PB83-207639
NOY ABATEMENT FOR STATIONARY SOURCES IN JAPAN
Faculty of Science and Engineering
Tokyo, Japan
May 83
U.S. DEPARTMENT OF COMMERCE
National Technical Information Service
NTIS
-------
PB83-207639
EPA-600/7-83-027
May 1983
NOZ ABATEMENT FOR STATIONARY SOURCES IN JAPAN
by
Jnmpei Ando
Faculty of Science and Engineering
Chuo University
Tokyo, Japan
Contract No. 68-02-3676
Project Officer
J. David Mobley
Emissions/Effluent Technology Branch
Industrial Environmental Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park. North Carolina 27711
Prepared for:
U.S. Environmental Protection Agency
Office of Research and Development
Washington. D.C. 20460
-------
TECHNICAL REPORT DATA
(Pit ate rcad /NUflu Ilo?IN on tile P11 epte b< fort’ c o ’nph SIng)
1 REPORT NO 2
EPA-600/7-83-027 I
4 TITLE AND SU6TITLE
NOx Abatement for Stationary Sources in Japan
3 RECIPIENT S ACCESSION NO
PB8 3 2 0 7 6 3 9
5 REPORT DATE
j y 1983 —-_________
6 PERFORMING ORGANIZATION CODE
7 AUTHORtS)
Jumpei Ando
8 PERFORMING ORGANIZATION REPORT NO
PERFQRMING OROANIZATION NAME AND ADDRESS
Faculty of Science and Engineering
Chuo University
Tokyo, Japan
1OPR( ’GRAM ELEMENT NO
11 CONTRACT/GRANT NO
68-02-3676
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
Final; 1/81 — 1/83
14 SPONSORING AGENCY CODE
EPA/600/13
15 SUPPLEMENTARY NOTES IERL—RTP project officer is J. David Mobley, Mail Drop
61, 919/541—2578. EPA—600/7—81—030 is an earlier related report. EPA-
6OO/7- 3-O28 Is the comparable report for S02 abatement in Japan. -
16 ABSTRA T The report is a compilation of information on the current statu
of NOx abatement technologies for stationary sources in Japan, where
strict ambient air quality standards for S02 and NOx mandate the use of
various air pollution control technologies. The author obtained this
information from electric power companies, various industries, and
developers of numerous technology processes, as well as from his own
original research in the field. The report focuses on the Combustion
Modification (CM) and Selective talytic Reduction (SCR) NOx abatement
technologies. Information is provided on the development status, pilot
and demonstration plant tests, technological problems, and costs asso-
ciated with the use of these technolgies in Japan. Detailed operating
data are given to describe the commercial operation of SCR plants.
17 KEY WORDS AND DOCUMENT ANALYSIS —
DESCRIPTORS h IDENTIFIERS/OPEN ENOED TERMS COSATI I I d L,toup
Pollution Pollution Control 13B
Nitrogen Oxides Stationary Sources 07B
Combustion Control Japan 21E
Catalysis Combustion Modifica— 07D
Reduction (Chemistry) tion 07C
Selective Catalytic
Reduction
i ,i5TR BuT ON STATEMENT
Release to Public
19 SECURITY CLASS rhis Report)
Unclassified
21 NO OF PAGES
, --t-
22 PRICE —
iI 11 EtASS frh,s page)
Unclassified
EPA Form 2220 1 (9 73)
•1
-------
NOTICE
This document has been reviewed in accordance with
U.S. Environmental Protection Agency policy and
approved for publication. Mention of trade names
or commercial products does not constitute endorse-
ment or recommendation for use.
ii
-------
ABSTRAc’r
Strict ambient air quality standrds for SO 2 and NO 1 in Japan mandate
the use of various air pollution control technologies. This report is a com-
pilation of information on the current status of NO 1 abatement technologies
for stationary sources in Japan. The author obtained this information from
electric power companies, various industries, and developers of numerous tech-
nology processes, as well as from his own original research in the field. The
report focuses on the combustion modification (CM) and Selective Catalytic
Reduction (S R) NO 1 abatement technologies. Information is provided on the
development status, pilot and demonstration plant tests, technological pro-
blems, and costs associated with the use of these technologies in Japan. De-
tailed operation data is given to describe the commercial operation of S R
plants.
AcIOWLEG EME NT
The author wishes to extend his deep thanks to many Japanese electric
power companies, industry representatives, and developers of S1 R processes,
catalysts, and low—NO 1 burners who provided useful information and reviewed
parts of this report. The author is particularly grateful to unihiko
Mouri of the Electric Power Development Company and Katsuya Nagata of
Waseda University who assisted in the preparation of Section 2. The author
also wishes to acknowledge the assistance of Nancy Gates, Randy Parmley, and
Gary Jones of Radian Corporation, Austin, Texas, in editing this report. In
addition, Radian Corporation’s word processing and graphics support is grate-
fully acknowledged. (Radian Corporation’s assistance was provided by
EPA under Contract No. 68—02—3171, Task No. 52.)
- iii
-------
TABLE OF CONTENTS
Page
.Abètract ..... iii
Acknowledgement .
List of Tables x i
List of Figures x vii
Summary ‘ xxx
SECTION 1 NO 1 EMISSIONS AND CONTROL 1
1.1 1ERGY AND NO 1 EMISSIONS 1
1.2 AMBIENT N02 STANDARDS 8
1.3 NO 1 EMISS ION STANDARDS FOR STATIONARY SOURCES 9
1.4 NO 1 CONTROL FOR MOBILE SOURCES 14
1.5 METHODS OF NO 1 ABATEMENT FOR STATIONARY SOURCES 16
1.6 LOCAL REGULATIONS AND NO 1 ABATEMENT 21
1.6.1 Introduction 21
1.6.2 Air Quality Monitoring Systems in the
Kanagawa Prefecture 23
1.6.3 NO 1 Emissions and Control in Kanagawa
Prefecture 23
1 .7 AIR POLLUTION PROBWIS RELATED TO NO 1 27
REP ENCES 32
SECTION 2 NO 1 ABATEMENT BY COMBUSTION MODIFICATION ( 1) 33
2.1 D’4TRODUCTION 33
2.1.1 Classification of Combustion Modification
(CM)Techniques 33
2.1.2 NO 1 from Pulverized Coal Burning 36
2.2 NO 1 ABATEMENT BY IEI—FW DF—QJ COAL BURNER 37
2.2.1 IHI-FW DF—C Burner 37
2.2.2 Test Furnace and Coal Composition 39
2.2.3 Test Results 39
2.3 NO 1 ABATEMENT BY BRK LOW—NO 1 COAL BURNER 46
2.3.1 Primary Gas Dual Air Register Burner 46
2.3.2 Test Results 51
2.4 MEl LOW—N() COAL. BURNERS 57
2.4.1 Separate Gas Recirculation Burner 57
2.4.2 PM Burner for Pulverized Coal 59
2.5 KEl LOW-NO1 COAL BURNER 70
2.5.1 Structure and Combustion Model 70
2.6 EPDC’S NO 1 ABATEMENT BY COMBUSTION MODIFICATION 75
2.6.1 Introduction 75
2.6.2 Tests of NO 1 Abatement by Combustion
Modification 78
2.6.3 Combustion Modification at Isogo Power
Station 81
2.6.4 Matsushima Power Station 85
Preceding page bJank
-------
TABLE OF CONTENTS (Continued) Page
2.6.5 Takehara Power Station 88
2.7 LOW-NOT BURNERS FOR GAS- AND OIL—FIRING 90
2.7.1 Introduction 90
2.7.2 Uigh—MixingBurner .. 90
2.7.3 Divided—Flame Burner 93
2.7.4 Self—Recirculating Burner 93
2.7.5 Staged—CombustionBurner 103
2.7.6 Off—Stoichiometric—Combustion Burner 120
2.7.7 Water—InjectionBurner 120
2.7.8 Combination Burner 128
2.8 COMBUSTION MODIFICATION FOR SPECIAL. NO 1 SOURCES
AND FUELS 138
2.8.1 Refuse Incinerators 138
2.8.2 Cement ilns . 142
2.8.3 Glass Melting Furnace 145
2.8.4 Use of Emulsified Oil for Small Boilers 147
2.8.5 Special Fuels and Processes 149
REFERENCES 154
SECTION 3 SELECTIVE CATALYTIC REDUCTION (SCR) OF NO 1 . 157
3.1 REDUCTION OF NO 1 BY AMMONIA 157
3.1.1 BasicReactions 157
3.1.2 Problems with SCR 158
3.1.3 Major Factors in Catalytic Reactions 159
3 .2 COMPOSITION AND PROPERTIES OF SCR CATALYSTS 163
3.2.1 History 163
3.2.2 Alumina—Based Catalysts 165
3.2.3 Fe03— Based 167
3.2.4 Ti02—BasedCatalyst 171
3 .3 LOW TEMPERATURE CATALYSTS 171
3.3.1 Introduction 171
3.3.2 Catalytic Activity Decrease Caused by
Ammonium Bisulfate Deposit 176
3.3.3 Use of Heating to Recover Catalytic Activity 179
3.3.4 Alkali Poisoning and Poisoning Countermeasures 183
3.3.5 Suary 187
3 .4 CATALYST SHAPE AND REACrORY TYPE 188
3.4.1 Introduction 188
3.4.2 Moving—Bed Reactor 189
3.4.3 Parallel Flow Reactors 192
3.4.4 Honeycomb and Plate Catalysts 198
3.4.5 Comparison of Catalysts and Reactors 202
vi
-------
TABLE OF CONTENTS (continued) Page
3 .5 PROBLEMS WITh SCR FOR COAL—FIRED BOILERS 206
3.5.1 Introduction 206
3.5.2 Fly Ash Erosion of the Catalyst 209
3.5.3 Dust Adhesion and Plugging 211
3.5.4 NO 1 Removal Efficiency for Flue Gas From
Coal 214
3.6 AMMONIUM BISULFATE DEPOSITION IN TEE AIR PREREATER 216
3.6.1 Formation and Behavior of Ammonium Bisulfate 216
3.6.2 Formation of SOi 220
3.6.3 Laboratory Tests on Bisulfate Deposition 220
3.6.4 Air Preheaters (Heat Exchangers) 227
3.6.5 Laboratory Corrosion Tests 235
3.6.6 Ammonium Bisulfate/Iron and Fly Ash Reaction
Products 240
3.6.7 Compounds in Air Preheater and Heat Exchanger
Deposits 240
3.6.8 Deposit Formation Reduction 242
3 .7 OThER PROBLEMS WIlE AMMONIA 243
3.7.1 Introduction 243
3.7.2 Dust Removal Improvement 245
3.7.3 Ammonia Bisulfate Deposition on S R Catalysts. 245
3.7.4 Contamination of Fly Ash by Ammonia 246
3.7.5 PlumeFormation 248
3.7.6 Effect of A onia on Flue Gas Desulfurization 249
3.7.7 Measurement of Ammonia in Flue Gas 251
REFERENCES 255
SECTION 4 SCR FOR FLUE GAS FROM UTILITY BOILERS 257
4.1 DESIGN AND PERFORMANCE OF SCR PLANTS 257
4.1.1 Development of S R for Utility Boilers 257
4.1.2 DesignofS Plants 262
4.1.3 SCRPerformance 266
4.2 SCR UNITS OF CHIJBU ELECTEIC POWER COMPANY 271
4.2.1 Introduction 271
4.2.2 ChitaPowerStation 271
4.2.3 S R Units at the Chita Station 275
4.2.4 S R Plants at Other Stations 277
4.3 SCR SYSTEM AT UDAMATSU STATION, CH1JGOKU ELECIRIC 280
4.3.1 Introduction 280
4.3.2 Sf .R Units 28].
4.3.3 SCR System Performance 286
4.3.4 Ammonium Bisulfate Deposition in the Air
Preheater 287
4.3.5 Economics 289
4.3.6 Evaluation 289
vii
-------
TABLE OF CONTENTS (continued) Page
4 .4 SCR SYSTEM AT THE SHIMONOSEKI PLANT, CHUGOKU ELECTRIC 291
4.4.1 Introduction 291
4.4.2 SCR System 292
4.4.3 Problems Related to Unreacted NRa 297
4.4.4 Evaluation 299
4.5 SQ SYSTEM AT TOMATO-ATSUMA STATION, HOKXAIDO ELECTRIC.... 300
4.5.1 Introduction 300
4.5.2 SQ System 302
4.5.3 Future Construction 306
4.5.4 Evaluation 306
4 .6 K.ANSAI ELECTRIC’ S S CR PLANTS 307
4.6.1 Large Scale Testing of S R 307
4.6.2 Commercial S R Plants 307
4.6.3 Evaluation 312
4 .7 SCR TESTS AT THE NA .KOSO STATION 312
4.7.1 Test Plant 312
4.7.2 Test Results 315
4.7.3 Evaluation 320
4.8 EPDC’S 5cR ACTIVITIES 321
4.8.1 Introduction 321
4.8.2 Pilot Plant Tests with I 321
4.8.3 DomonstrationPiant. 328
4.9 S R COSTS 332
4.9.1 Investment Costs 332
4.9.2 Estimated Investment and Annualized Costs 334
4.9.3 Power Generation and Flue Gas Treatment Costs 341
4.10 USE OF JAPANESE SQ TECHNOLOGY IN THE UNITED STATES 342
4.1.0.1 Introduction 342
4.10.2 Hitachi Zosen Technology Pilot Plant Tests a
the Mitchell Station 342
4.10.3 Evaluation 348
4.10.4 Costs of SQ for Coal—Fired Boiler Applications
in the U.S 355
R.EFER. 1cES 360
SECTION 5 SQ FOR INDUSTRIAL GAS SOURCES 361
5.1 GENERAL DESCRIPTIONS 361
5.1.1 SCR Units for Industrial Gas Sources 361
5.1.2 Economics 366
5.2 SODEGAURA REFINERY, FUH OIL 368
5.2.1 Introduction 368
5.2.2 SQ Unit for CO Boiler 368
5.2.3 SQ Unit for Oil—Fired Boiler 372
5.2.4 Evaluation 374
viii
-------
TABLE OF CONTENTS (continued) Pa&e
5 .3 KAWASAXI PLANT, NIPPON YAKIN 374
5.3.1 Introduction 374
5.3.2 Process Description 375
5.3.3 Cost
5.3.4 Evaluation 378
5.4 SCR FOR COKE OVEN FLUE GAS (NIPPON STEEL) 379
5.4.1 Introduction 379
5.4.2 Process and Operation 379
5.4.3 Evaluation 383
5.5 SCE USING AN IRON ORE CATALYST 383
5.5.1 Introduction 383
5.5.2 Process Description 384
5.5.3 Evaluation 386
5 .6 LOW TEMPERATURE CATALYST TESTS (KOBE STEEL) 386
5.6.1 Introduction 386
5.6.2 Catalysts and Plant Operation 387
5.6.3 Evaluation 389
5.7 ScR FOR INCINERATOR FLUE GAS 391
3.7.1 Introduction 391
5.7.2 Process Description 392
5,7.3 Econoinics..................... .. ..... 392
5.7.4 Evaluation . 394
5.8 0’l ER SCR UNITS 394
5.8.1 S R Unit at Ukishima Chemical, Petrochemical’s
Chiba Plant 394
5.8.2 Shindaikyowa Petrochemical’s Yokkaichi Plant 396
5.8.3 Kawasaki Steel’s Chiba Plant 396
5.8.4 S R for Diesel Engine Flue Gas 397
5.8.5 SCR for Gas Turbine Flue Gas 397
R.EFER ’i CEs 3 99
SECTION 6 OTBER PROCESSES FOR NO REMOVAL 401
6.1 CLASSIFICATION OF NOx REMOVAL PROCESSES 402
6.1.1 Dry Processes 403
6.1.2 Wet Processes . 404
6.1.3 Gas Composition Appropriate for Simultaneous
Removal Processes 405
6 .2 SELECTIVE NONCATAL TIC REDUCTION (SNR) 405
6.2.1 Introduction 405
6.2.2 Mill’s SNR Laboratory Studies 406
6.2.3 SNR Tests at MCI’s Boilers 409
6.2.4 Full—Scale SNR Test Unit at Chuba Electric’s
Chita Station 416
6.2.5 Exxon Thermal DeNOx Systems 421
6.2.6 Toho Gas Company’s SNR Units 424
6.2.7 Evaluation 424
ix
-------
TABLE OF CONTENTS (continued) Page
6.3 SNRJSCR COMBINED SYSTEMS 426
6.3.1 Introduction 426
6.3.2 Tokyo Electric’s SNRIScR Demonstration Plants 427
6.3.3 Kansaj Electric’s Full—Scale SNR Test 432
6.3.4 Evaluation 433
6 .4 MACI IN—FURNACE NOx REMOVAL PROCESS 433
6.4.1 Introduction . 433
6.4.2 Small—Scale Tests 435
6.4.3 Fundamental MACI Studies 437
6.4.4 Pilot Plant Tests with Corner—Firing Furnaces 443
6.4.5 Evaluation 444
6.5 ACrIVATED CARBON PROCESS 448
6.5.1 Introduction 448
6.5.2 Pilot Plant Test of the Activated Carbon
Process 449
6.5.3 Evaluation 453
6 .6 EBARA ELEC’IRON BEAM IRRADIATION PROCESS 455
6.6.1 Introduction 455
6.6.2 Ebara Process Pilot Plant Tests 455
6.6.3 Reaction Mechanism 463
6.6.4 Ebara Process Commercial Plant Assumptions 465
6.6.5 Evaluation 468
6.7 M POTASSIUWEDTA WET SIMULTANEOUS S0 1 /N0 1 REMOVAL
PROCESS . 470
6.7.1 Introduction 470
6.7.2 Process Description 470
6.7.3 RequirementsandCosts 473
6.7.4 Evaluation 476
6.8 NO OXIDATION CATALYST 478
6.8.1 Introduction 478
6.8.2 Catalyst and Reactor 478
6.8.3 Evaluation 479
6.9 OThER PROCESSES AND PLANTS 479
6.9.1 Copper Oxide Process for Simultaneous N0 1 /S0 1
Removal (Shell Process) 479
6.9.2 Activated Carbon Simultaneous Removal (Unitika
Process) 481
6.9.3 Molecular Sieve Process for NO Adsorption 481
6.9.4 Sumitomo—Fujikasui Wet Simultaneous Removal
Process 482
6.9.5 In—Furnace NO Removal (Three—Stage Combustion—
ilitachi—Zosen) 482
REFEREMCES 484
x
-------
TABLES
Number Pa&e
S—i Examples of Controlled and Uncontrolled NO 1 Concentrations
in Utility Boiler Flue Gas ‘X x vi
S—2 Examples of NO Regulations and Emissions from Utility
Boilers
S—3 CM Investment Cost xxxix
S—4 S R Costs for 700 MW New Boiler xli
SECTION 1
1—1 Japan’s Energy Supply: 1980—2000 3
1—2 State of Compliance with the New Nitrogen Dioxide Environ-
mental Quality Standard in 1978 12
1—3 Expansion of NO, Control for Stationary Sources 12
1—4 NO, Emission Standards for Major Stationary Sources (ppm) 13
1—5 Automobile Emission Standards in Japan 15
1—6 Automobile Emission Standards in the United States 15
1—7 Emission Standards for Diesel Engine Heavy Duty Vehicles (ppm). 16
1—8 NO 1 Control for Diesel Engine Passenger Cars (glkm) 16
1—9 Classification of NO Abatement Methods 18
I
1—10 Total Cost for anagawa Continuous Air Monitoring Center
(1970—1980) 24
1—11 NO Emissions and Regulations for Existing Utility Boilers
in 1 Yokohama City 25
1—12 NO Removal Plants in Kanagawa Prefecture 26
I
1—13 Ambient Air Quality Standards 27
1—14 Ni ber of Days in which Photochemical Smog Warning were
Issued 30
1—15 Gaseous Hydrocarbon Emissions from Stationary Sources 31
xi
-------
TABLES (continued)
Number Page
SECTION 2
2—1 Number of Boilers and Combustion Modifications Applied for
Oil and Gas as of April 1981 (Tokyo Electric) 35
2—2 NOx Abatement by Conbustion Modification ( 4) for Utility
Boilers
2—3 Examples of Combustion Modification for Coal—fired Utility
Boilers 36
2—4 Compositions of Coals Tested 41
2—5 Combustion Furnace Test Conditions for Factory Test 56
2—6 Main Features of Mlii Test Furnace (4 t/h Test Furnace) 61
2—7 Analysis of Coals Used with MRI Test Burner 62
2—8 EPDCsPowerStationsaudNO Regulations ... 78
x
2—9 Measurement Results of P.O.M. at Furnaces. .. 81
2—10 Specifications for Isogo 265 MW Pulverized Coal—fired Boiler
(No. 2) 82
2—11 Typical Analysis of Coal Used at Isogo Station 83
2—12 Modifications Made for Increased NO Control at the isogo
Power Unit . . 83
2—13 Specifications of Boilers at Matsushima Power Station 85
2—14 Specification of No. 3 Boiler of Takehara Station 88
2—15 NO Emission Levels Guaranteed with the MR Burner for Small
Sche Boilers 112
SECTION 3
3—1 Comparison of Catalyst’s Carrier Characteristics 164
3—2 Chemical Composition of Red Mud and Linomite Catalysts 167
3—3 Performance of Honeycomb Catalysts 175
3—4 Properties and Compositions of Catalysts (Fresh and Used) 181
xii
-------
TABLES (continued)
Number Page
3—5 Properties and Compositions of Catalyst (Fresh and Used, and
Heated) 181
3—6 Changes in TiOs—V205 Catalyst Caused by Alkali Addition and
By Alkali Addition and By Washing 187
3—7 S R Catalysts and Reactors Suitable for Various Gases 189
3—8 SIR Plants Using a Moving Bed Reactor 193
3—9 AV and SV of Square Honeycomb Catalyst 196
3—10 Specific Surface Area of Catalysts 197
3—11 Honeycomb and Plate Catalysts 198
3—12 Comparison of Catalysts and Reactors 203
3—13 Typical Examples of SCR For Boiler Flue Gases 204
3—14 Coimnercial ScR Plants for Coal—Fired Boilers in Japan 207
3—15 Comparison of High—Dust and Low—Dust Systems for Coal 209
3—16 Types of Air Preheaters (Heat Exchangers) 227
3—17 Chemical Composition of Steels 236
3—18 Effect of Additives on Corrosion of Low—Alloy Steel by
Ammoniun Bisulfate at 200°C for 75 Hours 236
3—19 Deposits and Gas Compositions 242
3—20 Compounds in Deposits in Heat Exchangers and Air Preheater 244
3—21 Two Major Methods Used in Japan for Continuous Analysis of
Nils in Flue Gas 252
SECTION 4
4—1 S R Plants for Utility Boilers in Japan and the United States.. 259
4—2 Chuba Electric’s S R Plants 272
4—3 Chita Station Boiler and NO Abatement Data 273
I
4—4 Emission Limits at the Chita Power Station 274
4—5 Data Summary on Babcock Hitachi SCR System 275
x i i i
-------
TABLES (continued)
Number Page
4—6 Design Data for the MEl ScR System Used with the No. 4 Boiler
at the Chita Power Station 278.
4—7 Boilers at the udamatsu Station 280
4—8 SCR Plant Specifications (Kudamatsu Station. Chugoku
Electric) 282
4—9 Capital Cost Breakdown of SCR System on No. 3 Boiler (700 MW)
Chugoku Electric, Kudamatsu Power Station 290
4—10 Operating Cost Breakdown of SCR System on No. 3 Boiler (700
‘IN) Chugoku Electric. udamatsu Power Station 290
4—11 Regulations for Shimonseki Station 291
4—12 Design Oata for the MEl S R System Installed at the
Shimonoseki Power Station 294
4—13 Design Conditions for S R Unit. Tomato—Atsa Station,
Kokkaido Electric Power Co 302
4—14 Primary Plant Equipment Specifications 303
4—15 Data, Tomato—Atsuma S R Unit, Hokkaido Electric Power Company.. 305
4—16 Flue Gas Composition and Fly Ash Properties SQ Pilot Plant,
Takehara Station 321
4—17 Fly Ash Composition 323
4—18 Description of S R Demonstration Plant, Takehara Station No.
1 Unit 330
4—19 Comparison of Existing and New Fans at Takehara Power Station.. 332
4—20 Investment Costs of S R Plants for Utility Boilers 333
4—21 Calculated Costs for and S R Plant for a New 700 MW Coal—
Fired Boiler 335
4—22 Calculated Costs for SCR for New 700 MW Oil— and Gas—fired
Boilers 336
4—23 Power Generation Costs fo a Coal—fired Boiler in Japan (1981).. 341
4—24 Principal Vendor Contacts for Japanese NO ScR. Technology 343
xiv
-------
TABLES (continued)
Nnmbe Page
4—25 Composition of Fly Ash 354
4—26 Estimated Capital Investment for a 500 M Y Application of the
Hitachi Zosen Process in the U.S 357
4—27 Estinated Average Annual Revenue Requirements for a 500 PI
Application of the Hitachi Zosen Process in the U.S 358
SECrI0N 5
5—1 ScR Plants for Oil, Petrochemical. and Gas Companies 362
5—2 SCR Plants for Steel and Metal Industries 363
5—3 SCR Plants for Other Industries 364
3—4 Investment Costs of Seven Industrial SCR Units 367
5—5 Operating Costs of SIR Units in 1979 369
5—6 Boilers and NO Abatement at Sodegaura Refinery 370
5—7 Permissable Emissions of NO and SO from the Sodeganra
x x
Refinery 370
5—8 Volume and Composition of Flue Gas to be Treated by S R at the
Sodegaura Refinery 370
5—9 S R Units at the Sodegaura Refinery 370
5—10 Nippon Steel S System Test Conditions 382
5—11 Mitsui Cost Calculation (Mitsui Engineering) 395
SECTION 6
6—1 NO Removal Process Other than SCR 401
x
6—2 Oxidation Reduction 404
6—3 Complex Absorption Process 404
6—4 The MEl Mechanism of the De—NO Reaction and Rate
of Reactions 408
6—5 Basic Formulas for the MEl Process De—NO Reaction and Rate
- , I
of Reactions 408
xv
-------
TABLES (continued)
Number Page
6—6 Boilers Used for the MHI Application Tests 410
6—7 SNR System Specifications. Unit No. 2, Chita Station, Chubu
Electric Power Co 416
6—8 Exxon/Tonen Eechnology Thermal DeNO Plants 422
6—9 Thermal DeNO Cost Figures Calculated by Tonen Technology 423
6—10 Tohu Gas Company’s SNR Units 425
6—11 Tokyo Electric’s SNR/SCR Demonstration Plants 428
6—12 Small Test Furnace Specifications 435
6—13 Composition and ileat Value of Coals 435
6—14 Furnace Specifications 443
6—15 M&CTPilotPlantTestConditions . .443
6—16 Flue Gas Composition During Pilot Plant Test 449
6—17 Test Conditions 449
6—18 Balance of NO , NH3, and SO (Percent of Inlet Concentration).. 461
x x
6—19 Estimated Utility Require nents per Ro Estimated for Ebara
Process System Applied to a Utility Boiler 467
6—20 Cost Comparison of Flue Gas Trea ent Process 467
6—21 Utility and Material Requirements and Costs of an MKX Process
System for a 1000 MWhr Boiler (1980) 474
6—22 Annual Cost of an MKK Process System for a 1000 MWhr Boiler.... 476
xv i
-------
FIGURES
Number Page
S-i SCR and FGD System Arrangements in Use in Japan xxxiii
SECTION 1
1—1 Primary Energy Suupply in Japan: 1960—1980 2
1—2 Annual NO 1 Emissions from Stationary Sources in Japan 5
1—3 Estimated increase in Electrical Generation Capacity 6
1—4 Preliminary Plan for Development of Coal Gasification and
Li.quefaction by the Sunshine Project 7
1—5 Changes in Annual Average Concentration of NOs (Air Pollution
Monitoring Stations) 10
1—6 Changes in Annual Average Concentration of Nitrogen Dioxide
(Automobile pollution monitoring stations) ii.
1—7 Estimated Total Amount of NO 1 Emissions from Road Traffic
in the Tokyo Bay Area 17
1—8 Number and Total Capacity of NO 1 Selective Catalytic Reduc-
tion (S R) Plants 20
1—9 Kanagawa Prefecture and Its Major Cities 22
1—10 Annual Average Ambient SOs Concentrations in 15 Major Cities
and Industrial Regions; and Number of Designated Air Pollu—
tionPatients 28
SECTION 2
2—1 DF—Ci Burner 38
2—2 Cross—section View of Test Furnace 40
2—3 Secondary Air Vane Opening vs. NO 1 42
2—4 Inner Vane Opening vs NO 1 43
2—5 Opening of Tertiary Air Damper vs NO 1 44
2—6 Effect of Fuel Injection Velocity on NO 1 45
2—7 NO 1 Reduction by Low—NO 1 Burner and Two—Staged Combustion 47
2—8 Nitrogen Content in Fuel vs. NO 1 48
2—9 NO 1 Reduction by Flue Gas Recirculation 49
x v ii
-------
FIGURES (continued)
Number Page
2—10 PG Dual Air Register Burner ..... 50
2—11 Results of Two—staged Combustion Test: NO 1 and CO 52
2—12 Effect of Flue Gas 02 Level on NO 1 Emissions in PG Dual
Register Burner 53
2—13 Effect of Boiler Load on NO 1 Emissions in PG Dual Register
Burner 54
2—14 Performance of the PG Dual Register Burner in the Test Fur-
nace 55
2—15 Configurations of Conventional and SIR Burners . 58
2—16 Flowsheet of the 4t/h Test Furnace 60
2—17 Comparison of NO 1 Formation with SCR Burner and Conventional
Burner 63
2—18 Effect of Primary—stage Stoichiome try on NO 1 (Fuel—rich side).... 64
2—19 Effect of Primary Stage Stoichiometry on NO 1 (Fuel—lean side) .... 65
2—20 Concept of Pulverized Coal—Fired Low NO 1 PM Burner 66
2—21 Structure of the Coal—fired PM Burner 67
2—22 NO 1 and Unburned Carbon in Fly Ash vs OFA (PM Burner) 68
2—23 NO 1 and Unburned Carbon in Fly Ash vs. Ex. 02 (PM Burner) 69
2—24 Comparison of NO 1 Emissions: Laboratory Test Results vs.
Field Data . 71
2—25 Structure of Original KEI Low—NO 1 Burner 72
2—26 Structure of DV Low—NO 1 Pulverized Coal Burner 73
2—27 Model of Low-NO 1 Pulverized Coal Combustion 74
2—28 NO 1 Emission Characteristics of Original IHI Burner 76
2—29 Comparison of NO 1 Emissions in DV and Original Burners 77
xvifl
-------
FIGURES (continued)
Number Page
2—30 Examples of Combustion Modification Tests Conducted by EPDC
and Boiler Manufacturers 79
2—31 NO 1 Concentration vs. P.O.M. Emission 80
2—32 IHI’s Boundary Air System and Burner Throat Cooling Device 84
2—33 Damper Open Ratio and NO 1 Concentration (Isogo) 86
2—34 NO 1 Concentrations During Boiler Operation (Isogo) 87
2—35 Heat Release Rate at Boiler Zone (%) 89
2—36 NFK — TRW Burner 91
2—37 Effect of NFK—TRW Burner on NO 1 Concentrations 92
2—38 1111 Divided Flame Burner 94
2—39 Daido Tokushuko Self—recirculation Burner 95
2—40 Effect of Self—recirculation Burner on NO 1 Reduction 96
2—41 Schematic Diagram of the Voltmetric Burner (Nippon S.T.
Johnson)
2—42 NPLBurner (Chugairo Kogyo) 98
2—43 YLAP Burner (Yokoi Kikai Kosakiyo) 99
2—44 Effect of NPL Burner on NO 1 Emissions 100
2—45 ONK Burner: (Sanrey Reinetsu) 101
2—46 NO 1 Emission Level of ONR Burner 102
2—47 RSNT Burner (Rozai Kogyo) 104
2—48 NO 1 Emission Levels of the RSNT Burner 105
2—49 XE Burner (Osaka Gas) 106
2—50 Effect of XE Burner on NO 1 Emission Levels 107
2—51 High Speed Burner (Tokyo Gas) 108
2—52 APOC Burner (Tokyo Gas) 109
xix
-------
FIGURES (continued)
Number Paae
2—53 Schematic of Fli Burner (Chugairo Kogyo) ... 110
2—54 Effect of the Fil Burner on NO 1 Reduction 111
2—55 Schematic of MR Burner 113
2—56 TS Burner 114
2—57 NO 1 Emission Level of the TS Burner in a Batch Forging
Furnace with a Capacity of 20+/Charge 115
2—58 Schematic of TGS Burner 116
2—59 NO 1 Reduction Effects of TOS Burner in a Batch Forging Fur-
nace with a Capacity of 2+/charge 117
2—60 Schematic of TZ Burner 118
2—61 NO 1 Emission Levels with the ‘12 Burner 119
2—62 Schematic of the DLS Burner ..o... ... 121
2—63 NO 1 Emission Levels with the DLS B irner in a Batch Forging
Furnace with a Capacity of 2+/charge 122
2—64 Smokeless PM Burner 123
2—65 NO 1 and Particulate Reduction with the Smokeless PM Burner 124
2—66 Atomizing Nozzles in the Off—Stoichiometric—Combustion Low—
NO 1 Burner for Oil 125
2—67 I—Type Throat in the Dual—Air Register Burner .... ... 126
2—68 NO 1 Reduction of the OSC Burner with I—Type Throat in a
Water Tube Boiler 127
2—69 Atomizer Tips and Distribution of Water Particles 129
2—70 NO 1 Reduction By Water—Injection Atomizer Tip in a Boiler With
a Capacity of 162 t/h 130
2—71 Schematic of JSR Burner 131
2—72 NO 1 Reduction with ISR Burner in a batch with a Capacity of
130+/h 2—73 Schematic of UN Burner 132
2—73 Schematic of UN Burner 133
xx
-------
FIGURES (continued)
Number Page
2—74 NO Emission Levels with UN Burner 134
I
2—75 Schematic of SRG Burner 135
2—76 Effect of SRG Burner on NO Reduction 136
I
2—77 Schematic of CVS Voitmetric Burner 137
2—78 Schematic of TCG Burner 139
2—79 Schematic of BLN Burner 140
2—80 NO Emission Level with . .N Burner in a Boiler with a
Ca acity of 1.3t/h 141
2—81 NO Reduction by Two—stage Combustion in a Refuse Incinerator.... 143
2—82 Flow Diagram of the GQ Process Used in Cement Kilns 144
2—83 The FT System Used in Both Oil— and Coal—Fired Cement Kilns 146
2—84 Emulsified Oil Production Systems 148
2—85 Reduction of Soot Emissions by Emulsified Oil 150
2—86 NO Reduction by Emulsified Oil in a Water Tube Boiler with
Ca acity of 12t/h 151
SECTION 3
3—1 Effect of Oxygen on NO Removal 158
3—2 Catalyst Size vs. Dust Deposition for Oil—Fired Dust 161
3—3 SCR for Oil—Fired Boiler Flue Gas at 350°C 161
3—4 NO Concentration vs. Catalyst Quantity 162
3—5 De—NO Efficiency vs. Catalyst Quantity 162
3—6 Criteria for Catalysts Used with Clean Gas 166
3—7 Deactivation of Catalyst on & — AlaO, Carrier by SO 1 166
3—8 Calcination Temperature vs. Initital Activity 169
xxi
-------
FIGURES (continued)
Number Paae
3—9 Effect of Additives on MO 1 Activity 169
3—10 Catalyst Test with Oil—Fired Boiler Flue Gas 170
3—11 Additional Catalyst Tests with Oil—Fired Boiler Flue Gas 170
3—12 Additives’ Effect on TiO—Based Catalyst 172
3—13 Activity of TiO—Based Catalyst 172
3—14 NO Removal and SOs Ozidation 173
I
3—15 Performance of Ti—V and Ti—W--V Catalysts 173
3—16 Effect of Potassium on TiOz—VzOs Catalyst 174
3—17 Test Results of Accelerated Deterioration Test with
Alkalimetal Sulfate 174
3—18 Low Temperature Catalyst with Zeolite Carrier 177
3—19 Catalytic Activity of Various Transition Metal Ion—Exchanged
Y—Type Zeolites for the NO—NBa Reaction 177
3—20 Decrease in Catalytic Activity of TiOs—Metal Oxide Catalysts
at-250°C 178
3—21 Decrease in Activity of TiOs—MoO—VsOs Catalyst at 250°C 178
3—22 Decrease in Activity of TiOz—MoO—VsOs Catalyst at 200°C 180
3—23 Decrease in Actiivity of TiOs—MoO—VsOs Catalyst at 250°C 180
3—24 Recovery in Activity of Contaminated Catalyst During Three—
Hour Heating Period 182
3—25 Effects of 450°C Thermal Treatment on Catalyst Contaminated
with Ammonium Bisulfate 182
3—26 Effect of Thermal Decomposition at 450°C 184
3—27 Effect of Alkali on TiOs—VsOs Catalyst 184
3—28 Effect of Alkali on TiOs—Metal Oxide Catalysts 185
3—29 Effect of Alkali on TiOz—NoO—Vz—Oa Catalyst 185
xxii
-------
FIGURES (continued)
Number Page
3—30 KZSO4 Poisoning and Recovery of the TiOz—Based Catalyst 186
3—31 Schematic of a Moving Bed Reactor 190
3—32 Moving Bed Reactor for a Large Amount of Gas 190
3—33 Cross Sections of Parallel Flow Catalysts 194
3—34 Installation of the Catalyst 195
3—35 Characteristics of Coated Honeycomb Substrate 199
3—36 Velocity Distribution in Each Section of a Channel of Plate
and Honeycomb Catalyst 201
3—37 Results of a Test Using a Honeycomb Catalyst with Flue Gas
Containing 380 mgI Nm3 of Fly Ash 205
3—38 Total Flue Gas Treatment Process for Coal—Fired Boilers 208
3—39 Change in Catalyst Erosion Rate with Time 210
3—40 Anticipated Weight Loss of Plate Catalyst vs. Operating Time 210
3—41 Results of Accelerated Erosion Test 212
3—42 Fly Ash from Coal—Fired Boilers in Japan 213
3-43 S R for Coal—Fired Boiler Flue Gas at 350°C Using Honeycomb
Catalyst 215
3—44 Air Preheater Elements 217
3—45 Formation Temperature of Ammonium Sulfate and Bisulfate 219
3—46 Melting Point of NH3—112S04 System 219
3—47 Relationship of Oz and SO, Concentrations 221
3—48 Relationship of SO, Concentration to Particulate Content 221
3—49 Apparatus for Amnionium Bisulfate Deposition Test 222
3—50 Tube Used for Ammonium Bisulfate Deposition Test 223
3—51 Gas Velocity and Deposition Ratio 225
xxiii
-------
FIGURES (Continued)
Number Page
3—52 NH4/S04 Ratio of Deposits Formed from Gases of Varying
Compositions in Different Tube Sections 225
3—53 NH Mole Ratio vs. Deposit Ratio . 226
3—54 Inlet S03 and NB3 Concentration vs. Deposit Ratio 226
3—55 Schematic of Ljungstrom Air Preheater 228
3—56 Temperatures of Gas, Air and Heating Element in a Ljungstrom
Air Preheater 229
3—57 Heating Element Configurations in Conventional and Modified
Ljungstrom Air Preheaters 231
3—58 Two Types of Heating Elements 232
3—59 Rothemuehle Air Preheater Used in a Pilot Test Plant 233
3—60 DepositsinAirPreheaterTubes . ........... 234
3—61 Corrosion of Steel by Ammoninm Bisulfate 237
3—62 Ammonium Bisulfate Corrosion Rate of Stainless Steel and
237
3—63 Corrosion of Mild Steel by Ammonium Bisulfate Showing the
Effect of Ammonium Sulfate 238
3—64 Effect of(NE4)2S04 on Corrosion of Mild Steel by Ammonium
Bisulfate 238
3—65 Corrosion of Mild Steel by Sulfuric Acid Containing Ammonium
Bisulfate .. ...... 239
3—66 Effect of Potassium Chloride on Corrosion of Mild Steel by
%oniu.m Bisulfate 239
3—67 Adsorption of Unreacted NB3 on Fly Ash 247
3—68 Three Most Popular Wet Process FGD Systems Used in Iapan 250
3—69 Comparison of N113 Analysis Methods 253
xx iv
-------
FIGURES (continued)
Number Page
SECTION 4
4—1 S Plant Layout 263
4—2 Flowsheet for Standard Boiler Denitrification Equipment Used
with Heavy Oil Combustion Gas 264
4—3 Ammonia Injection System 265
4—4 Economizer Outlet Temperature Adjustments 267
4—5 Soot Blower Used with an SCR Reactor 268
4—6 Expected Performance of an SCR Unit for a Utility Boiler 269
4—7 S Reactor and Catalyst Bed 276
4—8 S R Catalyst and Reactor Arrangement 279
4—9 Retrofit Construction of the SV R System for the No. 2 Boiler
atludainatsuStation 283
4—10 Flowsheet of the S R System for the No. 2 Boiler at
Kudamatsu Station 284
4—11 Fixed Bed Reactor for Oil—Fired Applications 285
4—12 Flue Gas Treatment System for No. 1 175 MW Coal—Fired Boiler 293
4—13 Isometric Flow Diagram of Boiler and St.R Reactor 295
4—14 S R Reactor and Catalyst Basket at Shimonoseki Power Station 296
4—15 Principle of Unreacted NU3 Analysis 298
4—16 Flowsheet for the De—NO System at the Tomato—Atsuma Station 301
4—17 Catalyst Used in the Tomato—Atsuma S R Unit 304
4—18 S R Reactor Retrofitted to Boiler Nos. 2, 4, 6, and 7 309
4—19 Performance of the S R Unit for the 250 MW LNG—Fired No. 6
Boiler 310
4—20 S R Reactor Retrofitted to Boiler Nos. 2, 3, and 4 311
4—21 ScR Reactor Retrofitted to Boiler No. 6, Himeji Station 313
xxv
-------
FIGURES (continued)
Number Page
4—22 Flowsheet of the Nakosa S R Test Plant 314
4—23 Operating Experience of High —Dust Dc—NO System 316
4—24 Operating Experience of Air Preheater vs. Soot Blowing 317
4—25 Operating Experience of Low—Dust De_NO System 318
4—26 Operating Experience of Air Preheater vs. Soot Blowing 319
4—27 Flowsheet of the Takehara S Pilot Plant 322
4—28 Effect of Reaction Temperature on NO Removal Efficiency,
S R Pilot Plant, Takehara Station 324
4—29 Effect of Nifi/NO Ratio on NO Removal Efficiency and
Unreacted NH3, S R Pilot Plant, Takehara Station 324
4—30 NO Removal Efficiency and Pressure Drop During 4000 hr
Test Period, S R Pilot Plant, Takehara Station 325
4—31. Temperature Swing Test Results. S R Pilot Plant, Takehara
Station 326
4—32 NRa Injection at Low Temperature and Recovery Test, Sca Pilot
Plant, Takehara Station 327
4—33 Equipment Modifications Made for S R System Installation at
EPDC’s Takehara Station 329
4—34 Arrangement of the S R Demonstration Plant, Takehara Station 331
4—35 Conventional and Modified Air Preheater Designs 331
4—36 Flue Gas Trea ent Systems 338
4—37 ScR Pilot Plant and Fly Ash Sampling Points, Mitchell
Station, Georgia Power Co 344
4—38 NO ON 600 Catalyst 345
4—39 Demonstration Test Results from the S R Pilot Plant, Mitchell
Station
4—40 NO Removal vs. Area Velocity 349
4—41 NO Removal vs. Mole Ratio 350
x
xxvi
-------
FIGURES (continued)
Number Pate
4—42 Scanning Electron MicroscopG Photographs of Fly Ash at Plant
Mitchell 3 2
4—43 Scanning Electron Microscope Photographs of Fly Ash at Plant
Mitchell 353
4—44 Parallel Flow Catalysts and their Dust Clogging Tendencies 356
SECTION 5
5—1 Flowsheet of CO Boiler and S Unit 371
5—2 Flowsheet of Oil—Fired Boiler and SCR Unit 373
5—3 Flue Gas Treatment System at the Kawasaki Plant, Nippon Yakin.... 376
5—4 Flowsheet of an SCR System for a Coke Oven 380
5—5 change in NO Removal Efficiency During Ueating Cycle 381
5—6 Operating Conditions and NO Removal Efficiency 381
5—7 FGD—ScR System for Flue Gas from Iron Ore Sintering Machine 385
5—8 Effect of Nibium on NO Removal 388
- x
5—9 Thermal Treatment System 390
5—10 Flue Gas Treatment System for an Incinerator 393
5—li. Odor Component Removal by Oxidation with the S R Catalyst 393
SECTION 6
6—1 ‘WI Laboratory Test Data 407
6—2 Calculation Procedure Used in the MHI Simulation Program 411
6—3 Results of a Computer Simulation of SNR De_NO Performance 412
6—4 Ammonia Injection Nozzles for Mizushima Plant, Unit No. 3 413
6—5 Test Results for Mizushima Plant’s Unit No. 3 414
6—6 MEl Nozzle Cooling System for Unit No. 4, Mizushima Plant 415
6—7 Locations of Nozzles in the No. 2 Boiler, Chita Station 417
xxv ii
-------
FIGURES (continued)
Number Page
6—8 Effects of NH3 Injection Position and Load on NO Reduction
Rates During Operation of the No. 2 Boiler at th Chita
Station 419
6—9 De—NO Performance Test Results for Unit No. 2 at the Chita
Stati n 420
6—10 Simplified Flowsheet of the Combined SNRJSCR System at Tokyo
Electric’sOhiStation 429
6—11 Schematic Diagram of the Combined SNRJScR System at Tokyo
Electric’s Yokosuka 430
6—12 Typical Operation Data for Tokyo Electric’s Combined SNRJSCR
System 431
6—13 Tokyo Electric’s Test Data: Effect of Boiler Load on the
SNRJSCR Combined System 431
6—14 Staged Combustion and Staged Combustion/MAC I Processes.......... . 434
6—15 Flowsheet of a Small—Scale MACT Test Plant . 436
6—16 change in NO Concentration Caused by MACT ...... 438
6—17 Gai Temperature, Oa Percentage and NO 1 Removal During MACI
Treatment ..... 438
6—18 NO Concentration Before and After MACT Treatment 439
I
6—19 NO Decomposition Ratio at Different Temperatures and 02
Concentrations 439
6—20 Inlet NO and Decomposition Ratio 440
6—21 Relationship Between and NO Decomposition for Various Fuels... 440
6—22 Conversion of NO to N ih by Secondary Fuel Addition 442
6—23 Conversion of Ni to NO by Air Addition for Complete
Combustion 442
6—24 03 Percentage and NO Concentration After Combustion Modi-
fication (CM) and CM 1 P1us MACT in a Corner Firing Furnace
Using Three Fuels 445
6—25 Results of MAC i ’ Pilot Plant Tests with Propane Fuel 446
xxvi i i
-------
FIGURES (continued)
Number Page
6—26 Results of MACT Pilot Plant Test with Heavy B Oil Fuel 446
6—27 Results of MACT Pilot Plant Tests with Coal Fuel 447
6—28 Results of MACT Pilot Plant Tests with Oil Containing
Pyridine
6—29 Flowsheet for the Activated Carbon Process Pilot Plant at
EPDC’s Takehara Station 450
6—30 Effect of Temperature and Gas Volume on SO and NO
I I
Removal Efficiencies 452
6—31 SO and NO Concentrations and Removal Efficiencies
Dnhng an 0OO flour Continuous Test Cycle 454
6—32 Flowsheet of an Ebara Electron Beam Process Pilot Plant 456
6—33 Dose Rate Distribution in Ebara Process Pilot Plant Reactor 458
6—34 Relationship Between Gas Rotation Ratio and SO and NO
I I
Removal Efficiencies 459
6—35 Effect of Dose on Removal Efficiencies and Exhaust NR3 460
6—36 Results of 600 Hours of Continuous Ebara Process Pilot Plant
Tests 462
6—37 Removal of SO and NO vs. Amount of Ammonia Added 464
I I
6—38 Layout Plan for a 1,000,000 Nm 3 /hr Commercial Ebara Process
Plant 466
6—39 Flowsheet of the MKK Potassium—EDTA Wet Simultaneous Removal
Process 471
6—40 Annualized Cost for Various Operating Conditions in an MK
Process System 475
6—41 Oxidation Ratio vs. Catalyst Type and Temperature 480
6—42 Hitachi Zosen’s Three Stage Combustion Test Furnace 483
6—43 Results of Three Stage Combustion Tests 483
xxix
-------
SUMMARY
NO Standards and AmbLent Concentrations
aecent air pollution control efforts in Japan have concentrated on
NO 1 abatement, since ambient SOi concentrations already have been drasti-
cally reduced in response to stringent standards. In 1978, the ambient air
quality standard for NOs was amended from 0.02 ppm to the range of 0.04 to
0.06 pp as a daily average. In regions with NOz concentrations above 0.06
ppm, the concentration will be reduced to 0.06 ppm by 1985. In regions where
NO 2 concentrations range from 0.04 to 0.06 ppm, efforts will be made to keep
the concentrations from substantially exceeding the present level. In areas
where the concentration falls below 0.04, efforts must be made to keep at that
level. The new NOs standard is relaxed compared with the previous standard,
0.02 ppmas a daily average, but is still more stringent than the U.S. stan-
dard of 0.05 ppm as a yearly average.
In regions with large cities such as Tokyo and Osaka, ambient NOs concen-
trations often exceed the standards, reaching 0.07—0.08 ppm as daily averages.
The prefectoral governments of Chiba, Kanagawa, and Nb have established even
more stringent regulations and plan to reduce NOz concentrations to 0.04 ppm
from the current 0.05—0.06 ppm level. Even in regions with NO 2 concentrations
below 0.04 ppm, NO 1 reduction is often required by local governments to
prevent any further increase.
Nearly 2 million tons/year of NO 1 is emitted in Japan, 60 of which
is derived from stationary sources and the rest from mobile sources. In large
cities such as Tokyo and Osaka, about 60% of the NO 1 is emitted from
mobile sources.
xxx
-------
NO 1 emissions from gasoline engine passenger cars manufactured since
1978 have been controlled by stringent regulations. The current limit is 0.25
gram/km. which amounts to 8% of the NO 1 emissions from cars in 1973.
NO 1 emissions from diesel—engined buses and trucks have been reduced to
about 50% of the 1974 level. For stationary combustion sources, emission
standards based on advanced combustion modification technology has been
applied to reduce NO 1 by 30—70%.
With these efforts, ambient N03 concentrations are no longer increasing
despite a continuing increase in the number of stationary and mobile sources.
However, it is difficult to lower current NO 2 concentrations in large cities
and industrial regions without more effective emission controls for diesel
engine cars and stationary sources.
NO. Reduction for Stationary Sources
Nearly all NO 1 emissions are produced by the combustion of fossil
fuels. In Japan, the major combustion fuel is heavy oil. This residue of the
atmospheric distillation of crude oil has been used at the rate of nearly 200
million kiloliters per year. Coal use decreased markedly between 1965 and
1975 and currently accounts for only 3% of the nation’s total energy supply.
However, coal consumption is expected to triple in the next 10 years. Im-
ported LNG also accounts for about 3% of the energy currently used and is ex-
pected to nearly triple in 10 years.
Large stationary sources such as utility boilers have reduced NO 1
emissions 50—70% by applying combustion modifications Cal) including low ex-
cess air combustion, staged combustion, flue gas recirculation and low—NO 1
burners. As a result, the NO 1 concentration in flue gas from utility
boilers is minimal———150—300 ppm for coal, 30—120 ppm for oil, and 40—60 ppm
for gas firing. Smaller boilers and furnaces have reduced NO 1 30—50% by
using low—NO 1 burners or by switching from heavy oil to kerosene fuel.
For additional NO 1 abatement, several flue gas treatment (FGT)
processes have been developed. Of all the processes, selective catalytic
xxxi
-------
reduction (ScR) which uses ammonia and a catalyst at 300_4000C to control
NO 1 , is presently the most advanced technology. Over 150 commercial SCR
plants are in operation to remove 80—90% of the NO 1 emissions. Selective
noncatalytic reduction (SNR) which uses ammonia at 800—l,000 C to remove 30—
50% of NO 1 emissions has been developed and applied to about 20 furnaces
and industrial boilers. Wet and dry simultaneous SO 1 and NO 1 removal
processes also have been developed but have not been applied commercially
except for several small units.
S R has been used most often for flue gas trea ent because of its
simplicity (which enables unattended operation), relatively high NO 1 re—
noval efficiency (80—90%), and relatively low cost. Most of the new coal—
fired utility boilers being planned will have S R units. SIR will also be
needed for some of the existing boilers even in regions with NO 1 concen-
trations below a 0.04 ppm daily average, due to local policies which forbid
any increase in NO 1 levels. For example, when a new boiler is installed
at a power station, not only the new boiler, but also some of the existing
boilers will be required to have S R units so that total NO 1 emissions
from the station do not increase.
S R is usually used with . For most boilers and furnaces, 1 is
applied first, followed by SCR in order to meet the stringent regulations.
For over 90% NO 1 reduction, the combination of 1 (to reduce 35—50% of the
NO 1 emissions) and S R (to remove 80—85% of the remaining NO 1 ) is
usually more economical than S€ R by itself.
Typical examples of uncontrolled and controlled NO 1 concentrations in
utility boiler flue gas are shown in Table S —i. Examples of NO 1 regula-
tions and emissions from utility boilers are shown in Table S—2.
A new combustion process, in—furnace NO 1 removal, has been developed
to remove about 50% of NO 1 by injecting a small portion of the fuel above
the flame, followed by air addition to assure complete combustion. By using
this process along with C%( for utility boilers, NO 1 may be reduced to 100
ppm for coal, 50 ppm for oil, and 20 ppm for gas.
xxxi i
-------
TABLE
S—i. EXAMPLES OF CONTROLLED AND UNCONTROLLED NOx CONCENTRATIONS IN
UTILITY BOILER FLUE GAS
Fuel
Outlet NO Concentration (ppm) Percent Control
Without Controlled Controlled TOTAL
I
Control, By By CM by by
ppm CM and SCR CM SCR CONTROL
Gas
Oil
Coal
200 50 10 83 95
300 100 20 ,/ i 80 93
600 250 50 58 80 91
TABLE S—2. EXAMPLES OF NO REGULATIONS AND EMISSIONS FROM
UTILITY BOILERS (ppm)
Fuel
Boiler Central Government Local Agreement Actual Emission
Gas
Existing 100 60 60 a
Gas
New 60 10
Oil
Existing 150 100
Oil
New 130 25 20 b
Coal
Existing 400 159 d 170 a
Coal
New 400 170 160 c
By combustion modification (CM).
By CM and selective catalytic reduction (ScR).
CBY CM and partial SGR.
dDesired by local government.
xxx iii
-------
Fluidized bed combustion, gasification, and liquefaction of coal all have
been tested in Japan, but these technologies are not as advanced there as they
are in the U.S. This is because Japan must import coal and because these
technologies are unable to meet the stringent Japanese NO 1 emission regula-
tions. Most of the new coal—fired boilers in Japan will use conventional pul-
verized coal combustion with Cd, SCR and FGD.
SCR and SCRJFGD Characteristics
In the past S’R has experienced several problems, but these have been
solved by recent improvements. The major problems were: 1) poisoning of the
catalyst by SO in the gas, 2) dust plugging of the catalyst, and 3) depo-
sition of ammonium bisulfate in the air preheater downstream of the S1 R re-
actor. Catalyst poisoning has been eliminated by using catalysts based on
Ti02 instead of Al203 or Fe203. The use of parallel—flow honeycomb, tube or
plate catalysts or a parallel passage reactor eliminates dust plugging. Ammo-
ninm deposits can be prevented by maintaining the concentration of anreacted
anmionia in the reactor outlet gas below S ppm and using a low—oxidation
catalyst. To do this, 0.82—0.95 mole NE3 is usually used per mole NO 1 to
obtain 8O—9O o NO 1 removal with less than 5 ppm unreacted NIh, while SOs
oxidation is kept below 1%.
S R and FGD system applications to boilers and other gas sources are
schematically shown in Figure S—i. SCR is easily applied to boiler economizer
outlet gas at 300—400°C as shown in the A portion of the figure. For SO—rich
gases, FGD may be applied downstream of SCR as shown in B. At an early stage
of development. SGR was applied downstream of RiD systems, as shown in C, to
protect the catalyst from SO 1 attack. System C, however, is expensive
since it requires large ammounts of energy for gas reheating. System B has
become popular as S0 1 —resistant catalysts have been developed.
xxxiv
-------
370 E I 370 I 150 I _______
A B] SC A J __
STACK
[ ] BOILER AH ] AIR PREIIE.ATEP H HEATER
ESP ELECTROSTATIC PRECIPITATOR
Ftgure S1 SCR and FGD System Arrangements In Use in Japan.
(numbers Indicate gas temperature in C)
B
C
D
1OA 2629
XXx V
-------
System D is often used for flue gas from a low—sulfur coal. In this sys-
tem, the boiler economizer outlet gas is first treated by a hot electrostatic
precipitator (ESP) and then by SCR and FGD. A cold ES? is not highly effi-
cient for flue gas from low—sulfur coal. For high—and medium—sulfur coals,
system B is preferable. System B may also be useful for low-sulfur coal pro-
vided that both the cold ES? and the FGD unit are designed for sufficient dust
removal.
Low—temperature catalysts have been used for 200_2500C gases such as that
produced by coke ovens, as shown in system E of Figure S— i. Since ammonium
bisulfate deposits on the catalyst at these low—temperatures, the catalyst
requires occasional heating to 400°C to renove the bisulfate.
When wet FGD is applied downstream of S R or SNR. the ammonia present in
the reactor outlet is caught by the FGD system and goes into the wastewater.
In some cases it may be necessary to use the activated sludge process to re-
move ammonia from the wastewater.
Cost of NO Abatement for Stationary Sources
The investment cost for a combustion modification system is shown in
Table S—3. Costs range from 400 to 800 per N& of flue gas or 1,200—
2,4001kW for 55—70% reduction using a combined low—NO 1 burner, staged com-
bustion, and flue—gas recirculation system.
TABLE S-3. Csl INVESTMENT COST
Method
NO removal (%)
Yen/Nm 3
Yen/Kw
I
Low-NO
burner
20—40
100—200
300—600
1 .2—2 .4
I
Combined
System ’
50—70
400—800
1200—2400
4.8—9.6
a 1 = 250
bLow-NO 1 burner, staged combustion and flue gas recirculation
xxxvi
-------
SCR costs for new 700 MW gas—, oil—, and coal—fired utility boilers are
shown in Table S—4. For cost estimation purposes, it was assumed that flue
gases leaving the boiler economizer at 330_4000C are treated in two equal
size reactors in parallel and that unreacted Nils is kept below 5 ppm. The
investment cost for 80% NO 1 removal is about 2,50O/kW for gas, 4,l00—
6,2001kw for oil, and 6,7OO—8,4O0/kW for coal. The cost differences are d ae
to the varying amounts of catalyst required. For example, a small amount of a
very active catalyst is used for gas strews while a larger amount of a less
active catalyst which is resistant to SO and dust erosion is used for
dirty flue gas (oil or coal streams). Compared with 80% removal, 90% removal
costs l5 more for gas, 25—30% more for oil, and 30% more for coal. The
investment cost of an S R system for an existing boiler is 10—50% more than
for a new boiler.
The dirtier the gas, the shorter the life of the catalyst. Therefore,
the annualized S R cost/kWhr is higher when S R is used with dirty gas. On
the other hand, the annualized cost per pound of NO 1 removed is lower with
NO 1 rich gas. The cost per pound for 80% removal is 10—17 percent lower
than the cost for 90% removal. Ninety percent NO 1 removal with a low
level of ,inreacted Nils (about 5 ppm) is not easy to obtain with a large amount
of gas from a utility boiler. The reason for this is that both gas velocity
and NO 1 concentrations vary across the duct at the reactor inlet.
The investment and annualized S R costs for 80% SOs removal for coal—
fired boilers are about one—third of those for 90% SO removal using the wet
lime/limestone FOD process. On the other hand, SCR is more expensive than CM.
Although the investment cost of S R for a gas—fired boiler is similar to that
of CM in the combined system (Table S—3), the annualized cost of SCR may be
considerably higher than ai, which has low operating costs. Therefore,
for NO 1 abatement, 01 should be used first and SCR should be used in combi-
nation with CM when CM alone is not sufficient to meet control regulations.
One 01 technique, flue gas recirculation, is relatively expensive and is not
highly efficient for coal. For this reason, flue gas recirculation may not be
useful when S R is applied to coal—fired boilers.
xxxv ii
-------
><
.4.
Ga a
TA1ILE 8—4. SCR COST FOR 700 1 1W 14EV BOILLII (1981 COST)
(7(0. BOILER UTILIZATION 4.292.000 Hwhr/year)
Low—S High-S Low-S
Oil Oil Coal
High-S High—S
Coal Coal
Fool
inlet NO ppm
60
100
200
300
300
600
Ca La lyst
Uo cycomb
Honeycomb
Honeycomb
Honeycomb
lype
pellet
Honeycomb
3.3
3.3
3.3
3.5
3.5
Cost. lO’)en/m’
3.0
3
3
2
2
2
life. year
4
90
80 90
NO removal. percent
80 90
80 90
80 90
80 90
80
Investment
5.11
6.23 7.93
6.69 8.51
7.26 9.10
8.44 10.76
(1.OOOby n/kV)
2.47 2.80
4.13
Annualized 1 b
0.44 0.36
0.59 0.76
0.65 0.82
0.81 1.04
(ycn/kWhr)
0.17 0.20
0.28
5 1n luding initial charge of catalyit, civil engineering. •nd test operation.
hinciuding l01 inLere st aiid 7 years depreciation
-------
The costs of other RT processes are uncertain because they have not been
used widely in continuous conimercial operation. However, experience with
Thermal DeNO 1 , a type of SNR used with an oil—fired utility boiler, indi-
cates that its cost is about half that of S R although the NOx removal
efficiency is also half as much (40% versus 80%).
Xxxix
-------
SECTION 1
NO EMISSIONS AND CONTROL
x
1.1 ENERGY AND NO EMISSIONS
Between 1960 and 1973 the energy supply in Japan increased sharply and
then dropped slightly during the 1974 oil crisis. Since 1975 the energy
supply has been increasing steadily (Figure 1—1). In 1980, 66.4 of Japan’s
total primary energy supply was dependent on imported oil (Table 1—1). The
Japanese gs vernment intends to restrict oil imports to 366 million kiloliters
after 1985, and to increase the use of other fuels, mainly atomic power and
coal. By 1990, 50% of the nation’s total energy supply will come from sources
other than imported oil.
Japan’s supply of domestic coal cannot exceed 20 million tons annually
because of its deep and small native mines; this necessitates the importation
of coal. In 1980, imported coal exceeded 70 million tons; nearly all of it
was used for coke production in the steel industry. Imported coal for fuel is
expected to increase rapidly in the next few years, reaching 22 million tons
in 1985, and 53 million tons in 1990.
Man—made NO emissions in Japan total about 2 million tons annually.
About 60% of the total emissions are derived from stationary sources and the
rest from mobile sources; in large cities, as much as 50—60% of the NO
comes from mobile sources.
I
-------
4—
OTHER
ATOMIC
_______ .- HYDRO
0
1960 1965 1970 1975 1980
Figure 1•1 Primary Energy Supply in Japan: 1960. 1980.
3—
2—
a
C
>-
C,
LU
z
LU
0
OIL
2
-------
TABLE 1—1. JAPAN’S ENERGY SUPPLY: 1980—2000 (MIT!, 1982)
_____ 1980 _____ ________
Amount Amount
(%)
aCounted as oil energy
bIflcludjflg solar energy, coal gasification. charcoal. etc.
- -
2000
( ?o)
Amount
(%)
Coal, total (10’ t)
Domestic
Imported
For Power Generation
92.4
18.1
74.3
21.3
(16.7)
153
18-20
133—135
66
(19.5)
200
(19.0)
Atomic Power (10’ kW)
15.7
(5.0)
46
(11.3)
90
(18.0)
Natural Gas, total (10’ kl) 5
Domestic (lOs m )
LNG (10’ t)
25.9
2.2
16.8
(6.0)
68
73
43
(11.5)
82
(11.0)
Ilydro Power (106 kW)
29.8
(5.6)
45.5
(5.0)
63
(5.0)
Geothermal (10’ k!)U
0.3
(0.1)
6
(1.0)
15
(2.0)
New Energy, etc. (10’ kl)b
0.7
(0.2)
15
(2.5)
65
(8.0)
Oil, total (106 kI)
Domestic
286.6
0.5
(66.4)
290
1,9
(49.2)
290
(37 .0)
TOTAL (106 hi)
429
(100)
590
(100)
770
(100)
-------
Figure 1—2 shows annual NOx emissions from large stationary sources.
The emissions from utility boilers add up to about 310,000 tons which account
for 34% of the emissions from large sources. As larger amounts of coal are
used the emissions from utilities may increase significantly (Figure 1—3).
Although extensive efforts have been made to reduce NO 1 from both mobile and
stationary sources, further NO 1 abatement efforts will be needed for future
coal—fired boilers.
New coal usage techniques, f1uidized —bed combustion (FBC), and coal gasi-
fication and liquefaction have been tested by many Japanese organizations.
Figure 1—4 shows the Japanese Government’s 1978 preliminary plan for develop-j-
ment of coal gasification and liquefaction. Called the Sunshine Project, this
work is being carried out in cooperation with various industries under the
leadership of the Japanese government. In addition, private companies have
been trying to develop these technologies by theniselves or through interna-
tional cooperation.
Development of FBC and coal liquefaction and gasification is not as ad-
vanced in Japan as it is in the U.S. Japan is handicapped by the necessity of
importing- coal and by stringent NO 1 emission regulations. These regulations
often require NO 1 removal from the exit flue gases of FBC or coal gasifica-
tion and liquefaction systems. In addition, each process has the following
disadvantages.
FBC; Japan has no land available for disposal of the process’ throwaway ash
containing lime and calcium sulfate. Although the absorbent may be regener-
ated and recycled, this approach may be too costly.
Gasification: For utility boilers, conventional pulverized coal combustion
with flue gas treatment by SCR of NO 1 , ESP, and FGD may be less costly and
more reliable than gasification with combined cycle power generation; both FGD
and SC1 have proved to be highly reliable in Japan. For industry, there has
been little need to use gas produced from coal. Japan plans to import in-
creasing amounts of LNG, a large portion of which will be burned at power
4
-------
Figure 1.2 Annual NO EmIssions from Stationary Sources.
5
-------
300
N HYDRO
I THERMAL
N NUCLEAR
H
200
H
H
C)
0. _____
0
100 1
14
I
N
N
N
0—
1975 1985 1990 1995
130-
0 OIL
L LNG
C. COAL
0
100 — 0
0
0
a.
L
0
50
L
0 _____
L
C
C
L C
0—
1975 1985 1990 1995
Figure 1.3 EstImated Increase in Electrical Power Generation Capacity.
6
-------
1974 — 1980 1981 — 1985 1986 — 1990 1990 —
Gasification
St 40t 250t 5 OOOt
Low calorie ) > /
High calorie _ ?. _> J .L_> l•000t
Liquefaction
Solvolysis 2.000t
—4
hydrogenation 2 .4t> 1.000 _ 3 0003 9000_15 000t)
Solvent & Hydrogen ___1. __> 3.0 00t == ====>
Figure 1—4. Preliminary Plan for Development of Coal Gasification and Liquefaction by the
Sunshine Project (Arrows Show Construction Pi riods of Plants and Coal Treating
Capacities per Day).
> Pilot Plant ) Demonstration Plant > Commercial Plant
-------
plants (Table 1—1 and Figure 1—3). It may be more appropriate for power
plants to use more coal so that industry and cities can use larger amounts of
LNG.
Liquefaction: It is not economically feasible to construct a liquefaction
plant to produce fuel oil in Japan because of the necessity of importing
coal. The primary objective of the liquefaction tests in Japan has been to
produce a binder for coke production. When fuel oil from coal becomes neces-
sary in the future, it would be more practical for liquefaction plants to be
constructed abroad near large coal mines and the fuel oil exported to Japan.
Given these disadvantages as well as the technical and environmental
problems, the Sunshine project plans shown in Figure 1—4 have been delayed, or
in some cases, terminated. Although gasification and liquefaction may be
needed in the future when the supply of oil and LNG becomes limited, further
technological and economic improvements are needed before these technologies
can be commercialized in Japan.
Many coal—fired boilers presently planned or under construction will use
pulverized coal. Most of these boilers will use a wet limestone—gypsum FGD
process and many will also use SCR.
1 .2 AMBIENT NO 2 STANDARDS
The environmental quality standard for NOi in Japan was amended in 1978
from 0.02 ppm as a daily average to a new criterion of “within the range of
0.04—0.06 ppm as a daily average.” The old standard was found to be overly
stringent and not achievable in industrial regions and large cities. Under
the new standard, NO concentrations above 0.06 ppm, must be reduced to 0.06
ppm by 1985. In regions with NOi concentrations below 0.04 ppm, efforts must
be made to keep NO from exceeding 0.04 ppm.
The new daily average standards correspond to the annual NO 2 average of
0.02—0.03 ppm, a criterion proposed by the Expert Subcommittee on NO 2 of the
Central Council for Environmental Pollution Control. At the same time it is
8
-------
expected that an hourly average of 0.1—0.2 ppm, a short term criterion of the
Subcommittee, can be met by the new standards.
There has been some criticism regarding the relaxation of the standards.
The new standards, however, are still much more stringent than those of the
U.S. and West Germany, which are equivalent to 0.05 ppm as a yearly average or
0.1 ppm as a daily average. As shown in Figures 1—5 and 1—6 and Table 1—2,
NO 2 concentrations at some Japanese monitoring stations exceed the new stan-
dards although they meet the U.S. standards (based on the new Saltzman coef-
ficient). Some of the prefectural governments have established more strict
ambient air quality standards for NOs—0.04 ppm as a daily average——and
have been trying to reduce NO 2 levels from the current 0.05—0.07 ppm to 0.04
ppm. Extensive NO 1 abatement efforts have been made so that both stationary
and mobile sources can meet these standards.
1 .3 NO EMISSION STANDARDS FOR STATIONARY SOURCES
The NOx emission standards for stationary sources were first promul-
gated by the Central Government in 1973 to apply to large boilers and fur-
naces. These standards were revised in 1975, 1977, and 1979 to become more
stringent and to cover a larger, more diverse number of emission sources
(Table 1—3)
The NO 1 emission standards for major stationary sources are shown in
Table 1—4. In addition to those shown in the table, emission standards have
been applied to many other sources such as driers, calciners, smelters, re-
formers, generators, etc. Except in the case of nitric acid plants, NO 1
from these sources are derived from combustion, and therefore, the standards
are based on the improved technology of combustion control described in Sec-
tion 2. Industrial efforts described in the following sections have enabled
these standards to be met.
Even with extensive NO 1 control for stationary sources and automobile
exhausts (Section 1.4), it is not possible to meet the NO 2 ambient air quality
standard in several regions which include large cities such as Tokyo and
9
-------
I I I
1970
72
0.028
0.027 0.027
I I I I I I
74
YEAR
76
78
80
FIgure 1.5 Changes In Annual Average Concentration of NO 2 . (Average of those
recorded at 15 general air pollution monitoring stations) (1).
70A2633
0.027
0.022
0.025
E
a.
a.
w
w
>
-I
z
z
0
z
0.03
flIV) —
v.’J’.
0.01
0
0.026
0.021
0.026
10
-------
0.05 -
0038 0040
0.04 -
uJ
0.03 -
uJ
>
z
0.02 —
0
z
0.01 —
I I I I
0
1972 74 76 78 80
YEAR
Figure 1.6 Changes in Annual Average Concentration of Nitrogen Dioxide.
(average of 26 automobile pollution monitoring stations) (1)
70A2634
11
-------
TABLE 1-2. STATE OF COMPLIANCE WITH THE NEW NITROGEN DIOXIDE ENVIRONMENTAL
QUALITY STANDARD IN 1978 (The data correspond to 98% of total
readings)
Descriptions of data
corresponding to 98
of the total readings
Tvue of
Monitoring Station
General Air
Pollution Monitoring
Station
Automobile Pollution
Monitoring Station
Number of
stations Percentage
Number of
stations Percentage
Recordings exceeding
0.06 ppm
75 7.6
77 40.5
Recordings in the range
0.04 — 0.06 ppm
233 23.8
92 48.4
Recordings below 0.04 ppm
673 68.6
21 11.1
Total
981 100
190 100
Source: Environment Agency, Japan
TABLE 1—3. EXPANSION OF NO 1 CONTROL FOR STATIONARY SOURCES
- 1973
1975
1977
1979
NO 1 sources Large boilers
Medium sized
Relatively
Small boilers
controlled and furnaces
boilers and
small boilers
and furnaces.
(additions HNO3 plants.
furnaces,
and furnaces.
Metal heating
each year)
Large cement
kilns and
coke ovens,
Medium sized
cement kilns
and coke ovens.
Waste inciner-
ators.
furnaces.
Gas generators,
calciners, etc.
Number of sources 1,500
3,400
13,000
105,000
controlled (total)
NO 1 emitted from 38
44
73
95
the controlled
sources (percent
of total NOx from
all stationary
sources)
12
-------
TABLE 1—4. NO 1 EMISSION STANDARDS FOR MAJOR STATIONARY SOURCES (pp3)
Sourcea
(O %)
Capac ity
(1000 /hr)
—1973
7375
75—77
7779
1979—
Boiler (gas)
(5%)
>500
100—500
40—100
10—40
<10
130
130
130
150
150
130
130
130
150
150
100
100
130
130
150
60
100
100
130
150
60
100
100
130
150
Boiler (oil)
(4%)
>500
100—500
40—100
10—40
5—10
<5
180
170
190
210
250
250’
180
180
180
230
250
25O
150
150
150
150
250
250’
130
150
150
150
iso
180
130
150
150
150
iso
180
Boiler (coal)
(6%)
>100
40—100
11—40
5—10
<5
480
600
600
480
680’
400
480
600
480
48O
400
480
480
480
480’
400
400
400
400
400
400
400
400
400
400
Sinteri g
llachine
>100
10—100
(10
260
270
300 b
260
270 b
300
260
270 b
300
220
220 b
300
220
220
220
Metal heating
furnace
(Radiant type)
(11%)
>100
40—100
10—40
5—10
<5
200
200
200
200
200 b
200
200
200
200 b
200
100
150
150
200
200 b
100
150
150
150
180
100
150
150
150
180
Pet ro1e
heating
furnace (6%)
>40
10 -40
5—10
(5
170
180
180
200
170
150
180
200
100
150
180
200
100
130
150
180
100
130
150
180
Cement Riln
(10 )
>100
(100
480
480
480
480
250
480
250
350
250
350
Glass furnace
(1S )
500
500
500
500
450
Waste
incinerator
(12 )
>40
(40
300 b
300 b
300
300
300 b
300 b
250
300 b
250
250
Coke oven
(7%)
>100
(100
350
350
350
350
200
350
170
170
170
170
Nitric acid
plant
200
200
200
200
200
5 To be net by 1984.
bTo be net by 1982.
13
-------
Osaka. Therefore, the Central Government (Environment Agency) decided to ap-
ply total emission regulations to Tokyo, Osaka, and Yokohama in order to fur-
ther increase NO 1 abatement and meet the air quality standard. Similar
total emission regulations for SO 1 have been successfully applied in several
regions and have resulted in the attainment of the SOi air quality standard.
The NO 1 regulations are more difficult to attain because over 5O of NO 1
emissions in most large cities are derived from mobile sources, while most
SO 1 emissions are from stationary sources. With implementation of the total
emission regulations as well as stringent local regulations, flue gas treat-
ment for NO 1 removal will be needed in addition to combustion modification.
1 .4 NO CONTROL FOR MCB LE SOURCES
The emission standards for gasoline engine passenger cars are shown in
Table 1—5. According to the 1978 standards, NO 1 discharge is controlled be-
low 0.25 g/km, on the average. Accordingly, permissable limits are set at
0.48 glkm, as shown in Table 1—5. At the same time, those in 11 modes are
limited to 6.0 g/test (average discharge 4.4 g/test). The 1978 standard is
responsible for a 60—70 percent reduction in NO 1 discharge compared with the
1976 regu lations and to over a 90 percent reduction compared with the time
when there were no regulations. The 1978 standards are much more stringent
than similar standards in the U.S. (Table 1—6).
It was predicted that automobile performance and fuel economy would
decline measurably with enactment of the 1978 standards. However, strenuous
efforts have been made to maintain and even to improve fuel economy while
maintaining performance. As the result of these efforts, fuel consumption
rates have improved considerably in the vehicles which satisfy the 1978 stan-
dards. Fuel economy for some models is lower than that of automobiles which
complied with the 1973 standards.
The NO 1 regulations for heavy duty diesel engine vehicles (buses and
trucks) have also been tightened (Table 1—7). Under the 1977 regulation,
NO 1 emissions from diesel vehicles were reduced by about 3O while those
14
-------
TABLE 1—5. AUTOMOBILE EMISSION STANDARDS IN IAPANa
BC Co NO 1
(g/km) (g/km) (g/km)
1973
3.80 (2.94) 28.00 (18.40)
1975
0.39 (0.25) 2.70 (2.10) 1.60 (1.20)
1976
0.39 (0.25) 2.70 (2.10) 0.84 (0.60)
1978
0.39 (0.25) 2.70 (2.10) 0.48 (0.25)
aFigures show
emissions.
allowable limits; those in parentheses indicate average
TABLE 1-6.
AUTOMOBILE EMISSION STANDARDS IN TUE UNITED STATESa
HC CO
(g/km) (g/km)
NO 1
(glkm)
1978
0.94 9.4
1.25
1979
0.94 9.4
1.25
1980
0.26 4.4
1.25
198].
0.26 2.1
0.63
aEstablished by the Clean Air Act of 1977.
15
-------
from gasoline engine and LPG engine heavy duty vehicles were reduced by about
45%—as compared with the uncontrolled emissions in 1973.
With these regulations and efforts to reduce traffic by 1O , the NO 1
‘missions from automobiles will be reduced by about 4O by 1983; but will in-
crease after 1984 due to projected increases in traffic (Figure 1—7). The
1978 standard for gasoline engine passenger cars will not be amended. Vow—
ever, NO 1 control for diesel engine vehicles will be tightened (Tables 1—7
and 1—8).
TABLE 1-7. EMISSION STANDARDS FOR DIES GINE HEAVY DUTY V ICLES (PPM)
CO
UC
NO
x
Auxiliary Direct
Chamber Type Injection Type
1974
980
670
590
1000
1977
980
670
500
850
1979
980
670
450
700
1981
450
540
1983
290
470
TABLE
1—8. NO
Current
CONIROL FOR
I
Regulation
DIESEL
Interim
ENGINE
Target
PASSENGER CARS (g/km)
Target
Medium
Size
1.2
0.9
0.6
Small
Size
1.0
0.7
0.5
1 .5 ME1 ODS OF NO ABATE!IENT FOR STATIONARY SOURCES
The major methods of NO 1 abatement for stationary sources are classifi-
ed in Table 1—9.
16
-------
(1)
300 /
/
250 - ,/‘
U)
200 —
• (2)
U’
2 (3)
U) -
E
w
150—
- (1) NO REGULATION
(2) NO MODIFICATION TO 1976 CONTROL
100 — (3) 1978 CONTROL ON PASSENGER CARS
(4) 1978 CONTROL ON PASSENGER CARS,
1977 CONTROL ON DIESEL-ENGINE CARS
AND HEAVYWEIGHT GASOLINE-ENGINE CARS
0 I • . I • • I • I • . I
1965 1970 1975 1980 1985 1990
YEAR
Figure 1.7 Estimated Total Amount of NO Emissions from Road
Traffic in the Tokyo Bay Area. (2)
17
70A2635
-------
TABLE 1—9. CLASSIFICATION OF NO ABATEMENT METHODS
I
Low Oxygen Combustion
Flue Gas Recirculation
Combustion Staged Combustion
Modification Off—Stoichiometric Combustion
Low NO Burners
Reduction of I
Formation
Use of Low— Change of Fuel
Nitrogen Fuel L Nitrogen Removal from Fuel
Selective Catalytic Reduction (SNR)
NO only Selective Noncatalytic Reduction (SNR)
In—furnace Reduction
Dry
Process Carbon Adsorption
SO 1 and NO Copper Oxide Adsorption
N0 L Electron Beam Radiation
Re oval
N0 only Oxidation Absorption
Wet I Oxidation Reduction
Process SO and NO L Absorption Reduction
x x
Combustion modification is the most reasonable first step for stationary
source NO 1 abatement. It is described in more detail in Section 2. Combus-
tion modification has been widely applied in Iapan and has reduced NO 1 by 20—
70 percent from each stationary source.
A widespread switch from high—nitrogen to low—nitrogen fuel has occurred
in .Tapan. Heavy oil, now the major fuel, usually contains 0.2—0.3 percent
nitrogen, and its combustion gas contains 100—130 ppm of NO 1 using the best
combustion control. Hydrodesulfurization of heavy oil, which removes 70—90
percent of sulfur, also removes 20—30 percent of the nitrogen and contributes
to NO 1 reduction by 20—30 ppm. Kerosene is used in many small stationary
sources.
18
-------
A considerable number of power plants constructed recently near large
cities burn LNG (imported liquid natural gas), which produces only 30—40 ppm
of NO 1 using the best combustion control. However, there is, a limit to the
supply of such clean fuel, and it is expensive as well.
Recently it has become necessary to use increasing amounts of coal,
particularly for power generation. Since coal contains over l !’ nitrogen and
produces flue gas with about 200 ppm NO 1 , even using combustion modifica-
tion, flue gas treatment may be needed for further NO 1 reduction.
any processes have been developed to remove NO from flue gas. Of
these, a selective catalytic reduction (SCR) process that uses NH3 and a cata-
lyst at 300—400°C to reduce NO 1 to N 2 has been the most widely used. SCR is
simple, has a high NO 1 removal efficiency of 80—9O , and does not produce by-
products which present difficult disposal problems. Although NO 1 can be
reduced to N a by other reducing agent gases such as Hz, CO, and CB., large
amounts of these gases are consumed by reacting with the 02 in flue gas. The
use of NHj is more suitable because it selectively reacts with NO 1 (Section
3) . S R was initially applied to industrial boilers and furnaces (Figure
1—8). Since 1978, many large 5CR units have been constructed for utility
boilers (Table 4—1, Section 4).
Selective noncatalytic reduction (SNR), N E 3 reduction without a catalyst,
also has been used commercially. SNR is simpler than SCR, but its NO 1 re-
moval efficiency rarely exceeds 50 percent and a considerable amount of NH3 is
emitted (see Section 6.2).
In—furnace NO 1 reduction has been developed recently and applied to
small oil—fired utility boilers to remove about 5O of NO 1 . With this
process, a small amount of fuel, about 10 of that used for the boiler, is in-
jected above the flame in the furnace and then has air added to it for com-
plete combustion. The size of this furnace is 10 to 20 larger than a conven-
tional furnace (see Section 6.4).
19
-------
200
150
100
50
0
YEAR
a.
E
z
>-
I-
0
a-
0
I-
0
I-
Figure 1.8 Number and Total Capacity of NO Selective Catalytic
Reduction (SCR) Plants: 1974. 1981.
250
60
C,,
I-
z
0
C,,
U-
0
w
z
-J
I-
0
I-
74 76 78
0
80 82
20
70A2636
-------
Among dry SO and NO 1 removal processes, carbon adsorption and
electron beam processes have been tested at a pilot plant while a copper oxide
adsorption process has been used for an industrial boiler. The carbon and
copper oxide processes use ammonia to decompose NO 1 to N a and HzO; the elec-
tron beam process forms ammonium nitrate and ammonium sulfate byproducts
which are used for fertilizer.
The disadvantage of wet processes is that they require extensive waste—
water treatment to eliminate nitrate and other compounds. Efforts have been
niade to recover useful byproducts and to minimize the amount of wastewater
produced.
1 .6 LOCAL REGULATIONS AND NO 1 ABATE 1ENT
1.6.1 Introduction
Prefectural governments and large city authorities have established pol-
lution control ordinances for stationary sources which are usually much more
stringent than those set by the Central Government. In addition, city
authorities and citizens groups often make agreements with industries for
further pollution control. A good example of the way in which local pollution
control regulatory agencies operate is the anagawa Prefecture.
Kan.agawa Prefecture adjacent to Tokyo has an area of 2,391 km 2 and
three major cities——Yokohama, Kawasaki, and Yokosuka with populations of
about 4 million, 1 million, and 400,000, respectively (Figure 1—9). These
cities have numerous industries including several electric power stations,
steel works, and oil refineries.
The first step towards pollution control was taken in 1951 with the
enactment of the anagawa Prefectural Workshop Pollution Prevention Ordinance.
The ordinance was revised in 1955, 1971, and 1978 to comprise the present
anagawa Prefectural Ordinance for Pollution Control.
21
-------
YAMANASHI
PREFECTURE
TOKYO
METROPOLIS
KANAGAWA
PREFECTURE
ODAWARA CITY
TOKYO
BAY
SHIZUOKA
PREFECTURE
SAGAMI
BAY
. I
PACI FIC
OCEAN
0 5 10 15 km
Figure 1.9 Kanagawa Prefecture and its Major CIties.
22
-------
1.6.2 Air Quality Monitorina Systems in the Kana awa Prefecture
The three major cities in Kanagawa have installed their own monitoring
systems with automatic analyzer—telemeter—computer systems for measuring,
recording, and analyzing pollutants in ambient air, automobile exhausts, and
stack emissions. For example, Yokohama City has 12 ambient air monitoring
stations for measuring SOi, suspended particulates, NO, NO 2 , and oxidants as
well as wind velocity and direction; eight automobile exhaust monitoring sta-
tions for suspended particulates, CO. NO, NO 2 , HC, oxidants, as well as the
number of passing cars; and stack gas monitoring systems to analyze SON,
NOT, and particulates at 40 major stationary sources. All of the analytical
results are transmitted to the Yokohama City Office on an hourly basis.
In addition, Kanagawa Prefecture has its own ambient air and automobile
exhaust monitoring stations in areas other than the three cities. Informa-
tion from a total of 72 monitoring stations, including the stations owned by
the cities, are transmitted to the Prefecture’s air monitoring center. The
annual costs for the center are shown in Table 1—10.
The initial air quality control efforts in Kanagawa focused on SOx
abatement. Since the ambient SOs concentrations now have been reduced to meet
the standard (0.04 ppm daily average) at most of the monitoring stations,
recent efforts have been concentrated on NO control.
The daily average NOs concentrations at some of the stations in the three
cities sometimes exceeds the 0.06 ppm national standard. Kanagawa Prefecture
has been making efforts to reduce the concentrations to 0.04 ppm by extensive
NO reduction.
1.6.3 NOx Emissions and Control in Kanagawa Prefecture
In the city of Yokohama. about 65T3 of the total NO emissions are de-
rived from stationary sources while about 30 come from automobiles. The lat-
ter is controlled by the national regulation; stationary source emissions are
regulated by prefectural ordinances and agreements between the Prefecture or
23
-------
TABLE 1—10. TOTAL COST FOR KANAGAVIA CONTINUOIJ AIR MONITORING CENTLH (1970—1980)
Measuring Sees — — Te1emetiy . ______ -— _________ ______
Purchase Operating Purchase Cpcrating Purchase Operating
Items/Year Cost ($1,000) Cost ($1,000) Cost ($1,000) Coal. ($1,000) Cost ($1,000) Cost ($1,000) TOTAL
1970 1624 23 50 1,697
1971 307 109 169 22 1 608
1972 192 144 670 26 154 2 1,183
1973 172 220 947 13 9 1,361
1974 144 176 74 23 4 30 451
1975 21 326 123 35 505
I ’ , )
1976 29 304 133 35 501
1977 455 361 158 32 33 1,039
1978 672 361 160 8 36 1,237
1979 301 399 149 7 39 1,108
1980 166 444 402 141 39 1,192
TOTAL 4,083 2,844 2,498 948 255 259 10.887
JOTAL 6,927 3,446 514 10,887
-------
City and industry. Yokohama has 1,165 plants which use more than 1 kiloliter
of oil per hour including 13 large plants (three utility power stations, three
oil refineries, two glass plants, etc.) that emit about 95% of the NO 1 from
all stationary sources. As a whole, a 64% reduction of the 1974 N0 emis-
sions must be achieved to attain 0.04 ppm as a daily average. Therefore,
60—70% NO 1 reduction of the 1974 emissions is requested of the large plants
and a 3O—5O9 reduction of the smaller plants. Since these reductions are not
easily achieved by combustion modification, a considerable number of plants
must remove NO 1 from flue gas. According.to the City’s initial plan, the
reduction was to be achieved by 1981. This time limit has been postponed 2
years because it was too difficult to meet.
Yokohama has 3 power stations—Yokohama Power Station which burns oil
and ?Iinamiyokohama Power Station (LNG) , both belonging to Tokyo Electric, and
EPDC’s Isogo Power Station which burns coal. The UO 1 emissions and regula-
tions which apply to these utility boilers are compared in Table 1—11.
TABLE 111. NO 1 EMISSIONS AND REG ]LATI0NS FOR EXISTING UTILITY
BOILERS IN YOKOHAMA CITY
Fuel
Before NO 1
abatement
YOKOHAMA CITY (PPM)
Present Central
level government
Yokohama
(Target)
Gas
150—250
40—80
130
50
Oil
250—350
80—150
180
80
Coal
400—600
160—200
400
159
aAchleved by combustion modification
The target NO 1 concentration chosen by Yokohama City is less than one—
half the emission standard set by the Central Government and is not easily
achievable by combustion modification alone. Yokosuka and awasaki Cities
also have set similarly stringent NO 1 target concentrations. In order to
meet these targets, 7 SCR plants are in operation for various gas sources,
25
-------
TAHLE 112. NOx REMOVAL PLANTS IN KANACI\ IA PREFECTURE
Gas Treated Year
Owner Plant Site Gas Source (1.000 Nm 3 /hr) Process Completed
a
Asahi Glass Yokohama Glass furnace 75 SCR 1976
Tokyo Electric Yokohama Oil—fired boiler 500 SNRb+SCR 1978
Yokosuka 1060 SNR+SCR 1978
Nippon Kokan I awasaki Sintering machine 1300 8CR 1979
AjinoLloto Kawasaki Oil—fired boiler 180 8CR 1978
Nippon Yakin Kawasaki Oil—fired boiler 15 SCR 1976
1,olio Gas Kawasaki Gas generator 10 SCR 1976
Tonen Seklyu Kawasaki Oil—fired boiler 423 SNRC 1976
Kagaku
Tonen Sekiyu Kawasaki Oil—fired boiler 423 SNRC 1977
Ku aku
Ton Nenryo Kawasaki CO boiler 314 S,IRC 1978
1oa Nenryo Kawasaki Pipestill furnace 254 SNRC 1978
Sc1ective catalytic reduction
Sd ective noncatalytic reduction
dSecti e noncatalytic reduction for emergency use
I)emonstration Plant. The SNR system was removed in 1980.
-------
while four plants using selective noncatalytic reduction (Thermal DeNOx)
have been installed for emergency use (Table 1—12).
The N0 concentrations in the flue gas from the coal—fired boilers at
EDPC’s Isogo Station, has been lowered from 400—500 ppm to 160—180 ppm by com-
bustion modification. Although the City has requested further NO reduction
the power station has no space to install a NO removal facility because FGD
plants were retrofitted in the narrow space beside the boilers. Therefore,
efforts have been continued for further NO abatement by combustion modifica-
tion.
1 .7 AIR POLLUTION PROBLEMS R ATED TO N0
In Japan, ambient SOi concentrations have been lowered substantially in
order to meet the stringent standards (Figure 1—10 and Table 1—13). On the
other hand, the number of designated air pollution patients have increased
(Figure 1—10) . Designated patients are the inhabitants of specific polluted
regions such as Yokohama, awasaki, Amagasaki and portions of Tokyo and Osaka
who have been diagnosed as air pollution victims. They receive compensation
from the government for medical care and living expenses.
TABLE 1-13. AMBIENT AIR QUALITY STANDARDS (ppm)
SOz NO2
Daily Yearly Daily Yearly
Japan 0.04 (0.02) 0 • 02 a b 001 a b
0.04—0.06 (0.02—0.03)
U.S.A. 0.03 0.05
West Germany 0.05 0.05
a 197 3—1978
bSince 1979
27
-------
0.06
0.04
0.02
0
60,000
60,000
40,000
20,000
0
1980
—0-— DESIGNATED AIR POLLUTION PATIENTS
1965
1970 1975
. SO 2
U)
z
w
I-
a-
U-
0
w
z
FIgure 1.10 Annual Average Ambient SO 2 Concentrations in 15
Major Cities and Industrial Regions and Number
of Designated Air Pollution Patients.
70A2638
E
a-
a-
C l )
z
0
I-
I-.
z
w
0
z
0
0
28
-------
The major reason for the increase in patients is the expansion of the
designated regions to include several regions of Tokyo and Osaka. As ambient
NO 2 concentrations have increased considerably between 1970 and 1974 and
slightly since 1978, some people claim that the increase in patients is caused
by ambient NO 2 concentrations. Although the NO 2 concentrations exceed the
standard in several regions, NO 2 may not be the direct cause of patient in-
creases; the concentrations are well below the U.S. standard.
A number of problems with photochemical smog were encountered in 1973—
1974 but these have decreased since that time (Table 1—14). The decrease may
be related to the decrease in hydrocarbon emissions (Table 1—15). Particulate
emissions from stationary sources also have been reduced substantially.
It should be noted that the annual average ambient sulfate conceutrations
in five major cities (Tokyo, awasaki, Osaka, Amagasaki, and Kxtakyushu),
decreased from 3O tg/& in 1970 to 13 ig/m 3 in 1973. Since then the concentra-
tions have increased slightly to 15 ig/m 3 in 1979, despite an on—going de-
crease in ambient SOi concentrations. This may be due to the increase in
diesel cars which emit sulfuric acid mists and also possibly to the slight
increase in NO which tends to promote the oxidation of SOz to sulfate.
Although sulfate concentrations since 1973 have not been high, the sul-
furic acid mists from diesel cars may affect human health in .Tapan where there
are many diesel buses and trucks and the sulfur content in the light oil for
diesel engines already is high (0.4% as compared with about 0.2% in the U.S.).
Sulfur abatement of the light oil and further control of particulates for
iesel—engine exhausts may be needed.
29
-------
TABLE 1—14. NUMBER OF DAYS IN flhICH PHOIOCIIEMJCAL SMOG IVALNINGS WERE 1 SLJED (1)
Area
Prefecture 1970 ‘71 ‘72 ‘73 ‘74 ‘75 ‘76 ‘77 ‘78 ‘79 ‘80 ‘81
Tokyo Bay
Saitama
Ch ib a
Tokyo
Kana gawa
23 15 45 29 44 15 26 36 8 15 8
19 21 28 26 33 21 7 14 11 13 8
7 33 33 45 26 41 17 21 22 12 13 14
11 31 30 26 27 17 12 18 19 10 11
Aichi
Mi e
TOTALS
6 3 2
3 1
1 1
1 1
7 98 176 328 288 260 150 187 189 84 72 57
Osaka Bay
Ky 01.0
Osaka
Ilyogo
Nara
7 17 17 11 6 9 5 1 5 3
4 18 26 27 23 25 25 16 12 10 12
7 19 23 19 11 3 4 2 1 1 1
1 6 3 9 3 3 3 1
Seto
Inland
Okayurna
3
14
16
5
1
5
8
Sea
Ilirashima
Yamaguchi
Tokushirna
Kagawa
Ehime
2
9
1
2
18
5
2
4
13
4
1
2
1
1
4
6
5
3
7’
9
3
1
6
1
Ise Bay
1 5 8 2
4 6 7
2 1
1
]nc1udiiig other vrcfectutes
-------
TABLE 1—15. GASEOUS H’YJ)UOCARBON EMISSIONS Fh ii; STATIONARY SOURCES (1)
Sources
Refinery
Service Tank
Service Station
Others
67,000
68 .100
62,800
100
1973
42 ,400
41,800
78,500
100
1978
4.0
3 .7
6 .9
TOTAL 1,301.200 100.0
1 iO.0 1,117,600 10().0 100.0
Pc trol eum
Amount Ratio fl.itio Amount Ratio Ratio
(t) (% of total) (‘h) (t) ( .) (%)
5.2
5.2
4.8
15.2
Painting
Petrochemical Plant
69,800
5.4
Tank
4,900
0.4
14.6
5.8
Inks
27,000
3,600
2.4
0.3
Production
1,600
0.1
1,500
0.1
4.1
Car
Painting
37,500
2.9
45,400
1.5
Ship
Painting
19.900
564,500
1.5 47.9 16,700
43.4 552 ,800
49.5
Others
Other Solvents
2.7
109,800
8.4
8.4
55.2
81,100
Cementing
42,500
3.3
32.800
2.9
4.7
Metal degreasing
85,000
6.5
52,800
Cleaning
116,500
9.0
22.7 106,000
7.2 7.2
20.3
-------
R ER 4CES
1. Environment Agency. ¶hite Paper on Environment, 1982 (in Japanese).
. Environment Agency. Recent Countermeasures for Air Pollution Control in
Japan, 1977 (After OECD fleview of Japanese Environmental Politics in
1976)
32
-------
SECTION 2
NO ABATE EENT BY COMBUSTION MODIFICATION (CM)
2 .1 INTRODUCTION
2.1.1 Classification of Combustion Modification (CM) Techniques
NO emissions from combustion processes consist of thermal NO formed
by the oxidation of nitrogen in air at high temperatures and fuel NO de-
rived from nitrogen compounds in fuel. Thermal NO 1 can be reduced by corn—
bustion modifications, such as decreasing the oxygen concentration in the
combustion regions; shortening the residence time of combustion gases in high
temperature zones; or lowering the flame temperature. Fuel NO 1 emissions
can be reduced by decreasing the oxygen concentration in the combustion reac-
tion zone and by switching to a low—nitrogen fuel.
Combustion modification techniques used widely in Iapan for NO 1 control
can be grouped into four categories:
• Modified operating conditions
1) Low—excess—air combustion
2) Change of fuel/air contacting in the combustion chamber
3) Reduction of heat load in combustion chamber (reduction of
output power)
4) Lowering of air preheating temperature
o Burner design modification
33
-------
• Modifications in combustion system design
1) Reduction of heat density in combustion chamber (increased
furnace volume)
2) Staged combustion
3) Flue gas recirculation
4) Water or steam injection
o Other methods
1) Fuel switching
2) Firing modification
any combinations of the above techniques have also been used. Although
operating conditions can be changed relatively easily in existing installa-
tions, the changes usually reduce NO emissions only slightly and often
cause operating problems. On the other hand, modifications in burner and
combustion system design are promising control techniques. Fewer problems are
encountered when design modifications are incorporated in new plants;
relatively large reconstruction costs are required when the modifications are
applied to existing in tallations.
In Japan, NO emission limits have been applied to over 100,000
sources including boilers, various furnaces and kilns, heaters, and incin-
erators, representing about 95 of the total emissions from all stationary
sources (Section 1.3). Virtually all of the these sources have reduced NOx
emissions by combustion modification. Larger sources use the combination of
staged combustion, flue gas recirculation (gas mixing) and low—NOt burners
(Table 2—1) while smaller sources reduce NOx primarily with low—NOx
burners and fuel switching.
With these efforts, NO concentrations in boiler flue gases have been
maintained at 30—120 ppm for gas—, 60—200 ppm for oil—, and 150—400 ppm for
coal—firing to meet the national emission standards and the more stringent
local regulations (Table 2—1) . For further NO 1 reduction, over 160 SCR
plants are in operation or under construction. Normally SCR is applied in
combination with combustion modification.
34
-------
TABLE 2—1. NUMBER OF BOILERS AND COMBUSTION MODIFICATIONS APPLIED FOR OIL
AND GAS AS OF APRIL 1981 (TOKYO ECTRIC)
Methods of Combustion Modification
Number of To
Boilers
tal Generating
Output (MW)
FGRa only
15
1,857
FGR + Two stage combustion
12
5,289
FGI + Low N0 burner
9
2,710
FGR + Two stage combustion
+ Low NO burner
32
13,605
Total
68
23,461
aFlue Gas Recirculation
TABLE 22. NO ABATEMENT
BY
COMBUSTION
MODIFICATION (CM)
FOR UTILITY
BOILERS
GAS
OIL
COAL
NO Control
200—300
250—350
500—800
GMa
TSCb
150—200
150—250
180—250
180—250
400—600
300—500
GM + TSC
100—150
120—200
250—400
GM + TSC + LNBC
50—90
80—140
150—300
Gas Mixing (Flue Gas Recirculation)
Two—Stage Combustion
Low—N0 Burner
Combustion modification has also been applied extensively in the U.S. and
European countries. Combustion modification technology is similar in most of
these countries. This section describes combustion modification with emphasis
Ofl 1owNO burners developed and used in Japan.
In addition, an advanced combustion modification process has been devel-
oped which reduces about 5O of the NO 1 emissions by injecting about 10T3 of
the fuel above the flame followed by air addition to assure complete combus-
tion. The process is also considered to be a flue gas treatment process and
is described in Section 6.4.
35
-------
2.1.2 NO 1 from Pulverized Coal Burning
The reduction of NO 1 in flue gas from pulverized coal combustion has
become very important in Japan because of the rapid increase in coal consump-
tion and high NO 1 concentrations in the combustion gas. Reduction of fuel
NO 1 is of particular importance because coal contains much larger amounts of
aitrogen (l—3 ) than does oil (0.l—O.5 ). The fuel NO 1 from coal consists
of volatile NO 1 derived from volatile nitrogen compounds in the coal and
char NO 1 derived from residual nitrogen compounds in the char. Abatement of
the volatile NO 1 can be accomplished efficiently by combustion control.
Char NO 1 control may not be accomplished as easily since it is less sen-
sitive to injection design modifications in the early stage of combustion.
Staged combustion is highly effective for NO 1 abatement for pulverized
coal flue gas because it reduces both fuel and thermal NO 1 . On the other
hand, flue gas recirculation (gas mixing) is less effective because it reduces
thermal NO 1 which is not as much of a problem for coal—burning facilities.
Various types of low—NO 1 burners have been developed and used in com—
bination with other combustion modification techniques such as staged combus-
tion and flue gas recirculation as shown in Table 2—3.
TABLE 2—3. EXAMPLES OF COMBUSTION MODIFICATION FOR COAL—FIRED UTILITY
BOILERS
Station
(Company)
Boiler
(M Y)
Coal
Source
Burner Type
NO 1 (ppm)
ISOGO (EPDC)
265 a
Domestic
DF_CNC
150200
Tomato—Atsuma
(Hokkaido)
35oI
Domestic
Dual Aird
Register
200
Iatsushima (EPDC)
500 b
Imported
SGRe
200—280
Staged combustion is used.
sta ed combustion and flue gas recirculation are used.
4 See Section 2.2.
See Section 2.3.
eSee Section 2.4.
36
-------
2 .2 NO 1 ABATEMENT BY IEI—FIV DF—CN COAL BURNER (1, 2. 3)
2.2.1 IllI—FW DF—CN Burner
Ishikawajima—Barima Eleavy Industries (IHI) and Foster 7heeler Corporation
of the U. S. have developed a low—NO 1 burner called the ]:HI—F 7 DF—CN (Dual
Flow Convergent Fuel Nozzle) burner. It is based on the following principles
for NO 1 reduction:
1) Minimum primary/secondary mixing prior to completion of the
combustion of volatile fraction.
2) Ignition stability at the injector.
3) Dilution of the secondary air with recirculating combustion
products.
The configuration of the burner is shown in Figure 2—1. The secondary air
sharing the largest portion of total combustion air is divided into an outer
and inner flow by an annular flow divider located between the outer sleeve of
the coal nozzle and throat ring. The flow quantity and/or the swirl intensity
of inner air is controlled by changing the opening of the circularly—linkaged
inner vane located just downstream of the secondary air vane. The inner
swirled air is supplied mainly to the burner axis zone, where the fuel—rich
combustion of volatile matter takes place and stabilizes the flame. The outer
swirled air is supplied along the periphery of the burner throat and controls
the mixing with char particles as well as strengthening the external recircu—
lation flow of combustion products.
The pulverized fuel is injected through a ring— shaped convergent nozzle,
so that fuel—rich combustion on the burner axis can take place. The inner
sleeve of the fuel nozzle can be moved toward and away from the furnace. The
velocity of the injected fuel may be varied to some extent by moving the inner
sleeve, and thus the initial mixing of fuel and air may be controlled.
37
-------
FUEL LEAN
COMBUSTION
STRENGTHENED EXTERNAL
Figure 2.1 DF .CN Burner.
38
-------
2.2.2 Test Furnace and Coal Composition
IHI conducted a combustion test using DF—CN type burners with a capacity
of 2,000 kg/hr on a test furnace installed at IHI ’s Aioi Works. The study
consisted of two stages. During the first stage, the characteristics of the
DF—CN burner were investigated. During the second stage, other combustion
modifications such as staged combustion and/or flue gas recirculation were
examined for use with the DF—CN burner. The test furnace is illustrated in
Figure 2—2. It was designed to have a geometrical similarity to the actual
furnace of the coal—fired boiler. Single—burner and overfire—air ports COAP)
for staged combustion were installed on the front wall. Maximum heat input to
the furnace is approximately 11 x 10’ kcal/hr. Secondary air is preheated by
a tubular gas—air heater to a wind box air temperature of almost 3000C.
The compositions of the coals used for these tests are shown in
Table 2—4.
2.2.3 Test Results
The -results of tests conducted with Japanese coal A are shown in Figures
2—3 through 2—6. The Oz concentration in the flue gas was about 4 .
Figure 2—3 shows that NO decreases as the swirl intensity of secondary
air is reduced by the opening of the secondary air vane. Figure 2—4 shows
that NO 1 decreases sharply when the opening of the inner vane is reduced; it
reaches a minimum before the vane is entirely closed. Figure 2—5 indicates
that the tertiary air stream has a strong effect on NC . The increase in
fuel injection velocity assisted by tertiary air is assumed to be the cause of
the decrease in NO 1 emissions.
The primary air/coal ratio and the position of the inner sleeve influence
NO 1 emissions by changing the fuel injection velocity. The effect of the
injection velocity on NO 1 removal is shown in Figure 2—6.
39
-------
VAPOR
_0 0 0
00000000
Ni
OBSERVATION
PORT
Figure 2•2 Crosa sectIon View of Test Furnace.
OPENING ii
OBSERVATION
PORTS
OVER
AIR PORT
FURNACE
N.W L.
Q
0
FURNACE
00000
4 5m (3m width)
-=- --------
2.5m
I _
BURNER —
STAiRW _
— FLUE GAS
5
40
-------
TABLE 2-4. COMPOSITIONS OF COALS TESTED
T ’ sj Coal
B B C
A
C A
D
Kind of Coal
Proximate Aiialysis
Higher Heating Value
(kcal/kg)
6.200
6.600
-5.410
6,870
7.220
6.740
6.460
Inherent MoLsture • %
5.29
1.61
1.72
3.72
4.06
3.13
4.53
Volatile Matter, %
43.9
39.1
35.8
28.7
41.4
34.6
32.5
Ash, %
13.9
18.4
31.1
10.4
5.44
12.7
12.8
Fixed Carbon, %
36.9
40.9
31.3
57.2
49.1
49.6
50.2
Ultimate Analysis
62.9
66.6
52.7
72.8
71.8
68.0
65.2
C,
H, %
5.1
5.0
4.4
4.1
5.0
4.3
4.1
N, %
1.1
0.9
0.7
1.0
1.4
1.4
2.3
S, %
0.26
2.31
1.37
0.81
0.43
0.60
0.98
0, %
10.66
4.84
7.45
6.74
11.62
9.44
9.43
41
-------
450
400—
350—
300—
250
I I I
—20 —10 0
OPENING OF SECONDARY AIR VANE (mm)
Figure 2-3 Secondary Air Vane Opening vs. Nor.
POSITION OF INN ER SLEEVE
E
0.
c.1
0
Co
0
I-
w
I-
0
w
0
0
x
0
z
+ 75%
BASE OPENING POSITION
42
-------
450 —
E
0.
400—
c .1
0
Co
0
I-
350-
I-
0
w
0
0
300—
250
i I I -I I
0 10 20 30 40 50
INNER VANE OPENING (mm)
FIgure 2•4 Inner Vane Opening vs. NOR.
43
-------
300—
I I
0 20
I
I
I
40
60
80
OPENING OF TERTIARY AIR DAMPER (0)
FIgure 2-5 Opening of Tertiary Air Damper vs. NOR.
250—
200—
E
0
w
I—
0
w
0
0
x
0
z
44
-------
SYMBOLS
SLEEVE
POSITION
A
—100
0
—50
.
±0
0
+25
+75
V
I I I
10 15 2.0
RATIO OF FUEL INJECTION VELOCITY
Figure 2•6 Effect of Fuel Injection Velocity on NOR.
S \
450—
400—
350—
300—
E
c’J
0
Co
0
I-
0
LU
0
0
‘C
0
z
45
70A2540
-------
The effects of staged combustion and flue gas recirculation on the per-
formance of the low— 4O burner were also studied. Three over—fire air ports
were installed—one of them straight above the burner and the others on either
side of the air port. Figure 2—7 shows that NO was reduced substantially
by lowering the burner stoichiometry; it reached about 150 ppm at 0.8 stoi—
chiometry. The relationship between the nitrogen content of the coal and the
NO concentration at different burner stoichiometries is shown in Figure
2—8. By using a lower stoichiometry, most of the fuel NO can be reduced to
Ni which reduces NO 1 to a low level even with a high—nitrogen coal. Figure
2—9 shows the effect of flue gas recirculation on NO 1 concentration. By
using the combination of a low—NO 1 burner, staged combustion, and flue gas
recirculation, NO 1 was reduced to about 100 ppm, even with high—nitrogen
coal.
Commercial application of the burner is described in Section 2.6.3.
2.3 NO 1 ABATEMENT BY BH LOWNO 1 COAL BUBNER (4,5,6)
2.3.1 Primary Gas Dual Air Register Burner
This low—NO 1 burner is based on the dual air register burner developed
by Babcock and Wilcox of the U.S. with an improvement made by Babcock—Hitachi
K.K. (BHX). The major difference between this burner and the dual air regis-
ter burner is that the former has a primary gas (P0) port for injection of
recirculated flue gas to reduce NO 1 formation. Figure 2—10 shows the con-
struction of the burner. The nozzle for pulverized coal is located in the
center of the air register and allows injection of pulverized coal which is
mixed with air (primary air) and sent into the furnace. The pulverized coal
injection nozzle has two concentric cylindrical sleeves. The inner of the
cylindrical paths formed by them is called the primary gas port and the outer
is called the secondary air port. The cylindrical path formed by the outer
sleeve of the secondary air port and the burner throat is called the tertiary
air port.
46
-------
Q
JAPANESE COAL A
B
C
S
IMPORTED COAL A
0
B
C
D
500
400
300
200
(3.3 4%)
100
1.3 1.2 1.1 1.0 0.9 0.8
BURNER STOICHIOMETRY
NOTE: Figures in Parentheses indicate Exhaust Gas °2•
Figure 2.7 NO Reduction by Low•NO Burner and Two.staged Combustion.
a.
a.
0
Ca
0
w
I-
0
w
0
0
x
0
z
(3.6-3.8%)
*
(4.8%)
47
-------
a,
• 70
/
‘I
“7
7
APPROX. 1.0
,1
‘V
1.0 1.5 2.0 2.5
NITROGEN CONTENT IN FUEL (ASH AND MOISTURE FREE) (%)
NOTE. Symbols in This Figure are the same as those in Figure 2-7.
Figure 2-8 Nitrogen Content in Fuel vs. NOR.
BURNER STOICHIOMETRY
APPROX. 1.2
V
0
E
0.
0
0
I -.
U i
I-
0
U i
0
0
‘C
C
z
400
300
200
100
‘I
0.5
48
-------
300 -
BURNER STOICHIOMETRY
APPROX 1 2
200
(0
0
I-
A
w
100- APPROX.08
FLUE GAS MIXING RATIO IN SECONDARY AIR (%)
NOTE: Symbols in this Figure are the same as those in Figure 2-7.
Figure 2-9 NO Reduction by Flue Gas RecirculatiOn.
49
-------
TERTIARY
PRIMARY AIR AND
PULVERIZED COAL
PRIMARY
GAS VANE
SECONDARY
AIR REGISTER
U i
Q
C
*
TERTIARY
AIR
INLET OF PRIMARY
GAS OR COLD AIR
SECONDARY
AIR VANE
Figure 2.10 PG DualAIr Register Burner.
-------
Combustion air from the windbox is supplied to the secondary and tertiary
air ports. Each port is equipped at its inlet with louvre dampers, called
secondary and tertiary registers, for distributing the air flow among these
ports. Each of the nozzles accomodates a specific amount of combustion air.
In addition, the PG type burner uses a primary gas port to introduce flue gas
from the economizer outlet and slow the rapid mixing of the fuel—rich flame
with secondary air.
2.3.2 Test Results -
Results of tests with a 250 MW natural circulation coal—fired boiler are
shown in Figure 2—11. The A portion of the figure shows the relationship
between the two—stage combustion ratio (air through the burner throat/theo-
retical air) and the N0 concentrations for a conventional circular burner
and the PG dual air register burner. Assuming that the NO concentration
with the conventional burner without two-stage combustion is 100%, the reduc-
tion ratio was 31% with two-stage combustion, 39% with the use of the PG dual
air register burner, and 63% with the combination of two—stage combustion and
the PG burner.
Figure 2—11 B gives an example of the relationship between CO concentra-
tion and the two-stage combustion ratio. CO emissions increased up to the two—
stage combustion ratio of 87 and then decreased. The variation in two-stage
combustion ratios causes variation in the flow ratios of the secondary and
tertiary air. Because of this variation the injection speed of the combustion
air at the burner throat also varies. On the other hand, the relative speed
between the air side and fuel side at the burner throat varies because the
fuel injection speed (primary air) is constant. The concentration of CO seems
to increase at zero relative speed.
Figures 2—12 and 2—13 show the effects of Oz levels in flue gas and
boiler loads on NO 1 emission levels. Figure 2—14 illustrates the results of
using different coals in a test furnace as shown in Table 2—5. The applica-
tion of primary gas along with staged combustion and flue gas recirculation
51
-------
100
80
60.
40
20
0 ‘80 90 100 110 120
COMBUSTION AIR THROUGH BURNER
(%, RATIO TO THEORETICAL AIR)
300
200
100
I ’
V
I I I I I
(A)
LOAD = 100%
0
0
uJ
>
I
>.
I
I
I-
0
>0
LLJ
I-
0
0
-J
w
>
I
I
I
80 90 100 110
COMBUSTION AIR THROUGH BURNER
(%, RATIO TO THEORETICAL AIR)
(B)
(I )
Results of Two .Staged Combustion Test: NO and CO.
52
LOAD = 100%,
I
QQZ
Ooz
U J IUJ
, 0>
0z z
z 0
8
I-
U i
- I
00
z
0
0
U i
Figure 2.11
70A2596
-------
70 0
60
4
2.5 2.75 3.0 3.25 3.5 3.75
EXCESS 02 AT ECONOMIZER EXIT (%)
Figure 2.12
OVER-FIRE AIR
PORT POSITION
05% OPEN
G40% OPEN
75% OPEN
Effect of Flue Gas 02 Level on NO Emissions
In PG Dual Register Burner.
0
I-
z
0
C ,)
C,)
w
0
z
53
-------
I-
OQ
w-J
zo-<
OOz
(1)
z o
0
0
70
60
50
40
30
20
20
FIgure 2-13
OVER-FIRE AIR
PORT POSITION
O 5% OPEN
20°/o OPEN
O 100% OPEN
Boiler Load (%)
Effect of Boiler Load on NO Emissions
in PG Dual Register Burner.
40 60 80 100
54
-------
BURNER TYPE
COAL
NO OTHER
CONTROL
2 STAGE
OMBUSTIO
2 STAGE
+ GM
2 STAGE
+ GM + PG
DUAL AIR
REGISTER
JAPANESE
X
PG DUAL AIR
REGISTER
JAPANESE
OVERSEASA
OVERSEASB
C
A
C
G
£
•
6
U
4
S
LU
zz
uJ
>
Ow
U. —
ow
zo
2 STAGE COMBUSTION 85% BURNER STOICHIOMET1RY
GM = GAS MIXING (FLUE GAS RECIRCULATION)
PG = FLUE GAS (8 5%) FED THROUGH THE BURNER
100 X
50
25
C.
100
k
FLUE GAS RECIRCULATION (°/ )
FIgure 2.14 Performance 01 the PG Dual Register Burner In the Test Furnace
0
75
WITH PG
55
-------
TABLE 2-5. COMBUSTION FURNACE TEST CONDITIONS FOR FACTORY TEST
1. Furnace -
Type: Horizontal water—cooled cylinder
Size: Inner diameter 2.2m
Length 9. Oni
Combustion capacity:
max. 1 3 x1 0 ’ kcal/h
2. Type of burner
(1) Dual air register
(2) P.G. Dual air register
3. Type of coal
(1) Japanese coal: 1,000 kg/h
(2) Overseas coal A: 930 kg/h
(3) Overseas coal B: 870 kg/h
4. Air temperature: 300°C
5. Furnace outlet Oz:3.09,
6. Fuel analysis:
Type of Coal
Japanese Coal
Overseas
Coal
A
B
Proximate Analysis
kcal/kg
6,040
6,494
6,900
GCV (Dry Base)
Moisture
!o
1.90
2.04
2.60
Volatile matter
¶o
32.29
25.64
28.20
Fixed carbon
%
40.91
55.74
57.30
Ash
%
24.90
16.58
12.00
Ultimate Analysis
¶o
1.0
0.83
0.70
Total sulfur
Carbon
59.86
67.76
71.98
Hydrogen
4.08
3.53
4.12
Nitrogen
1.09
1.61
0.90
Oxygen
8.07
8.41
10.02
Sulfur
0.64
0.53
0.42
Fluorine
ppm
106
1,581
—
56
-------
reduced NO 1 to about 25% of the level obtained with conventional combustion.
CO was not i-ncreased by using primary gas.
The performance of the PG—type burner will be discussed in Section 2.6
(Figure 2—3 5)
2.4 I LOW—NO 1 COAL BUP N S (7,8,9)
2.4.1 Sei,arate Gas Recirculation Burner
Mitsubishi Heavy Industries U I) developed a low—NO 1 separate gas
recirculation (SGR) burner, which is used with the new 500 MW coal—fired
boiler at EPDC’s Matsushima Station. Recently Mlii further improved the SGR
burner to produce a PM coal burner.
By conducting extensive fundamental studies, M I II has found the following
conditions to be important for NO 1 abatement.
1) The temperature in the volatile matter combustion zone should
be kept as high as possible and the combustibles should stay in
this zone as long as practicable.
2) The temperature of the char combustion zone should be
appropriately low and combustion should be completed in this
zone.
Figure 2—15 compares the SGR burner with a conventional burner. The SGR
burner has several features to satisfy the conditions described above. A
flame holder is provided at the outlet of the coal nozzle to maintain a stable
ignition and expedite the combustion of volatile matter. The coal nozzle and
each auxiliary air nozzle are located apart from one another and SGR nozzles
are placed between them to retard the diffusion of secondary air to the vola-
tile matter combustion zone.
57
-------
r
EI! iJ
U
I
P
I
Ii
RE E
U I I
- I
_____ AUX.
____ ____ (OIL) _____
__ LL
__ __ SGRL
COAL
_________ SGR
AUX.
______ (OIL) _______
CONVENTIONAL BURNER SGR BURNER
(Oil Burners are for Start up)
—AIR
- FLUE GAS
________ COAL
AND AIR
FLUE GAS
— AIR
U’
Co
Figure 2•15 ConfIguratIons of Conventional and SOR Burners.
-------
Extensive combustion tests have been carried out with a test furnace
shown in Fig.ure 2—16 and described in Table 2—6, using coals shown in Table 2—
7. As shown in Figure 2—17, NO 1 concentrations were 30—50% lower with the
SGR burner than with a conventional burner. NOx abatement was more
pronounced with the Japanese coal which had a low fixed carbon (FC)/volatile
matter (VM) ratio. As will be shown in Figure 2—24, data obtained with the
test furnace agree well with the data from commercial boilers.
2.4.2 P t Burner for Pulverized Coal
Figure 2—18 indicates that the concentration falls as the primary
air/coal ratio (kg/kg) decreases from 3 to 1. Figure 2—19 shows that the
NO 1 concentration is also lowered when the ratio increases from 3 to 7.
As shown in Figure 2—20, a ratio of 3-4 approximately corresponds to the
theoretical air for volatile matter in coal. NO 1 concentrations are lowered
with the smaller ratio because of the deficiency of Oz. On the other hand,
the ratio 7—8 corresponds to the theoretical air for coal. With the ratio of
between 3—4 and 7—8, fixed carbon in coal burns with deficient Os to form
NO 1 . -
A conventional burner may produce a primary air/coal ratio of 2—3 as
shown in Figure 2—20. If the primary air—coal supply is divided into two
streams with different ratios———Ci for a coal—rich concentrated nozzle and C 2
for coal—lean weak nozzle—--—for separate burning, the average NO 1 value
should be (N0 1 )PM as shown in the figure. This is the weighted mean of
(N0 1 )Ci and (N0 1 )C 2 and is substantially lower than (N0 1 )Co. The com-
bined use of the concentrated flame and weak flame may result in low—NO 1
combustion.
Figure 2—21 shows the configuration of the Pollution Minimum (PM) burner
for use with coal. The performance of the PM burner in the test furnace is
shown in Figures 2—22 and 2—23. A low NO 1 concentration, about 100 ppm or
below, was achieved by the furnace.
59
-------
0
IDF
Figure 2•16 Flowsheet of the 4tihr Test Furnace
70A2601
-------
x
I
I
TABLE 2—6. MAIN FEATuRES OF MIII TEST FUItNACE (4 t/hr Test Furnace)
Equipment Technical details
Test furnace Horizontally installed, water cooled, circular—walied steel
construction. 4.4 diameter. 20 m lt iigth
F.D. fan Capacity: 92Om /min x 800z u.Aq at 20°C, Motor: 160kw
No. 1 J.D. fan Capacity: 1140m /niin 8OOi rnAq at 100 0 C , Motor: 330kw
Fan No. 2 I.D. fan Capacity: 200m’/min 32OmmAq at 100°C. Motor: 30 kw
Exhaust fan Capacity: 260&/min l00atniAq at 100°C. Motor: 90kw
G.R. fan Capacity: 160&/min x 600 mznAq at 350°C, Motor: ]b5kw
Air Heater Ljungstrom type. Capacity: 750Nm’Iniin x (25 350°C)
Gas No. 1 Gas cooler Water—cooled tubular type, Capacity: l5Oma’/min x (600 500°C)
Cooler No. 2 Gas cooler Water—cooled tubular type, Capacity: 200r.m’/min x (700 120°C)
No. 3 Gas cooler Water cooled tubular type, Capacity: 100nni /min x (700 350°C)
Ash Cyclone separator Multi—cyclone type. Capacity: 1000nxii /min x 500°C
disposal Bag filter Gas area 613m’, Capacity: ]200nm /min x 120°C
Precipitator Gas velocity: 1—1.5m/s, Capa .ity: 2500ni 3 /h
Fuel feed Coal pulverizer Bowl mill type, Capacity: 2300h 8 /h x 8000ni 3 /h x 600aunAg
assembly P.C. bin Capacity: 20m’
P.C. feeder Table feeder type, Capacity: Max 4500kg/h
-------
TABLE 2—7. ANALYSiS OF COALS USED )IiII Mlii TEST DUkNEIt
Proximate analysis Ultii ate Anulysis
( surface moisture free) — (dry base )
Volatile Fixed 11 1EV
Matter Carbon Ash FC/VM N 0 0/N (surface moisture free)
Coals Harks ( h) (%) (%) (—) ( ) (%) (—) kcal/kg
Japanese I I 42.8 40.2 11.2 0.94 1.1 13.9 11.0 6390
I 31 .6 46.0 20.5 1.46 1.0 6.6 6.6 6450
J 26.3 59.7 7.5 2.27 1.7 11.5 7.2 6710
K 35.7 50.6 10.1 1.42 1.4 11.0 7.86 6930
Australian L 28.6 50.6 17.8 1.77 1.5 8.1 5.40 6410
M 34.6 52.6 10.9 1.50 1.5 8.3 5.53 7120
N .)
N 30.8 54.1 14.1 1.77 1.1 5.8 5.3 6730
Chinese 0 27.4 60.6 8.8 2.21 0.86 9.8 10.2 7100
P 25.2 56.1 15.5 2.23 1.9 8.4 4.42 6520
South African 0 33.9 49.4 12.8 1.46 1.7 10.9 6.41 6540
R 24.1 57.6 15.2 2.39 1.5 8.1 6.95 6490
S 23.7 57.9 14.2 2.44 1.6 9.3 7.05 6350
U.S. T 36.9 47.0 9.9 1.27 1.1 13.4 12.2 6060
-------
Figure 2.17 ComparIson of NO Formation with SGR Burner
and Convenlional Burner.
70A2602
E
a
a
0
C l )
Co
(‘1
0
z
FUEL RATIO (FC/VM)
63
-------
PRIMARY AIR/COAL RATIO (kg/kg COAL)
Figure 2 18 Effect of Primary Stage Stoichuometry on NO (FueI•rlch side)
a’
E
0
In
( 5
ID
0
0
z
-------
PRIMARY AIR/COAL RATIO (kg/kg COAL)
Figure 2 19 EI(ect of Primary Stage Stoichiometry on NO (Fuel lean side)
U i
E
0
a,
U)
Cu
C D
0
0
z
-------
FIgure 2-20
PRIMARY AIR/COAL RATIO (kg!kg COAL)
V.M : Volatile Matter
Concept of Pulverized Coal-fired Low NO PM Burner.
E
0.
0
0
z
(NOr) C 0
a’
a’
0
C 1 C 0 3-”4 C 2 7vB
COAL
70A2605
-------
OIL __ ___________
_ —II
____ SGR
___ AUX2 ___ IjII
CONC
_ I )
a-’ ____ SOR ___________
_____ WEAK
____ AUX1 ________ I
OIL _____________________
FRONT VIEW CROSS-SECTIONAL SIDE VIEW
I I
! Ij- -I
SEPARATOR
FIgure 2-21 Structure of the CoaI.Ilred PM Burnor
-------
SYM8OL
0
A
H
0
P
o
(CHINESE)
(JAPANESE)
(SOUTH AFRICAN)
OFA%
—
—
SGR%
0
0
0
6M/.
0
0
0
EX0 2 /.
40
40
40
load %
100
100
100
—-0- -0—-
I
—
-—-0— 0—-
— — — -A- -
10 20
OFA (%)
E
a
a
U i
C l )
03
C.
0
0
z
U )
>-
-j
I L
z
z
0
0)
0
U I
z
0)
z
OFA (%)
0
Figure 2-22 NO and Unburned Carbon in Fly Ash vs. OFA. (PM Burner)
-------
E
0
w
0
0
(P
( . 4
0
0
z
I
(I )
-J
U-
z
z
0
00
C)
uJ
z
00
z
0
300
250
—
200
—
150
100
—
/ — 2
A
50—__-__
—
—
SYMBOL
Q
A
0
COAL
0
(CHINESE)
H
(JAPANESE)
P
(SOUTH AFRICAN)
OFA%
20
20
20
SGR%
0
0
0
GM%
0
0
0
EXO 2 /.
—
—
—
LOAO%
100
100
100
15
—
10
\
—
0
\
5
0
..
—
\ —
‘
+; T
0 2 6 8 10
EX 02(°fo)
0
2 4 6
8 10
EX 02(%)
Figure 2-23 NO and Unburned Carbon In Fly Ash vs. Ex 02. (PM Burner)
-------
Figure 2—24 compares the NO 1 concentrations for a Combustion Engineer-
ing (CE) conventional burner, SGR burner and PM burner when used for tangen-
tial—firing commercial boilers and the test furnace. Since NO 1 concentra-
tions for test boilers agree well with those for commercial boilers, it is
expected that with use of a commercial PM burner. NO 1 concentrations may be
reduced to 60—80 ppm with Japanese coals and 100—150 ppm with imported coals.
2.5 HI LOW—NO 1 COAL BURNER (11,12)
2.5.1 Structure and Combustion Model
Kawasaki Beavy Industries ( I) initially developed a low—NO 1 coal
burner for pulverized coal as shown in Figure 2—25. Later I improved the
burner design and developed a vortex diffuse (VD) burner as shown in Figure
., _ ,
KEl’s burners are designed to reduce NO 1 efficiently by using low—NO 1
burners in combination with staged combustion. The primary air is swirled
intensely; much larger amounts of secondary and tertiary air are introduced to
provide slow contacting with the coal and the partial combustion products.
A combustion model is shown in Figure 2—27. In the primary combustion
zone (I of the figure), volatile matter begins to burn, producing high concen-
trations of NO 1 due to the combustion of nitrogen compounds in the coal. In
the secondary zone (II), the volatile matter continues to burn, the char
begins to burn, and Os concentrations decrease, producing a reducing atmos-
phere which causes the decomposition of the NO 1 to N 2 . The reducing atmos-
phere also exists in the tertiary zone (III), where the NO 1 concentrations
reach a minimum just before the over—fire air port for staged combustion. The
CO concentration remains high.
70
-------
#1,2
I CECONVENTIALBURNER SGR I PM.....J
BURNER BURNER
(HOKKAIDO E P Co SUNAGAWA #3, EPDC
EPOC TAKASAGO #1,2, MATSUSHIMA
MATSUSHIMA #1,2)
Figure 2.24 Comparison of NO Emissions: Laboratory
Test Results vs. Field Data. (10)
70A2609
E
0.
0.
C
Co
z
0
w
C / )
x
C
z
0 F A (%)
71.
-------
SEPARATE TYPE
WIND BOX
PULVERIZED COAL
AND PRIMARY AIR —
EGlSTER VANE
TERTIARY THROAT
SECONDARY THROAT
PRIMARY THROAT
L’ - -EL
-4
OIL = - = I _i — S IRL R
BURNER GUN
SECONDARY AIR TERTIARY AIR
AND EXHAUST GAS AND EXHAUST GAS
FIgure 2 25 Structure 00 OrigInal KHI Low.N0 1 Burner
-------
TERTIARY AIR
SECONDA5RY AIR
+ EXHAUST GAS
PULVERIZED COAL
+ PRIMARY AIR
OIL BURNER
GUN
PILOT
WATER COOLING
JACKET
1’
FIgure 2.26 Slructure of VO Low .NO 1 Pulverized Coal Burner
70A2611
-------
TWO-STAGE
COMBUSTION AIR
SECONDARY AIR (A 2 )
+ EXHAUST GAS
TERTIARY AIR (A 3 )
+ EXHAUST GAS
0 ’
(III) A 2 + A 3 .s- CHAR
‘I
I.
(IV)TSC AIR +
REMAINING CHAR
PULVERIZED COAL +
PRIMARY AIR (A 1 )
I
—1
A 3
A 2
A 2 __ ——----_ .
(II) A 2 + REMAINING
VOLATILE
(I) A 1 + VOLATILE
4,
Figur. 2-27 Model of Low -NOn Putverlz.d Coal Combustion.
-------
Upon injection of the staged air (IV), the remaining char, CO and other
reducing gases are burned. However, no significant amounts of fuel NO 1 or
thermal NO 1 are formed during the complete combustion because: 1) virtually
all of the nitrogen compounds in the coal already have been decomposed, and
2) the combustion temperature is not very high.
Figure 2—28 shows the test results using the original burner with
Japanese coal at the 3t/hr test furnace. The figure clearly illustrates the
efficient NO 1 reduction of the burner when used in combination with staged
combustion. Flue gas recirculation (gas mixing) further reduced NO 1 concen-
trations.
Figure 2—29 compares the performance of the original burner with that of
the improved YD burner. In the ‘ ID burner, the effect of the coc bined use of
staged combustion and gas mixing is greater; a very low 10 x concentration,
below 100 ppm, was achieved. Although the CO concentration in the flue gas
increased appreciably, it remained in an acceptable range.
2.6 EPDC’S NO 1 ABATENENT BY COMBUSTION MODIFICATION (10)
2.6.1 Introduction
Electric Power Development Co. (EPDC) is a joint venture of the Japanese
Government and nine major power companies. One of its objectives is the con-
sumption of domestic coal. EPDC has three power stations which use domestic
coal and recently constructed a new station (Matsushima) to use imported
coals. The stations and applicable NO 1 regulations are shown in Table 2—8.
75
-------
200
ISYMBOLI TSC I GM I
I Q% 0%
10128 101
lot 1301
100
0
—&---- - A
0
500-
WITH OUT TSC
AND GM
0,
/
II
6 300 WITH TSC
E
0’ /
.9
/
200
z /
/
100 — WITH TSC
AND GM
0 I _____ I I t I I
10 11 12 13 14 15
I I I TOTAL AIR RATIO
0 2 6
02 (%)
TSC TWO STAGE COMBUSTION GM GAS MIXING
Figure 2•28 NO 0 EmIssion CharacteristIcs of OrigInal KHI Burner.
76
-------
/
12 13
I I
2 4
14 15
TOTAL AIR RATIO
6 02 (°/o)
Figure 2•29 Comparison of NO 0 Emissions in VD and Original Burners.
7OA 26 14
200
E
0
0
(0
00
II
C ’.
0.
E
*
0
z
SYM
TSC
GM
BURNER
0
00/0
0%
ORIGINAL
TYPE
33
30
A
0
0
VD
TYPE
_
34
100 —
—0
- —
0
—
500 — WITHOUT TSC
AND GM
iA”
- ‘A
400 —
Ii
/ /
300 —
200 —
/
/
0
100 —
9’ WITH TSC ÷ GM
—
0
10
11
I i I i ___________
0
77
-------
TABLE 2-8. EPDC’S POWER STATIONS AND NOx REGIJLATIONS
Power Station
Isogo
Takasago
Takehara Matsushima
Boiler No.
No.1 No.2
No.1 No.2
No.1
No.3 No.1 No.2
Capacity, MW
265 265
250 250
250
700 500 500
Year completed
1967 1969
1968 1969
1967
1983 1980 1981
Type of coal
Domestic
Domestic
Domestic
Imported Imported
NO 1 regulation, ppm
Emission standard
Local agreement
400
is?
400
300
400 b
112
400 b 400
60 300
Actual
150—180
Below 300
240—280
Under negotiation.
After the completion of the No. 3 boiler in 1982. S R is to be applied.
2.6.2 Tests of NO 1 Abatement by Combustion Modification
EPDC has conducted extensive tests of combustion modification for NO 1
abatement jointly with major boiler manufacturers including Mitsubishi Heavy
Industries (MIII), Babcock Hitachi, Ltd (BilK), Ishikawajima—Harima Heavy Indus-
tries (IHI) and Kawasaki Heavy Industries (Kill). Some of the test results are
shown in Figure 2—30. Those tests indicate that by using a combination of low—
NO 1 burners with staged combustion and flue gas recirculation, NO 1 can be
reduced by 60—7 ———to 100—150 ppm with low—nitrogen coal and to 150—200 ppm
with a nitrogen—rich coal. A test conducted jointly with a company C indi-
cated that flue gas recirculation was not effective for NO 1 abatement with
coal, and that unburned carbon increased with the application of staged
combustion.
The relationship between polycyclic organic matter (POM) concentrations
and NO 1 concentrations is shown in Figure 2—31. POM generally increased
with the decrease in NO 1 caused by combustion modification. Table 2—9 shows
the POM compounds and their relationship to combustion modification.
78
-------
A COMPANY
B COMPANY
C COMPANY
100
wZ
ZO
5
/
-I
z z
O Ui
—L ii Z Z —Ui Z
(3 Z (3 Z (3
+ + UJ +
(3 (3 Z C.) C) o (.
O Ui C )) (1) 0 ‘ii n C) 0 Ui (1) C))
o z — >- 0 Z i- I— 0 Z — I-
TSC TWO-STAGE COMBUSTION
GA GAS RECIRCULATION
Figure 2.30 Examples of Combustion Modification Test8 Conducted
by EPDC and Boiler Manufacturers.
E COAL
(N = 1 7 °/a)
DCOAL
(N = 09°/a)
ECOAL
“ (N = 1 7 °/a)
ECOAL
“ (N = 1 7 °/a)
0 COAL
0 (N = 09°/a)
400
300
200
0 COAL
‘‘ (N = 07%)
> ::
10
0, .
,
0”
-o
/
79
-------
8.0
70
6.0
5.0
4.0
3.0
2.0
100
AT COMBUSTION
MO DI FICATION
AT NORMAL
COM BUSTIO N
400
(NOX concentration in test furnace is the average value
of maximum and minimum values.)
Figure 2.31 NO Concentration vs. P.O.M. Emission.
70A2616
E
z
0
0
TEST FURNACE
200
300
NO (ppm)
80
-------
TABLE 2-9. MEASUREMRNT RESULTS OF POM AT FURNACES (Unit jig/Nm 3 )
Combustion
POM Condition
Component
Normal
Combustion
2 Stage
Combustion
2 Stage +
FG Mix +
Primary Gas
PGM
Partial
Load
Naphthalene
0.22
0.18
0.22
0.24
0.24
Fluoranthene
——
0.40
0.88
0.78
0.84
Pyrene
0.65
0.64
0.87
0.87
0.84
Benzo (a) pyrene
0.48
0.59
0.66
0.58
0.62
Total P01.1
(Percentage when
Normal combustion
is 100)
1.99
(100)
2.62
(132)
2.46
(124)
3.15
(158)
3.49
(175)
Although the POM concentration increased with combustion modification, it
reached a much lower level than the values reported in the U.S. POM was
measured using the EPA No. 5 sampling method, modified by Battelle, and a
liquid chromatography analyzer. It is not clear whether the difference in
values is due to the analytical method (GC—MS was used in the U.S.) or to
differences in the type of coal and the size of boiler used.
2.6.3 Combustion Modification at Isogo Power Station
The Isogo Power Station, located in Yokohama City (population 4 million)
was asked by the city to reduce NO 1 emissions as much as possible. The sta-
tion had no space to install SCR units because FGD plants had already been
retrofitted in the narrow spaces beside the boilers. Therefore, extensive
efforts have been made to reduce NO 1 by combustion modification.
The specifications of the Isogo boilers are given in Table 2—10. The
boilers burn domestic low—sulfur coal which is relatively low in nitrogen and
volatile matter and is suitable for NO 1 abatement (Table 2—11).
81
-------
TABLE 2—10. SPECIFICATIONS FOR IS000 265 MW PULVERIZE) COAL-FIREI) BOILER
Type: IBI—FW single drum, radiant type,
natural circulation, reheat boiler
(indoor service)
Evaporation (at t.1.C.R.): 840,000 kg/h
Steam pressure (at M.C.R.)
Superheater outlet; 176 kg/cci 2 g
Reheater outlet: 34 kg/cm2g
Steam temperature (at M.C.R.)
Superheater outlet 571°C
Reheater outlet 571°C
Fuel: Bituminous coal (equipped with 509 MCR
heavy oil firing system)
£‘umber of burners: 24 set (4 rows and 3 stages on boiler
front and rear walls)
Draft System: Balance draft system
The boilers were completed in 1967 and 1969, and were modified in 1973
with an over—fire air port system for NO abatement. As a second step to
meet the increasingly stringent NO regulation, an IHI—FV DF—CN burner
(Section 2.2) and a boundary air system (Figure 2—32) were applied to the
No. 1 boiler in 1976 and to the No. 2 boiler in 1977. The boundary air pre-
vents the melting of the ash. As a third step, burners out of service (BOOS)
and steam injection from oil guns were tested on the No. 2 boiler in 1978.
Eight additional over—fire air ports also were installed near the side walls
and a division wall, based on tests conducted by tEl. In addition, steam
injection was carried out through oil burner guns for the No. 2 boiler. These
modifications were made in the spring of 1979 to reduce N0 below 180 ppm
(at 6 Y Oz) under routine operating conditions. The same modifications were
made on the No. 1 boiler in 1980. The results of these modifications are
summarized in Table 2—12.
82
-------
TABLE 2-11. TYPICAL ANALYSIS OF COAL USED AT ISOGO STATION
Higher heating value 6,200 kcal/kg
Ash 16.5%
Volatile matter 39.6%
Fixed carbon 39.4%
Sulfur O.4
Hydrogen 5.5
Nitrogen 1.0%
TABLE
2-12. MODIFICATIONS MADE FOR INCREASED MOX CONTROL
AT
THE IS000 POWER
UNIT
NO 1 Emission Level by Achieved NO 1
Month,
Year Regulation, etc. (ppm) Level (ppm)
Modifications
Before
May 600 570
None
1973
June,
1973 600 380 — 510
Two—Stage
combus tion
April,
1977 480 240 — 250
Low—NO 1 burner
Curtain Air
April,
1979 400 170 — 190
More rigorous
Two— Stage
combustion, etc
April,
1980 159 a 170 — 190
aYokoh a City’s Requirement
83
-------
K
-
1c:
INAIR/
OVER AIR PORT
L NER
BURNER THROAT COOLING DEVICE
Figure 2-32
IHI’s Boundary Air System and Burner Throat Cooling Device.
-------
Figure 2—33 shows the relationship between the overf ire air port open
ratio and NO 1 emissions. With the 50% opening, NO 1 is reduced to 150—180
ppm. By increasing the size of the opening, NO 1 can be further lowered, but
the super—heater tube metal temperature becomes excessively high. To prevent
overheating of the superheater, methods of modifying the superheater layout.,
were tested; unfortunately test results were not very successful. Figure 2—34
shows NO 1 and 03 concentrations during the actual operation of the boiler.
NO 1 is maintained around 170 ppm at full load and 140 ppm at the reduced
load. Although the NO 1 concentration increased during the load change, it
seldom exceeded 200 ppm.
2.6.4 Matsnshiina Power Station
Two new 500 MW coal—fired boilers were completed at the Matsushima Sta-
tion in late 1980, and went into commercial operation in 1981. The boilers
were constructed by MIII using tangential—firing and SGR burners (Section
2.4.1). The specifications of the boilers are shown in Table 2—13. NO 1
concentrations of 200—290 ppm were attained during test operation using im-
ported coal containing 1.7% nitrogen. In commercial operation, NO 1 has been
kept below 290 ppm to meet the local agreement of 300 ppm.
TABLE 2—13. SPECIFICATIONS OF BOILERS AT MATSUSHIMA POWER STATION
Output, MW 500 x 2 unit
Fuel Coal (imported)
Boiler Manufacturer MIII
Type of Boiler Supercritical pressure reheat type
U.P. boiler
NO 1 (guaranteed), ppm 285 (Os = 6%) (N = 1.7%)
NO 1 (target) same as guaranteed
Burner type SGR burner (illustrated in Section 2.4.1)
85
-------
300 —
— ——
—
PREDICTED
50 100%
OAP DAMPER OPEN RATIO
Figure 2•33
Damper Open Ratio and NO Concentration. (Isogo)
70A2618
WITHOUT
STEAM
E
a
a
0
c c
0
w
0
w
0
0
0
z
STEAM
250
200 —
150 —
100 —
50 —
0—
ACTUAL
0
86
-------
DATE FEB. 18, ‘80
___________ ± E±: E E E ElE E E.E E
HI
ii i 4 j— :::::::: :::j:::__ __
______ • —
1 1° :
I :;;; ; : : _ _;4;
1
— I 1- —t- t— t
__________
— .---4 -
tt—t --..---.- -
265 MW LOAD 140 MW UP 265 MW
REDUCTiON TtME (O’CLOCK) —
Figur. 2.34 NO Conc•ntratiofl$ Durthg BolIsr Op.ratlon. (Isogo)
87
-------
2.6.5 Takehara Power Station
The Takehara Station has a 250 MW coal—fired boiler (No. 1) and a 350 MW
oil—fired boiler (No. 2) . In addition, EPDC has begun construction of the
No. 3 Unit, a 700 MW coal—fired boiler. Many NO 1 abatement techniques for
the coal—fired boilers are planned, including the application of SCR. Speci-
fications for the No. 3 boiler are shown in Table 2—14. The boiler is being
constructed by Babcock—Hitachi (BK ) and is scheduled for completion late in
1982; it will begin commercial operation in March 1983.
TABLE 2—14. SPECIFICATIONS OF NO. 3 BOILER OF TAKEEARA STATION
Output, MW
Fuel
Eoiler 1 anufacturer
SCR Manufacturer
Type of Boiler
NO 1 (guaranteed
at boiler), ppm
NO 1 (target at
boiler), ppm
NO 1 (guaranteed
at stack), ppm
Burner type
700 x 1 unit
Coal (imported coal)
BilK
BilK
B&W Supercritical pressure reheat type
UP boiler
250 (O 6%, N 1.8%)
200 (02 = 6%, N = 1.8%)
60 and below (02 = 6%, N = 1.8%)
PG dual burner
The Takehara Power Station uses PG dual—air register burners (Section
2.3). Figure 2—35 shows the relationship between the heat release rate at the
burner zone and NO 1 concentrations with three types of burners——conven-
tional circular—air register, dual—air register, and PG dual—air register——
with or without two—stage combustion and gas mining. The No. 1 boiler uses a
dual—air register burner and produces about 200 ppm NO 1 using two—stage com-
bustion, and about 300—360 ppm NO 1 without two—stage combustion. A test
furnace with a higher heat release rate produced about 400 ppm NO 1 with the
dual—air register burner and around 350 ppm NO 1 with the PG—type burner when
staged combustion was not applied; it produced 100—200 ppm NO: with the
88
-------
700
(I ,
U)
I .,’
0
E
-j
‘U
>
“a
0
z
200
100
0
HEAT RELEASE RATE AT BURNER ZONE (%)
CB: CIRCULAR AIR
REGISTER BURNER
DRB: DUAL AIR
REGISTER BURNER
PG -DRB: PG TYPE-ORB
TSC: TWO-STAGE COMBUSTION
GM: GAS MIXING TO
SECONDARY AIR
Figurs 2.35 H.at R.Isss. Rats at BoIIsr Zons. (%)
600
500
50 60 70 80 90 100 110 120 130 140
89
-------
PG—type burner with staged combustion and gas mixing. Since the heat release
rate of the No. 3 boiler falls between that of the No. 1 boiler and the test
furnace, a NO concentration about 150 ppm or below is expected.
2.7 LOV1—N0 1 BURNERS FOR GAS AND OIL FIRING
2.7.1 Introduction
Various types of of 10w—NO 1 burners for gas— and oil—fired boilers have
been developed and used in Japan. These burners may be classified into the
following categories according to the principles of NO 1 abatement employed.
(1) sigh—mixing
(2) Divided—flame
(3) Self—recirculating
(4) Staged—combustion
(5) Off—stoichiometric—combustion
(6) Water—injection
(7) Combination
This section describes the major low—NO 1 burners available for commercial
gas— and oil—firing applications.
2.7.2 High—Mixing Burner
A high—mixing burner developed by Nippon Furnace I ogyo aisha, Ltd. is
shown in Figure 2—36. It is called a NFK—TRW burner, because it was developed
by Nippon Furnace jointly with TRW Inc. of the U.S., under a licensing agree-
ment with Civiltech Corp. of the U.S.
NO 1 concentrations for the burner using oil under various operating
conditions are shown in Figure 2—37. This burner reduces the formation of
thermal NO 1 by shortening the residence time of the combustion gas in high—
temperature zones.
90
-------
ATOMIZING
AIR OR STEAM
Figure 2•36
NFKTRW Burner.
DEFLECTOR
ATOMIZING
AIR OR STEAM
COMBUSTION
AIR
91
-------
80 I
60 - 100% LOAD -
LOAD
20
0 I I
1.0 1.2 1.4 1.6
EXCESS AIR FACTOR
(Fuel: Grade A fuel oil; air temperature: 25°C;
maximum fuel flow rate: 380 llhr)
Figure 2.37 Effect of NFK•TRW Burner on NO Concentrations.
70A2551
92
-------
2.7.3 Divided—flame Burner
A divided—flame burner developed by IHI is shown in Figure 2—38. In this
burner the flame is divided into several small flames by means of grooved
nozzle tips. This reduces the flame temperature and residence time in high—
temperature zones, thus reducing thermal NO 1 (13).
2.7.4 Self—recirculating Burner
In the self—circulating burner, a portion of the hot combustion gas is
recirculated to the initial stage of combustion by the jet force of the fuel
and/or air injection. Thus vaporization and burning of the fuel oil in a low
Oi concentration are achieved. Various burners of this type have been pro—
duced by many manufacturers.
Figure 2—39 shows the configuration of the burner developed by Daido
Tokushuko (Special Steel) Co. The NO 1 concentration with this burner is
shown in Figure 2—40. The burner was originally developed for metal—heating
furnaces firing oil or gas. A similar burner called the SPP has been devel-
oped by the Volcano Company.
The burner shown in Figure 2—41 produces a circulating flow of combustion
gas in the precombustion chamber by means of orienting the air flow with in-
tense swirls to the axis of the atomizer. It is called a voitmetric burner
and was commercialized by Nippon S. T. Iohnson Shokai Co., Ltd.
Similar burners, designed for kerosene firing, are shown in Figures 2—42
and 2—43. The Nl’L burner was developed by Chugairo Kogyo, Ltd., and the YLAP
burner by Yokoi Kikai Kosakujo, Ltd. The NO 1 levels with the NPL burner are
shown in Figure 2—44. These burners use low air—pressure atomizers with
various burner tile configurations.
Figure 2—45 shows the ONK burner with a rotary atomizer, developed by
Sanrey Rejnetsu Co., Ltd. Combustion gas is recirculated by the secondary air
jet at a high velocity. The burner’s performance is shown in Figure 2—46.
93
-------
FUEL INJECTION PORT
DIRECTION OF
FUEL SPRAY
DETAILS OF NOZZLE TIP
Figure 238
GROOVES
FUEL INJECTION
PORT
NOZZLE TIP
FUEL
HI Divided Flame Burner.
94
-------
IGNITION PLUG
Daido Tokushuko Seif-recirculatlon Burner.
PILOT BURNER
FUEL
FLAME
COMBUSTION
AIR
Figure 2-39
95
-------
150
GUNTYPE >
I za:
c.’l
o
0
. 100 - ROTARY TYPE - 0
0
0
I-
LU OILPRESSURETYPE Z
0
50-
I
O ,0
o
STEAM ATOMIZING TYPE 0
-j
I I
0.9 1.0 11 1.2 1.3
EXCESS AIR FACTOR
(Fuel: Grade A oil; fuel flow rate: 100 l/hr)
Figure 2•40 Effect of SeIf.recirculatlofl Burner on NO Reduction.
96
-------
(A) ATOMIZER
AIR
(B) FLOW PATTERN IN PRE.COMBUSTION CHAMBER
Figure 2.41 SchematIc Diagram of the Volimetric Burner (Nippon S.T. Johnson).
STEAM c
FUEL
xl
97
-------
PILOT BURNER
ATOMIZER
FUEL
INLET
FIgure 2.42
NPL Burner (Chugairo Kogyo).
BURNER TiLE
4
4-
98
-------
X 1 - X 2 SECTION
Figure 243
YLAP Burner (Yokol KIkai Kosakujo).
FUEL INLET
-------
50
FURNACE TEMPERATURE (°C)
(Fuel: kerosene; oxygen concentration ,n the outlet of
test furnace: 3%, fuel flow rate: 15 I/hr 55 I/hr)
Figure 2.44 Effect of NPL Burner on NO Emissions.
E
0.
0.
0
0
w
I-
0
w
0
0
0
z
40
30 -
20
10
0—
800
551/hr
501/hr
301/hr
151/hr
900 1000 1100 1200
1300
70A2557
100
-------
SECONDARY AIR
COMBUSTION GAS
ATOMIZING AIR NOZZLE
PRIMARY AIR NOZZLE
Figure 245
ONR Burner (Sanrey Relnetsu).
SPINNING CUP
WIND BOX
AIR NOZZLE
101
-------
220 —
180 -
140-
100 -
60 -
20
50 75 100
BOILER LOAD (%)
(Fuel flow rate: 700-1000 kg/hr; excess air factor: 1 10-1.15;
‘combustion air temperature: 20°C)
FIgure 2.46 NO EmIssion Level of ONR Burner.
70A2560
E
0.
0.
0
0
I-
w
0
w
0
0
x
-0
z
GRADE C HEAVY FUEL OIL
KEROSENE
I I
25
102
-------
Figure 2—47 shows the RSNT burner designed by Rozai Kogyo Co. for oil and
gas firing. An intense swirling flow of secondary air causes the recircula—
tion of the combustion gas in the narrow space confined by the burner tile.
NO 1 emission levels by fuel type are shown in Figure 2—48.
Figure 2—49 shows the XB burner developed by Osaka Gas Co. jointly with
Sanrei Reinetsu Co. for gas—firing applications. Several fuel injection noz-
zles are arranged circularly to shorten the flame. A low NO 1 level 1 20—40
ppm, is achieved, as shown in Figure 2—50.
The High Speed burner, developed by Tokyo Gas Co., is shown in Figure
2—51. Flue gas is recirculated by the momentum of the fuel flow and a large
amount of combustion gas is entrained by the high—velocity gas flow.
2.7.5 Staged—Combustion Burner
The staged—combustion burner may be classified according to the composi-
tion of the first—stage flame into two types: fuel—rich and air—rich. The
fuel—rich burner is similar to the standard staged combustion burner. Both
fuel NO 1 and thermal NO 1 are lowered by fuel—rich combustion followed by
complete combustion at relatively low temperatures. With the air—rich burner,
the first stage has an excessive Oa concentration and additional fuel is in-
jected during the second stage.
Fuel—Rich Type
Figure 2—52 shows the APOC burner developed by Tokyo Gas Company for gas—
firing applications. In the pre—combustion chamber, an excess air factor of
0.5 to 0.7 is used. NO 1 concentrations in the flue gas after complete com-
bustion are about 35 ppm (corrected to 5T Oi).
Figures 2—53 and 2—54 show the FU burner developed by Chugairo Kogyo,
Ltd., and its performance. A high NO 1 reduction efficiency, 75—80” , can be
achieved with this burner when used with certain types of furnaces.
103
-------
AIR
Figure 2.47 RSNT Burner (Rozal Kogyo).
4-
FUEL
I ’
RECIRCULATION ZONE
SWIRLER
TILE
104
-------
SYMBOL
FUEL
kca l/Nm 3
•
BUTANE
27,000
•
GRADE A OIL
10,000 (N
=
0.06%)
0
BUTANE -BFG
8500
£
BUTANE-BFG
4600
A
BUTANE-BFG
3300
0 COG+BFG
2500
FURNACE TEMPERATURE (°C)
Figure 2-48
NO Emission Levels of the RSNT Burner. (Thermal input: 2.4 x 10 b kcal Ih;
excess air factor 1.1; combustion air temperature: 250-350°C).
E
0 .
0.
(‘1
0
r
r
0
I-
0
w
I-
0
w
II
0
0
x
0
z
0
U i
100
50
20
10
900 1000 1100 1200 1300 1400
-------
R EC I RCU LATED
COMBUSTION GAS
RECIRCULATED
COMBUSTION GAS
Figure 2.49
XB Burner (Osaka Gas).
106
-------
I I
2 4 6 8
THERMAL IN PUT (kcai/hr)
Figure 2.50
Effect of XB Burner on NO Emission Levels.
I I — I I
8
E
0
I ’ ,
0
I-
a
LU
I-
0
U i
0
0
0
z
80
60 -
40 -
20 -
0
-6
CITY GAS FIRING, N0
-4
C l ,
L U
-J
LL
z
z
0
I-
I-
z
LU.
0
z
2 0
0
0
CITY GAS FIRING, 02
0
_0
10 x io 5
107
-------
PILOT BURNER &
Figure 2.51 HIgh Speed Burner (Tokyo Gas).
FUEL
PEEPING
SECONDARY AIR
FLUE GAS
I
AIR
108
-------
FIgure 2-52 APOC Burner (Tokyo Gas).
//
PRE-COMBUSTION CHAMBER
PRIMARY
AIR
AIR
109
-------
PRIMARY
Figure 2.53
Schematic of FH Burner (Chugairo Kogyo).
AIR
FUEL
GAS
110
-------
EXCESS AIR FACTOR
FIGURE 2.54
Effect of the FH Burner on NO Reduction.
1400
1350
1300
I I
0..
I I
0
0
uJw
I-
LLQ
w
E
0
•1
0
I-
z
0
I-
0
w
‘C
0
z
250
200
150
100
50 -
0
I I
— — —
CONVENTIONAL BURNER
L
-.
NO REDUCTION RATIO
100 -
80-
60 -
I
I
0 1
2 3
02 CONCENTRATION
IN
FLUE GAS (¼)
I I
I I
1.00
111
-------
Figure 2—55 shows the MR burner developed by MEl for package boilers with
a capacity of up to 30 t/hr. In addition to producing the effect of staged
combustion, this burner also reduces NO 1 by cooling the flame with an ad-
jacent water tube. The NO 1 levels guaranteed by MIII are shown in Table 2—
15.
The TS burner developed by Osaka Gas Co. and its performance are shown in
Figures 2—56 and 2—57.
TABLE 2—15. NOX EMISSION LEVELS GUARANTEED WITE TEE MR BURNER FOR SMALL
SCALE BOILERS
Fuel Nitrogen Content in Fuel (%) NO 1 Emission Level (ppm)
Grade C
Oil
0.20
150 a
Grade B
Grade A
Oil
Oil
0.14
0.03
130 a
690 a
Kerosene
Gas
0.01
—
a
6070 b
50
Corrected to 4% 02.
Corrected to 5% 02.
Air—Rich Type
Figure 2—58 shows the TGS burner developed by the Tokyo Gas Company. The
primary fuel gas is premixed with a large amount of excess air. The rest of
the gas is then added to the flame. The burner shows an appreciable NO 1
reduction as illustrated in Figure 2—59.
Figure 2—60 shows the Ti burner developed by the Daido Sanso Company. In
the primary combustion chamber, 40—60% of the fuel is fired with a large
amount of excess air——a factor of 1.6 to 2.4——and the secondary fuel in-
jected into the furnace is burned with an 02 concentration of 7—14%. The
overall excess air ratio is kept between 1.05 and 1.2. The performance of the
burner is shown in Figure 2—61.
112
-------
SECONDARY
N REDUCTION
REG ON
PRE.REACTION REGION
OF SPRAY
DIRECTION OF
SECONDARY SPRAY
FIgure 2.55 Schematic of MR Burner (Mitsubishi Heavy industries).
113
-------
SECONDARY
AIR NOZZLE
PRIMARY AIR NOZZLE
FUEL NOZZLE
GAS
Figure 2•56
TS Burner.
AIR
PEEP
HOLE
114
-------
- 130
- 120
1 l0
100
90
E
0.
0.
0
I
80
70
60
50
40
30
20
10
0
E
a.
a
c.’I
0
C ,)
C ’,
Lu
0
x
U i
13
12
11
10
9
8
7
6
5
4
3
2
1
0
I-.
0
FURNACE TEMPERATURE
1 2 3 4 5 6 7
PERIOD AFTER IGNITION (hr)
(Fuel: city gas; calorific value: 5000 kcal/Nm 3 ;
number of burners: 6 units)
Figure 2•57 NO Emission Level of the TS Burner in a Batch
Forging Furnace with a Capacity of 20 t/charge.
1400
1300
1200
1100
1000
900
800
700
600
500
400
0
a
Lu
I-
Lu
a.
Lu
I.-
Lu
0
z
U.
02
300
200
100 -
0
Ox
115
-------
FUEL GAS PREMIXED
WITH AIR
Figure 2.58
Schematic of TGS Burner.
FUEL GAS ———
116
-------
FURNACE TEMPERATURE (°C)
(Fuel: city gas; calorific value: 5000 kcal/Nm 3 ; number
of burners 4 units; thermal input: 2.4 x i0 5 kcal/hr)
Figure 2•59 NO Reduction Effects of TGS Burner in a Batch
Forging Furnace with a Capacity of 2t/charge.
70A2573
E
0.
0
0
700
800 900 1000 1100 1200 1300 1400
117
-------
SECONDARY FUEL
PRIMARY FUEL
Figure 260
Schematic of TZ Burner (Daldo Sanso).
118
-------
800 1000 1200 1400
FURNACE TEMPERATURE (°C)
(Total emission factor: 1.1, thermal input: 1.0 x - 1.5 x kcallhr)
Figure 2.61 NO Emission Levels with the 12 Burner.
E
c .
c i.
C)
70A2575
119
-------
2.7 .6 Off—Stoichimetric—Combust ion Burners
This type of burner first produces fuel—rich and fuel—lean flames both of
which have low NO 1 concentrations. The two flames are then mixed for com-
plete combustion.
Figure 2—62 shows the DLS burner developed by Osaka Gas Company for gas—
firing metal—heating furnaces. Its performance is shown in Figure 2—63. Com-
pared with the TS burner (Figures 2—56 and 2—57), the fuel gas in the DLS
burner is burned with much less excess Oi, and NO 1 concentrations are almost
constant during the operation period.
Figures 2—64 and 2—65 show the improved smokeless PM burner developed by
!llI, and its performance. P.ecirculated flue gas is injected between the air—
rich and fuel—rich flames to retard the mixing of the flames, thus causing
greater NO 1 reduction. Particulates, as well as NO 1 , can be reduced by
this burner.
Off—stoichiometric combustion can be attained by a special arrangement of
the fuel -injection holes in a burner. Figure 2—66 shows such an arrangement
in the burner developed by the Volcano Company. The larger holes produce the
fuel—rich flame while the smaller holes produce the fuel—lean flame.
Figure 2—67 shows the I—type throat in the dual—air register burner de-
veloped by Babcock Bitachi, Ltd. (Section 2.3). From 25—40% of the combustion
air is fed through the grooved parts while the rest is fed through the cir—
cular part to achieve the off—stoichiometric combustion. The effect of the
I—type throat using the off—stoichiocietric—combustion type atomizer with an
OSC tip is shown in Figure 2—68.
2.7.7 Water—Injection Burner
Water—injection burners have been used extensively in Japan because they
reduce N0 by lowering the flame temperature and, to some extent, produce
effects similar to emulsified oil (Section 2.8.4).
120
-------
NOZZLE
I-a
nrH r
GAS
Figur. 242 Sch.mallc of tb. DLS Bumr (Osaka Gas).
AIR
-------
E 100
0.
C.
0
0
i-60
IL l
I-
0
U i
0
20
0
z
Figure 2.83 NO Emission Levels with the DLS Burner In a Batch
Forging Furnace with a Capacity of 2ticharge.
0 1 2 3 4 5
PERIOD AFTER IGNITION (hr)
(Fuel city gas, caloritlc value 5000 kcalfNrn 3 )
6 7 8
TEMeERp TU
FURNP C
1400
1200
1000
800
600
400
200
0
LU
I
I-
I
LU
LU
I-
LU
0
z
I
U.
0
C I ,
LU
0 1
LU
122
-------
AIR-RICH FLAME
AIR-RICH FLAME
— UPPER COMPARTMENT
RECIRCULATED FLUE GAS
MIDDLE COMPARTMENT
RECIRCULATED FLUE GAS
LOWER COMPARTMENT
FUEL-RICH FLAME
Figure 2-64 Smokeless PM Burner (MHI).
-------
E
0.
C ’
0
n
0
I- .
0
LU
I-
0
LU
I
0
0
0
z
EXCESS 02 (%)
C,
E
z
E
LU
I-
-J
0
I - .
I
0.
(Fuel grade B heavy fuel oil, nitrogen content 0 13°/o. fire air
about 5°/a, recirculated flue gas including flue-gas recirculatiOn 100/0)
Figure 2-65 NO 0 and Particulate Reduction with the Smokeless PM Burner.
200
:;j ’— -
100 —1-
90 - f
80 1
70 —
60 1
—
5C —
S.
ibU
100
90
80
70
50
50
40
30
n
40—
30
n_
—
-
1
CONVENTIONAL BURNER
• NO,
o PARTICULATE
SMOKELESS PM BURNER
o NO 0
• PARTICULATE
2
3
124
-------
OIL
Figure 2.66 AtomizIng Nozzles In the Off.etolchlor
Low.NO Burner for Oil (Volcano Co.).
STEAM
(A)
(B)
70A2580
125
-------
Figure 2.67
-Type Throat In the Dual-Air Register Burner.
126
70A2581
-------
O STANDARD THROAT AND TIP
0 I.TYPE THROAT AND OSC TIP
(Fuel. grade C oil, thermal input 3 4 x 1O kcal/hr)
Figure 2.68 NO Reduction of the OSC Burner with I.Type
Throat In a Watewr Tube Boiler (15 t!hr).
E
0 .
0.
0
0
I-
w
0
w
0
0
x
0
z
0 25
RATIO OF PRIMARY AIR TO TOTAL AIR (%)
50
70A2582
127
-------
Burner tips for water injection, developed by Asahi Chemical Industry
jointly with Nippon Plant Engineering Co.. are shown in Figure 2—69. Oil and
water are ejected with steam in different ways with various types of tips.
The type—C tip generally gives the best results, reducing both NO 1 and par—
ticulates as shown in Figure 2—70. Similar types of tips also are used by
Osaka Gas Co.
The JSR burner, developed by the Iapan Synthetic Rubber Company, injects
water and oil from separate holes as shown in Figure 2—71. NO 1 reduction
with the burner is shown in Figure 2—72. The NO reduction is caused not
only by the lowering of the flame temperature which reduces thermal NON, but
also by the atmosphere produced by the reaction of water and oil which reduces
the fuel NO 1 .
The UN burner, shown in Figure 2—73, injects water into the atomizing
air pressurized at 1 to 2 kg/cm 3 . The relationship between the water—injected
burner and NO 1 emissions is shown in Figure 2—74. In addition, water injec-
tion reduces soot formation and thus enables low excess air combustion, some-
times resulting in energy Savings.
2.7.8 Combination Burner
There are several low—NO 1 burners which use a combination of the above
NO 1 reduction principles. For example, the SRG burner (Figure 2—75). devel-
oped by Nippon Furnace Kogyo Kaisha, Ltd., applies two—stage combustion and
self recirculation of flue gas. The self recirculation is attained by the
Coanda effect of the high—velocity combustion gas. These burners have been
used mainly for oil— and gas—fired furnaces at chemical plants and ref in—
eries. NO 1 levels produced by the SRG burner, in comparison with conven-
tional Tandem burners, are shown in Figure 2—76.
The CVS Voitmetric burner, shown in Figure 2—77, was developed by
Chugairo ogyo and is based on a similar NO 1 reduction principle.
128
-------
©
©
0
0
0
WATER
OIL
(A) STEAM - I
OIL
WATER
STEAM
OIL
(B) WATER -
OIL
STEAM
WATER —.
OIL
(C) STEAM
OIL
WATER
WATER
STEAM _____
(D) OIL
STEAM
- WATER
WATER
OIL
(E) STEAM
OIL
WATER
Figure 2-69 Atomizer Tips end Distribution of Water PartIcles.
STEAM
01
WATER
OIL
STEAM
OIL
STEAM —
OIL ’
WATER
INTERMEDIATE MIXING TYPE (A), (C), (D) AND (E)
INTERNAL MIXING TYPE (B)
129
-------
CONVENTIONAL TIP
WATER INJECTION TIP OF (C)
E E
z
E
z z
2 0
I-
I .- I-
z z
w uJ
0 0
z z
O 0
0 0
uJ U i
I-
_J
o 0
—
3. 3.
WATER -OIL RATIO (%) VOLUME
E NO LEVEL OF CONVENTIONAL TIP WITHOUT
NO REDUCTION TECHNIQUE
Q OFF.STOICHIOMETRIC COMBUSTION
TWO-STAGE COMBUSTION
O OFF-STOICHIOMETRIC COMBUSTION AND
TWO-STAGE COMBUSTION
FIGURES IN THE SYMBOLS INDICATE THE
EXCESS 02 CONCENTRATIONS
(Fuel grade Coil, fuel flow rate 11 Un, number of burners 9 units, 7 0A2584
Firing type front firing, combustion air temperature 115C)
FIgure 2.70 NO 1 Reduction by Water-Inlection Atomizer lip in a Boiler with a Capacity of 162 tftr
E
0.
0.
C ’
0
0
I-
Ui
I-.
0
Ui
0
0
0
z
E
0.
0.
N
0
0
Ui
I-
0
Ui
0
0
0
z
130
-------
OIL SPRAY
STEAM
WA
SPRAY
Figure 2.71 Schematic of JSR Burner (Japan Synthetic Rubber).
131
-------
E
0.
0.
N
0
0
0
0
I-
0
U i
0
Ui
0
0
0
z
(Fuel: grade C oil; firing type: front firing)
Figure 2.72 N0 Reduction with JSR Burner in a Boiler
with a Capacity of 130 f/hr.
70A2586
EXCESS 02 (%)
132
-------
x i - X 2 SECTION
ATOMIZING
Flgur. 2-73 Sch.matlc of UN Bumr.
133
-------
EXCESS AIR FACTOR
Figure 2.74 NO Emission Levels with UN Burner.
E
C.
a
C M
C
0
O
LU
I-
0
LU
0
0
‘C
0
2
1.0 1.2 1.4 1.6
1.8
134
-------
BURNER TILE
AIR
AND
TERTIARY AIR
PRIMARY AIR
Figure 2.75 SchematIc of SRG Burner. (Nippon Furnace Kogyo Kaisha )
70A2589
TERTIARY AIR
FUEL
135
-------
NO 0 VS AIR RATIO SPG BURNER
AND TANDEM BURNER
I I
AIR RATIO, ¼
Figure 2.76 Effect of SAG Burner on NO Reduction.
B FUEL OIL, 105 2/hr
0
200
- r 150
00
z
100
50
11 12
13
AIR RATIO, )
14 15
z
0
00
0
z
136
-------
PRIMARY
COMBUSTION CHAMBER
COMBUSTION AIR ( -
AIR REGISTER
ATOMIZER
PILOT BURNER
FIgure 2.77 SchematIc of CVS Voltmetric Burner (Chugalro Kogyo).
137
-------
Figure 2—78 shows the TCG burner developed by Osaka Gas Co. jointly with
Rjrakawa Tekko Co. Self recirculation of combustion gas is achieved by the
high—velocity secondary fuel, while off—stoichiometric combustion is attained
by separate injection of fuels.
The ELN burner developed by flosoyama Neki Co. is shown in Figure 2—79.
The burner applies flue gas recirculation and two—stage combustion. NO 1
levels with the burner are shown in Figure 2—80.
2 .8 COMBUSTION MODIFICATION FOR SPECIAL NO 1 SOURCES AND FUELS
2.8.1 Refuse Incinerators (14 )
Flue gas from refuse incinerators usually contains about 150 ppm NO 1
(corrected to l2 Oz). A large portion (70—80%) of the NOX is derived from
nitrogen compounds in the refuse. The conversion ratio of fuel nitrogen to
NOx is considered to be 8—12% under normal incineration conditions. In a
low excess air incinerator, a conversion ratio of 8% was reported.
NO 1 conversion characteristics for refuse incineration are similar to
those of oil and coal burning for the following reasons:
(1) Furnace temperatures do not affect the conversion ratio
appreciably.
(2) Types of nitrogen compounds in the refuse do not greatly
influence the conversion ratio.
(3) Generally, the larger the nitrogen content, the smaller the
conversion ratio. The conversion ratio, however, essentially
stays constant above a certain nitrogen content.
(4) The conversion ratio increases with the excess air factor.
Based on these characteristics, staged combustion is considered most
effective for abatement of NO 1 abatement in incinerator flue gas. Reducing
gases such as NB 3 , hydrocarbons, and HCN, fored at the first stage in the
138
-------
RECI RCULATED
COMBUSTION GAS
Figure 2-78
AIR
Schematic of TCG Burner (Osaka GaslHlrakawa Tekko).
GAS FOR
IGNITION
GAS
PRIMARY AND
SECONDARY AIR
TERTIARY
139
-------
— RECIRCULATED GAS
- L PRIMARY AIR CONTAINING
1 RECIRCULATED COMBUSTION GAS
— -4- - GAS
SECONDARY AIR
ih
AIR
Figure 2-79 SchematIc of HLN Burner (Hosoyama Nekfl.
140
-------
60 -
50
o
40
0
0
I-
30 1.5
0
o 1.4 <
w
EXCESS I. 1.3
o A
20 I, Z R
1.2 ci
0
U)
x
o 10-
z 1.1
_______________ 1.0
0 I I I I I I I I
0 50 100
BOILER LOAD (%)
(FueI city gas; calorif Ic value: 5000 kcallh)
Figure 2-80 NO Emission Level with HLN Burner in a Boiler
with a Capacity of 1.3 t!hr.
70A2594
141
-------
reducing atmosphere, decompose NO 1 and keep the NO 1 conversion ratio as
low as 5%.
Problems with staged combustion and modifications to remedy them are
shown below:
Problems Modifications
Increase of unburned carbon Increase of stoker area
Application of after—burning system
Control of furnace temperature Flue gas recirculation
Water injection
Steam injection
By making these modifications to two—stage combustion, NO 1 emissions
below 120 ppm (corrected to 12% Os) are achieved, as shown in Figure 2—81.
This emission level more than meets the emission standard of 250 ppm NO 1
(corrected to 12% Oa) for a new incinerator.
2.8.2 Cement Kilus (15)
Rotary kilns for cement production usually produce a flue gas containing
a high concentration of N0 1 —400--1,000 ppm—--even vith oil firing. This is
due to the long retention time of the gas at temperatures of 1000 1550°C and
the presence of a large excess of air. A new NSP kiln with suspension pre—
heaters and a gas generator (GG) or a flash furnace (FF) (Figures 2—82 and
2—83) has been developed and used in Japan. It has a high production efui—
ciency and produces NO 1 emissions below 200 ppm; the emission standard for
new kilns is 250 ppm (corrected to 10% 02).
A conventional rotary kiln uses all of the fuel from a burner located at
one end of the kiln to produce a wide range of high temperatures (1000—
1550°C). This assures the decomposition of calcium carbonate and the reaction
of lime with clay and other raw materials to form the cement clinker. With
the new types of kilns equipped with GG or FF systems, one half of the fuel is
injected into the gas generator or flash furnace (calciner). The furnace is
installed at the opposite end of the kiln to assure calcination of the feed
142
-------
150 200
o
250 300 350
BURNING RATE ON MAIN STOKER (kg/m 2 hr)
(Refuse feed rate 150 I/24hr)
FIgure 2.81 NO Reduction by Two.Staged CombustIon In a Refuso Incinerator.
200 F
0
150
c’1
0
I .-
0
w
Q
Lu
100
0
0
0
WITHOUT
TSC
/
0
0
0 A WITH
TSC
I
-J-
400
70A2544
143
-------
‘---- FLUE GAS
COMBUSTION
CHAMBER
DENITRIFICATION
CHAMBER
— — — -, AIR AND FLUE GAS FLOW
FUEL
CEMENT
CLINKER
Figure 2.82 Flow Diagram of the GG Process Used In Cement Kilns.
a . FEED FLOW
COOLER
70A2545
144
-------
and decomposition of the calcium carbonate at around 1,0000C. NO 1 formation
in the kiln is substantially lowered by the reduced heat load. With the GG
system, it is possible to use the reducing gas to remove NO 1 formed in the
denitrification chamber. The reducing gas is burned by the hot air leaving
the cement clinker cooler. NO 1 concentrations can be reduced below 150 ppm
(corrected to 1O9 Oi), easily meeting the emission standard.
Use of coal instead of oil with the GG system does not cause a substan-
tial increase in NO 1 since NO 1 can be removed in the denitrification cham-
ber by the reducing gas.
Mitsui Mining Co. has modified the FF system originally developed for oil
firing for use with coal, as shown in Figure 2—83. The major modifications
are as follows (16):
(1) A larger calciner (furnace) is used to ensure complete coal
combustion.
(2) A low—pressure—drop suspension preheater is used to deal with
the increase in gas volume caused by the use of coal rather
than oil.
(3) The feed is introduced into the calciner from both the bottom
and the side to ensure better calcination and the correct com-
bustion temperature for NO 1 reduction.
2.8.3 Glass Melting - - -
Glass melting furnaces burning gas or oil used to emit high levels of
NO 1 (above 500 ppm) due to: 1) high furnace temperatures of approximately
1,600°C, 2) relatively long gas—retention times needed to ensure the melting
and clarification of the glass, and 3) the high Oz concentration of the gas.
When sodium nitrate is used as a feed to promote clarification, the NO 1 con-
centration in the flue gas commonly exceeds 2,000 ppm. The use of nitrate,
however, has almost been eliminatd in Japan in order to reduce NO 1 emis-
sions.
145
-------
—* -FEED FLOW
AIR FROM
COOLER
Figure 2-83 The FF System Used In Both Oil and Coal-f Ired Cement Kline.
IR AND FLUE
— GAS FLOW
CALCINER.
AIR FROM ______
COOLER
OIL FIRING
COAL FIRING
146
-------
In furnaces not using nitrate, virtually all of the NO 1 present in the
flue gas is thermal NO 1 . Using the following measures, this thermal NO 1
is reduced in much the same manner as it is in other oil— or gas—fired indus-
trial furnaces.
(1) Suppression of excessively high temperatures
(2) Prevention of air leakage
(3) Application of low excess air combustion
(4) Strict control of 03 in flue gas using oxygen sensors
A mild flame with a slow reaction is considered an effective method of
NO 1 abatement for these furnaces. For oil—fired furnaces, the atomizer has
been changed from an air to an oil—pressure type. An oil—pressure atomizer
that produces a plane flame shape has been developed by the Asahi Glass
Company £ or NO 1 abatement purposes (17). The plane flame shape is reduced
by the slit in the atomizer tip. With the combination of this atomizer and an
adequately shaped port, a 40% NO 1 reduction can be achieved.
2.8.4 Use of Emulsified Oil for Small Boilers
Emulsified oil (oil—water emulsion) has been used as an efficient method
of NO 1 abatement for small boilers. In some cases emulsified oil also may
be effective for saving energy because it allows a reduction in excess air
without causing incomplete combustion.
Many companies have developed systems for producing emulsified oil.
These systems may be divided into two categories, as shown in Figure 2—84.
The line blending system generally is more advantageous than the service tank
system because of its simplicity. On the other hand, the service tank system
is compatible with many types of combustion equipment, although pump cavita-
tion may be an operational problem with this system. Different types of
emulsifying agents, in concentrations of 0.1 to 0.2 , are used with various
types of boiler fuels.
147
-------
OIL
LINE BLENDING SYSTEM
EMULSIFYING AGENT
WATER
M- MIXER
T TANK
F FURNACE SYSTEMS
Figure 2.84
Emulsified Oil Production Systems.
EMULSIFYING AGENT
WATER
OIL
SERVICE-TANK SYSTEM
P PUMP
148
-------
Upon combustion of the emulsified oil, the atomized oil droplets are sep-
arated into very small particles by sudden expansion of the water contained in
the emulsion. This increases the contact area for the oil and air, thus in-
creasing the combustion rate, and reducing thermal NO 1 formation. Since
water and steam can be used to slow the formation of soot as shown in Figure
2—85, excess air can be further reduced as shown in Figure 2—86 (18). There-
fore, an increase in thermal efficiency may be expected when using emulsified
oil for small boilers. The reducing atmosphere formed by the reaction of the
water and fuel promotes the reduction of NO 1 . Injection, agitation, and
static mixers all have been used to produce emulsified oil.
2.8.5 Special Fuels and Processes
Coal Oil Mixture (CON )
COM is a slurry composed of equal amounts of oil and pulverized coal. In
Yapan, EPDC and major boiler manufacturers have conducted studies of COM, be-
cause the slurry can reduce oil consumption without the handling problems
inherent in coal use. The combustion properties of COM compared with those of
oil are &s follows.
(1) The ignition characteristics of CON are similar to those of
oil, but the atomized particles are often finer.
(2) With CON the flame becomes longer; oil burns initially,
followed by the combustion of volatile matter, and
then by the combustion of the fixed carbon of the coal.
(3) The flame temperature with CON—firing is about equal to that
for oil firing.
(4) The CON burning rate is slower which means that a larger amount
of excess air is required. Because of this, the NO 1 concen-
tration becomes higher.
Because of COil’s higher nitrogen content compared to oil, staged combus-
tion is more effective than flue gas recirculation for NO 1 abatement pur-
poses. NO 1 concentrations below 200 ppm are achieved by staged combustion.
149
-------
0.4
GRADE C
OIL
GRADE B
OIL
BLEND OIL
OF B & C
I
SPECIFIC GRAVITY OF OIL
Figure 2-85
Reduclion of Soot Emissions by Emusllfled Oil.
• OIL ONLY
o EMULSIFIED
OIL
C.,
E
z
E
z
0
C ’,
C ’ ,
LU
I-
0
0
C’,
0.3
0.2
0.1
0
GRADE A
OIL
0.80 0.85 0.90 0.95
150
-------
150
OIL
/
/
/
/
/
/ X_EMIJLSIFIED
0
OIL
0
I
100
0
I-
U i
I-
0
U i
0
0
‘C
0
z
50
01 I I I I
0 2 4 6 8 10
EXCESS °2 (°/o)
(Fuel grade C oil, water mixing ratio 30%)
Figure 2.86 NO Reduction By Emulsified Oil in a Water Tube Boiler
with a Capacity of 12 tlh.
151
-------
Methanol
A feasibility study on the use of methanol by utility boilers has been
conducted (19) . Combustion tests using a small test furnace produced NO 1
emissions below 30 ppm under normal conditions and of about 20 ppm when flue
gas recirculation was applied.
Catalytic Combustion
Basic studies on catalytic combustion of gaseous fuels also have been
conducted. Catalytic combustion has the following advantages.
(1) Low—temperature combustion is possible, resulting in low NO 1
concentrations.
(2) A wide range of excess air factors can be used without
affecting the flame stability.
(3) Uniform temperature and composition of the combustion gas can
be obtained thus avoiding local overheating and formation of
high NO 1 concentrations which occur during conventional corn—
.bustion.
One test indicated that NO 1 levels below 20 ppm were attained even with
a combustion temperature of 1,6000C and a catalyst. A large volumetric heat
release rate was produced without forming unburned species such as CO and
hydrocarbons (20).
One disadvantage of catalytic combustion is that the catalyst can be phys-
ically unstable if subjected to repeated temperature variations.
Fluidized Bed Combustion
Fluidized bed combustion (FBC) of coal has the following advantages:
(1) Low SO 1 emissions can be achieved by adding limestone or
dolomite in the combustion zone.
152
-------
(2) Low NO 1 emissions may be achieved because of the low combus-
tion temperature (around l,0000C).
(3) Smaller boiler sizes are possible due to a higher heat transfer
efficiency and a higher heat release rate in the furnace.
(4) Low—grade as well as high—grade coals may be burned.
The Coal Mining Research Center, jointly with EPDC, Kill, and Babcock
Hitachi, constructed a 5 MW equivalent atmospheric FBC pilot plant at EPDC’s
Wakamatsu Station. The plant began operation in April 1981 (21). Preliminary
tests using domestic low—sulfur coal (0.2% S and 0.9% N) and limestone indi-
cated that when staged combustion was applied, NO 1 was reduced to 64 ppm.
Construction of a 75 MW demonstration plant will begin in 1983; the plant
should be operating by 1985.
A small commercial atmospheric FBC plant owned by the Sumitomo Colliery
Akabira Company (10 t/hr steam, 2.5MW equivalent) began operation in April
1980 (22). The furnace burns a low—grade coal (3,050 kcal/kg) recovered from
coal—wash water and uses silica sand as the fluidizing agent. The plant, con-
structed by Babcock Hitachi, is a modification of an oil pulverized coal—fired
boiler. NO 1 concentrations in the flue gas have ranged from 260—350 ppm.
However, fluidized bed combustion has the following disadvantages in
Japan:
(1) There is no land space for disposal of ash containing calcium
sulfate, lime, and other materials.
(2) Although tests have been conducted on a process to separate and
regenerate the SO 1 absorbent, there are both technical and
economic problems associated with the process.
(3) SO 1 and NO 1 concentrations in the flue gas may not be low
enough to meet the stringent local regulations. In Japan, both
FGD and SCR have been applied successfully to coal—fired
boilers to reduce SO 1 and NO 1 to very low levels.
153
-------
RffERE 4CES
1. Ikebe, 11. NO 1 Reduction by Advanced Coal Boiler, The U.S.—Iapan
Exchange of No 1 Control Technical Information Conference, Tokyo, May
1981.
2. Ikebe, U., et J,. NO 1 Reduction for Pulverized Coal Boiler by IKI—
F! Type Advanced Burner, The Thermal and Nuclear Power, (30):7, 1979, p.
27 (in Japanese).
3. Ikebe, H. Combustion Technologies for Coal Firing, The Energy Conserva-
tion, (33):3, 1981, p. 14 (in Japanese).
4. Babcock Hitachi, Ltd. Combustion System to Coal—Firing Boiler, The U.S.—
Japan Exchange of NO 1 Control Technical Information Conference, Tokyo.
May 1981.
5. Masuko, S., J. Low—NO 1 PG Burner for Pulverized Coal Firing,
The Thermal and Nuclear Power, (29):6, 1978, p. 29 (in Japanese).
6. Masuko, S., et 1. Combustion System for Coal Firing Boiler,
Hitachi Review , (62):4, 1980, P. 21 (in Japanese).
7. Mitsubishi Heavy Industries. Development of Pulverized Coal—Fired Low—
NO 1 PM Burner, The U.S.—Japan Exchange of NO 1 Control Technical
Information Conference, Tokyo, May 1981.
8. Kuniimoto, T., et al. Design of 500 MW Coal Firing Boiler for
Matsushima Thermal Power Plant, Mitsubishi Juko Giho, (17):2, 1980, p.
72, (in Japanese).
9. Sato, S.,, et al. 500 MW Coal Fired Units in Matsushima Power Sta-
tion, The Thermal and Nuder Power, (30):ll, 1979, p. 33, (in Japanese).
154
-------
10. Mouri, K. and Y. Nakabayashi. Japanese Technical Development for Combus-
tion NO Control, EPA/EPRI Joint NO Symposium, September 1980.
11. Kawasaki Heavy Industries. Summary of Low—NOx Combustion Technique for
Various Fuels Developed by EBI, The tJ.S.—Japan Exchange of NO Control
Technical Information Conference, Tokyo, May 1981.
12. Shinano, T., Development of Low—NOx Pulverized Combustion
Techniques, K.H.I. Technical Review, No. 76, 1980, p. 80 (in Japanese).
13. Ando, I. N0 Abatement for Stationary Sources in Japan, EPA—600/7—79—
205, U.S. Environmental Protection Agency, Research Triangle Park, NC,
August 1979.
14. Report on NO 1 Emission from Refuse Incinerators, Environment Agency,
March 1980 (in Japanese).
15. Shimosato, S., et al. Development of the GG Process for Reducing
NO 1 Emission in Cement Plants, Mitsubishi Juko Giho, (14):5, 1977, p.
54 (in Japanese).
16. Hirao, M. Use of Coal in Cement Industries, The Energy Conservation,
(33):3, 1981, p. 37 (in Japanese).
17. Official Bulletin of Patent, 1978—6313 (in Japanese).
18. Ihara, H. Production and Combustion of Emulsified Fuel Oil, The Energy
Conservation, (33):4, 1981, p. 22 (In Japanese).
19. Hoshizawa, K. Study on Methanol Combustion — A Technical Study of Metha-
nol Combustion for Power Plants, The Thermal and Nuclear Power, (28):4,
1977, p. 37.
155
-------
20. Fukuzawa, H., et A Survey of Characteristics and Thermal NO
Reduction of Catalytic Combustion, The Thermal and Nuclear Power, (33):3,
1981, p. 39.
21. Tamanuki, S., H. Katayama and N. Takami. 20th Pilot Plant of Fluidized
Bed Boiler, Third Coal Utilization Technology Symposium sponsored by Coal
Mining Research Center, September 1981 (in Japanese).
22. Nisawa, M., and H. Terada. Utilization of Sludge Coal as a Fuel for
Fluidjzed Bed Boiler, Ibid, September 1981 (in Japanese).
156
-------
SECTION 3
SELECTIVE CATALYTIC REDUCTION (SCR) OF NO 1
3 .1 REDUCTION OF NO 1 BY AMMONIA
3.1.1 Basic Reactions
Virtually all of the NO 1 in combustion gases is present in the form of
NO, while NO 1 from non—combustion sources, e.g., nitric acid plant tail gas,
consists of both NOi and NO. Although NO 1 can be reduced to Na with the use
of reducing gases such as H 4 , H , and CO, these gases are also readily con-
sumed by reaction with Oa. N113 is a more economical reducing agent for a gas
with a high 02/NO 1 ratio, because NB3 selectively reacts with NO 1 .
Sinde a small amount of Oa promotes the reaction of NRa with NO 1 , the
reaction is usually expressed by equations 1 and 2.
4NRa+4N0+Oz 4N 2 +6HaO (1)
4NR 3 +2NO +O a=3Na+6HzO (2)
Combustion gases usually contain at least one percent Oa, which is sufficient
for the reaction (Figure 3—1).
The optimum temperature for the reactions without a catalyst is 900—
1000°C. Above this temperature, a considerable portion of the NUa is con-
verted to NO 1 . Below this temperature, the reaction rate as slow. Selec-
tive noncatalytic reduction (SNR), the injection of N113 into a boiler or
furnace at 800_10000C, has been applied on a commercial scale, but NO 1
removal is usually limited to 30—5O at N113/NO mole ratios of 1—2 (Section
6.2).
157
-------
‘00 ____________________
1 . 05%
250 300 350 400 450
CATALYST TEMPERATURE C)
(SV 10000 hr - NO. 100 opm NH IN0 I 0)
P1gw. 3-I Effect of Oxygen on NO. Removal I)
Selective catalytic reduction (S R) uses a catalyst at 150_4500C and
achieves 80—90% NO 1 reduction with an NUs/NO mole ratio in the range of 0.85—
1. Various catalysts are effective in the temperature range of 300—400°C.
which is the temperature of utility boiler economizer outlet gas.
3.1.2 Problems with SCR
The serious problems encountered in the early stages of S R development
have largely been solved. Catalyst poisoning by SO 1 was solved by substi-
tuting base metal catalysts based on TiOs for AlzOs or FesOs. Catalyst
plugging problems in dust—laden gases were eliminated with the use of honey-
comb, plate, and tube (“parallel flow”) catalyst configurations or parallel
flow reactors. Soot blowing is also used to remove dust. Ammonium bisulfate
formation occurs when NB3, SOs, and HsO react due to temperatures below 300°C
caused by reduced boiler loads. Bisulfate deposits lower catalytic activity.
Ammonium bisulfate formation has been prevented by maintaining a gas
temperature above 3000C, either by adjustments in economizer efficiency or
with a hot gas bypass upstream of the economizer. Economizer energy losses
are recovered with an air preheater downstream of the SCR. Ammonium bisulfate
tends to deposit in the air preheater. The deposits can be prevented or mini-
mized by maintenance of low levels (<5 ppm) of NH 3 in the reactor outlet gas.
158
-------
Oxidation of SOz to S03 also occurs with some catalysts, increasing the forma-
tion of ammoniijm bisulfate. Low—oxidation catalysts have been developed to
address this problem. Finally, catalyst erosion by fly ash particles has been
prevented by maintaining moderate gas velocities, increasing catalyst hard-
ness, and employing the “dummy spacers” described in Section 3.4.
These improvements have increased the reliability and ease of operation
of SIR processes, resulting in the construction of over 150 commercial plants.
Most have been operated with few problems, although ammonium bisulfate deposi-
tion in the air preheater continues to occur at some plants.
Catalyst life in commercial SCR systems varies depending on boiler fuel.
For gas— and oil—fired boilers virtually no degradation has been evident after
more than 4 years of operation. In July 1982 coal—fired boilers had operated
for over 2 years with little catalyst degradation. The actual life of S R
catalysts may be as long as 6—7 years for gas—fired boilers, 5 years for oil—
fired boilers, and 3 years for coal—fired boilers.
Disposal of the spent catalysts is considered a potential problem because
of the presence of heavy metals. Virtually no catalysts had required re-
placement at the time this report was written, so there is little experience
with spent catalyst disposal. Since catalyst composition is confidential,
producers must recycle all of the spent catalysts to recover the heavy metals.
For example, titanium producers are expected to recover TiOz. The feasibility
of these recovery practices for exported catalysts is uncertain.
3.1.3 Maior Factors in Catalytic Reaction
Equation 4 shows the relationship between space velocity (SV) and NOx
removal rate:
log(l—x) = —kISV (4a)
where:
k = koSap (4b)
x = Reaction ratio (N0 removal ratio).
159
-------
k Reaction constant.
= A constant at a given temperature relating to the catalyst
property.
8 ap Apparent surface area. The surface area of the catalyst assuming
that the catalyst surface is smooth, not porous.
Space velocity is defined as the ratio of the gas volume passing through the
catalyst bed per hour to the volume of the catalyst itself.
A smaller SV and a larger k give a larger x. A smaller SV implies that a
larger amount of catalyst has been used. The reaction constant, k, is larger
with a larger k 0 and Sap The constant, k 0 , is larger when the catalyst
material is more active, while Sap is larger with a more closely spaced cata-
lyst. The latter, however, not only produces a larger pressure drop but makes
the catalyst more susceptible to dust plugging (Figure 3—2). Therefore,
usually a granular catalyst with a 3—6 mm diameter is used for clean gas while
a honeycomb configuration with a 5—8 mm channel diameter and a parallel plate
with a 4—7 mm clearance are used for dusty gas.
The inlet NR3/N0 1 mole ratio also influences the NO 1 removal effi-
ciency as shown in Figure 3—3. A larger mole ratio produces a higher effi-
ciency and a larger amount of unreacted ammonia leaving the reactor. In
practice, a mole ratio between 0.83 and 1.0 is generally used to attain 80—90%
NO 1 removal and maintain N113 emissions below 5 ppm.
For a catalyst with a given SV, NH3/N0 1 mole ratio and tempera-
ture, the NO 1 removal efficiency as well as ammonia reaction ratio changes
little with the inlet NO concentration. This indicates that both outlet
NO 1 and Nib concentrations increase with inlet NO 1 concentration. There-
fore. in order to maintain both NO removal efficiency and unreacted Nil 3 at
a constant level, a large amount of catalyst (smaller SV) is needed with an
increase in inlet NO 1 (Figure 3—4). The relationship between NO removal
efficiency and catalyst requirements at a constant inlet NO 1 concentration
and unreacted Nil 3 level is shown in Figure 3—5.
160
-------
100
0
80
I ::
20
0 2 4 6 8 10 12 4 16 18 20
DIAMETER OF GRANULAR CATALYST mm)
Figure 3•2 Catalyst Size vs Dust Deposition for Oil Fired Dust
Hitachi Ltd)
NH 3 INO MOLE RATIO
(honeycomb Catalyst low sulfur Oil inlet NO. 150 ppm
Figure 3 3 SCR for Oil Fired boiler Flue Gas at 350CC
I I
80
>
0
70
E
z
I-
0
U i
z
161
-------
3 -
z
a
I-
C l,
).
-j
I-
Q
LL.
0
0
z
a
Cl )
3.
-J
Q
U.
0
0
INLET NO CONCENTRATiON (ppm)
Figure 3.4 NO 1 ConcentTatlon vs. Catalyst Quantity. (2)
DE•N0 1 EFFICIENCY (°Jo)
Figure 3.5 Os-NO 1 Efficiency vs. Catalyst Quantity. (2)
DE -N0 1 EFFICIENCY CONSTANT
LEAKAGE NH 3 CONSTANT
75
50
(BASE)
I
C
I
100 200 300
o 500
INLET NO, 1 CONCENTRATiON CONSTANT
LEAKAGE NH 3 . CONSTANT
150
100
50
(BASE)
17
80 70 80 90 100
162
-------
3.2 COMPOSITION AND PROPERTIES OF SCR CATALYSTS
3.2.1 History
S R catalysts have been produced by many manufacturers. The four major
manufacturers are Catalyst and Chemicals, Inc., NGK Insulators Co., Sakai
Chemical, and Nihon Shokubai Kagaku which produce various types of ceramic
catalysts with honeycomb, ring, tube, and pellet configurations. StR process
developers such MHI, ml, KHI, JGC Corp., and Kurabo have conducted their own
studies on S R catalysts and have provided manufacturers with specifications
for use in catalyst production. Other process developers produce catalysts
themselves; Hitachi Ltd. and Hitachi Zosen make metal—based catalysts while
Sumitomo Chemical and Lobe Steel manufacture ceramic ones.
At an early stage in SQ development, catalysts based on alumina carriers
were used predominantly. At that time, &—AlaOs was popular as a heat resis-
tant carrier due to its large surface area and high catalytic activity. The
relationships between the surface areas of various carriers are shown below
(1):
6—A1203 > TiOs > ZrO > MgO ) a—A1203 > SiOi
Catalysts based on ferric oxide also have been used. Ferric oxide is
catalytically active and can be obtained both as a natural mineral and as a by-
product.
A problem associated with both alumina— and ferric oxide—based catalysts
is that they have been found to be poisoned by SO 1 in the flue gas. This
occurs because of the formation of sulfate which tends to plug the small pores
of the catalyst and reduce its surface area and activity. Although the sul-
fate can be decomposed by heating it to 700_8000C, the treatment requires a
substantial amount of energy.
163
-------
More recently, titania (TiOs)—based catalysts have become popular for
dirty gas treatment because of their high activity and resistance to SO 1 .
The relative SO 1 resistances of the catalyst carriers are shown below (1)
TiOa > SiOs > a—AlsOi ) a—AlsO
The properties of various catalyst carriers are compared in Table 3—1.
TABLE 3-1. COMPARISON OF CATALYST’ S CARRIER CHARACTERISTICS (KAWASAKI
HEAVY INDUSTRIES)
Characteristics
Material
Alumina
Fern c
Oxides
Silicates
Tuff
Titania
SO 1 resistivity
C
D
A
B
A
Initial activity
for DeNO 1
A
C
C
B
A
Physical strength
B
D
A
A
C
Durability
C
D
C
B
A
Cost of catalyst
C
B
C
A
D
Note: Symbols A — D indicate ratings with A being the best.
Adding VzOs to TiOs produces a highly active S R catalyst. However, a
drawback is that vanadium promotes the oxidation of SO 2 to S03 which causes
ammonium bisulfate deposits in the air preheater. Recent efforts to improve
SCR catalysts have concentrated on development of an active catalyst which is
not poisoned by SO 1 and has little SOz oxidation capability.
Low—temperature catalysts active at 150_2500C also have been evaluated
for use with low—temperature gases such as those from coke ovens and sin—
tering machines. The greatest problem with the low—temperature catalyst is
that ammonium bisulfate deposits on it. In practice, the deposits are removed
164
-------
by periodically heating the catalyst above 400°C. This heating process is
necessary to restore catalytic activity.
Ammonia decomposition catalysts were studied and used a few years ago but
have not been developed. This type of catalyst tends to be poisoned by S0
and to oxidize a considerable portion of S02 to S03. In recent practice,
unreacted NR3 has been reduced by using a low NE3/N0 1 mole ratio and an
increased amount of SQt catalyst.
3.2.2 Alumina—Based Catalysts
Most alumina carrier catalysts are in the form of &—Al 2 03 which has a
large surface area and high activity. This type of catalyst has been used
with clean gas containing little or no SO 1 . The performance of alumina—
based catalysts for clean gases is shown in Figure 3—6. A catalyst made of
alumina impregnated with Cr203 shows a high activity around 2500C, while one
with Fea0a shows a high activity around 4300C. A catalyst active at 300—400°C
can be obtained by using both Cr203 and Fe203 or ViOs and Cri0i.
The aiumina in & form readily reacts with S03 to form Alz(S04)3 with
reduced surface area and activity (see Figure 3—7). Even though a gas may
contain SOs and no SOi, a portion of SOs is oxidized to S03 on the catalyst
surface and reacts with A1 2 03. When the poisoned carrier is heated to 8000C,
Ala(S04)3 decomposes to regenerate A1203 and catalytic activity is restored.
These reactions are:
A1203 ÷ 3S03 = Alz(S04)3 (1)
A1203 + 3S0 2 + 3/202 = Alz(S04)3 (2)
800°C
Al2(S04)3 ———) Als0 ÷ 3S02 + 3/2 02 (3)
However, the heating requires a considerable amount of energy and is not
economical, Heating &—AlsOa to 1100°C, produces cz—Al203 which is resistant to
SO 1 . However, a—A1203 has a much smaller surface area and is less active
than &AlsOi. Since S0 1 —resistant catalysts based on TiOs have become popu-
lar, alumina—based catalysts now are seldom used for gas containing SO x.
165
-------
100
‘
8O- /
D
F
LU
&- I
0 BI l
z
A
I
z
B .
U i
20
______________ 10 -
I •3 0
200 300 400 500
REACTiON TEMPERATURE (C)
k Cr 2 0 3 -A1 2 0 3 D: Fe 2 O 3 AI 2 O 3
B: Pt•A1 2 0 3 E. Fe 2 0 3 .C 3 0 3 -A1 2 0 3
C. MeO .Ai 2 O 3 P V 2 0 5 -Cr 2 0 3 .A 1 2 0 3
Figure 3-6 Cilteria for Catalysts Used with Clean Ga.. (1)
160
N
4
UJ 120
4
LU
0
BC
LL
LL 40
LU
U, ______________________________
C 002 004 006 008 ala
SULFUR DEPOSIT (g)ICATALYST (g)
Figure 3-7 DeactIvation of Catalyst on — A 1 2 0 , Carrier by SO . (1)
166
-------
3.2.3 FesOs—Based Catalysts
Fe303—based ring—shaped catalysts were used several years ago at a few
S R plants but they were found to be poisoned by SOx. Nippon Kokan recently
began using a special type of iron ore as the S R catalyst to treat flue gas
from an iron sintering machine. Contaminated by SOx and dust, the catalyst.
along with other iron ores, is used for steel production. Catalyst and Chess—
icals, Inc. recently studied the production of a low—cost Fe203 catalyst using
red mud, a useless by—product of aluminum production, and lirnonite, a neutral
mineral mined in the Aso and Kurohime Mounta ins (3).
The chemical composition of the red mud and lissonite catalyst is shown
in Table 3—2. The red mud’s major component is highly crystalline a—Fe103;
limonite is composed of a—FeOOH and a—Fe203 with low crystallinity.
TABLE 3-2. CHEMICAL COMPOSITION OF RED MUD AND LIMONITE CATALYSTS (3)
(%, on dry basis)
Red Mud Limonite
FesOs 52—55 84—85
AlsO s 20—27 13
SiOs 7—10 13—14
TiOs 7—10 Trace
CaO 0—2.5 0.1
MgO — 0.15
NasO 3—5 0.3
K O Trace 0.3
SO4 0.3 0.3
MnO Trace 0.4
(HiO) About 80 30—40
During the first stage of catalyst production. the red mud is washed with
sulfuric acid and water to remove sodium. The red mud and limonite then are
combined with water, kneaded, shaped and calcined to produce pellets 3 mm in
diameter. The catalyst pellets (25 ml) are placed in a reactor tube. Then a
synthetic gas containing 500 ppm NO, 500 ppm SO s, 3% 03, 10 — 15% HaO, with
167
-------
600 or 750 ppm NH3 is passed through the catalyst bed at 200 380°C with an SV
of 10,000 or 6,000 hr .
Figure 3—8 shows the effect of the catalyst’s calcination temperature on
its initial activity (before SOx poisoning). The limonite was highly active
when calcined at 400°C and became much less so when calcination took place at
higher temperatures. Since SCR usually takes place at 300—400°C, calcination
of the catalyst at 400—500°C may be adequate to produce good activity and
sufficient physical strength. The red mud catalyst was not very active when
calcined at 400°C.
The effects of adding various compounds to the catalysts are shown in
Figure 3—9. Red mud’s activity was increased by the addition of ViOs and CuO
but decreased when SnOs was added. The optimtlm amount of VaOs was 5—10% of
the amount of the red mud. TiOz increased the activity of limonite.
Figure 3—10 shows the increase in SO. and decrease in N0 removal
efficiency which resulted when three types of catalysts were exposed to flue
gas from an oil—fired boiler. After varying exposure periods up to 1,200
hours, the- catalysts were subjected to both an S04 analysis and an N0
removal test. Limonite with TiOs calcined at 500°C showed a high initial
efficiency which decreased rapidly during exposure (with the concurrent
increase in SO 4 ). Virtually all of the Fe203 was converted to Fez(S04)3 after
200 hours exposure. Linonite calcined at 700°C with TiOi was inactive al-
though it became more active with an exposure of 200 to 600 hours. Red mud
calcined at 500°C was moderately active and remained that way during a period
of 1.200 hours.
Figure 3—11 shows the results of the exposure tests on red mud catalysts
which contained additives of TiO or TiOs and VzOs. The improved red mud
catalysts maintained fairly good activity even after 3,000 hours of exposure.
This indicates that it may be possible to produce a relatively inexpensive
SOxresistant catalyst using red mud.
168
-------
80
-j
>
0
uJ
0
z
400
500
600
CALCINATION TEMPERATURE (°C)
(350°C, SV 6,000 hr 1, NO 500 ppm, NH 3 700 ppm)
Figure 3.8 Calclnation Temperature vs Initial ActIvity. (3)
(SV 10.000 hr l)
70A2699
A PEDMUD9O ,V 2 0 5 10
B REDMUD9O,CuOlO
C RED MUD 90, MoO 3 10
D RED MUD
E RED MUD 90, Sn0 2 10
F LIMONITE
G LIMONITE 70, 1102 30
(CALCINED AT 500C)
Figure 3-9 Effect of Additives on NO Removal.
40
-J
‘C
>
0
w
0
z
TEMPERATURE (°C)
1O 27OO
169
-------
40
Figure 3.10 Catalyst Tests with Oll.Flred BoIler Flue Gas. (3)
60
40
a
I
U.’
20
0
0
350C
0 1 Q00 2000 3000
TEST PERIOD rnr
o RED MUD 80 1102 20 (5mm PELLET CALCINED AT 500°C
o RED MUD 89 7102 10 V 2 0 5 1(5mm PELLET CALCINJED AT 500CC)
A RED MUD (6mm PELLET CALCINED AT 500C(
FIgure 3.11 Additional Catalyst Tests with Oll.Fired Boiler Flue Gas (3)
400 600
TEST PERIOD (hr)
o c LIMONITE 70, 710230 (CALCINED AT 500C)
A A LIMONITE 70. Ti0 2 30 (CALCINED AT 700c)
0 0 RED MUD (CALCINED AT 500C)
-j
>
0
U.’
0
z
-J
>
0
U.’
I
0
z
170
-------
3.2.4 TiOs—Based Catalysts
Titania (Ti02), in the form of anathase, is catalytically active and has
a good resistance to SOS. Its activity can be increased by adding base
metal oxides. As shown in Figures 3—12 and 3—13, Vanadium CV) is a most
effective additive for this purpose. For this reason, a TiOs—based catalyst
containing 1—5% V 205 has been most popular for gas containing SOS. However,
a common problem associated with this catalyst is the oxidation of a portion
of the SO 2 to SO3. During cooling, the SOs combines with NEs and H 2 O to form
NH 4 HSO 4 which results in the plugging of the air preheater elements. The
oxidation ratio is larger with a lower SV and a smaller Nils/NOx mole ratio;
N E 3 works as a reducing agent to lower the oxidation ratio (Figure 3—14).
When Cu, Co, and Ce are used as additives to the TiOs—based catalyst, a
substantial amount of oxidation occurs; when V is used with Mo and W only a
slight amount of oxidation occurs.
Catalyst producers are making efforts to lower the oxidation ratio. A
common method is to replace a considerable portion of V with other metals such
as Mo or W (Figure 3—15). With the replacement, the NO removal efficiency
is slightly lowered while the oxidation ratio is lowered to a much greater
extent. Table 3—3 shows the TiOz—VsOs catalyst (108—H) and improved catalysts
(109 H — ill HC).
The TiOs—VzOs catalyst is poisoned by alkali compounds, particularly by
potassium at low temperatures (Figure 3—16). Figure 3—17 shows that a cata-
lyst using two kinds of selected base metals is poisoned to a lesser degree.
3 .3 LOW-TEMPERATURE CATALYSTS
3.3.1 Introduction
For some gas sources, such as coke ovens and iron—ore sintering machines,
the temperature of the gas stream requiring NOx removal is low (150—250°C).
Since heating the gas to 300—400°C requires a large amount of energy, low—
171
-------
>
0
U i
I
0
z
— I
>
0
LU
I
0
z
“ i i-
50
0
200
300
TEMPERATURE (C)
400
(metal oxide Ti0 2 = 5 100. 3 3mm pellet
NO 200 ppm. NH 3 200 ppm SO 2 500 ppm
02 1 5°/o H 2 0 10% SV = 10 000 hr )
Additives Effect on TlO•Based Catalyst (4)
Figure 3.12
(3mm pellets Sv = 15000 -1 NO 150 ppm
NN 180 ppm 02 4°/Q N 2 0 100/0 SO 2 1 000 ppm
Figure 3-13 Activity of TiO Based Catalyst (5)
w
Co
TEMPERATURE (CC)
172
-------
80-
2000 4000 6000
SPACE VELOCITY (I!h)
Figure 3.14 NO Removal and SO Oxidation. (6)
i .100-
-J
> 90
0
Lu
80
0
300 350 400
4
SO 2 2000 ppn
NO 150 250 ppm
02 40%
NH 3 INO, = I
TEMPERATURE (C)
FIgure 3.15 Performance at TI-V and TI.W.V Catalysts.
(Kawasaki Heavy Industries)
60-
>.
Q
z
U i
0
U..
U-
‘U
-J
>
0
uJ
0
z
40-
ATU RE.
GAST
NH 3 !NO MOLE RATIO
O : 103
— au 2.!6
A a. 0.70
20-
L
0
U,
0
I—
to
0
C d,
U-
0
z
2.0 o
C d)
LU
>
z
173
-------
100
-J
>
0
LU
0
z
/7
40
o K 2 S0 4 A K 2 C0 2 a KNO 3
0KC I KOH
0 I
200 250 300
TEMPERATURE (°C)
Figure 3-18 Effect of Potassium on T 10 2 .V 2 0 5 Catalyst. (4)
1.0
0.8
0.6
>
<0.4
0.2
0
0 20 40 60 80 100
ACCELERATED TEST HOURS (hr)
Figure 3.17 Test Results of Accelerated Deterioration
Test with Alkalimetal Sulfate. (7)
174
-------
TABLE 3—3. PERFORMANCE OF BONEYCOMB CATALYSTS (NIHON SHOKtJBAI KAGAKU)
(Gas Composition: NO 1 180—200 ppm, SOx 150—200 ppm), CO:
7—10q 0 , HaO 7—10 , Oz 7—l , NHj/NOx mole ratio: 1.0, Area
Velocity 1.0 m/hr)
Catalyst
No.
SV 1
(hr )
NO 1
Removal
( ) at
SOz oxidation ( ) at
350°C
380°C
300°C
350°C
380°C
108 Ha
4,830
92
96
97
2.1
4.0
109 Ha
4,830
87
94
95
0.5
1.0
110 11 a
4,830
77
85
89
0.05
0.1
11 a
4,830
84
92
94
0.2
0.4
109 Hcb
3,710
79
87
90
0.5
1.1
lii. uc”
3,710
76
84
88
0.2
0.4
a.Honeycomb for oil—fired flue gas: channel diameter 6 mm, wall thickness
1 .4 mm.
bHoneycomb for coal—fired flue gas: channel diameter 8 mm, wall thic ess
1 .8 mm.
175
-------
temperature catalysts active at 150_2500C, have been developed. However, the
following problems have occurred with this type of catalyst: (1) the reaction
rate is generally slow; (2) ammonium bisulfate deposits on the catalyst below
3000C; and (3) poisoning by potassium compounds occurs more seriously at lower
temperatures (Figure 3—16)
Kureha Chemical has developed a catalyst which is active at 150°C (8).
However, this catalyst, is seriously poisoned by SO in the gas stream and
therefore has not been used commercially. S0 1 —resistant catalysts active at
200_2500C also have been developed by several manufacturers. Even with an
SO 1 resistant catalyst, ammonium bisulfate deposits form, thereby lowering
catalytic activity. To increase its activity, the catalyst occasionally must
be heated above 400°C. This heating can be achieved either by mixing a hot
gas with the low—temperature gas or by removing the catalyst from the reactor
(moving bed) for thermal treatment.
Many of the early low—temperature catalysts were based on zeolite com-
bined with various metals, as shown in Figures 3—18 (5) and 3—19 (11). They
generally consumed less than 1 mol NH to 1 mol NO. but were no more efficient
than the T102—VzOs catalyst, except for those containing Pt and Pd (which are
readily poisoned by SO 1 ). Most of the recently developed low-temperature
catalysts are based on T102, except for those developed by Kobe Steel which
are alumina—based. (Section 5.6).
3.3.2 Catalytic Activity Decrease Caused by Ammonium Bisulfate Deposit
Figure 3—20 illustrates the decreased catalytic activity of Ti03—based
metal catalysts during a 100 hour period at 250°C. The decrease is due to
ammonium bisulfate deposition. The gas used for the test contained 1,000 ppm
SOz without S03. It is obvious that a portion of SOz is oxidized on the cata-
lyst surface to form the bisulfate. The decrease in catalytic activity was
substantial with the Ce catalyst, slightly less severe with the Cu, V, and Co
catalysts, and only slight with the Mo and W. Although the catalysts with Mo
and W additives were barely affected by the bisulfate, the initial activity of
these catalysts was much lower than that of the vanadium catalyst.
176
-------
(carrier • 13X, J 5A. V Na-V. £ NH 4 moldenite
3mm pellet SV 15.000 hr’)
Figure 3-18 Low-Temperature Catalyst with Zeolute Camer (5)
REACTION TEMPERATURE (C)
(NO 1 8 mi/mm NH 3 1 2 mtlmin He 57 mt/mm
contact time hg sec/mi)
NOTE The degree of ion exchange is as follows ) Cool) V 70°/a
2) CullI) V 68°/a 3) Pd(lll V 48°/o 4) P1(111 V 52°/a
5) Fe(lII) V 750/0 6) Ni(lI) V 800/a 7) Co/Ill/-V 95°/s
Figure 3-19 Catalytic Activity of Various Transition Metal ton Exchanged
V Type Zeolites br the NO NH 3 Reaction
-J
4
>
0
w
0
z
TEMPERATURE 1C1
0
z
I L .
0
z
0
(1,
IL l
>
z
0
0
100 200 300
400
177
-------
100
-J
>
0
w
0
z
90
80
70
0
oV SV = 10.000 t Mo SV = 5000. OW SV = 1700
oCu SV = 1700 Ce SV = 1500 XC0 SV = 1000
(NO 200 ppm NI -I 3 300 ppm 50, 1000 ppm
0-, 1 5°/a -4,0 100/a 3mm De et)
Figure 3.20 Decrease in Catalytic Activity of Ti0 2 -Metal Oxide Catalysts at 250°C. (4)
DO o 00 00 00—
°d?9 °O
:z2t —
0
80
70
A
0
0
9
0
Mo
tOO
95
90
50
tO
0
V
0
5
tO
50
90
tOO
SV
5000
5800
8000
10000
10000
10000
40 60 80
TEST PERIOD lhr)
20
(NO 200 porn NH 3 300 ppm. SO 2 1000 ppm 02 1 5°/a H 2 0 10°/a 3mm pellet)
Figure 3.21 Decrease in Activity 01 Ti0 2 -MoO-V 2 0 5 Catalyst at 250°C. (4)
TEST PERIOD (hr)
90
-J
>
0
0
z
100
178
-------
Figures 3—21 and 3—22 show the results of tests performed on the T102—MOO—
VzOs catalysts at 250° and 2000C, respectively. A more noticeable decrease in
catalytic activity occurred at 200°C than at 250°C. A catalyst with an Mo:V
ratio of 95:5 exhibited only a slight decrease in activity during a 50 hour
test period at 200°C, as well as during 100 hours at 250°C.
Figure 3—23 shows the decrease in catalytic activity which occurs when
the catalyst is used with a gas containing 200 ppm SO 2 and 50 ppm S03. This
decrease was considerably smaller than that with gas containing 1000 ppm SOs
without S03, as shown in Figure 3—21. These results indicate that a consider-
able portion of SOs is converted to S03 on the catalyst surface.
Table 3—4 shows the properties and chemical composition of TiOs—VsOs—MoO
catalysts before and after use at 250°C for 100 hours and 200°C for 50 hours.
With use, NH4 and S04 deposited on the catalysts, lowering the surface area
and decreasing the pore volume. Both the deposition and the lowered activity
occurred to a greater degree with the catalyst that contained more vanadium.
The NR4/S04 weight ratio ranged from 0.12 to 0.25 and the NHI/S04 mole ratio
from 0.6 to 1.33, indicating that the deposits were mostly ammonium bisulfate
(NB4HSO4) .-
3.3.3 Use of Heating to Recover Catalytic Activity
The activity of TiOs—based catalysts can be restored by heating them to
400_4500C. In one series of tests, a catalyst contaminated with ammouium
bisulfate was heated for three hours at 400°, 430°, 450° and 470°C and then
again used for SCR at 200°, 250° and 300°C (Figure 3—24). The activity of the
catalyst heated at 450° and 470°C was almost equal to that of the fresh cata-
lyst. Figure 3—25 shows that by heating the catalyst at 450°C for three
hours, virtually all of the N114 and SO. were removed and the catalytic acti-
vity was restored. Table 3—5 shows that when the used catalyst is heated at
450°C for three hours, both surface area and pore volume increased to approach
those of the fresh catalyst.
179
-------
A
0
()
9
Q
Mo
100
95
90
50
10
0
V
0
5
10
50
90
100
SV
3000
3000
4000
6000
6000
6000
(NO 200 ppm, NH 3 300 ppm, SO 2 1000 ppm,
02 1 5%, H 2 0 10%, 3mm pellet)
Figure 322 Decrease in Activity of T 10 2 .MoO-V 2 0 5 Catalyst at 200°C. (4)
70A2713
A___ _____ ___
80
V
Mo
SV
100
0
10000
A 10
90
10000
20 - 40 60 80
TEST PERIOD (hr)
(NO 0 200 ppm, NH 3 300 ppm SO 2 200 ppm 503 50 ppm,
02 1 5°/o, H 2 O 10°/a, 3mm pellet)
100
FIgure 3.23 Decrease in Activity of T 10 2 .MoO .V 2 0 5 Catalyst at 250°C. (Gas contains SO 3 ) (4)
70A 27 14
-J
>
0
uJ
0
z
0 10 20 30 40 50
TEST PERIOD (hr)
60
-J
>
0
U i
0
z
70
0
180
-------
TABLE 3—4. PROPERTIES AND COMPOSITIONS OF CATALYSTS (FRESH AND USED)
Reaction Surface Pore
Ratio Temp. Area Volume
V:Mo (°C) (m2/g) (mug)
Chemical Composition
NU 4 ( ) S0 4 (%) N1141S04
fa 31 0.27
100:0 250 D 17 0.18
— — —
1.3 10.9 0.12
2 OO 17 0.19
1.6 10.2 0.16
f 34 0.25
— — —
0:100 250 31 0.25
0.1 0.5 0.20
200 32 0.26
0.2 0.8 0.25
f 34 0.28
— — —
90:10 250 17 0.23
1.0 7.1 0.14
200 16 0.20
1.6 8.3 0.19
f 33 0.29
— — —
50:50 250 17 0.25
0.8 4.7 0.17
200 20 0.23
1.3 5.7 0.23
f 35 0.30
— — —
10:90 250 26 0.29
0.2 1.2 0.17
200 28 0.28
0.3 1.3 0.23
f 35 0.30
— — —
5:95 250 31 0.29
0.2 0.9 0.22
200 29 0.27
0.2 0.8 0.25
Fresh (before use)
After used at 250°C for 100 hours
C
After used at 200°C for 50 hours
TABLE 3—5. PROPERTIES AND COMPOSITIONS
OF CATALYSTS (FRESH, USED,
AND
HEATED
AT 450°C FOR REE HOURS)
Surface Pore
Ratio Area Volume
V:Mo Type (m 2 Ig) (mug)
Chemical Co position
N H4( ) S04(° ) NH4ISO4
Fresh 34 0.28
— — —
90:10 Used 20 0.23
0.9 4.3 0.21
Heated 27 0.30
— — —
Fresh 33 0.29
— — —
50:50 Used 20 0.24
0.7 3 .2 0.22
Heated 30 0.30
— — —
181
-------
100
400 430 450 470
HEATING TEMPERATURE (C)
Figure 3.24
100 -
Recovery In Aclivity of Contaminat
Thre. .Hour Heating PerIod. (9)
ed Catalyst
NH 4 REMOVAL
80
-
80
.
40
I
0 1 2 3
HEATING PERIOD (hr)
FIgure 3.25 Effects of 450°C Thermal Treatment on Catalyst
Contaminated with Ammonium Blsulfate. (9)
80
-J
< 80
0
w
0
z
40
NO: 180 ppm, NH 3 . 180 ppm,
SO 2 . 180 ppm, 02. 59%,
H 2 010%,SV 8000hr
20
I ’
0
182
-------
Figure 3—26 shows that the vanadium—rich catalyst is very active ini-
tially, but requires frequent heating because of rapid bisulfate deposition.
The Ti0 2 —Mo catalyst has a low initial activity which did not decrease during
100 hours of use; a catalyst with an Mo:V ratio of 90:10 has a fairly high
Initial activity which decreased slightly with 100 hours of use.
3.3.4 Alkali Poisoning and Poisoning Countermeasures
When alkali is added to the TiOa—V20s catalyst, the catalyst becomes less
active (Figure 3—27). Potassium proved to be even more poisonous to this
catalyst than sodium. Figure 3—28 compares the poisoning of various Ti02—
metal oxide catalysts. Mo and W catalysts showed a considerable amount of
poisoning. Figure 3—29 shows the poisoning of TiOl—MoO—V20s catalysts; poi-
soning was more extensive with a larger Mo:V ratio.
Tests have indicated that alkali compounds present in the contaminated
catalysts can be removed with a hot water wash or with aqueous solutions of
11202 and H2S04. Figure 3—30 shows how washing restores catalytic activity.
The TiOz—VzOs catalyst’s activity is partially restored by a hot water wash
and almost, completely restored after a wash with aqueous solutions of flz0i and
112S04. During the same test, the activity of the Ti02—MoO catalyst was not
restored by the wash. The activity of the TiOz—VzOs—MoO catalyst (V:Ho10:90)
recovered well after a wash with a mixed solution of H203 and 11 S04.
Table 3—6 shows the change in the Ti02—Vz0s catalyst caused by alkali
poisoning and the subsequent washing of the poisoned catalyst with hot water
and a mixture of H 2 03 and H2S04. After the wash, most of potassium was re—
moved and the pore volume was restored. The V205 decreased during washing
because it is slightly soluble in water and the Hz0 and 11 1 S04 solution. For
this reason, it may be necessary to add V205 to the catalyst after it has been
washed several times.
183
-------
1
80 V Mo SV
0 0 100 3000
70 alO 90 4000
1OC
50 50 6000
: \\\\\\
70 v Mo SV
1 10 906000
70 C 00 0 6000
0 100 200 300
TEST PERIOD (br)
NO. 200 ppm NH 300 ppm SO 2 1000 porn
1 5 °/o H,O 10°/a 200C 3mm pellet)
Figure 3.26 Effect of Thermal Decomposition
at 45O C. (4)
?0A2717
to
(200C)
0 OK 2 SO 4
C Na,30 4
0
0 10 20
ALKALI ADDED (°/o)
Figure 3-27 Effect of Alkali on T10 2 .V 2 0 5 Catalyst (4)
70A2718
184
-------
-J
0
LU
*
0
z
Figure 3.28
100
50
0
Effect of Alkali on Ti0 2 .Metal Oxide Catalysts.
-J
>
0
LU
*
0
z
VO 50 100
MolCO 50 0
RATIO OF Mo AND V
Figure 3.29 Effect of Alkali on T10 2 .M0OV 2 O 5 Catalyst. (4)
185
-------
-J
>
0
w
0
z
FIgure 3-30
K 2 S0 4 Poisoning and Recovery of the T 10 2 -Based Catalyst.
(M0V Indicates Mo:V = 90:10) (4)
70A2721
100
80
60
40
20
0
FRESH
POISONED
BY
K 2 S0 4
WASHED WITH
HOT H 2 0
H 2 0 2
H 2 S0 4
H 2 0 2 +
H 2 S0 4
®S
- ®
®
-
®
.
( )
e 5
S®
S
S
®
186
-------
TABLE 36. CHANGES IN ThE TjO -ViO CATALYST CAUSED BY ALKALI ADDITION AND
BY WASHING
Surface
Area
(m 2 Ig)
Pore
Volume
(mug)
VzOs
(%)
KaO
(%)
(A)
Fresh
31
0.27
4.7
0.03
(B)
Alkali
added
31
0.25
4.5
0.60
(C)
B was
washed
by
hot
water
31
0.28
4.1
0.17
(D)
B was
washed
by
lisOs
+ H2S04
30
0.27
4.0
0.10
3.3.5 Summary
The Ti02—VzOs catalyst is active even at low temperatures of 200—250°C.
However, ammonium bisulfate readily deposits on the catalyst at these temper—
atures and reduces catalytic activity due to the oxidation of SOa to S03.
Frequent heating of the catalyst to 450°C is needed to remove the bisulfate
and restore activity. A TiOz—MoO catalyst does not have as much of a problem
with ammonium bisulfate deposition but is not very active. TiOa—based cata-
lysts with MoO and V205 in a ratio of (90—95):(1O—5) may be suitable for low—
temperature applications because they are catalytically active and have no
problem with bisulfate deposition.
Alkali compounds, particularly those containing potassium, poison the
catalysts. This poisoning occurs most seriously at low temperatures. Al-
though the alkali contained in the catalyst can be removed by washing with hot
water or aqueous solutions, some VzO is lost in the wash. This means that
more V 3 0 5 must be added and the wash liquor must be treated. For these rea-
sons, it may not be appropriate to treat alkalt—rich gas with a low—tempera-
ture catalyst.
For boilers, the application of SCH to 300—400°C economizer outlet gas
may be preferable to using a low—temperature catalyst for the 150—200°C air—
187
-------
preheater outlet gas. This is true despite the fact that an existing boiler
requires considerable modification for application of S R to the economizer
outlet. -
For low—temperature gas such as that from coke ovens (200—250° ), the use
of a low—temperature catalyst is preferable to using a conventional catalyst
combined with gas heating to 300_3500C. Although the low—temperature catalyst
requires occasional heating up to 400°C, it can save a considerable amount of
ener - compared with continuous gas heating to 300 350°C.
3.4 CATALYST SHAPE AND REACIOR TYPE
3.4.1 Introduction
Various types of catalysts and reactors have been in commercial use in
Japan, as shown in Table 3—7.
Both spherical— and cylindrical—shaped granular catalysts (3—6 mm dia-
meter) have been used with clean gas. Since only the 0.3mm—thick surface
layer of the catalyst is effective for catalytic action, smaller catalysts are
more efficient. However, since small catalysts are susceptible to dust plug-
ging, ring or ring—tube catalysts. 10—30 mm in diameter, have been used with
gas containing a small amount of dust such as 20 mg/rn 3 or less.
A rough calculation indicates that when a gas containing 20 mg/rn 3 of dust
is sent through a catalyst bed for ten hours at an SV of 5,000 hr , the
volume of the dust in the gas is approximately equal to the volume in the
catalyst.
188
-------
TABLE 3—7. SCR CATALYSTS AND REACIDRS SUITABLE FOR VARIOUS GASES
Gas Type
Fuel
(Gas Source)
Dust
(g/Nm 3 )
Catalyst Shape
Reactor Type
Clean
Natural gas
Kerosene
Naphtha
Below
0.02
Granule, Ring
Fixed bed
Semi—
Dirty
Heavy oil
Coal with.D
hot ESP
0.02—0.2
Granule, Ring
Honeycomb,
Plate
Moving bed
Fixed bed
Dirty
(dusty)
Coal
15—25
Granule (in
elements)
Tube, Plate,
Honeycomb
Parallel
passage
Fixed bed
Very c
dirty
Glass furnace,
cement kiln
sintering
machine
Granule
Moving bed
Paralle1 flow type.
E1ect .0 5tatic precipitator.
Containing vapor of alkaline compounds which deposit on the catalyst.
Dusty gas can be treated either with a moving bed or a parallel flow
reactor. Moving bed reactors have been developed by several vendors. The
parallel passage reactor was originally developed by Shell for FGD and later
modified by JGC Corp. for S R. These reactors were first applied commercially
several years ago. More recently, parallel flow reactors using parallel flow
catalysts with honeycomb, plate and tube configuration have become quite
popular. Among the parallel flow catalysts. the honeycomb and plate types
have been used most widely.
3.4.2 Moving Bed Reactor
A basic sketch of a moving bed reactor is presented in Figure 3—31. The
granular catalyst is charged from the top of the reactor and moves down either
intermittently or continuously while the gas is sent through the catalyst
189
-------
CATALYST
GAS IN
PARTICULATES
Figure 3•31 Schematic of a Moving Bed Reactor.
CONVEYOR
FLUE GAS
OUTLET
(TO AIR
PREHEATER)
CATALYST
LAYER
Figure 3•32 Moving Bed Reactor tor a Large Amount of Gas. (liii)
MOVING BED
REACTOR
VIBRATING
SCREEN
CATALYST
CONVEYOR
190
-------
layer in a cross flow (Figure 3—31) . The catalyst discharged from the bottom
of the reactor is screened to remove dust and returned to the reactor. A
larger amount of dust in the gas being heated requires more frequent screening
and recycling to keep the pressure drop below a certain level. This results
in an increase in power consumption as well as catalyst loss. For practioal
purposes the dust content of the gas being heated by a moving bed reactor
should be no more than 200 mg/Nm 3
Moving bed reactors are best suited to dirty gas. such as that from glass
melting furnaces, which contains sodium sulfate vapor. The vapor deposits on
the catalyst. These reactors are also appropriate for low temperature gas in
which ammonium bisulfate deposits on the catalyst. The contaminated catalyst
can be withdrawn from the reactor, cleaned by water washing or heating, and re-
turned to the reactor without interrupting plant operation.
Moving bed reactors have a number of advantages over parallel flow reac-
tors:
(1) The catalyst can be changed without stopping the operation of
the system.
(2) The catalyst is usually granular and is considerably cheaper
than parallel flow catalysts. It also can be packed easily in
the reactor.
(3) The moving bed can remove up to about 70 of the dust in the
gas and may be well suited to gases for which 70 dust removal
is sufficient to meet regulatory requirements. These gases in-
clude those from coke ovens and relatively small oil—fired
boilers not equipped with dust removal facilities.
(4) Moving bed reactors may be suitable for low temperature gas or
dirty gas which causes deposition on the catalyst.
191
-------
On the other hand, moving bed reactors have the following disadvantages:
(1) The operation of a moving bed reactor is more costly and
troublesome than that of a fixed bed reactor. The latter has
no moving parts and usually does not require an operator.
(2) A low gas velocity (0.5—1.5 rn/see) is used to maintain a small
pressure drop in the reactor. Therefore, when a large amount
of gas is being treated, a special reactor design is needed to
obtain a large gas flow area and to maintain uniform gas flow
and movement of the catalyst (Figure 3—32)
(3) The reactor cannot easily treat gases containing more than 200
mg/Nm 3 of particulates.
(4) Attrition of the catalyst tends to occur.
Moving bed reactors have been used in many pilot plants and conmercially
as well. They also have been used at prototype plants to treat dirty gas from
glass melting furnaces and low temperature gas from coke ovens (Table 3—8)
For oil— and coal—fired boilers, parallel flow reactors have become more popu-
lar since they now have a longer life cycle.
3.4.3 Parallel Flow Reactors
Various parallel flow catalysts, including honeycomb, plate, and tube
types, have been developed (Figure 3—33). These 50—100 centimeter—long cata-
lysts are placed in containers, transported to the reactor site, and installed
as shown in Figure 3—34.
For parallel flow catalysts, “area velocity” (AV) is often used in place
of SV (space velocity). It is defined as:
AV = Volume of gas treated (Nm 3 /hr )
Apparent surface area of catalyst (ins)
192
-------
TABLE 3-8. SCR PLANTS USING A MOVING BED REACTOR
Process
Developer
User
Plant Site
Gas Source
Fuel
Capacity
(Nm 3 Jhr)
Completion
Kurabo
k’urabo
I lirakata
Boiler
110 a
30,000
1975
Asahi Glass
Asahi Glass
Keihin
Furnace
110
70,000
1976
J.G.C. Corp.
Catalyst & Chemicals
Wakamatsu
Furnace
110
10,000
1976
Mitsubishi 11.!.
Sumitomo Chemical
‘
Sodegaura
Boiler
HO
350,000
1976
hitachi Ltd.
Kawatetsu Chemical
Chiba
Coke Oven
COGb
500,000
1976
hitachi Ltd.
Chiyoda Kenzai
Kaizuka
Boiler
110
15,000
1977
Hitachi Ltd.
Nisshin Steel
Amagasaki
Boiler
110
19.000
1977
hitachi Ltd.
Nippon Oil & Fats
Amagasaki
Boiler
110
20,000
1978
Kobe Steel
Kansai N.K.
Amagasaki
Coke Oven
COG
104,000
1977
I - .
a 11 oil
b
Coke oven gas
-------
rn
ornr
fl flE
A: HONEYCOMB
(CERAMIC)
(SQUARE)
B: HONEYCOMB
(CERAMIC)
(HEXAGONAL)
/1
/1
C: HONEYCOMB
(METAL)
(WAVE TYPE)
0: PLATE (METAL)
00
0
F: TUBE (CERAMIC)
G: PARALLEL PASSAGE
Figur. 3.33 CroasSictions of ParaII•I Flow Catalysts. (actual slz.)
E MODIFIED PLATE (METAL)
194
-------
I BASKET
TRANSPORTATION
Figure 3-34 installation of the Catalyst (Hitachi Ltd)
CRANE BOOM
‘ 0
L I ’
UNIT OF BLOCK
INSTALLATION
-------
Apparent surface area is the surface area assuming that the catalyst
surface is smooth, without pores. Examples of the relationship between AV and
SV for a square honeycomb catalyst are shown in Table 3-9.
TABLE 3-9. AV AND SV OF SQUARE BONEYCOMB CATALYST
Channel
(mm
Diameter
)
Wall Thickness
(mm)
AV
(m/hr)
SV
(hr 1 )
5
1
-
5
10
2,300
5,600
7
1.5
5
10
1,900
3,800
10
2
5
10
1,400
2,800
Since the catalytic action occurs on a thin surface layer of the cata
lyst, a thinner catalyst is more efficient than a thicker one (Table 3—10).
Recent technology improvements have made it possible to produce honeycomb and
plate catalysts with a thin wall, 0.6—1 mm thick (10). This has resulted in
the widespread use of these catalysts, especially for utility boilers (Table 4—
1). By comparison, a tube catalyst (Figure 3—33, F) with a length of 100 cm
must be more than 2.5 mm thick to have sufficient mechanical strength; it is
also a less efficient catalyst. Tube catalysts have been tested with gas
containing a large amount of dust, such as the gas from a coal—fired boiler
which tends to erode the catalyst.
Another type of parallel flow reactor originally developed by Shell is
the parallel passage reactor (Figure 3—33, G). This reactor uses a small
catalyst, about 1 mm in diameter and 2—3 mm in length, packed in a 7—8 mm—
thick envelope made of steel gauze. The envelopes are placed in parallel in a
container with a clearance of 7—9 mm. The catalyst does not need much mechan-
ical strength and thus a very active catalyst can e used. iowever, inside
the envelope, the catalyst may not work as efficiently as it would on the
outside surface.
196
-------
TABLE 3-10. SPECIFIC SURFACE AREA OF CATALYSTS
Specific Surface
Surface Area
Dimension (mm) Catalyst Volume
Area (m2/m3)
Surface
Packed
Area
Volume
Granule (St here) Diameter
4 1500 1100
6 1000 740
8 750 555
10 600 444
Tube C b
10 7 1350 405
15 10.5 900 270
20 14 677 203
25 17.5 543 163
33 23.1 410 124
Honeycomb
10 7 1204 387
9 7 1866 437
14 10 908 278
20 14 602 193
7 5 1819 556
10 7 1204 388
14 10 909 277
20 14 601 194
Parallel Plate Thickness Clearance
1 5 2000 320
L::IIIFI .1 —1 1 8 2000 220
8 8 250 125
LIJri 10 12 200 91
Apparent surface area.
True volume of catalyst assuming that it has no pores.
COptimum separate packing.
197
-------
3.4.4 Honeycomb and Plate Catalysts
The composition, production methods, and characteristics of honeycomb and
plate catalysts are compared in Table 3—11. The most popular of these cata-
lysts is the solid or molded (C—S) ceramic honeycomb which is produced by
kneading, extruding, drying, and calcinating catalyst material such as Ti02
and ViOs. The production process is simple but requires a large amount of
material for the honeycomb’s relatively thick walls. The thick walls are
needed to maintain sufficient mechanical strength.
TABLE 3-11. HONEYCOMB AND PLATE CATALYSTS
Type
Production Method
Simpliary
of
Production
Mechanical
Strength
Variety of
Composition
Ceramic
Honeycomb
C—S
Solid (or molded) type
A
C
A
C—C
Ceramic carrier, coated
B
B
A
C—I
Ceramic carrier, impregnated
B
B
B
Metal Base d
M—C
Metal plate, coated
C
A
A
MG—C
Metal gauze, coated
B
B
A
M—E
Metal plate, etched
B
A
C
A = Superior B = Medium C = Inferior
The coated ceramic honeycomb (C—C) catalyst has thin walls (0.6—1.2 mm in
diameter). To produce the catalyst, a honeycomb substrate (support) with
walls 0.3—0.9 mm is formed. The substrate which consists of mullite
(3A12O32SiO2) or cordierite (2MgO 2AlzO3’5SiO ) is then coated with 0.15 mm
of catalyst material, dried and calcined again. This coated honeycomb has
been produced chiefly by NUK Insulators Co., and Kobe Steel. Compared with
solid honeycomb, the coated type uses less catalyst material but has greater
mechanical strength and a larger surface area, as shown in Figure 3—35.
198
-------
o o,
o o C ])
7 0 —600
a C )
E
z —
0 C)
500— -
I
c
C,
U-
.<
60— —500
-J C)
U I
0 z r
400— -140
z
cri ,Z -120
U I
0. I-
.0 —i ü ’
—400
-80
z
•60
• UI
300— -40
2
-20 0
T (SOLIDTYPEYff
50—
I I I I I
0.6 0.7 0.8 0.9 1.0 1.1 1.2
WALL THICKNESS OF SUBSTRATE (mm)
Figure 3.35
Characteristics of Coated Honeycomb Substrate.
(NGK Insulator Co.)
199
-------
A third type of ceramic honeycomb is produced by impregnating the ceramic
substrate with catalyst material. This type is not popular as an SCR catalyst
although it has been widely used for automobile exhaust treatment.
There are also plate or honeycomb catalysts which are produced by coating
thin metal plates (M—C) or metal gauze (MG—C) with catalyst material. The
thickness of the coating and the metal plate (or gauze) is 0.1—0.2 and 0.3—0.5
mm, respectively. The M—E catalyst is made by etching a special type of stain-
less steel plate (about 1 mm thick) to produce a reactive 0.15 mm thick sur-
face layer. Honeycomb catalysts made with metal plate (M—C and M—E) may have
greater mechanical strength than metal gauze—based and ceramic ones. On the
other hand, M—C catalysts need a special cementing material to prevent their
coating from coming off during thermal and mechanical shocks; M—E is limited
by its chemical composition.
The M—C plate has been produced mainly by Hitachi Ltd., which recently
also began manufacturing modified plate catalysts as shown in Figure 3—33, E.
To make the modified plate, a stainless steel plate is processed to form wire
netting. The netting is treated chemically to prevent SO s corrosion, coated
for f1exib tlity and formed in the shape of the air preheater element. The M—E
catalyst is produced by Hitachi Zosen, which recently began producing the MG-C
catalysts as well.
Figure 3—36 shows the gas velocity distribution in each section of a
channel of plate and honeycomb type catalysts. The velocity is higher at the
center and lower near the wall; it is particularly low in the small—angled
corners which are susceptible to dust plugging. Since the honeycomb catalyst
has many corners, its efficiency is not substantially higher than that of a
plate catalyst although honeycomb does have a larger surface area.
Among the types of honeycomb catalysts, the hexagonal variety is superior
to the square or wave types because of its larger corner angles (Figure 3—33)
However, the mechanical strength of the hexagonal type is less than that of
the square.
200
-------
PARALLEL PLATE
SQUARE
EQUILATERAL TRIANGLE
1
0 3 6 9 12 15
VELOCITY (mis)
VELOCITY (m/s)
50
25
n
03
6 9 12 15 18
VELOCITY (m/s)
Velocity Distribution In Each Section of a Channel of Plate and Honeycomb Catalysts.
(HitachI Ltd.)
0
Oz
00
-------
3.4.5 ComDarison of Catalysts and Reactors
The operation parameters and performance of catalysts and reactors are
compared in Tables 3—12 and 3—13. A moving bed reactor has a thin catalyst
bed through which a low velocity gas passes in order to minimize pressu.re
drop. Parallel flow reactors have much deeper beds and much higher gas velo-
cities. With parallel flow catalysts, especially those with large channel
diameters, a low gas velocity produces a low NOx removal efficiency due to
the laminar flow of the gas. Moreover, particulates tend to adhere to the
catalyst at a low gas velocity thereby reducing catalytic activity (Figure 3—
37). A high gas velocity causes a turbulent flow, reduces particulate adhe-
sion and increases NO 1 removal efficiency (Figure 3—37, B). On the other
hand, a high velocity also increases the pressure drop and the catalytic ero-
sion caused by particulates. Therefore, it is best to use a moderate gas
velocity with an appropriate channel diameter and bed depth suited to the
NO 1 removal efficiency required, dust content of the gas, fan capacity, etc.
As shown in Tables 3—12 and 3—13, a moving bed reactor can have a larger
SV than a parallel flow reactor. The moving bed reactors have not been widely
used becai se of their higher pressure drop and the possibility of mechanical
problems including catalyst attrition. Parallel flow reactors have no moving
parts and this allows unattended operation. Both honeycomb and plate cata-
lysts with thin walls are most commonly used because of their high efficiency
and low pressure drop.
A particular catalyst and reactor should be chosen according to the com—
position of the gas to be cleaned, nature of the dust, desired NO 1 removal
efficiency, catalyst life, and other factors.
202
-------
TABLE 3-12. COMPARISON OF CATALYS AND REACTORSa
a$ 0 _ 85 % NO removal with
gas. (SOx: 200—2000
bg/Nm for others.)
\te1Oc ty at 350—400°C in open column (superficial velocity)
volume (Nm’/hr)/catalyst volume (m’)
e’ ’ catalyst bed
Thickness of envelope
F )
0
Moving
bed
Ceramic
Molded
Honeycomb
Coated
Honeycomb
or Plate
Metal Base
Tube
Parallel
Passage
Catalyst size (mm)
Diameter
Wall thickness
Channel diameter
5—7
1—2
5—8
0.6—1.2
4—8
0.5—1
4—8
3—5
15—30
7 _ 9 e
7—9
Gas Velocity (m/sec) t ’
0.5—1.5
3—7
3—7
37
510
510
Bed depth (m)
0.3—0.6
1.5—3.5
1.2—3
1.2—3
35
3—5
SV (1000 hr_i)C
5—8
3—6
3.5—7
3.5—7
2—4
2.5—5
Pressure drop (mm
50—90
20—60
15—50
15—40
50—90
50—90
an NH /NOx mole ratio of 0.85—0.9 at 350—400°C for dirty
ppm. Particulates: 0.02—0.2 g/Nm’ for moving bed and 0.02—20
-------
1iuc gas Is pasSed through a hot—side ESP prior
Iluc gas with full dust load is treated by SCR.
di5SU1C drop through catalyst bed
fixed bed
IntermitLeflt moving bed
Continuous moving bed
rsj
0
TABLE 3—13. TYPICAL
EXAMPLES
OF SCR FOR
BOILER
Fuel
Dust
(gINm’)
Inlet
NO
(pp .)
NO
Remo al
(s.)
Nih/NO
mole ‘
RatIo
(‘.ranular Catalyst
Honeycomb Catalysf 1
Diameter
(mm)
Pressure
SV Drop
(lit ‘) (urn UuO
Cbnnnel
Diameter
(mm)
Wall
Thickness
(mm)
SV
(hr 1)
Pressure
Drop
(mmJIsO
Gas
0-0.01
100
80
90
0.90
1.00
5 d
5 d
18.000 50
13,500 65
5
5
1.2
1.2
9.500
7.000
30
40
Oul
(lows)
0.01-0.05
100
200
80
90
80
90
0.90
1.00
0.86
0.96
5
S
5°
S
11.000 70
8,000 85
9.000 80
6,500 95
6
6
6
6
1.4
1.4
1.4
1.4
5.300
3.800
4.700
3.300
40
50
45
55
Oil
(highs)
0.05—0.2
200
80
90
0.86
0.96
5
S
7.000 90
5,000 105
7
7
1.6
1.6
3.600
2.500
50
60
CoulU
0.05—0.2
300
80
90
0.85
0.95
5
5
6.500 95
4.500 110
7
7
1.
1.6
3.300
2.200
50
60
Cuala
15—25
300
80
90
0.85
0.95
Not
feasible
8
8
1.8
. 1.8
2,700
1.800
55
80
to 5CR.
-------
I I I
2 4 b 10 20 40
GAS VELOCITY (misec)
(A)
-J
>
0
‘U
0
z
GAS VELOCITY (mlsec)
(B)
Figure 3.37 Results of a Teat Using a Honeycomb Catalyst with Flue Gas Containing
380 mgINm 3 of Fiy Ash.
N)
C
(N
E
0
E
z
0
U)
Lu
I
I .-
U)
0
10-I -
ic r 2 -
-------
3 .5 PROBLEMS WITH SCR FOR COAL—FIRED BOILERS
3.5.1 Introduction
The application of SQ to flue gas from coal—fired boilers has become
more and more important in Japan because of the increase in the number of coal—
fired boilers and the relatively high NO concentration in their flue gas
200—400 ppm (depending on coal and furnace types) even with advanced com-
bustion modification. Table 3—14 lists commercial ScR plants for coal—fired
boilers.
SQ for coal—fired boiler flue gas has had the following problems:
(1) When coal—fired utility boilers use pulverized coal, the flue
gas contains 15—25 gINm 3 of fly ash. This fly ash is made up
of hard spherical particles (1—30 microns in diameter) which
erode the catalysts.
(2) Adhesion of the fly ash to the catalyst causes a decrease in
catalytic activity as well as plugging of the catalyst.
(3) A low—oxidation catalyst may be needed because the flue gas is
usually rich in SOi.
(4) The fly ash may become contaminated by ammonium compounds which
may present problems with the utilization and disposal of the
ash (Section 3.7.2).
There are two S R systems available for the trea ieut of flue gas——— high—
dust and low—dust, as shown in Figure 3—38. In the high—dust system. econo—
mixer outlet gas with a full dust load is passed through an SCR reactor, air
preheater, and then a cold—side electrostatic precipitator (cold ESP) or
baghouse. In the low—dust system. the economizer outlet gas is first treated
with a hot—side ESP (hot ESP) to reduce the dust content to 50—200 mg/Nm 3
206
-------
ItctroI It
New
lIitaubasbi heavy Industries
Babcock Hitachi
tawasukI heavy Industries
I ,hakuwsjacsa—llariaaa heavy Industries
hlo ncy comb
Plate
Tubc
Reactor is dciigned for 80% removal. which iw attained by increaain the amount of catalyst
r. )
Q
TABLE
3-14.
Capacity
ilanufac—
Year of
high Dust
Low Dust
NO 1 Rcmovsl %
(Deiign)
Power Coopany
Power Station
(11w)
NIB
turer
(‘atalyst
Completion or
Chugoku Electric
“
“
“
“
Shimonosoki
Shin Ube
Shin Ube
Shin lIbo
Ihizuahima
Ilazushima
173
75
75
156
125
156
K 5
K
K
K
K
B
1011°
MIII
MIII
M1I
tIll
Bhi
1 1 g
I I
II
“ I I
l
P
1980
1982
1982
1982
1984
1984
Uigb
111gb
11 1gb
111gb
11 1gb
high
55
65 k
65 k
63 k
65 k
63
Electric Power D.C.
“
T kchara
Takehara
250 s /a
250 z /a
700
B
Rb
N
1111
11 e
D I I
P
T 1
P
1981
1981
1983
Low
low
low
80
80
80
llokkaido Electric
Tomato
350 1
N
hi l l
P
1980
l.ow
80
Joban Kyodo Electric
“
Nakoso
“
600
600
N
N
Ulhl
1111
II
II
1982
1983
Ihigh
High
80
80
Kyusbu Electric
Minato
136
B
MIII
II
1983
11 1gb
50
Tohoku Electric
‘loboku Electric
Scodsi
Scndai
175
175
K
B
DII
1 111
p
P
1982
1982
Ihigh
high
56
56
-------
NH 3
IDF
BUF
PREHEATER
LOW DUST DeNO PROCESS
PREHEATER
HIGH DUST DeNO PROCESS
HOT DeNO
BOILER ESP AIR
FDF
0
STACK
NH 3
GGH WET-DeSO
IDF BUF
DeNO
COLD
ESP
BOILER AIR FDF
STACK
GGI-I WET-DeSO
FIgure 3-38 Total Flue Gas Treatment Process for Coal-Fired Boilers.
-------
(over 99% removal) and then passed through an SCR reactor and an air pre-
heater.
The advantages and disadvantages of the high—dust and low—dust systems
are compared in Table 3—15.
TABLE 3-15. COMPARISON OF HIGE—DUST AND LOW-DUST SCR SYSTEMS FOR COAL
High—Dust System
Low—Dust
System
Catalyst erosion by dust
More
Less
Dust adhesion to catalyst
Less
More
Ammonium bisulfate deposition
in air preheater
Less
More
Contamination of fly ash by
ammonium compounds
Can be prevented by
reducing unreacted
N H3
None
Cost of ESP
Less
More
Nature of ESP
Suitable to medium
and high sulfur coals
Suitable to
sulfur coal
low
With the low—dust system. catalyst erosion does not occur but the dust leaving
the hot ESP consists of fine, relatively alkaline ash which tends to adhere to
the surface of the catalyst. In addition, ammonium bisulfate deposits tend to
form in the air preheater. With the high—dust system the large amount of dust
prevents the deposit of bisulfate. On the other hand, fly ash caught by the
cold ESP may contain a noticeable amount of ammonium compounds, which can
cause problems in utilizing or disposing of the ash. Both high—dust and
low—dust systems have been used commercially in Japan (Table 3—15) . The high—
dust system may prove to have more advantages than the low—dust system if the
problem of ammonia contamination of the fly ash can be solved (Section 3.7).
3.5.2 Fly Ash Erosion of the Catalyst
With a parallel flow catalyst, erosion occurs mainly at the gas inlet
and, to a lesser degree, on the inside surfaces. Figure 3—39 compares the
209
-------
x 10—s
E
3
U i
I-
4
I
z
0
U,
0
I
Ui
U,
-j
4
I-
4
0
2.5
2.0
15
10
05
0
10.
T1ME(hr)
10A2133
Figure 3.39 Change In Catalyst Erosion Rate with Time. (6)
EXPERIMENTAL EQUATION W WEIGHT LOSS OF CATALYST (wt¼)
W KoV 29 .C t7 T 4 K CONSTANT
V GAS VELOCITY (mis )
C. DUST CONCENTRATION (giNm 3 )
I OPERATING TIME (hr)
Figure 3.40 AnticIpated Weight Loss ol Plate Catalyst vs. Operating TIme. (2)
0 2000 4000 6000 8000
20
a
U,
4
4
C.)
U.
0
U,
U,
0
I
C,
w
OPERATING TIME (hr)
7OA27 4
210
-------
erosion rates of conventional and improved tube catalysts (6) . With the
conventional molded catalyst a substantial amount of erosion occurs at the
beginning of operation and then continues slowly. The erosion can be signi-
ficantly reduced by hardening as much as 1/100 to 1/50 of the length of the
inlet portion of the catalyst by impregnation with a certain material (6).
Figure 3—40 shows the weight loss in a metal—based plate catalyst that is
caused by fly ash erosion (2) . The figure indicates that as far as erosion is
concerned, the catalyst may be useful for over 2 years; the weight loss in
20,000 hours is about 0.3% with flue gas containing 30 g/Nm 3 of dust at a
velocity of 4 rn/sec. The figure also shows an empirical equation which ex-
plains the relationship between the weight loss and the gas velocity, dust
concentration, and operation time. The equation indicates that the weight
loss associated with a flue gas (velocity: 6 m/sec) containing 20 g/Nm 3 of
dust is about 0.5% in 20,000 hours.
Figure 3—41 compares the erosion ratio of different types of honeycomb
catalysts (7). A conventional molded catalyst D is most susceptible to ero-
sion, while catalyst C, which contains an impregnated catalyst carrier, under-
goes less prosion because of the hardness of the carrier. Catalyst B, which
contains erosion—resistant material, undergoes even less erosion. In the case
of the improved catalyst B (B and dummy), erosion was eliminated altogether by
using a dummy layer which has the same cross section as the catalyst and is
placed upstream of it to absorb the impact and act as a flow straightener.
3.5.3 Dust Adhesion and Plugging
Fly ash from pulverized coal usually consists of spherical particles of
a hard glassy substance. The particles are generally 1—30 microns in diameter
and have fairly smooth surfaces as shown in Figure 3—42. Therefore, the dust
adhesion and plugging problem can be eliminated when a flue gas with full dust
load is sent down through parallel flow catalysts with a sufficient channel
diameter. An excessively high gas velocity can cause the catalyst to erode
and a large channel diameter makes the process less efficient. A gas velocity
211
-------
25
TEST CONDITIONS:
GAS VELOCITY 20 rn/sec
ANGLE 00
ASH LOADING 50 g/Nm 3
AVERAGE DIA. 20 microns
B .—
A _.-_______
B + DUMMY
EROSION
RESISTANT
CATALYST
/
/
/
I
/
/
/
COMMON
CATALYST
(IMPREGNATION
TYPE)
D
COMMON
CATALYST
(KNEADING
TYPE)
FIgure 3-41 Results of Accelerated Erosion Test. (7)
0
I-
z
0
Cl)
0
LU
20
15
10
5
0
/
212
-------
(B) ISOGO (LOW-SULFUR COAL)
Figure 3-42
Fly Ash From Coal-Fired Boilers in Japan
(Japanese Coals) .( 3OOO)
TR.produc from
L st available copy .
(Al TAKASAGO (MEDIUM-SULFUR COAL)
213
-------
around S rn/sec and a channel diameter of 5—8 mm for honeycomb, or clearance of
4—7 mm for the plate catalyst, are usually used to eliminate these problems.
On the other hand, the fly ash leaving a hot ESP consists mainly of
smaller particles which are often rich in alkaline components and have a
tendency to adhere to the catalyst surface, especially at the gas inlet.
Since the amount of ash leaving the hot ESP is not large, the dust adhesion
and plugging problem can be eliminated by occasional soot blowing.
The design and operation of SCR plants for coal—fired boilers in Japan
have been improved significantly so that there is generally no dust clogging
problem with the catalyst, even without soot blowing. The U.S. Environmental
Protection Agency (EPA) conducted StR demonstration tests with Hitachi Zosen’s
honeycomb catalyst at Georgia Power’s Mitchell Pilot Plant. Results of this
test showed that the fly ash adhered considerably on the catalyst surface even
with soot blowing, resulting in a decrease in NO removal efficiency and/or
an increase in the pressure drop. Although the adhesibility might be caused
by a unique condition at the pilot plant (Section 4.10.3) further studies are
needed on the adhesive properties of fly ash with different compositions.
It may not be feasible to use S R on plants fired with coal if a consi-
derable portion of the alkali compounds are volatilized by combustion. The
alkali components may condense on the catalyst and poison it to some extent.
3.5.4 NOx Removal Efficiency for Flue Gas from Coal
An example of StR for coal—fired boiler flue gas using a honeycomb cata-
lyst is shown in Figure 3—43. Generally speaking, the amount of catalyst
needed for coal—fired boilers is about twice the amount needed for low—sulfur
oil—fired boilers with equal power generation capacities. This is true for
the following reasons: 1) Flue gas volume is nearly 20% more for coal than
for oil given an equal power generation capacity. 2) Because of a higher SOs
concentration in the coal flue gas, a less active low—oxidation catalyst is
needed. 3) Since a large amount of fly ash in coal flue gas tends to cause
erosion, the catalyst has to be harder and less active. 4) In order to
214
-------
30
E
a.
20
uJ
-j
10
0
06 07 08 09 10 11 12
NH 3 /NO MOLE RATIO
(channel diameter 8mm, wall thickness 1 8mm, inlet NO 300 ppm,
- full dust load, pilot plant test)
Figure 3.43 SCR for Coal-Fired Boiler Flue Gas at 350CC Using Honeycomb Catalyst.
-J
>
0
LU
0
z
215
-------
prevent dust plugging, a larger channel diameter is needed (Table 3—13). 5)
Higher inlet NO 1 concentrations require a larger amount of the catalyst for
a given NO 1 removal ratio with a given concentration of unreacted NB3
(Figure 3—4)
In order to prevent the plugging of the air preheater with ammonium
bisulfate (Section 3.6), it is necessary to keep unreacted NH, below 10 ppm
for the high—dust system and below 5 ppm for the low—dust system. For the
high—dust system, it may be necessary to maintain the unreacted NH level
below 5 ppm in order to eliminate the contamination of fly ash by ammonium
compounds.
Figure 3—43 indicates that a 90 percent NO 1 removal efficiency with an
unreacted N Ih level below 5 ppm may be attained with the NR ,/N0 1 mole ratio
of 0.93 and an SY of 1,500 hr . When treating a large amount of flue gas
from a utility boiler, it may be difficult to attain 90 percent removal with 5
ppm unreacted Nll for two reasons. First, the flue gas velocity as well as
NO 1 concentration is not uniform in each portion of the large inlet duct and
reactor, and second, boiler load often fluctuates, resulting in the fluctua-
tion of the gas volume and NO 1 concentration. In order to maintain un—
reacted NB, below 5 ppm, an 80—85 percent NO 1 removal efficiency may be a
practical limit for a large coal—fired boiler.
3.6 AMMONIUM BISULFATE DEPOSITION IN TEE AIR PREHEATER
3.6.]. Formation and Behavior of Ammonium Bisulfate
While most problems with S1 R already have been solved, the greatest
remaining problem is the deposit of ammonium bisulfate in the air preheater or
heat exchanger causing plugging and corrosion (Figure 3—44). The problem is
significant, not only for S R but also for selective noncatalytic reduction
(SNR or Thermal DeNO 1 ) as well.
S03, NIh, and R20 present in hot gas combine to form ammonium bisulfate
(NE4HSO 4 ), on cooling:
216
-------
Figure 3-44
Plugged (Below)
tl ;pJ
IJEITlr4iutr
Air Preheater Elements: Clean (Above)
217
-------
NRa + S03 + RiO > NB4RSO4 (1)
(g) (g) (g)
The formation temperature is shown in Figure 3—45. Bisulfate will form
below approximately 2300 C in a gas containing about 10 ppm of both NRa and
so 3 .
Ammonium bisulfate has a melting point of 147°C (Figure 3—46). Between
the formation temperature and the melting point, bisulfate exists as a
corrosive melt. Below the melting point, it exists as a very hygroscopic
solid. A simple calculation shows that 10 ppm each of SO and NBa in gas from
a 500 MW boiler could account for the formation of 2 tons/day of bisulfate.
This amount is sufficient to cause problems such as plugging and corrosion in
an air preheater or heat exchanger. Many problems of this nature have been
encountered in a few S R facilities in Japan. The problem has not been as
serious for small boilers and furnaces that discharge flue gas at a tempera-
ture above about 2300C. However, for larger boilers and furnaces which re-
quire sufficient heat recovery from the flue gas, countermeasures must be
taken to eliminate the bisulfate problem.
Thereis some question about the composition of the substance which first
forms. Based on thermodynamic data, it appears that ammonium sulfate,
(NR 4 )2S04, deposits from NRa—rich gas. However, the sulfate may not be stable
at a high temperature. This author’s study indicates that (NH 4 ) R(SO 4 )z, a
double salt of (NR4)1S04 and NH 4 RSO4, forms at a high temperature under the
presence of excessive NH 3 (16).
Other studies indicate that when gas containing equal amounts of SO 1 and
NH 3 , is cooled, mists of sulfuric acid HSSO4 form prior to the formation of
bisulfate (13). Even though the mists form, they must absorb ammonia immedi-
ately, according to the NBa concentration. When a hot gas cools, the result
may be the formation of mists with varying compositions based on the gas
composition. The mists solidify either partly or entirely on further cooling
to form a mixture of sulfuric acid and ammonium bisulfate, bisulfate and
double salt, or double salt and ammonium sulfate as will be shown in Section
3.6.3.
218
-------
E
a.
a.
z
0
I-
z
U i
0
z
0
0
U,
I
z
1 5 10 50 100
SO 3 GAS CONCENTRATION (ppm)
Figure 3 45 Formation Temperature ot Ammonium Sultate and Busultate (12)
P DATA FROM REFERENCES
Ui
— - - - DIFFERENTIAL TRERMAL ANALYSIS
U i
150 —0---- MEASUREMENT OF LIQUID LINE
-J — —-— MEASUREMENT OF SOLID LINE
U i
07 08 09 10 11 12 13
NH 3 IH 2 SO 4 MOL RATIO
Figure 346 Melting Point 01 NH 3 -I4 2 S0 4 System (12)
219
500 1000
-------
Bisulfate, particularly when combined with sulfuric acid, is highly
corrosive and readily reacts with air preheater material and fly ash to form
various solid compounds (Sections 3.6.6 and 3.6.7). The reaction products
also cause plugging of the air preheater.
It should be noted that although the formation of compounds causes pro—
blems, at the same time it helps to reduce SO 3 emissions. Sulfuric acid mists
are far more hazardous than SOz and yet are not well captured with FGD.
Ammonium promotes deposit formation and prevents the emission of SO 3 or sul-
furic acid mists.
3.6.2 Formation of SO 3
The relationship of S03 and 02 concentrations in flue gas from the
combustion of heavy oil is shown in Figure 3—47. The figure indicates that
SO s increases with an increase in the Oz concentration in the gas. Therefore,
low—oxygen combustion to reduce NO also helps to reduce SO3. The use of less
oxygen, however, tends to increase the particulate content; particulates
increase as SOs decreases, as shown in Figure 3—48. Usually 1—3 percent of
the total 50 x in flue gas from heavy oil is present as SOs under good
combustion conditions. It has been shown that the SOs concentration in gas
from coal is lower than that in gas from heavy oil. This is because coal fly
ash is a better adsorbent of S03.
The catalyst for selective catalytic reduction of NO oxidizes a small
portion of SO 2 to S03. The actual oxidation ratio differs with catalyst
composition, temperature, and 02 and NBs concentrations (Section 3.2.4).
Although catalysts with a low oxidation ratio (below 1%) can be used, a
considerable amount of SOs still may form in SOs—rich gas.
3.6.3 Laboratory Tests on Bisulfate Deposition (12)
Figures 3—49 and 3—50 show an apparatus used to test ammonium bisulfate
deposition. Gases containing N E 3 , SOs, and IliO, are preheated and mixed at
4000C, and then sent through connected pyrex glass tubes 9.5 mm in inner
220
-------
100
70
50
30
E
0
(l
15
E
z
E
C ,,
i ii
I -
4
-J
C-)
I-
4
0
03 05 10 20 30
°2 (%)
(heavy oil S 1 3°’o) 14)
Figure 3 47 Relationship 01 02 and SO 3 Concentrations
400 -
200 -
100
50 -
0 20 40 60 80
SO 3 (%)
(heavy Oil S 1 3%, O in gas 03 3°” ) (14)
Figure 3 48 Relationship 01 SO 3 Concentration to Particulate Content
10
5
/
221
-------
FLOWMETER
H 2 S0 4
t’.)
t\)
I ’ -)
AIR
HEATER
DEPOSITION
TUBE
TUBE (B)
NH 3 IN 2 GAS
CYLINDER
OIL BATH
ELECTRIC FURNACE
SPIRAL TUBE FOR GAS
MIXING AND HEATING (C)
FIgure 3-49 Apparatus for Ammonlum Blsultate DepositIon Test
-------
________ MIXED GAS
IN LET
DEPOSITION TUBE
pyrex glass
95 mm ØID
120 mm%OD
NOMENCLATURE OF TUBES
Inlet tube A
straight tube
curved tube
outlet tube
Figure 3.50 Tube Used for Ammonlum BIsulIato Deposition Test.
iS, 2S, 3S, 4S, 5S
1C, 2C, 3C, 4C
z
THERMOCOUPLE
10A2744
-------
diameter and 1.2 m long. At the end of the test cycle these tubes are sub-
merged in an oil bath kept at different temperatures ranging from 120 to
2800C. After the gas has passed through them for a certain period of time,
the tubes are analysed for deposits.
Figure 3—51 shows the deposit ratio of S03 and NRa when a gas containing
200 ppm each of SO 3 and NRa with 10 percent RaO was sent through the appara-
tus. A maximum deposit ratio (nearly 100 percent) was obtained at a tempera-
ture of 200°C and a velocity of 10 rn/sec. It appears that the high gas veloc-
ity caused gas turbulence which in turn promoted deposit formation. The
deposits were mainly NH4USO4 with a small amount of (NR 4 )3H(S04)a; the NRa/SO.
mole ratio was almost 1.1.
Measurement of the deposits formed in each section of the tube at a gas
velocity of 6 rn/sec revealed that nearly all of the deposits formed within 60
cm of the inlet. This indicates that most of the deposits formed within 0.1
second after being introduced into the glass tube at 120—240°C.
Figure 3—52 shows the NHS/S04 mole ratios of the deposits formed in
each secti9n of the tube when gases with different mole ratios were passed
through the tube at a temperature of 160°C. Despite differing inlet gas
compositions the deposits formed at the inlet of the tube had ratios of about
1.1; the ratios of deposits formed downstream were closer to that of the inlet
gas. Figure 3—53 indicates that the deposit ratio was the highest when the
NRa/SO 3 mole ratio of the gas was around 1.1. The low melting point at the
1:1 ratio (Figure 3—46) may favor deposit formation. Figure 3—54 shows that
the deposit ratio was nearly 100%, even with low NBa and S03 (10 ppm each)
concentrations in the gas, although the tests using these low concentrations
may contain experimental errors.
The tests described above show that deposits form very rapidly when the
gas is cooled to the critical temperature, and that the deposit ratio of SOa
is usually slightly smaller than that of NB3. A portion of the S03 may form
fine mists of 112S04 which have a tendency to pass through the tube along with
224
-------
GAS COMPONENT
NH 3 200 porn
503 200 ppm
H Q 10 VOL°’c
120 150
10
6
200
OIL BATH TEMPERATURE (IC)
Figure 3.51 Gas Velocity and Deposition Ratio
N 11 3 /SO 3 = 1 5
250
300
GAS VELOCITY 6 Nmlsec
GAS COMPONENTS
NH 3 200 pDm constant
so 3 variable
H 2 0 10 voi°/o
OIL BATH TEMPERATURE i60 C
alter 3 hrs
20 40 60 ao 100
DISTANCE OF TUBE FROM GAS INLET (cm)
A iS IC 2S 2C 3S 3C 4S 4C 55
Figure 3-52 NH 4 ISO 4 Ratio of Deposits Formed from Gases of Varying Compositions
in Different Tube Sections
225
100
80
60 -
40 -
20 -
0 45 -
GAS VELOCITY (NmIsec)
l0
0
I-
4
I-
(1)
0
a.
w
I
1
0
(I ,
0
U)
I
1
20-
15
05 V
NH 1 ISO 3
15
C 12
o ii
o 10
08
• 05
05
•
120
z
-------
alter 3 Firs)
NH 200 ppm constant
60
OIL BATH TEMPERATURE 160C
GAS VELOCITY 6 Nm/sec
40 - GASCOMPONENTH 2 O 10voI
05 10 15
NH 3 /S0 3 MOLE RATIO
Figure 3-53 NH 3 ISO 3 Mole Ratio vs. Deposit Ratio
100
80
60
GAS VELOCITY 6 Nm/sec
H 2 O GAS CONCENTRATION 10 vOI°/o
OIL BATH TEMPERATURE 160C
40 NH 1 ISO 3 = 1 0
20
0
50 100 150 200
NH 3 503 GAS CONCENTRATION lOom)
Figure 3 54 Inlet SO 3 and NH 3 Concentration vs Deposit Raiio
100
80
SO 1 200 ppm constant
0
0
0
uJ
0
C)
N
20 -
0
0
U,
0
c i-
w
0
226
-------
the gas. The NH3 in the gas is immediately absorbed by the deposits, parti-
cularly by those in the liquid phase.
Because the deposits are very hygroscopic, a considerable amount of
liquid may be present even at temperatures between the melting point and
1OO C. The liquid phase may be the most corrosive.
3.6.4 Air Preheaters (Heat ExchangersI
There are two types of air preheaters or heat exchangers used to heat
flue gases. These are shown in Table 3—16.
TABLE 3-16. TYPES OF AIR PREHEATERS (HEAT EXCHANGERS)
1. Rotating
a. Ljangstrom (with rotating elements)
b. Rothemuehle (with rotating hoods)
2. Multi—tube (no moving parts)
Rotating preheaters have a better heat transfer efficiency and have been
used mainly for treating a large amount of gas while multi—tube preheaters are
simpler and have been used mainly for smaller gas sources.
For large boilers, Ljungstrom preheaters have been widely used. This
type of preheater has rotating elements as shown in Figure 3—55. Flue gas
leaves the boiler economizer (or SGR reactor) at 300_4000C and then passes
through one side of the elements and is cooled to 150_1600C, while air or cold
gas passes through the other side of the elements and is heated to 250_3000C.
Figure 3—56 shows the temperatures of the gas, air, and heating elements.
These temperatures usually fall into three zones: high—. intermediate—, and
low—temperature. The temperatures of the elements are between those of the
gas and air and fluctuate within a certain range as the elements rotate be-
tween the gas and air. For example, Figure 3—56 shows that at the boundary of
the intermediate— and low—temperature zones, the temperatures of the gas and
227
-------
AIR INLET
Figure 3•55 SchematIc of Ljungstrom Air Preheater.
228
-------
300 —
100
0
FIgure 3.56
Temperatures of Gas, Air and Heating Element in a
Ljungstrom Air Preheater.
0
a
w
I-
w
0
w
200— I
— 400
— 300
200
— 100
—0
LOW-TEMP
ZONE
INTER-
MEDIATE
ZONE
H IG H .TEM P
ZONE
229
-------
air are 230°C and 140°C respectively. At the same time the temperature of the
elements fluctuates between 200 and 170°C. These are the most favorable
temperatures for the formation of deposits.
The gas velocity through the heating elements is normally about 3 rn/sec.
The retention time for the gas in the preheater is approximately 1 second at a
temperature below 230°C. During this time most of the S03 or NH3 in the gas
(whichever is present in a lower concentration) may precipitate to form
ammonium bisulfate and related compounds while the rest of the S03 and NH3 is
released from the preheater.
An air preheater or heat exchanger usually has a soot blowing system on
the cold side of the elements (Figure 3—57A). However, this system cannot
effectively remove the deposits, particularly those formed between the ele-
ments in low—temperature and intermediate zones (Figure 3—57)
Recently an improved type of air preheater (Figure 3—57,B) has been used
commercially. This preheater has two elements, with hot— and intermediate—.
and low—temperature zones and a soot blowing system applied to both the cold
and hot sides of the heater. This preheater is considerably more costly than
conventional preheaters and is used with the low—dust SCR system to treat flue
gas from coal. (Section 3.5.1)
Figure 3—58 shows two of the most popular kinds of heating elements.
Tests have proven that the notched—flat type (Figure 3—58, A and Figure 3—44)
is easier to clean with soot blowing than the notched—undulated type (Figure 3—
60.B). Further studies are being conducted to develop heating elements which
minimize the bisulfate problem.
Another rotating air preheater, the Rothemuehie type, is shown in Figure
3—59. It has stationary elements with a rotating hood for hot gas. Ammoniurn
bisulfate deposits form in this preheater in the same way that they form in
a Ljungstrom preheater.
230
-------
(A) CONVENTIONAL PREHEATER
SOOT BLOWER
(B) IMPROVED PREHEATER
SOOT BLOWER
LOW AND INTERMEDIATE-TEMP. HIGH-TEMP SOOT BLOWER
Figure 3-57
Heating Element Configurations in Conventional and Modified Ljungstrom Air Preheaters.
DEPOSIT
r’-)
-------
NOTCHED
1 GAS
‘I
V
(A) NOTCHED-FLAT TYPE (N F)
12
NOTCHED
/
//
//
//
(B) NOTCHED-UNDULATE TYPE (N 0)
25
FIgure 3-58 Two Types of Heating Elements.
232
-------
AIR INLET-
HOT GAS ______
INLET
HEATING ELEMENT
1.2 NFTYPE
3,4 NU TYPE
DIMENSIONS IN mm
Figure 3-59 Rothemuehie Air Preheater Used in a Pilot Test Plant.
233
-------
BULK AIR BULK GAS BULK AIR BULK GAS
TEM PERATURE TEMPERATURE TEMPERATURE TEMPERATURE
(°C) (°C) (°C) (°C)
I 4 4 4
80
100
120
140
160
AIR
TUBE SURFACE TEMPERATURES
Figure 3•60
Deposits in Air Preheater Tubes.
234
-------
Figure 3—60 shows the temperatures of the gas 1 air, and tubes as well as
the deposits formed in a typical multi—tube air preheater. The preheater has
been used with a 10 MW equivalent oil—fired industrial boiler with an SCR
system. The 350°C gas leaves the S R reactor and passes through the tubes
with a velocity of 10 mlsec. The gas is then cooled to 220°C while air pre-
heated to 80°C is passed outside the tubes and heated to 160°C. The deposits
which formed inside the tube at temperatures below 200°C, ‘were found to be the
reaction product of ammonium bisulfate and the tube material.
3.6.5 Laboratory Corrosion Tests (15 )
Corrosion tests have been conducted using four different steel samples.
The chemical composition of the steels are shown in Table 3—17. Each sample
was dipped into melted ammonium bisulfate at 200°C. As shown in Figure 3—61,
low—alloy steel and SS 41 steel corroded slowly at the beginning and then
fairly rapidly while stainless steels corroded rapidly at the onset and then
more slowly after that. The corrosion rates of the stainless steels and
aluminum are shown in Figure 3—62. The figure indicates that aluminum under-
went little corrosion because it formed a protective coating.
Figures 3—63 and 3—64 illustrate that addition of ammonlum sulfate
reduces corrosion; the strong acidity of the bisulfate is weakened by the
sulfate. On the other hand, corrosion is substantially increased by the
addition of sulfuric acid as shown in Figure 3—65. Most of the liquid in the
tube reacted with the steel in about 60 hours and the corrosion (as shown by
the dotted line) almost stopped after that. The broken line of the figure
shows the maximum amount of corrosion possible when a sufficient amount of the
liquid is present.
Figure 3—66 shows that potassium chloride (KC1) promotes the corrosion
process. The effects of various additives on corrosion are shown in Table 3—
18. CaO, MgO, and Fe203 reduced the corrosion because they neutralized the
acidity of the liquid. A1303 and SiOi had little effect. Corrosion was
increased substantially by NH4C1, KCI, and Fez(S04)3. This was due presumably
to the following reactions:
235
-------
TABLE 3 i7.
C
ChEMICAL COMPOSITION OF STEELS ( )
Si Mn P S
Cr Ni Cu Mo
SS 41 Mild Steel
——
——
—— <0.05
<0.05
Low—Alloy Steel
0.07
0.27
0.36 0.075
0.017
0.36 0.10
0.28
—
304 Stainless Steel
0.05
0.60
0.92 0.026
0.005
18.32 9.15
——
—
316 Stainless Steel
0.06
0.59
1.06 0.031
0.004
16.90 12.50
0.29
2.29
TABLE
3—18.
EFFECT OF ADDITIVES ON CORROSION OF LOW-ALLOY
AMMONIUM BISULFATE AT 200° FOR 75 HOURS.
STEEL
BY
Additive None CaO
MgO Al 3 03 Fe3 Oa S10 2
N.) Amount of
Additive (%)
Corrosion
(gi na)
Corros ion
rate
(g I in 3 / Ii r)
I CC 1
NII 4 C1 Fe 2 (SO 4 )a
0
18.3
20.0
16.7
20.6
37.5
25.9
27.5
27.0
22 .50
15.48
11.87
7.68
6.36
24.37
19.90
14.39
10.44
27.64
26.83
254.8
267.6
58.72
170.8
344.7
421.2
0.30
0.21
0.16
0.10
0.08
0.32
0.27
0.19
0.14
0.37
0.36
3.40
3.53
0.78
2.28
4.60
5.62
-------
E
z
0
(I ,
0
a:
a:
0
0
U.
0
F-
z
0
22
20
18
16
z
2 14
C / )
0
a: 12
a:
0
0
CL
0
I—
z 8
0
6
4
2
304 STAINLESS /
STEEL ‘c
/
/
N,)
4 - . )
—1
I I I
PURE NH 4 HSO 4 MELT
200 C
/
/
I I I I I
PURE NH 4 HSO 4 MELT
50-
200 C
do— / -
LOW ALLOY STEEL /
3Qi
- / SS41MILD
20 - • STEEL
10 - / ‘•‘04 STAINLESS STEEL
— 0 316 STAINLESS STEEL
I I I I I
0 20 40 60 80 100 120 140 160
TIME (h)
/
/
/
od 4
,-_o
—0
0, 316 STAINLESS STEEL
7’
—
1100 ALUMINUM
.-- a-•2-- -
I I i
1 2 5 10 20 50 100 200 500
TIME (hil LOGARITHMIC
FIgure 3 62 Ammonlum Bisultate CorrosIon Rate 01
Slalnless Steel and Aluminium
FIgure 361 CorrosIon at Steel by Ammonlum Bisultate
-------
I I I I I I
SS 41 MILD STEEL
40 200 C
Z 30 - — 14 SS4 1MILDSIEEL
0
PURE NH 4 HSO>/ 200CC 25 HR
o 12
a: z
a:
( ) 0 U) 10-
0 / 0
20- 0
0 00
0 8-\
I-
N) z
/0 wt°/o (NH 4 ) 2 S0 4 0 \
( ) 0
0
_ 9 + NH 4 HSO 4 MELT 6 \
10 -,_:, 8
4
0
/ 8
< 2
— — — -
C’ - I I I I I I
0 20 40 60 80 100 120 140 160 ‘ 0 2 4 6 8 io 12 14
TIME (fir)
(NI- 1 4 ) 2 S0 4 CONCENTRATION ImoI°/o)
Figure 3 63 Corrosion 01 Mild Steel by Ammonium Blsultate Figure 364 Eflect 01 (NH 4 ) 2 S0 4 on Corrosion 01 Mild Steel
Showing the Effect 01 Ammonlum Sulfate by Aminonium Bisulfale
-------
U I U I
e
250 / - — — — — € 1
I
300 :
— 200C
z
E25O
0
75HR
200 SS41MILDSTEEL 0
150
z
o 200 -
0
SS 41 MILD STEEL
0
1 4w 1% FI 2 SO 4 + NH 4 HSO 4
U-
20(YC a:
2 100® 0 150-
D
2 0
o 0
i.- 100-
2
50-
0
50 -
4
0 I I o —8— I I
0 20 40 60 80 100 120 140 0 o abs oi 015 02
TIME (01)
KCI C0NCENTAATIO 4 (mOI°/)
FIgure 3 65 Corrosion of Mild Steel by Sulfuric Acid Figure 3 66 ElIect of Potassium Chloride on CorrosIon 01
Containing Ammonium Blsulfate Mild Steel by Ammonium Bisulfate
-------
NH4 H SO. + NH4C1 ————> (NH4)z SO. + HC1 (1)
2NH4 HSO 4 + 2KC1 ————> (NH4)z SO. + KSSO4 + 2HC1 (2)
2NH 4 KSO4 + Fes(S04)3 —> 2NK4Fe(S04)a ÷ HSSO4 (3)
3.6.6 Ammonipm Bisulfate/Iron and Fly Ash Reaction Products (16 )
The author has studied the reactions of ammonitun bisulfate with iron and
fly ash at temperatures ranging from 1500 — 300°C. When ammonium bisulfate
reacted with iron the major products were NH4FeH(SO4) H:0, (NH4)311(S04)z,
(NH4)3Fe(S04)3, and NH4Fe(S04)a. These products indicate that the following
reactions occurred:
Fe + 4NH .HSO . + 1/2 03 —> NE4FeH(S04)z. + (NH4)311(S04)z (1)
Fe + 3NH 4 HSO . + 3/4 Oz —> (N114)IFe(S04)s + 3/2 HaO (2)
Fe + 2NB 4 ESO. + 3/4 03 —> NE4Fe(S04)z + NH3 + 3/2 HzO (3)
The first reaction (1) occurred readily at the low temperatures (150°—
200°) while the third reaction (3) occurred readily at the higher temperatures
of 250°C o above. The intermediate temperatures promoted the second reaction
(2).
Upon heating, the aluminum and calcium in the fly ash reacted with
aonium bisulfate to form NH 4 A1(S0 4 )3 and CaSO4 by the following reactions:
A1203 + 4MH4HSO4 —) 2NH 4 A1(S04)1 + 2NH3 + 3HiO (4)
CaO + NH4HSO4 —) CaSO4 + NTh + RiO (5)
These reaction products are the major compounds which compri se the de-
posits found in air preheaters and heat exchangers.
3.6.7 Compounds in Air Preheater and Heat Exchanger Deposits
( 13 .16 )
This author also used X—ray diffraction to determine the composition of
240
-------
the deposits found in a heat exchanger and two air preheaters. Table 3—19
lists the samples. Two of the deposit samples (A—i and A—2) were obtained
from a commercial heat exchanger for flue gas from an oil—fired industrial
boiler. In this system the 60°C gas leaves the FGD system while the hot gas
leaves the S R reactor at 400°C. Samples B—i and B—2 were obtained from an
air preheater at a S R pilot plant which treats coal—fired boiler flue gas.
Samples C—i and C—7 were obtained from seven different temperature zones of a
commercial air preheater. The preheater was used with a flue gas from an oil—
fired utility boiler (low—sulfur oil) subjected to SNR (Thermal DeNOx).
The results of the analysis of these samples are shown in Table 3—20.
Samples A—i and A—2 were composed of NH4FeH(SO4)z H2O. This compound is
formed by reaction (1) described in Section 3.6.6. A—i also contained
FeSO 4 HzO which did not form by the reaction of NH 4 BSO4 and Fe. This
indicates that sulfuric acid condensed and reacted with Fe to form the com-
pound; the flue gas is rich in S03 and lean in NH3. Samples B—i and B—2
contained NH2A1(S0 4 )a and CaSO4 in addition to fly ash and NH 4 Fe(S04)2. The
aluminum and calcium compounds were formed by the reaction of NH 4 HSO 4 with fly
ash from coal (Section 3.6.6).
Sample deposits C—i through C—7 formed in a gas which was rich in N H 3
and lean in S03. (N114)2S04 was present in the deposits in the low-temperature
zones (C—i through C—3) 1 while (NH4)3H(S0 4 )z was present in the deposits in
the intermediate zones (C—2 through C—4). Reaction products of N11411S04 with
Fe were present in C—4 through C—7, indicating that corrosion occurred in the
high temperature zones. All of the deposits contained a large amount of
unburned carbon, which indicated that carbon seemed to increase the amount of
deposition and promoted plugging.
Tests on the three sets of samples also revealed that almost no NLHSO4
was present in the deposits. NH4HSO 4 is very corrosive and had reacted with
Fe and fly ash to form various compounds.
241
-------
TABLE 3—19. DEPOSITS AND GAS COMPOSITIONS
System
Sample No.
Location
Boiler Fuel
SOa(ppm)
NR3(ppm)
A
A—l,A—2
Shindaikoywa
Petrochemical’ s
Yokkaichi
Plant
ugh—Sulfur
Oil
30—40
10
B
B—1,B—2
EPDC’s Test
Facility
Medium—Sulfur
Coal
3—15
10—20
C
C—i — C—7
Chuba Electric’s
Chita Plant
Low—Sulfur
Oil
2—3
20—30
3.6.8 Deposit Formation Reduction
Ammonium bisulfate deposition may be reduced in 5 main ways: 1) reduc-
tion of unreacted NEs; 2) reduction of SOs; 3) combustion control to reduce
soot in oil—fired burners; 4) use of coal fly ash for cleansing; and 5)
selection of appropriate air preheater elements/soot blowing systems. Reduc-
ing unreacted NH3 is the most commonly used method of decreasing bisulfate
formation.- This can be accomplished by lowering the inlet NB3INOX mole
ratio to less than 1.0 and simultaneously using a large amount of catalyst.
Keeping unreacted NH 3 below 5 ppm will substantially reduce deposition but may
not solve the problem entirely. Some of the S R plants maintain an un.reacted
NH3 concentration of less than 2 ppm to minimize deposit formation. (Sections
4.4 and 5.2). On the other hand, with a low unreacted NE3 concentration, aci-
dic deposits with NBS/S03 mole ratios below 1.0 may form, causing corrosion.
S03 can be reduced in 3 ways: 1) by using low—sulfur fuel; 2) by
combustion control in order to use a mini.mum amount of excess air; and 3) by
using a low—oxidation SCR catalyst. When the gas is lean in SOs and rich in
Nh, (NR 4 )3H(S04)z will be formed at higher temperatures and (N114)sSO4 will
form at lower temperatures. Both of these compounds are less corrosive than
NH 4 11S04. The deposits may be removed quite easily with the soot blower.
242
-------
For oil—fired boilers, combustion control to reduce soot (unburned car-
bon) is a good way to decrease deposits. Soot tends to form deposits along
with the NH 3 and S03 compounds. thereby causing an increase in plugging. The
cleansing effect of coal fly ash can be observed in a boiler burning pulver-
ized coal with a full dust load. At an adequate gas velocity there will be
little or no bisulfate problem primarily because the bisulfate deposits on the
ash. In this case, however, the unreacted N113 concentration in the flue gas
must be kept low in order to reduce the contamination of the ash by the am—
monium compounds. Selecting the appropriate air preheater elements and using
an effective soot blowing system also will minimize ainmonium bisulfate deposi-
tion. If these measures are used, the air preheater may be operated contin-
uously for over a year without bisulfate problems.
However, considerable deposition may still occur in some of the SCR
plants which use high sulfur oil or a low—dust system for coal. The deposits
usually do not hinder the heat transfer efficiency to any great extent, but do
increase the pressure drop of the gas in the preheater. When the pressure
drop exceeds a certain level, it is necessary to interrupt the operation of
the preheater for a water wash. Usually about one day is needed for the wash,
which includes the cooling of the preheater, water washing 1 and drying. Since
the wash water becomes acidic when it mixes with the bisulfate and related
compounds (Table 3—20), wastewater trea ent may be needed for neutralization
and also to remove ammonia.
Recently, Chugoku Electric developed a new method of removing deposits
called thermal cleaning which may be superior to water washing. (Section
4.3.4)
3 .7 OTHER PROBLEMS WITH AMNONIA
3.7.1 Introduction
In addition to the bisulfate problem described in the previous section.
ammonia also has the following effects on flue gas treatment systems:
243
-------
r’.)
TAULE
Sample
Number
Temperature
Zone
Estimated
tcmp.(C)
Major
Medium
Minor
b
phi
A—i
Low (cold)
90—110
Mlh.FcII(S04)a
UsO
FeSO. hliO
Nhi.IISO.
1
A—2
i.ow(lnt. 5 )
110—130
IULsFeII(S0 .)a
UsO
Nhi. Fe(SOm)
( iihlu) Fe(S0 .)i
1
li—i
lot.
About 200
Fly Ash
(N1h,)sfc(SOi)s
NU .Ah(S0.)i.CaSOI
1
1 1—2
m l.
About 200
Fly Ash
(Nll.hFc(S0i)
Nih .Al(SO.)i.CSSOI
1
C—I
Low (cold)
About 120
Carbon
(N 114)3S04
3
C—2
Low (hot)
>120
Carbon
(N1i.) iS O.
(Nhi .) iI l(SO .) .
1
C—3
lot. (cold)
About 160
Carbon
(NhIi)iFe(S0 .)
(NH .)iH(S0.) .
1
C—4
In C. (lot.)
>160
Carbon
(NIhs)ihc(S0 .)i
(N1i .) .ll(SO .)i
1
C—S
lot. (hot)
About 200
Carbon
N1l .Fc(S0 .) i
(NhI .) sFc(SO.) .
1
C—6
high (cold)
>200
Carbon
N1I .Fc(S0 .)z
(NhI.hFc(SO.)i
1
C— i
high (hot)
About 250
Carbon
Nli .F e(S0 .)i
I
lntcrmediatc
Measured roughly with 15
percent aqueous solution.
-------
(1) Dust removal improvement (with ESP’s).
(2) Ammonium bisulfate deposition on the catalyst.
(3) Fly ash contamination.
(4) Plume formation.
(5) Effects on FGD system.
3.7.2 Dust Removal Improvement
In Japan, ammonia injection is often-used to improve ESP dust removal.
Ammonia is injected into oil—fired utility boiler flue gas as it leaves the
air preheater. This is done in order to improve the soot removal efficiency
of the ESP and to prevent corrosion of the ESP by sulfuric acid mists in the
gas.
For flue gas from coal, the dust removal efficiency may be increased by
using a small amount of ammonia——up to 50 ppm for high sulfur coal, 20 ppm for
low sulfur coal. Larger amounts of ammonia may cause excessive adhesion of
fly ash to ESP collector plates (17). With baghouses, ammonia leakage from
ScR systems could potentially require more frequent cleaning and bag replace-
ment (17).- These effects, however, have not really been demonstrated in Japan
because of the low ammonia emissions, normally below 5 ppm.
3.7.3 Amnionium Bisulfate Deposition on SCR Catalysts
Ammonium bisulfate and related compounds tend to deposit on SCR catalysts
even at temperatures above 300°C; they deposit in air preheaters at a tempera-
ture usually below 2500C. Deposition at the higher temperature nay be caused
by the high S03 concentration on the catalyst surface due to the oxidation of
SOz, as well as to the high N113 concentration in the reactor. The critical
temperature for deposition may be as high as 350°C with an SOx—rich gas and
a high—oxidation catalyst.
The bisulfate essentially lowers catalytic activity by coating the cata-
lyst surface and may even poison some types of catalysts. Most catalysts used
for SOx—containing gases are not poisoned, however, and their activity can
245
-------
be recovered when they are heated to 400 450°C and the bisulfate is volati—
lized. With low—temperature catalysts used at 200—250°C, deposit formation is
a serious problem and necessitates occasional heating (Section 3.3).
In utility boilers, the economizer outlet gas temperature usually ranges
from 3500 — 400°C but may drop to 300—330°C when the boiler load is lowered at
midnight, for example. A small amount of the bisulfate may deposit on the
catalyst but is volatilized as the temperature increases with an increased
load, thus causing no problem for SOx—lean gas. With SOx—rich gas, how-
ever, the NOn removal efficiency may be noticeably lowered when the gas
temperature is kept below the critical temperature for several hours.
In order to maintain the temperature above the critical level, an
economizer by—pass system has been used with some coal—fired utility boilers.
Oil—firing has been used for the same purpose at smaller gas sources. A
moving—bed reactor is used for treating a low—temperature gas of 200—250°C;
the contaminated catalyst is removed from the reactor and heated.
3.7.4 Contamination of Fly Ash by Ammonia
Ammonii.im bisulfate may deposit on fly ash in the gas stream (unreacted
N113) at temperatures below 250°C. Figure 3—67 shows the relationship between
unreacted ammonia (N113 leakage) and the ammonia content of fly ash caught by a
cold ESP downstream of an air preheater. The ammonia content increased almost
linearly until it reached about 800 mg/kg at about 20 ppm of unreacted NH 3 in
the gas. It did not increase significantly after that, presumably because of
the limited S03 content of the gas which combines with N113 to form the de-
posits. Figure 3—67 indicates that the ammonia content of the ash was nearly
200 mg/kg (0.02 ) with 5 ppm unreacted NH3 and around 600 mg/kg with about 15
ppm unreacted NH3.
When used in concrete or landfill materials, an NH3—rich ash may produce
an ammonia odor. This occurs when the alkaline component of the ash reacts
with water. This reaction can be expressed as:
246
-------
1000
•0
0
:::i
400—
z
0
0
I
Z 200 —
0 I I
0 20 40 60 80 100
UN REACTED NH 3 (ppm)
Figure 3.67 Adsorption of Unreacted NH 3 on Fly Ash. (6)
247
-------
2NU4HSOI + CaO + RiO = CaSO4 2HzO + Jfl3
When NH3 is present in the ash, it also nay cause water pollution pro-
blems. For these reasons it is best to maintain an acceptable level of NH 3 in
the ash. It has been found that an NH content below 0.02% works well, when
ScR reactor outlet gas contains 5 ppm NH 3 and 90% of the NH3 is caught by the
ash present in 20 grams/Nm 3 in flue gas (20 gm/Nm 3 ) the NH 3 content of the fly
ash is 0.14%.
Chugoku Electric’s Shimonoseki Power Station (Section 4.4) maintains
unreacted NH 3 below 2 ppm and has produced fly ash which can be used without
problems for cement or concrete, or for landfill material.
If necessary, NR3 may be removed from the fly ash at the power plant by
mixing it with a small amount of moisture. This causes reaction (1) to occur
and release NH ; the released NH3 can be emitted from the stack without any
trouble. Moistening the ash also can help prevent dust formation during ash
handling.
3.7.5 Plume Formation
Plume problems caused by ammonia in flue gas have been known to occur
when the gas from an ammonia—scrubbing FGD plant contains more than 10 ppm
NH 3 . On the other hand, there is no plume problem associated with flue gas
which has undergone SCR. This is true even when the gas has been cooled to
140°C and contains 20 ppm NH3. A plume has been observed during the winter
with flue gas containing 50 ppm NH 3 after selective noncatalytic reduction
(thermal De—NOx) (Section 6.2).
The difference in the plume formation associated with wet FGD and dry De—
NO 1 systems is due to differences in moisture content and temperature as
well as in the amount and particle size of ammonium compounds in the flue
gases. With wet FGD, the 55°C moisture—saturated gas contains a large amount
of SOi and a considerable amount of S03 before it is heated to 70_1200C. In
the dry DeNO 1 system. flue gas containing a smaller amount of moisture is
248
-------
cooled from about 400°C to 150°C in an air preheater. In the preheater nearly
all of the SO 3 combines with N113 to precipitate out as long as NH3 is present
in excess.
Plume problems may occur when flue gas containing a considerable amount
of N113 after dry DeNO is introduced into a wet FGD system. This is
described in the following section.
3.7.6 Effect of Ammonia on Flue Gas Desuifurization
The effect of ammonia on flue gas desulfurization varies with the FGD sys-
tem being used. Three types of wet process FGD systems popular in Japan are
shown in Figure 3—68. No. 1 shows limestone scrubbing with a prescrubber in
which over 90% of the ammonia in the gas is caught in the prescrubber by an
acidic liquor with a pU of about 1. A typical FGD system for a 175 MW coal—
fired boiler releases 5/hr of wastewater from the prescrubber. The
ammonia content of the wastewater is calculated at roughly 200 ppm, assuming
an inlet gas ammonia concentration of 3 ppm. The liquor is sent to a waste—
water treatment system where ammonia may be removed either by a conventional
activated sludge process or an ammonia stripping process, if necessary.
Ammonia stripping has been applied at Chubu Electric’s Owase Plant (18); the
activated sludge process will be applied at the EPDC’s Takehara Plant. The
prescrubber outlet gas contains only a small amount of ammonia (below 1 ppm),
most of which is caught by the scrubber but does not affect the scrubber sys-
tem.
No. 2 shows a two—stage limestone scrubbing system. The first scrubber
catches 60—70% of the ammonia in the flue gas while the second scrubber re-
moves another 10—20% of the ammonia, which adds up to a total removal effi-
ciency of 70—90%. The ammonia in the scrubber liquor may increase SOz removal
efficiency and may not produce any adverse effects on the scrubber system
(17). The by—product sludge, however, may cause environmental problems.
In Japan, sludge is oxidized to produce gypsum which is subsequently
centrifuged and washed with water for use as cement, wallboard, and other
249
-------
NO 1
NO 2
GAS - —
SLURRY
NO 3
NaH SO 3
GAS
WATER
Figure 3-68 Three Most Popular Wet Process FGD Systems Used in Japan
250
TREATED
GAS
GAS - —
GAS
r — — —
SECOND SCRUBBER
FIRST SCRUBBER
TREATED
GAS
GAS ¶- - —
-------
materials. Both the filtrate and the wash liquor are recycled in the scrubber
system and a portion of the scrubber liquor is sent to the wastewater treat-
ment system. If necessary, ammonia may be removed from the liquor to prevent
environmental problems.
In FGD systems No. 1 and No. 2, flue gas leaving the scrubber contains
little ammonia and will not cause a plume problem except in cases when inlet
gas containing over 20 ppm NH3 enters System No. 2.
No. 3 shows simple sodium scrubbing which produces throw—away sodium
sulfite or sulfate liquor. Approximately one—half of the NRs in the flue gas
is caught by the scrubber liquor. Removal of ammonia from the waste liquor
may be necessary under certain circumstances. (A plume can be observed at the
scrubber outlet of this system when the inlet flue gas contains more than
about 20 ppm N113.)
3.7.7 Measurement of Ammonia in Flue Gas
The measurement of unreacted NH 3 in flue gas is important both for con-
trol of NO removal efficiency and prevention of the problems described
above. However, this measurement can be difficult for the following reasons:
(1) Ammonia is usually present in a small concentration, below 10
ppm.
(2) The presence of NOR, SO 1 , etc. interferes with the measure-
ment of ammonia.
(3) At a large plant, the ammonia concentration may be different in
each part of the large duct, causing sampling problems.
The major analytical methods used to measure N113 in Japan are shown in
Table 3—21 and Figure 3—69, I and II. Method I is used to measure the concen-
trations of NO 1 and a total of NO 1 and NH3, after N113 is converted to NO;
the N B 3 concentration equals the difference between these two concentrations.
251
-------
TABLE 3—21.
BV0 MAJOR MEI1IODS USED IN JAPAN FOR CONTINUOUS ANALYSIS OF Nih IN FLUE GAS
Method No. I II
Principle
Oxidation of Nib to NO, followed
by Chemiluininescence
Ultraviolet ray absorption by Nib
Major Manufacturer
Shimazu Seisakujo
Fuji E1ectric 1
Anritsu Electric
Example of User
Chugoku Electric’s Kudamatsu Power
Station (Commercial SCR plant)
Electric Power Development Co.’s
Isogo Power Station (Ill! SCR pilot
plant)
Electric Power Development Co.’s Isogo
Power Station hitachi Zosen
process SCR pilot plant
Mitsubishi Heavy Industries. Takasago
Research Station
Catalytic converter of NR a to NO
as needed. Since a small portion of
Nib is converted to NOa, en taly tic
convertor of NOi to NO is also
needed.
For a gas rich in NO , the
analytical error may be large.
Interfered with SOi in the gas,
although it may be possible to
correct the analytical value by also
measuring SOz ,
The gas sampling tube should be
kept above 300°C to prevent the
condensation of N11 4 11S04 in the tube.
Maintenance
Not simple
Simple
Price of the
Analyzer
5.5 million yen including converter
4.9 million yen for Nil 3 analyzer
5.5 million yen for Nib and SOi
analyzer
r ’J
LJ
Probl ems
-------
II
I II
TUBE SHOULD BE
KEPT ABOVE 300°C
FOR NH 3
AND 502
ULTRAVIOLET. REMOTE
(ANRITSU)
ULTRAVIOLET. DIGA
N )
U i
U)
GAS
FOR NO
DETECTOR
A NH 3 NO
BNO 2 NO
C, C’ CHEMILUMINESCENCE
DIFFERENCE BETWEEN
C’ AND C SHOWS NH 3
FIgure 369 ComparIson of NH 3 Analysis Methods
-------
With this method there may be a large analytical error when the NO 1 concen-
tration is much higher than the NE3 concentration, as is often the case with
flue gases after SCR. A modified method for the determination of a low con-
centration of N113 has been developed, and will be described in Section 4.4.3.
Method II is used for direct analysis of NH . However, any SO 1 present
in the gas interferes with the measurement, requiring a special calibration in
order to take the interference into account. When using this method, the ana-
lytical error may be large if the gas contains a large amount of SO 1 and a
smaller amount of NH 3 . The gas sampling tube also must be kept above 300°C to
prevent ammonium bisulfate deposition which can cause additional analytical
errors.
Although the analysis itself is fairly accurate when Methods I and II are
used properly, gas sampling may cause errors since gas composition may not be
homogeneous throughout a large duct. The Direct In—Gas Analysis (DIGA) method
developed in the U.S. (Figure 3—69, III) may be superior from this point of
view. With the DIGA method a light source is placed on one side of the duct
and a detector on the other. Light passes through a pipe across the duct
which is slotted to allow gas flow. This method allows quick response which
eliminates sampling error. However, the accuracy of the N 3 measurement is
affected by the presence not only of SOa, but also of particulates, requiring
calibrations according to their concentrations.
254
-------
REFERENCES
1. Atsukawa, M., et al. Development of NO 1 Removal Processes with
Catalysts for Stationary Combustion Facilities, ’ Mitsubishi Technical
Bulletin, No. 124, Iay 1977. (In English)
2. Nartita, T., and H. Kuroda. Babcock—Hitachi NO 1 Removal Process for
Flue Gas from Coal—Fired Boilers, presented at the Joint Symposium on
Stationary Combustion NO 1 Control, Volume II (EPA—EPRI), October 1980.
3. Catalyst and Chemicals Inc. S0 1 —Resistant Inexpensive SCR Catalyst
from Red Mud, etc., Steel Fund Study Report No. 2—10, Steel Industry
Foundation for the Advancement of Environmental Protection Technology,
Tokyo, Japan. May 1978. (In Japanese)
4. Sakai Chemical. R&D for Low—Temperature S R Catalyst for Sintering
Machine, Steel Fund Study Report No. 2—20, Steel Industry Foundation for
the Advancement of Environmental Protection Technology, Tokyo, Japan.
March 1980. (In Japanese)
5. Mitsui Toatsu Chemical. Study on Low—Temperature ScR, Steel Fund Study
Report, No. 2—22, Steel Industry Foundation for the Advancement of Envt—
ronmental Protection Technology, Tokyo, Japan. March 1980. (In Japanese)
6. Itoh, II., and Y. Kajibata. Countermeasures for Problems in NO 1
Removal Processes for Coal—Fired Boilers, presented at the Joint
Symposium on Stationary Combustion NO 1 Control, Volume II (EPA—EPRI).
U.S. Environmental Protection Agency, Research Triangle Park, October
1980.
7. Sengoku, T., and B. M. Howell, et al. The Development of a Catalytic
NO 1 Reduction System for Coal—Fired Steam Generators, presented at the
Joint Symposium on Stationary Combustion NO 1 Control, Volume II (EPA—
EPRI), U.S. Environmental Protection Agency, Research Triangle Park,
October 1980.
8. Ando, I. NO 1 Abatement for Stationary Sources in Japan, EPA—600/7—79-
205, U.S. Environmental Protection Agency, Research Triangle Park, August
1979. p. 280.
9. Hitachi Zosen. R&D of Low—Temperature S R Catalyst for Sintering Machine
Flue Gas, Steel Fund Study Report, NO. 2—21, Steel Industry Foundation
for the Advancement of Environmental Protection Technology, Tokyo, Japan.
March 1980. (In Japanese)
10. Nakabayashi, Y., and S. Niwa. Characteristics of Cylindrical NO 1
Catalysts for Coal—Fired Boiler, presented at the EPRI NO 1 Symposium,
November 1979.
255
-------
11. Seiyama, T., et al. Catalytic Reduction of Nitric Oxides with Ammonia
over Transition Metal Ion—Exchanged Y Zeolites. Chemistry Letters,
chemical- Society of Japan, pp. 781—784, 1975. (In English)
12. Kato. S.. 3. Ando, and H. Tohata. Fundamental Study on Precipitation of
Ammonium Sulfate and Bisulfate, Report of Chemical Engineer’s Associa-
tion, December 1978. (In Japanese)
13. Burke, J.M., and LL. Johnson. An Investigation of Ammonium Sulfate!—
Bisulfate Formation and Deposition in Air Preheaters, U.S. Environmental
Protection Agency, January 1980.
14. Ando, 3. Fundamental Problems of Sulfur Oxides and Flue Gas
Desulfurization, Industrial Pollution Control, (12) No. 9 1976. (In
Japane se)
15. Ishikawa, T. Basic Study on Steel Corrosion in Ammonium Bisulfate
Melt. Report of Chemical Engineers Association, December 1978. (In
Japanese)
16. Ando, 3., and K. Takemura. The Deposits in Heat Exchanger after Flue Gas
Denitrification Using Ammonia. 3. Chem. Soc. Japan, No. 1, 160—162, 1980.
(In Japanese)
17. Noblett, 3., R. Glover, and T. Shirley, and M. Bebets. Impact of NO
Selective Catalytic Reduction Processes on Flue Gas Cleaning Systems,
U.S. Environmental Protection Agency, Research Triangle Park, NC,
December 1979. (In English)
18. Ando, 3. SOi Abatement for Stationary Sources in Japan, A—60O/778
210, U.S. Environmental Protection Agency, Research Triangle Park, NC,
November 1978. (In English)
256
-------
SECTION 4
SCR FOR FLUE GAS FROM 1]T]11TY BOILERS
4.1 DESIGN AND PERFORMANCE OF SCR PLANTS
4.1.1 Development of SCR for Utility Boilers
Following numerous pilot plant tests, the first prototype SCE plant was
constructed in 1977 at Kansai Electric’s Kainan Station. It was designed to
treat one—fourth of the flue gas from an existing 450 MW oil—fired boiler by
reducing NO 1 from 120 ppm to about 20 ppm. The plant uses a ring—type cat-
alyst in a fixed bed. The first commercial SCR plant for utility boilers
began operation in 1978 at Chubu Electric’s Chita Station. Using a pellet
catalyst, the plant reduces NO 1 in the flue gas of two 700 MW gas—fired
boilers from 50 ppm to 10 ppm.
Since 1978, parallel flow SCE catalysts, especially the honeycomb and
plate types, have significantly improved in quality and decreased in cost. In
1979 the first full scale S R plant using a honeycomb catalyst began operation
at Chugoku Electric’s ndamatsu Station. It reduced NO 1 in the flue gas of
an existing 375 oil—fired boiler from 130 to 20 ppm. All of the coercial
SCE plants constructed in 1979 for utility boilers use honeycomb or plate
catalysts, except for one plant which uses a tubular catalyst. Even for clean
gas containing little dust, the parallel flow catalyst with a thin wall is
superior to the pellet catalyst because of the lower gas pressure drop it
causes.
257
-------
Full—scale plants for testing a combined system of selective noncatalytic
reduction (SNR) and S R applied to existing oil—fired boilers were constructed
in 1978 by Tokyo Electric and Kansai Electric. The tests were subsequently
discontinued and the SNR system was removed leaving the SQ system (Section
6.3).
For coal—fired boilers, the first full scale S R plant began operation in
1980 at Chugoku Electric’s Shimonoseki Station. The station treats flue gas
from an existing 175 MW boiler using a honeycomb catalyst. In 1980, another
SCR plant began operation at Bokkaido Electric’s Tomato—Atsuma Station. It
uses a plate catalyst to treat one—fourth of the gas from a new 350 MW boiler.
Seven S R plants for coal—fired boilers began operation in 1981 and 1982.
By the end of 1982, the total S R capacity for utility boilers will
exceed 21,000 MW, including about 3,300 MW for coal—fired boilers. S R plants
with a capacity of treating more than 250,000/Nm 3 /hr of flue gas (75 MW equiv-
alent) from utility boilers are listed in Table 4—1. Almost all of the coal—
fired boilers which will begin operation after 1982 will need SCR, although
NO 1 can be reduced to 160—200 ppm using combustion modification alone.
Most of the S R plants in Japan are designed for 80—85 percent NO 1
removal with unreacted NU3 below about 5 ppm. Although it may be possible to
remove 90 percent or more of the NO 1 , 80—85 percent removal has been found
to be the optimum level because of the smaller pressure drop, smaller amount
of unreacted NR3, and lower SO2/S03 oxidation rate associated with it. Some
of the S R plants initially are using a small amount of the catalyst to remove
50—60 percent of NO 1 and will increase the amount of catalyst to remove 80—
85 percent when needed in the future.
All of the SCR plants for utility boilers have been or will be con—
structed by boiler manufacturers; 11111, BEK, ml, and I. ‘fflI, the largest
boiler manufacturer in Japan, builds the largest share of the ScRs; normally
the SCR plant is constructed by the boiler manufacturer.
258
-------
TABL.E 4-1. SCB PLANTS FOR UrILITY BOIL S IN 3APAN AND THE UNITED STATES (By 1984)
Ro teal it
N e.
Mltsabiohi flescy Zadnatries
“H.bcock Hitichi
7 1 .bik ..aJ i.e B .rz i Besay Isduitriel
Ba. . .ki Ue.vy Industries
Boiler
Cisuc it,
1000
Cit.—
Stirt
Poser
Cosp.ny
Statios
Nise
No. Feel
El
)&/br
N/B
Vindor
lyst
Up
Chobu Elecric
Chit.
4
Oi1. .s
700
2,000
B’
BEI
B 1
0 b
1979
5
G.e
100
1,910
Nb
BOB
0
1978
1978
.
6
Ci.
700
1 ,910
1983
1
OIl
375
1.100
181’
.
Ni .b l a .goys
6
011
500
1.320
B
BffI
P 1
4
OIl
375
1,100
B
BOB
P
1981
1982
3
011
313
1,100
I
B
1980
“
Shi,n , 1 oy i
3
OIl
220
650
B
B
I
B
1981
•
2
Oil
220
650
II
1981
“
.
5
011
220
650
B
0
1981
.
6
Oil
220
650
B
B
1980
At.isi
3
Oil
700
1.870
N
I I
B
1981
“
4
Oil
700
1.870
N
BOB
p
1983
.
Chit. No. 2
1
Gas
700
1.910
B
1983
2
Cs.
700
1.910
N
18!
B
1979
Cbaloku Electric
Budiastlu
2
Oil
375
1.125
8
IHI
0
1979
.
.
3
Oil
700
2.000
NB!
B
1980
j..keai
3
Oil
350
1.000
N
NB!
0
1980
‘
2
Oil
500
1.450
B
NBI
B
1980
Sbimono .ekI
I
Coil
175
580
NB!
0
1982
Sbin—Ubo
1
Coil
75
250
1982
“
•
2
Coil
15
250
B
NB!
NB!
B
1982
3
Coil
156
500
B
P
Hizesbims
1
Coil
12$
400
2
BOB
BEl
P
1984
1984
2
Coil
156
500
Bo aey cosb
1 0r . a a le
Plate
259
-------
TABLE 4—1 SCR PLANTS FOR UTILITY BOILESS IN JAPAN AND TB! UNITED STATES (By 1984) (Continned)
2.trof It
N. .
°Nit,nbj,bj Sitny lodnutri,,
u 1 8.bcock flitichi
‘ Iibtkiv.jini firm, Reavy 1nd stri .s
E..sskl Buoy Induatrie ,
flon.yen u ib
0rinnl.
Plate
Tub.
.6th SN! isa (sited.
Can so jly___
Po..r
Co.psny
Ststion
Na.,
Boiler
No. Foil
8!
1000
W ’Ihr
N/ a
Vendor
Cats—
lyit
Start
Up
El.ctrto Po.er DC
Takehuru
I
Coal
250z1/2
ZSOzlI2
400
400
2
2
888
DI
P
1
1981
1981
3
Coil
700
2.340
N
881
P
1983
Bokk.Ido Electric
To .sto—Atauasl
1
Coil
350z1/4
280
N
881
P
1980
Qokeriku Electric
Toys..
2
Oil
500
1.470
N
18!
8
1980
Jobun Electric
Nakoso
8
Oil.Coul
400
1.810
Nb
N!1°
R
1982
C
C
9
Oil.Coal
600
1.810
N
12 !
8
1983
C.osai Electric
Oust.
1
Oil
156
500
2’
I
H
1978
2
Oil
156
500
8
X
B
1981
C
3
Oil
156
500
2
I
8
1980
C
.
4
Oil
156
500
2
83!
B
1980
(sinai
1
Oil
450.1/4
300
831
P 1
1979
•
C
3
Oil
600
1.640
2
83!
8
1982
C
4
Oil
600
1.640
8
881
P
1981
C
Hi.uji I
4
Oil
156
500 k
28!’
H
1978
.
Risiji II
6
0... i1
600
1.640
8
131
8
1981
C
A .alaaski—
1
Oil
156
470
8
888
P
1978
C
bi 8 .uhi
2
Oil
156
470
8
881
P
1980
A.asausk l—
I
Oil
156
470
8
831
P
1981
“
No. 3
2
Oil
156
470
8
881
P
1980
C
C
3
Oil
156
470
8
881
P
1980
Ssk.lko
1
Gii. il
250
800
8
83!
8
1980
e
2
Ga..oil
250
800
8
83!
H
1981
C
4
Gas.oil
250
800
8
83!
B
1981
260
-------
TABLE 4—1 SCR PLANTS FOR UTILITT DOILR!S IN JAPAN AND THE UNITED STATES (By 1 984) (Conti need)
R.trof it
N..
)!lti biahi Resey lodnutries
Hsbcock Butichi
1sh1k...jima Usrios lierey Indoatrias
Eavesk) Benvy Indoatries
flon.ycoiib
1 0r son 1.
Plate
Tub.
cith SNR eat testid
‘Coal—Oil ‘l ) ar
Po .e
Station
Roller
Capacity
1000
Csta
Start
Coepany
Na ..
No.
Feel
N V (‘&Ihr N/B
Vendor lyot
Up
[ snail Electric
Sskoiko
5
Gsa,oil
250
800
B
RB!
P
1980
.
6
G... t1
250
800
B
89!
B
1979
“
7
Gsa,oll
250
800
B
89!
B
1981
a
1 10 1 15 .5
3
Oil
156
500
8
891
B
1981
‘
4
Oil
156
500
8
l 9
B
1981
hoots.. II
2
Oil
600
1.320
8
BR!
P
1981
“
C u uu ido
1
Oil
156
470
8
IHI
B
1980
-
•
2
Oil
156
470
R
mi
B
1981
“
Aioi
I
Oil
375
1.030
Nb
mi°
8
1982
2
Oil
375
1.030
N
ml
B
1982
a
3
Oil
375
1,030
N
Ifli
B
1982
‘
Gobos
1
Oil
600
1.600
N
1 (91°
B
1984
2
Oil
600
1.600
N
BR ! 6
Nb
1984
.
°
3
Oil
600
1.600
N
891
B
1984
Ipeahe Elootric
Shinkokoy.
3
(hi
600
1.640
N
891
0
1975
.
°
4
Ge.
600
1.640
N
891
0
1978
.
5
Gee
600
1.640
N
1 191
0
1982
.
Omen
2
Coal
156
500
8 ’
1191
B
1982
Ninato
1
Coal
156
500
8
1(91
B
1983
Tekoka Electric
R.N u 1 1 .ta
2
011
600
1.640
N
191
B
1981
“
Sandal
3
Coil
175
580
8
991
P
1982
Tokyo Electric
Yokosaks
4
011
350
1 • 060 k
1
891
B
1978
Tokyo Electric
Tokosaka
1
COIl 1
265
850
8
89!
B
1984
Tokyo Electric
Tokounks
2
CON 1
265
850
8
891
B
1984
( hI
2
Oil
350
1 , 060 k
B
191
9
1978
.
Tokuhs .s
1
Oil
156
500 k
OH!
P
1978
S. Celif. Ediaon
llnntlnSIon
2
Oil
215.1/2
350
8
I
8
1981
261
-------
For oil— and coal—fired boilers, MHI and IBI use a honeycomb catalyst
while BBK uses a plate catalyst (Figure 3—33). KflI has constructed an SCR
plant for a coal—fired boiler at EPDC’s Takehara Station which uses a tubular
catalyst. However, l BI probably will use honeycomb catalysts for future
plants.
Compared with FGD, S R is much simpler and costs about one—third to one
fourth less. Au SIR unit is also much easier to operate and does not require
an operator. When applied to oil—fired boilers. SIR plants have experienced
almost no problems except for ammonium bisulfate deposition in air preheaters
(Section 3.6).
4.1.2 Design of SCR Plants
A typical design of an S R plant for an oil—fired utility boiler is shown
in Figures 4—1 and 4—2. Flue gas leaving the boiler economizer outlet at 300—
400°C is mixed with ammonia and sent into an S R reactor. Although one reac-
tor may be adequate to treat flue gas from a 700 MW boiler, two reactors in
parallel also may be used with a boiler larger than 400 MW. The flue gas
leaving the reactor is passed through an air preheater, an ESP, and then a
scrubber, if necessary. The flue gas from an oil—fired boiler may be passed
horizontally or up through the reactor while that from a coal—fired boiler
must pass downward to eliminate dust plugging.
Figure 4—3 shows an example of an ammonia injection system. Generally,
the ammonia flow rate is controlled by the flue gas flow rate and the incoming
NOx concentration. In a large boiler it is difficult to precisely measure
the flue gas flow rate. The steam flow rate to the turbine is used as a mea-
sure of the amount of fuel being burned and the combustion gas produced. This
control method can be finely tuned using the outlet NO 1 level as a feedback
signal.
When a boiler is operated at a low load, a low—temperature economizer
outlet gas is produced. When this happens ammonat bisulfate deposits may
262
-------
PLAN
Figure 4.1 SCR Plant Layout (IHI)
DENITRIFICATION REACTOR
ESP
SIDE VIEW
263
-------
-
=
= air
= air
NH 3 also
conlrol signal
Flue Gas
LIQUID
AMMONIA TANK
---0
Figure 4 2 Flowaheel (or Standard Boiler Denitrilication Equipment Used with
Heavy Oil Combustion Gas (iHI)
1
I
LOAD
r BOILER
SIGNAL I I
I I
STEAM/AIR
0 ’
BOILER
LOAD
SIGNAL
L )JJ
STACK
STEAM/AIR
HEATER
11= = = = = = = ii
II
II
II
L ft
EVAPORATOR
-------
AMMONIA
DISTAl
ADDER
RELAY
( \
NO,
OUTLET
‘7
AIR SUPPLY
I ‘ —
MULTIPLIER =
TOTAL NO, INLET
\ iii)
NO ,‘ FLUE GAS
INLET FLOW
Figure 4.3 Ammonia Inlection System (1)
265
-------
decrease the SCR catalyst’s activity (Section 3.7.2). To eliminate the prob-
lem, a gas heating device can be installed on some boilers. The gas is heated
either by an economizer by—pass as shown in Figure 4—4A or by modifying the
economizer as shown in Figure 4—4B. The modified economizer has two heat
transfer compartments, one with high efficiency and the other with low effi-
ciency. Most of the gas is passed through the low—efficiency compartment at
the low load to maintain the gas temperature above a desired level. Using
such a device lowers the heat recovery efficiency at the economizer but allows
the heat to be recovered by the air preheater downstream of the SQ reactor,
thus avoiding a loss of energy.
The SIR reactor usually has a soot blowing system which cleans the cata-
lyst using hot air or steam. An example of a soot blowing system is shown in
Figure 4—5. However, a soot blower has not been used with most of the reac-
tors which utilize parallel flow catalysts.
4.1.3 SCR Performance
The typical performance of an SCR unit used with a utility boiler is
shown in FIgure 4—6. When the boiler load is reduced, the gas temperature is
substantially lowered and the NO 1 removal efficiency remains almost the
same. In this situation, the reduced flue gas volume results in a smaller SV,
or longer reaction time. The SOz oxidation ratio also remains essentially the
same despite a decrease in the boiler load. However, both unreacted NH3 and
pressure drop are lowered with the load.
Below a certain temperature, annonium bisulfate (or related compounds)
deposits on the catalyst and lowers catalytic activity. The critical tempera-
ture depends upon the inlet concentrations of SOx and N Ih and on the degree
of SOz oxidation by the catalyst. For a flue gas from a low—sulfur oil, the
critical temperature is typically around 3000C; for flue gases from coal and
high—sulfur oil, it is 20_400C higher. If the gas temperature remains below
the critical temperature for a few hours a day (for example during the low
boiler load at midnight) and is kept around 400°C for the remainder of the
266
-------
A: ECONOMIZER BY.PASS
I —
/
LARGE
ECONOMIZER
GAS FLOW, %
HIGH LOAD 50 50
LOW LOAD >50 <50
B: SPLIT ECONOMIZER ARRANGEMENT (3)
Figure 44 Economizer Outlet Temperature Adjustments.
SMALL
ECONOMIZER
267
70A2642
-------
SOOT
CATALYST BED
Figure 4 5 Soot Blower Used with an 5CR Reactor (1)
-------
GAS FLOW RATE PRESSURE DROP INLET NO, EFFICIENCY
(x 1O Nm 3 /h) (mmH 2 O) (ppm) (%)
C
0 )
u, 0 0 0
a u. 0 0 0 0
— — I I I I I I
rn
V
0
C,
-v
0
C,
“4
C) m (I ) — m
a (p Z m
0 C —
i i
rr I
U) \
o 0 z (P
a rn
- z
C)
a
-4 0 -(
U)
(A z -v
3 ‘9
C)
= 33
I I I I I I I I I I
w 0 - on a —
0 0 th 0
I
SO 2 CONVERSION SLIP NI- i 2
TO SO 3 (°/ ) (ppm)
8
a
0 0
(n
L I I -
0
U) U)
0 ( 7 1 0
a o 0
GAS TEMPERATURE (C)
(N H 3 )!(N O )
-------
day, the small amount of bisulfate formed on the catalyst at the low tempera-
ture will be removed at around 4000C. Thus, there will not be much of an
adverse effect on the overall NOx removal efficiency. If the gas tempera-
ture is maintained below the critical temperature for a longer period, the
bisulfate deposition may become pronounced. To eliminate this problem the gas
temperature should be maintained above a certain level, as previously de-
scribed in Section 4.1.2.
For most utility boiler SCR plants, an NE3/N0 1 mole ratio of 0.83—0.90
has been used to remove 80—85 percent of NOx while maintaining unreacted N
below 10 ppm. Soot blowing is seldom used and thus no operator is needed for
routine operation.
The remaining problem associated with S R is ammonium bisulfate deposi-
tion in the air preheater. In most of the S R plants the problem has been
solved by maintaining unreacted NE3 below 5 ppm, and using combustion control
to reduce S03, and low—oxidation catalysts. Some of the plants also use im—
proved air heater elements and intensive soot blowing to remove the deposits.
A small amount of ammonium bisulfate may still deposit despite these prac-
tices, but it does not affect SCR operation. If necessary, the deposits can
be removed when the boiler is shut down for its annual maintenance; in Japan
larger boilers are required by law to perform this yearly maintenance.
Water washing also has been used to remove the bisulfate deposits. About
24 hours must be allowed for washing, which includes time for cooling, wash-
ing, and drying. When water washing is necessary two or more times a year,
one of the two parallel air preheaters is shut down for each washing while the
boiler operates at a half load using the other air preheater. Recently a new
technology — thermal cleaning — was developed for air preheater cleaning (Sec-
tiOn 4.3.4).
270
-------
4.2 SCR PLANTS OF CHUBU ELECTRIC POWER COMPANY
4.2.1 Introduction
Chubu Electric Power Company is the third largest utility company in
Japan with power stations located in the Chubu area which is between the Kanto
Area (including Tokyo) and the Kansai Area (including Osaka). Since most of
its steam power generating stations are close to cities and industrial re-
gions, extensive air pollution countermeasures have been taken. These include
FGD and the use of low—sulfur fuels for SO 1 abatement, and combustion modifi-
cation. ScR, and selective noncatalytic reduction (SNR) for NO 1 abatement.
Chubu Electric’s S R plants are shown in Table 4—2. Since the three
plants constructed at the Chita Station in 1978 and 1980 worked so success-
fully, Chubu Electic installed nine more units in 1980 and 198 . and will con-
struct four more in 1982 and 1983.
4.2.2 Chita Power Station
The Chita Power Station, near Nagoya City (population, 2 million), has
six boilers (Table 4—3) which use the most advanced NO 1 control systems for
utility boilers in Japan.
Extensive combustion modification has been applied to the boilers at the
Chita Station, including staged combustion, flue gas recirculation, low excess
air combustion, and low NO 1 burners. The 350 ppm NO 1 concentration in
flue gas from oil—fired boilers is reduced to about 100 ppm (a 70 percent
removal rate) by combustion modification.
The No. 5 and No. 6 LNG—fired boilers began operation in 1978. In order
to obtain permits from the local authorities (Aichi Prefecture and Chita City)
for the construction of two new boilers, Chubu Electric promised to use LNG
fuel and to install SQ units to minimize NO 1 emissions. In addition, Chubu
Electric installed an SNR system for the No. 2 boiler and an SCR unit for the
271
-------
TABLE 4-2. CIJUBU FLErrRIC’S SCR PLANTS
Boiler
Station No.
Capacity
(MW)
NO
NIR Fuel Inlet’
(ppm)
Removal
(Percent)
Completion
Date
Outlet
Chita 5 700 N Gas 50 10 80 Mar. 1978
6 700 N Gas 50 10 80 Apr. 1978
4 700 P. Gas,oil 100 20 80 Mar. 1980
1 375 R Oil 135 70 50 Dcc. 1982
Nishinagoya 6 500 R Oil 100 20 80 Jul. 1980
4 375 R Oil 100 20 80 Mar. 1981
3 375 R Oil 100 20 80 Jan. 1982
NJ
Atsumi 3 700 N Oil 80 16 80 May 1981
4 700 H Oil 80 16 80 June 1981
Shinnagoyn 3 220 R Oil 75 50 33 Sept. 1980
2 220 R Oil 90 60 33 Apr. 1981
5 220 R Oil 75 50 33 Sept. 1981
6 220 R Oil 85 60 29 May 1981
Chita 1 700 N Gas 50 10 80 May 1983
No. 2 2 700 N Gas 50 10 80 Jul. 1983
Ncw or rctrofit.
-------
TABLE 4-3. QIITA STATION BOILER AND NO ABATEMENT DATA
I
N
o. 1
No. 2
No. 3
No. 4
No. 5
No. 6
Boiler
Capacity. MW
Boiler Maker
375
MHIa
375
M I II
500
MIII
700
M I II
700
BHKb
700
BHKb
Operation Start—up
1965
1966
1967
1974
1978
1978
Fuel
Oil
Oil
Oil
Oil .LNG
LNG
LNG
NO, ppm
NO limitation, ppm
Denitrification Plant
Process
120—140
150
120—140
100
SNRC
90—100
105
95—100
20
SCRd
35—50
11
sct
35—50
11
sci
Constructor
—
MIII
—
MIII
Hitachi
Hitachi
Completion
—
1977
—
1980
1978
1978
NO Concentration
I
after treatment,
ppm
—
80—100
—
15—20
7—10
7—10
NIh Emitted, ppm
—
10—15
—
Below S
Below 5
Below 5
a
bMtts )15 Heavy Industries
Babcock—hitachi
Se1ect1ve noncatalytic reduction
Selective catalytic reduction
-------
TABLE 4—4. EMISSION LIMITS AT ThE CIIITA POWER STATION
Item Regulation Compliance Technique
Water
COl), mgfI. 10 • Chemical treatment
TSS . mg/L < 15
Oil, mg/L < 1 e Rainwater to oil separator
pH 5.8 — 8.6
Air
NO As shown in 0 Combustion modifications (controlled mixing
X Iablc 4—1 and OFA)
I SCR with Nih
• SNR with Nil 3
sO o Low sulfur fuel
Ash, mg/MIni <20 0 ESP’s on boilers 1, 2, 3, and 4
Noise, plioiies ( 50 o Walls around transformer
• Fl) fan insulated
• Valves covered
-------
No. 4 boiler so that the total NO 1 emissions would not increase with the
addition of boilers No. 5 and No. 6. The SNR system is described in Section
6.2.4.
Table 4—4 shows the water, NO 1 , SOi, particulate, and noise emission
limits which apply to the Chita Power Station.
4.2.3 SCR Units at the Chita Station
Two full scale S R units for the new No. 5 and No. 6 700—MW LNG—fired
boilers were constructed by Babcock—Hitachi using a pellet catalyst supplied
by Hitachi Ltd. (Table 4—5). Flue gas (1,900,000 Nm 3 /hr per boiler) from the
boiler economizer at 350°C (full load) and 290°C (250 MW load) is injected
with air—diluted NH3 and sent into two parallel reactors. Each reactor mea-
sures 10.5 z 11.5 x 12.0 m and has four 11 cm—deep layers of a pellet catalyst
(Figure 4—7). The flue gas passes through one of the layers in which the SV
is 20,000 hr . Because of good combustion control the inlet NO 1 concen-
tration is only 35—45 ppm. By adding about 0.9 mol NH 3 to 1 mol NO 1 , over
80 percent NO 1 removal is attained while unreacted N113 is maintained below 5
ppm. The plant has operated without problems since its startup in April
1978. Since that time the catalyst has not been changed; a catalyst life of
over 5 years is expected. The installation cost for the plant was 2400 yen/kW
including the initial charge for the catalyst.
TABLE 4-5. DATA SUMMARY ON BABCOCK-HITACHI SCR SYSTEM —
No. 5 and No. 6 BOILER, CIIITA POWER STATION
Reactors per unit 2
Catalyst Type Pellet
Catalyst Size, mm 5 x 5
N113:NO Mole Ratio 0.9
1
NO Removal, design, SO
NO Inlet, ppm 35—45
NO Outlet, ppm 7—9
Catalyst Manufacturer Hitachi Ltd.
275
-------
Figure 4-7 SCR Reactor and Catalyst Bed. (No S and No 8 Boilers. Chits Station) (3)
276
-------
The S R unit for the existing No. 4 boiler at the Chita Station was
supplied by MHI and began continuous operation in March 1980. The design data
for the unit are shown in Table 4—6. The unit’s reactor and catalyst arrange-
ments are shown in Figuxe 4—8. Within the reactor, a coated hexagonal honey-
comb catalyst with a 6.8 mm channel diameter and a 1.2 mm wall thickness, is
stacked in four stages. Each stage consists of two layers of catalyst pack-
ages; each package contains 50 elements.
Since startup, the S R unit has operated without problems and has suc-
ceeded in reducing NO 1 from about 100 ppm to about 18 ppm by using about
0.85 mol NH3 to each mol of NOR. Unreacted N has been maintained at 1—2
ppm. For fuel, the No. 4 boiler uses both low—sulfur oil and LNG, mostly the
latter. There has been no ammonium bisulfate problem with the air preheater
because of the low SO 1 content of the flue gas and the very low level of
leakage NH 3
The cost of this S R installation is 2.5 billion yen (3,430 yen/kW) and
therefore is relatively expensive. This is a result of the required retrofit-
ting of extensive ductwork. The unit also uses a relatively large amount of
catalyst (SV 5,430 hr’ ) to treat flue gas from oil containing SOx and
particulates. with a low level of leakage N113.
4.2.4 SCR Plants at Other Stations
In 1980 and 1981, eight S R plants were completed at three of Chubu Elec-
tric’s stations — t’lishinagoya, Atsumi, and Shinnagoya — as shown in Table 4—2.
Two of the SCR plants at Atsumi are used with new 700—?flV oil—fired
boilers. These plants have square (grid) honeycomb catalysts with a 7 mm
pitch (channel diameter plus wall thickness) and an SV of 5,200 hr 1 . A
low-sulfur oil containing about 0.2 percent sulfur is burned with combustion
modification using a low—NO 1 burner to yield a flue gas containing about 100
ppm SO 1 and 80 ppm NO 1 . The gas is treated by S R to reduce NO 1 below
16 ppm. These plants have not experienced any problems to date.
277
-------
TABLE 4—6. DESIGN DATA FOR THE M III SCR SYSTEM USED WITU THE NO. 4
D0ILER AT TIlE CIIITA POWER STATION (3)
General
G.i Volume, Nm’/hr 1.960,000
NO Removal, percent 80
NO Inlet, ppn 100
NO Outlet, ppn < 20
Reactor
Tenperature. GC 388
NB; NO Mole Ratio 0.9
I
NB; Outlet. pp ( 10
Cataiy5 ;
Space Velocity. he’ 5430
Material Transition Metal Oxide
Shape Hexagonal Honeyconb
Pitch, __ 8
Manufacturer
278
-------
CATALYST ELEMENT
FORCED
DRAFT
FAN
2mm WALL
PITCH 8mm
Figure 4.8 SCR Catalyst and Reactor Arrangement (No 4 Boiler, Chita Station) (3)
AIR HEATER
EACH
50 ELEMENTS
5x5x2
279
-------
The five other S R plants are applied to existing boilers; all use honey—
comb catalysts. Those at the Nishinagoya Station achieve an 80 percent NO 1
removal rate, while those at the Shinnagoya Station are designed for about 30
percent NO 1 removal and use a smaller amount of catalyst placed in the
existing duct between the boiler economizer and the air preheater. All of
these S R plants have operated without problems.
4.3 SCR UNITS AT KUDAMATSU STATION, CIJUGOKU ELEC’IRIC
4.3.1 Introduction
Chugoku Electric’s Kudamatsu Station is located near the Tokuyama Indus-
trial Complex and has three oil—fired boilers as shown in Table 4—7.
TABLE 4—7. BOILERS AT THE KUDAMATSU STATION
Boiler
No.
Capacity
( MW)
Year of
Completion
NO 1
(ppm)
NO
allowed
(ppm)
(Nm /hr)
1
156
1964
140—160
160
70
2
375
1973
130—160
32
34
3
700
1979
110—130
26
•.jQ
TOTAL
1,231
154
The No. 1 boiler was completed in 1964 and the No. 2 boiler in 1973. In
order to obtain a construction permit for the No. 3 boiler, Chugoku Electric
agreed to maintain total NO 1 emissions below 154 Nm 3 /hr for the entire
station. To do this, SCR plants were installed not only for the No. 3 boiler
but also to remove 80 percent of NO 1 from the No. 2 boiler (Table 4—7).
1111 constructed both the boilers and the SCR units at Kudamatsu. A
honeycomb catalyst (square) is used in the SCR units. The SCR unit for the
No. 2 boiler has a single reactor and began operation in April 1979. The No.
280
-------
3 boiler’s S R unit has two reactors in parallel and began operation in
September 1979 when the boiler also began operation.
Chugoku Electric also made agreements with the local government regarding
S0 and particulate emissions from the Kudamatsu Station. Emissions of
SO 1 must be maintained below 373 Nin 3 /hr by using low—sulfur oils
0.8 percent S for the No. 1 boiler and 0.2—0.25 percent S for Nos. 2 and 3.
Particulate emissions from the station must be maintained below 20 mg/Nm 3 and
below 68.6 kg/hr by using combustion control and ESP.
4.3.2 SCR Units
Flowsheets for the No. 2 SQ unit are presented in Figures 4—9 and 4—10;
S R system specifications are presented in Table 4—8. Figure 4—9 also gives
the timetable for the retrofit construotion about one year was required for
the SCR installation. New ducts were connected to the existing economizer and
air preheaters during the annual boiler shutdown period in February and March
1979. The flowsheet of the SCR units for the No. 3 boiler is similar to the
flowsheets in Figures 4—1 and 4—2.
The reactors in the S R units each have two layers of a square ceramic
honeycomb catalyst with horizontal gas flow as shown in Figure 4—11. The
catalyst was produced by Sakai Chemical and Catalyst and Chemicals, Inc.
Pilot plant tests by mi indicate that the catalysts produced by the two corn—
panics meet the specifications equally well.
The 275_3800C flue gas from the boiler economizer is injected with
ammonia and introduced into the reactor. The gas is then sent through an air
preheater and ESP. In the S R plant for the existing No. 2 boiler, induced
fans were installed to compensate for the pressure drop caused by the S R
reactor. The plant for the new No. 3 boiler does not have a fan.
281
-------
TABLE 4-8. SCR PLANT SPECIFICATIONS (KUDAMATSU STATION, CIIUGOKIJ ELECTRIC)
No. 2 Boiler No. 3 Boiler
New or retrofit Retrofit New
Power generation capacity, MW 375 700
Flue gas volume, Nm 3 /hr 1,050,000 1,900,000
Number of SCR reactors 1 2
Type of catalyst Honeycomb Honeycomb
Number of catalyst layers 2 2
1
Space velocity, SV 5,500 5,500
Catalyst voltme , m 3 190 347
Flue gas temperature, O 275—385 275—380
Type of fuel Heavy oil Heavy oil
Sulfur content of oil, ! 0.25 0.25
Inlet NO, ppm (Oz 4! ) 150—160 110—130
N!13/NO mole ratio 0.90 ( 0 • 85 )a 0.90 ( 0 Q 5 )fl
NO removal, 3 ( 5 ] _ 83 )a 80 (81—83)
Leak NH 3 , ppm 10 ( 4 _ 6 )a
Catalyst life, year 1 ( 3 )a 1 ( 3 )a
a a
Pressure drop through reactor, mm IhO 102 (70) 102 (70)
Start—up April 1979 Sept. 1979
arigitres in parentheses are actual ones.
282
-------
1978 1979
VEAPI I I I I I I I I I I I I I I I
12345678 91011121 23456789
MONTH (JAN) (DEC) (JAN)
OPERATION SHUTDOWN OPERATION SHUTDOWN
BOILER
SCR A B.C
FOUNDATION I I
SCR I I I I I I
1978
OPERATION
DUCTS B,C
I- I
REACTOR I
IDF TEST WITH TEST WITH
CRUDEOIL HEAVYOIL
I I III I II
1979
FIgure 4.9 Retrof It Construction of the SCR System for the No. 2 Boiler at Kudamatsu Station
70A2647
RETROFIT EQUIPMENT
TO
(SEPT)
SCA DUCT A
FACILITIES I- I
283
-------
BOILER I
c] I BOILER — — i —
LOAD SIGNAL I TO
I I __ — ____L_J STEAM (10,000m 3 lmIn x
I •
I I
.. HEATER
L — — — — 1070mm H 2 0 x 2300 kw)
[ J - - 4
MIXER VAPORIZER AMMONIA
TANK
FLUE GAS APH AIR PREHEATER
AIR ESP ELECTROSTATIC PRECIPITATOR
— - — - — - AMMONIA
•f$+++ ++tttANALYSIS
- CONTROL
TOBOILER - —
SCR
AMMONIA REACTOR
STEAM
HEATER
(10.000 m 3 lmIn
1070 mm H 2 O
2300 Icw)
--8
STACK
(200m)
Figure 4-10 Flowshoet of I he 5CR System for the No. 2 BoIler at Kudamatsu Station.
70A2848
-------
CATALYST LAYER
FLUE GAS
& OUTL
IUNIT MODULE
Figure 4.11 Fixed Bed Reactor for Oil Fired Applications. (honzontal gas flow)
FLUE GAS
IN LET
HONEYCOMB MESH
285
-------
The correct amonnt of ammonia that should be injection is determined by
measuring the boiler load and the inlet NO 1 concentration. Outlet NO 1 and
ammonia concentrations are also measured to trim the system. NO 1 concentra-
tions are measured by chemilmninescence and the outlet NR concentration after
NO conversion is continuously analyzed by catalytic oxidation (Section 3.7.6).
Since this method of NRs analysis is not highly reliable, manual analysis is
periodically used for calibration.
4.3.3 SCR System Performance
IHI guarantees that the NO 1 removal efficiency of its S R systems will
be over 80% if an NEs/NOx mole ratio of 0.9 is used for more than a year.
By using a mole ratio of 0.9, an 85—87% NO 1 removal efficiency was attained
initially at Kudamatsu. For normal operation, a mole ratio of 0.85 has been
used for both systems. Currently the NO 1 removal efficiency of the system
is 81—82%, while in 1980 it was 82—83%. Unreacted N B 3 has been maintained at
about 5 ppm. Based on these results the catalyst life is expected to be more
than 4 years.
At a boiler load of one—third, the economizer outlet gas temperature
drops to 275°C. When the boiler load goes below one—third, ammonia injection
is discontinued. At this low boiler load level the NO 1 concentration is
already low enough to meet the established emission limit.
The pressure drop of the flue gas is 50—60 mm ilsO through the catalyst
beds and nearly 100 mm HaO through the entire SGR system . including the
reactor and the connecting ducts. No increase in the pressure drop has been
noticed during the more than 3—year operation period. Soot blowing systems
have been installed for the reactors but have not been used because the cata-
lyst beds have remained clean.
The No. 1 and No. 2 boilers at the Kudamatsu Station are operated by 10
persons per shift while the No. 3 boiler is operated by 4 persons per shift.
No additional personnel are needed to operate the SCR system. If necessary,
286
-------
the catalyst will be replaced during the annual maintenance shutdown of the
boiler.
Ammonia consumption is 80—150 t/month for the two S R plants. Two 65—ton
ammonia tanks have been installed at the station. Prior to the application of
ScR, ammonia was injected into the flue gas as it left the air preheater.
This was done in order to neutralize the sulfuric acid mist and to prevent the
corrosion of the ESP. It produced a flue gas containing nearly 10 ppm NH 3 .
The operation of the S R units at Kudamatsu has been trouble—free except
for the ammonium bisulfate problem in the air preheater which is described in
the following section.
4.3.4 Ammonium Bisulfate Deposition in the Air Preheater
The flue gases from the No. 2 and No. 3 boilers contain about 1 ppm SO
with about 100 ppm SO,. After the gas passes through the S R reactor, its SOi
concentration increases to 5—7 ppm. The oxidation ratio of S0 to SO is high
(4—6%), because the catalyst was designed before low—oxidation catalysts were
developed. In the air preheater, the flue gas containing about 5 ppm of NH,
is cooled to 145°C and 137°C in the No. 2 and No. 3 systems, respectively.
A substantial amount of ammonium bisulfate deposits in the air preheaters
of the SCR unit. In the No. 2 system, soot blowing is applied from the cold
end of the preheater using 14 kg/cm’ air for 2—hour periods, 3 times a day.
The pressure drop of the gas through the preheater is normally about 150 mm
llaO. After the system has been operating for three or four months, the pres-
sure drop begins to increase rapidly. When it reaches 300 mm H,O after about
four months operation, the boiler is shut down for cleaning of the preheater.
One washing requires 2,500 tons of water. The wash water containing ammonium
bisulfate is stored, neutralized at a rate of 50 t/hr and stored again in
preparation for reuse. One washing takes about 20 hours including the drying
of the preheater. Washing is usually done on weekends when the power demand
is low.
287
-------
The preheaters for the No. 3 boiler have soot blowing systems at both
cold and hot ends. Three times a day air blowing is applied from the cold end
of the preheater; steam soot blowing from the hot end is used once a day.
Even with the increased blowing, water washing is necessary only once every
two or three months. Five thousand tons of water are normally used to waA
both of the preheaters. Each preheater is washed while the other preheater is
in use and the boiler is operating at half load.
To solve the bisulfate deposition problem in the No. 3 system, a number
of measures were taken during the annual maintenance shutdown of the boiler in
late 1980. First, the soot blowing system at the cold end of the preheater
was modified to provide a stronger blowing action. Second, a small amount of
catalyst was added to the reactor in order to reduce the unreacted NU3 from 6—
8 ppm to 4—5 ppm. There was room in the catalyst container to hold the addi-
tional catalyst. Third. since dust tends to increase bisulfate deposition,
the dust content of the flue gas was reduced by about one—half. The amount of
dust was reduced by reducing the ratio of the flue gas recirculation and the
staged combustion. This also had the effect of increasing the N0 concen-
tration of the flue gas from 105—110 ppm to 115—120 ppm.
So far, these modifications have proven successful with no increase in
the pressure drop. The reinforced soot blowing at the cold end of the pre-
heater is considered to be very effective; soot blowing from the hot end has
been abandoned. There is no plan to add catalyst in the No. 2 system, al-
though its soot blowing system may be reinforced in the future.
Recently Chugoku Electric, jointly with ml, studied a new technology to
remove the deposits——thermal cleaning. With this type of cleaning, the hot
gas is passed through the preheater as usual while air is stopped to allow
heating of the preheater elements to about 3000C. This causes the partial
decomposition of the deposits. Intense soot blowing is applied during the
heating process to dislodge the deposits. Test results have thus far not been
disclosed.
288
-------
4.3.5 Economics
The total investment cost of the Kudamatsu SCR plant was 2.2 billion yen
for the No. 2 unit and 2.5 billion yen for the No. 3 unit. These costs
include civil engineering, labor, and performance testing. The No. 2 unit was
relatively expensive bcause of the complex ductwork needed for the retrofit-
ting process as well as the installation of the induced fans.
A breakdown of capital costs for the No. 3 S R system is given in Table
4—9. The entire No. 3 boiler and generator system cost 55 billion yen, includ-
ing 2.5 billion yen for the SCR plant and 1.2 billion yen for the ESP system.
In this case the S R system accounts for 4.5% of the total cost of the boiler—
generator system.
The annual cost for the No. 3 S R plant is given in Table 4—10. The cost
assumes 70% boiler utilization or actual power generation of 4,292,400 MWhr/
yr. The annualized cost for S R adds up to 0.28 yen/kWhr (assuming a 3—year
catalyst life) which is equivalent to about 1.3% of the power generation cost.
This cost would increase to 0.32 yen/kWhr if the catalyst life is 2 years, and
would decrease to 0.26 yen/k hr with a 4—year catalyst life. Although a small
amount of catalyst was added to the No. 3 reactor, as mentioned above, it is
not certain whether the addition was really needed to reduce bisulfate deposi-
tion. A 4—year catalyst life may be attained without the addition of the
catalyst.
The power consumption of the S R system represents 0.15% of the total
power generated by the boiler. This is less than one—tenth of the power con-
sumption of conventional FGD systems. The investment cost for the No. 3 plant
is 3,570 yen/kW, approximately one—fifth that of a conventional FGD system.
4.3.6 Evaluation
The smooth operation of the reactor and the long life of the catalyst at
Kudamatsu have demonstrated the reliability and utility of SCR for flue gas
from low—sulfur oil. The ammonium bisulfate deposition problem is largely due
289
-------
TABLE 4—9. CAPITAL COST BREAKDOWN OF SCR SYSTEM ON NO. 3 BOILER (700 MW)
CHIJGOKU ELECTRIC, KIJDAMATSII POWER STATION ( 1 = 25O) (3)
Capital Cost
Percent
Component l0 10’
Catalyst 1.1 4.4
44
Reactor 0.7 2.8
28
NIh System 0.24 0.96
10
Testing and Start—Up 0.16 0.64
6
Other 0.30 1.2
12
TOTAL 2.5 10.0
100
TABLE 4—10. OPERATING COST BREAKDOWN OF SCR SYSTEM ON
NO. 3 BOILER
(700 MW)
CBUGOKtJ ELECTRIC, KIJDALIATSU POWER STATION
(±1 Y250)
Cost/Year
Percent
Component 106 10’
Interest 250 1.00
21
Depreciation 321 1.28
27
Tax 35 0.14
3
Catalyst 367 1.47
31
Power 83 0.33
7
NB3 76 0.30
6
Others 62 0,25
5
TOTAL 1194 4.77
100
Basis: 7 years depreciation
10 interest
3 years catalyst life
Note: Japanese utility power company economics include the catalyst cost both
in the investment cost (which is depreciated) and in the operating
cost.
290
-------
to the high oxidation ratio of S02 to SOi which occurs with certain catalysts.
The new low—oxidation catalysts greatly reduce this problem.
The applicability of thermal cleaning may be related to the intensity of
the soot blowing and the specific composition of the deposits. Usually, bi—
sulfate reacts with preheater construction material to form various compounds.
Combinations of some of the compounds may be difficult to remove with thermal
cleaning.
4.4 ScR UNIT AT THE SHIMONOSEKI PLANT, CHUGOKtJ ELECTRIC
4.4.1 Introduction
The Shimonoseki Station of the Chugoku Electric Power Company has two
boilers——a 175 MW coal—fired boiler (No. 1) and a 400 MW oil—fired boiler
(No. 2). The air and water pollution regulations which apply to the station
are listed in Table 4—11.
TABLE 4—11. REGULATIONS FOR SHIMONOSEKI STATION
Pollutant
Regulated Level
-
Air Pollution Control
2.7 (Ground
ppm)
level
concentration
0.0047
k Value
SOT, Mm 3 /hr
Below 412
Particulates, kg/hr
Below 130
No. 1 boiler, mg/Nm 3
Below 200
No. 2 boiler, mg/Nm 3
Below 40
NOT, Nm 3 /hr
Below 330
No. 1 boiler, ppm
Below 350
No. 2 boiler, ppm
Below 170
Suspended particulates, mg/&
Below 0.2
7ater Pollution Control
5.8 — 8.6
pH
Suspended solids, kg/day
Below 12
mg/liter
Below 15
Normal hexane soluble material
kg/day
Below 0.8
mg/liter
Below 1
Chemical oxygen demand, kg/day
Below 12
mg/liter
Below 15
291
-------
The No. 1 boiler was completed in 1967 and burned coal and oil at a 25/75
ratio before .a full scale MIII limestone—gypsum FGD plant was completed in July
1979. After the FGD plant began operation, a 50/50 coal and oil ratio was
used. Although the NO 1 regulation may be met by combustion modification,
even with coal as the only fuel, it is likely that more stringent NO 1 reduc-
tion may be required in future. For this reason, Chugoku Electric decided to
install a full—scale S R unit to allow for the burning of coal as the sole
fuel. This unit, completed in March 1980, is the first full—scale SCR plant
for a coal—fired boiler in the world and is considered to be a demonstration
plant.
Figure 4—12 shows the combined ScR/FGD system for the No. 1 boiler. In
the system, the flue gas is first subjected to SCR at 350—400°C, then passed
through two trains of air preheaters and dust collectors (multi—cyclone and
ESP) . and finally subjected to FGD after passing through a heat exchanger.
The No. 2 boiler is a relatively new one and uses a high sulfur oil. The
boiler uses combustion modification for NO 1 abatement and an MIII limestone—
gypsum FGD unit to reduce SOs.
4.4.2 SCR System
The design data for the Shimonoseki StR system are given in Table 4—12.
The No. 1 boiler is used for the station’s base load and the 370°C gas
temperature at its economizer outlet is suitable for SI2R. The load is some-
times lowered to 25% which causes a drop in the gas temperature. Since ao—
nium bisulfate may deposit on the catalyst at the low temperature and thereby
lower catalytic activity, an economizer by—pass system was installed (Figure 4—
12). The system mixes a portion of hot gas with the economizer outlet gas at
the low load in order to maintain a gas temperature of about 350°C.
292
-------
SELECTIVE CATALYTIC REDUCTION OF NO
MULTICYCLONE
AIR PREHEATER
FORCED DRAFT FAN
ECONOMIZER
Figure 4-12
Flue Gas Treatment System for No. 1175 MW Coal-Fired Boiler.
(Shimonoseki Power Station, Chugoku Electric)
HEAT EXCHANGER
(GAS-GAS HEATER)
BOILER
t ’ )
SCR
MC:
APH
FDF:
ECO
70A2650
-------
TABLE 4-12. DESIGN DATA FOR THE M I II
POWER STATION
SCR SYSTEM INSTALLED AT THE SHIMONOSEKI
It em
Design Value
Flue gas flow rate
Inlet temperature
Inlet NO 1
NO 1 removal
NII /NO 1 mole ratio
NR3 emissions from reactor
Inlet SOz
Inlet S03
Reactor dimensions
Space velocity
Catalyst
550,000 Nm 3 /hr
3700 C
500 ppm
50% (Reactor sized for 80%)
<0.51
<10 ppm
1600 ppm
32 ppm
11.8m W x 8.5m D x 13.3m H
3000 hr
Honeycomb, solid square type.
10 mm pitch (channel diameter plus
wall thickness).
An S R reactor was installed adjacent to the boiler so that the treated
gas could be sent to the existing air preheater (Figure 4—13). The reactor
contains five horizontal layers of honeycomb catalyst (Figure 4—14) through
which the gas passes downward. On top of the first honeycomb layer is a dummy
layer made of inactive hard ceramic material with the same cross section as
the honeyc mb catalyst. This layer maintains a uniform parallel gas flow and
prevents the erosion of the catalyst by fly ash. There is an open space at
the bottom of the reactor where an additional catalyst layer could be placed
to attain a higher NO 1 removal efficiency.
The planning and design of the S R system at Shunonoseki began in July
1979; construction followed in October 1979. The boiler was modified and the
reactor connected during the boiler’s annual maintenance shutdown period be-
tween February and March 31, 1980. Since startup in April 1980, the boiler,
S R system, and FGD system all have been operated without problems. Normally,
a 0.51—0.56 mol NH3 to 1 mol NO 1 ratio has been used to remove 5O—55 ’ of
NO 1 . This meets the NO 1 regulation and keeps unreacted NB3 below 2 ppm.
The system can be operated without an attendant.
294
-------
Figure 4 13 isometrIc Flow Diagram of Boiler and SCR Reactor (Shimonoseka Power Station) (3)
N J
U I
-------
1 ___ _____________
1>JE J
/ ____
II > II 1 <
ITTT, _ _____ I ii i i
‘ 2OO
60)1 • _ 37OO —
CAIALYSI BASKEt
CON I AININO 144
IAIAI YSI ElEMENtS
28 5 .. L 75 JL _ _ JL 28?SJ
25 25 25
§
F - i
1
L
E T
“
-
I
N
-
E JF <
\Z OE YIEW
II
TOP VIEW
. —OtJMMY LAYER
— OPEN SPACE
OIn1enssons In nAil
FIgure 4 II SCA Reactor and Catalyst Basket Cl Stilmonoseki Power StatIon
-------
MEl guarantees a catalyst life of one year for its SQ system. At
Shinionoseki, the catalyst was used for more than two years without any degra-
dation, except for some very minor erosion observed at the inlet edge of the
first reactor layer. Three years of catalyst life is expected and the cata-
lyst may be replaced during the annual boiler maintenance period. The reactor
has a soot blowing system but it has not been used thus far because the cata-
lyst has remained clean.
In this system the air preheater also has been kept clean. Originally,
the air preheater contained dual—undulated elements for high— and medium—
temperature zones and notched flat elements for the low—temperature zone. It
also was equipped with a soot blowing system at the cold end which was oper-
ated for 2—hour periods, 4 times daily. When the SCR unit was installed, a
notched flat element was substituted for the dual—undulated element in the
medium temperature zone: the former can be more easily cleaned by soot blow-
ing. An additional soot blowing system was installed at the hot end of the
preheater. Since the startup of the SCR system, soot blowing from both ends
has been used for 2—hour periods, 4 times daily. Thus the air preheater has
been kept clean and has not caused an increase in the pressure drop.
The total investment cost of the Shimonoseki SCR system was about 2 bil-
lion yen in 1980 including 1.7 billion yen paid to the builder (3).
In the future, when 80% NO 1 removal is required, a small amount of
catalyst will be added so that 80% removal is attained.
4.4.3 Problems Related to tinreacted N03
The amount of ammonia injected in the SCR reactor at Shimonoseki is de-
termined by the flue gas volume and inlet NO 1 concentrations in a manner
similar to that used at other SCR plants. The concentration of the unreacted
N E 3 can be measured but has not been used to determine ammonia injection be-
cause the method of measurement is less reliable than those used for other
parameters. Recently the Shimonoseki station began using a new method of
analyzing unreacted Nib developed by Shimazu (Figure 4—15). A sample portion
297
-------
SCR — GAS TO AIR PREHEATER
REACTOR
________ .........................._...fJ NO ANALYZER
____ ‘ [ NO ANALYZER
SCR
CATALYST
Figure 4.15 PrInciple of linreacted NH 3 Analysis.
298
-------
of the reactor outlet gas is sent through the reactor (with the S R catalyst)
where all of the unreacted NH3 is converted to N and HaO after reacting with
the NO in the gas. The amount of N113 in the gas is calculated by determining
the difference between the NO concentrations in the gas which has been sent
through the reactor and that which has not been exposed to the catalyst. This
method seems suitable for analyzing a low level of unreacted NH3
The product fly ash at Shimoseki contains very little ammonia and can
readily be used for cement, concrete, or landfill material. It is assumed
that the ammonia contamination of the fly ash will not be extensive as long as
unreacted NU3 is kept below 5 ppm.
The unreacted NH ] is eventually caught by the FGD system downstream of
the air preheater. The NRa has not caused any problems in the FGD system or
with wastewater treatment because of its low concentration level.
The Shimonoseki Station’s No. 2 boiler uses a high sulfur oil. For this
reason, ammonia is injected into the flue gas as it leaves the air preheater
in order to prevent ESP corrosion and also to increase particulate removal
efficiency; Thus, the flue gas which is sent into the FGD system contains
about 5 ppm NRa, which is more than that found in the No. 1 boiler flue gas
after SCR. The FGD system for the No. 2 boiler, as well as the wastewater
treatment system, has had no problem with the ammonia.
4.4.4 Evaluation
The smooth operation of the S1.R system at Shimonoseki has demonstrated
the commercial applicability of the high dust system for coal—fired boiler
flue gas. The use of a dummy layer in the reactor successfully prevents cat-
alyst erosion and is responsible for a catalyst life expectancy of more than 3
years.
The gas temperature at the S R reactor has been kept above about 3300C.
The critical temperature at which ammonium bisulfate deposits on the catalyst
differs with the inlet SO concentration, the catalyst—induced oxidation
299
-------
ratio of SOa to SOa and other factors. It has been found that with flue gas
from heavy oil or crude oil, the oxidation ratio increases with time due to
the deposition on the catalyst of vanadium derived from oil. This increase in
oxidation may not occur with coal, particularly with the high dust system.
Those problems warrant additional future study at Shimonoseki.
Since the catalyst has proved to be free from dust plugging and erosion
problems, even for flue gas with a full dust load, MHI has decided to use a
more efficient catalyst with a 7—mm pitch for future high—dust system S R
plants. This efficient catalyst has already been used at S R plants for three
coal—fired boilers (75 MW, 75 MW, and 125 M V) at chugoku Electric’s Shin—IJbe
Station, which began operation in early 1982. By using an SV of 4,000
hr’ and a NU3/NOX mole ratio of 0.66, 65% of NO has been removed
with 2 ppm unreacted N113. The plants have operated successfully and the Un—
reacted NB3 has not caused any effect on the air preheater, FGD system, or fly
ash composition.
4.5 SCR SYSTEM AT TOMATO-ATSUMA STATION, UOKKAIDO ELECTRIC
4.5.1 Introduction
In October 1980, Hokkaido Electric completed a new 350 MW coal—fired
boiler at Tomato—Atsuma. The station is located in a new industrial district
developed by the Hokkaido government. The flue gas treatment system for the
Tomato—Atsuma boiler is shown in Figure 4—16. In accordance with local regu-
lations, NOx emitted from the boiler is less than 184 Nm 3 /hr (158 ppm)
SO is kept below 152 Nm 3 /hr (130 ppm). The boiler burns domestic low—
sulfur coal containing 0.3% sulfur and 1.2% nitrogen. A hot—side ESP has been
installed to maintain a high dust removal efficiency. This unit is the first
hot—side ESP installed in Japan for a utility boiler. At the plant, combus-
tion modification including staged combustion, flue gas recirculation, and a
low-NO burner, is used to reduce NO to 200 ppm. To meet the NO regulatory
requirement, one—fourth of the flue gas is treated by S R to achieve over 80%
NO removal. A limestone—gypsum FGD system has been installed which removes
over 90% of the S0 from half of the flue gas.
300
-------
HOT ESP
AIR
LIOUID AMMONIA
C
AMMONIA
VAPORIZER
DILUTION
CHAMBER
DILUTION
AIR
DE-NO REACTOR
Figure 4-16 Flowsheet for the De NO System al the Tomato Alsuma Station
(Hokkaldo Electric Power Co)
-------
The total investment cost for the power station at Tomato—Atsuma was 70
billion yen including 11—13 billion yen for pollution control facilities. The
boiler, SQL unit, and FOD unit all were constructed by Babcock Hitachi Ltd.
Since its startup in October 1980, the entire system has operated smoothly,
except for a few minor problems with the FOD wastewater treatment system.
4.5.2 SCR System
The design conditions and equipment specifications for the Tomato Atsuma
S R unit are listed in Tables 4—13 and 4—14, respectively. A plate catalyst
(1 mm thick with 10 mm pitch) produced by Hitachi Ltd. is used in the u.nit
(Figure 4—17).
TABLE 4—13. DESIGN CONDITIONS FOR SCR UNIT, TOMATO—ATSUMA STATION, HOKKAIDO
ELEC’IlLIC POWER COMPANY
Items
Design Base
Remarks
Gas flow
280,000 N&/hr
350 MW x 1/4
Inlet gas temperature
356°C
Inlet NOx concentration
Max. 200 ppm
Outlet NO 1 - concentration
(40 ppm
NO 1 removal efficiency
>80%
(Guaranteed)
Inlet dust load
45 mg/Nm 3
Slip ammonia concentration
<10 ppm
Inlet gas humidity
8.16 Vol. %
The 360°C flue gas leaving the boiler economizer is sent through the hot—
side ES? where its dust content is lowered from about 23 g/m to 45 mg/Nm 3 , a
removal efficiency of 99.8%. Then the gas is sent downward through an SCR
reactor. Gas velocity in the reactor is about 4 rn/sec. This rate is suf-
ficient to prevent both erosion and dust plugging of the catalyst. Eighty to
85 of the NOx is removed in the reactor by using a ratio of 0.85 to 0.9 mol
N113 to 1 mol NO (Table 4—15).
302
-------
TABLE 4—14. PRIMARY PLANT EQUIPMENT SPECIFICATIONS
Equipment Item Description
DeNOx Reactor Type Vertical Type Parallel Flow
Number 1
Size 6.6m W x 17m H x 8.8m L
Material Carbon Steel
Catalyst Type Plate Type (1 mm x 10 mm Pitch)
Number Two layers
Material Titanium base
DeNOx Fan Type Double Suction Turbo Fan
Number 1
Flow 11,900 m /min at 356°C
Pressure —340 mmfliO (Inlet)
—170 mmH2O (Outlet)
Motor 480 kW, 50 Hz, 10 P
Liquid Ammonia Type florizontal Cylinder with Pillow Support
Storage Tank Number 2
Capacity 10 t
Material SB46N2SR (carbon steel)
Ammonia Vaporizer Type Vertical Spiral Tube
Number 1
Capacity Max. 113 Kg/h
Material SS41, STPG38
Ammonia Accumulator Type Vertical Cylinder
Number 1
Capacity 0.5 m 3
Material SM41B
Ammonia Gas Type Horizontal Cylinder
Dilution Chamber Number 1
Material STPG3 S
303
-------
SHAPE OF CATALYST ELEMENT
GAS FLOW
0
CATALYST UNIT
Figure 4 17 Catalyst Used in the Tomato-Atsuma SCR Unit
-------
TABLE 4-15. DATA. TOMATO-ATSUMA SCR UNIT, HOIKAIDO ELECTRIC POWER COMPANY
Items
Design
Actual
Gas flow, Nm 3 /hr
280,000
287,000
Inlet gas temperature, OC
356
332
Inlet NO 1 concentration, ppm
200
181
(Oi 6%)
Outlet NO 1 concentration, ppm
40
34
(02 6%)
NO 1 removal efficiency, %
80
81.2
(NHs)I(N0 1 ) Mole Ratio
0.85
0.85
Slip ammonia concentration, ppm
10
1.9
Inlet gas humidity, vol %
8.16
10.0
The amount of inlet ammonia injected is determined by the inlet flue gas
volume and NO 1 concentration. Outlet NEs concentration is sometimes
measured manually; usually, the outlet NHi is kept below 5 ppm.
In May 1981, the entire system, including the boiler, was shut down for a
4—day scheduled inspection period. The SCR catalyst was found to be clean
although a soot blowing system had not been installed. In addition, the air
preheater had almost no deposits in it and required no cleaning. During the
nearly 2—year operation period, the system did not experience any increase in
the pressure drop of the preheater nor any decrease in NO 1 removal effi-
ciency. Although catalyst life is guaranteed for 1 year. it is expected that
the catalyst will be used for more than 2 years to maintain over 80% NO 1 re—
moval efficiency with less than 10 ppm leakage NEs.
The cleanliness of the air preheater is due to the following factors:
1. Low leakage NE in the gas (below 5 ppm) which is further
diluted when the gas is mixed with untreated flue gas before
entering the preheater.
2. Low SOx concentration in the gas, the low oxidation ratio
of SOs to S03 (about 0.2%) caused by the catalyst, and the
dilution by untreated flue gas.
305
-------
3. Low dust content of the gas.
4. The use of soot blowing.
4.5.3 Future Construction
In 1985, Hokkaido Electric will complete the construction of a 600—MW
coal—fired boiler at the Tomato—Atsuma station. This No. 2 boiler will be
constructed by ml; Hitachi Ltd will supply the turbines. Imported coal,
containing 1.2% S and 1.8% N, will be used with extensive combustion modifica-
tion to maintain NO below 170 ppm. This combination will enable the boiler
to comply with the regulations without requiring SCR. The flue gas leaving
the air preheater will be treated by a cold ESP and an MIII limestone—gypsum
FQD system. Although the No. 2 boiler will not have an S R reactor, a space
between the boiler economizer and air preheater will be reserved for future
installation of a high—dust SCR system.
4.5.4 Evaluation
The SCR unit at Tomato—Atsuma is the first commercial low—dust SCR unit
to use a hot ESP for flue gas from coal. The smooth operation of the plant
demonstrates that the low—dust system, as well as the high—dust system, is
feasible. The Hitachi plate catalyst is very efficient and removes over 80%
of NO with only a small amount of unreacted NBi.
Several years ago when Hokkaido Electric decided to use the low—dust
system, the high—dust system had not yet proved to be viable; now Hokkaido
Electric is interested in using the high—dust system. Although the company
currently has no plans to install an S R system for the No. 2 boiler, a high—
dust system will be installed if SCR proves to be necessary.
306
-------
4.6 KANSAI ELECrRIC’S S R PLANTS
4.6.1 Large Scale Testing of SCR
Kansai Electric, the second largest utility company in Japan, first
installed an S R test plant in 1977. The plant is capable of treating one—
fourth of the flue gas from an existing 450 MW oil—fired boiler at the Kainan
Station. Initially, a ring tube catalyst was used in a fixed bed. The flue
gas contains about 100 ppm each of SO and NO 1 with about 20 mg/Nm 3 of
particulates. The 330—380°C flue gas leaving the boiler economizer enters the
S R reactor where over 80% of NO 1 is removed and unreacted NBi is maintained
at around 10 ppm. At first, the reactor had dust plugging problems and as a
result, the ring tube catalyst was replaced by a Hitachi plate catalyst. With
the new catalyst, the S R system has experienced trouble—free operation.
In 1977, Kansai Electric also installed a Thermal DeNO 1 SNR system for
an existing 156 MW boiler at its Himeji Station. In 1978, a small amount of a
honeycomb SCR catalyst was installed in the duct between the boiler economizer
and the air preheater in order to increase NO 1 removal efficiency and reduce
unreacted NH3. The SNR unit was removed in 1979. Since then, about 30 3 of
the NO 1 has been removed by the catalyst and a small amount of ammonia has
been injected between the boiler economizer and the catalyst bed.
Both of the plate catalysts at Kainan and the honeycomb catalyst at
Himeji have been used with little degradation for about 4 years. Based on
these results, Kansai Electric has installed many commercial S R plants which
use honeycomb and plate catalysts (Table 4—1).
4.6.2 Commercial SCR Plants
As shown in Table 4—1, Kansai Electric has 25 commercial SGR plants for
existing oil— and gas—fired boilers and plans to install 6 plants for new oil—
fired boilers. Twenty—one plants which use honeycomb catalysts have been or
will be constructed by MEl and ml; Babcock Hitachi has or will construct
307
-------
another 10 plants which use plate catalysts. All of the S R reactors are
placed between the boiler economizer and air preheater.
As shown in Figure 4—18, the SCR reactors for the Sakaiko Station’s
No. 2, 4, 6 and 7 boilers (250 MW each) are each placed in a small space at
the economizer outlet. The reactors use MHI honeycomb catalysts with a space
velocity of 5,070 hr’ so that flue gases from both oil and gas burning
can be treated.
The S R system performance for the No. 6 boiler which fires LNG is shown
in Figure 4—19. The flue gas contains about 110 ppm NO 1 . By using 0.8 mol
NBa per mol of NO 1 , over 75 percent of the NO 1 is removed with less than 5
ppm of unreacted NHa. During 13,000 hours of operation, the S R system showed
only a slight decrease in NO 1 removal efficiency——from 77—78% at the begin-
ning to 76—77 at the end.
Figure 4—20 shows the retrofitting of the S R reactor for the existing
No. 2 boiler (156 MW) at ansai Electric’s Osaka Station. The S R reactor is
positioned just above the vertical flow preheater so that the gas is sent
downward through the reactor. S R reactors also have been installed for the
Nos. 3 and 4 boilers. These boilers burn 1% sulfur oil and produce flue gas
containing about 500 ppm SOa with a small amount of S03. To meet regulatory
requirements, a low oxidation honeycomb catalyst is used to remove 75% of
NO 1 and to maintain anreacted NBa below 5 ppm.
As of Iuly 1982, the SCR systems for boiler Nos. 2 and 4 have operated
without problems for about 9,000 hours. There has been no ammonium bisulfate
deposition in the air preheaters, wh .ch may be due to the low levels of un—
reacted N113 and S03.
Osaka Station’s No. 1 oil—fired boiler (156 MW) has a simple S R system
which uses a small amount of catalyst in the duct between the boiler econo-
mizer and air preheater. This system removes about 30% of NO 1 .
308
-------
Figure 4 16 SCR Reactor Retrofitted to Boliei Nos 2, 4, 6, and 7
(Sakaiko Station, Kansal Electric Power Company)
L )
0
0
FAN
-------
—J’,
>>.
WUJ
Ou-
z
400
350
300
U i
L U
—I--
0
X I — .
0<
Zcr
c .,w
ZO
1.0
0.5
.
-
0
90•
I
I
I
I
I
80
.
— -
-
70
.
60
.
50
‘
i
i
I
0 1000 2000 3000 4000 5000 6000
OPERATION PERIOD (hr)
(Figures plot data once a day at full load)
Figure 4.19 Performance of the SCA Unit for the 250 MW, LNG•Fired No. 8 Boiler.
(Sakalko Station)
70A2658
310
-------
STACK
-REACTOR
L )
FOD
FGD FAN
Figure 4 20 SCR Reactor Retrofitted to Boiler Nos 2. 3, and 4
(Osaka Station, Kansal Electric Power Company)
-------
Figure 4—21 shows the St R reactor which was retrofitted to the existing
No. 6 boiler (600 M’W) at Kansai’s flimeji No. 2 Station. The boiler burns
either oil or gas. 1111 constructed the S R plant which uses a honeycomb cat-
alyst; a large additional amount of ductwork was required to connect a large
reactor between the boiler economizer and air preheater. The boiler has two
trains of air preheater—ESP systems and therefore required an S R reactor for
each of the trains. The boiler burns LNG currently and the S R systems have
been operated without problems.
4.6.3 Evaluation
Kansai Electric’s S R units have been operated without problems. The
lack of a bisulfate deposition problem in the unit’s air preheaters may be due
to tne use of a catalyst with a low SOa oxidation capability and to a low
level of unreacted NH3
4.7 SCR TESTS AT TEE NAXOSO STATION (5)
4.7.1 Test Plant
Tokyo Electric Power Company, Tohoku Electric Power Company, Joban Joint
Power Company, and MHI conducted extensive testing of SCE with flue gas (4,000
Nm 3 /hr) from the No. 7 boiler at Joban’s Nakoso Station. The tests were run
from April 1979 to November 1980. The No. 7 boiler burned 40% coal and 60%
oil in the daytime and 60—70% coal/30—40% oil during the reduced load period
at night.
The ±lowsheet of the test plant Is shown in Figure 4—22. The plant con-
sisted of both high—dust and low—dust S R systems in combination with wet
limestone—gypsum FGD system. The high—dust system treated flue gas with a
full dust load of 6—10 g/Nm 3 . It used 2 downflow SCE reactors with honeycomb
or plate catalysts. In the high—dust system, a ceramic dummy spacer was
placed above the catalyst layer to protect the catalyst from fly ash erosion.
In the low—dust system, the flue gas first was sent through a hot ESP where
the dust content was reduced to 0.1—0.3 gram/Nm 3 . From there, the gas went
312
-------
STACK
i-
- REACTOR
FAN
Figure 4.21 SCR Reactor Retrofitted to Boiler No 6 (Himell Station,
Kansai Electric Power Company)
-------
FHOM CCC
INLET
NO 1 60 1 1 1 K
FLUE GAS
111011 OUST DE NO .
4S® NW T RI
TONO bOILER
ER OUTLEI
F -’
.p-.
LEAK
TYPE REGENERATIVE 4VERTICAL(
TYPE VERTICAL FLOW TYPE VERTICAL PLOW
CATALYST SOUARE HONEYCOMB CATALYST PLAIE
REPLACED BY HONEYCOMB
IN MARCH BS(
LOW DUST (IF NO.
_ _ _ _ _ _ _ =
lOW Nni T Tl
OUTLET
TYPE SPRAY
CATALYST PLATE
TYPE GRID
TYPE VERTICAL FLOW
CATALYST SOUARE HONEYCOMB
Flguie 4 22 Flowsheet of the Nakoso SCR Test Plant
OE SO
LIMESTONE GYPSUM PROCESS
TYPE REGENERATIVE
(VERI CAL l
-------
into 3 SQ reactors in parallel. Two of the reactors used either a honeycomb
or plate catalyst with horizontal gas flow while the third used a honycoinb
catalyst with gas downflow. A vertical flow air preheater was installed down-
stream of each of the reactors.
4.7.2 Test Results
Tests were conducted for a total of 10,600 hours, excluding monthly
shutdown periods for inspection. These shutdown periods typically lasted
several days.
Figure 4—23 summarizes 1979 test results for the high—dust system. An
NO 1 removal efficiency of 80% was attained throughout the test period. The
catalysts were kept clean and showed no degradation; the dummy layer success-
fully prevented catalyst erosion.
During the test period the pressure drop of the gas through the catalyst
bed remained at about 25 and 15 mm RiO for the main and sub reactors, respec-
tively, without soot blowing.
The air preheater contained dual—undulated elements for high— and inter—
mediate—temperature zones and a notched flat element for the low—temperature
zone. Soot blowing was applied once a day to keep the preheater clean. There
was no increase in the pressure drop through the preheater during the entire
10,600 hour test period. Unreacted N113 was maintained at 1—3 ppm. Results of
the first 5,000 hours of testing are given in Figure 4—24.
Figures 4—25 and 4—26 present results of tests on the low—dust system.
Without soot blowing, fine particles of fly ash deposited horizontally on the
inlet edge of the honeycomb catalyst. This caused an increase in the pressure
drop and a decrease in the NO 1 removal efficiency. The ash deposits were
removed by soot blowing. Deposition can be reduced by using a honeycomb with
a larger channel diameter or with a higher gas velocity, but these measures
will not prevent the large increase in the pressure drop which occurs when
315
-------
SV CHANGED
40 -______________________________________________ ______ _______ 3823Hr 4614Hr 4986Hr
20 lll OHr 1733Hr 2885Hr 2995Hr 3286Hr -
- E 20 __________ _________ __________
-J _____________
i— E
I- .
0’ <
-------
—.1
w
0
U)
U)
0
0
w —
a:
(1)
U)
w
a:
a-
SOOT BLOWER
TYPE SINGLE NOZZLE
MEDIUM 14 ATG STEAM
200
r NSPECTI 0NI INSPECTIONI
0
ONCE A DAY FROM DOWN STREAM
SOOTBLOWING i El ONCEADAY IONCEADAYI / 10 I ONCEADAYFROMDOWNSTREAM
FROM DOWN STREAM FROM BOTH SIDE L ONCE A DAY FROM BOTH SIDE
DUST CONTENT
6 lOg /Nm 3
3823
1110
I -J
I I I I
4500
TEST PERIOD (tir)
5000
Figure 4-24 Operating Experience of AIr Preheater vs Soot Blowing (high dust content)
-------
0
w
z
0
z
- 0
(1 )1
>-
..jE
x< E
‘-
-------
w
U)
U)
4
O
OE
w
U,
U)
Lu
SOOT BLOWING INONEI 3 TIMES A DAY
FROM DOWN STREAM
3 TIMES A DAY FROM BOTH (UP AND DOWN) STREAM
DUST CONTENT
300 mg/Nm 3
100
1501
300
100
996
2292
i
Vi
i
V
i
i
0
500 1000 1500 2000
2500
3000
3500 4000
TEST PERIOD (Hr)
200
100
0
( MATER IA I-lI J(
SOOT BLOWER
TYPE SINGLE NOZZLE
MEDIUM 14 ATG STEAM
€ —€3
ELEMENT CHANGE
&-@
& INSPECTION
INSPECTION
V
V
FIgure 4 26 Operating ExperIence ot Air Preheater vs Soot Blowing (low dust content)
-------
soot blowing is not used. About 80% NO 1 removal efficiency and unreacted
NE3 in the range of 1—3 ppm was achieved throughout the test period.
The air preheater experienced ammonium bisulfate problems when it con-
tained the standard types of elements. As shown in Figure 4—26, the pressure
drop increased markedly even with soot blowing three times a day. After 996
hours of operation, the preheater was washed with water and the elements in
the intermediate— and low—temperature zones were replaced with new notched—
flat elements which permit soot blowing to work more efficiently. After the
change of elements, any increase in pressure drop was prevented by applying
soot blowing three times a day from both cold and hot sides of the preheater.
4.7.3 Evaluation
Based on these tests, Tokyo Electric considers both the high—dust and low—
dust SQ systems to be commercially applicable to coal—fired boilers, the high
dust being superior to the low—dust. Two commercial SCR plants under con-
struction for the Nos. 8 and 9 coal/oil—fired boilers at the Nakoso Station
(600 MW each) will use the high—dust system.
Rowever, the high—dust system does have two problems associated with it
(1) contamination of fly ash by ammonia, and (2) a relatively low cold ESP
efficiency for fly ash from low-sulfur coal. As the tests have demonstrated,
contamination of the fly ash can be eliminated by maintaining unreacted NH3 at
a very low level. The problem of the low ESP efficiency can be solved by
applying wet process POD which is very effectcive at removing particulates.
The tests at Nakoso showed that a particulate concentration of 300 mg/Nm 3 can
be reduced to about 30 mg/Nm 3 by POD (5).
320
-------
4.8 EPDC’S SCR ACTIVITIES (6)
4.8.1 Introduction
In cooperation with five major SCR process developers, EPDC conducted
extensive pilot plant tests of both high— and low—dust S R systems with coal—
fired flue gas. For the full—scale demonstration plant (250 MW) and for a
commercial plant for a new 700—MW coal—fired boiler, EPDC tested the low—dust
system. This section was made in order to maintain a low dust emission level
with various types of imported coals and to eliminate ammonia contamination of
the fly ash.
4.8.2 Pilot Plant Tests with KEI (7 )
Figure 4—27 is the flowsheet of the Takehara S R pilot plant. In the
plant, the 150°C flue gas leaving the existing ESP is heated by a heat
exchanger and electric heater to 300°—400°C. The composition of the flue gas
and the properties of the fly ash are shown in Table 4—16. The composition of
the fly ash is shown in Table 4—17.
TABLE 4—16. FLUE GAS COMPOSITION AND FLY ASII PROPERTIES, SCR
PILOT PLANT, TAKEHARA STATION
Flue gas contents:
NO , ppm 200 — 410
SOL ppm 1100 — 1720
S03, ppm 1—2
Oz,% 5—10%
8—9 %
Dust, mg/Nm 3 46 — 255 mg/Nm3
Dust properties:
Particle size, .i 10 —
Specific gravity Ca. 2.1 (absolute)
321
-------
BOILER FLUE GAS
RATE SIGNAL — — —
8 DIJCI
r
A DUCT
I.. — — —
Figure 4 27 Flowsheet ol the Takehara SCR Pilot Plant
-------
TABLE 4—17. FLY ASH COMPOSITION.
SiOs
Fe 503
A lzOa
CaO
MgO
ViOs
56.91
4.92
21.5
1.57
1.28
0.04
Na O
KaO
PbO
MnO
Ti02
CuO
0.57
0.22
0.03
0.04
1.67
0.01
ZuO
NE
Cl
S04
C03
C
0.02
<0.01
Nil
0.77
0.44
3 .5
The pilot plant uses a vertical S R reactor with a gas down flow and a
tubular catalyst. The objective of the pilot plant test was to attain 80%
NO 1 removal efficiency with unreacted NB3 less than 5 ppm. The S02 oxida-
tion ratio caused by the catalyst was about 1%.
Figure 4—28 shows the relationship between the gas velocity in the reac-
tor (linear velocity), space velocity, gas temperature, and NO 1 removal
efficiency. Over 90% of NO 1 was removed at temperatures above 350°C with an
SV of 4,160 hr’ or less and an NE3/NO 1 mole ratio of 1.0, although a
considerable amount of u.nreacted Nfl3 was released.
Figure 4—29 shows that by using a mole ratio of 1.0 at the SV of 3,120
hr , unreacted N H3 exceeded 5 ppm, although the NO 1 removal efficiency
exceeds 90%. In order to keep the level of unreacted NH3 below 5 ppm, the
mole ratio should be smaller than 0.9, which will result in an NO 1 removal
efficiency below 90%. An SV smaller than 3,000 hr may be used in large—
scale StR units to achieve the same results. Figure 4—30 shows that catalytic
activity did not decrease during the 4,000 hour test period. Nor did the
pressure drop, which indicates that no dust plugging occurred.
Figures 4—31 and 4—32 illustrate the results of special tests using one—
third the normal amount of catalyst. In one of the tests, the gas temperature
was varied between 360° and 290°C with a temperature change rate of 3.3°C/mm.
for the descending rate and 3.1°C/mm. for the ascending rate (Figure 4—32).
323
-------
f ::
70- ________
LVNmIS SVH-
60- 2 2080
3 3120
Z 50- 4 4160
40 NH 3 /NO. = 1 0
300 350 400
REACTION TEMP (C)
Figure 4.2B EliecI of Reaction Temperature on NO Pemo ai Elilciency.
SCR Pilot Plant. TaMehara Station
2667
SV=3120M
rEMPERA1 JRE 350C/’
- 80
Q
z
UI
70.
I’
UI
-S
0
I
02 04 06 08 10 12
NH JNO 1 RATIO
Figure 4. Effect of NI4,1N0 1 Ratio on NO 1 Removal Efficiency
and Unrectsd NH , SCR Pilot Piant. Takehara Station
70A 2665
324
-------
LV 3 Nm/s
NH 3 /NO 1 0
TEMP. 350CC
70
60 ________________
<0
WUJCSJ 50
40
w 30
100
>>- 90
00
Z 80
wUJ
70
0u
Z 60
I I
1000 2000 3000 4000
TIME (hr)
Figure 4.30 NO Removal Efficiency and Pressure Drop During 4000 hr
Test Period, SCR Pilot Plant, Takehara Station.
325
-------
NO REMOVAL
EFFICIENCY (%)
OUTLET NO
COCNETRATION
FLUE GAS
TEMPERATURE
INLET NO
CONCE NTAAT ION
(ppm) (CC) (ppm)
SOX
CONCENTRATION
(ppm)
0 0
‘1
C
C.)
0
0 0
-1-4
(n
U)
-4
C a
m
C a
C
C l)
C)
0
-o
Ca
-------
?
—
0
70
NH 3 INO PATIO 1
SV=9360H
0 350C
I
I
z
60
I
U-
50
NH 3 FEEDING
DISCONTINUED
FOR ABOUT l7hr
-J
<
UJ
40
30
•
•
d
z
20
10
•
270C
I i I
0 50 100 150 200 250
0 20 40 60 80 100
TIME (hr)
Figure 4.32 NH 3 Inpectlon at Low Temperature and Recovery Test.
SCR Pilot Plant. Takehara Station
327
300CC
-------
No change in the NOx removal efficiency was observed before or after the
test. Figure 4—33 shows that the N0 removal efficiency decreased from 60%
to about 30% when the gas temperature was maintained at 270°C for 250 hours
due to ammonium bisulfate deposition on the catalyst. When the gas temper-
ature was raised to 300°C , the N0 removal efficiency increased to over 40%.
When the gas was heated even more (3500C), catalytic activity returned to its
initial level. These results indicate that the catalyst was resistant to
attack by the ammonium bisulfate.
4.8.3 Demonstration Plant
Basic information on the SCR demonstration plant at the Takehara Station
is presented in Table 4—18 and Figures 4—33 and 4—34. The No. 1 coal—fired
boiler (250 MW) has two trains of air preheater/cold ESP/ ID fans. After
passing through this system. all of the boiler flue gas is treated by wet
limestone/gypsum FGD. For the demonstration test, a system with two trains of
hot ESP/SCRJfans and modified air preheaters was installed. One of the SCR
reactors is manufactured by BHK and uses a plate catalyst similar to the one
used at Hokkaido Electric’s Tomato—Atsuma Station (Section 4.5) with an SV of
2,300 hr’4 the other is manufactured by I and uses a tubular catalyst
with an SV of about 2,000 hr .
The air preheater used in the demonstration has a modified arrangement of
elements and soot blowers as shown in Figure 4—35. Ammonium bisulfate tends
to deposit between the intermediate— and low—temperature zones in a conven-
tional air preheater and is difficult to remove by conventional soot blowing.
To solve that problem a combined intermediate— and low—temperature element was
installed which permits soot blowing to work efficiently. The new fans in-
stalled in the S R demonstration plant are compared with the existing ones in
Table 4—19. The increase in the capacity of the induced fans is designed to
compensate for the increase in the pressure drop caused by S R reactor and hot
ESP. An increased capacity in the forced draft fans is required by the large
distance between the fan and the boiler caused by the retrofitting of the SQ
system.
328
-------
TO
STACK
EXISTING
— — — — NEW
L: }
I AIR
PRE-
HEATER
—--i r I
I I
HOTL SCRI ø
ESPI I I I
L__J L...__J
I ID I
FAN I FGDI ‘S K
___J
Figure 4.33
MODIFIED ARRANGEMENT
Equipment Modifications Made for SCR System Installation
at EPDCs Takehara Station. (8)
EXISTING ARRANGEMENT
70A2672
-------
TAJIIE 4-18. DFSUIIIIION OF SCR I)UNONSI1 1ATION PLANT. TAKFJIARA STATION NO. I UNIT
Efficiency
(Guaranteed) Remarks
Item Manufacturer Type
(1) lusting Plant Reheat Type
Radiant
(a) Boiler 01 1K Boiler 810 i/h Fuel: coal
(b) Plant output — — 250 101
(2) Demonstration
Plant
(a) StE BIlK Plate Catalyst 400.000 Nm’/h >8O’ .
Kill lube Catalyst 400.000 N&Ih >80’u
(b) Air Prchcater Gadilius Long element 400.000 N&/h — New Air
Ljungstrom x 2 Preheater
(c) l iI. Removal 1CL Activated 1.000 m’/Day (10 ppm
from III) Sludge as total N
lilowdown
-------
41 /3/ I
fl BOILER
2/ HOT ESP
3/ SCR REACTOR
4/ AIR PREHEATER
5/ LIMESTONE/GYPSUM
6/ STACK
7/ NH STORAGE/
‘IA P C RIZATION
Figure 3 34 Arrangement ol the 5CR Demonstration Plant Takeflara Station EPOC 3
INTERMEDIATE I COLD SCOT
HOT I I I
25mm 5mm SLOWER
CONVENTIONAL ARRANGEMENT OF HEAT TRANSFER ELEMENTS
SOOT ______
_____ HOT [
BLOWER
COMBINED
INTERMEDIATE AND COLD
5 mm
BLOWER
MODIFIED OESIGN BY GADELIUS K K
F.qures 5110w clearanCe Deiween IIle elemenls I
Figure 4 35 Conventional and Mod,(ied Air Preheater Designs 19)
I)
‘6)
331
-------
TABLE 4-19. COMPARISON OF EX.ISTD G AND NEW FANS AT TAKEHARA POWER
STATION (KW)
Fan
Existing
New
Forced
Draft
(air)
630 x 2
1350 x 2
Induced
Draft
(flue gas)
1350 x 2
2350 x 2
The Takehara SCR units began operation in 198]. and have since been oper-
ated with an NO 1 removal efficiency over 80% and 5 ppm of unreacted Nib. An
ammonium bisulfate problem occurred in the air preheater during the initial
stage of operation when unreacted Nib was about 10 ppm.
The small amount of ammonia contained in the flue gas after SCR is caught
by the wet FGD system. The FGD scrubbing liquor also contains nitrates de-
rived from NO 1 . In order to reduce the nitrogen concentration to below 10
ppm in the FGD system wastewater, a conventional activated sludge treatment
system has been constructed by Hitachi Zosen.
The total investment cost for the demonstration plant was 8 million yen
including the cost of the hot ESPs, air preheaters, fans and wastewater treat-
ment system. This cost is relatively high, due to the expense of retrofit-
ting, as well as the cost of equipment required for the demonstration tests.
According to EPDC personnel, the investment cost for an S R system designed to
remove 80% of the NO 1 from coal—fired boiler flue gas was 7,000 yen/kW in
1981. Assuming a 2—year catalyst life, the annualized cost for this same sys-
tem was about 0.6 yen/kWhr (including a 7—year depreciation schedule).
4.9 ScR COSTS
4.9.1 Investment Costi
Table 4—20 shows the investment costs of 9 S R plants for utility boil-
ers. Chubu Electric ’s 2 SCR plants for LNG—fired boilers at its Chita Station
332
-------
TA1JLE 4-20. INVES1IIENT COSTS OF SCM PLANTS 10 K UTILITY BOiLERS
bketrof it
C
d t)c0ck hitachi
Mitsubishi lIi.avy Industries
IshikowaJIwEi—lharima heavy Industric
I.awasski heavy Industries Reactor baa a capacity (or 80 NO 1 removal.
1ncluding bollsr modifIcation for economizer by—peas.
Rough approziaiation
t- )
Company
Station
Fuel
Size
( IhiY)
N/R
N03 Outlet
removal ( ) NO 1 (ppm)
Vendor
Up
(yen/lW)
Chubu Llc trlc
Chita
I.NG
l.NG
Oil
700
700
700
Na
N. 0
R
80
80
80
8
8
18
11 1 1 c
WI
MIII
1978
1978
1980
2400
2400
3600
Chu 6 oku Lioctric
Kudamatsu
Shimonoseki
Oil
Oil
Coal
375
700
175
R
N
Ii
80
80
50
32
26
200
11110
III !
MIII
1979
1979
1980
5900
3200
97008
hlokkaido Electric
Tomato—Atauma
Coal
350z1/4
N
>80
>40
DII
1980
10000 h
LI’ I IC
Takehara
Coal
Coal
250z1/2
250x1/2
R
80
80
80
80
DII
KIllt
1981
1981
7000 h
7O0O l
-------
are less costly (2400 yen/kW) than the other plants because they use clean gas.
treatment which permits the use of small amounts of a pellet catalyst with an
SV of 20,000 hr’. The SIR plant for the existing oil—fired boiler at the
Chits Station cost 3,600 yen/kW using a honeycomb catalyst with an SV of 5,100
hr .
Chugoku Electric’s SIR plants for oil—fired boilers at the Kudamatsu
Station cost 5.900 yen/kW for the existing boiler and 3,200 yen/kW for the new
—i
boilers. Both systems use a honeycomb catalyst with an SV of about 5.100 hr
The relatively high cost of the plant for the existing boiler was a result of
the cost of the extensive ductwork required for retrofitting, and the instal—
lation of new fans needed to compensate for the pressure drop caused by the
S R system (Section 4.3). The S R plant for the existing oil—fired boiler at
the Chita Station did not require an additional fan because the existing
boiler fan was able to compensate for the increased pressure drop on its own.
The S R plant for the existing coal—fired boiler at the Shimonoseki
Station cost nearly 10,000 yen/kW including the cost for the boiler modifica-
tion needed for installation of the economizer by—pass system (Section 4.4).
Hokkaido Electric’s S R plant at the Tomato—Atsuma Station cost about 10,000
yen/kW and has a capacity of treating one—fourth of the flue gas from a new
350 MW boiler. The relatively high cost of this SCR system for a new boiler
may be due to the small size of the plant (88 MW equivalent) and the fact that
it removes 90% of NOx. EPDC’s two SCR plants for existing coal—fired
boilers at the Takehara Station cost about 7,000 yen/kW each. Generally
speaking, an SIR plant which achieves 80% NOx removal for a new coal—fired
utility boiler may cost 6,000 — 8.500 yen, depending on inlet S0 and N0
concentrations and the amount of unreacted Nib allowed.
4.9.2 Estimated Investment and Annualized Costs
Tables 4—21 and 4—22 show the estimated investment and annualized costs
of SCR for new 700—MW utility boilers burning coal, oil, and gas. In prepar-
ing these cost estimates the following assumptions were made, based on a study
of the costs of commercial and demonstration S R plants.
334
-------
TMII I. 4-21 (Al ( III All Ii I4ISIS I UK AN SIR l I AN I 111K A NE8 700 (11 cOAl —FIRI U lUll, 1k (7(1% ll,,i I liii I an I Ion 4,292 . 1 1 10 ) 1 1 6k rJ year. I 981 y o
(liar (.as 110w rate 2.300.000 U&/hr. boneyconb catslyat renI square type. 3 5 sutton y o/n ’. 2 year life)
0011cr basin l ’araneter 11 1gb S Coal Iligh S (oat Icon S (el low S Coal
t O) • 600 300 300 300
54) 5 ppn 2500 2500 600 60(1
i.Jttenlstas. No’ 15—23 15—25 1325’ 0 050 2O
Ilesign dais Catalyst sins. no + 1.6k ’ + 6 h 7.4 I 6 6 •
NI ). eninalons. pp 4_S 4 5 45 2_3
N I) rewoval. % itO 90 80 90 80 9(1 80 90
NIIX./N0 1 nolu wstio_ 5 0.82 0.92 0 83 (I 93 0 83 I I 93 0.82 0 92
Sparc velocIty. hr 2200 1600 2700 2000 3000 22(10 3000 22(10
Catalyst voluoc. & 3045 1438 852 1150 767 1(145 767 1045
Reactor prensarc drog. m/ii,(I 130 150 115 135 110 13(1 11(1 130
l ’ow.r consuoplios. % 0.20 0.24 0 18 0 22 0.11 0 20 (1 19 0 20
lnvcstnent cunls Catalyst. 10’ cn 3.66 5 03 2.98 4 02 2 68 3 66 2 68 3 66
Other. 10’ YCOb 2.23 2.50 2 10 2 35 2 00 2 JO 2 00 2 30
blat. 10’ ye 3 91 7.33 3.08 6 37 4.68 5 96 4.68 5.96
rotsi. yøn/IW 8440 10760 7260 9100 6690 14510 6690 8510
Annual .osts F ,aed coal. Ion y 0 G 0 92 1.13 083 099 0 77 0 95 0 77 0.95
Catalyst. 10’ yen 1 83 2.32 1.49 2 01 1 34 I 83 1 34 1 83
An,oonls. 10’ y a 0.53 0.39 0 27 11 30 0.27 I I JO 0 26 0 29
Otbr. 10’ yen 0 20 0.24 0 18 0 22 0 16 4) 20 0 16 0 20
lotsl. 10’ yea 3.48 4.48 2 77 3.52 2.54 3 28 2 53 3 27
Annualized cost. ycn/8whr 0.81 1.04 0.65 0 82 0 59 0 76 0 59 0 76
Cost cffcctlyeneas. yea/Nm’ of NO 1 reoovcd 514 388 826 924 149 861 749 861
dollarsllb of NO renovad 1 0.45 0.31 0.73 0.81 0.66 0 76 0 66 076
1’ercenl of powi.r 8 eneratcd by the boiler
Includtng civil engineering sad tast Operstioll
o1 catalyst and 25% of investueni eneludiag estalyat
Catalyst .051 divided by life (yesra)
(At 10(1 000 yea/ton
Mainly for poecr (18 yen/Kwhr)
8250. NO 5 as NO
CbaoacI d,noeter plea n .h thickness
8lth full dnst load
3 lcavlng a hot ESP
-------
TAIII 1 4-22 N 401*101) COSTS P09 1CR IO8 NOV 700 Ill OIL— *1111 OAS-l 11(U) SUILEIS (10% boll Cr utilizatIon 4.292.000
tmhrlyesr) (191(1 coats, fIqo jas floe rat. 2,000.000 Ib/bz. c .rs.Ic nqoaro boneycoamb and pollol catalysis)
NO • pp
Son.
Panti ulst.a. •II
Catalyst type Pellet
Catalyst nine. S
CstsIyoi life, peel 4
Catalyst coat. 10’ yea/& 3 0
Nit. .amtnstons. p 6—8
Ott raamooal. % 80 90
IIll /t40 eUle r.tlo I 00 l,l0
Space elocity, br 18000 13000
Catalyst voluam.. •‘ ill 124
Reactor pressare dro , /ll ,O 120 140
Poer consueptlon. % 0 Ii 0 20
Catalyst. l0
Other. o
Total. 10 ye
Total. yea/kR
Flied cost, 10’ ,V .c
Catalyst. 10’ yan
Aamamonla. l0 yfn
Other. 10’ yes
Total. 10’ yes
Annual lz .d .unt (yea/lW/br) 0.44 0 56 0 33 0 41
Cost efIe,.tuvrne,s (yon/P ’ of NU remoced) 938 101(7 719 794
(dollars/lb of 140 roamootd)’ 0.1(4 0.96 0 63 0 70
bP0 tflt 01 l1 5tl cncrsted by the ballet
ts Iudlng cn ml angImiocrIng and test operation
cstal)st and 23 . of lnnentrnt eaclndla 1 ( catalyst
Catalyst scat dmvldsd by lilt (y.srs)
fAt lOft 0(1(1 ysn/too
llstnly I.. Ions, 121 1 sn/kahr)
iii 8250 NIl as 14mm,
(h. ,ineI (Is cmtcr pies. call Ihlck,enn
With Call stoat toad
macmt a lust
Roller boats
Paraneter
Ui 1 b S Oil
to. S Oil be S Oil
Gas
lo)
0 ’
200
1.300
0.03-0 20
lbon•ycosbb
6.6 + 1.4
3
3.3
2—3
Ueai n daIs
Investment coat
Annual Coats
200
100
0 02-0.05
lloaeyco b
3 8+1 .2
3
33
4— i
60
0 OOS O.O 2
100
100
0 02-0.0S
lios.yco b
5 8.1.2
33
4—5
80
0.1(3
0.1(3
0 84
0 94
0 88
0 98
2800
2000
4300
3100
3100
3700
714
1000
465
645
392
341
120
140
lOS
ItS
tOO
110
2.36 3.30 I 33 2 13 1.29 1.78 0 33 0.46
2.00 2.23 1 70 I 90 I 60 I SO 1 40 1.50
4.36 5.33 3 23 4 03 2 89 3.58 I 73 1.96
6230 7930 4610 5160 4130 SIlO 2410 2800
0.73 059 058 069 053 063 038 042
0.79 1.10 0 SI 0 71 0 43 0 29 0.08 0 12
0.16 0.17 0 16 0 18 0 08 0 09 0.05 0 06
O 20 0.24 0 18 0 19 0 17 0 19 0 20 0.24
1.88 2 40 1.43 I 77 I 21 1.50 0 11 0 84
0 28 0.35 0.17 0.20
1213 1343 1203 1270
I 07 1.18 I OS 1.12
-------
(1) As shown in Figure 4—36, flue gas from the boiler economizer is
treated in the two trains of the S R system. In most systems,
the gas is first injected with ammonia and sent into the SIR
reactor. The exception is the low—dust system for coal (C in
Figure 4—36) in which the flue gas is sent through a hot ESP
and then into a reactor.
(2) A square honeycomb catalyst is used except in the case of the
the system for the gas—fired boiler which uses a pellet cata-
lyst. The channel diameter of the honeycomb is 7.4 mm for coal
with a full dust load, 6.6 mm for coal with the low—dust system
and high—sulfur oil, and 5.8 mm for low—sulfur oil. The cata-
lyst cost is 3.5 million yen/n 3 for coal, 3.3 million yen for
oil, and 3 million yen for gas. (Plate catalysts may be used
instead of honeycombs for the same or a slightly lower cost.)
(3) Unreacted N113 is maintained at 4—5 ppm for coal with a full
dust load and low—sulfur oil. It is maintained at 2—3 ppm for
coal with a low—dust system and high—sulfur oil because, in
these cases, ammonium bisulfate deposition in the air preheater
may become a serious problem if the unreacted NEi concentration
is any higher.
(4) NO removals of 8O and 9O were assumed, which actually re-
quire 8l—82 and 9l—92 removal, respectively. For 9O NO 1
removal, 35—4O’ o more catalyst is used than for 8O removal. A
higher inlet NO 1 concentration also requires a larger amount
of catalyst.
(5) The catalyst—induced SOi oxidation ratio is maintained below
O.5 for high—sulfur fuels and below 1 5 for low—sulfur fuels.
(6) The annual power generation is 4,292,499 MW (7O of full capac-
ity)
-------
AH TAcK
GAS AND LOW-SULFUR OIL
B TAGK
HIGH SULFUR OIL AND COAL (HIGH DUST)
C TAcK
COAL (LOW DUST)
B BOILER [ APH AIR PREHEATER
HE !J HOT ELECTROSTATIC PRECIPITATOR
Figure 4.36 Flue Gas Treatment Systems.
YOA 2575
338
-------
(7) The investment cost includes costs for reactor, catalyst,
ducts, ammonia tank and injection system, soot blowing system,
control system, civil engineering services, and test operation.
(8) The fixed capital cost includes a 25% investment cost excluding
catalyst expense and 10% interest for the initial charge of the
catalyst.
(9) The catalyst life is 2 years for coal, 3 years for oil, and 4
years for gas. The annualized cost includes the catalyst cost
divided by the catalyst life (years).
(10) The cost of ammonia is 100,000 yen ( 400)/ton. The power cost
is 18 yen/kWhr for coal and 21 yen/kWhr for oil and gas.
Table 4—21 shows that the investment cost for 80% NO 1 removal is 8,440
and 7,260 yen/kW for high—sulfur coal with 600 and 300 ppm inlet NO 1 con-
centrations, respectively. The cost for low—sulfur coal is 6,690 yen/kW for
both the high—dust and low—dust systems. Compared with the high—dust system,
the low—dust system uses the same amount of more efficient catalyst to main-
tain a lower level of unreacted NH3. The investment cost for 90% NO 1 re-
moval is about 30% higher than that for 80% removal. The catalyst cost makes
up over 55% of the investment cost for 80% removal and over 65% for 90% re-
moval.
The annualized cost for 80% NO 1 removal is 0.81 and 0.65 yen/kWhr for
high sulfur coal with 600 and 300 ppm inlet NO 1 concentrations, respective-
ly, and is 0.59 yen/kWhr for low—sulfur coal. Ninety percent NO 1 removal
costs about 30% more than 80 removal.
Although the annualized cost for NO 1 removal with 600 ppm inlet NO 1
is about 25% more than that for 300 ppm NO 1 , the cost effectiveness is much
lower ——— 45 cents vs. 73 cents per lb of NO 1 removed for 80% removal. With
90% NO 1 removal it is nearly 20% more cost effective to remove NO 1 from
gas containing 300 ppm than from gas with 600 ppm NO 1 .
339
-------
In actual practice it is difficult to consistently maintain 90 NO 1
removal with less than 5 ppm of unreacted NHi when treating a large amount of
NOxrich flue gas such as that from a large coal—fired boiler. This is true
for two reasons. First, the NO 1 concentration as well as the gas velocity
may be considerably different in different areas of the large duct and reactor
inlet. Second, the boiler load is not constant, resulting in a fluctuation in
both gas volume and NO 1 concentration, and the NO 1 analyzer has a delay
time of several minutes. In order to reduce NO 1 from a large boiler by 90%,
the combination of combustion modification for 35—50% removal and SCR for 80—
85% removal may be more practical and economical than the use of SCR by it-
self.
Table 4—22 lists the costs of SCR systems for oil— and gas—fired boilers.
The most common type of S R plant in Japan is used for 80% NO 1 reduction in
flue gas from an oil—fired boiler containing about 100 ppm NO 1 . The invest-
ment cost is 3,860 yenlkW, the annualized cost 0.27 yen/kW. NO 1 removal of
200 ppm inlet NO 1 in low—sulfur oil flue gas costs about 20% more than this;
removal of NO 1 from gas—fired boiler flue gas containing 60 ppm NO 1 costs
about half as much.
S R for boilers which burn high—sulfur oil may require about 60% more
catalyst than those which burn low—sulfur oil, given the same inlet NO 1 con—
centration and NO 1 removal ratios. This occurs for two reasons: (1) In the
high—sulfur oil system the catalyst is less reactive. This is because of the
low SOs oxidation ratio and the larger channel diameter needed for the larger
amount of dust. (2) It is necessary to maintain unreacted NEs at a lower
level in the high—sulfur oil system in order to reduce aonium bisulfate
deposition in the air preheater. Thus, the StR system costs for high—sulfur
oil—fired boilers are about 40% more than those for low—sulfur oil—fired
boilers and are similar to S R system costs for low—sulfur coal—fired boilers.
The investment cost for an SCR system applied to an existing boiler may
be 10—40% more than that for a new boiler because of the retrofitting require-
ments and the possible need for new fans. When an SCR system is applied to
150°C flue gas leaving an air preheater. the system investment cost may be 40—
340
-------
60% more and the annualized cost 50—80% more than that for SCR applied at the
economizer outlet because of the requirement for gas heating. Although low—
temperature catalysts which work at 150_2000C have been developed, these cata-
lysts experience amnionium bisulfate problems. Applying S R to the economizer
outlet is more economical than using low—temperature catalysts at the air pre-
heater outlet, even for existing boilers.
4.9.3 Power Generation and Flue Gas Treatment Costs
Table 4—23 gives estimated costs for a new coal—fired power station with
flue gas treatment systems in Japan. The total investment cost Is 220,00
yen/kW, 10% of which is for FGD. 32% for SCR , and 1.8 for ESP. The total
annualized cost, including the cost of flue gas treatment is 18 yenlkW. Of
the total amount, 10% is for FGD, 4% for SCR, and 0.5% for ESP. The annual-
ized cost for this coal—fired station is less than the cost of an oil—fired
power station that uses ESP without FGD and SCR.
TABLE 4-23. POWER GENERATION COST FOR A COAL-FIRED BOILER IN JAPAN (1981)
Investment cost (yen/kW)
ESP (99% removal) 4,000
SCR (80% NO removal) 7,000
FGD (90% SOa removal) 30,000
Total Power Plant (including boiler,
turbine, generator, coal treatment,
flue gas treatment, plant site, and
auxiliary facilities) 260,000
Annualized cost including capital cost (yen/kWhr)
ESP 0.1
SCR 0.7
FGD 1.9
Total power cost (including flue gas
treatment) 18.0
341
-------
4.10 USE OF JAPANESE SCH TECHNOLOGY IN THE UNITED STATES
4.10.1 Introduction
As Japanese SC technology has become more reliable and less costly some
U.S. utility companies have become interested in using it. Various Japanese
SCR processes have been licensed to U.S. firms (Table 4—24) and a number of
pilot and demonstration plants have been constructed.
The U.S. Environmental Protection Agency (EPA) sponsored an 0.5 MW pilot
plant test of the Hitachi Zosen SCR process at Georgia Power Company’s
Mitchell Station. The Electric Power Research Institute (EPRI) constructed a
2.5 MW pilot plant which uses Kill’s SCR system at Public Service Co. of
Colorado’s Arapahoe Station. Southern California Edison has constructed a
107.5 MW I Scil demonstration facility with a capacity of treating one—half
of the gas from a 215 MW oil—fired boiler at its Huntington Beach Station. In
addition, till’s SCH catalyst will be used at an SCH pilot plant for an indus-
trial boiler.
This section describes the pilot plant tests of the Hitachi Zosen process
at Georgia Power Company’s Mitchell Station.
4.10.2 Hitachi Zosen Technology Pilot Plant Tests at the Mitchell
Station (1 )
Hitachi Zosen received an order from EPA for a 1,700 NM 3 /hr (0.5 MW
equivalent) capacity demonstration SCH plant which would remove 90% of NO
from coal—fired boiler flue gas.
Qiemico Air Pollution Control Corporation, Hitachi Zosen’s North American
licensee, constructed the demonstration plant at Georgia Power Company’s
Mitchell Station. The boiler burns Kentucky coal containing about 1% sulfur.
For the test, flue gas (1,700 Nm 3 /hr) from the boiler economizer with a full
dust load was used. Since the gas cools while passing through a long duct
-J
-------
TABLE 4—24. PRThCIPAL V 4DOR CONTACTS FOR JAPANESE NO SCR TECUNOLOGY (3)
Process Japan Representative U.S. Licensee or Partner
Babcock— [ itachi
(Hitachi, Ltd.)
Ciroshi uroda
Kure Works. Babcock—
Hitachi R.R.
No. 6—9 Tal .aroMachi
Kure—shi. ijiroshima—Ken,
737. Japan
(0823) (21) 1161
Telex 6624—21 BIlK KRE—J
Greg T. Bielawski
Babcock and Wilcox
FPG—CSC
P.O. Box 351
Barbarton, Ohio 44203
216/753—451
Hi tachi—Zosen
(Hitachi Ship-
building and
Engineering)
Shingo Tanaka
Hitachi Shipbuilding &.
Engineering Co. . Ltd.
Palaceside Building
1—1, Bitotsubahsi. 1—Chone
Chiyoda—Ku. Tokyo, Japan
Tokyo (213) 6611
Telex 122363 • 324490
Iahikawaj ixa—
Eari a
Heavy
I odu St rica
(1 ) 12)
Nobno Aoki
Ishikawa3 i a—Earima Heavy
Industries Co., Ltd.
Tokyo Genboku gaikan
30—13 I 5—chrone, Tokyo
Koto—Ku. Tokyo, 135, Japan
(03) 649—1111
Telex (IBI CO) 322232
John Cvicker
Foster Wheeler Energy Corp
9 Peach Tree 11611 Road
Livingston, ! J 07039
201/ 533—2687
Kawasaki
Heavy
I ndus tries
( I)
Senji Niwa
Kawasaki Heavy Industries,
Ltd.
14, 2—Chome
EigaahikawasakiCho
Ikuta—Ku. Kobe. 650—91
Japan
Tel. Kobe (078) 671—5001
Telex 5624—032 I CPL I
Joy Manufacturing
Western Precipitation Div.
P.O. Box 2744
Los Angelea, CA 90051
213/240—2300
Mitsubishi
Heavy
Industries
(l I)
T.damase Sengokn
Mitsubish Heavy Industries.
Ltd.
Shin—Tamachi Bldg.
34—6. Shiba 5—chone
Minato—Ku. Tokyo 108
Japan
(03) 455—5711
Telex 322282 HISHIJU
128578 ) ITAM
Donald I. Frey
Manager. Fuel Systena
Engineering
CE Power Syete s
Conbustion Engineering,
Inc..
1000 Prospect Hill Road
Windsor. Conu. 16095
203/688—1911 Ext. 2241
343
-------
FIgure 4.37
SCR Pilot Plant and Fly Ash Sampling Points,
Mitchell Station, Georgia Power Co.
70A2676
(A)
(B)
NH 3
REACTOR
(C)
344
-------
Figure 4.38
NOXNON 600 Catalyst.
345
-------
(14” diameter), it is electrically heated to about 4000C and then injected
with ammonia before it passes down through the StR reactor (Figure 4—37).
A NOXNON 500 catalyst was used for the first and second pilot plant
tests. The catalyst was prepared by treating the surface of a special all oy
plate; it is similar in shape to the NOXNON 600 catalyst used for the third
test (Figure 4—38).
The first test using the NOXNON 500 catalyst and an NH31N0 1 mole ratio
of 1.0 initially produced a 90% N0 removal rate. However, this efficiency
decreased during the course of the 2,000—hour test run as the pressure drop
increased. Tests on the used catalyst showed that the decrease in efficiency
was not caused by poisoning of the catalyst but instead, by fly ash deposition
on the catalyst surface.
For the second test, a new NOXNON 500 catalyst was used and soot blowing
was added. Despite these measures, the NOx removal efficiency still de-
creased during the 2,000 hour test period.
A N0 0N 600 catalyst was used for the third test. The N0 0N 600 has a
stainless steel mesh base which is coated with catalyst material and shaped as
shown in Figure 4—38. The pitch of the NOXNON 600 catalyst is 14 mm whereas
the NOXNON 500 has a pitch of 8 mm. Because of the wider pitch of the former
catalyst, the reactor was lengthened for the third test to enable it to attain
90% removal. Use of this catalyst greatly reduced the reactor’s dust clogging
tendency.
Results of demonstration tests conducted from May through October 1980
using the NOXNON 600 catalyst are shown in Figure 4—39 (9). The results
represent over 3500 hours of operation. During the first 90 days of operation
the NO removal efficiency averaged over 90 percent. For the entire test
period of 160 days of continuous operation. the average NO reduction was
only slightly lower——89.8 percent. Figure 4—39 also illustrates that when it
remains in the range of 2100 to 2400 Nm 3 /hr, the flow rate has essentially no
effect on the S R system’s NOx removal efficiency.
346
-------
100
:
80-
n.u t
•‘
• . ,...c.t at • ,sn. •tt.,
a’ ;,.
. •
70
12
1 0
140
0 1• • 44 tflW i -t ø tt••
n
.t.t. ,,,t,t
4
4,’,it.
tttt,t.,tt.*mflhltttt 4”..
4 t t.t
ti_itt..
•-ti t
tt.
08
-
‘
C
0
NOTE ADDITIONAL LOW MOLE RATIO TESIS WERE RUN ON THESE TWO DAYS
C
4
JUNE JULY
Figure 4 39 Demonstration Test Results trom the SCR Pilot Plant, Mitchell StatIon,
GeorgIa Power Co. (9)
S
-J
>
0
UJ
0
z
0
I-
ft
0
z
I
z
E
0
I t
0
-J
U-
U.)
-I : .-
t ‘ , • Itflt lt 4 tti
C
1980 MAY
‘ C
C C
t Ct t
C
,4.• 4.’ e,int’t ,ni4flttne•t.r, lttt.Ottt•tittt•ttC .4 •tt tie...
.4..” 4
AUGUST SEPTEMBER OCTOBER
70A2678
-------
Since catalytic activity decreased slightly during the 3,500—hour test
period, a regeneration technique was developed to restore the activity to its
initial level.
Figures 4—40 and 4—41 show the results of tests using NO ON 600 with
different area velocities and NH3/N0 1 mole ratios. Over 90% NO 1 removal
was attained with a mole ratio of 1.0 and an area velocity smaller than 12
m 3 /m 2 hr. However, with the mole ratio of 1.0, unreacted NH3 exceeds 30 ppm.
One gas analysis at the pilot plant indicated that at 90% NO 1 removal,
unreacted NB3 reached about 50 ppm and that the S03 content of the gas was 8.4
ppm at the reactor inlet and increased to 20.7 ppm at the outlet. (10)
4.10.3 Evaluation
The SIR pilot plant tests at the Mitchell Station indicated that with a
90% NO 1 removal, both unreacted N113 and SOs levels are high and can contami-
nate the fly ash. To lower unreacted NUs to below 5 ppm, a mole ratio of 0.82—
0.83 should be used to remove 82—83% of NO 1 . In order to reduce S03, a low—
oxidation, and therefore less active, catalyst should be used. An 80% NO 1
removal efficiency appears to be a practical target for treating a large
amount of flue gas with a high NO 1 concentration.
The NOXNON 500 catalyst proved to have a considerable dust clogging
problem. With its wider pitch, the NOXNON 600 performed well, but still was
not free of the dust clogging problem. On the other hand, during Hitachi
Zosen’s SQt test at EPDC’s Isogo Station, the NO ON 500 catalyst experienced
no dust clogging problem. This indicates a difference in the nature of the
fly ash at the two test sites.
In order to determine the reason for this difference, the author tested
samples of fly ash obtained from three locations at the Mitchell Station pilot
plant (Figure 4—37, A, B, and C): the commercial ESP, the heater, and the
reactor.
348
-------
-J
>
0
LU
0
z
10
AREA VELOCITY M 3 1M 2 -Hr
MOLE RATIO 1 0
REACTOR TEMP 720F
INLET NO, CONCENTRATION 400-450 ppm
Figure 4.40 NO Removal vs. Area Velocity (1)
70A 2679
349
-------
CATALYST NOXNON 600
TEMP 700-710F
FLOW RATE 1500 scfm
1300 scfm
INLET NO, 400.500 ppm
Figure 4.41 NO Removal vs Mole Ratio (1)
-j
0
LU
0
z
E
w
(D
0.
0.
-J
z
MOLE RATIO NH 3 /NO ,
70A2680
350
-------
Figures 4—42 and 4—43 are scanning electron microscope photographs of the
samples. The author compared these samples with fly ash samples obtained at
S R plants in Japan. The chemical composition of the samples is shown in
Table 4—26. Ash A has a clean smooth surface similar to the Japanese ashes as
shown in Figure 4—42. By contrast, ashes B and C have many deposits on their
surface. Such deposits appear to give the ash an adhesive quality. Some of
the ash B particles contain crystals which have grown perpendicularly on their
surfaces as shown in Figure 4—42. This indicates crystal growth caused by gas
phase reaction.
The author presumes that these deposits on the ash, including the
well—grown crystals, were formed by the reaction of S03 and metallic vapor in
the combustion gas. It seems likely that a considerable amount of S03 is
formed in the heater at the Mitchell Station. This heater contains fine
stainless steel tubes which are at a temperature above 4000C. Stainless steel
has been known to act as a good oxidation catalyst of SOi to S03. Although a
similar heater has been used by Hitachi Zosen at Isogo. the S03 content of the
flue gases at the two stations are different———about 1,000 ppm at the Mitchell
Station and about 300 ppm at Isogo. At the Mitchell Station a much larger
amount of $03 may form on the surface of the hot tubes. It is on these same
tubes that the fly ash deposits and undergoes the reaction described above.
It has been found that various metallic components in coal vaporize upon
combustion and condense on the ash surface during cooling. As shown in Table
4—25. ashes B and C contain more than 10 times the amount of potassium that
the Japanese ashes contain; potassium is volatile at high temperatures. Other
metallic components also may be involved in the formation of the deposits.
The author presumes that the combination of those two factors caused
deposit formation and made the ash adhesive; the deposits have a much lower
melting point than that of the ash particles themselves and may be quite
adhesive at 400°C. The well—grown crystals may have been formed while the ash
remained on the surface of the stainless steel tube. Some of the ash par-
ticles may occasionally dislodge and deposit on the catalyst.
351
-------
l gim4 42J Scanning Electron Microscope Photographs
of Fly Ash at Plant Mitchell (I) (x3000)
(A) (FROM ESP)
(B) (FROM HEATER)
352
-------
(C) FROM REACTOR)
Scanning Electron Microscope Photographs
of Fly Ash at Plant Mitchell (II) (x3000)
(B) (FROM HEATER)
9 u?P43
-------
TABLE 4-25. COMPOSITION OF FLY ASH (%)
- Mi,tcbe
B
11 Stati
on
Isoao
C
(1)
(2)
Si0 2
48.0
47.8
45.1
44.5
A1sO
Fe203
32.1
12.1
31.3
10.7
21.5
5.2
22.2
5.7
T iOi
1.3
1.2
1.4
1.3
CaO
MgO
1.8
1.3
1.8
1.4
1.1
1.8
4.7
1.9
NaaO
0.3
0.3
0.9
0.6
KzO
S04
2.8
a
3.0
0.2
0.08
9.1
a
0.24
5.3
a
An example of dust clogging on an S R catalyst in Japan took place during
treatment of flue gas from an iron—ore sintering machine. The gas contained
large amounts of S03, FezOz dust, and potassium vapor which resulted in
deposits of a double salt and potassium ferric sulfate on the catalyst.
Since ash A has a clean smooth surface and is not adhesive, it is pos-
sible that the fly ash at the Mitchell Station would not be adhesive if the
stainless steel heater were not used. On the other hand, surfaces of ash
particles obtained at Isogo, Takasago, and the Mitchell Station (A) are not en-
tirely clean, but have many fine deposits, although the amount of deposition
is far less than that of ashes B and C. Further study on the surface proper-
ties of these ashes may be needed.
A gas analysis of the S03 concentrations at the inlet and outlet of the
SQ reactor at Mitchell indicates that the SOs oxidation ratio was about 1.2 .
Although the oxidation ratio is not very high, a low—oxidation catalyst may be
desirable for treating SOs—rich gas. In addition, oxidation at the catalyst
surface may have promoted dust clogging.
354
-------
Catalyst shape also has an effect on dust clogging. As shown in Figure 4—
44, the wave catalyst has many corners with small angles in which the actual
gas velocity Is slow. This results in dust deposition in the corners and soot
blowing does not efficiently remove these deposits. Triangle catalysts, such
as N0 ON 500 and 600 also have a tendency for dust plugging at the corners.
Hexagonal or simple plate catalysts are the best type of catalysts for avoid-
ing dust plugging.
Dust plugging also may be prevented by using a suitably—shaped low oxida-
tion catalyst along with an adequate gas velocity.
4.10.4 Costs of SCR for Coal—Fired Boiler Arnlications in the U .S .
Tables 4—26 and 4—27 show the estimated investment and annual costs of
SQ applied to a 500—MW coal—fired boiler in the U.S. for 90% removal using
Ritachi Zosen’s catalyst. The total investment cost equals 51/kW and is
close to the 10,760 yen for 43/kW shown in Table 4—21 for 90% NO 1 removal
with 600 ppm inlet NO 1 , although the basis of the cost calculation may be a
little different. The total annual cost for a 7000—hour S R operation period
at the 500MW plant is estimated at 11,l60,000. Assuming an average 90% load
during operation hours, the annualized cost accounts for 3.54 mil/kWhr , which
is slightly less than the cost shown in Table 4—21.
Tables 4—26, 4—27 and 4—21 show a variation in the estimated cost of the
catalyst. Tables 4—26 and 4—27 assume the Initial charge of the catalyst to
be 10.5O/kW combined with a catalyst life of 1 year. Table 4—21 assumes the
initial charge of the catalyst to be 5230 yen ( 20.9O)IkW with a catalyst life
of 2 years, yielding an annualized catalyst cost of l0.50/kW.
In Japan, the life of an SCR catalyst used with a coal—fired boiler is
usually guaranteed for one year although in practice it may last for 2 years.
Only one catalyst has been used for 2 years (at Shimonoseki) and it operated
at 50% removal. With regeneration, Hitachi Zosen’s catalyst may be used for 2
years which would mean a lower annualized cost. On the other hand, a larger
355
-------
— DUST
(A) WAVE TYPE
DUST
ilL
(C) SQUARE TYPE
r DUST
(B) TRIANGLE TYPE
(D) HEXAGONAL TYPE
(E) PARALLEL PLATE
(F) MODIFIED PLATE
FIgure 4.44 Parallel Flow Catalysts and their Dust Clogging Tendencies
356
-------
TABLE 4—26. ESTIMATED CAPITAL INVESTMENT FOR A 500 MW APPLICATION OF TEE
IIITACUI ZOSEN PROCESS IN TEE . 8 .a
%
of
total
direct
Investment, investment
Direct investmentb
NE3 storage and injection
Reactor section
Gas handling
Air preheater modifications
Sub—total direct investment (DI)
Services, utilities (0.06 x DI)
Total direct invesmtent (TDI)
Indirect Investment
Engineering design and supervision
Architect and engineering contractor
Construction expense
= 0.25 (TDI x 10_6)0 3
Contractor fees = 0.096 (TDI x 106)0.76
Total indirect investment (IDI)
Contingency = 0.2 (TDI + IDI)
Total fixed investment (TFI)
2.2
0.5
Allowance for startup and modifications
= (0.1) (TFI)
Interest during construction
= (0.12) (TFI)
Total depreciable investment
Land
Working capital
Royalty fee
2,040,000 14.9
2,447 ,000 _____
24,882 ,000
5,000
270,000
300,000 _____
25,657,000 186.9
aBasis: 500 ? lW new coal—fired power plant, 3.5% sulfur coal, 90% NO
removal. Midwest plant location. Represents project beginning mid—1977,
ending mid—1980. Average basis for scaling, mid—1979. Investment require-
ments for fly ash disposal excluded. Construction labor shortages with
overtime pay incentive not considered.
bEach item of direct investment includes total equipment costs plus installa-
tion labor, and material costs for electrical, piping, ductwork, foundations,
structural, instrumentation, insulation, and site preparation.
692,000
5.0
10,247,000
74.7
472,000
3.4
1,534,000
11.2
12,945,000
94.3
777.000
5.7
13,722,000
100.0
298,000
75,000
2,198,000
16.0
703,000
5.1
3,274,000
23.8
3,399,000
24.8
20,395,000
148.6
Other Capital Charges
TOTAL CAPITAL INVESTMENT
17.8
181.3
3.4
2.2
357
-------
TABI L 4-27. LSTJ MAlI 0 AVLRA(.E P.NNIJAL kEV1 4UIi I 1LQUII (flILNTS loll A 500 liii APPLICATiON OF
11IL III1ACIII 20 51 )4 PI1OCI;SS IN 11111 U.S.C (9)
Annual
Unit
Annual
%
of annual
Item
Quantity
Cost ( )
Cost ( )
revenue required
I)ircct Costs
Kuw mntcrIols
NIl 3
5.28.10 kg
0.165/kg
871.000
7.8
Cutalyst
5,250.000
iota! row materials
6,121,000
54.8
Conversion costs
Operating labor and supervison
8.760
labor bra,
12.50/
labor lirs.
110,000
1.0
Uti ii ties
Steam
85,800 (LI
1.90/CJ
163.000
1.5
Ilectricity
18,226,000 1 1 t h
0.029/kWh
529,000
4.8
heat credit
125.200 Gi
—1.90/CJ
(214.000)
(1.9)
Maintenance 0.04 i Till
549,000
4.9
Analysos
2.920
labor bra,
17.00/
labor hrs
50.000
0.4
‘total conversion costs
j ,I87 .000
‘lutal direct costs
7.308.000
65.5
Indirect Costs
Capital charges
Depreciation (0.06) (total
1,493,000
13.3
depreciable investment)
Average cost of capital = (0.086) s
2,207.000
19.8
(total capital investment)
0vcrh ads
Plant = (0.5) (conversion coats
355,000
3.2
minus utilities)
Adminsitrativa = (0.1)
11.000
0.1
(operating labor costa)
‘lotol indirect costa
4.066,000
36.4
Spent catalyst disposal
(214,000)
Gross average revenue requirement
11.160. QQ
100,0
ThIAL ANNUAL tL ’V1)’ 1UE REQUIREMENTS
11,160,000
100.0
a
500 MW new coal—fired power plant. 3.5% S cool. 90% NO 1 reduction, 90% SO 5 removal. Midwest
power plant location. 1980 revenue requirements. Remaining life of power plant 30 years.
Plant on line 7000 hr/yr. Plant bent rate equals 9.5 MG/kWh. Investment and revenue require-
ment for disposal of fly ash excluded. Total direct investment l3,722,00O; total depreciable
investment 24.882,0OO. and tot.! capital invostoent 24 ,651.OO0.
( . )
-------
amount of catalyst than that upon which these figures are based may be needed
to reduce unreacted NH3 to a level of 5 ppm or below.
In the U.S., the reduction of unreacted NH 3 may be important for coal—
fired boilers. This is because it is necessary to maintain an appropriate
ammonia content in the large amounts of fly ash and sludge produced by the
lime/limestone FGD process. Based on this assumption, catalyst costs in the
U.S. may be similar to those listed in Table 4—21.
359
-------
REFERENCES
1. Wiener, R., P. Winkler, and S. Tanaka. The Hitachi Zosen NO 1 Removal
Process Applied to Coal—Fired Boilers. In proceedings of the Joint
Symposium on Stationary Combustion NO x Control, Vol. II, IERL—RTP—1084,
U.S. Environmental Protection Agency, Research Triangle Park, NC,
October, 1980.
2. Narita, T., H. Kuroda, Y. Arkawa, and F. Nakajima. Babcock—Hitachi NO 1
Removal Process for Flue Gas from Coal—Fired Boilers. In Proceedings of
the Joint Symposium on Stationary Combustion NO 1 Control, Vol. II, IERL—
RTP—1084, U.S. Environmental Protection Agency, Research Triangle Park,
NC, October, 1980.
3. Jones, G.D., and J.D. Mobley. Selective Catalytic Reduction and NO 1
Control in Japan, EPA—600/7—81—030, U.S. Environmental Protection Agency,
Research Triangle Park, NC, March, 1981.
4. Sengoku, T., and 3.11. Howell, et al. The Development of a Catalytic
NO 1 Reduction System for Coal—Fired Boilers. In Proceedings of the
Joint Symposium on Stationary Combustion NO 1 Control, Vol. Il, IERL—RTP—
1084, US. Environmental Protection Agency, Research Triangle Park, NC,
October, 1980.
5. Tokyo Electric Power Company NO 1 Control Technology and Its Application
for Power Plants. U.S.—Japan NO 1 Information Exchange Conference,
Tokyo, May 1981.
6. Ando,I. SOa and NO 1 Abatement for Coal—Fired Boilers in Japan. In
Proceedings: Symposium on Flue Gas Desulfurization—Houston, October,
1980; Vol. I, EPA 600/9—81—019a, U.S. Environmental Protection Agency,
Research Triangle Park, NC, April, 1981.
7. Nakabayashi, Y., 11. Yngami, and K. Mouri. Development of Flue Gas
Treatment in Japan. In Proceedings of the Joint Symposium on Stationary
Combustion NO 1 Control, Vol. II, IERL—RTP—1084, U.S. Environmental
Protection Agency, Research Triangle Park, NC, October, 1980.
8. Kawasaki Heavy Industries. The Operation Results of the Dry—Type NO 1
Removal Pilot Plant for the Flue Gas from the Coal—Fired Boiler. U.S.
Japan NO 1 Information Exchange Conference. Tokyo, Japan, May, 1981.
9. Mobley, J.D., and J.M. Burke.
Flue Gas Treatment Technology.
Conference, Tokyo, Japan, May,
EPA’s Pilot Plant Evaluations of NO
U. S.—Japan NO 1 Information Exchange
1981.
360
-------
SECTION 5
SCR FOR INDUSTRIAL GAS SOURCES
5 .1 GENERJ L DESCRIPTION
5.1.1 SCR Units for Industrial Gas Sources
SIR units for industrial gas sources with a capacity greater than 10,000
Nm 3 /hr are listed in Tables 5—1 through 5—3. Of a total of 62 units, 58 are
commercial and four are test units (Nos. 33, 35, 65, and 72).
Most of the SCR units were constructed between 1975 and 1978, when the
ambient NOs standard was extremely stringent——0.02 ppm as a daily average.
This standard is equivalent to 0.01 ppm as a yearly average and is less than
half of the 0.05 ppm equivalent U.S. daily average standard. In order to
achieve this stringent standard, local governments reduced NO as much as
possible.
Few S R units for industrial gas sources have been constructed since 1979
partly because the ambient standard was relaxed in 1978 to 0.04—0.06 ppm as a
daily average, and partly because of the economic recession which slowed the
construction of new factories requiring NO flue gas trea ent.
Most of the SCR units use granular (or pellet) catalysts for the follow-
ing reasons: (1) many units treat clean flue gas; (2) parallel flow catalysts
were not well developed before 1978; and (3) some of the units use granular
catalysts in a moving bed. Moving bed systems can remove 7O—80 of the dust
361
-------
a
b 0 ’ Oil
Passage Reactor
fle y Comb
Granular
tubul ar
0 ’
I’-)
IMIIE 5-1. SCI1
PLANTS FOR
OIL. PEIROCIIEMICAL, AND
GAS COMPANIES (Gas
Flow Greater than 10,000 Nm’/hr)
No.
User
Plant Site
Gaa Source
Fuel
Capacity
(1.000 Nni’/hr)
New or
Retrofit
Vendor
Catalyst
Start Up
I
2
3
4
5
Oil (ompnnies
Fuji Oil
luji Oil
Idcmitsu Kosan
Kansai Oil
Kashimi Oil
Sodegaura
Sodcgaura
Anegasaki
Sakal
IC a shi ma
CO Boiler
11011cr
Furnacc
Hailer
Furnace
IIO
l O
iCC gas
HO
Offgas
70
200
350
150
58
R
N
Il
H
R
JGC
Mitsubishi H.!.
hitachi Zosen
Hitachi Zoscn
JGC
lJ
G
U
PP
1976
1978
1975
1979
1915
6
7
8
9
10
11
12
13
I’c t roch.icalCop g jgj
Chiba Pet. Ch m.
ldcnitiu Pet. Chi..m.
Mitsui Pet. ChLm.
Mitsubishi Pet. Chin.
Nihon Pet. Chuim.
Osaka i’ t. (hum.
Shindaikyowa let. Chcm.
Ukishiina Pit. (hem.
Chiba
Anogasaki
Chsbo
Yokkaichi
Chiba
Sakai
Yokkaiclii
Chaba
Boiler
Furnace
Boiler
Boiler
floihcr
Furnace
Boiler
Boiler
I tO
LI t
FCC Gaa
110
110
Offgas
110
110
10
300
240
150
100
91
440
260
H
It
Il
H
ft
H
K
N
Ube Kosau
Mitsui loatsu
lhltsua Engineering
hitachi Ltd.
Mitsubishi Kakoki
Mitsui Toatsu
hitachi Zosen
Mittui Engineering
U
G
G
G
T°
G
C
T
1975
1977
1975
1976
1976
1976
1975
1918
14
15.16
17.18
19
20
21.22
23
Gas Companies
Osaka Gas
Osaka Gas
loho Gas
loho (.ss
loho Gis
hobo Gas
lolo Gas
Senpoku
Senpoku
Cliiba
Chaba
Soromi
Soraml
Sorami
Generator
1 1011cr
Generator
Boiler
Reformer
Reformer
Holler
Naphtha
Gas
Niphtha
Kerosene
Naphtha
Napbths
Naphtha
64
30 x 2
42 a 2
30
19
31 a 2
23
K
R
R
Ii
R
It
K
Mitsubishi I I. !.
Mitsubishi 11.1.
Mitsubishi Kakoki
Mitsubishi Kakoki
Mitsubishi Kakoks
Sumitomo Cbem. Fng.
Sumitomo Chem. Fog.
G
G
G
C
G
C
G
1975
1976
1977
1977
1976
1977
1977
-------
lADlE 5—2. SCM I’J AN IS FOIl siri . ASI) METAL INDUSTRIES (Larger than 10.000 Nni’/hr)
Capacity New or
Fo. user Plant Site Gas SourLo Fuel (1.000 Rm 1 /hr) Retrofit Vendor Catalyst Start Up
5tccl l rodiici.r
31 Iiiwasaki SticI Cliibu Coke Ovi.ii 500 N hitachi Ltd. 1976
32 Eawasaki SICLI Chiaba Sintering 762 N Hitachi Zosen 1977
Macli inc
33 kobe St i Amagasaki Coke Oven 104 R k obe Steel C II 1971
34 Nippon lol tin Ogishima Sintering 1.320 It Nippon Kokan G?IIC 1979
Macli inc
35 Nippon St .l Kimizu Coke Ovi.n d 133 N JUC LI’ 1978
36 Nisshin Sti. 1 /unagasaki Boiler (110) 19 It Hitachi Ltd. GIN 1977
37 (oshin Stecl hlimej I Furnace (110) 71 R Hitachi Zosen C 1976
H tal Iiro,luccr
38 Nihon Satetsu Katsunia Furnace (HO) 10 It lfltachi Zosen C 1976
39 llippoii Yakin Kawosiki Boiler (110) 15 It Mitsubishi Kakoki RC 1976
40 St.itctsu kagaku Iloiler Iloiler (110) 15 It Seitetsu Kngnku It 1975
Grinu1ar oiaIyst. moving bed.
Granular aiii1yst (fixed bed).
Graiitilar Iron ore is ustd as catalyst in moving bed.
IIi.ivy oil
Replaced in 1977 with ring—tube catalyst stacked regularly to achieve parallel flow.
-------
Sat PIANfS 10R OilIER INOUS1BI ES (L. Ib ..r than 10.0(1(1
lLuvy oil.
ight oil.
a cl i .
(,inouli. in moving bed
lb i ii. y omb
lADLE 5-3.
Capacity
No. thor Plant Site Gas Source Fuel (1.000 Nm’Ihr)
New or
Retrofit Vendor Catalyst Start Up
Nm ’ /hr)
51.52 Aito
Chiba
Boiler
LNG
30 x 2
N
Mitsubishi II
II.!.
1978
1978
53 Ajinomoto
54 Asahi Glass
Kawa aki
Keihiu
Boiler
Furnace
H0
110
180
75
50
8
R
N
Asahi Glass
Kawesaki 11.1.
d
G Il
(3
1976
1977
55 Atomic Power Res.
Boiler
10
I I
JGC
(111
1976
56 C telyst Iliem.
Wakamatsu
Purnace
110
15
H
Babcock Hitachi
6
1976
57 Chiyoda Keuzat
Kalauka
Boiler
19
8
Lbnra
6
1977
58 Lliiyoda Kt.nzai
Atsugi
Boiler
lID
L IJ
30
N
ilitsubashl 11.1.
( 3
1978
59 l)ainippon Ink
Sakal
Boiler
10
N
hitachi Zosi.ia
G
1974
60 Green Cheuscal
Kasbima
Furnace
200
N
Sumitomo Clicm Lug.
G
1974
61 II N. Ilcihanol
Sodegauri
Ref urner
Ll
1.024
N
Babcock Ihituchi
6 1 1
1981
62 J. Nutional Railway
Knwasakl
Gas lurbinc
16
I i
Bibeock hlltdi .hli
G
1978
63 Kansai lIant
Amagasaki
Boiler
Kerosene
25
8
Kawas ki 11.1.
6
1978
64 Kswssaki II.!.
Akashi
Boiler
30
R
Kurabo
GM
1975
65 Kurabo
llsraksta
Boiler
11.0.
170
8
Mitsui loaL u ,
(3
1976
j .
66.67 Mitsui Toatsu
68 Nihon Ammonia
Osaka
Sodegaura
Reformer
Reformer
Butbanc
LfG
&
290
20
N
N
Sumitomo CIii.m Eng.
Uobcock HiLu..hi
6
(3
1975
1918
69 Nikon Oil & Fat
Amagasaki
Boiler
Butbane
323
N
Ihitsul loatsu
(3
1976
70 N N. ilcthnnol
Sodegaura
Reformer
Ll ’G
20
N
Mitsui Toaisu
6
1980
71 Si.kisui Chem.
Sakai
Boiler
30
Ii
SumitomO (hun. hug.
6
1973
72 Sumitomo Chum.
Sodegaura
Boiler
110
LPG
200
8
SumitomO Clii.m bug.
G
1975
73 Sumitomo (‘hem.
Niihaeua
Reformer
100 200
Ii
Suis i Lomo Clii.m Lug.
6
1975
74.75 Sumitomo Cbuu.
Ahega sabi
Boiler
LFG
250 & 350
K
Sumitomo (‘ki.mii. Lug.
G
1976
76.77 Sumitomo Chs.m.
78 lokyo Metrupolai.e
79 lini tika
Sodegaura
Sunanmichi
Uj i
Boiler
Incinerator
Diesel Engine
Sluggo
10
100
70
N
N
Mitsui Eng.
Unit ik - - -
11 e
( 3
1979
1918
-------
in the flue gas along with 80—90% of NOx and are suitable for certain gas
sources which do not have ESPs.
There are 8 industrial S R units which use parallel flow reactors. Nos.
1, 5. and 35 use a parallel passage reactor, Nos. 2 and 78 use a honeycomb
catalyst, and Nos. 10, 13, and 39 use tube or ring—tube catalysts. Since
parallel flow catalysts provide simple reliable operation, they are generally
preferred to the moving bed reactor with a granular catalyst except for cases
in which dust removal by the bed is important. A moving bed reactor may be
useful for very dirty gases such as flue gas from a glass melting furnace
which contains alkaline vapor or for low—temperature gas containing S0 ,
which tends to deposit sodium sulfate or ammonium bisulfate on the parallel
flow catalysts (Nos. 54 and 33).
Most of the SCR units have been operated without problems. However,
several units have been shut down or operated intermittently due to the 1978
relaxation of the NOz regulation and the downturn in the economy which reduced
N0 emissions made continuous operation of the SQ system unnecessary.
Unit Nos. 1, 3, 5, 76, and 77 are the only major units which have not
been in use for over one year. These five units were constructed in 1975 and
1976. at an early stage in the development of SCR. Unit No. 12 has been
operated intermittently. Unit Nos. 1, 3, and 12 require fuel for gas heating,
and therefore have high operating costs.
Some of the catalysts produced before 1976 either were not resistant to
SOx or experienced dust plugging problems; most of the catalysts produced in
1976 or later had few poisoning or plugging problems. For example, the Fez0i—
based ring catalyst charged in 1976 for Unit No. 39 was not satisfactory and
was replaced in 1977 with a TiOi—based ring—tube catalyst. The latter cata—
lyst has treated flue gas from an industrial boiler burning heavy oil (2.5%
sulfur) for over 4 years without replacement (Section 5.3). The honeycomb
catalyst of Unit No. 2 has been used for over 4 years without renewal (Section
5.2)
365
-------
Most of the units, particularly the smaller ones, remove 90% or more of
NO 1 using 1.0 — 1.2 mol NHi to 1 mol NO 1 . Although unreacted NH3 reaches
10 ppm in these units, an ammonium bisulfate problem is seldom encountered be-
cause the flue gas is usually discharged above 2500C. An ammonium bisulfate
problem did occur in a few larger plants including Unit No. 12 (Shindaikyowa),
which cools the gas below 250°C for better heat recovery. Units No. 2 and 12
recently used 0.6 — 0.7 mol N113 to 1 mol NO 1 to remove 60—70% of NO 1 .
These units leaked Nib below 1 ppm and did not have a bisulfate problem.
5.1.2 Economics
The investment costs of seven S R units are shown in Table 5—4. Costs
are given for the SCR reactor, the initial charge of catalyst, connecting
ducts, ammonia storage and injection facilities, NO 1 and Nib analyzers, and
control systems (unless otherwise noted). Some of the unit costs include
costs for the heat exchanger, heater, and fans for gas heating purposes.
In general, SCR units were expensive during the early stages of develop-
ment in 1975 and 1976. The costs were substantially lowered in 1977 and 1978
due to technology developments and have increased slightly since 1979 due to
inflation. The individual cost of Unit Nos. 1, 12, 39, and 65, which were
constructed in 1975 and 1976 exceeds 13,000 yen/kWhr although these units are
for existing boilers and include heating or test facilities. By contrast, the
two units constructed in 1978 for new boilers (Nos. 2 and 13) cost only 1,230
and 2,640 yen/kWh, respectively. Unit No. 78, constructed in 1979, is fairly
costly—- 14,500 yen/kW—because it was designed to remove both NO 1 and
odor from an existing incinerator and has a heat exchanger, heater, and fans.
As opposed to SCR units for utility boilers, units for industrial uses
treat various kinds and amounts of gas under a variety of conditions. There—
fore, although the operating costs of S R units for large new industrial
boilers may be close to those for new utility boilers, the investment costs
vary widely.
366
-------
bfhcU ate from ga volume (300
Parallel I’assage Reactor.
Inc1udIng heater.
honeycomb.
Ior catalyst and reactor.
Including fan, heat exchanger and heater.
lIoving bcd reactor.
1 Including t..et facilities.
Removes both NO 5 and offensive odor.
0 ’
TArn I 5-4,
INVESTMENT
COSTS OF
SEVEN INDUSTRIAL SCR
UNITS
No,
Owner
Gas Source
Capacity
(MW)a
NO
Rcmo al
( )
New or
Retrofit
Reactor
(Catalyst)
5V 1
(hr )
Cost
Year
(10’ yen)
(yen/kIt)
Completed
I
2
12
luji Oil
liiJi Oil
Shindaskyowa
11011cr
BoIler
BoIler
23
53
150
95
80
90
ft
N
ft
I I
Granule
4200
5,220
10.000
460
65
2.000
20870 C
1 .230
13.300
1976
1978
1975
13
39
65
Uklshiama
Nippon Ynkin
Kurabo
BoIler
Boiler
BoIler
87
5
10
85
90
90
N
ft
R
Tube
Ring Tube
NI l 8
3.900
10,000
200
66
130
2.64O
13 ’ 200 h
13,000
1978
1976
1975
78
Tokyo Metro.
IncInerator
33
9O
I l
II
4.000
480
ims ,s o
1979
Ne m’/hr = 1 MW).
-------
The operating costs for 8 units in 1979 are shown in Table 5—5. The
costs range from 1,400 — 2,000 yen/ki oil or 0.33 — 0.46 yen/kWhr for the four
units which do not require gas heating, and from 2,200 — 4,000 yen/ki or 0.49 —
0.93 yen/kWhr for the four units requiring gas heating. Unit No. 1 is the
most expensive and has been shut down since 1977.
In 1981, the operation costs for SCR units may be slightly higher than
those shown in Table 5—5 due to inflation. However, the application of S R
can be less expensive than the use of higher quality fuel for NO 1 abatement,
as will be discussed in Section 5.3.
5.2 SODEGAURA REFINERY, FUJI OIL
5.2.1 Introduction
Fuji Oil’s Sodegaura Refinery is located in the eiyo Industrial Region
near Cliiba City and Tokyo. This area is the largest industrial region where
stringent N0 and SOx regulations have been applied by the local govern-
ment. The Sodegaura Refinery now has five boilers as shown in Table 5—6. The
total boiler capacity was almost doubled between 1974 and 1977. During the
same time period the permissible total NO 1 emissions from the refinery were
cut almost in half, requiring extensive NO 1 abatement by combustion modifi-
cation and SCR (Tables 5—6 through 5—9).
5.2.2 SCR Unit for CO Boiler
The flow sheet of the CO boiler and the SQ unit at the Sodegaura Refin-
ery is shown in Figure 5—1. The SCR unit was constructed for the existing CO
boiler by JGC Corporation using JGC ’s catalyst and the Shell parallel passage
reactor. Since it was difficult to install the SCR reactor between the boiler
and the economizer, an inline heater first was installed to heat the 250_3000C
economizer outlet gas to 360_4000C. The SCR reactor then was installed down-
stream of the heater. The S R unit was completed in 1976 and was operated for
about 6 months with an NO 1 removal rate of 92—96% (Table 5—9). The pres-
sure drop through the SCR reactor was 100—120 mm IhO.
368
-------
TABLE 5—5. OPERATiNG COSTS OF SCR UNITS IN 1979 (Heavy Oil—Fired Boiler Including 7 Years
Depreciation)
NO
No
Owner
Capacity
(MW)
Reactor
(Catalyst)
Remo al
(9’o)
Cost
Reniarks
(yeii/kW)
(yen/kwhr)
1
2
Fuji Oil
Fuji Oil
23
53
pga
II
95
70
4,000
1.400
0.93
0.32
With gas
heating
13
Ukishima
87
Tube
85
1,500
0.33
36
Nisshin Steel
6.5
MBC
90
2.800
0.65
Ylith gas
heating
39
Nippon Yakin
5.0
.
Ring Tube
90
2.200
0.49
With gas
heating
63
Kansai Paint
5.3
Ring
90
3,600
0.84
With gas
heating
65
Kurabo
10
M D
90
2.000
0.46
69
Nihon Oil & Fats
6.7
MB
90
1,800
0.41
1’arallel passage reactor.
neyc0m
Moving bed.
-------
TABLE 5—6. BOILERS AND NO 1 ABATEMENT AT SODEGAURA REFINERY (1)
Boiler
Fuel
Steam Rate
(ton/hr)
NO 1
Abatement Methods
No.
No.
No.
No.
CO
4
5
6
7
Boiler
.
Oil
Oil
Oil
Oil
CO,
Oil
75
160
160
160—200
45
a
LNB
LNB, TSCL?,
LNB, TSC.
LNB, TSC,
SCR
FGRC
FOR
SCR
aL
ow—NO 1
burner.
bTwo_stage
combustion. CF1ue
gas recirculation.
TABLE 57.
PERMISSIBLE EMISSIONS
REFINERY
OF
NOx AND
SO 1
FROM THE
1974
1975
1976
Since 1977
NO 1 , Nm /hr
157
124
107
92
SO 1 , Nm /hr
375
163
145
134
TABLE 5—8.
VOLtThIE AND
SODEGAURA
COMPOSITION
REFINERY
OF
FLUE
GAS
ID
BE TREATED
BY
SCR AT THE
Boiler
Fuel
Gas Volume
(Nm 3 Ihr)
NW
Equivalent
NO 1
(ppm)
SO 1
(ppm)
Dust
(mg/Nm 3 )
CO
Boiler
CO
+
Oil
70,000
23
160—200
400600
60-70
No.
7 Boiler
Oil
200,000
67
110—120
200
20
TABLE 5—9.
SCR UNITS
AT THE
SODEGAURA
REFINERY
Boiler
Year
Vendor Completed
SV
(hr_i)
Inlet
NO (ppm)
NB 3 /NO
Ratio 1
Outlet
NO (ppm)
Removal
( )
CO Boiler
JGCa 1976
4,000
180
1.1—1.2
8—15
92—96
No. 7
ffl 1 b 1977
6,400
115
0.5—0.8
57—23
50—80
IGC Corporation (using parallel passage reactor).
Mitsubishi Heavy Industries (using honeycomb catalyst).
Leak
NB 3 (ppm)
10—20
<1
370
-------
TO HEAT
RECOVERY
L )
IN LINE
BURNER
Figure 5-1 Fiowsheet of CO Boiler and SCR Unit (Fuji Oil)
-------
Operation of the SCR unit was expensive (Table 5—5) and not energy—
efficient because of the fuel requirement for gas heating as well as the high
pressure drop. The S R unit was shut down after 6 months of operation, when
the operational load of the refinery dropped to 7O of capacity. At that
time, total NO emissions were lowered to meet the standard without operat-
ing the unit.
5.2.3 SCR Unit for Oil—Fired Boiler
A new oil—fired boiler SIR unit was constructed in 1978 by MRI. The unit
uses a NGK honeycomb catalyst and has a capacity of treating 200,000 Nm 3 /hr of
flue gas containing about 120 ppm NO 1 , 200 ppm SO 1 , and 20 mg/Nm 3 of dust
(Figure 5—2). The reactor exists in ductwork extended between the boiler and
economizer and treats gas at temperatures of 330_3700C.
This ScR unit can remove about 9O of NO 1 using 0.9 mol Nib to 1 mol
NO 1 . In order to eliminate the ammonium bisulfate problem with the econo-
mizer, the unit was operated from 1978 to 1980 with 0.7 mol Nib to 1 mol
NO 1 . This resulted in a NO 1 reduction of nearly 70 ’o and less than 1 ppm
unreacted Nib. Since 1980 it has operated with 0.5 mol Nib to 1 mol NO 1 and
obtained 50 NO 1 removal with less than 1 ppm unreacted Nib. A small amount
of ammonium bisulfate still deposits on the tubes of the economizer, but the
economizer can be operated without problems for more than a year at a time and
is washed with water during the annual maintenance shutdown period.
After unit startup. a heavy oil additive was used for possible boiler
operation improvement, and as a result a slight dust deposit was observed in
the catalyst bed. The dust deposit has since been eliminated by not using the
additive and applying soot blowing to the catalyst bed.
Catalytic activity tests indicated that the use of 1 mol Nfl3 to 1 mol
NO 1 resulted in 93.5 NO 1 removal with 5 ppm of unreacted Nib and less
than 1 SOs onidation during the first 2 years of operation. An 89.5! NO 1
372
-------
TO
STACK
OIL
ECONOMIZER
‘NH 3
ESP
FAN
Figure 5.2 Flowsheet of Oll .Fired oiIer and SCR Unit. (Fuji Oil)
373
-------
removal rate with 10 ppm unreacted Nib and about 5 S02 oxidation was reported
after 2 years of operation. After 3—1/2 years of operation, the NOx removal
efficiency was 86.4% and the catalyst was still active.
5.2.4 Evaluation
The StR unit for the CO boiler is one of the oldest in Japan and was
designed at an early stage in the development of S R technology when fuel was
still inexpensive. Today, if an SIR unit were installed on a similar boiler,
the reactor might be placed between the boiler and the economizer to treat the
flue gas at 350—450°C, or a low—temperature catalyst might be used for the gas
leaving the economizer at 250_3000C in order to save energy.
The SIR unit for the No. 7 oil—fired boiler is one of the oldest units
using a honeycomb catalyst. Both investment and operation costs are reason-
able (Tables 5—4 and 5—5). Although the catalyst shows a slight decrease in
activity and an increase in the SOa oxidation ratio with time, it has been suc-
cessfully used for over 4 years. The increase in the SO 1 oxidation ratio may
be caused by the deposition of flue gas vanadium on the catalyst. Improve-
ments in ScR catalysts since 1978 may result in better catalysts for this type
of application.
5.3 KAWASAKI PLANT. NIPPON YAKIN
5 .3 .1 Introduction
Nippon Yakin is an alloy producer with a stainless steel factory in
Kawasaki City adjacent to Tokyo. This area was very polluted about ten years
ago and has since set stringent regulations for pollution control. The SO
and NOx concentrations in the factory’s boiler flue gas are transr itted to
the City Pollution Control Center.
The factory has 5 boilers, 2 of which use high—sulfur heavy oil
(S2.5%), while the rest use gas. Each of the 2 oil—fired boilers has a capa-
city of producing 6 tons of steam per hour with a total gas flow rate of about
374
-------
15,000 Nm 3 /hr (5 MW). According to the regulation, NO 1 concentrations in
the oil—fired flue gas must be kept below 60 ppm. Since the flue gas leaving
the boiler contains about 250 ppm NO 1 , 1,000 ppm SO 1 , and about 100 mg/Nm 3
of dust, both SQ and FGD have been applied. To be safe, NO 1 is usually
reduced to about 20 ppm by ScR, while SO x is reduced to 20—25 ppm by sodium
scrubbing.
The other three boilers use LPG gas and maintain NO 1 concentrations
below 50 ppm by combustion modification.
5.3.2 Process Descrii,tion
A flowsheet of the combined ScR/FGD system used in the Kawasaki Plant is
shown in Figure 5—3. Flue gas leaving the boiler at 280—300°C is heated to
330°C by a Rothemuehie heat exchanger and then to 370°C by oil—firing. After
that, it is injected with ammonia, and then sent to an SQ reactor. The reac-
tor and the heat exchanger were constructed by Mitsubishi Kakoki Kaisha CMXX)
in 1976. The Sarc catalyst, a large ring—shaped catalyst based on Fe103 and
was used with a random packing during the first 6 months of operation.
This resulted in a considerable increase in the pressure drop due to dust
plugging, and a large decrease in catalytic activity. After 6 months the
catalyst was replaced by an improved Sarc catalyst which was used for an addi-
tional 6 months without success.
In September 1977, the reactor was modified and a rashig ring type cata-
lyst made by Catalyst and Chemicals, Inc., was charged. This catalyst has
inner and outer diameters of 25 and 35 mm, a height of 40 mm, and is made of
TiOz and ViOs. The catalyst was piled regularly to form tubes, inside and
outside of which the gas passes in a parallel flow.
The new reactor went into operation in September 1977, and has operated
for over four years without problems and without catalyst replacement. The
_1
volume of the catalyst is 4 m 3 , and the SV is 3,750 hr . By using 1.2 mol
3 to 1 mol NO , 92% of the NO is removed and unreacted NIh is kept arou.nd
I I
375
-------
FUEL OIL
(4 S OkIlmontt)
SO. 20 ppm
NO, 20 ppm
GAS 15,000 Nm 3 !hr
so : 1,000 ppm
NO, 250 ppm
FUEL OIL
‘15k1/month)
GAS GAS
I - , )
0 ’
OPEN
BY PASS
‘C
HEAT
NH 3
( SO Okg!month)
FIgure 5 3 Flue Gas Treatment System at the Kawasakl Plant, Nippon Yakin
-------
10 ppm. The pressure drop in the reactor was 210 mm HzO at startup and in-
creased to about 270 mm HiO after 4 years of operation. The NO 1 removal
efficiency has remained over 90% during the 5 year period.
Flue gas leaving the StIR reactor is passed through the heat exchanger
where it is cooled to 330°C. Although the gas contains considerable amounts
of ? W3 and SOS, ammonium bisulfate does not deposit in the heat exchanger
because of the high flue gas temperature. The gas is then cooled to 280°C by
a heat recovery system and subjected to FGD using a spray of sodium hydroxide
solution to reduce SO to about 20 ppm. The liquor discharged from the
scrubber contains sodium sulfite, sulfate, and a small amount of ammonia and
is sent to a water trea ent system and treated with other wastewaters from
the factory.
The construction of the SCR and FGD plants required 4 months. Five days
were needed to connect the plants to the boiler.
The boilers undergo a maintenance shutdown once a year. When one of the
oil—fired boilers is down for maintenance, the other is in operation. There-
fore, the ScR—FGD system has operated continuously for 5 years without total
shutdown.
5.3.3 Economics
In 1976, the Kawasaki plant’s S R system (Section 3.6) cost 73 million
yen including the cost of the heat exchanger and the heater, while the FGD
system cost 63 million yen. The catalyst produced by Catalyst and Chemicals,
Inc. cost 6 million yen (1.5 million yen/mi) in 1977. The operating cost of
the S R and FGD systems was 30 million yen in 1979, including about 16 million
yen for sodium hydroxide and 5 million yen for gas heating fuel. The 1979
annual costs are summarized below.
377
-------
Annual cost (million yen)
Capital cost (25 of investment) 34.0
Catalyst (4 year life) 1.5
Others 30.0
Total 65.5
Annual Fuel Oil Consumption of the Boilers 5,400 ki
Annualized cost 12,130 yen/ki oil
In 1979, heavy oil cost 33,000 yen/hi- while kerosene with the equivalent
heating value cost 50,000 yen. Since the cost difference was 17,000 yen/hi,
the use of heavy oil with the SCR—FGD system meant a 4,778 yen/ki saving. In
1981, heavy oil cost 50,000 yen/ki and kerosene with the equivalent heating
value cost 70,000 yen while the annualized cost of the flue gas treatment
system was estimated at 12,700 yen/ki. Therefore, the use of this system
resulted in substantial savings.
5.3.4 Evaluation
More than 4 years of continuous catalyst use may be the longest record
for an S R catalyst used to treat high sulfur oil flue gas. The smooth opera-
tion of Nippon Yakin’s awasaki plant is a good example of the reliability of
an SQ system for industrial boilers.
The catalyst used at the Kawasaki Plant was designed by Catalyst and
Chemicals Inc. more than 6 years ago. Further improvements in SCR catalysts
made in the past 5 years may make SCR even more reliable.
The heat exchanger has had no ammonitzm bisulfate problems even though the
gas leaving the S R reactor contains a considerable amount of S03 with about
10 ppm unreacted NB3, due to high gas temperatures (3300C). The gas is subse-
quently cooled to 280°C for heat recovery. Ammonium bisulfate will deposit
when the gas is cooled to lower temperatures and reduction of unreacted Nib
will be needed for additional heat recovery.
378
-------
In the future, Nippon Yakin may replace the ring—tube catalyst with a
tubular catalyst since the tubular type can be more easily packed.
5 .4 SCR FOR CORE OVEN FLUE GAS (NIPPON STEEL )
5.4.1 Introduction
Nippon Steel, Japan’s largest steel producer, Nippon Steel Oiemical and
JGC Corp. have conducted large—scale S R tests with coke oven flue gas at
Nippon Steel’s imitsui Works Plant. The coke oven flue gas has a flow rate
of 152,000 Nm 3 /hr and contains about 400 ppm NOT, 15 ppm SOx, and less
than 35 mg/Nm 3 of dust. The gas temperature ranges from 170—200°C. Since gas
heating to 350°C for S R is expensive, a low—temperature catalyst developed by
JGC Corp. and a parallel passage reactor, originally developed by Shell, are
used.
Because ammonium bisulfate gradually deposits on the catalyst and reduces
its activity, the gas is periodically (once every several days) heated to
above 350°C. The heating takes place in an inline heater for a few hours in
order to volatilize the deposits. The SCR unit began operation in March 1978,
and has operated since that time without trouble.
5.4.2 Process and Operation (3)
A flow sheet of the Nippon Steel SCR system is shown in Figure 5—4. The
flue gas from the coke ovens is injected with ammonia, sent through two trains
of inline heaters and S R reactors, and finally into a stack.
Figure 5—5 shows examples of the change in NOx removal efficiencies
with the heating cycle during treatment of a gas containing 50 or 200 ppm
S0 . The efficiency was lowered more rapidly with gas richer in SOx due
to the deposition of larger amounts of bisulfate. After being heated at 400°C
for two hours, the catalyst activity returned to its original level.
379
-------
COKE OVEN GAS
AND AIR
N I - i 3
OPEN
BY-PASS
Figure 5.4 Flowsheet 01 an SCR System for a Coke Oven.
380
-------
S
a,
z
C
z
l A S
(.1
z
0
0
100
75
\ \ 4\ \ \
25 -4- SO 2000 cm
-0- SO 2 5oopm
C
CATALYST JP 501 AT 250C
HEATING CONDITION 400C 2 flours
GASCOMPOSITION NO 200ppm NH 1 22Oppm
0, 3% 14,0 I0•/. N 2 balance
Flgure 5.6 O er.ting Conditions arid NO 1 Removal EffIciency (3)
70*282 1
-S
>
0
“S
0
z
3 200 400 600
OPERATION PERIOO (hrl
800
FIgure 5.5 Change In NO 1 Removal Efficiency Ouflng Heating Cycle. (4)
70*2920
rypE OF TEST
Y,
NO REMOVAL
1%) 90
STANDARD
2700
200
13
AT 180C
3500
180
1 2
LOWNH 3 (180C 1
3500
180
09
LESSFREOUENTHEATING
3500
180
09
WITNOUT HEATING
3500
170
09
WITHOUT HEATING
3500
170
07
381
-------
Tests were conducted continuously for one year from April 1978 through
March 1979 uiider varied conditions, as shown in Table 5—10.
TABLE 5-10. NIPPON STEEL SCR SYSTEM TEST CONDITIONS
Temperature
SY (hr t )
(°C)
170,
2,700,
180,
200
3,500
NR3/NOx
mole
ratio
1.3,
1.2,
0.9, 0.7
The NO x removal efficiencies with different operating conditions are
shown in Figure 5—6. With the standard condition of 200°C with an SV of 2,700
hr and a mole ratio of 1.3, the NO removal efficiency was 94—98%.
The efficiency was 84—92% at 180°C with a mole ratio of 1.2 and an SV of 3,500
hr’,and 78—88% with a mole ratio of 0.9. When gas heating to 400°C took
place over longer intervals, the NO 1 removal efficiency was lowered. Tests
were conducted using 170°C flue gas without gas heating. These tests indi-
cated that the removal efficiency was 60—70% with a mole ratio of 0.9 and
about 50% with a mole ratio of 0.7.
The NO 1 removal efficiency under the standard condition was 98% at the
beginning and 97% at the end of the tests, which indicated that there was lit-
tle, if any, degradation of the catalyst during the one—year test period.
Tests also have shown that, immediately after being heated to 400°C, the cata-
lyst produced a relatively low NO 1 removal efficiency for a short time
period. For example, the NOx removal efficiency was about 80% with 1 mol
Nifli to 1 mol NO 1 at 180°C immediately after the heating, and increased to
nearly 90% in a few hours. It appears that the catalyst absorbs more ammonia
at lower temperatures than at higher temperatures. A considerable amount of
ammonia is released from the catalyst when it is heated to 4000C. After being
heated to 170—200°C, the catalyst absorbs ammonia in the flue gas until satu-
ration is reached. Saturation results in a temporary shortage of ammonia
which can react with NO 1 in the flue gas.
382
-------
The pressure drop in the reactor was 40—50 mm 1120 both at the beginning
and the end of the test period indicating that there was no dust plugging
problem.
5.4.3 Evaluation
Tests of Nippon Steel’s SI R unit were conducted successfully with over
90% NOx removal efficiency under the standard conditions. Neither catalyst
degradation nor dust plugging was observed during the one—year test period.
Although periodic gas heating to over 350°C is needed to remove ammonium
bisulfate, the use of low—temperature catalysts may be less costly than the
use of a standard catalyst with continuous gas heating to 350°C. Another
advantage of the system is the simple thermal treatment in the fixed bed.
Operation of this system is much easier than using a moving bed with a cata-
lyst that is discharged from the reactor for NOx treatment. Rowever, low—
temperature catalysts may not be suitable for SOx—rich gas because of the
requirement for frequent gas heating.
Parallel passage reactors are expensive because they use many steel gauze
envelopes to hold the catalyst. On the other hand, smaller, very active
catalysts can be used because less physical strength is needed in these
reactors. Therefore, the reactor may be suitable for a low-temperature very
active catalyst.
Although the S R system at Kimitsu works well, it does require consider-
able investment costs, land space, and energy. At this time, Nippon Steel has
no plans to construct a commercial plant using the system.
5.5 SCR USING AN IRON ORE CATALYST
5.5.1 Introduction
Nippon Kokan, one of the largest steel producers in Japan, constructed a
full—scale SCR unit at its Keihin Works plant. The unit has a capacity of
383
-------
treating 1,320,000 Nm 3 Ihr of flue gas from an iron—ore sintering machine using
iron ore as the catalyst in a moving bed reactor.
The flue gas is first subjected to dust removal by an ES? and then FGD by
ammonia scrubbing, utilizing ammonia present in the coke oven gas (5). The
F D system began operation in 1976 and has operated successfully since that
time. When further gas treatment for NOx removal was requested by the local
government, Nippon okan conducted pilot plant tests of SCR using iron ore as
the catalyst (2) . Based on these tests, the full—scale StR unit was completed
and began operation in 1979.
5.5.2 Process Descrii,tion
The flow sheet of the gas treatment system at eikin is presented in
Figure 5—7. The 150°C flue gas from the sintering machine contains about 300
ppm SOS, 200 ppm NO 1 , and 500 mg/Nm 3 of dust. It passes through an ES?,
heat exchanger, and an FGD unit to remove over 90% of the SOx. The flue gas
leaving the FGD unit is further cleaned by a wet ESP and then heated to 110°C
by a heat exchanger. An StR unit was installed downstream of the FGD system.
The gas is heated to 380—400°C by a heat exchanger and a heater, injected with
ammonia, and sent into the moving bed reactor (SV=3,500 hr’) where 80% of
the NO 1 is removed. One specific kind of iron ore was found to be particu-
larly active as a catalyst.
Since the flue gas after FGD still contains small amounts of SO 1 , a
portion of the iron ore reacts to form a powdery sulfate. The ore catalyst
discharged from the reactor is screened and returned to the reactor; the iron
ore fines containing sulfate are returned to the sintering machine.
The SGR unit cost 6.7 billion yen including the price of the heat ex—
changer and the heater.
384
-------
(-h)
u - I
EXCHANGER
ESP
1’
380 - 400
H EAT
EXCHANGER
RN ACE
FUEL
Figure 5-7 FGDISCR System for Flue Gas from iron Ore Suntering Machine
(Keihun Works. Nippon Kokan)
-------
5.5.3 Evaluation
The iron ore—catalyst S R system may be best suited to flue gas treatment
at steel works because iron ore is inexpensive and can be utilized for steel
production after it is used as a catalyst. From this standpoint, the process
is quite economical,compared with standard S R processes for which the cost of
the catalyst comprises more than half of the total investment and operation
costs.
Treatment of flue gas from a sintering machine still may be fairly
expensive because the gas temperature is low———about 150°C at the sintering
machine outlet and even lower after wet FOD treatment——thus requiring a heat
exchanger and fuel for the heater.
Use of a low—temperature catalyst might reduce the fuel requirement to
some extent. JGC Corporation, with funding from the Steel Federation, con-
ducted tests of SIR with a low—temperature catalyst for flue gas leaving a
sintering machine (2). These tests were conducted at 290°C or above to avoid
ammonium bisulfate deposition and thus required extensive gas heating.
The system at the Keihin Works may be one of the best solutions for
treating flue gas from a sintering machine to meet stringent particulate,
S0 , and NO standards.
On the other hand, where the regulations are not very stringent, a moving
bed reactor with iron ore may be the best way to treat flue gas from a sinter—
ing machine directly without an ESP. This system removes 70—80% of the dust
and portions of the SOx and NOx, and requires only minimal gas heating.
5 .6 L0W-TEMPERATUP .E CATALYST TESTS (I 0BE STEEL)
5.6.1 Introduction
Kobe Steel, one of Japan’s largest steel and heavy machinery producers,
has developed low—temperature S R catalysts for 180—250°C gases containing 50—
386
-------
250 ppm N0 , and 50—1,000 ppm SOx. These catalysts have been tested at
several pilot plants; the largest test plant has a capacity of treating
100,000 n 3 /hr of flue gas.
Kobe Steel’s low temperature catalysts are based on a—AlzO or TiOi and
are characterized by containing niobium in addition to vanadium. Both pellet
and honeycomb catalysts have been produced and tested. Pellet catalysts in a
movable bed have been used for low temperature gases containing SO. Bisul—
fates deposit on the honeycomb catalyst which means that the catalyst fre-
quently must be discharged from the reactor for thermal treatment.
Unlike other process developers, Kobe Steel has found that the alumina—
based catalyst containing Ti, V. and Nb performs well for low temperature gas.
Kobe Steel has also developed a special technique for thermally treating a
catalyst to recover its initial properties.
5.6.2 Catalysts and Plant Operation
The effect of niobium on the alumina—based TiOi—ViOs catalyst is shown in
Figure 5—8.. Niobium increases NO 1 removal efficiencies especially well at
low temperatures. Extensive pilot plant tests have indicated that the alumina—
based catalyst is superior to the titanium—based catalyst for low temperature
gases. When a titanium—based catalyst was used for a 180_2500C gas containing
— i
about 50 ppm SO (SV: 5,000 hr ; N113/NO mole ratio:1), the initial NO re—
I I I
moval efficiency of 98% dropped to about 8O within 8—10 days of continuous
operation. Although catalytic activity can be recovered when the contaminated
catalyst is heated to 400—450°C, frequent heating consumes considerable
energy.
When the alumina—based catalyst is used under the sane conditions, the
initial 90% NO 1 removal efficiency drops very slowly and requires 3,000—
4,000 hours to reach 80%. The large specific surface area of the alumina—
based catalyst——about 150 mi/gram compared with about 50 mi/gram for the
titania—based catalyst———may be a factor which explains the long period of
387
-------
30V-XN b-25TT
220°C
100
90
4000 C
300°C
250°C
z
0
C l )
w
>
z
0
0
0
z
1
2
34
AMOUNT OF Nb (gil)
(A 1 2 O 3 -based V 2 0 5 -TiO 2 catalyst)
FIgure 5.8 Effect of Niobium on NO Removal.
388
-------
catalyst usefulness. Another reason may be the low oxidation rate of S02 to
S03 caused by the niobium—containing alumina—based catalyst.
Figure 5—9 shows the thermal treatment system used to treat the contami-
nated aluminum—based catalyst. The catalyst is heated to 55O 600°C in several
minutes to avoid degradation. With rapid heating for a short time, catalytic
activity is restored to its original level without a loss in physical
strength.
Kobe Steel’s large test plant (100,000 Nm 3 /hr) has been operated for over
4 years, removing 90—80% of NO 1 in the 200—220°C flue gas which contains
about 250 ppm NO 1 , 50 ppm SO 1 (including about 5 ppm of S03), and 10—50
mg/Nm 3 of dust. Thermal treatment is applied once every 3,000—4,000 hours.
The catalyst retains a sufficiently high activity and physical strength during
this time. Both the catalyst and thermal treatment system have been patented
in several countries including the U.S.
5.6.3 Evaluation
Alumina—based catalysts for SCR were popular several years ago but they
are no longer used for S0 1 —containing gas by most process developers and
users. Instead, S0 1 —resistant TiOs—based catalysts have been developed.
ScR operator experience indicates that alumina—based catalysts are often
poisoned by SO 1 to form aluminum sulfate when used at 300—400°C. The cata—
lysts must be heated to 800°C in order to decompose the aluminum sulfate and
recover their catalytic activity.
Kobe Steel’s long—term SQ test has proved that the alumina—based cata-
lyst is best suited to a low—temperature gas with a relatively low SO 1 con-
centration. It appears that aluminum sulfate is not formed at lower tempera-
tures, because aonia in the flue gas readily combines with SOi before the
S03 attacks the catalyst. Above 300°C, ammonia does not combine with SOs.
Thus, heating the catalyst for several minutes at 550—600°C can remove the
sulfate formed on it.
189
-------
CATALYST
I CATALYST INLET (DOUBLE DAMPERS)
2 CATALYST PACKED ZONE
3 CATALYST LEVEL DETECTOR
4 ROTARY VALVE
5 CATALYST OUT (DOUBLE DAMPERS)
6 DUST SEPARATOR
7 BLOWER
8 HEAT EXCHANGER
9 GAS HEATER
10 CATALYST HEATER
GAS FOR
.5
6
TREATED
CATALYST
Ftgure 5 9 Thermal Treatment System
390
-------
Another reason for the usefulness of the alumina—based catalyst is that
not much SOSISO3 oxidation occurs at low temperatures, although a large amount
of oxidation àccurs above 300°C. Some catalysts oxidize SOi considerably,
even at 250°C, which results in reduced activity (Section 3.3.2). The alumina—
based catalyst containing niobium does not appear to oxidize any SOs at 200—
220°C and yet is highly active.
The alumina—based catalyst can be used for 3.000—4,000 hours at a time
and for over 4 years without degradation when used with a low—temperature gas.
This amounts to a considerable energy and catalyst cost savings. The TiOs—
based catalyst produced by Kobe Steel seems to perform similarly to TiOs—based
catalysts manufactured by other producers.
When Kobe Steel’s system using the alumina—based catalyst is compared
with JGC’s system using a TiOs—based catalyst (Section 5.4). it can be seen
that the former requires higher heating temperatures——550600°C versus 400°C.
At the same time, the energy requirement may be much less for the Kobe Steel
system because of a lower required heating frequency and the fact that the
catalyst is heated in a separate vessel. However, overall the Kobe Steel
system is more complex than the .TGC system.
Further studies are needed to demonstrate the applicability of alumina—
based catalysts to gases with various compositions at various temperatures.
5.7 S R FOR NCINERATOR FLUE GAS
5.7.1 Introduction
In Tokyo. Mitsui Engineering constructed a commercial SCR unit to remove
both NO and odor from its sewage sludge incinerator which produces 100.000
Nm 3 /hr of flue gas. Prior to construction. Mitsui conducted extensive tests
at the request of the Tokyo Metropolitan Government. The flue gas was found
to contain over 100 ppm N0 with considerable amounts of SO 1 , HC1, various
kinds of dust, and other odorous components (including aldehydes) which are
difficult to remove by carbon adsorption or wet process oxidation.
391
-------
The government—owned incinerators have sodium scrubbing systems and wet
ESPs for gas cleaning. The treated gas contains about 100 ppm NO 1 , 5 ppm
each of SO 1 and HC1, and has a considerable odor. Mitsui Engineering con—
ducted tests on flue gases both before and after gas cleaning by the scrubber
and ESP. They found that a certain kind of SCR catalyst is especially useful
for odor removal and applicable to the cleaned flue gas, although there are
problems involved in treating the uncleaned flue gas directly. A commercial
S R unit with a capacity of treating 108,000 Nm 3 /hr of flue gas leaving the
sodium scrubber and wet ESP subsequently was constructed; it began operation
in October 1979.
5.7.2 Process Description (6)
The 108,000 Nm 3 /hr flue gas from the incinerator contains about 100 ppm
NO, about 5 ppm each of SO and HC1, less than 0.01 g/Nm 3 of dust and about
10% Oi, 8% COi, 7% HiO, and 75% Na at 40_500C. The flowsheet of the unit is
shown in Figure 5—10. The flue gas is heated to 280°C in a heat exchanger and
then in a heater to 3500C, injected with ammonia, and sent into a reactor
— i
containing a honeycomb catalyst at an SV of 4,000 hr and linear velocity of
7.7 rn/sec. Over 90% of the NO 1 and 80% of the odor are removed by using 1.1
inol NH 3 to 1 md NO 1 . No problems have been encountered since system start-
up.
Figure 5—11 shows test results for the removal of various odorous gaseous
compounds. A synthetic gas containing 10 ppm each of these compounds was sent
through the S R catalyst at different temperatures without ammonia addition.
Over 90% of the compounds (except ammonia) were removed at 300°C or below,
forming COi, HiO and SOa by oxidation. Tests with ammonia added to the gas
indicated that ammonia did not substantially affect the removal ratio of these
compounds.
5.7.3 Economics (6)
The Mitsui Engineering S R unit cost 480 million yen including the price
of the heat exchanger, heater, initial catalyst charge, and civil engineering
392
-------
INCINERATOR
ESP
80
-J
>
0
uJ
60
TEMPERATURE (•Cl
Figure 5-11 Odor Component Removal by Oxidation with the SCR Catalyst
SCRUBBER
BLOWER
REACTOR
Figure 5.10 Flue Gas Trealment System (or an Incinerator.
MM METHYLMERCAPTAN
MOS METhYLOISULPIOE
200
393
-------
services. The annual cost for flue gas trea ent, based on 8,000 hours of
operation in 1980, is shown in Table 5—li. The total treatment cost of 450
yen/1,000 3 of flue gas includes the cost of kerosene for gas heating at 209
yenll,000 Nm 3 . When SCR is not applied, the flue gas leaving the wet ESP must
be heated. Therefore, the actual cost increase caused by the SCR unit is
about 350 yen/i ,000 Nm 3 .
5.7.4 Evaluation
This S R unit may be the first one in the world to remove NO 1 and odor
simultaneously, and has various other NO 1 and odor removal applications. It
is known that the presence of NO 1 increases the odors caused by other con—
pounds. Therefore, NO 1 removal is also helpful in reducing odor. The
process may be considerably less expensive when applied directly to the flue
gas leaving the incinerator. Direct treathent eliminates not only the heat
exchanger and the heater, but also ammonia consumption, since the flue gas
usually contains ammonia sufficient for ScR.
5.8 OTHER SCR UNITS
5.8.1 SCR Unit at Ukishima Chemical Petrochemical’s Chiba Plant
For Ukishima Chemical Petrochemical, Mitsui Engineering constructed an
S R unit with a capacity of treating 260,000 Nm 3 of flue gas from a new oil—
— i
fired boiler. The SQ unit has a tubular catalyst with an SV of 4,000 hr
The unit begain operation in 1978 with an investment cost of 260 million yen
(2).
The 350—400°C flue gas from the boiler contains about 150 ppm NO 1 , 50—
300 ppm SO 1 , and 100—150 mg/Nm 3 of dust. By using 1 inol NH3 to 1 mol NO 1 ,
an 85 NO 1 removal rate has been achieved. The removal efficiency decreased
slightly during the first 2 years of operation. In 1980, one—third of the
catalyst was replaced with a honeycomb catalyst to maintain the high removal
efficiency. Mitsui Engineering currently prefers honeycomb to tubular cata-
lysts.
394
-------
:
‘
“
Plant Design Capacity
(A)
Capacity 100.000 Nn’Ih
(B) Operation Hours 8.000 6./year
Investment Cost
(1.000 Yen)
(C)
Equipment
357.000
(0)
Civil Work
48.000
(E)
Catalyst
75.000
(F) Total
Cost
80,000
Long Term Loan
(1.000 Yen)
(G)
405.000
Interest
7 ,
Short Tern Loan
(1.000 Yen)
(H)
82 925
Interest
9’(
Annual Quantity
Unit Price
Annual Enpense
(1.000 Yen)
{ Electricity
Eerosane
Ammonia
Tertiary treatment
J
Water
Catalyst
I
a’
a.
Consumables
Total Variable Cost
Equip.
D,precia tieS
Civil
Flied Property Tax
Zasuranee
o Labor
•LI
m
Repair
Total Fixed Coat
Total Direct Coat
360 kW 1 (8) kWh
14 Yen/kWh
40.320
380 1/h (B) 1
55 Yen/I
167.200
7 6 kg/h z (B) kg
180 Yen/kg
10.94$
15 t/h z (B)
5 Yen/I
600
25&/2 Years a 1.2 n 5
3.000 Yen/i
31.500
(C) a 0 005
1,785
258.349
(C) a 0.9 + 1
43.900
(0) a 0.9 + 20
2.160
((C) + (0)1 a 0 016 z 1/2
3.240
(17) a 0.005
2.400
3 Men a 2.500.000 Yen Man/Tear
1,500
((C) + (0)1 a 0.03
12 .150
73 .350
331.699
erhead for Bead Office
.1 Long Term Loan
• I
— Short Term Loan
0I
Total Indirect Cost
(17) a 0.01
4.800
(G) a 0.007 a ((1 * 1)/(2 e 7)1
16.200
(8) a 0.09
7.463
28.463
Total Annual Expense
360,162
Treatment Cost
450 Yes/i.000
‘m
TABLE ,- I ‘IITSUI COST CALCULATION (MITSUI 4GINE (ING)
73
I ,
a
3
a
-a
a
395
-------
5.8.2 Shindaikyowa Petrochemical’s Yokkaicbi Plant
In 1975, Hitachi Zosen constructed an SQ unit with a capacity of treat-
ing 400,000 Nm 3 Ihr of flue gas from an oil—fired boiler at Shindaikyowa
Petrochemical’s Yokkaichi plant. The flue gas is treated by sodium scrubbing
using the Weliman—Lord process (2).
The 55 600C flue gas leaving the scrubber contains about 150 ppm N0 ,
100 ppm SOx, and 50 mg/Nm 3 of particulates (including about 20 mg/Nm 3 of
sodium salts). The gas is heated to 330_3400C by a heat exchanger, then to
400—410°C by an oil burner, injected with ammonia and sent into a reactor with
a pellet catalyst. About 80% of the NO is removed by using 1 mol NH3 to 1
mol NOx, producing about 10 ppm of unreacted NH3. The treated gas is then
passed through a heat exchanger where it is cooled to 160_1700C. A serious
ammonium bisulfate deposit problem was encountered with the heat exchanger and
frequent water washings were required.
Since the relaxation of the NOx emission regulations in 1979, 0.6—0.7
mol NH 3 tol mol NOx has been used to remove 60—70% of NOx and maintain
unreacted N H3 below 1 ppm. This change eliminated the previous bisulfate de-
position problem.
5.8.3 flwasaki Steel’s Chiba Plant
In 1976, Hitachi Zosen constructed an SIR unit with a capacity of treat-
ing 762,000 Nm 3 /hr of flue gas from an iron ore sintering machine (2) at
Kawasaki Steel’s Chiba Plant. The flue gas is treated by wet lime scrubbing
and wet ESP, heated by a heat exchanger and heater to 400°C, and subjected to
S R. The S R unit has been operated without problems for 4 years. In late
1980, the catalyst was replaced, although it was still active. The new cata-
lyst (also supplied by Hitachi Zosen) works at lower temperatures and there—
fore saves energy.
396
-------
5.3.4 SCR for Diesel Engine Flue Gas
In 1978, the Unitika Company constructed an S R unit with a capacity of
treating 70,000 Nm 3 /hr flue gas from a power—generation diesel engine fueled
by 1.5% sulfur oil. The 350—400°C flue gas contains 900—1,100 ppm NOx, 600
ppm S0 and 100—150 mg/N.m 3 dust with l3 Oi.
The S R unit uses a 4 mm spherical TiOi—based catalyst in an intermittent
moving bed with an SV of 10,000 hr . Nearly 85% of the N0 is removed
using 0.85 mol NH 3 to 1 mol NOx and unreacted N113 is maintained at a barely
detectable level. The S R unit began operation in April 1978, and has been
operated since that time without trouble. About 10% of the catalyst is re-
placed annually.
After S R and heat recovery, the SO in the gas is adsorbed by acti-
vated carbon. The carbon is washed with water to recover sulfuric acid in the
same manner as has been done with the flue gas from an industrial boiler (5).
Unitika conducted tests of simultaneous N0 and SO removal by activated
carbon and ammonia with thermal regeneration of SOi (2) . However, the tests
were discontinued when carbon consumption was fou.nd to be unacceptably high.
5.8.5 SCR for Gas Turbine Flue Gas
Hitachi Ltd. and Babcock Hitachi constructed a combined—cycle power
generation unit using kerosene, and an SCR unit for Japan National Railway’s
Kawasaki Plant. The gas and steam turbines have a capacity of 97.1 MW and
44.2 MW, respectively, with a total power generation capacity of 141.3 MW.
They produce 1,024,000 Nm 3 /hr of flue gas. Since the flue gas temperature at
the gas turbine outlet is too h2.gh for SCR, the gas is passed through one of
the two boilers where it is cooled to 364°C, treated by SCR, and then sent
into another boiler.
The kerosene fuel contains about 0.05 sulfur and the flue gas contains
small amounts of SO and dust with about 75 ppm NO 1 . A moving bed reactor
with a pellet catalyst is used for SCR. The entire system was designed in
397
-------
1977, when Hitachi’s plate catalyst was not yet well developed. Since the
system is operated about 14 hours a day with a shutdown at night, the catalyst
undergoes startup heating and shutdown cooling every day. In 1977, Hitachi
was not confident about the reliability of the plant catalyst under such
conditions, but currently prefers the plate catalyst in a fixed bed rather
than a moving bed for this type of gas.
The system began operation in April 1981, and has operated without prob—
lems since then. About 8O of the NO is removed by using 1 mol NH3 to 1
mol NOT, which produces unreacted NE3 at a concentration of about 10 ppm.
Because the SO content of the gas is low, the N113 has not caused any
deposition problems in the downstream facilities.
398
-------
REFERENCES
1. Jones, G.] ., and J.D. Mobley. Selective Catalytic Reduction and NO
Control in Japan. EPA 600/7—81—030. U.S. Environmental Protection Agency,
Research Triangle Park, NC, March 1981.
2. Ando, 3. NO Abatement for Stationary Sources in Japan. EPA 600/7—79—
205, U.S. Environmental Protection Agency, Research Triangle Park, NC,
August 1979.
3. Mizutani, H., et al. Test Operation of S R for Coke Oven Flue Gas.
Seitetsukenkyu No. 300, 1980. (In Japanese)
4. Shiba, M. NO Removal from Dirty Gas by Special Reactor. Nenryo Oyobi
Nensho, Vol. 46, No. 9, 1979. (In Japanese)
5. Ando, 3. SOz Abatement for Stationary Sources in Japan. EPA 600/7—78—
210, U.S. Environmental Protection Agency, Research Triangle Park, NC,
November 1978.
6. Kudo, H., et al. Dry Process Techniques for the Removal of NO 1 and Odor
from Flue Gas of a Sewage Sludge rncinerator, Mitsui Engineering Technical
Bulletin, January 1981. (In Japanese)
399
-------
Section 6
OTHER PROCESSES FOR NO REMOVAL
6.1 CLASSIFICATION OF NO 1 REMOVAL PROCESSES
NO 1 removal processes other than selective catalytic reduction (SCR)
may be grouped into two general categories: wet and dry. The processes in
each category are listed in Table 6—1.
TABLE 61. NO 1 REMOVAL PROCESSES OTHER THAN SCR
Dry Process
NO Removal Only
Selective noncatalytic reduction (SNR)
Combination of SNR and SCE
MACT in—furnace NO 1 removal
Molecular sieve adsorption
Simultaneous NO 1 /SO Removal
Activated carbon adsorption
Electron beam radiation
Copper oxide absorption
Wet Process
NO 1 Removal Only
Sodium scrubbing (equimolecular absorption)
Oxidation absorption
Simultaneous NOx and SO 1 Removal
Oxidation reduction
Absorption using complexing agent
Several years ago when SCR technology had serious problems and was very
costly, many of these other processes were being actively tested. However,
Preceding page blank 401
-------
now that S R has been successfully applied and costs much less, some of these
processes have lost their attractiveness and been abandoned.
6.1.1 Dry Processes
NO 1 Removal Ouly
Next to ScR, selective noncatalytic reduction (SNR) has been used most
widely. Several relatively large SNR units and about 20 smaller units have
been constructed. SNR uses ammonia at 800—1 ,0000C without a catalyst to
remove 30—50% of NO 1 (Section 6.2). To increase the NO 1 removal effi-
ciency and to reduce the unreacted NH3 associated with SNR, a combination of
SNR and ScR using a small amount of catalyst has been tested with utility
boilers (Section 6.3).
A second dry NO 1 removal process is MACT in—furnace reduction, in which
a small amount of fuel (about 10% of total fuel) is injected above the flame
to remove about 50% of NO 1 . Commercial application of the process has begun
(Section 6.4).
A third process, molecular sieve adsorption, has been used with nitric
acid plant tail gas (Section 6.9.3).
Simultaneous N0 1 /S0 1 Removal
The Unitika Company tested an activated carbon process to remove 80—90
of SO 1 and NO 1 simultaneously but abandoned the process because of its
high carbon consumption. Tests of the activated carbon process also have been
conducted by EPDC jointly with Sumitomo Heavy Industries (Section 6.5) and
also by Mitsui Mining Company (Section 6.9.2).
Ebara Corporation has developed an electron beam radiation process to
remove 80—90% of SO 1 and NO 1 simultaneously (Section 6.6). The process is
unique in that it converts SO 1 and NO 1 to ammonium sulfate and nitrate
which can be used as fertilizer.
402
-------
Another process, the Shell copper oxide process, has been applied to an
industrial boiler where it removes about 90% of SOx and about 40% of NOx
(Section 6.9.1).
6.1.2 Wet Processes
NOn Removal Only
Essentially all of the NO in combustion gas is in the form of NO. NO
is not easily absorbed in solutions. NOi or an equimolecular mixture of NO
and NOi are fairly reactive with solutions although their reactivity is less
than that of SOi. The tail gas from a nitric acid plant contains both NOa and
NO which can be absorbed by a sodium hydroxide solution to produce sodium
nitrite by the following reaction:
2NaOH + NO + N02 = 2NaNOs + H20 (1)
This is the basis of the sodium scrubbing NO removal process.
However, the demand for by—product nitrite is small. There are several
small commercial units which first oxidize NO to N02, then treat the gas with
sodium scrubbing, and subject the resulting liquor to wastewater treatment.
An oxidation absorption process such as this can not be applied on a large
scale because it requires treatment of a large amount of wastewater containing
nitrite and nitrate. Such trea ent is very difficult.
Simultaneous Removal
Since 1973, many wet simultaneous removal processes have been developed.
The processes can be classified into two groups: oxidation reduction and
complex absorption. and are listed in Tables 6—2 and 6—3.
403
-------
TABLE 6—2. OXIDATION REDUC I0N PROCESSES AND PLANTS
Process Developer
Oxidizing
Agent Absorbent By—Product
Capacity
(Nm 3 /hr)
Sumitomo—Fujikasui
ClOs NaOH (NaNO3, NaC1,
Na,, SO.)
62,000
39,000
100,000
Sumitomo—Fujikasui
ClOs CaCO3 [ Gypsum, CaClz,
Ca(N03)2]
25,000
mi
03 CaCO3 Gypsum, Nz
5,000
MEl
03 CaCO3 Gypsum, NH3
2,000
Chiyoda
Oi CaCO3 Gypsum, Ca(NO )z
1,000
TABLE 6—3.
COMPLEX ABSORPTION PROCESS (USING EDTA AND
FCSO4)
Process Developer
Absorbent By—Product
Capacity
(Nm ’/hr)
Chisso Eng.
NH4OH (NH4) ,SO4
500
Asahi Chemical
NaOH, CaCO3 Gypsum. N,
600
Kureha Chemical
CH3COONa CaCO3 Gypsum, Nz
5,000
Kureha Chemical
KOH Sulfur, N,
2,000
MXX — HZ
KOR Sulfur, N,
Bench Scale
Using the oxidation reduction process, NO is first oxidized by chLorine
dioxide or ozone to NO,, which then is absorbed by a sodium salt solution or a
calcium carbonate slurry. SOa is also absorbed to form sulfite. The NOz
absorbed by the liquor is reduced by the sulfite to form Nz or NR3 while the
sulfite is oxidized to sulfate.
The Sumitomo—Fujikasni process, which uses ClOz and sodium or limestone
scrubbing, has been applied in four relatively small commercial units since
1973. However, its development has been halted because of the wastewater
treatment problem associated with it (1). The similar 1111, MEl, and Chiyoda
processes all use ozone to substantially reduce the wastewater treatment
404
-------
problem. However, these processes are no longer being developed because of
the high cost of ozone.
In order to oxidize NO to NOa economically, an NO oxidation catalyst has
been studied (Section 6.8).
The complex absorption processes listed in Table 6—3 use a liquor con-
taining EDTA and ferrous sulfate to absorb NO and form a complex compound.
SO2 is also absorbed to form sulfite, which in turn reduces the absorbed NO to
N2 or NE ,. Those processes do not require an oxidizing agent but have not
been commercialized because of their complexity and the requirement for a
large multi—stage absorber to remove NO at a high efficiency.
6.1.3 Gas Comi,osition A proyriate for Simultaneous Removal Processes
In general, a flue gas with a relatively low SO/NO 1 ratio is a good
candidate for dry simultaneous removal. In the electron beam process, a high
SO/NO ratio yields a byproduct which is rich in ammonium sulfate and lean in
ammonium nitrate, and is therefore a low grade fertilizer. With the copper
oxide process. a high SOx content necessitates frequent regeneration of the
cuprice sulfate by—product and lowers the NO removal efficiency. Carbon
consumption in the carbon process increases with the SO content of the gas.
On the other hand, a high SOz/NO ratio is preferable for the wet NOx
removal processes because an excessive amount of sulfite is needed to attain
a high NO 1 removal efficiency (1)
6.2 SELEC VE NONCATALUIC REDUCTION (SNR)
6.2.1 Introduction
At about 900°C ammonia (NH 3 ) reacts with NO 1 to form N2 and H:O:
6N0 + 4NH , = 5N3 + 611i0 (1)
4NO + 4N11 , + Oz = 4Nz + 6 HiO (2)
405
-------
Selective noncatalytic reduction (SNR) is based on this reaction. During the
l970s many organizations conducted pilot plant and demonstration tests using
ammonia injection (at 500_l,1000C) for NO 1 removal without a catalyst.
Mitsubishi Chemical Industries (MCI) and Mitsubishi Heavy Industries
(MHI) conducted large—scale SNR tests using MCI ’s six industrial boilers.
Based on the results of those tests, MIII constructed a large—scale SNR test
plant for Chubu Electric at its Chita Station. The plant was designed to
treat all the flue gas from a 375 MW existing oil—fired boiler. MHI also
built a test plant to treat all the flue gas from an existing 375 MW oil—fired
boiler at Tokyo Electric’s Yokosuka Station. MCI submitted a patent applica-
tion for its SNR process three months after a similar application by Exxon.
Subsequently, MCI made a licensing agreement with Exxon although the patents
have not been filed in Japan.
Tonen Technology also has made a licensing agreement with Exxon and has
constructed 8 co=ercial SNR units to treat flue gases from industrial boilers
and heating furnaces at oil refineries and petrochemical plants. Some of
these units are specifically for use during photochemical emergency periods
when more extensive NO 1 reduction is temporarily required (Section 6.2.5).
Mitsubishi Kakoki Kaisha (MKK) has developed its own SNR technology and
has constructed many small commercial SNR units for a gas company (Section
6.2.6).
6.2.2 MEl’s SNR Laboratory Studies (2 )
M I II conducted laboratory tests to determine the effects of reaction
temperature and residence time (reaction time) on NO 1 removal efficiencies.
The tests were conducted using a composite gas containing 200 ppm NO, 400 ppm
ME 3 and 2% Oz (Figure 6—1, A, B). At temperatures between 900 and 1,000°C the
reaction was at an optimum level for NO 1 removal and, in fact, produced a
90% NO 1 removal efficiency in 0.4 seconds. The principal reactions are
shown in equations (1) and (2) of Table 6—4. As can be seen in Figure 6—1. B
406
-------
z
0
U
uJ
r
z
0
z
z
0
0
C
w
I
0
z
700
WORKING GAS TEMPERATURE (C)
(A) EFFECT OF WORKING GAS TEMPERATURE
06
RESIDENCE TIME (sec)
(B) EFFECT OF RESIDENCE TIME
FIgure 6 -1 MHI Laboratory Test Data
407
-------
the reaction rate was low, below 8500C. Above 1,000°C the NOx removal effi-
ciency was lowered due to the partial conversion of NB to NO, as shown in
equation (6) of Table 6—4.
TABLE 6-4. THE MIII ME( ANISM OF THE DE—N0 1 REACTION
Main Reaction
NHi + 1/4 02 ————> NRa + 1/2 HaO Eq. (1)
NH3 + NO ————> Na + RaO Eq. (2)
Sub Reaction
Nh + 1/2 Os ————> 1/2 Na + 1120 Eq. (3)
NHa + Os -—--> NO + IIzO Eq. (4)
Overall De—NO Reaction
4N0 + 4Mb + Os ———> 4Ns + 6HsO Eq. (5)
4NH3 + 5 0 a ———> 4N0 + 611s0 Eq. (6)
4Mb + 30s -S———> 2Ns + 611s0 Eq. (7)
Reaction Formula for NOa
2NOa + 4 NH + Os - ——> 3Ns + 6HaO Eq. (8)
The NO removal efficiency at an actual plant may be much lower than
the laboratory rate due to insufficient reaction times and the difficulty
involved in obtaining a rapid uniform mix of ammonia and gas. In order to
study these problems, the reaction constants Ki. Ks and Ks (as shown in Table
6—5) were determined empirically for different temperatures. This makes it
possible for the NO/NHs reaction rate to be calculated using equations (12)
and (13) of Table 6—5.
TABLE 6-5. BASIC FORMULAS FOR THE M I II PROCESS DE—N0 1 REACTION AND RATE OF
REACTIONS
4N0 + 4Mb + Os — —> 4Ns + 6RaO Eq. (9)
4NHs + 50a -k—) 4N0 + 6HsO Eq. (10)
4 ’ffl + 3O —k—) 2Ns + 6HsO Eq. (11)
— d(NO ) =
— Ka(NB S] [ O s]Y Eq. (12)
— d(NRs ) =
+ (Ks + K3)(NB3] (O ] Eq. (13)
408
-------
Computer simulation programs have been developed to predict the NO 1
reduction in a large furnace. As shown in Figure 6—2, a cross section of the
furnace gas passage is divided into small sections which take into account the
ntimber of ammonia injection nozzles. Each section is divided into smaller
subsections both crosswise and in the direction of the gas flow so that dif—
ferential equations for diffusion and mixing can be solved.
The results of one computer calculation are shown in Figure 6—3. The
NO 1 removal efficiency is shown along the gas flow axis. In this case,
ammonia is injected into the flue gas at 1,050°C and drives the gas tempera-
ture down to 900°C in 0.14 sec. over a distance of 1.8 meters. This results
in an NO 1 removal efficiency of more than 50 percent.
6.2.3 SNR Tests at MCI’s Boilers (2 )
Since 1974, SNR tests have been conducted with MCI’s 6 existing oil—fired
boilers (Table 6—6). All of the SNR units were constructed by MEl. The No. 3
and No. 4 boilers are the largest ones, each with a capacity of 180 MW equiva-
lent.
As shown in Figure 6—4, ammonia injection nozzles, FF, FM, and FR with
cooling water tubes were placed in three locations in the No. 3 boiler and
also at the boiler wall (nozzle F). Figure 6—5, A shows the relationship
between boiler load and NO 1 removal efficiency when 2 mol NH3/mol NO 1 is
injected from each nozzle. The FM nozzle produced the highest efficiency
(50%) at full load, while nozzles F and FF produced relatively high effi-
ciencies at the lower load. As shown in the B portion of the figure, the
NO 1 removal efficiency reached 6O with 4 mol NH3/1 mol NO 1 .
Because the nozzles used cooling water, low—temperature corrosion occur-
red due to the condensation of sulfuric acid derived from the SO3 in the flue
gas. As shown in Figure 6—6, the nozzles for the No. 4 boiler use boiler
water for cooling and thus improve process economy and eliminate the possi-
bility either of corrosion or of thermal damage.
409
-------
TABLE 6—6. BOILERS USED FOR THE MEl APPLICATION THSTS
- Capacity
(Flue Gas
Unit Name Flow) Nm 3 /h
Boiler
Type
Fuel
Firing
N/R
N113
Injection
Nozzle
Start
of
Tests
Mizushima I — #1, MCI. 175,000
OT
L.S.
Oil
T
R
N N
Aug.
1974
Mizushima I — #2, MCI. 194,000
OT
LS.
Oil
T
R
BWN
Apr.
1977
Mizushima I — #3, MCI. 540,000
OT
U.S.
Oil
T
R
BWN
Jane
1975
Mizushima I = #4, MCI. 540,000
OT
U.S.
Oil
T
R
BWN
Oct.
1974
Yokkaichi #1, MCI. 200,000
OT
U.S.
Oil
T
R
SCN
May
1976
Yokkaichi, YD Boiler, 40,000
MCI.
NC
U.S.
Oil
H
R
SCN
Sept.
1975
Chita #2, Chubu 1,000,000
Electric (Unit Capacity
350 MW)
CC
L.S.
Oil
H
R
BWN
Feb.
1976
Yokosuka # 4 a 1,000,000
Tokyo Electric (Unit Capacity
350 MW)
CC
L.S.
Oil
H
H
BWN
March
1978
aThe Yokosuka plant uses also a catalyst (Section 6.3).
Remarks: Boiler Type: NC — Natural circulation
CC — Controlled circulation
OT — Once—through
Firing: I — Tangential, H — Horizontal
N/R: N — New unit, R — Retrofit
NH3 Injection Nozzle:
B YN — Box—type water cooled nozzle
SCN — Steam cooled nozzle
NCN — Non—cooled nozzle
410
-------
Figure 6.2 Calculation Procedure Used in the MHI Simulation Program.
DIFFUSION
MIXING ZONE
411
-------
CAVITY
JET DILUTION’
AREA
THE LOCATION
OF JET FLOW
DIFFUSION MIXING AREA
Figure 6-3 Results of a Computer Simulation of SNR DE-NO Performance.
HEATING SURFACES ZONE
60
>-
0
z
Lu
0
I L .
IL
Lu
-J
4:
>
0
LU
I
x
0
z
-1100 —
0
a
-1000 o
-900
-800 w
-J
LL
—700
0 200 400 600 800 1000 1200 1400 1600 1800 2000
DISTANCE OF GAS FLOW DIRECTION X (mm)
—
412
-------
NOZZLEF 3x3=9
NOZZLEFF 3x1=3
NOZZLEFM 4x1=4
NOZZLEFR 3x1=3
ECONOMIZER
Figure 6-4 Ammonia Inlection Nozzles for Mizushima Plant. Unit No 3
F
FLUE GAS
WALL
413
-------
>.
0
z
LLj
0
Li-
w
-J
4
>
0
w
0
z
>-
0
z
w
0
U.
U.
w
-J
4
>
0
w
0
z
10
NH 3 /NO,, MOLE RATIO
Figure 6 5 Test Results for Mizushima Plant’s Unit No 3
BOILER LOAD (%)
414
-------
EVAPORATOR
OUTLET
THE PORTION OF THE LOWER
NANGER EVAPORATOR TUBES
70A2832
U.U. U.
w
-J
0
z
NOZZLE
EVAPORATOR
INLET
Figure 6.6 MPH Nozzle Cooling System for Unit No. 4, Mizushima Plant.
415
-------
6.2.4 Full—Scale SNR Test Unit at Chubu Electric’s Chita Station (2,3)
Boiler and SNR Unit
In 1976, Chubu Electric installed a full—scale SNR test unit for the
existing No. 2 boiler at the Chita Station. The No. 2 boiler has a capacity
of 375 MW and burns 0.2% sulfur oil. Among the four boilers which existed at
Chita, the No. 2 boiler was chosen as being the most suited for SNR; it had
space available for ammonia injection nozzles with adequate temperature
ranges.
The specifications of the Chita SNR system are presented in Table 6—7.
Since the boiler load fluctuates between 375 and 120 MW, a total of 15 nozzles
were placed in two locations, front and rear, as shown in Figure 6—7.
TABLE 6-7. SNR SYSTEM SPECIFICATIONS, UNIT NO. 2, CRITA STATION, CHUBU
ELECTRIC POWER COMPANY
BOILER:
Unit Name Unit No. 2 of Chita Station, Chubu
Electric
Evaporation, t/h 1,225
Steam Pressure, kg/cm G 176
Steam Temperature, °C 571/541
SNR TEST FACILITY:
Treated Gas Flow Rate, Nm 3 /h 1,000,000 (375 MW rating)
Retrofit Period Aug. 1976 — Jan. 1977
NR3 Injection Nozzle 15 sets
(Fin—welded box type)
Booster Air Fan for NB3 1 set
dilution and transport
N113 storage and supply 1 set
SNR APPLICATION TEST START Feb. 1977
PREDICTED DE-N0 1 PERFORMANCE:
Target NO 1 Removal 40
Efficiency, %
416
-------
* Boiter Type Mitsubishi-CE Control. Radiant Reheat Divided
Evaporation 1225 t/h, 375 MW
Pressure 176 atg
Temperature 571(540 6C
Figure 67 Locations ot Nozzles in the No 2 Boiler Chita Station
*
REAR
NOZZLE
BURN ERS-
417
-------
As shown in Figure 6—8, the rear nozzles (injection point 2) are used
when the boiler load exceeds 290 MW, the front nozzles (point 1) are used with
a boiler load below 220 MW, and both sets of nozzles are used with a load of
between 220 and 290 MW. In this way, ammonia is always injected into the gas
within a suitable temperature range——930—1,030°C. When the boiler load goes
below 180 MW, ammonia injection is stopped. At this load level, the amount of
NO 1 produced is low enough to meet the regulations without control measures.
The SNR unit has operated almost continuously since February 1977 except
for a period when the boiler was shut down. Figure 6—9 indicates the perfor-
mance of the SNR system. By using 1.5 mol NE3/1 mol NO 1 , 35—45% of NO 1 is
removed with 15—25 ppm of unreacted NH3. With a higher inlet NO 1 concentra-
tion (about 150 ppm) the SNR unit produces a higher removal efficiency (45%).
When the nozzles were modified to have more holes the result was better utili-
zation of NR3. The same NO 1 removal rate was obtained using 1.3 mol N113
with nozzle modification and 1.5 mol Nil 3 without the modification. Recently
at the plant about 1 mol NH 3 has been used to produce about 35% efficiency
with 10 ppm unreacted N113.
Evaluation/Economics
The operation of the SNR system at Chita Station has been virtually
trouble—free except for problems associated with SO 1 and NH3 in the treated
flue gas. The boiler flue gas contains about 100 ppm SOs and 1—2 ppm S03
which tend to combine with NH3 and condense on cooling. This produces 2
problems. First, during the winter, when unreacted N i l 3 exceeded 40 ppm, a
plume was observed coming from the stack. Second, when the system began
operation the pressure of the air preheater soot blower was too low. As a
result, deposits in the preheater caused the pressure drop to increase from
100 to 140 mm ilsO in two months. Eventually, the boiler had to be shut down
for washing of the preheater. The deposition has been reduced by using high—
pressure soot blowing and less NH3 to reduce unreacted NIh.
418
-------
POINT
NH 3 INJECTION AT POINT
NH 3 INJECTION AT POINT®
/
K
ONLY FRONT
NOZZLES
(POINT ®
BOTH
NOZZLES
/ — —
.% ,‘ ONLYREAR
NOZZLES
‘\ (POINT®)
/
375
200 300
LOAD (MW)
Figure 6.8 Elfects of NH 3 Injection Position and Load on NO 1 Reduction Rates
During Operation of the No 2 Boiler at the Chita Station
L)
LU
I-
LU
0.
LU
a)
(3
U i
-J
U.
POINT
1200
1100
1000
900
800
-
100
60
40
20
z
0
I-
0
a
LU
0
z
419
-------
>-
0
z
w
0
U-
U-
LU
-J
0
LU
0
z
7C
60
c 05 10 15 20 25 30
NH 3 1NO MOLE RATIO
Rgure 6.9 Do NO Performance Test Results for Unit No.2 at the Chita Station
/
/
/
/
I I I I
420
-------
The SNR unit cost 600 million yen (1,600 yen/icE in 1976. This cost is
quite high due to the price of boiler modification for retrofitting of the SNR
unit. In general, both the costs and the NO 1 removal efficiency of SNR are
slightly less than half of those for SCR.
6.2.5 Exxon Thermal DeNO 1 Systems
Eight thermal DeNOx SNR systems which use the technology developed by
Exxon (Table 6—8) have been constructed at oil refineries and petrochemical
plants. In most of the boilers and furnaces there was available space in
which to install the ammonia injection nozzles at the optimum temperature of
900_1,0000C. The exception is at Mitsui Petrochmical’s industrial boiler at
Chiba. There the nozzles were installed where the temperature is 700—800°C
and hydrogen is used with ammonia to increase NO 1 removal efficiency. The
ammonia and hydrogen are diluted with steam; air cannot be used for dilution
because of the risk of explosion.
Mitsui Petrochemical’s SNR unit has nozzles in three locations where the
temperature is between 700 and 800°C. Two mol NH3 and 1—2 mol Hz to 1 mol
NO 1 are injected from the three locations according to the boiler load. As
such, the unit removes 30—45% of inlet NO 1 (100—150 ppm) with 10—15 ppm of
unreacted N H3. When the unit began operation in late 1975. the boiler burned
0.8% sulfur fuel oil and the air preheater became clogged with deposits in
about 700 hours. Since then either gas or cracked naphtha fuel containing
0.2% sulfur has been used to reduce the deposition. The unit was operated for
about a year and then shut down for the following reasons: 1) low NO 1
removal efficiency with high unreacted N , 2) high consumption of NH3 and Hi.
and 3) the fact that combustion modification can reduce NO 1 for a lower
cost.
Tonen Sekiyu Kagaku (Tonen Petrochemical) and Toa Nenryo each operate two
SNR systems (at Kawasaki and Wakayama, respectively) which meet current regula-
tions and remove 10—20% of NO 1 using 0.3—0.5 mol NH3 to 1 mol NO 1 . Unre—
acted NB3 has been maintained at 5 ppm in all four units. The regulated NO 1
limit is about 140 ppm at Kawasaki and 190 ppm at Wakayama.
421
-------
Plant
User Site Gas Source Fuel
Mitsui thiba Industrial Naphthu
Petrochemical boiler
Tonen Sokiyu Kawasaki Industrial Low—S
gagaku boiler oil. Gas
Tonen Seklyu awasaki Industrial Low—S
agaku boiler oil. Gas
Kyokyto Chiba Pipestill Low—S
Petroleum furnace oil. Gas
Toa Nenryo Kawasaki CO boiler CO
Toa Nenryo Kawasaki Ptpcstill Low-S
furnace oil, Gas
Toa Nenryo Wakayams Industrial Low—S
boiler oil
Toa Nenryo Wakayams Industrial Low—S
boiler oil
NH, is injected at three location..
NH, is Injected at one location.
is used at about 800°C or bolos.
Gas
Treated
(Nm’/hr)
Inlet
NO
(ppi )
Nih/NO
mole ‘
ratio
Use Gas
of temp.
II , (‘C)
NO
remo al
(%)
Date
of Status of
Start operation
120.000
130
1 _ 2 a
Yes 700—800
30—40
Dec. No operation
1975 since 1977
468.000
—150
03 _ 1 b
C 700—1000
10—50
Oct. Occasional
1976
468.000
—150
0 , 3 _ 1 b
10 5 c 700—1000
10—50
July Occasion.l
1977
160.000
80—100
2 0 b
105 C 700—1000
50
Oct. For emergency
1977
314 .000
200
12 b
Yes 700-1000
45—55
March For emergency
1978
254.000
155
1 _ 2 b
y 05 C 700—1000
50—55
March For emergency
1978
176.000
—200
0 • 3 _ 0 5 b
No 850—900
10—20
Aug. ContInuous
1978
176.000
—200
O.3 O.5 ’
No 850—900
10—20
Feb. Continuous
1981
-------
The 3 other SNR units listed in Table 6—8 have been used for emergency
purposes when total NO emissions from the plants exceed the regulation or
when photochemical smog warnings require further NO reduction. The units
each remove 45—55 t 7o of NOx using hydrogen and 1—2 NB3/mol NOx.
Economics
Tonen Technology calculated the typical cost figures for an SNR system
applied to an existing boiler and furnace; these are presented in Table 6—9.
TABLE 6—9. TEERMAL DeMO 1 COST FIGURES CALCULATED BY TONEN TECHNOLOGY
Boiler Furnace
Flue gas volume, Nm 3 /hr 500,000 200,000
(MW equivalent) 165 65
Inlet NOR, ppm 200 200
Outlet NO 1 , ppm 100 100
Investment cost (10’ yen), 80 90
Utility consumption
N113, kg/hr 80—120 30—45
Steam,t/hra 2.5—3.0 1.0—1.2
Hi, Nm3/hrb 0—80 0—40
Air may be used when Hz is not used.
H 2 is used when the gas temperature is below 850° C. Hz costs 10—20 yen/Nm 3 .
The investment cost may be less for a new boiler and furnace. fost of
the cost of SNR systems for existing boilers and furnaces is for retrofitting.
These modifications differ considerably with each boiler and furnace. Mitsui
Petrochemical’s SNR unit cost 140 million yen in 1975, including the cost of
423
-------
the NO 1 analyzer, and control system and royalty fees. The cost is equiva-
lent to 3,500 yen/kW (*14 1kW) and is fairly high due mainly to the installa-
tion of nozzles in 3 locations. The annual operating cost of the same unit
was 150 million yen including 7 years depreciation, interest, overhead, and
other similar costs. The annualized cost was 0.5 yenlkWhr.
6.2.6 Toho Gas Company’s SNR units (4)
MKK constructed many small commercial SNR units for the Toho Gas Company
at the Sorami and Chita plants near Nagoya City (Table 6—10). The units are
for kerosene—fired boilers, naphtha—fired naphtha evaporation furnaces, and
steam naphtha superheaters and each has a unit capacity of treating 3,200 —
10,700 Nm 3 /hr of flue gas.
The NO 1 removal efficiencies of these units are relatively high and
range from 54 to 62 percent for the boilers, 58 to 67 percent for the evapora-
tors, and 44 to 56 percent for the superheaters. No other information
regarding their operation is available.
It is presumed that the inlet NO 1 concentrations are in the range of
100—200 ppm, and that the relatively high removal efficiencies may be due to
the following: 1) the units’ small size which enables rapid and uniform
mixing of ammonia, 2) a relatively long reaction time in a suitable tempera-
ture range, in the case of the naphtha evaporator, 3) possible use of a large
NH3/NOx mole ratio (2.0 or more), and 4) no ammonium bisulfate problem even
with a large concentration of unreacted Nils, since the fuel contains little
sulfur.
6.2.7 Evaluation
SNR is the simplest method of flue gas treatment for N0 abatement and
is well suited to boilers and furnaces which require up to 50 NO 1 reduc-
tion. In order to remove NO 1 efficiently, ammonia must be mixed with the
flue gas very rapidly and uniformly. The mixed gas then should be maintained
for a certain time period, preferably more than 0.2 seconds, within a suitable
424
-------
TABLE 6—10. TOHO GAS COMPANY’ S SNR UNITS
Plant
site
Gas
Source
Fuel
Gas
treated
(Nm 3 /hr)
NOx
removal
(%)
Startup
date
Sorami
Boiler
Kerosene
3,490
62
Oct. 1976
“
“
sNr
NE
Naphtha
“
2,570
3,540
56
67
“
“
“
Boiler
Kerosene
3,840
54
Oct. 1977
•
SNS
Naphtha
4,362
44
1
NE
“
6,216
58
‘I
Boiler
Kerosene
3,870
54
Nov. 1977
1
SNS
Naphtha
4,762
44
1
“
NE
6,216
58
“
SNS
‘I
3,168
44
Dec. 1977
“
NE
“
4,166
58
“
Boiler
“
3,169
44
June 1978
1
NE
“
4,166
58
II
Boiler
Low—S oil
10,700
60
July 1978
“
SNS
“
3,800
62
1
NE
“
5,200
62
Chita
NBC
Kerosene
6,000
50
Oct. 1977
“
SNS
“
4,300
50
U
“
NH
u
3,200
50
Steam naphtha superheater
evaporator
Naphtha heater
425
-------
temperature range—--——900—l ,O 30 °C without H 2 and 700 9O0° with Ha. Hydrogen
may be used at refineries and petrochemical plants but is not appropriate for
use at power plants.
SNR also may be suitable for certain types of furnaces that have rela-
tively long gas retention times at the appropriate temperatures. Over 60
NO 1 reduction may be attained with these furnaces. On the other hand, a
removal efficiency of 30—40 percent may be a practical limit for utility
boilers which must maintain unreacted Nib at a low level even during load fluc-
tuations.
SNR works well in an emergency situation when 40—50 percent removal of
NO 1 is needed temporarily. In this application, however, the maintenance of
the ammonia injection nozzles may present a problem, since they must be kept
cool at all times. Mill’s cooling system which uses boiler water can be used
to reduce the cooling costs.
SNR also may be useful for treating very dirty gas such as gas from a
glass melting furnace or a municipal incinerator which contains par ticulates
and gases capable of poisoning S R catalysts.
A combined SNR/SCR system can be used to attain a higher NO 1 removal
efficiency (50—70 percent) than can be attained by SNR alone. This combina-
tion is described in the following section.
6.3 SNPJSCR COMBINED SYSTEMS
6.3.1 Introduction
Between 1978 and 1980, both Toyko Electric and Kamsai Electric conducted
large—scale tests of combined SNRJSCR systems applied to existing oil—fired
boilers. In the combined system. 1000°C ammonia is injected into the boiler
for SNR; the 300—400°C gas leaving the boiler economizer is sent through a
small amount of parallel—flow S R catalyst (located in the existing duct) for
further reduction of NO 1 and Nib. The combined system’s advantage is that
426
-------
it removes 50—60 percent of NO in flue gas from an existing boiler, re-
quires neither an SCR reactor nor an additional fan, and produces less than 10
ppm of unreacted N113.
6.3.2 Tokyo Electric’s SNR/SCR Demonstration Plants
Tokyo Electric’s three SNRJSCR demonstration plants are described in
Table 6—11. All of the combined units were installed for existing oil—fired
boilers which burn low-sulfur oil and produce a flue gas with 120—200 ppm
N0 .
The SNRISCR system for the No. 2 oil—fired boiler (350 NW) at Tokyo
Electric’s Ohi Power Station is shown in Figure 6—10. Fifty ammonia injection
nozzles, each with 40 holes, were installed at one location in the boiler
where the gas temperature is about 1,000°C at full load. A honeycomb cata-
lyst, (87 m 3 (SV 11,720 hr ), was installed in the duct between the econo-
mizer and the air preheater. For test purposes, honeycomb catalysts with
different compositions and channel diameters, ranging from 8 to 10 mm, were
used. The pressure drop through the catalysts bed was less than 30 mm 1120.
The combined system at Tokyo Electric’s Yokosuka Station is shown in
Figure 6—11. It has ammonia injection nozzles at two locations in order to
compensate for changes in the gas temperature.
The demonstration plants operated for 7,900 — 10,700 hours as shown in
Table 6—11. Similar results concerning N0 removal efficiency and unreacted
NB were obtained at all three plants. A typical example of a test result is
shown in Figures 6—12 and 6—13. By using 2.0—2.5 mol N113 to 1 mol NOx, the
NO 1 removal efficiency was 50—60 percent with unreacted N f l 3 below 10 ppm.
Roughly 20 percent of the NO was removed by SNR; the rest was removed by SCR.
The NO 1 removal efficiency was higher at the smaller boiler load as
shown in the Figure 6—13. The reduced load was unfavorable for SNR because of
the decrease in the gas temperature but increased the SCR efficiency because
427
-------
TABLE 6—11. TOKYO ELECTRIC’ S SNR/ SCR DEMONSTRATION PLANTS (5)
Yokosuka No.4
350
(1,076,000)
Low sulfur oil
7,900
Mitsubishi
Heavy
Industries
Ohi No.2
350
(1,020,000)
Low sulfur oil
Honeycomb
June 1978
— Sep. 1980
10,700
Ishikawaj ima
ilarima Heavy
Industries
Yokohama No.1
175
(531 ,000)
Low sulfur oil
Plate
Aug. 1978
— Sep. 1980
10,500
Babcock
Hitachi
0.1 — 0.2 0.1 0.1 — 0.2
Power Station
and Unit No.
Capacity in MW
( in 3 N/B)
Fuel
Sulfur content in
fuel (wt ° )
Dust in flue gas, mg/Nm 3
(at economizer outlet)
NO 1 in flue gas, ppm
NO 1 reduction efficiency
Residual ammonia
Ammonia injection
(N1H /N0 1 mole ratio)
Type of catalyst
Test Period
Operating Hours
Manufacturer
Approximately 50
200 — 120
More than 50 percent
Less than 10 ppm
1.5 — 2.5
Honeycomb
& plate
Apr. 1978
— Aug. 1980
28
-------
Figure 6.10 Simpllfled Flowsheet of the Combined SNRISCR System at Tokyo
Electric’s Oh, Station.
429
-------
Figure 6-li Schematic Diagram of the Combined SNRISCR System at Tokyo
Electric’s Yokosuka Station (2)
-NH 3
BOILER
NOZZLE CC
BOOSTER FAN
FOR NH 3
430
-------
NH 3 lNO MOLE RATIO
Figure 6-12 Typical OperatIon Data tot Tokyo Electrtc’s CombIned
SNR!SCR System. (2)
z
0
I-
U
C
uJ
0
z
6O
50
B
I
z
- I
L°________
25
50
I D
tOO
BOILER LOAD (%)
Figure 6.13 Tokyo Electric’s Test Data: Effect of Boiler Load
on the SNRISCR Combined System (5)
431
-J
>
0
w
0
z
z
-J
(I )
U i
-------
of the smaller SY (longer reaction time through the catalyst bed). The in-
creased SCR efficiency was more than sufficient to compensate for the decrease
in SNR efficiency.
The catalyst was effective not only for increasing NO 1 removal effi-
ciency but also for reducing unreacted Nib below 10 ppm. Unfortunately, the
air preheaters experienced ammonium bisulfate plugging. The only exception
was at the Ohi plant which only had a slight bisulfate problem probably be-
cause of the low sulfur content of its fuel. The air preheaters at the other
stations required water washing once every 3 or 4 months.
The demonstration tests were completed in September 1980. At that time
the SNR ammonia injection nozzles were removed. Since that time, at the Ohi
plant a small amount of ammonia (0.15 — 0.2 mol NHu/mol NO 1 ) has been
injected through the nozzles for S R to remove 15—20% of No 1 with 2—3 ppm of
unreacted NH3.
Use of the SNR systems was discontinued for several reasons. First, the
local NO 1 regulation can be met using combustion modification. For example,
the NO 1 regulation for the Ohi No. 2 boiler is 140 ppm. This level can be
achieved by using combustion modification, including staged combustion, flue
gas recirculation and a low—NO 1 burner. For safety reasons, 15—20% of NO 1
from the boiler is removed by SCR. Second, because of the high temperatures
in which they were used, the SNR ammonia injection nozzles caused maintenance
problems. Another reason was that in the SNR system the ammonia consumption
is high and the unreacted Nib is higher than it would be in a standard SCR
plant. This caused ammonium bisulfate deposition problems in two plants.
6.3.3 ansai Electric’s Full—Scale SNR Test
In October 1977, Kansai Electric installed a full—scale SNR test unit for
an existing 156 M’ N oil—fired boiler at the Himeji No. 1 Station. In July
1978, a small amount of S R catalyst was placed in the duct between the boiler
economizer and air preheater to increase NO 1 removal efficiency to 50—70%
and to reduce unreacted NH .
432
-------
In 1979, the SNR unit was removed primarily because of the maintenance
problems associated with the ammonia injection nozzles. Since that time, the
SCH catalyst successfully has removed about 30% of the NO 1 with little
catalytic degradation.
6.3.4 Evaluation
A combined SNRJSCR system was once considered ideal for removal of 50—60%
of NO 1 in flue gas from existing boilers. However, maintenance of the sys-
tem’s ammonia injection nozzles proved to be a serious drawback. For most
existing utility boilers, a standard S R reactor can be retrofitted between
the boiler economizer and the air preheater to remove over 80% of NO 1 .
Alternatively, a parallel flow catalyst placed in the existing dust can remove
30—40% of NO 1 .
The combination system still may be valuable for removing 50—60% of NO 1
from existing furnaces and boilers that have no space for installation of SCR
reactors.
6.4 MITSUBISHI ADVANCED COMBUSTION TECHNOLOGY (MACfl IN-FURNACE NO 1
REMOVAL PROCESS
6.4.1 Introduction
MIII developed the Mitsubishi Advanced Combustion Technology (MACT) in—
furnace process to remove about 50% of NO 1 from a boiler. In the process
the secondary fuel——about 10% of the fuel for the boiler——is used as a reduc-
ing agent. At temperatures above 1200°C the secondary fuel is injected into
the boiler combustion gas to create a reducing atmosphere. In this atmosphere
70—90% of NO 1 is converted to N2 and llzO and additional air is injected for
complete combustion (Figure 6—14) . Since a small amount of NO 1 is formed
during the complete combustion stage, the overall NO 1 removal efficiency of
the process is usually about 50%.
433
-------
OVER FIRE
AIR
PRIMARY
BURNERS —
COMPLETE
COMBUSTION
REDUCING
ATMOSPHERE
-
STAGED COMBUSTION
Figure 6-14 Staged Combustion and Staged CombustlonlMACl Processes.
ADDITIONAL
AIR
SECONDARY
BURNER
COMPLETE
COMBUSTION
REDUCING
ATMOSPHERE
II . ’
OVER-FIRE
AIR
PRIMARY
BURNERS
STAGED COMBUSTION
AND MACI
434
-------
After completing small—scale tests, MHI joined with Tokyo Electric to
conduct pilot plant tests using corner—firing furnaces. With a combination of
combustion modification and MACT, NO 1 concentrations were lowered to about
60 ppm for coal, 30 ppm for oil and 10 ppm for gas fuels. Tokyo Electric has
been using the MAC’.t/staged combustion system commercially for two small oil—
fired boilers (125 MW each) which were originally designed to burn coal. The
system has reduced NO 1 to 30—40 ppm.
6.4.2 Small—Scale Tests (6,7)
Figure 6—15 and Table 6—12 provide a description of the small MACT test
furnace. The compositions of the domestic low—sulfur coals which were used
for the test are shown in Table 6—13.
TABLE 6-12. SMALL TEST FURNACE SPECIFICATIONS
Furnace Diameter 0.8 m Length 2.2 m
Castable refractory—lined cylindrical furnace
Primary burner Propane (10 kgfhr) or pulverized coal (100 kg/hr)
Air at room temperature
Secondary burner Propane or pulverized coal 10 kg/hr
Air at room temperature is used for combustion.
Pulverized coal is carried by nitrogen gas.
TABLE 6—13. COMPOSITION AND HEAT VALUE OF COALS (200 mesh 90% pass)
Fixed
Ash Volatile Carbon Moisture iN S Heat Value
(%) ( ) (%) (%) (%) (kcal/kg)
Small
scale tests 13.2 42.2 38.4 6.2 1.0 0.28 6220
Pilot
plant tests 10.1 42.6 40.8 6.5 1.0 0.20 6400
Figure 6—16 illustrates test results when propane fuel was used. Propane
was burned in the primary burner producing a flue gas containing 140 ppm NO
and 1% 02. A reducing gas formed by the partial combustion of the propane
(10% of primary fuel) was introduced into the flue gas at 1,200°C; this
435
-------
FLOW METER
*Dlmensuons are in mm
* *Location ol the secondary burner is changeable
FIgure 6 15 Flowsheet of a Small-Scale MACT Test Plant
-------
reduced NO to 25 ppm. For complete combustion. additional air was introduced.
This had the effect of increasing NO to 70 ppm, resulting in an overall NO 1
removal efficiency of 50%.
Figure 6—17 shows that NO 1 was reduced when the secondary fuel was
injected above 750°C and that the NO 1 removal efficiency was higher when the
flue gas contained a lower Oz concentration before the secondary fuel
injection.
Figure 6—18 shows the results of a test using a low—NO 1 burner with
pulverized coal fuel. The NO concentration ranged from 100 to 150 ppm before
the secondary fuel was injected, and was reduced to 50 to 60 ppm by the MACI.
The gas at the stack contained 10 ppm CO, but no hazardous gases such as HCN.
6.4.3 Fundamental MA T Studies (6)
Figures 6—19 and 6—20 show the NO 1 decomposition ratio that occurred
with propane fuel before the introduction of the additional air for complete
combustion. The decomposition ratio was influenced by X , which is defined by
the following equation:
= Inlet 02 (02 in flue gas before secondary fuel was added )
Oa required for complete combustion of the secondary fuel
About 90% of NO was decomposed at 1,150 — 1,300°C when . was 0.18 and
0.53, while the decomposition ratio was only about 10% when was 2.2. At
1,300°C a reaction time of 0.1. second was sufficient for decompostion when X
was 0.53. Figure 6—20 shows that the decomposition ratio was not influenced
by the inlet NO concentration, which ranged from 30 to 1,000 ppm.
Figure 6—21 shows test results with different fuels. When X was between
0.15 and 0.9, 80—95% of NO was decomposed for all fuels including gas. oil,
and coal. The decomposition ratio was considerably lower when X was below 0.1
and above 1.0. This indicates that a small amount of Oz is needed for the
reaction and that excessive Oz prevents decomposition.
437
-------
SECONDARY
FUEL ADDITIONAL
COMBUSTION —
1500
1300
0
1100
900
700
500
150
100 0
0 6
z
50 4
0 C
NO
02
2 4 6 9 10 12 14 16 18
DISTANCE FROM INLET (ml
(Fuel propane)
Figure 8.16 Change In NO Concentration Cau,ad by MACT
02=1%
-- -•0- ....
--
t
Z 500 600 700 800 900 1000 1100 1200
GAS TEMPERATURE AT SECONDARY BURNER (C)
Figure 8.17 Gas Temperature. 02 Percentage, and NO. Removal
Ounng MACT Treatment
438
-------
I SO
100-
0 p TERMACT
50
1 2 3 4 5 6
02 IN GAS 1%)
(Fuel oulveruzed coal)
FIgure 6-18 NO, ConcentratIon Sot ore and After MACT Treatment
0
900 1000 1100 1200 1300 1400
REACTION TEMPERATURE (C)
(Inlet 02 1% inlet NO 92.103 DDT1, reaction time 02 sec
gas volume 5 Nllmtn tuel Dropane)
FIgure 6.19 110 Oecompceltlon Ratio at Different
Tempereturea and 02 ConcentratIons
439
-------
100
z
0
—
U,
0
a.
0
0
w
0
z
z
0
I-
U)
0
a-
0
0
U I
0
0
z
ruEl.
Ci i ,
C,H
C.H.,
Ciii
•
C a
• .—
• .
•• 0 0 ••
S _____ 3t
S ___ \.
0
___ \.
(1300C 02 sec 02 0 1 10°/i inlet NO about 80 ppm
gas volume 5 NI/minI
Figure 8-21 RelationshIp Between A and NO Decomposition for Vaflous Fuels.
KEY
0
.
.
S
0
0
V
T UP 1C1
I I SO
¶300
¶300
I SO
1300
‘Ito
¶300
80
60
40
20
n
0
KEY
V 018
0 053
1 22
20 100 500 1000
INLET NO (ppm) -
(Fuel propane 1300C, reaction time 02 sec
inlet 02 1 ‘/o gas volume 5 NI/minI
Figure 8-20 Inlet NO and Oecompo.tlon Ratio.
100
80
60
20
4 06 08 10 12 14
•
*
KEROSENE{ 1300
S
02 0
COAl. J ¶300
S
440
-------
The decomposition reactions are assumed to be these:
Cm ‘ Hn ‘ *
CmHn +.02 —) + CO + H O (1)
Ra di cal
NO + 1m’Hn’ ——> Cm”Rn” + Ni + HaO + CO (2)
NO + Cm’Hn’ —> Cm”En” + + 1120 + CO (3)
NH compounds
indicates radical
Hydrocarbon (secondary fuel) is decomposed by Oz to form hydrocarbon
radical (1), which reacts with NO to form Ni (2) and NHi or nitrogen hydrogen
compounds such as N113 and NBa (3)
Figure 6—22 shows the relationship between the amount of NO decomposed
and the amount of NHi formed from the NO. When less than 40 ppm NO was
decomposed, nearly all of the NO was converted to NHi; the conversion ratio
decreased as the amount of decomposed NO increased. This indicates that the
MACT process may work more efficiently with flue gas richer in NO.
The reactions which take place during complete combustion by air addition
are shownbelow:
Cm”fln’ t + 03 —) HaO + CO (4)
CO + Oz —> COi (5)
NBi + 02 —> Ni + RiO (6)
NHj + Oa —> NO + H:O (7)
The nitrogen hydrides, NHz and Nib, shown as Hi, may be converted to
both Ni and NO. The conversion of N ih to NO occurs at temperatures above
700°C as shown in Figure 6—23; the conversion ratio is higher with a higher Oi
concentration in the gas.
441
-------
NO DECOMPOSED (ppm)
Figure 6-22
Figure 6.23
Conversion of NO to NHI by
Secondary Fuel Addition.
REACTION TEMPERATURE (°C)
Conversion of NHI to NO by Air Addition
for Complete Combustion.
E
C.
a
w
0
U-
z
0
200
80
0
I-
z
0
(1)
w
>
z
0
0
=
0.1 %
900
1300
442
-------
6.4.4 Pilot Plant Tests with Corner—firing Fnrnaces(7 )
As shown in Table 6—14, pilot plant tests of MA T were conducted with two
corner—firing furnaces. The operating conditions and the coals used for the
pilot plant tests are shown in Tables 6—15 and 6—13, respectively.
TABLE 6—14. FURNACE SPECIFICATIONS
Fuel Gas and Oil Coal
Beat capacity (10’ ca1/hr) 1.5 2.0
Number of burners 12 (gas), 8 (oil) 12
Forced draft fan (m 3 /min) 450 (20°C) 920 (20°C)
(mmHsO) 800 800
Induced fan (m 3 /nin) 650 (300°C) 1140 (1000C)
(mmH2O) 200 800
Gas recirculation fan (m Imin) 427 (500°C) 760 (350°C)
(mmBaO) 400 600
Air preheater (Nm’/min) 430 (288°C) 750 (350°C)
TABLE
6—15. MACT PILOT PLANT TEST CONDITIONS
Fuel
Gas
Oil
Coal
Main burner
Type
Fuel (kg/hr)
Gas recirculation
Over fire air ( )
(%)
PM
240—800
0—30
0—20
PM
200—800
0—20
0—20
SGR
600—1200
0—10
0—20
Secondary burner
Fuel (kg/hr)
Gas rectrculation
(%)
0—150
0—15
0—150
0—15
0—150
5—15
Additional air (Ta)
0—20
0—20
0—20
Oi at furnace outlet (%)
1—3
1—3
2—4
443
-------
The furnace was operated manually. As shown in Figure 6—24, furnace
operation stabilized three to five minutes after the use of MACT began.
Figures 6—25 to 6—27 show the relationship between Os concentrations in
the flue gas (prior to the injection of the secondary fuel) and NO 1 concen-
trations with and without MACT. Using a combination of combustion modifica-
tion and MACT, NO was reduced to 7—10 ppm for gas, 40—50 ppm for oil, and 50—
70 ppm for coal. Figure 6—28 shows the effect of MA4T for oil—burning when
the nitrogen content of the oil was increased to 0.9% by adding pyridine. NO
was maintained at 70 ppm even for oil containing 0.9% N.
An analysis of both the flue gas and the fly ash leaving the HACT system
indicated that MACT did not cause any degradation in their composition and
properties.
6.4.5 Evaluation
The MACT in—furnace process requires the same investment and operation
expenditures as the process of combustion modification by staged combustion.
By combini ng combustion modification and MACI’, NO 1 in flue gas from a large
utility boiler may be reduced to about 100 ppm for coal, 40 ppm for oil, and
20 ppm for gas. Process operation appears to be problem—free.
A disadvantage associated with MACT is that it requires a furnace that
is about 10% taller than most conventional furnaces. For this reason, MACT
may not be applied to most existing boilers without a reduction in boiler
capacity. An exception to this is the type of the boiler used commercially by
Tokyo Electric. This boiler was originally designed for burning coal and oil
and has a very large furnace for oil—firing.
The MACT process may be useful for a new boiler, especially for a coal—
fired boiler in which NO 1 reduction to 100 ppm is needed. For 50% NO 1
reduction, the MACT may be more advantageous than selective noncatalytic
444
-------
CHANGE
CM CMANDMACT
30 -— 02
1 0 -
20_O ______
Z 20
30- N0,
N 40-
o so ______ N0
GAS
2C , hhhhu,thhlhhlthhhhj’h1
225 CHANGE
02- CM I I CMANDMACT
03- 200
04 N0
0 : 115 ——-—--—--—-———-—-.,.
— ‘a 1 \\
50 NO
lOlL I
25
2,0: I I I I I I
CHANGE
2 CM CMANOMACT
02,
300 -
0 NO
5.
N
O 6 200
100 [ J - - - —-- --
1 M1N
__a,
TIME
Figure 6-24 02 Percentage and NO Concentration After Combustion Modification (CM)
and CM Plus MACI in a Corner.Firung Furnace Using Three Fuels
445
-------
100
go
80
70
60
50
40
30
20
15
10
9
8
7
6
1 2 3 4
02 (¼)
GA GAS RECIRCULATION
OFA OVER FIRE AIR INJECTION
0 • —
—
COt _ 0
— 0 p
—
0
0
0
z
Figuce 6-25 Resuita of MACT Pilot Plant Teats with
Pvopan. Fuel. (624 kglhr)
140
120
100
0 1 2 3 4 5
02 (¼)
FIgure 6-26 ResultS of MACI Pilot Plant Teat witfl
Heavy B Oil Fuel. (N 0.12% 800 lThr
446
-------
250
200
0
° 150
S
100
0
z
50
02 (¾)
Figure 6-21 Results of MACT Pilot Plant Tests
with Coal Fuel (1150 kg!hr)
* 0 w58 2¾ before the addition of the secondary Fuel
Figure 6-28 Resulte of MACT Pilot Plant Tests with
Oil ContainIng Pyndlne
‘ 1 2 3 4 5 6 1
BASE (1)
(OFA = MIN)
GA MIN)
BASE (2)
0
a
E
0
z
(GA = 29 ’/o)
N IN OIL. (¾)
447
-------
reduction (Thermal DeNOx) because MACT does not require ammonia and there—
fore does not have an ammonium bisulfate problem. For additional NO abate-
ment, selective catalytic reduction (SCR) must be used with MACT.
IHi, Babcock Hitachi (BR) , and Tokyo Electric have tested a process
similar to MACT and have obtained about 50% NOx removal. In addition,
Hitachi Zosen has developed a similar process for industrial boilers (Section
6.9.5) (8). The MACT process may be best suited to corner—firing (tangential
firing) boilers. It is based on rapid uniform mixing of the secondary fuel
with combustion gas which can be attained easily in that type of boiler.
6.5 ACTIVATED CARBON PROCESS (EPDC — SRI PROCESS)
6.5.1 Introduction
EPDC and Sumitomo Heavy Industries (SRI) have conducted tests of the
activated carbon process with the objective of removing 95% of SO and 40%
of NO in flue gas from a coal—fired boiler. They have operated a pilot
plant with a capacity of treating 10,000 Nm 3 /hr of flue gas from a coal—fired
boiler at £PDC’s Takehara Station since November 1978.
In the activated carbon process, the SOz adsorbed by the carbon is regen—
erated by heating and converted to elemental sulfur by the R.ECLAUS process.
In the RE LAUS process, SOz is reduced to 112S and S using coal, and then the
HaS and the SOa react in the Claus furnace to form byproduct elemental sulfur.
Ammonia is added to the flue gas to: 1) increase SO removal efficiency,
2) reduce carbon consumption by thermal regeneration. and 3) remove a por-
tion of the N0 .
An activated carbon process demonstration plant sponsored by the Ministry
of International Trade and Industry will be completed in 1983. The plant will
be located at EPDC’s Matsushima Station and will have a capacity of treating
300,000 N&/hr of flue gas from a coal—fired boiler.
448
-------
6.5.2 Pilot Plat t Test of the Activated Carbon Process (9 )
The composition of the flue gas and the pilot plant test conditions at
EPDC’s Takehara Station are shown in Tables 6—16 and 6—17.
TABLE 6—16. FLUE GAS COMPOSITION DURING PILOT PLANT TEST
Range
Average
SOa, ppm
600—1000
900
NOT, ppm
200—400
250
O , %
5.0—8.9
6.6
CO 2 . %
12.0
12.0
HzO, %
8.1—9.0
8.4
Dust, mg/Nm 3
140—310
250
TABLE
6—17.
TEST CONDITIONS
S02, ppm
NO, ppm
Gas temperature, °C
Space velocity (SV),
hr 1
600—1000
200—350
120—150
500—900
A flowsheet of the pilot plant is presented in Figure 6—29. Ammonia is
added to the flue gas in the ratio of 0.5 mol ammonia/mol SO 1 and the gas is
sent horizontally through a moving bed in which granular activated carbon
moves downward. The SO 1 in the gas is adsorbed by the carbon to form sul-
furic acid:
SO 2 + 1120 + 1/202 = H1S04 (1)
S03 + 1130 = fl2S04 (2)
A portion of the ammonia reacts with the sulfuric acid to form ammonium
bisulfate while the rest of it is reacted with NO by the catalytic action of
the carbon:
449
-------
BOILER EXHAUST
STEAM
I
r
0
BOILER
FEED
AC lISA II
CARRO ll
ACTIVATED CARROll COAL
GAS
— — — — BOILER FEED WATER STEAM
SULFUR
SULFUR
PUMP
FIgure 629 Flowsheet Ioi the Activated Caibon Process Pilot Plant at EPOC
Iskehara STation (9)
-------
NH3 + HBSO4 = NH4HSO4 (3)
4N + 4N0 + 02 = 4Nz + 6H20 (4)
The relationship of reaction temperature to removal efficiency in the
system is shown in Figure 6—30. The flue gas is then sent through an ESP or a
baghouse and on to the stack. Loaded with ammonium bisulfate and sulfuric
acid, the carbon is discharged from the bed and sent into another moving bed
reactor (regenerator). There it is heated above 350°C by an inert gas pro-
duced by the incomplete combustion of fuel. The treated carbon is returned to
the adsorber and the SOz is regenerated by the following reactions:
2HBSO4 + C = 2SOa + 2flz0 + COB (5)
2NE4HSO4 + HiSO. = 3S02 + 6Hi0 + Nz (6)
The SOB—rich gas leaving the regenerator is ini.xed with a small amount of
air and sent to the bottom of the coal reactor, a shaft furnace charged with
coal. As a result of the partial combustion of the coal, the gas temperature
in the reactor reaches about 9000C. This causes the following reactions:
C + H O = a 2 + CO (7)
HB +S=HzS (8)
C+ SOB S+COa (9)
co+ S= COS (10)
About 80% of the SOB is reacted in the furnace; about 40 forms sulfur,
30% forms HiS, and 10% forms COS. The gas is then cooled to allow condensa-
tion of elemental sulfur, and is sent into two Claus furnaces in series. In
the furnaces, nearly all of the sulfur compounds are converted to elemental
sulfur:
2HzS + SOi = 3S + 2HzO (11)
COS+H:O=H iS+COa (12)
451
-------
90
80
70
50
40
30
20
10
0
0
-j
>
0
w
0
C ’)
0
0
-J
>
0
w
0
z
TEMPERATURE (°C)
FIgure 6-30 Effect of Temperature and Gas Volume on SO and NO
Removal Efficiencies. (NH 3 !SO = 0.5)(1O)
160
GAS VOLUME
(RATIO)
SO NO
(ppm) (ppm)
100
80
120
910 230
910 170
910 200
I I I
120 140
C
452
70A2856
-------
In actuality, a small amount of air is added to the gas leaving the coal
reactor in order to adjust the EzS/SOz ratio to 2. The gas leaving the Claus
furnace (containing small amounts of HiS, SOi and COS) is incinerated to
convert B2S and COS to SOi and is returned to the adsorber.
Figure 6—31 shows the S03 and NO removal efficiencies during an 8,000—
hour continuous test cycle at the pilot plant. The SO removal efficiency
ranged from 80 to 99% (90—95% removal for most of the time) while the N0
removal efficiency ranged from 20—40% (30% average). The activated carbon was
consumed at a rate of 0.5—1% per day. Ammonia was not detected in the flue
gas leaving the adsorber as long as the NH3/SOx mole ratio was kept below
0.5. The recovered sulfur contained a very small amount (less than 0.5%) of
impurities such as NR 4 RSO4 and NH4C1; the impurities subsequently were removed
by filtration to yield a byproduct sulfur with over 99.9% impurity.
6.5.3 Evaluation
The activated carbon process is advantageous in that is does not require
either water or gas reheating, while it simultaneously removes over 90 of
SOx and about 30% of NOx, and produces elemental sulfur. On the other
hand, the process does consume a large amount of carbon even with the addition
of ammonia. The cost of carbon may be one of the important factors to consi-
der in determining the commercial applicability of the process.
The NO removal efficiency of the process can be increased to 80—90% by
using a much larger amount of ammonia and a high quality carbon at 230°C.
This was demonstrated by the Unitika process pilot plant. The tinitika pro-
cess, however, was abandoned because of its large consumption of costly carbon
and ammonia. It may be possible to use less costly carbon and smaller amounts
of ammonia for about 30% NOr removal at a lower temperature.
453
-------
OPERATION PERIOD (hi)
FIgure 6-31 SO and NO Concentrations and Removal Efficiencies During an 8000 Hour
Continuous Test Cycle (9)
-j
>
0
U I
E
a
a
z
0
I-
I—
z
U I
0
z
0
C-)
I-
U i
-J
z
U i
f t-j
4000
5000
7000
-------
The activated carbon process may not be suitable for flue gas with a high
SO 1 content because both ammonia and carbon consumption increase with SO 1
concentration; it may be well suited for flue gas from coal or oil with less
than about 1% sulfur. The operation of the demonstration plant now under
construction at Matsushima should provide more valuable information on the
process. Sf1 also conducted tests on the feasibility of using 2 absorbers to
remove about 80% of NO 1 and 90% of SOs. Further study is needed to evaluate
this process.
6.6 FBARA ELECTRON BEAM IRRADIATION PROCESS
6.6.1 Introduction
Ebara Corporation, and the Japan Atomic Energy Research Institute, have
developed a process which simultaneously removes SOx and NO 1 from flue gas
by irradiating the gas with an electron beam and then collecting the resulting
particulates with an ESP. From 1974—1977, Ebara operated a pilot plant with a
capacity of treating 1,000 Nm 3 /hr of flue gas from an oil—fired boiler. Tests
at this early stage of development were not very successful. Later Ebara
began using ammonia, which substantially improved the process.
By irradiating the gas containing ammonia. over 90% of SO 1 and 80% of
NO 1 can be removed, and converted to fine crystals of ammonium sulfate
nitrate (double salt) which can be used for fertilizer. A larger pilot plant
with a capacity of treating 3,000 to 10,000 Nm /hr of flue gas from an iron—
ore sintering machine was operated from July 1977 through June 1978 at Nippon
Steel’s Yawata Works. The Ebara electron beam irradiation process has been
licensed to AVCO, USA.
6.6.2 Ebara Process Pilot Plant Tests (11, 12)
A flowsheet of the larger Ebara process pilot plant is shown in Figure
6—32. The most important pieces of equipment in the plant are: gas coolers
for indirect cooling (heat exchanger) and direct cooling (water spray), an
ammonia injection facility, an electron beam reactor surrounded by a concrete
455
-------
WAIER-
WA I (fl —
NI4 SUPPLYING
ROOM
U i
FIgure 6 32 Flowsheet of an Ebara Electron Beam Process Pilot Plant
(3.000 10,000 Nm 3 lhr)
-------
wall to prevent x—ray emissions, an ESPI and a control system. As shown in
Figures 6—32 and 6—33, two electron beam accelerators with a unit capacity of
750 kV, 60 mA were placed at opposite ends of a cylindrical reactor with a
diameter of 2.6 m. The distribution of the dose rate in the reactor is shown
in. Figure 6—33.
Since the dose rate was not n.niform throughout the reactor, an impeller
was placed at the reactor inlet to rotate the gas and homogenize the reaction.
A side stream of the gas from an iron—ore sintering furnace was treated in the
reactor. The gas contained about 200 ppm each of SOi and NOx (mainly NO),
40 mg/ 3 of dust, 15.5% Oz, 10.3% HzO, and 16.2% COi, and had a flow rate of
3,000—10,000 Nm 3 /hr. It was cooled to 70—90°C, injected with ammonia, and
sent into the reactor. In some of the tests, SO 2 , NO, and fly ash were added
to the gas in order to simulate flue gas from coal. Ammonia was used at an
average stoichiometry of 1.0 (2 mol NB3/mol SO 1 and 1 mol N113/mol NO 1 ).
For some of the tests, this stoichometry was changed.
The relationship between the gas rotation ratio and the NO 1 and SO 1
removal efficiencies is shown in Figure 6—34. The gas rotation ratio is
defined as:
revolutions of impeller (rym )
gas flow rate (Nm 3 /hr)
A maximum efficiency——over 95% for SO and 80% for N0 1 —was obtained at
the ratio of about 1/60, with 1.8 Mrad at 90°C.
The effect of the dose on removal efficiency and unreacted NH3 is shown
in Figure 6—35. SO 1 removal efficiency exceeded 90% at 1.3 Mrad at 70°C and
at 2 Mrad at 90°C, while NO 1 removal efficiency reached a maximum ———— about
80%—with 1.8 Mrad at 70°C or about 65% with 1.8 Mrad at 90°C. The decrease in
NO 1 removal efficiency at the higher dose level may be due to the formation
of NO 1 from N3 and Oz. The amount of unreacted Nils decreased with an incre-
asing dose and was almost zero above 1.8 Mrad at 70°C; there was a consider-
able amount of unreacted NHs at 90°C even with 2.4 Mrad. The tests shown in
457
-------
(HOAIZONTAL CROSS SECTIONI
ELECTRON BEAM
L i
FIgure 6 33 Dose Rate Distribution In Ebara Process Pilot Plant Reactor
* accelerating voltage = 600 kV
(VERTICAL CROSS SECTION)
550 fl50
Ui
1 1
14
I .- -J
I— uJ
ELECTRON BEAM
*
-------
GAS ROTATION RATIO
REVOLUTIONS OF IMPELLER (rpm )
GAS FLOW RATE (Nm 3 /h)
Figure 6.34 Relationship Between Gas Rotation Ratio and SO
and NO 1 Removal Efficiencies.
0
(1)
z
0
z
U-
0
-J
0
UJ
( rpm •
GAS ROTATION RATIO )
459
-------
E
0.
A
150 3:
z
I -.
An Co
I
x
1u w
DOSE (Mrad)
*
Figure 6.35 Effect of Dose on Removal Efficiencies and Exhaust NH 3
* gas rotation ratio 1/33
0
(I ,
0
z
0
z
Ll
0
-J
>
0
uJ
3:
460
-------
Figure 6—35 were conducted at a gas rotation ratio of 1/33 units. Better
results may be obtained with a ratio of 1/60.
Figure 6—36 shows the results of 600 hours of pilot plant operation at
1.5 Mrad and 90°C with a gas rotation ratio of 1/33. During this test period
the pilot plant was shut down twice due to the shutdown of the the sintering
machine. The inlet SO 1 concentration ranged from 180 to 250 ppm; the inlet
NO 1 concentration was in the range of 170 to 210 ppm. The SO 1 and NO 1
removal efficiencies ranged from 92 to 979, and 75 to 85% , respectively, while
unreacted Nib leakage at the reactor outlet ranged from 10 to 50 ppm. This
high level of unreacted ammonia was caused by the frequent fluctuations in
inlet SO 1 and NO 1 concentrations and by problems with the ammonia flow
meter.
The balance of NO 1 , Nib, and SO x in three pilot plant test runs is
shown in Table 6—18. Most of the NO 1 , Nib, and SOx were reacted and
caught by the ESP, but a large portion of the Nib and SO 1 did deposit in the
duct and a considerable amount of NO 1 was emitted from the recovery system.
The total amounts of Nib and SOx were slightly larger than 100% presumably
because of a sampling and analysis error. The NO 1 totals were considerably
larger than 100%. which indicated that irradiation was causing the formation
of NO 1 from Ni and Oi. Chemical analysis of the exhaust gas revealed the
presence of NiO in concentrations ranging from 10 to 20 ppm.
TABLE 6-18. BALANCE OF NO 1 , N113 AND SO 1 (PERCENT OF INLET CONCENTRATION)
Run No.
Component
Caught by ESP
Deposit in Duct
Emitted
Total
NO
81
6
49
136
1109
NH
SO
I
64
62
36
43
9
10
109
115
1111
NO
NH
SO
I
87
65
64
6
36
40
31
10
8
124
111
112
1114
NO
NH
SO
1
78
65
59
6
36
43
38
7
8
132
105
110
461
-------
JX bOOOOOOOOOOOOO ?NO
80-
0
60-
40
20 r
100 200
A- AAA A8-A A A 00 A AA A fl
oOOoo oP% O 0 0 0
p-A SO
4 4—•
SHUTDOWN OF SINTERING PLANT SHUTDOWN OF SINTERING PLANT
GAS FLOW RATE. 3000 Nm 3 Ihr
DOSE 1 5 Mrad
TEMP 80C
NH 3 INO mole ratio 1 0
GAS ROTATION RATIO 1/33 rpm.hrlNm 3
zV v
300 400
500
HOURS OF OPERATION
Figure 8-36 Results of 600 Hours of Continuous Ebara Process Pilot Plant Tests.
0
-Jo
0
oz
w
(SOI l
0
z
a
z
‘-cc
u )
z
0”
ADJUSTMENT
TROUBLE WITH NH 3
ADDITION SYSTEM
E
a
a
C,
I
z
I-
U)
D
4
I
3 <
U i
600
70A2873
-------
During the test period, the titanium foil at the reactor window was often -
broken. Later, a foil with a titanium alloy was used which lasted for over
two months. Ebara has developed a method of changing the foil without stop-
ping the operation of the accelerator.
An X—ray diffraction analysis of products caught by the ESP indicated
that the major compounds present were 2NE 4 NO 3 (NH4)2SO4, 3NTI4N03 NH4)
3S04 and (NH4)3S04. Chemical analysis of a product obtained from a gas con-
taining 220 ppm SOi and 235 ppm NO 1 with 700 ppm NH3 at 90°C revealed that
the product contained 33.6% S04, 29.4% N03, 23.6% N113, and 0.9% C.
The results of a pilot plant test using SO 1 — and N0 1 —rich gas are
shown in Figure 6—37 (13). About 80% of both SO 1 and NO 1 were removed at
1.8 Mrad with an ammonia stoichiometry of 1.0. With these conditions un—
reacted ammonia was about 50 ppm; the gas rotation ratio was poor (1/17)
because of a limited ammonia supply and a flue gas volume of only 1,500
Nm 3 /hr. It is probable that about 90% of both SO 1 and NO 1 can be removed
with a better rotation ratio, as shown in Figure 6—34.
For sbme of the tests, 0.1 — 0.5 g/Nm 3 of fly ash (equivalent to 95—75%
removal of fly ash from pulverized coal) were added to the system. The re—
suits indicated that the fly ash produced no adverse effect on the SO 1 and
NO 1 removal efficiencies (13).
6.6.3 Reaction Mechanism
Fundamental tests also were conducted on the reaction mechanism of the
Ebara process. Although the reactions are complex, the major one appears to
be the formation of oxidant radicals such as OH caused by the radiation of
water vapor. In this reaction SOi and NO are oxidized to form sulfuric acid
and nitric acid mists; the mists react with ammonia to form fine crystals of
sulfate and nitrate. Tests results also showed that the SOi removal effici-
ency is slightly higher when the gas is cooled by a water spray rather than by
a heat exchanger, the NO 1 removal was not significantly different with the
463
-------
x
0
z
z
x
0
U)
I L.
0
-J
>
0
w
Figure 6-37
E
0
C.,
I
z
I-
C l )
I
w
Removal of SO, and NO vs. Amount of Ammonia Added.
AMOUNT OF NH 3 ADDED
(RELATIVE TO STOICHIOMETRIC)
0.9 10 1.1 12 13
464
-------
two types of cooling. Presumably the fine water droplets in the spray combine
with S03 to form sulfuric acid mists which promote the formation of SOs from
SOB.
6.6.4 Ebara Process Commercial Plant Assumptions
Electron beam accelerators with capacities of 100 — 150 kW have been used
commercially for various purposes without problems. A larger accelerator with
a capacity of 500 Id? will be commercialized in Iapan in the near future. This
500 kW accelerator can treat 60,000 to 70,000 Nm 3 /hr of flue gas. Since it is
advantageous to use at least 2 accelerators with a reactor, an economical
minimum capacity for a reactor may be 120,000 to 140,000 Nin 3 /hr, which is
equivalent to 35 to 50 MW.
Figure 6—38 shows a layout plan for a commercial Ebara process plant with
a capacity of treating 1,0001000 Nm 3 /hr of flue gas from a utility boiler
(290 MW for coal and 350 MW for oil). The flue gas contains 60 ppm NO 1 ,
1,000 ppm SOx, and 100 mg/Nm 3 of dust. In such a plant the 150°C flue gas
is cooled by a heat exchanger to about 110°C and then indirectly cooled to
70°C with water tubes. The cooled gas is mixed with ammonia and treated by
two reactors in parallel (diameter approximately 4 m), and sent through an
ESP.
A total of 14—16 accelerators (500 kW each) may be used to remove 80% of
NO 1 and 90% of SOS. The treated gas will probably contain 120 ppm NO 1 ,
and 100 ppm SO 1 , and less than 20 ppm unreacted NE3 with about 15 mg/Nm 3 of
dust.
Ebara estimated the investment cost of the electron beam process to be
close to that of a conventional wet FGD system, provi.ded that a large acceler-
ator (500 Id?) is available at a reasonable cost. The estimated utility
requirements for the 1,000,000 Nm /hr plant are shown in Table 6—19. The
electron beam system’s total power consumption of 10.000 kWhr accounts for 2.9
to 3.4% of the power generated by the power plant. This power èonsumption is
465
-------
PLUE G# S
1 ’INOUCTION FAN
HEAT EXCHANGER
(OPTION
L 1 . 90 METERS
QUENCH
TOWER
-
ELECTRON
ACCELERATOR
REACTOR
CONVEYOR
DOUBLE EXHAUST
VALVE
Figure 6-38 Layout Plan for a 1.000.000 Nm 3 Ihr
Commercial Ebara Process Plant
466
-------
TABLE 6-19. ESTIMATED UTILITY REQUIREMENTS PER HOUR ESTIMATED FOR EBARA
PROCESS SYSTEM APPLIED TO A UTILITY BOILER
(FLUE GAS 1,000,000 Nm 3 /hr, 350 MW FOR OIL, 290 MW FOR COAL)
Electron beam accelerator 6,900 kWhr
Power Other 3,100 kWhr
Total 10,000 kWhr
Ammonia vaporization 1 ton
Steam Soot blowing 1 ton
Total 2 ton
Gas cooling 33 ton
Water Other 1 ton
Total 34 ton
Ammonia 0.94 of stoichiometric amount 2 ton
TABLE 6-20
(Cost estim
90% removed
. COST COMPARIS
ates established
; NO = 620 ppm,
ON OF FLUE GAS TEE
by Gibbs Hill,
90% removed; lime
ATMENT PROCESSE
Inc. , New York.
scrubber does
S (500 MWe) (13)
SO 2 = 1900 ppm.
not remove NO.)
Lime
AvcofEbara
Scrubber
Total capit
al cost IM
43
60
11kw
86
120
(1/MWh)
(llMWh)
.
Depreciatio
Variable op
a
a and overhegd
erating cost
2.4
2.5
3.4
2.8
Reheat cost
0
1.0
Total operating cost
Byproduct charge/(credit)
4.9
(0.6)
72 d
0.4
Net operati
ng cost
4.3
7.6
Total first
year cost
*14M
126M
80% load factor; annual cost = 20% of total capital.
Ammonia at 1 130/ton.
Sold as agricultural fertilizer at 1 15/ton.
“Does not include cost of disposal site and preparation.
467
-------
slightly more than to that of a combination of S R and wet FGD. S R requires
about 0.2% and FGD consumes 2 to 2.5% (excluding gas reheating) of the total
power generated by the plant.
In Table 6—20 estimated costs for the AVCO/Ebara process are compared
with the costs of lime scrubbing. The cost of the AVCO/Ebara process assumes
the use of simple water spraying for gas cooling and is considerably lower
than the cost of lime scrubbing.
6.6.5 Evaluation
The Ebara/AVCO electron beam irradiation process is unique and has the
following advantages:
(1) SO 1 and NO 1 are removed simultaneously.
(2) Since it is a dry process, wastewater treatment is not necessary.
(3) Ammonia, SO 1 and NO 1 are recovered as useful byproducts.
(4) The process is simple; the only problem so far has been the need
for occasional renewal of the foil at the accelerator window.
However, the process does have the following disadvantages:
(1) For treating flue gas from coal, a dust removal facility is needed
upstream of the reactor in order to obtain a useful byproduct.
(2) Gas cooling to about 70°C is needed to obtain a high N0 1 /S0 1
removal efficiency.
(3) Large amounts of N113 and NzO may be emitted for high removal effi-
ciencies from S0 1 —rich gas.
468
-------
Fly ash from coal has been used as fertilizer to supply micronutrients
such as zinc and boron to certain soils. For this reason small amounts of fly
ash in the Ebara process byproduct ammonium may be acceptable but large
amounts decrease the compounds’ fertilizer value.
Using a water spray for cooling is simpler and produces a higher SOz
removal efficiency than indirect cooling. The gas can be cooled to 700C
without producing any wastewater. Unfortunately water spraying may cause the
ammonium compounds to absorb moisture when a large amount of spray is used and
the moisture content of the gas also exceeds the critical humidity for the
compounds.
The amount of unreacted ammonia from the Ebara reactor may be reduced
with improved control of ammonia addition. This can be achieved by measuring
the concentration of outlet NE3 as well as that of inlet SOx and NOT. The
formation of N20 may be also reduced, but cannot be prevented. A substantial
concentration of NzO occurs naturally, so the emission of a small additional
amount will not produce any adverse environmental effect. Nonetheless, fur-
ther study is necessary to determine ways of reducing N20 formation.
Table 6—18 shows that a considerable portion of the byproducts deposited
in the system ductwork. This deposition may be eliminated by increasing the
gas velocity through the duct.
Although there are many simultaneous SO 1 /NO removal processes, the
Ebara process is the only one that may be commercially applicable for byprod-
uct recovery of SOx, NOx, and NH3. Additional tests using a larger pilot
plant with gases of various compositions will be necessary for definite proof
of this commercial potential.
469
-------
6.7 MKK POTASSIUM-EDTA WET SIMULTANEOUS S0 1 /N0 1 REMOVAL PROCESS
6.7.1 Introduction
The potassium—EDTA wet process simultaneously removes SO and NO 1
from flue gas by using a solution containing potassium salts, ferrous sulfate 1
and EDTA (ethylenediamine tetraacetic acid). NO 1 is converted to Ni and
SO 1 to elemental sulfur. The process was developed by Mr. Y. Kobayashi who
was associated with Mitsubishi Kakoki Kaisha (MXX) and is now with Ritachi
Zosen ( l IZ). The process resembles the Kureha KDSN process which also uses
potassium salts, ferrous sulfate, and EDTA to convert SO 1 to elemental sul—
fur(4). Each process has been studied independently.
The MKX process is a combination of several commercially proven steps
plus a few new steps; bench scale tests have been conducted on the new steps.
6.7.2 Process Description
Figure 6—39 presents a flowsheet of the MKX process. Flue gas is sent
into an absorber and treated with a solution containing KUCO3, KZCO3, FeSO4,
and EDTA. About 99 of SO 1 and 9OTo of NO 1 can be absorbed by the follow—
ing reactions:
KHCO3 + SOz —> K:nS03 + COZ (1)
KSCO3 + SOa —> K2S03 + COi (2)
K2SO3 + SOi + HaO ———) 2KHSO3 (3)
2N0 + KBSO3 + 4KHSO3 ———> 2NU(SO3K)z + K2SO4 + RiO (4)
2KHSO3 + 1/2O —> KzSIO6 + EzO (5)
EDTA and FeSO4 promote the absorption of NO by reaction (4) to form
potassium imidodisulfonate (Nll(SO3K)z). K2SO4, KZSIO6 and NB(SO3K)z are
crystallized in the liquor, separated by a thickener and filter, and then
heated in a two—stage flash dryer. The heater uses hot gas obtained by incin-
eration of a portion of the HiS that was produced by the reduction of the
470
-------
—4
i - I
0- Y
i D a
- — — — - GA5STREAM
so, IS— I
INCIN EHA 1 O ” sul ION
I I ECOVE N ?
.I ,OUIDS 1NEAM - 5OLTDST NEAM
FIgure 639 F loweheel ci the MKK Poleselum—EDTA Wet SImuIlenOocs Removal Process
-------
potassium salts. The first stage of heating is for drying. During the second.
stage at 3700C, KiSzO is decomposed to KZSO4 and SOz:
KzSzO ———) KiSO + S02 (6)
The SOi contained in the hot gas, as well as that released from the
KZSZO6, is absorbed by a solution containing KaCO, and KHCO3 (obtained by the
carbonation of KaS)
KZCO3 + SOa ——> KaSO8 + COs (7)
KBCO3 + SOi —> KHSOa + COa (8)
The liquor leaving the SOa absorber contains KflCO3, KZCO3 and KHSO3 and
is mixed with EDTA, KO11, and FeSO4, and sent into the main absorber. In the
main absorber SO 1 and NO 1 are removed simultaneously by the 50_600C liquor
(pll 6.8—7.0). A large (S03 + 11S03)/NO ratio is needed in order to attain a
high NO removal efficiency. Incineration of a portion of the recovered HiS
increases both the ratio and the NO removal efficiency.
The drying process produces a solid consisting of KZSO4 and
NB(SOiK)z. The solid is caught by a cyclone, jet milled into a very fine
powder of about 5gm, and then sent into a reduction furnace. In the furnace a
1,400—1,500°C hot gas instantaneously reduces the powder to KiS at 800—900°C.
The reducing gas is produced by the partial oxidation of fuel oil by oxygen
and steam:
12S04 + 4112 = KiS + 41120 (9)
KZSO4 + 4C0 = KiS + 4COi (10)
NH (SO3K)l + 6.5 Ha ——> KiS + HiS + 1/2 Ni + 6HzO (11)
NH (SO3K)2 + 6.5 CO ———> KiS + HiS + 1/2 Ni + 6COa (12)
KaS is dissolved in water and the solution is sent to the No. 1 carbona-
tion tower where it is treated with COi at 80—90°C (pH 9.5—12.5):
472
-------
KzS + i0 + COs —> KHCO3 + KHS (13)
2KaS + EsO + COs —> KsCOa + KITS (14)
The resulting solution is carbonated at 40—60°C (pH of 8—9.5) in the
No.2 tower:
KHS + COz + HsO ———) KRCOi + HiS (15)
In actual practice, 3 to 4 towers in series may be needed to attain a
high carbonation efficiency.
As it leaves the tower, the liquor contains K2CO3 and HCO3. Next it is
sent into the S02 absorber; the gas (containing HaS and COa) leaving the tower
is sent into an HiS absorber, where uS is selectively absorbed by an amine
solution at 40°C.
The gas which leaves the HaS absorber containing COa is returned to the
No. 2 tower. The liquor leaving the HiS absorber is sent to the HaS stripper,
where steam is injected to vaporize fl S. Most of the HaS is sent to a Claus
furnace where elemental sulfur is recovered and a portion of H 2 S is inciner-
ated. This produces a hot gas for the heating process. The amount of HaS
incinerated is adjusted according to the inlet Oz, SOs, and NO 1 concentra-
tions and the required NO 1 removal efficiency.
6.7.3 Requirements and Costs
Table 6—21 shows the amounts and costs of the estimated raw materials and
utilities needed for an MKX system designed to treat 150°C flue gas from a
1,000 MV! coal—fired boiler containing 4.5% Oi, 900 ppm SOS. 300 ppm NOR,
50 ppm HC1, and 100 mg/Nm 3 of fly ash. The system will remove 99% of SO 1
and 80% of NO 1 .
473
-------
TABLE 6—21. UTILITY AND MATERIAL REQUIREMENTS AND COSTS OF AN MIX PROCESS
SYSTEM FOR A MW COAL—FIRED BOILER (1980) (For 1000 MW hr)
Consumption
Unit
Cost Cost (1000 Yen)
EDTA
FeSO4
XOH
High S oil (asphalt)
Low S oil
Os (95%)
Power
Industrial Water
Cooling water
(sea water)
Others
282.9 kg
320 g
188 kg
8690 1
3450 1
6400 N&
29100 Kwhr
170 t
1250 t
600
5
150
50
57
9
10
2
yen/kg
yen/kg
yen/kg
yen/i
yen/i
yen/Nm 3
yen/Kwhr
yen/t
yen/t
149.7
1.6
29.0
434.5
196.6
75.5
261.9
1.7
2.5
8.0
TOTAL
1160.1
Recovered S
4280 kg
22
yen/kg
94.2
It is estimated that such a system produces 700 kg/hr of sludge contain-
ing fly ash as well as 2,030 kg/hr of wastewater. The wastewater is needed to
prevent the accumulation of chlorine in the system.
The investment cost is estimated at 18.8 billion yen or 18,800 yen/kW.
The annual cost for 7,000 hours full load operation is shown in Table 6—22.
The relationship between the annualized treatment cost and the inlet Os
percentage, NO 1 concentrations and NOx removal efficiency is shown in
Figure 6—40. The treatment cost Increases sharply with an increase In the
inlet concentrations and the removal ratio. An increase in the inlet SOs
concentration, unlike a similar increase in the inlet NO 1 concentration,
does not have a significant effect on the cost. Over 99% of SOs is absorbed.
even with 75% NO 1 removal. Although an increase in SOa necessitates the
treatment of a larger amount of potassium salts during the reduction and
carbonation steps, a high SOs/NOx ratio favors NO 1 removal and reduces the
amount of H z S needed to produce SOs. It is also estimated that when the HC1
concentration increases by 10 ppm, the cost subsequently increases by 0.037
yen/kWhr due to the larger amount of wastewater which must be treated.
474
-------
0
C
0
C
0
>-
(0
0
C.)
z
w
I-
U i
2 3
22
21
20
1 9
1 8-
1 7-
1 6-
1 5
1.4-
o 1 2 3 4
02 CONCENTRATION IN FLUE GAS (%)
Figure 6-40 Annualized Cost for Vaflous Operating Conditions In an MKK Process System.
/
CAPACITY 1000MW
HCI S oppm
SO 2 90 0ppm
SO 2 REMOVAL EFFICIENCY 99%
7000 hrs FULL LOAD/year
INTEREST 8%/year
I
I
I
/
I
I
/
I
I
-J
I
1/
10A2878
475
-------
TABLE 6—22. ANNUAL COST OF AN MKK PROCESS SYSTEM
FOR A 1000 MW COAL—FIRED BOILER
Item
Calculation Basis
Cost (10’
yen)
(1)
Capital Cost
Depreciation
Interest
Tax
Insurance
Total
Investment Cost x 0.9 x (1/7)
1 x 0.08 x 0.55
x 0.014 x 0.55
z 0.4 x 0.003
2,417.1
827.2
144.8
22.6
3,411.7
(2)
Direct Cost
Labor
Maintenance
-
8,000 x 24 (person)
Investment cost x 0.015
192.0
282.0
(3)
Chemicals
Utilities
1,116,100 x 7000
8,120.7
(4)
Related cost
(2) + (3) x 0.08
687.6
(5)
Total
(1) + (2) + (3) + (4)
12,694.0
(6)
By—product S
94,200 x 7,000
659.4
(7)
Difference
12,034.5
ANNU
ALIZED COST
(7) 7,000,000,000 (kWh
r)
1.719 yen
/kW’hr
6.7.4 Evaluation
The NIK system is a combination of several steps most of which have been
proven to be feasible through commercial operation or pilot plant tests. One
exception is the step in which &zSO4 and NB(SO3K)2 are instantaneously reduced
to iS. Basic chemical and thermodynamic studies related to the process have
been conducted, although the total system has not been tested yet at a pilot
plant.
One advantage that potassium scrubbing has over sodium scrubbiiig is that
EzSO , H(SO3K)3, and KzSiO are less soluble than the corresponding sodium
salts, and can be separated from the mother liquor. On the other hand, for
carbonation of KzS, a larger CO2/JIZS ratio is needed than for NaiS. because
ERCO3 is more soluble than NaHCOj and cannot be crystallized out.
476
-------
The use of oxygen instead of air in the MKK system substantially reduces
fuel consumption, gas volume, and reactor size. Moreover, the loss of COz is
eliminated because it is not necessary to purge Na as it is when air is used.
Thus, the overall cost may be reduced by using oxygen.
Simultaneous SO and NO removal and the conversion of SO to elemental
x x
sulfur are other advantages of the MIX process. However, the process does
require complex equipment including an oxygen generator, Claus furnace, and
wastewater treatment system in addition to the facilities shown in Figure 6—
39. Pilot plant tests on the reduction and carbonation steps are needed to
prove the feasibility of the process. Although the wastewater treatment
system appears to be complex, it may not prove to be a serious problem because
of the small quantity of wastewater produced.
The system has a silo for solids storage and a storage tank for the
absorbing liquor. This allows the continuous operation of the absorber and
reduction furnace even when other steps in the process must shut down.
The estimated investment and operation costs for an M system are
slightly lower than those for a combined SCR/FGD system. For nore accurate
cost estimates, a pilot plant test of the entire system is necessary. Since
90% NOx removal requires a large multi—stage absorber and a large pressure
drop, 30% removal may be more practical. Because its cost increases sharply
with increases in the inlet Oi and NOx concentrations, the process may
be best suited to flue gas from high—sulfur oil rather than from coal. An
increase in the SOa content of flue gas does not significantly increase the
operation cost since it aids in NO removal.
The MIX process, as well as the Kureha KDSN process, is sophisticated and
probably will not be used commercially in the near future. The ability of
both of these processes to byproduce elemental sulfur and to remove N0
without using ammonia is an attractive feature. This type of NO /SO
removal technology should be further studied as a possible alternative to the
SQ /FGD combination. The latter may experience problems when widely used, due
477
-------
to its ammonia consumption and the difficulty associated with disposal of its
byproduct sulfur compounds.
6.8 NO OXIDATION CATALYST
6.8.1 Introduction
Almost all of the NO 1 in combustion gases is present in the form of NO.
NO is only slightly absorbed in most solutions except for those which contain
a complexing agent such as EDTA with ferrous ion. NOB, on the other hand 1 is
readily absorbed in solutions. The equimolecular mixture of NO and NO is
also fairly well absorbed. In several wet NO 1 removal processes, most of
the NO is first oxidized to NOz using chemicals such as ozone or chlorine
dioxide. However, these chemicals are expensive or cause wastewater treatment
problems.
In the air NO is oxidized slowly to NOz. As the temperature increases
this reaction rate also increases, but a high temperature lowers the NO conver-
sion ratio in equilibrium. The theoretical NO/NO 2 conversion ratio in air is
about 80 a.t 300°C and 20% at 500°C. Catalytic oxidation of NO at a rela-
tively low temperature may cause the conversion of a considerable portion of
NO in a short time period.
6.8.2 Catalytic Oxidation Tests
In Japan, H. Tominaga has conducted extensive studies of catalytic oxida-
tion of NO to N02 (13). Tominagas’ preliminary tests indicate that zeolite
impregnated with heavy metals is effective as a catalyst. He used synthetic
zeolite (molecular sieve) 13X and 13Y after the Na ion was exchanged with
heavy metal ions. Heavy metal compounds of CrC13 61120, CoCli . 6HzO,
FeC13 . 6HaO, NiCli . 6HiO, and CuCli 211i0 were used for the ion exchange.
The exchange ratio was 67% with Cu (II), 44% with Cr (III), 74% with Co (II),
and 47% with Ni (II).
478
-------
A pyrex glass tube (13 mm inner diameter and 500 mm long) was used for
the fixed bed reactor. Mixtures of NO and O , diluted with He, were passed
through the reactor after being preheated. In some of the tests, the effects
of adding HsO and S02 to the gas were studied. Two grams of catalyst between
—i
10 mesh and 40 mesh were used for an SV of 30,000 hr , fouz grams for an SV
— i
of 15,000 hr . Results of tests using 13x catalysts are shown in Figure 6—
41.
It is interesting to note that in some of the tests with the Cr (III)
catalyst the conversion ratio exceeded 70 (above the theoretical value) at
400°C in the presence of water vapor.
6.8.3 Evaluation
At 400°C or below, the catalysts tested by Tominaga are poisoned by S03.
Although the poisoning is not noticeable at 450°C, the conversion ratio at
this temperature may be too low to be useful. For this reason, the process
does not seem to be applicable to SOs—rich gas.
The process may work well for gases which do not contain SOS. With
these gases it can be used to convert about 50% of NO to NOs at 350°—400°C , in
order to achieve equimolecular absorption.
6.9 OThER PROCESSES AND PLANTS
6.9.1 Copper Oxide Process for Simultaneous NO /SO Removal (Shell
Process )
In 1973, Showa Yok.kaichi Sekiyu Co. (SYS) installed at its Yok.kaichi
Refinery, a Shell Process FGD unit which uses copper oxide as the acceptor.
In the process, SOx absorption forms copper sulfate, which is treated with
hydrogen to regenerate Cu and concentrated SOs gas. The latter is sent to a
Claus furnace for production of elemental sulfur.
479
-------
100
0
z
U-
0
z
0
>(
0
z
LU
0
LU
0
80
40
20
500
REACTION TEMPERATURE (°C)
Figure 6.41 Oxidation Ratio vs. Catalyst Type and Temperature.
60
70A2878
480
-------
Since copper sulfate is a catalyst for S R of NO 1 , ammonia has been
added to flue gas since 1975 to remove about 40% of NO 1 and about 90% of
SO 1 to meet the local regulation(l). The operability of the unit, however.
has been low except that for one year of continuous operation in 1978. SYS
does not plan to install a new unit.
6.9.2 Activated Coke Simultaneous Removal (Mitsui Mining Process )
Mitsui Mining Company has developed a simultaneous S0 1 /N0 1 removal
process which uses activated coke produced by a special coal treatment. A
pilot plant with a capacity of treating 1,000 Nm 3 /hr of flue gas from an oil—
fired boilers at 150°C has been operated to remove over 80 of the NO 1 and
about 98% of the SOz. Coal fly ash was added to the flue gas to study the
effect of the ash. The pilot plant has two reactors in series; most of the
SOi is absorbed in the first reactor by the coke. A small portion of ammonia
is fed to the first reactor, but the major portion is fed to the second
reactor where most of the NO 1 is converted to Na and RiO by reacting with
the ammonia. SOz is further adsorbed in the second reactor.
The adsorbed SOz is regenerated by a hot gas. The recovered concen—
trated SOa may be used for sulfuric acid or liquid SOa production. Hitsui
Mining has conducted some studies on a method of producing elemental sulfur
from the SOs.
6.9.3 Molecular Sieve Process for NO Adsorption
Nissan Chemical Industries installed an NO 1 recovery plant which uses a
molecular sieve produced by Union Carbide Corp. USA, to treat 22,000 Nm 3 /hr of
tail gas from a nitric acid plant. The plant reduces the NO 1 concentration
in gas from 1,000 ppm to 50 ppm (1). The adsorbed NO 1 eventually is ther-
mally desorbed and returned to the nitric acid plant. The plant was commis-
sioned in 1976 and has operated successfully for over five years without
requiring any replacement of the original molecular sieve.
481
-------
6.9.4 Sumitomo—Fuiikasni Wet Simultaneous Removal Process
SumitomoMetal and Fujikasui Engineering have developed a wet simul-
taneous removal process which uses ClOz as the oxidizing agent and NaOH as the
absorbent. In 1973 and 1974 they constructed three medium—sized commercial
plants, as shown in Table 6—2. The process can remove about 90% of both S0
and NO 1 but may not be suitable for large plants, because a large amount of
NaOff is consumed and the waste liquor containing sodium chloride, nitrate, and
sulfate must be treated (1.).
Sumitomo and Fujikasui have also developed a modified process which uses
ClOz as the oxidizing agent and CaCO3 slurry containing a catalyst as the
absorbent. In 1976 they built a demonstration plant which can treat 25,000
Nm 3 /hr of flue gas from an iron ore sintering machine. The modified process
is less costly than the sodium process but still produces waste liquor con-
taining calcium cloride and nitrate which requires treatment (1). There has
been no further development of either of the processes.
6.9.5 In—Furnace NO 1 Removal (Three Stage Combustion — Hitachi Zosen ) (15)
Hitachi Zosen has conducted tests of in—furnace NO 1 removal, a process
they call “three stage combustion”. The process is similar to the MACT pro-
cess described in Section 6.4. Figure 6—42 shows a horizontal cylindrical
test furnace with a length of 7m and an inner diameter of 1.5m. The furnace
has a capacity of burning 200 kg/hr of pulverized coal. Some of the three
stage combustion test results are shown in Figure 6—43. The NO concentra-
tion was about 800 ppm with a conventional burner, 600 ppm with a staged—
combustion burner, and about 100 ppm when 30 of the coal was injected as the
secondary fuel, with air introduced for complete combustion. In all of the
tests, the CO in the flue gas was below 100 ppm and the amount of combustible
components in the fly ash was less than 5%.
Hitachi Zosen plans to conduct additional tests using a larger furnace
which can burn 2,000 kg/hr of pulverized coal.
482
-------
COMPLETE
OMBUSTiON COMBUSTION —
RODUCTION _j
SECONDARY ADDITIONAL
FUEL COAL) AIR
Figure 842 Hitachi Zosen’s Three Stage Combustion Test Furnace
70A 2884
CONVENTIONAL
BURNER
800
600
° STAGED COMBUSTION
BURNER
400
0
z THREE STAGE
COMBUSTION
2 4 6 8
02 IN FLUE GAS )0 )
FIgure 8.43 Reauils 01 Three Slag. ComOustion Teats
in lurnace combustion)
70A2887
483
-------
REFERENCES
1. Ando. 3. NOx Abatement for Stationary Sources in Japan, EPA—600/7—79—
205, U.S. Environmental Protection Agency, Research Triangle Park. NC,
August 1979.
2. Sengoku, T., K. Fukahori, and S. Kaneko. Non—Catalytic NO 1 Reduction
System MHIT for Steam Generator. US—JAPAN, NO 1 Information Exchange.
Tokyo, Japan, May 1981.
3. Central Technical Research Laboratory. Chubu Electric, Operating Experi-
ences of High—Temperature Noncatalytic DeNox Process at No. 2 unit of
Chita Thermal Power Station Report of IERE, 1978.
4, Report of NO 1 Abatement Technology, Japan Environnent Agency. April,
1978. (in Japanese)
5. Tokyo Electric Power Co. NO 1 Control Technology and its Application for
Power Plants. US-Japan NO 1 Information Exchange, Tokyo. Japan, May
1981.
6. Takahashi, K., et al. A Basic Study of Mitsubishi “MACT” In—Furnace NO 1
Removal Process, Mitsubishi Juko Giho, (17)6, 1980. (in Japanese)
7. Takahashi, K., et al. A Study on the Practical Application of Mitsubishi
“MACTa In—Furnace NO 1 Removal Process, Mitsubishi 1 juko Giho, (17)6,
1980. (in Japanese)
8. Okigami, N. Prospects of Combustion Technology in Coal—Fired Boilers,
Third Coal Utilization Technology Symposium by Coal Mining Research Cen-
ter, September 1981. (in Japanese)
9. Sumitomo Heavy Industries, Development of Dry Flue Gas Desuiftirization
Process. October 1980.
484
-------
10. Takenouchi, S.. Y. Fujii. and T. Atsumi. Application of Dry Desulfuriza—
tion Processes to Total Flue Gas Cleaning Systems, Third Coal Utilization
Technology Symposium sponsored by the Coal Mining Research Center, Sep-
tember 1981. (in Japanese)
11. Kawamura, K., et al. Pilot Plant Experiment of NO 1 and S02 Removal from
Exhaust Gases by Electron Beam Irradiation, Radiation Physical Chemistry,
(13)1979, pp. 5—12.
12. Kawamura, K., et al. Pilot Plant Experiment on the Treatment of Exhaust
Gas from a Sintering Machine by Electron Beam Irradiation, Environmental
Science and Technology, (14) :228, 1980.
13. Shui, V. L, and Kawamura, K. Removal of SO 1 and NO 1 from Flue Gases
by Electron Beam Radiation. Presented at the 73rd Annual Meeting of the
Air Pollution Control Association, Montreal, Quebec, July 1980.
14. Tominaga, H. Catalytic Oxidation and Reduction of NO in Furnace Gases
Using Fluidized Bed. NO 1 Fund Report 2—3 by Steel Federation, May,
1978. (in Japanese)
15. Okigami, N. Prospects of Combustion Technology in Coal—Fired Boilers,
Third Coal Utilization Technology Symposium by Coal Mining Research Cen-
ter, September 1981. (in Japanese)
485
------- |