Conceptual Design
and Cost Study
Sulfur Oxide Removal From
Power Plant Stack Gas
Use of Limestone in
Wet-Scrubbing Process
1969
-------
Sulfur Oxide Removal From
Power Plant Stack Gas
. Use of Limestone in
Wet-Scrubbing Process
Conceptual Design and Cost Study Series
Study No.2
Prepared for
National Air Pollution Control Administration
(0. S. Department of Health, Education, and Welfare)
By
Tennessee Valley Authority
Contract No. TV-29233A
-------
-- ---
. 1.1 I
STACK
t
aO\\.E.~
-"'\
POWER PLANT EQUIPPED WITH LIMESTONE-WET
SCRUBBING PROCESS FOR SULFUR OXIDE REMOVAL
SCRUBBING EQUIPMENT SHOWN IN COLOR
-------
Preface
On February 16, 1967, the National Center for Air
Pollution Control (now National Air Pollution Control
Administration), U. S. Department of Health, Education,
and Welfare, entered into a contract with the Tennessee
Valley Authority (TV A) for a series of conceptual design
and economic studies to be carried out by TV A on
processes for reducing sulfur oxide emissions from power
generation. The purpose is to evaluate objectively and
realistically the merits of different methods under
consideration for sulfur oxide control, with a common and
uniform basis used for comparison. The studies include (1)
literature surveys, (2) direct contacts with workers in the
field to supplement literature information, (3) specific
studies by specialists in power plant design, chemical plant
design, power plant operation, and pollutant dispersion to
supplement the conceptual design study, (4) market surveys
where appropriate, (5) quotations from vendors and
fabricators on major equipment items, and finally (6)
investment and operating cost estimates.
Work has proceeded on three processes: (1) limestone
injection (dry process), (2) use of limestone in a
wet-scrubbing process, and (3) ammonia scrubbing. The
limestone injection study has been completed and a report
issued: "Sulfur Oxide Removal from Power Plant Stack
Gas-Sorption by limestone or lime (Dry Process)"
(1968). The present report covers use of limestone in
wet-scrubbing Dfocesses.
The work has been divided in TV A as follows:
Project Supervision
Applied Research Branch (Division of Chemical
Development)
Cost Esti mates
Applied Research Branch
Design Branch (Division of Chemical Development)
Design
Mechanical Design Branch (Division of Engineering
Design
Design Branch
Report Preparation
Applied Research Branch
Research Staff (Office of Power)
Steam-Electric Generation Branch (Division of
Power Production)
Air Quality Branch (Division of Health and Safety)
Water Quality Branch (Division of Health and Safety)
A major part of the evaluation has been the analysis of
findings by other organizations who have worked on wet
limestone scrubbing. These findings, which are, of course,
the basis for the conceptual design, are used throughout the
body of the report as needed. In addition, to serve as a
general reference, the work of most of the organizations
quoted is summarized and presented in appendixes. In
addition to the published literature, several organizations
have supplied information directly that has been quite
helpful in the study; the contributions of the following are
acknowledged.
Central Electricity Generating Board (England)
Combustion Engineering, Inc. - Detroit Edison Company
Imperial Chemical Industries Limited (England)
National Air Pollution Control Administration (NAPCA)
Union Electric Company
Universal Oil Products Company, Air Correction Division
Wisconsin Electric Power, Inc.
Copies of the reports in the NAPCA-TV A conceptual
design series can be obtained by ordering from the
following address.
Clearinghouse for Scientific
and Technical Information
5285 Port Royal Road
Springfield, Virginia 22151
The reports are identified and priced as follows;
Title Number
Sulfur Oxide Removal from Power PB 178-972
Plant Stack Gas-Sorption by
limestone or lime (Dry Process)
Sulfur Oxide Removal from Power
Plant Stack Gas-Use of limestone
in Wet-Scrubbing Process
*Not yet assigned.
Price
$3.00
*
*
3
-------
SUMMARY. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Process problems. . . . . . . . . . . . . . . . . . . . . . . . . .
Study assumptions. . . . . . . . . . . . . . . . . . . . . . . .
Process equipment. . . . . . . . . . . . . . . . . . . . . . . . .
Costs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Conclusions and recommendations. . . . . . . . . . . .
INTRODUCTION. . . . . . . . . . . . . . . . . . . . . . . . . . .
HISTORY AND STATUS.....................
PROCESS CHEMISTRY AND KINETICS.........
PROCESS CONSIDERATIONS AND PROBLEMS. .
Point oflimestone addition. . . . . . . . . . . . . . . . . .
Dust removal. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Scaling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Corrosion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Limestone effectiveness. . . . . . . . . . . . . . . . . . . .
Solids disposal and water pollution. . . . . . . . " . .
Stack gas reheat.. ................".......
Intermittent operation. . . . . . . . " . . . . . . . . . . . .
STUDY ASSUMPTIONS AND DESIGN CRITERIA .
Plant location. . . . . . . . . . . . . . . . . . . . . . . . . . . .
Plant size. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Fuel type. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Sulfur content of coal. . . . . . . . . . . . . . . . . . . . . .
Limestone type. . . . . . . . . . . . . . . . . . . . . . . . . . .
Amount of limestone used. . . . . . . . . . . . . . . . . .
Particle size. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Boiler type. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Point of limestone addition. . . . . . . . . . . . . . . . . .
Dust collection. . . . . . . . . . . . . . . . . . . . . . . . . . .
Solids disposal. . . . . . . . . . . . . . . . . . . . . . . . . . . .
Stack gas reheating. . . . . . . . . . . . . . . . . . . . . . . "
Stack height. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
EQUIPMENT SELECTION AND DESCRIPTION..
Major alternatives. . . . . . . . . . . . . . . . . . . . . . . .
Equipment description. . . . . . . . . . . . . . . . " . . " .
Page
CONTENTS
7
7
8
8
8
9
ECONOMIC EVALUATION. . . . . . . . . . . . . . . . . . .
Basis of evaluation. . . . . . . . . . . . . . . . . . . . . . . . .
Process alternatives. . . . . . . . . . . . . . . . . . . . . . . .
Sorbent vendor and shipping costs. . . . . . . . . . . .
Overall process investment evaluation. . . . . . . . . .
Overall operating cost evaluation. . . . . . . . . . . . . .
10
RESEARCH AND DEVELOPMENT NEEDED. . . . .
Absorbent efficiency. . . . . . . . . . . . . . . . . . . . . . .
Scaling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Corrosion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Oxidation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Waste disposal. . . . . . . . . . . . . . . . . . . . . . . . . . . .
Intermittent operation. . . . . . . . . . . . . . . . . . . . .
Nitrogen oxide removal. . . . . . . . . . . . . . . . . . . . .
Partial removal of sulfur oxides. . . . . . . . . . . . . . .
11
13
16
16
18
19
20
20
21
22
26
CONCLUSIONS AND RECOMMENDATIONS. . . . .
REFERENCES. . . . . . . . . . . . . . . . ". . . . . . .
APPENDIX A:
Studies of Limestone-Wet Scrubbing. . . . . . . . . . . .. 67
London Power Company (London, England) "" 67
James Howden and Company Imperial
Chemical Industries (London, England. . . . . . .. 67
Tennessee Valley Authority
(Muscle Shoals, Alabama) '""""""""" 71
Scientific Research Institute (Soviet Union) ..... 72
Wisconsin Electric Power Company
and Universal Oil Products, Inc.,
Air Correction Division. . " . . . . . . . . . . . . . . . .. 72
Combustion Engineering, Inc.,
and Detroit Edison Company. . . . . . . . . . . . . .. 75
29
29
29
29
29
29
30
30
31
31
31
31
33
33
APPENDIX B: Water Quality Considerations.. .".
APPENDIX C: Cost Estimates. . " . . . . . . . . . . . . . . .
Detailed estimates. . . . . . . . . " . . . . . . . . . . . . . .
Determination of credit for corrosion reduction. .
Determination of operating and investment credits
for eliminating electrostatic precipitator. . . . . . .
34
34
40
APPENDIX D: Drawings.....................
Page
43
43
44
49
52
53
59
59
59
59
60
60
60
60
60
62
65
78
83
83
96
96
97
-------
A-I
1
2
3
4
Capital requirements. . . . . . . . . . . . . . . . . .
Operating cost. . . .. ............. ...
Plant operating data: Colbert steam plant
Maximum groundline concentrations of sulfur
dioxide at 18-mph wind speed. . . . . . . . .
Sulfur content of coal from various sources.
Typical analysis and physical properties
of Colbert County limestone. . . . . . . . . . .
Comparison of scrubber combinations. . . . .
Scrubber comparison. . . . . . . . . . . . . . . . . .
Comparison of demisters ...............
Investment and operating cost for
one- and two-stage TCA scrubber
systems in existing 200-mw unit. . . . . . . . .
Effect of absorbent type on operating
cost for Process B . . . . . . . . . . . . . . . . . . . .
Dust removal efficiency, investment, and
operating cost of precipitator-wet
scrubber combination in existing 200-mw
power unit. . . . . . . . . . . . . . . . . . . . . . . .
Cost of gas reheating methodsjyr . . . . . . . .
Quoted prices of raw limestone. . . . . . . . . .
Overall investment for scrubbing
facility in existing plant. . . . . . . . . . . . . . .
Overall operating cost ( existing units) .....
5
6
7
8
9
10
11
12
13
14
15
16
Design and operating data for the Howden-ICI
pilot plant. . . . . . . . . . . . . . . . . . . . . . . . .
TABLES
Page
8
9
23
A-2
24
31
C-l
C-2
C-3
31
37
38
39
45
C-4
C-5
C-6
47
47
50
51
C-7
C-8
C-9
C-lO
C-11
C-12
52
55
69
C-13
C-14
C-15
C-16
Operating and design data from Fulham
sulfur dioxide removal unit. . . . . . . . . . . .
Summary of estimated fixed investment
requirements: Process A- Limestone
injection-scrub bing
200-mw existing power unit. . . . . . . . . . . .
1,000-mw existing power unit. . . . . . . . . .
1 ,OOO-mw new power unit. . . . . . . . . . . . .
Summary of estimated fixed investment
requirements: Process B- Limestone
addition-scrubbing
200-mw existing power unit. . . . . . . . . . .
1,000-mw existing power unit. . . . . . . . .
1,000-mw new power unit. . . . . . . . . . . . .
Annual operating costs for limestone-
wet scrubbing power plant stack gas:
Process A- Limestone injection-scrubbing
200-mw existing unit, 2.0% S in coal
200-mw existing unit, 3.5% S in coal. . . . .
200-mw existing unit, 5.0% S in coal. . . . .
500-mw existing unit, 3.5% S in coal. . . .
1,000-mw existing unit, 3.5% S in coal. . . .
1 ,OOO-mw new unit, 3.5% S in coal. . . . . .
Annual operating costs for limestone-
wet scrubbing power plant stack gas:
Process B- Limestone addition-scrubbing
200-mw existing unit, 3.5% S in coal. . . . .
500-mw existing unit, 3.5% S in coal. . . . .
1,000-mw existing unit, 3.5% S in coal. . . .
1,000-mw new unit, 3.5% S in coal. . . . . .
Page
71
83
83
84
84
85
85
86
87
88
89
90
91
92
93
94
95
-------
2
Effect of pH of calcium sulfite-bisulfite
solution on S02 equilibrium
vapor pressure. . . . . . . . . . . . . . . . . . . . . .
Desupersaturation of calcium sulfate solution
at 50° C. by suspended CaS04 2H20 . . . . .
Effect of reheat temperature on
ground-level S02 concentration at
various scrubber efficiencies. . . . . . . . . . .
Cooling of plume by evaporation of
water condensed before emission. . . . . . . .
Distribution of thermal power plants in
the United States (actual and projected) ..
200-mw unit at TV A Colbert steam plant. . .
Typical turbulent contact absorber scrubber
Sulfur oxide absorption efficiency in
turbulent contact absorber. . . . . . . . . . .
Dust collection efficiency in turbulent
contact absorber. . . . . . . . . . . . . . . . . . . . .
Collection efficiency for I-micron
particles (impingement-type scrubber). ..
Cost of power plant stacks. . . . . . . . . .. ..
Effect of pressure drop on operating
cost of stack gas exhaust fan. . . . . . .. ..
Investment for various methods of
solids disposal. . . . . . . . . . . . '.' . . . . . . . . .
Operating cost for various methods of
solids disposal. . .. ..................
Effect of reheat temperature on
investment for stack gas reheating.. ....
Effect of reheat temperature on
operating cost of stack gas reheating. .. .
Effect of intermittent operation on cost
of reheating stack gas by combustion of
natural gas. . . . . . .. ...............
Bulk transportation cost for shipping
limestone by truck or rail. . . . . . . . . . . . .
Effect of power unit size on
limestone-wet scrubbing investment
Effect of sulfur content of coal
on limestone-wet scrubbing investment.
Effect of power unit size on
limestone-wet scrubbing operating cost
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Page
FIGURES
15
19
25
27
30
32
34
35
A-I
A-2
A-3
35
37
44
A-4
A-5
45
A-6
46
47
A-7
48
49
D-l
D-2
D-3
D-4
50
D-5
52
D-6
53
D-7
54
D-8
54
Page
22
Effect of sulfur content of coal on
limestone-wet scrubbing operating cost. .
Effect of limestone cost on
limestone-wet scrubbing operating cost. .
Effect of scrubber system operating time
on limestone-wet scrubbing operating
cost. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Effect of boiler load factor on
limestone-wet scrubbing operating cost. .
Effect of boiler load factor on
wet scrubbing vs dry process operating
cost. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
64
55
23
56
24
57
25
58
26
Howden-ICI pilot plant. . . . . . . . . . . . . . .. 68
Howden-ICI unit at Fulham power station.. 70
Effect of suspended solids in lime slurry
on mass transfer coefficient 'for S02
absorption. . . . . . . . . . . . . . . . . . . . . . . .. 73
Scrubber arrangement in Wisconsin
power tests. . . . . . . . . . . . . . . . . . . . . . . .
Percent S02 removal vs pH of
scrubber effluent. . . . . . . . . . . . . . . . . . .
Sulfur oxide removal unit tested by
Combustion Engineering and
Detroit Edison. . . . . . . . . . . . . . . . . . . . .. 76
Lime scrubbing system to be installed
on boiler unit 5 of Kansas Power and
Light Company. . . .. ...............
Flowsheet-Process A . . . . . . . . . . . . . . . . . .
Supplement to Figure D-l """""'"
Flowsheet-Process B . . . . . . . . . . . . . . . . . .
Limestone receiving, handling, and dry
injection facilities.. ................. 100
Limestone wet-grinding
facilities-Process B """"""""" 101
General arrangement-scrubber
area elevation view. . . . . . . . . . . . . . . . . .. 102
General arrangement-scrubber
area plan view. . . .. "" ...... .... 103
TCA scrubber for wet limestone
scrubbing process .............. ....104
74
75
77
97
98
99
-------
Summary
The growing problem of atmospheric pollution by sulfur
oxides has promoted a large amount of research and
development on removal from power plant stack gases. Of
the several processes that have been proposed, use of lime
or limestone in a wet scrubber is one of the more
promising. Although the sulfur oxides are recovered as
calcium sulfate or sulfite, which are likely to be useless
products in most instances, relative simplicity and low
investment have promoted interest in the method.
Moreover, the net cost penalty in producing power can be
predicted with fair accuracy, whereas the net cost in
recovery processes is uncertain because the market for the
products is subject to variation.
limestone-wet scrubbing can be used in two ways,
either by injecting limestone into the power plant boiler
and catching it in a wet scrubber after the air heater or by
introducing limestone ( or lime) directly into the scrubber
system. Injection of limestone into the boiler removes a
portion of the sulfur dioxide ahead of the scrubber,
provides protection from corrosion by sulfur trioxide and
alkali salts, and converts the limestone to quicklime, a more
reactive form for use in the scrubber. Introduction oflime
or limestone directly into the scrubber system eliminates
potential boiler operating problems such as abnormal
slagging and increased erosion and makes possible certain
procedures that prevent equipment scaling. Direct
introduction into the scrubber system was developed to a
commercial scale in England over 30 years ago; modern
emphasis is on the injection-scrubbing process.
The present study is one of a series of conceptual design
and cost studies being carried out by the Tennessee Valley
Authority (TV A) for the National Air Pollution Control
Administration (NAPCA). The purpose is to develop, for
various promising processes, the best design possible from
existing data, estimate capital and operating costs on a
uniform basis, and recommend further research and
development needed. This is the second study in the series;
the first was an evaluation of the dry limestone
method-injection of limestone into the boiler without
subsequent wet scrubbing. One of the objectives in the
present study is to compare the merits of the wet and dry
processes.
Other major work is being done in the area of limestone
injection and wet scrubbing. In a joint program tJetween
TV A and NAPCA, various research projects now under way
will culminate in large-scale field tests of the dry limestone
process. The results of this study will be useful in refining
evaluation of combined injection and wet scrubbing. Also,
as part of the NAPCA program, Bechtel Corporation is
preparing a preliminary design for a large pilot-scale study
of limestone injection-wet scrubbing in scrubbers of
various types.
In commercial application of the process, Combustion
Engineering, Inc., is testing wet-scrubbing systems of the
injection-scrubbing type at two power stations.
Process Problems
Results of experimental work demonstrate that a high
degree (90% or more) of sulfur dioxide removal from power
plant stack gases can be accomplished by scrubbing with a
slurry of lime or limestone. Pilot plant results indicate that
good utilization of absorbent can be expected when
limestone is calcined by injection into the boiler or when
lime is used directly. However, both full-scale and pilot
plant tests show that absorbent requirement increases if
uncalcined limestone (added in the scrubber system) is used
as . the absorbent. P~rn~!eIs_which - - affect degre.e- of
absorbent utilization-limestone type, particle size, scrubber
operating conditions, and dead-bur2illg (when limestone-is
injected in the boiler)-need further study.
Formation of sulfite and sulfate scale was reported to be
a major problem in early commercial use of wet scrubbing.
In the early work in England, a satisfactory method for
preventing scaling was developed for the process in which
lime or limestone is introduced directly into the scrubber
system; the main feature was introduction of the absorbent
into a delay tank after the scrubber. Similar scaling
problems have been encountered in current plant tests in
the country in which the limestone is injected into the
boiler; prevention of scaling in such a system remains to be
demonstrated.
Scrubbing with a water slurry cools the gases to the
saturation temperature, about 1250 F. Discharge of a cool
saturated plume is not consistent with current emphasis on
maximum plume rise for good dispersion. Also, there is
some opposition to a visible vapor trail from a power plant
stack. A study of the effect of reheat temperature on
dispersion showed that with high scrubbing efficiency,
reheat only to maintain temperatures above dew point to
the stack top should prevent excessive ground-level
concentration of sulfur dioxide. However, a higher degree
of reheat may be required to prevent increase in ground
concentration of other pollutants such as nitrogen oxides.
D~jJosalof large- quanjities of solid waste, 160,000toIM
per year for a 200-megawatt plant, jVill require additional
land area for settling ponds or for disposal in the solid
7
-------
form. Since the cost of recycling the sluice water is not
high, recycling rather than release to streams is indicated.
However, if high-calcium limestone is used as the absorberu.
the product solids should not be soluble enough in the
scrubber effluent to cause water quality proble~in
receiving streams of the size normally contiguous to power
plants. I[dolomiticJiInestone is used, the high solubility 2f
magnesium~sa1ts would in many cases make sluice wat~
recycling mand~tory.
Study Assumptions
The cost of limestone-wet scrubbing and, to some
extent, the type of equipment required, depend on several
factors such as sulfur content of the coal, limestone cost,
degree of grinding, power plant size and location, and
excess limestone (above the theoretical amount) used to
increase degree of sulfur oxide removal. It was necessary to
assume a set of conditions as a base case for the conceptual
design; in addition, the effect of varying each parameter
was analyzed and factors developed wherever possible to
relate the basic design and estimate to other conditions.
The principal basic conditions assumed are given in the
following tabulation:
Conditions Assumed for Study
200 megawatts
Pulverized coal
Boiler size
Fuel type
Sulfur content
of coal
Ash content
of coal
Plant location
Limestone type
3.5%
12%
Northwestern Alabama
High-calcium limestone
(94.9% CaC03 + MgCO 3)
$2.05 per ton (delivered)
70% -200 mesh
Limestone cost
Particle size
Amount of
absorbent
110% of theoretical when
injected into the boiler;
130% and 100% of theoretical,
respectively, when limestone
or lime is introduced into
the scrubber system
2500 F.
Reheat temperature
Cost comparisons made in the study include: existing
versus new plants, operation with and without electrostatic
precipitators in existing plants, alternate scrubber designs,
alternate waste disposal methods, and alternate reheat
methods and temperatures.
8
Process Equipment
The equipment needed for limestone injection followed
by wet scrubbing includes many of the items required in
the dry process. The main needs are a limestone-receiving
hopper, storage facilities, a dry grinding installation, an
injection system, a scrubber with mist eliminator, a gas
reheat system, and solid waste disposal facilities. For
limestone added directly in the scrubber system, a
wet-grinding facility was provided.
For the absorption step, a dual-train, two-stage,
mobile-bed scrubbing system was selected. The scrubber is
designed to handle the total dust load. Gas reheat is
effected by heat transfer from the gas ahead of the scrubber
to the scrubber exit gas by circulating water through
finned-tube exchangers.
Modification of the boiler induced-draft fans was
necessary to compensate for the increased pressure drop
(approximately 12 in. of water) in the system. Further
investment is required for additional solids-disposal
equipment and pond area to accommodate the increased
waste load.
Costs
Capital and operating cost requirements have been
calculated for such factors as unit size, sulfur content of
coal, type of process, raw material cost, and operating time
under the assumptions outlined above. The results are
summarized in Tables I and 2.
Capital costs include all equipment from limestone
!~~~1~~~~~~~~k~~~~~_------------------
Capital, $/kw
~...P.!!~!~_U:~I1~~!!.'L
13.05
Conditions
-------
Base casea
Exceptions to base case
2.0% S
5.0% S
Limited reheatb
To 2000 F. 10.52
To 1750 F. 9.47
Process BC 13.80
Process B (with lime)d 20.00
500 mw 10.85
1,000 mw 8.21
1,000 mw, process B 8.82
1~~~~~~~~~Q~~~~~~-----------_____~1~___-
11.70
14.30
aBase case assumes 200-mw unit, existing power plant, 3.5% sulfur
in coal, process A (injection-scrubbing), reheat to 2500 F. by heat
exchange, 99.5% dust removal, 95% 802 removal, and nonrecycle
of sluice water.
bReheat by direct fIring natural gas.
c Addition of limestone to the scrubber circuit; 85% 802 removal.
d Addition of lime (CaO) to the scrubber circuit.
elncludes credit for eliminating electrostatic precipitator.
-------
receiving to waste-solids disposal and stack gas reheat, plus
installation, engineering and contractor fees, and provision
for contingency. Since new units would not require the
usual electrostatic precipitator, an investment credit
equivalent to precipitator cost was taken. While there is
some possibility that a shorter stack might be adequate for
a new plant because most of the sulfur oxide is removed,
other considerations such as particulate and nitrogen oxide
emission as well as possible process upsets make this
questionable; therefore no credit was taken for reduction in
stack height.
Operating costs include raw material, labor,
maintenance, and overhead costs; appropriate capital
charges (including return on investment); assignable credits
!~~~~~9~~~~~~~~!~-----------------------
$/ton
of coal
1.31
_~!!.sL~wJl
0.49
Conditions
Base casea -------
Exceptions to base case
2.0% S
5.0% S
$1.0O!ton limestone
$4.00/ton limestone
Intermittent operationb
720 hr/yr (30 da)
4,000 hr/yr (Yo yr)
Reduced boiler operationc
(50% load factor)
Limited reheat
To 2000 F.
To 1750 F.
I ntermittent reheat to
2000 F.d (30 da/yr)
Sluice water recycled
Pro cess Be
Process B (with lime)f
500 mw
1,000 mw
1,000 mw, process B
1,000 mw, new power unitg
1,000 mw, new power unit.
_~~~n~~~~~~B~_____------~~~------_J~~---
aBase case assumes 200-mw unit, existing power plant, 3.5% sulfur
in coal, 8,000 tu/yr operation, process A (injection-scrubbing),
nonrecycle of sluice water, reheat to 2500 F. by heat exchange,
$2.05/ton limestone cost, 14.5% capital charge, 99.5% dust
removal, and 95% S02 removal.
bBoiler operating at full load; scrubber operating for period shown;
cost is average based on all coal burned and power produced/yr.
cBoiler operated intermittently or at reduced load; scrubber
operated when boiler is in operation.
dDirect heating with natural gas.
e Addition of limestone to the scrubber circuit; 85% S02 removal.
f Addition of lime (CaO) to the scrubber circuit.
gIncludes credit for eliminating electrostatic precipitator.
1.05
1.57
1.17
1.56
0.39
0.59
0.44
0.59
0.75
1.18
0.28
0.44
2.10
0.79
1.13
1.06
0.42
0.40
1.07
1.33
1.42
1.85
1.11
0.90
0.99
0.76
0.40
0.50
0.53
0.69
0.41
0.34
0.37
0.29
for eliminating precipitator usage (existing precipitators will
not be operated) and corrosion reduction; and charges for
thermal losses when limestone is injected into the boiler.
The cost of limestone is likely to vary widely, depending
on amount used, power plant location, and the shipping
mode and distance. In a survey of the limestone industry
presented in this study, it was indicated that the price range
will be roughly $1.00 to $3.00 per ton (delivered, -I/2-in.
size). A very large amount of limestone is required if the
process is operated continuously-on the order of 4S.Q,Q0Q..
tons per year for a 1000-megawatt plg.,fit. Under conditions
of large-scale usage, it may be possible to get special price
and shipping cost considerations.
Conclusions and Recommendations
L~~Qp.$:-wet scrubbing has some important
advantages ove7the dry limestone process, including better
~¥",,J'or removing sulfur oxides over 90% vs.
50-60%), les~ialmie"irerence""with bo ~ uperltion.
and generally lower operating cost for large power plants.
On the other hand, tM dry process releases the stack gas at
higher temperature (thereby giving better dispersion of
residual pollutants), requires less investment, is simpler to
operate, is more amenable to intermittent operation, and is
less likely to cause ~!~.L{)9_n. The wet process appears
to be the best choice for the larger power plants,
particularly those on baseload service. Moreover, if
potential problems can be worked out, the wet process may
have an economic advantage over the dry method in small
plants, for comparable degrees of sulfur oxide removal. The
dry method is more applicable to situations where a high
degree of sulfur oxide removal is not required and where
intermittent operation to reduce cost is feasible or the
boiler is a peak-load unit out of service from time to time.
The relatively low investment for the dry process reduces
the penalty of continuing capital charges when the unit is
out of operation.
Of the two wet-process variations, introduction of lime
or limestone into the scrubber system costs somewhat
more; lime is too expensive to consider and more absorbent
is required when limestone is used. However, the limestone
handicap may be removed by further research. The method
is attractive because of its proven effectiveness in
eliminating scaling.
Research and development should be pushed forward
vigorously on the limestone-wet scrubbing process because
of its promise as a basic method for preventing pollution by
sulfur oxides. '{he main areas neerlin~ invpdig~t;{)n are Q)
the basic cqemistnr. of the process. (2) methods -Lor
J>reve~ Bcaling JD£lJld.in~fuJl~cd~,,~!ol'mel1tQLa.P:d1D,g
1jme~.t£w~~ In.~t4.e_=SS:lllQ2.er ccsxs~em, = and- (3)!~s!~n~.2!:
eq~ipment a.rrangem~DJ~(oril1ternlitten_t operation and ~
PaItiaLsulfur oxide remov~
9
-------
Introduction
This report presents the second of a series of design and
cost studies made by TV A (Tennessee Valley Authority)
for NAPCA (National Air Pollution Control
Administration) on methods for removing sulfur oxides
from power plant stack gases. The fIrst covered sorption by
limestone 1 in a dry method (1); the present one is
concerned with use of limestone in a wet-scrubbing process.
Of the various atmospheric pollutants, sulfur dioxide is
generally regarded as one of the most serious.
Approximately 30 million tons of sulfur dioxide is emitted
to the atmosphere annually in the United States, and power
plants contribute about half. Hence the problem of
reducing sulfur dioxide emission from power plants is
receiving increased emphasis. Several ways of reducing
emission have been studied, including use of low-sulfur fuel,
fuel desulfurization, fuel gasification to give a clean fuel to
the boiler, and sulfur oxide removal from the stack gas. Of
these, stack gas treatment has received major attention
because of its several advantages and because it has the
greatest potential for near-future application.
Stack gas treatment processes that have been studied fall
into two classes-those in which the sulfur is recovered in a
useful and salable form and those of the nonrecovery
("throwaway") type. The throwaway methods, while giving
no return to help offset operating cost, have the advantages
of relative process simplicity and of avoiding the
complexities in chemical product marketing. The sulfur
oxides are sorbed by lime or limestone-the only sorbent
low enough in cost to consider-and the resulting calcium
sulfIte or sulfate discarded.
Limestone sorption processes are of two types-wet and
dry. Dry sorption involves injection of finely divided lime
IThroughout this report the word "limestone" is used in a general
sense to cover all limestones and dolomites, including high-calcium
limestone (mainly CaC03), dolomite (CaC03"MgC03)' and
compositions that give intermediate CaO:MgO ratios.
2The following "names" are used to distinguish between the three
process types and combinations thereof:
Dry limestone injection: No scrubber involved
10
or limestone into the boiler and collection of the calcium
sulfate product, along with the fly ash, in the usual
dust-removal equipment after the boiler. In the wet nroce~~,
. estone absorb fur oxides in a
scrubber. at the end of the ~" The e
introatic"@,"auectlxinto-'thescrubberre. circula. tin~~
- 4. I
~t can be in' ected i?to t .. .'. . ~~ss. e
main a vantage oi~JaU:er is that the iniUd li.rr1estone is
~~~ancli<.~..I~~ (which is
caught in the scrubber) is a bet1eL absorheq,t than
~ne; calcination in separate equipment would be
quite expensive.
The main advantage of the wet process over the dry
method is more complete utilization of the lime or
limestone. The longer retention time afforded by the wet
scrubber, plus the more favorable conditions for mass
transfer, give on the order of three to fivefold more
complete reaction of the absorbent. A further advantage is
that the wet scrubber can remove the fly ash as well as the
sulfur oxides. But offsetting these good points are the
higher investment for the wet process and the loss in stack
gas buoyancy resulting from cooling in the scrubber. Some
degree of reheating is generally considered necessary, even
with a high degree of sulfur oxide removal, to improve
dispersion of other pollutants such as nitrogen oxides and
residual dust. The .economic effects of these opposing
factors have been evaluated and are presented in this report.
Much of the background for use of limestone as a sulfur
oxide sorbent, e.g., limestone availability and mining
technology, was given in the earlier report,3 and has not
been repeated in the present one. It will be helpful to those
using the present report to have a copy of the previous one
available.
Limestone-wet scrubbing: General term to distinguish from dry
injection; no specification as to point of limestone addition.
Limestone injection-scrnbbing: Injection into boiler followed by
wet scrubbing.
Limestone-scrubber addition: Introduction into scrubber
recirculating liquor.
3See preface page 3 for title and ordering information.
-------
History and Status
The use of limestone or lime as an absorbent for sulfur
oxides in a wet process has been studied by several research
organizations over a period of about 30 years. Most of the
work has been small-scale, but some pilot plant and plant
tests have been made. In this section of the report the
previous work will be summarized as background for the
design and cost study.
As noted earlier, two methods for addition of limestone
to the system have been studied. Introduction of the
reactant directly into the scrubber system was studied first,
dating back to the late 1920's. Research programs have
been carried out by the British Electricity Authority ~'yrl
t ower Company (2, 3), 1m erial Chemical
I u ," eJ+ ,.._"c'";"",, '
(5 the S 'e "c Research Institute (S()viei:"lli~~~,..
an Wisconsin El~~!'S9IIl'pahD1J0hetwo steps
in limestone injection:scrubbing have been studied both
separately and together. Research on the injection step
(without scrubbing) has been conducted by many
organizations, particularly in Germany from 1950 to the
present; the history of this work is detailed in the dry
limestone injection report (1). In addition, TV A and
NAPCA are now studying the process in a program
involving a full-scale boiler. Both steps, injection and
scrubbing, were first studied together by Combustion
Engineering, Inc. (8).
All of these studies were concerned with removal of
sulfur oxides from power plant gases. In addition,
Murakami and Hori (9) have developed a process for
scrubbing tail gas from sulfuric acid plants and other
chemical operations in which a lime slurry is used as the
absorbent.
Summaries of most of the studies listed above are given
in Appendjx A.
The London Power Company did pilot plant work that
led to full-scale gas-washing plants at the Battersea and
Bankside power stations in London. Alkaline water
(alkalinity equivalent to 200 p.p.m. of sodium hydroxide)
from the Thames River was used for scrubbing the gas on a
once-through basis. During part of the program lime or
calcium carbonate (chalk) was added for pH control. The
system removed a high percentage of the sulfur dioxide
(about 90%) but the acidic effluent from the two stations
lowered the pH of the Thames to an undesirably low level
and thus prevented use of the process at additional plants.
The once-through system also had the drawbacks of
requiring a very large amount of water and of cooling the
gas to an unduly low temperature.
"v.,"- .,
-'..0.,-, .."..;...,~:;..;::;::....-~-.~,,,,..G-;<;~":;'~-
f.i'~
In another study, a process utilizing a circulating liquor,
ept alkaline by addition of lime or chalk, was developed by
James Howden and Co., Ltd., and Imperial Chemical
Industries Limited. The process was used commercially at
the Tir John (Swansea) and Fulham (London) power
stations. In these plants, both lime and finely ground chalk
were tested; lime reacted stoichiometrically with sulfur
dioxide but about 30% excess calcium carbonate was
required for comparable sulfur dioxide removal,
presumably because of slower reaction rate. About 98%
removal of both dust and sulfur dioxide was achieved.
Scalin was controlled but corrosion was a problem.
" ,', 'tlttle" further wor h~mrt'l~pl:1~"'iW'" m the
past three decades. The Tir John plant was not operated
very long and the Fulham plant was shut down during
World War II for military reasons (because the plume was
believed to be guiding enemy aircraft); it was not
rehabilitated after the war (2) because of" the cost and
because emphasis in sulfur oxide control had shifted to use
of high stacks. As far as is known, the Battersea and
Bankside once-through scrubbers are still in operation.
During brief pilot plant studies by TV A (5) in the early
1950's, a packed tower was used to scrub stack gas with a
10% slurry of pulverized limestone (95% -100 mesh). About
89% sulfur dioxide removal was obtained with
stoichiometric addition. Evaporative cooling of the inlet gas
was necessary for good scrubber performance.
~ot £.!a~~ s~n~bbeL::!udies w~~~.w J.9j,?.,,~~~9~
p'rogram beiw.een~istionsin-Ele.G.1r¥'.~~~L,~...ll.ni~al
0U%tlJ(;ts,.*{\ir,c;:h..Qy~i,QJh'pj.~JZ1.; A mobile-bed
~, Contag" Ahsmb.c.LL",\lt!Bime.oI.pon,.o£._'Ct~~g;,tS-¥~.
'~9:qi~~i!tt~uly.erized.fu~1~b9ij~J;.,.4,.W~~pnsin Electric's
Oak Creek station. Circulating a slurry of pulverized
limestone (90% -200 mesh) through the scrubber removed
about 85% of the sulfur dioxide and 99% of the fly ash. In
a separate study, dry injection of limestone was investigated
but the combination of dry injection followed by wet
scrubbing was not tested.
~~K~rm.~j}~JIori(91-de~el~p.!:~Ses,~~
10}V concentrations of sulfur dioxide from industrial waste
gi~es. Th~gas,SJ,L~~9.!'~ip...;)" sprar ~~jl<;(J,'!Y&.x.,~"'~t~
~eU19ved an!:tJbefi. p~i~~~
£~9H~,Lstilfite. T~~~1ililtecisc oxidized,::witQ.
a,fo~lZe~..~~1lc~~!,l>i~,~PS1.lm is.separatedl;!~,.ceIl,t1ifu~~~J~
~~EE~~_c,~..c]:e,!tY:X1!l~!,~,<:l J9,,1~tJl~f<1~~'{s. High-quality
11
-------
gypsum is produced. The method is reported to be in use in
two plants in Japan, one a sulfuric acid plant and the other
a chemical plant in which gases from an oil-fIred tunnel kiln
are scrubbed. Mltsubishi Hea.2Jn~u~!!i~'~'''~!.~'k~~g
the1'}ro'Ce.~s f~1Ly.&ejn~~lan~s> (10).
All of these studies were concerned with the scrubbing
step only and demonstrated that lime is a good absorbent,
capable of removing 90% or more of the sulfur oxides. Raw
limestone under similar scrubbing conditions gave
somewhat lower efficiency.
T~e COnCp.Dt of caI.c~~j,g1_~j?rioE t?S~~j~gl>¥
~p.iLU2!.0.1);le k-9jl~P:P;;l!t;:fljlx-~~~fJr-,stW_°1'()sed Q~
Wisconsin Electric but was first test~dhby (9):!1,b.u~tign
Engin~~~iEi~i~i)fcr~~ni~~J2
-------
Process Chemistry and Kinetics
Very lit,tle it!.J'};),@".W&~~GaJ"afld kiaebc.datahiQe, a reaction that is known to
occur in scrubbing sulfuric acid plant tail gas with lime
slurry (in the absence of carbon dioxide). No data appear to
be available on the competition between this reaction and
those listed above.
Or the system co~ PK JjmjJ,\J:J;",1Rjl,JJJ,W~J~~7~cIl!pbirfl'
wh(!re it ~ P'eneralJ~~4eI.e:<;lthatcarbonate isno.t
involved and that sulfJ.1::js thepJ;W,9rID;M;.¥~g,ta,I}t.
IjNH4);S00W:':~'~~~~' .,
J
(6)
Ammonia added to the recirculating
regenerates the sulfite.
NH4HS03 +NH3 -+(NH4)2S03
scrubber then
(7)
Such a system seems unlikely in lime scrubbing, however, as
c~um sul!!t~,~~~.¥-£9,!M1.
-Thu's the scrubber slurry confiims mainly CaC03,
CaSO 3, and CaSO 4 (plus MgSO 3 when dolomite is used) in
suspension and a variety of compounds and ions in
solution. Ib~ ~lub~:!.ationships in the com~x ~hase
s~eI]).,"".J,;~\.!J.,tj,!1g ,,:b.aye,w.~<~~~,put. F~er
*~~~JiQ!1"j~.,...,_~¥u.desirable,-,""~~-,,eo~
understandiIl,ghot4 .of,.,su1£ur .dio~de..~Q~tj,~~(;L.g~
~<~i~'b~;"c~~a1iDgc~"£~i- -PY...4gJM>4'~~~'r~L,,,~C}£i~!i!!&
s2tjqs. ' ,- .
Engineering considerations require that the absorbent
slurry be recirculated to the scrubber, with a side stream
drawn off to remove the "make" of calcium sulfite and
sulfate. ~~.~h.;:.~.~~l~.!_in~ s~l}r~~,~ha!l~~~s...~~~£~~!i?~
mf~il£'i=:::~~~~~;;'di~~~
d~~~~~~~Jl~~c~s... !~~1-C-~~-,,~o!..~~,~~
because accl2rpjp.g to ]>~a!s2n, etaJ.'J. the bicarbonate
~~~~.O"~" _..,,-,~~ .I.-"-"'>~-.~~,7 jO'~,J_'fi;I!'I'K:.mo"""i!J'.<'i>....~'V'"~!:
13
-------
reacting with the sulfu~!oxid&,Ve~I~~!,9_£X~~A<;.tiW;),Ji
~-C~il~m"9~~~~itb~~,~e, ~s
~Vin~ a stab~(~). ~r
'mes one :j, oweverl?1!,}oes ,tGnd tg mc~.~
particularly when the absgW,ent ~i!;\4~g,.~t~!~lp.@s~!.tj.pp,\:!.
The main effect of this, as re orted b. Nonh~bel I . t~
force calcium sulfite and ~,~.eout);r:J"q,}!tWJliJ~J tbe~
increase invol~~~_t2:a:~~¥=~~~.
sulfite decreased b¥=about 20%. No data were presented for
~<';'....- -~ ,"-'~.~.jt'~-----==.'"
calcium sulfate, but it seems likely that some sulfate
precipitation would take place also. ore data are needed
relatin~ pH in - tho! sys~~IIF-3'" ()'f',;~, Jlqui~
PJlise \11 t~,,~\ijy.Tfus is important in relation to the
scaling problem, a~5 t/Ju PIi~c~~~~.tc>",ie
~~~J1i.uS-:;JI!iJ~:llhe-oSCn;j!bmeFc..G.i!rci1Iif't
Oxidation of sulfite is another phase of the process
chemistry, and also one that is not well understood. It has
been generally considered that oxidation is desirable,
because if the sulfite is allowed to escape into watercourses
it will gradually oxidize there and reduce the oxygen
content of the water. Moreover, in the Howden-ICI tests
sulfite was said to interfere with prevention of sulfate
scaling. Oxidation before disposal was a major objective in
the British work.
Lessing q 3) founi. that sulfite wasr~~~..~)~j~~,~
solution hilt noL.a~ " ~CI~~ilati<2.n. Soll1hj~~
~u~te increases j,~..!?lt~I~,b~cc:ke~pi,!:1.gth~c.pHbelQ\V.
...... at the scrubber 0~1!~!,dh,~,,w~.Qf-su.lJite~wa4ilW,,,,~p
~e~t .mw~on. However, scrubbing efficiency drops
with pH, so good control to hold the pH below and as close
as possible to 6.4 was necessary. ~:,Jl
~ld.,Drns9'e~~ess..: o.t QX¥g~~:Wr.J~,;stackJjs (over
the amount required for SOz oxidation), <;!!!!Y~~SO%
o~~tio~.:wasF ~e'ied,;m~7fb~" ~Qwden-ICI-pi1ot- plaRJ
_h~r. When the scrubber liquor was passed through a
separate tower in countercurrent flow with air (about 2% of
the combustion gases), nearly complete oxidation was
obtained. IL was sp.~~~ted, ~:t ~~all . ~.!!I.J1~!.k-.Qf
man anese from me coat ~lt lliIlestone served asu au
oxi bon cat ysr:ricr!ii~;~:~~at~Uersea,~~~ati.Q~
01.. the e~t,,,~,,,tlke,,-presence of iron reduced sulfite
c~!AtiWlLb:y 11 -factor of i O.
In the pilot plant work by Combustion Engineering (8),
the scrubbing effluent contained about equal amounts of
sulfate and sulfite. Similar results were obtained in the
Howden-ICI work (4). Murakami and Hori (9) use a
pressure oxidation step to convert sulfite to sulfate, the
objective being to make salable calcium sulfate.
During studies of absorption of sulfur dioxide in lime
slurries, ~'~C,9~cl1Jded that tll~.ra,te.of0xidatiQl1
~~M~ c tQ :$ulfite 'wasproportional. to the amountef
.u~ oisulfite ill tne sCfubberJiquoI. iWttitiOilof:astTlalf
a'm!f}~~O,004%)of p-aminophenol reduced .oXiitt-
,t>i~~t-oftYefolct,
14
Other uestions can b
sulfite effect 0 oxidati e
~ation rate o~cium and ma!!nesium slllfite~ fl1rther
ways to inWt ~r prp~ Q?$i~4ir0n (analog~us .t? data
obtained in ammonia scrubbing), and deslrahll1tv of
?xidizing relatively i(lID'!luble m~gnesium sllijite to soluble
~. Few data are available that would help
ing any of these questions.
Litt1~ information is available also on the kinetics of the
various reactions and transfers that take place in the
scrubber-hydration of calcium oxide and dissolution of
calcium hydroxide in the boiler injection process,
. dissolution of limestone when the limestone is introduced
. into the recirculating liquor, possibly the absorption of
carbon dioxide in the scrubber, and, of course, transfer of
. . sulfur dioxide to the reaction site and reaction with the
absorbing species, whatever it is. Data on these must be
I f obtained before the controlling mechanism or mechanisms
.1
! can be determined.
For trans er of the gaseous solutes, a molecule of solute
must be moved from the main body of gas to the gas-liquid
interface, absorbed into the liquid (with or without
chemical reaction), and moved from the interface into the
liquid by diffusion or turbulence. Degree and rate of
absorption depend on the difference between the partial
pressure of the gaseous solute and its vapor pressure above
the absorbing liquor. S' entration of sulfur
d'o . , er lant stack gas is low -0.3 0 the
va~or p~~.~W ~~,jn~ soIJA1io~, m.yst,)e quit~
~~vj~"."J..."dr,\y-Wg.12~L iRr ,~~Qrption. A
complicating factor is that the temperature of the scrubbing
liquor governs the vapor pressure of absorbed sulfur dioxide
at a given concentration, and, without external cooling, this
~uor te:yrat~re is fIxed at about l~Y. by the heat
~onieni-'~t..Jlul!1.Jemne.W.~ of the stack gas. More~r,
low H 'ves ' va or ressure when lime is used as the
a2sorbe~!.k!h~ft:::of~pll,..~y,~gJ;. pressure 0 su ur
~~.-.:~.,.q~,~~~s.1!\fjte-l:~l.utions as
d~.w~Y::~~J!~~5?~,,~.~- shown in Eiam~ 1.
Therefore, the - ure must be ke t low b
maintainin the e fective conce e s ur
dioxide at a low eve, either by scrubbing with a very large
~ror w7~'~ was done at Battersea and Bankside or
by keeping t~~J,~~l>Y lime addiU~s was ~one in the
CX£li£pro"e.s~s.
Presumably, mass transfer of carbon dioxide is not a
problem since its concentration in the gas stream is
relatively high; however, no data are available on the
transfer rate.
Transfer rate of the calcium may be an important
consideration, especially when raw limestone or dolomite is
used. Here the solute molecule does not have to move
through a layer of molecules to the interface, as in transfer
of gases, but must be pulled away from neighboring
-------
2
C'I
I
E
E
.
'"
o
(/)
a..
4
5
6
pH
Figure 1. Effect of pH of calcium sulfite-bisulfite
solution on 802 equilibrium vapor presswoe
molecules to which it is much more tightly held than in a
gas mixture. Thus particle size, degree of crystallinity,
crystallite size, deposition of reaction products on particle
surfaces, and other factors enter. The kinetics' a
stiuatio ot ,been
:IC.~~'"
studied for the r~() ~2Q.)-CO~ -SO~:~QJX~.t~W'
Available data indicate that lime (CaO) does better than
raw limestone, presumably because the small particle size
resulting from calcination and subsequent hydration makes
it dissolve faster. Adequate data are not available, however,
to evaluate either this factor or other parameters that may
affect rate of dissolution.
Some data have been obtained on overall transfer rates
in pilot plant scrubbers. Chertkov (6) reported that K
(lb./sq. ft.-hr.-atm.) increases in direct proportlOn to the gas
flow rate at low values of the latter. The increase in K was
from 40 to 80 as the gas velocity was increased from 1 to
2.7 feet per second in scrubbing gases containing 0.2 to
0.3% sulfur dioxide in a grid-packed absorber with lime
slurry containing 32% solids. The coefficient also increased
with decreasing solids content in the scrubbing slurry; with
20% solids, the value of K was 160 at a gas velocity of 2.7
feet per second.
In summary, the chemistry and kinetics of the
limestone-wet scrubbing process are quite complex and
largely unexplored. Most of the investigations have been of
the applied type, with the result that the basic mechanisms
and reaction kinetics have not been studied adequately.
ta are needed es e'all on 1. the main rea f
~{~Yff'~\~~'~6f- f~fht~iii;,:~~~r_t~~\1i:~~k6£~:
relation!.hips mtfi~scrubb~q}>IiuiOll. -
""""""""~c:."; "~,3:.0'~;'-"- - - -, '- n ---- -" -"~
15
-------
Process Considerations and Problems
The preceding history of the limestone-wet scrubbing
process and the discussions of process chemistry and
kinetics leave several questions that must be resolved before
settling on a conceptual design. There are so many
considerations that it is difficult to define a basic process
that gives the best combination of the various parameters.
The considerations that bear directly on process selection
will be examined first.
Point of Limestone Addition
The major process choice to be made is between
injecting limestone into the boiler and introducing lime or
limestone into the scrubber circuit. The early workers
apparently never considered the possible advantages of
injecting into the boiler and worked rather on the problems
of introducing the absorbent into the recirculating scrubber
liquor. The practice has grown, however, of injecting small
amounts of a basic material, usually calcium or magnesium
oxide or carbonate, into boilers-not to remove sulfur
dioxide but rather to reduce boiler corrosion. Thus when
attention was turned again in recent years to reducing
pollution caused by sulfur oxide emission, injection- of
limestone into the boiler as a means of controlling both
pollution and corrosion gained favor. But although this
appears quite attractive, there are major potential
drawbacks that may offset the advantages. The large-scale
tests now under way at the Union Electric and Kansas
Power and Light plants should provide useful data, as will
the forthcoming tests of dry limestone injection by TV A
(1). However, since no large-scale tests of scrubbing with
absorbent introduced directly into the scrubber circuit are
planned, the Howden-ICI experience of 35 years ago must
be used (along with TV A unpublished data) for
comparison.
The good points of limestone injection-scrubbing can be
summarized as follows:
1. Better absorbent utilization. Limestone injected
into the combustion space of a boiler is calcined almost
instantaneously. This is a prerequisite for the dry
limestone injection method and is also helpful in the wet
process, because the freshly calcined lime passing
directly to the scrubber at the end of the system goes
into solution more rapidly than does raw limestone. As
noted earlier, those investigators who have compared
calcined and raw limestone all found that more
absorbent was required when uncalcined limestone was
used.
The limestone does not have to be calcined in the
16
boiler, of course; in the British work, lime made in the
usual manner, in commercial limekilns, was used.
However, cost estimates presented in the report of the
TV A dry limestone injection study showed that
calcining in the boiler is an economical procedure and
that the cost of calcining in separate facilities is
prohibitive.
Hence economics would dictate use of raw limestone
if the absorbent is introduced into the scrubber circuit
rather than into the boiler. On the basis of past work, it
must be assumed that more. absorbent would be required
by this procedure and that injection into the boiler
would therefore have a major advantage. However, there
does not appear to be any reason for raw limestone to be
inferior if it can be dissolved at an adequate rate; further
development work may make this possible.
2. Reduced boiler corrosion. The potential reduction
of both high- and low-temperature corrosion in the
boiler system as a result of limestone injection was
discussed in the dry process report. The tests at Detroit
Edison indicated that reaction of the limestone with
corrosive alkali salts, which normally reduce tube life in
the areas exposed to maximum heat, should reduce
high-temperature corrosion. Also, virtual elimination of
sulfur trioxide in the combustion gas should decrease
any corrosion in the air heaters caused by condensation
of sulfuric acid.
However, the stoichiometric amount of absorbent
needed for sulfur dioxide removal is far more than
needed to react with the relatively small ambunt of
alkali salts and sulfur trioxide present in the gas stream.
It should be possible to avoid the potential drawbacks of
boiler injection and still retain the advantage of
corrosion reduction by injecting only a small amount
into the boiler and adding the remainder in the scrubber
circuit. Some additional investment would be required
but it should not be large.
Injection of part or all of the limestone into the
boiler would have a major advantage in using one of the
more promising methods for stack gas reheating, that of
exchanging heat between gas before and after the
scrubber. Without limestone injection into the boiler,
the sulfur trioxide would condense in the form of
sulfuric acid as the gas cooled in the exchanger, and it
would be necessary therefore to use relatively expensive
construction material.
3. Sulfur dioxide sorption in the boiler. The
limestone injected into the boiler should react with part
-------
t,.so i
of the sulfur dioxide as it does in the dry injection
method. However, since the amount of limestone needed
for good overall sulfur dioxide removal in the wet
process is considered to be near stoichiometric, the
degree of sorption in the boiler would not be expected
to be more than 20 to 30% because of the intrinsic
inefficiency of the dry process. Getting this much
reaction in the boiler is desirable, since it does not have
to be done in the scrubber. The net effect is that the gas
entering the scrubber would contain essentially no sulfur
trioxide and less sulfur dioxide. This might reduce the
scaling problem, as the calcium s~e
boo r woul not ave to e preCIpitate in the scrubber.
Scrul:) er size and cliculation rates, owever, probab y
would not be benefitted.
It should be noted that even 20 to 30% sulfur dioxide
removal in the dry injection process is expected to
require very careful operation in regard to point, angle,
and velocity of injection and possibly very fine particle
size. Going to all this trouble may not be warranted in
the wet process because any sulfur dioxide not reacted
in the boiler will be caught in the scrubber. Hence the
degree of reaction in the boiler may be quite low in
practice.
Offsetting these advantages are the following potential
drawbacks, which in turn are advantages for introducing the
absorbent into the scrubber circuit.
1. Interference with boiler operation. One of the
major fears in regard to the dry limestone injection
process is that the large amount of limestone injected
will interfere with boiler operation, by excessive slagging
on boiler surfaces, interference with fuel combustion or
heat transfer, increased erosion, and perhaps other ways
not yet considered. The wet process has the advantage
that much less limestone will be injected; however, even
the stoichiometric amount constitutes a very large
introduction of foreign material into the boiler, roughly
equivalent to doubling the solids content of the gas
stream. Moreover, the reduced amount of limestone
added as compared to the dry process might actually
have an adverse effect, by reducing the slag fusion
temperatures and thereby aggravating the slagging
problem.
The possible adverse effects of limestone injection are
discussed in detail in the dry process report. Experience
from large-scale tests will be required to clarify the
situation. In the current plant tests of limestone
injection-scrubbing, no particular problems with boiler
operation have been encountered. However" a much
longer period of operation will be necessary for any
conclusions. The TV A-NAPCA dry process test program
will include a 6-month continuous run to evaluate the
problem.
2. Increased scaling in scrubber. Calcium sulfate and
sulfite are notorious for their pro ~~
ro solution as an erent scale on process
.equipment. The general scaling problem in limestone-
~bing is discussed in detail in a later section. For
the present discussion on effect of absorbent intro-
duction point, it will suffice to say that boiler injection
necessarily brings the absorbent directly into the
scrubber with the gas, whereas if limestone is introduced
into the scrubber circuit it can be added after the
scrubber. This could have an important effect on scaling
because as noted earlier, the pH increases at the point
where the lime or limestone first comes in contact with
the scrubber slurry, and as a result calcium sulfite and
sulfate are precipitated. If this occurs in the scrubber,
scaling can well result. If it occurs in the slurry circuit
after the scrubber, it can be made to take place in a hold
tank on the surfaces of existing crystals.
As will be discussed later, the method developed in
the Howden-ICI process for preventing scaling was quite
effective, whereas the Union Electric scrubber has scaled
severely. The difference may be in the point of lime
addition, for the British introduced it after the scrubbxrSO d 4.-
and before a hold tank. ~,~ lr..l
. . He.nce the~e is the possibility that the '5IJ
IllJectlOn-scrubbmg process may have a considerable and 'It... 3
unavoidable disadva~ because of calcium salt -.
~osition iR tl;l.I: o{'TlJhhp.T FlJrthl'T phnt I'YrPrlPQ\:~ i~ ~
necessary to pVJlaatt tIre problem. pn. .uJ..,,,
--- \)1/0 v t.rAJ
3. Heat required for calcination. Although injection ~
into the boiler is an economical way to calcine J j .
limestone, mainly because equipment requirement is .Pv-r-
low, the amount of heat required is the same as if the ~
calcination were done in separate equipment (except for ~
possible heat loss in off-gases in separate calcination). In
contrast, if limestotle is introduced directly into the
scrubber circuit, the heat requirement is eliminated. The
advantage for scrubber addition is not large but is
significant; from figures given in the dry process report,
the saving is estimated at $0.036 per ton of coal (for
3.5% sulfur in the coal and $0.26 per million B.t.u.).
4. Loss of unreacted absorbent in solids disposal. In
injection-scrubbing, the sidestream to the disposal pond
is withdrawn after the absorbent is introduced into the
scrubber circuit rather than just before the introduction
as in the Howden-lCI process. Therefore there may be
considerably more loss o'f umeacted absorbent to waste.
The loss would depend on rate of lime dissolution, rate
of slurry recirculation, and other factors.
It is concluded that there are not enough definitive data
available to favor one process to the exclusion of the other.
Therefore, conceptual designs for both limestone
17
-------
injection-scrubbing and limestone-scrubber addition will be
presented in this report.
Point of absorbent introduction is also important within
each process. For example, in the injection-scrubbing
method, the limestone can be added with the coal or
injected through the boiler wall above the flame area;
addition with the coal (in pulverized-fuel boilers) is more
convenient and probably more economical in most cases.
For the dry injection method, where operation must be
carefully optimized for maximum sorption in the boiler,
addition with the coal is questionable because of the danger
of "dead-burning," that is, loss of porosity because of the
high temperature to which the limestone is exposed during
\~nation.
)., A' ead-burning probably is less of a roblem in e wet
~ .).. rocess, eca rate of dissolution in the ..
. y~,.: the main consideration rather that rate of .
. . '0 in~ic e. However, it may well be that
,.D f dead-burning, which involves growth of CaO crystallites to
a size larger than in normal calcination, could slow down
dissolution rate to such a degree that absorbent utilization
would be reduced. Further data are needed. As noted
earlier, the limestone will be injected with the coal in one
of the Kansas Power and Ught installations.
Choice of point of addition when the absorbent is added
to the recirculating slurry presents less of a problem. ~
introduction point should obviousl be after the slurry has
Ie t t e scru er order to reduce calcium salt
precipitation in the scrubber.
~
Dust Removal
The ability of a wet scrubber to remove dust as well as
sulfur oxides gives limestone-wet scrubbing a considerable
advantage over the dry system for sulfur oxide removal.
Eliminating the usual dust-removal equipment gives a major
reduction in power plant cost that can be credited to the
wet-scrubbing system.
When a new plant is involved, the cost reduction is a
very strong incentive to combining removal of dust and
sulfur oxides, even though the presence of the dust
complicates design and operation of the scrubber
somewhat. The main disadvantage is that without an
electrostatic precipitator, the scrubber would have to be
operated continuously for dust removal and the option for
intermittent sulfur dioxide control would be complicated,
as discussed below.
When an existing plant is in question, most of the ash
and lime could be removed in the existing dust-collection
equipment, leaving the scrubber only that part that passes
through the dust collectors. In this case the lime-ash
mixture caught in the dust collectors would have to be
transferred to the scrubber, which probably could be easily
arranged since the mixture could be dumped directly from
18
the dust collector into the scrubber recirculating system.
The question is whether continued operation of the dry
dust collectors would be economical as compared with
removing them and putting the full dust load on the
scrubber.
Wet scrubbing for dust removal is seldom used in power
plants, mainly because of high pressure drop and gas
buoyancy loss. In the present case, however, these
drawbacks must be accepted anyway in removing the sulfur
dioxide. The remaining disadvantages of eliminating the dry
collectors are:
1. High dust loading on the scrubbers, which makes
some scrubber types unsuitable and may introduce
operating difficulties.
2. High solids content in the recirculating slurry. This
assumes that if the precipitator were retained the
ash-lime mixture would be fed into a delay or settling
tank and a thin slurry of controlled solids content drawn
off for feed to the scrubber. Control of solids content
was found important in the Howden-ICI work.
3. Possible interference with intermittent operation.
If the existing dust collectors were left in place, the
scrubbers could be shut down since the existin!
equipment could take over the dust-removal function.
Otherwise continuous scrubber operation would be
necessary to remove dust, the same situation as when
aust and sulfur oxide removal are combined in a new
plant. Costs could still be reduced by feeding lime only
when weather conditions were poor for plume
dispersion, but during the nonfeeding periods the
sorubbing water would rapidly become acidic, in which
case resistant construction material would be required
and it would probably be necessary to recycle the acidic
scrubber effluent from the pond.
4. Sea' robably would aggr te If
c llectors were left in, lime would not go directly
into the scrubber and therefore the calcium salt
precipitation in the scrubber mentioned earlier would be
avoided.
If limestone is introduced directly into the scrubber
rather than injected into the boiler, there is even less
justification for removing the dust collectors. In this system
the scrubber does not have to be designed for dust removal
(although residual dust from the dry collectors must be
considered in scrubber operation) and the ash is kept
separate from the lime (which may improve scrubber
operation and possibly make the reaction product more
useful as a byproduct).
The main advantage of removing the dry collectors is
eliminating the cost of operating them-assuming that the
4 Assuming that a practical equipment arrangement could be worked
out; see later discussion.
-------
scrubber operating cost is not increased to an equal or
higher degree by the dust loading. These factors are
analyzed in the cost section of this report.
Scaling
Whenever solutions containing calcium sulfate or sulfite
must be han in a rocess, trou e with scalin can be
e~ected. t is a major pro lem in the ertilizer industry
where, in the production of phosphoric acid by the wet
process, calcium sulfate formed by reaction of calcium
phospahte ore with sulfuric acid crystallizes in lines and on
equipment surfaces to the extent that the plant must be
shut down periodically for "boiling out." Scaling was the
only major problem encountered in development of the
Howden-ICI process and the current experience at Union
Electric indicates that it will be also in process
developments in this country.
Although solids in the recirculating scrubber suspension
serve as nuclei for crystal growth, nucleation can also occur
on scrubber surfaces with resulting scaling, particularly
sinc both calciu sulfate and sulfite form su ersaturated
sQlutions fro!D-umkh precipitation ~en whe~
degree of supersaturation is high. In other similar systems
o
o
o
0'
o
..
Q)
~ 400
~
ca
0.
c-
o
'';:::
::I
'0
'"
,=
%-suspended Caso4
, t',,",
\ 1% ',." "-
~ " '. "'.
., t', ',.. '.
~ 2% '...... --- ..~.
" -..-.. ..~.
. ........ -':---. ~
~ ........ -- ............ .. --:... - . --- -
. ~ .-. .. ---- ----- .
~ 5% ----
:',. -----....
. --
e. . . . . . . .. :-: -:-. :-.?: -po .. ..
t . . . --.-...........................................
7%
600
2
4
the concentration of dissolved salts in the recycling liquor
can be reduced by cooling outside the scrubber and
supersaturation within the scrubber thereby avoided.
However, the solubilities of both calcium sulfate and
calcium sulfite have an inverse temperature coefficient and
the method therefore is not applicable,
The composition of several sam les of scale
reported y Lessing (1
~
Ca(OH)z
CaCO 3
CaSO 3
CaSO 4
Nil
0-19.7%
0.8-14.3
62.8-97.2
~
~~~
~
~I
nil
6
8
10
Time, minutes
Figure 2. Desupersaturation of calcium sulfate
solution at 50° C. by suspended Caso4 . 2 Hz 0
19
-------
slurry returning to the scrubber as a means of providing
surface for salt deposition at all points in the system; (2)
use a relatively high slurry recirculation rate to minimize
buildup of supersaturation in the scrubber; and (3) provide
a special "scale resistant" packing (suspended wood plates)
in the lower section of the scrubber to prevent trouble from
the small degree of supersaturation developed at that point.
By use of this technique, the Howden-ICI workers were
able to eliminate completely the severe scaling that had
been their main problem in developing the process. The
plant at Fulham was operated for an extended period with
no evidence of scaling. It should be noted, however, as
pointed out earlier, that much of the freedom from scaling
may have been due to the fact that the absorbent was
introduced between the scrubber and the delay tank rather
than directly into the scrubber. Absorption of sulfur oxides
in the scrubber builds up the calcium sulfate and sulfite
concentration but the accompanying decrease in pH
increases solubility of sulfite and helps avoid
supersaturation in the scrubber. Introduction of the lime or
limestone increases pH, decreases solubility, and induces
precipitation or supersaturation. If the introduction is just
before the delay tank, precipitation and dissipation of
supersaturation occur at a point where they cause no
trouble.
The scaling reported (11) in the current tests at the
Union Electric installation possibly is caused by
introduction of the lime directly into the scrubber. Or
other factors identified by the British workers may be
involved; further data from the tests are needed to clarify
the situation. The mobile-bed scrubber type in use at Union
Electric was tested in the pilot plant programs at Detroit
Edison and Wisconsin Electric. Scaling was not reported to
be a problem; apparently the movement of the marbles or
spheres against each other prevents scaling in the bed, but
good liquid distribution is necessary to prevent inactive bed
areas in which solids deposition could occur. Moreover,
scaling can occur in areas not in contact with the mobile
bed.
Corrosion
Corrosion can be a major problem in wet scrubbi,ng to
remove sulfur oxides. Unles the pH of the scr bber
solution is carefully controlled ~ly neutral levels,
attack o~tal surfaces is serious unless expensive
all~d. orrosion 0 uct work and scru ers was
said to be a major problem in the early English work;
however, no corrosion data are given in the various
publications. The general statement was made that mild
steel is satisfactory for the scrubber shell if the pH is
controlled at a minimum of about 6.2 at the scrubber
outlet. A more resistant material was required for scrubber
packing, valves, pumps, and lines (at some points).
Corrosion was not a problem in the pilot plant work at
20
Wisconsin Electric; the scrubber was of fiber
glass-reinforced polyester construction, a material resistant
to both acid and alkaline attack. Use of this material for the
large scrubbers required in commercial application of the
process would be complicated by structural problems.
However, one scrubber manufacturer indicated ability to
produce such units. Costs would be somewhat higher than
for coated mild steel.
Corrosion data were not reported for the CE-Detroit
Edison tests. The stack gas treatment facilities installed by
Union Electric and Kansas Power and Light include
scrubbers fabricated from mild steel and coated internally
with a bitumastic lining. The support trays, internal piping,
and spray nozzles are of stainless steel construction.
Equipment external to the scrubber, including recirculating
piping, pumps, settling tank, and line to the pond, are of
mild steel.
Limestone Effectiveness
In the dry injection process, the properties of the
particular limestone used are quite important because of
the considerable differences that have been found between
limestone types in effectiveness for sulfur oxide sorption.
The reasons for this are not fully known; impurities
apparently playa part, possibly by blocking absorptive
surfaces, and some limestones appear less resistant to
dead-burning than others. In any event, selection and
procurement of an effective limestone is a major problem in
the dry process.
It has been generally considered that limestone
properties would not be nearly so important in the
wet-scrubbing process, particularly in the injection-
scrubbing version. Although some sorption occurs in the
boiler, it is incidental and actually is not needed; the boiler
needs only to accomplish calcination, for the scrubber is
quite capable of doing all the sulfur oxide removal, as in the
British work.
It~onceivable, howev!!, that the same proRer-3s
which cause poo or tion in the boiler could also reduce
the d' sso u Ion rate of the c . ed limestone when it
~ch~ e scrubb~ In the boiler the problem is
developing a large sorption surface in the lime particles and
getting the gaseous sulfur oxide into the particle interior
and into contact with the reactive surface; any limestone
property that reduces the surface or interferes with mass
transfer is undesirable. To some extent the same may be
true in the scrubber, where the problem is to get the
scrubber solution into contact with reactive surfaces of the
lime to get it dissolved. Insoluble impurities fused onto the
CaO surface, low porosity, large crystallite size of CaO, and
a shell of calcium sulfate on the particle all could delay
dissolution.
Offsetting this is the relatively long retention time in the
scrubber loop; a lime particle should have ample
-------
opportunity to dissolve because it circulates in the system
until it happens to be caught in the side stream going to the
solids-disposal pond. As noted earlier, however, the fact
that in the injection-scrubbing method the side stream is
withdrawn shortly after introduction of the lime into the
system increases loss of unreacted absorbent to waste and
therefore would aggravate the adverse effect of slow
dissolution.
No data e uestion. The only
integrated test of injection-scrubbing was in the work of
Combustion Engineering and no tests of limestone type
were reported. Further test work to evaluate differences in
limestone effectiveness are needed.
More information is available on effectiveness of various
limestones for the system involving absorbent addition to
the recirculating slurry. During the Howden-ICI work, it
was found that precipitated calcium carbonate was a
satisfactory absorbent but some natural chalks, even when
ground to minus 200 mesh, were not sufficiently reactive
unless used in considerable excess. In contrast, finely
ground limestone was a good absorbent in early pilot plant
tests at TV A, but precipitated calcium carbonate gave poor
results. In recent bench-scale tests at TV A, a chalk-type
limestone absorbed more sulfur dioxide (by a factor of
about 3), under the same scrubbing condition, than a hard,
dense, limestone of the prevalent type. In the pilot plant
work at Wisconsin Electric, pulverized limestone was an
effective absorbent.
These results indicate that differences in limestone
properties will be important in the scrubber circuit addition
method. It may be possible to use any type of limestone if
provisions are made to get it into solution before the slurry
enters the scrubber, but the conditions needed for
accomplishing this may vary between limestones. More
research is needed.
Solids Disposal and Water Pollution
The make of calcium sulfate and sulfite must be
removed from the scrubber circuit and conveyed to the
disposal point. In so doing, problems occur both in
transporting the waste and in disposing of the
accompanying water.
In the Howden-ICI process, the solids concentration in
the circulating system was controlled in the range of 10 to
15% by diverting a purge stream (from a point between the
scrubber and the point of lime addition) to a conical
settling tank from which an underflow containing 33 to
40% solids was withdrawn. The underflow was then filtered
on a rotary drum to reduce the moisture content of the
solids to 30-45%, depending on particle size. Partially
clarified overflow from the settling tank and filtrate from
the filter were returned to the process; solids were
discarded to a waste pile. Partial clarification was
considered desirable because the smaller particles returning
to the scrubber served as seeds for crystal growth. Filtration
rates were about 40 pounds per hour per square foot.
Sludge disposal was not studied during Combustion
Engineering's experimental work but concentration studies
were made by equipment manufacturers with mixtures of
fly ash and reaction products from the test program. The
Meramec installation includes a thickener for concentrating
the slurry to about 35% solids before it is discharged to the
fly ash pond; thickener overflow is returned to the
scrubber. At the Kansas Power and Light Lawrence station
the entire scrubber effluent is pumped to a settling pond
and the overflow returned to the scrubber.
No information is available on such oints as minimum
solids content or t e s urry to aVOl flow trou les in the
s uic e, erosion of umps an ines, and ~-
characteristics 0 - e dewatered s udge. ese are all
importan problems in the disposal of gypsum from
phosphoric acid plants in the fertilizer industry, where the
general practice is to sluice to a pond and by use of dikes to
build the gypsum pile to heights as much as 100 feet above
ground level.
The solids could be dewatered at the scrubber and
conveyed to the disposal pile by solids-handling equipment,
a procedure that would eliminate pumping problems. The
method is quite expensive, however, and is seldom prac-
ticed in disposing of gypsum from phosphoric acid plants.
A major problem in solids disposal is what to do with
the sluice water decanted or drained from the pond. It can
either be pumped back to the power plant and reused in the
scrubber or allowed to flow into an adjacent watercourse.
Recycling is much more desirable from the standpoint of
water pollution because it eliminates any release of salts to
watercourses other than from seepage or accidental pond
overflow. From the process standpoint, however, recycle is
undesirable both because it is more expensive and because
recycling the saturated solution may introduce operating
problems.
Hence water pollution must be weighed against the cost
and adverse process effects of recycling the water.5 The
cost of recycling can be estimated; it depends on the
distance from scrubber to pond. The effect on operation,
however, is difficult to predict. Increased scrubber scaling
might well result because the amount of fresh water added
to the recirculating slurry would be reduced; the fresh
water makeup normally used to replace the pond water
drained to the watercourse would be eliminated. Hence the
slurry entering the scrubber would have a higher degree of
supersaturation and scaling would be more difficult to
prevent.
5It should be noted, however, that recycling is not a perfect
solution to the pollution problem. In the phosphoric acid industry,
pond water recycle is often practiced but there is still trouble with
pond dike breakage and with pond flooding by heavy rains, with
resulting escape of the pond water to the watercourse.
21
-------
It should be noted in this respect that in the pilot plant
work reported by Pearson (4) most of the water was
removed from the solids and recycled. The resulting "mud"
contained only 35% water as compared with 60% in the
heaviest slurry considered suitable for pumping in the
present study. The overall fresh water makeup was 0.62 ton
per ton of coal burned, of which about 0.1 ton was from
water added to replace that lost in the solids; the remainder
replaced that consumed in humidifying the stack gas. The
pilot plant gave no scaling trouble under these conditions,
even with only 0.1 ton of water leaving the system with the
solids and replaced with fresh water; thus elimination of
this small amount by recycling might not have any
significant adverse effect.
Scaling in the sluice line might also be aggravated by
continual use of saturated solution as sluice water. If the
water is not returned, excess fresh water could be used in
sluicing to prevent scale; however, it should also be possible
to sluice with the fresh water makeup in a recycle pond
water system.
It is assumed that recycling the pond water is the
preferred course and that such a "closed loop" system will
be used unless the situation is such that the cost and
operating difficulty, coupled with a relatively low and
acceptable outflow of polluting salts, makes recycling
unjustified. The problems in deciding whether or not
recycling is justified are reviewed in Appendix B.
Stack Gas Reheat
One of the major problems to be dealt with in the
wet-scrubbing process is loss of stack gas buoyancy as a
result of cooling.6 Much emphasis has been placed in recent
times on use of high stacks (up to 1200 ft.) to improve
dispersion and reduce the possibility of localized high
pollutant concentrations. A reduction in stack gas
temperature by wet scrubbing will reduce both the
momentum and buoyancy which cause the plume to rise
above the stack after it is emitted. Thus, the effective stack
height and the plume dispersion will be reduced by wet
scrubbing. This should not be serious if essentially all of the
objectionable components are removed in the scrubber but
such efficiency is not likely, and, in any event, upset in
operating conditions could release large quantities of
pollutants, in which case the safeguard of good dispersion
through high discharge would be desirable.
Humidification of the stack gas is also objectionable
because condensation may occur and cause formation of a
visible plume which could have an undesirable
psychological impact on the public. The importance of the
visible moisture plume is difficult to assess. Power plants
have been operated with such plumes (in the U. S. and
France), presumably without difficulty. Also, large cooling
6 0
Normally to the wet bulb temperature, about 125 F.
22
towers, which are becoming rather common, produce
low-level plumes. It is concluded that the plume is
undesirable but that, other things being equal, no
considerable expenditure in eliminating it is justified.
These problems have been generally considered so
serious that after the first studies in England and the U. S.
the wet-scrubbing method fell into disfavor and emphasis
was shifted to dry methods. It is only recently that the
work of Combustion Engineering, Wisconsin Electric, and
others (including work in France and Czechoslovakia on
scrubbing with aqueous ammonia solutions) has brought
the method back into consideration.
The obvious solution to the problem is to reheat the gas
before release to the atmosphere, but this apparently has
been generally regarded as too expensive. No cost figures
have been published, however, to show that the cost would
be intolerable. Instead, cost estimates published recently by
Combustion Engineering indicate that both capital and
operating cost would be acceptable (8).
Very little test work on reheating has been reported.
Combustion Engineering operated the pilot plant with a
finned-tube exchanger inserted in the scrubber off-gas duct.
Solids accumulated on the surface but they were easily
removed by water washing. The system at Union Electric
includes finned-tube heat exchangers to extract heat from
the stack gas ahead of the scrubber (after the air heater)
and to reheat the exhaust. A circulating ethylene
glycol-water mixture is used to transfer the heat. Design
conditions provide for cooling the inlet gas to the scrubber
o 0
from 282 to 232 F. and for reheating from 123 to 173 F.
At the Kansas Power and Light installation, the scrubber
exhaust will be reheated by heat exchange with boiler feed
water.
There is general agreement that the cooled gas should be
reheated but no agreement at all on the degree of reheating
necessary or desirable. The consensus seems to be that it
need not be reheated to the original temperature but that it
should be heated at least to a temperature high enough to
avoid a visible plume under most atmospheric conditions.
The reheating problem is a complicated matter. The gas is
not only cooled in the scrubber but the composition is
changed radically; most of the sulfur oxides are removed (in
an efficient system), part of the nitrogen oxides are
removed,7 and the water vapor content is increased to
7 Data on nitrogen oxide removal reported by various organizations
are as follows:
Organization Removal of NOx, %
Howden-ICI 63-65
VOP-Wisconsin Electric 19-23
CE-Detroit Edison 20 (approx.)
The reason for the large difference between results is not known.
The burden of proof seems to be on the Howden-ICI data as they
are much higher than reported for other aqueous scrubbing
methods.
-------
saturation. The problem is how to predict maximum
ground-level pollutant concentration in the power plant
vicinity for the changed conditions in regard to gas
composition and temperature.
Effect of Stack Gas Reheat Temperature on Plume
Dispersion: TV A has carried out comprehensive studies of
plume dispersion from coal-fired, electric generating plants.
Based on this experience, the effect of stack gas
temperature on maximum sulfur dioxide concentration at
ground level was determined for a 200-megawatt unit with
a 300-foot stack. From full-scale field studies (16), the
plume rise ahove stack top was found to be:
f'..h
- (L)Z7
- C - 3
U
where
~h
C
F
plume rise above stack top
stability coefficient
flux due to buoyancy and
momentum = g Vsr z( ~~)
acceleration due to gravity
stack gas exit velocity
stack exit radius
excess stack gas temperature
over ambient temperature
ambient air temperature
average wind speed
g
Vs
r
6T
=
=
!a
u
=
The groundline SOz concentration at any point
downwind is given by the equation:
Q
X -------
= 21TCyCz U xm
where
exp [--~
2xm
~ yZ zZ)]
-- + -- (9)
C z C z
y z
x. = SOz concentration
Q = SOz emission
Cy' Cz' and ~ = diffusion coefficients
u = average wind speed
x = distance downwind
y = distance crosswind
z = height of plume (above the ground,
center of cross section)
By means of these two equations, the maximum
expected groundline sulfur dioxide concentrations were
determined for (1) stack gas temperatures (at scrubber or
reheater outlet) ranging from 125 to 250" F. at 25
intervals, (2) scrubber efficiencies of 95, 90, 85, and 80%,
(8)
(3) the plant operational conditions shown in Table 3, and
(4) wind speeds ranging from 6 to 30 miles per hour. The
critical wind speed (speed at which maximum SO z
groundline concentration occurs) ranged from 8 to 26 miles
per hour for the various combinations of conditions. The
calculations were based on neutral to slightly unstable
atmospheric temperatures, that is, near normal decrease in
temperature with altitude increase. It was also assumed that
the gas contained no mist at the point of emission from the
stack, and that therefore evaporative cooling of the plume
would not adversely affect dispersion.
Results of the calculations are shown in Table 4; the
values shown are based on a wind speed of 18 miles per
hour, as most of the conditions gave critical wind speeds
close to this value. The wind speed was held constant to
illustrate the effect of stack temperature alone on plume
rise.
The main conclusion from Table 4 is that maximum
ground concentration is increased by about 85% if the gas is
emitted at scrubber exit temperature (without reheating)
rather than at normal stack temperature. At high degrees
of removal, however, this increase is not highly significant
because the concentration is still quite low even after the
85% increase. For example, at 95% removal and 1250 F.
stack temperature the maximum ground concentration is
0.015 parts per million, which is equivalent to 91%
reduction in ground concentration as compared with the
!~~~~~~~~!..Qe.~~tl'!9...~'!..~:_~Q!~~.!~!~!!1_~~~!..(y~l!..1.!
Rated capacity, mw 200
Stack height, ft 300
Coal burned, Ib/hr 150,000
Stack diameter, ft 16.5
----~~Q~~J~~~~l~~~tJ~!~m~~L______-_JQ----
~l~!~Jli!ge~J?~@!!J!I!..:'E:
Entering Leavinga Stack gas
__~~l- -~~~-- "'y~I.Q~!Y....fp!
310 296.5 42.26
250 236.5 38.91
225 211.5 37.51
200 186.5 36.12
175 161.5 34.72
150 136.5 33.32
-_U~------______11L~_____--______11~~__-
~oal~~~~ -~--
Mo~ure 7]
Volatile matter 34.2
Fixed carbon 42.4
Ash 12.0
Sulfur _]~5-
-------------------------------_J]~~----
ayemperature checks show approximately 13.50 F. loss between
inlet and outlet of stack.
23
-------
0.16 parts per million for normal power plant operation
without scrubbing. At lower degrees of removal, however,
the 85% increase becomes more significant. Since the
ground concentration is directly proportional to the
amount of sulfur dioxide left in the gas (see equation 2), it
follows that at 50% removal (10 times as much S~
remaining in the gas as for 95% removal), the 0.015 parts
per million concentration becomes 0.15 parts per million,
or practically the same as the 0.16 parts per million at
normal stack temperature but with no sulfur dioxide
removal. In other words, 50% removal gives little or no
benefit if the gas is not reheated.
Reductions in ground concentration at various degree of
reheat are shown in Figure 3. This covers low degrees of
removal and thus shows the effect of gas cooling on ground
concentration of gas constituents such as nitrogen oxides
that are removed to only a limited degree in the scrubbing
process. If 30% removal of nitrogen oxides is assumed, the
gas would have to be reheated to 1900 F. to avoid an
increase in ground-level concentration as compared with
normal power plant operation.
The graph also allows evaluation of reheating by use of a
scrubber bypass (in systems where an existing electrostatic
precipitator is left in operation). If we assume bypass of
o
25% of the gas (at 310 F., the air heater outlet
temperature) and 90% removal of sulfur dioxide from the
remainder, the resulting gas temperature would be about
o
170 F. and the net degree of sulfur dioxide removal 67.5%.
The resulting reduction in ground-level concentration
would be only about 50%. In contrast, scrubbing all the gas
but with no reheat would give a concentration reduction of
82%. Hence bypassing of 3100 F. gas does not appear to be
an acceptable reheating method. Hotter gas from ahead of
the preheater could be used but taking heat from this source
would incur a fuel cost and would be equivalent to direct
Table 4. Maximum Groundline Concentrations of Sulfur
Qlo~l~~~jJ~~~~~l~~~Q~e~_____-------------
Plume rise S02 concentration, ppm, at
Stack gas above top of indicated % of S02 removed
.!.e.!Jl"'p-,,~~t -_st-Ef.~l.L- 112- _~IL -~~- _8JL -Q.~
310 402 0.008 0.016 0.024 0.032 0.16
250c 363 0.010 0.020 0.030 0.040 0.20
225 345 0.011 0.021 0.032 0.042 0.21
200 341 0.011 0.022 0.033 0.044 0.22
175 318 0.012 0.024 0.036 0.048 0.24
150 292 0.013 0.026 0.039 0.052 0.26
__Jl~~____---~~-----~Q.~_]~I~J~~~_~~~~~J~
aBased on plant data from table 3. In addition: total S02 in gas,
126 tons/day; S02 in gas, about 0.3%; approximately 90% of the
b sulfur in coal leaving the system in the stack gas. 0
At exit of last operating unit before stack; further loss of 13.5 F.
in ducts and stack.
~Highest practical reheat temperature.
Without reheat.
24
heating of the gas (see later cost estimate on this);
moreover, ducting would be a problem and the thermal
balance in existing plants would be affected.
Calculations were made for other stack heights also. For
normal power plant operation, cutting the height from 300
to 150 feet increased maximum ground concentration by
about 75%. At 80% removal and without reheat (stack
temperature of 1250 F.), the increase was 100%.
Conversely, if the stack height were 600 feet instead of
300, the influence of gas temperature would diminish;
ground-level concentration would be increased only about
o
45% by releasing the gas at 125 rather than 310 F., as
compared with 85% increase with the 300-foot stack. At
lower wind speeds, which are generally more critical in
regard to ground fumigation from high stacks, release of gas
at 1250 F. would be even less significant; the increase in
o
ground concentration over release at 310 F. would be only
about 25%. Thus as stack height is increased above the 300
feet assumed, reheat becomes less important.
It is concluded that the gas should be reheated after
scrubbing but that the degree of reheat desirable may vary
widely depending on the particular situation. As ncted
earlier, it is assumed in this study that reheating beyond
2500 F. would not be attempted because of the relatively
high cost for going to higher levels. The minimum
acceptable temperature probably is that required to avoid
presence of mist in the gas as it leaves the stack (see next
section), which would vary with power plant duct and stack
design and with efficiency of the mist eliminator. Between
these two levels the appropriate degree of reheat probably
will vary depending on local factors and on degree of sulfur
dioxide removal for which the unit is designed. Relatively
low sulfur dioxide removal should be coupled with
relatively high degree of reheat to keep ground
concentration at a reasonably low level; for example, 70%
removal would require reheat to about 2l5°F. to avoid
ground concentration reduction from falling below 60%.
Effect of Evaporative Cooling: In the scrubber, hot
unsaturated gas is intimately contacted with the scrubbing
medium to promote rapid sulfur dioxide absorption; as a
result water is evaporated from the liquid and the gas
becomes saturated. After saturation, further loss of heat
may uccur by contact with cooler surfaces, in which case
water vapor condenses in the form of mist droplets. The
amount of condensate formed is a function of the heat loss
from the gas after saturation. In addition, some droplets
will be present because the mist eliminator does not remove
all of the mist formed in the scrubber. Also, condensation
will occur in the stack when saturated gas is cooled by heat
loss from the stack.
After the gas leaves the stack, further heat loss to the
ambient air takes place and more condensation occurs.
Then as the plume rises higher and, in many cases, is
-------
-40
-20
20
Reduction in maximum ground-level concentration, %
o
250
225
60
8
Ll..
2000
(I,'
...
:J
....
co
...
Q)
a.
E
Q)
....
....
H!
175 {i
a::
7
8
150
40
125
100
60
80
Figure 3. Effect of reheat temperature on ground-level
802 concentration at various scrubber efficiencies
(Base: Maximum ground-level concentration at
3100 stack temperature and without scrubber)
subjected to wind energy, mixing with ambient,
unsaturated air occurs and the dilution soon gives an
unsaturated mixture, still within the confines of the plume.
The resulting evaporation, which continues until all the
droplets are evaporated and the plume becomes clear, cools
the plume-possibly to a temperature below that of the
surrounding air. If this develops the plume will have a
"negative buoyancy" and may sink to the ground faster
than if the evaporative cooling had not taken place.
The situation is different from that in normal power
plant operation, where the stack gas is quite unsaturated at
the stack exit temperature. The dew point is so far below
the stack temperature that dilution by unsaturated ambient
air drops the dew point further, to below ambient
temperature, before condensation can occur.
Scorer (17, 18) has treated the effects of heat of
condensation in offsetting the loss of buoyancy resulting
from evaporative cooling. For condensation occurring after
the gas leaves the stack, it is obvious that the amount of
heat evolved during condensation is equal to that absorbed
during evaporation. Hence the effects cancel each other and
the fmal temperature should be the same as for the usual
25
-------
situation-an unsaturated gas leaving the stack. However,
the heat of condensation makes the gas hotter than it
otherwise would be and therefore it rises higher than would
an unsaturated gas. Thus any downward movement
resulting from evaporative cooling would begin at a point
higher than unsaturated gas would reach. The faster rate of
cooling presumably would bring the plume down faster so
it is difficult to say what the net effect would be. It seems
possible, however, as pointed out by Scorer, that a
saturated plume would actually disperse better because of
latent heat effects.
The situation in wet scrubbing is complicated by the fact
that some of the condensation occurs before the gas leaves
the stack. Since the resulting mist does not contribute any
latent heat to the gas after it leaves the stack, its heat of
evaporation all goes to cooling of the plume. As a result the
ultimate plume temperature should be lower than for an
unsaturated plume leaving the stack at the same
temperature. Moreover, the mist increases bulk density and
hence reduces buoyancy.
Scorer treats this problem in terms of stack exit
temperature as compared with wet-bulb temperature, which
is a measure of the degree of condensation because any
difference between these temperatures is due to heat loss
through duct and stack walls and this cooling is the cause of
the condensation. Scorer developed a nomogram (Fig. 4)
for quantitative estimation of the cooling of a wet plume
by evaporation of water condensed before emergence-at
various emission temperatures. Plotted isopleths show the
degree of cooling by evaporation and can also be used to
determine the plume temperature after dilution. For
example, with lOa-fold dilution and a gas wet-bulb
temperature of 125° F., a 100F. temperature drop in the
saturated gas before emission would result in about a
1.3 ° F. decrease in plume temperature; with less dilution,
say la-fold, the plume would be cooled 13° F. Based on
results of plume dispersion studies, the lOa-fold diluticm is
probably representative of typical conditions and the effect
of negative buoyancy would be insignificant.
However, mist escaping through the mist eliminator
would add to the problem, particularly since it does not
contribute any heat of condensation at all. Mist from the
scrubber may well have been the cause for the situation in
the tests at the Bankside station in London, where,
according to Scorer, the plume could "almost invariably be
seen to obscure the chimney tops and usually reached the
ground within a few hundred yards when ambient
turbulence was not large."
Scorer's study is one of very few on the subject.
Inquiries to leading meteorologists in the air pollution field
indicate that the effect of evaporative cooling on plume rise
has received only minimal attention to date. However, with
the advent of large, natural-draft cooling towers, it is likely
that the subject will receive increasing attention.
26
It is concluded that evaporative cooling may have an
adverse effect on dispersion if there is any mist in the gas
when it leaves the stack. Present information, however,
does not allow a quantitative treatment of the resulting
plume cooling. Any evaporative cooling is likely to make
the formulas in the preceding section inaccurate, in the
direction of predicting lower ground-level concentrations
than would actually occur.
A further conclusion is that the gas shrnld be reheated
at least enough to ensure absence of mist at the point of
emission from the stack. Heating to a higher temperature
would not help in regard to the evaporative cooling effect,
but would, of course, increase plume buoyancy as discussed
in the previous section.
Intermittent Operation
In the dry process study it was assumed that limestone
injection could be used by power plants in two
ways-continuously where local conditions dictate and
occasionally where the main need is to avoid high ground
concentrations that develop only when atmospheric
conditions are adverse to good dispersion from the stacks.
Since the latter is a very good way to reduce the cost of
sulfur oxide control, amenability to intermittent operation
is a highly desirable characteristic of any control process.
Operating flexibility is somewhat limited in the
limestone-wet scrubbing process because the scrubber also
serves as a dust collector. In new plants the scrubber would
have to be operated whenever the power plant was on line
because there would be no other way to remove dust. It
would be possible, of course, to achieve intermittent sulfur
oxide removal by the same means as in the dry
process-shutting off the feed of limestone to the boiler.
The scrubber would continue to remove dust with water as
the scrubbing agent. As noted earlier, however, the scrubber
liquor would quickly become acidic; as a result the scrubber
and waste-disposal system would have to be built of
corrosion-resistant material. Moreover, any mist escaping
from the mist eliminator could corrode parts of the power
plant after the scrubber, including the stack.
In existing plants already equipped with electrostatic
precipitators, intermittent sulfur oxide removal could be
accomplished by (1) connecting the precipitator in parallel
with the scrubber and shifting the gas flow to the
precipitator when sulfur oxide scrubbing was not needed or
(2) connecting the precipitator and scrubber in series and
shutting off the scrubber liquor flow to stop sulfur oxide
removal.
The main drawback to parallel operation is difficulty in
operating and maintaining the very large shutoff dampers
required and the fact that expected leakage through the
dampers would probably cause troubles in the equipment
not in service at the time. Leakage of gas through the
-------
u..
o
ai
...
::J
....
CO
...
Q)
a.
E
Q)
....
jJ
::J
jJ 1 00
....
Q)
s
50
Emission temperature, 0 F.
100
150
200
200
Decrease in plume temperature: *
28000 F.
13500 F.
150
7500 F.
1250 F.
75° F.
50
*Evaporation of the amount of condensed
moisture indicated by the difference be-
tween the wet bulb and emission tempera-
tures would reduce the plume temperature
by the degree shown if no dilutiol'l occurred.
Figure 4. Cooling of plume by evaporation
of water condensed before emission (17)
27
-------
damper into the precipitator could have serious
consequences. The precipitator would be at or near ambient
temperature when out of service and the gas leaking into it
would be expected therefore to cool below the dew point
of sulfuric acid. Dust leaking through would accumulate
and possibly form sticky and corrosive deposits with the
sulfuric acid. Also, a temperature differential might develop
across the precipitator that could cause warpage of
electrical collectors.
It should be noted that use of dampers is included in the
conceptual design presented later, in which their function is
to close off one of the two scrubbers at low power plant
load (a single scrubber would be too large to operate
satisfactorily at as Iowa load as sometimes required of
power boilers). The dampers are included because there
appears to be no alternative, but precipitator corrosion and
caking is not involved. Gas leakage into the scrubbers may
not be a major problem since they are equipped with liquid
feed facilities and therefore could be flushed occasionally
with water or alkaline solution during nonoperating
periods.
Connecting the precipitator and scrubber in series
appears to be a much more practical arrangement for
intermittent sulfur removal in an existing plant. Problems
with putting the precipitator in and out of service would
thus be avoided and there would be no dampers to give
trouble. To shut down sulfur removal, the scrubber liquor
28
would be cut off and water introduced to flush out all salt
residue. The scrubber would then dry out and ride on the
line until sulfur oxide removal was required again.
Possible drawbacks to this include accumulation of
residual dust from the precipitator in the packing and on
surfaces. Perhaps an occasional flushing with water would
prevent any trouble from such accumulation. Another
potential problem would be the possibility of damage
to temperature-sensitive parts of the scrubber, e.g.,
polypropylene sphere packing or scrubber shell coating, by
the hotter gas. If this were found to be a major problem the
gas could be bypassed around the scrubbers to the stack.
Such an arrangement would bring back the damper problem
but perhaps it could be handled as outlined above. -
It is concluded that the limestone-wet scrubbing
process could be operated intermittently if several
anticipated problems were resolved. There is little doubt
that these problems can be worked out; a program of
engineering development very likely would provide suitable
solutions. However, the overall plant investment would be
increased in fitting the plant for intermittent operation and
there would be some increase in overall operating cost that
would tend to offset the saving in limestone. These extra
costs could so cut into the advantage of intermittent
operation that continuous operation would be preferable,
with the added benefit of assured pollution control at all
times.
-------
Study Assumptions and Design Criteria
The conceptual design in this study has been developed
for two basic processes:
Process A. Inj ection of pulverized limestone into a boiler
with subsequent scrubbing of the flue gases
to remove both fly ash and sulfur dioxide.
Process B. Introduction of lime 01 pulverized limestone
directly into the scrubber recirculating slurry
to remove sulfur dioxide; fly ash will also be
removed in the scrubber.
The effect of dust removal prior to scrubbing through
use of electrostatic precipitators is also evaluated for both
processes. These four arrangements should cover most
potential applications of limestone-wet scrubbing. Each
process is compared under uniform conditions and in
addition, the effects of variables which significantly affect
costs are presented.
Plant Location
As a basis for development of cost data, the TV A
Colbert power plant in northwestern Alabama (16 mi. west
of Muscle Shoals) was selected. This plant is located on the
Tennessee River, has an overall capacity of 1300 megawatts
(one 500-mw. and four 200-mw. units), and has
coal-burning boilers of the pulverized-fuel type. The same
plant was used as the cost basis in the dry limestone
injection study.
The Colbert location should serve adequately as a
general case. The main variables in plant location are
distance from limestone supply and availability of space for
reacted limestone disposal. In these respects the Colbert
plant is well located; limestone is available from operating
quarries about 10 miles away, and disposal space is not a
major problem because the plant is located in a farming
area where additional space is available and probably could
be purchased at reasonable cost. The same situation appears
to hold for many of the power plants in the United States.
The main exceptions are plants located in or near urban
centers, where quarry operation may be feasible only some
distance away and disposal space may command a premium
or even be unavailable.
To cover plant locations not as favorably located as the
one selected, the effect of freight cost on limestone
delivered price was reviewed; the results are presented in
the cost estimate section. Costs of disposal also were
evaluated.
Plant Size
A 200-megawatt unit size was selected as the base case
for the study. From the chart of power plant sizes in Figure
5, it can be seen that this is a typical midrange boiler size.
Estimates were also made for 500- and lOOO-megawatt
boilers.
Fuel Type
Coal was selected as the fuel because it is the prevalent
type. Many plants burn high-sulfur oil, however, and
limestone-wet scrubbing would be an appropriate method
for sulfur dioxide control. The main difference is that since
oil produces very little ash, oil-burning plants have limited
facilities for handling solids. Equipment for solids disposal
would be new for an oil-fired plant instead of an addition
to existing facilities for coal firing and therefore
procurement of land for pond construction might be a
problem.
Sulfur Content of Coal
The coal deposits in use vary widely in sulfur content as
shown in Table 5. Three levels, 2%, 3.5%, and 5%, were
selected for the study. It seems likely that coals of
relatively low sulfur content will be consumed rapidly in
the next few years and that the higher levels will become
prevalent. The 5% level was selected as being typical of
coals which might be used in processes designed for
recovery of sulfur in marketable form and was included as a
basis of comparison for future evaluation of recovery
processes. The 2 and 3.5% levels appear to bracket most of
the coals currently in use.
Limestone Type
The limestone selected is from a quarry near the Colbert
plant. Price was the main consideration; because of the very
low cost at the quarry, freight is a major part of the cost
delivered to the power plant.
The limestone is a high-calcium type; a typical analysis is
given in Table 6. The proved reserve from which it is being
quarried is large, estimated at 8 million tons of about 95%
calcium plus magnesium carbonate material. Such a reserve
would last some 20 to 30 years in supplying the needs of a
BOO-megawatt plant such as Colbert; other undeveloped
deposits in the area probably would extend the supply
indefinitely.
29
-------
150
125
Age of unit: (includes nuclear units)
II>
....
....
co
:f:
.2
.:>t.
c:
g 100
E
+- Less than 10 years
+- More than 10 years
ij
c:
co
.c
Q>
N
.en
.s:
..c:
....
.~
75
II>
....
.c
::J
'0
>
.'=
:0
co
a.
co
U
50
25
/j) co /j) /j) U"I /j) /j) /j) /j) a a /j) /j) /j) /j) a a a a
a /j) /j) """ /j) /j) /j) a> a a a> a> a> a> a a a a
a> M o::t .- M U"I ...... a> a .- M U"I """ a> a N U"I
.- N .- I I I I I
Megawatts -? I I I I I I I I I I .-
a a a a a a a a a a a a a
a a a a a a
a a a a a a a
N N o::t N o::t (0 co N o::t (0 co
Year L 1950-.J L 1960 --.-J 1970 1980
-?
Figure 5. Distribution of thermal power plants
in the United States (actual and projected)
Source: Power System Statements to FPC for 1950 and 1960
For process B, the cost of calcining the limestone before
introduction into the scrubber circuit was estimated as an
alternate to use of raw limestone.
Dolomite or dolomitic limestone may be a more
economical choice in other parts of the country. The main
difference is that pond water recirculation may be required
when dolomite is used but not when high-calcium limestone
is the absorbent. The cost of recycling was estimated.
Amount of Limestone Used
F or the base cases, 11 0% of the stoichiometric
requirement of limestone was assumed for process A
(injection-scrubbing) and 130% for process B (scrubber
30
addition). These levels are based on the experience of
others as discussed earlier in this report. The requirement
for process A is fairly well identified. The amount needed
for process B is less certain and should be investigated
further in experimental studies. For use of lime in process
B, 100% of stoichiometric was assumed-in keeping with
the British work.
Particle Size
For both processes, grinding to 70% minus 200 mesh
was assumed. For injection into the boiler, this level is
consistent with the experience reported and permits use of
conventional coal mills. For injection into the scrubber, wet
-------
grinding was assumed as it has the advantage of eliminating
the heat requirement for drying (needed with high-moisture
feed to a dry-grinding operation). No other sizes were
assumed as the effect of particle size on limestone
utilization is not well defined. It may be that future
experimental work will show that smaller particle size
improves absorption. The cost of fine grinding by dry
methods was discussed in the report of the dry limestone
process and it is planned to evaluate wet grinding to small
particle size in further studies.
Boiler Type
The boiler on which the study is based-one of the
200-megawatt units at the TV A Colbert Steam Plant (Fig.
6)-is of the pulverized-fuel type, that is, the coal is finely
ground and injected through burners into the boiler. This is
the most common type. The cyclone type differs in that
the coal is not finely ground and is burned in cyclones
(furnaces that discharge heated gas into the boiler proper
through ports) set into the side of the boiler. Most of the
ash in the cyclone type leaves the boiler as bottom slag,
whereas most of it leaves as fly ash in the pulverized-fuel
type. Thus there would be a lower total load of solids in the
!
-------
'~al~~~~
~~~'-f-
1~'.:~
~~}- "l~ ~' ~
tun plm
,.... ~
~. :,~
, \.r::,:~
I , -'.
I ".,
1:::/: M
r- ~~.~
I........ \ / -' '.....,..
\ ~ . "
: ~ /1 j''<, ..
;~ ~WI ._~-'
~::. .:. 1\ II ~ J f-
,:,1. '= ~ ~ ii \ ( I ~
T
.1
, ..." I
-- I
" I
I
,
,
:f
';I
.j,'
-
'.
;~ I d
r.~ : : : ",
",
,',
~
..:~~
."
,j:
,,'
'"
'"
'"
'"
II ~ :::"":':
/ ' :
/,,' :::
..'0' ::: ~
/
~/
/ -~-
/
T
, ,
rJ
. .
-
c
/
32
B
f:r
8 ~
J.
~
~~
.~
r UoI
- EL.~!L Q -f-DRllJrJ -
I~
.~.
.~
:~.
,.
o
,
o
~)
~~
o
~
i
I
, I
'~
~ l\ 1[:
1\\ \ II
f-
: I
1>-
'.
: ":111 I ;
- ":::!'liill,1111Iillli!~:;ill
~~;"I'''oIL'II:I'1
Ir'::~I"IIt~~I';
.- - .'1' I [II:';: II
r-- I': '.,
':ii:
Iii
I~
~
, .......
'"'
1
I Ir-
:1:11
"I l-
I, -
Ii
, "---L.=:
I
11
I
:1
,
1\
~.~--- -- ~--- ,~-
I [~,-l
~~
.' ,j ,.
1
- - --
. ,
1 ~ -.., 1
Hi
1
~
'---
"-
'---
II
-
\
0;
~ -
~ '"
~ I -
n ~I~~ ~ ~<'''? ~
. ,
.....
,
Figure 6. 200-megawatt unit at TV A Colbert steam plant
-------
equipment because of the known high cost of the latter.
Alternates evaluated were (1) pumping the total purge
stream (solids content of about 10%) to the pond and (2)
concentrating the solids to a 40% slurry in a thickener
before pumping. The liquid phase would be discharged to
streams, the current practice in power plants for ash sluice
water. Since in some situations, particularly when dolomite
or dolomitic limestone is used, recycling of sluice water
may be required, the cost of recycling was estimated also.
Procurement of additional space for solids storage during
10 years of operation was assumed.
Stack Gas Reheating
. 0
Reheatlllg of the cooled stack gas to 250 F. was
assumed. This level is the maximum practical for use of the
waste heat in the hot gas entering the scrubber, and would
approach the stack temperature from existing units
equipped with low-level economizers. Costs for lower levels
of reheat were also estimated.
The method of reheating was selected on the basis of
cost estimates on five different methods. The most
economical was use of heat exchangers to extract heat from
the stack gas ahead of the scrubber and to transfer heat to
the scrubber exhaust gas.
Stack Height
Removal of sulfur dioxide by wet scrubbing will greatly
reduce the potential hazard of localized fumigation. It can
be argued that removal of most of the sulfur oxide
elminates the need for a high stack, particularly in view of
the current trend to very expensive 800- to 1200-foot
stacks for new large power plants. However, discussions
with air quality specialists indicate that there will be
considerable opposition to any major stack height
reduction merely because one of the pollutants has been
removed to a major extent. Other objectionable
components, mainly nitrogen oxides and residual dust, are
removed only to a limited degree. Moreover, during upset
operating conditions or scrubber unit breakdown it would
be desirable to have the high stack available for good
dispersion. Therefore, no credit was taken for reduction in
stack height in the basic cost estimate. As a matter of
interest, stack costs are presented so that the effect on
overall cost of reducing stack height can be visualized.
33
-------
Equipment Selection and
Equipment required for the limestone-wet scrubbing
process can be divided into three major categories:
1. Limestone receiving, storage, and handling.
2. Gas scrubbing and reheat.
3. Solids disposal.
The major alternatives and type of equipment selected to
carry out category I were discussed in detail in the report
of the dry process. Equipment in this category for the
present conceptual design is described later in this section.
There are several alternative choices in selection of
equipment in categories 2 and 3.
Major Alternatives
Scrubber Type: A study of scrubber types was
made, on the basis of information from plant
visits, scrubber consultants, scrubber vendors, and
literature sources.
The scrubber type used in the Howden-ICI pilot
plant program and later at the Fulham plant was a
countercurrent, grid-packed tower designed for
low gas velocity (5 ft./sec.) and high-volume liquid
recirculation. The pressure drop in the scrubber
was only about I inch of water. High efficiency
was obtained; when the system was operated at a
pH of about 6.3, about 98% of the sulfur dioxide
and dust in the inlet gas were removed.
Pilot plant studies at Wisconsin Electric were
carried out in a 14-inch diameter, Turbulent
Contact Absorber (TCA; moving bed of hollow
plastic sphere) designed and manufactured by
Universal Oil Products Company, Air Correction
Division (Fig. 7). The majority of the sulfur
dioxide absorption tests were made with two
stages of scrubbing. The effect of liquid and gas
rates on absorption efficIency with pulverized
limestone (90% -200 mesh) used as the absorbent
is shown in Figure 8.
Data on particulate collection efficiency in the
Wisconsin Electric studies are given in Figure 9.
With two-stage scrubbing and a high ratio of liquid
to gas, more than 99% of the dust was removed;
inlet dust loadings were in the range of 1.5 to 2.5
grains per cubic foot. Pressure drop with two
stages of scrubbing was about 6 inches of water.
The scrubber design used in the Combustion
Engineering - Detroit Edison work was also a
mobile-bed type (Hydro-Filter, manufactured by
34
Description
National Dust Collector Corporation). Instead of hollow
spheres, glass marbles were used in the bed. A single-stage
bed gave 6-inch pressure drop and removal of about 98% of
the sulfur dioxide and 99.5% of the dust.
Gas outlet
--
----
--
.::::::
....
......
.......
......
.........
..........
.........
.......... "'-.
~@P ......... ~ ~
Liquid ~ ~-~- -
.... inlet ~
~
......
~
~
::::::
t
Gas inlet
Figure 7. Typical turbulent
contact absorber scrubber
-------
90
*
>.
(,J
c
Q.)
'u
:E
Q.)
15 80
'...
Co
.....
o
-------
Thus only two types of scrubbers have been tested and
both gave good dust and sulfur oxide removal efficiencies.
The wooden grids used in the Howden-ICI work are seldom
used in scrubbers today. The mobile-bed type is a fairly
recent development, particularly applicable to scrubbing
with a slurry because of its self-cleaning action.
Other types of scrubbers also should be applicable. In
evaluating them, the following characteristics may be set
down as desirable:
1. Low pressure drop (6 in. of water or less)
2. High efficiency for removal of sulfur dioxide
(85-95%) and particulates (minimum 99.0%).
3. Turbulent mixing of the liquid phase and
suspended solids to promote dissolution of calcium
oxide or calcium carbonate.
4. Resistance to plugging by scaling or by settled
solids from the circulating slurry.
5. Adequate turndown ratio (ratio of design volume
rate to minimum volume rate at which scrubber will
operate satisfactorily) to allow operation during reduced
boiler load.
The following types of scrubbers have been considered
in this study:
1. Packed8
a. Countercurrent flow
b. Co current flow
c. Crossflow
2. Sieve tray
a. Countercurrent without downcomers
b. Crossflow impingement plate with downcomers
3. Venturi
4. Spray
a. Cyclonic
b. Spray tower
5. Mobile-bed
6. Orifice
The factors that exert the most influence on scrubber
design are inlet dust loading, gas volume, pressure drop, and
volumetric turndown.
With limestone injection in the boiler and no dust
collection ahead of the scrubber, the dust loading will be
about 7 to 9 grains per cubic foot. If a dust collector is used
upstream of the scrubber the inlet dust loading is reduced
to about I grain per cubic foot or less.
In most cases, scrubber dust-removal efficiency is stated
on a weight basis but the percentage of particles removed is
likely to be lower. Most scrubbers, regardless of type, will
effectively remove particles lO microns in diameter or
larger but removal of smaller particles is necessary for good
8packed with modern type of packing. The wood slat packing used
in the Howden-ICI work was not considered because the surface
area is low and therefore very large scrubbers would be required.
36
plume appearance since about 30% of a typical fly ash is
less than 10 microns in size. A 100-gram sample of this ash
would have approximately the following particle
distribu tion:
Particle size Weight in fly Number of
micronsa --~~.!-\L_- _2~'!!E!!S_-
--------
0.5 0.13 8.1 x 1011
1 0.37 2.9 x 1011
2 1.5 1.5 x 1011
3 2.5 7.4x1010
4 3.5 4.3 x 1010
5 1.5 9.5 x 109
6 3.5 12.9 x 109
7 3.5 8.1 x 109
8 4.5 7.0 x 109
9 4.0 4.4 X 109
10 5.0 4.0 X 109
20 20.0 2.0 X 109
30 12.0 3.5 x 108
40 8.0 1.0 X 108
50 8.0 5.0 X 107
70 9.0 1.7 X 107
100 3.0 2.4 X 106
+100b 10.0 2.9 x 106
----------------------------------------
aAssumes particles smaller than 0.5 micron are 0.5 micron in
diameter; particles smaller than 1 micron and larger than 0.5
micron are 1 micron in diameter, etc., to 100 microns.
b Assumes all +100-micron particles to be 150 microns in size.
The main question in regard to this is whether a scrubber
that removes the required 99% by weight will give adequate
performance on the basis of number of particles removed.
The difficulty is that no data of this type are available for
wet scrubbing of the fly ash-lime-calcium sulfate mixture
involved. Information from scrubber manufacturers
indicates that various types will remove over 99% of the
dust (by weight) at about 6 inches pressure drop, and from
70 to 90% of the particles (based on the distribution given
above). Operation at higher pressure drop would give better
removal of small particles (Fig. 10) but the cost would
increase quite rapidly at high efficiencies.
High weight removal efficiency was demonstrated in the
Combustion Engineering tests with a mobile-bed scrubber,
in which 99.5% removal was accomplished at 6 inches
pressure drop. No data are available on removal of the finer
particles. However, removal higher than 99% would
ordinarily be regarded as quite good without reference to
number of particles. Therefore, it has been assumed that
the more efficient scrubbers are acceptable at 6-inch
pressure drop.
The volume of gas to be scrubbed is quite large; stack gas
rates for 200-, 500-, and 1000-megawatt units are 545,000,
1,362,000, and 2,725,000 actual cubic feet per minute,
-------
100
95
~
>-
u
~ 90
'0
i
UJ
85
80
4
8
20
12
16
t::. P, inches of water
Figure 10. Collection efficiency for 1-micron
particles (impingement-type scrubber)
respectively, at 130°F., the estimated scrubber exhaust
temperature. To simplify gas distribution, the number of
scrubbers per boiler should be as small as possible; it is
assumed that not more than four would be used. Only a
few manufacturers are presently capable of supplying
scrubbers as large as this would require. (There is a need for
pushing scrubber technology to larger sizes.) The turbulent
bed scrubbers are the only type offered at this time in sizes
larger than 500,000 cubic feet per minute. One
manufacturer of the crossflow impingement type can build
units to handle flows up to 500,000 cubic feet per minute.
Table 7 shows scrubber combinations for the three
power plant sizes. The turndown ratio for the scrubbers is
based on operation of the power plant at reduced load, as
low as 30% of design; a typical 200-megawatt unit has six
Table 7. Com...p.arison of Scrubber Combinations -
--------- ---------------VOiumetnc-------
Design Gas flow turndown ratio Approx.
Plant Assumed capacity/ @ max. required for number of
size number of scrubber turndown ea scrubber scrubb1f
-~~- ~q~l!~~ J~J~fl.L -~g~~- j!!.QI!!!!.a.1i.Q!!. _f!!.3J:1.!!L-
200 1 545,000 164,000 3.33 2
200 2 272,500 164,000 1.67c 4
200 4 136,250 164,000 1.67c 6
500 2 681,250 410,000 1.67 2
500 4 340,625 410,000 1.67c 3
l~Q~___~--_J~lJ~~-~~~@]____l~I~_____-~--
aAssuming scrubber system would be operated at a minimum
capacity of 30% of design.
bPresently known manufacturers for scrubbers of the sizes required.
cTurndown ratio required for each scrubber when using only
one-half of the scrubbers at minimum system flow.
37
-------
coal pulverizers and can be operated with as few as two of
them, giving a load of about 60 megawatts.
A comparison of different scrubber types is given in
Table 8, showing the indicated effectiveness of each for
operation with or without dust collection ahead of the
scrubber and with limestone either injected into the boiler
or added directly into the scrubber circuit. For operation
with low dust loading, packed scrubbers (both crossflow
and cocurrent), cyclonic spray, and spray towers would be
suitable. For use without dust removal prior to scrubbing,
the mobile-bed type provides the best combination of
ability to handle high dust loading, small particle removal
with low pressure drop, and turbulent action which
promotes dissolution of the absorbent and also prevents
plugging. Either a two-stage TeA scrubber or a one-stage
Hydro-Filter scrubber operating at about 6 inches of water
pressure drop should remove 99% of the dust and 95% of
the sulfur dioxide when limestone is injected into the
boiler. Addition of limestone in the scrubber is assumed to
reduce the sulfur dioxide removal efficiency to 85%;
operation with a pressure drop of only 3 inches of water
would further reduce the removal to 70% and dust.
collection efficiency to 95%.
The mobile-bed type was selected for the economic
studies. Use of two two-stage TeA units designed for 6
inches of water pressure drop was assumed for both
processes A and B. Hydro-Filter scrubbers should also be
acceptable; however, in the Hydro-Filter the scrubbing
liquor enters at the bottom through upward sprays and is
carried up through the marble bed by the momentum of
the gas after which it flows back through downcomers
through the bed and to the scrubber outlet. In contrast, the
liquor enters at the top in the TeA scrubber and operation
is fully countercurrent. This should give lower retention
time and therefore possibly less scaling in the scrubber.
Another variation that might be considered is use of a
two-stage scrubbing system as a means of avoiding
introduction of the lime directly into the scrubber (in
process A) and thereby reducing scaling. The first stage
would be designed primarily to remove the solids, which
would then be introduced into the recirculating liquor
going to the second stage. For example, a venturi scrubber
!~~~~~~E~E~~~~~~~~Q~--------------------------------------------------------------
Approx. max. Approx. dust-removal Resistance
inlet gas efficienciesa Expected pressure 0 issolution to
Usual dust loading fmlCrOri-2-=-rTiTcron-5":ffiTcron drop for 95% S02 of solids
~"'pJj~!i.Q.'!. _9!!!!'!.s1f.!:~- .I?~!l~~ .I?~!l~~ l!~r!i~~ !.e.!'l.il~,!!,j~_W~!!!..r:. -~~!a..!!~- _1!.!~gJ!~~b-
_~c..r:.~E~JY.PJ!-
Countercurrent
packedd
Co current
packedd
Crossflow
packe~
Countercurrent
sieve
Crossflow
38
Approx.
turndown
ratio c
------
GA 1.2 0 0 95 8 G NR 1.2
GA 2 0 0 95 6 G UC 5.0
GA 3-4 0 0 95 4 G UC 5.0
GA & PR 10 85 92 95 6 G UC 1.2
GA & PR
GA & PRa
100
+100
93
97
96
99
sieve
SF venturi
Splash plate
venturi
Cyclonic
spray
Spray tower
Mobile bed,
Hydro-fi Iter
Mobile bed,
TCA GA & PR 25 92 98 99+ 6-7 E Ee 1.5
QdfL~_________E!!________25 35 75 93 6-8 NR E 1.2
~When operating at pressure drop requir~dfo~95%~ulf-;;di~~de~~~~v~~---------------------------------------
For a system in which calcium compounds are precipitating.
CFatio of design volume rate to minimum allowable volume rate.
dThese scrubbers will not remove particles smaller than 3 microns unless nucleation is effected by condensing water vapor.
;Provided spray nozzles do not plug.
For hi£hly soluble gases only. --------------------------------
Key:E-excellent; C=gocid;F-=-faiI;1,fR -=- not recommended; G"A=-gas absorption; PR = particulate removal; Y -yes; N -no; UC- uncertain.
GA & PRa
+100
97
99
GA & PRa
GA & PRf
15-20
15
30
20
70
60
GA & PR
25
92
98
98-99
99+
6
12
G
NR
NR
E
1.2
1.3
99+
12
E
2.0
UC
94
90
Ee
Ee
2.0
4.0
4-6
3-5
F
F
99+
Ee
1.5
E
6%
-------
might be used for the fIrst stage; it would have good
dust-collection efficiency but would not be expected to
dissolve the lime very well, which would minimize calcium
salt precipitation in the unit. The collected ash-lime
mixture would then be introduced into the slurry
circulating through a second scrubber designed to dissolve
the absorbent and with adequate retention time to absorb
the sulfur oxides. This unit could be of the spray tower of
mobile-bed type.
No data are available for evaluating this arrangement.
Mist Eliminators: The removal of liquid entrained in the
scrubber exit gas serves three purposes:
1. Reduces the load on the stack gas reheater.
2. Decreases the deposition of liquid in the fan and in
ducts located downstream from the scrubber.
3. Reduces amount of solids discharge by removing
the dissolved solids contained in the mist which, after
reheating, would be emitted to the atmosphere as dust.
The mist emitted by a wet scrubber would be comprised of
liquid particles ranging from about 10 microns to 60
microns in diameter; particles smaller than 10 microns
would not normally be produced unless water were
condensed from the flue gas. With no mist removal from
the TCA scrubber exhaust, the gas would contain about
0.012 pound of entrained liquid per cubic foot of gas. Use
of a 95% efficient demister would reduce the heat required
for reheat by about 3 million B.t.u. per hour or about 4%
of the total required for reheat to 250°F.
The mist eliminator should have the following
characteristics:
1. A removal efficiency of at least 70% for lO-micron
liquid particles and an overall effIciency of about 95%;
above this the high capital cost is not justified by the
advantages.
2. The ability to flush (with the collected mist) any
undissolved solids collected in the separator.
Several types of mist eliminators were evaluated. The
TCA and Hydro-Filter scrubbers both are equipped with an
impingement vane type of mist eliminator. The shape of the
vanes and their arrangement cause impaction and
coalescence of the mist. Other types of entrainment
separators evaluated were the swirl vane, wire mesh (York),
fiber bed (Brink), cyclonic, and packed bed (6 to 12 in. of
Tellerettes, Pall Rings, or others).
A comparison of the mist eliminators is shown in Table
9. From this it appears that the vane types offer the lowest
investment and best plugging resistance at acceptable
pressure drops. The overall removal efficiency of the vane
types should be about 95%. The Brink type of mist
eliminator is more efficient than justified for this
application. The wire mesh and packed-bed entrainment
separators could be used but should be equipped with a
means for flushing with fresh water to remove solids
collected on the mesh or packing. The cyclonic type (swirl
chamber or tangential scrubber outlet) would probably not
give an overall efficiency of 95%.
An impingement vane mist eliminator was selected for
use with the TCA scrubber in the conceptual design.
Equipment for Reheating Gas: Alternate methods
studied for reheat of the stack gas were:
1. Install a combustion system at the base of the
power plant stack for burning natural gas, oil, or coal
and mix the combustion products with the scrubber exit
gas.
The main advantages for this method are low
investment, flexibility in degree of reheat, minimum
added pressure drop, low maintenance, and good
reliability. Disadvantages are fuel cost, introduction of
objectionable components (SO 3, SO 2, ash) into the gas,
an.d fuel supply problems. Natural gas is the most
expensive fuel but the added cost might be justified on
the basis of convenience and clean combustion.
However, gas would not be available at many power
plants. Oil would be less expensive but receiving, storing,
and handling would be a problem. Use of coal would
result in the lowest fuel cost but would add more sulfur
!~~~~~~~~~d~o~_~~~nm~~~___---------------------------------------------------------
Removal efficiency Pressure Approx. cost Resistance
for 10-micron drop, in. for 200-mw to solids
---_.Jj'Q~----- -~~tP~!!~~~1'!...- ~.!~!~~- ln~~!!.ali.Q!!JJ- ...pl~gj.!!!L
Impingement vane 70 0.4 3,500 Good
Swirl vane 70 0.3 3,000 Good
Wire mesh (York)b 98-99 0.8 12,000 Fair
Fiberbed(Brink) 100 7.0 75,000 Poor
Packed bedc 90 0.1-0.2 9,000 Fair
~~~~~---------------____[~L~_____---------~JJ~~--------------~~@-------_._----~~o~---
aRcquired for removal efficiency shown for lo-micron particles.
bSix-inch-thick bed.
CEight inches of packing.
39
-------
dioxide and ash to the stack gas. The fly ash emission
could be minimized by fIring the coal on a grate stoker.
2. Bypass the scrubber with part of the gas stream
and mix this gas with the scrubber exit gas.
This procedure requires minimum investment and has
essentially no operating cost. As discussed earlier,
however, bypass of the 310° F. gas from a point after the
air heater increases ground-level concentration rather
than decreasing it. Bypass of gas from upstream of the
air heater would be more effective but this would reduce
boiler efficiency and possibly give rise to problems in
removing the dust.
3. Use heat exchangers for direct transfer of heat
from the scrubber inlet gas to the exhaust gas.
With this method, heat that would be wasted is
recovered. Further advantages are reduction in the
amount of water required for evaporative cooling, a
corresponding reduction in gas volume, and, except for
maintenance, no labor requirement. Disadvantages are
the large heat exchanger required (because of low
transfer coefficient and temperature differential), high
pressure drop, and possibility of fouling-which would
lead to low efficiency and high maintenance cost. For
addition of limestone in the scrubber circuit, corrosion
by sulfur trioxide would be a problem with mild steel
exchangers.
4. Use a cyclic-liquid heat exchange system with heat
transfer from the inlet gas to treated water and from the
water to the scrubber exhaust gas.
The better heat transfer coefficient would permit use
of smaller exchangers than those required for gas-gas
exchange and the smaller surface would reduce pressure
drop and maintenance.
5. Heat with steam from the turbine cycle in a heat
exchanger at the scrubber outlet.
This method would require additional coal firing in
the boiler to generate the extra steam and modification
of the turbine to allow higher than normal extraction
rates. Extensive modifIcation of existing units would be
impractical, but in a new plant a system could be
included to provide the steam required.
Use of steam for reheat would require relatively small
heat exchangers installed only on the scrubber discharge,
where the gas is relatively clean. Corrosion, fouling, and
pressure drop would be minimized. The main
disadvantage is the added fuel requirement.
6. Use a cyclic system comprised of heat exchange
towers where liquid or solid particles are sprayed into
the gas stream ahead of the scrubber and the sensible
heat gained by the particles is transferred to the scrubber
exit gas in a similar chamber.
With this system, there would be no heat exchangers
to foul or corrode and the pressure drop would be low.
With a liquid, partial dust removal could be effected by
40
filtering or centrifuging the liquid from the "hot" tower.
However, a low vapor pressure over the liquid would be
required to prevent carryover to the scrubber and the
liquid should be nonflammable or have a high kindling
temperature to prevent fire hazard. Use of solid particles
would require a material with good abrasion resistance
to withstand the rough handling.
Insufficient information is available to prepare a cost
estimate on this reheat method.
The cyclic-liquid heat exchange method of reheat was
chosen for the conceptual design; a cost estimate (see
economics section) showed it to have the best combination
of investment and operating costs.
With process B, the gas entering the scrubber will
contain sulfur trioxide whereas with limestone injection the
sulfur trioxide is reacted with lime in the boiler. The
portion of the heat exchanger at the scrubber inlet exposed
to gas at a temperature below the dew point of sulfur
trioxide (about two-thirds of the surface) was assumed to
be constructed from a corrosion-resistant alloy steel.
Addition of sufficient limestone in the boiler to react with
the sulfur trioxide would probably prevent the corrosion
problem but adding absorbent at two points would
complicate the process.
Equipment Description
The plant arrangement for receiving and handling
limestone is the same as presented in the report of the dry
process. Limestone is received by truck shipment, dumped
into an unloading hopper, and conveyed to the storage silo.
For process A, the material is conveyed from storage to a
surge hopper over the mill, ground, blown by the
mill-sweeping air to a feed bin equipped with a bag filter
vent, passed through a variable-speed screw feeder, fed into
the air injection pipe by a rotary valve, and collected along
with fly ash in the scrubber (Appendix D, Fig. D-l and
D-2). For process B, the limestone is conveyed from storage
to a surge bin, weighed, ground in a wet ball mill, pumped
as a slurry to a hold tank, and metered into the scrubber
recirculation tank (Appendix D, Fig. D-3). The equipment
items described below are for process A (Appendix D, Fig.
D-4). The same equipment, except for slight differences in
size, is used for process B except where indicated
(Appendix D, Fig. D-S).
Unloading Hopper: The unloading hopper is a 20- by
20-foot concrete hopper with sides sloping to a center
opening above a belt conveyor 12 feet below grade; the top
of the hopper is at grade for dump truck receipts. The
hopper will accommodate about 50 tons of limestone or
about three truckloads.
-------
Unloading Conveyor: The unloading conveyor transports
stone from the unloading hopper to a storage silo. The
conveyor is a truss-frame, troughed-belt type 160 feet in
length with a 24-inch belt.
Storage Silo: Incoming limestone is stored in a steel silo
25 feet in diameter with 30-foot straight sides. The silo will
accommodate about 600 tons of limestone or about a
60-hour supply for 110% stoichiometric treatment,
assuming 3.5% sulfur in the coal; this reserve is sufficient
for operation over a weekend without delivery. The storage
silo discharges onto a transfer conveyor belt.
Transfer Belt: Limestone is conveyed from the storage
silo to a surge hopper by a 230-foot belt of construction
identical to the unloading conveyor.
In process B, limestone is discharged from the storage
silo onto an 80-foot-Iong belt conveyor which empties into
a 20-foot high, continuous bucket elevator with a capacity
of 15 tons per hour. The elevator discharges into a surge
hopper above the mill feeder; the hopper and feeder are
similar to those described for process A. The feeder
discharges into another elevator (same specifications)
which empties into the mill.
Surge Hopper: A surge hopper is provided above the mill
to protect the injection system from minor delays in
operation of the transfer system. The hopper is a mild steel
tank 12 feet in diameter with IS-foot straight sides and a
conical bottom. The storage capacity is equivalent to 8
hours' supply.
Limestone Feeder: Rate of feeding limestone to the mill
is controlled by a beit-type feeder with a capacity of 15
tons per hour.
Pulverizer: The pulverizer, which reduces the limestone
from minus 1/4 inch to 70% minus 200 mesh, is a Band W
type E bowl mill of the type used for grinding coal. The
mill has a capacity of 15 tons per hour when grinding coal
to 70% minus 200 mesh. Air for classifying and drying in
the mill and transporting pulverized limestone to the feed
hopper is pulled from the duct that supplies preheated air
to the coal burners.
For process B, a Denver ball mill, 7-foot diameter by 10
feet long (measured inside the wear plate liners) and
designed to operate in closed circuit with a 60-inch
diameter Model 150 Denver spiral classifier, was selected to
wet grind the required 10 tons of limestone per hour from
1/2 by 0 inch to 70% minus 200 mesh. The mill is equipped
with a 300-horsepower motor. With a larger drive (350 hp.)
and larger classifier (72-in. diameter), the mill could be used
to produce a finer product, about 95% minus 325 mesh,
with less than 10% increase in operating costs.
Limestone is fed continuously to the mill along with
recycled oversize; makeup water is added in the mill. The
mill discharge slurry flows to a spiral classifier where coarse
particles settle and are recycled to the mill. The onsize
particles do not settle quickly enough to be picked up by
the spiral and are carried out of the classifier in the liquid
overflow as a 10% slurry. The recycle to product ratio is
about 7: 1.
Ground Limestone Feed Tank: The ground limestone
feed tank serves both as a receiver for the pulverizer and a
supply for the injection system. The tank is a mild steel,
cone-bottomed unit 8 feet in diameter and 10 feet high
equipped with a bin-vent type bag filter capable of handling
300 cubic feet per minute of air. An exhaust fan is provided
to compensate for the pressure drop across the filter. The
tank is mounted on load cells for rate checks-
For process B, a 150,000-gallon tank is provided to store
ground limestone as a 10% slurry ready for addition to the
scrubber recirculation tanks. The storage tank, equipped
with an agitator, will accommodate an 8-hour supply of
limestone to allow for minor maintenance of the mill.
injection System: Ground limestone is metered into the
injection system through a 6-inch screw feeder equipped
with a variable-speed drive; capacity of the injection system
is 15 tons per hour. The feeder discharges through a 6-inch
rotary valve into a mixing nozzle. This single metering
system supplies absorbent for all the coal burned;
instrumentation for ratio control is provided. The limestone
is transported in dense phase (0.07-0.08 lb. air/lb. of solids)
to the injection point through a flow splitter, a cylindrical
tank with six 2-inch pipe outlets equipped with orifice
plates. The flow of air and entrained solids is divided into
six streams and conveyed through separate 2-inch pipes to
flow nozzles that project through the boiler wall; the
nozzles can be pivoted in a vertical plane to vary the angle
of injection. Separate rotary compressors (maximum of 10
p.s.i.g. discharge pressure) will provide up to 400 cubic feet
per minute each for conveying and injecting the limestone;
the injection air will be added to the transport stream in
mixing tubes just ahead of each nozzle.
Duct Modification: Existing mechanical dust collectors
are removed and new ducts (8 by 13 ft.; insulated mild
steel) provided to direct the gas flow from the two air
heaters to the scrubbers, bypassing the precipitator. Blanks
inserted at flanges on the precipitator inlet would have to
be removed with rigging equipment if activation of the
precipitators were needed. Discharge ducts (7 by 12 ft.) are
provided to return the gas to the existing precipitator
discharge duct. The general arrangement of ducts and
scrubbers is shown in Appendix D (Figs. D-6 and D-7).
41
-------
Scrubbers: I\t full load, each of the two scrubbers will
handle 272,500 actual cubic feet of gas. The units are
two-stage, countercurrent, Turbulent Contact Absorbers
manufactured by the Air Correction Division of Universal
Oil Products Company (Appendix D, Fig. D-8). The
scrubber shell has a 15 - by 20-foot rectangular cross
section, i~ 40 feet high, and is constructed of mild steel
lined with a 1/8-inch bitumastic coating. Internal support
grids and spray nozzles (for liquid distribution) are of
stainless steel construction. The bottom of the vessel serves
as the receiver for the scrubber liquor, which drains by
gravity flow into a recirculation tanle The mobile bed is
composed of 1-lj2-inch-diameter polypropylene spheres.
The scrubbers are designed for 6-inch pressure drop with an
LjG volume ratio of 50.
At reduced load, one of the scrubbers will be shut down
by closing louvered-type dampe]s installed at the scrubber
inlet.
Separate recirculation tanks are provided for each
scrubber. Each has a capacity of 45,000 gallons, is 22 feet
in diameter by 14 feet high, and is equipped with a cone
bottom for discharge of solids from the system through a
variable-speed pump. Separate circulating pumps deliver
8,000 gallons per minute at 11O-foot head with
300-horsepower drives; installed spares are provided. The
system is. designed for 10% solids in the slurry. A pH
monitor is installed in the circulation tanks.
Reheat System: Mild steel, finned-tube heat exchangers
are installed in the duct work on each side of the scrubber.
The tubes have an outside diameter of 2 inches and
3j4-inch fins spaced five per inch. Based on a coefficient of
8 B.t.u. per hour per square foot, 145,000 square feet of
surface area is required. The tubes are 20 feet long and art
arranged in 8- by 16-foot tube bundles.
Circulating water (treated boiler water) is used to
transport heat from one exchanger to the other. In a cold
climate, antifreeze would need to be added for periods
when the system was shut down, such as in peak-load
operation. The design temperature profile is:
42
Heat removal, ° F.
Gas
Liquid
Reheat,O F.
Gas
Liquid
300-160
139-275
118-250
275-139
Liquid recirculation rate is 600 gallons per minute.
For process B, approximately 100,000 square feet of the
exchanger at the scrubber inlet is constructed from Corten,
a steel alloy containing nickel, chromium, and manganese.
Induced - Draft Fans: The overall increase in pressure
drop is 12 inches of water (6 in. in scrubber, 4 in. in reheat
system, and 2 in. in duct work). Existing fans (1200-hp.
motors) on both trains of the exhaust system are replaced
with new fans (1500-hp. motors) designed for 26-inch static
pressure. In some installations the required static pressure
could probably be met by equipping existing fans with
larger, higher speed motors. In others, installation of a
modified impeller in an existing fan housing, in addition to
a drive change, might suffice.
Waste Disposal: Rate of scrubber effluent discharge will
be controlled with a variable-speed pump connected to the
conical bottom of each recirculation tank. The pumps
discharge into a surge tank which serves as a supply for a
1000-gallon-per-minute slurry pump. Slurry is pumped
1500 feet through a new 12-inch line to the storage pond
area. Water flush connections are provided at the slurry line
inlet for use if solids settling is a problem.
Storage Pond: The impounded volume for disposal of
ash was increased to accommodate the additional solids
load resulting from limestone scrubbing. It was assumed
that acreage would be available at a price consistent with
current land value. For continuous operation of the
scrubber system with gas from combustion of coal
containing 3.5% sulfur, about 30 additional acre-feet will be
required annually. Assuming impoundment to a height of
10 feet, which is current practice for fly ash at Colbert, 3
acres more per year are needed. Cost of pond development
for 10 years of operation is included in the investment
estimate.
-------
Economic Evaluation
The major cost factors in the limestone-wet scrubbing
process are examined in this section. Included are
evaluation of process alternatives, raw material purchase
and shipping costs, and overall process investment
requirements and operating costs. Representative curves are
included where applicable and detailed cost estimates are
given in Appendix C.
Basis of Evaluation
The assumptions and design criteria discussed previously
form the basis for evaluating the two processes under
consideration-limestone injection followed by wet
scrubbing (process A) and lime or limestone addition
directly into the scrubber circuit (process B). To get a
uniform basis for comparing these, estimates were first
made to determine the most economical of the several
alternatives in regard to number of scrubber stages, type of
solids disposal, limestone versus lime for process B,
precipitator-scrubber operation, and stack gas reheat. Also
examined were the cost of system pressure drop and stack
gas reheat temperature so that their effect could be
identified. Finally, after making the basic comparison, the
effects on overall process economics were determined for
such variables as unit size, sulfur content of coal, onstream
tim\::, and raw material price.
The operating costs reported include appropriate capital
charges and overheads and the investments include
engineering costs, contractor fees, and contingency charges.
Since operating cost is handled in this way, differences in
investment between process alternatives are given proper
weight.
Fixed Capital Charges: Suggested guidelines for applying
fixed charges to cost evaluations of private and public
power production and distribution systems are outlined by
the Federal Power Commission (FPC) (20). A further
treatment of the economics of the regulated power industry
is given by Phillips (21).
In the FPC report, the components of fixed charges are
given as cost of capital, depreciation, interim replacements
of equipment having short life, insurance, and taxes. These
are applied as a percentage of power plant investment. The
term "cost of capital" is defined as the annual percent of
investment which a company should receive to maintain its
credit, pay a return to the owners, and ensure attraction of
money for future needs. As used in the FPC formula for
fixed charges, cost of capital includes the interest on bond
debt and a return on equity investment.
The fixed charges applied throughout this evaluation
were adapted from the guidelines suggested by FPC; the
investment percentages are shown below:
Annual % of
l~~rLl!)~~tJ:!lJ!!!~
Existing New
_!!.I!lL- !!~t
Depreciation (average of 20-yr life
assumed for existing units; 35-yr
life for new power unit), straight
line method
Interim replacements (equipment having
between 20- and 35-yr life)
Insurance
Cost of capital (capital structure
assumed to be 50% debt and 50% equity)
Bonds at 6% of average
undepreciated investment
Equity at 11 % of average
u ndepreciated investment
Taxes
Federal
State
5.00
2.85
0.25
0.65
0.25
1.50 1.50
2.75 2.75
2.80 2.80
l1Q lJQ
14.50 13.00
Total fixed charges
Credits for Reduction in Investment and Operating Cost
in Parts of the Power Plant Other Than the Scrubber: In
most existing fossil fuel-fired power plants the particulate
matter in the stack gas is removed by mechanical dust
collectors and/or electrostatic precipitators and the gaseous
contaminants are dispersed into the atmosphere by high
stacks. In this evaluation it is assumed that existing power
units are equipped with electrostatic precipitators (about
90% efficiency) and 300-foot stacks. If a new power unit
were to be built, it would in all probability have a more
efficient electrostatic precipitator (probably at least 99%)
and a higher stack.
When a limestone-wet scrubbing process is added to a
power unit, other types of dust collectors are not required.
In the case of existing units equipped with an electrostatic
precipitator, an operating cost credit could be claimed if
the precipitator were not operated. For new power units,
both an operating credit and an investment credit would be
valid if a precipitator were not installed. Such credits are
taken in this evaluation where appropriate.
An investment credit mayor may not be justified for
reduction in stack height. The cost per foot of such stacks
43
-------
increases rapidly with height (Fig. 11), and a limited
reduction in height would give a disproportionately large
reduction in investment. As noted earlier, it has been
assumed that no credit for lower stack investment would
normally be justified. In cases where justification exists, the
appropriate investment to be credited can be obtained from
Figure 11. Although this curve does not show directly the
effect of unit size, size is included, in effect, because the
curve is based on actual costs of a variety of units in which
stack height increased as the unit size increased.
Credit has also been taken for reduced corrosion of
boiler heat-exchanger surfaces when process A is used. The
corrosion reduction credit is applied to maintenance cost
only since it is assumed that the amortization life of the
power plant will remain unchanged.
Process Alternatives
To provide a basis for overall evaluation, the following
process alternatives were evaluated:
1. Number of Scrubber Stages: Both one- and
two-stage scrubber units, with each stage designed for a
pressure drop of approximately 3 inches (water), have
4,000
Concrete stack with steel liner
....
.J::.
0>
.jjj
.J::. 3,000
~
(,)
c:
been considered. Because of the higher pressure drop,
the two-stage scrubber will remove more of the dust and
sulfur dioxide from the stack gas; however, the
investment and operating cost will be greater than for a
single-stage unit.
The two-stage unit was chosen because of its higher
gas-scrubbing efficiency. It should be recognized,
however, that applications will probably exist for
single-stage scrubbing, e.g., power plants in which more
than 3 inches of added pressure drop would place an
intolerable operating or economic burden, or situations
where the higher degree of sulfur dioxide and dust
removal is not required. An economic analysis of the
investment and operating costs for both one- and
two-stage units in a 200-megawatt existing power facility
is given in Table 10 along with expected dust and sulfur
dioxide removal efficiencies. In addition, the cost of
system pressure drop over the range of 14 to 32 inches
water is shown graphically in Figure 12.
2. Solids Disposal: Current methods of solids disposal
at coal-burning power plants range from sluicing to
nearby ash ponds to dry transport, sometimes for long
distances, by rail, truck, or barge. The operating costs
800
1000
200
400 600
Stack height, ft
Figure 11. Cost of power plant stacks
44
-------
associated with these methods range from $0.50 to
$4.00 per ton of solids or higher. Since the
limestone-wet scrubbing process more than doubles the
quantity of solids to be disposed of (from 52,000 to
about 157,000 tons/yr. from a 200-mw. unit burning
pulverized coal containing 12% ash and 3.5% S and with
110% stoichiometric CaCO 3 injection), its use would be
most applicable to those plants that have low-cost
methods of solids disposal. Generally, these would be
plants having on-site disposal ponds and available
additional land area. A plant of this type was assumed
for the evaluation.
I~!1!.1~_!!1~~~~~!!.L'!!1i1J~p"e.!'l.JlJ1JI.J~_~1!.
C
.~
~ 80
CIJ
c-
O
co
:J
C
C
«
40
200 mw existing power unit
540 M acfm @ 1300 F
Gas density - 0.067 Ib/d
Maintenance - 1.5% investment
Capital charges - 14.5% of investment
Power cost - $0.004/kwh
8000 hrs/vr operation
Normal operating cost and pressure drop in power unit
12
16.
32
20 24 28
Fan pressure head, inches water gauge
Figure 12. Effect of pressure drop on operating
cost of stack gas exhaust fan
0.267
]
c
...
:J
.c
co
o
(.)
0.200 c
o
....
----
~
....-
en
o
(.)
C>
C
.~
co
0.133 ~
o
co
:J
C
C
«
0.067
36
45
-------
800
40% slurry-
No water return
en
~
.!!!
o
-c
"I-
o
-c
I::
~ 600
o
..r::.
...
...-
I::
II.>
E
...
~
>
I::
400
157,000 tons/yr solids (ash included)
Disposal pond included
200
2.0
0.5 1.0 1.5
Scrubber-to-pond distance, miles
Figure 13. Investment for various
methods of solids disposal
Three methods of on-site solids disposal were
compared for various scrubber-to-pond distances. These
were (1) direct pumping from scrubber to pond of dilute
waste slurry (10% solids) with no return of water, (2)
direct pumping to pond of 10% slurry with the pond
serving as a settling basin before return of the water for
reuse, and (3) thickening to a 40% slurry and pumping
to the waste pond with no return of water. The costs
shown are based on the amount of solids that would be
produced by a 200-megawatt unit using process A with
coal containing 3.5% sulfur and 12% ash. A value of
$0.10 per 1000 gallons is placed on the water.
The results are shown in Figures 13 and 14. Method
1, direct pumping of 10% slurry with no return of water,
requires the least investment, has the lowest operating
cost for distances over 0.6 mile, and, incidentally, is the
method used for handling fly ash in most coal-burning
power plants. It was selected as the method to be used in
the further estimates.
3. Use of Calcined Limestone in Process B: If calcined
limestone were used in place of raw limestone in process
B, more efficient scrubbing would be realized with a
resultant reduction in material requirements and
46
solids-handling and disposal cost. However, based on the
cost of calcining determined in the dry process study,
the higher price of calcined limestone more than offsets
any savings due to higher scrubbing efficiency (Table
11 ).
4. Precipitator-Scrubber Operation: Addition of a
wet-scrubbing process to a power plant already equipped
with an electrostatic precipitator requires a choice
between operating and bypassing the precipitator. Since
process operating and investment costs are affected, an
appraisal of the alternatives was made with the results
shown in Table 12.
The results indicate that operating the precipitator
would increase costs for process A and would offer little
advantage in process B. Therefore further estimates for
existing power units are based on inactivating the
precipitators.
5. Stack Gas Reheat: Five different methods (see
earlier description) of reheating the scrubbed and cooled
stack gas were evaluated: (1) direct heating with natural
gas, (2) indirect heat exchange between the stack gas and
a heat-transfer liquid, (3) indirect heat exchange
-------
'"
Jg 150
o
"0
.....
o
'"
"0
c::
!j!
:J
o
.I::.
....
....- 100
'"
o
u
C')
.S:
....
co
....
Q)
C-
o
co
:J
c::
c::
«
200
1 0% slurry-water returned
t
40% sturry-
No water return
t
-x---
--X---~
50
8000 hr/yr operation
157,000 tons/yr solids (ash included)
Water - $0.10/1000 gal
0.5 1.0 1.5
Scrubber-to-pond distance, miles
Figure 14. Operating cost for various
methods of solids disposal
Table 11. Effect of Absorbent Type on Operating Cost for
Process Sa
--------------------Raw---------(a~1fd---
limestone, 130% limestone, 100%
stoichiometric stoichiometric
I@=r!!~-IQ[O~i!i~ .fQ.~~i-1Q@jJl~
96,000 480,000 75,000 375,000
Limestone, tons/yr
Annual operating cost, $
Raw material
Handling
GrindingC
Calcinationd
Solids disposal
196,000 985,000 154,000 770,000
80,000 240,000 66,000 180,000
35,000 100,000
- 370,000 900,000
1:t1LQ!J.Q !Q!J.L0.!)!! 110,0.!)!! ]]~!!Q.!L
_J~~l__________i~~Q!J.Q_1J£~Q.OJ)_IQ.~Q!J.Q_~l~~Q!J.Q
a3.5% sulfur in coal, limestone at $2.05/ton, and 14.5% capital
charges.
b Assumes limestone purchased and calcined at power plant.
cGrinding cost based on figure 9 in Sulfur Oxide Removal from
Power Plant Stack Gas-Sorption by Limestone or Lime (Dry
Process), Tennessee Valley Authority, Muscle Shoals, Alabama, p.
37, 1968, with capital charges adjusted from 11% to 14.5% of
investment as used in this report.
dCalcination cost based on figure 11 in above report; capital charges
adjusted from 11% to 14.5% of investment as used in this report.
0.333
0.250
~
c::
....
:J
.!J
co
o
u
.....
o
c::
o
~
Y)
0.167 tf
o
u
C')
c::
.;:;
co
....
Q)
c-
o
0.083
co
:J
c::
c::
«
2.0
Table 12. Dust Removal Efficiency, Investment, and
Operating Cost of Precipitator-Wet Scrubber Combination
~-~~~~~£~~m~_~~~~_~~f______------------
Additional Additional
oper. cost new invest.
for systemb for system c
--_J_-- ___1__-
System
...P!.I!f~SS - QQ!!!.
-------
1,000
440 M scfm @ 3000 F leaving air heater
Process A - Limestone injection
Plant size, 200 mw
800
'"
....
.E!
o
"C
....
o
'"
"C
I:
~
:J
o
~
.....
600
.....-
I:
Q)
E
.....
'"
Q)
>
I:
400
200
Gas-liquid heat exchange --*
t
Direct heating with natural gas
100
130
160 190
Reheat temperature, 0 F
220
250
Figure 15. Effect of reheat temperature
on investment for stack gas reheating
between the gas streams before and after scrubbing, (4)
indirect heating with steam (in a new power plant
designed to produce excess steam), and (5) direct
heating with coal in a stoker system. In addition, the
effects of degree of reheat on operating cost and
investment were determined for two of the
methods-natural gas heating and indirect liquid-gas heat
exchange (Figs. 15 and 16).
A 200-megawatt unit, coal containing 3.5% sulfur,
o .
and reheat to 250 F. were assumed for the companson
of reheat methods; the 250 of. was selected arbitrarily
because higher levels are quite expensive by the heat
exchange methods and do not appear justified from the
standpoint of stack gas dispersion. The costs of pressure
drop and of quench water to precool at the scrubber
were included in the estimate, as these differ with
method used. The results (Table 13) show that liquid-gas
48
heat exchange is the most economical and it was selected
for the further estimates.
o
Gas reheat to temperatures lower than 250 F. may
be acceptable in some applications of the wet-scrubbing
process. Adjustments can be made to the overall
investment and operating costs (Figs. 19 and 21), which
o .
are based on reheat to 250 F., through use of Flgures 15
o
and 16. For example, for reheat to 190 F. in a
200-megawatt unit by the gas-liquid heat exchange
method, the difference in investment (from Fig. 15)
between 2500 and 190"F. is $600,000 or $3.00 per
kilowatt, which should be subtracted from the overall
investment; likewise the operating cost (from Fig. 16)
would be reduced by $120,000 or $0.20 per ton of coal
burned.
If it were necessary to reheat to temperatures higher
than 2500 F., a combination of reheat methods would be
-------
300
en
....
.!2
o
"0
....
o
en
-g 200
<0
en
::J
o
..c:
....
8000 hrs/yr - Process A
440 M scfm @ 300° F leaving air heater
Natural gas @ $0.40/mscf
Capital charges - 14.5% of investment
Plant size, 200 mw
tf
o
u
g>
.~ 100
Q)
a.
o
Cii
::J
C
C
«
100
130
160 190
Reheat temperature, of
250
.500
c
o
....
--
Y}
0.250 ~.
u
C)
c
'';::::;
<0
....
Q)
a.
O~ 125 ~
<0
::J
C
C
«
220
Figure 16. Effect of reheat temperature
on operating cost of stack gas reheating
best for economy. The indirect method would be used
to reheat up to 230°-240°F. and the natural gas- or
coal-fired stoker methods for heating to higher levels.
For example, for reheating to 300°F. in a 200-megawatt
unit, the operating cost would be $130,000 per year for
the indirect portion of the heating and $180,000 per
year for direct natural gas for a total of $310,000 per
year or $0.52 per ton of coal burned.
In some situations, reheat may be required only when
adverse atmospheric conditions reduce stack gas dispersion.
For such intermittent operation, the low investment, direct
heating methods would be applicable because the cost of
fuel, which puts these methods at a disadvantage for
continuous use, would be eliminated during nonoperating
periods. The effect of reheating operating time (with
natural gas) on overall operating costs is shown in Figure
17. Operating cost would be $1.10 per ton of coal with
reheat 10% of the time compared with $1.31 for
continuous reheat by the cyclic heat exchange method (Fig.
21). However, if the natural gas unit were operated as much
as 50% of the time there would be little or no advantage
over continuous reheat by heat exchange.
Natural gas was selected for the comparisons in Figures
15 and 16 because it is well proven, flexible, clean,
convenient for adding to an existing plant, and low in
investment. The coal stoker method is not well proven and
reduces the effective sulfur dioxide removal by introducing
sulfur dioxide from the coal into the gas. Steam-gas heat
exchange is also more economical but is applicable mainly
in new power plants because of the steam required.
Sorbent Vendor and Shipping Costs
Vendor Cost: To determine the cost of raw limestone
for use as sorbent in processes A and B, a series of fo.b.
price quotations were obtained from vendors located
mostly in the eastern United States. The prices quoted are
shown in Table 14 and cover varying particle size and
calcium oxide and magnesium oxide content. Much of this
information was obtained in the previous dry process study
and is repeated here for convenience.
It should be noted that these quoted prices are based on
costs in existing plants and on an amount of material
(50,000 tons/yr.) considerably less than that which would
be used by a modern large power plant. If a long-term
contract for a very large quantity were entered into, then
the vendor might be justified in putting in new equipment
and quoting a special low price.
It is not possible, of course, to estimate very closely the
purchase price under a long-term contract for a very large
quantity of limestone, say, the 386,000 tons per year
needed for a WOO-megawatt unit (at 3.5% S in the coal and
11 0% stoichiometric). A good figure could be obtained
only by actual negotiations in a specific situation. However,
discussions with limestone producers have supplied a rough
49
-------
1,000
800
Process A
Existing 200-mw unit
3.5% S in coal
Stack gas reheat to 2500 F,
by direct-fired natural gas
8000 hrs operation
of scru bber
system
1.50
en
....
.!!!
o
"'0
....
o
"'0
CIJ
c:
....
:J
.J:I
V>
"'0
c:
~
:J
o
..c:
....
....-
V>
o
u
en
c:
....
co
....
CIJ
g- 400
600
Cost without reheat facilities installed
co
o
u
....
o
c:
o
....
1.00 *
....-
V>
o
u
en
c:
....
co
....
CIJ
c-
o
co
:J
c:
c:
co
co
:J
c:
c:
co
co
....
CIJ
>
o
0.50 e
CIJ
>
o
200
20 40 60 80
Percent of operating time in which reheating unit is operated
(100% = 8000 hrs/yr)
100
Figure 17. Effect of intermittent operation on cost
of reheating stack gas by combustion of natural gas
!~~~l~~~~_~_~a~B~b~~~~_~~h~~DIT~__---------------------------------------------------
labor, $/ton
Fan and maintenance Capital coal Reheat
l~~'!!!.gl!~ Fue1 $ J!Q.~~,-t 1t_I!Y~.h~.!!,-t ~!!a~~j- lQ.~l! .!!!!.r.!!~!L l!!y!~,,-t
Direct, natural gas 241,000° 3,100 5,300 38,400 293,800 0.49 265,000
Indirect, liquid-gas c 31,000 23,000 131,200 185,200 0.31 905,000
Indirect, gas-gas c 55,000 64,200 333,000 452,200 0.75 2,300,000
Indirect, steam-gas i47,600d 14,000 16,800 60,900 247,300 0.41 420,000
Qk~~oalgQ~~____l@l]Q~___-]JQ~------~~Q~-------Q~£@----_JQ~~@-----_J~~-------_!~~@Q--
a 0
b200-mw power plant; reheat to 250 F.; 440 M scfu
Natural gas at $OAO/MM Btll.
~ndirect liquid-gas and gas-gas utilize waste heat in exhaust gases.
eSteam at $0.30/MM Btu from 200-mw plant designed for excess process steam.
Coal at $0.20/MM Btu.
50
-------
!~~~1~~9~~~~Y~~~-~~~~~~l~~1~n~_____--------------------------------------------------
P-l!!tj(~~SE!
Mesh size % through ---.Q.!!I.!!~1''!...._-- ----_£~j:J-,-!l.!Ql1..L.!!\!!~------
l!1.!!~~.!!!!. JIb!~n_,!!~!!. !;.'!..O- MllQ Q.f_'!!.a.!I[@L ~L~.9.i1!!Q.MllQ
% in. 100 51.3 1.7 1.35 2.55
200 70 53.5 0.8 3.85 7.10
16 100a 55.3 0.4 2.85 5.10
300 62b 55.3 0.4 2.85c 5.10
300 80 55.3 0.4 5.85 10.50
300 98 55.3 0.4 6.60 11.80
300 99.6 55.3 0.4 7.60 13.65
60 50d 54.0 1.65 3.05
325 75e 54.5 0.6 4.00 7.25
% in. 100 51.7 3.6 2.00 3.60
3 in. 100 30.6 21.5 1.50 2.90
200 48f 30.6 21.5 2.50 4.80
Kimballton, Va. 200 82 54.0 2.0 3.00 5.35
Stephens City, Va. 200 75 54.8 0.3 2.75 5.00
Strasburg, Va. 3/8 x 1/8 in. 100 55.0 0.4 2.00 3.60
Adams, Mass. 200 98g 54.0 0.7 10.00 18.30
1~!~!~~~YL~lih_________1Q~____-----~~-------_J~~-_____~1------_J-,-@-----______l~£~___-
~1 % -so mesh; sized product.
100% -35 mesh.
~LOW price due to product being surplus.
90% -10 mesh.
e97.S% -100 mesh.
fS7% -40 mesh.
g29% -15 microns; 11% -4 microns.
Location
Cherokee, Ala.
Longview, Ala.
Ste. Genevieve, Mo.
Roberta, Ala.
Coleman, Fla.
Russellville, Ky.
Durbin, O.
estimate of the potential price. The consensus for a
500,000-ton-per-year supply was about $1.15 per ton; the
range was $0.71 to $1.50 per ton.
H is concluded that for relatively small limestone
consumption, for example, the 44,000 tons per year needed
for a 200-megawatt unit with 2% sulfur coal and 110%
stoichiometric injection, or the 36,000 tons
(approximately) needed for a 1000-megawatt unit under
the same conditions but operated on an intermittent
injection basis, the quoted price range of crushed limestone
(maximum size, 1/4 in.), $1.35 to $2.00, is a good figure.
For larger consumption, however, particularly for the 1000-
to 2000-megawatt or larger central stations, much lower
cost can be expected. The ultimate might be a situation in
which a limestone plant of very large capacity, located
adjacent to a navigable river, supplied several large power
plants upstream and downstream by barge; this is similar to
the supply situation on the Great Lakes, where a large
limestone company sells crushed limestone in large
quantities at $0.73 per ton f.o.b. ship.
Shipping Cost: Freight costs of limestone by rail or
truck transportation are given in Figure 18. Except for
localized situations, rail shipment is more economical than
truck shipment.
Barging costs are much less than for rail or truck but vary
so widely that a curve of cost versus distance would be
misleading. Barging is suited mainly to large-volume
shipment on a contracted minimum annual tonnage basis,
and, of course, to shipping and receiving points on or close
to navigable water. Where such a combination exists, very
low rates can be obtained; TV A, for example, ships coal in
large quantities at an average of about 2 mills per ton-mile.
The conditions that make this rate possible may be
somewhat favorable, however; the following costs have
been estimated as generally applicable to barge shipping of
limestone:
_T.!J!!~YL
150,000
1,000,000
j)l!lP.I!!!}9-~~t.!l!Ql1
200_'!!L
0.75
0.65
]Q~!!)l
0.90
0.80
!I!!L'!!L
0.50
0.45
There should be added to the barge rates the cost of
unloading at the power plant, since there is little or no
unloading cost for trucks of hopper-bottom rail cars. A
typical barge unloading cost, based on power plant
experience, is $0.16 to $0.17 per ton at an unloading rate
of 1000 tons per hour.
51
-------
8
"0
Q.)
Q.
Q.
£;
'"
ro 6
.;:
Q.)
....
ro
E
c
o
....
{;9:4
Rail-l00,000 Ib minimum shipment
Truck-37,OOO Ib minimum shipment
..;
'"
o
()
....
£;
0'>
'i!)
'-
:; 2
::;
co
50
100
t
Rail
150
Distance, miles
200
250
300
Figure 18. Bulk transportation cost for
shipping limestone by truck or rail
Shipping cost can also be reduced by the unit-train
method-that is, a special train carrying only limestone and
proceeding direct from quarry to power plant. This practice
is also generally based on a contracted minimum annual
tonnage, and the cost varies widely with conditions. TV A,
for example, receives coal by unit-train shipment at rates
ranging typically from 4.7 mills per ton-mile for 343-mile
shipment to 9.2 mills for 114 miles. Minimum annual
tonnage for such contracts is I to 2 million tons; minimum
trainload tonnage is about 5000 tons.
In comparison with Figure 18, which is based on single
carloads, the shipping cost for the unit train is about 30%
lower.
As previously mentioned, a Colbert County, Alabama,
location has been assumed for the purpose of evaluating the
economics of limestone-wet scrubbing. Based on a
quotation from a vendor located in the Colbert County
area, the assumed f.o.b. cost of 1/4- by O-inch limestone
used in the overall process evaluations presented herein is
$1.35 per ton. Trucking cost to the power plant would be
$0.70 per ton, giving a delivered cost of $2.05 per ton of
raw limestone.
Overall Process Investment Evaluation
The ovenill fIxed investments for processes A and B in
existing plants of various sizes (3.5% S in coal) are
summarized in Table 15 and shown graphically in Figure
19. In addition, curves in Figure 19 show the investment as
dollars per kilowatt of power generation capacity, a
52
Table 15. Overall Investment for Scrubbing Facility in
~~~Q~~~~L______------------------------
-------------~~~g~-~~-------------
---j]Q---- ----@Q---- ---~~~---
Process ___.L- _:lliw.- _--1..-_....$.lkw ---1..---$fuv
A 2,610,000 13.05 5,425,000 10.85 8,210,000 8.21
~---~~~~~Q_l~~~_3~~~~~_J!~~-~&~~~0~--~~~
common power industry method of expressing investment.
Investment for process A includes equipment for
limestone receiving, storage, conveying, dry pulverizing,
injection into the boiler; two-stage gas scrubbing; indirect
liquid-gas cycle reheat of scrubbed flue gas to 250°F.;
direct pumping of 10% waste solids to pond with no return
of water; and incremental pond storage of the product
solids. Investment for process B includes the same
equipment as for process A except that wet grinding and
slurry preparation replace dry pulverizing and injection into
the boiler.
The overall investments for processes A and B installed
in new 1000-megawatt plants are shown below. Also given
is the effective investment after credit is taken for
eliminating the electrostatic precipitator (99% efficiency
Process
A
B
Overall investment for scrubbing
-----_J~~@Yl~~~~~~tU~Q~~~l_____-
-----~.f!!!~------ ~f!~...Pl.~lPjJ~-tQ!..q~c!lt
---j--- _~k~- ---j---- -jl!<~
7,620,000 7.62 6,320,000 6.32
8,270,000 8.27 6,970,000 6.97
-------
10
3.5% S in coal
Existing power unit
8
V)
~
.!!!
o
"0
....
o
~ 6
.2
ProcessB
~
E
+..;
!::
CI>
E
t;
CI>
E 4
co
~
CI>
>
o
2
200
400
600
Power unit size, megawatts
800
1000
Figure 19. Effect of power unit size on
limestone-wet scrubbing investment
assumed). This estimate shows the double advantage of
using the wet-scrubbing method in new rather than old
power plants. Actual investment is about 9% lower in a new
1000 megawatt plant and the credit for eliminating the pre-
cipitator gives a further reduction of 14%, or 23% overall.
In all cases, the investment required for process A is less
than that for process B. Major differences are in raw
material and waste handling due to the different raw
material requirements of the two processes (process A,
110% stoichiometric; process B, 130% stoichiometric) and
in the requirement that a portion of the heat transfer
surface for process B stack gas reheat be of alloy steel for
corrosion resistance, whereas process A requires only
regular materials.
The effect on investment of sulfur content of coal
burned, in a 200-megawatt existing unit equipped with
16
>
....
.0
co
Q.
~
C>
!::
12 .~
~
CI>
!::
CI>
C>
~
.::t!
--
~
....-
!::
CI>
8 E
....
V)
CI>
>
!::
co
~
CI>
>
o
4
1200
process A, is shown in Figure 20. The investment is only
slightly sensitive to sulfur content because the gain is
limited mainly to size economy of raw material and
waste-handling systems. Doubling the sulfur content in the
coal (2-4%) increases the total investment only 20%.
Overall Operating Cost Evaluation
A summary of the overall operating costs of processes A
and B at various unit sizes (existing plants and 3.5% S in
coal) is presented in Table 16 and shown as curves in Figure
21. The data relate both total annual cost and cost per ton
of coal burned to the unit size.
The overall operating costs for processes A and B used in
a new 1000-megawatt power unit are given below. These
values reflect the capital charge credit for precipitator
53
-------
e 1 ,000
Q)
>
o
54
'"
....
~
o
-c
....
o 3
'"
c
.2
E
...-
c
Q)
E
1;:
~ 2
.!:
=
C1I
....
Q)
>
o
~ 3,000
~
o
-c
....
o
'"
-c
c
~
::J
o
..c
...
'gj- 2,000
(.)
Cj)
c
...
C1I
....
Q)
C-
o
C1I
::J
C
C
C1I
4
Process A - Limestone injection
200-mw existing power unit
2
3
% Sulfur in coal
Figure 20. Effect of sulfur content of coal
on limestone-wet scrubbing investment
Existing units
8000 hrs/yr operation
Stack gas reheat to 2500 F
3.5% S in coal
110% 510 ichiometric CaG
200
400
600 800
Power unit size, megawatts
Figure 21. Effect of power unit size on
limestone-wet scrubbing operating cost
4
5
1000
6
2.00 ~
(.)
....
o
c
o
..t::
~
1.50 ~-
(.)
Cj)
c
...
C1I
....
Q)
c-
o
1.00 ~
c
c
«
0.50
1200
-------
investment and the economy of a large, new installation as
previously described.
Process
A
B
Overall operating cost
______11~0]~~~~~~~~~~~1____-
Total annual Cost, $/ton
___.£~1j_- -Q!5&!!,!J;I!!r:D~1i
2,327,100 0.775
2,601,000 0.866
The effect on operating costs of sulfur content in the
coal burned, in existing 200-megawatt units using process
A, is shown in Figure 22. Doubling the sulfur content of
the coal (for example from 2-4%) increases the overall cost
by 37%.
Operating cost sheets detailing the above estimate
summaries are given in Appendix C.
The method of deriving the assignable credit for
electrostatic precipitator operation is described in
Appendix C. Also included in Appendix C is a discussion of
the credit for corrosion reduction when limestone is
injected into a boiler as in process A. The method for
calculating thermal loss due to limestone injection is
described in the dry process report.
'"
....
12
o
"0 1 ,000
.....
o
Process A- Limestone injection
200-mw existing power unit
8000 hrs/yr operation
110% stoichiometric CaO
Stack gas reheat to 2500 F
'"
"0
c:
~
~
o
..c:
.... 800
....'
'"
o
u
C)
c:
.';::;
o
0.50
4
5
6
Figure 22. Effect of sulfur content of coal on
limestone-wet scrubbing operating cost
55
-------
4
Process A - Limestone injection
Existing units
8,000 hrs/yr operation
3.5% S in coal
110% stoichiometric CaD
1.33
3
1.00
-c
en Q)
... c
.!!! ...
;j
0.67 ::
..... co
o
o U
en .....
C 0
.Q c
= 0
E -
.......
(j)
tf --
o en
~3 2.008
.S CI
C
- .;;
co
... co
Q) ...
~ 1.67 (I)
o ~
o
~ co
;j
c ;j
c c
co 1.33 15
= 2
co =
... co
Q) ...
> Q)
o >
1.000
.67
200-mw unit
1.50
1.00
2 3
Cost of limestone, delivered, $/ton
4
5
Figure 23. Effect of limestone cost on
limestone-wet scrubbing operating cost
56
-------
Process A-Limestone injection
Existing unit
3.5% S in coal
Stack gas reheat to 2500 F
110% stoichiometric CaG
Boiler operated 8000 hrs/yr
3,000
\II
....
~
.g 1 ,500
.....
o
\II
"0
C
~
:J
o
~
.....
.....-
\II
o
<.>
CI
.5
.....
o
20
60 80
Percent of operating time (100% = 8000 hrs/yr)
Figure 24. Effect of scrubber system operating time
on limestone-wet scrubbing operating cost
1.50
1.00
-g
c
....
:J
0.50::
.....
o
c
o
.....
--
*
o
.....-
\II
o
CJ
CI
c
.;;
o
1.00
0.50
1 0
57
-------
1.875
5
1.500
4
.I::.
:s:
~
--
.!!!
.-
E
tf 1.125 ~ 3
o c:
o :;
g> .!:J
0';:; co
co 0
~ 0
c. ....
o 0
- c:
co 0
~ 0.750 ~ 2
c: ~
co
co
....
Q)
>
o
0.375
58
Existing units
Process A-limestone-wet scrubbing
3.5% S in coal
Stack gas reheat to 2500 F
Scrubber operated whenever boiler is on stream
110% stoichiometric Cao addition
100
Figure 25. Effect of boiler load factor on
limestone-wet scrubbing operating cost
-------
Research and Development Needed
Many of the assumptions for the design and cost studies
were made with limited information. In order to improve
the accuracy of estimates, further information is needed in
the following areas.
Absorbent Efficiency
Further data are needed on the several parameters
involving properties and degree of utilization of the
absorbent:
1. Calcitic versus dolomitic limestone.
2. Effect of overheating limestone injected into the
boiler.
3. Particle size.
4. Variations between limestone types.
5. Oxidation of sulfite to sulfate.
6. Dissolution rate of absorbent, particularly of lime-
stone introduced directly into scrubber recirculating
liquor.
Some of these data will be generated by the full-scale
wet-scrubbing systems being installed commercially. Also, a
pilot plant wet limestone scrubbing program to be
conducted under a NAPCA contract should produce
valuable data on absorption efficiency.
Research is needed particularly on the problems of using
the limestone-scrubber addition process. If good absorbent
efficiency could be obtained, the process could well be
superior to the injection-scrubbing method-in better boiler
operation, lower operating cost, and avoidance of scaling.
Possible approaches to increasing the rate of dissolution,
which presumably is the limiting factor, are as follows:
1. Grinding to small particle size.
2. Increasing the delay time between mixing of the
limestone feed with scrubber effluent liquor and
introduction of the resulting slurry into the scrubber.
3. Introducing carbon dioxide (as stack gas) into the
slurry during or after limestone addition.
4. Prescrubbing with the makeup water to give an
acidic solution for dissolving the limestone.
The problem of absorbent efficiency, as well as all the
other problems of the limestone-wet scrubbing method,
would be more amenable to solution if more were known
about the chemistry involved. Research is urgently needed
on both reaction kinetics and equilibrium compositions.
The latter is currently under study at TV A, in a computer
program in which known equilibrium constants are being
used to determine equilibrium concentrations in the
five-component system CaO-COrS02 -SO 3-H20. Further
work of this type is needed, especially with MgO added,
and the mechanisms of transfer of reactants into the liquid
phase should be established.
Scaling
Although a satisfactory method for preventing scaling
was developed by Howden-ICI, scaling and plugging is a
major problem in the current tests by Union Electric and
Combustion Engineering. Research and development should
proceed along the following lines:
1. Determine if introduction of the absorbent directly
into the scrubber is a major cause of scaling.
2. If so, test arrangements for separating the lime
from the gas ahead of the scrubber and introducing it
into the scrubber liquor circuit after the scrubber. For
an existing plant, the existing precipitator could be used
to separate the lime. For a new plant, two-stage
scrubbing with a venturi used to remove the lime in the
first stage should be tested.
3. Develop further data on the Howden-ICI process
(introduction of absorbent into the recirculating liquor)
to determine if it is adaptable to the more efficient
modern types of scrubbers and to various types of
limestone.
4. Study further the chemistry of the system in
respect to equilibrium composition, degree of
supersaturation, rate of desupersaturation, and
adherence of precipitated salts to various types of
scrubber surfaces.
5. Test dolomitic
scaling problem is
magnesium salts.
limestones to determine if the
affected by the presence of
Corrosion
The effect of limestone injection on high-temperature
corrosion in the boiler will be studied in full-scale tests of
the dry limestone process (discussed in the dry process
report). However, low-temperature corrosion may occur in
the heat exchanger installed before the scrubber if the
limestone is added in the scrubber liquor circuit. It is
assumed that corrosion will be no worse than in operation
of a low-level economizer (22) and that use of Corten alloy
will be acceptable. Data to confirm this are needed.
Control of scrubber solution pH is necessary for
corrosion protection of the circulating system. Use of
corrosion-resistant lining will reduce the hazard of damage
to the scrubbers but pH variation could cause corrosion to
pumps and piping that would not be seriously attacked by
59
-------
the normal scrubber solution at a pH of about 6.2. Plant
operating data are needed to evaluate this potential
problem.
Oxidation
On the basis of experience reported, approximately half
of the sulfur dioxide removed in the scrubber is oxidized to
sulfate. The remainder forms calcium sulfite which could,
under some conditions, cause an unacceptable depletion of
dissolved oxygen in receiving streams. A study of methods
for promoting oxidation in the scrubber is needed, plus
determination of the effect of the more complete oxidation
on scrubber performance.
A consideration beyond the scope of the present study is
the possibility of regenerating the reaction products to
reduce limestone requirement and to recover the sulfur
value of the sulfur dioxide. NAPCA is sponsoring work in
this area. Regeneration of calcium sulfite should be more
practical than decomposition of calcium sulfate because the
energy requirement is lower. If regeneration is an objective,
a scrubbing process that would inhibit oxidation would be
desirable. Use of oxidation retardants might have merit,
particularly for the process in which limestone is added in
the scrubber circuit and calcium sulfate formation in the
boiler therefore is avoided.
Waste Disposal
It has been assumed that the general practice of sluicing
fly ash to settling ponds and overflowing clarified effluent
to a receiving stream could be used for limestone wet
scrubber effluent if desired. Further information is needed
on the composition and other properties of the slurry to
evaluate this assumption. For example, the sulfite-sulfate
distribution in the reaction products will have a major
effect on the content of dissolved solids, as the sulfates of
calcium and magnesium are about 50 and 70 times as
soluble, respectively, as the sulfites. Moreover, the particle
size distribution of the suspended solids is not well defined
and therefore settling rates cannot be accurately predicted.
Abrasive characteristics of the fly ash-lime mixture are also
uncertain. Finally, the feasibility of recycling the pond
water, which will likely be desirable in most cases and
essential in some, should be explored.
Intermittent Operation
The limestone-wet scrubbing method would be greatly
improved if it could be made more amenable to
intermittent operation. The best opportunity for this is in
an existing plant, where the existing precipitator could be
operated in series with the scrubber to remove dust so that
liquor flow to the scrubber could be shut off when sulfur
oxide recovery was not required. The feasibility of
operating the scrubber in this way should be explored.
For a new plant, where the scrubber would be designed
60
to remove both dust and sulfur oxides, liquor flow could
not be cut off. However, the lime feed could be stopped
and the dust scrubbed out with the resulting sulfurous acid.
The feasibility of this should be studied.
Nitrogen Oxide Removal
Although the main objective of limestone-wet
scrubbing is sulfur dioxide removal, dust is also removed
and if nitrogen oxide absorption could be improved enough
to give adequate removal the wet process would virtually
solve the power plant emission problem. Present
information indicates that about 20% absorption of
nitrogen oxides is all that can be expected with the current
process. It might be possible to improve absorption by
identifying the pertinent operating conditions and adjusting
them to promote nitrogen oxide removal, or by adding
another scrubbing stage in which an absorbent with higher
affinity for nitrogen oxides could be used.
Work in this area would be a logical extension of
on-going or planned research and development efforts on
use of limestone in a wet-scrubbing process.
Partial Removal of Sulfur Oxides
One of the major questions in regard to the
limestone-wet scrubbing process is how it compares
economically with the dry injection method. It is difficult
to compare the two, however, because the wet process is
quite efficient, capable of removing 95% or more of the
sulfur oxides, whereas the dry method probably cannot do
better than 50 to 60%.
Actual cost of operation obviously is not a good basis
for comparison because it does not take into account the
higher efficiency of the wet process. A more logical
approach is to design or operate the wet-process unit for
only 50 to 60% removal and compare the resulting costs.
Comparison on this basis is complicated. It should be
noted in the first place that unless full reheat is assumed the
wet process must operate at higher than the 50% efficiency
assumed for the dry process to give the same ground
concentration. For reheat to 2500 F., removal by the wet
process would have to be about 60%. However, the
technical feasibility of only 60% removal may be
questionable. A scrubber designed for low sulfur dioxide
removal would not be a good dust collector and reducing
the limestone addition to an efficient scrubber could cause
acidic conditions that would likely result in both poor
absorption and corrosion. Use of a low-efficiency scrubber
in a plant equipped with an electrostatic precipitator might
be practical, but scrubber design and operation could well
be difficult. A delicate balance would have to be
maintained to ensure the proper degree of reaction between
sulfur dioxide and lime; if much of the excess sulfur
dioxide reacted, the system would become acidic.
In a multiplant station, use of the wet-scrubbing process
-------
might permit installation of gas-cleaning facilities on only
half of the units as a means of matching the assumed 50%
removal for the dry method installed on all the units. For
example, if only one unit of a station made up of two
200-megawatt units were equipped with wet scrubbing, the
overall removal (assuming 95% removal in the wet scrubber)
would be 47.5% and cost per ton of coal would be reduced
by 40% or so as compared with the dry system. However,
this would require good mixing of the untreated and
treated plumes before they reached the ground, which
would not be assured for all atmospheric conditions.
Moreover, if the plumes did mix, the resulting single plume
should behave in accordance with the curves in Figure 3,
and the temperature drop induced by the treated plume
would have an adverse effect on the degree to which
ground-level concentration is reduced. Assuming an average
combined plume temperature of 2750 F., ground
concentration would be reduced only by about 38%.
Hence the assumed 50% sulfur dioxide removal by the
dry process, and the accompanying 50% reduction in
ground concentration, could not be duplicated in a wet
scrubbing installation merely by scrubbing the gas from
alternate boiler units. The difficulty could be avoided by
passing about 15% of the gas from the unequipped boiler to
the scrubbing system on the other boiler and sizing the
scrubber accordingly. This would bring in other problems,
and trying to evaluate them would appear to be carrying
the string of conjecture too far.
Another approach (for an existing plant) is to divide the
gas from the air heater into two streams, one (about 60% of
the total gas) passing through the scrubber and the other
through the existing precipitator. The two streams would
then be recombined and passed into the stack.9 The
estimated operating cost of such a system is $0.82 per ton
of coal (200-mw. unit) as compared with $1.05 for the dry
process.
While this is a promising potential reduction in cost, the
arrangement would require that the ash-lime mixture from
the precipitator be introduced into the recirculating
scrubber slurry (to use the umeacted lime), which could
introduce operating problems because all the fly ash would
then go into a scrubber much smaller than for the normal
design. Moreover, for a new plant the advantage of
eliminating the precipitator could not be realized.
The problem would be simpler if process B (limestone
introduced into the scrubber circuit) were used, as the ash
from the precipitator could then be sent directly to the
pond: However,' process B is more expensive than
injection-scrubbing and would require that more gas be
passed through the scrubber, the latter because sulfu~
dioxide removal in the boiler would not be obtained and
therefore the concentration in the gas stream diverted
through the precipitator would be higher.
It is concluded that partial removal of sulfur oxides by
wet scrubbing probably is feasible but that potential
problems would have to be worked out. If partial removal
were acceptable as the basis for design, a power producer
could put a scrubbing unit in parallel with an existing
precipitator to give the degree of sulfur oxide removal
required by pollution authorities. Successful development
of such an arrangement would make wet scrubbing more
economical than the dry process, even for small plants, and
therefore development work should be carried out. Pilot
plant testing may be feasible, but tests in an operating unit
would be preferable.
----------
9The precipitator and scrubber could not be operated in series, with
the gas stream split before the scrubber, because the overloaded
precipitator would not remove all the dust and therefore the
stream bypassing the scrubber would carry dust out the stack.
61
-------
Conclusions and Recommendations
The main advantages of the limestone-wet scrubbing
process are (1) capability of highly efficient removal of
both sulfur dioxide and dust, (2) low investment as
compared with processes that recover sulfur products, and
(3) since a throwaway product is produced, no involvement
of the power company in marketing problems.
Disadvantages are (1) cooling of the plume, (2) potential
contribution to water pollution problems, (3) need for
solids disposal, and (4) no return to offset increased
operating cost. Compared with the dry limestone process,
wet scrubbing is more efficient and incurs lower waste
disposal costs because of more efficient utilization of
limestone. However, investment is higher and the process is
less adaptable to intermittent operation, particularly for
new plants in which the scrubber would be the only dust
collector installed.
Investment and operating costs are slightly higher when
limestone is added directly into the scrubber circuit, mainly
because of the higher limestone requirement assumed and
the higher cost of the corrosion-resisting reheating
equipment required. Use of lime would reduce the
absorbent requirement but the cost of calcining limestone
to lime outside the boiler is prohibitive. The advantage of
introducing the absorbent into the scrubber circuit is that
potential boiler operating problems inherent in the
limestone injection method are avoided, scaling is easier to
control, and operating cost may be lower.
The adverse effect of plume cooling is difficult to
evaluate. A study of the effect of stack temperature on
plume disperson shows that a high reduction in ground-level
sulfur oxide concentration is obtained, even with a cooled
plume, if the scrubber operates efficiently and removes
most of the sulfur oxides (say, 80% or more) from the gas.
However, as the removal of sulfur oxides (or any other
pollutant) is decreased the reduction in ground
concentration drops rapidly and at about 45% removal the
ground concentration is not reduced at all (for a plume at
scrubber exit temperature); in other words, the plume
cooling completely nullifies pollutant removal in the
scrubber. This is significant mainly in regard to nitrogen
oxides, which are not removed to any great extent in the
scrubber. If the plume is not reheated, the scrubbing would,
in effect, increase the ground concentration of nitrogen
oxides.
A further consideration is that adverse effect of plume
cooling will quite likely be increased if there is any mist
present in the gas as it leaves the stack; mist can be present
because of condensation between the scrubber and stack
62
top and because of incomplete removal of scrubber spray in
the mist eliminator.
It is recommended, therefore, that the stack gas be
reheated at least enough to ensure absence of mist in the
gas leaving the stack. Whether or not reheating to a higher
temperature is justified, and what temperature level should
be selected, will depend on the particular pollution
situation involved. The high cost of reheating to higher than
about 250°F., however, would make temperatures above
this level difficult to justify.
A comparison between the wet and dry limestone
processes is difficult to make because of the lower
efficiency of the dry process, assumed to be no higher than
50 to 60%. In situations where a high degree of sulfur oxide
removal, say over 80%, is essential, only the wet method
would be suitable. In contrast, there may be many
situations in which the lower degree of removal is adequate
and either process could be used. In comparing the
economics of the two, however, it is difficult to quantify
the advantage of the higher degree of sulfur oxide removal
in the wet process. Hence the economics will first be
compared without attaching any value to degree of
absorption.
For continuous operation, wet scrubbing appears to be a
better choice than the dry process for the larger plants. For
a 1000-megawatt existing plant operating 8000 hours per
year and burning coal containing 3.5% sulfur, investment
for the dry process (200% stoichiometric limestone) would
be $3,950,000 or $3.95 per kilowatt and the operating cost
would be $2,747,900 per year or $0.91 per ton of coal
burned. Use of the wet-scrubbing system in a similar
situation (with 110% of the stoichiometric amount of
limestone injected into the boiler and the stack gas reheated
to 25(f F.) would more than double the investment
($8.21jkw.) but the total operating cost for continuous
operation would be $0.90 per ton of coal. Since capital
charges are included in the operating cost, the higher
investment is not a handicap unless the unit is operated on
an intermittent basis.
Hence wet scrubber economics, considering the other
advantages of the method, are favorable for a large plant;
for a new plant they would be even better because credit
could be taken for eliminating precipitator investment. The
disadvantages-lower stack temperature and possibly more
water pollution-are difficult to quantify. If air quality
considerations dictated full reheat to normal stack
temperature (about 30<1' F.), the overall cost would be
increased to $1.11 per ton of coal. And if it developed that
use of dolomite made pond water recycling mandatory
-------
because of stream pollution, the cost would be increased in
areas where dolomite is plentiful but high-calcium
limestone is not.
On the other hand, the cost of wet scrubbing could be
much lower than the $0.90 per ton of coal given above if
lower stack temperature were acceptable. If it is assumed
that the minimum acceptable temperature is that required
to avoid the presence of water mist in the gas leaving the
stack, and that this temperature is, say, 175" F . (the exact
temperature varies with plant design and operating
conditions), then the wet-scrubbing cost for the
1000-megawatt unit would be only $0.78 per ton of coal.
And if a shorter stack should be acceptable to pollution
control authorities, the resulting credit would reduce the
overall cost even further.
These considerations show the wide range of costs that
might be involved in applying the wet-scrubbing process to
various plants, which makes economic generalization
difficult. In contrast, the dry process is much less affected
by cost variables.
The wet process would gain in the comparison if costs
were reduced by operating at lower efficiency, on the basis
that if, say, 50% removal by the dry process were
acceptable in a particular situation, then comparable
e(ficiency should be adequate for the wet process. The
problems of partial removal were discussed in the research
and development section. Several plant arrangements can be
considered but each introduces problems. The best
approach for an existing plant appears to be operation of
the precipitatur in parallel with a scrubber designed to
handle about 60% of the gas by the injection scrubbing
method; the only apparent problem is that all the ash
would end up in the recirculating liquor of the relatively
small scrubber. This may not be a serious problem, and the
savings-$0.82 per ton of coal against $1.05 for the dry
process-are considerable.
The dry system has the advantages of simpler operation,
less tendency to operational upset, and having operating
steps that are similar to those already practiced in the
power plant (grinding, injection, dust removal). The wet
process is a somewhat complicated chemical operation,
requiring closely controlled operation of a type foreign to
power plant practice. However, these differences are not
subject to quantitative evaluation.
If intermittent operation were feasible as a means of
reducing cost, the dry system should be superior (assuming
that 50 to 60% removal is acceptable). The dry process cost
for intermittent operation (total of 30 daysfyr. in 200-mw.
unit) is $0.47 per ton of coal for all coal burned per year;
see dry process report). As noted earlier, the mechanics of
intermittent operation would present some problems in the
wet-scrubbing method. Moreover, even if technically
feasible, cost reduction for the 30-day operation would
reduce total operating cost only to $0.75.
The same reasoning applies to older plants (usually of
small size) that operate intermittently as peak-load units.
The effect of boiler load factor on cost of sulfur oxide
removal for the limestone-wet scrubbing process as
compared with the dry process is shown in Figure 26.
Because of the lower investment required, the dry process is
not penalized as much by low boiler load factor as is the
wet method. For example, the dry process increases cost of
power by about 0.41 mill per kilowatt hour at 100% boiler
load factor and by 0.56 mill at 50% load factor, an increase
of 0.15 mill between the two levels. In contrast, the
increase in cost from 100 to 50% load factor for the wet
process is about 0.28 mill.
It is concluded that both the wet and dry processes have
their place, the wet method particularly for large, new
plants and the dry technique for small, peak-load plants or
for any plant so located that intermittent operation is
acceptable. Hence the wet process becomes the basic
method for comparing with recovery processes, which also
are more appropriate for large baseload plants than for
units operated intermittently. The costs presented in this
report indicate that because of the high and increasing cost
of low-sulfur fuel, limestone-wet scrubbing will generally
be the most economical method for sulfur oxide
control-short of entering into the uncertainties and
complications of recovery. In other words, for large plants
the power producer likely will adopt wet scrubbing unless a
recovery process shows promise of being more economical,
that is, a net cost sufficiently lower to make the gamble
inherent in recovery processes acceptable.
The present study suggests that the forthcoming pilot
plant program should include study of the following:
1. Variation of absorption efficiency with limestone
type and with calcination conditions in the boiler.
2. Variation of absorption efficiency with scrubber
type.
3. Effect of particle size on absorbent utilization.
This is of particular interest for direct addition of
limestone in the scrubber circuit.
4. Use of the delay tank principle in preventing
scaling.
5. Corrosion in the system in extended operation.
6. Effect of oxidation promoters and inhibitors.
7. Chemical and physical properties of scrubber
effluent.
It is recommended also the other research and
development work suggested herein be carried out,
particularly in regard to testing plant arrangements for
intermittent operation and for partial sulfur oxide removal.
63
-------
64
1.875
1.500
~
:J:
~
--
.!!!
'E
tf 1. 125 ~ 3
o c
u :;
~ .J:J
.~ co
~ 8
cu
a.
o
....
o
c
CtJ 0
~ 0.750 -t: 2
c ~
CtJ
CtJ
...
cu
>
o
0.375
5
4
Dry process-limestone injection -+
200-mw existing unit
3.5% S in coal
Limestone addition when boiler is operating
200% stoichiometric limestone injection
for dry process
110% stoichiometric addition for wet process
~Wet process-limestone scrubbing
20 40 60 80
Boiler unit load factor (100% = 8760 hrs/yr, 200-mw load)
100
Figure 26. Effect of boiler load factor on
wet scrubbing vs dry process operating cost
-------
References
1. Tennessee Valley Authority. Sulfur Oxide Removal
from Power Plant Stack Gas: Conceptual Design and
Cost Study. 1. Sorption by Limestone or Lime: Dry
Process (1968).
2. Rees, R. .L. 1. Inst. Fuel XXV (148), pp. 350-357
(March 1953).
3. Hewson, G. W., Pearce, S. L., Pollitt, A., and Rees, R.
L. Soc. Chem. Ind. (London), Chem. Eng. Group, Proc.
~$!9.,pp. 6,7~~2- (1933):. ~~"""'-"'"""~ ,,-' ,',' n, .~}i~"i
4~ pears6iCr.'T~~onhe'beliQ~~;:;£~ e~~:~~(1 :3~:- N. ~j
5. Tennessee Valley Authoriiy~1>rogr~ss' Report Assembly
No. 47A, pp. 52-53,62-63 (February 1954-September
195 5) (unpublished).
6. Chertkov, B. A. Khim. Prom. No.7, pp. 533-536
(1962).
7. Pollock, W. A., Tomany, J. P., and Frieling, G. Mech.
Eng. 89 (8), pp. 21-25 (August 1967).
8. Plumley, A. L., Whiddon, O. D., Shutko, F. W., and
Jonakin, J. "Removal of S02 and Dust from Stack
Gases." Paper presented at American Power Confer-
ence, Chicago, Illinois, April 25-27, 1967.
9. Murakami, K., and Hori, S. "Removal of Sulfur Dioxide
in Exhaust Gas" (unpublished report).
10. Mitsubishi Heavy Industries, Ltd. (Japan). "S02
Removal Plant." Sales brochure DA-234 (1.5) (1967).
11. Jonakin, 1. Panel Discussion: Sulfur Oxide Control
Process-How Soon? American Society of Mechanical
Engineers, 89th Winter Annual Meeting and 3rd Energy
Systems Exposition, New York, New York, December
1-5,1968.
12. Nonhebel, G. Trans. Farada~so~.32,~P' 1291-129~'1
I ",0,936)."" ':':-"':'--""S:3"rc",'("""~.1"',,,,,,.,,,,,,,- --.. '
13. Lessing, R. Journal of The Society of Chemical Indus-
try, Transactions and Communications 57, pp. 373-388
Novembe J~~)~,- -~.~"-.--=~~ "
, ,~.i.~~1.,~~!. nB. A. Teploe~~rg,,,~ q~» PP'9J-~~c (-!g~71r:I",
$~-t5: Cheitkov,' B. nk Zllur:-PtfkT. ~KJi[iii':--7:J (8); - pp.
~, , '17",08-1714 (1960). --, ',,-, , ',',',"C'=,---'-' r,' "1" -'. ,--,-"--"-,,,, ',-'.,-,,',', :-4,
i -_." ,- -0=,-'-.-- ,._~~Ic'" ,"h~~~"'" ". '., . ,,' . I'~'!",,"," ""4 ,,,,' ,,'
"16. Thomas, F. W., Carpenter,S. B.~and'Cofibatigi1;W. C.'
(Air Quality Branch, Division of Health and Safety,
Tennessee Valley Authority). "Plume Rise Estimates
for Electric Generating Stations." Accepted for publica-
tion by The Royal Society, London, England.
17. Scorer, R. S. Intern. J. Air Pollution 1, pp. 198-220
(1959).
18. Scorer, R. S. Air Pollution, pp. 86-94, Pergamon Press,
London (1968).
19. Field, J. H., Brunn, L. W., Haynes, W. P., and Benson,
H. E. Bureau of Mines Report of Investigations 5469, p.
3, U.S. Department of the Interior (1959).
20. Federal Power Commission. Instructions for Estimating
Electric Power Costs and Values, p. 24, Washington, D.
C. (1960).
21. Phillips, C. F., Jr. The Economics of Regulation,
Richard D. Irwin, Inc., Homewood, Illinois (1965).
22. Wiedersum, G. C., Jr., Brockel, W. E., and Sensenbaugh,
J. D. "Corrosion and Deposits in Low-Level Econ-
omizers" (ASME Paper No. 61-WA-138). Paper pre-
sented at the Winter Annual Meeting, New York, New
York, November 26-December 1, 1961, of The
American Society of Mechanical Engineers.
65
-------
66
-------
Appendix A:
Studies of Limestone-Wet
Scrubbing
London Power Company (London, England)I,2
The London Power Company in the early 1930's
developed a process for controlling dust and sulfur oxide
emission from the Battersea power station in London.
Cyclone collectors were used to remove most of the dust
from the boiler exhaust gases. The gases were then passed
through vertical steel scrubber elements installed in the
main flue where they were contacted in transverse flow
with water from the Thames River. Slow corrosion of the
steel provided easily and relatively cheaply the iron salts
that promoted oxidation of sulfite in the effluent, which
was considered necessary to prevent depletion of dissolved
oxygen in the river. The effluent was aerated after discharge
to further promote oxidation and was finally filtered in
gravel beds before discharge back into the river.
The gases passed out of the main flue into the downtakes
which discharged into flues leading to the chimneys.
Further washing was accomplished on steel scrubber
elements installed in the down takes and in wood grid
scrubbers in the flues. At the top of the final scrubbing
unit, the gases were contacted with a suspension of finely
ground chalk. The chalk (calcium carbonate) was
satisfactory as a source of alkalinity (presumably because of
reaction with carbon dioxide in the gas to form soluble
bicarbonate), but addition of lime was said to cause scaling
by reaction with calcium bicarbonate in the river water.
Finally, the gases were passed through spray eliminators
and discharged to the chimneys.
The effluent from the second and final washing steps was
oxidized by aeration but was not filtered because all dust
was removed in the first section. The total effluent was
mixed with about 15 to 20 times its volume of condenser
cooling water before it was returned to the river.
The process removed about 90% of the sulfur dioxide in
the flue gas and virtually all of the dust.
The effectiveness of the process as a pollution control
method was a controversial matter. A downward path of
the washed gas leaving the stack was often observed
especially when the wind speed was low and the gase~
passed over a relatively cool region. The value of gas
washing was questioned since the gases reached the ground
much nearer to the chimney than under usual conditions
----------
lRees, R. L. "The Removal of Oxides of Sulfur from Flue Gases."
J. Inst. Fuel, XXV (148), pp. 350-356 (March 1953).
2Hewson, G. W., Pearce, S. L., Pollitt, A., and Rees, R. L. "The
Application to the Battersea Power Station of Researches into the
Elimination of Noxious Constituents from Flue Gases, and the
Treatment of Resulting Effluents." Soc. Chern. Ind. (London),
Chern. Eng. Group, Proc. 15, pp.67-99 (1933),
and therefore were less diluted by atmospheric turbulence.
From theoretical considerations it was concluded that at
least 80% of the sulfur dioxide would have to be removed
before maximum ground concentration near the chimney
would be reduced. At low wind speeds, observation showed
that conditions were worse than predicted.
The large amount of water required and the excessive
cooling of the gases make the process unattractive;
therefore, no further details will be presented. A detailed
account of plant design and operation is given by Hewson,
et al.
James Howden and Company - Imperial
Chemical Industries (London, England)3
Also in the early 1930's, the Howden-ICI noneffluent
lime scrubbing process was developed in England. The
process as envisioned consisted of three main
operations: lime treatment, flue gas scrubbing, and
separation of solids, The plant requirements were:
1. The lime section-comprising lime or chalk storage,
mixing apparatus, and slurry stock.
2. The scrubbing section-consisting of absorption
towers and recirculation system.
3. Solids separation section-comprising settlers and
filters (or centrifuges), facilities for returning the
clarified liquor to the liquor circulation system, and
storage for wet solids.
The major problems encountered during process design
were (1) finding a nonchoking scrubber packing with high
absorption characteristics and suitable scaling resistance,
and (2) reducing the scaling potential of calcium sulfate and
sulfite in the recirculating solution. The last of these was
the most difficult. Theoretical considerations presented by
the Howden-ICI researchers on the scaling problem follow,
Addition of lime or chalk to the circulating liquor, which
contains dissolved carbon dioxide, was said to result in
formation of soluble calcium bicarbonate, the active
absorbing agent for sulfur oxides in the wet system.
Formation of the bicarbonate from suspended calcium
carbonate is continuous because of depletion of dissolved
bicarbonate by reaction with sulfur oxides. The calcium
sulfite and sulfate formed are relatively insoluble
compounds and are precipitated from the circulating liquor.
The normal tendency is for these compounds to deposit on
absorbing surfaces in the scrubber system with resulting
scaling and plugging.
----------
3pearson, J. L., Nonhebel, G., and Ulander, P. H. N. "The Removal
of Smoke and Acid Constituents from Flue Gases by a
Non-Effluent Water Process." J. Inst. Fuel VIII (39), February
1935.
67
-------
0'\
00
DEDUSTED
GAS
NOZZLE PLATE
AIR VENT
FILM
DISTRIBUTORS
GRID PACKING
PRIMARY
ELEMENTS-
MAKEUP
PUMPING
TANK
COMPRESSED
AIR
DELAY TANK
DRAIN
SLUDGE
TA-NIC
LIME
PUMP
LIME MIXING
TANK
Figure A-1. Howden. ICI pilot plant
-------
Calcium sulfite and calcium sulfate form relatively stable
supersaturated solutions so that precipitation takes place
very slowly, particularly when the solution is only slightly
supersaturated. Because of this, it is difficult to prevent
scaling by forcing the precipitation to take place away from
the scrubber surfaces. Lessing4 found, however, that this
could be accomplished by (1) carrying calcium sulfate and
sulfite crystals in the recirculating liquor (minimum content
of 3% each) to provide a "harmless" surface for deposition,
(2) holding the slurry in a delay tank for a period to allow
dissipation of supersaturation on the recirculating crystals,
and (3) using "nonscaling" packing (suspended wood
plates) in the lower part of the scrubber.
A pilot plant was constructed at Billingham in early 1933
to study the scrubbing process. A schematic diagram of the
system is shown in figure A-I. A 4-foot-square,
grid-packed absorption tower was used to scrub a portion
of the flue gas from a pulverized fuel boiler. Operation with
and without dust remoyai prior to the scrubber was studied.
During extended tests, 97 to 99% sulfur dioxide removal
and 97 to 98% dust removal were obtained; the absorption
rate was found to be controlled almost entirely by the gas
film resistance. Design and operating data are shown in
table A-I.
Because of the high solids content (up to 20%) in the
circulating liquor it was found advisable to: (1) limit liquid
flow to 7 feet per second in mild steel pipes, (2) rubber line
pipe bends, and (3) use axial flow pumps with rubber-lined
casings and rubber-coated impellers.
Provided the pH of the liquor was maintained above a
pH of 6.2, corrosion of the scrubber shell was negligible.
Although the unit was operated almost continuously for 20
months, the ~-inch mild steel plate showed no measurable
metal loss. Mild steel packing near the gas inlet was
appreciably corroded; wood, alloy steel, and brass packings
were satisfactory.
Use of the delay tank (2.5 min. retention) plus carrying
crystals in suspension virtually eliminated scaling. After
3,500 hours of operation, the packing remained as clean as
when it was new.
Lime was used in the first tests but calcium carbonate
was used for over 1,000 hours in the later tests. About 30%
excess calcium oxide was required when calcium carbonate
(finely ground chalk) was used; with lime, the
stoichiometric amount was adequate. Precipitated calcium
carbonate was also a satisfactory absorbent. The lime or
chalk was added continuously to a line just ahead of the
delay tank. The increase in pH from 6.2 to 6.8 reduced the
solubility of calcium sulfite and sulfate and promoted rapid
----------
4Lessing, R. "The Development of u Process of Flue Gas Washing
Without Effluent." J. Soc. Chern. Ind., Trans. and Com., pp.
373-388 (November 1938).
crystallization. Tests showed that, at 122° F., the solubility
of calcium sulfite is 0.637 grain equivalents per gallon at pH
6.2 and 0.518 at 6.8.
Table A-1. Design and Operating Data for the Howden-ICI
~~1f@~____-------------------------
Packing
No. of primary elements, 3.7 x 2.88 ft
No. of wooden grids, 1 in. deep, '% in. pitch, 4 ft z
Depth of grid packing, ft
Total surface area of grid packing, ftz
Total surface area of grid packing, plus
primary elements, ftz
Gas velocities
At inlet into scrubber (220° F.), ft/sec
Immediately under packing (220° F.), ft/sec
At exit through spray eliminators
(120° F.), ft/sec
Pressure drop over whole tower, in. water ga.
Pressure drop over packing only, in. water ga.
Analysis of coal fired, %
Ash
S
Operating conditions (pulverized fuel firing)
Equivalent rate of coal fired at 12%% COz Ib/hr
Effective sulfur in coal fired, %
Maximum gas rate, c.f.m. at N.T.P.
Circulating water rate, g.p.m.
Delay time, min.
COz in flue gas, %
Inlet gas temperature, of.
Exit gas temperature and circulating
liquor temperature, of.
Typical concentration of suspended solids, %
CaS04,2HzO
CaS03'2HzO
CaC03
Ash
Total
Alkalinity of liquor to methyl orange at
scrubber exit (pH 4.5), gr. equiv./gal
pH at scrubber exit hopper
pH at scrubber inlet (head tank)
Lime addition
Chalk addition
Gas analysis, gr./ft3 at N.T.P.
Inlet
Dust
Total sulfur
Outlet
Dust
Total sulfur
Alkali consumption, maximum % of theoretical
-~Q~~~lQL~OJYE~~~i______---------------]~
13
41
3.5
1,820
2,098
13.8
4.5
5.9
1.2
0.8
13
1.7 -2.0
1,100
1.7 -2.0
3,150
260-440
2.5-3.5
12-14
220
118
3.2
2.9
0.4
7.5
14.0
0.2
6.1
6.9
6.5
6-13
0.5-0.8
0.1-0.2
0.01-0.02
69
-------
i i
a::
o
! ::)
o
a:: :J
IIJ a..
....I ::)
IIJ 0 ~
)II::
a.. CD - a: IIJ a:
- - - 0 r
IIJ
-------
On shutting down, it was necessary to wash the packing
to prevent crystal formation on the packing. If through
maloperation scaling occurred, chemical cleaning could be
accomplished by substituting sodium carbonate for calcium
carbonate. The sodium salt reacted with calcium sulfate and
sulfite to form calcium carbonate which was dissolved by
carbon dioxide in the circulating liquor.
The application of the scrubbing process to a
commercial plant, the Fulham power station, was described
by Parker and Clarke.5 A schematic diagram of the
istallation is shown in figure A-2. Design conditions are
given in table A-2.
Efficiency of the plant was quite good. The sulfur
content of the exit gas was as low as 0.006 grain of sulfur
per cubic foot (99+% removal) compared with the
permissible value of 0.03. Only mere traces of dust (less
than 5 microns) were discharged to the atmosphere. Use of
the delay tank technique was completely successful in
controlling scale formation.
Several maintenance problems developed. Rubber lining
was difficult to maintain, particularly with lined pumps.
When unlined mild steel pumps were used corrosion and
erosion were serious. Use of pumps with shrouded, chrome
steel impellers reduced the problem. The original wood
packing was reinforced by brass strips and it was found that
these brass strips were attacked by dissolved oxygen in the
liquor and that dissolved zinc and copper acted as a catalyst
to oxidation in the system. This produced corrosion in
various parts of the system, and particularly on the mild
steel washer tanks. The dissolved zinc and copper also
caused pump corrosion through galvanic action.
Elimination of the brass strips reduced corrosion to about
one-third the original value. Addition of an organic
inhibitor (molasses) further reduced the problem.
Tennessee Valley Authority
(Muscle Shoals, Alabama)6
During early work (1953) by TV A on recovery of sulfur
dioxide from power plant stack gases, a pilot plant was
constructed, primarily to study an ammonia-scrubbing
process. The plant was designed to scrub about 350
standard cubic feet per minute of gas produced by burning
approximately 2 pounds per minute of pulverized coal
containing 3.5 to 4.0% sulfur; sulfur dioxide concentration
in the gas was about 0.3%. The scrubber was constructed of
mild steel and was 2 feet in diameter and packed with 8
feet of 2-inch ceramic Raschig rings. Gas-liquid flow was
countercurrent with recirculated liquid introduced through
----------
Sparker, W. C, and Clarke, H. J. 1l1st. Civ. El1grs. 9, pp. 28-36
(1937-1938).
6 Tennessee Valley Authority. Progress Report Assembly No. 47 A,
February 1954 to September 1955, pp. 72-74 (unpublished).
a distributor just above the packing. A water spray chamber
was provided ahead of the scrubber for cooling of the gas
by humidification from 3000 to 1250 F.
Late in the pilot plant program a test was made for the
purpose of investigating the use of a slurry of limestone for
removing the sulfur dioxide from combustion gases. The
first material tested was a precipitated calcium carbonate
(whiting). With slurries containing 10 to 20% solids, less
than 25% of the sulfur dioxide was removed even with 50%
excess calcium carbonate. The pH of the slurry decreased
rapidly from an initial value of 7.3 to about 4 despite the
presence of excess absorbent.
With a natural limestone (45.5% CaO) from the middle
Tennessee area, a 10% slurry of 95% minus 100-mesh
particles removed 89% of the sulfur dioxide. Makeup
limestone was added in the stoichiometric quantity
Table A-2. Operating and Design Data from Fulham
~~!~~Q~~~B~~~~~_~jt_-------------------
Coal fired (calorific value approx.
10,900 Btu/lb)
Gas circulated
CO2 content of gas
Sulfur content of coal
Weight of liquor in circulation
Weight of ash and unburnt carbon caught
in the plant
Solids formed, assuming 50% oxidation
Total solids in suspension, with 5%
suspended gypsum in recirculating liquor
Rate of circulation of washing liquor
Temperature of flue gases entering
the plant
Temperature of flue gases leaving
the plant
Humidity of gases leaving the washer
Makeup water required
Rate of flow of liquor to the
settling plant
Rate of flow of returned or clarified
liquor from the settling plant
Complete weight of one unit in working
order, exclusive of delay pipes
Total cross section of a washer
Number of cells/boiler
Maximum permissible gas flow through one
scrubber at 2400 F.
Total capacity of the delay tanks
Delay period
Final volume of gas/boiler leaving the
washer at 1240 F.
Lime requirements, assuming an excess
-~J~~~~~~D~~~~i~L~2~~~~_____LL@J~B~___-
31,050 Ib/hr
118,900 cfm
12.5%
1.7%
200 tons, approx.
781 Ib/hr
3,685 Ib/hr
13%
11,500 gpm
2400 F.
1240 F.
100%
32.7 gpm
496 gpm
445 gpm
300 to ns
400 ft2
16
118,900 cfm
6,300 ft3
3.5 min
105,000 cfm
71
-------
required to react with all of the sulfur dioxide in the gas.
Results from tests with this material are shown below:
fi.!!~r:.c!lI!l1il!nJ_a11!.
9!11J}T)i!, G!llJtlT)i!'l {ft~ l
5 1.6
10 3.2
p!t
6.5
6.6
~Q2_r~f!1~v_a~ ~
81.5
89.0
The above tests were made with gas that had been
cooled by humidification to about 1250 F. In tests without
cooling, the gas entered the scrubber at about 3000 F. and
removal of sulfur dioxide decreased to 54%; pH of the
slurry was 5.0. Use of large excesses of limestone did not
result in improved absorption nor increased pH. The slurry
temperature was only 50 higher (1300 vs 1250) than when
humidified gas was used and, therefore, probably was not
responible for the poor removal of sulfur dioxide. When
the humidifier was used, most of the sulfur trioxide was
removed in the humidifier effluent water.
In additional tests with redwood grids rather than rings
as packing in the scrubber, two additional limestones and
one dolomite were tested; the supply of stone used in the
earlier tests had been nearly exhausted. None of these
materials were satisfactory even when used in excess. Sulfur
dioxide removal efficiency varied from 6 to 15% and pH of
the slurry in each case was about 4.5.
A series of bench-scale exploratory tests was made in an
effort to determine why the limestones used in the last
pilot plant tests were ineffective in removing sulfur dioxide.
Measurements were made of the amount of sulfur dioxide
removed from simulated flue gas when the gas was bubbled
through 10% slurries of various stones. The different
limestones varied in reactivity but the stone which gave
good results in the early pilot plant tests was not
appreciably more reactive than the other types tested. It
was concluded that absorption was more closely related to
the ratio of gas rate and liquid volumes than to the type of
limestone used.
The low-priority, exploratory program on limestone
scrubbing was terminated so that work could continue on
the more promising ammonia process.
Scientific Research Institute
(Soviet Union) 7
In studies by Chertkov and coworkers at the Scientific
Research Institute in Russia, mass transfer coefficients
during absorption of sulfur dioxide from gases by lime
suspensions were calculated from data obtained during tests
in a pilot-scale (2,500 c.f.m.) grid-packed absorber.
----------
7 Chertkov, B. A. "Coefficients of Mass Transfer During Absorption
of S02 from Gases by Lime Suspension." Khim. Prom. No.7, pp.
533-36 (1962).
72
Absorption from gases containing 0.2 to 0.3% sulfur
dioxide and about 6.5 grains per cubic foot of fly ash varied
from 96 to 99% when the pH of the circulating liquor was
held at 6.1 to 6.2. The mass transfer coefficient was
directly proportional to the gas velocity for these
conditions, under which the equilibrium vapor pressure of
sulfur dioxide was nearly zero at the scrubber outlet.
With decrease in pH the equilibrium pressure increased
sharply from zero at pH 8.0 (pure calcium sulfite) to about
2.3 millimeters Hg, the same as the inlet gas, at a pH of
about 4.5. The mass transfer coefficient decreased from
150 to 120 pounds per square foot per hour per
atmosphere as the pH of the absorbing solution decreased
from 6.3 to 5.3; gas velocity was held constant. At pH
values higher than 6, the liquid phase resistance was
practically zero and the general coefficient of mass transfer
could be equated to the partial coefficient of the gas phase.
The effect of solids content of the scrubbing solution on
the transfer coefficient at various gas flow rates is shown in
figure A-3. The strong influence of solids content was in
part a function of slurry viscosity but also was related to
the type of packing. Use of finely divided packing instead
of the grid type used would have reduced the effect of
solids on the transfer coefficient.
Wisconsin Electric Power Company and Universal
Oil Products, Inc., Air Correction Division8
A study of reducing sulfur dioxide emissions from power
plant stack gas by lime scrubbing was initiated by Wisconsin
Electric Power Company in 1964; the Air Correction
Division of Universal Oil Products Company, the
manufacturer of the Turbulent Contact Absorber (TCA)
used in the project, contributed to the program.
A pilot TCA scrubber was installed on a 120-megawatt
boiler (Combustion Engineering design) at Wisconsin
Electric's Oak Creek station. The scrubber, constructed of
fiber glass-reinforced polyester, was 14 inches in diameter
and ~5 feet in overall height with provision for four
scrubbing sections each containing a bed of
1 Yz-inch-diameter hollow polypropylene spheres. Gas
entered the bottom of the scrubber and slurry was sprayed
into the top. A spray nozzle was installed near the gas inlet
to adiabatically cool the incoming gas and to assist in
removal of fly ash. A chevron-type mist eliminator was
provided at the top of the scrubber. A diagram of the pilot
plant equipment is shown in figure A-4.
The absorbent slurry was prepared in a feed tank and
metered through a rotameter to the slurry recirculation
pump. Dust loadings, sulfur dioxide samples, and pitot tube
traverses were taken at inlet and outlet sample points.
----------
8pollock, W. A., Tomany, J. P., and Frieling, G. "Flue-Gas
Scrubber." Mech. Eng. 89 (8), pp. 21-25 (August 1967).
-------
E
! 120
...:
~
~
-
C"
~
--
.ri
...; 80
c:
Q)
'<:)
iE
Q)
o
u
.....
Q)
-
II>
c:
~ 40
II>
II>
co
~
160
28%
32%
2
Gas velocity, ft/sec
3
Figure A-3. Effect of suspended solids in lime slurry
00 m... "'0"., ,oe!!;,;.o' 10' SO, ''''O'P~ .jf~ ~O; ~7 tJ~
IjJ~~ a~ .!JI~1 ~i ~
6>' 9(r ~
The majority of tests for sulfur dioxide absorption were
made with two stages of scrubbing. Both sodium and
calcium carbonate (limestone) were used as absorbents.
The absorbent efficiency with sodium carbonate ranged
from 89.6 to 94.4% as the liquid rate was increased from 20
to 40 gallons per minute per square foot, at a constant gas
rate of 750 feet per minute. Efficiencies with limestone
were 79.8 and 83.6% at 30 and 50 gallons per minute per
square foot, respectively, at a gas rate of 750 feet per
minute. With one stage of scrubbing, about 20% of the
nitrogen oxides was removed; efficiency was independent
of the absorbent used.
Inlet dust loadings ranged from 1.5 to 2.5 grains per
standard cubic foot. At a gas velocity of 920 feet per
minute with two-stage operation, 96.9, 98.9, and 99.5% of
the dust was removed at liquid rates of 20, 25, and 30
gallons per minute per square foot, respectively. With one
scrubber stage, dust collection efficiency was about 5%
lower than with two stages under comparable conditions.
Pressure drop was approximately 3 inches of water with
one stage and 6 inches with two stages, at a liquid rate of
40 and a gas rate of 750.
During operation of the pilot scrubber, inspection of the
interior surfaces indicated only a very thin film of calcium
carbonate and sulfate which could be easily wiped off. No
problems with plugging were encountered. The spheres
showed no signs of attrition or excessive wear after more
than 500 hours of operation.
The process visualized in these tests included injection of
limestone into the boiler to calcine the absorbent and
increase its reactivity. The injection step was tested but not
at the Oak Creek station where the scrubber tests were
made.
73
-------
RETAINING
GRIND
SPHERES
SUPPORT
GRID
74
INLET
GAS DUCT
--..
FAN
SAMPLING
STATION
DE-ENTRAINMENT
SECTION
OBSERVATION
PORT
" FOUR STAGE 0
" PILOT TCA
'"
"
"
"
"
.........
SPRAY
-.......... NOZZL
.........
..........
ROTAMETER %
MAKEUP
WATER
TO
WASTE
SURGE
TANK
SUMP
+
en
a:::
iaJ
.....
iaJ
~
o
z
~
~
Figure A-4. Scrubber arrangement in Wisconsin power tests
CHEMICAL
FEED
TANK
ROTAMETER
RECIRCULATION
PUMP
-------
100
~ 90
o
E
Q)
....
N
o
(/)
?f!. 80
5
6
-
Recycle
water rate
6 25 gpm
o 20 gpm
o 15 gpm
7
pH
8
9
Figure A-5. Percent 802 removal vs pH of scrubber effluent
Combustion Engineering, Inc.,
and Detroit Edison Company, 9,10
A system for removal of fly ash and sulfur dioxide from
power plant stack gas by lime scrubbing has been studied
by Combustion Engineering and Detroit Edison; the work
was reported in early 1967. The process is similar to that
tested by Wisconsin Electric Power; limestone or dolomite
is injected into the boiler to improve absorbent efficiency
by calcination, followed by wet scrubbing to remove ash
and lime from the gas and to promote reaction of sulfur
oxides with the lime.
A small pilot plant was first constructed and tested.
Controlled amounts of sulfur dioxide, calcined or
uncalcined dolomite, and fly ash were added to flue gas (at
2500 F.) from a natural gas-fired boiler and the mixture
(1,000 c.f.m.) was passed through a scrubber. With a dust
loading of 6.6 grains per standard cubic foot, a spray water
rate of 20 gallons per minute resulted in 98%
dust-collection efficiency and 99% sulfur dioxide removal
at a pH of 9.2 (125% stoichiometric calcined dolomite).
Only about half this efficiency was obtained with raw
dolomite.
The next phase of the study was conducted jointly by
Combustion Engineering and Detroit Edison at the latter's
----------
9Plumley, A. L., Whiddon, O. D., Shutko, F. W., and Jonakin, J.
"Removal of S02 and Dust from Stack Gases." Paper presented
at American Power Conference, Chicago, Illinois, April 25-27,
1967.
10Plumley, A. L., Jonakin, J., Martin, J. R., and Singer, J. G.
"Removal of S02 and Dust from Stack Gases." Combustion 40
(1), pp. 16-23 (July 1968). '
St. Clair power plant. The objective was to determine if the
pilot unit data could be conflImed on a commercial-size
unit. Dolomite was introduced on a full-scale basis to one
furnace of a 325-megawatt twin furnace unit (tangentially
fired). The dolomite was ground in one of the coal
pulverizers and injected through the top burners of the
boiler.
Approximately 3,000 actual cubic feet per minute of
exhaust gas (about 1 % of total) was withdrawn after the air
heater and passed through a fluid bed (glass marble)
scrubber (National Dust Collector Hydro-Filter). The gas
flowed upward together with a spray of recirculated slurry
through the marble bed. The clean gas passed through a
demister to the stack and the slurry drained back through
the bed and to a settling tank.
Continuous tests were made for periods ranging up to 5
days. No scaling or plugging problems were reported. Sulfur
dioxide removal efficiency dropped 2 to 5% when the angle
of injection into the boiler was changed so as to expose the
dolomite to higher temperature. Efficiency increased from
about 90% to 98% as the liquid to gas ratio was increased
from 7 to 12 gallons per 1,000 cubic feet. The effect of
scrubber effluent pH is shown in figure A-5; a pH of at least
6 was required for good absorption. It was found that the
pH increased when the effluent was held in a mixing tank
for a period prior to discharge to the settling tank. The
sludge from the settling tank contained about equal
amounts of calcium sulfate and calcium sulfite and the
liquid phase was nearly saturated with magnesium sulfate.
About 20% of the nitrogen oxides in the flue gas was
removed in the scrubber.
From a tabulation of data from various runs, optimum
75
-------
conditions were: scrubber pH, 6.5; slurry rate, 7.6 gallons
per 1,000 cubic feet; gas velocity, 530 feet per minute;
pressure drop, 6 inches of water; dolomite feed, 11 0% of
stoichiometric. Under these conditions, the system removed
98% of the sulfur dioxide and 99.5% of the dust. A
schematic diagram of the injection-scrubbing system is
shown in figure A-6. (Although a gas reheating system is
shown, this equipment was not included in the test
facility.)
In a later paper by Plumley et al. (Combustion
Engineering), commercial installations of the scrubbing
system are described. The first unit was installed at Unit
No.2 in the Union Electric Company's Meramec station
(St. Louis) and was started up in November 1968. The
140-megawatt boiler on which the unit was installed is
tangentially fired and burns pulverized rnidwestern
bituminous coal containing about 3.4% sulfur and 9% ash.
The scrubbers were installed above the precipitators, which
were blanked off and bypassed. Limestone is fed to an
existing mill at a rate equivalent to 110% of stoichiometric.
A vertical baffle was installed in the bunker to provide
storage space for limestone. Induced-draft fans were
modified for increased draft loss by installing new wheels to
operate at 880 r.p.m. instead of 700 r.p.m.; larger motors
were required.
The process is also being installed at the Lawrence,
Kansas, plant of the Kansas Power and Light Company, on
FURNACE
COAL
! ADDITIVE
r ,
I I
I
I I
l J
, ,
y
AIR
HEATER
MILL
Units 4 and 5. Unit 4, a 125-megawatt tangentially fired
boiler, has been operating since 1960; the scrubber system
was completed late in 1968. Unit 5 is a 420-megawatt
installation now under construction, with operation
planned for early 1971.
Tests were conducted at the Lawrence Unit 4 to
determine the feasibility of combining limestone with coal
prior to grinding. Results showed that:
1. Coal and limestone segregated less than 5% in the
bunker.
2. Pulverizer power requirement for the mixture was
about the same as for coal.
3. Furnace operation was improved by firing the mix.
4. More than 30% of the sulfur was removed by
in-furnace reactions and by pyrite rejection.
The limestone and coal will be premised in Unit 4 system
but Unit 5 will be fitted with separate injection facilities. A
system flow diagram for Unit 5 is shown in Figure A-7.
The Meramec and Lawrence installations are guaranteed
to remove 99% of the dust and to reduce sulfur dioxide to a
level equivalent to burning coal containing 0.5% sulfur. The
Meramec unit has separate additive feed, uses a
liquid-couple stack gas reheating circuit, and has a clarifier
for concentration of solids. The two Kansas Power and
Light units will have hot water reheaters and the fly ash
pond will be used for slurry clarification.
TO STACK
STACK GAS
REHEAT
SYSTEM
RECYCLE
AND
MAKEUP
WATER
TO
DISPOSAL
Figure A-G. Sulfur oxide removal unit tested by
Combustion Engineering and Detroit Edison
76
-------
STACK
I.D. FAN
STACK GAS
REHEATER
....-
~
SETTLING
POND
,.-r
/
I. D. FAN
STACK GAS SCRUBBER
-....]
-....]
Figure A-7. Lime scrubbing system to be installed on
boi(er unit 5 of Kansas Power and Light Company
-------
Appendix B:
Water Quality Considerations
The feasibility of releasing dissolved or suspended
calcium salts into watercourses depends on how much the
content of these impurities in the water course is increased,
which in turn depends on the amount of the particular
pollutant released in the effluent, the content in the
watercourse before influx of the effluent, and the water
volume flow of the watercourse. Various regulatory
agencies and advisory groups have set or recommended
limits above which the concentration of an impurity in a
watercourse should not be increased. Such limits for the
solids and dissolved salts released from a limestone-wet
scrubbing process are summarized herein. Also presented is
an analysis, based on these limits, of the degree of pollution
by limestone-wet scrubbing; for this it is assumed that the
power plant involved is a 200-megawatt unit, the coal
contains 3.5% sulfur, the scrubber is designed for 95%
removal of sulfur oxides, the limestone contains no
magnesium, and 110% of the stoichiometric amount of
absorbent is used. The following quantities are involved.
Total solids to pond
CaS04
CaS03
CaC03
Inert
Fly ash
Water of hydration
48,800 tons/yr
27 ,400
6,720
3,860
52,000
_2J1J~[
160,000 tons/yr
or 40,000Ib/hr
Free water (10% slurry) 360,000Ib/hr
Dissolved solids to watercourse
CaS04
CaS03
CaC03
890 Ib/hr
14
7
1. Suspended solids
Most regulatory agencies do not place quantitive
limits on either suspended solids or turbidity for raw
water used as a source for domestic water, but do
generally specify that concentrations shall not be
sufficiently high as to be objectionable or interfere
ith normal treatment processes. Both turbidity and
suspended solids can be reduced to acceptable limits
for most purposes by conventional treatment
methods; however, high concentrations increase the
cost of processing the water by increasing chemical
requirements and the volume of sludges to be
disposed of. In some cases head losses through the
filters are increased with a resulting increase in the
frequency of filter backwashing.
78
Settleable suspended solids often blanket stream
bottoms, killing fish eggs and young and destroying
much of the normal benthic biota which serve as food
organisms required for propagation and growth of
fish. Turbidity in streams also interferes with light
penetration and limits or diminishes the
photosynthetic effects necessary for the primary
productivity of fish-food organisms. The Committee
on Water Quality Criteria of the Federal Water
Pollution Control Administrationl recommends a
limit of 50 Jackson Turbidity units in warm water
streams. TV A biological surveys indicate that ash
discharge from power plant ash-settling ponds is not
sufficient to adversely affect the normal benthic
environment of the river channel or otherwise cause
water quality problems. Whether or not this situation
can be maintained with limestone-wet scrubbing will
depend on settling characteristics of the lime-calcium
sulfate-ash slurry and the degree to which pond area is
increased to handle the increased solids load.
2. Total dissolved solids (2,600 mg.!l. in effluent from
200-mw. unit, from quantities given in above
tabulation)
The following information is quoted from Water
Quality Criteria, issued by the Resources Agency of
California?
(a) Domestic Water Supplies. The 1962 USPHS
Drinking Water Standards specify that the total
dissolved solids should not exceed 500 mg./l. if
more suitable supplies are, or can be made
available. This limit was set primarily on the basis
of taste thresholds. The 1958 WHO International
Standards set the "permissible limit" at 500 mg./l.
and the "excessive limit" at 1500 mg.fl., but no
"maximum allowable limit" is given. The 1961
WHO European Standards do not include limits
for total dissolved solids. It is generally agreed that
the salt concentration of good, palatable water
should not exceed 500 mg./I.; however, higher
concentrations may be consumed without harmful
physiological effects and may indeed even be more
beneficial. Each water with a total salt
--------
IFederal Water Pollution Control Administration. Water Quality
Criteria (Report of the National Technical Advisory Committee to
the Secretary of the Interior). U.S. Government Printing Office,
Washington, D. C. (April 1, 1968).
2McKee, J. E., and Wolf, H. W. Water Quality Oiteria. The
Resources Agency of California, State Water Quality Control
Board, Publication No. 3-A (1963).
-------
concentration over 1000 mg./l. should be judged
on the basis of the local situation, alternative
supplies, and the reaction of the local population.
(e) Fish and Aquatic Life. It has been reported that
among inland waters in the United States
supporting a good mixed fish fauna, about 5%
have a dissolved-solids concentration under 72
mg./l.; about 50% under 169 mg./l., and about
95% under 400 mg./l.
Water quality objectives adopted by TV A for
surface water in the Tennessee River basin include
a limit of 500 milligrams per liter of dissolved
solids, which would require a dilution of about 5: 1
for the 2600 milligrams per liter of dissolved solids
from the 200-megawatt unit-easily attainable
with even a relatively small receiving stream for
the effluent flow involved (about 1.5 cu. ft./sec.
for the 200-mw. unit). For larger power plants, say
a total of 2000 megawatts at one site, the required
dilution of 50: 1 would also be practicable in most
situations. The flow of diluting water required
would be 75 cubic feet per second, in comparison,
the approximate average flow past various TV A
power plants is 51,000 cubic feet per second for
the Colbert plant (on the Tennessee River), 18,000
cubic feet per second at Gallatin (Cumberland
River), and 4,300 at Bull Run (Clinch River).
3. pH (no lower than 6.2)
TV A has proposed upper and lower pH limits for the
waters of the Tennessee Valley of 8.5 and 6.5,
respectively. The pH values required for industrial
water supplies are covered in some detail in the
FWPCA Water Quality Criteria report. The proposed
TV A criteria should be adequate for all of these but a
few extreme cases. For limestone-wet scrubbing, the
scrubber effluent should have a pH no lower than 6.2
for good scrubber operation; since the pH of
watercourses normally is considerably higher than this
(7.5 in the Tennessee River), dilution should be
adequate, even by small watercourses, to reach the
minimum pH level of 6.5. Any detrimental effects of
pH should be confined to a small mixing zone in the
vicinity of the waste outfall.
4. Calcium (Ca++, 756 mg.!1.; 1,860 mg.!1. expressed as
Caco3 hardness)
Calcium in water supplies has been suspected of
producing undesirable physiological reactions but no
definite relationship has yet been established. The
major concern is related more to its deleterious effects
in uses such as washing, bathing, and laundering and
to the problem of incrustations on cooking utensils
and water heaters. Some industrial processes require a
very soft or low calcium water. Costs of conditioning
water for such industries is directly related to source
concentration.
Thus the major detrimental effect of calcium
hardness is economic. TV A has proposed the
following limitation on water hardness:
There shall be no substance added to the waters
that will increase the hardness to such an extent
as to appreciably impair usefulness as a source
of water supply or interfere with other
reasonable and necessary uses of the water.
Rivers in the Tennessee Valley on which steam plants
are located generally range from 50 to 125 milligrams
per liter hardness as calcium carbonate. In general, a
concentration of 100 milligrams per liter would not
be objectionable for most uses if this concentration
were not significantly greater than the normal range
for the area. Thus with a background concentration of
88 milligrams per liter hardness (median between 50
and 150) and an effluent concentration of 1,860
milligrams per liter, a dilution ratio of about 150: 1
would be required for a 100-milligram-per-liter limit.
This is quite easily obtained for the 200-megawatt
plant. However, other sources of hardness (such as fly
ash) could elevate the effluent concentration.
Moreover, a 2,000-megawatt plant would require a
2,250-cubic-foot-per-second flow of diluting water, or
over 50% of the average (Clinch River) flow. This
would likely require a pipeline and possibly a diffuser
into the river to prevent delay in mixing and
consequent high concentration along the near river
bank for some distance downstream.
5. Calcium Sulfate (2,500 mg.!!.)
No limits have been placed on calcium sulfate by
regulatory agencies. The following summary of
beneficial uses and reported acceptable concentrations
is quoted from McKee and Wolf.
(a) Domestic Water Supplies. High calcium sulfate
concentrations in water are disadvantageous for
most household uses, but for drinking purposes,
300 mg/l or more is not harmful. The taste
threshold of calcium sulfate has been reported to
be 250 to 900 mg/l. Calcium sulfate significantly
increases plumbosolvency; raising the calcium
sulfate concentration from 25 to 250 mg/l
increased by 10 percent the amount of lead
dissolved from lead pipes.
(b) In general, calcium sulfate is beneficial in the
brewing industry, for it helps to maintain the
acidity of the wort and therefore causes more
complete coagulation of albuminous matter. It
also reduces the solubility of the bitter substances
of the hop.
79
-------
(c) Irrigation. Gypsum in irrigation waters improves or
restores the permeability and tilth of soil having an
unfavorable sodium ratio.
(d) Stock and Wildlife Watering. A saturated solution
of calcium sulfate, used as the sole source of
drinking water for rats, permitted satisfactory
growth. A concentration of 2400 mg/l of calcium
sulfate permitted normal growth and reproduction
among rats.
(e) Fish and Other Aquatic Life. Trama reported that
a saturated solution of calcium sulfate at 20° C did
not produce significant mortalities among
bluegills; but a later report from the same
organization indicated that this concentration was
the 96-hour TLm at 18-20° C in soft water for the
bluegill sunfish. It was also reported that 3200
mg/l caused a 50 percent reduction in the rate of
growth of the diatom, Navicula seminulum. Using
highly turbid water at 21-25° C and the
mosquitofish (Gambusia affinis) as the test animal,
Wallen et al indicated that the 96-hour TLm was
greater than 56,000 mg/l, a concentration much
larger than the solubility.
TV A has proposed no specific limits for calcium
sulfate; however, the concentration is limited by the
allowable sulfate concentration and the permissible
added calcium hardness.
6- Sulfates (SO:, 1,770 my.!1.)
McKee and Wolf discuss sulfate content as follows.
(a) Domestic Water Supplies. The 1962 Drinking
Water Standards of the USPHS recommend that
sulfates not exceed 250 mg/l, except where a more
suitable supply is not available. This limit does not
appear to be based on taste or physiological effects
other than a laxative action toward new users.
Public water supplies with sulfate contents above
this limit are commonly and constantly used
without adverse effects. The 1958 WHO
International Standards established a "permissible
limit" of 200 mg/l and an "excessive limit" of 400
mg/l, but set no maximum allowable limit. The
1961 WHO European Standards include a
recommended limit of 250 mg/l for sulfates.
Korotchenok reported on heavily mineralized
drinking waters in western Turkmenia (USSR),
noting that no outbreaks of disease have been
ascribed to waters in which the content of sulfate
did not exceed 1295 mg/l. In British Somaliland
well waters used for human consymption contain
very high sulfates, many having 2000-3000 mg/l.
One village is using water of 4400 mg/l sulfates. A
survey in North Dakota indicated that water
containing less than 600 mg/l of sulfates is usually
80
safe. In reviewing the literature, Moore quotes
Sollman to the effect that concentration of 1000
mg/l of sulfates in water is harmless. A cathartic
dose is 1.0 to 2.0 grams, or a liter of water
containing 1000 to 2000 mg/l of sulfates.
Sulfates appear to have no detrimental effect
on the corrosion of brass fittings in domestic water
systems, nor do concentrations less than 200 mg/l
increase plumbosolvency.
Whipple is quoted by Moore to the effect that
the taste thresholds of sulfate salts were as
follows:
Sodium sulfate
Calcium sulfate
Magnesium sulfate
200-500 mg/1
250-900 mg/1
400.600 mg/1
(b) Industrial Water Supplies. Limiting or threshold
concentrations, or optimum ranges, for sulfates are
assembled below:
_JJ~l!~I!!-I~.Q.~!!I.!!I}!I.!!.~!!!Il~~.!!lJlL!_-
-~-p~ -~]Q4... _M.!l~..o~ - 1J~]Q4
100-500 100-200 100
!!I.!!!!s.!.I'l!I1J!.l'.Qf!J~
Brewing
Carb beverages 250
Concrete corrosion 25
Ice making
Milk industry
Photo processes
Sugar making 20
I~~~~_______l~~____-----------------
300
130-300
300
60
100
TV A has proposed a limit of 150 milligrams per liter
for sulfates to provide a reasonably satisfactory water
for both domestic and industrial use. Normal water
treatment has little effect on sulfate concentration.
Assuming a normal background of 15 milligrams per
liter sulfate and an effluent of 1,800 milligrams per
liter, dilution requirement would be approximately 12
to 1. Again this should be no problem on major
streams.
7. Sulfite (SO~, 27 my.!1.)
The major concern in regard to high sulfite
concentration is oxidation to sulfate, thus lowering
the dissolved oxygen content of the receiving water. If
we assume 95% saturation of the effluent 8Vith
dissolved oxygen, a temperature of 85° F (29.4° C),
and complete conversion of sulfite to sulfate prior to
discharge to the receiving stream, the dissolved
oxygen content of the effluent would be reduced to
1.6 milligrams per liter. Assuming 95% of saturation
in a 77° F receiving stream (7 mg./l. of dissolved
oxygen), no significant reduction in the dissolved
oxygen of the stream would occur and a dilution ratio
of only 1.7: 1 would be required to maintain a
-------
minimum dissolved oxygen content of 5.0 milligrams
per liter. No problem should arise from such
discharges.
As noted earlier, this treatment of the dissolved solids
problem is based on the assumed use of calcitic limestone.
There may be also a considerable dissolved content of
magnesium sulfite and magnesium sulfate if dolomite is
used as the absorbent. Although data indicate little or no
reaction of magnesium oxide in the boiler, reaction can
occur in the scrubber and also in the disposal pond. Tests at
TV A showed that solid magnesium oxide in the pond could
react with calcium sulfate and atmospheric carbon dioxide
to form magnesium sulfate. In tests with a mixture of 20
grams MgO, 80 grams CaS04 '2H2 0, and 300 milliliters H20,
contact with carbon dioxide resulted in 70% dissolution of
the magnesium and 65% of the sulfate.
The effect of dissolved magnesium has been treated by
McKee and Wolf.
(a) Domestic Water Supplies. Magnesium is an essential
mineral element for human beings; the daily
requirement of magnesium is about 0.7 gram.
Magnesium is considered relatively non-toxic to man
and not a public health hazard because, before toxic
concentrations are reached in water, the taste
becomes quite unpleasant. At high concentrations,
magnesium salts have a laxative effect, particularly
upon new users, although the human body can
develop a tolerance to magnesium over a period of
time.
The 1946 USPHS Drinking Water Standards
recommended a limit of 125 mg/l but there is no
limit in the 1962 standards. The 1958 WHO
International Standards have a "permissible limit" of
50 mg/l and an "excessive limit" of 150 mg/l, but no
maximum allowable concentration. The 1961 WHO
European Standards have a recommended limit of
125 mg/l, but if the sulfate exceeds 250 mg/l, the
magnesium is limited to 30 mg/I.
The taste threshold for magnesium (in MgS04) has
been reported as 100 mg/l and for the average
individual it is given as about 500 mg/I.
The negative correlation between hardness in
water and cardiovascular disease does not appear to
hold for magnesium as it does for calcium; yet one
investigator reports the favorable use of magnesium
sulfate to treat such cases and claims that
magnesium, rather than calcium, is the beneficial
element in reducing cardiovascular attacks.
(e) Fish and Other Aquatic Life. The relative
concentrations of magnesium and calcium in water
may be one factor controlling the distribution of
certain crustacean fishfood organisms, such as
copepods, in streams. Hart et ai. cite a report that
among U.S. waters supporting a good fish fauna,
ordinarily 5 percent have less than 3.5 mg/l of
magnesium; 50 percent have less than 7 mg/l; and 95
percent have less than 14 mg/I.
Magnesium chloride and nitrate can be toxic to
fish in distilled water or tap water at concentrations
between 100 and 400 mg/l as magnesium. However,
magnesium chloride, nitrate, and sulfate, at
concentrations between 1000 and 3000 mg/l as
magnesium have been tolerated for 2-11 days. Some
fresh-water fish have been found in very saline lake
water containing over 1000 mg/l of magnesium as
well as additional sodium and calcium salts.
TV A has proposed no specific limit on magnesium.
Magnesium sulfite or sulfate formation, either in the
scrubber or pond, could give relatively large concentrations
of sulfite and sulfate in the pond effluent water because of
the high solubility of these salts. If we assume 50%
oxidation of sulfite to sulfate in the scrubber, the solids to
water ratio in the sluice given in the earlier tabulation, and
use of a straight dolomite (no excess calcium), all of the
magnesium sulfate and almost half of the magnesium sulfite
formed would go into solution. The resulting
concentrations would be 4,140 milligrams per liter of
Mg++, 4,730 milligrams of SO~, and 12,700 milligrams of
SO:. The total salt concentration would be on the order of
22 grams per liter.
These are very high concentrations that would make
release of the pond water to streams very questionable,
particularly from large plants located on the smaller rivers.
The large increase in sulfite concentration would be an
especially severe problem. The effluent from the settling
pond would be devoid of dissolved oxygen and would
contain a sufficiently high residual concentration of sulfite
to cause a measurable reduction of dissolved oxygen in the
receiving stream. Although the oxygen reduction would not
be appreciable for a 200-megawatt plant, the effluent from
a much larger plant, 2,000 megawatts for example, would
result in a marked effect-likely to exceed critical levels at
minimum stream flows. The total oxygen reduction cannot
be accurately predicted since it depends on the quantity
and quality of receiving water and the rate of sulfite
oxidation in such water. However, the total oxygen
required to completely oxidize the sulfite from the
2,000-megawatt plant is approximately 76,000 pounds per
day or a quantity sufficient, for example, to cause a
dissolved oxygen reduction of over 3 milligrams per liter in
the Clinch River at average flows, assuming immediate
reaction and complete mixing. The reaction is not
immediate and may, in fact, be quite slow; thus the actual
reduction would be somewhat smaller.
As another example, the Paradise plant in the TV A
system is now rated at 1,400 megawatts and will be
81
-------
expanded to 2,500 in 1969. The plant is located on the
Green River in central Kentucky which has an average flow
of about 8,000 cubic feet per second. If dolomite wet
scrubbing were used on all units, discharge of the effluent
would significantly reduce the dissolved oxygen content of
the river and the sulfate level would increase by about 50
milligrams per liter.
The situation is quite different from that in the dry
process, where little or no formation of magnesium sulfate
(or sulfite) apparently takes place. The main concern in the
dry process is that MgS04 will gradually form by reaction in
82
the pond between Mg(OH)2 and CaS04 2H2 O. The
possible extent of this reaction in practice has not been
determined.
If a high-calcium limestone is used, there should be
many situations in which release of the dissolved calcium
salts to watercourses would be acceptable. The choice then
would be between the cost and difficulty of recycling and
the undesirable release of even an acceptable amount of
dissolved salts. If dolomitic limestone is used, the number
of acceptable situations for pond water release will be
drastically reduced.
-------
Appendix C:
Detailed Estimates
Table C-1. Summary of Estimated Fixed Investment
Requirements: Process A - Limestone Injection-
~~~~~~~~--------------------------------
(200-mw existing power unit)
!!J~~tfT!!!!!!.~
Cost Estimates
Table C-2. Summary of Estimated Fixed Investment
Requirements: Process A - Limestone Injection-
~q~~~~~----------------------------
(1 ,OOO-mw existing power unit)
!!J~~tfT!!!.!!!.~
Yard improvements Yard improvements
Road and general yard modifications 20,000 Road and general yard modifications 32,000
Limestone storage and handling facilities Limestone storage and handling facilities
Concrete foundations and conveyor tunnel 40,000 Concrete foundations and conveyor tunnel 114,000
Receiving hopper, storage silo, conveyor Receiving hopper, storage silo, conveyor
supports and bridges, and powerhouse supports and bridges, and powerhouse
storage silo 50,000 storage silo 127,000
Powerhouse structural steel revisions 3,000 Powerhouse structural steel revisions 6,000
Elevating and conveying in powerhouse 50,000 Elevating and conveying in powerhouse 127,000
Pulverizing and air system equipment 85,000 Pulverizing and air system equipment 285,000
Injection equipment (dense phase) 35,000 Injection equipment (dense phase) 85,000
Two-stage wet scrubbers with pumps, Two-stage wet scrubbers with pumps,
structures, piping, and hold tank 500,000 structures, piping, and hold tank 1,630,000
Solids disposal system Solids disposal system
Equipment and piping 100,000 Equipment and piping 364,000
Disposal pond and land 150,000 Disposal pond and land 323,000
Stack gas reheat system incl uding supports, Stack gas reheat system including supports,
pumps, and piping 585,000 pumps, and piping 2,125,000
Central control room and equipment 85,000 Central control room and equipment 245,000
Electrical and water distribution 149,000 Electrical and water distribution 463,000
Painting and insulation 30,000 Painting and insulation 120,000
Construction facilities --@)IJIQ Construction facilities _.fI.DJIJIQ
Total direct cost 1,932,000 Total direct cost 6,316,000
Engineering design 193,000 Engineering design 494,000
Contractor fees and overhead 292,000 Contractor fees and overhead 955,000
Contingency allowance _1~3.LQ1JQ Contingency allowance -_1.'!P1IJ!!!
_J~~UuQkgl~~~mWl______----------~[W~~~
aBasis:
Sulfur in coal, 3.5%
Stack gas reheat to 2500 F. by indirect liquid-gas method
Direct solids disposal as 10% slurry (no return of water),
scrubber-to-pond distance of 1 mi.
_I~~~~~~~w~~tfT!!!.!!t_-------------_J~~~]QQ
aBasis:
Sulfur in coal, 3.5%
Stack gas reheat to 2500 F. by indirect liquid-gas method
Direct solids disposal as 10% slurry (no return of water),
scrubber-to-pond distance of 1 mi.
83
-------
Table C-3. Summary of Estimated Fixed Investment
Requirements: Process A - Limestone Injection-
~~~~~~~---------------------------------
(1 ,OOO-mw new power unit)
!!l~~stfllei1tJ
Yard improvements
Road and general yard modifications
Limestone storage and handling facilities
Concrete foundations and conveyor tunnel
Receiving hopper, storage silo, conveyor
supports and bridges, and powerhouse
storage silo
Elevating and conveying in powerhouse
Pulverizing and air system equipment
Injection equipment (dense phase)
Two-stage wet scrubbers with pumps,
structures, piping, and hold tank
Solids disposal system
Equipment and piping
Disposal pond and land
Stack gas reheat system including supports,
pumps, and piping
Central control room and equipment
Electrical and water distribution
Painting and insulation
Construction facilities
20,000
100,000
115,000
100,000
250,000
85,000
1,580,000
362,000
323,000
2,125,000
230,000
445,000
110,000
_1~0-,-0JIJ!
Total direct cost
6,095,000
Engineering design
Contractor fees and overhead
Contingency allowance
370,000
745,000
_1LO-,-0JIJ!
Total project investment
7,620,000
Ele~trostatic precipitator creditb
Ua!!!L(lOJl)
Total chargeable project investment
-_!~~~~~~E~~fu~[~~__--------____~l~~OJIJ!
aBasis:
Sulfur in coal, 3.5%
Stack gas reheat to 2500 F. by inclirect liquid-gas method
Direct solids disposal as 10% slurry (no return of water),
scrubber-to-pond distance of 1 mi.
b99% effective electrostatic precipitator
84
Table C-4. Summary of Estimated Fixed Investment
Requirements: Process B - Limestone Addition-
~~~~~~~~--------------------------------
(200-mw existing power unit)
Yard improvements
Road and general yard modifications
Limestone storage and handling facilities
Concrete foundations and conveyor tunnel
Receiving hopper, storage silo, conveyor
supports and bridges, surge storage,
and feed conveyors
Wet grinding ball mill and classifier
Slurry storage and pumping
Two-stage wet scrubbers with pumps, piping,
structures, and hold tank
Solids disposal system
Equipment and piping
Disposal pond and land
Stack gas reheat system including supports,
pumps, and piping
Central control room and equipment
Electrical and water distribution
Painting and insulation
Construction facilities
Total direct cost
Engineering design
Contractor fees and overhead
Contingency allowance
!.!l~~~~!!~!
25,000
45,000
75,000
115,000
35,000
505,000
120,000
161,000
610,000
85,000
170,000
40,000
__6J!1IJ!(l
2,046,000
205,000
304,000
-~-111J!(l
_!~~~~~~0i~~!~~!_---------------~~~J~Q--
aBasis :
Sulfur in coal, 3.5% 0
Stack gas reheat to 250 F. by inclirect liquid-gas method
Direct solids disposal as 10% slurry (no return of water),
scrubber-to-pond distance of 1 mi.
-------
Table C-5. Summary of Estimated Fixed Investment
Requirements: Process B - Limestone Addition-
~~Y~~n~~____----------------------------
(1 ,OOO-mw existing power unit)
Yard improvements
Road and general yard modifications
Limestone storage and handling facilities
Concrete foundations and conveyor tunnel
Receiving hopper, storage silo, conveyor
supports and bridges, surge storage,
and feed co nveyors
Wet grinding ball mill and classifier
Slurry storage and pu mping
Two-stage wet scrubbers with pumps, piping,
structures, and hold tank
Solids disposal system
Equipment and piping
Disposal pond and land
Stack gas reheat system including supports,
pumps, and piping
Central control room and equipment
Electrical and water distribution
Painting and insulation
Construction facilities
Total direct cost
Engineering design
Contractor fees and overhead
Contingency allowance
!!J~~!..fl!!!.!!!.j
32,000
135,000
190,000
275,000
90,000
1,700,000
390,000
425,000
2,410,000
260,000
524,000
150,000
_11!9-L1!9.Q
6,781,000
549,000
990,000
_E.IL°-Lf!!I.Q
_J~~l~~~0j~!~~~~----------------~~£~f!!I.Q
aBasis:
Sulfur in coal, 3.5%
Stack gas reheat to 2500 F. by indirect liquid-gas method
Direct solids disposal as 10% slurry (no return of water),
scrubber-to-pond distance of 1 mi.
Table C-6. Summary of Estimated Fixed Investment
Requirements: Process B - Limestone Addition-
~~Y~~n~~____----------------------------
(1 ,OOO-mw new power unit)
Yard improvements
Road and general yard modifications
Limestone storage and handling facilities
Concrete foundations and conveyor tunnel
Receiving hopper, storage silo, conveyor
supports and bridges, surge storage,
and feed conveyors
Wet grinding ball mill and classifier
Slurry storage and pu mping
Two-stage wet scrubbers with pumps, piping,
structures, and hold tank
Solids disposal system
Equipment and piping
Disposal pond and land
Stack gas reheat system including supports,
pumps, and piping
Central control room and equipment
Electrical and water distribution
Painting and insulation
Construction facilities
Total direct cost
Engineering design
Contractor fees and overhead
Contingency allowance
Total project investment
Electrostatic precipitator creditb
!!J~~S!..fl!!!.!!!.j
25,000
125,000
170,000
275,000
90,000
1,600,000
390,000
425,000
2,390,000
245,000
476,000
120,000
_£@-Lf!!I.Q
6,621,000
400,000
809,000
_iiO-Lf!!I.Q
8,270,000
!lJ!!ILQPJll
Total chargeable project investment
-_l~~~~~~~illL~~~@~_---------____~~L~f!!I.Q
aBasis:
Sulfur in coal, 3.5%
Stack gas reheat to 2500 F. by indirect liquid-gas method
Direct solids disposal as 10% slurry (no return of water),
scrubber-to-pond distance of 1 mi.
b99% effective electrostatic precipitator
85
-------
Table C-7. Anooal Operating Costs for limestone-Wet Scrubbing Power Plant Stack
g~~-~~~~~~-~~~~~~~~~~~~~~~p~~~~---------------------------
(200-mw existing unit, 2.0% S in coal)
Total $Iton
annual of coal
~~nJJ~LC{!J!!n!i!Y... _$fuDJ1- _f-~UL J>JJ[!I~ci
Direct Costs
Delivered raw material
Limestone, 95% C03 43.2 M ton 2.05/ton 88,600 0.148
Conversion costs
Operating labor and
supervision 13,150 man-hr 4.00/man-hr 52,600 0.088
Utilities
Water 144,000 M gal 0.10/M gal 14,400 0.024
Electricity 8,800,000 kwh 0.004/kwh 35,200 0.059
Maintenance
Labor and material 72,500 0.121
Analyses 2,190 hr 7.50/hr _1~1!l0- 0.027
Subtotal conversion costs 191,100 0.319
Subtotal direct costs 279,700 0.467
Indirect Costs
Capital charges, 14.5% of fixed investment 339,300 0.565
Overhead
Plant, 20% of conversion costs 38,200 0.064
Administrative, 10% of operating labor __~2Q.Q.. _0~Q9~
Subtotal indirect costs 382,800 0.629
Total annual operating cost before credits 662,500 1.096
Precipitator operating credit -17,900 (0.030)
Thermal effect of raw limestone injection
on operating cost of power generation +11,500 0.019
Maintenance credit for corrosion reduction
in boiler -18,000 (0.030)
_J~~lg~~~~~fl~~L~~@ti~c2g_____-------_____~3~J~Q.__LQ.~_-
aBasis:
Coal burned, 600,000 tons/yr
90% effective electrostatic precipit8tor not in operation, two-stage scrubbing
Stack gas reheat from 11t to 250 F., indirect liquid-gas method
Limestone added, 11 0% stoichiometric calcium oxide
Ground limestone injected, 75% -200 mesh
Coal, 0.75 lb/kwh
Unit on-stream time, 8,000 hr
Direct solids disposal-scrubber to pond- 1 mile, 10% solids slurry
North Alabama plant location
Capital investment $2,340,000 fIxed; $12,000 working
86
-------
Table C-8. Annual Operating Costs for Limestone-Wet Scrubbing Power Plant Stack
g!~-~!~C~~-~~h~~toE~l~~~~~~~i~~~~~_------------------------
(200-mw existing unit, 3.5% S in coal)
~1}!1~~Lc{!!!I.P1i1'L
_~~Ili!...
Total
annual
_f.~!..j-
$/ton
of coal
burned
-----
Direct Costs
Delivered raw material
Limestone, 95% C03 77.2 M ton 2.05/ton 158,300 0.264
Conversion costs
Operating labor and
supervision 14,500 man-hr 4.00Iman-hr 58,000 0.097
Utilities
Water 252,000 M gal 0.10/M gal 25,200 0.042
Electricity 9,280,000 kwh 0.004/kwh 37,100 0.062
Maintenance
Labor and material 78,300 0.131
Analyses 2,190 hr 7.50/hr _J!l~.QQ. 0.027
Subtotal conversion costs 215,000 0.359
Subtotal direct costs 373,300 0.623
Indirect Costs
Capital charges, 14.5% of fixed investment 378,500 0.631
Overhead
Plant, 20% of conversion costs 43,000 0.072
Administrative, 10% of operating labor __MQ.Q... 0.010
Subtotal indirect costs 427,300 0.713
Total annual operating cost before credits 800,600 1.336
Precipitator operating credit -17,900 (0.030)
Thermal effect of raw limestone injection
on operating cost of power generation +19,400 0.032
Maintenance credit for corrosion reduction
in boiler -18,000 (0.030)
_~~~1~~~~~~~~~L~~@ti~~~-----------______L8~J.QQ.__l~@_-
aBasis:
Coal burned, 600,000 tonsfyr
90% effective electrostaticcPrecipitator not in operation, two-stage scrubbing
Stack gas reheat from 118 to 2500 F., indirect liquid-gas method
Limestone added, 11 0% stoichiometric calcium oxide
Ground limestone injected, 75% -200 mesh
Coal, 0.75 lbfkwh
Unit on-stream time, 8,000 hr
Direct solids disposal-scrubber to pond-l mile, 10% solids slurry
North Alabama plant location
Capital investment $2,610,000 fIxed; $20,000 working
87
-------
Table C-9. Annual Operating Costs for Limestone-Wet Scrubbing Power Plant Stack
g~~-~~~~~~~-~~~JQ~~~l~~Q~~~~~~~~~-------------------------
(200-mw existing unit, 5.0% S in coal)
_Ann!:!~...9!!~..!i1Y-
_$LI!IJ.iL
Total
annual
-~
-------
Table C-10. Annual Operating Costs for Limestone -Wet Scrubbing Power Plant Stack
g~~_~9~~~~=_llm~~~Q~ml~~Q~~£~~QID~~____-----------------------
(500-mw existing unit, 3.5% S in coal)
_P..!:I!!~L
-------
Table C-11. Annual Operating Costs for Limestone-Wet Scrubbing Power Plant Stack
~~~-~9~~~~=Jlm~~QQ~ID1~~Qn~~W9QID~~____-----------------------
(1,OOO-mw existing unit, 3.5% S in coal)
_A..!1B!!..!! !!,~Q..
791 ,000
0.264
131,000
0.044
88,200 0.029
135,000 0.045
238,000 0.079
-~~Q.Q.. _O~QJ&-
641,500 0.213
1,432,500 0,477
1,190,500 0.396
128,300 0.043
_J~Q.Q.. _0~Q9.1-
1,331,900 0.440
2,764,400 0.917
-66,000 (0.022)
+97,000 0.032
-78,000 (0.026)
2,717 ,400 0.901
------------------------------------------------------------
aBasis:
Coal burned, 3,000,000 tons/yr
90% effective electrostaticcPrecipitator not in operation, two-stage scrubbing
Stack gas reheat from 118 to 2500 F., indirect liquid-gas method
Limestone added, 110% stoichiometric calcium oxide
Ground limestone injected, 75% -200 mesh
Coal, 0.75 Ib/kwh
Unit on-stream time, 8,000 hr
Direct solids disposal-scrubber to pond-l mile, 10% solids slurry
North Alabama plant location
Capital investment $8,210,000 fIxed; $80,000 working
90
-------
Table C-12. Annual Operating Costs for limestone-Wet Scrubbing Power Plant Stack
g~~-~9~~~~=Jlm~~~Q~wl~~Q~~~~9~ID~~__-------------------------
(1,OOO-mw new unit, 3.5% S in coal)
_~B~i!Lg!!.aB!i!Y
-$L~~L
Total
annual
-~~!..j-
$/ton
of coal
J:>.!:!!:D~~
Direct Costs
Delivered raw material
Limestone, 95% C03 386.0 M ton 2.05/ton 791,000 0.264
Conversion costs
Operating labor and
supervision 32,740 man-hr 4.00/man-hr 131,000 0.044
Utilities
Water 1,260,000 M gal 0.07/M gal 88,200 0.029
Electricity 45,000,000 kwh 0.003/kwh 135,000 0.045
Maintenance
labor and material 220,000 0.073
Analyses 6,570 hr 7.50/hr _~g2Q.Q- _O--,-QJ.2-
Subtotal conversion costs 623,500 0.207
Subtotal direct costs 1,414,500 0.471
Indirect Costs
Capital charges, 13.0% of fixed investment 990,600 0.330
Overhead
Plant, 20% of conversion costs 124,900 0.042
Administrative, 10% of operating labor _J~Q.Q- 0.004
-----
Subtotal indirect costs 1,12&600 0.376
Total annual operating cost before credits 2,543,100 0.847
Precipitator operating-investment credit -283,600 (0.095)
Thermal effect of raw limestone injection
on operating cost of power generation +97,000 0.032
Maintenance credit for corrosion reduction
in boiler -78,000 (0.026)
--~~~~~~~~~~~~~~~~~~~~~~-~~------------______lL~~L~O~-__~I~~-
aBasis:
Coal burned, 3,000,000 tonsfyr
99% effective electrostaticcrrecipitator replaced by two-stage scrubbing
Stack gas reheat from 118 to 2500 F., indirect liquid-gas method
Limestone added, 110% stoichiometric calcium oxide
Ground limestone injected, 75% -200 mesh
Coal, 0.75 lbfkwh
Unit on-stream time, 8,000 hr
Direct solids disposal-scrubber to pond-l mile, 10% solids slurry
North Alabama plant location
Capital investment $7,620,000 fIxed; $80,000 working
-------
Table C-13. Annual Operating Costs for Limestone-Wet Scrubbing Power Plant Stack
~~~-~~£~~~=~lm~~QP~_~~~~Q~~gM~~~g~---------------------------
(200-mw existing unit, 3.5% S in coal)
_A--DJ:!!:,-!!lgLLaJ:!!l~
-$L~IJl!...
Total
annual
-~~~j-
$/ton
of coal
J:>.!:!!:...~
-------
Table C-14. Annual Operating Costs for limestone-Wet Scrubbing Power Plant Stack
~~~-~9~~~~~~lm~!~~_~~~~Q~~~~~~mg~____-----------------------
(500-mw existing unit, 3.5% S in coal)
_A...!1JJI!'!!.gI,LaJJ!!!Y
-$LI!f!i~
Direct Costs
Delivered raw material
limestone, 95% C03
Conversion costs
Operating labor and
supervision
Util ities
Water
Electricity
Maintenance
labor and material
Analyses
233.2 M ton
2.05/ton
24,670 man-hr 4.00/man-hr
713,000 M gal
23,500,000 kwh
0.10/M gal
0.004/kwh
4,380 hr
7.50/hr
Subtotal conversion costs
Subtotal direct costs
Indirect Costs
Capital charges, 14.5% of fixed investment
Overhead
Plant, 20% of conversion costs
Administrative, 10% of operating labor
Subtotal indirect costs
Total annual operating cost before credits
Precipitator operating credit
Total chargeable annual operating cost
Total
annual
-~~~j-
$/ton
of coal
J>.!:!!:!I~ct
478,000
0.319
98,700
0.066
71 ,300 0.048
94,000 0.063
180,700 0.120
_~I~~Q. 0.022
-----
477,500 0.319
955,500 0.635
845,400 0.564
95,500 0.064
--Q~~Q. 0.007
-----
950,800 0.635
1,906,300 1.270
-44,000 (0.029)
1,862,300 1.241
------------------------------------------------------------
aBasis:
Coal burned, 1,500,000 tons/yr
90% effective electrostatic"precipitg.tor ~ot.in op~ra~ion, two-stage scrubbing
Stack gas reheat from 118 to 250 F", mduect liqUId-gas method
Limestone added, 130% stoichiometric calcium oxide
Ground limestone added, 75% -200 mesh
Coal,0.75Ib/kwh
Unit on-stream time, 8,000 hr
Direct solids disposal-scrubber to pond-1 mile, 10% solids slurry
North Alabama plant location
Capital investment $5,830,000 fixed; $50,000 working
-------
Table C-15. Annual Operating Costs for limestone-Wet Scrubbing Power Plant Stack
~~~-~~~~~~~~Mn~!~Q~~~~~Q~~g~~Qmg~____-----------------------
(1 ,OOO-mw existing unit, 3.5% S in coal)
Total $/ton
annual of coal
1\...!1.!!~C!!..9.~a!J!i!Y _$L~QlL -~~!t.j l>~ !:Di!.recipitator not in operation, two-stage scrubbing
Stack gas reheat from 118 to 2500 F., indirect liquid-gas method
Limestone added, 130% stoichiometric calcium oxide
Ground limestone added, 75% -200 mesh
Coal,0.751bfkwh
Unit on-stream time, 8,000 hr
Direct solids disposal-scrubber to pond-l mile, 10% solids slurry
North Alabama plant location
Capital investment $8,820,000 fixed; $96,000 working
94
-------
Table C-16. Annual Operating Costs for Limestone-Wet Scrubbing Power Plant Stack
~~~-~9~~~~~~lm~~~~_~~~g~~~g~~Qm~~____-----------------------
(1,000-mw new unit, 3.5% S in coal)
_A_I1B~alg~a.D!i!Y
_$L~Ill!...
Total
annual
_.Q.~tJi-
$/ton
of coal
J>.!:!!:rI~
-------
Determination of Credit for Corrosion Reduction
Injection of limestone into the boiler is expected to
reduce sulfur trioxide concentration in the exhaust gases to
a very low value. Since this will greatly reduce the
formation of acidic salts at high temperature and sulfuric
acid at low temperature, corrosion of the boiler heat
transfer surfaces will be reduced. The expected corrosion
reduction will result in a cost credit to be applied in the
evaluation of process A.
As indicated by TV A experience and the Federal Power
Commission report! the average maintenance cost for a
boiler is $OA5/ton of coal burned. From TV A accounts of
maintenance distribution, it was estimated that expected
corrosion reduction could be valued at $O.ll/ton of coal
burned. However, corrosion reduction must be weighed
against increased surface erosion due to the limestone. A
rough estimate of this offsetting effect indicates that no
more than about 30% of the above credit could be saved.
Therefore, a credit of $0.03/ton of coal burned was applied
in evaluating process A.
Determination of Operating and Investment Credits
for Eliminating Electrostatic Precipitator
In outlining the basis for evaluation of limestone-wet
scrubbing processes, it was assumed that existing fossil
fuel-burning units are equipped with 90% effective
electrostatic precipitators and that new units would be
equipped with at least 99% effective precipitators if no
other means of stack gas cleaning were to be used. If wet
scrubbers are added to the system for better sulfur dioxide
removal, the need for the electrostatic precipitators
becomes questionable. In the section "Economic
----------
IFederal Power Commission. Steam-Electric Plant Construction
Cost and Annual Production Expenses, Nineteenth Annual
Supplement (1966).
96
Evaluation" an economic appraisal of existing units using
various precipitator-scrubber combinations indicated that it
would be economical to shut down the existing
precipitators and remove the dust with the scrubbers alone.
Of course, the installation of electrostatic precipitators on
new units would not be required if a wet-scrubbing process
were used.
F or the use of processes A and B on existing units, an
operating credit was taken; for new units, both an operating
and investment credit were taken. The estimated credits are
shown below:
Precipitator Annual Credits
Labor and overheads
Power
Maintenance
Y:3J~Ln~~
2JJ.!Lrnw L.QQQLQVL
$ 3,600 $10,000
10,700 38,000
3,600 18,000
Newb
LQQQTILVL
$ 13,000
75,600
26,000
Total credit
for existing
precipitators
17,900
66,000
Capital charges
at 13% of new
investmentC
169,000
Total credit for
-~}Y-p,!,~!pitator - - $283 600
:?O% efficien;-----------------------_::'L_-
99% efficient.
cl,OOO-mw new unit; 3,300,000 actual c.f.m.; 3.5% g in coal;
estimated project cost of 99% effective precipitator-$1,300,000.
-------
,.
BOILER
PROCESS
WATER
SURGE
TANK
16
PULVERIZED LIMESTONE 2
PULVERIZED COAL
FAN
IBA
S
PUMPS
---------
ECONOMIZER
STACK
.ccD-
AIR
HEATER
8
PRECIPITATOR
STACK
GAS
COOLER
STACK
GAS
REHEATER
"
7
'---'
ASH
HOPPER
RECIRCULATION
TANK
PUMPS
ISA
~-cJ)--
PRECIPITATOR
SA
- ---- -- - - --
-- SCRUBBER
lOA
STACK
GAS
COOLER
RECIRCULATION
TANK
STACK
GAS
REHEATER
FAN
14
PUMPS
13
SETTLING POND
TYPICAL OPERATING CONDITIONS FOR LIME SCRUBBING PROCESS WITH A 200 MEGAWATT STEAM PLANT- 3.5" SULFUR IN COAL
\0
--.)
STREAM NO. I 2 3 . . . 7 . 8 SA 9 It 14 108104. 11811A 12 a 12A 13 ,. " 16816A 17817A 188 leA 19 It 19A
COAL LIMESTONE COMBUSTION COMBUSTION GAS GAS GAS GAS GAS GAS MAKE-UP SLURRY POND..I POND POND RECYCLE SCRUBBER ~A~tICA~ HOT WATER
DESCRIPTION 70 70 AIR TO AIR TO 70 70 70 70 70 70 WATER 70 SOLIDS WATER OVERFLOW SLURRY TO SLURRY TO ~ ESJ:g~E~~
BOILER BOtl.£R AIR HEATER BOilER ECONOMIZER AIR HEATER COOLERS SCRUBBER REHEATER STACK POND CCUMlA..ATK)N LOSS WATER SCRUBBER RECIR. TANK COOLER
RATE, LBS IHR 150M f9.JM ',9(/M I,728M 1,169'" ',869M 2,0$6 M 1,028M 1,04'M 1,046M 200M /96M 39'" 124M 228M 4,200M 4,196M 273M 273M
SCFM - 4'9M 378", 400M WOM .,40M 220M 227M 227M
GPM - - - 400 373 248 4.6 8M 7,992 ..6 .~
PARTICULATES, LBS IHR - 0 0 26.6M 26.6M 26.6M I~.~M O.l7M o 17M 40 -
TEMPERATURE. ~F AMBIENr AN8IENr I/O 610 890 70. 310 /77 120 2~0 6. '20 6. 6. 120 120 '48 280
SPECIFIC GRAVITY I.G I.O~ 1.0 I.O~ 10. 0.98 0.93
VISCOSITY, CPS 0.019 oo~o 1.05 1.05 0.44 0.20
UNDISSOLVED SOLIDS, "" 0 10 10 10.5
pH 7
CALCULATED ANALYSIS OF GAS FROM AIR HEATER NOTES: SYMBOL IN TABLE
J. CALCULATIONS BASED ON: 3 PARTICULATE RATES IN LBS IHR. SHOULD BE ADDED TO
CONCENTRATION PARTI C ULATES o. 110% STOICHIOMETRIC LIME STACK GAS RATES TO GET TOTAL STREAM RATE (1.0. TO N----- THOUSAND
(VOLUME '" (GRAINS/SCF) b. % SULFUR IN COAL (DRY BASIS) DETERMINE THE TOTAL RATE OF STREAM S, ADD 1,869M
,. 12 % ASH COAL (AS FIRED BASIS) 826.6 M)
" S IN COAL " S IN COAL d. 92 % OF SULFUR IN COAL EVOLVES AS S02 4 STREAMS SUCH AS 8 a eA, 9 a 9A HAVE EOUAl RATES
20 3 . 5 0 2 0 35 50 .. 75 % OF ASH IN COAL EVOLVES AS FLY ASH 8. PHYSICAL PROPERTIES, BUT ONLY THE INFORMATION FOR
t. MISCELLANEOUS INFORMATION FROM SCRUBBER MANUFACTURERS
SULFUR DIOXIDE 0.09 016 0.2:1 ,. 987 % REMOVAL OF PARTICULATES TO SCRUBBERS ONE SCRUBBER SYSTEM IS LISTED IN THE TABLE
h 9' % TOTAL S02 REMOVAL 5. GAS DUCT BYPASSES ARE SHOWN AS DASHED LINES
WATER VAPOR 1.84 789 7.94 I LIMESTONE TO BOILER CONTAINS 10% WATER AND 9S% CoCO!
NITROGEN 74.5~ 7436 '''.18 (DRY BASIS). 6 PRECIPITATORS ARE TO BE USED ONLY WHEN SCRUBBERS
OXYGEN 486 ".84 4.82 ARE SHUTDOWN AND BYPASSED
CARBON DIOXIDE 12.68 1215 128:1 - 2. POND WATER LOSS CALCULATIONS: EVAPORATION FROM
:112 380 U. S. 0 A TECH BULLETIN 271, ASSUME e2 IN I YR. RAINFALL;
FL Y ASH - - - 390 ASSUME 1.'0-8 CM I SEC SEEPAGE
CoO 8 Co SO.. - '8' :1.2~ 465
Figure D-1. Flowsheet - Process A
;;t>
~
~
(t)
~
0-
~.
~
o
o
I-t
~
~
S'
(JQ
r:/'J
-------
\0
00
TYPICAL OPERATING CONDITIONS FOR LIME SCRUBBING PROCESS WITH A 200 MEGAWATT STEAM PL ANT - 2" SULFUR IN COAL
STREAM NO 1 2 3 4 5 6 7 868A 989A 10810A 11811A 12612A 13 14 15 16816A 17817A 18 a 18A 19819A
COAL LIMESTONE COMBUSTION COMBUSTION GAS GAS GAS GAS GAS GAS MAKE UP SLURRY POND POND POND RECYCLE SCRUBBER tt~~A~:T~~ HOT WATER
DESCRIPTION TO TO AIR TO AIR TO TO TO TO TO TO TO WATER TO SOLIDS WATER OVERFLOW SLURRY TO SLlRRY TO TO STACK GAS
BOilER BOILER AIR HEATER BOILER ECONOMIZER AIR HEATER COOLERS SCRUBBER REHEATER STACk POND ACCUMJLATJON LOSS wATER SCRUBBER REC IR TANK COOLER REHEATER
RATE, LBS.lHR. 150M 10.8M 1,911M 1,724M 1,862M 1,862'" 2,049M I,024M 1,041M 1,041M 145M 139M 21.8M 9'M 1~9M 4,200M 4,I94M 211M 271M
sen., 419M 378M 399M 399M 439M 219M 226M 226M
GPM 290 265 162 316 6M 7,989 552 563
PARTICULATES, LBSJ HR - - 0 0 209M 209M 209M lOAM Ol3M 013M - - - - 30 - - -
TEMPERATURE.8F AMBIENT AMBIENT 110 610 690 705 310 177 119 250 65 119 65 65 119 119 146 260
SPECIFIC GRAVITY 1.0 1.05 1.0 1.05 1.05 0.98 0.93
VISCOSITY, CPS 0.019 0.030 105 1.05 0.44 0.20
UNDISSOLVED SOliDS. '"I. 0 10 10 10.3
pH 7
TYPICAL OPERATING CONDITIONS FOR LIME SCRUBBING PROCESS WITH A 200 MEGAWATT STEAM PLANT- 5 " SULFUR IN COAL
STREAM NO 1 2 3 4 5 6 7 888A 98 9A t0810A 11811A 128 12A 13 14 '" !6a16A 17 a 17A 18 al8A 19 al9A
COAL LIMESTONE COMBUSTION COMBUSTION GAS GAS . GAS GAS GAS GAS MAKE -up SLURRY POND POND POND RECYCLE SCRUBBER WARM WATER HOT WATER
DESCRIPTION TO TO AIR TO AIR TO TO TO TO TO TO TO WATER TO ~G~~~,OI WATER OVERFLOW SLURRY TO SLURRY TO TO STACK GAS TO STACK GAS
BOILER BOILER AIR HEATER BOILER ECONOMIZER AIR HEATER COOLERS SCRUBBER REHEATER STACK PONO LOSS WATER SCRUBBER RECIR. TANK COOLER REHEATER
RATE, LBS I HR. 150M 21.7M 1,911M 1,128M 1,876M 1816M 2,063M 1,031M 1.0~2M 1,052'" 255M 251M 502M 124M 328M 4,200M 4,I96M 215M 215M
SCFM 419M 318M 40lM 40lM 441M 220M 229M 229M - - -
GPM - - - - - - 510 476 - 248 656 6M 7,992 560 591
PARTICULATES, LBS IHR - - 0 0 32.3M 32.3M 32.3M 16.IM 0.211.1 0.21M - - - - 60 - - - -
TEMPERATURE, -F AMBIENT AMBIENT 110 610 690 705 310 179 122 250 65 122 65 65 122 122 150 260
SPECIFIC GRAVITY 10 105 1.0 105 105 096 093
VISCOSITY, CPS 0.019 0.030 1.05 1.05 0.415 0.20
UNDISSOLVED SOLIDS, 8f. 0 10 10 106
pH 7
Figure D-2. Supplement to Figure D-1.
-------
20
BOILER PROCESS
WATER 19
19A
9 PUMPS
ECONOMIZER - ----- - -- STACK
PULVERIZED COAL .co>-
AIR S
HEATER
PRECIPITATOR STACK
GAS 12
7 COOLER SURGE DELAY
'----/ TANK TANK
ASH 13
HOPPER
PUMPS
20A
I 9A
I - ----- - -- --
L.cD- SA IDA
PRECIPITATOR
STACK 12 A FAN
GAS
COOLER DELAY
15 SURGE
TANK TANK
13A
16
!. LIMESTONE PUMPS
14
SETTLING POND 4 ROCK
~
HOPPER
CLASSIFIER
TYPICAL OPERATING CONDITIONS FOR LIME SCRUBBING PROCESS WITH A 200 MEGAWATT STEAM PLANT- 3.5% SULFUR IN COAL
2- STAGE SCRUBBERS, B5% SOl REMOVAL
STREAM NO. 1 2 3 4 . . 1 8 a 8,. 9 & 9A loa lOA Ila IIA 128 12A 136134 14 ,. ,. 17817A 18 Eli 18A 19 a 19A 20. 20A
COAL. Llt.!ESTONE COMBUSTION COMBUSTION GAS GAS GAS GAS GAS GAS LIMESTONE MAKE-UP SLURRY POND POND POND RECYCLE SCRUBBER WARM WATER HOT WATER
TO TO AIR TO AIR TO TO TO TO TO TO TO SLURRY TO TO SOLIDS WATER OVERFLOW SLUR-RY TO SLURRY TO TO STACK TO STACK
BOilER HO PPER AIR HEATER BOILER ECONOMIZER AIR HEATER COOLERS SCRUBBER REHEATER STACIC DELAY TANK WATER POND ACCUMULATION LOSS WATER SCRUBBER SURGE TANK GAS COOLER REHEATER
RATE, LBS IHR /50M 2~JM 1,9/IM 1,"24'" 1,8&3M /,86JM 2,0$0 AI I,02$M 1,042M 1,042'" 278M /89M 208M '1'4'" 1241r1 250M 5,59JM 5,584'" 271M 271'"
SCFM 419 M '78'" 399'" 399'" 439111 219M 223M 223111
GPM ... '70 59. ..8 '00 107M 10.6M .O! ...
PARTI CUL ATE S, LBS.I HR. 0 0 13.3'" " 3 III 133M 6.7M 90 90 .0 -
TEMPERATURE, .F AMBIENT AMBIENT /10 610 890 70' JlO 17. 118 200 6. 118 .. 6. 118 118 146 280
SPECIFIC GRAVITY 1.23 1.0 1.05 10 10. 1.03 098 095
VISCOSITY, CPS 00/9 0030 10. 1.05 04' 020
UNDISSOLVED SOLIDS, % 55 0 10 10 I 104
pH 7
NOTES:
CALCULATED ANALYSIS OF GAS FROM AIR HEATER
r. CALCULATIONS BASED ON:
a. 130 % STOICHIOMETRIC LIME
b % SULFUR IN COAL fDRY BASis)
C, 12 % ASH COAL (AS FIRED BASIS)
d. 92 % OF SULFUR IN COAL EVOLVES AS S02
.. 75 % OF ASH IN COAL EVOLVES AS FLY ASH
r. MISCELLANEOUS INFORMATION FROM SCRUBBER MANUFACTURERS
9. 90% REMOVAL OF FLY ASH IN PRECIPITATORS WHEN USED
h % REMOVAL OF PARTICULATES IN SCRUBBERS. 98.7
I "" TOTAL S02 REMOVAL: SEE ABOVE
J LIMESTONE TO SYSTEM CONTAINS 10"" WATER AND 95% CaCO,
(DRY BASIS!. ASSUME 2% LOSS
2. POND WATER LOSS CAL.CULATIONS: EVAPORATION FROM
U.s. 0 A TECH BULLETIN 211; ASSUME 52 IN IYR. RAINFALL
ASSUME I. 10-8 CM 1 SEC SEEPAGE
CONCENTRATION PARTICULATES
(VOLUME .,,) (GRAINS I SCF)
SULFUR DIOXIDE 0.22
WATER VAPOR 7. 76
NITROGEN 74 58
OXYG:N 4.89
CARBON DIOXIDE 12.55
FLT ASH '59
\0
\0
, PARTICULATE RATES IN LBS I HR. SHOULD BE ADDED TO
STACK GAS RATES TO GET TOTAL STREAM RATE (I ;. TO
DETERMINE THE TOTAL RATE OF STREAM 5, ADD 1863M
a 13.5 M)
4 STREAMS SUCH AS 8 Bi 8A, 981 9A HAVE EQUAL RATES
& PHYSICAL PROPERTIES, BUT ONLY THE INFORMATION FOR
ONE SCRUBBER SYSTEM S LISTED IN THE TABLE.
S GAS DUCT BYPASSES ARE SHOWN AS DASHED LINES
SYMBOL IN TABLE
M--_u THOUSAND
Figure D-3. Flowsheet - Process B
-------
-
o
o
I
I
P-~
,
-~-~'l-
--- '
-- ~':1 I
// -l i I :
, , .
, ,
,
/'
----.--- -
:~
,r ..
\.
/::>
~
. ------- '
----
-
--<~-
,--~-
.----'
~
-<'"'
~
.~
~
Figure D-4. Limestone receiving, handling, and dry injection facilities
-------
-
o
-
t
i
o
ELEVATOR
SURGE ~IN
~
I
I I
I I
t I I
I I
'- I _I
, -
I -J.- -
'-,L-- -
PUMPS
Figure D-5. Limestone wet-grinding facilities - Process B
SURG~ TANK
S'fS"Tt.~
lIe.e.\~G
--+ ro 5CR
-.---.
-------
....
o
N
POWEIZHOU
-------
-~~-
5TACA:::.
~PO:O FEED
~UMP5
~ ~-
I
-l-
t
I
RECIRCULATION
TANK.
-
o
w
ICECIRCULA TION
PUMPS
ELt'CTIZOSTAT/C
PI2EC/PITA TO/Z5
6AS COOLER
.----+-
-----------
II
I
II
III~ I \ '
111r:tIX'
Ilia I - ",
I 111:1 1 'f "
II
I
II
II
II
I. 1:). FAN
SCI2U""D<..
Figure D-7. General arrangement-scrubber area plan view
FLUE GAB DUCT
~
\tI
..J
....
o
IQ
-------
!}'-O"
'I"
15'-0"
3'-0=1
'I'
9'-0"
..-
o
.j:.
A-,
7' 0'
" - ~I
-
-------------
~~~er~Q -]Jr~=-=
TOP STAGE
---------------
---------------
BOTTOM STAGE
--------------
-------------
[:xJ
A--1
.
o
,
o
(\j
.
o
.'
.
o
.'
I/)
'0
.;,
WATER
INLET
.
o
.'
I/)
(\j
.
o
I
.0
~
:{
'0
-'
ID
GAS
INLET
r
12'-0"
"I
------------------
ENTRAINMENT ELIMINATOR
RETAINING GRID
SPRAY NOZZLES
SUPPORT GRID
SUPPORT GRID
SECTION A-A
20" I PS DISCHARGE
Figure D-8. TCA scrubber for wet limestone scrubbing process
WATER
OUTLET
------- |