumetric
flow of flue gases from fossil fuel combustion sources.
The technology of measuring subsonic gas flow is not
rapidly changing at this time, and the existing techniques have been
well documented, including the magnitude of expected errors for lab-
oratory conditions using clean, dry gas. Where field conditions allow
the rigorous implementation of some standard test procedure, e.g.,
ASTM, ASME, ASH RAE, 8SI, etc., the flow measurement should be ac-
curate to well within 5%, with the accuracy dependent upon the parti-
cular measuring instrument used.
In many cases, flue gas flow measurements will have
to be made in gas streams where the flow may be asymmetric, varying
in time, particle laden, or of low velocity. The accuracy of measure-
ments made under these conditions cannot be estimated with confidence.
In fact, the accuracy will depend significantly on the skill and
care of the investigator.
4.3.2 Pitot Tube Traverse
The pitot-static tube traverse is the most commonly
used method of measuring gas flows in flues and ducts. In this tech-
nique, a pitot-s~atic tube (see Figure 4-6) is inserted through a
breaching in the duct wall and a number of velocity measurements are
50
WALDEN RESEARCH CORPORAtIQ"
-------
A
?!.', 8U
?
.1 r8JIAM[TU'
idj .'J-;I(J
16 .
-. T
:,"
32 ROD
5"; 16 D
. - - - - --
A
8 HOLES -0.04 DIAMETER EQUALLY
~,SPACED FREE FROM BURRS
'~.
SECTION A-A
INNER TUBING
~' a.D. x 21 BaS GAGE COPPER
OUTER TUBING
fu'O.D. 18 Bas GAGE COPPER
Figure 4-6.
A standard pi tot-static tube (525).
51
WALDEN RESEARCH CORPORATION
-------
made across the duct in a plane perpendicular to the flow direction.
From the velocity measurements and locations, the integral in Equation
(3) is evaluated by numerical or graphical techniques to give the volu-
metric flow.
4.3.2.1
Velocity Measurements
Interpretation of velocity measurements using
a pitot-static tube depends upon the relationship
2
liP = K 1/2 pv
(4)
where liP is the differential pressure between the static ports and
the pitot tube; v is the gas velocity; p is the mass density of the
gas, and K is a calibration factor of the pitot-static tube. Values
of K have been measured for various types of pitot-static tubes and
for most engineering work (gas speed errors within 1%) K can be taken
as unity. Some observed values of K are presented in Table 4-4.
Equation (4) and the values of K given above
apply only to pitot-static tubes with a zero angle of attack. The
effects of non-zero angles of attack are shown in Figure 4-7. Where
the direction of flow is not known within 200 to 300, or where turbu-
lence exists which changes the angle of attack within these limits,
the hemispherical nosed probe will still measure gas velocity to within
2.5% (v ex pl/2).
A practical limitation of the pitot-static
tube for measuring flue gas velocity exists for low velocities due
to the limiting sensitivity of the manometers which can be conveni-
ently used in field work. For a calibrated inclined manometer, the
limiting threshold is about 15 feet per second for an error within
1% (606). Measurements of lower velocities will result in larger
errors.
A Staubscheibe pitot tube (Figure 4-8)
which has a K greater than unity is sometimes used to measure low
velocities.
52
WALDEN RESEARCH CORPORATION
-------
TABLE 4-4
PITOT-STATIC FACTORS AT LOW AIR SPEEDS (606)
Air Speed Tapered-Nose Tube Hemispherical-Nose Tube
(Air at 150C * *
and 760 mm Reyno 1 ds No. K Reynolds No. K
Hg)
(ft/sec)
2 330 1 -020 335 1-055
4 655 0-989 670 1-006
6 985 0-995 1000 1 -001
8 1310 0-992 1335 0-996
10 1640 0-991 1670 0-992
12 1970 0-992 2005 0-991
14 2295 0-995 2340 0-992
16 2625 0-998 2675 0-996
18 2950 0-999 3005 0-999
20 3280 1-000 3340 1 -001
*
The Reynolds numbers for the two instruments at the same air speed
differ slightly because the outside diameters were not the same;
they were 0-307 in for the tapered nose and 0-312 for the hemi-
spherical nose.
53
WALDEN RESEARCH CORPORATION
-------
(L0
'--...
(L"
Figure 4-7.
~
---
1'15
1.10 --
I
I '
----1- u~ -----1---
--l--t-- : -~--- -+--- --
I
1.05
1.00
I
I
I ' i I I
-p~=- dl~erenl~a~pr~:: a~~~w-u~---r-
Po = differenllal pressure al O. yaw I ~ Ellipsoidal nose
)'90 -- I I I -----~
I' I
I
f)'9~
0'85
,
30
':'ngle nf yaw >/I. degrees
Effect of yaw on pitot-static combinations
(606)
--=1
-':=~-=-~
Figure 4-8.
Staubscheibe Pi tot Tube
54
WALDEN RESEARCH CORPORATIQ~
-------
However, K varies appreciably with velocity
for this device as shown in Figure 4-9, and each tube must be cali-
brated over the velocity range of the intended measurement.
1.59
~-In.tubes
;;...-
1.58
1'57 t
1.56
K <==
1.55
~54
1.53
'.52
0 10 20 30 40 50 60 70 80
Air speed. tt/see
Figure 4-9.
Calibration Factors, K, for Staubscheibe Pi tot Tube (606)
Since K varies rapidly with velocity. it is likely that intrinsic er-
rors will exist for measurement of turbulent flows.
The Staubscheibe tube is frequently recommended
for velocity measurements in particulate laden gas streams because it
does not have small static ports which may be plugged (607,525). How-
ever, purging systems can be constructed for the pi tot-static tubes
which force compressed air through the static ports before making a
velocity measurement (608). In locations with particulate-laden gases
where compressed air is readily available, the purged pitot-static
tube is preferable to the Staubscheibe pi tot-tube.
4.3.2.2 Traverse
Measuring gas stream velocities at various
points across. a flue or duct is a velocity traverse. From the geom-
etry of the ducts and the geometry of the traverse point, the gas
velocities may be used to calculate the volumetric flow. This
55
WALDEN RESEARCH CORPORATION
-------
calculation is the evaluation of the integral in Equation (3). The
integral may be evaluated graphically; however, the most common tech-
nique is to divide the test section into a number of equal area zones
and determine the mean velocity in each zone. The velocities for each
zone are averaged and the volumetric flow is given by
v = vA
(5)
where v is the average velocity and A is the area of the test section.
Discussion of techniques for dividing test
sections into equal area zones will be limited to ducts of circular
and rectangular cross section, since sampling in ducts of other forms
is rare. For ducts of unusual shape, the volumetric flow should be
determined by graphical integration for accurate results.
The number of test zones that a flue is
divided into will depend upon the uniformity of velocity distribu-
tion and the accuracy desired, not upon the size of the duct, since
for any two similar ducts (different only in size) with similar velo-
city distributions, an equal number of velocity readings will be re-
quired to determine the average velocities with equal accuracies.
However, in practice, the size of the pitot tube will limit the
number of velocity measurements in small ducts.
4.3.2.2.1
Equal Area Division of Ducts of
Circular Cross Section
The tangential method divides
a duct of circular cross section into n equal zones, a cricu1ar
central zone and (n-1) annular zones (Figure 4-10). Each zone
is divided into two equal area annular parts and the velocity
measurement is made at the radius of the boundary between the equal
area parts. The mathematical derivation of the division of a circu-
lar cross section by this technique is found in Ower (606). The
method of dividing a circular duct by this technique is shown in
Figure 4-10.
56
WALDEN RESEARCH CORPORAflON
-------
-=---. 11
.. =rr ~
\ + 0: ~ ~
. 1"::; ~ "!
-L ~ ~ 'I i
~.li.u
IIOT<: .Inaicat.. Point..
0", location 0"
Sampling Tub.
.CROSS SECTION OF CIRCULAR CUCT"
Formula for determining 10Lation of points in circular duct
- ~2R'(2P-C
1'.-
11
where 1'p = distance from center of duct to point p
R = radius of duct
p = sampling point number. To be numbered
from center of duct outward. AlI four points
on same circumference have same number.
11 = total number of points
::'IloTE: 1'. wilI be in same units as R.
Example: Duct radius = R; 20 points tota!.
Distance to point 3 = r".
1',= ~2R.(2~3-1) = ~2~5 =~ .5R"
1'.= 0.707R
Figure 4-10.
Tangential Method for Duct Division (609)
The log-linear method is an al-
ternative which gives higher accuracy. The circular cross section is
again divided into equal area annular zones, but the velocity is not
arbitrarily measured at center of area of each zone. Instead, the
measurement points are calculated on the basis of an empirical
analysis of the flow through circular pipes. The development of this
method, including the determination of the measurement points may be
found in Ower (606). The results are summarized in Table 4-5.
For fully developed flow,
Winternitz (606) found that a four point log-linear traverse gave
an error of less than 0.5%; whereas, the 10 point tangential method
overestimated the mean velocity by about 1%. For non-fully developed
flow, the 10 point tangential technique was somewhat better than
the four point log-linear, but, an eight point log-linear method
was superior to the 10 point tangential method. The 6 point log-
linear method will give results with an error less than 1% in flow
distributions as asymmetric as shown in curve A of Figure 4-11 (606).
57
WALDEN RESEARCH CORPORATION
-------
TABLE 4-5
LOCATION OF MEASURING POINTS FOR LOG-LINEAR METHOD (606)
No. of Measuring
Points Per Diameter
Distance from Wall in Pipe Diameters
4
6
0.043, 0.290, 0.710, 0.957
0.032, 0.135, 0.321, 0.679, 0.865,
0.968
0.021, 0.117, 0 '184, O' 345
0'655, 0,816, 0'883, 0.979
0.019, 0.076, 0,153, 0,217, 0.361
0'639,0'783, 0'847, 0'924, 0.981
8
10
Because the log-linear method pro-
vides better accuracy than the tangential method for an equal number
of measurements, it is recommended for velocity traverses in ducts
of circular cross section. At the present time, the log-linear method
is in general use in Britain (610,537).
Because of velocity gradient ef-
fects across the diameter of a pitot tube head, pitot tubes should
have a diameter less than 1/30 the diameter of the duct when the al-
lowed error is to be less than 1% (606).
4.3.2.2.2 Equal Area Division of Ducts of
Rectangular Cross Section
The technique for dividing a
rectangular duct into equal area zones is to divide the section into
a number of geometrically similar rectangular zones and to measure
the velocity at the centroid of each zone. The rules for the num-
ber of zones are more arbitrary than for circular ducts. As for
58
WALDEN RESEARCH CORPORATION
-------
'.61
-t- --- + -- /X->O<\x
i ! AI: \
~'- , ru/'y~~+\
'.0 +...-4' +-+;-*-;-+_. -
Xx/Xx i i I \
"" "---\t> ._-. I +
0.6 X
o 0.2
'.4----
0.4
0.6
0.8
'.0
Ratio of wall dIStance 10 diameter, V/d
Fi gure 4-11.
Types of Asymmetric Velocity Distribution in Pipes
circular ducts, the accuracy of the traverse will depend on the uniformity
of the flow and the number of velocity measurements; however, it is a
convention to increase the number of sampling points with duct size.
The recommended number of test points are given in Table 4-6 as a func-
tion of the cross section area of the duct.
British Standards 1042 (615) rec-
ommends a division into at least 16 zones, with five velocity measure-
ments in each corner zone, and three velocity measurements on each
wall zone. The velocity in each zone is averaged before averaging
the velocities over the duct. No zone should be greater than 36 inches2,
which means that more than 16 zones are necessary for ducts with cross
sectional areas over 4 feet2. The BSI Traverse plan is shown in Fig-
ure 4-12.
The National Engineering Labora-
tory (U.K.) has found that errors of 2% or more can occur for certain
types of asymmetric flow distributions when the 16 part, 48 point
traverse is used (606).
59
WALDEN RESEARCH CORPORATION
-------
TABLE 4-6
TEST POINTS FOR RECTANGULAR DUCTS
A. Haaland (607t
Cross Section Area
Squa re Feet
Number of Test Points
2 to 25
4
12
Less than 2
Greater than 25
20 or more
B. ASTM (614)
Inside Cross Sectional
Area of Flue, ft2
Minimum Number of Test Points
*
1
to
3
4
6 - 24
2< to 12
12<
More than 24
60
WALDEN RESEARCH CORPORATION
-------
f--- -
I
r
I
I
b
j
r
---------..- -
,)
__m u~
I
L
-~ ~ .. +
-+- .... + .- + + + -+
.. .... ..... .....
+ .- + + + + + +
+ -+- + + + + + -+-
+ + -.. +
-.. -+- + + + -+- + +
+ -+- + +
r----_O/4 b/24
I
b/4
1
~,
0/24
~ -t- (//24
T
Arrangemer,t of pitot tube
positions for each corner
panel of the airway
Figure 4-12.
Traverse Plan for Rectangular Duct (610)
Because of velocity gradient effects
across the diameter of the pitot tube head, it is usually recommended
that the pi tot tube diameter should be less than 1/30 the shortest side.
This value is probably based on the work done with circular ducts
(see previous section).
4.3.2.2.3
Flow Variations During Traverse
Frequently the flue gas flow from
a combustion source will not be constant during the time needed to
traverse the duct. A reference pitot-static tube should then be
pl~:~ed in a fixed positi~n in the duct to provide a reference velocity
61
WALDEN RESEARCH CORPORATION
-------
pressure. With each reading of velocity pressure during the traverse,
the reference pressure should be recorded. Each individual velocity
pressure reading of the traverse is then reduced to the standard ref-
erence pressure Ps by multiplying it by the ratio Ps/P~, where P~ is
the reference pressure observed concurrently with that particular tra-
verse velocity pressure.
4.3.2.3 Calculations of Volumetric Flow from Velocity
Pressure
The relationship between velocity pressure
and velocity is given by Equation (4). However, this equation is not
very convenient for calculation. A more convenient form is Equation
( 6) be 1 ow :
v = 2.90 J2~.~~ T
(6)
where
v = gas velocity (feet/second)
h = velocity pressure (inches of water)
T = absolute temperature (degrees Rankine)
P = absolute pressure of flue gas (inches of mercury)
G = specific gravity of flue gas referred to that of air
K = calibration factor of the pitot tube (for a standard
type pi tot-static tube K may be taken as unity)
The volumetric flow, if the gas temperature is
constant across the duct, is found from Equations (5) and (6).
V = 2.90A~2~.~~ T
~ I Ih
where the term l/n r~is the average of the square root of the velo-
city heads for an eqUa13area trarerse of n test points. V is the
volumetric flow in feet -second- , and A is the inside cross sec-
tional area of the duct in feet2.
62
WALDEN RESEARCH CORPORATION
-------
4.3.3 Low Velocity Flow Measurement
4.3.3.1
Micromanometer - Pitot Tube
When measurements of gas velocities below about
15 feet per second and accuracies better than 1% are necessary. a microm-
eter may be used to extend the gas velocity range of a standard pitot-
static tube down to about 5 feet/second with an accuracy within 1%. The
difficulty is that micromanometers (Figure 4-13) are more delicate than
simple inclined manometers, and taking the velocity head readings is more
tedious.
Electronic manometers are available which com-
bine the sensitivity of micromanmometers with the operational simplicity
of the.inc1ined manometer. A number of these instruments are briefly
described below.
4.3.3.2
Electronic Micromanometers
The following companies manufacture electronic
pressure measurement equipment (not complete listing):
1.
MKS Instruments, Inc.
45 Middlesex Turnpike
Burlington, Mass.
2.
Datametrics Division
CGS Scientific Corporation
127 Coolidge Hill Road
Watertown, Mass.
3.
Datasciences Corporation
Instrument Systems Division
9601 Canoga Avenue
Chatsworth, California
The first two instruments use a pressure sensor
that depends upon a change in capacitance due to a change in diaphragm
geometry which is a function of the pressure differential across the
diaphragm. The last instrument is similar except that the change in
diaphragm geometry is detected by a change in reluctance.
63
WALDEN RESEARCH CORPORATION
-------
-- ---
LEVELING
SCREWS -
CAL8RATt:D-- DIAL-HANDLE \
0.001. DIVIIIONS
.. . - - - ---- -
COARSE
ELEVATION
SCALE
0.1. DIVISIONS
.
,
COARSE FLUID
LEVEL INDICATING
TUBE
INCLINED
TUBE
20'1 SLOPE
,02.1.0. ' ,
ELEVA1'ING
RIDER
- SCREW
20 THO. I IN.
Fi gure 4-13.
NACA Micromanometer
Both the MKS and CGS instruments have pres-
sure ranges suitable for measurement of air velocities (at STP) from
less than 2 to about 200 ft/sec using a standard pi tot-static tube.
The price range of these instruments is $2,500 to about $3,500 de-
pending on digital or meter movement display.
The sensitivity of the Datasciences instru-
ment does not offer much improvement over an inclined manometer. The
most sensitive full scale deflection corresponds to an air velocity
of about 20 ft/sec, which provides a useful low limit of about 6
ft/sec.
64
WALDEN RESEARCH CORPORATION
-------
4.3.3.3 Hot-Wire Anemometer
For low velocity measurements in non-particulate
laden gas streams, hot-wire anemometers may be used. Shielded hot-wire
anemometers are commercially manufactured that will return to calibration
when the shield is wiped clean of particles. H. Trissley & Co. Ltd.
makes an instrument of this type which is accurate to within 0.5% for
velocities of 5 down to less than 0.4 feet/second.
temperature of the
the manufacturer1s
Hot-wire anemometers can be sensitive to the
gas streams; therefore, it is necessary to stay within
suggested temperature ranges.
4.3.3.4
Vane Anemometers
Vane anemometers are also employed to measure
low speed flow. Figure 4-14 shows two configurations of vane anemometers.
The dial of a vane anemometer is marked off in length units (usually feet)
of air passing by the instrument. To measure velocity, it is necessary to
take the time interval for a number of feet of air to pass the instrument
and calculate the velocity.
Vane anemometers as shown above are suitable
for measuring velocities in the range of about 1 to 10 feet/second. How-
ever, each instrument must be calibrated and the calibration should be
frequently checked. The velocity measured must be corrected for the gas
density by the following:
v = vi M
where v is the true velocity, v. is the observed velocity, p is the density
1
of the gas measured and Po is the density of the gas used for calibration.
For errors in volumetric flow less than 1%, this
instrument should not be used in ducts smaller than about 8 times the
diameter of the instrument (606j.u- Traverses are made according to
either the tangential rule or log-linear scale. Table 4-7 shows the
65
WALDEN RESEARCH CORPORATION
-------
.--- -- ------ _.._--
------.---"---
~
':,.
,'" '¥. 't,
Figure 4-14.
Vane Anemometers
limiting conditions for different sized anemometers in terms of the
ratio of the duct, D, to the anemometer diameter, d.
4.3.3.5 Plate Orifice Meter
Volumetric flow measurements can be made, es-
pecially in small ducts, by measuring the differential static pressure
across a flow constriction. The common types are plate orificest
shaped nozzles, and venturi tubes. The latter two types of constr1c-
tions are not readily applied to manual methods of source testing.
The plate orifice, Figure 4-15, is a much simpler device and may be
66
WALDEN RESEARCH CORPO~.
-------
TABLE 4-7
OIAMETER RATIOS FOR VANE ANEMOMETERS (606)
Number of Oistance of Nearest Point Minimum Old Rati 0
Measuring to Wa 11
Points Per Tangential Log-Linear Tangential Log-Linear
Diameter Rule Rule Rule Rul e
4 0.0670 0.0430 7.5 11 .6
6 0.0430 0.0320 11.5 15.6
8 0.0320 0.0210 15.4 23.8
10 0.0260 0.0190 19.8 26.3
adapted to small flues. There are two types of plate orifices, the stan-
dard and the calibrated. Standard orifices have been well investigated
and when accurately constructed do not need calibration. The ASME stan-
dard orifice is reported to be in error less than 1% (611); while the
British Standard orifice has a specified accuracy to within 1% in 95
out of 100 cases (606). Further details about these devices may be
found in References 609 and 610. The standard orifices have the draw-
back that in flue gases their carefully machined sharp edged orifices
will corrode and change performance characteristics.
----
~---
dl dz
- -L~=- - H_-
---------
-- - -- ----
--~~~~~~~=:----- ---t-
-, ..,:=.:----
fF~;:::::
'.:- -_J ---
Figure 4-15.
Flow Through a Plate Orifice
67
WALDEN RESEARCH CORPORATION
-------
A more practical device is the calibrated ori-
fice. This device is made by placing a thin orifice disc across a duct
and calibrating the pressure differential against a standard technique
such as a pitot traverse. It is convenient for measuring pulsing and
time-changing flow because volumetric flow is presented in one reading.
4.3.4 Comparison of Volumetric Flow Measuring Methods
A comparison of volumetric flow measuring methods is
presented in Table 4-8. A standard pitot-static tube traverse should
be used wherever possible. When this is impossible and another method
is selected, the instrument should be carefully calibrated if accurate
results are expected.
4.3.5 Indirect Determinations of Volumetric Flow
4.3.5.1
Stoichiometric Method
An indirect technique to compute flue gas
flow is by material balance. For these computations, it is necessary
to know fuel composition, fuel flow rate, combustion air and exit
particulate composition (to determine unburnt carbon content). This
technique is based on the conservation of matter and the gas laws.
The computational methods are aptly described i~ any standard stoich-
iometry textbook [e.g., Lewis and Radasch (612)}: Whereas such a technique
,
certainly has a theoretical advantage in that no velocity probes are
necessary, it does require accurate measurements of the input rates
which may be difficult to obtain.
4.3.5.2 Dilution Technique
Dilution is an indirect technique that can
be used to obtain volumetric flow through a duct. In this technique
an inert tracer gas of known concentration is injected into the flue
gas streams at a known flow rate. At a point far enough downstream
for complete mixing of the tracer with the flue gas, the flue gas
is sampled for the tracer gas concentration.
68
WALDEN RESEARCH CORPORATION
-------
TABLE 4-8
COMPARI~ON OF VOLUMETRIC FLOW MEASURING METHODS
Applicable Flue Expected Error For
Vel oci ty Size (Character- Volumetric Flow (Well
Measurement Method Range istic Length) Developed Flow)
Standard Pi tot- a, >15 30x Pitot Dia. 2%
ftjsec
b, >5
ftjsec
Staubscheibe a, >12 Same as Above
Pi tot Traverse ftjsec
Hot-Wi re >0.4 Several Times 2%
0'\ Anemometer ftjsec Larger Than
1.0 Probe Size
Vane Anem- >1 ft/sec 8x Diameter
ometer <10 ftjsec of Anemometer
Standard Plate Practically all '\,1%
Orifi ce
Rema rks
:E
»
r
C
I'TI
Z
::u
I'TI
C/I
I'TI
»
::u
(')
:J:
(')
o
::u
(3
::u
~
o
z
Calibrated
Plate Orifice
Same as Above
a, Inclined manometer,
b, Micromanometer,
Pitot-static tubes are
the practical standard
for velocity measurements.
a, See above
Accuracy depends on cali-
bration
Clean gas stream required,
fast response time, re-
quires frequent calibration
low gas temperature
Frequent calibration & clean
gas, stream required
No calibration needed, short
lifetime in flue gases, must
be constructed for particu-
lar flue
Accuracy depends on calibra-
tions, more rugged than
standard plate orifice, con-
venient where flow cycles,
introduces pressure drop in
flow system
-------
The volumetric flow in the duct is then calcu-
lated from the equation:
v = qi (Ci/Cs - 1)
where V is the flue gas volumetric flow, q. is the tracer gas injection
1
flow rate, C. is the tracer gas concentration at the injection point
1
and Cs is the tracer gas concentration at the sampling point.
To utilize this technique, Cs must be measured
accurately, since to determine V within 1%, Cs must be measured within
1 %.
70
WALDEN RESEARCH CORPORATION
-------
5. SAMPLING METHODS FOR SULFUR OXIDES
5. 1
Introduction
The key to satisfactory analysis is collection of a representa-
tive sample, selection of the proper probe, collecting medium, and analy-
tical method. Oxidation of S02 + S03 (catalytically) is a factor to be
considered both in the probe and in the collecting medium employed. For
example, if the probe is heated to very high temperatures, catalytic oxi-
dation of S02 can occur (526). Condensation of H2S04 and water in the
probe can lead ,to low S03 (and S02) values. Interferences with the analy-
tical method may be avoided by the proper choice of collecting medium.
Thus, the sampling system (probe, collecting medium, analytical method,
etc.) has to be considered as a whole in order to obtain truly precise
and accurate results. A complete sampling system for the sulfur oxides
designed during this program is given in Appendix 4.
These factors are discussed in detail below.
5.2 Probes
The material of construction is a significant factor in select-
ing probe components. The selection of a sampling probe must include
consideration of the mechanical properties of the equipment as a func-
tion of the temperature range and chemical environment to be encountered.
Other factors involved in the selection of flue gas probes such as cost,
frequency of maintenance, accessibility. etc., have been mentioned by
Dixon (505). The British Standards Institution (379) and ASME (502) de-
scribe probes, collecting equipment and methods commonly used for flue
gas sampling. A water-cooled multipoint probe designed by Chojnowski
(500) can be used for taking three simultaneous gaseous samples.
Probes may be divided into two classes, viz., simple* and
temperature-controlled. Provision must be made for a preliminary
*
A simple probe is not temperature controlled.
71
WALDEN RESEARCH CORPORATION
-------
filter, (quartz or pyrex wool) to remove combustion generated particu-
lates which may interfere with the analysis or change the composition
of the flue gas. Typical power plant sampling points are described
in Table 5-1. (Also see Figure 4-4.) Typical probes are described
in more detail in the following sections.
TABLE 5-1
TYPICAL POWER PLANT SAMPLING POINTS
Sampling Location
Temperature Range
(oF)
Purpose
*
In furnace above burner
Before air heater
700-800
(More important for NOx
process control
studies)
For checking combustion
control
2000-3000
Exit of air heater
300-400
Emission inventory or
heat balance
*
Not in mission of present program
5.2.1
Simple Probes
A simple probe may be a stainless steel tube (170) or
a stainless steel tube with a quartz or pyrex insert (76). The lat-
ter has been used with some success where the flue gas temperature is
in the vicinity of 5000F (76). The problem commonly encountered in
sampling with an unheated probe is that cooling of the flue gas in the
probe and/or sampling lines may lead to condensation. For example,
for an S03 concentration of 10 ppm, the dewpoint of the flue gas is
2700F (Figure 5-1). In a series of tests at a coal-fired power
plant, up to 50% of the total S03 was found in the probe and sampling
lines (527). Therefore, the probe and sampling line have to be
72
WALDEN RESEARCH CORPORATION
-------
330
310
/
/
V
- P. MULLER (CALCULATED) V
0 THIS WORI< EXPERIMENTAL - - -- V
. PARTIAL PRESSURE MEASUREMENT
~
/ 0
/
/ ...,
/
.~/;'
/
,/ - -
,,"'"
/' --
,,/
V --
290
IL
. 270
f-"
Z
o
Q.
~ 250
o
Z30
ZIO
190
001
0.1
10 10
SOJ (H, SO.) IN FLUE GAS, PPM
100
Figure 5-1.
Sulfuric Acid Dewpoint as a Function
of S03 Concentration (502)
73
WALDEN RESEARCH CORPORATION
-------
quantitatively washed into the collection vessel. As a result of this
cumbersome procedure, the use of this probe is not recommended.
5.2.2 Temperature Controlled Probes
5.2.2.1 Heated
Most of the sampling in power plants is done
in the duct work after the air heater where the temperature is in the
vicinity of 3500F. For this application, the heated probe described
by ASME (502), BSI (381) (Figure 5-2) is most suitable (80,81,119,217,
165,221,218,160,215,207). In operation, the probe (quartz or pyrex
tube inside a stainless steel case) is electrically heated to a tem-
perature (500-6000F) well above the acid dewpoint. Heating of the
sampling lines from the probe to the collector is generally neces-
sary to prevent condensation. A heated probe is recommended for flue
gas temperatures less than 8000F.
5.2.2.2 Water-Cooled
When sampling at temperatures above 8000F,
several changes in the probe design are needed. Pyrex inserts have
to be replaced with quartz and the probe has to be cooled to ap-
proximately 600-7000F to prevent changes in the composition of the
flue gas. Water cooled probes (Figure 5-2) with quartz inserts have
been used by B&W (425), BSI, and ASME. These water cooled probes
have been used for taking HOx samples in the furnace where tempera-
tures approach 25000F. (This problem is not considered further in
this program, since it exceeds the temperature range of interest.)
5.3 Collection Methods for S03
5.3.1
Introduction
Most literature references on collection methods
for S03 deal with concentrations less than 50 ppm, characteristic
74
WALDEN RESEARCH CORPORATION
-------
2" PIPE PLUG DRILLED TO FIT INSULATION
I" STAINLESS STEEL TUBE \
!!. .. vveOR " t" STAINLESS STEEL \ RESISTANCE WIRE ....,TAPE INSULATION
/ . , e '"BE ' " ,
~~E~~~," . ~. . ~jeuil~~~$-\~~ijt~~~-'o:O~~~~~"'~1\T~
,( , . I I I I
....../ I I,'
I~ I- t WELD - I 12 MM vveOR I / 16 MM vveOR
I I I I FILTER I
~n_----6"--- ---4c-- --- - u- -- - - -12"--- - - -- -- --- - 4.----2'-6"- - -- ~ u - - - - 6"- - - ---J
~--------- - --- - -- - -- -- ------ - - ---4'-6"------- -- -- -- - - -- --- -- - - - - - -- - - --J
STAINLESS STEEL VYCOR SAMPLING PROBE
~-_. --
Water connections
70
Water Jacket
Leads~
Transpa-rent silica tube
porcelain beads covered with asbestos paper
Asbestos wrapping
Spherical silica
joint equivalent
to size NO.18/9
of 8.5. 2761
I~
-L
~
approx.
'I2o.d.
Nichrome
\ 72
winding 26 5WG, (O.46mm)
Sampling Probe @
. Heated probe unit with water-cooled jacket
Alumina cement
Figure 5-2.
Heated and Cooled Probes
75
WALDEN RESEARCH CORPORATION
-------
of fossil fuel plants. These measurements have been made for diagnostic
purposes (i.e., corrosion control, etc.) since the 503 concentration is
only about 1-2% of the total sulfur oxides emitted. This low 503 con-
centration limits the collection techniques to those which concentrate
the sample, e.g., absorption, rather than grab sampling techniques.
A summary of general sampling methods for gaseous pollutants is given
in Table 5-2. The other principal S03 sources are sulfuric acid plants
and smelters. Techniques used for sulfuric acid plants have been de-
scribed by MCA (95) and are generally those developed for fossil fuel
combustion. No new collection techniques appear to have been developed
for the smelting industry (496).
The major problem in quantitative collection of S03 in
power plant effluents is that 502' present in a large excess, is easily
oxidized, leading to high S03 values and/or poor precision and accuracy.
Most collection methods for the sulfur oxides have been based upon
physical separation of S02 and S03 (by differential absorption) to re-
duce the magnitude of the oxidation problem. Techniques in which both
oxides are collected without separation (such as absorption in caustic
solution) hQve not been widely used since accuracy is generally poor.
S03 collection techniques may be divided by approach
into two different classes: absorption and condensation methods. The
former depend on the solubility of H2S04 in aqueous solutions. In-
hibitors, such as alcohols, have generally been added to prevent the
oxidation of dissolved 502' Condensation methods depend upon cooling
of the flue gas (controlled or uncontrolled) to a temperature where
H2504 aerosol is formed and subsequent collection on a sintered glass
frit or filter paper.
In principle, solid adsorbents, reactive or unreactive,
may be used for collection of S03' We have found no references to
analyses of effluents for S03 utilizing this technique.*
*
This approach may indeed be useful for SO and/or NO measurements
but we do not recommend attempts to devel~p such a t~chnique for S03'
76
WALDEN RESEARCH CORPORATION
-------
=E
~
C
fT1
2
::0
fT1
CIJ
~
::0
()
:r:
()
o
::0
2J
::0
~
o
:z
TABLE 5-2
COMPARATIVE SUMMARY OF SAMPLING DEVICES
(542)
.......
.......
TYPE OF PRINCIPLE INSTRUMENTS APPLICATION METHOD OF SKILL IN SKILL IN REMARKS
CONTAMINANT OF METHOD QUANTITATION OPERATION QUANTITATIOJ
Absorption Bubbler trains Chemic/li. Some Efficient for collection of reactive gases. Efficieney..
Analysis for fine particulate matter and gases of low or s""
solubility. Simple commercially available equ;pmst.
Absorption Spray co::tact"Jr Chemical Considerable Efficiency low for gases of Iow.or slow solubility. ~..
Ana!ysis efficicnt than bubhlers fnr part'Mtlate mRtter. Permitll
large ratio of volume gas scrubbed to liquid used ill-
ereasiq sensitivity of test. Spray losses should ..
considered. Equipment not commercially available.
Atbc rption Tubes of cartridges Weight Considerable ConsideraWe Highly efficient for collection cJf wide variety of ~
Ii lied with char- General Chemical for reactive and non-reactive. Chemical analysis require
coal. silica or Analysis Trace dmorption under pre-tested conditions. Equip.....
Gaseous alumina gel. Atmospheres Quantit18 mut be constructed to meet test conditioDB. Mer
ordi...rily commercially available.
Condensation Fr~ut Chemical Considerable Less Hi8'hly efficient for relath-ely non-volatile vaporiMIII
Ducts Analysis for liquids. Requires low temperature refrigeration Ii...W
Gross air or "dry ice." Apparatus may be assembled f-
Stacks Amauntll commercially available component'S.
Bottle Chemical sOme Bfficient for gases of slow solubility- Quick and ....
Absorption collection (Low Analysis "n_t method. No power source needed at _pi..
. _ition. Gas volumes limited to aize of available eo....
and concentrations) 'talnen. Equipment readily availahle.
mechanical Plastic bag Chemical Little Uaeflll as light weight large volume container. Far
collection Analysis trace quantities wall adsorption must be eonsid"""
retention and adsorbed gases removed by solution in reagnt.
Not recommended for particulate matter. Not..-
mereially available.
-------
Since S03 is collected as H2S04' analysis may be per-
formed by methods specific for sulfate or the acid content may be det-
ermined by neutralization methods. These methods are discussed in
detail in Section 5.4.
5.3.2 Absorption in a Liquid
5.3.2.1 80% Isopropanol
Alyea and Backstrom (306) found that several
higher alcohols (isopropyl, sec-butyl and benzyl) prevented the oxi-
dation of dissolved S02. Flint (131) selected isopropanol because
it is miscible with water, its inhibiting action extends over long
periods of time, and it may be employed as a constant boiling mix-
ture. An additional advantage is that the solubility of S02 in IPA
solutions is considerably lower than in distilled water.
The collection of S03 from power plant flue
gases by absorption in 80% IPA is the most widely used collection
method (160,215,218,221,263,78,516,425,52,145,300,527). The IPA solu-
tion is used as a prescrubber for S03' and S02 is collected in H202
or another absorbing solution. A drawback of the technique is that
the S02 dissolved in the solution must be removed immediately by
purging to prevent oxidation of sulfite ion to sulfate.
Flint (131) determined the collection effici-
ency for H2S04 aerosol in 80% IPA. He found that the collection is
quantitative with two collectors and a filter all in series (Table
5-3), although the efficiency of each of the three elements is quite
low. Although a number of different collection devices have been
used, little data on S03 collection efficiency has appeared. Cor-
bett (160,215) and others (218,221) use Flint's apparatus (Figure
5-3a); Seidman (78) uses three lamp sulfur absorbers in series, while
others use midget bubblers (559,516,425,527).
78
WALDEN RESEARCH CORPORATION
-------
TABLE 5-3
COLLECTION EFFICIENCY OF S03 IN 80% IPA (131)
Test 1 2 3 4 5 Average
% Bubbler 1 29 40 47 44 48 42
% Bubbler 2 50 46 42 47 45 46
% Fi lter 18 13 12 9 7 12
Tota 1 97 99 101 100 100 100
Juntgen (530) reported a "drastic" decrease in
the collection efficiency of Flint's Apparatus at ~l ~pm. Since this work
was performed by comparison to another collection device, the absolute
efficiency was not determined.
As a result of the importance of IPA absorp-
tion methods and the scarcity of absorption efficiency determinations,
we have, carri ed out 1 aboratciry measurements for several important col-
*
lectors as a function of flow rate.
S03 collection efficiencies were determined
in a simulated combustion effluent (3% 02' 'v 10% H20, 87% N2) main-
tained at flue temperatures (300oF) up to the collectors. The col-
lection devices evaluated were:
a. midget impingers
b. midget bubblers
c. lamp sulfur absorbers
These devices were charged with 80% isopro-
panol-20% water and immersed in an ice bath. Known volumes of the
*
The delivery system is described in Appendix 5 of this report.
79
WALDEN RESEARCH CORPORATION
-------
/ Carrying case
b
......
b ( 76 )
--~- -
,
,
,
..-- Coke-fired
boiler
a (218 )
Figure 5-3.
-----..
----
7______--- .:
C ".
Heating tapes
dnd insulation
Sintered
glass disk
Inlet
d (379)
.
'0
.'
...
,
-,... ...
hcll.. ....
f
"".-
f
v.c- .....
f
Drr-!I-
c (310)
Collection Devices for 503
80
E
WALDEN RESEARCH CORPORATION
-------
simulated flue gas were drawn through the collectors using critical
fices as metering devices. The total sample volume collected was 15
liters for the midget impingers and bubblers and 30 liters for the
lamp sulfur absorbers. The S03 collected was analyzed as sulfate by
the barium chloranilate colorimetric procedure. The results for the
three devices tested are described below:
ori -
Midget Impingers - The collection system under
test was made of two midget impingers in series followed by a high ef-
ficiency (millipore) glass fiber aerosol) filter. Each impinger was
filled with ]5 cc of the 80% isopropanol-20% water solution. Collec-
tion efficiency was qbtained as a function of flow rate, by determina-
tion of the fraction of the total S03 (impingers + filter) in the first
and in both impingers. The results obtained are summarized in Table
5-4.
Within the precision of the results, a high
and constant (90%) collection efficiency is obtained for the primary
impinger in the range 0.5-3.0 lpm. Within this range the observed
efficiency of two series impingers is a constant 95%, significantly
below quantitative collection, however.
The last column of Table 5-4 is the collection
efficiency for two impingers calculated on the basis of the measured
efficiency of the primary impinger taken to be constant at a given
flow rate. Comparison of observed efficiencies for two series im-
pingers to the calculated values clearly shows that the effective ef-
ficiency of the second impinger has fallen to about 50% (in the range
0.5-3.0 lpm). The most direct interpretation of these results is
that the collection efficiency of the midget impinger decreases as
the concentration decreases. However, examination of the very wide
concentration ranges included for a given mean (0.5 and 3.0 lpm, for
example) does not show pronounced concentration dependence. (See
also following section on concentration dependence.)
81
WALDEN RESEARCH CORPORATION
-------
~
1:7
",
Z
::u
",
C/J
~
::u
o
x
8
::u
~
;
;Q
TABLE 5-4
COLLECTION EFFICIENCY OF MIDGET IMPINGERS FOR S03a
Collection Efficiency, %
Flow Rate Number of b
Mean S0Pm
LPM Replicates Conc. p
0.5 3 208 :t 140
1.0 3 220 :t 20
(X)
N 3.0 2 170 :t 80
5.0 3 93 :t 20
d
Calc.
P. c
rlmary
BothC
90 :t 2 94 :t 2 99
87 :t 3 93 :t 3 98
94 :t 2 97 :t 1 99.6
70 :t 5 84 :t 2 91
a. 15 liter sample collected in 15 ml 80% IPA at OOC
b. mean S03 concentration for all runs = 173 ppm
c. ranges are average deviations
d. calculated on basis of constant efficiency measured for primary impinger at given
flow rate
-------
At the highest flow rate (5 lpm) a signifi-
cant decrease in the collection efficiency of the impinger is observed.
The collection efficiency of the second impinger remains, however, at
only 50%! It is difficult to reconcile this result with a simple in-
terpretation of the low efficiency of the second impinger (e.g., a
non-wetted fraction of the original aerosol). However, too much em-
phasis should not be placed on this single value, since it was obtained
in the reduced efficiency range, and other factors may come into play.
Midget bubblers do not show this behavior (see below).
Midget Bubblers - The collection system
under test was made of two midget bubblers in series, again followed
by a high efficiency filter. The bubblers were each charged with 15
ml of 80% IPA solution. The observed collection efficiencies are
given in Table 5-5 as a function of flow rate.
The collection efficiencies for a single bub-
bler are significantly lower than for a single midget impinger at all
*
flow rates. In the range 0.5-1.0 lpm, however, the observed ef-
ficiencies of two bubblers in series are the same as for two series
impingers, within the precision of the results. In striking contrast
to the impinger re~ults, the efficiency of both bubblers at a given
flow rate remains constant over the entire flow range, although the
efficiency for a single bubbler drops from 73% to 35% over the range
measured.
Lamp Sulfur Absorbers - The lamp sulfur ab-
sorbers tested were of the NAPCA design, which includes a spray tower.
The absorber (but not the spray tower), was immersed in an ice bath
with 30 ml IPA solution in the absorber and 10 ml in the spray tower.
A high efficiency filter was placed downstream to complete the system
under test. Analyses were performed of the combined absorber and
*
This result must be qualified by noting that the mean S03 concentration
is less than half the mean concentration for the impinger tests.
83
WALDEN RESEARCH CORPORATION
-------
TABLE 5-5
COLLECTION EFFICIENCY OF MIDGET BUBBLERS FOR S03a
Collection Efficiency, %
Flow Rate Number of Mean SO b,c PrimaryC Both c Bothd
LPM Replicates Conc. p~m (Exp) (Exp)
0.5 4 1 00 :t 30 73 :t 6 94 :t 2 93
1.0 2 75 :t 25 73 :t 5 93 :t 2 93
ex> 3.0 5 49 :t 20 58 :t 6 83 :t 10 82
+:>
5.0 2 29 :t 8 35 :t 5 60 :t 8 58
~
r-
o
rrI
Z
::tJ
rrI
C/)
~
::tJ
o
x
8
::tJ
~
XI
~
~
a. 15 liter sample collected in 15 ml 80% IPA at OOC
b. mean S03 concentration for all runs = 66 ppm
c. ranges are average deviations
d. calculated on basis of constant efficiency measured for primary bubbler at given
flow rate
-------
spray tower solutions and separately of the filter to obtain collection
efficiencies. The results obtained are shown in Table 5-6 as a function
'.of flow rate.
TABLE 5-6
COLLECTION EFFICIENCY OF LAMP SULFUR ABSORBER FOR S03a
Flow Ra te Number of Mean SOg Conc.
LPM Replicates ppm ,c
1 2 130 :t 1 00
3 3 120 :t 60
5 2 90 :t 30
Coll ecti onc
Efficiency, %
95 :t 5
96 :t 2
95 :t 2
a. 30 liter sample collected in 30 ml 80% IPA at OoC in absorber and 10ml
80% IPA at ambient in spray tower
b. mean S03 concentration for all runs = 114 ppm
c. ranges are average deviations
The collection efficiency of the lamp sulfur
absorber is constant over the range 1-5 lpm and equal to the best values
obtained for two series impingers or bubblers over the more limited flow
ranges for efficient collection which apply to those devices.
Summary of Collection Efficiency Results - A
comparison of the observed collection efficiencies for the three sys-
tems tested is given in Figure 5-4 as a function of flow rate. Maxi-
mum efficiency of 95% is observed for all trains (excluding the fil-
ter). The (two) midget bubblers are most sensitive to loss of ef-
ficiency with increasing flow rate. The collection efficiency of
the lamp sulfur absorber is, however, constant in the range 1-5 1pm.
Concentration Dependence of the Collection
Efficiency of Midget Impingers - We have investigated the collection
85
WALDEN RESEARCH CORPORATION
-------
100
~
~
s::
aJ
or-
U
or-
If-
If-
UJ
s::
0
-
~
u 5
(X) aJ
~ .....
.....
0
u
L.S.
1.0
3.0 3.0
Flow Rate (t/min)
o lamp sulfur absorber (S03 ~114 ppm)
[J two midget impingers (S03 ~173 ppm)
~ two midget bubblers (S03 ~ 66 ppm)
5.0
Figure 5-4.
Collection Efficiencies for SO~ as a Function of Flow Rate
-------
efficiency of midget impingers as a function of S03 concentration at
5 lpm (where collection efficiency for two impingers has dropped from
the maximum value). The results obtained are shown in Figure 5-5.
The upper straight line represents the efficiency for two series im-
pingers. The data have been fitted by simple linear regression analy-
sis. The slope is significantly different from zero at the 85% con-
fidence level but suggests that the concentration dependence (at 5
lpm) is relatively small.
The lower curve of Figure 5-5 for a single
impinger is quite different from the two series impinger results and
shows a rapid decrease in efficiency with decreasing concentration.
Since there is considerable scatter in the data at 30 ppm S03' the
reliability of the results may be questioned. Taken at face value,
however, these results imply considerably increased efficiency at
low concentrations for the second impinger compared to the first.
This is the direct reverse of the behavior observed at high concen-
trations (S03 ~ 170 ppm).
Conclusions - Collection efficiencies are
less than quantitative for all conditions investigated if a final
filter is not used. Further work should be conducted at S03 con-
centrations in the 10-20 ppm range to determine collection effici-
encies as a function of flow rate.
5.3.2.2 Boiling Water
The principle of this method is that S02
and S03 can be separated by their differential solubilities in boiling
water (337); S02 is essentially insoluble, but S03 is quite soluble
in boiling water. No data are available for this promising col-
lection method.
5.3.2.3 Aqueous Sodium Hydroxide
S02 and S03 are collected together in a de-
vice such as that illustrated in Figure 5-3b. The collection
87
WALDEN RESEARCH CORPORATION
-------
~
s:::
QJ
0,...
U
or-
l+-
I+-
lJ.J
s:::
o
.,...
+-'
U
QJ
..--
..--
o
u
100
50
Collection Efficiency, % = 74 + 0.127 (Conc. S03' ppm)
(two impi ngers)
o
Fi gure 5- 5.
~
~
~~
o
50
100
ppm S03
o
1:::1
single impinger
two impingers in series
------- -- -
Collection Efficiency of Midget Impingers as a
Function of Concentration (Sampling Rate = 5 t/min)
88
-------
technique is susceptible to both the oxidation problem and to many in-
terferences in subsequent analysis (see following sections). Many in-
hibitors such as benzaldehyde, mannitol, furfural, and glycerol, have
been tried in an attempt to prevent oxidation of 502' but none of these
appears to have solved this problem. In Table 5-7 we compare results
of 503 measurements obtained by collection in cuastic (NaOH) to both
80% isopropanol and distilled water scrubbing solutions. In each case
the caustic scrubbing results are approximately 50% higher than the
others. These data indicate that sulfite oxidation may be a major prob-
lem in caustic scrubbing. No data on the collection efficiency of 503
in NaOH solutions has been located.
5.3.2.4 Distilled Water
Gillham's (161) results indicate that when
502 is bubbled through aqueous acid, over a 24-hour period the frac-
tion oxidized is approximately 25%. If the oxidation rate is linear
it would amount to l%/hr. which is significant. If 503/50x = .01,
the apparent S03 content of a 15-minute sample would be increased by
25%. This method has not been widely used for collection of S03 in
the presence of S02 since the sulfite oxidation contribution cannot
be adequately calculated (as Gillham suggests). The system is also
not specific since the solubility of S02 in water is high (requiring
another correction for most analytical methods). This technique
could be used to collect S03 where little S02 is present (e.g., a Mon-
santo Cat-ox stream) since S03 can be collected quantitatively (161,
528) in distilled water.
5.3.3
Condensation Methods
Condensation methods depend upon the presence of water
vapor to form H2S04 mist by reaction with S03 and are, therefore, limited
to a moist stream such as a combustion effluent. In the controlled, as
well as the uncontrolled condensation methods, S02 and S03 are collected
separately.
89
WALDEN RESEARCH CORPORATION
-------
TABLE 5-7
COMPARISON OF 503 ANALYSES AS A FUNCTION OF COLLECTION METHOD
ppm SOT * ppm S03 **
80% IPA Col ection Caustic Collection
(548) (NaOH) (548)
Inlet 8 16
Outlet 11 16
Inlet 14 16
Outl et 13 14
Inl et 7 18
Outlet 12 17
Inlet 8 12
Outlet 13 19
Average 11 16
ppm SO~ ppm 503
Distilled ater Caustic Collection
Collection (161) (161)
4 8
10 9
6 13
5 8
Average 6 9.5
*
Determined by titration with Ba(C104)2 and thorin indicator
**
Determined by the Berk and Burdick method
90
WALDEN RESEARCH CORPORATION
-------
5.3.3.1
Controlled Condensation
Johnstone (529) was perhaps the first to rec-
ognize that the dewpoint of the flue gas was a function of the S03 con-
centration. Although a considerable amount of work had been done on
the relation of the flue gas dewpoint to the S03 concentration, it
wasn't until the theoretical work of Muller (495) and the experimental
work of Lisle and Sensenbaugh (80) that this was clearly established.
The correlation of the experimental results with calculation is shown
in Figure 5-1. By reducing the temperature below 2000F, one can re-
move the S03 (as H2S04) from a (wet) flue gas essentially quantitatively.
It is well known that when air containing
S03 and moisture is rapidly cooled, a sulfuric acid aerosol is formed.
Goodeve (506) and Kerrigan and Snajberk (509) have shown that H2S04
mist can be collected quantitatively with a filter. Using this prin-
ciple, Knol (511 ,403), Kantor (508), Hissink (119,415) and others have
used controlled condensation of flue gas (at 60-900C) to form H2S04
*
aerosol and collected it on a filter.
Goksoyr and Ross (81) collected the sulfuric
mist on a sintered glass frit (Figure 5-6). This technique has gained
favor recently (35,80,127,119,55,103). The major advantage of control-
led condensation is that S03 is quantitatively condensed at a tempera-
ture (60-900C) above the water dewpoint. S~ is therefore not retained
in the S03 collector and the two species are easily separated, essen-
tially eliminating the problem of S02 oxidation.
A number of investigators have determined
that quantitative collection of sulfuric acid mist is obtained by the
controlled condensation technique (80,81,119,415). Typical results
of controlled condensation collection are given in Table 5-8.
* .
A modification of this technique described by Schnelder (104,264) uses
a tube of sodium chloride to filter the H2S04 aerosol.
91
WALDEN RESEARCH CORPORATION
-------
GAS SI\MPLE
IN
OVERFLOW
TUBe
STAINUSS STEEL WATER RUBBER BUNG /
UNCOOLED JACKET 6O-9O'>C ~." /
QUARTZ PROBE PRESS FIT / " ~
'\ 810 ~~L~= . . .~~~;l f '>
SIUC:" WOOL ~ t\, J;<\ '", OUT
SULPHU~ TR!OXIDE/ "LID
COLLECTOR "'-
ASBESTOS RETAINING
PAPER SCREW
U)
N
:IE
;»
r
o
IT!
:z
::0
IT!
en
IT!
;»
::0
()
:t:
()
o
::0
cg
::0
:.!t
is
---z;
Figure 5-6.
50 Controlled Condensation Apparatus
3
-------
----"
i
I
f ~reri. I
ment I
I
I
A
B
C
D
E
F
G
H
I
J
K
l
M
N
o
p
Q
R
S
T
U
TABLE 5-8
S03 COLLECTION EFFICIENCY USING
CONTROLLED CONDENSATION (415)
SO.
(0) in gas I' (h) i
mi\ture determinl'd
p.p.m.
47
44
'10'0
<),4
37
39
10.8.
9'4
4.8
4,7
3,0
0,98
42
44
44
39
40
42
56
15
12
--I 45
I n'6
. 9,6
I 39
I 41
11 3
I .
I' ~::
4,7
2.8
I 0,94
I 44
I 44
I 41
. 18
, 37
I 42
57
I IS
I 12
(c) in gas
100 b i mi1\ture
a I
I 97
101"
I 106
I 102
104
105
I 105
100
99"
I 101"
93
I 96
104
! 101"
! 93
47
91
100
102
100
100
693
732
733
690
695
701
838
717
638
702
899
750
760
732
737
716
112R
1104
15('()
1700'
1405
--- --- -- - --
so.
I (d)I
determined J
p.p.ro. I
I
I
713
721
735
693
699
724
853
771
661
741
900
762
760
721
725
625
1152
992
16(,()
1770
1430' .
__on --~~~---
'-----~-'-
Remarks
lood I
C '
103 .
I~}: no dust filler in the apparatus.
100
100]'
103 ,
102 I
107 .. .
103 dust filler 10 position.
105 .
100 i
102
1 ()() ; tcmp. beforc filter e 270 C; behind filter 50 C.
98 300 C; .. 50 C.
98 250 C ; .. 80 C.
87 ...... 250- C ; .. .. 140 C.
J02} \\ilh niter paper temp, near filter 130 C.
90! ., 'I' n n 90 C.
:~} with fluc gas in apparatus; G4 filter.
102 with flue gas; filter paper.
93
WALDEN RESEARCH CORPORATION
-------
5.3.3.2 Uncontrolled Condensation
This method is a modification (379,217,165,178,
310,170) of Flint's technique (absorption in 80% IPA). The collection
vessel may consist of a double sintered glass frit (Figure 5-3c) or a
single frit such as that adopted by the British Standards Institute
(Figure 5-3d). The stack gas is passed into a vessel at ambient tem-
perature. The resulting (uncontrolled) cooling produces H2S04 mist
which is collected on a glass frit. H2S04 is constantly removed from
the frit by a flowing stream of 80% IPA-20% water. The major difficulty
is that water condenses with the H2S04 mist. Since S02 is very soluble
in water, the problem of oxidation of S02 remains. This collection
technique is, thus, not basically different from the 80% IPA absorption
method previously discussed. The complexity and fragility of the re-
quired apparatus are a definite drawback in field sampling. No data
on the absolute efficiency have been found.
5.3.3.3 Dewpoint Method
This technique provides at best an estimate of
the S03 concentration. The probe for colllection of H2S04 (Figure 5-7)
is inserted into the stack. Compressed air is passed through the probe
to establish a temperature gradient to initiate H2S04 aerosol formation
on the surface. The temperature is measured at various points along the
probe. After a known time has elapsed, the probe is removed and each
section washed and titrated with standard base. From the analysis and
temperature profile, the dewpoint, and therefore the S03 concentration,
can be determined.
The British Standards Institute gives the re-
producibility of the dewpoint temperature as !60C. At 10 ppm 503 the
precision is !7 ppm and the coefficient of variation about 70%. Clearly,
this is a very imprecise method and will not be further considered.
94
WALDEN RESEARCH CORPORATION
-------
T.,'..(01lll,.8 III.. ,'.ncH...
-..111.'
r===". ': -------:::>;-~l'-<
~ II ---, ..--1 ~,/ ~
~. ' , , '-- _.:- -
Figure 5-7.
Acid-Deposition Probe (Part Section) (531)
5.4 Collection Methods for S02
5.4. 1
Introduction
The literature search conducted for S02 collection
methods has included examination of all areas where analysis of gas-
eous sulfur compounds is of significance. In addition to fossil fuel
combustion sources, relevant sampling techniques have been employed
in ambient (air) analysis and in the petroleum, chemical, smelting,
and pulp and paper industries. Although the objectives of a number
of sampling techniques developed for purposes other than heat and
power generation may be different from those required for fossil fuel
sources, we have attempted to make our discussion complete, in order
to indicate possible areas for future development.
In the following sections, we describe aqueous and
solid absorbents which have been utilized in the past or which ap-
pear promising as a result of peripheral application. We do not,
however, attempt to "invent" new collection methods for which the
potential appears low. There is, thus, for example, no discussion
of organic liquids which have recently been examined for possible
application to flue gas scrubbing (510).
95
WALDEN RESEARCH CORPORATION
-------
The concluding section discusses the adaptation of
ambient methods by dilution techniques.
5.4.2 Aqueous Absorbents
5.4.2.1
Hydrogen Peroxide
A.
Description of Method
This has been the most widely used collec-
tion method for S02 (78,80,81,125,119, 131,218, 35,221,403, 516,
521).
The sample, after removal of S03 in a
prescrubber, is bubbled through a solution of hydrogen peroxide which
oxidizes S02 to sulfuric acid. The resulting H2S04 has been deter-
mined by a variety of methods, either specific for sulfate or by ti-
tration to a selected pH (see Section 5.5).
B.
Areas of Application
Power plants and ambient air sampling.
C. Applicable Concentration Range
From 0.01 ppm (ambient sampling) to 2000
ppm (using 1-3% H202) for source sampling.
96
WALDEN RESEARCH CORpORATION
-------
D.
Precision and Accuracy
Varies with analytical method (see sam-
pling statistics).
E.
Interferences
Interferences depend on the determination
step. If a method specific for sulfate is used, the only interfer-
ences are oxidizable sulfur compounds. Where a non-specific method
is used, e.g., titration with standard base, strong acid or basic
species may interfere.
F.
Comments
The main reasons for wide acceptance are
high collection efficiency (514,515), ease of interfacing the scrub-
bing solution with analytical methods, good reproducibility of re-
sults, and negligible interference from weakly acidic species such
as C02 or organic acids.
G.
Conclusions
This procedure has a wide range of appli-
cation; interferences are minor. Total sulfur oxides can be deter-
mined easily by collecting S02 and S03 in the peroxide solution or, if
desired, S02 can be collected alone after removal of S03'
5.4.2.2 Sodium Hydroxide
A.
Description of the Method
The LA APCD (1) collects S02 in sodium
hydroxide after prior removal of S03 by the controlled condensation
method. Most users (76,204,210,195,252) collect both S02 and S03
simultaneously in a bubbler containing sodium hydroxide. In the
latter method, an inhibitor must be added to prevent oxidation of
97
WALDEN RESEARCH CORPORATION
-------
sulfite ion to sulfate. Berk and Burdick (210) employ benzyl alcohol,
para-amino phenol hydrochloride and mannitol as inhibitors. Wickert
(252) uses formaldehyde and Haller (384) and others (195,312) add
glycerol to prevent oxidation of sulfite.
B. Areas of Application
Power plants.
C. Applicable Concentration Range
Wide.
D. Precision and Accuracy
Will depend on the analytical method sel-
ected, but generally poor. Preliminary analysis of field samples in-
dicates that the precision may be as poor as I 7% for duplicate samples.
E.
Interferences
One must correct for the amount of C02
absorbed in the caustic solution in order to obtain the total volume
of gas sampled. In the Berk and Burdick procedure, C02 presents a
problem in the analytical method. This is corrected for by titration
to pH 4.1. However, organic acids have been found to interfere with
the Berk and Burdick procedure (76). Caustic solution is a good
scrubber for acidic components in the flue gas, e.g., N02' N203 and
HCl (for coal fired units) which interfere with many analytical
methods.
F.
Comments
Berk and Burdick (210) and API (204)
titrate an aliquot of the caustic solution with acid to obtain total
S02 and S03. Benzidine chloride is added to a separate aliquot to
determine S03 by the benzidine sulfate method. Smith (76) oxidizes
98
WALDEN RESEARCH CORpORATI~
-------
the caustic solution with peroxide and determines total S02 and S03
by the benzidine sulfate method. He also describes a "modified"
Shell method (76) which follows collection in caustic solution, by
oxidation with H202 and titration with barium ion using thorin in-
dicator. Wickert (252) and LA APCD (1) oxidize the sulfite to sulfate
to determine sulfate gravimetrically by precipitation with BaC12.
Atkin (195) determines S02 as sulfite by reaction with fuchsin and
forma1depyde (the precursor of the West-Gaeke procedure). Axford
and Sugden (312) determine sulfite polarographically.
G.
Conclusions
This approach is subject to a wide vari-
ety of interferences and errors in the collection and analysis of
samples. (See Section on sampling statistics.) Oxidation of sulfite
is a problem (as may be seen from the variety of inhibitors used)
which is still not solved.
5.4.2.3
Iodine
A.
Description of Method
The gas is bubbled through a standardized
starch-iodine solution until the solution is just deco10rized. The
gas volume needed for decolorization is measured by flowmeter or in-
tegrating meter. A modification of the standard procedure increases
the KI concentration from 25 to 125 g/l to reduce the 12 vapor pres-
sure (410). This is claimed to prevent 12 loss and increase method
precision. The reaction with S02 is:
12 + S02 + 2H20 + 2HI + H2S04
B. Areas of Application
acid plants
paper mills
This method is widely used in sulfuric
(Reich Test) (95,192), power plants (170,215,504,379) and
(410).
99
WALDEN RESEARCH CORPORATION
-------
C.
Applicable Concentration Range
Has been used over a range of 200-2000
ppm (379).
D.
Precision and Accuracy
Accuracy and precision of the Reich test
are influenced by gas pressure, gas temperature, gas flow rate, and
accuracy of volume measurement. BSI (379) reports a precision of fl%
at 2000 ppm but at 200 ppm precision falls off to f5%. The accuracy
may be seriously affected by interferences in the flue gas. (See
Section 6.2.)
E.
Interferences
Any compound which reacts with iodine or
iodide will interfere. This includes aldehydes, unsaturates and many
other organic compounds as well as H2S and mercaptans. Oxidants such
as ozone, nitrogen oxides. and to some extent oxygen will reoxidize
iodide to iodine and lead to false results. The oxides of nitrogen
are a major problem since nitrate reacts with iodide ion in acid
solution as follows (545):
2N02- + 21- + 4H+ + 2NO + 12 + 2H20
F.
Comments
The advantages of iodine absorption are
speed and ease of determination (S02 concentration can be directly
measured as a function of volume through-put). The disadvantages are
sensitivity to light, non-specificity due to reaction of 'iodine with
other compounds in the sample, some degree of endpoint uncertainty.
iodine volatility and the need to use freshly standardized solutions.
100
WALDEN RESEARCH CORpORAl
-------
G.
Conclusions
This method works best in a well defined
system, free of NOx interferences such as a sulfuric acid plant. It
can be used where a fast approximation of S02 content is desired, the
S02 level is fairly high, and the operators are aware of the limita-
tions. We do not recommend its use as an accurate method in a power
plant because of the reaction with the oxides of nitrogen.
5.4.2.4 Sodium Tetrachloromercurate
A.
Description of Method
Hendrickson et al (13), in a manual for
the sampling and analysis of kraft mill recovery gases. recommend
sampling stack effluent for S02 by trapping with a sodium tetrachloro-
mercurate solution. Determination is done by the West-Gaeke method. (140)
Graue, Gradtke, and Nagel (128) in Germany
collect S02 by taking a grab sample of the stack gas in a container
of precisely known volume and diluting the sample down to ambient level
with clear air. The diluted sample is collected in a bubbler con-
taining sodium tetrachloromercurate solution. Sample analysis is done
by the West-Gaeke method.
Borgwardt, Shigehara and Hartlage (499)
collect a small volume of stack gas using a displacement technique.
The sampling apparatus is heated to prevent condensation. They then
transfer the gas sample to a syringe containing 20 cc of tetrachloro-
mercurate solution. The S02 is determined by the West-Gaeke pro-
cedure. Borgwardt et al compare the West-Gaeke results to both the
*
Shell method and gravimetric method (absorption in NaOH). The W.G.
* .
Collection in peroxide and titration with Barium Chloride using
Thorin indicator~
101
WALDEN RESEARCH CORPORATION
-------
results are 8-14% lower than the Shell method and 4-8% lower than the
gravimetric method for different sets of runs. These low values may
be explained by side reactions occuring when stack gas is absorbed
directly in the tetrachloromercurate solution (128).
B.
Areas of Application
Kraft mill recovery gases (13), flue
gas (128).
C.
Applicable Concentrations Range
Can be applied to any sample concen-
tration if diluted to a final S02 concentration less than 5 ppm.
D.
Precision and Accuracy
Unknown.
E.
Interferences
ion is a
bleaches
Any species that reacts
potential interference. (Metals, sulfides,
the West-Gaeke color.
with mercuri c
etc.) N02
F-
Comnents
Side reactions have been }.eported to oc-
cur with undiluted gas in the scrubbing solution (128,265). There-
fore, the effluent must be diluted. (No detailed explanation given.)
As a result of the required sample dilution, minor components such as
S03' have to be measured separately. The necessity for grab sampling
rules out the advantages inherent in continuous sampling. In addition,
the dilution process is complicated and time consuming.
The sodium tetrachloromercurate approach,
which is specific for S02' has always been combined with the West-Gaeke
method for S02 measurement, requiring low concentrations of S02 in the
102
WALDEN RESEARCI-I CORPORATION
-------
final sample to be analyzed. It is not clear at this time, whether modi-
fication can be made to permit collection of S02 at high concentration
levels directly from flue gas samples.
G.
Conclusion
On the basis of present work this method
is not attractive. However, since it is one of the few methods which
is specific for S02' it should be closely examined.
5.4.2.5 Stannous Chloride (385)
A.
Description of Method
The sample is collected in a bubbler con-
taining SnC12-HCl solution with reduces S02 to H2S, The H2S formed
reacts with a solution of ammonium molybdate to give a solution of
molybdenum blue, which is measured colorimetrically.
B.
Areas of Application
Stratmann (385) describes a method for
the determination of S02 in the presence of H2S in both flue gas and
ambient air samples, where the amount of H2S is small relative to the
S02 concentration. A means to eliminate any H2S in the sample by
bUbbling the entire sample through a solution of ammonium molybdate
is suggested. S02 is said not to react with this solution.
C.
Applicable Concentration Range
Has been used for ambient ~nd stack concen-
trations.
D.
Precision and Accuracy
Reported precision is better than ilO%
(duplicate analysis) for ambient sampling and better than i2% (dupli-
cate analysis) for stack sampling.
103
WALDEN RESEARCH CORPORATION
-------
E.
Interferences
H2S oxidants, and metallic ions. The lat-
ter will complex with the H2S formed; oxidants, including oxygen, will
destroy the effectiveness of the stannous chloride.
F.
Comments
Many potential sources of trouble and
needs fine control before use.
G.
Conclusions
to apply
chloride
This appears to be a difficult procedure
to stack gas samples because of the instability of stannous
solutions.
5.4.2.6 Sodium Acetate (116)
A.
Description of Method
S02 has been separated from "H2S and other
sulfides found in kraft pulp mill effluent by selective absorption of
the S02 in 0.2 N NaC2H302' The method is based on the relative acid
strengths of sulfurous acid> acetic acid »H2S, Thus, S02 will be
absorbed by the acetate solution while H2S will not.
B.
Areas of Application
Kraft pulp mill effluent.
C.
Applicable eoncentration Range
Absorption is said to be quantitative at
S02 concentrations up to 1.5 g/M3.
D.
Precision and Accuracy
Total method error reported to be 5-10%.
104
WALDEN RESEARCH CORPORATION
-------
E.
Interferences
Acid and basic gases (and other sources
depending on the analytical procedure selected).
F.
Comments
The method requires close control of pH,
thus the presence of acidic or basic gases is a serious problem.
G.
Conclusions
This method does not offer any advantages
over hydrogen peroxide collection for collection of a sample from a
fossil-fuel effluent.
5.4.2.7 Collection Efficiency of Aqueous Scrubbers
Little information is available on the col-
lection efficiency for the various absorbing solutions used for kraft
mill effluents. However, there are data on the collection efficiencies
of the major scrubbing solutions commonly used for fossil fuel com-
bustion (H202' 12' and NaOH). For caustic and iodine collection
systems, bubblers are commonly used. With peroxide, where collec-
tion efficiency is considerably higher, the design of the device
is not as critical and a variety of absorbers such as illustrated
in Figure 5-8 have been used. Comparative efficiencies of selected
devices for the several major scrubbing media are given in Table
5-9. It is seen that quantitative collection (99%) is readily
obtained in all instances employing two series scrubbers.
105
WALDEN RESEARCH CORPORATION
-------
.
.
.
SImple Bubbler Ablorben.
41
.
Bubhler Ahsorben wltb DiBullfl.
.
Figure 5-8.
Scrubber Configurations (542)
106
~ -==.-== ~~-_-: -===-= - ---
II
.
Spiral- T)"P8 Abiorberl.
-------
TABLE 5-9
EFFICIENCIES OF ABSORBING DEVICES AND REAGENTS FOR 502
Flow Absorbing
Device Rate Reagent Eff. % Ref.
Large Impinger 1 cfm NaOH 80-90 503
H202 90-100 515
12 90 503
Fritted Bubbler 5 11m NaOH 95 503
H202 95+
Multijet Bubbler 10-15 H202 98-100 515
11m
Midget Impinger .5 11m H202 90-95 516
107
WALDEN RESEARCH CORPORATION
-------
In order to determine flow rate dependence, 502
collection efficiencies were measured for two midget impingers in series
each containing 15 ml of freshly prepared 1% hydrogen peroxide solution.
The dilution system used is given in Appendix 5. Total sample volumes
ranged from 15 to 30 liters. Sulfate analyses were performed by the
barium chloranilate method.
For high efficiency scrubbers, the collection
efficiency may be determined by analysis of both impi~gers to obtain the
fraction collected in the first, independent of source description. For
a given collection medium, the collection efficiency is assumed to be a
function of S02 concentration, total flow rate, temperature and device
geometry (pressure, except as influenced by flow pressure drop, constant).
As a preliminary experiment to determine repro-
ducibility in the system and to investigate possible variability between
impingers, five consecutive runs were conducted at constant flow rate,
temperature and S02 concentration with five different midget impingers of
identical design.* The results (Table 5-10) show that the collection ef-
ficiency is not a strong function of impinger parameters. The standard
deviation of the collection efficiency determined experimentally (~2%)
is roughly in the range expected from the precision of the analytical
and sampling methods combined. Any variation between impingers, for this
limited population, is, therefore, small.
The collection efficiency has been determined
as a function of flow rate at two concentration levels. The observed
collection efficiencies (calculated from the fraction collected in the
first of two series impingers) are given in Table 5-11 and plotted in
Figure 5-9 as a function of flow rate. At 875 ppm, the decrease in
collection efficiency with increasing flow rate, although apparent, is
sufficienc1y gradual so that essentially quantitative (99%) collection
efficiency is maintained to flow rates of 5 1iter/min for two impingers
in series. At the higher concentration, 1740 ppm, cOllection efficiency
drops more rapidly (at rates above 1.5 1iter/min) so that the 99% col-
lection efficiency limit for two series impingers occurs at about
*Ace Glass, Inc., Cat. No. 7531.
108
WAlDEN RESEARCH CORPORA110N
-------
TABLE 5-10
COLLECTION EFFICIENCIES OF SINGLE MIDGET IMPINGERS FOR S02a
Impinger
Collection Efficiency (%)b
1
2
3
4
5
mean:!: standard
deviation
94
99
94
97
97
96 :!: 2.2
a t = 22 :!:'loC; flow rate 0.5 liter/min; 502 concentration
~ 875 ppm in air; 15 m1 1% H202
b determined by amount found in second midget impinger
109
WALDEN RESEARCH CORPORATION
-------
-'
-'
a
~
r
o
1"1
Z
::0
1"1
C/)
1"1
,.
::0
(")
:x:
g
::0
~
::0
~
'6
"%
TABLE 5-11
COLLECTION EFFICIENCY OF MIDGET IMPINGER5 FOR
502 IN PEROXIDEa
Replicates
502' ppm
5
3
3
3
875
3
2
3
1740
a--T = 22°C; 15 m1 1% H202
Flow Rate
(1 iter/min)
ppm 502
1st impinger 2nd impinger
Tota 1 502'
ppm
890
871
859
854
1710
1675
1650
Collection Efficiency
100 x #1/(#1 + #2)
96
95
92
91
96
90
86
0.5
1.5
3.0
5.0
853
831
794
774
37
40
65
80
1.5
3.0
5.0
1640
1525
1420
70
150
230
-------
.....
95
;,.I!
..
»>
u
s::::
CI.I
....
U
....
'+-
'+- 90
LiJ
s::::
0
....
oj.)
u
CI.I
......
......
0
u
85
100
G) 875 ppm
& 1740 ppm
1
"
'"
"
"
~"
"
"
2
3
"
"
"
"
4
5
---- flow rate, liter/min
Figure 5-9.
Collection Efficiency of a Single Midget Impinger
for S02
111
-------
3 liter/min. Blowover of peroxide solution was observed at flow rates
> 3 liter/min, leading to possible loss of sample.
'\.
The collection efficiencies of single impingers
were also determined at 39°C and found to be somewhat better at the
higher temperature ('\.98% for both 875 and 1740 ppm S02).
It is concluded that the collection efficiency
is not a very strong function of either sample flow rate or S02 concen-
tration in the regions of interest. The recommended flow rate for quan-
titative collection in two series impingers is 0.5-3.0 liter/min.
5.4.3 Solid Adsorbents
5.4.3.1
S il i ca Ge 1
A.
Description of the Method
a) Untreated Silica Gel - Silica gel has
been used as an adsorbent for S02 both in ambient air and in paper mill
stack effluents. This technique is used in Swedish paper mills (154) as
a rapid quantitative determination for S02 up to 100 ppm. A sample is
adsorbed at room temperature and desorbed at 120-150°C by heating the gel.
Detection is by gas chromatography.
b) Treated Silica Gel - Patterson and
Mellon (182) and Kitagawa and Kobayashi (376) have developed treated
gels such as ammonium vanadate-gel, iodate or periodate-gel and have in-
vestigated other inorganic and organic-gel formulations. Kanno (125) has
tested Kitagawa detector tubes in power plant effluents.
A separation of H2S and S02 has also
been effected by using a silica gel column impregnated with silver sul-
fate and potassium hydrogen phosphate (396). Ammonia is the only seri-
ous interference - the gel must be acidic or S02 is retained as well as
the H2S. S02 is determined by H202 oxidation followed by Thorin titra-
tion. However, this approach needs more careful study before being
recommended for widespread usage.
112
WALDEN RESEARCH CORPORATION
-------
Stratmann (137) developed a method for
S02 by adsorption on silica gel followed by reduction to H2S at 700-900°C
over a platinum catalyst. The H2S was then determined by the molybdenum
blue method.
B.
Areas of Application
Ambient air (376), paper mills (154), stack
effluents.
C.
Applicable Concentration Range
Can be applied to total range of interest.
D.
Precision and Accuracy
Using the silica gel-silver sulfate approach
(396), precision is reported to be about ~10% and accuracy is indicated
to be 85-95% for amounts of S02 from 1-11 mg in the presence of up to
1200 ~g H2S,
No data are given on untreated silica col-
1ection (154).
As a result of the surprisingly good deter-
minations obtained by Kanno in stack effluents (CV ~ 5%, see Section 6.4).
we have compared the Kitagawa detector tube to peroxide collection for a
laboratory S02-air mixture. The results obtained are given in Table 5-12
and plotted in Figure 5-10.
The measurements were made at three S02
levels with and without the moisture tubes recommended for flue gas
analysis. Under the conditions tested and within the limitations im-
posed by the restricted range of the experiments, we conclude:
1) Without moisture tubes, the S02 de-
tector (Kitagawa) tubes give higher readings than actual.
113
WALDEN RESEARCH CORPORATION
-------
TABLE 5-12
COMPARISON OF DETECTOR TUBEa TO PEROXIDE
IMPINGER COLLECTION FOR S02b
Actualc ppm S02 Measured
ppm S02 Moisture Tube with Detector Tube
660 No 800
660 Yes 600
1680 No 1900
1680 No 2100
1680 Yes 1100
2170 Yes 1400
2170 Yes 1600
2170 No 2500
a Unico Environmental Instruments, Inc., Fall River, Massachusetts,
Cat. No. 103 bf, Kitagawa Detector Tubes for Sulfur Dioxide in Flue
Gas (0.02-0.30%)
b These measurements were made with synthetic, dry samples at room
temperature. For sampling of flue gases, the manufacturer recom-
mends a probe temperature greater than 120°C, although the detector
tubes are not heated. Calibration is stated to be unaffected for
detector tube temperatures in the range 0-40°C.
c As determined by titration of peroxide impinger.
114
WALDEN RESEARCH CORPORATION
-------
QJ
.Q
:::I
+J
~
o
+J
U
QJ 0
+J 0
QJ Ln
"C .....
~
.Q
C\I
o
CI)
a.
Q. 0
o
o
.....
o
o
Ln
C\I
o
o
o
C\I
o
o
Ln
Figure 5-10.
/
~
[!]
unit
slope
/
o
/'
/@
/
,,/
@
/'
(!) moisture tube
~ without moisture
tube
500
1000
2000
1500
ppm S02 by impinger collection
Comparison of Detector Tube and Peroxide Collection
Analysis for S02
115
-------
2) With moisture tubes, the apparent 502
concentration is lower than actual.
3) Although the precision and accuracy of
the detector tube is not nearly as good as Kanno reports, the technique
is attractive because of its simplicity.
E.
Interferences
Depends on impregnating
ever, none are totally specific for sulfur oxides so
potential interferences including: H2S, unsaturated
bon monoxide, aldehydes, water, etc.
agents used. How-
that there will be
hydrocarbons, car-
F.
Comments
There appears to be no advantage in using
untreated silica gel for collecting S02' The advantages of using treated
silica gel, i.e., detector tubes for the determination of 502 are the
rapid color response obtained indicative of a given S02 concentration and
the simplicity and rapidity of the method.
G.
Conclusions
These methods would be of some utility where
H2S is present but offer no advantage over peroxide method for use in
fossil fuel combustion effluents. Detector tubes, because of their
simplicity and immediate readout, may be of use as a simplified method.
5.4.3.2 Adsorption on Molecular Sieve or Ion-Exchange
Resins
A.
Description of Method
Sulfur dioxide is collected on the semi-dry
solid substrate and then determined by one of many techniques.
B.
Areas of Application
Unknown
116
WALDEN RESEARCH CORPORATION
-------
C. Applicable Concentration Range
Unknown
D. Precision and Accuracy
Unknown
E. Interferences
Unknown
F. Comments
Papers by Krejcar (231) in Czechoslovakia,
and Cole and Shulman (139) and Layton and Youngquist (393) report on the
sorptive characteristics of S02 on ion exchange resins. Data indicate
502 is quantitatively adsorbed up to the breakthrough point by both
weakly and strongly basic resins. The breakthrough point depends upon
the concentration of S02 and the linear velocity of gas. Gas phase ad-
sorption on fresh resin is essentially irreversible until a monolayer
cover has been added. Reversible sorption in the gas phase is possible
using saturated resin (adsorption and desorption isotherms are identical),
but the process is highly diffusion-limited. Bienstock and Field (185)
at the Bureau of Mines have also worked on S02 adsorption and absorption.
Their findings indicated molecular sieve 13x gives an adsorption of 7
grams of S02 per 100 grams of adsorbent which compares quite favorably
with adsorption by activated carbon (8 grams 502/100 grams adsorbent).
Removal by desorption of S02 is not considered in this paper.
The emphasis to date has been to use ion-
exchange or molecular sieve as a tool to remove 502 from gas streams.
Therefore, there is not any information on this approach as a technique
for measurement of sulfur oxides.
G.
Conclusion
Ion-exchange resins appear to offer an in-
teresting alternative for collection of S02. This is especially true if
117
WALDEN RESEARCH CORPORATION
-------
S02 and S03 can be collected together
tion techniques; however, this cannot
time.
and then removed by aqueous elu-
be considered as a method at this
5.4.3.3 Reactive Solid Sorbents
In this section, we briefly discuss the poten-
tial applicability of reactive solid sorbents to the collection and sub-
sequent determination of 502'* An example of a reactive sorbent is the
lead dioxide candle, which has been widely used as a passive (integrating)
ambient sulfation rate monitor (96). Very recently, NAPCA-PCE has sup-
ported major programs to evaluate metal oxides (493,492) and other in-
organic compounds (491) as possible sorbents for application to new
processes for removal of 502 from flue gases. It is quite feasible on
the basis of that work to select promising sorbents from the somewhat
different point of view of analytical collection. Readout (following
collection) will require analysis specific for sulfate, as in the lead
dioxide candle technique.
A.
Carbonates and Hydroxides
From the analytical point
(491) reaction classes may be reduced to the following
maximum free energy of reaction) (Table 5-13).
of view, FMC's
(on the basis of
The minimum value of Kp, Kp (min.), has been
recalculated for an equilibrium effluent concentration of 1 ppm S02 for
an effluent gas composition, which is satisfactory for analysis for S02
concentrations down to the order to 50 ppm. Utilizing these criteria,
and noting that one is considering isolated reactions rather than com-
peting multiple equilibria, we note the following potential sorbents at
400oK. The temperature (rv250°F) has been selected for operation internal
to the duct providing major advantages of representative sampling and
*We ha~e previously discussed reversible adsorbents and detector tubes
(Sectlons 5.4.3:1 and 5.4.3.2). The latter are reactive indicating
sorbents on an lnert support .
118
WALDEN RESEARCH CORPORATION
-------
=E
~
r
o
IT1
Z
;:0
IT1
C/)
~
;:0
(")
::I:
(")
o
;:0
<3
;:0
~
o
z
TABLE 5-13
STANDARD METAL COMPOUND REACTION CLASSES
Class Reaction Reaction Kp Kp (min.) x 10-4
No.
I M C03 + S02 t M S03 + C02 PCO IPSO 14
2 2
II M (OH)2 + S02 t M S03 + H20 PH O/PSO 7
2 2
IV M C03 + 1/2 02 + S02 t M S04 + C02 PCO IPSO' p1/2 81
2 2 °2
. <: p1/2
V M (OH)2 + 1/2 02 + S02 t M S04 + H20 PH OIPSO . 40
2 2 °2
--'
--'
~
-------
increased reaction rates. For all the reactions selected, Kp (min.) is
in the range 105_106 for the conditions stated.
The data of Table 5-14 show quite clearly
that carbonates of the strongly basic oxides (Class I) are thermodynamic-
ally feasible as sorbents and should be evaluated for analytical applica-
tion particularly as hydrates stable in an effluent environment. Only
the classical "strong bases" are theoretically suitable as 502 sorbents
and are, of course, not applicable as such in an atmosphere containing
C02 at an appreciable partial pressure (Class II). Class IV reacti~ns
are all favorable, reflecting the relative acid strength of 503 vs C02'
but are not of real interest since they require a catalytic high temper-
ature oxidizer section to obtain appreciable reaction rates. Class V
reactions have the same drawback.
It is of interest to
bonate impregnated filter paper has recently been
sive sampler for ambient 502 and N02 (490,489).
note that alkali car-
investigated as a pas-
B.
Oxides
Thomas et a1 (493) conducted extensive thermo-
dynamic analysis (and some kinetic studies) of metal oxides as potential
(reversible) sorbents for 502. From the different perspective of collec-
tion for analysis, one can again isolate a large number of potential
solid absorbents which provide the desired low equilibrium vapor pressures
of 502. Two categories of oxides, oxidants and non-oxidants, may be used
as sorbents for the oxides of sulfur. The first group includes Pb02' Mn02'
Ce02' and possibly Cr03' although the latter is hygroscopic. Pb02 has been
used extensively for ambient sulfur oxide analysis. Although the candle
method of analysis works very well, the formation of Pb504 is slow. Mn02
and Ce02 appear to be the most promising oxidizing sorbents. Procedures
for producing the sorbent must be standardized to obtain reproducible ac-
tivity, since the reaction rate depends upon the method of preparation (493).
120
WALDEN RESEARCH CORpORATION
-------
TABLE 5-14
POTENTIAL 5°2 "ANALYTICAL" SORBENTS AT 4000K
Reactant{s) CO; OH - CO;+02 OH-+02
Class (FMC) I II IV V
Lo910 K (min) 5.147 4.835 5.908 5.603
Element
Ba + + + +
Ca + + +
Co ? 0 + 0
Cu (II) 0 + 0
Cu (I) 0 + 0
Fe (II) 0 + 0
Fe (II I) 0 + 0
-"~I
Li + + + +
Mg + +
Mn (II) +
Mn (II I) 0 + 0
K + 0 + 0
Na + + + +
Sn
-------
The oxides, which simply react with 502 and
50 to form sulfites and sulfates respectively, appear less promising.
3
Oxides have to be chosen, which are stable in an effluent atmosphere and
which do not form carbonates preferentially. BaO, Zr02' CdO, PbO, Bi203'
and Ce203 appear to be possibilities (493). Appreciable sulfite formation
may not occur except at elevated temperatures, requiring an auxiliary fur-
nace. Absorption begins at a relatively low temperature on Cr203 and 5n02
(mixed) but even then it is slow below 300°C. The oxides listed above were
not experimentally investigated as sorbents because of the high decomposi-
tion temperatures of their sulfites and sulfates or because of dispropor-
tionation of the sulfite, and their usefulness is therefore unknown.
However, BaO forms a sulfite between 30 and 800°C, and may well be a
good oxide sorbent.
A number of mixed oxides have also been in-
vestigated (492) in which an active sorbent is precipitated on an inert
support, exposing a large sorbent area to the sorbate. The best of
these, investigated by AVCO are CuO supported on alumina or silica and
manganese oxide, supported on alumina. For these materials, adsorption
is fast. Mn-Al-2, for example, at 300°C adsorbs at an initial rate of
400 gms 502/100 gm sorbent/hr/mole fraction of 502' (Mn-Al-2 is manganese
oxide, supported on aluminum oxide, where Mn-Al-2 denotes the method of
precipitation.) From the time of half absorption capacity exhaustion
(13 min. at 0.3% 502)' it may be estimated that the Mn-Al-2 composition
may be a suitable analytical sorbent for grab sampling; i.e., for short
duration (approx. 1 min.) samples where the absorption rate remains high.
It is important to note that these materials have been selected on the
basis of engineering criteria for flue gas cleaning and have not been
optimized for sampling.
As a result of the potential of solid sor-
bents for both simplified and high precision methods, we have briefly
examined the behavior of Pb02 as a sorbent for 502'
122
WALDEN RESEARCH CORPORATION
-------
C.
Experimental Studies of the Collection
Efficiency of S02
We have conducted a preliminary investiga-
tion of the collection efficiency for S02 by solid Pb02' The initial
experiment, utilizing Research Appliance Corporation Pb02 (for sulfation
candles) was unsuccessful since an unacceptably large pressure drop de-
veloped across a 6-cm column. A coarser grade (Fisher Scientific,
Special Micro Grade, 12-20 mesh) proved completely satisfactory.
The Pb02 was loosely packed into 6 mm i.d.
glass tubing and secured with glass wool plugs. The S02-in-air test mix-
ture was passed through the Pb02 adsorber and then through an impinger
containing peroxide to collect the remaining S02' The calculated S02
concentrations obtained from the dilution system were checked before and
after each days run by collection of S02 in two peroxide impingers in
series.
The collection efficiency of the Pb02 ad-
sorber was determined from the known S02 concentration, the volume of
gas sampled and the amount of S02 found in the series impinger. The
results obtained are summarized in Table 5-15. At ambient temperature,
quantitative collection is observed (within the precision of the experi-
ment) for a 12-cm column at 875 ppm S02 and for a 15-cm column at 1740
ppm, both at a flow rate of 1.0 liter/min.
These results suggest a surprisingly high
rate of S02 absorption, for this specific Pb02' and confirm the high
potential of solid adsorbents for S02 collection.
Gully et al (616) have investigated solid
sorbents for the purpose of removing acid gases from space cabin atmo-
spheres. They utilized standard concentrations of about 10-20 ppm for
various gases including S02 and determined the exit gas concentrations
as a function of time. The collection efficiency may be calculated from
the initial ratio of the inlet/outlet concentrations. They used heated
(~300°C) columns 1-2 em in length for their measurements and gas flow
123
WALDEN RESEARCH CORPORATION
-------
TABLE 5-15
a b
COLLECTION EFFICIENCY OF SOLID Pb02 FOR S02 '
S02 Length of Pb02 Collection Efficiency
Run Concentration (cm) (%)c
1 875 9 92
2 875 12 98
3 875 12 99
4 875 12 99
5 1740 15 98
6 1740 15 98
a Fisher Special Micro Grade, 12-20 mesh
b Sampling rate = 1 liter/min for 15 minutes at 23°C
c Determined from amount in series impinger
rates of ~300 m1/hour. The flow rates in our Pb02 experiments were about
100 times greater as were the column lengths so that residence times are
comparable in both cases. A summary of efficiencies for collection of
502 calculated from their results is given in Table 5-16. Even at ~350°C,
the Li, Ca, and Sa carbonates do not provide quantitative collection of 502'
50dfum carbonate and bicarbonate provide useful collecfion efficiencies
at the elevated temperature studied. Manganese dioxide appears quite
promising since the collection efficiency is quantitative (99%). Gully
indicates further that studies of Mn02 at room temperature also show high
efficiency but no data are provided to support this statement. It is
clear from both our brief studies and the work of Gully et a1 that the
oxidative solid sorbents are quite promising for collecting the sulfur
oxides.
D.
Conclusion
Although it is theoretically feasible to
consider fixation of sulfite and sulfate and subsequent determination of
124
WALDEN RESEARCH CORPORATIO~
-------
TABLE 5-16
COMPARISON OF COLLECTION EFFICIENCIES OF SOLID SORBENTS FOR SO a
2
ppm S02 Collection Efficiency (%)b 50rbent
13-14 40 CaC03
12 47 Li 2C03
10 48 BaC03
14 95 NaHC03
15 91 Na2C03
15 99 Mn02
a T = 342°C, flow rate 300 ml/hr
b Collection efficiency =
( C - C )
i Ci f x 1 00
where Ci and Cf are the inlet and outlet concentrations
each species, it is our present belief that the most promising immediate
prospect for reactive adsorbents is for simple total SO determination
x
(possibly +NO ).
x
5.4.4 Dilution Methods
One may, in principal, dilute a stack effluent contain-
ing 502 and 503 down to ambient levels and use one of the analytical tech-
niques described by Hocheiser (96). However, for a ten-fold, or more
likely, a hundred-fold sample dilution, minor components such as 503 will
be lost.
Several different dilution techniques may be employed.
One dilution technique for S02 (described in Section 5.4.2.4), which involves
taking a grab sample and diluting this with air, provides only the instantan-
eous pollutant concentration. The grab sampler has to be heated while
125
WALDEN RESEARCH CORPORATION
-------
collecting and diluting the sample to prevent condensation of moisture
and subsequent loss of the 502.
Dilution techniques have been used frequently in auto-
mobile exhaust studies. One such technique is the variable dilution
technique described by Rose et al (497). Another method which could be
used involves two calibrated critical orifices. The stack gas is di-
luted with air, then collected in bubblers containing various absorbing
solutions (96). Thus, the sample could be collected over a long enough
period of time to obtain a good estimate of the average plant output.
The dilution section should be heated to prevent condensation.
The major drawbacks to dilution methods used with manual
methods of analysis are:
1. Loss of minor components in stack gases.
2. Collection efficiencies of various absorbing
solutions usually decrease at lower concentrations.
3. "Clean" dilution air is required.
4. The apparatus is complex and cumbersome.
5. Interferences not commonly found for ambient methods
may be a problem at stack concentrations (this situation may be very
critical for coal-fired plants).
6. Precision and accuracy of the wet chemical methods
usually decrease with concentration.
As a result of these problems, we do not recommend this
technique.
5.5 Comparison of Analytical Methods
5.5. 1
Introduction
All practical methods for determination of 50 depend
1 . . 3
upon ana YS1S ln an aQU~OUS solvent. The resulting H2S04 may be
measured as sulfate or titrated as acid by the variety of techniques
described below. For S02 collection, many methods include an oxidation
126
WALDEN RESEARCH CORPORATIQN
-------
reaction either in the absorbing solution or subsequent to absorption
and determination as sulfate.
However, S02 may also be determined as sulfite, although
this method has not been widely used. (Another option for the analysis
of S02 is reduction to H2S with SnC12 and determination by the molybdenum
blue technique.. See Section 5.4.2.5.) The alternatives are discussed
in following sections.
5.5.2 Analytical Methods for the Determination of Sulfate Ion
or Sulfuric Acid
5.5.2.1
Titration with Barium Ion and Selected Indica-
tors (52,78,145,169,303,308,345,353,129,196,193.
Also see Section 5.5.4.1)
A.
Description of Method
A solution containing sulfate ions is ti-
trated with a standard barium chloride or barium perchlorate solution
using an adsorption indicator (such as thorin). The titrant is prefer-
ably Ba(C104)2 due to its greater solubility in alcoholic solution. The
problem of coprecipitation is diminished by using the perchlorate instead
of the chloride anion (196).
B.
Areas of Application
Fritz and Yamamura (303) developed this
method for the analysis of boiler water and raw and treated city water.
The technique has been extended to stack gas sampling by Seidman at Shell
Development Co. (78), Fielder and Morgan (169) and others (52,145,300).
A microchemical method for sulfur oxides in ambient air (308) also uses
barium ion titration.
C.
Applicable Concentrations
down to 10 ppm,
upper limit unknown
Tetrahydroxy quinone 10 ppm up to 30,000 ppm
Alizarin Red S no range given
Thorin
127
WALDEN RESEARCH CORPORATION
-------
D.
Precision and Accuracy
Thorin
fO.4% at 25 ~g S03/m1
f4% at 2.5 ~g S03/m1
f2.4%
about 1% indicated
Tetrahydroxy quinone
A1 izarin Red S
E.
Interferences
All barium sulfate precipitates are subject
to both cation and anion interferences due to coprecipitation. Cation re-
moval is necessary for all three indicators since colored complexes are
formed with the indicators which mask the endpoint of the titration. Sul-
fite and phosphate are reported to cause serious interference with the
thorin titration (303). Phosphate interference with thorin is removed by
precipitation with MgC03' Phosphate interference with tetrahydroxy
quinone is removed by pH adjustment.
F.
Comments
Alizarin Red Sand tetrahydroxy quinone are
reported to give poor endpoint definition and are not usable to as low a
concentration as thorin (303). In addition, thorin has been found to be
more precise than Alizarin Red S (196).
G.
Conclusion
Thorin appears to be the best of the three
indicators compared.
5.5.2.2 Titration with Barium Ion Using Other Substituted
Naphthalene Disulfonic Acid Derivatives (SNDAD)
as Indicators (60,392,394,395,430,405)
A.
Description of Method
Alcohol or acetone solutions of sulfate ions
are titrated using SNDAD. The titrant is usually Ba(C10) but with
arsenazo III (2,7-bis (azo-2)-phenyl arsono-l ,8-dihYdrOx~n~Phtha1ene-3,6-
128
WALDEN RESEARCH CORpORATION
-------
disu1fonic acid disodium salt) a mixture of barium and lead acetate is
emp 1 oyed (617 ,430).
B.
Areas of Application
Any solution containing sulfate ions.
Bredesinsky (392,394,395,405) has used SNDAD's to determine the sulfur
content of both organic and inorganic compounds after conversion to sul-
fate. Kanno (125) and Fukui (60) have used the arsenazo III method for
S02 analysis in stack gases.
C.
Applicable Concentration Range
Range is claimed to be greater than for
thorin (60) but no lower limit is given.
D.
Precision and Accuracy
Not given for arsenazo III.
indicators, cr = to.OB - 0.14% (394).
Other SNDAD
E.
Interferences
No interference is claimed from H202' Cl-,
N02' or NO;. No interference was found from sulfite ion added in 300-
fold excess of the sulfate ion concentration. Solution has to be kept
below pH 3 (60).
F.
Conunents
Titrations with arsenazo III or other SNDADls
are claimed to be superior to thorin. Lack of interferences and greater
ease of endpoint detection are two good reasons for the superiority
claim (60).
G.
Conclusion
to be very useful for a
actually as easy to see
Arsenazo III (or related) indicator appears
titration procedure (provided the endpoint is
as stated, but see Section 5.5.4.1A).
129
WALDEN RESEARCH CORPORATION
-------
5.5.2.3 Titration with NaOH (35,55,80,81,127,180,403)
A.
Description of Method
A solution containing sulfuric acid formed
by the oxidation with H202 is titrated with standard sodium hydroxide to
a selected endpoint. Bromphenol blue (35,55,80,81,127) and methyl red (180)
have been used as indicators. The amount of titrant used determines the
solution acidity and thus the sulfate ion concentration, provided no ad-
ditional acidic species are present.
B.
Areas of Application
This method has been used in both chemical
plants (180) and power plants {35,55,80,81,127,403,430}.
C.
Applicable Concentration Range
Wide. Lower limit unknown. JIS K103 (430)
does not recommend use of this method below 500 ppm S02 because of the
problem of interferences from acidic components in flue gases.
D.
Precision and Accuracy
Precision - 1% or less; accuracy - depends
on interferences present.
E.
Interferences
This is an acid-base titration; NH3' oxides
of nitrogen, and other basic or acidic species in the gas sample will in-
fluence the volume of titrant needed to produce a color change.
F.
Comments
Since this is an acid-base reaction, it is
not a specific method, and thus is not desirable for use where high ac-
curacy is desired.
130
WALDEN RESEARCH CORpORATION
-------
G.
Conclusion
Due to non-specificity, this method is not
as desirable as titration with barium ion.
5.5.2.4 Colorimetry Using Barium Ch10rani1ate (153,159,
164,271,273,311,361,424,425,430. Also see
Section 5.5.4.2 and Appendix 3.)
A.
Description of Method*
Barium ch10rani1ate is added to a buffered
alcoholic solution containing sulfate ions. The sulfate ions react with
the barium ch10rani1ate to yield barium sulfate and the acid ch10rani1ate
ion. The amount of acid ch10rani1ate liberated is proportional to the
sulfate ion concentration. Buffering is required because the solution
absorbance is a function of pH. The reaction is run in 50-80% alcohol to
decrease the solubilities of barium sulfate and barium ch10ranilate.
B. Areas of Application (153,164,271,273,311,
360,424,425)
This method is being used for flue gas
sampling, ambient air sampling (159), and as a meanS of determining
traces of sulfur in naphthas by lamp combustion (361). At least two
instruments have been constructed using the barium chloranilate method
for S03 in power plants (153,437).
C. Applicable Concentration Range
The method is claimed to be applicable from
2-400 ppm sulfate (164). Determination at 330 nm rather than 530 nm is
reported to allow detection down to 0.06 ppm (311).
D.
Precision and Accuracy
Precision of 1-3% and 99% accuracy has been
reported for determinations made at 530 nm (164,273).
*See Appendix 3 for full description.
131
WALDEN RESEARCH CORPORATION
-------
Precision of 5-20% is reported for the 330 nm
determination (311).
E.
Interferences
Cation contamination must be either prevented
by a particulate filter in the sampling line or removed by an ion exchange
resin, before reaction with the barium ch10rani1ate. Interference from
cations is due to colored species formed with the ch10ranilate ion which
decrease the available amount of ch10rani1ate in the mixture and may also
interfere with the absorbance measurement. Common anions such as nitrate
and chloride are said not to interfere (164). The presence of 0.15% H202
has been reported to give no interference (424).
F.
Conments
The method is simple to perform, has several
pH options, and two wavelength readouts, depending on the needs of the
analyst. Instrumental readout aids in minimizing operator error.
G.
Conclusion
unskilled operators.
for sulfate analysis.
The analysis is easily performed, even with
This appears to be one of the best available methods
5.5.2.5 Turbidimetry (160,165,170,215,218,221)
A.
Description of Method
Excess barium chloride solution is added to
a solution containing sulfate ions and the resulting turbidity determined.
B.
Areas of Application
Used in England by several power companies
for the determination of 502 and 503 in flue gas (see references above).
132
WALDEN RESEARCH CORPORATION
-------
C. Applicable Concentration Range
Unknown.
D. Precision and Accuracy
i2.7% at 1.3 mg/100 ml concentration (165)
i6.6% at 0.1 mg/100 ml concentration (170)
i4% at 20.2 mg S03/100 ml (160)
E.
Interferences
BaS03 will precipitate along with BaS04 un-
less it is removed prior to precipitation. Other interferences result
from the usual coprecipitation and occlusion problems associated with
BaS04 precipitation (196,303).
F.
Comments
The method has several drawbacks including
the need for good control over the temperature, rate of BaC12 addition,
amount and degree of agitation, and exact settling time after mixing be-
fore taking a turbidity reading. There is poor reproducibility between
laboratories (see Precision and Accuracy above).
G.
Conclusion
This method is not desirable as a result of
its many inherent drawbacks.
5.5.2.6 Conductivity (161,162,337)
A.
Description of Method
The conductivity of a solution containing
H2S04 is measured, and used to determine the sulfate ion concentration.
B.
Areas of Application
Power plants (161,162,337).
133
WALDEN RESEARCH CORPORATION
-------
C.
Interferences
Conductivity is a non-specific method which
will measure total solution conductance, hence any anion or cation other
than H+ ion from H2S04 can be considered an interference.
D.
Comments
Detailed consideration was not undertaken
since conductivity is not a manual method. It is clearly necessary to
confine this non-specific technique to closely controlled comparisons.
5.5.2.7 Precipitation Reactions (1,204,210,263,269)
A.
Description of Method
A sulfate solution is precipitated as the
barium or benzidine salt. The barium salt is filtered, dried, and
weighed as BaS04 (1,263). The benzidine salt may be filtered, redis-
solved, diazotized and then coupled with an ethylene diamine derivative
(269), or filtered, redissolved in a known excess of standard caustic
and back-titrated with standard acid solution (76,204,210).
B.
Areas of Application
Used in power plants (1,76,204) and for the
determination of sulfate in waste water (269).
C.
Applicable Concentration Range
Upper limit not given, micro determination
by benzidine sulfate method reported down to 15 ~g of sulfate.
D.
Precision and Accuracy
One author claims 98-99% accuracy and 0.1%
precision (269).
134
WALDEN RESEARCH CORPORATION
-------
E.
Interferences
The BaS04 method suffers from coprecipita-
tion and occlusion of nitrate and chloride anions (196,303). Phosphate
is reported to interfere in the benzidine sulfate procedure (269).
F.
Comments
The methods are tedious and time consuming.
Barium precipitation requires heating the sulfate solution while adding
titrant and maintaining the solution at approximately 60°C for two or more
hours while the precipitate digests. Benzidine sulfate is relatively
water soluble and thus susceptible to losses during filtration.
G.
Conclusions
The speed and accuracy (under field condi-
tions) of this type of determination is poor compared to the colorimetric
or titrimetric procedures.
5.5.2.8 Titrations Involving Ethylenediaminetetraacetic
Acid (EDTA) (333,354)
A.
Description of Method
In one variation {354}, a known excess of a
barium salt is added to a sulfate solution and the excess barium at pH 10
is titrated with an EDTA solution. Another variation is to precipitate
BaS04 with BaC12' filter the solution, and dissolve the precipitate in an
excess of ammoniacal EDTA. The solution is titrated with a magnesium solu-
tion to an eriochrome T endpoint (333).
B. Areas of Application
Natural water analysis - boiler feed.
C.
Applicable Concentration Range
Not given.
135
WALDEN RESEARCH CORPORATION
-------
D.
Precision and Accuracy
Accuracy of 92% at 20 ppm level, accuracy
of 99% at 50 ppm level. Standard deviations for these levels are given
as :t1.4% and :t1.0%. Standard deviation at 200 ppm given as :t3.4% (354).
E.
Interferences
Difficulties include interaction of EDTA
with the barium sulfate precipitate, unsatisfactory behavior of the in-
dicator, and changeable behavior of the EDTA-barium complex. pH control
is also necessary (333).
F.
Conclusions
This method is not as good as either the
barium chloranilate or the thorin procedures.
5.5.2.9 Summary and Conclusion
In summary, our examination of the methods now
being used for the measurement of sulfate ion concentration indicates:
A. Thorin and arsenazo III are the best in-
dicators for a titration procedure since they provide good precision and
accuracy. are not subject to major interferences (which cannot readily be
eliminated) and may be applied over a wide range of concentration.
B. NaOH titration is non-specific and there-
fore not desirable for a very precise and accurate method.
C. Colorimetry using barium chloranilate is
one of the best overall methods for the determination of sulfate.
D. Turbidimetry has many drawbacks which ex-
clude it from further consideration.
E. Conductivity is non-specific in addition to
not being within the scope of the present work.
136
WALDEN RESEARCH CORPORATION
-------
F. Precipitation reactions have drawbacks such
as coprecipitation which exclude them from further consideration.
G. Reactions involving EDTA have serious prob-
lems to overcome before being considered as useful.
The two methods of choice, therefore, are
titration with barium using thorin (or other) indicator and the barium
chloranilate colorimetric procedure.
5.5.3 Analytical Methods for the Measurement of Sulfite Ion
Concentration
5.5.3.1
Iodine Titration by Decolorization of a Standard
Iodine Solution (95,170,192,215,410,504)
(See Section 5.4.2.3.) A variation of iodine
titration is to use excess iodine and titrate the excess with standard
thiosulfate solution. All other comments in Section 5.4.2.3 apply here
as well.
5.5.3.2 Iodine Titration of NaHS03 Solution (384)
A.
Description of Method
A caustic solution containing NaHS03 is
acidified, and titrated with a standard iodate solution. Glycerol (5%)
is added to the caustic to prevent sulfite oxidation to sulfate during
collection.
B.
Area of Application
Sulfuric acid plants (384).
C. Applicable Concentration Range
No lower limit given; highest concentration
stated is 0.9 gm S02 absorbed in 100 ml of 10% NaOH solution.
137
WALDEN RESEARCH CORPORATION
-------
D.
Precision and Accuracy
Accuracy is indicated to be 99.5%.
Precision
is about 1-2%.
E.
Interferences
See Section 5.4.2.3.
F.
Comments
The method uses the reaction 2S02 + K103 +
2HC1 + 2S03 + KC1 + H20 + IC1. Presence of some KI initially forms some
iodine, which is observed in a chloroform globule in the bottom of the
solution flask. The endpoint is observed by the disappearance of the
purple color when an excess of KI03 is present. Vigorous shaking is
necessary to maintain good precision. Only standard solution required
is KI03'
G.
Conclusion
The major source of error is in the use of
caustic for collection of the S02'
5.5.3.3 West-Gaeke Colorimetric Determination of S02
( 1 3 , 1 28 )
A.
Description of Method
Gas containing S02 is bubbled through a solu-
tion of O.lM sodium tetrachloromercurate (TCM). The TCM reacts with S02
to form dich1orosu1fitomercurate ion. Reaction of the ion with formal-
dehyde and acid-bleached pararosaniline dye forms a red-purple solution
of pararosaniline methyl sulfonic acid. Concentration is determined
by colorimetry.
B. Applicable Concentration Range
0.005-5 ppm.
138
WALDEN RESEARCH CORPORATION
-------
C.
Precision and Accuracy
Reported as flO% from 0.005-0.10 ppm.
tYrij~Y incr~a$~~ with concontration.
Ac-
D.
Interferences
03 interference can be eliminated by gas
filtration through FeS04 crystals. N02 interference can be eliminated by
the addition of sulfamic acid or o-toluidine to the collected sample be-
fore addition of formaldehyde and pararosaniline. Heavy metal inter-
ference is eliminated by addition of EDTA (ethylene diamine tetraacetic
acid disodium salt) to the TCM solution.
E.
Comments
Color stability is reported to be independent
of temperature is the range of 11-30oC and to be stable for up to 3 hours.
F.
Conclusion
The method appears well suited to the selec-
tive determination of S02 at ambient levels. Stack samples containing
502 concentrations higher than 5 ppm should be diluted down to 5 ppm or
less to use this procedure. The S03 content of the sample after dilution
would be extremely small, thus a second sample would be needed if an
503 determination is desired.
5.5.3.4 Acid Titration - NaHS03 Solution (204)
This is the method described by Berk and
Berdick. The S02 is collected by bubbling through a standard NaOH solu-
tion. After a known volume of gas is bubbled through the solution, the
excess caustic is titrated to pH 4.1 with standard acid. This method
is subject to a number of interferences described in Section 5.4.2.4.
5.5.3.5 Conclusion
None of these techniques offers any advantage
over oxidation of S02 and determination as sulfate.
139
WALDEN RESEARCH CORPORATION
-------
5.5.4 Laboratory Investigation of Analytical Methods for Det-
ermination of Sulfate Ion
5.5.4.1 Titration of Sulfate Solutions with Barium
Perchlorate
A.
Screening of Indicators
Several indicators have been reported (394)
to be superior to thorin for the titration of sulfate ion with barium ion.
The main reason given for superiority over thorin is a more readily seen
endpoint (better color change differentiation), and in the case of arsenazo,
a claim is made for fewer interference problems. Of those reported, only
three, Sulfonazo III, Methylsulfonazo III, and Arsenazo III, were avail-
able (Aldrich Chemical Co.).
Sulfonazo III and methylsulfonazo III are
reported to change from wine-red to blue when barium ion is in excess.
The solutions as prepared (0.1% in deionized water) were blue. Repeated
attempts to produce a red color in the presence of sulfate failed, so
that no further work has been done with these two indicators.
Arsenazo III is reported to give a red to
blue endpoint color change (617). The blue color is due to the presence of
lead ion in the titrant. We have prepared several bottles of arsenazo III
solution and of the titrant as described in the paper by Fukui (617).
The dye is digested in water overnight, then
filtered and the filtrate stored in a dark bottle. The titrant is a mixture
of barium acetate and lead acetate in 80% isopropyl alcohol. We observe
a purple endpoint, which is no easier to see than the thorin endpoint.
A series of four titrations showed a coefficient of variation of 1.6%
as compared to 0.5-0.7% for thorin. (See E following.) No further
work has been done with this indicator-titration system. Of the in-
dicators studied, thorin appears to be the best. (See below.)
B.
Thorin Indicator
Solutions of 80% isopropanol containing
4 drops of 0.2% thorin indicator were titrated with 0.006 N Ba(C10 )
4 2
140
WALDEN RESEARCH CORPORATION
-------
solution in 80% isopropanol to determine blank reproducibility and to
observe the color change at the endpoint. The change is from yellow
to pink and is reported by Fritz and Yamamura (303) to be quite sharp.
We found the endpoint to be nebulous, the color going from yellow
through orange to pink. Solutions containing known amounts of sulfate
were used to standardize the Ba(C104)2 solution; the standard deviation
of the titration of four samples was 0.6%.
Barium titrant concentrations of 0.005M
have been recommended by Fritz. Nine titrations at this concentration
yielded a precision (coefficient of variation = standard deviation x
100/mean) of 1.0%. Endpoint visibility was improved by addition of 1-2
drops of methylene blue solution (0.0125 g/100 m1 water) in addition to
the two drops of thorin.
However, remaining endpoint visibility
problems suggested an increase in the titrant strength to O.OlM barium.
At this concentration endpoint, detection becomes simple; a coefficient
of variation of 0.5-0.7% is obtained. (See Section E following.) The
O.OlM concentration is, therefore, recommended.
C.
Effects of Solution pH on Thorin Titration
Samples were prepared at several apparent
pHis by the addition of perch10ric acid or 0.25M magnesium acetate to
solutions of sulfuric acid in 80% isopropyl alcohol. Magnesium ion is
claimed not to cause an interference in the precipitation reaction (345).
a.
Procedure
1. Pipet Xm1 0.0054 H2S04 into a 100 m1 beaker (X = 5 or 8).
2. Add 40 m1 isopropanol and dilute to 50 m1 mark with H20.
pH with dilute HC104 or 0.25M magnesium acetate.
3. Add 2 drops 0.2% thorin and 1 drop 0.0125% methylene blue.
4. Titrate to a pink endpoint using O.OlM barium perchlorate adjusted
to pH 3.5 with dilute perch10ric acid.
Adjust
141
WALDEN RESEARCH CORPORATION
-------
levels are given below.
Sul fate Level
pH
ml of O.0054M H2S04 -
5 4.39
4.36
4.36
8
2.77
2.82
2.77
The results at two different sulfate
ml Titrant Required
2
3
4 5
4.23 4.17
4.18
4.11 3.88
2.59 2.63
2.59 2.58
4.36
4.34
4.35
2.76
2.81
2.72
4.34*
2.77*
*
Mean values obtained from precision experiments following.
The titration may. therefore, be conducted
at pH's or 1,2, and 3 with very little error. However, if the pH of
the solution is 4 or more a negative error results. This is in agree-
ment with the literature which states that thorin functions best at an
apparent pH of 2 to 3, with a usable range of 1.5 to 3.5 (196).
D.
Effect of Temperature on Thorin Titration
Samples were titrated with O.OlM barium
perchlorate to a thorin endpoint after temperature equilibration at
200e and 300e. Solution temperatures rose and fell one degree at each
temperature during the titration. Essentially no change in titer was
observed. Data is as follows:
200e 300e
4.24 ml titrant 4.22 ml titrant
4.25 4.21
4.24 4.21
No temperature restriction is, therefore,
necessary in the titration procedure.
142
WALDEN RESEARCH CORPORATION
-------
E. Precision of Titration Using Thorin as
Indi cator
a. Procedure
1. Pipet X m1s of 0.0054M H2S04 into a 100 m1 beaker (X = 5 or 8).
2. Add 40 ml of isopropyl alcohol and dilute to 50 m1 mark with water,
adjust pH to 3 with dilute HC104 or 0.25M magnesium acetate.
3. Add 2 drops 0.2% thorin and 1 drop 0.013% methylene blue.
4. Titrate to pink endpoint using O.OlM barium perchlorate adjusted to
pH 3.5 with dilute perchloric acid.
b. m1s of Titrant Required for X = 5
5 ml buret used
2.75
2.74
2.74
2.70
2.73
2.71
2.70
mean value = 2.73
standard deviation
cv* = :to.7%
= :to.02
c. m1 of Titrant for X = 8
5 m1 buret used
4.32
4.38
4.33
4.35
4.34
4.35
4.33
4.36
4.32
4.35
mean value = 4.34
standard deviation = :to.02
cv* = :to.5%
*
coefficient of variation
143
WALDEN RESEARCH CORPORATION
-------
F.
Study of Potential Interferences in the
Titration Procedure
A number of cations and anions have been re-
ported to interfere in the titration of sulfate ion with barium ion. Pre-
sumably, cationic interference from particulates would be essentially nil
if the particulate filter in the probe is working properly. However, a
small cation exchange column, used in the procedure, would eliminate any
cations which get by the filter or are carried through in the gas phase.
To simulate interferences for this experi-
ment, we prepared solutions of a number of materials, then shook the
solutions with at least 1 ml wet volume of Dowex 50W-XB ion exchange resin
to exchange the solution cations with the hydrogen ions on the resin.
Aliquots of the solutions were then taken for use in the interference
study.
a.
Procedure
1.
50W-XB resin washed with 3N HC1, then rinsed with distilled water
until effluent showed no Cl- when checked with AgN03'
1 + ml (wet volume) of the resin. was added to 100 ml of potential
interference at 0.2,0.1, or 0.05M concentration (1% and 0.5% solu-
tions of H202 were used.
Mixtures were shaken vigorously for at least one minute, then allowed
to stand.
2.
3.
4.
5 ml aliquots of the anion solutions were pipetted into 100 ml beakers.
a. Add 5 ml 0.0054M H2S04'
b. Add 40 ml isopropanol.
c. Add 2 drops thorin + 1 drop methylene blue solution.
5. Titrate with O.OlM Ba(C104)2 to appearance of pink color.
The results obtained are given in the
table below.
144
WALDEN RESEARCH CORPORATION
-------
Anion
ml s. of Ti trant
at 0.1 M
ml s. of Ti trant
at 0.05M
nitrate
fluoride
chloride
bicarbonate
2.08
2.49
2.67 (2.03 at 0.2M)
2.45
2.70 (very difficult endpoint
to see)
ppt.
ppt.
-- (2.65 at 0.2M)
-- (2.66 at 1%)
2.38
4.53
phosphate
oxalate
forma 1 dehyde
hydrogen peroxide
ppt.
ppt.
2.66
2.65 (0.5%)
Control was 5 ml H2S04 + 5 ml H20 + 40 ml isopropanol
2.66 ml of titrant
2.64
2.63
The most serious interferences are seen
to be oxalate and phosphate ions. Oxalate ion was chosen to be represent-
ative of oxidized organic material. Phosphate was included in the list
because it is known to be a serious interference to the determination of
sulfate ion in streams and natural waters and if present in a stack-gas
sample, could lead to considerable analysis error.
5.5.4.2 Determination of Sulfate by the Barium
Chloranilate Method*
The pH of the chloranilate solution is a funda-
mental parameter which must be defined in the development of the color-
metric method. As shown below pH 5.6 was selected as an optimum value.
In this section the experimental basis for this selection is first des-
cribed and followed by the description of method parameters at pH 5.6.
Additional experimental 'work at other pHis is described in Part III,
Activities Report.
*
Additional laboratory work and discussion of this method is given in
Appendix 3.
145
WALDEN RESEARCH CORPORATION
-------
A.
pH Dependence
Kanno et a1 (360) state that the absorbance
of the ch10rani1ate solution has a relatively strong pH dependence up to
a pH of 5.2, but not at higher values. In order to compensate for vari-
ations in sample pH it is necessary to employ a buffer with capacity
adequate to maintain absorbance within the desired limits. For this
purpose, it is desirable to operate in the pH range which shows least
absorbance sensitivity.
Kanno's findings have been checked on buffer
solutions prepared from 0.2M sodium acetate and 0.2M acetic acid solutions.
The results are shown below.
E.!:!. Absorbance vs H20
4.6 0.505
5.2 0.462, 0.465
5.4 0.455
5.6 0.451
5.8 0.449, 0.449
6.0 0.443, 0.446
6.4 0.447
6.6 0.449
5.8 Blank 0.013
The developed color becomes relatively inde-
pendent of solution acidity at pH 5.6. A sensitivity loss of ~ 10% occurs
(from pH 4.6). Experimental optimization of other major factors (at pH
5.6) is described below.
B. Color Development as a Function of Mixing
Time
Two experiments were performed to establish the
minimum mixing time needed for color development to stop, or decrease to a
barely discernible rate at room temperature.
146
WALDEN RESEARCH CORPORAT,ION.
Jiioi
-------
All solutions contained:
5 ml of K2S04 solution (2,500 ug/ml)
5 ml of pH 5.6 buffer
15 ml of sulfate-free water
25 ml of ethanol
0.2 gm of barium chloranilate
The solutions (except for 0 and l-minute
samples) were shaken on a Burrell Wrist-Action Shaker at the maximum
rate obtainable. The O-minute solution was simply swirled several times,
then centrifuged. The one-minute solution was hand shaken for one min-
ute at approximately the same rate as the solutions on the shaker. The
results are indicated below:
Absorbance (vs H20)
Time (mi ns . t Run 1 Run 2
o 0.394
1 0.455
2 0.459
5 0.462
10 0.462 0.470
15 0.463
20 0.471 0.475
25 0.468
30 0.471 0.478
60 0.480
15 (Blank) 0.028
30 (Blank) 0.029
A 20-minute mixing time was selected based on the above data.
C.
Color Development as a Function of Temperature
Hand shaking of the solutions was neces-
sary to avoid the awkwardness of thermostatting the solutions while on
the shaker. To determine relative efficiencies to samples were hand
shaken for 20 minutes at room temperature and two samples machine
,4
147
WALDEN RESEARCH CORPORATION
-------
shaken for 20 minutes at room temperature.
foll ows:
The absorbances obtained were as
Hand Shaking
0.452
0.453
Machine Shaking
0.454
0.455
Having found that hand shaking was satis-
factory, six samples were hand shaken in constant temperature baths, three
at 200C and three at 300C. Measured absorbances on these solutions were:
200C 300C
0.449 0.460
0.450 0.460
0.452 0.459
Due to an oversight, the absorbances were
not remeasured after the solutions reached room temperature. However,
previous measurements suggest that the changes on reaching room tempera-
ture will be = 0.002 or 0.003; thus, temperature correction is not neces-
sary if the sample and calibration curve solutions are mixed at a tempera-
ture differential of say 50C.
D.
Amount of Barium Ch10ranilate Required
Three similar solutions were prepared for
mixing with only the amount of barium ch1orani1ate differing. Total
solution volume was 50 m1. Results follow:
GMS Barium Ch10rani1ate
Absorbance vs H20
0.1
0.2
0.3
0.459
0.470
0.470
Thus, 0.2 gm of barium ch10ranilate ap-
pear sufficient to react with the sulfate present in the 50 m1 volume.
148
WALDEN RESEARCH CORPORATION
-------
E. Adherence to the Beer-Lambert Law at pH 5.6
The solutions were prepared from a stock
solution of K2S04 at a concentration of 2,500 ug S04=/m1. Final concen-
trations in the range 25 to 450 ug S04=/m1. The solutions were mixed on
a shaker at room temperature for 20 minutes, centrifuged at highest
speed attainable with an International centrifuge (clinical model) for
5 minutes, then read on a Beckman DU at 530 nm vs H20 using 1 cm cells.
The results are given below.
Concentration (Final Solutiont
ug S04-/m1
25
50
100
250
350
450
Absorbance
0.078
0.117
0.199
0.468
0.647
0.980
The plot of these data show linear dependence of absorbance on concen-
tration to at least a concentration of 350 ug su1fate/m1 for solution
of K2S04.
F.
Buffer Capacity
The buffer capacity of the pH 5.6 sodium
acetate-acetic acid buffer was determined by diluting known concentra-
tions of H2S04 to a fixed volume of 20 m1 adding 5 m1 of pH 5.6 buffer
and measuring the pH of the solution. An apparent pH was also measured
after the addition of 25 m1 of ethanol to the sulfate-acetate solution.
A duplicate set of solutions was prepared in volumetric flasks, barium
ch10rani1ate added and the absorbance measured after shaking and centri-
fuging. The results are given below.
149
WALDEN RESEARCH CORPORATION
-------
Volume of pH Before pH After "pH" After Absorbancea
0.025M H2S04 Addi ng Buffer Adding Buffer Adding Ethanol
o ml 5.6 5.6 6.6 0.023
1 ml 2.7 5.4 6.4 0.108
2 ml 2.4 5.2 6.2 0.199
3 ml 2.3 5.1 6.1 0.299
4 ml 2. 1 5.0 5.9 0.396
5 ml 2.1 4.9 5.9 0.491
--' 0.719b
(J1 7 ml 2.0 4.6 5.6
a 1.09b
10 ml 1.9 4.2 5.3
a Absorbances obtained from separate experiment performed in the same way.
b Not done at same times as 0.5 ml set.
~
....
C
ITI
Z
::0
ITI
en
~
::0
()
:r
()
o
::0
cg
::0
~
i5
z
-------
A plot of absorbance versus ml H2S04 does
not deviate from linearity until the pH after buffering fell to 4.6.
This correlates well with earlier work which showed absorbance sensitivity
below pH 5 (Section A).
These results accentuate the need for good
control of pH. Adjustment of pH prior to buffering the solution has,
therefore, been included in the procedure.
G.
Description of Proposed Method
The procedure used is as follows:
a. A volume of solution containing sulfate ions is pi petted into a 100
ml beaker.
b. Adjust pH to 5-6 with 0.2M sodium hydroxide.
Add five ml of pH 5.6 buffer (0.2M sodium acetate to which is added
sufficient 0.2M acetic acid to give a pH of 5.6).
d. Total volume is adjusted to 25 ml with sulfate-free water.
c.
e. Add 25 ml of ethanol.
f. Pour contents of beaker into 100 ml volumetric flask.
g. Add 0.2 g barium chloranilate.
h. Shake on wrist-action shaker for 20 minutes.
i.
Centrifuge in 15 ml centrifuge tubes at maximum RPM (or filter).
Decant into 1 cm cells and measure absorbance at 530 nm vs H20.
j.
H.
Precision of the Analytical Method
Using the procedure described in H (except
that K2S04 at the 250 ug/ml level is used rather than H2S04 and pre-
liminary neutralization), we have generated the following precision data:
151
WALDEN RESEARCH CORPORATION
-------
Absorbance vs H20 for Ten Separate Runs
0.473
0.472
0.477
0.475
0.473
0.477
0.475
0.471
0.473
0.478
Average = 0.474
Standard Deviation (lcr) = 0.002
Relat,'ve E - 0.002 x 100 - +0 4%
rro r - 0 . 474 - - . 0
1.
Interferences in the Barium Chloranilate
Method (pH 5.6)
a. Solutions of potential anionic inter-
ferences were prepared at 0.2M, O.lM, and 1% (in case of H202) concentra-
tions:
1. Procedure followed was:
a. 5 ml K2S04 solution.
b. 5 ml potential interference.
c. Adjust pH.
d. 5 ml pH 5.6 acetate buffer.
e. 25 ml ethanol and shake well.
f. Dilute to 50 ml with water.
g. Add 0.2 gm barium chloranilate and shake on machine for 20 minutes.
h. Centrifuge 5 minutes.
i. Measure absorbance in 1 cm cell at 530 nm vs H20.
b. Data
Material Added
Absorbance of Solution with Indicated Initial
Concentration Level of Interference
KN03
KCl
NaHC03
C).2M
0.472
0.453
0.495
O.lM
0.464
0.461
0.478
152
WALDEN RESEARCH CORPORATION
-------
Material Added
Absorbance of Solution with Indicated Initial
Concentration level of Interference
NaF
NaHS03
K2S04
0.468
0.473 (1%
solution)
not done
0.810
O.lM
0.685
0.875
0.465
0.468 (1%
solution)
not done
0.865
Kl04
Na2C204
CH20
H202
0.2M
1.43
references:
run 1 -- (0.2M solutions)
run 2 -- (O.lM solutions)
0.517*
0.508*
*
Prior runs out of the
0.470 as absorbance.
served here.
same batch of K2S04 solution consistently gave
No reason is known for the different value ob-
Again as with the titration procedure, oxalate
and phosphate are the two largest interferences.
153
WALDEN RESEARCH CORPORATION
-------
6.
STATISTICS OF FIELD SAMPLING AND ANALYSIS
6.1
Introduction
In this section we discuss analytical errors in general terms
and summarize the scattered literature references, as well as some here-
tofore unpublished data on the precision and accuracy of the analytical
and sampling procedures, separately or combined, depending upon the
source material. Finally we analyze in some detail the few comparative
evaluations of methods conducted as field trials.
The statistical analyses presented in this section have been
simplified in order to obtain suggestive results rapidly. For example,
the linear regression model employed here assumes that no portion of the
error is attributable to the independent (x) variable, although more
realistic models are available (532) and will be utilized in future work
where desirable. We believe this procedure is justifiable for quali-
tative results in a field where such comparisons have been conspicuously
absent.
6.1.1
Accuracy
In most manual methods, the pollutant of interest is
first trapped and then made to reduce or oxidize another species, react
to form a color, act as an acid or base, etc. Because these are all
general, class or type, reactions it is difficult to obtain specificity
and eliminate side reactions and/or interferences with other species
which lead to a loss in accuracy. This is particularly true for highly
labile species where there is opportunity for the gas to react before
an analysis can be completed. An ultimate but not necessarily attain-
able, goal of the analytical chemist is the use of a method which will
not disturb the system during measurement and which will provide com-
plete specificity.
Another factor which makes the accuracy of manual (or
instrumental) techniques not generally predictable is the wide range
of matrices encountered in combustion processes. To properly evaluate
accuracy one must carefully assess each matrix to which the method is
154
WALDEN RESEARCH CORPORATION
-------
to be applied. One hopes for and expects high accuracy but in practice
90-95% is reasonable for field sampling work.
It is very difficult to determine the accuracy of a
method in the field. However, a few workers have correlated stoichio-
metric S02 concentrations calculated from the sulfur content of the
fuel with measured S02 concentrations. If we assume that sulfur emis-
sion is quantitative, we can use this technique as a measure of the ac-
curacy under field conditions. This approach is discussed in Section 6.4.
6.1.2
Precision
Many factors contribute to imprecise results. These in-
clude reaction chemistry, operator errors, and equipment variability.
Various steps in the overall procedure, including a1iquoting, dilutions
to the mark, etc. each add to the imprecision. Techniques which depend
upon spectrophotometric read-out will have an inherent imprecision of up
to several percent (for a simple analytical instrument) resulting from
the inability of the operator to read the absorbance dial to any higher
precision. In addition to these factors there exist problems associated
with sample collection and sample instability and, finally. those of in-
terlaboratory variations.
Ultimately, the desired precision of the mean, which is
a function also of replication, must be related to the resources allo-
cated and optimization of the total emissions determination program.
6.2 Precision Obtainable in Field Sampling of Sulfur Oxides
Since the following data were obtained at power plants, the ac-
curacy of the methods involved was not determined in an absolute sense.
However, the precision (sampling and analysis combined) of these methods
observed under field conditions, can be computed from replicate measure-
ments.
We have made some recent measurements of S02 ana S03 concen-
trations at an oil-fired power plant (516). The sampling system and
155
WALDEN RESEARCH CORPORATION
-------
procedures were similar to those used previously (527) except that
electrically heated probes and sampling lines were used. S03 was col-
lected in midget impingers containing 80% isopropanol; S02 was collected
in midget impingers containing 3% hydrogen peroxide. The barium chloran-
ilate method was used for analysis of both S02 and S03' In five consecu-
tive tests, we obtained a coefficient of variation (standard deviation x
100/mean) for S03 (sampling and analysis) of !ll% (at 10 ppm). The
coefficient of variation for the S02 measurement (sampling and analysis)
was considerably better, !2.6% (at 1500 ppm). Seidman (78) obtained a
coefficient of variation of 1.5% for 33 S02 samples (sampling and analy-
sis) collected in peroxide and analyzed by barium titration using thorin
indicator, in good agreement with the precision of our colorimetric
analyses.
Corbett (347) utilized 80% isopropanol (IPA) absorption for
S03 and peroxide collection for S02 with turbidimetric readout (160)
for both. Data were obtained from about 30 tests run in duplicate at
an oil-fired power plant. Corbett1s values for the standard deviations
were used directly. We have computed the mean SO concentrations and
x
the coefficients of variation for the duplicate samples only. The re-
sults are given below:
Species
Mean SO
x
Concentration, ppm
Standard Deviation, ppm
S02
503
997
12.2
73.3
2.6
CV%
7.4
21.4
The coefficient of variation for S02 and 503 includes sampling
errors as well as source variation. The more severe sampling problem
for 503 is reflected in the much larger value of its coefficient of
variation. The large values of the coefficients of variation for SO
2
and 503 are presumably a result of the imprecise analytical method
(turbi dimetry) .
E550 Research Ltd. (520) has done a statistical analysis of
field data for 503 determined by collection in 80% isopropanol and
analyzed by turbidimetry. They obtain a standard deviation of 2.8 ppm
503 for the total method (sampling and analysis) of which the error due
156
WALDEN RESEARCH CORPORATION
-------
to the analytical method contributes (0.6 ppm, and sampling error 2.7
ppm). Thus the larger contribution to the total error is in sampling.
If we assume a mean of 15 ppm 503 (for oil-fired combustion) the total
coefficient of variation becomes !19%, in good agreement with Corbett's
observations.
Nine duplicate 503 samples obtained at a coal-fired power
plant collected in NaOH and analyzed by the benzidine method (548) show
the same precision as that obtained with isopropanol collection and
turbidimetric readout, CV = t19% at a mean 503 concentration of 61
ppm. 5ignifica~tlY, however, the mean 503 concentration (61 ppm) for
NaOH collection is considerably higher than the 10-30 ppm range
usually obtained at large fossil fuel fired plants. Gillham has com-
pared measurements of 503 collected in distilled water and caustic solu-
tion and found that caustic (NaOH) collection produced 503 results
which were nearly twice as high as the aqueous acid collection method
for 503' This bears out our earlier conclusion (5ection 3) that sta-
bilization of sulfite ion is a major problem in caustic collection
of 503'
We have obtained unpublished field data from Combustion
Engineering (498) on determinations of 503 collected by the controlled
condensation method, and 502' collected in peroxide. Analyses were
performed by titration with standard NaOH solution using bromphenol
blue indicator. The data were obtained at an oil-fired plant during
dolomite addition tests. We have grouped the 503 data (27 tests run
in triplicate) into high and low 503 concentrations ranges and calcu-
lated the precision statistics given in Table 6-1 below.
*Both 50 snd 50 are collected together, leading to the possibility of
oxidati6n of sutfite and consequent poor precision and accuracy in the
sulfate analysis.
157
WALDEN RESEARCH CORPORATION
-------
TABLE 6-1
PRECISION STATISTICS FOR S03 BY CONTROLLED CONDENSATION METHOD
*
No. of Triplicates
Mean
S03 Cone. (pprn)
Std. Dev. (ppm)
CV%
10
16
5.7
12.2
0.36
0.49
6.3
4.0
*
1 set di scarded
Combustion Engineering conducted 20 S02 tests with three or
more replicates. These data were also grouped into high and low con-
centration ranges. The calculated precision statistics are given in
Table 6-2 below:
TABLE 6-2
PRECISION STATISTICS FOR S02 BY PEROXIDE COLLECTION
Mean
Total No. S02 Cone. (ppm) Pooled
No. of Tests of Runs and Range Std. Dev. (ppm) CV%
9 30 664 35.0 5.3
(580-890)
11 33 1100 24.2 2.2
(930-1170)
The decreased precision at the lower S02 concentration is sig-
nificant. Above 1000 pprn the precision is comparable to that obtained
with methods specific for sulfate, and in good agreement with the pre-
cision of I2% reported by the Bay Area APCD (145) for this method. The
poorer precision at the lower S02 concentration is probably due to inter-
ferences from acidic species in the flue gas. The Japanese Industrial
Standard (430) does not recommend use of standard NaOH titration for
158
WALDEN RESEARCH CORPORATION
-------
S02 collected in peroxide at concentrations below 500 ppm because of
this problem. Thus, NaOH titration has a limited range below which
it should not be used if accurate results are expected.
In general non-specific analytical methods for sulfate, such
as the iodine method (see fOllowing) or peroxide collection followed
by NaOH titration, may provide results of good precision at high 502
concentrations but lose precision at low S02 concentrations, where
interferences appear to become significant. We have not yet located
any results at low S02 concentrations for peroxide collection followed
by analysis by the barium chloranilate or barium ion titration. We
may, however, utilize Corbett's (347) work which employs an imprecise
method specific for sulfate (turbidimetry). Selecting only the S02
concentrations below 1000 ppm we obtain the following statistics:
14
S02 Cone.
Ranqe (ppmt
200-900
Mean S02
Cone. (ppmt
685
Pooled Std.
Dev. (ppm)
52
CV%
No. of Duplicates
7.6
Comparison of these statistics to those for the complete range
of Corbett's work presented previously (p. 156) reveals that~ although
the mean S02 concentration was reduced from 997 ppm, the coefficient
of variation remains essentially constant (7.6 compared t~ 7.4%).
The conclusion of this indirect argument is that methods which are
specific for sulfate appear to provide nearly the same precision over
a broad range of S02 concentrations, although non-specific methods
yield poorer precision at low S02 concentrations.
Sulfur Oxide statistics reported by the British 5taDdards
Institute are given in Table 6-3 (379).
S02 is determined by collection in standard iodine solution
and measurement of the volume of gas required to decolorize the solu-
tion. S03 is collected in 80% isopropanol and titrated with BaC12
using thorin indicator.
159
WALDEN RESEARCH CORPORATION
-------
TABLE 6-3
BSl SULFUR OXIDE STATISTICS
Sulfur dioxide
Standard deviation (10 results): 10 ppm at 200 ppm by volume
20 ppm at 2000 ppm by volume
Accuracy: +5 percent at 200 ppm by volume
t5 percent at 2000 ppm by volume
Su lfur tri oxi de
Standard deviation (10 results): 0.2 ppm at 5 ppm by volume (169)
0.2 ppm at 20-50 ppm by volume
Accuracy: +10 percent at 5 ppm by volume
t5 percent at 20-50 ppm by volume
Lower limit of detection: 0.5 ppm by volume
The standard deviation cited for S03 at 5 ppm in the BSI docu-
ment (374) appears to be in error and is corrected above. The source
cited by BSI, Fielder and Morgan (169) give the standard deviation (of
the analytical method alone) for 5 ppm (equivalent) S03 as 0.2 ppm (not
.02 ppm as given in reference 379).
At 2000 ppm S02' the iodide-iodine method has a reported stand-
ard deviation of 20 ppm yielding a coefficient of variation of 1%. This
is the error in determination of the total volume (reported as t5 sec
for a 10 minute run, reference 170). The "accuracy" (not defined) stated
as t2% may include factors not explicitly stated. At 200 ppm S02' the
coefficient of variation becomes 5%, which is stated "accuracy". Al-
though it is clear that accuracy and precision should decrease at lower
concentration, these results are not self-consistent, since no addi-
tional factors appear in the "accuracy" at low concentrations. As a
result of the vague definitions employed and apparent inconsistencies,
we cannot properly evaluate these data. It is difficult to accept the
160
WALDEN RESEARCH CORPORATION
-------
results at 2000 ppm S02 as representative of the precision to be ex-
pected for a non-specific method subject to many interferences.
For S03 the value given for "accuracy" probably represents
the errors in sampling and analytical methods combined. If this is
true, the value obtained at 5ppm S03 (!10%) is in good agreement with
our results (!11%) using the barium ch10rani1ate method.
Data on S02 collection by absorption in NaOH was obtained at
a coal-fired power plant (548). The NaOH solution, after collection,
was divided into three a1iquots. Each aliquot was analyzed by a dif-
ferent analytical method. The methods used were the benzidine, the
*
modified Shell (barium ion titration with thorin indicator), and the
Berk and Berdick. These methods have been described in detail by
Smith et a1 (76). The mean S02 concentrations as well as the coef-
ficient of variation for each analytical method are given in the table
below for a1iquots of 5 duplicate samples.
TABLE 6-4
COMPARISON OF ANALYTICAL METHODS FOR S02 AFTER CAUSTIC ABSORPTION
Analytical Method
Mean S02 Cone. (ppm)
CV%
Benzidine
"modified Shell II
Berk & Berdick
(HC1 titration with
bromphenol blue indicator)
1362
1356
1780
6.8
5.3
3.3
*
The sample is collected in caustic rather than peroxide, and subsequently
oxidized.
161 '
WALDEN RESEARCH CORPORATION
-------
The most precise analytical results for caustic scrubbing are
apparently obtained with the Berk and Berdick (B&B) method. However,
the mean S02 concentration obtained by the B&B method is 30% higher than
the S02 concentration obtained by either of the other two techniques. We
point out again that these are analyses of aliquots of the same NaOH so-
lution and should yield identical results for all three methods. These
results suggest that the accuracy of the B&B method is quite poor, at
least in this case.
Another significant feature of the above results is that the
modified Shell method gives a coefficient of variation of 5.3% compared
to 1.5% [results of Seidman (78)] when peroxide collection is used. This
again emphasizes the importance of the collection medium.
6.3 Comparison of Sampling Errors and Analytical Errors in the
Determination of Sulfur Oxides
In the previous section, the precision of both sampling and
analysis was combined. It is of interest to know the magnitude of error
in each component (sampling and analysis) in order to minimize the error
in each process. Sampling and analytical errors can be resolved by de-
termination of the errors due to the analytical method only and assuming
that all other errors are due to sampling.
The combined sampling and analytical error (precision) for 80%
IPA absorption with analysis by the barium ch10rani1ate is !ll% (Section
6.2). The analytical error for the barium chloranilate method is !4% at
10 ppm S03 (516). We may determine the error in sampling ~ from the
difference between the total CV and the CV due to the analytical method
as foll ows:
162
WALDEN RESEARCH CORPORATION
-------
(CVS)2 = (CVT)2 - (CVA)2
(CVS)2 = (11)2 - (4)2
CVS = 10%,
where CVS is the coefficient of variation attributable to sampling pre-
cision. The largest fraction of the error (87%) is due to sampling.
The precision (sampling and analysis) of the controlled con-
densation collection method for S03 is considerably better (CV = 6%)
than 80% IPA absorption.
Since the error in the analytical method (NaOH titration) is
about the same as colorimetric analysis, we have:
(CVS)2 = (6.0)2 - (4.0)2
CVS = 4.5%
The collection method is, thus, an important factor in the overall reli-
ability since the sampling error in the 80% IPA absorption method is
nearly twice the sampling error for the controlled condensation method.
This reduced error for S03 collection by controlled condensation is not
surprising since the problem of oxidation of dissolved S02 is eliminated.
The precision of the barium chloranilate analytical method for
S02 determined from duplicate analyses of aliquots of 23 stack samples
(516) is !2% (at an absorbance of 0.5). The coefficient of variation
found for S02 sampling and analysis was !2.6%. If we assume no source
variation we may estimate the error in S02 concentration resulting from
sampling from these values as follow~:
(CVT)2 = (CVA)2 + (CVs)2
(2.6)2 = (2.0)2 + (CVs)2
(CVS) = 1.3%
163
WALDEN RESEARCH CORPORATION
-------
Thus, if 2% of the total error is due to the analytical method, 1.3%
can be attributed to the sampling error for S02. This result is not
unexpected in view of the high concentration of S02 and the quantita-
tive collection of S02 in hydrogen peroxide. For peroxide collection
followed by barium ion titration with thorin indicator we obtain 1.5%
for CVT and 1.0% for CVA'* whence CVS for this method is 1.1%, in good
agreement with the results above. The sampling error for S02 is com-
parable to the ~nalytical error, in contrast to S03 determination by
absorption.
6.4 Comparison of Stoichiometric to Measured S02 Field Results
We have located two sets of data for which stack analyses
have been systematically compared to S02 flue gas concentrations esti-
mated from the sulfur content of the fuel. These data may be used to
estimate the accuracy of various methods in field sampling. Unpublished
data on a comparison of S02 analytical results conducted in 1963 at a
coal-fired utility boiler by NAPCA personnel using the Shell Develop-
ment Method, have been furnished by R. Larkin of NAPCA (548). Stoich-
iometric sulfur dioxide values were determined from total volumetric
flow and sulfur content of the fuel.
To obtain a measure of the accuracy of the Shell method, we
have conducted a simple regression analysis of the Shell data (10
points) versus the stoichiometric values, taken as the dependent vari-
able. This procedure is, of course, as dependent upon the accuracy of
the calculation as upon the method of analysis. (See footnote, Section
6.2.)
The regression equation obtained is:
Yst = 40 + 0.86 Xs
*
NPACA data on 12 aliquots of S02 stack samples (548).
164
WALDEN RESEARCH CORPORATION
-------
Yst = 1108 ppm
SE = 79 ppm
CV = 7.1%
A scatter diagram for the data and the regression equation
(solid line) are shown in Figure 6~1. The relatively low value of the
coefficient of variation is encouraging in view of the severe sampling
problems encountered in a coal-fired utility.
6.5 Conclusions
S03 collected by absorption in a liquid is relatively imprecise
(and inaccurate) as a result of oxidation of dissolved S02. The coeffi-
cient of variation observed is in the range ilO-20% (sampling and analy-
sis) depending on the analytical method. The controlled condensation
method, which eliminates the problem of oxidation of dissolved S02'
yields considerably better precision. The coefficient of variation for
this technique was found to be i5% for sampling and analysis. The con-
trolled condensation method is the most precise and/or accurate collec-
tion technique for S03 from fossil fuel combustion sources.
Caustic collection for S02 yields poor precision presumably as
a result of scrubbing of interfering flue gas components. With barium
ion titration and thorin indicator, a coefficient of variation of 1.5%
(sampling and analysis) is achieved for peroxide scrubbing of S02.
Caustic scrubbing followed by barium ion titration yields considerably
poorer results, CV = 5.3% (sampling and analysis). Since the analytical
method is the same, caustic scrubbing techniques do not appear attractive.
The iodine-iodide scrubbing solution technique lacks the necessary quanti-
tative field data for assessment and the many probable interferences pre-
clude its selection as a precise and accurate method for combustion sources.
Peroxide collection is the most widely used technique because the inter-
ferences are minimal and it yields precision better than i3% in field use.
NaOH titration coupled with peroxide collection provides good precision
(CV = 2.2%) at high (1100 ppm) S02 concentrations, but at lower concentra-
tions the precision is considerably poorer (CV = 5.3%, at 664 ppm S02).
165
WALDEN RESEARCH CORPORATION
-------
2000
1500
1000
500
o
500
y = 40 + 0.86Xs (ppm)
st
o
o
1000
1500
2000
Measured S02 (PHS-Shell Method), ppm
Figure 6-1.
Comparison of Stoichiometric (calculated)
to Measured S02 Concentrations (coal-fired
boiler)
166
WALDEN RESEARCH CORPORATION
-------
I~?: .,
This is characteristic of a non-specific method for sulfate determina-
tion.
The accuracy of the best field methods for S02 as judged by
crude comparison to limited stoichiometric calculations apparently falls
in the range of 3-8%. Subjective evaluation suggests 5 to 10% as the
range to be expected in the emission determination for S02 at ~ 1000 ppm.
There is as yet no basis for determining the accuracy of S03
field measurements.
167
WALDEN RESEARCH CORPORATION
r.,
.;. !,~
.,.
-------
7.
RECOMMENDED METHODS
7 - 1
Collection Methods for S03
Our objective is the selection of (manual) collection and analy-
tical procedures which will gi~e the highest possible precision and ac-
curacy for S03 in the range 5 to 10-300 ppm. It is thus important to
choose a system with a high (and therefore, reproducible) collection ef-
ficiency. The controlled condensation technique has been shown to
have a high collection efficiency in several different configurations (80,
81,403,411). The primary basis of selection is, however, the observed
greatly improved precision of the controlled condensation method com-
pared to IPA absorption techniques.
A number of objections have been expressed to the field use of
the condenser method. These include:
1.
2.
3.
4.
Difficulty in maintaining temperature
Filter plugging
Fragility of equipment
Possible failure in dry process streams
The first objection may be overcome by using a constant boiling
liquid in the condenser. The only auxiliary equipment needed in this
case is a heating tape, eliminating the need for a temperature control-
ler and thus improving equipment portability. Combustion Engineering
uses a simple thermostat for temperature control. The second problem
is of a major consequence since the important coal-fired plants pro-
duce an effluent with high particulate loading. Redesign of the samp-
*
ling probe to divert the larger particles would reduce the particulate
loading on the filter. A quartz or pyrex pre-filter is satisfactory
for most situations. The third objection can be solved by using a
stainless steel jacket around the condenser (80,502). The last problem
*
For example, to include a miniature cyclone.
i68
WALDEN RESEARCH CORPORATION
-------
does not, of course, apply to combustion flue gases for which the
condenser technique was developed. For dry process stream 503 measure-
ments, the Flint technique should be compared to condenser methods at
low temperatures.
On the basis of the preceding, we rate collection methods
for S03 in the following order:
l.
2.
Controlled Condensation
Absorption in 80% Isopropanol
Other methods are either unsatisfactory or untested. A com-
plete sampling system which may be used for either method is described
in Appendix 4.
7.2 Collection Methods for S02
The most widely used collection methods for S02 have been based
upon aqueous scrubbing media. Solid sorbents appear to have potential
for development as specialized collection techniques, but no directly
useable experimental data have yet been obtained. Dilution methods are
inherently complex and offer no obvious advantages.
Three major aqueous media have been widely used in fossil
fuel effluent collection, viz. peroxide, caustic and iodine. The re-
maining aqueous media utilized primarily in kraft mills and ambient
sampling all present numerous problems and few advantages for fossil
fuel use.
Both caustic and iodine collection methods have been shown
to be subject to interference resulting in poor precision, particularly
at low concentrations and/or in a "dirty" environment.
Collection in hydrogen peroxide is the most widely used
nique and yields the best precision in field use. An additional
vantage is that any sulfate method may be used for analY5is as a
tech-
ad-
result
169
WALDEN RESEARCH CORPORATION
-------
of the ease of peroxide decomposition.
the method of choice for 502'
A sampling system for 502 collection is described in Appendix
Peroxide collection is clearly
4.
7.3 Analytical Methods
The barium chloranilate method, and the Ba(C104)2 titration
with thorin (or related) indicator provide the greatest selectivity and
highest accuracy and/or precision for manual methods of sulfate
analysis.
The major difficulty of the titrametric method using thorin
indicator is the poor endpoint detectability. A number of different
indicators have been tested but it is not clear that they offer sig-
nificant improvement. Endpoint detectability may be improved by
photometric readout (169,352).
The bariu~ chloranilate technique is recommended because
it eliminates the need for subjective judgement on the part of the
operator, is relatively rapid, and can be applied over a wide concen-
tration range. The Shell Development method (barium ion titration) is
recommended as a result of the good statistical correlation observed
in field tests, where experienced analysts are available. Detailed
procedures for both methods are appended.
Analyses based upon determination of sulfite ion are not as
desirable as sulfate methods. Even the West-Gaeke procedure for low
concentrations of 502' which offers the advantage of specificity, can-
not be advantageously applied to combustion effluent analysis as a
result of the complexity of the required apparatus and poorly under-
stood interferences.
7.4 Simplified Methods
A simplified procedure may be obtained by determination of
total SOx in peroxide by elimination of the 503 scrubber. This technique
170
WALDEN RESEARCH .CORPORATION
-------
has been utilized by B.S.I. (618). There are intriguing possibilities for
development of simplified SO methods based upon reaction with reactive
x
solid sorbents such as alkali carbonates or supported oxidants such as
Mn02 on alumina. Indicating silica gel (detector tubes) also appear
promising if proper controls are maintained.
A simplified procedure for the determination of SO by cal-
x
lection in peroxide followed by determination of total acidity is ap-
pended.
7.5 Correlation of Combustion Source Type with Selection of
Sampling and Analysis Techniques
The selection and application of the recommended methods for
sampling and analysis for the oxides of sulfur in stationary fossil
fuel combustion sources is generally straightforward since a single
method is applicable to all sources for a given contaminant. On the
other hand, relatively high level R&D planning is required in some
instances, e.g., the ab initio determination of S02 emissions from a
large furnace with maximum precision and accuracy.
The principal factors to be correlated are:
l.
2.
3.
4.
The precision and accuracy sought
The concentration of the species sought
Source size
Transients
The precision and accuracy sought while primarily mission-
oriented will determine the selection of method, i.e., precision or
simplified. Should emission standards be set in terms of permissible
lbs/hr, large sources will require higher precision analyses than
small sources.
The major source factors proper are related to species concen-
tration and operating transients. Species concentration, taken in the
171
WALDEN RESEARCH CORPORATION
-------
general sense, here, includes spatial and temporal variance. As shown
in Section 1.0, the primary interest for emissions determination is in
relatively large sources. Thus, the special problems of small sources
have been treated only briefly.
Gross species concentrations (S02' S03 and SOx) are of pr~mary
importance for determination of sampling duration in order to obtain
a sample size adequate for the precision sought. For S02 and SOx'
simple stoichiometric calculations (see procedures) allow ~ priori
determination of sampling duration required for desired sample size.
For S03 sampling in high sulfur environments (residual oil and coal
fired) a safe rule is to assume that 1% of the calculated SO will
x
appear as S03 since experience shows this to be a lower bound in well
controlled combustion at low excess air. Should sampling duration
become inconveniently long, the various high sensitivity options of
the barium ch10ranilate procedure may be applied. The IPA-thorin
procedure is not recommended for low S03 concentrations for this reason.
Neither S03 procedure has been designed for low sulfur (distillate
fuel oil) sources since this application is remote.
A distinction in the dispersion of a population of S02
measurements may be expected between coal and residual oil-fired
sources if the samples are taken for short periods (say 15 minutes)
over an interval of hours. This results from the greater short term
variance of the sulfur content of coal compared to the more homo-
geneous residual oil supply (Section 4.2.5). A high precision emis-
sion determination for a coal-fired source may, therefore, require
either flue gas sampling over an extended period (week) or correlation
of short term samples with representative (as burned) coal samples.
Where such coal sampling and analysis is feasible, it is recommended
as a simpler procedure than long-term flue gas sampling.
172
WALDEN RESEARCH CORPORATION
-------
8.
CONCLUSIONS
The primary conclusion that may be drawn from this analysis of high
precision manual sampling methods for S02 and S03 is that the statistical
basis for evaluation of existing methods is fragmentary, but reasonably
self-consistent. The best present methods for S02 analysis, in the usual
range of concentration, i.e., 500-2500 ppm, appear capable of yielding a
precision of t3% in stack analysis. The accuracy of these determinations
should approach t5%, but the statistical basis for this value is yet weak.
Peroxide collection methods for S02 appear best; caustic scrubbing is
clearly the poorest major technique. The barium chloranilate analysis for
sulfate should provide a major improvement in the elimination of the
judgment error associated with the barium ion titrations of precision.
Analysis for S03' in the usual range of interest, 10-50 ppm, is con-
siderably more difficult and less precise than that for S02' as is gen-
erally known. For IPA absorption, precision of the determination is
generally in the range 10-20%. The condenser method provides consider-
able improvement in precision. Precision of t5% at 10 ppm S02 has been
obtained.
It is recommended that considerable effort be devoted to collection
of significant sampling and analysis statistics for improved methods.
173
WALDEN RESEARCH CORPORATION
-------
78.
*
LITERATURE CITED
1.
J. R. Taylor, W. D. Holland, R. D. MacPhee, and K. H. Schoenemann,
Edltors Labora~ethods Air Pollution Control District, County
Of' Los Angeles (1958). '
R. F. Davis and W. E. O'Neill, "Determination of Oxides of Nitro-
gen in Diesel Exhaust Gas by a Mod~fied Saltzman Method~" U:S.
Dept. of the Interior, Bureau of Mlnes Report of Investlgatl0ns
6790, 5 pp. (1966).
9.
13. A. C. Stern, Editor, Air Pollution, Volume III, "Sources of Air Pol-
lution and Their Control," Academic Press, New York and London,
866 pp. (1968).
22.
S. T. Cuffe, "Air Pollutants from Power Plants," Archives of Environ-
mental Health ~ (3), 128-33 (1963).
G. K. M. Wang, "An Instrument for Determining Sulfur Oxides in Flue
Gases," Combusti on 38 (11), 46-49 (May 1967).
35.
51. W. O. Negherbon, "Sulfur Dioxide, Sulfur Trioxide, Sulfuric Acid
and Fly Ash: Their Nature and Their Role in Air Pollution,"
Edison Electric Institute, New York, EEl Publication No. 66-900
(June 1966).
52.
"Determination of Sulfur Dioxide and Sulfur Trioxide in Stack Gases,"
Shell Development Co., Analytical Dept., Emeryville Method Series
4816/59a, 6 pp. (1959).
55. Ulrich Fauth and W. Schille, "Gaseous and Solid Emissions from Oil-
Fi red Stoves," Staub 27 (6), 1-11 (1967).
60. S. Fukui, "Determination of Sulfur Oxides in Stack Gases by the
Arsenazo III Method," Japan Analyst J.!, 838-842 (1965).
74.
R. W. Gerstle, S. T. Cuffe, A. A. Orning, and C. H. Schwartz, "Air
Pollutant Emissions from Coal-Fired Power Plants," Report No.2,
Journal of the Air Pollution Control Association 15 (2), 59-64
(1965). -
76. J. F. Smith, J. A. Hultz, and A. A. Orning, "Sampling and Analysis
of Flue Gas for Oxides of Sulfur and Nitrogen," U.S. Dept. of the
Interior, Bureau of Mines Report of Investigations 7108, 21 pp.
(April 1968).
E. B. Seidman, "Determination of Sulfur Oxides in Stack Gases,"
Analytical Chern. 30 (10), 1680-1682 (1968).
*
These references were selected from a larger bibliography and are not
necessarily consecutive.
174
WALDEN RESEARCH CORPORATION
-------
LITERATURE CITED (continued)
80.
E. S. Lisle and J. D. Sensenbaugh, liThe Determination of Sulfur Tri-
oxide and Acid Dew Point in Flue Gases," Combustion 36 (7), 12-16
(January 1965). -
81. H. Goksoyr and K. Ross, liThe Determination of Sulfur Trioxide in
Flue Gases," J. of the Institute of Fuel 35, 177-179 (1962).
93. J. L. Mills, K. D. Luedtke, et al., "Emissions of Oxides of Nitro-
gen from Stationary Sources in Los Angeles County," Los Angeles
County, Air Pollution Control District Report No.4, 34 pp. (July
1961) .
95.
"Atmospheric Emissions from Sulfuric Acid Manufacturing Processes, II
U.S. Dept. of Health, Education and Welfare, Public Health Service
Publication No. 999-AP-13, 127 pp. (1965).
96. S. Hochheiser, "Methods of Measuring and Monitoring Atmospheric Sul-
fur Dioxide," U.S. Dept. of Health, Education and Welfare, Public
Health Service Publication No. 999-AP-6 (August 1964).
10-3. V. P. Popov and S. M. Grushetskaya, "Determination of the Concentra-
tion of Sulfur Trioxide in Products of Combustion," Industrial Lab.
32 (1), 26 (1966).
104. W. Schneider, "Simple Apparatus for the Determination of Sulfur
Oxides in Smoke and Exhaust Gases," Chemische Technik (Berlin) 20
(5), 299-300 (1968). -
116. P. Colombo, D. Corbetta, et al., "Critical Discussion on the Analyt-
ical Methods for Mercaptan and Sulphur Compounds," TAPPI 40 (6), 490-
498 (1957).
119. M. Hissink, "Determination of S03 and S02 in (Flue) Gases," Chemisch
Weekblad 58, 460-462 (1962).
125. S. Kanno, S. Fukui, et al., "Research on Sampling Method and Compari-
son of Several Analytical Methods for Determination of Sulfur Oxides
in Flue Gases," Japan Analyst li (11),1013-1018 (1965).
127. J. P. Chory, "Method for Quantitative Determination of S02 and S03
in Flue Gases," Brennstoff-Warme-Kraft li, 601-603 (Dec. 1962).
128. G. Graue, W. Gradtke, and H. Nagel, "Modification of the Pararosani-
line Method for the Measurement of Sulphur Dioxide Emission," Staub
25 (12), 525-527 (1965).
129. V. Hraseova and J. Vocel, "Colorimetric Determination of Sulfur Di-
oxide in Combustion Products," Energetika J!. (7), 344-346 (1966).
175
WALDEN RESEARCH CORPORATION
-------
131.
137.
139.
LITERATURE CITED (continued)
D. Flint "A Method for the Determination of Small Concentrations of
S03 in the Presence of Larger Concentrations of S02," J. of Soc. Chern.
Indus. 67, 2 (January 1948).
H. Stratmann, "Microanalytical Methods for Determinin~ Sulfur Di-
oxi de in the Atmosphere," Mi krochimi ca Acta, 668-678 (1954).
R. Cole and H. L. Shulman, "Absorbing Sulfur Dioxide on Dry Ion Ex-
change Resins -- For Reducing Air Pollution," Indus. and Engineer-
ing Chern. ~ (10), 859-860 (1960).
140. P. W. West and G. C. Gaeke, "Fixation of Sulfur Dioxide as Disulfito-
mercurate (II) and Subsequent Estimation," Analytical Chern. 28 (12),
1816-1819 (1956).
145.
Regulation 2, Bay Area Air Pollution Control Board, 1480 Mission St.,
San Francisco (Adopted May 4, 1960, Revised pages published January 1,
1962) .
153. J. W. Laxton, P. J. Jackson, "Automatic Monitor for Recording SUlphur
Trioxide in Flue Gas," J. of the Instit. of Fuel 37,12-17 (1964).
154. Av Kjell Andersson and Jan G. T. Bergstrom, "Determination of Hydro-
gen Sulfide and Sulfur Dioxide at Low Concentrations," Svensk Papper-
stidning 70 (23), 805-808 (December 15, 1967).
159. S. Kanno, "The Colorimetric Determination of Sulfur Oxides in Atmo-
sphere," Inter. J. of Air & Water Poll. 1, 231-233 (Jan. 1959).
160.
161.
162.
164.
165.
169.
P. F. Corbett, "Phototurbidimetric Method for the Estimation of S03
in the Presence of S~," J. Soc. Chern. Indus. 67,227-230 (1948).
E. W. F. Gillham, "The Determination of Oxides of Sulphur in Flue Gas,"
J. Soc. Chern. Indus. 65, 370-372 (1946).
E. W. F. Gillham, "The Automatic Estimation of Sulphur Trioxide in
Flue Gas," J. Soc. Chern. Indus. 65,405-409 (1946).
R. J. Bertolacini and J. E. Barney, II, "Colorimetric Determination
of Sulfate with Barium Chloranilate," Analytical Chern. 29 (2), 281-
282 (February 1957). -
D. H. Napier and M. H. Stone, "The Collection and Turbidimetric Es-
timation of Sulphur Trioxide in Flue Gases," J. App. Chern. 8, 787-
793 (December 1958). -
R. S. Fielder and C. H. Morgan, "An Improved Titrimetric Method for
Determining Sulphur Trioxide in Flue Gas," Analytica Chimica Acta 23
(6), 538-540 (1960).
176
WALDEN RESEARCH CORPORATION
-------
LITERATURE CITED (continued)
170. R. S. Fielder, P. J. Jackson, and E. Raask, liThe Determination of
Sulphur Trioxide and Sulphur Dioxide in Flue Gases," J. of Instit.
of Fuel 33, 84-89 (1960).
178. A. W. Fletcher, liThe Determination of Sulphur Trioxide in Gases,"
Chemistry and Industry, 777-778 (1954).
180. J. B. Lombardo, "Analysis of Sulfuric Acid Contact Plant Exit Gas,"
Analytical Chern. 25 (1), 154=160 (1953).
182. G. D. Patterson and M. G. Mellon, "Determination of Sulfur Dioxide
by Color-Changing Gels," Analytical Chern. 24 (10), 1586-1590 (1952).
185. D. Beinstock and F. J. Field, "Bench-Scale Investigation on Removing
Sulfur Dioxide from Flue Gases," J. of Air Poll. Control Assoc. 10
(2), 121-125 (April 1960). -
192. G. Dragt and K. W. Greenan, "Contact Sulfuric Acid Manufacture -
Evaluation of the Reich Test," Analytical Chern. 14 (11), 883-885
(November 15, 1942). -
193. R. T. Sheen and H. L. Kahler, "Direct Titration of Sulfates -- Fur-
ther Studies with Tetrahydroxyquinone as an Internal Indicator,"
Analytical Chern. ~ (1), 127-130 (March 15, 1936).
195. S. Atkin, "Determination of Sulfur Dioxide in Presence of Sulfur
Trioxi de," Analyti cal Chern. 22 (7), 947-948 (July 1950).
196. J. S. Fritz and M. Q. Freeland, "Direct Titrimetric Determination of
Sul fate, II Ana lyti ca 1 Chern. 26, 1593-1595 (1954).
198. J. L. Mills, K. D. Luedtke, P. Wool rich and L. Perry, "Emissions of
Oxides of Nitrogen from Stationary Sources in Los Angeles County, II
Report No.3: Oxides of Nitrogen Emitted py Medium and Large
Sources, Los Angeles County Air Pollution Control District, 51 pp.
(Apri 1 1961).
204. Manual on Disposal of Refinery Wastes, Vol. V: Sampling and Analysis
of Waste Gases and Particulate Matter, American Petroleum Institute,
Division of Refining, 1271 Avenue of the Americas, New York 10020
(1954) .
207. P. F. Corbett and W. M. Crane, "Review Series No. 111 Determination
of S02 and S03 in Flue Gases," The British Coal Utilization Res.
Assoc. J&. (1),1-11 (1952).
210. A. A. Berk and L. R. Burdick, "A Method of Test for S02 and S03 in
Gases," U.S. Dept. of the Interior, Bureau of Mines Report of Inves-
tigations 4618, 9 pp. (January 1950).
177
WALDEN RESEARCH CORPORATION
-------
LITERATURE CITED (continued)
215. P. F. Corbett, "The Determination of S02 and S03 in Flue Gases,"
J. of Instit. of Fuel 24, 247-251 (November 1951).
217.
218.
221.
P. H. Crumley, H. Howe, and D. S. Wilson, "The Aut0'!1atic Determina-
tion of Sulphur Trioxide in Flue Gases,1I J. of Instlt. of Fuel 11,
378-382 (September 1958).
R. A. Mott and K. Parramore, liThe Determination of Sulphur Trioxide
in Flue Gases: I," J. of Instit. of Fuel 30,123-126 (1957).
R. E. Matty and E. K. Di eh 1, "Measuri ng Fl ue-Gas S02 and S03' II Power
101 (11),94-97 (1957).
231. E. Krejcar, "Anion-Exchange Resins as Sorbents of Sulphur Dioxide,"
Chemicky Prumysl li (2), 77-79 (1965).
252.
263.
K. Wickert, liThe Determination of S02' S03 and H2S in Flue Gases,"
Brennstoff-Warme-Kraft II (10), 449-451 (1960).
W. Radmacher and K. Hadicke, "Determination of S02 and S03 in Flue
Gases,1I Staub ~ (6),174-176 (1958).
264. W. Schneider, "Simple Methods for Determining S02 and S03 in Flue
Gases," The British Coal Utilization Research Assoc. 16 (l), 1-11
(1952). -
265. A. Radwanska, IIDetermination of S02 and S03 in Flue Gases," Gaz,
Woda I Technika Sanitarna !L (5), 171-173 (1967).
269. B. Klein, "Microdetermination of Sulfate,1I Indust. and Eng. Chern.,
Analytical Ed. ~ (8), 536-537 (1944).
271. J. E. Barney, "Determination of Sulphate with Barium Chloranilate,"
Talanta II (4), 425-426 (1965).
273.
286.
R. M. Carlson, R. A. Rosell, and W. Vallejos, IIModification to In-
crease Sensitivity of Barium Chloranilate Method for Sulfate,"
Analytical Chern. 39 (6), 688-690 (1967).
P. H. Crumley and A. W. Fletcher, liThe Emissions of Sulphur Gases
from a Domestic Solid-Fuel Appliance,1I J. of the Instit. of Fuel
30, 608-612 (November 1957).
300. Shell Development Co. Method for the Determination of S02-S03 as
Adapted by the New Jersey State Dept. of Health, Air Pollution
Control Program, 2 pp.
303. J. S. Fritz and S. S. Yamamura, IIRapid Microtitration of Sulfate,1I
Analytical Chern. 27 (9), 1461-1464 (1955).
178
WALDEN RESEARCH CORPORATIOI
-------
LITERATURE CITED (continued)
306. H. N. Alyea and H. L. J. Backstrom~ 1I0xidation of Sulfite Solutions,1I
Amer. Chern. Soc. J. ~, 108-109 (January 1929).
308. L. Dubois, C. J. Baker, T. Teichman, A. Zdrojewski, and J. L. Monkman,
liThe Determination of Sulfuric Acid in Air: A Specific Method,1I
Mikrochemica Acta (2)~ 269-279 (1969).
310. A. W. Fletcher, IIAn Improved Apparatus for the Determination of Sul-
phur Tr'ioxide in Gases,1I Chern. and Ind. (16),303-304 (April 28,1956).
311. R. J. Bertolacini and J. E. Barney, II, IIUltraviolet Spectrophoto-
metric Determination of Sulfate, Chloride, and Fluoride with Chlor-
anilic Acid,1I Analytical Chern. 30 (2), 202-205 (February 1958).
312. D. W. E. Axford and T. M. Sugden, liThe Estimation of Sulphur Tri-
oxide in its Mixtures with Sulphur Dioxide by the Method of Ampero-
metric Titration,1I J. of Chern. Soc. (London), 901-903 (1946).
333. A. M. G. Macdonald, IIAnalysis for Industry,1I The Ind. Chemist, 345-
348 (July 1960).
337. J. Seith and H. G. Heitmann~ IIlmprovements in or Relating to the De-
termination of Gas Mixtures,1I British Patent 970,609 (August 13, 1963).
345. E. Hakoila and p. Noponen, IIEffects of Several Ions on the Direct
Volumetric Determination of Sulphate 10n,1I Acta Chemica Scandinavica
Ii (4), 947-952 (1965).
347. P. F. Corbett, liThe S03 Content of the Combustion Gases from an Oil-
Fired Watertube Boiler,1I J. Instit. Fuel 26, 92-106 (August 1953).
352. H. H. Bovee and R. J. Robinson, IISpectro-Visual Method for Deter-
mining Endpoints -- Application to Titration of Soluble Su1fate,1I
Analytical Chern. 29 (9), 1353-1355 (September 1957),
353. J. S. Fritz, S. S. Yamamura and R. J. Richard, IITitration of Sulfate
Following Separation with A1umina,1I Analytical Chern. 29 (1), 158-161
(January 1957).
354. J. R. Munger, R. W. Nipp1er and R. S. Ingols, IIV01umetric Determina-
tion of Sulfate 10n,1I Analytical Chern. ~ (11), 1455-1457 (November
1950) .
360. S. Kanno, S. Fukui, S. Naito, and M. Kaneko, et al., IISome Improve-
ment on Barium Chlorani1ate Method,1I Eisel Kagaku 13 (4), 217-218
(1967) .
361. R. W. Klipp and J. E. Barney, II, IIDetermination of Sulfur Traces in
Naphthas by Lamp Combustion and Spectrophotometry,1I Analytical Chern.
IL, 596-597 (1959).
179
WALDEN RESEARCH CORPORATION
-------
374.
LITERATURE CITED (continued)
Fuel Flue Cases: A Digest of Their Properties, Behavior, and Utiliza-
tion, C. G. Segeler, Editor, American Gas Association, N.Y., 198 pp.
-( 1941).
376. T. Ki tagawa and Y. Kobayashi, "Gas Analysi s by Means of Detector
Tubes. XI. Rapid Method for the Determination of Sulfur Dioxide,"
J. of Chem. Soc. of Japan, Ind. Chern. Soc. 57, 103-105 (1954).
378. A. A. Orning, C. H. Schwartz, and J. F. Smith, "Minor Products of
Combustion in Large Coal-Fired Steam Generators," Amer. Sco. of Mech.
Engineers, Paper 64-WA/FU-2 (November 1964).
379.
Methods for the Sampling and Analysis of Flue Gases Part 4: Miscel-
laneous Analysis, British Standards Institution BS 1756: Part 4:
68 pp. (1963).
381. Methods for the Sampling and Analysis of Flue Gases Part 1: Methods
of Sampling, British Standards Institution BS 1756: Part 1: 31 pp.
(1963) .
384. P. Haller, "The Determination of Sulfites, and of Sulfur Dioxide in
Gaseous Mixtures," J. of Soc. of Chern. Ind. 38 (5), 52-56T (1919).
385.
H. Stratman, "Determination of Sulfur Dioxide and Hydrogen Sulfide in
the Atmosphere," Mitteilingen-Vereinigung Der Groskesselbesitzer (23),
396-400 (1953).
392. B. Budesinsky, "Barium Titration of Sulfate with Chlorophosphonazo
III as Indicator," Microchemical J. li, 242-248 (1969).
393.
394.
L. Layton and G. R. Youngquist, "Sorption of Sulfur Dioxide by Ion
Exchange Resins," Ind. & Eng. Chern., Process Design and Dev. 8 (3),
317-324 (July 1969). -
B. Budesinsky and L. Krumlova, "Determination of Sulphur and Sul-
phate by Titration with Barium Perchlorate, II Analytica Chimica Acta
39, 375-381 (1967).
395. B. Budesinsky and D. Brzalova, "Determination of Organic Sulfur and
Sulfates with Dibromosulfonazo III and Dimethylsulfonazo III,"
Chemist-Analyst 55, 110 (1966). .
396.
403.
Carl-Elis Bostrom and C. Brosset, "A Method for Simultaneous Deter-
mination of H2S and S02 in Flue Gases," Atmos. Environ. 3 (4), 407-
416 (July 1969). -
B. P. Knol,"Improvements in Determination of S03 and S02 in Combus-
tion Gases," Rivesta De Combustibili i, 542-545 (l960).
180
WALDEN RESEARCH CORPORATION
-------
LITERATURE CITED (continued)
405. B. Budesinsky, "Modification of Flask Method of Sulfur Determina-
tion -- Detennination of Sulfates with Sulfonazo 111," Analytical
Chern. 37 (9), 1159 (August 1965).
410. H. Konen, "Determination of Sulfur Dioxide and Sulfur Trioxide in
Gases from Sulfur Burners by Means of Iodine," Das Papier 8, 90-91
( 1 954 ) . -
411. Anon., "Standard Methods for Continuous Analysis and Automatic Re-
cording of the Sulfur Dioxide Content of the Atmosphere," American
Society for Testing and Materials Designation: D 1355-60, 330-44
ASTM Standards, Pt. 23 (October 1968).
415. M. Hissink, "An Instrument for Determining Sulfur Oxides in Flue
Gases," J. of the Instit. of Fuel (London) 36, 372-376 (Sept. 1963).
424. W. Reid, Private Communication, Battelle Memorial Institute, Columbus,
Ohio (1969).
425. D. Barnhard, Private Communication, Babcock & Wilcox, Alliance, Ohio
(1969) .
"Analytical Methods for Determination of Total Sulfur Oxides in Flue
Gases, II Japanese Ind. Standard K0103 (1963).
437. M. E. Gales, W. H. Kaylor, and J. E. Longbottom, "Determination of
Sulphate by Automated Colorimetric Analysis," Analyst, London 93,
97-100 (1968).
430.
459.
Editors of Power, Power-Generation Systems, McGraw-Hill, New York
(1967) .
460. L. Schnidman, Editor, Gaseous Fuels, 2nd Ed., Amer. Gas Assoc., New
York (1954).
461. R. L. Chass and R. E. George, J. Air Poll. Cont. Assoc. }Q, 34-43
(1960).
462. F. E. Vandaveer and C. G. Segeler, Ind. Eng. Chern. 37, 816-820 (1945);
see also correction Ind. Eng. Chern. 44, 1833 (1952).
463. J. A. Danielson, Editor, Air Poll. Eng. Manual, USDHEW, PHS, NCAPC,
999-AP-40, Cincinnati (1967).
464. D. H. Barnhart and E. K. Diehl, "Control of Nitrogen Oxides in Boiler
Fl ue Gases by Two-Stage Combus ti on, II JAPCA }Q, 397-406 (1960).
465. G. Gould, "Fonnation of Air Pollutants," Power 104, 86-88 (1960).
181
WALDEN RESEARCH CORPORATION
-------
467.
468.
469.
LITERATURE CITED (continued)
J. D. Sensenbaugh and J. Jonakin, "Effect of COnDustion Conditions
on Nitrogen Oxide Formation in Boiler Furnaces," ASME Paper 60-WA-
334 (1960).
R. P. Hangebrauck, D. J. von Lehmden, and J. E. Meeker, "Emission of
Polynuclear Hydrocarbons and Other Pollutants from Heat-Generation
and Incineration Processes," JAPCA li, 267-278 (1964).
H. C. Austin and W. L. Chadwick, "Control of Air Pollution from Oil-
Burning Power Plants," Mech. Engr. 82,63-66 (1960).
470. J. D. Sensenbaugh, "Air Pollution Problems of Public Utilities,"
Presented at the Fifth Annual Meeting, New England Section APCA,
Bloomfield, Conn. (May 10, 1961).
471. H. C. Austin, "Atmospheric Pollution Problems of the Public Utility
Industry, ". Informative Report No.1, JAPCA lQ., 292-294 (1960).
C. I. Harding, E. R. Hendrickson and B. B. Sundarssan, "Atmospheric
Pollution #28," National Council for Stream Improvement, New York
(1965) .
473. Matheson Gas Data Book (1961).
472.
474.
G. Keppeler, Z. Angew. Chern., £1, 577 (1908).
475. Allied Chern. Corp., "Sulfan" (Stabilized S03) (1968).
476. W. E. Young and A. E. Hershey, Corrosion 11, 725 (1957).
477.
C. Ficai, Giorn. Chim. Ind. Applicata lQ., 199 (1928).
478. E. B. Miller, Chern. Met. Eng. 23, 1155 (1920).
479. W. Latimer, Oxidation Reduction Potentials, 2nd Ed., Prentice Hall,
N. Y. ( 1952 ) .
480. J. R. Ehrenfeld, et al., "Systematic Study of Air Pollution from
Fossil Fuel Combustion Equipment," Final Report, Contract No. CPA
22-69-85, Walden Research Corp. (1970).
481. Handbook of Chemistry and Physics, Chern. Rubber Pub. Co. (1959).
483. ~~u~~~:e~o~~dI~'P;;~:~~ ~~~~~i~: Organic and Inorganic Com-
182
WALDEN RESEARCH CORpORATION
~;:.'z "
-------
lITERATURE CITED (continued)
484. W. Burnside, W. G. Marskell and J. M. Miller, J. Inst. Fuel 29,
261 (1956).
485. A. levy and E. L. Merryman, "Sulfur Chemistry and Its Role in Cor-
rosion and Deposits," Trans. ASME, p. 374 (1965).
488.
D. R. Stull, et al., JANAF Thermochemical Tables, Dow Chern. Co.
(1965-1968).
489. S. Fubui and S. Kanno, presented at the Second International Confer-
ence on Clean Air, london, England (1966).
490. M. A. Waller and N. A. Huey, "Evaluation of a Static Monitor of
Atmospheric Activity of Sulfur Oxides, Nitrogen Dioxide and
Chloride, II APCA Paper #69-90, presented at the 62nd Annual APCA
Meeting (1969).
491. FMC Corp., "Applicability of Inorganic Solids Other than Oxides to
the Development of New Processes for Removal of S02 from Flue Gases,"
NAPCA Contract No. PH 22-68-57 (1969).
492. AVCO Corp., "A Survey of Metal Oxides as Sorbents for Oxides of
Sulfur," NAPCA Contract No. PH 86-67-51 (Feb. 1969).
493. A. D. Thomas, et a1., "Applicability of Metal Oxides to the Develop-
ment of New Processes for Removi ng S02 from Fl ue Gases, II NAPCA Con-
tract No. PH 86-68-68, Tracor Co. (July 1969).
494. TVA, "Sulfur Oxide Removal from Power Plant Stack Gases," Report
prepared for NAPCA (May 1969).
495. P. Muller, "A Contribution to the Problem of the Action of H2S04
on the Dew Poi nt Temperature of Fl ue Gases, II Chern. Eng. Tech. n.,
345 (1959).
496. K. Nelson, Amer. Smelting and Refining Co., Salt lake City, Utah,
Private Communication (1969).
497. L. Broering, et al., "Automotive Mass Emission Analysis by a Vari-
able Dilution Techniquej" Paper No. 67-200 Presented at the 60th
Annual APCA Meeting (1967).
498. J. Martin, Combustion Engineering, Windsor, Conn., Private Communica-
tion (1969).
499.
R. Borgwardt, R. Shigehara, and T. Hartlage, "Detel~mination of S02
from Stacks Using the West Gaeke Method of Analysis," Air Pollution
Training Manual, Dept. of Health, Education and Welfare (1966).
183
WALDEN RESEARCH CORPORATION
-------
LITERATURE CITED (continued)
500.
Chojnowski, "A Multipoint Solid and Gas Sampling Probe, II Inst. Fuel
J. 37, 79 (1964).
502. ASME, "Flue and Exhaust Gas Analysis," Report No. PTC 19.10 (1968).
503. Anon., American Public Health Assoc. Yearbook, p. 92 (1939-1940).
504. Alexander, et al., "Acid Deposition in Oil Fired Boilers," J. Inst.
Fuel (London) 34, 52 (1961).
505. W. S. Dhon, II Se 1 ect i on of Proper Fl ue Gas Probes," ISA Journal!!,
42 (December 1961).
C. F. Goodeve, II Remova 1 of Mi st by Centri fuga 1 Methods, II Trans.
Farad. Soc. 32, 1218 (1936).
507. M. B. Jacobs, liThe Analytical Toxicology of Industrial Inorganic
Poisons," Interscience (1967).
506.
508.
C. V. Kanter, R. G. Lunche and A. P. Fuderich, "Techniques of Test-
ing for Air Contaminants from Combustion Sources, II APCA J. E..' 191
(1967) .
509. J. V. Kerrigan and K. Snajberk, Anal. Chern. 32, 1168 (1960).
510. Battelle Memorial Institute, Applicability of Organic Liquids to
the Development of New Processes for Removing S02 from Flue Gases,"
NAPCA Contract PH 22-68-19 (March 1969).
511. B. P. Knol, (Paper No. 54 of Central Tech. Instit.) Presented at the
Symposium on Fuel Oil Combustion, Sponsored by Esso Res. Ltd. (Nov.
1959) .
512. Wasser, et al., "Effects of Air Fuel Stoichiometry on Air Pollution
from Oi 1 Fi red Furnaces," APCA J. !!!, 332 (1968).
514.
P. Urone, J. B. Evans and C. M. Noyes, Anal. Chern. 37, 1104 (1965).
515. M. D. Thomas and R. E. Amtower, "Gas Dilution Apparatus for Prep.
Reproducible Dynamic Gas Mixture in any Desired Complexity," J.
Air Poll. Control Assoc. ~, 618 (1966).
516. J. N. Driscoll, Walden Research Corp., Unpublished Results (1969).
517.
R. Stevens, A. O'Keefe and G. Ortman, "Absolute Calibration of a
Falem Photometric Detector ...," Env. Science and Tech. 3, 652
(July 1969). -
184
WALDEN RESEARCH CORpORATION
cj;t~.,;""""
-------
LITERATURE CITED (continued)
R. T. Walsh, "Combustion Equipment," in Air Poll. En9. Manual,
J. A. Danielson, Editor, NCAPC, PHS Pub. No. 999-AP-40,
Cincinnati, Ohio (1967).
520. J. Percival, Esso Res. and Eng., Linden, N.J., Private Communica-
ti on (l969).
518.
"Reference Book of Nationwide Emissions, June 1969 Estimates,"
Bureau of Criteria and Standards, Div. of Air Quality and Emission
Data, NAPCA, Durha, N.C.
522. From 1966 Data Cited in "Control Techniques for Sulfur Oxide Air
Pollutants," U.S. DHEW, PHS, NAPCA, Washington, D.C., p. 4 (Jan.
1969).
521.
523.
For a quantitative discussion, see J. R. Hamm, et al., "... Physi-
cal Methods of Separation ...," Final Report to NAPCA, Section 3.3
(Jan. 1969).
524. F. C. Luxl, "Analyzing and Control of OXYgen in Boiler Flue Gas,"
Paper No. 61-WA-340, ASME Annual Meeting (1961).
525. B. D. Bloomfield, in Air Pollution, Vol. II, A. C. Stern, Editor,
Academic Press, N.Y. 11968).
526. Barrett, et al., "S03 Formation in a Noncatalytic Combustor," Pre-
sented at ASME Meeting, Paper No. 65-WA/CD-l (1965).
527. A. W. Berger, et al., "Study of Reactions of Sulfur in Stack Plumes,"
GCA Tech. Rpt. 68-19-G, Contract No. PH-86-67-125 (1969).
528. J. N. Driscgll and P. Warneck, "Primary Process in the Photolysis of
S02 at 1849A," J. Phys. Chern. 72, 3736 (l968).
529. H. F. Johnstone, "An Electrical Method for the Determination of Dew
Point of Flue Gases," Univ. Ill. Eng. Expt. Station Circular 20
(1929).
530. H. Juntgen, "Sulfur Balance and Sulfur Trioxide Equilibrium in Flue
Gases," Erdol Kohle}&., 119 (l963).
531. P. A. Alexander, et al., "An Air Cooled Probe for Measuring Acid
Deposition in Boiler Flue Gases," J. Inst. Fuel 33, 31 (1960).
532. H. H. Ku, Editor, Precision Measurement and Calibration, Vol. I,
Statistical Concepts and Procedures, NBS Special Publication 300,
Washington, D.C. (l969).
533. V. J. Altieri, "Gas Analysis and Testing of Gaseous Materials,"
American Gas Assoc. (1945).
185
WALDEN RESEARCH CORPORATION
-------
LITERATURE CITED (continued)
534. AGA Gas Measurements Manual, Pub. by American Gas Assoc. (1963).
537. P. G. W. Hawksley, et al., "Measurement of Solids in Flue Gases,"
British Coal Utilization Research Association, Leatherhead (1961).
539.
II ASH RAE Guide and Data Book - Systems and Equipment," ASHRAE, N.Y.
(1967) .
540. American Conf. of Govt. Indust. Hygienists,"Air Sampling Instru-
ments for Evaluation of Atmospheric Contaminants," ACGIH, 1014
Broadway, Cincinnati, Ohio (1966).
541. Arthur D. Little, Inc., "Investigation of Sampling Procedure Re-
quirements," AGA, Inc., N.Y. (1957).
542. Manufacturing Chemist Assoc., Inc., "Air Pollution Abatement
Manual ," Washington, D.C. (1952).
543.
545.
547.
548.
549.
Industrial Gas Cleaning Institute, Inc., "Test Procedures for Gas
Scrubbers," Pub. No.1, IGCI, Box 448, Rye, N.Y. 10580 (1964).
Kol toff and Sandell, "Textbook of Quantitative Inorgani c Analysi s, II
The MacMillan Co., p. 617 (1947).
1968 Book of ASTM Standards, Part 23, ASTM, Philadelphia.
R. Larkin, NAPCA, Cincinnati, Ohio, Private Communication (1969).
C. F. Junge and T. G. Ryan, Quart. J. Roy. Meteor. Soc. 84, 46
(1958). --
551. W. S. Smith, "Atmospheric Emission from Fuel Oil Combustion ... ,"
PHS Pub. No. 999-AP-2, Cincinnati, Ohio (1962).
552. W. S. Smith and C. W. Gruber, "Atmospheric Emissions from Coal Com-
bustion - An Inventory Guide," PHS Pub. No. 999-AP-24 (1966).
553. R. Dennis and R. H. Bernstein, "Engineering Study of Removal of
Sulfur Oxides from Stack Gases," Reprint for APC, GCA TR-68-15-G
(1968) .
555. A. Rawdon, Riley Stoker Corp., Worcester, Mass., Private Communica-
t ion (1969).
556. ABMA Reports for 1968.
557.
559.
, Bureau of Census, Dept. of Commerce, Form M34N (1967).
F. E. Gartrell, et al., "Atmospheric Oxidation of S02 in Coal Burn-
ing Power Plants," AIHA J. 24, 173 (1963).
186
WALDEN RESEARCH CORPORATION
-------
LITERATURE CITED (continued)
571. Sherwood and Pigford, Absorption and Extraction, McGraw Hill Book
Co., New York (1952).
606.
E. Ower and R. C. Pankhurst, Measurement of Air Flow, Pergamon
Press (1966).
607. H. H. Haaland, Editor, "Methods for Determination of Velocity, Vol-
ume, Dust and Mist Contents of Gases,1I Bulletin WP-50, 7th Ed.,
Western Precipitation Div., Joy Manufacturing Co., Los Angeles (1968).
608.
609.
R. M. Jamison, V. W. Hanson and O. M. Arnold, Air Eng. I, 26 (1965).
"Determining Dust Concentration in a Gas Stream,1I PTC 27-1957, The
Amer. Soc. of Mech. Engrs., N.Y. (1957).
610.
IIMethods of Testing Fans for General Purposes,1I Part 1, B.S. 848,
British Standards Institution, London (1963).
612. W. K. Lewis and A. H. Radash, Industrial Stoichiometry, McGraw Hill,
N. Y. (1926).
613. Prof. Paul Giever, University of Michigan, Private Communication
(1969) .
614. ASTM 0-22 Subcommittee VI, Tentative Standard Method for Sampling
Stacks (1970).
615.
II Flow Measurement, II B. S. 1042:
London (1951).
1943, British Standards Institution,
616. A. J. Gulley, et a1., II Removal of Acid Gases ...,11 NASA Rpt. No.
CR-1388 (July 1969).
617. S. Fukui, Ei sei Kagaku lQ., 102 (1964).
618. Methods for the Sampling and Analysis of Flue Gases Part 5: Semi-
Routine Analysis, British Standards Institution B.S. 1756: Part 5,
22 pp. (1963).
187
WALDEN RESEARCH CORPORATION
-------
APPENDIX 1
REFERENCE DATA ON THE COMBUSTION EFFLUENT ENVIRONMENT
.~
1. Coal:
2. Coal:
3. Coal:
4 . Oi 1 :
5. Oil:
6. Oi 1 :
7. Gas:
8. Gas:
9: Gas:
Contents
Large
Intermediate
Small
Large
Intermediate
Small
Large
Intermediate
Small
189
WALDEN RESEARCH CORPORATION
-------
SIZE CATEGORIES FOR STATIONARY COMBUSTION SOURCES
Large Intermedi ate Small
BTU Per Hour >5xlOB 3xl05_5xlOB <3xl05
Boiler Horsepower >15,000 10-15,000 <10
Pounds of Steam >500,000 350-500,000 <330
Per Hour
Megawatts >50 <50MW Not Applicable
190
WALDEN RESEARCH CORpORATION
-------
FUEL: COAL
SIZE: LARGE
LOAD (#/hr)
VELOCITY OR
VOLUME
TEMPERATURE (OF)
02 (%)
C02 (%)
CO
.....
\0
.....
S02 }
S03
::2 S
HC
:Ii:
»
r
C
",
Z
:xl
",
en
",
»
:xl
()
:I:
()
o
:xl
"'tI
o
:xl
~
<5
z
PARTICULATE
CYCLE &
TRANSIENTS
REMARKS
SOURCE
.
u
> 11 0 ,000 cfm
275-475°F
<25% XS
0.02#106 BTU(e)
0.5#/T (f)
38S#/T (f)
20#/T (f)
0.007#/106 BTU(e)
0.2#/T (f)
e = estimating factor
f = emission factor
552
100-1400 ppm
0.17-2.5
#/106 BTU
470,471 cited
552
650-1460 ppm
1.1-2.6
#/1 06 BTU
several boiler
designs
469
1,100,000
255-280
6.1-7.1
(41-50% XS)
12.1-12.4
8-25 ppm
1450-2240 ppm
14 ppm
242-403 ppm
10-20 ppm
74
875,000
250-260
5.2-6.0
(34-40% XS)
12. 5- 1 3.9
0-10 ppm
1370-2350 ppm
2-16 ppm
520-664 ppm
2-8 ppm
74
-------
FUEL: COAL
SIZE: LARGE
LOAD (#/hr) 960,000 700,000 830,000 640,000 full load
(75%) (75%) (75%) (partial load)
VELOCITY OR
VOLUME
TEMPERATURE (OF) 279-284 241-264 220-250 240-248
02 (%) 4.6-4.9 4.2-4.5 6.6-7.4 5.8-5.9
(28-30% XS) (27% XS) (45-53% XS) (37-38% XS)
C02 (%) 14. 3- 14.4 14.6-14.8 11 .9- 12.5 12.8-13.3
CO 8-30 ppm 4-5 ppm 0.005-0.44
#/1 06 BTU
-.I S02 620-1470 ppm 600-1300 ppm 1500-1790 ppm 1100-1820 ppm
\b (different fuels)
N
S03 11-13 ppm 10-13 ppm 6-14 ppm 19-22 ppm
NO --
l 0.38-2.5#/106
397 ppm 304-346 ppm 125-217 ppm 386-521 ppm
N02 BTU ~0.28-1.9
-- #/1 0 BTU)
HC 13-14 ppm 6-8 ppm 0.001-10#/106
BTU
:E no load data
»0
r-
1:1
fT1 PARTI CULATE
z
::0
fT1
(J) CYCLE &
fT1
»0 TRANSIENTS
::0
0
:I:
0 REMARKS
0
::0
C3 SOURCE 74 74 74 74 22 ci ted in 552
::0
~
0
z
-------
~
»
r
o
I'TI
Z
::0
I'TI
en
I'TI
»
::0
()
:I:
8
::0
cg
::0
~
o
z
FUEL: COAL
SIZE: LARGE
LOAD (#/hr)
VELOCITY OR
VOLUME
TEMPERATURE (OF)
°2 (%)
C02 (%)
CO
S02
...... S03
\C
tJ.) NO }
N02
HC
PARTI CULATE
CYCLE &
TRANSIENTS
REMARKS
SOURCE
150-000-1,100,000
10-34% XS
0-110 ppm
(1 value @ 780 ppm)
0.1-4.1% of S as S03
650-1420 ppm
200-670 ppm
464
378
-------
FUEL: COAL
SIZE: INTERMEDIATE
LOAD (#/hr) U 149,000#/hr 108 ,OOO#/hr 0.3-147xl06 BTU/hr
(full) (75% load)
VELOCITY OR .211 0,000 CFM
VOLUME
TEMPERATURE (OF) 425-750 305-310 301 -316
02 (%) 25-75% XS 4.5-6.6 6.3-6.7
(25-45% XS) (42-46% XS)
C02 (%) 11 .8-14. 1 12.2-12.5
CO 0.1#/106 BTU(e) <0.1-0.51#/106 BTU
3-50#/T (f)
..... S02 1260-1580 ppm 1530-1840 ppm
\0 38S#/T (f)
~ S03 9 ppm 6-10 ppm
-
NO ......,
r 8-20#/T (f) 329-393 ppm 308-347 ppm 0.30#/106 BTU
N02 .---1
HC 0.05#/106 BTU(e) 0.005-0.1#/106 BTU
1-10#/T (f)
PARTI CULATE
~
r CYCLE &
c
rr1 TRANSIENTS
z
::0
rr1 REMARKS (e) = estimating factor partial load
en
rr1
» (f) = emission factor
::0
(')
:I:
(') SOURCE 552 74 74 468
o
::0
a
:0
~
5
z
-------
" FUEL: COAL
SIZE: SMALL
LOAD (#/hr) U 26,000 BTU/hr 0.006-0.115xl06 BTU
VELOCITY OR <100 CFM
VOLUME
TEMPERATURE (OF) 750-800
02 (%) 75-100% X5
C02 (%)
CO 2#/106 BTU(e) 1. 1-3. 5#/106 BTU
502 400 max with coal
100 ppm max with coke
--' 503 3-4 ppm coal None
\0 6-8 coke
(J1
L 5% of 5 as S03
J
NO 1 0.11-0.36#/106 BTU
3 ppm max
0.0014-0.047#/106 BTU
!
N02 -I
HC 0.5#/106 BTU(e) 0.12-0.73#/106 BTU
PARTI CULATE
~ CY CL E & 2 hr (coal stove) cycle
r-
0
rrJ TRANSIENTS Coke slower (4-5 hrs)
z
:::0
rrJ REMARKS (e) = estimating max SOx early in
(F)
~ factor combustion
:::0
(')
:r
(') SOURCE 552 286 468 460
o
:::0
2J
:::0
~
(5
z
-------
FUEL: OIL 1000 BHP
SIZE: LARGE load factor 85% (usual)
LOAD (#/hr) Usual Range Recormnended Extreme Range 175 MW
VELOCITY OR
VOLUME
TEMPERATURE (OF)
02 (%)
C02 (%)
CO 0.3 ppm 0- 100 ppm
0.005#/1000# 0- 1.7#/1000#
502 (440-520)S ppm 510S ppm (52-520)S ppm
(17-19.9)S#/1000# 19.65#/1000# (2.0-20)S#/1000#
--' 503 6-24 ppm 18 ppm 0-76 ppm
\0
CTI (0.063-0.69)S 0.30S#/1000# (0.063-2.9)S
#/1 000# #/1000#
NO
180-700 ppm 210 ppm .0-1020 ppm 310-915 ppm 330-915 ppm
(f, tangenti a 1)
470 ppm 28#/1000#
(f, hori z.)
NO --
2
~ HC 0.4#/1000# 0-5#/1000#
r
0 PARTICULATE 0.025-0.060 gr/SCF 0.033 gr/5CF 0.005-0.205
I'T1
Z gr/5CF
:0
I'T1 0.82-1.8#/1000# 1 #/1 000# 0.15-6.3
Cfl
I'T1 #/1000#
;'I>
~
:I: CYCLE &
(')
0 TRAN5 I ENTS
:0
a
:0 REMARKS
~
(5
'Z SOURCE 551 551 551 469 466 cited in 551
-------
FUEL: OIL
SIZE: LARGE
LOAD (#/hr) >1000 MW >1000 MW 175MW >1000 BHP
VELOCITY OR
VOLUME
TEMPERATURE (OF)
02 (%)
C02 (%)
CO
S02
S03
--' NO 1
1.0
'i 500-700 ppm 100-900 ppm 310-915 ppm 275-600 p~m
*0.78#/10 BTU
\
1 (average for 554 samples)
NO .J 14.2#/1000# (18,3000 BTU/H)
2
HC
PARTICULATE
~ CYCLE &
,.
r- TRANSIENTS
0
rrI
Z
:0 REMARKS * good
rrI
(f)
~ SOURCE 470 ci ted ; n 551 471 465 c; ted ; n 551 198 cited in 551
:0
(")
:r
(")
0
:0
~
:0
~
0
2
-------
FUEL: OIL
SIZE: LARGE
LOAD (#/hr)
VELOCITY OR
VOLUME
TEMPERATURE (OF)
02 (%)
C02 (%)
CO
S02
--'
\D
ex>
S03
NO
--,
N02 -
HC
PARTI CULATE
~
,...
c
",
Z
:0
",
en
",
~
:0
(')
::I:
8
:0
~
:0
~
'6
CYCLE &
TRANSIENTS
REMARKS
SOURCE
90%S-?S02
1-5% S02-?S03
NO f(XS air)
NO~f(load)
100-900 ppm
0.02-0.04
gr/SCF
551
cited in 551
1 00%S-?S02
1 - 2%S-?SO
*2200 pp~ S02
0.14#/1000#
* 4% S/oi1
471
1-5%-+503
160-699 ppm
465
467
-------
FUEL: OIL
SIZE: LARGE
LOAD (#/hr) 5xl08 BTU/hr
VELOCITY OR
VOLUME
TEMPERATURE (OF)
02 (%)
C02 (%)
CO
S02
S03
-'
\0 NO I
\0
~ 230#/hr 320-920 ppm
N02 J
HC
PARTI CULATE
CYCLE &
TRANSIENTS
~ REMARKS
~
r
1:1
ITI 463 464
z SOURCE
:;0
ITI
CII
ITI
~
:;0
(')
:I:
(')
0
:;0
-0
0
:;0
~
(5
z
-------
FUEL: OIL
5IZE: INTERMEDIATE
LOAD (#/hr) <1000 BHP <1000 BHP <1000 BHP
Recommended Usua 1 Range Extreme Range
VELOCITY OR
VOLUME
TEMPERATURE (OF)
02 (%)
C02 (%)
CO 1 5 P pm 0-120 ppm 0-11 00 ppm
0.25#/1000# 0-2#/1 000# 0- 194#/ 1000#
502 (510)5 ppm (0-520)5 ppm (365-520)5 ppm
(19.6)5#/1 000# (0-20) 5#/1 000# (14-20 )5#/1 000#
N
C> 503 (0-6.5)5 ppm (0-6.8)5 ppm
C> 5.25 ppm
0.255#/1000# (0.31 )5#/1 000# (0-3.4)5#/1000#
NO } 320 ppm 0.140 ppm 0-630 ppm 47 ppm 0.065-230#/hr
9#/1 000# 0-4#/1000# 0-18#/1000# 0.33#/106BTU
N02
HC
PARTICULATE 0.049 gr/5CF 0.033-0.13 0-0.33 gr/5CF
:E gr/SCf
»
r- 1. 5#/1000# 1-4#/1000# 0-10#/1000#
0
rrt
Z
~ CYCLE &
rrt TRAN5IENTS
en
rrt
»
~
(") REMARKS
:r
8 SOURCE 551 551 551 93 463
~
~
::0
~
'6
z
-------
FUEL: OIL
SIZE: INTERMEDIATE
LOAD (#/hr) <20xl06 BTU/hr 1100#steam/hr 60 HP 200 HP 350 HP
VELOCITY OR
VOLUME
TEMPERATURE (oF)
02 (%) 65% XS 210% XS 94% XS
C02 (%)
CO 0.01% 0.02 0
S02 355 ppm 11 ppm 1 7 ppm
S03 1.6 ppm 5.6 ppm 0
N
0
--' NO
} usually not 300 ppm 100% load 47 ppm 21 ppm 72 ppm
over 100 ppm 50%
25%
N02
HC
PARTICULATE 0.069 gr/SCF 0.14 gr/SCF 0.014 gr/SCF
::E CYCLE &
~ TRANSIENTS
r
0
ITI
:z REMARKS 1 %S ! #2 O.97%S!#2
::tJ O.42S!#2
ITI
CIJ
~ SOURCE 463 464 461
(")
:J:
(")
0
::tJ
~
::0
~
(5
:z
-------
FUEL: OIL
SIZE: INTERMEDIATE
lOAD (#/hr) 100 HP 200 245 120 125 245
VELOCITY OR
VOLUME
TEMPERATURE (OF)
02 (%) 290% XS 370% XS 115% XS 68% XS 180% XS 43% XS
C02 (%)
CO 0 0.002 0.002 0.003 0 0
S02 98 ppm neg. 1 02 ppm 414 ppm 264 ppm 397 ppm
N S03 1 .4 ppm 0 0.5 ppm 4.7 ppm 3.2 ppm 0.4
a
N
NO i
ppm 36 55 33 368 128 387
N02
HC
PARTI CUlATE
gr/SCF 0.071 0.10 0.041 0.074 0.11 0.064
CYCLE &
~ TRANSIENTS
r-
c O. 71 %S , #2 0.55%S,#2 0.21%S,#2 1.0%S,#6
ITI REMARKS 1. 78%S ,#6 0.44%S,#6
z
;;0
ITI SOURCE 461 461 461
C/J
ITI
>
;;0
(")
:I:
(")
0
;;0
(g
::0
!i
Q
-------
FUEL: OIL
SIZE: INTERMEDIATE:
LOAD (#/hr) 425 HP 460 HP 500 HP 21x106 BTU/hr 200 HP 225 HP 257 HP 350 HP
VELOCITY OR L. ~
VOLUME
TEMPERATURE (oF) 486 avo
02 (%) 11 0% XS 107% XS 92% XS 2.7-10.3
C02 (%) 10.3%
CO (%) O. 001 0 0 N-40 ppm
S02 ppm 700 362 594 160-997
N S03 ppm 6.7 2.2 3.6
0
w NO ppm } 61-127 ppm
275 199 256 0.31#/106 BTU
N02 ppm 9.8-37 ppm
HC
PARTICULATE
gr/SCF 0.28 0.039 0.045 12.5#/1000 gal
CYCLE &
~ TRANSIENTS
,
c 0.06%S, 0.078%S, 1 . 39%S ,
ITI REMARKS
:z
::c #6 #6 #6
fT1
fI)
~ SOURCE 461 459
::c
(')
I
(')
0
::c
~
::c
~
0
:z
-------
FUEL: OIL
SIZE: SMALL
LOAD (#/hr)
VELOCITY OR
VOLUME
TEMPERATURE (OF)
02 (%)
C02 (%)
CO
N S02
0
.j::>.
S03
NO J
~
N02
HC
~ PARTICULATE
r-
0
1"1
Z
:;0. CYCLE &
1"1 TRANSIENTS
en
1"1
,.
:;0 REMARKS
(')
::I:
(')
a SOURCE
:;0
cg
:;ID
~
C5
350,000 BTU/hr
negligible
.065#/hr
463
1 gph
5.5-12.8%
(0-150% XS)
5.9-9.6%
0.86-96.6 gm/Kg
10,000 ppm @ 0% XS Air extreme
1,000 ppm @ 25% XS (Best)
1 .46-1 .96 gm/ Kg
<0.02 gm/Kg
<0.05 ppm
85-92% of NO
x
1.08-2.41 gm/Kg
40-52 ppm, 46@ 25% XS air (equi1ib. should be 150-650 ppm)
25 gm/Kg = 2700 ppm HC extreme
300 ppm @ 25% XS to zero @ 175% XS
o . 04 - 3 3 . 40 gm/ Kg
On for 11.4 min
Every 34 mi n .
0.098%, #2
512
-------
FUEL: OIL
SIZE: SMALL
LOAD (#/hr) 0.17xl06 BTU/hr 6 5000 K ca 1 /hr
O. 18xl 0
VELOCITY OR
VOLUME
TEMPERATURE (OF)
02 (%)
C02 (%)
CO CO is high when soot 2600 ppm on
high 1 i ghti ng
200-1000 after
1 0 mi n
S02 3
N 200-800 mg/m STP
0
<.n 3
S03 30 mg/m @ max load 0.75-1.75%
SO +S03
0-3% variation shown vs XS air ab6ut 20% of S
in soot
NO 1 0.44#106 BTU 0.1#106 BTU
N02 ..J
:E HC
»0 8.6 mg/~3+6.35
r PARTICULATE
0
I'T1
z 13 mg/m +7.65
::0
I'T1
en CYCLE &
I'T1
»0 TRANSIENTS
::0
0
::r
0 REMARKS vaporizing burner atomi zi ng
o
::0
(3 SOURCE 459 459 55 55
::0
~
0
z
-------
FUEL: OIL
SIZE: SMALL
LOAD (#/hr) 0.66-2.5 gph
VELOCITY OR
VOLUME
TEMPERATURE (OF) 362-1025
682 average
02 (%)
C02 (%)
CO N-1500 ppm
S02 23.3-194 ppm
N S03
a
Q)
NO 3.9-76 ppm
N02 N-26.4
HC
PARTICULATE
CYCLE &
TRANSIENTS
~ REMARKS
>
r
0
JTI SOURCE
z
::0
JTI
CII
~
::0
(1
:r
(1
0
::0
;g
::0
~
(5
z
-------
FUEL: GAS
SIZE: LARGE
LOAD (#/hr)
VELOCITY OR
VOLUME
TEMPERATURE (0 F)
N
o
~
02 (%)
C02 (%)
CO
S02
S03
NO
")
'r
\
N02 -'
HC
PARTICULATE
~
»
r
o
,.,
2
::0
,.,
en
,.,
»
::0
(")
::x:
(")
o
::0
~
::0
~
o
2
CYCLE &
TRANSIENTS
REMARKS
SOURCE
190-1350 ppm
100#/hr @ 5x108 BTU/hr
180 ppm-1100 ppm
469
463
464 cited in 463
-------
FUEL: GAS
SIZE: INTERMEDIATE
LOAD (#/hr) 3x105_5x108 BTU/hr 107 BTU/ hr 6
9.3xlO
VELOCITY OR
VOLUME
TEMPERATURE (OF)
02 (%)
C02 (%)
CO
S02
S03
N
a NO -
co
8.47 ppm
0.06#106 BTU 0.14#/106 BTU
NO - 0.03#/hr-100#/hr 90 ppm
2 0.16#/106 BTU
HC
PARTICULATE
CYCLE &
~ TRANSIENTS
~
roo
0
fTI REMARKS
:z
::0
rn SOURCE 198 463 462 459
~
::u
(')
:r
(')
0
::u
~
:;0
~
(5
:z
-------
FUEL: GAS
SIZE: SMALL
LOAD (#/hr) 350,000 BTU/hr 5 0.17xl06 BTU/hr
0.9xlO BTU/hr
VELOCITY OR
VOLUME
TEMPERATURE (OF)
02 (%)
C02 (%)
CO
502
503 negligible in none
N small equipment
0
\D
NO I 34-62 ppm
'r (50 average) 0.09#106 BTU
\ 0.093#/1 0 BTU) 0.03#/hr 50 ppm
N02 J
HC
PARTICULATE
.:E CYCLE &
~ TRANSIENTS
r
0
fI'1
Z REMARKS
:u
fI'1
(f) SOURCE 93 463 462
~ 460 459
::0
(")
:r
(")
0
::0
~
::0
~
(5
z
-------
APPENDIX 2
LITERATURE SEARCH
A comprehensive search of the literature was conducted to learn
of all methodology which might be applicable to this program. This in-
cludes methods currently in use, those used in the past which seemed ap-
propriate, and information from other areas which seemed relevant.
In order to perform this study we made a thorough search of the
literature both past and present. Thus, we covered the journals, ab-
stracts, indices and bibliographies shown in Tables 2-1 and 2-2. By this
approach we feel we covered all journals in the western world and many
others which might have any pertinence to this problem. To supplement
this, we requested and obtained a literature search by APTIC on sampling
and analytical methods for the pollutants of interest.
As articles of interest were uncovered by the literature re-
search staff, the papers were studied by analytical chemists and signifi-
cant information abstracted for further evaluation. As the study progres-
sed, group discussions were held with chemists with relevant experience
in order to both evaluate existing methods and seek fresh alternatives.
211
WALDEN RESEARCH CORPORATION
-------
TABLE A2-1
PRIMARY LITERATURE SOURCES UTILIZED IN THIS PROGRAM
Air Pollution Control Abstracts
1956-1969
Air Pollution Titles
Thru 1969
American Petroleum Institute
Proceedi ngs
1962-1969
1954-1969
Analytical Abstracts
Applied Science and Technology
Index
Chemical Abstracts
1939-1969
1907-1968
Chemical Titles
Thru 1969
Engineering Index
Fuel Abstracts and Current Titles
1939-1969
1960-1969
SAE Progress in Technology Series
1961-1967
212
. WALDEN RESEARCH CORPORATION
-------
TABLE A2-2
BIBLIOGRAPHIES
Air Pollution Publications: A Selected Bibliography
(DHEW) 1963-1966
DDC Bibliography on Air and Water Pollution
(1968) AD 679210
Research into Sampling, Analysis and Monitoring of
Gaseous Pollutant Emissions from Stacks
Literature Searches I and II (Engineering
Found) 1963, 1965
Sulfur Oxides and Other Sulfur Compounds
A. G. Cooper, DHEW 1965
213
WALDEN RESEARCH CORPORATION
-------
APPENDIX 3
DISCUSSION AND ADDITIONAL STUDIES OF THE BARIUM
CHLORANILATE METHOD FOR SULFATE
A.
DISCUSSION
1. Bertolacini and Barney (164) developed a method which con-
sists of adding 0.3 g of barium chloranilate to a 50% ethanolic solu-
tion containing sulfate ion. The solution is buffered to pH 4 due to
absorbance dependence on pH. Measurement is done colorimetrically at
530 nm.
a. Range is reported to be 2-400 ppm sulfate (2 ppm
using 5 cm cells).
b. Accuracy and precision is given as approximately 1%.
c.
exchange column.
Cationic interference is eliminated by use of an ion
Anion interference said to be negligible.
d. Time of mixing after addition of barium chloranilate
was 15 minutes. Absorbance increased an additional 5% in 24 hours.
Total time for analysis is 20 minutes.
2. Bertolacini and Barney (311) have done further work on
the use of barium chloranilate in the determination of sulfate ion
and found that acid chloranilate solutions have about 30 times greater
absorption at 332 nm than at 530 nm. As in paper (1), solutions were
put through cation exchange columns, buffered to pH 4 in 50% alcoholic
solution and 0.2 g barium chloranilate added. Solution shaken for 15
minutes, then filtered. Filtrate absorbances were read at 332 nm in
1 cm cells against a reagent blank.
215
WALDEN RESEARCH CORPORATION
-------
a.
Range - Down to 0.06 ppm in 1 cm cell, upper limit
not given.
b. Accuracy and Precision - Standard deviation of 0.2
ppm over a range of 0.98 to 3.86 ppm standards. (5-20% relative)
c. Interferences - Dish washing
nitrate ion produces an absorbance equivalent
fite does not interfere.
detergents. 2500 ppm
to 0.3 ppm sulfate. Sul-
d.
Time - About 30 minutes.
3. Carlson, Rosell and Vallejos (273) have made further
modifications to the Bertolacini and Barney procedures by making use
of solution equilibrium at varying pHis. They, therefore, add the
buffer (a mixture of phosphoric acid and potassium dihydrogen phos-
phate) after filtering the ch1orani1ate solution. Measurements were
done at 530 nm. The calibration curve is non linear.
a.
Range not reported.
b.
of known amount.
Accuracy and Precision - Accuracy is within 3%
c.
Interferences - Removed by ion exchange column.
1 hour.
d. Time - Sample and barium ch1orani1ate mixed for
Total time is thus about 1 hour, 15 minutes.
4. Klipp and Barney (361) have modified the Bertolacini
and Barney procedure by changing the buffer from potassium acid
phthalate to a 0.02 M solution of sodium acetate-acetic acid. Buf-
fer capacity has thus been increased to eliminate post filtration re-
adjustment of pH.
216
WALDEN RESEARCH CORPORATION
-------
a. Range - Sulfur in naphthas determined over range of
1-400 ppm 0-40 ppm calibration curve run at 330 nm, 40-400 ppm cali-
bration curve run at 530 nm.
b. Accuracy and Precision - Standard deviation
3 ppm at 150 ppm sulfur level, 1 ppm at 25 ppm sulfur level.
relative)
of about
(2-4%
c.
Interferences - Removed by ion exchange.
d.
Time - About 20 minutes (after oxidizing sulfur to
sulfate).
5. A paper by Kanno et al (360) recommends a pH change from
4.0 to 5.2 to give more reliable results using the Bertolacini and Barney
procedure.
6. Laxton and Jackson (153) automated the Bertolacini and
Barney procedure. A silica wool probe filter was used to screen out
cationic interferences. No buffer was used. The sample is collected
in 80% isopropanol, and the sulfuric acid-isopropanol solution is
passed through a 5-10 mm bed of barium chloranilate supported on a
sintered glass filter. A second finer filter removes precipitated
barium sulfate and fine particles of barium chloranilate. The filtered
solution flows into a cell for measurement.
a. Range - Beer's Law plot up to 100 ~g/ml 503 has only
slight curvature at lower end.
b. Color Development - 94% of absorbance resulting from
12 hour solution-reagent contact is achieved in the time needed for
solution to reach the cell.
c. Interferences - Dissolved 502 at concentrations of 80
~g/ml caused less than 0.1% change in solution color.
d.
Time - About 5 minutes.
217
WALDEN RESEARCH CORPORATION
-------
B.
ADDITIONAL LABORATORY STUDIES
1.
Introduction
In Section 5.5.4.2 of this report, the principal laboratory
investigations of the barium chloranilate system at the adopted pH 5.6
were described. In this section we report on additional work conducted
on this method, principally of a preliminary or auxiliary nature.
A study was conducted to evaluate some variations in the
barium chloranilate colorimetric method for sulfate, specifically buffer-
ing systems given in the literature (164,311,153,271,273) to determine:
a.
adherence to Beer's law
sensitivity
time and amount of mixing
absorbance measurement
needed prior to
b.
c.
Work on the acetate-acetic acid buffer system (pH 4.6)
showed adherence to Beer's law up to at least 250 ~g SO~/ml of solution.
The sensitivity was 0.002 absorbance units per ~g. Initial measurements
revealed that the curve deviates from Beer's law between 250 ~g/ml and
500 ~g/ml concentration. At 0.002 absorbance units/~g S04/ml solution,
the method is sensitive enough to determine 8 mg/M3 of S03 in flue gas
(assuming no interferences).
Preliminary work on the phosphate-phosphoric acid buffer
system (pH 1.8) indicated that the system may have greater sensitivity
than the acetate-acetic acid system, but may not be as useful since the
operable range would be limited. A single run on the non-buffered sys-
tem using 80% isopropanol solvent indicated a decrease in sensitivity
compared to the acetate-acetic acid system.
2.
pH Dependence
Three of the literature methods were run for comparison pur-
poses. The methods were those conducted at pH 1.8 (273), 4.0 (164), and
4.6 (361). Results indicate little difference between the pH 4.0 and 4.6
218
WALDEN RESEARCH CORPDF
-------
systems. The pH 1.8 system appears to be much more sensitive than either
the 4.0 or the 4.6 system, e.g., 250 ~g SO;/ml gives the following ab-
sorbancies vs water at 530 nm after 15 minutes mixing.
Absorbance Absorbance
1ili Corrected for Blank of Blank
1.8 1.64 o. 107
4.0 0.520 0.013
4.6 0.500 0.013
Since the applicable dynamic concentration range would be
more limited with the pH 1.8 system due to the steepness of the slope
(0.006 absorbance units/~g), we have ruled out the pH 1.8 system.
A comparison of data for the pH 4.0 and 4.6 systems is as
follows: Solutions were shaken (hand shaking 10 shakes every 2 minutes)
for 15 minutes followed by five minutes centrifuging to separated unre-
acted barium chloranilate and precipitated barium sulfate from the cen-
trifugate. (An International Clinical Model centrifuge was used - 6-
place head, 15 ml tubes at highest speed setting.) The absorbance was
measured vs a water blank on a Beckman DU spectrophotometer in 1 cm cells
at 530 nm. All water used for solution preparation was percolated through
a bed of Amberlite MB-3 ion exchange resin to remove sulfate ion from the
water. The calibration data obtained for both pH ranges are given below:
Concentration
~
o
50
125
250
500
Absorbance
pH 4.0 pH 4.6
0.013
0.127
0.281
0.533
0.778
0.013
0.116
0.267
0.513
0.752
3.
Removal of Precipitate Prior to Color Readout
Solutions were initially filtered through Whatman No. 42
paper after reaction with the barium chloranilate. However, this is
219
WALDEN RESEARCH CORPORATION
-------
quite slow and the possibility of concentration changes while filtering
suggested we find an alternative. A comparison of filtration vs centrifu-
gation on an International Clinical Model showed a difference of 0.005
absorbance units between filtered and centrifuged solutions at the 0.500
absorbance level (or a 1% difference on the samples measured). Since
centrifugation was faster and gave comparable results, we have since fol-
lowed the practice of centrifuging rather than filtering solutions.
220
WALDEN RESEARCH CORpORA
-------
APPENDIX 4
SAMPLING SYSTEM FOR SO AND NO
x x
Equipment for sampling SO (and NO ) gases from combustion
x x
flue gases has been designed. The objectives of the design were first
to provide high precision and secondly to make the equipment convenient
to transport, install and operate at field locations. The system is
comprised of the following three modules:
1.
2.
3.
dual probe module
SOx and NOx collection module
control module
The above modules constitute a complete system for sampling
for S03' S02 and NOx in the flue gas.
Probe Module - A schematic of the probe module is shown in
Figure A4-1. The features of this module are:
1.
dual probes, a 1/4 inch diameter stainless steel probe
for NOx mechanically coupled to 11 mm pyrex probe for SOx.
pyrex probe electrically heated to prevent condensation
of S03 in the sampling line.
2.
3.
stack adapter assembly that allows various probe in-
sertion depths.
filter in the probe to remove particulates from the SOx
sample gas.
4.
5. quick attachment tees for NOx grab samplers.
6. pyrex socket joint for connection to collector module.
221
WALDEN RESEARCH CORPORATION
-------
N
N
N
1/4" Diameter
Stainless Steel Probe
for NOx
Heating
Wire
Pyrex Socket Joint
11 ter
~.
11 rnm Dia.
Pyrex
1 inch
Stainless
Tube
6'
~
,...
C
1"1
Z
::a
1"1
en
~
::a
(')
:I:
8
::a
~
;;0
PROBE MODULE
Figure A4-l
-------
SOX Collector Module - The collector module depicted in Figure
A4-2 has two principal parts, the S03 condenser and the S02 impinger
train. The S03 condenser is a pyrex coil with a glass frit filter at-
tached to the downstream end. This assembly with an electric heating
coil and thermostat is encased in a two inch diameter copper water
jacket. The thermostat will maintain the water temperature at l400F.
The upstream end of the condenser assembly terminates in a pyrex ball
joint which mates with the socket joint on the probe module. The down-
stream end of the condenser assembly also terminates in a pyrex ball
joint which mates with a socket joint on the S02 impinger train assembly.
The S02 impinger train assembly consists of two midget im-
pingers modified with ball and socket joint connectors.
The SO collection module is mechanically connected to and
x
supported by the probe through a right-angle bracket. The S03 con-
denser assembly and S02 impinger train are separately attached to the
right-angle bracket.
The SO collection module is pneumatically connected to the
x
control module through rubber vacuum hose.
Control Module - The control module is shown in Figure A4-3
The principal components of this module are:
1.
2.
3.
4.
vacuum pump
critical orifice meter
SO probe heater power supply
x
power control panel
The vacuum pump is a Gast Model 1531. This pump is capable of
achieving a static vacuum of 26" Hg, and maintaining a pressure drop suf-
ficient for critical flow across a 3i/min critical orifice when connected
to the sampling system.
223
WALDEN RESEARCH CORPORATION
-------
N
N
~
~
r
o
IT!
2
::0
IT!
(J)
IT!
>
::0
(")
::J:
(")
o
::0
~
::0
>.
Probe
Module
=- CID~ ~
- - - -.
-....
--1' :;. ... ,
I, ... ,
-~ ...
Right-Angle
Bracket
- - - -
Heater Coil
I
I I
I "
I I
--~
Thermostat
- - - - - -
I' .... - ... I ,- ....
'1" I" "
"1/, " I ,,/--
. I ... ... 1 I " / -
-' ... - , , -' , - ,
- - --
- - - -
I I
@
I I
------ -
---------
- - -
- - -
-----
- - -
Condenser
Coil
COLLECTOR MODULE
Fi qure A4-2
- - - -
- - -
----..-"'
I I
~
I I
I
I
.-r
()Ol I Ii,
~ --< "LU~.
ili'~' - 1
I
-------
~
o
r>1
Z
:;v
r>1
CJ)
~
:;v
(')
::J:
(')
o
:;v
~
:;v
~
o
z
N
N
(J1
Uti 1 ity
Space
Variable
Transformer
(i]1
[j]'
Power Panel
Utility
Space
Critical
Orifice
Figure A4-3. Control Module (front panel not shown).
Vacuum
Pump
Vacuum Gauges
Hose Connector
Mi 11 i pore Fi lter
Thermometer
-------
The critical orifice meters (Millipore) are available as stock
items in the sampling rates of interest for this system (0.5, 1.0 and
3.0 /min). Vacuum gauges are provided upstream and downstream of the
critical orifice in order to monitor the critical flow condition, i.e.,
~P > 7.611 Hg. A thermometer is placed in the sampling line just up-
stream of the critical orifice. The temperature from this thermometer
and the pressure from the upstream pressure gauge may be used to convert
the sample gas volume to a volume at other conditions. A Millipore
filter is placed .in the sample line just upstream of the critical orifice
to protect the orifice against plugging.
A variable transformer provides the heater power supply for the
50 probe. During test of the sampling systems, this transformer is ad-
x
justed to maintain the 50 probe temperature above the condensation temp-
x
erature of 503' The control will be locked to prevent misadjustment in
field operations.
The pump motor, the variable transformer and the power line to
the 503 condense~ are plugged into the control panel and controlled by
the separate switches.
5ystem Operation - The probe module is fitted to a stack or
flue and the sampling lines connected as previously described. Power
cords are connected between the 50 probe heater and the variable trans-
x
former and between the 503 condenser and the power control panel. Water
is added to the 503 condenser jacket and peroxide solution added to the
502 impingers. The probe heater and 503 condenser are switched on at
the control panel and time allowed for them to come to operating temp-
erature. Then, the pump is started from the control panel switch and
the time recorded. During the sampling period, the operator checks ~P
across the critical orifice and records the pressure and temperature
values. At this time, he may collect NO grab samples in an evacuated
x
2~ flask through the NOx probe. At the end of the sampling period, the
226
VVALDEN RESEARCH CORPoR~
-------
pump is switched off and the time recorded. The SO sample is calcu-
x
lated from the time interval and the flow rate value of the critical
orifice. The S03 collected in the condenser and the S02 collected in
the impinger are transferred to sample bottles and the system may be
recycled for replicate samples.
227
WALDEN RESEARCH CORPORATION
~.
-------
APPENDIX 5
LABORATORY DILUTION SYSTEMS
A.
S02
A simple dilution system for producing high concentrations (>200
ppm) of S02' NO (and N02) was constructed (Figure A5-1).
The dilution system contains a pump for delivery of dilution air;
calibrated flow meters for air, S02' NO and N02; and a 5-liter mixing ves-
sel.
All tubing used in the original equipment was either glass or
viton. Air flow through the system is 10 to 15 liters/min. After mix-
ing, a small fraction of the total (0.5 to 5 liters/min) is withdrawn
from the exhaust.
A series of blank runs following high concentration S02 absorp-
tion efficiency tests showed significant S02 desorption was occurring.
Since the dilution system was constructed of glass and viton tubing, we
concluded that significant desorption of S02 from viton was occurring.
The viton tubing (Figure A5-2) in the collection system was replaced with
glass tubing. This eliminated the S02 background. All test results were
obtained with this system.
B. S03 DILUTION SYSTEM
The S03 dilution system constructed is considerably different
from the S02 dilution system because of the high reactivity and low
volatility of S03 and the addition of water vapor injection to simulate
a combustion effluent.
The system constructed was all glass (Figure A5-3). Dry air
passes through a flow meter into a thermostated flask containing solid
S03* (vapor pressure at room temperature is about 0.1 atm). The dry
*
Obtained as stabilized liquid (Sulfan) from Allied Chemical Corp.
229
WALDEN RESEARCH CORPORATION
-------
."
......
1.0
C
,
CD
»
01
I
......
oU')
............
3
U')"O :
0......
N CD
t:Uc..
::s ...... ".
c........
c
:z c-t
0""'.
0
::s
N VI
W «
0 VI
cT
CD
3
......
0
,
::T
......
1.0
::T Air
n
o
::s
n
CD
::s
c-t
~ '
t:U
r c-t
0 ......
,., 0
z ::s
XI VI
,.,
en
,.,
>
XI
n
:r
n
0
XI
'"tI
C
X1
....
Viton tubing
, To exhaust
Viton tubing
5L mixing
vessel
j
'"
to vacuum pump
midget impingers
containing peroxide
solution for 502
-------
Flow
Meter
N
W
. --'
(stainless steel tee)
Vtton O-ring seals
. - (expansi on
coil)
=:
>
r
o
",
Z
::0
",
CJ)
",
>
::0
(')
:r
(')
o
::0
-V"
o
::0
»
-I
o
Z
-
Figure A5-2.
e" aus t
baH' &&
socket. joi n t
" Vi ton
mi dget i mpi nger
5 liter flask
Dilution System.
-------
SU3 Carrier
Flowmeter
Dryi ng
Tube
N
W
N
Room Air
Source
Exhaust Flow
Straight Glass Mixing TU~
S03 -Sample
Fl ask
Cooling
Bath
Impi ngers half-
filled with Water
A A. /
/ "V "V~ -
Flow
Heating Tape
Figure A5-3. S03 Delivery System
Sampling Flow
Collection Device
Millipore Filter
Critical Orifice
Vacuum
Pump
-------
S03 air mixture leaving the flask is mixed downstream with moist air
saturated with water at a known temperature. The entire flow system,
with the exception of the S03 vessel, is electrically heated to prevent
condensation of H2S04 or H20.
Since both the S03 and moist air supplies are metered prior to
treatment, the ambient temperature calibration factors may be used to
establish the flow rates. S03 concentrations may be established by con-
trol of temperature and/or dilution ratios. The water vapor concentra-
tion of the output stream was determined using weighed CaC12 absorption
tubes and/or impingers cooled in a salt-ice bath. The moisture content
was in the range 8 to 12% (in good agreement with the desired values for
a simulated combustion effluent).
The delivery system for S03 was originally constructed with a
spiral mixing section. This was replaced by a straight run of electric-
ally heated glass tubing to obtain improved heat transfer and eliminate
a severe condensation problem.
Difficulties in obtaining low concentrations (50 to 150 ppm S03)
were solved by crystallizing the (a) solid phase from the stabilized liquid
(Sulfan). This reduced the vapor pressure by an order of magnitude. The
503 concentration is, however, still not well-controlled in this apparatus.
Even with the reservoir thermostated, run to run variations by a factor of
10 were not uncommon.
233
WALDEN RESEARCH CORPORATION
-------
APPENDIX 6
BIBLIOGRAPHY FOR HYDROGEN SULFIDE
At the initial outset of our work, we were required to perform a
literature search on existing analytical methods applicable to the de-
termination of hydrogen sulfide (H2S) in combustion effluents. However,
this task was changed early in the program in favor of emphasis on car-
bon monoxide.
This report is intended to cover the limited amount of work that
was spent on H2S. Thus, it is not intended in any sense to be complete -
either with regards to methodology or the bibliography.
The methodology for H2S can be grouped into a number of categories:
colorimetric, titrimetric, and turbidimetric. H2S undergoes a variety of
chemical reactions (oxidation-reduction, acid-base, complexation, etc.)
and all of these have been employed in developing the methods currently
available. Based on our brief exposure, it appears that a number of
these methods have sufficient sensitivity and precision for use in
analysis of fossil fuel combustion products. However, none are truly
specific for H2S so that accuracy would be questionable.
On the following pages are presented the literature which was found
for the determination of H2S. It is divided into sub-categories by method
as well as being presented at the end as a bibliography by author.
235
WALDEN RESEARCH CORPORATION
-------
~VIEWJ EVALUATION AND DISCUSSION PAPERS
The following papers provide reviews of the procedures for H2S.
RPl
Risenfie1d, F.C. and Orbach, H.K. "Methods for Determining Hydrogen Sulfide
in Gases". Petroleum Engineer, 25(6) (1953) C32-8. (CA 49: 12197, 1955).
RP2
Brychta, Miros1av and Strobl, Jiii. "Hydrogen Sulfide Determination in
Illuminating Gas". Paliva, 32 (1952)p. 113-116. (CA 50: 8173, 1956).
RP3
Jacobs, M.B. The Chemical Analysis of Air Pollutants, p. 182-194.
Interscience Publishers, Inc. N.Y. 1960.
RP4
Smith, A.F., Jenkins, D.G. and Cunningworth, D.E. "Measurement of Trace
Quantities of Hydrogen Sulphide in Industrial Atmospheres". Journal of
Applied Chemistry, 11 (Sept. 1961) p. 317-29.
RP5
Lahmann, Erdwin. "Measurement of Gaseous Sulfur Compounds .in the Atmosphere".
Erdoe1 und Koh1e, 18(10) (1965) p. 796-800. (CA 64: 1966).
RP6
Jensen, George A., Adams, Donald F. and Stem, Harry. "Absorption of H2S and
Methyl Mercaptan from Dilute Gas Mixtures". Journal of Air Pollution
Control Association, 16(5) (1966) p. 248-53. (CA 65: 1966).
RP7
Bamesber3er, W.L. and Adams, D.F. "Improvements in the Collection of Hydrogen
Sulfide in Cadmium Hydroxide Suspension". Environmental Science and
Technology, 3(3) (March 1969) p. 258-61.
236
-------
MBl
MB2
MB3
MB4
MBS
MB6
,',. .
PROCEDURES BASED ON METHYLENE BLUE REACTION
One of the more sensitive and selective procedures for H2S is based on trapping
in zinc acetate solution and reaction of the zinc sulfide with ferric ion, and
p-amino dimethyl aniline in acid solution. The resultant product is blue and
can be measured spectrophotometrically. Of the references listed below, the
paper by Gustafsson (MB5) provides a good discussion of the color reaction,
while the others are oriented towards a specific application.
Kosior, K.E.A. "Determination of Hydrogen Sulfide in Natural Gas".
Canadian Chemist~ and Process Industry, 32 (1948) p. 925-6, 929.
(CA 43: 835, 1949 .
Fogo, James K. and Popowsky, Milton. "Spectrophotometric Determination of
Hydrogen Sulfide-Methylene Blue Method". Analytical Chemistry, 21
(1949) p. 732-4. (CA 43: 7375, 1949).
Sands, A.E., Grafius, M.A., Wainwright, H.W., and Wilson, M.W. "The
Determination of Low Concentrations of Hydrogen Sulfide in Gas by the
Methylene Blue Method." U.S. Bureau of Mines, Report of Investigations,
no. 4547 (1949) 19 pp. (CA 44: 307e, 1950).
Jacobs, Morris B., Braverman, M.M., and Hochheiser, Seymour. "Ultramicro-
determination of Sulfides in Air". Analytical Chemistry, 29 (1957)
p. 1349-51.
Gustafsson, Lilly. "Determination of Ultramicro Amounts of Sulphate as
Methylene Blue. I. The Colour Reaction". Talanta, 4 (1960) p.227-35.
Choudens, C. de. "Quantitative Spectrophotometric Analysis for Sulphur
Dioxide and Hydrogen Sulphide in Gaseous Effluents from Recovery Units
of a Sulphate-Pulp Mill". Revue Association techq. Industrie
Papetiere, 22(2) (1968) p. 113-121. (Ca 17: 1216, Aug. 1969).
237
-------
REACTIONS WITH IODINE
As a reducing agent, H2S is a natural for determination with an oxidant.
Iodine has been used for this purpose for a long time. The major variations
in methods based on determination with iodine are in how the H2S is collected
and prepared for titration (or reaction) with 12. Some employ a metal ion
and then react with 12 (11, 17, 19, Ill, 113, also 114), while others use 12
as the absorbent (13t 16, 18, 110) and back titrate with thiosulfate. The
various references uncovered on the use of iodine are:
11
Zhdanov, V.V. "Rapid Methods for Determination of Hydrogen Sulfide in
Coke-Oven Gas". .Zavodskaya Laboratoriya Zhurnal, 6 (1937)
p. 1448-51. (Ca 32: 2323, 1938).
12
Cherepennikov, A.A. "Hydrogen Sulfide, Carbon Dioxide and Sulfur Contents
of Natural Gas from the Bashneft Oil Wells". Neftyanoe Khozyaistvo,
2 (1937) p. 68-9. (CA 32: 43114' 1938).
13
Krafft, Hans. "The Chemical Laboratory in the Service of Safety in Mining".
Montan. Rundschau, 31(20) (1939) p. 561-7. (CA 34: 59589' 1940).
14
Payer, Theo and Lehrenkrauss, Adolf. "Determination of Hydrogen Sulfide and
Hydrocyanic Acid in Manufactured Gas". Gas und Wasserfach, 82 (1939)
p.7l3-l5. (CA 34: 6096, 1940).
IS
Strada, Mario, and Macri, Antonio. "Rapid Method for Determining Hydrogen
Sulfide in Technical Gas Mixtures". Annali di Chimica Applicata,29
(1939) p. 64-8. (CA 33: 85246' 1939). .
16
Wilson, Stuart H. "The Analysis of Hot-Spring Gases". New Zealand Journal
of Science and Technology, 20B (1939) p. 233-48. (CA~: 67458' 1939).
17
Shaw, Joseph A. "Rapid Determination of Hydrogen Sulfide and Mercaptan
Sulfur in Gases and in Aqueous Solutions". Industrial and Engineering
Chemistry, Analytical Edition, 12 (1940) p. 668-71. (CA 35: 53a, 1941).
1a
Kraft, Janos. "The Chemical Laboratory as an Aid to Mine Safety".
Banyzszati es Kohaszati Lapok, 75 (1942) p. 113-17. (CA 1I:4l53a,1943).
19
Anon. "Test Procedure for Determining H2S and Mercaptan Content of Natural
Gas". American Gas Journal, 162(6) (1945) p. 47,60. (CA 39: 34136'
1945). -
no
Kitano, Yasushi and Takakuwa, Hidematsu. "Determination of Hydrogen Sulfide
and Sulfur Dioxide in Air. 1. Errors in Iodometry". Japan Analyst, 3
(1954) p.7-10. (CA 48: 69l3e, 1954).
238
-------
111
112
113
114
REACTIONS WITH IODINE (CONTINUED)
Wickert, K. "The Determination of S02, S03' and HZS in Flue Gases".
Brennstoff-Warme-Kraft, 12 (1960) p. 449-51. (CA 55: 1961).
Cave, G.C.B. "The Collection and Analysis of Odorous Gases from Kraft Pulp
Mills.!. Theoretical Considerations". Tappi, 46 (1) (1963) p. 1-5.
(CA 59: 8967, 1963).
Shul'gina, E., Arutyunova, A.Kh., Blyumshtein, A.E. "Determination of HZS
in Gases". Neftepererabotka i Neftekhimiya, Nauchn.-Tekn.Sb., 3 (1964)
p. 26-9. (CA 61: 1964).
"Separation and Determination of Mercaptans, Hydrogen Sulfide,
Organic Sulfides, and Organic Disulfides, and Organic Disulfides in an
Air Stream". Bay Area Air Pollution Control Distr~ct, San Francisco,
California, Method M-l.
239
-------
INDICATOR TUBES
A variety of Approaches have been taken to develop indicator tubes for H2S,
These include: a paper impregnated with potassium hydroxide, potassium
zincate and glycerol (IT9); lead acetate tiles (IT7); lead acetate paper
(ITl, IT2, ITlO, IT12); carriers impregnated with copper salts (IT4, IT5);
silver cyanide (IT6); bismuth nitrate (ITll); lead acetate (IT14); metallic
silver fibers (IT15). References found are:
ITI
Eymann, Constanz. "Determination of Hydrogen Sulfide and Hydrocyanic
Acid in Gases". Gas und Wasserfach, 81 (1938) p. 484-8. (CA B.:
8112, 1938). .
IT2
Sliva, Vitezslav. "The Determination of Hydrogen Sulfide in Illuminating
Gas". Plyn Voda a Zdravotni Technika, 18 (1938) p. 49-51. (CA 34:
42552' 1940).
IT3
Littlefield, John B. "Determination of Constituents Such as Hydrogen
Sulfide in Gaseous Atmospheres". (to Mine Safety Appliances Co.) U.S.
2,174,349. Sept. 26. (CA 34: 6926' 1940).
IT4
Smith, Bengt. "Quantitative Analysis of Mixtures of Hydrogen Sulfide and
Sulfur Dioxide". Transactions of Chalmers University of Technology,
Gothenburg, no. 150 (1954) 19pp. (CA 49: 804li, 1955).
IT5
Heuschkel, G. "Analysis of Gases Containing Hydrogen Sulphide and Sulphur
Dioxide by Means of Indicator Tubes". Erdol und Kohle, 14(6) (1961)
p. 467-68. (AA~: June 1962).
IT6
Schumann, H., Lovenstrin, K., Blaschke, H. of VEB Chemiewerk Coswig.
"Method and Device for the Determination of Hydrogen Sulphide in
Gases". British Patent 1,047,700; date app1. 2/9/63. (AA 14: 5842,
Sept. 1967). -
IT7
Gilardi, Edward F. and Manganelli, Raymond M. "A Laboratory Study of a
Lead Acetate-Tile Method for the Quantitative Measurement of Low
Concentrations of Hydrogen Sulfide". Journal of the Air Pollution
Control Association, 13(7) (1963) p. 305-9.
ITS
Vol 'berg, N. She and Gershkovich, E.!. "Use of Solid Adsorbents in
Industrial Sanitation Chemistry". Materialy k Nauchn. SessH
Posvyashch. 40-1etiyu Gos. Nauchn.-Issled. Inst. Gigieny Truda
Profzabolevanii, Leningrad, Sb., 1964, 56. (CA 64: 1966).
Huygen, C. "The Sampling of Hydrogen Sulfide in Air with Impregnated
Filter Paper". Analytica Chimica Acta, 30(6) (1964) p. 556-564.
i
IT9
240
-------
ITlO
ITll
ITl2
ITl3
ITl4
lTl5
INDICATOR TUBES (CONTINUED)
High, M.D. and Horstman, S.W. "Field Experience in Measuring Hydrogen
Sulfide". "American Industrial Hygiene Association Journal, 26(4)
(July-August 1965).
Sinkevish, O.V. and Noshchenko, A.E. "Determination of the Hydrogen
Sulfide Content of the Air". USSR 174,002 (CloG Oln) Aug. 6, 1965.
(CA 64: 1966).
Leidnitz, Kurt. "Determination of Hydrogen Sulfide in Coke-oven Gas".
Gas und Wasserfach, 106 (1965) p. 1204-6. (CA 66: 1967).
Grosskopf, Karl. "Tube Method in the Analytical Laboratory". Fortschr.
Chern. Forsch, 5(3) (1966) p. 530-67. (CA~: 1966). (A review.)
Ryashentseva, M.A. and Afanas'eva, Yu.A. "Adsorption-Chemical Determination
of Hydrogen Sulfide and Sulfur Dioxide in Gas Mixtures". Khim Tekhnol.
Top1. Masel, 12(3) (1967) p. 61-3. (CA~: 1967).
Falgout, D.A. and Harding, C.l. "Determination of H2S Exposure by Dynamic
Sampling with Metallic Silver Filters". Journal of the Air Pollution
Control Association, 18(1) (1968) p. 15-20.
241
-------
TURBIDIMETRIC METHODS
Trace quantities of H2S can be estimated from the amount of precipitate
formed after reaction with a metal ion. Copper (TM1), cadmium (TM2, TM3,
TMS), bismuth (TM3) , and arsenic (TM4) have been used for this purpose.
The turbidity can be estimated visually or via use of a spectrophotometer.
References found to this approach were:
TMl
Kastner, E.P. "The Photocolorimetric Determination of Hydrogen Sulfide in
the Air". Journal of Applied Chemistry, 12 (1939) p. 1097-1103.
(CA 34: 3203, 1940).
TM2
Bergstr(jm, H. and Trobech, K.G. "Investigations of Black Liquor". Svensk
Papperstidning, 42 (1939) p. 554-7. (CA 34: 1171, 1940).
TM3
Field, E. and 01dach, C.S. "Determination of Hydrogen Sulfide in Gases".
Industrial and Engineering Chemistry, Analytical Edition, 18 (1946)
p. 665-7. (CA 41: 54, 1947).
TM4
Ethrington, C.G., Warren, H., and Marsden, F.C. "Colorimetric
Determination of Small Amounts of Hydrogen Sulfide in Effluent
by Means of the Spekker Absorptiometer". Analyst, 75 (1950)
p. 209-11. (CA 44: 6345g, 1950).
Gases
TMS
Lahmann, Erdwin, and Prescher, Karl E.
Determination in the Atmosphere".
(CA 64: 1966).
"Intermittent Hydrogen Sulfide
Staub, 25(12) (1965) p. 527-8.
242
-------
TRl
TR2
TR3
TR4
TR5
TR6
TR7
TRB
TR9
T1ll0
MISCELLANEOUS METHODS WITH A TITRATION READ OUT
In addition to those based on an iodine titration, a number of other methods
have been developed with a titration read-out. Some are based on oxidation -
eerie (TRl), permanganate (TR2). Others depend on reaction of H2S with a
metal ion and determination of the excess metal - mercury (TRS), cadmium (TR8,
TRIO) and zinc (TR9, TRlO). A few utilize direct titration and a potentio-
metric end-point (TR3, TR6), another conductometric titration after collection
in hydroxide (TR7), while another depends on catalysis for selectivity (TR4).
The total list of references in this general category is:
Lilly, R.M. and Chesnutt, N.P. "Determining Hydrogen Sulfide Content of Gas
in the Field". Oil and Gas Journal, 36(14) (1937) p.52,5S,58,70.
(CA 31: 8885, 1937).
Seuthe, Ad. "Determination of Hydrogen Sulfide in Coke-Oven Gas". Chemiker-
Zeitung/Chemische Apparatur, 65 (1941) p. 59. (CA 35: 3416, 1941).
Felicetta, V.F., Peniston, Q.P., and McCarthy, J.L. "Determination of
Hydrogen Sulfide, Methyl Mercaptan, Dimethyl Sulfide, and Disulfide in
Kraft Mill Process Streams". Canadian Pulp and Paper Industry, 5(12)
(1952) p.16,18,20,22,24,26-7,30,41. (CA!!2: 5115, 1953).
Gershkovich, E.E. "Catalytic Reactions in Industrial Sanitary Chemistry".
Trudy Nauchnoi Sessii Leningradskogo Nauchno - Issledovatel'skogo
Instituta Gigieny Truda i Profzabolevanii, 1958, p.153-8. (pub. 1959).
(CA~: 1962).
HOffmann, E. "Mercurimetry in the Microquantitative Analysis of Iodides,
Cyanides, and Sulfides in Solution and Hydrocyanic Acid and Hydrogen
Sulfide in Gases". Zeitschrift fuer Ana1ytische Chemie, 169 (1959)
p. 258-63. (CA 54: 1960).
Tamele, M.W., Irvine, V.C., and Ryland, L.B. "Potentiometric Determination
of Sulfide Ions and the Behavior of Silver Electrodes at Extreme
Dilution" . Analytical Chemis try, 32 (1960) p. 1002-7.
Oehme F. "High-frequency Titrimetric Determination of Hydrogen Sulphide
, "
and Thiols in Technical Gases after Absorption in Alkaline Solution.
Erdtl1 und Kohle, 13 (1960) p. 394-96. (AA~: 1961).
Baranenko S.E. and Krivosheeva, V.I. "Tri1onometric Determination of Hydrogen
, "G i
Sulfide in Natural, Accessory, and other Gases. Voprosy Razvitiya azovo
Promyshlennosti Ukrainskoi SSR, Kiev.Sb., (1963)p.300-3. (CA 61: 1964).
Ba1abanoff, L. and Soto, L. "Rapid Determination of Hydrogen Sulfide with
EDTA". Chimie Ana1ytique, 46 (1964) p. 90-92. (AA 12: 1965).
Zugrlvescu, P.G. and Zugrivescu, M.A. "Determination of Trace Impurities
(Acetylene, Phosphine, Arsine, Hydrogen Sulfide and Carbon Dioxide) in
Gases". Revista de Chimie, 17(11) (1966) p. 704-5. (AA 1St Feb. 1968).
243
-------
MISCELLANEOUS COLOR REACTIONS
Many different reactions have been employed to H2S which end up with a
color read-out. Some of these are based on the well known starch-iodine
color (CR1). Lauth's violet (CR6). nitroprusside (CR7). and molybdenum
blue (CR8. CR9). The complete bibliography found for this section was:
CRl
Blohm. Clyde L., and Riesenfeld. Fred C. "Simultaneous Determination of
Hydrogen Sulfide and Carbon Dioxide in a Continuous Gas Stream".
Industrial and Engineering Chemistry. Analytical Edition. 18 (1946)
p. 373-6. (CA 40: 43158. 1946).
CR2
.Taramasso. M. and Piccinini. A. "Determination of Hydrogen Sulfide in
Light Petroleum Gases". Rivista dei Combustibili. 9 (1955) p.933-9.
(CA 50: 9724. 1956).
CR3
Maksimov. V.F. "Volatile Sulfur Compounds in a Kraft Pulp Mill". Trudy
Leningradskogo Tekhnologicheskogo Instituta imeni Lensoveta, 5
(1958) p. 19-22. (CA~: 1962).
CR4
Chetkowaka. M.. Gallus-Olender. J.. and Strzeszewska. 1. "Continuous
Determination of Hydrogen Sulfide in Air". Chemik. 14 (1961)
p. 384-6. (CA 56: 1962).
CR5
Gavrilets. E.S. and Demchuk. M.V. "Determination of CO? NH3. and H2S
in the Air of Living Quarters". Naukovi Pratsi. L vivskii
Zooveterinarnii Institut. 11 (1961) p. 89-96. (CA 59: 12076. 1963).
CR6
Murray. F.E. and Raynor. H.B. "A Procedure for Sampling and Analysis for
Hydrogen Sulfide in Kraft Mill Stack Gases". Tappi. 44(3)
(March 1961) p. 219-21.
CR7
Mokhov. L.A. and Matveeva. S.A. "Colorimetric Determination of Hydrogen
Sulfide in Air". Laboratornoe Delo. 8(3) (1962) p. 44-47. (CA~:
1962).
CR8
Buck. M. and Stratmann. H. "Determination of Hydrogen Sulfide in the
Atmosphere". Staub. 24 (July 1964) p. 241-50. (APCA Abstracts X
(8). January 1965).
CR9
Buck. M. and Geis. H. "Measurement of Hydrogen Sulphide in the Atmosphere.
Joint Determination of Hydrogen and Sulfur Dioxide". Staub.
Reinha1tung der Luft. 26 (Sept. 1966) p. 379-384. (Fuel Abstracts.
!: 2785. 1967).
244
-------
HI
M2
M3
M4
M5
M6
M7
M8
K9
MISCELLANEOUS
Other techniques that have been used for H S include: polarography (Ml, MS,
M9), cou1ometry (M4), x-ray absorption (M2t, fluorescence (M3, M7), catalysis
(MS), and precipitation (M6). The references are:
prchlik J. "Polarographic Methods Applied to Illuminating Gas".
Voda, 30 (1950) p. 303-8. (CA 45: 5908, 1951).
Palwa a
Quiram, Ernest R. "New Way to Continuous Hydrogen Sulfide Analysis".
Petroleum Refiner, 38(7) (1959) p. 143-4. (CA 53: 1959).
Wronski, M. "The Use of Mercurated Phenolphthalein and Fluorescein in
Chemical Analysis". Zeits~hrift fuer Analytische Chemie, 175(6)
(1960) p. 432-36. (AA~: Feb. 1961).
Ceskoslovenska Akademie Ved. "Electrochemical Gas Analysis". Belg.
624,208, Feb. 14, 1963; Czech. App1. Oct. 27, 1961; 12 pp. (CA 58:
1963).
Gershkovich, E.E. "Catalytic Reactions in Industrial and Sanitation Chemistry".
Novoe v Oblasti Sanitarno-Khimicheskogo Analiza, 1962, p. 53-63.
(CA 59: 12076, 1963).
Demus, H. and Liebert, H. "Analysis of a Mixture of Hydrogen Sulphide,
Carbon Dioxide, Carbon Disu1phide and Air". Faserforschung und
Texti1technik, 13(8) (1962) p. 376-77. (AA 10: May, 1963).
Andrew, T.R. and Nichols, P.N.R. "The Determination of Hydrogen Sulfide in
the Atmosphere". Analyst, 90(1071) (1965) p. 367-70. (CA 63: 1965).
Capuano, Ita10 A. "Continuous Polarographic Analysis for Low Hydrogen
Sulfide Concentrations". (to Union Carbide Corp.) U.S. 3,304,243
(C1.204-1), Feb. 14, 1967. 5 pp.
Wolf, F. and Langen, H. "Combined Trace Analysis of H2S and COS in 802-
containing Waste Gases". Chemie-Ingenieur-Tecknik, 39(16) (1967)
p. 945-8. (CA 67: 1967).
245
-------
17
:9
LR9
tP7
rR8
1M2
Rl
~P2
~R8
~R9
18
[12
Bibliography by Author
Andrew, T.R. and Nichols, P.N.R. "The Determination of Hydrogen Sulfide in
the Atmosphere". Analyst, 90(1071) (1965) p. 367-70. (CA 63: 1965).
Anon. "Test Procedure for Determining H2S and Mercaptan Content of Natural
Gas". American Gas Journal, 162(6) (1945) p.47, 60. (CA 39: 34136' 1945).
Ba1abanoff, L. and So to , L. "Rapid Determination of Hydrogen Sulfide with
EDTA" . Chimie Ana1ytique, 46, 1964, p. 90-92. (AA g: 1965).
Bamesberger, W.L. and Adams, D.F. "Improvements in the Collection of
Hydrogen Sulfide in Cadmium Hydroxide Suspension". Environmental Science
and Technology, 3(3), March 1969, p. 258-61.
Baranenko, S.E. and Krivosheeva, V.l. "Trilonometric Determination of
Hydrogen Sulfide in Natural, Accessory and Other Gases". Voprosy
Razvitiya Gazovoi Promyshlennosti Ukrainskoi SSR, Kiev. Sb., 1963
p. 300-3. (CA 61: 1964).
Bergstrom, H. and Trobeck, K.G. "Investigations of Black Liquor".
Papperstidning, 42 (1939) p. 554-7. (CA 34: 1171, 1940).
Svensk
Blohm, Clyde L., and Riesenfeld, Fred C. "Simultaneous Determination of
Hydrogen Sulfide and Carbon Dioxide in a Continuous Gas Stream".
Industrial and Engineering Chemistry, Analytical Edition, 18 (1946)
p. 373-6. (CA 40: 43158 1946).
Brychta, Miroslav and Strobl, JiiL "Hydrogen Sulfide Determination in
Illuminating Gas". Paliva, 32 (1952) p. 113-116. (CA 50: 8173, 1956).
Buck, M. and Stratmann, H. "Determination of Hydrogen Sulfide in the
Atmosphere". Staub, 24 (July 1964) p. 241-50. (i\PCA Abstracts X (8),
January 1965).
Buck, M. and Geis, H. "Measurement of Hydrogen Sulphide in the Atmosphere.
Joint Determination of Hydrogen and Sulfur Dioxide". Staub,Reinha1tung
der Luft, 26 (Sept. 1966) p. 379-384. (Fuel Abstracts !: 2785, 1967).
Capuano, Italo A. "Continuous Polarographic Analysis for Low Hydrogen Sulfide
Concentrations". (to Union Carbide Corp.) U.S. 3,304,243 (Cl. 204-1),
February 14, 1967. 5 pp.
Cave, G.C.B. "The Collection and Analysis of Odorous Gases from Kraft Pulp
Mills. 1. Theoretical Considerations". Tappi, 46(1) (1963) p. 1-5.
(CA 59: 8967. 1963).
246
-------
Bibliography (Cont'd)
14
Ceskoslovenska Akademie Ved. "Electrochemical Gas Analysis". Belg.
624,208. Feb. 14, 1963; Czech. Appl. Oct. 27, 1961; 12 pp. (CA 58:
1963) .
[2
Cherepennikov, A.A. "Hydrogen Sulfide, Carbon Dioxide and Sulfur Contents
of Natural Gas from the Bashneft Oil Wells". Neftyanoe Khozyaistvo,
2 (1937) p. 68-9. (CA 32: 43114 1938).
:R4
Chetkcwaka, M., Gallus-Olender, J., and Strzeszewska, 1. "Continuous
Determination of Hydrogen Sulfide in Air". Chemik, 14 (1961) p.384-6.
(CA~: 1962).
m6
Choudens, C. de. "Quantitative Spectrophotometric Analysis for Sulphur
Dioxide and Hydrogen Sulphide in Gaseous Effluents from Recovery Units
of a Sulphate-Pulp Mill". Revue Association techq. Industrie Papetiere,
22(2) (1968) p. 113-121. (CA 17: 1216, Aug. 1969).
16
Demus, H. and Liebert, H. "Analysis of a Mixture of Hydrogen Sulphide,
Carbon Dioxide, Carbon Disulfide and Air". Faserforschung und
Textiltechnik, 13(8) (1962) p. 376-77. (AA 10: May 1963).
~4
Ethrington, C.G., Warren, H. and Marsden, F .C. "Colorimetric Determination
of Small Amounts of Hydrogen Sulfide in Effluent Gases by Means of the
Spekker Absorptiometer". Analyst, 75 (1950) p. 209-11. (CA 44: 6345g,
1950).
:Tl
Eymann, Constanz. "Determination of Hydrogen Sulfide and Hydrocyanic Acid
in Gases". Gas und Wasserfach, 81 (1938) p. 484-8. (CA E: 8112, 1938).
[TlS Falgout, D.A. and Harding, C.1. "Determination of H2S Exposure by Dynamic
Sampling with Metallic Silver Filters". Journal of the Air Pollution
Control Association, 18(1) 1968. p. 15-20.
rR3
Felicetta, V.F., Peniston, Q.P., and McCarthy, J.L. "Determination of
Hydrogen Sulfide, Methyl Mercaptan, Dimethyl Sulfide, and Disulfide in
Kraft Mill Process Streams". Canadian Pulp and Paper Industry,S (12)
(1952) p.16,18,20,22,24,26-7,30,4l. (CA 47: 5115, 1953).
rM3
Field, E. and Oldach, C. S. "Determination of Hydrogen Sulfide in Gases".
Industrial and Engineering Chemistry, Analytical Edition, 18 (1946)
p.665-7. (CA41: 54, 1947).
1B2
Fogo, James K. and Popowsky, Milton. "Spectrophotometric Determination of
Hydrogen Sulfide.. Methylene Blue .Method". ~alytical Chemistry, 21
(1949) p. 732-4. (CA 43: 7375, 1949).
:as
Gavri1ets, E.8. and Demchuk, M.V. "Determination of C02' NH). and H2S in the
Air of Living Quarters". .. Naukovi Pratsi, L'vivskii Zooveterinarnii
Institut, 11 (1961) p. 89-96. (CA 59: 12076, 1963).
247
-------
Bibliography (Cont'd)
Gershkovich, E.E. "Catalytic Reactions in Industrial Sanitary Chemistry".
Trudy Nauchnoi Sessii Leningradskogo Nauchno - Iss1edovate1'skogo
Instituta Gigieny Truda i Profzabo1evanii, 1958, p. 153-8 (1959).
(CA 21.: 1962).
Gershkovich, E.E. "Catalytic Reactions in Industrial and Sanitation Chemistry".
Novoe v Oblasti Sanitarno-Khimicheskogo Analiza, 1962, p. 53-63.
(CA 59: 12076, 1963).
Gilardi, Edward F. and Mangane11i, Raymond M. "A Laboratory Study of a Lead
Acetate-Tile Method for the Quantitative Measurement of Low Concentrations
of Hydrogen Sulfide". Journal of the Air Pollution Control Association,
13(7) (1963) p. 305-9. (CA 59: 9236, 1963).
Grosskopf, Karl. "Tube Method in the Analytical Laboratory". Fortschr. Chern.
Forsch, 5(3) (1966) p. 530-67. (A review). (CA 65: 1966).
Heuschke1, G. "Analysis of Gases Containing Hydrogen Sulphide and Sulphur
Dioxide by Means of Indicator Tubes". Erdo1 und Koh1e, 14(6) (1961)
p. 467-68. (AA~: June 1962).
)
High, M.D. and Horstman, S.W. "Field Experience in Measuring Hydrogen Sulfide".
American Industrial Hygiene Association Journal, 26(4) (July-Aug. 1965)
p. 366-73.
Hoffmann, E. "Mercurimetry in the Microquantitative Analysis of Iodides,
Cyanides, and Sulfides in Solution and Hydrocyanic Acid and Hydrogen
Sulfide in Agses". Zeitschrift fuer Ana1ytische Chemie, 169 (1959)
p. 258-63. (CA 54: 1960).
Huygen, C. "The Sampling of Hydrogen Sulfide in Air with Impregnated Filter
Paper". Ana1yticaChimicaActa, 30(6) (1964) p. 556-64.
Jacobs, M.B. The Chemical Analysis of Air Pollutants, p. 182-194.
Interscience Publishers, Inc. N.Y. 1960.
Jacobs, Morris B., Braverman, M.M., and Hochheiser, Seymour. "U1tramicro-
determination of Sulfides in Air". Analytical Chemistry, 29 (1957)
p. 1349-51).
Jensen, George A., Adams, Donald F. and Stern, Harry. "Absorption of H2S and
Methyl Mercaptan from Dilute Gas Mixtures". Journal of Air PollutIon
Control Association, 16(5) (1966) p. 248-53. (CA 65: 1966).
Kastner, E.P. "The Photoco1ormetric Determination of Hydrogen Sulfide in the
Air". Journal of Applied Chemistry, 12 (1939) p. 1097-1103. (CA 34:
3203, 1940). -
248
-------
110
MBl
13
18
RP5
TM5
IT12
.tU
IT3
CR3
CR7
CR6
TR7
14
Bibliography (Cont'd)
Kitano Yasushi and Takakuwa. Hidematsu. "Determination of Hydrogen Sulfide
a~d Sulfur Dioxide in Air. I. Errors in Iodometry". Japan Analyst. 3
(1954) p. 7-10. (CA 48: 6913e. 1954).
Kosior. K.E.A. "Determination of Hydrogen Sulfide in Natural Gas".
Canadian Chemistry and Process Industry, 32 (1948) p. 925-6, 929.
(CA 43: 835, 1949).
Krafft. Hans. "The Chemical Laboratory in the Service of Safety in Mining".
Montan. Rundschu, 31(20) (1939) p. 561-7. (CA 34: 59589; 1940).
Kraft, Janos. "The Chemical Laboratory as an Aid to Mine Safety".
Banyzszati es Kohaszati Lapok, 75 (1942) p. 113-17. (CA 37: 41538' 1943).
Lahmann, Erdwin. "Measurement of Gaseous Sulfur Compounds in the Atmosphere".
Erdoe1 und Koh1e, 18(10) (1965) p. 796-800. (CA 64: 1966).
Lahmann, Erdwin, and Prescher, Karl E.
Determination in the Atmosphere".
(CA 64: 1966).
"Intermittent Hydrogen Sulfide
Staub, 25(12) (1965) p. 527-8.
Leidnitz, Kurt. "Determination of Hydrogen Sulfide in Coke-oven Gas".
Gas und Wasserfach, 106 (1965) p. 1204-6. (CA 66: 1967).
Lilly, R.M. and Chesnutt, N.P. "Determining Hydrogen Sulfide Content of Gas
in the Field". Oil and Gas Journal, 36(14) (1937) p. 52,55,58,70.
(CA 31: 8885, 1937).
Littlefield, John B. "Determination of Constituents Such as Hydrogen
Sulfide in Gaseous Atmospheres". (to Mine Safety Appliances Co.)
2,174,349. Sept. 26. (CA 34: 6926' 1940).
U.S.
Maksimov, V.F. "Volatile Sulfur Compounds in a Kraft Pulp Mill". Trudy
Leningradskogo Tekhnologicheskogo Instituta imeni Lensoveta, 5 (1958)
p. 19-22. (CA 56: 1962).
Mokhov, L.A. and Matveeva, S.A. "Colorimetric Determination of Hydrogen
Sulfide in Air". Laboratornoe Delo, 8(3) (1962) p. 44-47. (ca 57: 1962).
Murray, F.E. and Raynor, H.B. "A Procedure ofr Sampling and Analysis for
Hydrogen Sulfide in Kraft Mill Stack Gases". Tappi, 44(3) (March 1961)
p. 219-221).
Oebem, F. "High-frequency Titrimetric Determination of Hydrogen Sulphide and
Thiols in Technical Gases after Absorption in Alkaline Solution". Erdol
und kohle, 13 (1960) p.394-96. (AA!: 1961).
Payer, Theo and Lehrenkrauss. 'Adolf. "Determination of Hydrogen Sulfide and
Hydrocyanic Acid in Manufactured Gas". Gas und Wasserfach. 82 (1939)
~. 113-15. (CA 1!1 6096' 1940).
249
-------
Bibliography (Cont'd)
Prchlik, J. "Polarographic Methods Applied to Illuminating Gas".
a Voda, 30 (1950) p. 303-8. (CA 45: 5908, 1951).
P alwa
Quiram, Ernest R. "New Way to Continuous Hydrogen Sulfide Analysis".
Petroleum Refiner, 38(7) (1959) p. 143-4. (CA 1l: 1959).
Risenfield, F.C. and Orback, H.K. "Methods for Determinging Hydrogen Sulfide
in Gases". Petroleum Engineer, 25(6) (1953) C32-8. (CA 49: 12197, 1955).
~
Ryashentseva, M.A. and Afanas'eva, Yu.A. "Adsorption-Chemical Determination
of Hydrogen Sulfide and Sulfur Dioxide in Gas Mixtures". Khim. Tekhnol.
Topl. Masel., 12(3) (1967) p. 61-3. (CA~: 1967).
Sands, A.E., Grafius, M.A., Wainwright, H.W., and Wilson, M.W. "The
Determination of Low Concentrations of Hydrogen Sulfide in Gas by the
Methylene Blue Method". U.S. Bureau of Mines, Report of Investigations,
no. 4547, (1949) 19 pp. (CA 44: 307e, 1950).
Schwnann, H., Lobenstein, K., Blaschke, H. of VEB Chemiewerk Coswig. "Method
and Device for the Determination of Hydrogen Sulphide in Gases". British
Patent 1,047,700; date apple 2/9/63. (AA 14: 5842, Sept. 1967).
Seuthe, Ad. "Determination of Hydrogen Sulfide in Coke-oven Gas". Chemiker-
Zeitung/Chemische Apparatur, 65 (1941) p. 59. (CA~: 3416, 1941).
Shaw, Joseph A. "Rapid Determination of Hydrogen Sulfide and Mercaptan
Sulfur in Gases and in Aqueous Solutions". Industrial and Engineering
Chemistry, Analytical Edition, 12 (1940) p. 668-71.
(CA 35: 538' 1941).
Shul'gina, E., Arutyunova, A.Kh., Blyumshtein, A.E. "Determination of H2S
in Gases". Neftepererabotka i Neftekhimiya, Nauchn.-Tekn.Sb., 3 (1964)
p. 26-29. (CA 61: 1964).
1
Sinkevish, O.V. and Noshchenko, A.E. "Determination of the Hydrogen Sulfide
Content of the Air". USSR 174,002 (CloG Oln) Aug. 6, 1965. (CA 64: 1966).
Sliva, Vitezslav. "The Determination of Hydrogen Sulfide in Illuminating
Gas". Plyn Voda a Zdravotni Technika, 18 (1938) p. 49-51. (CA 34:
42552 1940).
Smith, Bengt. "Quantitative Analysis of Mixtures of Hydrogen Sulfide and
Sulfur Dioxide". Transactions of Chalmers University of Technology.
Gothenburg, 150 (1954) 19pp. (CA 49: 804li, 1955).
Smith, A.F., Jenkins, D.G. and Cunningworth, D.E. "Measurement of Trace
Quantities of Hydrogen Sulphide in Industrial Atmospheres". Journal of
Applied Chemistry, 11 (Sept. 1961) p. 317-29.
250
-------
15
TR6
CR2
ITS
III
16
'19
M3
II
TRIO
114
Bibliography (Cont'd)
Strada Mario and Macri, Antonio. "Rapid Method for
S~1f1de in Technical Gas Mixtures". Annal! di
(1939) p. 64-68. (CA 33: 85246, 1939).
Derermining Hydrogen
Chimica Applicata, 29
Tamele, M.W., Irvine, V.C., and Ryland, L.B. "Potentiometric
of Sulfide Ions and the Behavior of Silver Electrodes at
Dilution". Analytical Chemistry, 32 (1960) p. 1002-7.
Determination
Extreme
Taramasso, M. and Piccinini, A. "Determination of Hydrogen Sulfide in
Light Petroleum Gases". Rivista dei Combustibili, 9 (1955) p. 933-9.
(CA 50: 9724, 1956).
Vol'berg, N.Sh. and Gershkovich, E.l. "Use of Solid Adsorbents in Industrial
Sanitation Chemistry". Materialy k Nauchn.Sessii Posvyashch.40-letiyu
Gos. Nauchn.-Issled. Inst. Gigieny Truda i Profzabolevanii, Leningrad,
~, 1964, 56. (CA 64: 1966).
Wickert, K. "The Determination of 502' 503 and H2S in Flue Gases".
Brennstoff-Warme-Kraft, 12 (1960) p. 449-51. (CA 55: 1961).
Wilson, Stuart H. "The Analysis of Hot-Spring Gases". New Zealand Journal of
Science and Technology, 20B (1939) p. 233-48. (CA 33: 67458, 1939).
Wolf, F. and Langen, H. "Combined Trace Analysis of H2S and COS in S02-
containing Waste Gases". Chemie-Ingenieur-Tecknik, 39(16) (1967) p.945-
948. (CA 67: 1967).
Wronski, M. "The Use of Mercurated Phenolphthalein and Fluorescein in
Chemical Analysis". Zeitschrift fuer Analytische Chemie, 175(6) (1960)
p. 432-36. (AA!: Feb. 1961).
Zhdanov, V.V. "Rapid Methods for Determination of Hydrogen Sulfide in Coke-
Oven Gas". Zavodskaya Laboratoriya .Zhurnal, 6 (1937) p. l448-5l.
(CA 32: 2323, 1938).
ZugrAvescu, P.G. and Zugravescu, M.A. "Determination of Trace Impurities
(Acetylene, Phosphine, Arsine, Aydrogen Sulphide and Carbon Dioxide) in
Gases". Revista de Chimie, 17(11) (1966) p. 704-5. (AA.!1: Feb. 1968).
"Separation and Determination of Mercaptans, Hydrogen Sulfide,
Organic Sulfides, and Organic Disulfides in an Air Stream". Bay Area
Air Pollution control District, San Francisco, Calif. Method M-l.
251
------- |