United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park NC 27711
EPA-34O/1-83-O17
January 1983
Stationary Source Compliance Series
vvEPA
Kraft Pulp Milj
Inspection Guide
-------
-------
EPA-340/1-83-017
Kraft Pulp Mill
Inspection Guide
by
PEDCo Environmental, Inc.
11499 Chester Road
Post Off ice Box 46100
Cincinnati, Ohio 45246-0100
Contract No. 68-01-6310
Work Assignment No. 65
PN 3 660-1-65
John R: Busik, Project Officer
Robert C. Marshall, Task Manager
U.S. ENVIRONMENTAL PROTECTION AGENCY
Stationary Source Compliance Division
401 M Street, S.W.
Washington, D.C. 20460
• January 1983
-------
DISCLAIMER
This report was prepared by PEDCo Environmental, Inc. Cincinnati, Ohio,
under Contract No. 68-01-6310, Work Assignment No. 65. It has been reviewed
by the Stationary Source Compliance Division of the Office of Air Quality
Planning and Standards, U.S. Environmental Protection Agency and approved for
publication. Approval does not signify that the contents necessarily reflect
the views and policies of the U.S. Environmental Protection Agency. Mention
of trade names or commercial products is not intended to constitute endorse-
ment or recommendation for use. Copies of this report are available from the
National Technical Information Services, 5285 Port Royal Road, Springfield,
Virginia 22161.
ii
-------
CONTENTS
Figures
Tables
Acknowledgment
Executive Summary
1. Introduction
1.1 Purpose and Scope
1.1.1 Level I
1.1.2 Level II
1.1.3 Level III
1.1.4 Level IV
1.2 Continuous Compliance
1.3 Organization of Inspection Guide
1.4 Industry Overview
1.5 Regulation Under the Clean Air Act
1.5.1 State Implementation Plan
1.5.2 Federal standards of performance of new sources
2. General Preparatory and Preinspection Procedures
2.1 File Review
2.2 Safety Precautions
2.2.1 Exposure to hydrogen sulfide
2.2.2 Exposure to chlorine
2.3 Safety and Inspection Equipment
2.4 Preentry Observations
2.5 On-Site Inspection Checklists
3. Kraft Pulping Processes
3.1 Wood Handling Department
viii
xiii
xvi
xvii
1
2
2
3
4
5
7
13
15
16
21
21
22
32
32
35
36
37
53
53
111
-------
CONTENTS (continued)
Page
3.1.1 Process description
3.1.1.1 Debarking
3.1.1.2 Chipping
3.1.1.3^ Knotting and screening
3.1.1.4 Storage and transfer
3.1.2 Sources of emissions and control
3.1.3 Malfunctions
3.1.4 Inspection of wood handling department
3.2 Pulping Department
3.2.1 Process description
3.2.1.1 Pulp digester
3.2.1.2 Batch digester (blow)
3.2.1.3 Digester relief and turpentine recovery
3.2.1.4 Continuous digester
3.2.1.5 Pulp washing
3.2.1.6 Black liquor concentration (evaporation)
3.2.1.7 Condensate stripping
3.2.1.8 Black liquor oxidation
3.2.2 Sources of emissions and control
3.2.2.1 Digester and blow tanks
3.2.2.2 Washer hood vents
3.2.2.3 Evaporator condenser
3.2.2.4 Condensate stripping
3.2.2.5 Black liquor oxidation
3.2.2.6 TRS scrubbers
3.2.2.7 Incineration systems
3.2.2.8 Lime kiln incineration
3.2.3 Malfunctions
3.2.3.1 Digester relief systems
3.2.3.2 Digester blow system
3.2.3.3 Multiple-Effect evaporators
3.2.3.4 Black liquor oxidation
3.2.3.5 TRS vent control system
• 3.2.4 Inspection of pulping department
3.2.4.1 Digester
3.2.4.'2 Digester relief
53
53
57
60
63
65
70
70
72
72
72
74
78
81
81
84
91
97
99
99
101
101
104
104
104
104
104
107
107
107
110
111
111
111
114
115
iv
-------
CONTENTS (continued)
3.2.4.3 Brown stock washers
3.2.4.4 Multiple-Effect evaporators
3.2.4.5 Black liquor oxidation
3.2.4.6 Condensate stripping
3.3 Chemical Recovery
3.3.1 Recovery boiler
3.3.1.1 Process description
3.3.1.2 Sources of emissions
3.3.1.3 Control «
3.3.1.4 Boiler malfunctions
3.3.1.5 ESP malfunctions
3.3.1.6 Inspection of recovery boiler
3.3.2 Smelt dissolving tank
3.3.2.1 Process description
3.3.2.2 Sources of emission and control
3.3.2.3 Malfunctions
3.3.2.4 Inspection of the smelt dissolving tank area
3.4 Causticizing Department
3.4.1 Process description
3.4
3.4
3.4.1.3
3.4.1.4
1.1
1.2
Green liquor preparation
White liquor preparation
Lime mud washing
Calcining
3.4.2
3.4.3
3.4.4
3.4.5
Emission sources
Control
Malfunctions
Inspection procedures
3.5 Power Boilers
3.5.1 Process description
3.5.1.1 Gas- and oil-fired boilers
3.5.1.2 Coal-fired power boilers
3.5.1.3 Wood-fired power boilers
Page
115
115
122
122
122
122
126
134
145
164
167
183
208
208
210
215
215
223
223,
225
225
227
227
228
229
233
236
243
244
246
247
249
-------
CONTENTS (continued)
3.5.2 Sources of emissions
3.5.2.1 Gas-fired boilers
3.5.2.2 Oil-fired boilers
3.5.2.3 Coal-fired boilers
3.5.2.4 Bark boilers
3.5.3 Control techniques
3.5.3.1 Gas-fired boilers
3.5.3.2 Oil-fired boilers
3.5.3.3 Coal-fired boilers
3.5.3.4 Bark boilers
3.5.4 Malfunctions
3.5.4.1 Mechanical collectors
3.5.4.2 Scrubbers
3.5.4.3 Fabric filters
3.5.4.4 Electrostatic precipitators
3.5.5 Inspection of power boilers
3.5.5.1
3.5.5.2
3.5.5.3
3.5.5.4
3.5.5.5
3.6 Other Sources
Opacity
Transmissometer data
Boiler operating conditions
Flue gas volume
Control equipment inspections
3.6.1 Bleach plant
3.6.1.1 Process description
3.6.1.2 Sources of emission and control
3.6.1.3 Malfunctions
3.6.1.4 Inspection of bleach plants
3.6.2 Raw material handling systems
3.6.2.1 Process description
3.6.2.2 Sources of emissions and control
3.6.2.3 Malfunctions
3.6.2.4 Inspection
Page
249
249
249
250
252
252
252
253
253
255
255
261
264
267
270
274
275
275
276
279
281
285
285
285
293
294
294
295
295
295
297
298
VI
-------
CONTENTS (continued)
4. Compliance Determinations
4.1 Establishing a Baseline
4.1.1 Recovery boiler
4.1.2 Smelt tank
4.1.3 Lime kiln
4.1.4 Slaker
4.1.5 Turpentine condenser and multiple-effect
evaporators
4.1.6 Blow tank and hot water accumulator
4.2 Calculation of Emission Rates
4.2.1 TRS sources
4.2.2 Emissions from recovery boilers
4.2.3 Power boilers
4.3 Stack Test Methods
4.3.1 Particulate sampling
4.3.2 TRS sampling
4.3.3 SOp sampling
4.3.4 Visible emissions
Appendix A Summary of State regulations
Appendix B EPA Reference Methods 1-5, 17
Appendix C EPA Reference Methods 16, 16A
Appendix D EPA Reference Method 6
Appendix E EPA Reference Method 9
Page
307
309.
312
312
312
315
315
315
318
318
318
319
322
322
323
324
324 '
A-l
B-l
C-l
D-l
E-l
vii
-------
FIGURES
Number
" v
1-1 Geographic Location of Kraft Pulp Mills
2-1 Checklist for Obtaining Information During a File Review
2-2 Wet Scrubber Inspection Data Sheet
2-3 Mechanical Collector Inspection Data Sheet
2-4 Electrostatic Precipitator Inspection Data Sheet
2-5 Fabric Filter Inspection Data Sheet
3-1 Kraft Pulping Process
3-2 Drum Debarking
3-3 Bag Debarker
3-4 Ring Barker
3-5 Cutterhead Barker
3-6 Hydraulic Barker
3-7 Knife Barker
3-8 Norman Disk Chipper
3-9 Drum Chipper
3-10 Horizontal Parallel Chipper
3-11 Vibratory Chip Screen
3-12 Gyratory Chip Screen
3-13 Hydraulic Dump Truck for Chips
3-14 End-dump Freight Car and Unloading Platform for Chips
Page
8
23
39
4.3
45
49
54
56
56
57
58
59
59
61
62
62
64
65
66
67
vm
-------
FIGURES (continued)
Number
3-15 Rotary Freight Car Dryer for Chips (Link Belt Limited)
3-16 Typical Material and Air Flow for Small Woodyard
3-17 Pulping Process
3-18 Typical Digester Operating Curves
3-19 Typical Digester Charging Room Floor
3-20 Hot Water Accumulator Showing Primary and Secondary
Condensers
3-21 Typical Batch Digester Steam Flow Rate During Blow
3-22 Worksheet for Calculation of Blow Weight (Steam)
3-23 Digester Relief and Turpentine Recovery System
3-24 Odor Compounds in Relief Gas after Turpentine Condenser
as a Function of Condenser Outlet Temperature
3-25 Continuous Digester Flow Sheet
3-26 Vacuum Washer Flow Sheet
3-27 Pressure Washers Flow Sheet
3-28 Diffusion Washer Flow Sheet
3-29 Multiple-Effect Long-Tube Vertical Evaporators
(Backward Feed)
3-30 Multi-Effect Vacuum Evaporation Plant Flow Sheet
3-31 Chart of Evaporator Temperatures
3-32 Contaminated Condensates Air Stripping Plant Flow Sheet
3-33 Contaminated Condensates Steam Stripping Plant Flow
Sheet
3-34 Stripping Effciency for Different Steam-Condensate
Ratios with 10 Theoretical Plants
3-35 Agitated Air Sparging System for Black Liquor Oxidation
Page
68
69
73
75
76
77
78
79
80
82
83
85
86
87
89
90
92
94
95
96-
98
ix
-------
FIGURES (continued)
Number Page
3-36 Champion Two Stage Unagitated Strong Black Liquor
Oxidation System 98
3-37 Vaporsphere Flow Equalization Gas Holders 102
3-38 Floating Cover Flow Equalization Gas Holders 102
3-39 Hotwell Gas Scrubber for 100 Metric Tons Per Hour (9400
gpm) Evaporation Plant for H«S-Separation of 95 Percent
or More c 105
3-40 Noncondensable Gas Incineration System 106
3-41 Kraft Batch Digester Blow Gas Flow After Condensing and
without Equalization 109
3-42 Cross Section of B&W Recovery Boiler 127
3-43 Difference in Air Systems in U.S. Recovery Boiler Designs 129
3-44 Cyclone Evaporator 130
3-45 Cascade Evaporator 130
3-46 Venturi Evaporator 131
3-47 Three Types of Indirect Contact Evaporators 133
3-48 Effect of Solids Firing Rate on Reduced Sulfur Emissions
and Steam Generation Efficiency 136
3-49 Effect of Black Liquor Solids Concentration 137
3-50 Effect of Black Liquor Heating Value 137
3-51 Bed Temperature as a Function of Total Air 139
3-52 Bed Temperature as a Function of Primary Air 139
3-53 Effect of Total Air Supplied to the Unit 140
3-54 Theoretical Loss of Particulate as a Function of
Percentage Increases in Primary Air 141
3-55 Effect of Primary Air Temperature 142
3-56 Effect of Sulfur-Sodium Ratio in the B;lack Liquor 143
3-57 Effect of Chlorine in Black Liquor 144
x
-------
FIGURES (continued)
Number Page
3-58 Basic Processes Involved in Electrostatic Precipitation 146
3-59 Typical Wet-Bottom ESP with Heat Jacket 148
3-60 Electrical Diagram for ESP T-R Set 149
3-61 Typical Weighted-Wire ESP with Drag Bottom 151
3-62 Rigid-Frame Design 152
3-63 Wet-Bottom ESP 153
3-64 Drag Chain Assembly 154
3-65 MIGI Rapper Cross Section 156
3-66 Internal Falling-Hammer Rapper Design 157
3-67 Design SCA and Efficiency of 20 Recovery Boiler ESP's 158
3-68 Superficial Velocity Versus Year Installed 159
3-69 ESP Instrumentation Diagram 161
3-70 Positions of Measuring Instruments 162
3-71 Typical Secondary Current Pattern for Unit Experiencing
Salt Cake Buildup 168
3-72 Example of a Plugged Distribution Plate 174
3-73 . Typical Rapper Layout on a Modern Two-Chamber Precipitator 177
3-74 Typical Pattern Generated by Insulator Tracking 180
3-75 Example of Severe Corrosion of Collection Plates 181
3-76 Example of 6-minute Average Opacity Pattern 188
3-77 Typical Opacity Monitor Output with Severe Rapping
Reentrainment Losses 189
3-78 Method of Calculating Additional Moisture in the Flue Gas
Stream Due to Direct-Contact Evaporator 196
3-79 Optimum Secondary Current Distribution in ESP Serving Kraft
Recovery Boiler, Assuming Uniform Rapping and Wire Size
in All Fields 200
xi
-------
FIGURES (continued)
Number
3-80
3-81
3-82
3-83
3-84
3-85
3-86
3-87
3-88
3-89
3-90
3-91
3-92
3-93
3-94
3-95
3-96
3-97
3-98
3-99
4-1
Secondary Current Pattern for Two ESP Chambers ; Chamber A
is Having Maintenance Problems that Limit Power Input
Equilibrium Diagram for a Na2C03-NA2S System
Smelt Dissolving Tank with Water Sprays
*
Smelt Dissolving Tank with Steam Shatter Jets
Wire Mesh Pad Used in Smelt Dissolving Tank Vents
Low-Energy Entrainment Scrubber for use on Smelt Dissolving
Tank Vent (Ducon Dynamic Scrubber)
Typical Causticizing Flow Diagram
Slaker-Classifier used in Typical Causticizing Plant
H9S Emission from the Lime Kiln Related to NA?S Level in
^the Lime Mud
HgS Emission Related to Percent 02 in the Flue Gas
HS Emission Related to the Moisture Content in the Lime
Coal Size Distribution for Firing in a Spreader Stoker
Plugged Inlet Vane
Plugged Outlet Tube
Plugged Collecting Tube
Fabric Filter Bag Attachment Methods
Fly Ash Resistivity Curve
Flow Diagram of a Three-Stage Bleach Plant: CEH
Cutaway of a Vacuum Washer with Short Drop Leg
Chlorine Dioxide Generating System: Solvay
Calculation of Kraft Recovery Boiler ESP Efficiency
Page
201
209
211
212
213
214
224
226
230
230
231
258
263
265
266
269
272
289
290
292
320
-------
TABLES
Number Page
1-1 Location of Kraft Pulp Mills in the United States 9
1-2 Comparison of Light Extinction Terms 15
2-1 Summary of Reported Human Health Effects of Hydrogen
Sulfide 33
2-2 Summary of Reported Human Health Effects of Inhalation
of Chlorine 34
3-1 Checklist for Inspection of Wood Handling Systems 71
3-2 Typical Digester Liquor Requirements 74
3-3 Main Components of Typical Kraft Mill Condensates 93
3-4 Typical Kraft Mill Condensate Compositions, Mean Values
for 10 Mills 93
3-5 Summary of TRS Control Options for Pulping Department 99
3-6 Gas Flow Rates from Batch Digester 100
3-7 Typical Ranges of Digester Noncondensable Gas Flow Rates 100
3-8 Flammability Limits in Air for Kraft Noncondensable Gases 103
*
3-9 Flame-Spreading Velocities of Air-Mercaptan Mixtures 103
3-10 Typical Ranges of Evaporator Noncondensable Gas Flow Rates 103
3-11 Malfunctions that may Occur in Digester Relief Turpentine
Recovery Systems 108
3-12 Malfunctions that may Occur in Digester Blow Tank Hot .
Water Accumulator Systems 110
3-13 Malfunctions that may Occur in Multiple-Effect Evaporator
Systems 112
xiii
-------
TABLES (continued)
Number
3-14 Malfunctions that may Occur in Black Liquor Oxidation
Systems
3-15 Malfunctions that may Occur in Noncondensable Gas
Incineration System
3-16 Checklist for Inspection of Digester Blow Systems
3-17 Checklist for Inspection of Digester Relief Systems
3-18 Checklist for Inspection of Brown Stock Washer Systems
3-19 Checklist for Inspection of Multiple-Effect Evaporator
Systems
3-20 Checklist for Inspection of Black Liquor Oxidation
Systems
3-21 Checklist for Inspection of Condensate Stripping Systems
3-22 Summary of the Effects of Key Recovery Boiler Operating
Parameters
3-23 Recovery Boiler Operating Parameters to be Recorded
During Performance Tests or Inspections
3-24 Parameters to be .Measured by the Inspector During Level III
Inspection of Recovery Boiler ESP
3-25 Malfunctions that may Occur in Smelt Tank Particulate
Control Systems
3-26 Typical Lime Kiln Mass Balance
3-27 Power Boiler Operating Parameters to be Recorded During
Performance Tests or Level III Inspections
3-28 Common Letter Designations used for Bleach Agents
3-29 Bleaching Sequences for Sulfate Pulp
3-30 Bleaching Sequences for Hardwood Sulfate Pulp
3-31 Inspection Checklist for use in Bleach Plant
3-32 Inspection Checklist for'Material Handling Systems
4-1 Summary of the Effects of Recovery Boiler and ESP Operating
Parameters on Particulate and TRS Emission Rates
Page
11:3
113
116
118
119
120
123
125
184
191
207
215
228
277
287
288
291
296
299
313
xiv
-------
Number
4-2
4-3
4-4
4-5
TABLES (continued)
Summary of the Effects of Smelt Tank and Venturi Scrubber
Operating Parameters on Particulate and TRS Emission Rates
Summary of the Effects of Lime Kiln and Scrubber Operating
Parameters on Particulate and TRS Emission Rates
Summary of the Effects of Shaker and Venturi Scrubber
Operating Parameters on Particulate Emission Rate
Summary of the Effects of Turpentine Condenser and Multiple-
Effect Evaporator Operating Parameters on TRS Emission
Rate
' Page
314
316
317
317
xv
-------
ACKNOWLEDGMENT
This report was prepared for the U.S. Environmental Protection Agency by
PEDCo Environmental, Inc.t Cincinnati, Ohio. Mr. Robert Marshall was the EPA
Task Manager. Mr. Thomas Ponder served as the Project Director, and Mr. Ronald
Hawks was the Project Manager. The principal authors were Mr. Ronald Hawks,
Mr. Gary Saunders, Mr. Douglas Orf, and Mr. David Dunbar.
xvi
-------
EXECUTIVE SUMMARY
The purpose of this inspection guide is to provide the necessary technical
information and data to support State and local inspectors in the evaluation
of both new and existing kraft pulp mills with respect to continuing compliance.
The guide includes a great deal of information on the overall operation of
the many processes in a kraft pulp mill. This detailed process information
is presented to provide the inspector with enough information to allow him or
her to be conversant with plant personnel regarding the major aspects of a
given source or process. By having this basic understanding of each major
process, the inspector will be able to ask informed questions of plant per-
sonnel and to evaluate the source's program of continuous compliance.
In addition to providing detailed process information, the guide also
provides information on the various malfunctions that can occur within each
process and the affect that these malfunctions will have on the overall
emissions from the process and in many cases the ability of the mill to
continuously comply with the applicable Federal, State, or local regulations.
In addition to the discussion of malfunctions presented in each of the major
process subsections in Section 3, the reader is referred to the following
tables that provide a summary of the potential malfunctions associated with
each major process and the effect that these malfunctions may have on the
operation of the mill and the emissions from the given process:
Table
3-11
3-12
3-13
3-14
3-15
3-25
Process
Digester Relief Turpentine Recovery System
Digester Blow Tank Hot Water Accumulator
System
Multiple-Effect Evaporator Systems
Black Liquor Oxidation Systems
Noncondensable Gas Incineration System
Smelt Tank
Page no.
108
110
112
113
113
' 215
xvn
-------
Because the purpose and scope of State and local agency inspections vary
and the agency manpower is limited, the guide presents a discussion of four
levels of compliance inspections. Each level is increasingly more complex in
terms of the information that should be obtained and the evaluations to be
conducted. Although the increasing complexity of each inspection level
naturally increases the time required to perform each inspection, the benefits *
of the more complex level III inspections are substantial because the cause
of most,malfunctions and the lack of continuous compliance are not easily
identified by Level I or II Inspections.. Briefly, Level I Inspections only
consist of visual emission evaluations. Level II Inspections are more detailed
and require the inspector to document some of the process operating parameters
that determine the allowable emission rate. Basically this level of inspection
is commonly referred to as a "walk-through inspection" because the inspector
walks through the plant without physically measuring any parameters or conditions.
He or she does, however, record any data that may be available as a result o.f
plant instrumentation on the rates of the major pieces of process equipment,
the type and generating characteristics of the control equipment, major
maintenance activities, the type and quality of fuel-consumed, etc.
Level III inspections require that the inspector actually conduct some
of his or her own measurements. Therefore, he or she should have equipment
for measuring flue gas temperature, oxygen content and velocity, scrubber and
fabric filter pressure drop, and fan motor current and revolutions per minute.
The inspector may use the data he or she collects to calculate gas volumes
through the control devices and to compare these values with design values.
In general during a Level III inspection, the inspector should record infor-
mation from plant instruments that monitor process and control device operating
conditions including black liquor properties, black liquor firing rates, steam
flow, steam pressure, furnace drafts, and such control device parameters as
power input level, pressure drop, and flue gas velocity.
Level IV Inspections are similar to Level III Inspections except that
the former is conducted during a performance compliance stack test. The in-
formation obtained during a Level IV Inspection is used to produce a compara-
tive baseline for future Level III Inspections so that relationship between
certain parameters and emissions can be established to permit both inspector
and the plant personnel to monitor certain parameters to determine if the
source is complying with the applicable regulation.
xviii
-------
To assist the inspector in conducting the inspection, the guide presents
a series of checklists. The checklists for the most part are used with Level
III Inspections. The general information listed at the top of the checklist
can, however, be used for a Level I Inspection if the basic information on
the source is not already available. The checklist can also be used for
Level II Inspections where the plant has a considerable amount of instrumen-
tation available for monitoring various process or control equipment parameters.
The reader is referred to the following checklists or data sheets that help
to identify the type of information that should be obtained during a Level II
or III Inspection:
Figure Item Page
2-1 File Review 23
2-2
2-3
2-4
2-5
Table
3-1
3-16
3-17
3-18
3-19
3-20
3-21
3-23
3-24
3-27
3-31
3-32
Wet Scrubber
Mechanical Collector
Electrostatic Precipitator
Fabric Filter
Item
Wood Handling Systems
Digester Blow Systems,
Digester Relief Systems
Brown Stock Washer Systems
Multiple-Effect Evaporator Systems
Black Liquor Oxidation Systems
Condensate Stripper Systems
Recovery Boiler
Recovery Boiler Electrostatic Precipitator
Power Boiler
Bleach Plant
Material Handling System
39
43
45
49
Page
71
116
118
119
120
123
125
191
207
277
296
299
xix
-------
The guide also provides (Section 4) a detailed discussion on what the
inspector should do with all the information obtained during the inspection.
Procedures are presented on how to use the parameters that have been recorded
in conducting the necessary calculations and how the field observations can
be used to make judgements about continuous compliance. Certain process
parameters can be used to monitor compliance of specific sources between
compliance tests. The reader is referred to the following tables that provide
information on the effects of various parameters on the overall operation of
the kraft pulp mill, the particulate.and TRS emissions, and the ability of
the source to maintain continuous compliance.
Table Process page
4-1
4-2
4-3
4-4
4-5
Recovery Boiler and Electrostatic Precipitator 312
Smelt Tank and Venturi Scrubber 313
Lime Kiln and Scrubber 315
Slaker and Venturi Scrubber 315
Turpentine Condenser and Multiple-effect
Evaporator 315
In general, the particulate and S02 emission rates from the recovery
boiler are interrelated because the primary method of S02 and TRS control is
to convert these pollutants to sodium sulfate, which increases the particulate
emission rate. The primary recovery boiler parameters that affect particulate
emissions are: firing rate, primary air rate, excess air, smelt bed tempera-
ture, ESP power, ESP superficial velocity, and flue gas oxygen. A shift in
many of these parameters indicates an increase in emissions.
The primary parameters that may be used to determine compliance from the
smelt dissolving tank are connected with the rate of particulate generated and
the condition of the control devices. Specifically the rate of generation of
particulate is related to smelt rate (i.e., boiler firing rate and reduction
efficiency) and the amount of particle reentrainment. The condition of the
control device is related to such variables as superficial velocity and water
flow rate.
Uncontrolled particulate emission rates from the kiln are primarily
affected by parameters that affect the superficial velocity through the kiln,
xx
-------
the particle size of the kiln dust, rate of evolution of volatile particulate,
and the feed rate to the kiln. The superficial velocity is a function of kiln
firing rate and temperature profile. The rate of evolution of volatile par-
ticulate is related to slurry feed rate and the amount of soda present in the
slurry. Parameters affecting the control device are liquid-to-gas ratio,
pressure drop, and particle size. TRS emissions are related to mud washing
efficiency (% sodium sulfide expressed as Na20), flue gas oxygen, and lime
mud slurry moisture. The cold end temperature has an effect on TRS emission
levels. Both the kiln excess air and temperature profile down the kiln in-
fluence the residence time and oxidation rate of TRS compounds where the kiln
is used as a control device.
The rate of green liquor and calcium oxide reacted in the slaker has the
i
strongest effect on uncontrolled particulate emissions. The amount of heat
released (i.e., steam generated) and the degree of agitation are related to
the reaction rates. The condition of, the scrubber (i.e., water flow rate,
liquor gas ratio, and pressure drop) also affects the emission rate.
The rate of TRS emissions from multiple-effect evaporators and the
turpentine condenser is primarily a function of noncondensable gas volume
and tail gas condenser final temperature. The rate of TRS emissions from
the hot water accumulator is a function of digester operation, the condition
of the primary and secondary condensers, and blow gas volume. The parameters
are generally so interrelated that a single parameter analysis is not effec-
tive in predicting emissions. Generally, however, TRS emissions will increase
if the condensers are plugged.
Finally, the guide presents a detailed discussion of establishing a
baseline that involves documenting all pertinent operating parameters as they
relate to the emission characteristics of the mill. This includes both
process and control equipment parameters.
The baseline may be used for several.purposes. First, for existing
sources, baseline values may be obtained prior to a stack test to assist in
establishing representative operating conditions. The normal range of values
may be recorded during a period prior to a test, and these values may be
specified in a testing protocol to establish representative conditions or
used as a starting point in negotiating the testing protocol with the plant.
Comparison of documented compliance test parameters with those specified in
xxi
-------
the protocol helps to establish whether the process and control equipment
were operating at the specified representative conditions. Second, for new
sources, the initial compliance test establishes the operating parameter
values that correspond with the measured emission rate. These values can
then be compared with the design values. This provides a fixed reference
point for comparison to future operating data. Third, the values of the
baseline parameters provide data for evaluating routine inspection data. By
knowing the effects of the various process and control equipment parameters
on emissions, one can make comparisons to evaluate the direction and magnitude
of any changes in performance. Fourth, documentation of the baseline data
will assist in setting specific ranges on important parameters for possible
inclusion in an operating permit (if required by the agency). Finally, the
baseline test provides a fixed reference point for comparing long-term per-
formance trends. Proper evaluation of the baseline data may assist in the
establishment of preventive maintenance schedules as well as provide an indi-
cation of any design or installation problems. In addition, the rate at
which the normal operating parameters may vary from baseline values may assist
the agency in scheduling routine inspections and periodic compliance tests.
Baselining should only focus on those parameters that have a documented
affect on the emission levels rather than on all possible parameters that
might influence the emission levels. Collection of data that have no signifi-
cance can be inefficient and counterproductive. Considerable effort can be
involved in recording and analyzing all process and control equipment data
normally available at a facility. The inspector must be selective as to which
data to collect.
The use of the baseline for documenting deviations from normal conditions
requires the establishment of a logic system for each process or control
device operating parameter used. A substantial change in the parameter is
evaluated based on its impact on the overall emission levels.
Once a baseline has been established the data obtained during a Level III
Inspection are usually sufficient to allow the inspector to negotiate correc-
tive action with respect to the process and control equipment without the
expense of conducting a performance stack test. Many deficiencies may be
corrected as a result of increased or redirected maintenance activities. The
ability of inspectors to negotiate such corrective action varies from agency
xx ii
-------
to agency. The inspector must operate within his agency's guidelines with
regard to negotiating compliance agreements or issuing notices of violation.
In some cases the corrective action can be completed before the notices can
be drafted and formally issued.
xxi ii
-------
-------
SECTION 1
INTRODUCTION
Kraft pulp mills have\been and continue to be a major source of particu-
late matter, sulfur dioxide (S02), and total reduced sulfur (TRS) compound
emissions. Because the processes within a kraft pulp mill are numerous and
very complex, the potential emissions are a function of a number of interre-
lated process variables. Over the past several years, a number of concerns
have been raised regarding the ability of the various sources within the kraft
pulp mill to achieve continuous compliance.
1.1 PURPOSE AND SCOPE
This inspection guide is intended to provide the necessary technical in-
formation and data to support State and local inspectors in the evaluation of
both new and existing kraft pulp mills. The guide includes, among other
things, a brief process description of each emission source. This descrip-
tion is intended to provide the inspector with a basic understanding of the
technical aspects of each emission source or process. The process descrip-
tion includes the major steps and. substeps and operating conditions that can
have substantial impact on both uncontrolled and controlled emission rates.
It also should provide the inspector with enough information to allow him or
her to be conversant with plant personnel regarding the major aspects of a
given source or process. Having this basic understanding of each process
makes it easier for the inspector to ask informed questions of plant per-
sonnel or to seek more detailed data from the technical literature. This
guide also should provide the inspector with the necessary information to
support certain permit stipulations or more extensive engineering evaluations.
Certain portions of the guide provide information on those process parameters
that may be used to monitor compliance of specific sources between compliance
tests. Changes in these parameters may be used to document the need for more
extensive operation and maintenance (O&M) to ensure continued compliance. In
-------
those cases where a clearly defined cause and effect cannot be established
for a given source, certain changes in various operating parameters can be
used to support the need for a compliance test. This method of source evalua-
tion is commonly referred to as "baselining." The baseline is generally
established during a period of known compliance, typically during a compliance
stack test, and includes information on both the process conditions and key
control equipment operating parameters.
Through an understanding of potential malfunctions, the inspector often
can work with plant personnel to reduce both the frequency and duration of
excess emissions and to identify and correct those malfunctions that limit
production and cause excessive maintenance costs.
Based on the knowledge that the purpose and scope of agency inspections
vary and that agency manpower is limited, four levels of compliance inspections
have been developed. Each level is increasingly more complex in terms of the
evaluations that must be conducted. The increasing complexity of each level
naturally increases the time required to perform the inspection. The benefits
of the more complex inspections are substantial, however, because most mal-
functions and periods of noncompliance are not easily identified by Level I
or II Inspections.
The following subsections briefly describe the types of activities asso-
ciated with each level of inspection. More details regarding a given source
or process are presented in Section 3.
1.1.1 Level I
A Level I Inspection consists of a visual emission evaluation according
to procedures set forth in EPA Method 9—Visual Determination of the Opacity
of Emissions From Stationary Sources. The inspector compares the opacity with
local agency emission standards by using appropriate averaging times and any
exclusion periods that may be set forth in the State or local agency's regu-
lations.
1.1.2 Level II
A Level II Inspection is more detailed and requires the inspector to
document some of the process operating parameters that determine allowable
emission rates. The additional information includes capacity and operating
-------
rates of the major pieces of process equipment, the type and operating
characteristics of control equipment, major maintenance activities, the type
and quantity of fuel consumed, process flow diagrams, and types and quantities
of raw materials consumed. This level of inspection is generally referred
to as a "walk-through inspection" because the inspector just walks through
the plant without physically measuring any operating parameters or conditions.
If, however, oxygen monitors, temperature charts, and etc., are available,
these data should be recorded. Although the inspector does obtain whatever
data the plant has available on the operation of the source, which are valuable
in documenting compliance, this level of inspection may overlook serious mal-
functions or excess emission periods.
1.1.3 Level III s
The Level III Inspection is the most complex level of inspection that
is conducted as part of a routine inspection program. This level is more
complete and requires substantial time, both in terms of preparation and actual
execution of the inspection. The additional data acquired during a Level III
Inspection result in a more accurate assessment of Q&M practices. These data
are invaluable in developing baseline information and in determining source
compliance. Properly conducting a Level III Inspection requires that the in-
spector have equipment for measuring flue gas temperature, oxygen content,
and velocity; scrubber and fabric filter pressure drop; and fan motor current
and revolutions per minute.
The inspector may use the data he or she obtains along with any plant
data and information to calculate gas volumes through control devices and to
compare these flue gas values with the design values.
The inspector should also record information from plant instruments that
monitor process and control device operating conditions. These variables may
include black liquor properties (% solids, heat value), black liquor firing
rates, steam flow, steam pressure, and furnace drafts and such control device
parameters as power input level, pressure drop, and flue gas velocity.
Plant personnel routinely record operating conditions, and the inspector
should compare current conditions with historic values. Serious deviations
from established norms should be used to evaluate emissions trends or to
determine if a compliance test is warranted.
-------
1.1.4 Level IV
A Level IV Inspection is similar to a Level III Inspection except that
the former is conducted during a performance compliance stack test. The in-
formation obtained during this inspection is used to produce a comparative
baseline for future Level III Inspections.
1.2 CONTINUOUS COMPLIANCE
The Clean Air Act of 1970 required all States to prepare a state imple-
mentation plan (SIP) that set forth how the State intended to attain and main-
tain the National Ambient Air Quality Standards (NAAQS). Each SIP must con-
tain among other things the necessary legal authority and emission limitations
to ensure attainment and maintenance of the NAAQS.
The Clean Air Act of 1970 and the initial SIP's developed under that Act
placed primary emphasis on initial compliance of sources with a set of specified
emission limitations that reduced emissions sufficiently to attain the NAAQS.
Less attention and consideration were initially given in the formulation of
control strategies that would preserve or maintain the air quality once attain-
ment was achieved.
Control agencies have long recognized that initial compliance does not
necessarily mean future or continuing compliance. Faced with limited re-
sources, attainment dates mandated by legislation-, emission limits set by
regulations, and already established source compliance dates, the agencies
understandably placed a greater emphasis, on promoting initial source compliance
to ensure attainment of NAAQS. Now that .most sources have either demonstrated
initial compliance or are in the midst of a compliance-oriented program, con-
siderable concern has been raised within the air pollution control community
with respect to whether a sotfrce is operating and maintaining its control equip-
ment. Some concern has also been raised with respect to whether a source is
complying with the applicable emission limit on a continuous basis. In many
cases a source can fine tune its control system and make the necessary ad-
justments to comply with an emission limit during-a stack test conducted to
certify compliance with the applicable emission limit. Once these tests have
been completed, however, the control system may begin to deteriorate and the
source may no longer be in compliance with the applicable emission limit.
-------
Reasons for the possible deterioration of the control system include lack
of good O&M procedures, poor or virtually no maintenance, poor design, lack of
understanding on the part of the control equipment operator, lack of reliable
instrumentation, poor recordkeeping or little or no evaluation of the records
that are kept, or lack of any desire to make the control equipment operate
properly.
State and local agency officials are deeply concerned with the lack of
^
continuous compliance because of its potential impact on the ability of the
State or local agency to attain and maintain the NAAQS. As a result, many
State and local agencies are looking for ways to improve their existing
surveillance, inspection, and enforcement programs to encourage sources; to
properly operate and maintain their control equipment; to maintain adequate
records and to use these records to avoid significant operating problems; and
to continuously comply with all applicable emission limits and visible emis-
sion standards.
1.3 ORGANIZATION OF INSPECTION GUIDE
The inspection guide is basically divided into three major sections.
Section 2 outlines the activities that must take place before the inspector
enters the plant. The inspector should review the files that the agency has
compiled on the types and quantities of various process equipment, type and
sizes of control equipment, results of emission stack tests, and information
from previous inspections. These files also should contain a history of com-
plaints, malfunctions, visible emission evaluations, and previous compliance
status. This section provides detailed checklists to assist the inspector in
obtaining the necessary information from the files.
Section 2 also contains information concerning various safety precautions
that need to be taken during the inspection and describes the type of protec-
tive clothing the inspector should wear. As indicated in the description of
Level III Inspections, a variety of process monitoring equipment is necessary
to obtain operating parameters for use in compliance determination. This sec-
tion describes each piece of monitoring equipment and its purposes.
Section 3 is by far the largest and the most useful in terms of present-
ing information that the inspector can use during the actual inspection. This
-------
section is divided according to the following six major processes or systems
within a kraft pulp mill:
o Woodhandling
o Pulping
Chemical recovery
Causticizing
Power boilers
o
o
o
Other sources
A description and process flow sheet are provided for each of these major
systems. Because kraft pulping is a complex operation, some of the process
descriptions are quite detailed. This detail is necessary to provide the
inspector with a sufficient understanding of the process for him or her to
communicate effectively with plant personnel.
As noted earlier, the purpose of the compliance inspection is to obtain
information that will allow estimation of the level of atmospheric emissions.
Because many factors affect emissions, Section 3 also includes the major
chemical reactions associated with each major source of emissions and the
techniques used to control these emissions. Typical ranges of emissions are
provided, along with collection efficiencies, flow rates, temperatures, pressure
drops, power levels, etc., for the control equipment.
Although systems are designed to operate in a particular manner, frequent
malfunctions can occur and can lead to improper equipment operation and,
utlimately, to control equipment deterioration and excess emissions. As a
result, common malfunctions and their causes, symptoms, and results are dis-
cussed for each major process or system within the kraft pulp mill.
Section 3 also describes the procedures to be used during the inspection
of specific process equipment and associated control devices. Where appropriate,
checklists and tables indicating typical levels of operation are provided.
The specific items that need to be observed and the specific types of infor-
mation that need to be recorded for each process and control device are also
provided as a further aid to the inspector.
Section 4 sets forth what the inspector should do with all of the infor-
mation gathered during the inspection. This section also outlines how to use
-------
the parameters that have been recorded in conducting the necessary calcula-
tions and how the field observations can be used to make judgment decisions.
The table of contents for this inspection guide has been expanded to per-
mit the inspector to easily locate the necessary information regarding the
various processes within the kraft pulp mill, the control devices used to
minimize the emissions, the inspection procedures to be used for the partic-
ular process or control device, the parameters that affect emissions, and the
malfunctions that may occur. In addition, the inspector should also consult
the rather extensive list of figures and tables that provide photographs or
sketches of process and control equipment, inspection checklists, and other
valuable information that will aid in the inspection of the source and the
subsequent evaluation of the overall source's performance with respect to
continuous compliance.
1.4 INDUSTRY OVERVIEW
In 1979, 25 sulfite mills in the United States produced a total of
1.8 x 106 tons of pulp, 50 neutral sulfite semichemical (NSSC) mills produced
a total of 4.1 x 106 tons of pulp, and 121 kraft mills (located in 28 states)
fi
produced in excess of 38 x 10 tons of pulp. Figure 1-1 shows the relative
location of the 121 kraft mills by State and Table 1-1 lists the name and
exact location of these mills.
The type of pulping technique used depends on the end product the type
of wood available, and the general economics of the situation. The three
major pulping techniques are described briefly. The sulfite technique generally.
uses limestone and sulfur to produce the digestion liquor. Liquid sulfur is
burned to form S02» which is cooled and passed through a limestone-packed
tower. Because of its low cost, it is not necessary to recover the chemicals
from the liquor when calcium (limestone) is used. A significant water pollu-
tion occurs, however, when the chemicals are discharged. If sodium, ammonium,
or magnesium bases are used, however, some byproduct recovery is practical for
economic reasons. This pulping technique is simple compared with that required
for kraft pulping. The pulp produced by the sulfide or acid pulping technique,
which is light in color, is used for fine paper and tissue. Further processing
2
of the pulp yields rayon or cellulose acetate.
-------
00
NUMBER OF HILLS/STATE:
AlABAW . U NUUANO . !
ARIZONA . 1 MICHIGAN -t
ARKANSAS - C MINNESOTA - t
CALIFORNIA • 4 MISSISSIPPI - 4
FIWIM . S MONTANA - t
GEORGIA . t|
IDAHO . 1
KENTUCKY - t
UMISIMM II
NUDE - 7
HEM HAMPSHIRE - I
HEM TOOK . I
MOTH CAROLINA - S
OHIO - Z
OKLAHOMA - I
OREGON - J
KHNSYIVMIA . ]
SOUTH CAROLINA - 4
TENNESSEE - 2
TEXAS - i
VIRGINIA - 4
HASHINGTCH - 7
WISCONSIN . 4
Figure 1-1. Geographic location of kraft pulp mills.
-------
TABLE 1-1. LOCATION OF KRAFT PULP MILLS IN THE UNITED STATES
Company name
Location
Allied Paper
James River-Dixie
Northern, Inc.
Champion
Container Corp.
Alabama Kraft
Gulf States
Alabama River Pulp
Hammermill
International Paper
Kimberly Clark
MacMillan Bloedel
Scott
Union Camp
Southwest Forest
Georgia Pacific
Great Northern
Green Bay
International Paper
International Paper
Weyerhaeuser
Crown Simpson
Fiberboard (Louisiana Pacific)
Louisiana Pacific
Simpson Lee
Alton Box
Container Corp.
Georgia Pacific
Southwest Forest Industries
Procter & Gamble
St. Joe Paper Company
(continued)
Jackson, Alabama
Butler, Alabama
Courtland, Alabama
Brewton, Alabama
Mahrt, Alabama
Demopolis, Alabama
Claiborne, Alabama
Selma, Alabama
Mobile, Alabama
Coosa Pines, Alabama
Pine Hill, Alabama
Mobile, Alabama
Montgomery, Alabama
Snowflake, Arizona
Crossett, Arkansas
Ashdown, Arkansas
Morrilton, Arkansas
Camden, Arkansas
Pine Bluff, Arkansas
Pine Bluff, Arkansas
Fairhaven, California
Antioch, California -
Samoa, California
Anderson, California
Jacksonville, Florida
Fernandina Beach, Florida
Palatka, Florida
Panama City, Florida
Foley, Florida
Port St. Joe, Florida
-------
TABLE 1-1 (continued)
Company name
Location
St. Regis
St. Regis
Continental Forest
Continental Forest
Brunswick
Georgia Kraft
Georgia Kraft
Gilman
Great Northern
Interstate
ITT Rayonier
Owens-Illinois
Union Camp
Potlach
Western Kraft
Westvaco
Boise Cascade
Boise Southern
Continental Forest
Crown Zellerbach
Crown Zellerbach
Georgia Pacific
International Paper
International Paper
01 in
Pineville
Western Kraft
Diamond International
Georgia Pacific
International Paper
Lincoln
Scott Paper
(continued)
Jacksonville, Florida
Pensacola, Florida
Augusta* Georgia
Port Wentworth, Georgia
Brunswick, Georgia
Krannert, Georgia
Macon, Georgia
St. Mary, Georgia
Cedar Springs, Georgia
Riceboro, Georgia
Jesup, Georgia
Valdosta, Georgia
Savannah, Georgia
Lewiston, Idaho
Hawesville, Kentucky
Wickliffe, Kentucky
DeRidder, Louisiana
Elizabeth, Louisiana
Hodge, Louisiana
Bogalusa, Louisiana
St. Francisville, Louisiana
Port Hudson, Louisiana
Bastrop, Louisiana
Springhill, Louisiana
West Monroe, Louisiana
Pineville, Louisiana
Campti, Louisiana
Old Town, Maine
Woodland, Maine
Jay, Maine
Lincoln, Maine
Skowhegan, Maine
10
-------
TABLE 1-1 (continued)
Company name
Location
Oxford
S.D. Warren
Westvaco
Mead
Scott
Boise Cascade
Potlach
International Paper
International Paper
International Paper
St. Regis
Champion *
Brown
International Paper
Champion
Federal
Champion
Weyerhaeuser
Weyerhaeuser
Grief
Mead
Weyerhaeuser
American Can
Boise Cascade
Crown Zeller bach
Georgia Pacific
International Paper
Western Kraft
Weyerhaeuser
Rumford, Maine
Westbrook, Maine
Luke, Maryland
Escanaba, Michigan
Muskegon, Michigan
International Falls,
Minnesota
Cloquet, Minnesota
Moss Point, Mississippi
Natchez, Mississippi
Vicksburg, Mississippi
Monticello, Mississippi
Missoula, Montana
Berlin-Gorham, New
Hampshire
Ticonderoga, New York
Canton, North Carolina
Riegelwood, North Carolina
Roanoake Rapids, North
Caroli na
New Bern, North Carolina
Plymouth, North Carolina
Massillon, Ohio
Chillicothe, Ohio
Valliant, Oklahoma
Halsey, Oregon
St. Helens, Oregon
Clatskanie, Oregon
Toledo, Oregon
Gardinier, Oregon
* Albany, Oregon
Springfield, Oregon
(continued)
-------
TABLE 1-1 (continued)
Company name
Location
Appleton
P.H. Glatfelter
Penntech
Bowater
International Paper
South Carolina
Westvaco
Bowater
Packaging
Champion
International Paper
Owens-Illinois
Southland
Southland
Tempie-Fostex
Chesapeake
Continental
Union Camp
Westvaco
Boise Cascade
Crown Zellerbach
Crown Zellerbach
Longview
St. Regis
Weyerhaeuser
Weyerhaeuser
Consolidated
Great Northern
Hammermill
Mosinee
Roaring Springs,
Pennsylvania
Spring Grove, Pennsylvania
Johnsonburg, Pennsylvania
Catawba, South Carolina
Georgetown, South Carolina
Florence, South Carolina
Charleston, South Carolina
Calhoun, Tennessee
Counce, Tennessee
Pasadena, Texas
Texarkana, Texas
Orange, Texas
Houston, Texas
Lufkin, Texas
Evadale, Texas
West Point, Virginia
Hopewell, Virginia
Franklin, Virginia
Covington, Virginia
Wallula, Washington
Camas, Washington
Port Townsend, Washington
Longview, Washington
Tacoma, Washington
Everett, Washington
Longview, Washington
Wisconsin Rapids, Wisconsin
Nekoosa, Wisconsin
Kaukauna, Wisconsin
Mosinee, Wisconsin
12
-------
The NSSC technique uses sodium sulfite and sodium carbonate or bicarbonate
as digestion chemicals. The cooking chemicals are formed by burning sulfur to
form S02 and contacting the gas stream with sodium carbonate in an absorption
tower to yield sodium sulfite. Because the cooking chemicals are usually not
recovered, their disposal becomes a water pollution problem. At best, only
partial recovery of the organics is possible, despite the availability of
several techniques for this purpose unless the NSSC technique is part of a
kraft pulp mill where the NSSC liquor is often recovered in the kraft process.
The pulp, quality and pulp yield are very high. The pulp is used primarily for
a corrugating medium because the fibers have a very high crush strength.
Hardwoods are primarily used to obtain this type of pulp. The NSSC process
does not produce odors because of the lack of sulfide ions in the cooking
liquor. The principal emission from this technique is SC^-
The kraft technique, which will be explained in detail in Section 3, re-
generates the white liquor used for digestion because of the expense of the
cooking chemicals. Although most species of wood can be pulped, the kraft
process generally uses soft woods. The fiber produced is dark in color, but
it can be used for fine white papers after the pulp is bleached. It is the
3
most widely used pulping technique, despite its distinctively odorous dis-
charge.
1.5 REGULATION UNDER THE CLEAN AIR ACT
The Clean Air Act of 1970 gave the EPA the responsibility and authority
to control air pollution in the United States and its territories. One of the
responsibilities included under Section 109 of the Act was the promulgation of
National Ambient Air Quality Standards (NAAQS). Section-110 of the Act re-
quired the States to adopt and submit to EPA their plans for attaining and
maintaining the NAAQS in all regions of the State. Thus, each State had to
decide which existing emission sources should be controlled and to what extent.
In addition, Section 111 of the Clean Air Act gave EPA the authority to
develop performance standards for new stationary sources. The New Source
Performance Standards (NSPS) established at a national level apply to both new
and modified sources. The NSPS must reflect the degree of emission reduction
achievable through the application of the best system of continuous emission
13
-------
reduction that the EPA Administrator (taking into consideration the cost of
achieving such emission reduction, and any nonair quality, health,, and en-
vironmental impact and energy requirements) determines has been adequately
demonstrated for a source category.
Section lll(d) of the Clean Air Act also gives EPA the authority to
establish a procedure similar to that provided by Section 110 under which each
State shall submit to EPA a plan that establishes standards of performance for
any existing source for any air pollutant for which air quality criteria have
not been issued or which is not included on a list published under Section
108(a) or 112 but to which a standard of performance would apply if such
existing source were a new source.
In addition to limiting the mass emission rate, State and local regula-
tions also limit plume opacity for certain sources. Allowable plume opacities
vary with the jurisdictions. Although a maximum allowable opacity of 20 percent
is common, nearly all jurisdictions allow for periodic excursions under cer-
tain conditions. In most States the visible emission regulations are general
in wording and broad in terms of overall applicability. Although some States
or local agencies have identified specific source categories to which a ;
particular opacity standard would apply, the opacity requirement in most
States applies to practically all stationary sources. The most common termi-
nology for applicability is "all existing sources." Other terms include
"existing equipment," "existing facilities," "stationary sources," "process
operations," and "existing installations." The use of "all existing sources"
essentially makes all sources subject to the requirement. Although a few
States have opacity requirements that specifically apply to fuel-burning
sources or kraft pulp mills, most States have a general opacity requirement
that applies to all sources, including fuel-burning sources.
Visible emission regulations vary widely across the United States. The
visible emission or opacity requirements are almost evenly divided between
the Ringelmann chart and opacity. Some State or local agencies specify only
the Ringelmann number; others specify the Ringelmann number and an equivalent
opacity limit; and still others specify only an opacity limit. Light
transmittance (incident light flux/light flux leaving the plume), opacity,
4
and Ringelmann are compared in Table 1-2.
14
-------
TABLE 1-2. COMPARISON OF LIGHT EXTINCTION TERMS
Light transmittance, %
0
0.20
0.40
0.60
0.80
1.00
Plume opacity, %
100
80
60 .
40
-'. 20
0
Ringelmann
number
5
4
3
2
1
0
Most State and local agencies allow visible emissions to exceed a pre-
scribed standard for some finite period of time, usually 3 to 5 minutes in
any one hour, although a few agencies allow an excursion for 6 to 8 minutes
in any one hour. A few agencies define the excursion (or exception) as the
total minutes in a day that the standard may be exceeded. Only a very few
agencies do not allow any excursion above the prescribed standard.
The one exemption that applies almost universally to opacity standards
is the exclusion of uncombined water vapor from the opacity reading. This
exemption is usually worded as follows, "where the presence of uncombined
water is the only reason for failure of an emission to meet the requirements,
A
such sections of the opacity requirements shall not apply."
1.5.1 State Implementation Plans
Pursuant to the Clean Air Act, each State must adopt and submit to EPA a
plan that provides for attainment and maintenance of the NAAQS in all areas
of the State. The State Implementation Plan (SIP) must include emission
limitations, schedules, timetables, and any other measures that may be necessary
to ensure attainment and maintenance of the NAAQS. Each State determines the
mix of emissions limitations that would be applicable to the sources within
the State.
Because a kraft pulp mill is composed of several sources that emit particu-
late, S02, and TRS, a number of regulations within a given State may be applica-
ble to some sources within a kraft pulp mill. Table A-l in Appendix A sum-
marizes the applicable regulations for selected States where kraft pulp mills
are known to be located.
15
-------
A review of Table A-l indicates that certain sources (such as lime kilns,
smelt tanks, recovery boilers, and power boilers) generally are covered by
specific regulations that apply to these processes, whereas other processes
are covered under the general S02 or particulate process regulations. In
addition to subjecting processes to particulate and S02 requirements, some
States also require that TRS to be limited to so many pounds per ton of air-
dried pulp.
1.5.2 Federal Standards of Performance for New Sources
The Clean Air Act requires that EPA develop standards of performance for
new stationary sources of significant air pollution. These standards, commonly
known as NSPS, are based on the best system of continuous emission reduction
that has been adequately demonstrated, taking into account such nonair-quality
impacts as economics and energy. It should be noted that these regulations
take the form of standards—not just emission limits. Thus, an NSPS provi-
tion may require monitoring, process modification, or even specific emission
reduction methods.
The applicable NSPS under Subpart BB—Standards of Performance for
Kraft Pulp Mills—apply to the following affected facilities within a kraft
pulp mill: digester system, brown stock washer system, multiple-effect evapo^
rator system, black liquor oxidation system, recovery furnace, smelt dissolving
tank, lime kiln, and condensate stripper system. In pulp mills where kraft
pulping is combined with neutral sulfite semichemical pulping, the provisions
of this subpart are applicable when any portion of the material charged to an
affected facility is produced by the kraft pulping operation. The NSPS for
-kraft pulp mills indicates the following:
Process
Recovery furnace
Smelt dissolving tank
Lime kiln
STANDARD OF PARTICULATE MATTER
Limit
0.10 g/dscm (0.044 gr/dscf) 35% opacity
0.1 g/kg (0.2 Ib/ton) black liquor solids
0.15 g/dscm (0.067 gr/dscf) corrected to 10% 02
when gaseous fuel is burned
0.30 g/dscm (0.13 gr/dscf) corrected to 10% 02
when liquid fossil fuel is burned
16
-------
STANDARD FOR TRS
Straight kraft recovery furnace
Smelt dissolving tank
Lime kiln
Digester system, brown stock washer
system, condensate stripper system,
or black liquor oxidation system
that are controlled by a means other
than combustion
Digester system, brown stock washer
system, multiple-effect evaporator
system, black liquor oxidation
system, or condensate stripper
system
5 ppm by volume on a dry basis, cor-
rected to 8% 02
0.0084 g/kg (0.168 Ib/ton) black
liquor solids (dry weight)
8 ppm by volume on a dry basis, cor-
rected to 10% 02
5 ppm by volume on a dry basis, cor-
rected to the actual 02 content of
the untreated gas stream
5 ppm by volume on a dry basis, cor-
rected to 10% 02*
*Unless the gases are combusted in a lime kiln subject to the NSPS provisions;
or the gases are combusted in a recovery furnace subject to the NSPS provi-
sions; or the gases are combusted with other waste gases in an incinerator or
other device, or are combusted in a lime kiln or recovery furnace not subject
to the NSPS provisions and are subjected to a minimum temperature of 1200 F
for at least 0.5 second; or the owner or operator has demonstrated to the
Administrator's satisfaction that incinerating the exhaust gases from a new,
modified, or reconstructed black liquor oxidation system or brown stock
washer system in an existing facility is technologically or economically not
feasible. Any exempt system will become subject to the provisions of this
subpart if the facility is changed so that the gases can be incinerated.
Subpart B - Adoption and Submittal of State Plans for Designated Faci-
lities indicates that after promulgation of a standard of performance for the
control of a designated pollutant from an affected facility the Administrator
shall publish a guideline document containing information pertinent to control
of the designated pollutant from designated facilities. The guideline docu-
ment shall provide the following information:
o Information concerning known or suspected endangerment of public
health or welfare caused by the designated pollutant.
o A description of the emission reduction systems that have, been
adequately demonstrated.
o Information on the degree of emission reduction.that is achievalbe
with each system.
17
-------
o Incremental periods of time normally expected to be necessary for
the design, installation, and startup of identified control systems.
o An emission guideline that reflects the application of the best
system of emission reduction that has been "adequately demonstrated
for designated facilities.
o Such other information that may assist the State in developing a
State plan.
Within nine months after notice of availability of a final guideline docu-
ment, each State shall adopt and submit a plan for the control of the de-
signated pollutant to which the guideline document applies. Each plan shall
include emission standards and associated compliance schedules. Except as
provided by 60.24(f), the emission standards shall be no less stringent than
the corresponding emission guideline specified in Subpart C. Where the
Administrator has determined that a designated pollutant may cause or con-
tribute to endangerment of public welfare but that adverse effects on public
health have not been demonstrated, States may balance the emission guidelines,
compliance times, and other information against other factors of public con-
cern in establishing emission standards. In addition, on a case-by-case basis
for particular designated facilities the State may provide for the applica-
tion of a less stringent emission standard provided that the State demonstrates
with respect to such facility:
o Unreasonable cost of control resulting from plant age, location, or
basic process design.
o Physical impossibility of installing necessary control equipment.
o Other factors specific to the facility that make the application of
a less stringent standard significantly more reasonable.
TRS from a kraft pulp mill is a designated pollutant under lll(d) and 40 CFR
60 Subpart B. The recommended TRS emission limits for kraft pulp mills
published by EPA on May 22, 1979 (44 FR 29828) are as follows:
Digester 5 ppm
Multiple-Effect Evaporators 5 ppm
Straight Recovery Furnace 5 ppm
System Designed for Low TRS
Emissions
18
-------
All other Straight Kraft Recovery
Furnace Systems
Cross Recovery Furnace Systems
Lime Kiln Systems
Condenser Stripper System
Smelt Dissolving System
20 ppm
25 ppm
20 ppm
5 ppm
0.084 g/kg of black liquor solids
(dry weight)
19
-------
REFERENCES FOR SECTION 1
1. U.S. Environmental Protection Agency. Standards Support and Environ-
mental Impact Statement, Vol. I: Proposed Standards of Performance for
Kraft Pulp Mills (Revised). EPA-450/2-76-014a, September 1976.
2. U.S. Environmental Protection Agency. Technology Transfer. Environ-
mental Pollution Control, Pulp and Paper Industry, Part I, Air., EPA-
625/7-76-001, October 1976.
3. Kirk-Othmer. Encyclopedia of Chemical Terminology, 2nd Ed. Vol. 16,
1970.
4. U.S. Environmental Protection Agency. Analysis of State and Federal
Particulate and Visible Emission Combustion Sources. EPA-450/2-81-080,
November 1981.
20
-------
SECTION 2
GENERAL PREPARATORY AND PREINSPECTION PROCEDURES
Preparation is the key to a successful inspection of a kraft pulp mill.
Such preparation includes:
o Becoming familiar with the various types of process and control
equipment used
o Reviewing past operating practices
o Procuring and testing the necessary inspection equipment to be sure
it is working properly
o Inspecting the plant's exterior to obtain information about oper-
ating practices
o Advising key plant personnel well in advance so that they are avail-
able to answer questions and take part in the inspection. The
cooperation of key plant personnel is critical to the success of
the inspection.
Advanced preparation on the part of the inspector can save valuable time for
both the inspector and plant personnel. A well-informed and prepared inspec-
tor generates a degree of confidence that makes plant personnel more inclined
to provide information critical to completing a comprehensive plant inspec-
tion.
2.1 FILE REVIEW
Baseline operating parameters for both process and control equipment
should be obtained from information filed at the agency prior to the inspec-
tion. The typical source file should contain information on permit
activity, previous inspections, or emission stack tests. The files generally
provide descriptions of such characteristics as the size, throughput, and
efficiency of the process and control equipment, which give the inspector
a perspective on the overall layout and operation of the mill. The source
files also contain information on citizen complaints, equipment malfunctions,
21
-------
opacity levels, and the overall compliance status of the various sources at
the mill. This information helps to establish a history of the processes and
helps the inspector to focus attention on those processes with continuing com-
pliance problems.
File data on previous emission stack tests and other inspections should
be used to establish a baseline for comparison with future inspections or
stack tests. Such information also helps to establish a normal operating '
range for both process and control equipment and permits the inspector to
readily note any deviations.
The sample checklist in Figure 2-1 should be used to gather information
from the agency files about previous abatement activities.
2.2 SAFETY PRECAUTIONS ;
Safety precautions must be practiced during plant inspections because
heavy equipment movement, high-temperature process equipment, high-pressure
steam, toxic gases, and noise are common. Because kraft pulp mills have a
significant number of high-temperature processes and many of the processes
are wet, extreme caution should be taken to avoid burns from high-temperature
processes and the possibility of slipping and falling around the wet processes.
There are also several specific processes that are of concern in terms
of the inspector's overall safety. The operation of debarkers and chippers
present safety hazards because of the potential for small bark and wood
particles to be ejected from these processes at extremely high rates of speed.
In addition, these processes are extremely noisy, which makes it difficult to
hear any vehicles that may be moving through the mill.
Another special concern is the potential for an explosion of the smelt
dissolving tank or recovery boiler. When molten smelt at a temperature of
approximately 1600°F is added to smelt dissolving tanks containing water and
green liquor, the instantaneous transfer of heat between the smelt and the
liquid produces an explosion of steam. These explosions can be on the surface
of the tank or deep within the tank. Surface explosions result in a shower
of smelt and green liquor that can be hazardous because of the high tempera-
ture of the material. Explosions deep within the tank, however, may blow the
tank lid off or split the tank at the seam. Both types of explosion can result
22
-------
NAME OF COMPANY
ADDRESS
PLANT CONTACT
PREVIOUS INSPECTIONS:
Date Process(es)
Comment(s)
Date
Process(es)
Comment(s)
Figure 2-1. Checklist for obtaining information during
a file review. (Continued)
23
-------
Date
Process(es)
Comment(s)
Date
Process(es)
Comment(s)
Figure 2-1. Checklist for obtaining information during
a file review. (Continued)
24
-------
STACK TEST: (attach specific test results)
Emission rate
Date
Process(es)
Process rate
Actual
Allowed
COMPLAINTS:
Date
Source
Comment
MALFUNCTIONS:
Date
Source
Comment
Figure 2-1. Checklist for obtaining information during
a file review. (Continued)
25
-------
PROCESS INFORMATION:
B.
RAW MATERIAL RECEIVING (Wood yard)
Round wood
Chips
Sawdust
DEBARKING
Type
Number
CHIPPER
Type
Number
CHIP UNLOADING
Truck
CHIP TRANSFER
Conveyor
DIGESTERS
Type
batch
Continuous
Type
Process rate
_tons/h
Process rate
_tons/h
Rail
Barge
Pneumatic
Receiver
Control type
Process rate
tons/h
Number
Cook time
Pulp rate
Volume
min
ton/h
ft*
Cooks per day
Number of blow tanks
Volume ft"
Condensers primary __
secondary
TRS control
Number __
Pulp rate ______
yes
no
TRS control
Scrubber
tons/h
yes
no
Incinerator
Figure 2-1. Checklist for obtaining information during
a file review. (Continued)
26
-------
BLACK LIQUOR OXIDATION
Weak
Air
Type
Strong
Oo
Inlet sulfidity
Efficiency
TRS control
Type
g/liter Outlet sulfidity
g/liter
yes
no
Scrubber
Incinerator
H. MULTIPLE EFFECT EVAPORATOR
Number of lines
Number of effects
Inlet black liquor solids (BLS)
Feed rate M Ib/h
TRS control yes
Outlet BLS
no
RECOVERY BOILER (complete for each boiler)
Boiler manufacturer
Design old
Evaporator type _,
new
Cascade
Cyclone
Venturi
Noncontact
Inlet BLS to evaporator
BLS to guns 3
Firing rate BLS
BL
Rated capacity
Liquor heat value _
Auxiliary fields
Particulate control
M Ib/h
gpm
Ib steam/h
BLS/h
BLS/day
tons air-dried pulp (TADP)/day
_ Btu/lb BLS
_ Natural gas Residual oil
.- ESP
Scrubber
Figure 2-1. Checklist for obtaining information during
a file review. (Continued)
27
-------
0. SMELT DISSOLVING TANK (complete for each tank)
Number of tanks
Smelt rate
Control
tons/h
Mesh pads Scrubbing media
Low energy scrubber
Venturi scrubber
None
LIME KILN (complete for each kiln)
Capacity
Fuel
Heat input
Product rate
Slurry solids
Slurry feed rate
Slurry soda
tons/day
10° Btu/h
_ tons/day
gpm
Particulate control
TRS control yes
Venturi
no
Low energy scrubber
SLAKER (complete for each slaker)
Capacity tons/h
Green liquor
CaO
Control
Showers
Low energy scrubber
M. POWER BOILERS (complete for each boiler)
Type
Heat input
Fuel
Control
10° Btu/h
Coal Wood
Multicyclone
Venturi
ESP
Baghouse
Oil
Gas
Figure 2-1. Checklist for obtaining information during
a file review. (Continued)
28
-------
N. BLEACH PLANT
Capacity
TADP/day
Number of stages
Bleach chemicals
Control
C12
cio2
Other
Packed bed
Other
Figure 2-1. Checklist for obtaining information during
a file review. (Continued)
29
-------
in death or serious injury to those in the area. If green liquor represents a
high percentage of the liquid, the explosion is more violent and can occur at
a higher temperature because the salts in the green liquor lower the boiling
point of the liquid. The higher the temperature of the liquid, the less likely
and less violent will be the explosion.1'2
Recovery boiler explosions may occur for many reasons. The most common
is water/smelt interaction as a result of water tube leaks, firing of weak
liquor, or the improper use of.water or air lances. Between 1969 and 1973
there were 23 recovery boiler explosions caused by smelt-water contact and
three caused by auxiliary fuel firing practices.
It should be noted that the possibility of a recovery boiler explosion
does exist and the inspector should be familiar with the specified evacuation
routes and procedures before entering the recovery boiler area.2
Turpentine vapors and other organic mists emitted from the digesters also
create an explosion hazard. Scrubbers can be used to reduce this explosion
potential.
A major safety concern at kraft pulp mills is the entry into a confined
area. The cardinal rule for entering a confined area is "never trust your
senses." What may appear to be a harmless situation may well be a potential
threat. The three most common conditions constituting a threat are:
o Oxygen deficiency ' . i
o Presence of combustible gases and vapors ;
o Presence of toxic gases and vapors.
An inspector should always anticipate that any one or a combination of
the above conditions might exist in a confined area such as ductwork, stack,
open tanks, penthouse, or the internal portion of a wet scrubber or an ESP.
Tests for flammability, oxygen deficiency, and toxicity must be made before
an inspector enters a confined area. No one factor will provide more safety
than the knowledge of the potential threats that may exist within the area
to be inspected. Armed with this knowledge, the inspector can take appropriate
precautions and use the proper equipment to minimize any potential dangers.
Many liquids used and handled in the mill are corrosive or a skin irri-
tant. Care must be taken to avoid contact with such chemical reagents or
bleach chemicals as: sodium hydroxide (NaOH), sulfuric acid (HgSOj, black
30
-------
liquor, and white liquor. Many solids handled at the mill are also hazardous
to the skin and eyes and contact should be avoided. The inspector should
wear appropriate eye protection (glasses and side shields) when conducting
inspections in the mill.
Because pulping is carried out at elevated temperatures and steam is used,
accidental contact with heated surfaces may occur. The inspector should be
constantly aware of the location of piping, duct work, or equipment that may
present a potential hazard. Protective clothing (long sleeved shirts, gloves,
etc.) should be worn to minimize burns as a result of accidental contacts with
hot surfaces.
The inspector should remove jewlery, ties or other loose objects before
entering the work area. Because of the close quarters in many mills and the
possible contact with moving equipment during the process of taking measure-
ments, any object that could become easily entangled should be removed.
In addition to the general concerns noted above, if an inspector is con-
ducting an internal inspection of a wet scrubber or an ESP, he/she should:
o Observe interlock procedures
o Observe the confined entry procedures of isolation lockout
o Watch footing
o Never work alone
o Wear protective equipment
o Deenergize unit before entry
o Purge unit before entry
o Use grounding straps
o Never enter full or partially full hoppers (wet- or dry-bottom).
Lastly, the inspector should be aware of and obey all safety requirements
set forth by plant'personnel. Many mills have their own safety procedures
that must be obeyed; therefore, the inspector should meet briefly with plant
personnel regarding any additional safety concerns or requirements that the
mill has established with respect to access to certain equipment or areas of
the plant and any special safety equipment that may be required.
31
-------
2.2.1 Exposure to Hydrogen Sulfide
Because of the number of sources in the mill that produce TRS emissions
and the potential for the inspector to come in contact with these gases, the ;
following is provided to inform the inspector of the effects of hydrogen
sulfide exposure. Hydrogen sulfide is a colorless gas with an obnoxious odor
at low concentrations. In humans, it can cause headache, conjunctivitis,
sleeplessness, pain in the eyes at low concentrations, and death at high con-
centrations. Hydrogen sulfide is extremely toxic to humans. Initial exposure
is through the respiratory tract, from which the hydrogen sulfide is carried
by the blood stream to various body organs. Hydrogen sulfide in the blood
can block oxygen transfer at high concentrations.- Prolonged exposure can
also result in enzyme poisoning and irreversible nerve tissue damage. ' At
high concentrations (>1,000,000 yg/m ) exposure causes death by paralysis of
the respiratory center. Table 2-1 summarizes the health effects associated
with human exposure to hydrogen sulfide.
The reported odor threshold for hydrogen sulfide varies from 1 to 45
yg/m . At 500 yg/m , the odor is distinct; at 4,000 to 8,000 yg/m3, it
is offensive and intense. At 30,000 to 50,000 yg/m , the odor is very strong
fi *3
but tolerable; at greater than 320,000 yg/m , the smell is less pungent
because the olfactory nerves become paralyzed. Continued exposure to high
concentrations results in a distinct reduction of odor perception as a result
3 ?
of olfactory fatigue. At concentrations above 1,120,000 yg/m, no odor may
be sensed and death can occur rapidly. Dulling of the olfactory nerves
constitutes a major danger to the inspector who is exposed to moderate to high
concentrations for extended periods of time.
2.2.2 Exposure to Chlorine
The inspector should be particularly cautious while inspecting the bleach
plant, where chlorine and chlorine dioxide may be used as bleaching agents.
The extraction and washing of bleached pulp exposes the pulp to ambient,air,
where residual chlorine gas may be lost. Although these sources are hooded
and vented, the possibility of exposure still exists in some cases.
Chlorine, a dense, greenish-yellow gas with a distinctive irritating
odor, is a very strong oxidizing agent. This highly corrosive gas is
extremely hazardous to all life forms.
32
-------
TABLE 2-1. SUMMARY OF REPORTED HUMAN HEALTH EFFECTS OF
HYDROGEN SULFIDE7
Concentration,
yg/m3
Effect
1 - 45
10
150
500
15,000
30,000
30,000 - 60,000
150,000
270,000 - 480,000
640,000 - 1,120,000
1,160,000 - 1,370,000
> 1,500,000
Odor threshold.
Threshold or reflex effect on eye.
Smell slightly perceptible.
Smell definitely perceptible.
Minimum concentration causing eye irritation.
Maximum allowable occupational exposure for
8 hours.
Strongly perceptible but not-intolerable
smell. Minimum concentration causing lung
irritation.
Olfactory fatigue in 2 to 15 minutes; irrita-
tion of eyes and respiratory tract after
1 hour; death in 8 to 48 hours.
No serious damage for 1 hour but intense local
irritation; eye irritation in 6 to 8
minutes.
Dangerous concentration after 30 minutes or
. less.
Rapid unconsciousness, respiratory arrest, and
death, possibly without odor sensation.
Immediate unconsciousness and rapid death.
Human sensitivity to chlorine gas varies greatly. Its main effect is an
irritating and corrosive attack on the mucus membranes of the eyes, nose,
throat, and respiratory tract.10 High concentrations of chlorine can damage
the lungs and result in pneumonia,11 emphysema, and bronchitis. In very
12 13
high concentrations, damage may be severe enough to cause suffocation. '
Table 2-2 summarizes the health effects attributed to chlorine exposure.
33
-------
TABLE 2-2. SUMMARY OF REPORTED HUMAN HEALTH EFFECTS OF
INHALATION OF CHLORINE9
Concentration,
PPM
Exposure time
Effect
< 1.0
1
3.3
3-5
3-6
4
5
5
5
10
> 10
14-21
20
50
100
Several hours
Working
conditions
30-60 minutes
< 1 minute
0.5-1.0 hour
< 30 minutes
30-60 minutes
< 1 minute
Objective symptoms of irritation.
Slight; symptoms after several hours
exposure.
Risk to health or life; impossible
working conditions.
Tolerable for short periods of time
without objective evidence or
injury.
Stinging or burning sensation in the
eyes, nose, and throat; sometimes
headache due to irritation of the
nasal sinuses.
Maximum amount that can be inhaled
for 1 hour without serious
disturbances.
Slight smarting of the eyes and irra-
tation of the nose and throat.
Premature aging; those exposed suffer
from disease of bronchi and become
predisposed to tuberculosis; teeth
corrode from hydrochloric acid;
inflammation or ulceration of the
mucous membrane of the nose occurs.
Does not endanger life.
Noxious effect; impossible to breathe
after several minutes.
Severe coughing and eye irritation.
Immediate and delayed.
Dangerous.
Endangers life.
Immediately fatal.
Cannot be tolerated for longer than
1 minute.
34
-------
o
The odor threshold reported in the literature is below 1,000 yg/m (0.33
ppm), but it may be lower when acting in combination with hydrogen chloride. '
The 8-hour threshold limit value (TLV) established by the National Institute
of Occupational Safety and Health (NIOSH) is 3 mg/m3 (1 ppm).16 A peak of
0.5 ppm/15 min has been recommended but not adopted. The inspector should
wear respiratory protection while in the bleach plant area and when working
near process equipment and control devices. The plant will usually supply
respirators for visitors and inspectors to use.
2.3 SAFETY AND INSPECTION EQUIPMENT
While conducting an inspection of the pulp mill, the inspector should
use the appropriate protective clothing and safety equipment and follow all
company rules and recommendations. Because of the potential for flying pieces
of wood and bark particles, the inspector should wear safety glasses with
sideshields for protection. The debarkers, chippers, and various conveying
systems are the major sources of these flying wood or bark particles. Hear-
ing protection such as ear plugs should also be used because of the high
noise levels around debarkers and chippers. Steel-toed shoes and a hard hat .
are required for protection against overhead hazards and heavy objects. A
nonslip sole shoe with ankle support is'recommended. A long-sleeved shirt,
gloves, and trousers should be worn for protection from the high-temperature
processes and the rough texture of the delivered raw materials (wood). Neck-
ties, hair ribbons, rings, etc., should be removed prior to the actual inspec-
tion. Dust and mist respirators should be available and used around potentially
dusty operations. In some cases a gas mask may be required. In a few cases,
the inspector should use a self-contained breathing apparatus when required
^to enter a confined area.
The equipment or instruments used during an inspection vary according to
the time allotted and the level of the inspection. For example, a detailed
Level III Inspection involving several days at the mill requires the following:
a pitot tube and manometer for measuring the gas stream flow from each pro-
cess, a manometer for measuring the pressure drop across the appropriate
control equipment, a thermometer or thermocouple for measuring stack gas
temperatures, a wet bulb/dry bulb thermometer and psychrometric chart for
determining moisture, a tachometer for measuring fan speed, an ammeter for
35
-------
measuring fan motor current, an oxygen meter for determining concentration of
the exhaust gases, and a Fyrite or Orsat for determining gas composition.
Some general pieces of equipment are also necessary, including 1) a flash
light, 2) a stopwatch, 3) a tape measure, 4) a pressure gauge, and 5) a ;
water-flow device. ;
For a less detailed Level II Inspection, the inspector may only need a
camera, a compass, and a stopwatch. During a Level II Inspection, when the
inspector does not make any measurements, he or she should obtain readings
from plant instruments (e.g., pressure in the digesters, fuel firing rates,
temperatures of stack gases) and review the data.
If permitted on mill property, a camera can be useful to document ex- i
cessive opacity levels and to provide graphical descriptions of problems
arising from poor maintenance and housekeeping, missing bags, or control
equipment components, and the relative location of certain sources. Immed-
iately after taking a photograph, the inspector should make a log book
notation describing the situation represented in each photograph and the
time, date, weather conditions, and pertinent directional information.
A compass is useful for determining directions of sources relative to
each other, to the sun, and to the inspector. A stopwatch for timing visible
emissions observations is-also useful. •
2.4 PREENTRY OBSERVATIONS
Before entering the mill property or while moving from one process opera-
tion to another, the inspector can gain considerable information by preentry
observation of the mill. Sources of fugitive dust can sometimes best be
observed outside the mill. The inspector should also note the weather con-
ditions (especially precipitation and windspeed) during and, prior to the
inspection.
Exterior observations also provide an opportunity to observe the general
housekeeping practices of the mill and give the inspector an overall picture
of the mill layout for comparison with information obtained from the files.
The inspector also.can get an idea of the level of activity by observing the
raw material and pulp moving operations, mill traffic, and processes.
36
-------
While outside the mill property, the inspector can normally use a camera
to photograph excessive visible emissions. Some state laws, however, prohibit
the use of a camera witho'ut the prior permission of the mill. Data regarding
any photographs that may be taken (e.g., date, time of day, weather conditions,
position relative to the source) must be recorded immediately. In all cases,
while on plant property, the inspector must get permission from plant personnel
before taking photographs.
Visible emission observations are important in determining the operating
conditions of some processes and their associated control equipment. When
observing opacity from stacks, the inspector should follow the procedures of
Federal EPA Method 9 or appropriate State agency method. Windspeed, sky con-
dition, and other weather data should be recorded for future use because the
reading may be challenged in court. A diagram is also important in identify-
ing the particular source being observed (e.g., the No. 3 coal-fired power
boiler) and the observer's position in relation to the sun and the source.
The inspector should record opacity readings on the observation form for
a specified duration, depending on the local requirements. Although the
regulation may specify a plume opacity below a certain average for a 6-minute
period, the inspector may want to take the reading for a longer period, say
30 minutes, and determine if a 6-minute period may have exceeded the applica-
ble limit.
Opacity readings are sometimes best obtained before the inspector enters
the mill for the inspection or'after he/she leaves the plant property. The
inspector should compare the visual measurements with values obtained by using
the plant's continuous emission monitoring equipment (if available) for the
same time period. The frequency of calibration of these continuous emission
monitors should be noted.
If the agency's policy is to provide the mill with a copy of the opacity
readings taken during the inspection, the plant official receiving the copy
should sign and date the original of the opacity readings.
2.5 ON-SITE INSPECTION CHECKLISTS
During the on-site inspection, the inspector may find it useful to have a
series of checklists to record the information obtained during the inspection.
37
-------
Figures 2-2 through 2-5 are examples of the types of check lists that can be
used to obtain information on the process and abatement equipment; Addi-
tional copies of certain sections of the check lists should be completed if
there are multiple sources for each process (i.e., three recovery boilers,
three lime kilns, etc.).
38
-------
LOCATION
DESIGNATION
INSPECTOR(S) :
CLAIMED CONFIDENTIAL Yes
DATA SHEET NO.
DATE
No
A. DESCRIPTIVE INFORMATION
Wet Scrubber Type
Manufacturer
Model Number
Date Installed
Process/Source Controlled _
Particulate Characteristics
B. COMPONENT INFORMATION (Describe if applicable)
1. Gas Pretreatment:
Presaturator
Cyclones
Settling Chamber
Other
Demister:
Cyclone
Chevron
Fiberous Mat
Other
Pumps:
Number
Recirculation
Pump Manufacturer
Recirculation
Pump Rated Horsepower
Recirculation Pump Type
Figure 2-2. Wet scrubber inspection data sheet. (Continued)
39
-------
Data Sheet No.
Preparer
Confidential:"
Yes
No
B. COMPONENT INFORMATION (continued)
4. Fan/Motor (Specify)
Fan Manufacturer
Blade Type: Radial _____
Drive: Direct
Backward
Forward
Damper Position
Motor Manufacturer
Model No.
Belt
Rated Horsepower
Location: Forced Draft
Induced Draft
5. Instrumentation (Check if Applicable)
Differential
Pressures:
Temperatures;
pH:
Flow Rates;
Motor Current:
mroat
Separator
Demister
Gas Outlet
Gas Inlet
Liquor Inlet
Liquor Outlet
in.
in.
in.
°F
°F
°F
°F
Recirculation
Exit Liquor
Fan Motor Current
Other
Nozzle Pressure
Recirculation
Makeup
Purge
Fan
Pump
qpm
gpm
anm
a
a
H20
H20
H20
Figure 2-2. Wet scrubber inspection data sheet. (Continued)
40
-------
Data Sheet No.
Preparer
Confidential: Yes
No
B. COMPONENT INFORMATION (continued)
6. Materials of Construction (Specify type and gauge)
Presaturator
Throat \
Scrubber Shell
Trays/Bed Supports
Mist Eliminator
Fan Housing
Figure 2-2. Wet scrubber inspection data sheet. (Continued)
41
-------
Data Sheet No.
Preparer
Confi dential: Yes
No
C. DIAGRAM
1. Sketch wet scrubber system. (Show all major components and processes
controlled.)
2. Sketch wet scrubber layout
Figure 2-2. Wet scrubber inspection data sheet. (Continued)
42
-------
LOCATION
DESIGNATION
DATE
DATA SHEET NO.
INSPECTOR(S)
CLAIMED
CONFIDENTIAL
Yes
No
A. DESCRIPTIVE INFORMATION
Mechanical Collector Type
Cyclone
Cyclone Bank
Multiclone
Settling Chamber
Double Vortex Cyclone
Other (describe)
Manufacturer
Model Number
Date Installed
Process/Source Controlled
Particulate Characteristics
B. COMPONENT INFORMATION
1,
Cyclone
Diameter of Body
Cone Length
ft
ft
Material of Construction
Gauge of Metal
Number of Cyclones
Hoppers
Number
Slope
Insulation: Yes
Heated: Yes
Vibrators:
Yes
No
No
No
Figure 2-3. Mechanical collector inspection data sheet. (Continued)
43
-------
Data Sheet No.
Preparer
B.
Confidential:Yes _
COMPONENT INFORMATION (continued)
3. Solids Removal (Check applicable items and provide dimensions)
No
4.
Rotary Valves
Flapper Valves
Screw Conveyors
Pneumatic Conveyors
Free Fall
Fan/Motor
Fan Manufacturer
Model Number
Blade Type:
Drive:
Radial
Direct
Backward
Belt
Forward
Motor Manufacturer
Model Number
Rated Horsepower
RPM
Location: Forced Draft
Induced Draft
C. SYSTEM LAYOUT
Figure 2-3. Mechanical collector inspection data sheet. (Continued)
44
-------
LOCATION
DESIGNATION
INSPECTOR(S)
CLAIMED
CONFIDENTIAL Yes
DATE SHEET NO.
DATE
No
A. DESCRIPTIVE INFORMATION
Electrostatic Precipitator Type
Manufacturer
Model No.
Date Installed
No. of TR Sets
No. of Fields in Series
Plate Length, Height
Plate Area
Plate Spacing
Number of chambers
Bypass System: Yes
ft
ft2
in.
No
B. ELECTROSTATIC PRECIPITATOR LAYOUT SKETCH (each square 2 ft x 2 ft)
Figure 2-4. Electrostatic precipitator inspection data sheet. (Continued)
45
-------
C. COMPONENT INFORMATION
1.
2.
Gas Distribution
Perforated Plates: Inlet
Turning Vanes: Inlet
Rappers
Discharge Wires
Type
Number
Manufacturer
Air Pressure (if applicable)
Collection Plates
Type :
Number
Manufacturer
Data Sheet No.
Preparer
Confidential:Yes
No
Outlet
Outlet
Air Pressure (if applicable)
Distribution Plates/Vanes
Type ,
Number
Manufacturer
Air Pressure (if applicable)
Figure 2-4. Electrostatic precipitator inspection data sheet. (Continued)
46
-------
Data Sheet No.
Preparer
Conf i dentTaT: Yes
No
C. COMPONENT INFORMATION (continued)
Layout of Rappers (Show discharge wire rappers as D, collection plate
rappers as C, and distribution plate rappers as X.)
3.
Hoppers
Number _
Length _
Slope
Level Indicator Height
Insulation: Yes
Heating: Yes
ft
degrees
ft from top
No
No
Figure 2-4. Electrostatic precipitator inspection data sheet. (Continued)
47
-------
c.
Data Sheet No.
Preparer
Confidential:Yes
No
COMPONENT INFORMATION (continued)
4. Dry Bottom (Recovery Boiler)
Drag Chains: Longitude
Latitude
6.
7.
8.
Number of Chains
Wet Bottom (Recovery Boiler)
Number of Agitators
Drag Chain Motor Current
Drag Chain Speed rpm
Dust Discharge to Screw
amps
Liquid level indicator Yes
Liquid temperature
Flow in gal/min
Percent solids in
Instrumentation:
Primary Current
Primary Voltage
Secondary Current
Secondary Voltage
No
Flow out
Out
gal/min
Gas Inlet Temperature
Level Indicator Alarm
Hopper Heater Indicator
Light
Penthouse Heater Indicator
Light
Shell
Material of Construction
Insulation: Yes
No
Opacity Monitor
Number
Manufacturer
Model Number
Location
Power Supply
Type
Pulse System: Yes _
Digital Control Mode
Other
No
Figure 2-4. Electrostatic precipitator inspection data sheet. (Continued)
48
-------
LOCATION
DESIGNATION
INSPECTOR(S)
CLAIMED
CONFIDENTIAL Yes
No
DATA SHEET NO.
DATE
DESCRIPTIVE INFORMATION
Fabric Filter Type
Manufacturer
Model No.
Cloth Type and Weight
Area
Air-to-cloth Ratio
Pressure Drop Range
Operating Temperature
Collection Efficiency
Type of Bag Cleaning:
Shaker
ft2
ACFM/ft2
in. H20
Pulse Jet
Reverse Air
Frequency of Bag Cleaning
Bypass System: Yes
No
B. FABRIC FILTER LAYOUT SKETCH
Figure 2-5. Fabric filter inspection data sheet. (Continued)
49
-------
C. COMPONENT INFORMATION
1.
2.
3.
Hoppers
Number _
Length _
Shape
ft
degrees
Level Indicator Height _
Insulation: Yes No
Heating: Yes No __
Shell
Material of Construction
Insulation: Yes No
Opacity Monitor
Number
ft from top
Manufacturer
Model Number
Location
Figure 2-5. Fabric filter inspection data sheet. (Continued)
50
-------
REFERENCES FOR SECTION 2
1. Nelson, W. A New Theory to Explain Physical Explosions. TAPPI, March
1973.
2. Taylor, Malcolm L., and Howard S. Gardner. Causes of Recovery Boiler
Explosions. TAPPI, November 1974.
3. Petri, H. The Effects of Hydrogen Sulfide and Carbon Disulfide. Staub
2i(2)-:64, 1961.
4. Denrnead, C. F. Air Pollution by Hydrogen Sulfide from a Shallow Pollu-
ted Tidal Inlet. Auckland, New Zealand. Clean Air Conference, Uni-
versity of New South Wales, 1962.
5. Permissible Emission Concentrations of Hydrogen Sulfide. Subcommittee
on Effects of Hydrogen Sulfide of the Committee on Effects of Dust and
Gas of the Verein Deutscher Ingenieure Committee on Air Purification,
VDI 2107, 1960.
6. Patty, F. A. (Ed.). Industrial Hygiene and Toxicology, Vol. I. (New
York: Interscience, p. 896, 1958.)
7. U.S. Department of Health, Education, and Welfare. Preliminary Air
Pollution Survey of Hydrogen Sulfide. A Literature Review. Contract
No. PH 22-68-25, October 1969.
8. Adams, D. F., and F. A. Young. Draft Odor Detection and Objectionability
Thresholds. Washington State University Progress Report on U.S. Public
Health Service Grant, 1965.
9. U.S. Department of Health, Education, and Welfare. Preliminary Air
Pollution Survey of Chlorine Gas. A Literature Review. Contract No.
PH 22-68-25, October 1969.
10. Heyroth, F. F. Chlorine, CL2, in Industrial Hygiene and Toxicology,
Vol. II, 2nd ed. (New York: Intersicence, 1963.)
11. Kowitz, T. A., et al. Effects of Chlorine Gas upon Respiratory Function'.
Arch. Environ. Health 14:545. 1967.
12. Manufacturing Chemists Association. Chemical Safety Data Sheet SD-80,
Chlorine. Washington, D. C. 1960.
13. The Chlorine Institute, Inc. Chlorine Manual. New York. 1959.
51
-------
14. Beck, H. Experimental Determination of Olfactory Threshold for Some
Important Irritant Gases (Chlorine, Sulfur Dioxide, Ozone, Netrosyl-
sulphuric Acid) and Their Manifestations of the Effect of Law Concen-
trations in Maa, Thesis, Wurgburg. 1959.
15. Takhirov, M. T. Determination of Limits of Allowable Concentrations of
Chlorine in Atmospheric Air. U.S.S.R. Literature on Air Pollution and
Related Occupational Dieseases 3:119. 1960.
16. U.S. Department of Health and Human Services. NIOSH/OSHA Pocket Guide
to Chemical Hazards. September 1978.
52
-------
SECTION 3
KRAFT PULPING PROCESSES
There are five major processes and several minor processes within the
kraft pulp mill. The major processes are wood handling, pulping, chemical
recovery, causticizing, and power generation. The minor processes include
bleaching and raw material handling. Figure 3-1 is a simplified diagram of
the kraft pulping process. This section presents descriptions of these
processes and discussions regarding the sources of emissions, the control
techniques, the potential malfunctions, and the inspection techniques to be
used for each process.
3.1 WOOD HANDLING DEPARTMENT
This subsection describes the processes used to receive, prepare, and
transport logs to the pulping department. These processes include debarking,
chipping, screening, transfer, feed preparation, and storage. Most of the
emissions in the wood handling department are fugitive. The control measures
generally consist of containment and water.
The major malfunctions generally involve the pluggage of the chip trans-
fer system and discharge of the chips from the cyclone exit. The inspection
of the wood handling department consists of a visual evaluation of the fugi-
tive dust sources and a verification that the prescribed containment has not
been removed. Inspections of the wood handling department are generally
Level I Inspections.
3.1.1 Process Description
The major processes within the wood handling department are described
briefly.
3.1.1.1 Debarking—
Depending on the location of the pulp mill with respect to the wood
supply, wood is transported to the mill by railcar, truck, or barge in the
53
-------
WOOD CHIPS
EVAPORATOR
GASES
STEAM
i
1
Figure 3-1. Kraft pulping process.
54
-------
form of chips, sawdust, shavings, or roundwood (logs). Roundwood, generally
in four to eight foot lengths, is usually debarked before it is used in pulp
manufacturing. Prior to the debarking, various techniques can be used to soften
the bark. These include:
o Spraying the logs with water
o Soaking the logs in ponds
o Steaming the logs in special chambers
After the logs are softened, they must be slashed or cut into 4 to 8 foot
lengths.
Because bark has very little useful fiber, and it contains dirt that re-
duces the overall pulp quality, debarking is used to improve the quality of
the pulp. Several types of debarking machines are commonly used: drum, bag,
ring, cutterhead, hydraulic, and knife debarkers. One of the most widely
used is the drum debarker (Figure 3-2). The logs are fed into an open-ended
horizontally rotating cylinder shell that is 8 to 16 ft in diameter and 22 to
75 ft long. The tumbling action of the logs removes the bark by abrasion.
The logs move toward the discharge end during the process, and the debarked
logs are expelled. The bark removed from the logs falls through slits located
along the entire length of the cylinder. Inadequately debarked logs are re-
turned for reprocessing.
The debarking drums can be constructed of either staved sections or en-
closed vat sections. Staved sections promote tumbling of the logs against
each other, which removes loose pieces of bark. The enclosed vat sections are
used to soak the logs in a pool of warm water to thaw the bark at mills located
in areas where freezing temperatures are common. If the bark is to be used
as boiler fuel, presses are often used to remove the moisture.
Figure 3-3 depicts a bag debarker that abrades the bark from the logs by
jostling and rotating. The removed bark falls through holes in the base of
the machine, and operation can be batch or continuous, wet or dry. Water is
frequently used, to thaw frozen bark.
Ring debarkers, shown in Figure 3-4, remove bark by a scraping action.
Individual logs are fed axially through a rotating ring having equally spaced
arms held under pressure. The removed bark is either flung from the machine
2
or allowed to fall onto a conveyor below.
55
-------
Figure 3-2. Drum debarking.
Source:
Figure 3-3. Bag debarker.
Joint Textbook Committee of the Paper Industry, the Pulping of Wood,
1969. Reproduced with permission of McGraw-Hill.
56
-------
Source:
Figure 3-4. Ring barker.
Joint Textbook Committee of the Paper Industry, the Pulping of Wood,
1969. Reproduced with permission of McGraw-Hill.
Figure 3-5 shows a cutterhead machine. Bark is removed from one log at
a time by the milling action of the cylindrical cutterhead. Cutting edges
are mounted on the longitudinal axis, which rotates parallel to the axis of
the log. The removed bark is carried away by a conveyor.
Hydraulic barkers, depicted in Figure 3-6, blast the bark from the log
by directing a high-pressure water stream at the surface. The logs are de-
barked one at a time. The water then washes away the removed bark.
Figure 3-7 shows a knife debarker that uses radially mounted knives to
' strip bark and wood from the surface of the log. The knives are mounted on a
2
rotating disk.
If the debarked logs have diameters too large to enter the feed spout of
the chipping equipment, they are sent to a splitter that quarters them. De-
barked logs that are not needed immediately are stored for later use.
3.1.1.2 Chipping—
After the logs have been debarked, they must be reduced in size so that
cooking chemicals can easily penetrate the wood fibers to separate lignin and,
carbohydrates from the cellulose. Chippers, which will accept 4- to 8-foot
57
-------
Source:
Figure 3-5. Cutterhead barker.
Joint Textbook Committee of the Paper Industry, the Pulping of Wood,
1969. Reproduced with permission of McGraw-Hill. '>-
58
-------
Figure 3-6. Hydraulic barker.
Figure 3-7. Knife barker.
Source: Joint Textbook Committee of the Paper Industry, the Pulping of Wood,
1969. Reproduced with permission of McGraw-Hill.
59
-------
debarked logs, use powerful high-speed rotating knives to reduce the wood to
a uniform size (about 5/8 to 3/4 inch long by 1/8 inch thick).
Among the variety of chippers used, the most common is the Norman disk
chipper (Figure 3-8). The close spacing of the knives holds the log steady
and prevents bouncing in the feed spout. In operation, each knife moves through
a section of the log in the same direction as its bevel. The logs are cut
, against the grain at an angle of 35 to 52 degrees relative to their axis. At
every point along the cutting edge, the bevel is directed toward the edge of •
the next chip slot. Every point along the knife edge exerts the same force
in pulling the log into the disk. This technique reduces the bruising of the
chip fibers and results in a stronger pulp.
Drum chippers (Figure 3-9) consist of an open-ended drum that rotates at
20 to 40 rpm. The drum chipper is fitted with several equally spaced cutting
knives on the periphery that cut parallel to the grain of the log. The cut-
ting assemblies consist of front knives with three cutting edges that cut '.
trapezoidal grooves in the wood to form chips. The remaining wood is sliced ;
off by a trailing knife to form additional chips.
The spiral horizontal parallel (HP) chipper (Figure 3-10) cuts parallel ;
to the grain. Wood is fed into the disks mounted on a horizontal shaft
arranged in a V shape. Two or more rows of knives are mounted on each disk
in a spiral formation.
3.1.1.3 Knotting and Screening--
After passing through the chipper, the wood contains fines, slivers, and
oversized chips. A screening operation is performed to separate rejects and
fines from properly sized chips. Fines are generally less than 1/8 inch in
diameter and are usually pulped separately for low-strength paper, or are
burned in the bark boilers. Oversized chips are separated and removed for
rechipping or recrushing and rescreening.
Rechipping often damages the chips and results in increased fines.
Several types of rechippers are currently being used, but the chips receive
less damage if the rechippers use knives rather than blunt hammers.
Screening generally involves two decks. The upper deck captures the
oversized chips and the lower deck retains proper sized chips. Fines drop
into a collection hopper below the lower deck. The kinds of screening equip-
ment, classified according to movement mechanism, are rotary-drum, shaker,
60
-------
SE6MENTAL KNIFE HOLDER
-DISK
LOG
CLAMPING BOLT
CHIP SLOT
STEEL INSERT
CHIP SLOT
SPOUT
PULL-
ANGLE
BEDKNIPE
FRONT FRAME
Figure 3-8. Norman disk chipper.
s
Source: Reproduced with permission of Carthage Machine Company, Inc.
61
-------
Figure 3-9. Drum chipper.
Source: Reproduced with permission of Koehring-Waterous Ltd.
Source:
Figure 3-10. Horizontal parallel chipper.
Reproduced with permission of Joint Textbook Committee of the Paper
Industry, the Pulping of Wood, 1969. '
62
-------
vibratory, and gyratory. Vibratory and gyratory are the most efficient and
the ones most often used. Vibratory screens (Figure 3-11) are inclined at
approximately 25 degrees from the horizontal;f Gyratory screens (Figure 3-12)
are inclined at 4 degrees from horizontal, and they use r-ubber balls that
deflect against the underside of the wire mesh deck to dislodge chip particles
27
that plug the openings. *
3.1.1.4 Storage and Transfer-
Chip feeders operating at a constant and uniform rate are used to move
the chips from storage. Various equipment is used to accomplish this task,
including a traveling screw, a rotary plate, a stoker, a multiple screw, and
parallel drag chains. Each has its own technique for moving the chips.
Pneumatic conveying is frequently used when the chips are stored outside.
A pneumatic,system consists of a pipeline with a high-velocity airstream that
is generated by a blower. The chips are fed into the airstream and transported
to the desired point of exit.
Chips, shavings, and sawdust arrive at the mill in freight cars, semi-
trailers, or trucks and must be removed to storage areas for future use. Some
of the techniques used to transport the chips, shavings, and sawdust are:
o Chip digger with suction unloading (for boxcars)
o Rotating nozzle suction unloading (open-top cars)
o Hydraulically operated truck dumper
o Rotary freight car dumper with gravity-dump pit
o End-dump freight car with pit.
Once the chips are removed from their transporting vehicle, they can be
transported to bins, outside piles, or silos by belt conveyor, bucket elevators,
or pneumatic systems. Usually the bins are loaded by overhead conveyors and
unloaded by a slot at the bottom that uses a traveling screw.
Pnuematic systems use cyclones as primary receivers to separate chips
from the transport gas stream. The cyclones are 8 to 14 ft in diameter and
typically operate at gas volumes of 4000 to 11,000 acfm. The inlet velocities
are on the order of 100 to 160 ft/s. In this application, the cyclones are
used as product transfer and are not considered an air pollution control de-
vice. Typical transport rates are 50 to 100 tons/h.- The cyclone may be
63
-------
"OS" SPRING SUSPENSION
ASSEMBLIES
FEED PLATE
IIUBBEM COVERED SUPPORT MAILS
TENSIONINO STRIPS
"K" SPRING FLOOR
MOUNTING WITH STABILIZERS
•CUttN END SUPPORT
DtSCHAMQE PANEL LIP
Figure 3-11. Vibratory chip screen.
Source: Reproduced with permission of W. S. Tyler Company.
64
-------
INTAKE
onrr, SAWDUST
AND OTHER FINES
USABLE OOPS
DISCHARGE OF
OVERSIZE CHIPS
Figure 3-12. Gyratory chip screen.
Source: Reproduced with permission of ROTEX, Inc.
used to transfer the material to screens, the hammer mill, or the chip storage
bins.9
Figure 3-13 shows a typical end-dump rail car unloading system, and Figure
3-14 shows a hydraulic truck dump system. Chips received in rail cars can be
unloaded by a rotary rail car dump (Figure 3-15).
Figure 3-16 shows the normal material and air flow at a small wood-
0
handling yard.
3.1.2 Sources of Emissions and Control
Most of the emissions from a woodyard, except for those from pneumatic
conveying systems, are fugitive. In general, control measures consist of con-
tainment of sources and the use of water on haul roads and other traveled
areas. Water may also be used on the debarkers to reduce dust and to wash the
logs.
Chips that are received dry (shaving and saw dust) are also potential
fugitive sources. Water is effective in reducing the emissions at transfer
points.
65
-------
Figure 3-13. Hydraulic dump truck for chips.
Source: Reproduced with permission of Screw Conveyor Corporation,
66
-------
Figure 3-14. End-dump freight car and unloading platform for chips.
Source: Joint Textbook Committee of the Paper Industry, the Pulping of Wood,
1969. Reproduced with permission of McGraw-Hill.
67
-------
Figure 3-15. Rotary freight car dryer for chips.
Source: Joint Textbook Committee of the Paper Industry, the Pulping of Wood,
1969. Reproduced with permission of McGraw-Hill.
68
-------
EXIST.
CTl
ID
EXIST.
EXIST. TRUCK « RAIL CAR
CHIP UNLOADING SYSTEM
E~JE~3
-Jft tMIIMi M
600
TONS: SHIR
RING SHOWER I OVERSIZE
(WOOD WASHER) CONV.
EXIST.
BLOWER
TO MILL
50 TONS/HR
(EMERGENCY ONLY)
FINES TRUCKED
TO LANDFILL
Figure 3-16. Typical material and air flow for small woodyard.
-------
In urban areas, more control is needed because of the potential nuisance
to surrounding property owners. As a result, several plants within urban areas
have paved the haul roads and chemically stabilized or sprayed the storage
areas. Periodic sweeping and/or wetting of these haul roads has been effec-
tive in reducing emissions resulting from vehicular traffic. Minimiaing the
dust around the plant side has the added benefit of reducing the carryover of
dust and dirt into the pulping process.
Emissions from the pneumatic conveying cyclone are generally controlled
by the use of water sprays. These water sprays can reduce cyclone emissions
by 95 percent. Grain loadings from a chip transfer cyclone (14 foot diameter
and 10,000 acfm) handling 100 tons/h of dry chips can range from 0.0003 to
0.034 gr/dscf. The lower value is associated with a cyclone using water sprays.
A typical water injection rate is approximately 4.5 gpm (0.5 gal/1000 acfm).
Green chip transfer through the same system generate an emission rate of 0.002
gr/dscf.9
3.1.3 Malfunctions
Malfunctions that can occur in the wood handling department involve the-
pluggage of the chip transfer systems and the discharge of the chips from the
cyclone exit. These malfunctions usually cause a fallout problem on the plant
premises, which normally can be corrected quickly.
When water sprays are used to control fugitive dust from the transfer and
debarker systems, provisions must be made to prevent freezing during the
winter months.
3.1.4 Inspection of Wood Handling Department
The inspection of the wood handling department consists of a visual evalua-
tion of the fugitive dust sources and a verification that the prescribed con-
tainment has not been removed; If water spray systems are used, the inspector
should verify the location of spray nozzles and visually determine if the
water spray pattern is adequate.
The water flow rate should be recorded along with the water supply pres-
sure for each system. Table 3-1 provides a checklist for process weight and
control variables of the emission sources in the wood handling department. '.
Generally the information listed in Table 3-1 can be obtained during a Level:
I or II inspection.
70
-------
TABLE 3-1. CHECKLIST FOR INSPECTION OF WOOD HANDLING SYSTEMS
Debarker:
Chipper:
Screens:
Rechipper:
Truck unloading:
Type
Process rate
Control method_
Water flow rate
Water pressure _
Type Process rate
Control method
Water flow rate
Water pressure
Type. Process rate
Reject rate
Control method
• Process rate
Control method
Type Process rate
Control method
Rail car unloading: Type
Process rate
Control method
Pneumatic transfer: (more than 1 repeat)
Process from
Process rate
Gas volume
Cyclone diameter
Inlet deminsions
to
tons/h
acfm
ft
X
Inlet area ft
Water spray flow rate
Duct leaks Yes No
in.
gpm
_gpm
psig
_gpm
psig
71
-------
3.2 PULPING DEPARTMENT
The pulping process involves many separate processes, including the
cooking of the chips to produce pulp, recovery of cooking chemicals, and con-
centration of spent liquors. For simplicity, all of these processes are in-
cluded in this subsection.
The primary emissions from the pulping department are TRS compounds which
can be controlled by 1) reduction in quantity through process operation, !
2) scrubbing with fresh water or alkaline solutions, or 3) incineration. The
malfunctions in the pulping department involve the digester relief systems,
digester blow system (hot water accumulator and blow tank), multiple-effect
evaporator system, black liquor oxidation system, and the TRS Vent Control ;
System. In general, all the major malfunctions associated with the above
systems will tend to increase TRS emissions in one way or another. Several
tables are presented in Section 3.2.3 that identify the potential malfunctions
for each type of system, the primary effects of these malfunctions, and the
subsequent results in terms of TRS emissions.
The inspection of the pulping department is generally based on information
concerning equipment specifications, process procedures, process weights, and/
or control equipment bypass and malfunction. Because the inability to obtain
physical measurements with hand held equipment in the pulping department, in-
spections in the pulping department are generally limited to Level I or II.
A series of check lists are presented in Section 3.2.4 that will aid the in-
spection in obtaining whatever information is available for the major systems
in the pulping department.
3.2.1 Process Description
3.2.1.1 Pulp Digester— ;
«
The two major components of pulpwood are cellulose and lignin. The fibers
of cellulose are bound by the lignin in the wood. Figure 3-17 presents a
schematic of the pulping process. Wood chips are added to a reaction vessel
(digester) containing cooking chemicals comprised of about 66 percent sodium
hydroxide (NaOH) and 33 percent sodium sulfide (Na2S). Some recycled black
liquor is also added to the vessel. The mixture is heated to 340° to 350°F
under 100 to 135 psig of steam for lh to 3 hours. This process removes the
lignin and carbohydrates from the pulp to free the cellulose. Sodium
72 ;
-------
CO
SECONDARY
CONDENSER
11 HOT WATER
10 COOLING WATER
(5) RELIEF GAS
(8) TURPENTINE
© FOUL CONDENSATE
(?) BLOW GAS
HEAT
EXCHANGER
5 HOT WATER
A WARM WATER
© BLOW CONDENSTATE BLEED
2 FRESH WATER
(?) PULP-LIQUOR
POINTS OF POSSIBLE ODOR RELEASE ARE ENCIRCLED,
Figure 3-17. Pulping process.
12
-------
hydroxide breaks down the fibers in the wood. Sodium sulfide, by double de-
composition reaction, breaks down to liberate NaOH as it is consumed. Thus
Na2S acts as a reservoir for NaOH so that excessive amounts are not added,
which would weaken the pulp. Sodium sulfide is responsible for the strength
attributable to kraft pulp.
The digestion process can be either batch or continuous, and the digester
can be heated directly or indirectly. The digester generally has a volume of
between 3000 and 6500 ft3 and will normally produce between 8.5 and 17.9 tons
of unbleached air-dried pulp per blow. The digesters are charged with white
liquor, chips, and recycled black liquor. Table 3-2 shows the typical liquor
O
requirements for bleached paper and bag paper. The liquor requirements, pulp
yield, pulp strength, and cooking time are mill-specific.
TABLE 3-2. TYPICAL DIGESTER LIQUOR REQUIREMENTS2
Pulp type
Bleached papers
Bag paper
Total liquor per
blow, ftVTADP
100 - 235
80 - 170
White liquor per
blow, ftVTADP
80 - 120
65 - 115
Black liquor per
blow, ftVTADP;
5 - 110
5-90
A typical digester temperature/pressure chart for repeated digester batch
cycles is shown in Figure 3-18.2 The total number of potential digester blows
per day may be calculated from the number of digesters and the average digester
cycle time, including charging, cooking, and blowing.
Chips are charged to the digester from overhead chip bins in the pulp i
building. Minor steam is lost and odors are generated during charging. Re-
movable charge chutes are used to transfer the chips. Figure 3-19 shows a
typical digester charging room floor.
3.2.1.2 Batch Digester (Blow)—
During batch digestion, the pulp is held at elevated temperature and
pressure. To remove the pulp from the tanks, the digester bottom is opened
and the tank pressure is relieved to a blow tank. The process is referred to
as a digester blow. ;
74
-------
Figure 3-18. Typical digester operating curves.
Source: Joint Textbook Committee of the Paper Industry, the Pulping of Wood,
1969. Reproduced with permission of McGraw-Hill.
75
-------
Figure 3-19. Typical digester charging room floor.
Source: Joint Textbook Committee of the Paper Industry, the Pulping of Wood,
1969. Reproduced with permission of McGraw-Hill.
Because the temperature and pressure of the digester contents are quite
high, the rapid release of pressure flashes a portion of the water to steam.
The amount flashed depends on the blow rate and the initial and final temperature.
A large pressure vessel (blow tank) is provided for separation of the pulp
from the spent cooking chemicals (black liquor) and to allow expansion volume
for the evolved steam. For economical operation of the digester system, the
blow tank is equipped with a hot water accumulator (Figure 3-20), which may
have both primary and secondary condensers to recover heat from-the flashed
steam. The primary condenser is generally a direct-contact type and the
secondary condenser is a tube and shell design.
Calculation of the exhaust volume from the blow can be based on liquor
volume, system mass balance, and specific heats of the materials. After con-
densing, the gas volume is much lower, and the gas contains equilibrium moisture
at 212°F and 1 atm plus any noncondensable gases and trapped air.
76
-------
BLOW SECONDARY
BLOW
STEAM
PRIMARY
CONDENSER
BLOW
CONDENSATE
GAS
CONDENSER
HOT WATER
4 WARM WATER
COLD WATER
Figure 3-20. Hot water accumulator showing primary
and secondary condensers.12
The digester blow is a significant air pollution source of TRS compounds.
The reaction of the wood with white liquor sulfide generates hydrogen sulfide
(H2S), turpentine, and traces of methanol (CHgOH), ethanol (CHgCHgOH), and
acetone (CH2COCH3). The composition and quantity of digester gases will
differ between batch digesters and continuous digesters. Generation rates will
also differ depending on wood type, white liquor sulfidity, final cooking
temperature, and cooking time.
The typical blow volume is 90 percent steam. The noncondensable portion
3 12
of the blow volume is about 95 ft /ton of pulp. The blow rate is not con-
stant and reaches a peak during a portion of the blow period. Figure 3-21
shows a typical digester steam flow pattern. The primary and secondary
condensers are designed for the peak flow rate.
The digester blow volume may be calculated from the digester operating
conditions and a liquor mass balance. Most of the information required to
complete the calculation is available from the pulp cooking records and/or
digester design data. At the end of the cooking period, the digester is
77
-------
100 -
90 -
80 -
70-
50 '
LU
40 -
w
5 30 -
§ 20-
10 -
0
h 2500 I
»_
L 2000
- 1500
1000
500
10 20
TIME, MINUTES
30
Figure 3-21. Typical batch digester steam flow rate during blow.
13
discharged into the blow tank and the temperature drops from 350 to 220°F.
Some steam is flashed during the blow period. The basic work sheet for cal-
2
culating blow weight is presented in Figure 3-22. The calculation represents
a balance between total heat available from the liquor and total heat availa-
ble from the pulp and digester shell.
3.2.1.3 Digester Relief and Turpentine Recovery—
The digester cooking cycle generates organic gases as a result of volati-
lization of wood oils. For maintenance of acceptable cooking pressures, the
digester must be periodically vented to the atmosphere. The relief gases are
passed through a liquor separator (cyclone) to remove any entrained liquor
and fiber (Figure 3-23). From the separator the gases are passed through a
surface condenser» which removes the condensable gases (primarily turpentine
and water). The condensate is passed to a decanter, where the water and
organics are allowed to separate. Crude turpentine is drained off to storage-
2 12 14
and shipped to refiners. * ' ;
78
-------
Weight and specific heat of black liquor discharged:
Lignin, etc.
Water in chips
Cooking liquor
Black liquor
15,800 Ib
26,200 Ib
59,000 Ib
30.000 Ib
Total 131,000 Ib
Relief vapor -1.850 Ib
Total weight of black liquor discharged 129,150 Ib
Liquor specific heat 0.855
Weight and specific heat of moisture-free pulp discharged
Pulp weight
Pulp specific heat
Weight and specific heat of digester shell
Shell weight
Shell specific heat
Mass balance - Blow heat available, 10 Btu
Black liquor
Pulp
Heat from digester shell
(129,150)(0.855)(350-220)
( 14,400)(0.33)(350-220)
( 63,270)(0.117)(350-255)
Total heat
Latent heat of steam at 220°F
Weight of steam flashed (excluding 1t- /-on nnn
noncondensable gases) at 220°F = 1a'g°"'uuu
965
14,400 Ib
0.33
63,270 Ib
0.117
= 14.355
= 0.620
= 0.705
= 15.680
965
= 16,250 Ib/blow
Figure 3-22. Worksheet.for calculation of blow weight (steam).
79
-------
RELIEF COMPUTING
AND CONTROL SYSTEM
DECANTER
STRAINER
LIQUOR
SEPARATOR
CONDENSATE
CRUDE TURPENTINE
BLOW VALVE
JXJ
CONTROL SYSTEM OF TURPENTINE COLLECTION
Figure 3-23. Digester relief and turpentine recovery system.15
80
-------
Typical turpentine yields from soft wood pulping are on the order of 2
to 4 gal/TADP. The turpentine yields from hardwood pulping are minimal and
recovery is not practiced. Continuous digesters also generate turpentine as
a result of continuous relief during pulp blowing. The yield and quality of
turpentine from continuous digesters is generally inferior.
The amount of condensate produced by softwood pulping is mill-specific,
but is generally about 540 Ib/ton of pulp. The volume of noncondensable gases
3 12
produced is on the order of 32 ft /ton of pulp.
The amount of noncondensable organic emissions present in the condenser
vent is a function of the condensate outlet temperature. Figure 3-24 shows
the relationship between TRS emissions and condenser outlet gas temperature.
17
3.2.1.4 Continuous Digester--
Continuous digestion presents less of an air pollution problem than batch
digestion because the gas flow is regular. Spent liquor from the digester is
expanded (flashed) in two stages. The flash from the primary tank is used to
impregnate chips in a presteaming vessel prior to digestion. Steam from the
secondary flash tank is used for various purposes, such as turpentine recovery,
hot water production, additional impregnating steam, etc. In both batch and
continuous digestion when soft woods are digested, the condensate undergoes a
secondary process that separates the turpentine from water. Figure 3-25 depicts
12
this process.
The water that condenses from the digestion process is used for brown
stock (pulp) washing. Prior to being washed, however, the pulp goes through
a knotting operation that further dilutes the fiber to lh percent by weight
and removes oversized chips that have not been completely impregnated with
white liquor. These chips are sent back through the digester, and the water
is drained into filtration tanks.
3.2.1.5 Pulp Washing--
After cooking, the pulp contains a considerable quantity of chemicals
and dissolved waxes, oils, and fatty acid salts. These chemicals are re-
coverable when converted to their original form through the recovery process.
Pulp from the blow tank is mixed with white liquor and black liquor in
the blow tank and transferred to the pulp washing lines. The pulp washing
lines are of the displacement-washing type. In this process the liquor and
81
-------
20 •
"E
^ 15-
to
CD
z
o 10 -
|
H
0
8 5-
n -
* .
o
o
• ° x
X *
•
X
X
X* ,
0
o o
i
A 0 0 °
50 60 70 80 90
CONDENSER OUTLET TEMPERATURE, °C
• HYDROGEN SULFIDE
o METHYL MERCAPTAN
X DIMETHYL SULFIDE
A DIMETHYL DISULFIDE
Figure 3-24. Odor compounds in relief gas after turpentine
condenser as a function of condenser outlet temperature.10 '
12,17
82
-------
14 STEAM
15 COND.
PRESTEAMING if"5^
VESSEL
13 WHITE LIQUOR
12 CHIPS
© VENT
11 HOT WATER
10 COOLING WATER
(§) TURPENTINE
(7) FOUL CONDENSATE
l6 FLASH STEAM
(|0) WEAK LIQUOR
^ © PULP + LIQUOR
OT) WASH LIQUOR
POINTS OF POSSIBLE ODOR RELEASE ARE ENCIRCLED
Figure 3-25. Continuous digester flow sheet.12
83
-------
chemicals are displaced in the pulp by fresh water. The washing is accomplished
in stages, with waste liquor being transferred countercurrent to the pulp flow.
The washing usually takes place on rotary drum filters, and air is used to
maintain a pressure differential over the washed pulp sheets.
The hot black liquor is exposed to large quantities of air and generates.
large volumes of steam. The exposure of the liquor to ambient air has two
effects: 1) the reduced sulfur in the black liquor is partially oxidized,
and 2) a portion of the TRS compounds is flashed through the washer vents.
Three types of pulp washers are commonly used: vacuum washers, pressure
washers, and diffusion washers. In the vacuum washer, the most common (Figure
12
3-26), the washer system has three sections: a drum, liquor tanks, and a
foam tank. The foam tank and drum, which are vented to the atmosphere, are
sources of TRS emissions. The drums are typically hooded in groups and vented
directly through the mill building roof. The volume of gases exhausted from
the drum hood is substantial (48,000 to 190,000 acfm/TADP) and the concentra-*
tion of TRS gases is low (25 to 30 ppm).12
The pressure washer is similar to the vacuum washer, except that the i
drums are enclosed with recirculating pressure blowers (Figure 3-27). The
quantity of air vented from the drum is smaller (less ambient air exhausted
from the building), but the TRS levels in the gases are higher. Because the
gas volume is smaller, TRS emissions from pressure washers can generally be
controlled by incineration.
Diffusion washing usually is accomplished in a closed reactor (Figure
3-28) without air, which reduces liquor oxidation and odor emissions. The
washers may be batch or continuous.
Black liquor recovered from the washers is generally between 10 and 15
percent solids by weight and is only partially oxidized. The liquor also
contains recoverable amounts of tall oil soap, waxes, and fatty acids.
The washer lines contain multiple stages that have specific design pulp
process weights and liquor flow rates. Operation at excessive pulp rates can
result in incomplete chemical removal from the pulp and/or reduced chemical
recovery through the washing liquor.
3.2.1.6 Black Liquor Concentration (Evaporation)—
The most common method of concentrating the black liquor from the brown '
stock washers is indirect contact with steam. Most sources prefer a multiple-
effect evaporator (MEE), which is generally composed of three to seven stages.
84
-------
HOOD
VENT
—*[
4 WASH
3 WASH
2 WASHED PULP
PULP
VENT
QO) WEAK
LIQUOR
POINTS OF POSSIBLE ODOR RELEASE ARE ENCIRCLED
Figure 3-26. Vaccum washer flow sheet.
12
85
-------
1
r
'
LIQ.
TANK
1
,r
LIQ.
TANK
1
«
LIQ.
TANK
t
r
^
c
t
r
ft*
FOAM
TANK
-«-»•
(5) VENT
4 WASH
3 WASH
2 WASHED PULP
(?) VENT
GO) WEAK LIQUOR
POINTS OF POSSIBLE ODOR RELEASE ARE ENCIRCLED
Figure 3-27. Pressure washers flow sheet.
12
86
-------
1
1
! i
DIFF.
WASH
1
! i
DIFF.
WASH
2
1
•«•••
I
i
i 4r
DIFF.
WASH
3
^. 9 UIACU
..fe. (Toi WFAK
BATCH DIFFUSION WASHERS FLOW SHEET
(5) VENT
WEAK LIQUOR
3 WASH
2 WASHED PULP
1 PULP + UOUOR
POINTS OF POSSIBLE ODOR RELEASE ARE ENCIRCLED.
CONTINUOUS DIFFUSION WASHERS FLOW SHEET
Figure 3-28. Diffusion washer flow sheet.12
87
-------
18
Each line in the evaporator is normally equipped with condensate flashing,
liquor preheating, hot water generating, degassing, tail steam condensing,
and vacuum generating equipment (steam ejectors). Methods of liquor movement
can be falling film, rising film, or forced circulation. The majority of the
effects are a long-tube vertical and a shell design. As Figure 3-29 shows,
the most concentrated liquor is heated by high-pressure steam. The resulting
steam is then used to heat the liquor in the next effect. Each effect operates
at successively lower pressure, thus lowering the boiling point of the liquor.
Evaporated water vapor from the last effect is condensed in either
barometric condensers (direct contact) or tube and shell noncontact (Figure
12
3-30) condensers with steam ejectors. The vapor removal rate of the con-
densers must be high enough to maintain a vacuum above the vapor space in
the last effect. Noncondensable gases are vented from the tail gas condenser '
with a small steam ejector.
Storage tanks are used between the middle effects (liquor solids 28 to 30%)
to allow decanting and skimming of tall oil soap from the concentrated liquors.
The size (volume/height) of the soap skim tanks is defined by the vertical
velocity of soap micelles in the liquor. The vertical velocity is affected
18 19
by liquor pH, temperature, and micelle size. *
The storage tanks and condensate hot wells, which are generally vented
to the atmosphere, are a source of TRS emissions. Noncondensable gases
generated in each effect must be vented to maintain heat transfer coefficients
and to prevent corrosion of the shell. The venting may be accomplished by
either single- or two-stage systems. Single-stage venting consists of the
direct venting of each effect to a central manifold. The gases are typically
vented to the secondary condenser, and noncondenables are released through the
12
hot well or steam jet ejector.
In two-stage venting the first two stages after the liquor feed stage are
vented. The noncondensable gases (5 to 15% of the total vapor flow) are
vented to separate condensers that are then vented to the main vacuum system.
Black liquor generally enters the MEE line at 12 to 15 percent solids
and leaves the last effect at 48 to 55 percent solids. Flow rate (gallons/
minute) and liquor weight through the effects are measured in total liquor
flow in and out of the line plus liquor percent solids in and out of the line.
88
-------
VAPOR
PROCESS
STEAM
CONDENSATE
THICK BLACK LIQUOR
WEAK BLACK LIQUOR
Figure 3-29. Multiple-effect, long-tube vertical evaporators
(backward feed).*
89
-------
STEAM
20
22
'4^) Ub
SEC'Y CONDENSATES
18 WARM WATER
17 FRESH WATER
16 FRESHWATER
14 FRESHWATER
15 STEAM
VENT
THICK
LIQUOR
WEAK
LIQUOR
SOAP
6+7+81)
TAILCONDENSATES
POINTS OF POSSIBLE ODOR RELEASE ARE ENCIRCLED.
Figure 3-30. Multi-effect, vacuum evaporation plant flow sheet.12
-------
Normally, the steam flow to the effect is also measured. Figure 3-31 shows a
typical temperature chart for liquor steam system.
The amount of noncondensable gases vented from the hot well and tail gas
condenser may vary from mill to mill, depending on the tightness of the system
and the amount of air inleakage into the system.20'21 Published data indicate
qas volumes between 10 and 400 ft3/TADP for hot wells. A reasonable value
• 3 12
for a southern softwood is on the order of 32 ft /TADP.
3.2.1.7 Condensate Stripping—
Condensates from the digesters and evaporators contain odorous compounds
that may be lost to the atmosphere during water treatment and storage. As a
result, many new mills operate condensate treatment plants to lower water
treatment biological oxygen demand (BOD) and reduce the potential loss to the
atmosphere.
The organic compounds are divided into two classes, those that are wood
oil bases containing no sulfur and those that are of the TRS type. The latter
class generally has a very low odor threshold and is maladorous. Table 3-3
• 22 23 24
lists the main components of general mill condensate. ' *
The composition and amounts of condensates are a function of process
operation, pulp yield, cooking chemicals, cooking time, and dilution. Table
3-4 presents species concentration data obtained from 10 mills.
There are several ways to control condensate odor problems: reduction
in quantity, segregation of clean condensate, weak liquor oxidation, chlori-
nation, and air or steam stripping. Figure 3-32 shows a typical air
stripping condensate plant flow sheet.12 The condensates are acidified and
passed through a separation column. Overall TRS removal efficiency can be as
12
high as 95 percent.
Steam stripping is similar, but the column is steam-heated and equipped
with primary and secondary noncontact condensers (Figure 3-33). The column
is generally a bubble-cap type containing 10 to 20 trays. '
Removal efficiency is a function of the steam-to-condensate ratio
(Figure 3-34). The pH of the condensate feed also influences H2S removal
rates.26'27 Noncondensable gases are vented from the condensers and usually
28
incinerated in the lime kiln or power boiler;
91
-------
Figure 3-31. Chart of evaporator temperatures.
Source: Joint Textbook Committee of the Paper Industry, the Pulping of Wood,
1969. Reproduced with permission of McGraw-Hill.
92
-------
oo 03 04.
TABLE 3-3. MAIN COMPONENTS OF TYPICAL KRAFT MILL CONDENSATES"'"'
No.
1
2
3
4
5
6
7
8
Component
CH3pH
CH3CH2OH
CH2COCH3
Turpentine
(Pinene)
H2S
CH3SH
CH3SCH3
CH3SSCH3
Boiling point,
°r
64.7
78.5
56.5
154
-59.6
7.6
37.5
117
(°F)
(148.5)
(173.3)
(133.7)
(309)
(-75)
(45.7)
(99.5)
(243)
BOD,
kg/ kg
1.00
1.23
0.67
-
0.60
0.07
0.31
0.61
Sulfur,
%
0
0
0
0
94
67
52
68
Odor threshold,
PPb
100,000
10,000
100,000
-
0.4-5
0.4-3
1-10
2-20
TABLE 3-4. TYPICAL KRAFT MILL CONDENSATE COMPOSITIONS, MEAN VALUES
FOR 10 MILLS23
Condensate
compound
H2S
CH3SH
CH3SCH3
CH3SSCH3
Total S
CH3OH
CH3CH2OH
CH3COCH3
Total BOD
Turpentine
decanter,
mg/liter
90
250
400
130
. 550
6,500
1,600
160
860
Means: Digester
blow, mg/liter
60
80
70
50
180
4,300
500
40
490
Evaporator
effects ,
mq/ liter
40
10 '
5
5
51
10,000
60
6
1,070
Evaporator
hotwel 1 ,
mq/ liter
100
40
7
15
135
1,000
40
10
1,060
93
-------
STRIPPING
COLUMN
8)VENT
5) FOUL CONDENSATES
STORE
TANK
Jl
2 REST ACID
L_r
4 AIR
BLOWER
•> 1 CLEAN CONDENSATES
POINTS OF POSSIBLE ODOR RELEASE ARE ENCIRCLEa
i ?
Figure 3-32. Contaminated condensates air stripping plant flow sheet.
94
-------
SECONDARY i
CONDENSER '
PRIMARY
CONDENSER
8 VENT
7 FRESH WATER
6 WARM WATER
5 TURPENTINE PHASE
HEAT
EXCHANGER
4 STEAM
2 REST ACID
1 CLEAN CONDENSATES
Figure 3-33.
Contaminated condensates steam stripping
plant flow sheet.12
95
-------
o
§
o
U_
U_
LU
LJ
or
100 -T
80 -
60
40 -
20 -
2 4 6 8
STEAM/CONDENSATE, AS %
10
Figure 3-34. Stripping efficiency for different steam-condensate
ratios at 10 theoretical plants.12
96
-------
3.2.1.8 Black Liquor Oxidation—
If a contact evaporation is used in the chemical recovery process, the
black liquor that is separated from the pulp during the washing cycle is nor-
mally oxidized with air to control the sulfide level and prevent the release
of odorous compounds. As Figure 3-35 shows, the process consists of counter-
currently passing the black liquor through an air stream by means of a porous
diffuser,29 sieve tray tower,30 packed tower or agitated air sparge.
The
first two techniques are more efficient than the last two. If an inexpensive
source of molecular oxygen is available, it can provide a very efficient means
of oxidation. The oxidation reaction consists of converting sodium sulfide
(Na2S) to sodium thiosulfate (Na2S203) according to the reaction: 2Na2S +
202 + H20 -> Na2S203 + 2NaOH. The benefits derived from black liquor oxida-
tion are that it increases the black liquor solids and therefore, improves
the multiple-effect evaporation process, reduces odors, reduces the corrosion
rate of metal evaporating surfaces, reduces the chemical makeup requirements
for Na2S04 and CaO, and increases the sulfidity of the white and green liquor,
which affects pulp yield and quality. It may, however, reduce the heating
value of the black liquor, and foaming may occur because of the presence of
fatty acid sodium salts, particularly when using softwoods. The use of mo-
lecular oxygen in place of air, however, has proved to be effective in con-
trolling this problem. The oxidation may be carried out as weak liquor (prior
to MEE) or as strong liquor after concentration. Figure 3-36 shows a typical
strong liquor oxidation system. •
The oxidation efficiency of the process is measured by the change in
sulfide (Na2S) of the black liquor in grams/liter. Conversion rates in.
excess of 99 percent are common.35'36 Acceptable sulfidity levels in the
liquor are on the order of 0.1 g/1 to achieve TRS emission limits from the
recovery boiler. The efficiency is a function of residence time, liquor mix-
ing, liquor height, air flow rate, and inlet Na2S concentration. The oxygen
requirement for conversion is generally measured in cubic feet of 02/ton of
Na2S, pounds of 02/ton of pulp, or cubic feet of 02/gal of black liquor.
Emissions of TRS generally come from the tank vent and are on the order
of 10 to 500 ppm. The gas volume is between 16,000 and 48,000 ft /TADP.12
.97
-------
FOAM
BREAKER
EXHAUST
AIR
AGITATOR
AIR
BLACK
LIQUOR
INLET
BLACK
LIQUOR
EXIT
TURBINE
AERATOR
Figure 3-35. Agitated air sparging system for black liquor oxidation.12
BLACK
LIQUOR
EXIT
L JL_I
NUMBER 2
QXIQIZER
STRONG BLACK
LIQUOR STORAGE
BLACK
LIQUOR
INLET
BLOWER PUMP
BLOWER PUMP
»
Figure 3-36. Champion two stage unagitated strong black liquor
oxidation system.31*
98
-------
3.2.2 Sources of Emissions and Control
11 •— ' ' '- ...-••-•-•- r,, * JT _. ^,-
Emissions from the pulping department consist primarily of odorous gases.
These gases are generally controlled by one of three methods: 1) reduction in
quantity through process operation, 2) scrubbing with fresh water or alkaline
solutions, or 3) incineration of noncondensable gases. Table 3-5 summarizes
the control options that are available for each of the emission sources within
the pulping department.
TABLE 3-5. SUMMARY OF TRS CONTROL OPTIONS FOR PULPING DEPARTMENT
12
Emission source
Control option
Digester gases (blow and relief)
Washer vents
Evaporator gases
Condensate stripping
Condensate water
Black liquor oxidation (tower vent)
Tall oil reactor vent
Incineration
Incineration3
Scrubbing
Incineration
Incineration
Steam stripping
Air stripping
Incineration
Scrubbing
Diffusion and pressure washers.
3.2.2.1 Digester and Blow Tanks--
Digester and blow tank noncondensable gases are generally low in volu-
metric flow rates, but high in such TRS compounds as methyl mercaptan,
dimethyl sulfide, and dimethyl disulfide. The emissions from the digester and
blow tanks can range from 0.5 to 5.0 Ib of sulfur per ton of pulp (10 to
1000,000 ppm volume).
Gases from the relief system pass through the turpentine recovery con-
denser and decanting system,.and "the flow is relatively constant at low volu-
metric flow rates. When batch digesters are used, the emissions from the blow
tank are periodic and can occur up to 120 times per day (depending on number
of digesters and the cooking time). The time during which peak flow occurs,
99
-------
however, is only K) to .20 minutes per digester cycle.13 The volume of gas
released depends on the digester volume, the gas temperature, the moisture !
content, the condition of hot water accumulator condensers, and the amount of
air and inert gases in the system.
The amount of gas generated also may be affected by upsets in the system.
The normal practice is to blow only one digester at a time. If an upset occurs,
a simultaneous blow of two digesters will generally overload the primary and
secondary condensers and result in an increase in noncondensable gas treat-
ment volume. Table 3-6 presents typical volumes during average, maximum, and
upset conditions.
13
TABLE 3-6. GAS FLOW RATES FROM BATCH DIGESTER13
Condition
Average
Maximum
Upset
Digester blow
ft3/blow
1,600
4,000
125,000
ft3/TADP
128
320
10,250
The amount of flow also varies between digester types, with the continuous
digester generating the lowest emissions. Table 3-7 presents comparative data
12
for both batch relief and continuous digester systems.
TABLE 3-7. TYPICAL RANGES OF DIGESTER NONCONDENSABLE GAS FLOW RATES12
Source
Flow rate ft3/TADP
Batch blow
Batch relief
Continuous
15,200 - 203,500
10 - 3,040
20 - 200
Because of the nature of the digester TRS emissions, scrubbing is not an
effective control option. Scrubbing is effective in removing acidic TRS species,
but it is not effective in removing other TRS species. For effective
100...
-------
destruction of all the gases, the noncondensable stream usually is incinerated
in either a lime kiln or a wood-waste-fired boiler.
Because of the range of flow rates expected from the batch digester
blow, it is not economically feasible to design for peak flow conditions. In
most systems a gas accumulator is used to collect the blow gases and retain
37
them to provide uniform flow to the incinerator.
The most common type of flow equalization system is the vaporsphere
(Figure 3-37). The sphere contains a diaphragm and flow control system for
venting collected gases to the incinerator. Vaporspheres are typically 10,000
ft3.17'37
A second type of gas holder is a floating-cover system in which a water
seal is used (Figure 3-38). Either system must be maintained to keep
the gas concentration below or above the flammability limits. Table 3-8
presents the acceptable flammability limits.38 Duct design and flash back
systems are used to prevent explosions. Table 3-9 provides typical flame-
go
spreading velocities.
3.2.2.2 Washer Hood Vents-
Existing hood vents over vacuum washers are not generally controlled
because of the low concentrations of the TRS in the gas stream, the large gas
volume, and the high amounts of water vapor. TRS control is required of new
washers under NSPS. The limit is 5 ppm by volume corrected to actual oxygen
content of the gas stream. Emissions from pressure-type washers and diffusion
washers may be condensed to remove water and noncondensable gases and then
vented to the lime kiln or wood-waste-fired boiler to be incinerated. Because
the volumeric flow rate of the gas stream is constant, the use of a gas .equali-
zation system is not required.
3.2.2.3 Evaporator Condenser--
Gasesxfrom multiple-effect evaporators (tail gas and hot well) are
generally condensed to remove water and condensable organics. The remaining
gas may be scrubbed or incinerated in the lime kiln. The volume of gas re-
quiring treatment varies greatly !and depends on evaporator design, condenser
design, condenser temperature, and the amount of infiltrated air. Table 3-10
12
presents a range of gas volumes for the two types of condensers.
101
-------
Diaphragm
Vacuum
Pressure
Relief
Relief Gases
Sliding
Weight
Flow
Control
Figure 3-37. Vaporsphere flow equalization gas holders.17'37
Blow
By-pass
Vent
Vacuum
Pressure
Relief
System
Handling
System
Figure 3-38. Floating cover flow equalization gas holders.17'37
102
-------
TABLE 3-8. FLAMMABILITY LIMITS IN AIR FOR KRAFT NONCONDENSABLE GASES
38
H2S
CH3SH
CH3SCH3
Terpene
Concentration, % by volume
Lower
4.3
3.9
2.2
0.8
Upper
45.0
21.8 ~
19.7
•*••
TABLE 3-9. FLAME-SPREADING VELOCITIES OF AIR-MERCAPTAN MIXTURES
39
Mercaptan
% by vol.
18.9
22.8
23.1
23.7
25.5
25.7
Flame velocity,
ft/s
1.8
1.5
1.3
1.2
0.6
0.5
TABLE 3-10. TYPICAL RANGES OF EVAPORATOR NONCONDENSABLE GAS FLOW RATES
12
Source
Surface condenser
Jet condenser
Flow rate, ft3/TADP
20 to 420
10 to 100
The primary sulfur compounds that are emitted from the MEE are H2S and
CH3SH. The concentrations range from 0.2 to 3.0 Ib sulfur per ton of pulp
(300 to 25,000 ppm volume). The alkaline scrubber is effective in reducing
TRS compounds that are acidic (H2S and CH3SH), but it is not effective in re-
moving CH3SCH3 and CH3SSCH3.17
103
-------
3.2.2.4 Condensate Stripping-- :
Noncondensable gases from the air stripping column or steam stripping ''
column, which are generally low in volume, are incinerated in the lime kiln or
wood-waste-fired boilers. :
3.2.2.5 Black Liquor Oxidation—
The gas stream from the oxidation tower vent generally contains a low
concentration of TRS and is saturated with water vapor. This gas stream is
generally not controlled. If control is necessary, however, a primary con- i
denser is usually used to remove the water vapor. The gas stream volume is
generally too large to be incinerated in lime kilns, but it can be vented to
and incinerated in a wood-waste-fired boiler.
3.2.2,6 TRS Scrubbers-
Scrubbing a noncondensable gas stream to remove TRS compounds is generally
accomplished with packed-bed scrubbers using a countercurrent flow of an ;
alkaline liquid (white water). The packing medium can be gravel, stone, or
packing rings. The scrubbers can be constructed of mild steel or stainless
steel (304 or 316L). The stainless steels are used where corrosion is
expected. .
Figure 3-39 shows a typical two stage packed bed scrubber that is appli-
cable to noncondensable gases from multiple-effect evaporators. General
liquor-to-gas ratios are on the order of 300 gal/1000 acfm of noncondensable
gas. The superficial gas velocity is approximately 50 ft/min through the
packed section. A mist eliminator is generally used to remove mist carryover.
3.2.2.7 Incineration Systems—
A noncondensable gas incineration system may consist of a flow equaliza-
tion system, condenser scrubber, rupture discs, fan, flow recorder, condensate
traps, bypass vent system, flame arrestors, and the incinerator. The in-
cinerator may be the lime kiln, separate incinerator, or wood-waste-fired
boiler (Figure 3-40).
3.2.2.8 Lime Kiln Incineration—
The noncondensable gases are introduced either into a dedicated burner or
into the primary air' system of the kiln burner for effective mixing with air
to ensure complete combustion. Systems using a dedicated burner are generally
40
104
-------
SPRAY
NOZZLES"
SCRUBBED GASES
A
4- _
T
600mm
600
•mm.
600mm
t
600mm
600mm
600mm
1
/ \
^.rSv^
\
WHITE LIQUOR
RETURN
I.Q 50mm
(2 in.)
I. D. 150mm (6 in.)
I.Q 600mm (24 in.)
MIST ELIMINATOR
25mm (I in.) RASCHIG RING
I.D. 150mm (6 in.)
HOTWELL GASES
MAX. FLOW - lOOOm'/h
(590ft5/min)
COOLED WHITE LIQUOR
MAX. FLOW - 50 mVh
(30ft3/min)
Figure 3-39. Hotwell gas scrubber for 100 metric tons per hour 9400 gpm)
evaporation plant for H2S-Separation of 95 percent or more.40
105
-------
o
a>
Gases From Multiple-
Effect Evaporators
Gases From Turpentine
Condenser
Relief Gases
from,
Kamyr Digester
Blow & Relief
Gases From 4
Batch Digesters
Defibrator Off-
Gases
Heat
Accumulator
Entrapment
Separator
Floating-Cover
Gas Holder
Vent
t
J
Figure 3-40. Noncondensable gas incineration system.
1-7
-------
used on a lean gas collection system. Rich gas streams are generally incinerated
in the kiln. The dilution factor of the gases entering the kiln is about 50
to 1, and the gas velocity is at least 30 ft/s.12'41
The gases are incinerated in the lime kiln at temperatures between 2200°
and 2550°F. Residence times are generally greater than 0.5 s above 1000°F and
may exceed 1.5 s in longer kilns. The S02 generated from the combustion is
adsorbed by the lime to form calcium sulfite (Ca2 SOg).
3.2.3 Malfunctions
The following subsections discuss process and control equipment malfunc-
tions that have an impact on uncontrolled and controlled emission rates.
3.2.3.1 Digester Relief Systems--
Most of the operating problems that occur with digester relief systems
have an impact on pulp quality, but some also have an impact on atmospheric
emissions. Entrainment of air into the digester system as a result of improper
charging can produce higher than normal noncondensable gas volumes. In some
mills excessive liquor carryover and foaming can cause plugging of the system
2
and fouling of the turpentine condenser. Foaming can occur as a result of
resins in softwoods and entrained air. Table 3-11 summarizes the malfunc-
tions that can occur in the digester relief system.
2,42
3.2.3.2 Digester Blow System—
The primary malfunction that occurs in digester blow systems is the
fouling of the hot water accumulator condensers or pluggage of the blow tank.'
As the pluggage progresses the ability of the condenser to condense the flashed
steam is reduced. This results in increased noncondensable gas volumes and
overloading of the vent control system and gas accumulator. Serious malfunc-
tion of the condenser may cause vent gas system to be bypassed as a result of
top much pressure. Figure 3-41 shows the effect of heat exchanger malfunc-
17 42
tions on digester noncondensable gas volumes. '
Simultaneous blowing of two or more digesters can produce high noncon-
densable gas volumes and reduce the condenser's efficiency. Because of this
increased volume, the vent system may be bypassed to the atmosphere.
If the condensers become seriously fouled, the back pressure in the digester
blow system can cause fugitive emissions through duct flanges, access hatches,
107
-------
TABLE 3-11. MALFUNCTIONS THAT MAY OCCUR IN DIGESTER RELIEF
TURPENTINE RECOVERY SYSTEMS2'1*2
Malfunction
Primary effect
Result
Liquor carryover
Low water flow rate to
turpentine condenser
Failure to close di-
gester valve after blow
Digester relief line plug-
gage
Trapped air in chips due
to poor digester filling
Liquor foaming when cooking
resinous woods
f-
Temperature condenser plug-
gage or fouling
Increased condenser water
temperature
Fouling of digester blow
line
Digester overpressure
and emergency bypass
relief or premature
digester blow which
will increase TRS
emissions
Reduced condenser heat
transfer and increased
noncondensable gas
temperature yielding
increased equilibrium;
TRS vapor pressure and
TRS emissions
Increased equilibrium
TRS vapor pressure and
TRS emissions
Loss of digester cook-
ing liquor, overpres-
sure during blowing,
increased digester blow
volume which will in-
crease TRS emissions
108
-------
I
20
io H
o
23
j
V( t) dt = 550m3/blow
MEAN - 1400 m3/h
25
5 10 15 20
TIME, MINUTES
CASE I. NORMAL OPERATIONS
40 -
^ 30 -
*>.
V
WQ 20 -
1 »H
23
)5 V( t ) dt = 300m3/blow
20
25
0 5 10 15
TIME, MINUTES
CASE 2. MALFUNCTION OF BLOW HEAT RECOVERY HEAT EXCHANGERS
50 -
40 -
30 -
20 -
10 -
0
23
5 V(t) dt = 4500m3/blow
\^
MEAN-ll,700m3/h
5
20
25
10 15
TIME, MINUTES
CASE 3. INCREASING MALFUNCTION OF HEAT EXCHANGERS
Figure 3-41.
Kraft batch dige'ster blow gas flow after condensing
and without equalization.17
109
-------
etc. Bypass of the primary condenser has been observed as a result of severe
corrosion of the blow tank and associated ductwork. It should be noted that
the concentration of TRS in these streams is very high, and the inspector should
wear appropriate respirators in these areas when fugitive losses are noted.
The blow tanks are pressure-rated vessels and can be weakened by corrosion
and repeated overpressure. Table 3-12 summarizes the malfunctions that may
occur in the digester blow system.
TABLE 3-12. MALFUNCTIONS THAT MAY OCCUR IN DIGESTER BLOW
TANK HOT WATER ACCUMULATOR SYSTEMS2*17>*2
Malfunction
Primary effect
Result
Fouling of primary hot
water accumulator, pri-
mary and secondary con-
densers
Low water flow rate to
primary and secondary
condensers or hot water
accumulator
Simultaneous digester
blow
Corrosion of blow tank
and ductwork
Reduced heat transfer and
loss of condensate
Increased condenser water
temperature
Overloading of primary and
secondary condenser and/or
TRS incineration system
Fugitive condensable gases
Increased digester blow
volume and increased
TRS emissions
Increased equilibrium
TRS vapor pressure and
TRS emission rates
Increased noncondensable
gas volume, carryover
of condensate into vent
gas system, overpres-
sure bypass to at-
mosphere, loss of cook-
ing chemicals and in-
creased TRS emissions
Increased TRS emissions
as a result of vent
system by pass
3.2.3.3 Multiple-Effect Evaporators—
The primary malfunctions that affect evaporators are associated with foul-
2 12 21
ing of the evaporator body and ambient air infiltration. ' ' Both can
result in reduced evaporator efficiency. Ambient air infiltration results
in increased noncondensable gas volumes and reduced condenser performance,
which can result in low water temperature and increased equilibrium TRS vapor
pressures.43 In systems that use alkaline scrubber noncondensable gas control,
110
-------
loss of water flow-, reduction in alkali content, or an increase in gas volume
can result in lower mass transfer rates and increased TRS emission rates.
Table 3-13 summarizes the malfunctions that may occur in multiple-effect
? i? /n M
evaporator systems. §1*HfH
3.2.3.4 Black Liquor Oxidation-
Malfunctions in the black liquor oxidation system generally result in
reduced oxidation efficiency (high outlet sulfidity) which produces elevated
TRS concentrations from the recovery boiler. Malfunctions include reduced
air flow volume, pluggage of oxygen injection system, increased black liquor
flow rates, liquor foaming, and increased inlet liquor sulfidity. Table 3-14
33
summarizes the malfunctions that may occur.
3.2.3.5 TRS Vent Control System--
The TRS vent control system collects noncondensable gases from the blow
tank, hot water accumulator, and other miscellaneous systems. Increases in
gas volume may exceed system design and limit the ability of the sources to
be evacuated. Increased flow rates result in higher static pressure losses
through the lines and may require larger gas-moving equipment.
For safety reasons, the concentration of the organic vapor must be main-
tained either below the lower explosive limits~or above the upper explosive
limits. Flash-back flame protection devices must be used to prevent explosions.
Entrained moisture must be removed from the gas stream to prevent cooling
or blowout of the incinerator flame. Large variations in flow must be mini-
mized to stabilize flame temperature and maintain acceptable residence time.'
Table 3-15 summarizes the malfunctions associated with the noncondensable
gas incineration system.
3.2.4 Inspection of Pulping Department
Because emissions from the pulping department are gaseous in nature and
the gases are potentially hazardous at high concentrations, physical measure-
ment and evaluation of emission sources are limited.
Inspections of these sources are generally based on information concerning
equipment specifications, process procedures, process weights, and/or control
equipment bypass and malfunction. The ability of the inspector to measure
TRS concentrations with hand-held equipment is extremely limited. Errors in
45
111
-------
TABLE 3-13. MALFUNCTIONS THAT MAY OCCUR IN MULTIPLE-EFFECT
EVAPORATOR SYSTEMS2»12»"3»*"
Malfunction
Primary effect
Result
Evaporator fouling and
scaling
Ambient air inleakage
into evaporator body
Low condenser water
flow rate
High inlet condenser
water temperature
Reduced scrubber water
flow rate
Increased scrubber gas
volume
Scrubber packing flow
channeling
Liquor foaming
Separation of soap
from liquor
Carbonaceous deposits
in evaporator body
Reduced evaporator effi-
ciency
Increased noncondensable
gas volume
Increased condenser outlet
water temperature
Increased condenser outlet
water temperature
Reduced liquor-to-gas ratio,
reduced mass transfer and
adsorption
Increased superficial veloc-
ity through scrubber, re-
duced liquor-to-gas ratio,
reduced mass transfer and
adsorption, increased mist
eliminator carryover
Reduced liquor-to-gas con-
tact, reduced mass transfer
and adsorption
Liquor carryover and re-
duced heat transfer
Liquor carryover and re-
duced heat transfer
Liquor carryover and re-
duced heat transfer
Reduced boiler effi- .
ciency and increased
TRS emissions
Reduced condenser
efficiency and in-
creased TRS emissions
Increased condenser
equilibrum TRS vapor
pressure and TRS emis-
sions
Increased condenser
equilibrium TRS vapor
pressure and TRS emis-
sions
Increased TRS emissions
and decreased removal
efficiency
Increased TRS emissions
and decreased removal
efficiency
Increased TRS emissions
and reduced removal
efficiency
Reduced evaporator
efficiency, decreased
final liquor solids
attainable and in-
creased TRS emissions
Reduced evaporator
efficiency and increased
TRS emissions . '
Reduced evaporator
efficiency and in-
creased TRS emissions
112
-------
TABLE 3-14. MALFUNCTIONS THAT MAY OCCUR IN BLACK LIQUOR OXIDATION
SYSTEMS
Malfunction
Primary effect
Result
Reduced air flow volume
through oxidation tank
Plugging of air sparge
Increased liquor flow
rate
Liquor foaming
Increased inlet liquor
sulfidity
Reduced oxidation of sodium
sulfide
Stratification of liquor air
column and reduced contact
Decreased liquor residence
time and oxygen adsorption
Foam carryover limits
system liquor volume and
blowing rates
Increased outlet liquor
sulfidity and TRS emis-
sions
Increased outlet liquor
sulfidity and TRS emis-
sions
Increased outlet liquor
sulfidity and TRS emis-
sions
Increased outlet liquor
sulfidity and TRS emis-
sions
Increased outlet liquor
sulfidity and TRS emis-
sions
TABLE 3-15. MALFUNCTIONS THAT MAY OCCUR IN NONCONDENSABLE GAS
INCINERATION SYSTEM
Malfunction
Primary effect
Result
Increased flow volume
Operation between lower
and upper explosive
limits
Low gas flow velocity
Entrained moisture
Exceeding system design,
reduced residence time
Operation below flame
propagation velocity
Flame blowout, reduced
flame temperature
Fugitive TRS emissions,
increased TRS emissions
due to incomplete com-
bustion
Potential for explosion
Potential for explo-
sion and/or fire
Increased TRS emissions
as a result of incom-
plete combustion and
potential for explo-
sion
113
-------
measurement due to interference are a major concern. In addition, some sources
are vented through small pipes at high velocity, which makes it extremely
difficult to measure flow.
Because the emissions are gaseous, visible emission evaluations are not
useful except for fugitive emissions in the area between the blow tank and
hot water accumulator. In this area the presence of a leak is observed as a
condensed water plume.
Special care must be taken in such areas to avoid confined spaces in which
toxic concentrations of hydrogen sulfide may accumulate. The inspector should
carefully avoid any contact with process vent plumes that may contain TRS ;
gases.
The following discussions concern information that should be documented
for each process step to allow calculation of process weights, operating cycles,
and control equipment performance. In many instances, the data required do
not apply to the calculation of emission rates from a specific source, but are
necessary for analysis of other process operations in the chemical recovery
cycle.
3.2.4.1 Digester—
The inspector should determine the cooking cycle of the digester, the
charge weight, the white liquor charge weight, the liquor sulfidity, and the
digester volume. The number of cooks per day is an indication of the level :
of operation, assuming reasonably constant pulp yield.
The volume of the blow tank and blow time, together with the calculated
blow volume, can be used to calculate the size of primary and secondary con- '
densers in the hot water accumulator. The occurrence of simultaneous digester
blows can overload the system. The inspector should observe the blowing opera-
tion to determine if condensable gases are being lost from ducts as a result
of overpressure in the system. These fugitives bypass the TRS nbncondensable
gas system.
Malfunctions concerning the blow tank system occur primarily as a result
of condenser plugging. The final equilibrium vapor pressure of TRS gas in the
gas stream is defined by the secondary condenser temperature and gas stream
pressure. In cases of condenser plugging, the heat exchange rate across the
condenser and the outlet water temperature are reduced. The inspector should
114
-------
review operating charts for the condenser water temperature differential (inlet/
outlet) to determine if a trend is occurring that would indicate fouling.
In some systems the noncondensable gas volume is monitored and recorded
by automatic instruments. The inspector should review these data to~determine
if peak gas flow is within the design specifications of the incineration system.
It is suggested that the inspector request and maintain a process flow diagram
for the noncondensable gas system. Table 3-16 provides an inspection checklist
for the digester system.
3.2.4.2 Digester Relief—
The digester relief system contains significant quantities of TRS compounds
and turpentine. The amount of turpentine is a function of cooking variables,
wood species, and time of year. The resinous woods produce the highest quantity
of turpentine. To evaluate the Operation of the relief system, it is necessary
to know the relief gas flow volume and cycle time. The peak loading and highest
temperature represent the most critical period. The digester relief gases are
condensed in a surface condenser. The inspector should determine the final
outlet gas temperature of the condenser (inlet/outlet), and the vacuum if steam
ejectors are used. Changes in outlet gas temperature can indicate higher
emission levels as the vapor pressure increases. Table 3-17 is an inspection
checklist for the digester relief system.
3.2.4.3 Brown Stock Washers--
The inspection of brown stock washers consists of determining source loca-
tion, process weight, and gas volumes. Emission levels are generally low and
not controlled. Table 3-18 is an inspection checklist for the brown stock
washer system.
3.2.4.4 Multiple-Effect Evaporators--
The inspection of multiple-effect evaporators consists of documenting the
emission points, process weights, and control equipment used. Where condensers
are used, the inspector should record the cooling water flow rate, condenser
surface area, gas stream pressure (where ejectors are used), and the differ-
ential water temperature (inlet/outlet). Table 3-19 is an inspection check-
list for multiple-effect evaporators.
115
-------
TABLE 3-16. CHECKLIST FOR INSPECTION OF DIGESTER BLOW SYSTEMS
Digesters
Number of digesters
Chip charge weight tons
White liquor charge weight
ft°/batch
Digester volume ft
Cook time min
Blow time min
Cooks per day
Blow Tank
3
Blow tank volume ft
Pre-blow digester temp. °F
Final blow tank temp. °F
Blow time min
Calculated blow volume lb/hc
Hot Water Accumulator
Hot water accumulator condensers
Primary
Secondary
Primary condenser temperature _
Water flow
Type
Type
o
F
gpm
Secondary condenser temperature
Water flow
Surface area
D TRS Control
gpm
ft2
Noncondensable gas volume
Temp.
Pressure
acfm
in H20 (vacuum)
(continued)
116
-------
TABLE 3-16 (continued)
Vapor equalizer Type
Volume
fr
Noncondensable gas treatment method
Incineration
Scrubbing
Incinerator type
Catalyst
Direct fire
Lime kiln
Wood-waste boiler
Fossil fuel boiler
Recovery boiler
Dilution ratio at burner
Explosion protection yes
Type: Flame arresters
Bypass vent
Rupture discs
no
117
-------
TABLE 3-17. CHECKLIST FOR INSPECTION OF DIGESTER RELIEF SYSTEMS
Number of digesters
Relief flow rate
acfm
Noncondensable gas flow
Condenser type
Water flow
acfm
Or-
gpm
Inlet temperature _
Outlet temperature
Surface area
ft'
Operating pressure
Turpentine recovery _
Noncondensable gas treatment
in. H20 (vacuum)
gal/TADP
gph
scrubber
incinerator
none
118
-------
TABLE 3-18.. CHECKLIST FOR INSPECTION OF BROWN STOCK WASHER SYSTEMS
Number of wash lines
Number of stages per line
Washer type vacuum
pressure
batch
diffusion _
continuous
Process weight
TADP/day
Outlet weak black liquor flow
M Ib/h
gpm
Outlet weak black liquor solids
Exhaust hood gas flow rate
Exhaust hood gas temperature
acfm
0,-
Control method
incineration
condenser
none
Noncondensable gas flow rate
Condenser water flow rate
Condenser water inlet temp
Condenser water outlet temp
Condenser surface area
acfm
gpm
Or
ft'
119
-------
TABLE 3-19. CHECKLIST FOR INSPECTION OF MULTIPLE-EFFECT EVAPORATOR SYSTEMS
A.
Multiple-Effect Evaporators
Number of evaporator lines _
Number of effects
Manufacturer
Inlet conditions
Black liquor flow
Black liquor solids
Outlet conditions
Black liquor flow _
Black liquor solids
Soap recovery _
Control type: Tail gas
Hot well
Noncondensable gas control
Scrubber
Incineration
None
Weak black liquor oxidation
B. Noncondensable gas scrubber
Type
Superficial velocity
Liquor to gas ratio
Liquor type
Liquor flow rate
Pressure drop
gpm
M Ib/h
gpm
M lb/he
%
Ib/h
yes
no
ft/min
gal/1000 acfm
water
white liquor
alkaline solution
gpm • •
in. H20
(continued)
120
-------
TABLE 3-19 (continued)
C.
Outlet gas temperature
Liquor alkalinity
PH
Tail Gas Condenser/Hot Well Condenser
Number of stages
Condenser type barometric
Condenser area
Pressure
Water flow
Inlet temp.
' Outlet temp.
Condenser inlet flow
Temperature
Noncondenser gas flow
surface
ft2
in. H20 (vacuum)
gpm
acfm
°F
acfm
Or-
121
-------
3.2.4.5 Black Liquor Oxidation—
Because the black liquor oxidation system determines the emission of TRS
gases from direct-contact recovery boiler systems, the evaluation of the system
is directed toward the final liquor sulfidity. The inspection is directed
toward determining the operating parameters that affect the oxidation effic- :
iency of the system. Table 3-20 is an inspection checklist for black liquor
oxidation systems.
3.2.4.6 Condensate Stripping—
The inspection of the stripping system consists of measuring such key
operating variables as condensate pH, steam flow, air flow, and condensate
flow rate. Table 3-21 is an inspection checklist for stripping systems.
3.3 CHEMICAL RECOVERY
The two major sources within the chemical recovery portion of the kraft
pulp mill are the kraft recovery furnace or boiler and the smelt dissolving
tank. The kraft recovery boiler is used to combust spent liquor from the
pulping process. The smelt dissolving tank, located below the recovery boiler,,
is a large vessel that continuously receives a molten mixture of sodium sul-
fide and sodium carbonate from the floor of the recovery boiler.
3.3.1 Recovery Boiler
The following subsections describe the kraft recovery boiler process,
identify the sources of emissions from a recovery boiler, discuss the con-
trol techniques used to minimize emissions from the recovery boiler, discuss
the possible malfunctions associated with operation of the recovery boiler,
and present inspection procedures for the recovery boiler and associated con-
trol equipment.
The particulate matter and TRS emission rates from the recovery boiler
depend on several interrelated boiler operating variables; e.g., firing rate,
black liquor heat valve, black liquor concentration, total combustion air,
primary air, black liquor sulfur-to-sodium ratio, primary air temperature, char
bed temperature, and black .liquor chlorine content. Section 3.3.1.2 provides,
a detailed discussion of how these variables affect the uncontrolled emission
rate of the boiler and the overall performance of the electrostatic precipita-
tor (ESP) used to control the particulate emissions from a recovery boiler.
122
-------
TABLE 3-20. CHECKLIST FOR INSPECTION OF BLACK LIQUOR OXIDATION SYSTEMS
A.
Black Liquor Oxidation
Oxidation type weak
Air
strong
Oxidation method
Weak black liquor
oxidation (BLOX)
Oxygen
Porous plate diffusers
Sieve tray tower
Packed tower
Agitated air sparging
Rotating fluid contactor
Other
Strong BLOX
Liquor flow rate
Single-stage unagitated air sparging
Two-stage unagitated air sparging
Agitated air sparging
gpm
M Ib/h
Liquor solids content _
Residence-'time h
Inlet liquor sulfidity _
Outlet liquor sulfidity
Oxidation efficiency
Air flow volume
_g/liter
_ g/liter
acfm
Tower vent TRS control
Control type
yes
no
Condenser
Scrubber
Incineration
(continued)
123
-------
TABLE 3-20 (continued)
gpm
0,-
Condenser type _
Surface area _
Water flow
Inlet temp
Outlet temp °F
Operating pressure
Noncondensable gas volume
Temperature
Scrubber type
Superficial velocity
Liquor/gas ratio
Liquor type
in. H20 (vacuum)
acfm
ft/min
gal/1000 acfm
Liquor flow rate
Pressure drop
Water
White liquor
Alkaline solution
gpm
in. H20
Outlet gas temperature
Liquor alkalinity
PH
Incinerator type
Catalyst
Direct fire
Lime kiln
Wood-waste boiler
124
-------
TABLE 3-21. CHECKLIST FOR INSPECTION OF CONDENSATE STRIPPING SYSTEMS
Condensate Stripping Systems
Type Air
Steam
Foul condensate flow rate
Air Stripping
Column type: Tray
Number of trays
Tray distance
Column diameter
Air volume
M Ib/h
gpm
Other
in.
in.
Condensate inlet temperature
Condensate pH ________
Steam stripping
Column type: Tray
Number of trays
Tray distance
Column diameter
Steam flow
Other
in.
in.
Ib/h
Condensate inlet temperature
Condensate pH .
125
-------
TRS and S02 emissions are controlled by the optimization of process variables
that cause sulfurous gases to become chemically combined with sodium to form
a particulate emission which is collected by the ESP.
The malfunctions that the increase emissions from the recovery boiler are
generally divided into two areas: 1) those that occur as a result of furnace
operation, and 2) those that occur as a part of control equipment operation.
Several malfunctions that result from improper boiler operating practices also
have an impact on ESP performance. Section 3.3.1.4 provides a detailed dis-
cussion of both boiler and ESP malfunctions that affect emissions. A summary
table (Table 3-22 on page 184) is provided that lists the key parameters and :
potential malfunctions and their affect on the overall emission rates associated
with recovery boiler operation.
A considerable amount of information is presented in Section 3.3.1.6 on
the inspection of recovery boilers and ESP's. The specific kinds of data that
should be collected (Table 3-23 on page 191) during the inspection is identified
along with a discussion of the procedures and calculations that should be used
to evaluate the inspection data to assess the overall continuous compliance
status of the recovery boiler operation.
3.3.1.1 Process Description--
The kraft recovery boiler or furnace is an indirect water-walled
steam generator used to produce steam and to recover inorganic chemicals
from spent cooking liquors. The boiler consists of a large vertical combus-
tion chamber lined with water tubes.. The heat exchanger section typically
consists of a low-pressure boiler, superheater, and economizer. Figure 3-42
shows a cross section of a modern Babcock and Wilcox (B&W) boiler.12 The fuel
used in the boiler is spent concentrated cooking liquor (black liquor). The
liquor in the burners has a solids content of between 60 and 70 percent (depend-
ing on wood species and yield) and is made up of organic and inorganic frac-
47 40
tions. ' The organic fraction contains lignin derivatives, carbohydrates,
2 49
soap, and waxes. » The inorganic portion consists primarily of sodium
sulfate.
Black liquor is sprayed into the furnace at an elevated level in the com-
bustion chamber through a number of steam or mechanical atomizing nozzles. The
suspended liquor is burned as it falls through the combustion zone. The follow-
ing are the major steps in the combustion process:46
126
-------
Furnace
Slag Screen
Tertiary Air Ports
& Windbox
Black Liquor
Oscillator Burner \
/Oscillator Burner
Secondary Air Ports
/& Windbox
Pin Stud
Upper Limit
Smelt Spouts
&Hood
Primary Air Ports
& Windbox
Green-Liquor
Recirculation
Pumps
Dissolvingj|j Dissolving Tank
Figure 3-42. Cross section of B&W recovery boiler. '
127
-------
o The liquor is dehydrated to form a char.
o The char is burned in a bed at the bottom of the furnace.
o The ash (inorganic portion) remaining after combustion of the char
is exposed to active reducing conditions to convert sodium sulfate ;
and other sodium-sulfur-oxygen compounds to sodium sulfide. ;
o The organic materials in the upper section of the furnace are
oxidized to complete combustion.
Reduced inorganic material (smelt), which consists of a mixture of sodium sul-
fide and sodium carbonate, is continuously drained from the furnace. The
ratio of sodium sulfide to sodium carbonate depends on the temperature and
the ratio of sulfur to sodium in the fired liquor.
Combustion of the char begins on the hearth of the furnace. Air for
combustion is supplied through air ports located in the furnace walls. The
primary air supply is used to initiate char combustion. The primary air
supply, which is introduced in the lower portion of the char, is kept to a
minimum to maintain the necessary reducing conditions to convert the ash to
sodium sulfide. The secondary air supply is located at a higher level in the
furnace to create the oxidizing condition necessary to control the char bed
height. A tertiary air supply may be used to complete combustion in the upper
levels of the furnace and thereby eliminate reduced sulfur compounds. As the;
char bed is burned, the inorganic ash is liquified and drained to the furnace
hearth, where it is reduced.
Figure 3-43 shows the location of air ports for the two major American-
manufacturered recovery boilers. One uses primary, secondary, and tertiary
air in finite zones or levels, and the other uses primary air with secondary
12
air introduced tangentially. Combustion gases produced by the burning of
the liquor are passed through the heat exchanger section of the boiler before
being exhausted to a particulate control device. The gases are cooled to
about 805°F in the boiler tube bank before passing into the economizer.
Temperatures of the gas leaving the economizer, which are about 750°F, are
reduced in either indirect-contact or direct-contact evaporators.
There are three types of direct contact evaporators: cyclone, venturi,
and cascade. Figures 3-44, 3-45, and 3-46 show the basic design for each.
Cyclone evaporators concentrate the black liquor by placing it in contact
128
-------
BABCOCK 8 WILCOX
Steam
12
COMBUSTION ENGINEERING
fSteam
LEGEND
I. Furnace
2. Smelt Spouts
3. Black Liquor Spray Nozzles
4. Primary Air Supply
5. Secondary Air Supply
6. Tertiary Air Supply •
7. Position of Char Bed Burners for Oil or Gas
8. Normal Configuration of Char Bed
8*. Same at Low Primary Air Flow and Pressure
9. Screen Tubes
10. Superheater
II. Boiler Tube Bank
12. Exit to Economizer
SECTION A-A
Figure 3-43. Difference in air systems in U.S. recovery boiler designs.
12
129
-------
WALL-WETTING
NOZZLES
CYCLONE
EVAPORATOR
BLACK-UQUOR
RECIRCULATINS
PUMPS
MECHANICAL
POWER
STRAINER.
UOUOR TRANSFER
TO SALT CAKE
MIX TANK
ILACK-UOUOR
INLET SPRAYS '
FLUE
GASES
Figure 3-44. Cyclone evaporator.
48
Figure 3-45. Cascade evaporator.
130
12
-------
FLUE GAS
FROM BOILERrA
CLEAN GAS OUTLET TO FAN
WALL WASH
RECYCLE LIQUOR
60-70% SOLIDS
190°F
LIQUOR TO BOILER
60-70% SOLIDS
YCLONIC SEPARATOR
MAKE-UP
WATER
LIQUOR FROM
MULTIPLE-EFFECT
EVAPORATOR
40-50% SOLIDS
Figure 3-46. Venturi evaporator.
48
131
-------
with the high-temperature gas stream by using the wetted wall of the cyclone.48
The cyclone removes approximately 50 percent of the uncontrolled particulate
from the gas stream. A venturi evaporator concentrates black liquor by placing
the flue gas in contact with liquor through the generation of liquor droplets.
The droplets are generated through the shearing action of the gas stream as
it passes a weir into which the liquor is being pumped. Venturi evaporators
remove approximately 85 percent of the uncontrolled particulate emissions
generated by the furnace when operated at 4 to 5 in. H20 pressure drop. In
the cascade evaporator, a thin film of liquor coats several tubes rotated
through the flue gas stream. This type of evaporator will generally increase ;
the black liquor soTids content from 48 to 65 percent. The rate of evapora-
tion is related to the flue gas temperature and the cascade rotation rate.
The evaporator can be operated beyond design rates without substantial process
upsets, and can remove up to 50 percent of uncontrolled particulate emissions
from the furnace.
As a result of the direct contact of flue gases with the black liquor in
the evaporator, there is considerable stripping of TRS compounds (i.e., generally
H2S) from the liquor. The loss of TRS compounds is primarily the result of a
reaction between carbon dioxide (C02) and sodium sulfide in the liquor. Gases
from the recovery boiler usually contain 12 to 15 percent C02. When the liquor
is exposed to the gas stream, it absorbs the C02 and the liquor pH is de-
creased. The amount of H2S released from the reaction (Na2S + C02 + H?0 ->
Na2 C03 + H2S) increases as the pH is decreased and the concentration of sodium
sulfide is increased. One method of reducing sodium sulfide concentration is
to convert it to sodium thiosulfate through the,use of black liquor oxidation.
Sulfur dioxide in the flue gas stream also can result in the release of
HgS as a result of decreasing pH. In many operations, however, the level of
S02 is too low to have any significant effect on the H2S emissions.
The amount of oxygen in the flue gas stream has an effect on the genera-
tion of TRS gases from the combustion process but appears to have negligible
effect on the stripping of H2S from the evaporator. :
There are three types of indirect contact evaporators. Figure 3-47 shows
the basic concept employed by each.1 In general, these evaporators evaporate
the liquor by use of a noncontact tube and shell design. Because the black
132
-------
B & W HIGH SOLIDS SYSTEM WITH NO DIRECT CONTACT EVAPORATOR
ADDED
Ur-
i
BLACK l
LIQUOR
~l
•i
PRIMARY
PARTICULATE
CONTROL
DEVICE
•
SECONDARY
PARTICULATE
CONTROL
DEVICE
(OPTIONAL)
CIRCULATION
EVAPORATOR
CE SYSTEM WITH NO FLUE GAS DIRECT CONTACT EVAPORATOR
s I CONTACT ,
^L«- -»JEVAPORATOR |___
! I BLACK LIQUOR
I •« •» mm MB ^ •• *
CE SYSTEM WITH NO DIRECT CONTACT EVAPORATOR
^*—JrF
BLACK LIQUOR
FORCED 1
CIRCULATION '-
EVAPORATOR
Figure 3-47. Three types of indirect contact evaporators.
133
-------
liquor in a noncontact evaporator design does not come in contact with the ,
flue gas, stripping of TRS compounds is prevented. Evaporated water is con-
densed by using a tail gas condenser. Noncondensable gases are directed to
lime kilns or into the furnace primary air system for incineration.
The high temperatures in the furnace char zone result in a partial vapori-
zation of sodium and sulfur from the smelt. The fume is removed from the
furnace with the combustion gases and condenses to a fine particulate consist-
ing of sodium sulfate (Na2S04) and sodium carbonate (Na2C03).
Modern recovery boilers are sized for two process conditions: 1) the
heat input to the furnace, and 2) the weight of the chemicals to be recovered.
Both of these conditions affect the heat release rate as a function of
'3 2
furnace volume (Btu/ft ) and furnace cross section or hearth area (Btu/ft ).
Typical design values are on the order of 9800 Btu/ft (furnace volume) and
o Cf\
900,000 Btu/ft (hearth area). The exact dimensions of the furnace depend
on the elemental composition, solids content, heat value, sulfidity, and
chloride content of the black liquor. Deviations of 10 percent or greater in
50
design variables should be investigated to ensure maximum efficiency. The
manufacturers generally consider the boiler to be overloaded when the
firing rate (Btu/h) exceeds 120 percent of the rated value.
Operation outside of design values can increase emission rates as well
as reduce thermal efficiency as result of tube fouling and reduced smelt
recovery.
3.3.1.2 Sources of Emissions-- •
The uncontrolled partfculate matter and TRS emission rates from the
boiler depend on several interrelated boiler operating variables; e.g., .firing
rate [pounds of black liquor solids (BLS) per hour], black liquor heat value
(Btu/pound BLS), black liquor concentration (% solids), total combustion air
(excess air), primary air (% of total air), black liquor sulfur-to-sodium
ratio (S/Na2), primary air temperature (°F), char bed temperature (°F), and
black liquor chlorine content (55).51»52,53,54 In add1tion to affect1ng the
uncontrolled emission rates, these variables can also reduce ESP performance.
The following discussion addresses both of these effects.
It should be noted that reductions in TRS and S02 emissions from kraft
furnaces result primarily from the optimization of process variables that
cause sulfurous gases to become chemically combined with sodium to form a
134
-------
particulate emission. Operation of the boiler under these process conditions
can also result in increased uncontrolled particulate emissions in the form of
sodium sulfate.
Firing Rate—The firing rate of a kraft recovery boiler is measured in
pounds of black liquor solids per unit of time (either pounds BLS/24 hours or
pounds BLS/hour). Given a specific heat value of the black liquor solids,
percent solids in the liquor, and elemental composition of the liquor, the
flue gas volume produced and boiler heat input can be defined. The firing
rate of a recovery furnace is often increased to increase pulp production
rates. Increasing the firing rates, usually requires an increase in the
gallons per minute of the fired liquor. The firing rate is limited by the
pumping capacity of the system based on the liquor temperature and viscosity.
The-boiler combustion chamber is designed (i.e., sized) according to the
expected heat release rate and volume of flue gas at maximum firing condi-
tions.54 An increase in the flue gas volume above design conditions causes an
increase in vertical gas velocity through the furnace combustion zone and an
increase in particulate emissions resulting from the entrainment of black
liquor droplets and char particles. Thus, particulate emissions tend to in-
crease with increased velocity. The loss of sodium-based particulate may
increase as the flue gas volume is increased because of the higher gas
velocities and temperatures on the hearth of the furnace. The release of
sodium in the flue gas from the char bed increases as flue gas volume increases
because of favorable diffusion conditions.55 The release of sodium is also
related to the temperature of the bed and is primarily the result of evapora-
tion. The rate of evaporation depends on the diffusion conditions (gas
CfT C~J '
velocity) in the zone between the flue gas and the char bed. V ,
An increase in flue gas volume also increases the velocity passing
through the steam tubes and decreases the efficiency of heat transfer. This
decrease in thermal efficiency reduces the overall boiler efficiency and in-
creases the stack temperature. Because of the increased gas volumes and their
effect on char bed temperature, TRS emissions also tend to increase with
boiler firing rate. Figure 3-48 shows the effect of firing rate on TRS and
steam generation rates for a typical boiler.
135
-------
30
35 40
DRY SOLIDS FIRED, 1()3 Ib/hr
45
2.0
Figure 3-48. Effect of solids firing rate on reduced sulfur
emissions and steam generation efficiency.l '
Based on computer models of recovery boiler operations, the particulate
emissions (gains per dry standard cubic foot) increase sharply with increases
in the heat value and solids content of the black liquor (Figures 3-49 and
3-50). These changes are primarily the result: of changes in the heat produc-
tion rate of the char bed.54
Because the heat value and solids content of black liquor are dependent
on a number of process variables in the pulping process (e.g., digesters,
evaporators, wood species, and harvest conditions), day-to-day firing condi-
tions and liquor properties may vary significantly.
Increased firing rates in the recovery furnace can increase flue gas
flow rates. The amount of flue gas produced by the combustion of a specific
black liquor can be calculated by developing a factor similar to an F-factor,
which is defined by the elemental composition of the liquor. The F-factor is
the measured flue gas volume (dry standard cubic feet per minute corrected to
0% flue gas oxygen) divided by the boiler black liquor solids firing rate
(pounds Bis/minute). The value of the F-factor varies from mill to mill
136
-------
600
.-, 500
M
-H
i
£ 400
I
I 300
01
a* 200
100
J_
S.O
4.5
4.0
3.S
58 62 66 70
Black Liquor Solids Conwntration, ft
3.0
Figure 3-49. Effect of black liquor solids concentration.
54
2000
- 2
5200 5600 6000 6400 6800
HMting Value of Solids to Rirnac*, Btu/lb
Figure 3-50. Effect of black liquor heating value.
54
137
-------
because of the variation in species, pulp yield, makeup chemicals, and evaporator
CO
operation. This value, however, is reasonably constant for a specific mill.
Knowledge of the F-factor allows periodic calculation of flue gas
volumes to ensure that the boiler is not exceeding design values. For a more
accurate calculation of velocities in the combustion chamber and the total
flue gas volumes, the F-factor must be corrected for water evaporated from
the fired liquor, water of combustion, temperature, excess air, and miscel-
laneous additions such as steam from soot blowing.
Estimates of uncontrolled emission rates as a function of furnace firing
rate are boiler-specific. Generally, the uncontrolled emission rate of a
typical indirect-contact, recovery boiler is 8.0 gr/dscf. A baseline uncon-
trolled emission rate for use in making future comparisons generally can be
obtained from the vendor's acceptance performance stack test on the ESP. In
summary, increased firing rates increase both the flue gas volume and the un-
controlled particulate emission levels.
Char Bed Temperature—As the firing rates of a recovery boiler are in-
creased, the temperature and diffusion conditions in the char bed tend to in-
crease. This in turn leads to an increase in uncontrolled particulate emis-
sions from the boiler. A qualitative discussion of the mechanisms by which
particulate emissions are increased is presented in the remainder of this sub- ;
section. Dehydrated liquor (char) on the furnace hearth is burned at a high
temperature to allow the inert portion of the liquor to be melted and drained
from the hearth. A mixture of sodium sulfide and sodium carbonate must be
maintained under reducing conditions to prevent oxidation to sodium oxides.
Under normal operation, elemental sodium is vaporized and reacts to form
Na?0. The rate of evaporation depends on the char bed temperature and the
diffusion conditions in the smelt zone. As the sodium evaporates from the
bed, it reacts with oxygen in the primary air zone to form Na20. The Na20
reacts with C02 to form Na2C03.
55
The char bed temperature also determines the rate of sulfur released to
the flue gas. Sulfur is commonly present in the flue gases as S, H2S, S02, or
S03. The higher temperatures favor the formation of S02 and S03- Sulfur in
the form of S and H2S reacts with excess oxygen in the oxidizing zones of the
furnace to form S02 and SOg. The Na2C03 reacts with the S02 to form Na2S03,
138
-------
which is later oxidized to Na2S04. Typically, the Na2$04 deposits on the heat
exchanger surfaces (screen tubes, superheater, and boiler tubes) and must be
removed through continuous sootblowing. These deposits reduce the heat transfer
and the overall efficiency of the boiler. This increases the gas temperature
entering the ESP, as well as the superficial velocity. As a result, there is
a tendency to overfire the boiler to achieve the required steam flow.
Computer models of smelt bed temperatures indicate that the smelt bed
temperature is a function of total combustion air and primary air. Figures
3-51 and 3-52 show smelt bed temperature as a function of total and primary
54
air, respectively.
**
I
so
45
40
35
30
2000
1900
1800
100 110 120 130 440
Total Air to Uiit, parent of theoretical
1700
150
Figure 3-51. Bed temperature as a function of total air.
54
Priauy Air, fnotat of total
1600
60
Figure 3-52. Bed temperature as a function of primary air.
54
Boiler Excess Air—The amount of combustion air supplied to the furnace
influences the chemical-thermodynamic equilibrium at the smelt bed. This
determines the amount of sodium that is volatilized and ultimately lost from
139
-------
the smelt bed. Figure 3-53 shows the theoretical loss of particulate from
the furnace as the total air to the furnace is increased (primary air is fixed
at 45%).54
12
S 10
1 S
I
I
1500
1000
8
I
a"
500
\H»2304 in Fun
100
80 *fc
60 *^.
a
40
Figure 3-53. 'Effect of total air supplied to the unit.
6 ;TJ
:j
tst
3 "*
2 f
1 |
0 *
54
The amount of boiler excess air needed for complete combustion is
normally between 110 and 125 percent of theoretical air (stoichiometric air).
A minimum level of excess oxygen (1 to 2%) must be maintained to ensure com-
plete combustion and reduce the formation of
12
When the amount of excess
air is above 125 percent (5% 02 in the flue gas), the formation of S03 in-
creases.59 The formation of S03 and H2S04 can be increased considerably if a
residual fuel oil with a high content of vanadium pentoxide (V205) is fired in
combination with the black liquor. The SO, is absorbed in the particulate
12
at low temperatures, which makes the particulate sticky. This sticky par-
ticulate fouls heating surfaces in the economizer and reduces heat transfer
rates. The deposits may result in a high draft across the economizer. The
particulate also causes severe operating problems when collected on the plates
of the ESP. The sticky salt cake (particulate) cannot be removed effectively
by boiler soot blowers or removed from the ESP plates by normal rapping in-
tensity.60
Higher than normal excess air increases the volume of flue gas, which re-
sults in increased furnace vertical velocities and reduced heat transfer ef-
ficiencies. If the high excess air is caused by an increase in primary air
in the char bed zone, sodium evaporation will increase along with the particu-
late emission rate. When the high excess air is caused by increased secondary
air, the formation of S03 in the flue gas is more likely.
140
-------
In summary, the firing of the boiler at high excess air has three effects:
1) increased particulate emission rates on an uncontrolled basis, 2) in-
creased flue gas volume to the ESP, and 3) formation of S03 that reduces ESP
power as a result of salt cake buildup on both the plates and the wires.
Primary Air--Primary air is required to provide complete combustion and
to maintain the temperature in the char bed to prevent a condition called
"blackout." The amount of air is a compromise between maintaining sufficient
combustion and reducing abnormally high vertical velocities in the furnace.
The total air volume (secondary plus primary) must be high enough to produce
complete combustion, but be limited so as to reduce the vertical velocity in
12
the furnace and total flue gas volume. A higher velocity increases the
release of sodium and sulfur from the char bed because of the increase
in diffusion of the vapor from the bed. The higher volume also increases the
combustion rate of char, which increases bed temperature. Figure 3-54 shows
the theoretical loss of particulates as a function of percentage increases in
primary air. This produces an accumulation of deposits on the heating surface
of the boiler (after cooling of the gas stream and condensation of fume). This
accumulation causes an increase in the particulate emissions.
"
12
10
*
i «
I 2
• 1500
i
3 1000
.6
I 500
8
3
100
80
60
40
20
0
Figure 3-54. Theoretical loss of particulate as a function of percentage
increases in primary air.
5
-------
Secondary Air—The total amount of primary and secondary air required
for combustion is between 110 and 125 percent of theoretical air (stoichiometric
air). The normal limits are between 2 and 5 percent excess 02-
The secondary air should be a minimum of 40 percent of the total air
(maximum of 65 percent of the theoretical air).12 In boilers with a high char
bed, the secondary air has two purposes. The primary purpose is to complete
combustion of CO gas released from the char bed as it moves up the furnace
walls.61 The secondary purpose is to provide primary air in the center of
the furnace to burn the char bed.
Primary Air Temperature—The primary air temperature has a direct in-
fluence on the smelt bed temperature. An increase in the primary air temperature
has the same effect as an increase in the primary air volume. Figure 3-55
shows the effect of primary air temperature.54 In modern designs, the primary
air is heated with indirect steam to approximately 400°F to increase the bed
temperature. This reduces the TRS emissions from the bed and increases the
emission of Na2S04.50 Changes in primary air- port design are also used to
increase velocities, which help to stabilize bed temperature through better
air penetration.
600
soo
3
J <°°
I
ft 300
I
a **>
rf*
100
I
1 r
100 200 300 400
Priauy Air T«pwatux«, *F
500
Figure 3-55. Effect of primary air temperature.
54
142
-------
Sulfur-to-Sodlum Ratio—As the ratio of sulfur to sodium in black liquor
increases, the amount of sodium in the smelt that is available to combine
with sulfur gases is reduced. This results in an increase in S02 and TRS
-emissions and a decrease in sulfidity across the furnace. As a result, the
smelt has a lower sulfidity (Figure 3-56). At sulfidity levels in excess of
en cfl
30 percent it is difficult to comply with applicable TRS emission standards. '
1
8000
6000
I
jf 4000
M
n
I
sf 2o°°
8 3
6 S.
§
r
4 y
100
if
53
I
80 -
60
40
20
— — Reported,by Fukui
0.2 0.4 0.6 0.8 1.0 1.2
Molar Ratio S/Jtaj in Black Liquor
1.4
Figure 3-56. Effect of sulfur-sodium ratio in the black liquor.
54
143
-------
Chlorine in Black Liquor—Chlorine can enter the pulp process through
wood-born contaminants, chemicals, or mill wastes. The most common sources
are saltwater log storage and bleach plant effluent. The presence of small
amounts of chlorine increases the loss of particulates from the smelt bed.
The chlorine combined with sodium (NaCl) is highly volatile. Figure 3-57 shows
the theoretical loss of sodium chlorides as a function of percent chlorine in
54
the fired liquor. The loss of HC1 from the furnace can increase corrosion
in the ESP and ductwork if significant cooling occurs as a result of inleakage
or improper insulation.
~ 800
.3
a
600
I
§ 400
•H
K
200
s
40
30 -
20 -
10 -
01234
Oilorin* in Black Liquor, percent (w»t buis)
Figure 3-57. Effect of chlorine in black liquor.
54
144
-------
3.3.1.3 Control —
The primary method of controlling participate emissions from recovery
furnaces is electrostatic precipitation. At one time venturi scrubbers were
used for primary and secondary particulate control on some designs; however,
since their use has been discontinued, the following discussion is limited to
ESP's.
Equipment Design—The three basic processes in electrostatic precipita-
tion are 1) the transfer of an electric charge to suspended particles in the
flue gas, 2) the establishment of an electric field for removing the particles
to a collecting electrode, and 3) removal of the particles from the ESP with
as little loss to the atmosphere as possible. Figure 3-58 illustrates the
fi?
basic processes involved in electrostatic precipitation.
Particulate matter is collected in an ESP by means of an electrical
charge placed on the particles and a grounding surface of opposite charge.
The particles move in an electrostatic field created by electrodes operating
at a negative potential ranging from 30,000 to 80,000 volts. The electrodes
consist of wires placed perpendicular to the gas flow between collecting plates.
As the potential on the electrodes is increased, electrons are released that
charge the particles passing between the wires and plates.
Depending on the size of the wire and its roughness, shape, and dust
coating, changes occur in the voltage at which the electrons begin to flow
(as measured by secondary current). The cloud of electrons surrounding the
electrode is called a corona, and the voltage at which current occurs is de-
fined as the corona initiation voltage.
The charged particles move to the collecting plates and bleed their charge
through the dust layer. The rate of release of this charge is indicated by
the secondary current. When the dust concentration in the gas stream is low,
the relationship between secondary voltage and secondary current is an expres-
sion of the electrical resistance of the system.
The secondary voltage and resulting secondary current may be increased
until the potential is high enough to allow a spark to occur directly through
the gas stream to the plate. This results in power loss without the charging
of particles. In older ESP designs, the optimum utilization of power (highest
efficiency) occurred at moderate spark rates (>50/min). With the use of new
145
-------
FREE
ELECTRONS
REGION OF\
CORONA GLOW\
e
jELECTRONS
CORONA GENERATION
ELECTRON
DUST
PARTICLE
GAS
MOLECULE
CHARGING
-© WIRES
TURBULENT
GAS FLOW
>>\ RAPPING
N SYSTEM
COLLECTION
COLLECTING
PLATE
;HOPPER
ASH REMOVAL
SYSTEM
REMOVAL
Figure 3-58. Basic processes involved in electrostatic precipitation."
62--
-------
digital control systems that sense the occurrence of sparking, optimum power
levels can be maintained without sparking or at a moderate spark rate (<50/min).
fi?
Figure 3-59 shows the general arrangement of an ESP. The space between
the plates is called the gas passage or lane. Although the spacing between
plates varies with manufacturer, it is typically between 9 and 10 in. The
electrodes (wires) are placed at equal distances between the plates (at 8-
in. intervals in the direction of gas flow). The wires may be individually
supported and tensioned by use of weights or rigidly held in a frame. The
electrode cross section may be round, square, twisted, 'barbed wire, or strained
barbed wire. The shape and diameter determine the corona initiation voltage.
In general, the resistance of the particle to conduct its charge to the
collecting surface is defined as the particle resistivity. The high moisture
content of the gas stream from recovery furnaces results in a highly con-
ductive particle (low resistivity). This low resistivity allows the particle
to release the charge quickly at the plate and to be effectively removed from
the plate with moderate plate rapping.
Power is supplied to the ESP by rectified high-voltage transformers.
The rectifiers may be full-wave or double haIf-wave design, depending on the
63
sectionalization (chambers) of the ESP (Figure 3-60).
The ESP is divided into separate collection surfaces arranged in series
to-allow increases in power levels as dust is removed from the gas stream and
to allow individual field rapping. The total power input to each successive
field increases until the transformer-rectifer (T-R) set limit is reached.
Power usage by each section (field) is indicated by the product of primary
voltage and current of the transformer and by the product of secondary
voltage and current of the rectified power supply. The primary power level
is not considered a true measurement of ESP power because of inefficiencies
and losses in the transformer and rectifier. Although the secondary power
readings indicate the power delivered to the ESP, in some cases they do not
measure the power delivered to the particle charging mechanism because of
insulator tracking, distribution losses, etc.
Because of the nature of the charging process, the collection efficiency
is affected by the length of time the particles are in the electrostatic
field (time of treatment). Because the physical dimensions of the ESP are
fixed, the time of treatment is determined by the total gas flow rate, box
147
-------
GROUND SWITCH BOX-
• ON TRANSFORMER
TRANSFORMER-
RECTIFIER
HEAT JACKET
PERFORATED
DISTRIBUTION
PLATES
DISCHARGE
ELECTRODE
DISCHARGE
ELECTRODE
VIBRATOR
COLLECTING
ELECTRODE
RAPPERS
TOP HOUSING
ACCESS DOOR
TOP HOUSING
HOT ROOF
CCESS DOOR
BETWEEN
COLLECTING
PLATE SECTIONS
COLLECTING ELECTRODES
WET BOTTOM
Figure 3-59. Typical wet-bottom ESP with heat jacket.62
148
-------
R.-L
Low
Voltage
Input
High Voltage
Transformer
—s*.
Bridge
Rtetifitr (Static)
One or More
But Section*
FULL-WAVE CIRCUIT SCHEMATIC
Usually
Linear Reactor
Two Separate Bus Sections
or Fields
Low
Voltage
Input
DOUBLE HALF-WAVE CIRCUIT SCHEMATIC
Figure 3-60. Electrical diagram for ESP T-R set.
63
149
-------
65
length (number of fields), and cross-sectional area; i.e., the time of treat-
ment is defined by the superficial velocity between the plates and the box
length. The deviation from the average velocity is limited to +10 percent of
the mean value.
The design amount of power to be delivered to the ESP in each field is
sometimes specified in terms of the length of the wire electrodes. The inlet
field is typically sized for 0.2 milliamp (mA)/ft of wire, and the outlet
field may be sized for up to 0.10 mA/ft of wire. Accordingly, the inlet field
T-R sets are sized to be consistent with anticipated power consumption. The
plate area used per 1000 acfm of gas being treated is one convenient measure
of the potential collection capability of the ESP. This term is defined as
the specific collection area in square feet/1000 acfm.
There are two methods of supporting the discharge electrode in recovery
boiler ESP's. In the first method, referred to as the weighted-wire design,
each electrode is individually supported and tensioned between the plates
(Figure 3-61). In the second method, referred to as a rigid-frame design,
the electrode is attached to a rigid frame between the plates (Figure 3-62).
A modification of the second design is a rigid-pipe electrode system in which
the corona is generated on the tip of spikes attached to a vertical pipe.
Collected particulate can be removed from the ESP in three ways. In the
first method, referred to as a wet-bottom ESP, the salt cake is allowed to
fall into an agitated pool of black liquor in the bottom of the ESP (Figure
65
3-63). In the second method, referred to as a dry- or drag-bottom ESP, the
salt cake is allowed to fall onto the flat bottom of the ESP shell, where a
drag chain physically moves the material to a discharge screw (Figure 3-64).
The third method of dust removal consists of a pyramid-shaped hopper with
rotary air locks and slide-gate discharges. This design is not often used in
recovery boiler ESP designs because the hopper tends to plug.
To remove the particulate from the discharge electrodes and collection
plates, the ESP collecting surfaces are vibrated or rapped. Rapping equip-
ment falls into three main categories: electrical, physical, and pneumatic.
The electrical rappers consist of MIGI (magnetic impulse gravity impact)
rappers and vibrators. The physical rappers are rotating falling hammers,
either internal or external to the ESP. The pneumatic rappers are air
65
150
-------
Figure 3-61. Typical weighted-wire ESP with drag bottom.
65
151
-------
Figure 3-62. Rigid-frame design.65
152
-------
Figure 3-63. Wet-bottom ESP.
153
65
-------
IQ
ro
00
en
o
-s
Id
O
Q>
_l.
3
fD
cr
-------
cylinder hammers. Figure 3-65 shows a MIGI rapper, and Figure 3-66 shows a
typical internal falling hammer rapper design.
Buildup on the plates resultsS1n low collection efficiency due to re-
duced power input to the ESP. The effect is most prevalent when the flue gas
temperatures are below 300°F and a combination of high boiler excess air and
ambient air inleakage is occurring.
In general, ESP control efficiency is determined by the initial design
of the unit and the operating characteristics of the source. The major design
factors include plate area, superficial velocity, and sectionalization of the
unit and size of the T-R set. Figure 3-67 shows the relationship between
O
design efficiency and the design specific collection area (ft /1000 acfm) for
modern ESP's.66 It should be noted that increasing the plate area above 400
ft2/1000 has a diminishing effect with regard to improving performance, pri-
marily because of the predominance of factors not related to corona power,
such as hopper sneakage and rapping losses. To minimize the effects of rap-
ping loses and to improve residence time, manufacturers have been designing
the newer ESP's with lower superficial velocities. The accepted maximum
velocity in current technology is about 3.5 ft/s. Figure 3-68 shows the
design velocity of 20 randomly selected units.
It should be noted with increased treated gas volume, an increase in gas
temperature can increase the superficial velocity beyond design values and
reduce the effective SCA and collection efficiency.
Instrumentation—Optimum performance of an ESP depends on effective con-
trol of the operating parameters that can vary because of changes in the
physical characteristics of the ESP or the flue gas stream being treated.
Recent ESP installations generally are equipped with instrumentation for moni-
toring and recording the major operating parameters. The operator and in-
spector should thoroughly understand the function of each instrument and
associated records to evaluate ESP performance.
Instrumentation for a kraft pulp recovery boiler ESP generally consists
of monitors for ESP power input, flue gas temperature, oxygen content, and
opacity, rapper operation, and the particulate discharge system. The in-
struments are generally located in proximity to the ESP unit. When a plant
has more than one ESP, the instrumentation for all the ESP units can be housed
in a centrally located control.
155
-------
Figure 3-65. MIGI rapper cross section.
65
156
-------
Figure 3-66. Internal falling-hammer rapper design.
157
-------
100
99
o
98
97
96
DEUTSCH-
ANDERSON
MATTS-OHNFELDT
/
100
0
200 300 400 500 600
DESIGN SCA, ft2/1000 acfm
Figure 3-67. Design SCA and efficiency of 20 recovery boiler ESP's.66
158
-------
6SI
SUPERFICIAL VELOCITY, ft/s
vo
tn
oo
—r
10
-5
fD
CO
CT>
CO
GO
£=
•O
n>
I?.
o
O)'
fD
o
o
*1,
4?
fD
5
PO
CO
vo
CJ1
O
en
C7I
en
o
tn
£=
CO
fD
O)
-S
fit
vo
o
vo
tn
• •
fD
Q.
CT>
CT>
vo
oo
o
-------
In the ESP instrumentation block diagram shown in Figure 3-69,62 the
ESP has four T-R sets, each of which has two bus sections. Primary voltage
and primary current are measured for each T-R set. Secondary voltage, secon-
dary current, and spark rate are measured for each bus section. Figure 3-70
shows the positions of various instruments in an ESP circuit. The ESP in this
figure has four primary voltmeters and four primary ammeters; the secondary-
side instrumentation consists of eight secondary voltmeters, eight secondary
ammeters, and eight spark rate meters.
The function of each instrument determines its relative location in the
circuit. The primary ammeter, for example, is always located ahead of the
transformer to indicate the current available for transformation. Temperature
is measured at the ESP inlet, and opacity is measured at the outlet.
Primary-side meter readings only indicate the general condition of the
ESP and cannot be used for complete diagnosis of internal conditions while
the ESP is in operation. Nevertheless, an experienced operator generally can
use the primary readings to indicate wire breakage, severe plate buildup, and
discharge problems.
Energization power for an ESP is supplied at the primary side of the T-R
set. The alternating current electrical power is normally 220 or 460 volts.
The voltmeter dial for primary voltage shows a range of 0 to 480 volts.
Normal operation is usually between 220 to 460 volts. A temporary deviation
of +5 percent from the rated supply voltage is fairly common. The primary-
side voltmeter is located ahead of the T-R set, but after the power control
circuit, linear reactor, and feedback network. This positioning ensures
measurement of the regulated voltage available at the T-R set.
The ammeter on the primary side indicates the current drawn by the T-R
set. The primary ammeter is located between the T-R set and the power con-
trol circuit, linear reactor, and feedback network.
Secondary-side instrumentation indicates the power input to an individual
ESP section. The secondary-side instrumentation usually consists of a volt-
meter, an ammeter, and a spark rate meter for each bus section. Secondary-side
meters are used to produce current/voltage relationships that characterize the
internal conditions of the ESP. This allows maintenance personnel to concen-
trate on the most severe problem areas.
160
-------
CTl
g3E=3E=)E^
SI SR PV PI SV SI SR
SV SI SR PV PI SV SI SR
ACITY
INDICATOR
IINLET CAS TEMPERATURE INDICATOR
SV: SECONDARY VOLTAGE
SI: SECONDARY CURRENT
SR: SPARK RATE
PV: PRIMARY VOLTAGE
P!: PRIMARY CURRENT
Figure 3-69. ESP instrumentation diagram.
62
-------
4M V
i w
CO HI
-o o
CONTROL
TRANS.
cr>
ro
AUTOMATIC CONTROL HOOULC
H.V.
I VOLTAGE ) TRASSF.
\NETER
ESP
Figure 3-70. Positions of measuring instruments.
68
-------
Secondary voltmeters are calibrated in kilovolts to measure the high
voltage of the power input to the discharge electrode. The secondary volt-
meter is located between the rectifier output side and discharge electrodes
to indicate the direct current voltage across the discharge electrodes.
The secondary-side ammeter indicates the current that is being supplied
to the discharge electrodes in one section of the ESP. The secondary current,
which is produced by stepping down the primary current in the transformer,
is measured in mi Hi amps.
The spark rate meter is a major indicator of ESP performance. The spark
rate indicates the rate of sparks in a single ESP section.
Certain instruments external to the ESP are also important in the over-
all operation of the ESP. These include opacity monitors, oxygen monitors,
inlet gas temperature indicators, and discharge system monitors.
Opacity is limited under the NSPS regulations for kraft pulp mills, and
new kraft pulp mills are required to install and operate continuous opacity
monitors. A correlation can be developed between opacity and mass emissions
during the performance test that allows the operator to use opacity as an
indicator of overall performance with respect the mass emission rate.
The level of oxygen in the flue gas is an indication of excess gas
volume due to boiler excess air and air infiltration into the ductwork and
the ESP. Because of the relationship between ESP performance and total gas
volume, high oxygen content in the flue gas is an indication of reduced
performance. The optimum levels-of excess oxygen should be identified during
the performance test and used as a baseline for determining when maintenance
may be needed.
Temperature is usually measured at the exit of the recovery boiler.
The temperature of the flue gas affects corona power and ESP efficiency; i.e.,
low temperatures can cause acid and moisture condensation, which lead to
corrosion and eventual structural failure.
Instrumentation associated with the discharge system varies with the
type of discharge system being used, i.e., wet-bottom or drag-chain. The
instrumentation for wet-bottom ESP's usually consists of a float to indicate
the level of the black liquor in the wet bottom and a flow meter to indicate
volumetric flow. Some mills also have .ammeters to indicate whether the pumps
163
-------
that transport the black liquor are working properly and tachometers to
indicate the revolutions per minute of the agitators.
For drag chains, the only two parameters that are usually measured are •
the revolutions per minute of the drag chain and the current to the motor
that operates the drag chain. These instruments are basically used to tell
whether the drag chain is operating properly. Too much tension on the drag
chain can cause the chain to break, and too little tension can cause particu-
late to build up in the bottom of the ESP. If the particulate continues to
build up, it can cause the chain to break. In addition, significant buildup
that goes unnoticed for a period of time can eventually ground out the system.
The basic instrumentation associated with rapping is a meter to indicate
the rapper has "fired." Information on the internal operation of the rapper
and rapper intensity is extremely difficult to obtain without shutting down
the ESP.
3.3.1.4 Boiler Malfunctions-
Malfunctions that increase emissions can be divided into two areas:
1) those that occur as a result of furnace operation, and 2) those that occur
as a part of control equipment operation. Several malfunctions that result
from improper boiler operating practices also may have an impact on ESP
performance.
The following is a summary of the major boiler operating or design
parameters and associated malfunctions that may Increase emissions.
Firing Rate--The firing rate of the boiler in pounds of black liquor
solids per hour defines both the heat input to the boiler and standard cubic •
feet of flue gas volume generated. The flue gas volume and heat input change
with the chemistry and the heat content of the liquor. At defined chemistry
and heat values, predictable volumes of a flue gas may be calculated. The
F-factor for black liquor varies from mill to mill because of species, pulp
yield, and makeup chemicals, but is reasonably constant for specific mills.
An increase in firing rates above design (flue gas volume) results in increased
vertical gas velocity through the combustion zone of the furnace and an in-
crease in particulate emissions due to the entrainment of black liquor droplet's
12
and char particles and increased char bed temperature.
1,53
164
-------
Participate emissions also increase at high firing rates because of the
increased volume of primary air. The release of sodium in the flue gas from
the char bed increases as the flue gas volume increases as a result of favor-
54
able diffusion conditions. The release of sodium is related to the tempera-
CO
ture of the bed and is primarily the result of evaporation. The rate of
evaporation depends on the diffusion conditions (gas velocity) in the zone
between the flue gas and the char bed.
The firing rate has a dual effect. It influences the flue gas volume
treated by the ESP and the uncontrolled particulate emission rates.
Char Bed Temperature — The char bed temperature appears to have an effect
on the evaporation rate of sodium to the flue gas. The amount of sodium that
is distributed to the flue gas increases sharply with temperature, following
12
a curve similar to a vapor pressure curve.
The sodium evaporates from the bed as elemental sodium reacts with oxygen
in the primary air zone to form Na^O. The Na^O reacts with C02 to form Na-pCO
The amount of sodium released from the bed does not appear to be dependent on
the sodium content of black liquor solids.
The temperature in the bed also determines the release of sulfur to the
flue gas. The sulfur is commonly present in flue gases as S, H2S, or S02.
The lower temperatures favor the formation of S and H2S. Typical char bed
temperatures are 1700° to 2200°F.54
The sulfur that is released reacts with excess oxygen in the oxidizing
zones in the furnace to form S0. The
reacts with the S02 to form
2. ^
Na2S03 that is later oxidized to Na2S04. '
It should be noted that in some designs the combustion of the char may
be accomplished on the boiler, wall. In these designs no char bed (profile)
is maintained, and the smelt drains from the walls onto a sloped hearth for
removal.
Primary Air — The primary air is required to provide complete combustion
and to maintain the temperature in the char bed to prevent a condition called
"blackout."
The amount of air is a compromise between maintaining sufficient combus-
tion and reducing abnormally high vertical velocities in the furnace. The
increased velocity results in an accumulation of deposits on the heating
165
-------
surface of the boiler (after cooling of the gas stream and condensation of
fume). This accumulation causes an increase in the particulate emissions.
The increased rate of primary air (particularly at high velocities) in-
creases the release of sodium and sulfur from the char bed because of a
rq
favorable increase in diffusion of the vapor from the bed. The increased
volume also increases the combustion rate of char, which increases bed
temperature. Particulate emissions increase sharply when the amount of
primary air exceeds 45 percent of the total air volume.
Prolonged operation at low primary air volumes can result in increased
char bed height, which must be reduced by an increase in bed temperature.
The most common method of reducing the bed height is to increase the primary
and secondary air volumes. This can cause the smelt ratio to be altered as
oxidation and temperature conditions are changed.
The increased combustion can increase the flue gas volume from the
boiler to a point where it is greater than the amount calculated from the
instantaneous firing rate of the boiler. For the volumes to be equal to in-
stantaneous firing conditions, the char bed combustion rate must be at equi-
librium with the amount of char deposited on the bed. When the bed is being
reduced in height, the flue gas and particulate emission rates will be greater
than those predicted by the black liquor firing rate.
Secondary Air—The total amount of primary and secondary air required
for combustion is 110 percent of theoretical air (stoichiometric air). The
normal limits are between 2 and 5 percent excess Oy-
The secondary air should be a minimum of 40 percent of the total air
1 9
(maximum of 65 percent of the theoretical air). In boilers with a high char
bed, the secondary air has two purposes.- The primary purpose is to complete
combustion of CO gas released from the char bed as it moves up the furnace
walls. The secondary purpose is to provide primary air in the center of the
furnace to burn the char bed.
The total air volume (secondary plus primary) must be high enough to
produce complete combustion, but it must also be limited to reduce the verti-
cal velocity in the furnace and total flue gas volume.
Boiler Excess Aii—The amount of boiler excess air needed for complete
combustion is typically between 110 and 125 percent of theoretical air
166
-------
(stoichiometric air). When the amount of excess air is above 125 percent (5%
Og in flue gas), formation of S03 increases. ' The S03 is absorbed in the
particulate at low temperatures, which makes it sticky. This sticky particu-
late fouls heating surfaces in the economizer and reduces heat transfer
rates. The deposits may result in a high draft across the economizer. The
particulate also causes severe operating problems when collected on the
plates of the ESP. , The sticky salt cake cannot be removed effectively by
boiler soot blowers or removed from the ESP plates by normal rapping inten-
sity.
Buildup on the plates causes low collection efficiency as a result of
i
reduced power input to the unit. The effect is most prevalent when flue
gas temperatures are below 300°F and a combination of high boiler excess air
and ambient air inleakage is occurring.
This condition can be identified in the ESP when high secondary voltage
(> 50 kV) and low secondary current (< 100 mA) are observed in the inlet fields.
Figure 3-71 shows a typical secondary current pattern for a unit experiencing
salt cake buildup as a result of high excess air (SO, formation).
The formation of SO, and H2S04 can increase considerably when a residual
fuel oil with a high content of vanadium pentoxide (V90r) is fired in combi-
70
nation with the black liquor.
The firing of the boiler at high excess air has three effects: 1) in-
creased particulate emission rates; 2) increased flue gas volume to the ESP;
and 3) formation of SO., which reduces ESP power because of salt cake buildup
on both the plates and the wires.
3.3.1.5 ESP Malfunctions--
The following is a discussion of the major operating and design param-
eters and associated malfunctions that have an impact on ESP performance.
Operation and maintenance (O&M) of ESP's is a broad subject involving
all aspects of ESP performance. It cpvers all component parts and all oper-
ating conditions. In general, maintenance is considered to be the routine
analysis and replacement of components parts that have failed because of age
or abuse. Maintenance requirements may be increased as a result of poor
operating practices or reduced through superior system design. Detailed and
exhaustive maintenance practices do not, however, necessarily yield superior
or exceptional ESP performance.
167
-------
600
500
400
LU
o:
O
or
«x
o
o
CJ
LU
C/7
300
200
100
SOUTH
2
FIELDS
Figure 3-71. Typical secondary current pattern for unit
experiencing salt cake buildup.71
168
-------
Poor design or operating practices that cause a high level of maintenance
activity are not necessarily corrected by that maintenance activity. Because
the symptom is often treated, rather than the real problem, this discussion
centers around good operating practices and major design problems instead of
routine maintenance activities. It is imperative, however, that a good
inspection and diagnostic program be used to identify problem areas and to
direct corrective action. Such action may include changes in operating
practices, the basic system of design, and maintenance schedules.
In general, the ESP recovery boiler system must be operated within the
scope of its design variables. Deviation from the initial design for inlet
temperature, inlet grain loadings, gas volume, superficial velocity, corona
power, and velocity distribution can adversely affect expected performance
levels. The following discussion deals with the importance of several oper-
ating parameters and the major causes of poor performance.
Gas Volume—The gas volume treated by an ESP is defined by the boiler
size (heat input), excess air, temperature of the gas stream, inleakage
through duct flanges and across cascades, and black liquor composition.
Several of these variables, e.g., black liquor composition, are fixed over a
very narrow range. The other parameters have a wider range of variability
and depend on boiler operation, boiler age, or lack of maintenance.
Changes in those variables that increase the gas volume treated by the
ESP decrease ESP removal efficiency. Because the cross-sectional area of the
ESP is fixed, any increase in gas volume causes a corresponding increase in
superficial velocity and a decrease in SCA. Performance of the ESP has been
improved in recent years by increases in SCA through the use of more fields.
The SCA has also been increased by increasing the cross-sectional area of the
box to reduce superficial velocity. Currently, ESP1s are being designed with
SCA's in excess of 400 ft2/1000 acfm and superficial velocities near 3.5
ft/s. Increasing the operating flue gas volume above the design effectively
reduces the size of the ESP and reduces its performance.
The amount of flue gas generated by a pound of black liquor solids is a
function of the elemental composition of the solids. The organic components
are carbon, hydrogen, oxygen, and sulfur. The nonorganic or ash portion of
the solids is also important because it combines with the oxygen, sulfur
169
-------
dioxide, and C02 to form sulfates and carbonates. The major elements in the
ash portion are sodium and inert oxides.
Based on a combustion analysis, an F-factor for black liquor can be de-
veloped. The F-factor (F.) for a typical sulfate pulping black liquor is
7? '
approximately 51.07 dscf/lb BLS. This value assumes stoichiometric con-
ditions and does not include inleakage or soot blowing. The amount of water
vapor created by the combustion of hydrogen in the liquor can be calculated
by use of the same method, and is defined as F_. The F for a typical sulfate
v5p
black liquor is approximately 7.99 scf/lb BLS.
If the oxygen content and temperature are known, the flue gas volume can
be calculated from the boiler firing rate with the appropriate corrections
for the moisture in the flue gas. Moisture in the flue gas results from
moisture in the fired liquor, moisture from the direct-contact evaporator,
and moisture added by soot blowing.
If a more exact or plant-specific value is desired, the F-factor can be;
calculated from the stack test gas volume, the flue gas oxygen content, and
the firing rate of black liquor solids.
Using this method, the operator, inspector, or environmental personnel
can determine, on a day-to-day basis, the flue gas volume being treated by
the ESP without the expense of using a Pitot tube to determine stack flue gas
volume.
When flue gas oxygen increases above the normal ranges, the source of
inleakage should be identified immediately, and appropriate repairs should
be made to reduce the inleakage. Failure to reduce the inleakage not only
results in excess emissions, but the cooling effect of the ambient air causes
reduced power input, excess sparking, and corrosion.
For inspection purposes, the oxygen content of the flue gas stream can
be obtained by using portable instruments and by taking multiple readings
across the stack. Care must be taken to avoid inleakage stratifications that
may occur along the duct walls; this value would not be representative of the
major portion of the duct cross section.
Corona Power--The power used by the ESP in charging the dust particles
is measured in watts and is the product of secondary voltage and secondary
current. The particulate removal efficiency depends on the rate of particle
170
-------
charging and on the electrical field potential in which the particle is placed.
•It becomes imperative that the maximum corona power be maintained in each
field to achieve the desired performance.
The corona power in a recovery boiler ESP increases from inlet to outlet
as dust is precipitated out of the gas stream. Most units are designed with
an expected power consumption level and an expected secondary current in each
field. The power consumption is expressed in watts and the T-R set is sized
accordingly. Because there is a loss of energy in the transformer and in the
rectification process, the secondary-side power levels do not agree with the
primary-side power levels. The conversion efficiency ranges from 65 to 80
percent depending on the size and design of the T-R set and the ratio of T-R
set operating point (milliamps) to T-R rating (milliamps). The conversion
efficiency is highest when the T-R set rating and operating point are closely
matched.73
The secondary current for design purposes is expressed in milliamps per
foot of electrode in each field. The values generally are in the range of
0.02 mA/ft in the inlet field and increase in successive fields up to about
90 percent of the T-R set rating.
The secondary voltage applied to the electrode is maximized to provide a
strong field strength (kilovolts) and is limited by electrode-to-plate clear-
ance or T-R set rating.
The limiting level of corona power input is determined by the dust
loading, gas temperature, moisture content, and wire/plate clearances. The
amount of power delivered is typically expressed in terms of the physical
size of the ESP (watts per square foot) and is based on the amount of flue
gas being treated (watts/1000 actual cubic feet per minute). The efficiency
of the unit is directly related to the power delivered to the gas stream,
i.e., watts/1000 actual cubic feet per minute.
Because many studies have indicated a strong relationship between corona
power and collection efficiency, it is necessary to monitor and maintain a
high level of corona power. The inlet fields remove 90 percent of the par-
ticulates; therefore, maximum power levels must be maintained in the inlet
fields. The optimum power distribution increases linearly from inlet to
outlet.
171
-------
To prevent corona power degradation due to misalignment, dust cake
buildup, or T-R set failure, the boiler operator should record hourly T-R
meter readings on the boiler log sheet and note any deviations from normal
readings.
Typically, ESP and boiler operating conditions are not recorded during
the stack test period. Without these data, a comparative baseline (watts/1000
actual cubic feet per minute) cannot be established. The comparative base-
line allows an accurate evaluation of the day-to-day operation. Long-term.
degradation of corona power levels, which can occur because of the loss of
rapper effectiveness, increases in flue gas volume, misalignment, or changes
in T-R set controllers, is seldom noticed over a period of months, even
though the overall efficiency may be decreasing. In most cases, immediate
short-term failures such as rapper control loss, T-R set controller failure,
wire breakage, drag chain failure, or insulator failure are indicated by
changes in corona power levels, which can occur over a few minutes or several
hours.
Daily review of corona power levels by supervisory personnel and compari-
son with normal values can permit a rapid and correct diagnosis of maintenance
problems before they result in excess emissions or catastrophic failure of
the unit.
Most recovery boiler ESP's are designed with two parallel chambers, which
allows internal maintenance to be performed on one chamber while the other
chamber carries the gas volume. Maintaining compliance with emission limits,
however, requires that the boiler load be reduced to keep the total gas
volume compatible with the reduction in collection plate area because all the
gas is passing through one chamber.
In those systems that have a single T-R set per field, the T-R set is
installed in a double half-wave design. The T-R set controller is typically
designed to maximize power input based on secondary-side operating param-
eters. Power input to the unit is limited by the side with the lowest spark
point (i.e., closest clearance, heaviest dust cake buildup, or a section with
a cold air or oxygen stratification in a single lane). For determining if
one chamber is limiting total power input, alternate T-R taps should be
grounded and the power, to each chamber evaluated. Major deviations in voltage
172
-------
or current levels at the T-R set limit or spark point indicate clearance
problems, rapper failures, or salt cake buildup on the plates.
Gas Distribution—For optimum performance of an ESP, the velocity of
gas passing through each gas passage or lane must be approximately equal.
According to the Industrial Gas Cleaning Institute (IGCI), the maximum
deviation allowed across the ESP inlet face is +10 percent of the mean veloc-
ity. This distribution of velocity is achieved by using inlet turning vanes
and distribution plates. Figure 3-72 shows a typical perforated gas distri-
bution plate that has moved from the vertical plane and become plugged.
Erosion of the vanes and plates or pluggage of the distribution plate causes
a deviation in the gas distribution. A serious deviation can reduce corona
power because of stratification of the gas stream.
The gas volume to each chamber is balanced and modified by inlet and
outlet dampers. Buildup on these dampers combined with pluggage of the dis-
tribution plate can cause imbalance in chamber gas volume. A Pitot tube
should be used to make periodic checks of the gas volume to each chamber.
Balancing of the gas volume based on damper position or static pressure drop
can be very misleading.
Drag Chains—Particulate collected on the plates and wires is removed by
rappers and falls to the bottom of the ESP. A drag chain assembly is used
for continuous removal of the particulate cake from the bottom of the ESP via
a discharge screw conveyor. Failure to remove the salt cake can result in a
buildup in the electrical field and a short in the T-R set, which can cause
serious permanent misalignment of the wire frame assembly.
Typical malfunctions of the drag chain system involve chain breakage,
misalignment of the drags, sprocket failure, and/or motor failure. Failures
of the chains and sprockets require isolation of the chamber to make the
necessary internal repairs.
Two types of drag chains are used, longitudinal and lateral. The longi-
tudinal drag-chain system moves across electrical fields in the direction of
gas flow. Sneakage of the gas below the treatment zone between fields is
prevented by baffles located between fields at a point above the drag chain.
The baffle does not completely isolate the field, and some sneakage occurs
when the drag chains are in motion. Failure of the baffles as a result of
173
-------
••'••
*1*5*
'•!•*
•.;.•• -<-.»«^^i • -T:. , • "--g"** ..• .. •••^r-,^j,--^^^:'j«
-.:?•.•••• .-'?-•'- •••^•:-^/.'yf^,-sm^
+ --/••''.• -.-•-;•. -r""-s^p
>v':: V:/:^^:^^^&
|: ,.;! •- ••: - ^VmV^fil*S
i.-^;-"1^^ ^K ' • •'•V ^F-p- -• '- - .. ' -.-".• •- . .:.'•-•;*..'-'• b-^^ '^- . •;• •:•....•.-:'. .:.•'•-•-..>-.>;;,. ;.:-.-f..^-.I'^y'-V?^^
*';t^***Xv-^:
(iS*^:^!^^:
> - > • ' r~T !P V£*
••••' ' '•"•-:'"'.-'''iS®
:;fe'^S^^
y.'' ;,, ''•T/^>VV'J^V-'''^^^0**W
i»J-
:•'• >^:
&«<•,
Figure 3-72. Example of a plugged distribution plate.
174
66
-------
corrosion or misalignment results in increased emissions due to sneakage
below the treatment zone.
The lateral drag-chain system consists of an individual chain system in
each field that moves perpendicular to the gas flow. Positive baffling is
provided between each field to prevent sneakage.
Wet-Bottom ESP—In most recovery boiler ESP designs, salt cake rapped
from the plates is allowed to fall into a pool of black liquor maintained in
the bottom of the ESP. Agitators are used to dissolve the particulate in the
liquor. The black liquor that has been concentrated in the direct- or
indirect-contact evaporator is pumped into the ESP bottom, where it is en-
riched with the collected salt cake. The agitation is slow and a crust of
salt cake is normally maintained on the liquor surface. A continuous flow of
liquor is maintained in the bottom and is controlled by the boiler firing
rate. The level is adjusted to maintain a positive seal under the antisneakage
baffles between fields. If the agitation or mixing efficiency is poor, salt
cake may build up in the corners of the bottom and possibly cause bridging
and grounding of the electrical discharge system (i.e., clinker formation).
Because overfilling the bottom can ground the system, the liquor level must
be monitored continuously. Corrosion of the baffles between the fields and
the shell wall is also a serious problem in wet-bottom ESP's.
Rappers—Rapping is required in three areas in an ESP: 1) the collec-
tion plates, 2) the discharge wires, and 3) the distribution plates.
Collected salt cake must be removed from the plates to maintain power
input levels (secondary current) .in each field. Rapping effectiveness
depends on the salt cake properties, the temperature, and the transmittahce
of energy through the anvil/support system to the plates. Loose or corroded
linkages dampen the vibration and result in ineffective removal of salt cake
in the lower portion of the plates.
Salt cake must be^removed from the wire to reduce the effective wire '
diameter and keep the corona initiation voltage low, which increases the
field strength and the power input. Rapping of the wires is similar to plate
rapping, but fewer rappers are used per field. Effective rapping is accom-
plished by creating a standing wave in the wire during the rapping process.
As in plate rapping, efficient transmittal of rapping forces is required to
175
-------
remove the collected salt cake. As a result, loose connections and corrosion
must be prevented to ensure effective rapping.
Distribution plates must be rapped to remove deposits and prevent mal-
distribution of the gas flow. Generally, a rigid connection is made between
the rapper shaft and the center of the perforated plate.
Figure 3-73 shows a typical rapper pattern on a modern ESP using external
top rapping. The number of.rappers per field and per chamber depends on
rapper size and plate design. Usually there are more plate rappers than
discharge rappers. The number of rappers is based on.the plate area and the
^y
linear feet of discharge wire per field.
Plate and wire frame rapping can be accomplished by magnetic impulse
gravity impact (MIGI) rappers, pneumatic cylinders, or external or internal
rotating hammers.
Rapping frequency depends on dust loading and ESP design. Typically,
the inlet fields are rapped more frequently than the outlet fields because
more dust is collected there. Rapping intensity may be changed to remove
sticky or cohesive dust. Overrapping can result in fracture of the dust cake
and resulting rapper reentrainment puffs or a snowing condition.
Rapping intensity in MIGI rappers can be changed by increasing the
voltage to the integral DC coil (increasing lift) or by changing the weight
of the piston. Generally, the piston weight is 20 pounds, and the lift
height can be adjusted to yield an impact of between 6 and 20 ft/lb. The
amount of plate area per rapper is defined by the design.
Typical failures of MIGI rappers occur as a result of coil failure
(i.e., open coil).
Pneumatic impulse rappers are double-action air cylinders with integral
pistons. At a set pressure, the frequency and impact force of the rapper are
defined. Various kinds of rapper failures can cause the piston to freeze.
The most common failure results from water in the compressed air lines.
Water causes sludge or rust, which restricts the piston motion. Condensation
as a result of air compression must be removed through the use of traps,
dehydration units, or desiccant units. Failure to install or maintain drying
equipment is the main cause of pneumatic rapper failure.
Failure of rappers can also occur as a natural result of piston seal
wear. A normal preventive maintenance program of replacing seals can mini-
mize this method of failure.
176
-------
X X
0 0
X X
X X
0 0
X X
X X
o o
X X
X X
0 0
X ' X
X X
0 0
X X
X X
0 0
X X
X X
0 0
X X
X X
0 0
X X
X X
X
Figure 3-73. Typical rapper layout on a modern two-chamber precipitator.
177
-------
Pneumatic rappers are activated by opening an air valve for a preset
period of time. The valve is opened by an electrically activated solenoid.
Although* solenoid failure is infrequent, it can cause rapper failure.
Vibrators consist of two plates separated by an air gap. The upper
plate is attached to an electric coil. Passage of an alternating current
through the coil creates a strong magnetic field in the upper plate, which
causes the plates to come together with great force. Upon impact, the elec-
tric circuit is broken and the gap is opened through the use of compression
springs. The rapid closing and opening of the circuit results in a high-
frequency vibration in the rapper body, which is attached to the rapper rod.
Rapping intensity can be increased by changing the voltage to the vibrators.
Failure of the rapper occurs as a result of spring failure, a coil short, or
a change in air gap setting.
Falling-hammer rappers consist of a number of swinging hammers attached
to a rotating shaft. As the shaft rotates, the hammer is brought into an
elevated position, from which it falls in an arc and strikes an anvil attached
to a number of plates or wire frames. The force applied to the anvil is de-
termined by the weight of the hammer, which can be changed by increasing the
mass. The frequency of the strike is determined by the rotation rate of the
hammer shaft. Rapper failures occur as a result of hammer/anvil misalignment,
hammer fracture, hammer/shaft connection failure, and shaft chain-drive
failure. The frequency, duration, and pattern of rapper activation are de-
termined by either electrical or mechanical control systems.
One mechanical control system consists of a synchronous motor with an
electrical contact on the edge of a wheel. Electrical contacts located along
the perimeter of the wheel are attached to individual rappers. The pattern
of rapping can be changed by adjusting the location of the leads. The total
time between rapping cycles can be adjusted by changing the wheel rotation
rate. Failure of the control system can occur if the contacts become loose,
burnt, or misaligned.
A second type of mechanical rapper control system consists of individual
contacts for each rapper, which are activated by a cam. A number of cam/
switches are attached to a shaft with the cam lobes offset. Rotation of the
cam shaft results in a preset rapper frequency and pattern. The cycle time
can be changed by changing the rotation rate. System failures result from
178
-------
burning of cam switch contacts, cam breakage and wear, and synchronous motor
failure.
Several types of solid-state rapper control systems are in use. Most of
these systems allow changes in frequency, intensity, duration, and pattern of
individual rappers or group of rappers. Failures of these rapper controls
result from thermal decomposition of solid-state components (resistors,
transistors, diodes, transformers,, etc.), loose wires and contacts, and
failure of breakers.
Insulators—Insulators are used to isolate the high-voltage distribution
system from the ESP shell. Cracking of the insulator as a result of moisture
or electrical tracking can cause a loss of a field (T-R section).
Tracking occurs when dust or moisture accumulates on the insulator
surface. The deposit provides a conductive layer through which the high
voltage may pass to the ground. The circuit created generates heat on the
surface of the insulator that may eventually result in insulator cracking.
Depending on its location, insulator failure can cause a short that trips out
a T-R set.
Insulator tracking may be, identified through use of an air load or gas
load test. A typical pattern generated by insulator tracking is shown in
Figure 3-74.74
Clearance Between Plates and Wires—The clearance between plates and wires
must be maintained at the design tolerances to prevent premature sparking and
to allow maximum power input. Reduction in clearance can occur if the upper
or lower wire frame is not aligned with the plates. The alignment is changed
by adjusting the standoff insulator position. Clearance problems in wire
frame designs can occur as a result of- warpage of the wire frame or misalign-
ment of the wire frame alignment system.
Plate-to-wire clearance can also be reduced as a result of plate warp-
age. Minor warpage problems can be solved by physical straightening. Severe
clearance problems can be corrected by removal of wires in the area of the
warpage. Care must be taken to minimize the removal of wires within a single
passage or lane.
Plate warpage can be caused by corrosion, unequal heating during startup,
fire in the ESP, or air inleakage. Severe uncorrectable plate warpage that
limits corona power input may require replating of the unit.
179
-------
1000
900
800
700
^ 600
UJ
I 500
>-
I 400
CD
O
UJ
CO
300
200
100
0
__ A * SPARKOVER
A
REFERENCE
CURVE
LEAKAGE
COMPONENT —1
10 15
20
25
30
35 40
45
50
SECONDARY VOLTAGE, kV
Figure 3-74. Typical pattern generated by insulator tracking.
180
74
-------
Corrosion—Corrosion appears to be the most serious maintenance problem
in the long-term operation of recovery boiler ESP's. Corrosion attacks the
ESP shell and internal components. Advanced corrosion is accelerated by air
inleakage through corroding areas in the ductwork, around access doors, or in
areas near the liquor-flue gas interface in wet-bottom units. Corrosion in
internal areas causes a reduction in the removal of collected particulate
from plates and wires because rapper effectiveness is severely impaired. A
survey by the Technical Association of the Pulp and Paper Industry (TAPPI) of
19 noncontact recovery boilers installed between 1974 and 1979 indicated that
63 percent had some corrosion problems and 26 percent had severe corrosion
75
problems. Based on the operating conditions of the 19 boilers, the average
temperature of those with serious corrosion problems was 361°F. The average
temperature of those reporting no corrosion problems was 384°F. Figure 3-75
shows an example of severe corrosion of collection plates in .an ESP.
Figure 3-75. Example of severe corrosion of collection plates.
181
-------
76
Corrosion in a mild steel component has the appearance of multiple, thin
metal layers separated by areas of bright orange granular material. As
oxidation occurs, the metal expands greatly in volume. The loss of a few
thousandths of an inch may expand to 1/2 inch as a result of crystal growth
and restructuring. This expansion, when in a confined area such as between
plate mounting brackets or structural components, can cause a great deal of
stress and force components apart.
Corrosion occurs at a faster rate in the colder areas of the ESP.
Localized cooling occurs when heat loss through the shell is highest, i.e.,
where outside stiffeners or structural columns are attached to the shell.
The primary corrosive agent in kraft recovery boiler,ESP's is sulfuric
acid. Flue gases from the boiler contain hLO vapor with a high concentration
7fi
of SOg. A portion of the S02 is converted to S03 in the boiler. At tem-
peratures below 415°F, 99 percent of the S03 vapor combines with the water
present to form sulfuric acid vapor (H2S04). As the temperature of the gas
stream is reduced, the H2S04 vapor becomes saturated and forms an acid mist
The temperature at which the sulfuric acid mist condenses on a cool
surface is defined as the acid dewpoint. The dewpoint actually is the begin-
ning of the saturation process, and the exact acid dewpoint temperature
cannot be determined. The rate of condensation reaches a maximum at 40° to
60°F below the theoretical dewpoint.
Severe corrosion occurs when water condenses on surfaces and dilute
acidic solutions are formed. The rate of corrosion is aggravated by air in-
leakage.
The dewpoint temperature of uncombined water is different from the acid
dewpoint. The typical water dewpoint in recovery boiler flue gas streams is
165°F, which must be avoided.
The most severe corrosion in wet-bottom ESP's occurs in the area above
the liquid level and below the treatment zone. This area is baffled and is
not a part of the main gas volume passing through the ESP. Vapors from the
black liquor in the bottom are extremely corrosive. The activity of the
78
vapors increases with high oxygen and sodium sulfide concentrations. Also,
water vapor raises the local dewpoint temperature. The temperature of the
black liquor, which is normally below 180°F, results in a cool shell tempera-
ture surrounding the liquor. This cool shell temperature causes a gradual
182
-------
79
decrease in shell temperature between the treatment zone and liquor level.
The lower wall temperature is usually below the acid dewpoint and near the
moisture dewpoint.
Maintenance of a uniformly high temperature in the ESP is important in
reducing the rate of corrosion. The temperature may be increased by reducing
air inleakage, insulating the shell, and heating the shell.
The temperature within the ESP is not uniform, and variations in the
shell occur as a result of contact with structural members, degree of in-
sulation, exposure, and orientation. Areas of low gas circulation in the
ESP typically have the lowest temperature and highest rate of corrosion.
In general, a well-constructed, well-insulated steel shell ESP will
experience minimum-corrosion at flue gas temperatures above 350°F.
Units with a flue gas temperature between 243°F and 265°F have problems
in the areas of highest heat loss. Units with temperatures below 300°F re-
quire supplemental heating.
The amount of heating required varies with flue gas temperature,
degree of insulation, and other environmental factors such as wind loss
and degree of exposure. The heat requirements are generally 500,000 Btu/h.
Table 3-22 summarizes the parameters and potential malfunctions that
.affect emission rates. As noted in the table, many parameters are inter-
related and may have more than one effect. For this reason, the inspector
should be able to confirm as many operating parameters as possible during the
performance tests to define deviations from normal operation or practice.
3.3.1.6 Inspection of Recovery Boiler--
This subsection summarizes the activities associated with the inspection
of kraft recovery boilers and ESP's. It also identifies the kinds of data
that should be collected during an inspection and the procedures that should
be used to evaluate these data.
Both plant personnel and regulatory agencies make inspections of process
and control equipment. Although the thoroughness and scope of these inspec-
tions may be vastly different, they have a common goal--to determine the com-
pliance of the source with applicable State and Federal emission limits con-
tained in the SIP and the NSPS.
To achieve the stated goal of determining the compliance status of
the source, the inspection must be well planned, and sufficient time must be
183
-------
TABLE 3-22.
SUMMARY OF THE EFFECTS OF KEY RECOVERY BOILER OPERATING
PARAMETERS
Parameter
Factor
influenced
Effect
Firing rate
Char bed tempera-
ture
Primary air
Char bed height
Secondary air
Excess air
Firing residual
oil containing
high sulfur
(continued)
Flue gas volume
Vertical velocity in com-
bustion zone
Primary air volume
Sodium and sulfur vapor
pressure
S02 content of flue gas
Flue gas volume
Vertical velocity in com-
bustion zone
Sodium and sulfur evapora-
tion rate due to tempera-
ture and diffusion condi-
tions
Char bed temperature
Rate of primary air
Vertical velocity in corn-
bustiorf zone
Excess air
Vertical velocity in com-
bustion zone
Flue gas volume
Increased S03 formation
Increased SO3 formation
184
Changes ESP efficiency
Changes uncontrolled inlet
grain loading
Changes diffusion condi-
tions and affects rate of
evaporation of sodium from
smel t
Changes uncontrolled par-
ticulate grain loading
Changes composition of dust
cake
Changes ESP efficiency
Changes uncontrolled par-
ti cul ate emission rate
Changes uncontrolled par-
ticulate emission rate
Changes uncontrolled par-
ti cul ate emission rate
Changes uncontrolled par-
ti cul ate emission rate and
flue gas volume to ESP
Changes uncontrolled par-
ticulate emission rate
Increases flue gas volume
Changes uncontrolled par-
ticulate emission rate
Changes ESP efficiency
Changes particulate com-
position; reduces ESP power
input and efficiency
Changes particulate com~
position; reduces ESP power
input and efficiency
-------
TABLE 3-22 (continued)
Parameter
Factor
influenced
Effect
Primary tempera-
ture
Black liquor
chloride content
Black liquor
heat value
Sodium-sulfur
ratio
Flue gas volume
Corona power
Corona initiation
voltage
Corona power dis-
tribution within
chamber
Corona power per
chamber
Superficial
velocity
Flue gas oxygen
at ESP inlet
(continued)
Smelt bed temperature
Smelt bed elemental
equilibrium
Flue gas volume
Char bed temperature
Smelt bed elemental
equilibrium
Superficial velocity
Specific current density
Specific corona density
(W/1000 acfm)
Buildup of dust on elec-
trodes
Buildup of dust on elec-
trodes
Plate alignment
Buildup of dust on elec-
trodes
Plate alignment
Dust reentrainment during
rapping
Increased total gas volume
Flue gas volume
Increased S03 formation
Flue gas temperature
185
Changes uncontrolled par-
ti cul ate emission rate
Changes TRS emission rate
Changes uncontrolled par-
ti cul ate emission rate
Changes uncontrolled par-
ti cul ate rate
Changes ESP efficiency
Changes uncontrolled par-
ti cul ate rate
Changes ESP efficiency
Changes ESP efficiency
Changes ESP efficiency
Changes ESP efficiency
Changes ESP efficiency
Changes ESP efficiency
Changes ESP efficiency
Changes dust properties and
increases plate deposits;
reduces ESP power and effi-
ciency
Increases sparking, which
reduces power and effi-
ciency
-------
TABLE 3-22 (continued)
Parameter
Flue gas tempera-
ture
Factor
influenced
Acid dewpoint
Moisture dewpoint
Effect
Corrosion
Structural failure
Increases air inleakage
Reduces corona power
Reduces ESP efficiency
provided to acquire the necessary data on boiler and ESP operating conditions.
This section outlines the overall scope of'the inspection and the procedures
the inspector may use to determine the compliance status of the kraft recovery
boiler with the particulate emission limits defined by applicable State or
Federal regulations. The approach the inspector should take is to identify
those operating conditions or variables that indicate operation outside the
accepted norms for a particular boiler/ESP system. Normal values or condi-
tions are established during the initial performance stack test or are based
on the accepted state-of-the-art. This approach, which is generally referred
to as a "baseline approach" to source evaluation, may be used by both source
personnel and regulatory agencies to initiate a more detailed analysis or to
trigger a performance stack test to verify compliance with the emission stanr
dard.
The early identification of operation and maintenance (O&M) problems
reduces the extent and the occurrence of excess emissions and allows the
plant to schedule outages or make on-line adjustments to maintain production
and operate within the prescribed emission limits.
Most inspectors make visible emission observations in accordance with
EPA Method 9 and record the production rate of the boiler. This level of in-
spection is usually not adequate for determining compliance or evaluating the
maintenance procedures of a recovery boiler. The inspector must visually in-
spect the ESP and record pertinent operating variables for both the boiler
and the ESP. Key parameters also must be measured to allow the inspector to
determine if the ESP is being operated within design limits and if there is
sufficient reason to require documentation of compliance through a stack test.
186
-------
The following is a summary of specific things that should be checked
during the inspection.
Opacity—The inspector should conduct an EPA Method 9 observation of the
recovery boiler plume,,either at a point just before steam condensation (in
the case of a detached water vapor plume) or just after steam dissipation (if
the plume is attached). The readings should be made at 15-second intervals
and averaged over a 6-minute period. The total evaluation time should be at
least 30 minutes and preferably over one ESP rapper cycle (outlet field rap-
pers). The 6-minute averages should be plotted to identify any cyclic pat-
tern. An example of such a pattern is shown in Figure 3-76.
Transmissometer Data--In units equipped with opacity monitors, the in-
spector should record the current 6-minute average opacity and review the
previous 4 hours of monitor output to determine if a cyclic pattern is occur-
ring. To ensure that the output values are accurate, the inspector should
request the plant to place the monitor in the calibration mode with respect
to zero and span. As part of the initial monitor certification, the inspector
should have data available on the recorder scale factors and effective stack
diameter. Average opacity readings from the Method 9 observation should be
compared with average transmissometer readings for identical periods. A major
deviation between the values may indicate possible monitor error. It should
be noted that the manual and instrument methods are not equivalent. In sources
that have real-time monitor output, instantaneous opacity spikes (rapper re-
entrainment) will generally be included in the 6-minute average and the value
will generally be higher than that obtained by the manual method. The in-
spector should note the frequency and magnitude of rapper spikes and determine
if a pattern is occurring (inlet to outlet field rapper pattern). The opacity
and rapper reentrainment pattern in each chamber should be evaluated if
separate monitors are installed in each duct. Figure 3-77 shows a typical
monitor output with severe rapping reentrainment losses. The opacity should
be compared with a typical baseline value for the boiler during known emis-
sion periods (i.e., performance tests). The opacity data should be used to
evaluate conditions in the ESP.
Serious deviations in opacity between chambers can indicate gas flow
maldistribution, an increase in penetration through one chamber as a result
of rapper failure, inleakage, or low power input.
187
-------
88T
AVERAGE OPACITY, %
ro
o
CO
o
en
o
CQ
c
CO
rn
1
-a
a>
o
-h
3
C
rt-
n>
Q)
n>
cu
U3
fD
o
•o
o>
o
—1.
4?
•a
EU
(D
-s
•
CO
o
t-i CO
3>
m
en
m f*
m
tn
(Tl
-------
00
Figure 3-77. Typical opacity monitor output with- severe rapping reentrainment losses,
-------
Boiler Operating Conditions—The inspector should record such boiler
operating conditions as steam flow (103 pounds per hour), flue gas oxygen
(%), and flue gas temperature (°F) at the time of the inspection. By com-
paring these current operating conditions with the historic baseline obtained
during the performance test, the inspector can determine if the boiler is
operating at normal production levels. Major deviations from normal values
should be evaluated with respect to their impact on ESP performance and TRS
and particulate emission levels. Operating conditions, continuous emission
monitors (TRS and opacity), and ESP operating conditions should be used to
determine if the boiler is in compliance with applicable standards.
Boiler operating data that should be measured during a performance test
or an inspection are plant-specific. Each boiler is usually custom-designed
and erected with a unique instrument and control system package. The level
of instrumentation is specified by the design engineer and purchaser (plant
engineering department). Based on the size of the boiler, its cost, and the
experience of the purchaser, the instrument package may range from a straight-
forward package to one that is very complex. In general, a minimum amount of
instrumentation is necessary for safe operation of the boiler, and all facil-
ities will have this level of instrumentation. More complex instrument
packages can include an automated computer control system that allows the
source to optimize combustion and increase the overall efficiency of the
operation.
Most critical boiler parameters are recorded on continuous strip charts,
or circular chart recorders, and copies may be obtained after the stack test
(at the end of the day) to provide the necessary documentation. Most mills
require the boiler operator to record key parameters at set intervals on a
log sheet or in a log book. The log sheet is typically divided into the fol-
lowing general measurement areas: black liquor, auxiliary fuels, forced air,
furnace drafts, gas temperatures, feedwater, steam, and miscellaneous items.,
Table 3-23 lists the items or conditions that must be recorded during
the stack test or inspection. The list is based on a typical boiler and
would require adjustment for individual installations.
Integrator readings also should be recorded at the beginning and end
of each test run for the following parameters:
190
-------
TABLE 3-23. RECOVERY BOILER OPERATING PARAMETERS TO BE RECORDED
DURING PERFORMANCE TESTS OR INSPECTIONS
Parameter
Black liquor
Auxiliary fuels
Forced air
Variable
Liquor flow
Black liquor pressure
Gun size
Number of guns
Black liquor temperature
BLS to guns
BL flow to ESPa
BLS to ESPa
BL flow to evaporator3
BLS to evaporator9 .
Oil flow
Number of guns
Oil pressure
Oil temperature
Natural gas rate
Primary air flow
Primary air pressure
Primary air temperature
Secondary air flow
Secondary air pressure
Secondary air temperature
Tertiary air flow
Tertiary air pressure
Tertiary air temperature
Total air flow
Units
gal /mi n
(103 Ib/h)
psig
None
None
°F
%
gal /mi n
. (103 Ib/h)
V
la
gal /mi n
(103 Ib/h)
%
gal/h
None
psig
°F
103 fts/h
scf/min
in. H20
°F
scf/min
in. H20
°F
scf/min
psig
°F
scf/min
(continued)
191
-------
TABLE 3-23 (continued)
Parameter
Furnace drafts
Gas temperatures
Feedwater
Steam
Chemicals
Miscellaneous
Variable
Furnace
Superheater outlet
Boiler outlet
Economizer outlet
ID fan inlet
Preci pita tor inlet
Superheater outlet
Boiler outlet
Economizer outlet
Evaporator outlet
ID fan outlet
ESP inlet
Flow
Pressure
Temperature
Flow
Drum pressure
Superheater temperature
Salt cake makeup
Flue gas oxygen (boiler outlet)
Black liquor heat value
Units
in. H20
in. H20
in. H20
in. H20
in. H20
in. H20
°F
°F
°F
°F
°F
°F
(103 Ib/h)
psig
°F
(103 Ib/h)
psig
°F
Ib/min
%
Btu/lb BLS
To be used with correction factors to calculate BLS to guns where not
measured directly.
192
-------
o Black liquor flow (10 pounds or gallons)
o Steam flow (pounds)
o Steam used in soot blowing (pounds)
o Oil flow (pounds or gallons)
o Natural gas flow (103 cubic feet).
Based on data obtained from the log, steam tables, integrator readings,
and boiler design data, the following values should be calculated:
o Average steam flow for each run
3
o Average BL fired (10 pounds/hour) for each test run
o
o Average BLS fired (10 pounds/hour) for each test run
o Heat input (10 Btu/hour) to the boiler for each test
run for each fuel fired (BL, oil, natural gas)
o Average boiler output (10 Btu/hour) for each test run
o Boiler thermal efficiency (heat output/heat input)
for each test run
o Boiler excess air (percent)
o Pounds of steam per pound of BLS fired.
ESP Power Levels—The inspector should record ESP power levels (primary
current, primary voltage, secondary current, secondary voltage) for all ESP
fields and chambers. The spark rate should be estimated from a manual count-
ing of meter deflections. The inspector also should plot ESP functions by
field (inlet to outlet) for each chamber. Deviations from optimum values
determined from baseline or normal values should be used to determine internal
ESP conditions and to analyze potential emission levels. If recent V-I
curves are not available, the inspector should request the plant environ-
mental engineer or electrician to produce a V-I curve for each field. Data
from the V-I curves should be used to target the inspection of the rappers,
the gas distribution system, and local cooling and to check for inleakage.
Serious deviations from normal values should be evaluated with respect to
their impact on potential emission levels. Opacity data for each chamber are
helpful in determining the effect of corona power levels. In general, the
efficiency of the ESP follows the pattern of the Deutsch-Anderson equation or
Matts-Ohnfeldt equation for prediction of emissions as a function of corona
power and gas flow rates.
193
-------
ESP Rappers—The inspector should determine if discharge electrode,
collection plate, and distribution plate rappers are functioning. Particular
attention should be given to areas identified as having decreased corona
power or areas where higher corona initiation voltages were noted. Initial
evaluation of the rappers should be based on sound intensity and/or failure
to activate.
Flue Gas Volume—Because ESP performance is affected by total gas volume,
the inspector should make an estimate of the volume based on black liquor
firing rate, flue gas oxygen, and temperature. Most plants monitor flue gas
oxygen at the economizer outlet rather than at the ESP outlet. An estimate
of the flue gas volume must be based on ESP outlet conditions. The inspector
should be equipped with portable temperature measurement equipment (i.e.,
thermometer or thermocouple) and portable oxygen measurement equipment (i.e.,
Fyrite oxygen analyzer). The flue gas volume may be calculated from a plant-
specific F-factor (dry standard cubic feet/pound BLS) with correction for
flue gas oxygen5 moisture, and temperature.. Temperature and oxygen measure-
ments should be made at the outlet of each chamber where possible (accessible).
The following presents the method used to calculate the flue gas volume
at the ESP inlet or outlet. It should be noted that the corrections for the
flue gas oxygen are for dry standard gas volume not wet gas volume.
Q =
BLS
20.9
dry 120.9 - %02
460) °R
528°F
where
BLS = black liquor solids firing rate to the boiler.
Fd » F-factor for black liquor solids in dscf/lb BLS.
% Og s oxygen content of flue gas at ESP inlet in percent.
F = standard cubic feet of water vapor generated from combustion
of hydrogen per pound of black liquor solids.
FE = standard cubic feet of water vapor evaporated in direct
contact evaporator.
F = standard cubic feet of water vapor added to flue gas stream
as a result of soot blowing.
194
-------
TS = temperature of the flue gas at the ESP inlet in °F.
The amount of water evaporated in a direct-contact evaporator may be
determined by using liquor flow rates and liquor solids content entering and
leaving the unit (see Figure 3-78). This method is a simple mass balance
based on the assumption that the total amount of solids does not change in
the evaporator. This is not strictly true because the liquor does adsorb
salt cake from the flue gas stream. This effect is considered negligible,
however, in calculation of the water lost. If a more exact estimate is
desired, adsorption rates may be estimated depending on flue gas volume (actual
cubic feet/minute) and the uncontrolled boiler dust loading (grains/actual
cubic feet per minute). In general, a cascade type evaporator may remove 50
percent of the uncontrolled particulate. In most units this will increase
the total liquor solids mass by less than 5 percent and usually will result
in less than a 2 percent error in the true gas stream moisture.
If a more exact or plant-specific value is desired, the F-factor can be
calculated from stack test gas volume (dry standard cubic feet/minute), flue
gas oxygen content, and firing rate of black liquor solids.
Using this method, the operator, inspector, or plant environmental per-
sonnel can make a day-to-day determination of the flue gas volume being
treated by the ESP without the expense of conducting stack flue gas volume
determinations with a Pitot tube.
When flue gas oxygen increases above the normal ranges, the source of
inleakage should be identified immediately and appropriate repairs made to
reduce the inleakage. Failure to reduce the inleakage will not only cause
excess emissions, but because of the cooling effect of the ambient air, "will
also cause low-power input, excess sparking, and corrosion.
The amount of steam used in soot blowing is not generally measured, but
based on discussions with ESP and boiler manufacturers and limited data from
pulp mills, the value is estimated to be 8 to 10 percent of the rated steam
81
flow of the boiler. The value is expressed in pounds of water vapor per
minute, which must be converted to standard cubic feet per minute. The values
of F and F . may be compared with the values obtained during a stack test
as a check on the validity of the derivation. The variables affecting these
values are too numerous to list here, but they include wood species and mix,
process step variables, quantity of inorganic salt cake recycled to the
195
-------
FLUE GAS (IN)
FLUE GAS (OUT)
f
IQUOR (OUT)
EVAPORATOR
BLACK LI QUO
where
A =
i s
B *
water evaporated (Ib/min) = A p. - B p
gallons of black liquor to the evaporator
density of black liquor into the evaporator in Ib/gal
gallons of black liquor from the evaporator
density of black liquor out of the evaporator in Ib/gal
% BLS.
A P.. ( TT^TT-) = B
% BLS.
where
% BLS.. = solids content of black liquor entering evaporator
% BLSQ = solids content of black liquor leaving evaporator
Figure 3-78. Method of calculating additional moisture in the flue gas
stream due to direct-contact evaporator.
196
-------
recovery boiler, percent solids in the black liquor, and heating value of the
black liquor.
When deriving the black liquor F-f actor from stack tests or by theoretical
equations, it is convenient to work in terms of standard cubic feet of gas be-
cause it allows for addition of values without constant correction for different
gas conditions. The ESP, however, must be analyzed at the actual gas condi-
tions (i.e. , at the measured temperature and oxygen content). Once established,
the values of F. and FtQ+ai tend to remain relatively constant provided no
significant changes in the process occur.
The use of an established F-factor to determine gas flow through the ESP
requires relatively little calculation. Only the following are needed: value
of the F-factor (dry), firing rate of the black liquor, percent BLS, density
of the black liquor, and the temperature and oxygen content of the flue gas.
The ESP dimensions can be obtained from engineering drawings (blueprints).
Using these dimensions will usually produce superficial velocity values that
are slightly lower than actual values. The area input into the calculations
does not account for the cross-sectional area blocked by the plates and wires.
The calculated value should be in the range of 2.5 to. 4.0 ft/s, and the lower
values generally are recommended. Obviously, as the superficial velocity in
an ESP decreases, treatment time will increase. Also, if the superfical
velocity exceeds 8 ft/s, not only will the treatment time drop, but reentrain-
ment of captured particulate may occur as a result of the high velocity strip-
ping material off the ESP plate. Thus, it is important to consider the gas
volume through the ESP. Gas volume is especially critical if there is a pos-
sibility of high excess air levels resulting from air inleakage or improper
boiler operation, or if high gas volumes could occur from overfiring the
recovery boiler.
Another value that should be checked is the actual SCA. This value re-
o
lates the total available plate area to the gas volume (ft / 1000 acfm), and
when compared with design or baseline values, indicates ESP performance
capabilities. Generally, an increase in the SCA (actual) means improved
performance, but other factors are involved; therefore, a comparison of actual
SCA with design or baseline values is not meaningful by itself.
ESP Corona Power— The evaluation of ESP performance is based primarily
on T-R electrical readings from both primary and secondary meters. The
197
-------
inspector can define most of the problem areas and determine the ESP performance
level externally through the use of a combination of calculations and the
evaluation of trends present in the data.
Secondary meter readings are preferred for an evaluation of ESP perform-
ance and diagnosis of ESP problems because they tend to reflect more closely
the power input characteristics of the ESP. For example, the secondary
voltage at which secondary current is initiated can be used to determine if
ash is building up on the wires and if insulators are tracking. These data
allow adjustment of the vibrator intensities for optimum power input and
indicate the need for cleaning or replacing insulators. The secondary cur-
rent/voltage curve also can be used to diagnose clearance problems and indi-
cate plate warpage and/or wire frame misalignment on a section-by-section
basis.
No voltage on the secondary side may indicate an open primary circuit.
The circuit breaker may be open or tripped or a reactor secondary may be
open. High voltage on the primary side and no voltage on the secondary side
may be due to a faulty, open, or disconnected ESP; an open bus; or a faulty
rectifier. Low voltage on the secondary side coupled with low voltage on the
primary side could be the result of leaks in the high voltage insulation,
buildup of dust in the discharge system, excessive dust on electrodes, or
CO
swinging electrodes.
No secondary current and no secondary voltage indicate an open primary
circuit. Irregular secondary current coupled with low secondary voltage
indicates a high-resistance short in the circuit, possibly due to excessive
fi°
dust or arcing.
Because primary meters can provide an-indication of the voltage and
current going to the ESP, they can be used if they are the only meters
available to obtain a more indirect measure of ESP performance. For example,
no voltage on the primary side may indicate an open primary circuit. The
circuit breaker may be open or tripped, or the reactor secondary may be open.
High primary voltage may indicate an open transformer primary or improper
connection of an ESP, a faulty, open, or disconnected ESP, an open bus, or a
faulty rectifier. Low primary voltage may indicate leaks in the high-voltage
en
insulation, excessive dust on electrodes, or swinging electrodes.
198
-------
An ammeter reading that shows no primary current associated with no
primary voltage indicates an open primary circuit, which may be due to open
circuit breakers or an open reactor secondary. Irregular primary current
coupled with low primary voltage indicates a high-resistance short in the
circuit. The' causes of this condition include electrode short, excessive
dust on collecting surfaces, excessive dust on electrodes, support insulator
CO
arcing, and the possible presence of foreign material.
When recording and evaluating ESP meter readings, the inspector should
realize that trends in the voltage and current levels are the first indicators
of ESP performance levels. Specifically, a gradual rise in the ESP secondary
current levels should be apparent as one progresses from the inlet to the
outlet fields (Figures 3-79 and 3-80). The secondary current level seldom
exceeds 250 milliamperes (mA) in the inlet fields. Typical designs call for
a current level of 0.02 mA per linear foot of wire length in the inlet fields.
Sparking in the inlet fields is evidenced by deflection of the meters during
a spark. When the meters indicate progression toward the outlet field,
secondary current levels should increase and sparking should decrease, with
almost no sparking occurring at the outlet. Most ESP outlet field T-R cur-
rent levels should be at least 85 percent of the T-R set current rating
(e.g., if the secondary current rating was 1000 mA on the outlet T-R, a
design reading of at least 850 mA is normally expected) and in the range of
0.06 to 0.08 mA per linear foot of discharge wire.
Another trend the inspector, may note is a gradual decrease in the secon-
dary voltage from inlet to outlet. Larger ESP's (with five fields or more)
often experience lower voltages in the inlet field due to sparking, an in-
crease in voltage in the middle fields; and then a slight decrease in voltage
in the outlet fields.
The gas entering the inlet field of the ESP contains the greatest
concentration of particulate, and the greatest quantity of particle charging
occurs in this field. Consequently, a great many electrons are captured
during charging, and the rate of charge transfer from the discharge to the
collection electrodes is lessened because the mass of the particles migrating
toward the plates is considerably larger than that of the electrons or
charged molecules (ion mobility). In addition, more force is required to
199
-------
LU
CC
O
§
t/1
J_
2 3
FIELD NUMBER
0.10
0.08
0.06
0.04 >•
, cc.
C£.
U
QC
o
<_)
LU
tn
0.02
Figure 3-79. Optimum secondary current distribution in ESP serving kraft
recovery boiler, assuming uniform rapping and wire size in all fields.
200
-------
ee.
§
C/5
>CHAMBER B
•CHAMBER A
FIELD
Figure 3-80. Secondary current pattern for two ESP chambers; Chamber A
is having maintenance problems that limit power input.
201
-------
move the electrons from the corona discharge to the collecting plates because
the charged particles in the interelectrode space act as a large electro-negative
cloud that repels the electrons. This is known as the "space charge" effect
and is usually present only in the first and sometimes the second electrical
field. The greatest sparking occurs in this field because this large cloud
of charged particles tends to form numerous paths for interelectrode gas
breakdown (spark formation). Thus, operation in this field tends to be a
balance between very high electrical field strengths and reduction of spark
rates to moderate levels without excessive waste of electrical power.
As the gas moves through the ESP and the particles migrate to the plate,
fewer particles remain to be captured and to inhibit the flow of electrons.
Therefore, less force (voltage) is required to obtain a high current flow.
Usually the increase in current is much greater than the decrease in voltage,
and the net effect is an increase in power input from ESP inlet to outlet.
The relationship or ratio between voltage and current levels for each
T-R is not constant throughout the ESP (from inlet to outlet). Changes in
the voltage-current relationship are due to capacitance and resistance char-
acteristics of the particulate, as well as to inefficiencies in the circuitry
of the T-R set at various operating levels. The relationship between the
primary and secondary voltage and current levels will not be constant from
inlet to outlet for the same reasons. Thus, general relationships between
primary and secondary meter readings are difficult to establish and will
change with dust characteristics. Over very narrow operating ranges within a
T-R, however, the relationship between primary and secondary voltage and
current levels may be considered linear. This relationship is useful in
evaluating ESP performance if a secondary meter is out of service and the
corresponding primary meter on the other side of the transformer is operating.
Although trends in the voltage and current levels are important in
evaluating ESP performance, the corona power input provides one of the most
useful indicators. Secondary meters are preferred in determining corona
power because they are more indicative of actual power input to the ESP.
Primary meters may be used for this calculation, however, if they are the
only meters available, provided an appropriate efficiency factor is used.
Corona power calculations are simply the product of the secondary voltage
202
-------
multiplied by the secondary current to yield watts of power to the ESP field
from the T-R. This calculation should be made for each field of the ESP.
*
When both primary and secondary meters are available, the primary voltage and
current levels should be multiplied to yield primary power Input to the T-R
in watts. The value produced by the primary meter must be higher than that
obtained from the secondary meter. If the secondary power product (corona
power) is higher than the primary power input, then the values on the meters
are incorrect. Isolating the malfunctioning meter, however, may be very
difficult.
Electrical losses occur in the T-R during increases in voltage and rec-
tification to an unfiltered DC wave form. In addition, losses in the T-R
control circuitry reduce the efficiency of transferring from primary input
power to secondary corona power. This efficiency factor for the T-R may be
calculated through the use of the ratio of secondary power to primary power:
T-R efficiency = secondary power
J primary power
(Eq. 1)
This value usually ranges from 0.55 to 0.85, although values up to 0.90
are occasionally observed. In general, the value of the T-R efficiency in-
creases as the T-R approaches its rated output current level. Thus, T-R
efficiencies tend to be lower in the inlet fields (0.55 to 0.60) and higher
in the outlet fields (0.80 to 0.85) because of limitations imposed on the
electrical operating characteristic by particulate load and space charge
effect. A value of 0.70 to 0.75 is usually appropriate as an average for all
fields in the ESP.
Another item that should be checked in multichambered ESP's with T-R's
for each chamber is a balance of power across the ESP. The secondary current
level and power input for each field should be approximately equal across the
ESP. Some relatively small differences may occur because of rapper sequenc-
ing, slight gas flow imbalances, or differences in internal alignment. These
differences should not be large in most ESP designs, however, as chambers
with equal gas flow and equal power levels give the best ESP performance.
Once the power levels have been calculated and the patterns checked as
previously discussed, the corona power from each T-R should be added and
totalled to yield total corona power to the ESP (in watts). When multi -
chambered ESP's are installed with T-R's serving each individual chamber,
203
-------
the corona power levels should be totalled for each chamber (to indicate
balanced power), and an overall total corona power level should be calcu-
lated for all chambers.
Baseline test values, if available, should be compared with the actual
secondary current'and corona power levels. If the gas volume through the ESP
is nearly identical to the test value and the meter readings are also nearly
the same, then ESP performance is usually similar to that observed during the
stack test. It is usually not possible to compare design power levels to
performance levels, as these are usually not included in most design informa-
tion. A key indicator of ESP performance is the specific corona power, which
provides a useful value for determining whether ESP performance has changed
significantly.
Specific corona power is calculated by the following equation:
total corona power (watts)
Specific corona power = total gas volume (1000 acfm)
This value may be calculated for the entire ESP or for individual chambers.
The volume obtained by the modified F-factor (divided by 1000) is substituted
into this equation. In general, the higher the value of the specific corona
power, the higher the ESP removal efficiency. Thus, one may determine whether
ESP performance would be expected to increase or decrease by evaluating the
specific corona power.
In general, the specific corona power needed to meet NSPS for kraft re-
covery boilers is usually above 400 watts per 1000 acfm, although acceptable
performance may be obtained with lower specific corona power values if there
are no major problems with power distribution, inleakage, or rapper operation.
Equation 2 indicates that a decrease in gas flow through the ESP may lead to
an increase in performance (constant corona power is assumed). In this case,
however, the corona power is generally not constant, but increases with de-
creasing gas velocity (volume) because the bulk of the particles do not
penetrate far enough into the ESP to inhibit power input (the opposite occurs
with an increase in the gas volume). Thus, a decrease in gas volume may
substantially increase ESP performance. This relationship provides an impor-
tant reason for evaluating firing rates and excess air levels. It should be
noted, however, that poor performance is possible (but less likely) with
204
-------
specific corona power levels of more than 750 watts/1000 acfm and that specific
corona power provides a useful but not a sole indicator of ESP performance.
Another useful indicator of ESP performance is power density (watts/ft2
plate area). The value for the power density may be calculated by the equation:
(Eq. 3)
Power density = corona power input
ft plate area
This calculation may be performed for each field by substitution of the
T-R corona power and the plate area associated with that T-R into the equa-
tion. The increasing trend from inlet to outlet that should be evident may
range from 0.25 W/ft2 at the inlet to 5 W/ft2 in the outlet field. This
calculation is useful for ESP's that do not employ equally sized fields.
The trend of secondary current and corona power is not smooth, but this cal-
culation will help "normalize" the values. A corresponding calculation may
be performed on the secondary current (current density, mA/ft2) to normalize
the data.
The substitution of total corona power and total plate area into the
equation will usually determine the overall power density. Values of 1 to
3.0 W/ft are typical and usually indicate good performance. As with specific
corona power, however, performance may be poor even though the power density
may be high.
A baseline test can be used to calculate ESP efficiency. The baseline
test is used to calculate total corona power, gas volume, and efficiency or
penetration (1 - efficiency). The values from the baseline test can be used
along with a modified version of the Deutsch-Anderson equation to calculate
a constant that may be used in subsequent calculations:68
p = e-0.06 K(Pc/Q)
U
(Eq. 4)
where
Pt = penetration
K = constant
PC = corona power, watts
Q = gas flow, 1000's acfm
205
-------
This equation relates corona power input and gas flow to ESP performance.
The original form of the Deutsch-Anderson equation assumed maximum and non-
varying field density and related particle migration rate to the SCA of the ,
ESP. The particle migration rate, however, was a function of corona power
input; thus, corona power changes the particle migration rate and the ESP
efficiency. This equation is useful in predicting changes in efficiency
provided that wide variations in the specific corona power have not occurred
or the power distribution within the ESP is not seriously altered.
The value of "k" will typically fall in the 0.1 to 0.25 range for most
kraft recovery boilers. Once the value of the constant has been calculated,
it can be used in subsequent calculations in which inspection data are used.
Because the equation usually will overpredict performance within narrow
operating ranges, the efficiency predicted is based on the assumption that
all the power is used to collect particulate. As substantial power decreases
occur, however, the equation may substantially underpredict performance
because the equation becomes very sensitive to variations in the specific
corona power. In fact, the value of the constant begins to change with sub-
stantial increases or decreases of specific corona power, and the predicted
performance may be incorrect by as much as half an order of magnitude when
applied over a very wide range of specific corona power. The value of the
constant and the behavior of the ESP will also change with inoperative T-R's
(particularly those in the inlet fields). The equation will, however, pro-
vide a gross indication of a performance shift.
The underlying limitation to the equation is the assumption that the
power input and the changes in power input are the only factors that affect
ESP performance once the SCA and gas volume are fixed, and that the value of
the constant and the effective migration velocity are linearly related.
Actually, the relationship in the value of the exponent is probably a power
••
function similar to that presented in the Matts-Ohnfeldt equation, which
makes the equation less sensitive to variations in power input. Using an
equation at conditions different from baseline values still requires care,
however.
In summary, evaluation of ESP data includes plotting of secondary current
to discern an increase from inlet to outlet; calculating corona power from
206
-------
primary or secondary meters; evaluating the balance in the readings between
parallel fields; calculating total corona power, power density, and specific
corona power; and comparing these values to baseline readings. If conditions
are nearly identical, performance is likely to be similar. For small changes
of specific corona power, an indication of the magnitude of any performance
shift may be estimated by use of a modified form of the Deutsch-Anderson
equation. Care must be exercised not to use this equation when gross changes
in ESP operation have occurred unless the values are modified appropriately.
ESP External Conditions—The inspector should visually inspect the ESP
shell to determine the condition of the insulation and note any major points
of oxygen inleakage and corrosion. The external inspection should be based
on the analyses of flue gas oxygen and ESP power levels. Table 3-24 summarizes
the parameters that the inspector should measure as part of an ESP inspection.
TABLE 3-24. PARAMETERS TO BE MEASURED BY THE INSPECTOR DURING LEVEL III
INSPECTION OF RECOVERY BOILER ESP
Parameter
Flue gas oxygen
Flue gas temperature
Primary voltage
Primary voltage
Secondary voltage
Secondary voltage
Location
ESP inlet
ESP outlet
ESP inlet
ESP outlet
Each field
' Each field
Each field
Each field
Units
01 .
to
of
h
nF
°F
Volts
Amperes
. Kilovolts
Mi li amperes
Record RevJew--The inspector should review the records maintained by the
plant environmental engineer with regard to component failures, outages, and
trends to determine the performance of the unit between inspections. Typical
records that should be reviewed include weekly inspection reports, quarterly
inspection reports, annual internal inspection reports, air-load and gas-load
V-I curves, work orders, wire breakage location charts, and the number of
component failures (e.g., rappers, insulators, wires). To interpret the
207
-------
maintenance records adequately, the inspector should request a copy of the
ESP specifications and ESP construction blueprints (arrangement, wire frame,
plate detail, etc.).
3.3.2 Smelt Dissolving Tank
The following subsections describe the smelt dissolving process, identify
the major emission sources, discuss the control techniques for minimizing
emissions from the smelt dissolving tank, discuss the possible malfunctions
associated with the operation of the smelt dissolving tank, and present the
inspection procedures for the smelt dissolving tank and associated control
equipment.
The particulate emissions from the smelt dissolving tank are generally
controlled with low-energy scrubbers. The smelt dissolving tank malfunctions
that affect emissions are generally associated with the various types of
scrubbers that are used to control the particulate emissions. Table 3-25 on
page 215 summarizes the malfunctions and identifies their potential impact.
The inspection of the smelt dissolving tank area usually involves a Level
II Inspection because of the limited accessibility of sampling locations. The
level of inspection depends a great deal on the availability of instrumenta-
tion and access to the control equipment. A detailed discussion of the items
to be covered in Level I, II, and II inspections for smelt dissolving tanks
is presented in Section 3.3.2.4.
3.3.2.1 Process Description--
Molten smelt composed of sodium sulfide and sodium carbonate is drained
from the recovery furnace hearth through smelt spouts. The smelt is dis-
charged into a water-filled vessel referred to as the dissolving tank.
"Green liquor" is formed by the incorporation of the smelt into water through
quenching.
The melting point of the smelt is defined by the composition, i.e., per-
cent Na2C03 and Na2S. Figure 3-81 presents an equilibrium diagram for the
system.12*82 The hearth is generally operated at a high temperature to ensure
free-flowing smelt. Temperatures in the smelt spout are typically between
1600° and 2000°F.
208
-------
LJ
QC
UJ
o.
2I56°F
I380°F
1300
No-CO, % 100 80 60 40
0 2O 4O GO
20
80
COMPOSITION OF SMELT, % BY WEIGHT
Figure 3-81. Equilibrium diagram for a Na2C03-NA2S system
12,82
The dissolving tank is a cylindrical metal tank in which green liquor is .
formed by dissolving the smelt. The tank is vented to provide relief for the
steam produced by the quenching of the smelt.
Green liquor levels are maintained several feet below the top of the
tank to allow for expansion and steam liberation. High liquor levels result in
carryover of liquor droplets and particulate matter into the tank vent system.
The tank is agitated with compressed air or mechanical agitation to aid in
mixing. The water-cooled smelt spout enters the tank through a doghouse
enclosure installed on the top of the tank.
Steam or recirculated green liquor is used in the smelt spout to shatter
the smelt before it contacts the liquor surface. Incomplete shattering can
result in violent steam explosions as large droplets enter the water and
209
-------
release excessive quantities of steam. The major causes of explosions in
smelt tanks (which constitute a serious and potentially dangerous problem)
are smelt sulfidity, lack of shatter efficiency, lack of smelt reduction
efficiency, sodium chloride content, and sodium hydroxide.
Figure 3-82 and 3-83 show two designs for smelt dissolving tanks that
O"3
use water sprays and/or steam shatter jets. Gases that are generated from
the tank (primarily water vapor, entrained air, and TRS) are vented through a
natural draft stack or with an induced draft fan.
Because of the turbulence in the space above the liquor, significant
amounts of particulate are emitted with the steam. The uncontrolled emission
rate from the tank varies greatly depending on quench rate, tank design,
exhaust volume, and smelt chemistry. Some data indicate a strong relationship
exists between steam shatter flow and carryover. The location and amount of
84
steam used also has an effect on TRS emissions.
By proper attention to tank design, steam shatter jet location, steam
flow, and vent control, many mills are able to operate with minimum abatement
equipment on the smelt dissolving tank stack.
3.3.2.2 Sources of Emission and Control —
Particulate emissions from the smelt tank are controlled by a number of
low-energy scrubbing systems. The simplest system consists of a wire mesh
pad mist eliminator equipped with a back-flush system to remove collected
particulate (Figure 3-84). The pads are about 1 ft thick and are generally
made of stainless steel wire. Low vertical velocities through the pad must
be maintained to prevent channeling and/or liquor reentrainment (<20 ft/s).
Water flow is used on an intermittent basis. Usage can be determined by time
or pressure drop. Typical pressure drops are 1 to 2 in. H^O and back-flush
rates are 3 to 4 gal/1000 acfm of gas. Expected particulate removal effic-
iencies can range from 70 to 90 percent.
Higher removal efficiencies may be achieved by use of low-energy spray
towers or packed-bed scrubbers. The packed-bed scrubbers are generally used
after low-energy scrubbers to remove TRS emissions. The scrubbing liquor is
usually weak wash water.
The use of low-energy entrainment scrubbers with packed beds for par-
ticulate removal has achieved efficiencies above 95 percent. Typical liquor-
to-gas ratios are 4 to 8 gal/1000 acfm at a pressure drop of 4 in. H20 (Figure
3-85).85
210
-------
LEGEND
A — dissolving tonk
8 — furnace
C — vent stack
0 — green liquor
E — air line for agitation
F — dog house
G — smelt spout
H — circulated green liquor shatter spray
Figure 3-82. Smelt dissolving tank with water sprays.
83
211
-------
LEGEND
A — dissolving tonk
B — furnace
C — vent stack
D — green liquor
E — oir line for agitation
F — dog house
G — smelt spout
H — circulated green liquor spray
j — steam shatter spray
K — troy
L—vent
Figure 3-83. Smelt dissolving tank with steam shatter jets.
83
212
-------
oo
(O
c
0>
CO
I
00
n>
in
TJ
cu
D.
n>
D.
(D
t/)
O
_l
<
_J«
3
sr
3
*T
-------
Gas Outlet
Eliminator Section
Iji^^^iiijII^^^^^^^^^I^J ^M^M
Sludge Outlet
Figure 3-85. Low-energy entrainment scrubber for use on smelt
dissolving tank vent (Ducon dynamic scrubber).
214
-------
Low-energy venturi scrubbers also can be used on tank vents. Common
design pressure drops are 6 to 8 in. HpO at liquor-to-gas ratios of 8 to 10
gal/1000 acfm. Removal efficiencies can exceed 99 percent.
3.3.2.3 Malfunctions--
Typical smelt dissolving tank malfunctions involve reduced efficiency of
the abatement system. Table 3-25 describes malfunctions associated with each
type of control device.
TABLE 3-25.
MALFUNCTIONS THAT MAY OCCUR IN SMELT TANK
PARTICULATE CONTROL SYSTEMS
Control system
Malfunction
Result
Mesh pad
Low-energy scrubber
Venturi scrubber
Low water flow
Maldistribution of back
flush water
Pad pluggage
High gas flow rates
Low water flow
Pluggage of packed bed or
nozzles
Pluggage or demister
Throat wear
Low water flow rate
Poor water distribution
Pluggage of pad
Pluggage of pad
Channeling and bypass
High pressure drop and
droplet reentrainment
Low efficiency
High pressure drop and/
or channeling
Liquor reentrainment
Low pressure drop
Low pressure drop
Low pressure drop
3.3.2.4 Inspection of the Smelt Dissolving Tank Area--
The smelt dissolving tank area is usually inspected in conjunction with
the recovery boiler area because of the proximity of the smelt dissolving
tank(s) to the recovery boiler and the accessibility to control instrumentation,
which is usually installed in the recovery boiler control room. The instru-
mentation is normally limited to process-related parameters such as the pro-
duction rate, composition (or density), and temperature of the green liquor,
but some instrumentation may be available for the control equipment.
215
-------
The control equipment used on smelt dissolving tank vents usually consist
of mesh-pad mist eliminators or low-energy venturi or cyclone scrubbers.
These particulate control devices are well suited to this operation because
a wet saturated gas stream is involved and the particles generated by the
smelt dissolving tanks tend to be relatively large and easy to remove with
low energy input. Scrubbing liquor is usually allowed to flow down the stack
walls directly to the smelt dissolving tank. The use of fiber-reinforced
plastic (FRP) components is common to avoid corrosion problems associated
with high moisture conditions. In general, the control equipment used on the
smelt dissolving tank vents is not very complex.
The vent gases contain water droplets, Na2S, and NaC03 particles generated
by the quenching of the molten smelt leaving the bottom of the recovery boiler.
The continuous flow of liquid smelt is not, however, allowed to be quenched '
en masse in the smelt dissolving tank. Rather, steam shattering nozzles are
typically applied to the smelt spout to form particles of smelt that are
easier to cool and dissolve because they have a larger surface area. As the
molten smelt is cooled and steam evolves under relatively turbulent con-
ditions, however, smaller particles of smelt are carried out of the dissolving
tank in the vent gases. These are the particles that the control equipment
is attempting to collect.
The three levels of inspection that have been defined for this area are
outlined below. Limitations of accessibility to measurement points will
often limit the inspector to a Level II Inspection with some of the elements
of a Level III Inspection. Much depends on the availability of instrumen-
tation and access to the control equipment.
Level I Inspection—Observing visible emissions is of limited value for
the smelt dissolving tank stacks. The three general limitations are 1)
uncombined water vapor produced by quenching of the smelt usually causes the
formation of a condensed steam plume, which forces observations to be made
beyond the point of steam plume dissipation; 2) the stack outlet is usually
near other outlets associated with the recovery boiler operation, which may
interfere with the ability to read the opacity'of the individual plume; and
3) it is difficult to select an observation site that will provide optimal
background and contrast conditions and establish the proper sun-source orien-
tation without interference from other sources.
216
-------
Observing visible emissions observation is useful, however, when large
changes in the opacity occur. Such changes may indicate a change in the
operation of the smelt dissolving tank or control equipment. For example, a
change in the positioning or pressure of the shattering nozzles may cause
a shift in the size of the smelt particles formed. If smaller particles are
formed, they may be more difficult to collect in the control equipment at a
fixed energy input level and greater penetration may occur. In addition, a
shift in particle size may enhance the light-scattering characteristics of
the particulate so that opacity would increase as a result of both finer
particulate and higher concentrations. A failure in the control equipment
might produce similar results. Thus, the visible emissions observation of
the Level I Inspection may be used as a screening tool to determine areas
that may warrant further investigation.
Level II Inspection—The Level II Inspection involves a combination of
observation of the opacity of the stack plume and a physical check of the air
pollution control equipment to confirm operation (although equipment operating
parameters are not measured). Operating data from available instrumentation
are obtained for both the process and the control equipment. Typical process-
related information that should be recorded include green liquor production
rate (tons/hour, gallons/minute, or gallons/hour), liquor composition or
density, liquor temperature, and smelt dissolving rate (from recovery boiler).
Typical control equipment parameters that may be monitored include pressure
drop, stack temperature, and the flow rate, pressure, and (in some cases)
temperature of the scrubber water. Because no measurements are made of these
parameters and the inspection includes only visual confirmation of component
operation (pumps, fans, etc.), the inspector must base his or her subjective
assessment of the accuracy of these data on this monitoring effort. If the
instruments providing the data are inadequate or malfunctioning, the result
may be an incorrect assessment of the control equipment compliance status.
To the extent possible, the values obtained during the inspection should
be compared with values from the design, baseline, or previous inspections to
determine if there has been a significant change in performance or in the
number of equipment malfunctions. With the low-energy equipment that is
being used, performance can change significantly with only a slight shift in
217
-------
pressure drop. For example, a change from a 6-in. to a 4-in. H-O pressure
drop in a low-energy venturi is significant, whereas with much higher operating
pressures this shift might not be considered significant. If particle size
is not very small, however, a shift in pressure drop may not significantly
affect emissions. It is usually the identification of the causal factors of
the parameter change that determines the magnitude of performance changes.
These causal factors are not always easily identified through a Level II
Inspection.
Level III Inspection—The detailed inspection procedures associated with
a Level III Inspection are slightly different for each control device used to
minimize the emissions from the smelt dissolving tank, but all of the control
devices have common elements. The characteristics of each type of control
equipment are discussed to aid the inspector in the diagnosis of operation
and/or maintenance problems.
The operating parameters that are common to all the pieces of control
equipment include smelt production rate, green liquor production rate, liquor
composition or density, and*liquor temperature. The control equipment param-
eters the inspector should obtain include the scrubber pressure drop, stack
temperature, scrubber water flow rate, scrubber water pressure, and, in some
cases, scrubber water temperature. These parameters, along with the obser-
vation of visible emissions, are the same components that makeup a Level II
Inspection. The difference between Level II and III Inspections is that in
the latter the inspector actually measures some of these parameters. In
addition, the inspection is keyed toward identifying operating problems that
characterize control equipment performance.
A most useful parameter for diagnosis of the control equipment operation
is the measurement of volumetric flow rate of the gas into the control equip-
ment. The most accurate method of obtaining this parameter is to select a
suitable location for performing a Pi tot traverse and determining volumetric
flow rate. The gas density will be somewhat less than air density because
the gas will contain a substantial fraction of water vapor at (typically)
saturated conditions. The volumetric flow rate will be related to the quantity
of air inle'akage around the smelt dissolving tank, the production rate of
smelt, and the smelt temperature leaving the furnace (affecting the steam
evolution rate). Using fan operating parameters to determine gas volume
218
-------
(where fans are used) is usually impractical because the density of air and
water vapor differs significantly and because entrained water droplets can
increase the energy requirements of the fan.
The value obtained by use of the Pi tot tube should be compared with the
design gas volume or a value obtained from a previous stack test. If pro-
duction rates are comparable, the volumetric flow rates should be similar
(within +15 percent) unless changes have occurred elsewhere in the system.
Another key parameter that should be measured is pressure drop. Although
this parameter is typically monitored by a magnehelic gauge, the taps should
be removed from the sampling locations and cleaned, and the inspector should
check the measurements with his or her own gauge. In some instances, the
pressure taps are connected to recording and controller instrumentation that
regulate water flow rates and alarm systems. Taking care not to interfere
with these systems, the inspector should measure pressure drop through sepa-
rate taps that do not interfere with active systems. Before removing taps,
the inspector should check with plant supervisory personnel concerning any~
union rules that might require union personnel to perform this task.
The principal operating problem to be detected in the mesh-pad control
and low-energy venturi scrubbers is one of improper distribution of scrubbing
water or the gas stream within the control device. These maldistribution
problems tend to allow a portion of the gas stream to bypass the zone where
particulate collection mechanisms are operating. In addition, maldistribu-
tion problems are often more crucial in the operation of this equipment
because such small energy inputs are typical. For,example, mesh-pad mist
eliminators tend to operate at around a pressure drop of 2 inches ^0, venturi
scrubbers operate with a pressure drop of 6 to 8 inches H20, and cyclone
wet-fan scrubbers operate with a pressure drop of between 4 and 8 inches H20.
These small energy inputs make the control equipment sensitive to maldistri-
bution problems, whereas larger pressure drops tend to reduce the sensitivity
to distribution problems. If the scrubbers are operating properly, however,
larger energy inputs are usually not necessary.
A change in pressure drop in these low-energy devices is normally signif-
icant and indicates that the performance of the control equipment has changed.
Fortunately, the particulate size is large by air pollution control standards,
which helps reduce the sensitivity to changes in pressure drop. The purpose
219
-------
of the Level III Inspection is to assess the causes of these changes and
determine their significance with regard to emission rates.
For mesh-pad eliminators, the usual indication of operating problems is
an increase in pressure drop, usually caused by an increase in the pad's
resistance. Most mesh-pad systems use continuous back-flushing sprays through -
the bed to minimize buildup of particulate, but some systems use a pressure
drop or timer-controlled back-flush system. The presence of the mesh pad and
the water flowing through it normally place some resistance on gas flow. If
the back-flush system fails or water distribution is not adequate, particulate
buildup will occur in the bed and cause the pressure drop to increase.
Failure of the water spray system will usually cause a relatively rapid
increase in pressure drop. A decrease in water flow rate to near zero flow
(because of line or nozzle pluggage, increased water pressure, or an inoperative
pump) also could be indicative of mesh pad pluggage resulting from failure of
,the water spray system. Insufficient water flushing volume or inadequate
coverage of the mesh pad may cause a partial pluggage to occur over a longer
time period. The existence of localized pluggage will cause an increase in
the local gas velocity through the bed, which will increase the pressure drop
and could increase the penetration and reentrairiment of droplets and particles
through the control equipment. When conditions permit and plant personnel
allow, the access door to the mesh pad should be opened to determine if the
spray nozzles are providing adequate coverage.
Low-energy venturi scrubbers also are very sensitive to improper water
distribution in the throat area. In a practical sense, these low-energy
scrubbers, which operate in the range of 6 to 10 inches pressure drop, are
more sensitive to slight changes in pressure drop and water distribution than
venturi scrubbers operating in the 18- to 45-inch range. The low energy
associated with these scrubbers means the available impaction energy is low,
and the poor distribution allows the impaction zone to be bypassed. These
scrubbers typically use spray nozzles to achieve the gas-liquid contact
required in the scrubber throat, as the low-energy impact may not be high
enough to achieve gas atomization of the scrubbing liquid and good throat
coverage. -The relatively large particle size from this emission source,
however, favors low energy input to the scrubber.
220
-------
When inspecting low-energy venturi scrubbers, the inspector should check
the gas flow rate, the scrubbing liquid flow rate, and the pressure drop. A
small change in pressure drop (i.e., a change of 1 inch H20 or more) should
be considered significant, particularly if it is a decrease. It is difficult
to discuss all of the possible malfunctions and their effect on performance.
An increase in the pressure drop, liquid-to-gas ratio, or throat velocity,
however, will generally improve performance, provided good liquid-to-gas
contact is established in the scrubber. There is no easy method by which an
inspector can establish whether good water distribution is being provided to
the scrubber during normal operation unless separate valves are provided for
each nozzle. If the scrubber is equipped with individual valving to allow
for individual nozzle inspection and cleanout, each nozzle could be turned
off, removed, inspected, cleaned (or replaced, if necessary) and returned to
service. Turning off a nozzle will produce only a slight change in water
flow rate and water pressure, but a change in pressure drop should be evident.
If any nozzle is turned off without having any effect on the scrubber per-
formance, that nozzle may be plugged or severely eroded and could be causing
a maldistribution of the scrubbing liquor.
Many plants use a cyclone spray scrubber. This scrubber, actually a
hybrid of .a cyclone spray scrubber and a wet fan collecting device, is also a
relatively low-energy device with typical pressure drops of 6 to 12 inches.
The device takes advantage of particle inertia to promote impaction of the
particle with the walls of the scrubber and the fan rather than using impac-
tion of the particle onto water droplets in the gas stream (although this
mechanism does play a role in the collection of the particulate). The water
applied in this scrubber is applied primarily as a particulate-transporting
medium once the particulate has reached a surface inside the scrubber. This
scrubber is much less susceptible to pluggage or problems with improper water
distribution than the scrubbers previously discussed.
The gas is usually introduced tangentially into the scrubber to remove
the large particles and liquid droplets. A spin vane system also may be used
at the inlet to begin the inertia! separation process. Water is introduced
into the scrubber to minimize pluggage and erosion of the spin vane as well
as to provide a transport medium for the large particles collected in this
221
-------
portion of the scrubber. The gas stream is then passed to a wetted radial-
blade centrifugal fan, where all but the smallest particles are collected.
Significant inertia! forces are applied to the particles and water droplets,
and their inertia tends to force them to the outer wall of the fan housing,
where the liquid is drained from the scrubber. Some particles become impacted
on the fan blades and are removed by the water applied to the fan wheel, and
some small droplets are created by the shattering force applied by the rotating
fan wheel. The remaining water droplets in the gas stream are then separated
by cyclonic action through a spin vane separator.
Most of the energy applied to the particles is supplied by the rotation
of the fan. The fan must provide the energy to overcome the energy losses of
the ductwork, the cyclonic separation of the particulate and water droplets,
and the acceleration of the water and particulate in the fan. The usual
procedure is to determine the pressure drop across the entire scrubber from
inlet to outlet. Using fan curves to calculate gas volumes through the
scrubber is not possible because of the direct application of water to the
fan blades. The use of Pitot traverses is usually not possible either because
of the presence of cyclonic flow after the scrubber. Thus, the only indicators
available to the inspector for assessment of performance are pressure drop,
water flow rate to the scrubber, and fan motor current.
Although less susceptible to pluggage than mesh-pad scrubbers, pluggage
that occurs in cyclone scrubbers will generally increase the pressure drop
across the scrubber because of an unbalancing and increased resistance to gas
flow through the scrubber. A decrease in liquid to the fan may not have much
effect on the pressure drop across the scrubber, but such a decrease tends to
be indicated by a decrease in the fan motor horsepower because of the apparent
decreased "gas" density on the fan. Thus, while operating parameters are
limited, those that are available may indicate scrubber performance changes.
As mentioned previously, the low energy input of the scrubber limits the
ability of the scrubber to collect fine particulate; however, the particulate
generated is generally large. Changes to smaller particle size because of
changes in smelt characteristics or positioning,of steam smelt shattering
nozzles may" cause problems in particulate collection efficiency, and when
combined with other malfunctions, can cause performance to deteriorate sig-
nificantly.
222
-------
3.4 CAUSTICIZIN6 DEPARTMENT
The two major emission sources in the causticizing department are the
slakers and the lime kilns. The major source of particulate emissions in the
causticizing department is the lime kiln which is typically controlled by a
venturi scrubber. The malfunctions associated with lime kilns can generally
be divided into those that are process-related and those that are related to
scrubber performance and operation. Section 3.4.4 provides a detailed dis-
cussion of lime kiln and venturi scrubber malfunctions. Section 3.4.5 pro-
vides a discussion of the inspection procedures for lime kilns and slakers.
In general, Level III inspections are usually limited by production schedules,
availability of instrumentation, and accessability to process and control
equipment.
3.4.1 Process Description
The economical operation of the alkaline pulping process requires the
recovery and reuse of the sodium and sulfur compounds used in the c'ooking
process. The liquor produced by the smelt dissolving tanks (green liquor)
contains a mixture of sodium carbonate and sodium sulfide. Because sodium
carbonate cannot be used in the cooking process, it must be converted to a
usable form, NaOH.
The causticizing process involves the conversion of Na2C03 to NaOH
through the use of calcium oxide (CaO) (Figure 3-86).2 Causticizing occurs
in two steps: 1) reaction of the calcium oxide with water and 2) reaction of
the calcium hydroxide with the sodium carbonate.
The first reaction', which is referred to as slaking, liberates a con-
siderable amount of heat. The reaction is written as follows:
' CaO + H20 ->• Ca(OH)2 + 486 Btu/lb
The second reaction, which is referred to as causticizing, yields sodium
hydroxide and calcium carbonate. This reaction is written as follows:
Ca(OH)2 + Na2C03 * 2NaOH + CaC03
In normal practice both these reactions occur simultaneously in the slaker
tank. Because the second reaction is reversible, all of the sodium carbonate
cannot be converted to sodium hydroxide. The completeness of the reaction is
defined as the causticizing efficiency.
223
-------
GREEN LIOOOR
TTJ^
o OC
Figure 3-86. Typical causticizing flow diagram.
224
-------
The efficiency may be determined by the following equation:
Efficiency (%) =
NaOH
NaOH
Na2C03
x 100
The efficiency is a function of the concentration of sodium sulfide in the
green liquor. Higher concentrations of sodium sulfide reduce conversion
efficiency. Unreacted compounds represent an unreacted circulating load in
the liquor system. Sodium sulfide in the liquor hydrolyzes to form sodium
85 *
sulfide and sodium hydrosulfide.
For recovery of the calcium used in the slaking process, calcium carbonate
is separated from the"white liquor, dewatered, and disassociated under intense
heat. The reaction is referred to as lime burning or calcining and proceeds
according to the following reaction:
CaCOo •* CaO + COp
The causticizing process is composed of many subprocesses, including
green liquor preparation, white liquor preparation, and lime mud handling.
3.4.1.1 Green Liquor Preparation--
Green liquor produced by the smelt tank contains insoluble impurities
known as dregs, which must be removed. Dregs, which are composed of carbonaceous
2
material, silica, and metallic sulfides, are removed in the green liquor
clarifier through settling and decanting. The solution is washed in the
dregs washer, where the dregs are removed. The wash water is then pumped to
oc
the mud washer. .
3.4.1.2 White Liquor Preparation--
Clarified green liquor is pumped to the green liquor storage tank for
introduction into the slaker. Calcium oxide is introduced into the slaker
(reburn plus pebble lime) and reacts to form sodium hydroxide and calcium
carbonate according to the previously stated reactions. The slaker contains
a clarifying section in which unreacted material (grit) is removed by a
87
mechanical rake (Figure 3-87).
Lime feed to the slaker is generally no more than 85 percent CaO. The
reaction is carried to equilibrium in the causticizers (up to three may be
used). White liquor containing lime mud is pumped from the causticizers to
225
-------
VtNT STACK OPCNIMC"
.FLANGED HPt CONNECTION FOR
OVCRfLOW CONNECTION TO
CITHCR3IDC OF MACHINE
OORRCOMIXCR
RING SEAL-
awffis"
Bath*
.CONNCCTMC HOC
FMNT KAKC MANCCR
FRONT LINK
•GKIT OIKHAKCC
-
"- -LI
SECTIONAL CievATION
Figure 3-87. Slaker-classifier used in typical
causticizing pi ant.87
226
-------
the white liquor clarifier, where the mud and liquor are separated. The
clarified liquor is then pumped to the digesters.
3.4.1.3 Lime Mud Washing--
Because the lime mud slurry contains sodium, which can adversely affect
the calcination process, it is washed in dilution washers to reduce the
sodium and sulfide content. Waste wash liquor is returned to the smelt
dissolving tank as weak liquor. Lime mud washers may be decanting, belt-
filter, precoat-filter, or pressure-filter washers. The lime mud must be
dewatered before being introduced to the lime kiln. Centrifuge or vacuum
filter dewatering systems are typical.
Final soda content is between 0.1 and 2.5 percent by weight (as Na?0),
88
and dewatered mud is between 60 and 70 percent solids.
3.4.1.4 Calcining—
Calciriing takes place on fluid-bed calciners or in rotary, direct-
fired kilns. Rotary kilns used in kraft mills are 8 to 13 feet in diameter
and 100 to 400 feet in length. The kilns slope downward from the feed end to
the discharge end at an angle of about 3/4 in. per foot of length. The feed
mud is moved through the kiln by slow rotation, usually between 0.60 and 1.33
rpm. The rate of material movement down the kiln is determined by the angle
of the lime mud respose, kiln length, kiln slope, rotation speed, and kiln
diameter.
Processes carried out in the kiln are water evaporation, mud heating,
and calcination. Feed end temperatures are between 300° and 500°F and dis-
charge end temperatures are between 2000° and 2400°F.
Energy for calcining is supplied by direct firing countercurrent to the
material flow. Fuels are normally residual and distillate oils, natural gas,
and coal, but waste oils, turpentine, and tall oil soaps also may be used.
Because of the presence of inert material in the lime mud, product yield
(CaO) will not equal theoretical values. A typical mud may contain up to 15
percent inert material. Normal yield, based on inlet feed weights, may be on
the order of 45 to 50 percent. In well-operated systems, fuel rates are in
the range of 7 to 7.5 x 106 Btu/ton of product.
227
-------
3.4.2 Emission Sources
Particulate emissions from the kiln are the result of particle suspen-
sion and entrainment due to the flow of combustion gases and carbon dioxide.
Fine particulate is generated by the vaporization and condensation of sodium
based compounds. The amount of fine particulate generated is related to the
soda content of the feed.
Secondary and primary air for fuel combustion are provided at the burner
end of the kiln. The burner-end hood is held at a slight negative pressure
(0.1 in. HpO). The air supply rate is controlled to maintain an exit gas
oxygen content of 1.5 to 3.0 percent (7 to 15 percent excess air). Feed-
end drafts can be significantly higher, depending on kiln diameter and length
and the weight of chains used.
Table 3-26 presents a mass balance of a typical 260 tons/day kiln. The
dust in the mud feed is recycled lime dust from primary dust collectors at
~i
the kiln exit.
TABLE 3-26. TYPICAL LIME KILN MASS BALANCE89
Lime mud feed, tons/day
CaC03
Dust
Inerts
Water
Product rate, tons/day
CaO
Inerts
Emission rate, tons/day
CaC03
CaO
Heat input, 106 Btu
Heat input/ton product, 10 Btu
Kiln size, ft
3
Kiln volume, ft
Kiln gas volume, acfm
Kiln gas temperature, °F
Retention time, min
417.857
28.331
26.000
101.366
234.000
26.000
28.331
0.000
85.069
7.8525
12 x 275
24,971
68,000
450°
129.430
228
-------
Lime mud crystals that precipitate from the white liquor in the causti-
cizers have Na2S that cannot be removed during mud washing. The removal rate
depends on the sulfidity of the wash water and the ratio of soluble to in-
soluble sulfide in the mud. As a result, Na2$ is carried into the kiln, where
it is exposed to an atmosphere containing approximately 18 percent C02. The
C02, adsorbed on the mud surface, lowers the mud pH, which allows the sulfide
to be converted to H2S. The rate of H2$ that is released is a function of the
Na2S content of the mud, mud moisture, 02 content of the gas stream, and the
flue gas temperature. Figures 3-88, 3-89, and 3-90 show the effect of these
variables on H2$ emission levels.
3.4.3 Control
88
The control of particulate emissions from the lime kiln is provided
almost exclusively by venturi scrubbers. Although impingement plate scrubbers
and ESP's have been used in some locations, the venturi scrubber offers
significant economic advantages. First, venturi scrubbers are more compact
in size than other scrubbers of comparable efficiency, which means less
capital cost and space requirements. Second, plugging is less of a problem
than it is with other scrubbers. Although plugging and erosion of lines and
nozzles are possible in a venturi scrubber, they can tolerate much higher
levels of solids in the scrubbing liquor (10 to 20%) than that which can be •
tolerated by an impingement plate scrubber (1 to 3%). This usually means
less clarifying, treatment, and makeup water requirements for the venturi
scrubber. The single largest disadvantage of the venturi scrubber is the
energy requirements to maintain a high, pressure drop. Pressure drops between .
20 to 35 inches hUO, which require large fans and motors, are not uncommon
for venturi scrubbers applied to lime kilns.
Most venturi scrubbers are constructed out of 316 and 316L stainless
steel to prevent corrosion. Most designs also have a "flooded elbow" variable
throat that allows for additional pressure drop control. The flooded elbow
at the exit of the venturi scrubber divergent section provides a trough of
water to capture the larger water droplets and particulate that exit the
scrubber and prevents abrasion of the elbow at the turn. The throat of the
venturi may be circular, rectangular, or oval, and the throat area can be
229
-------
f "
o
SI
MOISTURE - 0.4 g/i •( *y CaCO.
TEMPERATURE- SOO*F
OXYGEN- 2,3X
TOO »OO
.S IN LIME MUD
?b of dry CoCO,
Figure 3-88.
H2$ emission from the lime kiln related to Na?S level
in the lime mud.88 '
en
#
•M.S -1001*4 ,< +,
MOSTU*E-0.4e/l| of »j OCO,
Tf MR.- SOO V
OXYGEN IN FLUE GAS — %
Figure 3-89. H2S emission related to percent 02 in the flue gas.
88
230
-------
100-
90-
80-
70 -
60
so!
N«,S-ioom/o of ft c«cc,
TEMP. - 800 V
orroEN- 1.6%
ta a*
MOISTURE • LIME MUD
Orom/grom of *y CoCO,
Figure 3-90. H2$ emission related to the moisture content in the lime mud.
88
controlled by hinged plates in the throat or by a "bob" that moves up and
down in the effective throat area. Pressure drop is usually controlled by
controlling the throat area (hence, throat velocity), not by controlling the
water flow rate to the scrubber.
Water is usually introduced to the scrubber through an overflow Weir
system above the throat to avoid the use of nozzles, which may become plugged
or eroded as a result of the solids in the scrubbing water. Throat coverage
by the water is somewhat low because the water flows down the walls of the
scrubber to the throat. This problem is more than offset, however, by the
avoidance of nozzle problems. Some designs do use nozzles to introduce water
droplets into the throat, but adequate maintenance capabilities must be
provided for cleanout and inspection of the nozzles while the scrubber is
operational.
Particle impaction on the water droplet is the dominant collection mech-
anism in the venturi scrubber. The high-velocity gas entering the throat
(gas atomization) shatters the water introduced into the scrubber into "fine"
droplets with a large effective surface area. These water droplets provide
impact targets for the particles that cannot follow the gas streamlines
around the droplets, and the particles collide with the water droplets.
231
-------
Particles with diameters of 25 to 75 ym are much easier to collect than
particles in the 5- to 10-ym range. Compared with the gas stream, the water
droplets have essentially no velocity when introduced into the scrubber
throat. Impaction is the major collection mechanism at the point of intro-
duction into the scrubber throat. The velocity of water droplets, however,
accelerates through the throat of the scrubber, and the effectiveness of the
impaction mechanism decreases as the water velocity approaches the gas
velocity. Smaller water droplets, higher throat velocities, higher liquid-
to-gas ratios, and the resulting higher pressure drops usually mean higher
collection efficiencies.
Diffusion is another collection mechanism in the venturi scrubber. This
mechanism applies only to very fine particles, generally less than 0.05 ym,
where Brownian motion influences the particle motion and particles tend not
to follow gas streamlines. This mechanism is usually operative in the
divergent section of the venturi where water droplets in the gas stream are
travelling at nearly the same velocity. •
Unfortunately, neither impaction nor diffusion is very effective in a
certain range of particle sizes (0.1 to 1 ym). In this range, the particles
are small enough to follow gas streamlines around the droplets, but are too
large to allow diffusion'to be an effective collection mechanism. If a
significant portion of the particles entering the scrubber falls in this
range, very high pressure drops may be required for adequate control of
particulate emissions. Particles in this size range are the ones that most
effectively scatter light and cause an observable opacity. Lime mud with a
high soda content (as a result of inefficient washing) can produce a large
quantity of particles in this size range. Increased pressure drop or process
modification may be the only solution to this problem.
In many mills, emissions from the lime slaker are often ignored and no
controls are provided. The emissions of lime dust and steam from the lime
sleker and the causticizing of green liquor to white liquor can be controlled
with low-energy scrubbers. When slaker emissions are controlled, the usual
equipment consists of cyclone scrubbers or dynamic scrubbers using cyclone
scrubbers with a wet fan arrangement similar to that used for smelt dissolving'
tank vents.
232
-------
Scrubbers in this application usually need only 4 to 6 in. of pressure
drop to be effective. .The dominant collection mechanism is inertial separa-
tion; i.e., inertial forces on the particle cause it to separate from the gas
stream, either in the cyclone scrubber or in the wet fan arrangement. Scrub-
bing liquid may be water, weak white liquor, or clarified green liquor. The
scrubbing liquor is usually returned directly to the lime slaker for use in
the lime recycle loop.
Scrubbers used on causticizers and lime slakers usually must be hooded or
placed in covered enclosures to ensure adequate capture efficiencies and to
minimize ventilation requirements. Scrubber components are normally stain-.
less steel to minimize corrosion problems.
3.4.4 Malfunctions
The malfunctions and problems in the lime kiln area can generally be
divided into those that are process-related (i.e., kiln operation) and those
that are related to scrubber performance and operation. The process-related
problems generally will influence the 'scrubber performance by affecting both
the temperature and the particle size entering the scrubber. These problems
may be the result of equipment failures, but they are most often caused by a
process change.
The most common problem in the lime kiln process is inadequate washing
of the lime mud before its introduction into the kiln. As discussed in Sec-
tion 3.4.1, an interaction occurs as a result of material production rates,
design parameters, yield efficiencies, and quantities of inerts in the
caustizing area. From an air pollution control standpoint, problems begin
when lime kiln mud from the causticizing of green liquor to white liquor is
not washed effectively before it is introduced into the kiln. Residual
sodium compounds, particularly sodium sulfide (Na2S) and sodium hydroxide,
remain in the lime mud after causticizing, and the mud must be washed and
filtered to remove as many of these constituents as possible. Because it is
impossible to remove all of these compounds, the lime kiln can become a
significant generator of fine particles and H2S if process conditions are
right.
The sodium compounds are easily volatilized in the high-temperature
flame end of the kiln, and they tend to leave the kiln as an uncondensed
233
-------
alkali material. Unless the temperature of the gas stream is reduced before
it reaches the inlet of the scrubber, most of the alkali will remain in the
gas phase until it is quenched in the scrubber throat. Unfortunately, alkalies
tend to form in the 0.1- to 1-ym range, which makes them difficult to collect
in a venturi scrubber. One possible solution is to quench the gas stream by
use of a presaturator. This allows the particles to grow before they enter
the inlet of the scrubber. A-better solution to this problem, however, may
be to change the washing characteristics to reduce the sodium carryover in
the lime mud.
The generation of fine particulate is a symptom of sodium-based compound
carryover, but the principal compound in the lime mud that is related to
potential H2S emissions is Na2S. (See Figure 3-88.) The conditions of high
C02 levels and high temperatures can promote the breakdown of Na2S to form
H2S.88 The concentration of Na2S in the lime mud is not the only variable
that affects the generation of H2$, however. The presence of oxidizing
conditions (02) and the temperature also affect H2S emission rates. Low
temperatures or low 02 contents enhance the formation of H2S, whereas high v
temperatures and higher 02 levels cause equilibrium shifts to form S02.
Thus, short kilns with higher operating temperatures have less potential to
emit H2S than do longer kilns at the same production rate and lower tempera-
tures. Kilns operating with oxygen enrichment to increase production rates
may also produce less H2S because of high .operating temperatures in the kiln.
If the 02 at the exit end of the kiln is 0. to 5 percent, sufficient excess
02 must be provided to avoid H2$ generation if residual oil is burned.
In addition to inefficient or insufficient lime mud washing, potential
contamination can come from the washing liquor itself. If the incoming
washing liquor is contaminated with sodium compounds, particularly Na2$ from
other portions of the mill, then washing will be unable to reduce the content
below the level of the incoming wash solution. The eventual solution to this
problem may require repiping the supply lines to avoid the contaminating
source(s).
Scrubber operating problems typically involve failure to maintain pres-
sure drop, difficulty in supplying scrubber liquor to the scrubber, or a
buildup of solids in the liquor. Because some of these problems are inter-
related, identifying a specific problem can be difficult.
234
-------
Most scrubber controllers use pressure drop as an indicator of perform-
ance. Failure to maintain the scrubber pressure drop at a preset point can
cause operating problems at either extreme. High pressure drops can cause a
fan limiting condition that prohibits the fan from moving the required quantity
of gas. This results in fugitive emissions from the kiln, which operates
under a slight negative pressure. Other failures result from the inability
to maintain adequate pressure drop through the scrubber. This condition may
be caused by failure of the pressure drop controller system, a water distri-
bution problem, or the abrasion and erosion of scrubber internal components,
all of which usually decrease scrubber performance. The latter two problems
can usually only be identified by an internal inspection, whereas a failure
in the controller system can normally be identified by a lack of response to
a change in pressure drop set point.
The easiest method for controlling pressure drop in a variable-throat
venturi scrubber is by controlling the throat area and the effective throat
velocity. Although control of the water flow rate is an alternative means
of controlling pressure drop (usually the only alternative in fixed-throat
venturi scrubbers), it is not often used because it is more difficult and
less sensitive to error adjustment. Controlling both the throat velocity and
the water flow rate is generally not successful because of the interactive
relationship of the two parameters on pressure drop. Attempting to control
both parameters typically results in unacceptable or unstable oscillations in
pressure drop or extremely slow response times.
Water distribution is usually not a problem in these scrubbers because
nozzles are seldom used. When nozzles are used, however, provisions must be
made to isolate and extract each nozzle for inspection, cleaning, and replace-
ment during normal operation. Although inspection of each nozzle may cause a
temporary maldistribution of liquid in the scrubber, it is better than a
longterm failure. The maldistribution of liquid over a period of time can
cause excessive abrasion of the throat components resulting from localized
increases in velocity with abrasive particulate. Inadequate covered throat
area will allow the gas to "channel" to the area of least resistance.
The solids levels in the recycled scrubber liquid can cause erosion of
pump impellers and pipes. Wear of pump impellers can decrease the volume and
velocity of scrubbing liquid to the scrubber, which can decrease scrubber
235
-------
performance and cause pluggage of liquid supply lines. High solids levels
can also cause nozzle pluggage in addition to nozzle erosion.
High solids levels also can cause problems with resuspension of particu-
late in the gas stream where the hot gas stream contacts the liquid at the
throat. The evaporation of the liquid allows both the suspended and dissolved
solids to be resuspended in the gas stream. The suspended solids level can
be controlled by using a bleed-off stream with fresh water makeup to prevent
high concentration levels. The problem also can be helped somewhat by quench-
ing the gas stream with fresh water. Use of water with high dissolved solids
content can allow the formation of a submicron fume that is difficult to
collect at normal pressure drops. A bleed-off stream and fresh water makeup
can also help to reduce problems with high suspended solids. A material
balance typically shows that a site specific equilibrium level can be established
at which particulate will begin to become resuspended in the gas stream. It
is desirable to maintain solids levels below those levels that increase resus-
pension.
Malfunctions in the lime slaking area are generally limited to inadequate
distribution of the scrubbing liquid in the scrubber. Because these are
low-energy devices, the malfunctions tend to be related to scrubber pump
failure or nozzle pluggage. The other failure mechanism is generally related
to the inability to capture the emissions because hatches, covers, or hoods
are not properly aligned.
3.4.5 Inspection Procedures
Control of the lime kiln particulate emissions is usually accomplished
through the use of a venturi scrubber. Although other types of scrubbers are
sometimes used, most of the newer lime kilns have venturi scrubbers operating
in the range of 20 to 35 inches pressure drop. For this reason, the discussion
of the inspection techniques in the lime kiln area is limited to venturi
scrubbers.
As discussed previously, the lime kiln provides the second chemical
recycle loop in the kraft pulping process, which allows green liquor to be
converted to white liquor for use in digestion of wood chips. The particulate
generated by the lime kiln is primarily calcium carbonate and calcium oxide.
Energy requirements for the scrubber depend on the size distribution of the
236
-------
particles leaving the kiln and entering the scrubber. Effective collection
of smaller particles requires higher pressure drops. Scrubbers are usually
fabricated of 316 stainless steel and are designed with a flooded elbow to
minimize wear at the exit of the scrubber.
The three levels of inspection for this control device are defined.
Ability to conduct a complete Level III Inspection is usually limited by
production schedules, availability of instrumentation, and accessability to
equipment.
Level I Inspection—The Level I Inspection of the lime kiln area has the
same limitations as those for other wet plumes: 1) uncombined water vapor
caused by evaporation in the scrubber usually causes the formation of a con-
densed steam plume, forcing observations to be conducted beyond the point of
steam plume dissipation; 2) the selection of an observation point may not
provide good or optimal background and contrast conditions for the establish-
ment of the proper sun-source orientation; and 3) when several kilns and
scrubbers are located close together, the plumes may interfere with each
other. In some cases, the exhaust from several scrubbers may be combined in
a single stack.
All scrubbers will not exhibit an attached plume at the stack exit, even
though the gas stream is at saturated conditions. The energy input by the
fan in high-energy scrubbers may provide just enough energy to revaporize the
condensed water droplets and cause an unattached steam plume to form that may
facilitate opacity readings at the stack exit.
A change in opacity from the baseline condition or a previously established
operating condition indicates that some parameter or group of operating
parameters has changed, which warrants a more detailed inspection of the
equipment. Therefore, a change in opacity levels may be used to establish
the need for conducting more detailed inspections. Without further operating
data, however, the cause of the opacity change is difficult to assess with a
Level I Inspection.
Level II Inspections—The Level II Inspection entails a combination of
visible emissions observation and a physical check of the air pollution
control equipment by use of plant instrumentation to confirm its operation.
Usually the inspector does not take measurements in a Level II Inspection.
237
-------
If instrumentation is properly located and operated, however, the data pro-
vided can be useful in evaluating the performance of-the venturi scrubber.
Operating data from both the process equiment and the control equipment
should be obtained during the inspection. Typical data that should be col-
lected include kiln production rate (as CaO), lime mud feed rate, fuel firing
rate, kiln exit temperature, kiln rotation rate, scrubber inlet and outlet
temperatures, scrubber water flow rates, and pressure drop across the scrubber
throat.
The quantity of flue gas produced by the lime kiln is related to the kiln
production rate, the conversion efficiency of CaC03 to CaO, the quantity of fuel
fired in the kiln, and the quantity of excess air allowed into the kiln during
combustion. The oxygen levels are often monitored at the kiln exit to assist
in the firing of kiln. In addition, if the kiln is undersized with respect to
the rest of the plant capacity and the kiln becomes the limiting equipment in
the mill production rate, oxygen enrichment of the combustion zone may sometimes
be used to increase the heat release rates in the kiln, which increases the
production rate. Because all these parameters vary on a daily basis, the
variable-throat venturi scrubber is normally used to control particulate
emissions. Control of both water flow rates and the throat area usually
results in an unstable controlling operation because of the effect the reaction
and interconnection of the two operating parameters has on pressure drop. If
a constant pressure drop is maintained across the scrubber regardless of
production rate, the control equipment will maintain its collection efficiency
for smaller particles at both normal and reduced production rates. At reduced
production rates, the uncontrolled particulate concentration also may be
lower because of reduced .gas velocities through the kiln. The inspector
should be concerned with any significant change in scrubber pressure drop
and/or any large change in scrubber water flow rates.
The inspector should check the operation of the scrubber pumps, any
water pressure indicators, and the water flow into any settling pond or tank.
Because not all sources will be equipped with water flow rate indicators,
this check will ensure that' water is flowing into scrubber; however, it will
not ensure that water is being evenly distributed in the scrubber throat or
that the flow rates are correct.
238
-------
A significant change in both the visible emissions and the values for
pressure drop or water flow rate usually indicates a change in the mechanical
operation of the scrubber. Examples of mechanical changes might include a
change in controller set point for pressure drop, a change in liquid flow
rate due to a change in valve settings, pump wear, plugged nozzles, or wear
and abrasion of the scrubber throat. Changes in opacity without changes in
pressure drop or water flow to the scrubber, however, may be related to a
change in process operation or to chemical balances in the scrubber liquid.
This condition would be caused by changes in particle size distribution
rather than by a mechanical change in energy input to the scrubber. The
additional measurements required to evaluate.the nature of these changes
would constitute a Level III Inspection.
Level III Inspection—In addition to the aspects of the Level II Inspec-
tion, a Level III Inspection of the lime kiln involves supplemental measure-
ments to assure the collection of correct scrubber data. The values obtained
during the inspection are compared with those of the design and/or baseline
scrubber parameters to assess scrubber performance. Some measurements are
difficult to make, however, because of the design aspects of the scrubber.
As a result, the inspector must decide which values should be obtained on the
basis of the type of evaluation to be performed.
Two important parameters (pressure drop across the scrubber and visible
emissions) have already been discussed. A change in these parameters may
indicate the need for further investigation. Although plant instrumentation
should provide a reading of the pressure drop, the inspector should also
measure the pressure drop to confirm the accuracy of this reading. Measure-
ment of the pressure drop across the scrubber should be taken just before the
converging section and just after the diverging section of the venturi scrubber.
Because a measurement after the cyclone separator will include static pressure
loss associated with mist elimination in the gas stream, the value would not
represent the energy expended in the throat of the venturi scrubber. If the
design of the scrubber includes a flooded elbow, measurement of the pressure
after the throat may not be possible. Measurements taken just beyond the
elbow will provide reasonably good data if the inspector allows for about a
1-in. loss associated with the turning of the gas stream at the elbow.
239
-------
Because the particulate collection efficiency improves with higher
throat velocities (particularly at smaller particle diameters), the gas
volume through the scrubber is a useful parameter. Although it is difficult
to know what the effective throat area is during operation, a good estimate
of the range may be calculated for most scrubbers. Unfortunately, it is
difficult to determine gas volume through a venturi scrubber. The use of an
F-factor method is not very successful because of the absorption of C02 into
the scrubber liquor. Many stacks have tangential inlets that introduce
cyclonic flow in the stack, but installing a fan after a cyclone separator
usually destroys any cyclonic flow present after the separator. Because many
designers use the stack as a final mist eliminator, however, the gas is
introduced tangentially. Unless flow straightening devices are used, the
cyclonic flow will render velocity traverses by a Pitot tube inaccurate. The
measured velocity will usually have a positive bias unless the cyclonic flow
is severe, in which case the bias will be negative. Thus, a Pitot traverse
may not be possible.
The alternative, the use of fan curves, also may not be acceptable for
several reasons. First, the gas stream is usually saturated with water
vapor, which decreases gas density. Second, the presence of C02 and water
droplets increase the effective density of the gas stream. If it is assumed
that no water droplets are entrained in the gas and that the decrease in gas
density because of water vapor is offset by the presence of C02, the gap must
be treated as if it were air at the scrubber outlet temperature. This
approach will provide an estimate of gas volume through the scrubber only
if no water is injected into the fan.
Another reason fan curves may not provide acceptable results is because
of cyclonic flow from the cyclone separator at the inlet of the fan. Fan
curves and tables are based on smooth transitions and no cyclonic flow into
the fan inlet. The presence of cyclonic flow at the fan inlet in the direction
of fan rotation will change both the static pressure/volume relationships and
the horsepower/gas volume relationships much the same as a reduction in fan
speed would. This change occurs because the difference in angular velocities
between the fan rotating speed and the gas rotation speed is reduced. If the
gas is travelling cyclonically in the same direction as the fan rotation, less
energy is required to accelerate the gas to the rotating discharge velocity
240
-------
of the fan. The inspector cannot ascertain the effect on fan performance
without making in-depth tests that are beyond the scope of this guide and
outside the normal inspection procedure.
If a Pi tot traverse can be performed before the gas enters the scrub-
ber, an estimate of the gas through the scrubber can be calculated by
assuming that the gas is adiabatically cooled to saturated conditions. The
procedures for this calculation may be found in any engineering handbook.
Basically, the calculation includes the decrease in actual gas volume due
to cooling and the increase in gas volume due to evaporation of liquid
water to water vapor. If the volume through the scrubber is assumed to be
the inlet value at the high temperature, the predicted throat velocity
would be too high. A simplifying assumption is that the gas stream is
cooled and saturated immediately upon entering the scrubber throat and the
resulting gas volume is that which passes through the throat area to yield
an average throat velocity. If this value is calculated, it should be
compared with design values (if available). The nominal values for throat
velocity are between 5,000 and 15,000 cm/s, with values between 10,000 and
15,000 cm/s most common for high-energy venturi scrubbers.
Another useful parameter in the determination of venturi scrubber per-
formance is the liquid-to-gas ratio. To find this value, the inspector
must have both the volumetric flow rate of the scrubbing liquor through the
scrubber and the flue gas volume. He or she must rely on plant instrumen-
tation to determine the liquid flow rate', as no inexpensive portable in-
strumentation is available. Plant personnel should periodically check the
calibration of this instrumentation to increase the overall confidence in
the data. (The usual method is to calculate the ratio between the liquid
volume and the gas volume.) Typical values for the lime kiln scrubber are
10 to 20 gallons/1000 acfm of gas. Higher values usually mean better per-
*
formance and higher pressure drop. Physical measurement of the' liquid
stream is not possible because of the large quantities involved (between
600 and 1500 gallons per minute).
As with other scrubbers, good liquid distribution across the throat is
necessary to provide adequate scrubber performance. Improper coverage be-
cause of nozzle pluggage will allow a portion of the gas stream to bypass the
gas-liquid contact zone and result in a lower pressure drop and lower
241
-------
collection efficiency. Water flow may remain essentially constant, but the
maldistribution will probably prevent the scrubber from reaching its normal
or maximum pressure drop. Most larger scrubbers are not equipped with spray
nozzles so as to avoid this problem. Also, the use of internal overflow
weirs allow the increasing velocity at the throat to atomize the liquid to
droplets.
Even when pressure drop and water flow rate are at the normally accepted
levels, there are two items that can cause an opacity problem: total sus-
pended solids and total dissolved solids in the scrubbing liquor. Because
the kiln is a high-temperature source (exit temperatures of between 400° and
650°F), the resuspension problems associated with high solids levels may be
magnified by this process.
After scrubbing, the water contains captured particulate that must be
processing further. Because it is ususally not economical or permissible
to use the water on a once-through basis, the water is recirculated through
the scrubbing system. One of the advantages of a venturi scrubber is that
it will tolerate much higher levels of suspended solids than other scrubbers
(e.g., an impingement scrubber) because it is less susceptible to pluggage.
Adequate retention time, however, must be provided in the settling or
clarifying tanks for particulate settling and for maintaining the suspended
solids at an acceptable level (15 to 30% by weight). If excessive levels
of suspended solids are allowed to build up, wear of scrubber components such
as pipes, pumps, and nozzles (if any) will increase. In addition, the particle
may come out of liquid suspension during vaporization of the liquid in the
scrubber throat and form a particulate that can pass through the scrubber to
the stack. An equilibrium is eventually established, as the particulate must
exit the scrubber either through the clarifying step or through the stack.
Some of the solids dissolve and go into solution with the scrubbing
liquid.* The quantity that goes into solution is limited by the solubility -
of the various compounds. The actual level of dissolved solids may be
governed by the solubility limit, but a equilibrium level is usually estab-
lished by the addition of makeup water to account for water losses in the
scrubber. When the dissolved solids reach a certain level, the water vapori-
zation in the scrubber throat causes the formation of a very fine particle
that produces a visible opacity. If a change in opacity is noted, water samples
242
-------
should be obtained to determine if the suspended and dissolved solids levels
also have changed. Although the dissolved solids level is site-specific,
values of 3 to 5 percent or less by weight should prevent the generation of
fine aerosols.
Part of the problem with solids involves evaporation of the scrubber
liquor and the suspension of particles in the gas stream. Although the
evaporation cannot be prevented, proper clarifier retention time, adequate
water bleed-off and makeup rates, and possible scrubber changes can signifi-
cantly reduce the opacity problems .that may be associated with the solids
levels.
If all the parameters compare favorably with the previously established
values (design or baseline test conditions), the performance of the scrubber
is likely to be similar. Changes in parameters should help identify the prob-
lem or indicate that performance has improved. The key parameter is the scrub-
ber pressure drop and its relationship to the other operating parameters.
3.5 POWER BOILERS
In most mills, the kraft recovery boiler could produce a significant
portion of the mill's energy requirements for process steam and electricity.
Because its primary responsibility is chemical recovery, however, the
recovery boiler does not have a very high thermal efficiency. Therefore,
power boilers are used to supplement the steam and electricity requirements.
Based on a number of economic and regulatory considerations, the number of
boilers and the type of fuel burned varies from mill to mill. A given mill
may have anywhere from one to eight power boilers burning wood, oil, coal,
natural gas, or a combination of fuels. This subsection discusses their
potential operating problems along with the major sources of emissions, avail-
able control techniques, potential malfunctions, and inspection techniques
for power boilers and associated control equipment.
The major emissions from power boilers are particulate matter, S02, NOX,
and C02. The quantity of each pollutant produced is a function of the fuel
characteristics, the firing method, and the combustion characteristics for
each boiler. In general, only coal- and bark-fired boilers have particulate
matter control devices. Although there may be numerous malfunctions that can
243
-------
occur in a power boiler, the two most common operational problems are fuel
quality and establishment of proper excess air levels.
As with the inspection of the recovery boiler, the inspection of the power
boiler involves the identification and evaluation of those operating param-
eters or variables that indicate operation outside the norm for both the boiler
and the control device. Section 3.5.5 discusses in detail the major points
to be covered in inspecting both the boiler and the various types of particu-
late matter control devices. Table 3-27 on page 277 identifies the particular
parameters and variables that should be recorded during a Level III Inspection
of the power boiler.
3.5.1 Process Description
The power boiler is designed for the efficient production of high-
pressure steam for conversion into electricity through a turbine/generator;
the steam exhausted from the turbine is used for process operation (process
steam heating, operating turbine driven equipment, etc.). Because steam pro-
duction is the only responsibility of the boiler, it can be designed to
maximize heat transfer and to minimize heat losses through the stack or other
sources.
Regardless of the fuel, the basic combustion and heat transfer mech-
anisms are the same in the various types of boilers. Fuel characteristics
dictate such boiler design considerations as furnace volume, heat release
rate, heat transfer surface areas, and tube spacing. The simplified combus-
tion process involves the mixing of fuel and combustion air in the furnace.
The addition of adequate combustion air, mixing (turbulence), and a suffi-
ciently high temperature cause the fuel to ignite, and the heat generated by
the oxidized fuel is released in the furnace zone. Some of the heat is used
to sustain combustion, but most of it is used to transfer heat to the boiler
tubes and to generate steam.
Three heat transfer mechanisms are important to the boiler operation:
radiation, convection, and conduction. In the furnace zone where the fuel is
combusted, radiation is the dominant heat transfer mechanism. The heat
absorbed by the tubes is conducted through the tube to the boiler water to
produce steam. As the combustion flue gas leaves the furnace zone to pass
through the other tube sections, the gas is cooled because of radiant heat
244
-------
loss and the radiant heat transfer decreases. The gas temperature drops from
2200° to 2400°F in the flame zone down to 1500° to 1600°F before the gas enters
the boiler tubes, where convective heat transfer dominates. The zones where
the superheater, economizer, and air preheater are placed rely primarily on
convective heat transfer.
The boiler feedwater enters the economizer under pressure (typically 650
to 1100 psi) to be preheated by the flue gas passing around the economizer
tubes. The feedwater then enters the steam drum arrangement from which it
passes to the tubes in the furnace walls. As the preheated feedwater passes
through the tubes in the radiant heat transfer zone, it boils and forms steam.
When the .steam and entrained water droplets leave the furnace walls, they
enter steam drums that separate the water droplets from the steam. The steam,
which is saturated, is then passed through the superheater to extract more
heat from the combustion gases. The superheater is needed not only to maintain
boiler efficiency but also to prevent the turbines from operating with a com-
bination of steam and water droplets. The superheater is located at the
boiler combustion zone outlet, where both radiant and convective heat transfer
occur.
The steam exiting the superheater is typically limited to a maximum
operating temperature of 1050°F because of the boiler tube metal required to
maintain the discussed pressure ranges. At most mills, all the superheated
steam is piped to a common header that is fed to one or more turbine/generators
for electrical power generation. The energy extracted by the turbine reduces
the temperature and the pressure of the steam. Typical exit pressures are
between 150 and 250 psig to prevent formation of water droplets in other equip-
ment where steam is used (above the saturated steam conditions). The loop is
completed when the steam is cooled, condensed, and recycled back to the boiler
economizer.
Not all boilers are equipped with economizers. Multiple-drum boilers,
which have one or more steam drums and several mud drums, usually allow the
water to boil in the furnace tube walls and the tubes between the mud drums
and the steam drums. The steam drums are still used to separate the steam
from the water droplets, and a superheater section is still used to increase
the energy content of the steam. The thermal efficiency of boilers operating
without economizers, however, is generally not as high as these equipped with
245
-------
economizers. As a compromise to reduce the thermal stack losses, most boilers
are equipped with air preheaters that preheat the combustion air before it is
introduced into the furnace; this improves combustion efficiency and increases
the radiant heat released in the boiler furnace zone. Specific design features
for each type of boiler are discussed.
3.5.1.1 Gas- and Oil-Fired Boilers—
The design of boilers for firing oil or gas are very nearly identical,
and most boilers can fire either fuel interchangably. These boilers are
usually physically smaller than boilers having the same production rate but
firing solid fuels because the acceptable heat release rates and velocities
through the tube .banks are higher. The significant quantities of ash associated
with solid fuels accounts for the differences in physical design.
The oil/gas boilers used in paper mills can be either shop-assembled
or field-assembled, depending on the steam requirements. Shop-assembled
"package" boilers are usually selected when additional steam requirements are
small. These boilers are generally in the 25,000 to 250,000 Ib/h steam range.
They are small, easily shipped by railroad, and relatively inexpensive. When
larger steam capacity is needed, field-erected boilers in the range of 250,000
to 600,000 Ib/h of steam are usually the norm. As the name implies, the
boiler components are assembled at the construction site, but in some cases
modular components can be shop assembled to reduce the cost.
The fuel characteristics and the assembly method affect the design param-
eters for oil/gas fired boilers. Typical parameters of importance are the
heat release rates (based on furnace volume and radiant heat transfer area),
flue gas velocity through the tubes, and tube spacing, all of which also
affect the physical size of the boiler. Typical heat release values for oil/
o
gas boilers are 200,000 Btu/h/ft of effective projected radiant heat transfer
90
area, regardless of whether the boiler is shop- or field-erected. The shop-
erected boilers, however, generally have volume-related heat release rates
approximately twice those of field-erected units. Field-erected boilers are
generally designed with heat release rates of 25,000 to 50,000 Btu/h/ft of
furnace volume, whereas shop-assembled package boilers will be designed at
O On
50,000 to 100,000 Btu/h/ft . The type of fuel, size limitations, and steam
requirements will affect the value selected. Natural gas and distillate
246
-------
oils have higher values. The use of somewhat "dirtier" fuels like No. 6
residual oils will dictate lower values.
The type of fuel also affects the acceptable velocity through the tubes
and the tube spacing. Typical velocities through the tubes are approximately
100 ft/s because the flue gas contains very few abrasive particulates. Tube
spacings are generally 1 to 2 in., depending on the location in the boiler.
Somewhat wider spacings are provided when residual oil is fired. Draft losses
from the close tube spacing and high velocities are usually controlling factors.
3.5.1.2 Coal-Fired Power Boilers--
The use of field-erected coal-fired boilers is the alternative to firing
premium fuels (oil and natural gas) in a power boiler. Coal-fired boilers
are physically larger than their oil/gas counterparts of similar capacity.
They can be stoker-fired or pulverized-coal-fired, depending on the steam re-
quirements. In most mills, these boilers are either a spreader stoker with
travelling grate or pulverized-coal-fired unit.
Spreader-stoker boilers are used when a combination of coal and wood is
fired. Spreader-stoker boilers are limited by grate width and depth, grate
heat release rates, and furnace heat release rates. Practical stoker boiler
designs limit the size of the boiler to between 50,000 and 400,000 Ib/h steam.
Typically the boilers range in size from 100,000 to 250,000 Ib/h steam. If
the boiler is coal-fired, up to 450,000 Ib/h is possible (although these are
usually low-pressure, 150 psig steam).
The physical limitation of most stoker boiler designs is a combination
of grate size and grate heat release rates. For a spreader-stoker, the
practical limitation for the grate size is approximately 21 ft in depth. For
higher heat release, however, the furnace must get even wider. Grate widths
of 32 feet or greater present a physical limitation on the boiler size. The
grate heat release rates can usually be in the range of 500,000 to 750,000
0 Of]
Btu/h/ft of grate area. Lower values are more conservative and are con-
sistent with unusual design. The higher value is a maximum, and operation at
this level may cause operational difficulties if coal is not of the correct
quality. The furnace heat release rates for spreader-stokers are limited to
3
25,000 to 32,000 Btu/h/ft of furnace volume; the lower value is more con-
90
servative. . Coals that have lower ash fusion temperatures will generally re-
quire lower heat release rates.
247
-------
When the steam capacity of the spreader-stoker is inadequate and coal
quality problems'are expected, or when combination firing of coal with wood
is not planned, the selection of pulverized-coal (p-c) firing may be an
alternative. From an economic standpoint the break-even point for cost be-
tween stoker and pulverized-coal-firing is in the 200,000 to 250,000 Ib/h
range. Spreader-stokers rely on burning the coal on the grate, whereas in
p-c firing, the dry, finely ground (70% through 200 mesh) coal is burned in
suspension.
Although nearly any type of coal can be burned in a p-c boiler, the
boiler must be designed to handle the kind(s) of coal it will normally burn.
Differences in the ash content, the ash fusion temperature, the heat content,
and grindability of the coal will affect the overall design of the boiler.
Unlike oil/gas firing, abrasion and slagging are of concern in both the radiant
furnace zone and the convective boiler passages. Therefore, p-c boilers are
designed with the aim of preventing slagging. In general, boilers firing a
Western subbituminous coal with high slagging possibilities will be larger
than a comparable-capacity boiler firing a cleaner coal. Heat release rates
must be kept below the slagging temperatures, and this requires larger furnace
volumes for a given capacity.
Three parameters usually help define p-c boiler furnace dimensions: the
2
volumetric heat release, the heat release per ft of projected radiant heat
fi 7
transfer surface, and the plane area heat release (10 Btu/h/ft of plane area
of the furnace).
coals.
Higher heat release values are acceptable for low slagging
3
Typical values are 22,000 Btu/h/ft for volumetric heat release rate,
120,000 Btu/h/ft2 of effective radiant transfer area, and 2.1 x 106 Btu/h/ft
of boiler plane area. Values are lower for coals with high slagging poten-
tial. Typical values are 15,000 Btu/h/ft3, 70,000 Btu/h/ft2, and 1.4 x 106
7 Qf) Q1
Btu/h/ft plane area. ' These numbers govern the physical size of the
furnace zone.
The velocity through the convective tube banks in a coal-fired boiler
(typically 50 ft/s) is approximately one-half that in oil/gas boilers.
Tube spacings also are much larger, with spacings ranging from 16 inches in
some superheaters down to about 2 inches in the boiler outlet zones. Foul-
ing, plugging, and tube abrasion are of some concern when firing solid
fuels, however.
248
-------
3.5.1.3 Wood-Fired Power Boilers--
Wood-fired boilers, which are normally the travelling-grate type, usually
burn the bark from the debarking drum. Wood may also be fired in combination
with coal or in a spreader-stoker using gas or oil. The heat-release limita-
tions for spreader-stokers described in the coal-fired boiler section are
generally applicable to bark boilers. The bark is generally mass-fed onto a
travelling grate, and the depth of the bark and the grate travel rate control
the steam production rate. As with a mass-fed travelling grate boiler, the
response to load changes is slow. Firing of supplemental fuel will aid in
the swing-load capabilities of this kind of boiler. The boiler load is
usually limited to between 75,000 and 200,000 Ib/h.
3.5.2 Sources of Emissions
Emissions from all power boilers consist primarily of particulate, S02,
?. The quantity of each pollutant produced is a function of the
fuel characteristics, the firing method, and the combustion characteristics
for each boiler.
NOX, and
3.5.2.1 Gas-Fired Boilers—
Gas-fired boilers are usually the simplest boilers from the standpoint
of design and operation. Emissions generated by this "clean fuel" are gen-
erally limited to NO . Because the dominating factor in the production of
/\
NO is the flame temperature and the level of excess air, staged combustion
is used to meet excess air requirements and to reduce flame temperature by
reducing turbulence in the combustion zone. This diffusion-limited flame is
much more acceptable in reducing NOV emissions than highly turbulent, high-
X
excess-air flames. Low excess air levels (^ 15%) also promote higher boiler
efficiencies. The emissions of CO, hydrocarbons, and particulate are
negligible if complete combustion occurs.
3.5.2.2 Oil-Fired Boilers-
Fuel oil combustion generates emissions of particulate matter, S02, and
NO . Although combustion characteristics are very similar to natural gas fir-
ing, the potential for pollutant generation is greater. The two most commonly
used oils are No. 2 distillate oil and No. 6 residual oil. Normally No. 2
oil contains less residual ash and sulfur than the No. 6 oil.
249
-------
To a certain extent, combustion characteristics control the level of
participate emissions from oil-fired boilers. The oil must be atomized and
mixed with combustion air as it enters the furnace zone. Improper atomization
mixing will result in poor combustion and the generation of particulate in
addition to that present in the oil as a residual ash. The particulate gen-
erated as a result of poor combustion will be a combination of fine carbon . ,
particles and other unburned hydrocarbons. Inadequate excess air levels will
cause similar problems. Particulate generation rates will otherwise be
proportional to the ash content of the oil.
As with gas-fired boilers, NOX emissions from oil-fired boilers are gen-
erally controlled by combustion modification. Staged-combustion and diffusion-
limited flames reduce peak flame temperature and NO formation by limiting
A
excess air. Highly turbulent flames with high excess air levels promote higher
levels of NOX emissions. In addition, some oils contain residual nitrogen
compounds that can tend to increase NO production. In general, tangential
or corner firing of oil-fired boilers produces less NOX than wall-fired
boilers.
The generation of S02 is generally proportional to the level of sulfur
in the fuel, and most of the sulfur is converted to S02 upon combustion.
With sufficiently high temperatures and excess air some of the S02 will con-
vert to S03 and form sulfuric acid mist upon cooling and contact with water
vapor. This problem is more serious with residual oils that contain .vanadium
(generally 75 ppm or more). The acid mist forms very small, light-scattering
particles that can cause a significant opacity problem. Acid smuts may also
be produced when acids condense.on boiler heating surfaces. Control of excess
air, switching from high-sulfur/high-vanadium oils, or the use of an inhibitor
to reduce the catalytic action of vanadium on sulfur are the usual control
options.
3.5.2.3 Coal-Fired Boilers—
Coal-fired boilers produce significant quantities of particulate,
sulfur oxides, and nitrogen oxides. The level of each is related to the fir-
ing method, combustion efficiency, the pollution-control equipment, and the
fuel characteristics.
250
-------
The level of sulfur oxides produced is generally proportional to the
sulfur content in the coal. The formation of an acid mist is generally not
a problem unless the sulfur content of the coal is above 2.5 percent. The
sulfur content of the coal will also affect boiler design and operating
characteristics (sulfur affects the ash fusion temperature) and the choice of
particulate control equipment. The acid mist problem can be controlled, in
part, by the control of boiler excess air.
The generation of NOV is strongly related to the combustion method and
X
combustion controls. Free nitrogen in the coal also can be a significant
contributor to overall NO emissions.
A
Most NO controls deal with the control
A
of combustion air to the burners (p-c firing). Staged combustion or off-
stoichiometric firing produces lower peak flame temperatures by limiting the
amount of air available for combustion. This generally produces longer flames
that are diffusion-limited, which means longer reaction times at lower
temperatures for complete combustion. Combustion air is provided in excess
of stoichiometric requirements (usually 15 to 30 percent), but the air is
gradually introduced to the flame. It is very difficult to minimize the
quantity of NOV because of nitrogen in the fuel.
A
In general, tangential or
corner-fired p-c boilers will produce less NO than wall-fired units.
A -
Control of NO in stoker boilers is also a function of controlling both
A
overfire and underfire air. In addition, the proper placement of overfire
air nozzles is critical to the overall efficiency of the boiler. The operat-
ing principle, however, is the same. Generally, the underfire air should be
below the stoichiometric rate, which produces a zone in the top of the fuel
bed that is oxygen-deficient and causes distillation and reduction of volatiles
from the bed. When the underfire gases contact the overfire air at the
high temperatures, combustion is completed. The overfire air should provide
just enough excess air and turbulence to complete the combustion process
without causing excessive stack heat losses.
The particulate emission rate is a function of the coal ash content and
the firing method. Pulverized-coal boilers will produce more uncontrolled
ash than stoker boilers because the coal is fired in suspension. Between 70
and 85 percent of the ash in the coal will exit the boiler as fly ash.
Considerably less ash will exit with the flue gas from a stoker-fired boiler
(approximately 30 to 50%). In addition, the particulate produced by stoker
251
-------
boilers tends to be much coarser than the ash from p-c boilers and will
generally contain considerably more combustible materials. These factors
affect the selection of control equipment for these boilers.
3.5.2.4 Bark Boilers—
The characteristics of bark boilers are similar to those of coal-fired
stoker boilers except with regard to sulfur oxides emissions. Because bark
contains virtually no sulfur, emissions of S02 are negligible when bark is
fired alone. Thus, nitrogen oxides, particulate, and unburned hydrocarbons
are the potential emissions from bark boilers.
Particulate emissions from bark boilers are relatively low because the
bark is mass-burned on the grate rather than "thrown" on the grate. Also,
the ash quantities in bark can be significantly less than those in some coal
supplies. Thus, overall potential emissions are less when compared with coal
firing alone.
Nitrogen oxide emissions potential and control for bark boilers are
similar to those for coal stoker boilers. More care is required, however,
for proper adjustment of the overfire and underfire air, because the high
moisture content of the bark (up to 50 percent by weight) can cause combus-
tion problems. It should be noted that this moisture may be beneficial from
a NOV standpoint because it lowers the effective flame temperature. Good
A
overfire air distribution and sufficient residence times in the furnace zone,
however, are needed for complete combustion. This can make control of NO
A
and particulate emissions very difficult. In addition, the unburned hydro-
carbons will tend to form very fine, sticky particulate. Using natural gas
or oil as a supplemental fuel over the fuel bed sometimes minimizes these
problems provided boiler design heat release rates are not exceeded.
3.5.3 Control Techniques
The kind of control selected for power boilers .depends on fuel type, the
method of combustion, fuel and ash characteristics, and the costs of the
equipment options available to meet the prescribed .emission requirements.
3.5.3.1 Gas-Fired Boilers--
No add-on control equipment is applied to natural-gas-fired boilers
because particulate and S0? emissions are negligible and NOV is controlled by
£ * X
combustion modification.
252
-------
3.5.3.2 Oil-Fired Boilers-
Depending on the emission requirements, oil-fired boilers may have add-on
particulate control equipment. Oil-fired boilers that fire No. 2 distillate
oil do not require particulate controls. Boilers that fire residual oil
efficiently also may not require controls. If controls are required, they
usually include multicyclones, small ESP's, and scrubbers. Multicyclones are
generally effective in collecting particles down to 5 to 10 ym in size at 3
to 4 in. pressure drop. Unfortunately, multicyclones are effective in a very
narrow range of gas volume around the design point. Small ESP's also can be
used, but the initial capital cost is high compared to that of other control
options. Scrubbers also may be used, but high operating costs can discourage
this option. One extra feature that a scrubber can add is its ability to re-
move a portion of the SOp emissions, especially with the use of an alkaline
scrubbing medium.
3.5.3.3 Coal-Fired Boilers--
Coal-fired boilers require the application of control equipment to meet
particulate emission standards. Multicyclones, scrubbers, ESP's, or fabric
filters are acceptable control techniques for spreader-stoker boilers. These
same control options are applicable to p-c fired boilers except that the
multi cyclone would be used only in combination with either an ESP or fabric
filter, not as a separate control device.
The use of a scrubber for controlling particulate is more practical for
a spreader-stoker boiler than for a p-c boiler because the stoker boiler pro-
duces larger particles. Also, because the mass loading from the spreader
stoker is considerably less, pressure drop and liquid volumes to the scrubber
are smaller. The scrubber must be designed to avoid resuspension of particu-
late matter as a result of the high suspended and dissolved solids in the
scrubbing liquor. Also, scrubber components must be designed to withstand
possible corrosion.
Electrostatic precipitators can be used on either stoker or p-c fired
boilers. Design specifications for the two types will differ considerably,
however. Stoker boilers generally produce an ash that is composed of carbon
and other hydrocarbons, of which 30 to 50 percent is combustible material.
This can produce a low-resistivity ash that is difficult to retain on the
253
-------
collection plates. Although the actual surface area required to collect the
o
particulate may be low (100 to 175 ft /1000 acfm) and the power input may be
high, the ESP must be designed with low superficial velocities to minimize
particle reentrainment.
The opposite situation may occur with p-c boilers equipped with ESP's.
Because combustion is usually more complete in a p-c boiler, the other con-
stituents in the coal play a major role in determining ash characteristics.
High resistivity can be of major concern, particularly when low-sulfur/low-
sodium coals are burned. If high resistivities are encountered, the ash tends
to be very tenacious when collected in the ESP. This makes the ash difficult
to remove from the plates. High-resistivity ash causes a significant voltage
drop across the dust layer, which can cause sparking, back-corona effects,
and reduced power input to the ESP. Potential solutions are to change the
coal supply; to change operating,temperatures to produce better resistivity
characteristics; to use such conditioning agents as ammonia, water, or SO,;
or to size the ESP to operate at low power input with adequate treatment
times.
Fabric filters are often used to avoid the problems with resistivity.
This control technique is suitable for both stoker and p-c fired boilers.
Particular attention, however, must be paid to coal quality and firing prac-
tices of stoker-fired boilers to avoid bag blinding problems due to unburned
particles of carbon and sticky hydrocarbons. The fabric filter is usually
®
designed for high-temperature operation, and either fiberglass or Teflon
bags are used for particulate collection. The two types of bag cleaning
mechanisms are reverse-air and pulse-jet. The reverse air mechanism is
p
generally limited to an air-to-cloth ratio (A/C) of 2.5 acfm/ft of cloth
area. The pulse-jet mechanism can operate in a somewhat higher range (A/C
ratio of 4.0 to 4.5) because of its higher available cleaning energy. A pre-
cleaner such as a multicyclone or a simple impaction baffle plate is usually
employed to remove the larger, more-abrasive particulate. The normal operat-
ing pressure drop range for fabric filter is 3 to 6 in. of H^O.
The use of a mechanical collector or multicyclone as the sole control de-
vice is generally limited to stoker fired boilers. The larger particulate and
the lower uncontrolled emission rates make this a suitable control option for
some sources. These devices, however, operate most effectively at or near the
design gas volume. Thus, care must be taken not to exceed the gas volume.
254
-------
3.5.3.4 Bark Boilers—
Participate emissions from bark boilers have been controlled by either
mechanical collectors or scrubbers, and sometimes both. More recently some
bark boilers have been equipped with ESP's to control emissions, but operating
experience is relatively limited. Although bark boilers behave much as coal-
fired stoker boilers do, a major concern is the generation of fine particulate
as a result of unburned hydrocarbons leaving the furnace. These particles are
difficult to control when a scrubber or multicyclone is used. Therefore, com-
bustion control is generally more effective. These particles are less difficult
to collect in a properly sized ESP, although they are potentially sticky.
This potential stickiness almost precludes the use of fabric filters.
Variable-throat scrubbers are generally used to maintain a preset pres-
sure drop over a variety of gas flow rates. Multicyclones are less tolerant
to flow rate changes and perform better with a stable flow near design gas
volume conditions. Pressure drop requirements for a scrubber may be high (15
to 25 in. H20), whereas 3 to 4 in. H20 pressure drop is typical for the multi-
cyclone. The two major disadvantages of scrubbers are the space requirements
and the disposal of the scrubbing liquor containing the collected ash.
3.5.4 Malfunctions
Numerous malfunctions can occur in power boilers, but the two most com-
mon operational problems are fuel quality and establishment of proper excess
air levels. Often the excess ai.r level is much higher than needed for com-
plete combustion. This situation can decrease boiler efficiency, increase
the amount of fuel fired to develop a given quantity of steam, and increase
emissions.
The presence of the extra nitrogen and oxygen in the excess combustion
air in combination with the combustion gases generally produces a dilution
effect. Although the peak flame temperature may increase, the average flame
temperature decreases because an extra quantity of gas must be heated. This
decrease in average temperature decreases the radiant heat transfer to the
furnace walls. In extreme cases, the extra gas volume will cause incomplete
combustion because of the average decrease in temperature and increase in the
vertical velocity out of the boiler, which causes unburned fuel to be carried
out of the furnace zone. This is contrary to combustion principles, which
255
-------
indicate that a decrease in temperature requires an increase in reaction time.
As the flue gas leaves the flame zone, it begins to lose heat and starts to
cool.
. Increased gas volume resulting from increased excess air means that the
velocity of the flue gas through the convective tube passes in the boiler must
increase because the available volume between the tubes is fixed. An increase
in velocity of the flue gas over the tubes will increase the convective heat
transfer rate. Unfortunately, the tube area is fixed, and the flue gas
passes through the convective heat transfer zone faster than normal, and the
increased heat transfer rate in the convective section of the boiler never
quite offsets the decrease in heat transfer in the radiant zone. Thus, more
heat is lost from the boiler during high excess air conditions than during
normal low excess air firing. "The usual indicator of heat loss is an increase
in stack temperature and oxygen levels. It should be noted that controlling
the excess air is one way that boiler operators can use to control heat
transfer rates in various portions of the boiler. Improper adjustment or
failure to calibrate the controls periodically can reduce boiler efficiency
and make steam adjustments difficult.
Minimum excess air levels differ for the various fuels, ranging from 10
to 15 percent for natural gas, 10 to 25 percent for oil, 20 to 30 percent for
p-c boilers, 30 to 50 percent for spreader-stoker boilers, and 75 to 100
percent for bark boilers. The minimum excess air levels for highest boiler
efficiency are generally those at which small levels of carbon monoxide begin
to appear. Boiler excess air is controlled by the use of C02 or 02 monitors
at the outlet of the radiant heat zone. Typically, CO levels are maintained
at approximately 100 ppm. Maximum levels of 400 ppm are usually established
to prevent explosions of CO pockets within the boiler due to incomplete
combustion of the fuel.
Another boiler problem that shows similar symptoms to high excess air
levels is the failure of sootblowers. Sootblowers may use steam or compressed
air to clean deposits of the boiler tubes. The loading and characteristics
of the ash dictate sootblowing requirements. Continuous sootblowing may be
required on a p-c boiler, whereas a cycle of once per shift may be sufficient
for oil-fired boilers. Failure to blow the soot from the boiler tubes allows
deposits to form, which reduces the heat transfer rate through the tubes.
256
-------
The resulting decrease in efficiency is characterized by an increase in stack
temperature with the 02 or C02 levels remaining the same.
The mixing of natural gas and combustion air entails relatively few
problems. With the firing of oil or solid fuel, however, provisions must be
made for effective mixing of combustion air and fuel without excessive turbu-
lence or excessive quantities of excess air. In oil firing it is necessary
to atomize the fuel in the burner. This is usually accomplished by using
steam or air at a pressure of at least 30 psig. The finer the atomization,
the better the combustion, because there is more droplet surface area on
which the combustion processes can take place. Failure to check routinely
for proper atomizing can result in poor combustion and excessive carbon
carryover.
In coal-fired boilers, coal quality can be a serious problem. Problems
vary according to the firing method, but the effect of coal quality is equally
great on emissions and combustion efficiency. Because of their fundamental
differences, stoker and p-c firing are discussed separately.
Spreader-stoker boilers are very sensitive to coal characteristics. For
example, for proper combustion, the size of the coal must be within the range
of 1-1/4 inches to fines. (See Figure 3-91.) These size distributions
represent coal delivered to the boiler, not that received from the suppliers.
The spreader-stoker must have some fine coal because 30 to 50 percent of the
coal is burned in suspension and the larger coal falls to the grate to form a
fuel bed. The mid-sized to larger coal lumps are necessary to form a fuel
bed of the proper porosity for good combustion. If excessive fines are
present in the coal, too much coal is burned in suspension and carried out of
the boiler, which causes poor combustion (high carbon loss) and excessive
particulate carryover. Too much handling of the coal can create high levels
of fines. On the other hand, the lack of fines can have an adverse effect on
combustion because of excessive burning on the grate, unstable flame conditions,
and high carbon loss. The coal must also be evenly distributed over the
grates from side to side and front to rear.
Coal sizing is not a big problem in p-c firing except for the maximum
size. The maximum top size for most pulverizers is about 2 inches. Very
high quantities of fines may cause some problems with some pulverizers,
but most are easily capable of producing the required fineness (70 to 75
257
-------
95
A ft
oU
/U
CA
OU
^n
40
30
25
O)
5 20
" 15
O)
§ 10
£ 8
1—
•4-> 6
G
V. \ \ _v
•a 11 -i co AC
-------
percent through a 200-mesh screen). The fundamental difference between the
p-c boiler and the stoker is that in p-c firing the coal is dried, crushed,
and burned in suspension, whereas in stoker firing it is fired as is. Sig-
nificant changes in coal characteristics, however, can cause problems.
Most p-c fired boilers installations have at least one extra pulverizer
to allow routine maintenance to be performed without reducing the boiler
load. Selecting the number of pulverizers to handle the desired load is
based on the heat content and grindability of the coal because these charac-
teristics define the quantity of coal that must be ground per hour, minute,
day, etc., and the rate at which the coal can be ground to the desired fine-
ness. They also define the minimum pulverizer capability required and the
necessary redundancy for maintenance. Because pulverizers are expensive,
excess capacity is held to a minimum. If the grindability of the coal de-
creases (more difficult to grind) or the heat content of the coal decreases
significantly, or both, the existing pulverizer capability may not be suffic-
ient to provide the necessary fuel input to maintain steam rate. This is
often indicated by pulverizers operating at full capacity and the boiler still
being unable to reach full load.
Changes in ash content and other characteristics of the coal can signif-
icantly affect boiler operation. Increased ash content may increase the
sootblowing requirement and where sootblowing capabilities are marginal, it
can increase boiler stack losses due to inefficient heat transfer. A more
serious problem is an increase in slagging potential of the ash. This may
entail derating of the boiler to lower heat release rates to prevent slagging
and increasing the sootblowing requirements. Boiler operating conditions may
become difficult to control because of a sticky ash coating that reduces heat
transfer efficiency. These sticky ashes can be very difficult to remove,
and they can cause damage to the boiler tubes due to localized hot spots.
These problems may be the result of changes in a coal seam or the supplying
coal mine. Coal blending can sometimes lead to similar problems because the
euectic ash formed by the blended coal may have an ash fusion temperature
lower than either coal on its own.
Improvement in coal quality also can lead to operating problems in the
p-c boiler. If the boiler was designed for a slagging coal and a nonslagging
coal is burned, the furnace walls will be too clean for the design conditions.
259
-------
A higher heat release rate can be tolerated under these conditions, but too
much radiant heat is usually adsorbed, which makes it difficult for,the
superheater to obtain the high superheated steam temperatures. Turning off
or reducing sootblowing helps alleviate this problem.
Fineness of the coal from the pulverizers should be checked weekly.
Failure to feed a fine coal to the burners may cause incomplete combustion of
coal, which allows unburned, raw coal to pass out of the boilers to the
control equipment and represents inefficient boiler operation. Excessively
fine coal may result in excessive pulverizing energy being expended and the
possible taxing of the pulverizer capacity at full load conditions. When the
coal is Eastern bituminous, the test should show that 70 to 75 percent of the
pulverized coal would pass through a 200-mesh screen. When the coal is
Western sub-bituminous, somewhat less fineness (60 to 65 percent) is needed
because of the "noncaking" properties of these coals.
A tube leak will cause boiler shutdown, and it can affect the control
equipment operations. Tube leaks that are most serious from the standpoint
of affecting control equipment operation are waterwall, boiler tube, and
economizer tube leaks. Significant quantities of water can escape into the
flue gas and cause possible pluggage of multicyclones and fabric filters and
make ash removal from ESP plates difficult. Superheater tube leaks are not
as serious in terms of their effect on control equipment because the steam is
superheated rather than saturated.
In cold weather conditions, freezing of the fuel, particularly coal
and wood bark, can prohibit proper flow of fuel to the boilers. The fuel
can hang up in chutes, hoppers, or feeders, and in the case of coal shipped
by rail, can freeze up in the coal railcars. On stoker boilers, hang ups
in the fuel feed system can cause improper distribution of fuel on the
grates. When this occurs, underfire air can channel to these uncovered
portions of the grates because there is less resistance to air flow. This
reduces the underfire air to other portions of the grate. This reduction
may cause distortion of the grates because both the ash layer and the
underfire air help protect the grates, the former by providing an insulat-
ing layer on the grate surface and the latter by providing grate cooling.
In addition, the channeling of the air destroys the air/fuel ratio and in-
creases emissions. Residual oil tanks and fuel lines are generally heated
260
-------
to maintain the proper oil viscosity. Failure of these heating systems re-
sults in the inability to pump the oil.
Each control equipment category has a characteristic group of mal-
functions related to the type of process being controlled and to the design
characteristics of the equipment itself. The malfunction mechanisms that
have been outlined in previous sections are discussed only briefly. Mal-
functions in control equipment not used elsewhere in the mill are discussed
in more detail.
3.5.4.1 Mechanical Collectors--
Mechanical collectors, specifically multicyclones do not have many
applications in the paper mill. They are generally used to control emissions
from stoker-fired coal, wood, and combination coal/bark-fired boilers. In a
few cases they are used to control emissions from oil-fired boilers. They
also can be used as precleaning devices for other control equipment, such as
fabric filters and ESP's. Because mechanical collectors are relatively
simple to operate, they do not have many failure mechanisms. Nevertheless,
they require periodic checks and maintenance. There are several methods in
which multicyclones fail to give the desired performance; some relate to
boiler operation and others are caused by mechanical failure of components.
Multicyclones are designed to operate in a narrow range around one
design gas volume and uniform particulate density. When excess air levels
are significantly higher than design, however, the flue gas temperature and
the firing rate are increased to compensate for decreased efficiency. As a
result, the flue gas volumetric flow rate through the multicyclone can be
substantially higher than design. This will lead to an increased pressure
drop through the tubes and the application of increased inertia! separation
forces to the particles. Unfortunately, there is a practical limit to the
amount of pressure drop and inertia! forces that can be applied. At high gas
volumes, fan capacity may be taxed and the increased pressure drop across the
multicyclone may limit the quantity of gas that may be moved. When this
"fan-limited" condition exists, the boiler produces more flue gas than the
fan can handle, and the boiler operates under positive static pressure. Most
boilers are designed to operate under slight negative pressure (between 0.1
and 0.15 inches H20). In addition, the inertia! forces applied to the partic-
ulate may be wasted because of a particle "bounce" phenomenon that occurs at
261
-------
very high tube velocities. Operation at the low end of the gas volumes re-
sults in low inertia! forces.
Another problem lies in expecting the multicyclone to collect very fine
particulate. This equipment is not very efficient at collecting particles
below 10 ym in size, which is one reason why collection of particulate from
p-c boilers by multicyclones alone is generally not acceptable. Because
stoker-fired boilers usually produce large particles, multicyclones are very
effective in their control. If combustion problems cause significant quantities
of unburned carbon or hydrocarbons to be produced (as a result of fuel quality
or air distribution problems), however, these particles are likely to pass
through the mechanical collector rather than being collected.
Fly ash reinjection is not a satisfactory approach either. If reinjectipn
is used, the larger particles (char) with higher carbon contents should be
screened and reinjected into the boiler, and the fine particulate should be
disposed of. Reinjection of all of the multicyclone ash means the ash must
either leave as boiler ash or be reduced down in size until it is capable of
passing through the collector. Little operating efficiency can be gained by
using this method.
The material captured by the mechanical collector should not be stored
in the hopper. Ideally, the ash will be removed by continuous discharge, •
which minimizes the possibility of hopper pluggage. Hopper pluggage can
cause buckling of the hopper or softening of the tubes due to annealing with
hot ash around the tubes. Hammering on hoppers until they become deformed
does not help future performance and can actually cause future hopper bridging
problems by establishing a "shelf" in a previously smooth hopper wall.
If the collector is located on the inlet side of the fan and is under
negative static pressure, the hopper should be sealed by a rotary airlock,
double-gate flapper valves, or some other isolation valve to prevent air
inleakage through the hopper discharge. Inleakage through the discharge can
cool the ash so that it does not flow as easily. Inleakage can also resus-
pend the captured ash into the gas stream and destroy or disturb the separa-
tion vortex established in each tube.
The multicyclone tubes are also capable of plugging. The tubes can
become plugged in the outlet tube, which will cause the gas entering the tube
to flow out through adjacent tubes and disturb their vortex (Figure 3-92).
262
-------
til
1
§
III
o
o
Figure 3-92. Plugged inlet vane.
263
-------
The inlet vanes of the multicyclone can also become plugged and disrupt the
vortex pattern (Figure 3-93). And finally, the discharge of a tube can
become plugged and allow material to enter the tube and the particulate to be
resuspended (Figure 3-94). Because approximately 90 percent of the pressure
drop is established across the inlet vanes, a number of tubes could plug
without any significant change in the collector pressure drop.
Another malfunction mechanism commonly observed in mechanical collectors
is lack of routine inspection and failure to replace vanes and tubes within
the collector. A maximum period between tube checks should be one year.
These components are subjected to high abrasion forces and cannot perform
their tasks as effectively when they are worn.
The last malfunction, one that has been encountered occasionally, is
improper sealing between the inlet and outlet of the mechanical collector at
the tube sheet. Specifically, gasket material used to seal the tubesheet and
tubes must be selected to withstand the normal peak operating temperatures
expected. Improper selection or installation of the gaskets will result in
the bypass of particulate around the tubes.
3.5.4.2 Scrubbers—
The use of a scrubber is generally limited to stoker-fired bark and/or
coal boilers, although it can be extended to p-c boilers and oil-fired
boilers. The operating problems and characteristics are generally similar
to those found on venturi scrubbers applied to lime kilns. It is suggested
that the section on malfunctions (Section 3.3) be reviewed for a discussion
of scrubber malfunctions.
Scrubber performance is affected by poor pressure drop maintenance, .im-
proper water flow rates, high suspended and dissolved solids, pluggage and
erosion of pipes and nozzles, pump wear, and to the inability to capture fine
condensable particulate. These problems, particularly with bark boilers, are
related to incomplete combustion of carbon and hydrocarbons produced in the
fuel bed and are analogous to the evolution of alkali materials from the
kiln. Unfortunately, these hydrocarbons may not exist as a particulate until
they enter the scrubber throat and are cooled by the contact with the scrub-
bing liquid (evaporative cooling). In addition, the use of a presaturator to
cool the gas stream and allow particulate growth may enhance particulate col-
lection, but the hydrocarbons may still remain in the very fine particle range
264
-------
Figure. 3-93. Plugged outlet tube.
265
-------
PLUGGED COLLECTING CONE
Dust Discharge
ro
en
CQ
fC
co
(Q
CO
o
o
fD
O
3
UD
CT
fD
-------
and pass through the scrubber. Very high pressure drops may be required to
collect the particulate and prevent a residual blue, haze characteristic of
hydrocarbon emissions. Proper combustion will also help minimize the problem
of hydrocarbon emissions. ,
3.5.4.3 Fabric Filters-
Fabric filters are generally applied to p-c boilers and sometimes to
stoker-fired coal boilers. They are applied to minimize emissions or avoid
resistivity problems associated with some ESP applications. The capital
costs of fabric filters are generally less than those for ESP's, but operating
costs are usually somewhat higher.
Failures and malfunctions in a fabric filter can be related to four
general categories: failure of the cleaning system, failure of the dust
discharge system, failure of the fabric, and a malfunction in the boiler.
Some of the problems are related to two or more areas. Malfunctions in any
of these areas can affect the overall performance and can cascade to other
problems in the fabric filter.
An area of major concern when applying fabric filters to stoker-fired
boilers is that of maintaining proper coal properties and combustion in the
boiler. As has been mentioned previously, the operation of a spreader stoker
is sensitive to the size of the coal, particularly the quantity of fines
in the coal. If excessive fines are present, the excess carbon carryover can
blind the fabric with a very small, sticky particulate that makes cleaning
difficult. The result of this blinding is excessively high pressure drops
that can tax the gas-moving capabilities of the fan. This will eventually
cause a load reduction due to fan-limiting conditions. Eventually, the fabric
filter will have to be removed from service and the bags replaced. Most
boilers cannot operate effectively with a pressure drop of 10 to 14 in. H20.
Another problem related to the high carbon carryover is the potential of
fines in the fabric filter system. If oxygen levels are high enough in the
presence of the carbon carryover, the result can be the ignition of the car-
bon to a high temperature glow if a spark is carried to the baghouse. This
dD
can destroy a Teflon bag or take the abrasion-resistant finish off a
fiberglass bag. If the finish is removed from a fiberglass bag by high
temperature conditions, it will become brittle and abrade itself to form
pinhole leaks and tears in the bag.
267
A fabric filter may be successfully applied to a spreader-stoker boiler
if combustion air and coal quality are carefully monitored. Although these
parameters are not problems on p-c fired boilers, similar problems may occur
if the air balance is upset in one of the burners. The major concern with
these boilers is the temperature of the flue.gas.
As in other control equipment, dust discharged into the hoppers should
not be stored for any period of time. A continuous dust discharge from the
hoppers is the best approach. If allowed to build up in the hopper, the dust
can close off the cleaning contribution of a module and increase the A/C
ratio, pressure drop, and abrasion of the other bags. Also, hopper outlets
should be sealed against air inleakage to prevent resuspension of the partic-
ulate into the gas stream or excessive cooling of the fabric filter.
Proper installation of the bags is very important to the overall per-
formance of a fabric filter. Bags should be the proper material, weight, and
size. If bags do not meet the basic specifications, they will not survive in
the environment to which they are subjected. Bag tension and handling of the
bags also affect the bag life. Improperly tensioned bags or bags that are
mishandled will wear out and develop leaks.
Some provision should be made for abrasion resistance in the fabric fil-
ter. The use of a precleaning device to remove the largest, most abrasive
particulate will generally improve bag life. The larger particulate is gen-
erally unable to follow the gas flow and will follow a course that causes it
to impact with the bottom of the bags opposite the inlet. It is usually the
bottom 18 inches of the bag that has the.worst abrasion damage. Fabric fil-
ters that have reverse-air cleaning (where the filtration is from inside to
outside of the bag), long thimbles, and cuffs on the bag minimize wear
(Figure 3-95).
Cleaning system failures can cause significant problems. Because the
cleaning mechanism governs the maximum A/C ratio it also controls the cleaning
frequency to some extent. In general, the higher the A/C ratio, the more
frequent and more energetic the cleaning energy requirements. Failure to
clean the fabric means an increase in the thickness of the dust layer (which
does most of the filtering) and an increase in pressure drop. If prolonged,
this situation can lead to serious long-term consequences. The response of
the control systems on the boiler would be to maintain gas flow rate by opening
268
-------
THIMBLE AND CLAMP RING DESIGN
POOR BETTER
CLAMP-
SAG
.IKCREASED
«r ABRASION
SHORT CUFF
rf
«aftr«.... , >
J <--H
.
BAG
1
I
LONG
j THIMBLE ,
J
I
*
> LONG CUFF AND
"^REDUCED ABRASION
^^
eu ri ny / TUBE SHEET '
POOR
BAG
CUFF
WITH
SNAP
R1N6
J
" 6AS
FLOW
POOR
SNAP RING DESIGN
SHORT CUFF
NO THIHBLE
INCREASED
„/ABRASION
•BETTER
BAG
CUFF WITH
SNAP RING
V6AS I
LONG CUFF AND
REDUCED ABRASION
TUBE SHEET
AND
THIMBLE
fiAS FLOW
Figure 3-95. Fabric filter bag attachment methods.
269
-------
a da'mper. Thus, gas flow would likely be the same (for awhile) at a higher
pressure drop across the fabric. Unfortunately, this pressure drop can mean
that fine particulate can be drawn further into the fabric weave (because of
the higher available energy drop) and cause the particulate to remain even
after the cleaning system is repaired. This partial blinding of the bag can
cause a permanent increase in pressure drop until the bags are replaced. This
can happen on both reverse-air and pulse-jet cleaning systems, although it is
usually more serious on pulse-jets. This problem is usually detected by
a slight increase in pressure drop if one module or row is involved and the
cleaning system fails to activate. The use of individual manometers or
magnehelic gauges on individual compartments of multicompartmented baghouses
will not detect this problem because static pressures and flow will balance
out as the air will follow the line of least resistance through the baghouse.
Another problem (limited mostly to pulse-jet fabric filters) is the presence
of moisture and oil in the compressed air supply. Air-line dryers and oil traps
should be provided to avoid blowing moisture and oil into the bags at 90 to 120
psig. The moisture and oil will combine with the particulate to blind the
bags and cause a high pressure drop. Provisions for blowdown and other design
changes will help minimize the problem, but periodic checks of compressor rings
and seals in combination with the dryer will virtually eliminate the problem.
3.5.4.4 Electrostatic Precipitators—
The malfunctions of an ESP applied to a power boiler are generally the
same as those on ESP's applied to the recovery boiler, with two notable excep-
tions: 1) fly ash resistivity must be considered in this application, and
2) ash discharge designs for fly ash are different than for salt cake, although
problems caused by the failure of the discharge system are nearly identical.
As discussed in other sections of this guide, an ESP charges particles
through a corona discharge process, and under the force of an electric field,
moves the particles to the collection plate. At the collection plate, the
particulate must leak some of its charge to the plate to complete the electrical
circuit in the ESP. The resistivity of the particulate (measured in ohm-cm)
governs how easily the charge is transferred through the dust layer to the
plate. The particulate must be capable of retaining some charge so that it
will remain on the plate until it is rapped off.
270.
-------
The optimum range of particle resistivity in an ESP is 1 x 108 to 1 x
10 ohm-cm. This range allows for good collection of particulate under a
strong electric field and results in high particle migration rates. The par-
ticles retain enough charge in this resistivity range to remain on the plates
under the force of the electric field and to allow electrons to pass rela-
tively easily to the plates. It is when the resistivity begins to go outside
this range that operational problems begin.
Resistivity of particulate is not a constant value; it depends on several
factors, including the chemical composition of the dust, the presence of
moisture and other gases, and the temperature of the gas. As Figure 3-96
shows, resistivity varies over a wide range of temperatures, and the peak value
for fly ash resistivity occurs in the 300-350°F range. Unfortunately, this
happens to be the range in which most boilers are operated. If the resistivity
falls outside the optimum range, serious problems may develop. At low
temperatures, electrons tend to be conducted over the surface of the particle
because of the condensation of the conducting species on the particulate. At
high temperatures, the thermal activity of the electrons within the particulate
allow the electrons to pass through the particulate in a mechanism called
bulk conductivity. At intermediate temperatures, neither mechanism is very
effective and the composite effect is an increased resistivity.
Low resistivity can usually be associated with high levels of carbon in
the particulate. The particles are charged in the interelectrode space be-
tween the discharge electrode and plates, as usual. During migration to the
plate, however, the particles leak too much charge to the plate and the
electric field strength in the dust layer is low. In the absence of a strong
charge, the particle is free to become resuspended in the gas stream (reen-
trainment) and to be recollected. The general approach to low resistivity is
to design for low superficial velocities and to minimize rapping and perhaps
ammonia conditioning to reduce reentrainment potential. The electric power
input to the ESP is usually high, but without the proper resistivity, per-
formance can be poor. If low resistivity is caused by poor combustion, com-
bustion problems should be solved.
The other extreme, high resistivity, is of much more concern in p-c fired
boilers, where combustion is usually complete. Fly ash constituents, coal
sulfur content, and temperature play a major role in governing the ash
.271
-------
RESISITIV1TY
.13
10
10
- *J1
« 10
ce
10
.10
Minimal
Lavals,
I I 1
Duo to Bulk
Conductivity
Only
0 100 200 300 400 600 600
Gas T«mperatura, °F
^_____ Du» to Combinod
J Bulk and Surface
Conductivity
Charge*
Putici*
». ,-"-•» Surface Conductivity
^-N-'
Bulk Conductivity
Figure 3-96. Fly ash resistivity curve.
272
-------
resistivity. In general, low-sulfur coals have high potential to cause
operatin problems with high resistivity.
The high-resistivity condition does not affect the ability of the particle
to accept a charge. The problems begin when that particle reaches the dust
layer at the plate. The particles cannot release and transfer the charge
easily, and the charge that is transferred requires considerable energy. At
the high voltage drops across a relatively thin dust layer, the gas between
the particles begins to break down much like the corona discharge occurring at
the discharge electrode. Because very high voltage drops are present across
both the dust layer and the interelectrode space, the localized breakdown
causes a spark to occur. Many of the breakdowns occur through the dust layer,
and high resistivity is characterized by high sparking rates throughout the
ESP. The critical resistivity is generally around 2 x 10 ohm-cm, where these
localized breakdowns occur, and performance continues to deteriorate as resis-
tivity increases. During normal operation, random sparking generally does
not cause problems, even though the voltage field collapses during the sparks
and all the power in that field is channelled to the spark. If sparking is
serious, the T-R controls limit sparking by limiting the strength of the voltage
field and current flow. This reduces the particle charging rate and the
migration rate in the ESP's. The voltage and current in the ESP will continue .
13
to drop as the resistivity climbs into the 1 x 10 ohm-cm range.
The sparking, which is a symptom of a phenomenon known as back corona,
finally can no longer take place because operating voltages are too low to
propagate a spark across the interelectrode space. The voltage drop across
the dust layer, however, is still high enough to cause a breakdown in the
dust layer, and the dust layer produces significant positive particles that
tend to cancel the charging process and effectively reduce the ESP to the
performance of a settling chamber. Back-corona is usually characterized by
low voltages and high currents, as well as a particular V-I curve relation-
ship. If the ESP is to operate with any efficiency, it must operate at the
very low levels, below back-corona onset.
The absence of electrically conducting species such as sodium and sulfur
oxides in the form of sulfuric acid help enhance the resistivity of the fly ash.
The absence of these species increases the potential for problems with high
273
-------
resistivity, particularly when the ash contains high levels of calcium. If
coal supplies vary significantly, resistivity problems may appear periodically.
The most obvious of several potential solutions is to change the coal
supply. Other possible solutions include 1) designing the ESP sufficiently
large to handle the high-resistivity conditions, 2) adding conditioning agents
such as water or S03 to the gas stream to enhance the surface conductivity
characteristics, 3) adding sodium compounds such as sodium carbonate to the
coal, 4) adjusting the operating temperature to avoid the resistivity problems
or moving the ESP to a different location (e.g., before the air preheater)
where temperatures make resistivity more acceptable, or 5) adjusting and
modifying T-R sets to take advantage of the new pulse-energization technology.
Each solution must be examined in terms of the site-specific application.
Bark boiler ash apparently creates no resistivity problems in an ESP.
Neither does oil-fired boiler ash because there is sufficient residual carbon
in the ash to keep resistivity to acceptable levels. In the case of oil-fired
boilers, however, there can be a problem with excessive levels of carbon in
an ESP. High temperature, sufficient oxygen, and sparking can lead to ESP
fires, which can cause catastrophic failure of the ESP because burning
temperatures are hot enough to warp the plates and frame guides. The fires
can occur in hoppers, on the plates, or both. This is one reason why ESP's,
like fabric filters, are usually bypassed or deenergized on startup.
3.5.5 Inspection of Power Boilers
This section summarizes the activities associated with power boilers and
the associated control equipment. It also identifies the data that should be
collected during a Level III inspection and the procedures that should be used
to evaluate these data.
The approach taken during the inspection is to identify those operating
parameters or variables that indicate operation outside the norm for a particu-
lar boiler/control equipment combination. Normal values or conditions are
established during the initial performance test or are based on the accepted
state-of-the-art. This approach to source evaluation may be used to determine
if a more detailed evaluation or performance stack test is required to verify
compliance with emission standards.
The following subsections summarize specific areas that should be checked
during the inspection.
274
-------
3.5.5.1 Opacity--
The inspector should observe opacity according to EPA Reference Method
9. The observation should be made for at least 30 minutes to determine if
any cyclic patterns are present. If further evaluation is warranted, the plume
should be observed over a continuous period to identify any "puffing" problems
when ESP's or fabric filters are used. The latter is a separate activity that
is conducted after the Method 9 observation. It should be noted that opacity
observations of boilers firing only natural gas should not be necessary,
although a smoking natural gas boiler indicates combustion problems.
3.5.5.2 Transmissometer Data—
If units are equipped with opacity monitors (transmissometers), the in-
spector should record the current 6-minute average opacity and review the previous
4 hours of monitor output to determine if a cyclic pattern is occurring.
To ensure that the output values are accurate, the inspector should request
the plant to place the monitor in the calibration mode with respect to zero
and span. As part of the initial monitor certification, the inspector should
have data available on the recorder scale factors and effective stack diameter.
Average opacity readings from the Method 9 observation should be compared with
average transmissometer readings for identical periods. A major deviation
between the values may indicate possible monitor error. It should be noted
that the manual and instrument methods are not equivalent. In sources that
have real-time monitor output, instantaneous opacity spikes will generally
be included in the 6-minute average." Therefore, the 6-minute averages ob-
tained from the monitor will generally be higher than those obtained by the
manual method because the duration of the spikes are too short to be ob-
served by the inspector. For ESP's with transmissometers, the inspector should
note the frequency and magnitude of rapper spikes and visually determine if
an inlet-to-outlet field rapper pattern is occurring. If separate monitors
are installed in each duct, the inspector should evaluate the opacity and
rapper reentrainment pattern in each chamber. The opacity should be compared
with a typical baseline value for the boiler during known emission periods
(i.e., performance tests). The opacity data should be used to evaluate ESP
conditions.
Serious deviations in opacity between chambers can indicate gas flow
maldistribution, an increase in penetration through one chamber as a result
of rapper failure, inleakage, or low power input.
275
-------
For fabric filters, the inspector should note the frequency and magnitude
of spikes after the cleaning cycles. These can indicate which compartment or
row may have a leak because some slight rise in opacity will normally occur
after cleaning. Once the dust cake has reestablished itself over the holes,
the opacity should decrease.
Opacity monitors are rarely used with scrubbers because of the wet plume
conditions; however, they are used occasionally following multicyclones.
3.5.5.3 Boiler Operating Conditions—
The inspector should record pertinent boiler operating data to determine
boiler operating conditions during the Level III inspection. These values may
be used to determine if the boiler is operating at normal production levels
and to compare with the historic baseline data obtained during performance
tests. Major deviations from normal values should be evaluated with respect
to their effect on control equipment performance and emission levels.
Boiler operating data that should be measured during a performance test
or an inspection are plant-specific. Each boiler is usually custom-designed
and erected with a unique instrument and control system package. The level
of instrumentation is specified by the design engineer and purchaser (plant
engineering department). Based on the size of the boiler, its cost, and the
experience of the purchaser, the instrument package may range from an extremely
straightforward package to one that is quite complex. In general, a minimum
amount of instrumentation is necessary for safe operation of the boiler, and
this level of instrumentation will be present in all facilities. More complex
instrument packages can include an automated computer control system, which
allows the source to optimize combustion and increase the overall efficiency
of the operation.
Most critical boiler parameters are recorded on continuous strip charts
or circular chart recorders, 'and copies may be obtained after the stack test
(at the end of the day) to provide the necessary documentation. Most mills
require the boiler operator to record key parameters at set intervals on a
log sheet or in a log book. The log sheet is typically divided into the fol-
lowing general measurement areas: fuels, forced air, furnace drafts, gas
temperatures, feedwater, steam, and miscellaneous items (e.g., pulverizers).
Table 3-27 lists the items or conditions that must be recorded during
the stack test or Level III inspection. The list is based on a typical boiler
and would require adjustment for individual installations.
276
-------
TABLE 3-27. POWER BOILER OPERATING PARAMETERS TO BE RECORDED
DURING PERFORMANCE TESTS OR LEVEL III INSPECTIONS
Parameter
Variable
Units
Fuels
Forced air
Furnace drafts
Oil flow
Number of guns
Oil pressure
Oil temperature
Natural gas rate
Coal feed
Primary air flow
Primary air pressure
Primary air temperature
Secondary air flow
Secondary air pressure
Secondary air temperature
Overfire air flow
Overfire air pressure
Damper setting
Underftre air flow
Underfire air pressure
(each windbox)
Damper setting
Total air flow
Furnace
Boiler inlet
Boiler outlet
Economizer outlet
ID fan inlet
Control equipment inlet
gal/h
None
psig
103 ft3/h
ton/h
scf/min
in. H20
°F
scf/min
in. H20
°F
scf/min
in. H20
scf/mi n
in. H20
scf/min
in. H20
in. H20
in. H20
in. H20
in. H20
in. H20
(continued)
277
-------
TABLE 3-27 (continued)
Parameter
Variable
Units
Gas temperatures.
Feedwater
Steam
Miscellaneous
Boiler outlet
Economizer outlet
Evaporator outlet
ID fan outlet
Control equipment inlet
Flow
Pressure
Temperature
Flow
Drum pressure
Superheater temperature
Flue gas oxygen (boiler outlet)
Pulverizer motor current
Pulverizer speed
ID fan motor current
FD fan motor current
(l(r Ib/h)
psig
Or-
Ib/h)
psig
amps
rpm
amps
amps
Integrator readings also should be recorded at the beginning and ending
of each test run for the following parameters:
o Fuel flow (pounds or gallons)
o Steam flow (pounds)
o Steam used in soot blowing (pounds)
Based on data obtained from the log, steam tables, integrator readings,
and boiler design data, the following values should be calculated:
o Average steam flow for each run
o Average fuel fired for each test run
o Heat input (106 Btu/h) to the boiler for each test run for
each fuel fired (oil, natural gas, coal, wood)
278
-------
0
0
Average boiler output (106 Btu/h) for each test run
Boiler thermal efficiency (heat output/heat input)
for each test run
o Boiler excess air (percent)
3.5.5.4 Flue Gas Volume--
Since control equipment performance is affected by total gas volume, the
inspector should estimate the volume based on fuel firing rates, flue gas
oxygen content, and temperature. Most plants monitor flue gas oxygen at the
economizer or the furnace outlet for combustion control. Measurements during
the Level III should be made at the outlet of the control equipment to esti-
mate flue gas volume (except for scrubbers). The flue gas volume can be cal-
culated by using an F-factor method and the heat input rate to the boiler.
The inspector should be equipped with a portable temperature-measuring device
(e.g., a thermometer or thermocouple) and portable oxygen-measuring equipment
(e.g., a Fyrite oxygen analyzer or Orsat). The contribution of multiple fuels
may be determined if their percent contribution to the overall heat input rate
is known and the gas volumes of each fuel can be calculated. Because all mea-
surements and calculations are at the same condition, the volumes from combus-
tion of each fuel may be added to determine the total gas volume. The F-factor
can be calculated from the fuel analysis, but standard tables are available
with values for a variety of fuels. It should be noted that oxygen measure-
ments are made on a dry-basis and excess air corrections are based on a dry-
basis.
For fuels such as coal, oil, or natural gas, the equation is as follows:
Q = 10° Btu/min
20.9
d 20.9-%0,
w
(Tm + 460)
528
where
Q =
Fd =
'fay ~
gas volume in actual cubic feet per minute, acfm
dry F-factor for fuel, dscf/106 Btu
percent oxygen in flue gas
F = wet F-factor for fuel, wscf/10 Btu
279
-------
Tm = measured temperature of flue gas, °F
The heat input can be calculated if the fuel input rate and the heating value
of the fuel are known. For example, for coal:
106 Btu/min = Ib/min x 106 Btu/lb of coal
or oil:
106 Btu/min = gal oil/min x 106 Btu/gal
The factor of (FW - Fd) accounts for the extra gas volume due to moisture
produced during combustion plus typical moisture in the fuel. It must be
added after correction of the gas volume for excess air because the 02 mea-
surement is on a dry-basis.
For bark or wood firing, the equation becomes somewhat more complex be-
cause the moisture percent of these fuels is very high and the contribution
of the moisture from evaporation to the flue gas is significantly higher than
that produced by combustion. For wood or bark, an equation is as follows:
fi r 9n Q T (T_ + 460)
Q = 10° Btu/min Fd x 20 glg0 + (Moisture Correction) m528
This equation should be used when the values have the same definitions as above
and the moisture correction is
Moisture Correction =
fractional % H20 in fuel by weight x 21.41
(1- fractional % H20 in fuel) x 106 Btu/lb of dry fuel
Data on wood and bark usually give weight percent of moisture and heating
value per pound of dry fuel. The value of 21.41 is the conversion factor
(minus units) to convert 1 Ib of water to 1 Ib of water vapor volume (i.e.,
1 Ib H20 occupies 21.41 scf of volume).
For scrubbers, measurements should be made at the inlet of the scrubber
and then corrected for the decrease in temperature and increase in water vapor
through the adiabatic cooling of the gas stream.
When fuel measurements are not available, the boiler heat input rates
may be estimated if the steam conditions, feedwater conditions, and boiler ef-
ficiency are known. Boiler efficiency can be determined from boiler efficiency
280
-------
charts if temperature and excess air levels are known. The difference between
heat content of the steam and the feedwater represents an approximate heat
output of the boiler (these values may be determined from standard steam
tables). Once heat output and boiler efficiency have been calculated, heat
input to the boiler can be calculated as follows:
UQ . T__ll4. _ Heat Output
Heat Input - Boiler Efficiency
This value can then be used for the F-factor calculations.
3.5.5.5 Control Equipment Inspections—
ESP's--The ESP's applied to power boilers are very similar to those of
recovery boilers with respect to their layout and key parameters to be calcu-
lated. A complete discussion of the data to be taken, the calculations, and
the significance of these parameters were presented earlier. The major dif-
ference between ESP's on power boilers and recovery boilers is the possibility
of resistivity problems on the power boilers.
Low resistivities are generally not expected except on stoker-fired
boilers and boilers firing high-sulfur coal. The electrical signs of operat-
ing problems are high voltage, current, and power input levels that show the
appropriate patterns from the inlet to the outlet of the ESP. If reentrain-
ment problems are apparent, superficial velocities are normal, and applied
rapping forces are normal, then a low resistivity problem may exist. Checks
should be made to see if changes have occurred in temperature or fuel quality
that could cause a change in resistivity.
For ESP's not designed to handle high resistivities, a change in resis-
tivity to higher values can significantly impair performance. Higher resistivity
may be caused by a change in coal sulfur content, a change in other ash con-
stituents, or a change in the temperature. Whatever the cause, the performance
usually deteriorates. As ash becomes more tenacious and difficult to remove
from the plates, power levels decrease and sparking increases throughout the
ESP. In severe cases, virtually no normally expected increase in power or
current levels occurs from inlet to outlet. Secondary current levels may just
approach 50 percent of maximum secondary current ratings on the outlet fields
rather than the typical 85 to 100 percent levels. When fly ash resistivity
has increased to the point where sparking no longer occurs because of a strong
281
-------
back corona effect, currents will be very high at low voltages and performance
will be poor. The voltage-current curve will show distinct negative slopes,
indicating back-corona. This condition may be temporary, or it may persist
for some time if residual ash continues to cling to the plates.
Fabric Filters—The calculations and measurements that can be made on a
fabric filter are somewhat limited. Symptoms of operating and/or design
problems may appear during the inspection. For safety reasons, however, an
internal inspection of the fabric filter is not always possible without
shutting the boiler down. Internal inspections are usually most effective
for determining fabric filter problems.
The first parameter of interest is the opacity. Unless a condensable
plume is present, the average opacity should be low. After each cleaning
cycle, opacity will generally increase slightly as the bags are placed back
in service with the dust cake removed. Some seepage occurs until the filter-
ing action of the dust cake is restored. A significant increase in opacity
could indicate a pinhole leak in a given module or a row of bags. The length
of time required to restore* emissions to their previous levels is an indication
of the severity of the problem.
The pressure drop may give some indication of the severity of any holes
in the bags. The presence of a few pinhole leaks is unlikely to affect
the total pressure drop. Larger leaks, however, can lower the pressure drop,
and complete failure of bags can cause the pressure drop to approach 0.5 to
1 in.
An increase in pressure drop may be the result of increased gas flow
through the fabric filter or greater dust layer resistance to the gas flow.
An F-factor calculation should indicate any shifts in gas flow rate through
the fabric filter. If no apparent shift in gas volume has occurred, blind-
ing of the fabric due to failure of the cleaning mechanism or some other
related cause may be suspect. Increased gas volumes through the fabric filter
can also lead to gradual bag blinding caused by the deeper penetration of the
particulate into the fabric weave, which cleaning cannot remove. The result
is usually a gradually increasing pressure drop as the bags slowly become
blinded. Failure of the cleaning system or injection of moisture into the
fabric filter will usually cause rapid increases in pressure drop.
282
-------
If no instruments are available to determine pressure drop, static
pressure taps at the inlet and outlet of the fabric filter should provide a
basis for determining pressure drop. If instruments are available, they
should be checked to see if they are operating properly. Inlet pressure taps
tend to become plugged because of the dirty gas stream. The taps also should
be checked to be certain that they are drilled through the walls; occasionally,
this is forgotten during construction. Sometimes only one tap is connected.
If only the inlet tap is connected, readings may appear to be in the proper
range when they actually are quite erroneous.
Individual manometers or magnehelic gauges on each compartment generally
will not indicate problems within a specific compartment because the pressure
drop will equalize across compartments. Because the gas flow through
the compartments will follow the path of least resistance, flow rates through
the various compartments may be unequal, although pressure drops may be the
same.
The hopper discharge should be checked for plugged or damaged hoppers.
Screw conveyers, rotary airlocks, and other ash removal systems should operate
continuously or at least on a frequent cycle. PIuggage of the hopper can
allow the ash to build up well into the bags and cause the bags to be shut
off from the gas flow. This will increase the effective A/C ratio in the
fabric filter and the pressure drop.
The cleaning system should be checked for proper operation, and each
compartment or bag row should be cleaned. If the time between cleaning cycles
is too long, cleaning mechanism should be checked in the manual mode. Pulse-
jet systems should fire with a resounding thud, with compressed air pressures
of 90 to 120 psig. Reverse-air systems should isolate each compartment and
the reverse-air and dwell cycles should be sequenced to allow flexing and
release of the dust cake under gentle conditions (no "popping" of the bags).
A more definitive diagnosis of problems within a fabric filter requires
compartment isolation and internal access to the baghouse with the appropriate
safety equipment. This may be possible during scheduled outages or, if good
isolation and purge systems exist. Typical key points in an internal inspec-
tion include proper installation and tensioning of bags, the presence and
patterns of deposits on the "clean side" of the fabric filter, location and
283
-------
integrity of the baffle plate, apparent bag/hopper piuggage, moisture or oil
problems blinding the bag, and evidence of high temperatures in the fabric
filter.
Scrubbers—Because the venturi scrubbers applied to power boilers are
nearly identical to those applied to lime kilns, the reader is referred to
the section on Lime Kiln Scrubber Inspections for a discussion of items to
observe or calculate. There is a difference, however, in the operating pH of
these two scrubbers. Lime kiln scrubbers operate under alkaline conditions
and power boilers operate under acidic conditions. As a result of the acidic
conditions, the power boiler scrubber may have to be more carefully engineered
to prevent corrosion problems. Other problems with condensable matter, solids
buildups, and fine particle generation will be similar for a power boiler.
Multicyclones—Inspections of multicyclones are relatively limited
because of restricted access to equipment and the limited number of key operat-
ing parameters to be evaluated. Other than checking pressure drop across the
multicyclone, checking for proper-hopper discharge, and confirming that gas
flow rates are near nominal design levels, more detailed checks for proper
operation require internal access to the multicyclone. This will require
scheduling a visit during a boiler outage.
Multicyclone opacity levels usually provide less information about equip-
ment performance than the opacity levels for other pieces of control equipment.
The multicyclone is normally unable to collect the smaller light scattering
particles and, therefore a higher level of opacity is possible than with ESP's
or fabric filters. Because the multicyclone only collects the larger sized
particles, little or no observable shift in opacity may be noted even though
performance has decreased. As a result of a malfunction, however, the change
in opacity level may indicate a change in fuel or combustion characteristics
because these can affect the distribution of fine particles entering the
multicyclone.
Pressure drop across the multicyclone is generally a poor indicator of
performance and internal multicyclone conditions in the normal pressure drop
range (i.e., 2 to 4 in. H20). When very low or very high pressure drops are
encountered, it provides an indication that something is wrong inside the
multicyclone and that maintenance is required. Small shifts in pressure drop,
284
-------
(<0.5 in. H20) however, have little meaning in evaluating performance. Usually
higher pressure drops mean improved performance, but this is dependent on the
initial design parameters.
3.6 OTHER SOURCES
Several miscellaneous sources at a kraft pulp mill involve the movement
of material and/or the treatment of pulp. These sources include the bleach
plant and the raw material handling systems. The major emissions from the
bleach plant are chlorine gas and chlorine dioxide. The chlorine and chlorine
dioxide are hooded, vented, and passed through a packed bed scrubber using a
caustic solution. The major malfunctions at the bleach plant involve the bed
pluggage and channeling and reduction in water flow to the packed bed scrub-
ber. Because of the toxic nature of the emissions from the bleach plant the
inspection is generally limited to a review of process and control equipment
specifications. A checklist to aid in inspecting the bleach plant is provided
in Table 3-31 on page 296.
The major emissions from the raw material handling system are fugitive
particulate. In general, the fugitive emissions are controlled by hooding
and venting the emissions to a fabric filter. Most malfunctions in the raw
material handling area are the result of low capture velocity or fabric filter
failure. The inspections for the raw material handling area involve a Level
I or Level II Inspection. A checklist to aid in inspecting the raw material
handling system is provided in Table 3-32 on page 299.
3.6.1 Bleach Plant
3.6.1.1 Process Description--
Due to its nature, the brightness of kraft pulp is low. In order to use
the pulp for finer quality white paper, additional lignin must be removed.
The lignin may be removed through chlorination, oxidation, or reduction.
Chemicals used in chlorination bleaching are chlorine gas and chlorine dioxide.
Oxidation may be accomplished by use of hypochlorites, peroxides,
sodium chlorite, and peracetic acid. Reducing agents include sodium sulfide,
sodium bisulfide, sodium dithionite, zinc dithionite, and borohydrides. '
In the pulp mill, bleaching is accomplished in aqueous solutions under
controlled conditions. The pulp is bleached and extracted in successive steps.
After bleaching, the decomposed lignin is removed by alkaline washing.
285
-------
The following is a discussion of the bleach processes commonly used in
kraft mills.
Chlorination—Chlorination is accomplished through the use of chlorine
gas or chlorine dioxide that is injected into the stock solution and allowed
to react in a chlorination tower. The pH of the solution must be maintained
below 2. The retention time is generally between 20 and 60 minutes at a
stock concentration of 3 to 4 percent. Usage rates are of the order of 3 to
8 percent of pulp weight.2 Extraction is accomplished by alkali washing in
successive stages.
Alkali extraction—Alkali is used to remove residual bleaching agents
and freed lignin. Bleach products are soluble in alkaline solutions at a pH
of 10 to 11. The alkaline solution also dissolves resinous materials, pentosons,
and carbohydrates. Typical chemicals used are hydroxides, sulfides of alkali
metals, and calcium hydroxide.
Extraction is generally accomplished at stock concentrations of 3 to 16
percent and require 60 to 120 minutes for completion. The most common extrac-
tion chemical is sodium hydroxide. Solution strength is between 0.5 and 3.0
percent on weight of pulp (air dried unbleached tons).2
Hypochiorites—Hypochiorites may be used in second or third stage bleach
steps after chlorination. Hypochlorite bleaches are not specific to lignin
and react with the pulp cellulose. Bleaching is typically carried out at a
pH of 6 to 9.5. Stock concentration ranges from 12 to 16 percent. Reten-
tion time may be 2% hours for the first stage and 4 hours for the second
2
stage. The solution pH may be buffered by use of caustic soda, sodium
carbonate, milk of lime, or lime.
Chlorine dioxide—Chlorine dioxide may not be shipped and must be
generated at the mill site. It is toxic, corrossive, and explosive. Above
50°C it experiences explosive decomposition. Typical application rates are
0.3 to 1.2 percent on weight of pulp at 70°C and at a pH between 3 and 7.2
Peroxides—Peroxides are generally used as a final bleach step. Appli-
cation rates are 0.1. to 0.25 percent on weight of pulp. Pulp concentrations
are between 12 and 15 percent. Retention times may be as hfgh as 5 hours at
80 to 85°C and at a pH of'9 to 11.5.
286
-------
Figure 3-97 shows a typical three-stage bleach line used for sulfate
pulp. The line consists of a chlorination tower (Clg). chlorination washer,
caustic extraction, caustic washer, hypochlorite tower, and hypochlorite
washer.92'96
Bleach towers are either up-flow or down-flow. The height and diameter
are specified to provide the required retention time. Vacuum washers are
frequently used and are generally covered to reduce chlorine lost to the work
93 97
area. A vacuum washer is shown in Figure 3-98. '
Bleach stages and process are generally identified by use of a shorthand
notation with a series of letters. Table 3-28 lists the most common de-
signation for each chemical step. A slash (/) between steps indicates
successive additions of the agents without washing. Agents placed in paren-
thesis ( ) indicate simultaneous application of agents. Table 3-29 is a
2
list of the most common bleaching sequences used for sulfate pulps.
TABLE 3-28. COMMON LETTER DESIGNATIONS USED FOR BLEACH AGENTS2
A Acid treatment
C Chlorination
D Chlorine dioxide
E Alkaline extraction
H Hypochlorite
HS Dithionite
P Peroxide
PA Peracetic acid
W Water soak
( ) Simultaneous addition
/ Successive addition without washing
287
-------
TABLE 3-29. BLEACHING SEQUENCES FOR SULFATE PULP
*CED
*CEDED
*CEHH
*CEHD
CEDD
C/HEDD
*CEH
*CEHEH
CC/HEHH
C(EH)HEH
CEHHP
CC/HPH
CH(EH)D
CEH(EP)H
*CEHDP
CEHDED
*CEHDP
*CEHEDP
CC/HEHEH
CEHEHH
CEHEHD
CEHHDED
CC/HED/H
CEHCHDED
*HCEH
*CHEH
*CHED
Bleaching sequences for hardwood sulfate pulp.
288
-------
FRESH WATER
00
LEGEND
-A- REMOTE OPERATED VflLVE
GATE VALVE
CHECK VALVE
-©- FLOW INDICATOR OR CONTROLLER
i .% AIR DRY CONSISTENCY
*3 CHLOROMIX
TOWER
Figure 3-97. Flow diagram of a three-stage bleach plant: CEH.92
Source: Reproduced with permission of Joint Textbook Committee of the Paper Industry.
-------
93
Figure 3-98. Cutaway of a vacuum washer with short drop leg.
Source: Reproduced with permission of Sandy Hill Corporation.
290
-------
The bleaching sequence and chemical type determine the brightness of the
pulp. Table 3-30 shows the most common sequences and the brightness attain-
able with each sequence.
. TABLE 3-30. BLEACHING SEQUENCES FOR HARDWOOD SULFATE PULP2
Brightness
75
75 to 80
80 to 85
85 to 90
90 plus
Sequences
CEH
CEHH,
CHEH,
CEHD,
CEHEDP
Chlorine dioxide manufacture--Because of
CED, HCEH
CEHEH, CED, CEHDP, CEHD
CHED, CEDED, CEHDP
the explosive nature of chlorin<
dioxide, it cannot be safely liquified or transported. Mills that use the
material produce the gas on site by one of four commercial methods: Solvay,
Mathieson, Rapson, and Day Kesting. The Solvay and Mathieson are the most
common processes.
The Solvay process reacts sodium chlorate, sulfuric acid, methanol, and
air under pressure to form chlorine dioxide, sodium sulfate, and formic acid
(Figure 3-99). The gas is adsorbed in a packed-bed scrubber and placed
in storage. In the Mathieson process, sodium chlorate, sulfur dioxide, and
sulfuric acid are reacted to form chlorine dioxide and sodium sulfate. The
gas is adsorbed in a dual-stage packed-bed scrubber and placed in storage.
Chlorine cells—Chlorine is a greenish-yellow gas that is nonexplosive,
nonflammable, very toxic, and a Tung irritant. The gas is heavier than air
(2.5 times more dense than air) and may collect in pockets where air circu-
lation is poor.
Chlorine is received at the mill as a dry liquid and is converted to gas
through vaporization for use in bleach towers. When placed in water, the gas
forms hypochloric and hypochlorous acids. The acids are extremely corrosive
and at 150°C rapidly attack carbon steel.
291
-------
COOLER |—I
COLD
COLD
NO. 1
REACTOR
JACKET
WATER
HOT
H,0
X,
N
R
^
CHILLED WATER
ABSORBER
NO. 2
REACTOR
H,0
AIR
TO RECOVERY
CIOSOLUTION
STORAGE
TO PROCESS
Figure 3-99. Chlorine dioxide generating system.
Source: Reproduced with permission of Taylor Instrument Company.
292
-------
Many plants produce chlorine on site for use in bleaching. The most
common process used for manufacture is the electrolysis of a sodium chloride
solution. The reaction is expressed as follows:
2 NAC1 + 2 H20 + electric current -»• C12 + H2 + 2 NaOH
About 12 plants manufacture chlorine by this process. The two major
electrolysis systems are diaphragm and mercury cell.
In the diaphram plant, a solution of sodium chloride is passed through
the cell and unreacted salt is resaturated and returned to the reactor. Spent
liquor from the cell is concentrated in multiple-effect evaporations and
cooled to allow crystallization of the salt. Caustic is separated and the
salt filtered and washed. Chlorine released in the reaction is cooled, dryed
with sulfuric acid, compressed and liquified, and placed in storage. Chlorine
from the reactor has a purity of 97 to 98 percent before cleanup. Hydrogen
is also recovered from the cell and either oxidized or collected for sale.
In the mercury cell, the iron cathode is replaced by a thin layer of
mercury. Metallic sodium that is formed in the reaction is adsorbed on the
mercury. Analgam is passed through a packed tower of graphite where it reacts
with water to form sodium hydroxide. The freed mercury is returned to the
cell. The hydrogen that is produced is cooled and scrubbed for use as fuel
for the chemical reaction. The system must be maintained under positive
pressure to prevent air inleakage.
Hypochlorite manufacture—Many plants produce hypochlorite bleaches on-
site by either batch or continuous process. Calcium hypochlorite is manu-
factured when calcium hydroxide reacts with chlorine. Sodium hypochlorite is
formed by the reaction of sodium hydroxide with chlorine. Sodium chloride is
98
a waste reaction product and therefore must be removed.
3.6.1.2 Sources of Emission and Control--
Chlorine gas and chlorine dioxide represent a potential danger in pulp
mills. The gases are toxic, heavier than air, corrosive, and lung irritants.
Chlorine dioxide gas is also explosive. In the bleach process, residual
chlorine must be removed through washing in vacuum washers. The gases in
these systems contain traces of chlorine and chlorine dioxide. The bleach
tower, washers, and seal boxes are hooded and vented to remove these gases
293
-------
from the work place. The rate of ventilation varies depending on the tightness
of the system and number of sources tied to each system.
Control is accomplished by passing the gas stream through a packed bed
scrubber using a caustic solution. Typical superficial velocities through
the bed are 8 to 10 ft/s, and the liquor-to-gas ratios are 3 and 5 gal/1000
acfm. The scrubbers are generally made of fiber reenforced plastics (FRP)
and are designed for low temperature. Under normal conditions, the scrubbers
can obtain removal efficiencies of 98 to 99 percent. Final concentration of
ClOg in the gas stream is a function of water temperature and C102 concen-
tration (g/1) in the water.12'99'100
3.6.1.3 Malfunctions— .
Bleach plant emissions are controlled by hooding, ventilation systems,
and packed-bed scrubbers. Malfunctions to the hooding and ventilation systems
normally do not occur because these sources are in work areas. Loss of hood
velocity and/or ventilation rates are usually quickly corrected because of
constant monitoring of the work atmosphere required for worker protection.
Malfunctions in this area are therefore generally associated with the packed-bed
scrubber. The most common problems involve bed pluggage and channeling, and
reduction in water flow. Channeling is typically .caused by bed pluggage as
a result of pulp carry-over into the ventilation system or shifting of the
bed packing. Channeling can also occur as a result of poor water distribution
within the packed bed. As a result of the channeling, a disproportionate
amount of gas is passed through a reduced bed cross section. The increased
velocity results in water entrainment and reduced collection efficiency.
Liquor carryover can also occur at high liquor-to-gas ratios as a result of
bed flooding.
3.6.1.4 Inspection of Bleach Plants--
Inspection in the bleach plant is limited to process and control equip-
ment specifications (Level I or Level II). Because of the toxic nature of the
emissions, the inspector should avoid breathing the contaminated gas streams
and use appropriate respirators while working in this area. Plant safety re-
gulations must be strictly obeyed. Failure to observe safety requirements
can result in damage to eyes, skin, and lung tissue.
294
-------
The inspector should document the physical arrangement of the bleach line
(towers, washers, etc.) and the chemicals used in each step. Rate of applica-
tion of each bleach or extraction agent should be documented based on weight
of pulp. The pulp tonnage should also be documented. Points of ventilation
should be noted, and the rate of ventilation should be determined.
The design specifications of the packed-bed scrubber should be obtained.
Specific design variables including cross-sectional area, liquor flow rate,
superficial velocity, and gas volume should be calculated.
Specific malfunctions such as pluggage or bed flooding should be deter-
mined based on the presence of mist carry-over on high scrubber pressure drop.
High chlorine penetration can be observed as a bluish-green gas emitted from
the scrubber stack. Table 3-31 is a checklist for use in the inspection of
the bleach plant.
3.6.2 Raw Material Handling Systems
3.6.2.1 Process Description— • ] . -
Many sources in the mill generate fugitive emissions as a result of
material handling. These sources are not unique to specific mills, but the
number and magnitude of sources vary from mill to mill. These sources include
pebble lime silos, hot lime conveying and storage, salt cake unloading and
storage, starch/clay unloading and storage, salt unloading and storage, and
pigment unloading and storage.
The rate of emission and process weight of these sources are highly
variable. The material can be transferred through the use of bucket elevators,
drag chain conveyors, screw conveyors, or a pneumatic conveying system. The
sources may either be contained or open depending on mill design and age.
3.6.2.2 Sources of Emissions and Control--
Hot lime conveyor—Hot lime from the lime'kiln burner end hood must be
cooled before being placed in storage or returned to the slaker. The lime is
typically cooled on an open drag chain conveyor. This conveyor, because of the
presence of dusty lime, thermal drafts, and chain movement, is a source of
fugitive particulate. Mills have controlled the emissions from this area by
containment of the conveyor and hooding and venting to a fabric filter. The
filter usually controls the hot lime elevator and silos. The filter must be
designed to handle high temperatures as well as abrasive material. Typical
295
-------
Stage No.
TABLE 3-31. INSPECTION CHECKLIST FOR USE IN BLEACH PLANT
Chemical rate on weight of pulp
Ventilation points
Control type
Gas volume
Bed cross section
Superficial velocity
Liquor flow rate
Liquor-to-gas ratio _
Pressure drop
Mist carry-over
Visible plume
Plume color
acfm
ft
ft/s
gpm
gal/1000 acfm
in H20
yes
yes
yes
no
no
no
296
-------
bag materials are Nomex or polyester. Systems usually are of the pulse-jet
2
design and operate at air-to-cloth ratios of 4 to 6 acfm/ft .
Raw material silos—Mills receive a number of raw materials for use in
the pulp and paper departments. These materials include salt cake, lime,
starch, resin, and pigments. The materials are received in a dry condition
by either truck or rail and are generally placed in silos. Silos may also be
used for day storage within the process area. Pneumatic conveyors are generally
used to transport the products. Conveying gases that are vented from the silos
can contain considerable amounts of particulates. Uncontrolled emission rates
can be as high as 20 gr/acfm of vented gas. The amount of gas vented can be
highly variable depending on process weight and silo size. Most systems are
between 300 and 500 acfm. The most common control device is the fabric filter,
which is usually mounted on top of the receiving silo. The vented gases are
generally between 70 to 100°F and are dry. Typical air-to-cloth ratios are
2
between 4 and 6 acfm/ft .
3.6.2.3 Malfunctions--
Most malfunctions in the material handling area occur as a result of low
capture velocity or fabric filter failure. Capture velocity may be reduced
as a result of high-filter pressure drop (reduced gas volume) or infiltration
of air into the ventilation system between the hood and fabric filter. In
this case, the filter is operating at design gas volume, but the source capture
velocity is very low.
Filter failure occurs as a result of the failure of the filter cleaning
system, fabric structure, or dust discharge system. Cleaning system failures
may occur as the result of failures in: pulse diaphragms, timers, solenoids,
or an air compressor. Water in the compressed-air system commonly results in
diaphragm failure and problems with dust cake release. Failure to properly
clean the fabric results in high filter pressure drop and reduced ventilation
volume.
Fabric failure occurs as a result of fabric abrasion, chemical damage
or temperature excursion. Most fabric failures occur as pin holes in the
fabric structure. Holes result from abrasion as the fiber strength is re-
duced. Particle penetration occurs though pinholes as the gas stream passes
through the orifice. As a result of the pressure drop across the orifice, the
297
-------
velocity through the opening is very high. The loss of one bag in the filter
can reduce filter efficiency by several percent.
Failure of the dust discharge system results in buildup in the hoppers
that eventually causes complete failure of the system. Dust discharge failure
occurs as a result of component failure such as screw conveyors or air locks.
Inleakage of water through flanges, welds, and door gaskets causes hopper
buildup and ptuggage.
3.6.2.4 Inspection—
During a Level III inspection the inspector should determine the process
weight of the product being transferred. Depending on the ventilation system
design, the inspector should determine the system flow rate and calculate the
collector air-to-cloth ratio. If the process is a high-temperature source
(hot lime conveyor), the inlet gas temperature should be measured.
Operation of the filter cleaning system should be checked by listening
for the operation of the compressed-air pulse-cleaning system (pulse jet col-
lectors). Visible emission after pulse firing indicates that a pin hole may
exist in the fabric for a given row of bags. The rotary air lock and screw
conveyors should be checked to determine if any hopper pluggage is occurring.
If the system can be shut down, the inspector should open and inspect
the clean side of the collector tube sheet to determine if penetration is
occurring. Table 3-32 is a checklist for use in a Level III inspection of
material handling systems using fabric filters.
298
-------
TABLE 3-32. INSPECTION CHECKLIST FOR MATERIAL HANDLING SYSTEMS
Source name
Product
Transfer rate
tons/h
Fabric filter type
Cloth area
Gas volume
Inlet gas temp
Air-to-cloth ratio
Pressure drop
Cleaning system operating
ft2
acfm
°F
acfm/f t2
in. HoO
yes no
Rows not cleaning , , ,
Internal inspection
Clean side deposits
Pin holes
Oil /water or bags
Hopper discharge
Type screw
Functioning yes
yes
yes
yes
air lock
no
no
no
no
Hood capture
Visible emissions
Hood condition
yes
Duct work
Corrosion
Inleakage
yes
yes
good
no
poor
no
no
299
-------
REFERENCES FOR SECTION 3
1. U.S. Environmental Protection Agency. Atmospheric Emissions From the
Pulp and Paper Manufacturing Industry. EPA-450/1-73-002, September 1973.
2. Joint Textbook Committee of the Paper Industry. Pulp and Paper Manu-
facture, Vol. I. The Pulping of Wood, 1969.
3. Carthage Machine Company, Inc.
4. Koehring-Waterous Ltd.
5. Goderhamm Maching Mfg. Co.
6. The W. S. Tyler Company.
t
7. ROTEX, Inc.
8. Elliott, R. D., and W. H. deMontmorency. The Transportation of Pulp-Wood
Chips by Pipeline, Pulp Paper Res. Inst. Can. WR Ind. 144, 1963.
9. Hawks, R. In-house Engineering Data. PEDCo Environmental, Inc.
10. Screw Conveyor Corporation.
11. Link Belt Limited.
12. U.S. Environmental Protection Agency. Technology Transfer. Environmental
Pollution Control, Pulp and Paper Industry, Part I, Air. EPA-625/7-76-001,
October 1976.
13. Kock, P. A. Treating Kraft Digester Waste Gases. M. S. Thesis, Chemical
Engineering Department, Helsinki Technical University, Finland.
September 12, 1972.
14. Miller, J. T. Methods for Managing Batch Digesters. Manager Pulping
System Division, Rader Company, Inc.
15. Vinnis, V., and T. Kinnula. A Method to Control the Turpentine Recovery
Process of Batch Kraft Pulp Digesters. TAPPI Engineering Conference,
1981.
16. Hrulfiord, B. F., and D. F. Wilson. Turpentine Concentrations in Kraft
Mill Condensate Streams. Pulp and Paper Mag. of Can. Vol. 74, No. 6,
June 1973.
300
-------
17. Kelski, R. Kraft Mill Odor Abatement by Condensate Stripping and Waste
Gas Incineration. M. S. Thesis, Wood Industry Department, Helsinki
Technical University, Finland. September 18, 1969.
18. Ellerbe, R. W. Why, Where, and How U.S. Mills Recover Tall Oil Soap.
Paper Trade Journal, June 25, 1973,
19. Propst, M. Improved Techniques for Soap Recovery. . TAPPI Engineering
Conference 1981.
20. Edwards, L., and R. Baldus. Evaluation and Design of Multiple Effect
Evaporation Systems for Kraft Black Liquor.
21. Lankenau, H. G., and J. T. Badyrka. Multiple Effect Evaporators - Problem
and Troubleshooting. TAPPI Engineering Conference 1981.
22. Nylander, G. Report on Forest Industry Waste Waters. Svensk Papperstidning
67(15):565-572, August 1964.
23. Leornados, G., D. Kendall, and N. Barnard. Odor Threshold Determinations
of 53 Odorant Chemicals. JAPCA 19:91-95, February 1969.
24. Wilby, F. V. Variation in Recognition Odor Threshold of a Panel. JAPCA
19:96-100, February 1969.
25. Backstrom, B., H. Hellstrom, and F. Kommonen. Purification of Malodorous
Sulfur Containing Condensates from Turpentine Separation, Digester Blow
and Spent Liquor Evaporation at the Oy Kaukas Ab, Kraft Mill. Paperi
ja Puu 52(3):113-120, 1970.
26. Morgan, I. P., and F. E. Murray. A Comparison of Air and Steam Stripping
as Methods to Reduce Kraft Pulp Mill Odor and Toxicity from Contaminated
Condensate. Pulp and Paper Magazine of Can. 73(5):62-66, May 1972.
27. Papic, M. M., A. D. Mclntyre, and J. G. Dunsmore. Stripping of H2S and
CH3SH from Aqueous Solutions. Pulp and Paper Magazine of Can. Vol. 74,
No. 10, October 1973.
28. Rowbottom, B., and G. Wheeler. Stripping-Incineration System Cuts TRS
Emissions at Cornwall Pulp and Paper Magazine of Can. Vol. 76, No. 2,
February 1975.
29. Collins, T. T. The Oxidation of Sulfate Black Liquor and Related
Problems. TAPPI 38:172A-175A, August 1955.
30. Trobeck, K. G. The BT System for Soda and Heat Recovery in Sulfate Pulp
Mills. Paper Trade Journal 133(15):40-48, April 20, 1960.
31. Bealkowsky, H. W., and C. G. Dellaas. Stabilization of Douglas Fir Kraft
Black Liquor. Paper Mill News 74(35):14-22, September 1, 1951.
301
-------
32. Waltker, J. E., and H. P. Amberg. Odor Control in the Kraft Pulp
Industry. Chemical Engineering Progress 66:73-80, March 1970.
33. Blosser, R. 0., and H. B. H. Cooper. Survey of Black Liquor Oxidation
Practices in the Kraft Industry. NCASI Atmospheric Pollution Technical
Bulletin No. 39. National Council of the Paper Industry for Air and
Stream Improvement, Inc., New York, New York, December 1968.
34. Padfield, D. H. Control of Odor from Recovery Units by Direct Contact
Evaporative Scrubbers with Oxidized Black Liquor. TAPPI 56:83-86,
January 1973.
35. Van Donkelaar, A. Air Quality Controls in a Bleached Kraft Mill. Pulp
and Paper Magazine of Can. 69(18):69-73, September 20, 1968.
36. Shah, I. S., and W. D. Stephenson. Weak Black Liquor Oxidation: Its
Operation and Performance. TAPPI 51:87A-94A, September 1968.
37. Sarkonen, K. V, B. F. Hrutfiord, L. N. Johonson, and H. S. Gardner.
Kraft Odor. TAPPI, Vol. 53, No. 5, May 1970.
38. Perry, J. H. (ed.). Chemical Engineers Handbook, 3rd Edition.
York, McGraw Hill Book Company, pp. 1585-1586, 1950.
New
39. Ghisoni, P. Elimination of Odors in a Sulfate Pulp Mill. TAPPI 37:
201-205, May 1955.
40. The Venemark-Design White Liquor Scrubber. Swedish Patent 226 789,
Stockholm, Sweden.
41. Morrison, J. L. Collection and Combustion of Noncondensable Digester
and Evaporator Gases. TAPPI, Vol. 52, No. 12, December 1969.
42. Martin, G. C. Fiber Carryover With Blow Tank Exhaust. TAPPI, Vol. 52,
No. 12, December 1969.
43. Douglass, I. B., M. Lee, R. L. Weichman, and L. Price. Sources of Odor
in the Kraft Process. TAPPI, Vol. 52, No. 9,^September 1969.
44. Berry, L. R. Black Liquor Scaling in Multiple Effect Evaporators. TAPPI,
Vol. 49, No. 4, April 1966.
45. Blosser, R. 0., and H. B. H. Cooper. Current Practices in Thermal
Oxidation of Noncondensable Gases in the Kraft Industry. Atmospheric
Pollution Technical Bulletin No. 34, NCASI, New York, New York, 1967.
46. Babcock and Wilcox. -Steam/Its Generation and Use. 1978.
47. Passinen, K. Chemical Composition of Spent Liquors. Proceedings of the
Symposium on Recovery of Pulping Chemicals. Helsinki, Finland, 1968.
302
-------
48. Control of Atmospheric Emissions in the Wood Pulping Industry. U.S.
Department of Health, Education, and Welfare. Contract Mo. CPA 22-
69-18. March 1970.
49. Rydholm, S. A. Pulping Processes. Intersciences Publishers, New York
1965. •
50. Personal Communication. J. Blue, Babcock and Wilcox Company.
51. Thoen, G. N., G. G. DeHaas, R. G. Tallent, and A. S. Davis. Effect of
Combustion Variables on the Release of Odorous Compounds From a Kraft
Recovery Furnace. TAPPI, Vol. 51, No. 8, August 1968.
52. Clement, J. L., J. H. Caulter, and S. Suda. B&W Kraft Recovery Unit
Performance Calculations. TAPPI, Vol. 46, No. 2, February 1963.
53. Borg, A., A. Teder, and Bjorn Warnquist. Inside a Kraft Recovery Furnace •
Studies on the Origins of Sulfur and Sodium Emission. TAPPI Environmental
Conference, 1973.
54. Bhada, R. K., H. B. Lange, and H. P. Markant. Air Pollution From Kraft
Recovery Units - The Effect of Operational Variables. TAPPI Environmental
Conference, 1972.
55. Bauer, F. W., and R. M. Dorland. Canadian Journal of Technology 32:91,
1954.
56. Teller, A.-J., and H. R. Amberg. Considerations in the Design for TRS
and -Particulate Recovery from Effluents of Kraft Recovery Furnaces. TAPPI
Environmental Conference.
57. Lange, H. B,, D. P. Pierce, and J. W. Kisner. Emissions From a Kraft
Recovery Boiler - The Effects of Operational Variables. TAPPI, Vol. 57,
No. 7, July 1974.
58. Hawks, R., and G. Saunders. Unpublished Data. PEDCo Environmental, Inc.
59. Lang, C. J., G. G. DeHass, J. V. Gommis, and W. Nelson. Recovery Furnace
Operating Parameter Effects on S02 Emissions. TAPPI 56:115, June 1973.
60. Unpublished data, Kopper's Corporation.
61. Chamberlain, B., E. Lafkrantz, A. Smith, and R. Wostradowski. Eliminate
Recovery Furnace H2S Emissions by Controlling CO. Pulp and Paper Magazine
of Canada, Vol. 79, No. 2, February 1978.
62. Szabo, M. F., and Y. M. Shah. Inspection Manual for Evaluation of
Electrostatic Precipitator Performance. EPA-340/1-79-007, March 1981.
63. Katz, J. The Art of Electrostatic Precipitation. Precipitator Tech-
nology, Inc. 1981.
303
-------
64. Hawks, R. Unpublished data. PEDCo Environmental, Inc.
65. Kopper's Corporation, Baltimore, Maryland.
66. PEDCo Environmental, Inc. Identification of Parameters That Affect the
Particulate Emissions From Recovery Boilers. June 1982.
67. Personal communication. Kopper's Corporation, Baltimore, Maryland.
68. White, H. J. Electrostatic Precipitation of Flyash, Part I. Journal
of the Air Pollution Control Association, January 1977.
69. LoCicero, P. M., and P. E. Sjolseth. Operating Experiences With the Ace
Recovery Furnace Odor Control System. TAPPI Environmental Conference.
70. Weinaug, R. J. Jr. Early Experiences With a B&W Low Odor Recovery
System. TAPPI Environmental Conference, 1972.
71. Hawks', R. Unpublished data. PEDCo Environmental, Inc.
72. Saunders, G. Unpublished data. PEDCo Environmental, Inc.
73. Personal communication. Buell Envirotech, Inc.
74. A Manual for the Use of Electrostatic Precipitators to Collect Flyash
Particles. EPA-600/8-80-025, May 1980.
75. Henderson, J. S. Final Survey Results for Noncontact Recovery Boiler
Electrostatic Precipitators. J. E. Sirrine Co. TAPPI, Vol. 63, No. 12,
December 1980.
76. Gooch, V. P. Low Temperature Corrosion by Sulfuric Acid in Power Plant
Systems. Southern Research ESP Symposium, February 1971.
77. Rylands, J. R., and J. R. Jenkinson. The Acid Dew Point. Journal of
Institute of Fuel, June 1974.
78. Stockman, L., and A. Tansen. Gvensk Paperstidn 62:907-914 (1959)
Abstr. Bull. Inst. Paper Chem., 30;1164-1165 (1960). The Paper Industry,
p. 215, June 1960.
79. Zarfoss, J. R. Clean Air From Paper Mill Recovery Boilers Without
Corrosion. Environmental Elements Corporation, Baltimore, Maryland.
National Association of Corrosion Engineers, IGCI. Atlanta, Georgia,
1976.
80. Hawks, R. Unpublished data. PEDCo Environmental, Inc.
81. Personal communication. J. Blue, Babcock and Wilcox.
82. Timmerman, J. Physio-Chemical Constant of Primary Systems in Concentrated
Systems, Vol. 3, Intersciences Publishers, New York, 1960.
304
-------
83. Sallack, J. A. An Investigation of Explosions in the Sode Smelt Dissolving
Operation. Canadian Pulp and Paper Association, Technical Section,
June 1955.
84. Hawks, R. Unpublished data. PEDCo Environmental, Inc.
•
85. Campbell, A. J. Factors Affecting White Liquor Quality: Green Liquor
Concentration, Drugs Concentration and Lime Dosage. Pulp and Paper
Magazine of Canada. Vol. 82, No. 4, April 1981.
86. Kramm, D. J. Selection and Use of the Rotary Lime Kiln and Its
Auxilliaries - 11. Paper Trade Journal, August 21, 1972.
87. Dorr-Oliver, Inc.
88. Blosser, R. 0., A. L. Caron, R. P.. Fisher, M. E. Franklin, W. J.
Gillespie. Factors Affecting TRS Emissions From Lime, Kilns. TAPPI
Environmental Conference.
89. Hawks, R. Unpublished data. PEDCo Environmental, Inc.
90. Schwieger, B. Power Magazine, Vol. 121, No. 2, February 1977.
91. Burback, et al. Combustion Engineering, Power Magazine. December 1977.
92. Three Stage Bleach Plant, Improved Machinery Co.
93. Vacuum Washer. Sandy Hill Corporation.
94. Solvay Chlorine Dioxide Generating System. Taylor Instrument Companies
and Allied Chemical Company.
95. Hawks, R. Unpublished data. PEDCo Environmental, Inc.
96. Rapon, W. H., C. B. Anderson, and D. W. Reeve. The Effluent-Free
Bleached Kraft Mill, Part VI, Substantial Substitution of C102 for C12
in the First Stage of Bleaching. TAPPI Alkaline Pulping Conference
1975.
97. Gall, R. J., H. D. Partridge, D. J. Josyka, and G. R. Roseman. Sequential
Chlorination - Its Impact on the Environment. TAPPI Environmental
Conference, 1975.
98. Reeve, D. W., and W. H. Rapson. The Recovery of Sodium Chloride From
Bleached Kraft Pulp Mills. Pulp and Paper Magazine of Canada, Vol. 71,
No. 13, July 3, 1970.
99. Ekono Oy, Helsinki, Finland.
100. Haller, I. F., and W. W. Northgraves, TAPPI 38:199, April 1955.
305
-------
-------
SECTION 4
COMPLIANCE DETERMINATIONS
Historically, the level of inspections practiced by State and local
agencies consisted of visible emission evaluations and a quick walkthrough
of the plant. The purpose of these inspections was to provide a quick deter-
mination of compliance that could be easily documented and defended, and
that would require a minimum of resources. This level of inspection, previously
referred to as Level I or II, was effective in documenting major occurrences
of noncompliance especially with respect to visible emission standards.
In general, these inspections are effective screening tools in determining
potential visible emission violations or potential noncompliance with such
permit stipulations as firing rate, process rate, or fuel characteristics
(sulfur, ash, etc.). Most agencies use the visible emission standard in lieu
of requiring a stack test to determine compliance with an applicable particulate
emission standard. As a result of this practice, many sources were unofficially
allowed to operate at particulate emissions levels above the applicable regu-
latory limit (i.e., up to the point of violation of visible emission standards).
In addition, visible emission observations were rarely made when stack tests
were conducted to determine compliance with the particulate standards.
In many cases, certification stack tests are conducted under optimum
process and control equipment conditions. These conditions often do not re-
flect normal day-to-day operating practices or conditions. A review of stack
tests conducted on recovery boilers indicates a serious -deficiency in recording
process or control equipment parameters that would allow determination of re-
presentative conditions. For these and other reasons, it is apparent that
the use of visible emissions and annual certification stack tests are not an
effective method of determining the compliance status of many sources in a
kraft pulp mill.
Although the above methods ensure a certain level of compliance and
maintenance of control devices at least on a yearly basis, many operating
307
-------
and maintenance problems occur between inspections and stack tests that can
result in serious noncompliance and which have an adverse impact on the am-
bient air quality. It has been shown that the application of a more compre-
hensive evaluation of process and control equipment can document serious
operation and maintenance problems that in many cases have not been previously
2
noted by plant personnel.
The more comprehensive inspection technique is referred to as a Level III
Inspection. This level of inspection helps to ensure continuous compliance
and is based on the application of basic engineering logic to process equip-
ment and control equipment operation. Level III Inspections require a basic
understanding of the process and control equipment variables or parameters
that influence emissions and a thorough understanding of the factors that
influence control equipment performance. The inspector is asked to document
a number of operating parameters or variables and compare changes in these
parameters with a known reference point. This is very similar to the procedures
that are used in evaluating visible emissions. For example, an increase in
opacity above a given standard can be an indication of an increase in particu-
late emissions.
Level III Inspections typically rely on more than a single indicator
of performance unless the emission rate is dominated by a single variable.
Major changes in ESP power input or major increases in ESP gas volume are
considered strong indicators of ESP performance. In most applications a
number of parameters are. used to support the determination of compliance or
noncompliance. When the parameters are contradictory, a stack test is re-
quired to certify compliance. If a stack test is conducted, the inspector
must be assured that the test is conducted under the identical conditions
(process and control equipment) that were observed at the time of noncompliance.
The use of comprehensive inspections requires that the inspector have an
established baseline for each parameter during a period of known compliance.
These parameters are generally established during a performance stack test
or are based on specific design conditions. Once the baseline is established,
the inspector can return to the plant at a later date and determine the degree
of continuous compliance by evaluating certain parameters.
To ensure complete and accurate documentation of parameters or variables
during the stack test, the inspector should observe stack tests on all sources
308
-------
within his inspection jurisdiction. Although the primary purpose for
observing the emission test is to verify the representativeness of the test
and that acceptable testing procedures are being followed, the process opera-
tion during the test is of critical interest to the inspector from several
aspects.
During the initial compliance test, the inspector usually can determine
the range of process and control equipment parameters that the plant operator
and control equipment supplier consider optimum to achieve compliance with
the applicable emission standards. This information is useful not only for
establishing representative operating conditions during a specific test, but
is also useful in selecting or evaluating operating permit conditions and
in assisting the inspector in evaluating future performance. The overall
process of establishing a benchmark for the process operation is commonly
referred to as "baselining."
4.1 ESTABLISHING A BASELINE
Establishing a baseline involves documenting all pertinent operating
parameters as they relate to the emission characteristics of the source.
This includes both the process and control equipment parameters. The base-
line provides a fixed point of operation or a narrow range of operating para-
meters against which other determinations can be made. The concurrent emission
test provides a documented emission rate(s) that may be correlated with process
and control equipment operating characteristics derived during the test. The
baseline test is useful in conducting subsequent routine inspections.
The baseline may be used for several purposes. First, for existing
sources, baseline values may be obtained prior to a stack test to assist in
establishing representative operating conditions. The normal range of values
may be recorded during a period prior to a test, and these values may be speci-
fied in a testing protocol to establish representative conditions or used as
a starting point in negotiating the testing protocol with the plant. Compari-
son of documented compliance test parameters with those specified in the pro-
tocol helps to establish whether the process and control equipment were operat-
ing at the specified representative conditions. Second, for new sources, the
initial compliance test establishes the operating parameter values that
309
-------
correspond with the measured emission rate. These values can then be compared
with the design values. This provides a fixed reference point for comparison
to future operating data. Third, the values of the baseline parameters pro-
vide data for evaluating routine inspection data. By knowing the effects of
the various process and control equipment parameters on emissions, one can
make comparisons to evaluate the direction and magnitude of any changes in
performance. Fourth, documentation of the baseline data will assist in setting
specific ranges on important parameters for possible inclusion in an operating
permit (if required by the agency). Finally, the baseline test provides a
fixed reference point for comparing long-term performance trends. Proper
evaluation of the baseline data may assist in the establishment of preventive
maintenance schedules as well as provide an indication of any design or in-
stallation problems. In addition, the rate at which the normal operating param-
eters may vary from baseline values may assist the agency in scheduling routine
inspections and periodic compliance tests.
The types of data that should be recorded are dependent on site-specific
factors such as the type of source or process, and the control equipment in-
stalled. For that reason, the agency observer and the inspector (if different
from the observer) should become aware of the site-specific factors that affect
the emissions, and take steps to obtain that data. In some instances it will
be difficult to separate and correlate all effects of the variations in the
process and control equipment parameters. By acquiring as complete an under-
standing as possible of the facility production process and its controls
prior to the stack test, however, the agency observer can frequently identify
key parameters that will have the most influence on emission levels. It is
strongly recommended that the baselining only focus on those parameters that
have a documented affect on the emission levels rather than on all possible
parameters that might influence the emission levels. Collection of data that
have no significance can be inefficient and counterproductive. Considerable
effort can be involved in recording and analyzing all process and control equip-
ment data normally available at a facility. The inspector must be selective
as to which data to collect. Also, certain data may be of a proprietary nature
and considered to be confidential business data that will require special
handling and safekeeping. It is best not to incur this responsibility if it
can be avoided.
310
-------
Although it 1s suggested that the agency collect all pertinent data, the
agency should.be certain that 1) the data are needed and a change in its value
has an effect on the operation of the source and 2) the data are accurate to
the point that the recorded value has some meaning. When an operation and
maintenance program or permit to operate is issued based on a range of process
and control equipment parameters, the agency must be able to measure these
parameters. As a result* maintenance and calibration of the key parameter
instrumentation is needed prior to the performance test. Additional instru-
mentation may also be needed in some cases.
During the baseline source test, the agency observer must essentially
conduct an inspection to obtain all the operating parameters for evaluation.
The only major difference between this and any routine inspection is that
emission testing is occurring simultaneously. Thus, the observer cannot
spend all of the time observing the test because of other responsibilities.
Process and control equipment parameters should be checked throughout the
test. Data should be obtained for the week previous to and the week after the
stack test to demonstrate representative conditions during the test. In many
cases two agency inspectors or observers may be needed to observe the baseline
testing. The services of the field inspector responsible for that facility is
strongly recommended to enhance the field inspector's knowledge and relation-
ship with the facility.
The use of the baseline for documenting deviations from normal conditions
requires the establishment of a logic system for each process or control device
operating parameter used. A substantial change in the parameter is evaluated
based on its impact on the overall emission levels. For example, an increase
in recovery boiler firing rate would be evaluated because of its impact on the
uncontrolled emission rate from the boiler.
The technical data provided in previous sections on control device mal-
functions and process conditions are intended to provide the inspector with
the technical .background to evaluate the emission level changes. The evalua-
tion may be subjective in many instances. In other cases, sufficient technical
data are available to accurately predict the emission levels.
The data obtained are usually sufficient to allow the inspector to
negotiate corrective action with respect to the process and control equip-
ment without the expense of conducting a performance stack test. Many
.311
-------
deficiencies may be corrected as a result of increased or redirected
maintenance activities. The ability of inspectors to negotiate such correc-
tive action varies from agency to agency. The inspector must operate within
his agency's guidelines with regard to negotiating compliance agreements or
issuing notices of violation (NOV). In some cases the corrective action can
be completed before the notices can be drafted and formally issued.
The following is a summary of key operating parameters that the inspector
may use to evaluate the overall performance of the process/control equipment
for each major piece of process or control equipment.
4.1.1 Recovery Boiler
The particulate and SCL emission rates from the recovery boiler are inter-
related because the primary method of S02 and TRS control is to convert these
pollutants to sodium sulfate, which increases the particulate emission rate.
The primary parameters that affect particulate emissions are: firing rate,
primary air rate, excess air, smelt bed temperature, ESP power, ESP superficial
velocity, and flue gas oxygen. A shift in many of these parameters indicates
an increase in emissions. In many cases these shifts in parameters can be
used in support of a requirement to take corrective action or to conduct a
performance stack test. Table 4-1 summarizes the effects of recovery boiler
and ESP operating parameters on particulate and TRS emission rates.
4.1.2 Smelt Tank
The primary parameters that may be used to determine compliance from the
smelt dissolving tank are connected with the rate of particulate generated and
the condition of the control devices. Specifically the rate of generation of
particulate is related to smelt rate (i.e., boiler firing rate and reduction
efficiency) and the amount of particle reentrainment. The condition of the
control device is related to such variables as superficial velocity and water
flow rate. Table 4-2 summarizes the effects of smelt tank and venturi scrubber
operating parameters on particulate and TRS emission rates.
4.1.3 Lime Kiln
Uncontrolled particulate emission rates from the kiln are primarily
affected by parameters that affect the superficial velocity through the kiln,
the particle size of the kiln dust, rate of evolution of volatile particulate,
' 312
-------
TABLE 4-1. SUMMARY OF THE EFFECTS OF RECOVERY BOILER AND ESP OPERATING
PARAMETERS ON PARTICULATE AND TRS EMISSION RATES
Parameter
Firing rate
Primary air
Excess air
Smelt bed temperature
ESP power input
ESP superficial ve-
locity
Flue gas oxygen
Primary air tempera-
ture
Visible emissions
Black liquor sulfidity
Change
Increase
Increase
Increase
Increase
Decrease
Increase
Increase
Decrease
Increase
Increase
Effect on parti cul ate
emission rate
Increase
Increase
Increase
Increase
Increase
Increase
Increase
Decrease
Increase
None
Effect on TRS
emission rate
Increase
Decrease
Decrease
Decrease
None
None
Decrease9
Increase
None
Increase
llf increase in oxygen is a result of an increase in primary air volume.
313
-------
TABLE 4-2. SUMMARY OF THE EFFECTS OF SMELT TANK AND VENTURI SCRUBBER
OPERATING PARAMETERS ON PARTICULATE AND TRS EMISSION RATES
Parameter
Firing rate (smelt
rate)
Shatter jet steam rate
Mesh pad superficial
velocity
Mesh pad back flush
rate (caustic)3
Packed bed water flow
(caustic)3
Venturi scrubber water
flow
Pressure drop
Visible emissions
Change
Increase
Increase
Increase
Decrease
Decrease
Decrease
Decrease
Increase
Effect on particulate
emission rate
Increase
Increase
Increase
1
Increase
Increase
Increase
Increase
Increase
Effect on TRS
emission rate
Increase
Increase
None
Increase
Increase
None
None
None
It should be noted that all smelt vent control devices do not use caustic.
In some cases only water is used.
,314
-------
and the feed rate to the kiln. The superficial velocity is a function of kiln
firing rate and temperature profile. The rate of evolution of volatile par-
ticulate is related to slurry feed rate and the amount of soda present in the
slurry. Parameters affecting the control device are liquid-to-gas ratio,
pressure drop, and particle size^ TRS emissions are related to mud washing
efficiency (% sodium sulfide expressed as Na20), flue gas oxygen, and lime
mud slurry moisture. The cold end temperature has an effect on TRS emission
levels. Both the kiln excess air and temperature profile down the kiln in-
fluence the residence time and oxidation rate of TRS compounds where the kiln
is used as a control device. Where sulfides are present in an insoluble form
and they cannot be removed through washing, lime mud oxidation may be required.
Table 4-3 summarizes the effects of lime kiln and scrubber operating parameters
on particulate and TRS emission rates.
4.1.4 Slaker
The rate of green liquor and calcium oxide reacted in the slaker has the
strongest effect on uncontrolled particulate emissions. The amount of heat re-
leased (i.e., steam generated) and the degree of agitation are related to the
reaction rates. The condition of the scrubber (i.e., water flow rate, liquor
gas ratio, and pressure drop) also affects the emission rate. Table 4-4 sum-
marizes the effects of slaker and venturi scrubber operating parameters on
particulate emission rates.
4.1.5 Turpentine Condenser and Multiple-Effect Evaporators
The rate of TRS emissions from multiple-effect evaporators and the
turpentine condenser is primarily a function of noncondensable gas volume and
tail gas condenser final temperature. The equilibrium vapor pressure of TRS
in the gas stream is also a function of temperature. Table 4-5 summarizes the
effects of turpentine condenser and multiple-effect evaporator operating
parameters on TRS emission rates.
4.1.6 Blow Tank and Hot Water Accumulator
The rate of TRS emissions from the hot water accumulator is a function of
digester operation, the condition of the primary and secondary condensers, and
blow gas volume. The parameters are generally so interrelated that a single
315
-------
TABLE 4-3., SUMMARY OF THE EFFECTS OF LIME KILN AND SCRUBBER OPERATING
PARAMETERS ON PARTICULATE AND TRS EMISSION RATES
Parameter
Production rate
Firing rate
Kiln temperature
(firing end)
Kiln oxygen
Slurry moisture
Mud sodium sulfide
content (Na20)
Scrubber L/6 ratio
Scrubber pressure drop
Scrubber throat
velocity
Water maldistribution
venturi throat
Visible emissions
Change
Increase
Increase
Increase
Decrease
Increase
Increase
Decrease
Decrease
Decrease
Increase
Effect on parti cul ate
emission rate
Increase
Increase
Increase
Decrease
Decrease
Increase
Increase
Increase
Increase
Increase
Increase
Effect on TRS
emission rate
Increase
None
None
Increase
Increase
Increase
None
None
None
None
None
316
-------
TABLE 4-4. SUMMARY OF THE EFFECTS OF SLAKER AND VENTURI SCRUBBER
OPERATING PARAMETERS ON PARTICULATE EMISSION RATE
Parameter
Causticizing rate
Shower water flow
Venturi L/G ratio
Venturi pressure drop
Visible emissions
Venturi throat velocity
Change
Increase
Decrease
Decrease
Decrease
Increase
Decrease
Effect on parti cul ate
emission rate
Increase
Increase
Increase
Increase
Increase
Increase
TABLE 4-5. SUMMARY OF THE EFFECTS OF TURPENTINE CONDENSER AND MULTIPLE-
EFFECT EVAPORATOR OPERATING PARAMETERS ON TRS EMISSION RATE
Parameter
Change
Effect on TRS
emission rate
Tail gas condenser in-
let water temperature
Tail gas condenser
water flow rate
Noncondensable gas
volume
Vent gas temperature
Increase
Decrease
Increase
Increase
Increase
Increase
Increase
Increase
317
-------
parameter analysis is not effective in predicting emissions. Generally,
however, TRS emissions will increase if the condensers are plugged.
4.2 CALCULATION OF EMISSION RATES
In most cases the determination of compliance is based on a series of
primary control performance indicators that may not generally be quantifiable.
In specific applications, however, the inspector may make reasonable estimates
of emission rates using these indicators. The following methods have been
demonstrated to be effective in calculating or estimating the emission rates
for various emission sources.
4.2.1 TRS Sources
Because the final temperature and pressure of noncondensable gases define
the partial vapor pressure of TRS compounds emitted from the condenser, these
values taken in conjunction with gas flow can be used to calculate the emis-
sion rate. As discussed in Section 3, the concentration of TRS gases is
directly related to the condenser outlet temperature. If the inspector can
determine the gas volume and temperature, the emission rate in pounds per hour
o o
may be calculated by the product of concentration and flow (i.e., g/m x m /min
x 60 min T 453.6 g/lb). This method can be applied to such sources as the
turpentine vent and multiple effect evaporators.
4.2.2 Emissions from Recovery Boilers
The efficiency of the ESP serving the recovery boilers can be determined
by using variables that define the power input to the system and the gas
volume being treated. The equation that is generally applied is a modified
version of the Deutsch-Anderson equation. The equation contains a constant
that must be estimated or calculated from a previous baseline performance
test. A detailed discussion of the use and limitations of the equation are
provided in Section 3.3. In order to apply this method, the inspector must
be able to determine the flue gas volume passing through the ESP and the
power input to the ESP. The inspector must also be able to determine that the
unit is in reasonably good operating condition (no gross gas maldistribution,
power distribution imbalances, high resuspension rates, or high rapper re-
entrainment) because certain occurrences can cause the Deutsch-Anderson equa-
tion to over-predict efficiency.
318
-------
The value to be used in the equation is determined from previous base-
line tests or it may be estimated. Typical values for k range from 0.1 to
0.25.5 It is usually advantageous to request that stack tests be performed
over a range of boiler loads (gas volume) and at several ESP power input
levels. This allows the inspector to determine if the value of k remains
reasonably constant over the range of normal operating conditions. The use
of the Matts-Ohnfeldt version of the efficiency equation has been shown to
provide a better correlation where units are designed for high specific corona
power levels (> 500 W/1000 acfm).
Figure 4-1 shows a method of calculating ESP efficiency based on migra-
tion velocity and plate area. This method is useful in reviewing design
capabilities of an ESP but may not be accurate for evaluating a unit that is
currently in service. This method is limited because the migration velocity
is not constant over the operating range of the unit. As the power input to
the unit is decreased, the migration velocity is reduced. The lower section
of Figure 4-1 shows the method to be used in calculating emissions using ESP
power input and gas volume. Power input is calculated from secondary meter
readings. The gas volume is calculated from a modified F-factor or measured
with a Pitot tube. The most serious deficiency in applying this method is the
determination of the uncontrolled emission rate from the boiler. As stated
previously, the uncontrolled particulate emission rate is not constant and
is a function of the boiler combustion characteristics. The emission rate is
a compromise between particulate as sodium sulfate and S02- If the level
of S02 can be determined to be fixed over the operating range of interest,
the particulate emission rate may b* assumed to be reasonably constant (Sec-
tion 3.3.1). If inlet performance stack tests are available, they should be
used in the mass emission rate calculation.
4.2.3 Power Boilers
Emission rates from power boilers using an ESP may be calculated in a
similar manner to the one used for recovery boilers. The restrictions and
cautions are the same in applying the equations. In general, the values of k
and migration velocity are different from those applied to recovery boilers
(Section 3.5).
319
-------
Given:
Boiler:
ESP:
Manufacturer B-W
Indirect contact
1200 ADTP/day
Firing rate 350 gpm at 61.5% BLS (149,820 Ib/h BLS)
Kopper's
Wet bottom
2 chambers
3 fields „
Plate area = 151,200 ft
Superficial velocity = 3.11 ft/s
SCA = 386 ft2/ 1000 acfm
Gas volume = 392,000 acfm (6533.3 acfs)
Gas temperature = 415°F
Migration velocity =8.0 cm/s (0.262 ft/s)
Uncontrolled emission rate =3.2 gr/acfm
Calculate:
1. Allowable emission (AL)
AL = 3 lb/3000 Ib BLS
= 149.821b/h
2. Uncontrolled Emission Rate (UN) based on inlet loading
UN = (392.000 acfm) (3. 2 gr/acfm) _ fin .
7000 gr/lb bu min
3.
UN = 10,752 Ib/h
Actual Emission per Deutch-Anderson equation for ESP efficiency
(AE) '
a. Design efficiency for ESP per Deutch-Anderson equation:
(1 -
100
where n = collection efficiency, %
w = migration velocity, ft/s
A = electrode collecting area, ft
v = gas volume, acfs
Figure 4-1. Calculation of kraft recovery boiler ESP efficiency.
320
-------
b.
-[0.26
n = (1 - e
n = (1 - e"6'017) 100
n = (1 - 0.002437) 100
n = 99.75%
AE = UN x (100 - n)/100
) 100
AE = (10,752)
AE = 26.20 Ib/h
Actual Efficiency for ESP using Modified Deutsch-Anderson equation
n = 100 [1 - e-°-06 k (p/C»]
given: power input 135,000 watts
gas volume 392,000 acfm
k = 0.135 from previous stack tests
f - n fte/n'iocwl35,000x1
n = i . e-0.06(0.135)( 3g2 )J
n =
n =
1 - e
-3.099
100
1 - 0.04501 -100
n = 95.49%
AE = UN x (100 - collection efficiency)/100
AE - (10,752) (
AE = 484.6 Ib/h
Figure 4-1. Calculation of kraft recovery boiler ESP efficiency. (Continued)
321
-------
4.3 STACK TEST METHODS
Each State or local agency has adopted stack test requirements that may
be used to demonstrate compliance with parti oilate, TRS, S02, and visible
emission standards. The following is a description of the federally approved
reference methods.
4.3.1 Particulate Sampling
Particulate sampling is generally accomplished through the use of
Reference Method 5 or 17. Method 17 uses isokinetic sampling with mass weight
determined by filter catch. The filter is at gas stream equilibrium tempera-
ture and moisture conditions within the source stack. Method 5 requires
extractive gas filtering with the gas stream passing the filter to be no
greater than 250°F + 48°F. Both methods use isokinetic sampling methods
with gas moisture determination by condensation in the impingers and adsorption
in the silica gel. Sample volume is measured on a dry basis with a dry gas
meter.
The use of an impinger catch in calculating the mass emission rates vary
from state to state with several States requiring back-half inclusion and
some only counting material in the first impinger.
For most direct-contact recovery boilers, the gas stream temperature is
close to the required temperature for Method 5 filters, and generally there
is reasonable agreement between Methods 5 and 17. In noncontact systems the
gas temperature is higher (>425°F) and the gas stream must be cooled in
Method 5 sampling. Because of the presence of condensables, the Method 5
catch may be higher than a comparable Method 17 catch.
In order to account for the higher gas temperature and to adjust the filter
catch, Method 17 requires the addition of 0.004 gr/dscf to the measured value.
A maximum stack temperature of 400°F should also be specified.
The inclusion of a back-half catch results in an increase in the measured
mass emission rate. A portion of this catch is the result of condensable
particulate passing through the filter, but in some cases the weight may be
due to artifact formation.
In general, there are no technical problems in applying Method 5 or 17
to any particulate-emitting source in the mill. Strict adherence to the
322
-------
sample methods Is required, however, to give reproducible and enforceable
results.
The most serious deficiencies in the sampling protocol of the pulp mill
stack tests fall into two areas 1) failure to fully document process conditions
as required under 40 CFR 60, Appendix A, to prove representativeness of the
test conditions and 2) sampling of scrubber stacks where severe tangential flow
is occurring. Item one has been discussed in other sections of this manual
and does not require further comment.
In regard to item two, most kraft mills use low-energy dynamic or venturi
scrubbers that use cyclone separators to remove water droplets from the gas
stream. The separator creates a highly tangential flow pattern that almost
without exception cannot be sampled in accordance with the requirement of EPA
Reference Methods 1 and 2. Straightening vanes or other methods must be used
to reduce the tangential flow to acceptable levels to complete the test.
Reference Methods 5 and 17 are provided in Appendix B.
4.3.2 TRS Sampling •
TRS sampling is generally accomplished through Methods 16 and 16A. Method
16 is a semicontinuous method that determines sulfur compound emissions using
the principle of gas chromatographic separation and flame photometric detec-
tion. The gas sample is extracted from the gas stream and diluted with clean
dry air. A portion of the sample is analyzed for hydrogen sulfide (H2S),
methyl mercapton (MeSH), dimethyl sulfide (DMS), and dimethyl disulfide (DMDS).
The total of these compounds is defined as TRS. Moisture, carbon monoxide,
carbon dioxide, particulates, and S02 are interferences for this method and
must be quantified or removed from the gas stream before analysis. Method
16 is provided in Appendix C for reference.
Method 16A is the preferred method for TRS sampling because of its
simplicity. The gas sample is extracted from the stack, and S02 is selectively
removed from the sample by use of a citrate buffer solution. The gas stream
is then oxidized and analyzed as S02 using a barium-thorin titration procedure.
The analysis for S02 is identical to that used in Reference Method 6. Method
16A is also provided in Appendix C.
323
-------
4.3.3 SO2 Sampling
Sampling for S02 is generally accomplished through use of a wet chemistry
method that selectively adsorbs and reacts with S02. The Reference Method
is EPA Method 6. A gas sample is extracted from the gas stream, and the S02
and sulfuric acid mist are separated. The S02 is measured by the barium-
thorin titration method. Interferences with the method are free ammonia,
water soluble cations, and fluorides. Cations and fluorides are removed by a
glass wool filter and an isopropanol bubbler. Ammonia interferes by forming
a particulate sulfide that reacts with the indicator. The presence of ammonia
can be detected by observing a white particulate precipitant in the isopropanol
bubbler. Method 6 is provided in Appendix D.
4.3.4 Visible Emissions
Compliance with visible emission standards is usually documented through
the use of Method 9. The opacity is determined visually by a qualified
observer. In general, the observer visually determines the opacity of an
emitted plume at 15-second intervals for a period of 6 minutes or as required
by agency policy. The averaging time for the observations in Method 9 is 6
minutes. The averaging time varies from state to state, and an exclusion
period above the standard is usually permitted. The presence of uncombined
or condensed water is an interference, and the method cannot be applied in
these circumstances. Observations must be made before the point of steam
condensation (detached plume) or after the point of steam dissipation.
Observations must be conducted with the sun angle behind the observer within
a specified angle and with the plume imposed on a contrasting background.
Method 9 is provided in Appendix E for reference.
324
-------
REFERENCES FOR SECTION 4
1. PEDCo Environmental, Inc. Identification of Parameters That Affect the
Participate Emissions from Recovery Boilers. June 1982.
2. PEDCo Environmental, Inc. Development of Pilot Inspection System for
Virginia Air Pollution Control Commission, Interim Results. November
1982.
3. Saunders, G., and B. DeWees. Observing and Establishing Plant Operating
Baseline Conditions During Compliance Emission Tests. PEDCo Environ-
mental, Inc.
4. White, H. 0. Electrostatic Precipitation of Flyash, Part I. Journal
of the Air Pollution Control Association. January 1977.
5. Hawks, R., and 6. Saunders. Unpublished data. PEDCo Environmental, Inc.
6. U. S. Environmental Protection Agency. Standards of Performance for
New Stationary Sources. EPA-340/1-82-005, June 1982.
325
-------
-------
APPENDIX A
SUMMARY OF STATE REGULATIONS
A-l
-------
-------
TABLE A-l. SUMMARY OF SELECTED VISIBLE EMISSION, PARTICULATE, S02, AND TRS EMISSION LIMITS
APPLICABLE TO SOURCES WITHIN A KRAFT PULP MILL
State
Alabama
Arizona
Participate
general
process - .
regulations
Class II— exist-
ing and all new
E = 4.10 p°'67
P < 60,000 Ib/h
E = 55.0P0-1-40
P > 60,000 Ib/h
Class I"
exlsting
E = 3.59 P0-62
P £60,000 Ib/h
E = 17.31 P0-16
P > 60,000 Ib/h
E = 4.10P0'67
P <. 60,000 Ib/h
E = 55.0P°-^40
P > 60,000 Ib/h
Visible 2
emission
20% (I)
40% (2)
so2—
general
process
10% of Sentering
the process
S0,--fuel
3
burning
Class I
1.8 lb/106 Btu
Class II
4 lb/106 Btu
New
0.8 Ib S02/106
Btu
Oil or coal
existing
1.0 Ib S02/106
Btu
oil or coal
Particulate--,
fuel burning
Class I—exist-
ing and all new
E = 1.38 Q"°'44
Class II—
existing
E = 3.109 q-°-589
'
Q < 4200 x 106
Btu/h
E = 1.02QU-769
Q > 4200 x 106
Btu/h
E = 17.0Q0'432
Particulate—
kraft pulp
mills4
Rec. furn.
4 Ib/TADP
Smelt tank
0.5 Ib/TADP
lime kiln
0.1 Ib/TADP
Total reduced
sulfur— kraft
pulp mills4
1.2 Ib/TADP5
(continued)
-------
TABLE A-1 (continued)
State
Arkansas
Florida
Georgia
Par t1culate
general
process - ,
regulations
P * 100 Ib/h,
E = 2.5 Ib/h;
P * 1.0 x 106
Ib/h;
E = 100 Ib/h;
P = 1.0 x 106
Ib/h;
E = 700 Ib/h
E = 3.59P0'62
P < 60,000 Ib/h
E = 17.31P0'16
P > 60,000 Ib/h
Visible -
eralsslon
New
20% (1)
Existing
40% (2)
New
20* (1)
Existing
40% (2)
Carbonaceous
fuel
30% (1.5)
New
20% (1)
Existing
40% (2)
so2~
general
process
S0,--fue1
3
burning
> 250 x 106 Btu/h
Liquid fossil
2.75 Ib S02/106
Btu
Solid fossil
6.17 Ib S02/106
Btu
Liquid fossil
< 100 x 106 Btu/h
2.6% S
> 100 x 106 Btu/h
3.0% S
Partlculate--,
fuel burning
Uses general pro-
cess weight rate
for solid fuel
-
Fossil fuel
> 250 x 106 Btu/h
0.1 lb/106 Btu
(2-h avg)
Carbonaceous fuel
> 30 x 106 Btu/h
0.2 lb/106 Btu
Applies to non-
attainment
areas
New
< 10 x 106 Btu/h
0.5 lb/106 Btu/h
10-250 x 106
Btu/h
Partlculate--
kraft pulp
Mills*
Source by source
Rec. furn.
80 - 800 Ib/h
Smelt tank
25 - 55 Ib/h
Lime kiln
40 - 110 Ib/h
3 lb/3000 BLS
Total reduced
sulfur— kraft
pulp mills*
New
•1 ppm or 0.03 Ib/
3000 Ib/BLS
Existing
17.5 ppm or 0.05
lb/3000 Ib BLS
Rec. furn.
Old - 20 ppm
(8% 02)
Hew - 5 ppm
Cross
25 ppm
(continued)
-------
TABLE A-1 (continued)
State
Georgia (cont.)
Idaho
Particulate
general
process - ,
regulations
E - 4.10P0'67
P <.60,000 Ib/h
E = 55.0P0-1-40
P > 60,000 Ib/h
Visible -
emission
New
20% (1)
Existing
40% (2)
so2-
general
process
New
20% (1)
Existing
40% (2)
SO,— fuel
3
burning
Solid fossil
c
< 100 x 10° Btu/h
2.5% S
> 100 x 106 Btu/h
3.0% S
Existing
< 10 x 106 Btu/h
0.07 lb/106 Btu/h
10-206o x 106
Btu/h
E . 0.7 (10)0.202
> 2000 x 106
Btu/h
0.24 lb/106 Btu
Residual oil
1.75% S
Distillate oil
Grade 1—0.3% S
Grade 2—0.5% S
Coal
1.0% S
Particulate—,
fuel burning
E • 0.5 (IP-)0'5
> 250 x 106 Btu/h
c
0.10 Ib/lO? Btu
Q 1 10 x 106 .
Btu/h
0.6 lb/106 Btu/h
Q >. 15,000 x 106
Btu/h
0.12 lb/106
Btu/h
Particulate—
kraft pulp
mills4
'
Total reduced
sulfur— kraft
pulp mills4
Smelt tank
0.0168 Ib TRS/ton
pi c
Di_J
Lime kiln
,40 ppm (10% 02)
>
(continued)
-------
TABLE A-1 (continued)
Kentucky
Partlculate
general
process - ,
regulations1
New
E ' 3.59P0'62
P <_ 60,000 Ib/h
E - 1731P0'16
P > 60,000 Ib/h
Existing
E - 55.0P-0-11
40
0.02 gr/ft3
97X efficiency
Visible *
emission
Hew--20Z
40X 2-m1nute
In any 1 h
Existing
40S
sor-
general
process
S0,--fuel
3
burning
County Class I
Liquid fuel
Y»7.7223X-°-4106
Solid fuel
YM3.8781X"0-4434
County Class II
Liquid fuel
Y»9.4644X"0l374°
Solid fuel
Y=14.1967X'°-3740
County Class III
Liquid fuel
Y'8.060X"0<2436
Solid fuel
Y=12.2539X'°'2432
County Class IV
Liquid fuel
Y«7.3639X"°'1347
Solid fuel
Y=10.8875X'°'1338
County Class V
Liquid fuel
Y-8.0189X"0'1260
articulate— ,
fuel burning
riorlty I
-0.9634Q-0-2356
riorlty II
.1.2825X-0-2330
riorlty III
Y=1.3152X-0'2159
Partlculate—
Scraft pulp
mills*
Rec. furn.
3.5 Ib/ton
Lime kiln
1.0 Ib/ton
Smelt tank
0.5 Ib/ton
Total reduced
sulfur—kraft
pulp mills*
Rec. furn.
15 ppm arlth. avg
40 ppm for more
than 60 tnln. In
any 24 h
Thermal oxidation
98X efficiency
(continued)
-------
TABLE A-1 (continued)
State '
Kentucky (cont.)
Particulate
general
process - ,
regulations
-
Visible 2
emission
<
so2-
general
process
S0,,~fuel
3
burning
Solid fuel
Y=12.028«-°-1260
NGW
Liquid
0.8 lb/106 Btu
(2h avg.)
Solid
1.2 lb/106 Btu
(2h avg.)
Existing
Priority I
Same as new
Priority II
Liquid
1.5 lb/106 Btu
Solid
2.0 lb/106 Btu
Priority III
Liquid
2.0 lb/106 Btu
^ Solid
3.5 lb/106 Btu
Particulate—,
fuel burning
Particulate—
kraft pulp
mills*
Total reduced
sulfur—kraft
pulp mills"
>
(continued)
-------
TABLE A-l (continued)
State
Louisiana
Ma'ine
Maryland
Partlculate
. general
process - .
regulations
E - 4.10P0'67
P <. 60,000 Ib/h
E - 55.0P0'^40
P > 60,000 Ib/h
E = 3.59P0'62
P <_ 60,000 Ib/h
E = 17.31P0'16
P > 60,000 Ib/h
Hew
0.05 gr/SCFD
Existing
E = 55. OP0' "40
P < 60.000 Ib/h
0.05 gr/SCFD
Visible .
emission
20% (1)
40X (2)
20% (I)
^so2-
general
process
2000 ppm
New
500 ppm
Existing
2000 ppm
S0,~fuel
3
burning
2000 ppm
Area 1 - 6
2.5X S
Area 7
after 11/1/75
1.5% S
after 11/1/85
1.0% S
>. 13 x 106 Btu/h
Residual oil
2.0% S
Distillate oil
Process gas
0.3% S
Partlculate--,
fuel burning
0.6 lb/106 Btu
3-10 x 106 Btu/h
0.6 lb/106 Btu
> 150 x 106
Btu/h
0.3 lb/106 Btu
New
13 i Q <. 25 x
106 Btu
25 < 0 <_ 250 x
106 Btu/h
Partlculate—
kraft pulp
mlllsl
Rec. furn.
4 Ib/TADP
Smelt tank
0.5 Ib/TADP
Lime kiln
1.0 Ib/TADP
Rec. furn.
4 Ib/TADP
Smelt tank
0.5 Ib/TADP
Lime kiln
1.0 Ib/TADP
Total reduced
sulfur—kraft
pulp mills'
0.6 lb/TODP6
>
00
(continued)
-------
TABLE A-l (continued)
Stdte
Maryland (cont.)
Michigan
Partlculate
general
process - ,
regulations1
P > 60,000 Ib/h
E = 4.10P0'67
P £60,000 Ib/h
E = 55.0P0-1-40
P > 60,000 Ib/h
Visible -
emission
(2)
•-
so2-
general
process
SO,— fuel
3
burning
>_ 13 unit actual
and >_ 100
plant design
capacity
Solid fuel
3.5 Ib S02/
106 Btu
< 500,000 Ib
steam/h
coal 1.5% S,
2.4 Ib S02/106
Btu
oil 1.5* S,
1.7 lb/106 Btu
> 500,000 Ib
steam/h
coal 1.0* S,
1.6 lb/106 Btu
oil 1.0*. S
1.1 lb/106 Btu
Particulate— 3
fuel burning
E = log 10
[6.597538Q"0-3]
> 250 x 106 Btu/h
6.1 x 106 Btu/h
Existing
< 10 x 106 Btu/h
0.6 lb/106 Btu/h
> 10 x 106 Btu/h
E = 1.025985Q"0'23
Pulverized coal
<_ 100 x 103 Ib
steam/h
0.3 lb/1000 Ib gas
Other
'articulate—
kraft pulp
mills'!
297
0-100 0.65 lb/1000 Ib gas
100-300 0.64-0.45 lb/1000 Ib gas
Total reduced
sulfur— kraft
pulp mills'
(continued)
-------
TABLE A-1 (continued)
State
Minnesota
Partlculate
general
process -
regulations1
E = 4.10P0'67
P <. 60,000 Ib/h
E ' 55.0P0ll*40
P > 60,000 Ib/h
Visible
emission2
New
201 (1)
Existing
60S (3)
so2-
general
process
S02~fuel
burning3
New
< 250 x 106 Btu/h
Liquid
2.0 lb/106 Btu
Solid
4.0 lb/106 Btu
> 250 x 105 Btu/h
Liquid
0.8 lb/106 Btu
Solid
1.2 lb/106 Btu
Existing
< 250 x 106 Btu/h
Liquid
2.0 lb/106 Btu
Solid
4.0 lb/106 Btu
> 250 x 106 Btu/h
Liquid
1.6 lb/106 Btu
Solid
3.0 lb/106 Btu
Partlculate—
fuel burning3
New
0.4 lb/106 Btu
Existing
0.6 lb/106 Btu
Particulate—
kraft pulp
mills*
Total reduced
sulfur—kraft
pulp mills''
(continued)
-------
TABLE A-1 (continued)
State
Mississippi
Montana
Participate
general
process - .
regulations
E=4.10P°'67
E = 4.10P0-67
P £60,000 Ib/h
E = 55. OP0-1-^
P > 60,000 Ib/h
Visible 2
emission
40% (2)
New
20%
Existing
(2)
so2-
general
process
New
500 ppm
Existing
2000 ppm
Concentrations
shall not exceed
2 ppm at any
time
.
SO,— fuel
3
burning
New
Mod. units
2.4 Ib S02/106
Btu
Existing
< avg annual .
emission rate
for 1970 units
1.0 1b/106 Btu
Particulate—,
fuel burning
£ 10 x 106 Btu/h
0.6 lb/106 Btu
> 115 x 106 Btu/h
0.1 lb/106 Btu
New
£ 10 x 106 Btu/h
0.6 lb/106 Btu
100 x 106 Btu/h
0.35 lb/106 Btu
1000 x 106 Btu/h
0.2 lb/106 Btu
>. 10,000 x 106
Btu/h
0.12 lb/106 Btu
Particulate—
kraft pulp
mills*
Rec. furn.
4 Ib/TADP
Smelt tank
0.5 Ib/TADP
Lime kiln
1.0 Ib/TADP
Total reduced
sulfur--kraft
pulp mills*
Rec. furn.
0.087 lb/1000 Ib
BLS or 17.5 ppm
(continued)
-------
TABLE A-1 (continued)
State
Pennsylvania
(cent.)
Partkulate
general
process -
regulations1
Visible
emission'
.
soz~
genera]
process
S02--fue1
burning3
> 250 x 106 Btu/h
Inner—O.es
Outer-- 1.2%
Fuel oil --Com.
No. 2 & lighter
Inner— 0.2%
Outer— 0.3%
No. 4,5,6, and
heavier
Inner- -0.5%
Outer— 1. OX
Noncommercial
Inner— 0.62
Outer-1.2%
Partlculate—
fuel burning3
Partlculate—
kraft pulp
mills''
Total reduced
sulfur—kraft
(continued)
-------
TABLE A-1 (continued)
State
Montana (cont.)
New Hampshire
Particulate
general
process - .
regulations
New
E = 4.10P°'67
P < 60,000 Ib/h
nil
E = 55. OP" -40
P > 60,000 Ib/h
Existing
EV5.05P°'67
P £60,000 Ib/h
E = 66. OP0' ^48
P > 60,000 Ib/h
Visible -
emission
New
20%
Existing
C
< 250 x 10° Btu
40%
>' 250 x 106 Btu
20%
. S02"
general
process
SO,— fuel
3
burning
Solid
2.8 Ibs S/106
Btu after
4/15/80
1.5 Ibs/ 106 Btu
liquid
#2 - 4% S
#4 - 11 S
#5 4 #6 - 2% S
Particulate—,
fuel burning
Existing
<. 10 x 106 Btu/h
0.6 lb/106 Btu
100 x 106 Btu
0.4 lb/106 Btu
1000 x 106 Btu
0.28 lb/106 Btu
>. 10,000 x 106
Btu
0.19 lb/106 Btu
New
Q < 10 x 106
Btu/h
£
0.6 lb/10° Btu
10 < Q < 250 x
106 Btu/h
E=1.0286Q~°-2341
Q > 250 x 106
Btu/h
0.1 lb/106 Btu
Particulate—
kraft pulp
mills*
Rec. furn.
4 Ib/TADP
Smelt tank
0.5 Ib/TADP
Lime kiln
1.0 Ib/TADP
Total reduced
sulfur— kraft
pulp mills4
2 Ibs/TADP
V-»
CO
(continued)
-------
TABLE A-1 (continued)
State
Hew Hampshire
(cent.)
New York
Participate
general
process - ,
regulations
E = 0.24P0'665
P < 100,000 Ib/h
E«3gp0.08250
P > 100,000 Ib/h
Visible ,
emission
20% (1)
so2«
general
process
S0,,--fuel
<• 3
burning
By area
New York City
Fuel oil
0.2* to'l.OX
Coal— 0.2 to
0.6 lb/106 Btu
Rest of State
New
> 250 x 106
Btu/h
Fuel oil— 0.75X S
Coal
0.6 Ib S/106 Btu
New and Existing
< 250 x 106
Btu/h
Fuel oil -23 S
Coal— 2.5 lb/106
Btu S max or
1.9 lb/106 Btu
S avg.
Particulate--,
fuel burning
Existing
0 <. 10 x 106
Btu/h
0.6 lb/106 Btu
10 < 10,000 x
106 Btu/h
E=0.8803Q-°<1665
Q > 10,000 x 106
Btu/h
0.19 lb/106 Btu
1-10 x 106 Btu/h
0.62 lb/106 Btu
10-10,000 x 106
Btu/h
E = LOO.'0'22
Coal installa-
tions
< 250 x 106 Btu/h
Prior to 8/11/72
E - LOO.'0'22
Coal and oil In-
stallations
> 250 x 106 Btu/h
After 8/11/72
0.1 lb/106 Btu
50-250 x 106 Btu/f
011-0.2 lb/106
Btu
Particulate-
kraft pulp
mills4
•
Total reduced
sulfur— kraft
pulo wills*
(continued)
-------
TABLE A-1 (continued)
State
North Cdrol Ififl
Ohio
Particulate
general
process - .
regulations
E = 4.10P0'67
P < 60,000 Ib/h
E » 55.0P0<1-40
P^ 60,000 Ib/h
E = 4.10P0'67
P <_ 60,000 Ib/h
E = SS.OP0'1^
P > 60,000 Ib/h
Visible 2
emission
New
20% (1)
Existing
40% (2)
20% (1)
so2-
general
process
S0,--fuel
3
burning
2.3 Ib SOZ/106
Btu
t
Source specific
*
Part1culate--,
fuel burning
Coal, oil
Q < 10 x 106 Btu/h
0.6 lb/106 Btu
10-10,000 Btu/h
E = 1.09Q-0-2594
Q >. 10.000 Btu/h
0.1 lb/106 Btu
Hood
Q < 10 x 106 Btu/h
0.7 lb/106 Btu
Q > 10 x 106 Btu/h
E = 1.1698Q-0-2230
Priority I Region
Q <. 10 x 106 Btu/h
0.6 lb/106 Btu
Q >_ 1000 x 106
Btu/h .
0.1 lb/106 Btu
Priority II & III
Q <_ 10 x 106 Btu/h
0.6 lb/106 Btu
Q >. 1000 x 106
Btu/h
0.15 lb/106 Btu
articulate—
kraft pulp
mills*
Rec. furn.
3.0 Ib/TAOP
Smelt tank
0.6 Ib/TADP
lime kiln
O.S Ib/TADP
Total reduced
sulfur— kraft
pulp mills'*
'
m
(continued)
-------
TABLE A-1 (continued)
State
Oklahoma
Participate
general
process - .
regulations1
E * 4.10P0'67
P <. 60,000 Ib/h
E = 55.0P0<1i40
P > 60,000 Ib/h
Visible
emission'
20%
so2-
general
process
Kraft- pulp
18 Ib/TADP
S02--fue1
burning3
Particulate—
fuel burning3
< 10 x 106 Btu/h
0.6 lb/106 Btu
100 x 106 Btu/h
0.35 lb/106 Btu
1000 x 106 Btu/h
0.2 lb/106 Btu
" >. 10,000 x 106
Btu/h
0.1 lb/106 Btu/h
Wood
< 10 x 106 Btu/h
0.60 lb/106 Btu
10-1000 x 106
Btu/h
0.5 lb/106 Btu
1000-10,000
0.35 lb/106 Btu
>. 10,000 x 106
Btu/h
0.15 lb/106 Btu/h
Combined wood/
fossil
> 250 x 106 Btu
0.1. lb/106 Btu
Particulate—
kraft pulp
mills4
Total reduced
sulfur—kraft
pulp mills4
CTl
(continued)
-------
TABLE A-1 (continued)
State
Oregon
Particulate
general
process -
regulations1
P = 100 Ib/h
0.46 Ib/h
P = 10,000 Ib/h
10.0 Ib/h
P < 60,000 Ib/h
E = 55.0P0-1-40
Visible
emission'
New
20% (1)
Existing
40% (2)
so2~
general
process
1000 ppm
S02--fuel
burning-*
Residual oil
1.75% S
Distillate
Grade 1 0.3%
Grade 2 0.5%
Coal
1% S
New
150 < Q <. 250 x
106 Btu/h
Liquid
1.4 lb/106 Btu
Solid
1.6 lb/106 Btu
> 250 x 106 Btu/h
Liquid
0.8 lb/106 Btu
Solid
1.2 lb/106 Btu
Particulate—
fuel burning3
New
<_ 10 x 106 Btu/h
0.27 lb/106 Btu
> 1000 x 106
Btu/h
0.05 lb/106 Btu
Existing
£ 10 x 106 Btu/h
0.55 lb/106 Btu
> 1000 x 106
Btu/h
0.27 lb/106 Btu
Particulate--
kraft pulp
mills'*
Rec. furn.
4 Ib/TADP
Smelt tank
0.5 Ib/TADP
Lime kiln
1 Ib/TADP
Total reduced
sulfur—kraft
pulp mills4
0.5 Ib/TADP
or 17.5 ppm
-
I
M
>>l
(continued)
-------
TABLE A-1 (continued)
State
Pennsylvania
Participate
general
process - ,
regulations
0.04 gr/ft3 or
A * 6000 E
E = 150,000 -
300,000
0.02 gr/ft3
E > 300,000 ft3
Visible 2
emission
202
3 minutes
1n any In
60% any
time
so2-
general
process
500 ppm
SO.-fuel
3
burning
3.0 lb/106 Btu
< 50
A = 5.1E0'14
50 - 2000
1.8 lb/106
> 2000
Allegheny Co.
1.0 lb/106 Btu
< 50
A = 1.7 E0-14
50 - 2000
.6 lb/106 Btu
> 2000
SE Penn Air Basin
< 250 x 106 Btu/h
Inner— 1.0%
Outer— 1.2%
Participate--,
fuel burning
< 50 x 106 Btu/h
0.4 lb/106 Btu
50-600 x 106
Btu/h
A = 3.6E-0-56
> 600 x 106 Btu/h
0.1 lb/106 Btu
Partlculate—
kraft pulp
mills'*
Total reduced
sulfur— kraft
pulp wins'1
,
>
(->
00
(continued)
-------
TABLE A-1 (continued)
Parti cul ate
general
process - ,
regulations1
E = 4 10P0.67
P £60,000 Ib/h
p = cc tip" -4(1
P > 60,000 Ib/h
NPU
F^3.59P°-62
P <• fin nnn ih/h
E = 17.31P0'16
P-> fin nnn ih/h
Existing
E - 4.10P0'67
Visible 2
emission
2058
Existing
40%
New
20% (1)
Existing
40% (2)
Wood
> 100 x 106 Btu/h
40%
so2-
general
process
S0,~fuel
3
burning
Class I—
<_ 10 x 106 Btu/h
3.5 lb/106 Btu
> 10 x 106 Btu/h
2.3 lb/106 Btu
Class II—
< 1000 x 106
Btu/h
3.5 lb/106 Btu/h
>_ 1000 x 106 Btu
2.3 lb/106 Btu
Class III--
3.5 lb/106 Btu
< 1000 x 106
Btu/h
1.6 - 5 lb/106
Btu
> 1000 x 106
Btu/h
1.2-5 lb/106 Btu
Particulate— ,
fuel burning
Q < 1300 x 106
0.6 lb/106 Btu
> 1300 x 106
Btu/h
E = 57.84Q-0'637
Except existing
prior to 2/11/71
Q < 10 x 106
Btu/h
0.8 lb/106 Btu
Mau
lien
Q <_ 10 x 106
Btu/h
0-6 lb/106 Btu
10-250 x 106
Btu/h
E= 0.6(i|)°'5566
Particulate--
kraft pulp
mills4
Rec. furn.
2.75 Ib/TADP
Smelt tank
1.0 Ib/TADP
Lime kiln
1.0 Ib/TADP
Rec. furn.
3 Ib/TADP
Smelt tank
0.5 Ib/TADP
Lime kiln
1 Ib/TADP
Total reduced
sulfur— kraft
pulp mills4 -
Rec. furn. (8% 00)
old 20 ppm
new 5 ppm
cross 25 ppm
Dig. system,
evap. system, and
stripper system
5 ppm
Lime kiln
20 ppm (10% 02)
Smelt tank
0.0084 g/ kg BLS
l-»
ID
(continued)
-------
TABLE A-1 (continued)
State
Tennessee (cont.)
Texas
Particulate
general
process -
regulations1
P < 60,000 Ib/h
E = SS.OP0'1^
P > 60,000 Ib/h
E = 3.12P0"985
P < 40,000 Ib/h
E = 25.4P0'287
P > 40,000 Ib/h
Visible
emission2
New
20%
Existing
30%
so2-
general
process
S02~fuel
burninq3
Liquid
440 ppm S02
Solid
3.0 lb/106 Btu
i
Particulate—
fuel burninq3
>. 250 x 106 Btu/h,
0.1 lb/106 Btu
Hood
New
Q < 25 x 106
Btu/h
0.33 gr/sdcf
Q > 100 x 106
Btu/h
0.20 gr/sdcf
Existing
Q < 50 x 106
Btu/h
0.33 gr/sdcf
Q >. 100 x 106
Btu/h
0.1 gr/sdcf
E = 0.048 (stack
flow rate
acfm)-0'62
Particulate—
kraft pulp
mills"
Total reduced
sulfur—kraft
pulp mills4
ro
o
(continued)
-------
TABLE A-1 (continued)
State
Virginia
Washington
Participate
general
process - .
regulations
E = 4.10P0'67
P £60,000 Ib/h
E = 55.0P0-1-40
P > 60,000 Ib/h
New
0.1 gr/sdcf
Existing
0.2 gr/sdcf
Visible -
emission
20%
•35%
Cf\
so2-
general
process
500 ppm
S0,~fue1
3
burning
AQCR 1-6
2.64Q
AQCR 7
Liquid or gas
1.06 Q
Solid
1.52Q
1000 ppm
Particulate--,
fuel burning
Q < 10 x 106
Btu/h
0.6 lb/106 Btu
10-10,000 x 106
Btu/h
E=1.096Q-°'2594
> 10,000 x 106
Btu/h
0.1 lb/106 Btu
AQCR 7
< 100 x 106 Btu/h
0.3 lb/106 Btu
100-10,000 x 106
Btu/h
, - on-0.2386
E = 0.9Q
>. 10,000 x 106
Btu/h
0.1 lb/106 Btu
New
0.1 gr/sdcf
'Existing
0.2 gr/sdcf
Partlculate--
kraft pulp
mills4
Rec. furn.
3 Ib/TADP
Smelt tank
0.75 Ib/TADP
L1me kiln
1 Ib/TADP
Slaker tank
0.3 Ib/TADP
Rec. furn.
0.23 g/dcm
(0.10 gr/dscf)
Smelt tank
0.15 g/kg
(0.30 Ib/ton)
of solids fired
Total reduced
sulfur— kraft
pulp mills*
1.2 Ib/TADP
(daily avg./
quarter)
-
L1me kiln
Not exceed 8 ppm
for two consecu-
tive hours.
Nbt exceed 50 ppm
for a daily aver-
age. After
1/1/85 shall not
exceed 20 ppm for
a daily average.
(continued)
-------
TABLE A-l (continued)
State
Washington (cont.)
Wisconsin
Parti cul ate
general
process - .
regulations
E = 3.59P0"62
P < 60,000 Ib/h
E = 17.31P0'16
P > 60,000 Ib/h
Visible 2
emission
Mew
20%
Existing
40%
sor-
general
process
S0,--fue1
3
burning
> 250 x 106 Btu/h
Liquid
0.8 lb/106 Btu
Solid
1.2 lb/106 Btu
Partlculate— 3
fuel burning
New
<. 250 x 106 Btu/h
0.15 lb/106 Btu
> 250 x 106 Btu/h
0.10 lb/106 Btu
Existing
ASH APS-1
Max 0.6 lb/106
Btu
Partlculate—
kraft pulp
mills'*
Lime kiln
0.30 g/dcm
(0.13 gr/dscf)
All other sources
0.23 g/dscm
(0.10 gr/dscf)
Fugitive emis-
sions—reasonable
precautions.
Note: All mills
are required to
meet RACT
Total reduced
sulfur— kraft
pulp mills4
17.5 ppm
Rec. furn.
DO
ro
Footnotes
1. E, allowable emissions In Ib/h; P, process weight rate 1n tons/h.
2. X )• Rlngelmann number.
3.
4.
5.
6.
E, or Y allowable emissions In Ib/h; Q, or X heat Input 1n 10° Btu/h.
Rec. furn, recovery furnace; evap system, multi-effect evaporator system; dig system, digester system; TADP, ton air dried pulp;
BLS, black liquor solid.
ppm expressed as H,S on a dry gas basis.
TODP, ton oven-dried unbleached pulp.
-------
APPENDIX B
EPA REFERENCE METHODS 1-5, 17
B-l
-------
-------
Title 40—Protection of Environmirrt
APPENDIX A—REFERENCE METHODS
The reference methods in this appendix
are referred to in § 60.8 (Performance Tests)
and § 60.11 (Compliance With Standards
and Maintenance Requirements) of 40 CFR
Part 60, Subpart A (General Provisions).
Specific uses of these reference methods are
described in the standards of performance
contained in the subparts, beginning with
Subpart D.
Within each standard of performance, a
section titled "Test Methods and Proce-
dures" is provided to (1) identify the test
methods applicable to the facility subject to
the respective standard and (2) identify any
special instructions or conditions to be fol-
lowed when applying a method to the re-
spective facility. Such instructions (for ex-
ample, establish sampling rates, volumes, or
temperatures) are to be used either in addi-
tion to, or as a substitute for procedures in a
reference method. Similarly, for sources
subject to emission monitoring require-
ments, specific instructions pertaining to
any use of a reference method are provided
in the subpart or in Appendix B.
Inclusion of methods in this appendix is
not intended as an endorsement or denial of
their applicability to sources that are not
subject to standards of performance. The
methods are potentially applicable to other
sources; however, applicability should be
confirmed by careful and appropriate evalu-
ation of the conditions prevalent at such
sources.
The approach followed in the formulation
of the reference methods involves specifica-
tions for equipment, procedures, and per-
formance. In concept, a performance specifi-
cation approach would be preferable in all
methods because this allows the greatest
flexibility to the user. In practice, however.
this approach is impractical in most cases
because performance specifications cannot
be established. Most of the methods de-
scribed herein, therefore, involve specific
equipment specifications and procedures,
and only a few methods in this appendix
rely on performance criteria.
Minor changes in the reference methods
should not necessarily affect the validity of
the results and it is recognized that alterna-
tive and equivalent methods exist. Section
60.8 provides authority for the Administra-
tor to specify or approve (1) equivalent
methods, (2) alternative methods, and (3)
minor changes in the methodology of the
reference methods. It should be clearly un-
B-3
-------
Chapter I—Environmental Protection Agency
App. A
derstood that unless otherwise identified all
such methods and changes must have prior
approval of the Administrator. An owner
employing such methods or deviations from
the reference methods without obtaining
prior approval does so at the risk of subse-
quent disapproval and retestlng with ap-
proved methods.
Within the reference methods, certain
specific equipment or procedures are recog-
nized as being acceptable or potentially ac-
ceptable and are specifically identified in
the methods. The items identified as accept-
able options may be used without approval
but must be identified in the test report.
The potentially approvable options are cited
as "subject to the approval of the Adminis-
trator" or as "or equivalent." Such poten-
tially approvable techniques or alternatives
may be used at the discretion of the owner
without prior approval. However, detailed
descriptions for applying these potentially
approvable techniques or alternatives are
not provided in the reference methods. Also,
the potentially approvable options are not
necessarily acceptable in all • applications.
Therefore, an owner electing to use such po-
tentially approvable techniques or alterna-
tives is responsible for: (1) assuring that the
techniques or alternatives are in fact appli-
cable and are properly executed; (2) includ-
ing a written description of the alternative
method in the test report (the written
method must be clear and must be capable
of being performed without additional in-
struction, and the the degree of detail
should be similar to the detail contained in
the reference methods); and (3) providing
any rationale or supporting data necessary
to show the validity of the alternative in the
particular application. Failure to meet these
requirements can result in the Administra-
tor's disapproval of the alternative.
METHOD 1.—SAMPLE AND VELOCITY TRAVERSES
POR STATIONARY SOURCES
1. Principle and Applicability
1.1 Principle. To aid in the representa-
tive measurement of pollutant emissions
and/or total volumetric flow rate from a
stationary source, a measurement site where
the effluent stream is flowing in a known di-
rection is selected, and the cross-section of
the stack is divided into a number of equal
areas. A traverse point is then located
within each of these equal areas.
1.2 Applicability. This method is applica-
ble to flowing gas streams in ducts, stacks,
and flues. The method cannot be used
when: (1) flow is cyclonic or swirling (see
Section 2.4), (2) a stack is smaller than
about 0.30 meter (12 in.) in diameter, or
0.071 m2(113 in.2) cross-sectional area, or (3)
the measurement site is less than two stack
or duct diameters downstream or less than a
half diameter upstream from a flow disturb-
ance.
The requirements of this method must be
considered before construction of a new fa-
cility from which emissions will be meas-
ured; failure to do so may require subse-
quent alterations to the stack or deviation
from the standard procedure. Cases involv-
ing variants are subject to approval by the
Administrator, U.S. Environmental Protec-
tion Agency.
2. Procedure
2.1 Selection of Measurement Site. Sam
pling or velocity measurement is performec
at a site located at least eight stack or due'
diameters downstream and two diameter
upstream from any flow disturbance such a
a bend, expansion, or contraction in th>
stack, or from a visible flame. If neeessarj
an alternative location may be selected, at r
position at least two stack or duct diameter
downstream and a half diameter upstrear'
from any flow disturbance. For a rectangi.
lar cross section, an equivalent diamete
(D,) shall be calculated from the followin.
equation, to determine the upstream an
downstream distances:
2LW
~(L+W)
where Z.=length and W=width.
2.2 Determining the Number of Traver-
Points.
B-4
-------
CO
C71
0.5
50
40
O
a.
LU
V)
cc
LU
30
oc
LU
00
Z
20
z 10
1.0
1.5
* FROM POINT OF ANY TYPE OF
DISTURBANCE (BEND. EXPANSION. CONTRACTION, ETC.)
2.0
2.5
\
A
1
T
"f
-B
i
1
1
- -
t
i
'DISTURBANCE
MEASUREMENT
:- SITE
DISTURBANCE
*
TJ
DUCT DIAMETERS DOWNSTREAM FROM FLOW DISTURBANCE (DISTANCE B)
10
m
I
-------
Chapter I—Environmental Protection Agency
2.2.1 Particulate Traverses. When the
eight- and two-diameter criterion can be
met, the minimum number of traverse
points shall be: (1) twelve, for circular or
rectangular stacks with diameters (or equiv-
alent diameters) greater than 0.61 meter (24
in.); (2) eight, for circular stacks with diam-
eters between 0.30 and 0.61 meter (12-24
in.); (3) nine, for rectangular stacks with
equivalent diameters between 0.30 and 0.61
meter (12-24 in.).
When the eight- and two-diameter crite-
rion cannot be met, the minimum number
of traverse points is determined from Figure
1-1. Before referring to the figure, however,
determine the distances from the chosen
measurement site to the nearest upstream
and downstream disturbances, and divide
each distance by the stack diameter or
equivalent diameter, to determine the dis-
tance in terms of the number of duct diame-
ters. Then, determine from Figure 1-1 the
minimum number of traverse points that
corresponds: (1) to the number of duct di-
App. A
ameters upstream; and (2) to the number of
diameters downstream. Select the higher of
the two minimum numbers of traverse
points, or a greater value, so that for circu-
lar stacks the number is a multiple of 4, and
for rectangular stacks, the number is one of
those shown in Table 1-1.
TABLE 1-1. CROSS-SECTION LAYOUT FOR
RECTANGULAR STACKS
Number of traverse points
g ''
. 12 '.
16
20
25
30
3Q
42 «•
49
Matrix layout
3x3
4x3
4x4
5x4
5x5
6x5
6x6
7x6
7x7
B-6
-------
pa
50
DUCT DIAMETERS UPSTREAM FROM FLOW DISTURBANCE (DISTANCE A)
0.5 1.0 1.5 2.0
40
O
a.
LU
CO
> 30
20
Z
S
Z 10
T
2.5
\
T
A
_t
t
8
A.
—
i
I
DISTURBANCE
MEASUREMENT
- SITE
DISTURBANCE
&
2 34 5 6 7 8 9
DUCT DIAMETERS DOWNSTREAM FROM FLOW DISTURBANCE (DISTANCE B)
10
f
I
Figure 1-2. Minimum number of traverse points for velocity* (nonparticulate) traverses.
-------
Chapter 1—Environmental Protection Agency
APR. A
2.2.2 Velocity (Non-Particulate) Tra-
verses. When velocity or volumetric flow
rate is to be determined (but not participate
matter), the same procedure as that for par-
tlculate traverses (Section 2.2.1) is followed,
except that Figure 1-2 may be used instead
of Figure 1-1.
2.3 Cross-sectional Layout and Location
of Traverse Points.-
2.3.1 Circular Stacks. Locate the traverse
points on two perpendicular diameters ac-
cording to Table 1-2 and the example shown
in Figure 1-3. Any equation (for examples,
see Citations 2 and 3 in the Bibliography)
that gives the same values as those in Table
1-2 maybe used in lieu of Table 1-2.
For participate traverses, one of the diam-
eters must be in a plane containing the
greatest expected concentration variation,
e.g., after bends, one diameter shall be in
the plane of the bend. This requirement be-
comes less critical as the distance from the
disturbance increases; therefore, other di-
ameter locations may be used, subject to ap-
proval of the Administrator.
In addition for stacks having diameters
greater than 0.61 m (24 in.) no traverse
points shall be located within 2.5 centi-
meters (1.00 in.) of the stack walls; and for
stack diameters equal to or less than 0.61 m
(24 in.), no traverse points shall be located
within 1.3 cm (0.50 in.) of the stack walls.
To meet these criteria, observe the proce-
dures given below.
2.3.1.1 Stacks With Diameters Greater
Than 0.61 m (24 in.). When any of the tra-
verse points as located in Section 2.3.1 fall
within 2.5 cm (1.00 in.) of the stack walls, re-
locate them away from the stack walls to:
(1) a distance of 2.5 cm (1.00 in.); or (2) a
distance equal to the nozzle inside diameter,
whichever is larger. These relocated tra-
verse points (on each end of a diameter)
shall be the "adjusted" traverse points.
Whenever two successive traverse points
are combined to form a single adjusted tra-
verse point, treat the adjusted point as two
separate traverse points, both in the sam-
pling (or velocity measurement) procedure,
and in recording the data.
TRAVERSE
POINT
1
2
3
4
S
DISTANCE.
% of diameter
4-4
14.7
29.5
70.5
85.3
95.6
Figure 1-3. Example showing circular stack cross section divided into
12 equal areas, with location of traverse points indicated.
TABLE 1-2. LOCATION OF TRAVERSE POINTS IN CIRCULAR STACKS
[Percent of stack diameter from inside wall to traverse point]
Traverse point number on a diameter
,
3
6 1™.. 1 -
7 «»»«.«.«..««.«
Number of traverse points on a diameter—
2
14.6
85.4
4
6.7
25.0
75.0
83.3
6
4.4
14.6
29.6
70.4
85.4
95.6
8
3.2
10.5
19.4
32.3
67.7
80.6
89.5
10
2.6
8.2
14.6
22.6
34.2
65.8
77.4
12
2.1
6.7
11.8
17.7
25.0
35.6
64.4
14
1.8
5.7
9.9
14.6
20.1
26.9
36.6
16
1.6
4.9
8.5
12.5
16.9
•22.0
28.3
18
1.4
4.4
7.5
10.9
14.6
18.8
23.6
20
1.3
3.9
6.7
9.7
12.9
16.5
20.4
22
1.1
3.5
6.0
8.7
11.6
14.6
18.0
24
1.1
3.2
5.5
7.9
10.5
13.2
16.1
B-8
-------
App. A
Title 40—Protection of Environment
TABLE 1-2. LOCATION OF TRAVERSE POINTS IN CIRCULAR STACKS—Continued
[Percent of stack diameter from inside wall to traverse point]
Traverse point number on a diameter
8 „
g
10 :...
11
12
13
14
15
16
17 ;
18 „
19
20
21
22
23
24
Number of traverse points on a diameter—
2
4
6
8
96.8
10
85.4
91.8
97.4
12
75.0
82.3
88.2
93.3
97.9
14
63.4
73.1
79.9
85.4
90.1
94.3
98.2
16
37.5
62.5
71.7
78.0
83.1
87.5
91.5
95.1
98.4
18
29.6
38.2
61.8
70.4
76.4
81.2
85.4
89.1
92.5
95.6
98.6
20
25.0
30.6
38.8
61.2
69.4
75.0
79.6
83.5
87.1
90.3
93.3
96.1
98.7
22
21.8
26.2
31.5
39.3
60.7
68.5
73.8
78.2
82.0
85.4
88.4
91.3
94.0
96.5
98.9
24
19.4
23.0
27.2
32.3
39.8
60.2
67.7
72.8
77.0
80.6
83.9
86.8
89.5
92.1
94.5
96.8
98.9
2.3.1.2 Stacks With Diameters Equal to
or Less Than 0.61 m (24 in.). Follow the pro-
cedure in Section 2.3.1.1, noting only that
any "adjusted" points should be relocated
away from the stack walls to: (1) a distance
of 1.3 cm (0.50 in.); or (2) a distance equal to
the nozzle inside diameter, whichever is
larger.
2.3.2 Rectangular Stacks. Determine the
number of traverse points as explained in
Sections 2.1 and 2.2 of this method. From
Table 1-1, determine the grid configuration.
Divide the stack cross-section into as many
equal rectangular' elemental areas as tra-
verse points, and then locate a traverse
point at the centroid of each equal area ac-
cording to the example in Figure 1-4.
If the tester desires to use more than the
minimum number of traverse points,
expand the "minimum number of traverse
points" matrix (see Table 1-1) by adding the
extra traverse points along one or the other
or both legs of the matrix; the final matrix
need not be balanced. For example, if a 4x3
"minimum number of points; matrix were
expanded to 36 points, the final matrix
could be 9x4 or 12x3, and would not neces-
sarily have to be 6x6. After constructing the
final matrix, divide the stack cross-section
into as many equal rectangular, elemental
areas as traverse points, and locate a tra-
verse point at the centroid of each equal
area.
The situation of traverse points being too
close to the stack walls is not expected to
arise with rectangular stacks. If this prob-
lem should ever arise, the Administrator
must be contacted for resolution of the
matter.
2.4 Verification of Absence of Cyclonic
Flow. In most stationary sources, the direc-
tion of stack gas flow is essentially parallel
to the stack walls. However, cyclonic flow
may exist (1) after such devices as cyclones
and inertial demisters following venturi
scrubbers, or (2) in stacks having tangential
inlets or other duct configurations which
tend to induce swirling; in these instances,
the presence or absence of cyclonic flow at
the sampling location must be determined.
The following techniques are acceptable for
this determination.
o
o
0
o
r
0
1".
o
/ 1
1
---t-
0 j
I
l~
1 .
0 1
4
o
_
0
" ~™~ ™*"
o
Figure 1 -4. Example showing rectangular stack cross
section divided into 12 equal areas, with a traverse
point at centroid of each area.
Level and zero the manometer. Connect a
Type S pitot tube to the manometer. Posi-
tion the Type S pitot tube at each traverse
point, in succession, so that the planes of
the face openings of the pitot tube are per-
pendicular to the stack cross-sectional
plane; when the Type S pitot tube is in this
position, it is at "0° reference." Note the dif-
ferential pressure (Ap) reading at each tra-
verse point. If a null (zero) pitot reading is
obtained at 0° reference at a given traverse
B-9
-------
Chapter I—Environmental Protection Agency
\
point, an acceptable flow condition exists at
that point. If the pitot reading is not zero at
0* reference, rotate the pltot tube (up to
±90' yaw angle), until a null reading is ob-
tained. Carefully determine and record the
value of the rotation angle (a) to the near-
est degree. After the null technique has
been applied at each traverse point, calcu-
late the average of the absolute values of a;
assign a values of 0* to those points for
which no rotation was reauired, and include
these in the overall average. If the average
value of a is greater than 10% the overall
flow condition in the stack is unacceptable'
and alternative methodology, subject to the
approval of the Administrator, must be used
to perform accurate sample and velocity tra-
verses.
App. A
3. Bibliography
1. Determining Dust Concentration in a
Gas Stream, ASME. Performance Test Code
No. 27. New York, 1957.
2. Devorkin, Howard, et al. Air Pollution
Source Testing Manual. Air Pollution Con-
trol District. Los Angeles, CA. November
1S63.
3. Methods for Determination of Velocity,
Volume, Dust and Mist Content of Gases.
Western Precipitation Division of Joy Man-
ufacturing Co. Los Angeles, CA. Bulletin
WP-50.1968.
4. Standard Method for Sampling Stacks
for Partlculate Matter. In: 1971 Book of
ASTM Standards, Part 23. ASTM Designa-
tion D-2928-71. Philadelphia, Pa. 1971.
5. Hanson, H. A., et al. Particulate Sam-
pling Strategies for Large Power Plants In-
cluding Nonuniform Plow. USEPA, OBD,
ESRL, Research Triangle Park, N.C. EPA-
600/2-76-170, June 1976.
6. Entropy Environmentalists, Inc. Deter-
mination of the Optimum Number of Sam-
pling Points: An Analysis of Method 1 Crite-
ria. Environmental Protection Agency, Re-
search Triangle Park, N.C. EPA Contract
No. 68-01-3172. Task 7.
METHOD 2—DETERMINATION OP STACK GAS
VELOCITY AND VOLUMETRIC PLOW RATE
(TYPE S PITOT TUBE)
1. Principle and Applicability
1.1 Principle. The average gas velocity in
a stack is determined from the gas density
and from measurement of the average veloc-
ity head with a Type S (Stausscheibe or re-
verse type) pitot tube.
1.2 Applicability. This method is applica-
ble for measurement of the average velocity
of a gas stream and for quantifying gas
flow.
This procedure is not applicable at mea-
surement sites which fail to'meet the crite-
ria of Method 1, Section 2.1. Also, the
method cannot be used for direct measure-
ment in cyclonic or swirling gas streams;
Section 2.4 of Method 1 shows how to deter-
mine cyclonic or swirling flow conditions.
When unacceptable conditions exist, alter-
native procedures, subject to the approval
of the Administrator, U.S. Environmental
Protection Agency, must be employed to
make accurate flow rate determinations; ex-
amples of such alternative procedures are:
<1) to Install straightening vanes; (2) to cal-
culate the total volumetric flow rate stoi-
chiometrically, or (3) to move to another
measurement .site at which the flow is ac-
ceptable.
2. Apparatus
Specifications for the apparatus are given
below. Any other apparatus that has been
demonstrated (subject to approval of the
Administrator) to be capable of meeting the
specifications will be considered acceptable.
B-10
-------
App. A
Title 40—Protection of Environment
1.10 • 2.S4 cm*
(9.75-1.0 in.)
•SUGGESTED (INTERFERENCE FREE)
PITOT TUBE THERMOCOUPLE SPACING
Figure 2-1. Type S pilot tube manometer assembly.
2.1 Type S PItot Tube. The Type S pitot
tube (Figure 2-1) shall be made of metal
tubing (e.g. stainless steel). It is recommend-
ed that the external tubing diameter (di-
mension D, Figure 2-2b) be between 0.48
and 0.95 centimeters (%« and % inch). There
shall be an equal distance from the base of
each leg of the pitot tube to its face-opening
plane (dimensions PA and PB Figure 2-2b); it
is recommended that this distance be be-
tween 1.05 and 1.50 times the external
tubing diameter. The face openings of the
pitot tube shall, preferably, be aligned as
shown in Figure 2-2; however, slight misa-
lignments of the openings are permissible
(see Figure 2-3).
The Type S pitot tube shall have a known
coefficient, determined as outlined in .Sec-
tion 4. an identification number shall be as-
signed to the pitot tube; this number shall
be permanently marked or engraved on the
body of the tube.
B-ll
-------
r
Chapter I—Environmental Protection Agency
TRANSVERSE
TUBE AXIS
App. A
\
FACE
-•— OPENING-*-
PLANES
A-SIDE PLANE
LONGITUDINAL
TUBE AXIS *"
t
\
Dt
t
A
B
- I
PB I
NOTE:
1.05 Dt*£P< 1.50 Dt
B-SIDE PLANE
(b)
A ORB
(c)
Figure 2-2. Properly constructed Type S pitot tube, shown in: (a) end view; face opening planes perpendicular
to transverse axis; (b) top view; face opening planes parallel to longitudinal axis; (c) side view; both legs of
equal length and centerlines coincident, when viewed from both sides. Baseline coefficient values of 0.84 may
be assigned to pitot tubes constructed this way.
B-12
-------
APP.A
Title 40—Protection of Environment
TRANSVERSE
TUBE AXIS
LONGITUDINAL
TUBE AXIS
B FLOW
(c)
B FLOW
(d)
(e)
B-13
-------
Chapter I—Environmental Protection Agency
App. A
(f)
(g)
Figure 2-3. Types of face-opening misalignment that can result from field use or improper construction of
Type S pltot tubes. These will not affect the baseline value of Cp(s) so long as al and o2 10°, /31 and 02
S; z 0.32 cm (1/8 In.) and w 0.08 cm (1/32 in.) (citation 11 in Section 6).
A standard pitot tube may be used instead
of a Type S, provided that it meets the
specifications of Sections 2.7 and 4.2; note,
however, that the static and impact pres-
sure holes of standard pitot tubes are. sus-
ceptible to plugging in particulate-laden gas
streams. Therefore, whenever a standard
pitot tube is used to perform a traverse, ade-
quate proof must be furnished that the
openings of the pitot tube have not plugged
up during the traverse period; this can be
done by taking a velocity head (Ap) reading
at the final traverse point, cleaning out the
impact and static holes of the standard
pitot tube by "back-purging" with pressur-
ized air, and then taking another Ap read-
ing. If the Ap readings made before and
after the air purge are the same (±5 per-
cent), the traverse is acceptable. Otherwise,
reject the run. Note that if Ap at the final
traverse point is unsuitably low, another
point may be selected. If "back-purging" at
regular intervals is part of the procedure,
then, comparative Ap readings shall be
taken, as above, for the last two back purges
' at which suitably high Ap readings are ob-
served.
2.2 Differential Pressure Gauge. An in-
clined manometer or equivalent device is
used. Most sampling trains .are equipped
with a 10-in. (water column) Inclined-verti-
cal manometer, having 0.01-in. H,O divisions
on the 0-to 1-in. inclined scale, and 0.1-in.
H,O divisions on the 1- to 10-in. vertical
scale. This type of manometer (or other
gauge of equivalent sensitivity) is satisfac-
tory for the measurement of Ap values as
low as 1.3 mm (0.05 in.) HaO. However, a dif-
ferential pressure gauge of greater sensitiv-
ity shall be used (subject to the approval of
the Administrator), if any of the following
is found to be true: (1) the arithmetic aver-
age of all Ap readings at the traverse points
in the stack is less than 1.3 mm (0.05 in.)
HaO; (2) for traverses of 12 or more points,
more than 10 percent of the individual Ap
readings are below 1.3 mm (0.05 in.) HaO; (3)
for traverses of fewer than 12 points, more
than one Ap reading is below 1.3 mm (0.05
in.) HjO. Citation 18 in Section 6 describes
commercially available instrumentation for
the measurement of low-range gas veloci-
ties.
As an alternative to criteria (1) through
(3) above, the following calculation may be
performed to determine the necessity of
using a more sensitive differential pressure
gauge:
where:
Ap<=Individual velocity head reading at a
traverse point, mm HaO (in. HaO).
n=Tptal number of traverse points.
B-14
-------
App. A
Title 40—Protection of Environment
K=0.13 mm H2O when metrip units are used
and 0.005 in H,O when English units are
used.
If T is greater than 1.05, the velocity head
data are unacceptable and a more sensitive
differential pressure gauge must be used.
NOTE: If differential pressure gauges other
than inclined manometers are used (e.g.,
magnehelic gauges), their calibration must
be checked after'each test series. To check
the calibration of a differential pressure
gauge, compare Ap readings of the gauge
with those of a gauge-oil manometer at a
minimum of three points, approximately
representing the range of Ap values in the
stack. If, at each point, the values of Ap as
read by the differential pressure gauge and
gauge-oil'manometer agree to within 5 per-
cent, the differential pressure gauge shall
be considered to be in proper calibration.
Otherwise, the test series shall either be
voided, or procedures to adjust the meas-
ured Ap values and final results shall be
used subject to the approval of the Adminis-
trator.
2.3 Temperature Gauge. A thermocou-
ple, liquid-filled bulb thermometer, bimetal-
lic thermometer, mercury-in-glass thermom-
eter, or other gauge, capable of measuring
temperature to within 1.5 percent of the
minimum absolute stack temperature shall
be used. The temperature gauge shall be at-
tached to the pitot tube such that the
sensor tip does not touch any metal; the
gauge shall be in an interference-free ar-
rangement with respect to the pitot tube
face openings (see Figure 2-1 and also
Figure 2-7 in Section 4). Alternate positions
may be used if the pitot tube-temperature
gauge system is calibrated according to the
procedure of Section 4. Provided that a dif-
ference of not more than 1 percent in the
average velocity measurement is introduced,
the temperature gauge need not be attached
to the pitot tube; this alternative is subject
to the approval of the Administrator.
2.4 Pressure Probe and Gauge. A piezo-
meter tube and mercury- or water-filled U-
tube manometer capable of measuring stack
pressure to within 2.5 mm (0.1 in.) Hg is
used. The static tap of a standard type pitot
tube or one leg of a Type S pitot tube with
the face opening planes positioned parallel
to the gas flow may also be used as the pres-
sure probe.
2.5 Barometer. A mercury, aneroid, or
other barometer capable of measuring at-
mospheric pressure to within 2.5 mm Hg
(0.1 in. Hg) may be used. In many cases, the
barometric reading may be obtained from a
nearby national weather service station, in
which case the station value (which is the
absolute barometric pressure) shall be re-
quested and an adjustment for elevation dif-
ferences between the weather station and
the sampling point shall be applied at a rate
of minus 2.5 mm (0.1 in.) Hg per 30-meter
(100 foot) elevation increase or vice-versa
for elevation decrease.
2.6 Gas Density Determination Equip-
ment. Method 3 equipment, if needed (see
Section 3.6), to determine the stack gas dry
molecular weight, and Reference Method 4
or Method 5 equipment for moisture con-
tent determination; other methods may be
used subject to approval of the Administra-
tor.
2.7 Calibration Pilot Tube. When calibra-
tion of the Type S pitot tube is necessary
(see Section 4), a standard pitot tube is used
as a reference; The standard pitot tube
shall, preferably, have a known coefficient,
obtained either (1) directly from the Nation-
al Bureau of Standards, Route 270, Quince
Orchard Road, Gaithersburg, Maryland, or
(2) by calibration against another standard
pitot tube with an NBS-traceable coeffi-
cient. Alternatively, a standard pitot tube
designed according to the criteria given in
2.7.1 through 2.7.5 below and illustrated in
Figure 2-4 (see also Citations 7, 8, and 17 in
Section 6) may be used. Pitot tubes designed
according to these specifications will have
baseline coefficients of about 0.99±0.01.
2.7.1 Hemispherical (shown in Figure 2-
4), ellipsoidal, or conical tip.
2.7.2 A minimum of six diameters
straight run (based upon D, the external di-
ameter of the tube) between the tip and the
static pressure holes.
2.7.3 A minimum of eight diameters
straight run between the static pressure
holes and the centerline of the external
tube, following the 90 degree bend.
2.7.4 Static pressure holes of equal size
(approximately 0.1 Z», equally spaced in a
piezometer ring configuration.
2.7.5 Ninety degree bend, with curved or
mitered junction.
2.8 Differential Pressure Gauge for Type
S Pitot Tube Calibration. An inclined mano-
meter or equivalent is used. If the single-ve-
locity calibration technique is employed (see
Section 4.1.2.3), the calibration differential
pressure gauge shall be readable to the
nearest 0.13 mm H,O (0.005 in. H,O). For
multivelocity calibrations, the gauge shall
be readable to the nearest 0.13 mm H,O
(0.005 in HaO) for Ap values between 1.3 and
25 mm HaO (0.05 and 1.0 in. H,O), and to
the nearest 1.3 mm H,O (0.05 in. H3O) for
Ap values above 25 mm H»O (1.0 in. H,O). A
special, more sensitive gauge will be re-
quired to read Ap values below 1.3 mm H,O
10.05 in. H,O] (see Citation 18 in Section 6).
B-15
-------
00
I
t
CURVED OR
MITEREO JUNCTION
STATIC
HOLES
(-0.10)
HEMISPHERICAL
TIP
Figure 2-4. Standard pitot tube design specifications.
-------
App. A
3. Procedure
3.1 Set up the apparatus as shown In
Figure 2-1. Capillary tubing or surge tanks
installed between the manometer and pitot
tube may be used to dampen Ap fluctu-
ations. It is recommended, but not required,
that a pretest leak-check be conducted, as
follows: (1) blow through the pitot impact
opening until at least 7.6 cm (3 in.) H.O ve-
locity pressure registers on the manometer;
then, close off the impact opening. The
pressure shall remain stable for at least 15
seconds; (2) do the same for the static pres-
sure side, except using suction to obtain the
minimum of 7.6 cm (3 in.) H,O. Other leak-
check procedures, subject to the approval of
the Administrator may be used.
3.2 Level and zero the manometer. Be-
cause the manometer level and zero may
Title 40—Protection of Environment
drift due to vibrations and temperature
changes, make periodic checks during the
traverse. Record all necessary data as shown
in the example data sheet (Figure 2-5).
3.3 Measure the velocity head and tem-
perature at the traverse points specified by
Method 1. Ensure that the proper differen-
tial pressure gauge is being used for the
range of Ap values encountered (see Section
2.2). If it is necessary to change to a more
sensitive gauge, do so, and remeasure the Ap
and temperature readings at each traverse
point. Conduct a post-test leak-check (man-
datory), as described in Section 3.1 above, to
validate the traverse run.
3.4 Measure the static pressure in the
stack. One reading is usually adequate.
3.5 Determine the atmospheric pressure.
B-17
-------
Chapter I—Environmental Protection Agency
App. A
PLANT.
DATE.
.RUN NO.
STACK DIAMETER OR DIMENSIONS, m(in.)
BAROMETRIC PRESSURE, mm Hg (in. Hg)
CROSS SECTIONAL AREA. m2(ft2)
OPERATORS
PITOTTUBEI.D.NO.
AVG. COEFFICIENT. Cp = .
LAST DATE CALIBRATED.
SCHEMATIC OF STACK
CROSS SECTION
Traverse
Pi. No.
Vel. Hd.,Ap
mm (in.) H20
Stack Temperature
ts,°C<°FI
Averajt
T$, °K (°B)
,
P9
mm Hg (in.Hg)
Vf Ap
Figure 2-5. Velocity traverse data.
B-18
-------
App. A
Title 40—Protection of Environment
3.6 Determine the stack gas dry molecular
weight. For combustion processes or proc-
esses that emit essentially CO,, O,, CO, and
Na> use Method 3. For processes emitting es-
sentially air, an analysis need not be con-
ducted; use a dry molecular weight of 29.0.
For other processes, other methods, subject
to the approval of the Administrator, must
be used.
3.7 Obtain the moisture content from
Reference Method 4 (or equivalent) or from
Method 5.
3.8 Determine the cross-sectional area of
the stack or duct at the sampling location.
Whenever possible, physically measure the
stack dimensions rather than using blue-
prints.
4. Calibration
4.1 Type S Pitot Tube. Before its initial
use, carefully examine the Type S pitot
tube in top, side, and end views to verify
that the face openings of the tube are
aligned within the specifications illustrated
in Figure 2-2 or 2-3. The pitot tube shall
not be used if it fails to meet these align-
ment specifications.
After verifying the face opening align-
ment, measure and record the following di-
mensions of the pitot tube: (a) the external
tubing diameter (dimension D,, Figure 2-2b);
and (b) the base-to-opening plane distances
(dimensions PA and PB, Figure 2-2b). If D, is
between 0.48 and 0.95 cm (¥ie and % in.) and
if PA and Pe are equal and between 1.05 and
1.50 Di, there are two possible options: (1)
the pitot tube may be calibrated according
to the procedure outlined in Sections 4.1.2
through 4.1.5 below, or (2) a baseline (isolat-
ed tube) coefficient value of 0.84 may be as-
signed to the pitot tube. Note, however, that
if the pitot tube is part of an assembly, cali-
bration may still be required, despite knowl-
edge of the baseline coefficient value (see
Section 4.1.1).
If D,, PA, and Pa are outside the specified
limits, the pitot tube must be calibrated as
outlined in 4.1.2 through 4.1.5 below.
4.1.1 Type S Pitot Tube Assemblies.
During sample and velocity traverses, the
isolated Type S pitot tube is not always
used; in many instances, the pitot tube is
used in combination with other source-sam-
pling components (thermocouple, sampling
probe, nozzle) as part of an "assembly." The
presence of other sampling components can
sometimes affect the baseline value of the
Type S pitot tube coefficient (Citation 9 in
Section 6); therefore an assigned (or other-
wise known) baseline coefficient value may
or may not be valid for a given assembly.
The baseline and assembly coefficient
values will be identical only when the rela-
tive placement of the components in the as-
sembly is such that aerodynamic interfer-
ence effects are eliminated. Figures 2-6
through 2-8 illustrate interference-free
component arrangements for Type S pitot
tubes having external tubing diameters be-
tween 0.48 and 0.95 cm (%e and % in.). Type
S pitot tube assemblies that fail to meet any
or all of the specifications of Figures 2-6
through 2-8 shall be calibrated according to
the procedure outlined in Sections 4.1.2
through 4.1.5 below, and prior to calibra-
tion, the values of the intercomponent spac-
ings (pitot-nozzle, pitot-thermocouple, pitot-
probe sheath) shall be measured and record-
ed.
NOTE: Do not use any Type S pitot tube
assembly which is constructed such that the
impact pressure opening plane of the pitot
tube is below the entry plane of the nozzle
(see Figure 2-6b).
4.1.2 Calibration Setup. If the Type S
pitot tube is to be calibrated, one leg of the
tube shall be permanently marked A, and
the other, B. Calibration shall be done in a
• flow system having the following essential
design features:
B-19
-------
Chapter I—Environmental Protection Agency
APP.A
1*.
TYPE S PITOT TUBE
t x> 1.90 cm (3/4 in.) FOR Dn-1.3 cm (1/2 in.)
SAMPLING NOZZLE
A. IOTTOM VIEW; SHOWING MINIMUM PITOT-NOZZLE SEPARATION.
SAMPLING
PROIE
SAMPLING
NOZZLE
•"
JJL
STATIC PRESSURE
OPENING PLANE
IMPACT PRESSURE
OPENING PLANE
' TYPES
PITOT TUBE
NOZZLE ENTRY
PLANE
•. SIDE VIEW: TO PREVENT PITOT TUBE
FROM INTERFERING WITH GAS FLOW
STREAMLINES APPROACHING THE
NOZZLE. THE IMPACT PRESSURE
OPENING PLANE OF THE PITOT TUBE
SHALL BE EVEN WITH OR ABOVE THE
NOZZLE ENTRY PLANE.
R-20
-------
THERMOCOUPLE
-0-
TYPE SPITOT TUBE
SAMPLE PROBE
THERMOCOUPLE
Z> S.Mem
Win.) "
i
rc - c ^
•?
>
TYPE SPITOT TUBE
, SAMPLE PROBE
CD
CO
I
ro
Figure 2-7. Proper thermocouple placement to prevent interference;
Dr between 0.48 and 0.95 cm (3/16 and 3/8 in.).
TYPE SPITOT TUBE
ilii I
SAMPLE PROBE
Y >7.62 cm (3 in.)
3
Figure 2-8. Minimum pitot-sample probe separation needed to prevent interference; |
n+ hotu/ppn n 48 anri 0 95 rm (3/16 and 3/8 in.). ~
-------
Chapter I—environmental Protection Agency
App. A
4.1.2.1 The flowing gas stream must be
confined to a duct of definite cross-sectional
area, either circular or rectangular. For cir-
cular cross-sections, the minimum duct di-
ameter shall be 30.5 cm (12 in.); for rectan-
gular cross-sections, the width (shorter side).
shall be at least 25.4 cm (10 in.).
4.1.2.2 The cross-sectional area of the
calibration duct must be constant over a dis-
tance of 10 or more duct diameters. For a
rectangular cross-section, use an equivalent
diameter, calculated from the following
equation, to determine the number of duct
diameters:
».=
2 7,ir
Kquaticm 2-1
where:
A=Equivalent diameter
L=Length
W= Width
To ensure the presence of stable, fully de-
veloped flow patterns at the calibration site,
or "test section," the site must be located at
least eight diameters downstream and two
diameters upstream from the nearest distur-
bances.
NOTE: The eight- and two-diameter crite-
ria are not absolute; other test section loca-
tions may be used (subject to approval of
the Administrator), provided that the flow
at the test site is stable and demonstrably
parallel to the duct axis.
4.1.2.3 The flow system shall have the ca-
pacity to generate a test-section velocity
around 915 m/min (3,000 ft/min). This ve-
locity must be constant with time to guaran-
tee steady flow during calibration. Note that
Type S pitot tube coefficients obtained by
single-velocity calibration at 915 m/min
(3,000 ft/min) will generally be valid to
within ±3 percent for the measurement of
velocities above 305 m/min (1,000 ft/min)
and to within ±5 to 6 percent for the mea-
surement of velocities between 180 and 305
m/min (600 and 1,000 ft/min). If a more
precise correlation between Cf and velocity
Is desired, the flow system shall have the ca-
pacity to generate at least four distinct,
time-invariant test-section velocities cover-
ing the velocity range from 180 to 1,525 m/
min (600 to 5,000 ft/min), and calibration
data shall be taken at regular velocity inter-
vals over this range (see Citations 9 and 14
in Section 6 for details).
4.1.2.4 Two entry ports, one each for the
standard and Type S pitot tubes, shall be
cut in the test section; the standard pitot
entry port shall be located slightly down-
stream of the Type S port, so that the
standard and Type S impact openings will
lie in the same cross-sectional plane during
.calibration. To facilitate alignment of the
pitot tubes during calibration, it is advisable
that the test section be constructed of plex-
iglas or some other transparent material.
4.1.3 Calibration Procedure. Note that
this procedure is a general one and must not
be used without first referring to the special
considerations presented in Section 4.1.5.
Note also that this procedure applies only
to single-velocity calibration. To obtain cali-
bration data for the A and B sides of the
Type S pitot tube, proceed as follows:
4.1.3.1 Make sure that the manometer is
properly filled and that the oil is free from
contamination and is of the proper density.
Inspect and leak-check all pitot lines; repair
or replace if necessary.
4.1.3.2 Level and zero the manometer.
Turn on the fan and allow the flow to stabi-
lize. Seal the Type S entry port.
4.1.3.3 Ensure that the manometer is
level and zeroed. Position the standard pitot
tube at the calibration point (determined as
outlined in Section 4.1.5.1), and align the
tube'so that its tip is pointed directly into
the flow. Particular car should be taken in
aligning the tube to avoid yaw and pitch
angles. Make sure that the entry port sur-
rounding the tube is properly sealed.
4.1.3.4 Read A",,,, and record its value in a
data table similar to the one shown in
Figure 2-9. Remove the standard pitot tube
from the duct and disconnect it from the
manometer. Seal the standard entry port.
4.1.3.5 Connect the Type S pitot tube to
the manometer. Open the Type S entry
port. Check the manometer level and zero.
Insert and align the Type S pitot tube so
that its A side impact opening is at the same
point as was the standard pitot tube and is
pointed directly into the flow. Make sure
that the entry port surrounding the tube is
properly sealed.
4.1.3.6 Read Ap. and enter its value in the
data table. Remove the Type S pitot tube
from the duct and disconnect it from the
manometer.
4.1.3.7 Repeat steps 4.1.3.3 through
4.1.3.6 above until three pairs of Ap readings
have been obtained.
4.1.3.8 Repeat steps 4.1.3.3 through
4.1.3.7 above for the B side of the Type S
pitot tube.
4.1.3.9 Perform calculations, as described
in Section 4.1.4 below.
4.1.4 Calculations.
4.1.4.1 For each of the six pairs of Ap
readings (i.e., three from side A and three
from side B) obtained in Section 4.1.3 above,
calculate the value of the Type S pitot tube
coefficient as follows:
B-22
-------
App. A
PITOTTUBE IDENTIFICATION NUMBER:
CALIBRATED BY:, :
Title 40—Protection of Environment
DATE:
RUN NO.
1
2
3
"A" SIDE CALIBRATION
Apstd
cm H20
(in. HaO)
Apfe)
cmH20
-------
Chapter I—Environmental Protection Agency
App. A
'Ap.,.1
P*
Equation 2-2
where:
C,d)=Type S pitot tube coefficient
C,(,U)=Standard pitot tube coefficient; use
0.99 if the coefficient is unknown and
the tube is designed according to the cri-
teria of Sections 2.7.1 to 2.7.5 of this
method.
Ap.u=Velocity head measured by the stand-
ard pitot tube, cm H,O (in. H,O)
Ap.=Velocity head measured by the Type S
pitot tube, cm H,O (in H,O)
4.1.4.2 Calculate Cp (side A), the mean A-
side coefficient, and C? (side B), the mean B-
slde coefficient: calculate the difference be-
tween these two average values.
4.1.4.3 Calculate the deviation of each of
the three A-side values of CPo) fronvQ, (side
A), and the deviation of each B-side value of
Cft,) from Cf Cf (side B). Use the following
equation:
Deviation=Cpl.)—CP(A or B)
Equation 2-3
4.1.4.4 Calculate S. the average deviation
from the mean, for both the A and B sides
of the pitot tube. Use the following equa-
tion:
a (side A or B) =—
Equation 2-4
4.1.4.5 Use the Type S pitot tube only if
the values of S (side A) and S (side B) are
less than or equal to 0.01 and if the absolute
value of the difference between CP (A) and
Cf (B) is 0.01 or less.
4.1.5 Special considerations.
4.1.5.1 Selection of calibration point.
4.1.5.1.1 When an isolated Type S pitot
tube is calibrated, select a calibration point
at or near the center of the duct, and follow
the procedures outlined in Sections 4.1.3
and 4.1.4 above. The Type S pitot coeffi-
cients so obtained, i.e., Cp (side A) and CP
(side B), will be valid, so long as either: (1)
the isolated pitot tube is used; or (2) the
pitot tube is used with other components
(nozzle, thermocouple, sample probe) in an
arrangement that is free from aerodynamic
interference effects (see Figures 2-6
through 2-8).
4.1.5.1.2 For Type S pitot tube-thermo-
couple combinations (without sample
probe), select a calibration point at or near
the center of the duct, and follow the proce-
dures outlined in Sections 4.1.3 and 4.1.4
above. The coefficients so obtained will be
valid so long as the pitot tube-thermocouple
combination is used by itself or with other
components in an interference-free arrange-
ment (Figures 2-6, and 2-8).
4.1.5.1.3 For assemblies with sample
probes, the calibration point should be lo-
cated at or near the center of the duct; how-
ever, insertion of a probe sheath into a
small duct may cause significant cross-sec-
tional area blockage and yield incorrect co-
efficient values (Citation 9 in Section 6).
Therefore, to minimize the blockage effect,
the calibration point may be a few Inches
off-center if necessary. The actual blockage
.effect will be negligible when the theoreti-
cal blockage, as determined by a projected-
area model of the probe sheath, is 2 percent
or less of the duct cross-sectional area for
assemblies without external sheaths (Figure
2-10a), and 3 percent or less for assemblies
with external sheaths (Figure 2-10b).
4.1.5.2 For those probe assemblies in
which pitot tube-nozzle interference is a
factor (i.e., those in which the pitot-nozzle
separation distance fails to meet the specifi-
cation illustrated in Figure 2-6a), the value
of C?M depends upon the amount of free-
space between the tube and nozzle, and
therefore is a function of nozzle size. In
these instances, separate calibrations shall
be performed with each of the commonly
used nozzle sizes in place. Note that the
single-velocity calibration technique is ac-
ceptable for this purpose, even though the
larger nozzle sizes (>0.635 cm or 'A in.) are
not ordinarily used for isokinetic sampling
at velocities around 915 m/min (3,000 ft/
min), which is the calibration velocity; note
also that it is not necessary to draw an Iso-
kinetic sample during calibration (see Cita-
tion 19 in Section 6).
4.1.5.3 For a probe assembly constructed
such that its pitot tube is always used in the
same orientation, only one side of the pitot
tube need be calibrated (the side which will
face the flow). The pitot tube must still
meet the alignment specifications of Figure
2-2 or 2-3, however, and must have, an aver-
age deviation (6) value of 0.01 or less (see
Section 4.1.4.4).
B-24
-------
03
I
ro
en
(a)
ESTIMATED
SHEATH
BLOCKAGE
ElxW "I
lUCT AREAj
(b)
x 100
Figure 2-10. Projected-area m.odels for typical pitot tube assemblies.
a
5*
8.
f
f
•
-------
Chapter I—Environmental Protection Agency
App. A
Figure 2-10. Projected-area models for typi-
cal pitot tube assemblies.
4.1.6 Field Use and Recalibration.
4.1.6.1 Field Use.
4.1.6.1.1 When a Type S pitot tube (iso-
lated tube or assembly) is used in the field,
the appropriate coefficient value (whether
assigned or obtained by calibration) shall be
used to perform velocity calculations. For
calibrated Type S pitot tubes, the A side co-
efficient shall be used when the A side of
the tube faces the flow, and the B side coef-
ficient shall be used when the B side faces
the flow; alternatively, the arithmetic aver-
age of the A and B side coefficient values
may be used, irrespective of which side faces
the flow.
4.1.6.1.2 When a probe assembly is used
to sample a small duct (12 to 36 in. in diame-
ter), the probe sheath sometimes blocks a
significant part of the duct cross-section,
causing a reduction in the effective value of
Cr(j}. Consult Citation 9 in Section 6 for de-
tails. Conventional pitot-sampling probe as-
semblies are not recommended for use in
ducts having inside diameters smaller than
12 Inches (Citation 16 in Section 6).
4.1.6.2 Recalibration.
4.1.6.2.1 Isolated Pitot Tubes. After each
field use, the pitot tube shall be carefully
reexamined in top, side, and end views. If
the pitot face openings are still aligned
within the specifications illustrated in
Figure 2-2 or 2-3, it can be assumed that the
baseline coefficient of the pitot tube has not
changed. If, however, the tube has been
damanged to the' extent that it no longer
meets the specifications of Figure 2-2 or 2-
3, the damage shall either be repaired to re-
store proper alignment of the face openings
or the tube shall be discarded.
4.1.6.2.2 Pitot Tube Assemblies. After
each field use, check the face opening align-
ment of the pitot tube, as in Section
4.1.6.2.1; also, remeasure the intercompon-
ent spacings of the assembly. If the inter-
component spaclngs have not changed and
the face opening alignment is acceptable, it
can be assumed that the coefficient of the
assembly has not changed. If the face open-
Ing alignment is no longer within the speci-
fications of Figures 2-2 or 2-3, either repair
the damage or replace the pitot tube (cali-
brating the new assembly, If necessary). If
the intercomponent spacings have changed,
restore the original spacings or recalibrate
the assembly.
4.2 Standard pitot tube (if applicable). If
a standard pitot tube Is used for the velocity
traverse, the tube shall be constructed ac-
cording to the 'criteria of Section 2.7 and
shall be assigned a baseline coefficient value
of 0.99. If the standard pitot tube is used as
part of an assembly, the tube shall be in an
Interference-free arrangement (subject to
the approval of the Administrator).
4.3 Temperature Gauges. After each
field use, calibrate dial thermometers,
liquid-filled bulb thermometers, thermocou-
ple-potentiometer systems, and other
gauges at a temperature within 10 percent
of the average absolute stack temperature.
For temperatures up to 405° C (761° F), use
an ASTM mercury-in-glass reference ther-
mometer, or equivalent, as a reference; al-
ternatively, either a reference thermocouple
and potentiometer (calibrated by NBS) or
thermometric fixed points, e.g., ice bath and
boiling water (corrected for barometric pres-
sure) may be used. For temperatures above
405° C (761° F), use an NBS-calibrated refer-
ence thermocouple-potentiometer system or
an alternate reference, subject to the ap-
proval of the Administrator.
If, during calibration, the absolute tem-
peratures measured with the gauge being
calibrated and the reference gauge agree
within 1.5 percent, the temperature data
taken in the field shall be considered valid.
Otherwise, the pollutant emission test shall
either be considered invalid or adjustments
(if appropriate) of the test results shall be
made, subject to the approval of the Admin-
istrator.
4.4 Barometer. Calibrate the barometer
used against a mercury barometer.
5. Calculations
Carry out calculations, retaining at least
one extra decimal figure beyond that of the
acquired data. Round off figures after final
calculation. •,
5.1 Nomenclature.
A=Cross-sectional area of stack, m2(ft2).
Bua=Water vapor in the gas stream (from
Method 5 or Reference Method 4), pro-
portion by volume.
CP=Pitot tube coefficient, dimensionless.
Kf=Pitot tube constant,
Q7 JH |" (g/g-mole) (mm Hg) "I'/'
secL (°K)(mmH20) J
for the metric system and
_ft_ P( Ib/lb-mole) (in. Hg)-]'«
sect (°R)(in. H2O) "J
for the English system.
Ma—Molecular weight of stack gas, dry basis
(see Section 3.6) g/g-mole (Ib/lb-mole).
AT,=Molecular weight of stack gas, wet
basis, g/g-mole (Ib/lb-mole).
Bv,) +18.0 Bu,
B-26
-------
App. A
Title 40—Protection of Environment
Equation 2-5
Ptar=Barometric pressure at measurement
site, mm Hg (in. Hg).
PC = Stack static pressure, mm Hg (in. Hg).
P. -Absolute stack gas pressure, mm Hg (in.
Hg).
Equation 2-6
P,ui=Standard absolute pressure, 760 mm
Hg (29.92 in. Hg).
Q«i=Dry volumetric stack gas flow rate cor-
rected to standard conditions, dscm/hr
(dscf/hr).
£,= Stack temperature, °C (°P).
71=Absolute stack temperature, °K, (°R).
=273+t for metric
Equation 2-7
=460+2. for English
Equation 2-8
TM= Standard absolute temperature, 293 °K
(528° R)
v,= Average stack gas velocity, m/sec (ft/
sec).
AP= Velocity head of stack gas, mm H,O (in.
HaO).
3,600= Conversion factor, sec/hr.
18.0= Molecular weight of water, g/g-mole
(Ib/lb-mole).
5.2 Average stack gas velocity.
Equation 2-9
5.3 Average stack gas dry volumetric flow
rate.
C.a=3,600 U-
»t(i
Equation 2-10
6. Bibliography
1. Mark, L. S. Mechanical Engineers'
Handbook. New York, McGraw-Hill Book
Co., Inc. 1951.
2. Perry, J. H. Chemical Engineers' Hand-
book. New York. McGraw-Hill Book Co.,
Inc. 1960.
3. Shigehara, R. T., W. P. Todd, and W. S.
Smith. Significance of Errors in Stack Sam-
pling Measurements. U.S. Environmental
Protection Agency, Research Triangle Park,
N.C. (Presented at the Annual Meeting of
the Air Pollution Control Association, St.
Louis, Mo., June 14-19,1970.)
4. Standard Method for Sampling Stacks
for Particulate Matter. In: 1971 Book of
ASTM Standards, Part 23. Philadelphia, Pa.
1971. ASTM Designation D-2928-71.
5. Vennard, J. K. Elementary Fluid Me-
chanics. New York. John Wiley and Sons,
Inc. 1947.
6. Fluid Meters—Their Theory and Appli-
cation. American Society of Mechanical En-
gineers, New York, N.Y. 1959.
7. ASHRAE Handbook of Fundamentals.
1972. p. 208.
8. Annual Book of ASTM Standards, Part
26. 1974. p. 648.
9. Vollaro, R. F. Guidelines for Type S
Pitot Tube Calibration. U.S. Environmental
Protection Agency. Research Triangle Park,
N.C. (Presented at 1st Annual Meeting,
Source Evaluation Society, Dayton. Ohio,
September 18,1975.)
10. Vollaro, R. F. A Type S Pitot Tube
Calibration Study. U.S. Environmental Pro-
tection Agency, Emission Measurement
Branch, Research Triangle Park, N.C. July
1974.
11. Vollaro, R. F. The Effects of Impact
Opening Misalignment on the Value of the
Type S Pitot Tube Coefficient. U.S. Envi-
ronmental Protection Agency, Emission
Measurement Branch, Research Triangle
Park, N.C. October 1976.
12. Vollaro, R. F. Establishment of a Bas-
line Coefficient Value for Properly Con-
structed Type S Pitot Tubes. U.S. Environ-
mental Protection Agency, Emission Mea-
surement Branch, Research Triangle Park
N.C. November 1976.
13. Vollaro, R. F. An Evaluation of Single-
Velocity Calibration Technique as a Means
of Determining Type S Pitot Tubes Coeffi-
cient. U.S. Environmental Protection
Agency, Emission Measurement, Branch, Re-
search Triangle Park N.C. August 1975.
14. Vollaro, R. F. The Use of Type S Pitot
Tubes for the Measurement of Low Veloci-
ties. U.S. Environmental Protection Agency,
Emission Measurement Branch, Research
Triangle Park N.C. November 1976.
15. Smith, Marvin L. Velocity Calibration
of EPA Type Source Sampling Probe.
United Technologies Corporation, Pratt and
Whitney Aircraft Division, East Hartford,
Conn. 1975.
16. Vollaro, R. F. Recommended Proce-
dure for Sample Traverses in Ducts Smaller
than 12 Inches in Diameter. U.S. Environ-
mental Protection Agency, Emission Mea-
surement Branch, Research Triangle Park
N.C. November 1976.
17. Ower, E. and R. C. Pankhurst. The
Measurement of Air Flow, 4th Ed., London,
Pergamon Press. 1966.
18. Vollaro, R. F. A Survey of Commercial-
ly Available Instrumentation for the Mea-
B-27
-------
Chapter I—Environmental Protection Agency
APP.A
surement of Low-Range Gas Velocities. U.S.
Environmental Protection Agency, Emission
Measurement Branch, Research Triangle
Park N.C. November 1976. (Unpublished
Paper)
19. Gnyp. A. W., C. C. St. Pierre, D. S.
Smith, D. Mozzon.'and J. Steiner. An Ex-
perimental Investigation of the Effect of
Pitot Tube-Sampling Probe Configurations
on the Magnitude of the S Type Pitot Tube
Coefficient for Commercially Available
Source Sampling Probes. Prepared by the
University of Windsor for the Ministry of
the Environment, Toronto, Canada. Febru-
ary 1975.
METHOD 3—GAS ANALYSIS FOR CARBON DIOX-
IDE, OXYGEN, EXCESS AIR, AND DRY MOLEC-
ULAR WEIGHT
1. Principle and Applicability
1.1 Principle. A gas sample is extracted
from a stack, by one of the following meth-
ods: (1) single-point, grab sampling; (2)
single-point, integrated sampling; or (3)
multi-point, Integrated sampling. The gas
sample is analyzed for percent carbon diox-
ide (COi), percent oxygen (Oa), and, if neces-
sary, percent carbon monoxide (CO). If a
dry molecular weight determination is to be
made, either an Orsat or a Pyrite * analyzer
may be used for the analysis; for excess air
or emission rate correction factor determi-
nation, an Orsat analyzer must be used.
1.2 Applicability. This method is applica-
ble for determining CO, and O2 concentra-
tions, excess air, and dry molecular weight
of a sample from a gas stream of a fossil-
fuel combustion process. The method may
also be applicable to other processes where
it has been determined that compounds
'Mention of trade names or specific prod-
ucts does not constitute endorsement by the
Environmental Protection Agency.
other than CO*, O2, CO, and nitrogen (n,)
are not present in concentrations sufficient
to affect the results.
Other methods, as well as modifications to
the procedure described herein, are also ap-
plicable for some or all of the above deter-
minations. Examples of specific methods
and modifications include: (Da multi-point
sampling method using an Orsat analyzer to
analyze individual grab samples obtained at
each point; (2) a method using CO, or O*
and stoichiometric calculations to determine
dry molecular weight and excess air; (3) as-
signing a value of 30.0 for dry molecular
weight, in lieu of actual measurements, for
processes burning natural gas, coal, or oil.
These methods and modifications may be
used, but are subject to the approval of the
Administrator, U.S. Environmental Protec-
tion Agency.
2. Apparatus
As an alternative to the sampling appara-
tus and systems described herein, other
sampling systems (e.g., liquid displacement)
may be used provided such systems are ca-
pable of obtaining a representative sample
and maintaining a constant sampling rate,
and. are otherwise capable of yielding ac-
ceptable results. Use of such systems is sub-
ject to the approval of the Administrator.
2.1 Grab Sampling (Figure 3-1).
2.1.1 Probe. The probe should be made of
stainless steel or borosilicate glass tubing
and should be equipped with an in-stack or
out-stack filter to remove particulate matter
(a plug of glass wool is satisfactory for this
purpose). Any other materials inert to Oa,
CO,, CO, and Ni and resistant to tempera-
ture at sampling conditions may.be used for
the probe; examples of such material are
..aluminum,.copper, quartz glass and Teflon.
2.1.2 Pump. A one-way squeeze bulb, or
equivalent, is used to transport the gas
sample to the analyzer.
2.2 Integrated Sampling (Figure 3-2).
2.2.1 Probe. A probe such as that de-
scribed in Section 2.1.1 is suitable.
B-28
-------
App. A
PROBE
-FILTER (GLASS WOOL)
Tills 40—Protection of Environment
FLEXIBLE TUBING
/^
TO ANALYZER
SQUEEZE BULB
Figure 3-1. Grab-sampling train.
RATE METER
AIR-COOLED
CONDENSER
PROBE
\
FILTER
(GLASS WOOL)
PUMP
QUICK DISCONNECT-
VALVE
RIGID CONTAINER'
BAG
Figure 3-2. Integrated gas-sampling train.
2.2.2 Condenser. An air-cooled or water-
cooled condenser, or other condenser that
will not remove O,, CO,, CO, and N,, may be
used to remove excess moisture which
would interfere with the operation of the
pump and flow meter.
2.2.3 Valve. A needle valve Is used to
adjust sample gas flow rate.
2.2.4 Pump. A leak-free, diaphragm-type
pump, or equivalent, is used to transport
sample gas to the flexible bag. Install a
small surge tank between the pump and
B-29
-------
r
Chapter I—Environmental Protection Agency
App, A
rate meter to eliminate the pulsation effect
of the diaphragm pump on the rotameter.
2.2.5 Rate Meter. The rotameter, or
equivalent rate meter, used should be capa-
ble of measuring flow rate to within ±2 per-
cent of the selected flow rate. A flow rate
range of 500 to 1000 cmVmin is suggested.
2.2.6 Flexible Bag. Any leak-free plastic
(e.g.. Tedlar. Mylar, Teflon) or plastic-
coated aluminum (e.g., aluminized Mylar)
bag) or equivalent, having a capacity con-
sistent with the selected flow rate and time
length 'Of the test run, may be used. A ca-
pacity in the range of 55 to 90 liters is sug-
gested.
To leak-check the bag, connect it to a
water manometer and pressurize the bag to
5 to 10 cm H,O (2 to 4 in. H,O). Allow to
stand for 10 minutes. Any displacement in
the water manometer indicates a leak. An
alternative leak-check method is to pressur-
ize the bag to 5 to 10 cm H,O (2 to 4 in. H,O)
and allow to stand overnight. A deflated bag
indicates a leak.
2.2.7 Pressure Gauge. A water-filled U-
tube manometer, or equivalent, of about 28
cm (12 in.) is used for the flexible bag leak-
check.
2.2.8 Vacuum Gauge. A mercury mano-
meter, or equivalent, of at least 760 mm Hg
(30 in. Hg) is used for the sampling train
leak-check.
2.3 Analysis. For Orsat and Fyrite ana-
lyzer maintenance and operation proce-
dures, follow the instructions recommended
by the manufacturer, unless otherwise spec-
ified herein.
2.3.1 Dry Molecular Weight Determina-
tion. An Orsat analyzer or Fyrite type com-
bustion gas analyzer may be used.
2.3.2 Emission Rate Correction Factor or
Excess Air Determination. An Orsat analyz-
er must be used. For low CO, (less than 4.0
percent) or high O. (greater that 15.0 per-
cent) concentrations, the measuring burette
of the Orsat must have at least 0.1 percent
subdivisions.
3. Drv Molecular Weight Determination
Any of the three sampling and analytical
procedures described below may be used for
determining the dry molecular weight.
3.1 Single-Point, Grab Sampling and
Analytical Procedure.
3.1.1 The sampling point in the duct
shall either be at the centroid of the cross
section or at a point no closer to the walls
than 1.00 m (3.3 ft), unless otherwise speci-
fied by the Administrator.
3.1.2 Set up the equipment as shown in
Figure 3-1, making sure all connections
ahead of the analyzer are tight and leak-
free. If and Orsat analyzer is used, it is rec-
ommended that the analyzer be leaked-
checked by following the procedure in Sec-
tion 5; however, the leak-check is optional.
3.1.3 Place the probe in the stack, with
the tip of the probe positioned at the sam-
pling point; purge the sampling line. Draw a
sample into the analyzer and immediately
analyze it for percent CO, and percent O>.
Determine the percentage of the gas that is
Ni and CO by subtracting the sum of the
percent CO3 and percent O» from 100 per-
cent. Calculate the dry molecular weight as
indicated in Section 6.3.
3.1.4 Repeat the sampling, analysis, and
calculation procedures, until the dry molec-
ular weights of any three grab samples
differ from their mean by no more than 0.3
g/g-mole (0.3 Ib/lb-mole). Average, these
three molecular weights, and report the re-
sults to the nearest 0.1 g/g-mole (Ib/lb-
mole).
3.2 Single-Point, Integrated Sampling
and Analytical Procedure.
3.2.1 The sampling point in the duct
shall be located as specified in Section 3.1.1.
3.2.2 Leak-check (optional) the flexible
bag as in Section 2.2.6.. Set up the equip-
ment as shown in Figure 3-2. Just prior to
sampling, leak-check (optional) the train by
placing a vacuum gauge at the condenser
inlet, pulling a vacuum of at least 250 mm
Hg (10 in. Hg), plugging the outlet at the
quick disconnect, and then turning off the
pump. The vacuum should remain stable for
at least 0.5 minute. Evacuate the flexible
bag. Connect the probe and place it in the
stack, with the tip of the probe positioned
at the sampling point; purge the sampling
line. Next, connect the bag and make sure
that all connections are tight and leak free.
3.2.3 Sample at a constant rate. The sam-
pling run should be simultaneous with, and
for the same total length of time as, the pol-
lutant emission rate determination. Collec-
tion of at least 30 liters (1.00 ft3) of sample
gas Is recommended; however, smaller vol-
umes may be collected, if desired.
3.2.4 Obtain one integrated flue gas
sample during each pollutant emission rate
determination. Within 8 hours after the
sample is taken, analyze it for percent CO,
and percent O, using either an Orsat analyz-
er or a Fyrite-type combustion gas analyzer.
If an Orsat analyzer is used, it is recom-
mended that the Orsat leak-check described
in Section 5 be performed before this deter-
mination; however, the check is optional.
Determine the percentage of the gas that is
Ni and CO by subtracting' the sum of the
percent CO3 and percent O, from 100 per-
cent. Calculate the dry molecular weight as ,
indicated In Section 6.3.
3.2.5 Repeat the analysis and calculation
procedures until the Individual dry molecu-
lar weights for any three analyses differ
from their mean by no more than 0.3 g/g-
mole (0.3 Ib/lb-mole). Average these three
molecular weights, and report the results to
the nearest 0.1 g/g-mole (0.1 Ib/lb-mole).
B.-30
-------
APP- A
3.3 Multi-Point, Integrated Sampling and
Analytical Procedure.
3.3.1 Unless otherwise specified by the
Administrator, a minimum of eight traverse
points shall be used for circular stacks
having diameters less than 0.61 m (24 in.), a
minimum of nine shall be used for rectangu-
lar stacks having equivalent diameters less
than 0.61 m (24 in.), and a minimum of
twelve traverse points shall be used for all
other cases. The traverse points shall be lo-
cated according to Method 1. The use of
fewer points is subject to approval of the
Administrator.
3.3.2 Follow the procedures outlined in
sections 3.2.2 through! 3.2.5, except for the
following: traverse all sampling points and
sample at each point for an equal length of
time. Record sampling data as shown in
Figure 3-3.
4. Emission Rate Correction Factor or
Excess Air Determination
NOTE: A Fyrite-type combustion gas ana-
lyzer is not acceptable for excess air or emis-
Title 40—Protection of Environment
sion rate correction factor determination,
unless approved by the Administrator, if
both percent CO, and percent O, are meas-
ured, the analytical results of any of the
three procedures given below may also be
used for calculating the dry. molecular
weight.
Each of the three procedures below shall
be used only when specified in an applicable
subpart of the standards. The use of these
procedures for other purposes must have
specific prior approval of the Administrator.
4.1 Single-Point, Grab Sampling and
Analytical Procedure.
4.1.1 The sampling point in the duct
shall either be at the centroid of the cross-
section or at a point no closer to the walls
than 1.00 m (3.3 ft), unless otherwise speci-
fied by the Administrator.
4.1.2 Set up the equipment as shown in
Figure 3-1, making sure all connections
ahead of the analyzer are tight and leak-
free. Leak-check the Orsat analyzer accord-
ing to the procedure described in Section 5.
This leak'Check is mandatory.
TIME
TRAVERSE
PT.
AVERAGE
Q
1pm
%DEV.a
avf
(MUST BE < 18%)
Figure 3-3. Sampling rate data.
4.1.3 Place the probe in the stack, with
the tip of the probe positioned at the sam-
pling point; purge the sampling line. Draw a
sample into the analyzer. For emission rate
correction factor determination, immediate-
ly analyze the sample, as outlined in Sec-
tions 4.1.4 and 4.1.5, for percent CO, or per-
cent Oa. If excess air is desired, proceed as
follows: (1) immediately analyze the sample,
as in Sections 4.1.4 and 4.1.5, for percent
COa, Oa, and CO; (2) determine the percent-
age of the gas that is N: by subtracting the
B-31
-------
Chapter I—Environmental Protection Agency
App. A
turn of the percent CO>. percent O,, and per-
cent CO from 100 percent; and (3) calculate
percent excess air as outlined in Section 6.2.
4.1.4 To Insure complete absorption of
the CO., Oi, or if applicable, CO, make re-
peated passes through each absorbing solu-
tion until two consecutive readings are the
same. Several passes (three or four) should
be made between readings. (If constant
readings cannot be obtained after three con-
lecutive readings, replace the absorbing so-
lution.)
4.1.5 After the analysis is completed,
leak-check (mandatory) the Orsat analyzer
once again, as described in Section 5. For
the results of the analysis to be valid, the
Orsat analyzer must pass this leak test
before and after the analysis.
NOTE: Since this single-point, grab sam-
pling and analytical procedure in normally
conducted in conjunction with a single-
point, grab sampling and analytical proce-
dure for a pollutant, only one analysis is or-
dinarily conducted. Therefore, great care
must be taken to obtain a valid sample and
analysis. Although in most cases only CO,
or O« Is required, It is recommended that
both COi and O. be measured, and that Ci-
tation 5 in the Bibliography be used to vali-
date the analytical data.
4.2 Single-Point, Integrated Sampling
and Analytical Procedure.
4.2.1 The sampling point in the duct
•hall be located as specified in Section 4.1.1.
4.2.2 Leak-check (mandatory) the flexi-
ble bag as In Section 2.2.6. Set up the equip-
ment as shown In Figure 3-2. Just prior to
sampling, leak-check (mandatory) the train
by placing a vacuum gauge at the condenser
inlet, pulling a vacuum of a least 250 mm Hg
(10 in. Hg), plugging the outlet at the quick
disconnect, and then turning off the pump.
The vacuum shall remain stable for at least
0.6 minute. Evacuate the flexible bag. Con-
nect the probe and place it in the stack,
with the tip of the probe positioned at the
sampling point; purge the sampling line.
Next, connect the bag and make sure that
all connections are tight and leak free.
4.2.3 Sample at a constant rate, or as
specified by the Administrator. The sam-
pling run must be simultaneous with, and
for the same total lengh of time as, the pol-
lutant emission rate determination. Collect
at least 30 liters (1.00 fts) of sample gas.
Smaller volumes may be collected, subject
to approval of the Administrator.
4.2.4 Obtain one integrated flue gas
sample during each pollutant emission rate
determination. For emission rate correction
factor determination, analyze the sample
within 4 hours after it is taken for percent
CO. or percent O, (as outlined In Sections
4.2.5 through 4.2.7). The Orsat analyzer
must be leak-check (see Section 5) before
the analysis. If excess air Is desired, proceed
as follows: (1) within 4 hours after the
sample is taken, analyze it (as in Sections
4.2.5 through 4.2.7) for percent CO., O,, and
CO; (2) determine the percentage of the gas
that is Na by subtracting the sum of the per-
cent CO,, percent O,, and percent CO from
100 percent; (3) calculate percent excess air,
as outlined in Section 6.2.
4.2.5 To insure complete absorption of
the CO,, O,, or if applicable, CO, make re-
peated passes through each absorbing solu-
tion until two consecutive readings are the
same. Several passes (three of four) should
be make between readings. (If constant
readings cannot be obtained after three con-
secutive readings, replace the absorbing so-
lution.)
4.2.6 Repeat the analysis until the fol-
lowing criteria are met:
4.2.6.1 For percent CO., repeat the ana-
lytical procedure until the results of any
three analyses differ by no more that (a) 0.3
percent by volume when CO, is greater than
4.0 percent or (b) 0.2 percent by volume
when CO3 is less than or equal to 4.0 per-
cent. Average the three acceptable values of
percent CO, and report the results to the
nearest 0.1 percent.
4.2.6.2 For percent Oa, repeat the analyt-
ical procedure until the results of any three
analyses differ by no more than (a) 0.3 per-
cent by volume when O, is less than 15.0
percent or (b) 0.2 percent by volume when
O. is greater than or equal to 15.0 percent.
Average the three acceptable values of per-
cent O, and report the results to the nearest
0.1 percent.
4.2.6.3 For percent CO, repeat the ana-
lytical procedure until the results of any
three analyses differ by no more than 0.3
percent. Average the three acceptable
values of percent CO and report the results
to the nearest 0.1 percent.
4.2.7 After the analysis is completed,
leak-check (mandatory) the Orsat analyzer
once again, as described in Section 5. For
the results of the analysis to be valid, the
Orsat analyzer must pass this leak test
before an after the analysis.
NOTE: Although in most Instances only
CO, or O, is required, it is recommended
that both CO, and O, be measured, and that
Citation 5 in the Bibliography be used to
validate the analytical data.
4.3 Multi-Point, Integrated Sampling and
Analytical Procedure.
4.3.1 Both the minimum number of sam-
pling points and the sampling point location
shall be as specified in Section 3.3.1 of this
method. The use of fewer points than speci-
fied is subject to the approval of the Admin-
istrator.
4.3.2 Follow the procedures outlined in
Sections 4.2.2 through 4.2.7, except for the
following: Traverse all sampling points and
R-32
-------
App. A
Title 40—Protection of Environment
sample at each point for an equal length of
time. Record sampling data as shown in
Figure 3-3.
5. Leak-Check Procedure for Orsat Analysers
Moving an Orsat analyzer frequently
causes it to leak. Therefore, an Orsat ana-
lyzer should be throughly leak-checked on
site before the flue gas sample is introduced
into it. The procedure for leak-checking an
Orsat analyzer is:
5.1.1 Bring the liquid level in each pi-
pette up to the reference mark on the capil-
lary tubing and then close the pipette stop-
cock.
5.1.2 Raise the leveling bulb sufficiently
to bring the confining liquid meniscus onto
the graduated portion of the burette and
then close the manifold stopcock.
5.1.3 Record the meniscus position.
5.1.4 Observe the menicus in the burette
and the liquid level in the pipette for move-
ment over the next 4 minutes.
5.1.5 For the Orsat analyzer to pass the
leak-check, two conditions must be met.
5.1.5.1 The liquid level in each pipette
must not fall below the bottom of the capil-
lary tubing during this 4-minute interval.
5.1.5.2 The meniscus in the burette must
not change by more than 0.2 ml during this
4-minute interval.
5.1.6 If the analyzer fails the leak-check
procedure, all rubber connections and stop-
cocks should be checked until the cause of
the leak is identified. Leaking stopcocks
must be disassembled, cleaned, and re-
greased. Leaking rubber connections must
be replaced. After the analyzer is reassem-
bled, the leak-check procedure must be re-
peated.
6. Calculations
6.1 Nomenclature.
j|&=Dry molecular weight, g/g-mole Ub/lb-
mole).
%EA=Percent excess air.
%COa=Percent CO2 by volume (dry basis).
%Oj=Pereent O, by volume (dry basis).
%CO=Percent CO by volume (dry basis).
%Ni=Percent Na by volume (dry basis).
0.264=Ratio of Oa to N, in ah-, v/v.
0.280=Molecular weight of N, or CO, divid-
ed by 100.
0.320=Molecular weight of O* divided by
100.
0.440=Moleeular weight of CO, divided by
100.
6.2 Percent Excess Air. Calculate the per-
cent excess air (if applicable), by substitut-
ing the appropriate values of percent O?,
CO, and N3 (obtained from Section 4.1.3 or
4.2.4) into Equation 3-1.
%02-0.5%CO I
N2 (%02-0.5 %CO) J1UU
Equation 3-1
NOTE: The equation above assumes that
ambient air is used as the source of O, and
that the fuel does not contain appreciable
amounts of N2 (as do coke oven or blast fur-
nace gases). For those cases when apprecia-
ble amounts of N? are present (coal, oil, and
natural gas do not contain appreciable
amounts of Na) or when oxygen enrichment
is used, alternate methods, subject to ap-
proval of the Administrator, are required.
6.3 Dry Molecular Weight. Use Equation
3-2 to calculate the dry molecular weight of
the stack gas
M((=0.440(%COJ)+0.320(%Oa)+
0.280(%Na+%CO)
Equation 3-2
NOTE: The above equation does not consid-
er argon in air (about 0.9 percent, molecu-
lars weight of 37.7). A negative error of
about 0.4 percent is introduced. The tester
may opt to include argon in the analysis
using procedures subject to approval of the
Administrator. ,
7. Bibliography .
1. Altshuller, A. P. Storage of Gases and
Vapors in Plastic Bags. International Jour-
nal of Air and Water Pollution. 6:75-81.
1963.
2. Conner, William D. and J. S. Nader. Air
Sampling with Plastice Bags. Journal of the
American Industrial Hygiene Association.
25:291-297. 1964.
3. Burrell Manual for Gas Analysts, Sev-
enth edition. Burrell Corporation, 2223
Fifth Avenue, Pittsburgh, Pa. 15219. 1951.
4. Mitchell, W.. J. and M. R. Midgett. Field
Reliability of the Orsat Analyzer. Journal
of Air Pollution Control Association 26:491-
495. May 1976.
5. Shigehara, R. T., R. M. Neulicht, and
W. S. Smith. Validating Orsat Analysis Data
from Fossil Fuel-Fired Units. Stack Sam-
pling News. 4(2):21-26. August, 1976.
METHOD 4—DETERMINATION OF MOISTURE
CONTENT IN STACK GASES
1. Principle and Applicability
1.1 Principle. A gas sample is extracted
at a constant rate from the source; moisture
is removed from the sample stream and de-
termined either volumetrically or gravime-
trically.
1.2 Applicability. This method is applica-
ble for determining the moisture content of
stack gas.
Two procedures are given. The first is a
reference method, for accurate determina-
B.-33
-------
Chapter I—Environmental Protection Agency
APP.A
lions of moisture content (such as are
needed to calculate emission data). The
second is an approximation method, which
provides estimates of percent moisture to
kid in setting isokinetic sampling rates prior
to a pollutant emission measurement run.
The approximation method described
herein is only a suggested approach; alter-
native means for approximating the mois-
ture content, e.g., drying tubes, wet bulb-dry
bulb techniques, condensation techniques,
stoichiometrlc calculations, previous experi-
ence, etc., are also acceptable.
The reference method is often conducted
simultaneously with a pollutant emission
measurement run; when it is, calculation of
percent isokinetic, pollutant emission rate,
etc., for the run shall be based upon the re-
sults of the reference method or its equiva-
lent; these calculations shall not be based
upon the results of the approximation
method, unless the approximation method
is shown, to the satisfaction of the Adminis-
trator, U.S. Environmental Protection
Agency, to be capable of yielding results
within 1 percent H»O of the reference
method.
NOTE: The reference method may yield
questionable results when applied to satu-
rated gas streams or to streams that contain
water droplets. Therefore, when these con-
ditions exist or are suspected, a second de-
termination of the moisture content shall
be made simultaneously with the reference
method, as follows: Assume that the gas
stream is saturated. Attach a temperature
sensor [capable of measuring to ±1° C (2*
F)] to the reference method probe. Measure
the stack gas temperature at each traverse
point (see Section 2.2.1) during the refer-
ence method traverse; calculate the average
stack gas temperature. Next, determine the
moisture percentage, either by: (1) using a
psychrometric chart and making' appropri-
ate corrections if stack pressure is different
from that of the chart, or (2) using satura-
tion vapor pressure tables. In cases where
the pyschrometric chart or the saturation
vapor pressure tables are not applicable
(based on evaluation of the process), alter-
nate methods, subject to the approval of the
Administrator, shall be used.
2. Reference Method
The procedure described in Method 5 for
determining moisture content is acceptable
as a reference method.
2.1 Apparatus. A schematic of the sam-
pling train used in this reference method is
shown in Figure 4-1. All components shall
be .maintained and calibrated according to
the procedure outlined in Method 5.
B-34
-------
CO
on
FILTER
(EITHER IN STACK
OR OUT OF STACK)
STACK
WALL
CONDENSER-ICE BATH SYSTEM INCLUDING
SILICA GEL TUBE—j
MAIN VALVE
AIR-TIGHT
PUMP
V
>
o
o
Figure 4-1. Moisture sampling train-reference method.
-------
Chapter I—Environmental Protection Agency
APP.A
2.1.1 Probe. The probe Is constructed of
stainless steel or glass tubing, sufficiently
heated to prevent water condensation, and
Is equipped with a filter, either in-stack
(e.g., a plug of glass wool Inserted Into the
end of the probe) or heated out-stack (e.g.,
as described in Method 5), to remove partic-
ular matter.
When stack conditions permit, other
metals or plastic tubing may be used for the
probe, subject to the approval of the Admin-
istrator.
2.1.2 Condenser. The condenser consists
of four impingers connected in series with
ground glass, leak-free fittings or any simi-
larly leak-free non-contaminating fittings.
The first, third, and fourth impingers shall
be of the Greenburg-Smith design, modified
by replacing the tip with a 1.3 centimeter
(V4 inch) ID glass tube extending to about
1.3 cm (V4 in.) from the bottom of the flask.
The second impinger shall be of the Green-
burg-Smith design with the standard tip.
Modifications (e.g., using flexible connec-
tions between the Impingers, using materi-
als other than glass, or using flexible
vacuum lines to connect the filter holder to
the condenser) may be used, subject to the
approval of the Administrator.
The first two impingers shall contain
known volumes of water, the third shall be
empty, and the fourth shall contain a
known weight of 6- to 16-mesh indicating
type silica gel, or equivalent desiccant. If
the silica gel has been previously used, dry
at 175° C (350° P) for 2 hours. New silica gel
may be used as received. A thermometer, ca-
pable of measuring temperature to within 1°
C (2* F), shall be placed at the outlet of the
fourth impinger, for monitoring purposes.
Alternatively, any system may be used
(subject to the approval of the Administra-
tor) that cools the sample gas stream and
allows measurement of both the water that
has been condensed and the moisture leav-
ing the condenser, each to within 1 ml or 1
g. Acceptable means are to measure the con-
densed water, either gravlmetrically or volu-
metrically, and to measure the moisture
leaving the condenser by: (1) monitoring the
temperature and pressure at the exit of the
condenser and using Dalton's law of partial
pressures, or (2) passing the sample gas
stream through a tared silica gel (or equiva-
lent desiccant) trap, with exit gases kept
below 20' C (68' P), and determining the
weight gain.
If means other than silica gel are used to
determine the amount of moisture leaving
the condenser, it Is recommended that silica
gel (or equivalent) still be used between the
condenser system and pump, to prevent
moisture condensation In the pump and me-
tering devices and to avoid the need to make
corrections for moisture in the metered
volume.
2.1.3 Cooling System. An ice bath con-
tainer and crushed ice (or equivalent) are
used to aid in condensing moisture.
2.1.4 Metering System. This system in-
cludes a vacuum gauge, leak-free pump,
thermometers capable of measuring tem-
perature to within 3° C (5.4° P), dry gas
meter capable of measuring volume to
within 2 percent, and related equipment as
shown in Figure 4-1. Other metering sys-
tems, capable of maintaining a constant
sampling rate and determining sample gas
volume, may be used, subject to the approv-
al of the Administrator.
2.1.5 Barometer. Mercury, aneroid, or
other barometer capable of measuring at-
mospheric pressure to within 2.5 mm Hg
(0.1 in. Hg) may be used. In many cases, the
barometric reading may be obtained from a
nearby national weather service station, in
which case the station value (which is the
absolute barometric pressure) shall be re-
quested and an adjustment for elevation dif-
ferences between the weather station and
the sampling point shall be applied at a rate
of minus 2.5 mm Hg (0.1 in. Hg) per 30 m
(100 ft) elevation increase or vice versa for
elevation decrease.
2.1.6 Graduated Cylinder and/or Bal-
ance. These items are used to measure con-
densed water and moisture caught in the
silica gel to within 1 ml or 0.5 g. Graduated
cylinders shall have subdivisions no greater
than 2 ml. Most laboratory balances are ca-
pable of weighing to the nearest 0.5 g or
less. These balances are suitable for use
here.
2.2 Procedure. The following procedure
is written for a condenser system (such as
the impinger system described in Section
2.1.2) incorporating volumetric analysis to
measure the condensed moisture, and silica
gel and gravimetric analysis to measure the
moisture leaving the condenser.
2.2.1 Unless otherwise specified by the •
Administrator, a minimum of eight traverse
points shall be used for circular stacks
having diameters less than 0.61 m (24 in.), a
minimum of nine points shall be used for
rectangular stacks having equivalent diame-
ters less than 0.61 m (24 in.), and a mini-
mum of twelve traverse points shall be used
in all other cases. The traverse points shall
be located according to Method 1. The use
of fewer points is subject to the approval of
the Administrator. Select a suitable probe
and probe length such that all traverse
points can be sampled. Consider sampling
from opposite sides of the stack (four total
sampling ports) for large stacks, to permit
use of shorter probe lengths. Mark the
probe with heat resistant tape or by some
other method to denote the proper distance
into the stack or duct for each sampling
point. Place known volumes of water in the
first two impingers. Weigh and record the
B-36
-------
App. A
weight of the silica gel to the nearest 0.5 g,
and transfer the silica gel to the fourth im-
pinger; alternatively, the silica gel may first
be transferred to the impinger, and the
weight of the silica gel plus impinger record-
ed.
2.2.2 Select a total sampling time such
that a minimum total gas volume of 0.60
scm (21 scf) will be collected, at a rate no
greater than 0.021 mVmin (0.75 cfm). When
both moisture content and pollutant emis-
sion rate are to be determined, the moisture
determination shall be simultaneous with,
and for the same total length of time as, the
pollutant emission rate run, unless other-
wise specified in an applicable subpart of
the standards.
2.2.3 Set up the sampling train as shown
in Figure 4-1. Turn on the probe heater and
(if applicable) the filter heating system to
temperatures of about 120° C (248° P), to
prevent water condensation ahead of the
condenser; allow time for the temperatures
to stabilize. Place crushed ice in the ice bath
container. It is recommended, but not re-
quired, that a leak check be done, as follows:
Disconnect the probe from the first im-
pinger or (if applicable) from the filter
holder. Plug the inlet to the first impinger
(or filter holder) and pull a 380 mm (15 in.)
Hg vacuum; a lower vacuum may be used,
provided that it is not exceeded during the
test. A leakage rate in excess of 4 percent of
the average sampling rate or 0.00057 mV
min (0.02 cfm), whichever is less, is unaccep-
table. Following the leak check, reconnect
the probe to the sampling train.
2.2.4 During the sampling run, maintain
a sampling rate within 10 percent of con-
Till* 40—Protection of Environment
stant rate, or as specified by the Adminis-
trator. For each run, record the data re-
quired on the example data sheet shown to
Figure 4-2. Be sure to record the dry gas
meter reading at the beginning and end of
each sampling time increment and when-
ever sampling is halted. Take other appro-
priate readings at each sample point, at
least once during each time increment.
2.2.5 To begin sampling, position the
probe tip at the first traverse point. Imme-
diately start the pump and adjust the flow
to the desired rate. Traverse the cross sec-
tion, sampling at each traverse point for an
equal length of time. Add more ice and, if
necessary, salt to maintain a temperature of
less 20° C (68° F) at the silica gel outlet.
2.2.6 After collecting the sample, discon-
nect the probe from the filter holder (or
from the first impinger) and conduct a leak
check (mandatory) as described in Section
2.2.3. Record the leak rate. If the leakage
rate exceeds the allowable rate, the tester
shall either reject the test results or shall
correct the sample volume as in Section 6.3
of Method 5. Next, measure the volume of
the moisture condensed to the nearest ml.
Determine the increase in weight of the
silica gel (or silica gel plus impinger) to the
nearest 0.5 g. Record this information (see
example data sheet, Figure 4-3) and calcu-
late the moisture percentage, as described in
2.3 below.
2.3 Calculations. Carry out the following
calculations, retaining at least one extra
decimal figure beyond that of the acquired
data. Round off figures after final calcula-
tion.
B-37
-------
I
;ecAnw
OPERATOR
BATf
RUN NO.
AMIIEKT TEMPERATURE.
IAROMETRIC PRESSURE-
moiE LENGTH n(W
SCHEMATIC OF STACK CROSS SECTION
CO
00
TRAVERSE POINT
NUMIER
TOTAL
SAMPLING
TIME
(ei.min.
AVERAGE
STACK •
TEMPERATURE
•C(«F|
PRESSURE
DIFFERENTIAL
ACROSS
ORIFICE METER
(AH),
•"•(in.) H{0
METER
READING
GAS SAMPLE
VOLUME
m) Ift3)
AVm
»'(tt3)
GAS SAMPLE TEMPERATURE
AT DRY GAS METER
INLET
(Tmj.l.'Ct'F)
An.
A*
OUTLET
(TiiMtl.'CCF)
**
TEMPERATURE
OF GAS
LEAVING
CONDENSER OR
LAST IMPINGER.
•C(«F)
Figure 4-2. Field moisture determination-reference method.
I
i
I
\
a
5*
Jr
-------
APR. A
FINAL
INITIAL
DIFFERENCE
IMPINGE;:
VOLUME.
ml '
SILICA GEL
WEIGHT.
9
Figure 4-3. Analytical data • reference method.
Title 40—Protection of Environment
where:
_K\=0.001333 m'/ml for metric units
=0.04707 ftVml for English units
2.3.3 Volume of water vapor collected in
silica gel.
2.3.1 Nomenclature.
B „=Proportion of water vapor, by volume,
in the gas stream.
Af«,=Molecular weight of water, 18.0 g/g-
mole (18.0 Ib/lb-mole).
/>»,=Absolute pressure (for this method,.
same as barometric pressure) at the dry
gas meter, mm Hg (in. Hg).
P,,,j=Standard absolute pressure, 760 mm
Hg (29.92 in. Hg).
j£=Ideal gas constant, 0.06236 (mm Hg)
(m3)/(g-mole) (°K) for metric units and
21.85 (in. Hg) (fts)/(lb-mole> (°R) for
English units.
T«= Absolute temperature at meter, °K
CR).
T.w=Standard absolute temperature, 293°
K (528"R).
Vw=Dry gas volume measured by dry gas
meter, dcm (dcf).
A Vm=Incremental dry gas volume measured
by dry gas meter at each traverse point,
dcm (dcf).
V»<«=Volume of water vapor condensed
corrected to standard conditions, scm
(scf).
V«.f<«d>=Voiume of water vapor collected in.
silica gel corrected to standard condi-
tions, scm (scf).
V/=Pinal volume of condenser water, ml.
V,=Initial volume, if any, of condenser
water, ml.
Wi=Final weight of silica gel or silica gel
• plus impinger, g.
Wi=Initial weight of silica gel or silica gel
plus impinger, g.
Y =Dry gas meter calibration factor.
p«=Density of water, 0.9982 g/ml (0.002201
Ib/ml).
2.3.2 Volume of water vapor condensed.
(Vr-Vi)i>»RTmt.t
Equation 4-2
where:
K,=0.001335 m»/g for metric units
=0.04715 ftVg for English units
2.3.4 Sample gas volume.
VmP*
Equation 4-3
where:
JT,=0.3858 °K/mm Hg for metric units
=17.64 °R/in. Hg for English units
NOTE: If the post-test lead rate (Section
2.2.6) exceeds the allowable rate, correct the
value of Vm in Equation 4-3, as described in
Section 6.3 of Method 5.
2.3.5 Moisture Content.
_
•"*
Vice («td) + Vv it (»td) .
c <«td) + Vir., (ltd) + V« (•")
Equation 4-4
Equation 4-1
NOTE: In saturated or moisture droplet-
laden gas streams, two calculations of the
moisture content of the stack gas shall be
made, one using a value based upon the
saturated conditions (see Section 1.2), and
another based upon the results of the im-
pinger analysis. The lower of these two
values of B „ shall be considered correct.
2.3.6 Verification of constant sampling
rate. For each time increment, determine
the AVm. Calculate the average. If the value
B-39
-------
Chapter I—Environmental Protection Agency
APP.A
tor any time Increment differs from the
average by more than 10 percent, reject the
results and repeat the run.
3. Approximation Method
The approximation method described
below Is presented only as a suggested
method (see Section 1.2).
3.1 Apparatus.
3.1.1 Probe. Stainless steel glass tubing,
sufficiently heated to prevent water conden-
sation and equipped with a filter (either in-
stock or heated out-stack) to remove partlc-
ulate matter. A plug of glass wool, inserted
into the end of the probe, is a satisfactory
filter.
3.1.2 Impingers. Two midget impingers,
each with 30 ml capacity, or equivalent.
3.1.3 Ice Bath. Container and ice, to aid
in condensing moisture in impingers.
3.1.4 Drying Tube. Tube packed with
new or regenerated 6- to 16-mesh indicating-
type silica gel (or equivalent desiccant), to
dry the sample gas and to protect the meter
and pump.
3.1.5 Valve. Needle valve, to regulate the
sample gas flow rate.
3.1.6 Pump. Leak-free, diaphragm type,
or equivalent, to pull the gas sample
through the train.
3.1.7 Volume Meter. Dry gas meter, suffi-
ciently accurate to measure the sample
volume within 2%, and calibrated over the
range of flow rates and conditions actually
encountered during sampling.
3.1.8 Rate Meter. Rotameter, to measure
the flow range from 0 to 3 1pm (0 to 0.11
cfm).
3.1.9 Graduated Cylinder. 25 ml.
3.1.10 Barometer. Mercury, aneroid, or
other barometer, as described in Section
2.1.5 above.
3.1.11 Vacuum Gauge. At least 760 mm
Hg (30 in. Hg) gauge, to be used for the sam-
pling leak check.
3.2 Procedure.
3.2.1 Place exactly 5 ml distilled water in
each Impinger.
Leak check the sampling train as follows:
Temporarily insert a vacuum gauge at or
near the probe inlet; then, plug the probe
inlet and pull a vacuum of at least 250 mm
Hg (10 in. Hg). Note; the time rate of
change of the dry gas meter dial; alterna-
tively, a rotameter (0-40 cc/min) may be
temporarily attached to the dry gas meter
outlet to determine the leakage rate. A leak
rate not In excess of 2 percent of the aver-
.age sampling rate is acceptable.
NOTE: Carefully release the probe inlet
plug before turning off the pump.
B-40
-------
HEATED PROBE
SILICA GEL TUBE
RATE METER,
VALVE
co
FILTER
(GLASS WOOL)
ICE BATH
MIDGET IMPINGERS
PUMP
Figure 4-4. Moisture-sampling train • approximation method.
3
3T
f
8.
IW
-------
LOCATION.
TEST
COMMENTS
DATE
OPERATOR
BAROMETRIC PRESSURE
5
a
•a
m
3
I
I
-p.
ro
CLOCK TIME
GAS VOLUME THROUGH
METER. (Vm).
m3 (ft3)
RATE METER SETTING
m3/min. (ft3/min.)
METER TEMPERATURE.
°C (°F)
Piniirn 4_R Pialrl mrhic+ura rla+Armina+i/\n _ or\nr/\vimat!nn nnathrkrl
1
to
3
•3
•o
-------
App. A
Title 40—Protection of Environment
Figure 4-5. Field moisture determination-
approximation method. .
3.2.2 Connect the probe, insert it into the
stack, and sample at a constant rate of 2
1pm (0.071 cfm). Continue sampling until
the dry gas meter registers about 30 liters
(1.1 ft9) or until visible liquid .droplets are
carried over from the first impinger to the
second. Record temperature, pressure, and
dry gas meter readings as required by
Figure 4-5.
3.2.3 After collecting the sample, com-
bine the contents of the two impingers and
measure the volume to the nearest 0.5 ml.
3.3 Calculations. The calculation method
presented is designed to estimate the mois-
ture in the stack gas; therefore, other data,
which are only necessary for accurate mois-
ture determinations, are not collected. The
following equations adequately estimate the
moisture content, for the purpose of deter-
mining isokinetic sampling rate settings.
3.3.1 Nomenclature.
B«™=Approximate proportion, by volume,
of water vapor in the gas stream leaving
the second impinger, 0.025.
Bwt=Water vapor in the gas stream, propor-
tion by volume.
Aft=Molecular weight of water, 18.0 g/g-
mole (18.0 Ib/lb-mole).
Pm= Absolute pressure (for this method,
same as barometric pressure) at the dry
gas meter.
Pii=Final volume of impinger contents, ml.
V/=Initial volume of impinger contents, ml.
V«= Dry gas volume measured by dry gas
meter, dcm (def).
Vm(«d)=Dry gas volume measured by dry gas
meter, corrected to standard conditions,
dscm (dscf).
V»c<«=Volume of water vapor condensed,
corrected to standard conditions, scm
(scf).
p«=Density of water, 0.9982 g/ml (0.002201
Ib/ml).
Y=Dry gas meter calibration factor.
3.3.2 Volume of water vapor collected.
where: \
_(Vf-V,)p,rKT.tl
' «•'.— s »7
= Kl(Vf-Vi)
Kquation 4-5
K,=0.001333 m'/ml for metric units
=0.04707 ftVml for English units.
3.3.3 Gas volume.
V.P.
Kquation 4-(i
where:
#,=0.3858 °K/mm Hg for metric units
= 17.64 °R/in. Hg for English units
3.3.4 Approximate moisture content.
r "*m
Equation 4-7
4. Calibration
4.1 For the reference method, calibrate
equipment as specified in the following sec-
tions of Method 5: Section 5.3 (metering
system); Section 5.5 (temperature gauges);
and Section 5.7 (barometer). The recom-
mended leak check of the metering system
(Section 5.6 of Method 5) also applies to the
reference method. For the approximation
method, use the procedures outlined in Sec-
tion 5.1.1 of Method 6 to calibrate the me-
tering system, and the procedure of Method
5, Section 5.7 to calibrate the barometer.
5. Bibliography
1. Air Pollution Engineering Manual
(Second Edition). Danielson, J. A. (ed.). U.S.
Environmental Protection Agency, Office of
Air Quality Planning and Standards. Re-
search Triangle Park, N.C. Publication No.
AP-40. 1973.
2. Devorkin, Howard, et al. Air Pollution
Source Testing Manual. Air Pollution Con-
trol District, Los Angeles, Calif. November,
1963.
3. Methods for Determination of Velocity,
Volume Dust and Mist Content of Gases.
Western Precipitation Division of Joy Man-
B-43
-------
Chapter I—Environmental Protection Agency
App. A
ufocturing Co., Los Angeles, Calif. Bulletin
WP-50.1968.
METHOD ^DETERMINATION OF PARTICULATE
ISitissibNS FROM STATIONARY SOURCES
1. Principle and Applicability
1.1 Principle. Particulate matter is with-
drawn isokinetically from the source and
collected on a glass fiber filter maintained
tt a temperature in the range of 120±14° C
(248±25* F) or such other temperature as
specified by an applicable subpart of the
standards or approved by Administrator,
OJS. Environmental Protection Agency, for
* particular application. The participate
mass, which includes any material that con-
denses at or above the filtration tempera-
ture, is determined gravimetrically after re-
moval of uncombined water.
1.2 Applicability. This method is applica-
ble for the determination of particulate
emissions from stationary sources.
2. Apparatus
2.1 Sampling Train. A schematic of the
sampling train used in this method is shown
in Figure 5-1. Complete construction details
are given in APTD-0581 (Citation 2 in Sec-
tion 7); commercial models of this train are
also available. For changes from APTD-
0581 and for allowable modifications of the
train shown in Figure 5-1, see the following
subsections.
The operating and maintenance proce-
dures for the sampling train are described in
APTD-0576 (Citation 3 in Section 7). Since
correct usage is important in obtaining valid
results, all users should read APTD-0576
and adopt the operating and maintenance
procedures outlined in it, unless otherwise
specified herein. The sampling train con-
sists of the following components:
B-44
-------
TEMPERATURE SENSOR
PROBE
TEMPERATURE
SENSOR
/•r
IMPINGER TRAIN OPTIONAL, MAY BE REPLACED
BY AN EQUIVALENT CONDENSER
*
•o
HEATED AREA
PITOTTUBE
PROBE
THERMOMETER
FILTER HOLDER
en
i
en
REVERSE-TYPE
PITOT TUBE
THERMOMETER
PITOT MANOMETER
ORIFICE
IMPINGERS ICE BATH
BY-PASS VALVE
THERMOMETERS
VACUUM
GAUGE
MAiN VALVE
DRY GAS METER AIR-TIG.HT
PUMP
Figure 5-1. Particulate-samplinK train
CHECK
VALVE
VACUUM
LINE
HI
I
I
-------
Chapter I—Environmental Protection Agency
App. A
2.1.1 Probe Nozzle. Stainless steel (316)
or glass With sharp, tapered leading edge.
The angle of taper shall be 30° and the
taper shall be on the outside to preserve a
constant internal diameter. The probe
nozzle shall be of the button-hook or elbow
design, unless otherwise specified by the Ad-
ministrator. If made of stainless steel, the
nozzle shall be constructed from seamless
tubing; other materials of construction may
be used, subject to the approval of the Ad-
ministrator.
A range of nozzle sizes suitable for isokin-
etic sampling should be available, e.g., 0.32
to 1.27 cm <% to % in.)—or larger if higher
volume sampling trains are used—inside di-
ameter (ED) nozzles in increments of 0.16 cm
(%• in.). Each nozzle shall be calibrated ac-
cording to the procedures outlined in Sec-
tion 5.
2.1.2 Probe Liner. Borosilicate or quartz
glass tubing with a heating system capable
of maintaining a gas temperature at the exit
end during sampling of 120±14° C (248±25°
P), or such other temperature as specified
by an applicable subpart of the standards or
approved by the Administrator for a partic-
ular application. (The tester may opt to op-
erate the equipment at a temperature lower
than that specified.) Since the actual tem-
perature at the outlet of the probe Is not
usually monitored during sampling, probes
constructed according to APTD-0581 and
utilizing the calibration curves of APTD-
0576 (or calibrated according to the proce-
dure outlined in APTD-0576) will be consid-
ered acceptable.
Either borosllicate or quartz glass probe
liners may be used for stack temperatures
up to about 480' C (900' P) quartz liners
shall be used for temperatures between 480
and 900* C (900 and 1,650° F). Both types of
liners may be used at higher temperatures
than specified for short periods of time, sub-
ject to the approval of the Administrator.
The softening temperature for borosilicate
is 820° C (1,508° P), and for quartz it is
1,500° C (2,732° P).
Whenever practical, every effort should
be made to use borosilicate or quartz glass
probe liners. Alternatively, metal liners
(e.g., 316 stainless steel, Incoloy 825,* or
other corrosion resistant metals) made of
seamless tubing may be used, subject to the
approval of the Administrator.
2.1.3 Pitot Tube. Type S, as described in
Section 2.1 of Method 2, or other device ap-
proved by the Administrator. The pltot tube
shall be attached to the probe (as shown in
Figure 5-1) to allow constant monitoring of
the stack gas velocity. The impact (high
pressure) opening plane of the pitot tube
•Mention of trade names or specific prod-
uct does not constitute endorsement by the
Environmental Protection Agency.
shall be even with or above the nozzle entry
plane (see Method 2, Figure 2-6b) during
sampling. The Type S pitot tube assembly
shall have a known coefficient, determined
as outlined In Section 4 of Method 2.
2.1.4 Differentia Pressure Gauge. In-
clined manometer or equivalent device
(two), as described in Section 2.2 of Method
2. One manometer shall be used or velocity
head (Ap) readings, and the other, for orifice
differentia pressure readings.
2.1.5 Filter Holder. Borosilicate glass,
with a glass frit filter support and a silicone
rubber gasket. Other materials of construc-
tion (e.g., stainless steel, Teflon, Viton) may
be used, subject to .approval of the Adminis-
trator. The holder design shall provide a
positive seal against leakage from the out-
side or around the filter. The holder shall
be attached immediately at the outlet of the
probe (or cyclone, it used).
2.1.6 Filter Heating System. Any heating
system capable of maintaining a tempera-
ture around the filter holder during sam-
pling of 120±14° C (248±25° F), or such
other temperature as specified by an appli-
cable subpart of the standards or approved
by the Administrator for a particular appli-
cation. Alternatively, the tester may opt to
operate the equipment at a temperature
lower than that specified. A temperature
gauge capable of measuring temperature to
within 3° C (5.4° P) shall be installed so that
the temperature around the filter holder
can be regulated and monitored during sam-
pling. Heating systems other than the one
shown in APTD-05'81 may be used.
2.1.7 Condenser. The following system
shall be used to determine the stack gas
moisture content: Four implngers connected
in series with leak-free ground glass fittings
or any similar leak-free non-contaminating
fittings. The first, third, and fourth im-
pingers shall be of the Greenburg-Smith
design, modified by replacing the tip with
1.3 cm (Va In.) ID glass tube extending to
about 1.3 cm (V4 in.) from the bottom of the
flask. The second, impinger shall be of the
Greenburg-Smith design with the standard
tip. Modifications (e.g., using flexible con-
nections between the impingers, using mate-
rials other than glass, or using flexible
vacuum lines to connect the filter holder to
the condenser) may be used, subject to the
approval of the Administrator. The first and
second Implngers shall contain known quan-
tities of water (Section 4.1.3), the third shall
be empty, and the fourth shall contain a
known weight of silica gel, or equivalent
desiccant. A thermometer, capable of meas-
uring temperture to within 1° C (2° F) shall
be placed at the outlet of the fourth im-
pinger for monitoring purposes.
Alternatively, any system that cools the
sample gas stream and allows measurement
of the water condensed and moisture leav-
B-46
-------
App. A
Title 40—Protection of Environment
ing the condenser, each to within 1 ml or 1 g
may be used, subject to the approval of the
Administrator. Acceptable means are to
measure the condensed water either gravi-
metrically or volumetrically and to measure
the moisture leaving the condenser by: (1)
monitoring the temperature and pressure at
the exit of the condenser and using Dalton's
law of partial pressures; or (2) passing the
sample has stream through a tared silica gel
(or equivalent desiccant) trap with exit
gases kept below 20° C (68° P) and determin-
ing the weight gain.
If means other than silica gel are used to
determine the amount of moisture leaving
the condenser, it is recommended that silica
gel (or equivalent) still be used between the
condenser system and pump to prevent
moisture condensation in the pump and me-
tering devices and to avoid the need to make
corrections for moisture in the metered
volume.
NOTE: If a determination of the partieu-
late matter collected in the impingers is de-
sired in addition to moisture content, the
impinger system described above shall be
used, without modification. Individual
States or control agencies requiring this in-
formation shall be contacted as to the
sample recovery and analysis of the im-
pinger contents.
2.1.8 Metering System. Vacuum gauge,
leak-free pump, thermometers capable of
measuring temperature to within 3° C (5.4°
F), dry gas meter capable of measuring
volume to within 2 percent, and related
equipment, as shown in Figure 5-1. Other
metering systems capable of maintaining
sampling rates within 10 percent of isokine-
tic and of determining sample volumes to
within 2 percent may be used, subject to the
approval of the Administrator. When the
metering system is used in conjunction with
a pitot tube, the system shall enable checks
of isokinetic rates.
Sampling trains utilizing metering sys-
tems designed .for higher flow rates than
that decribed in APTD-0581 or APDT-0576
may be used provided that the specifica-
tions of this method are met.
2.1.9 Barometer. Mercury aneroid, or
other barometer capable of measuring at-
mospheric pressure to within 2.5 mm Hg
(0.1 in. Hg). In many cases the barometric
reading may be obtained from a nearby na-
tional weather service station, in which case
the station value (which is the absolute
barometric pressure) shall be requested and
an adjustment for elevation differences be-
tween the weather station and sampling
point shall be applied at a rate of minus 2.5
mm Hg (0.1 in. Hg) per 30 m (100 ft) eleva-
tion increase or vice versa for elevation de-
crease.
2.1.10 Gas Density Determination Equip-
ment. Temperature sensor and, pressure
gauge, as described in Sections 2.3 and 2.4 of
Method 2, and gas analyzer, if necessary, as
described in Method 3. The temperature
sensor shall, preferably, be permanently at-
tached to the pitot tube or sampling probe
in a fixed configuration, such that the,tip of
the sensor extends beyond the leading edge
of the probe sheath and does not touch any
metal. Alternatively, the sensor may be at-
tached just prior to use in the field. Note,
however, that if the temperature sensor is
attached in the field, the sensor must be
placed in an interference-free arrangement
with respect to the Type S pitot tube open-
ings (see Method 2, Figure 2-7). As a second
alternative, if a difference of not more than
1 percent in the average,velocity measure-
ment is to be introduced, the temperature
gauge need not be attached to the probe or
pitot tube. (This alternative is subject to the
approval of the Administrator.)
2.2 Sample Recovery. The following
items are needed.
2.2.1 Probe-Liner and Probe-Nozzle
Brushes. Nylon bristle brushes with stain-
less steel wire handles. The probe brush
shall have extensions (at least as long as the
probe) of stainless steel. Nylon, Teflon, or
similarly inert material. The brushes shall
be properly sized and shaped to brush out
the probe liner and nozzle.
2.2.2 Wash Bottles—Two. Glass wash
bottles are recommended; polyethylene
wash bottles may be used at the option of
the tester. It is recommended that acetone
not be stored in polyethylene bottles for
longer than a month.
2.2.3 Glass Sample Storage Containers.
Chemically resistant, borosilicate glass bot-
tles, for acetone washes, 500 ml or 1000 ml.
Screw cap liners shall either be rubber-
backed Teflon or shall be constructed so as
to be leak-free and resistant to chemical
attack by acetone. (Narrow mouth glass bot-
tles have been found to be less prone to
leakage.) Alternatively, polyethylene bottles
may be used.
2.2.4 Petri Dishes. For filter samples,
glass or polyethylene, unless otherwise spec-
ified by the Administrator.
2.2.5 Graduated Cylinder and/or Bal-
ance. To measure condensed water to within
1 ml or 1 g. Graduated cylinders shall have
subdivisions no greater than 2 ml. Most lab-
oratory balances are capable of weighing to
the nearest 0.5 g or less. Any of these bal-
ances is suitable or use here and in Section
2.3.4.
2.2.6 Plastic Storage Containers. Air-
tight containers to store silica gel.
2.2.7 Funnel and Rubber Policeman. To
aid in transfer of silica gel to container; not
necessary if silica gel is weighed in the field.
2.2.8 Funnel. Glass or polyethylene, to
aid in sample recovery.
B-47
-------
Chapter I—Environmental Protection Agency
APR. A
2.2 Analysis. For analysis, the following
equipment is needed.
2.3.1 Glass Weighing Dishes.
2.3.2 Desiccator.
2.3.3 Analytical Balance. To measure to
within 0.1 mg.
2.3.4 Balance. To measure to within 0.5 g.
2.3.5 Beakers. 250 ml.
2.3.6 Hygrometer. To measure the rela-
tive humidity of the laboratory environ-
ment.
2.3.7 Temperature Gauge. To measure
the temperature of the laboratory environ-
ment.
3. Reagents
3.1 Sampling. The reagents used in sam-
pling are as follows:
3.1.1 Filters. Glass fiber filters, without
organic binder, exhibiting at least 99.95 per-
cent efficiency <<0.05 percent penetration)
on 0.3-mlcron dioctyl phthalate smoke parti-
cles. The filter efficiency test shall be con-
ducted in accordance with ASTM standard
method D 2986-71. Test data from the sup-
plier's quality control program are suffi-
cient for this purpose. In sources containing
SO, or SO,, the filter material must be of a
type that is unreactlve to SO, or SO,. Cita-
tion 10 in Section 7 may be used' to select
the appropriate filter.
3.1.2 Silica Gel. Indicating type, 6 to 16
mesh. If previously used, dry at 175° C (350°
F) for 2 hours. New silica gel may be used as
received. Alternatively, other types of desic-
cants (equivalent or better) may be used,
lubject to the approval of the Administra-
tor.
3.1.3 Water. When analysis of the mate-
rial caught in the impingers is required, dis-
tilled water shall be used. Run blanks prior.
to field use to eliminate a high blank on test
samples.
3.1.4 Crushed Ice.
3.1.5 Stopcock Grease. Acetone-insoluble,
heat-stable silicone grease. This is not neces-
sary if screw-on connectors with Teflon
sleeves, or similar, are used. Alternatively,
other types of stopcock grease may be used,
subject to the approval of the Administra-
tor.
3.2 Sample Recovery. Acetone-reagent
grade, <0.001 percent residue, in glass bot-
tles—is required. Acetone from metal con-
tainers generally has a high residue blank
and should not be used. Sometimes, suppli-
ers transfer acetone to glass bottles from
metal containers; thus, acetone blanks shall
be run prior to field use and only acetone
with low blank values (<0.001 percent) shall
be used. In no case shall a blank value of
greater than 0.001 percent of the weight of
acetone used be subtracted from the sample
weight.
3.3 Analysis. Two reagents-are required
for the analysis:
3.3.1 Acetone. Same as 3.2.
3.3.2 Desiccant. Anhydrous calcium sul-
fate, indicating type. Alternatively, other
types of desiccants may be used, subject to
the approval of the Administrator.
4. Procedure
4.1 Sampling. The complexity of this
method is such that, in order to obtain reli-
able results, testers should be trained and
experienced with the test procedures.
4.1.1 Pretest Preparation. All the compo-
nents shall be maintained and calibrated ac-
cording to the procedure described in
APTD-0576, unless otherwise specified
herein.
Weigh several 200 to 300 g portions of
silica gel in air-tight containers to the near-
est 0.5 g. Record the total weight of the
silica gel plus container, on each container.
As an alternative, the silica gel need not be
preweighed, but may be weighed directly in
the impinger or sampling holder just prior
to train assembly.
Check filters visually against light for ir-
regularities and flaws or pinhole leaks.
Label filters of the proper diameter on the
back side near the edge using numbering
machine ink. As an alternative, label the
shipping containers (glass or plastic petri
dishes) and keep the filters in these contain-
ers at all times except during sampling and
weighing.
Desiccate the filters at 20±5.6° C <68±10'
F) and-ambient pressure for at least 24
hours and weigh at intervals of at least 6
hours to a constant weight, i.e., 0.5 mg
change from previous weighing; record re-
sults to the nearest 0.1 mg. During each
weighing the filter must not be exposed to
the laboratory atmosphere for a period
greater than 2 minutes and a relative hu-
midity above 50 percent. Alternatively
(unless otherwise specified by the Adminis-
trator), the filters may be oven dried at 105*
C (220° F) for 2 to 3 hours, desiccated for 2
hours, and weighed. Procedures other than
those described, which account for relative
humidity effects, may be used, subject to
the approval of the Administrator.
4.1.2 Preliminary Determinations. Select
the sampling site and the minimum number
of sampling points according to Method 1 or
as specified by the Administrator. Deter-
mine the stack pressure, temperature, and
the range of velocity heads using Method 2;
it is recommended that a leak-check of the
pitot lines (see Method 2, Section 3.1) be
performed. Determine the moisture content
using Approximation Method 4 or its alter-
natives for the purpose of making isokinetic
•sampling rate settings. Determine the stack
gas dry molecular weight, as described in
Method 2, Section 3.6; if integrated Method
3 sampling is used for molecular weight de-
termination, the integrated bag sample
shall be taken simultaneously with, and for
R-48
-------
App. A
the same total length of time as, the partic-
ulate sample run. •'
Select a nozzle size based on the range of
velocity heads, such that it is not necessary
to change the nozzle size in order to main-
tain isokinetic sampling rates. During the
run, do not change the nozzle size. Ensure
that the proper differental pressure gauge
is chosen for the range of velocity heads en-
countered (see Section 2.2 of Method 2).
Select a suitable probe liner and probe
length such that all traverse points can be
sampled. For large stacks, consider sampling
from opposite sides of the stack to reduce
the length of probes.
Select a total sampling time greater than
or equal to the minimum total sampling
time specified in the test procedures for the
specific industry such that (1) the sampling
time per point is not less than 2 min (or
some greater time interval as specified by
the Administrator), and (2) the sample
volume taken (corrected to standard condi-
tions) will exceed the required minimum
total gas sample volume. The latter is based
on an approximate average sampling rate.
It is recommended that the number of
minutes sampled at each point be an integer
or an integer plus one-half minute, in order
to avoid timekeeping errors. The sampling
time at each point shall be the same.
In some cirumstances, e.g., batch cycles, it
may be necessary to sample for shorter
times at the traverse points and to obtain
smaller gas sample volumes. In these cases,
the Administrator's approval must first be
obtained.
4.1.3 Preparation of Collection Train.
During preparation and assembly of the
sampling train, keep all openings where con-
tamination can occur covered until just
prior to assembly or until sampling is about
to begin.
Place 100 ml of water in each of the first
two impingers, leave the third impinger
empty, and transfer approximately 200 to
300 g of preweighed silica gel from its con-
tainer to the fourth impinger. More silica
gel may be used, but care should be taken to
ensure that it is not entrained and carried
out 'from the impinger during sampling.
Place the container in a clean place for later
use in the sample recovery. Alternatively,
the weight of the silica gel plus impinger
may be determined to the nearest 0.5 g and
recorded.
Using a tweezer or clean disposable surgi-
cal gloves, place a labeled (identified) and
weighed filter in the filter holder. Be sure
that the filter is property centered and the
gasket properly placed so as to prevent the
sample gas stream from circumventing the
filter. Check the filter for tears after assem-
bly is completed.
When glass liners are used, install the se-
lected nozzle using a Viton A O-ring when
stack temperatures are less than 260° C
Title 40—Protection of Environment
(500° P) and an asbestos string gasket when
temperatures are higher. See APTD-0576
for details. Other connecting systems using
either 316 stainless steel or Teflon ferrules
may be used. When metal liners are used,
install the nozzle as above or by a leak-free
direct mechanical connection. Mark the
probe with heat resistant tape or by some
other method to denote the proper distance
into the stack or duct for each sampling
point.
Set up the train as in Figure 5-1, using (if
necessary) a very light coat of silicone
grease on all ground glass joints, greasing
only the outer portion (see APTD-0576) to
avoid possibility of contamination by the
silicone grease. Subject to the approval of
the Administrator, a glass cyclone may be
used between the probe and filter holder
when the total participate catch is expected
to exceed 100 mg or when water droplets are
present in the stack gas.
Place crushed ice around the impingers.
4.1.4 Leak-Check Procedures.
4.1.4.1 Pretest Leak-Check. A pretest
leak-check is recommended, but not re-
quired. If the tester opts to conduct the pre-
test leak-check, the following procedure
shall be used.
After the sampling train has been assem-
bled, turn on and set the filter and probe
heating systems at the desired operating
temperatures. Allow time for the tempera-
tures to stabilize. If a Viton A O-ring or
other leak-free connection is used in assem-
bling the probe nozzle to the probe liner,''
leak-check the train at the sampling site by
plugging the nozzle and pulling a 380 mm
Hg (15 in. Hg) vacuum.
NOTE: A lower vacuum may be used, pro-
vided that it Is not exceeded during the test.
If an asbestos string is used, do not con-
nect the probe to the train during the leak-
check. Instead, leak-check the train by first
plugging the inlet to the filter holder
(cycone, if applicable) and pulling a 380 mm
Hg (15 in. Hg) vacuum (see Note immediate-
ly above). Then connect the probe to the
train and leak-check at about 25 mm Hg (1
in. Hg) vacuum; alternatively, the probe
may be leak-checked with the rest of the
sampling train, in one step, at 380 mm Hg
(15 in. Hg) vacuum. Leakage rates in excess
of 4 percent of the average sampling rate or
'•"0.00057 m'/min (0.02 cfm), whichever is less,
are unacceptable.
The following leak-check instructions for
the sampling train described in APTD-0576
and APTD-0581 may be helpful.'Start the
pump with bypass valve fully open and
coarse adjust valve, completely closed. Par-
tially open the coarse adjust valve and
slowly close the bypass valve until the de-
sired vacuum is reached. Do not reverse di-
rection of bypass value; this will cause water
B-49
-------
Chapter I—Environmental Protection Agency
APP.A
to back up into the filter holder. If the de-
sired vacuum is exceeded, either leak-check
»t this higher vacuum or end the leak-check
as shown below and start over.
When the leak-check is completed, first
slowly remove the plug from the inlet to the
probe, filter holder, or cyclone (if applica-
ble) and immediately turn off the vacuum
pump. This prevents the water in the im-
plngers from being forced backward into the
filter holder and silica get from being en-
trained backward into the third impinger.
4.1.4.2 Leak-Checks During Sample Run.
If, during the sampling run, a component
(e.g., filter assembly or Impinger) change be-
comes unecessary, a leak-check shall be con-
ducted immediately before the change is
made. The leak-check shall be done accord-
ing to the procedure outlined in Section
4.1.4.1 above, except that it shall be done at
a vacuum equal to or greater than the maxi-
mum value recorded up to that point in the
test. If the leakage rate is found to be no
greater than 0.00057 mVmin (0.02 cfm) or 4
percent of the average sampling rate
(whichever is, less), the results are accept-
able, and no correction will need to be ap-
plied to the total volume of dry gas metered;
if, however, a higher leakage rate is ob-
tained, the tester shall either record the
leakage rate and plan to correct the sample
volume as shown in Section 6.3 of this
method, or shall void the sampling run.
Immediately after component changes,
leak-checks are optional; If such leak-checks
are done, the procedure outlined in Section
4.1,4.1 above shall be used.
4.1.4.3 Post-test Leak-Check. A leak-
check is mandatory at the conclusion of
each sampling run. The leakcheck shall be
done in accordance with the procedures out-
lined In Section 4.1.4.1, except that it shall
be conducted at a vacuum equal to or great-
er than the maximum value reached during
the sampling run. If the leakage rate is
found to be no greater than 0.00057 mVmin
(0.02 cfm) or 4 percent of the average sam-
pling rate (whichever is less), the results are
acceptable, and no correction need be ap-
plied to the total volume of dry gas metered.
If, however, a higher leakage rate is ob-
tained, the tester shall either record the
leakage rate and correct the sample volume
as shown in Section 6.3 of this method, or
shall void the sampling run.
4.1.5 Participate Train Operation.
During the sampling run, maintain an iso-
kinetic sampling rate (within 10 percent of
true isokinetic unless otherwise specified by
the Administrator) and a temperature
around the filter of 120±14° C (248±25° P),
or such other temperature as specified by
an applicable subpart of the standards or
approved by the Administrator.
For each run, record the data required on
a data sheet such as the one shown in
Figure 5-2. Be sure to record the initial dry
gas meter reading. Record the dry gas meter
readings at the beginning and end of each
sampling time increment, when changes in
flow rates are made, before and after each
leak-check, and when sampling is halted.
Take other readings required by Figure 5-2
at least once at each sample point during
each time increment and additional read-
ings when significant changes (20 percent
variation in velocity head readings) necessi-
tate additional adjustments in flow rate.
Level and zero the manometer. Because the
manometer level and zero may drift due to
vibrations and temperature changes, make
periodic checks during the traverse.
Clean the portholes prior to the test run
to minimize the chance of sampling deposit-
ed material. To begin sampling, remove the
nozzle cap, verify that the filter and probe
heating systems are up to temperature, and
that the pitot tube and probe are properly
positioned. Position the nozzle at the first
traverse point with the tip pointing directly
into the gas stream. Immediately start the
pump and adjust the flow to isokinetic con-
ditions. Nomographs are available, which
aid in the rapid adjustment of the isokinetic
sampling rate without excessive computa-
tions. These nomographs are designed for
use when the Type S pitot tube coefficient
is 0.85+0.02, and the stack gas equivalent
density (dry molecular weight) is equal to
29db4. APTD-0576 details the procedure for
using the nomographs. If CP and Ma are out-
side the above stated ranges do not use the
nomographs unless appropriate steps (see
Citation 7 in Section 7) are, taken to com-
pensate for the deviations.
B-50
-------
PLANT
LOCATION
OPERATOR,
DATE
RUN NO
SAMPLE BOX NO..
METER BOX NO._
METERAH®
C FACTOR
AMBIENT TEMPERATURE.
BAROMETRIC PRESSURE-
ASSUMED MOISTURE. X_
PROBE LENGTH,™ (ft)
PITOT TUBE COEFFICIENT, Cp.
SCHEMATIC OF STACK CROSS SECTION
NOZZLE IDENTIFICATION NO
AVERAGE CALIBRATED NOZZLE DIAMETER, cm (in.).
PROBE HEATER SFTTIMr.
LEAK RATE,m3/mii>.(cfm)
PROBE LINER MATERIAL
STATIC PRESSURE, mm Hg (in. Hg)_
FILTER NO
CO
I
en
TRAVERSE POINT
. NUMBER
TOTAL
AVERAGE
SAMPLING
TIME
(0), min.
VACUUM
mm Hg
(in. Hg)
STACK
TEMPERATURE
-------
Chapter I—Environmental Protection Agency
App.A
When the stack is under significant nega-
tive pressure (height of impinger stem),
take care to close the coarse adjust valve
before inserting the probe into the stack to
prevent water from backing into the filter
holder. If necessary, the pump may be
turned on with the coarse adjust valve
closed.
When the probe is in position, block off
the openings around the probe and porthole
to prevent unrepresentative dilution of the
gas stream.
Traverse the stack cross-section, as re-
quired by Method 1 or as specified by the
Administrator, being careful not to bump
the probe nozzle into the stack walls when
sampling near the walls or when removing
or inserting the probe through the port-
holes; this minimizes the chance of extract-
ing deposited material.
During the test run, make periodic adjust-
ments to keep the temperature around the
filter holder at the proper level; add more
ice and, if necessary, salt to maintain a tem-
perature of less than 20' C (68° F) at the
condenser/silica gel outlet. Also, periodical-
ly check the level and zero of the mano-
meter.
If the pressure drop across the filter be-
comes too high, making isokinetic sampling
difficult to maintain, the filter may be re-
placed in the midst of a sample run. It is
recommended that another complete filter
assembly be used rather than attempting to
change the filter itself. Before a new filter
assembly is installed, conduct a leak-check
(see Section 4.1.4.2). The total particulate
weight shall Include the summation of all
filter assembly catches.
A single train shall be used for the entire
sample run, except in cases where simulta-
neous sampling is required in two or more
separate ducts or at two or more different
locations within the same duct, or, in cases
where equipment failure necessitates a
change of trains. In all other situations, the
use of two or more trains will be subject to
the approval of the Administrator.
Note that when two or more trains are
used, separate analyses of the front-half
and (If applicable) impinger catches from
each train shall be performed, unless identi-
cal nozzle sizes were used on all trains, in
which case, the front-half catches from the
individual trains may be combined (as may
the impinger catches) and one analysis of
front-half catch and one analysis of im-
pinger catch may be performed. Consult
with the Administrator for details concern-
ing the calculation of results when two or
more trains are used.
At the end of the sample run, turn off the
coarse adjust valve, remove the probe and
nozzle from the stack, turn off the pump,
record the final dry gas meter reading, arid
conduct a post-test leak-check, as outlined
In Section 4.1.4.3. Also, leak-check the pitot
lines as described in Method 2, Section 3.1;
the lines must pass this leak-check, in order
to validate the velocity head data.
4.1.6 Calculation of Percent Isokinetic.
Calculate percent isokinetic (see Calcula-
tions, Section 6) to determine whether the
run was valid or another test run should be
made. If there was difficulty in maintaining
isokinetic rates due to source conditions,
consult with the Administrator for possible
variance on the isokinetic rates.
4.2 Sample Recovery. Proper cleanup
procedure begins as soon as the probe is re-
moved from the stack at the end of the sam-
pling period. Allow the probe to cool.
When the probe can be safely handled,
wipe off all external particulate matter near
the tip of the probe nozzle and place a cap
over it to prevent losing or gaining particu-
late matter. Do not cap off the probe tip
tightly while the sampling train is cooling
down as this would create a vacuum in the
filter holder, thus drawing water from the
impingers into the filter holder.
Before moving the sample train to the
cleanup site, remove the probe from the
sample train, wipe off the silicone grease,
and cap the open outlet of the probe. Be
careful not to lose any condensate that
might be present. Wipe off the silicone
grease from the filter inlet where the probe
was fastened and cap it. Remove the umbili-
cal cord from the last impinger and cap the
impinger. If a flexible line is used between
the first impinger or condenser and the
filter holder, disconnect the line at the
filter holder and let any condensed water or
liquid drain into the impingers or condens-
er. After wiping off the silicone grease, cap
off the filter holder outlet and impinger
inlet. Either ground-glass stoppers, plastic
caps, or serum caps may be used to close
these openings.
Transfer the probe and filter-impinger as-
sembly to the cleanup area. This area
should be clean and protected from the
wind so that the chances of contaminating
or losing the sample will be minimized.
Save a portion of the acetone used for
cleanup as a blank. Take 200 ml of this ac-
etone directly from the wash bottle being
used and place it in a glass sample container
labeled "acetone blank."
Inspect the train > prior to and during dis-
assembly and note any abnormal conditions.
Treat the samples as follows:
Container No. 1. Carefully remove the
filter from the filter holder and place it in
Its identified petri dish container. Use a pair
of tweezers and/or clean disposable surgical
gloves to handle the filter. If it is necessary
to fold the filter, do so such that the partic-
ulate cake is inside the fold. Carefully trans-
fer to the petri dish any particulate matter
and/or filter fibers which adhere to the
filter holder gasket, by using a dry Nylon
B-52
-------
App. A
bristle brush and/or a sharp-edged blade.
Seal the container.
Container No. 2. Taking care to see that
dust on the outside of the probe or other ex-
terior surfaces does not get into the sample,
quantitatively recover particulate matter or
any condensate from the probe nozzle,
probe fitting, probe liner, and front half of
the filter holder by washing these compo-
nents with acetone and placing the wash in
a glass container. Distilled water may be
used instead of acetone when approved by
the Administrator and shall be used when
specified by the Administrator; in these
cases, save a water blank and follow the Ad-
ministrator's directions on analysis. Perform
the acetone rinses as follows: ,
Carefully remove the probe nozzle and
clean the inside surface by rinsing with ac-
etone from a wash bottle and brushing with
a Nylon bristle brush. Brush until the ac-
etone rinse shows no visible particles, after
which make a final rinse of the inside sur-
face with acetone.
Brush and rinse the inside parts of the
Swagelok fitting with acetone in a similar
way until no visible particles remain.
Rinse the probe liner with acetone by tilt-
Ing and rotating the probe while squirting
acetone into its upper end so that all inside
surfaces will be wetted with acetone. Let the
acetone drain from the lower end into the
sample container. A funnel (glass or poly-
ethylene) may be used to aid on transfer-
ring liquid washes to the container. Follow
the acetone rinse with a probe brush. Hold
the probe in an inclined position, squirt ac-
etone into the upper end as the probe brush
is being pushed with a twisting action
through the probe; hold a sample container
underneath the lower end of the probe, and
catch any acetone and particulate matter
which is brushed from the probe. Bun the
brush through the probe three times or
more until no visible particulate matter is
carried out with the acetone or until none
remains in the probe liner on visual inspec-
tion. With stainless steel or other metal
probes, run the brush through in the above
prescribed manner at least six times since
metal probes have small crevices in which
particulate matter can be entrapped. Rinse
the brush with acetone, and quantitatively
collect these washings in the sample con-
tainer. After the brushing, make a final ac-
etone rinse of the probe as described above.
It is recommended that two people be
used to clean the probe to minimize sample
losses. Between sampling runs, keep brushes
clean and protected from contaminations.
Title 40—Protection of Environment
After ensuring that all joints have been
wiped clean of silicone grease, clean the
inside of the front half of the filter holder
by rubbing the surfaces with a Nylon bristle
brush and rinsing with acetone. Rinse each
surface three times or more if needed to
remove visible partieulate. Make a final
rinse of the brush and filter holder. Careful-
ly rinse out the glass cyclone, also (if appli-
cable). After all acetone washings and par-
ticulate matter have been collected in the
sample container, tighten the lid on the
sample container so that acetone will not
leak out when it is shipped to the labora-
tory. Mark the height of the fluid levcel to
determine whether or not leakage occured
during transport. Label the container to
clearly identify its contents.
Container No. 3. Note the color of the in-
dicating silica gel to determine if it has been
completely spent and make a notation of its
condition. Transfer the silica gel from the
fourth impinger to its original container
and seal. A funnel may make it easier to
pour the silica gel without spilling. A rubber
policeman may be used as an aid in remov-
ing the silica gel from the impinger. it is not
necessary to remove the small amount of
dust particles that may adhere to the im-
pinger wall and are difficult to remove.
Since the gain in weight is to be used for
moisture calculations, do not use any water
or other liquids to transfer the silica gel. If
a balance is available in the field, follow the
procedure for container No. 3 in Section 4.3.
Impinger Water. Treat the impingers as
follows; Make a notation of any color or
film in the liquid catch. Measure the liquid
which is in the first three impingers to
within ±1 ml by,using a graduated cylinder
or by weighing it to within ±0.5 g by using a
balance (if one is available). Record the
volume or weight of liquid present. This in-
formation is required to calculate the mois-
ture content of the effluent gas.
Discard the liquid after measuring and re-
cording the volume or weight, unless analy-
sis of the impinger catch is required (see
Note, Section 2.1.7).
If a different type of condenser is used,
measure the amount of moisture condensed
either volumetrically or gravimetrically.
Whenever possible, containers should be
shipped in such a way that they remain up-
right at all times.
4.3 Analysis. Record the data required on
a sheet such as the one shown in Figure 5-3.
Handle each sample container as follows:
B-53
-------
Chapter I—Environmental Protection Agency
Plant
App. A
Date.
Run No..
Filter No..
Amount liquid lost during transport
Acetone blank volume, ml
Acetone wash volume, ml
Acetone blank concentration, mg/mg (equation 54).
Acetone wash blanx, mg (equation 5-5)
CONTAINER
NUMBER
1
2
TOTAL
WEIGHT OF PARTICULATE COLLECTED,
mg
FINAL WEIGHT
^>-NCREASE' 9 - VOLUME WAT€R. ml
1 g/ml
B-54
-------
App. A
Title 40—Protection of Environment
Container No. 1. Leave the contents in the
shipping container or transfer the filter and
any loose particulate from the sample con-
tainer to a tared glass weighing dish. Desk*_
cate for 24 hours in a desiccator containing
anhydrous calcium sulfate. Weigh to a con-
stant weight and report the results to the
nearest 0.1 mg. For purposes of this Section,
4.3, the term "constant weight" means a dif-
ference of no more than 0.5 mg or 1 percent
of total weight less tare weight, whichever is
greater, between two consecutive weighings,
with no less than 6 hours of desiccation
time between weighings.
Alternatively, the sample may be oven
dried at 105° C (220° P) for 2 to 3 hours,
cooled in the desiccator, and weighed to a
constant weight, unless otherwise specified
by the Administrator. The tester may also
opt to oven dry the sample at 105° C (220°
P) for 2 to 3 hours, weigh the sample, and
use this weight as a final weight.
Container No. 2. Note the level of liquid in
the container and confirm on the analysis
sheet whether or not leakage occurred
during transport. If a noticeable amount of
leakage has occurred, either void the sample
or use methods, subject to the approval of
the Administrator, to correct the final re-
sults. Measure the liquid in this container
either volumetrically to ±1 ml or gravime-
trically to ±0.5 g. Transfer the contents to a
tared 250-ml beaker and evaporate to dry-
ness at ambient temperature and pressure.
Desiccate for 24 hours and weigh to a con-
stant weight. Report the results to the near-
est 0.1 mg.
Container No. 3. Weigh the spent silica
gel (or silica gel plus impinger) to the near-
est 0.5 g using a balance. This step may be
conducted in the field.
"Acetone Blank" Container. Measure ac-
etone in this container either volumentrical-
ly or gravimetrically. Transfer the acetone
to a tared 250-ml beaker and evaporate to
dryness at ambient temperature and pres-
sure. Desiccate for 24 hours and weigh to a
constant weight. Report the results to the
nearest 0.1 mg.
NOTE: At the option of the tester, the con-
tents of Container No. 2 as well as the ac-
etone blank container may be evaporated at
temperatures higher than ambient. If evap-
oration is done at an elevated temperature,
the temperature must be below the boiling
point of the solvent; also, to prevent "bump-
ing," the evaporation process must be close-
ly supervised, and the contents of the
beaker must be swirled occasionally to
maintain an even temperature. Use extreme
care, as acetone is highly flammable and
has a low flash point.
5. Calibration
Maintain a laboratory log of all calibra-
tions.
5.1 Probe Nozzle. Probe nozzles shall be
calibrated before their initial use in the
field. Using a micrometer, measure the
inside diameter of the nozzle to the nearest
01.025 mm (0.001 in.). Make three separate
measurements using different diameters
each time, and obtain the average of the
measurements. The difference between the
high and low numbers shall not exceed 0.1
mm (6.004 in.). When nozzles become
nicked, dented, or corroded, they shall be
reshaped, sharpened, and recalibrated
before use. Each nozzle shall be permanent-
ly and uniquely identified.
5.2 Pitot Tube. The Type S pitot tube as-
sembly shall be calibrated according to the
procedure outlined in Section 4 of Method
2.
5.3 Metering System. Before its initial
use in the field, the metering system shall
be calibrated according to the procedure
outlined in APTD-0576. Instead of physical-
ly adjusting the dry gas meter dial readings
to correspond to the wet test meter read-
ings, calibration factors may be used to
mathematically correct the gas meter dial
readings to the proper values. Before cali-
brating the metering system, it is suggested
that a leak-check be conducted. For meter-
ing systems having diaphragm pumps, the
normal leak-check procedure will not detect
leakages within the pump, for these cases
the following leak-check procedure is sug-
gested: make a 10-minute calibration run at
0.0057 m Vmin (0.02 cfm); at the end of the
run, take the difference of the measured
wet test meter and dry gas meter volumes;
divide the difference by 10, to get the leak,
rate. The leak rate should not exceed
0.00057 m Vmin (0.02 cfm).
After each field use, the calibration of the
metering system shall be checked by per-
forming three calibration runs at a single,
intermediate orifice setting (based on the
previous field test). With the vacuum set at
the maximum value reached during the test
series. To adjust the vacuum, insert a valve
between the wet test meter and the inlet of
the metering system. Calculate the average
value of the calibration factor. If the cali-
bration has changed by more than 5 per-
cent, recalibrate the meter over the full
range of orifice settings, as outlined in
APTD-0576.
Alternative procedures, e.g., using the ori-
fice meter coefficients, maybe used, subject
to the approval of the Administrator.
NOTE: If the dry gas meter coefficient
values obtained before and after a test
series differ by more than 5 percent, the
test series shall either be voided, or calcula-
tions for test series shall be performed using
whichever meter coefficient value (i.e.,
before or after) gives the lower value of
total sample volume.
B-55
-------
Chapter I—Environmental Protection Agency
App.A
6.4 Probe Heater Calibration. The probe
heating system shall be calibrated before its
initial use in the field according to the pro-
cedure outlined in APTD-0576. Probes con-
structed according to APTD-0581 need not
be calibrated if the calibration curves in
APTD-0576 are used.
5.5 Temperature Gauges. Use the proce-
dure in Section 4.3 of Method 2 to calibrate
In-stack temperature gauges. Dial thermom-
eters, such as are used for the dry gas meter
«nd condenser outlet, shall be calibrated
against mercury-In-glass thermometers.
5.6 Leak Check of Metering System
Shown in Figure 5-1. That portion of the
sampling train from the pump to the orifice
meter should be leak checked prior to initial
use and after each shipment. Leakage after
the pump will result in less volume being re-
corded than is actually sampled. The follow-
ing procedure is suggested (see Figure 5-4):
Close the main valve on the meter box.
Insert a one-hole rubber stopper with
rubber tubing atached into the orifice ex-
haust pipe. Disconnect and vent the low side
of the orifice manometer. Close off the low
side orifice tap. Pressurize the system to 13
to 18 cm (5 to 7 in.) water column by blow-
ing into the rubber tubing. Pinch'off the
tubing and observe the manometer for one
minute. A loss of pressure on the mano-
meter indicates a leak in the meter box;
leaks, if present, must be corrected.
5.7 Barometer. Calibrate against a mer-
cury barometer.
6. Calculations
Carry out calculations, retaining at least
one extra decimal figure beyond that of the
acquired data. Round off figures after the
final calculation. Other forms of the equa-
tions may be used as long as they give equiv-
alent results.
B.-56
-------
tn
VACUUM
GAUGE
RUBBER
TUBING
BLOW INTO TUBING
UNTIL MANOMETER
READS 5 TO 7 INCHES
WATER COLUMN ORIFICE
MANOMETER
AIR-TIGHT
PUMP
I
a
o
figure 5-4. Leak check of meter box.
-------
Chapter I—Environmental Protection Agency
APP.A
6.1 Nomenclature
X.=Cross-sectional area of nozzle, m2 (ft2).
BM=Water vapor in the gas stream, propor-
tion by volume.
G=Acetone blank residue concentration,
mg/g.
c,=Concent rat Ion .of participate matter in
stack gas, dry basis, corrected to stand-
ard conditions, g/dscm (g/dscf).
7=Percent of isokinetic sampling.
Zo=Maximum acceptable leakage rate for
either a pretest leak check or for a leak
check following a component change;
equal to 0.0057 m'/min (0.02 cfm) or 4
percent of the average sampling rate,
whichever is less.
l4=Individual leakage rate observed during
the leak check conducted prior to the
"i"1" component change (4=1, 2, Z....n),
m'/mln (cfm).
Ip=Leakage rate observed during the post-
test leak check, mVmin (cfm).
m.=Total amount of particulate matter col-
lected, mg.
if»=Molecular weight of water, 18.0 g/g-
mole (18.01b/lb-mole).
ma=Mass of residue of acteone after evapo-
ration, mg.
P^=Barometric pressure at the sampling
site, mm Hg (in. Hg).
P,=Absolute stack gas pressure, mm Hg (in.
Hg).
P,u=Standard absolute pressure, 760 mm
Hg (29.92 in. Hg).
H=Ideal gas constant, 0.06236 mm Hg-m3/
•K-g-mole (21.85 in. Hg-ftVR-lb-mole).
Tm=Absolute average dry gas meter tem-
perature (see Figure 5-2), °K (°B).
T,—Absolute average stack gas temperature
(see Figure 5-2), °K (°R).
71IU,=Standard absolute temperature, 293° K
(528' R).
V,=Volume of acetone blank, ml.
V«=Volume of acetone used in wash, ml.
Vif=Total volume of liquid collected in im-
pingers and silica gel (see Figure 5-3),
ml.
V»,= Volume of gas sample as measured by
dry gas meter, dcm (dscf).
Vn (,,,<)=Volume of gas sample measured by
the dry gas meter, corrected to standard
conditions, dscm (dscf).
lk<«
-------
App. A
and substitute only for those leakage rates
(Li or Lp) which exceed La.
6.4 Volume of water vapor.
Equation 5-2
Title 40—Protection of Environment
6.7 Acetone Wash Blank.
»F. = C.V..p.
Equation 5-5
where:
#,=0.001333 m'/ml for metric units
=0.04707 ftVml for English units.
6.5 Moisture Content.
Ba.,-
Equation 5-3
NOTE: In saturated or water droplet-laden
gas streams, two calculations of the mois-
ture content of the stack gas shall be' made,
one from the impinger analysis (Equation 5-
3), and a second from the assumption of
saturated conditions. The lower of the two
values of Ba shall be considered correct. The
procedure for determining the moisture
content based upon assumption of saturated
conditions is given in the Note of Section 1.2
of Method 4. For the purposes of this
method, the average stack gas temperature
from Figure 5-2 may be used to make this
determination, provided that the accuracy
of the in-stack temperature sensor is ±1° C
(2° F).
6.6 Acetone Blank Concentration.
6.8 Total Particulate Weight. Determine
the total partieulate catch from the sum of
the weights obtained from containers 1 and
2 less the acetone blank (see Figure 5-3).
NOTE: Refer to Section 4.1.5 to assist in
calculation of results involving two or more
filter assemblies or two or more sampling
trains.
6.9 Particulate Concentration.
c.= (0.001 ff/mg)
6.10 Conversion Factors:
Equation 5-6
From
scf
g/ft'
g/ft3
g/fts.
To
m*
gr/fl5.
lb/fts
Multiply by
0 02832
15.43.
2 205X10"3
35 31
Equation 5-4
6.11 Isokinetic Varition.
6.11.1 Calculation From Raw Data.
100 T,[K3Vle
606v.P,An
) ( Pbtr+ Aff/13.6)]
Equation 5-7
where:
#,=0.003454 mm Hg-m3/ml-°K for metric
units.
=0.002669-in. Hg-ft'/ml-'R for English
units.
6.11.2 Calculation From Intermediate
Values.
--K4
T V
* »vi
m («td)
Equation 5-8
where:
,ff.=4.320 for metric units
=0.09450 for English units.
6.12 Acceptable Results. If 90 percent /
<110< percent, the results are acceptable.
If the results are low in comparison to the
standard and / is beyond the acceptable
range, or, if / is less than 90 percent, the Ad-
ministrator may opt to accept the results.
B-59
-------
Use Citation 4 to make judgments. Other-
wise, reject the results and repeat the test.
7. Bibliography
1. Addendum to Specifications for Inciner-
ator Testing at Federal Facilities. PHS,
NCAPC. Dec. 6, 1967.
2. Martin, Robert M. Construction Details
of Isokinetic Source-Sampling Equipment.
Environmental Protection Agency. Re-
search Triangle Park, N.C. APTD-0581.
April 1971.
3. Rom, Jerome J. Maintenance, Calibra-
tion, and Operation of Isokinetic Source
Sampling Equipment. Environmental Pro-
tection Agency. Research Triangle Park,
N.C. APTD-0576. March, 1972.
4. Smith, W. S., R. T. Shigehara, and W.
P. Todd. A method of Interpreting Stack
Sampling Data. Paper Presented at the 63d
Annual Meeting of the Air Pollution Con-
trol Association, St. Louis, Mo. June 14-19,
1970.
5. Smith, W. S., et al. Stack Gas Sampling
Improved and Simplified With New Equip-
ment. APCA Paper No. 67-119.1967.
6. Specifications for Incinerator Testing at
Federal Facilities. PHS, NCAPC. 1967.
7. Shigehara, R. T. Adjustments in the
EPA Nomograph for Different Pitot Tube
Coefficients and Dry Molecular Weights.
Stack Sampling News 2:4-11, October, 1974.
8. Vollaro, R. F. A Survey of Commercially
Available Instrumentation For the Measure-
ment of Low-Range Gas Velocities. U.S. En-
vironmental Protection Agency, Emission
Measurement Branch. Research Triangle
Park, N.C. November, 1976 (unpublished
paper).
9. Annual Book of ASTM Standards. Part
26. Gaseous Fuels; Coal and Coke; Atmos-
pheric Analysis. American Society for Test-
ing and Materials. Philadelphia, Pa. 1974.
pp. 617-622.
10. Felix, L. G., G. I. Clinard, G. E. Lacey,
and J. D. McCain. Inertial Cascade Impac-
tor Substrate Media for Flue Gas Sampling.
U.S. Environmental Protection Agency. Re-
search Triangle Park, N.C. 27711, Publica-
tion No. EPA-600/7-77-060. June 1977. 83 p.
B-60
-------
APR. A
METHOD 17—DETERMINATION OF PARTICULATE
EMISSIONS PROM STATIONARY SOURCES (!N-
STACK FILTRATION METHOD)
Introduction
Particulate matter is not an absolute
quantity; rather, it is a function of tempera-
ture and pressure. Therefore, to prevent
variability in particulate matter emission
regulations and/or associated test methods,
the temperature and pressure at which par-
ticulate matter is to be measured must be
carefully defined. Of the two variables (i.e.,
temperature and pressure), temperature has
the greater effect upon the amount of par-
ticulate matter in an effluent gas stream; in
most stationary source categories, the effect
of pressure appears to be negligible.
In method 5, 250° F is established as a
nominal reference temperature. Thus,
where Method 5 is specified in an applicable
subpart of the standards, particulate matter
is defined with respect to temperature. In
order to maintain a collection temperature
of 250° F, Method 5 employs a heated glass
sample probe and a heated filter holder.
This equipment is somewhat cumbersome
and requires care in its operation. There-
fore, where particulate matter concentra-
tions (over the normal range of temperature
associated with a specified source category)
are known to be independent of tempera-
ture, it is desirable to eliminate the glass
probe and heating systems, and sample at
stack temperature.
This method describes an in-stack sam-
pling system and sampling procedures for
Title 40—Protection of Environment
use in such cases. It is intended to be used
only when specified by an applicable sub-
part of the standards, and only within the
applicable temperature limits (if specified),
or when otherwise approved by the Admin-
istrator.
1. Principle and Applicability.
1.1 Principle. Particulate matter is with-
drawn isokinetically from the source and
collected on a glass fiber filter maintained
at stack temperature. The particulate mass
is determined gravimetrically after removal
of uncombined water.
1.2 Applicability. This method applies to
the determination of particulate emissions
from stationary sources for determining
compliance with new source performance
standards, only when specifically provided
for in an applicable subpart of the stand-
ards. This method is not applicable to stacks
that contain liquid droplets or are saturated
with water vapor. In addition, this method
shall not be used as written if the projected
cross-sectional area of the probe extension-
filter holder assembly covers more than 5
percent of the stack cross-sectional area (see
Section 4.1.2).
'2. Apparatus.
2.1 Sampling Train. A schematic of the
sampling train used in this method is shown
in Figure 17-1. Construction details for
many, but not all, of the train components
are given in APTD-0581 (Citation 2 in Sec-
tion 7); for changes from the APTD-0581'
document and for allowable modifications
to Figure 17-1, consult with the Administra-
tor.
B-61
-------
TEMPtRATVtE MOTUK
IEKOR FILTER HOLDER
«.y>1Jc»(t;iW
HWlPISERTRAtllOrriOIIAI., MAY in REPLACED
IVAN EQUIVALENT COKDEIiSER
THERMOMETER
O3
I
ORIFICE MANOMETER
* SUGGESTED (INTERFERENCE-FREE) SMCINGS
DRV GAS METER
•§
r
I
i
•a
Figure 17-1. Paniculate-Sampling Train, Equipped with In-Stack Filter.
-------
App. A
The operating and maintenance proce-
dures for many of the sampling train com-
ponents are described in APTD-0576 ^Cita-
tion 3 in Section 7). Since correct usage is
important in obtaining valid results, all
users should read the APTD-0576 document
and adopt the operating and maintenance
procedures outlined in it, unless otherwise
specified herein. The sampling train con-
sists of the following components:
2.1.1 Probe Nozzle. Stainless steel (316)
or glass, with sharp, tapered leading edge.
The angle of taper shall be 30° and the
taper shall be on the outside to preserve a
constant internal diameter. The probe
nozzle shall be of the button-hook or elbow
design, unless otherwise specified by the Ad-
ministrator. If made of stainless steel, the
nozzle shall be constructed from seamless
tubing. Other materials of construction may
be used subject to the approval of the Ad-
ministrator.
A range of sizes suitable for isokinetic
sampling should be available, e.g., 0.32 to
1.27 cm (.VB to % In)—or larger if higher
volume sampling trains are used—inside di-
ameter (ID) nozzles in increments of 0.16 cm
(Vie in). Each nozzle shall be calibrated ac-
cording to the procedures outlined in Sec-
tion 5.1.
2.1.2 Filter Holder. The in-stack filter
holder shall be constructed of borosilicate
or quartz glass, or stainless steel; if a gasket
is used, it shall be made of silicone rubber,
TeHon, or stainless steel. Other holder and
gasket materials may be used subject to the
approval of the Administrator. The filter
holder shall be designed to provide a posi-
tive seal against leakage from the outside or
around the filter.
2.1.3 Probe Extension. Any suitable rigid
probe extension may be used after the filter
holder.
2.1.4 Pitot Tube. Type S, as described in
Section 2.1 of Method 2, or other device ap-
proved by the Administrator; the pitot tube
shall be attached to the probe extension to
allow constant monitoring of the stack gas
velocity (see Figure 17-1). The impact (high
pressure) opening plane of the pitot tube
shall be even with or above the nozzle entry
plane during sampling (see Method 2,
Figure 2-6b). It is recommended: (1) that
the pitot tube have a known baseline coeffi-
cient, determined as outlined in Section 4 of
Method 2; and (2) that this known coeffi-
cient be preserved by placing the pitot tube
in an interference-free arrangement with re-
spect to the sampling nozzle, filter holder,
and temperature sensor (see Figure 17-1).
Note that the 1.9 cm (0.75 in) free-space be-
tween the nozzle and pitot tube shown in
Figure 17-1, is based on a 1.3 cm (0.5 in) ID
nozzle. If the sampling train is designed for
sampling at higher flow rates than that de-
scribed in APTD-0581, thus necessitating
the use of larger sized nozzles, the free-
Title 40—Protection of Environment
space shall be 1.9 cm (0.75 in) with the larg-
est sized nozzle in place.
Source-sampling assemblies that do not
meet the minimum spacing requirements of
Figure 17-1 (or the equivalent of these re-
quirements, e.g.. Figure 2-7 of Method 2)
may be used; however, the pitot tube coeffi-
cients of such assemblies shall be deter-
mined by calibration, using methods subject
to the approval of the Administrator.
2.1.5 Differential Pressure Gauge. In-
clined manometer or equivalent device
(two), as described in Section 2.2 of Method
2. One manometer shall be used for velocity
head (Ap) readings, and the other, for ori-
fice differential pressure readings.
2.1.6 Condenser. It is recommended that
the impinger system described in Method 5
be used to determine the moisture content
of the stack gas. Alternatively, any system
that allows measurement of both the water
condensed and the moisture leaving the con-
denser, each to within 1 ml or 1 g, may be
used. The moisture leaving the condenser
can be measured either by: (1) monitoring
the temperature and pressure at the exit of
the condenser and using Dalton's law of
partial pressures; or (2) passing the sample
gas stream through a silica gel trap with
exit gases kept below 20° C (68° F) and de-
termining the weight gain.
Flexible tubing may be used between the
probe extension and condenser. If means
other than silica gel are used to determine
the amount of moisture leaving the con-
denser, it is recommended that silica gel still
be used between the condenser system and
pump to prevent moisture condensation in
the pump and metering devices and to avoid
the need to make corrections for moisture
in the metered volume.
2.1.7 Metering System. Vacuum gauge,
leak-free pump, thermometers capable of
measuring temperature to within 3° C (5.4°
F), dry gas meter capable of measuring
volume to within 2 percent, and related
equipment, as shown in Figure 17-1. Other
metering systems capable of maintaining
sampling rates within 10 percent of isokine-
tic and of determining sample volumes to
within 2 percent may be used, subject to the
approval of'the Administrator. When the
metering system is used in conjunction with
a pitot tube, the system shall enable checks
of isokinetic rates.
Sampling trains utilizing metering sys-
tems designed for higher flow rates than
that described in APTD-0581 or APTD-0576
may be used provided that the specifica-
tions of this method are met.
2.1.8 Barometer. Mercury, aneroid, or
other barometer capable of measuring at-
mospheric pressure to within 2.5 mm Hg
(0.1 in. Hg). In many cases, the barometric
reading may be obtained from a nearby na-
tional weather service station, in which case
B-63
-------
Chapter I—Environmental Protection Agency
App. A
Che station value (which is the absolute
barometric pressure) shall be requested and
in adjustment for elevation differences be-
tween the weather station and sampling
point shall be applied at a rate of minus 2.5
mm Hg (0.1 in. Hg) per 30 m (100 ft) eleva-
tion increase or vice versa for elevation de-
crease.
2.1.9 Oas Density Determination Equip-
ment. Temperature sensor and pressure
gauge, as described in Sections 2.3 and 2.4 of
Method 2, and gas analyzer, if necessary, as
described in Method 3.
The temperature sensor shall be attached
to either the pitot tube or to the probe ex-
tension, In a fixed configuration. If the tem-
perature sensor is attached in the field; the
sensor shall be placed in an interference-
free, arrangement with respect to the Type
8 pitot tube openings (as shown in Figure
17-1 or in Figure 2-7 of Method 2). Alterna-
tively, the temperature sensor need not be
attached to either the probe extension or
pitot tube during sampling, provided that a
difference of not more than 1 percent in the
average velocity measurement is introduced.
This alternative is subject to the approval
of the Administrator.
2.2 Sample Recovery.
2.2.1 Probe Nozzle Brush. Nylon bristle
brush with stainless steel wire handle. The
brush shall be properly sized and shaped to
brush out the probe nozzle.
2.2.2 Wash Bottles—Two. Glass wash
bottles are recommended; polyethylene
wash bottles may be used at the option of
the tester. It is recommended that acetone
not be stored in polyethylene bottles for
longer than a month.
2.2.3 Glass Sample Storage Containers.
Chemically resistant, borosilicate glass bot-
tles, for acetone washes, 500 ml or 1000 ml.
Screw cap liners shall either be rubber-
backed Teflon or shall be constructed so as
to be leak-free and resistant to chemical
attack by acetone. (Narrow mouth glass bot-
tles have been found to be less prone to
leakage.) Alternatively, polyethylene bottles
may be used.
2.2.4 Petri Dishes. For filter samples;
glass or polyethylene, unless otherwise spec-
ified by the Administrator.
2.2.5 Graduated Cylinder and/or Bal-
ance. To measure condensed water to within
1 ml or 1 g. Graduated cylinders shall have
subdivisions no greater than 2 ml. Most lab-
oratory balances are capable of weighing to
the nearest 0.5 g or less. Any of these bal-
ances is suitable for use here and in Section
2.3.4.
2.2.6 Plastic Storage Containers. Air
tight containers to store silica gel.
2.2.7 Funnel and Rubber Policeman. To
aid In transfer of silica gel to container; not
necessary if silica gel is weighed in the field.
2.2.8 Funnel. Glass, or polyethylene, to
aid In sample recovery.
2.3 Analysis.
2.3.1 Glass Weighing Dishes.
2.3.2 Desiccator.
2.3.3 Analytical Balance. To measure to
within 0.1 mg.
2.3.4 Balance. To measure to within 0.5
mg.
2.3.5 Beakers. 250 ml.
2.3.6 Hygrometer. To measure the rela-
tive humidity of the laboratory environ-
ment.
2.3.7 Temperature Gauge. To measure
the temperature of the laboratory environ-
ment.
3. Reagents.
3.1 Sampling.
3.1.1 Filters. The in-stack filters shall be
glass mats or thimble fiber filters, without
organic binders, and shall exhibit at least
99.95 percent efficiency ( 0.05 percent pene-
tration) on 0.3 micron dioctyl phthalate
smoke particles. The filter efficiency tests
shall be conducted in accordance with
ASTM standard method D 2986-71. Test
data from the supplier's quality control pro-
gram are sufficient for this purpose.
3.1.2 Silica Gel. Indicating type, 6- to 16-
mesh. If previously used, dry at 175° C (350°
F) for 2 hours. New silica gel may be used as
received. Alternatively, other types of desic-
cants (equivalent or better) may be used,
subject to the approval of the Administra-
tor.
3.1.3 Crushed Ice.
3.1.4 Stopcock Grease. Acetone-insoluble,
heat-stable silicone grease. This is not nec-
essary if screw-on connectors with Teflon
sleeves, or similar, are used. Alternatively,
other types of stopcock grease may be used,
subject to the approval of the Administra-
tor.
3.2 Sample Recovery. Acetone, reagent
grade, 0.001 percent residue, in glass bottles.
Acetone from metal containers generally
has a high residue blank and should not be
used. Sometimes, suppliers transfer acetone
to glass bottles from metal containers.
Thus, acetone blanks shall be run prior to
field use and only acetone with low blank
values ( 0.001 percent) shall be used. In no
case shall a blank value of greater than
0.001 percent of the weight of acetone used
be subtracted from the sample weight.
3.3 Analysis.
3.3.1 Acetone. Same as 3.2.
3.3.2 Desiccant. Anhydrous calcium sul-
fate, indicating type. Alternatively, other
types of desiccants may be used, subject to
the approval of the Administrator.
4. Procedure.
4.1 Sampling. The complexity of this
method is such that, in order to obtain reli-
able results, testers should be trained, and
experienced with the test procedures.
4.1.1 Pretest Preparation. All compo-
nents shall be maintained and calibrated ac-
B-64
-------
App. A
cording to the procedure described in
APTD-0576, unless otherwise specified
herein.
Weigh several 200 to 300 .g portions of
silica gel in air-tight containers to the near-
est 0.5 g. Record the total weight of the
silica gel plus container, on each container.
As an alternative, the silica gel need not be
preweighed, but may be weighed directly in
its impinger or sampling holder just prior to
train assembly.
Check filters visually against light for ir-
regularities and flaws or pinhole leaks.
Label filters of the proper size on the back
side near the edge using numbering ma-
chine ink. As an alternative, label the ship-
ping containers (glass or plastic petri dishes)
and keep the filters in these containers at
all times except during sampling and weigh-
ing.
Desiccate the filters at 20±5.6° C (68±10°
F) and ambient pressure for at least 24
hours and weigh at intervals of at least 6
hours to a constant weight, i.e., 0.5 mg
change from previous weighing; record re-
sults to the nearest 0.1 mg. During each
weighing the filter must not be exposed to
the laboratory atmosphere - for a period
greater than 2 minutes and a relative hu-
midity above 50 percent. Alternatively
(unless otherwise specified by the Adminis-
trator), the filters may be oven dried at 105°
C (220° P) for 2 to 3 hours, desiccated for 2
hours, and weighed. Procedures other than
those described, which account for relative
humidity effects, may be used, subject to
the approval of the Administrator.
Title 40—Protection of Environment
4.1.2 Preliminary Determinations. Select
the sampling site and the minimum number
of sampling points according to Method 1 or
as specified by the Administrator. Make a
projeeted-area model of the probe exten-
sion-filter holder assembly, with the pitot
tube face openings positioned along the cen-
terline of the stack, as shown in Figure 17-2.
Calculate the estimated cross-section block-
age, as shown in Figure 17-2. If the blockage
exceeds 5 percent of the duct cross sectional
area, the tester has the following options:
(Da suitable out-of-stack filtration method
may be used instead of in-stack filtration; or
(2) a special in-stack arrangement, in which
the sampling and velocity measurement
sites are separate, may be used; for details
concerning this approach, consult with the
Administrator (see also Citation 10 in Sec-
tion 7). Determine the stack pressure, tem-
perature, and the range of velocity heads
using Method 2; it is recommended that a
leak-check of the pitot lines (see Method 2,
Section 3.1) be performed. Determine the
moisture content -using Approximation
Method 4 or its alternatives for the purpose
of making isokinetic sampling rate settings.
Determine the stack gas dry molecular
weight, as described in Method 2, Section
3.6; if integrated Method 3 sampling is used
for molecular weight determination, the in-
tegrated bag sample shall be taken simulta-
neously with, and for the same total length
of time as, the particular sample run.
B-65
-------
Chapter I—Environmental Protection Agency
App. A
STACK
WALL
IN-STACK FILTER-
PROBE EXTENSION
ASSEMBLY
ESTIMATED
BLOCKAGE
.(*)
PSHADED AREA]
DUCT AREAJ
X 100
Figure 17-2. Projected-area model of cross-section blockage (approximate average for
a sample traverse) caused by an in-stack filter holder-probe extension assembly.
B-66
-------
APP. A
Title 40—Protection of Environment
Select a nozzle size based on the range of
velocity heads, such that it is not necessary
to change the nozzle size in order to main-
tain isokinetic sampling rates. During the
run, do not change the nozzle size. Ensure
that the proper differential pressure gauge
is chosen for the range of velocity heads en-
countered (see Section 2.2 of Method 2).
Select a probe extension length such that
all traverse points can be sampled. For large
stacks, consider sampling from opposite
sides of the stack to reduce the length of
probes.
Select a total sampling time greater than
or equal to the minimum total sampling
time specified in the test procedures for the
specific industry such that (1) the sampling
time per point is not less than 2 minutes (or
some greater time interval if specified by
.the Administrator), and (2) the sample
volume taken (corrected to standard condi-
tions) will exceed the required minimum
total gas sample volume. The latter is based
on an approximate average sampling rate.
It is recommended that the number of
minutes sampled at each point be an integer
or an integer plus one-half minute, in order
to avoid timekeeping'errors.
In some circumstances, e.g., batch cycles,
it may be necessary to sample for shorter
times at the traverse points and to obtain
smaller gas sample volumes. In these cases,
the Administrator's approval must first be
obtained.
4.1.3 Preparation of Collection Train.
During preparation and assembly of the
sampling train, keep all openings where con-
tamination can occur covered until just
prior to assembly or until sampling is about
to begin.
If impingers are used to condense stack
•gas moisture, prepare them as follows: place
100 ml of water In each of the first two im-
pingers, leave the third impinger empty,
and transfer approximately 200 to 300 g of
preweighed silica gel from its container to
the fourth impinger. More silica gel may be
used, but care should be taken to ensure
that it is not entrained and carried out from
the impinger during sampling. Place the
container in a clean place for later use in
the sample recovery. Alternatively, the
weight of the silica gel plus impinger may
be determined to the nearest 0.5 g and re-
corded.
If some means other than impingers is
used to condense moisture, prepare the con-
denser (and, if appropriate, silica gel for
condenser outlet) for use.
Using a tweezer or clean disposable surgi-
cal gloves, place a labeled (identified) and
weighed filter in the filter holder. Be sure
that the filter is properly centered and the
gasket properly placed so as not to allow the
sample gas stream to circumvent the filter.
Check filter for tears after assembly is com-
pleted. Mark the probe extension with heat
resistant tape or by some other method to
denote the proper distance into the stack or
duct for each sampling point.
Assemble the train as in Figure 17-1, using
a very light coat of silicone grease on all
ground glass joints and greasing only the
outer portion (see APTD-0576) to avoid pos-
sibility of contamination by the silicone
grease. Place crushed ice around the im-
pingers.
4.1.4 Leak Check Procedures.
4.1.4.1 Pretest Leak-Check. A pretest
leak-check is recommended, but hot re-
quired. If the tester opts to conduct the pre-
test leak-check, the following procedure
shall be used.
After the sampling train has been assem-
bled, plug the inlet to the probe nozzle with
a material that will be able to withstand the
stack temperature. Insert the. filter holder
into the stack and wait approximately 5
minutes (or longer, if necessary) to allow
the system to come to equilibrium with the
temperature of the stack gas stream. Turn
on the pump and draw a vacuum of at least
380 mm Hg (15 in. Hg); note that a lower
vacuum may be used, provided that it is not
exceeded during the test. Determine the
leakage rate. A leakage rate in excess of 4
percent of the average sampling rate or
0.00057 m'/min. (0.02 cfm), whichever is
less, is unacceptable.
The following leak-check instructions for
the sampling train described in APTD-0576
and APTD-0581 may be helpful. Start the
pump with by-pass valve fully open and
coarse adjust valve completely closed. Par-
tially open the coarse adjust valve and
slowly close the by-pass valve until the de-
sired vacuum is reached. Do not reverse di-
rection of by-pass valve. If the desired
vacuum is exceeded, either leak-check at
this higher vacuum or end the leak-check as
shown below and start over.
When the leak-check is completed, first
slowly remove the plug from the inlet to the
probe nozzle and immediately turn off the
vacuum pump. This prevents water from
being forced backward and keeps silica gel
from being entrained backward.
4.1.4.2 Leak-Checks During Sample Run.
If, during the sampling run, a component
(e.g., filter assembly or impinger) change be-
comes necessary, a leak-check shall be con-
ducted immediately before the change is
made. The leak-check shall be done accord-
ing to the procedure outlined in Section
4.1.4.1 above, except that it shall be done at
a vacuum equal to or greater than the maxi-
mum value recorded up to that point in the
test. If the leakage rate is found to be no
greater than 0.00057 mVmin (0.02 cfm) or 4
percent of the average sampling rate
(whichever is less), the results are accept-
able, and no correction will need to be ap-
plied to the total volume of dry gas metered;
B-67
-------
Chapter I—Environmental Protection Agency
App. A
if, however, s higher leakage rate is ob-
tained, the tester shall either record the
leakage rate and plan to correct the sample
volume as shown in Section 6.3 of this
method, or shall void the sampling run.
Immediately after component changes,
leak-checks are optional; if such leak-checks
are done, the procedure outlined in Section
4.1.4.1 above shall be used.
4.1.4.3 Post-Test Leak-Check. A leak-
check Is mandatory at the conclusion of
each sampling run. The leak-check shall be
done in accordance with the procedures out-
lined in Section 4.1.4.1, except that it shall
be conducted at a vacuum equal to or great-
er than the maximum value reached during
the sampling run. If the leakage rate is
found to be no greater than 0.00057 m'/mln
(0.02 cfm) or 4 percent of the average sam-
pling rate (whichever Is less), the results are
acceptable, and no correction need be ap-
plied to the total volume of dry gas metered.
If, however, a higher leakage rate Is ob-
tained, the tester shall either record the
leakage rate and correct the sample volume
as shown in Section 6.3 of this method, or
shall void the sampling run.
4.1.5 Particulate Train Operation.
During the sampling run, maintain a sam-
pling rate such that sampling is within 10
percent of true isokinetic, unless otherwise
specified by the Administrator.
For each run, record the data required on
the example data sheet shown in Figure 17-
3. Be sure to record the initial dry gas meter
reading. Record the dry gas meter readings
at the beginning and end of each sampling
time increment, when changes in flow rates
are made, before and after each leak check,
and when sampling is halted. Take other
readings required by Figure 17-3 at least
once at each sample point during each time
increment and additional readings when sig-
nificant changes (20 percent variation in ve-
locity head readings) necessitate additional
adjustments in flow rate. Level and zero the
manometer. Because the manometer level
and zero may drift due to vibrations and
temperature changes, make periodic checks
during the traverse.
E-68
-------
App. A
Title 40—Protection of Environment
U
cc
8
ce
a.
S
£
|
1
*
i
i-
i
s
K
r
i
u
2
E
t-
(9
Z
UJ
_J
Z
o
PROBE EXTEN
u
»-
a
z
E
C
5
c
UL
t-
z
UJ
c
u
R
a
z
e
x
1
CC
G
i
t
=
!'s
o **
X *"
i
.£
'E
in
I*
<
CC
X
<
UJ
«J
H
•S
=
f
. E
• c
- uf
GC
3
STATIC PRESS
i
•
g
i
g
u
<
g
u
E:
?
1
n.
U
u.
u.
UI
s
m
O
K '
SoSj
fi Ji
^
' i|
*i
UJ °
§5
^
ac P CA ui
fpi
0
Xffl*
S*"
• •
i£
«C
•
\
I
3 til
s^
IS
.
5 =
£"?
io
jji
E?l
ilS
" E •—
s
e
3
™ -E
r ffS,
s °u
III
is
UJ JH
r
O1.
^
O)
«
I
UJ
U
<
?
I
T
B-69
-------
Chapter I—Environmental Protection Agency
App. A
Clean the portholes prior to the test run
to minimize the chance of sampling the de-
posited material. To begin sampling, remove
the nozzle cap and verify that the pitot tube
and probe extension are properly posi-
tioned. Position the nozzle at the first tra-
verse point with the tip pointing directly
into the gas stream. Immediately start the
pump and adjust the flow to isokinetic con-
ditions. Nomographs are available, which
aid in the rapid adjustment to the isokinetic
sampling rate without excessive computa-
tions. These nomographs are designed for
use when the Type S pitot tube coefficient
is 0.85±0.02, and the stack gas equivalent
density (dry molecular weight) is equal to
28 ±4. AFTD-0576 details the procedure for
using the nomographs. If C, and Md are out-
side the above stated ranges, do not use the
nomographs unless appropriate steps (see
Citation 7 In Section 7) are taken to com-
pensate for the deviations.
When the stack is under significant nega-
tive pressure (height of impinger stem),
tajce care to close the coarse adjust valve
before Inserting the probe extension assem-
bly Into the stack to prevent water from
being forced backward. If necessary, the
pump may be turned on with the coarse
adjust valve closed.
When the probe is in position, block off
the openings around the probe and porthole
to prevent unrepresentative dilution of the
gas stream.
Traverse the stack cross section, as re-
quired by Method 1 or as specified by the
Administrator, being careful not to bump
the probe nozzle into the stack walls when
sampling near the walls or when removing
or Inserting the probe extension through
the portholes, to minimize chance of ex-
tracting deposited material.
During the test run, take appropriate
steps (e.g., adding crushed ice to the im-
pinger ice bath) to maintain a temperature
of less than 20° C (68° P) at the condenser
outlet; this will prevent excessive moisture
losses. Also, periodically check the level and
zero of the manometer.
If the pressure drop across the filter be-
comes too high, making Isokinetic sampling
difficult to maintain, the filter may be re-
placed in the midst of a sample run. It is
recommended that another complete filter
holder assembly be used rather than at-
tempting to change the fliter itself. Before a
new filter holder Is installed, conduct a leak
check, as outlined In Section 4.1.4.2. The
total particulate weight shall include the
summation of all filter assembly catches.
A single train shall be used for the entire
sample run, except in cases where simulta-
neous sampling is required in two or more
separate ducts or at two or more different
locations within the same duct, or, in cases
•where equipment failure necessitates a
change of trains. In all other situations, the
use of two or more trains will be subject to
the approval of the Administrator. Note
that when two or more trains are used,' a
separate analysis of the collected particu-
late from each train shall be performed,
unless identical nozzle sizes were used on all
trains, in which case the particulate catches
from the individual trains may be combined
and a single analysis performed.
At the end of the sample run, turn off the
pump, remove the probe extension assembly
from the stack, and record the final dry gas
meter reading. Perform a leak-check, as out-
lined In Section 4.1.4.3. Also, leak-check the
pitot lines as described in Section 3.1 of
Method 2; the lines must pass this leak-
check, in order to validate the velocity head
data.
4.1.6 Calculation of Percent Isokinetic.
Calculate percent isokinetic (see Section
6.11) to determine whether another test run
should be made. If there Is difficulty in
maintaining isokinetic rates due to source
conditions, consult with the Administrator
for possible variance on the isokinetic rates.
4.2 Sample Recovery. Proper cleanup
procedure begins as soon as the probe ex-
tension assembly is removed from the stack
at the end of the sampling period. Allow the
assembly to cool.
When the assembly can be safely handled,
wipe off all external particulate matter near
the tip of the probe nozzle and place a cap
over it to prevent losing or gaining particu-
late matter. Do not cap off the probe tip
tightly while the sampling train is cooling
down as this would create a vacuum in the
filter holder, forcing condenser water back-
ward.
Before moving the sample train to the
cleanup site, disconnect the filter holder-
probe nozzle assembly from the probe ex-
tension; cap the open Inlet of the probe ex-
tension. Be careful not to lose any conden-
sate, if present. Remove the umbilical cord
from the condenser outlet and cap the
outlet. If a flexible line is used between the
first Impinger (or condenser) and the probe
extension, disconnect the line at the probe
extension and let any condensed water or
liquid drain into the impingers or condens-
er. Disconnect the probe extension from the
condenser; cap the probe extension outlet.
After wiping off the silicons grease, cap off
the condenser inlet. Ground glass stoppers,
plastic caps, or serum caps (whichever are
appropriate) may be used to close these
openings.
Transfer both the filter holder-probe
nozzle assembly and the condenser to the
cleanup area. This area should be clean and
protected from the wind so that the chances
of contaminating or losing the sample will
be minimized.
Save a portion of the acetone used for
cleanup as a blank. Take 200 ml of this ac-
B.-70
-------
APR. A
Title 40—Protection of Environment
etone directly from the wash bottle being
used and place it in a glass sample container
labeled "acetone blank."
Inspect the train prior to and during dis-
assembly and note any abnormal conditions.
Treat the samples as follows:
Container No. 1. Carefully remove the
filter from the filter holder and place it in
its identified petri dish container. Use a pair
of tweezers and/or clean disposable surgical
gloves to handle the filter. If it is necessary
to fold the filter, do so such that the partic-
ulate cake is inside the fold. Carefully trans-
fer to the petri dish any particulate matter
and/or filter fibers which adhere to the
filter holder gasket, by using a dry Nylon
bristle brush and/or a sharp-edged blade.
Seal the container.
Container No. 2. Taking care to see that
dust on the outside of the probe nozzle or
other exterior surfaces does not get into the
sample, quantitatively recover particulate
matter or any condensate from the probe
nozzle, fitting, and front half of the filter
holder by washing these components with
acetone and placing the wash in a glass con-
tainer. Distilled water, may be used instead
of acetone when approved by the Adminis-
trator and shall be used when specified by
the Administrator; in these cases, save a
water blank and follow Administrator's di-
rections on analysis. Perform the acetone
rinses as follows:
Carefully remove the probe nozzle and
clean the inside surface by rinsing with ac-
etone from a wash bottle and brushing with
a Nylon bristle brush. Brush until 'acetone
rinse shows no visible particles, after which
make a final rinse of the inside surface with
acetone.
Brush and rinse with acetone the Inside
parts of the fitting in a similar way until no
visible particles remain. A funnel (glass or
polyethylene) may be used to aid in trans-
ferring liquid washes to the container. Rinse
the brush with acetone and quantitatively
collect these washings in the sample con-
tainer. Between sampling runs, keep
brushes clean and protected from contami-
nation.
After ensuring that all joints are wiped
clean of silicone grease (if applicable), clean
the inside of the front half of the filter
holder by rubbing the surfaces with a Nylon
bristle brush and rinsing with acetone.
Rinse each surface three times or more if
needed to remove visible particulate. Make
final rinse of the brush and filter holder.
After all acetone washings and particulate
matter are collected in the sample contain-
er, tighten the lid on the sample container
so that acetone will not leak out when it is
shipped to the laboratory. Mark the height
of the fluid level to determine whether or
not leakage occurred during transport.
Label the container to clearly identify its
contents.
Container No. 3. If silica gel is used in the
condenser system for mositure content de-
termination, note the color of the gel to de-
termine if it has been completely spent;
make a notation of its condition. Transfer
the silica gel back to its original container
and seal. A funnel may make it easier to
pour the silica gel without spilling, and a
rubber policeman may be used as an aid in
removing.the silica gel. It is not necessary to
remove the small amount of dust particles
that may adhere to the walls and are diffi-
cult to remove. Since the gain in weight is to
be used for moisture calculations, do not use
any water or other liquids to transfer the
silica gel. If a balance is available in the
field, follow the procedure for Container
No. 3 under "Analysis."
Condenser Water. Treat the condenser or
impinger water as follows: make a notation
of any color or film in the liquid catch.
Measure the liquid volume to within +1 ml
by using a graduated cylinder or, if a bal-
ance is available, determine the liquid
weight to within ±0.5 g. Record the total
volume or weight of liquid present. This in-
formation is required to calculate the mois-
ture content of the effluent gas. Discard the
liquid after measuring and recording the
volume or weight.
4.3 Analysis. Record the data required on
the example sheet' shown in Figure 17-4.
Handle each sample container as follows:
Container No. 1. Leave the contents in the
shipping container or transfer the filter and
any loose particulate from the sample con-
tainer to a tared glass weighing dish. Desic-
cate for 24 hours in a desiccator containing
anhydrous calcium sulfate. Weigh to a con-
stant weight and report the results to the
nearest 0.1 mg. For purposes of this Section,
4.3, the term "constant weight" means a dif-
ference of no more than 0.5 mg or 1 percent
of total weight less tare weight, whichever is
greater, between two consecutive weighings,
with no less than 6 hours of desiccation
time between weighings.
Alternatively, the sample may be oven
dried at the average stack temperature or
105° C (220° F), whichever is less, for 2 to 3
hours, cooled in the desiccator, and weighed
to a constant weight, unless otherwise speci-
fied by the Administrator. The tester may
also opt to oven dry the sample at the aver-
age stack temperature or 105° C (220° F),
whichever is less, for 2 to 3 hours, weigh the
sample, and use this weight as a final
weight.
B-71
-------
Chapter I—Environmental Protection Agency
H»nt__
Date
App. A.
Run No..
Filter No.
Amount liquid lost during transport
Acttone blank volume, ml
Acetone wash volume, ml
Aettone Mack concentration, mg/mg (equation 17-4)
Acetone wash blank, mg (equation 17-5J
CONTAINER
NUMBER
1
2
TOTAL
WEIGHT OF PARTICULATE COLLECTED.
mg
FINAL WEIGHT
Z^^^^d
TARE WEIGHT
^xd
Less acetone blank
Weight of particulate matter
WEIGHT GAIN
FINAL
INITIAL
LIQUID COLLECTED
TOTAL VOLUME COLLECTED
VOLUME OF LIQUID
WATER COLLECTED
IMPINGER
VOLUME.
ml
SILICA GEL
WEIGHT,
g
fT" ml
* CONVERT WEIGHT OF WATER TO VOLUME BY DIVIDING TOTAL WEIGHT
INCREASE BY DENSITY OF WATER (1g/ml).
Figure 17-4. Analytical data.
INCREASE., 8VOLUMEWATB|
1 g/ml
B-72
-------
App. A
Container No. 2. Note the level of liquid in
the container and confirm on the analysis
sheet whether or not leakage occurred
during transport. If a noticeable amount of
leakage has occurred, either void the sample
or use methods, subject to the approval of
the Administrator, to correct the final re-
sults. Measure the liquid in this container
either volumetrically to ±1 ml or gravime-
trically to ±0.5 g. Transfer the contents to a
tared 250-ml beaker and evaporate to dry-
ness at ambient temperature and pressure.
Desiccate for 24 hours and weigh to a con-
stant weight. Report the results to the near-
est 0.1 mg.
Container No. 3. This step may be con-
ducted in the field. Weigh the spent silica
gel (or silica gel plus impinger) to the near-
est 0.5 g using a balance.
"Acetone Blank" Container. Measure ac-
etone in this container either volumetrically
or gravimetrically. Transfer the acetone to a
tared 250-ml beaker and evaporate to dry-
ness at ambient temperature and pressure.
Desiccate for 24 hours and weigh to a con-
stant weight. Report the results to the near-
est 0.1 mg.
NOTE: At the option of the tester, the con-
tents of Container No. 2 as well as the ac-
etone blank container may be evaporated at
temperatures higher than ambient. If evap-
oration is done at an elevated temperature,
the temperature must be below the boiling
point of the solvent; also, to prevent "bump-
ing," the evaporation process must be close-
ly supervised, and the contents of the
beaker must be swirled occasionally to main-
tain an even temperature. Use extreme care,
as acetone is highly flammable and has a
low flash point.
5. Calibration. Maintain a laboratory log
of all calibrations.
5.1 Probe Nozzle. Probe nozzles shall be
calibrated before their initial use in the
field. Using a micrometer, measure the
inside diameter of the nozzle to the nearest
0.025 mm (0.001 in.). Make three separate
measurements using different diameters
each time, and obtain the average of the
measurements. The difference between the
high and low numbers shall not exceed 0.1
mm (0.004 in.). When nozzles become
nicked, dented, or corroded, they shall be
reshaped, sharpened, and recalibrated
before use. Each nozzle shall be permanent-
ly and uniquely identified.
5.2 Pitot Tube. If the pitot tube is placed
in an interference-free arrangement with re-
spect to the other probe assembly compo-
nents, its baseline (isolated tube) coefficient
shall be determined as outlined in Section 4
of Method 2. If the probe assembly is not in-
terference-free, the pitot tube assembly co-
efficient shall be determined by calibration,
using methods subject to the approval of
the Administrator.
Title 40—Protection of Environment
5.3 Metering System. Before its initial
use in the field, the metering system shall
be calibrated according to the procedure
outlined in APTD-0576. Instead of physical-
ly adjusting the dry gas meter, dial readings
to correspond to the wet test meter read-
ings, calibration factors may be used to
mathematically correct the gas meter dial
readings to the proper values.
Before calibrating the metering system, it
is suggested that a leak-check be conducted.
For metering systems having diaphragm
pumps, the normal leak-check procedure
will not detect leakages within the pump.
For these cases the following leak-check
procedure is suggested: make a 10-minute
calibration run at 0.00057 m'/min (0.02
cfm); at the end of the run, take the differ-
ence of the measured wet test meter and
dry gas meter volumes; divide the difference
by 10, to get the leak rate. The leak rate
should not exceed 0.00057 m'/min (0.02
cfm).
After each field use, the calibration of the
metering system shall be checked by per-
forming three calibration runs at a single,
intermediate orifice setting (based on the
previous field test), with the vacuum set at
the maximum value reached during the test
series. To adjust the vacuum, insert a valve .
between the wet test meter and the inlet of
the metering system. Calculate the average
value of the calibration factor. If the cali-
bration has changed by more than 5 per-
cent, recalibrate the meter over the full
range of orifice settings, as outlined in
APTD-0576.
Alternative procedures, e.g., using the ori-
fice meter coefficients, may be used, subject
to the approval of'the Administrator.
NOTE: If the dry gas meter coefficient
values obtained before and after a test
series differ by more than 5 percent, the
test series shall either be voided, or calcula-
tions for the test series shall be performed
using whichever meter coefficient value
(i.e., before or after) gives the lower value of
total sample volume.
5.4 Temperature Gauges. Use the proce-
dure in Section 4.3 of Method 2 to calibrate
in-stack temperature gauges. Dial thermom-
eters, such as are used for the dry gas meter
and condenser outlet, shall be calibrated
against mercary-in-glass thermometers.
5.5 Leak Check of Metering System
Shown in Figure 17-1. That portion of the
sampling train from the pump to the orifice
meter should be leak checked prior to initial
use and after each shipment. Leakage after
the pump will result in less volume being re-
corded than is actually sampled. The follow-
ing procedure is suggested (see Figure 17-5).
Close the main valve on the meter box.
Insert a one-hole rubber stopper with
rubber tubing attached into the orifice ex-
B-73
-------
Chapter I—Environmental Protection Agency
haust pipe. Disconnect and vent the low side
of the orifice manometer. Close off the low
side orifice tap. Pressurize the system to 13
to 18 cm (5 to 7 in.) water column by blow-
ing into the rubber tubing. Pinch off the
App. A
tubing and observe the manometer for one
minute. A loss of pressure on the mano-
meter indicates a leak in the meter box;
leaks, if present, must be corrected.
B-74
-------
RUBBER
TUBING
RUBBER ADICIPE
STOPPER °RIFICE
VACUUM
GAUGE
CO
en
BLOW INTO TUBING
UNTIL MANOMETER
READS 5 TO 7 INCHES
WATER COLUMN
ORIFICE
MANOMETER
MAIN VALVE
CLOSED
AIR-TIGHT
PUMP
Figure 17-5. Leak check of meter box.
a
5'
f
i
-------
Chapter I—Environmental Protection Agency
App.A
5.6 Barometer. Calibrate against a mer-
cury barometer.
6. Calculations. Carry out calculations, re-
taining at least one extra decimal figure
beyond that of the acquired data. Round off
figures after the final calculation. Other
forms of the equations may be used as long
as they give equivalent results.
6.1 Nomenclature.
A,=Cross-sectional area of nozzle, m2 (ft2).
B.,=Water vapor in the gas stream, propor-
tion by volume.
C.=Acetone blank residue concentration,
mg/g.
c,=Concentration of participate matter in
stack gas, dry basis, corrected to stand-
ard conditions, g/dscm (g/dscf).
I=Percent of isokinetic sampling.
L,=Maximum acceptable leakage rate for
either a pretest leak check or for a leak
check following a component change;
equal to 0.00057 m'/min (0.02 cfm) or 4
percent of the average sampling rate,
whichever is less.
L,=Indivldual leakage rate observed during
the leak check conducted prior to the
"!"•" component change (1=1, 2, 3 ... n),
mVmin(cfm).,
L,=Leakage rate observed during the post-
test leak check, m Vmin (cfm).
m,,=Total amount of particulate matter col-
lected, tng.
Mw=Molecular weight of water, 18.0 g/g-
mole (18.0 Ib/lb-mole).
m.=Mass of residue of acetone after evapo-
ration, mg.
Ptar=Barometric pressure at the sampling
site, mm Hg (in. Hg).
P,=Absolute stack gas pressure, mm Hg (in.
Hg).
P.u«=Standard absolute pressure, 760 mm
Hg (29.92 in. Hg).
R=Ideal gas constant, 0.06236 mm Hg-mV
•K-g-mole (21.85 in. Hg-ft V°R-lb-mole).
T.,—Absolute average dry gas meter tem-
perature (see Figure 17-3), °K (°R).
T.=Absolute average stack gas temperature
(see Figure 17-3), °K OR).
T,u=Standard absolute temperature, 293°K
(528'R).
V«=Volume of acetone blank, ml.
V.w=Volume of acetone used in wash, ml.
Vic—Total volume of liquid collected in im-
pingers and silica gel (see Figure 17-4),
ml. .
V«,=Volume of gas sample as measured by
dry gas meter, dcm (dcf).
Vm(,u)=Volume of gas sample measured by
the dry gas meter, corrected to standard
conditions, dscm (dscf).
V.(,ui)=Volume of water vapor }n the gas
sample, corrected to standard condi-
tions, scm (scf).
v,«Stack gas velocity, calculated by Method
2, Equation 2-9, using data obtained
from Method 17, m/sec (ft/sec).
W.=Weight of residue in acetone wash, mg.
Y=Dry gas meter calibration coefficient.
AH=Average pressure differential across
the orifice meter (see Figure 17-3), mm
H»O (in. HaO).
p,=Density of acetone, mg/ml (see label on
bottle).
»w=Density of water, 0.9982 g/ml (0.002201
Ib/ml).
8- Total sampling time, min.
0,=Sampling time interval, from the begin-
ning of a run until the first component
change, min.
#,=Sampling time interval, between two
successive component changes, begin-
ning with the interval between the first
and second changes, min.
0B=Sampling time interval, from the final
(n"1) component change until the end of
the sampling run, min.
13.6=Specific gravity of mercury.
60=Sec/min.
100=Conversion to percent.
6.2 Average dry gas meter temperature
and average orifice pressure drop. See data
sheet (Figure 17-3).
6.3 Dry Gas Volume. Correct the sample
volume measured by the dry gas meter to
standard conditions (20° C, 760 mm Hg or
68° F, 29.92 in. Hg) by using Equation 17-1.
Vm(std) = Vin
Equation 17-1
where:
K,=0.3858° K/mm Hg for metric units;
17.64' R/in. Hg for English units.
NOTE: Equation 17-1 can be used as writ-
ten unless the leakage rate observed during
any of the mandatory leak checks (i.e., the
post-test leak check or leak checks conduct-
ed prior to component changes) exceeds L,.
If LP or L, exceeds L., Equation 17-1 must be
modified as follows:
(a) Case I. No component changes made
during sampling run. In this case, replace
Vm in Equation 17-1 with the expression:
B-76
-------
App. A
Title 40—Protection of Environment
-
(b) Case II. One or more component
changes made during the sampling run. In
this case, replace Vm in Equation 17-1 by the
expression:
L1 ' La>
'
-------
Chapter I—Environmental Protection Agency
K.=4.320 for metric units; 0.09450 for Eng-
lish units.
6.12 Acceptable Results. If 90 percent
-------
APPENDIX C
EPA REFERENCE METHODS 16..16A
C-l
-------
-------
APR. A
Title 40—Protection of Environment
METHOD 16—SEMICONTINUOUS DETERMINA-
TION OF SULFUR EMISSIONS PROM STATION-
ARY SOURCES
Introduction
The method described below uses the
principle of gas chromatographic separation
and flame photometric detection. Since
there are many systems or sets of operating
conditions that represent usable methods of
determining sulfur emissions, all systems
which employ this principle, but differ only
In details of equipment and operation, may
be used as alternative methods, provided
that the criteria set below are met.
1. Principle and Applicability.
1.1 Principle. A gas sample is extracted
from the emission source and diluted with
clean dry air. An aliquot of the diluted
sample is then analyzed for hydrogen sul-
fide (HjS), methyl mercaptan (MeSH), di-
methyl sulfide (DMS) and dimethyl disul-
fide (DMDS) by gas chromatographic
-------
Chapter I—Environmental Protection Agency
APR. A
from the sample. In the example system,
SO, is removed by a citrate buffer solution
prior to GC injection. This scrubber will be
used when SO, levels are high enough to
prevent baseline separation from the re-
duced sulfur compounds.
Compliance with this section can be dem-
onstrated by submitting chromatographs of
calibration gases with SO, present in the
same quantities expected from the emission
source to be tested. Acceptable systems
shall show baseline separation with the am-
plifier attenuation set so that the reduced
sulfur compound of concern is at least 50
percent of full scale. Base line separation is
defined as a return to zero ± percent in the
interval between peaks.
4. Precision and Accuracy.
4.1 OC/PPD and Dilution System Cali-
bration Precision. A series of three consecu-
tive injections of the same calibration gas,
at any dilution, shall produce results which
do not vary by more than ±5 percent from
the mean of the three injections.
4.2 GC/FPD and Dilution System Cali-
bration Drift. The calibration drift deter-
mined from the mean of three injections
made at the beginning and end of any 8-
hour period shall not exceed ± percent.
4.3 System Calibration Accuracy. Losses
through the sample transport system must
be measured and a correction factor devel-
oped to adjust the calibration accuracy to
100 percent.
5. Apparatus (See Figure 16-1).
5.1. Sampling.
6.1.1 Probe. The probe must be made of
inert material such as stainless steel or
glass. It should be designed to incorporate a
filter and to allow calibration gas to enter
the probe at or near the sample entry point.
Any portion of the probe not exposed to the
stack gas must be heated to prevent mois-
ture condensation.
5.1.2 Sample Line. The sample line must
be made of Teflon,' no greater than 1.3 cm
(V4) Inside diameter. All parts from the
probe to the dilution system must be ther-
mostatically heated to 120° C.
5.1.3 Sample Pump. The sample pump
shall be a leakless Teflon-coated diaphragm
type or equivalent. If the pump is upstream
of the dilution system, the pump head must
be heated to 120* C.
5.2 Dilution System. The dilution system
must be constructed such that all sample
contacts are made of inert materials (e.g.,
stainless steel or Teflon). It must be heated
to 120* C. and be capable of approximately a
9:1 dilution of the sample.
5.3 SO, Scrubber. The SO, Scrubber is a
midget impinger packed with glass wool to
'Mention of trade names or specific prod-
ucts does not constitute endorsement by the
Environmental Protection Agency.
eliminate entrained mist and charged with
potassium citrate-citric acid buffer.
5.4 Gas Chromatograph. The gas chro-
matograph must have at least the following
components:
5.4.1 Oven. Capable of maintaining the
separation column at the proper operating
temperature ±1° C.
5.4.2 Temperature Gauge. To monitor
column oven, detector, and exhaust tem-
perature ±1° C.
5.4.3 Flow System. Gas metering system
to measure sample, fuel, combustion gas,
and carrier gas flows.
5.4.4 Flame Photometric Detector.
5.4.4.1 Electrometer. Capable of full scale
amplification of linear ranges of 10"9to 10~4
amperes full scale.
5.4.4.2 Power Supply. Capable of deliver-
ing up to 750 volts.
5.4.4.3 Recorder. Compatible with the
output voltage range of the ele'ctrometer.
5.5 Gas Chromatograph Columns. The
column system must be demonstrated to be
capble of resolving the four major reduced
sulfur compounds: HaS, MeSH, DMS, and
DMDS. It must also demonstrate freedom
from known interferences.
To demonstrate that adequate resolution
has been achieved, the tester must submit a
Chromatograph of a calibration gas contain-
ing all four of the TRS compounds in the
concentration range of the applicable stand-
ard. Adequate resolution will be defined as
base line separation of adjacent peaks when
the amplifier attenuation is set so that the
smaller peak is at least 50 percent of full
scale. Base line separation is defined in Sec-
tion 3.4. Systems not meeting this criteria
may be considered alternate methods sub-
ject to the approval of the Administrator.
5.6 Calibration System. The calibration
system must contain the following compo-
nents.
5.6.1 Tube Chamber. Chamber of glass or
Teflon of sufficient dimensions to house
permeation tubes.
5.6.2 Flow System. To measure air flow
over permeation tubes at ±2 percent. Each
flowmeter shall be calibrated after a com-
plete test series with a wet test meter. If the
flow measuring device differs from the wet
test meter by 5 percent, the completed test
shall be discarded. Alternatively, the tester
may elect to use the flow data that would
yield the lower flow measurement. Calibra-
tion with a wet test meter before a test is
optional.
5.6.3 Constant Temperature Bath. Device
capable of maintaining the permeation
tubes at the calibration temperature within
±0.1' C.
5.6.4 -Temperature Gauge. Thermometer
or equivalent to monitor bath temperature
within ±r C.
6. Reagents.
C-4
-------
APR. A
Title 40—Protection of Environment
6.1 Fuel. Hydrogen (Hi) prepurified
grade or better.
6.2 Combustion Gas. Oxygen (O.) or air,
research purity or better.
6.3 Carrier Gas. Prepurified grade or
better.
6.4 Diluent. Air containing less than 50
ppb total sulfur compounds and less than 10
ppm each of moisture and total hydrocar-
bons. This gas must be heated prior to
mixing with the sample to avoid water con-
densation at the point of contact.
• 6.5 Calibration Gases. Permeation tubes,
one each of HiS, MeSH, DMS, and DMDS,
agravimetrically calibrated and certified at
some convenient operating temperature.
These tubes consist of hermetically sealed
PEP Teflon tubing in which a liquified gas-
eous substance is enclosed. The enclosed gas
permeates through the tubing wall at a con-
stant rate. When the temperature is con-
stant, calibration gases Governing a wide
range of known concentrations can be gen-
erated by varying and accurately measuring
the flow rate of diluent gas passing over the
tubes. These calibration gases are used to
calibrate the GC/PPD system arid the dilu-
tion system.
6.6 Citrate Buffer. Dissolve 300 grams of
potassium citrate and 41 grams of anhy-
drous citric acid in 1 liter of deionized water.
284 grams of sodium citrate may be substi-
tuted for the potassium citrate.
7. Pretest Procedures. The following proce-
dures are optional but would be helpful in
preventing any problem which might occur
later and invalidate the entire test.
7.1 After the complete measurement
system has been set up at the site arid
deemed to be operational, the following pro-
cedures should be completed before sam-
pling is initiated.
7.1.1 Leak Test. Appropriate leak test
procedures should be employed to verify the
integrity of all components, sample lines,
and connections. The following leak test
procedure is suggested: For components up-
stream of the sample pump, attach the
probe end of the sample line to a ma- no-
meter or vacuum gauge, start the pump and
pull greater than 50 mm (2 in.) Hg vacuum,
close off the pump outlet, and then stop the
pump and ascertain that there is no leak for
1 minute. For components after the pump,
apply a slight positive pressure and check
for leaks by applying a liquid (detergent in
water, for example) at each joint. Bubbling-
indicates the presence of a leak.
7.1.2 System Performance. Since the
complete system is calibrated following each
test, the precise calibration of each compo-
nent is not critical. However, these compo-
nents should be verified to be operating
properly. This verification can be performed
by observing the response of flowmeters or
of the GC output to changes in flow rates or
calibration gas concentrations and ascer-
taining the response to be within predicted
limits. In any component, or if the complete
system fails to respond in a normal and pre-
dictable manner, the source of the discrep-
ancy should be identified and corrected
before proceeding.
8. Calibration. Prior to any sampling run,
calibrate the system using the following
procedures. (If more than one run is per-
formed during any 24-hour period, a calibra-
tion need not be performed prior to the
second and any subsequent runs. The cali-
bration must, however, be verified as pre-
scribed in Section 10, after the last run
made within the 24-hour period.)
8.1 General Considerations. This section
outlines steps to be followed for use of the
GC/FPD and the dilution system. The pro-
cedure does not include detailed instruc-
tions because the operation of these systems
is complex, and it requires a understanding
of the individual system being used. Each
system should include a written operating
manual describing in detail the operating
procedures associated with each component
in the measurement systerii. In addition, the
operator should be familiar with the operat-
ing principles of the components; particular-
ly the GC/FPD. The citations in the Bib-
liography at the end of this method are rec-
ommended for review for this purpose.
8.2 Calibration Procedure. Insert the per-
meation tubes into the tube chamber.
Check the bath temperature to assure
agreement with the calibration temperature
of the tubes within ±0.1° C. Allow 24 hours,
for the tubes to equilibrate. Alternatively
equilibration may be verified by injecting
samples of calibration gas at 1-hour inter-
vals. The permeation tubes can be assumed
to have reached, equilibrium when consecu-
tive hourly samples agree within the preci-
sion limits of Section 4.1.
Vary the amount of air flowing over the
tubes to produce the desired concentrations
for calibrating the analytical and dilution
systems. The air flow across the tubes must
at all times exceed the flow requirement of
the analytical systems. The concentration in
parts per million generated by a tube con-
taining a specific permeant can be calculat-
ed as follows:
C = K
where:
Equation 16-1
C-5
-------
APP.A
Till* 40—Protection of Environment
10.2 Recallbratlon. After each run, or
after a series of runs made within a 24-hour
period, perform a partial recalibration using
the procedures in Section 8. Only HjS (or
other permeant) need be used to recalibrate
the GC/FPD analysis system (8.3) and the
dilution system (8.5).
10.3 Determination of Calibration Drift.
Compare the calibration curves obtained
prior to the runs, to the calibration curves
obtained under paragraph 10.1. The calibra-
tion drift should not exceed the limits set
forth in subsection 4.2. If the drift exceeds
this limit, the intervening run or runs
should be considered not valid. The tester,
however, may instead have the option of
choosing the calibration data set which
would give the highest sample values.
11. Calculations.
11.1 Determine the concentrations of
each reduced sulfur compound detected di-
rectly from the calibration curves. Alterna-
tively, the concentrations may be calculated
using the'equation for the least square line.
11.2 Calculation of TRS. Total reduced
sulfur will be determined for each anaylsis
made by summing the concentrations of
each reduced sulfur compound resolved
during a given analysis.
TRS=2 (H.S. MeSH, DMS, 2DMDS)d
Equation 16-2
where:
TRS=Total reduced sulfur in ppm, wet
basis.
HiS-Hydrogen sulfide, ppm.
MeSH=Methyl mercaptan. ppm.
DMS - Dimethyl sulfide, ppm.
DMDS=Dimethyl disulfide, ppm.
d-Dilution factor, dimensionless.
11.3 Average TRS. The average TRS will
be determined as follows:
Average TRS=
Average TRS=
Average TRS=Average total reduced suflur
in ppm, dry basis.
TRS|=Total reduced sulfur in ppm as deter-
mined by Equation 16-2.
N=Number of samples.
B».=Praction of volume of water vapor in
the gas stream as determined by refer-
ence method 4—Determination of Mois-
ture in Stack Gases (36 PR 24887).
11.4 Average concentration of individual
reduced sulfur compounds.
Equation 16-3
where:
S,=Concentration of any reduced sulfur
compound from the ith sample injec-
tion, ppm.
C=Average concentration of any one of the
reduced sulfur compounds for the entire
run, ppm.
N=Number of injections in any-run period.
12. Example System. Described below is a
system utilized by EPA in gathering NSPS
data. This system does not now reflect all
the latest developments in equipment and
column technology, but it does represent
one system that has been demonstrated to
work.
12.1 Apparatus.
12.1.1 Sampling System.
12.1.1.1 Probe. Figure 16-1 illustrates the
probe used in lime kilns and other sources
where significant amounts of particulate
matter are present, the probe is designed
with the deflector shield placed between the
sample and the gas inlet holes and the glass
wool plugs to reduce clogging of the filter
and possible adsorption of sample gas. The
exposed portion of the probe between the
sampling port and the sample line is heated
with heating tape.
12.1.1.2 Sample Line 9U inch inside diam-
eter Teflon tubing, heated to 120° C. This
temperature is controlled by a thermostatic
heater.
12.1.1.3 Sample Pump. Leakless Teflon
coated diaphragm type or equivalent. The
pump head is heated to 120° C by enclosing
it in the sample dilution box (12.1.2.4
below).
12.1.2 Dilution System. A schematic dia-
gram of the dynamic dilution system is
given in Figure 16-2. The dilution system is
constructed such that all sample contacts
are made of inert materials. The dilution
system which is heated to 120° C must be ca-
pable of a minimum of 9:1 dilution of
sample. Equipment used in the dilution
system is listed below:
12.1.2.1 Dilution Pump. Model A-150
Kohmyhr Teflon positive displacement
type, nonadjustable 150 cc/min. ±2.0 per-
cent, or equivalent, per dilution stage. A 9:1
dilution of sample is accomplished by com-
bining 150 cc of sample with 1,350 cc of
clean dry air as shown in Figure 16-2.
12.1.2.2 Valves. Three-way Teflon sole-
noid or manual type.
C-6
-------
Chapter I—Environmental Protection Agency
App. A
12.1.2.3 Tubing. Teflon tubing and fit-
tings are used throughout from the sample
probe to the GC/PPD to present an" inert
surface for sample gas.
12.1.2.4 Box. Insulated box, heated and
maintained at 120" C, of sufficient dimen-
sions to house dilution apparatus.
12.1.2.5 Plowmeters. Rotameters or
equivalent to measure flow from 0 to 1500
ml/mm ±1 percent per dilution stage.
12.1.3 SO2 Scrubber. Midget impinger with
15 ml of potassium citrate buffer to absorb
SO, in the sample.
12.1.4 Gas Chromatograph Columns.
Two types of columns are used for separa-
tion of low and high molecular weight
sulfur compounds:
12.1.4.1 Low Molecular Weight Sulfur
Compounds Column (GC/FPD-I).
12.1.4.1.1 Separation Column. 11 m by
2.16 mm (36 ft by 0.085 in) inside diameter
Teflon tubing packed with 30/60 .mesh
Teflon coated with 5 percent polyphenyl
ether and 0.05 percent orthophosphoric
acid, or equivalent (see Figure 16-3).
12.1.4.1.2 Stripper or Precolumn. 0.6 m
by 2.16 mm (2 ft by 0.085 in) inside diameter
Teflon tubing.
12.1.4.1.3 Sample Valve. Teflon 10-port
gas sampling valve, equipped with a 10 ml
sample loop, actuated by compressed air
(Figure 16-3).
12.1.4.1.4 Oven. For containing sample
valve, stripper column and separation
column. The oven should be capable of
maintaining an elevated temperature rang-
ing from ambient to 100° C, constant within
±r c.
12.1.4.1.5 Temperature Monitor. Thermo-
couple pyrometer to measure column oven,
detector, and exhaust temperature ±1' C.
12.1.4.1.6 Flow System. . Gas metering
system to measure sample flow, hydrogen
flow, and oxygen flow (and nitrogen carrier
gas flow).
12.1.4.1.7 Detector. Flame photometric
detector.
12.1.4.1.8 Electrometer. Capable of full
scale amplification of linear ranges of 10"s
to 10"' amperes full scale.
12.1.4.1.9 Power Supply. Capable of deli-
vering up to 750 volts.
12.1.4:1.10 Recorder. Compatible with
the output voltage range of the electrom-
eter.
12.1.4.2 High Molecular Weight Com-
pounds Column (GC/FPD-II).
12.1.4.2.il Separation Column. 3.05 m by
2.16 mm (10 ft by 0.0885 in) inside diameter
Teflon tubing packed .with 30/60 mesh
Teflon coated with 10 percent Triton X-305,
or equivalent.
12.1.4.2.2 Sample Valve. Teflon 6-port gas
sampling valve equipped with a 10 ml
sample loop, actuated by compressed air
(Figure 16-3).
12.1.4.2.3 Other Components. All compo-
nents same, as in 12.1.4.1.5 to 12.1.4.1.10.
12.1.5 'Calibration. Permeation tube
system (figure 16-4).
12.1.5.1 Tube Chamber. Glass chamber
of sufficient dimensions to house perme-
ation tubes.
12.1.5.2 Mass Flowmeters. Two mass
flowmeters in the range 0-3 1/min. and 0-10
1/min. to measure air flow over permeation
tubes at ±2 percent. These flowmeters shall
be cross-calibrated at the beginning of each
test. Using a convenient flow rate in the
measuring range of both flowmeters, set
and monitor the flow rate of gas over the
permeation tubes. Injection of calibration
gas generated at this flow rate as measured
by one flowmeter followed by injection of
calibration gas at the same flow rate as
measured by the other flowmeter should
agree within the specified precision limits.
If they do not, then 'there is a problem with
the mass flow measurement. Each mass
flowmeter shall be calibrated prior to the
first test with a wet test meter and thereaf-
ter, at least once each year.
12.1.5.3 Constant Temperature Bath. Ca-
pable of maintaining permeation tubes at
certification temperature of 30° C. within
±0.1° C.
12.2 Reagents
12.2.1 Fuel. Hydrogen (H,) prepurified
grade or better.
12.2.2. Combustion Gas. Oxygen (O3) re-
search purity or better.
12.2.3 Carrier Gas. Nitrogen (N2) prepuri-
fied grade or better.
12.2.4 Diluent. Air containing less than
50 ppb total sulfur compounds and less than
10 ppm each of moisture and total hydro-
carbons, and filtered using MSA filters
46727 and 79030, or equivalent. Removal of
sulfur compounds can be verified by inject-
ing dilution air only, described in Section
8.3.
12.2.5 Compressed Air. 60 psig for GC
valve actuation.
12.2.6 Calibrated Gases. ' Permeation
tubes gravimetrically calibrated and certi-
fied at 30.0° C.
12.2.7 Citrate Buffer. Dissolve 300 grams
of potassium -citrate and 41 grams of anhy-
drous citric acid in 1 liter of deionized water.
284 grams of sodium citrate may be substi-
tuted for the potassium citrate.
12.3 Operating Parameters.
12.3.1 Low-Molecular Weight Sulfur
Compounds. The operating parameters for
the GC/FPD system used for low molecular
weight compounds are as follows: nitrogen
carrier gas flow rate of 50 cc/min, exhaust
temperature of 110° C, detector temperature
of 105° C, oven temperature of 40° C, hydro-
gen flow rate of 80 cc/min, oxygen flow rate
of 20 cc/min, and sample flow rate between
20 and 80 cc/min.
C-7
-------
App. A
12.3.2 High-Molecular Weight „ Sulfur
Compounds. The operating parameters for
the GC/FPD system for high molecular
weight compounds are the same as in 12.3.1
except: oven temperature of 70° C, and ni-
trogen carrier gas flow of 100 cc/min.
12.4 Analysis Procedure.
12.4.1 Analysis. Aliquots of diluted
sample are injected simultaneously into
both GC/PPD analyzers for analysis. GC/
FPD-I is used to measure the low-molecular
weight reduced sulfur compounds. The low
molecular weight compounds Include hydro-
gen sulfide, methyl mercaptan, and di-
methyl sulfide. GC/FPD-II is used to re-
solve the high-molecular weight compound.
The high-molecular weight'compound is di-
methyl dlsulf ide.
12.4.1.1 Analysis of Low-Molecular
Weight Sulfur Compounds. The sample
valve is actuated for 3 minutes in which
time an aliquot of diluted sample is injected
into the stripper column and analytical
column. The valve is then deactivated for
approximately 12 minutes In which time.
the analytical column continues to be fore-
flushed, the stripper column Is backflushed,
and the sample loop is refilled. Monitor the
responses. The elution time for each com-
pound will be determined during calibra-
tion.
12.4.1.2 Analysis of High-Molecular
Weight Sulfur Compounds. The procedure
IB essentially the same as above except that
no stripper column Is needed.
13. Bibliography.
Title 40—Protection of Environment
13.1 O'Keeffe, A. E. and G. C. Ortman.
"Primary Standards for Trace Gas Analy-
sis." Analytical Chemical Journal, 38,760
(1966).
13.2 Stevens, R. K., A. E. O'Keeffe, and
. G. C. Ortman. "Absolute Calibration of a
Flame Photometric Detector to Volatile
Sulfur Compounds at Sub-Part-Per-Million
Levels." Environmental Science and Tech-
nology, 3:7 (July, 1969).
13.3 Mulick, J. D., B. K. Stevens, and R.
Baumgardner. "An Analytical System De-
signed to Measure Multiple Malodorous
Compounds Related to Kraft Mill Activi-
ties." Presented at the 12th Conference on
Methods in Air Pollution and Industrial Hy-
giene Studies, University of Southern Cali-
fornia, Los Angeles, CA. April 6-8,1971.
13.4 Devonald, R. H., R. S. Serenius, and
A. D. Mclntyre. "Evaluation of the Flame
Photometric Detector for Analysis of Sulfur
Compounds." Pulp and Paper Magazine of
Canada, 73,3 (March, 1972).
13.5 Grimley, K. W., W. S. Smith, and R.
M. Martin. "The Use of a Dynamic Dilution
System in the Conditioning of Stack Gases
for Automated Analysis by a Mobile Sam-
pling Van." Presented at the 63rd Annual
APCA Meeting in St. Louis, Mo. June 14-19,
1970.
13.6 General Reference. Standard Meth-
ods of Chemical Analysis Volume III A and
B Instrumental Methods. Sixth Edition.
Van Nostrand Reinhold Co.
C-8
-------
Chapter I—Environmental Protection Agency
App. A
•o
10
5
f.
to
i
o
u
M
cn
«>
I-
in
a>
1
0.
C-9
-------
TOCC/FPDAIALYZEM
19:1
FILTER
(CLASS WOOL)
O
(-»
O
•p
>
FLOVVMETER
DIAPHRAGM
PUMP
(HEATED)
Figure 16-2.Sampl!ng and dilution appiratui
DILUTION BOX HEATED
TOIOfl'C
VEMT
s.
m
I
-------
Chapter I—Environmental Protection Agency
App.A
C-ll
-------
TO raSTRUMEHTS
AND
DILUTION SYSTEM
O
I
CONSTANT
TEMPERATURE
BATH
DILUENT
DRIER ^-A0'g
NITROGEN
GLASS
CHAMBER
PERMEATION
TUBE
•p
>
Figure 16-4. Apparatus for field calibration.
i
f
I
8.
m
I
-------
Chapter I—Environmental Protection Agency
App.A
OS
0«*
Itu
GC
X
-j a.
II
I HI
;o
• c
'.9
c
e
1
o
in
a
iE
SV3
NOI1VU8I1V3
C-13
-------
Federal Register / Vol. 46, No. 117 / Thursday, June 18, 1981 / Proposed Rules
31905
1. By revising paragraph (d)(l) of
f 60.285 to read as follows:
5 60 285 Test methods and procedures.
* *
(dj* " *
(1) Method 18 or, at the discretion of
Iho owner or operator, Method 16A for
the concentration of TRS,
*****
2, By amending Appendix A by adding
a new method as follows: '
Appendix A—Reference Methods
*****
Method 1GA. Determination of Total Reduced
Sulfur Emissions From Stationary Sources
(Impinger Technique)
1. Applicability and Principle—1.1
Applicability. This is° an alternative method
to Method 16 for determining total reduced
sulfur (TRS) compounds from recovery
furnaces, lime kilns, and smelt dissolving
tanks at kraft pulp mills. The TRS compounds
include hydrogen sulflde, methyl mercaptan,
dimethyl sulflde, dimethyl dlsulfide, and
other reduced sulfur compounds (e.g.,
cabonyl sulflde, if present). Therefore,
Method 16A might yield higher TRS
concentrations than Method 16.
The minimum detectable limit of the
method has been determined to be 0.04 ppm
TRS (compounds with single sulfur atom)
when sampling at 2 liters/min for 60 minutes.
For an analytical accuracy of at least ±5
percent, a minimum sulfur dioxide (SO?) mass
of 500 ug should be collected. The upper
concentration limit of the method generally
exceeds all encountered TRS levels from
kraft pulp mills.
1.2 Principle. A gas sample is extracted
from the sampling point in the stack. SO2 is
selectively removed from the sample using a
citrate buffer solution. Then reduced sulfur
compounds are oxidized and analyzed as SOa
using the barium-thorin titration procedure of
Method 6.
Z. Apparatus—2.1 Sampling. The sampling
train is shown in Figure 16A-1. The apparatus
is the same as listed in Method 6, except as
listed below. Other designs 'are acceptable
provided that the sampling system meets .the
performance check of Section 5.
2.1.1 SO2 Scrubber. Two midget impingers
in series packed with glass wool to eliminate
entrained mist and charged with potassium
citrate-citric acid buffer.
2.1.2 Combustion Tube. Quartz glass with
an expanded combustion chamber of 22 to 25
mm and at least 30.5 cm long. The tube ends
shall have an outside diameter of about 6 mm
to accept Teflon tubing or Swagelok fittings.
2.1.3 Combustion Tube Furnace. A
• furnace of sufficient size to enclose the
combustion chamber of the combustion tube
with a temperature regulator capable of
maintaining the temperature at 815 ±15°C.
2.1.4 Rate Meters. Rotameters, or
equivalent, capable of measuring flow rate to
within 2 percent of the selected flow rate.
2.1.5 Probe Brush. Nylon bristle brush
with stainless steel wire handle. The brush
shall be properly sized and shaped and of
sufficient length to brush out the entire length
of the probe.
2,2 Sample Recovery. Same as in Method
6, Section 2.2.
2.3 Analysis. Same as in Method 6,
Section 2.3, except a 10-ml buret with 0.1-ml
graduations is required and the
spectrophotorheter is not needed.
3. Reagents—Unless otherwise indicated,
all reagents must conform to the
specifications established by the Committee
on Analytical Reagents of the American
Chemical Society. Where such specifications
are not available, use the best available
3.1 Sampling. Th'e following reagents are
needed;
3.1.1 Water. Same as Method 6, Section
3.1.1.
3.1.2 Hydrogen Peroxide, 3 percent. Same
as Method 6, Section 3.1.3 (40 Ml is needed
per sample).
3.1.3 Citrate Buffer. Dissolve 300 g of
potassium citrate (or 284 g of sodium citrate)
and 41 g of anhydrou citric acid in 1 liter of
deionized distilled water.
3.1.4 Calibration Gas. Hydrogen sulflde in
nitrogen (30 to 50 ppm) stored in aluminum
cylinders. Verify the concentration by-
Method 11.
3.1.5 Combustion Gas. Air or oxygen
containing less than 50 ppb total sulfur
compounds and less than 10 ppm total
hydrocarbons.
3.2 Sample Recovery and Analysis.
Deionized distilled water (as in 3.1.1) and the
same reagents as in Method 6, Section 3,3,
are required.
4. Procedure—4.1 Sampling.
4.1.1 Preparation of Collection Train. For
the SOa scrubber, measure 20 ml citrate
buffer solution into each of'two midget
impingers with glass wool packed in topi For
the Method 6 part of the train, measure 20 ml
of 3 percent hydrogen peroxide into each of
the first two midget impingers. Leave the final
midget impinger dry. Assemble the train as
shown in figure 16A-1. Place the SO2
scrubber as close to the stack wall as
practical. Adjust the probe heater to a
temperature sufficient to prevent water
condensation. Maintain the oxidation furnace
at 815°C. Place crushed ice and water around
the impingers.
BILLING CODE 6560-26-M
C-14
-------
«•*=[ HEATED
PROBE
r>
t—i
en
S02SCRUBBER
THERMOMETER
OXIDATION
TUBE
IMPINGERS
TUBE FURNACE
I
jH1
>
"ICE BATH
SILICA GEL
DRYING TUBE
h
THERMOMETER
ft
Figure 16A-1. Sampling train
NEEDLE
VALVE
PUMP
•a
CD
9.
n
CD.
)_&
s°
co
CD
TJ
O
CD
CD
P*
CD
CO
-------
Federal Register / Vol. 46. No. 117 / Thursday. June 18. 1981 / Proposed Rules 31907
4.1.2 Leak-Check Procedure. Same as
Method 6. Section 4.1.2.
4.1.3 Sample Collection. Same as Method
6. Section 4.1.3, except for the following:
Adjust the sample flow to a constant rate of
approximately 2.0 lilers/min (±10 percent) as
Indicated by the rotameter. Other constant
flow rates may also be usgd provided its
acceptability is checked as in Section 5.
Collect the sample for 60 minutes. The 15-
mlnulo purge of the train following collection
need not be performed.
In Method 16, a sample run is composed of
10 individual analyses (injects) performed
over a period of not less than 3 hours or more
than 6 hours. For Method 16A to be
consistent with Method 16, the following may
be used to obtain a sample run: (1) collect
three GO-mfnute samples or (2) collect one 3-
hour sample with a total gas sample volume"*"
of 120 liters either intermittently (equal
samples, equally spaced) or continuously
over 3 hours.
After collecting the sample, disconnect the
probe and,tubing from the SOi scrubber and
allow to cool. Before conducting the next run,
do the following. Clean the inside surface of
the probe using a nylon brush and deionized
distilled water from a wash bottle.until the
rinse shows no visible particles. Replace the
probe filter. Thoroughly rinse the sample line
connecting the probe to the scrubber until all
visible particles are removed.
4,2 Sample Recovery. Disconnect the
implngcra. Replace the SO, scrubber contents
and the glass wool if saturated with solution
for subsequent runs. Pour the contents of the-
midget impingers of the Method 6 part of the
train into a leak-free polyethylene, bottle for
shipment. Rinse the three midget impingers,
the connecting tubes, and the sample line
between the furnace and the first impinger
with deionized distilled water, and add the
washings to the same storage container.
Mark the fluid level. Seal and identify the
sample container.
4.3 Sample Analysis. Note level of liquid
in container, and confirm whether any
sample was lost during shipment; note this on
analytical data sheet. If a noticeable amount
of leakage has occurred, either void the
sample or use methods, subject to the
approval of the Administrator, to correct the
final results.
Transfer the contents of the storage
container to a 100-ml graduated cylinder.
Rinse the container with deionized distilled
water and add to the cylinder. Measure the
volume, and pour into a 250-ml Erlenmeyer
flask. Using the cylinder, add sufficient 100
percent isopropanol to give a final sample
concentration of 80 percent (v/v) isopropanol.
Add four to six drops of thorin indicator and
titrate to a pink end point using 0.0100N
barium perchlorate. Run a blank with each
series of samples.
Note.—Protect the 0.0100 N barium
perchlorate solution from evaporation at all
times.
5. Calibration—5.1 Metering System,
Thermometers, Rotameters, Barometer, and
Barium Perchlorate Solution. Fellow the same
calibration procedure as in Method 6,
Sections 5.1 to 5.5, respectively.
5.2 System Performance Check. Using HjS
cylinder gas and combustion gas (as specified
in Sections 3.1.4 and 3.1.5), generate a series
of samples in the suspected concentration
range of TRS in the stack. Using the set-up
shown in Figure 16A-2, take at least two 30-
minute samples to determine system
performance'efficiency. Use the cylinder
regulator to set the combustion gas rotameter
flow rate to the desired level. Adjust the H2S
regulator to .a slightly higher than desired
flow rate to ensure excess gas for the system.
With the pump valve completely closed, turn
on the pump and open the valve slowly until
a 2 liter/min flow rate (or other selected flow
rate) is obtained. Observe the pressure
control vessel while opening the valve and
during the sampling run to maintain an
excess flow. The samples must be
transported through the entire sampling
system hi the normal manner. Compare the
resulting measured concentration to the
known concentration' by subtracting the
corrected volume of combustion gas from the
corrected total sample volume and treating as
in Section 6.3. The sampling system is .
considered acceptable when two consecutive
samples of calibration gas produce results
which do not vary by more than ±5 percent
from their mean, and this mean-value is
within ±15 percent of the known value.
BILLING CODE 6560-26-M
C-16
-------
o
i
PRESSURE
CONTROL'
VESSEL
H2S
COMBUST
GAS
TUBE FURNACE
SOzSCRUBBER
£2
g
a.
9.
3 PERCENT HzOz
PUMP.
H
3.
10
Figure 16A-2. Calibration System
3
o
•a
o
-------
Federal Register / Vol. 46. No. 117 / Thursday. June 18. 1981 / Proposed Rules 31909
Conduct this performance check before the
test to validate the test procedure, sampling
system and tester. In addition, field
validation samples shall be taken to monitor
tosses in the probe due to absorption by stack
components. Perform this test by collecting a
known HiS sample (in the applicable
concentration range) after each third field
•ample and before cleaning the probe.
Introduce the gas into the probe and collect
in the usual manner. The obtained
concentration shall be within ±15 percent of
the known value. Otherwise, void the
previous three samples or make corrections
by dividing the sample concentration by the
fraction of recovery if the losses are between
0-20 percent. Substitute a field audit sample
for ono known sample during the collection
period if available. Such audit samples are
usually available from the Quality Assurance
Division, Environmental Monitoring Systems
Laboratory, U.S. Environmental Prbtection
Agency, Research Triangle Park, North
Carolina 27711. ,
6. Calculations—Carry out calculations,
retaining at least one extra decimal figure
beyond that of the acquired data. Round off
figures after final calculation.
8.1 Standard Dry Sample Gas Volume.
Using Equation 6-1 of Method 6. calculate the
dry samplo gas volume V^dm) at standard
conditions'.
0,2 TRS Concentration as SOj. Calculate
tho TRS concentration in ppm as SOa by
using Equation 6-2 of Method 6, except use
KSO*-12.020fi/meq.
7. Bibliography—7.1 Curtis, F. and G.D.
McAlfster. Development and Evaluation of an
Oxidation/Method 6 TRS Emission Sampling
Procedure. Emission Measurement Branch,
Emission Standards and Engineering
Division, OAQPS, Environmental Protection
Agency, Research Triangle Park, North
Carolina 27711. February 1980.
7.2 Blosser, R.O., H.S. Oglesby, and A.K.
Jain. A Study of Alternate SOj Scrubber
Designs Used for TRS Monitoring. A Special
Report by the National Council of the Paper
Industry for Air and Stream Improvement,
Inc., New York. N.Y. July 1977.
7.3 Gellma. I. A Laboratory and Field
Study of Reduced Sulfur Sampling and
Monitoring Systems. Atmospheric Quality
Improvement Technical Bulletin No. 81.
National Council of the Paper Industry for Air
and Stream Improvement, Inc., New York,
N.Y. October 1975.
7.4 Annual Book of ASTM Standards.
Part 31; VViiter. Atmospheric Analysis.
American Society for Testing and Materials.
Philadelphia, Pennsylvania, 1974. pp. 40-42.
[FR Doc, 81-18096 Filed 6-17-81; 8:45 am)
BILLING CODE 6560-26-M
C-18
-------
APPENDIX D
EPA REFERENCE METHOD 6
D-l
-------
-------
Chapter I—Environmental Protection Agency
App. A
METHOD 6—DETERMINATION or SULFUR DIOX-
IDE EMISSIONS FROM STATIONARY SOURCES
1. Principle and Applicability
1.1 Principle. A gas sample is extracted
from the sampling point in the stack. The
sulfuric acid mist (including sulfur trioxide)
and the sulfur dioxide are separated. The
sulfur dioxide fraction is measured by the
barium-thorin titration method.
1.2 Applicability. This method is applica-
ble for the determination of sulfur dioxide
emissions from stationary sources. The
minimum detectable limit of the method
has been determined' to .be 3.4 milligrams
(mg) of SCVm3 (2.12x10-' lb/ft3). Although
no upper limit has been established, tests
have shown that concentrations as high as
80,000 mg/m3 of SO2 can be collected effi-
ciently in two midget impingers, each con-
taining 15 milliliters of 3 percent hydrogen
peroxide, at a'rate of 1.01pm for 20 minutes.
Based on theoretical calculations, the upper
concentration limit in a 20-liter sample is
about 93,300 mg/m,.
Possible interferents are free ammonia,
water-soluble cations, and fluorides. The ca-
tions and fluorides are removed by glass
wool filters and an isopropanol bubbler, and
hence do not affect the SO, analysis. When
samples are being taken from a gas stream
with high concentrations of very find metal-
lic fumes (such as in inlets to control de-
vices), a high-efficiency glass fiber filter
must be used in place of the glass wool plug
(i.e., the one in the probe) to remove the
.cation interferents.
Free ammonia interferes by reacting with
SOa to form particulate sulfite and by react-
ing with the indicator. If free ammonia is
present (this can be determined by knowl-
edge of the process and noticing white par-
ticulate matter in the probe and isopropanol
bubbler), alternative methods, subject to
the approval of the Administrator, U.S. En-
vironmental Protection Agency, are re-
quired.
D-3
-------
App. A
Title 40—Protection of Environment
D-4
-------
Chapter I—Environmental Protection Agency
App. A
2. Apparatus
2.1 Sampling. The sampling train is
shown in Figure 6-1, and component parts
are discussed below. The tester has the
option of substituting sampling equipment
described in Method 8 in place of the
midget impinger equipment of Method 6.
However, the Method 8 train must be modi-
fied to include a heated filter between the
probe and isopropanol impinger, and the op-
eration of the sampling train and sample
analysis must be at the flow rates and solu-
tion volumes defined in Method 8.
The tester also has the option of deter-
mining SOj simultaneously with particulate
matter and moisture determinations by (1)
replacing the water in a Method 5 impinger
system with 3 percent peroxide solution, or
(2) by replacing the Method 5 water im-
pinger system with a Method 8 isopropanol-
filter-peroxide system. The analysis for SO,
must be consistent with the procedure in
Method 8.
2.1.1 Probe. Borosilicate glass, or stain-
less steel (other materials of construction
may be used, subject to the approval of the
Administrator), approximately 6-mm inside
diameter, with a heating system to prevent
water condensation and a filter (either in-
stack or heated outs tack) to remove particu-
late matter, including sulfuric acid mist. A
plug of glass wool is a satisfactory filter.
2.1.2 Bubbler and Impingers. One midget
bubbler, with medium-coarse glass frit and
borosilicate or quartz glass wool packed in
top (see Figure 6-1) to prevent sulfuric acid
mist carryover, and three 30-ml midget im-
pingers. The bubbler and midget impingers
must be connected in series with leak-free
glass connectors, silicone grease may be
used, if necessary, to prevent leakage.
At the option of the tester, a midget im-
pinger may be used in place of the midget
bubbler.
Other collection absorbers and flow rates
may be used, but are subject to the approval
of the Administrator. Also, collection effi-
ciency must be shown to be at least 99 per-
cent for each test run and must be docu-
mented in the report. If the efficiency is
found to be acceptable after a series of
three tests, further documentation is not re-
quired. To conduct the efficiency test, an
extra absorber must be added and analyzed
separately. This extra absorber must not
contain more than 1 percent of the total
SO,.
2.1.3 Glass Wool. Borosilicate or quartz.
2.1.4 Stopcock Grease. Acetone-insoluble,
heatstable silicone grease may be used, if
necessary.
2.1.5 Temperature Gauge. Dial thermom-
eter, or equivalent, to measure temperature
of gas leaving impinger train to within 1° C
(2' F.)
2.1.6 Drying Tube. Tube packed with 6-
to 16-mesh indicating type silica gel, or
equivalent, to dry the gas sample and to
protect the meter and pump. If the silica gel
has been used previously, dry at 175° C (350°
F) for 2 hours. New silica gel may be used as
received. Alternatively, other types of decis-
sants (equivalent or better) may be used,
subject to approval of the Administrator.
2.1.7 Valve. Needle valve, to regulate
sample gas flow rate.
2.1.8 Pump. Leak-free diaphragm pump,
or equivalent, to pull gas through the train.
Install a small surge tank between the
pump and rate meter to eliminate the pulsa-
tion effect of the diaphragm pump on the
rotameter.
2.1.9. Bate-Meter. Rotameter, or equiva-
lent, capable of" measuring flow rate to
within 2 percent of the selected flow rate of
about 1000 cc/min.
2.1.10 Volume Meter. Dry gas meter, suf-
ficiently accurate to measure the sample
volume within 2 percent, calibrated at the
selected flow rate and conditions actually
encountered during sampling, and equipped
with a temperature gauge (dial thermom-
eter, or equivalent) capable of measuring
temperature to within 3° C (5.4° F).
2.1.11 Barometer. Mercury, aneroid, or
other barometer capable of measuring at-
mospheric pressure to within 2.5 mm Hg
(0.1 in. Hg). In many cases, the barometric
reading may be obtained from a nearby na-
tional weather service station, in which case
the station value (which is the absolute
barometric pressure) shall be requested and
an adjustment for elevation differences be-
tween the weather station and sampling
point shall be applied at a rate of minus 2.5
mm Hg (0.1 in. Hg) per 30 m (100 ft) eleva-
tion increase or vice versa for elevation de-
crease.
2.112 Vacuum Gauge and Rotameter. At
least 760 mm Hg (30 in. Hg) gauge and 0-40
cc/min rotameter, to be used for leak check
of the sampling train.
2.2 Sample Recovey.
2.2.1 Wash bottles. Polyethylene or glass,
500'ml, two.
2.2.2 Storage Bottles. Polyethylene, 100
ml, to store impinger samples (one per
sample).
2.3 Analysis.
2.3.1 Pipettes. Volumetric type, 5-ml, 20-
ml (one per sample), and 25-ml sizes.
2.3.2 Volumetric Flasks. 100-ml size (one
per sample) and 1000 ml size.
2.3.3 Burettes. 5- and 50-ml sizes.
2.3.4 Erlenmeyer Flasks. 250 mi-size (one
for each sample, blank, and standard).
2.3.5 Dropping Bottle. 125-ml size, to add
indicator.
2.3.6 Graduated Cylinder. 100-ml size.
2.3.7 Spectrophotometer. To measure ab-
sorbance at 352 nanometers.
3. Reagents
-------
App. A
Title 40—Protection of Environment
Unless otherwise indicated, all reagents
must conform to the specifications estab-
lished by the Committee on Analytical Rea-
gents of the American Chemical Society.
Where such specifications are not available,
use the best available grade.
3.1 Sampling.
3.1.1 Water. Deionized, distilled to con-
form to ASTM specification D1193-74, Type
3. At the option of the analyst, the KMnO.
test for oxidizable organic matter may be
omitted when high concentrations of organ-
ic matter are not expected to be present.
3.1.2 Isopropanol, 80 percent. Mix 80 ml
of Isopropanol with 20 ml of deionized, dis-
tilled water. Check each lot of Isopropanol
for peroxide impurities as follows: shake 10
ml of Isopropanol with 10 ml of freshly pre-
pared 10 percent potassium iodide solution.
Prepare a blank by similarly treating 10 ml
of distilled water. After 1 minute, read the
absorbance at 352 nanometers on a spectro-
photometer. If absorbance exceeds 0.1,
reject alcohol for use.
Peroxides may be removed from Isopro-
panol by redistilling or by passage through
a column of activated alumina; however,
reagent grade Isopropanol with suitably low
peroxide levels may be obtained from com-
mercial sources. Rejection of contaminated
lots may, therefore, be a more efficient pro-
cedure.
3.1.3 Hydrogen Peroxide, 3 Percent.
Dilute 30 percent hydrogen peroxide 1:9 (v/
v) with deionized, distilled water (30 ml is
needed per sample). Prepare fresh daily.
3.1.4 Potassium Iodide Solution, 10-Per-
cent. Dissolve 10.0 grams KI in deionized,
distilled water and dilute to 100 ml. Prepare
when needed.
3.2 Sample Recovery.
3.2.1 Water. Deionized, distilled, as in
3.1.1.
3.2.2 Isopropanol, 80 Percent. Mix 80 ml
of Isopropanol with 20 ml of deionized, dis-
tilled water.
3.3 Analysis.
3.3.1 Water. Deionized, distilled, as in
3.1.1.
3.3.2 Isopropanol, 100 percent.
3.3.3 Thorin Indicator. l-(o-arsonopheny-
lazo)-2-naphthol-3,6-disulfonic acid, diso-
dium salt, or equivalent. Dissolve 0.20 g in
100 ml of deionized, distilled water.
3.3.4 Barium Perchlorate Solution, 0.0100
N. Dissolve 1.95 g of barium perchlorate tri-
hydrate [Ba(ClO«)r3H,O] in 200 ml distilled
water and dilute to 1 liter with isopropanol.
Alternatively, 1.22 g of [BaCl,-2H,O] may be
used instead of the perchlorate. Standardize
as in Section 5.5.
3.3.5 Sulfuric Acid Standard, 0.0100 N.
Purchase or standardize to ±0.0002 N
against 0.0100 N NaOH which has previous-
ly been standardized against potassium acid
phthalate (primary standard grade).
4. Procedure.
4.1 Sampling.
4.1.1 Preparation of collection train.
Measure 15 ml of 80 percent Isopropanol
into the midget bubbler and 15 ml of 3 per-
cent hydrogen peroxide into each of the
first two midget impingers. Leave the final
midget impinger dry. Assemble the train as
shown in Figure 6-1. Adjust probe heater to
a temperature sufficient to prevent water
condensation. Place crushed ice and water
around the impingers.
4.1.2 Leak-check procedure. A leak check
prior to the sampling run is optional; how-
ever, a leak check after the sampling run is
mandatory. The leak-check procedure is as
follows:
Temporarily attach a suitable (e.g., 0-40
cc/min) rotameter to the outlet of .the dry
gas meter and place a vacuum gauge at or
near the probe inlet. Plug the probe inlet,
pull a vaccum of at least 250 mm Hg (10 in.
Hg), and note the flow rate as indicated by
the rotameter. A leakage rate not in excess
of 2 percent of the average sampling rate is
acceptable.
NOTE: Carefully release the probe inlet
plug before turning off the pump.
It is suggested (not mandatory) that the
pump be leak-checked separately, either
prior to or after the sampling run. If done
prior to the sampling run, the pump leak-
check shall precede the leak check of the
sampling train described immediately above;
if done after the sampling run, the pump
leak-check shall follow the train leak-check.
To leak check the pump, proceed as follows:
Disconnect the drying tube from the probe-
impinger assembly. Place a vacuum gauge at
the inlet to either the trying tube or the
pump, pull a vacuum of 250 mm (10 in.) Hg,
plug or pinch off the outlet of the flow
meter and then turn off the pump. The
vacuum should remain stable for at least 30
seconds.
Other leak-check procedures may be used,
subject to the approval of the Adminstrator,
U.S. Environmental Protection Agency.
4.1.3 Sample collection. Record the ini-
tial dry gas meter reading and barometric
pressure. To begin sampling, position the tip
of the probe at the sampling point, connect '
the probe to the bubbler, and start the
pump. Adjust the sample flow to a constant
rate of approximately 1.0 liter/min as indi-
cated by the rotameter. Maintain this con-
stant rate (±10 percent) during the entire
sampling run. Take readings (dry gas meter,
tempertures at dry gas meter and at im-
pinger outlet and rate meter) at least every
5 minutes. Add more ice during the run to
keep the temperture of the gases leaving
the last impinger at 20° C (68° F) or less. At
the conclusion of each run, turn off the
pump, remove probe from the stack, and
record the final readings. Conduct a leak
D-6
-------
Chapter I—Environmental Protection Agency
App. A
check as in Section 4.1.2 (This leak check is
mandatory.) If a leak is found, void the test
run, or use procedures acceptable to the Ad-
ministrator.to adjust the sample volume for
the leakage. Drain the ice bath, and purge
the remaining part of the train by drawing
clean ambient air through the system for 15
minutes at the sampling rate.
Clean ambient air can be provided by
passing air through a charcoal filter or
through an extra midget impinger with 15
ml of 3 percent H,Oj. The tester may opt to
simply use ambient air, without purifica-
tion.
4.2 Sample Recovery. Disconnect the im-
pingers after purging. Discard the contents
of the midget bubbler. Pour the contents of
the midget impingers into a leak-free poly-
ethylene bottle for shipment. Rinse the
three midget impingers and the connecting
tubes with deionized, distilled water, and
add the washings to the same storage con-
tainer. Mark the fluid level. Seal and identi-
fy-the sample container.
4.3 Sample Analysis. Note level of liquid
in container, and confirm whether any
sample was lost during shipment; note this
on analytical data sheet. If a noticeable
amount of leakage has occurred, either void
the sample or use methods, subject to the
approval of the Administrator, to correct
the final results.
Transfer the contents of the storage con-
tainer to a 100-ml volumetric flask and
dilute to exactly 100 ml with deionized, dis-
tilled water. Pipette a 20-ml aliquot of this
solution into a 250-ml Erlenmeyer flask, add
80 ml of 100 percent isopropanol and two to
four drops of thorin indicator, and titrate to
a pink endpoint using 0.0100 N barium
perchlorate. Repeat and average the titra-
tion volumes. Run a blank with each series
of samples. Replicate Mirations must agree
within 1 percent or 0.2 ml, whichever is
larger.
NOTE: Protect the 0.0100 N barium perch-
lorate solution from evaporation at all
times.
5. Calibration
5.1 Metering System.
5.1.1 Initial Calibration. Before its initial
use in the field, first leak check the meter-
ing system (drying tube, needle valve, pump,
rotameter, and dry gas meter) as follows:
place a vacuum gauge at the inlet to the
drying tube and pull a vaccum of 250 mm
(10 in.) Hg; plug or pinch off the outlet of
the flow meter, and then turn off the pump.
The vaccum shall remain stable for at least
30 seconds. Carefully release the vaccum
gauge before releasing the flow meter end.
Next, calibrate the metering system (at
the sampling flow rate specified by the
method) as follows: connect an appropriate-
ly sized wet test meter (e.g., 1 liter per revo-
lution) to the inlet of the drying tube. Make
three independent calibration runs, using at
least five revolutions of the dry gas meter
per run. Calculate the calibration factor, Y
(wet test meter calibration volume divided
by the dry gas meter volume, both volumes
adjusted to the same reference temperature
and pressure), for each run, and average the
results. If any Y value deviates by more
than 2 percent from the average, the meter-
ing system is unacceptable for use. Other-
wise, use the average as the calibration
factor for subsequent test runs.
5.1.2 Post-Test Calibration Check. After
each field test series, conduct a calibration
check as in Section 5.1.1 above, except for
the following variations: (a) the leak check
is not to be conducted, (b) three, or more
revolutions of the dry gas meter may be
used, and (c) only two independent runs
need be.made. If the calibration factor does
not deviate by more than 5 percent from
the initial calibration factor (determined in
Section 5.1.1), then the dry gas meter vol-
umes obtained during the test series are ac-
ceptable. If the calibration factor deviates
by more than 5 percent, recalibrate the me-
tering system as in Section 5.1.1, and for the
calculations, use the calibration factor (ini-
tial or recalibration) that yields the lower
gas volume for each test run.
5.2 Thermometers. Calibrate against
mercury-in-glass thermometers.
5.3 Rotameter. The rotameter need not
be calibrated but should be cleaned and
maintained according to the manufactu-
turer's instruction.
5.4 Barometer. Calibrate against a mer-
cury barometer.
5.5 Barium Perchlorate Solution. Stand-
ardize the barium perchlorate solution
against 25 ml of standard sulfuric acid to
which 100 ml of 100 percent isopropanol has
been added.
6. Calculations
Carry out calculations, retaining at least
one extra decimal figure beyond that of the
acquired data. Round off figures after final
calculation.
6.1 Nomenclature.
CK,=Concentration of sulfur dioxide, dry
basis corrected to standard conditions,
mg/dscm (Ib/dscf).
N=Normality of barium perchlorate titrant,
milliequivalents/ml. -
P,»r=Barometric pressure at the exit orifice
of the dry gas meter, mm Hg (in. Hg).
POA=Standard absolute pressure, 760 mm
Hg (29.92 in. Hg).
Tm= Average dry gas meter absolute tem-
perature, °K (°R).
T,M=Standard absolute temperature, 293° K
(528° R).
Vc—Volume of sample aliquot titrated, ml.
Vm=Dry gas volume as measured by the dry
gas meter, dcm (dcf).
D-7
-------
App. A
V»c,u)-Dry gas volume measured by the dry
gas meter, corrected to standard condi-
tions, dscm (dscf).
Vwta-Total volume of solution in which the
sulfur-dioxide sample is contained, 100
ml.
Vi— Volume of barium perchlorate titrant
used for the sample, ml (average or rep-
licate titrations).
V<>*= Volume of barium perchlorate titrant
used for the blank, ml.
y=Dry gas meter calibration factor.
32.03 ^Equivalent weight of sulfur dioxide.
6.2 . Dry sample gas volume, corrected to
standard conditions.
r,eJ\ / °bmA jr y '•m °hmr
~Tl) \J\TJ~K*Y Tm
Equation 6-1
Title 40—Protection of Environment
4. Patton, W. F. and J. A. Brink, Jr. New
Equipment and Techniques for Sampling
Chemical Process Oases. J. Air Pollution
Control Association. 13:162.1963.
5. Rom, J. J. Maintenance, Calibration,
and Operation of Isokinetic Source-sam-
pling Equipment. Office of Air Programs,
Environmental Protection Agency. Re-
search Triangle Park, N.C. APTD-0576.
March 1972.
6. Hamil, H. F. and D. E. Camann. Col-
laborative Study of Method for the Deter-
mination of Sulfur Dioxide Emissions from
Stationary Sources (Fossil-Fuel Fired Steam
Generators). Environmental Protection
Agency, Research Triangle Park, N.C. EPA-
650/4-74-024. December 1973.
7. Annual Book of ASTM Standards. Part
31; Water, Atmospheric Analysis. American
Society for Testing and Materials. Philadel-
phia, Pa. 1974. pp. 40-42.
8. Knoll, J. E. and M. R. Midgett. The Ap-
plication of EPA Method 6 to High Sulfur
Dioxide Concentrations. Environmental
Protection Agency. Research Triangle Park,
N.C. EPA-600/4-76-038. July 1976.
where:
.Ki—0.3858' K/mm Hg for metric units.
-17.64' R/in. Hg for English units.
6.3 Sulfur dioxide concentration.
Equation 6-2
where:
.Ki-32.03 mg/meq. for metric units.
=7.061 x!0"slb/meq. for English units.
7. Bibliography
1. Atmospheric Emissions from Sulfuric
Acid Manufacturing Processes. U.S. DHEW,
PHS, Division of Air Pollution. Public
Health Service Publication No. 999-AP-13.
Cincinnati, Ohio. 1965.
2. Corbett, P. F. The Determination of
SO, and SO, in Flue Gases. Journal of the
Institute of Fuel. 24:237-243,1961.
3. Matty, R. E. and E. K. Diehl. Measuring
Flue-Gas SOi and SOS. Power. 101: 94-97.
November 1957.
'Mention of trade names or specific prod-
ucts does not constltue endorsement by the
Environmental Protection Agency.
D-8
-------
APPENDIX E
EPA REFERENCE METHOD 9
E-l
-------
-------
App. A
METHOD 8—VISUAL DETERMINATION OF THE
OPACITY OF EMISSIONS PROM STATIONARY
SOURCES
Many stationary sources discharge visible
emissions into the atmosphere; these emis-
sions are usually in the shape of a plume.
This method involves the determination of
plume opacity by Qualified observers. The
Title 40—Protection of Environment
method includes procedures for the training
and certification of observers, and proce-
dures to be used in the field for determina-
tion of plume opacity. The appearance of a
plume as viewed by an observer depends
upon a number of variables, some of which
may be controllable and some of which may
not be controllable in the field. Variables
which can be controlled to an extent to
which they no longer exert a significant in-
fluence upon plume appearance include:
Angle of the observer with respect to the
plume; angle of the observer with respect to
the sun; point of observation of attached
and detached steam plume; and angle of the
observer with respect to a plume emitted
from a rectangular stack with a large length
to width ratio. The method includes specific
criteria applicable to these variables.
Other variables which may not be control-
lable in the field are luminescence and color
contrast between the plume and the back-
ground against which the plume is viewed.
These variables exert an influence upon the
appearance of a plume as viewed by an ob-
server, and can affect the ability of the ob-
server to accurately assign opacity values to
the observed plume. Studies of the theory
of plume opacity and field studies have
demonstrated that a plume is most visible
and presents the greatest apparent opacity
when viewed against a contrasting back-
ground. It follows from this, and is con-
firmed by field trials, that the opacity of a
plume, viewed under conditions where a
contrasting background is present can be as-
signed with the greatest degree of accuracy.
However, the potential for a positive error is
also the greatest when a plume is viewed
under such contrasting conditions. Under
conditions presenting a less contrasting
background, the apparent opacity of a
plume is less and approaches zero as the
color and luminescence contrast decrease
toward zero. As a result, significant negative
bias and negative errors can be made when
a plume is viewed under less contrasting
conditions. A negative bias decreases rather
than increases the possibility that a plant
operator will be cited for a violation of opac-
ity standards due to observer error.
Studies have been undertaken to deter-
mine the magnitude of positive errors which
can be made by qualified observers while
reading plumes under contrasting condi-
tions and using the procedures set forth in
this method. The results of these studies
(field trials) which involve a total of 769 sets
of 25 readings each are as follows:
(1) For black plumes (133 sets at a smoke
generator), 100 percent of the sets were read
with a positive error* of less than 7.5 per-
•Por a set, positive error = average opac-
ity determined by observers' 25 observa-
Continued
E-3
-------
Chapter I—Environmental Protection Agency
App. A
cent opacity; 99 percent were read with a
positive error of less than 5 percent opacity.
(2) For white plumes (170 sets at a smoke
generator, 168 sets at a coal-fired power
plant, 298 sets at a sulfuric acid plant), 99
percent of the sets were read with a positive
error of less than 7.5 percent opacity; 95
percent were read with a positive error of
less than 5 percent opacity.
The positive observational error associat-
ed with an average of twenty-five readings is
therefore established. The accuracy of the
method must be taken into account when
determining possible violations of applicable
opacity standards.
1. Principle and applicability.
1.1 Principle. The opacity of emissions
from stationary sources is determined visu-
ally by a qualified observer.
1.2 Applicability. This method is applica-
ble for the determination of the opacity of
emissions from stationary sources pursuant
to § 60.1Kb) and for qualifying observers for
visually determining opacity of emissions.
2. Procedures. The observer qualified in
accordance with paragraph 3 of this method
shall use the following procedures for visu-
ally determining the opacity of emissions:
2.1 Position. The qualified observer shall
stand at a distance sufficient to provide a
dear view of the emissions with the sun ori-
ented In the 140° sector to his back. Consist-
ent with maintaining the above require-
ment, the observer shall, as much as possi-
ble, make his observations from a position
such that his line of vision is approximately
perpendicular to the plume direction, and
when observing opacity of emissions from
rectangular outlets (e.g. roof monitors, open
baghouses, noncircular stacks), approxi-
mately perpendicular to the longer axis of
the outlet. The observer's line of sight
should not Include more than one plume at
a time when multiple stacks are involved,
and in any case the observer should make
his observations with his line of sight per-
pendicular to the longer axis of such a set of
multiple stacks (e.g. stub stacks on bag-
houses).
2.2 Field records. The observer shall
record the name of the plant, emission loca-
tion, type facility, observer's name and af-
filiation, and the date on a field data sheet
(Figure 9-1). The time, estimated distance
to the emission location, approximate wind
direction, estimated wind speed, description
of the sky condition (presence and color of
clouds), and plume background are recorded
on a field data sheet at the time opacity
readings are initiated and completed.
2.3 Observations. Opacity observations
shall be made at the point of greatest opac-
ity in that portion of the plume where con-
tlons—average opacity determined from
transmlssometer's 25 recordings.
densed water vapor is not present. The ob-
server shall not look continuously at the
plume, but instead shall observe the plume
momentarily at 15-second intervals.
2.3.1 Attached steam plumes. When con-
densed water vapor is present within the
plume as it emerges from the emission
outlet, opacity observations shall be made
beyond the point in the plume at which con-
densed water vapor is no longer visible. The
observer shall record the approximate dis-
tance from the emission outlet to the point
in the plume at which the observations are
made.
2.3.2 Detached steam plume. When water
vapor in the plume condenses and becomes
visible at a distinct distance from the emis-
sion outlet, the opacity of emissions should
be evaluated at the emission outlet prior to
the condensation of water vapor and the
formation of the steam plume.
2.4 Recording observations. Opacity obser-
vations shall be recorded to the nearest 5
percent at 15-second intervals on an obser-
vational record sheet. (See Figure 9-2 for an
example.) A minimum of 24 observations
shall be recorded. Each momentary observa-
tion recorded shall be deemed to represent
the average opacity of emissions for a 15-
second period.
2.5 Data Reduction. Opacity shall be de-
termined as an average of 24 consecutive ob-
servations recorded at 15-second intervals.
Divide the observations recorded on the
record sheet into sets of 24 consecutive ob-
servations. A set is composed of any 24 con-
secutive observations. Sets need not be con-
secutive in time and in no case shall two sets
overlap. For each set of 24 observations, cal-
culate the average by summing the opacity
of the 24 observations and dividing this sum
by 24. If an applicable standard specifies an
averaging time requiring more than 24 ob-
servations, calculate the average for all ob-
servations made during the specified time
period. Record the average opacity on a
record sheet. (See Figure 9-1 for an exam-
ple.)
3. Qualifications and testing.
3.1 Certification requirements. To receive
certification as a- qualified observer, a candi-
date must be tested and demonstrate the
ability to assign opacity readings in 5 per-
cent increments to 25 different black
plumes and 25 different white plumes, with
an error not to exceed 15 percent opacity on
any one reading and an average error not to
exceed 7.5 percent opacity in each category.
Candidates shall be tested according to the
procedures described in paragraph 3.2.
Smoke generators used pursuant to para-
graph 3.2 shall be equipped with a smoke
meter which meets the requirements of
paragraph 3.3.
The certification shall be valid for a
period of 6 months, at which time the quali-
E-4
-------
App.*
Title 40—'Protection of Environment
flcation procedure must be repeated by any
observer in order to retain certification.
3.2 Certification procedure. The certifica-
tion test consists of showing the candidate a
complete run of SO plumes—25 black plumes
and 25 white plumes—generated by a smoke
generator. Plumes within each set of 25
black and 25 white runs shall be presented
in random order. The candidate assigns an
opacity value to each plume and records his
observation on a suitable form. At the com-
pletion of each run of 50 readings, the score
of the candidate is determined. If a candi-
date fails to qualify, the complete run of 50
readings must be repeated in any retest.
The smoke test may be administered as part
of a smoke school or training program, and
may be preceded by training or familiariza-
tion runs of the smoke generator during
which candidates are shown black and white
plumes of known opacity.
3.3 Smoke generator specifications. Any
smoke generator used for the purposes of
paragraph 3.2 shall be equipped with a
smoke meter installed to measure opacity
across the diameter of the smoke generator
stack. The smoke meter output shall display
iiistack opacity based upon a pathlength
equal to the stack exit diameter, on a full 0
to 100 percent chart recorder scale. The
smoke meter optical design and perform-
ance shall meet the specifications shown in
Table 9-1. The smoke meter shall be cali-
brated as prescribed in paragraph 3.3.1 prior
to the conduct of each smoke reading test.
At the completion of each test, the zero and
span drift shall be checked and if the drift
exceeds ±1 percent opacity,,the condition
shall be corrected prior to conducting any
subsequent test runs. The smoke meter
shall be demonstrated, at the time of instal-
lation, to meet the specifications listed in
Table 9-1. This demonstration shall be re-
peated following any subsequent repair or
replacement of the photocell or associated
electronic circuitry including the chart re-
corder or output meter, or every 6 months,
whichever occurs first.
TABLE 9-1—SMOKE METER DESIGN AND
PERFORMANCE SPECIFICATIONS
Parameter
a. Light source..
b. Spectral response of
photocell.
c. Angle oi view
d. Angle of projection .:
e. Calibration error ...
f. Zero and span
drift±1% opacity, 30
minutes.
g. Response time
Specification
Incandescent lamp operated «l
nominal rated voltage.
Photopic (daylight spectral re-
sponse of the human eye-
reference 4.3).
15* maximum total angle.
15* maximum total angle.
±3% opacity, maximum.
5 seconds.
3.3.1 Calibration. The smoke meter is cali-
brated after allowing a minimum of 30 min-
utes warmup by alternately producing simu-
lated opacity of 0 percent and 100 percent.
When stable response at 0 percent or 100
percent is noted, the smoke meter is adjust-
ed to produce an output of 0 percent or 100
percent, as appropriate. This calibration
shall be repeated until stable 0 percent and
100 percent readings are produced without
adjustment. Simulated 0 percent and 100
percent opacity values may be produced by
alternately switching the power to the light
source on and off while the smoke generator
is not producing smoke.
3.3.2 Smoke meter evaluation. The smoke
meter design and performance are to be
evaluated as follows:
3.3.2.1 Light source. Verify from manufac-
turer's data and from voltage measurements
made at the lamp, as installed, that the
lamp is operated within ±5 percent of the
nominal rated voltage.
3.3.2.2 Spectral response of photocell.
Verify from manufacturer's data that the
photocell has'a photopic response; i.e., the
spectral sensitivity of the cell shall closely
approximate the standard spectral-luminos-
ity curve for photopic vision which is refer-
enced in (b) of Table 9-1.
E-5
-------
Chapter I—Environmental Protection Agency
App. A
1
C£
UJ
UJ
h~
i
i«
S 2 «
fe
i
g
ggii
«-• _j trt (_?
}_•-*(/)}
ex- u. »-«•"-•
uj u. 2: o
C* -i
CO O CD
O O O
£
g; C5 in P s: =
S S " a ? S
A
*
S
i i!
* iS
I II
Z •>£
S SI
fC H-<
£»••
*
s g«
S-
«
'
•- P»J 5
4* C VI V)
O HJ •*" **~ W ^^
do §
21
E-6
-------
App. A
COMPANY
LOCATION
TEST NUMBET
DATE
Title 40—Protection of Environment
FIGURE 9-2 OBSERVATION RECORD TAGE OF
OBSERVER
TYPE FACILITY
POINT OF EHISSIOUT
Hr.
M1n.
0
1
?
T
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
1<>
20
21
??
23
24
25
26
27
?R
29
Seconds
IT
IB
:•)<)
j
45
StEAM PLUME
(check 1f applicable)
Attached
' ' :
Detached
COMMENTS
' »
3.3.2.3 Angle of view. Check construction
geometry to ensure that the total angle of
view pf the smoke plume, as seen by the
photocell, does not exceed 15°. The total
angle of view may be calculated:from: =2
tan"'d/2L, where 6=total angle of view;
d=the sum of the photocell diameter+the
diameter of the limiting aperture; and
L=the distance from the photocell to the
limiting aperture. The limiting aperture is
the point in the path between the photocell
and the smoke plume where the angle of
view is most restricted. In smoke generator
smoke meters this is normally an orifice
plate.
3.3.2.4 Angle of projection. Check con-
struction geometry to ensure that the total
angle of projection of the lamp on the
smoke plume does not exceed 15'. The total
angle of projection may be calculated from:
E-7
-------
Chapter I—Environmental Protection Agency
App.A
«»2 tan" 'd/2L, where 0= total angle of pro-
jection; d= the sum of the length of the
limp filament + the diameter of the limit-
Ing aperture; and L= the distance from the
lamp to the limiting aperture.
FIGURE 9-2 OBSERVATION RECORD
(Continued)
COMPANY
LOCATION •
TEST NUMBlT
MTE
PAGE
OF
OBSERVER .
TYPE FACILITY
POINT OF EMISSTOTJT
Hr.
M1n.
30
31
32
K
34
3S
36
'"'37
38
39
4rt
41
*5~
44
4S'
46
47
'48
49
SO
•M
"S?
^3
fi4
™«tt
56
«S7
«iR
<>9
Seconds
ff
1£>
30
4b
STEAM PLUME
(check 1f applicable)
Attached
Detached
COWIENTS
.
3.3.2.5 Calibration error. Using neutral-
density filters of known opacity, check the
error between the actual response and the
theoretical linear response of the smoke
meter. This check Is accomplished by first
calibrating the smoke meter according to
3.3.1 and then inserting a series of three
neutral-density filters of nominal opacity of
20, 50, and 75 percent In the smoke meter
pathlength. Filters calibrated within ±2
percent shall be used. Care should be taken
when Inserting the filters to prevent stray
E-8 •
-------
App.A
Title 40—Protection of Environment
light from affecting the meter. Make a total
of five nonconsecutive readings for each
filter. The maximum error on any one read-
ing shall be 3 percent opacity.
3.3.2.6 Zero and span drift. Determine the
zero and span drift by calibrating and oper-
ating the smoke generator in a normal
manner over a 1-hour period. The drift is
measured by checking the zero and span at
the end of this period.
3.3.2.7 Response time. Determine the re-
sponse time by producing the series of five
.simulated 0 percent and 100 percent opacity
values and observing the time required to
reach stable response. Opacity values of 0
percent and 100 percent may be simulated
by alternately switching the power to the
light source off and on while the smoke gen-
erator is not operating.
4. References.
4.1 Air Pollution Control District Rules
and Regulations, Los Angeles County Air
Pollution Control District, Regulation IV,
Prohibitions, Rule 50.
4.2 Weisburd, Melvin I., Field Operations
and Enforcement Manual for Air, U.S. Envi-
ronmental Protection Agency, Research Tri-
angle Park, N.C., APTD-1100, August 1972,
pp. 4.1-4.36.
4.3 Condon, E.U., and Odishaw, H., Hand-
book of Physics, McGraw-Hill Co., N.Y.,
N.Y., 1958, Table 3.1, p. 6-52.
E-9
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-340/1-83-Q17
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Kraft Pulp Mill Inspection Guide
5. REPORT DATE
February 1983 (preparation)
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Ronald Hawks/Gary Saunders
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
PEDCo Environmental, Inc. .
505 S. Duke St., Suite 503
Durham, North Carolina 27701
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-01-6310'
Task No. 65
12. SPONSORING AGENCY NAME AND ADDRESS
U.S. Environmental Protection Agency
Stationary Source Compliance Division
Jfeshinoton. D.C. 20460
13. TYPE OF REPORT AND PERIOD COVERED
Final report
14. SPONSORING AGENCY CODE
18. SUPPLEMENTARY NOTES
EPA Project Officer for this report was Robert Marshall, telephone: (202) 382-2862
10, ABSTRACT
This manual presents technical data on kraft pulp mill processes and control
equipment design and application. The manual also includes inspection checklists
for use by agency personnel in evaluating process parameters and control equipment
operating conditions. Major emphasis is placed on baseline analyses and detection
and elimination of operation- and maintenance-related problems.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Operation and maintenance
Kraft pulp mills
Particulate
TR 5
Recovery boilers
Lime kilns
10. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (TillsReport)
Unclassified
21. NO. OF PAGES
405
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Fotm 2220-1 (R»v. 4-77) PREVIOUS EDITION is OBSOLETE
-------
-------
-------
-------
United States
Environmental Protection
Agency
Office of Air Quality Planning and Standards
Stationary Source Compliance Division
Washington D.C. 20460
Official Business
Penalty for Private Use -
$300
Publication No. EPA- 340/1-83-017
Postage and
Fees Paid
Environmental
Protection
Agency
EPA 335
If your address is incorrocl. please change on (he above label;
tear off; and return to the abova address
If you do not dosiro toconttnua receiving this technical report
series, CHECK HERE D; tsar olf label; and return it to the
above address.
------- |