U.S. Environmental Protection Agency Industrial Environmental Research EPA-600/7-76-017
Office of Research and Development Laboratory
Research Triangle Park, North Carolina 27711 Q CtOb6T 1976
PRELIMINARY ENVIRONMENTAL
ASSESSMENT OF THE CAFB
Interagency
Energy-Environment
Research and Development
Program Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S.
Environmental Protection Agency, have been grouped into seven series.
These seven broad categories were established to facilitate further
development and application of environmental technology. Elimination
of traditional grouping was consciously planned to foster technology
transfer and a maximum interface in related fields. The seven series
are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from
the effort funded under the 17-agency Federal Energy/Environment
Research and Development Program. These studies relate to EPA's
mission to protect the public health and welfare from adverse effects
of pollutants associated with energy systems. The goal of the Program
is to assure the rapid development of domestic energy supplies in an
environmentally—compatible manner by providing the necessary
environmental data and control technology. Investigations include
analyses of the transport of energy-related pollutants and their health
and ecological effects; assessments of, and development of, control
technologies for energy systems; and integrated assessments of a wide
range of energy-related environmental issues.
This document is available to the public through the National Technical
Information Service, Springfield, Virginia 22161.
-------
EPA-600/7-76-017
October 1976
PRELIMINARY
ENVIRONMENTAL ASSESSMENT
OF THE CAFB
by
Arthurs. Werner, Charles W. Young, Mark I. Bornstein,
Robert M. Bradway, Michael T. Mills, and Donald F. Durocher
GCA Corporation
GCA/Technology Division
Bedford, Massachusetts 01730
Contract No. 68-02-1316, Task 14
Program Element No. EHB537
EPA Task Officer: Samuel L. Rakes
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
ABSTRACT
This document presents the results of a preliminary environmental assess-
ment of the Chemically Active Fluid Bed (CAFB) process. All waste streams
contributing air, water and solid waste pollutants were evaluated in terms
of emission rates and potential environmental effects. As part of this
investigation, a field sampling and laboratory analysis program was car-
ried out to compile an emissions inventory of the CAFB pilot plant at the
Esso Research Centre, Abingdon (ERCA), England. In addition to the en-
vironmental assessment, an economic evaluation of the CAFB relative to
alternative residual oil utilization techniques is presented. Finally,
recommendations are made for further control research and development to
be carried out at the CAFB demonstration plant in San Benito, Texas.
iii
-------
CONTENTS
Page
Abstract iii
List of Figures ix
List of Tables xvi
Acknowledgments xxi
Sections
I Executive Summary . 1
Overview 1
Conclusions 6
References 10
II Introduction 11
The Chemically-Active Fluid Bed Process 11
Program Objectives 12
Report Organization 12
References 15
III Process Description 16
Introduction 16
Overview 16
Fuel Feed System 27
Limestone Handling System 28
iv
-------
CONTENTS (continued)
Sections Page
Gasifier 29
Regenerator 33
Spent Solids Handling System 36
TM
FW Resox Off Gas Treatment System 36
Boiler 37
References 39
IV Emissions Estimates 40
Introduction 40
Input Materials 41
Fugitive Air Emissions From Oil Storage and Handling 46
TM
Fugitive Air Emissions From Resox Coal Storage and
Handling 48
Fugitive Air Emissions From Limestone Storage and
Handling 48
Trace Element Emissions 50
TM
Water Emissions From Resox Coal Storage 53
TM
Emissions From Resox Solid Waste 55
Emissions Associated With Spent Regenerator Stone 56
Emissions and Environmental Effects of Condenser Cooling 57
Emissions From Boiler Water Treatment and Boiler
Slowdown 66
References 69
V Field Test Program and Laboratory Results 72
Introduction 72
Field Sampling Protocol 73
v
-------
CONTENTS (continued)
Sections Page
Field Analysis 79
Laboratory Analyses 79
Field Test Program 90
Field Test Results 93
Laboratory Results 101
Summary 172
References 173
VI CAFB Air Quality Impact Assessment For The La Palma Retrofit 174
Introduction 174
Variables Affecting Ambient Concentrations 175
Dispersion Modeling Analysis 179
References 204
Appendixes
A Process Description and Emissions Estimates for the
Coal-Fired CAFB 205
Process Description: 10 MW Demonstration Plant 205
Emissions Estimates: 10 MW Demonstration Plant 208
Process Description: 250 MW Coal-Fired CAFB 209
Emissions Estimates: 250 MW CAFB 214
References 220
B Comparison of the CAFB With Other Residual Oil Utilization
Techniques 221
Introduction 221
Residual Desulfurization 222
vi
-------
CONTENTS (continued)
Appendixes Page
Environmental Impacts of Desulfurization Techniques 231
References 235
C Process Descriptions and Flow Diagrams of Residual Oil
Desulfurization Techniques 236
Flexicoking 236
Gulf HDS 240
RCD Isomax 243
Residue Desulfurization (BP Process) 246
Resid Hydroprocessing (Standard Oil Co. Indiana) 250
LC-Fining 254
i
Resid Ultrafining 255
Go-Fining (Exxon Research and Engineering Co.) 259
Residfining (Esso Research and Refining Company) 263
Residue Hydrodesulfurization > 265
Hydrodesulfurization, Trickle 268
IFF Resid and VGO Hydrodesulfurization 271
Demetalization/Desulfurization 275
Delayed Coking 276
VGO/VRDS Isomax 279
Shell Gasification Process 284
References 289
D Economics and Process Parameters of Alternative Residual
Oil Utilization Technologies 291
Introduction 291
vii
-------
CONTENTS (continued)
Appendixes Page
Process Parameters of Feedstock Desulfurization 291
Economics of Feedstock Desulfurization Processes 291
Economic Comparison of FGD, HDS and CAFB 295
References 301
viii
-------
LIST OF FIGURES
No.
b
1 Generalized Schematic of the CAFB 2
2 Unit Operations Flow Diagram of the ERCA Pilot Plant 18
3 Unit Operations Flow Diagram of the FW Demonstration Plant 20
4 Unit Operations Flow Diagram of the FW 250 MW Plant 23
5 Gasifier-Regenerator Schematic 30
6 Limestone Feed. Broadband ESCA Scan 45
7 Aerial Photograph of the La Palma Power Station 58
8 Fixed Bed Ion Exchange System 67
9 CAFB Pilot Plant ' 74
10 Pilot Plant Stack 75
11 Stack Sampling Ports 77
12 Adsorbent Sampling System 78
13 Hi-Vol Filter. Broadband ESCA Scan 85
14 Aluminum Substrate. Broadband ESCA Scan
15 Vanadium Metal ESCA Scan 87
16 Vanadium Pentoxide ESCA Scan 88
17 Carbon Is Binding Energies 89
18 Stack Particulate Size Distribution, .Run No. 3. Fuel Oil
Gasification 102
19 Stack Particulate Size Distribution, Run No. 5. Bitumen
Gasification 103
20 Log-Normal Particulate Size Distributions 104
21 Bitumen, LC Fraction 1 IR Spectrum 108
22 Bitumen, LC Fraction 2 IR Spectrum 108
ix
-------
LIST OF FIGURES (continued)
No.' . Page
23 Bitumen, LC Fraction 3 IR Spectrum 109
24 Bitumen, LC Fraction 4 IR Spectrum 109
25 Bitumen, LC Fraction 5 IR Spectrum 110
26 Bitumen, LC Fraction 6 IR Spectrum 110
27 Bitumen, LC Fraction 7 IR Spectrum 111
28 Bitumen, LC Fraction 8 IR Spectrum 111
29 Flue Gas From Bitumen Gasification, Run No. 7, LC
Fraction 1 IR Spectrum 115
30 Flue Gas From Bitumen Gasification, Run No. 7, LC
Fraction 3 IR Spectrum 115
31 Flue Gas From Bitumen Gasification, Run No. 7, LC
Fraction 4 IR Spectrum 116
32 Flue Gas From Bitumen Gasification, Run No. 7, LC
Fraction 5 IR Spectrum 116
33 Flue Gas From Bitumen Gasification, Run No. 7, LC
Fraction 6 IR Spectrum 117
34 Flue Gas From Bitumen Gasification, Run No. 7, LC
Fraction 8 IR Spectrum 117
35 Stack Cyclone Material From Bitumen Gasification, Run No. 5,
LC Fraction 1 IR Spectrum 119
36 Stack Cyclone Material From Bitumen Gasification, Run No. 5,
LC Fraction 2 IR Spectrum 119
37 Stack Cyclone Material From Bitumen Gasification, Run No. 5,
LC Fraction 3 IR Spectrum 120
38 Stack Cyclone Material From Bitumen Gasification, Run No. 5,
LC Fraction 4 IR Spectrum 121
39 Stack Cyclone Material From Bitumen Gasification, Run No. 5,
LC Fraction 5 IR Spectrum 122
-------
LIST OF FIGURES (continued)
No. Page
40 Stack Cyclone Material From Bitumen Gasification,
Run No. 5, LC Fraction 6. IR Spectrum 122
41 Stack Cyclone Material From Bitumen Gasification,
Run No. 5. .Broadband ESCA Scan 128
42 Stack Sampling Train Filter Material From Bitumen
Gasification, Run No. 5. Broadband ESCA Scan 129
43 Stack Cyclone Material From Bitumen Gasification,
Run No. 5. Vanadium ESCA Scan 132
44 Stack Sampling Train Filter Material From Bitumen
Gasification, Run No. 5. Vanadium ESCA Scan 133
45 Stack Cyclone Material From Bitumen Gasification,
Run No. 5. Sulfur ESCA Scan 135
46 Stack Sampling Train Filter Material From Bitumen
Gasification, Run No. 5. Sulfur ESCA Scan 136
47 Regenerator Bed Material From Bitumen Gasification,
Run No. 5, LC Fraction 1. IR Spectrum 138
48 Regenerator Bed Material From Bitumen Gasification,
Run No. 5, LC Fraction 2. IR Spectrum 138
49 Regenerator Bed Material From Bitumen Gasification,
Run No. 5, LC Fraction 3. IR Spectrum 139
50 Regenerator Bed Material From Bitumen Gasification,
Run No. 5, LC Fraction 4. IR Spectrum 139
51 Regenerator Bed Material From Bitumen Gasification,
Run No. 5, LC Fraction 5. IR Spectrum 140
52 Regenerator Bed Material From Bitumen Gasification,
Run No. 5, LC Fraction 6. IR Spectrum 140
53 Regenerator Bed Material From Bitumen Gasification,
Run No. 5, LC Fraction 7. IR Spectrum 141
54 Regenerator Bed Material From Bitumen Gasification,
Run No. 5, LC Fraction 8. IR Spectrum 141
xi
-------
LIST OF FIGURES (continued)
No.
55 Regenerator Bed Material From Bitumen Gasification,
Run No. 5. Broadband ESCA Scan 143
56 Regenerator Bed Material From Bitumen Gasification,
Run No. 5. Carbon ESCA Scan 145
57 Regenerator Bed Material From Bitumen Gasification,
Run No. 5. Sulfur ESCA Scan 146
58 Stack Sampling Train Filter Catch During Bitumen Combustion
and Fresh Limestone Feeding, Run No. 6. Broadband ESCA Scan 148
59 Stack Sampling Train Filter Catch During Bitumen Combustion
and Fresh Limestone Feeding, Run No. 6. Carbon ESCA Scan 149
60 Stack Sampling Train Filter Catch During Bitumen Combustion
and Fresh Limestone Feeding, Run No. 6. Sulfur ESCA Scan 150
61 Stack Cyclone Material From Fuel Oil Gasification, Run No. 4,
LC Fraction 1. IR Spectrum 152
62 Stack Cyclone Material From Fuel Oil Gasification, Run No. 4,
LC Fraction 2. IR Spectrum 152
63 Stack Cyclone Material From Fuel Oil Gasification, Run No. 4,
LC Fraction 3. IR Spectrum 153
64 Stack Cyclone Material From Fuel Oil Gasification, Run No. 4,
LC Fraction 4. IR Spectrum 153
65 Stack Cyclone Material From Fuel Oil Gasification, Run No. 4,
LC Fraction 5. IR Spectrum 154
66 Stack Cyclone Material From Fuel Oil Gasification, Run No. 4,
LC Fraction 6. IR Spectrum 154
67 Stack Cyclone Material From Fuel Oil Gasification, Run No. 4,
LC Fraction 7. IR Spectrum 155
68 Stack Cyclone Material From Fuel Oil Gasification, Run No. 4,
Broadband ESCA Scan 159
69 Stack Cyclone Material From Fuel Oil Gasification, Run No. 4,
Vanadium ESCA Scan 160
70 Stack Cyclone Material From Fuel Oil Gasification, Run No. 4,
Sulfur ESCA Scan 161
xii
-------
LIST OF FIGURES (continued)
No. Page
71 Stack Cyclone Material From Fuel Oil Gasification,
Run No. 4. Carbon ESCA Scan 162
72 Stack Sampling Train Filter Material From Fuel Oil Gasification,
Run No. 4. Surface Broadband ESCA Scan 164
73 Stack Sampling Train Filter Material From Fuel Oil Gasification,
Run No. 4. Subsurface Broadband ESCA Scan 165
74 Regenerator Bed Material From Fuel Oil Gasification,
Run No. 4. Broadband ESCA Scan 168
75 Regenerator Bed Material From Fuel Oil Gasification,
Run No. 4. Sulfur ESCA Scan 169
76 Regenerator Bed Material From Fuel Oil Gasification,
Run No. 4. Carbon ESCA Scan . 170
77 Annual Surface Wind Roses 177
78 Annual Mean Daily Solar Radiation 178
79 Annual Mean Windspeeds and Resultant Wind Directions 180
80 Isopleths (m x 10 ) of Mean Annual Morning Mixing Heights 181
2
81 Isopleths (m x 10 ) of Mean Annual Afternoon Mixing Heights 182
82 Horizontal Dispersion Coefficient as a Function of Distance
for Pasquill's Stability Types 186
83 Vertical Dispersion Coefficient as a Function of Distance
for Pasquill's Stability Types 187
84 S02 Concentrations Versus Downwind Distance for Stability
Class 1 192
85 S02 Concentrations Versus Downwind Distance for Stability
Class 2 193
86 S02 Concentrations Versus Downwind Distance for Stability
Class 3 194
87 S02 Concentration Versus Downwind Distance for Stability
. Class 4 195
xiii
-------
LIST OF FIGURES (continued)
No.
88 S02 Concentration Versus Downwind Distance for Stability
Class 5 196
89 SO, Concentration Versus Downwind Distance for Stability
Class 6 197
90 S02 Concentration Versus Downwind Distance for Stability
Class 1 (Plume Rise Retardation Included) 200
91 S02 Concentration Versus Downwind Distance for Stability
Class 2 (Plume Rise Retardation Included) 201
B-l Typical Two-Stage Claus Sulfur Plant 224
C-l Flexicoking Unit 237
C-2 Flexicoking Products 239
C-3 The Gulf HDS Process 242
C-4 Typical RCD Isomax Unit Flow Diagram 245
C-5 BP Residue Desulfurization Process 248
C-6 Resid Hydroprocessing - Standard Oil Co., Indiana 253
C-7 LC-Fining Process Flow Diagram 256
C-8 Residual Ultrafining 258
C-9 Go-Fining 261
C-10 Schematic of the Residfining Process 264
C-ll Residue Hydrodesulfurization Flow Diagram 266
C-12 Hydrodesulfurization, Trickle Flow, Flow Diagram 270
C-13 IFP Resid and VGO Desulfurization Flow Diagram 272
C-14 Demetalization/Desulfurization Flow Diagram 275
C-15 Simplified Flow Diagram for Delayed Coking • 278
C-16 VGO/VRDS Flow Diagram 281
xiv
-------
LIST OF FIGURES (continued)
No. Page
C-17 Combined Cycle/Shell Gasification Process 285
C-18 Shell Gasification Power Generation Block Diagram 286
xv
-------
LIST OF TABLES
No.
1 ERCA Pilot Plant Mass Flow Rates 19
2 Mass Flow Rates for FW 10 MW Oil-Fired CAFB Demonstration
Plant 21
3 Mass Flow Rates for FW 250 MW Oil-Fired CAFB Design 24
4 Analysis of Fuel Oil Used by ERCA 42
5 Analysis of Bitumen Used by ERCA 44
6 Analysis of Limestone Used by ERCA 44
7 "Typical" Fuel Oil to be Used at the FW Demonstration Plant 46
8 Volatile or Toxic Trace Elements in Oil and Stone 51
9 Comparison of Worst Case Emission Estimates With Air
Quality Goals 53
10 Composition of Drainage From Coal Piles 54
11 Summary of Potential Environmental Impacts From the
La Palma Station Cooling Towers 59
12 Water Effluent Standards 61
13 Residual Chlorine Recommendations 61
14 Chemicals Used in Recirculative Cooling Water Systems 62
15 Cooling Tower Corrosion and Scale Inhibitor Systems 62
16 Classes of Organic Compounds Eluting in Each Liquid
Chromatography Fraction, and Their Approximate IR
Detection Limits 81
17 Elemental Sensitivity Factors for the ESCA . 83
18 Summary of CAFB Pilot Plant Operating Modes During Test
Program 91
19 Representative Pilot Plant Operating Temperatures 91
20 Summary of Sampling Activity 92
xvi
-------
LIST OF TABLES (continued
No. Page
21 Leached Stone Samples 93
22 Field Test Results: Run 1 Fuel Oil Gasification 94
23 Field Test Results: Run 2 Fuel Oil Gasification 94
24 Field Test Results: Run 3 Fuel Oil Gasification 95
25 Field Test Results: Run 4 Fuel Oil Gasification 95
26 Field Test Results: Run 5 Bitumen Gasification 96
27 Field Test Results: Run 6 Bitumen Combustion 96
28 Field Test Results: Run 7 Bitumen Gasification 96
29 Summary of Stack Emissions 97
30 Sample Analyses 106
31 Distribution of Material and Function Groups in Bitumen 112
32 Health and Ecological Effects and MEGS of Organic
Compound Classes 113
33 Distribution of Material and Functional Groups in Stack
Gas Effluent: Run No. 7 118
34 Distribution of Extractable Organic Material and Functional
Groups in Stack Cyclone Particulate: Run No. 5 123
35 Mass Spectrographic and Atomic Absorption Spectrometric
Analysis of Stack Cyclone Particulate: Run No. 5 125
36 Surface and Subsurface Concentrations of Stack Particulate
Collected During Bitumen Gasification 131
37 Surface Concentrations of Gasifier Bed, Gasifier Cyclone
and Knockout Baffle Particulate 134
38 Distribution of Extractable Organic Material and Functional
Groups in Spent Stone: Run No. 5 142
39 Surface Concentrations of Spent Stone Particles, Run No. 5 144
xvii
-------
LIST OF TABLES (continued)
No. Page
40 Surface Concentrations of Stack Particulate: Run No. 6 147
41 Distribution of Extractable Organic Material and Functional
Groups in Stack Cyclone Particulate: Run No. 4 156
42 Mass Spectrographic and Atomic Absorption Spectrometric
Analysis of Stack Cyclone Particulate: Run No. 4 157
43 Surface and Subsurface Concentrations of Stack Particulate
Collected During Fuel Oil Gasification: Runs 1 to 4 166
44 Surface and Subsurface Concentrations of Particulate Col-
lected on Impactor Substrates: Run No. 4 167
45 Surface Concentrations of Spent Stone Particles: Run No. 4 171
46 Surface and Subsurface Elemental Compositions of Leached
Stone Samples 171
47 '..'. Relation of Pasquill Stability Classes to Weather Conditions 185
48 National Ambient Air Quality Standards 198
49 Texas Ambient Particulate Standards 199
50 Allowable Particulate Emission Rates For Specific Flow
Rates 202
A-l Mass Flow Rates for FW 10 MW Coal-Fired CAFB Demonstration
Plant 206
A-2 Mass Flow Rates for FW 250 MW Coal-Fired CAFB Design 211
A-3 Emission Factors for Coal Drying 214
A-4 Power Plant Coal Ash Compositions 217
A-5 Selected Trace Elements in Ash (ppm) 217
B-l Summary Description of Flue Gas Desulfurization Processes 228
B-2 Oil-Fired Utility Boilers in the United States Employing
Flue Gas Desulfurization 230
B-3 FGD Environmental Impact, tons/yr 231
xviii
-------
LIST OF TABLES (continued)
No.
C-l Economics of Flexicoking Process 240
C-2 Reduced Crudes to HDS Units 241
C-3 Economics of Gulf HDS 242
C-4 Development of Gulf HDS 243
C-5 Yields for ROD Isomax Processing of Kuwait Reduced Crude
to 1.0 wt Percent Sulfur 244
C-6 Yields for RCD Isomax Processing of Kuwait Reduced Crude
to 0.7 wt Percent Sulfur 244
C-7 Yields for RCD Isomax Processing of Kuwait Reduced Crude
to 0.3 wt Percent Sulfur 245
C-8 Economics of RCD Isomax Process 247
C-9 Residue Desulfurization Pilot-Plant Data 249
C-10 Economics of BP Process 250
C-ll HDS Yield Data 252
C-12 Economics of HDS Process 254
C-13 Development of LC-Fining Process 255
C-14 Desulfurization of Kuwait Atmospheric Bottoms 257
C-15 Resid Ultrafining Desulfurization Costs 260
C-16 Resid Ultrafining West Texas Sour Desulfurization Costs 260
C-17 Go-Fining Yields at 90 Percent Desulfurization Level 262
C-18 Economics of Go-Fining 262
C-19 Economics of Residfining Process 264
C-20 Residue HDS Product Yields 267
C-21 Economics of Residue HDS 268
xix
-------
LIST OF TABLES (continued)
No.
C-22 Typical Results From Hydrodesulfurization of Thermal
Cracker Gas Oil 269
C-23 Economics of Hydrodesulfurization, Trickleflow Process 271
C-24 Feed Specs and IFF Process Performance 273
C-25 Yields From Kuwait Residue 273
C-26 Typical Economics of IFP HDS Process 274
C-27 Cost Comparison - Desulfurization Versus Demetalization/
Desulfurization 277
C-28 Low Sulfur Fuel Oil Production From Arabian Light Residuum 280
C-29 VGO Isomax Plants 282
C-30 Investment Summary 283
C-31 Processing Costs 283
C-32 Typical Product Gas Composition 288
C-33 Power Generation Cost - 200 MW Study 288
D-l Process Parameters of High Metals Feedstock Desulfurization
Techniques 292
D-2 Process Parameters of Residual Oil Feedstock Desulfuriza-
tion Techniques 293
D-3 Economics for Other Residual Oil Utilization Techniques 296
D-4 Comparative Capital Costs of the CAFB, HDS and FGD
Processes 298
D-5 Comparative Operating Costs for the CAFB, HDS and
Processes, $/yr 299
D-6 1972 Projected Costs for the CAFB and FGD Processes 300
xx
-------
ACKNOWLEDGMENTS
The authors gratefully acknowledge the guidance and support provided
by the Project Officer, Mr. Samuel Rakes. We thank Dr. Robert Statnick,
Process Measurements Branch, Industrial Environmental Research Labora-
tory, U.S. Environmental Protection Agency, for assistance with planning
and completing the field test and laboratory analysis programs. The
authors also express their gratitude to the following people: Dr. Graham
Johnes, Mr. Z. Kowszun, Dr. Gerry Moss and their staff at the Esso Re-
search Centre, Abingdon (ERCA), for their cooperation during the field
test program; Mr. Richard McMillan and Mr. Frank Zoldak of the Foster-
Wheeler Energy Corporation (FW) for helpful discussions; Dr. Peter Jones
of Battelle Columbus Laboratories for consultations regarding organic
sampling and analysis; and Dr. Paul Larson and Mr. John Rendina of the
GCA/McPherson Instrument Division for performing the ESCA analyses of
pilot plant samples. Finally, we acknowledge the following GCA/Technology
Division staff members: Mr. John Langley and Mr. Stephen Brenan for
assistance with the field test program, and Ms. Dorothy Sheahan and
Ms. Susan Field for assistance in preparing this report.
xxi
-------
SECTION I
EXECUTIVE SUMMARY
OVERVIEW
This document presents the results of a preliminary environmental assess-
ment of the Chemically Active Fluid Bed (CAFB) process. The CAFB is a
technique whereby high sulfur, high metal residual oil is vaporized in a
fluidized bed of lime to produce a low Btu, low sulfur product gas which
is then burned in a conventional boiler to generate electrical energy.
Most of the sulfur and metals contained in the oil feed are captured by
the lime. This spent lime is subsequently processed to recover sulfur.
At present the only existing CAFB unit is a 2.93 MW pilot plant at the
Esso Research Centre, Abingdon, England facility. Foster Wheeler Energy
Corporation (FW) is in the final design and procurement stages of a
10 MW retrofit demonstration plant to be constructed in San Benito,
Texas, at the La Palma Power Station of the Central Power and Light
Company. In addition, FW has developed a conceptual design for a
2
250 MW commercial scale unit.
Figure 1 is a generalized schematic diagram of the CAFB showing principal
unit operations and material flows. Limestone and oil are fed continuously
into the .gasifier at a Ca (limestone)/S (oil) molar ratio of unity.
Limestone (CaCO^) is rapidly converted to lime (CaO) and C0~ and the
lime is maintained in a fluidized state by a preheated air/flue gas
mixture. The air input rate is equal to roughly 20 percent of stoichio-
metric with respect to oil. Fuel oil is consecutively vaporized,
-------
c
OIL
c
LIMESTONE
STACK
CYCLONE
1
?
COOLING
TOWER
BOILER
COAL
CYCLONE
GASIFIER
RESOX
TM
COAL
ASH
REGENERATOR
SPENT
STONE
-^-SULFUR
Figure 1. Generalized schematic of the CAFB
-------
oxidized, cracked and reduced at 870°C (1600°F) to produce a low Btu gas.
Over 80 percent of the input feed sulfur is removed by the lime. The
gas travels from the gasifier through cyclones for particulate removal
and then into a boiler for combustion. The boiler flue gas encounters
a knockout baffle and another cyclone before entering the stack.
Lime is continuously cycled between the gasifier and the regenerator where
the roughly 7 percent of the lime which is sulfided in the gasifier is
oxidized to CaO. Sulfur dioxide produced in the regenerator is fed to the
boiler stack in the pilot plant or reduced to elemental sulfur by coal in
the demonstration and commercial plants. Some spent lime is continuously
withdrawn from the regenerator and retained for disposal. To maintain
sulfur removal efficiency, an equivalent amount of limestone is continuously
added to the gasifier.
The CAFB generates pollutants to air, water and land. The primary source
of air emissions is the boiler stack but fugitive emissions from feed ma-
terial storage and handling are also present. Water effluents are similar
to those found in conventional combustion systems and include boiler blow-
down and cooling tower outputs. Disposal of spent, sulfided limestone is
a major environmental problem. A substantial amount of work has and is
being carried out to develop an environmentally sound method for stone
disposal. 3
In view of the extensive efforts in the area of solid waste treatment and
disposal, the study reported upon here concentrated primarily on air emis-
sions and to a lesser extent on effluents to the other media. Character-
ization of the multi-media effects of the CAFB involved theoretical en-
gineering and emission calculations for all three CAFB development stages
and an extensive field measurement and laboratory analysis program for the
pilot plant.
The preliminary theoretical phase of the study utilized CAFB pilot plant
data, engineering data developed by FW for the demonstration and com-
mercial units, reports dealing specifically with the CAFB and the general
-------
literature to project emission levels from the CAFB demonstration plant
and proposed commercial sized facility. Fugitive air emissions were
identified as resulting from the storage and handling of oil, limestone
and coal, the latter material being used in the FW RESOX™ process to
reduce sulfur dioxide emanating from the regenerator to elemental sulfur,
and from cooling tower discharges. The fugitive oil vapor emission rate
is 104 kg (250 lb)/tank fillup for the demonstration plant and 13 kg/s
(103 Ib/hr) for the commercial facility. One of the two fuels used at the
pilot plant, bitumen, was found possibly to contain polycyclic organic
matter (POM); thus emissions from storage of this material, as well as
from other oil feeds, must be investigated further. The fugitive dust
emission rate from limestone storage and handling is projected to be
3 mg/s (0.024 Ib/hr) at the demonstration plant and from 1.9 to 30.6 g/s
(13 to 244 Ib/hr) at the commercial unit, the extremes corresponding to
uncontrolled and baghouse contained crushing emissions. Uncontrolled
emissions from coal utilization are estimated to be 6.7 x 10~^ mg/s
(5.3 x lO"4 Ib/hr) and 0.89 mg/s (7.1 x 10'3 Ib/hr) for the demonstration
and commercial plants respectively. Cooling tower drift losses at the
demonstration plant are estimated to be beween 6 and 50 x ICT1* m3/s (1.2
and 4.8 cfm) with an evaporative loss of 0.064 m3/s (135 cfm).
Discharges to ambient water will come from coal pile runoff, cooling tower
blowdown and boiler blowdown. Because the CAFB demonstration plant will
utilize an existing boiler at the La Palraa Power Station, cooling tower
and boiler blowdown effluents will be unaffected by CAFB retrofit. At the
demonstration plant RESOX™ coal will be stored in a bin; hence no runoff
is expected. Runoff from coal storage for the commercial plant will depend
upon the specific site, but is estimated to be roughly 212 m3/year
(7500 ft3/year).
Solid waste disposal requirements will depend upon marketability and
TM
disposal options for spent stone, RESOX coal ash and elemental sulfur.
The demonstration plant will generate 0.07 kg/sec (557 Ib/hr) of spent
stone and the commercial facility will produce O.<91 kg/sec (7,190 Ib/hr)
-------
of this material. As noted earlier, disposal of this solid waste is the
major environmental problem associated with the CAFB.
The bulk of the sampling and analysis program carried out in conjunction
with pilot plant operating during December 1975, was directed toward
quantifying stack emissions. Samples were collected during seven separate
runs: four fuel oil gasification runs, two bitumen gasification runs and
one combustion/startup bitumen run. The field measurement program en-
tailed on-site quantification of SC>2, 863, NOX, H2S, total particulate and
particulate size distributions. In addition, vapor and particulate samples
were collected for subsequent chemical analyses. Sulfur dioxide emission
rates for fuel oil gasification averaged 0.63 lb/106 Btu, 80 percent of
the New Source Performance Standard (NSPS) for oil-fired steam generators.
Bitumen gasification under conditions of saturated gasifier bed stone
(caused by clogging in the gasifier-regenerator stone transfer duct) re-
sulted in an S02 emission rate of 1.6 lb/106 Btu. Sulfur trioxide emis-
sion rates averaged 0.023 lb/106 Btu for these same three runs. Total
nitrogen oxide emissions ranged from 0.067 to 0.085 lb/106 Btu, roughly
25 percent of the NSPS for oil-fired boilers. No significant H2S was
detected in any run. Total particulate emissions ranged from 0.063 to
0.10 lb/106 Btu for normal gasification (the NSPS is 0.1 lb/106 Btu).
During fresh stone feed this rate increased to 0.19 lb/106 Btu due to
attrition of fresh particles. Particulate size distribution measurements
made under gasification conditions for both fuel oil and bitumen feeds
indicated roughly one third of the escaping stack particulate is in the
respirable size range.
Laboratory analysis of stack particulate employing spark source mass
spectrometry (SSMS), atomic absorption spectroscopy (AA) and electron
spectroscopy for chemical analysis (ESCA) demonstrated that vanadium, which
is bound in a mixture of oxides, is emitted at a rate of almost 90 percent
of the EPA established critical value. No other trace element emissions
were found to be of concern. Both particulate and gaseous stack samples
-------
were also analyzed for organic functional groups by the procedure out-
lined by the EPA Level 1 protocol. Flue gas analysis results indicated
that emissions of hydrocarbons, quinone and carbonyl compounds are po-
tentially of concern.
The results of the field measurement program were used in conjunction with
meteorological and topographical characteristics of the San Benito area to
project ambient loadings of SC>2 and particulate in the vicinity of the
demonstration plant. These projections were compared with State of Texas
regulations and found to be in compliance with state requirements.
Finally, the CAFB was compared with alternative residual oil utilization
techniques: feed stock desulfurization and flue gas desulfurization (FGD).
Of 17 feedstock treatment processes, only three are capable of handling
high sulfur and high metals content oil. Comparisons were also developed
comparing projected capital and operating costs of the CAFB, FGD and feed
stock desulfurization which show that for commercial size facilities, FGD
appears to be the most economical of the three options. However, the
only existing FGD unit on an oil-fired plant is a MgO scrubber which is
almost four times as expensive as published projected costs for FGD units.
CONCLUSIONS
Air
Priority Problems -
• Reduction of stack particulate emissions. Total stack particulate
emissions from oil-fired operation of the pilot plant, 30 percent
of which are in the respirable size range, were only slightly
lower than the New Source Performance Standard (NSPS) for oil-
fired boilers. During stone feed/start-up these emissions con-
siderably exceeded the NSPS. The vanadium concentration of these
particulates is such that the vanadium emission rate is only
slightly lower than the Multi-Media Environmental Goal (MEG) for.
this element. Under coal-fired operation of the CAFB, proposed
for the demonstration plant, the particulate emission problem may
be even more pronounced. Foster-Wheeler is designing more
-------
efficient cyclones than were installed at the pilot plant.
Extensive particulate emission rate measurements at the demon-
stration plant should be undertaken for all operating modes
and for all fuels.
Reduction of SC>2 emissions during abnormal operating conditions.
Blockage of the gasifier-regenerator transfer duct causes satura-
tion of gasifier bed stone and a resultant increase in SC>2 emis-
sions. Operation of the CAFB in this mode for extended time
periods should be avoided. Continuous SC>2 monitoring is
recommended.
Problems Needing Further Study, But Which Could be Important -
Detailed investigation of organic stack emissions. Flue gas
analyses indicated the possible presence of quinone, carbonyl
compounds and aliphatic hydrocarbons in sufficient quantitites
to produce ambient concentrations in the neighborhood of the
MEG's for these species. Organic emissions are highly dependent
on gasifier and boiler operating conditions and should be analyzed
with greater specificity than was possible in the present study.
Measurement of fugitive emissions from oil storage. Polycyclic
organic matter (POM) was tentatively identified as a constituent
of bitumen. Fugitive air emissions of these compounds from
storage and handling of bitumen present a potential environmental
hazard. Additional characterization of these emissions is
required.
Areas Not Definable Because of Lack of Data -
• Fugitive emissions from RESOX™ coal and ash handling and storage.
• Fugitive emissions from storage and handling of spent stone.
Areas Probably Not Important But Requiring Checking -
• Fugitive emissions from limestone handling.
• Cooling tower emissions.
-------
Areas Definitely Not a Problem -
• NOX stack emissions. Measurements of NOX emissions for three
separate runs were about 25 percent of the NSPS for oil-fired
boilers.
• Trace elements other than vanadium. Stack emission rates of no
element other than vanadium approached creating ambient levels on
the order of the MEG for that element.
Water
Areas Not Definable Because of Lack of Data -
Chemical composition of boiler blowdown, cooling tower blowdown
and RESOX™ coal pile runoff. Effluents from the first two
categories will be unaffected by CAFB retrofit to existing boilers.
Coal pile runoff characteristics will be coal type and site
specific.
Solid Waste
Priority Problem -
Environmentally acceptable disposal of spent stone. The demonstra-
tion plant will generate 6000 kg/day (13,000 Ib/day) and a 250 MW
commercial size unit 79,000 kg/day (173,000 Ib/day) of sulfided,
metal containing lime which must be treated before being disposed
of by selling, using as landfill or dumping in the ocean.
Problem Needing Further Study But Which Could be Important -
• Environmentally acceptable disposal of RESOX^M coal ash. Approxi-
mately 1600 kg/day (3600 Ib/day) of ash will be produced at the
demonstration plant and 22,000 kg/day (48,000 Ib/day) will be
generated at a 250 MW facility.
-------
Area Probably Not Important by Requiring Checking -
Environmentally acceptable disposal of elemental sulfur. The
RESOXTM unit will produce 2600 kg/day (5640 Ib/day) of sulfur at
the demonstration plant and 35,000 kg/day (76,000 Ib/day) at a
commercial 250 MW units. Forster-Wheeler plans to sell this
material if a market can be found.
-------
REFERENCES
1. Craig, J. W. T., G. L. Johnes, Z. Kowszun, G. Moss, J. H. Taylor, and
D. E. Tisdall. Chemically Active Fluid-Bed Process for Sulphur
Removal During Gasification of Heavy Fuel Oil - Second PHase. Esso
Research Centre, Abingdon, Berkshire, England. U.S. Environmental
Protection Agency, Research Triangle Park, N.C. Report Number EPA-
650/2-74-109. November 1974. 589 p.
2. Chemically Active Fluid Bed Process (CAFB) Preliminary Process Design
Manual. Foster Wheeler Energy Corp., Livingston, N. J. U.S. Environ-
mental Protection Agency, Research Triangle Park, N.C., Contract
Number 68-02-2106. December 1975. 185 p.
3. Keairns, D. L., R. A. Newby, E. J. Vidt, E. P. O'Neill, C. H. Peterson,
C. C. Sun, C. D. Buscaglia and D. H. Archer. Fluidized Bed Combustion
Process Evaluation (Phase 1 - Residual Oil Gasification/Desulfurization
Demonstration at Atmospheric Pressure) Volumes I and II. Westinghouse
Research Laboratories, Pittsburgh, Pa. U.S. Environmental Protection
Agency, Research Triangle Park, N.C. Report Number EPA-650/2-75-027a.
March 1975. 578 p.
10
-------
SECTION II
INTRODUCTION
THE CHEMICALLY ACTIVE FLUID BED PROCESS
The Chemically Active Fluid Bed (CAFB) process was developed by the Esso
Research Centre, Abingdon (ERCA), England as a means to generate electrical
energy from high sulfur, high metal heavy fuel oil. Fuel oil is fed con-
tinuously into a fluidized bed of limestone maintained at 870 C (1600 F)
by preheated, substoichiometric air. The fuel oil entering the gasifier
is vaporized, oxidized, cracked and reduced to produce a low-Btu, low-
sulfur gas which is then burned in a conventional boiler. Sulfur contained
in the oil initially forms various gaseous compounds which then react with
the bed limestone to yield solid calcium sulfide. The sulfided lime is
cycled to a regeneration unit where it is oxidized to produce CaO which
is returned to the gasifier and SO,, which is sent to a sulfur recovery unit.
An additional feature of the CAFB process is that the gasifier bed material
adsorbs vanadium, nickel and sodium contained in the fuel oil, thus
limiting air emissions of these trace elements.
At present the only existing CAFB unit is a 2.93 MW pilot plant at the
ERCA facility. Fos.ter-Wheeler Energy Corporation (FW) is in the final
o
design and procurement stages of a 10 MW retrofit demonstration plant
to be constructed in San Benito, Texas, at the La Palma Power Station
of the Central Power and Light Company. In addition, FW has developed
o
a conceptual design for a 250 MW commercial scale unit.
11
-------
PROGRAM OBJECTIVES
The objective of this study was to conduct a preliminary environmental
assessment of the CAFB. The results of this program provide guidance on
measures which must be taken to minimize the environmental impact of the
CAFB and suggest follow-up investigations which should be undertaken to
insure the environmental acceptability of this process.
To attain these goals, a systematic evaluation of all waste streams from
the CAFB was made and a process emissions inventory was compiled. These
data were derived from engineering estimates and from an extensive pilot
plant field sampling and laboratory analysis program. Emission rates de-
termined for the pilot plant were then used to predict pollutant loadings
for the CAFB demonstration plant and for the proposed commercial unit. To
provide a long-term overview proposed coal-fired operation of the CAFB is
also evaluated. The emissions data are compared with legal requirements
and quantifiable health and ecological effects and sources of concern are
noted. As part of this latter task, procedures for calculating incre-
mental ambient air loadings are outlined and used to compare projected
S09 and particulate emissions from the demonstration plant with federal
and State of Texas regulations.
In addition to the preparation of the environmental assessment, a pre-
liminary economic assessment was completed which compares the investment
and operating costs of a commercial CAFB facility with the costs of al-
ternative residual oil utilization techniques: flue gas desulfurization
and feedstock desulfurization.
REPORT ORGANIZATION
Section III provides process descriptions of the ERCA pilot plant, FW de-
monstration unit and the proposed FW 250 MW commercial facility. Each
development stage is broken down into its component unit operations and
schematic flow diagrams are developed and waste streams identified.
12
-------
Emissions estimates developed from engineering evaluations and worst case
analyses are presented in Section IV. These projections concentrate on
waste streams which were not investigated as part of the field test program.
Section V, which is the crux of this report, describes the field test
program and subsequent laboratory analytical studies carried out to
characterize stack gas and particulate emissions and solid waste efflu-
ent. The results of these studies are presented and intrepreted in terms
of potential environmental impact.
Section VI discusses the meteorological and topographical characteristics
of a source which control the transport of air pollutants emitted from
a stack. The models developed here are then applied to SO- and particu-
late emissions from the La Palma Electric Generating Station.
Conclusions and recommendations for future work are presented in the
Executive Summary, Section I.
Appendix A considers coal-fired operation of the CAFB and highlights dif-
ferences in operating parameters and potential loadings between this
mode and oil-firing.
The final three appendixes constitute the comparative economic evaluation
of the CAFB. Appendix B discusses the operating characteristics and
potential emissions from flue gas desulfurization and from the three
feedstock desulfurization procedures capable of processing high metal
content residual oil.
Appendix C provides process descriptions and flow diagrams of 15 residual
oil desulfurization techniques identified by GCA as being either in com-
mercial operation or potentially viable.
13
-------
The economic comparison between the CAFB, flue gas desulfurization and
residual oil feedstock desulfurization is presented in Appendix D.
14
-------
REFERENCES
1. Craig, J.W.T., G.L. Johnes, Z. Kowszun, G. Moss, J.H. Taylor, and
D.E. Tisdall. Chemically Active Fluid-Bed Process for Sulphur
Removal During Gasification of Heavy Fuel Oil - Second Phase.
Esso Research Centre, Abingdon, Berkshire, England. U.S. Environ-
mental Protection Agency, Research Triangle Park, N.C. Report
Number EPA-650/2-74-109. November 1974. 589 p.
2. Chemically Active Fluid Bed Process (CAFB) Preliminary Process
Design Manual. Foster Wheeler Energy Corp., Livingston, N.J.
U.S. Environmental Protection Agency, Research Triangle Park,
N.C., Contract Number 68-02-2106. December 1975. 185 p.
15
-------
SECTION III
PROCESS DESCRIPTION
INTRODUCTION
This chapter summarizes, to the degree of detail requisite to an emissions
assessment, the technical aspects of the CAFB process. The process de-
scriptions consider each of three development stages, the ERCA pilot plant,
the FW 10 MW demonstration unit presently approaching final design specifi-
cations and the FW conceptual design of a 250 MW commercial unit. Operating
123
and engineering design parameters have been culled from Esso ' ' and
4
Foster Wheeler reports, from conversations with representatives of these
organizations and during a site visit and subsequent sampling operation
at the ERCA pilot plant. This section considers oil gasification only;
proposed coal gasification is discussed in Appendix A.
OVERVIEW
In the CAFB process heavy fuel oil is consecutively vaporized, oxidized,
cracked and reduced in a fluidized bed of lime to produce a low Btu
gas. This gas, from which over 80 percent of the sulfur has been re-
moved by the lime, travels from the gasifier through cyclones for
particulate removal and then into a boiler for combustion. The boiler
flue gas encounters a knockout baffle and another cyclone before entering
the stack. Lime is continuously cycled between the gasifier and the
regenerator where lime sulfided in the gasifier is oxidized to CaO. Sul-
fur dioxide produced in the regenerator is fed to the boiler stack or
16
-------
chemically treated to recover sulfur. Some spent lime is continuously
withdrawn from the regenerator and retained for disposal. To maintain
sulfur removal efficiency, an equivalent amount of limestone is continu-
ously added to the gasifier.
Figure 2 is a schematic diagram of the ERCA pilot plant. Input and out-
put streams to and from each unit operation are labeled and their mass
flows and characteristics are given in Table 1. The quantities listed
in this table are those projected at steady-state conditions. Parameters
will vary during start-up, shut-down and other atypical operating modes.
These variations, as pertinent to emission rates, are discussed in the
sections describing specific unit operations.
The Foster Wheeler demonstration plant shown schematically in Figure 3
contains, in addition to the basic gasifier and regenerator units, a RESOX™
system for sulfur recovery from the regenerator off gas, a spent solids
handling system and a coal storage and feed system for coal gasification.
As noted earlier coal gasification will be discussed in Appendix A.
Mass flows and stream conditions listed in Table 2 are based on FW
design parameters.
The proposed design for a 250 MW CAFB unit is illustrated in Figure 4
The general design is similar to the demonstration plant but more complex
in terms of number of unit operations and number of modules required
for each unit operation. Stream conditions listed in Table 3 are again
based on FW projections.
The remainder of this section treats each CAFB unit operation separately
and describes the variation in that unit operation for each stage of
development. Waste streams are identified but discussion as to their
nature is presented in Sections IV and V.
17
-------
00
LIMESTONE
FEED
HOPPER
GASIFIES AIR BLOWERS
Figure 2, Unit operations flow diagram of the ERCA pilot plant
-------
Table 1. ERCA PILOT PLANT MASS FLOW RATES
Process stream
1. Oil feed to gasifier
2. Limestone feed to gasifier
3. Gasifier to regenerator stone transfer
4. Regenerator to gasifier stone transfer
5. Product gas to cyclone
6. Cyclone solids return to gasifier
7. N_ gas to solids transfer lines
8. Product gas to boiler
9. Air to regenerator
10. Spent solids from regenerator
11. Regenerator off gas to cyclone
12. Regenerator off gas, cyclone to stack
13. Flue gas from boiler
14. Flue gas recirculated to gasifier
15. Flue gas to Tuyere Blower
16. Recycled flue gas from cyclone
17. Flue gas and air to gasifier
18. Flue gas to stack
19. Solids from boiler flue gas cyclone
20. Solids from recycled flue gas cyclone
21. Solids from regenerator off gas cyclone
22. Start up kerosene to gasifier
23. Stack emissions
24. Fuel injection air
Mass
kg/sec
0.04
0.003
0.11
0.11
0.16
0.0006
0.16
0.01
0.002
0.01
0.01
0.50
0.03
0.02
0.02
0.10
0.50
0.0005
0.50
0.01
flow rate,
(Ib/hr)
(288)
(25)
(860)
(850)
(1,279)
(4.5)
(1,279)
(65)
(14)
(63)
(63)
(4,000)
(250)
(125)
(125)
(800)
(4,000)
(4)
(4,000)
(45)
Temperature,
°C (°F)
88 (190)
850 (1,560)
850 (1,560)
1,050 (1,920)
1,050 (1,920)
43 (110)
19
-------
LIMESTONE STORAGE d FEED SYSTEM
to
O
uitcs
OCLIV
—
_ ^lltf^ToilC
• CCEIVER
riLTE*
Y
SURGE
BIN
V
>TIC _^ I
K»T »/0\ 4
RESOX OFF GAS TREATMENT SYSTEM
MODOCT LOW
•A* TO BURKCR8
OIL STORAGE a FEED SYSTEM
•DELIVERY
SPENT SOLIDS SYSTEM
COAL STORAGE & FEED SYSTEM
COAL
Figure 3. Unit operations flow diagram of the FW demonstration plant
-------
Table 2. MASS FLOW RATES FOR FW 10 MW OIL-FIRED
CAFB DEMONSTRATION PLANT
Process stream
1. Limestone to gasifier
2. Product gas from gasifier
3. Gasifier to regenerator stone
transfer
4. Regenerator to gasifier stone
transfer
5. Flue gas to pulsed solid transfer
lines
6. Regenerator of f-gas : total
S02
COz
N2
7. Water or steam injection
8. Regenerator of f-gas after cyclone
and cooling
9. Coal to RESOX™ reactor
10. Hot solids from RESOX™ reactor
11. Waste solids from RESOX™ quench
vessel
12. Hot air to RESOX™ reactor
13. Influent gas to RESOX™ reactor
14. Elemental sulfur from RESOX™
15. Return steam
16. Water to sulfur condenser
17. RESOXTM tail gas
18. Condensed liquid sulfur
19. Fugitive dust from coal handling
system
20. Air to start up heater
21. Air to start up heater
22. Air to RESOX™ reactor
Mass flow rate,
kg /sec (Ib/hr)
0.12
7.52
4.86
4.83
0.5
0.52
0.09
0.02
0.41
0.07
0.52
0.04
0.02
0.02
0.03
0.25
0.66
0.03
(975)
(59,660)
(38,500)
(38,275)
(4,000)
(4,140)
(724)
(128)
(3,288)
(575)
(4,140)
(300)
(150)
(150)
(253)
(2,000)
(5,250)
(253)
Temperature,
°C (°F)
871 (1,600)
171 (340)
1,038 (1,900)
1,038 (1,900)
1,038 (1,900)
1,038 (1,900)
649 (1,200)
760 (1,400)
149 (300)
0
149 (300)
100 (212)
160 (320)
21
-------
Table 2 (continued). MASS FLOW RATES FOR FW 10 MW OIL-FIRED
CAFB DEMONSTRATION PLANT
Process stream
23. Cooling water for RESOX™ solid
waste
24. Steam from quench vessel
25. Regenerator spent solids
26. Regenerator off -gas cycloned solids
27. Air to spent solids cooler
28. Cooled solids
29. Cooler exhaust to cyclone
30. Cooled solids to storage
31. Air emissions from spent solids
cooler
32. Cycloned solids to storage
33. Solids to storage
34. Solid waste from storage silo
35. Air emissions from solids storage
silo
36. Air to gasifier and regenerator
37. Flue gas recycled from stack
38. Boiler stack emissions
39. Flue gas to coal distributing
conveyor
40. Influent gas to gasifier: total
Air
Flue gas
Tail gas
41. Air and flue gas refenerator
42. Coal to distributing conveyor
43. Coal to gasifier
44. Oil to gasifier
45. Fugitive limestone handling
emissions
Mass flow rate,
kg/sec (Ib/hr)
0.01
0.07
0.69
0.22
4.50
1.89
23.50
5.96
3.93
1.37
0.66
0.56
1.47
(50)
(557)
(5,510)
(1,746)
(35,610)
(15,000)
(186,000)
(47,280)
(31,140)
(10,890)
(5,250)
(4,470)
(11,630)
Temperature,
Ofi fQi?\
\j \ r )
38 (100)
177 (350)
482 (900)
171 (340)
171 (340)
121 (250)
22
-------
NJ
II GASIFIER-REGENERATOR SYSTEM
Figure 4. Unit operations flow diagram of the FW 250 MW plant
-------
Table 3. MASS FLOW RATES FOR FW 250 MW OIL-FIRED CAFB DESIGN
Process stream
a
Mass flow rate,
kg/sec (Ib/hr)
Temperature,
°C
10. Limestone to dryer
11. Off-gas from limestone dryer to
baghouse
12. Air emissions from baghouse
13. Solids collected by baghouse
14. Limestone to crusher
15. Fugitive dust emissions from lime-
stone crusher
16. Limestone from crusher
17. Limestone to gasifier modules
18. Fuel oil from short term storage
19. Fuel oil from heating pumping set
20. Oil injection air
21. Fuel oil to gasifier modules
24. Product gas to quad cyclone
25. Product gas to boiler
26. Solids returned from quad cyclone
27. Gasifier to regenerator stone
transfer
28. Regenerator to gasifier stone
' transfer
29. Regenerator off-gas to twin-
cyclones
30. Spent solids from regenerator
31. Regenerator off gas from twin-
cyclones
32. Regenerator off gas from cooler
33. Air to RESOXTM reactor
1.59 (12,590)
19.03 (150,900)
97.67 (774,500)
97.67 (774,500)
63.0 (499,500)
62.71 (497,240)
6.69 (53,070)
0.91
(7,190)
2.05 (16,220)
121 (250)
871 (1,600)
871 (1,600)
1,038 (1,900)
649 (1,200)
24
-------
Table 3 (continued). MASS FLOW RATES FOR FW 250 MW OIL-FIRED CAFB DESIGN
Process stream
34. Gas to RESOX™ reactor
35. Coal to RESOXTM reactor
36. Solid waste from RESOX™ reactor
37. Sulfur gas from RESOX™ reactor
38. Water to solids quench vessel
39. Steam from solids quench vessel
40. Solid waste from quench vessel
41. Gaseous effluent from ash storage
42. Air emissions from ash storage vent
filter
43. Solids from vent filter to ash
storage
44. Solid waste from ash storage
45. Water to sulfur condenser
46. Steam from sulfur condenser
47. Tail gas from sulfur condenser
recycled to gasifier
48. Liquid sulfur to storage
49. Liquid sulfur waste from storage
50. Solids from regenerator and twin-
cyclones
51. Air to solids cooler
52. Air to spent solids storage
53. Air to spent solids storage
54. Solids from solids cooler
55. Solids to storage
56. Exhaust from solids cooler to
cyclone
57. Cycloned solids cooler exhaust to
stack
- - ' -
Mass flow rate,
kg/sec (Ib/hr)
0.50 (4,000)
0.25 (2,000)
0.41 (3,290)
8.58 (68,000)
Temperature,
OG (op)
760 (1,400)
149 (300)
100 (212)
149 (300)
160 (320)
38 (100)
177 (350)
482 (900)
25
-------
Table 3 (continued). MASS FLOW RATES FOR FW 250 MW OIL-FIRED CAFB DESIGN
Process stream
Mass flow rate,
kg/sec (Ib/hr)
Temperature,
oc (oF)
58. Cycloned solids cooler exhaust to
coal and limestone dryers
59. Cycloned solids to storage
60. Solids to storage
61
62
63
Exhaust from storage to vent
filters
Air emissions from vent filters
Solids from vent filters to storage
64. Solid waste from storage
65. Flue gas from boiler to stack
66. Air emissions from stack
67. Flue gas recycled to gasifier
68. Fugitive vapor emissions from long
term fuel oil storage
69. Fugitive vapor emissions from short
term fuel oil storage
70. Air to gasifier
71. Air to regenerator
304.91
304.91
17.81
0.01
0.01
50.95
7.31
(2,418,200)
(2,418,200)
(141,250)
(103)
(103)
(404,020)
(58,000)
171 (340)
Process streams 1 through 9, 22 and 23 are applicable to coal-firing and
are presented in Appendix A.
26
-------
FUEL FEED SYSTEM
ERCA Pilot Plant
Fuel Oil — Fuel oil, stored in an external vented storage tank, is passed
through an oil immersion heater before being fed to the gasifier.
Bitumen — Bitumen is stored in a four-compartment tank car having a total
3
capacity of 63.6 m (18,000 imperial gallons). Each compartment is heated
with gas oil to bring the bitumen to a temperature which will allow it to
flow through the feed lines to the gasifier. The gas oil is stored in a
3
0.18 m (50 imperial gallon) drum.
Kerosene — Kerosene, used as the startup fuel, is stored in an underground
1.8 m3 (501
feed line.
3
1.8 m (500 imperial gallon) tank from which it is pumped into the fuel oil
Emissions — The principal emissions from the fuel feed systems are fugitive
vapors escaping from storage tank venting systems. In addition, there may
be some leakage of liquid fuels during tank fillup (evidence of leakage
was noted about the bitumen tank car). Seepage of this sort is localized
and easily confinable.
FW Demonstration Plant
Oil and Pitch — Fuel will be delivered to the plant in heated tank cars
3
and stored in a heated 378.5 m (100,000 gallon) tank. Oil is transfer!
to the gasifier through two headers located adjacent to the gasifier.
Kerosene — Startup fuel is stored in a separate tank and fed to the
gasifier through the oil delivery system.
Emissions — Considerations similar to those in the pilot plant system
apply here.
27
-------
FW 250 MW Unit
Fuel Oil — Fuel oil will be delivered by rail in heated tank cars and
pumped into short-term and long-term storage tanks. The long-term tank
will be designed for 3 weeks storage and the short-term tank for 2 days
supply. Oil from the short-term tank will be pumped to a heating/pumping
set which will bring the oil to the temperature and pressure required
in the gasifier oil supply header. A portion of the feed oil will be
returned to the short-term storage tank for temperature control.
Kerosene — This system is similar to that of the demonstration plant.
Emissions — Considerations similar to those in the pilot plant apply here.
LIMESTONE HANDLING SYSTEM
ERCA Pilot Plant
Limestone is delivered to the plant in bags and transferred to a ground
level hopper. A pneumatic system transports the stone to an upper hopper
from which it is periodically dropped into a weigh feeder. The limestone
then moves by gravity into the gasifier. Fugitive dust will escape during
hopper loading and stone feed.
FW Demonstration Plant
Limestone will be delivered to the plant by truck and offloaded to a
storage bunker designed to contain a 13 day stone supply. Baghouse filters
attached at the top of the bunker are designed to abate fugitive dust
emissions. Limestone is to be transported through a rotary feeder-airlock
valve, a pneumatic transfer line and finally into a pressurized surge
bin from which another rotary feeder-airlock valve will inject stone into
28
-------
the gasifier. Fugitive dust emissions will be generated by these limestone
handling operations.
FW 250 MW Unit
High calcium limestone will be conveyed to either a 6 day supply pile or
to a dead storage pile containing a 30 day supply. Sorbent will be dried
in a fluidized dryer prior to crushing. During start-up, hot gases for
drying will be provided by combusting coal in the furnace section of the
dryer. At steady state operation, hot exhaust from the spent solids
cooler will be used for coal drying. The dried sorbent will then be
transferred to a limestone crusher. Crushed limestone, sized at less than
than 3.2 mm (1/8 in.), will then be transferred to the gasifier modules.
As presently designed this group of unit operations will produce fugitive
dust emissions from limestone transfer, storage and crushing and fugitive
gases from coal combustion.
GASIFIER
General Description and Chemistry
The basic components of the CAFB process are the gasifier and regenerator.
Figure 5 schematically illustrates the interaction between these unit
operations. Limestone and fuel oil are added to the gasifier at an approx-
imate Ca:S molar ratio equal to one. ERCA pilot plant studies indicate a
sulfur removal efficiency (SRE) of at least 80 percent based on this
stone/feed makeup ratio. Air is fed into the gasifier at 20 to 23 percent
stoichiometric in order to partially oxidize the fuel oil and produce a
temperature 871 C (1600 F) suitable for vaporization and cracking of the
fuel. Flue gas from the boiler at approximately 171°C (340 F) is recir-
culated to the gasifier for temperature control. A product gas is produced
which has a heating value of approximately 1665 kcal/kg (3000 Btu/lb).
The predominant reactions taking place in the gasifier are as follows:
29
-------
Oil thermal cracking -»• C + H2 4- hydrocarbons + H2S + CS2 -I- COS
CaO
CaS 4-
CaO + COS
CaS + CO,
CaO + 1/2CS,
CaS + 1/2CO,
The equilibria for these reactions are well to the right. Approximately
7 percent of the input limestone as calcium oxide is reduced to calcium
sulfide on each pass of stone through the gasifier.
LIMESTONE
DESULFURIZED
PRODUCT GAS
REACTED STONE
REGENERATED STONE
OFF GAS
REGENERATOR
SPENT MATERIAL
AIR RECYCLED
FLUE GAS
AIR
Figure 5. Gasifier-regenerator schematic
ERCA studies also indicate that approximately 95 percent of the vanadium,
75 percent of the nickel and 40 percent of the sodium contained in the
fuel oil are captured by bed stone.
30
-------
Startup is accomplished by heating the unit slowly by kerosene combustion
until the appropriate gasification temperature is reached. Fresh lime-
stone is added toward the end of this period until the requisite bed
depth is attained. During limestone addition, stone attrition and calcin-
ing, which leave the bed in the form of CaO, result in appreciable par-
ticulate and C02 formation.
A possible upset condition, which in fact occurred several times during
the GCA sampling program at ERCA, is clogging of the gasifier-regenerator
stone transfer system. This situation results in saturation of the
gasifier stone with the consequent decrease in SRE.
ERCA pilot plant
The process streams and mass flow rates associated with the ERCA pilot
plant are shown in Figure 2 and listed in Table 1. These quantities
are based upon a product gas flow rate of 0.16 kg/s (1279 Ib/hr).
The gasifier used in the pilot plant is circular,in plan and consists of
a cylindrical and conical section over its height. It is 0.73 m (28 in.)
O O '
in diameter at the top and has a total volume of 1.08 m (38.1 ft ).
Fuel oil enters through a single entrance port situated above the air
distribution mechanism. The quantity of air introduced into the gasifier
is 20 to 23 percent of the stoichiometric amount required to completely
oxidize the carbon in the fuel oil. In addition, flue gas is recirculated
to the gasifier for temperature control. Product gas passes through a
cyclone adjacent to the gasifier before entering the boiler. A solids
drain transports collected particulate matter back into the gasifier.
ERCA studies have defined the most important factors influencing SRE
to be bed depth and stone sulfur content. The static bed depth should
be greater than 38 cm (15 in.) and the content of sulfur in the stone
less than 4 percent. Recent analysis by ERCA indicates that water added
to the gasifier can be detrimental to SRE.
31
-------
FW Demonstration Plant
The specific process streams and mass flow rates specified for the Foster
Wheeler design are shown in Figure 3 and listed in Table 2. The
quantities are based upon development of 8.3 kg/s (65,800 Ib/hr) of
3 3
product gas with a higher heating value of 1735 kcal/m (195 Btu/ft )
2
The gasifier proposed by Foster Wheeler has a floor area of 14.6 m
(149 ft2) and an internal height of 3.66 m (12 ft). Limestone will be
3
fed into the gasifier from an adjacent pressurized surge bin of 1.4 m
3
(50 ft ) volume. A variable speed rotary feeder will inject limestone
at a height of approximately 1.2 m (4 ft) above the level of the chamber
floor. The expanded limestone bed depth will be maintained at 0.91 m
(3 ft). The particle diameter of the limestone feed ranges from 0.6 to
3.2 mm (0.024 to 0.126 in.).
Fuel oil will be fed into the gasifier chamber by way of two headers,
both of which subdivide into 15 injection nozzles. Each nozzle enters
into 1 of 30 oil injection combustion pits which are spaced evenly over
2
the gasifier floor. The pits are square in plan with an area of 0.1 m
2
(1 ft ) and a depth of 12.7 cm (5 in.). Air will be injected at a rate
of 22 percent of the stoichiometric amount required for complete combustion
TM
of the fuel oil. Flue gas and RESOX tail gas are to be recirculated to
the gasifier for temperature control and removal of residual sulfur gas,
respectively.
The gaseous mixture will enter the plenum below the gasifier floor before
entering the nozzle distribution system. Five hundred and ninety stainless
steel air/flue gas nozzles are distributed evenly over the floor area
at a spacing of 15.2 cm (6 in.). Four nozzles enter through the bottom of
each oil injection combustion pit in order to provide uniform interaction
between the fluid limestone bed and fuel oil.
32
-------
FW 250 MW Unit
The gasifier will consist of two modules comprised of three cells each.
Each cell is designed as a mockup of the gasifier/regenerator unit pro-
posed for the 10 MW demonstration plant. An individual cell has a floor
area of 23.2 cm2 (250 ft2) as compared to a floor area of 13.9 m2
(149 ft2) designed for the 10 MW demonstration plant.
Product gas from each cell will pass through a quad cyclone (four
cyclones in parallel) before firing the steam generating unit. Refractory
lined return pipes will convey collected solids back to the cells by
gravity.
Each gasifier module will be approximately 17.7 m (58 ft) high with
the two bottom cells 6.1 m (20 ft) and the top cell 5.5 m (18 ft)
in height. Each gasifier cell will be 3.34 m (11 ft) wide and 7.54 m
(24.75 ft) long. The vertical distance from the air distribution grid
to the ceiling of the cell is to be 4.6 m (15 ft). Ducts leading to the
quad cyclones exit from the top of each gasification cell. During gasi-
fication, oil will be injected into each cell through 50 injection pipes.
The oil injection combustion pits and air distribution system are as
described for the 10MW demonstration plant. Each oil injection pipe will
22 22
serve 0.5 m (5 ft ) of the total cell floor area of 23.2 m (250 ft ).
REGENERATOR
General Description and Chemistry
The regeneration step is accomplished in a reaction vessel adjacent to the
gasifier. Limestone comprised of approximately 93 percent CaO and 7 percent
CaS is fed to the regenerator where it reacts with a stoichiometric quantity
of air by the reactions:
33
-------
CaS + 3/2 02 s CaO + S02
CaS
CaS + 3CaSO. s 4CaO + 4SO_
4^ 2
In addition, carbon deposited on the stone during gasification (approximately
0.3 percent by weight) is oxidized to CO-.
The off gas from the regenerator contains S0«, CO,, and N- derived from the
influent air. Spent solid material consists of approximately 94 percent CaO,
2.5 percent CaSO,, and 3.5 percent CaS. In the Foster Wheeler demonstra-
tion plant and 250 MW unit, off gas will be transported to the RESOX™
system for recovery of elemental sulfur, and spent solids will be conveyed
to a solids cooler and storage bin. These components are shown sche-
matically in Figures 3 and 4. At the ERCA pilot plant regenerator off gas
passes through a cyclone and then into the boiler stack. The stone
transfer rate indicated in Tables 1 and 2 for the ERCA pilot plant and
the FW demonstration plant are based on a factor of 3.3 kg of stone
transferred per kg of oil fed to the gasifier. The S02 volume in the
regenerator off gas is equivalent to 0.031 kg of elemental sulfur per kg
of oil input to the gasifier.
ERCA Pilot Plant
The regenerator used in the ERCA pilot plant is contained in a refractory
concrete block. The axis of the regenerator is offset 0.69 m (27 in.)
from the central axis of the gasifier. The diameter of the regenerator
is 0.25 m (10 in.) at the top and the height of the unit is 3.35 m (132 in.).
A nitrogen gas system is used to pulse solids through transfer pipes
which run between the gasifier and regenerator at the bottom of each unit.
34
-------
Regenerator off gas flows through a cyclone for particulate removal and
then directly to the stack for atmospheric discharge. Spent solid mate-
rial is stored during pilot plant operation and subsequently discarded.
FW Demonstration Plant
The regenerator will be housed within the same structure as the gasifier,
the two vessels being separated by a partition. The plan area of the
2 2
regenerator is 1.8 m (19.3 ft ) and the height is 3.66 m (12 ft). Stone
transport is to be accomplished by way of two transfer conduits in the
separation wall. A set of flue gas nozzles will be included in each
transfer slot in order to maintain continuous material flow through the
duct. A division wall within the regenerator prevents short circuiting
of spent stone back into the gasifier prior to complete regeneration.
TM
Regenerator off gas will pass through a cyclone and then into the RESOX
system. Spent stone will be sent to a solids handling system for eventual
disposal or reuse.
FW 250 MW Unit
The regenerator units will be cast monolithically with the gasifier cells
in each module. A solids transfer system will be housed in the gasifier/
regenerator division wall in order to pulse solids between the two
chambers. The floor area of each regeneration unit (one per gasifier
cell) will be 0.87 m (34.4 in.) by 3.3 m (129.4 in.). The regenerator
off gas containing S02 will pass through two refractory lined cyclones
prior to entering the RESOX™ reactor. Collected solids and spent
material will be transferred to a solids cooler and stored.
Emissions
Of the three CAFB development projects only the ERCA pilot plant regen-
erator produces waste streams which enter the environment, directly. The
35
-------
regenerator off gas stream is composed primarily of SCL, CX^ and N£.
The spent stone is a mixture of CaO, CaS, CaSO,, carbonaceous material
and trace metals in various chemical forms.
SPENT SOLIDS HANDLING SYSTEM
FW Demonstration Plant and 250 MW Unit
the spent solids handling system designed by Foster Wheeler will confine
solids continuously withdrawn from the regenerator and particulate matter
collected by the regenerator off gas cyclones. The combined material will
be cooled to approximately 177 C (350 F) by heat exchange with air in a
fluidized bed cooler. Cooled solids will then be transported by pneumatic
conveyor to a storage silo from which spent material will eventually be
removed in closed dump trucks to disposal sites.
Emissions
Hot air from the fluidized bed cooler will pass through a cyclone before
becoming a waste stream to the atmosphere. This stream will be primarily
nitrogen and oxygen but may also contain CO,,, SO- and lime particulate.
The solid waste stored in the silo will be primarily CaO with small amounts
of CaS, CaSO, and trace quantities of metallic oxides and carbonaceous
material.
TM
FW RESOX OFF GAS TREATMENT SYSTEM
Demonstration Plant and 250 MW Unit
TM
The RESOX system is a proprietary system developed by Foster Wheeler
to reduce S0_ to elemental sulfur. Details of the process are not in the
public domain and thus only a cursory description of the unit operations
involved can be given. Regenerator off gas passes through a cyclone and
36
-------
f\ TM
is then cooled to 650 C (1100 F) before entering the RESOX reactor where
the S0_ in the off gas reacts with anthracite coal (carbon content~92 per-
cent) via the reaction:
Preheated air is fed to the reactor to maintain the temperature at about
760°C (l400°F). Foster-Wheeler estimates that 70 percent of the influent
S0_ is reduced to elemental sulfur. Coal ash from the reactor is quenched
with water and stored for later disposal. Gaseous elemental sulfur formed
in the reactor is condensed and the resultant liquid sulfur is stored for
possible resale. Tail gas exiting from the condenser will be returned to
the gasifier for reaction with limestone.
Emissions
TM
The principal waste stream associated with the RESOX system is coal ash
remaining after reduction in the reactor. Present plans call for storage
of the ash and subsequent disposal. In addition, fugitive dust may be
generated during coal handling and storage.
BOILER
ERCA Pilot Plant
Product gas from the gasifier undergoes combustion ina2.9MW(10x
10 Btu/hr) pressurized water tube boiler. A mechanical draft cooling
tower is used to dissipate cooling water circulating to the condenser.
Boiler flue gas containing roughly 5 percent oxygen exist through a knock-
out baffle and cyclone before entering the boiler stack. About 5 percent
of the boiler flue gas is bled off through a baghouse unit and returned
to the gasifier for temperature control. The remaining Hue gun i:xUn
the top of the stack at a rate of approximately 5.7 m3/s (1200 ft3/min).
37
-------
FW Demonstration Plant
Product gas from the gasifier will undergo combustion in steam generator
Unit No. 4 of the La Palma Power Station. This 20 MW oil and gas fired
unit will be retrofit with two burners designed to handle 7.52 kg/s
(59,660 Ib/hr) of product gas, thus allowing half load firing with product
gas from the 10 MW CAFB or full load firing using both natural gas and
product gas. Unit No. 4 produces up to 31.5 kg/s (250,000 Ib/hr) of steam
at 446°C (835°F) and 4.76 x 106 Pascals (675 psig). Preheated air enters
the boiler at 232°C (450°F) and flue gas enters the boiler stack, after pas-
sing through the air preheater heat exchanger at 191 C (375 F). Flue
gas will be recycled to the gasifier for temperature control at the rate
of 0.66 kg/s (5,250 Ib/hr). The remainder of the flue gas, approximately
34.1 kg/s (270,000 Ib/hr) will exit through the power station stack.
FW 250 MW Unit
A boiler has not yet been selected for the utility retrofit.
Emissions
Boiler flue gas is the primary source of air emissions from all CAFB units
and is treated in detail in Sections IV and V of this report in which
actual measurements of pilot plant stack emissions are discussed and pro-
jections given for the demonstration plant and 250 MW unit. Other sources
of atmospheric and water emissions are the cooling towers and boiler
blowdown and treatment associated with each plant.
38
-------
REFERENCES
1. Craig, J.W.T., G.L. Johnes, G. Moss, J.H. Taylor, and D.E. Tisdall.
Study of Chemically Active Fluid Bed Gasifier for Reduction of
Sulphur Oxide Emissions (Final Report, June 1970 to March 1972).
Esso Research Centre, Abingdon, Berkshire, England. U.S. Environ-
mental Protection Agency, Research Triangle Park, N.C. Report
Number EPA-R2-72-020. June 1972. 334 p.
2. Craig, J.W.T., G.L. Johnes, Z. Kowszun, G. Moss, J.H. Taylor, and
D.E. Tisdall. Chemically Active Fluid-Bed Process for Sulphur
Removal During Gasification of Heavy Fuel Oil — Second Phase.
Esso Research Centre, Abingdon, Berkshire, England. U.S. Environ-
mental Protection Agency, Research Triangle Park, N.C. Report
Number EPA-650/2-74-109. November 1974. 589 p.
3. CAFB Operators Manual. Esso Research Centre, Abingdon, Berkshire,
England. November 1975. 44 p.
4. Chemically Active Fluid Bed Process (CAFB) Preliminary Process
Design Manual. Foster Wheeler Energy Corp., Livingston, N.J.
U.S. Environmental Protection Agency, Research Triangle Park,
N.C. Contract Number 68-02-2106. December 1975. 185 p.
39
-------
SECTION IV
EMISSIONS ESTIMATES
INTRODUCTION
This section and the next probe the chemical and physical properties of
the waste streams identified in the preceding section. The emissions
assessment discussion is divided into two parts: the first, presented here,
contains emissions estimates for waste streams not sampled by GCA; the second
half, described in Section V, consists of a detailed presentation of the
protocols and results of the field test program conducted by GCA at the
ERCA CAFB pilot plant.
The emissions estimates calculated in this section are derived from several
sources:
• CAFB pilot plant process data and log sheets;
• Emissions projections prepared by Foster-Wheeler
for the demonstration and commercial CAFB plants;
• Reports dealing specifically with the CAFB process;
• General literature on process emissions.
Studies by ERCA, ' Westinghouse, »^ Foster Wheeler^ and others have concen-
trated on two areas: stack SO- emissions and sulfate, sulfide and trace
metal concentrations of spent regenerator stone. In addition, these
reports contain detailed discussions of the environmental and economic
acceptability of various options proposed for stone disposal.
40
-------
Reports by EPA contractors dealing with spent stone from fluidized-bed
combustion of coai6»7 provide the basis for stone characterization reported
in this study. These workers have concentrated on sulfur and trace metal
content of stone. To complement these studies GCA collected spent stone
samples and had them analyzed for organic functional groups and surface
elements. These results are presented in Section V.
Although fugitive air emissions from oil, coal, ash and limestone storage
and handling and water emissions from cooling towers and boiler effluent
are not unique to the CAFB, they are discussed here to provide a complete
emissions assessment. Analyses of these emissions are based on general
systems** with some amplification of factors peculiar to the CAFB or to
conditions associated with the San Benito area. In addition, worst case
analyses for flue gas emissions, based upon input material composition
and feed rates tabulated in the next subsection, are presented and com-
pared to legal requirements and known health and ecological effects in-
formation where appropriate.
INPUT MATERIALS
During the latest operation of the ERCA pilot plant, Run No. 10 November-
December 1975, both No. 6 fuel oil (atmospheric bottoms) and bitumen (vacuum
bottoms) were gasified in the CAFB. The fuel oil, from Venezuelan crude,
had been used in previous ERCA runs. The bitumen had not. Chemical and
physical analyses of both fuels are presented in Tables 4 and 5. Table 6
presents an extensive breakdown of elements found in the limestone used
during Run No. 10. The concentrations reported in this table were deter-
mined by ERCA using atomic absorption spectroscopy (AA) and neutron
activation analysis (NAA). In addition, Figure 6 is an ESCA spectrum of
the limestone particulate surface. Surface abundances are also listed in
Table 6.
See Section V.
41
-------
Table 4. ANALYSIS OF FUEL OIL USED BY ERCA
Elements and properties
ca
Ha
sa
NS
va
Nia
Naa
b
CaD
b
Si
b
K
b
Sn
Cd (In)C
Pbb
Znb
Fea
b
Al
CrC
d
Mg
Mn
b
Sb
b
P
b
Mo
b
Cu
Concentration or value
85.3 ± 0.2 %
11.3 ± 0.1 %
2.5 ± 0.01 7o
0.35 ± 0.02 7,
307 ±2.2 ppm
41 ± 2.5 ppm
39 ± 3.3 ppm
26 ± 3.1 ppm
20 ± 3.9 ppm
17 ± 7.8 ppm
17 ± 9 ppm
7 ± 1.4 ppm
3.6 ± 0.2 ppm
2.7 ± 0.5 ppm
2.7 ± 0.33 ppm
2.1 ± 0.6 ppm
1 . 4 ± 0 . 14 ppm
1-10 ppm
0.9 ± 0.18 ppm
0.73 ± 0.31 ppm
0.47 ± 0.18 ppm
0.44 ± 0.36 ppm
0.39 ± 0.13 ppm
42
-------
Table 4 (continued). ANALYSIS OF FUEL OIL
USED BY ERCA
Elements and properties
Concentration or value
As
Rbl
B
Co"
Bab ' -
Srb
Csb
cid
Gab
TeC
Ged
Specific gravity
Conradson Carbon0
a
Asphaltenes
£
Heating value
a
0.3-3 ppm
0.27 ± 0.10 ppm
0.25 ± 0.15 ppm
0.22 ± 0.08 ppm
0.22 ± 0.015 ppm
0.2 ± 0.04 ppm
0.16 ± 0.01 ppm
0.092 ± 0.088 ppm
0.090 ppm
0.06 - 0.6 ppm
0.024 ± 0.009 ppm
<1 ppm
<0.074 ppm
0.958 ± 0.001
10.8 ± 0.3
5.45 ± 0.22 %
10.3 kcal/gm
(18,530 Btu/lb)
Reference 2, p. 539.
By Spark Source Mass Spectrometry (SSMS); per-
formed for PMB/EPA by Northrup Services, Inc.
c
By Neutron Activation Analysis (NAA); performed
for ERCA by the U.K. Atomic Energy Establishment,
Harwell.
By Atomic Absorption (AA) spectroscopy; performed
by ERCA.
From ERCA.
43
-------
Table 5. ANALYSIS OF BITUMEN USED BY ERCA
Elements and properties
Concentration or value
Nia
Phenolics
Aromatics - possibly POM°
Conradson Carbon
a
Specific gravity
Heating value3
Viscosity3
3.75 %
550+50 ppm
74 ppm
Present
Present
9.71
1.0185
9.9 kcal/gm (17,900 Btu/lb)
376.3 cs (135°C)
203.6 cs (150°C)
75.0 cs (175°C)
Private connnunication,, ERCA.
From LC/IR organic functional group analysis (see
Section V) performed by EPA and Battelle Columbus
Laboratories.
Table 6. ANALYSIS OF LIMESTONE
USED BY ERCA3
Element
Cn
Mg"
Sib
Alb
Feb
Srb
Kb
Bab
Clb
Nab
Nlb
Cd or Inc
Mnc
Sbc
Ib '
Pb
Tlb
Tec
Crc
La"
Coc
Vb
Surface 0
Surface Cd
Surface Cad
co3'/cd
Concentration
71.52
0.2 - 2*
600 - 6000 ppm
200 - 2500 ppm
200 - 2000 ppm
100 - 1000 ppm
100 - 1000 ppm
30 - 300 ppm
10 - 100 ppm
10 - 100 ppm
< 50 ppm
29+6 ppm
22+1 ppm
< 10 ppm
1-10 ppm
1-10 ppm
0.6 - 6 ppm
2+0.2 ppm
2+0.4 ppm
0.3-3 ppm
0.3 + 0.01 ppm
0.06 - 0.6 ppm
49.5 X
38.9 Z
11.6 %
0.5
All results except surface ele-
ments and CO */C from ERCA.
u ' 3
Atomic Absorption (AA) spectroscopy.
GNeutron Activation Analysis (NAA) per-
formed the U.K. Atomic Energy Estab-
lishment, Harwell.
Electron Spectroecopy for Chemical
Analysis (ESCA) (see Figure 6).
44
-------
Ln
O
UJ
oc |_
O
O
1000
800
600 400
BINDING ENERGY, eV
200
Figure 6. Limestone feed. Broadband ESCA scan
-------
These data are used throughout this report to make worst case emission
analyses, engineering estimates of emission rates and to normalize the
field program results obtained at the pilot plant. Finally, typical
characteristics of the fuels to be utilized in the FW demonstration plant
are presented in Table 7 to provide a basis for projected emissions from
that facility and from the 250 MW unit.
Table 7.
"TYPICAL" FUEL OIL TO BE USED AT
THE FW DEMONSTRATION PLANT
Elements and
properties
C
H
S
0
N
Moisture
Ash
Specific gravity
Heating value
Concentration
or value
84.43%
10.58%
2.67%
1.68%
0.37%
0.2%
0.07%
0.9765
10.3 kcal/gm
(18,423 Btu/lb)
FUGITIVE AIR EMISSIONS FROM OIL STORAGE AND HANDLING
Fugitive evaporative losses from liquid storage tanks depend on several
factors:
• Vapor pressure of the liquid
• Temperature variations within the tank
• Height of vapor space
• Tank diameter
• Filling and emptying frequency
• Condition and type of tank.
46
-------
For fixed roof storage tanks the largest emissions result from emptying
and filling operations (working losses) and from breathing losses associated
with thermal expansion, pressure fluctuations and continuous vaporization.
General formulas' for estimating both working and breathing losses have
been developed by the American Petroleum Institute. In general, the
breathing losses are one to two orders of magnitude less than working
losses and will not be considered here.
The working loss rate is given by:
' 180 + N
W - 1000 D m P
6N
2
where W = working loss in lb/10 gal throughput
D = oil density in Ib/gal
-4
m = empirical factor estimated to be 1.5 x 10 for residual oil
P = vapor pressure at the bulk oil temperature
N = number of tank refills per year.
For the demonstration plant, the following values are assumed: D = 8.1 lb/
gal; P = 4.6 psia; N = 126 refills/year based on continuous operation.
Thus W = 2.3 lb/103 gallon throughput or 230 lb vapor/tank refill (104 kg
vapor/tank refill).
For the commercial system the short term tank will hold a. 2-day oil supply
and will be refilled every 24 hours during one 8-hour shift. Thus D and P
are the same as above but N = 182.5. The working loss W becomes 1.85 lb/
10 gallon throughput or 827 lb vapor/fillup (376 kg/fillup). Assuming
this working loss is distributed evenly over the 8-hour shift, the fugitive
oil emission rate is equal to 13 kg/s (103 Ib/hr). This emission rate
for the short term tank is equally applicable to fillup from either rail
car or from the long term tank.
47
-------
Similar considerations apply to kerosene tanks and the long-term tank in
the 250 MW unit. However, these tanks will be filled up infrequently.
In Section V, the chemical nature of fugitive emissions from bitumen
storage and handling are discussed in more detail.
TM
FUGITIVE AIR EMISSIONS FROM RESOX COAL STORAGE AND HANDLING
At the demonstration plant crushed coal will be delivered by truck and stored
in a silo. Coal will be transferred by a vibrating feeder and bucket ele-
vator to a feed bin directly over the reactor. The only information avail-
TM
able regarding RESOX coal handling for the 250 MW unit is that front-end
loaders will transport coal from the stock pile to the reactor.
Particulate emissions from coal piles are influenced by wind speed, pile
surface area, coal density, and the prevailing precipitation - evaporation
index. The dust emission factor from coal piles is estimated to be equal
to 0.59 mg/kg-yr (0.00118 lb/ton-yr).8 Wind erosion from stationary coal
piles represents only 1/3 of total particulate emissions from coal storage
Q
and handling;0 therefore this factor is multiplied by 3 to derive the total
emission rate from coal storage, conveying, and feeding.
TM
The annual RESOX coal throughput at the demonstration plant will be
co 7
approximately 1.2 x 10 kg (1.3 x 10 tons) and 1.6 x 10 kg (1.8 x
4
10 tons) at the 250 MW unit. Using these values in conjunction with the
emission factor given above, uncontrolled fugitive dust emission rates
from RESOX™ coal will be 6.7 x 10 mg/s (5.3 x 10"^ Ib/hr) at the demon-
_o
stration plant and 0.89 mg/s (7.1 x 10 Ib/hr) at the commercial facility.
FUGITIVE AIR EMISSIONS FROM LIMESTONE STORAGE AND HANDLING
FW Demonstration Plant
At the demonstration plant fugitive limestone dust will be released by
storage and transport operations. Emission rates for these unit operations
48
-------
are difficult to estimate but calculations based upon published' empirical
rates for rock handling processes may be appropriate. Total uncontrolled
emission rates due to screening, conveying and handling are estimated to
be 5 g/kg (10 Ib/ton). Although no figures are given for the percentage
of this total which falls into the suspended particulate range, approxi-
mately 50 percent of the uncontrolled losses from rock crushing settle
out in the immediate vicinity of that operation. Applying this 50 percent
factor to the above emission rate sets this at 2.5 g/kg (5 Ib/ton). In
addition, Foster-Wheeler plans to incorporate a filter over the limestone
surge bin. This control device would remove approximately 99 percent of
the fugitive dust,^ lowering the controlled emission rate to 25 rag/kg
(0.05 Ib/ton). Applying this factor to the limestone feed rate of
0.123 kg/s (975 Ib/hr) yields an emission rate of 3 mg/s (0.024 Ib/hr).
Additional fugitive dust emissions from limestone storage should be negli-
gible by comparison. 'There is no drying unit designed for the 10 MW Demo
and, therefore, no related fugitive emissions.
FW 250 MW Unit
The proposed design for the 250 MW commercial unit calls for crushing and
drying of limestone in addition to handling and storage. Estimates of
emission factors for these operations can be obtained from AP-42' factors
for lime manufacturing. This publication indicates that primary and
secondary crushing operations generate particulate emissions of 15.5 g/kg
(31 Ib/ton) and 1 g/kg (2 Ib/ton) respectively. Because the analyses
presented here reflect worst case situations, it will be assumed that the
factor for primary crushing is applicable. The preliminary FW design
does not include a baghouse over the crushing unit although such a control
device is indicated as an adjunct to the dryer. A baghouse filter would
reduce crushing emissions by 99 percent to 0.16 g/kg (0.31 Ib/ton). There-
fore, at a limestone feed rate of 1.6 kg/s (12,600 Ib/hr) a worst case
analysis predicts uncontrolled crushing emissions will be 25 g/s (200 Ib/hr)
and controlled emissions will be 0.25 g/s (2 Ib/hr).
49
-------
No specific data are available regarding air emissions from limestone
drying. For a worst ease analysis comparative emissions expected from
calcining operations may be illustrative of the order of magnitude involved.
The emission factor given by AP-42^ for rotary kiln calcining is 100 g/kg
(200 Ib/ton). With use of a baghouse filter, as proposed by FW for the
250 MW unit these emissions will be reduced by about 99 percent to 1 g/kg
(2 Ib/ton). At a feed rate of 1.6 kg/s (12,600 Ib/hr) a worst case
analysis of limestone drying predicts an emission rate of 1.6 g/s
(12.6 Ib/hr).
Additional emissions due to limestone screening, conveying and handling
if unabated by a control system would be roughly 4 g/s (31.5 Ib/hr) or
0.04 g/s (0.32 Ib/hr) if covered by a baghouse unit. This is estimated
from applying the emission factor derived for the 10 MW Demo to the lime-
stone feed rate stipulated for the 250 MW plant.
Therefore, total fugitive air emissions at the 250 MW unit resulting from
limestone storage, handling and drying operations will fall in the range
of 1.9 g/s (13 Ib/hr) to 30.6 g/s (244 Ib/hr).
TRACE ELEMENT EMISSIONS
Trace element emissions from the fuel oil combustion can present environ-
mental impacts by several pathways:
• Enrichment - Toxic elements (e.g., Pb, V) can volatize
and selectively condense on small particulates in the
combustion process. These enriched fine particulates
are doubly problematical in that they are difficult to
control at the stack exit and once released, they can
readily penetrate deeply into the lung.
• Vaporization - Some toxic compounds are sufficiently
volatile to be emitted from the combustor in the gas
phase (e.g., Hg, F, Se).
• Formation of carcinogenic compounds - Compounds of cer-
tain trace elements (Cr, Ni) are carcinogenic. These
emissions are of particular concern because quantitative
correlations between ambient concentrations of these
species and health effects have not been established.
50
-------
The unique feature of the CAFB limiting trace element emissions is that
2
the limestone bed acts as a sink for these species (e.g., V, Ni, and Fe).
In this fashion the gasifier itself functions as a control device for trace
element emissions.
There are very few analyses available of the trace element content of
limestone and residual oil or bitumen. The trace element content of
residual oil may vary greatly depending on its origin. The analyses
presented in Tables 4 and 5 for trace element composition of residual
oil and limestone will be used for the estimates calculated here.
Table 8 lists "those elements found in oil or stone which are either
volatile or toxic.
Table 8. VOLATILE OR TOXIC
TRACE ELEMENTS IN
OIL AND STONE
Cadmium
Cobalt
Arsenic
Lead
Scandium
Tellurium
Iron
Vanadium
Zinc
Ant imony
Chromium
Copper
Fluorine
Nickel
The major source of trace element emissions through the stack is feed oil
rather than limestone for two reasons:
• The oil/limestone feed ratio is greater than 10 to 1.
Of the trace elements listed in.Table 8 only iron
is an order of magnitude more abundant in limestone
than in oil;
• Trace elements in the sorbent are contained in a
limestone matrix as the fairly unreactive oxide
or carbonate (see Table 6); thus they will have much
lower emission factors than the more volatile forms
of trace elements (such as sulfides) encountered
on the fuel.
51
-------
To determine possible environmental impacts of trace element emissions
a worst case analysis can be made, assuming that all trace elements
in the fuel feed exit through the stack. If these emission rates.can be
shown to produce negligible environmental impacts, then trace element
emissions will not be of concern in the CAFB.
Emission factors at the top of the stack for those elements called out in
Table 8 are tested in Table 9. It has been estimated1^ that ground level
ambient concentrations in the vicinity of the stack are on the order of
0.1 percent of those at the top of the stack. These ground level values
are also listed in Table 9. To judge the potential environmental impact
of these trace element emissions these ambient loadings should be com-
pared with maximum acceptable ambient air concentrations or Multi-Media
Environmental Goals (MEGS) established by EPA.11 These factors are deter-
mined from Threhold Limit Values (TLV's)12 set by OSHA by the following
formula.
MEG = (8/24)(0.01) TLV
The factor 8/24 adjusts the 8-hour workday OSHA standard to 24-hour ex-
posure, and the factor 0.01 provides a margin of safety for those people
who are less healthy than the average industrial worker. Both TLV's and
MEGS are listed in Table 9. For a given element to be of potential concern
its ambient concentration must exceed its MEG.
Applying this criterion, vanadium, cadmium and nickel are the only trace
elements whose worst case emission rates may be of concern. Previous ERCA
studies, however, have shown that almost all fuel vanadium and three-
quarters of the nickel are picked up by the gasifier bed material. In the
ERCA analysis of fuel oil (Table 4) cadmium could not be distinguished
from indium. Thus it is not at all clear that cadmium is present in the
oil to any significant extent. Although the worst case analyses make no
assumption about the physical form of trace elements exiting the stack,
most of these elements will in fact condense on particulate surfaces as
52
-------
Table 9. COMPARISON OF WORST CASE EMISSION ESTIMATES WITH AIR
QUALITY GOALS
Element
As
Cd
Co
Cr
Cu
F
Fe
Ni
Pb
Sb
Te
V
Zn
Concentration at
top of stack,
mg/m3
0.19
0.45
0.013
0.090
0.025
0.014
2.0
2.62
0.23
0.047
0.064
19.65
0.17
Ambient concentration,
yg/m3
0.19
0.45
0.013
0.090
0.025
0.014
2.0
2.62
0.23
0.047
0.064
19.65
0.17
TLV
mg/m3
0.5
0.05
0.1
1.0
1.0
2.0
1.5
1.0
0.15
0.5
0.1
0.5
5.0
MEG
yg/m3
1.7
0.17
0.33
3.3
3.3
6.7
5.0
3.3
0.5
1.7
0.33
1.7
16.7
the stack gas cools. Particulate chemical composition is discussed in
detail in Section V.
WATER EMISSIONS FROM RESOX™ COAL STORAGE
Surface run-off from natural precipitation constitutes the primary source
of potential contamination of surface waters due to coal storage. The
pollution potential of coal pile runoff depends upon local precipitation,
pile area, storage foundation material and storage pile coating. Coal pile
runoff usually has a low pH and a high concentration of dissolved solids
including iron, magnesium, and' sulfate. Aluminum, sodium, manganese,
and other metals may also be present in undesirable amounts. Coal pile
drainage contains dissolved metallic salts in the concentration range
shown in Table 10. The variability of drainage composition reflects the
53
-------
Table 10. COMPOSITION OF DRAINAGE FROM
COAL PILES13
Alkalinity
BOD
COD
Total solids
Total suspended solids
Total dissolved solids
Ammonia
Nitrate
Phosphorus
Turbidity
Acidity
Total hardness
Sulfate
Chloride
Aluminum
Chromium
Copper
Iron
Magnesium
Sodium
PH
Concentration, mg/la
15
3
100
1,500
20
700
0.4
0.3
0.2
6
10
130
130
20
825
0
1.6
0.4
90
160
2.2
- 80
- 10
- 1,000
- 45,000
- 3,300
- 44,000
- 1.8
- 2.3
- 1.2
- 505
- 27,800
- 1,850
- 20,000
- 480
- 1,200
- 16
- 3.9
- 2.0
- 180
- 1,260
- 8.0
Appropriate for all values except pH.
54
-------
variety of coals used as well as the rate of rainfall. No specific data
TM
was found for anthracite, the type of coal to be used for the RESOX at
the demonstration plant. During heavy rainfall the level of dissolved solids
in the runoff will be high initially and will rapidly decrease. When
rainfall is light, the long retention time may allow more diffusion, and
hence more chemical reaction to occur and result in higher pollutant
concentrations. Meteorological conditions in the San Benito area produce
relatively brief periods of heavy rainfall during the late summer and
early fall and little other precipitation.
Accurate assessment of the runoff associated with the 250 MW unit must
await site selection and specifics of the coal storage pile. An order of
magnitude estimate of this emission can be made from general correlation
i ^
data.1-3 Assuming a 30-day supply of coal is kept on hand, the storage
pile will hold up to 1.4 x 10 kg (3 x 10 Ib). This corresponds to a
3 43
volume of roughly 850 m (3 x 10 ft ) which will be assumed to be contained
2 2
in a pile of area 186 m (2,000 ft ) and height 4.6 m (15 ft). At an
annual rainfall of 114 cm (45 in.) the yearly runoff would be 212 m3
(7500 ft3).
TM
EMISSIONS FROM RESOX SOLID WASTE
Spent fuel effluent from the RESOX system amounts to 0.02 kg/s (150 Ib/hr)
at the demonstration plant and 0.25 kg/s (2,000 Ib/hr) at the 250 MW unit.
Foster Wheeler plans to market this material which is approximately
75 percent carbon and 25 percent ash as a low sulfur solid fuel with a
heating value of about 5800 kcal/kg (10,500 Btu/lb). If this material is
not marketable a number of disposal possibilities including ponding and
landfill will have to be considered. Air, water and leachate emissions
from these options should be carefully evaluated if such disposal will be
required.
TM
The other solid product from the RESOX unit operations is sulfur.
Foster Wheeler also plans to market this material. Nevertheless, sulfur
55
-------
processing and handling operations must be evaluated for their environ-
mental impacts.
EMISSIONS ASSOCIATED WITH SPENT REGENERATOR STONE
Solid waste will be emitted from the regenerator in both Foster Wheeler
designs and will be transported to the spent solids handling system where
the solids will be cooled with air and pneumatically transported to a
spent solids storage silo. The cooler air exhaust is vented to a cyclone
and collected solids are sent to the storage silo. To reduce air emis-
sions, the storage silo exhaust will pass through vent filters.
Foster Wheeler has considered the prospect of marketing the spent solid
material. If marketing is not possible, the waste material must be dis-
posed of in an environmentally acceptable manner. As is the case with
sulfur, unit operations associated with marketing must be carried out in
an environmentally acceptable manner.
Spent stone from the CAFB cannot be disposed of as a solid land-
fill in an environmentally acceptable manner without further treatment.
The stone consists of from 3 to 5 percent CaS which will react with
moisture in the air to liberate H2S. The H2S will be oxidized in the
atmosphere to S02. This S02 will add to the S02 emissions from the CAFB
unit and the whole system could exceed federal SO- standards. For example,
Westinghouse has determined that if 90 percent of the fuel sulfur is re-
tained in the bed and 70 percent of the waste sulfide is converted to
sulfate, then the total emissions from the CAFB and waste disposal pile
would exceed the current federal SO emission standard (0.8 Ibs SO./10 Btu)
3
after 12 years assuming a 6 percent sulfur annual loss rate. Clearly,
the waste stone must be treated to remove the sulfide or render it inert.
56
-------
4
Westinghouse has been investigating several methods for spent stone
processing prior to disposal. These methods include:
• Dry sulfation - reacting stone with S02 and 02 at 870°C
(1600°F) to produce a product containing 90 percent
CaS04 and 10 percent CaO.
• Missing stone with coal fly ash and hot pressing.
• Wet slurrying with carbonation - reacting spent lime with
water and C02 to produce CaCOs and H2S.
Three possible disposal options are also being considered: sale of processed
stone, land filling and ocean dumping. As yet no combination of processing
and disposal has been shown to be environmentally acceptable.
EMISSIONS AND ENVIRONMENTAL EFFECTS OF CONDENSER COOLING
The La Palma power station condensers are cooled by six mechanical draft
cooling towers. *•* These cooling towers are visible in Figure 7 which
is an aerial photograph of the power plant* The FPC Form 67 data stipu-
lates a cooling water recirculation rate of 10.8 m^/s (380 ft^/s) in
order to service the entire 230 MW of plant capacity.14 xhe use of the
10 MW CAFB demonstration plant should have negligible effect on the over-
all quantity and characteristics of cooling water withdrawal, recirculation,
and discharge. A summary of potential environmental impacts produced by
the La Palma cooling towers are presented in Table 11.15
thermal Discharge
3 3
Makeup water is required at a rate of 0.09 m /s (3.15 ft /s) and discharged
oo1 3
at 0.025 m /s (0.9 ft /s), reflecting an evaporative water loss of 0.06 m /s
(2.25 ft /s). The cooling water experiences a temperature rise of 9°C
(16°F) as it circulates past the condensers. Thermal discharge to the water
environment will depend upon whether blowdown is performed at the cold
side or hot side of the cooling system. If blowdown is done on the hot
side, a conservative estimate of heat rejection to the ambient water is
10 percent of the heat content of the recirculating cooling water.13 At
57
-------
Figure 7. Aerial photograph of the La Palma Power Station
(from Foster Wheeler)
58
-------
Table 11. SUMMARY OF POTENTIAL ENVIRONMENTAL IMPACTS FROM THE
LA PALMA STATION COOLING TOWERS1^
Atmospheric effects
Hydrologic and aquatic effects
Other effects
Visible plume:
visual obstruction, ground
shading, and reduction in
visibility to various modes
of transportation.
Ground fog:
potential hazard to land and
water transportation and
nuisance to nearby communi-
ties.
Icing:
hazard to land transportation
and ice accumulation on
nearby structures and utility
wires.
Drift deposition:
potential damage to biota,
acceleration of corrosion of
nearby structures, and con-
tamination of soil and water
bodies.
Cloud formation:
visual obstruction and .poten-
tial local weather modifica-
tions.
Precipitation and snow
augmentation:
potential local weather modi-
fications.
Blowdown:
potential increase of water
temperature near discharge
point, contamination of
surface-water and ground-
water supplies, potential
increase of soil salinity.
Water consumption:
potential depletion of
surface-water and ground-
water resources.
Seepage and leakage water:
same effects as blowdown
discharges.
Intake screen devices:
impingement or entrapment
of aquatic life.
Transport through condensers
and circulation pumps:
damage to aquatic organisms,
Discharge systems:
disturbance to aquatic
communities due to mechan-
ical forces and turbulence.
Land use.:
large land areas
required for each of
the cooling systems.
Sound levels:
nuisance to nearby
residents and tran-
sient observers.
Aesthetics:
unsightly to nearby
residents and tran-
sient observers.
59
-------
a AT of 9°C (16°F) and a discharge rate of 0.025 m3/s (0.9 ft3/s), the heat
discharge to the ambient water is equal to 23 kcal/s (7.8 x 106 Btu/day).
The effect of heat discharge to ambient water is the reduction in the
dissolved oxygen concentration, which can cause migration of aquatic
species, fish kills, and a reduction in the capacity for natural stream
purification.
Cooling Tower Slowdown Wastewater Discharge^
Federal regulations require that pollutants discharged in cooling tower
blowdown will not exceed the concentrations noted in Table 12.
A recent EPA document*° requires even more stringent limitations on
effluent residual chlorine discharged into fresh water. Table 13
illustrates these specifications. The allowable residual chlorine con-
centration thus depends upon whether the cooling water is discharged to
the tidal estuary portion of the Rio Grande or to the Gulf of Mexico.
As cooling water evaporates, all dissolved and suspended solids are con-
centrated in the cooling stream. The solubility of the constituents at
specific temperature and pH limits the degree of concentration. Preci-
pitation of solids onto metal surfaces can occur and is prevented by
injecting chemical additives for control of scale, corrosion, and algae,
slime, and fungi buildup. Tables 14*° and 15^ illustrate the type and
concentration of chemicals mixed into cooling tower water.
The FPC Form 67 summary1^ for the San Benito Plant states that 3 tons/yr
of chlorine are added to the circulating cooling water in order to control
the fouling of metal surfaces with microorganism growth. Disregarding
any chlorine reaction results in a residual chlorine discharge of 3.4 mg/1,
which is an order of magnitude higher than the limits noted in Table 12.
This is an extremely conservative estimate based solely on a worst case
analysis and actual free chlorine discharge will be much lower than
60
-------
Table 12! WATER EFFLUENT STANDARDS
Pollutant
Free available chlorine
Zinc
Chromium
Phosphate
Other corrosion
inhibiting materials
1-day
maximum
concentration,
mg/1
0.5
1.0
0.2
5.0
30-day
average
concentration,
mg/1
0.2
1.0
0.2
5.0
Limit to be established
on a case by case basis
Table 13. RESIDUAL CHLORINE RECOMMENDATIONS
17
Type of chlorine use
Residual chlorine
concentration, mg/1
Degree of protection
Continuous
Intermittent — 2 hrs/day
<0.01
<0.002
<0.2
<0.04
Protects trout and salmon
and other important fish
food organisms. Poten-
tially lethal to more
sensitive species.
Protects most aquatic
organisms.
Protects trout and salmon.
Protects most aquatic
organisms.
61
-------
Table 14. CHEMICALS USED IN RECIRCULATIVE
COOLING WATER SYSTEMS18
Use
Corrosion Inhibition or scale
prevention In cooling towers
Blocides In cooling towers
pH control In cooling towers
Dispersing agents In
cooling towers
Biocides In condenser cooling
water systems
Chemical
Organic phosphates
Sodium phosphate
Chromates
Zinc salts
Synthetic organlcs
Chlorine
Hydrochlorouo acid
Sodium hypochlorlte
Calcium hypochlorlte
Organic chromates
Organic zinc compounds
Chlorophenates
Thlocyanates
Organic sulfurn
Sulfurlc acid
Hydrochloric acid
Lignlne
Tannins
Polyacrylonltrtle
Polyacrylamlde
Polyacryllc acids
Polyacryllc acid salts
Chlorine
Hypochlorltes
Sodium pentachlorophenate
Table 15. COOLING TOWER CORROSION AND
SCALE INHIBITOR SYSTEMS13
Inhibitor system
Concentration of chemical
additives in recirculating
water, mg/1
1. Chromate
2. Chromate + Zinc
3. Chromate + Zinc +•
Phosphate (Inorganic)
4. Zinc + Phosphate
(Inorganic)
5. Phosphate (inorganic)
6. Phosphate (organic)
7. Organic bloclde
200 - 500 mg/1 Cr04
17 - 65 mg/1
8 - 35 mg/1
10 - 15 mg/1
8 - 35 mg/1 Zn++
30 - 45 n.g/1 P04E
8-35 mg/1 ZnW-
15 - 60 mg/1 P0/+s
15 - 60 mg/1 P04B
15-60 mg/1 P04S
3-10 mg/1 organic8
30 mg/1 chlorophenol
5 mg/1 sulfone
1 mg/1 thiocyanate
62
-------
specified limits because most of the added chlorine will be chemically
bound to other species contained in the cooling water.
Dissolved species in, cooling water may be naturally occurring or introduced
as corrosion.inhibitors, biocides, pH controls, and dispersants. When
the concentration of these ions exceeds solubility limits, salt will
precipitate. .The solubility of some salts decreases when the temperature
rises. Salts exhibiting this characteristic are likely to precipitate
and form scale on hot condenser tube walls and reduce heat transfer. The
most common way to control scale formation is to blowdown a portion of the
circulating water stream and replace it with fresh water so that the ion
concentration in the circulating water does not reach saturation at any
time. Blowdown (B) is a function of cooling water makeup quality. As
shown below, the volume of cooling water makeup (M) required is equal to
the sum of the volume of cooling water lost as blowdown (B), drift (D),
evaporation (E), and seepage or leakage (S). S is very small in comparison
to the other volume parameters and can be neglected without significantly
affecting calculated volumes.
M-B+D + E+S
It follows that the volume of blowdown is a function of makeup water
quality and can be determined from the following expression.
E - (S 4- D)(C - 1)
B ~ C - 1
13
where C = cycles of concentration (dimensionless).
Cycles of concentration is the number of times that the solute species
can be concentrated before one particular constituent concentration ex-
ceeds a critical level. C can be increased as influent water quality
increases. This qualitatively illustrates the degree to which influent
63
-------
water quality can degrade prior to falling below acceptable levels. The
equation shows that for a constant rate of evaporation, drift and seepage
the required blowdown decreases as C increases. The equation represents
a tradeoff between external feedwater treatment and internal chemical
conditioning needs.
For average quality cooling water makeup, the value for C is conventionally
kept between 4 and 6. For extremely high quality cooling water makeup,
C values of 15 and above may be employed. When saline cooling water is
used, C generally ranges between 1.2 and 1.5.^
Cooling Tower Drift
Warm moist air discharged from cooling towers contains water droplets
which range in diameter from a few to several hundred micrometers. Those
droplets greater than 20 urn in size are considered as drift and smaller
droplets constitute fog. Whereas fog is relatively pure condensed water
vapor, drift droplets contain the same concentration of dissolved chemicals
as the circulating cooling water.19,20
Cooling tower characteristics which affect drift rates1^ include:
• Volume of circulating water in the system per unit time
• Tower features (height, diameter, and characteristics of
drift eliminators for natural-draft tower; height, cell
diameter, characteristics of drift eliminators, and
number of cells for mechanical draft tower)
• Drift flux and droplet size distribution
• Exit temperature
• Efflux velocity
Smaller size water droplets remain in the cooling tower plume for longer
time periods than larger heavier droplets. As the heavier droplets fall
64
-------
out they are affected by atmospheric turbulence. Atmospheric character-
istics which affect drift deposition include:^
• Ambient temperature,
• Relative humidity,
• Atmospheric stability,
• Mixing layer depth,
• Wind speed and direction, and
18
• Precipitation.
Cooling tower drift losses vary between 0.005 and 0.02 percent of the
cooling tower circulation rate.13 This amounts to 6 to 50 x 10" m /s
(1.2 to 4.8 cfm) drift discharged to the atmosphere from the mechanical
draft cooling towers in use at the La Palma Power Station.
Fogging
Plumes from cooling towers have the potential to produce conditions of
fogging and icing. Normally the plume will mix with the ambient air and
not inhibit visibility. However, during thermal inversions and periods
of high humidity and low temperature, the plume can become bounded close
to the ground surface and cause fogging. Fogging is generally limited to
'the cooling tower site (within ~600m (2000 feet) of the tower). The pro-
bability of occurrence is higher with mechanical draft than natural
draft cooling towers.^'
Water Consumption
The mechanical draft cooling towers in use at the La Palma station cool
•*
primarily by latent heat transfer; only about 25 percent of heat loss
•A
is through sensible heat transfer.^ The FPC reports that 0.064 m Is
o
(2.25 ft /s) of water is evaporated by the La Palma cooling towers.^
65
-------
EMISSIONS FROM BOILER WATER TREATMENT AND BOILER SLOWDOWN
Boilers in steam-electric power plants require that makeup water be added
to steam condensate return in order to compensate for recirculating water
lost during boiler blowdown, steam soot blowing, venting, gland and
boiler tube leakage. The required quantity and quality of feedwater is a
function of boiler operating pressure and heat transfer rate. It is not
anticipated that feedwater requirements and water emissions will change
after the 10 MW CAFB is retrofitted to Unit No. 4 at the La Palma Power
Plant. '
The La Palma Plant uses fixed bed demineralizers for treatment of feed-
91'
water used in boiler units 4 and 6. x This is an ion exchange process
in which undesirable ions such as calcium and magnesium react with a
polymeric resin and are removed from the feedwater. Both positive and
negative ions are removed by cation and anion exchange resins. Cation
resins are generally synthetic polymeric materials containing ion groups
such as SO_H . Common anion exchange resins are synthetic amines.22
A typical cation exchange reaction is:
2H+
where R represents the cation exchange resin.
When the exchange resin's capacity for collecting more cations is exhausted,
it is regenerated by passing a 2 to 10 percent ILSO, solution through the
bed; i.e.,
p-. D i Oil **» TJ .D -L Pa
v»o. • IV T £11 m_ n.n l\ "t" \jci
66
-------
INLET
METER
t
XJ-
WASHWATER COLLECTOR
-M-
BACKWASH INLET
OUTLET
BACKWASH
OUTLET
M
-M-
RINSE
OUTLET
ION-EXCHANGE
UNIT —<
EXCHANGE
MATERIAL
RINSE
WATER
DISCHARGE
SUPPORTING
BED
REGENERANT
TANK
Figure 8. Fixed bed ion exchange system
23
Anion exchange replaces undesirable anions with hydroxide ions according to:
SO
I, + R-(OH)2^±R'S04 +
20H
Regeneration is accomplished by passing 5 to 10 percent solution of sodium
hydroxide through the bed:
R-S04 + 20H~^R-(OH)2 + SC>4=
Actual treatment involves a number of steps. The feedwater is passed
through the resin bed until an excess of contaminant appears in the
effluent. Following such breakthrough, the bed is backwashed and the
resin regenerated and rinsed. The bed is then ready for another treatment
cycle.22 Figure 8 schematically illustrates in ion exchange treatment
unit.23
67
-------
Backwashing is performed after breakthrough for a period of about 10 min-
utes at a flow rate of 3.4 to 4.8 liters per second per square meter (5 to
7 gallons per minute per square foot) of bed area. This step removes any
accumulated dirt and loosens the resin to prevent flow channeling during
subsequent treatment cycles. After regeneration, excess regeneration
solution and spent solution is rinsed from the bed. The total volume of
rinse water required is approximately 3.34 x 103 liters per m3 (25 gallons
per ft3) of bed volume. The waste materials carried in the rinse water
are primarily sodium, calcium and magnesium chlorides or sulfates, plus
i ^
excess sulfuric acid or alkali (NaOH) used for regeneration.0
Boiler Slowdown
In order to maintain dissolved and suspended solids below specified levels,
a portion of the circulating boiler water is periodically or continuously
discharged from the system. If solids are allowed to accumulate they
will eventually precipitate onto heat transfer surfaces and cause ef-
ficiency and structural integrity to deteriorate.
Pollutants discharged with boiler blowdown include suspended and dissolved
solids, hardness, phosphates, and alkalinity. Total dissolved solids
content ranges between 10 and 100 mg/1. At La Palma, hydrazine is added
to condensate return for corrosion prevention and it is estimated that
blowdown pH ranges between 9.5 and 11 and contains ammonia at a concen-
tration of 1 to 2 mg/1.13
68
-------
REFERENCES
1. Craig, J. W. T., G. L. Johnes, G. Moss, J. H. Taylor, and D. E. Tisdall.
Study of Chemically Active Fluid Bed Gasifier for Reduction of
Sulphur Oxide Emissions. Final Report, June 1970 to March 1972.
Esso Research Centre, Abingdon, Berkshire, England. U.S. Environmental
Protection Agency, Research Triangle Park, N.C. Report Number EPA-R2-
72-020. June 1972. 334 p.
2. Craig, J. W. T., G. L. Johnes, Z. Kowszun, G. Moss, J. H. Taylor, and
D. E. Tisdall. Chemically Active Fluid-Bed Process for Sulphur
Removal During Gasification of Heavy Fuel Oil - Second Phase. Esso
Research Centre, Abingdon, Berkshire, England. U.S. Environmental
Protection Agency, Research Triangle Park, N.C. Report Number EPA-
650/2-74-109. November 1974. 589 p.
3. Keairns, D. L., D. H. Archer, R. A. Newby, E. P. O'Neill, and E. J.
Vidt. Evaluation of the Fluidized-Bed Combustion Process. Volume IV -
Fluidized-Bed Oil Gasification/Desulfurization. Westinghouse Research
Laboratories, Pittsburgh, Pa. U.S. Environmental Protection Agency,
Research Triangle Park, N.C. Report Number EPA-650/2-73-048d.
December 1973. 328 p.
4. Keairns, D. L., R. A. Newby, E. J. Vidt, E. P. O'Neill, C. H. Peterson,
C. C. Sun, C. D. Buscaglia and D. H. Archer. Fluidized Bed Combustion
Process Evaluation. (Phase 1 - Residual Oil Gasification/Desulfuriza-
tion Demonstration at Atmospheric Pressure) Volumes I and II. Westing-
house Research Laboratories, Pittsburgh, Pa. U.S. Environmental Pro-
tection Agency, Research Triangle Park, N.C. Report Number EPA-650/
2-75-027a, b. March 1975. 578 p.
5. Chemically Active Fluid Bed Process (CAFB) Preliminary Process Design
Manual. Foster-Wheeler Energy Corporation, Livingston, N.J. U.S.
Environmental Protection Agency, Research Triangle Park, N.C. EPA
Contract Number 68-02-2106. December 1975. 185 p.
6. Murthy, K. S., H. Nack, E. H. Hall, H. R. Hazard, K. D. Kiang, P. S. K.
Choi, and G. R. Smithson, Jr. Engineering Analysis of the Fluidized-
Bed Combustion of Coal. Final Report. Battelle Columbus Laboratories,
Columbus, Ohio. U.S. Environmental Protection Agency, Research
Triangle Park, N.C. EPA Contract Number 68-02-1323, Task Order
Number 6. May 1974. 253 p.
7. Fennelly, P. F., D. F. Durocher, H. Klemm, and R. R. Hall. Preliminary
Environmental Assessment of Coal-Fired Fluidized Bed Combustion. GCA
Corporation, GCA/Technology Division, Bedford, Massachusetts. U.S.
Environmental Protection Agency, Research Triangle Park, N.C. EPA
Contract Number 68-02-1316, Task Order Number 13. May 1976. 125 p.
69
-------
8. Surprenant, N. F., R. Hall, S. Slater, T. Suza, M. Sussman, and
C. Young. Preliminary Emissions' Assessment of Conventional Stationary
Combustion Systems. Volume II - Final Report. GCA Corporation, GCA/
Technology Division, Bedford, Massachusetts. U.S. Environmental Pro-
tection Agency, Research Triangle Park, N.C. Report Number EPA 600/
2-76-046b. March 1976. 531 p.
9. Compilation of Air Pollutant Emission Factors. Second Edition. U.S.
Environmental Protection Agency, Research Triangle Park, N.C. Publica-
tion Number AP-42. March 1975.
10. Cowheard, C., M. Marcus, C. M. Guenther, and J. L. Spigarelli. _>. . .
Hazardous Emission Characterizations of Utility Boilers. U.S. Environ-
mental Protection Agency, Publications Number EPA-650/2-75-066.
July 1975.
11. Eimutuc, E. C., et al. Source Assessment, Prioritization of Stationary
Air Pollution Sources. Model Description. Monsanto Research Corpora-
tion, St. Louis, Missouri. U.S. Environmental Protection Agency,
Research Triangle Park, N.C. Report Number EPA 600/2-76-032a.
February 1976.
12, American Conference of Governmental Industrial Hygienists. Threshold
Limit Values for Chemical Substances and Physical AGents in the Work-
room Environment With Intended Changes for 1974. Copyright 1974.
13. Development Document for Effluent Limitations Guidelines and New
Sources Performance Standards for the Steam Electric Power Generating
Point Source Category. U.S. Environmental Protection Agency,
Washington, D.C. Report Number EPA 400/1-74-029-a, Group I. October
1974. 840 p.
14. Steam-Electric Plant Air and Water Quality Control Data for the Year
Ended December 31, 1972. Based on FPC Form Number 67, Summary
Report. Federal Power Commission. March 1975.
15. Roffman, A. Environmental, Economic, and Social Considerations in
Selecting a Cooling System for a Steam Electric Generating Plant.
Published in Cooling Tower Environment - 1974 by the U.S. Energy
Research and Development Administration. 1975.
16. Steam Electric Power Generating Point Source Category - Effluent
Guidelines and Standards. Env. Reporter. 135:0541. July 11, 1975.
17. Reviewing Environmental Impact Statements - Power Plant Cooling
Systems, Engineering Aspects. Environ Prot Technol Ser. U.S.
Environmental Protection Agency, EPA Report Number 660/2-73-016.
October 1973.
18. Aynsley, E. and M. R, Jackson. Industrial Waste Studies - Steam
Generating Plants. Draft Final Report. U.S. Environmental Protection
Agency. Contract Number EPA-WQO, Number 68-01-0032. May 1971.
70
-------
19. Roffman, A. and R. E. Grimble. Drift Deposition Rates From Wet
Cooling Systems. Published in Cooling Tower Environment - 1974
by the U.S. Energy Research and Development Administration. 1975.
20. The State-of-the-Art of Measuring and Predicting Cooling Tower Drift
and Its Deposition. J Air Pollut Control .Assoc. 24/9):855-859.
21. McMillan, R. Personal communication. Foster Wheeler Energy Corpora-
tion. April 28, 1976.
22. Eckenfelder, W. W. Industrial Water Pollution Control. New York,
McGraw Hill-Book Company, 1966. p. 110-117.
23. Strauss, S.: D. Water Treatment. Power. June 1973.
71
-------
. SECTION V
FIELD TEST PROGRAM AND LABORATORY RESULTS
INTRODUCTION
The field test program carried out by GCA at the ERCA CAFB pilot plant
during the period November 24 to December 11, 1975, was directed primarily
toward flue gas and particulate emissions and secondarily toward solid
waste effluents. The goal of this effort was to characterize as completely
as possible, within the economic and time constraints of the project,
the physical and chemical properties of the emissions from the boiler
stack. Specific details of the field test program evolved during the
course of the testing, being largely dependent on the operating param-
eters of the pilot plant.
Coincident with the planning and pre-test site visit was the announcement
of the "multilevel phased approach" to source sampling and analysis
by the Process Measurements Branch of IERL. Because this project is
a preliminary environmental assessment of a facility which heretofore
had not been subjected to a comprehensive emissions assessment a decision
was made early on in the project to combine measurements of the "criteria
pollutants" using the standard EPA methods with the Level 1 approach to
determination of organic and inorganic emissions. The rationale for
this approach is that the CAFB demonstration plant and proposed commercial
unit will have to meet local, state and federal emission standards and
new source performance standards for particulate, NO , SO , and CO.
X X
Furthermore, because this study has as one of its goals the generation
of recommendations for more comprehensive testing of the Foster-Wheeler
72
-------
demonstration plant currently being designed, that extensive Level 1
information for a number of operating conditions would be more valuable
than detailed data on a limited number of specific pollutants.
The GCA field team collected flue gas samples during normal gasification
of bitumen and fuel oil, as well as during startup and abnormal (clogged
gasifier/regenerator stone transfer system) operation. Actual sampling
was accomplished during seven different pilot plant runs. In addition
to stack gas samples, the GCA team collected spent regenerator stone,
leached stone, gasifier bed stone, cyclone fines and fuel and limestone
feed samples for subsequent laboratory analysis.
The following subsections detail the sampling and analytical techniques
employed in the field and in the laboratory, the pilot plant operating
conditions and the results of the test program.
FIELD SAMPLING PROTOCOL
As shown schematically in Figure 8 flue gas leaves the boiler via a
68.6 cm (2.25 ft) diameter duct from which approximately 5 percent of the
flow is diverted through an experimental baghouse. The remaining flue
gas encounters a knockout baffle and cyclone and then exhausts through
a 68.6 cm (2.25 ft) diameter stack. Three ports (labeled A in Figure 9)
spaced 120 degrees apart are located approximately six diameters upstream
from the flue gas entry into the stack. Figure 10 is a photograph of
the stack showing the locations of the sampling ports, the cyclone and
the knockout baffle. Two of the stack ports are 3-inch BSP and one is
2-inch BSP. Figure 11 is a closeup picture of two of the three ports.
Installation of a fourth port to allow for perpendicular traversing would
have weakened the structure.
73
-------
KNOCKOUT
BAFFLE
95%
DISPOSAL
AIR
Figure 9. CAFB pilot plant
-------
I
Figure 10. Pilot plant stack
-------
Particulate sampling was accomplished using a standard RAG train constructed
2
according to the procedures outlined in EPA Method 5. Due to the positions
of the installed ports, eight point traverses were taken on two diameters
120 degrees apart. The train was modified slightly to allow for sampling
of gaseous organic species (see below).
Particulate size distribution measurements were taken with a University
of Washington eight stage instack impactor using ungreased substrates.
A single point was sampled isokinetically for sufficient time (15 to
30 minutes) to collect a weighable quantity on each stage.
2
In addition to particulate, flue gas was sampled for NO by Method 7,
2 2 x
S0_/S0« by Method 8 and H.S by Method 11. An Orsat analyzer was used to
measure CO, CO. and 0-.
To collect gaseous organic species the RAG train was modified by placing
a gas adsorbent column between the filter and the impingers. This
column, shown in Figure 12, was developed by Battelle Columbus Laboratories
and made available to GCA for this program. Flue gas, after passing through
the filter, is cooled to slightly above its dew point and then passes
through a cell containing Tenax GC adsorber. This polymer reportedly
collects all organic gases C, and above. After sampling for approximately
1 hour the adsorbent columns were capped and stored in darkness to await
laboratory analysis.
76
-------
Figure 11. Stack sampling ports
-------
oo
FLOW DIRECTION
8-MM GLASS
COOLING COIL
GLASS WATER
JACKET
ADSORBENT-
s
RETAINING SPRING -i
\ ' \
"• - -.' .: •' -">s^ V
GLASS FRITTED
DISC
GLASS WOOL PLUG-
FRITTED STAINLESS STEEL DISC
15-MM SOLV-SEAL JOINT
-JT
1L
Figure 12. Absorbent sampling system
-------
FIELD ANALYSES
Analyses for total particulate, SO./SO-, N0x, H2S and particle size dis-
tribution were performed on-site in ERCA laboratories by GCA personnel.
The procedures outlined in EPA Methods 5, 7, 8 and 11 were followed for the
analyses of total particulate, NO , SO and KLS, respectively. To preclude
X A £•
degradation, all standards, except barium perchlorate and potassium
dichromate which are stable for long periods of time, were prepared at
ERCA.
LABORATORY ANALYSES
Three general types of analytical procedures were applied to oil, flue gas,
particulate and solid waste samples collected during the field test program:
organic functional group identification; trace element quantification; and
surface element and inorganic compound quantification. Organic functional
group and trace element analyses were performed according to the procedures
outlined in the EPA Level 1 protocol; surface analysis is more properly
a Level 2 technique. Each analytical technique is described below.
*
Organic Functional Group Analysis
In this procedure, sample extracts are separated into eight fractions by
liquid chromatography (LC), evaporated.to dryness, weighed, redissolved
and analyzed by infrared spectrpscopy. _ Methylene chloride was used to
extract oil, particulate and spent stone samples; "pentane was used to
extract organic vapors adsorbed on the Tenax polymer.
Liquid chromatographic separation into eight fractions is accomplished by
transferring the extract to an LC column and eluting sequentially with the
following solvent mixtures:
These analyses were performed by Battelle Columbus Laboratories under
subcontract and by the Process Measurements Branch of EPA.
79
-------
(1) 25 ml 60/80 petroleum ether
(2) 25 ml 20% methylene chloride in 60/80 petroleum ether
(3) 25 ml 50% methylene chloride in 60/80 petroleum ether
(4) 25 ml methylene chloride
(5.) 25 ml 5% methyl alcohol in methylene chloride
(6) 25 ml 20% methyl alcohol in methylene chloride
(7) 25 ml 50% methyl alcohol in methylene chloride
(8) 25 ml methyl alcohol.
Table 16 indicates the classes of organic compounds eluting in each frac-
tion and their detection limits based upon the total sample extract. After
collection from the LC column each fraction is reduced to dryness using
a Kuderna-Danish evaporator and air evaporation and then weighed to deter-
mine the amount of organic material in each fraction.
The dried fractions are then redissolved in methylene chloride and subjected
to IR analysis. The IR spectra are then scanned for functional group peaks.
*
Trace Element Analysis
Stack particulate, spent stone, fuel oil, gasifier bed stone, gasifier
cyclone fines, and knockout baffle material were analyzed for elemental
composition using low precision (± 200 percent) spark source mass spectrom-
etry (SSMS). This technique is sensitive to 70 elements. To calibrate
the SSMS results, some elements were quantified by higher precision atomic
absorption (AA) spectroscopy. Interference of organic ions with low atomic
weight elements is well known in SSMS as are losses of volatile compounds.
Thus uncertainties of values derived for light elements such as fluorine,
sodium and sulfur may be higher than the indicated precision.
This work was performed by Battelie Columbus Laboratories and Aculabs
under subcontract to GCA and by Northrup Services under contract to EPA.
80
-------
Table 16. CLASSES OF ORGANIC COMPOUNDS ELUTING IN EACH LIQUID
CHROMATOGRAPHY FRACTION., AND THEIR APPROXIMATE IR
DETECTION LIMITS
Fraction
Compound type
Approximate IR sensitivity
1
2
3
Aliphatic hydrocarbons
Aromatic hydrocarbons
POM
PCB
Halides
Esters
Ethers
Nitro compounds
Epoxides
Phenols
Esters
Ke tones
Aldehydes
Phthalates
Phenols
Alcohols
Phthalates
Amines
Amides
Sulfonates
Aliphatic acids
Carboxylic acid salts
Sulfonates
Sulfoxides
Sulfonic acids
Sulfonic acids
1-10 yg
1-10 yg
0.1-1 yg
0.1-1 yg
0.1-1 yg
0.1-1 yg
0.1-1 yg
0.1-1 yg
81
-------
Surface Analysis
A number of particulate and solid samples were investigated for surface
elements and inorganic compounds using X-ray photoelectron spectroscopy
(XPS) also known as electron spectroscopy for chemical analysis (ESCA).
In ESCA a high energy X-ray beam (for the analyses reported here the MgKa
line having an energy of 1253.6eV was used) impinges on a solid knocking
out core electrons from atoms on the solid surface. The resulting electrons
pass through an energy analyzer and are pulse counted by a particle mul-
tiplier. The binding energy of the electrons is then calculated from the
energy of the incident X-ray, the spectrometer work function and the measured
electron kinetic energy. Binding energy ranges can be uniquely associated
with specific precursor elements. In fact, ESCA is sensitive to all elements
in the periodic table. An additional feature of ESCA spectra is that the
1 precise electron binding energy in a known range is a function of the
valence state of the atom of interest. For example, sulfur combined as
sulfate can be differentiated from sulfur as sulfide. In addition, because
core electron ejection cross sections are relatively independent of valence
state, the ratio of the areas under the peaks corresponding to sulfate
and sulfide is a measure of the sulfate-sulfide surface concentration ratio.
A further consequence of the independence of cross section upon valence
state is that the relative concentrations of all elements on a surface
can be calculated from known sensitivity values. Table 17 lists sensi-
tivity factors applicable to the GCA/McPherson ESCA 36 instrument calculated
4
from published photoionization cross sections.
All samples analyzed by ESCA in this study were first scanned over the entire
electron binding energy range (broadband scan) to identify those elements
present in concentrations greater than 0.1 to 1 percent (the sensitivity
of ESCA to any one element is a function of the photoionization cross
section of the most intense core electron emission of that element).
82
-------
Table 1.7. ELEMENTAL SENSITIVITY. FACTORS.
FOR THE ESCA36
Element
0
C
N
K
S
Fe
F
Ca
Cl
Si
Pb
Al
Sb
As
Na
Cu
Sn
P
V
Mg
Cr
Cd
Electron
Is
2s
Is
Is
2s
2p
2p
2p
3P
Is
2p
2s
2p
2s
2p
4d5/2
4f
2s
2P
3d3/2
4p
4d
3p
3d
2s
2p
KLL
2p
LMM
3d
2s
2p
2p3/2
2s
2p
2p
3d
Sensitivity factor
430.8
36.6
241
340
436
956
496.8
618.7-
381
562
1215
275
567
394.2
596
459.8
4317 (±20%)
226
258.5
3787
154.6
1331 (±25%)
364
540
283.7
170.2
1381
589
1060
2990 (±25%)
243.6
479.9
1026.3
220
167.4
743 . 6
2820.7
83
-------
These broadband spectra were then analyzed to yield surface concentrations
of all identifiable elements.
Some of the filter samples and impactor substrate samples had relatively
light covering thus exposing portions of filter and aluminium foil to the
X-ray beam. ESCA scans of these bare materials are shown in Figures 13
and 14.
The compound forms of surface vanadium and sulfur are also of interest in
this study. They were investigated by scanning the binding energy ranges
corresponding to the ejection of the 2p electron of vanadium and the 2p
electron of sulfur. Figures 15 and 16 show the spectra of V2<35 and
vanadium metal used as standards to bound the vanadium valence range between
+5 and 0. The asymmetric bimodal structure of each spectrum is due to the
presence of two spin-orbit states, 2p-,2 and 2p- ,„, and not to two different
oxidation states. In all vanadium analyses the position of the larger
2p . peak was used as the comparison position. The oxygen ion peak, present
in all vanadium scans, was used to calibrate the binding energy scale.
Similarly, sulfate and sulfide bound the sulfur valence state scale between
+6 and -2.
In addition to the vanadium and sulfur scans, the Is peak of carbon was
scanned over the energy range between 275 and 295 eV. This scan serves
two purposes: the position of the carbon Is peak at 284.8eV corresponding
to hydrocarbons (the major carbon peak) calibrates the energy scale; and the
size of the carbonate peak at 289.leV indicates the surface concentration
of this species relative to organic carbon species. In addition, the shape
of the main carbon peak is indicative of hetero-atom substitution of the
hydrocarbon species. Figure 17 displays the binding energies of carbon
Is electrons ejected from various carbon compounds.
To supplement the,bulk SSMS analyses, high energy argon ions were used to
o
etch away surface layers exposing strata 20 to 100 A deep. The exposed
sample layers were then rescanned over the entire binding energy range and
84
-------
VF
00
JO
w
O
UJ
<
fL
13
O
O
tn
O
1000
_L
_L
800
600 40O
BINDING ENERGY, eV
200
Figure 13. Hi-vol filter. Broadband ESCA scan
-------
O9
o
»
Ul
O
O
J-
_L
—
o
J.
1000
_L
I I
VAL
CO
-------
V205( STANDARD)
o
UJ
oo
o
o
t-f-5
_L
_L
540
532
524 516
BINDING ENERGY, eV
508
50O
Figure 15. Vanadium metal ESCA scan
-------
V FOIL-2min SPUTTERING WITH Ar+
(STANDARD)
CO
CO
JO
o
uT
I-
o
o
_L
_L
_L
540
532
524 516
BINDING ENERGY, eV
5O8
500
Figure 16. Vanadium pentoxide ESCA scan
-------
291
290
289
RCOOH
•CaC03
>
OJ
UJ
z
UJ
5
2
00
288
287
286
285
284
283
Cr(CO)6
0
C = 0
0
CBr
HYDROCARBONS
GRAPHITE
Fe3C
SiC
•CCI.COH
282
vc
TiC
281
Figure 17. Carbon Is binding energies
89
-------
the resultant elemental concentrations were compared with the surface
values arid the SSMS analyses.
FIELD TEST PROGRAM
Stack sampling was carried out during three distinct operating conditions
of the pilot plant: fuel oil gasification; bitumen gasification; bitumen
combustion (startup). In all, seven sampling runs were made, four during
fuel oil gasification, two during bitumen gasification and one during
bitumen combustion. The number and duration of the tests were limited by
pilot plant up time. Fuel oil gasification runs (Runs 1 to 4) approached
"normal" operation; however, frequent cyclone malfunction occurred, re-
sulting in variable particulate emissions. The first bitumen gasification
run (Run 5) was rendered "abnormal" by clogging of the gas ifier-regenerator
stone transfer system with consequent buildup of sulfided stone in the
gasifier. Later during this same run the transfer system was purged and
fresh stone addition commenced. Startup operation (Run 6) consisted of
bitumen combustion accompanied by fresh stone feed. The final test run
(Run 7) was carried out during bitumen gasification and fresh stone
feeding.
Table 18 summarizes pilot plant operating modes for Runs 1 to 7. Fluc-
tuations in operating conditions occurred during several of the runs.
The row labeled "Stone feed" refers to continuous operation; because of
blockages in the gasifier-regenerator transfer system, stone was added at
the beginning of Runs 1 to 4 after which the continuous feed system was
shut down. Table 19 lists representative operating (temperatures for all
runs acquired from pilot plant log sheets. A comprehensive listing of all
operating temperatures, pressures and feed rates is not presented here but
will be published in a forthcoming ERCA/EPA report.
Table 20 summarizes field tests and samples collected during the field
trip. The first group of emissions were measured or collected at the stack
sampling port. The bottom group except for leached stone were acquired from
ERCA personnel and retained for laboratory analysis.
90
-------
Table 18. SUMMARY OF CAFB PILOT PLANT OPERATING
MODES DURING TEST PROGRAM
Run
1
.2
3
4
5
6
7
Date
12/04/75
12/05/75
12/06/75
12/08/75
12/09/75
12/10/75
12/11/75
_ .a
Fuel
No. 6 oil
No. 6 oil
No. 6 oil
No. 6 oil
Bitumen
Bitumen
Bitumen
Fuel heat input
2.16 x 104 kcal/s
(5.13 x 10 Btu/hr)
2.16 x 104 kcal/s
(5.13 x 10 Btu/hr)
2.16 x 104 kcal/s
(5.13 x 10 Btu/hr)
2.16 x 104 kcal/s
(5.13 x 10° Btu/hr)
2.38 x 10, kcal/s
(5.66 x 10 Btu/hr)
2.38 x 104 kcal/s
(5.66 x 10 Btu/hr)
2.38 x 104 kcal/s
(5.66 x 10 Btu/hr)
Gasifier
operating mode
Gasification
Gasification
Gasification
. Gasification
Gasification
Combustion
Gasification
Stone
feed
Off
Off
Off
Off
On
On
On
Feed rate: 2.27 1/s (36 gal/hr).
Based on»fuel oil specific gravity of 0.958 and heating value of
9.94 x 10 kcal/kg (1.79 x 104 Btu/lb) and bitumen specific gravity
of 1.0185 and heating value of 1.029 x 104 kcal/kg (1.853 x 104 Btu/lb).
Table 19. REPRESENTATIVE PILOT PLANT
OPERATING TEMPERATURES3
Unit operation
Gasifier
Regenerator
Bitumen feed
Oil feed
Gasifier air feed
Top of stack
Temperature ,
°C (°F)
900
950
160
85
150
110
(1652)
(1742)
(320)
(185)
(302)
(230)
From pilot plant log sheets.
91
-------
Table 20. SUMMARY OF SAMPLING ACTIVITY
Type of sample or test
SO
X
NO
X
H2S
Total particulate
Particulate sizing
°2
co2
CO
Moisture
Organic stack gases
Gasifier bed
Regenerator bed
Q
Left-hand cyclone
a
Right-hand cyclone
Knockout baffle
Stack cyclone
Bitumen
Fuel oil
Limestone
Leached stone
Run
1
X
X
X
X
X
X
X
X
X
2
X
X
X
X
X
X
X
X
X
X
X
X
X
3
X
X
X
X
X
X
X
X
4
X
X
X
X
X
X
X
X
X
5
X
X
X
X
X
X
X
X
X
X
X
X
X
X
6
X
X
X
X
X
X
7
X
X
X
X
X
X
Other
xb
xb
X
These cyclones are located between the gasifier and
boiler.
Obtained during pre-sampling site survey September 1975.
92
-------
The leached stone material is spent regenerator stone from a 1974 pilot
plant run which was.sintered and placed in leaching buckets on the ground
near the plant. Six leached stone samples identified in Table 21 were
collected.
Table 21. LEACHED STONE SAMPLES
Sample
identification
SSI
SS2
SS3
SS4
SS5
SS6
Sintering
temperature ,
. °C
1500
1500
1400
1400
1300
1300
Sintering
time,
hours
3
1
3
1
3
1
FIELD TEST RESULTS
Tables 22 to 28 present the results of field analyses from Runs 1 to 7.
Table 29 summarizes the emission measurements for SO , NO
particulate, each of which is discussed below.
H2S and total
SO
Sulfur dioxide emissions during fuel oil gasification were approximately
0.65 lb/10 Btu (300 ppm), almost 20 percent below the New Source Per-
formance Standard (NSPS) for oil-fired steam generators. During these
runs the SQ2 concentration :in the regenerator off-gas ranged between 4 and
5 percent. This value is in good agreement with those reported by ERCA
and with Foster-Wheeler projections of 0.64 and 0.78 lb/10 Btu for the
demonstration plant and commercial CAFB unit, respectively.
93
-------
Table 22. FIELD TEST RESULTS: RUN 1
FUEL OIL GASIFICATION
Emission or
parameter
Flue gas flow rate
Temperature at
sampling port
Moisture
co2
°2
CO
N0x
S02
so3
H2S
Total partlculate
Rate or value
0.56
0.102 g/dscm
0.775 g/dscm
0.027 g/dscm
<7 x 10" g/dscm
0.117 g/dscm
dscm/s (1180 dscfm)
108°C (226°F)
8.3 %
13.0 %
1.0 %
0.1 %
(53.5 ppmv) 0.085
(292 ppmv) 0.643
(8.3 ppmv) 0.023
0.097
lb/106 Btu
lb/106 Btu
lb/106 Btu
lb/106 Btu
Table 23. FIELD TEST RESULTS: RUN 2
FUEL OIL GASIFICATION
Emission or
parameter
Flue gas flow rate
Temperature at
sampling port
Moisture
co2
°2 '•'
CO
NO
S02
so3
V
Total participate
Rate or value
0.51 dscm/s (1088 dscfm)
109°C (229°F)
8.6 %
12.0 %
3.8 %
0.1 %
0.087 g/dscm (45.7 ppmv) 0.067
0.089 g/dscm (305 ppmv) 0.619
0.026 g/dscm (7.9 ppmv) 0.020
<7 x 10"5 g/dscm
0.073 g/dscm 0.056
lb/106 Btu
lb/106 Btu
lb/106 Btu
lb/106 Btu
94
-------
Table 24. FIELD TEST RESULTS: RUN 3 FUEL OIL GASIFICATION
Emission or
parameter
Flue gas flow
rate
Temperature at
sampling port
Moisture
C00
t.
°2
CO
H»S
£.
Total
particulate
Rate or value
0.60 dscm/s (1265 dscfm)
111°C (231°F)
9.7%
12.2%
2.8%
0.2%
3.17 x 10~4 g/dscm (0.23 ppmv) 2.63 x 10~4
0.080 g/dscm 0.067
lb/106 Btu
lb/106 Btu
Table 25. FIELD TEST RESULTS: RUN 4 FUEL OIL GASIFICATION
Emission or parameter
Rate or value
Flue gas flow rate
Temperature at sampling port
Moisture
co2
°2
CO
Total particulate
0.58 dscm/s (1238 dscfm)
133°C (271°F)
9.0%
11.6%
5.3%
0%
0.106 g/dscm 0.092 lb/106 IJtu
95
-------
Table 26. FIELD TEST RESULTS: RUN 5 BITUMEN GASIFICATION
Emission or parameter
Rate or value
Flue gas flow rate
Temperature at sampling port
Moisture
co2
CO
0.49 dscm/s (1040 dscfm)
138°C (281°F)
9.5%
12.0%
3.9%
0%
N0x
so2
so3
Total particulate
0
2
0
0
.111
.194
.037
.141
g/dscm
g/dscm
g/dscm
g/dscm
(58.
(828
(11.
4
1
ppmv)
ppmv)
ppmv)
0.
1.
0.
0.
079
562
026
101
lb/10
lb/10
lb/10
lb/10
D
6
6
6
Btu
Btu
Btu
Btu
Table 27. FIELD TEST RESULTS: RUN 6 BITUMEN COMBUSTION
Emission or parameter
Rate or value
Flue gas flow rate
Temperature at sampling port
Moisture
co2
°2
CO
Total particulate
0.56 dscm/s (1193 dscfm)
80°C (176°F)
2.4%
12.0%
3.9%
0%
0.056 g/dscm 0.104 lb/106 Btu
Table 28. FIELD TEST RESULTS: RUN 7 BITUMEN GASIFICATION
Emission or parameter
Rate or value
Flue gas flow rate
Temperature at sampling port
Moisture
co2
°2
CO
Total particulate
0.51 dscm/s (1090 dscfm)
127°C (261°F)
7.8%
12.0%
3.9%
0.112 g/dscm 0.192 lb/10 Btu
96
-------
Table 29, SUMMARY OF STACK EMISSIONS
Run
1
2
3
4
5
6
7
Fuel
Fuel oil
Fuel oil
Fuel oil
Fuel oil
Bitumen
Bitumen and
stone feeding
Bitumen
NO
X
ppm
53.5
45.7
58.4
lb/106 Btu
0.0851
0.0671
0.0791
so2
ppm
292
305
828
lb/106 Btu
0.6431
0.6191
1.5624
so3
ppm
8.3
7.9
11.1
lb/106 Btu
0.023
0.0201
0.0263
H2S
ppm
<0.05
<0.05
0.23
lb/106 Btu
2 . 6xlO"4
Total
particulate
ppm
lb/106 Btu
0.0971
0.0561
0.063
0.0921
0.101
0.1046
0.1921
VO
-------
Bitumen gasification during Run 5 produced an SO- emission of 1.56 lb/10
Btu (828 ppm). Two factors contributed to this elevated discharge. First,
the sulfur content of the bitumen is 50 percent higher than that of the
residual oil. Therefore, about 50 percent of the additional SCL greater
than 300 ppm can be attributed to fuel sulfur content. The second and
more important factor was the presence of saturated limestone in the
gasifier due to gasifier-regenerator transport duct clogging. This factor
was reflected in the regenerator off-gas which contained only 1 percent
S0». Although fresh limestone was added to the gasifier at about 11:00 a.m.
on December 9, stack sampling was performed earlier in the day when the
sulfur recovery efficiently (SRE) was abnormally low.
Sulfur trioxide emissions, for which there are no NSPS, increased only
about 40 percent in response to stone saturation. The mechanism for SO
formation in combustion systems is not yet established. Three pathways
have been proposed:
• Gas-phase reaction between SO. and 02;
• Catalytic oxidation on particulate sur
• Gas-liquid reaction on water droplets.
Of the four species necessary for S0» formation, water, oxygen and par-
ticulate increased only slightly from Run 3 to Run 5. The fact that SO
increased by only 40 percent in the presence of an almost 300 percent in-
crease of SO* could be taken to indicate that the rate of the reaction to
form the trioxide is less than first order in S02.
NO emissions result from two reactions occuring within the gasifier and
the boiler:
*
In fact SOo is made commercially by passing SO,, and 0~ over a V^Oc
catalyst.
98
-------
Fuel bound N + 1/2 0 NO = Fuel-N conversion
and
Atmospheric N2 + 02 ^ 2ND = Thermal fixation
Both reactions' require a'high temperature envi-ronment. -Euel-N conversion
proceeds at normal combustion temperatures and is weak function of temper-
ature. Thermal fixation, on the other hand, is highly temperature-
dependent, with the rate of NO formation increasing significantly above
980°C (1800°F).
In the relatively low temperature, 900°C (.1.65Q°F) .environment of the
gasifier the thermal fixation reaction is very inefficient. In fact,
studies of NO ' formation in fluidized bed combustion of coal in which
X . '
the oxygen concentrations is in excess of stoichiometric indicate that
at this temperature almost all NO produced is formed from fuel nitrogen
conversion. Measured stack NO emissions include not only NO formed in
X X
the gasifier but also that produced in the high temperature boiler, where
thermal fixation is likely the primary source of NOX.
The average NOV emission rate during Runs 1, 2 and 5 of 53.5 ppm is
X
considerably lower than the low end of the emission rate range found
for conventional oil and gas-fired boilers. This measured rate is also
about one-fourth of the NSPS for-oi-1-fired boilers and one-third of the
NSPS for gas-fired boilers. Furthermore, the relative invariance of the
three measurements suggests a low correlation between the NO emission
X
rate and temperature, excess oxygen, bed stone history and fuel.
Several factors may contribute-to the low absolute NO emission rate. The
X
reducing atmosphere may severely inhibit oxidation of fuel nitrogen. It
has also been suggested that limestone might catalytically aid in the
decomposition of NO'or react directly with NO. The presence of nitrogen
on the surface of some of the smaller particulate (which is noted later in
the section) is consistent with this latter mechanism. It is also
99
-------
possible that reduced nitrogen species, such as NH-, formed in the gasi-
fier pass through the boiler without reacting. It is more difficult to
explain away the apparently small amount of NO formed by thermal fixation
in the boiler. The rate of tube thermal fixation reaction is strongly
affected by boiler design and firing characteristics which cannot easily
be evaluated.
Two uncertainties are thus apparent. What is the fate of the fuel bound
nitrogen which is not converted to NO ? How will the rate of the thermal
X
fixation reaction be affected by the particular characteristics of the
boiler to be used in conjunction with the demonstration plant?
§2-
No hydrogen sulfide was detected during the first two oil gasification
runs and only a quarter of ppm was found in the third run. Thus, H.S
does not present a pollution problem for the CAFB.
Particulate
The primary source of particulate emissions from the CAFB is gasifier bed
stone which passes through the two internal cyclones, the knockout baffle
and the stack cyclone. The NSPS for particulate emissions from oil-fired
boilers is 0.1 lb/10 Btu. Inspection of Table 29 shows that during
oil gasification two of the four runs produced emissions only a few per-
cent below the standard. The NSPS was exceeded during bitumen gasifica-
tion and combustion; the final bitumen gasification run exceeded the par-
ticulate standard by a factor of two. Two factors must be invoked to
understand the variation in and magnitude of particulate emissions:
• Cyclone efficiency
• Fresh stone feed.
As noted earlier, cyclone malfunction occured frequently during all runs.
In addition, ERCA personnel reported that the cyclones installed at the
100
-------
pilot plant were very old and were not designed specifically for the CAFB
system. ERCA estimates stack cyclone efficiency of 50 percent. The
high emission rate from Run 7 can be attributed to an unusually high
fresh stone feed rate. This stone feed rate will be typical of CAFB
start-up procedure. Fresh stone undergoes attrition as it enters the
gasifi,er while being transformed from carbonate to oxide. This start-up
condition will normally occur in conjunction with gasifier combustion
(see Section II) but was employed during Run 7 to compensate for the
buildup of saturated stone due to the clogging of the gasifier-regenerator
stone transfer duct. Similar stone addition occurred during pilot plant
Run 5, but particulate sampling took place earlier in the day.
Figures 18 and 19 are bar graphs of particulate size distributions for
fuel oil and bitumen gasification, respectively. Figure 20 presents
these data in log-normal format. These distributions indicate that a sub-
stantial fraction of the particulate emissions are in the respirable range
and hence of primary concern. The large respirable fraction is typical
of conventional cyclones and may be expected in emissions from the de-
monstration plant which will also employ cyclones for particulate control.
It is difficult to predict particulate loading and size distribution for
the demonstration plant and 250 MW unit. Foster-Wheeler claims cyclone
design efficiencies of 98 percent, but extensive testing during normal
gasification and startup will be necessary to establish actual efficien-
cies. The abnormally high particulate emission rate observed during fresh
stone feed at the pilot plant will have to receive special attention in
the demonstration program.
LABORATORY RESULTS
Three types of laboratory analyses were performed on the samples listed
in Table 20. The decisions regarding which samples to analyze by which
technique were made based upon: importance of information to be gained;
availability of previous analyses; cost of analysis; availability of
101
-------
o
(•O
30
25
2O
s
5
is
10
RUN #3
j 1 1_ _1_ 1 i
I j I 1 | I j i 1 i i
O.I 0.2 O.3 0.5 O.7 I 2 3456 789IO 20 30 SO 7O IOO
AERODYNAMIC DIAMETER, MICROMETERS
Figure 18. Stack particulate size distribution, Run No. 3.
Fuel oil gasification
-------
o
u>
MASS PERCENTAGE
_ _ r» r
D 01 O Oi O t
i i
i i
i i i i i i j
i i i i i i 1
i i i i i
i i i 1 i ii
RUN #5
J
i i i 1 i i
i
0.2 0.3 0.5 0.7 I 2 3 4 5 6 7 8 9IO
AERODYNAMIC DIAMETER, MICROMETERS
20 30 40
Figure 19. Stack particulate size distribution, Run No. 5.
Bitumen gasification
-------
30
20
CO
OC a
m o
Ul f.
X 6
o
oc
2 4
(E
Ul
Ul
2
O
O
I
z
1.0
o °'8
tr 0.6
ui
0.4 -
0.3 -
i i i
i i
xo
o x
ox
LEGEND
o RUN #3
x RUN # 5
i i i i
O.I I 10 20 40 60 80 90 97
PERCENTAGE OF MASS LESS THAN OR EQUAL TO STATED SIZE
Figure 20. Log-normal particulate size distributions
104
-------
additional analytical support. Because the emphasis in the present
study is on boiler stack emissions, laboratory.work was directed toward
characterization of the organic and inorganic chemical nature of stack
particulate and gases. Nevertheless, to extend the investigations by
ERCA and Westinghouse on the nature of spent stone, regenerator bed and
gasifier bed samples were analyzed for organic functional group and for
surface elements and compounds. Table 30 lists the types of analyses per-
formed on samples collected during the field trip (see also Table 20).
Bitumen sample results and fuel oil sample results are presented separately.
below.
Bitumen Gasification
Bitumen - The organic functional group composition of bitumen was in-
vestigated to assess the potential effects of fugitive emissions from
storage and handling (see also Section IV). As with all organic analyses
reported in this section, the liquid chromatography - infrared spectro-
scopy (LC/IR) technique described earlier was employed. Figures 21
through 28 are the IR spectra of the eight separable fractions. The
distribution of material among the eight fractions is listed in Table 31.
Because bitumen is entirely extractable, this distribution is effectively
a complete organic analysis of the fuel.
Of the groups tentatively identified in bitumen, POM, phenol and quinone
are of particular concern as fugitive species. No MEGS ahve been es-
tablished for POM, but in general, any amount of these carcinogenic spe-
cies is considered dangerous. Table 32 summarizes the health effects of
several classes of organic compounds and lists their MEGS. Phenol and
quinone emissions present potential problems because of their relatively
high vapor pressures at the temperature at which bitumen is handled; the
low MEG of quinone makes emissions of this compound particularly pernicious.
In fact, the high temperature of bitumen storage and handling dictates
Through PMB/EPA Contract No. DA-6-99-H606A for ESCA analyses.
105
-------
Table 30. SAMPLE ANALYSES
Sample
Bitumen
Fuel oil
Limestone
Regenerator bed stone
Run no. 4
Run no. 5
Gasif;Ler bed stone
Run no. 2
Run no. 5
Left-hand cyclone particulate
Run no. 2
Run no. 4
Right-hand cyclone particulate
Run no. 5
Knockout baffle particuiate
Run no. 3
Run no. 5
Stack cyclone particulate
Run no. 4
Run no. 5
Gaseous effluent
Run no. 4
Run no. 6
Run no. 7
0
leached stone
Type of analysis
Code
.BIT
F05
LSS
RB8
RB9
GB5
GB9
LH5
LH8
RH9
K06
K09
SC8
SC9
GE8
GEO
GE1
SSI
SS3
SS5
Organic
X
X
X
X
X
X
X
X
X
X
trace element
X
X
X
X
X
X
X
X
X
X
Surface
X
X
X
X
X
X
X
X
X
X
X
X
X
106
-------
Table 30 (continued). SAMPLE ANALYSIS
Sample
Method 5 train, filter catch
Run no. 1
Run no. 2
Run no. 3
Run no. 4
Run no. 5
Run no. 6
Run no. 7
Impactor substrates
Run no. 3
Stage 1
Stage 2
Stage 3
Stage 4
Stage 5
Stage 6
Stage 7
Backup filter
Impactor substrates
Run no. 5
Stage 1
Stage 2
Stage 3
Stage 4
Stage 5
Stage 6
Stage 7
Backup filter
Type of analysis
Code
UW61
UW62
UW63
UW64
UW65
UW66
UW67
UW68
UW91
UW92
UW93
UW94
UW95
UW96
UW97
UW98
Organic
Trace element
Surface
X
X
X
X
X
X
X
X
X
X
X
X
X
• x
X
X
X
X
X
X
X
X
X
See Table 21.
107
-------
WAVELENGTH, mieroM
5 6
4000 3500 3000 2500 2000 1800 1600 1400 1200 1000 800 600 400 200
FREQUENCY, em-l
Figure 21. Bitumen, LC fraction 1. Peak at 2920 cm"1 and peaks at
~1450 cm"1 and 1370 cm"1 are from CH3, CH2- Bands at
~1600 cm"1 and 1690 are typical of asphaltic materials.
1690 cm~l band is from gross mixture of carbonyls whereas
1600 cm"1 band is due to structures such as highly con-
densed aromatics and quinones
2.5
100
^
£ 60
o 60
z
£ 40
WAVELENGTH, microns
8 9 10 12 • IS 20 30 40
at
CO
20
4000
3500 3000 2500 2000 1800 1600 1400 1200 1000 800 GOO 400 200
FREQUENCY, cm-'
Figure 22. Bitumen, LC fraction 2. See Figure 21. Bands between
700-900 cm ^ indicate aromatic compounds (possibly POM)
108
-------
2.5
100
i BO
WAVELENGTH, miCfOM
5 6 7 8 9 10 12 15 20 3040
*
40
20
0
4000
i |i FT r]nii 17111 i ( T i i iiii i i i i ii \\ |i ii i i i r i i r T i nr i n p
3500 3000 2500 2000 1800 1600 1400 1200 1000 600 600 400 200
FREQUENCY, crH
Figure 23. Bitumen, LC fraction 3. See Figure 21
WAVELENGTH .microns
8 9 10 12 15 20 30 "0
4000 3500 3000 2500 2000 1800 1600 1400 1200 1000 800 600 400 200
FREQUENCY, em'1
Figure 24. Bitumen, LC fraction 4. See Figure 21. This spectra
indicates asphaltic and carbonyl compounds
109
-------
WAVELENGTH, micron*
I i i » i Kj i i i
3500 3000 2500
i I I I I I I i I i I ' i
2000 1600 1600
I I 1 I 1 M I I I I I I 1 ' M I I T I I I I
1400 1200 1000 600 600 400 200
FREQUENCY, cnH
Figure 25. Bitumen, LC fraction 5. Strong band at ~3400 cm
indicates -OH. Band between 1200-1300 cnT1
this might be present as phenol
-1
suggests
2.5
°h
4000
WAVELENGTH, microns
5 678
9 10 12
20 30 40
I i ii i i vj
3500 300CT
FREQUENCY,cm-I
Figure 26. Bitumen, LC fraction 6. See Figure 21. Band at
1025 cm~l probably from SiO impurity
110
-------
2.5
100
WAVELENGTH .microns
45 6
7 B 9 10 12 15 20 30 40
i i i i j_ i i i i
80
Ul
=r 60
K-
5 40
I*
o
4000 3500 3000 2500 2000 1800 1600 1400 1200 1000 800 600 400 200
FREQUENCY, em-'
Figure 27. Bitumen, LC fraction 7. See Figure 21.
1025 cm"1 probably from SiO impurity
0 -.
4000
WAVELENGTH .microns
8 9 10 12 15 20 30 40
i i I i I i i I i I i i'I I | i r i | i i i i i i I I 'i ' I i i i i i'i i r i n | ' ' n ' T-T
3500 3000 2500 2000 1800 1600 1400 1200 1000 800 600 400 200
FREQUENCY, em-'
Figure 28. Bitumen, LC fraction 8. See Figure 21.- Band at
1025 cm"1 probably from SiO impurity
111
-------
Table 31. DISTRIBUTION OF MATERIAL AND FUNCTION GROUPS
IN BITUMEN
Fraction
1
2
3
4
5
6
7
8
Total
Weight, yg
8,900
890
810
350
280
1,400
210
450
13,290
Percent
67.0
6.7
6.1
2.6
2.1
10.5
1.6
3.4
100
Functional groups
1.
2.
3.
4.
5.
6.
7.
8.
Aliphatic hydrocarbons ,
asphaltenes , carbonyls ,
highly condensed aromatics ,
' quinones .
Aromatics (possibly POM)
Asphaltenes, carbonyls
Phenol
112
-------
Table 32. HEALTH AND ECOLOGICAL EFFECTS AND MEGS OF
ORGANIC COMPOUND CLASSES
Generic class
Hydrocarbons
Ex: Ethylene
Alcohols
Ex: Ethyl alcohol
Phenols
Ex: Phenol
Phthalate
Ex: Phthalic anhydride
ICsters and curboxyllc acids
Ex: Acetic acid
N-Heteroaromatic
Ex: Pyridine
Quinone
C=0 Containing species
Ex: Formaldehyde
Carboxylic acid salt
Kx: Acetic acid, nickel (II)
salt
Effects
Threshold effects on plants - reduced
growth, premature senescence and re-
duced flowering and fruit production.
Irritant to eyes and mucous membrane.
Repeated contact produces dry, scaly,
and fissured dermatitis. Causes
Intoxication when inhaled at
high concentrations.
Primary irritant having strong corro-
sive properties for all body tissue.
Acute poisoning mainly characterized
by central nervous tissue manifesta-
tions. Pneumonia, renal and hepatic
damage frequently follow phenol
intoxication.
In a pure state, it is not an irri-
tant, but in contact with water, the
caustic phthalic acid is formed.
Irritation may produce conjunctivitis
contact dermatitis, atrophy of the na-
sal mucous membrane, loss of sense of
smell and hoarsness.
Bronchitis, emphysema and asthma may
occur.
High concentration of vapor produce
conjunctivitis, dental errosion and
nasal irritation. On contact, glacial
acetic acid produces painful burns,
repeated contact produces fissured
dermatitis. Inhalation may lead to
bronchitis and pulmonary edema
Irritating to eyes, nose and throat.
acute exposure produces flushing of the
face and narcotic effects of nausea,
vomiting and dizziness. Kffects of
chronic exposure include headaches,
nervousness, and insomnia.
Condensation of vapor on eyes produces
conjunctivitis lacrimatlon, photophobia,
corneal strains, ulceratlons and opaci-
ties. In animals ingestlon of quinnnc •
produces convulsions, respiratory diffi-
culties, hypotension and asphyxia
Irritating to conjunctivia and mucous
membranes of upper respiratory tract.
Ingestlon may result in gastrointentinal
Irritation. Respiratory depression and
death.
Metallic nickel and its soluble salts
are toxic to animals due more to gas-
trointestinal irritation than to any
specific toxicity chronic inhalation of
nickel dust produce tumors. Ingestlon
of nickel by animals reduces repro-
duction and growth rates.
Phase
Gas
Gas
Gas
Particulate
and
gas
Gas
Gas
Gas
Particulate
MEG
ug/ml
6.3 mg/m
63 Mg/m
Ref.
17
18
18
40 ug/m
18
83
18
50 us/m
1.3 Ug/m
20 ug/m
18
18
18
19
113
-------
that fugitive emission sampling for gaseous organic compounds should be
undertaken at the demonstration plant. This point is particularly im-
portant because of the variability of fuels which will be available to
the CAFB.
Flue Gas - Spectra of gaseous organic stack emissions collected during
Run 7 are shown in Figures 29 through 34. Spectra not shown contain
no peaks other than those corresponding to aliphatic hydrocarbons (pre-
sent in all fractions) or to the ubiquitous silicon oil impurity.
Table 33 contains the distribution of material between the eight
fractions and lists species identified in each. The bulk of the gaseous
emissions is a mixture possibly containing disubstituted amide, N-hete-
roaromatics, doubly conjugated ketones and quinone. Additional, Level 2
organic analysis will be necessary to identify this material.
3
Gaseous effluent was collected for 53 minutes during which time 0.56 m
3
(19.7 ft ) of flue gas was pulled through the absorbent column. The
concentration of organic species (
-------
WAVELENGTH, micron*
2.5
_ 100
"c
I 80
uT
2 60
8 9 10 12 15 20 3040
_i i _ i i i i i i
GEI-I
»/>
<
40
20
I I I I I
4000 3500 3000 2500 2000 1800 1600 1400 1200 1000 800 600 400 200
FREQUENCY, cnH
Figure 29. Flue gas from bitumen gasification, Run No. 7,
LC fraction 1. Peaks at 2920, 1450 and 1370 in-
dicate aliphatic hydrocarbons. Structure at
lower frequencies is due to silicon oil impurity
2.5
100
.80
60
40
20
0
WAVELENGTH, micron*
4 5 6 7 8 9 10 12 15 20 3040
ii i I i i i i i i i r IT i i T i i i i i r i n i i i i i i i i i i r
4000 3500 3000 2500 2000 1800 1600 1400 1200 1000 800 600 400 200
FREQUENCY, em-'
Figure 30. Flue gas from bitumen gasification, Run No. 7, .
LC fraction 3. Complex spectrum suggests: (1)
disubstituted amide, (2) N-heteroaromatic, (3)
doubly conjugated ketone, or (4) quinone
115
-------
2.5
WAVELENGTH .microns
45 671
9 10 12 15 20 30 40
i i i i i i i
100
Is
5 80
*•
&
•S3 60
t 40
a
Z 20
4000
I 1 I | "I I
3500 3000 2500 2000 1800 1600 1400 1200 1000 800 600 400 200
FREQUENCY, em-'
Figure 31. Flue gas from bitumen gasification, Run No. 7, LC
fraction 4. See Figure 30
2.5
_ 100
7s
•*
I 80
uf
z 60
»-
5 40
I M
WAVELENGTH .microns
5 678
9 10 12 15 20 30 40
4000
3500 3000 2500 2000 1600 1600 1400 1200 1000 800 600 400 200
FREQUENCY,em-'
Figure 32. Flue gas from bitumen gasification, Run No. 7 LC
fraction 5. See Figure 30. Band at 3400 cm~^
suggests presence of alcohol or carboxylate
116
-------
WAVELENGTH, micront
5 6 7 8 9 10 12 15 20 3040
I '"' I I i i | i i i | I i i | i I'l | i i i | 'i i i | i ' »
4000 3500 3000 2500 2000 1800 1600 1400 1200 1000 800 600 400 200
FREQUENCY, em'"
Figure 33. Flue gas from bitumen gasification, Run No. 7, LC
fraction 6. Peak at 3400 cm"1 indicates carbo-
xylate group. Peak at 1640 cm~l indicates doubly
conjugated ketone. Peaks between 600-800 cm"1
indicate aromatics
2.5
WAVELENGTH, micront
5 6 7 8 9 10 12 15 20 30 40
100
: 80
3 60
z
S 40
a
to
20
GEI-8
i i i i | I I I I M I I I I I I I I 1 »ii 1 MI | |i ! | i I M I I i | i i i i I I i M ' ' I ' ' I
4000 3500 3000 2500 2000 1800 1600 1400 1200 1000 800 600 400 200
FREQUENCY,em-'
Figure 34. Flue gas from bitumen gasification, Run No. 7, LC
fraction 8. Possible traces of carboxylic acid
salts. Band between 1000-1100 cm"1 probably from
impurity
117
-------
Table 33. DISTRIBUTION OF MATERIAL AND FUNCTIONAL
GROUPS IN STACK GAS EFFLUENT: RUN NO. 7
Fraction
1
2
3
4
5
6
7
.8
Total
Weight, yg
1,700
60
1,600
13,000
1,100
250
74
32
17,816
Percent
9.5
0.3
9.0
73.0
6.2
1.4
0.4
0.2
100
Functional groups
Aliphatic hydrocarbons
Complex mixture
Complex mixture
Alcohol or carboxylate
Carboxylate, doubly
conjugated ketones, aromatics
Carboxylic acid salts
118
-------
WAVELENGTH, mi
20 30 40
4000 3500 3000 2500 2000 1800 1600 1400 1200 1000 600 600 400 200
FREQUENCY,em-l
Figure 35. Stack cyclone material from bitumen gasification,
Run No. 5, LC fraction 1. Peak at ~2920 cm and
peaks at 1450 cm"1 and 1370 cm"-'- are from CH,, CH9.
This indicates presence of aliphatic hydrocarbons.
Peak at 1730 cm"1 is from C=0
2.5
100
eo
at 60
I 40
i/>
« 20
WAVELENGTH, mic
6 7 8 9 10 12 15 20 3040
4000 3500 3000 2500 2000 1800 1600 1400 1200 1000 800 600 400 200
FREQUENCY,em-l
Figure 36. Stack cyclone material from bitumen gasification,
Run No. 5, LC fraction 2. Peaks at ~2920 cm"1 and
1730 cm"1 indicate presence of aliphatic esters
119
-------
2.5
100
60
.
60
5 20
at
WAVELENGTH, microns
45 6
7 8 9 10 12 15 20 JO 40
4000 3500 3000 2500 2000 1800 1600 1400 1200 1000 BOO 600 400 200
FREQUENCY, em-'
Figure 37. Stack cyclone material from bitumen gasification,
Run No. 5. LC fraction 3. Peaks at 1730 cm"1 and
~1500 cm"1 indicate presence of aliphatic carbonyl
compounds
120
-------
2.5
_ 100
I 80
uT
2 60
<
1 40
1M
0
WAVELENGTH .mj
5 6 7
8 9 10 12 15 20 30 40
I 1 I 1| I I
ii
i i i
4000 3500 3000 2500 2000 1600 1600 1400 1200 1000 600 600 400 200
FREQUENCY,cm-l
WAVELENGTH, microns
4000 3500 3000 2500 2000 1800 1600 1400 1200 1000 800 600 400 200
FREQUENCY, em-'
Figure 38. Stack cyclone material from bitumen gasification,
Run No. 5, LC fraction 4. Peak at 3400 cm"1 in-
-1
indicate
dicates -OH; peaks between 600-800 cm"
aromatics; peak at 1730 cm"1 indicates carbonyls.
Peak at 1500 cm"1 and complexity of spectrum be-
tween 1000-1300 cm"! indicates possible presence
of phthalates, phenols, or alcohols
121
-------
2.5
100
. 60
60
40
20
WAVELENGTH .micron*
4 5
' '
6 ? 6 9 10 12
_J 1 1 1 i.i
20 30 40
o
ts>
T l i i | i i i l | l l I i | I I I I | I I l | I i l | l l i i i
4000 3500 3000 2500 2000 1800 1600 1400
FREQUENCY, cm-l
I I I l l | l l l | l l i | i i i | lit
1200 1000 600 600 400 200
Figure 39. Stack cyclone material from bitumen gasification,
Run No. 5, LC fraction 5. See Figure 38.
2.5
100
60
60
WAVE LENGTH, microns
8 9 10
12
15
20 30 40
' '
I 40
-------
Table 34. DISTRIBUTION OF EXTRACTABLE ORGANIC MATERIAL AND
FUNCTIONAL GROUPS IN STACK CYCLONE PARTICULATE:
RUN NO. 5
Fraction
1
2
3
4
5
6
7
8
Total
Weight, pg
70
14
28
140
230
66
16.
90
654
Percentage
10.7
2.1
4.3
21.4
35.2
10.1
2.4
13.8
100
Functional groups
Aliphatic hydrocarbons ,
Aliphatic esters
Aliphatic carbonyls
Aromatics, carbonyls,
phthalates , phenols ,
As in fraction 4
Carbonyls, alcohols
carbonyl
alcohols
123
-------
cyclone material is representative of particulate emissions (this is
probably a poor assumption because substantial condensation of organic
gases on particulate occurs after the stack cyclone where the flue gas has
cooled down significantly), 654 yg of organic recovery corresponds to an
O _C f.
organic particulate loading of 0.12 mg/m or 7 x 10 lb/10 Btu. This
organic emission rate, when compared with the health effects data in
Table 32 does not appear to be a potential problem.
The stack cyclone particulate sample was also analyzed for bulk elemental
composition by the methods discussed earlier. The results of this analysis
is presented in Table 35. The total particulate loading during Run 5
3
was 0.141 gm/m . Multiplying the concentrations listed in Table 35
by this number yields the concentration of trace elements in the flue gas.
For all metals, the result is less than the worst case analyses emission
*
factor listed in Table 9.
The only .element whose particulate abundance is larger than the worst
3 3
case prediction is fluorine 0.06 mg/m versus 0.014 mg/m . Fluorine was
also found in the analysis by ERCA of stack cyclone particulate from a
previous pilot plant run (private communication); however in that case
the fluorine concentration was between 6 and 60 ppra (compared to 450 ppm
here). In the present case three possible explanations for "violation"
of the worst case result can be given. Worst case analyses displayed in
Section IV were based upon analyses for fuel oil and limestone reported
in Tables 4 and 6. No trace element analysis is available for bitumen;
fluorine may be much more abundant in this fuel than in No. 6 oil.
It is also possible that ERCA's analysis of limestone is in error. Their
analysis indicates an upper limit of 2 ppm for fluorine but also indi-
cates the presence of an interference. If fluorine were present at a
Although the values in Table 9 were calculated based upon No. 6 oil
as the fuel, trace metal concentrations in bitumen will not differ
significantly.
124
-------
Table 35. MASS SPECTROGRAPHIC AND ATOMIC ABSORPTION SPECTROMETRIC
ANALYSIS OF STACK CYCLONE PARTICULATE "RUN NO. 5
Q
Element
CaC
Sb
V0
Si
Na
Mg
Nic
Fec
F
K
Al
Cl
Ti
Ba
Cr
Cu
Sr
Zn
P
Mn
Co
Pb
Mo
Li
Ge
B
Br
Zr
Se
Concentration, ppmw or percent
14.1 %
3.83
1.04
0.49
0.43
0.32
0.22
0.12
450 ppmw
340
340
120
63
55 • . ' • .
51
33
32
30
21
21
17
7.8
5.0
4.3
4.1
2.3
2.2
1-5
1.3
125
-------
Table 35 (continued).
MASS SPECTROGRAPHIC AND ATOMIC
ABSORPTION SPECTROMETRIC ANAL-
YSIS OF STACK CYCLONE PARTICU-
LATE RUN NO. 5
a
. Element
. I
. Rb
Ce
Yb
Ga
. Bi
Ta
Cd
Sn
W
Hf
Tl
Y
La
Th
Dy
Sm
Be
Nb
Nd
Pr
Concentration, ppmw or percent
1.3. ppmw
1.1
0.6
<0.5
0.4
0.4
0.4
0.3
0.3
<0.3
<0.3
0.2
0.2
0.2
<0.2
<0.2
<0.2
<0.12
0.1
0.1
<0.1
Elements not listed are
<0.1 ppm, not detected.
Determined by wet chemistry.
c
Determined by atomic absorp-
tion spectrometry.
Used as internal standard.
126
-------
level of 2 ppm the worst case analysis would still yield an upper limit
of also 0.025 tng/m . Another explanation for the apparently high fluorine
content of particulate is that the fluorine concentration measured by
SSMS is artificially high due to interfering contributions to apparent
mass 19 by organic ions. Nevertheless, if the fluorine concentration
reported in Table 35 is correct, the resultant ambient loading is still
too low to be of concern (see Table 9).
In Section IV it was pointed out that vanadium, cadmium and nickel are the
only metals whose worst case emission factors are of concern. The actual
3 3
vanadium emission factor is 0.141 gm/m x 0.0104 =1.5 mg/m which is
3
1.5 yg/m at ground level or 88 percent of its MEG. This is equivalent
to 3.4 percent of the vanadium content of bitumen. This finding is very
critical because the vanadium emission factor might increase during pro-
longed operation with the stone transfer system clogged. Thus, the claim
that bed material accumulates almost 100 percent of the fuel vanadium is
somewhat misleading because particulate emissions which are representative
of bed stone contain this significant quantity of vanadium. Cadmium and
nickel emission factors are much less than their MEGS and, hence, need no
additional control.
To pursue the nature of the particulate emissions further, ESCA spectra
were taken of these stack cyclone particulates as well as of material
caught by the hi-vol filter and that deposited on each stage of the im-
pactor. Figures 41 and 42 are broadband scans of stack cyclone (SC9) and
filter (FS9) pa'rticulate. It is apparent that both samples are heavily
coated with carbonaceous material. This coating is the result of in-
complete combustion coupled with deposition of organic material at all
stages of the process, particularly in the cooler stack region. Table 36
summarizes the surface elemental abundances of these samples as well as
results of scans of each impactor substrate stage (particulate size de-
creases from UW91 to UW98). In addition to analyzing surface properties,
o
several impactor substrate samples' were sputtered down to ~ 100 A and
rescanned. The results of these spectra are labelled "subsurface" in
127
-------
fo
CO
3
>%
w
U
o
uT
o
o
600
480
360 240
BINDING ENERGY, eV
120
Figure 41. Stack cyclone material from bitumen gasification, Run No. 5.
Broadband ESCA scan
-------
FS9
o
u
o
u
600
480
360 Z40
BINDING ENERGY, eV
120
Figure 42.
Stack sampling train filter material from bitumen gasification,
Run No. 5. Broadband ESCA scan
-------
Table 36. The subsurface scans indicate that the bulk of the carbon on
the small particulate is on or near the surface, reinforcing the hypo-
thesis that a significant portion of the organic material condenses in
the stack. (The results for the smallest particulate, UW98, are anomalous
in this regard. This may indicate that these particles are in fact mostly
carbonaceous material, rather than attributed stone.) It is also inter-
esting to note the relatively high surface sulfur concentrations. Again
this could be due to condensation of sulfur oxides in the stack.
Vanadium surface and subsurface concentrations are on the order of bulk
values (see Table 35), with subsurface values appearing higher due to
removal of surface carbon. Sodium is considerably more abundant on the
surface than in bulk. Surface enrichment of sodium is well known and
is due to vaporization of sodium compounds in the gasifier and subsequent
condensation of these species in the cooler stack. A similar surface en-
richment phenomenon has been found for vanadium but is not evident from
the particulate results. However, it will be noted later that surface
vanadium in gasifier bed material and larger particulate (that captured
by the gasifier cyclones) is less than 0.2 percent, thus indicating that
smaller particulate surfaces are preferentially enriched in vanadium.
Also included in Table 36 are the results of broadband scans of
filter particulate collected during Run 7 (FS1). The surface abundances
on sample FS1 are almost identical to those from FS9. The similarity
between subsurface and surface abundances on FS1 differs from the results
of the impactor substrate studies but is consistent with the results of
filter particulate collected during fuel oil gasification (which
is discussed later in this section).
To determine the compound form of vanadium the cyclone and filter sam-
ples were scanned over the biriding energy range corresponding to ejection
of the 2p electron of vanadium. These spectra are shown in Figures 43
and 44. Comparison of these spectra with standards V^O,- and vanadium
metal (Figures 15 and 16) indicates that a mixture of oxides presumably
130
-------
Table 36. SURFACE AND SUBSURFACE CONCENTRATIONS OF STACK PARTICULATE
COLLECTED DURING BITUMEN GASIFICATION
Element
0
V
M
C
Ha
S
Ca
SC9a
Surface
12.8
1.1
-
80.8
0.8
3.1
1.5
Sample, Z abundance
PS9a
Surface
34.6
2.4
-
49.8
2.9
7.7
2.6
UW9ia
Surface
34.2
0.7
3.3
52.4
1.4
7.9
-
Sub-
surface
61.0
1.0
-
31.5
0.9 .
4.2
1.4
BW92a
Surface
28.8
0.4
2.7
60.7
1.0
6.4
-
UW93"
Surface
37.4
0.5
2.3
50.1
1.3
8.3
-
Sub-
surface
67.9
0.9
-
26.1
0.9
2.9
1.2
aRun 5.
bRim 7.
UW948
Surface
32.1
0.6
3.1
56.2
1.5
6.5
-
OW95a
Surface
34.4
0.4
2.2
55.1
1.1
7.0
-
Sub-
surface
65.9
0.9
-
28.1
0.8
2.8
0.7
DW96a
Surface
32.1
0.4
3.2
55.6
1.6
7.1
.-
OW97a
Surface
35.4
0.5
2.7
52.9
1.6
6.9
-
Sub-
surface
63.8
1.3
-
30.5
1.3
2.4
0.7
UW988
Surface
14.1
0.3
-
82.4
0.8
2.5
-
Sub-
surface
10.5
1.9
-
83.9
0.8
2.8
-
FSlb
Surface
36.0
2.1
-
48.5
2.5
6.2
4.8
Sub-
surface
33.1
1.3
-
51.5
1.8
'5.1
6.9
-------
SC9
c
3
o
UJ
CJ
to
o
o
J_
540
532
524 516
BINDING ENERGY, eV
508
Figure 43. Stack cyclone material from bitumen gasification, Run No. 5.
Vanadium ESCA scan
500
-------
o
w
2a
o
ul
oc
o
54O
532
524 516
BINDING ENERGY, eV
508
Figure 44. Stack sampling train filter material from bitumen gasification,
Run No. 5. Vanadium ESCA scan
500
-------
V 0,., V~0- and V0? are present. From an environmental impact perspective,
additional specificity is unimportant because all three compounds are
equally toxic. However, the surface content of vanadium (1 to 2 per-
cent) is of particular concern in that it is reasonable to assume that
surface vanadium compounds on particles embedded in the lung will attack
tissue more readily than bulk molecules.
The other element, besides vanadium, of environmental interest is sulfur.
Figures 45 and 46 are scans in the sulfur 2p binding energy region. Sur-
face sulfur on the smaller particulate is all bound as sulfate whereas
in particulate captured by the cyclone roughly 75 percent is sulfate and
25 percent is sulfide. This difference is not unexpected if it is assumed
that a substantial fraction of surface sulfur is formed by reaction in
the stack between particulate cations (calcium in this case) and gas phase
sulfur dioxide and trioxide.
Gasifier bed, internal cyclone and knockout baffle material - Particulate
in these three categories were also analyzed by ESCA. Table 37 contains
the results of broadband scans of gasifier bed material (GB9) right hand
gasifier cyclone catch (RH9) and stack knockout baffle particulate (K09).
Table 37. SURFACE CONCENTRATIONS OF GASIFIER BED,
GASIFIER CYCLONE AND KNOCKOUT BAFFLE
PARTICULATE
Element
0
Ca
C
S
Na
Sample, % surface abundance
GB9
47.2
12.9
38.3
1.6
-
RH9
46.8
11.9
40.0
1.3
-
K09
27.6
6.1
63.0
3.0
0.2
134
-------
O
O
i i • i i I T r r
SC9
U)
tn
"c
3
O
UJ
I I
I I
175
171
167 163
BINDING ENERGY, eV
159
155
Figure 45. Stack cyclone material from bitumen gasification, Run No. 5,
Sulfur ESCA scan
-------
FS9
in
'c
3
O
lo
O
LJ
cr
O
O
175
171
167 163
BINDING ENERGY, eV
159
155
Figure 46. Stack sampling train filter material from bitumen gasification,
Run No. 5. Sulfur ESCA scan
-------
Comparison of columns GB9 and RH9 indicates that material captured by
the gasifier cyclones is representative of bed material. Knockout baffle
particulate is more appropriately compared with material captured by the
stack cyclone (SC9). The larger material from the knockout baffle has a
somewhat smaller carbon coating, less surface sodium and surface vanadium
below 0.2 percent as would be expected from the previous discussion.
These findings are similar to .those, encountered for fuel oil gasification
samples.
Spent Stone - Regenerator bed material, representative of CAFB solid
waste, from Run 5 was analyzed for organic components, bulk elements
and surface elements-and compounds. Figures 47 through 54 contain the
spectra of organic material extracted' from bed stone', table 38 summarizes
the spectral identification and gives the distribution of material among
the eight chromatographic fractions. A large variety of compounds are
present in the bed material, roughly one-third hydrocarbons and two-thirds
oxygenated species. The potential environmental impact of these compounds
will depend on the method of disposal and upon the type of predisposal
treatment. The effects of leachate containing compounds such as phenols,
aromates carbonyls and esters.-would have to be determined. However, the
low abundances of most of the compounds found present coupled with pro-
per disposal does not appear to present any readily apparent deleterious
environmental effects.
Bulk elemental analysis of bed material (now shown) indicates that of
the major metal elements found in bitumen V, Ni and Na have much lower
abundances in this material t.han in stack particulate. Only iron has
the same concentration in both samples. This finding is consistent with
the mechanisms proposed earlier for enrichment of V and Na in small
particles.
.The bulk analysis is confirmed by the ESCA scan shown in Figure 55
which shows neither vanadium nor sodium. Thus the surface abundances
of both these elements is less than 0..1 percent in the bed stone.
137
-------
2.5
WAVELENGTH, micron*
5 6
7 8 9 10 12 15 20 3040
I I I 1 J I i 1 i
too
c
I 80
Ul
z 60
-x
k-
t-
2 40
| 20
0
RB9-I
4000 3500 3000 2500 2000 1800 1600 1400 1200 1000 800 600 400 200
FREQUENCY,em-l
Figure 47. Regenerator bed material from bitumen gasification,
Run No. 5, LC fraction 1. Peaks at 2920 cm"1 and
~1370 cm~i and 1450 curl are from CH CH2 and in-
dicate aliphatic hydrocarbons
2.5
WAVELENGTH, micron*
45 6
8 9 10 12 15 20 30 40
4000 3500 3000 2500 2000 1800 1600 1400 1200 1000 800 600 400 200
FREQUENCY, cm-'
Figure 48. Regenerator bed material from bitumen gasification,
Run No. 5, LC fraction 2. See Figure 47
138
-------
WAVELENGTH, microns
5 6
8 9 10 12 15 20 30 40
3500 3000 2500 2000 I BOO 1600 1400 1200 1000 800 600 400 200
FREQUENCY, era-'
Figure 49. Regenerator bed material from bitumen gasification,
Run No. 5, LC fraction 3. See Figure 47. The peak
at 1730 cm indicates an ester, C=0
2.5
100
"c
if 80
V
a
S 60
z
t 40
a
CO
5 20
WAVELENGTH .micront
^ 6
7 8 9 10 12 15 20 3040
i i i i i i i i i
4000 3500 3000 2500 2000 1800 1600 1400 1200 1000 800 600 400 200
FREQUENCY, em-'
Figure 50. Regenerator bed material from bitumen gasification,
Run No. 5, LC fraction 4. See Figure 47. Peak
at 1730 cm~l indicates the carbonyl group, C=0. Peak
at ~3400 cm~l could be from alcohol or carboxylate
139
-------
2.5
_ 100
"c
*» •
I 80
•
ul
i 60
a 40
CO
« 20
WAVELENGTH, microns
5 6 7 8 9 10 12 15 20 3040
I i
4000 3500 3000 2500 2000 1800 1600 1400 1200 1000 800 600 400 200
FREQUENCY, enH
Figure 51. Regenerator bed material from bitumen gasification,
Run No. 5, LC fraction 5. See Figure 50
2.5
WAVELENGTH, microns
7 8 9 10 12 15 20 30 40
• I.. 1 1 1 1—1_
100
Ti
5 80
«*
A
a 60
<
f 40
a
VI
5 20
oe
RB9-6
4000
3500 3000 2500 2000 1800 1600 1400 1200 1000 800 600 400 200
FREQUENCY, cm''
Figure 52. Regenerator bed material from bitumen gasification,
Run No. 5, LC fraction 6. Peaks at 2920 cm"1 are
from CH3, CH2 and the peak at 1730 cm"1 is from C=0.
The peak at 850 cm"1 and the number of bands be-
tween 1000 and 1600 cm"1 suggest the presence of aro-
matic carbonyl compounds. Peak at 3400 cm"1 suggests
phenol or carboxylic acids
140
-------
2.5
WAVE LENGTH .microns
5 67
9 10 12 15 20 3040
100
.80
60
40
5* 20
I I L.
RB9-7
4000
3500 3000 2500 2000 1600 1600 1400 1200 1000 800 600 400 200
FREQUENCY, cnH
Figure 53. Regenerator bed material from bitumen gasification,
Run No. 5, LC fraction 7. Trace quantities of ali-
phatics and carbonyl compounds may be present
WAVELENGTH, mieroni
8 9 10 12 15 20 30 40
°r
4000 3500 3000 2500 2000 1800 1600 1400 1200 1000 800 600 400 200
FREQUENCY, em-'
Figure 54. Regenerator bed material from bitumen gasification,
Run No. 5, LC fraction 8. See Figure 52
141
-------
Table 38. DISTRIBUTION OF EXTRACTABLE ORGANIC MATERIAL AND
FUNCTIONAL GROUPS IN SPENT STONE: RUN NO. 5
Fraction
1
2
3
4
5
6
7
8
Total
Weight, yg
330
64
82
110
85
290
57
210
1,228
Percentage
26.9
5.2
6.7
9.0
6.9
23.6
4.6
17.1
100
Functional groups
Aliphatic hydrocarbons
Aliphatic hydrocarbons
Esters
Carboxylate
Carbonyls, alcohol
As in fraction 4
Carbonyls, aromatic carbonyls ,
phenol, carboxylic acid
Carbonyls
As in Fraction 6
142
-------
RB9
tn
"E
J3
w
O
Ul
cc
z
o
o
I
2S
600
480
360 240
BINDING ENERGY, eV
120
Figure 55. Regenerator bed material from bitumen gasification,
Run No. 5. Broadband ESCA scan
-------
o
Furthermore, scans of spent stone etched to a depth of ~ 100 A did not
show any vanadium. Surface abundances of the four elements observed on
the surface are listed in Table 39. The carbon Is electron scan presented
in Figure 56 shows a substantial concentration of carbonate. This is not
unexpected of a material subject to severely oxidizing conditions and is
consistent with the substantial carbonyl presence found in the organic
analyses and with the finding by ERCA that regenerator bed material is
heavily carbon coated.
Table 39. SURFACE CONCENTRATIONS
OF SPENT STONE PARTI-
CLES, RUN NO. 5 .
Element
0
C
Ca
S
Abundance, %
18.7
74.2
5.1
2.0
It is also interesting to note that surface sulfur, shown in Figure 57,
is all in the form of sulfide. This is presumably a reflection of the
particular conditions of temperature, oxygen feed rate and past history
of this stone. Later in this section it is noted that sulfur on spent
stone collected during oil gasification during Run 4 is evenly distribu-
ted between sulfate and sulfide. Because regenerator bed stone will
undergo further treatment before disposal or sale, the state of surface
sulfur on stone leaving the regenerator is not directly of environmental
importance.
Bitumen Combustion/Start-Up
The only sample available for laboratory analysis during bitumen combustion
was the Method 5 filter. Because of the small quantity of particulate
collected surface analysis was the only technique employed to characterize
144
-------
RB9
3
>»
w
O
!o
w
o
Ul
O
o
I
295
291 287 283 279
BINDING ENERGY, eV
Figure 56. Regenerator bed material from bitumen gasification,
Run No. 5. Carbon ESCA scan
275
-------
i-
z
3
O
O
I 1 I
RB9
c
3
O
ui
I
175
171
167 163
BINDING ENERGY, cV
159
155
Figure 57. Regenerator bed material from bitumen gasification,
Run No. 5. Sulfur ESCA scan
-------
this sample. Figure 58 is a broadband ESCA scan of the stack sampling
train filter particulate. Table 40 lists the surface abundance cal-
culated from this scan.
..Table 40. SURFACE CONCENTRATIONS OF STACK PARTICULATE RUN NO. 6
Element
0
C
Ca
Na
S
Abundance, %
40.4
49.3
8.5
1.0
1.9
The most striking feature of this spectrum is the absence of vanadium.
It may be noted in Tables 26 and 27 that the temperature at the stack
sampling port is almost 60°C lower in Run 6 than in Run 5. It can there-
fore be assumed that the temperature in boiler and cyclones was con-
siderably lower in Run 6. Therefore, vanadium oxide condensation on par-
ticulate surfaces occurred well before the stack and was largely covered
up by subsequent deposition of other species such as C02 reacting with
lime to form CaCOo-
Evidence for this explanation is provided in Figure 59 which shows
that a substantial portion of surface carbon (-25 percent) is in the form
of carbonate. An additional factor contributing to the relatively high
proportion of surface carbonate is increased combustion efficiency in the
CAFB under conditions of high excess air. Finally, Figure 60 shows
that essentially all surface particulate sulfur is bound as sulfate.
This is expected under combustion conditions because of the low probabi-
lity of calcium sulfide formation.
147
-------
FSO
-P-
oo
c
3
w
O
UJ
55
(E
O
o
600
480
Co2p
IS
360 240
BINDING ENERGY, eV
02S
120
Figure 58. Stack sampling train filter catch during bitumen combustion and
fresh limestone feeding, Run No. 6. Broadband ESCA scan
-------
FSO
o>
•*-
"E
3
VO
o
o
J_
_L
_L
295
291
287 283
BINDING ENERGY, eV
279
275
Figure 59. Stack sampling train filter catch during bitumen combustion and
fresh limestone feeding» Run No. 6. Carbon ESCA scan
-------
FSO
c
3
jQ
O
UJ
IT
I-
O
O
175
171
167 163
BINDING ENERGY. eV
159
155.
Figure 60. Stack sampling train filter catch during bitumen combustion and
fresh limestone feeding, Run No. 6. Sulfur ESCA scan
-------
Fuel Oil Gasification
Stack Particulate - Stack cyclone particles collected during Run 4 were
the only samples from fuel oil gasification analyzed for organic func-
tional groups, bulk elemental composition and surface chemicals. In gen-
eral, the results of the analyses of samples collected during fuel oil
gasification are similar to those from bitumen gasification samples.
Figures 61 through 67 contain the infrared spectra of the first seven
chromatographic fractions from the extract of stack cyclone particulate.
Table 41 summarizes the spectral identifications and presents the weight
distribution among the eight fractions. The functional groups identified
in this sample and their relative amounts are similar to that found in
stack cyclone particulate collected during bitumen gasification (see
O
Table 34). The total condensed organic loading 0.2 mg/m is the same
and thus does not appear to represent any significant environmental hazard.
Table 42 contains the results of bulk elemental analysis of the stack
cyclone particles. These results are similar to the bitumen stack par-
ticulate analysis. The ratios between sulfur, vanadium and nickel in
the two sets of particulate are roughly equal to the ratios of those
elements in the two fuels. Fluorine is even more abundant in this sam-
ple than in the bitumen particulate; the hypotheses advanced in that dis-
cussion apply here as well. The substantial chlorine concentration in
these particles is unexpected but lower than a worst case analyses pre-
diction based upon the chlorine composition in limestone. As with bitu-
men emissions, vanadium is the only element of potential concern for the
reasons suggested in that discussion.
A number of particulate samples collected during the 4 days of fuel oil
combustion were investigated using ESCA. Figure 168 is a broadband scan
of stack cyclone particulate (SC8) from Run 4. This spectrum is similar
to that of bitumen stack cyclone material (SC9) in Figure 41. Figures 69
through 71 are detailed scans of vanadium, sulfur and carbon. Surface
151
-------
2.5
WAVELENGTH, micron*
4 56 7 8 9 10 12 15 20 30 40
4000 3500 3000 2500 2000 1800 1600 1400 1200 1000 BOO 600 400 200
FREQUENCY, cnH
Figure 61. Stack cyclone material from fuel oil gasification,
Run No. 4, LC fraction 1. Trace quantities of
aliphatic hydrocarbons
2.5
WAVELENGTH, microns
5 6
7 8 9 10 12 IS 20 30 40
• iii t i i 1—1_
100
80
,
so
40
20
vt
4000
"1 It I I I Till 1 I I i T | •
3500 3000 2500 2000 1800 1600 1400 1200 1000 800 600 400 200
FREQUENCY, cm-'
Figure 62. Stack cyclone material from fuel oil gasification,
Run No. 4, LC fraction 2. Peak at ~2920 cm'1 and
peaks at 1450 cm"1 and 1370 cm"1 are from CH3, CH2,
while the peak at 1730 cm"1 indicates the carbonyl'
group C=0. This suggests the presence of aliphatic
esters. Peaks between 1100 cm'1 and 1500 cm"1 indi-
cate presence of other C=0 containing species
152
-------
2.5
100
g. 80
3 60
40
20
4
i
WAVELENGTH, rojcron»
5 6
7 8 9 10 12 15 20 30 40
i i i i i i i i
4000 3500 3000 2500 2000 1800 1600 1400 1200 1000 800 600 400 200
FREQUENCY, cm'l
Figure 63. Stack cyclone material from fuel oil gasification,
Run No. 4, LC fraction 3. Peaks at ~2920 cm"1 and
~1730 cm"1 indicate presence of aliphatic ester
2.5
WAVELENGTH, microns
5 678
9 10 12
20 30 40
i •
°P
4000 3500 3000 2500 2000 1800 1600 1400 1200 1000 800 600 400 200
FREQUENCY,em-'
Figure 64. Stack cyclone material from fuel oil gasification,
Run No. 4, LC fraction 4. Peak at ~2920 cm"1 and
peaks at 1450 cm"1 and 1370 cm"1 indicate aliphatics,
while peak at 1730 cm indicates C=0. Sample pro-
bably contains aliphatic esters, ketones, or aldehydes.
Broad band between 1000-1100 cm"1 probably comes from
Si02 impurity
153
-------
2.5
100
c
? 80
5 60
z
<
£ 40
20
WAVELENGTH .micron*
7 8 9 10 • 12 15 20 30 40
_i j L-_J I L I I I
4000
I I I 1 1 I I I
3500 3000 2500 2000 1800 1600 1400 1200 1000 800 600 400 200
FREQUENCY,em-'
Figure 65. Stack cyclone material from fuel oil gasification,
Run No. 4, LC fraction 5. Presence of carbonyl
compounds suggested by structure 1730 cm"1
2.5
100
.80
60
40
20
WAVELENGTH, microns
8 9 10 12 15 20 30 40
I I I 1 1 1 L
SC8-6
4000
I f I I I 1 \. I 7 I
3500 3000 2500 2000 1800 1600 1400 1200 1000 800 600 400 200
FREQUENCY, cm-l
Figure 66. Stack cyclone material from fuel oil gasification,
Run No. 4, LC fraction 6. Peaks at ~2920 cm"1,
1450 cm"1, and 1370 cm~l indicate aliphatic esters.
Peaks at ~3400 cm'1 and between 1100-1300 cm'1
suggest presence of carboxylates or alcohols
154
-------
2.5
100
•£
«>
t; eo
•>
Ok
g GO
t 40
i
5 20
WAVELENGTH, microns
5 6 7 6 9 10
12 15 20 30 40
4000
3500 3000 2500 2000 1800 1600 1400 1200 1000 800 600 400 200
FREQUENCY, cm-'
Figure 67. Stack cyclone material from fuel oil gasification,
Run No. 4, LC fraction 7. Peak at 1730 cm suggests
mixture of carbonyl compounds. Broad band between
1000-1100 cm"1 probably from Si02 impurity
155
-------
Table 41. DISTRIBUTION OF EXTRACTABLE ORGANIC MATERIAL
AND FUNCTIONAL GROUPS IN STACK CYCLONE PAR-
TICULATE: RUN NO. 4
Fraction
1
2
3
4
5
6
7
8
Total
Weight, yg
97
210
28
120
180
200
35
0
870
Percentage
11.2
24.1
3.2
13.8
20.7
23.0
4.0
0
100
Functional groups
Aliphatic hydrocarbons,
carbonyls, aliphatic esters,
C=0 contain species
Aliphatic esters,
C=0 species, aliphatic esters
ketones, aldehydes
Carbonyls
aliphatic esters,
carboxylates, alcohols
Carbonyls
. "
156
-------
Table 42. MASS SPECTROGRAPHIC AND ATOMIC ABSORPTION
SPECTROMETRIC ANALYSIS OF STACK CYCLONE
PARTICIPATE: RUN NO. 4
Element3
Cac
Sb
Vc
.Mg
Si
Cl
c
Fec
F
Al
c
Ni
Na
K
Sr
Ba
Ti
P
Zn
Br
Mn
Pb
Cu
Cr
Mo
B
Co
Zr
I
Li
Concentration, ppmw or percent
34.5 %
2.13
0.80
0.32
0.21
0.16
0.15
0.13
0.13
0.10
850 ppmw
340
180
150
150
96
80
60
50
47
44
31
14
5.0
5.0
3.0
2.9
2.1
157
-------
Table 42 (continued),
MASS SPECTROGRAPHIC AND ATOMIC ABSORPTION
SPECTROMETRIC ANALYSIS OF STACK CYCLONE
PARTICIPATE: RUN NO. 4
a
Element
Se
W
Sn
Ce
Cd
La
Ge
Y
Rb
Yb
Ga
Ta
Bi
Hf
Nb
Nd
Dy
Sm
Th
Be
Pr
Tl
Concentration, ppmw or percent
1.3ppmw
1.1
1.0
1.0
0.8
0.8
0.6
0.5
0.5
<0.5
0.4
0.4
<0.3
<0.3
0.2
0.2
0.2
<0.2
<0.2
<0.12
0.1
<0.1
Elements not listed are <0.1
ppm, not detected.
Determined by wet chemistry.
^
Determined by atomic absorp-
tion spectrometry.
Used as internal standard.
158
-------
I I I I I
SC8
o
LL)
O
o
02S
600
480
360 240
BINDING ENERGY, eV
120
Figure 68. Stack cyclone material from fuel oil gasification,
Run No. 4. Broadband ESCA scan
-------
I I I I I I I I
SC8
ON
O
E
3
.0
O
UJ
z
^
O
J_
540
532
Figure 69.
524 . 516
BINDING ENERGY, eV
508
Stack cyclone material from fuel oil gasification,
Run No. 4. Vanadium ESCA scan
500
-------
SC8
c
3
J3
O
uJ
I-
oc
t-
O
u
I
I
175
171
167 163
BINDING ENERGY, eV
159
155
Figure 70.
Stack cyclone material from fuel oil gasification,
Run No. 4. Sulfur ESCA scan
-------
3
>»•
k
O
!o
o
UJ
oc _
ID
O
O
295
291
279
Figure 71.
287 283
BINDING ENERGY, eV
i,
Stack cyclone material from fuel oil gasification, Run. No. 4.
275
Carbon ESCA scan
-------
vanadium is, as was the case during bitumen gasification, present in a
mixture of oxides. Sulfur is all in the form of sulfate, reflecting the
large quantity of excess oxygen (5.3 percent) in the flue gas. The carbon
Is spectrum is asymmetric toward the higher binding energy region reflect-
ing the high proportion of oxygenated organic species found by the organic
analysis.
Broadband scans of particulate captured by the filter in Runs 1 to
4 were also taken. Figures 72 and 73 are broadband surface and subsur-
face spectra of filter particulate collected during Run 4. Spectra of
samples from Runs 1 to 3 are not shown but are summarized in Table 43.
Surface and subsurface elemental compositions in Runs 1, 2 and 4 are
remarkably similar. The filter catch in Run 3 has a substantially larger
carbon abundance which, in fact, increases slightly from the surface to
o
a depth of ~ 100 A. This anomalous behavior does not correlate to any
other emission properties measured during Run 3.
Table 44 lists the elemental abundances found on particulate collected by
the impactor. As was the case with bitumen samples, impactor particulate
is more heavily carbon coated than particulate collected on the hi-vol
sampling train filter. This may be merely a reflection of the heavier
carbon coating on the pure aluminum substrates than on the filter material
(see Figures 13 and 14), since impactor substrate coverage is light. The
appearance of silicon on several of the substrates is most likely due to
left over silicon grease from early impactor runs. The most surprising
observations are the lack of any detectable vanadium, even on the sub-
surface scans and the relatively small quantities of sulfur and sodium.
Taken together, these results suggest that substantial deposition of car-
bonaceo-us material occurs in the impactor as the flue gas cools
Spent Stone - The only analyses of regenerator bed material are the sur-
face spectra shown in Figures 74 through 76. Relative surface element
abundances listed in Table 45 are identical to those found in bitumen
spent stone. The carbonate/organic carbon ratios are also the same. The
163
-------
FS8
to
O
o>
'£
3
O
a
I
1000
800
600
BINDING
400
ENERGY,eV
200
Figure 72.
Stack sampling train filter material from fuel oil gasification,
Run No. 4. Surface broadband ESCA scan
-------
FS8
5 min SPUTTER
CO
O
(0
'c
3
O
I—
O
UJ
55
tr.
z
13
O
O
1000
800
600 400
BINDING ENERGY, eV
200
Figure 73.
Stack sampling train filter material from fuel oil gasification,
Run No. 4. Subsurface broadband ESCA scan
-------
Table 43. SURFACE AND SUBSURFACE CONCENTRATIONS OF STACK PARTICIPATE
COLLECTED DURING FUEL OIL GASIFICATION; RUNS 1 TO 4
Element
0
V
Ca
C
Na
S
F
C£
N
Sample, % abundance
SC8a
Surface
22.2
0.4
2.9
68.0
1.0
5.4
-
-
-
FS8a
Surface
39.8
1.7
4.7
43.3
3.3
7.2
-•
-
-
Sub-
surface
38.9
1.6
8.5
42.3
2.5
6.2
-
-
-
FS6b
Surface
19.3
0.9
0.8
70.8
2.1
. 4.5
0.8
-
0.9
Sub-
surface
12.4
0.8
1.8
79.8
1.6
3.6
-
-
-
FS5&
Surface
43.2
1.6
6.1
37.2
3.1
8.9
-
-
-
Sub-
surface
47.7
1.2
9.9
30.2
3.9
8.2
-
0
-
FS4d
Surface
41.1
0.6
5.3
40.5
2.5
8.3
1.7
-
-
Sub-
surface
44.5
0.3
10.4
36.6
1.6
4.5
1.2
0.9
-
Run 4.
5Run 3.
:Run 2.
Run 1.
-------
Table 44. SURFACE AND SUBSURFACE CONCENTRATIONS OF PARTICIPATE COLLECTED ON IMPACTOR
SUBSTRATES: RUN NO. 4
Element
0
Ca
C
Na
S
N
F
Si
ca
V
Sample, % abundance
UW61
Surface
22.4
-
72.6
0.8
1.6
1.5
-
1.1
-
-
UW62
Surface
20.5
-
75.8
0.4
1.4
-
-
1.9
-
-
UW63
Surface
25.2
0.6
69.3
0.9
1.6
1.3
0.3
0.7
-
• -
Sub-
surface
70.4
1.7
24.0
0.9
2.2
-
-
-
0.8
-
UW64
Surface
24.8
0.3
70.2
0.9
1.1
1.9
-
-
1.0
-
UW65
Surface
22.7
-
73.7
0.6
1.1
0.9
-
1.0
-
-
UW66
Surface
33.9
-
60.8
0.9
2.1
1.8
0.5
-
-
-
Sub-
surface
39.2
-
58.6
0.7
1.5
-
-
-
-
-
UW67
Surface
25.9
-
69.1
0.2
1.1
1.3
-
1.3
0.6
-
UW68
Surface
10.8
85.4
1.2
2.1
-
-
-
-
0.4
-------
RB8
oo
JD
O
LU
1-
o:
f-
2
O
O
^-v^-\
Ca
2S
Ca.
"IS
'2P
600
480
360 Z40
BINDING ENERGY, eV
120
Figure 74. Regenerator bed material from fuel oil gasification,
Run No. 4. Broadband ESCA scan
-------
c
3
Till
RB8
o
Id
t-
-------
RB8
u>
O
o
295
291
287 283
BINDING ENERGY, eV
279
Figure 76. Regenerator bed material from fuel oil gasification,
Run No. 4. Carbon ESCA scan
275
-------
only difference between the two spent stone samples is that surface sul-
fur here is distributed evenly between sulfate and sulfide. As noted
earlier, this distinction is of modest environmental interest.
Table 45. SURFACE CONCENTRATIONS OF
SPENT STONE PARTICLES:
RUN NO. 4
Element
0
C
Ca
S
Abundance, %
19.8
73.4
5.2
1.6
Leached Stone
The three leached stone samples collected from the outdoor buckets were
analyzed by ESCA for surface and subsurface elements. Results of these
analyses are presented in Table 46.
Table 46. SURFACE AND SUBSURFACE ELEMENTAL COMPOSITIONS OF
LEACHED STONE SAMPLES
Sample, % abundance
Element
0
Ca
C
S
Si
SSI
Surface
52.8
12.0
33.2
1.8
-
Sub-
surface
55.8
15.1
26.2
1.6
-
SS3
Surface
52.4
12.3
33.3
1.9
-
Sub-
surface
59.0
18.3
21.3
1.4
-
SS5
Surface
53.8
13.3
31.0
1.9
-
Sub-
surface
63.1
20.6
15.1
1.3
1.3
171
-------
Comparison of these results with the analyses of spent stone in Tables 39
and 45 shows that these samples have a much lower carbon coating and an
equivalent surface concentration of sulfur. Because sintered, but unleached
stone was not available for analysis, effects of leaching and sintering
cannot be separately evaluated.
SUMMARY
Boiler stack gas and stack particulate emissions and solid waste efflu-
ents from fuel oil gasification, bitumen gasification and bitumen com-
bustion were sampled and analyzed. The following points summarize the
results of environmental interest.
• Stack NOX emissions are consistently much lower than
New Source Performance Standards (NSPS).
• Under normal operating conditions SO emissions are
lower than NSPS.
• Saturated gasifier stone causes SOX emissions to exceed
NSPS.
• Under normal operating conditions particulate emissions
are just barely lower than NSPS.
• During fresh stone feeding particulate emissions exceed
NSPS.
• Vanadium is the only trace element whose emission rate
presents a potential problem.
• Stack gas and particulate organic emission rates do not
present a potential problem.
• Fugitive air emissions from bitumen storage and handling
may contain POM.
172
-------
REFERENCES
1. Dorsey, J.A., C.H. Lochmuller, L.D. Johnson, and R.M. Statnick.
Guidelines for Environmental Assessment Sampling and Analysis
Programs. Environmental Protection Agency, Research Triangle
Park, N.C. Draft Final. March 1976.
2. Standards of Performance for New Stationary Sources. Code of
Federal Regulations, 40 CFT, Part 60, May 23, 1975.
3. Jones, P.W., A.P. Graffeo, R. Detrick, P.A. Clarks, and R.J. Jakobsen.
Technical Manual for Analysis of Organic Materials in Process
Streams. Battelle Columbus Laboratories, Columbus, Ohio. Environ-
mental Protection Agency, Research Triangle Park, N.C. Report
Number EPA-600/2-76-072. March 1976. 81 p.
4. Berthou, H. and C.K. Jorgensen. Relative Photoelectron Signal
Intensities Obtained With a Magnesium X-Ray Source. Anal Chem.
47:482-488. March 1975.
5. Title 40-Protection of the Environment. Part 60 Standards of
Performance for New Stationary Sources. Federal Register
36(247)-.24876. December 23, 1971.
6. Jonke, A.A., E.L. Carls, R.L. Jarry, M. Haas, W.A. Murphy, and
C.B. Schoffstoll. Reduction of Atmospheric Pollution by the Applica-
tion of Fluidized-Bed Combustion. Argonne National Laboratory,
Argonne, Illinois. Report Number ANL/ES-CEN-1001. June 1969. p. 65.
7. Fennelly, P.F., D.F. Durocher, A.S. Werner, M.T. Mills, S.M. Weinstein,
A.H. Castaline, and C.W. Young. Environmental Assessment Perspectives.
GCA/Technology Division, Bedford, Mass. U.S. Environmental Protection
Agency, Research Triangle Park, N.C. Report Number EPA-600/2-76-069.
p. 239. March 1976.
8. Gafafer, W.M. (ed). Occupational Diseases, A Guide to Their
Recognition, U.S. HEW, Public Health Service Publication No.
PHS-1097. Reprinted June 1966, U.S. Government Printing Office,
Washington, D.C.
9. Lucrey, T.D., B. Venugopal and D. Hutcheson. Heavy Metal Toxicity
Safety and Hormology. George Thieme Publishers, p. 56-57. 1975.
10. Linton, R.W., A. Loh, D.F.S. Natusch, C.A. Evans, and P. Williams.
Surface Predominance of Trace Elements in Airborne Particles.
Science 191:852-854. February 27, 1976.
11. Waters, M.D., D.E. Gardner and D.L. Coffin. Cytotoxic Effects of
Vanadium and Other Metals In Vitro. Paper presented at the Twelve
Annual Meeting Society of Toxicology. New York, New York.
March 18-22, 1973.
173
-------
SECTION VI
CAFB AIR QUALITY IMPACT ASSESSMENT
FOR THE LA PALMA RETROFIT
INTRODUCTION
In a comprehensive environmental assessment the next step after compiling
the emissions inventory is to calculate the incremental loadings to the
local ambient air, water and soil resulting from the output of all pro-
cess waste streams. These incremental loadings should then be compared
with known human health and ecological effects data in order to assess
the environmental acceptability of the process. A complete environmental
impact evaluation of this sort is well beyond the scope of this prelimi-
nary study.
Rather than being an attempt at a full superficial environmental impact
analysis, this Section presents a discussion of meteorological and topo-
graphical characteristics of an area which influence the transport of
pollutants emitted from a point source. Special emphasis is placed upon
the most significant parameters for the La Palma Power Station. This
general review of dispersion characteristics in the vicinity of the plant
is followed by a detailed diffusion modeling analysis of the expected SO-
and particulate levels after the installation of the CAFB. Finally, the
results of this analysis are compared with Texas emission and ambient air
standards.
174
-------
VARIABLES AFFECTING AMBIENT CONCENTRATIONS
Emission Characteristics
There are a number of point source emission characteristics which affect
the subsequent transport for a given pollutant. These include the stack
height, stack diameter, gas exit velocity and gas temperature. These
parameters, together with the ambient temperature and stability index,'
are used in the calculation of buoyant plume rise, which is responsible
for a greater degree of plume dilution due to an increased effective
stack height. A greater source height will also ensure a lesser degree
of plume depletion due to dry deposition upon the ground surface. It is
sometimes necessary to study the relationship of each source to nearby
structures in the area due to their influence in the processes of plume
rise retardation or downwash.
Another emission characteristic of interest would be the time variation
of the emission rate over a daily, weekly or seasonal period. For exam-
ple, if a given sector of the population is sensitive to short term
episodes of elevated concentrations, then an hourly distribution of emis-
sion rates would be of interest, whereas the long term effects of wet
and dry deposition of pollutants in the vicinity of a source would re-
quire only average annual emission values.
Topographical Characteristics
The general nature of the landscape will exert a significant influence
upon the atmospheric transport of pollutants. The channeling of atmos-
pheric pollutants by topographical features such as ridges and valleys
is a well known phenomenon. Areas situated near a lake or ocean will in
general experience a lower dilution of pollutants due to the resulting
higher atmospheric stability when the wind blows across the land from the
175
-------
cooler body of water. Areas having a greater elevation than the base of
a stack will usually be exposed to greater pollutant concentrations.
Since the area surrounding the La Palma plant is characterized by rather
flat terrain, the pollutant transport process is not likely to depend
upon topography.
Climatological Characteristics
For the Lower Rio Grande Valley of Texas, the most significant climato-
logical characteristics in terms of transport and diffusion are the pre-
vailing wind direction and the amount of solar insolation. Figure 77
illustrates the frequency distributions of surface wind direction for a
number of weather stations in the U.S. The annual "wind rose" for Browns-
ville, Texas, the closest station to the La Palma facility, shows that
for the most part the winds are confined to the east-southeast, southeast
and south-southeast directions. The few occasions when northwest winds
are present are confined exclusively to the winter months. The strong
southeasterly flow off the Gulf of Mexico is driven by high pressure off
the southeastern coast and the northeast gulf and is reinforced by a sea-
breeze which develops during the late morning. The fact that the wind is
predominantly from the southeast quadrant will lead to elevated concen-
trations northwest of the plant for averaging times greater than 1 hour.
the solar radiation incident at the surface of the earth is another
parameter required for the analysis of concentration levels due to point
source emissions. The annual mean daily solar radiation (Langleys) is
given in Figure 78 for a number of weather stations throughout the country.
The value of 442 Langleys reported for Brownsville would argue for a higher
frequency of unstable atmospheric conditions than for stations in the
northeastern part of the country. As indicated later in this Section
(Figures 84 through 89), these unstable conditions will lead to higher
concentrations for receptors located near the plant and reduced levels at
greater distances from the source.
176
-------
SURFACE WIND ROSES. ANNUAL
n 4f-
( Hint «ot^.x .1/y
•EZG
Figure 77. Annual surface wind roses
-------
oo
'r<^r-•••••---^ '~^~~:T~~~~-~MEAN DAILY SOLAR RADIATION (Langieys), ANNUAL-
i >7"|, '^:*7'S< ''3'r'">>^irh-------"^~-.J£
"l--. ' ••"°'i' /•.... j?j:' ••^^-_^^.';p,v.'
•• j j-
-------
As indicated by the modeling analysis (see below), the predicted con-
centrations will be quite sensitive to windspeed because this parameter
is used in the determination of plume rise, stability class and the
amount of plume dilution. On an annual basis, however, the surface
windspeed will average about 5.4 m/s (12 miles per hour) (see Figure 79)
which is only slightly greater than the average for the entire country.
The remaining two climatological variables which will affect pollutant
dispersion in the vicinity of the La Palma site are the mixing depth and
the ambient temperature. The mixing depth may be roughly defined as the
atmospheric boundary layer near the earth's surface in which the turbulent
diffusion mechanisms predominate. In response to daytime heating of the
land surface, the depth of this layer may exceed 1 or 2 kilometers, but
will be considerably reduced during nighttime hours. The top of this
layer, marked by a discontinuity in the temperature profile, acts as a
barrier to the vertical migration of material released within the layer.
The mixing height data presented in Figures 80 and 81 indicate that the
La Palma site is characterized by lower afternoon mixing depths and greater
morning mixing depths than the country as a whole. The actual manner in
which these mixing depths enter into the dispersion calculations is ex-
plained in the modeling analysis section. The relatively high ambient
temperature for the La Palma site (annual average, 23 C (74 F)) will re-
sult in a slightly reduced plume rise as compared with areas in the north-
ern half of the U.S., but this effect would mean only a few percent change
in the annual concentration.
DISPERSION MODELING ANALYSIS
Description of Modeling Techniques
This section addresses the quantitative evaluation of short-term S02
and suspended particulate (TSP) levels in the vicinity of the La Palma
facility for the CAFB configuration. In the case of SOo, worst 1-hour
179
-------
r " /«•>£•$" "-'---^ -~~"--:-^l_J> RE VAILING "DlRECTl6N'~AND MEAN SPEED "(M.P .H .")" OF "WINDj
1 i '/£-..,,. , ""•7r-:-J-._ "t— —i ! ANNUAL ! '
! > ^fi "^ f-c "~"~^-—-~i^...
00
o
I, ^f I-'^LJ^ Ifr "\j I ^
^
/ T~< /, : /^-i; \S~H\ \ ! |i \. x
•S. / •"" ,'ih. / '-—-V "I -X~_ I i •<. n\ :>
~. ^r ?')%' ' 6 ,'-r-\ i MSnJllM
, r---^ i N ^. / r I f^TTrr1"^
•^.-. ^-i-f^-i is*-*-- U-'-~,-o4—:
"•-I '^""s ~~~~~~——--X__; ••'"1 r., /" —C^..~i >.-m,niu i' |'^>
p^' \v. !/». rfcr^i-f—i---* -4-,.fip
lO
>
L
El^
'0
l»w*
^v^i"
^^¥
\
\
I U-^—8
r*-*riLiiu A- ..--^^f. *J
tot >*«_-—_-:*ari-y*
r"'/x;H>^ JU1^*
.Ci ' WM..W- >0
V'or* W •>
^.A;r°\-
£..-*.
1 •• jf* *^
-1 '•*-'
jr\\
.'»-\\
/ N Out
-J \
*--\
'10 6-
c.^.
sv
-vi
1 = « \
*#\
S-^V-
or - «
-i NOTE:
11
UL:
Arrows fly with wind.
Figure 79. Annual mean windspeeds and resultant wind directions
-------
oo
6 5 4
7
Figure 80. Isopleths (m x 10 ) of mean annual morning mixing heights'
-------
00
1X5
8 12
14 12
Figure 81. Isopleths (m x 10^) of mean annual afternoon mixing heights''
-------
predicted concentrations can be extrapolated to highest 3-hour and
24-hour averages through the application of peak to mean ratio statis-
tics derived from an analysis of air quality data collected in the
vicinity of isolated point sources. Highest 24-hour averages for TSP
can be determined in the same manner. The expression used for evalua-
tion of 1-hour pollutant concentrations downwind of a point source is
3 4
the Gaussian plume equation ' given by
Q exp
-y
X(x,y,z)=
2a
(x)]
2 IT a (x) oz(x) u
exp I-
exp-
2a (x)
(1)
where
x = distance along plume axis (m)
y = horizontal distance from plume axis (m)
z = distance above surface (m)
o
X(x,y,z) = concentration of pollutant (g/m )
ay(x), az(x)
effective emission rate of pollutant distance x
(g/sec)
horizontal and vertical dispersion coefficients for
a particular atmospheric stability (A,B,C,D,E,F)
u = windspeed at source height (m/sec)
h(x) = effective emission height at distance x (m).
The variation of a and a with x for each of the six stability classifi-
y z 3
cations (A to F) has been determined from a number of experiments based
upon low level releases of tracer material and do not strictly apply to
elevated sources or for downwind distances greater than about 5 km. The
usual procedure, however, is to assume that these results are approxi-
mately true for greater source heights and that they may be extrapolated
to longer distances. The choice of a given stability will depend upon
windspeed, cloud cover and sun elevation. The basis for the selection
183
-------
is given in Table 47. The variation of a and o with distance is shown
in Figures 82 and 83. The second exponential term in brackets on the
right side of Equation (1) is an "image" point source contribution which
is required to meet the zero flux boundary condition at the ground sur-
face (z = 0). The effective source strength Q.(x) will be different than
the strength Q.(Q) at the point of emission due to wet deposition, dry
fallout, and chemical transformation. The effective stack height h(x)
will be greater than the actual 'stack height h due to the buoyancy of the
plume. The expression for h(x) for stabilities A through D is given by
h(x) = hQ + Ah (2)
where Ah = 1.6F1/3 u"1 x2/3 for x g 3.5x*
Ah = 1.6F1/3 u'1 (3.5x*)2/3 for x > 3.5x*
x* = 14F5/8 when F < 55 m4/sec3
x* = 34F2/5 when F £ 55 n4/sec3
2 /Ts - Te
F = gwr I ^—-
2
g = gravitational acceleration (m/sec )
w = stack gas ejection velocity (m/sec)
r = radius of stack (m)
T = stack gas temperature (°K)
S
T = air temperature (°K).
184
-------
Table 47. RELATION OF PASQUILL STABILITY CLASSES
TO WEATHER CONDITIONS
Surface wind
speed (at 10 m) ,
m sec~l
< 2
2-3
3-5
5-6
> 6
Day
Incoming solar radiation
Strong
A
A-B
B
C
C
Moderate
A-B
B
B-C
C-D
D
Slight
B
C
C
D
C
Night
Thinly overcast
or
> 4/8 low cloud
E
D
D
D
£ 3/8
cloud
F
E
D
D
The neutral class, D, should be assumed for overcast conditions
during day or night.
A - Extremely unstable
B - Moderately unstable
C - Slightly unstable
D - Neutral
E - Slightly stable
F - Moderately stable
Pasquill stability classes o
A, extremely unstable 25.0°
B, moderately unstable 20.0
C, slightly unstable 15.0°
D, neutral 10.Oc
E, slightly stable 5.0C
F, moderately stable 2.5C
185
-------
z
Ul
o
o
o
in
cc.
a.
CO
z
o
M
ac
o
4»IO°
A-EXTREMELY UNSTABLE I
B-MODERATELY UNSTABLE -
C-SLIGHTLY UNSTABLE
D-NEUTRAL
E-SLIGHTLY STABLE
F- MODERATELY STABLE -
DISTANCE FROM SOURCE ,m
Figure 82. Horizontal dispersion coefficient as a function
of distance for Pasquill's stability
186
-------
z
UJ
o
UJ
o
o
z
o
CO
a
ui
a
to
4
O
3
2
10
10
10'
10
10'
A-EXTREMELY UNSTABLE
B-MODERATELY UNSTABLE
C-SLIGHTLY UNSTABLE
D-NEUTRAL
E-SLIGHTLY STABLE
F-MODERATELY STABLE
il
I i
1 I I i i
IOJ
10*
10-
DISTANCE FROM SOURCE, m
Figure 83. Vertical dispersion coefficient as a function
of distance for Pasquill's stability types^
187
-------
For stability classes E and F the plume rise becomes
1/3
- 2.9 (£-) (3)
(=)
u g d9
where s = i- ^
e
~ = 0.02 °K/m for stability E
dz
£§. = 0.035 °K/m for stability F
dz
The windspeed (u) at source height (h ) may be related to the windspeed
(u ) measured at a standard distance (h ) above ground level according to
the following power law:
/h \
U = U ( 7— I
m\hm/
(A)
where the exponent p depends upon the stability class.
The presence of the mixing boundary may be accounted for by the incorpora-
tion of multiple image sources as was done to satisfy the zero flux con-
dition at ground level. Equation (1) may then be generalized to give
188
-------
Q exp
2
-y.
X(x,y,z) =
2ir a (x) a (x)
y ^
exp I - -'
(z - h(x))2
2 a/(x)
+ exp
(\ n
(z + h(x))2\ + y^
2 a 2(x) / L-J
I 1=1
exp I -
(z - h(x) - 2jL)'
2 az2(x)
exp _
+ exp I-
(z + h(x) - 2jL)'
2 a 2(x)
(z + h(x) + 2jL)'
2 a 2(x)
z
L. (z - h(x) + 2jL)'
2 a 2(x)
Z
(5)
where L = depth of the mixing layer (m)
n = number of images considered
In practice, only the first few image terms contribute significantly to
the overall ambient concentration. For distances greater than 2 x ,
where x^ is given by a (x ) = 1.6L, Equation (5) may be approximated by
L Z L
Q exp
/21T a (x) u L
y
x > 2 XT
(6)
This discussion of diffusion modeling techniques has so far neglected
the aerodynamic effects caused by buildings adjacent to the stack. Low
exit velocities and the presence of nearby buildings may result in a
189
-------
reduction of plume rise due to the low pressure in the wake of the stack
or building. In certain cases the plume may actually be brought to the
ground a short distance downwind of the stack. Effective stack height
Q
corrections due to these effects have been estimated by Briggs. The
first correction in stack height is due to the stack aerodynamic effect.
h' = h + 2 (w/u - 1.5)D (7)
where h = actual stack height (m)
w = stack gas exit velocity (m/sec)
u = windspeed (m/sec)
D = stack diameter (m)
^<1.5
u —
The next stack height correction will depend upon building height (h, )
and is given by one of the following three expressions:
Case 1.
U" — V, ' 4 -P V, -» 1 ^ li
h" = h' if h > 2.5 h, (8)
Case 2.
h" = 0 if h' <_ 1.5 hb * (9)
Case 3.
h" = 2h' - 2.5 hfe if 1.5 hb < h' <_ 2.5 hb (10)
The standard plume rise correction due to buoyancy effects is then applied
to cases 1 and 3. For case 2 no buoyancy term is added, but an initial
2
dilution volume of cross-sectional area h, is assumed.
b
190
-------
Model Calculations
The following model input parameters will be used for air quality pre-
dictions at the La Palma plant:
Q (source strength): S02> T.SP = 17 g/sec, 2.0 g/sec
r (stack radius at outlet) = 1.14 m
w (stack gas exit velocity) =6.55 m/sec
T (stack gas temperature) = 455 K
s
T (ambient temperature) = 297°K
h (stack height) = 38.7 m
h, (building height) = 15.5 m
D
L (mixing height) = 1300 m.
Figures 84 through 89 display hourly plume centerline SC>2 concentra-
tions calculated by use of Equation (1) and the input parameters listed
above. Concentration predictions are developed for a wide range of at-
mospheric stabilities and windspeeds. The results indicate that maximum
3
1-hour concentrations would be approximately 100 yg/m and that maximum
1-hour particulate levels would be a factor of 8 lower. Because concen-
trations for longer average times will be lower than the 1-hour values,
these results indicate that both the federal and state ambient air qual-
ity standards will be met (see Tables 48 and 49).
These input data can also be used to determine whether plume rise retar-
dation or downwash is likely to be significant at La Palma. According
to Equation (7), the greatest stack height reduction due to plume rise
retardation will be 6.8 m. Therefore, complete downwash is not possible
according to the restriction given by Equation (9). Partial downwash is
possible, however, according to Equation (10). Evaluating Equations (7)
191
-------
o
i — -i
a:
cc
u_
o :
10"
O HIND SFcEO=i.5 M/SEC
A^WIND SPEED=2.0 M/SEC
.+ WIND 5FEED=2.5 M/5EC
X WIND SPEED = 3.'0 M/SEC
1 6 7 8~~ 9~~ \ Q°
DOWNWIND DISTANCE CKM)
5 6 7 8 9
Figure 84. S09 concentrations versus downwind distance for stability class 1
-------
U)
j
O XI:-.:0 :.fctO=c'. 0 M/SEC
A WIND 3PEEQ=3.0 M/SEC
4- WIND 3PEED = 4..0 M/SEC
x HIND 3PEEO = 5-. 0 M/SEC
10'
3
9
- 6 V
QQVfNWINtr DISTflNCE
-| FT i f
6 .7 89
Figure 85. S0« concentration versus downwind distance for stability class 2
-------
(Tj. WfNu iPEEO-cJ.O M/3EC
A WIND 3PEED=5.0 M/5EC
+ WIND SPEEDS.0 M/SEC
x WIND SPEEO=11.0 M/SEC
0 WIND SFEED=14.0 M/SEC
VO
4--
cc
cc
LUu,
o
~s e T T
- -3
"e~
8
'1C1
DOWNWINCT DISTflNCE (KM)
Figure 86. S09 concentration versus downwind distance for stability class 3
-------
o
(is HI HO SPEEQ-2. G H/StC
A HIND SPEEO=6.0 M/ScC
+ HIND 3PEED.-10.0 M/SEC
x HIND SPEED-- 14.0 M/SEC
HIND SPEED=18.0 M/SEC
v
DOHNiHIND DISTflNCE (KM)
Figure 87. S09 concentration versus downwind distance for stability class 4
-------
Q WIND SPEEO---2. 0 M/SEC
A WIND SPEED=3.0 M/SEC
+ WIND SPEEDS. 0 M/SEC
x. WIND SPEED=5.0 M/SEC
o
cr
a:
UJu.
^'^
o
o
10U
,
3 6 7 « 9 'iQ1
DOWNWIND DISTfiNCE (KM)
—r-
7
Tltf
Figure 88.
concentration versus downwind distance for stability class 5
-------
o
O HIND SFEED=2.0 h/5EC
A HIND 5FEED=3:0 M/SEC
+ WIND 5PEED=4.0 M/SEC
x WIND SPEEOS.O M/SEC
cr
a:
o
•LOT
6^ 7 8 5 \(f
DOWNWIND DISTflNCE (KMi
Figure 89. SO concentration versus downwind distance for stability class 6
-------
Table 48. NATIONAL AMBIENT AIR QUALITY STANDARDS
Pollutant
Suspended
particulates
Sulfur oxides
measured as S0?
Carbon monoxide,
CO
Photochemical
oxidants
Hydrocarbons
Nitrogen dioxide,
NO.
2
Measurement classification
Annual geometric mean
maximum 24-hour average, 1/yr
Annual arithmetic mean
maximum 24-hour average, 1/yr
maximum 3-hour average, 1/yr
Maximum 8-hour average, 1/yr
maximum 1-hour average, 1/yr
Maximum 1-hour average, 1/yr
Maximum 3-hour average, 6-9 am, 1/yr
Annual arithmetic mean
Primary standards
/ 3 a
yg/m
75
260
80
365
—
_
-
160
160
100
b
ppm
—
—
0.03
0.14
—
9
35
0.08
0.24
0.05
3 c
mg/m
-
—
_
-
—
10
40
—
-
—
Secondary standards
/ 3 a
yg/m
60
150
_
260
1300
-
—
160
160
100
b
ppm
-
—
0.02
0.10
0.50
9
35
0.08
0.24
0.05
mg/m c
-
—
—
-
—
10
40
-
-
_
VO
oo
Micrograms per cubic meter.
Parts per million (T = 25°C, P = 760 mmHg).
"Milligrams per cubic meter.
-------
and (10) using the model input data, the following expression for the
source height which incorporates both plume rise retardation and down-
wash results:
h" = 24.97 + 59'74 for u > 4.37 m/sec
u —
h" = 38.65 for u < 4.37 m/sec.
Table 49. TEXAS AMBIENT PARTICULATE STANDARDS
Cone entr at ion
averaging
time
5 hours
3 hours
1 hour
Maximum
concentrations ,
yg/m3
100
200
400
After incorporating the above plume rise retardation effects, and repeat-
ing the calculation of plume centerline concentrations for stabilities
1 and 2, it was found that the maximum 1-hour concentrations are not sig-
nificantly increased (see Figures 90 and 91).
9
Texas emission standards for particulates are given in terms of the
effluent flow rate according to Table 50. The flow rate at the La Palma
facility is approximately 57,000 acfm with an associated particulate emis-
sion rate of 15 Ib/hr, a configuration which falls well within the limits
set forth in Table 50. Since the unit at La Palma will only have a heat
input of 210 million Btu per hour, it does not fall under the following
regulation for oil- or gas-fired steam generators:
105.32 No person may cause, suffer, allow or permit
emissions of particulate matter from any oil or gas
fuel fired steam generator with a heat input greater
than 2500 million Btu per hour to exceed 0.1 Ib. per
million Btu heat input maximum 2-hour average.
199
-------
fi-j WIND DfEED=1.5 M/SEC,H=38.65 M
£. WIND SPEED=2.0 M/SEC,H=38.65 M
+ WIWD SPF:ED = 2.5 M/SEC, H = 38. 65 M
x XINO SPEEO-3.0 M/SEC,H=38.65 M
o
to
o
o
<£
cc
UJa
10"
/
/ /
v /
/
' /
/
i
~i r
5 6 7 8 9 10 2
DOWNUIND DISTRNCE (KM)
~r~
T
-r~
T
-i r~
8 9
\tf
Figure 90. SC>2 concentration versus downwind distance for stability class 1
(plume rise retardation included)
-------
o
S o
cc
cc
O
o
o WIND
A WIND
+ WIND
x WIND
SPEED=2.0
SPEED=3.0
SPEED=y.O
SPEED=5.0
M/5EC,H=38.65 M
M/SEC,H=38.65 M
M/SEC,H=38.65 M
M/SEC>H = 36.9 M
•T-
8
—i—
9
10"
—t—
6
~T T~
8 9
DOWNWIND DISTflNCE (KM)
Figure 91. SC>2 concentration versus downwind distance for stability class 2
(plume rise retardation included)
-------
Table 50. ALLOWABLE PARTICULATE EMISSION RATES
FOR SPECIFIC FLOW RATES
Effluent flow rate,
acfm
1,000
2,000
4,000
6,000
8,000
10,000
20,000
40,000
60,000
80,000
100,000
200,000
400,000
600,000
800,000
1,000,000
Rate of emission,
Ib/hr
3.5
5.3
8.2
10.6
12.6
14.5
22.3
34..2
44.0
52.6
60.4
92.9
143.0
184.0
219.4
252.0
.For a coal-fired power plant the following regulation applies for par-
ticulate emissions:
105.31 No person may cause, suffer, allow, or permit
emissions of particulate matter from any solid fossil
fuel fired steam generator to exceed 0.3 Ibs. per mil-
lion Btu heat input maximum 2-hour average.
For the La Palma plant, this translates into an allowable emission rate
of 63 Ib/hr which is well in excess of the projected value of 15 Ib/hr.
The Texas SO- emission regulation for liquid fuel-fired steam generators
states that SO flue gas concentrations may not exceed 440 ppm. This
202
-------
condition will be met for La Palma since the S0? flue gas concentration
will be about 290 ppm. The regulation for coal-fired plants states that
S0« emissions shall not exceed 3.0 pounds per million Btu heat input.
For the La Palma plant this translates into an allowable emission rate
for S02 of 627 Ib/hr which is well above the projected rate of 135 Ib/hr.
203
-------
REFERENCES
1. Climatic Atlas of the United States. U.S. Department of Commerce.
Environmental Science Service Administration. Environmental Data
Service, June 1968.
2. Holzworth, G.C. Mixing Heights, Wind Speeds, and Potential for
Urban Air Pollution Throughout the Contiguous United States.
Office of Air Programs Publication No. AP-101. Environmental
Protection Agency. Office of Air Programs. Research Triangle
Park, North Carolina, January 1972.
3. Gifford, F.A., Jr. An Outline of Theories of Diffusion in the Lower
Layers of the Atmosphere. Chapter 3, Meteorology and Atomic Energy
1968 (D. Slade, ed.). United States Atomic Energy Commission.
Report No. USAEC-TID-24190, 1968.
4. Pasquill, F. Atmospheric Diffusion. London, D. Van Nostrand
Company, Ltd, 1962.
5. Briggs, G.A. Plume Rise. AEC Critical Review Series. United
States Atomic Energy Commission. Report No. TID-25075, 1969.
6. Turner, D.B. Workbook of Atmospheric Dispersion Estimates. U.S.
Department of Health, Education and Welfare, Consumer Protection and
Environmental Health Service, National Air Pollution Control Ad-
ministration, Cincinnati, Ohio. Public Health Service Publication
Number 999-AP-26, Revised 1969.
7. Busse, A.D. and J.R. Zimmerman. User's Guide for the Climatological
Dispersion Model. U.S. Environmental Protection Agency, Raleigh,
North Carolina. Publication No. EPA R-4-73-024, December 1973.
8. Briggs, G.A. Diffusion Estimation for Small Emissions. U.S.
Department of Commerce, No. AA-ERL-ARATDL. Contribution No. 79.
Oak Ridge, Tennessee, May 1973.
9. Environment Reporter. Texas Clean Air Act.
204
-------
APPENDIX A
PROCESS DESCRIPTION AND EMISSIONS ESTIMATES
FOR THE COAL-FIRED CAFB
This section discusses the differences in process operation and emissions
associated with the coal-fired CAFB alternative advanced by Foster-Wheeler.
Both the 10 MW demonstration plant and 250 MW unit are considered. Operat-
ing conditions and emissions will be similar to oil-firing but additional
unit operations such as coal crushing and drying and additional problems
of ash handling and increased particulate emissions must be considered.
PROCESS DESCRIPTION: 10 MW DEMONSTRATION PLANT
The plant is designed to operate in the same manner as described for oil-
firing in Section III. The process flow diagram given in Figure 3
applies to coal-firing as well. Table A-l is a listing of the mass flow
rates associated with coal-firing of the 10 MW demonstration plant.
Coal will be removed from the coal pile and transported to two storage
bunkers by a vibrating conveyor. Coal will be withdrawn from the storage
bunkers and sent to a crusher to produce a size gradation of 100 percent
< 1/2 inch, 88 percent < 1/4 inch, and 18 percent < 30 mesh. Crushed coal
will be transferred into a 6-hour intermediate storage silo and withdrawn
in two separate streams by gravimetric feeders. The coal will be trans-
ported to vibrating tables which are pressurized with flue gas recirculated
from the boiler. The solid fuel will then feed into the gasifier through
24 3-inch diameter coal needles, with 12 needles on each side of the
gasifier chamber.
205
-------
Table A-l. MASS FLOW RATES FOR FW 10 MW COAL-FIRED
CAFB DEMONSTRATION PLANT
Process stream
1. Limestone to gasifier
2. Product gas from gasifier
3. Gasifier to regenerator stone transfer
4. Regenerator to gasifier stone transfer
5. Flue gas to pulsed solid transfer lines
6. Regenerator off-gas (total)
SO 2
CO 2
N2
7. Water or steam injection
8. Regenerator off -gas after cyclone and
cooling
9. Coal to RESOX™ reactor
10. Hot solids from RESOX™ reactor
11. Waste solids from RESOX™ quench
vessel
12. Hot air to RESOX™ reactor
13. Influent gas to RESOX™ reactor
14. Elemental sulfur from RESOX™
15. Return steam
16. Water to sulfur condenser
17. RESOX™ tail gas
18. Condensed liquid sulfur
19. Fugitive dust from coal handling
system
20. Air to start up heater
21. Air to start up heater
22. Air to RESOX™ reactor
23. Cooling water for RESOX solid waste
24. Steam from quench vessel
25. Regenerator spent solids
26. Regenerator off -gas cycloned solids
Mass flow rate,
kg/s - (Ib/hr)
0.29 (2,300)
8.05 (63,800)
7.94 (63,000)
7.54 (59,800)
0.99 (7,820)
0.09 (730)
0.03 (260)
0.03 (260)
1.00 (7,900)
0.05 (390)
0.11 (880)
Temperature
°C (°F)
206
-------
Table A-l (continued). MASS FLOW RATES FOR FW 10 MW COAL-FIRED
CAFB DEMONSTRATION PLANT
Process stream
27. Air to spent solids cooler
28. Cooled solids
29. Cooler exhaust to cyclone
30. Cooled solids to storage
31. Air emissions from spent solids
cooler
32. Cycloned solids to storage
33. Solids to storage
34. Solid waste from storage silo
35. Air emissions from solids storage
silo
36o Air to gasifier and regenerator
37. Flue gas recycled from stack
38. Boiler stack emissions
39. Flue gas to coal distributing
conveyor
40. Influent gas to gasifier (total)
Air
Flue gas
Tail gas
41. Air to regenerator
42. Coal to distributing conveyor
43. Coal to gasifier
44. Oil to gasifier
45. Fugitive limestone handling
emissions
Mass flow rate,
kg/s (Ib/hr)
5.75 (45,600)
4.07 (32,300)
0.68 (5,400)
1.00 (7,900)
0.69 (5,500)
2.41 (19,100)
2.41 (19,100)
Temperature
°C (°F)
207
-------
TM
The remainder of the system including regenerator, RESOX , solids handling,
and limestone feed is identical to that described for oil-firing in Sec-
tion III.
EMISSIONS ESTIMATES: 10 MW DEMONSTRATION PLANT
Differences between emissions from coal-firing and oil-firing include air
and water emissions from coal handling, an increase in solid waste from ash
production, and a potential increase in particulate emissions from the
stack.
Emissions from Coal Handling
Coal handling air emissions will emanate from the coal storage pile, coal
conveyors and feeders, coal crushers, and coal dryers. Coal drying is not
intended for the demonstration plant but is included in the 250 MW proposal
and will be discussed in that subsection.
Air emissions from coal storage depend upon wind speed, coal pile surface
area, degree of containment, coal density, and the prevailing precipitation-
2
evaporation index. The Midwest Research Institute has estimated a partic-
ulate emission factor of 0.018 g/kg (0.0036 Ib/ton) which includes losses
from coal storage, handling and feeding. Based on an average coal usage
rate of 2.4 kg/s (19,000 Ib/hr), the particulate emission from coal storage,
-3
handling and feeding will be equal to 4.3 x 10 g/s (0.034 Ib/hr).
Air emissions from coal crushing vary depending on whether the operation is
wet or dry and on the type of containment and control practices. There is
very limited data available for the prediction of particulate emissions.
An uncontrolled emission factor of 0.25 g/kg (0.5 Ib/ton) is adapted from
3
estimates for crushing of rock. Coal has different fracturing charac-
teristics than rock but this is the only reasonable estimate of emissions
available. Application of this factor results in a particulate emission
rate of 0.6 g/s (4.8 Ib/hr) from the crushing of coal.
208
-------
Table A-2. MASS FLOW RATES FOR FW 250 MW COAL-FIRED
CAFB DESIGN
Process stream3
Mass flow rate,
kg/s (Ib/hr)
Temperature
oc 0F)
1. Coal from storage to dryer
2. Exhaust from dryer to cyclone
3. Exhaust from cyclone to scrubber
4. Air emissions from scrubber
5. Solids collected by cyclone
6. Coal from dryer to crusher
7. Fugitive dust emissions from crusher
8. Coal from crusher
9. Coal to coal:limestone blenders
10. Limestone from storage to dryer
11. Off-gas from limestone dryer to
baghouse
12. Air emissions from baghouse
13. Solids collected by baghouse
14. Limestone to crusher
15. Fugitive dust emission from lime-
stone crusher
16. Limestone from crusher
17. Limestone to gasifier modules
22. Limestone and coal from blenders
23. Limestone and coal from vibrating
feeders to gasifier modules
24. Product gas to quad cyclone
25. Product gas to boiler
26. Solids returned from quad cyclone
27. Gasifier to regenerator stone
transfer
28. Regenerator to gasifier stone
transfer
29. Regenerator off-gas to twin-
cyclones
28.73 (227,800)
0.29 (2,280)
0.09
0.01
0.20
(680)
(70)
(1,600)
28.73 (227,800)
3.91 (31,000)
104.4 (827,500)
104.4 (827,500)
110.5 (876,000)
107.4 (851,500)
12.80 (101,500)
-------
Table A-2 (continued). MASS FLOW RATES FOR FW 250 MW COAL-FIRED
CAFB DESIGN
Process stream
a
Mass flow rate,
kg/s (Ib/hr)
Temperature
°C (°F)
56. Exhaust from solids cooler to cyclone
57. Cycloned solids cooler exhaust to
stack
58. Cycloned solids cooler exhaust to
coal and limestone dryers
59. Cycloned solids to storage
60. Solids to storage
61. Exhaust from storage to vent filters
62. Air emissions from vent filters
63. Solids from vent filters to storage
64. Solid waste from storage
65. Flue gas from boiler to stack
66. Air emissions from stack
67. Flue gas recycled to gasifier
70. Air to gasifier
71. Air to regenerator
72. Liquid waste from coal dryer scrubber
8.92 (70,700)
52.93 (419,700)
11.04 (87,550)
0.63 (5,000)
Process streams 18 to 21, 68, and 69 are applicable to oil firing and are
presented in Section III.
213
-------
EMISSIONS ESTIMATES: 250 MW CAFB
Emissions from Coal Handling
Air emissions associated with coal handling include storage pile emissions,
conveying and feeding emissions, and crushing and drying emissions.
The particulate emission factor for coal piles is taken as 0.0018 g/kg
(0.0036 Ib/ton) as noted earlier in conjunction with the 10 MW demonstra-
tion plant. Two piles will be maintained to provide 33 days of storage.
At a coal feed rate of 28.7 kg/s (2730 tons/day), the yearly particulate
emission from coal storage, handling and feeding will be 0.052 g/s
. (3600 Ib/hr).
The particulate emission rate for coal crushing is estimated to be 0.25 g/kg
(0.5 Ibs/ton) of coal processed. At a coal feed rate of 28.7 kg/s (2730 tons
day), the particulate emission from coal crushing at the 250 MW plant will
be 7.2 g/s (250 tons/yr).
Air emissions from coal drying are higher than those produced at other
points in the coal processing cycle. Table A-3 presents estimates of un-
controlled emissions from three types of coal dryers.
Table A-3. EMISSION FACTORS FOR
COAL DRYING
Type of dryer
Fluidized bed
Flash
Multilowered
Uncontrolled
emissions,3 Ib/ton
20
16
25
The following collection efficiencies are
applicable for control with cyclones:
Cyclone collection efficiency 70%
Multicyclones 85%
Cyclone and wet scrubber 99-99.9%
214
-------
Exhaust from the coal dryer proposed for the 250 MW installation will pass
through a cyclone and scrubber. Therefore, particulate emissions from the
drying units can be expected to be on the order of 0.1 g/kg (0.2 Ib/ton)
of coal processed. This amounts to 2.9 g/s (100 tons/yr) based on a
coal feed rate of 28.7 kg/s (2730 tons/day).
A 33-day coal storage requirement is anticipated for the 250 MW plant. Based
on an assumed storage pile height of 4.5 m (15 ft) and a yearly precipitation
of 0.5 m (20 in), the total runoff from coal storage is approximately
3
5700 m /yr (1,500,000 gallons/yr). The dissolved solid concentrations given
in Table 10 will be applicable for the 250 MW proposal.
Other Emissions
TM
Particulate air emissions resulting from RESOX coal storage will be
_2
approximately 2.19 mg/s (1.7 x 10 Ib/hr), based on an emission factor
of 1.77 ng/kg-yr (0.0036 Ib/ton-yr). Emissions from limestone handling
will amount to 4.65 g/s (36.9 Ib/hr), based on a total particulate emission
rate of 1.19 g/kg (2.38 Ib/ton). The comments made in the previous section
regarding stack particulate emissions from the demonstration plant pertains
to the 250 MW unit as well.
TM
Water emissions from the RESOX coal storage pile are expected to be the
same as calculated for oil-firing of the 250 MW commercial unit and
amounts to 212 m3 (7500 ft3).
Solid Waste
Coal-fuel operation of the CAFB will result in a larger solid waste output
than will oil-fuel operation. This additional material will be produced
as bottom ash mixed with spent regenerator stone and as effluents from
scrubbers and cyclones used in the coal feed system. The discussion pre-
sented below of the environmental impact associated with disposal of this
solid waste is based upon the substantial amount of work which has been
done on the environmental effects of disposal of solid waste from fluidized
bed combustion of coal and from flue gas desulfurization.
215
-------
Spent solids extracted from the regenerator will have high sulfide and
sulfate content. Three predisposal treatment methods are presently being
developed, including oxidation and sintering, mixing of stone with coal
ash and hot pressing, and wet slurrying. These processes are discussed
in some detail in Section IV.
After preliminary treatment of spent stone or ash, three options exist
for subsequent handling. The material can be used as landfill, discharged
to a holding pond, or recycled. In the first two instances, it is important
to assess potential air and water pollutant emissions.
Air emissions will be a problem mainly with landfilling and not with solids
discharge to water. In the case of coal ash, it is estimated that wind
erosion particulate losses from ash disposal sites will amount to 1 Ib/ton
of ash discarded. Typical chemical compositions of coal ash are given in
4
Table A-4. Trace elements which may be combined with the ash are listed
in Table A-5.
Air emissions from land disposal of spent sorbent stone will consist of
particulate matter and gaseous sulfur compounds. The two hazardous com-
pounds of concern are calcium sulfide and calcium sulfate. Sulfide reacts
with moisture in the air to form H»S which is subsequently oxidized to S00.
The water environment can be adversely affected regardless of whether spent
stone and ash is landfilled or discharged to some type of settling basin.
Pollutants are discharged to groundwater and surface waters from landfills
by leaching. BCURA and Pope, Evans and Robbins have investigated the
properties of the leachate obtained from a fly ash-stone effluent. BCURA
found that, although CaO, MgO and CO. contents of the leachate varied, all
their samples showed common features:
• High pH (10.5 to 11.6)
• High or complete extraction of sulfate
• Negligible extraction of magnesium
216
-------
Table A-4. POWER PLANT COAL ASH COMPOSITIONS
Constituent
Silica (Si02)
Alumina (Al-O-)
Ferric Oxide (Fe_0~)
Lime (CaO)
Potassium Oxide (K_0)
Magnesia (MgO)
Sodium Oxide (Na20)
Titanium Dioxide (TiO )
Sulfur Trioxide (SO.)
Carbon (C) and volatiles
Boron (B)
Phosphorus (P)
Uranium (U) and Thorium (Th)
% by weight
30-50
20-30
10-30
1.5-4.7
1.0-3.0
0.5-1.1
0.4-1.5
0.4-1.3
0.2-3.2
1.0-4.0
0.1-0.6
0.01-0.3
0.0-0.1
Table A-5. SELECTED TRACE ELEMENTS IN ASH (ppm)
Element
Arsenic
Mercury
Antimony
Selenium
Cadmium
Zinc
Manganese
Boron
Bar ium
Beryllium
Nickel
Chromium
Lead
Vanadium
Fly ash
15
0.03
2.1
18
<0.5
70
150
300
5000
3
70
150
30
150
3,
Bottom ash
3
<0.01
0.26
1
<0.5
25
150
70
1500
<2
15
70
20
70
Actual trace element composition will vary
widely depending on boiler type and coal
composition.
217
-------
Theis in a study on the potential trace metal contamination of water
through fly ash disposal has made the following assertions. At the normal
pH range of natural waters, the hydroxide of some metals (Hg, Pb, Cu, Cr,
Cd, Zn) controls their solubility. At elevated pH, carbonate may control
solubility. In general, trace metals display drastically decreased
solubilities with increasing pH. In the pH range 7 to 8.5, only Zn and
Cd could be considered soluble. Arsenic is generally very soluble.
Theis presented the relationship between solubility and pH. Manlock
has also studied leachate solubility-pH relationships.
The elements of major concern are therefore limited to As, Se, V, and
Cd. However, since complexes may form which would increase the
solubility of the metals, Pb and Hg, at least, may also be of concern in
leachates. Theis found for example that addition of EDTA of his ash
4
samples increased the solubility of all elements but mercury.
Rossof and Rossi have investigated possibly toxic elements in scrubber
sludges. Although the composition of a scrubber sludge is different than
that of the spent stone from a regenerator or overhead from the combustor,
in general, the same elements are present and should be affected by pH
and ionic species present in a similar manner.
The studies done by BCURA and PER with partially sulfated lime (bed mate-
rial) have shown that the leachate is highly alkaline. Since metal solu-
bility increases with decreasing pH, however, leaching occurring in an
acidic environment will result in higher trace element concentrations.
Thus, it is unlikely that the spent stone or overhead will produce an
acidic leachate. Increased solubility may, however, occur by means of
complex formation. No data were available on complex formation in the
leachate from either spent stone or bed material.
Pollutant emissions to the ambient water environment also occur when ash
or spent stone is discharged to a settling basin. The overflow from the
218
-------
basin contains a finite percentage of suspended and dissolved solids
influent to the basin. It has been estimated that the total suspended
solids concentration of ash pond overflow from electric utility power
o
plants averages approximately 100 ng/1.
When coal ash is added to water an immediate reduction in pH and dissolved
9 10
oxygen occurs. A study by Rohrman indicates that nitrogen and phos-
phorus are detectable in ash holding basins in dissolved form at a level
of 0.1 to 1.0 ng/1. Approximately five times as much phosphorus may be
present in suspended form. These nutrients will enhance plant and bacterial
growth in the settling pond and may have an effect on ambient water after
overflow.
The addition of spent stone to water results in contamination with calcium
and magnesium oxide, calcium sulfide, sulfate and carbonate, and magnesium
sulfite. This may result in the production of heat, formation of H_S,
and flotation of agglomerated nonsettleable solids.
219
-------
REFERENCES
1. Chemically Active Fluid Bed Process (CAFE) Preliminary Process Design
Manual. Foster Wheeler Energy Corp., Livingston, N.J. U.S. Environ-
mental Protection Agency, Research Triangle Park, N.C., Contract
Number 68-02-2106. December 1975. 185 p.
2. Development of Emission Factors for Fugitive Dust Sources. MRI.
U.S. EPA Report No. 450/3-74-037. 1974.
3. Compilation of Air Pollutant Emission Factors. EPA Publication AP-42.
U.S. Environmental Protection Agency. April 1973.
4. Solid Waste Disposal. Radian Corporation. U.S. EPA Report No. 650/
2-74-030. May 1974.
5. Pressurized Fluidized Bed Combustion. National Research Development
Corporation, London. Report to Office of Coal Research. R & D Report
Number 85.
6. Pope, Evans and Robbins, Inc. Multicell Fluidized Bed Boiler Design,
Construction and Test Program. Interim Report No. 1. Publication
Number PER-570-74 for Office of Coal Research. August 1974.
7. Theis, T.L. The Potential Trace Metal Contamination of Water Resources
Through the Disposal of Fly Ash. Presented at 2nd National Conference
on Complete Water Reuse.
8. Development Document for Effluent Limitations Guidelines and New Source
Performance Standards for the Steam Electric Power Generating Point
Source Category. U.S. EPA Report No. 440/l-74-029a. October 1974.
9. Surprenant, N., et al. Preliminary Emissions Assessment of Conventional
Stationary Combustion Systems. Volume II - Final Report. U.S. EPA
Report No. 600/2-76-046b. March 1976.
10. Rohrman, F.A. Analyzing the Effect of Fly Ash on Water Pollution.
Power. 76-77, August 1971.
220
-------
APPENDIX B
COMPARISON OF THE CAFB WITH OTHER RESIDUAL
OIL UTILIZATION TECHNIQUES
INTRODUCTION
Three alternatives are available for the combustion of high sulfur residual
oil in an environmentally acceptable manner:
• In-situ desulfurization
• Precombustion desulfurization of feedstock
• Flue gas desulfurization (FGD).
The CAFB process is the only potentially viable in-situ technique identified
by GCA. A number of flue gas desulfurization schemes presently exist which
are applicable to both coal-fired and oil-fired boilers. Residual oil
desulfurization techniques produce a variety of solid, liquid and gaseous
fuels. Of the many desulfurization options, only three are competitive
with the CAFB in their ability to handle high metal as well as high sulfur
content feedstocks.
The alternative processes are examined in some depth in this and the
following two appendices. This appendix is a general overview of
desulfurization technology and describes associated unit operations
required for sulfur recovery. Flue gas desulfurization (FGD) is pre-
sented and existing systems are identified. The final subsections of
this appendix compares the environmental impacts of hydrodesulfurization
and FGD. Appendix C provides process descriptions and flow charts of
potentially viable and currently used individual residual oil
221
-------
desulfurization techniques. Appendix D contains a summary and comparison
of the economics associated with all desulfurization options.
RESIDUAL DESULFURIZATION
Present technology has allowed the refiner to produce low sulfur solid,
liquid and gaseous fuels from high sulfur feedstocks. For example:
coking produces solid, liquid and gaseous fuels; hydrodesulfurization
(HDS) produces a liquid fuel; and a procedure called partial oxidation
produces a low Btu gas. Although the specific reaction steps vary, most
feedstock desulfurization techniques are based on the reaction of hydrogen
with oil in the presence of a catalyst, as is shown in the general
reaction,
[fuel sulfur] + H2 CatalyStr H2S + [clean fuel!
H9S is evolved in the desulfurization process. Consequently, residual
desulfurization is a two step process: (1) desulfurization of the resid-
ual oil with the resulting formation of KLS; and (2) the disposal or
recovery of the H^S process stream in an environmentally acceptable manner.
The metals content of the feed is the most significant variable influencing
processing cost and desulfurization efficiency. High molecular-weight
organometallic compounds of vanadium and nickel are found in most crude-oil
residues. Under the reaction conditions necessary for desulfurization,
some of these complex molecules decompose resulting in the deposition of
vanadium and nickel on the surface of the desulfurization catalyst. Over
months of continuous operation, metal accumulation causes a reduction in
catalytic desulfurization activity. High metals feedstock (metals in ex-
cess of 150 to 200 ppm) will, in most cases, require some form of feed
demetalization.
222
-------
In general, residual oils of low to moderate metals content (Ni plus V
content less than 100 ppm) and sulfur content as high as 6 percent can be
directly desulfurized to yield heavy oil products containing as little as
0.5 wt percent sulfur. For higher metal content feeds and lower sulfur
product fuel oil levels, modified techniques, such as Flexicoking or
demetalization/desulfurization, will be required.
In addition to the CAFB only three systems (L.C. Fining, demetalization/
desulfurization and Flexicoking in conjunction with a HDS unit) have been
designed to effectively handle high metal content feedstocks. L.C. Fining
and demetalization/desulfurization are both hydrodesulfurization techniques.
The Flexicoking process is an extension of the fluid coking process. Pro-
cess descriptions and flow diagrams of all three processes are found in
Appendix C. Other systems are capable of desulfurizing high metal feeds,
but at a higher operating cost. Process descriptions of these systems
are also presented in Appendix C. The feed to the CAFB will consist of
high sulfur and high metal resid, thus this section will compare the CAFB
only with those systems capable of handling a similar feed. Economic and
process data for all 16 desulfurization techniques considered are presented
in Appendix C.
H..S Removal
Because feedstock desulfurization generates tLS as a process stream, it
is necessary to dispose of or convert this gas to a useful product. The
most commonly practiced method is the conversion of H_S to elemental sulfur
3
by means of a Glaus Plant. A flow diagram of a typical two stage Glaus
sulfur plant is shown in Figure B-l.
223
-------
THERMAL STAGE CATALYTIC
;S + jo2-»sc STAGES
INCINERATOR STACK
O +• A
/ \ "*>•* T T
\. S |__ 1 * ^
^"*— ^ " ** f ] i • • '"• 1
AIR • 1 1
H20 ' H20 '
1
S
APPROXIMATE SULFUR 6(
^A
I
/f
i
H^
7
)%
so2 + z
i
s
?,
-J
2
H2S
.
•
>
1
5%
^3
1
0
X
f
L«.
cz
r-
J
54 2
f
...— J
1
— — ,
'I
7
H-0 -f- A lO.OOOppmTO
z 30,000 ppm
4
PMfl..
AIR ^ ^
%
YIELD
Figure B-l. Typical two-stage Glaus sulfur plant
The most common method of concentrating and collecting H_S involves
washing the product gas with a water solution containing an amine. The
rich solution is then steam stripped, driving off the H_S, and regenerating
the absorbing solution. A typical composition of the gas taken from an
amine regenerator is:
H2S
co.
2
HC
H_0 vapor
80
2
0.5
5
- 93%
- 10%
- 2%
- 10%
The H-S present in the gas may then be converted to elemental sulfur by
the following reaction scheme:
H2S
S0_ + H_ (thermal combustion)
3 S + 2 H 0 (thermal and catalytic)
(overall) 3 H0S + On
+ 3 S +
224
-------
The overall efficiency of a Glaus plant is 90 to 97 percent." Maximum
sulfur conversion in a Glaus plant is limited because:
• The Glaus reaction is reversible and is limited by
chemical equilibrium;
• A very significant portion (25 percent) of the sulfur
passes through the system in relatively unreduced
form - carbonyl sulfide and carbon disulfide.
Glaus Tail Gas Cleanup — The tail gas from a typical sulfur plant contains
about one-third water vapor, 5 to 15 percent CCL, 2 to 4 percent sulfur com-
pounds (H?S, S0_, COS and CS~), and the balance nitrogen. The SO- concen-
tration in the tail gas is .10,000 to 30,000 ppm. In order to produce a
stack gas with less than 250 ppm SO content, the overall sulfur plant
must be at least 99.9 percent efficient. This efficiency is not possible
with present technology unless a tail gas cleanup plant is also used.
Several systems are available for tail gas cleanup: the Beavon Sulfur
Removal Process, the Cleanair Sulfur Process, and the IFP process. The
investment and operating costs for the Beavon and Cleanair Process are
approximately equal to the original cost of the sulfur removal plant. The
IFP process is approximately one half the cost of the original sulfur
plant but is not as efficient as the first two (99.0 percent versus 99.9
percent).
Beavon Sulfur Removal Process - The Beavon Sulfur Removal Process, de-
veloped by Ralph M. Parsons Company and Union Oil Company of California,
is capable of limiting S02 emissions to 40 to 80 ppm depending on the
efficiency of the preceding Glaus Plant. In this process the Glaus plant
tail gas is mixed with hot combustion gas produced by burning fuel gas
with air. The resulting reducing mixture is passed through a catalytic
reactor similar to that in a Glaus plant. The sulfur is hydrogenated to
H^S on a cobalt/molybdate catalyst. Water is condensed from the gas in
a heat exchanger. The cooled gas stream is passed to a Stretford section
225
-------
in which H_S is removed from the gas and converted to elemental sulfur.
The cost of this system is approximately equal to the original cost of
the Glaus plant.
Stretford Process — The Stretford Process consists of a gas washing system
wherein the gas is contacted countercurrently with an alkaline washing
solution (sodium carbonate). Hydrogen sulfide is removed from the gas
stream and is oxidized to elemental sulfur. The sulfur is formed as a
finely dispersed solid in the circulating solution. The reduced solution
is then oxidized by air blowing which simultaneously removes the sulfur
by froth flotation. The oxidized solution is returned to the gas wash
system to repeat the cycle. The sulfur slurry is fed to an autoclave
where heat is applied to dry and melt the sulfur. Liquid sulfur of
greater than 99.5 percent purity is obtained.
Cleanair Sulfur Process — The Cleanair Sulfur Process developed by
J.F. Pritchard and Co. and Texas Gulf Sulfur Co. is capable of producing
a gas effluent containing less than 250 ppm of S0?. This system is com-
posed of three process stages, two of which are proprietory and are not
fully explained in the literature:
• Stage 1 converts essentially all of the S02 to elemental
sulfur with some additional conversion of H2S to elemental
sulfur;
• Stage 2, which is the Stretford process (the same process
* used in the Beavon process) converts the remaining H_S to
elemental sulfur;
• Stage 3 is an important step in controlling COS and CS9
emissions from the Glaus tail gas. Concentrations of
these two compounds are reduced to less than 250 ppm
equivalent S02- Carbon disulfide and carbonyl sulfide
are the prime precursors of high SCL concentrations in
Glaus tail gas treating systems.
The cost of this system is similar to the Beavon process and is approx-
imately equal to the original cost of the Glaus plant.
226
-------
IFF Process - The third Claus tail gas treatment process is the Institute
Francais du Petrole (IFF) system. Claus tail gas at about 127°C (260°F)
is injected into the lower section of a packed tower, where a solvent
containing catalyst is circulated countercurrently, resulting in maximum
liquid-gas contact. Product sulfur accumulates at the bottom of the
tower and is continuously removed. Some solvent is lost by evapora-
tion through the top of the column and therefore must be replaced. Cata-
lyst is also pumped to the tower to maintain a constant concentration.
Due to this system's inability to handle COS and CS9, emissions are approx-
imately 1500 ppm S0_; however, the original investment is only one-half
of that required for the Beavon or Cleanair Sulfur Process.
Flue Gas Desulfurization
Numerous processes have been proposed for flue gas desulfurization (FGD).
However, only the six systems outlined in Table B-l have gained acceptance
in the United States. Three more systems are in the prototype stage of
development; the Foster Wheeler-Bergbon Forsching process, the Thorough-
bred 101 process and the Shell Flue Gas Desulfurization process.
Most FGD systems now in use are operating on coal-fired boilers. Only
two plants, the City of Key West, Stock Island Plant and the Boston
Edison Mystic 6 Plant (see Table B-2) use FGD on oil-fired boilers. The
Stock Island Plant has had considerable difficulty during operation neces-
sitating extensive downtime. The Mystic 6 Plant is the only full size
system (a demonstration plant) using magnesia wet-scrubbing on an oil-
fired utility generating unit.
227
-------
Table B-l. SUMMARY DESCRIPTION OF FLUE GAS DESULFURIZATION PROCESSES
Process
Lime/limestone
scrubbing
Double alkali
process
Magnesium oxide
scrubbing
Classification/
operating principles
Throwaway process/
vet absorption in
scrubber by slurry;
insoluble sulfites
and sulfates dis-
posed of as waste.
Throwaway process/
wet absorption in
scrubber; reac-
tants and reaction
products soluble;
reaction products
precipitated and
removed from re-
cycled react ant
solution outside
of scrubber; most
common reactant
sodium sulfite.
Regenerative pro-
cess/wet absorp-
tion by magnesium
oxide slurry; fly
ash removed prior
to or after scrub-
bing; magnesium
oxide regenerated
by calcining with
carbon; S02 by-
product can be
converted to sul-
fur ic acid or
sulfur.
S02 particulate
efficiency
Up to 90 percent
S02 renoval/99
percent fly ash
removal by most
scrubbers.
High efficiency
> 90 percent S02
removal/high par-
ticulate removal.
90 percent SC>2
removal/partlcu-
late removal as
required by
prescrubber.
Development status
Most studied but
reliability ques-
tionable due prim-
arily to scaling;
16 full scale
units in operation
or planned for
start-up by 1977.
Active area but no
full scale demon-
stration as yet;
G.M. installed a
unit on a coal-
fired boiler in
February 1974;
several sulfate
removal schemes
under study.
One full scaled
unit on test at
Boston Edison
150 MH oil-fired
unit.
Application
Old but pre-
ferably new
power plants;
coal- or oil-
fired.
As above with
potentially
lower cost
and greater
ease of oper-
ation favor-
ing some In-
roads into
smaller
plants .
Similar to
lime /lime-
stone but
oil-fired
boilers will
not require
particulate
control up-
stream of
scrubber.
Implementation
An additional
40 to 60 units
forecast for
installation by
1977 ; forecast
appears optimis-
tic; 4 to 5 years
lead time needed
for new plants;
3 years for re-
trofit of old
plants.
Research-Cottrell
estimates $600
million a year
market by 1979;
a second genera-
tion lime/
limestone system;
lead times as
above for power
plants.
Ho known plans for
immediate imple-
ment at ion ; 1 ead
times as for lime/
limestone systems.
Advantages
Cheapest of
existing pro-
cesses ; elim-
ination of
particulate
control re-
quirement.
Potentially
cheaper, sim-
pler and more
reliable than
lime/limestone
system.
May be more
reliable than
lime/limestone
process; no
known waste
disposal prob-
lems ; regen-
eration facil-
ity need not
be located at
utility.
Disadvantages
Technical reliability doubt-
ful; waste and water pollu-
tion problems; reheat of
scrubber exit gases needed;
supply and handling of large
volumes of reactant may be
problems.
Similar to above and all
throw away systems.
Cost of regeneration;
marketing of sulfur
products; reheat.
N>
N5
O)
-------
Table B-l (continued). SUMMARY DESCRIPTION OF FLUE GAS DESULFURIZATION PROCESSES
Process
Wellman-Lord
Citrate system
Catalytic
oxidation
«.
Classification/
operating principles
Regenerative pro-
cess/sodium base
scrubbing with
sulfite to pro-
duce bisulfite;
regeneration in
an evaporative
crystallizer;
sulfate formed
either purged
or removed by
selective crys-
tallization.
Regenerative pro-
cess/flue gas
washed to re-
move particles
and 803, cooled
and absorbed in
sodium citrate-
citric acid
solution in
packed tower;
solution then
reacted with
hydrogen sul-
fide to form
sulfur.
Regenerative
process/cata-
lytic oxida-
tion by V205 at
850 to 900°F to
convert SC^ to
S03 followed by
condensation to
form 70 to 80
percent HoSO^ .
Variation of con-
tact process ap-
plied to dilute
gases.
S(>2 particulate
efficiency
> 90 percent S02
removal particu-
late removal by
prescrubber.
> 95 percent S02
removal/part icu-
late removal as
required.
85 to 90 per-
cent S(>2 recov-
ery/high parti-
culate effici-
ency needed to
avoid plugging
and fouling of
catalyst.
Development status
Reliably operated
(> 9000 hours) in
Japan. Full scale
demonstration
scheduled at North-
ern Indiana Public
Service coal-fired
115 MW boiler;
sulfate removal
vital to success.
New development
by Bureau of Mines;
now testing 1000
cfm pilot plant;
also 2000 cfm unit
in Terre Haute,
Indiana. High
potential.
Two-year test
period on 15 MW
boiler; also test
on 100 MW boiler
of Illinois Power
Company; relia-
bility not demon-
strated.
Application
As above.
As above.
New plants,
oil or coal.
Implementation
As above .
As above .
As above .
Advantages
More reliable
than lime/
limestone sys-
tem based on
Japanese ex-
perience ; sim-
plicity of
unit opera-
tions in re-
generator;
waste dis-
posal prob-
lems reduced .
High effici-
ency ; econo-
mic ; no inter-
mediate SO 2
regeneration;
high reliabil-
ity; poten-
tially most
attractive of
viable pro-
cesses.
Relatively
simple and
known tech-
nology ; min-
imal mechan-
ical opera-
tions; no
relevant re-
heat require-
ments.
Disadvantages
Some bleed of solution
to remove undesirable
reaction products a
source of water pollu-
tion, otherwise as above.
Marketing of sulfur;
reheat .
_
Expensive; poor quality sul-
furic acid; poor reliability
with appreciable downtime;
extra ducting to avoid prob-
lems associated with ESP fail-
ures and high temperature
gases.
VO
-------
ENVIRONMENTAL IMPACTS OF DESULFURIZATION TECHNIQUES
Environmental Impacts of FGD
The environmental impacts of five flue gas desulfurization techniques are
discussed below. These techniques are:
1. Limestone slurry scrubbing;
2. Lime slurry scrubbing;
3. Magnesia slurry scrubbing;
4. Sodium solution - S02 reduction;
5. Catalytic oxidation
Emissions and effluent data are based on a 500 MW power plant. The fuel
is coal containing 3.5 percent sulfur with the FGD system assumed to have
a 90 percent efficiency. Coal is used in this comparison instead of resid-
ual oil because it is the only fuel with sufficient environmental data for
FGD. The solid waste and particulate matter generated when firing residual
oil will be less than for coal firing. Table B-3 lists the pollutants which
are incompletely converted or generated as byproducts from each system.
Table B-3.
FGD ENVIRONMENTAL IMPACT& TONS/YR
Limestone slurry
scrubbing
Lime slurry
scrubbing
Magnesia slurry
Sodium solution
scrubbing
SO- reduction
Catalytic oxidation
Particulate
1,280
4,276
1,968
3,077
96
so2
1,921
1,847
2,884
4,225
144
NO
X
448
431
673
1,089
34
Solid waste
156,444
156,442
386
35,002
(32,700-sulfur)b
55
Water
soluble
-
-
110,400b
(H2S04)
1,300
(Na2S04)b
109,900
CV
-------
Limestone Slurry Process — A considerable quantity of CaSCL/CaSO, solid
waste is generated approaching as much as 4 times the weight of the sulfur
removed. Wastes discharged to settling ponds are reported to have poor
settling properties and may lead to difficulty when reclaimnng the land for
future use. Potential runoff from the ponding site could lead to addi-
tional water pollution problems.
Lime Slurry Process — Characteristics and problems associated with the
lime slurry process are similar in nature to the limestone slurry process.
The only difference is that an additional 3000 tons of particulates are
produced from the production of lime, which may, however, be generated
offsite.
Magnesia Slurry Process — This process is also similar to the two preced-
ing FGD methods with the exception that the by-products (MgSCL/MgSO,) are
regenerated, thus eliminating the large quantities of solid waste. The
regeneration step requires additional process water and fuel thus producing
additional emissions.
Sodium Solution Process — Although this process is considered to be a
regenerative process, a great amount of Na2SO, by-product is produced.
This process requires a large amount of steam and water resulting in the
largest quantity of airborne pollutants among the five processes.
Catalytic Oxidation Process — This process is the cleanest and least
energy intensive of all five processes with no by-products generated
other than marketable sulfuric acid.
The only recent data for the environmental impacts of a residual oil-fired
boiler using a FGD system is the Boston Edison Mystic 6 Station. This
demonstration plant is for a 150 MW magnesia-wet scrubbing system. The
design of the facility is based on firing 2.5 percent sulfur fuel. Spent
material is sent to an off-site MgO regeneration plant capable of producing
-------
50 tons per day of sulfuric acid. The system is able to recover 91.7 per-
cent of the inlet SO- and can control particulate emissions by 57 percent.
Sources of emissions from this demonstration plant include:
— MgO losses (total average loss of 0.37 tons/day over
13 day test program)
• Stack
• Centrifuge washing
• Centrifuge case leaks
• Pump packing gland leaks
• Absorber overflow
• MgO slurry tank blow-down
• MgO slurry tank overflow
• Centrate tank overflow
• Solids loss at dryer feed end
• Dust losses at dryer I.D. fan
• Dust loss at expansion joints
• Spillage at MgO feeder
• Spillage at MgSO belt galley
• Spillage at truck loading point
— Waste water
• Process water
• Cooling water
— Solids buildup in regenerated MgO
• Vanadium
• Nickel
• Ash
Environmental Impacts of Residual HDS
Possible environmental problem areas from HDS are:
• Catalyst disposal (including vanadium and nickel deposits)
233
-------
• Vanadium, nickel and other trace constituents
in desulfurized fuel
e Various waste water streams
» Glaus and tail gas cleanup emissions
o NH» from amine scrubber
o Catalyst disposal from Glaus and tail gas cleanup process
o COS emissions
o CS_ emissions
234
-------
REFERENCES
1. Koehler, G. and J.A. Burns. The Magnesia Scrubbing Process as Applied
to an Oil-Fired Power Plant. Chemical Construction Company, New York,
N.Y. U.S. Environmental Protection Agency. Report Number EPA-600/
2-75-057. October 1975.
2. Keairns, D.L., D.H. Archer, R.A. Newby, E.P. O'Neill, and E.J. Vidt.
Evaluation of the Fluidized-Bed Combustion Process. Volume IV —
Fluidized-Bed Oil Gasification/Desulfurization. Westinghouse Research
Laboratories, Pittsburgh, Pa. U.S. Environmental Protection Agency,
Research Triangle Park, N.C. Report Number EPA-650/2-73-048d.
December 1973. 328 p.
3. Beers, W.D. Characterization of Glaus Plant Emissions. Process
Research, Inc., Cincinnati, Ohio. U.S. Environmental Protection
Agency. Report Number EPA-R2-73-188. April 1973. 173 p.
4. Surprenant, N.S., R. Hall, S. Slater, T. Suza, M. Sussman, and C. Young.
Preliminary Emissions Assessment of Conventional Stationary Combustion
Systems. Volume II — Final Report. GCA Corporation, GCA/Technology
Division, Bedford, Mass. U.S. Environmental Protection Agency,
Research Triangle Park, N.C. Report Number EPA 600/2-76-046b.
March 1976. 531 p.
5. Yan, C.J. Evaluating Environmental Impacts of Stack Gas Desulfurization
Processes. Environ Sci Technol. 10_:54-58, January 1976.
235
-------
APPENDIX C
PROCESS DESCRIPTIONS AND FLOW DIAGRAMS OF RESIDUAL
OIL DESULFURIZATION TECHNIQUES
FLEXICOKING
Flexicoking is the advanced EXXON fluid coking process with coke gasifica-
tion. It produces very low sulfur fuel oil blendstocks (0.4 wt percent)
from a wide range of residuum feeds. About 99 percent of a typical vacuum
residuum is converted to liquid and gaseous fuel products and about 95 per-
cent of the total sulfur in the residuum feed is removed and recovered as
elemental sulfur. The remaining 1 percent feed is then converted into a
low-sulfur coke purge containing the bulk of the metals contained in the
feed. Approximately 50 percent of the combined nitrogen in the feed is
converted to N_.
Process Description
The flow diagram of a Flexicoker unit is shown in Figure C-l. Vacuum
residuum is cracked at 482 to 538°C (900 to 1000°F) and about 1.7 bar
(10 psig) in a fluidized coke bed, yielding a wide range of gaseous and
liquid products plus coke. Vapor products leave the reactor and are
quenched in a scrubber where entrained coke is removed and a heavy recycle
feed is condensed. The final hydrocarbon products are then separated in
a conventional fractionator.
In the Flexicoking process, coke from the conventional coker reactor
circulates through the heater vessel and gasifier where the coke is gasified
by steam and air or oxygen. The heat required for the residuum cracking
236
-------
SCRUBBER
FRACTIONATOR
HEATER
GASIFIER
VENTURI
SCRUBBER
N>
OJ
LOW SULFUR
REACTOR
Figure C-l. Flexicoking unit
-------
reaction is supplied by the sensible heat removed from the gasifier product
gas and the hot solids stream circulating between the gasifier and heater.
The hot product gas is then cooled in a waste heat boiler, scrubbed to
remove fines and desulfurized. The fines are a metals-rich residue,
containing 99 percent of the metals in the feed, and are thus a potentially
valuable by-product for sale to the metallurgical industry.
About 20 to 25 percent of the total feed sulfur is liberated as H-S in
the reactor and appears in the off 'gas. Essentially all of the H~S is
removed from this gas by amine scrubbing. Over 90 percent of the sulfur
present in the liquid products, amounting to 40 to 45 percent of the total
sulfur in the feed, is removed by hydrotreating. The remaining 30 to
40 percent of the feed sulfur is concentrated in the reactor coke.
In the heater/gasifier the majority of coke-sulfur is gasified with the
coke. About 97 percent of the coke-gas sulfur will be present as H?S
which can be removed by commercially available processes (e.g., a Glaus
plant). The total sulfur content of the resulting fuel gas can easily
be reduced to about 250 ppm, which is equivalent to a 0.3 wt percent
sulfur fuel oil. The remainder of the coke sulfur, less than 1 percent
of the feed sulfur will be found in the solids purge (see Figure C-2).
Stage of Development
Exxon Research and Engineering Company has recently operated the world's
first Flexicoker. It is rated at 750 bbl/day and converts vacuum residue
and tar materials into liquid and gaseous products. Although it is not
of commercial scale, it is larger than a conventional pilot-plant. As
of March 1975 the unit had been operated for approximately 6 months. The
first commercial Flexicoker, rated at 22,000 bbl/sd is under construction
at Toa Oil Company's Kawashi refinery in Japan. It is due for start-up
the first quarter of 1976.
238
-------
s^onpoad gui;>iooTxa-[d '3-0
03 dn :
'(TT° T3n3 3U3-[BA-jnba 5 % 3« Z'O) uldc' OZ.1 U6q3
SutssaooJd uiB3J3sunop Uf uoT3BZ7anj-[ns3p %
03 pazjanj^nsap aq UBO scS 35^03 aqj,
aq UBD ^ooqspaaj Uf g 30 °^
uo paseq uofjBZTanj-fnsap aaajB PUB aaojsq
isn,i •
XasAonoa
stcqau; aoi P33J •
jauinsuoo XB^OI oq
sa-[BS ao ^anj Xaau-jjaa •
saqBUfaSTp ^0 ssS QK
03 uoisasAuoa aaqqanj •
S9-[BS lanj s «°T •
sanqoBjnuBiu 3^5 •
Suipua^q SB§ OK •
jan; Xjaufjay •
aan33Bjnupin H •
XasAooaa SB8-dl •
•[anj Xaaufja^ •
uoiq-tsodsTp 3-[qTsso,j
anj-[ns \\ >
UinT pBUBA %9~
( OM «°T)
jDs/nqg 521
S mdd 0^1>
uin-fpBUBA mdd 5>
IdV SI~
Bs %c-o>/s %i-e
Bs °ao'o>/s %i-o
jos/nqg OOC'T
ES TTN/%^
safqf[Bnb qonpoaj
T A Sl/% 3" 01
% 3« ei
paaj uo PI3TA
qonpoaj
aaTj7SB8/aa3B3H
1 1 1 ^f»
u> p
4> O
ft
SB8 3>JOO " *
?« 0
^x O
Pi-
(J0c/f-0nqr) "
-[T° seS ja^oo
* A odd 068
s we
5 aaqqnjos »DO 5-92
Uo09£ 0) /joqoBB^ (-KtoOSO'T)
cqsqdBU a3Hoo 4 — paaj annpisaa
oinnoBA
^3 8B8 a033B3H
CO
cs
-------
Economics
The economics of the Flexicoking process are presented in Table C-l.
Table C-l. ECONOMICS OF FLEXICOKING PROCESS
Investment, battery limit onsite, (Basis: Direct
material and labor 2nd quarter 1973 U.S.
Gulf Coast), $ per bpsd capacity 650-800a
Typical requirements, units/bbl, feed: (140-180)
Steam export, (600 psig), Ib (140-180)
Steam required, (150 psig), Ib 50-70
Electricity, kwh 14-18
Water, cooling, gal 8.10
Water, boiler feed, gal 20-30
Air, scf 12-16
Updated economic data are presented in Appendix D.
GULF HDS3'5"8
The Gulf Hydrodesulfurization (HDS) process can upgrade high sulfur at-
mospheric residuum to low sulfur fuel oil, to minor amounts of low
sulfur naphtha, and to middle distillate. Several different crudes
have been charged to commercial units either individually or as mixtures.
A listing of these crudes is presented in Table C-2. The Gulf HDS process
offers the flexibility of producing a wide range of low sulfur products
(0.1 to 1.0 wt percent) depending upon the number of catalytic reactors
installed in the system. A one reactor system (Type II) using Kuwait
53 percent reduced crude can produce a product with 1 percent sulfur, a
two reactor system (Type III) can produce a 0.3 percent product and a
three reactor system (Type IV) can produce a product with 0.1 percent sulfur
240
-------
Table C-2. REDUCED CRUDES TO HDS UNITS
Crudes charged (separately or in mixtures) to commercial units:
Kuwait Iranian Heavy (20%)
Murban Sumatra
Nigerian Arabian Light
Forcados
Additional possible crudes that could be charged include:
West Texas Kirkuk
Arabian Medium Ratawi
Arabian Heavy Khafji
Safaniya Iranian Light
Zubair Rostam
Darius Alaskan
Process Description
A process diagram of the Gulf HDS Unit is shown in Figure C-3. The second
and third reactors are present depending on the percent sulfur desired
in the final product. The reactor charge consists of fresh filtered feed
from a desalted crude, recycle gas and makeup hydrogen. This mixture
is heated to 343 to 454°C (650 to 850°F) prior to entering the reactor.
Hydrogen rich gas from the reactor is flashed into a high presuure separator.
The separator gas is purified prior to being recycled back to the reactor.
The liquid bottoms from the high pressure separator pass through a low
pressure separator to remove H_S and fuel gas. The remainder enters a
fractionator for the separation of naphtha, middle distillate and fuel
oil.
Economics
Economic data shown in Table C-3 are based on Kuwait 53 percent reduced
crude.
241
-------
SULFUR
FUEL
ACID GAS I
LIGHT H.C.
Figure C-3. The Gulf HDS process — Type IV process uses three
reactors whereas Type II uses only the first and
Type III uses the first two'
Table C-3. ECONOMICS OF GULF HDS
Type unit
% Feed sulfur
Q
Typical requirements
(Basis: 50,000 bpsd)
Unit cost, U.S. Gulf Coast, $ MM
Hydrogen consumption, scf/bbl
feed
Utilities, average
Power shaft, kW
Steam, 50 psig, M Ib/hr
Fuel, MM Btu/hr
Water, cooling, 20 F rise, gpm
Condensate, gpm
II
1.0
21
515
12,300
67
140
5,200
50
III
0.5
27
III
0.3
740
16,500
75
190
6,200
65
IV
0.1
32
900
17,500
81
190
6,200
65
Cost estimates as published May 1973 — updated economic data are
presented in Appendix D.
242
-------
Stage of Development
The development of the Gulf HDS process is presented in Table C-4.
Table C-4. DEVELOPMENT OF GULF HDS
Company
Date on-stream
Type unit
Charge stock (design)
Charge capcity, bpsd
Fuel oil product sulfur,
wt %
Cycles per year
Nippon
Mining
12/22/69
I
28,000
1.0
2
Idemitsu
3/5/72
II
Kuwait re
40,000
1.0
2
Okinawa
4/26/72
II
duced cru
38,000
1.2a
2
Mitsubishi
Oil
Under
construction
II
in
45,000
1.0
1
Design charge is 800 F (38%) Kuwait reduced crude. When charging
650°F+ Kuwait reduced crude, fuel oil product sulfur is 1.0 percent.
RCD ISOMAX
3,9-12
In the RCD Isomax process residual oils of low to moderate metals content
(nickel and vanadium content less than 100 ppm) and sulfur as high as
5 to 6 wt percent are desulfurized directly to yield a heavy fuel oil
product containing as little as 0.3 wt percent sulfur. Yields for RCD
Isomax processing of Kuwait reduced crude to 1, 0.7, and 0.3 wt percent
sulfur are shown in Tables C-5, C-6, and C-7.
243
-------
Table C-5.
YIELDS FOR RCD ISOMAX PROCESSING OF
KUWAIT REDUCED CRUDE TO 1.0 WT
PERCENT SULFUR
Feed oil
Chem H? (scfb)
NH3
H2S
Cl
C2
C3
C4
C - 400°F
400°F+
Total
Wt - 7o
100.00
0.95
0.09
3.14
0:23
0.09
0.10
0.06
1.64
95.60
100.95
LV - 7o
100.00
(600)
0.1
2.0
99.6
101.7
°API
16.3
51.2
22.4
S, wt - 7o
3.92
0.01
0.95
Table C-6. YIELDS FOR RCD ISOMAX PROCESSING OF KUWAIT
REDUCED CRUDE TO 0.7 WT PERCENT SULFUR
Feed oil
Chem H2 (scfb)
NH3
H2S
C1~C4
C5 - 400°F
400°F+
Total
Wt - 70
100.00
0.98
0.11
3.83
0.45
1.10
95.49
100.98
LV - 7o
100.00
(614)
1.35
99.51
100.86
°API
15.8
58.9
22.1
Sulfur,
wt - 7o
4.1
0.01
0.71
Nitrogen,
wt - %
0.23
0.19
Viscosity,
cst@
122°F
330
74
244
-------
Table C-7. YIELDS FOR RCD ISOMAX PROCESSING OF KUWAIT
REDUCED CRUDE TO 0.3 WT PERCENT SULFUR
Feed oil
Chem H2 (scfb)
NH3
H2S
Crc4
C5-400°F
400°F
Total
Wt - %
100.00
1.18
0.13
4.06
0.89
2.61
93.49
101.18
LV - %
100.00
(750)
3.30
98.69
101.99
°API
15.8
55
24.0
Sulfur,
wt - %
4.1
0.01
0.3
Nitrogen,
wt - 7»
0.23
0.13
Viscosity,
cst @
122°F
330
50
Process Description
A schematic flow diagram of a typical RCD Tsomax process is shown in
Figure C-4. The basic elements employed in this process are similar to
those used in many distillate hydrodesulfurization units.
REDUCED
CRUDE
COMPRESSOR
-o-
RECYCLE H.
COMPRESSOR
SEPARATOR
GASES
- HYDROGEN
GASOLINE
MIDDLE
DISTILLATE
^
-A.
I
LOW SULFUR
FUEL OIL
Figure C-4. Typical RCD Isomax unit flow diagram
245
-------
Reduced crude, makeup and recycle hydrogen are pretreated prior to
entering the RCD reactor. Depending on the design, multiple series flow
and/or multiple reactor trains may be used. Reactor effluent is cooled
and directed to a high pressure separator where recycle hydrogen and a
liquid product are recovered. Separator liquid is sent to a low pres-
sure flash drum where the major portion of dissolved hydrogen and light
co-product gases are flashed off. The flashed liquid is charged to a
fractionator for separation into individual products or sent to a product
stripper for flash point control without separate recovery of distillate
product.
Stage of Development
As of September 1974 three RCD Isomax units were on-stream and two others
were being designed. Total capacity of the five units is 175,000 bpsd.
Economics
The economics of the RCD Isomax process are presented in Table C-8.
RESIDUE DESULFURIZATION (BP PROCESS)13'14
The British Petroleum Company has developed a process to reduce the sul-
fur content of Kuwait atmospheric residue from 4 to 1 wt percent. A
50,000 bbl/sd unit has been designed with the following goals:
• High desulfurization activity;
• High tolerance to metals accumulation;
• Low cracking activity to give high fuel-oil yields;
• Low denitrogenation activity;
• Low hydrogen consumption;
• Low catalyst cost.
246
-------
Table C-8. ECONOMICS OF RCD ISOMAX PROCESS
Basis:
Utility values:
50,000 bpsd charge
330 operating days per year
Electricity 1C kWh
Fuel 42 C MM Btu
Cooling water 2C /1,000 gal
Steam, 73C /1,000 Ib
Investment, 10 $
Operating costs
Direct operating costs
Labor
Utilities
Catalyst and royalties
Maintenance, taxes and
insurance @ 6% of plant cost
Subtotal
Indirect operating costs
Administrative at 100% of labor
Interest and depreciation
8%, 10 years
Subtotal
Total operating costs, ex. 11
Hydrogen, at 60C/Mcf
Grand total
1% sulfur
in product
21.0
$/SD
624
5,146
3,150
3,818
12,738
624
10,062
10,686
23,424
21,000
44,434
C/bbl
1.2
10.3
6.3
7.6
25.4
1.2
20.1
21.3
46.7
42.0
86.7
0.3% sulfur
in product
31.0
$/SD
772
5,428
4,511
5,636
16,347
772
14,853
15,625
31,972
27,000
58,972
C/bbl
1.5
10.9
9.0
11.3
32.7
1.5
29.7
31.2
63.9
54.0
117.9
Economic data published May 1973
presented in Appendix D.
— updated economic data are
247
-------
Process Description
A flow diagram of the BP Residue Desulfurization process is shown in
Figure C-5.
rTO HtS
REMOVAL
FRACTIONATOR
DESULFURIZEO
RESIDUE
Figure C-5. BP Residue Desulfurization process
Atmospheric residue feed and hydrogen-rich recycle gas are brought to
the required reaction temperature and are passed through a guard chamber
prior to entering the main reactor. Recycle gas is also sent to the main
reactor as a quench.
The reaction effluent passes to a separation system and the liquid prod-
ucts are recovered by distillation. To prevent the deposition of ammo-
nium sulfide, water is injected into the recycle gas stream. The recycle-
gas treatment process involves the partial removal of methane and hydrogen
248
-------
sulfide necessary for maintaining an adequate hydrogen partial pressure
in the reaction section.
Stage of Development
As of 1971 several pilot plant tests had been performed. Typical product
yields obtained at catalyst mid-life are shown in Table C-9. The catalyst
used in this process was selected not on the basis of regenerability but
rather on its ability to maintain selective desulfurization activity, its
low cost and its high tolerance for metals deposition.
Table C-9. PILOT-PLANT DATA
Yield on feedstock, wt %
Yield on feedstock, vol %
Specific gravity, 60/60°F
Flash point, °F
Sulfur content, wt %
Viscosity kinematic at 50°C centistokes
Viscosity kinematic at 60°C centistokes
Viscosity kinematic at 77°C centistokes
Viscosity kinematic at 99°C centistokes
Pour point, °F
Conradson carbon residue, wt 7»
Asphaltenes , wt 70
Ash content, 'wt %
Iron content, ppm
Vanadium content, ppm
Nickel content, ppm
Sodium content, ppm
Nitrogen content, ppm
Desulfurization, wt %
Demetalization (V + Ni only) , wt %
Denitrogenation, wt %
Filtered
feed
0.955
245
3.87
220
126.3
57.2
26.1
65
9.3
2.2
0.010
3
48
16
17
1,975
75
77
21
Total fuel-
oil product,
> 177°C
96.73
99.74
0.915
245
0.96
55
35.5
19.6
10.7
45
4.5
0.7
0.001
3
10
5
6
1,555
75
77
21
249
-------
Economics
The economics of the BP process are presented in Table C-10.
Table C-10. ECONOMICS OF BP PROCESS
Sulfur content, wt %
Yields, vol %
Light gasoline (C -180°F)
Heavy gasoline (180-350°F)
Kerosene (350-438°F)
Gas oil (438-626°F)
Fuel oil residue ( > 626°F)
Total
Chem H» consumption, scf/bbl
Investment (Basis: 50,000-bpsd unit
to desulfurize Kuwait atmospheric
residue, estimated erected cost -
materials and direct labor - mid
1973. U.K. location, excluding
initial catalyst charge), $ per bbl
charge
Operating cost (exluding capital charges
and allowing no credits for recovered
sulfur or distillate produced), $ per
bbl fuel oila
Feed
4.0
4.6
95.4
100.0
Products > 350°F
1.0
0.4
0.6
1.5
10.5
88.7
101.7
625
355
1.15
0.5
0.6
1.
1.9
11.0
87.8
102.4
835
410
1.50
0.3
1.0
2.8
4.1
13.9
81.3
103.1
1050
1.90
Based on hydrogen at $1 per 1,000 scf and fuel at $1.50 per MM Btu —
updated economic data are presented in Appendix D.
RESID HYDROPROCESSING (STANDARD OIL CO. INDIANA)
15,16
A fixed bed catalytic hydrodesulfurization process developed by the
Standard Oil Company of Indiana uses a proprietary catalyst that resists
poisoning by sulfur, nitrogen, metals, coke-forming materials and other
250
-------
troublesome constituents of resids. Typical feedstocks consist of atmos-
pheric and vacuum resids from Khafji, Gach Saran, Cyrus, Jobo, Darius,
El Morgan, Kuwait, West Texas and Mid-Content crudes. Low sulfur fuels
ranging upward from about 0.3 percent sulfur can be obtained from these
crudes. The catalyst utilized in the process enables desulfurization to
be carried out at large hydrogen partial pressures and low catalyst usage.
The catalyst is also highly tolerant to metal contaminants and is speci-
fically designed to overcome pressure drop problems. Table C-ll presents
yield data for four types of resid feeds.
Process Description
A process diagram of the hydrodesulfurization process is shown in Figure C-6.
Makeup recycle hydrogen is combined with resid feedstock prior to entering
a prereaction furnace. Heated material from the furnace passes through
the multibed reactor and then into a high pressure separator where vapors
are separated from the liquid. The bottoms from the high pressure separa-
tor are further processed in a low pressure separator, where additional
liquid and vapor streams are generated. The vapor streams from both
separators are scrubbed for H_S removal and are used as recycle gas in the
reactor. The hydrocarbon liquid streams from the two separators are
fractionated into desulfurized resid and lighter products. The desulfurized
resid can be blended with other fuels or processed further to recover
gas oil.
Economics
The economics of the hydrodesulfurization process are presented in
Table C-12.
251
-------
Table C-ll. HDS YIELD DATA
Ni
wt %
C5-360°F, vol %
360-650°F, vol 7.
650°F+, vol 7.
Gas oil, vol 7o
Residuum, vol 7«
Product quality, 650°F+
Sulfur, wt 7.
°API
Ni + V, ppm
Viscosity, cs at 122°F
o
Pour point, F
Ramscarbon, wt 7.
Kuwait
15.1
4.02
69
400
55
0
560 600
0.46 0.87
1.4 1.9
8.7 11.7
91.0 88.0
1.0 0.5
19.9 21.4
16 9
270 140
25 25-
Sour West Texas
15.4
3.65
41
300
80
13.3
700
1.4
3.0
26.8
72.1
57.1
15.0
Gas oil DeS. resid
0.23 0.7
24.8 9.8
0.1 31
72 80a
100 155b
1.0 16
Khaf ji
12.3
4.47
141
3,000
70
0
780
0.70
2.2
10.9
89.5
0.65
20.5
57
270
60
Gach
Saran
14.6
2.55
258
650
75
0
400
0.80
2.4
11.8
88.0
0.40
20.5
67
260
65
aAt 250°F.
Softening point.
-------
FRACTIONATOR
to
Ul
CO
REACTOR(S)
FURNACE
RECYCLE GAS
COMPRESSOR
MAKEUP GAS
COMPRESSOR
(-
- ^
A ?
r-J~— , 4
( >nV 3
^ •A VVV \
1 2
v.
wa^H /*^ •• i* •'
WATER ^A,
X
/
S
i.
(
J
\
<- \
^
«
X
s,
s
s
1
I
'
1
1 Ulft
J s
RESID FEED
*£
COLD
HIGH PRESSURED.
SEPARATOR
HOT
HIGH PRESSURE
SEPARATOR
LOW PRESSURE
SEPARATOR
MAKEUP HYDROGEN
LIGHT HYDROCARBON
RESID
Figure C-6. Resid Hydroprocessing - Standard Oil Co., Indiana
-------
Table C-12. ECONOMICS OF HDS PROCESS3
Investment, (Basis: desulfurizing 20,000 to 40,000 bpsd
of Kuwait atmos. resid to 1.0 wt % sulfur in 650°F+
product, January 1973, Gulf Coast), $ per bpsd capacity 560-620
Typical requirement, unit per bbl feed
Electricity, kWh 4.4
Steam, Ib 2.6
Fuel, M Btu 86
Water, cooling, gal , 160
Water, process gal 4.2
Catalyst $ 0.08
Average hydrogen consumption scf 560
a
Includes amine recovery and regeneration.
Updated economic data are presented in Appendix D.
LC-FINING10'15'29
The LC-Fining process can be used for the desulfurization of atmospheric
residuum, vacuum bottoms and other heavy oils. Efficient hydrocracking
is employed to convert heavy gas oils or residues into lighter fractions.
Process Description
Hydrogen and heavy oils are reacted in a fluid bed consisting of vapor
and liquid in which solid catalyst particles are maintained in random
motion by continuous upflow of the liquid phase. Two types of catalyst
can be used, a 1/32-inch extrudate or a fine powder. The extrudate
form requires an internal liquid recycle to expand the catalyst bed.
The powder form is fed into the reactor mixed with the fuel oil and does
not require the internal pumped liquid recycle for fluidization. At
equilibrium operation, catalyst leaving the reactor with the fuel product
is replaced by adding catalyst with the feed. Catalyst replacement is
done on a daily basis. Because the bed is in a continuous motion, contact
254
-------
between the catalyst and the oil is greatly improved, resulting in longer
catalyst life and the capability to process high metals feedstock.
Figure C-7 is a schematic 9f the LC-Fining process.
Stage of Development
The development of the LC-Fining process is presented in Table C-13.
Table C-13. DEVELOPMENT OF LC-FINING PROCESS
Commercial installations
Unit
Lake Charles, La.
Shuaiba, Kuwait
Salamanca, Mexico
Lake Charles, La.
Kashima, Japan
Capacity, bpsd
6,000
28,800
18,500
25,000
10,000
Status
In operation
In operation
In operation
In design
In design
Years
12
7
2
Economics
The economics of the LC-Fining process are presented in Table C-14.
RESID ULTRAFINING3'19
A proprietary process developed by Amoco, called Resid Ultrafining, has
been used to desulfurize numerous resids of widely varying properties
in both bench scale and large pilot plant equipment.
Process Description
Resid feed is preheated then combined with recycle hydrogen and heated to
reactor inlet temperature in a furnace (see Figure C-8). Desulfurization
takes place in a multibed reactor where an intermediate gas quench
255
-------
rs>
Ul
HYDROGEN
H-OIL
REACTORS
_y
Al
V ,
OIL
A/W
FURNACE
CHARGE
-o
RECYCLE HYDROGEN
STEAM
GA!
STABILIZER
LOW SULFUR^
FUEL OIL
Figure C-7. LC-Fining process flow diagram
-------
Table C-14. DESULFURIZATION OF KUWAIT ATMOSPHERIC BOTTOMS
Objective: Production of 1.0 wt % sulfur in 650 F product
Throughput: 40,000 bpsd (stream factor of 0.9)
Feed Inspection: 650°F+, 15.0 °API, 4.05 wt 7, S, 49.6 vol 7. 975°F+
Yields
Wt 7. Vol 7, API Wt 7. S
(On fresh feed)
H2S
NH3
Cl
C2
C4
400
650
975°
C3
- 400
- 650
- 975
F+
0
o
o
F
F
F
3.
0.
0.
0.
0.
3.
23.
38.
30.
4
1
6,
6
6
6
7
2
2
—
—
4.
26.
40.
29.
5
6
8
8
54
33
25
13
-
—
.0
.0
.1
.0
<0
0
0
1
-
—
.1
.2
.5
.7
650°F+
101.0 101.7 24.4 0.8
68.4 70.6 19.8 1.0
Hydrogen consumption —
Chemical
Losses3
Total:
Catalyst replacement —
650 scf/bbl
260 scf/bbl
910 scf/bbl
8c per barrel of feed oil
Estimated investment —
Installed cost LC-fining unit 30.0 MM$
Initial catalyst charge 0.6 MM$
Royalty 2.8 MM$
Total: 33.4 MM$
Utilities -
Power, kW 6,300
Heat @ 75% efficiency, MM Btu/hr 219
Recoverable heat,c MM Btu/hr Kf8
Cooling water @ A25°F, gpm 4,590
Labor, operators/shift 3
This is with the use of a purge system for purification of recycle gas.
This figure includes major equipment, material, piping, labor, purchasing,
engineering, field expenses, a contractor's fee of 6% and 57, for contin-
gencies. Product fractionation is not included. Hydrogen is assumed
available at 300 psig and 95 mol 7o purity. The investment is calculated
in U.S. dollars on a 2nd quarter, 1975 basis at a Gulf Coast location.
Enthalpy in reactor liquid stream above 400°F.
257
-------
ho
ui
00
COMPRESSOR
COMPRESSOR
MAKEUP HYDROGEN
•AAA/— ->
FURNACE
RESID FEED
RE ACTOR (S)
WASH.
GAS
SCRUBBING
H9S RECOVERY
FUEL GAS (SULFUR-FREE)
WATER ££
\
x
k
}
r
A. HOT
JHIGH
PRESSURE
^J SEPARATOR
' rf
s^
£
^
>
f
p
. '
COLD
HIGH
^PRESSURE
^SEPARATOR
f
\
^ .
\
±
f^\ NAPHTHA .
I ^
I DISTILLATE >,.
| DESULFURIZED .
RESID
SOUR WATER ^
LOW PRESSURE
SEPARATOR
Figure C-8. Residual Ultrafining
-------
maintains the temperature and stability of the catalyst. A hot high-
pressure separator splits the reactor effluent into a vapor and a bottom
stream. The vapors are condensed in an exchanger system leading to a
cold high-pressure separator. The hydrogen rich gas is scrubbed with
amine to remove hydrogen sulfude and is recycled back to the reactor.
Bottoms from the hot high-pressure separator are mixed with the condensed
material from the cold high-pressure separator and then flashed in a low-
pressure separator. Liquid from this separator is ultimately fractionated
into fuel gas, naphtha, distillate and desulfurized resid. The gas streams
from the low-pressure separator and the fractionator are scrubbed to pro-
duce a sulfur-free fuel gas.
Preliminary data based on bench-scale runs has shown that catalyst life
can be expected to last at least 9 months. Deposits of coke as well as
nickel and vanadium sulfides have a tendency to shorten catalyst life.
Stage of Development
Numerous resids of widely ranging properties have been desulfurized in
both bench-scale and large pilot plant equipment.
Economics
The economics of the Resid Ultrafining process are presented in Tables C-15
and C-16.
^ 20-92
GO-FINING (EXXON RESEARCH AND ENGINEERING CO.)
Go-Fining is a proprietary process for handling high boiling virgin and
cracked gas oils. It is a fixed bed system and operates at pressures of
286 to 562 bar (400 to 800 psig).
259
-------
Table. C-15. RESID ULTRAFINING DESULFURIZATION COSTS,
BASIS: 40,000 bpsd
Resid
650*°^ product sulfur, v;t %
On-site investment, MM$
Catalyst charge, MM$
Off-site investment, MM$
Total investment, MM$
Cost, C/bble
Hydrogen
Utilities and chemicals
Catalyst
Labor
Investment related
Total, C/bbl
Xhaf ji
1.0
17.3
1.1
5.7
24.1
33.5
14.5
15.2
1.7
41.1
106.0
West Texas
sour
1.0
13.0
0.4
4,4
17.8
25.5
13.7
5.1
1.7
30.5
76.5
Difference
-
4.3
0.7
1.3
6.3
8.0
0.8
10.1
-
10.6
29.5
78 percent S recovery.
74 percent S recovery.
°Current U.S. Gulf Coast cost (published May 1973) -updated
economic data are presented in Appendix D.
Includes working capital.
£>
Per barrel of charge.
Table C-16. WEST TEXAS SOUR DESULFURIZATION COSTS,
BASIS: 40,000 bpsd
1 Q
650 F product sulfur, wt %
c
On-site investment, MM?
Catalyst charge, MM$
Off-site investment, MM$
Total investment, MM$
Costs, C/bbl
Hydrogen
Utilities and chemicals
Catalyst
Labor
Investment related
Total, c/bbl
0.3U
17.3
1.1
5.7
24.1
34. G
14.5
15.2
1.7
41.1
107.1
i.ob
13.0
0.4
4.4
17.8
25.5
13.7
5.1
1.7
30.5
76.5
Difference
4.3
0.7
1.3
6.3
9.1
0.8
10.1
10.6
30.6
92 percent S recovery.
3_.
74 percent S recovery.
c,d,e
See footnotes for Table C-15.
260
-------
Process Description
Fuel oil and hydrogen rich gas are preheated and fed into a catalytic
desulfurization reactor. After being cooled in a heat exchanger, the
hydrogen gas is separated from the oil, desulfurized, and either recycled
or used in another part of the plant. The desulfurized oil is ultimately
stripped of small amounts of low-boiling products and used directly as
fuel or stored as a low-sulfur blending stock. Figure C-9 is a schematic
of the Go-Fining process.
RECYCLE H,
H-S
OESULFUHIZED
FUEL OIL
Figure C-9. Go-Fining
Depending upon the metals content of the feedstock, the catalyst may be
regenerated for longer life and lower operating cost. Typical product
yields are presented in Table C-17.
261
-------
Table C-17. GO-FINING YIELDS AT 90 PERCENT DESULFURIZATION LEVEL
(HIGHER LEVELS MAY BE OBTAINED)
Crude source
Feed boiling range, F
°API
Sulfur, wt %
Average yields
C,-C0 (including H0S), wt %
4 o 2.
C, vol %
C5-400°F, vol %
400°F+, vol %
Sulfur, wt %
Chemical hydrogen
consumption, scf/bbl
Kuwait
22.2
3.05
, 3.1
0.08
0.8
99.0
0.3
280
Arabian
light
(?r-r> /I
23.1
2.28
2.3
0.07
0.6
99.2
0.23
220
Khafji
21.7
2.97
3.1
0.09
0.8
99.0
0.3
300
• Gach
Saran
22.4
1.91
2.0
0.08
0.7
100.0
0.19
220
Economics
Hydrogen consumption ranges from 220 to 300 scf/bbl depending on the
amount of sulfur removed and operating pressure. Economic data
are presented in Table C-18.
Table C-18. ECONOMICS OF GO-FINING
Economics of go-fining0
Investment, $
Fuel fired, 1,000 Btu
Power, kW
Cooling water, gal
Basis: total erected cost, 1971 Gulf Coast
(includes initial charge of catalyst)
Per barrel
of feed
100-220
20-40
1-2
200-350
Updated economic data are presented in Appendix D.
262
-------
Stage of Development
As of September 1972, there were approximately 390,000 bpsd of Go-Fining
capacity in units ranging from 15,000 to 80,000 bpsd. An additional
total capacity of 580,000 bpsd were in the planning, design, or construc-
tion stage.
RESIDFINING (ESSO RESEARCH AND REFINING COMPANY)3'22'23
This is a proprietary process for the hydrodesulfurization of atmospheric
residue for the production of low sulfur fuel oil.
Process Description
Residfining is a fixed-bed system operating at pressures of approximately
700 bar (1000 psig). The residual oil to be treated and hydrogen-rich
gas are preheated before entering the desulfurization reactor. Following
heat exchange and cooling, the hydrogen-rich gas is separated from the
fuel oil and recycled or used in another process. The desulfurized oil
is stripped of small amounts of low-boiling products generated in the
reaction and is used directly as fuel or stored for low-sulfur blending
stock.
The proprietary catalyst used rejects many of the asphaltenes contained
in the residuum. It has been tested using residuum feeds containing
30 to 200 ppm nickel and vanadium with satisfactory results. A schematic
is shown in Figure C-10.
Economics
Long catalyst life at low pressure is a significant economic determinant.
Operating 700 bar (1000 psig) as opposed to 1400 bar (2000 psig) results
in a lower operating cost due to reduced investment, reduced hydrogen
consumption, and reduced energy consumption. Economic data are presented
in Table C-19.
263
-------
O
MAKEUP
HYDROGEN
RECYCLE H2
O
ATMOSPHERIC
RESIDUUM
REACTOR
1,000 piig
D-a
COMPRESSOR
H2S
SWEET
FUEL
DESULFURIZED
PRODUCT
Figure C-10. Schematic of the residfining process
Table C-19. ECONOMICS OF RESIDFINING PROCESS
Economics of
residfininga
. o
Investment, $
Power consumption, kW
Fuel fired, 1,000 Btu
Cooling water, gal
Per barrel
of feed
330-500
1.2-1.4
50-60
200-250
Economic data published September
1972.
Total erected cost: 1971 Gulf Coast;
inclusive of catalyst. Updated eco-
nomic data are presented in Appendix D.
Stage of Development
As of September, 1972, two units were in the design stage.
264
-------
RESIDUE HYDRODESULFURIZATIOl
The Badische Anilin-und-Soda-Fabrik AG and Institut Francais du Petrole
process is used to remove sulfur, nitrogen and metallic contaminants from
heavy feedstocks. Typical charges to the system are atmospheric residue,
vacuum'residue and total crude oil.
Process Description
A flow diagram of the process is shown in Figure C-ll. The feedstock and
hydrogen-rich gas plus recycle are preheated in a heat exchanger using
the reactor products. The heated charge then enters the fixed bed reactor.
After passing through the catalyst bed the reaction products are cooled
and sent to separators where the product is desulfurized and separated
from the unreacted hydrogen and light hydrocarbons. The product stream
is then stabilized in a stripper column.
Three process schemes have been designed to allow for differences in
product sulfur content, stream factor and by-product utilization:
• Vacuum gas oil desulfurization - a deep desulfurization
of the vacuum gas oil (VGO) and the blending of it with
the vacuum residue;
• Indirect desulfurization and solvent deasphalting - the
topped crude is distilled, the VGO is deeply desulfurized,
the vacuum residue is deasphalted and then desulfurized;
• Direct desulfurization of the topped crude.
See Table C-20 for product yields.
Stage of Development
Not reported.
265
-------
t-o
REACTOR
-V\A.
\/
FURNACE
MAKEUP HYDROGEN
FRES,
FEED
SEPARATORS
RECYCLE
STRIPPER
PURGE
-> PURGE
LIGHT
STEAM DISTILLATE
U0:
PRODUCT
Figure C-ll. Residue hydrodesulfurization flow diagram
-------
Table C-20. RESIDUE HDS PRODUCT YIELDS
Feedstock:
Specific gravity
ASTM distill., IBP, °F
Sulfur content, wt %
Pour point, F
Desulfurization rate, °L
Yields (mid-run), wt% on feed:
H2S + NH3
C1'C4
C - 302°F
302-482°F
482 - 662°F
662°F
Total
Product quality:
Gas-oil, 482 - 662°F
Specific gravity
Sulfur content, wt%
Pour point, F
Fuel oil, 662°F+
Specific gravity
Sulfur content, wt%
Pour point, F
Ratawi
crude
0.985
563
5.1
54
80
4.50
0.80
0.60
2.75
10.50
82.50
101.20
0.870
0.065
10.5
0.941
1.20
54
Kuwait
crude
0.969
574
4.1
54
80
3.60
0.35
0.30
2.50
10.40
82.81
100.96
0.867
0.045
10.5
0.924
0.95
49
267
-------
Economics
The economics of the Residue Hydrodesulfurization process are presented
in Table C-21.
Table C-21. ECONOMICS OF RESIDUE HDS
A detailed engineering study for a unit treating 45,000 bpsd
of Kuwait atmospheric residue at a desulfurization rate of
80 percent gives:a
Investment, $ per bpsd capacity
Erected battery limits 344
Catalyst, first charge 19
Typical requirement, units per bbl feed
Electricity, kWh 3.2
Steam (medium pressure), Ib 25
Fuel (absorbed heat), M Btu 99
Hydrogen consumption, scf 650
Catalyst life, ultimate, months 12
a
Economic data published September 1972.
Updated economics data are presented in Appendix D.
HYDRODESULFURIZATION, TRICKLE
FLOW25
Hydrodesulfurization, Trickle Flow, improves the quality of petroleum
fractions ranging from kerosene to heavy gas oil, as well as vacuum
flashed distillate by the removal of sulfur and by the hydrogenation of
unsaturated components.
Process Description
As shown in Figure C-12, feedstock, combined with hydrogen-rich make-up
and recycle gas, is passed through a feed/effluent heat exchanger prior
to entering a furnace, where the temperature is raised to 332 to 400°C
268
-------
(630 to 750°F). The heated charge is then passed through the reactor in
a trickle flow. After being cooled the product is flashed in a high-
pressure separator at a temperature of 38 to 49 C (100 to 120 F) or for
extra heavy gas oils at 149 to 177°C (300 to 350°F). The liquid product
is pumped to a work-up section where H~S and dissolved gases are removed.
The gas leaving the high-pressure separator is used as recycle gas.
Typical yields from the HDS of thermal cracker gas oil are shown in
Table C-22.
Table C-22. TYPICAL RESULTS FROM HYDRODESULFURIZATION OF
THERMAL CRACKER GAS OIL (380-650°F FRACTION)
Specific gravity 20°/4°C
Sulfur content, wt %
Bromine number, g/100 g
Maleic anhydride value, mg/g
o
Pour point, C
Cloud point, C
Desulfurization, %
Chemical H_ consumption, scf/bbl
% sulfur removal
Feedstock
0.8469
1.33
23
5.2
-13
-9
88
Product
0.8326
0.16
1
—
-16
-9
88.0
315
Stage of Development
At the end of 1973, 82 units with a combined capacity of 1,050,000 bpsd
were operating.
269
-------
FRESH GAS |—,
/f U-*"
->*
RECYCLE GAS
KNOCK
OUT
HYDROGEN 1 I DRUM
MAKEUP
-------
Economics
The economics of the HDS, Trickle Flow, process are described in Table C-23,
Table C-23. ECONOMICS OF HYDRODESULFURIZATION PROCESS
Typical requirement, unit per bbl middle distillate
Electricity, kwh 1.2
Steam (200 psig), Ib 9.6
Fuel, M Btu 52.8
Water, cooling (30°F rise), gal 260
Catalyst consumption, Ib 0.01
IFF RESID AND VGO HYDRODESULFURIZATION3
The Institut Francais du Petrole's (IFF) HDS process is a catalytic fixed
bed operation. This process can be used to improve heavy petroleum stocks
by removal of sulfur, nitrogen and metallic contaminants. Typical charges
to the reactor are atmospheric residues, vacuum residue and total crude
oil. Desulfurization can reach 85 percent.
Process Description
A flow diagram of the IFF hydrodesulfurization process is shown in
Figure C-13.
Feed and makeup hydrogen are mixed with a portion of the recycle gas and
are then fed down through the catalyst beds. The remaining portion of the
recycle gas is used as a temperature regulating quench in the reactor. The
reaction products are cooled and sent to a high-pressure separator where
hydrogen-rich gas is removed and recycled to the reactor. The product is
stabilized in a stripper column where light ends and residual H S are
removed.
In this process, IFF employs a cobalt molybdate catalyst in the form of
extrudates 1.5 mm in diameter and 3 to 6 mm long.
271
-------
RECYCLE GAS
MAKEUP HYDROGEN
FEED
REACTOR
STRIPPER
'COLUMN
HIGH
PRESSURE
SEPARATOR
-*• GAS
+. LIGHT
DISTILLATE
PRODUCT,
400 F. + LOW-
SULFUR FUEL OIL
Figure C-13. IFF resid and VGO desulfurization flow diagram
272
-------
Tables C-24 'and C-25 give feed and product specifications for a Kuwait
residue.
Table C-24. FEED SPECS AND IFF PROCESS PERFORMANCE
Kuwait residue
Gravity
Sulfur
Nitrogen
Metals
Conradson carbon
Asphaltenes
Viscosity at 210°F
Pour point
ASTM distribution IBP
5%
50%
Desulfurization rate
Hydrogen chemical consumption
Catalyst ultimate life
15.5 API
4.1 wt %
2,500 wt ppm
63 wt ppm
9.5 wt %
2.6 wt 7»
160 SUS
52°F.
572°F.
716°F.
1,013°F.
89%
760 scf/bbl
9 months
Table C-25. YIELDS FROM KUWAIT RESIDUE
Yields, mid run
Crc4
C5-400°F
400°F+
Long residue 400 F+
Gravity
Sulfur
Flash point
Metals
Viscosity at 210°F
Asphaltenes
3.85 wt% on feed
0.55 wt% on feed
3.0 wt% on feed
93.8 wt% on feed
(=99.50 vol% on feed
24.8
0.50 wt%
300°F.
17 ppm
80 SSU
0.6 wt%
273
-------
Stage of Development
As of September 1972 two plants were in operation, one in Japan and the
other in the Near East.
Economics
The economics of the IFF HDS process are presented below in Table C-26.
Table C-26. TYPICAL ECONOMICS OF IFF HDS PROCESS (WITH
IRANIAN LIGHT ATMOSPHERIC RESIDUE)3
Charge:
Plant capacity:
400°F+ cut:
Investment:
C/bbl feed:
Hydrogen
Catalyst
Utilities0
Investment related
Sulfur6
Totals
Net charges:
67.5 C/bbl of feed
68.2 C/bbl of fuel 400°F
650 F IBP
25 API
2.5 wt % S
40,000 b/sd
0.3 wt % S
99 vol % yield on feed
$20,000,000
Cost
24.0
11.0
5.5
31.4
71.9
Credit
4.4
4.4
As of September 1972 updated economic data are presented in Appendix D.
b35C/Mscf.
CPower iC/kW, fuel 25C/MM Btu, steam 0.08C/lb. Light products counted
for fuel.
20 percent/year of investment cost including amortization, interest,
maintenance, labor and overhead.(includes amine washing, sulfur plant
and gas cleaning).
e$15/long ton.
274
-------
DEMETALIZATION/DESULFURIZATION
26
As the metals content (principally vanadium) increases in residual feed-
stock, the cost of desulfurizing increases due to larger reactor volumes
and higher catalyst usage. When processing residual fuel oils with vana-
dium contents above 200 to 300 ppm, the economics may favor a scheme of
demetallization/desulfurization.
Process Description
This method uses a recently developed system employing two different
catalysts (see Figure C-14). The first catalyst system uses a demetal-
lization ebullating bed reactor containing an inexpensive natural catalyst.
This reactor is followed by one or more desulfurization ebullating bed
reactors containing a conventional Co-Mo catalyst. The demetallization and
desulfurization reactor designs are similar. The advantage of this tech-
nique lies with the natural catalyst used for demetallization, the cost
of which is about 5 to 10 percent of the Co-Mo catalyst.
FOR DEMETALLIZATION/
DESULFURIZATION
FEED OIL
HYDROGEN
DEMETALIZED
OIL a H2
A
o
o *2
-------
Stage of Development
Several pilot plant studies have been performed successfully using medium-
metals-content (400 ppm vanadium) atmospheric bottoms and Boscan high
metals crude (1100 ppm vanadium).
Economics
The economics of the demetalization/desulfurization system are presented
in Table C-27.
DELAYED COKING23'27
The process of Delayed Coking upgrades heavy residuals or bottom of the
barrel materials into more valuable distillate products and coke. By
1980 the production of coke is expected to exceed 4.5 x 10 kg (50,000 tons)
per day. Delayed coking accepts as feed material a full range of reduced
crude oils, shale oil, Athabasca bitumen, gilsonite, coal tar pitch, and
asphalt. Needle coke for electrodes in aluminum manufacture is produced
as a side product from aromatic and refractory stocks, such as catalytic
cycle oils and thermal tars.
Process Description
A flow diagram of a simplified delayed coking and fractionation section
is shown in Figure C-15. The feed material is fed directly to the bottom
section of the fractionator where material lighter than the desired end
point of the heavy gas oil is flashed off. The remaining material from
the bottom of the fractionator is combined with recycle oil and is pumped
to the coking heater where it is rapidly heated to above 482°C (900°F).
The liquid-vapor mixture then leaves the coking heater and enters a coke
drum.
276
-------
Table C-27. COST COMPARISON - DESULFURIZATION VERSUS DEMETALIZATION/DESULFURIZATION
Throughput, b/sd
Type operation
Feedstock data
Gravity, °API
V, ppm
Space velocity
Vo/hr/Vr (Case 1 = 1.00)
Hydrogen cons., scf/bbl
Investment, $ million
Major processing cost
Catalyst cost, C/bbl
Hydrogen, C/bblc
In ve stment, C/bbl
Total, C/bbl
Venezuelan medium-metals
atmospheric residuum
25,000
Desulfurization
11.8
375
0.49
720
19.8
42
54
24
120
25,000
Demetallization/
desulfurization
12.7
398
0.31
680
25.2
13
51
31
95
Venezuelan high-metals
crude, Boscan
25,000
Desulfurization
10.4
1,100
0.53
1,140
21.8
62
86
26
174
25,000
Demetallization/
desulfurization
10.4
1,100
0.49
1,030
23.9
11
77
29
117
Article published June 1975, updated economic data are presented in Appendix D.
Investment includes demetallization (if any)/desulfurization sections at a Gulf Coast location.
Hydrogen is assumed to be from steam-methane reforming at 75C/1.000 scf.
Investment payout over 10 years in C/bbl based on 0.90 on-stream factor.
-------
ACCUMULATOR
N3
-~J
00
FRACTIONATOR
8IO°F
k
^
COKE
DRUMS
Jk
30 PSIG
1
CONDENSATE
DRUM
COKE
£r~*
START]
OX
FEED
GAS
UNSTABILIZED
GASOLINE
GAS OIL
Figure C-15. Simplified flow diagram for delayed coking
-------
A coking unit usually has two drums, one on stream while the other
is being decoked. The coke units are usually designed so that each one
operates on a 48-hour cycle. The overhead vapors from the coke drum
enter the lower section of the fractionating tower for separation into
gas, gasoline, gas oils and recycle stock.
Stage of Development
Delayed Coking has been used extensively in the petroleum industry for
several years. Coking capacity by the end of the seventies is expected
to grow to between 4.1 x 107 and 4.5 x.10 kg (45,000 and 50,000 tons)
per day. As advancements in operating techniques are made, a wider range
of feed stocks /will be utilized. Several units are currently operating
successfully outside the U.S. and are designed for a coal tar pitch feed.
Economics
No data available.
12 28 29
VGO/VRDS ISOMAX ' '
The combination of a vacuum gas oil desulfurizer (VGO Isomax) with a
vacuum residuum desulfurizer (VRDS Isomax) is often an attractive alterna-
tive to direct desulfurization of atmospheric residuum (RDS Isomax). The
VGO/VRDS is an extension of RDS technology; the major difference involving
feed stock and type of catalyst.
Process Description
The separation of atmospheric residuum into a gas-oil fraction and a
vacuum-tower bottom combined with desulfurization of each stream indivi-
dually requires 35 percent less hydrogen than direct desulfurization of
atmospheric residuum as is the case in the RDS process. In addition,
279
-------
the yield of heavy fuel oil (659 F ) is 2 to 3 percent higher in the
VGO/VRDS combination process.
Vacuum gas oil is processed at considerably lower pressures than the
residue, which 'leads to a very selective hydrodesulfurization with minimum
hydrogen-consuming side reactions. Table C-28 shows the yields from
Arabian light residuum using VGO, VRDS and RDS processes. As indicated
in this table, the VGO process produces 0.1 percent sulfur product which
when combined with the VRDS product yields an overall 0.5 percent sulfur
fuel oil. Figure C-16 presents a flow diagram of a VGO/VRDS process.
Table C-28. LOW SULFUR FUEL OIL PRODUCTION
FROM ARABIAN LIGHT RESIDUUM3
Process
Feed sulfur, wt %
Product sulfur, wt %
Product yields
C1-C4, wt 7o
H2S, NH3, wt 7o
C5+, wt 7o
C5+, LV 7o
Hydrogen consumption
Scf/bbl
Scf/lb sulfur
VGO
2.3
0.1
0.59
2.44
97.51
100.6
330
47
VRDS
4.1
1.28
0.56
3.00
97.34
102.0
720
71
VGO+
VRDS
2.9
0.5
0.58
2.55
97.46
101.0
450
56
RDS
2.9
0.5
0.58
2.55
97.67
101.5
550
69
Chevron hydrotreating process yields (middle of run),
Stage of Development
The development status of VGO Isomax plants is presented in Table C-29.
280
-------
OO
C-I50°F.
NAPHTHA
HYDROTREATER
REFINERY PROCESS FUEL
I50°-350°F
ARAB. LIGHT
CRUDE
TURBINE FUEL
350°-540° F. DIESEL
0.45% S
540-1,020° F.
_\k
>LOW SULFUR
FUEL OIL
1.2% S
Figure C-16. VGO/VRDS flow diagram
-------
Table C-29. VGO ISOMAX PLANTS
Company
On-stream
Chevron Oil Co.
Fuji Oil Co.
Koa Oil Co.
Koa Oil Co.
Nippon Petroleum Refining Co.
Subtotal
Engineering and construction
Asia Kyoseki Co.
Nippon Petroleum Refining Co.
Nippon Petroleum Refining Co.
Tohoku Oil Co.
Bahrain Petroleum Co.
Kashima Oil Co.
Unannounced
Unannounced
Subtotal
Total
Location
Salt Lake City, Utah
Sodeguara, Japan
Marifu, Japan
Osaka, Japan
Negishi, Japan
Sakaide, Japan
Negishi, Japan
Muroran, Japan
Sendai, Japan
Bahrain, Arabian Gulf
Kashima, Japan
Capacity,
b/sd
5,200
23,000
8,000
12,000
40,000
88,200
15,000
28,000
40,000
35,000
50,000
25,000
60,000
36,000
289,000
377,200
As of September 1972.
Economics
Tables C-30 and C-31 present the differences in investment and processing
costs for a VGO/VRDS process and a RDS process.
282
-------
Table C-30. INVESTMENT SUMMARY
Feed: 86,000 b/cd Arabian light
650°F+ residuum
Product: 350 F fuel -oil containing
0.5 % sulfur
Processing scheme
On-plot investment,
$ millions,
relative to RDS - VGO/VRDS
Crude unit
H2 plant .
VGO, VRDS
Total
+ 7
- 3
-11
- 7
U.S. Gulf Coast estimates for Jan-
uary 1975. List does not include the
process equipment common to both
cases.
Table C-31. PROCESSING COSTS
Feed: 86,000 b/cd Arabian light
650°F+ residuum
Product: 350 F fuel oil containing
0.5 7o sulfur
Processing scheme VGO/VRDS
Amortization, $/bbl F.O.
relative to RDS , -0.05
Operating costs, $/bbl F.O.
relative to RDS
Catalyst
Hydrogen
Utilities
Other
Total
0
-0,20
-0.04
•tO. 09
-0.20
a
Includes only on-plot investment
for crude unit, hydrogen plant, and
hydrotreaters.
Utility costs are: fuel $13/bbl
equivalent fuel oil; steam $2/1,000
Ib; cooling water and process water
$0.26/1,000 gal; power $0.027/kWh.
283
-------
30 31
SHELL GASIFICATION PROCESS '
The Shell Gasification Process (SGP) in conjunction with a combined cycle
is based on a new application of SGP developed in the Amsterdam research
laboratories of Royal Dutch Shell during the early 1950's.
Process Description
The SGP involves the partial combustion of heavy, sulfur-containing resi-
dual fuels and heavy crude oils to produce a mixture of hydrogen and
carbon monoxide. Hydrogen sulfide produced during this reaction is readily
removed to yield a sulfur-free (5 ppm) fuel gas, which is used for power
generation in a typical combined cycle. Figure C-17 contains a schematic
diagram of this SGP/combined cycle process.
A simplified SGP flow diagram is shown in Figure C-18. The hydrocarbon
charge and the oxidant are preheated and fed to the reactor. Hot reactor-
effluent gas, containing about 3 percent of the feed as soot, is passed
to a waste-heat boiler, producing high-pressure saturated steam. High
heat-transfer rates assure that the temperature of the gas leaving the
waste-heat boiler is close to that of the steam produced in the boiler.
The design and construction of the waste-heat boiler are such that the
surface remains clean for an indefinite period (without using external
cleaning devices). The waste-heat boiler of the Shell prototype unit has
been in operation since 1956 and never has been cleaned on the gas side.
This waste-heat boiler can be designed for steam pressure up to about
1 kbar (1,500 psig).
Gas Cleanup
The "crude" gas leaving the waste-heat boiler at temperatures around
177 C (350 F) is then passed to the carbon-removal system. In this system
284
-------
OIL 525 MW
1.766 «I06 Blu/hr
CONDENSATE
PREHEATER
I5MW
AIR
200 MW 4
ELECTRICAL
OUTPUT
STACK
CONDENSATE
Figure C-17. Combined cycle/shell gasification process
285
-------
STEAM
PREHEATERS
HIGH PRESSURE
STEAM
to
oo
TO POWER PLANT
WASTE HEAT BOILER
CARBON SLURRY
SEPARATOR
_>. FUEL GAS TO SULFINOL UNIT
HYDROCARBON FEEDSTOCK-
CARBON
SLURRY
OL
Ul
03
CQ
u
V)
**.
cc.
UJ
_]
o
o
u
C.V
FRESH WATER
CARBON-FREE
,, CIRCULATION WATER
PELLETIZER
CARBON PELLETS
HOMOGENIZER
WASTE
WATER
Figure C-18. Shell gasification power generation block diagram
-------
bulk removal of the carbon is accomplished by contact of the gas with
water. The remaining product gas has less than 5 ppm of carbon. The
carbon produced in the gasification is recovered as a soot in a water slurry
(carbon content 1 percent to 2 percent by weight). In most cases, it
would not be possible to dispose of this untreated carbon slurry. There-
fore, Shell has developed a technique for removing carbon from the slurry
permitting the water to be reused. Depending on the metals content of the
feedstock and the economics and maintenance policy of the process operator,
Shell claims that up to 100 percent of the soot can be recycled to extinc-
tion with the fresh feed.
Sulfur in the feedstock is converted primarily to l^S and traces of COS.
The carbon-free product gas is treated in a Shell Sulfinol or ADIP process
unit, where the sulfur compounds and most of the CO- are absorbed. The
desulfurized gas is said to contain typically less than 5 ppm of sulfur.
The acid-gas effluent from the Sulfinol unit is fed to a Glaus process
unit, which recovers elemental, salable sulfur.
Either oxygen or air can be used as the oxidant depending on the desired
heating value in the product gas. Nitrogen present in the air will act
as a moderator for temperature control in the reactor. In either case,
steam is injected into the reactor for further temperature control. Air
3
3 3
oxidation produces a low-heating-value 1068 kcal/m (120 Btu/ft ) fuel gas,
while oxygen feed produces a medium-heating-value gas 2670 kcal/m
3
(300 Btu/ft ). Typical product-gas c
gasification are shown in Table C-32.
3
(300 Btu/ft ). Typical product-gas compositions for air and oxygen
287
-------
Table C-32. TYPICAL PRODUCT GAS COMPOSITION
Hydrogen
Carbon monoxide
Methane
Nitrogen
Argon
Sulfur
Total
7o vol, dry basis
02
oxidation
48.0
51.0
0.6
0.2
0.2
5 ppm
100.0
Air
oxidation
12.0
21.0
0.6
66.0
0.4
5 ppm
100.0
Economics
The economics of the Shell gasification process are presented in Table C-33,
Table C-33. POWER GENERATION COST - 200-MW STUDY*
Gross output, Mw 200.0
Power consumed, Mw 4.7
Net power output, Mw 195.3
Overall efficiency, 7= 38.0
Capital cost, $ millions (1972)
Fuel-processing unit 18.2
Power-generation unit 31.4
Total capital cost 49.6
Operating cost, mills/kwhr
Sulfur credit @ $10/ton (0.06)
Catalysts and chemicals 0.06
Water costs 0.40
Operating labor @ $83,500/job (4 operators) 0.20
Maintenance @ 3% of capital 0.85
Local overhead @ 10070 labor plus 257» maintenance 0.41
Taxes and insurance @ 17» of capital 0.29
Total net Operating cost 2.15
Fuel cost (X = dollar cost per bbl of oil) 1.52X
«a
Yearly average value. Actual capacity is 117= higher to
compensate for 90% stream factor.
Cost data published February 1973. Basis 10,000 bpsd;
updated economic data are presented in Appendix D.
288
-------
REFERENCES
1. Flexicoking Passes Major Test. Oil and Gas J. 53-56, March 10, 1975.
2. Flexicoking. Hydrocarbon Process. September 1974.
3. Hydrodesulfurization-Technology Takes on the Sulfur Challenge.
Oil and Gas J. September 11, 1972.
4. Flexicoking: An Advanced Fluid Coking Process. API Proceedings,
Division of Refining, N.Y., N.Y., 1972.
5. HDS. Hydrocarbon Process. September 1974. x
6. To Make Low-Sulfur Resids. Hydrocarbon Process. May 1973.
7. Low Sulfur Fuel Oil Production - Gulf Hydrodesulfurization Process,
API Proceedings, Division of Refining 1971, San Francisco, California.
8. Commercial Development of HDS Gulf HDS Process. API Proceedings,
Division of Refining 1973, Philadelphia, Pennsylvania.
9. Desulfurize Kuwait Reduced Crude. Hydrocarbon Process. May 1973.
10. RCD ISOMAX Production Route to Today's and Tomorrow's Low Sulfur
Residual Fuels. AIChE Symp Ser. Recent Ado in Air Pollution
Control. 70:38, 1974.
11. Recent Operating Results with RCD Isomax, API Proceedings, Division
of Refining, Philadelphia, PennysIvania, 1973.
12. Clean Fuels Through New Isomax Technology. API Proceedings, Division
of Refining, Philadelphia, Pennsylvania, 1973.
13. Residue Desulfurization. Hydrocarbon Process. September 1974.
14. New BP Process Desulfurizes Resid. Oil and Gas J. October 11, 1971.
15. Resid Hydroprocessing. Hydrocarbon Process. September 1974.
16. New Way to Desulfurize Resids. Hydrocarbon Process. November 1970.
17. H-Oil. Hydrocarbon Process. September 1974.
18. Communication with Cities-Service Research and Development Co.,
N.J. February 1976.
19. Pilot Plant Proves Resid Process. Hydrocarbon Process. , May 1973.
289
-------
20. Go-Fining and Residfining. Hydrocarbon Process. September 1974.
21. Go-Fining Goes Low Pressure. API Proceedings. Division of Refining,
N.Y., N.Y., 1972.
22. Economics of Resid Processing. API Proceedings. Division of Refining,
San Francisco, California, 1971.
23. Dealyed Coking. Hydrocarbon Process. September 1974.
24. Residue Hydrodesulfurization. Hydrocarbon Process. September 1972.
25. Hydrodesulfurization, Trickle Flow. Hydrocarbon Process. September
1974.
26. Demetallization Cuts Desulfurization Costs. Oil and Gas J.
June 30, 1975.
27. Delayed Coking - What You Should Know. Hydrocarbon Process. June 1971.
28. RDS and VRDS Isomax. Hydrocarbon Process. September 1972.
29. Resid Hydroprocessing Options Multiply with New Technology. Oil and
Gas J. May 19, 1975.
30. New Process Gasifies High Sulfur Resid. Electr World. February 1, 1973,
31. The Generation of Clean Gaseous Fuels From Petroleum Residues, Shell
Department Co. Presented at AIChE Meeting. Tulsa, Oklahoma, March 11-
13, 1974.
290
-------
APPENDIX D
ECONOMICS AND PROCESS PARAMETERS OF ALTERNATIVE RESIDUAL
OIL UTILIZATION TECHNOLOGIES
INTRODUCTION
This section summarizes the technical details and economics of the sys-
tems presented in the preceding two appendices. The concluding portion
of this section compares the costs of the CAFB with the costs of feed-
stock desulfurization and flue gas desulfurization.
PROCESS PARAMETERS OF FEEDSTOCK DESULFURIZATION
Table D-l summarizes the feed types, desulfurization efficiencies and
hydrogen and water requirements of the three feedstock desulfurization
techniques capable of handling high metal feedstock. Table D-2 presents
the same data for the other processes described in Appendix C.
ECONOMICS OF FEEDSTOCK DESULFURIZATION PROCESSES
Cost data presented in Appendix C are taken directly from literature
published by system developers. It is difficult to accurately compare
process costs for the following reasons:
• Differences in feedstock
— Source
— Sulfur content
— Metals content
291
-------
Table D-l. PROCESS PARAMETERS OF HIGH METALS FEEDSTOCK DESULFURIZATION TECHNIQUES
N3
Process
Flexicoking
Dene talizat ion/
Desulfurization
L. C. Fining
Feed
type
Iranium
heavy
Bachaquero
W. Texas
Venezuelan
high metals
crude
Venezuelan
medium metals
attn resid
Kuwait
atm resid
Gach
Saran
atm resid
7. S
feed
3.43
3.66
4.6
5.6
2.8
4.05
4.05
2.6
2.6
7. S
product
7. S
removal
0.2 equiv. 94% equiv.
0.2 equiv. 95% equiv.
0.2 equiv. 967. equiv.
1.27
0.64
1.0
0.5
1.0
0.5
77
77
75
88
62
81
Metals
feed,
ppm
525
1040
137
Ni - 85
V - 1100
Ni - 57
V - 398
Ni - 15
V - 49
Ni - 15
V - 49
Ni - 45
V - 165
Ni - 45
V - 165
Metals
product,
ppm
5
10
1
_
.
_
H2
consump-
tion,
scf/bbl
-
-
-
1140
680
910
1030
540
630
Hater usage
20-30 gal/bbl cooling
12-16 gal/bbl boiler feed
20-30 gal/bbl cooling
12-16 gal/bbl boiler feed
20-30 gal/bbl cooling
12-16 gal/bbl boiler feed
•-
-
4590 gal/min A25°F cooling
4860 gal/min A25°F cooling
5410 gal/min A25°F cooling
5830 gal/min A25°F cooling
-------
Table D-2. PROCESS PARAMETERS OF RESIDUAL OIL FEEDSTOCK DESULFURIZATION TECHNIQUES
Process
HDS-Gulf
RCD Isomax
Universal Oil
Products Co.
Residue
Desulfurization
BP
Residue
Hydroprocessing
Standard
Oil Co.
Residue
Ultrafining
Amoco
Feed
type
Kuwait
Kuwait
Direct I
Direct II
Modified
Direct III
Kuwait
Kuwait
Khafyi
W. Texas
Sour
7. S
feed
3.8
3.8
3.8
3.92
U.I
3.9
3.9
3.9
4.0
4.0
4.0
4.02
4.02
4.47
3.85
3.85
7. S
product
1.0
0.3
0.1
1.0
0.3
1.0
0.5
0.32
.1.0
0.5
0.3
1.0
0.5
1.0
1.0
0.3
% S
removal
75
92
97
74
93
74
87
92
75
88
93
75
88
78
74
92
Metals
feed,
ppm
60
60
60
Ni - 15
V - 47
Ni - 15
V - 47
Ni - 15
V - 45
Ni - 15
V - 45
Ni - 15
V - 45
Ni - 13
V - 49,
Ni - 13
V - 49
Ni - 13
V - 49
69
69
Ni - 93
V - 32
Ni - 25
V - 16
Ni - 25
V - 16
Metals
product,
pom
0.2
<0.1
<0.1
_
-
-
-
-
-
-
-
-
-
Ni - 6
V - 18
Ni - 4
V - 13
Ni - 3
V - 12
16
5
-
H2
consump-
tion,
scf/bbl
515
740
900
600
750
550
770
850
625
835
1050
560-620
560-620
580
420
600
Mater usage
214 gal/bbl cooling A20°F
288 gal/bbl cooling A20°F
355 gal/bbl cooling A20°F
-
-
-
-
-
-
.
-
160 gal/bbl cooling 4.2 gal/bbl process
160 gal/bbl cooling 4.2 gal/bbl process
- •
'
-
-------
Table D-2 (continued). PROCESS PHYSICAL PARAMETERS OF RESIDUAL OIL FEEDSTOCK
DESULFURIZATION TECHNIQUES
VO
Process
Go-fining
Exxon
Res id fining
Exxon
Residue
HDS
Badische
Anilin-und
Soda-Fabrik
A.G.
HDS - Trickle
Flow
IFF Residue
and V(X> HDS
Institute
Francais du
Petrole
Feed
type
Arab
Heavy
Athabasca
sands
Gach
Saran
Arab
Heavy
Kuwait
Thermal
cracker gas
Iranian
Light
Atmospheirc
% S
feed
2.96
3.97
2.5
4.19
4.1
1.33
2.5
7. S
product
0.1
0.11
0.3
0.3
0.95
0.16
0.3
% S
removal
97
97
88
93
77
88
88
Metals
feed,
ppm
-
-
220
120
-
-
Metals
product,
ppm
-•
-
-
-
-
-
H2 .
consump-
tion,
scf/bbl
410
975
625
915
650
232
Water usage
3.0-50 gal/bbl cooling
30-50 gal/bbl cooling
150-300 gal/bbl cooling
150-300 gal/bbl cooling
-
260 gal/bbl cooling A30°F
37.7 gal/bbl cooling
-------
• Differences in amortization rates
• Differences in assumed costs of hydrogen
and other materials
— For example, the reported price of hydrogen varies
between 25C and $1.00 per thousand standard cubic
feet of gas
• Differences in plant sizes
• Unpublished assumptions regarding labor costs,
transportation costs, etc.
With these caveats in mind Table D-3 is presented to summarize the econo-
mics of the processes described in Appendix C. The column labeled "GCA
estimated operating costs" represents an attempt to report process costs
on a common basis in 1975 dollars. The figures in this column were cal-
culated based upon estimated operating costs of 75c per thousand SCF
of H2 and 15 percent investment related costs.
It is difficult to put the capital costs in Table D-3 on the same basis.
However, the investment related costs always include process equipment
costs and may or may not include direct labor costs or fee. Therefore,
when comparing costs in Table D-3, process descriptions in Appendix C
should be consulted to insure that capital costs are calculated on the
same basis.
ECONOMIC COMPARISON OF FGD, HDS AND CAFB
Westinghouse has generated both operating and capital costs for a lime-
32
stone scrubbing FGD unit based on a previous comprehensive study of
33
FGD economics prepared by EPA/TVA. Westinghouse has also generated
the HDS operating and capital costs on the same basis as the FGD costs.
The capital costs for FGD and HDS are given in Table D-4 for two dif-
ferent size plants. Foster-Wheeler projects a cost of $23,975,905 for
their 250 Mw oil gasifier including engineering and fee. As can be
seen from Table D-4, FGD presents the smallest capital cost while CAFB
and HDS are roughly equivalent.
295
-------
Table D-3. ECONOMICS FOR OTHER RESIDUAL OIL UTILIZATION TECHNIQUES
Process
HDS-Gulf
RCD-Isomax
Residue Desulf.
Bp process
Res id Hydro-
processing
Standard
Oil Co.
Go -fin ing
Res id -
fining
Residue HDS
• Feed
type
Kuwait
Type II
Type III
Type IV
Kuwait
Direct I
Direct II
Modified
Direct III
Kuwait
Arab Heavy
Athabasca
tar sand
Gach
Saran
Arab
Heavy
Cost
basis,
bpsd
50,000
50,000
50,000
50,000
50,000
40,000
40,000
40,000
50,000
20,000-
40,000
18,000-
95,000
55,000
(Avg.)
Kuwait 45,000
i
% S
feed
3.8
3.8
3.8
3.92
4.1
3.9
3.9
3.9
4.0
4.0
4.0
4.02
2.96
3.97
2.5
4.19
4,1
% S
product
1.0
0.3
0.1
1.0
0.3
1.0
0.5
0.32
1.0
0.5
0.3 .
1.0
0.1
7. S
removal
74
92
97
74
93
74
87
92
75
88
93
75
97
O.li 97
0.3
0.3
0.95
88
93
77
Investment
Total
MM, $
28.1
41.5
21.80
29.32
35.31
28.6-
31.6
Per bbl,
capacity,
$
1.58
2.10
2.43
1.69
2.49
1.63
2.21
2.65
1.36
1.57
-
2.14
2.37
$80-150/bpsd
capacity
$300-750/bpsd
capacity
1.38
Operating cost
Total
MM, $
Per bbl,
capacity,
$
-
-
-
1.30
1.73a
0.71a
0.98
1.10
1.49
1.95
2.47
•
GCA estimated
operating cost.
S/bbl
0.85
1.20
1.58
1.41
1.91
1.15
1.59
1.78
_
-
-
1.08
_
! -
_
,
0.90
References
1, 2, 3, 4, 5
,6, 7, 4, 8, 9
10, 11
12, 13
•
14, 15, 4, 16
15, 4, 14
17
jor difference in operating cost is price of hydrogen 60^/MSCF versus 25f/MSCF.
-------
Table D-3 (continued). ECONOMICS FOR OTHER RESIDUAL OIL UTILIZATION TECHNIQUES
to
VO
Process
HDS-Trickle
flow
IFF Res id
and VGO HDS
Resid ultra-
fining
Amoco
Shell
Gasification
process
Delayed coking
VGO/VRDS Isomax
Flexicoking
Demetalization/
Desulfurization
L.C. Fining
Feed
type
Khafyi
W. Texas
Sour
W. Texas
Sour
Vene-
zuelan
metals
crude
Vene-
zuelan
metals
atm
re sid
Kuwait
atm
re sid
Cost
basis,
bpsd
13,000
(Avg.)
40,000
40,000
40,000
10,000
conversion
of vacuum
Resid at
$2/bbl
No econo-
mic data
20,000
20,000
20,000
25,000
25,000
40,000
Gach
Saran
% S
feed
1.33
2.5
4.47
3.85
3.85
5.0
3.43
3.66
4.6
5.6
2.8
4.05
4.05
2.6
% S
product
0.16
0.3
1.0
1,0
0.3
% S
removal
88
88
78
74
92
0.2 equiv.
94% equiv.
0.2 equiv.
957. equiv.
0.2 equiv.
9651 equiv.
1.27
0.64
1.0
0.5
1.0
2.6 ! 0.5 '
77
77
75
88
62
81
Investment
Total
MM
26.8
19.8
26.8
24.4
17.9
20.4
23.7
23.9
25.2
33.4
34.3
29.8
38.9
Per bbl,
capacity,
$
2.01
2.01
1.49
2.01
2.73
3.09
3.60
2.89
3.05
2.53
2.60
2.26
2.95
Operating cost
Total
MM
Per bbl,
• capacity,
$
0.99
1.16
0.84
1.17
$0.79/MM
Btu of gas
or
$3. 57 /bbl
1.51
2.22
3.23
0.24
0.34
0.48
1.17
0.95
GCA estimated
operating cost
$/bbl
.
1.28
1.60
1.17
1.63
-
-
-
-
-
1.32
1.11
1.21
1.35
1.04
1.28
References
18
4
19, 4
20, 21
22, 23, 24, 25, 9
26, 27, 4, 28
29
30, 4, 31
-------
Table D-4, COMPARATIVE CAPITAL COSTS OF THE CAFB, HDS AND FGD PROCESSES
32
NJ
VO
00
Process equipment in place
Process materials and labor
Total directs
Distribu tables
Subtotal
Indirect costs
Total bare cost
Contingency
Fee
Total process investment
New i.D. fan0
d
Burner costs
Total investment
Start-up costs
Interest during construction
Ideal capital costs
$/kW
CAFB
250 MW,
20% A/F
?23,975,905
96
Grass roots HDS unit
with H2 production
(9000 bbl/d unit)
(supply for 225 MW)
$ 7,761,000a
10,439,000a
. $18,200,000
2,750,000
$20;950,000
2,480,000
$23,430,000
1,820,000
910,000
$26,160,000
200,000 (cat.)
$26,360,000
l,054,000e
2,109,000
$29,523,000
131
Grass roots HDS unit
with H£ production
(45,000 bbl/d unit)
(supply for 1125 MW)
$21,492,000a
28,908,000a
$50,400,000
7,600,000
$58,000,000
6,867,000
$64,867,000
5,040,000
2,520,000
$72,427,000
2,000,000 (cat.)
$72,427,000
2,977,000e
5,954,000
$83,358,000
74
Limestone
scrubbing
200 MW unit
$3,254,000a
4,377,000a
7,631,000
l,153,000b
$8,784,000
l,040,000b
$9,824,066
763,000
382,000
$10,969,000
$10,969,000
878,660
878,000
$12,725,000
64
Limestone
scrubbing
500 MW unit
$ 6,350,000a
8,542,000a
$14,892,000
2,250,000
$17,142,000
2,029,000
$19,171,000
1,489,000
745,000
$21,405,000
$21,405,000
1,712,000
1,712,000
$24,829,000
50
aProportioried from total directs for CAFB 20% A/F 200 MW unit.
bAdjust to CAFB/SWEC %.
cAllowance.
SWEC allowance.
eHalf normal charge due to advanced stage of HDS development.
From reference 34.
-------
Table D-5. COMPARATIVE OPERATING COSTS FOR THE CAFE, HDS AND FGD PROCESSES, $/yra'32
to
VD
VO
Limestone or catalyst
Labor and supervision to operate
Steam
Water
Power
Maintenance
Labor costs
Capital charges
Plant overhead
Labor overhead
Total
H2SO, or sulfur credit
Fuel for process heat
Net
Mills /kWh
c/106 Btu
CAFE
250 MW,
20% A/F
$ 250,000
149,800
Neg.
Neg.
722,833
809,400
27,000
3,572,410
403,716
15,000
$5,950,149
3.40
. 34.0
Grass roots HDS unit
with H2 production
(9000 bbl/d unit)
(supply for 225 MW)
$ 356,400
149,800
38,500
12,800
311,800
728,000
45,600
4,398,900
257,300
15,000
$6,314,100
(280,700)
1,739,600
$7,773,000
4.36
43.6
Grass roots HDS unit
with H2 production
(45,000 bbl/d unit)
(supply for 1125 MW)
$1,782,000
149,800
192,300
64,200
1,559,200
2,016,000
45,600
12,420,300
805,400
15,000
19,049,800
(1,403,300).
8,698,200
$26,344,760
2.96 •
29.6
Limestone
scrubbing
200 MW unit
$ 200,000
210,200
138,000
8,000
315,000
610,500
45,600
1,896,000
265,500
21,000
$3,709,800
2.65
26.5
Limestone
scrubbing
500 MW unit
$ 500,000
210,200
345,000
20,000
787,400
1,191,400
45,600
3,699,500
519,900
21,000
$7,340,000
2.10
21.0
aBasis: 7000 hr/yr, 2.5% S oil, $4/ton stone, 70C/1000 Ib STM, Ic/kWh, 8c/1000 gal. H20, 14.9%/yr capital charges,
$1.85/liter for catalyst, $6/ton of H2S04, Oil and $1.53/MM Btu for reheat, maintenance, at 8%/yr limes, 4%/yr CAFB,
CAT-Ox and HDS as % of total direct investment, labor at $8/man-hour, plant overhead at 20% of 0 & M costs, labor
overhead at 10% of direct labor costs.
-------
Operating costs for a 250 MW CAFB unit can be extrapolated from Westing-
house's data. The operating costs for HDS, FGD and CAFB are given in
Table D-5. The predicted operating costs for FGD are less than either
the CAFB and HDS.
It would appear from these most up-to-date predictions that the CAFB
process is not cost competitive with the limestone scrubbing FGD pro-
cess. However, it should be noted that the costs presented in Tables D-4
and D-5 are projections and should be viewed skeptically. Indeed, pro-
34
jections from an earlier Westinghouse report presented in Table D-6
show that FGD should be twice as expensive as the CAFB process.
There are, at present, no actual cost figures available for FGD using
limestone scrubbers on oil-fired boilers. The only economic data avail-
able for oil-fired boilers using FGD is for the Boston Edison plant with
a magnesia scrubber. The results show that in order for this system to
be economically competitive, a $3/bbl difference must exist between the
32
cost of high sulfur and low sulfur fuel oil. Since low sulfur oil can
be prepared from high sulfur feedstock by HDS for under $3/bbl, it appears
that FGD using a magnesia scrubber is a costly way of meeting air pollu-
tion standards for S02 emissions.
This may also be true for FGD using limestone scrubbing when actual cost
figures for FGD on oil-fired boilers or for the CAFB are evaluated.
Table D-6. 1972 PROJECTED COSTS FOR THE CAFB AND FGD PROCESSES,
c/106 Btu, 370,000 Ib steam/hr, new installation34
Low sulfur oil
CAFB
Conventional with
wet scrubbing
Capital
charges
4.58
13.92
28.02
Desul-
furiza-
tion cost
26.0
Labor
5.2
11.7
11.7
Sorbent
-
2.6
4.5
Power
0.7
1.3
1.3
Solids
disposal
-
1.6
17.5
Total
36.48
31.12
63.02
300
-------
REFERENCES
1. Commercial Development of HDS Gulf HDS Process. API Proceedings,
Division of Refining 1973, Philadelphia, Pennsylvania.
2. HDS. Hydrocarbon Process. September 1974.
3. To Make Low-Sulfur Resids. Hydrocarbon Process. May 1973.
4. Hydrodesulfurization-Technology Takes on the Sulfur Challenge.
Oil and Gas J. September 11, 1972.
5. Low Sulfur Fuel Oil Production - Gulf Hydrodesulfurization Process,
API Proceedings, Division of Refining 1971, San Francisco, California.
6. Desulfurize Kuwait Reduced Crude. Hydrocarbon Process. May 1973.
7. RCD ISOMAX Production Route to Today's and Tomorrow's Low Sulfur
Residual Fuels. AIChE Symp Ser. Recent Ado in Air Pollution
Control. 70_:38, 1974.
8. Recent Operating Results with RCD Isomax, API Proceedings, Division
of Refining, Philadelphia, Pennsylvania, 1973.
9. Clean Fuels Through New Isomax Technology. API Proceedings, Division
of Refining, Philadelphia, Pennsylvania, 1973.
10. Residue Desulfurization. Hydrocarbon Process. September 1974.
11. New BP Process Desulfurizes Resid. Oil and Gas J. October 11, 1971.
12. Resid Hydroprocessing. Hydrocarbon Process. September 1974.
13. New Way to Desulfurize Resids. Hydrocarbon Process. November 1970.
14. Economics of Resid Processing. API Proceedings. Division of Refining,
San Francisco, California, 1971.
15. Go-Fining and Residfining. Hydrocarbon Process. September 1974.
16. Go-Fining Goes Low Pressure. API Proceedings. Division of Refining,
N.Y., N.Y., 1972.
17. Residue Hydrodesulfurization. Hydrocarbon Process. September 1972.
18. Hydrodesulfurization, Trickle Flow. Hydrocarbon Process. September
1974.
19. Pilot Plant Proves Resid Process. Hydrocarbon Process. May 1973.
301
-------
2.0.'; New Process Gasifies High Sulfur Resid. Electr World.. February 1, 1973.
''-,.'••' '
21. The Generation of Clean Gaseous Fuels From Petroleum Residues, Shell
Department Co. Presented at AIChE Meeting. Tulsa, Oklahoma, March 11-
13, 1974. ' .. .' '
• .\ •
22. Delayed Coking..- Hydrocarbon Process. September 1974.
23. Delayed Coking - What You Should Know. Hydrocarbon Process. June 1971.
24. RDS and VRDS-.Isomax. Hydrocarbon Process. September 1972.
25. Resid Hydroprocessing Options Multiply with New Technology. Oil and
Gas J,; May.19, 1975.
26.- Flexicoking Passes Major Test. Oil and Gas J. 53-56, March 10, 1975.
.27.' Flexicoking.' Hydrocarbon Process. September 1974.
* • ' * '
28. Flexicoking: An Advanced Fluid Coking Process. API Proceedings,
'..;; Division of Refining, N.Y. , N.Y. , 1972.
29. Demetallization Cuts Desulfurization Costs. Oil and Gas J.
June 30, 1975.
30. H-Oil. Hydrocarbon Process. September 1974.
•31. Communication with Cities-Service Research and Development Co.,
N.J. February 19/6.
32. Yan, C.J. Evaluating Environmental Impacts of Stack Gas Desulfuriza-
tion Processes. Environ Sci Technol. 10_:54-58, January 1976.
302
-------
TECHNICAL REPORT DATA
(Please read InUructions on the reverse before completing)
. REPORT NO. 2.
EPA-600/7-76-017
. TITLE ANDSUBTITLE
PRELIMINARY ENVIRONMENTAL ASSESSMENT OF
THE CAFB
3. RECIPIENT'S ACCESSION-NO.
5. REPORT DATE
October 1976
6. PERFORMING ORGANIZATION CODE
A s Werner, C.W. Young, M.I. Bornstein,
R.M. Bradway, M.T. Mills, and D. F. Durocher
8. PERFORMING ORGANIZATION REPORT NO.
GCA-TR-76-18-G
. PERFORMING ORGANIZATION NAME AND ADDRESS
GCA Corporation
GCA/Technology Division
Bedford, Massachusetts 01730
10. PROGRAM ELEMENT NO.
EHB537
11. CONTRACT/GRANT NO.
68-02-1316, Task 14
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Task Final; 8/75-5/76
14. SPONSORING AGENCY CODE
EPA-ORD
s.SUPPLEMENTARY NOTES jERL_RTP Task officer for this report is S. L. Rakes, Mail Drop
61, 919/549-8411 Ext 2825.
16. ABSTRACT
repOrt; gives results of a. preliminary environmental assessment of the
Chemically Active Fluid Bed (CAFB) process. All waste streams contributing air,
water, and solid waste pollutants were evaluated in terms of emission rates and
potential environmental effects. As part of the investigation, a field sampling and
laboratory analysis program was carried out to compile an emissions inventory of
the CAFB pilot plant at the Esso Research Centre, Abingdon, England. In addition
to the environmental assessment, the report presents an economic evaluation of the
CAFB relative to alternative residual oil utilization techniques. Finally, it
recommends that further control research and development be carried out at the
CAFB demonstration plant in San Benito, Texas.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
Pollution
Assessments
Combustion
Fluidized Bed Processing
Pollution Control
Stationary Sources
Environmental Assess-
ment
Chemically Active Fluid
Bed Process
13B
14B
2 IB
13H,07A
18. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
324
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
303
------- |