v/EPA
United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park NC 27711
EPA-450/3-79^001
January 1979
Air
Electric Utility Steam
Generating Units -
Flue Gas Desulfurization
Capabilities as of
October 1978
-------
EPA-450/3-79-001
Electric Utility
Steam Generating Units -
Flue Gas Desulfurization Capabilities
As of October 1978
by
B.A. Laseke, Jr.
PEDCo Environmental Inc.
11499 Chester Road
Cincinnati, Ohio 45246
Contract No. 68-02-2811
EPA Project Officer: Kenneth Woodard
Emission Standards and Engineering Division
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air, Noise, and Radiation
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
January 1979
-------
This report has been reviewed by the Emission Standards and Engineering Division of the Office of Air
Quality Planning and Standards, EPA, and approved for publication. Approval does not signify that the
contents necessarily reflect the views and policies of the Environmental Protection Agency, nor does
mention of trade names or commercial products constitute endorsement or recommendation for
use. Copies of this report are available through the Library Services Office (MD-35), U.S. Environ-
mental Protection Agency, Research Triangle Park, N.C. 27711, or from National Technical Informa-
tion Services, 5285 Port Royal Road, Springfield, Va. 22161.
Publication No. EPA-450/3-79-001
-------
ABSTRACT
This study updates the previously published final report,
"Flue Gas Desulfurization System Capabilities for Coal-Fired
Steam Generators," EPA-600/7-78-032b, published in March 1978.
This assessment was made by reviewing the changes and develop-
ments in the technology since the preparation of the March 1978
report. A substantial increase in the number and capacity of
operational FGD systems, plus the additional operational exper-
ience obtained by previously identified operational systems, have
resulted in a substantial increase in the amount of design and
performance information. Most notably, these include dependa-
bility (availability, operability, reliability, and utilization)
data, removal efficiency data (sulfur dioxide and particulate),
operating problem and solution data, results from various re-
search, development, and demonstration programs, and process and
design innovations for new systems. Virtually all of the FGD
operating experience gained to date has been with the wet-phase,
nonregenerable, lime/limestone processes. As a direct result of
this previous experience, the systems committed for operation
within the next 3 to 5 years also show an overwhelming preference
for lime/limestone processes. Analysis of the current status
of the technology indicates that the design and operating exper-
ience gained with the first and second generation FGD systems
has resulted in improved design and operation of subseouent
installations. Because FGD systems that are being engineered
and/or erected will incorporate many or all of these design
innovations, even better performance can be expected without
substantial cost increase.
iii
-------
CONTENTS
Abstract
Acknowledgment xv
Figures v
Tables
1. Introduction 1_1
1.1 Application of FGD Technology in the United
States 1-1
1.2 Application of FGD Technology in Japan 1-8
1.3 Comparison of FGD Technology Development in
the United States and Japan 1-17
2. Operational FGD Systems 2-1
2.1 Lime Slurry FGD Systems 2-1
2.2 Limestone Slurry FGD Systems 2-45
2.3 Wellman Lord FGD Systems 2-142
2.4 Other FGD Systems 2-155
3. Continuous SO2 Monitors 3-1
4. Performance and Dependability of FGD Systems 4-1
4.1 Current Technological Trends 4-1
4.2 Recent Trends in the Dependability of the
Technology 4-33
4.3 Trends in Manufacturer's Guarantees 4-36
4.4 Summary of Sulfur Dioxide Removal Efficiency 4-38
Appendices
A. Domestic Lime Slurry FGD Scrubbing Systems A-l
B. Domestic Limestone Slurry FGD Scrubbing Systems B-l
C. Summary of Changes: August 1977 to October 1978 C-l
IV
-------
FIGURES
Number
1-1 Projections of coal-fired electrical generating
capacity from 1975 to 1990 and FGD capacity
from 1975 to 1986. 1-5
1-2 FGD systems in Japan from 1970 through 1977. 1-14
1-3 Trend in average ambient SO2 concentration in
Japan from 1965 through 1975. 1-16
2-1 Process flow diagram for Green River 1, 2, and 3. 2-4
2-2 Scrubber system operability - Green River 1, 2,
and 3. 2-8
2-3 Schematic of the process lines and major com-
ponents of the Bruce mansfield air quality
and waste disposal system. 2-12
2-4 FGD system availability, Pennsylvania Power,
Bruce Mansfield 1. 2-20
2-5 FGD system availability, Pennsylvania Power,
Bruce Mansfield 2. 2-21
2-6 Conesville 5 and 6, simplified process diagram. 2-27
2-7 Scrubber system operability, Columbus and Southern
Ohio Electric, Conesville 5. 2-28
2-8 Scrubber system operability, Columbus and Southern
Ohio Electric, Conesville 6. 2-30
2-9 Simplified process flow diagram, Cane Run 4. 2-35
2-10 Simplified process flow diagram, Cane Run 5. 2-35
2-11 Scrubber system operability, Louisville Gas and
Electric, Cane Run 4. 2-39
2-12 Scrubber system operability, Louisville Gas and
Electric, Cane Run 5. 2-41
-------
FIGURES (continued)
Number Page
2-13 Availability history of La Cygne 1, Kansas City
Power and Light. 2-51
2-14 Flow diagram of one of the eight FGD modules at
La Cygne 1. 2-52
2-15 La Cygne limestone wet scrubbing system. 2-53
2-16 Simplified process diagram for Northern States
Power Co., Sherburne 1 and 2 FGD system. 2-56
2-17 Simplified diagram of a Sherburne scrubber module 2-57
2-18 Availability history of Sherburne 1 and 2,
Northern States Power Company 2-64
2-19 Simplified process flow diagram, Cholla 1 FGD
system. 2-77
2-20 Simplified process flow diagram of Module A,
Cholla 1 FGD system. 2-78
2-21 Reliability history of Cholla 1, Arizona Public
Service, from January 1974 through December 1975 2-88
2-22 Reliability of Cholla 1, Arizona Public Service,
from January 1976 through December 1977. 2-89
2-23 Reliability history of Cholla 1, Arizona Public
Service, from January through August 1978. 2-90
2-24 Cutaway view of a Duck Creek 1 FGD scrubber module 2-94
2-25 Duck Creek 1 FGD system scrubbing circuit. 2-97
2-26 Simplified process flow diagram of Duck Creek 1
power plant and emission control system. 2-98
2-27 Simplified process flow diagram of one of the two
Lawrence 4 scrubbing modules. 2-113
vi
-------
FIGURES (continued)
Number Page
2-28 Simplified process flow diagram of one of the two
Lawrence 5 scrubbing modules 2-114
2-29 Simplified process flow diagram of Southwest 1,
Springfield City Utilities. 2-118
2-30 A schematic of TVA Widows Creek 8 FGD System. 2-123
2-31 A simplified process flow diagram of TVA Widows
Creek 8 FGD system. 2-124
2-32 Availability of TVA Widows Creek 8 FGD system. 2-129
2-33 Simplified flow diagram of Martin Lake 1 and 2
FGD systems. 2-135
2-34 Scrubber system availability, Northern Indiana
Public Service, Dean H. Mitchell 11. 2-145
2-35 Process diagram of FGD module at Montana Power Co.
Colstrip 1 and 2. 2-159
2-36 FGD system availability, Montana Power, Colstrip 1. 2-168
2-37 FGD system availability, Montana Power, Colstrip 2. 2-169
2-38 Process diagram of an FGD system, Nevada Power Co.,
Reid Gardner Station. 2-181
2-39 FGD system availability, Nevada Power Co.,
Reid Gardner. 2-186
2-40 FGD system availability, Nevada Power Co.,
Reid Gardner. 2-187
2-41 FGD system availability, Nevada Power Co.,
Reid Gardner 3. 2-188
2-42 Jet bubbling reactor. 2-190
2-43 Process flow diagram of prototype plant, Gulf
Power Co., Sholz Station. 2-194
vii
-------
FIGURES (continued)
Number Page
2-44 Dry phase collection with nahcolite injection. 2-197
2-45 Process flow diagram of two-stage dry scrubber/SO2
absorber. 2-200
2-46 Carborundum/DeLaval dry phase collection system. 2-203
2-47 ,Joy Manufacturing/Niro Atomizer dry phase collection
process. 2-204
4-1 Average plant FGD dependability versus startup date. 4-34
A-l Summary of Four Corners 2A Scrubber Module Isolation
Testing: Effect of pH on SC>2 removal. A-6
Vlll
-------
TABLES
Number Page
1-1 Number and Capacity of U.S. Utility FGD Systems
in August 1977 and October 1978 1-3
1-2 Capacity of Committed U.S. Utility FGD Systems
as October 1978 1-7
1-3 Number and Capacity of FGD Systems in Japan as
of January 1978 1-9
2-1 Major Domestic Operational Lime Slurry FGD Systems 2-2
2-2 Green River Power Station, Operational Data
for 1977 through September 1978 2-6
2-3 Green River Power Station, Operational Data
for September 1977 to September 1978 2-7
2-4 Summary of Sulfur Dioxide Emission Rates 2-9
2-5 FGD Systems Data for Bruce Mansfield 1 and 2
Shippingport, Pennsylvania 2-11
2-6 Performance Data on Bruce Mansfield 1 FGD System,
Pennsylvania Power 2-18
2-7 Performance Data on Bruce Mansfield 2 FGD System,
Pennsylvania Power 2-19
2-8 FGD System Data for Conesville 5 and 6,
Conesville, Ohio 2-25
2-9 Performance Information on Conesville 5,
Columbus and Southern Ohio Electric 2-29
2-10 Performance Data on Conesville 6, Columbus
and Southern Ohio Electric 2-31
2-11 FGD Systems Data for Cane Run 4, Louisville,
Kentucky 2-34
2-12 FGD Systems Data for Cane Run 5, Louisville,
Kentucky 2-36
ix
-------
TABLES (continued)
Number Page
2-13 Operating Data on Cane Run 4, Louisville Gas
and Electric 2-40
2-14 Operating Data on Cane Run 5, Louisville Gas
and Electric 2-42
2-15 Major Domestic Operational Limestone FGD
Installations 2-46
2-16 La Cygne Station Unit No. 1 4-Hour Full Load and
Stack Emission Test 2-54
2-17 Sherburne 1 Performance Summary: May 1976 to
September 1978 2-65
2-18 Sherburne 1 Performance Summary From October 1977
to September 1978 2-67
2-19 Sherburne 2 Performance Summary from April 1977
to September 1978 2-68
2-20 Sherburne 2 Performance Summary from October 1977
to September 1978 2-69
2-21 FGD Systems Data for Cholla 1, Joseph City,
Arizona 2-72
2-22 FGD Systems Data for Cholla 2, Joseph City,
Arizona 2-73
2-23 Cholla 1 Performance Summary January 1974 to
September 1978 2-85
2-24 FGD Systems Data for Duck Creek 1, Canton,
Illinois 2-93
2-25 Results of the D-Scrubber Module Test 2-103
2-26 FGD Systems Data for Lawrence 4, Lawrence, Kansas 2-106
2-27 FGD Systems Data for Lawrence 5, Lawrence, Kansas 2-107
-------
TABLES (continued)
Number Page
2-28 FGD Systems Data for Southwest 1, Springfield,
Missouri 2-117
2-29 FGD Systems Data for Windows Creek 8, Bridgeport,
Alabama 2-121
2-30 FGD System Performance of TVA Widows Creek 8 2-130
2-31 Preliminary Scrubber Sulfur Dioxide Removal Test
Data for Widows Creek 8 2-131
2-32 FGD Systems Data for Martin Lake 1, Tatum, Texas 2-133
2-33 FGD Systems Data for Martin Lake 2, Tatum, Texas 2-134
2-34 Domestic Operational Wellman Lord FGD Installations 2-143
2-35 Performance Data for Dean H. Mitchell 11, Northern
Indiana Public Service 2-147
2-36 FGD Systems Data for San Juan 1 and 2 Waterflow,
New Mexico 2-152
2-37 FGD Systems Data for Colstrip 1, Colstrip,
Montana 2-157
2-38 FGD Systems Data for Colstrip 2, Colstrip,
Montana 2-158
2-39 Background Information for SO? Emissions at
Colstrip 1 and 2 2-164
2-40 Tests of SO2 Emissions at Colstrip 1 2-166
2-41 Tests of SO- Emissions at Colstrip 2 2-167
2-42 Scrubber Availability for Colstrip 1 and 2 2-170
2-43 FGD System Data for Milton R. Young 2, Center,
North Dakota 2-172
XI
-------
TABLES (continued)
Number Page
2-44 FGD Systems Data for Reid Gardner 1,
Moapa, Nevada 2-177
2-45 FGD Systems Data for Reid Gardner 2,
Moapa, Nevada 2-178
2-46 FGD Systems Data for Reid Gardner 3,
Moapa, Nevada 2-179
2-47 FGD Performance Data, Nevada Power Co.,
Reid Gardner 1 2-183
2-48 FGD Performance Data, Nevada Power Co.,
Reid Gardner 2 2-184
2-49 FGD Performance Data, Nevada Power Co.,
Reid Gardner 3 2-185
2-50 FGD Systems for Scholz 1 and 2, Chattahoochee,
Florida 2-192
2-51 Dry Powder Nahcolite Injection Versus Two-Stage
Soda Ash System 2-202
3-1 Identification of FGD Units with Continuous
SO2 Monitors 3-2
3-2 Summary of the Performance of Continuous S02
Monitors 3-6
3-3 Companies Known to Record Continuous S02
Monitoring Data 3-9
4-1 Identification of Plants in Figure 4-1 4-35
4-2 U.S. Utility FGD Systems Achieving or Exceeding
90% Sulfur Dioxide Removal 4-39
A-l Summary of Four Corners 2A Scrubber Module Iso-
lation Testing Service in February and March 1976 A-5
A-2 FGD System Data for Conesville 5 and 6, Conesville,
Ohio A-9
xii
-------
TABLES (continued)
Number Page
A-3 FGD System Data for Elrama Power Station,
Elrama, Pennsylvania A-12
A-4 FGD System Data for Phillips Power Station,
South Height, Pennsylvania A-15
A-5 FGD System Data for Hawthorn 3 and 4, Kansas
City, Missouri A-19
A-6 FGD System Data for Green River 1, 2, and 3,
Central City, Kentucky A-21
A-7 FGD System Data for Cane Run 4, Louisville,
Kentucky A-25
A-8 FGD System Data for Cane Run 5, Louisville,
Kentucky A-28
A-9 FGD System Data for Mill Creek 3, Louisville,
Kentucky A-30
A-10 FGD System Data Milton R. Young 2, Center,
North Dakota A-37
A-ll Background Information for SO- Emissions at
Colstrip 1 and 2 A-43
A-12 Tests of S02 Emissions at Colstrip 1 A-44
A-13 Tests of SO2 Emissions at Colstrip 2 A-45
A-14 FGD System Data for Bruce Mansfield 1 and 2,
Shippingport, Pennsylvania A-47
A-15 Summary of Emission Control System at Bruce
Mansfield 3 A-50
A-16 FGD Systems Data for Huntington 1, Price, Utah A-58
B-l FGD System Data for Tombigbee 2 B-5
B-2 FGD System Data for Apache 2 B-8
XI11
-------
TABLES (continued)
Number Page
B-3 FGD System Data for Cholla 1 B-13
B-4 FGD System Data for Cholla 2 B-14
B-5 FGD System Data for Duck Creek 1 B-20
B-6 FGD System Data for Scholz 1 and 2 B-25
B-7 FGD System Data for Petersburg 3 B-29
B-8 FGD system Data for La Cygne 1 B-35
B-9 FGD System Data for Jeffrey 1 B-38
B-10 FGD System Data for Lawrence 4 B-43
B-ll FGD System Data for Lawrence 5 B-44
B-12 FGD System Data for Sherburne 1 B-49
B-13 FGD System Data for Sherburne 2 B-50
B-14 FGD System Data for Winyah 2 B-56
B-15 FGD System Data for R. D. Morrow 1 B-60
B-16 FGD System Data for Southwest 1 B-63
B-17 FGD System Data for Widows Creek 8 B-68
B-18 FGD System Data for Martin Lake 1 B-75
B-19 FGD System Data for Martin Lake 2 B-76
B-20 FGD System Data for Monticello 3 B-82
xiv
-------
ACKNOWLEDGMENT
This report was prepared under the direction of Mr. Richard
W. Gerstle. The principal authors were Mr. Bernard A. Laseke,
Jr., Mr. Michael T. Melia, Mr. Michael P. Smith, and Mr. T. S.
Koger. Project Officer for the Environmental Protection Agency
was Mr. Kenneth Woodard. The authors appreciate the contribu-
tions made by Mr. Woodard, and also Mr. Kenneth R. Durkee of the
Environmental Protection Agency.
xv
-------
SECTION 1
INTRODUCTION
Early in 1977, the U.S. Environmental Protection Agency
(EPA) undertook a program to review and revise the existing New
Source Performance Standards (NSPS) limiting sulfur dioxide (SO2)
emissions from large fossil-fuel-fired steam generators. The
existing emission regulation was first promulgated in 1971. .Jks..a
result of the Clean Air Act (CAA) ammendments of 1977, the EPA
was mandated to review this regulation in light of new technology.
The EPA's Office of Air Quality Planning and Standards (OAQPS)
had prime responsibility for the review, which formed a basis for
determining revisions to the existing NSPS level.
To develop background information for support of this
review, the EPA contracted with PEDCo Environmental to study the
capabilities of flue gas desulfurization (FGD) technology. Th(is
study was conducted between May 1977 and February 1978. A final
report, "Flue Gas Desulfurization System Capabilities for Coal-
Fired Steam Generators," EPA-600/7-78-032 a and b, was published
in March 1978.
The most recent information published in the March 1978
report was as August 1977. This report discusses changes in FGD
technology that occurred from August 1977 to October 1978.
1.1 APPLICATION OF FGD TECHNOLOGY IN THE UNITED STATES
1.1.1 Current Level of Development
As of October 1978, there were 143 utility FGD systems with
an equivalent electrical generating capacity of 61,732 MW in
operation, under construction, or planned in the United States.
1-1
-------
Of these systems, 46 were operational (16,054 MW); 39 were under
construction (16,728 MW); and 58 systems were planned (28,950
MW) .l
A number of plants that will use FGD systems are omitted
because such information is not ready for public release. From
55 to 60 systems with an equivalent electrical generating capac-
ity between 36,000 and 41,000 MW fall into this category.2
To date, 16 systems representing approximately 1555 MW have
been shut down. Some (425 MW), however, continue to remove fly
ash and some SO- (from 35 to 50 percent), because alkaline addi-
A
tives are still added to the scrubbing solution for pH control.
1.1.2 Changes From August 1977 to October 1978
As of August 1977, there were 125 utility FGD systems with
an equivalent electrical generating capacity of 53,352 MW in
operation, under construction, or planned. Of these systems, 29
were operational (8,914 MW); 28 were under construction (11,810
MW); and 68 systems were planned (32,628 MW) .
The total number and capacity of U.S. utility FGD systems in
August 1977 and October 1978 are provided in Table 1-1, which
summarizes changes in these totals. This table shows that the
number of operational FGD systems increased by approximately 60
percent and the capacity of such systems, by 80 percent. It also
indicates that the number and capacity of FGD systems under con-
struction each increased by approximately 40 percent. According
to Table 1-1, the number of FGD systems in the planning stage
(contract awarded, letter of intent signed, requesting or evalu-
ating bids, and considering FGD) decreased by approximately 15
percent; and the capacity of such systems decreased by 11 percent,
If, however, the FGD systems that are omitted at the present time
because information about them is not ready for public release
are taken into account, the number of FGD systems in the planning
stage increases to at least 113; and the capacity of planned
systems, to 65,000 MW. These values represent approximately a 66
1-2
-------
TABLE 1-1. NUMBER AND CAPACITY OF U.S. UTILITY FGD SYSTEMS IN
AUGUST 1977 AND OCTOBER 1978
Status
Operational
Under construction
Planned:
Contract awarded
Letter of intent
signed
Requesting or
evaluating bids
Considering FGD
Total
August 1977
No.
29
28
23
5
5
35
125
MW
8,914
11,810
11,880
1,892
2,825
16,031
53,352
October 1978
No.
46
39
23
1
6
28
143
MW
16,054
16,728
12,450
240
3,650
12,610
61,732
Changes
No.
+17
+11
-4
+1
-7
+18
MW
+7,140
+4,918
+570
-1,652
+825
-3,421
+8,380
1-3
-------
percent increase in the number of systems and a 125 percent
increase in capacity.*
A summary of the individual units that changed status
between August 1977 and October 1978 is provided in Appendix C of
this report.
Several other trends have been detected during the interim
period. As Figure 1-1 shows, coal-fired electrical generating
capacity in the United States increased by 6 percent from 250,000
MW in late 1977 to approximately 265,000 MW by late 1978. Figure
1-1, which was developed from several sources, also indicates
that the application of FGD systems increased from 3.5 percent of
the total coal-fired capacity in September 1977 to 6 percent a
year later.
More low-sulfur coal is being fired than before. Low-sulfur
coal is defined as coal that, when combusted, will emit no more
than 516 ng/J (1.2 lb/106 Btu) of SC>2, the existing NSPS limit.
High-sulfur coal is defined as coal that will cause higher emis-
sions. Approximately 90 percent of the August 1977 operating
capacity (8914 MW) came from units firing high-sulfur coal, but
only 81 percent of the operating capacity 1 year later was
supplied by such units.
Approximately 50 percent of the current committed FGD
capacity is designed to meet existing NSPS. This was true of
only 42 percent of committed FGD capacity in August 1977.
1.1.3 Process Selection
Chemical process in this report refers to the reagent used
to scrub SO2 from the flue gas stream. The major generic chemi-
cal processes available for commercial application in the United
States rely on limestone, lime, sodium, and magnesium. These
reagents are used in eight major process designs: direct lime-
stone, direct lime, sodium carbonate, magnesium oxide, sodium/
*
As already mentioned, from 55 to 60 systems with an equivalent
electrical generating capacity between 36,000 and 41,000 MW are
planned, but not yet publicly announced. The lower estimates
for both number and capacity have been used to calculate the
increase suggested.
1-4
-------
450 r
I I I I I I I I I
CO
o
- 250 h-
o
-------
thermal regeneration (Wellman Lord), sodium/calcium double
alkali, aqueous carbonate, and citrate.
Table 1-2 summarizes the systems that are operational, under
construction, and planned according to committed process type as
of October 1978. Of the total active committed capacity of
50,827 MW, calcium-based (limestone/lime) process designs account
for 44,980 MW, or 88.5 percent of the total.
The utility industry has preferred limestone to lime pro-
cesses, primarily because of economic considerations. This
preference is expected to continue for the next few years. In
August 1977, only 4047 MW of FGD operating capacity relied on
limestone; and 4237 MW, on lime.8 By October 1978, however, 8734
MW of operating capacity came from limestone systems; and 6070
MW, from lime systems. These values constitute 54 percent and 38
percent of the total current operating capacity, respectively.
Table 1-2 indicates a strong preference for limestone also among
systems under construction, and an even stronger one among
systems still being planned.
There are several reasons for the predominance of limestone
and lime processes. They have had the most development and
commercial operating experience; the first commercial applica-
tions of calcium-based FGD technology on flue gas from coal-fired
th<
10
Q
units began in 1968. Also, these processes do not require the
sophisticated, multiple loops of many regenerable processes.
Limestone is available in practically all parts of the United
States. It is a major commodity, ranking second after sand and
gravel, and the annual production of limestone is from 5.4 to 6.4
Gg (600 to 700 million tons). The capital and annual costs of
limestone and lime processes are generally much less than those
12
of other processes. Furthermore, limestone and lime processes
have been designed and used to meet the existing NSPS. These
processes have demonstrated an ability to maintain such removal
levels when both high- and low-sulfur coals are fired.
1-6
-------
TABLE 1-2. CAPACITY OF COMMITTED U.S. UTILITY FGD SYSTEMS
AS OCTOBER 1978
(MW)
Process
Limestone
Lime
Lime/limestone
Sodium carbonate
Magnesium oxide
Wellman Lord
Double alkali
Aqueous carbonate3
Citrateb
Total
FGD capacity
Operational
8,734
6,070
20
375
120
735
0
0
0
16,054
Under
construction
8,497
5,050
330
509
600
180
1,102
400
60
16,728
Planned
11,488
4,461
330
0
726
940
0
100
0
18,045C
Total
28,719
15,581
680
884
1,446
1,853
1,102
500
60
50,825C
Includes nonregenerable dry collection process design and
regenerable process design configurations.
This system is being installed at St. Joseph Minerals' G.F.
Wheaton Plant and is considered a utility FGD system because
the plant is connected by a 25-MW interchange to the Duquesne
Light Company.
The processes for all planned systems are not known. Thus, the
totals in this table are less than those in Table 1-1.
1-7
-------
1.2 APPLICATION OF FGD TECHNOLOGY IN JAPAN
1.2.1 Current Level of Development
At the beginning of 1978, there were over 500 major FGD
systems in service in Japan with a combined gas capacity greater
than 89,000,000 Nm3/h (52,500,000 scfm). This capacity is
equivalent to 25,000 MW of electrical generating capacity.*
Table 1-3 summarizes the systems by supplier, process, and
14 l 5
capacity. '
Approximately 500 more operational sodium systems, repre-
senting a combined gas capacity of approximately 10,000,000 Nm /h
(5,900,000 scfm), are not identified in Table 1-3.16 These
additional systems bring the total combined FGD operating capac-
ity in Japan to 99,000,000 Nm /h (58,400,000 scfm), which is
equivalent to 28,000 MW of electrical generating capacity.
Table 1-3 indicates that approximately 47 percent of the FGD
operating capacity in Japan relies on the direct lime/limestone
process that produces gypsum as a byproduct, 16 percent on the
indirect lime/limestone (double alkali) process that also pro-
duces gypsum, 13 percent on regenerable processes that produce
sulfuric acid, elemental sulfur, and ammonium sulfate, and 24
percent on once-through sodium processes that produce sodium
sulfite or sulfate. Direct lime/limestone systems average
427,000 Nm /h (252,000 scfm), or approximately 120 MW; indirect
lime/limestone systems, 291,000 Nm /h (172,000 scfm) or 80 MW;
regenerable systems, 378,000 Nm3/h (223,000 scfm), or 110 MW; and
once-through sodium systems, 59,600 Nm /h (35,200 scfm), or 15
MW.17'18
Approximately half of Japan's total FGD capacity is installed
at utility power plants. By the end of 1978, there were 51
utility FGD systems with an equivalent electrical generating
capacity of 12,649 MW in operation, under construction, or
planned. Of these systems, 47 were operational (11,224 MW); and
*
For general considerations, 1 MW is equivalent to approximately
3560 Nm3/h (2100 scfm).
1-8
-------
TABLE 1-3.
NUMBER AND CAPACITY OF FGD SYSTEMS IN
JAPAN AS OF JANUARY 1978a
VD
Plant constructor
Mitsubishi Heavy Industries
(MHI)
Ishikawajima U.I. (IHI)
Hitachi, Ltd.
Mitsubishi Kakoki (MKK)
Kawasaki Heavy Industries
Tsukishima Kikai (TSK)
Chiyoda Chemical Engineering
& Construction
Oji Koei
Fuji Kasui Engineering
Kurabo Engineering
Mitsui Miike-Chemico
Ebara Manufacturing
Nippon Kokan (NKK)
Kureha Chemical
Showa Denko
Gadlius
Sumitomo (SCED)-Wellman
Mitsui Metal Engineering
Kobe Steel
Japan Gasoline
Dowa Engineering
Niigata Iron Works
Mitsui Shipbuilding
Sumitomo Heavy Industries
Total
Direct
1 ime/ 1 ime s to ne
No.
33
17
13
2
4
1
7
-t
3
4
5
1
94
lOSNm-Vh
18,270
14,445
6,940
256
756
3,954
2,744
245
1,006
1,125
330
40,171
Indirect
1 ime/ 1 imes tone
No.
6
4
14
5
11
1
5
1
47
lO^NmVh
5,450
398
4,459
413
1,914
150
453
185
13,422
Regenerable
processes
No.
2
13
1
1
1
2
6
2
1
1
30
10JNm3/h
590
6,478
88
18
500
1,990
1.288
130
125
150
11,357
Once- through
sodium
No.
3
79
15
41
7
40
57
6
106
10
6
8
5
8
1
392
lb3Nn»3/h
292
4,351
603
913
256
4.042
4,280
270
3,751
1,167
62
1,431
1,372
1,291
160
24,241
Total
No.
36
96
30
56
17
46
14
57
13
112
5
21
12
8
5
8
6
6
5
2
5
1
1
1
563
103Nn>3/h
18,562
8,796
8,133
7,643
6,380
4,528
4,459
4,280
4,224
4,182
3,244
3,081
2,447
1,431
1,372
1,291
1,288
1,136
1,125
455
500
185
160
150
89,138
a The byproduct of direct lime/limestone and indirect lime/limestone systems is gypsum.
-------
4 were under construction or planned (1425 MW). Table 1-4
provides a summary of the committed full-scale FGD systems
21
applied to power plants in Japan. Table 1-5 indicates the
capacity of steam power plants and FGD systems in-service or
19 20
planned in Japan. '
As shown in Table 1-5, nine major utility companies, largely
using imported oil, have produced about 70 percent of the total
steam electric power generated in Japan. The Electric Power
Development Company (EPDC) has been the primary consumer of
domestic coal for power production, although coal-fired utilities
account for less than 3 percent of the total capacity (including
plants under construction) of 87,465 MW. Oil- and coal-fired
units provide about 75 percent of this total; hydroelectric
21 22
units, 20 percent; and nuclear units, 5 percent. '
Tokyo Electric, Kansai Electric, and Chubu Electric supply
power to the largest cities and industrial complexes in Japan.
As Table 1-5 shows, these companies use such low-sulfur fuels as
naphtha and liquid natural gas in heavily polluted sections of
their service areas. Recent regulations require large power
plants in designated regions to keep SO,, stack emissions below 50
ppm, so that low-sulfur fuel must be used for reheat. More
stringent standards may reduce the use of FGD as an effective
control strategy. Hokuriku Electric, however, has power plants
located far from large cities and uses FGD on a large percentage
* -4. -^ 23,24
of its capacity.
1.2.2 FGD Trends
Application of FGD systems in Japan has increased rapidly
o c J c.
since 1972, as indicated by Figure 1-2. ' This growth can be
largely attributed to elimination of the previous large differ-
ence in cost between FGD and low-sulfur fuels and increasing
27 28
confidence in the reliability of FGD systems. '
Table 1-4, however.- shows that only four new systems are
under construction or planned. One cause of this declining
growth rate is the reduction of ambient SO2 concentrations in
1-10
-------
TABLE 1-4. OPERATIONAL AND PLANNED UTILITY FGD SYSTEMS IN JAPAN
Power company
Tohoku
Tokyo
Chubu
Hokuriku
Kanaai
,
Chugoku
Hokkaido
Power station
Shinsendai
Hachinohe
Niigata
Niigata H.
Akita
Kashima
Yokosuka
Nishinagoya
Owase
Toyama
Fukui
Nanao
Sakai
Amagasaki
Osaka
Kainan
Hizushima
Tamashima
Shimonoseki
(coal)
Higashitomakoma i
Boiler
No.
2
4
4
1
3
3
1
1
1
1
1
1
8
2
1
3
2
4
4
2
3
2
2
1
1
MW
600
250
250
600
350
600
265
220
375
500
350
500
250
156
156
156
156
156
600
156
500
350
400
1753
500
FGD
MM
150
125
125
150
350
150
133
220
375
250
350
500
63
35
121
156
156
156
156
150
100
500
350
400
175
250
Process developer
Kureha-Kawasaki
Mitsubishi H.I.
Wellman-MKK
Mitsubishi H.I.
Kureha-Kawasaki
Hitachi-Tokyo
Mitsubishi H.I.
Wellman-MKK
Mitsubishi H.I.
Chiyoda
Chiyoda
Not decided
Sumitomo H.I.
Mitsubishi H.I.
Mitsubishi H.I.
Mitsubishi H.I.
Babcock-Hitachi
Babcock-Kitachi
Babcock-Ilitachi
Mitsubishi H.I.
Babcock-Hitachi
Babcock-Hitachi
Babcock-Hitachi
Mitsubishi H.I.
Mitsubishi H.I.
Not decided
Absorbent, precipitant
Na2S03, CaO>3
CaO
Na2S03
CaC03
Na2S03, CaCOs
Carbon, CaC03
CaC03
Na2SO3
CaO
CaO
H2SO4, CaCO3
H2SO4, CaCO,
II2SO4, CaCO3
Carbon
CaO
CaO
CaO
CaCO3
CaC^3
CaC03
CaO
CaC03
CaC03
CaC03
CaCO,
CaCOj
CaCO3
By product
Gypsum
Gypsum
H2SO4
Gypsum
Gypsum
Gypsum
Gypsum
H2S04
Gypsum
Gypsum
Gypsum
Gypsum
H2S04
Gypsum
Gypsum
Gypsum
Gypsum
Gypsum
Gypsum
Gypsum
Gypsum
Gypsum
Gypsum
Gypsum
Gypsum
Gypsum
Year of
completion
1974
1974
1976
1976
1977
1972
1974
1973
1976
1974
1975
1978
1972
1973
1975
1976
1975
1975
1976
1974
1974
1975
1976
1976
1979
1981
aCoal-fired boilers. Others are for oil-fired boilers.
(continued)
-------
TABLE 1-4 (continued)
Power company
Shikoku
Kyushu
EPDC
Niigata
Showa
Toyama
Hizusbima
Sumitomo
Sakata
Pukui
Power station
Anna
Sakaide
Karita
Karatsu
Ainoura
Buzen
Takasago (coal)
(coal)
Isogo (coal)
Takehara (coal)
Matsushima (coal)
(coal)
Niigata
Ichihara
Toyama
Mizushima
Niihama
Sakata
Fukui
Boiler
No.
3
3
2
2
3
1
2
1
2
1
2
1
1
1
2
1
1
5
1
5
3
1
2
1
MW
450
450
375
375
500
375
500
500
500
250«»
250a
265a
250*
SCO*
500a
350
150
250
250
156
156
350
350
250
FGD
MW
450
450
188
188
250
250
250
250
250
250
250
265
250
500
500
175
150
250
250
156
156
350
350
250
Process developer
Kureha-Kawasaki
Kureha-Kawasaki
Mitsubishi H.I.
Mitsubishi H.I.
Mitsubishi H.I.
Mitsubishi H.I.
Mitsubishi H.I.
Kureha-Kawasaki
Kureha-Kawasaki
Mitsui-Chemico
Mitsui-Chemico
Chemico-IHI
Babcock-Hitachi
Not decided
Not decided
MHI
Showa Denio
Babcock-H i tach i
Chiyoda
Mitsubishi H.I.
IHI
Mitsubishi H.I.
Mitsubishi H.I.
Not decided
Absorbent, precipitant
Na2S03, CaCO3
CaO
CaCO,
CaCO3
CaCC>3
CaCO 3
Na2SO,, CaCO,
NA2S03, CaC03
CaCO,
CaCO 3
CaC03
CaCO3
CaCO 3
CaC03
N32SO3, CaSO3
CaC03
H2SO4, CaC03
CaO
CaC03
CaC03
CaC03
CaCO 3
Byproduct
Gypsum
Gypsum
Gypsum
Gypsum
Gypsum
Gypsum
Gypsum
Gypsum
Gypsum
Gypsum
Gypsum
Gypsum
Gypsum
Gypsum
Gypsum
Gypsum
Gypsum
Gypsum
Gypsum
Gypsum
Gypsum
Gypsum
Gypsum
Year of
completion
1975
1975
1974
1976
1976
1976
1976
1977
1978
1975
1976
1976
1977
1980
1980
1975
1973
1976
1975
1975
1975
1976
1977
1977
H
I
M
NJ
^oal-fired boilers. Others are for oil-Cired boilers.
-------
TABLE 1-5. CAPACITIES OF STEAM POWER PLANTS AND FGD SYSTEMS IN JAPAN
I
M
U>
Power company
Hokkaido
Tohoku
Tokyo
Chubu
HoXuriku
Kai.sai
Chuyoku
Shikoku
Kyushu
EPDC
Niigata
Showa
Toyama
Mizushima
Sumitomo
Sakata
Fukui
Others
Total
Power generation, MW
Existing
'1..270
4,275
19,167
9,933
..«..->.
-. 'j, 572
3,7''7
2,687
5,000
1,430
350
550
750
462
368
0
250
5,512
68r365
Under
construction
1,225
850
4,400
3,800
500
1,200
1,800
450
2,200
1,000
350
0
0
0
250
700
0
375
19,100
Total
2,495
5,125
23,567
13,733
2,214
11,872
5,577
3,137
7,200
2,430
700
550
750
462
618
700
250
5,887
87,645
FGD, MW
Existing
0
900
283
970
1,100
993
1,350
900
1,626
1,015
175
400
250
156
156
700
250
0
11,224
Under
construction
250
0
0
0
0
0
175
0
0
1,000
0
0
• o
0
0
0
0
0
1,425
Total
250
900
283
970
1,100
993
1,525
900
1,626
2,015
175
400
250
156
156
700
250
0
12,649
a Includes units in the planning stage as well as those currently under construction.
-------
1200
1000H
800k
CO
CO
>-
CO
CtL
UJ
CO
0
1970 1971
1972
1973 1974 1975 1976 1977
Figure 1-2. FGD systems in Japan from 1970 through 1977.
-------
large cities and industrial districts to a level between 0.02 and
0.03 ppm, which almost achieves the annual standard of 0.016 ppm
2Q
(see Figure 1-3). Another cause is the recent downturn in the
Japanese economy, which has restricted the building of new
plants; the majority of the FGD systems were built between 1970
and 1975, when the Japanese economy was growing rapidly. '
As of the beginning of 1978, desulfurized oil was in such over-
supply that its cost was only $19 to $22 per kiloliter more than
high-sulfur oil; thus, burning desulfurized oil was not more
expensive than firing high-sulfur oil with scrubbers and avoided
32
the problems of finding capital.
The market for byproduct gypsum from FGD operations has
become saturated; and throwaway operation is unattractive to
Japanese utilities because of the sludge disposal problem. The
declining growth rate of FGD capacity may also be attributed in
part to more stringent regulations governing nitrogen oxides
(NO ) emissions. Postcombustion removal of nitrogen oxides is
X
required; and catalytic reduction with ammonia, the favored
method of removal. Such systems work much better with low-sulfur
oil than high-sulfur oil because sulfur trioxide poisons the
34
catalyst. Furthermore, the more stringent NO emission stan-
J^
dards have spurred the development of technology for simultaneous
removal of NO and SO,,. Instead of installing separate systems
X fc
for NO and SO9 control, much of Japanese industry is waiting for
35 36
this new technology to be demonstrated. '
Some factors may promote FGD development and application in
Japan. The Federal Government is encouraging shifts from oil to
imported and domestic coal. If such shifts occur, scrubbers will
probably be required, because it would be difficult to find coal
with a low enough sulfur content to meet SO2 emission standards,
which in many areas require oil with a sulfur content no higher
37
than 0.2 or 0.3 percent. Also, the cost of oil is likely to
increase as soon as supply and demand level out; and FGD costs
are declining because of process design innovations and competi-
38
tion. New uses (e.g., in the cement industry) for gypsum are
1-15
-------
0.06
i
H
cr\
0
1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975
Figure 1-3. Trend in average ambient S02 concentration in Japan from 1965 through 1975
-------
39
being found and could encourage the growth of FGD capacity.
Finally, regulations are becoming so tight that in some areas the
equivalent of "best available technology" and "non-deterioration"
criteria are being applied. Scrubbers have the advantage that
40
any degree of SO2 removal can be attained.
1.3 COMPARISON OF FGD TECHNOLOGY DEVELOPMENT IN THE UNITED
STATES AND JAPAN
FGD systems have been used more in Japan than in the United
States because of Japan's serious air pollution problems in the
1960's. These problems resulted from high concentrations of
industry near major Japanese population centers.
Approximately 1000 FGD plants having a combined equivalent
capacity of 28,000 MW are presently in operation in Japan. In
the utility sector, approximately 66,000 MW of power generation
is produced by fossil fuel combustion, 2 to 3 percent of which is
provided by coal. Approximately 16 percent of the current
fossil-fuel-fired electrical generation is controlled by FGD.
Projections indicate that the FGD-controlled capacity will drop
41 42
to approximately 14.5 percent by the end of 1981. '
The application of FGD systems in the U.S. utility sector
has grown and will continue to do so. Such application of FGD
systems increased from 3.5 percent of the total coal-fired power
generating capacity in September 1977 to 6 percent 1 year later.
Current projections* indicate that FGD capacity could be anywhere
from 17 to 24 percent of the total coal-fired utility capacity by
1985.43'44
FGD technology is working well in Japan on both utility and
industrial applications. The performance of the coal-fired
units, with 90+ percent SO2 removal efficiency and 95+ percent
By 1985, coal-fired capacity is projected to total 363,200 MW;
and committed FGD capacity, to reach 61,732 MW, or 17 percent
of this total. If the revised NSPS require all new coal-fired
utility boilers to be controlled by FGD systems after 1983, the
FGD capacity by 1985 will be 87,000 MW, or 24 percent of the
total.
1-17
-------
availability, has not appreciably differed from that of oil-fired
or industrial counterparts.
The success of lime/limestone scrubbing in Japan had been
attributed partially to open loop operation created by purging
large quantities of process liquids. The extent of blowdown,
however, is variable. Often, no more water is removed from a
Japanese lime/limestone system through the gypsum produced and
liquid purged than from a typical closed loop U.S. system through
ponding sludge. '
There are a number of differences between Japan and the
United States in FGD technology development and applications.
Japanese system suppliers and operators recognize that FGD
systems involve chemical processes requiring carefully controlled
operation by personnel specifically trained for this purpose.
In Japan, the user expects and demands that the FGD system
supplied perform with very high reliability. Such demands have
caused scrubber systems to be initially more expensive than in
the United States, but have reduced subsequent operating problems
49 50
and maintenance costs. '
The sulfur content range of the coal burned in Japanese
utility and industrial boilers is lower than that of the coal
fired in many U.S. power generating systems. Japanese FGD
systems have been generally used on flue gases with an SO. inlet
concentration between 400 and 2300 ppm. Although this range
occurs in many U.S. utility FGD systems, there is no experience
51 52
in Japan with the high-sulfur coals burned by many U.S. utilities. '
Japan employs a stringent, continuous monitoring and en-
forcement program to insure that utility and industrial sources
53 54
are in continuous compliance with environmental regulations. '
In some areas of Japan where the ambient concentration of
SO7 is considered to be sufficient to cause health problems, an
SO, emission tax has been levied. Proceeds from such taxes have
^
been used for the care of patients diagnosed as having pollution-
related illnesses, such as chronic bronchitis. '
1-18
-------
Finally, a cooperative spirit appears to exist between
Japanese industry and regulatory agencies. The need for SO- con-
trols was critical in the late 1960's, when Japan's FGD installa-
tion program began to accelerate. The national goal of a cleaner
environment has been accepted by the utility industry, which has
tried to buy the best scrubbers available and to operate them
properly.57'58
1-19
-------
REFERENCES FOR SECTION 1
1. In-house information gathered by PEDCo Environmental for the
EPA utility FGD survey program. Ccntract No. 68-01-4147,
Task No. 52. November 1978.
2. Personal Communication with R. RAO. Research Cottrell,
Marketing projection based upon data published by the
National Electric Reliability Council. April 1978. pp.
146-163.
3. Laseke, B.A., Jr., and T.W. Devitt. Status of Flue Gas
Desulfurization Systems in the United States. Presented at
the 71st AIChE Meeting, Miami Beach, Florida. November 14,
1978. p. 3.
4. Ibid. p.4.
5. PEDCo Environmental. Summary Report - Flue Gas Desulfuriza-
tion Systems - June-July 1977. Prepared for the U.S.
Environmental Protection Agency, Contract No. 68-01-4i47,
Cincinnati, Ohio. July 1977. p. 241, 243-244.
6. Melia, M., et al. EPA utility FGD survey August-September,
1978, Preliminary Report. Prepared for U.S. Environmental
Protection Agency, Contract No. 68-02-2603, PEDCo Environ-
mental, Cincinnati, Ohio. November 1978. pp. vii-viii.
7. Op. cit. No. 1
8. Laseke, B.A., Jr., and T.W. Devitt. Status of Flue Gas
Desulfurization Systems in the United States. Presented at
the Flue Gas Desulfurization Symposium, Hollywood, Florida.
November 9, 1977- pp. 2-10.
9. Laseke, B.A., Jr. EPA Utility FGD Survey: December 1977-
January 1978. EPA-600/7-78-051a, PEDCo Environmental,
Cincinnati, Ohio. March 1978. p. 212.
10. Op. cit. No. 3. pp. 24, 27.
11. Slack, A.V., and G.A. Hollinden. Sulfur Dioxide Removal
From Waste Gases, Second Edition. Noyes Data Corporation,
Park Ridge, New Jersey. 1975. pp. 43, 44.
1-20
-------
12. Kaplan, N., M.A. Maxwell, T.W. Devitt, and B.A. Laseke,
Status of Flue Gas Desulfurization on Utility and Industrial
Boilers in the United States and Japan. Presented at the
71st AIChE Meeting, Miami Beach, Florida. November 13,
1978. pp. 24-31.
13. Op. cit. No. 1.
14. Maxwell, M.A., H.W. Elder, and T.M. Morasky. Sulfur Oxides
Control Technology in Japan. Preliminary Report Prepared
for Honorable Henry M. Jackson, Chairman, Senate Committee
on Energy and Natural Resources. Interagency Task Force.
June 30, 1978. pp. 14, 15
15. Op. cit. No. 12. p. 34.
16. Op. cit. No. 14. p. 14.
17. Ibid.
18. Op. cit. No. 12. p. 34.
19. Op. cit. No. 14. pp. 16-19.
20. Op. cit. No. 12. pp. 36-39.
21. Op. cit. No. 14. pp. 16-19.
22. Op. cit. No. 12. pp. 36-39.
23. Op. cit. No. 14. pp. 16-19.
24. Op. cit. No. 12. pp. 36-39.
25. Op. cit. No. 14. pp. 3 -19.
26. Op. cit. No. 12. pp. 36-39.
27. Op. cit. No. 14. pp. 16-19.
28. Op. cit. No. 12. pp. 36-39.
29. Op. cit. No. 14. p. 9.
30. Ibid. p. 12.
31. Op. cit. no. 12. p. 32.
32. Slack, A.V. Technology Development for Power Plant Emission
Control, Survey of Developments in Japan - May-June 1977.
Prepared by SAS Corporation for PEDCo Environmental. August
15, 1977. p. 1.
1-21
-------
33. Ibid.
34. Ibid.
35. Op. cit. No. 14. p. 12.
36. Op. cit. No. 12. p. 32.
37- Op. cit. No. 14. p. 2.
38. Ibid.
39. Ibid.
40. Ibid.
41. Op. cit. No. 14. pp. 16-19.
42. Op. cit. No. 12. pp. 36-39.
43. Op. cit. No. 3. pp. 6, 8, 9.
44. Op. cit. No. 12. pp. 16, 17, 19
45. Op. cit. No. 14. p. 20.
46. Op. cit. No. 12. pp. 38, 39.
47. Op. cit. No. 14. p. 23.
48. Op. cit. No. 12. p. 41.
49. Op. cit. No. 14. p. 23.
50. Op. cit. No. 12. p. 41.
51. Op. cit. No. 14. p. 23.
52. Op. cit. No. 12. p. 41.
53. Op. Cit. No. 14. p. 24.
54. Op. Cit. No. 12. p. 41.
55. Op. Cit. No. 14. p. 24.
56. Op. Cit. No. 12. pp. 41, 42.
57. Op. Cit. No. 14. p. 24.
58. Op. Cit. No. 12. p. 42.
1-22
-------
SECTION 2
OPERATIONAL FGD SYSTEMS
2.1 LIME SLURRY FGD SYSTEMS
As of September 30, 1978, there were 17 operational lime
slurry FGD systems in the United States treating the flue gas
from utility boilers. These systems represent a total net gen-
erating capacity of 6060 MW. A list of the domestic operational
lime slurry FGD installations is presented in Table 2-1 and these
*
units are described in Appendix A. The previously identified
units, Green River 1, 2, and 3 and Bruce Mansfield 1 and 2, are
updated in the following section with recently operational high
performance systems added in Section 2.1.2.
2.1.1 Domestic Lime Scrubbing Units; Previously Identified
Operational Systems
Since September 1977, various design modifications have been
made at the previously identified operational units. These modi-
fications, as well as updated performance information, are pre-
sented in the following two sections.
2.1.1.1 Green River 1, 2, and 3—
Since startup of the FGD system at the Green River Power
Station in September 1975, certain operational problems have
necessitated design modifications. With the exception of outage
time caused by severe winter weather from November 1977 through
March 1978, system operability has been greater than 95 percent
since completion of the modifications. '
Modifications—As previously mentioned, one of the most
severe problems encountered with the FGD system was the frequent
The Shawnee Station of Tennessee Valley Authority is not included
in Table 2-1 because of its experimental use of both lime and
limestone.
2-1
-------
TABLE 2-1. MAJOR DOMESTIC OPERATIONAL LIME SLURRY FGD SYSTEMS
Company
Columbus and Southern
Ohio Electric
Columbus and Southern
Ohio Electric
Duquesne Light
Duquesne Light
Kansas City Power and
Light
Kansas City Power and
Light
Kentucky Utilities
Louisville Gas and
Electric
Louisville Gas and
Electric
Louisville Gas and
Electric
Louisville Gas and
Electric
Minnkota Power
Cooperative
Montana Power
Montana Power
Pennslyvania Power
Pennsylvania Power
Utah Power and Light
Station/Unit
Conesville 5
Conesville 6
Elrama 1 thru 4
Phillips 1 thru 6
Hawthorne 3
Hawthorne 4
Green River 1,2, and
Cane Run 4
Cane Run 5
Mill Creek 3
Paddy ' s Run 6
Milton R. Young 2
Colstrip 1
Colatrip 2
Bruce Mansfield 1
Bruce Mansfield 2
Huntington 1
Size,8
MW
400
400
510
410
100
100
3 64
178
183
425
65
450
360
360
825
825
405
Initial
startup
date
1-77
6-78
10-75
7-73
11-72
8-72
9-75
8-76
12-77
8-78
4-73
9-77
11-75
8-76
4-76
7-77
5-78
New or
retrofit
New
New
Retrofit
Retrofit
Retrofit
Retrofit
Retrofit
Retrofit
Retrofit
New
Retrofit
New
New
New
New
New
New
Coal
sulfur
content, *
4.7
4.7
2.0
2.0
2.0
2.0
3.7
3.75
3.75
3.75
3.75
0.7
0.8
0.8
4.7
4.7
0.5
Design sulfur
dioxide removal
efficiency/ *
89.5
89.5
83
83
70
70
80
85b
85b
85b
80b
75C
60C
60°
92.1
92.1
80
Actual sulfur
dioxide removal
efficiency, *
80 - 90
86 - 89
80 - 99
75
75
95.0
95.0
to
I
tsi
a Net with FGD
b Carbide lime
0 Lime/alkaline fly ash
-------
high loading of acid mist in the scrubber exit gas stream. Such
loading caused acid condensation and rainout in the stack and
immediate plant area. Along with the necessary stack repairs,
replacement of the radial vane mist eliminator with a chevron
type was completed. The reheat system has also been installed by
American Air Filter. This system utilizes extraction steam from
an adjacent unit (No. 5) to heat ambient air. This air is then
injected into the scrubbed gases before they exit through the
scrubber stack. A temperature boost of approximately 28°C
(50°F) is provided.
Other minor modifications have been made to solve various
problems encountered. The smaller balls in the mobile bed con-
tactor were replaced with larger balls to reduce ball migration,
the rubber-lined pump impellers were replaced with Ni-hard im-
pellers, and the lime slaker's degritter was redesigned to reduce
grit pick up.
Figure 2-1 presents a simplified process flow diagram of the
Green River facility. '
Performance—A plant strike at the Green River Power Station
from July 1977 through early November 1977 prevented boiler and
scrubber operations for that period (100 percent FGD availability).
After a return to service, severe winter weather caused numerous
freezeups and outage time. System availability was 86 percent in
November 1977, but dropped to 23 percent in January 1978. The
FGD system was shut down during February and March because emergency
conditions made it necessary for utility personnel to concentrate
their efforts on power generation rather than FGD system repair.
The boiler, however, logged a total of 1341 hours during the two
month period. After being returned to service, the FGD system
showed increasing availability from 41 percent in April to 73
percent in June, with a system operability of 95 percent or
greater through September. Minor problems were encountered when
the screens on the suction side of the.pumps that pump slurry
from the preparation area to the nozzles began plugging because
2-3
-------
to
I
AIR
STEAM
ELECTRICAL
GENERATING
UNIT NO. 1
ELECTRICAL
GENERATING
UNIT NO. 2
BOILER NO. 1
i.o.
FAN
MECHANICAL
COLLECTORS
BOILER NO. 2
fS
M
I-°-O
FAN I/
STACK
-n
X^SCRUBBER^X
STACK
BOILER NO. 3
FA*
BYPASS
DAMPERS
CHEVRON X
> * •* *• "*\ ~
""""I /%?
MAKEUP MATER
.SCRUBBER
/BOOSTER
FAN
MAKEUP WATER
FROM
MOBILE BED
.CONTRACTOR .p
IATER\/^
POND
VENTURI
REACTANT ADDITION
cs*=> eJe=> c»L->
BLEED TO POND
LIME
STORAGE
SILO
MAKEUP UATER
MAKEUP UATER
RECYCLE r^ MX/HOLD TANK
^^
SPARE
SPARE
SPARE
Figure 2-1. Process flow diagram for Green River 1, 2, and 3.
-------
of large grit. Routine maintenance was performed on the system
from July through the middle of August. Recurring plugging
problems in September made it necessary for the FGD system to be
operated at one third of total capacity. Index values for
average system availability, operability, reliability, and utili-
zation for 1977 and 1978 (through September) are presented in
Table 2-2. Table 2-3 presents these index values and averages
for the period September 1977 to September 1978. System oper-
ability is plotted from July 1977 to September 1978 (Figure
2-2).7'8
Removal efficiency—During the period October 17-19, 1978,
PEDCo Environmental, Inc., conducted a series of flue gas tests
at the Green River Plant of Kentucky Utilities. The test program
was conducted to determine scrubber efficiency and compliance
with the State of Kentucky's air pollution regulations for par-
ticulate matter and sulfur dioxide emissions. The sulfur dioxide
tests were run simultaneously at the inlet and outlet to the
scrubber.
In addition to emission data, flue gas volume, temperature,
moisture, carbon dioxide, and oxygen content were also measured.
Coal samples were also taken by plant personnel for proximate and
ultimate analysis. All the tests were run under normal boiler
operating conditions as determined by plant personnel.
A summary of the sulfur dioxide emissions is presented in
Table 2-4.
Sulfur dioxide emissions at the outlet ranged from 482 to
641 ng/J (1.12 to 1.49 lb/106 Btu) and averaged 555 ng/J (1.29
lb/10 Btu); average concentration was 488 parts per million
(ppm). This resulted in an average sulfur dioxide removal
efficiency of 83.3 percent for the test period. The outlet flow
rate averaged 95 m /s (200,754 dry scfm) at an average temp-
erature of 154°C (310°P). Moisture content averaged 13 percent
with an average composition of 11.3 percent carbon dioxide and
9
8.2 percent oxygen.
2-5
-------
TABLE 2-2.
GREEN RIVER POWER STATION, OPERATIONAL DATA
FOR 1977 THROUGH SEPTEMBER 1978
Year
1977
Month
January
February
March
April
May
June
July
August
September
October
November
December
Total
1978
January
February
March
April
May
June
July
August
September
Total
A
Hours
in
period
744
672
744
720
744
720
744
744
720
744
720
744
8760
744
672
744
720
744
720
744
744
720
6552
13
Hours FfiD
system
available
698
243
0
288
736
720
744
744
720
744
634
375
6646
722
672
744
296
474
524
99
454
546
4531
C
Hours FCD
called
upon
744
266
0
167
527
34
0
0
0
0
332
596
2666
537
0
0
295
473
524
99
207
298
D
Hours FGD
system
operated
698
243
0
164
513
34
0
0
0
0
301
375
2328
170
0
0
296
474
524
99
205
298
2433 1 2066
Parameters
E
Hours
boilers
operated
744
266
0
167
527
34
0
0
0
0
332
596
2666
537
672
669
295
474
525
103
207
303
3785
B/A
Avail-
ability, '.
94
36
0
40
99
100
100
100
100
100
88
50
76
97a
100a
iooa
41
64
73
13
61
76
69
D/C
Reli-
ability, %
94
91
0
98
97
100
0
0
0
0
91
63
87
32
0
0
100
100
100
100
99
100
85
D/E
Oper-
ability, %
94
91
0
98
98
100
0
0
0
0
91
63
87
32
0
0
100
100
100
96
99
98
55
D/A
Util *a-
tioi,, %
94
36
0
23
69
5
0
0
0
0
42
50
27
23
0
0
41
64
73
13
28
41
32
Because of emergency conditions, the utility chose to concentrate their maintenance crews on power generation rather
than FGD operation. Under normal conditions the relatively minor FGD system problems could have been solved more
quickly, and the system could have been considered available throughout most of this period.
-------
TABLE 2-3. GREEN RIVER POWER STATION, OPERATIONAL DATA
FOR SEPTEMBER 1977 TO SEPTEMBER 1978
.
Year
1977
1978
Month
September
October
November
December
January
February
March
April
May
June
July
August
September
Total
A
Hours
in
period
720
744
720
744
744
672
744
720
744
720
744
744
720
9480
B
Hours FGD
system
available
720
744
634
375
722
672
744
296
474
524
99
454
546
7004
C
Hours FGD
called
upon
0
0
332
596
537
0
0
295
473
524
99
207
298
3261
D
Hours FGD
system
operated
0
0
301
375
170
0
0
296
474
524
99
205
298
2742
Parameters
E
Hours
boilers
operated
0
0
332
596
537
672
669
295
474
525
103
207
303
4713
B/A
Avail-
ability, %
100
100
88
50
97*
100a
100*
41
64
73
13
61
76
74
D/C
Reli-
ability, %
0
0
91
63
32
0
0
100
100
100
100
99
100
84
D/E
Oper-
ability, %
0
0
91
63
32
0
0
100
100
100
96
99
98
58
D/A
Utiliza-
tion, %
0
0
42
50
23
0
0
41
64
73
13
28
41
29
to
a Because of emergency conditions, the utility chose to concentrate their maintenance crews on power generation rather
than FGD operation. Under normal conditions, the relatively minor FGD system problems would have been solved more
quickly, and the system could have been considered available throughout most of this period.
-------
to
I
CO
c
-------
TABLE 2-4. SUMMARY OF SULFUR DIOXIDE EMISSION RATES
•
Test
No.
1
2
3
4
5
6
Date
1978
10/19
10/19
10/19
10/19
10/19
10/19
Average
Emission rates (outlet)
Concentration ,
ppma
424
449
557
533
504
463
488
fcg/s
(lb/h)
0.10 (839)
0.11 (888)
0.14 (1101)
0.13 (1055)
0.12 (996)
0.12 (917)
0.12 (966)
ng/J
(lb/106 Btu)
482 (1.12)
507 (1.18)
632 (1.47)
598 (1.39)
572 (1.33)
529 (1.23)
555 (1.29)
Emission rates (inlet)
Concentration ,
ppma
2611
2464
2556
2628
2587
2759
2600
kg/s
(lb/h)
0.72 (5701)
0.68 (5379)
0.70 (5581)
0.72 (5738)
0.71 (5648)
0.76 (6024)
0.72 (5678)
ng/J
(lb/106 Btu)
2915 (6.78)
2794 (6.50)
2876 (6.69)
2958 (6.88)
2936 (6.83)
3130 (7.28)
2936 (6.83)
Removal
efficiency, %
86.1
83.8
80.0
81.9
83.1
84.2
83.3
10
I
vo
Parts per million by volume.
-------
2.1.1.2 Bruce Mansfield 1 and 2—
On October 1, 1977, Bruce Mansfield 2 began commercial
operation and brought the current net generating capacity of the
Bruce Mansfield Power Station to 1650 MW (1834 MW gross). This
unit is similar in design to Bruce Mansfield 1, which includes a
Chemico wet lime scrubbing system.
The system consists of six parallel, two-stage, scrubbing
trains. Each train includes a variable-throat venturi scrubber,
a wet induced-draft fan, and a fixed-throat venturi absorber.
The scrubbing trains are arranged in two groups of three. Flue
gas from the three trains in each group flows together into an
oil-fired reheat chamber and then is discharged to the atmosphere
through a 290-m (950-ft) stack. The stack, which serves both
operating units, contains four carbon steel flues for receipt of
the discharge gases from the four oil-fired reheat chambers.
The lime used in the scrubbing operations is a proprietary
reagent, known as Thiosorbic lime, supplied by the Dravo Cor-
poration. This lime, which contains 2 to 6 percent magnesium
oxide, offers the advantage of increased sulfur dioxide removal
efficiency and allows a subsaturated mode of operation.
The flue gas cleaning wastes produced by the scrubbing
systems are treated and disposed of in an environmentally accept-
able manner in a waste disposal system designed and built by the
Dravo Corporation. The waste disposal system consists of a
pumping and treatment facility, a transportation facility, and a
containment area. In the pumping and treatment facility, a
cementitious stabilizing agent, Calcilox, is added to the scrubber
thickener underflow. This mixture is then pumped by pipeline to
a disposal area approximately 11 km (7 mi) west of the power
plant. The disposal area is a ravine with an earthen dam that
creates a reservoir into which the waste slurry is pumped and
deposited on the valley floor under a covering of water.
Additional data for Bruce Mansfield 1 and 2 are presented in
Table 2-5. Figure 2-3 represents a schematic of the process
lines and major components of the FGD systems.
2-10
-------
TABLE 2-5.
FGD SYSTEMS DATA FOR BRUCE MANSFIELD 1 AND 2
SHIPPINGPORT, PENNSYLVANIA
Unit rating (gross), MW
(net), MW
Fuel
Average fuel characteristics:
Heating value, kJ/kg (Btu/lb)
Ash, %
Moisture, %
Sulfur, %
Chloride, %
FGD process
FGD system supplier
Application
Status
Startup date:
Initial
Commercial
Design removal efficiency:
Particulate, %
Sulfur dioxide, %
Actual removal efficiency:
Particulate, %
Sulfur dioxide, %
Water loopa
Total water makeup,
liters/s per net MW
(gal/min per net MW)
Sludge disposal
1834
1650
Coal
27,593 (11,863)
15.11
5.53
2.44
Lime
Chemico
New
Operational
December 1975 (Unit 1)
June 1976 (Unit 1);
October 1977 (Unit 2)
99.8
92.1
99.9
95.0
Open
Stabilized sludge disposed in
an offsite dammed reservoir
For the purposes of this report, closed water loop is defined
as the amount of makeup water entering the system equalling the
water exiting the system in the settled sludge and humidified
flue gas.
2-11
-------
I
t-*
10
Figure 2-3. Schematic of the process lines and major components of the
Bruce Mansfield air quality and waste disposal system.
-------
Problems and solutions—Startup and subsequent operation of
the Bruce Mansfield scrubbing and waste disposal systems have
been accompanied by several design, mechanical, and chemical
problems, especially in the scrubbing systems and related equip-
ment. The utility, in conjunction with the system supplier, has
conceived and implemented solutions to many of these problems.
Some of the problems and resulting modifications were identified
in the previous report.
The scrubbing systems were originally designed so that five
of the six scrubbing trains installed on each unit could handle
total boiler gas flow at a slightly reduced particulate and
sulfur dioxide collection efficiency. Actual operation has
shown, however, that all six scrubbing trains are necessary when
the unit is operating at full load. This has eliminated the
option of servicing one train over a short period of time without
load cutbacks.
The reheaters have never worked properly. At maximum
operating conditions, a resonance pattern created by the oil
burners produced severe duct vibration. The shock wave created
by the oil injection nozzle matched the reasonance frequency of
the ducts. This vibration was so severe that, if permitted to
continue, it would have cracked the ducts and shaken them loose.
The oil injection nozzles were modified by the manufacturer to
eliminate this shock wave. Although this modification was suc-
cessful, the reheaters still cannot fire at a rate that provides
the design temperature increase of 22°C (40°F).
Mist eliminator performance has been a major problem area in
system operation. Although the problems encountered have re-
sulted from complex chemical, mechanical, and design factors, the
major contribution has come from design factors. Mist eliminator
scaling was encountered very early and prompted modification to
•i
the mist eliminator wash system. Moreover, tests conducted in
late 1976 indicated that mist carryover from the mist eliminators
was approximately 7 g/m (3 gr/scf), which is three times higher
2-13
-------
than the maximum design value. Pennsylvania Power and Chemico
experimented with second-stage vertical mist eliminators. Duct
diameter and spatial restrictions caused these mist eliminators
to experience high flow velocities, on the order of 15 m/s (50
ft/s). One experimental vertical mist eliminator was installed
and collapsed because of structural failure. Another vertical
configuration was developed that would operate at lower gas
velocities. Concurrent with this research, model studies per-
formed by Chemico indicated that excessive carryover resulted
when pressure drops exceeding 0.2 kPa (0.75 in. H2O) developed
across horizontal mist eliminators. No carryover was evident
when pressure drops were maintained at 0.1 kPa (0.5 in. H2O) or
less. A new wash system was devised and is activated when the
pressure drop across the mist eliminator exceeds 0.07 kPa (0.3
in. H20). Where the problem is not corrected and the pressure
drop continues to rise, the module is taken out of service; and
the mist eliminator is manually cleaned.
Failure of stack liners is also considered a design problem.
Failure of the original liner material applied to the carbon
steel (Cor-Ten) flues has resulted in half-load operation for 1
year on both units. Originally, the two flues on Unit 1 were
lined with a flaked glass lining. The first 61-m (200-ft) sec-
tion of each flue was lined with a troweled-on flaked glass
material approximately 0.15 cm (60 mils thick). The remaining
213-m (700-ft) section was lined with a sprayed-on flaked glass
material approximately 0.05 cm (20 mils) thick. The bubbles,
widespread failure, and resulting corrosion were most severe in
the top section. For want of other viable alternatives, the
utility replaced the sprayed-on liner with the troweled-on
material. Several test patches were also inserted in one of the
flues of Unit 2. Another inspection of the flues in the spring
of 1978 revealed that one of the flues for Unit 2 had developed a
crack approximately three-quarters of the way around because of
liner failure and acid corrosion attack. This crack had extended
to 90 percent of the circumference by the time the utility re-
2-14
-------
paired it by applying metal cladding to the failed area. The
utility contracted Carnegie Mellon Institute to investigate this
problem thoroughly. Their findings and the results of the test
patch program indicate that a completely suitable liner material
does not exist. Of all the material evaluated, CXL-2000, developed
by Pullman Kellogg, holds the most promise for long-term service.
The utility has used this material for relining three of the four
flues to date. In addition, they plan to reline the remaining
flue with this material when necessity dictates. Because of this
material's ability to withstand a wet corrosive environment,
the reheaters have been shut down permanently.
Another design problem concerns the operation of the wet
induced-draft fans. Although these fans have been beset by a
number of problems that are a combination of chemical, mechanical,
and design related factors, the major problem encountered has
been failure of the construction materials. The pH at this
location has been measured at approximately 2.0. The fan casings
(constructed of rubber-lined carbon steel), the scrolls (con-
structed of carbon steel), and the hubs (constructed of carbon
steel) have been damaged extensively by corrosion and erosion.
The utility is now replacing many of these with components con-
structed of more sophisticated alloys such as Carpenter 20 or
Inconel 625.
Many of the chemical problems that beset the scrubbing
system and related equipment were caused by a faulty pH monitoring
netwprk during the early phases of operation. Primary diffi-
culties involved flow sampling location and glass probe breakage,
which caused pH to be controlled manually during much of the
initial operation stage. This manual control in turn caused
subsequent problems such as scale formation, plugging, and acid
corrosion. The pH monitors were relocated to a different posi-
tion in the recirculation circuit, and sampling procedures were
modified. The results have been excellent. The pH is controlled
within the very narrow band from 6.8 to 7.2. Magnesium ion
2-15
-------
concentration is maintained at a minimum level of 1500 ppm in the
liquid circuit. Sulfur dioxide removal efficiency levels are
consistently above the 92.1 percent design value. Finally, the
modules are operating without any substantial development of hard
scale (gypsum) or plugging, which often affected mist eliminator
performance.
Although the system has been plagued by a number of minor
problems such as pump and valve failures, they have caused very
little outage time.
Waste disposal system—Generally, the operation of the waste
disposal system has proceeded without major incident. Some
problems, however, have been encountered.
The system has never operated in a closed water loop mode.
Supernatant from the reservoir is being discharged into the Ohio
River. One major reason is the great requirement of fresh makeup
water in scrubbing operations (e.g., in the mist eliminator
wash). Another is the quality of the supernatant resulting from
the stabilization process. The pH of the supernatant ranges
from 11.0 to 11.5; the design pH, however, is 8.0. There has
thus been a decreasing emphasis on using the supernatant return
for slurrying and washing. One cause of the high pH is the use
of more stabilizer (Calcilox) than was anticipated when the
system was designed.
Core samplings of the stabilized waste material covering the
reservoir floor indicate different strata of material with
varying physical characteristics. Such strata resulted from not
varying the Calcilox feed rate with varying thickener underflow
characteristics, especially during the initial operation, when
stabilizer was added manually. This situation could prevent the
use of the reservoir as a site for future building construction.
The reservoir is now being filled faster than the design
rate, based on the station's current capacity factor. If this
trend continues, the site may not have ample capacity for three
units with a 35-year service life.
2-16
-------
Performance—In September 1977, Bruce Mansfield 1 was taken
down for six days in order to install rigging in the IB chimney
flue in preparation for flue liner repairs. As a result of these
repairs; only three of the scrubber trains were returned to
service when the unit was brought back on line. The repairs
continued through March 1978, at which time the three modules
were returned to service. Miscellaneous problems encountered
through June 1978 included leaks in fan housings and ducts that
required repair. FGD system dependability factors for Bruce
Mansfield 1 are summarized in Table 2-6.
The Bruce Mansfield 2 FGD system was declared available for
commercial operation on October 1, 1977. Total system avail-
ability during the first month of operation was 80 percent, al-
though premature failure of several test patches in the 2B flue
stack lining necessitated repairs shortly after startup. Avail-
ability reached a high of 97 percent in January 1978, despite
freezing that occurred in the process piping caused by severe
winter weather. The weather also caused problems with the
induced-draft fan coolers during February. Major repair of the
2B stack flue lining began in early April 1978 and lasted through
June 1978; the repair work reduced system availability.
FGD system performance information for Bruce Mansfield 2 is
listed in Table 2-7. System availability for both Bruce Mans-
12 13
field 1 and 2 is depicted graphically in Figures 2-4 and 2-5. '
f
Removal efficiency—Chapter 123.11 of the Pennsylvania
regulations governing the Bruce Mansfield units limits partic-
ulate emissions to 0.019 g/m (0.0175 gr/scf), and sulfur dioxide
emissions to 258 ng/J (0.6 lb/10 Btu) of heat input to the boil-
er. Actual particulate emissions, as measured by the utility,
are 13 ng/J (0.03 lb/10 Btu), below the standard. At Bruce
Mansfield 1, continuous monitoring data indicated sulfur dioxide
removal efficiencies ranging from 60 to 94 percent as a result of
pH control problems. Since better control of scrubbing pH has
been attained, sulfur dioxide removal efficiency has consistently
improved.
2-17
-------
TABLE 2-6. PERFORMANCE DATA ON BRUCE MANSFIELD 1 FGD SYSTEM,
PENNSYLVANIA POWER
1977
1978
September
October
November
December
January
February
March
April
May
June
Boiler
hours
558
720
720
626
331
514
689
720
457
0
Availability,
%
66
46
48
49
33
37
66
82
78
67
Oper ability,
%
75
48
48
49
33
47
72
82
75
0
Utilization,
%
58
46
48
44
19
37
66
82
46
0
to
I
M
00
-------
TABLE 2-7. PERFORMANCE DATA ON BRUCE MANSFIELD 2 FGD SYSTEM,
PENNSYLVANIA POWER
1977 October
November
December
1978 January
February
March
April
May
June
Boiler
hours
595
581
607
391
672
494
713
270
417
Availability,
%
80
72
93
97
89
82
50
50
34
Oper ability,
%
78
74
92
85
75
83
44
50
50
Utilization ,
%
66
60
77
53
75
56
44
18
34
K)
I
M
vo
-------
I
to
100
90
80
70
60
50
40
30
20
10
0
SEPT '77 OCT NOY DEC JAN '78 FEB MAR APR MAY JUNE
Figure 2-4. FGD system availability , Pennsylvania Power, Bruce Mansfield 1
availability index reflects the hours the FGD system is available for
operation (whether operated or not) divided by the total hours in the period
-------
to
I
ro
OCT '77 NOV DEC JAN '78 FEB MAR APR MAY JUNE
Figure 2-5. FGD system availability*, Pennsylvania Power, Bruce Mansfield 2
aThe availability infex reflects the hours the FG^ -system is available for
operation (whether operated or not) divided by the total hours in the period,
-------
Sulfur dioxide emission rate tests were conducted on Units
1 and 2 by the Pennsylvania Department of Energy Resources and by
an independent testing firm hired by Pennsylvania Power. Results
indicated sulfur dioxide emissions as low as 64 ng/J (0.15 lb/10
Btu) .
In May 1977, a program conducted by the U.S. Environmental
Protection Agency involved gathering continuous monitoring data
for sulfur dioxide emission rates at Bruce Mansfield 1. Partial
monitoring began in late August 1977, with the first complete
data beginning in mid-September. One of the chimney flues was
undergoing liner repairs during the test period, causing the unit
to be operated at half load with only three scrubber modules
available. The sulfur dioxide removal statistics for a 24-hour
averaging period indicated that when pH control was a severe
problem, the mean removal efficiency was 81.4 percent, with a
standard deviation of 1.057 (20 test points were used). A second
mean sulfur dioxide removal efficiency obtained from 11 data
points gathered during improved pH control was 85.3 percent, with
14 15
a standard deviation of 1.029. '
2.1.2 Domestic Lime Scrubbing Units: Recently Operational
Systems
2.1.2.1 Conesville 5 and 6—
The Conesville Power Station of Columbus and Southern Ohio
Electric is located on the Muskingum River in Conesville, Ohio.
The plant has a current capacity of 2055 MW (gross). Conesville
1, 2, and 3 have a combined capacity of 433 MW (gross) and share
a common stack. Conesville 4 is rated at 800 MW (gross), and
Conesville 5 and 6 are each rated 411 MW (gross). Conesville 4,
5, and 6 each have a separate stack.
The Conesville 5 and 6 steam generators are dry-bottom, pul-
verized-coal-fired boilers that were supplied by Combustion Engi-
neering. A mixture of high-sulfur Ohio coals is burned in each.
The sulfur content ranges from 4.2 to 5.1 percent; the ash con-
tent, from 12 to 19 percent; and the heating value, from 24,000
2-22
-------
to 26,000 kJ/kg (10,300 to 11,200 Btu/lb). Forty percent of the
coal is delivered by a conveyor from a coal mine complex 11 km (7
mi) from the plant site. The remainder is trucked in from
southeast Ohio.
Each unit develops a flue gas flow of 658 m /min (1,393,893
acfm) at a temperature of 147°C (296°F). The stack for each is
acid-brick-lined and is 244 m (800 ft) tall.16'17
Pollution control—The air pollution control systems at
Conesville 5 and 6 each consist of a Research Cottrell cold-side
ESP, followed by two Universal Oil Products (UOP) sulfur dioxide
absorber modules in parallel. Each ESP is disigned for 99.65
percent particulate removal efficiency; and each Turbulent
Contact Absorber (TCA) system, for 89.6 percent sulfur dioxide
removal efficiency. Each system is designed for an outlet sulfur
dioxide loading of 430 ng/J (1.0 lb/106 Btu of heat input).
Boiler induced-draft (ID) fans are located immediately downstream
of the ESP's.
Following the ID fans, the flue gas enters the two parallel
TCA scrubbing trains. Each absorber is capable of handling 60
percent of the flue gas flow. A presaturator section lowers the
flue gas temperature from 141°C (286°F) to 107°C (125°F) and pro-
vides some initial sulfur dioxide removal. The gas then enters
the neoprene-lined, carbon steel absorber modules. At Conesville
5, one stage of 3.8-cm (1.5-in.) plastic balls provides a con-
tacting surface between the lime slurry and the flue gas. At
Conesville 6, two stages of plastic balls provide the contacting
surfaces in each absorber module. Following each module, the
flue gas passes through a fiberglass entrainment separator and
two horizontal banks of chevron-type mist eliminators. The
bottoms of the trap-out trays are washed intermittently, and the
lower mist eliminators are washed continually with recycled pond
water. The flue gas from the parallel absorber trains then
enters the 244-m (800-ft), Ceilcote-lined stack. Following each
boiler.ID fan is a bypass breeching around the entire scrubber
2-23
-------
loop. Each module can be bypassed independently. No stack gas
reheat is currently employed at either unit.
Dravo Thiosorbic lime from Maysville, Kentucky, is utilized
in the UOP scrubber modules at a stoichiometric ratio of 1.1.
The calcined, pelletized lime has a nominal particle diameter of
4.45 cm (1.75 in.), a magnesium oxide (MgO) content from 3 to 8
percent, and a calcium oxide (CaO) content from 90 to 95 percent.
The lime slaker discharges the 20 percent solids slurry into an
agitated lime slurry sump, where it is retained for 5 minutes
before being transferred to the lime slurry storage tank, which
handles the surge requirements of the absorption system. The
transfer of slurry from the storage tank to the TCA recycle tanks
is accomplished by variable-speed pumps that respond to changing
sulfur dioxide concentrations and boiler load conditions via a pH
monitor. The scrubbing liquor contains from 7 to 12 percent
solids and is recirculated by four pumps (one standby) per unit.
Each pump is rated at 757 liters/s (12,000 gal/min). Scrubber
outlet pH is 5.8, and pH in the recycle tank is approximately
6.8.
A bleed stream of spent reaction products is continuously
withdrawn from the recycle tanks and pumped to the thickener
system. The thickener is 30 m (100 ft) in diameter and 4 m (14
ft) deep in the center. Here the reaction product slurry is
concentrated to an underflow composition of approximately 40
percent solids. This underflow is cycled to fixation facilities
supplied by IU Conversion Systems, Inc. (IUCS). At these facil-
ities, the underflow is thickened, vacuum-filtered, and mixed
with a blend of dry fly ash and lime to form a 73 percent solid
substance (IUCS Poz-o-Tec). The product is currently discharged
to a diked pond measuring 4.32 million m (3500 acre-ft).
The wastewater pond receives ash sluice water, cooling water
blowdown, and water from the sludge treatment plant. This system
is not operating closed loop at the present time. Design infor-
18
mation is presented in Table 2-8.
2-24
-------
TABLE 2-8. FGD SYSTEM DATA FOR CONESVILLE 5 AND 6,
CONESVILLE, OHIO
Unit rating (gross), MW
(net), MW
Fuel
Average fuel characteristics:
Heating value, kJ/kg (Btu/lb)
Ash, %
Moisture, %
Sulfur, %
Chloride, %
FGD process
FGD system supplier
Application
Status
Startup date:
Initial
Commercial
Design removal efficiency:
Particulate, %
Sulfur dioxide, %
Actual removal efficiency:
Particulate, %
Sulfur dioxide, %
Water loop
Total water makeup,
liters/s per net MW
(gal/min per net MW)
Sludge disposal
411
375
Bituminous coal
25,240 (10,850)
7.5
4.7
Lime (Thiosorbic)
Air Correction Division, UOP
New
Operational
January 1977
June 1978
99.6
89.5
Open
0.08
1.25
Stabilized sludge disposed
of in an onsite landfill
2-25
-------
Figure 2-6 presents a simplified process flow diagram of the
Conesville FGD system.
Performance history—The A-module of Conesville 5 was placed
in service in November 1977, after startup was delayed by a fire.
Conesville 5 was taken out of service in December 1977 for a
scheduled 3-month overhaul, and the carbon steel stack liner was
replaced with an acid brick liner.
Conesville 5 was returned to service in March 1978. Problems
were then encountered with the thickener. An excess of floccu-
lant was caused by a malfunction of the feed system, and the tank
had to be drained.
From March through September 1978, system availability
averaged 44 percent, with a high of 66 percent during April. A
low of 18 percent in August was caused by scale that built up in
the mist eliminator and required removal, and by outage time for
replacement of some of the balls in the TCS regions. Figure 2-7
illustrates system operability from September 1977 to September
1978. Table 2-9 lists additional performance information.
Conesville 6 was placed in service in June 1978. Initial
problems included a loss of control of the lowered dampers to the
bypass system and back pressure buildups that caused automatic
boiler shutdown. In July, the damper controls were adjusted in
an attempt to correct the problem. Plugging problems have been
experienced with the fiberglass reinforced plastic (FRP) transfer
line from the thickener to the IUCS system. The utility has
reported that the bypass control dampers as well as the sludge
line tend to be high maintenance areas.
Through September 1978, Conesville 6 system availability
averaged 61 percent. A high of 77 percent occurred in July; and
a low of 50 percent in June. Since startup, monthly operating
hours have increased from 174 in June to 372 in September.
System operability is plotted in Figure 2-8 and additional
performance data is indicated in Table 2-10. '
2-26
-------
TO STACK - aUE 6ASES
MAKEUP HATER
(POND MATER)
10
I
OVERFLOW
TO ASH POND
FROM OTHER
RECYCLE TANK
LIME FROM
STORAGE
SERVICE
HATER
TO IUCS
(Note:
Figure 2-6. Conesville 5 and 6, simplified process diagram.
The second stage of packing in the Conesville 5 TCA towers has been removed)
-------
I
M
00
OCT NOV DEC JAN FEB MAR APR HAY JUN JUL AUG SEP
Figure 2-7. 'Scrubber system operability , Columbus and Southern
Ohio Electric, Conesville 5.
The operability index reflects the hours the FGD system was
operated divided by boiler operating hours in the period.
-------
TABLE 2-9. PERFORMANCE INFORMATION ON CONESVILLE 5, COLUMBUS AND SOUTHERN OHIO ELECTRIC
October 1977
November
December
January 1978
February
March
April
May
June
July
August
September
Boiler
hours
559
715
0
0
379
716
720
720
727
667
707
Availability,
%
16
22
95
0
0
20
66
53
39
55
18
58
FGD 33
Oper ability,
%
12
18
72
0
0
18
58
48
34
50
20
45
rstem
Reliability,
%
12
18
69
0
0
59
62
48
34
50
20
54
Utilization,
%
9
18
24
0
0
9
58
46
34
48
18
44
10
I
to
vo
-------
100
90
80
70
. 60
>-
CO
CL.
O
50
40
30
20
10
I
I
JUN JUL AUG SEP
1978
Figure 2-8. Scrubber System Operability , Columbus and Southern Ohio
Electric, Conesville 6.
aThe operability index reflects the hours the FGD system was operated
divided by boiler operating hours in the period.
2-30
-------
TABLE 2-10. PERFORMANCE DATA ON CONESVILLE 6,
COLUMBUS AND SOUTHERN OHIO ELECTRIC
June 1978
July
August
September
Boiler
hours
524
502
642
706
Availability ,
%
50
76
55
62
Oper ability,
%
41
28
55
52
Reliability,
%
42
48
74
55
Utilization,
%
36
19
48
52
to
I
U)
-------
2.1.2.2 Cane Run 4 and 5—
The Cane Run power station, located in Louisville, Kentucky,
is operated by the Louisville Gas and Electric Company. The
plant has six electric power steam generating units, providing
a total steam turbine net generating capacity of 992 MW.
Cane Run 4 and 5 are coal-fired boilers with continuous net
generating capacities of 178 and 183 MW, respectively. Cane Run
4 has a maximum power generation capacity of 190 MW; and Cane Run
5, a maximum power generating capacity of 200 MW. The coal
burned at both units has an average heating value of 26,750 kJ/kg
(11,500 Btu/lb) and average sulfur and ash contents of 3.8 and
15.5 percent, respectively.
Pollution control—The emission control system at Cane Run 4
consists of an electrostatic precipitator (ESP) upstream of a wet
scrubbing system. The ESP provides primary particulate control
while the wet scrubbing system provides additional particulate
removal and primary sulfur dioxide control.
The FGD system consists of two identical, parallel, mobile
bed absorber modules designed and installed by the American Air
Filter Company (AAF). Each module is lined with precrete and
Placite 4005 and utilizes polyurethane balls 3.18 cm (1.25 in.)
in diameter as the packing material. The total scrubber system
gas flow is 346 m /sec (734,000 acfm) at a temperature of 160°C
(320°F). The liquid-to-gas ratio in each absorber varies from 7
liters/m3 (55 gal/103 ft3) to 9 liters/m3 (65 gal/103 ft3).
The wet scrubbing system utilizes a slurry of carbide lime,
a waste product from a nearby acetylene manufacturing plant. The
hydrated lime contains from 90 to 92 percent calcium hydroxide,
2.0 to 2.5 percent silica, and 3 to 8 percent calcium carbonate.
It also contains 0.1 percent magnesium oxide.
From the FGD system, the scrubbed gases pass through a two-
bank chevron-type mist eliminator lined with Plasite 4005 before
exiting through a 76-m (-250-ft) stack lined with precrete.
Direct combustion reheat using oil as the combustion fuel pro-
vides a temperature boost of 28°C (50°F) to the exiting gases.
2-32
-------
Spent absorbent from the FGD system is discharged to a
thickener that is 23 m (75 ft) in diameter. The thickener
underflow is dewatered by a vacuum filter, mixed with fly ash and
lime, and disposed of in an onsite, clay-lined pond. The dis-
posed material is 6 percent fly ash, 24 percent bottom ash, 33
percent sulfur dioxide waste, 3 percent unreacted reagent, and 33
percent water. The system operates with an open water to which
makeup water is added at 6 liter/s (100 gal/min).
Additional system design information is presented in Table
2-11. Figure 2-9 provides a simplified process flow diagram.
The emission control system at Cane Run 5 was supplied by
Combustion Engineering and consists of two spray towers down-
stream of the existing electrostatic precipitator, which provides
primary particulate control. Carbide lime is utilized as the
scrubbing reagent and is injected into each spray tower through a
series of three spray stages. At full load, the gas velocity is
3 m/s (10 ft/s), with a pressure drop of approximately 0.75 kPa
(3 in. H2O) across the absorber. The FGD system capacity is 330
m3/sec (700,000 acfm) at 154°C (310°F). The absorber liquid-to-
3 33
gas ratio is from 8 to 13 liters/m (60 to 100 gal/10 ft ) . An
inline steam reheat system provides a temperature boost of 22°C
(40°F) to the exiting flue gas. The FGD system produces 16
liters/sec (248 gal/min) of spent absorbent, which is discharged
at 20 percent solids to a thickener 33.5 m (110 ft) in diameter.
The sludge is ultimately stabilized and disposed of in an onsite
clay-lined disposal pond.
Additional system information is presented in Table 2-12.
22 23 24 25
Figure 2-10 provides a simplified process flow diagram. ' »* »*
Performance history—After startup of the FGD system at Cane
Run 4, many problems made it necessary for certain design modifi-
cations to be performed. Major problem areas included poor gas
flow distribution and excessive pressure drop, which caused the
booster fans to lack sufficient capacity and the boiler output to
be limited to a maximum between 150 and 155 MW. The unit has a
2-33
-------
TABLE 2-11. FGD SYSTEMS DATA FOR CANE RUN 4,
LOUISVILLE, KENTUCKY
Unit rating (gross), MW
(net), MW
Fuel
Average fuel characteristics:
Heating value, kJ/kg (Btu/lb)
Ash, %
Moisture, %
Sulfur, %
Chloride, %
FGD process
FGD system supplier
Application
Status
Startup date:
Initial
Commercial
Design removal efficiency:
Particulate, %
Sulfur dioxide, %
Actual removal efficiency:
Particulate, %
Sulfur dioxide, %
Water loop
Total water makeup,
liters/s per net MW
(gal/min per net MW)
Sludge disposal
190
178
Coal
26,750
15.5
3.8
(11,500)
Lime (carbide)
American Air Filter
Retrofit
Operational
August 1976
99.0
85.0
99.0
86-89
Open
0.04
(0.6)
Stabilized sludge disposed of in
an onsitie sludae tjond
2-34
-------
:NCIKB
HIST
ELIMINATOR
(CHEVRON)
Figure 2-9. Simplified process flow diagram, Cane Run 4
ELECTROSTATIC
PRECIPITATOR
•OILER
FLUE.
CAS
"TO STACK
-STEAM
REHEAT
LINE SLURRY
FEED
TRANSFER
t *
REACTION
TANK
V
RECYCLE
PUNP
^
>
I
1 TNtCKI
;[
Figure 2-10. Simplified process flow diagram, Cane Run 5
2-35
-------
TABLE 2-12. FGD SYSTEMS DATA FOR CANE RUN 5,
LOUISVILLE, KENTUCKY
Unit rating (gross), MW
(net), MW
Fuel
Average fuel characteristics:
Heating value, kj/kg (Btu/lb)
Ash, %
Moisture, %
Sulfur, %
Chloride, %
FGD process
FGD system supplier
Application
Status
Startup date:
Initial
Commercial
Design removal efficiency:
Particulate, %
Sulfur dioxide, %
Actual removal efficiency:
Particulate, %
Sulfur dioxide, %
Water loop
Total water makeup,
liters/s per net MW
(gal/min per net MW)
Sludge disposal
200
183
Coal
26,750
15.5
3.8
(11,500)
Lime
Combustion Engineering
Retrofit
Operational
December 1977
June 1978
99.0
85.0
Open
Stabilized sludge disposed in
an rma-i
2-36
-------
gross rating of 190 MW. Modifications included the installation
of turning vanes in the flooded elbow and base of the scrubbers.
The mist eliminator was converted to a chevron type by cutting
away portions of the mist eliminator wheel.
Another problem was the low liquid-to-gas ratio of approx-
3 33
innately 4 liters/m (30 gal/10 ft ) . At high temperatures, the
original spray nozzle casings expanded so much that the spinner
vanes stuck out the front and caused the nozzle to plug. To
correct this problem, the nozzles were replaced with ones con-
structed of sturdier ceramic material. More spray headers were
also installed above the mobile beds, and the pump capacity was
increased. The liquid-to-gas ratio was brought to approximately
8 liters/m3 (60 gal/103 ft3).
Other modifications were made to help improve overall sulfur
dioxide removal efficiency. Underbed sprays were added to im-
prove circulation of the mobile bed balls; and existing pH
meters were replaced so that the pH could be monitored more
precisely.
The modifications to the FGD system were completed in July
1977. Early in August, the system was tested for compliance with
Jefferson County and Federal sulfur dioxide air emission regula-
tions. The modifications enabled the system to meet the Jefferson
County removal requirement of 85 percent and the Federal standard
of 515.9 ng/J (1.2 lb/106 Btu). The testing was handled by EPA
personnel and indicated a sulfur dioxide removal efficiency
between 86 and 89 percent for coal containing from 3.3 to 3.4
percent sulfur. This efficiency is equivalent to an outlet
emission value of 343.9 ng/J (0.8 lb/10 Btu).
The operability of the FGD system was greater than 90 per-
cent from August through December 1977. After a lime feed line
froze late in December, the system operated intermittently
through January 1978; but there were no scrubber breakdowns.
A coal strike forced the unit to be shut down during February
1978. It came back on line in March, and no FGD system forced
outages have been reported through September 1978.
2-37
-------
Figure 2-11 illustrates system operability from September
1977 through September 1978. System operating information is
presented in Table 2-13.
Initial operation of the FGD system at Cane Run 5 began on
December 29, 1977. Some minor modifications were necessary for
controls that were not operating properly. Cane Run 5 was re-
moved from service in February 1978 because of a coal strike and
was restarted in March. The FGD system logged approximately 91
hours in March, although initial startup problems still caused
intermittent operation. From April through September 1978,
operability was 80 percent or greater. Trouble with the reheat
steam coils caused some outage time. The welds at the end of
each bank of coils have been fouling since initial operation.
Performance tests were conducted from April through May
1978. Reports by the utility indicate that the tests were
inconclusive because of poor handling of data and improper test
procedures.
FGD system operability is illustrated in Figure 2-12 with
2fi 27 28
additional performance data presented in Table 2-14. ' '
2-38
-------
to
I
OJ
\D
100
90
80
70
60|-
a.
°
40
30
20
10
UNIT OUTAGE DUE
TO COAL SHORTAGE
1
L
I
SEP OCT NOV DEC
1977
JAN FEB MAR APR
month/year
MAY
1978
JUN JUL
AUG
SE
Figure 2-11. Scrubber system operability3, Louisville Gas and Electric, Cane Run 4
aThe operability index reflects' the hours the FGD system was operated divided by
boiler operating hours in the period.
-------
TABLE 2-13. OPERATING DATA ON CANE RUN 4,
LOUISVILLE GAS AND ELECTRIC
September 1977
October
November
December
January 1978
February
March
April
May
June
July
August
September
Period
hours
720
744
720
744
744
672
744
720
744
720
744
744
720
Boiler
hours
529
677
483
715
—
0
-
303
352
720
687
744
138
Scrubber
hours
524
662
454
608
-
0
249
303
115
715
678
701
138
Performance factors, %
Operability
99
98
94
85
67
0
-
100
35
99
99
94
100
Utilization
99
89
63
82
67
0
34
47
12
99
91
94
19
to
I
-------
100
90-
80
70
60
5 50
a.
o
40
30
20
10
I
APR MAY JUN JUL AUG SEP
1978
i .
Figure 2-12. Scrubber system operability , Louisville Gas
and Electric, Cane Run 5.
The operability index reflects the hours the FGD system was
operated divided by boiler operating hours in the period.
2-41
-------
TABLE 2-14. OPERATING DATA ON CANE RUN 5,
LOUISVILLE GAS AND ELECTRIC
April 1978
May
June
July
August
September
Period
hours
720
744
720
744
744
720
Boiler
hours
669
432
685
632
540
609
Scrubber
hours
648
364
590
506
464
485
Performance
Operability
97
84
86
80
86
80
factors, %
Utilization
90
49
82
68
62
67
to
I
Ji.
to
-------
REFERENCES FOR SECTION 2.1
1. Melia, M.f et al. EPA Utility FGD Survey August-September
1978. Preliminary report. Prepared for the U.S. Environ-
mental Protection Agency under Contract No. 68-02-2603,
PEDCo Environmental, Inc., Cincinnati, Ohio. November 1978.
p. 86.
2. Ibid. pp. 22-82.
3. Laseke, B.A., Jr. EPA Utility FGD Survey: December 1977-
January 1978. EPA-600/7-78-051a. PEDCo Environmental, Inc.,
Cincinnati, Ohio. March 1978. p. 72.
4. Op. cit. No. 1. pp. 42-43.
5. Laseke, B.A., Jr. Survey of Flue Gas Desulfurization Systems
Green River Station, Kentucky Utilities. EPA-600/7-78-048e.
PEDCo Environmental, Inc., Cincinnati, Ohio. March 1978.
pp. 20-24.
6. Beard, J.B. Scrubber Experience at the Kentucky Utilities
Company Green River Power Station. In: Proceedings of the
Symposium on Flue Gas Desulfurization-Hollywood, Fl., Novem-
ber 1977, (Volume 1), EPA-600/78-058a. March 1978.
pp. 248-249.
7. Op. cit. No. 3. p. 72.
8. Op. cit. No. 1. pp. 42-43.
9. Campbell, R.L. Emissions Testing Report, Kentucky Utilities
Green River Station, Central City, Kentucky. PEDCO Environ-
mental, Cincinnati, Ohio. October 1978.
r
10. Laseke, B.A., Jr. Survey of Flue Gas Desulfurization Sys-
tem: Bruce Mansfield Station, Pennsylvania Power Company.
Preliminary report. Prepared for the U.S. Environmental
Protection Agency under Contract No. 68-02-2£03, PEDCo
Environmental, Inc., Cincinnati, Ohio. November 1978.
pp. vii-x.
11. Ibid. pp. 50-55.
12. Op. cit. No-. 3. pp. 119-122.
13. Op. cit. No. 1. pp. 63-66.
2-43
-------
14. Kelly, W.E., et al. Air Pollution Emission Test: Volume I:
First Interim Report, Continuous Sulfur Dioxide Monitoring
at Steam Generators. Emission Measurement Branch Report No.
77 SSP 23A, the U.S. Environmental Protection Agency, Research
Triangle Park, North Carolina. August 1978. pp. ix-2, 51.
15. PEDCo in-house files.
16. Devitt, T., et al. Flue Gas Desulfurization System Capa-
bilities for Coal-Fired Steam Generators. EPA-600/7-78-
032b. PEDCo Environmental, Inc., Cincinnati, Ohio. March
1978. pp. A-15.
17. PEDCo in-house files.
18. Op. cit. No. 16. pp. A-15 through A-17.
19. Op. cit. No. 3. p. 32.
20. Op. cit. No. 1. pp. 27-29.
21. PEDCo in-house files.
22. Ibid.
23. Op. cit. No. 16. pp. A-9 through A-ll.
24. Laseke, B.A., Jr. (PEDCo Environmental, Inc.). Notes on
Trip to Louisville Gas and Electric Company's Cane Run
Station. March 31, 1978.
25. Laseke, B.A., et al. Utility FGD Costs: Reported and
Adjusted Costs for Operating FGD Systems. Preliminary
report. Prepared for the U.S. Environmental Protection
Agency under Contract No. 68-02-2603, PEDCo Environmental,
Inc., Cincinnati, Ohio. September 1978. Form 9, p. 3.
26. Op. cit. No. 3. p. 76.
27. Op. cit. No. 1. pp. 4.4-45.
28. Op. cit. No. 24.
2-44
-------
2.2 LIMESTONE SLURRY FGD SYSTEMS
There are currently 20 operational units totalling 8514 MW
(net) in the United States that are equipped with FGD systems.
A list of the domestic operational units employing FGD
systems appears in Table 2-15. These units are described in
Appendix B, although some unit descriptions lack detail as a
result of the unavailability of data from the utility or system
supplier. The previously identified units, La Cygne 1 and
Sherburne 1 and 2, are brought up to date in Section 2.2.1; and
more recently operational systems are described in Section 2.2.2.
2.2.1 Domestic Units; Previously Identified Operational Systems
Since September 1977, a number of design and maintenance
modifications have taken place on the previously identified FGD
systems at La Cygne 1 and Sherburne 1 and 2. These units are
addressed in the text that follows.
1-4
2.2.1.1 Kansas City Power and Light, La Cygne
Background—As a result of continuing modifications and
improved operating procedures, the FGD system availability at La
Cygne 1 has steadily improved since startup in 1973. Without
time deductions for modifications, FGD system availability can
easily average 95 percent. With the addition of the eighth
module in April 1977, continuous daytime load capability now
exceeds 800 MW without appreciably affecting scrubber operation.
The results of performance tests in August 1977 indicated
that module gas flow was still below maximum design capability,
that the induced- and forced-draft fans were loaded well below
rating, and that most subsystems were in proper balance. Proper
maintenance improved limestone utilization and control of scaling.
Because of the utility's effort to improve operating proce-
dures and stabilize equipment operation, the scrubber operating
and maintenance force was increased.
The text that follows summarizes the operating history and
performance of the scrubbing system, the problems encountered and
solutions, the current research and development, and the future
initiatives that are planned.
2-45
-------
TABLE 2-15. MAJOR DOMESTIC OPERATIONAL
LIMESTONE FGD INSTALLATIONS
Company
Alabama Electric
Arizonia Electric Power
Arizonia Public Service
Arnonia Public Service
Central Illinois Light
Indianapolis Power and
Light
Gulf Power
Kansas Power and Light
Kansas Power and Light
Kansas Power and Light
Kansas Power and Light
Northern States Power
Co.
Northern States Power
Co.
South Carolina Public
Service
Southern Mississippi
electric
Springfield City
Utilities
Tennessee Valley
Authority
Texas Utilities
Texas Utilities
Texas Utilities
Stat lon/un i t
Tomb i qbee 2
Apachee 2
Choi la 1
Cholla 2
Duck Creek 1
Petersburg 3
Sholt 1 and 2
LaCygne 1
Jeffrey 1
Lawrence 4
Lawrence 5
Sherburne 1
Sherburne 2
Hinyah 2
H.D. Morrow 1
Southwest 1
Widows Creek 8
Martin Lake 1
Martin Lake 2
Monticello )
S,,.',
MH
2JS
200
US
250
400
515
23°
820
680
115
400
680
680
2S8
180
173
SSO
7SO
750
750
Initial
startup
date
8-78
9-78
10-71
6-78
8-78
10-78
8-78
2-73
8-78
12-68
11-71
3-76
1-77
7-77
8-78
4-77
S-77
10-77
S-78
5-78
Mew >r
retrofit
New
New
Retrofit
New
New
New
Retrofit
New
New
Retrofit
New
New
New
New
New
New
Retrofit
New
New
New
Coal
sulfur
content, t
1.15
0.7
0.4 1
u.< 1
2.5 }
3.25
2.0
5.U
0.3
0.5
0.5
0.8
0.8
1.0
l.u
3.5
j.7
1.0
0.9
0.9
Design sulfur
dioxide removal
efficiency, t
60
85
92b
75
• 5
• 5
90
76
to
76
52
50d
50"
70
15
(0
10
60
71
74
Actual sulfur
dioxide reenval
efficiency, «
50 - «0
91. «
10.1
90*
50 - 55
50 - 55
15
92
15 - 94
71. »
* Net with PCD (except where otherwise indicated).
b The A-side train includes the packed absorber with 1 IMS tone slurry circulated through the moduls.
SO2 removal for the A-side is 92 percent. The B-side train does not include packing or limestone
slurry and removes approximately 25 percent of the SO}. Total system removal efficiency is 58.5
percent.
FGD system size (prototype unit).
The FGD system also collects fly ash rich in CaO which contributes to SOj resoval.
2-46
-------
Problems and modifications—La Cygne 1 has undergone a
number of modifications in an effort to improve system perfor-
mance. Although many problems have remained unresolved and more
problems are surfacing, some have been eradicated.
Improved maintenance procedures in early 1977 (acid flushing,
sonic cleaning, and periodic water backflushing) have produced
linear pH response, reduced limestone usage by approximately 30
percent, and improved scaling control. In the past, it was
nearly impossible to operate the inline glass cells that monitor
pH without caking the limestone during shutdown or eroding the
cells during operation with high concentrations of fly ash in the
slurry.
The redesigning of the original mist eliminator wash system,
which operated with sprays of blended freshwater and recycled
pondwater, has eliminated plugging and scaling. The intermittent
system was removed; and the continuous system, which delivered
water at 8.8 liters/s (140 gal/min) , was moved from underneath to
above the first mist eliminator of the two-stage system. More
nozzles were added to the continuous wash system, so that it
delivers water at a rate of 15 liters/s (230 gal/min). As of
September 1978, the utility was studying some new two-stage and
three-stage mist eliminators installed in the the scrubbing
system. More efficient mist elimination would lower the fre-
quency of reheater cleaning and reduce loss of slurry and related,
FGD-promoted, particulate emissions.
The induced-draft (ID) fans are located downstream from the
reheater. Carryover to the blades of these fans continues to
require regular fan washings. Each fan now requires cleaning
every 4 to 7 days. A new cleaning process, a "spinning" wash
with low-pressure hoses, has made the high-pressure wash seldom
necessary. The scrubber system includes six parallel induced-
draft fans, one of which is held as a spare on a rotating basis.
Fans are cleaned when they are out of service as spares.
After approximately 5.5 years of commercial operation, some
materials that have not been modified are wearing out. Rubber
2-47
-------
pipe linings and rubber-lined pumps have been an increasing
maintenance problem. Rubber linings that tear out cause damage
in other piping or pumps, plug nozzles, and allow the steel pipes
to wear through. Wear did not present a significant problem in
the past; however, the constant flow of abrasive slurry has
slowly eroded some parts of the system. Module availability will
suffer as repairs become necessary.
Corrosion of carbon steel in the ductwork, dampers, induced-
draft fan rotors and housings, breeching, and stack liner is
still a major problem. Burning high-sulfur coal in conjunction
with module outages can create high sulfur trioxide concentra-
tions on these surfaces. This "cold-end corrosion" damage
requires extensive surveillance by maintenance engineers, and
unit outages should be planned to take into consideration temper-
atures and time requirements for application of special coatings.
The scrubber operating and maintenance force has been in-
creased to 54 people with the addition of one electrician and two
technicians. The continued improvements in operating procedures
and stable equipment operation should permit analysis of improved
chemistry and control parameters. The utility has reported that
the increase in staff is necessary if any further progress is to
be made, particularly in maintaining pH cells and sulfur dioxide
analyzers.
There have also been increased demands upon maintenance
personnel to accumulate, record, and evaluate operating data on
water saturation trends, limestone utilization, draft fan wear
rates, failures of reheater steam tubes and lined pumps, replace-
ments of nozzles and rubber-lined pipe, use of spare parts/ etc.
The operators are also busy updating operational and special
instructions and reviewing safety and training procedures.
Other proposed modifications include the following:
Installation of an improved steam source to increase
the supply for module reheater service from 32,000 to
54,000 kg/h (70,000 to 120,000 Ib/h) and permit addi-
tional steam bundles for optimum reheat.
2-48
-------
0 Construction of an additional sludge pond for deposit
of scrubber spent slurry for approximately 30 years.
Adding another disposal pond might also allow clear
water to be recycled to the scrubber for improved
chemistry.
0 Installation of a second mist eliminator in the remain-
ing five modules and evaluation of the possible addi-
tion of a third mist eliminator. (This was not possi-
ble in the past because of the early problems with
scaling.)
0 Initiation of a study on the submicrometer fly ash and
sulphuric acid mist that pass through the scrubber
without being collected.
0 Continuation of work to devise a better method to clean
inline reheat tubes without taking equipment out of
service.
Performance—La Cygne 1's FGD system availability has stead-
ily improved since startup in 1973. The averages have been 76.3,
84.3, 92.0, and 92.5 percent for 1974, 1975, 1976, and 1977,
respectively. From January through September 1978, the average
availability was 93.2 percent.
Sulfur dioxide removal efficiency averaged 77 percent with
individual modules averaging from 65 to 80 percent. Particulate
emissions from the plant have met EPA and Kansas State require-
ments. However, work is continuing to reduce the particulate
emissions from this unit.
Current operating experience reports have indicated no major
scrubber problems. No FGD-related forced outages were reported
for La Cygne 1 from September 1977 to September 1978 (latest data
available for this report). Outages reported were a result of
boiler and turbine problems and, in one case, a low demand for
power. During the scheduled and unscheduled (forced) boiler
outages scrubber work was carried out. The flue gas cleaning
system has primarily required routine maintenance. The only
modifications have been a steam coil bundle replacement in the
reheat system and the mist eliminator modifications mentioned
earlier.
2-49
-------
The improved mist elimination is expected to yield better
system performance, lower the frequency of reheater cleaning, and
reduce the loss of slurry and related FGD-promoted particulate
emissions. Figure 2-13 summarizes La Cygne 1 availability data.
As shown, availability for September 1977 through September 1978
averaged about 94 percent. Figure 2-14 shows an updated flow
diagram of one of the eight FGD modules at La Cygne 1 and Figure
2-15 shows a diagram of the overall installation.
Sulfur dioxide removal efficiency for the whole installation
averaged 77 percent in an August 1977 test, and individual modules
averaged between 65 and 80 percent. Table 2-16 shows the test
results.
5—8
2.2.1.2 Northern States Power- Sherburne 1 and 2
Background—The operations of the Sherburne scrubbing sys-
tems have been rather successful. The success has, in part,
resulted from the continual efforts by Northern States Power
(NSP) and Combustion Engineering to improve scrubber performance
through design and operation modifications, some of which were
included in the previous report.
The text that follows summarizes the operating history and
performance of the scrubbing system, problems encountered and
solutions, current research and development, and the future
initiatives that are planned.
Problems and solutions—The performance of the scrubber
systems has exceeded minimum design expectations and has steadily
improved over the course of the operating period (up to September
1978). This improvement has come from the accumulation of oper-
ating experience, the resolution of specific problems, and design
modifications undertaken to resolve or prevent operating problems.
Many problems have been encountered, and some continue to reduce
system availability. A discussion of the major problems and
solutions and modificiations that have resulted is given in the
following paragraphs.
2-50
-------
10
100
90
CO
-------
TO STACK
FLUE GAS FROM
AIR HEATER
HYDROCLONE
SLUDGE
TO POND
I
STEAM :,v=V^
REHEATER
HOT AIR FROM (NOT USED IN
AIR HEATER "D" MODULE)
MIST ELIMINATOR
.. CONTINUOUS WASH
VENTURI
SIEVE
TRAYS
S02 ABSORBER
SCREEN-
\
IT-I 1
VENT
D-i-0
D-K3
-WATER WASH
STAGE
I
RECIRCULATION TANK
WATER FROM POND
WATER MAKEUP
LIMESTONE SLURRY
Figure 2-14. Flow diagram of one of the eight
FGD modules at La Cygne 1.
2-52
-------
fO
I
en
oo
SOOT BLOWER
ABSORBER pnNn
RETIRCULATION D^D
PUMPS RETURN
AIR
HEATERS TO
RECIRCULATIO
TANKS
ENTURI
SCRUBBER
VENTURI
RECIRCULATION
PUMPS
HYDROCLONE
CLASSIFIER
LIMESTONE
BUNKER
LIMESTONE
SLURRY x
STORAGE^
BOILER
SETTLING POND
WET BALL
MILL
Figure 2-15. La Cygne limestone wet scrubbing system.
-------
TABLE 2-16. LA CYGNE STATION UNIT NO. 1 4-HOUR FULL LOAD AND STACK EMISSION TEST
Date: August 26, 1977
Time: 11:00 a.m. - 12:00 midnight
Load range: 800+ MW
Ambient temperature: 34 °C (94°F)
NOX emission:
348 ng/J (0.81 lb/10 Btu)
Average SO2 removal: 77*
Particulate emission: 91.6 nq/J (0.213 lb/106 Btu)
Gas flow indicated, m3/s (acfm)
Throat position
Reheat temperature, "C (°F)
Venturi pressure change, kPa
(in. H20)
Reheater pressure change, kPa
(in. H20)
Absorber mist eliminator
pressure change, kPa (in. H2O)
Reheat outlet damper position,
% open
ID fan load, Amps, (maximum)
Inlet damper position of ID fan
% open
FD fan load. Amps, (maximum)
Laboratory pH
Sulfite concentration, g/liter
Carbonate concentration, g/liter
SOj removal efficiency, %
Inlet SO2, ppm
Outlet SO2, ppm
TEST RUNS
A
0.169(400)
Open
77(170)
1.2(5)
0.62(2)
1.6(6.5)
50
380
42
490
5.45
60.4
50.3
80.0
4600
920
B
0.165(350)
Open
88(190)
1.4(5.5)
1.4(5.5)
1.4(5.5)
100
420
42
470
5.7
72.4
75.6
82.1
4600
825
C
0.179(380)
Open
66(150)
1.2(5)
1.1(4.5)
2.5(10)
96
380
32
430
5.55
101.0
53.1
74.9
4600
1150
D
0.189(400)
Open
88(190)
1.2(5)
1.1(4.5)
1.9(7.5)
3fi
400
36
5.7
74.1
54.4
64.3
4600
2285
E
0.166(352)
Open
85(185)
1.2(5)
1.2(5)
1.7(7)
100
470
36
5.58
70.0
59.4
76.4
4600
1085
F
0.179(380)
Open
82(180)
1.2(5)
0.63(2.55)
1.6(6.5)
52
470
40
5.77
43.9
83.8
72.1
4600
1285
G
0.175(370)
Open
71(160)
1.2(5)
1.1(4.5)
2.0(8)
100
5.72
43.9
68.1
73.1
4600
1235
H
0.173(366)
Open
77(170)
1.2(5)
1.4(5.5)
1.7(7)
100
5.29
63.6
42.5
74.8
4600
1160
I
Ul
-------
Original scrubber equipment included slurry line strainers,
comminuters, and perforated plates that were designed to remove
large chunks and solid particles from the slurry circuit and thus
prevent nozzle plugging. The arrangement of this equipment for
each module is illustrated in Figure 2-16. The failure of the
equipment to perform its function has contributed to nozzle
plugging, the most significant problem affecting system avail-
ability.
The original strainers were Zurn duplex units, located on
the discharge of the slurry spray pumps; these units were de-
signed to remove solid particles larger than 6.4 mm (0.25 in.) in
diameter. In operation, however, frequent episodes of strainer
plugging, mechanical failures, and inability to remove large
particles caused severe nozzle plugging and low system avail-
ability. The duplex design also resulted in impractical opera-
tion and maintenance requirements. After extensive, futile
efforts to remedy these problems, the strainers were abandoned in
favor of a new design.
The reaction tanks were originally equipped with a corn-
minuter and an internal perforated plate to capture and grind
large chunks of solids that accumulated in the slurry circuit.
The equipment proved to be unnecessary because large chunks are
not normally generated within the module. The plate also tended
to be a convenient site for scale formation, which eventually
plugged it. These experiences, along with the decision to
design a new strainer, led to the removal of the comminuters and
the lower portion of the perforated plate.
Combustion Engineering installed new intank strainers, con-
sisting of large, semicircular perforated plates fitted around
the suction side of the spray pumps. Each strainer is equipped
with an oscillating and retractable wash lance for periodic
backwashing. Figure 2-17 shows a diagram of the redesigned
intank strainer and wash lance arrangement.
One strainer, constructed of carbon steel, was installed in
September 1976 and resulted in an immediate improvement. All the
2-55
-------
N)
i
m
er>
FLUE GAS
TO
SCRUBBER
REHEATER
OXIOIZER
FLUE GAS
TO STACK
RIVER
ID FAN
MIST
ELIMINATOR
OVERFLOW POTS
AND MARBLE BED
REACTION
TANK
TANK
THICKENER
UNDERFLOW
PUMP
SPRAY
PUMP
LIMESTONE —,
nrtNLur —
WATER
i
WEIR
1
THICKENER
— ^-
\
RE
GIF
tc
•
i
SLURP
TANK
-A
SLURRY
PUMP
ASH POND
RETURN PUMP
Figure 2-16. Simplified process diagram for Northern States Power Co.,
Sherburne 1 and 2 FGD system.
-------
REHEATER OUTLET It a
REHEATER INLET ttjOTT H
SPRAY PUMP
OUTLET GAS DUCT
OUTLET DAMPER
REHEATER
AND
WASH BLOWERS (4)
MIST ELIMINATOR
(2 STAGES)
MARBLE BED
ft
ft; POTS
UNDERBED .
SPRAY HEADERS (9)
NOZZLES
REACTION TANK
| INTANK
•STRAINER
GAS
INLET
OVERFLOW
(48)
0=D
INLET SOOT BLOWER
^SLURRY HEADER AND
NOZZLES (26)
ROD-DECK
VENTURI
HOLDING
TANK
TOP-ENTRY
AGITATOR
SIDE-ENTRY
AGITATORS (2)
Figure; 2rl7. Simplified diagram of a Sherburne scrubber module
2-57
-------
modules were equipped with the new design by March 1977. The
remedy has not, however, been a complete success. Nozzle plug-
ging persists, primarily because of scale formation that is
sheared off and carried to the nozzles from the strainer and
piping headers. Further, the tendency of the strainers to plug
activates the wash lance; and repeated, efficient backwashing has
caused strainer material failures (erosion of carbon steel plate)
and a dilution of slurry solids that adversely affects process
chemistry. The first problem has been corrected by using 316
stainless steel perforated plates. The second is being corrected
by devising a wash lance with a different supply pressure and
capacity.
Erosion of the rod-deck venturi housing and rods has been a
continual problem. The housing erosion has been traced to the
direct impingement of slurry from the spray nozzles. Sacrificial
wear plates of 316L stainless steel construction have been
fastened to the inside of the converging section of the venturi.
The rubber-coated, 316L stainless steel rods failed shortly after
installation. As a temporary solution to this problem, sacrifi-
cial wear plates of stainless steel angle iron were welded to the
bottom rod deck. The cause of the premature rod wear has been
linked to the large pressure drops across the rod-deck venturi
scrubbing stage. This stage was originally designed to operate
at a pressure drop of 2.1 kPa (8.5 in. H2O). To meet particulate
guarantees and lessen the opacity of stack emissions, the pres-
sure drop was increased to 3.3 kPa (13 in. H2O).
The Ni-Hard impellers and wear plate of the original slurry
spray pumps were designed by Worthington to supply slurry at 353
liters/s (5600 gal/min). They must be inspected and/or replaced
every 4000 to 6000 hours. On selected modules, Northern States
Power (NSP) is evaluating the use of 28 percent chrome-iron
internals and rubber-lined internals for the spray pumps. The
evaluation is still in progress; several modules on both systems
are being refitted with rubber-lined pumps having higher flow
rates.
2-58
-------
The slurry recirculation and effluent bleedoff piping is
lined with rubber (LaFavorite) up to the entry into the spray
headers (Hetron FRP). Failures have been numerous. The rubber
is sheared off in chunks by the abrasive slurry and carried to
the nozzles and spray headers. Such piping design errors as too
many reducers and sharp bends create part of the problem by
causing accelerated wear and failure at these points. The util-
ity first tried to solve the problem by removing the lining/ but
removal produced accelerated wear and erosion of the unprotected
carbon steel piping. The performance of fiber-reinforced plastic
(FRP), stainless steel, and rubber-lined spool pieces is currently
being evaluated.
Numerous failures have also occurred in the thickened slurry
underflow piping. The original fiberglass piping has been re-
placed with Ni-Hard piping. The mist eliminator wash lances,
which are retractable, and soot blower washers are also con-
structed of fiberglass; and NSP is evaluating a stainless steel
construc-tion for this application.
Material failures and modifications have affected the
arrangement and construction of the spray nozzles in the rod-deck
venturi. The nozzles were originally designed to extend into the
gas stream to ensure proper spray distribution and to minimize
the buildup of deposits on the rods. The nozzles themselves,
however, provided convenient sites for solids deposition and had
to be replaced with a new design that did not extend into the gas
stream; but this modification resulted in improper spray distri-
bution and subsequent buildup on the rods. New nozzles, designed
to cover the rod decks adequately and yet not extend into the gas
stream, have been installed and are functioning properly.
Nozzles constructed of Nordell (a proprietary plastic) and
carbon steel orifices were originally installed in the slurry re-
circulation lines of the modules. These components failed after
a short time and were replaced with ceramic nozzles and orifices.
Although initial perfprmance after this change was encouraging,
2-59
-------
the nozzles eventually failed because of erosion in the tangen-
tial cone forming chamber. Different nozzle types have been sub-
stituted for evaluation.
The main body of the marble bed scrubber module is lined
with a flaked fiberglass material (Ceilcote) that has been
troweled on the carbon steel shell from the slurry level to the
reheat section. Numerous failures of this coating have been
encountered in the modules of Sherburne 2. Pinholes have formed,
and chunks of coating have fallen off. Because these problems
have been encountered only in Sherburne 2, they are believed to
stem from improper application. The manufacturer has repaired
the liners at no cost to NSP.
The sidewalls above the slurry level in the venturi area
have been eroded by slurry cascading down from the rods and
impinging on the sloped surfaces. Failure of the flaked fiber-
glass in this area led to the addition of 316L stainless steel
wear plates.
Some minor material failures in the reheater were observed
during the early stages of operation. This was the result not of
external or internal corrosion, but of external erosion caused by
initial weld failures and compounded by washing after the initial
failure. A straightening bar that caused weld failures by adding
undue stress was removed. No failures have occurred since this
modification.
The slurry thickening and sludge disposal system has had
several problems since startup. During the winter of 1976-1977,
for example, the rake drive on Sherburne 1's thickener kicked out
on torque overload. The situation was not detected for a consid-
erable period, and the entire thickener filled with sludge.
Slurry from both units was then temporarily piped to one thick-
ener (Sherburne 2 was not in commercial service at the time),
which was able to accommodate the extra volume with a minimal
loss in efficiency.
Solids are thickened in a ratio of 2.5 to 1. The thickeners
were designed for an inlet of 10 percent solids, an outlet from
2-60
-------
25 to 30 percent solids, and an overflow with no solids. As the
equipment is now operated, part of the recycled water is used to
sluice down an ash hopper in the boiler economizers. Average
thickener inlet is less than 5 percent solids; thickener bottoms
are about 12 percent solids; and overflow is about 1 percent. No
flocculants are required.
Other problems have included plugging in the sludge pump
suction piping and valves, plugging in the dead leg pipes,
failures and freezing in the sludge transport lines because of
hydraulic and drainage problems, and electrical malfunctions of
valve operators. A complete design review was performed, and the
following modifications have been or are being made:
0 Modification and relocation of thickener sludge pump
discharge piping.
0 Installation of slower operators on selected valves in
sludge piping to reduce water hammer.
0 Installation of larger pumps and redesign of piping in
the scrubber pump house.
0 Leveling the sludge lines to the fly ash pond.
0 Installation of surge restraints (accumulators) on
sludge lines.
0 Connecting slurry inlets to thickeners, so that either
unit can discharge into either thickener.
Research development and future initiatives—Northern States
Power and Combustion Engineering are cooperating in a scrubber
improvement program that encompasses two primary objectives: (1)
optimization of the operation of the Sherburne 1 and 2 scrubber
systems and (2) development of the process design and operating
parameters for Sherburne 3 and 4. Specific goals included in the
first objective are:
0 Increasing particulate and sulfur dioxide removal
efficiencies.
0 Reducing manpower and operating costs. At present, 54
persons are employed to operate and maintain the
systems. The instrument technician crews (4) and
cleaning crews (21) have been unexpectedly large.
2-61
-------
0 Developing a more efficient and durable rod-deck ven-
turi scrubber section.
0 Improving control applications. The utility has al-
ready installed and activated a computer monitoring and
control system.
Several equipment modifications are being implemented as a
result of the goals. These modifications include:
0 Installation of a separate effluent bleedoff pump for
each module to replace the present procedure of bleed-
ing off the slurry spray discharge line.
0 Modification of the rod-deck venturi section with
different materials and design and with rods 11.4 cm
(4.5 in.) in outer diameter.
0 Continuation of the testing of different materials for
the rods, pumps, piping, and nozzles.
Module 101 of Sherburne 1 has been modified for the research
program to develop process design and operating parameters for
Sherburne 3 and 4. Specific equipment modifications include:
0 Removal of the marble bed and conversion to a spray
tower design.
0 Operating the rod-deck venturi scrubbing stage with a
pressure drop from 3.8 to 4.3 kPa (15 to 17 in. H2O).
0 Modification of the rod decks to use rods 18 cm (7 in.)
in outer diameter in a slightly different configura-
tion.
0 Incorporation of a bulk entrainment separator before
the mist eliminator to improve particulate removal. A
wash tray will also be added upstream of the mist
eliminator to test its effect on mist eliminator
cleanliness.
0 Testing the effect of the liquid-to-gas (L/G) ratio on
sulfur dioxide and particulate removal efficiencies.
The L/G ratio for Module 101 will be doubled by cou-
pling in the slurry spray pump and piping from another
module for a short time.
Operating history and performance—Sherburne 1 was first
fired in early 1976 and was placed in service on March 26, 1976.
2-62
-------
Commercial operation of the unit began on May 1, 1976. Sherburne
2 was first fired in late 1976 and was placed in service on
January 25, 1977. Commercial operation began on April 1, 1977.
The scrubbing systems were started and certified commercial
at the same time as their respective power generating units. For
Sherburne 1, system availability average 55 percent during the
first 3 months of operation after commercial startup. During the
remaining 4 months of 1976, however, availability rose to an
average of 94 percent. Thus average availability in 1976 was 90
percent. In 1977, the annual average availability for the
Sherburne 1 scrubber system was 93 percent. From January through
September 1978, the availability averaged 93 percent.
For Sherburne 2, scrubber system availability averaged 93
percent from April through December 1977. From January through
September 1978, the availability averaged 94 percent. The
slightly better availability of the scrubber system on Sherburne
2 since commercial startup may be attributed to the design and
operating experience gained by the utility and system supplier
with Sherburne 1. Figure 2-18 shows the average monthly FGD
system availability since January 1977 for Sherburne 1 and 2.
Tables 2-17, 2-18, 2-19, and 2-20 show the data from which Figure
2-18 was derived.
Another index by which to measure the performance of the
scrubber systems is the kilowatt-hour (kWh) output of the gener-
ating units. Because the systems cannot be bypassed, their
availability limits kWh output. The average capacity factor for
Sherburne 1 during its 29 months of commercial operation (May
1976 through September 1978) was 72.7 percent. The average
capacity factor for Sherburne 2 for its 18 months of commercial
operation (April 1977 through September 1978) was 78.8 percent.
Thus, the station capacity factor was 75.8 percent, which closely
corresponds to NSP's capacity estimate of 80 percent for the
period from 1976 to 1980. The monthly kWh outputs for both units
2-63
-------
to
I
CO
-
co
70
I I
• SHERBURNE 1
o SHERBURNE 2
I I I I
'jAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY JUN
1977 1978
Figure 2-18. Availability history of Sherburne 1 and 2, Northern States Power Company
-------
TABLE 2-17. SHERBURNE 1 PERFORMANCE SUMMARY: MAY 1976 TO SEPTEMBER 1978
Month/year
May 1976
Jun. 1976
Jul. 1976
Aug. 1976
Sep. 1976
Oct. 1976
Nov. 1976
Dec. 1976
Average
Jan. 1977
Feb. 1977
Mar. 1977
Apr. 1977
May 1977
Jun. 1977
Jul. 1977
Aug. 1977
Sep. 1977
Unit
output
MWh
318,050
372,450
269,700
421,110
349,470
385,610
470,820
480,920
383,220
386,060
395,210
510,650
460,100
195,790
127,290
476,490
357,380
416,430
Unit
capacity
factor
0.64
0.75
0.54
0.85
0.70
0.68
0.95
0.97
0.77
0.78
0.80
1.0
0.93
0.39
0.26
0.96
0.72
0.84
Boiler
operation,
h
657
688
512
705
566
606
720
722
647
607
609
743
718
312
248
736
640
686
Total FGD
system
availability,
%
86
84
84
94
95
93
93
95
90
90
91
95
95
92
92
97
95
95
Module operability, %
101
62
80
46
87
97
83
88
78
94
89
47
84
96
76
93
85
86
102
83
62
93
90
84
80
84
82
75
99
92
65
48
77
66
89
85
103
81
71
51
93
96
87
87
81
99
64
95
92
92
75
92
66
88
104
59
81
84
76
96
79
80
79
76
96
93
95
87
75
90
55
92
105
72
80
83
76
95
92
71
81
96
64
93
96
95
30
92
81
88
106
90
68
76
79
30
80
97
74
77
99
95
62
96
87
94
90
62
107
57
81
71
85
74
93
91
79
70
81
88
73
81
58
91
83
83
108
69
75
84
79
76
89
95
81
92
62
93
91
98
44
17
79
90
109
60
79
81
85
91
69
94
80
81
98
95
58
96
61
94
72
86
110
75
63
76
80
81
78
88
77
40
93
83
93
78
80
83
85
77
111
72
91
87
92
100
73
73
84
75
98
78
90
35
0
78
90
77
112
67
69
91
96
87
93
75
83
95
61
72
88
87
83
83
66
82
to
I
o\
Ul
-------
TABLE 2-17 (continued)
Month/year
Oct. 1977
Nov. 1977
Dec. 1977
Average
Jan. 1978
Feb. 1978
Mar. 1978
Apr. 1978
May 1978
Jun. 1978
Jul. 1978
Aug. 1978
Sep. 1978
Average
Unit
output,
MWh
360,850
412,360
325,050
368,640
374,290
366,200
423,220
464,520
380,010
414,670
394,510
416,930
185,740
380,010
Unit
capacity
factor
0.73
0.83
0.65
0.74
0.75
0.74
0.85
0.94
0.77
0.84
0.79
0.84
0.37
0.77
Boiler
operation,
h
609
705
557
598
648
636
676
713
635
717
694
742
357
646
Total
system
availability,
«
88
92
93
93
92
92
92
95
95
93
95
91
97
94
Module operability, %
101
70
77
97
83
83
0
71
92
61
50
82
64
89
66
102
81
91
90
80
63
93
83
87
86
84
76
65
62
78
103
95
42
95
83
88
92
64
87
85
85
71
73
77
80
104
97
86
94
86
73
89
89
44
86
85
74
63
77
76
105
63
79
71
79
84
74
90
81
89
62
75
65
58
75
106
35
82
92
88
84
85
83
85
64
78
52
80
82
77
107
87
87
26
76
92
89
62
91
62
55
75
81
68
75
108
99
91
67
77
64
88
89
86
83
8'
63
73
68
87
109
96
89
96
85
91
76
97
92
82
88
62
63
80
81
110
39
89
98
78
80
86
71
91
71
82
72
73
80
78
111
58
92
95
72
88
88
79
87
87
72
66
64
55
76
112
96
52
90
80
82
87
90
52
79
95
73
83
75
80
to
I
-------
TABLE 2-18. SHERBURNE 1 PERFORMANCE SUP1MARY FROM OCTOBER 1977 TO SEPTEMBER 1978
Month/year
Oct. * 1977
Nov. 1977
Dec. 1977
Jan. 1978
Feb. 1978
Mar. 1978
Apr. 1978
Nay 1978
Jun. 1978
Jul. 1978
Aug. 1978
Sep. 1978
Average
Unit
output,
MNh
360.850
412,360
325,050
374,290
366,200
423.220
464,520
380,010
414,670
394,510
416,930
185,740
341,310
Unit
capacity
factor
0.73
0.83
0.65
0.75
0.74
0.85
0.94
0.77
0.84
0.79
0.84
0.37
0.76
Boiler
operation,
h
609
705
557
648
636
676
713
635
717
694
742
357
641
Total
system
availability,
%
88
92
93
92
92
92
95
95
93
95
91
97
93
,
Module operability, %
101
70
77
97
83
0
71
92
61
50
82
64
89
70
102
81
91
90
63
93
83
87
86
84
76
65
62
80
103
95
42
95
88
92
64
87
85
85
71
73
77
80
104
97
86
94
73
89
89
44
86
85
74
63
77
80
105
63
79
71
84
74
90
81
89
62
-75
65
58
74
106
35
82
92
84
85
83
85
64
78
52
80
82
75
107
87
87
26
92
89
62
91
62
55
75
81
68
73
108
99
91
67
64
88
89
86
83
83
63
73
68
80
109
96
89
96
91
76
97
92
82
88
62
63
80
84
110
39
89
98
80
86
71
91
71
82
72
73
80
78
111
58
92
95
88
88
79
87
87
72
66
64
55
78
112
96
52
90
82
87
90
52
79
95
73
83
75
80
to
I
-------
TABLE 2-19. SHERBURNE 2 PERFORMANCE SUMMARY FROM APRIL 1977 TO SEPTEMBER 1978
Month/year
Apr. 1977
May 1977
Jun. 1977
Jul. 1977
Aug. 1977
Sep. 1977
Oct. 1977
Nov. 1977
Dec. 1977
Average
Jan. 1978
Feb. 1978
Nar. 1978
Apr. 1978
Hay 1978
Jun. 1978
Jul. 1978
Aug. 1978
Sep. 1978
Average
Unit
output.
MWh
442,170
421,020
423,680
482,890
366,500
423,280
449,140
43S.360
443,500
420,840
387,190
367,080
483,750
436,420
70,070
326,780
393,610
384,400
369,500
360,640
Unit
capacity
factor
0.89
0.85
0.85
0.77
0.74
0.85
0.90
0.88
0.89
0.85
0.78
0.74
0.97
0.88
0.14
0.66
0.79
0.77
0.80
0.73
Boiler
operation,
h
697
644
720
602
675
717
684
715
733
687
682
620
744
719
120
572
697
695
720
619
Total
system
availability,
t
92
91
96
97
93
94
95
91
93
94
92
92
97
92
91
95
95
93
96
94
Module operability, t
201
0
33
92
96
88
89
98
85
53
70
91
83
82
70
97
77
87
88
72
83
202
95
100
76
87
75
61
89
93
93
85
75
85
92
82
94
46
89
100
82
83
203
93
44
78
86
67
82
87
68
94
78
64
55
90
90
80
41
62
48
70
67
204
87
95
79
97
82
90
86
80
81
86
72
91
83
84
90
67
93
29
61
74
205
83
94
89
96
84
81
62
93
89
86
74
89
78
91
90
62
90
81
74
81
206
84
100
74
88
35
82
99
73
95
81
67
76
85
83
89
62
64
72
64
74
207
94
76
67
88
81
52
98
75
93
80
91
71
91
84
90
72
86
64
82
81
' 208
74
98
88
95
80
88
93
66
83
85
88
89
62
86
92
78
67
87
72
80
209
91
94
88
57
56
56
96
94
62
77
77
85
83
78
28
60
64
54
75
67
210
87
92
78
93
79
87
96
77
82
86
72
81
78
90
91
62
81
76
78
79
211
85
91
45
94
71
73
70
91
90
79
73
97
88
67
78
76
73
80
82
79
212
86
98
85
87
33
86
81
65
92
79
84
60
89
as
14
75
71
71
68
69
to
I
a\
oo
-------
TABLE 2-20. SHERBURNE 2 PERFORMANCE SUMMARY FROM OCTOBER 1977 TO SEPTEMBER 1978
Month/year
Oct. 1977
Nov. 1977
Dec. 1977
Jan. 1978
Feb. 1978
Mar. 1978
Apr. 1978
May 1978
Jun. 1978
Jul. 1978
Aug. 1978
Sep. 1978
Average
Unit
output i
Mtfh
449,140
435,360
443,500
387,190
367,080
483.750
436,420
70,070
326,780
393,610
384,400
396,500
381,150
Unit
capacity
factor
0.90
0.88
0.89
0.78
0.74
0.97
0.88
0.14
0.66
0.79
0.77
0.80
0.77
Boiler
operation,
h
684
715
733
682
620
744
719
120
572
697
695
720
642
Total
system
availability,
%
95
91
93
92
92
97
92
91
95
95
93
96
94
Module operabilit)
201
98
83
53
91
83
82
70
97
77
87
88
72
82
202
89
93
93
75
85
92
82
94
46
89
100
82
85
203
87
68
94
64
55
90
90
80
41
62
48
70
71
204
86
80
81
72
91
83
84
90
67
93
29
61
76
205
62
93
89
74
89
78
91
90
62
90
81
74
81
206
99
73
95
67
76
85
83
89
62
64
72
64
77
207
98
75
93
91
71
91
84
90
72
86
64
82
83
, »
208
93
66
83
88
89
62
86
92
78
67
87
72
80
209
96
94
62
77
85
83
78
28
60
64
54
75
71
210
96
77
82
72
81
78
90
91
62
81
76
78
80
211
70
91
90
73
97
88
67
78
76
73
80
82
80
212
81
65
92
84
60
89
85
14
75
71
71
68
71
to
I
o\
\0
-------
*
have tended to increase, and Sherburne 2's output has tended to
exceed that of Sherburne 1. These trends reflect the design and
operating experience that has been gained.
Several performance tests and continuous monitoring data
have demonstrated that removal efficiencies met the standard
emission regulations. Scrubber systems typically remove about 99
percent of the inlet particulate matter [6.9 g/m3 (3.0 gr/ft3)
inlet, 0.09 g/m3 (0.04 gr/ft3) outlet]1" and from 55 to 60 percent
of the inlet sulfur dioxide (700 ppm inlet, 300 ppm outlet). The
only compliance problems have been with opacity requirements.
Even when the systems have met or exceeded design particulate
control levels, opacity has typically ranged from 40 to 45 per-
cent, which is about twice the regulatory limit.
2.2.2 Domestic Limestone Scrubbing Units; Recently Operational
Systems
9—12
2.2.2.1 Arizona Public Service, Cholla 1 and 2
The Cholla Steam Electric Station of Arizona Public Service
is located in an arid desert region of Navajo County, Arizona,
near Joseph City- The area surrounding the station is relatively
flat and sparsely populated. There are no other major industries
in the area.
Cholla now operates only two of the five planned steam tur-
bine generating units. Both Cholla 1 and 2 have wet bottom,
pulverized coal-fired steam generators supplied by Combustion
Engineering. Cholla 1 is a 115-MW (net) unit and was put in
commercial service in May 1962. Cholla 2 is a 250-MW (net) unit
that began operation in June 1978, but as of October 1978 had not
yet been declared commercial.
Arizona Public Service is increasing the station's capacity
from the current 365 MW (net) to 1315 MW (net). Units 3 and 4,
now under construction, are scheduled for commercial startup in
June 1979 and June 1980 and are rated at 250 and 350 MW (net),
Sherburne 2's monthly output dropped off for the first three
quarters of 1978 because of outages and overhausl.
Dry basis.
2-70
-------
respectively. Unit 5, now in the planning stage, is scheduled
for commercial startup in June 1983. Like Cholla 1 and 2, Cholla
3, 4, and 5 will consist of pulverized-coal-fired units supplied
by Combustion Engineering.
The plant burns low-sulfur subbituminous coal shipped by
rail from the McKinley Mine near Gallup, New Mexico. A typical
analysis of this coal shows the heating value to be 24,190 kJ/kg
(10,400 Btu/lb), the sulfur content to be 0.5 percent, the chlor-
ide content to be 0.025 percent, the ash content to be 13.5 per-
cent, and the moisture content to be 15 percent.
Cholla 1 is equipped with mechanical collectors upstream
from the FGD system.
A Research Cottrell FGD system was retrofitted on Cholla 1
and began operation in October 1973.
Current emission regulations require that three of the other
four units at Cholla be equipped with FGD systems. Arizona
Public Service consequently awarded two separate contracts to
Research Cottrell for limestone slurry FGD systems on Units 2 and
4. Unit 3 will include only an ESP for the control of particu-
late emissions. The emission control strategy for Unit 5 has not
yet been determined.
Cholla 2 consists of four modules for the control of partic-
ulate and sulfur dioxide. The FGD system can handle 100 percent
of the flue gas and started operation simultaneously with the
boiler in June 1978. Cholla 3, an identical 250-MW (net) unit,
is scheduled to go on line in June 1980. Tables 2-21 and 2-22
summarize FGD system data for Cholla 1 and 2.
Pollution control—Cholla 1 is equipped with mechanical
collectors upstream from the FGD systems. The mechanical collec-
tors, Research Cottrell multicyclones, are designed to remove 80
percent of the inlet particulate matter. If the design efficiency
is achieved, loading at the outlet of the mechanical collectors
3 3
should be approximately 2.2 g/m (2.0 gr/ft ).
The FGD system on Cholla 1 consists of two modules, A and B.
Each module includes a 316L stainless steel flooded-disc venturi
2-71
-------
TABLE 2-21. FGD SYSTEMS DATA FOR CHOLLA 1,
JOSEPH CITY, ARIZONA
Unit rating (gross), MW
(net), MW
Fuel
Average fuel characteristics:
Heating value, kj/kg (Btu/lb)
Ash, %
Moisture, %
Sulfur, %
Chloride, %
FGD process
FGD system supplier
Application
Status
Startup date:
Initial
Commercial
Design removal efficiency:
Particulate, %
Sulfur dioxide,3 %
Actual removal efficiency:
Particulate, %
Sulfur dioxide,a%
Water loop
Total water makeup, j^
liter/s per net MVP*
(gal/min per net MW)
Sludge disposal
119
115
Subbituminous coal
23,609 (10,150)
13.5
15.0
0.5
0.025
Limestone slurry
Research Cottrell
Retrofit
October 1973
December 1973
99.7
58.5
99.7
50-60
Open
0.066
(1.04)
Unstabilized sludge disposed of in
an onsite unlined pond
a Sulfur dioxide removal efficiencies represent the overall efficiency.
The utility has no scrubber bypass.
b Net MW.
2-72
-------
STABLE 2-22. FGD SYSTEMS DATA FOR CHOLLA 2,
JOSEPH CITY, ARIZONA
Unit rating (gross), MW
(net), MW
Fuel
Average fuel (characteristics:
Heating value, kJ/kg (Btu/lb)
Ash, %
Moisture, %
Sulfur, %
Chloride, %
FGD process
FGD system supplier
Application
Status
Startup date:
Initial
Commercial
Design removal efficiency:
Particulate, %
Sulfur dioxide ,a%
Actual removal efficiency:
Particulate, %
Sulfur dioxide, %
Water loop
Total water makeup,
liter/s per net MW
(gal/min per net MW)
Sludge disposal
250
Subbituminous coal
23,609 (10,150)
13.5
15.0
0.5
0.025
Limestone slurry
Research Cottrell
New
June 1978
99.7
75.0
Open
0.066
(1.04)
Unstabilized sludge disposed of in
an onsite unlined oond
Sulfur dioxide' removal efficiencies represent the overall
efficiency. The utility has no scrubber bypass.
2-73
-------
scrubber, a cyclonic mist eliminator, an absorber tower constructed
of 316L stainless steel, and a final mist eliminator. The ab-
sorber on Module A includes packing for removal of the sulfur
dioxide with circulating limestone slurry. The absorber tower in
Module B is a hollow spray tower, and limestone slurry is not
circulated through it. Each module treats approximately one-half
of the total boiler flue gas. The entire FGD system is designed
to treat 227 m3/s (480,000 acfm) of flue gas at 136°C (276°F).
Actual boiler flue gas flow to both modules (at 115-MW generating
capacity) measures approximately 189 m /s (400,000 acfm). In
addition, bypass leakage around the FGD system amounts to 8 m /s
(17,000 acfm).
The sulfur dioxide absorber in Module A includes a fixed
plate and conical hat separator and packing. The fixed plate and
conical hat separator are constructed of 316L stainless steel.
The Munters packing, which consists of a fixed matrix of rigid
sheets of polypropylene, has a high specific surface area and low
pressure drop of only 0.12 kPa (0.5 in. H20). The superficial
gas velocity of the sulfur dioxide absorber is 2.1 m/s (6.9
ft/s), and the L/G ratio is 6.5 liters/m3 (48.9 gal/103 ft3).
The mist eliminators in each absorber tower are arranged
horizontally in two stages. The first stage is a chevron-type,
two-pass, polypropylene mist eliminator, located approximately
3.7 to 4.6 m (12 to 15 ft) above the absorber packing. The
design configuration of the second-stage, four-pass, polypropylene
mist eliminator differs only slightly from that of the first
stage. The distance between stages is approximately 1.2 m (4
ft). Vane spacing is 3.8 cm (1.5 in.) in the first stage and
18.1 cm (8.1 in.) in the second stage. On each tower, both
stages of the mist eliminator are washed on a timed cycle with
makeup water from plant wells. The quadrants of each mist elim-
inator stage are sprayed sequentially for 45 seconds every 30
minutes with makeup water at 520 kPa (60 psi). The flow rate of
the makeup water to the mist eliminator is approximately 15
liters/s (240 gal/min).
2-74
-------
She11-and-tube heat exchangers in each module raise the
temperature of the scrubbed gas from 50°C (122°F) to 72°C (162°F)
before it is discharged to the atmosphere. Each reheater con-
sists of two bundles of 316L stainless steel bare tubes with an
outside diameter of 2.5 cm (1.0 in.). The heating medium is high
pressure steam that is extracted from the boiler stream drum and
reduced in pressure from 13.2 MPa (1900 psig) to 1.8 MPa (250
psig). The reheater rating is approximately 84 GJ/h (8 million
Btu/h). Reheater steam power requirements are equivalent to
approximately 2 MW of electrical capacity. Six steam soot
blowers are operated for 5 minutes every 8 hours (once per shift)
to clean the tubes.
The reheated, scrubbed gases are discharged through carbon
steel ducts to the main stack. The ducts from each module enter
the stack at points directly opposite each other. The stack
shell is constructed of brick-lined concrete.
Choila 2 controls particulate and sulfur dioxide emissions
with a four-module limestone slurry FGD system supplied by
Research Cottrell. This new unit has a design particulate
removal efficiency of 99.7 percent and a design sulfur dioxide
removal efficiency of 75 percent. The four scrubbing trains are
arranged in parallel. Each train consists of a packed tower
absorber preceded by a flooded-disc scrubber that handles 25
percent of the flue gas (100 percent of the flue gas is scrubbed)
The scrubber trains have a 33 percent design capacity, so that
the unit is capable of operating at full load with only three
trains. At full load, each train can treat 185 m /s (392,000
acfm) of flue gas at 50°C (122°F). Cholla 2 includes an inline
steam coil reheat system that provides a temperature increase of
22°C (40°F). Spent scrubbing slurry from these open loop FGD
systems is disposed of in an onsite, unlined pond. The plant
site has no facilities for sludge storage or fixation. Because
of the area's light rainfall and the high evaporation rate,
wastewater discharge into receiving waters is not a problem. No
liquor is recirciilated back to the process.
2-75
-------
Cholla 1 uses about 5 MW from its gross generating capacity
of 126 MW to operate the emission control system. Another 6 MW
is required to operate auxiliary unit equipment. Energy require-
ments for Cholla 2 are not yet available.
FGD system—A simplified process flow diagram of the entire
FGD system is shown in Figure 2-19. A simplified process flow
diagram of Module A, which provides the primary sulfur dioxide
control, is shown in Figure 2-20.
Flue gas from the boiler induced-draft (ID) fans is pres-
surized by two booster fans to a static pressure of approximately
6.2 kPa (25 in. H2O) and flows downward through the throat of the
venturi-type, flooded-disc particulate scrubber. Limestone
slurry flows over the disc and is atomized as it is sheared by
the gas stream at the edge of the disc. Slurry is injected
tangentially through nozzles on the inside wall of the venturi
scrubber shell above the tapered throat. The orifice is formed
by the annular space between the circumference of the horizontal
disc and the wall of the tapered duct section in the throat. The
disc is adjusted within the tapered duct to increase or decrease
the area of the orifice. Thus, gas pressure and particulate
scrubbing efficiency can be controlled. The saturated, scrubbed
flue gas passes through a cyclonic mist eliminator, where solids
from collected fly ash, limestone slurry, and reaction products
are separated from the gas stream before it enters the absorber.
A diagram of a Cholla flooded-disc venturi and cyclonic mist
eliminator is provided in Figure 2-20.
Gas from the cyclonic mist eliminator enters the absorber
tower near the base. In Module A only, it contacts the limestone
slurry on the surface of the wetted film Munters packing, which
is 0.6 m (2 ft) thick.
The packed tower section is separated from the cyclonic mist
eliminator by a plate containing a conical hat. This arrangement
permits the flue gas to leave the cyclonic mist eliminator and
enter the packed spray section, but it prevents the spent lime-
2-76
-------
nm
-N
iouvn
-Nf
LOUVCt
STEAM
SMtll ADO TUBI
UOSTEt
FM
$0,A1SOMCI
•
^
i —
—
H U 1 Lftjvn
STEAH
>
4-
1 •*!
r r^ &
MS
to
I
»l
AISMUR
BIST
CLIHIMATM
NMWU.
HOSTEI (U1UOTIK
FAN
IIMESTCNE SLUMT MAKEUP
UATEI NMCEIIP
7
\
PAATtfULATE vAAAAjl
SCRU8BW
MBOUIE A 1
ASM.LADEM 1
HATER
1
,
sd
TA
<
i
2
OK
s
PHI
.$
STACK
' 1
ro
TAN
N
I
s
1C
s
\
^-
1
'
MIKE
1
--
S
ASH-
VHTI
TMU
PARTICIPATE
XMHtER 1
* L
NOOUU 1
.ADEN
•(
| | SUM
I"
J
ETA« |
10
ran
Figure 2-19. Simplified process flow diagram, Cholla 1 FGD system.
-------
INLET GAS
FROM MECHANICAL
COLLECTORS
BOOSTER
FAN
BYPASS
DAMPER
I
FLOODED DISC
SCRUBBER
to
I
~sl
oo
^^
H
i_T7i
\7
CYLCONIC
MIST
ELIMINATOR
SLUDGE
HOLDUP
TANKS
EXIT GAS
TO STACK
REHEATER
IMPINGEMENT
MIST ELIMINATORS
*+ MAKEUP WATER
(FROM WELL)
PACKING
CONICAL SLURRY
SEPARATOR
FDS
DISCHARGE
LIMESTONE
MAKEUP WATER
"(FROM y£LL)
1X1
PRE-EXISTING
ASH DISPOSAL
POND
FDS SLURRY TANK
TOWER TANK
Figure 2-20. Simplified process flow diagram of
Module A, Choi la 1 FGD system.
-------
stone slurry in the packed spray section from combining with the
spent scrubbing solution in the flooded-disc venturi (see Figure
2-19). Thus, fly ash cannot enter the absorber tower.
The scrubbed gas passes through a set of mist eliminators
(one set per absorber) and is reheated before it is discharged to
the atmosphere through the main stack. The mist eliminators are
of the slat (special baffle design) impingement type; they are
constructed of polypropylene and arranged horizontally (vertical
gas flow) in two stages. Reheat is provided by a set of steam-
heated, shell-and-tube heat exchangers (one set per module).
Each set of reheaters contains two bundles of tubes, which raise
the temperature of the saturated gas stream from 49°C (120°F) to
71°C (160°F) before it passes through a duct to the brick-lined,
concrete stack.
Limestone is added to Module A of the FGD system at approx-
imately 110 percent of the stoichiometric requirement for reac-
tion with the sulfur dioxide in the flue gas. Part of the
circulated liquor in the sulfur dioxide absorber is diverted to
the flooded-disc scrubber tank (common to both modules) to
maintain the pH between 4 and 5 in the particulate control system
(flooded-disc venturi). The liquid level in this tank is main-
tained by pumping the excess spent liquor to one of two surge
tanks (sludge holdup tanks) before it is discharged to a pre-
existing pond.
Dampers in the FGD system are arranged so that Module B
alone or Modules A and B together can be bypassed. Module A
alone can only be bypassed for short periods because limestone,
which enters the system in the Module A absorber tower tank, is
used to keep the operating pH of the entire particulate scrubber
system between 4 and 6. The inlets to both modules from the
booster fans are connected by a common suction header. Flue gas
flow control to both modules is maintained by balancing the
module fans through amperage control. Figure 2-19 presents a
diagram of the Cholla 1 FGD facility.
2-79
-------
Design parameters—The FGD system at Cholla is designed by
Research Cottrell to treat 227 m3/s (480,000 acfm) of flue gas at
136°C (276°F). Actual boiler flue gas flow to both modules (at
115-MW generating capacity) is approximately 189 m /s (400,000
acfm). In addition, bypass leakage around the FGD system amounts
to 8 m3/s (17,000 acfm).
The flooded-disc particulate scrubbers are constructed of
316L stainless steel and are 1.8 m (6 ft) in diameter and 13.7 m
(45 ft) high. The pressure drop on each is 2.5 kPa (15 in. H2O).
Each scrubber operates with a liquid recirculation rate of about
137 liters/s (2170 gal/min) at full load, which is equal to an
L/G ratio of 1.4 liters/m3 (10.1 gal/103 ft3) at 50°C (122°F).
Two-thirds of the scrubbing solution used in the flooded-disc
scrubbers is introduced through the hollow shaft of the flooded
disc; the remainder is sprayed through tangential nozzles on the
vessel wall.
The absorber towers are constructed of 316L stainless steel.
They are 6.7 m (22 ft) in diameter and 21.3 m (70 ft) high. The
sulfur dioxide absorber in Module A includes a fixed plate and a
conical hat separator and packing, which are also constructed of
316L stainless steel.
Limestone milling facilities—Most of the ground limestone
for the FGD system is supplied by the Superior Company in Phoenix,
Arizona. The grade supplied, known as "red wall" limestone, is
at least 75 percent minus 200-mesh. Chemical composition speci-
fications call for a minimum calcium oxide content of 52.5 percent,
a guaranteed minimum calcium carbonate content of 95 percent, and
maximum magnesium carbonate and silica contents of 0.5 and 1.0
percent, respectively.
The finely ground limestone is stored in an onsite silo from
which it is discharged at 9 kg/min (20 Ib/min) into a slurry
preparation tank at the base of the silo. The fresh limestone
slurry is introduced into the FGD system through the sulfur
dioxide absorber recirculation tank.
2-80
-------
Arizona Public Service is installing a limestone grinding
facility on the plant grounds. This facility will be able to
meet limestone requirements for Units 1 and 2. It will consist
of a ball mill capable of grinding the 0.6-cm (0.25-in.) lime-
stone rock delivered to the plant by rail so that 75 percent is
less than 200-mesh.
Modifications—Since startup, the FGD operating procedures
have been modified somewhat. The most important changes have
occurred in the process control area. The continuous pH sensors
were eliminated, and manual wet techniques were adopted once per
shift. The original density meters were replaced with nuclear
units. The utility has adopted the practice of water purging of
plugged sensing lines. The mist eliminator wash system, originally
designed to spray each quadrant with service water for 12 seconds
every 8 minutes, has been changed to spray each quadrant 45
seconds every 30 minutes.
The only major change in the scrubbing system design that
has been incorporated into the Cholla 2 scrubbing system is the
use of Inconel 625 in the reheater tubes; 316L stainless steel is
used in Cholla 1.
Problems and solutions—Startup and operation of the Cholla
1 FGD system have been accompanied by many problems. An analysis
of these problems reveals that most were related to process
design rather than process chemistry. The utility operators and
the FGD system designer have conceived and implemented solutions
to many of these problems. The major problems and solutions are
discussed in the following paragraphs.
Scale accumulated on top of and inside the cavity of the
shaft's stuffing box in the flooded-disc scrubber.. These scale
deposits were discovered early enough to prevent binding of the
shaft. Modifying the assembly of the stuffing box and rein-
stalling it in an inverted position (the cavity at the bottom so
it cannot accumulate solids) delayed binding. Eventually,
however, the shaft did freeze and had to be cleaned out. Other
2-81
-------
minor scale accumulations on top of the shaft dome and around the
tangential nozzles of the flooded-disc scrubber did not obstruct
the flow of limestone slurry or flue gas through the scrubber.
Dilute sulfurous acid condensate caused corrosion in the
expansion joints above the reheaters of both FGD modules and on
the top row of tubes near the tube sheet on Module B. To prevent
recurrence of this corrosion problem, the carbon steel ductwork
upstream of the reheaters in the Modules A and B was insulated
with a flaked glass liner (Ceilcote); and the Cor-Ten steel
expansion joints were replaced with rubber expansion joints. The
corroded tube bundle was replaced; and to prevent acid condensate
from reaching the new tubes, a trough was installed to divert any
condensate from the tube bundles. It is important to note that
corrosion of the reheater by sulfurous acid occurred only in
Module B (the module without packing), which has a low sulfur
dioxide removal efficiency. Presumably, the higher sulfur
dioxide removal efficiency of Module A (the packed tower) pre-
vents significant formation of sulfurous acid condensation.
Evidence of chloride attack was noted in the liquid-gas
centrifugal separator shell below the absorber. To remedy this
problem, Research Cottrell coated the interior of the vessel with
an epoxy material; it eroded in spots and had to be repaired.
The epoxy material also eroded and disbonded below the scrubber
disc. Acid-resistant brick was installed in this lower section
of the absorber and has held up for more than six months.
Evidence of additional chloride attack has been noted on
Module B reheater tubes, probably because of the chlorides that
are introduced in the well water used to prepare makeup slurry.
The spray distribution deflector above the flooded disc failed
because of stress-corrosion cracking. The deflector was rede-
signed by Research Cottrell, and the new design is holding up.
Extensive corrosion has recently occurred in the ductwork
leading from the Module B absorber tower exhaust elbow to the
reheat tube bundle. The utility has recoated the elbow several
2-82
-------
times with a Ceilcote liner. An application problem caused
repeated failure of the liner. This problem has still not been
fully resolved.
Harmonic vibrations occurred in the reheaters. The vibra-
tions were attributed to the vortex effect of an inadequate
transition of duct size from the absorber outlet to the reheater
shell. To remedy the situation, cross baffles were installed at
the reheater entrance. Vibrations also occurred in the Module B
booster fan because of uneven scale buildup on the fan blades
when the unit was idle. The blades were sandblasted, cleaned,
and rebalanced to eliminate these vibrations.
Sediment built up several times in dead spaces in pipelines
and valves of idle pumps, as well as in process lines. Buildup
occurred when slurry velocities in the pipe were low (during
periods of reduced operating rate). This problem was resolved by
redesigning some pipes to eliminate potential dead pockets. To
prevent valve freezing because of sediment buildup, some valves
were repositioned and flushout lines were installed.
Some pipe liners eroded (e.g., in the absorber tower pump
inlet piping). The erosion was sometimes caused by unsatisfac-
tory liner materials and sometimes by high flow velocities
through pipes and fittings. The rubber lining in some pipes
cracked, primarily because of defects in fabrication. Piping
modifications helped to reduce the erosion problem.
Burning a lower grade of coal (22 percent ash and 0.7 per-
cent sulfur) in the boiler has been accompanied by some plugging
in the mist eliminator and tower packing in Module A. Arizona
Public Service has not yet verified whether this plugging is
related to the lower grade of coal. If this buildup of material
continues, it appears that the life span of the packing and the
mist eliminator may be reduced as much as 50 percent.
The FGD system can accommodate the boiler down to a 50-MW
load without major problems occurring. Constant flow is main-
tained in the liquid circuit to prevent solids deposition in the
pipelines. A turndown below 50 MW requires that liquid flow be
2-83
-------
modulated and increases the probability of solids accumulating in
the pipelines because of the reduced liquid flow velocity.
A number of additional minor problems, have been encountered
and resolved by normal maintenance and engineering practices.
Among these problems are pump failures, expansion joint failures,
vessel lining failures (requiring recoating), malfunction of
solenoid valves in the mist eliminate*- wash system, reheater
steam leaks, gas damper adjustments, corrosion, erosion, scaling,
and liquid leaks in tanks, valves, and pipelines.
Since September 1977, no major design modifications have
been necessary at Cholla 1. From October 1977 through September
1978, only routine overhauls and general maintenance have occurred.
Scrubbing system problems have involved a recycle pump expansion
joint failure and leaks in the flooded-disc scrubber tank header
and venturi section. A-side and B-side reliabilities have
averaged 96 percent and 95 percent, respectively, since September
1977. Table 2-23 shows the performance history (reliability) of
Cholla 1 since January 1974 when the figures became available.
Figures 2-21, 2-22, and 2-23 show a graph of the module reliabil-
ities. Because of Cholla 2's recent startup, typical performance
information was not available.
Cholla 2 began initial operation in June 1978, and all four
scrubbing trains are in service. Cholla 2 has experienced
reasonant vibrations in the slurry piping since the unit became
operational. Research Cottrell has been injecting nitrogen into
the slurry lines to dampen the vibration, but no permanent solu-
tion to the problem has yet been found. In addition to the
vibration problem, the utility has observed peeling of the corro-
sion-resistant lining in the downcomer area of one absorber
module.
Removal efficiencies for Cholla 1 have been 99.7 percent for
particulates and from 50 to 60 percent for sulfur dioxide.
Removal efficiencies are not yet available for Cholla 2.
2-84
-------
TABLE 2-23,
CHOLLA 1 PERFORMANCE SUMMARY JANUARY 1974
TO SEPTEMBER 1978
Month/ year
January 1974
Feburary 1974
March 1974
April 1974
May .1974
June 1974
July 1974
August 1974
September 1974
October 1974
November 1974
December 1974
Average.
January 1975
February 1975
March 1975
April 1975
May 1975
June 1975
July 1975
August 1975
September 1975
October 1975
November 1975
December 1975
Average
Reliability, %
Module A
97
100
100
66
98
100
97
97
95
83
100
100
94
98
96
88
48
100
97
95
98
84
100
100
Module B
90
94
66
57
99
100
92
97
99
68
98
100
88
99
99
65
40
100
98
100
97
55
80
100
91 85
(continued)
2-85
-------
TABLE 2-23 (continued)
Month/year
January 1976
February 1976
March 1976
April 1976
May 1976
June 1976
July 1976
August 1976
September 1976
October 1976
November 1976
December 1976
Average
January 1977
February 1977
March 1977
April 1977
May 1977
June 1977
July 1977
August 1977
September 1977
October 1977
November 1977
December 1977
Average
Reliability, %
Module A
99
99
76
64
100
100
100
56
96
98
89
72
99
72
100
87
100
97
97
100
100
100
97
Module B
99
98
100
39
98
100
100
56
98
100
89
93
99
93
100
87
100
99
99
100
100
98
91
93 97
(continued)
2-86
-------
TABLE 2-23 (continued)
Month/ year
January 1978
February 1978
March 1978
April 1978
May 1978
June 1978
July 1978
August 1978
Average
Reliability, %
Module A
97
99
74
100
87
100
99
100
Module B
91
88
74
100
98
100
100
95
95 93
2-87
-------
NJ
I
CO
CO
MODULE A
MODULE B
I
I
I
I
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
1974 1975
Figure 2-21. Reliability history of Cholla 1, Arizona Public Service,
from January 1974 through December 1975.
-------
I
00
vo
100
90
80
70
i*j
. 60
>-
—>
3 50
40
30
20
LU
OL
10k
i i i i i i i i i i i i i i i i i i i i i i l
MODULE A
MODULE B
i i i i i i l i i i I I \ I l I I 1 1 1 1 1 1 L
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY JUN JUL AU6 SEP OCT NOV DEC
1976 1977
Figure 2-22. Reliability of Cholla 1, Arizona Public Service,
from January 1976 through December 1977.
-------
100
90
80
70
60
ri 50
CO
40
30
20
10
1 I I I
• MODULE A
• MODULE B
I I I i I I i i
JAN FEB MAR APR MAY JUN JUL AUG
1978
Figure 2-23. Reliability history of Cholla 1, Arizona
Public Service, from January through August 1978.
2-90
-------
2.2.2.2 Central Illinois Light, Duck Creek i13"16—
Background—The Duck Creek Plant is a new coal-fired power
generating station owned and operated by the Central Illinois
Light Company (CILCo). The plant is situated in an unreclaimed
strip mining area near Carrton, Illinois. The current capacity
of the plant is 416 MW (gross), which is provided by one coal-
fired power generating unit. The electric power generating
facilities at Duck Creek 1 consist of a Riley Stoker pulverized-
coal-fired steam generator and associated turbine generator.
Duck Creek 1 was placed in commercial service on June 1, 1976.
Duck Creek 1 is designed to fire high-sulfur, bituminous
Illinois coal having average sulfur and ash contents of 3.3 and
18.0 percent, respectively. During non-FGD operations, low-
sulfur Colorado coal is burned. Three more units of identical
capacity are planned for installation at the station. Duck Creek
2, 3, and 4 are scheduled for commercial operation in 1982, 1989,
and 1992, respectively, and will bring the total station capacity
to 2064 MW (gross).
Primary particulate control is provided by two parallel
ESP's designed by Pollution Control-Wather to have a removal
efficiency of 99.8 percent. Primary sulfur dioxide (SCO control
is provided by a Riley Stoker/Environeering limestone slurry FGD
system consisting of four parallel modules, each capable of
handling 25 percent of capacity. Overall (design) SG>2 removal
efficiency is 85.3 percent when burning 4 percent sulfur coal.
The modules are vertical, rod-deck, venturi scrubbers. Each
module can accommodate 167 m /s (353,900 acfm). All or part of
the flue gas can be bypassed around the scrubber modules during
outages or emergencies by manipulating bypass dampers and module
isolation dampers. The bottom ash, fly ash, and scrubbing wastes
2
are disposed of in an onsite 263,000 m (65-acre) sludge pond
lined with a natural impermeable material.
Spent scrubbing slurry is bled from the scrubber recircula-
tion lines as a 15 percent solids slurry containing reaction
products, unreacted limestone, and collected fly ash. The spent
2-91
-------
slurry is transferred to a waste collection tank, where it is
combined with liquid waste streams from plant sumps and dis-
charged to the pond. The waste solids present in the spent
scrubbing slurry settle in the pond, and the supernatant is
returned to the plant for reuse.
Duck Creek 1 uses about 10 MW from its gross generating
capacity of 416 MW to operate the FGD facility. Another 6 MW are
required to operate auxiliary unit equipment.
Duck Creek 1 data are summarized in Table 2-24.
FGD system—Primary sulfur dioxide control is provided by
four parallel, limestone slurry, rod-deck scrubber modules
supplied by Riley Stoker/Environeering and designed to remove 85
percent of the inlet sulfur dioxide. The modules are constructed
with carbon steel shells, Hastelloy-C internals, and 316L stain-
less steel rods (all internals are Hastelloy-C or 316L stainless
steel). The rod-deck scrubber is a proprietary, second-genera-
tion design scrubbing vessel marketed as a Venturi-Sorber Scrubber,
The vertical, multistage scrubber provides countercurrent contact
between gas and liquid and contains a series of rod decks arranged
vertically on staggered centers. The rods in each deck are 2.5
cm (1 in.) in diameter and spaced 2.5 cm (1 in.) apart. Each
module is designed to accommodate a flue gas volume of 167 m /s
(353,900 acfm).
Each scrubber module has a separate set of mist eliminators
arranged in a tilted vertical position in the horizontal dis-
charge ducts. The mist eliminators are equipped with a freshwater
wash system that consists of a common washdown tank with spray
pumps and piping for each mist eliminator. The wash system can
of deliver 55 liters/s (885 gal/min) of freshwater to each mist
eliminator. The water is sprayed in the second mist eliminator,
collected in the washdown tank, and reused in the first mist
eliminator.
Figure 2-24 presents a cutaway view of a module. It shows
the overall arrangement, as well as the internals.
2-92
-------
TABLE 2-24. FGD SYSTEMS DATA FOR DUCK CREEK 1,
CANTON, ILLINOIS
Unit rating (gross), MW
(net), MW
Fuel
Average fuel characteristics:a
Source
Heating value, kJ/kg (Btu/lb)
Ash, %
Moisture, %
Sulfur, %
Chloride, %
FGD process
FGD system supplier
Application
Status
Startup data:
Initial
Commercial
Design removal efficiency:
Particulate, %
Sulfur dioxide, %
Actual removal efficiency:
Particulate, %
Sulfur dioxide, %
Water loop
Total water makeup,
liter/s per net MW
(gal/min per net MW)
Sludge disposal
416
400
Bituminous coal
Illinios
24,523 (10,543)
9.12
18.0
3.30
0.03
Colorado
24,750 (10,640)
6.97
0.41
Limestone slurry
Riley Stoker/Engineering
New
July 1976
August 1978
99.8
85.3
Open
0.094
1.49
Unstabilized sludge disposed of
in an onsite lined pond
Design fuel specifications for high-sulfur Illinois coal.
Boiler and ESP commercial operation in June 1976. One FGD module
commenced operation in July 1976; full commercial operation with
all four FGD modules commenced in August 1978.
2-93
-------
MIST ELIMINATORS
SLURRY
SPRAY HEADS
GAS INLET
QUENCH
STATION
ROD DECKS (8)
ROD DECK (1)
SPENT
SLURRY
Figure 2-24. Cutaway view of a Duck Creek 1 FGD scrubber module,
2-94
-------
A reheat system is not included in the FGD system design.
The stack is protected from the acidic effects of the saturated
gas by an acid-resistant liner. The "wet stack" is 150 m (500
ft) tall and contains a Cor-Ten steel flue lined with flaked
glass. Four bottom hoppers are included in the stack for collec-
tion of moisture and slurry droplets that fall out of the flue
gas.
Limestone arriving at the plant is delivered to the dead
storage area, the live storage area, or the limestone grinder
building. The dead storage area holds a 90-day supply; and the
live storage area, a 3-day supply- Limestone delivered to the
grinder building is stored in a feed bin having a 24-hour capacity
The limestone delivered to the storage bin is 1.9 cm (3/4
in.) and must be ground so that 90 percent of it is less than 200
mesh. Grinding occurs at 10 kg/s (40 tons/h) in one of two (one
operational, one spare) wet ball mills to which the stone is
supplied by a weigh feeder. Fresh makeup water is fed to the
ball mill at 14 liters/s (220 gal/min) under maximum conditions
(100 percent boiler load, 4 percent sulfur coal). The milled
limestone is discharged to a slurry tank for collection and
pumped to a slurry storage tank through a classifier to insure
that the proper particle size has been achieved. The effluent
from the mill system is a 40 percent solids slurry; it is re-
tained in the slurry storage tank for 100 minutes before it is
added to the liquid scrubbing circuit through the scrubber
recirculation tanks.
Process description-Flue gas exits the boiler at 1,140 m /s
(2,415,000 acfm) and 446°C (835°F) and passes through half-size
air preheaters. The flue gas then enters two Pollution Control-
Walther ESP's connected in parallel; each ESP treats 50 percent
of the total gas flow. The ESP's are designed to remove 99.8
percent of inlet particulates when the inlet gas loading is 14.5
mg/m (6.34 gr/ft ). The gas is discharged from the ESP's at 717
m /s (1,520,000 acfm) and 135°C (275°F) and enters a manifold
2-95
-------
supplying four induced-draft fans. These fans, which overcome
draft loss in the boiler as well as the ESP's and FGD system, are
connected in parallel to a common duct that distributes the gas
to each scrubber module in the FGD system or to the bypass duct.
Flue gas enters the base of each scrubber module, where it
is quenched to adiabatic saturation conditions. The quenched gas
flows upward through nine successive stages of rod decks, where
contact is made with the scrubbing slurry in a countercurrent
fashion. The scrubbing slurry sprayed from the top of each
module flows downward through the rod decks. The rod decks
provide sites for intimate contact between gas and slurry. They
enhance mass transfer of the sulfur dioxide into the liquid phase
and thus promote sulfur dioxide removal. The design rate of gas
inlet to the scrubbing system is 668 m /s (1,415,600 acfm) at
135°C (275°F). The design rate of gas outlet from the system is
572 m3/s (1,211,000 acfm) at 53°C (127°F).
The cleaned, saturated gas stream in each module exits the
spray zone and turns 90 degrees to pass through a horizontal mist
elimination section. Entrained droplets of moisture and slurry
are removed from the gas stream by two stages of chevron-type,
slanted vertical mist eliminators located in the horizontal
discharge duct of each module. Scrubbed, saturated gas exits the
FGD system and enters the 150-m (500-ft) "wet stack" through the
breeching section.
Figure 2-25 shows the Duck Creek 1 FGD system scrubbing cir-
cuit for a given module. Figure 2-26 is a simplified process
flow diagram of the Duck Creek 1 power plant and emission control
system.
Performance—CILCo initially installed only one of the four
scrubber modules at Duck Creek 1 in order to evaluate its effec-
tiveness on high-sulfur coal before proceeding with the design,
installation, and operation of the remaining scrubber modules.
The first scrubber module (D-scrubber) was completed in June
1976 and placed in service on July 1, 1976. It operated inter-
2-96
-------
MIST
ELIMINATION
S02 ABSORPTION
ZONE
ROD-DECK
SCRUBBER
4-OPERATIONAL
GAS INLET-
LIMESTONE
SLURRY
AGITATOR
AGITATOR
STORAGE
TANK
SLURRY
TRANSFER
PUMP
1 - OPERATIONAL
3 • SPARE
MAKEUP
WATER
SPRAY
PUMPS
8-OPERATIONAL
4-SPARE
MIST ELIMINATOR
WASHDOWN TANK
4-OPERATIONAL
BLEED TO MASTE
COLLECTION TANK
SCRUBBER
RECIRCULATION
TANK
4-OPERATIONAL
RECIRCULATION
PUMPS
8-OPERATIONAL
4-SPARE
Figure 2-25 . Duck Creek 1 FGD system scrubbing circuit,
2-97
-------
TO ITMX
to
I
10
00
Figure 2-26. Simplified process flow diagram of Duck Creek 1
power plant and emission control system.
-------
mittently throughout the fall and winter and for approximately 1
month in the spring of 1977. The purpose of this operation was
to verify process chemistry and design. During this period of
service, several problems that made subsequent design modifica-
tions necessary became apparent. The modifications made included
revamnina the mist eliminator piping and wash system, revampinq
the slurry circulation system, replacing natural rubber lining
with neoprene, modifying values and flush/drain systems, and pro-
viding additional mixers for greater agitation. The remaining
modules were installed between April 1977 and July 1978. Durina
this time, the utility burned low-sulfur coal to meet the sulfur
dioxide emission standard of 516 ng/J (1.2 lb/10 Btu). Initial
startup of the entire PGD system commenced on July 23, 1978.
Because the Duck Creek 1 FGD system has only recently
attained commercial operating status, operating data are limited.
However, the data obtained during the D-scrubber test (removal
efficiencies, problems, solutions, and necessary desicm modifica-
tions) are discussed in the remainder of this section.
Duck Creek 1 commenced commercial operation on June 1, 1976.
D-scrubber was initially placed in the flue gas path on July 1,
1976. Limited operation during the balance of July and August
was caused primarily by such construction defficiencies as bad
welds, faulty pipe hangers, and scrubber slurry leaks. D-scrubber
was taken out of the gas.path to resolve these problems and
placed back on September 9. It operated approximately 360 hours
(noncontinuously) during the balance of the month and 385 hours
(noncontinuously) in October. During the autumn, several major
operating problems were encountered, especially sealing massive
mist eliminator scale, spray nozzle and pipe plugging.- and mate-
rials failure. The module remained out of service throughout
December 1976 and all of January and February 1977 because of a
scheduled 6-month boiler/turbine overhaul. During this outage, a
number of modifications were made to the scrubber to correct the
operating problems. The unit was placed back in service in mid-
march, and D-scrubber operated almost continuously for 350 hours
2-99
-------
during the balance of the month. Testing during April and May
1977 was to concentrate on operating the automatic control loops,
but was prematurely terminated because of installation difficul-
ties. D-scrubber was taken out of service, although Duck Creek 1
remained in service with the boiler firing low-sulfur Colorado
coal. The ESP's remained in service and removed particulates
from the flue gas with the aid of sulfur trioxide gas injection.
D-scrubber and the other scrubber modules were not olaced back in
the flue gas until July 23, 1978.
From July through September 1978, the system operated
intermittently. Modifications were made to the slurry transfer
tank after it was found to be underdesigned. The FGD system had
plugging problems caused by coal fines in the slurry (a common
unloading and transfer system is used for coal and limestone).
By the August-September 1978 period, the FGD system had achieved
an average availability of only 46 percent.
Problems and solutions—The interim testing of the D-scrubber
module revealed several chemical, mechanical, and desian problems
and prompted a number of modifications to the system. These
problems are discussed briefly in the paragraphs that follow.
Many of the chemical problems that beset the D-scrubber
module and ancillary equipment were caused or aggravated by an
incomplete instrumentation/control network. The sophisticated
automatic control system could not be put in service during the
early stage of operation.
Frequently, scaling and plugging occurred in the mist elim-
inators. Although these problems were attributed primarily to
the lack of automatic controls, the wash system was modified to
provide more efficient washing. Specifically, the polyvinyl
chloride materials used in the washwater piping that feeds water
from the washdown tank to the spray nozzles for each mist elim-
inator stage were-replaced with fiber-reinforced plastic. Also,
an additional spray header was added to the wash system to
provide more thorough rinsing.
2-100
-------
Another chemical problem was the widespread corrosion of the
Ceilcote 151 flaked glass liner that was sprayed on the Cor-Ten
steel stack flue to a thickness of 0.5 mm (20 mils). Inspection
of the liner after the D-scrubber test program revealed blister-
ing and acid corrosion, as well as subseouent widespread deteri-
oration of the flue. This problem seemed to stem largely from
the intermittent and partial scrubbing load which caused gas
conditions to vary widely through the stack. Premature failure
of the liner occurred because operating conditions exceeded the
design conditions specified for the materials. Two other factors
may have contributed to liner failure: the liner material itself
(it is no longer offered by the supplier for stack lining appli-
cations) and the absence of a reheat system.
The utility has repaired areas where cracked and peeled
liner exposed bare metal surface, but information on the success
of these repairs is not available. Other U.S. utility FGD sys-
tems with wet stacks have also experienced widespread corrosion
of liners, flues, and stacks. Such corrosion has required
extended outages for repairs and modifications.
Many mechanical problems involved premature pump lining
failures and damper leakage. Originally, all the slurry recircu-
lation and transfer pumps were lined with natural rubber; and
pump cavitation, which occurred frequently, caused the linings to
be stripped from the casings. To correct this problem, CILCo
replaced the natural rubber linings in the slurry recirculation
pumps with neoprene linings and the linings in the remaining
slurry pumps with reinforced natural rubber linings.
The utility also equipped all the slurry pumps with a flush-
out system. Because the circulating fluid is a slurry (with 15
percent solids in the recirculation and discharge lines and from
40 to 55 percent solids in the transfer lines), solids settle out
when flow is stopped. If they settle out in the pump, the pump
impeller and lining can be damaged on startup. Therefore, a
flush system was installed to purge the pump with freshwater
whenever the system is not in service.
2-101
-------
Several design deficiencies were observed to have caused or
aggravated the chemical and mechanical problems. These defi-
ciencies are summarized below:
0 The ceramic spray nozzles in the scrubber spray heads
were originally of the spinner vane type. Repeated
plugging of these nozzles prompted replacement with
orifices in the flow lines (open pipe arrangement) and
splash plates on the top rod deck.
0 Much of the mist eliminator fouling was attributed to
an unexpected high carryover of slurry solids in the
gas stream. These solids were deposited on the mist
eliminators, where they caused fouling, increased
pressure drops, and inefficient operation. Eventually
the scrubber module had to be shut down to clean the
mist eliminators. The problem was corrected by modify-
ing the rod decks. This modification increased the
pressure drop across the scrubber, reduced the entrain-
ment of slurry solids in the gas stream, and reduced
fouling in the mist eliminator.
0 Freshwater flush and drain systems were added to all
the slurry storage tanks, recirculation tanks, and pipe
lines to purge them of solids that settle out during
periods of inactivity.
0 Erosion of piping valves in the scrubber recirculation
lines was eliminated by moving the valves closer to the
recirculation tanks. Freshwater flush and drain sys-
tems have also helped to extend valve life.
0 An improper gas velocity profile in the scrubber con-
tributed to some of the problems. Riley Stoker/
Environeering is now attempting to determine the actual
profile and necessary corrective action.
0 More agitation was added to all the slurry tanks to
maintain solids suspension in the slurry circuit,
minimize solids settling, and promote reaction chem-
istry.
Removal efficiency—Because the FGD system has attained its
commercial operating status so recently, typical sulfur dioxide
removal 'efficiencies for full-scale operations are not available;
however, sulfur dioxide removal efficiency was measured on the D-
scrubber module during the interim test. The results (summarized
in Table 2-25) indicate that the removal efficiency was 91.6
2-102
-------
percent, which exceeds the design maximum guarantee value of 85.3
percent. This measurement was taken for sulfur dioxide inlet
concentrations of 3000 ppm.
TABLE 2-25. RESULTS OF THE D-SCRUBBER MODULE TEST
Gas capacity, m /s (acfm)
Sulfur dioxide inlet concentration, ppm
Pressure drop, kPa (in. H2O)
L/G ratio, liters/m3 (gas/103 ft3)
Sulfur dioxide outlet concentration, ppm
Sulfur dioxide removal efficiency, %
140 (300,000)
3000
2.2 (8.8)
2.2 (50)
252
91.6
Particulate removal efficiency measurements taken during the
D-scrubber test program also suggested that the scrubber was
removing as much as 70 percent of the inlet particulate matter
after it had passed through the upstream ESP's, even though the
scrubber is not designed to provide any particulate removal
capability beyond that of the emission control system.t The
utility and system supplier indicate that this phenemenon may be
attributed to the ionization or agglomeration of the small
is i fi
particles in the upstream ESP's. '
17-22
2.2.2.3 Kansas Power and Light, Lawrence 4 and 5
The Lawrence Energy Center is 625-MW (gross) power generat-
ing station owned and operated by the Kansas Power and Light
Company and located in Lawrence, Kansas. The station consists of
five units, the first of which was built in 1939. Lawrence 1 is
a 10-MW turbine powered by extraction steam from Lawrence 5.
Lawrence 2 and 3 are oil/gas-fired peaking units rated at 30 and
60 MW, respectively. Lawrence 2 was placed in service in 1950;
* The efficiency guarantee applied to 4 percent sulfur coal.
t The FGD system' is guaranteed not to add any particulate loading
to the discharge gas stream as measured at the outlet of the ESP's
2-103
-------
and Lawrence 3; in 1956. Lawrence 4 and 5 are multiple-fuel-
fired units that now exclusively fire coal and are rated at 125
and 400 MW, respectively. Lawrence 4 was first placed in service
in 1959; Lawrence 5, in 1971. The steam generators for Lawrence
4 and 5 are balanced draft, tangentially fired, multiple fuel
units supplied by Combustion Engineering. Lawrence 5 produces
1272 Mg (2,805,000 Ib) per hour of superheat steam at 540°C
(1005°F) and 18.1 MPa (2620 psi).
The coal burned at Lawrence is low-sulfur subbituminous
Wyoming coal with a heating value of 23,000 kJ/kg (10,000 Btu/lb).
It contains on average 0.5 percent sulfur, 10 percent ash, and 12
percent moisture. Lawrence 4 consumes approximately 45 Ma (50
tons) per hour of coal at full load; and Lawrence 5, approximately
145 Mg (150 tons) per hour.
To meet air emission regulations of the Department of
Health and Environment of the State of Kansas and the U.S. EPA,
each unit is equipped with a wet limestone scrubbing system
consisting of two parallel scrubber modules for the control of
particulates and sulfur dioxide. These Combustion Engineering
systems represent a second generation design replacement of the
limestone furnace injection systems originally installed on these
units in 1968 and 1971.
The maximum particulate emission allowed by the regulations
of the Department of Health and Environment of the State of
Kansas is 43 ng/J (0.1 lb/10 Btu) of heat input to the boiler.
As measured by the system supplier and utility during performance
tests Lawrence 4 emits only 34 ng/J (0.08 lb/10 Btu) of heat
input to the boiler.
The maximum sulfur dioxide emission allowed by the regula-
tions of the Department of Health and Environment of the State of
Kansas is 129 mg/J (0.3 lb/106 Btu) of heat input to the boiler.
As measured by the system supplier and the utility during per*
formance tests, Lawrence 4 emits from 6.5 to 13 mg/J (0.015 to
0.03 lb/106 Btu) of heat input to the boiler.
2-104
-------
Lawrence 4 and 5 share a common limestone storage and
preparation facility and a common waste disposal facility. Waste
solids in the Lawrence 4 slurry circuits are removed from the
scrubbing system by a liquid staging, forced oxidation, and
thickening. After treatment, the waste stream has a solids con-
tent of 35 percent and is conveyed to a network of three disposal
ponds. The supernatant from these ponds is returned to the
process and blended with thickener overflow in the recirculation
tank. Although Lawrence 4 and 5 share the same sludge disposal
ponds, Lawrence 5 is not equipped with a liquid staging and
thickening system. Spent slurry is forcibly oxidized by air
sparging and bled from the system by effluent bleed pumps, which
discharge underflow from the reaction tank directly to the ponds.
Supernatant is returned to the process and added directly to the
reaction tank.
Lawrence 4 uses about 10 MW of its gross generating capacity
of 125 MW to operate the emission control systems and auxiliary
station equipment; and Lawrence 5, approximately 20 MW of its
gross generating capacity of 420 MW.
FGD system data are summarized on Tables 2-26 and 2-27.
FGD system—Lawrence 4 and 5 are equipped with tail end wet
limestone scrubbing systems. Each operational scrubbing system
consists of two parallel two-stage scrubber 'modules to control
particulates and sulfur dioxide. Every module consists of a
rectangular, variable-throat, rod-deck venturi scrubber arranged
in series with slurry hold tanks, mist eliminators, inline re-
heaters, and isolation and bypass dampers. Ducts are provided so
that the modules can be bypassed when oil or natural gas is
burned in the boilers. Both systems share a common limestone
storage and preparation facility and a common waste disposal
facility.
The scrubbing systems were designed and supplied by Combustion
Engineering. They represent a second-generation design replace-
ment of the limestone furnace injection and tail end scrubbing
systems originally installed on these boiler in 1968 and 1971.
2-105
-------
TABLE 2-26. FGD SYSTEMS DATA FOR LAWRENCE 4,
LAWRENCE, KANSAS
Unit rating (gross), MW
(net), MW
Fuel
Average fuel characteristics:
Heating value, kj/kg (Btu/lb)
Ash, %
Moisture, %
Sulfur, %
Chloride, %
FGD process
FGD system supplier
Application
Status
Startup date:
Initial3
Commercial
Design removal efficiency:
Particulate, %
Sulfur dioxide, %
Actual removal efficiency:
Particulate, %
Sulfur dioxide, %
Water loop
Total water makeup,
liter/s per net MW
(gal/min per net MW)
Sludge disposal
125
115
Coal
23,000
9.8
11.8
0.55
(10,000)
Limestone slurry
Combustion Engineering
Retrofit
January 1976
98.9
73.0
99+
96-98
Closed
Unstabilized sludge disposed of in
an onsite unlined pond
Second generation FGD system.
2-106
-------
TABLE 2-27. FGD SYSTEMS DATA FOR LAWRENCE 5,
LAWRENCE, KANSAS
Unit rating (gross), MW
(net), MW
Fuel
Average fuel characteristics:
Heating value, kJ/kg (Btu/lb)
Ash, %
Moisture, %
Sulfur, %
Chloride, %
FGD process
FGD system supplier
Application
Status
Startup date:
Initiala
Commercial
Design removal efficiency:
Particulate, %
Sulfur dioxide, %
Actual removal efficiency:
Particulate, %
Sulfur dioxide, %
Water loop
Total water makeup,
liter/ per net MW
(gal/min per net MW)
Sludge disposal
420
400
Coal
23,000
9.8
11.8
0.55
(10,000)
Limestone slurry
Consulting Engineering
New
April 1978
98.9
52.0
Open
Unstabilized sludge disposed of
in an onsite unlined pond
Second generation FGD system.
2-107
-------
The original limestone furnace injection and tail end
scrubbing system was retrofitted on Lawrence 4. It operated from
November 1968 until mid-September 1976 and accumulated approx-
imately 27,000 hours of service. The unit was shut down at that
time to perform a scheduled turbine overhaul. During the over-
haul, construction and erection of the new scrubber modules were
completed. The new system went into service in early January
1977.
The original limestone furnace injection and tail end
scrubbing system was installed as new equipment on Lawrence 5 and
operated from November 1971 until March 20, 1978. Approximately
23,000 hours of service time was accumulated before the unit was
shut down to complete the tie-in of the new scrubbing system into
the flue gas path. The new scrubber modules were erected direct-
ly behind the existing system, which remained in service during
construction of the new system. Because the new system was
designed to use the existing reaction tank, spray pumps, induced-
draft fans, and stack, an outage of 6 weeks was required to
complete installation. The new system went into service on April
14, 1978.
The original scrubbing systems installed on Lawrence 4 and
5 were identical in basic design and operation. Each system
included facilities for pulverizing limestone and injecting it
into the boiler furnace chamber for calcination. The flue gas
transported the calcined linestone and fly ash to the scrubber
modules for particulate and sulfur dioxide scrubbing. The
cleaned gases then passed through mist eliminators, reheaters,
and induced-draft fans before being discharged through the stacks
to the atmosphere.
The Lawrence 4 scrubbing system consisted of two scrubber
modules. The Lawrence 5 scrubbing system was originally equipped
with six modules, and two more were added soon after startup.
All the modules were identical in size; each was designed to
handle approximately 70 m /s (150,000 scfm) of flue gas. Each
module had a single marble bed of Pyrex glass marbles that were
2-108
-------
1.9 cm (0.75 in.) in diameter. The beds were approximately 9 cm
(3.5 in.) thick and included overflow pots for drainage of spent
slurry into the recirculation tanks.
Each module was also equipped with mist eliminators and
reheaters. Two stages of horizontal, chevron-type mist elimin-
ators were situated approximately 1.5 m (4.5 ft) above the marble
bed. Four rows of carbon steel finned tube reheat bundles were
situated approximately 6.5 m (20 ft) above the second mist
eliminator stage. Under the mist eliminators were automatic
retractable wash lances that sprayed pond-return water at 1.1 MPa
(150 psig) for one hour each day. The reheaters were also
equipped with a self-cleaning system in which compressed air
between 0.65 and 0.80 MPa (80 and 100 psi) was blown from lances
for 3 minutes six times each day.
Each Lawrence 4 module was connected through an induced-
draft fan to a separate 36-m (120-ft) carbon steel stack. All
eight Lawrence 5 modules were connected through a common duct to
a 114-m (375-ft) stack. Originally, all the modules were equipped
with bypass ducts and hydraulic seal dampers. Extensive corro-
sion and plugging necessitated their removal from both modules of
Lawrence 4.
The limestone scrubbing systems now in service on Lawrence 4
and 5 are second-generation units that encompass the same general
equipment layout. This layout consists of a common limestone
storage and preparation facility, two rod-deck venturi scrubbers,
two spray tower absorber modules, and a common waste-disposal
facility.
Process—Flue gas from the boiler passes through the air
heater and is conveyed by new ductwork to two scrubber modules.
Each module can handle 50 percent of total capacity and consists
of a rectangular, variable-throat, rod-deck venturi scrubber
arranged in series with a spray tower absorber. Each module is
equipped with reaction tanks, mist eliminators, reheaters, a
bypass duct, and bypass and isolation dampers.
2-109
-------
Flue gas enters the scrubbers at 190 m /s (403,000 acfm) and
138°C (280°F). Each rod-deck venturi scrubber is comprised of a
converging gas section and rod section. The converging gas
section directs the flue gas downward to the rod-deck section,
which consists of two staggered levels of rubber-coated fiber-
glass rods. The rod section is 0.9m (3 ft) wide and 7m (23 ft)
long. The outer diameter of each rod is 16.8 cm (6.625 in.); the
rods are located on 33-cm (13-in.) centers. The vertical spacing
between the two rows of rods is regulated automatically according
to gas load to insure a constant pressure drop across the rod
section.
Limestone slurry is sprayed continuously into the rod-deck
scrubber by nonatomizing, fan-type, spray nozzles around the
perimeter of the throat area. The rod section provides a site
for intimate gas-slurry contact and thus facilitates particulate
and sulfur dioxide removal. Spent slurry from the rod section
falls into a collection tank located directly below the venturi
section. The collection tank has a liquid capacity of 190,000
liters (5,000 gal) and holds slurry approximately 14 minutes for
completion of chemical reactions. Slurry is recycled at 227
liters/s (3600 gal/min) from the collection tank to rod-deck
scrubber by two slurry recirculation pumps, the one operational,
the other spare.
After passing through the rod-deck venturi, the flue gas
makes a 90-degree turn and traverses the spray tower approach
where the gas makes another 90-degree turn. The saturated,
cooled gas enters the spray towers at 165 m /s (349,000 acfm) and
552°c (124°F). Additional sulfur dioxide scrubbing occurs as the
gas flows upward through the open towers, where it is contacted
by slurry through two levels of countercurrent sprays. Each
spray level is comprised of four internal spray headers, and each
spray header contains four spray nozzles. The first spray level
is situated 3 m (10 ft) above the inlet duct; and the second
spray level, 3 m (10 ft) above the first spray level.
2-110
-------
Spent slurry feeds by gravity into a reaction tank located
directly below each spray tower. The reaction tank has a capacity
of 262,000 liters (69,200 gal) and holds slurry approximately 10
min. for completion of chemical reactions and dissolution of
fresh limestone additive. One pump recycles slurry from the
reaction tank to the spray tower at 335 liters/s (5300 gal/min).
Entrained droplets of moisture and slurry picked up by the
flue gas stream are removed in a mist elimination section 3 m (10
ft) above the spray zone. Each mist eliminator has an A-frame
design and is comprised of a bulk entrainment separator followed
by two stages of chevron vanes. Each mist eliminator is equipped
with an intermittent, high-pressure wash system which sprays the
top of the bulk entrainment separator and bottom of the first
mist eliminator stage with blended makeup water.
After the mist eliminators, the saturated gas stream is
reheated by inline, hot water, carbon steel reheaters. One
reheater is provided for each spray tower, and each reheater
consists of four rows of circumferentially finned tubes arranged
in a staggered fashion. The heating medium is hot water from the
feed water deaerator of the boiler. Two half-track soot blowers,
located upstream of the reheaters, are used once every four hours
of service to clean the reheaters with compressed air. The
reheaters boost the temperature of the gas stream approximately
11°C (20°F). The reheated gas stream flows through the discharge
ducts leading to the induced-draft fans and stacks, which dis-
charge the gas to the atmosphere.
The Lawrence 5 scrubbing system closely resembles the
Lawrence 4 scrubbing system. Two scrubbing modules, each con-
sisting of a rod-deck scrubber in series with a spray tower
absorber, are provided to treat 100 percent of the flue gas from
the steam generator. In addition, the system shares the lime-
stone handling and preparation facilities and the sludge disposal
ponds used by Lawrence 4. Several major features, however, are
different:
2-111
-------
0 Gas enters the Lawrence 5 scrubbing system at 1858 m /s
(3,937,000 acftn) and 149°C (300°F) . Because the flue
gas flow and temperature are greater than those for the
Lawrence 4 system, the modules are significantly larger.
0 The flue gas is contacted by a single level of sprays
in the spray tower.
0 A single reaction tank receives the spent slurry from
both modules, as well as the fresh limestone slurry
introduced into the systems. (Slurry is recycled to
the rod-deck scrubber and spray towers of both modules
after a 10-minute retention period.)
0 Lawrence 5 is not equipped with a thickener or any
other dewatering device for separation of solids from
liquid prior to disposal.
Figures 2-27 and 2-28 are simplified process flow diagrams for
the second generation Lawrence 4 and 5 emission control systems.
Performance history—No lost boiler capacity has been
reported at Lawrence 4 since January 1977, when the limestone
slurry spray towers began operation. Although availability has
been 100 percent, Lawrence 4 experienced some freezing in January
1978 in the thickener underflow discharge piping and consequent
clarifier plugging. The freezeup problem continued through March
1978, but FGD operations continued because two fire hoses 7.6 m
(3 in.) in diameter were used to pump the underflow solids to the
disposal pond. The utility reported no unit outages through the
middle of September 1978. The unit was down the last part of
September 1978 for a scheduled turbine/boiler outage, and no FGD
operations were required. The particulate removal efficiency for
Lawrence 4 has exceeded 99 percent, and the sulfur dioxide
removal efficiency has reached from 96 to 98 percent. Performance
factors (availability, operability, etc.) for Lawrence 4 and 5
are not reported by the utility.
Construction continued into the first quarter of 1978 at
Lawrence 5 on the replacement of the original scrubbing system,
which continued to operate until March 20, 1978. It was then
taken off the line, so that the new system could be tied into the
2-112
-------
INLET-
FLUE GAS
TO OTHER
MODULE
BYPASS
STACK
+H- fj
I
MIST
ELIMINATOR
VANES
O
GAS
REHEATER
ROD SECTION
ROD
SCRUBBER
SPRAY
TOWER
ABSORBER
to
i
M
M
LJ
ADDITIVE
COLLECTION BLEED
iSTRAINER
TANK
SPRAY
PUMPS
EFFLUENTi
BLEED PUMP1
POND RETURN
WATER
RECIRCi
PUMPST
RECIRCULATION
TANK
,.
EFFLUENT BLEED
FROM OTHER
I MODULE
REACTION
TANK
STRAINERl
WEIR
OVERFLOW
UNDERFLOW
PUMPS
Figure 2-27. Simplified process flow diagram of one of the two
Lawrence 4 scrubbing modules.
-------
TO
STACK
ID FANS (2)
OUTLET
DAMPER
REHEATER
BYPASS
FLUE GAS FROM
AIR PREHEATER
REHEATER
BLOWER
MIST ELIMINATOR
BLOWER
STRAINER
WASHER (TYP)
STRAINER
WASH LINE (TYP)
ADDITIVE
(FROM MILL)
~1
L-0-
ADDITIVE
FEED PUMPS
(2)
n Iy!REACTION;
°~nlj TANK |
STRAINERS!
ABSORBER
SPRAY PUMP
ADDITIVE
STORAGE
TANK
EFFLUENT
BLEED
PUMP
COMMINUTOR
TO
POND
Figure 2-28. Simplified process flow diagram of one of the two
Lawrence 5 scrubbing modules.
2-114
-------
gas path. The new system began operation on April 14, 1978, and
has continued with 100 percent availability; no forced outages
have been reported through September 1978. Typical removal effi-
ciencies were not available for Lawrence 5.
2.2.2.4 Springfield City Utilities, Southwest I23"27—
The Southwest Power Station of Springfield City Utilities is
a one-unit 173-MW (net) station approximately 8 km (5 mi) south-
west of Springfield, Missouri. A second identical unit is
planned. Southwest 1 consists of a Riley Stoker pulverized-coal-
fired boiler with a 194-MW (gross) turbine generator. Operation
began in April 1977, and the unit was declared commercial in
September 1977.
The coal burned at this unit is bituminous high-sulfur
Kansas coal with a heating value of 29,100 kJ/kg (12,500 Btu/lb) .
It has a sulfur content of 3.5 percent and an ash content of 13
percent.
A four-field, cold-side ESP designed to be 99.7 percent
efficient provides primary particulate control for Southwest 1.
A limestone slurry FGD system provides sulfur dioxide control.
Both ESP and FGD systems are supplied by the Air Correction
Division of Universal Oil Products.
The FGD system consists of two parallel turbulent contact
absorber (TCA) modules that scrub 100 percent of the flue gas (50
*
percent each). Southwest 1 also has facilities for limestone
crushing and preparation and sludge handling. One or both FGD
modules can be bypassed during an emergency or malfunction by the
use of dampers. The sulfur dioxide removal efficiency of the FGD
facility is designed to be 80 percent. Cleaned, saturated flue
gas is discharged to the atmosphere through the main stack.
Flue gas cleaning wastes are discharged from the recircu-
lation tanks to a thickener, where solids are separated; the
clarifier liquor is recycled. Sludge from the thickener under-
flow is dewatered by a rotary drum vacuum filter; and the filter
2
cake is trucked to an onsite 40,000-m (10-acre) valley landfill.
2-115
-------
The Southwest Power Station uses about 9 MW of its gross
generating capacity of 194 MW to operate the emission control
system. Another 12 MW is required to operate auxiliary station
equipment.
FGD systems data are summarized in Table 2-28.
FGD systems—The FGD systems consists of two parallel TCA
modules with bypass ducting to allow one or both modules to be
bypassed when necessary. Each TCA module contains two stages of
packing that now consists of solid rubber balls 3.2 cm (1.25 in.)
in diameter. The design does not include a reheat system for the
cleaned saturated gas. The gas is ducted directly to a Ceilcote-
lined stack.
Process—Flue gas from the steam generator passes through
the ESP for particulate removal, and is boosted through the two
TCA modules by induced-draft fans. The TCA modules are designed
for gas at 314 m3/s (665,000 acfm) and 149°C (300°F). Flue gas
flows into the base of one of the TCA modules, passes through a
presaturator, and rises through two levels of packing consisting
of solid spheres 3.2 cm (1.25 in.) in diameter; the limestone
slurry is contacted in a countercurrent fashion. Each module is
equipped with a separate recirculation tank where reactions are
completed and fresh slurry is added. Flue gas cleaning wastes
are discharged from these tanks. The cleaned saturated flue gas
then passes mist eliminators before it is discharged to the
atmosphere through the 117-m (384-ft) main stack.
A simplified process flow diagram for Southwest 1 appears in
Figure 2-29.
Problems and solutions—Since this unit began operation in
April 1977 a number of modifications have taken place. The main
ones have been as follows:
0 During a shutdown from October 1977 through January
1978, the scrubbing system outlet damper was replaced
with 316L stainless steel. The hollow walled plastic
spheres having heavier walls. The lining in the
scrubber discharge ducts was replaced with a Ceilcote
2-116
-------
TABLE 2-28. FGD SYSTEMS DATA FOR SOUTHWEST 1,
SPRINGFIELD, MISSOURI
Unit rating (gross), MW
(net), MW
Fuel
Average fuel characteristics:
Heating value, kJ/kg (Btu/lb)
Ash, %
Moisture, %
Sulfur, %
Chloride, %
FGD process
FGD system supplier
Application
Status
Startup date:
Initial
Commercial
Design removal efficiency:
Particulate, %
Sulfur dioxide, %
Actual removal efficiency:
Particulate, %
Sulfur dioxide, %
Water loop
Total water makeup,
liter/s per net MW
(gal/min per net MW)
Sludge disposal
194
173
Bituminous coal
29,100 (12,500)
13.0
6.0-12.0
3.5
0.3
Limestone slurry
Air Correction Division, UOP
New
April 1977
September 1977
99.7
80.0
99.8
92.0
Closed
Unstabilized:.sludge dewatered and
disposed of in an onsite landfill
2-117
-------
SCRUBBER INLET
TEST PORT LOCATION
TOP OF STACK
ELEVATION 1645'
TEST PORT
ELEVATION 1516'
PREC1PITATOR
INLET TEST
PORT LOCATION
NJ
I
00
B" ABSORBER
MODULE
l_TOP OF OUTLET BREECHING
ELEVATION 1372'-2 3/8"
L-TOP OF BYPASS BREECHING
ELEVATION 1307'- B 1/2"
GRADE ELEVATION 1261'
"A" ABSORBER
NODULE
FIITU Oil
(NIUD HIT*
FITASN
UUVIU
DISMMll
SCRUBBER INLET
TEST PORT LOCATION
Figure 2-29. Simplified process flow diagram of
Southwest 1, Springfield City Utilities.
-------
lining. Insulation was installed around the seal air
fan between the inlet dampers, and the piresaturators
were relined with high-molybdenum steel.
0 Between February and March 1978, an expansion joint
failed between the induced-draft fan and the B-module
absorber. The absorber was bypassed during this period.
0 A fiber-reinforced plastic liner failed between April
and May 1978. A pump failure and a gas damper rupture
also occurred. Thus, only one of the two TCA modules
could be operated.
0 In June 1978, the mist eliminator wash system was
changed from a separate closed loop for each module to
a common system for both modules. During June, the FGD
facility was plagued with instrumentation problems; and
the B-module was inoperable because of the unrectified
problem with the expansion joint.
By September 1978, the unit had achieved an average total
FGD system availability of only 25 percent. The B-module expan-
sion joint had still not been corrected, and another ball failure
in the TCA packing necessitated a replacement of the plastic
spheres with solid rubber balls. From July through September
1978, the instrumentation problems were largely rectified,
although some slurry line plugging occurred.
A compliance test was conducted on September 14 and 15,
1977. Federal and state emission regulations require that
Southwest 1 limit particulate emissions to 43 ng/J (0.1 lb/10
Btu) of heat input, sulfur dioxide emissions to 516 ng/J (1.2
lb/10 Btu) of heat input, and nitrogen oxide emissions to 301
ng/J (0.7 lb/10 Btu) of heat input. The test showed an average
ESP inlet dust loading of 3143 ng/J (7.31 lb/106 Btu) and outlet
loading of 7.91 ng/J (0.0184 lb/10 Btu); thus the average ESP
efficiency was then 99.75 percent.
The average inlet sulfur dioxide concentration was 2726 ng/J
(6.34 lb/10 Btu); and the average outlet sulfur dioxide concen-
tration, 226 ng/J (0.556 lb/10 Btu). The average sulfur dioxide
efficiency was thus 91.7 percent. The average nitrogen oxides
emission was 229 ng"/J (0.762 1!
below the emission regulation.
emission was 229 ng"/J (0.762 lb/10 Btu), which was significantly
2-119
-------
2.2.2.5 Tennessee Valley Authority, Widows Creek 828~32—
Widows Creek Power Station of Tennessee Valley Authority
(TVA) is an eight-unit station located near Bridgeport, Alabama.
Units 1, 2, 4, 5, and 6 are rated at 140 MW (gross) each. Unit
3 is rated at 150 MW (gross). Units 7 and 8 are rated at 575 and
550 MW (gross), respectively. The total station generating
capacity is 1978 MW (gross).
The electric power generating facilities at Widows Creek 8
consist of a Combustion Engineering, tangentially fired balanced-
draft, coal-fired steam generator with associated turbine gener-
ator. Widows Creek 8 was first placed in service in 1963 and has
a net generating capacity of 516 MW.
The coal burned in this unit is bituminous, high-sulfur (3.7
percent) coal with a heating value of 26,200 kJ/kg (11,270
Btu/lb) and an ash content of 17 percent.
A 50 percent (actual) efficient ESP supplied by Koppers
provides primary particulate control. Secondary particulate
control and primary sulfur dioxide control are provided by a
limestone slurry scrubbing system designed and installed by TVA,
with the scrubber modules fabricated by Polycon. The unstab-
ilized waste material from the scrubbing operations is disposed
of in an onsite unlined pond.
Widows Creek 8 uses about 26 MW of its gross generating
capacity to operate the emission control system. Another 8 MW is
required to operate auxiliary station equipment.
The FGD systems data are summarized in Table 2-29.
FGD system—The scrubbing system consists of four parallel
trains. Each train includes a variable-throat venturi scrubber
followed by a multigrid tower absorber. The venturi sections are
constructed by 316L stainless steel; the multigrid towers, of
carbon steel (shells) with a rubber lining and 316L stainless
steel on the lower sloping (downcomer) sections, where the
slurry falling from the grids is funneled to the recycle system.
Inside each tower are five grids. The first, third, and fifth
2-120
-------
TABLE 2-29. FGD SYSTEMS DATA FOR WIDOWS CREEK 8,
BRIDGEPORT, ALABAMA
Unit rating (gross), MW
(net), MW
Fuel
Average fuel characteristics:
Heating value, kj/kg (Btu/lb)
Ash, %
Moisture, %
Sulfur, %
Chloride, %
FGD process
FGD system supplier
Application
Status
Startup date:
Initial
Commercial
Design removal efficiency:
Particulate, %
Sulfur dioxide, %
Actual removal efficiency:
Particulate, %
Sulfur dioxide, %
Water loop
Total water makeup
liter/s per net MW
(gal/min per net MW)
Sludge disposal
550
516
Coal
26,200
17.0
10.0
3.7
(11,270)
Limestone slurry
Tennessee Valley Authority
Retrofit
May 1977
January 1978
99.5
80.0
99.5+
85-94
Open
Unstabilized sludge disposed of
in an onsite unlined pond
2-121
-------
grids are made of 316L stainless steel, the second and fourth, of
fiber-reinforced plastic. Plans for one of the towers call for
the installation of packing that will be supported by the grids.
The scrubbing system includes a four-pass vertical chevron
mist eliminators constructed of 316L stainless steel. The mist
eliminator system is equipped with a continuous wash system.
Reheat of the wet gas is provided by an indirect hot-air
steam coil reheat system that raises the wet gas temperature by
28°C (50°F), to a point well above the dewpoint.
Limestone handling and milling equipment includes a shaker
for truck and rail unloading, a live storage silo, a wet ball
mill, and a cyclone classifier. The grinding system is designed
to grind 45 Mg, (50 tons) per hour of dry limestone, so that 90
percent of it is less than 200 mesh. The unloading and storage
facilities through the live storage silo are designed to accommo-
date a future limestone scrubbing facility on Widows Creek Unit
No. 7.
Figure 2-30 shows a schematic of the scrubbing facility at
Widows Creek 8; and Figure 2-31, a simplified process flow
diagram of this system.
Process—Flue gas entering the scrubbing system from the
ESP's passes through an induced-draft fan with a capacity of 190
o
m /s (400,000 acfm). The gas then enters one of the four variable-
throat Venturis. Each venturi is approximately 7m (23 ft) wide
and 8.5 m (28 ft) deep at the variable throat. Approximately 10
percent of the sulfur dioxide is removed in the Venturis. The
wet flue gas then enters a multigrid tower through which it rises
at 3.7 m/s (12 ft/s) and encounters the slurry in a counter-
current fashion. After turning through a four-pass vertical
chevron-type mist eliminator, the gas joins air drawn from the
powerhouse and pumped through steam coils that heat the air to
204°C (400°F). The stack gas temperature is 79.4°C (175°F).
2-122
-------
REHEATER
ENTRAPMENT SEPARATOR
to
U)
FLUE GAS DUCT
FROM POWERHOUSE
EXISITNG ESP'S
VENTURI CIRCULATION TANKS
ABSORBER CIRCULATION TANKS
ABSORBER CIRCULATION PUMPS
Figure 2-30. A schematic of TVA Widows Creek 8 FGD System.
-------
fo
I
M
to
, FROM
. \TRAINS
"-1 (B,C, AND D
-J
REHEATER
STEAM
ENTRAPMENT
SEPARATOR
SLURRY PUMP
SEAL WATER
HEADER
Q
FAN
FROM B. C. AND D TRAIN
VENTURI CIRCULATION TANK*! 1
TO SETTLING POND
VENTURI
CIRCULATION
TANK AND
PUMPS
ABSORBER
CIRCULATION
TANK AND
PUMPS
EFFLUENT SLURRY
SURGE TANK AND PUMPS
TO TRAINS
B.C. AND DJ >_
RIVER WATER
" PUMPS '
FROM POND WATER
RECYCLE PUMPS
FROM LIMESTONE SLURRY
TO TRAINS
B,C, AND D
LIMESTONE SLURRY
STORAGE TANK
LIMESTONE SLURRY
FEED PUMPS
Figure 2-31. A simplified process flow diagram of
TVA Widows Creek 8 FGD system.
-------
Performance—Since the scrubbing systems began operations in
May 1977, the utility has reported a number of operational
difficulties.
Because of poor ESP performance, the induced-draft fans must
handle a gas with a high grain load, from 11.4 to 13.7 g/m (5 to
6 gr/ft ). The high loading, effectively sandblasts the rotors
and the resulting erosion causes fan inefficiency and imbalance.
Excessive vibration necessitates removing the fan from service.
The fan housing also suffers severe erosion.
Each scrubber train has three bottom-entry guillotine
dampers, one each for the gas inlet, outlet, and bypass ducts.
Initial problems with the dampers resulted from corrosion of the
316 stainless steel seals. Because of the absence of housings
and bonnets on the dampers and because the FGD system is pres-
surized, the dampers have been leaking ash and flue gas exces-
sively. The drive mechanisms below the guillotine dampers have
collected ash in the threads, boots, and jack screw assemblies.
As of April 1978, the utility reported that the dampers were
inoperable except with extraordinary effort, because of electri-
cal component failure, deterioration of boots, and drive mech-
anism jamming.
Holes and tears have developed at numerous places in the
expansion joint fine-ply fabrics.
The most serious problem affecting the scrubber operations
have been the failure of rubber linings. In mid-June 1977, an
absorber inspection revealed that rubber lining on the downcomer
of Train C had come loose and that bare metal was exposed.
Patches of loose rubber and bare metal were found in other
trains, but only in the sloping sections of the absorber and
venturi hoppers. It was believed that the peeling and bubbling
resulted from the extremely cold weather during the curing period
in January 1976. Replacement material was obtained, and repair
work was carried out in early August 1977.
On September 26, 1977, an inspection of the trains during a
boiler outage revealed that rubber lining on Train C was again
2-125
-------
missing and that the slurry had worn a hole through the carbon
steel material of the venturi section hopper wall sloping towards
the downcomer. A temporary wear plate was installed to permit
operation until the scheduled boiler maintenance outage in
October. Other sections of loose rubber were found during this
same inspection. These defects were all corrected during the
October outage.
On December 20, 1977, an inspection revealed approximately
2 2
9.3 m (100 ft ) of bare metal showing in the four trains. As
before, the lining failures were confined to the sloping areas of
the absorber and venturi hoppers. Repairs began on December 27,
and Unit 8 came back on line in early January 1978.
On January 24, 1978, an inspection showed that large sections
of liner were missing and that holes were appearing in the carbon
steel shell. The utility decided to replace the rubber lining in
the sloping sections with 316L stainless steel plate welded to
the carbon steel shell.
Evidence of rubber liner failures has also been found in the
centrifugal pumps that send slurry from the ball mill product
tank to the classifiers. The liners in these pumps have required
high maintenance, possibly because of poor cyclone performance or
improper pump speeds.
The sumps in the ball mill house and near the spent slurry
surge tanks have been flooding. Limestone spills down into these
tanks, and limestone particles settle at the bottoms of the
sumps. Consequently, it has been difficult to empty waste from
these sumps to the pond.
The wet grind, rubber-lined ball mill, designed for a
capacity of 45 Mg/h (50 tons/h), was not functioning as intended.
During initial shakedown, rock and ball rejection occurred at
rates above 18 Mg/h (20 tons/h). Because coarse reject material
was falling into the mill slurry product tank, a chute was
installed to convey the material into a barrel and avoid contam-
ination of the finely ground product. Inspection of the product
2-126
-------
end of the ball mill revealed that the helix designed to retain
balls and rock in the grinding chamber was inadequate. A per-
forated plate was spot welded to the existing helix and extended
over the trommel screen; this plate has satisfactorily resolved
the problem.
Multigrids and absorber nozzles have presented problems.
The absorber section of each train has five trays of multigrids.
Originally, the bottom three of these were 316 stainless steel;
and the top two were fiber-reinforced plastic. Poor spray
distribution from nozzles eroded the plastic grids; and the
stainless steel grids were brought to the top, just below the
spray nozzles. Some adjustment in the nozzle cones improved the
spread of the spray, but a satisfactory flow pattern had not yet
been attained at the time of this report. Because of the nozzle
design and the slurry flow, the nozzle spray cones have fallen
off or have worn loose from part of their supports. The failure
seems less a result of erosion than the intense vibration that
wears away the cone supports.
Many instrument problems have resulted from poor initial
placement or inappropriate application. Particularly crucial are
the pH monitors, which are submerged flow-through meters. These
meters give readings on the absorber slurry flowing to the
nozzles. The transmitters associated with the pH meters vibrated
excessively when placed near the top of the absorber circulation
tanks. Relocation of the meters to ground level and replacement
of the reference electrode liquid seems to have improved reli-
ability and performance. There has also been poor control of the
feed rate of fresh limestone slurry because of a nonlinear
response from a linear control signal.
Pressure drop instruments for the gas are excessively
sensitive to ambient temperature changes and often go out of
calibration. Annubars for gas flow and outlet sulfur dioxide
monitors are located downstream of the mist eliminator section of
the scrubber and -have been unreliable as a result of moisture and
particulate plugging. Bracing and reverse air purging have
2-127
-------
improved instrument performance somewhat. Scaling and plugging
have not been a significant problem, because the pond-return
water has not been saturated with calcium sulfate or chloride.
There has been nozzle and header plugging in the venturi
section, partly because of excessive sulfates in the slurry.
These sulfates occur because of poor pH control, which results
from unreliable pH meters and poor control of fresh limestone
feed.
Heavy solid particles that solidify upon drying appear to be
settling in the venturi nozzle header. Until a steady state is
reached and a complete chemical analysis of the system can be
made, the cause of this settling cannot be identified.
Since May 1977, the average availability of the Widows Creek
8 FGD system has been 64 percent. Figure 2-32 shows the avail-
ability of the system from May 1977 through April 1978; and Table
2-30, the performance factors for the system during the same
period.
The particulate removal efficiency at Widows Creek 8 is 99.5
percent. In preliminary tests, the sulfur dioxide removal effi-
ciency has range from 85 to 94 percent. Table 2-31 presents the
results of these tests.
2.2.2.6 Texas Utilities, Martin Lake 1 and 233"42—
Martin Lake Steam Electric Station of Texas Utilities con-
sists of four 750-MW (gross) units located in Tatum, Texas.
Martin Lake 1 and 2 are operational units. Martin Lake 3 is
under construction, and Martin Lake 4 is in the planning stage.
The electric power generating facilities at Martin Lake 1
and 2 consist of Combustion Engineering lignite-fired steam
generators with 750-MW (gross) turbine generators. Martin Lake 1
began operation in August 1977, Martin Lake 2, in May 1978.
The coal burned at this station is low-sulfur (0.9 percent
sulfur) Texas lignite with an average heating value of 17,170
kj/kg (7,380 Btu/lb) and an ash content of 8.0 percent.
2-128
-------
100
90
80
70
' 60
d 50
CO
40
30
20
10
0
o
LU
O
CO
O
O
t/)
MAY JUN JUL AUG SEP OCT NOV DEC JAN FEB MAR APR
'1977 1978
Figure 2-32. Availability of TVA Widows Creek 8 FGD system,
2-129
-------
TABLE 2-30. FGD SYSTEM PERFORMANCE OF TVA WIDOWS CREEK 8
Month/year
May 1977
June 1977
July 1977
Aug. 1977
Sep. 1977
Oct. 1977
Nov. 1977
Dec. 1977
Jan. 1978
Feb. 1978
Mar. 1978
Apr. 1978
Availability,
%
2
11
25
26
75
86
56
88
55
60
69
Operability.-
%
3
18
28
30
71
88
97
96
54
66
83
Reliability,
%
30
71
88
59
96
61
59
67
Utilization,
%
2
11
25
26
60
50
56
79
47
58
62
2-130
-------
TABLE 2-31. PRELIMINARY SCRUBBER SULFUR DIOXIDE
REMOVAL TEST DATA FOR WIDOWS CREEK 838
Compliance Test Data
Test Number
SC>2 mass rate (average of
2 runs), kg/h (Ib/h)
Coal rate, Mg/h (tons/h)
Heating value (as fired) , kJ/kg
(Btu/lb)
Test emission (average of
2 runs), ng/j (lb/106 Btu)
1
1,859 (4,095)
181 (199)
26,180 (11,270)
396 (0.92)
2
899
184
25,902
189
(1,980)
(203)
(11,150)
(0.44)
3
1,501
167
24,903
361
(3,305)
(184)
(10,720)
(0.84)
FGD Efficiency and Viaual Opacities
Month
November
December
January
February
Coal sulfur
content , %
3.4
2.9
2.1
2.2
Visual opacity,
% Average
28
20
12
15a
SO,,
inlet
2398
2359
2138
2138
ppmv
Outlet
219
163
227
317
Efficiency SO2
Removal (Avg . ) , %
91
94
89
85
Only 1 day (February 3, 1978)
2-131
-------
Cold-side ESP's designed to be 99.4 percent provide primary
particulate control for Martin Lake 1 and 2. Sulfur dioxide is
controlled by limestone slurry FGD systems. The ESP's and FGD
systems on both units were supplied by Research Cottrell.
The PGD system on each unit consists of six parallel packed
spray tower absorbers that scrub a maximum of 75 percent of the
boiler flue gas with an efficiency of 95 percent (design) to
achieve an overall sulfur dioxide removal efficiency of 71
percent.
Spent scrubbing slurry (15 percent solids) is discharged
from each absorber tower into a common stream that flows into a
gravity thickener 43 m (140 ft) in diameter. The scrubbing
wastes are concentrated to a solids content of 35 percent and the
thickener underflow is fed to one of three centrifuges for
additional dewatering. Filter cake with a solids content between
68 and 70 percent is discharged from the centrifuges into a
Muller-type blender. The cake is combined with fly ash collected
in the ESP to form a material that can be conveyed by truck and
dumped. Railcars receive the blend for ultimate disposal in a
landfill.39'40'41
Martin Lake 1 and 2 each use about 10 MW of their gross
generating capacities for emissions control. Data about the FGD
systems at Martin Lake 1 and 2 are summarized in Tables 2-32 and
2-33. A simplified process flow diagram is shown in Figure 2-33.
FGD system—Each of the six parallel absorber towers includes
a three-stage sulfur dioxide removal configuration followed by a
two-stage mist elimination system. Reheat is provided by the hot
flue gas that goes through the ESP's, but bypasses the absorbers..
In a duct before the stack, this unscrubbed gas rejoins the
saturated flue gas from each absorber.
The system design includes a separate limestone milling and
handling facility for each unit. Each facility includes a lime-
stone storage bin, weigh feeder, ball mill, wet cyclone sepa-
rator, slurry prepration tank, and related slurry pumps and
piping.
2-132
-------
TABLE 2-32. FGD SYSTEMS DATA FOR MARTIN LAKE 1,
TATUM, TEXAS
Unit rating (gross), MW
(net), MW
Fuel
Average fuel characteristics:
Heating value, kj/kg (Btu/lb)
Ash, %
Moisture, %
Sulfur, %
Chloride, %
FGD process
FGD system supplier
Application
Status
Startup date:
Initial
Commercial
Design removal efficiency:
Particulate, %
Sulfur dioxide, %
Actual removal efficiency:
Particulate, %
Sulfur dioxide, %
Water1 loop
Total water makeup,
liter/'s per net MW
(gal/min per net MW)
Sludge disposal
750
Lignite
17f166 (7,380)
8.0
33.0
0.9
Limestone slurry
Research Cottrell
New
August 1977
99.4
Absorbers, 95.0; overall 70.5
Closed
0.05
(0.73)
Stabilized, dewatered sludge
disposed of in an onsite landfill
2-133
-------
TABLE 2-33. FGD SYSTEMS DATA FOR .MARTIN LAKE 2,
TATUM, TEXAS
Unit rating (gross), MW
(net), MW
Fuel
Average fuel characteristics:
Heating value, kj/kg (Btu/lb)
Ash, %
Moisture, %
Sulfur, %
Chloride, %
FGD process
FGD system supplier
Application
Status
Startup date:
Initial
Commercial
Design removal efficiency:
Particulate, %
Sulfur dioxide, %
Actual removal efficiency:
Particulate, %
Sulfur dioxide, %
Water1 loop
Total water makeup,
liter/s per net MW
(gal/min per net MW)
Sludge disposal
750
Lignite
17,166 (7,380)
8.0
33.0
0.9
Limestone slurry
Research Cottrell
New
May 1978
99.4
Absorbers, 95.0; overall 70.5
Closed
0.05
(0.73)
Stabilized, dewatered sludge dis-
posed of in an onsite landfill
2-134
-------
FAN OUTLET DAMPER
TOMER INLET DAKPER
TOMER OUTIH
I OAKPER
FROM SOILED
AIR PREHEATERS
"-
LOCATI
- nm
HERE/
a
t
w
ME
ITE
L
—
OF
R
t
»
4
FO
—
R FUTUt
TOUERS
U)
Ul
FLY ASH
MET CYCLONES
. PIMP SEAL HATER,
BALL MILL COOLING
AND HOSE STATIONS
MAKEUP HATER TO
LIMESTONE AREA
SERVICE MATER ASH POND
(LAKE WATER) RETURN MATER
(RECYCLE HATER)
Figure 2-33. Simplified flow diagram of Martin Lake 1 and 2 FGD systems.
-------
Process—The processes description provided applies to both
Martin Lake 1 and 2. The Martin Lake 1 and 2 flue gas cleaning
facilities are separate systems installed on separate units.
Flue gas exits the boiler, passes through the air preheater,
and enters the ESP's at 1495 m3/s (3,167,000 acfm) and 168°C
(335°F). From the ESP's the gas flows into a header for four
induced-draft fans that draw the gas into a second header feeding
the six absorbers and the bypass duct. The flowrate through each
absorber and through the bypass duct is controlled by balancing
dampers at every junction. Flue gas enters the base of each
absorber tower at 177.2 m3/s (375,350 acfm) and 168°C (335°F).
The gas rises through a cyclonic quenching chamber into the
absorber section of the tower at 142.9 m /s (302,870 acfm) and
57°C (135°F); it goes through a layer of sprays, packing, and
another layer of sprays. After passing a horizontal two-stage
mist eliminator system near the top of the absorber tower, the
saturated, cleaned gas exits the tower at 144.2 m /s (305,500
acfm) and 57°C (135°F) and mixes with hot, bypass gas at 541.2
m3/s (1,146,650 acfm) and 168°C (335°F) from the second header
duct. The combined gas stream enters the stack at 1264.2 m /s
(2,678,760 acfm) and 97°C (207°F).
Performance—Since Martin Lake 1 began operation in August
1977, a number of problems have occurred; and many modifications
have been made. Information is largely unavailable for Martin
Lake 2. It is assumed, however, that the Unit 1 modifications
have been taken into consideration for possible incorporation in
the Unit 2 design.
The utility has had trouble isolating individual absorbers.
Dampers provided for the Martin Lake FGD system include a single-
louver bypass damper, two consecutive louver dampers at each
tower inlet, and single-link-louver dampers at each tower outlet.
Even with the addition of seal air blowers between the two tower
inlet dampers, isolation of individual towers for personnel
access and maintenance has been extremely difficult. Areas of
malfunction have included bearings, seal strips, and linkages.
2-136
-------
All gas-side expansion joints have been replaced because the
original joints could not withstand exposure to slurry and
deteriorated. The replacement joints were performing adequately
as of October 1978.
Slurry line and spray nozzle plugging that required pipe
disassembly and mechanical cleaning was discovered during shut-
downs in early 1978. Such plugging was partly attributable to
accumulated construction debris, pieces of fallen wetted film
contactor (WFC) packing materials, and WFC support beams from
October and November 1977.
Sonic flow switches, installed to monitor flow deviations in
key process loops were found to be inadequate for an electric
generating station. They were eliminated, where possible, or
replaced with magnetic flowmeters.
Submerged pH meters in the quencher instrument wells exper-
ienced calibration drifts and electrical problems. After an
extensive test program, the meters were replaced by flow-through
devices with ultrasonic cleaners.
Sonic devices at liquid level in the quencher instrument
wells experienced problems related to electronics, poor calibra-
tion, slurry foaming, fouling of transducers, and excessive
variations in liquid level. These problems were addressed; and
the devices have functioned adequately since October 1977,
although modifications were still being made at the time of this
report because of transducer problems.
Poor gas distribution at the absorber tower inlets led to
the relocation of the gas flowmeters at the tower outlets. High
amounts moisture from the air and probe plugging were thought to
have contributed to the poor performance of the meters.
Electric on-off valve operators on reagent feed and density
control valves experienced numerous failures in early operation.
Some electric operators were replaced with pneumatic operators,
and others were modified to improve performance.
The original fiber-reinforced plastic packing support beams
were replaced with 316L stainless steel beams in January and
2-137
-------
February 1978. The beams had failed completely in one tower and
were on the verge of failure in several other towers when the WFC
packing was removed from all towers on November 7 and 8, 1977.
Failure of these beams resulted from inadequate structural
design, but was accelerated because solids accumulated in the
WFC. The accumulations of solids were caused by operating the
FGD system out of chemical balance. Such operation resulted from
malfunctioning pH devices and on-off valve operators, as well as
the difficulty in manual control that these malfunctions created.
Poor slurry distribution to the WFC may also have accelerated
this buildup. Later operation with 0.6 m (2 ft) of WFC packing
and with the new pneumatic valve operators produced insignificant
buildup in the WFC and mist eliminator packings.
Scale buildup at the cyclonic inlets of the absorber towers
was discovered in November 1977. Model testing showed that
modifications to the inlets would allow operation of the towers
for a full year without cleaning. These modifications have been
i
successful and will be made on all towers.
Martin Lake 1 and 2 underwent acceptance tests during
August 1978. As of October 1978, the results had not been
reported. Research Cottrell has, however, reported experimental
sulfur dioxide removal efficiencies. Absorber towers with 1.2 m
(4 ft) of WFC packing have achieved sulfur dioxide removal effi-
ciencies greater than 99 percent. Towers with no WFC packing
removed from 80 to 85 percent of all sulfur dioxide at peak tower
gas velocities. This testing, prompted the installation of 0.6 m
(2 ft) of WFC packing in each tower and the increase of maximum
tower gas throughput by 10 percent. Typical performance informa-
tion is unavailable for Martin Lake Station.
2-138
-------
REFERENCES FOR SECTION 2.2
1. PEDCo in-house files.
2. McDaniel, C.F. LyCygne Station Unit No. 1 Wet Scrubber
Operating Experience. Presented at Utility Scrubber Con-
ference, Denver, Colorado. March 29-30, 1978.
3. Laseke, B.A., Jr. EPA Utility FGD Survey: December 1977 -
January 1978. EPA-600/7-89-051a. PEDCo Environmental,
Inc., Cincinnati, Ohio. March 1978. pp. 55-59.
4. Melia, M., et al. EPA Utility FGD Survey: August-September
1978. Preliminary report. Prepared for the U.S. Environ-
mental Protection Agency under Contract No. 68-02-2603.
PEDCo Environmental, Inc., Ci ^innati, Ohio. November 1978.
p. 38.
5. Devitt, T., et al. Flue Gas Desulfuriz^iion System Capabil-
ities for Coal-fired Steam Generators, Volume II: Technical
Report. EPA-600/7-78-032b. PEDCo Environmental, Inc.,
Cincinnati, Ohio. March 1978. pp. 3-89 through 3-95.
6. PEDCo in-house files.
7. Laseke, B.A., Jr. Survey of Flue Gas Desulfurization
Systems: Sherburne County Generating Plant, Northern States
Power Company. Preliminary report. Prepared for the U.S.
Environmental Protection Agency under Contract No. 68-02-
2603. PEDCo Environmental, Inc., Cincinnati, Ohio. December
1978. pp. 2, 5, 7, 8, 13-15, 17-20, 22, 24, 26, 29, 33, 36,
38, 43-49, 52, 54, 56, 57, 58-60.
8. Op. cit. No. 4. pp. 59-62.
9. PEDCo in-house files.
10. Laseke, B.A., Jr. Survey of Flue Gas Desulfurization
System: Cholla Station, Arizona Public Service Co. EPA-
600/7-78-048a. PEDCo Environmental, Inc., Cincinnati, Ohio.
March 1978. pp. 2, 4, 6, 9, 11, 13, 22, 23, 26, 27, 29.
11. Op. cit. No. 4. pp. 4, 5, 24, 25.
2-139
-------
12. Laseke, B.A., Jr., et al. Utility FGD Costs: Reported and
Adjusted Costs for Operating FGD Systems. Preliminary
report. Prepared for the U.S. Environmental Protection
Agency under Contract No. 68-02-2603. PEDCo Environmental,
Inc., Cincinnati, Ohio. September 1978. Form 1.
13. Op. cit. No. 3. pp. 166-168.
14. Laseke, B.A., Jr. Survey of Flue Gas Desulfurization
Systems: Duck Creek Station, Central Illinois Light
Company. Preliminary report prepared for the U.S. Environ-
mental Protection Agency under Contract No. 68-02-2603.
PEDCo Environmental, Inc., Cincinnati, Ohio. December 1978.
pp. 2, 4, 5, 7, 10, 11-14, 16, 18, 21, 23, 28, 32, 35, 37,
44-46, 48-50.
15. Op. cit. No. 4. pp. 6, 26.
16. PEDCo in-house files.
17. Op cit. No. 12.
18. PEDCo in-house files.
19. Op. cit. No. 3. pp. 60-67.
20. Op. cit. No. 4. pp. 9, 10, 40, 41.
21. Power Engineering. New Generating Plants. Harrington,
Illinois. May 1978. p. 8.
22. Laseke, B.A., Jr. Survey of Flue Gas Desulfurization
System: Lawrence Energy Center, Kansas Power and Light
Company. Preliminary report. Prepared for the U.S. Environ-
mental Protection Agency under Contract No. 68-02-2603.
PEDCo Environmental, Inc., Cincinnati, Ohio. January 1979.
pp. vii-ix, 2, 4, 6, 7, 8, 14, 17, 19, 21-23, 25, 26, 28, 29,
31, 35, 42, 51, 52-54, 56, 57, 65-70.
23. PEDCo in-house files.
24. Op. cit. No. 3. pp. 130-132.
25. Op. cit. No. 4. pp. 19, 72, 73.
26. Op. cit. No. 12.
27. Universal Oil Products,.Air Correction Division. Stack
Emission Compliance Test Report for the City Utilities of
Springfiled, Missouri, Southwest Power Station. Des Plaines,
Illinois. September 1977.
28. PEDCo in-house files.
2-140
-------
29. Wells, W.L., W.B. Muirhead, and J.H. Buckner. TVA's Exper-
ience with Limestone Scrubbers at the 550-MW Widows Creek
Unit 8. Presented at American Power Conference, Chicago,
Illinois. April 24-26, 1978.
30. Op. cit. No. 3. pp. 141-143.
31. Op. cit. No. 4. pp. 19, 77, 78.
32. Op. cit. No. 12.
33. PEDCo in-house files.
34. Ballard, B., and M. Richman. FGD Systems Operation at
Martin Lake Steam Electric Station. Presented at the Joint
Power Generation Conference, Dallas, Texas. September 10-
13, 1978. pp. 1-17.
35. Op. cit. No. 3. pp. 144, 145;
36. Op. cit. No. 4. pp. 20, 21, 79.
37. Personal communications with D. Haverlah and R. Leard, Texas
Air Control Board.
2-141
-------
2.3 WELLMAN LORD FGD SYSTEMS
As of September 30, 1978, there were three operational
Wellman Lord systems treating flue gas from U.S. utility boilers.
These systems represent a total net generating capacity of ap-
proximately 735 MW, whereas the one previously reported utility
application of a Wellman Lord system represented a net generating
capacity of only 115 MW. A list of the domestic operational
Wellman Lord FGD installations is presented in Table 2-34. The
unit described in detail in the previous report, Dean H. Mitchell
11, is updated in Section 2.3.1.1; and the recently operational
San Juan systems are described in Section 2.3.2.1.
2.3.1 Domestic Wellman Lord Scrubbing Units; Previously
Identified Operational Systems
2.3.1.1 Northern Indiana Public Service Company, Dean H.
Mitchell 11—
No major system design modifications have taken place at the
Dean H. Mitchell Power Station since the previous report.
Performance—FGD operations resumed at Dean H. Mitchell on
June 13, 1977, after a 5-month outage caused by a boiler-related
mishap. Several hundred hours of total system operation at
partial and full loads were logged in June and July.
Performance tests by the EPA began on August 29, 1977. and
were completed on September 14, 1977- The test period included 12
days of operation at a flue gas equivalent of 92 MW and 3.5 days
of operation at a flue gas equivalent of 110 MW. During this
period, 91 percent of the sulfur dioxide was removed from the
flue gas. Coal containing 3 percent sulfur was burned, and and
207 Mg (204 long tons) of elemental sulfur were recovered. The
mean inlet sulfur dioxide concentration of 25 samples was 2709
ng/J (6.3 lb/106 Btu); and the mean outlet sulfur dioxide concen-
tration, 288 ng/J (.67 lb/106 Btu). All performance criteria
were met, including those for control of sulfur dioxide and
particulate emissions, consumption of raw materials and electric
2
power, and quality of recovered sulfur.
2-142
-------
TABLE 2-34. DOMESTIC OPERATIONAL WELLMAN LORD FGD INSTALLATIONS
Company
Northern Indiana
Public Service
Public Service of
New Mexico
Public Service of
of New Mexico
Station/unit
Dean H. Mitchell 11
San Juan 1
San Juan 2
Size*.
MM
115
314
306
Initial
startup
date
Nov/76
April/78
July/78
New or
retrofit
Retrofit
New
Retrofit
Coal
sulfur,
%
3.5
0.8
0.8
Design sulfur
dioxide resoval
efficiency. %
90.0
85.0
85.0
Actual sulfur
dioxide reaoval
efficiency, »
91.0
*Net with FGD
-------
After the performance tests, a demonstration test year was
begun. The purpose of the test year was to collect data in order
to characterize the operation of the plant boiler and scrubber
facilities. During October 1977, the FGD unit logged 132 hours
of operation with interruptions occurring to make repairs to the
sulfur dioxide reduction section, the evaporator, and the boiler.
In November 1977, the FGD system operated for 18 consecutive
days. Data indicated an average sulfur dioxide removal effi-
ciency of 90 percent, and 290 Mg (285 long tons) of sulfur
recovered. A boiler tube leak during November interrupted FGD
operations. Maintenance was performed in the evaporator toward
the end of the month.
An FGD system outage began in early December because of
abnormal boiler operating conditions related to high silica
levels in the feedwater. The outage lasted through the end of
February 1978; and maintenance was performed on the absorber
solution regeneration section, the evaporator circulating pump,
and the sulfur dioxide superheater piping.
The FGD system logged ten days of operation in March before
shutdown of the boiler for repair of the coal grinding mills and
precipitators. Poor quality coal was a major cause of the
problem with the grinding mills. The outage lasted through the
end of April, when a new supply of high-sulfur coal was obtained.
The FGD system operated intermittently through the end of
July because of various problems. In May, only 11 days of opera-
tion were logged because the flue gas isolation dampers failed.
Other problems included plugging of an entrainment separator in
the sulfur dioxide reduction unit and an imbalance of the booster
blower.
In August, the boiler operated continuously on high-sulfur
coal, and the FGD system logged 707 hours of operation. With the
exception of one 2-hour outage caused by a governor malfunction
on the sulfur dioxide compressor drive turbine, the FGD system
remained operational through September 12, 1978, when the unit
was shut down for an annual overhaul. Figure 2-34 illustrates
2-144
-------
NOV DEC JAN FEBMAR APR MAY JUN JUL AUG—SEP
Figure 2-34. Scrubber system availability , Northern Indiana
Public Service, Dean H. Mitchell II.
The availability index reflects the hours the FGD system is
available for operation (whether operated or not) divided by
the total hours in the period.
2-145
-------
system availability from November 1977 through September 1978.
Additional performance data is presented in Table 2-35.
Future initiatives—Of primary concern during the one year
test period at the Dean H. Mitchell Station is the collection and
evaluation of performance and economic data.
Northern Indiana Public Service Company (NIPSCO) will be
comparing the Wellman Lord/Allied Chemcial FGD system operating
and economic characteristics with sulfur dioxide emission control
alternatives. The other three generating units at Dean H.
Mitchell Station are presently burning low-sulfur western coal.
Thus, firing low-sulfur coal can be compared with using the
Wellman Lord/Allied Chemical FGD system. Other testing now under
consideration is discussed below.
In 1972 when the FGD plant was designed, there was little or
no concern that the process required the use of approximately 396
3 3
m (14,000 ft ) of natural gas per long ton of sulfur produced.
However, the cost and availability of natural gas have changed
drastically since the system was designed, and another reductant
is needed to alleviate this unanticipated disadvantage of the
system. The EPA and NIPSCO are considering joint investigation
of alternate gaseous reductants, including gases generated by the
gasification of coal. It is visualized that hydrogen and carbon
monoxide sources, such as low Btu gas, would be suitable for the
Dean H. Mitchell Station. If a suitable technology is identified
that shows economic and technical promise for demonstration with
the Wellman Lord/Allied Chemical FGD plant, the current test
period may be extended to include the demonstration of alternate
gaseous sources for the reduction of sulfur dioxide to sulfur in
the Allied sulfur dioxide reduction installation.
The ultimate reductant source is the abundant supply of
coal. Allied Chemical has developed technology for the direct
reduction of sulfur dioxide to sulfur utilizing a wide range of
steam coals. A long-range demonstration program to make this
technology commercially available is also under consideration for
the NIPSCO site.
2-146
-------
TABLE 2-35. PERFORMANCE DATA FOR DON H. MITCHELL II,
NORTHERN INDIANA PUBLIC SERVICE
1977 November
December
1978 January
February
March
April
May
June
July
August
September
Hours in
period
720
768
720
720
720
720
720
720
720
720
720
Hours
available
531
379
576
336
648
0
368
97
43
707
319
Hours
called
on to
operate
596
272
0
0
281
288
529
521
443
720
321
Hours
operated
428
0
0
0
215
0
263
3
17
707
319
Availabil ity,
%
74
49
80
47
90
0
51
13
6
98
44
Reliability,
%
72
0
77
0
50
1
4
98
99
Utilization,
%
90
0
0
0
30
0
37
0
2
98
44
-------
The EPA has indicated an interest in extending the demon-
stration period beyond one year to assess the effect of the
addition of a fourth Koch valve tray to the absorber. Prelim-
inary studies suggest that a fourth tray would permit sulfur
dioxide removal efficiencies in excess of 95 percent. Removal
efficiencies in this range could allow the utility to meet the
NSPS for sulfur dioxide of 516 ng/J (1.2 lb/106 Btu) and bypass
some untreated flue gas to reheat the treated flue gas. In
addition, the demonstration of high-efficiency absorption capa-
bility would be pertinent if the Federal NSPS for sulfur dioxide
were made more stringent.
A second area of interest in the study of high-efficiency
absorber operation is the evaluation of a novel tray design
developed by Merix Corporation. Such a novel tray may be in-
stalled as the fourth tray in the absorber to demonstrate in
field application that the tray can operate effectively and
efficiently without recirculating scrubbing solution. The
developer claims that the makeup sodium carbonate solution alone
will provide sufficient flow to support satisfactory operation of
the tray.
The first year of demonstration testing will determine the
ability of the FGD system to follow the typical boiler operating
conditions. The test program is important to demonstrate the
plant's versatility, reliability, and operability. Operating
costs will also be logged for comparison with the costs of other
sulfur dioxide control alternatives. An extension of the test
period would permit optimization of the energy and economic
demands of the plant.
The EPA Industrial Environmental Research Laboratory at
Research Triangle Park has initiated a program for the compre-
hensive environmental assessment of conventional combustion
processes. One aim of the program is to compare a well con-
trolled utility boiler firing coal with a well controlled utility
boiler firing oil. Dean H. Mitchell 11 is being considered for
investigation as the well controlled coal-fired boiler. At the
2-148
-------
sites chosen, the EPA will conduct comprehensive characteriza-
tions of all inlet and outlet streams and pollution control
equipment. The feedstream and emissions characterization data
will be used to assess the environmental, economic, social, and
energy impacts of the combustion system and associated pollution
control equipment.
2.3.2.1 Public Service Company of New Mexico, San Juan 1 and 2—
The San Juan Power Station is owned and operated by the
Public Service Company of New Mexico and is located in Waterflow,
New Mexico. The plant currently has two operational steam
electric generating units producing a total net generating
capacity of 620 MW.
San Juan 1 and 2 are base-loaded units with net generating
capacities of 314 and 306 MW, respectively. The gross rating of
both units is 350 MW. The coal burned at San Juan has an average
sulfur content of 0.8 and ash content of 20 percent; it is fired
at 181,440 kg/h (400,000 Ib/h). San Juan 1 developes a flue gas
flow of 416 m3/s (882,000 scfm) at 27°C (80°F); and San Juan 2,
3 67
a flow of 374 m /s (792,000 scfm) at the same temperature. '
8 9
Pollution control ' —The emission control systems at San
Juan 1 and 2 were supplied by Davy Powergas and utilize the
Wellman Lord process for removal of sulfur dioxide from the flue
gas. Each system consists of four scrubber modules, three of
which are required to handle full boiler load operations and one
of which is a spare. A hot-side electrostatic precipitator (ESP)
precedes each scrubber for primary particulate control and
provides a 99.8 percent removal efficiency. The FGD system inlet
particulate loading (dry) is 0.121 g/m (0.053 gr/ft ) at San
3 3
Juan 1 and 0.18 g/m (0.079 gr/ft ) at San Juan 2. The inlet
sulfur dioxide concentration for each averages 850 ppm, with a
range from 490 to 1200 ppm. The dry booster blower on each
delivers the flue gas through a variable-throat venturi to the
absorber at a design pressure drop of 7.7 kPa (31 in.H2O). The
flue gas is cooled and saturated in the venturi by water and
2-149
-------
slurry recirculated to the venturi sprays at 225 liters/a (3,570
gal/min). Fly ash captured by the scrubbing solution is purged
continuously from the system and buried in an 11.3-km2 (2800-
acre) mine disposal facility.
The absorption of sulfur dioxide from the flue gas takes
place in a five-stage absorber. Each stage consists of a valve
tray and a collector tray. A sodium sulfite solution reacts with
the sulfur dioxide to form sodium bisulfite. Each absorber was
designed for a flow of 215 m /s (455,000 acfm) at a pressure drop
of 4 kPa (17 in.H0O). The design liquid-to-gas ratio of each
3 33
absorber is 0.23 liter/m (1.7 gal/10 f t ) . A mist eliminator
removes entrained liquid droplets, and an indirect steam reheat
system provides a temperature boost of 28°C (50°F) to the exiting
flue gas.
The product solution collected on the bottom of the collec-
tor tray of each absorber overflows to the absorber surge tank.
From the tank, the solution is pumped through a filter to ensure
that no fly ash enters the evaporator system. A small sidestream
of the filtered solution is sent to the purge treatment area to
remove the sodium sulfate. The purge treatment equipment con-
sists of four chilled-wall crystallizers. A slurry of sodium
sulfate crystals forms in these crystallizers and is then removed
in a centrifuge. The bulk of the product solution is pumped to
the evaporator for regeneration of the sodium sulfite.
The evaporation system consists of a forced circulation
vacuum evaporator. The filtered solution is recirculated in the
evaporator, where steam at 345 kPa (50 psig) is used to evaporate
the water from the sodium bisulfite solution. When sufficient
water has been removed, sodium sulfite crystals form and precipi-
tate. Sulfur dioxide is removed with the overhead vapors. The
slurry formed by the sodium sulfite crystals is withdrawn contin-
uously to the dump/dissolving tank, where oondensate from the
evaporator is used to dissolve the crystals in the solution,
which is pumped back to the top stage of the absorber.
2-150
-------
Water vapor is removed from the sulfur dioxide in water-
cooled condensers. The sulfur dioxide is compressed by a liquid
ring compressor for introduction to the Allied Chemical facility
for reducing sulfur dioxide. The gas stream is about 85 percent
sulfur dioxide; the remainder, mostly water vapor.
Sodium lost as sulfate in the purge treatment system is re-
plenished by the addition of sodium carbonate to the absorber
solution. Soda ash is brought to the plant in trucks and trans-
ferred to the 101 Mg (100 ton) storage bin by a pneumatic con-
veying system. It is metered to the 462,000-liter (122,000-gal)
slurry tanks by a bin activator and belt feeder. The soda ash
slurry is pumped to the absorber feed tank by parallel centrifugal
pumps.
The small sidestream of filtered solution from the absorber
is pumped to the four chilled-wall crystallizers, where sodium
sulfate crystals form. The crystallized slurry is centrifuged to
extract the sodium sulfate crystals, and the clear solution is
returned to the evaporator feed system. The sodium sulfate
crystals are melted and fed to a stream-heated dryer, whose
discharge product is stored in a bin until loaded in trucks for
shipment. Any gases that evolve from the purge treatment are
chemically scrubbed and vented to the atmosphere.
The compressed sulfur dioxide is fed to the Allied Chemical
sulfur dioxide-reduction plant, where it is reacted with natural
gas. The resulting elemental sulfur is condensed and stored in
molten form for shipment. Waste gases are burned in an incin-
erator to convert any hydrogen sulfide to sulfur dioxide and
returned to the absorber inlet.
The sulfur regeneration rate of the San Juan unit is 0.96
Mg/h (0.94 tona/h). Table 2-36 provides additional system design
data.
Performance history —Initial PGD operation began at San
Juan 1 on April 8, 1978. After startup, the system was operated
with only two of the necessary three modules online. Thus, one-
third of the flue gaa was bypassed until the end of September,
2-151
-------
TABLE 2-36. FGD SYSTEMS DATA FOR SAN JUAN 1 AND 2
WATERFLOW, NEW MEXICO
Unit rating (gross), MW
(net) , MW
Fuel
Average fuel characteristics:
Heating value, kJ/kg (Btu/lb)
Ash, %
Moisture, %
Sulfur, %
Chloride, %
FGD process
FGD system supplier
Application
Status
Startup date:
Initial
Commercial
Design removal efficiency:
Particulate, %
Sulfur dioxide, %
Actual removal efficiency:
Particulate, %
Sulfur dioxide, %
Water loop
Total water makeup,
liters/s per net MW
(gal/min per net MW)
Sludge disposal
San Juan 1 314
San Juan 2 306
20
0.8
Wellman Lord
Davy Powergas
Unit 1 New
Unit 2 Retrofit
Unit 1 April 1978
Unit 2 Julv 1978
99.5
85.0
Closed
Elemental sulfur produced
2-152
-------
when modifications necessitated by an unusually large pressure
drop access the Venturis were performed. Adjustable plumb bobs
within the modules were moved further away from the Venturis to
improve the gas flow and reduce the pressure drop.
Initial operation at San Juan 2 began late in August 1978.
Early in September, the system was run in an integrated mode with
three modules. Initial problems were encountered with the
booster boiler control damper. Because of the recent startup of
the San Juan units, typical operating hours and reliability
parameters are not yet available.
2-153
-------
REFERENCES FOR SECTION 2.3
1. Melia, M., et al. EPA Utility FGD Survey August-September
1978. Preliminary Report. Prepared for the U.S. Environ-
mental Protection Agency under Contract 68-02-2603. PEDCo
Environmental, Cincinnati, Ohio. November 1978. p. 92.
2. Kelly, W.E., et al. Air Pollution Emission Test: Volume I:
First Interim Report, Continuous Sulfur Dioxide Monitoring
at Steam Generaters. Emission Measurement Branch Report No.
77 SPP 23A. The U.S. Environmental Protection Agency,
Research Triangle Park, North Carolina.
3. Op. cit. No. 1. pp. 57-58.
4. Laseke, B.A., Jr. EPA Utility FGD Survey: December 1977-
January 1978. EPA-600/7-78-051a. PEDCo Environmental,
Cincinnati, Ohio. March 1978. p. 106.
5. Link, W.F., and W.H. Ponder. Status Report on the Wellman-
Lord/Allied Chemical Flue Gas Desulfurization Plant at
Northern Indiana Public Service Company's Dean H. Mitchell
Station. In: Proceedings of the Symposium on Flue Gas
Desulfurization-Hollywood, FL, November 1977 (Volume).
EPA-600/7-78-058a. March 1978.
6. PEDCo in-house files.
7. Op. cit. No. 1. pp. 16, 18.
8. Devitt, T., et al. Flue Gas Desulfurization System Capa-
bilities for Coal-Fired Steam Generators. EPA-600/7-78-032b.
PEDCo Environmental, Cincinnati, Ohio. March 1978. pp.
3-276 through 3-282.
9. Personal communication between R. Mullens, Public Service
Company of New Mexico, and H.J. Young, Edison Electric
Institute. March 21, 1975.
10. Op. cit. No. 1. pp. 68-69.
2-154-
-------
2.4 OTHER FGD SYSTEMS
2.4.1 Domestic Alkaline Fly Ash/Lime Scrubbing Units
At the time of the previous report publication, only one
utility, Montana Power, was using alkaline fly ash as the primary
absorbent for sulfur dioxide in flue gas from a full-scale
commercially operating utility boiler. The alkaline fly ash/lime
scrubbing system at Montana Power's Colstrip station was dis-
cussed in Appendix A of the earlier report; that discussion has
been updated in Appendix A of this report. The text that follows
describes the background and performance of the Colstrip facility
and of the new Milton R. Young facility, which uses alkaline fly
ash/lime scrubbing.
2.4.1.1 Montana Power, Colstrip 1 and 2 —
Colstrip Power Station of Montana Power Company is currently
a two-unit station (with plans for two additional units) located
near Colstrip, Montana, approximately 160 km (100 miles) east of
Billings, Montana. Colstrip 1 and 2 are identical new 358-MW
(gross) units. The planned units, Colstrip 3 and 4, are each
rated at 700 MW (gross). Colstrip 1 began operation in September
1975 and was declared commercial in November 1975. Colstrip 2
began operation in May 1976 and was declared commercial in August
1976.
The electric power generating facilities at Colstrip 1 and
2 consist of pulverized-coal-fired steam generators with turbine
generators.
The coal burned at these units is subbituminous low-sulfur
coal with a heating value of 20,570 kJ/kg (8,843 Btu/lb). The
sulfur content ranges from 0.4 to 1 percent and averages 0.77
percent. The ash content varies from 6.1 to 12.6 percent.
Primary particulate and sulfur dioxide control is provided
by an alkaline fly ash/lime scrubbing system supplied by A. D.
Little and Combustion Equipment Associates. This system consists
of three scrubber modules, each capable of handling 40 percent of
the boiler flue gas capacity. The scrubbing system is designed
2-155
-------
to remove 99.5 percent of the particulate matter and 60.0 percent
of the sulfur dioxide from 100 percent of the boiler flue gas.
The scrubbing wastes, consisting of unreacted fly ash, calcium
sulfate, and calcium sulfite, are discharged to an initial hold-
ing basin adjacent to the scrubber plant and then pumped 4.8 km
(3 miles) to a permanent sludge disposal pond.
The Colstrip station uses about 24 MW of its current gross
generating capacity of 716.8 MW to operate the emission control
system. Another 32.8 MW are required to operate auxiliary
station equipment.
FGD system data are summarized in Tables 2-37 and 2-38.
A simplified process flow diagram is shown in Figure 2-35.
FGD system—Each scrubbing system at Colstrip consists of
three parallel scrubber modules. Each scrubber includes a
variable-throat venturi for particulate and sulfur dioxide
removal and a spray absorption section for additional sulfur
dioxide removal. The venturi and spray column operate with a
recirculating slurry containing 12 percent solids. The venturi
operates with a pressure drop of 4.2 kPa (17 in.tUO). The
venturi section is designed for a liquid-to-gas ratio (L/G) of 2
3 33
liters/m (15 gal/10 ft ); and the spray absorbtion section, for
an L/G ratio of 2.4 liters/m (18 gal/10 ft ). A wash tray
system has been provided below the chevron mist eliminators to
dilute the entrainment and prevent plugging of the mist elimin-
ators. Fans are located downstream of the scrubber and operate
on dry reheated gas. Colstrip 1 was the first commercial system
to utilize the alkalinity of the fly ash for a major portion of
the sulfur dioxide removal. Supplemental lime, which is added as
required for pH control, is pumped from the slaker to the re-
action tank in the base of the scrubber modules. Flue gas
reheat is provided by an inline steam coil system that increases
the temperature of the saturated gas above the dew point.
Process flow—Boiler flue gas exits the air reheater and
enters the scrubbing system at 675 m /s (1,430,000 acfm) and
2-156
-------
TABLE 2-37. FGD SYSTEMS DATA FOR COLSTRIP 1,
COLSTRIP, MONTANA
Unit rating (gross), MW
(net), MW
Fuel
Average fuel characteristics:
Heating value, kj/kg (Btu/lb)
Ash, %
Moisture, %
Sulfur, %
Chloride, %
FGD process
FGD system supplier
Application
Status
Startup date:
Initial
Commercial
Design removal efficiency:
Particulate, %
Sulfur dioxide, %
Actual removal efficiency:
Particulate, %
Sulfur dioxide, %
Water loop
Total water makeup,
liters/s per net MW
(gal/min per net MW)
Sludge disposal
358
330
Subbituminous coal
20,570
8.6
23.9
0.77
(8,843)
Alkaline fly ash and lime slurry
A.D. Little/Combustion Equipment
Associates
New
September 1975
November 1975
99.5
60.0
99.5
75.0
Closed
Unstabilized sludge is disposed of in
an offsi-hf* unlin^d nnncl
2-157
-------
TABLE 2-38. FGD SYSTEMS DATA FOR COLSTRIP 2,
COLSTRIP, MONTANA
Unit rating (gross), MW
(net), MW
Fuel
Average fuel characteristics:
Heating value, kj/kg (Btu/lb)
Ash, %
Moisture, %
Sulfur, %
Chloride, %
FGD process
FGD system supplier
Application
Status
Startup date:
Initial
Commercial
Design removal efficiency:
Particulate, %
Sulfur dioxide, %
Actual removal efficiency:
Particulate, %
Sulfur dioxide, %
Water loop
Total water makeup,
liters/s per net MW
(gal/min per net MW)
Sludge disposal
358
330
Subbituminous coal
20,570 (8,343)
8.6
23.9
0.77
Alkaline fly ash and lime slurry
A.D. Little/Combustion Equipment
Associates
New
May 1976
August 1976
99.5
60.0
99.5
75.0
Closed
Unstabilized sludge is disposed of in
an nffg-j-t-pi jun-linpd nond
2-158
-------
01
VO
FLUE
GAS
PLENUM
BLEED
MIST
ELIMINATORS
WASH TRAY
V
RECYCLE PUMPS
EMERGENCY WATER
00-SEAL WATER SUPPLY
T^LUNDERSPRAY
MAKEUP
WATER
TRAY RECYCLE
STEAM
*-CLEAN FLUE GAS EE7V
FLY ASH POND
V!
POND
RETURN
7
WASH TRAY POND
Figure 2-35. Process diagram of FGD module at Montana Power Co. Colstrip 1 and 2.
-------
144°C (291°F). Flue gas enters one of the three plenum sections,
where initial fly ash removal occurs. Flue gas is ducted to the
top of the module and flows down into the venturi section, which
forms a column through the center of the module. The gas is
saturated, and most of the particulate matter is removed in the
venturi section. Initial sulfur dioxide removal also occurs.
The venturi is equipped with a variable throat to maintain a con-
stant pressure drop. Each venturi throat has a throat a 4.6 m
(15-ft) in diameter. The throat is constructed of carbon steel
and lined with abrasion-resistant bricks. The plumb bob, which
controls the pressure drop, is made of 316 stainless steel. The
gas flows down from the venturi toward the reaction tank and
abruptly turns 160 degrees; it then rises through the spray zone,
which forms a ring around the venturi section. The rising gas
contacts a recycle spray of absorption slurry. The slurry from
the venturi and the absorber spray is collected and held in the
base of the module; it is recirculated at an L/G ratio of 2
liters/m3 (15 gal/103 ft3) for the venturi and 2.4 liters/in3 (18
3 3
gal/10 ft ) for the absorber spray. The absorber section is
constructed of carbon steel and lined with flaked glass poly-
ester. After the spray zone, the gas encounters a 316 stainless
steel wash tray that removes most of the entrained slurry and
water droplets before the mist eliminator.
After the wash tray, the gas rises up through a four-pass
horizontal chevron and a mesh pad mist eliminator, both con-
structed of Norel plastic. The gas leaves the absorption section
at a temperature of about 49°C (120°F) and enters an in-line
steam coil reheat system that boosts the wet gas temperature 28
to 42 (50 to 75°F). The reheated gas then passes through the dry
induced-draft fans and is discharged to atmosphere from the top
of the 152-m (500-ft) stack.
Performance—The overall performance of the FGD systems at
Colstrip 1 and 2 has been good, although certain problems were
encountered. The monitors for stack opacity and nitrogen oxides,
as well as those for sulfur dioxide on the inlets and outlets of
2-160
-------
the scrubber modules, have exhibited erratic behavior. A cor-
rective program, however, has improved monitor data. The inline
pH probes were eroding and loosing sensitivity because of de-
posits on the elements. Slurry density monitors were also
erratic, and a test program was undertaken to isolate these
devices from vibration. Both slurry density and pH, key operat-
ing parameters, had to be taken manually.
Because of deposits (not calcium sulfate or calcium sulfite
scale) on the wet-dry interface at the entrance to the scrubber
modules, system availability was reduced significantly. Tests
showed that there was a maldistribution of the dust entering each
module and that gas velocity varied enormously from one side to
the other of each entrance. A model study of the liquid flow on
the tangential shelf above the venturi indicated that gas flow
turning vanes in the duct elbow above each scrubber module and
liquid guiding vanes and baffles on the tangential shelf were
required to reduce buildup at the wet-dry interface. One module
was modified with such encouraging results that modifications
were made to the remaining five modules. Actual operating expe-
rience from the modules so equipped has shown a significant
reduction in cleaning time and has increased the time increment
between inspection and cleanings.
The supplementary alkali feed system was difficult to use
because of line plugging and equipment failure. The lime system
was modified to be simple to use and less prone to plugging.
Pinch valves were placed next to the scrubber, on each feed line,
so that solids accumulation is automatically cleared as these
valves are operated. Although the alkaline ash captured by the
scrubber has generally kept scrubber pH within the control range,
it was hoped that better control of the lime system would provide
more constant pH and allow optimal sulfur dioxide removal.
Failures of the protective flaked glass lining on the carbon
steel vessel walls and ducts occurred. These failures were
evident prior to a major temperature excursion at Colstrip 1 in
October 1976. The excursion followed a complete station power
2-161
-------
blackout and the failure of the emergency quench water supply
system for the scrubbers to operate. The flaked glass lining and
the plastic mist eliminators were seriously damaged at Colstrip
1 and accounted for low scrubber availability during November and
December 1976.
The motor of an induced-draft fan failed during November and
December 1976 independently of the temperature excursion. This
failure also reduced availability. However, if the fan motor
failure had not been considered, the scrubber availability at
Colstrip 1 would have been 88.5, 94.0, and 81.6 percent for
October, November, and December, respectively. Problems with
other induced-draft fan motors early in 1977 reduced scrubber
availability.
The availability of 63 percent at Colstrip 2 during May 1977
reflects fan motor repairs and the modifications to reduce build-
up at the wet-dry interface.
Inspection of an induced-draft fan rotor during the spring
of 1977 revealed cracks in the center plate next to the blades.
Also other fans were then cleaned and inspected. When cracks
were found or when it was suspected that cracks might form, the
faulty portions of fans were ground and welded; and stiffener
plates were added. Availabilities during the summer of 1977
reflected this repair work.
Quick clean basket strainers were added to the suction pip-
ing of the two main recycle pumps of each vessel to replace the
startup strainers contained in a pipe spool. This replacement
has allowed deposits and foreign materials to be removed before
they clog spray nozzles and has also decreased maintenance and
downtime on the vessles.
A program was undertaken to evaluate abrasion-resistant
protective lining materials that could be used where the flue gas
makes a 180-degree turn after the venturi downcomer and passes
through the absorption sprays. The absorption spray slurry
erodes the flaked glass lining, which has been replaced at least
once in each module since startup.
2-162
-------
A program for evaluating corrosion-, temperature-, and
abrasion-resistant materials in other areas of the scrubber was
also undertaken. Small test patches and full linings were tried
on several scrubber modules.
Plugging of spray nozzles under the wash trays on scrubber
modules of both units occurred during May 1977 and forced the
plants to shut down. The major cause of the problem was piping
that had broken in the wash tray pond and allowed material from
the pond to be ingested. A parallel pipeline and pumping system
was installed to bring the units back on line, and repairs were
made to the original piping.
The scrubber ponds are reclaimed by dredging sludge to a
disposal pond approximately 4.8 km (3 miles) from the plant site.
Liquid from the pond is returned to the scrubber ponds when the
dredge is not operating. The scrubber sludge deposits in deltas
in the ponds and contains from 55 to 60 percent solids. The
material hardens so well that the dredge has had difficulty in
reslurrying the sludge. Because of the residual alkalinity, the
deposited sludge becomes like cement. Improved cutters for the
dredge and a delumper ahead of the dredge pump are expected to
increase the solids handling capacity. The wash tray pond
filled with solids more rapidly than had been expected, and the
removal of these solids has been a problem. Each pond at the
plant has its own chemical identity based on water balances.
Because dredge cannot be used to empty the wash tray pond without
adversely affecting the plant operation or pond water balances,
a less efficient clamshell operation was put into use. Plans are
being developed for the addition of another wash tray pond, so
that one may be dewatered and cleaned while the other is in
operation.
Many emission tests have been conducted on the Colstrip
units. During the spring of 1977, EPA compliance tests for
particulates, sulfur dioxide, and nitrogen oxides were completed
on both plants; and emission monitor certification testing was
conducted after the compliance tests. Table 2-39 gives background
2-163
-------
TABLE 2-39.
BACKGROUND INFORMATION FOR SO.
EMISSIONS AT COLSTRIP 1 AND 253 *
Emissions of S02
kg/h (Ib/h)
ng/J (lb/106 Btu)
ppm
Maximum emissions allowed
by NSPS for a 358-MW unit
Scrubber guarantee for a
358-MW unit
Projected emissions from
pilot plant for a 358-MW
unit firing:
Coal with a sulfur
content of 0.78 percent
and an ash content of
8.19 percent
Coal with a sulfur
content of 1.0 percent
and an ash content of
12.85 percent
1844 (4063)
1537 (3386)
633 (1394)
940 (2071)
516 (1.2)
430 (1.0)
510
425
176 (0.41)
185
262 (0.61)
260
2-164
-------
information about sulfur dioxide emissions at Colstrip 1 and 2.
Tables 2-40 and 2-41 show the results of the tests for sulfur
dioxide and indicate that sulfur dioxide emissions are well below
the maximum allowed by the guarantee and Federal standards.
Although efficiencies cannot be accurately calculated because
scrubber inlet data have not been taken, it appears that a sulfur
dioxide removal efficiency between 70 and 75 percent is being
achieved.
Figures 2-36 and 2-37 show the availabilities of Colstrip 1
and 2, respectively. Table 2-42 lists the data upon which
Figures 2-36 and 2-37 are based and includes qualifications for
data when necessary.
2.4.1.2 Minnkota Power Cooperative, Milton R. Young 2 —
The Milton R. Young Station is owned and operated by Minnkota
Power Cooperative and Square Butte Electric Cooperative. It is a
two-unit station in Center, North Dakota. Milton R. Young 1 is
an existing 240-MW (gross) unit; Milton R. Young 2, a new 477-MW
(gross) unit.
The electric power generating facilities at Milton R. Young
2 consist of a Babcock and Wilcox cyclone steam generator with a
turbine generator. Milton R. Young 2 began operation in June
1977 and was declared commercial in June 1978.
The coal burned at Milton R. Young power station is low-
sulfur (0.7 percent sulfur) North Dakota lignite with a heating
value of 15,000 kJ/kg (6,500 Btu/lb). Its aah content is 7
percent.
Primary particulate control is provided by a 99.6 percent
efficient cold-side electrostatic precipitator supplied by
Wheelabrator-Frye. Sulfur dioxide is controlled with an FGD
system supplied by A. D. Little and Combustion Equipment Asso-
ciates. The FGD system utilizes a solution of solubilized
alkali obtained from the fly aah collected in the ESP's and
supplemented by lime. The FGD system is designed to remove 85
percent of- the sulfur dioxide from 85 percent of the boiler flue
gas for an overall aulfur dioxide removal efficiency of approx-
imately 70 percent. - ir_
*—J.O5
-------
TABLE 2-40. TESTS OF S02 EMISSIONS AT COLSTRIP 1 3
Date
2/76
4/76
7/76
9/76
12/76
1/77
5/77
6/77
Coal as received
Gross MW
353a
210b
184b
186b
223b
230b
331a
340a
Heating value,
kJ/kg (Btu/lb)
20,092 (8638)
20,611 (8861)
20,485 (8807)
20,080 (8633)
19,524 (8394)
20,280 (8719)
20,587 (8851)
20,673 (8888)
Ash content,
%
9.03
7.79
8.49
7.93
8.54
8.17
8.41
8.96
Sulfur
content, %
0.83
0.71
0.64
0.62
0.94
0.77
0.61
0.91
Emissions of SO-
kg/h
db/h)
720 (1587)
207 (456)
117 (258)
125 (275)
408 (898)
350 (770)
251 (552)
528 (1164)
ng/J
(lb/106 Btu)
206 (0.48)
99 (0.23)
64 (0.15)
64 (0.15)
185 (0.43)
150 (0.35)
69 (0.16)
140 (0.325)
ppm
197
87
53
56
154
133
72
122
to
I
M
Three scrubber modules were online.
Two scrubber modules were online.
-------
TABLE 2-41,
TESTS OF S02 EMISSIONS AT COLSTRIP 2
53
Date
10/76
11/76
12/76
3/77
6/77
Coal as received
Gross MWa
331
327
324
335
305
Heating value,
kJ/kg (Btu/lb)
19,464 (8368)
19,734 (8484)
20,213 (8690)
20,327 (8739)
20,709 (8929)
Ash content,
%
7.96
7.86
7.87
7.96
8.48
Sulfur
content, %
0.56
0.59
0.64
0.63
0.72
Emissions of SO_
kg/h
(Ib/h)
594 (1309)
316 (696)
354 (780)
301 (664)
271 (597)
ng/J
(lb/106 Btu)
181 (0.42)
99 (0.23)
107 (0.25)
86 (0.20)
81 (0.19)
ppm
178
83
98
84
67
to
a
Three scrubber modules were online.
-------
a\
oo
100
90
80
70
'. 60
50
40
30
20
10
100
90
80
70
* 60
i—
_j
m 50
_J
5 40
30
20
10
JAN FEB MR APR MAY JUN JUL RUG SEP OCT NOV DEC JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
1976 1977
JAN FEB MAR APR MAY JUN JUL
1978
Figure 2-36. FGD system availability, Montana Power, Colstrip 1.
-------
to
I
VO
100
90
80
70
". 60
>•
»-
S so
I 40
30
20
10
i i i i '' '' ' _ I - 1 - 1 - 1 - 1 - 1 - 1
1 - '
100
90
80
70
•V 60
5 SO
i*
30
20
10
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY JUN JUL AU6 SEP OCT NOV DEC
1976
JAN FEB MAR APR MAY JUN JUL
1978
Figure 2-37. FGD system availability, Montana Power, Colstrip 2.
-------
TABLE 2-42. SCRUBBER AVAILABILITY FOR COLSTRIP 1 AND 2
Month/year
Jan. 1976
Feb. 1976
Mar. 1976
Apr. 1976
May 1976
Jun. 1976
Jul. 1976
Aug. 1976
Sep. 1976
Oct. 1976
Nov. 1976
Dec. 1976
Average
Jan. 1977
Feb. 1977
Mar. 1977
Apr. 1977
May 1977
Jun. 1977
Jul. 1977
Aug. 1977
Sep. 1977
Oct. 1977
Nov. 1977
Dec. 1977
Average
Jan. 1978
Feb. 1978
Mar. 1978
Apr. 1978
May 1978
Jun. 1978
Jul. 1978
Average
1
90
98
98
74
97
0
93
95
89
80
63
74
79
93
95
83
85
87
85
93
93
96
96
98
91
96
100
92
100
66e
76
96
89
2
100
100
99
96
98
90
94
93
96
83
94
97
85
63
88
91
81
9°h
77b
98
98
87
97
95c
87^
99
97
96
94
In general, scrubber availability for Colstrip is the total
module hours available divided by three, and multiplied by the
hours of the month (multiplied by 100). Because of extensive
scheduled outages, the values from May through August 1976
are based only upon days of operation.
This value is a conservative estimate because the scrubbing
system was only unavailable during part of boiler outage where
scrubber maintenance was performed. This figure penalizes the
scrubbing system for the entire outage.
This value is based upon the 17 days of operation before the
unit was shut down for a scheduled annual overhaul.
This value is based upon 8 days of operation in April after a
unit overhaul.
This value is based upon 9.92 hours of operation on two scrub-
bers, while unit 1 was being brought back on line after comple-
tion of its annual overhaul. The ID fan motor was not available
at unit startup on the 1A module.
2-170
-------
The spent scrubbing solution is discharged to a recycle tank
at the bottom of the spray towers, where the chemical reactions
are completed. Flue gas cleaning wastes (unused reagent, sulfite/
sulfate salts, fly ash) are discharged from the recirculation
loop to a thickener, where the waste solids settle out. The
overflow is returned to the process and used for preparation of
the fly ash slurry. The thickener underflow has a maximum solids
content of 40 percent and is piped to a vacuum thickener building,
which houses two rotary drum vacuum filters for additional solids
dewatering. Filter cake with a solids content of 70 percent is
produced at 45 Mg/h (50 tons/h) and transported by conveyor to
three 32-Mg (35-ton) off-the-road trucks, which haul it to a mine
landfill for final disposal. Water recovered in the vacuum
filter is also used for preparation of the fly ash slurry.
During emergency conditions (mechanical malfunction and high flow
rates), sludge from the vacuum filters can be piled outside the
vacuum filter building for handling by front-end loaders and off-
the-road trucks.
Milton R. Young 2 uses about 9 MW of its gross capacity to
operate the sulfur dioxide control systems. Another 60 MW is
required to operate auxiliary station equipment.
FGD system data are summarized in Table 2-43.
FGD system—Sulfur dioxide is controlled by FGD system
consisting of two parallel spray towers and a chevron mist elim-
inator system near the top of each module. The wet gas is
reheated with hot particulate-cleaned flue gas. Alkaline fly ash
is fed to a silo from the ESP's of Units 1 and 2, and the lime
slurry reagent is only used when necessary.
Process flow—Design criteria and operating parameters for
the FGD system were determined during an extensive test program
at a pilot plant with a gas flow of 2.4 m /s (5,000 acfm). The
program was conducted by Minnkota Power Cooperative and the
system suppliers in conjunction with the Minnesota Power and
Light Company and the Grand Forks Energy Research Center.
2-171
-------
TABLE 2-43. FGD SYSTEM DATA FOR MILTON R. YOUNG 2,
CENTER, NORTH DAKOTA
Unit rating (gross), MW
(net) , MW
Fuel
Average fuel characteristics:
Heating value, kJ/kg (Btu/lb)
Ash, %
Moisture, %
Sulfur, %
Chloride, %
FGD process
FGD system supplier
Application
Status
Startup date:
Initial
Commercial
Design removal efficiency:
Particulate, %
Sulfur dioxide, %
Actual removal efficiency:
Particulate, %
Sulfur dioxide, %
Water loop
Total water makeup,
liters/s per net MW
(gal/min per net MW)
Sludge disposal
477
408
Lignite
15,000 (6,500)
8.0
38
0.7
Lime with alkaline fly ash
Arthur D. Little/Combustion
Equipment Associates
New
Operational
June 1977
June 1978
99.6
75.0
Open
0.1
1.56
Unstabilized sludge disposed
of in a landfill
2-172
-------
Boiler flue gas at 849 m3/s (1.8 million acfm) and 171°C
(34°F) first passes through the ESP. Inlet particulate loading
ranges from 1.6 to 3.4 g/m (0.7 to 1.50 gr/ft ) and averages
approximately 2.6 g/m (1.15 gr/ft ). The ESP is designed to
remove 99.5 percent of particulates, so that the outlet loading
3 3
averages 0.01 g/m (0.005 gr/ft ).
After the ESP, 85 percent of the flue gas is routed to the
FGD system. When coal with a maximum sulfur content of 1.3
percent is burned, approximately 1900 ppm (dry) sulfur dioxide
enters the FGD system. The flue gas contacts the scrubbing
solution in a series of vertical spray zones. The scrubbing
solution contains solubilized calcium oxide, magnesium oxide, and
sodium oxide from the fly ash; lime is added to the solution to
increase sulfur dioxide removal and pH control. At the maximum
sulfur dioxide inlet concentration, the absorbers are approxi-
mately 85 percent efficient. Because 85 percent of the flue gas
stream is treated, the total removal efficiency is approximately
70 percent; and the outlet sulfur dioxide concentration, 535 ppm
(dry).
When the boiler fires coal with a maximum sulfur content of
1.3 percent, the FGD system operates at a liquid-to-gas (L/G)
ratio of 11 liters/m3 (80 gal/103 ft3). The pH of the absorbent
ranges from 6.4 to 6.5, and the level of suspended solids is 12
percent. Lime is added at 3.6 Mg/h (4 tons/h).
After passing the spray zone of the absorbers, the scrubbed
gases enter chevron-type mist eliminators. A wash tray beneath
each mist eliminator dilutes any entrainment to prevent plugging,
scaling, and eventual restriction of gas flow. The wash trays
are joined to separate recirculation loops and tanks to ensure
that the wash water remains low in suspended solids and dissolved
salts. To prevent plugging, sprays irrigate the mist eliminators
with a mixture of makeup water and clarified liquor from each
tray loop. Also, sprays flood the underside of the trays and the
areas between each tray and the top recycle spray below it.
2-173
-------
The scrubbed gas stream is reheated by mixing it with the
bypassed flue gas stream at 171°C (340°F). This raises the
temperature of the saturated scrubbed gas stream from 57 to 74°C
(135 to 165°F); and the gases are discharged to the atmosphere
through the stack.
Performance—FGD operations at Milton R. Young 2 began in
September 1977 and were intermittent, mainly because of the
severe winter weather. The freezing and rupture of many lines
caused an outage during the first part of December for the
installation of heat tracing in the liquid circuit. Other
initial problems were reported with the guillotine gas dampers
and flowmeters.
An emergency unit shutdown that occurred on December 5,
1977, as a result of turbine bearing damage, made it necessary to
postpone compliance testing scheduled during late December or
early January. The unit came back on line on February 21, 1978;
but it was discovered that a forced-draft fan in the FGD system
had an oil leak and a shaft alignment problem. The fan was taken
off line for repairs, and one module of the FGD system was taken
out of service. The repaired fan was placed back in service on
April 10.
Other minor problems were encountered through September
1978, Large particles from the vacuum filter caused some rubber
lining downstream to peel; modifications to correct this problem
are being studied. Also, the guillotine damper chain drives were
discovered to be underdesigned and had to be replaced.
Compliance testing took place at the unit during the week of
June 5, 1978. Official results of the testing are not yet avail-
able.
2.4.2 Domestic Throwaway Sodium Scrubbing Units
The sodium scrubbing process has proven very successful for
desulfurization of flue gas from coal-fired utility steam gener-
ators at specific sites.
2-174
-------
Sodium scrubbing becomes cost effective where all of the
following conditions apply:
0 Emission regulations call for high sulfur dioxide
removal efficency from flue gas containing a relatively
low concentration of sulfur dioxide.
0 Waste material (sodium salts) can be disposed of with-
out the need for expensive pretreatment.
0 The sodium reagent material (trona) can be obtained at
a relatively low cost.
Plants that are in arid western regions near natural trona beds
or mines and that burn low-sulfur western coal are ideal candi-
dates for sodium scrubbing.
2.4.2.1 Nevada Power Company, Reid Gardner 1, 2, and 3
Reid Gardner Power Station of Nevada Power Company is a
three-unit station (with plans for a fourth unit) in Clark County,
near Moapa, Nevada. Reid Gardner 1, 2 and 3 are identical 125-MW
(gross) units. Reid Gardner 4 is planned to be a 250-MW (gross)
unit. The boilers of Reid Gardner 1, 2, and 3 were placed in
service in 1965, 1968, and 1976, respectively. The FGD systems
began operation in May 1974 (Units 1 and 2) and July 1976 (Unit
3).
The electric power generating facilities at Reid Gardner
Power Station consist of Foster-Wheeler pulverized-coal-fired wet
bottom steam generators with 125-MW (gross) turbine generators.
The coal burned at this station is medium-to-low-sulfur (0.5
to 1.0 percent sulfur) coal with an average heating value of
29,000 kJ/kg (12,450 Btu/lb) and ash content between 8.0 and 10.0
percent.
Primary particulate control is provided by Research-Cottrell
multiclones that remove about 75 percent (actual) of the inlet
particulate matter. Sulfur dioxide and secondary particulate
control is provided by a sodium carbonate scrubber system supplied
by Arthur D. Little and Combustion Equipment Associates; this
system is designed to remove 90 percent of all sulfur dioxide and
brings overall particulate removal efficiency to greater than 99
percent.
2-175
-------
The scrubber effluent is neutralized to a pH of 7.5 with
soda ash or trona and sent to a sealed settling pond for removal
of fly ash and insolubles. The settl'ng pond effluent is con-
veyed to a sealed evaporation pond, where all of the water is
evaporated. The settling pond and the evaporation pond are both
sealed with a clay slurry. There is an underground line from the
settling pond to the evaporation pond.
Reid Gardner Station uses about <*5 MW of its current gross
generating capacity of 375 MW to operate the emissions control
system and auxiliary station equipment.
FGD system data on Reid Gardner 1, 2, and 3 are summarized
in Tables 2-44, 2-45, and 2-46.
FGD system—The FGD system consists of two parallel venturi
scrubbers and a single-tray wash tower fed by the twin-throat
venturi system. The venturi throats are constructed of Incoloy
825; the rest of the venturi system is made of rubber-lined
carbon steel. The sieve plate in the absorber is 316L stainless
steel. The outlet ducts are lined with Plasite 4004-S epoxy.
The piping, pumps, and process vessels are rubber lined.
There is a horizontal mist eliminator that has radial vanes
and is 15 cm (6 in.) deep by 9.1 m (30 ft) in diameter. The mist
eliminator is near the top of the absorber tower.
Reheat of the wet cleaned gas is provided by an indirect
steam coil reheat system constructed of carbon steel. This
system heats ambient air and injects it into the exiting flue gas
stream.
Reid Gardner 1, 2, and 3 share the same facility for reagent
handling and storage. This facility includes storage and sand
removal equipment, a blender, a mix tank, a shaker, and related
equipment.
Process flow—The hot flue gas from the boiler passes
through mechanical collectors (multiclones) that remove about 75
percent of the fly ash. The flue gas is then drawn through a
2,200-KW (3,000-hp) fan, splits into two streams, and enters the
2-176
-------
TABLE 2-44. FGD SYSTEMS DATA FOR REID GARDNER 1,
MOAPA, NEVADA
Unit rating (gross), MW
(net), MW
Fuel
Average fuel characteristics:
Heating value, kJ/kg (Btu/lb)
Ash, %
Moisture, %
Sulfur, %
Chloride, %
FGD process
FGD system supplier
Application
Status
Startup date:
Initial
Commercial
Design removal efficiency:
Particulate, %
Sulfur dioxide, %
Actual removal efficiency:
Particulate, %
Sulfur dioxide, %
Watef loop
Total water makeup,
liters/s per net MW
(gal/min per net MW)
Sludge disposal
125
110
Coal
29,000 (12,450)
8
5.0-6.0
0.5-1.0
0.05
Sodium carbonate slurry
A.D. Little/Combustion Equipment
Associates
Retrofit
March 1974
April 1974
99.0
85.0
99+
85-95
0.078
(1.24)
Unstabilized sludge disposed of in
__aji__onsjite_golar evaporation pond
2-177
-------
TABLE 2-45. FGD SYSTEMS DATA FOR REID GARDNER 2,
MOAPA, NEVADA
Unit rating (gross), MW
(net), MW
Fuel
Average fuel characteristics:
Heating value, kj/kg (Btu/lb)
Ash, %
Moisture, %
Sulfur, %
Chloride, %
FGD process
FGD system supplier
Application
Status
Startup date:
Initial
Commercial
Design removal efficiency:
Particulate, %
Sulfur dioxide, %
Actual removal efficiency:
Particulate, %
Sulfur dioxide, %
Water loop
Total water makeup,
liters/s per net MW
(gal/min per net MW)
Sludge disposal
125
110
Coal
29,000 (12,450)
8
5.0-6.0
0.5-1.0
0.05
Sodium carbonate slurry
A.D. Little/Combustion Equipment
Associates
Retrofit
March 1974
April 1974
99.0
85.0
99+
85-95
0.078
(1.24)
Unstabilized sludge is disposed of in
i j-,e;n1 ar *» vaor a
2-178
-------
TABLE 2-46. FGD SYSTEMS DATA FOR REID GARDNER 3,
MOAPA, NEVADA
Unit rating (gross), MW
(net), MW
Fuel
Average fuel characteristics:
Heating value, kj/kg (Btu/lb)
Ash, %
Moisture, %
Sulfur, %
Chloride, %
FGD process
FGD system supplier
Application
Status
Startup date:
Initial
Commercial
Design removal efficiency:
Particulate, %
Sulfur dioxide, %
Actual removal efficiency:
Particulate, %
Sulfur dioxide, %
Water loop
Total water makeup,
liters/s per net MW
(gal/min per net MW)
Sludge disposal
125
110
Coal
29,000 (12,450)
8.0
5.0 - 6.0
0.5 - 1.0
0.05
Sodium carbonate slurry
A.D. Little/Combustion Equipment
Associates
New
June 1976
July 1976
99.0
85.0
99+
85-95
Open
0.078
(1.24)
Unstabilized sludge disposed of in
an onsite solar evaporation pond
2-179
-------
twin-throat venturi system. The gas is quenched with circulated
scrubbing liquor sprayed from tangential nozzles on the walls of
the Venturis and exits into a sump, where the reduction in
velocity causes the liquor droplets to fall. A liquid-to-gas
ratio of 1.3 liters/m3 (10 gal/103 ft3) and a pressure drop of
about 3.7 kPa (15 in.H2O) are maintained in the venturi. The
scrubbed gas at 540°C (130°F) tangentially enters the cylindrical
droplet separator tower, rises through the tower, and bubbles
through a single-sieve tray. The tray is flooded with clear
water from the ash pond at 0.16 liters/m (1.2 gal/10 ft ) of
gas. This tray is intermittently washed with fresh water sprayed
against the tray bottom. The cleaned gas continues up the tower
and passes through a single pass, horizontal mist eliminator with
radial vanes. No mist eliminator washing is provided. The
saturated gases are reheated by direct mixing with hot air, and
the combined stream is discharged to the atmosphere from a 61-m
(200-ft) stack.
The FGD system water is supplied from the ash handling pond.
The water comes originally from wells and from the Muddy River.
Trona, a low-grade ore containing 60 percent sodium carbonate, is
normally used as the scrubbing agent. The sodium carbonate
slurry is pumped to a clarifier, where insoluble impurities
settle out. The clarified sodium carbonate liquor is injected
into the venturi recirculation loop. The remainder of the makeup
water (pH 11) is injected in the absorber circuit through the
tray recycle tank. Slowdown from the venturi recycle tank is
mixed with the alkaline clarifier underflow stream and adjusted
to pH 7 in the postneutralization tank. Spent liquor from the
2
postneutralization tank is pumped to a 24,300-m (6-acre) ash
settling pond, and the overflow from this pond goes to a 182,000-
2
m (45-acre) evaporation pond. No liquor is recycled to the
modules from these ponds, and the pond evaporation rate is equal
to the scrubber effluent discharge rate.
Figure 2-38 shows a simplified process flow diagram of a
Reid Gardner FGD system.
2-180
-------
COAL
PULVERIZER
to
I
M
00
FURNACE
MECHANICAL
OUST
COLLECTOR
(MULTICLONES)
BOILER
F.D. FAN
BOILER
I.D. FAN
PURGE
TO ASH POND
TWIN
VENTURI
THROAT
SCRUBBER!
TO CHIMNEY
i
BYPASS „
~s?(w«
ASH POND PUMP
TEAM
REHEATER AIR FAN
EVAPORATION PONDS
ON MESA
MATER FROM
ASH POND
ASH POND
ASH AND WATER
SOOA ASH
SLURRY PUMP
Figure 2-38. Process diagram of an FGD system, Nevada Power Co., Reid Gardner Station,
-------
Performance—There were many problems during the startup of
the FGD systems. These problems included improper guillotine
damper operations, liquid level control difficulties resulting
from undersized control valves (Units 1 and 2), accumulation of
rubber fragments in the system from the rubber linings, fly ash
plugging in the reject slurry and postneutralization lines,
plugging in the instrument sensing line, plugging in the water
control valves of the Venturis, and fly ash interference with pH
meter electrodes. The units also experienced corrosion where
rubber lining failures occurred and flaking of Ceilcote liner in
the scrubber outlet ducting. Another minor problem was excessive
leaking in the recirculation pump seals.
The problems with the guillotine dampers were resolved by
installing redesigned drive train components. The underdesigned
liquid level control valves were replaced where necessary. The
water control valves of the Venturis were flushed every shift to
eliminate plugging.
After shakedown and debugging of the FGD systems were
largely completed, performance improved, particularly at Reid
Gardner 2 and 3. Tables 2-47, 2-48, and 2-49 list the perform-
ance factors for Reid Gardner 1, 2, and 3 from startup to August
1978. Figures 2-39, 2-40, and 2-41 show system availabilities.
The Reid Gardner Power Station emissions control systems
were designed for overall particulate and sulfur dioxide removal
efficiencies of 99.0 and 85 percent, respectively. Actual par-
ticulate removal efficiencies have exceeded 99 percent; and
actual sulfur dioxide removal is typically 84 percent, although
efficiencies as high as 95 percent have been achieved by increas-
ing soda ash makeup.
17 18
2.4.3 Domestic Chiyoda Thoroughbred 121 Systems '
The first domestic Chiyoda Thoroughbred 121 system began
operation as a prototype unit on August 30, 1978, at the Scholz
Power Plant of Gulf Power. This unit uses a limestone slurry
process that is a second generation innovation of the CT-101
2-182
-------
TABLE 2-47.
FGD PERFORMANCE DATA, NEVADA POWER CO.,
REID GARDNER la
Month/year
Jan. 1975
Feb. 1975
Jul. 1975
Sep. 1975
Oct. 1975
Jan. 1976
Feb. 1976
Mar. 1976
Apr. 1976
Aug. 1976
Sep. 1976
Oct. 1976
Nov. 1976
Dec. 1976
Averaqe
Jan. 1977
Feb. 1977&
Mar. 1977b
Apr. 1977C
May 1977C
Jun. 1977
Jul. 1977
Aug. 1977
Sep. 1977
Oct. 1977
Nov. 1977
Dec. 1977
Average
Jan. 1978
Feb. 1978
Mar. 1978
Apr. 1978
May 1978
Jun. 1978
Jul. 1978
Aug. 1978
Sep. 1978
Average
Availability, %
65
95
81
91
99
15
97
95
98
87
93
84
72
0
0
0
0
47
99
19
84
99
79
85
49
56
97
28
100
97
89
100
94
99
84
Operability, %
85
78
60
29
78
72
91
75
97
96
84
92
90
67
0
0
0
0
45
99
21
73
99
82
82
47
68
75
58
97
92
89
99
93
X99
86
Reliability, %
57
89
99
100
94
95
98
81
88
79
87
0
0
0
0
44
99
19
69
99
79
83
48
69
94
58
100
96
100
100
94
97
90
Utilization, %
78
14
25
75
39
13
49
89
87
71
80
59
59
0
0
0
0
44
99
19
35
85
78
71
41
56
43
28
75
78
89
99
89
87
72
No data are available for April through June 1975, August 1975,
or November through December 1975; a boiler outage occurred
from May through July 1976.
Rubber liner failure in the FGD system caused ait outage.
Scrubber liner failure caused an outage; the failure occurred
because- of a faulty boiler air preheater.
2-183
-------
TABLE 2-48.
FGD PERFORMANCE DATA, NEVADA POWER CO.,
REID GARDNER 2a
Month/year
Feb. 1975
Jul. 1975
Sep. 1975
Oct. 1975
Nov. 1975
Jan. 1976
Feb. 1976
Mar. 1976
Apr. 1976
May 1976
Jul. 1976
Aug. 1976
Sep. 1976
Oct. 1976
Nov. 1976
Average
Feb. 1977
Mar. 1977
Apr. 1977
May 1977
Jun. 1977
Jul. 1977
Aug. 1977
Sep. 1977
Oct. 1977
Nov. 1977
Dec. 1977b
Average
Jan. 1978
Feb. 1978
Mar. 1978
Apr. 1978
May 1978
Jun. 1978
Jul. 1978
Aug. 1978
Sep. 1978
Average
Availability, %
72
84
67
85
99
91
95
94
95
52
83
93
92
95
99
94
40
87
74
76
86
0
76
67
93
98
100
100
100
80
81
100
91
Operability, %
90
85
77
87
99
66
86
60
77
83
81
75
96
95
58
78
75
86
95
85
93
41
91
67
80
0
71
74
92
89
98
100
92
82
93
100
91
Reliability, %
66
86
62
83
98
96
94
94
95
51
83
88
97
95
98
94
40
87
71
76
0
75
67
92
97
100
100
100
79
82
97
90
Utilization, %
69
62
83
62
83
53
68
79
57
71
91
88
50
70
56
76
95
60
83
40
79
64
80
0
63
67
87
80
44
97
92
74
81
94
00
No data are available for March through' June 1975, December 1975,
or June 1976; boiler outages occurred in December 1976, January
1977, and November 1977.
ID fan was repaired.
2-184
-------
TABLE 2-49. FGD PERFORMANCE DATA, NEVADA POWER CO., REID GARDNER 3
Month/year
Jul. 1976
Aug. 1976
Sep. 1976
Oct. 1976
Nov. 1976
Dec. 1976
Average
Jan. 1977
Feb. 1977
Mar. 1977
Apr. 1977
May 1977
Jun. 1977
Jul. 1977
Aug. 1977
Sep. 1977
Oct. 1977
Nov. 1977
Dec. 1977
Average
Jan. 1978
Feb. 1978
Mar. 1978
Apr. 1978
May 1978
Jun. 1978
Jul. 1978
Aug. 1978
Sep. 1978
Availability, %
45
56
49
22
28
99
50
98
81
44
86
63
98
73
70
90
99
89
88
82
100
96
97
97
97
100
100
98
100
Operability, %
46
43
46
29
80
97
57
98
74
51
85
26
63
73
63
90
90
88
92
74
100
95
97
89
77
96
80
98
100
Reliability, %
70
50
46
21
29
99
53
98
72
52
85
21
64
73
63
99
99
89
96
76
100
95
97
97
96
96
100
98
97
Utilization, %
42
43
44
21
29
97
46
91
63
44
83
16
62
"73
50
90
61
87
93
64
100
88
96
87
66
95
78
97
32
to
I
M
00
(Jl
-------
to
I
M
00
10 -
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY JUN JUL AUG SEP
1976 1977 1978
Figure 2-39. FGD system availability, Nevada Power Co., Reid Gardner.
-------
ro
I
M
60
I I I I I I I I 1 I 1 I
10 -
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY JUN JUL AUG SEP
1976 1977 1978
Figure 2-40. FGD system availability, Nevada Power Co., Reid Gardner.
-------
00
00
I I I I I I I I I I I I i I I I I I I I I I I I I I
10 -
JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY JUN JUL AUG SEP
1976 1977 1978
Figure 2-41. FGD system availability, Nevada Power Co., Reid Gardner 3
-------
process previously described. Laboratory experiments and pilot
plant operations utilizing this process were first undertaken at
the Kawasaki Research and Development Center in Japan. The
economic recession in Japan and the applicability of the process
to coal-fired boilers prompted Chiyoda International to install
a prototype plant in the United States. All the equipment for
the prototype unit was made by modifying the equipment left from
the CT-101 pilot plant studies done at the Scholz Power Station.
2.4.3.1 Gulf Power, Scholz 1 and 219'20 —
The Scholz Power Plant is located in Chattahoochee, Florida,
approximately 64.4 km (40 miles) northwest of Tallahassee. The
plant consists of two steam electric generating units that pro-
vide a combined net generating capacity of 80 MW. The gross
rating of each unit is 47 MW. Both units have pulverized-coal-
fired, dry bottom boilers manufactured by Babcock and Wilcox. A
3
flue gas flow of 50 m /s (106,000 acfm) is developed by each.
The coal burned at the plant has an average heating value of
29,075 kJ/kg (12,500 Btu/lb) sulfur content of 2.5 percent, ash
content of 11 percent, and moisture content of 5 percent. Coals
with various sulfur contents, however, will be burned during
prototype operations. Each unit burns an average of 17.2 Mg (19
tons) of coal per hour. A Buell cold-side electrostatic precipi-
tator (ESP) designed for 99.5 percent removal of particulate
matter is located immediately downstream of each boiler. The
temperature of the flue gas exiting the ESP ranges from 135 to
168°C (275 to 335°F). The ESP can be selectively de-energized to
vary the particulate loading to the FGD system during experimental
operations.
FGD system—The Thoroughbred 121 process developed by
Chiyoda International utilizes a limestone slurry absorbent with
a new jet bubbling reactor illustrated in Figure 2-42. The
reactor is constructed of stainless steel and is 7 m (24 ft) in
diameter.- The flue gas enters a relatively shallow liquid layer
(the jet bubbling zone) through vertical spargers. The open ends
2-189
-------
WATER
FLUE GAS
AIR
CLEAN GAS
GYPSUM SLURRY
LIMESTONE SLURRY
Figure 2-42. Jet bubbling reactor.
2-190
-------
of these spargers are submerged from 10 to 41 cm (4 to 16 in.)
below the liquid surface. The velocity of the flue gas ranges
from 5 to 20 m/s (16 to 66 ft/s) and causes it to entrain the
surrounding liquid, so that a jet bubbling or froth layer is
created. The resulting large gas-liquid interface provides
effective transfer of sulfur dioxide. The gas contact time
ranges from 0.5 to 1.5 seconds.
In the reaction zone (immediately below the jet bubbling
zone), the liquid is moderately agitated by air bubbling and
mechanical agitation. Oxidizing air is introduced into the
reactor by air spargers at levels from 200 to 300 percent those
of stoichiometric requirements. Excess oxidizing air, sufficient
residence time, and adequate suspended solids promote gypsum
crystal growth. The gypsum settles to the bottom of the tank at
a minimum rate of 5 cm/s (10 ft/h), and a bleed stream of gypsum
is continuously drawn off. The gypsum is discharged from the
reactor in a slurry containing from 10 to 25 percent solids. The
slurry is pumped to a gypsum stack where the solids settle out by
gravity, and the supernatant (stack overflow) is pumped back to
91 99
the process. '
There is a vertical, two-stage, double-pass chevron mist
eliminator. The scaling problems frequently encountered with
mist eliminators of conventional limestone systems should be
avoided with the jet bubbling reactor because of the high gas
velocity and resulting small amount of mist produced. Mainte-
nance for the mist eliminator will include a 1-minute wash of the
first stage every 2 weeks.
The scrubbed gases exit through a 23-m (76-ft) fiber-rein-
forced plastic stack. The outlet temperature is 54°C (130°F).
No reheat is provided.
Additional system design information is presented in Table
2-50.23
Process chemistry—The chemistry of this new process is
similar to that of conventional limestone scrubbing processes,
2-191
-------
TABLE 2-50. FGD SYSTEMS DATA FOR SCHOLZ 1 AND 2,
CHATTAHOOCHEE, FLORIDA
Unit rating (gross), MW
(net), MW
Fuel
Average fuel characteristics:
Heating value, kj/kg (Btu/lb)
Ash, %
Moisture, %
Sulfur, %
Chloride, %
FGD process
FGD system supplier
Application
Status
Startup date:
Initial
Commercial
Design removal efficiency:
Particulate, %
Sulfur dioxide, %
Actual removal efficiency:
Particulate, %
Sulfur dioxide, %
Water loop
Total water makeup,
liters/s per net MW
(gal/min per net MW)
Sludge disposal
47 (each)
40 (each)
Coal
29,075 (12,500)
11.0
5.0
2.5 (5.0 maximum)
Limestone
Chiyoda International
Prototype (retrofit)
August 1978
99.5
90.0+
Closed
FGD waste in the form of gypsum is
in -hhe t*-xi gfiina onsite pond
2-192
-------
except that virtually all sulfite produced in the reactor is
forcibly oxidized to sulfate. Only trace amounts of sulfite
remain after the oxidation process.
The overall reaction equation for the system is:
SO2 + CaC03 + 1/2 02 +
The principal chemical reactions in the jet bubbling (froth)
zone are:
+ S02(aq)
SO0(aq) + H00 + H
f, £• ^~ *• -J
HS03~ - S03~ + H+
SO3+ 1/2 O2 (aq) j
CaC03U) ^ CaC03(aq)
CaC03(aq) + H+ j Ca++
HCO3~ + H+ j H2O + C02
Ca++ + S04~~ J CaS04
CaSO4 + 2H2O j CaSO4'2H2O
24
The reaction equations in the reaction (liquid) zone are:
02(g) + 02(aq)
S03"~ + 1/2 02(aq) * SO4~"
j CaC03(aq)
CaCO-(aq) + H+ j Ca++ + HCO ~
J *3
Ca++ + SO "~ + 2H00 -»• CaSO.,»2H00
4 2 «- 42
Ca++ + SO4j CaSO4
CaS04 + 2H20 + CaS04-2H20 (4-)
95
Process flow —A schematic process flow diagram is pre-
sented in Figure 2-43. After the ESP, the flue gas flows into
the jet bubbling reactor at a typical dry particulate loading of
2-193
-------
JET BUBBLING REACTOR
STACK
WATER
Figure 2-43. Process flow diagram of prototype plant, Gulf Power Co., Sholz Station,
-------
0.09 g/m3 (0.04 gr/ft ). The FGD system is designed to handle 50
percent of the flue gas from either boiler, or 25 m /s (53,000
acfm). The system has an equivalent electrical capacity of 23
MW. The temperature of the gas as it enters the reactor ranges
from 135 to 168°C (275 to 335°F). In the reactor, it is quenched
with water and sparged into the limestone absorbent through an
array of vertical sparges to produce the froth layer. Sulfur
dioxide is absorbed in this layer to produce sulfite, which is
completely oxidized to sulfate. The scrubbed gases pass through
the chevron-type mist eliminator and are discharged from the 23-m
(76-ft) fiber-reinforced plastic stack at 54°C (130°F). This
prototype unit is designed to remove more than 90 percent of the
sulfur dioxide from the flue gas.
26 27
Performance ' —Operation of this experimental unit began
on August 30, 1978. An availability of greater than 99 percent
since startup has been reported. No problems have yet been
encountered. The sulfur dioxide removal efficiency of the
*
system has averaged approximately 93 percent from startup through
the end of September.
2.4.4 Domestic Dry Phase Collection "/stems
Integrated processes designed to remove two or more pollu-
tants simultaneously from the flue gas of a coal-fired boiler
have commanded considerable interest because of economic and
operational reasons. The success of fabric filters in removing
particulates from the flue gases of coal-fired boilers has moti-
vated investigations into the feasibility of using the same
equipment to control both particulates and sulfur dioxide. Tests
have been conducted with the injection of dry powdered nahcolite
(a mineral form of sodium bicarbonate) into the flue gas stream
and onto fabric filter bags for the removal of sulfur dioxide.
The bicarbonate, when heated, gives off water and carbon dioxide
to form a porous carbonate which has a large specific surface
area and reacts readily with sulfur dioxide. Nahcolite was
chosen as the source of bicarbonate because there are large
2-195
-------
reserves, principally in shale deposits in northern Colorado. It
was anticipated that nahcolite could be mined at a relatively low
cost and delivered as a usable raw product with a purity between
70 and 80 percent.
During test in mid-1974 at the Nucla Station of Colorado-Ute
X
Electric Association, dry nahcolite powder was injected into the
flue gas stream and onto existing fabric filter bags. These
tests provided performance data for full-scale operation of a
small power plant. The sulfur dioxide removal efficiency was
approximately 69 percent; and the utilization of the alkali
content of the injected material, 56 percent. Figure 2-44 il-
lustrates the process flow of a dry phase nahcolite injection
system.
Built in features of the fabric filter particulate collec-
tion hoppers and the method of alkali injection made nahcolite
drop out roughly 28 percent and thus significantly reduced the
effective sorbent utilization. Test data suggested that there
should be an optimum condition for injection of alkali which
would promote maximum utitlization of absorbent.
In November 1974, the Otter Tail Power Company conducted
five tests to determine whether or not a fabric filter would be
feasible not only for the collection of solid particulates from
lignite, bat also for sulfur dioxide removal. Tests were per-
formed at a single point in the duct at the inlet to a cold-side
precipitator serving Unit 3 at the Hoot Lake Station in Fergus
Falls, Minnesota. A pilot fabric filter with an air-to-cloth
ratio of 2.3:1 was used. The filter could handle 0.002 m /s (5
acfm) of gas. Various precoat materials consisting of lignite
ash (collected from precipitator hoppers) and alkaline additives
(nahcolite, calcium oxide, and calcium hydroxide) were tried.
The program was under the general supervision of the Otter
Tail Power Company and was directed by Wheelabrator-Frye, Inc.
Members of the Grand Forks Office of the U.S. Bureau of Mines
(now within the U.S. Department of Energy) measured fly ash
2-196
-------
to
I
PRECIPITATOR
BOILER f "I
I I J
* * *
PULVERIZED AIR FLY ASH
COAL HEATER
T L ,
ACU 1 | . ,_. ,._
1 v , _., M 4 ROTARY
VunDDCD/ W rvjIMKI
\HUrrtly » FFPHFR
\__^^^^^ ^^^ 1 UUUL.I\
T GROUND HOPPER
T NAHCOLITE ORE 1 ) ROTARY
BOTTOM ASH (SLUDGE OR V FEEDER
CONTINUOUS FEED) \ { ^
y^lAHCOLITE £=d~
_y~| ^-ZT? °^ BLOWER
J[ CRUSHING GR
TRUCK DUMP AND *^ MODULE M(
SURGE BIN 1 1
A rn
GROUND
NAHCOLITE
ORE
' (PRECOAT)
RAfiHOUSE
*" '
\ /
I CONVEYOR
FLY ASH
^™ Wa,.cri
iiaoou^
1 NAHCOLITE
ORE
J"\ AIR
FAN
^^T
12
1 T0
\ LANDFILL
INDING ROTARY FEEDER
DDULE *"|
JL
v
BLOWER T
/ — 1 "
HOPPER
ROTARY
FEEDER
i
/STACK!
Figure 2-44. Dry phase collection with nahcolite injection.
-------
partic-_late concentration, and Wheelabrator-Frye conducted the
field -^sting of the pilot baghouse.
Sulfur dioxide removal efficiency was 94 percent at Hoot
Lake 3 immediately after the fabric filter was precoated with
nahcolito; average efficiency was 46 percent. Precoating with
calciun oxide and calcium hydroxide were less efficient. The
Nucla -est results and other factors motivated the Otter Tail
Power Company and its partners to undertake further tests with
the dry powder injection. In late 1976, a relatively large pilot
progra- was instituted at the Leland Olds plant of the Basin
Electrir Power Cooperative. The aim of the test was to obtain
%
data or. fabric filter performance under conditions similiar to
those t>..it will exist at the 400-MW (net) lignite fired Coyote
Station of otter Tail Power that is now slated to begin operation
in 1981. A description of the pilot plant facility is provided
" s 9Q
below.'" "*9
2.4.4.1 Basin Electric Power Cooperative, Lelands Olds Pilot
Plant30-37__
The initial pilot installation at the Leland Olds Power
Station imrned lignite in a boiler with a cyclone furnace similar
to the f:iture Coyote installation. The installation consisted of
two modules, each with six full-size filter bags. Each bag was
29.2 cr. ill.5 in.) in diameter and 9 m (30 ft) tall. The bags
were arranged to receive a slipsteam of flue gases at approxi-
mately 1.4 m /s (3000 acfm). The program was based principally
on the w^rk done by Wheelabrater-Frye at the Nucla Station, but
was modified on the basis of experience by the Superior Oil
Company, the nahcolite supplier.
The results were better than those of the Nucla tests. The
percentage of sulfur dioxide removed in relation to the percent-
age of srrbent utilized (the R:U factor) increased from 69:56 at
Nucla first to 83:77 and latter to 90:60 at Lelands Olds. As the
later te^t shows, the sulfur dioxide removal efficiency of 90
percent was achieved at some sacrifice in sorbent utilization.
2-198
-------
Steadily advancing techniques improved performance of the
nahcolite dry powder additive system at Leland Olds. For low-to-
medium-sulfur coals, this system appeared to offer simple, low
cost operation at low investment. There was, however, growing
concern regarding the availability of the nahcolite reactant.
As late as the spring of 1977, the three potential suppliers
of nahcolite were experiencing various difficulties in meeting
government requirements for excavation in the nahcolite resource
areas of northern Colorado. These difficulties prompted further
review of available options.
It was found that using a spray dryer would eliminate the
exclusive need for nahcolite. Thus, Wheelab.rator-Frye and
Atomics International undertook a joint test program at Lelands
Olds with a two-stage system combining a spray dryer and fabric
filter. The spray dryer was the first stage; it accomplished
alkali injection, as well as primary sulfur dioxide removal. The
downstream fabric filter functioned as a second stage sulfur-
dioxide absorber and flue gas particulate collector. The pilot
plant facilities at Leland Olds were modified to accommodate the
operation of a spray dryer 2 m (7 ft) in diameter with the
existing fabric filter modules. Figure 2-45 shows a simplified
process flow diagram of the two-stage dry scrubber/sulfur dioxide
absorber.
The test program had three main aims. The first was to
demonstrate that the process was a viable method of desulfurizing
flue gases from a lignite-fired utility boiler with a cyclone
furnace. The second aim was to compare the effectiveness of soda
ash, trona, and the less costly slurries of lime, limestone, and
fly ash. Finally, a week-long, round-the-clock endurance test
was planned to demonstrate the continuous operability of the
spray dryer/fabric filter system and to uncover any limitations
of extended operation.
A 2-month program included tests with sodium carbonate at
flue gas flows between 0.5 and 2.0 m /s (1000 and 4500 acfm) and
sulfur dioxide concentrations from 400 to 2300 ppm. Sodium car-
2-199
-------
to
I
N>
O
O
WASTEWATER
FROM
ASH POND
FEED AT L/G RATIO
OF APPROXIMATELY
0.04 liter/m3
(0.3 gal/103ft3)
CLEAN
FLUE GAS
ID FAN
FIRST-STAGE
SPRAY DRYER
SECOND-STAGE
FABRIC FILTER
FEED
PUMP
DRY PRODUCT
FOR DISPOSAL
Figure 2-45. Process flow diagram of two-stage dry scrubber/SO2 absorber,
-------
bonate solutions were fed to the spray dryer at varying rates to
produce temperature drops across the spray dryer between 50°C
(90°F) and 90°C (170°F). Actual operating temperature of the
fabric filter were between 74°C (165°F) and 110°C (230°F). No
degradation of sulfur dioxide removal efficiency was observed;
nor were any serious equipment shortcomings uncovered during the
week-long test of continuous operation.
A comparison of nahcolite injection with two-stage soda ash
removal is shown in Table 2-51.
As a result of the pilot testing, Wheelabrator Frye and
Atomics International have been awarded a turnkey contract for a
sulfur dioxide and particulate removal system at the 410-MW
Coyote 1 Station of the Otter Tail Power Company. The gas flow
3
at Coyote 1 is 892 m /s (1,890,000 acfm). The system will use
aqueous carbonate and is expected to begin operation in May 1981.
Montana Dakota Utilities will be the station operator.
After completion of the Wheelabrator Frye/Atomics Interna-
tional program, testing was conducted at Lelands Olds by Car-
borundum using a DeLaval separator. The system employs a bag-
house and a dry scrubber in combination with a wet slurry sorbent.
The wet sorbent is first atomized and then mixed with the flue
gas to combine with the sulfur dioxide. The heat of the gas
quickly dries the newly formed waste product, which is captured
with other particulates in the baghouse. Either calcium or
sodium sorbents can be used; the choice depends on site-specific
requirements. The tests with this system were conducted on gas
i s
34
with a flow from 5 to 7 m /s (10,000 to 12,000 scfm). Figure 2-
46 illustrates the Carborundum/DeLaval system.
A third set of tests at Leland Olds was undertaken by Joy
Manufacturing with the Niro Atomizer spray absorption system.
This system can be used for removal of hydrogen chloride and
other gases, as well as sulfur dioxide. The process is similar -
to conventional spray drying techniques and is outlined in Figure
2-47. It involves a feed preparation system that depends upon
the absorbent utilized, a transportation system for delivering
2-201
-------
TABLE 2-51. DRY POWDER NAHCOLITE INJECTION VERSUS
TWO-STAGE SODA ASH SYSTEM3
Fabric filter only (dry powder Nahcolite injection)
Stoichiometric
ratio
0.5
1.0
1.5
SO 2 removal, %
Spray
dryer
Fabric
filter
42
74
89
Total
42
74
89
Utilization, %
Spray
dryer
Fabric
filter
85
74
59
Total
85
74
59
Spray dryer/filter (two-stage soda ash)
Stoichiometric
ratio
0.5
1.0
1.5
SO2 removal, %
Spray
dryer
40
82
86
Fabric
filter
8
10
12
Total
48
92
98
Utilization, %
Spray
dryer
80
82
57
Fabric
filter
16d
10e
8f
Total
96
92
65
Sulfur dioxide concentration from 800 through 2800 ppm.
Temperature entering fabric filter was approximately 143°C
(290°F).
Temperature entering fabric filter was approximately 93°C
(200°F).
Actual utilization of unreacted absorbent is 80%.
Actual utilization of unreacted absorbent is 48%.
Actual utilization of unreacted absorbent is 19%.
2-202
-------
DRY SCRUBBER
FLUE GAS
149°C (399°F)
LIME SLURRY
BAGHOUSE
WASTE: WASTE:
Ca2S03, CaS04 Ca2S03,
n
CLEAN
FLUE
GASES
ID FAN
STACK
FLY ASH
Figure 2-46. Carborundum/DeLaval dry phase collection system.
2-203
-------
FLUE GAS
FROM BOILER
L.L.r\ V^
SPRAY
ABSORBER
CLEAN GAS
TO STACK
S02
ABSORBANT
SLURRY
DUST COLLECTOR
DRY PARTICULATE PRODUCT
Figure 2-47. Joy Manufacturing/Niro Atomizer dry phase
collection process.
2-204
-------
the feed from the preparation site to the atomization assembly, a
rotary or spinning wheel atomizer for absorbent atomization in
the process chamber- a gas distribution system for obtaining the
correct gas-liquid mixing within the chamber, and dust collection
equipment. These items remove sulfur dioxide and particulates
and produces a dry residue that is disposed of along with flyash
wastes without further treatment. As in normal dryer operation,
temperatures within the absorption system are closely controlled.
The outlet temperature is held enough above the dew point to
allow for a temperature drop across the dust collection equipment
without causing operational problems. As a result of the Leland
Olds project, Joy Manufacturing has received a contract from
Basin Electric Power Cooperative for a dry scrubbing system to be
installed at the Antelope Valley Station near Beulah, North
Dakota. Startup of unit 1 is scheduled for mid-1981 and will be
nominally rated at 455 MW.
A fourth pilot testing program at another Basin Electric
Station was conducted by Babcock and Wilcox. As a result of this
testing, Basin Electric has awarded Babcock and Wilcox a contract
for a dry phase sulfur dioxide and particulate removal system to
be installed at the 550-MW (net) Laramie River Station Unit 3.
Startup of this unit is scheduled for January 1982.
2-205
-------
REFERENCES FOR SECTION 2.4
1. PEDCo in-house files.
2. Laseke, B.A., Jr. EPA Utility FGD Survey: December 1977 -
January 1978. EPA-600/7-78-051a. PEDCo Environmental,
Cincinnati, Ohio. March 1978. pp. 84-87.
3. Melia, M., et al. EPA Utility FGD Survey: August-September
1978. Preliminary Report. Prepared- for the U.S.'Envion-
mental Protection Agency under Contract €8-02-2603. PEDCo
Environmental,-Cincinnati,,Ohio.. November 197-8. pp. 12,
49, 50.
4. Laseke, B.A., Jr. Utility FGD Costs: Reported and Adjusted
Costs for Operating FGD Systems. Prepared for the U.S.
Environmental Protection Agency under Contract 68-02-2603.
PEDCo Environmental, Cincinnati, Ohio. September 1978. Form 13,
5. Berube, D.T., and C.D. Grimm. Status and Performance of the
Montana Power Company's Flue Gas Desulfurization System.
In: Proceedings of the Symposium on Flue Gas Desulfurization-
Hollywood, Fl., November 1977 (Volume I). EPA-600/7-78-
058a. March 1978. pp. 277-291.
6. PEDCo in-house files.
7. Op. cit. No. 2. pp. 81-83.
8. Op. cit. No. 3. pp. 12, 48.
9. Op. cit. No. 4.
10. PEDCo in-house files.
11. Op. cit. No. 3. pp. 88-98.
12. Op. cit. No. 3. pp. 12, 13, 51-56.
13. Op. cit. No. 4.
14. LaMantia, C.R., et al. Application of Scrubbing Systems to
Low Sulfur/Alkaline Ash Coals. EPRI FP-595. Electric Power
Research Institute, Palo Alto, California. December 1977.
pp. V-9 through V-12, B-31 through B-42.
2-206
-------
15. Vaughn, C.F. Flue Gas Scrubbing with Sodium Carbonate (Wet
Scrubbers and Their Controls and SO2 Operating Experiences).
Presented at PCEA Engineering and Operating Conference, Los
Angeles, California. March 17-18, 1977.
16. Rosenberg, H.S., et al. Status of Stack Gas Control Tech-
nology. EPRI 209. Electric Power Research Institute, Palo
Alto, California. August 1975. Section 9.
17. Melia, M., et al. EPA Utility FGD Survey August-September
1978. Preliminary Report. Prepared for the U.S. Environ-
mental Protection Agency under Contract 68-02-2603. PEDCo
Environmental, Cincinnati, Ohio. November 1978. p. 9.
18. Idemura, H., T. Kanai, and H. Yanagioka. Jet Bubbling Flue
Gas Desulfurization Process. In: Proceedings of the Second
Pacific Chemical Engineering Congress (PAChEC '77). American
Institute- of Chemical Engineefs:, New York, New York. "1977.
pp. 365-370.
19. Edwards, R.A., and R.E. Rush. Operational Experience with
Three 20-MW Prototype Flue Gas Desulfurization Processes at
Gulf Power Company's Scholz Electric Generating Station.
20. PEDCo in-house files.
21. Idemura, H., and D.D. Clasen. Limestone/Gypsum Jet Bubbling
Scrubbing System. In: Proceedings of the Symposium on Flue
Gas Desulfurization-Hollywood, Fl., November 1977 (Volume
1). EPA-600/7-78-058a. March 1978. pp. 838-840.
22. Op. cit. No. 18. pp. 365-366.
23. PEDCo in-house files.
24. Op. cit. No. 18. p. 367.
25. Ibid. p. 369.
26. Op. cit. No. 17- p. 34.
27. PEDCo in-house files.
28. Estcourt, V.F., et al. Tests of a Two-Stage Combined Dry
Scrubber/SO- Absorber Using Sodium or Calcium. Presented at
the 40th Annual Meeting of the American Power Conference,
Illinois Institute of Technology, Chicago, Illinois. April
26, 1978. pp. 4-7.
29. Mcllvaine Co. Full-scale Baghouse to be Used For SO.
Removal. The Fabric Filter Newsletter, l(13):2-3, November
10, 1976.
2-207
-------
30. Op. cit. No. 28. pp. 7-8.
31. Ibid. pp. 8-10.
32. Melia, M. , et al. EPA Utility FGD Survey August-September
1978. Preliminary Report. Prepared for the U.S. Environ-
mental Portection Agency under Contract 68-02-2603. PEDCo
Environmental, Cincinnati, Ohio. November 1978. p. 15.
33. Mcllvaine Co. WP/Niro Guaranteeing 9Ci S0_ Removal with
Lime. The Fabric Filter Newsletter, 31:5, May 10, 1978.
34. Ilcllvaine Co. Interest Increasing in Dry SO? Removal. The
Fabric Filter Newsletter, 32:1-2, June 10, 1978.
35. Mcllvaine Co. Filter to be Used in Dry Process. The Fabric
Filter Newsletter, 26:3-4, December 10, 1977.
36. Mcllvaine Co. Bright Future For Spray Dryers. The Fabric
Filter Newsletter, 36:2, October 10, 1978.
37. PEDCo in-house files.
2-208
-------
SECTION 3
CONTINUOUS SO2 MONITORS
Continuous SO2 monitoring instruments are standard equipment
in domestic utility FGD installations. As many as 46 instal-
lations may have SO2 monitoring instruments, and 38 have been
positively identified as having them. Not all of the identified
units are being routinely monitored (e.g., Apache 2, Duck Creek
1, Conesville 5 and 6, Elrama 1 through 4, Phillips 1 through 6,
Petersburg 3).
Most of these units have sampling points both before and
after the FGD system (Conesville 5 and 6; La Cygne 1; Green River
1, 2, and 3; Cane Run 4 and 5; Mill Creek 3; Paddys Run 6; Sher-
burne 1 and 2; Eddystone 1A; Widows Creek 8; Shawnee 10A and 10B;
and Huntington 1). A few installations sample the stack only
(Apache 2, Winyah 2, R.D. Morrow 1, Martin Lake 1), and some
installations sample all three locations (Tombigbee 2, Petersburg
3, Milton R. Young 2, San Juan 1 and 2). Table 3-1 presents a
summary of the units known to have continuous monitoring instru-
ments and their sampling points. The SO- concentrations moni-
tored are not recorded at these units, however.
The performance of these instruments has often been affected
by such problems poor protection of probes, lack of maintenance,
lack of spare parts, and inefficient operation of mist elimina-
tors which allow plugging of outlet montior lines. These prob-
lems have occurred at Apache 2, Duck Creek 1, Conesville 5 and 6,
Elrama 1 through 4, Phillips 1 through 6, Petersburg 3, San Juan
1 and 2, Winyah 2, and Southwest 1. Nevertheless, it appears
that satisfactory instrument operation can be attained if a
preventive maintenance program is developed and implemented. Such
a program could involve: (1) routine instrumentation system in-
3-1
-------
TABLE 3-1. IDENTIFICATION OF FGD UNITS WITH CONTINUOUS SO2 MONITORS
Company
Alabama Electric
Arizona Electric Power
Arizona Public Service
Arizona Public Service
Central Illinois Light
Oolunbus and Southern
Ohio
Coluvbus and Southern
Ohio
Duquesne Light
Duquesne Light
Gulf Power
Indianapolis Power and
Light
Kansas City Power and
Light
Kansas City Power and
Light
Kansas City Power and
Light
Kansas Power and
Light
Kansas Power and
Light
Kanaa, Po«er and
Kentucky Utilities
Louisville Gas and
Electric
lousiville Gas and
Electric
Louisville Gas and
Electric
Plant and unit
Tbnbiobee 2
Apache 2
Cholla 1
Cholla 2
Duck Creek 1
Conesville 5
Conesville 6
Elraraa 1 through
6
Phillips 1 through
6
Sholz 1 and 2
Petersburg 3
Hawthorn 3
Hawthorn 4
la Cygne 1
Jeffrey 1
Lawrence 4
Lawrence 5
Green Rivers
1, 2, and 3
Cane Run 4
Cane Run 5
Mill Creek 3
Monitor
type
Photometric
Photometric
Photometric
Photometric
Photometric
Photometric
Photometric
Photometric
Photometric
Photometric
Photometric
Photometric
Photometric
Photometric
Photometric
Photometric
«„ i ..,...,1 „ j ..
UIL1UBU.1C
Photometric
Model
Unknown
Unknown
460
460
460
460
Unknown
Unknown
Unknown
Unknown
460
1819
460
460
460
460
460
460
Manufacturer
Leeds & Northrup
Environmental Products
Dupont
Dupont
Dupont
Dupont
Teak Siegler
Environmental
Instruments
Environmental
Instruments
Thermo Electron Corp-
Environmental Data Corp.
Dupont
Environmental Products
Dupont
Dupont
Dupont
Dupont
Dupont
Dupont
Sampling points and comments
Scrubber inlet and scrubber outlet; stack
Stack
FGD inlet and outlet
No longer in service; utility searching
for replacement
Inlets and outlets of scrubber modules
Inlets and outlets of scrubber nodules
No longer in service; utility seraching
for replacement
No longer in service; utility searching
for replacement
Inlets and outlets of scrubber modules;
stack
Ir lets 0^ .-cruttKH mur'-ile;
Sc:irubt>?r redd* outlet ncvsr tunctioned
properly
Inlet a-id outl''*: of scrubor nochilo
Inlets and outlets ot scrubbp.c i.oduies
Inlets and outlets of scrubber nodules
Inlets and outlets of scrubber nodules;
stack
U)
I
to
-------
TABLE 3-1. CONTINUED
Company
Louisville Gas and
Electric
Nimkota Power Go-op
Montana Power
Montana Power
Nevada Power
Nevada Power
Nevada Bower
Northern Indiana
Public Service
Northern States Power
Northern States Power
Pennsylvania Power
Pennsylvania Power
Philadelphia Electric
Public Service Company
of New Mexico
Public Service Company
of New Mexico
South Carolina Public
Service
Southern Mississippi
Electric
Springfield City
Tennessee Valley
Authority
Tennessee Valley
Authority
Plant and unit
Paddy's Run 6
Milton R. Young 2
Colstrip 1
Colstrip 2
Reid Gardner 1
Reid Gardner 2
Reid Gardner 3
Dean H. Mitchell
11
Sherbume 1
Sherbume 2
Bruce Mansfield 1
Bruce Mansfield 2
Eddystone 1A
San Juan 1
San Juan 2
Winyah 2
R.D. Morrow 1
Southwest 1
Widows Creek 8
Shawnee 1QA and 10B
Monitor
type
Photonetric
Photometric
Photometric
Photometric
Photonetric
Photonetric
Photometric
Photometric
Photometric
Photometric
Photometric
Photometric
Photometric
Photometric
Photometric
Photometric
Photometric
Model
460
Digata 1400
Series
(stack)
460
460
460
460
460
460
460
460
Unknown
Unknown
Unknown
SM800
1-2822
400
400
Manufacturer
Dupont
Bnviromental Data
Corp.
Dupont
Dupont
Dupont
Dupont
Dupont
Dupont
Dupont
Dupont
Dupont
Dupont
environmental Products
TJW Siegler
Environmental Products
Dupont
Dupont
Sampling points and Garments
Scrubber inlet and outlet
Scrubber inlet and outlets; stack
Each module inlet; each bank outlet
(before and after reheater)
Each nodule inlet; each bank outlet
(before and after reheater)
Scrubber inlet and outlet
Scrubber inlet and outlet; stack
Scrubber inlet and outlet; stack
Stack
Stack
Scrubber inlet; stack
Scrubber inlets and outlets
Scrubber lints and outlets
CO
I
u>
(continued)
-------
TABLE 3-1. CONTINUED
Ooopany
Texas Utilities
Texas Utilities
Texas Utilities
Utah Power and Light
Plant and unit
Martin Lake 1
Martin Lake 2
Monticello 3
Huntingdon 1
Monitor
type
Photometric
Photometric
Photometric
Photometric
Photometric
Model
460
463
460
463
460
Manufacturer
Dupont
Dupont
Dupont
Dupont
Dupont
Sampling points and ooranenta
Inlets of scrubber nodules
Stack
Inlets of scrubber modules
Stack
Inlet and outlet of scrubber nodules;
bypass
OJ
I
-------
spection (the frequency of which is site dependent); (2) frequent
instrument calibration; (3) maintaining an adequate stock of
vendor recommended spare parts; (4) training maintenance person-
nel to make field repairs of instruments; and (5) adaptation of a
standard system to a unique problem at the site. Similar types
of programs are in effect at Green River 1, 2, and 3; Cane Run 4
and 5; Mill Creek 3; Paddys Run 6; Sherburne 1 and 2; and Hunting-
ton 1. Table 3-2 summarizes the operating performance of instru-
ments at various installations.
Only a few installations listed in Table 3-1 are known to
record data from their continuous SO2 monitors. The small number
is due either to poor instrument performance or to recent start-
up. Apparently, none of the monitoring sytems have been certi-
fied. Most don't monitor O2 and moisture entering and leaving
the scrubber. Some don't monitor inlet and outlet SO2 concen-
trations. Table 3-3 identifies utilities that have recorded
continuous monitoring data.
3-5
-------
TABLE 3-2. SUMMARY OF THE PERFORMANCE OF CONTINUOUS S00 MONITORS
Company
Plant and unit
Startup date
U>
I
Alabama Electric
Arizona Electric Power
Arizona Public Service
Central Illinois Light
Columbus and Southern Ohio
Duquesne Light
Gulf Power
Indianapolis Power and Light
Kansas City Power and Light
Kansas Power and Light
Kentucky Utilities
Louisville Gas and Electric
Tombigbee 2
Apache 2
Choila 1
Cholla 2
Duck Creek 1
9/78
8/78
10/73
6/78
7/78
Conesville 5
Conesvilie 6
Elrama 1 through 4
Phillips 1 through 6
Sholz 1 and 2
Petersburg 3
Hawthorn 3
Hawthorn 4
LaCynge 1
Jeffrey 1
Lawrence 4
Lawrence 5
Green River 1, 2, and 3
Cane Run 4
1/77
6/78
10/75
7/73
8/78
10/77
11/72
8/72
2/73
8/78
12/68
11/71
9/75
fl/76
No particular problems have been encountered
The instrument has never worked properly because of
electronic problems; the unit was returned to
the factory for repairs
The original units have been abandoned because of
poor performance; the instruments were part of
the original control system, but because of
poor performance, they were disconnected and
.neglected; the utility is now searching for
replacements
The monitors were never maintained and have
fallen into disrepair
The instruments were never started
The original units have been abandoned; the utilty
is now searching for replacements
The original units have beer abandoned; the
utility is now searching for replacements
The stack monitor has not worked properly; parts
are now on order.
No particular problems have been encountered with
the instruments when regular maintenance is
provided
The units present no major problems when they are
regularly maintained; future placement of
sampling points in the stack will depend on regu-
latory agency requirements
(continued)
-------
to
I
Coapany
•
Minnkota Power Corp.
Montana Power
Nevada Power
Northern Indiana Public
Service
Northern States Power
Pennsylvania Power
Philadelphia Electric
Public Service Co. of New
Mexico
Plant and unit
Cane Run 5
Mill Creek 3
Padd<- "s Run
Milton R. Young 2
Colstrip 1
Colstrip 2
Reid Gardner 1
Reid Gardner 2
Reid Gardner 3
Dean H. Mitchell 1
Sherburne 1
Sherburne 2
Bruce Mansfield 1
Bruce Mansfield 2
Eddystone 1A
San Juan 1
Startup date
12/77
8/78
4/73
9/77
11/75
8/76
4/74
4/74
7/76
11/76
3/76
4/77
4/76
7/77
9/75
4/78
The units present no major problems when they
are regularly maintained; future placement
of sampling points in the stacl-s will depend
on regulatory agency requirements
The units present no major problems when they
are regularly maintained; the utility is now
planning to place sampling points in the stack
to meet regulatory agency requirements
The units present no major problems when they are
regularly maintained
Initial electronic operational problems were
encountered, however, the problems were corrected
when the instruments were modified by adding
automatic gain controls; the instruments are now
operating satisfactorily
Problems are kept to a minimum by using a technical
staff that regularly maintains the instruments;
therefore, operation is kept at a maximum
Problems are kept to a minimum by using a technical
staff that regularly maintains the instruments;
therefore, operation is kept at a maximum
The original instruments were Lear Siegler units,
which never performed satisfactory; they have
since been temporarily replaced with Dupont units,
while the utility decides what instruments to use
permanently
(continued)
-------
Company
Plant and unit
Startup date
U>
I
CO
Public Service Co. of New
Mexico
South Carolina Public
Service
Springfield Mississippi
Electric
Springfield City Utilities
San Juan 2
Winyah 2
R.D. Morrow 1
Southwest 1
7/78
7/77
8/78
4/77
Tennessee Valley Authority
Texas Utilties
Utah Power and Light
Widows Creek 8
Shawnee IDA, 10B
Martin Lake 1
Martin Lake 2
Monticello 3
Huntington 1
5/77
4/77
8/77
5/78
5/78
5/78
The original instruments were Leer Siegler units
that never performed satisfactorly; they have
since been replaced with the Dupont units
temporarily, while the utility decides what in-
struments to use on a permanent basis
The instruments have operated poorly; pluggage
problems are the most common problem; preventive
maintenace has not been a high priority
The instrument is now at the factor for updating
modifications; no problems were encountered
during the short period of operation
The instrument has never operated satisfactorily
due to electronic problems; initial pluggage
problems have been rectified by regular
cleaning; Environmental Products Corp. is now
trying to rectify the low dependability of the
instrument
The instruments have given dependable service
when properly maintained
The instruments have given dependable service
when properly maintained
The stack instruments have given dependable service.
The majority of analyzer downtime has been due to
malfunctions of the strip chart data recorder. Probe
failures have been caused by metal fatigue due to
turbulance and corrosive atmosphere downstream of
the FGD system
An initial plugging problem was encountered, but is
now kept to a minimum with regular inspection and
cleaning
-------
TABLE 3-3. COMPANIES KNOWN TO RECORD CONTINUOUS
SO2 MONITORING DATA
Company
Unit
Kentucky Utilities
Louisville Gas and Electric
Louisville Gas and Electric
Louisville Gas and Electric
Minnkota Power
Northern States Power
Public Service of New Mexico
Tennessee Valley Authority
Utah Power and Light
Green River I, 2, and 3
Cane Run 4
Cane Run 5
Mill Creek 3
Milton R. Young 2
Sherburne 1 and 2
San Juan 1 and 2
Shawnee 10A and 10B
Huntington 1
3-9
-------
SECTION 4
PERFORMANCE AND DEPENDABILITY OF FGD SYSTEMS
This section summarizes the current status of FGD technology
and highlights recent technological developments regarding sulfur
dioxide removal efficiency and system dependability. The materi-
al contained in this section supplements the material contained
in Section 4 of the previous report "Flue Gas Desulfurization
System Capabilities for Coal-Fired Steam Generators", March 1978
(EPA-600/7-78-032b) . All of the documentation in this section is
presented by reference to specific pilot, prototype, demonstra-
tion, and full-scale facilities. Information for these facilities
is presented in detail in Section 2 and Appendices A, B, and C of
this report. Much of this information was obtained directly from
the owning/operating utility companies, system suppliers, and
system designers. Additional information was obtained from
technical reports published by the system operators, suppliers,
designers, and other research firms to supplement the information
obtained through direct contacts.
Analysis of this information indicates that the design and
operating experience gained with first and second generation FGD
systems (both experimental and full-scale) has resulted in
improved design and operation of subsequent installations.
Because FGD systems that are being engineered and/or erected will
incorporate many or all of these design innovations, even better
sulfur dioxide removal and dependability can be expected without
substantial cost increase.
4.1 CURRENT TECHNOLOGICAL TRENDS
As indicated in Section 1 of this report, virtually all of
the FGD operating experience gained to date has been with the
wet-phase, nonregenerable, lime/limestone processes. As a
4-1
-------
direct result of this previous experience, the systems committed
for operation within the next 3 to 5 years also show an over-
whelming preference for lime/limestone processes.
Although there have been and still are problems associated
with lime/limestone FGD systems, most of these problems have been
identified and solved. In addition, there is a better under-
standing of methods for reducing the severity of those problems
not yet completely solved. The inherent deficiencies of using
relatively unreactive materials such as limestone or lime in the
treating of gas streams that can vary widely in composition and
quantity over short periods of time preclude the solution of some
operating difficulties encountered through system design innova-
tions alone, and a combination of good design, operator atten-
tion, maintenance, and experience is required.
4.1.1 Process Chemistry
In lime and limestone FGD systems, severe chemical problems
have been manifested in the form of scaling and corrosion. In
many instances, the presence of one of these problems stimulates
the presence or severity of the other. The phenomenon of scaling
is closely linked to sulfur dioxide removal and reagent usage.
4.1.1.1 Scaling—
Scaling refers to the uncontrolled chemical deposition of
calcium-bearing solids on the internal components of such FGD
equipment as scrubbers, mist eliminators, reheaters, nozzles,
pipes, and pumps. Three types of scale occur in lime/limestone
systems: sulfite, carbonate, and sulfate. The appearance of
either sulfite or carbonate scaling does not usually require
shutdown for cleanout. This scaling can be removed immediately
by lowering the liquor pH, which drives the calcium sulfite and
calcium carbonate back into solution.
Sulfate scale, however, is generally considered to be a more
serious problem because of the cement-like nature of calcium
sulfate (calcium sulfate dihydrate or gypsum). Once formed, this
scale generally must be removed by mechanical cleaning, which
requires shutdown. Sulfate scaling usually results from oxida-
4-2
-------
tion in the system (e.g., in the scrubber, reaction tanks, and
thickeners). Laboratory experiments, pilot testing, prototype
testing, and experience with full-scale systems have shown that
heterogeneous crystallization (the condition necessary for sul-
fate scaling) is not critical until the relative saturation level
"23
reaches approximatley 1.4 to 1.5. ' There have been several
important findings regarding sulfate scaling, and many measures
to avoid its occurrence have been or are being incorporated in
full-scale systems in service, under construction, or planned.
These measures are summarized as follows:
Two basic modes of operation have been identified for
lime/limestone FGD systems that enable scale-free operation:
coprecipitation and desupersaturation.
Coprecipitation involves the removal of calcium sulfate from
4
the system as part of a calcium sulfite/sulfate solid solution.
If a system operates so that the maximum oxidation in the slurry
circuit is 15 to 16 percent, the scrubbing liquor remains sub-
saturated with respect to calcium sulfate, and no hard scale
occurs. If the degree of oxidation exceeds the 15 to 16 percent
level, more calcium sulfate is formed in the slurry circuit than
can leave the system in a coprecipitated form. This causes the
system to operate supersaturated and to reach relative saturation
levels approaching the critical 1.4 to 1.5 values, resulting in
heterogenous crystallization and the formation of hard scale
within the system.
Desupersaturation involves the removal of calcium sulfate
from the system through the use of calcium sulfate (gypsum) seed
crystals,, which provide nucleation sites for the nonscale-forming
homogeneous precipitation of calcium sulfate. The seed crystals
are used to control sulfate scaling in a system operating in a
supersaturated mode. Crystal growth occurs on the seed crystals,
the sulfate is removed from the system as gypsum, and super-
o
saturation levels are kept below the critical 1.4 to 1.5 levels.
Experimentation and operation have shown that lime slurry
systems usually operate subsaturated with respect to sulfate
4-3
-------
through coprecipitation because of its high degree of reactivity
(compared to limestone). In most cases, lime dissolution occurs
within 20 to 30 seconds after introduction into the hold tanks.
Thus, less retention time is required for the flue gas in the
scrubber and the liquor in the hold tank to complete the reaction
with sulfur dioxide. This minimizes the amount of sulfite oxi-
dation that can occur in the system and keeps the level below the
15 to 16 percent coprecipitation requirement. Limestone systems,
however, generally operate supersaturated with respect to calcium
sulfate because limestone is less reactive than lime. Thus, more
retention time is required in the scrubber and hold tank to
complete the reaction with sulfur dioxide. This increases the
amount of sulfite oxidation that can occur in the system, re-
quiring the circulation of gypsum seed crystals to promote homo-
9
geneous crystallization for desupersaturation.
Use of magnesium additives—A number of additives have been
tested and used to improve the chemistry of lime/limestone slurry
systems with respect to scaling, sulfur dioxide removal, and
reagent utilization. Foremost among these approaches is the use
of magnesium additives to beneficiate the scrubbing slurry.
Generally, increasing the magnesium ion concentration increases
the liquid-phase alkalinity of the scrubbing slurry, increases
the amount of sulfite and sulfate the scrubbing system can hold
without exceeding solubility requirements, suppresses calcium
concentration in the scrubbing slurry, and suppresses the rate of
oxygen pickup relative to sulfur dioxide pickup from the flue gas
stream. The overall effect is a subsaturate operation with
higher removal efficiencies and high utilizations. Operating
experience with magnesium addition at the Shawnee test facility
and EPA-sponsored characterization tests at Louisville Gas and
Electric, as well as full-scale operating experience at Phillips
and Elrama stations of Duquesne Light, the Bruce Mansfield
station of Pennsylvania Power, and the Conesville station of
Columbus and Southern Ohio Electric are briefly summarized below:
4-4
-------
Shawnee - Increasing the effective magnesium ion con-
centration in the scrubbing liquor (from 0 to 9000 ppm)
strongly increased sulfur dioxide removal. At an
effective magnesium ion concentration of 2000 ppm, lime
slurry sulfur dioxide removal efficiencies increased by
15 to 20 percent (in the venturi/spray tower). At
effective magnesium ion concentrations of 0, 5000, and
9000 ppm, average sulfur dioxide removals were 77, 84,
and 94 percent, respectively (for the TCA tower).
Increasing the magnesium ion concentrations did not
always result in subsaturated operation with respect to
calcium sulfate in both lime and limestone systems.
For both lime/magnesium ion scrubbing and limestone/
magnesium ion scrubbing, gypsum scaling occurred when
the gypsum saturation of the scrubbing liquor was as
low as 80 percent, compared to about 120 percent with
no effective magnesium ion. The increase in concentra-
tion of sulfite and sulfate introduced by adding mag-
nesium despresses calcium ion concentration and can
cause unexpected gypsum scaling in the absorber.
Although the calcium ion concentration is depressed and
will be at a low value in the slurry at the scrubber
inlet, dissolution of calcium sulfite or calcium car-
bonate in the scrubber can cause a great increase in
gypsum saturation across the scrubber.
12
Paddy* s Run - The use of commercial lime with an
addition of 55 percent slurry of magnesium hydroxide
yielded an effective magnesium ion concentration of
4000 ppm. During the course of the test, the magnesium
oxide concentration was gradually lowered to 2000 ppm.
Sulfur dioxide removals of 99.7 to 99.9 percent were
achieved with inlet sulfur dioxide loadings of 2150 to
2230 ppm and outlet loadings of 1 to 5 ppm. These
removal efficiencies were accompanied by calcium sul-
fate relative saturations approaching zero. Maintain-
ing the effective magnesium ion concentration in the
2400 to 3000 ppm range provided the best control for
maintaining high sulfur dioxide removals and low cal-
cium sulfate relative saturation levels.
Phillips - Experimental work at the Duquesne Light
Phillips Power Station involving magnesium modified
lime supplied by Dravo (Thiosorbic lime: 2 to 6 per-
cent magnesium oxide) was conducted late in 1974 and
from October to December 1975. Approximately 160 MW of
flue gas was supplied to two single-stage venturi
scrubbers. One scrubber was maintained at constant
load, wheareas the other was maintained at station
cyclic load. Magnesium oxide concentrations of 6, 8,
and 10 percent in the lime corresponded to an effective
4-5
-------
magnesium ion concentration of 1900 to 4000 ppm in the
scrubbing slurry. Results of the tests indicated that
sulfur dioxide removal as high as 88 percent was
achieved with 10 percent magnesium oxide (3000 to 4000
ppm magnesium ion) and 77 percent sulfur dioxide
removal with 6 percent magnesium oxide. Previous
operatoh using pebble quickline indicated the system
was achieving sulfur dioxide removal efficiencies in
the 50 to 60 percent range. The scaling that had pre-
viously plagued the system when using high calcium
commercial lime did not occur. Low sulfate relative
saturation levels were achieved, and no hard or soft
scale accumulation was observed.
As a result of the successful test program at Phillips,
Duquesne Light modified their existing FGD systems at
Elrama and Phillips to use 7 to 8 percent magnesium
oxide lime to remove 83 percent of the sulfur dioxide
from 2.1 percent sulfur coal. Full-scale operation
using the magnesium-modified lime began at both sta-
tions in late March 1978. Compliance tests conducted
by Alleghany County in late August 1978 indicated that
Phillips was well within the allowable sulfur dioxide
emission regulations of 258 ng/J (0.6 lb/10^ Btu).
14
0 Bruce^ Mansfield - Operation of the full-scale systems
at Bruce Mansfield indicate that emission regulations
are being met or bettered using Thiosorbic lime, which
yields an effective magnesium ion concentration of 2000
ppm in the scrubbing liquor. Following optimization of
the pH-reagent feed control system, subsaturated
operation with respect to calcium sulfate was achieved
and no accumulation of scale deposits in the scrubbers
has been observed.
° Conesville - The Conesville FGD systems use Thiosorbic
lime^as the scrubbing reagent, which yields an effec-
tive magnesium ion concentration of 2000 ppm in the
scrubbing liquor to achieve high sulfur dioxide re-
movals (89.5 percent design for 5.0 percent sulfur
coal) and subsaturated operation with respect to cal-
cium sulfate.
Pullman Kellogg offers a proprietary reagent known as
catalytic limestone which contains 3 to 27 percent soluble
magnesium sulfate. Although only bench scale and pilot plant
operating experience has been obtained, results indicated that
high sulfur dioxide removals (90 to 100 percent) were achieved at
low relative calcium sulfate saturation levels.
4-6
-------
17 18
sively researched by TVA and TVA/EPA. ' This research has
concentrated on the use of benzole acid and, more recently,
adipic acid at the Shawnee test facility. These acids are
generally stronger than carbonic acid but weaker than sulfurous
acid. This has two effects: first, it assists in limestone
dissolution by lowering carbonate ion back pressure, which means
that more alkalinity in the form of bicarbonate enters the
scrubber in soluble form; and second, because sulfurous acid is
a stronger acid, the benzoate or adipate ion acts as a base in
19
the sulfur dioxide absorption step. Thus, the addition of
organic acid acids increases the total liquid phase alkalinity of
the scrubbing liquor in much the same fashion as an increase in
20
alkalinity due to magnesium ion. Intensive testing with
adipic acid was recently begun at the TVA/EPA Shawnee alkali
scrubbing test facility. In July 1978 initial test runs were
performed without adipic acid to establish base lines for both
lime and limestone scrubbing. Adipic acid testing has continued
throughout the balance of the year. Some of the major preliminary
21
findings are summarized below:
1. Lime slurry scrubbing in the two-stage prototype con-
sistently achieved 96 to 99 percent sulfur dioxide
removal with 1400 ppm of adipic acid in the venturi and
1600 ppm in the spray tower loops. The base case lime
run in this system showed 66 percent sulfur dioxide
removal.
2. Lime slurry scrubbing in the TCA prototype achieved
approximately 80 percent sulfur dioxide removal with
400 ppm of adipic acid, compared with a 67 percent
sulfur dioxide removal for the base case lime run.
3. Limestone slurry scrubbing in the two-stage prototype
consistently achieved sulfur dioxide removals higher
than 90 percent with 2100 ppm of adipic acid in the
venturi and 1500 ppm in the spray tower loops, compared
with a 57 percent sulfur dioxide removal for the base
case limestone run.
4. Limestone slurry scrubbing in the TCA prototype con-
sistently achieved sulfur dioxide removals higher than
90 percent with 750 and 1500 ppm of adipic acid, com-
pared with a 71 percent sulfur dioxide removal for the
base case limestone run.
4-7
-------
5. The preliminary results of the Shawnee test program
indicate higher removal efficiencies, higher utiliza-
tions, and lower scaling potential, which substantially
agrees with initial expectations. Specifically, it has
been most effective when used in conjunction with
forced oxidation and upstream particulate precollection
(fly ash removal by an ESP). One negative aspect, how-
ever has been the unexpectedly high deterioration or
decomposition of adipic acid that takes place in the
scrubber. Actual feed ra+-.es of adipic acid were two to
three times higher than could be accounted for in the
system discharge sludge.
Design changes—A number of process design innovations have
been developed to eliminate scaling and increase sulfur dioxide
removal efficiency and reagent utilization. Forced oxidation is
one technique that has been successfully piloted and used in
commercial installations. Forced oxidation has been pilot-plant
tested at EPA-IERL/RTP and Shawnee. A number of demonstration
and full-scale installations have also used forced oxidation,
including Sherburne 1 and 2 (Northern States Power), Lawrence 4
and 5 (Kansas Power and Light), Widows Creek 8 (Tennessee Valley
Authority), Cholla 1 (Arizonia Public Services),* and Will County
1 (Commonwealth Edison).* The use of forced oxidation has been
considered a major factor in achieving high sulfur dioxide re-
movals while enabling the systems to run calcium sulfate scale-
free.
Forcibly oxidizing all the sulfite to sulfate and desuper-
saturating by circulating seed crystals has been especially suc-
cessful in controlling scale in limestone systems treating flue
gas with low-sulfur dioxide loadings (Sherburne 1 and 2, and
Lawrence 4 and 5). This occurs because the lower the inlet
sulfur dioxide concentration, the higher the oxidation frac-
22 23 24
tion. ' ' Thus, a lesser degree of forced oxidation is re-
quired to drive all the sulfite to sulfate in order to insure
efficient desupersaturation and sulfate scale-free operation.
Systems treating gas from low-sulfur western coal can operate at
25
natural oxidation levels in excess of 90 percent. Pilot studies
Short-term experimental basis.
4-0
-------
show that 600 ppm loadings yielded 70 percent oxidation, whereas
26
2500 ppm loadings resulted in 12 percent oxidation.
Operating changes—Several operating parameters have an
important impact on scaling.
1. Control of pH. Excluding all other factors, the differ-
ence in optimum operating pH affects the performance of
lime/limestone slurry systems by introducing two
problems: operation at low pH promotes the formation
of hard calcium sulfate scale, and operation at high pH
promotes the formation of soft calcium sulfite scale.
Hard scale forms as a function of slurry pH because the
solubility of calcium sulfite increases dramatically as
pH decreases, and the calcium sulfate concentration
decreases slightly as pH decreases. Thus, calcium
sulfate is more likely to precipitate out as a hard
scale at lower pH. To avoid hard scale formation, the
slurry pH should be maintained at high levels. The
solubility of calcium sulfite decreases rapidly as pH
increases, however, promoting the formation of soft
27
calcium sulfite scale. Therefore, optimum pH levels
for lime and limestone are generally maintained in the
8.0 to 8.5 and 5.5 to 6.0 ranges, respectively. La Cygne
and Bruce Mansfield are examples of full-scale systems
where equipment limitations (pH sensor type and measure-
ment location) prevented accurate pH measurements.
Hence, frequent losses of chemical control were en-
countered, resulting in periodic massive scaling.
Resolution of these pH-related equipment limitations
resulted in immediate improvement in reagent utiliza-
tion and sulfur dioxide removal and virtual elimination
28 29
of any scaling. '
2. Solids Level. If all other variables are held con-
stant, increases in slurry solids levels increase the
amount of seed crystal area available for homogeneous
crystallization. This is especially true for systems
4-9
-------
that control sulfate levels by desupersaturation. For
these systems, an optimum amount of slurry seed crystals
is maintained in the system by maintaining an optimum
level of slurry solids. Thus, when the solids level
drops, the seed crystal level drops correspondingly,
resulting in the impairment or loss of homogeneous
crystallization, the onset of heterogeneous crystalliza-
tion, and subsequent scale development. Examples of
full-scale systems that utilize this method of control
are Sherburne 1 and 2 and Lawrence 4 and 5. These
systems desupersaturate by converting all the sulfite
to sulfate (with the aid of forced oxidation); and
precipitating the calcium sulfate with the aid of
gypsum seed crystals. These systems have also en-
countered minor episodes of sulfate scaling, which in
every case was due to dilution of the slurry solids
level after the mist eliminator wash water rate was
increased to improve cleanliness. The increased levels
of makeup water in the system decreased the slurry
solids level and the seed crystal level, impairing
desupersaturation and resulting in scale formation in
the systems. Reestablishment of slurry solids levels
31
prevented further episodes of scaling.
Liquid-to-gas Ratio. If all other variables are held
constant, increasing the L/G reduces the sulfur dioxide
pickup per volume of scrubbing liquor. Thus, the
relative saturation increase in the scrubber liquor can
be reduced. Higher L/G's also reduce the requirement
for alkaline solids dissolution because more liquid-
phase alkalinity will be available. Therefore, the
requirement for calcium dissolution and its resulting
effect upon calcium sulfate relative saturation level
32
can be reduced by using high L/G's.
4-10
-------
4.1.1.2 Summary of Scaling Problems—
In summary problems connected with scaling, sulfur dioxide
removal, and reagent utilization with respect to lime/limestone
slurry FGD systems are essentially limited to the deposition of
calcium sulfate solids (gypsum) in the slurry circuit as a result
of oxidation in the scrubber vessels, reaction tanks, and thick-
ener. This scaling, which affects system availability, can be
described in terms of relative saturation levels. Critical
relative saturation levels are about 1.4 to 1.5. At this level,
heterogeneous crystallization and uncontrolled depostion of
gypsum on equipment internals can occur. Several important
considerations and findings have been investigated in recent
years, and corrective measures have been developed that have
minimized or virtually eliminated gypsum scaling as a major
operational problem. Foremost among these recent developments
are two basic modes of chemical operation that can maintain a
scale-free environment: coprecipitation and desupersaturation.
Coprecipitation effectively removes calcium sulfate from the
system as part of a calcium sulfite/sulfate solid solution. This
can be achieved when the maximum oxidation in the slurry circuit
is approximately 15 to 16 percent. Desupersaturation effectively
removes calcium sulfate from the system through the use of cal-
cium sulfate seed crystals, which provide nucleation for the non-
scale-forming homogenous precipitation of calcium sulfate. The
identification of these two basic chemical modes of operation and
the tendency of lime and limestone slurry systems to operate in
either one or the other has given rise to a number of techniques
that improve their capability to operate scale-free while achiev-
ing higher sulfur dioxide removals and higher reagent utiliza-
tions. Such techniques involve the use of additives such as
magnesium and organic acids, which improve reagent reactivity,
reagent utilization, and sulfur dioxide removal, and enhance the
potential for scale-free operation. The use of additives in
commercial applications has been confined to beneficiation with
magnesium. Experimental work at Shawnee with magnesium and
4-11
-------
adipic acid has also showed correspondingly improved operation
for both lime and limestone slurry systems. The use of forced
oxidation for limestone slurry systems that scrub flue gas
resulting from the burning of low-sulfur western coal flue gas
has been very successful in maintaining high sulfur dioxide
removals and scale-free operation. In addition to the chemical
and process design scale-control methods, a number of operating
parameters such as pHf slurry solidb, and L/G have been deter-
mined to have a pronounced impact on scaling, utilization, and
sulfur dioxide removal. Factorial analyses have demonstrated
that optimum operating parameters for lime and limestone slurry
systems can prevent or correct the incidence of scaling while
maintaining high utilizations and sulfur dioxide removals. To
summarize, sufficient information is now available so that good
engineering design can now insure when the use of lime or lime-
stone is more appropriate and what chemical and physical measures
and operating parameters are necessary to achieve high sulfur
dioxide removals and scale-free operation for both high and low
sulfur coals.
4.1.1.3 Corrosion—
In simple terms, corrosion is the dissolving or wearing away
of the metal surfaces of FGD equipment. Although corrosion is a
chemical phenomenon, this phenomenon cannot be analyzed in a
strictly chemical fashion as was scaling. Analysis of corrosion
involves consideration of design, construction, and fabrication
(see Process Design, Section 4.1.2). Nevertheless, the phenom-
enon of corrosion with respect to lime/limestone slurry FGD
systems and design and operating measures taken to minimize or
eliminate corrosion problems are discussed briefly in the balance
of this section.
Two basic corrosive agents are present that have very ser-
ious consequences from the standpoint of FGD system performance:
(1) acids containing sulfur in the form of sulfurous and sulfuric
acids and (2) soluble mineral matter in the form of chloride.
These two basic types of corrosive agents give rise to a number
4-12
-------
of specific types of corrosion: general corrosion, pitting,
crevice corrosion, intergranular corrosion, stress-corrosion
cracking, and erosion-corrosion.
A number of design, construction, fabrication, and operation
measures have been developed that successfully minimize the rate
of corrosion or prevent its appearance altogether. These mea-
sures are summarized briefly below.
A selective process design approach has been developed by
the major system suppliers that allows highly corrosive envi-
ronments to be isolated to discrete areas of the FGD system.
Such an approach involves the segregation of the scrubbing loop
into separate multiple loops so that a different set of chemical
conditions is maintained for quenching or prescrubbing, absorp-
tion, and mist eliminator washing (wash trays). In such designs,
the quencher or prescrubber bears the first full brunt of the
incoming hot flue gas. It encounters the total chloride content
from the fuel fired without dilution, and it also is the area in
which low-pH, chloride-laden, return water is used freely in
lime/limestone slurry systems. Isolation of such a corrosive
environment to the quencher or prescrubber is advantageous
because this area is small enough and discrete enough that it can
be constructed of chloride-resistant materials without drasti-
cally increasing costs. Alloys that have been tested and spec-
ified for full-scale systems are listed in ascending order of
molybelmum content, pitting resistance, and cost: 317L stainless
steel, Incoloy Alloy 825, Hastelloy G, Inconel Alloy 625, and
Hastelloy C-276. Systems operating at Cholla, Duck Creek, Reid
Gardner, Conesville, Petersburg, Southwest, and Martin Lake (to
mention a few) have used these materials successfully in small
amounts in the quench or prescrubbing area.
In many of the initial lime/limestone slurry systems, the
incoming hot gas contacted the reactive absorbant suspension,
resulting in the accumulation of solids at the wet/dry interface.
These deposits provided convenient sites for the accumulation of
chloride at concentrations approaching 50,000 ppm. The result
4-13
-------
was severe episodes of pitting, stress corrosions, crevice
34
corrosion, stress-corrosion cracking, and errosion-corrosion.
This problem area has been largely overcome by better control of
process chemistry, the use of self-cleaning devices, the highly
selective use of superior construction materials, and the use of
multiple-loop designs. Better control of process chemistry
eliminates the formation of corrosion in the system and thus
prevents the formation of chloride Ion host sites. A number of
systems are equipped with soot blowers in the approach ducts to
the scrubber modules so that the inevitable buildup of solids at
the wet/dry interface can be cleaned automatically and periodi-
cally. The Lawrence and Sherburne systems have such self-clean-
ing devices. '
Most suppliers now prefer 316L stainless steel for the
scrubber modules. This material has demonstrated superior
resistance of corrosion, errosion, and scale development compared
with carbon steel, 304 stainless steel, and 304L stainless
steel. The preference for 316 and 316L stainless steel is based
primarily on the smooth mating surfaces and molybdenum content of
these steels. The former attribute minimizes the presence of
crevices that provide convenient sites for buildup of soluble
mineral materials (e.g., chloride). The molybdenum content (2.50
to 2.75 percent minimum) of stainless steel increases corrosion
resistance to sulfurous and sulfuric acids and also increases
resistance to localized attack such as pitting and crevice cor-
rosion. One utility specifies 317 low-carbon stainless steel,
which has a minimum molybdenum content of 3 to 4 percent (Ten-
37
nessee Valley Authority, Widows Creek 7). Full-scale systems
that have successfully used 316 and 316L stainless steel as the
primary material of construction for the scrubber modules are
Cholla, La Cygne, Lawrence, Will County, and St. Clair.38'39'40'41'4
Mist eliminators have been very susceptible to corrosion.
Most suppliers now recommend the use of fiberglass-reinforced
plastics, polypropylene, and corrugated plastics over stainless
steels and other alloys in mist eliminators because these ma-
terials are relatively lightweight, inexpensive, and superior in
4-14
-------
43
resistance to corrosion. Reheaters have been especially
susceptible to corrosion, and the trend is toward indirect hot-
air reheat because in-line reheat systems have been subject to
corrosion and plugging in the tubes. The corrosion in some cases
has been so severe that even the heartier alloys have been
unsatisfactory under many operating conditions.
Particulate precollection using an upstream electrostatic
precipitator (ESP) or fabric filter (FF) minimizes a number of
chemical-related problems associated with simultaneous or two-
stage wet scrubbing systems. Most notably, these include mini-
mizing the potential of corrosion attack and scale formation
within the FGD system. In this latter regard, some metallic
components in the fly ash have been indentified as catalytic oxi-
dation sites for the promotion of sulfite to sulfate oxidation in
the scrubbing slurry, thus resulting in sulfate saturation in the
scrubbing slurry and uncontrolled gypsum scale formation.
Problems associated with erosion-corrosion are reduced because
the slurry is not as abrasive as it would be if it contained
collected fly ash. Two-stage scrubbing systems, which collect
fly ash as well as sulfur dioxide, often have slurry solids level
as high as 18 to 20 percent (e.g., La Cygne). Hence, erosion-
corrosion problems and the materials required to minimize these
45
problems are accentuated.
If reliable particulate precollection in the form of ESP's
or FF's is provided, the balanced-draft fan or FGD booster fan
may be placed upstream of the FGD system since the erosive nature
of hot, fly-ash-laden flue gas will have been largely eliminated.
Such a design improves overall system dependability. Many of the
fans on the initial systems operated on cool, saturated gas,
which resulted in acid attack, corrosion, solids deposition
erosion, and scale development on the fan blades and housing, and
often led to forced outages.
4.1.1.4 Summary of Corrosion Problems—
In .summary, problems connected with corrosion in lime/lime-
stone slurry scrubbing systems can be limited to the action of
4-15
-------
the basic corrosive agents present in FGD systems: sulfur-con-
taining acids in the form of sulfurous and sulfuric acid and
soluble mineral matter in the form of chloride. The specific
types of corrosion that have been encountered in scrubbing
systems are general corrosion, pitting, crevice corrosion, inter-
granular corrosion, stress-corrosion cracking, and erosion-
corrosion. A number of design, construction, fabrication, and
operation measures have been developed to decrease the rate of
corrosion or prevent its appearance altogether. System suppliers
and designers have developed a selective process design approach
that relegates highly corrosive environments to discrete areas of
the FGD system. These discrete areas (such as prescrubbers or
quenchers) are small enough so that they can be constructed of
exotic construction materials without having a severe impact on
the overall capital cost of the system. Other design, construc-
tion, and operating procedures have also been adopted to minimize
corrosion, and sufficient data are now available for a better
understanding of corrosion problems in FGD systems. Procedures
have been developed to insure that corrosion does not seriously
affect removal efficiency or dependability.
4.1.2 Process Design
Several advances in the process design of lime/limestone
slurry FGD systems have improved system dependability and sulfur
dioxide removal. These advances and methods developed to reduce
problems with major FGD equipment are described below.
47
4.1.2.1 Emission Control Strategy—
In recent years the trend in design of particulate and
sulfur dioxide emission control systems is toward combined ESP/FGD
or FF/FGD strategies over simultaneous or two-stage wet scrubbing
strategies. This preference in emission control strategy design
is due to the high reliability afforded by ESP's and FF's, which
enables selective bypass of scrubber modules without having to
reduce the load or shut the unit down. Other benefits include:
The potential for corrosion at wet/dry interfaces and
erosion-corrosion in the FGD systems is minimized.
4-16
-------
0 The potential for scale formation and plugging in the
FGD system is minimized.
0 Exotic construction materials can be used more select-
ively and in less amounts.
0 Balanced-draft and booster fans can precede rather than
follow the FGD system.
0 Sludge blending and stabilization processes which use
dry fly ash as an additive are permitted.
4.1.2.2 Equipment Design Improvements—
Specific design and operating improvements for FGD-related
equipment are as follows:
Balanced-draft or booster fans —In addition to placement
of these fans upstream of the FGD system, another development is
the use of variable-pitch, axial flow fans. The main advantage
of this design is its consistently higher efficiency (versus
centrifugal fans) over the entire boiler operating range, which
results in a substantial power savings. Other advantages are
superior flow control, arrangement flexibility, easy access and
maintenance, less severe construction requirements, and increased
design reliability.
Dampers—Bypass and isolation dampers are used to regulate
the flow of flue gas into and around the FGD system. The primary
purpose of isolation is to continue unit operation while the
scrubber modules are under maintenance. Efficient and reliable
dampers allow maintenance crews to service the modules in an
efficient and timely manner. Common designs include slide-gate
(guillotine), single-blade butterfly, and multiblade parallel
49
(louver) dampers. Corrosion and erosion of various types of
dampers and damper seals have been common. In some cases, dampers
have failed or been so inefficient that the modules could not be
maintained during bypass situations. Because of such problems,
TVA considered removing or welding open all the guillotine
dampers on their Widows Creek scrubbing system. The current
trend is toward two-stage louver dampers having a pressurized
seal-air system that maintains a positive pressure between them.
4-17
-------
Pressurized seal air increases the parasitic energy demand of the
system because of increased fan power requirements/ but it con-
tributes significantly to successful damper operation.
Scrubbers—Several recent design innovations increased
dependability and removal efficiency of scrubbers. Cooling the
gas to its adiabatic saturation temperature prior to contact with
the scrubbing slurry increases sulfur dioxide removal capability
and minimizes the potential for scaling and corrosion at the
slurry/gas interface area. Presaturators or quenchers were not
incorporated into the design of many of the initial FGD systems,
partly because they had venturi scrubbers and packed-bed scrubbers
incorporated in them for first-stage or simultaneous removal of
particulate in the FGD system. In such systems, the incoming,
hot, pollutant-laden flue gas contacted the suspended reactive
absorbant and resulted in solids accumulation and subsequent
corrosion. For this reason, presaturators or quenchers that use
clear liquor or spent slurry for the absorbing stage (multiple
slurry loops) are now used (or plans call for their use) in
52
systems that include dry-phase particulate precollection.
The trend in lime/limestone slurry FGD systems is away from
venturi and packed-bed designs to spray towers and combination
towers. The venturi design has been abandoned largely because
the small liquid/gas contact time resulted in relatively low
sulfur dioxide absorption. Scaling, plugging, and corrosion of
internals have occurred frequently in packed-bed designs (fixed
and mobile) and tray towers. Spray towers, on the other hand,
have few internal components in the gas liquid contact zone, and
therefore offer the potential for greater dependability because
there are fewer sites for deposition of solids in the form of
54
scale, collected fly ash, and unused reagent. A number of past
and current operations have or are utilizing spray tower designs,
including Four Corners (demonstration, terminated), Mohave
(demonstration, terminated), St. Clair (demonstration, termi-
nated) , Highgrove (prototype, terminated), Shawnee (prototype,
operational), Lawrence (full-scale, operational), M. R. Young
4-18
-------
(full-scale, operational), and Sherburne (prototype,* opera-
tional). To date spray tower operation has been very successful.
High dependability and sulfur dioxide removal has been reported
on almost all the above-mentioned systems; however, several
limitations have been encountered. Mass transfer limitations and
cost tend to restrict spray tower design applications so that
only low- and medium-sulfur coal can be used in conventional
lime/limestone slurry systems.
The greater tendency for slurry carryover in spray towers
requires either increased tower height or special mist eliminator
designs (wash trays, bulk entrainment separators), both of which
increase capital and annual cost requirements. These limita-
tions have given rise to the development of combination towers.
These towers combine the features of venturi, packed, tray, and
spray towers into one module. Examples of combination towers are
Research Cottrell's spray/packed tower, Combustion Equipment
Associates' venturi/spray tower, and Universal Oil Products'
tray/packed tower. These designs offer greater flexibility in
that operating conditions can be segregated into discrete areas
of the scrubber, allowing separate chemical and physical condi-
tions to be maintained. This permits the use of the double-loop
slurry concept in which low-pH liquor contacts the entering flue
gas in an initial scrubbing stage, where some sulfur dioxide
removal takes place and unused reagent is consumed. High-pH
liquor is contacted with the gas in the second scrubbing stage,
where the bulk sulfur dioxide removal takes place. Spent slurry
from this stage is discharged to the first stage. Fresh makeup
*
One of the 12 scrubber modules installed on Sherburne 1 has
been converted to a spray tower as a test to determine design
and operating parameters for Sherburne 3 and 4, two full-scale
units for which spray tower absorbers are planned.
High-sulfur-coal spray tower applications in service at Cane
Run and scheduled for Bruce Mansfield and Thomas Hill use
special reagents (carbide lime, magnesium-promoted lime, and
limestone) to compensate chemically for mass transfer limita-
tions.
4-19
-------
reagent is added only to the second stage. This type of design
takes advantage of the concept used for developing the Weir
scrubber where the highest gas sulfur dioxide concentration
(inlet) contacts the lowest liquor alkalinity and the highest
liquor alkalinity contacts the lowest sulfur dioxide concentra-
tion (outlet). ' Performance has verified the potentially
hig'h removals and utilizations afforded by such designs. Borgwardt
has reported 97 percent limestone utilization and 81 percent
sulfur dioxide removal in a limestone-scrub Ing, forced-oxidation
58
pilot plant, which is of the double-loop, two-stage design.
Texas Utilities has reported 99 percent sulfur dioxide removal
and high utilization in the spray/packed tower limestone slurry
59
scrubbers at Martin Lake. Montana Power has reported higher
than design sulfur dioxide removal and reagent utilization in the
alkaline fly-ash, lime-slurry scrubbers at Colstrip, which are of
the venturi/spray tower design.
Reaction tanks—Coinciding with gas-side staging is liquid-
side staging, in which hold tanks are arranged in series to
simulate plug flow reactor designs. (A plug flow design is one
that allows the reacting liquor to flow through the reactor
without the occurrence of any backmixing. A plug flow situation
can be simulated by arranging agitated tanks in series.) This
design concept was originally piloted by Borgwardt at IERL-RTP
and further tested at Shawnee, where sulfur dioxide removal was
increased from 70 to 79 percent by changing the reaction tank
design from one agitated tank to three agitated tanks of equal
6°
volume in series for limestone slurry scrubbing. *" Full-scale
applications that incorporate liquid-side staging include Law-
rence 4 (supplied by Combustion Engineering), Cholla 1, and
Martin Lake 1 (both supplied by Research Cottrell).63'64/65 All
these applications are low-sulfur, limestone-slurry systems.
Sulfur dioxide removals and dependabilities greater than 90
percent have been reported.
66 67
Mist elimination ' —Chevron and baffle-type mist elimina-
tors still continue to be the exclusive designs used in U.S.
4-20
-------
utility FGD systems. A number of different designs have been
tested, including wire-mesh/ tube-tank, gull-wing, ESP, and
radial-vane. The performance and economics associated with these
and other design alternatives indicate that emphasis on the use
of baffle and chevron types will continue. The popularity of
these separators is due primarily to design simplicity, adequate
collection efficiency for medium to large size drops, relatively
low pressure drop, wide-open construction, easy access for main-
tenance, design flexibility, and relatively low cost. Within
these two preferred types of mist eliminators, a number of
specific design, construction, and operation improvements have
been implemented.
1. Chevron designs of continuous-vane construction predom-
inate over noncontinuous-vane construction because of
greater strength and lower cost.
2. Multiple-stage designs predominate over single-stage
designs. This tendency is process-sensitive in that
limestone systems, which predominate in the United
States, generally require two or three stages for
effective mist entrainment separation. La Cygne -and
Lawrence are examples of limestone systems that employ
multiple stages, including precollectors.
3. Single-stage designs are successful for lime systems
because of the superior reactivity of lime and the cor-
respondingly higher utilization.
4. The number of passes per stage also tends to be process-
sensitive in that four-pass designs are generally used
for lime and three-pass designs for limestone. More
passes are required for lime systems because of the
single-stage design, less passes are required for
limestone because of the multiple-stage design.
5. Fiberglass-reinforced plastics, polypropylene, and
corrugated plastics are now used in almost every
operational system and specified for use in almost
every planned system. These materials are preferred
because they are relatively lightweight, inexpensive,
and superior in resistance in corrosion. Potential
problems associated with high temperature have been
minimized by specifying materials that can withstand
exposure to 205°C (400°F).
4-21
-------
6. Vane spacings of 3.8 to 7.6 cm (1.5 to 3.0 in.) are
generally used in single or first stages and 2.2 to 2.5
cm (0.9 to 1.0 in.) for second stages.77 Multiple
staging permits the use of finer spacings, which pro-
vide increased mist-separation capability for smaller
particles.
7. The horizontal configuration (vertical gas flow) is
still widely used because of adequate performance (to
date), operational and design simplicity, and lower
capital cost. The vertical configuration offers a
number of advantages over the horizontal configuration.
For example, reentrainment due to the gas flow opposing
the path of the drainage is eliminated, and limitations
on wash water quality and quantity (as well as wash
direction) are eliminated. Two systems (Widows Creek
and Duck Creek) recently started operations using
vertical configurations. Initial performance results
indicate adequate operation performance and no major
operating problems to date.
8. Special features, such as hooks and pockets, are
desirable for prevention of reentrainment.
9. Bulk separation devices, impingement plates, single
baffle deflectors, and gas direction changes are be-
coming integral parts of mist eliminators because of
increased removal efficiency and design flexibility.
10. Wash and knock-out trays have been incorporated into a
number of mist eliminators to conserve fresh water and
increase (or extend) the quantity of water available
for washing.
11. Wash systems utilizing blended water consisting of pond
return water or thickener overflow and fresh water are
used over other strategies (total return or total
makeup). Intermittent high-pressure, high-velocity
wash systems are preferred to continuous wash systems
because of impact on water balance and chemistry.
12. Optimum distances between stages are generally 1 to 1.5
m (4 to 5 ft), and freeboard distances are 1 to 1.5 m
(4 to 5 ft). The former is the minimum permitting easy
access for maintenance. The latter is the distance at
which carryover can be minimized without drastically
increasing tower height and pressure drop.
13. Superior overall operation is obtained when fly ash is
collected prior to the scrubbing system. This is
because nonfly-ash scrubbing systems usually have a low
slurry solids content, and the lower the slurry solids
4-22
-------
content, the less likely the tendency for mist elimi-
nator fouling.
Reheaters—A pronounced preference for stack gas reheat
versus no reheat (wet stack) is still evident for those systems
68
in service and committed for future operation. The wet stack
systems at Cane Run and Green River have been abandoned in favor
of reheat because of corrosion, plume dispersion, and plume
visibility problems. ' Stack liner repairs have also been re-
quired at Duck Creek. Reheater malfunction at Bruce Mansfield
contributed to stack liner failures and acid rainout problems
72
encountered at that station. These problems are more pro-
nounced at high-sulfur coal applications because of the higher
sulfur dioxide loadings. Concerning reheater design and con-
73 74
struction, a number of developments have been made: '
1. Among the systems that have operated or are currently
in service, in-line reheat has proved to be the most
popular strategy.
2. The trend in reheat strategies, as evidenced by FGD
systems scheduled for immediate and future operation,
is away from in-line and direct combustion methods and
toward indirect hot air reheat. The rationale for this
trend is largely dependent on the problems encountered
in the former and the necessity for oil or natural gas
in the latter. In-line reheat systems have been sub-
ject to corrosion and plugging in the tubes. The
corrosion in many cases has been so severe that even
the heartier alloys have been unsatisfactory under many
operating conditions. Many of these problems have been
attributed to upstream mist eliminator inefficiency and
inadequate self-cleaning techniques (soot blowers).
3. A number of the major system suppliers still recommend
in-line reheaters, especially when minimization of the
parasitic energy demand is in order. It has been
determined that corrosion of high-alloy materials is
attributed to stress corrosion caused by chloride,
whereas carbon steel is more susceptible to acid corro-
sion caused by high sulfur dioxide concentrations.
Therefore, if low sulfur/low chloride, low sulfur/high
chloride, or high sulfur/low chloride environments can
be accurately identified, in-line reheaters may be used
successfully.
4-23
-------
4. Indirect hot air reheat has the undesirable effect of
increasing the parasitic energy demand of the FGD
system as well as the overall capital cost.
5. Bypass reheat is used for low-sulfur coal FGD applica-
tions when the required degree of reheat is not seri-
ously constrained by emission standards.,
6. The use of efficient mist eliminators reduces the load
on the reheat system by removing water droplets from
the flue gas stream.
7. ATfs of 14° to 28°°C (25° to 50°F) adequately prevent
downstream water condensation.
Solids separation —The major development in this area
involves the increased emphasis placed on the mechanical separa-
tion techniques of clarification, centrifugation, and vacuum
filtration, and decreased emphasis on interim ponding. Formerly,
an interim pond was relied on to fulfill three functions: clar-
ification, dewatering, and temporary or final sludge storage.
The realization that a single pond cannot perform all three has
spurred the development of the other techniques. Furthermore,
increasing emphasis on off-site disposal for landfill and struc-
tural fills, plus increased emphasis on attaining closed water-
loop operations, has also stimulated the use of clarifiers,
centrifuges, and vacuum filters. In addition to these techni-
ques, several installations are experimenting with or using
forced-oxidation strategies to enhance solids settling properties,
to improve process chemistry, and to decrease sludge disposal and
land requirements.
Process control and instrumentation —Because of the complex
nature of lime/limestone scrubbing chemistry, which has been the
primary source of operating problems in full-scale systems, pro-
cess control is considered a crucial item. The following is a
brief review of some of the essential findings and innovations in
the development of process control technology:
0 Four process variables have been identified as critical
from the standpoint of process control: flue gas flow,
slurry pH, slurry solids, and liquid level.
4-24
-------
0 Virtually all of the operating full-scale systems reg-
ulate reagent feed rate by controlling slurry pH. This
works by using a pH sensor to provide a control signal
for modulating the flow of reagent to the FGD system in
a feedback control mode. The pH signal regulates the
position of control valves for controlling the rate of
reagent feed.
0 The major problems encountered with pH control systems/
include sensor plugging, calibration drift, breakage,
false indication, improper location of sensor, and
erosion/corrosion damage.
0 Sufficient operating experience has been obtained so
that most of the reagent feed control problems have
been identified. Once identified, these problems have
been resolved through various design modifications
and/or new operating and maintenance procedures.
0 Concerning selection of hardware, for pH control it has
been noted that dip-type sensors are more successful
than in-line sensors because they are easier to clean
and calibrate. In-line, flow-through sensors are
generally subject to more wear and abrasion and gener-
ally require more frequent maintenance.
0 Other reagent feed control systems have been or are
being evaluated on full-scale systems. One type
involves feed-forward reagent control on inlet flue gas
flow rate and sulfur dioxide concentration with trim
provided by slurry pH.
Another type involves control of reagent feed rate
using the outlet sulfur dioxide as the control vari-
able. Limited success has been reported for both of
these systems, primarily because of the difficulty in
obtaining accurate and consistent readings from sulfur
dioxide gas analyzers. This has been especially diffi-
cult on high-sulfur coal applications.
4.1.2.3 Summary of Process Design—
Process design innovations on lime/limestone slurry scrub-
bing systems involve emission control strategy and specific
equipment design improvements.
With respect to the approach to emission control, ESP/FGD
and FF/FGD strategies predominate over simultaneous or two-stage
scrubbing strategies for particulate and sulfur dioxide removal.
Dry-phase particulate precollection offers a number of benefits,
4-25
-------
including the following: the scrubber module can be bypassed
during maintenance or forced-outage situations without having to
shut down or reduce the load; potential for corrosion, erosion,
and scaling in the system is minimized; exotic construction
materials for the FGD system can be used more selectively and in
less amounts; fans can precede rather than follow the FGD system;
and sludge blending and stabilization processes using dry fly-ash
additives are permitted.
With respect to equipment design improvements, several
specific design and operating improvements have been made in a
number of major FGD-related equipment areas. Briefly, these
include the use of variable-pitch axial-flow fans, pressurized
seal-air louver dampers, spray and combination scrubber towers,
liquid-side staging, multiple staging and vertical mist elimina-
tor arrangements, indirect hot-air reheaters, mechanical solids/
liquid separation techniques, automatic pH-reagent feed control
systems, 316L stainless steel construction materials for the
scrubbers, and CXL-2000 lining for the stack flues. Analysis of
the general and specific process design trends indicates that
systems are designed to incorporate increased flexibility and
reliability. Specifically, the tendency is toward sparing of
modules and ancillary components and designing less interdepen-
dent systems (i.e., systems in which major unit operations are
not strongly affected by upstream or downstream performance). An
example of this latter design trend is the removal of the re-
heater from direct contact with the flue gas stream (in-line),
thus making it less sensitive to mist eliminator performance.
Also, many of these systems are installed on new, base-loaded
units that are designed to fire coal from specific sources. This
results in a flue gas with more constant and stable character-
isitcs. This aids FGD system dependability because the system
does not have to respond to a dramatic variation in flue gas flow
rate and composition. In many of the original FGD applications,
the systems were required to operate on widely varying loads
(cycling and peak), and varying coal types (low-sulfur western,
4-26
-------
high-sulfur eastern, and blends), situations that often required
them to respond to conditions beyond their process control capa-
bility. Consequently, variations in the reagent feed rate, loss
of chemical control, and the incidence of chemical and mechanical
problems resulted in numerous forced outages.
4-27
-------
REFERENCES FOR SECTION 4.1
1. Cheremismoff, P.N., and Fellman, R.T. Optimizing SO2
Scrubbing Processes. Power Engineering. October 1974.
pp. 54-56.
2. Corbett, W.E., Hargrove, O.W., and Merill, R.S. A Sum-
mary of the Effects of Important Chemical Variables Upon
the Performance of Lime/Limestone Wet Scrubbing System.
Prepared by Radian Corporation for EPRI. FP-639, Research
Project 630-3. December 1977. p. 3-4.
3. Gogineni, M.R., and Maurin, P.G., Sulfur Oxides Removal
by Wet Scrubbing - Application to Utility Boilers. Pre-
sented at Frontiers of Power Technology Conference, Still-
water Okalhoma, October 1-2, 1975. p. 4.
4. Op. cit. No. 2. p. 3-9.
5. Op. cit. No. 2 pp. 3-11, 3-12.
6. Op. cit. No. 3. p. 4.
7. Op. cit. No. 2. p. 3-13.
8. Op. cit. No. 3. p. 4.
9. Op. cit. No. 2. pp. 2-9, 2-10, and 3-11.
10. Op. cit. No. 2. pp. 2-14, 2-15, 2-16, and 3-8.
11. Head, H.N. EPA Alkali Scrubbing Test Facility: Advanced
Program, Third Progress Report. EPA-600/7-77-105. Bechtel
Corporation, San Francisco, California. September 1977.
pp. 1-5, 1-7, and 5-25.
12. Van Ness, R.P. Louisville Gas and Electric Company Scrub-
ber Experiences and Plans. Presented at the Flue Gas De-
sulfurization Symposium sponsored by the U.S. Environmental
Protection Agency, Hollywood, Florida, November 8-11, 1977.
p. 7.
4-28
-------
13. Nelson, R.L., and O'Hara, R.D. Operating Experiences at
the Phillips and Elrama Flue Gas Desulfurization Facili-
ties. Presented at the 2nd Pacific American Chemical
Engineering Conference, Denver, Colorado, August 1977.
p. 309.
14. Laseke, B.A. Survey of Flue Gas Desulfurization Systems:
Bruce Mansfield Station, Pennsylvania Power Company. Pre-
liminary Draft. PEDCo Environmental, Cincinnati, Ohio
pp. 53-54.
15. Laseke, B.A., et al. EPA Utility FGD Survey: December
1977-January 1978. EPA-600/7-78-051a, March 1978. PEDCo
Environmental, Cincinnati, Ohib. pp. 29-32.
16. M.W. Kellogg Engineering Information Series. Improving
Mass Transfer Characteristics of Limestone Slurries by
Use of Magnesium Sulfate. Reprinted from Environmental
Science & Technology, June 1976.
17. Rochelle, G.T. Process Synthesis and Innovation in Flue
Gas Desulfurization. EPRI FP-463-SR. July 1977. p. 2-51.
18. Borgwardt, R.H. Significant EPA/IERL-RTP Pilot Plant
Results. Presented at the EPA Industry Briefing Conference.
August 29, 1978.
19. Op. cit. No. 2. p. 2-16.
20. Ibid.
21. Melia, M. et al. EPA Utility FGD Survey: October-November
1978. Preliminary Draft, January 1979. PEDCo Environ—
mental, Cincinnati, Ohio. pp. 78-79.
22. Laseke, B.A. Survey of Flue Gas Desulfurization Systems:
Sherburne County Generating Station, Northern States Power
Company. Preliminary Draft. PEDCo Environmental, Cincinn-
ati, Ohio. pp. 42-43.
23. Laseke, B.A. Survey of Flue Gas Desulfurization Systems:
Lawrence Energy Center, Kansas Power and Light Company.
Preliminary Draft. PEDCo Environmental, Cincinnati, Ohio.
pp. 49-50, 53.
24. Op. cit. No. 2. p. 3-6.
25. Ibid.
26. Ibid.
4-29
-------
27. Op. cit. No. 1.
28. Op. cit. No. 14.
29. McDaniel, C.F. La Cygne Station Unit No. 1 Wet Scrubber
Operating Experience. Presented at the Utility Scrubber
Conference, Denver, Colorado. March 29-30, 1978.
30. Op. cit. No. 2. pp. 3-4, 3-5, and 3-6.
31. Op. cit. No. 23. pp. 53-54.
32. Op. cit. No. 2. p. 3-13.
33. The Key Technical Issues of SO2 Scrubbing. Vol. 2-June 1976.
Prepared by Research-Cottrell, Bound Brook, New Jersey.
34. Laseke, B.A., and Devitt, T.W. Status of Flue Gas Desul-
furization Systems in the United States. Presented at
the 71st Annual AIChE Meeting, Miami Beach, Fla., November
12-16, 1978. pp. 32-33.
35. Op. cit. No. 22. pp. 26, 31, 32, and 52-55.
36. Op. cit. No. 23. p. 42.
37. Op. cit. No. 34. p. 42.
38. Laseke, B.A. Survey of Flue Gas Desulfurization Systems:
Choila Steam Electric Station, Arizona Public Service.
EPA-600/7-78-048a. March 1978. Prepared by PEDCo Environ-
mental, Cincinnati, Ohio pp. 11-13.
39. Laseke, B.A. Survey of Flue Gas Desulfurization Systems:
LaCygne Power Station, Kansas City Power and Light. EPA-
600/7-78-048b. March 1978. Prepared by PEDCo Environ-
mental, Cincinnati, Ohio. pp. 20, 33, 34, 37.
40. Op. cit. No. 23. pp. 36, 37.
41. Laseke, B.A. Survey of Flue Gas Desulfurization Systems:
Will County Power Station, Commonwealth Edison. EPA-
600/7-78-048d. March 1978. Prepared by PEDCo Environ-
mental, Inc., Cincinnati, Ohio. pp. 12, 31.
42. Laseke, B.A. Survey of Flue Gas Desulfurization Systems:
St. Clair Station, Detroit Edison Co. EPA-600/7-78-048c. .
March 1978. Prepared by PEDCo Environmental, Inc., Cin-
cinnati, Ohio. pp. 13, 15.
4-30
-------
43. Op. cit. No. 34. p. 34.
44. Op. cit. No. 34. p. 36.
45. Op. cit. No. 34. p. 30.
46. Op. cit. No. 34. p. 31.
47. Laseke, B.A. Affidavit for Case No. C2-78-786. United
States of America vs. Ohio Edison Company and Duquesne
Light Company.
48. Op. cit. No. 34. p. 31.
49. Lime Slurry Data Book. Preliminary Draft. Prepared for
EPA and EPRI by PEDco Environmental, Inc., Cincinnati,
Ohio. Section 5.4.11, p. 39.
50. Wells, W.L., et. al. TVA's Experience with Limestone
Scrubbers at the 550-MW Widows Creek Unit 8. Presented
at the American Pover Conference, Chicago, Illinois.
April 25-26, 1978.
51. Op. cit. No. 49 p. 39.
52. Op. cit. No. 34. pp. 32,33.
53. Op. cit. No. 49. p. 23.
54. Op. cit. No. 34. p. 33.
55. The Key Technical Issues of SO? Scrubbing. Vol. 4-January
1977. Prepared by Research-Cottrell, Bound Brook, New
Jersey.
56. Op. cit No. 2. pp. 2-14, 2-14.
57. The Key Technical Issues of SO? Scrubbing. Vol. 2-June
1976. Prepared by Research-Cottrell, Bound Brook, New
Jersey.
58. Op. cit. No. 2. p. 2-13.
59. Ballard, B., and M. Richman. FGD Systems Operation at
Martin Lake Steam Electric Station. Presented at the
Joint Power Generation Conference, Dallas, Texas.
September 10-13, 1978.
4-31
-------
60. Berube, D.T., and C.D. Grinmu Status and Performance of
the Montana Power Company's Flue Gas Desulfurization
System. In: Proceedings of the Symposium on Flue Gas
Desulfurization, Hollywood, Fla., November 1977 (Volume IJ .
EPA-600-7-058a. March 1978. pp. 277-291.
61. Op. cit. No. 2. p. 2-13.
62. Ibid.
63. Op. cit. No. 23. p. 47.
64. Op. cit. No. 38. pp. 8,9,10.
65. Op. cit. No. 59.
66. Op. cit. No. 34. pp. 33,34,35.
67. Op. cit. No. 49. Section 4.7.10., pp. 59-60.
68. Op. cit. No. 34. p. 35.
69. Laseke, B.A. Survey of Flue Gas Desulfurzation Systems:
Green River Station, Kentucky Utilities. EPA-600/7-78-048e,
March 1978. Prepared by PEDCo Environmental, Inc., Cin-
cinnati, Ohio. pp. 23,24.
70. Op. cit. No. 12. pp. 2,3,4.
71. Laseke, B.A. Survey of Flue Gas Desulfurization Systems:
Duck Creek Station, Central Illinois Light Company. Pre-
liminary Draft. Prepared by PEDCo Environmental, Inc.,
Cincinnati, Ohio p. 48.
72. Op. cit. No. 14. pp. 50,51,52,53,
73. Op. cit. No. 34. pp. 35,36,37.
74. Op. cit. No. 49. Section 4.11.11, pp. 41,42.
75. Op. cit. No. 34. pp. 38,39.
76. Op. cit. No. 34. pp. 40,41.
4-32
-------
4.2 RECENT TRENDS IN THE DEPENDABILITY OF THE TECHNOLOGY
The figures used in the earlier report and those used here
were developed from one or a combination of two or more of the
four performance factors used to characterize the operation of an
FGD system in the EPA Utility FGD Survey reports, but availabil-
ity/ operability, reliability, and utilization are now combined
and addressed (as a group) as dependability. These figures are
developed from the most appropriate parameter or parameters
reported by a given utility when more than one is reported or
from whatever parameters are available when only one utility is
reporting.
As was mentioned in the earlier report, the design and
operation of newer FGD systems benefit from the experience gained
on older units. As expected and as shown in Figure 4-1, depend-
ability improves as newer FGD units are put into operation.
Figure 4-1 also shows the cumulative FGD dependability percentage
as a function of FGD startup dates. Table 4-1 identifies the
units plotted in this figure.
Improved dependability can be seen in two ways. Because the
low performance of systems has been improving over the years, the
slope, although positive, is decreasing. The y-intercept clearly
has increased from roughly 40 percent in the March 1977 report to
53 percent in the updated September 1978 graph. Also the cumula-
tive dependability of the newer systems has in general been
greater than the linear regression line (i.e., the points lie
above the line) , with only a few very poor performers pulling the
high end of the regression line downward. The slope should
decrease as the average dependability approaches 100 percent with
newer FGD installations and as the dependability of early units
4-33
-------
*>.
i
GJ
£>.
^100
LLJ
^ 90
5 80
fe* >•
.* 70
>- Q
5 ^ 60
>-• UJ
OQ U.
|~ 50
I ^ 40
UJ
-------
TABLE 4-1. IDENTIFICATION OF PLANTS IN FIGURE 4-1
1. Will County No. 1
Commonwealth Edison
2. Mystic No. 6
Boston Edison
3. Hawthorn No. 4
Kansas City Power and Light
4. Hawthorn No. 3
Kansas City Power and Light
5. La Cygne No. 1
Kansas City Power and Light
6. Paddys Run No. 6
Louisville Gas and Electric
7. Cholla No. 1
Arizona Public Service
8. Mohave 2A
Southern California Edison
9. Reid Gardner No. 2
Nevada Power
10. Mohave 1A
Southern California Edison
11. Reid Gardner No. 1
Nevada Power
12. Scholz No. 1A
Gulf Power
13. Scholz IB and 2B
Gulf Power
14. Green River Unit Nos. 1, 2, 3
Kentucky Utilities
15. Colstrip No. 1
Montana Power
16. Sherburne County Station No. 1
Northern States Power
17. Bruce Mansfield No. 1
Pennsylvania Power
18. Cane Run No. 4
Louisville Gas and Electric
19. Colstrip No. 2
Montana Power
20. Reid Gardner No. 3
Nevada Power
21. Conesville No. 5
Columbus & Southern Ohio Electric
22. Lawrence 4a
Kansas Power and Light
23. Sherburne County Station No. 2
Northern States Power
24. Widows Creek No. 8
Tennessee Valley Authority
25. Bruce Mansfield No. 2
Pennsylvania Power
26. Martin Lake lb
Texas Utilities
Estimate based on PEDCo Environmental in-house files. The
newly modified system began operating in 1977.
Estimate based on Reference 1 and PEDCo Environmental in-house
files.
4-3-r
-------
improves. The dependability coordinate of the median point on
the graph is increasing as expected with improved FGD technology.
4.2.1 Availability of Selected Systems
Figure 4-2 is an updated version of the average availability
graph shown in the March 1978 report. This figure also shows the
range in average monthly systems (except where otherwise indi-
cated) for a number of selected FGD systems that historically
have shown high average performance. The figure also shows the
annual average (except where otherwise indicated) for the systems.
4.3 TRENDS IN MANUFACTURERS' GUARANTEES1
Many of the initial systems purchased by the utilities were
accompanied by nebulous process guarantees. In most instances
these guarantees proved to be less than binding because of lack
of specificity in covering the possible range of operating condi-
tions. Currently, however, the major system suppliers are willing
to supply detailed guarantees in several important areas, includ-
ing sulfur dioxide removal, particulate removal, acid mist in outlet,
waste stream quality/quantity, power consumption, water consump-
tion, reagent consumption, reheat energy consumption, and system
availability- Some general developments in each of these guar-
antee areas are briefly summarized in the following:
4.3.1 Sulfur Dioxide Removal
Previously, guarantees were written to specify only sulfur
dioxide removal for a specific coal sulfur content. The removal
level was such that, for the specified coal, sufficient sulfur
dioxide would be removed from the flue gas to meet the applicable
regulation. Problems were encountered with this strategy because
coal sulfur content and flue gas flow rate deviations can affect
the required sulfur dioxide removal and operation of the system.
Bidders are now being required to guarantee meeting the applica-
ble regulation over the entire range of operating conditions set
forth in the design basis.
4.3.2 Particulate Removal
For most systems coming on stream, particulate collection
4-36
-------
100
90
80
70
60
50
40
30
20
co
IN.
cn
en{^
to
cn
cn ir>
*~ cn
co
[
t£ ^
VO
«
O
oo
r»
en
en
MH\JKL AVERAGE
AVAILABILITY
FOR ENTIRE
FGD SYSTEM
RANGE IN INDIVIDUAL
MONTHLY MODULE
AVAILABILITY
CO
r-»
en
5"
r- oo
r» r
cn i
Su>
ui
3 o
§3
oo
en
LaCYGNE 1 CHOLLA 1* COLSTRIP 1 SHERBURNE 1 CANE RUN COLSTRIP 2 REID SHERBURNE 2 BRUCE
(5.4% S) (0.6% S) (0.8% S) (0.8% S) 4t (0.8X S) GARDNER 3 (0.8% S) MANSFIELD 2
(3.4-4.OX S) (0.5-l.OXS) (5.0% S MAX)
Figure 4-2. Average availability for selected FGD systems.
* Reliability
I Operability
j With the exception of Sherburne 1 and 2, these system*
-------
occurs in an ESP or fabric filter upstream of the FGD system.
Particulate emission is still an area of concern, however, be-
cause of slurry carryover and acid mist. Suppliers are now asked
to request a guarantee that the particulate loading at the FGD
outlet be no greater than that at the inlet (ESP or fabric
filter outlet). If particulate removal is provided in the
scrubbing system (e.g. simultaneous or two-stage), guarantees
like that for sulfur dioxide removal (Section 4.3.1) are appli-
cable.
4.3.3 Acid Mist
Although this is a nebulous area because of difficulty in
qualification and quantification, mist loading at the scrubber
discharge is being requested because of its impact on particulate
emissions and the operation of downstream equipment (reheater
loading, corrosion, etc.).
4.3.4 Power Consumption
Power consumption guarantees now specify a maximum percent-
age of the total plant power to be used by the FGD system.
4.3.5 Reheat Energy, Reagent, and Water Consumption
Normally, the system suppliers are requested to provide a
quaranteed maximum consumption rate for each of these parameters.
4.3.6 Sy s tern Ava i1abi1ity
Guarantees concerning system availability are now frequently
required of suppliers. Availabilities of 90 percent or greater
are now offered by most major suppliers.
4.4 SUMMARY OF SULFUR DIOXIDE REMOVAL EFFICIENCY
A complete list of U.S. utility FGD systems that have
achieved 90 percent or greater sulfur dioxide removals is pro-
vided in Table 4-2. This table identifies all pilot, prototype,
demonstration, and full-scale FGD systems according to utility,
station, unit, nature of application, and removal value(s). It
also provides information qualifying the nature of the removal
values listed.
4-38
-------
TABLE 4-2.
U.S. UTILITY FGD SYSTEMS ACHIEVING OR EXCEEDING
90% SULFUR DIOXIDE REMOVAL
Utility
Station
Unit
Application
% SO2 Removal
i
u>
'.£>
Arizona Public Service
Arizona Public Service
Boston Edison
Central Illinois Light
Columbus & So. Ohio Elec.
Commonwealth Edison
Detroit Edison
Duquesne Light
Gulf Power
Gulf Power
Gulf Power
Gulf Power
Kansas Power & Light
Kentucky Utilities
JLouisville Gas & Electric
LouisvilTe Gas & Electric
Nevada Power
Nevada Power
Nevada Power
Northern Indiana Public Serv,
Pennsylvania Power
Pennsylvania Power
Philadelphia Electric
Potomac Electric & Power
Public Service of New Mexico
Southern California Edison
Southern California Edison
Southern California Edison
Springfield City Utilities
Tennessee Valley Authority
Tennessee Valley Authority
Tennessee Valley Authority
Texas Utilities
Cholla
Four Corners
Mystic
Duck Creek
Conesville
Will County
St. Clair
Phillips
Scholz
Scholz
Scholz
Scholz
Lawrence
Green River
Cane Run
Paddy's Run
Reid Gardner
Reid Gardner
Reid Gardner
D.H. Mitchell
Bruce Mansfield
Bruce Mansfield
Eddystone
Dickerson
San Juan
High Grove
Mohave
Mohave
Southwest
Shawnee
Shawnee
Widows Creek
Martin Lake
1
5
6
1
5
1
6
1-6
1A
2A
1B&2B
1&2
4
1,2&3
4
6
1
2
3
11
1
2
1
3
1
1
1
2
1
10A
10B
8
1
Full-scale
Demonstration
Demonstration
Demonstration*
Full-scale
Full-scale
Demonstration
Full-scale
Prototype
Prototype
Prototype
Prototype
Full-scale
Demonstration
Full-scale
Demonstration
Full-scale
Full-scale
Full-scale
Full-scale
Full-scale
Full-scale
Demonstration
Demonstration
Full-scale
Prototype
Demonstration
Demonstration
Full-scale
Prototype
Prototype
Full-scale
Full-scale
92
95
90
92
90
92
90+
95
95
95
93 i
96-98.
m
m
m
90
99
90+
90+
90+
95"
95n
95
90
90+
90+
95
95
92P
95-99
95-9
94
(continued)
-------
TABLE 4-2 (continued)
Module A only. Total systems removal is 58.5 percent.
Performance test value.
One of four modules operating during an 8-month test.
Performance test value.
Performance test value estimate. Actual value not available.
Module A performance value during one series of tests.
g Thirty-day performance test value.
Two-stage scrubbing train test performance value.
Performance test value.
"T -1 Initial performance data based on American Air Filter strip chart data,
0 Initial performance data.
Performance data during limited test period.
Performance test values.
n Performance value based on Pennsylvania Power strip chart data.
Performance test value.
p Performance test value. Design removal is 80 percent.
^ Performance test value. Design removal is 75 percent.
r Scrubber removal values. Total system removal is 70 percent.
-------
An analysis of the systems identified in Table 4-2 indicates
that approximately half of them are experimental pilot, proto-
type, and demonstration units. Of the full-scale units listed,
only five (Conesville, Mitchell, Bruce Mansfield 1 and 2, and San
Juan) are required to remove 90 percent of the sulfur dioxide or
better at maximum coal sulfur content. The remaining full-scale
systems are included because of short-term test results (Lawrence,
Green River, Cane Run, Reid Gardner, Southwest, and Widows Creek)
or because of modular performance* (Cholla, Will County, Phillips,
and Martin Lake). Nineteen are medium- to high-sulfur coal
applications, 12 are low-sulfur coal applications, and 2 are
medium- to high-sulfur oil applications.
* Both short-term and long-term performance periods,
4-41
-------
SECTION 4.3
REFERENCES
1. Lime Slurry Data Book. Preliminary Draft. Prepared for
EPA and EPRI by PEDCo Environmental, Inc., Cincinnati, Ohio,
Section 5.3, pp. 7,8,9,10,11,12,13,14,15.
4-42
-------
APPENDIX A
DOMESTIC LIME SLURRY FGD
SCRUBBING SYSTEMS
A-l
-------
This appendix summarizes data on the lime slurry FGD scrub-
bing systems presented in Section 2.1 and describes the other
systems in this country that use lime or lime with alkaline fly
ash.
The following FGD systems are described:
0 Arizona Public Service - Four Corners 5
0 Columbus and Southern Ohio Electric Company - Conesville
5 and 6
0 Duquesne Light Company - Elrama Power Station
0 Duquesne Light Company - Phillips Power Station
0 Kansas City Power and Light - Hawthorn 3 and 4
0 Kentucky Utilities - Green River 1, 2, and 3
0 Louisville Gas and Electric - Cane Run 4
0 Louisville Gas and Electric - Cane Run 5
0 Louisville Gas and Electric - Mill Creek 3
0 Louisville Gas and Electric - Paddy's Run 6
0 Minnkota Power Cooperative - Milton R. Young 2
0 Montana Power - Colstrip 1 and 2
0 Pennsylvania Power - Bruce Mansfield 1 and 2
0 Tennessee Valley Authority - Shawnee 10A and 10B
0 Utah Power and Light - Huntington 1
A-2
-------
ARIZONA PUBLIC SERVICE
FOUR CORNERS 5
POLLUTION CONTROL1'2
The FGD system retrofitted on Four Corners 5 in February
1976 was a Weir horizontal crossflow spray scrubber operated in a
test program at the Mohave Generating Station of the Southern
California Edison Company. The system was used in a prototype
scrubbing program through December 1976 and has been previously
described. No formal FGD operations have taken place at Four
Corners since completion of the 10-month test.
FUTURE PLANS3'4
To meet current state regulations, New Mexico Air Pollution
Control officials have determined that 67.5 percent of the sulfur
dioxide (SO2) must be removed from the emissions at Four Corners.
At present, Four Corners 1, 2, and 3 each operate two Chemico
venturi scrubber modules for particulate control. Because of the
high alkalinity of the fly ash, these scrubbers remove roughly 30
percent of the S02 from the flue gas.
Tests were carried out on scrubber 2A (Four Corners 2)
between February 28 and March 10, 1976, to determine whether the
venturi scrubbers at Units 1, 2, and 3 could be upgraded to
achieve the necessary additional SO2 control. The program was
designed to characterize the influence of several operating
parameters, especially slurry pH, on SO2 removal efficiency.
Normal scrubber design was modified to allow the addition of a
lime slurry directly into the scrubber bottom, and the desired pH
was attained by varying the lime slurry flowrate. Figure A-l
A-3
-------
shows the effect of pH on S02 removal efficiency; and Table A-l
summarizes the important parameters and results.
These results, together with additional data from testing in
mid-1977, prompted Arizona Public Service to plan compliance with
SC>2 removal regulations by increasing the alkalinity of the
scrubbing solution in the three venturi scrubbers with more lime.
Full-scale operation is expected to begin in 1979.
At Four Corners 4 and 5, plans call for FGD systems to be
installed and start operation in 1982. United Engineers has been
selected to supply the systems, although a scrubbing process has
not yet been chosen. The utility is currently evaluating infor-
mation from the prototype scrubbing program to help ascertain the
emission control strategy.
A-4
-------
Table A-l. SUMMARY OF POUR CORNERS 2A SCRUBBER MODULE ISOLATION
TESTING SERVICE IN FEBRUARY AND MARCH 19764
Test
number
1
2
3
4
5
6
7
e
Ratio of liquid
-to gas, ,
11ters/mJ (gal/KP ft3)
2.9 (22)
2.7 (20)
2.8 (21)
2.8 (21)
3.1 (23)
2.7 (20)
2.7 (20)
3.9 (29)
Liquid rate.
llters/s (gal/nln)
31S (5000)
293 (4650)
300 (4750)
312 (4950)
322 (5100)
328 (5200)
341 (5400)
410 (6500)
Ventur-1 pressure
drop, In. HjO
18.3
20.4
19.3
19.4
20.2
9.3
15.7
19.5
Slurry
solids, I
2.5
5.4
10.6
3.5
5.6
3.2
2.0
Slurry
recycle pH
3.6
4.0
4.0
6.0
7.0
4.5
8.6
Inlet
S02, ppn
600
520
625
575
785
600
635
580
S02
removal, I
25
24
31
47
63
23
22
82
n
CaCOj
1.44 x
ID'6
1.32 >
ID'5
1.02 x
ID'5
0.108
2.71
9.94 x
lO'5
29.2
Slurry recycle
»lat1ve saturation
CaS03-l/2H20
7.56 x 10"3
0.068
.067
8.64
6.39
0.247
6.23
CaS04-2H20
1.78
1.63
1.66
1.93
1.05
1.12
0.51
Partial pressure
of S02 over
slurry recycle.
Pa (at»)
2.5 (/.5 x 10's)
2.9 (2.9 x 10"5)
3.5 (3.5 x 10'5)
2.4 x 10"2(2.4 x ID"7!
2.6 x 10'4(2.6 x 10'9
9.1 x 10"'(5.0 x 10"6]
1.8 x 10"7
(1.8 x 10'12)
Total number
of scrubbers
operating
6
4
4
2
2
2
2
2
Rate of
scale deposit
c«/«' per yr (In/ft2 per yr)
136 (4.96)
105 (3.84)
22 (0.80)
0 (0.00)
0 (0.00)
0 (0.00)
0 (0.00)
22 (0.81 )b
Lime
utilization.
127
65
a
63
4 Realistic line utilization could not be computed due to slaker operational problems.
b Scale appeared to be caUlun carbonate.
-------
100
TEST NO. 8
75
5:
o
UJ
Qi
CM
O
CO
50
25
TEST NO. 1
TEST NO. 3
NO. 4
TEST NO. 6
0
-TEST NO. 2
I
I
5 6 7
SCRUBBER SLURRY pH
8
Figure A-l.
Summary of Four Corners 2A Scrubber
Module Isolation Testing: Effect
of pH on SO2 removal.^
A-6
-------
COLUMBUS AND SOUTHERN OHIO ELECTRIC COMPANY
CONESVILLE 5 AND 6
BACKGROUND INFORMATION5
The Conesville Power Station is located on the Muskingum
River, near Coshocton, in northeast Ohio. The plant's current
capacity is 2055 MW (gross). Conesville 1,2, and 3 have a com-
bined capacity of 433 MW (gross) and share a common stack.
Conesville 4 is rated at 800 MW (gross) and Conesville 5 and 6
are each rated at 411 MW (gross). Conesville 4,5, and 6 each
have a separate stack.
The Conesville 6 steam generator is a dry-bottom, pulver-
ized-coal-fired unit supplied by Combustion Engineering (CE). A
mixture of high-sulfur (from 4.2 percent to 5.1 percent sulfur)
Ohio coals is burned with a heating value from 24,000 to 26,000
kJ/kg (10,300 to 11,200 Btu/lb) and ash content from 12 to 19
percent. A conveyor delivers 40 percent of the coal from a mine
complex 11 km (7 miles) from the plant site. The remainder is
trucked in from southeast Ohio.
POLLUTION CONTROL6
The air pollution control system at Conesville 6 is identi-
cal in design to the system originally installed at Conesville 5
It consists of two parallel turbulent contact absorber (TCA)
towers supplied by the Air Correction Division of Universal Oil
Products (UOP). The towers are preceded by a Research Cottrell
(RC) cold-side electrostatic precipitator (ESP). Each module
consists of two stages of 3.8-cm (1.5-in.) solid rubber balls
that provide the turbulent contacting surface between the lime
A-7
-------
slurry and the flue gas. The modules were each originally
designed to handle 60 percent of the flue gas flow.
The Conesville 5 FGD system has been modified by removing
the second TCA stage of each of the two modules because it was
discovered that the SO design efficiency of 89.5 percent could
be met without the second stage. Less power is thus consumed by
the FGD system, and it is hoped that a higher system availability
can be achieved.
Detailed information about Conesville 5 and 6 is presented
in Table A-2.
PERFORMANCE HISTORY7'8
The A-module of Conesville 5 began operation in November
1977 after startup was delayed by a fire. Conesville 5 was taken
out of service for a scheduled 3-month overhaul in December 1977,
when the carbon steel stack liner was replaced with an acid brick
liner.
After Conesville 5 was returned to service in March 1978,
problems were encountered with the thickener. An excess of
flocculant was caused by a malfunction of the feed system, and
the tank had to be drained.
From March through September 1978, system availability
averaged 44 percent; a high of 66 percent occurred during April.
A low of 18 percent in August was caused by scale buildup in the
mist eliminator that required removal, and by outage time for
replacement of some of the balls in the TCA regions.
Conesville 6 was placed in service in June 1978. Initial
problems included a loss of control of the bypass system louvered
damper, and backpressure buildups causing automatic boiler shut-
downs. In July, the damper controls were adjusted in an attempt
to correct the problem. Plugging problems have been experienced
with the fiber-reinforced plastic transfer line from the thick-
ener to the fixation system supplied by IU Conversion Systems
(IUCS). The utility has reported that the bypass control dampers,
as well as the sludge line, typically require high maintenance.
A-8
-------
TABLE A-2. FGD SYSTEM DATA FOR CONESVILLE 5 AND 6,
CONESVILLE, OHIO
Unit rating (gross), MW
(net), MW
Fuel
Average fuel characteristics:
Heating value, kJAg (Btu/lb)
Ash, %
Moisture, %
Sulfur, %
Chloride, %
FGD process
FGD system supplier
Application
Status
Startup date:
Initial
Commercial
Design removal efficiency:
Particulate, %
Sulfur dioxide, %
Actual removal efficiency:
Particulate, %
Sulfur dioxide, %
Water loop
Total water makeup,
liters/s'per net MW
(gal/min per net MW)
Sludge disposal
411
375
Bituminous coal
25,240 (10,850)
15.1
7.5
4.7
Lime (thiosorbic)
Air correction division, UOP
New
Operational
January 1977
June 1978
99.6
89.5
Open
0.08
1.25
Stabilized sludge disposed
of in an onsite landfill
A-9
-------
From June through September 1978, system availability aver-
aged 61 percent. A high of 77 percent occurred in July; a low of
50 percent occurred in June. Monthly operating hours have in-
creased from 174 in June to 372 in September.
FUTURE PLANS9
Columbus and Southern Ohio Electric has plans for two new
units to come on line in 1983. Poston 5 and 6 are each rated at
375 MW and will burn high-sulfur (approximately 2.5-percent
sulfur) Ohio coal. FGD systems are being considered for both
Poston 5 and 6, but neither a vendor nor a process has been
selected.
A-10
-------
DUQUESNE LIGHT COMPANY
ELRAMA POWER STATION
POLLUTION CONTROL10
The emission control system at the Elrama Power Station
consists of five single-stage variable-throat venturi scrubbers
arranged in parallel. Each module is designed to treat 255 m /s
(540,000 ft3/min) of flue gas at 151°C (303°F). The venturi
scrubbers are designed to accommodate four boilers and are lo-
cated downstream of mechanical collectors and ESP's, which are
installed in series on each of the four boilers. The scrubbed
gas enters a 4100-kW (5500-hp) induced-draft fan, and are dis-
charged to an external mist eliminator vessel. A common duct
transfers the gas through a direct oil-fired reheat system, to a
wet gas stack approximately 122 m (400 ft) high. The sludge
produced is stabilized by the Poz-O-Tec method of IU Conversion
Systems (IUCS) and hauled to an offsite landfill.
Table A-3 presents additional design information.
PERFORMANCE HISTORY
More sludge handling equipment was placed in service in
April 1978. The two original thickeners, each 7.6 m (25 ft) in
diameter, were replaced by two thickeners 36.6 m (120 ft) in
diameter; and the IUCS equipment was upgraded. Full-time opera-
tion of the scrubber system with thiosorbic lime containing from
8 to 9 percent magnesium oxide (MgO) began after the end of a
coal strike in late March 1978. Initial testing indicated that
the unit was well within particulate emission standards, but the
SO_ emission tests were not handled properly and will be redone.
A-ll
-------
TABLE A-3. FGD SYSTEM DATA FOR ELRAMA POWER STATION,
ELRAMA, PENNSYLVANIA
Unit rating (gross), MW
(net), MW
Fuel
Average fuel characteristics:
Heating value, kJ/kg (Btu/lb)
Ash, %
Moisture, %
Sulfur, %
Chloride, %
FGD process
FGD system supplier
Application
Status
Startup date:
Initial
Commercial
Design removal efficiency:
Particulate, %
Sulfur dioxide, %
Actual removal efficiency:
Particulate, %
Sulfur dioxide, %
Water loop
Total water makeup, liters/s per MW
(gal/min per MW)
Sludge disposal
510
494
Coal
26,400 (11,350)
21
2.2
Lime (thiosorbic)
Chemico
Retrofit
Operational
October 1975
September 1976
(all five modules)
99
83
Open
Stabilized sludge trucked
to an offsite landfill
A-12
-------
The fourth boiler was tied into the FGD system in January
1978. Full-scale scrubber operations, however, were hampered by
low demand and the unavailability of coal. A coal strike forced
Elrama to shut down on February 11, 1978; and repairs and modi-
fications were performed on the FGD system. Boiler exit dampers
were lined with 316 stainless steel (SS) on areas of high ero-
sion; and holes in the wet gas duct were repaired and relined
with Ceilcote. After returning to service in late March, plug-
ging within the mist eliminator became a problem. Low pH result-
ing from failures of the lime handling and slurry preparation
systems was a primary cause of the plugging, and it was necessary
to shut down the lime mixing basin to clean out excessive grit
and solids buildup.
The utility is also investigating ways to improve the water
balance by intermittently blending thickener supernatant for the
mist eliminator wash with clean service water.
A-13
-------
DUQUESNE LIGHT COMPANY
PHILLIPS POWER STATION
POLLUTION CONTROL14
The pollution control equipment at the Phillips Power Sta-
tion consists of four parallel scrubbing trains that are situated
downstream of the mechanical collectors and ESP's, and are in-
stalled in series on each of the six boilers. Each train is a
Chemico single-stage variable-throat venturi scrubber designed to
treat 255 m3/s (540,000 acfm) of flue gas at 151°C (303°F). One
train has a second-stage scrubber after a wet induced-draft fan
that was designed to provide additional removal of SO-. Duquesne
Light intended to install second-stage scrubbers on the remaining
three modules/ but tests have indicated that single-stage scrub-
bers can achieve the required SO2 removal efficiency of 83 per-
cent by using thiosorbic lime in the existing Venturis. Dravo is
supplying the Phillips Power Station with thiosorbic lime con-
taining from 8 to 9 percent MgO; and some design modifications
have been implemented, such as the addition of more lime storage
and slaker capacity, and a third thickener 23 m (75 ft) in diam-
eter. Sludge is treated by the IUCS Poz-O-Tec method and trucked
to an offsite landfill.
The scrubbed gases are discharged to an external mist
eliminator and through a common duct to a direct oil fired reheat
system. The stack is a wet gas type approximately 122-m (400-ft)
high. Table A-4 presents additional design information.
PERFORMANCE HISTORY15'16
Tests with thiosorbic lime to check SO2 removal efficiencies
began late in May 1977, but were terminated prematurely because
A-14
-------
TABLE A-4. FGD SYSTEM DATA FOR PHILLIPS POWER STATION,
SOUTH HEIGHT, PENNSYLVANIA
Unit rating (gross), MW
(net), MW
Fuel
Average fuel characteristics:
Heating value, kJ/kg (Btu/lb)
Ash, %
Moisture, %
Sulfur, %
Chloride, %
FGD process
FGD system supplier
Application
Status
Startup date:
Initial
Commercial
Design removal efficiency:
Particulate, %
Sulfur dioxide, %
Actual removal efficiency:
Particulate, %
Sulfur dioxide, %
Water loop
Total water makeup, liters/s per MW
(gal/min per MW)
Sludge disposal
410
387
Coal
26,400 (11,350)
18.2
2.2
Lime
Chemico
Retrofit
Operational
July 1973
99
83
Open
Stabilized sludge
disposed of in an
offsite landfill
-------
of leaking in a duct leading to the stack. The problem was
corrected; and tests continued, although occasional problems with
the slaker and feeder caused interruptions. Full-time use of
thiosorbic lime as the scrubbing reagent began in late March
1978, when the Phillips Power Station returned to service after a
2-month coal strike. Tests conducted by Allegheny County in late
August 1978 indicated that the Phillips Station was in compliance
with the regulation limiting emissions of SO0 to 258 mg/J (0.6
6
lb/10 Btu). Thus, the system was removing approximately 83
percent of the inlet SO- when coal containing 2 percent sulfur
was burned.
Problems have occurred because of grit buildup in the lime
handling and slurry preparation systems. Occasional failure of
these systems has caused low-pH swings and mist eliminator
plugging. It was thus necessary to shut down the lime mixing
basin to clean out excessive grit and solids. Water balance
problems have also been attributed to grit buildup, and ways to
improve the water balance with thickener supernatant have been
investigated.
Another problem has been insufficient dry fly ash to mix
with the sludge. The product from the IUCS system is not as
solid as desired and has leaked from transport trucks en route to
the final disposal site. Replacement of the present cyclone
collectors is being investigated as a solution the problem.
A-16
-------
KANSAS CITY POWER AND LIGHT
HAWTHORN 3 AND 4
BACKGROUND INFORMATION17
The Hawthorn Power Station of Kansas City Power and Light is
located in a heavily industrialized area on the north bank of the
Missouri River in East Kansas City, Missouri. The plant contains
five coal-fired power generating units. Hawthorn 1 and 2 are
considered peaking units, each nominally rated at 80 MW. Haw-
thorn 3 and 4 are cyclic boilers, each nominally rated at 100 MW,
when burning coal. These four boilers were built and placed in
service between 1950 and 1955. Hawthorn 5 was placed in service
in early 1970; it operates as a base-load unit at a rated capa-
city of 480 MW.
The steam generators at the Hawthorn Power Station are all
dry-bottom, pulverized-coal-fired units designed and manufactured
by Combustion Engineering, which also manufactured the FGD
systems for this plant.
Two grades of coal are burned. One has a heating value of
26,500 kJ/kg (11,400 Btu/lb), ash content of 14 percent, and
sulfur content of 3 percent. The other has a heating value of
22,800 kJ/kg (9800 Btu/lb), ash content of 11 percent, and sulfur
content of 0.6 percent.
POLLUTION CONTROL18'19
Only Hawthorn 3 and 4 are fitted with FGD systems. Origi-
nally, both FGD systems operated by furnace injection of ground
limestone rock followed by a flue gas wet scrubbing system in
which both the S02 and the furnace-calcined limestone were scrub-
bed and allowed to react in a reaction tank. In Feburary 1977, a
A-17
-------
conversion was made to tail-end lime slurry systems for both
units. Each system now includes two marble-bed scrubber modules
that can be bypassed in case of an emergency. Reheat is provided
by a finned-tube steam system downstream of a chevron-type mist
eliminator. The total gas capacity of each FGD system is 306
m /s (648,000 acfm) at 149°C (300°F). Spent slurry from each
system is discharged to a clarifier tank 35 m (115 ft) in diame-
ter. Overflow from this tank flows to a clean well tank, where
makeup water from the Missouri River is added. The discharge
from the clear well tank is recycled to the FGD system. The
underflow from the clarifier is pumped without treatment to a
160-acre pond that is also used for disposal of fly ash from the
other boilers.
Additional system design data are presented in Table A-5.
PERFORMANCE HISTORY20'21
Tail-end lime scrubbing operations began early in 1977 for
Hawthorn 3 and 4. Particulate testing soon after startup indi-
cated that the new systems were well within the emission standard
of 73 ng/J (0.17 lb/10 Btu). Tests of S02 removal efficiency
have not been run because the emission limitation regulation is
not established for this station.
Problems with process chemistry were encountered, but
quickly corrected. Shortly after startup, the FGD systems were
pulled off line so that the underbed spray headers could be
replaced with 316L SS headers. A strike at the Hawthorn Plant in
mid-1978 has hampered recordkeeping, but the FGD system has
remained operational.
A-18
-------
TABLE A-5,
FGD SYSTEM DATA FOR HAWTHORN 3 AND 4,
KANSAS CITY, MISSOURI
Unit rating (gross), MW
(net), MW
Fuel
Average fuel characteristics:
Heating value, kJ/kg (Btu/lb)
Ash, %
Moisture, %
Sulfur, %
Chloride, %
FGD process
FGD system supplier
Application
Status
Startup date:
Initial
Commercial
Design removal efficiency:
Particulate, %
Sulfur dioxide, %
Actual removal efficiency:
Particulate, %
Sulfur dioxide, %
Water loop
Total water makeup, liters/s per MW
(gal/min per MW)
Sludge disposal
100 each
Coal
24,770 (10,650)
12.5
2.0
Lime
Combustion Engineering
Retrofit
Operational
August 1972
November 1972
99
70
Open
Unstablized sludge disposed
of in unlined pond
A-19
-------
KENTUCKY UTILITIES
GREEN RIVER 1,2, AND 3
POLLUTION CONTROL22'23
Acid fallout from the plume at Green River 1,2, and 3 dam-
aged employees' automobiles and the superstructure of the nearby
substation, so that design modifications were necessary. Ameri-
can Air Filter was authorized to install an indirect hot-air
reheat system to raise the stack gas temperature by 28°C (50°F).
The system utilizes extraction steam from an adjacent unit to
heat ambient air, which is injected into the stack to increase
the exhaust temperature. The radial-vane mist eliminator was
modified to reduce formation of acid mist and fouling, but this
modification was not as effective as was hoped. The unit has
since been replaced with a chevron-type mist eliminator.
Other modifications included replacement of the smaller
balls in the mobile bed contacter with larger balls to reduce
ball migration, replacement of the rubber-covered pump impellers
with Ni-hard impellers, and redesigning the degritter to reduce
grit pickup.
Table A-6 presents additional system design data.
PERFORMANCE HISTORY 4~26
A strike at the Green River Plant by operating personnel
began in July 1977 and caused the plant to be shut down through
November 1977, although the FGD system was available for opera-
tion. When restarted, the winter weather caused numerous freeze-
ups and resulting outages. In January 1978, the sludge line to
the pond froze; and the system had to be shut down after 170
hours of operation. Because of emergency conditions, the utility
A-20
-------
TABLE A-6. FGD SYSTEM DATA FOR GREEN RIVER 1, 2, and 3,
CENTRAL CITY, KENTUCKY
Unit rating (gross), MW
(net), MW
Fuel
Average fuel characteristics:
Heating value, kJ/kg (Btu/lb)
Ash, %
Moisture, %
Sulfur, %
Chloride, %
FGD process
FGD system supplier
Application
Status
Startup date:
Initial
Commercial
Design removal efficiency:
Particulate, %
Sulfur dioxide, %
Actual removal efficiency:
Particulate, %
Sulfur dioxide, %
Water loop
Total water makeup,
liters/s per net MW
(gal/min per net MW)
Sludge disposal
64
61
Coal
25,120 (10,800)
13.4 (high-sulfur coal)
12.1 (high-sulfur coal)
4.0 (high-sulfur coal);
1.0 (low-sulfur coal)
Lime
American Air Filter
Retrofit
Operational
September 1975
June 1976
99.5
80.0
Closed5
4.7
75
Unstabilized sludge disposed
of in an onsite sludge pond
aSystem water loop verified with utility during Oct.-Nov. 1978
EPA Utility FGD Survey Report update.
A-21
-------
concentrated on electrical generation; and the FGD system was not
returned to service until April.. During the outage, torn gaskets
and other weather-related problems were corrected. In May, the
screens on the suction side of the pumps that deliver the slurry
from the preparation area to the nozzles kept plugging. The
utility reported that this is a recurring problem caused by large
grit in the slurry- Other FGD system outages through September
1978 have been for routine maintenance and correction of bleed
pump problems.
From November 1977 through September 1978, the scrubber was
available 5540 hours and operated 2912 hours. The boilers were
in service 4713 hours. Thus, system availability, operability,
reliability, and utilization in the 11-month period were 63, 58,
82, and 34 percent, respectively.
From October 17 through 19, 1978, PEDCo Environmental con-
ducted tests at the Green River Station to determine scrubber
efficiency. These tests indicated that the SO? removal effici-
ency averaged 81.2 percent. The average outlet SO2 concentration
was 555 ng/J (1.29 lb/106 Btu).
A-22
-------
LOUISVILLE GAS AND ELECTRIC
CANE RUN 4
POLLUTION CONTROL27"29
Many problems were encountered at Cane Run 4 during the FGD
system's initial operating phase, especially poor gas flow dis-
tribution and excessive pressure drop. As a result, the booster
fans lacked sufficient capacity; and the maximum boiler output
ranged from 150 to 155 MW, although the gross rating is 190 MW.
To remedy these problems, as well as the other problems described
below, a number of major system modifications were made during a
3-month overhaul period which extended from April 18, 1977, to
July 17, 1977.
Turning vanes were installed in the flooded elbow and the
base of each scrubber; and the mist eliminator was converted to a
chevron type by cutting away portions of the mist eliminator
wheel. At high temperatures, the original spray nozzle casings
expanded so that the spinner vanes stuck out the front and caused
the nozzle to plug. To correct this problem, the nozzles were
replaced with ones constructed of sturdier ceramic material.
More spray headers were also installed above the mobile beds, and
the pump capacity was increased. The L/G ratio was thus in-
3 33
creased to approximately 8 liters/m (60 gal/10 ft ) .
Other modifications were made to improve overall SO2 re-
moval efficiency. Underbed sprays were added to improve circu-
lation of the mobile bed balls; and existing pH meters were
replaced with new meters so that the pH could be monitored more
precisely. "Black-lime", which comes from a carbide slag opera-
tion and contains from 2 to 4 percent MgO, was also put into use.
A-23
-------
The stack at Cane Run 4 was approximately 25 years old and
constructed of acid brick. No stack liner or reheat system was
provided in the original FGD system design, and condensate in the
stack was causing acid corrosion. Louisville Gas and Electric
thus installed direct oil-fired heaters at the base of the stack
in conjunction with the turning vanes to provide a good flow and
mix. These heaters were designed to increase stack gas tempera-
ture from 22° to 28°C (40° to 50°F). The brickwork inside the
stack was removed and replaced with wire mesh covered with gunite,
Also, the duct work between the mist eliminator and the stack was
relined with acid-resistant concrete. Table A-7 presents addi-
tional design information.
PERFORMANCE HISTORY30'31
The modifications to the FGD system were completed in July
1977- Early in August, the system was tested for compliance with
SO_ emission regulations; it met the Jefferson County removal
requirement of 85 percent and the Federal standard of 515.9 ng/J
(1.2 lb/10 Btu) of heat input. The testing was handled by EPA
personnel and indicated an SO- removal efficiency between 86 and
89 percent for coal containing from 3.3 to 3.4 percent sulfur
coal. This efficiency amounted to outlet emissions of 343.9 ng/J
(0.8 lb/106 Btu).
The operability of the FGD system was greater than 90 per-
cent from August through December 1977. In January 1978, after a
lime feed line froze late in December, the system operated
intermittently.
A coal strike forced the generating unit to be shut down
during February 1978. It came back on line in March, and no FGD
system forced outages have been reported through September 1978.
A-24
-------
TABLE A-7. FGD SYSTEM DATA FOR CANE RUN 4,
LOUISVILLE, KENTUCKY
Unit rating (gross), MW
(net), MW
Fuel
Average fuel characteristics:
Heating value, kJ/kg (Btu/lb)
Ash, %
Moisture, %
Sulfur, %
Chloride, %
FGD process
FGD system supplier
Application
Status
Startup date:
Initial
Commercial
Design removal efficiency:
Particulate, %
Sulfur dioxide, %
Actual removal efficiency:
Particulate, %
Sulfur dioxide, %
Water loop
Total water makeup,
liters/s per net MW
(gal/min per net MW)
Sludge disposal
190
178
Coal
26,750 (11,500)
15.5
3.8
Lime (carbide)
American Air Filter
Retrofit
Operational
August 1976
99
85
Open
0.04
0.6
Stabilized sludge disposed
of in an onsite sludge pond
A-25
-------
LOUISVILLE GAS AND ELECTRIC
CANE RUN 5
BACKGROUND INFORMATION32
The Cane Run Power Station in Louisville, Kentucky, is
operated by the Louisville Gas and Electric Company. The station
has six units, which provide a total steam turbine generating
capacity of 992 MW (net).
Cane Run 5 is a coal-fired boiler with a continuous capacity
of 183 MW (net) and 200 MW (gross). The coal burned has an
average heating value of 26,750 kJ/kg (11,500 Btu/lb), an average
sulfur content of 3.8 percent, and an average ash content of 15.5
percent.
POLLUTION CONTROL33'34
The emission control system for Cane Run 5 was supplied by
Combustion Engineering and consists of two spray towers down-
stream of the existing electrostatic precipitator (ESP), which
provides primary particulate control. Carbide lime is the
scrubbing reagent and is injected into each spray tower through a
series of three spray stages. At full load, the gas velocity is
3 m/s (10 ft/s), with a pressure drop of approximately 0.75 kPa
(3 in. H20) across the absorber. The FGD system capacity is 330
m3/s (700,000 acfm) at 154°C (310°F). The absorber L/G ratio
3 3
ranges from 8 to 13 1/m (60 to 100 gal/ft ). An inline steam
reheat increases stack gas temperature by 22°C (40°F).
The FGD system produces 16 liters/s (248 gal/min) of spent
absorbent, with a solids content of 20 percent. This spent
slurry is discharged to a thickener 33.5 m (110 ft) in diameter.
A-26
-------
After stabilization, the sludge is disposed of in an onsite clay-
lined pond.
Additional FGD system information is presented in Table
A-8.
PERFORMANCE HISTORY35
The FGD system at Cane Run 5 began operation on December 29,
1977. Some minor modifications were necessary for controls that
were not operating properly. Cane Run 5 was removed from service
in February 1978 because of a coal strike and was restarted in
March. The FGD system operated approximately 91 hours in March,
although startup problems still caused intermittent operation.
From April through September 1978, operability was 80 percent or
greater. Trouble with the reheat steam coils caused some outage
time; the welds at the end of each bank of coils have been foul-
ing since startup.
Performance tests were conducted during the period from
April through May 1978. Reports by the utility indicate that the
tests were inconclusive because of poor handling of original data
and improper test procedures.
FUTURE PLANS36
Louisville Gas and Electric is installing a double alkali
scrubbing system at Cane Run 6. The system is being supplied by
A. D. Little/Combustion Equipment Associates and is expected to
begin initial operation in December 1978.
A-27
-------
TABLE A-8. FGD SYSTEM DATA FOR CANE RUN 5,
LOUISVILLE, KENTUCKY
Unit rating (gross), MW
(net), MW
Fuel
Average fuel characteristics:
Heating value, kJ/kg (Btu/lb)
Ash, %
Moisture, %
Sulfur, %
Chloride, %
FGD process
FGD system supplier
Application
Status
Startup date:
Initial
Commercial
Design removal efficiency:
Particulate, %
Sulfur dioxide, %
Actual removal efficiency:
Particulate, %
Sulfur dioxide, %
Water loop
Total water makeup, liters/s per MW
(gal/min per MW)
Sludge disposal
200
183
Coal
26,750 (11,500)
15.5
3.8
Lime (carbide)
Combustion Engineering
Retrofit
Operational
December 1977
June 1978
99
85
Open
Stabilized sludge disposed
of an onsite sludge pond
A-28
-------
LOUISVILLE GAS AND ELECTRIC
MILL CREEK 3
BACKGROUND INFORMATION37'38
The Mill Creek Power Station in Louisville, Kentucky, is
owned and operated by Louisville Gas and Electric. The plant has
three operational units, which provide a total capacity of 1085
MW (net). A fourth unit currently under construction will bring
overall plant capacity to 1580 MW (net).
Mill Creek 3 is a coal-fired boiler that has a capacity of
425 MW (net) and was supplied by Babcock and Wilcox. The coal
burned has an average heat content of 26,750 kJ/kg (11,500 Btu/lb),
sulfur content of 3.8 percent, and ash content of 11.5 percent.
POLLUTION CONTROL39"41
The emission control system at Mill Creek 3 was installed by
American Air Filter (AAF). It is similar in design to, although
larger than, the system at Cane Run 4. Carbide lime is the
scrubbing reagent. Design modifications found necessary at Cane
Run 4 have been incorporated at Mill Creek. Four mobile-bed
absorbers are preceded by ESP's also supplied by AAF. The scrub-
ber linings for the entire FGD system are Precrete applied to
carbon steel. The system operates in an open loop mode with
fresh water makeup at 9.5 liters/s (150 gal/min). A steam reheat
system is provided for the exiting flue gas, and the scrubbing
waste is stabilized with fly ash and lime.
Additional system design data are presented in Table A-9.
A-29
-------
TABLE A-9. FGD SYSTEM DATA FOR MILL CREEK 3,
LOUISVILLE, KENTUCKY
Unit rating (gross), MW
(net), MW
Fuel
Average fuel characteristics:
Heating value, kJ/kg (Btu/lb)
Ash, %
Moisture, %
Sulfur, %
Chloride, %
FGD process
FGD system supplier
Application
Status
Startup date:
Initial
Commercial
Design removal efficiency:
Particulate, %
Sulfur dioxide, %
Actual removal efficiency:
Particulate, %
Sulfur dioxide, %
Water loop
Total water makeup,
liters/s per net MW
(gal/min per net MW)
Sludge disposal
425
Coal
26,750 (11,500)
11.5
3.8
Lime (carbide)
American Air Filter
New
Operational
August 1978
99
85
Open
0.02
0.35
Stabilized sludge disposed
of in a sludge pond
A-30
-------
PERFORMANCE HISTORY42
The FGD system at Mill Creek 3 began operation on August 13,
1978, and logged 576 hours in September, for an average opera-
bility of 81 percent. Initial problems were encountered with
some piping made of fiber-reinforced plastic (FRP); and certain
pumps experienced bearing and shaft failures.
FUTURE PLANS43
Lime/limestone FGD systems have been ordered from Combustion
Engineering for Mill Creek 1 and 2, each of which is 330 MW
(net). The Mill Creek 1 system is expected to be put into opera-
tion in January 1981; the Mill Creek 2 system, one year later. A
compliance schedule for these two units has been submitted to the
Jefferson County Air Pollution Control District.
A mobile-bed absorber similar to that at Mill Creek 3 will
be installed by AAF at Mill Creek 4 (a coal-fired unit currently
under construction). The capacity of Mill Creek 4 will be 495 MW
(net), and FGD operations are expected to begin in June 1980.
A-31
-------
LOUISVILLE GAS AND ELECTRIC
PADDY'S RUN 6
PERFORMANCE HISTORY44"47
After startup in the spring of 1973, the FGD system at
Paddy's Run 6 operated with so few problems that the EPA spon-
sored a test program there to determine some of the factors-
attributing to the success. Tests began on October 25, 1976, and
were carried out in four phases.
The initial phase involved characterizing the Paddy's Run
system as it normally operated, with carbide lime as the absor-
bent. After steady state operating conditions were established,
key performance parameters were monitored and used to calculate
SO2 removal rates, sulfite/sulfate precipitation rates, lime
utilization efficiencies, and the sulfite oxidation rate. It was
noted that the system oxidation rate was low, with less than 10
percent of the SO2 removed from the flue gas being oxidized to
sulfate.
Phase two of the testing began in mid-March 1977- Commer-
cial lime was used as the alkaline additive rather than carbide
sludge. The major goal of this phase was to identify differences
in operation that might result from changing the additive.
Although SO,, removal rates were not affected, the system operated
in or near a supersaturated mode most of the time, and produced
large masses of gypsum under varying conditions.
The third and fourth testing phases overlapped. Phase three
was to include hold tank modifications to determine whether high
lime utilization efficiencies could be attained with short hold
tank residence times. Phase four was to determine the affects of
chloride and magnesium levels on subsaturated gypsum operation.
A-32
-------
The fourth phase began with the addition of a slurry con-
taining 55 percent magnesium hydroxide from Dow Chemical to the
commercial lime. With this addition, the liquid phase magnesium
remained in the 4000 ppm range, and very high SO2 removal rates
were achieved, while gypsum saturations approached zero. As the
magnesium level was lowered to 2000 ppm, calcium and relative
saturation levels began to rise tremendously, and it was neces-
sary to add more magnesium to the system. A magnesium level
between 2400 and 3000 ppm was found to be the best level for
system control.
During August 1977, the third phase occurred. There were
tests to eliminate the reaction tank and to reduce the reaction
time from 35 minutes to a period between 3 and 5 minutes. No
serious problems occurred; however, it could not be determined
whether calcium sulfite scaling was promoted during the short
duration of the tests. The solids content of thickener underflow
decreased with the reduced reaction tank residence.
The completion of the fourth phase of testing on August 31,
1977, involved adding calcium chloride and reducing reaction time
to produce conditions similar to ones from firing high-chloride
coal. Initially, the chloride level was 9000 ppm, which was then
reduced to 3000 ppm. To compensate for chloride addition,
magnesium levels were increased to 3500 ppm. The SO. removal
efficiencies were around 99 percent, and gypsum relative satur-
ation levels were very low. No operational problems were encoun-
tered. Thus, high levels of chlorides can be controlled with
high levels of magnesium.
In September 1977, Paddy's Run was no longer needed to
supply electricity and was taken out of service until April 1978,
when it was operated for only a few hours. Intermittent opera-
tion totaling 8 to 10 days occurred in June and July. In Sep-
tember 1978, the unit was placed in'service for two weeks, when a
new type of flocculant was tested in the thickener. The results
of the test may affect the type of flocculant used in the future
at other Louisville Gas and Electric units.
A-33
-------
MINNKOTA POWER COOPERATIVE
MILTON R. YOUNG 2
dfi
BACKGROUND INFORMATION
The Milton R. Young Station in Center, North Dakota, is
owned and operated by Minnkota Power Cooperative and Square Butte
Electric Cooperative. The station has two units, providing a
total capacity of 690 MW.
Milton R. Young 2 has a maximum gross capacity of 477 MW and
a maximum net capacity of 408 MW. The boiler was manufactured by
Babcock and Wilcox, and placed in service in 1977. The maximum
flue gas flow is 954 m3/s (2,021,000 acfm) at 179°C (355°F).
The fuel burned is lignite, with an average heating value of
15,000 kJ/kg (6,500 Btu/lb), sulfur content of 0.7 percent, and
ash content of 7 percent. The moisture content is 38 percent.
POLLUTION CONTROL49
The emission control system for this unit was supplied by A.
D. Little/Combustion Equipment Associates. It consists of two
spray towers that utilize alkaline fly ash and lime supplement as
the scrubbing reagents. Primary particulate control is provided
by a cold-side ESP from Wheelabrator-Frye. Design criteria and
operating parameters for the FGD system were determined during an
extensive test program at a pilot plant with a gas flow of 2.4
m /s (5,000 acfm). The program was conducted by Minnkota Power
Cooperative and the system suppliers in conjunction with the
Minnesota Power and Light Company and the Grand Forks Energy
Research Center.
Boiler flue gas at 849 m3/s (1.8 x 106 acfm) and 171°C
(34°F) first passes through the ESP. Inlet particulate loading
A-34
-------
ranges from,1.6 to 3.4 g/m (0.7 to 1.50 gr/ft ) and averages
approximately 2.6 g/m (1.15 gr/ft ). The ESP is designed to
remove 99.5 percent of participates, so that outlet loading
averages 0.01 g/m (0.005 gr/ft ).
After the ESP, 85 percent of the flue gas is routed to the
FGD system. When coal with a maximum sulfur content of 1.3
percent is burned, approximately 1900 ppm (dry) SO2 enters the
FGD system. The flue gas contacts the scrubbing solution in a
series of vertical spray zones. The scrubbing solution contains
solubilized calcium oxide, magnesium oxide, and sodium oxide from
the fly ash collected in the ESP's; lime is added to the solution
to increase SO2 removal and pH control. When the SO2 inlet
concentration is 1900 ppm, the absorbers remove approximately 85
percent of the SO2 from the scrubbed gas. Because 85 percent of
the flue gas stream is treated, the total removal efficiency is
approximately 70 percent and the overall outlet SO2 concentration
is 535 ppm (dry).
When the boiler fires coal with a maximum sulfur content of
1.3 percent, the FGD system operates at an L/G ratio of 11
3 33
liters/m (80 gal/10 ft ). The pH of the absorbent ranges from
6.4 to 6.5, and the level of suspended solids is 12 percent.
Lime is added at 3.6 Mg/h (4 tons/h).
After passing the spray zone of the absorbers, the scrubbed
gases enter Chevron-type mist eliminators. A wash tray beneath
each mist eliminator dilutes any entrainment to prevent plugging,
scaling, and eventual restriction of gas flow. The wash trays
are joined to separate recirculation loops and tanks to ensure
that the wash water remains low in suspended solids and dissolved
salts. To prevent plugging, sprays irrigate the mist eliminators
with a mixture of makeup water and clarified liquor from each
tray loop. Also, sprays flood the underside of the trays and the
areas between each tray.
The scrubbed gas stream is reheated by mixing it with the
bypassed flue gas at 171°C (340°F). This raises the temperature
A-35
-------
of the saturated scrubbed gas from 57° to 74°C (135° to 165°F);
and the gases are discharged to the atmosphere through the stack.
The spent scrubbing solution is discharged to a recycle tank
at the bottom of the spray towers, where the chemical reactions
are completed. Flue gas cleaning wastes, including unused rea-
gent, sulfite/sulfate salts, and fly ash, are conveyed to a
thickener, where the waste solids settle out. The overflow is
returned to the process and used foi preparation of the fly-ash
slurry. The thickener underflow has a maximum solids content of
40 percent and is piped to a vacuum thickener building, which
houses two rotary-drum vacuum filters for additional dewatering.
Filter cake with a solids content of 70 percent is produced at 45
Mg/h (50 tons/h) and transported by conveyor to three 32-Mg (35-
ton) off-the-road trucks, which haul it to a mine landfill for
final disposal. Water recovered in the vacuum filters is also
used for preparation of fly-ash slurry. During emergency periods
of mechanical malfunction or high flowrates, sludge from the
vacuum filters can be piled outside the vacuum filter building
for handling by front-end loaders and off-the-road trucks.
Additional system design data are presented in Table A-10.
PERFORMANCE HISTORY50'51
FGD operations at Milton R. Young 2 began in September 1977
and were initially intermittent, mainly because of problems
arising from the severe winter weather. The freezing and rupture
of many lines caused an outage during the first part of December
for the installation of heat tracing in the liquid circuit.
Other initial problems were reported with the guillotine gas
dampers and flow meters.
An emergency shutdown that occurred on December 5, 1977, as
a result of turbine bearing damage, made it necessary to postpone
compliance testing scheduled during late December or early Jan-
uary. The unit came back online on February 21, 1978; it was
discovered that a forced-draft fan in the FGD system had an oil
A-36
-------
TABLE A-10. FGD SYSTEM DATA FOR MILTON R. YOUNG 2,
CENTER, NORTH DAKOTA
Unit rating (gross), MW
(net) , MW
Fuel
Average fuel characteristics:
Heating value, kJ/kg (Btu/lb)
Ash, %
Moisture, %
Sulfur, %
Chloride, %
FGD process
FGD system supplier
Application
Status
Startup date:
Initial
Commercial
Design removal efficiency:
Particulate, %
Sulfur dioxide, %
Actual removal efficiency:
Particulate, %
Sulfur dioxide, %
Water loop
Total water makeup,
liters/s per net MW
(gal/min per net MW)
Sludge disposal
477
408
Lignite
15,000 (6,500)
8.0
38
0.7
Lime with alkaline fly ash
Arthur D. Little/Combustion
Equipment Associates
New
Operational
June 1977
June 1978
99.6
75.0
Open
0.1
1.56
Unstabilized sludge disposed
of in a landfill
A-37
-------
leak and a shaft alignment problem. The fan was taken offline
for repairs, and one module of the FGD system was taken out of
service. The repaired fan was placed back in service on April
10.
Other minor problems were encountered through September
1978. Large particles from the vacuum filter caused some rubber
lining downstream to peel; modifications to correct this problem
are being studied. Also, the guillotine damper chain drives were
discovered to be underdesigned and had to be replaced.
Compliance testing took place at the unit during the week of
June 5, 1978. Official results of the testing are not yet avail-
able.
A-38
-------
MONTANA POWER
COLSTRIP 1 AND 2
PERFORMANCE HISTORY52
The overall performance of the FGD systems at Colstrip 1 and
2 has been good, but some problems have been encountered. The
monitors for SO2 on the inlets and outlets of the scrubber
modules have exhibited erratic behavior. A corrective program,
however, has improved monitor data. The inline pH probes were
eroding and loosing sensitivity because of deposits on the
elements. Slurry density monitors were also erratic, and a test
program was undertaken to isolate these devices from vibration.
Both slurry density and pH, key operating parameters, had to be
taken manually.
Because of deposits (not calcium sulfate or calcium sulfite
scale) at the entrance to the scrubber modules, system avail-
ability was reduced significantly. Tests showed that there was
maldistribution of the dust entering each module and that gas
velocity varied enormously from one side to the other of each
entrance. A model study of the liquid flow on the tangential
shelf above the venturi indicated that gas flow turning vanes in
the duct elbow above each scrubber module and liquid guiding
vanes and baffles on the tangential shelf would be required to
reduce deposit buildup at each scrubber entrance. One module was
modified with such encouraging results that modifications were
made to the remaining five modules. Actual operating experience
from the modules equipped has shown a significant reduction in
cleaning time and has increased the time increment between
inspection and cleanings.
A-39
-------
The supplementary alkali feed system was difficult to use
because of line plugging and equipment failure. The lime system
•»
was modified to a simpler design that would be less prone to
plugging. Pinch valves were placed next to the scrubber on each
feed line so that solids accumulation is automatically cleared as
these valves are operated. Although the alkaline ash captured by
the scrubber has generally kept scrubber pH within the control
range, it was hoped that better control of the lime system would
provide more constant pH and allow optimal SO~ removal.
Failures of the protective flaked-glass lining on the carbon
steel vessel walls and ducts occurred. These failures were
evident prior to a major temperature excursion at Colstrip 1 in
October 1976. The excursion followed a complete station power
blackout and the failure of the emergency quench water supply
system for the scrubbers to operate. The flaked-glass lining and
the plastic mist eliminators were seriously damaged at Colstrip 1
and accounted for low scrubber availability during November and
December 1976.
The motor of an induced-draft fan failed during November and
December 1976 independently of the temperature excursion. This
failure also reduced availability during November and December.
Problems with other induced-draft fan motors early in 1977 also
reduced scrubber availabilities.
The availability of 63 percent for Colstrip 2 during May
1977 reflects fan motor repairs and the modifications to reduce
deposits at the scrubber entrances.
Inspection of an induced-draft fan motor during the spring
of 1977 revealed cracks in the center plate next to the blades.
All other fans were then cleaned and inspected. When cracks were
found or when it was suspected that cracks might form, the faulty
portions of fans were ground and welded; and stiffener plates
were added. Availabilities during the summer of 1977 reflected
this repair work.
Quick clean basket strainers were added to the suction
piping of the two main recycle pumps of each vessel to replace
A-40
-------
the startup strainers contained in a pipe spool. This replace-
ment has allowed deposits and foreign materials to be removed
before they clog spray nozzles and has also decreased maintenance
and downtime on the vessels.
A program was undertaken to evaluate abrasion-resistant
protective lining materials that could be used where the flue gas
makes a 180-degree turn after the venturi downcomer and passes
through the absorption sprays. The absorption spray slurry
erodes the flaked-glass lining/ which has been replaced at least
once in each module since startup.
A program for evaluating corrosion-, temperature-, and
abrasion-resistant materials in other areas of the scrubber was
also undertaken. Small test patches and full linings were tried
on several scrubber modules.
Plugging of spray nozzles under the wash trays on scrubber
modules of both units occurred during May 1977 and forced the
plants to shut down. The major cause of the problem was piping
that had broken in the wash tray pond and allowed material from
the pond to be ingested. A parallel pipeline a'nd pumping system
was installed to bring the units back on line, and repairs were
made to the original piping.
The scrubber ponds are reclaimed by dredging sludge to a
disposal pond approximately 3 miles from the plant site. Liquid
from the disposal pond is returned to the scrubber ponds when the
dredge is not operating. The scrubber sludge deposits in deltas
in the ponds and contains 55 to 60 percent solids. The material
hardens so well that the dredge has had difficulty in reslurrying
the sludge. Because of the residual alkalinity, the deposited
sludge becomes like cement. Improved cutters for the dredge and
a delumper ahead of the dredge pump are expected to increase the
solids handling capacity- The wash tray pond filled with solids
more rapidly than had been expected, and the removal of these
solids had been a problem. Each pond at the plant has its own
chemical identity based on water balances. Because a dredge
cannot- be used to empty the wash tray pond without adversely
A-41
-------
affecting the plant operation or pond water balances, a less
efficient clamshell operation was put into use. Plans are being
developed for the addition of another wash tray pond, so that one
may be dewatered and cleaned while the other is in operation.
Many emission tests have been conducted on the Colstrip
units. During the spring of 1977, EPA compliance tests for
particulates, SO0, and NO were completed on both plants; and
^ X
emission monitor certification testing was conducted after the
compliance tests. Table A-ll gives background information for
the tests of S02 emissions and Tables A-12 and A-13 show the
results of these tests. Data from full-scale operations agrees
well with pilot plant data and shows that SO2 emissions are well
below the maximum allowed by the guarantee and Federal standards.
Although efficiencies cannot be accurately calculated because
scrubber inlet data have not been taken, an SO2 removal effi-
ciency from 70 to 75 percent and a particulate removal efficiency
of 99.5 percent, predicted from pilot plant information, are
apparently being achieved.
A-42
-------
TABLE A-ll. BACKGROUND INFORMATION FOR SO.
EMISSIONS AT COLSTRIP 1 AND 253
Emissions of SO2
kg/h (Ib/h)
ng/J (lb/106 Btu)
ppm
Maximum emissions allowed
by NSPS for a 358-MW unit
Scrubber guarantee for a
358-MW unit
Projected emissions from
pilot plant for a 358-MW
unit firing:
Coal with a sulfur
content of 0.78 percent
and an ash content of
8.19 percent
Coal with a sulfur
content of 1.0 percent
and an ash content of
12.85 percent
1844 (4063)
1537 (3386)
633 (1394)
940 (2071)
516 (1.2)
430 (1.0)
510
425
176 (0.41)
185
262 (0.61)
260
A-4,3
-------
TABLE A-12.
TESTS OF S02 EMISSIONS AT COLSTRIP 1
53
Date
2/76
4/76
7/76
9/76
12/76
1/77
5/77
6/77
Coal as received
Gross MW
353a
210b
184b
186b
223b
230b
33la
340a
Heating value,
kJ/kg (Btu/lb)
20,092 (8638)
20,611 (8861)
20,485 (8807)
20,080 (8633)
19,524 (8394)
20,280 (8719)
20,587 (8851)
20,673 (8888)
Ash content,
%
9.03
7.79
8.49
7.93
8.54
8.17
8.41
8.96
Sulfur
content, %
0.83
0.71
0.64
0.62
0.94
0.77
0.61
0.91
Emissions of SO2
kg/h
db/h)
720 (1587)
207 (456)
117 (258)
125 (275)
408 (898)
350 (770)
251 (552)
528 (1164)
ng/J
(lb/106 Btu)
206 (0.48)
99 (0.23)
64 (0.15)
64 (0.15)
185 (0.43)
150 (0.35)
69 (0.16)
140 (0.325)
ppm
197
87
53
56
154
133
72
122
Three scrubber modules were online.
Two scrubber modules were online.
-------
TABLE A-13.
TESTS OF S02 EMISSIONS AT COLSTRIP 2
53
Date
10/76
11/76
12/76
3/77
6/77
Coal as received
Gross MWa
331
327
324
335
305
Heating value,
kJ/kg (Btu/lb)
19,464 (8368)
19,734 (8484)
20,213 (8690)
20,327 (8739)
20,709 (8929)
Ash content,
%
7.96
7.86
7.87
7.96
8.48
Sulfur
content, %
0.56
0.59
0.64
0.63
0.72
Emissions of SO,,
kg/h
db/h)
594 (1309)
316 (696)
354 (780)
301 (664)
271 (597)
ng/J
(lb/106 Btu)
181 (0.42)
99 (0.23)
107 (0.25)
86 (0.20)
81 (0.19)
ppm
178
83
98
84
67
I
.£.
Ul
Three scrubber modules were online.
-------
PENNSYLVANIA POWER
BRUCE MANSFIELD 1 AND 2
POLLUTION CONTROL54"56
The emission control system for Bruce Mansfield 2 was manu-
factured by Chemico Air Pollution Control Company. It is a
venturi wet-scrubbing system identical in design to the system
installed on Bruce Mansfield 1. Table A-14 presents additional
system design information for Bruce Mansfield 1 and 2.
PERFORMANCE HISTORY55'56
^ Initial operation of the emission control system at Bruce
Mansfield 1 began on December 11, 1975; and commercial operation,
on July 1, 1976. Initial operation of the emission control
system at Bruce Mansfield 2 started in July 1977; and commercial
operation, on October 1, 1977.
Many major design, mechanical, and chemical problems accom-
panied the initial and subsequent operation of the emission
control systems. These problems included corrosion, scaling,
mist eliminator inefficiency, reheater vibration, pH control
failures, stack liner failures, and induced-draft fan problems.
Several of these problem areas and the resulting system modifi-
cations were discussed previously.
Faulty pH monitoring produced many chemical problems.
Difficulties with flow sampling and glass probe breakage forced
unreliable manual control of pH during much of the initial oper-
ation stage; and this manual control caused such problems as
scaling, plugging, and acid corrosion. Since the relocation of
pH monitors in the recirculation circuit and the modification of
sampling procedures, results have been excellent, with the pH
A-46
-------
TABLE A-14. FGD SYSTEM DATA FOR BRUCE MANSFIELD 1 AND 2,
SHIPPINGPORT, PENNSYLVANIA
Unit rating (gross), MW
(net) , MW
Fuel
Average fuel characteristics:
Heating value, kJ/kg (Btu/lb)
Ash, %
Moisture, %
Sulfur, %
Chloride, %
FGD process
FGD system supplier
Application
Status
Startup date:
Initial
Commercial
Design removal efficiency:
Particulate, %
Sulfur dioxide, %
Actual removal efficiency:
Particulate, %
Sulfur dioxide, %
Water loop
Total water makeup, liters/s per MW
(gal/min per MW)
Sludge disposal
917
825
Bituminuous coal
26,700 (11,500)
12.9
7.0
3.0
Lime
Chemico
New
Operational
December 1975
June 1976
99.8
92.1
Open
Stabilized sludge disposed
of in a landfill
A-47
-------
remaining between 6.8 and 7.2. Also, the modules have operated
without any substantial deposits of hard scale (gypsum) or plug-
ging problems, which often hurt mist eliminator performance.
Continuous SO2 analyzers (DuPont 460 Photometric) were
provided upstream and downstream of the modules, but have not
operated well.
Although many problems have been resolved, difficulties with
the stack liners and induced-draft fans have severely limited
system availability. Stack liner failures have required Bruce
Mansfield 1 and 2 to operate at half capacity for long periods.
After investigating new stack liner materials, Pennsylvania Power
decided to use CSL 2000 (Pullman/Kellogg). The IB and 2B flues
have been relined; and the 2A flue is being relined. It is hoped
that the 1A flue, which is still lined with flake glass and
showing signs of failure, will last another year.
To avoid limited availability v.'hen stock problems occur, the
utility is manifolding the four flues together, so that one can
be bypassed for repair without necessitating a load reduction or
shutdown.
The scrubbing systems were originally designed so that five
of the six scrubbing trains installed on each unit could handle
total boiler gas flow at a slightly reduced particulate and S02
removal efficiency. Operation has shown, however, that all six
trains are necessary when each unit is operating at full load.
Thus, a train cannot be taken out of service without load cutbacks,
C-7 CO
FUTURE PLANS '
Bruce Mansfield 3, a 917-MW (gross) coal-fired unit, is
currently under construction alongside Bruce Mansfield 1 and 2.
Commercial operation is scheduled for April 1980. The emission
control system for Bruce Mansfield 3 is designed and supplied by
Pullman Kellogg and is dramatically different from those on Bruce
Mansfield 1 and 2. Following recent trends, ESP's will provide
particulate removal upstream of a lime-based, spray-tower FGD
A-48
-------
system. There will be dry forced-draft fans upstream of the
spray towers; and high component redundancy will increase overall
system reliability. Redundant components will include a spare
booster fan, precipitator, and spray tower, as well as one spare
stage per spray tower. The ESP's are designed to remove 95 per-
cent of particulates; and the spray towers, to collect additional
particulates simultaneously with SO . The low-efficiency ESP's
offer substantial capital savings and permit a simpler design
than high efficiency ones. Another important feature of the
Bruce Mansfield 3 system design is the plan to line the 183-m
(600-ft) stack with an Inconel-625 alloy. The choice of this
exotic alloy stems from the nearly disastrous lining failures at
Bruce Mansfield 1 and 2. Flue gas cleaning wastes from Bruce
Mansfield 3 will be disposed of in the existing waste disposal
system.
Table A-15 summarizes FGD system data for Bruce Mansfield 3.
A-49
-------
TABLE A-15. SUMMARY OF EMISSION CONTROL SYSTEM
AT BRUCE MANSFIELD 358
Rating (gross), MW
Fuel
Heating value, kJ/kg (Btu/lb)
Sulfur content, percent
Ash content, percent
Particulate emission rate,
ng/J (lb/106 Btu)
Sulfur dioxide emission rate,
ng/J (lb/106 Btu)
Emission controls:
Particulate
Sulfur dioxide
FGD process
FGD system supplier
Absorber type
Number of absorbers
Number of ESP's
Gas capacity, m /s (ft /min)
Pressure drop, kPa (in. H20)
Gas reheat:
Type
Temperature change, °C (°F)
Gas bypass capability
Startup date (commercial)
917
Coal
27,700 (11,900)
2.6 - 4.75
9.5 - 19.7
32 (0.075)
258 (0.6)
Electrostatic precipitators and
spray towers absorbers
Spray tower absorbers
Lime
Pullman Kellogg
Spray tower
5 (1 spare)
4 (1 spare)
1110 (2355)
7.0 (28.0)
Oil-fired
22 (40)
No
April 1980
.A-50
-------
TENNESSEE VALLEY AUTHORITY
SHAWNEE 10A AND 10B
PERFORMANCE HISTORY59
During September 1977, a 10-day test was conducted on the
TCA system with an automatic limestone feed control that operated
at stoichiometric ratios determined by the flue gas flow rate and
the inlet S02 concentration. The system operated well.
In mid-November 1977, a 1-month test of the venturi/spray
tower system's reliability was started. There was forced oxida-
tion at two scrubber stages, and limestone slurry with a high
percentage of fly ash was used. The flue gas rate to the scrub-
ber system was varied according to the boiler load, which ranged
between 100 and 155 MW. The test ended after 965 operating hours
and demonstrated the reliability of the system. With the inlet
SO2 concentration varying from 2500 to 3400 ppm, the average SO
removal efficiency for the entire run was 86 percent.
Another 1-month reliability run on the venturi/spray tower
system started in mid-December. Lime scrubbing was used in this
test. Except for the higher pH at the spray tower inlet and the
higher alkali utilization of a lime system, the operating condi-
tions and test results were similar to those of the limestone
reliability run.
Two tests were conducted on the TCA system. In these tests,
the three beds of nitrile foam spheres were replaced by 23 layers
of Ceilcote plates, which altogether were 117 cm (46 in.) deep.
At the full gas flow rate of 14 m /g (30,000 acfm), the plates
removed slightly less S02 than the beds of spheres, each of which
was 12.7 cm (5 in.) deep and divided into four grids. However,
at 8 itr/g (18,000 acfm), the SO- removal efficiency of the
A-51
-------
C^ilcote plates was a few percentage points higher than that of
the foam spheres.
Forced oxidation tests were conducted on the TCA system with
an air sparger. Limestone slurry with a high percentage of fly
ash was used. Air sparging and limestone addition occurred
sometimes in the same tank and sometimes in separate tanks.
Almost total sulfite oxidization was achieved in both configura-
tions when the stoichiometric ratio of oxygen to absorbed SO was
1.8 and the pH of the oxidation tank was between 5.4 and 5.7.
The depth of the slurry oxidation tank was 5 m (18 ft).
During February 1978/ the effect of the slurry level in the
air sparged oxidation tank was investigated. Lime slurry with a
high percentage of fly ash was used. Almost complete sulfite
oxidation (98 percent) was achieved when the slurry was 4 m (14
ft) and 5 m (18 ft) deep in the oxidation tank and the stoichio-
metric ratio of oxygen to absorbed SO was 1.8. An air stoichio-
metric ratio of nearly 3.8 was needed to produce comparable
oxidation when the slurry level dropped to 3 m (10 ft).
New tests started on March 1 on the venturi/spray tower
system. To improve SG>2 removal efficiency, MgO was added to the
spray tower slurry loop of the two-loop operation, with forced
oxidation in the venturi loop. Limestone slurry with a high
percentage of Ify ash was used. Because of the shortage of coal
caused by the miners' strike, coals from different sources were
burned in the boiler. As a result, inlet SO2 concentration
fluctuated at much as tenfold (from 350 to 3500 ppm). The MgO
tests continued through early May 1978. With MgO addition, over-
all S02 removal was 96 percent at an inlet SO2 concentration of
2300 ppm; without MgO addition, it was 86 percent at an inlet S02
concentration of 1600 ppm. The S02 removal by venturi alone was
30 percent with MgO addition, about the same as the case without
MgO addition. The MgO tests showed that almost complete sulfite
oxidation could be achieved with the stoichiometric ratio of
oxygen to absorbed SO2 was as low as 1.3.
A-52
-------
Forced oxidation was also conducted on the limestone slurry
bleed stream from the venturi/spray tower system. A single
effluent hold tank was used for both venturi and spray tower.
Enough MgO was added to the effluent hold tank to maintain mag-
nesium ion (Mg++) concentration of 5000 ppm. A slurry stream was
taken from the scrubber downcomer and sent to an oxidation tank
into which air was sparged. About 2 liters/s (30 gal/min) of
slurry was recycled from the oxidation tank to the effluent hold
tank to control pH in the oxidation tank and provide gypsum seeds
in the scrubber slurry. Final system bleed was withdrawn from
the oxidation tank. At an average oxidation tank pH of 6, sul-
fite oxidation averaged 98 percent. Filter cake solids content
was 85 percent, similar to that obtained when two scrubber loops
were operated. However, the settling rate of slurry solids was
only 0.4 cm/min (0.16 in./min), compared to approximately 0.8
cm/min (0.31 in./min) for the two-loop operation. The settling
rate for unoxidated slurry containing Mg++ normally did not
exceed 0.1 cm/min (0.04 in./min) when the solids content of the
filter cake ranged from 50 to 60 percent.
In March 1978, the TCA system was operated with both lime
and limestone slurries to which MgO was added. Flue gas with a
high percentage of fly ash was scrubbed. These further tests
were conducted to resolve some of the inconsistent results ob-
tained during lime/MgO and limestone/MgO tests from April through
November 1976. It was suspected that air had leaked through the
scrubber downcomer in some of the earlier runs and caused the
high sulfite oxidation and gypsum saturation mentioned above.
For the tests in 1976, the inlet SO2 concentrations had
usually been high (greater than 3000 ppm), whereas the inlet SO2
concentrations for recent runs had been about 2500 ppm. At the
higher inlet SC>2 concentration and -S02 make-per-pass, the sulfite
oxidation level had determined whether the slurry was unsaturated
or supersaturated with gypsum. Severe scaling occurred during
operation in the gypsum-saturated mode. Although recent runs
have been conducted in the gypsum-saturated mode, no significant
A-53
-------
scaling has occurred because of the lower inlet SO2 concentration
and lower SO~ make-per-pass.
Forced oxidation of the bleed stream from the venturi/spray
tower system continued through mid-June 1978. There were two
test runs in which MgO was added to maintain an effective Mg++
concentration of 5000 ppm in the scrubber slurry. No stream was
recycled to the effluent hold tank from the oxidation tank, in
which the slurry level was ekpt at 5 m (18 ft). Sulfite oxida-
tion averaged 96 percent or higher when the air stoichiometric
ratio of oxygen to absorbed SO2 was 1.6. The pH of the oxidation
tank varied from 5.4 to 5.6, whereas that of the effluent hold
tank was between 5.2 and 5.4. The solids content of the filter
cake was 85 percent, and the solids settling rate was between 0.4
and 0.5 cm/min (0.16 and 0.2 in./min).
Both scrubber systems were shut down for 2 weeks beginning
on June 19, 1978, because of a scheduled boiler outage, during
which the flue gas duct from the 244-m (800-ft) stack was rerouted
to the small stack at Shawnee 11.
In July 1978, new lime and limestone tests began. Adipic
acid, an organic acid pH buffer, was added to the slurries of
both scrubbers to improve SO2 removal efficiency after some
initial runs were conducted without adipic acid to establish the
base cases for S02 removal with lime and limestone scrubbing.
The venturi/spray tower system was operated with two loops, with
forced oxidation in the first loop; the TCA system was operated
with one loop, with no forced oxidation. Preliminary results
showed that S02 removal efficiency was consistently between 96
and 99 percent in the venturi/spray tower system when the adipic
acid concentration was approximately 1600 ppm in the venturi and
1400 ppm in the spray tower. This efficiency compared favorably
with the 66 percent removal for the base case lime run without
adipic acid produced an S02 removal efficiency of nearly 80 per-
cent, compared to 67 percent for the base case.
The testing with adipic acid continued through August and
September 1978. Both the venturi/spray tower and TCA systems
A-54
-------
used limestone slurry with a high percentage of fly ash. The TCA
system was operated without forced oxidation; the venturi/spray
tower system had forced oxidation in two loops. As in the lime
tests with adipic acid, significant improvement in SO_ removal
efficiency was observed. Under typical conditions, SO2 removal
higher than 90 percent could be achieved by the venturi/spray
tower system when adipic acid concentration was approximately
2100 ppm in the venturi and 1500 ppm in the spray tower. Without
adipic acid, the S02 removal efficiency was only 57 percent.
Sulfite oxidation and waste sludge dewatering properties did not
appear to be affected by adipic acid. In the TCA system, an
adipic acid concentration between 750 and 1500 ppm yielded SO
removal efficiencies higher than 90 percent, compared to 71
percent for a base case run without adipic acid. In both
scrubber systems, the slurry pH needs to be above 5.0, the upper
pH buffer point of adipic acid, for the additive to be fully
effective. Deterioration or decomposition of adipic acid
apparently takes place in the scrubber. The amount of adipic
acid added to the slurry was from two to three times more than
the amount discharged in the sludge.
A-55
-------
UTAH POWER AND LIGHT
HUNTINGTON 1
BACKGROUND INFORMATION60
The Huntington Power Station, owned and operated by Utah
Power and Light, is located in Price, Utah. This station oper-
ates two boilers with a combined electric generating capacity of
810 MW (net).
Huntington 1 is a pulverized-coal-fired boiler manufactured
by Combustion Engineering and placed in service early in 1978.
This unit is nominally rated at 405 MW, with a gross rating of
430 MW. The coal burned has an average sulfur content of 0.5
percent and ash content of 10.0 percent.
POLLUTION CONTROL61
The FGD system for Huntington was supplied by Chemico and
consists of four parallel spray towers utilizing a lime slurry
for SO- removal. Each module was designed for a gas flow of 833
m3/s (1,742,000 acfm) at 130°C (266°F) and has an L/G ratio of 6
3 33
liter/m (43 gal/10 ft ). The spray towers are constructed of
carbon steel lined with polyester flaked glass. Primary particu-
late control is provided by a cold-side ESP manufactured by
Buell/Envirotech. During operation a damper is used to bypass
from 10 to 20 percent of the flue gas. This bypass flue gas is
then used in conjunction with an inline steam reheat system to
increase the temperature of the existing flue gas between 25°C
and 28°C (45°F and 50°F). A horizontal chevron-type mist elimi-
nator constructed of polypropylene is downstream of each module.
From the mist eliminators the scrubbed gases are ducted to a
common 183-m (600-ft) acid-brick-lined stack.
A-56
-------
The spent absorbent from each module is pumped to a flat
bottom thickener that is 18 m (60 ft) in diameter and constructed
of carbon steel with a flaked-glass lining. The underflow has a
solids content between 25 and 30 percent and is pumped to a
vacuum filter, where it is dewatered to approximately 60 percent
solids. This product is mixed with fly ash and trucked to an
onsite landfill for final disposal.
The FGD system is designed to operate in a closed water loop
with thickener overflow recycled to the modules. To compensate
for water losses from evaporation and entrainment, additional
water is added to the system at 19 liters/s (300 gal/min).
Design information is presented in Table A-16.
PERFORMANCE HISTORY62
Initial FGD operation began at Huntington 1 on May 10, 1978.
Since startup only minor instrumentation problems have been ex-
perienced. From June through September 1978, the FGD system
logged 2224 operating hours; and the boiler, 2491 operating
hours. Thus, average FGD system operability was 89 percent.
Although operability began at 65 percent in June, it improved to
98 percent in July and was 100 percent for August and September.
FUTURE PLANS63
Utah Power and Light has awarded contracts to Chemico to
supply lime FGD systems for two new 400-MW units, Emery 1 and 2.
The systems will be designed with an S02 removal efficiency of 80
percent for low-sulfur Utah coal. Primary particulate control is
to be provided by an ESP upstream of each scrubber. The sludge
will be stabilized with fly ash and disposed of on the plant
site. Operation is expected to begin in December 1978 at Emery
1. The scheduled startup date for Emery 2 is June 1980.
A-57
-------
TABLE A-16. FGD SYSTEMS DATA FOR HUNTINGTON 1,
PRICE, UTAH
Unit rating (gross), MW
(net), MW
Fuel
Average fuel characteristics:
Heating value, kJ/kg (Btu/lb)
Ash, %
Moisture, %
Sulfur, %
Chloride, %
FGD process
FGD system supplier
Application
Status
Startup date:
Initial
Commercial
Design removal efficiency:
Particulate, %
Sulfur dioxide, %
Actual removal efficiency:
Particulate, %
Sulfur dioxide, %
Water loop
Total water makeup,
liters/s per net MW
(gal/min per .net MW)
Sludge disposal
430
405
Coal
10
0.5
Lime
Chemico
New
Operational
May 1978
99.5
80.0
Closed
0.05
0.74
Stabilized sludge disposed
of in an onsite landfill
A-58
-------
APPENDIX A
REFERENCES
1. Devitt, T., et al. Flue Gas Desuifurization System Capa-
bilities for Coal-Fired Steam Generators. EPA-600-7-78-032b.
PEDCo Environmental, Cincinnati, Ohio. March 1978. pp. A22,
A23.
2-. -Melia, M., et al. EPA Utility FGD Survey August-September
1978. Preliminary Report. Prepared for the U.S. Environ-
mental Protection Agency under Contract 68-02-2603. PEDCo
Environmental, Ci-. cinnati, Ohio. November 1978. p. 5.
3. Ibid.
4. Hargrove, O.W., and F.B. Meserole. Phase I SO2 Test Program
Results: Scrubber 2A, Unit No. 2, Four Corners Power Plant,
Fruitland, New Mexico. Prepared for the Arizona Public Service
Co. Radian Corp., Austin, Texas. May 2, 1977. pp. 4-8, 32-45,
5. Op. cit. No. 1. p. 15.
6. PEDCo in-house files.
7. Laseke, B.A., Jr. EPA Utility FGD Survey: December 1977-
January 1978. EPA-600-7-78-051a. PEDCo Environmental,
Cincinnati, Ohio. March 1978. p. 32.
8. Op. cit. No. 2. pp. 27-29.
9. Op. cit. No. 2. pp. 7-8.
10. O'Hara, R., and R.L. Nelson. Operating Experience at the
Phillips and Elrama Flue Gas Desulfurization Facility. In:
Proceedings of the Second Pacific Chemical Engineering
Congress (PAChEC '77). American Institute of Chemical
Engineers, New York, New York. 1977. pp. 308-309.
11. Smith, M.P. (PEDCo Environmental, Inc.) Telephone Conver-
sation with R. O'Hara (Duguesne Light Company). November
27, 1978.
A-59
-------
12. Op. cit. No. ~ ,. pp. 36-37.
13. Op. cit. No. Z. pp. 30-31.
14. Op. cit. No. l%v pp. 308-309.
15. Op. cit. No. '. p. 42
16. Op. cit. No. 2. pp. 32-33.
17. Isaacs, G.A. , ,v.id F.K. Zado. Survey of Flue Gas Desulfur-
ization Systems: Hawthorn Station, Kansas City Power and
Light Co. EFA-i-,50-2-75-047h. PEDCo Environmental, Inc.,
Cincinnati, Ohio. September 1975. p. 2-1.
18. Op. cit. No. 1-. pp. 2-1, 2-2.
19. Op. cit. No. T. p. 10.
20. Ibid. pp. 45-5>4.
21. Op. cit. No. I. p. 9-2.
22. Beard, J.B. Scrubber Experience at the Kentucky Utilities
Company Green >iver Power Station. In: Proceedings of the
Symposium on ~1 ;:e Gas Desulfurization, Hollywood, Florida,
November 1977, .Volume 1). EPA-600-78-058a. March 1978.
pp. 248-249-
23. Laseke, B.A., ." v. Survey of Flue Gas Desulfurization
Systems: Greer River Station, Kentucky Utilities. EPA-600-
7-78-048e. PZ?v%o Environmental, Inc., Cincinnati, Ohio.
March 1978. p. 24.
24. Op. cit. No. 7. pp. 71-72.
25. Op. cit. No. 2. pp. 42-43.
26. Campbell, R.L. Emission Testing Report, Kentucky Utilities
Green River Starion, Central City, Kentucky. PEDCo Environ-
mental, Inc., C\ncinnati, Ohio. October 1978. p. 2.
27. Hartman, J.S. .PEDCo Environmental, Inc.) Internal Memo
about Telephone conversation with R.P. Van Ness (Louisville
Gas and Electr.,- Company). July 15, 1977.
28. Yerino, L.V. .TEDCo Environmental, Inc.) Internal Memo
about inspectio^ of Louisville Gas and Electric Company's
Cane Rune Station: Unit 4 Sulfur Dioxide Scrubber Status.
April 25, 1977.
A-60
-------
29. Laseke, B.A., Jr. (PEDCo Environmental, Inc.) Notes on
trip to Louisville Gas and Electric Company's Cane Run
Station. August 17, 1976.
30. Op. cit. No. 7. pp. 75-76.
31. Op. cit. No. 2. p. 44.
32. Op. cit. No. 2. p. 11.
33. Op. cit. No. 27.
34. Laseke, B.A., Jr. (PEDCo Environmental, Inc.) Notes on
trip to Louisville Gas and Electric Company's Cane Run
Station. March 31, 1978.
35. Op. cit. No. 2. p. 45.
36. Op. cit. No. 2. p. 11.
37. PEDCo in-house files.
38. Op. cit. No. 2. p. 11.
39. Ibid.
40. PEDCo in-house files.
41. Van Ness, R.P. Louisville Gas and Electric Company Scrubber
Experiences and Plans. In: Proceedings of the Symposium on
Flue Gas Desulfurization, Hollywood, Florida, November 1977
(Volume 1). EPA-600/7-78-058a. March 1978. pp. 235-245.
42. Op. cit. No. 2. p. 46.
43. Ibid. p. 11.
44. Corbett, W.E., and O.W. Hargrove, Jr. Results of EPA
Sponsored Characterization Tests of Louisville Gas and
Electric Company's Paddy's Run Flue Gas Desulfurization
System. In: Proceedings of the Second Pacific Chemical
Engineering Congress (PAChE 1977). American Institute of
Chemical Engineers, New York, New York. 1977. pp. 336-337.
45. Op. cit. No. 41.
46. Op. cit. No. 7. p. 80.
47- Op. cit. No. 2. p. 47.
48. Op. cit. No. 7. p. 82.
A-61
-------
49. Ibid.
50. Ibid. p. 83.
51. Ibid. p. 48.
52. Berube, D.T., and C.D. Grumm. Status and Performance of the
Montana Power Company's Flue Gas Desulfurization System.
In: Proceedings of the Symposium on Flue Gas Desulfuriza-
tion, Hollywood, Florida, November 1977 (Volume 1). EPA-
600-7-058a. March 1978. pp. 277-291.
53. Ibid. p. 290.
54. Op. cit. No. 7. p. 121.
55. Laseke, B.A., Jr. Survey of Flue Gas Desulfurization
Systems: Bruce Mansfield Station, Pennsylvania Power
Company. Preliminary Report. Prepared for the U.S.
Environmental Protection Agency under Contract 68-02-2603.
PEDCo Environmental, Inc., Cincinnati, Ohio. November 1978.
pp. 50-55.
56. Laseke, B.A., Jr. (PEDCo Environmental, Inc.) Telephone
conversation with R. Forsythe (Pennsylvania Power Company).
December 1, 1978.
57. Op. cit. No. 55. p. 55.
58. Ibid. p. 56.
59 PEDCo in-house files.
60. Ibid.
61. Ibid.
62. Op. cit. No. 2. p. 82.
63. Ibid. p. 21.
A-62
-------
APPENDIX B
DOMESTIC LIMESTONE SLURRY
FGD SCRUBBING SYSTEMS
B-l
-------
This appendix summarizes information on limestone slurry FGD
scrubbing systems presented in Section 2.2 and describes the
other systems in this country that use limestone or limestone
with alkaline fly ash.
The following FGD systems are described:
0 Alabama Electric Cooperative - Tombigbee 2
0 Arizona Electric Power Cooperative - Apache 2
0 Arizona Public Service - Cholla 1 and 2
0 Central Illinois Light Company - Duck Creek 1
0 Gulf Power Company - Sholz 1 and 2
0 Indianapolis Power and Light Company - Petersburg 3
0 Kansas City Power and Light - La Cygne 1
0 Kansas Power and Light Company - Jeffrey 1
0 Kansas Power and Light Company - Lawrence 4 and 5
0 Northern States Power Company - Sherburne 1 and 2
0 South Carolina Public Service Authority - Winyah 2
0 Southern Mississippi Electric - R.D. Morrow 1
0 Springfield City Utilities - Southwest 1
0 Tennessee Valley Authority - Widows Creek 8
0 Texas Utilities - Martin Lake 1 and 2
0 Texas Utilities - Monticello 3
The EPA test facility at Shawnee uses both lime and lime-
stone. It is described in Appendix A.
B-2
-------
ALABAMA ELECTRIC COOPERATIVE
TOMBIGBEE 2
BACKGROUND INFORMATION '2
Tombigbee Power Station of Alabama Electric Cooperative is a
three-unit station near Leroy, Alabama. Tombigbee 1 is an exist-
ing 240-MW unit, and Tombigbee 2 and 3 are identical 255-MW,
(gross) units. Tombigbee 3 is under construction.
The electric power generating facilities at Tombigbee 2
consist of a Riley Stoker pulverized-coal-fired steam generator
with a 225-MW (net) turbine generator. Tombigbee 2 began initial
operation in September 1978.
The coal burned at Tombigbee 2 is a bituminous low-to-
medium-sulfur (from 0.8 percent to 1.5 percent sulfur) coal, its
heating value ranges from 25,600 to 27,900 kJ/kg (11,000 to
12,000 Btu/ Ib); and its ash content, from 10 to 16 percent.
POLLUTION CONTROL
A Research Cottrell electrostatic precipitator (ESP), de-
signed to be 99.3 percent efficient, provides primary particulate
control for Tombigbee 2. Sulfur dioxide (SO.) is controlled with
a limestone slurry FGD system supplied by Peabody Process Sys-
tems. The FGD system consists of two parallel spray tower ab-
sorbers that scrub 70 percent of the boiler flue gas with a
guaranteed S02 removal efficiency of 85 percent. The design also
includes a single-stage chevron mist eliminator system and a flue
gas bypass reheat system that allows 30 percent of the hot flue
gas to bypass the spray towers and rejoin the cleaned saturated
gas in a single duct to the stack. The overall SO2 removal
efficiency at Tombigbee 2 is designed to be 59.5 percent.
B-3
-------
Spent scrubbing slurry containing 8 percent solids is piped
to a clay-lined pond at a rate of 38 liters/s (600 gal/min).
Table B-l summarizes the FGD data for Tombigbee 2.
PROCESS DESCRIPTION3
Flue gas exits the boiler economizer at 396°C (745°F) and
enters a four-field hot ESP with an efficiency of 99.3 percent.
The cleaned gas enters the boiler air preheater at 385°C (725°F)
through two parallel ducts from the ESP. Each gas stream is
drawn through an induced-draft fan that forces the gas into a
common header feeding the SO? absorber towers and the bypass duct
at 144°C (291°F). Of the 450 m3/s (953,000 acfm) of gas entering
the header, 30 percent bypasses the absorbers; and 35 percent
passes through each tower.
Approximately 157 m /s (334,000 acfm) of gas is fed into the
base of each spray tower. The flue gas turns 90 degrees and
rises through trays as limestone slurry is sprayed down, counter-
current to the gas flow. The cleaned, saturated gas passes
through a horizontal, single-stage chevron mist eliminator into a
common duct to the stack at 255 m /s (540,000 acfm) and 54°C
(130°F). The scrubbed gas joins the bypassed gas and the com-
bined stream enters the 120-m (400-ft) stack at approximately 387
m3/s (820,000 acfm) and 81°C (178°F).
PERFORMANCE HISTORY
Tombigbee 2 started FGD operations in September 1978. As of
October 1978, the unit was performing shakedown and debugging
operations. Because of the unit's recent startup, typical per-
formance information is not available.
FUTURE PLANS4
Alabama Electric Cooperative is installing an identical
unit, Tombigbee 3, scheduled to commence operation in June 1979.
Like Tombigbee 2, Tombigbee 3 will include a limestone slurry FGD
B-4
-------
TABLE B-l. FGD SYSTEM DATA FOR TOMBIGBEE 2
Location
Rating (net), MW
Fuel
Heating value, kJ/kg (Btu/lb)
Ash, percent
Moisture, percent
Sulfur, percent
FGD process
FGD system supplier
Application
Startup date (initial)
Design removal efficiency
Particulate (ESP), percent
Sulfur dioxide, percent
Absorbers
Overall
Water loop
Total water makeup,
Liter/s per net MW (gal/min
per net MW)
Sludge disposal
Leroy, Alabama
255
Bituminous coal
25,600-27,900 (11,000-12,000)
10-16
16
0.8-1.5
Limestone
Peabody Process Systems
New
September 1978
99.3
85
59.5
Open
0.08 (1.19)
Unstablized sludge is disposed
of in an onsite, clay-lined pond
B-5
-------
system supplied by Peabody Process Systems and designed to scrub
70 percent of the flue gas. The boiler construction is currently
50 percent complete.
B-6
-------
ARIZONA ELECTRIC POWER COOPERATIVE
APACHE 2
BACKGROUND INFORMATION5
Apache Power Station of Arizona Electric Power Cooperative
is a three-unit station in Cochise, Arizona. Apache 1 is an
existing 100-MW (gross) unit, and Apache 2 and 3 are identical
245-MW (gross) units. Apache 3 is under construction.
The electric power generating facilities at Apache 2 consist
of a Riley Stoker dry bottom pulverized-coal-fired steam gener-
ator with a 200-MW (net) turbine generator. Apache 2 began
operation in September 1978.
The coal burned at Apache 2 is a bituminous, low-sulfur
(from 0.5 percent to 0.8 percent sulfur) coal. Its heating value
ranges from 23,300 to 25,600 kJ/kg (10,000 to 11,000 Btu/lb), and
its ash content is 10 percent.
POLLUTION CONTROL
An ESP designed to be 99.56 percent efficient and supplied
by the Air Correction Division of Universal Oil Products provides
primary particulate control for Apache 2. Emissions of SO_
are controlled with a Research Cottrell limestone slurry FGD
system, which consists of two parallel packed-tower absorbers.
Each tower contains a layer of packing approximately 1.2 m (4 ft)
thick. Apache 2 scrubs 100 percent of the flue gas with an SO2
removal efficiency of 85 percent to meet the State emission
regulation of 344 ng/J (0.8 lb/106 Btu).
Spent scrubbing slurry is pumped to an unlined pond. An-
other .unlined pond is used for disposal of bottom and fly ash.
The life expectancy of each pond is 30 years.
Table B-2 summarizes the FGD data for Apache 2.6
B-7
-------
TABLE B-2. FGD SYSTEM DATA FOR APACHE 2
Location
Rating (gross), MW
(net), MW
Fuel
Heating value, kJ/kg (Btu/lb)
Ash, percent
Sulfur, percent
FGD process
FGD system supplier
Application
Startup date (initial)
(commercial)
Design removal efficiency
Particulate (ESP), percent
Sulfur dioxide, percent
Absorbers
Overall
Water loop
Total water makeup,
liter/s per net MW (gal/min
per riet MW)
Sludge disposal
Cochise, Arizona
245
200
Bituminous coal
23,300-25,600 (10,000-11,000)
10
0.5-0.8
Limestone
Research Cottrell
New
August 1978
November 1978
99.4-99.6
85
85
Open
0.852 (13.5)
Unstabilized sludge is disposed
of in an unlined pond
B-8
-------
PROCESS DESCRIPTION6
After the flue gas exits the boiler, it passes into an ESP
with an efficiency of 99.56 percent and enters the preheater
system for boiler air. The flue gas is fed into the base of two
parallel absorbers designed to handle 190 m /s (400,000 acfm) of
flue gas at 132°C (270°F). The gas rises through a spray zone in
the tower and a 1.2-m (4-ft) layer of packing. Mist elimination
is provided by two stages of horizontal mist eliminators situated
above the packing layer in each tower. After cleaning, the
saturated flue gas enters the 122-m (400-ft) lined stack.
PERFORMANCE HISTORY
Apache 2 began FGD operation in August 1978, and commercial
operation commenced in November 1978. As of October 1978, the
unit was performing shakedown and debugging operations. Because
of the unit's recent startup, performance information is not
available.
FUTURE PLANS7
Arizona Electric Power Cooperative will install an identical
unit, Apache 3, scheduled to begin operation in April 1979.
Apache 3 will include a limestone slurry FGD system supplied by
Research Cottrell that will scrub 100 percent of the flue gas
with an SO- removal efficiency of 85 percent.
B-9
-------
ARIZONA PUBLIC SERVICE
CHOLLA 1 AND 2
BACKGROUND INFORMATION8
The Cholla Steam Electric Station of Arizona Public Service
is in an arid desert region of Navajo County, near Joseph City
Arizona.
Only two of the five planned steam turbine generating units
at Cholla are in operation.
Both Cholla 1 and 2 have wet bottom pulverized-coal-fired
steam generators supplied by Combustion Engineering. Cholla 1 is
a 115-MW (net) unit and started commercial service in May 1962.
Cholla 2 is a 250-MW unit that began operation in June 1978, but
had not been declared commercial as of October 1978.
Arizona Public Service is increasing the station's capacity
from the current 365 MW (net) to 1315 MW (net). Cholla 3 and 4,
now under construction, are scheduled for commercial startup in
June 1979 and June 1980 and are rated at 250 and 350 MW (net),
respectively. Now in the planning stage, Cholla 5 is scheduled
for commercial startup in June 1983. Like Cholla 1 and 2f Cholla
3,4, and 5 will consist of CE puliverized-coal-fired units.
The plant burns low-sulfur, subbituminous coal that is
shipped by rail from the McKinley mine near Gallup, New Mexico.
This coal has a typical heating value of 24,190 kJ/kg (10,400
Btu/lb), sulfur content of 0.5 percent, chloride content of 0.025
percent, ash content of 13.5 percent, and moisture content of 15
percent.
Cholla 1 is equipped with Research Cottrell multicyclones
upstream from the FGD system that are designed to remove 80
B-10
-------
percent of the inlet particulate matter. It also has a retro-
fitted Research Cottrell, FGD system that began operation in
October 1973.
Current emission regulations require that three of the other
four units at Cholla be equipped with FGD systems. Arizona
Public Service has awarded contracts to Research Cottrell for
limestone slurry FGD systems on Units 2 and 4. Unit 3 will
include only an ESP for the control of particulate emissions/ and
the emission control strategy for Unit 5 has not yet been deter-
mined.
Cholla 2 consists of four modules to control particulate and
SO2. The FGD system can handle 100 percent of the flue gas and
commenced operation simultaneously with the boiler in June 1978.
Cholla 3/ an identical 250-MW (net) unit, is scheduled to go on
line in June 1980.
POLLUTION
Cholla 1 is equipped with Research Cottrell multicyclones '
and an FGD system consisting of two modules, A and B. Each
module includes a flooded-disc venturi scrubber, a cyclonic mist
eliminator, an absorber tower, and a final mist eliminator. The
absorber on Module A includes packing to remove SO2 with cir-
culating limestone slurry. The absorber tower in Module B is a
hollow spray tower, and limestone slurry is not circulated
through it. Hence, any SO_ removal in Module B occurs in the
venturi only. Each module treats approximately 50 percent of the
total boiler flue gas. The entire FGD system is designed to
treat 227 m3/s (480,000 acfm) of flue gas at 136°C (276°F) .
Actual boiler flue gas flow to both modules (at 115-MW generating
o
capacity) measures approximately 189 m /s (400,000 acfm). In
addition, bypass leakage around the FGD system amounts to 8
m /sec (17,000 acfm).
The SO- absorber in Module A includes a fixed plate and
conical hat separator and packing. The packing, which consists
of a fixed matrix of rigid sheets of polypropylene, has a high
B-ll
-------
specific surface area and pressure drop of only 0.12 kPa (0.5 in.
H2O). The superficial gas velocity of the SO absorber is 2.1
m/s (6.9 ft/s), and the liquid-to-gas (L/G) ratio is 6.5 li-
ters/m3 (48.9 gal/1000 ft3).
The mist eliminators in each absorber tower are arranged
horizontally in two stages. The first stage is a chevron-type,
two-pass, polypropylene mist eliminator, located approximately
3.7 to 4.6 m (12 to 15 ft) above the absorber packing. The
design configuration of the second-stage, four-pass, polypro-
pylene mist eliminator differs only slightly from that of the
first stage.
A set of shell-and-tube heat exchangers in each module
raises the temperature of the scrubbed gas from 50°C (122°F) to
72°C (162°F) before it is discharged to the atmosphere. The
reheated, scrubbed gases are discharged through carbon steel
ducts to the main stack.
Cholla 2 controls particulate and SO emissions with a four-
module limestone slurry FGD system also supplied by Research
Cottrell. This new unit is designed to remove 99.7 percent of
particulates and 75 percent of SO,,. The four scrubbing trains
are arranged in parallel. Every train consists of a packed-tower
absorber preceded by a flooded-disc scrubber that handles 25
percent of the flue gas (100 percent of the flue gas is scrub-
bed) . Each train, however, can scrub up to 33 percent of the
total flue gas, so that the unit is capable of operating at full
load with only three trains. At full load, each train can treat
185 m3/s (392,000 acfm) of flue gas at 50°C (122°F). Cholla 2
includes an inline steam coil reheat system that increases flue
gas temperature by 22°C (40°F). Spent scrubbing slurry from this
open loop FGD system is disposed in an onsite, unlined pond.
Tables B-3 and B-4 summarize the FGD data for Cholla 1 and
2, respectively.
B-12
-------
TABLE B-3. FGD SYSTEM DATA FOR CHOLLA 1
Location
Rating (gross), MW
(net), MW
Fuel
Heating value, kJ/kg (Btu/lb)
Ash, percent
Moisture, percent
Sulfur, percent
Chloride, percent
FGD process
FGD system supplier
Application
Startup date (initial)
(commercial)
Design removal efficiency
Particulate, percent
Sulfur dioxide, percent3
Actual removal efficiency
Particulate, percent
Sulfur dioxide, percent3
Water loop
Total water makeup,
liter/s per net MW (gal/min
per net MW)
Sludge disposal
Joseph City, Arizona
119
115
Subbituminous coal
23,609 (10,150)
13.5
15.0
0.5
0.025
Limestone
Research Cottrell
Retrofit
October 1973
December 1973
99.7
58.5
99.7
50-60
Open
0.066 (1.04)
Unstabilized sludge is disposed
of in an onsite, unlined pond
These are total system removal values. The A-side module, which
contains the sulfur dioxide absorber, removes 92 percent.
B-13
-------
TABLE B-4. FGD SYSTEM DATA FOR CHOLLA 2
Location
Rating (gross), MW
Fuel
Heating value, kJ/kg (Btu/lb)
Ash, percent
Moisture, percent
Sulfur, percent
Chloride, percent
FGD process
FGD system supplier
Application
Startup date (initial)
Design removal efficiency
Particulate, percent
Sulfur dioxide, percenta
Water loop
Total water makeup,
liter/s per net MW (gal/min
per net MW)
Sludge disposal
Joseph City, Arizona
250
Subbituminous coal
23,609 (10,150)
13.5
15.0
0.5
0.025
Limestone
Research Cottrell
New
June 1978
99.7
75.0
Open
0.066 (1.04)
Unstabilized sludge is disposed
of in an onsite, unlined pond
Because the utility has no scrubber bypass, this represents
the overall removal efficiency.
B-14
-------
PROCESS DESCRIPTION
Flue gas from the Choila 1 boiler induced-draft fans is
pressurized by two booster fans and then flows downward through
the throat of the venturi-type, flooded-disc particulate scrub-
ber. Limestone slurry flows out over the disc and is atomized as
it is sheared by the gas stream at the edge of the disc. Slurry
is also injected tangentially through nozzles on the inside wall
of the venturi scrubber shell above the tapered throat. The
saturated, scrubbed flue gas then passes through a cyclonic mist
eliminator, where solids from collected fly ash, limestone slur-
ry, and reaction products are separated from the gas stream
before it enters the absorber.
Gas from the cyclonic mist eliminator enters the absorber
tower near the base. In Module A only, it contacts the limestone
slurry on the surface of the wetted-film packing, which is 0.6 m
(2 ft) thick.
The scrubbed gas passes through a set of mist eliminators
(one set per absorber) and is reheated before it is discharged to
the atmosphere through the main stack. The mist eliminators are
slat (special baffle design) impingement type, constructed of
polypropylene and arranged horizontally (vertical gas flow) in
two stages. Reheat is provided by a set of steam-heated, shell-
and-tube heat exchangers (one set per module).
Limestone is added to Module A of the FGD system at a rate
of approximately 110 percent of the stoichiometric requirement
for reaction with the SO- in the flue gas. Part of the circu-
lated liquor in the SO2 absorber is diverted to the flooded-disc
scrubber tank (common to both modules) to maintain the pH between
4 and 5 in the particulate control system. The liquid level in
this tank is maintained by pumping the excess spent liquor to one
of two surge tanks (sludge holdup tanks) before it is discharged
to a pond. The plant site has no facilities for sludge storage
or fixation. Because of the area's light rainfall and a high
evaporation rate, wastewater discharge into receiving waters is
not a problem. Therefore no liquor is recirculated back to the
process. B_15
-------
Dampers in the FGD system are arranged so that either Module
B alone or Modules A and B together can be bypassed for a long
period. Module A alone can only be bypassed for a short time
because limestone, which is used to control the operating pH of
the entire particulate scrubber system, enters the system in the
Module A absorber tower tank.
PERFORMANCE HISTORY11
Since September 1977, no major design modifications have
been necessary at Cholla 1. The period from October 1977 through
September 1978 is characterized by routine overhauls and general
maintenance. The main problems during this period have been the
failure of a recycle pump expansion joint, leaks in the flooded-
disc scrubber tank header, and leaks in the venturi section. A-
side and B-side reliabilities have averaged 96 percent and 95
percent, respectively.*
Cholla 2 began operation in June 1978, and all four scrub-
bing trains are in service. Cholla 2 has experienced reasonant
vibrations in the slurry piping since the unit became opera-
tional. To reduce the vibrations, Research Cottrell has been
injecting nitrogen into the slurry lines; but no permanent solu-
tion to the problem has yet been found. In addition to the
vibration problem, some peeling of the corrosion-resistant lining
has occurred in the downcomer area of one absorber module.
FUTURE PLANS12'13
Arizona Public Service plans to install an FGD system on
Cholla 4 and 5. Cholla 3 will include an ESP for particulate
control only. Cholla 4 will utilize a single module FGD system
for partial scrubbing, and Cholla 5 may be required to use a
scrubbing system to remove 98 percent of the SO2 from the flue
Reliability values are provided because this is the only index
reported by the utility for this system.
B-16
-------
gas. Cholla 4 and 5 are scheduled to begin operations in June
1980 and 1983, respectively.
Arizona Public Service will also control SO- emissions to a
greater degree at its Four Corners Station, where the operational
particulate scrubbers at Units 1, 2, and 3 are being upgraded for
additional SO. removal. The particulate scrubbers at these units
currently remove approximately 30 percent of the SO2 emitted from
the boilers. The Four Corners Station, which consists of five
units, is required to achieve a total SO- removal efficiency of
67.5 percent. The control strategy has not yet been finalized
for Four Corners 4 and 5. Four Corners 1, 2, and 3 should begin
upgraded FGD operations sometime in 1979. Four Corners 4 and 5
will not begin operations until 1982 at the earliest.
B-17
-------
CENTRAL ILLINOIS LIGHT COMPANY
DUCK CREEK 1
BACKGROUND INFORMATION
The Duck Creek Plant is a new coal-fired power generating
station owned and operated by the Central Illinois Light Company.
The plant is situated in an unreclaimed strip-mine area near
Carrton, Illinois. The current capacity of the plant is 416 MW
(gross), which is provided by one unit placed in commercial
service on June 1, 1976.
Duck Creek 1 is designed to fire bituminous, high-sulfur
(3.3 percent sulfur), Illinois coal with an average ash content
of 18 percent. When the FGD system is not used, low-sulfur
Colorado coal is burned.
Three more units of identical capacity are planned for
installation at the station. Duck Creek 2, 3, and 4 are sched-
uled for commercial operation in 1982, 1989, and 1992, respec-
tively and this will bring total station capacity to 1664 MW
(gross).
POLLUTION CONTROL
Primary particulate control is provided by two parallel
ESP's designed to remove 99.8 percent of particulates. Primary
SO~ control is provided by a limestone slurry FGD system con-
sisting of four parallel modules, each supplying 25 percent of
capacity- Overall (design) SO2 removal efficiency is 85.3 per-
cent when burning 4 percent sulfur coal. The modules are verti-
cal, rod-deck, venturi-sorber scrubbers. Each module can accom-
modate 167 m /s (353,900 acfm). All or part of the flue gas can
B-18
-------
be bypassed around the scrubber modules during outages or emer-
gencies by manipulating bypass dampers and module isolation
dampers.
Table B-5 summarizes the FGD data for Duck Creek 1.
PROCESS DESCRIPTION
Leaving the boiler at 1,140 m3/s (2,415,000 acfm) and 446°C
(835°F), flue gas passes through half-size air preheaters and
into two parallel ESP's supplied by Pollution Control-Walther Co.
The ESP's are designed to remove 99.8 percent of particulates
when the inlet gas loading is 14.5 mg/m (6.34 gr/scf) . The
discharge gas from the ESP's enters a manifold with four induced-
draft fans that overcome draft loss in the boiler, the ESP's, and
the FGD system. These fans are connected in parallel to a common
duct that distributes the gas to each module in the FGD system or
to the bypass duct.
At the base of each scrubber module, flue gas is quenched to
adiabatic saturation conditions. The quenched gas flows upward
through nine stages of rod decks where contact is made with the
scrubbing slurry in a countercurrent fashion. The scrubbing
slurry is sprayed from the top of each module and flows downward
through the rod decks. By providing for intimate contact between
the gas and slurry, the rod decks enhance transfer of SO into
the liquid phase and thus promote SO2 removal.
The cleaned, saturated gas stream in each module exits the
spray zone, turns 90 degrees, and passes through a horizontal
mist elimination section. Entrained droplets of moisture and
slurry are removed from the gas stream by passage through two
stages of chevron-type, slanted-vertical mist eliminators located
in the horizontal discharge duct of each module.
Scrubbed, saturated gas exits .the FGD system and enters the
stack through the breeching section without reheat. The "wet
stack" is a 150-m (500-ft) chimney that contains a Cor-Ten steel
flue lined with flake glass. Four bottom hoppers are included in
B-19
-------
TABLE B-5. FGD SYSTEM DATA FOR DUCK CREEK 1
Location
Rating (gross), MW
(net), MW
Fuela
Heating value, kJ/kg (Btu/lb)
Ash, percent
Moisture, percent
Sulfur, percent
Chloride, percent
FGD process
FGD system supplier
Application
Startup date (initial)
(Commercial)
Design removal efficiency
Particulate (ESP), percent
Sulfur dioxide, percent
Water loop
Total water makeup,
liter/s per net MTV (gal/min
per net MW)
Sludge disposal
Canton, Illinois
416
400
Bituminous coal
24,523 (10,543)
9.12
18.0
3.30
0.03
Limestone
Riley Stoker/Environeering
New
July 1976
August 1978
99.8
85.3
Open
0.094 (1.49)
Unstabilized sludge is disposed
in an onsite, lined pond
Design fuel specifications for high-sulfur Illinois coal.
Boiler and ESP began commercial operation in June 1976. One
FGD module commenced operation in July 1976, and full commercial
operation of all four FGD modules commenced in August 1978.
B-20
-------
the stack for collection of moisture and slurry droplets that
fall out of the flue gas because of a difference in the velocity
of the droplets and velocity of the gas.
Spent scrubbing slurry is bled from the scrubber recircu-
lation lines. The solids content of this slurry is 15 percent
and includes reaction products/ unreacted limestone, and fly ash.
The spent slurry is transferred to a waste collection tank where
it is combined with liquid waste streams from plant sumps and
then discharged to an onsite pond for final disposal.
PERFORMANCE HISTORY14"16
Duck Creek 1 commenced commercial operation on June I/ 1976.
The D-scrubber module was initially placed in the flue gas path
on July I, 1976. Operation during July and August was limited
because of such construction deficiencies as bad welds, faulty
pipe hangers, and leaks in the scrubber system. To resolve these
problems, the D scrubber was taken out of the gas path; it was
placed back on September 9th and operated noncontinuously for
approximately 360 hours in September and 385 hours in November.
From September through November, a number of major operating
problems were encountered, especially with massive scaling of the
mist eliminators, plugging of the spray nozzles and pipes, and
failure of materials. The module remained out of service in
December, January, and February because of a scheduled 6-month
overhaul of the boiler and turbine. During this outage, a number
of modifications were made to the scrubber to correct the oper-
ating problems encountered. The unit was placed back in service
in mid-March 1977, and the D scrubber operated almost contin-
uously for 350 hours during the balance of the month. Testing of
the automatic control loops, during April and May, was premature-
ly terminated because of installation difficulties that caused
the D scrubber to be taken out of service. Duck Creek 1 remained
in service with the boiler firing low-sulfur Colorado coal and
the ESP's removing particulates with the aid of sulfur trioxide
gas injection. The D scrubber, along with the other scrubber
B-21
-------
modules, was placed back in the flue gas path on July 23, 1978.
Since July 1978, the system has operated intermittently.
Modifications were made to the slurry transfer tank after it was
found to be underdesigned. Because a common unloading and trans-
fer system is used for coal and limestone, coal fines have been
present in the limestone slurry and produced plugging problems in
the FGD system. From August through September 1978, the average
availability of the FGD system was only 46 percent.
Because the FGD system has only recently begun commercial
operation, SO2 removal efficiencies for full-scale operations are
not available. However, SO0 removal efficiency was measured for
^
the D-scrubber module during the interim test. The test indi-
cates that the module's removal efficiency was 91.6 percent, well
above the design maximum guarantee of 85.3 percent when coal
containing 4 percent sulfur is fired. This measurement was taken
for SO2 inlet concentrations of 3000 ppm.
Measurements of particulate removal efficiency during the
test of the D scrubber suggest that the scrubber was removing as
much as 70 percent of the inlet particulate matter that passed
through the upstream ESP's. Although the FGD system is guaran-
teed not to add any particulates to the gas stream discharged
from the outlet of the EPS's, the scrubbers are also not designed
to provide any additional particulate removal capability to the
emission control system. The utility and system supplier indi-
cate that the additional particulate removal may be attributed to
the ionization or agglomeration of particulates during passage
through the upstream ESP's. As a consequence, collection of
these particles in the downstream scrubber is greatly enhanced.
FUTURE PLANS18'19
The utility is currently evaluating bids for the construc-
tion of a second 400-MW (nominal) unit to begin operation in
1982. Duck Creek 2 will use ESP's for particulate control and a
limestone slurry or double alkali process for SO- control.
B-22
-------
GULF POWER COMPANY
SCHOLZ 1 AND 2
BACKGROUND INFORMATION20
The Scholz steam power plant of Gulf Power Company is a two-
unit station in Chattahoochee, Florida. Scholz 1 and 2 are 47.5-
MW (gross) units, each of which can produce 40 MW (net) at
maximum capacity.
The electric power generating facilities at the Scholz steam
power plant consist of two pulverized-coal-fired steam generators
supplied by Babcock & Wilcox with associated turbine generators.
Scholz 1 and 2 were placed in service in 1953.
The coal burned at this station generally is high-sulfur
(2.5 percent sulfur) coal having a heating value of 29,000 kJ/kg
(12,400 Btu/lb) and an ash content of 14 percent. A variety of
coals, however, are burned during prototype testing; and sulfur
content is as high as 5 percent at times.
POLLUTION CONTROL21
An ESP designed to be (97.7 percent efficient) supplied by
Buell (now Buell-Environmental) provides primary particulate
control for the Scholz installation. The SO2 is controlled with
a prototype limestone slurry FGD system supplied by Chiyoda
International. The Scholz units have tried a number of experi-
mental FGD processes. The current Thoroughbed 121 system is an
upgraded model using components from a Thoroughbred 101 prototype
also supplied by Chiyoda International.
The FGD system consists of a jet bubbler tank that scrubs
28.8 percent of the flue gas. The system design includes a
B-23
-------
double-pass chevron mist eliminator. No reheat is provided. Gas
exits the FGD system and passes through a test stack.
Scrubbing waste, in the form of a relatively stable gypsum,
is bled from the bottom of the jet bubbler tank. The gypsum,
with a solids content between 10 and 25 percent, is pumped to a
gypsum pond where the solids settle out. The supernatant is
pumped back to the process.
FGD data are summarized in Tabxe B-6.
PROCESS DESCRIPTION22'23
The Thoroughbred 121 prototype system is designed to accom-
modate flue gas from a coal-fired steam generator having an
equivalent electrical capacity of 23 MW. Portions of flue gas
from Scholz 1 and 2 pass through a header into the FGD system.
The flue gas flows through a port slightly above the center of
the jet bubbler into a ring-like chamber around the tank's
circumference. It then travels downward through a shallow slurry
layer in pipe-like vertical ducts (vertical sparges) in the jet
bubbler zone. The open ends of the vertical sparges are sub-
merged 10 to 41 cm (4 to 16 in.) below the slurry surface. The
flue gas contacts the slurry in the froth created by the high-
velocity flue gas as it comes from the sparges. Oxidizing air
introduced into the tank through a horizontal duct near the tank
base promotes gypsum crystal growth from the spent reagent in the
tank. Gypsum settles to the bottom of the tank and is drawn off
in a continual bleed stream. The cleaned, saturated gas rises
above the slurry froth around the ring of vertical sparges and
exists through a center duct near the top of the tank. After
passing through the mist eliminator system, the flue gas is
ducted without reheat to a 23-m (76-ft) stack used for the
scrubbing system only.
PERFORMANCE HISTORY24'25
The Thoroughbred 121 prototype began operation at the Scholz
station on August 30, 1978. An availability of greater than 99
B-24
-------
TABLE B-6. FGD SYSTEM DATA FOR SCHOLZ 1 AND 2a
Location
Rating (gross), MW
Fuel
Heating value, kJ/kg (Btu/lb)
Ash, percent
Sulfur, percent
Chloride, percent
FGD process
FGD system supplier
Application
Startup date (initial)
Design removal efficiency
Particulate (ESP), percent
Sulfur dioxide, percent
Water loop
Sludge disposal
Chattahoochee, Florida
40 (each)
Coal
29,000 (12,400)
14
2.5 (5% maximum)
0.15
Limestone
Chiyoda International
Retrofit
August 1978
99.7
Open
FGD waste in the form of
gypsum is stacked in the
existing onsite pond
Flue gas from both units passes through a header into this
prototype FGD system. Only 28.75 percent of the flue gas
generated by this 80-MW (gross) station is treated.
B-25
-------
percent has been reported since then. As of October 1978, no
operating problems had been reported.
B-26
-------
INDIANAPOLIS POWER AND LIGHT COMPANY
PETERSBURG 3
BACKGROUND INFORMATION26
The Petersburg Power Station, owned and operated by the
Indianapolis Power and Light Company/ is located near Petersburg,
Indiana, approximately 160 km (100 miles) southwest of
Indianapolis. The station consists of three coal-fired, steam-
turbine generating units, which have a combined capacity of 1180
MW. Petersburg 1 is rated at 220 MW; Petersburg 2, at 430 MW;
and Petersburg 3, at 530 MW. The utility is planning the instal-
lation of another 530-MW unit, which is scheduled to be in
service by April 1982. Petersburg 4 will bring the station's
combined power generating capacity to 1710 MW. Of the three
operational units, only Petersburg 3 is required to meet the
current Federal New Source Performance Standards for SO. emissions
of 516 ng/J (1.2 lb/106 Btu). This unit is fitted with an FGD
system.
The Petersburg 3 boiler is a Combustion Engineering unit
designed to fire bituminous high-sulfur (4.5 percent sulfur)
Indiana coal with a heating value of 25,000 kJ/kg (10,750 Btu/lb),
ash content from 8 to 15 percent, and moisture content from 10 to
16 percent.
POLLUTION CONTROL27
The emission control system on Petersburg 3 consists of two
ESP's designed by Research Cottrell to be 99.3 percent efficient
and a limestone slurry FGD system supplied by the Air Correction
Division of Universal Oil Products. The FGD system consists of
four parallel modules that scrub 100 percent of the flue gas. In
B-27
-------
each module, there are three beds of turbulent contact absorbers
(TCA's). The SO removal efficiency for the system is designed
to be 85 percent. The system includes two stages of chevron-type
mist eliminators in each module and an indirect hot-air steam-
coil reheat system that raises the temperature of the cleaned,
saturated gas 17°C (30°F).
Spent scrubbing slurry from the system is thickened and
stabilized in a fixation facility supplied by International
Utility Conversion Systems before disposal in an onsite unlined
pond. A ball mill with a capacity of 38 Mg/h (42 tons/h) pro-
vides onsite limestone preparation.
Table B-7 summarizes the FGD data for Petersburg 3.
2R
PROCESS DESCRIPTION
Approximately 900 m /s (1,900/000 acfm) of flue gas at 137°C
279°F) enters two Research Cottrell cold-side ESP's for primary
particulate control. The ESP's remove 99.3 percent of the inlet
particulate matter, to produce an outlet loading of 0.0572 g/m
(0.025 gr/acf) or 43 ng/J (0.1 lb/106 Btu). Two forced-draft
fans then boost the gas through four absorbers supplied by the
Air Correction Division of Universal Oil Products, each of which
accommodates 25 percent of the flue gas. Every module is capable
of treating 220 m /s (465,000 acfm) of flue gas. The gas is fed
into the base of the absorber towers and rises through three
stages of mobile bed packing, where intimate contact occurs with
the limestone slurry. The pressure drop through each module is
1.8 kPa (7.0 in. H.O); the superficial gas velocity, 4.42 m/s
(14.5 ft/s); and the L/G ratio, 6.7 liters/in3 (50 gal/103 ft3).
Two pumps, each of which can handle 44,700 liters/min (11,800
gal/min), provide limestone slurry for every module; and there is
one spare pump per two modules. Cleaned saturated gas passes
through two stages of chevron mist eliminators in each module at
approximately 177 m3/s (375,000 acfm) and 48°C (118°F). This
gas contains 385 ppm of SO. and is then reheated to 65°C (148°F),
B-28
-------
TABLE B-7. FGD SYSTEM DATA FOR PETERSBURG 3
Location
Rating (gross), MW
(net), MW
Fuel
Heating value, kJ/kg (Btu/lb)
Ash, percent
Moisture, percent
Sulfur, percent
FGD process
FGD system supplier
Application
Startup date (initial)
Design removal efficiency
Particulate (ESP), percent
Sulfur dioxide (absorbers),
percent
Water loop
Total water makeup,
liter/s per net MW (gal/min
per net MW)
Sludge disposal
Petersburg, Indiana
532
515
Bituminous coal
25,005 (10,750)
8-15
10.5-16.5
4.5
Limestone
Air Correction Division, UOP
New
December 1977
99.3
85
Closed
0.1 (1.7)
Stabilized sludge is disposed
of in an onsite, unlined pond
B-29
-------
The reheat is provided by heating ambient air with bundles of
steam tubes located outside the gas ducts. The gases then dis-
charge to a 188-m (616-ft) stack.
Spent scrubbing solution is discharged from the TCA modules
to a waste slurry recycle tank, where chemical reactions are
completed. Flue gas cleaning wastes, which include calcium
sulfite, calcium sulfate, calcium carbonate, and fly ash, are
bled from the absorbers and discharged to a thickener, where the
waste solids settle out. Thickener overflow is returned to the
process; and thickener underflow, with a solids content of 35
percent, is discharged to an International Utility Conversion
Systems installation for chemical fixation. The stabilized
sludge is transported by conveyor to trucks that deliver the
material to an onsite, unlined, disposal pond approximately 300 m
(1000 ft) from the fixation area. Clarified water is returned to
the process for further use.
PERFORMANCE HISTORY29
The initial two TCA modules were put in operation by
December 12, 1977; the remaining two, by the end of the same
month. During the winter of 1977, temporary enclosures had to be
erected around segments of the FGD system until heat tracing
could be installed to prevent freezing in the liquid circuit. As
of October 1978, the system was still undergoing design modifi-
cation as shakedown and debugging operations continued. Since
December 1977, the utility has encountered problems with a
recycle tank agitator, the fly-ash removal system, the liquid
circuit, and the control valve and piping areas. Reliable per-
formance data have not yet been made available by the utility.
FUTURE PLANS30
Indianapolis Power and Light has awarded a contract to
Research Cottrell for a limestone FGD system on Petersburg 4,
which is scheduled to begin operations in April 1982. The S00
B-30
-------
removal efficiency at Petersburg 4 is expected to be 80 percent
for the same type of high-sulfur coal burned in Petersburg 3.
Sludge will be dewatered and mixed with fly ash to produce a dry,
stabilized product. Plant construction began in December 1977.
B-31
-------
KANSAS CITY POWER AND LIGHT
LA CYGNE 1
PERFORMANCE HISTORY31'32
Because of continuing modifications and improved operating
procedures, the availability of the La Cygne 1 FGD system has
steadily improved since startup in 1973. The averages were 76.3,
84.3, 92.0, and 92.5 percent for 1974, 1975, 1976, and 1977,
respectively. From January through September 1978, the average
availability was 93.2 percent. With the addition of the eighth
module in April 1977, continuous daytime load capability now
exceeds 800 MW without appreciably affecting scrubber operation.
A performance test on August 26, 1977, indicated that module
gas flow remained below maximum design capability, that the
induced- and forced-draft fans were loaded substantially below
rating, and that most subsystems were properly balanced. The
total SO2 removal efficiency averaged 77 percent, with the
averages for individual modules ranging from 65 to 80 percent.
Particulate emissions from the plant have met EPA and Kansas
State requirements, although work is continuing to achieve
further reductions in such emissions from this unit.
Limestone utilization was greatly improved in early 1977
with improved pH control. In the past, it was almost impossible
to maintain inline glass cells without caking the limestone
during shutdown or eroding the cells during operation with the
high concentration of fly ash. Proper maintenance, including
acid flushing, sonic cleaning, and periodic water backflushing,
has produced reliable pH readings, reduced limestone use by
approximately 30 percent, and improved control of scaling.
B-32
-------
Mist eliminator plugging or scaling has not been a problem
at La Cygne since the intermittent wash system was eliminated and
the continuous wash system was moved from underneath to above the
first mist eliminator. More nozzles were added to the continuous
wash system, so that its water delivery rate increased from 8.8
liters/s (140 gal/min) to 15 liters/s (230 gal/min).
Carryover to the blades of the induced-draft fans continues
to require regular washings. Each fan must now be cleaned every
4 to 7 days. The high-pressure wash is seldom necessary any
more. A "spinning" process with low-pressure hoses has been very
effective in cleaning fans while they are out of service. The
washings are usually done as preventive maintenance, but fans
must be taken out of service when bearing vibrations become
excessive.
Rubber pipe and pump linings have been an increasing main-
tenance problem. After approximately 5 years of operation, some
materials that have not been modified are wearing out. Rubber
linings that tear out may damage other pipes, pumps, or plug
nozzles and allow the steel pipes to wear through.
Corrosion of carbon steel in the duct work, dampers, induced-
draft fan rotors and housings, breeching, and stack liner con-
tinues to be a major problem. The burning of high-sulfur coal in
conjunction with module outages can result in enormous concentra-
tions of sulfurous and sulfuric acid on these surfaces. Diligent
surveillance by maintenance engineers is needed to prevent such
"cold-end corrosion" damage, and unit outages are planned to take
into consideration temperatures and time requirements for apply-
ing special coatings.
The scrubber operating and maintenance force is being in-
creased to 54 people by adding one electrician and two tech-
nicians. Continued improvements in operating procedures and
stable equipment operation should permit analysis of improved
chemistry and control parameters. The current effort to maintain
pH cells and S02 analyzers may require an increase in the manpower
force. There are also increased demands on present maintenance
B-33
-------
personnel to accumulate, record, and evaluate operating data on
water saturation trends, limestone utilization, wear rates of
draft fans, failures of reheater bundles, and lined pumps,
replacement of nozzles and rubber-lined pipe, use of spare
parts, etc. The operators are also busy updating operational and
special instructions and reviewing safety and training procedures.
FGD data for La Cygne 1 are summarized in Table B-8.
FUTURE PLANS33
Proposed modifications and activities include the following:
0 Installation of an improved steam source to permit
additional steam bundles for optimum reheat.
0 Construction of an additional sludge pond for deposit
of spent scrubber slurry for approximately 30 years.
Adding another disposal pond might also allow clear
water to be recycled to the scrubber for improved
chemistry.
0 Installation of a second mist eliminator in the remain-
ing five modules and evaluation of the addition of a
third mist eliminator.
0 Initiation of a study on the submicron fly ash and
sulfuric acid mist that pass through the scrubber
without being collected.
0 Continuation of work to devise a better method to clean
inline reheat tubes without taking equipment out of
service.
B-34
-------
TABLE B-8. FGD SYSTEM DATA FOR LA CYGNE 1
Location
Rating (gross), MW
(net) , MW
Fuel
Heating value, kJ/kg (Btu/lb)
Ash, percent
Moisture, percent
Sulfur, percent
Chloride, percent
FGD process
FGD system supplier
Application
Startup date (initial)
(commercial)
Design removal efficiency
Particulate, percent
Sulfur dioxide, percent
Water loop
Total water makeup,
liter/s per net MW (gal/min
per net MW)
Sludge disposal
La Cygne, Kansas
874
820
Subbituminous coal
21,000-22,600 (9,000-9,700)
25
8.6
5
0.027
Limestone
Babcock and Wilcox
New
February 1973
June 1973
99.5
80
Open
0.088 (1.4)
Unstabilized sludge is disposed
of onsite in an unlined pond
B-35
-------
KANSAS POWER AND LIGHT COMPANY
JEFFREY 1
BACKGROUND INFORMATION34'35
Kansas Power and Light (KP&L) Company is developing a new
2880-MW (gross) coal-fired power generating station known as the
Jeffrey Energy Center in Pottawatomie County, near Belvue,
Kansas. The station is composed of four 720-MW (gross) units, of
which the first two are scheduled for commercial operation in
October 1978 and June 1980 and the latter two, in 1982 and 1984.
All units will fire low-sulfur (0.32 percent sulfur) Wyoming coal
having a heating value of 18,900 kJ/kg (8125 Btu/lb), an ash
content of 5.8 percent, and a moisture content of 30.0 percent.
The pulverized-coal-fired steam generators and emissions
control systems for Jeffrey 1 and 2 are designed and supplied by
Combustion Engineering; and the turbine generators, by Allis
Chalmers. Jeffrey 1 began operation in August 1978. Jeffrey 2
is currently under construction.
POLLUTION CONTROL36
The emissions control system for each unit includes the
following: an over-fire air system at the tangential burners for
control of nitrogen oxides; two cold-side ESP's with an effi-
ciency of 99.0 percent and crossover ducts upstream and down-
stream of the ESP's for control of particulates; and four induced-
draft fans and six pressurized vertical spray towers for control
of SO2.
The six spray towers, one of which is a spare, are designed
to remove SO2 from 75 percent of the flue gas. The remaining 25
percent of the flue gas is bypassed around the towers to provide
B-36
-------
reheat to the scrubbed gas before discharge to the atmosphere
through separate 183-m (600-ft) stacks. Each spray tower has two
spray headers located 4 and 8 m (13 and 26 ft) above the gas
inlet and is equipped with louver isolation and bypass dampers
that permit module isolation from the gas path during periods of
inactivity (reduced load, spare duty, or maintenance).
Limestone is used for SO2 removal. The limestone is re-
ceived as rock at the plant and ground by wet ball mills, each
capable of handling 11 Mg/h (12 tons/h). There are three mills
for Jeffrey 1 and 2, leaving one ball mill as a spare.
Four reaction tanks are provided per FGD system. Two of
these tanks serve four spray towers with each tank shared by two
towers. The other two reaction tanks serve only one spray tower
each.
A mixing chamber is provided for each system to permit
mixing and drying of the scrubbed gas stream with the bypass gas
stream before discharge to the atmosphere. The mist eliminators
are identical in design and construction to those at KP&L's
Lawrence station, which are A-frame, two-stage mist eliminators
constructed of fiber-reinforced plastic (FRP) with a bulk-entrain-
ment separator. Also like the Lawrence station, the modules are
constructed of 316 low-carbon stainless steel.
Spent absorbent collected in the reaction tanks is bled to
a common transfer tank as a slurry with a solids content of 10
percent. The sludge is currently mixed with bottom ash and
disposed onsite in the bottom ash pond. Water returned from the
pond is blended with cooling tower blowdown and used as makeup in
the reaction tanks.
Table B-9 summarizes FGD data for Jeffrey 1.
PERFORMANCE HISTORY37
Jeffrey 1 began operations in August 1978. As of October
1978, the unit was performing shakedown and debugging operations.
Because of the unit's recent startup, reliable performance infor-
mation is not available.
B-37
-------
TABLE B-9. FGD SYSTEM DATA FOR JEFFREY la
Location
Rating (gross), MW
(net), MW
Fuel
Heating value, kJ/kg (Btu/lb)
Ash, percent
Moisture, percent
Sulfur, percent
Chloride, percent
FGD process
FGD system supplier
Application
Startup date (initial)
Design removal efficiency
Particulate (ESP), percent
Sulfur dioxide, percent
Absorbers
Overall3
Water loop
Total water makeup,
liter/s per net MW (gal/min
per net MW)
Sludge disposal
Belvue, Kansas
720
680
Coal
18,900 (8,125)
5.8
30.0
0.32
0.01
Limestone
Combustion Engineering
New
October 1978
99.0
80
60
Closedb
0.04 (0.6)
Unstabilized sludge is disposed
of in an onsite, unlined pond
Twenty-five percent of the flue
system.
Closed water loop operation for
utility during the Oct. - Nove.
Report update.
gas bypasses the scrubbing
this system verified by the
1979 EPA Utility FGD Survey
-------
FUTURE PLANS38
Jeffrey 2 is scheduled to commence operation in June 1980
The unit is currently under construction.
B-39
-------
KANSAS POWER AND LIGHT COMPANY
LAWRENCE 4 AND 5
BACKGROUND INFORMATION39
The Lawrence Energy Center is an existing 625-MW (gross)
power generating station owned and operated by KP&L in Douglas
County, near Lawrence, Kansas. The station consists of five
units. Lawrence 1 was built in 1939 and is a 10-MW turbine now
powered by extraction steam from Lawrence 5. Lawrence 2 and 3
were placed in service in 1950 and 1956; they are peaking units
that can fire either oil or gas and are rated at 30 and 60 MW,
respectively. Lawrence 4 and 5 began operation in 1959 and 1971.
Although originally multiple-fuel-fired units, they now exclu-
sively fire coal. These cyclic-load units are rated at 125 and
400 MW, respectively. The steam generators for Lawrence 4 and 5
are balanced-draft, tangentially fired units supplied by Combus-
tion Engineering. Lawrence 5 produces 1272 Mg (2,805,000 Ib) per
hour of superheat steam at 540°C (1005°F) and 18.1 MPa (2620
psi) .
The coal burned at Lawrence is low-sulfur (0.5 percent sul-
fur) subbituminous Wyoming coal with an average heating value of
23,000 kJ/kg (10,000 Btu/lb), ash content of 10 percent, and
moisture content of 12 percent. Lawrence 4 consumes at full load
approximately 45 Mg (50 tons) per hour of coal; and Lawrence 5,
approximately 145 Mg (150 tons) per hour.
40 41
POLLUTION CONTROL u/
Lawrence 4 and 5 are equipped with tail-end wet limestone
scrubbing systems. In each operational scrubbing system, there
are two parallel two-stage modules for the control of particulates
B-40
-------
and SO2. Every module consists of a rectangular, variable-throat
rod-deck venturi scrubber followed by a spray tower absorber.
Each system is also equipped with slurry hold tanks, mist elimina-
tors, inline reheaters, and isolation and bypass dampers. Bypass
ducts are provided so that the modules can be bypassed during
periods when oil or natural gas may be burned in the boilers.
Both systems share a common limestone storage and prepara-
tion facility and a common waste disposal facility. Waste solids
in the slurry circuits of Lawrence 4 are removed from the scrub-
bing system by liquid staging, forced oxidation, and thickening.
After treatment, the waste stream has a solids content of 35
percent and is conveyed to a network of three disposal ponds.
The supernatant from these ponds is returned to the process and
blended with thickener overflow in the recirculation tank.
Although Lawrence 4 and 5 share the same sludge disposal ponds,
Lawrence 5 is not equipped with a liquid staging and thickening
system. Spent slurry is forcibly oxidized by air sparging and
bled from the system by effluent bleed pumps, which discharge
underflow from the reaction tank directly to the ponds. Super-
natant is returned to the process and added directly to the
reaction tank.
The scrubbing systems were designed and supplied by Combus-
tion Engineering. They represent a second-generation design
replacement of the systems first installed on these boilers in
1968 and 1971.
The original limestone furnace-injection and tail-end
scrubbing system was retrofitted on Lawrence 4 and operated from
1968 until mid-September 1976. Approximately 27,000 hours of
service was accumulated before the unit was shut down to perform
a scheduled turbine overhaul. During the overhaul, erection of
the new scrubber modules was completed. The new system went into
service in early January 1977. In the last part of September
1978, the unit was down for a scheduled turbine/boiler outage,
and no FGD operations were required. Particulate removal effi-
ciency at Lawrence 4 has been in excess of 99 percent; and
removal efficiency, as high as 96 to 98 percent.
B-41
-------
The original limestone furnace-injection and tail-end scrub-
bing system was installed as new equipment on Lawrence 5 and
operated from November 1971 until March 20, 1978. Approximately
23,000 hours of service time was accumulated before the unit was
shut down for the tie-in of the new scrubbing system into the
flue gas path. The new scrubber modules were erected directly
behind the existing system, which remained in service during
construction of the new system. Because the new system was
designed to use the existing reaction tank, spray pumps, induced-
draft fans, and stack, an outage of 6 weeks was required to
complete installation. The new system began operation on April
14, 1978. Information is not available about typical particulate
and SO2 removal efficiencies at Lawrence 5.
Tables B-10 and B-ll summarize the revised FGD data for
Lawrence 4 and 5, respectively.
PROCESS DESCRIPTION
The following paragraphs describe the Lawrence 4 scrubbing
system. Although the Lawrence 5 scrubbing system is similar in
design and operation, a number of major features are different.
These differences are discussed at the end of the section.
Flue gas from the boiler passes through the air heater and
is conveyed by new duct work to two scrubber modules. Each
module can handle 50 percent of total capacity and consists of a
rectangular, variable-throat, rod-deck, venturi scrubber followed
by a spray tower absorber. Also, each module is equipped with
mist eliminators, reheaters, two reaction tanks, a bypass duct,
and bypass and isolation dampers.
Flue gas enters the scrubbers at 190 m3/s (403,000 acfm) and
138°C (280°F). Each rod-deck venturi scrubber is comprised of a
converging gas section and a rod section. The converging gas
section directs the flue gas downward to the rod-deck section,
which consists of two staggered levels of rubber-coated fiber-
glass rods. Limestone slurry is sprayed continuously into the
scrubber by nonatomizing, fan-type, spray nozzles around the
A-42
-------
TABLE B-10. FGD SYSTEM DATA FOR LAWRENCE 4
Location
Rating (gross), MW
(net), MW
Fuel
Heating value, kJ/kg (Btu/lb)
Ash, percent
Moisture, percent
Sulfur, percent
FGD process
FGD system supplier
Application
Startup date
Design removal efficiency
Particulate, percent
Sulfur dioxide, percent
Actual removal efficiency
Particulate, percent
Sulfur dioxide, percent
Water loop
Sludge disposal
Lawrence, Kansas
125
115
Coal
23,000 (10,000)
9.8
11.8
0.55
Limestone
Combustion Engineering
Retrofit
January 1976
98.9
73.0
99+
96-98
Closed
Unstabilized sludge is disposed
of in an onsite, unlined pond
B-43
-------
TABLE B-ll. FGD SYSTEM DATA FOR LAWRENCE 5
Location
Rating (gross), MW
(net), MW
Fuel
Heating value/ kJ/kg (Btu/lb)
Ash, percent
Moisture, percent
Sulfur, percent
FGD process
FGD system supplier
Application
Startup date (initial)
Design removal efficiency
Particulate, percent
Sulfur dioxide, percent
Water loop
Sludge disposal
Lawrence, Kansas
420
400
Coal
23,000 (10,000)
9.8
11.8
0.55
Limestone
Combustion Engineering
New
April 1978
98.9
52.0
Closed
Unstabilized sludge is disposed
of in an onsite, unlined pond
B-44
-------
perimeter of the throat area. The rod section provides a site
for intimate gas-slurry contact and thus facilitates particulate
and SO2 removal. Spent slurry from the rod section falls into a
collection tank located directly below the venturi section.
Slurry is recycled at 227 liters/s (3600 gal/min) from the col-
lection tank to the rod-deck scrubber by two slurry recirculation
pumps, the one operational, the other spare.
After passing through the rod-deck venturi scrubber, the
flue gas makes a 90-degree turn and traverses the spray tower
approach before making another 90-degree turn. The saturated,
cooled gas enters the spray towers at 165 m /s (349,000 acfm) and
at 52°C (124°F). Additional SO2 scrubbing occurs as the gas
flows upward through the open towers, where it is contacted by
slurry through two levels of countercurrent sprays.
Spent slurry feeds by gravity into a reaction tank directly
below each spray tower. One pump recycles slurry from the
reaction tank to the spray tower at 335 liters/s (5300 gal/min).
Entrained droplets of moisture and slurry are removed in a
mist elimination section in the spray towers. Each mist elimi-
nator has an A-frame design and is comprised of a bulk-entrain-
ment separator followed by two stages of chevron vanes.
After the mist eliminators, the saturated gas stream is
reheated by inline carbon steel rcheaters. One reheater is
provided for each spray tower, and each reheater consists of four
rows of circumferentially finned tubes arranged in a staggered
fashion. The heating medium is hot water from the feed water
deaerator of the boiler. The reheaters increase the temperature
of the gas stream approximately 11°C (20°F). The reheated gas
stream flows through the discharge ducts leading to the induced-
draft fans and stacks, which discharge the gas to the atmosphere.
The Lawrence 5 scrubbing system closely resembles the
Lawrence 4 system. Two scrubbing modules, each consisting of a
rod-deck scrubber with a spray tower absorber, are provided to
treat 100 percent of the flue gas from the steam generator. In
B-45
-------
addition, the system shares the limestone handling and prepara-
tion facilities and the sludge disposal ponds used by Lawrence 4.
Several major features, however, are different:
0 Inlet gas enters the Lawrence 5 scrubbing system at
1858 m3/s (3,937,000 acfm) and 149°C (300°F). Because
the flue gas flow and temperature are greater than
those for the Lawrence 4 system, the modules are
significantly larger.
0 The flue gas is contacted by a single level of slurry
sprays in the spray tower.
0 A single reaction tank receives the spent slurry from
both modules, as well as the fresh limestone slurry
introduced into the system.
0 Lawrence 5 is not equipped with a thickener or any
other dewatering device for separation of solids from
liquid prior to disposal.
42-44
PERFORMANCE HISTORY
No lost boiler capacity has been reported at Lawrence 4
since January 1977, when the limestone slurry spray towers were
started. Although availability has been 100 percent, Lawrence 4
experienced some freezing in the thickener underflow discharge
piping and consequent clarifier plugging in January 1978. The
freezeup problem continued through March 1978, but FGD operations
continued because two fire hoses were used to pump the underflow
solids to the disposal pond. The utility reported no outages
through the middle of September 1978.
Construction continued into the first quarter of 1978 at
Lawrence 5 on the replacement of the original FGD system, which
continued to operate until March 20, 1978. It was then pulled
off the line so that the new system could be tied into the gas
path. The new system began operation on April 14, 1978, and has
continued with 100 percent availability; no forced outages have
been reported through September 1978.
B-46
-------
NORTHERN STATES POWER COMPANY
SHERBURNE 1 AND 2
BACKGROUND INFORMATION45
The Sherburne County Generating Plant is a new four-unit
3160-MW (gross), coal-fired station located in Sherburne County,
near Becker, Minnesota. The station is wholly owned and operated
by the Northern States Power (NSP) Company, which serves the
energy needs of Minnesota, Wisconsin, North Dakota, and South
Dakota. Sherburne 1 and 2 are currently operational. Sherburne
1 was initially placed in service on March 16, 1976, and began
full commercial operation on May 1, 1976. Sherburne 2 was
initially placed in service on January 25, 1977, and started full
commercial operation on April 1, 1977. Sherburne 3 and 4 are
currently in the planning stage and are scheduled to begin opera-
tion in May 1984 and May 1987, respectively.
Sherburne 1 and 2 are each rated at 720 MW (gross) and fire
low-sulfur (0.8 percent sulfur) subbituminous coal from the
Colstrip and Sarpy Creek areas of southern Montana. This coal
has a heating value of 19,800 kJ/kg (8,500 Btu/lb), ash content
of 9 percent, and moisture content of 25 percent.
POLLUTION CONTROL46
To meet particulate and SO2 emission regulations promulgated
by the Minnesota Pollution Control Agency, each unit is equipped
with a two-stage wet scrubbing system using alkaline fly ash and
limestone. The scrubbing systems were designed and supplied by
CE; and each consists of 12 two-stage scrubber modules, 11 of
which are required for full-load Operation. The modules are
i
arranged in a four-by-three matrix. Each module includes a
B-47
-------
variable-throat rod-deck venturi scrubber, a single-stage marble-
bed absorber, a two-stage chevron mist eliminator, and an inline
hot-water reheater. Downstream of the modules are four induced-
draft fans, one for each group of three modules. The induced-
draft fans operate in tandem with two forced-draft fans to com-
pensate for gas-side draft losses in the unit. All spent scrub-
bing slurry is concentrated in a thickener and disposed of in an
onsite, clay-lined settling pond, from which supernatant is
returned to each scrubbing system for additional use.
Alkalinity for S02 comes from the calcium oxide in the
collected fly ash and calcium carbonate in the limestone additive,
Raw limestone rock is received at the plant and ground in wet
ball mills to form a slurry with a solids content of 4 percent.
This slurry is added to the internal reaction tank of each
scrubber module.
The Sherburne scrubbing systems were designed for a minimum
availability of 90 percent while achieving the following particu-
late and SO2 control levels: particulate matter in the flue gas
exiting the scrubber should not exceed 1 percent of the inlet
content or 0.09 g/m (0.04 gr/scf), whichever is greater; SO2 in
the flue gas exiting the scrubber should not exceed 50 percent of
the inlet content or 200 ppm, whichever is greater.
Tables B-12 and B-13 summarize the FGD data for Sherburne
1 and 2.
47 48
PROCESS DESCRIPTION '
Flue gas from the steam generator leaves the air preheaters
at approximately 154°C (310°F) and enters the scrubbing system
through one of the 12 scrubber modules. After passing the inlet
damper, the flue gas goes into the module's first stage, the
rectangular-throat venturi section, which contains two parallel
rows of horizontal rods perpendicular to the flue gas flow. The
slurry spray is introduced just prior to these rods. With the
narrowing of the venturi throat and quick increase in the gas
velocity, particulate capture and SO2 absorption by slurry drop-
lets are enhanced. B-48
-------
TABLE B-12. FGD SYSTEM DATA FOR SHERBURNE 1
Location
Rating (gross), MW
(net), MW
Fuel
Heating value, kJ/kg (Btu/lb)
Ash, percent
Moisture, percent
Sulfur, percent
FGD process
FGD system supplier
Application
Startup date (initial)
(commercial)
Design removal efficiency
Particulate, percent
Sulfur dioxide, percent
Water loop
Total water makeup,
liter/s per net MW (gal/min
per net MW)
Sludge disposal
Becker, Minnesota
720
680
Subbituminous coal
19,300 (8,300)
9.0
25.0
0.8
Alkaline fly ash with limestone
Combustion Engineering
New
March 16, 1976
May 1, 1976
99
50
Open
0.071 (1.13)
Forcibly oxidized sludge is
disposed of in an onsite,
clay-lined settling pond
B-49
-------
TABLE B-13. FGD SYSTEM DATA FOR SHERBURNE 2
Location
Rating (gross), MW
(net), MW
Fuel
Heating value, kJ/kg (Btu/lb)
Ash, percent
Moisture, percent
Sulfur,- percent
FGD process
FGD system supplier
Application
Startup date (initial)
(commercial)
Design removal efficiency
Particulate, percent
Sulfur dioxide, percent
Water loop
Total water makeup,
liter/s per net MW (gal/min
per net MW)
Sludge disposal
Becker, Minnesota
720
680
Subbituminous coal
19,300 (8,300)
9.0
25.0
0.8
Limestone with alkaline fly ash
Combustion Engineering
New
January 25, 1977
April 1, 1977
99
50
Open
0.071 (1.13)
Forcibly oxidized sludge is
disposed of in an onsite,
clay-lined settling pond
B-50
-------
The flue gas, however, still contains a significant portion
of the inlet SO-. It then travels into the main part of the
*•
scrubber, where contact occurs with both the slurry on the glass-
marble bed and the slurry being sprayed to the bed by underbed
spray nozzles.
The flue gas next passes through the two-stage mist elimi-
nation section, consisting of three passes of chevron vanes per
stage. The temperature of the gas at the mist eliminator exit is
55°C (131°F). To prevent condensation of the equilibrium mois-
ture, the gas passes upwards through an inline reheater with
finned tubes heated by hot-water heat. The reheater increases
the scrubbed gas temperature to approximately 77°C (171°F). The
flue gas leaves the reheater and travels through the outlet duct
work to induced-draft fans, two outlet breeching sections, and
finally the flues in the stack (one flue per unit).
There are four impoundment areas: the bottom ash pond, fly
ash pond, recycle basin, and holding basin. These areas function
as the plant's primary solid and liquid waste disposal system.
Bottom ash from the steam generators is disposed of in one diked
impoundment pond, and flue gas cleaning wastes, in another. Both
ponds are clay-lined at the bottom and contain a clay core in the
dikes to minimize seepage. Flue gas cleaning wastes are dis-
charged to the fly ash pond at approximately 38 Mg (42 tons) of
solids per hour at full load.
4Q ^n
PERFORMANCE HISTORY '
The Sherburne 1 and 2 scrubbing systems began operation
simultaneously with the power generating units and were certified
commercial on May 1, 1976, and April 1, 1977, respectively.
Operation during and after startup has revealed a number of
chemical, mechanical, and design problems. The major problems
included failure of the inline slurry recirculation strainers
with consequent severe plugging of nozzles, erosion of the rods
and compartment in the rod-deck venturi scrubbers, erosion of the
sidewalls of the internal reaction tank associated with the
B-51
-------
venturi scrubbers, plugging in the mist eliminators, corrosion of
reheater tubes, premature failure of the rubber lining in the
slurry piping, premature failure of the recirculation pump
impellers, and premature failure of the protective fiberglass
lining in some scrubber modules.
Virtually all problems have been or are being resolved
through system design modifications. These modifications and the
additional operating experience gained by the utility have in-
creased system availability. Since September 1977, the Sherburne
1 and 2 scrubbing systems have achieved system availabilities
averaging 93 and 94 percent, respectively- These averages are
well above the minimum design availability level of 90 percent.
Performance tests and continuous monitoring data at Sherburne
1 and 2 have demonstrated compliance with emission regulations.
Typically, the systems have removed on a dry basis approximately
99 percent of the inlet particulate matter, with an inlet loading
of 6.9 g/m (3.0 gr/scf) and an outlet loading of 0.09 g/m (0.04
gr/scf). From 55 to 60 percent of the inlet SO2 has also been
removed, with a typical inlet SO2 content of 700 ppm and outlet
content of 300 ppm. The only problem encountered to date has
been compliance with opacity requirements. Even when the systems
have met or exceeded design particulate control levels, opacity
has usually ranged from 40 to 45 percent, or about twice the
regulatory limitation.
FUTURE PLANS51'52
Sherburne 3 and 4, each of which is rated at 860 ME (gross),
will be erected alongside Sherburne 1 and 2 and equipped with CE
wet scrubbing systems using alkaline fly ash and limestone. Each
system will include eight two-stage scrubber modules, seven of
which will be required for full-load operation. The modules will
utilize rod-deck venturi scrubbers for particulate control and
vertical countercurrent spray towers for absorption of SC^-
These systems are designed to remove 99.5 percent of inlet
B-52
-------
particulates and 80 percent of inlet SO2. Presently, NSP and CE
are conducting tests on module 101 of Sherburne 1 to determine
the design and operating characteristics of the Sherburne 3 and 4
scrubbing systems.
B-53
-------
SOUTH CAROLINA PUBLIC SERVICE AUTHORITY
WINYAH 2
BACKGROUND INFORMATION
The Winyah Power Station of South Carolina Public Service
Authority (SCPSA) is located in Georgetown, South Carolina.
South Carolina Public Service now operates two of the four units
planned for the station. Winyah 1 and 2 are rated at 315 and 280
MW (gross), respectively. Units 3 and 4 are identical 280-MW
(gross) units planned for commercial operation in 1980 and 1981.
The electric power generating facilities at Winyah 2 consist
of a pulverized-coal-fired steam generator and turbine generator.
Winyah 2 began operation in July 1977.
The coal burned at this unit is medium-sulfur (from 1.0 per-
cent to 1.2 percent sulfur), bituminous Virginia coal with a
heating value of 26,800 kJ/kg (11,500 Btu/lb), ash content of
13.5 percent, and moisture content from 6.5 to 7.0 percent.
POLLUTION CONTROL
A cold-side ESP designed by RC to be 99.4 percent efficient
provides primary particulate control for Winyah 2. The S02 is
controlled with a limestone slurry FGD system supplied by Babcock
and Wilcox.
The FGD system consists of one scrubbing train, which
includes a venturi scrubber and tray tower absorber. The FGD
system scrubs 50 percent of the flue gas with a guaranteed SO2
removal efficiency of 69 percent. Cleaned saturated flue gas
from the FGD system is mixed with hot bypass flue gas for reheat.
When the thickener is in use, spent slurry with a solids
content of 35 percent is discharged at 5.85 Mg/h (6.45 tons/h)
B-54
-------
and pumped to an onsite, unlined, diked pond, capable of holding
the wastes from 20 years of operation. The utility, however, is
currently bypassing the thickener circuit, so that the solids
content of the discharged waste is far below 35 percent.
FGD system data are summarized in Table B-14.
C ^ C A
PROCESS DESCRIPTION '
Flue gas exits the ESP 384 m /s (814,030 acfm) and 132°C
(270°F) with an outlet particulate loading of 43 ng/J (0.1 lb/106
Btu). A forced-draft fan boosts 50 percent of the flue gas
through the scrubbing train. The gas then enters the venturi at
192 m3/s (407,000 acfm) and 132°C (270°F). In the venturi,
additional fly ash is removed, and initial SO~ removal takes
place. The design features of the venturi include a 0.7 kPa (3
in. H2O) pressure drop, a superficial gas velocity of 27 m/s (90
ft/s), and an L/G ration of 1.9 liters/m3 (14.4 gal/103 ft3).
The saturated gas flows at 159 m /s (338,000 acfm) and 52°C
(126°F) to the tray tower, where SO2 removal is completed. The
design features of the absorber include a 1.1 kPa (4.5 in. H_O)
pressure drop, a superficial gas velocity of 3.2 m/s (10.5 ft/s),
and an L/G ratio of 6.4 liters/m3 (47.5 gal/103 ft3). The
cleaned gas flows through the mist eliminator to a common duct
with the hot bypassed flue gas for reheat before discharge to the
stack. The scrubbing solution is a limestone slurry with a
solids content of 20 percent. The slurry is fed to the FGD
system at 130 percent stoichiometry. Particulate removal effici-
ency has averaged 99.55 percent; and total S02 removal efficiency,
35 percent (935 ppm absorber inlet, 294 ppm absorber outlet).
PERFORMANCE HISTORY55'56
Winyah 2 began FGD operations in July 1977 and has experi-
enced virtually 100 percent availability. As of October 1978,
the utility had reported only minor scaling and plugging, which
was rectified by general maintenance. A forced outage related to
B-55
-------
TABLE B-14. FGD SYSTEM DATA FOR WINYAH 2
Location
Rating (gross), MW
(net), MW
Fuel
Heating value, kJ/kg (Btu/lb)
Ash, percent
Moisture, percent
Sulfur, percent
FGD process
FGD system supplier
Application
Startup date (initial)
Design removal efficiency
Particulate, percent
Sulfur dioxide, percent
Absorbers
Overall (50 percent bypass)
Actual removal efficiency
Particulate
Sulfur dioxide
Absorbers
Overall (50 percent bypass)
Water loop
Total water makeup,
liter/s per net MW (gal/min
per net MW)a
Sludge disposal
Georgetown, South Carolina
280
258
Coal
26,800 (11,500)
13.5
6.5-7.0
1.0-1.2
Limestone
Babcock and Wilcox
New
July 1977
99.4
69
34.5
99.4
85.0
42.5
Open
0.045 (0.714)
Unstabilized sludge is disposed
of in an onsite, unlined pond
At 140-MW FGD capacity.
B-56
-------
the FGD system occurred in January 1978 because of problems with
the S02 continuous analyzers. The SO. analyzers required several
sample line replacements and eventually heat tracings around the
units.
Performance data for availability, operability, reliability,
and utilization have not yet been disclosed by SCPSA.
CT co
FUTURE PLANS '°
Two additional units, each rated at 280 MW (gross), are
planned by SCPSA at Winyah Station. Winyah 3 and 4 will use RS
steam generators. Winyah 3 will use ESP's from RC for particu-
late collection and an FGD system from Babcock and Wilcox for SO_
removal. Winyah 3 will scrub 100 percent of the flue gas and,
unlike Winyah 2, will include an indirect steam-coil reheat
system that will heat ambient air and inject it into the FGD
system exit. Winyah 3 is scheduled to start up in May 1980.
Winyah 4 is scheduled to commence operation in May 1982.
B-57
-------
SOUTHERN MISSISSIPPI ELECTRIC
R. D. MORROW 1
BACKGROUND INFORMATION59
R. D. Morrow Power Station of Southern Mississippi Electric
is a two-unit station (Unit 2 is under construction) in Hattisburg,
Mississippi. R. D. Morrow 1 and 2 will be identical 200-MW
(gross) units.
The electric power generating facilities at R. D. Morrow 1
consist of an RS pulverized-coal-fired steam generator with
turbine generator. R. D. Morrow 1 began operation in August
1978.
The coal burned at this unit is medium-sulfur (1.0 percent
sulfur) coal with a heating value of 27,900 kJ/kg (12,000 Btu/lb)
and an ash content of 8 percent. The coal (as burned) specifi-
cations will be altered when the coal prewashing facility that is
under construction goes into operation.
POLLUTION CONTROL60
A Buell-Envirotech ESP with a design efficiency of 99.6
percent provides primary particulate control for R. D. Morrow 1.
The SO~ is controlled with a limestone slurry FGD system supplied
by Riley Stoker/Environeering. The system consists of a venturi
rod scrubber with a mild steel (carbon steel) shell and a glass-
flake liner (including the outlet duct work). The inlet duct
work made of carbon steel. The scrubber includes seven layers of
stainless steel rods that are 1.9 cm (0.75 in.) in diameter and
spaced roughly 2.5 cm (1 in.) apart. The scrubber is 27 m (90
ft) high, 12 m (40 ft) wide, and 3 m (10 ft) deep. The ESP
removes particulates from 100 percent of the flue gases, and the
B-58
-------
FGD module scrubs 62 percent of the flue gas with a guaranteed
SO2 removal efficiency of 85 percent. The FGD module includes a
mist eliminator near the flue gas exit and a reheat system that
allows 38 percent of the hot flue gas to bypass the scrubber and
rejoin the cleaned saturated gas at the scrubber gas exit in a
single duct to the stack. The scrubber pressure drop is 2 kPa (8
in. H2O). The overall SO2 removal efficiency is designed to be
53 percent.
The slurry circulated through the scrubber is maintained at
a minimum pH of 5 and a maximum solids level of 20 percent. The
spent slurry is concentrated in a thickener, and the thickened
underflow is stabilized with fly ash in a pug mill after being
dewatered by a vacuum filter to a solids content of 60 percent.
Stabilized sludge is then trucked to an offsite landfill. The
thickener overflow is recirculated back to the FGD system.
Table B-15 summarizes FGD system data for R. D. Morrow 1.
PROCESS DESCRIPTION61
Flue gas exits the boiler and flows into the hot-side ESP,
designed to remove 99.6 percent of particulates. The flue gas
then passes through the boiler air preheater. An induced-draft
fan draws 62 percent of the hot gas off to the scrubbing system
at 192.1 m3/g (406,940 scfm) and 132°C (270°F). The 38 percent
of the flue gas that bypasses the scrubbing system is used to
reheat the wet scrubbed gas so that the total stack gas tempera-
ture remains above the dew point and acid fallout is avoided.
The flue gas enters the base of the rod-deck scrubber
module and flows upward. After turning a 90-degree angle, it
encounters limestone slurry in a countercurrent fashion through
the turbulence created by the sprays, the venturi rods, and the
gas flow. The cleaned, saturated gas turns 90 degrees back
toward a vertical mist eliminator section and exits the scrubber
at 165.9 m3/g (351,455 scfm) and 52°C (126°F).
B-59
-------
TABLE B-15. FGD SYSTEM DATA FOR R.D. MORROW 1
Location
Rating (gross), MW
(net) , MW
Fuel
Heating value, kJ/kg (Btu/lb)
Ash, percent
Moisture, percent
Sulfur, percent
FGD process
FGD system supplier
Application
Startup date (initial)
Design removal efficiency
Particulate (ESP), percent
Sulfur dioxide, percent
Absorber
Overall (62 percent of gas
scrubbed)
Water loop
Total water makeup,
liter/s per net 'MW (gal/min
per net MW)
Sludge disposal
Hattisburg, Mississippi
200
180
Coal
27,900 (12,000)
8.0
8.0
1.0
Limestone
Riley Stoker/Environeering
New
August 1978
99.6
85
52.7
Closed
0.11 (1.77)
Stabilized*, dewatered sludge
is trucked to an offsite
landfill
B-60
-------
PERFORMANCE HISTORY62'63
The FGD system at R. D. Morrow 1 began operation in August
1978. As of October 1978, the unit was still involved with
shakedown and debugging operations. Because of the recent
startup, typical performance information is not available.
Boiler tube leaks kept the unit off-line until November I, but no
FGD-related outages have been reported (100 percent availability),
FUTURE PLANS64'65
Southern Mississippi Electric is installing another unit
scheduled to begin operation in January 1979. R. D. Morrow 2
will be identical to R. D. Morrow 1, with a 200-MW steam gener-
ator, Buell-Envirotech ESP designed to be 99.6 percent efficient,
and limestone slurry rod-deck scrubber treating 62 percent of the
flue gas. Construction is approximately 85 percent complete.
B-61
-------
SPRINGFIELD CITY UTILITIES
SOUTHWEST 1
BACKGROUND INFORMATION66
The Southwest Power Station of Springfield City Utilities is
a one-unit station (with plans for a second identical unit)
approximately 8 km (5 miles) southwest of Springfield, Missouri.
Southwest 1 is rated at 194 MW (gross), with generating facili-
ties that consist of a Riley Stoker pulverized-coal-fired steam
generator with a turbine generator. Southwest 1 began operation
in April 1977 and was declared commercial in September 1977.
The coal burned at this unit is bituminous high-sulfur (3.5
percent sulfur) Kansas coal with a heating value of 29,100 kJ/kg
(12,500 Btu/lb) and an ash content of 13 percent.
POLLUTION CONTROL67'68
A four-field, cold-side ESP designed to be 99.7 percent
efficient provides primary particulate control for Southwest 1,
and SO2 is controlled by a limestone slurry FGD system. The ESP
and FGD systems are supplied by the Air Correction Division of
UOP. In the FGD system, there are two parallel TCA modules, each
of which can handle 50 percent of total capacity. These modules
scrub 100 percent of the flue gas. Southwest 1 also has facili-
ties for limestone crushing and preparation and sludge handling.
One or both FGD modules can be bypassed during an emergency or
malfunction by the use of seal-air gas dampers. The S02 removal
efficiency of the FGD facility is designed to be 80 percent.
Scrubbed, saturated flue gas is discharged to the atmosphere
through the main stack.
B-62
-------
TABLE B-16. FGD SYSTEM DATA FOR SOUTHWEST 1
Location
Rating (gross), MW
(net), MW
Fuel
Heating value, kJ/kg (Btu/lb)
Ash, percent
Moisture, percent
Sulfur, percent
Chloride, percent
FGD process
FGD system supplier
Application
Startup date (initial)
(commercial)
Design removal efficiency
Particulate (ESP), percent
Sulfur dioxide, percent
Actual removal efficiency
Particulate (ESP), percent
Sulfur dioxide, percent
Water loop
Sludge disposal
Springfield, Missouri
194
173
Bituminous coal
29,100 (12,500)
13
6-12
3.5
0
Limestone
Air Correction Division, UOP
New
April 1977
September 1977
99.7
80.0
99.8
92.0
Closed
Unstabilized sludge is
dewatered and disposed of in
an onsite landfill
B-63
-------
Flue gas cleaning wastes are discharged from the recircula-
tion tanks to a thickener, where solids are separated; the
clarified liquor is recycled. Sludge from the thickener under-
flow is dewatered by a rotary drum vacuum filter; and the filter
2
cake is trucked to an onsite 40,000-m (10-acre) valley landfill.
FGD data are summarized in Table B-16.
PROCESS DESCRIPTION69'70
Flue gas from the steam generator passes through the ESP for
particulate removal and is boosted through the two TCA modules by
induced-draft fans. The TCA modules are designed for 316 m /s
(665,000 acfm) of gas at 149°C (300°F). Flue gas flows into the
base of one of the TCA modules, passes through a presaturator,
and rises through two levels of packing consisting of spheres 3.2
cm (1.25 inch) in diameter; limestone slurry is contacted in a
countercurrent fashion. Each module is equipped with a separate
recirculation tank where reactions are completed and fresh slurry
is added. Flue gas cleaning wastes are discharged from these
tanks. The cleaned saturated flue gas then passes through mist
eliminators before it is released into the atmosphere through the
117-m (384-ft) main stack. There is no reheat system.
PERFORMANCE HISTORY71'72
Federal and State emission regulations require that South-
west 1 limit emissions of particulate to 43 ng/J (0.1 lb/10 Btu)
heat input, SO to 516 ng/J (1.2 lb/10 Btu) of heat input, and
6
nitrogen oxide (NO ) to 301 ng/J (0.7 lb/10 Btu) of heat input.
X
Compliance tests were conducted on September 14 and 15, 1977.
These tests showed an average ESP inlet dust loading of 3143 ng/J
(7.31 lb/106 Btu) and outlet loading of 7.91 ng/J (0.0184 lb/106
Btu); the average ESP efficiency was thus 99.75 percent. The
average inlet S00 concentration was 2,726 ng/J (6.34 lb/10 Btu);
6
and the average outlet SO2 concentration, 226 ng/J (0.526 lb/10
Bty). Typically, therefore, the scrubber removed 91.7 percent of
B-64
-------
SO-. The average NO emissions were found to be 229 ng/J (0.672
6 "
lb/10 Btu), which was below the emission regulation requirement.
Since September 1977, a number of modifications have taken
place. During a shutdown from October 1977 through January 1978,
the scrubbing system's outlet damper was replaced with 316L
stainless steel (SS). A ball failure prompted the replacement of
the hollow plastic spheres in the TCA packing with plastic
spheres having thicker walls. The discharge ducts were relined
with Ceilcote; insulation was installed around the seal air fan
between the inlet dampers; and the presaturators were relined
with high-molybednum steel.
Between Feburary and March 1978, an expansion fount failed
on B-module between the induced-draft fan and absorbers. The
scrubber was bypassed to keep the steam generator facility oper-
ating.
An FRP-liner failure occurred between April and May 1978, as
well as a pump failure and a gas damper rupture. Thus, only one
TCA module could be operated.
In June 1978, the mist eliminator wash system was changed
from a separate closed loop for each module to a common system
for both modules. The FGD facility was then plagued with instru-
mentation problems; and the B-module was inoperable because of
the unrectified problem with the expansion joint.
By September 1978, the unit had achieved an average total
FGD system availability of only 25 percent. The B-module expan-
sion joint had still not been corrected, and another ball failure
in the TCA packing necessitated replacement of the plastic
spheres with solid rubber balls. From July through September
1978, the instrumentation problems were largely rectified,
although some slurry line plugging occurred.
FUTURE PLANS73
Springfield City Utility's plans call for a second 194-MW
(gross) unit at Southwest Power Station. At the time of this
report, projections for Unit 2 indicated a 1981 startup.
B-65
-------
TENNESSEE VALLEY AUTHORITY
WIDOWS CREEK 8
BACKGROUND INFORMATION74
Widows Creek Power Station of Tennessee Valley Authority
(TVA) is an eight-unit station located near Bridgeport, Alabama.
Units 1, 2, 4, 5, and 6 are each rated at 140 MW (gross). Unit 3
is rated at 150 MW (gross). Units 7 and 8 are rated at 575 and
550 MW (gross), respectively. The total station generating
capacity is 1975 MW (gross).
The electric power generating facilities at Widows Creek 8
consist of a Combustion Engineering tangential-fired, balanced-
draft, coal-fired steam generator with associated turbine gen-
erator. Widows Creek 8 was first placed in service in 1963 and
has a net generating capacity of 516 MW.
The coal burned in this unit is bituminous, high-sulfur (3.7
percent sulfur) coal with a heating value of 26,200 kJ/kg (11,270
Btu/lb) and an ash content of 17 percent.
POLLUTION CONTROL75'76
A 50 percent (actual) efficient ESP supplied by Koppers
provides primary particulate control. Secondary particulate
control and primary SO- control are provided by a limestone
scrubbing system designed and installed by TVA, with the scrubber
modules fabricated by Polycon.
The scrubbing system consists of four parallel trains. Each
train includes a variable-throat venturi scrubber followed by a
multigrid tower absorber. The venturi sections are constructed
of 316L SS; the multigrid towers, of carbon steel (shells) with a
rubber lining and 316L SS on the lower sloping (downcomer)
B-66
-------
sections, where the slurry falling from the grids is funneled to
the recycle system. Inside each tower are five grids. The
first, third, and fifth grids are made of 316L SS; the second and
fourth, of fiberglass reinforced plastic. Plans for one of the
towers call for the installation of packing that will be supported
by the grids.
The scrubbing system includes a four-pass vertical chevron
mist eliminator constructed of 316L SS. The mist eliminator
system is equipped with a continuous wash system.
Reheat of the wet gas is provided by an indirect hot-air
steam-coil system that raises the wet gas temperature by 28°C
(50°F), to a point well above the dewpoint.
Actual particulate removal efficiency at Widows Creek 8 is
99.5 percent. Actual SO2 removal efficiency ranges from 85 to 94
percent.
The. FGD system data are summarized in Table B-17.
PROCESS DESCRIPTION77'78
Flue gas entering the scrubbing system from the ESP's passes
through an induced-draft fan with a capacity of 190 m /s (400,000
acfm). The gas then enters one of the four variable-throat
Venturis. Each venturi is approximately 7 m (23 ft) wide and 8.5
m (28 ft) deep at the variable throat. Approximately 10 percent
of the SO2 is removed in the Venturis. The wet flue gas then
enters a multigrid through which it rises at 3.7 m/s (12 ft/s)
and encounters the slurry in a countercurrent fashion. After
turning through a four-pass vertical chevron-type mist elimina-
tor, the gas joins air drawn from the powerhouse and pumped
through steam coils that heat the air to 204°C (400°F). The
stack gas temperature is 79.4°C (175°F).
7Q_OO
PERFORMANCE HISTORYty °*
Since the scrubbing system began operations in May 1977, the
utility has reported a number of operational difficulties.
B-67
-------
TABLE B-17. FGD SYSTEM DATA FOR WIDOWS CREEK 8
Location
Rating (gross), MW
(net) , MW
Fuel
Heating value, kJ/kg (Btu/lb)
Ash, percent
Moisture, percent
Sulfur, percent
FGD process
FGD system supplier
Application
Startup date (initial)
(commercial)
Design removal efficiency
Particulate (overall), percent
Sulfur dioxide, percent
Actual removal efficiency
Particulate (overall), percent
Sulfur dioxide, percent
Water loop
Sludge disposal
Bridgeport, Alabama
550
516
Coal
26,200 (11,270)
17
10
3.7
Limestone
Tennessee Valley Authority
Retrofit
May 1977
January 1978
99.5
80.0
99.5+
85-94
Open
Unstabilized sludge is
disposed of in an onsite,
unlined pond
B-68
-------
Because of poor ESP performance, the induced-draft fans must
handle gas with a high grain load, 11.4 to 13.7 g/m (5 to 6
gr/scf). This load and the high rotor tip speed effectively
sandblast the rotors. The resulting erosion causes fan ineffi-
ciency and imbalance. Excessive vibration necessitates removing
the fan from service. The fan housing also suffers severe
erosion.
Each scrubber-absorber train has three bottom-entry guillo-
tine dampers, one each for the gas inlet, outlet, and bypass
ducts. Initial problems with the dampers resulted from corrosion
of the 316 SS seals. Because of the absence of housings and
bonnets on the dampers and because the FGD system is pressurized,
the dampers have been leaking ash and flue gas excessively. The
drive mechanisms below the guillotine dampers have collected ash
in the threads, boots, and jack screw assemblies.
The utility reported that the dampers were inoperable,
except with extraordinary effort, because of electrical component
failure, deterioration of boots, and drive mechanism jamming.
Holes and tears have developed at numerous places in the expan-
sion joint fine-ply fabrics.
The most serious operating problem has been the failure of
rubber linings. A mid-June 1977 absorber inspection revealed
that rubber lining on the downcomer of Train C had come loose;
and bare metal was exposed. Patches of loose rubber and bare
metal were found in other trains, but only in the sloping sec-
tions of the absorber and venturi hoppers. Replacement material
was obtained and repair work was carried out in early August
1977.
A September 26, 1977, inspection of the trains during a
boiler outage revealed that rubber lining on Train C was again
missing and that the slurry had worn a hole through the carbon
steel material of the venturi section hopper wall sloping towards
the downcomer. A temporary wear plate was installed to permit
operation until the scheduled boiler maintenance outage in October
B-69
-------
These defects, as well as other sections of loose rubber that
were found during this same inspection, were all corrected during
the October outage.
A January 24, 1978, inspection again revealed that large
sections of liner were missing and that holes were appearing in
scrubber train carbon steel shells. The utility decided to
replace the rubber lining in the sloping sections with 316L SS
plate welded to the carbon steel shell.
Evidence of rubber liner failures has also been found in the
centrifugal pumps that send slurry from the ball mill product
tank to the classifiers. The liners in these pumps have required
high maintenance, possibly because of poor cyclone performance or
improper pump speeds.
The sumps in the ball mill house and near the spent slurry
surge tanks have been flooding. Limestone spills down into these
tanks, and limestone particles settle at the bottoms of the
sumps. It has thus been difficult to empty waste from these
sumps to the pond.
The wet grind, rubber-lined ball mill, designed for a
capacity of 45 Mg/h (50 tons/h) was not functioning as intended.
During initial shakedown of the system, excessive rock and ball
rejection occurred at rates above 18 Mg/h (20 tons/h). Because
coarse reject material was falling into the mill slurry product
tank, a chute was installed to convey the material into a barrel
and avoid contamination of the finely ground product. Inspection
of the product end of the ball mill revealed that the helix
designed to retain balls and rock in the grinding chamber was
inadequate. A perforated plate was spot welded to the existing
helix and extended over the trommel screen; this plate has
satisfactorily resolved the problem.
Multigrids and absorber nozzles have presented problems.
The absorber section of each train has five trays of multigrids.
Originally, the bottom three of these were 316 SS, and the top
two were FRP- Poor spray distribution from nozzles created
obvious erosion of the FRP grids; and the SS grids were brought
B-70
-------
to the top, just below the spray nozzles. Some adjustment in the
nozzle cones improved the spread of the spray but a satisfactory
flow pattern had not yet been attained at the time of this
report. Because of the nozzle design and slurry flow, spray
cones have fallen off or have worn loose from part of their
supports.
Many instrument problems have resulted from poor initial
placement or inappropriate application. Particularly crucial are
the pH monitors, which are submerged flow-through meters. These
meters give readings on the absorber slurry flowing to the
nozzles. The transmitters associated with the pH meters vibrated
excessively when placed near the top of the absorber circulation
tanks. Relocation of the meters to ground level and replacement
of the reference electrode liquid seems to have improved relia-
bility and performance. There has also been poor control of the
feed rate of fresh limestone slurry because of a nonlinear
response from a linear control signal.
Pressure drop instruments for the gas are excessively
sensitive to ambient temperature changes and often go out of
calibration. Annubars for gas flow and outlet SO2 monitors are
located downstream of the mist eliminator section of the scrubber
and have been unreliable as a result of moisture and particulate
plugging. Bracing and reverse air purging have improved instru-
ment performance somewhat.
There has been nozzle and spray header plugging in the
venturi section, partly because of excessive amounts of calcium
sulfate in the slurry. This occurred because of poor pH control,
which results from unreliable pH meters and poor control of fresh
limestone feed.
Heavy solid particles that solidify upon drying appear to be
settling in the venturi nozzle header. Until a steady state is
reached and a complete chemical analysis of the system can be
made, the cause of this settling cannot be identified.
Since May 1977, the average FGD system availability has been
64 percent.
B-71
-------
FUTURE PLANS83'84
The TVA is currently retrofitting a Combustion Engineering
scrubbing system on Widows Creek 7, so that operations can begin
in October 1980. The facility under construction is a limestone
slurry spray tower system.
The TVA also has plans for FGD systems on their Paradise and
Johnsonville stations.
B-72
-------
TEXAS UTILITIES
MARTIN LAKE 1 AND 2
BACKGROUND INFORMATION85
Martin Lake Steam Electric Station of Texas Utilities
consists of four 750-MW (gross) units located in Tatum, Texas.
Martin Lake 1 and 2 are operational units. Martin Lake 3 is
under construction, and Martin Lake 4 is in the planning stage.
The electric power generating facilities at Martin Lake 1
and 2 consist of Combustion Engineering lignite-fired steam
generators with 750-MW (gross) turbine generators. Martin Lake 1
began operation in August 1977; Martin Lake 2 in May 1978.
The coal burned at this station is low-sulfur (0.9 percent
sulfur) Texas lignite with an average heating value of 17,170
kJ/kg (7,380 Btu/lb) and an ash content of 8.0 percent.
fifi
POLLUTION CONTROL
Cold-side ESP's designed to be 99.4-percent efficient pro-
vide primary particulate control for Martin Lake 1 and 2. The
SO2 is controlled with limestone slurry FGD systems. The ESP's
and FGD systems on both units were supplied by RC.
The FGD system on each unit consists of six parallel packed
spray tower absorbers that scrub a maximum of 75 percent of the
boiler flue gas with an efficiency of 95 percent (design) to
achieve an overall SO2 removal efficiency of 71 percent. Each
absorber tower includes a three-stage SO2 removal configuration
followed by a two-stage mist elimination system.
Reheat is provided by the hot flue gas that goes through the
ESP's, but bypasses the absorbers. In a duct before the stack,
this unscrubbed gas rejoins the saturated flue gas from each
absorber. _ _
B— / o
-------
Spent scrubbing slurry from the FGD system is dewatered and
mixed with fly ash and transported by rail to a landfill.
FGD system data for Martin Lake 1 and 2 are summarized in
Tables B-18 and B-19, respectively.
Q 7_O Q
PROCESS DESCRIPTION °*
The process description provided applies to both Martin Lake
1 and 2. The Martin Lake 1 and 2 flue gas cleaning facilities
are separate systems installed on separate units.
Flue gas exits the boiler, passes through the air preheater,
and enters the ESP's at 1495 m3/s (3,167,000 acfm) and 168°C
(335°F). From the ESP's, the gas flows into a header for four
induced-draft fans that draw the gas into a second header feeding
the six absorbers and the bypass duct. The flowrate through each
absorber and through the bypass duct is controlled by balancing
dampers at every junction. Flue gas enters the base of each
absorber tower at 177.2 m3/s (375,350 acfm) and 168°C (335°F).
The gas rises through a cyclonic quenching chamber into the
absorber section of the tower at 142.9 m /s (302,870 acfm) and
57°C (135°F); it goes through a layer of sprays, packing, and
another layer of sprays. After passing a horizontal two-stage
mist eliminator near the top of the absorber tower, the saturated,
2
cleaned gas exits the tower at 144.2 m /s (305,500 acfm) and 57°C
2
(135°F) and mixes with hot, bypass gas at 541.2 m /s (1,146,650
acfm) and 168°C (335°F) from the second header duct. The com-
bined gas stream enters the stack at 1264.2 m /s (2,678,760 acfm)
and 97°C (207°F).
Spent scrubbing slurry with a solids content of 15 percent
is discharged from each absorber tower into a single stream that
feeds a gravity thickener 43 m (140 ft) in diameter. The scrub-
bing wastes are concentrated to a solids content of 35 percent,
and the thickener underflow is fed to one of three centrifuges
for additional dewatering. Filter cake with a solids content
between 68 and 70 percent discharge from the centrifuges into a
B-74
-------
TABLE B-18. FGD SYSTEM DATA FOR MARTIN LAKE 1
Location
Rating (gross), MW
Fuel
Heating value, kJ/kg (Btu/lb)
Ash, percent
Moisture, percent
Sulfur, percent
FGD process
FGD system supplier
Application
Startup date (initial)
Design removal efficiency
Particulate (ESP), percent
Sulfur dioxide, percent
Absorbers
Overall
Water loop
Total water makeup,
liter/s per netMW (gal/min per
net MW)
Sludge disposal
Tatum, Texas
750
Lignite
17,166 (7,380)
8.0
33.0
0.9
Limestone
Research Cottrell
New
August 1977
99.4
95.0
70.5
Closed
0.05 (0.73)
Stabilized, dewatered sludge
is disposed of in an onsite
landfill
B-75
-------
TABLE B-19. FGD SYSTEM DATA FOR MARTIN LAKE 2
Location
Rating (gross), MW
Fuel
Heating value, kJ/kg (Btu/lb)
Ash, percent
Moisture, percent
Sulfur, percent
FGD process
FGD system supplier
Application
Startup date (initial)
Design removal efficiency
Particulate (ESP), percent
Sulfur dioxide, percent
Absorbers
Overall
Water loop
Total water makeup,
liter/s per net MW (gal/min oer
net MW)
Sludge disposal
Tatum, Texas
750
Lignite
17,166 (7,380)
8.0
33.0
0.9
Limestone
Research Cottrell
New
May 1978
99.4
95.0
70.5
Closed
0.05 (0.73)
Stabilized, dewatered sludge
is disposed of in an onsite
landfill
B-76
-------
Muller-type blender. At this point the cake combines with fly ash
collected in the ESP to form a material that can be conveyed by
truck and dumped. Railcars receive the blend for ultimate dis-
posal in a landfill.
9n QI
PERFORMANCE HISTORY Uf*
Since Martin Lake 1 began operations in August 1977, a
number of problems have occurred; and many modifications have
been made. Information is largely unavailable for Martin Lake 2.
It is assumed, however, that the Unit 1 modifications have been
taken into consideration for possible incorporation in the Unit
2 design.
The utility has had a great deal of trouble isolating
individual absorbers at Martin Lake. Dampers provided for the
Martin Lake FGD system include a single-louver bypass damper, two
consecutive louver dampers at each tower inlet, and single link-
louver dampers at each tower outlet. Even with the addition of
seal air blowers between the two tower inlet dampers, isolation
of individual towers for personnel access and maintenance has
been extremely difficult. Areas of malfunction have included
bearings, seal strips, and linkages.
All gas-side expansion joints have been replaced because the
original joints could not withstand exposure to slurry and dete-
riorated. The replacement joints were performing adequately as
of October 1978.
Slurry line and spray nozzle plugging that required pipe
disassembly and mechanical cleaning was discovered during shut-
downs in early 1978. Such plugging was partly attributable to
accumulated construction debris, pieces of fallen wetted film
contactor (WFC) packing, and WFC support beams from October and
November 1977.
Sonic flow switches, installed to monitor flow deviations in
key process loops were found to be inadequate for an electric
generating station. They were eliminated, where possible, or
replaced with magnetic flowmeters.
B-77
-------
Submerged pH meters in the quencher instrument wells exper-
ienced calibration drifts and electrical problems. After an
extensive test program, the meters were replaced by flow-through
devices with ultrasonic cleaners.
Sonic devices at liquid level in the quencher instrument
wells experienced problems related to electronics, poor cali-
bration, slurry foaming, fouling of transducers, and excessive
variations in liquid level. These problems were addressed; and
the devices have functioned adequately since October 1977,
although modifications were still being made at the time of this
report because of transducer problems.
Poor gas distribution at the absorber tower inlets led to
the relocation of the gas flowmeters at the tower outlets. High
amounts of moisture from the air and probe plugging were thought
to have contributed to the poor performance of the meters.
Electric on/off operators on reagent feed and density
control valves experienced numerous failures in early operation.
Some electric operators were replaced with pneumatic operators,
and others were modified to improve performance.
The original FRP packing support beams were replaced with
316L SS beams in January and February 1978. The FRP beams had
failed completely in one tower and were on the verge of failure
in several other towers when the WFC packing was removed from all
towers on November 7 and 8, 1977.
Failure of these beams resulted from inadequate structural
design, but was accelerated because solids accumulated in the
WFC. The accumulations of solids were caused by operating the
FGD system out of chemical balance. Such operation resulted from
malfunctioning pH devices and on/off valve operators and the
difficulty in manual control that these malfunctions created.
Poor slurry distribution to the WFC may also have accelerated
this buildup. Later operation with 0.6 m (2 ft) of WFC packing
and with the new pneumatic valve operators produced insignificant
buildup in the WFC and mist eliminator packings.
B-78
-------
Scale buildup at the cyclonic inlets of the absorber towers
was discovered in November 1977. Model testing showed that modi-
fications to the inlets would allow operation of the towers for
a full year without cleaning. These modifications have been
successful and will be made on all towers.
Martin Lake 1 and 2 underwent acceptance tests during
August 1978. As of October 1978, the results had not been
reported. Experimental SO2 removal efficiencies, however, have
been reported by RC. Absorber towers with 1.2 m (4 ft) of WFC
packing have achieved SO2 removal greater than 99 percent.
Towers with no WFC packing removed from 80 to 85 percent of SO2
at peak tower gas velocities. This testing prompted the installa-
tion of 0.6 m (2 ft) of WFC packing in each tower and the in-
crease of maximum tower gas throughput by 10 percent. Typical
performance information is unavailable for Martin Lake Station.
FUTURfc PLANS92'93
Texas Utilities has plans for two more 750-MW units at
Martin Lake Station. The startup dates for Martin Lake 3 and 4
are December 1978 and November 1982, respectively.
Texas Utilities also plans an additional 750-MW unit,
Forest Grove 1, near Athens, Texas, scheduled for startup in
1982.
B-79
-------
TEXAS UTILITIES
MONTICELLO 3
BACKGROUND INFORMATION94
Monticello Steam Electric Station of Texas Utilities is a
three-unit facility located in Mt. Pleasant, Texas. Monticello
1 and 2 are 575-MW (gross) units. Monticello 3 is a 750-MW
(gross) unit.
The electric power generating facilities for Monticello 3
consist of a CE lignite-fired steam generator with 750-MW (gross)
turbine generator. Monticello 3 began operation in May 1978.
The coal burned in this unit is medium-sulfur (1.5 percent
sulfur) lignite with a heating value of 15,100 to 17,500 kJ/kg
(6,500 to 7,500 BTU/lb) and an ash content of 18.9 percent.
95
POLLUTION CONTROL
Two parallel ESP's designed to be 99.5 percent efficient and
supplied by Pollution Control-Walther provide primary particulate
control for Monticello 3. The S0_ is controlled with a limestone
slurry FGD system supplied by Chemico.
The FGD system consists of three parallel spray towers that
scrub 100 percent of the flue gas with a design SO2 removal
efficiency of 74.0 percent. The design includes a single-stage
horizontal mist eliminator and two emergency bypass ducts. The
saturated, cleaned gas is reheated by an indirect hot-air-injec-
tion system prior to discharge. The reheat system consists of
two separate parallel units, each of which uses a fan to force
ambient air through steam coils. The hot air stream is mixed
with the scrubber discharge gases in a common duct leading from
the modules to the stack.
B-80
-------
Stabilized sludge is disposed in an onsite lined pond.
FGD system data are summarized in Table B-20.
PROCESS DESCRIPTION96
Flue gas exists the boiler air heater at 1650 m /s (3,490,000
acfm) and 168°C (335°F) and passes through the ESP's. The gas
exits the ESP's through four parallel ducts to a header, from
which it is drawn by three parallel induced-draft fans at 1618
3
m /s (3,428,000 acfm) and 168°C (335°F). The three gas streams
enter a second duct at 1586 m /s (3,359,700 acfm) and 176°C
(348°F). The second duct feeds the three spray towers and two
emergency bypass ducts. Flue gas flows to the base of each spray
tower at 520 m3/s (1,100,000 acfm) and 176°C (348°F); it passes
through four banks of spray headers and contacts the limestone
slurry in a countercurrent fashion. The scrubbed gas passes
through a single-stage horizontal mist eliminator and exits the
absorber at 428 m /s (907,000 acfm) and 61°C (141°F); it passes
to a common duct serving all three modules and the two emergency
bypasses. The cleaned saturated gas is reheated by two parallel
indirect steam coil reheaters that feed heated ambient air into
the common duct prior to the stack at the rate of 164 m /s
(346,000 acfm) and 149°C (300°F) each. The gas stream enters the
stack at 1592.3 m3/s (3,373,912 acfm) and a 70.8°C (159.5°F).
Q7 Qfl
PERFORMANCE HISTORY ''*°
Monticello 3 began operation in May 1978 with only one of
the three spray tower absorbers operating. As of October 1978,
reports indicated that the unit was fully operational; but the
compliance test had not been initiated. Typical performance
information has not been reported by the Utility.
B-81
-------
TABLE B-20. FGD SYSTEM DATA FOR MONTICELLO 3
Location
Rating (gross), MW
Fuel
Heating value, kJ/kg (Btu/lb)
Ash, percent
Sulfur, percent
Chloride, percent
FGD process
FGD system supplier
Application
Startup date (initial)
Design removal efficiency
Particulate (ESP), percent
Sulfur dioxide, percent
Water loop
Total water makeup,
liter/s per net MW (gal/min per
net MW)
Sludge disposal
Mt. Pleasant, Texas
750
Lignite
15,100-17,500 (6,500-7,500)
18.9
1.5
0.4
Limestone
Chemico
New
May 1978
99.5
74.0
Closed
0.046 (0.729)
Stabilized sludge is dis-
posed of in an onslte lined
sludge pond
B-82
-------
APPENDIX B
REFERENCES
1. PEDCo in-house files.
2. MeIIvane Co. Details on Alabama Electric Coop. The Wet
Scrubber Newsletter, 29:7, November 30, 1976.
3. PEDCo in-house files.
4. Melia, M., et. al. EPA Utility FGD Survey: August-September
1978. Preliminary Report. Prepared for the U.S. Environ-
mental Protection Agency under Contract No. 68-02-2603.
PEDCo Environmental, Inc., Cincinnati, Ohio. November 1978.
p.4.
5. PEDCo in-house files.
6. Op. cit. No. 4.
7. PEDCo in-house files.
8. Laseke, B.A., Jr. Survey of Flue Gas Desulfurization System:
Cholla station, Arizona Public Service Co. EPA-600/7-78-
048a. PEDCo Environmental, Inc., Cincinnati, Ohio. March
1978. pp. 2, 4, 6, 9, 11, 13, 22, 23, 26, 27, 29.
9. Ibid.
10. Op. cit. No. 4. pp. 4 and 24-25.
11. Ibid.
12. Op. cit. No. 4.
13. Op. cit. No. 4. pp. 4 and 24-25.
14. Laseke, B.A., Jr. Survey of Flue Gas Desulfurization Systems:
Duck Creek Station, Central Illinois Light Company. Pre-
liminary draft prepared for the U.S. Environmental Protection
Agency under Contract No. 68-02-2603. PEDCo Environmental,
Inc., Cincinnati, Ohio. December 1978. pp. 2, 4, 5, 7, 10,
11-14, 16, 18, 21, 23, 28, 32, 35, 37, 44, 45, 46, 48-50.
«
15. Op. cit. No. 4. pp. No. 6, 26.
B-83
-------
16. Ibid.
17. PEDCo in-house files.
18. Op. cit. No. 14. pp. 2, 4, 5, 7, 10-14, 18, 21, 23, 28, 32,
35, 37, 44-46, 48-50.
19. Op. cit. No. 14. pp. 6, 26.
20. PEDCo in-house files.
21. Ibid.
22. Idemura, H., T. Kanai, and H. Yanagioka. Jet Bubbling Flue
Gas Desulfurization Process. In: Proceedings of the Second
Pacific Chemical Engineering Congress (PAChEC *77) . American
Institute of Chemical Engineers, New York, New York. 1977.
pp. 365-370.
23. Idemura, H., and D.D. Clasen. Limestone/Gypsum Jet Bubbling
Scrubbing System. In: Proceedings of. the-Symposium on Flue
Gas Desulfurization-Hollywood, Fl., November 1977 (Volume
1). EPA-600/7-78-058a. March 1978. pp. 338-840.
24. Op. cit. No. 4. p. 34
25. PEDCo in-house files.
26. Ibid.
27. Ibid.
28. Laseke, B.A., Jr. EPA Utility FGD Survey: December 1977 -
January 1978. EPA-600/7-78-051a. PEDCo Environmental,
Inc., Cincinnati, Ohio. Ma ah 1978. p. 44.
29. Op. cit. No. 4. pp. 9, 35.
30. Ibid.
31. McDaniel, C.F. LaCygne Station Unit No. 1 West Scrubber
Operating Experience. Presented at Utility Scrubber Con-
ference, Denver, Colorado. March 29-30, 1978.
32. Op. cit. No. 4. p. 38.
33. Op. cit No. 31. p. 6.
34. PEDCo in-house files.
B-84
-------
35. Power Engineering. New Generating Plants. Harrington,
Illinois. May 1978. p. 8.
36. Laseke, B.A., Jr. Survey of Flue Gas Desulfurization
System; Lawrence Energy Center, Kansas Power and Light
Company. Preliminary report prepared for the U.S.
Environmental Protection Agency under Contract No. 68-02-
2603. PEDCo Environmental, Inc., Cincinnati, Ohio.
January 1979. pp. vii-ix, 2, 4, 6, 7, 8, 14, 17, 19, 21-23,
25, 26, 28, 29, 31, 35, 42, 51-54, 56, 57, 63-70.
37. Op. cit. No. 4. pp. 10, 38.
38. Ibid.
39. PEDCo in-house files.
40. Ibid.
41. Op. cit. No.-36.
42. Op. cit. No. 28. p. 63.
43. Op. cit. No. 4. p. 40.
44. Ibid. p. 41.
45. PEDCo in-house files.
46. Laseke, B.A., Jr. Survey of Flue Gas Desulfurization
Systems: Sherburne County Generating Plant, Northern
States Power Company. Preliminary report prepared for
the U.S. Environmental Protection Agency under Contract
No. 68-02-2603. PEDCo Environmental, Inc., Cincinnati,
Ohio. December 1978. pp. 2, 5, 7, 8, 13-15, 17-20, 22, 24,
26, 29, 33, 36, 38, 43-49, 52, 54, 56-60.
47. Ibid.
48. PEDCo in-house files.
49. Op. cit. No. 46.
50. Op. cit. No. 4. pp. 59-62.
51. PEDCo in-house files.
52. Op. cit. No. 4. pp. 14, 15.
53. Op. cit. No. 28. pp. 128, 129.
54. PEDCo in-house files.
55. Op. cit. No. 28. . pp. 128, 129.
B-85
-------
56. Op. cit. No. 4. p. 70.
57. Op. cit. No. 4. p. 17.
58. PEDCo in-house files.
59. Ibid.
60. Ibid.
61. Ibid.
62. Op. cit. No. 4. pp. 18, 71.
63. PEDCo in-house files.
64. Op. cit. No. 64. pp. 18, 71.
65. PEDCo in-house files.
66. Ibid.
67. Ibid.
68. Op. cit. No. 28. pp. 130-132.
69. PEDCo in-house files.
70. Op. cit. No. 28. pp. 130-132.
71. Op. cit. No. 4. pp. 19, 72, 73.
72. Universal Oil Products, Air Correction Division. Stack
Emission Compliance Test Report for the City Utilities
of Springfield, Missouri, Southwest Power Station. Des
Plaines, Illinois. September 1977.
73. PEDCo in-house files.
74. Ibid.
75. Ibid.
76. Op. cit. No. 28. pp. 142, 143.
77. Ibid.
78. PEDCo in-house files.
79. Ibid.
80. Op. cit. No. 28. pp. 142, 143.
B-86
-------
81. Op. cit. No. 4. pp. 77, 78.
82. Wells, W.L., W.B. Muirhead, and J.H. Buckner. TVA's
Experience with Limestone Scrubbers at the 550-MW Widows
Creek Unit 8. Presented at American Power Conference,
Chicago, Illinois. April 24-26, 1978.
83. PEDCo in-house files.
84. Op. cit. No. 41. pp. 77, 78.
85. PEDCo in-house files.
86. Ibid.
87. Ibid.
88. Op. cit. No. 4. pp. 20, 21, 79, 80.
89. Ballard, B., and M. Richman FGD Systems Operation at Martin
Lake Steam Electric Station. Presented at the Joint Power
Generation Conference, Dallas, Texas. September 10-13,
1978.
90. Ibid.
91. PEDCo in-house files.
92. Ibid.
93. Op. cit. No. 4. pp. 20, 21, 79, 80.
94. PEDCo in-house files.
95. Ibid.
96. Ibid.
97. Ibid.
98. Op. cit. No. 4. pp. 21, 81.
B-87
-------
APPENDIX C. SUMMARY OF CHANGES: AUGUST 1975 TO OCTOBER 1978
' Period
Company
Station, unit
July 31 . 1977
Alabama Electric
Tonblgbeo 2
Tonblgbee 3
Arizona Electric
Apache 2
Arizona Public Service
Cholla 2
Choi la 4
Four Corners 1
Four Corners 2
Four Corners 3
Four Corners 4
Four Corners S
Associated Electric
Thomas Hill 3
Basin Electric Power
Antelope Valley 1
Antelope Valley 2
La ramie River 1
Laramle River 2
Laranle River 3
Big River Rural Electric
Green 2
Operational
No.
29
+1
+1
+1
Board of Municipal Utilities
Slkeston 1
m
'8,914
225
200
250
Under
construction
No.
28
-1
+1
-1
-1
+1
+1
+1
+1
+1
HW
11,810
225
225
200
250
350
570
570
240
235
Contract
awarded
No.
23
-1
+1
+1
+1
+1
+1
-1
-1
-1
-1
NH
11,810
225
755
755
670
455
455
570
570
240
235
Letter of
Intent
No.
5
MH
1.892
Requesting/
evaluating bids
No.
5
+1
m ^
2,8?5
550
Considering
No.
35
-1
+1
+1
+1
-1
-1
-1
-1
HH
16.031
350
175
175
229
755
755
455
550
Terminated9
No.
16
NH
1,488
Total
No.
125
+1
+1
+1
+1
+1
+1
NH
53,352
175
175
229
670
455
240
-------
APPENDIX C (continued)
Period
Company
Station, unit
Brazos Electric Power
San Miguel 1
Central Illinois Light
Duck Creek 1
Duck Creek 2
Cincinnati Gas & Electric
East Bend 2
Colorado Ute
Craig 1
Craig 2
Columbus ft So. Ohio Elec.
Conesville 6
Commonweal th Edison
Will County 1
Delmarva Power
Delaware City 1.2.S 3
Eastern Kentucky Power
Spur lock 2
General Public Utilities
Coho 1
Seward 7
Operational
No.
+1
+1
-1
MM
400
400
167
Under
construction
No.
+1
-1
+ 1
+1
-1
+1
HW
400
400
450
450
400
180
Contract
awarded
No.
-1
+1
-1
-1
-1
+1
MU
4no
600
450
450
no
son
Letter of
Intent
No.
MU
Requesting/
evaluating bids
No.
+1
MW
400
Considering
No.
-1
-1
-1
+1
+1
MW
400
600
500
800
800
Terminated
No.
+1
MW
167
Total
No.
+1
+1
MW
800
800
o
I
10
(continued)
-------
APPENDIX C (continued)
Period
Company
Station, unit
Gulf Power
Sholz IB ft 2B
Hoosler Energy
Nero* 1
Heron 2
Indianapolis Power
and Light
Petersburg 3
Petersburg 4
Kansas Power ft Light
Jeffrey 1
Lakeland Utility
Nclntosh 3
LoulsvlUe Gas ft Electric
Cane Run 5
Mill Creek 1
Mill Creek 2
Mill Creek 3
Hill Creek 4
Minnesota Power ft Light
Clay Boswell 4
Minnesota Power
M.R. Young 2
Operational
No.
+1
+1
+1
+1
+1
+1
MM
23
530
680
183
425
450
Under
construction
No.
+1
+1
-1
+1
-1
-1
+1
-1
+1
+1
-1
MW
490
490
530
530
680
183
330
425
495
500
450
Contract
awarded
No.
+1
+1
-1
MU
350
330
495
Letter of
Intent
No.
-1
MM
500
Requesting/
evaluating bids
No.
-1
-1
MM
490
490
Considering
No.
-1
-1
-1
MW
530
330
330
Terminated*
No.
MU
Total
No.
+1
+1
HU
23
350
o
OJ
(continued)
-------
APPENDIX C (continued)
O
Period
Conpany
Station, unit
Nevada Power
Reid Gardner 4
New England Electric
Bray ton Point 3
Otter Tall Power
Coyote 1
Pacific Gas ft Electric
Fossil 1
Fossil 2
Pacific Power ft Light
J1m Brldger 4
Pennsylvania Power
Bruce Nawnsfleld 3
Potomac Electric & Power
Dicker son 1
Dlckerson 2
Public Service of Indiana
Gibson 3
Gibson 4
Gibson 5
Public Service of New
Mexico
San Juan 1
San Juan 2
San Juan 3
San Juan 4
Operational
No.
+1
«•!
NU
314
306
Under
construction
No.
+1
+1
+1
-1
-1
MW
400
509
825
375
306
Contract
awarded
No.
-1
-1
+1
+1
MW
509
825
468
472
Lette
Inte
No.
-1
-1
r of
nt
MM
250
468
Reqy
evalua
No.
-1
-1
+1
estlng/
ting bids
HW
650
650
650
Const
No.
+1
-1
-1
+1
+1
-1
-1
-1
derlnq
MW
250
650
400
800
800
190
190
472
Terminated*
No.
MW
To
No.
»1
-1
+ 1
+ 1
-1
-1
-1
-1
+ )
tal
HW
250
650
800
800
190
190
650
650
650
(continued)
-------
APPENDIX C (continued)
Period
Company
Station, unit
Salt River Project
Coronado 1
Coronado 2
Semlnole Electric
Semlnole 1
South Cardlne Public
Service
Hlnyah 3
Southern Illinois Power
Marlon 5
Southern Indiana Gas &
Electric
A.B. Brown 1
Southern Mississippi
Electric
R.D. Morrow 1
Southerwestern Electric
Power
Perkey 1
Springfield Hater, Light
and Power
Dull man 3
Tampa Electric
Big Bend 4
Operational
No.
+1
MW
180
Under
construction
No.
+1
+1
+1
-1
MM
350
350
250
180
Contract
awarded
No.
-1
-1
+1
-1
+1
+1
MM
350
350
300
250
720
190
Letter of
Intent
No.
MM
Requesting/
evaluating bids
No.
-1
MU
300
Considering
No.
+1
+1
-1
+1
MU
600
300
190
450
Terminated3
No.
MW
To
"No7~
+1
+1
+1
+1
+1
tal
MM
600
300
720
190
450
o
U1
(continued)
-------
APPENDIX C (continued)
Period
Company
Station, unit
Tennessee Valley Authority
Johnsonville 1
Paradise 1
Paradise 2
Widows Creek 7
Texas Power & Light
Sandow 5
Texas Utilities
Forest Grove 1
Martin Lake 1
Martin Lake 2
Martin Lake 3
Montlcello 3
United Power Association
Coal Creek 1
Coal Creek 2
Utah Power & Light
Emery 2
Hunting ton 1
Wisconsin Power & Light
Columbia 2
October 31, 1978
Dperational
No.
+1
+1
+1
+1
46
MW
793
793
750
415
16,054
Under
construction
No.
+1
+1
-1
-1
+ 1
-1
+1
+1
-1
39
MW
600
575
793
703
793
750
545
545
415
16,723
Contract
awarded
No.
+1
-1
-1
-1
+1
*1
23
MW
545
793
545
545
400
527
12,450
Letter of
Intent
No.
MW '
Req
evali
"No. "
+ 1
! +1
,
-1
1
I
-1
1
527
240
uesting/
atinq bids
MW
650
650
545
+ 1 750
Considering
No.
-1
-1
MW
575
750
i
6
1
3.C50
28
12,610
Terminated3
No.
MW
!
1
17
1,555
Total
No.
+ 1
+ 1
tl
MW
600
650
650
+ 1
143
400
61,732
o
I
a\
a Terminated category includes prototype or demonstration systems that onerated for a limited period of time. A number of these systems
are still in service, removing primarily particulates. The units in this category are not included in the total category.
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-450/3-79-001
2.
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Electric Utility Steam Generating Units
Flue Gas Desulfurization Capabilities As of
October 1978
6. REPORT DATE
January 1979
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
B.A. Laseke, Jr., M.T. Melia, M.T. Smith & T.J. Koger
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
PEDCo Environmental, Inc.
11499 Chester Road,
Cincinnati, Ohio 45246
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-02-2811
12. SPONSORING AGENCY NAME AND ADDRESS
Deputy Assistant Administrator for Air Quality Planning
and Standards, Office of Air, Noise and Radiation
U.S. Environmental Protection Agency
Research Triangle Park. N.C. 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
EPA/200/04
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This study updates the previously published final report, "Flue Gas
Desulfurization System Capabilities for Coal-Fired Steam Generators," EPA-600/7-78-032b
published in March 1978. This assessment was made by reviewing the changes and
developments in the technology since the preparation of the March 1978 report. A
substantial increase in the number and capacity of operational FGD systems, plus the
additional operational experience obtained by previously identified operational systems
have resulted in a substantial increase in the amount of design and performance
information. Most notably, these include dependability (availability operability,
reliability, and utilization) data, removal efficiency data (sulfur dioxide and
particulate), operating problem and solution data, results from various research,
development, and demonstration programs, and process and design innovations for new
systems. Virtually all of the FGD operating experience gained to date has been with
the wet-phase, nonregenerable, lime/limestone processes. As a direct result of this
previous experience, the systems committed for operation within the next 3 to 5 years
also show an overwhelming preference for lime/limestone processes. Analysis of the
current status of the technology indicates that the design and operating experience
gained with the first and second generation FGD systems has resulted in improved
design and operation of subsequent installations. Because FGD systems that are being
engineered and/or erected will incorporate many or all of these design innovations,
17.
KEY WORDS AND DOCUMENT ANALYSIS
a.
DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED TERMS C. COS AT I Field/Group
Electric Power Generation
Reliability
Control Equipment
Sulfur Dioxide
13 B
18. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
20. SECURITY CLASS (Thispage)
Unclassified
451
22. PRICE
EPA Form 2220-1 (R«v. 4-77) PREVIOUS EDITION is OBSOLETE
------- |