United States Industrial Environmental Research
Environmental Protection Laboratory
Agency Research Triangle Park NC 27711
Technology Transfer
c/EPA Summary Report
Sulfur Oxides Control
Technology Series:
Flue Gas Desulfurization
Dual Alkali Process
u. $
N. ). 08817
tit,-'. •'
-------
-------
Technology Transfer EPA 625/8-80-004
Summary Report
Sulfur Oxides Control
Technology Series:
Flue Gas Desulfurization
Dual Alkali Process
October 1 980
LIBRARY
U S. ENVIRONMENTAL PfittTECTlOH
EDISON, H. J- 08817
This report was developed by the
Industrial Environmental Research Laboratory
Research Triangle Park NC 27711
-------
Absorber, oil preheater, and flue gas ducting
-------
Introduction
The Environmental Protection
Agency (EPA) is studying dual alkali
(or double alkali) flue gas desulfur-
ization (FGD) as part of an extensive
program of FGD technology
development. In this throwaway
process (Figure 1), sulfur dioxide
(S02) is removed from the flue
gas by a soluble sodium-based
scrubbing liquor. The collected SO2
is precipitated as calcium sulfite
(CaS03), calcium sulfate (CaS04), or
a mixed crystal of both salts, and is
purged from the system.
Currently over 50 dual alkali systems
are operating in the United States
and Japan, and several more are
under construction. Since the
early 1970's, dual alkali installations
have controlled SO2 emissions
from various sources: utility and
industrial boilers (coal and oil fired),
and ore roasting and coking facilities.
The EPA has been active in demon-
strating and testing the dual
alkali FGD process. Initial studies
were conducted at the EPA Industrial
Environmental Research Laboratory
at Research Triangle Park, North
Carolina. Subsequently, EPA
performed laboratory and pilot plant
studies of the process. This early
work was expanded later to include
prototype testing of a utility
application of dual alkali FGD. A
20-MW FGD prototype system was
constructed by Combustion
Equipment Associates/Arthur D.
Little (CEA/ADL) for Southern
Company Services, Inc., at the
Scholz plant of Gulf Power Company.
This installation operated for
1 7 months, from 1 975 to 1 976, and
was shut down after the successful
demonstration of the technology.
Industrial application of the process
was evaluated at General Motors'
Chevrolet-Parma plant near
Cleveland, Ohio, from August 1974
to May 1976. The operation of
the 32-MW FGD facility was tested
intensively over three 1-month
periods.
Research to date has established
dual alkali FGD as feasible for con-
trolling S02 emissions. The
capabilities of the process at full
scale are being assessed at Louisville
Gas and Electric Company's
300-MW Cane Run No. 6 Boiler,
where testing of a dual alkali system
is underway.
This summary report provides a
basic understanding of the dual alkali
FGD process. More detailed informa-
tion can be found in the literature
cited at the end of the report.
Key
Flue gas/off-gas
Cleaned flue gas
Absorption liquor
Sulfur products
Other systems
Flue
gas
Desulfunzed
flue gas
S0x-nch
sludge
Disposal
Figure 1.
Typical Dual Alkali FGD System
-------
Process Description
The dual alkali FGD process con-
sists of four basic steps:
• Flue gas pretreatment (optional)
• S02 absorption
• Absorbent regeneration
• Solid/liquid separation and
solids dewatering
Figure 2 illustrates the process flow
for a typical dual alkali FGD system.
During pretreatment, flue gas from
the boiler is routed through an
electrostatic precipitator (ESP) to
remove particles (fly ash) upstream
of the absorber. Pretreatment can
also involve wet scrubbing alone or
in series with the ESP for particle
and chloride removal. Pretreatment
is not always necessary in dual
alkali FGD; its use depends on site-
specific conditions such as fuel
characteristics and cost consid-
erations.
The flue gas then flows to an absorber
and is brought in contact with
a recirculating solution containing
an equilibrium mixture of sodium
sulfite (Na2SO3), sodium bisulfite
(NaHS03), sodium hydroxide (NaOH),
sodium carbonate (Na2C03), and
sodium bicarbonate (NaHC03).
Sulfur dioxide diffuses into this
solution and reacts with the sodium-
based alkali to form soluble
sulfur oxide (SOJ salts, which are
drawn off in the scrubber effluent.
Desulfunzed flue gas leaves the
absorber; it is reheated, if necessary,
and is exhausted through the stack
to the atmosphere. The SOx-nch
scrubber effluent is routed to the
absorbent regeneration system to be
reacted with lime or limestone.
The soluble sulfur oxide salts are
precipitated as:
• Calcium sulfite hemihydrate
(CaS03 • 1/2H20)
• Gypsum (CaS04 • 2H20)
• A mixed crystal form of both salts
The precipitation reaction also
regenerates the sodium-based alkali
for recycle to the absorber.
The precipitated SOX salts are
separated from the scrubbing liquor
and concentrated for disposal in
the solid/liquid separation and
solids dewatering step. The solids
settle out of the slurry in a clarifier-
thickener; they are dewatered further
in a vacuum filter or centrifuge,
and are washed to recover sodium
before disposal. The clear liquor
overflow from the clarifier-thickener
is combined with makeup sodium
and water and returned to the
absorption system.
Pretreatment
Pretreatment removes particles
from the flue gas upstream of the
absorber, by use of an ESP. Pre-
treatment also prevents large scale
contamination of the recirculating
liquor used during absorption.
In some cases, a ventun prescrubber
may be used to remove chlorides,
particles, and some sulfur dioxide.
The prescrubber generates an
acidic waste stream that must be
neutralized with lime or limestone
before disposal. In dual alkali
systems, a prescrubber can create
water balance problems; however, it
may be needed for applications
burning high chloride coals.
Absorption
Sulfur dioxide can be absorbed
from the flue gas in a tray tower
absorber (Figure 3a). The flue
gas enters the bottom of the absorber
and flows countercurrent to the
sodium-based scrubbing liquor,
which absorbs S02 and is recycled
through the absorber. A bleed stream
of the S02-rich scrubbing liquor
is withdrawn continuously from the
absorber and routed to the absorbent
regeneration system. A mist
eliminator removes entrained
liquor from the scrubbed gas stream,
which is then routed to the stack.
-------
Key
Flue gas/off-gas
Cleaned flue gas
Absorption liquor
Sulfur products
Other systems
Wash water
Makeup water
S0x-nch sludge
Disposal
Figure 2.
Dual Alkali FGD Process
Sulfur oxides in the flue gas are
absorbed primarily by Na2S03
dissolved in the scrubbing liquor.
The major absorption reaction is:
SOg2 + S02 + H20 -» 2HS03 (1)
Other bases are present in the
scrubbing liquor and react with the
flue gas sulfur dioxide. Scrubbing
liquor from lime-based regeneration Sodium makeup to the system, usually
systems may contain OH", which in the form of Na2C03, also absorbs
reacts with S02 as follows:
20H- + S02 — SOJ2 + H20
OH- + S02 -» HSOJ
(2)
(3)
S02:
COjM
2HS
HCOa-
- 2S02 +
°3 + C°:
f S02 ->
H20 ->
,1
HSOs + C02 T
(4)
(5)
-------
To stack
Mist eliminator —
\ Flue gas ^
Solid
waste to
disposal
Vacuum fitter
Flue gas/off-gas
Cleaned flue gas
Absorption liquor
Sulfur products
Other systems
Figure 3.
Dual Alkali FGD: (a) Tray Tower Absorber, (b) Absorbent Regeneration System, and (c) Solid/Liquid Separation and
Solids Dewatering System
-------
Dual alkali systems are usually
classified as dilute or concentrated,
depending on the concentration of
active alkalis in the scrubbing
liquor. Active alkalis are those ions
in the scrubbing liquor that par-
ticipate in the S02 absorption
reactions; they include COj2, HC03,
OH~, SOs2, and HS03. Although
it does not react directly with SO2 in
these systems, HSOj is consid-
ered an active alkali because it
can be converted to an active form by
reaction with lime or limestone.
The concentration of sodium ion
(Na+) in solution that is associated
with the active alkali is known
as active sodium.
Dilute systems usually have active
sodium concentrations less than
0.1 5 molar; in concentrated systems
the concentrations are greater
than 0.15 molar. The distinction is
important in determining appropriate
methods for sulfate removal.
An important side reaction in the
scrubber is oxidation of sulfite
or bisulfite to sulfate:
2S032 + 02 -» 2SC-42
2HS03+02
(6)
(7)
The sulfate thus formed interferes
with S02 removal. Controlling
this reaction is important in the
operation of dual alkali systems.
Absorbent Regeneration
In the regeneration step the sodium
alkali is regenerated, the collected
SOX salts are precipitated and
removed, and provision is made for
sulfate removal. A continuous bleed
stream is withdrawn from the
absorber and sent to the absorbent
regeneration system (Figure 3b).
The regenerant (lime or limestone)
is mixed with the scrubber effluent
in chemical mix tanks.
The regeneration reactions differ
according to whether slaked lime
[Ca(OH)2] or limestone (CaCO3) is
the regenerant. When slaked lime is
used, the reactions are:
Ca(OH)2 + 2HSC-3 — S032
+ CaS03 • 1/2H20 i + %H20 (8)
Ca(OH)2 + SOg2 + 1/2H20 —
20H- + CaS03 • 1/zH20 I (9)
Ca(OH)2 + SC>42 + 2H20 ^
20H-+CaS04-2H2O I (10)
For limestone the reactions are:
CaCO3 + 2HSC-3 —
S032 + CaS03 • 1/2H20 1
+ C02 T + y2H2O
(x + y)CaC03 + xSOj2
+ (x + y)HSO3 +zH20
(11)
+ xCaS04 • yCaS03 • zH20 I
+ xS032 (12)
Equation 12 is postulated based on
known products and the pH at
which the reaction occurs.
Because limestone is less soluble
than calcium sulfite, OH~ will not be
regenerated from SO^2 in these
systems.1 Therefore, the S02
absorption reactions of Equations 2
and 3 do not occur in limestone
systems.
Solid/Liquid Separation and Solids
Dewatering
The slurry of CaS03, CaS04, and
mixed crystal solids is low in
insoluble solids as it leaves the
regeneration system. It is fed to the
center well of the clarifier-thickener
(Figure 3c) to be concentrated.
Waste product solids produced during
regeneration settle from the slurry
liquor. Effluent from the bottom
of the thickener is then drawn off
and fed to a vacuum filter for further
dewatering.
The solids cake formed on the
vacuum filter is washed by several
water sprays. The wash water
removes up to 90 percent of the
occluded soluble salts from the cake
and returns them to the system.
This step reduces sodium losses,
and therefore sodium carbonate
makeup requirements, and lowers
the leaching potential of the solid
waste. The mixed filtrate and wash
liquor are returned to the thickener.
Clear liquor overflow from the
clarifier-thickener is collected in a
holding tank, which supplies
absorbent liquor to the absorption
system. Water can be added to the
holding tank to make up the dif-
ference between total system water
lost and total water entering
from other sources.
Makeup sodium is supplied by
adding Na2C03 to the clarifier-
thickener. In dilute systems, where
IMa2C03 may also be used as a
softening agent, some CaC03
does precipitate and is removed
with the solids from the clarifier-
thickener. Very little, if any,
CaC03 precipitates in concen-
trated systems.
Integrated System
The foregoing steps are part of the
integrated system. Figure 4 shows
how they relate to one another to
form a complete dual alkali FGD
process.
-------
To stack
Mist eliminator —
Solid
waste to
disposal
Vacuum filter
Flue gas/off-gas
Cleaned flue gas
Absorption liquor
Sulfur products
Other systems
Figure 4.
General Dual Alkali FGD Process
-------
Design Considerations
A complete discussion of the design
considerations involved in the
construction and operation of a dual
alkali FGD system is beyond the scope
of a summary report. The following
discussion, however, contains
sufficient information to permit a
macroscopic analysis of the process.
Pretreatment
It is sometimes necessary to pretreat
the flue gas to remove chlorides
and particles (fly ash).
Chlorides are undesirable because
they are absorbed from the flue
gas and accumulate in the scrubbing
liquor, causing stress corrosion of
the scrubbing equipment. Further-
more, chloride buildup as sodium
chloride (NaCI) in the scrubbing
liquor can lower the solubility of
sodium sulfate (Na2S04), which can
precipitate as scale.2
As a rule the only mechanism
for chloride removal is by occlusion
with the solid wastes. When the
wastes are washed to recover
sodium, however, chlorides are
returned to the system. A prescrubber
for chloride removal may be
desirable, therefore, at installations
firing high chloride coals in spite
of associated water balance
problems. As an alternative, down-
stream equipment can be made
from materials resistant to chloride
attack, such as 317 stainless steel
or lined carbon steel.
Particles from the flue gas can
accumulate in the recirculating
scrubbing liquor and interfere with
SO2 absorption. Conventional
particle control devices include
ESP's, wet scrubbers, cyclone
collectors, and fabric filters.
The CEA/ADL pretreatment system
at Gulf Power Company's Scholz
plant consists of a high efficiency
ESP and a variable throat venturi
prescrubber. The venturi can
be used for particle control and
for absorption; it can be operated on
a separate liquor loop or in series
with the absorber liquor loop.
Experiments conducted at the Scholz
plant during June and July of
1976 assessed the short term effects
of high inlet particle loadings
on system performance. These
tests were performed with the ESP
entirely out of service. The particle
removal efficiency of the venturi
prescrubber operating in series with
a tray absorber was 98.9 percent.3
The CEA/ADL pretreatment system
at Louisville Gas and Electric's
Cane Run Station does not include
a prescrubber. Flue gas from the
boiler ESP is routed directly to a
presaturation chamber in the
absorber.
Absorber
Of primary consideration in absorber
design are the quantity of gas to
be cleaned, the sulfur dioxide
and particle content, and the desired
SO2 removal efficiency. These
factors determine the type of
absorber, the number and size of
absorbers, and the number of
gas/liquid contact stages. Thus, they
affect capital and operating costs
significantly.
Two kinds of absorbers are commonly
used in dual alkali FGD: the venturi
scrubber and the tray tower absorber.
The venturi is desirable where
particle control is necessary because
it can be used alone for both
sulfur dioxide and particle removal.
If further S02 removal (greater
than 95 percent) is desired, a venturi
can be used in combination with
a tray tower absorber at a small
increase in cost.4
If only a modest reduction in particle
loading is needed, the absorber
will function primarily as an SO2
removal device rather than a high
energy particle scrubber. In cases
of this kind, low-energy tray tower
absorbers are suitable.4
-------
The CEA/ADL scrubbing system at
the Scholz plant consisted of a
variable throat venturi operating in
series with a multitray-tower
absorber. The absorbers were
designed to receive 75,000 actual
ft3/min (2,100 actual m3/min) of
flue gas. The system was installed
on the No. 1 boiler and handled
about 50 percent of the flue gas
generated by firing 0.9-5.1 percent
sulfur coal.
A venturi scrubber was included at
the Scholz plant to test simultaneous
particle and S02 removal. Later
the system was modified to operate
with the venturi alone by diverting
the regenerated scrubbing liquor
(usually fed to the tray tower
absorber) to the venturi.
The General Motors scrubbing system
at Parma was designed primarily
to remove sulfur dioxide rather than
particles. To meet this criterion
the absorbers were designed as
columns with three valve trays.
Four absorbers operating in parallel
received 240,000 actual ft3/imin
(6,800 actual m3/min) of flue
gas from four boilers firing 2.5 percent
sulfur coal. Each absorber was
designed to control the emissions
from its boiler; there was no crossing
over from one absorber to another
while the boiler was operating. As
is typical of industrial installations,
the General Motors boilers operated
at variable rates, responding not
only to process requirements that
vary by hour, shift, and day of the
week, but also to seasonal space-
heating requirements. Accordingly,
there was variation m the number
of absorbers used.
Two parallel tray tower absorbers
have been installed on Louisville
Gas and Electric's Cane Run No. 6
boiler, which fires coal containing
3.5-6.3 percent sulfur. The de-
sign gas flow rate to the system is
1.065 X 106 actual ft3/min (30,1 60
actual m3/min), equivalent to
the boiler peak load capacity of
300 MW. Each absorber is sized
to handle 60 percent of the design
gas flow rate. If necessary, the
system can be turned down to
20 percent of peak load. At levels
less than 60 percent of the design
capacity, the system can be operated
with one or both absorbers.
The S02 removal efficiency of
a given absorber depends on several
operating conditions. Theoretically,
the upper limit of removal efficiency
is determined by the equilibrium
partial pressure of the SO2 above
the scrubbing liquor. Contemporary
computer models can use extensive
vapor pressure data collected
in the 1930's for SO2 over sulfite/
bisulfite solutions.5-6 The data
are modeled over the full range
of liquor compositions expected
during various operating conditions.
Standard chemical engineering
practices can then be used in design-
ing the appropriate absorber.5
In practice, S02 removal efficiency
also depends on a number of other
variables of operation, including
the S02 concentration of the
inlet flue gas, the pH of the scrubbing
liquor, liquid to gas (L/G) ratios, and
the absorber pressure drop. In
general, SO2 removal efficiency
increases with greater absorber pres-
sure drops and higher L/G ratios,
although the latter may not hold
at very high gas rates.7-8 Higher
removal efficiencies are also possible
with higher inlet S02 concentrations,
because these conditions create a
greater driving force for S02 mass
transfer.
In most high sulfur coal applications,
removal efficiencies approaching
99 percent can be achieved con-
tinuously in low energy tray
tower absorbers.4 The Cane Run
absorbers are designed for 95 percent
removal efficiency when handling
coal containing more than 5 percent
sulfur (inlet SO2 concentrations
greater than 3,471 ppm). If the coal
contains less than 5 percent sulfur,
the Cane Run absorbers will con-
trol S02 emissions to less than
200 ppm (dry).
The General Motors tray tower
absorbers at Parma demonstrated
SO2 removal efficiencies of 90
percent over extended periods. These
high rates were accomplished at
fairly low S02 inlet values of
650-1,600 ppm, which were caused
by the high excess air rates of the
boilers and the intermediate levels of
sulfur (2.5 percent) in the coal burned.
At the Scholz plant, removal efficien-
cies averaging 95 percent were
achieved with the venturi and tray
tower operating in series. When the
venturi was used alone, a removal
efficiency of about 90 percent
was attained.
For a given scrubber configuration
and inlet SO2 level, the SO2 removal
efficiency can be adjusted by
varying the operating pH of the
scrubbing liquor.4 Below pH 6.0,
scrubber efficiency drops rapidly.7-8
Above pH 8.5, removal efficiencies
are high; however, if the pH is
greater than around 9.0, carbon
dioxide absorbed from the flue gas
may precipitate as CaC03 scale.
Scrubber pH should be kept between
6 and 9, therefore, for most efficient
SO2 removal without risk of scale
formation. Acceptable removal can
be obtained with a pH of approxi-
mately 6.5. At this value, the
scrubbing solution is highly buffered
and responds well to rapid changes
in flue gas inlet conditions.9
Sulfate in the scrubbing liquor—
formed by oxidation of sulfite or
bisulfite as described in Equations 6
and 7—has received a great deal
of attention. This product does
not participate in S02 removal and is
difficult to convert back to an
active form. Unless sulfate is
removed from the system at a rate
roughly equal to that at which it
forms, it will precipitate as calcium
sulfate scale or sodium sulfate crystals.
Considerable design attention
is given to methods of reducing
sulfate concentrations in the
8
-------
Dual alkali reactors and regeneration slurry system thickener
scrubbing liquor. One approach is to
limit liquor residence time in the
scrubber. This practice effectively
reduces oxidation of sulfite to
sulfate, because up to 90 percent
of the oxidation in a dual alkali
system typically occurs in the
scrubber.9
Reheating
It may be necessary to reheat the
desulfurized flue gas to prevent
water vapor from condensing after
the gas is ejected from the stack to
the atmosphere. Indirect reheating
uses steam to heat a fresh air
stream, which is then mixed with the
treated flue gas. This method was
used at General Motors' Parma plant.
The Cane Run dual alkali system
uses a direct method of reheating,
in which hot gas from fuel oil
combustion is injected into the
stream of treated flue gas.
In some instances, untreated flue
gas can provide all or part of the
reheating. This approach is possible
where SO2 removal efficiencies
are high and overall S02 emissions
regulations are met by the com-
bined stream of treated and
untreated gas.4
Absorbent Regeneration
In designing for absorbent regenera-
tion it is important to consider
precipitation of absorbed S02 as
calcium solids, sulfate removal,
and whether lime or limestone is to
be the regenerant. These considera-
tions affect the configuration of
the chemical mix reactor tanks
and the reaction time needed.
Sulfate can be removed from dilute
dual alkali systems in the form of
gypsum (CaS04 • 2H20) according
to the following reaction:
SC>42 + Ca(OH)2 + 2H20 t^ 20H~
+ CaS04-2H2Oi (13)
-------
In concentrated systems, sulfate ions
coprecipitate with calcium sulfite.
The waste product in these systems
is a mixed crystal of calcium sulfate
and calcium sulfite. The relative
amounts of calcium sulfate and
calcium sulfite that are formed de-
pend on the ratio of sulfate to
sulfite in the reactor liquor. It is
reported that enough sulfate can be
removed to accommodate oxidation
rates as high as 25-30 percent
of the S02 absorbed. Moreover, the
reaction self-adjusts over a range
of oxidation rates, as shown in
the prototype application of the
CEA/ADL system at the Scholz plant10
Sulfate ion also can be removed
in the presence of high sulfite
concentrations when sulfuric acid
is added to the system in a separate
reactor. The acid dissolves calcium
sulfite, increasing the concentra-
tion of calcium ions in solution
sufficiently to exceed the solubility
product of calcium sulfate. A
solution pH of around 2.5 to 3.0 is
needed for the desired sulfate
removal. The added sulfuric acid
does, however, significantly
increase lime consumption, and
therefore may be economically
undesirable in applications with very
high oxidation rates.5
The choice between a dilute system
producing gypsum and a concen-
trated system producing mixed
crystals depends primarily on oxida-
tion rates in the scrubber. As a rule
concentrated systems are best
suited to applications with low
oxidation rates: typically those in
which medium to high sulfur coals
are burned with little excess air.
Dilute systems are preferable where
oxidation rates are high. Burning
low sulfur coal promotes high
rates of oxidation because of the
high 02/SOX ratio in the flue
gas. Boiler firing using excess air
has a similar effect.
There are other factors to be weighed
in considering dilute dual alkali
systems. Because the concentration
of active alkali is lower, more
scrubbing liquor must be circulated
and system costs rise accordingly.
Also, the concentration of dissolved
calcium is higher than in concen-
trated systems because calcium
level in dilute systems must be
raised before calcium sulfate is pre-
cipitated. The higher calcium level
is conducive to scaling in the
scrubber. The problem of gypsum
scaling in dilute systems can be
controlled by softening the scrubber
liquor with sodium sulfite or sodium
carbonate.
Efforts have been made to use
limestone rather than lime as a
regenerant in dual alkali systems.10-11
Limestone is available near most
industrial sites and is considerably
cheaper than lime. Laboratory
and pilot studies, however, have
demonstrated major problems as
well as success in using limestone.
Impurities in limestone, especially
magnesium, seriously impair the
settling properties of the solids.
Also, because limestone is less
reactive than lime, a longer reaction
time is needed. Calcium utilization
rates are also lower, and the pro-
portion of sulfate in mixed-crystal
solids is smaller than in systems
regenerated with lime.
Raising temperatures to increase
reaction rates might improve
the settling and dewatering properties
of solids. Furthermore, magnesium
could be precipitated from a
slip stream. Such added complexity,
however, would probably nullify
the economic advantage of using
limestone instead of lime.10 For
the present, therefore, lime appears
to be superior to limestone as a
regenerant for dual alkali systems.
EPA intends to conduct prototype
testing of a 20-MW limestone
system at the Scholz plant in 1980.
Solid/Liquid Separation and
Dewatering
Major design considerations for
the solid/liquid separation and
dewatering system include ease of
solids settling and sodium con-
servation. Ease of settling is important
because nonsettling solids create
plugging problems and interfere
with equipment operation. Sodium
conservation is desirable and
should focus on the dewatering and
filter cake washing steps.
The key to meeting these goals is
the formation of large, compact
crystals in the reactor. The kind of
crystal formed depends largely
on features of the regeneration sys-
tem. The gypsum precipitated in
dilute dual alkali systems is more
easily dewatered than the finer
grained mixed crystal produced by
concentrated systems.
Low concentrations of oxidizable
sulfur (below 0.1-0.2 molar) and
sulfate levels of 0.50 to 0.75 molar
were found by tests to promote good
crystal growth in dilute systems.
Solids were recycled to a level of
4 to 5 percent to enhance crystalliza-
tion. Design for solids recycling
is widely recommended,5'7 but it
does increase the demands on the
dewatering system.
At General Motors' Parma plant,
the filter cake was low in solids
because of excess calcium hydroxide,
low gypsum content, and high
sulfite content. To increase the
solids air was sparged into the reactor
system to oxidize the sulfite not
oxidized in the scrubber system.
Because this step lowered system
pH, part of the underflow from the
first clarifier was recycled to
allow the excess hydroxide in the
solids to raise the pH. (Recycle
also provides seed crystals, which
reduce the possibility of sulfate
supersaturation and subsequent
scaling.) Unfortunately, the high
solids recirculation rate overloaded
the clanfiers; the overload resulted
in frequent solids overflow and
interfered with settling.12
Solids recycling was later eliminated
at Parma. The scrubbing liquor
was oxidized upstream of the
reactors. Slaked lime was added
constantly to the first of the two
10
-------
reactors, increasing reaction
time. Hydroxide was measured
frequently to avoid feeding excess
lime. As a result, the solids were
consistently low in moisture content.
Many dilute dual alkali systems,
including Parma, use two clarifier-
thickeners, with sodium makeup
going to the second.5-12'13 This
practice separates the softening
reaction from the regeneration
reaction.
Sodium consumption can be
measured in moles of sodium con-
sumed per mole of sulfur removed
from the system. A value of 0.05
mole of sodium makeup per mole of
sulfur removed appears to be a
reasonable design target based
on present U.S. technology.5
Thickened slurry from the clarifier-
thickener is routed to a rotary
drum vacuum filter where the
solids are filtered to a cake contain-
ing 50 to 80 percent solids. On
the filter the cake is washed with
water sprays to remove occluded
soluble salts, and the salts are
returned to the system along with the
clarified liquor from the thickener.
This technique reduces sodium
losses and minimizes sodium
carbonate makeup.
Filters do not as a rule require ex-
cessive maintenance. The rotary
drum vacuum filter at the Scholz
plant, however, was the greatest
source of trouble in the system.3
The filter, made of plastic and
fiberglass, was subject to erosion,
and to frequent failures at stress
points. These problems can be
avoided if stainless steel or other
suitable material is used in filter
construction or, more generally, by
better filter design.
11
-------
Environmental
Considerations
Dual alkali FGD can achieve S02
removal efficiencies of greater
than 90 percent. The CEA/ADL
system at the Scholz plant, for
example, has operated at S02
removal efficiencies of about 95
percent and has demonstrated the
capability for more than 99 percent
removal.3
The process is also capable of
particle removal. The venturi
scrubber at the Scholz plant averaged
a 98.9-percent particle removal
efficiency with the ESP out of service.
The dual alkali process can, however,
create other pollution problems.
Because the scrubbing system
contains soluble salts, it must be
operated closed loop, with no liquid
effluent, to prevent water pollution.14
That is, water cannot be added to
the system at a rate exceeding
normal water losses. Fresh water
is added for many purposes, including
flue gas saturation, pump seals,
demister washing, slurry makeup,
waste product washing, and
tank evaporation. Water should
leave the system, however, only by
evaporation in hot flue gas, by
occlusion with the solid waste, and
as water of crystallization in the
solid waste. When a prescrubber
is used, the hot flue gas evaporates
water in a liquor loop separate
from the rest of the dual alkali system.
Therefore water is lost in the
rest of the system only with the
solid waste. This relatively small
loss may not allow enough water
to be added for such practices
as waste product washing, demister
washing, pump seals, and slurry
makeup.
The solubles can create water
pollution problems.15 The salts
can be leached from the disposed
sludge by percolation and water
runoff, and can contaminate surface
or ground water. Techniques (such
as washing) to reduce the amount
of soluble sodium salts have
been tested. At Parma soluble
sodium salts in the filter cake dropped
from 2.4 percent to 1 percent as
a result of effective cake washing.
It appears, however, that there will
be inevitably higher concentra-
tions of these salts in sludge from
dual alkali FGD than in sludge
from lime or limestone processes.
Sludge fixation can alleviate
this problem by decreasing the
permeability of the sludge.5 Dual
alkali sludge can be fixed using
essentially the same techniques as
those applied to lime and limestone
sludges. One approach is to mix
dry fly ash with the waste to raise
the solids content sufficiently
for compaction.
Alternatively, untreated sludge may be
disposed of in well-designed, lined
ponds to prevent seepage.15 The
sludge at the Scholz plant was
dumped from a truck into a narrow
pond equipped with a polyethylene
liner underlaid with natural clay.
Sludge disposal also can cause
land use problems because non-
settling sludges make land reclama-
tion difficult.15 Although the
waste product may appear dry, the
highly porous or spongelike calcium
sulfite crystals can retain a great
deal of water. These rather fragile
crystals break under pressure and
release the water. Thus, calcium
sulfite sludge is thixotropic;
that is, it tends to become fluid with
vibration or stress.
To prevent problems of water
pollution and land use, therefore,
an environmentally acceptable solid
waste should be nontoxic and
nonthixotropic. It should be low in
soluble solids and moisture, and
its compressive strength should
be high.5
12
-------
Status of Development
The dual alkali FGD process was
developed to overcome disad-
vantages inherent in lime and lime-
stone scrubbing (such as scaling)
while retaining the cost advantages
of a throwaway system. Several
variations on dual alkali systems
have been tested and studied
extensively during the past few years
by equipment vendors, potential
users, and EPA.13
Table 1 summarizes the operating
and planned full-scale dual alkali
systems in the United States. Of
these, 11 are operational (1,473-MW
equivalent) and 7 more are in the
design or construction stage (1,852-
MW equivalent). Installation of
General Motors' dilute mode dual
alkali FGD process at Parma in 1 974
represented the first full-scale
industrial boiler application of this
technology in this country.
In Japan, the Showa Denko KK
Company and Kureha Chemical
Industry Company/Kawasaki Heavy
Industries have developed sodium-
based dual alkali FGD systems
through pilot plant testing, and now
operate relatively large full-scale
systems. By the end of 1977,
approximately 47 dual alkali FGD
plants, with an average plant capacity
of 96-MW, were operating in
Japan.16 Approximately 45 percent
of the total dual alkali FGD plant
capacity in Japan is for utility
boilers (primarily oil fired); the
remaining capacity is for industrial
boilers, sintering plants, smelters,
and sulfunc acid plants.4
Japanese dual alkali FGDtechnology
is characterized by processes that
have unlimited oxidation tolerance
and that use limestone as a re-
generant.
Sodium storage silo and dead storage pond
13
-------
Table 1 .
Dual Alkali FGD Installations in the United States3
Company and location
Operating:
Caterpillar Tractor Co..
East Peona IL . . , ...
Joliet IL .
Mapleton IL. . . ...
Morton IL
Mossville IL . . . . . ...
Central Illinois Public Service, Newton IL .
Firestone Tire and Rubber Co., Pottstown PA. .
PMC/Industrial Chemical Division, Modesto CA . .
General Motors Chevrolet, Parma OH
Louisville Gas and Electric Louisville KY . . .
Southern Indiana Gas and Electric, West Frank-
lin IN
Under construction or in design:
Arco Polymers Inc Monaca PA
C.A.M. (Carbide-Amoco-Monsanto), Houston
TX . .
Caterpillar Tractor Co Mapleton IL
Dupont Inc Athens GA ... .....
Gnssom Air Force Base, Bunker Hill IN . ...
Northern Indiana Public Service, Wheatfield IN.
Schahfer No 17
Schahfer No 18
Completed, not operating: U.S. Gypsum Corporation,
Oakmont PA ....
Feed gas origin
Coal-fired industrial
boiler system
Coal-fired industrial
boiler system
Coal-fired industrial
boiler system
Coal-fired industrial
boiler system
Coal-fired industrial
boiler system
Coal-fired utility
boiler
Coal-fired industrial
boiler
Reduction
kiln
Coal-fired industrial
boiler
Coal-fired utility
boiler
Coal-fired utility
boiler
Coal-fired industrial
boiler
Industrial boiler
system
Coal-fired industrial
boiler system
Coal-fired industrial
boiler system
Coal-fired industrial
boiler system
Coal-fired utility
boiler
Coal-fired utility
boiler
Coal-fired industrial
boiler system
% S
3.2
3.2
3.2
3.2
3.2
4.0
2.5-3.0
(d)
2.5
4.8
45
3.0
NA
3.2
1.5
3.0-3.5
3.2
3.2
NA
Gas
volume
treated
(1 ,000
stdft3/mm)
210
67
131
38
140
1,150
8.07
20
1284
554
500
305
1,300
105
280
32
842
842
19.3
MW
equivalent
105
34
65
19
70
575
4
10
64
277
250
152
650
52
140
16
421
421
10
Sulfur
removal
(%)b
90
90+
90
90
90+
95
905
95+
90
95
85
90
NA
90
90
NA
NA
NA
NA
Active
alkahc
C
D
C
C
C
C
C
C
D
C
C
C
NA
C
NA
C
NA
NA
D
Startup
date
1978
1974
1979
1978
1975
1979
1975
1971
1974
1979
1979
1980
1984
1980
1985
1979
1983
1985
NA
aAs of April 1979 for industrial boiler systems, and as of December 1979 for utility boilers.
bFigures for plants not yet operating represent design targets.
CC = concentrated; D = dilute
dNot applicable
Note —NA = data not available.
SOURCES: Tuttle, J., A. Patkar, S. Kothari, D Osterhout, M. Hefflmg, and M. Eckstein, EPA Industrial Boiler FGD Survey: First Quarter 1979, EPA 600/7-
79-067b, Apr. 1979. Smith, M., M. Meha, and N Gregory, EPA Utility FGD Survey: October-December 1979, EPA 600/7-80-029a, Jan. 1980.
14
-------
System Requirements
Raw Materials and Utilities
The dual alkali FGD process con-
sumes sodium, usually in the form
of soda ash (Na2CO3), and calcium,
in the form of lime or limestone.
Its utility needs are for energy
and process water. Table 2 shows
the estimated annual raw material and
utility requirements for three dual
alkali systems.
Sodium must be added to the
system to replace that lost in the
washed cake. Sodium consumption
is a minor factor in system operating
cost, but can have significant
environmental consequences if the
sodium can be leached from the
waste. Based on present U.S. tech-
nology, a value of 0.05 mole sodium
makeup (0.025 mole Na2C03) per
mole of sulfur removed is a reasonable
design target for a concentrated
system burning coal with more than
3 percent sulfur and cleaning
fuel gas with a relatively low oxygen
content. In Japan, makeup values
have been reported as low as
0.02 mole of Na per mole of
sulfur removed. During the early
months of operation at Parma, when
the filter cake was not washed,
the sodium loss was above 0.1 mole
Na2 per mole of SOX in the cake.
After cake washing was initiated, the
loss was as low as 0.028 mole Na2
per mole of SOX for an extended
1-month average.12
Calcium consumption is specified in
terms of calcium (or lime or lime-
stone) stoichiometry as moles of
calcium added per mole of sulfur
removed. A calcium stoichiometry of
0.98 to 1.0 is a reasonable design
target for concentrated dual
alkali systems.5 The overall lime
stoichiometry at the Scholz plant has
been 0.95 to 1.0 mole lime per mole
of sulfur removed.
Low energy consumption is a
major advantage of dual alkali sys-
tems. Design targets in the range
of 1 to 2 percent of power plant
generation are possible if the need
for stack gas reheat is excluded
and efficient upstream particle
collection is assumed. An increase
of 50° F (28° C) from reheating
the stack gas increases the design
target to roughly 3 percent of the
total power generated.5 The
CEA/ADL prototype unit at Scholz
consumed energy equivalent to
about 2.5 to 3.0 percent. The
process at Cane Run (excluding
reheat) will require about 1.2 percent
of the peak power generated by
Unit No. 6. Approximately 60 percent
of this energy is needed for the
booster fans, 10 percent is for
reheater fans, and 30 percent is for
the rest of the system. Including
oil for reheat, the total energy
requirements for the system amount
to about 3.0 percent of the peak
power generated.4
Table 2.
Estimated Annual Raw Material and Utility Requirements for the Dual Alkali
FGD Process
Requirement
New coal-fired plant
300 MW
500 MW 1,000 MW
Raw materials (1,000 tons).
Lime . .
Utilities-
Reheat (106 Btu)
Electricity (106 kWh)
47 02
3 85
296 000
120
54
78.22
641
493 000
190
90
156.02
12 78
986 000
393
179
Note.—3 5% sulfur coal; 90% S02 removal, 7,000 h/yr operating time; on-site solids disposal, stack
gas reheat to 175° F.
SOURCE: PEDCo Environmental, Inc., computerized FGD cost program, July 19, 1979.
15
-------
Flue gas/off-gas
Cleaned flue gas
Absorption liquor
Sulfur products
Other systems
Na2C03Q
storage and
mixing
187 ft
Figure 5.
Concentrated Dual Alkali Installation Requirements
Energy requirements for reheat can
be reduced by using methods
other than oil firing. In some cases,
untreated flue gas can provide
all or part of the reheating.
Because the dual alkali process
operates in a closed loop to avoid
water pollution, water is lost only
by evaporation and by occlusion
or crystallization with the solid waste.
A recent study estimates an annual
water rate of 230 X 106 gal/yr
(871,000 m3/yr) for a 500-MW unit
operating 7,000 h/yr. This rate
is equivalent to 0.066 gal/kWh
(0.90 rrvYkJ).17
Installation Space and Land
A dual alkali FGD system requires
equipment similar to that required
for a lime or limestone scrubbing
system. Retrofit problems might be
less severe with dual alkali systems
because the flue gas contact
equipment can be smaller and re-
generation can be carried out
at some distance from the scrubber.15
The installation space required for
a lime scrubbing system on a new
500-MW unit has been estimated at
about 41,200 ft2 (3,800 m2), of
which 42 percent represents the
scrubber system and 58 percent
represents the materials handling
and feed preparation system.18
A comparable dual alkali system
should require space similar to or
perhaps somewhat less than this
estimate.
Figure 5 shows a general plan for
a concentrated dual alkali installation.
Operation m a dilute mode would
require a second clanfier-thickener
roughly the same size as the first.
A significant land area is needed
for sludge disposal. Less land
is required than for lime or limestone
systems, however, because dual
alkali systems use less calcium
and produce dryer sludges. It has
been estimated that about 0.2 acre-
ft/MW-yr (246.6 m3/MW-yr) would
be required to dispose of sludge
from a dual alkali S02 removal
process.19
16
-------
Costs
Dual alkali FGD systems now appear
to be economically competitive
with wet alkali lime/limestone
slurry scrubbing systems for reasons
that include:
• Lower scrubber L/G ratio
• Lower scrubber pressure drop
• Simpler scrubber design
• Less exotic construction materials
• Solid waste with better handling
properties
An FGD system can vary widely in
estimated and actual costs depending
on the assumptions, conditions
of operation, options included,
degree of redundancy, and other
factors.
Table 3 presents estimated annual
operating costs for a dual alkali
FGD system. The table identifies
specific components and gives
examples of the contribution of each
component to the annual operating
cost.
Table 4 shows the capital and
annual operating costs for dual alkali
systems installed on different
Calcium sulfite/sulfate solids separation and dewatering tank
17
-------
Table 3.
Annual Operating Costs for a Dual Alkali FGD System on a New 500-MW Boiler
Annual operating costs
Component
Annual quantity
Unit cost ($)
$1,000 mills/kWh
Direct costs:
Delivered raw materials:
Lime 63,600 tons 42.00/ton 2,671.20 0.763
Sodium carbonate 6,060 tons 90.00/ton 545.40 0 156
Total raw materials 3,216.60 0.919
Conversion costs.
Utilities:
Steam 489,300 X 106 Btu 2.00/106 Btu 978.60 0.280
Process water 241.5X106gal 012/1,000 gal 29.00 0.001
Electricity 29.1 X 10e kWh 0.029/kWh 843.90 0.241
Total utilities 1,851.50 0522
Operating labor and supervision 34,500 man-hours 12.50/man-hour 431.30 0.123
Maintenance: labor and material 1,027.60 0.294
Analyses 4,560 man-hours 17.00/man-hour 77 50 0 022
Total conversion costs 3,387.90 0.961
Total direct costs 6,604.50 1.880
Indirect costs:
Capital charges:
Depreciation, interim replacements, and insurance at 6% of total de-
preciable investment 2,911.80 0832
Average cost of capital and taxes at 8.6% of total investment. . 4,347.40 1.242
Overheads:
Plant, 50% of conversion costs less utilities 768.20 0.219
Administrative, 10% of operating labor 43.10 0.012
Total indirect costs 8,070.50 2.305
Total annual operating costs 14,675.00 4.185
Note.—Midwest plant location, 1980 revenue requirements. 30-yr remaining plant life. 7,000 h/yr operating time. 1.5 X 10 tons coal burned,
9,000 Btu/kWh, 3.5% sulfur. Stack gas reheat to 175° F. 34,560 tons/yr sulfur removal. 144,690 tons/yr solids disposal. Investment and revenue require-
ment for removal and disposal of fly ash excluded. Total direct investment, $26,750,000; total depreciable investment, $48,530,000; total capital
investment, $50,551,000. Meets emission regulation of 1.2 Ib SO2 per 106 Btu.
SOURCE: Torstrick, R. L, L. J. Henson, and S. V. Tomhnson, "Economic Evaluation Techniques, Results and Computer Modeling for Flue Gas
Desulfunzation," In Proceedings: Symposium on Flue Gas Desulfunzation—Hollywood, FL, EPA 600/7-78-058a, Mar. 1978
sizes and types of boilers. The
costs are subject to variation and
depend on a number of site-specific
factors. Any specific situation
can be compared with the factors
used as a base. Each location
should be evaluated for availability
and cost of raw materials, energy
sources, physical plant, disposal
criteria, and other environmental
considerations. For example, the
total estimated capital cost for
the dual alkali system on Louisville
Gas and Electric's Cane Run Unit
No. 6 is $17,379,000 ($57.9/kWh).
Total annual operating cost, with
reheat, is estimated at $5,142,600
(3.27 mills/kWh). These costs are
calculated in 1977 dollars for a
300-MW plant (gross peak load)
with 60 percent annual load,
9,960 Btu/kWh (10,510 kJ/kWh),
3.8 percent sulfur coal, and 94.2
percent S02 removal.4
18
-------
Table 4.
Estimated Capital and Operating Costs for Dual Alkali FGD
System characteristics
Size
(MW)
200
200
500
500
500
500
1,000
1,000
500
Application
Existing
New
Existing
New
New
New
Existing
New
Existing
Fuel
Type
Coal
Coal
Coal
Coal
Coal
Coal
Coal
Coal
Oil
% S
3 5
3 5
3.5
20
3 5
50
3 5
3 5
2 5
Plant life
(yr)
20
30
25
30
30
30
25
30
25
S02
removal0
S
S
S
S
S
S
S
S
R
Total capital
investment3
$106
26.01
2548
53.67
42.11
5055
57 58
85.49
7902
4026
$/kW
130.0
127.4
107.4
84.2
101.1
115.2
85.5
790
80.5
Annual operating costs'"
$106
7.553
7.169
15442
1 1 335
14676
17.742
25 751
24 148
11 128
mills/kWh
5.40
5.12
441
324
4.19
507
3 68
3.45
3.18
a1 979 dollars. Minimum in-process storage, only pumps are spared
b1980 revenue requirements Power unit operating 7,000 h/yr.
°S = meets 1 2-lb S02/106 Btu heat input emission regulation R = 0.8-lb S02/106 Btu
Note —Midwest location On-site sludge disposal No fly ash removal and disposal No
SOURCE: Tomlmson, S. V , F M. Kennedy, F A Sudhoff, and R L Torstnck, Definitive SOK
Citrate FGD Processes, EPA-600/7-79-1 77, Aug 1979
heat input allowable emission.
overtime pay.
Control Process Evaluations • Limestone, Double Alkali, and
19
-------
References
1 LaMantia, C. R., R. R. Lunt, J. E.
Oberholtzer, E. L. Field, and N.
Kaplan. "EPA-ADL Dual Alkali
Program—Interim Results." In Pro-
ceedings: Symposium on Flue
Gas Desulfurization, Atlanta.
EPA-650/2-74-126. Nov. 1974.
2Bloss, E. H., J. Wilhelm, and
W. J. Holhut. "The Buell Double-
Alkali S02 Control Process." In
Proceedings: Symposium on Flue
Gas Desulfurization, New Orleans.
Vol I. EPA-600/2-76-136a. 1976.
3Rush, R. E., and A. E. Reed. "Op-
erational Experience With Three
20 MW Prototype Flue Gas
Desulfurization Processes at Gulf
Power Company's Scholz Electric
Generating Station." In Proceedings:
Symposium on Flue Gas Desulfur-
ization—Hollywood, FL Vol. I.
EPA-600/7-78-058a. Mar. 1978.
*VanNess, R. P., R. C. Somers,
T. Frank, J. M. Lysaght, I. L.
Jashnani, R. R. Lunt, and C. R.
LaMantia. Executive Summary
for Full-Scale Dual Alkali Demon-
stration at Louisville Gas and
Electric Co. —Preliminary Design
and Cost Estimate. EPA-600/7-78-
01 Oa. Jan. 1978.
5Kaplan, N. "Introduction to Double
Alkali Flue Gas Desulfurization
Technology." Proceedings:
Symposium on Flue Gas Desulfur-
ization, New Orleans. Vol. 1.
EPA-600/2-76-136a. Mar. 1976.
6Lowell, P. S., D. M. Ottmers, T. I.
Strange, K. Schwitzgebel, and D. W.
DeBerry. A Theoretical Description
of the Limestone Injection Wet
Scrubbing Process. Austin TX,
Radian Corporation, June 1970.
7Cornell, C. F., and D. A. Dahlstrom,
"Performance Results on a 2500
ACFM Double-Alkali Plant."
Presented at the 66th Annual AlChE
Meeting, Philadelphia PA, Nov. 1973.
8Phillips, R. J. Sulfur Dioxide Emis-
sion Control for Industrial Power
Plants. Warren Ml, General Motors
Technical Center, 1971.
9LaMantia, C. R., R. R. Lunt, R. E.
Rush, T. M. Frank, and N. Kaplan.
"Operating Experience—CEA/ADL
Dual Alkali Prototype System
at Gulf Power/Southern Services,
Inc. In Proceedings: Symposium
on Flue Gas Desulfurization, New
Orleans. Vol. I. EPA-600/2-76-136a.
Mar. 1976.
10LaMantia, C. R., R. R. Lunt, J. E.
Oberholtzer, E. L Field, and J. R.
Valentine. Final Report: Dual
Alkali Test and Evaluation Program.
3 vols. EPA-600/7-77-050a-c.
May 1977.
"Oberholtzer, J. E., L N. Davidson,
R. R. Lunt, and S. P. Spellenberg.
Laboratory Study of Limestone
Regeneration in Dual Alkali
Systems. EPA-600/7-77-074.
July 1977.
12lnteress, E. Evaluation of the
General Motors' Double Alkali SO2
Control System. EPA-600/7-77-
005. Jan. 1977.
13Kaplan, N. "An Overview of
Double Alkali Processes for Flue
Gas Desulfurization." In Proceed-
ings: Symposium on Flue Gas
Desulfurization, Atlanta. EPA-650/
2-74-126. Nov. 1974.
20
-------
14Ellison, W., S. D. Heden, and
E. G. Kominek. "System Reliability
and Environmental Impact of S02
Scrubbing Processes." Presented
at the Coal and Environment
Technical Conference of the
National Coal Association,
Louisville KY, 1 974.
15Ponder, W. H. "Status of Flue
Gas Desulfurization Technology for
Power Plant Pollution Control."
Presented at the Thermal Power
Conference, Washington State
University, Oct. 1 974.
16Ando, J. "Status of SO2 and NOX
Removal Systems in Japan." In
Proceedings: Symposium on Flue
Gas Desulfurization—Hollywood,
FL Vol. 1. EPA-600/7-78-058a.
Mar. 1978.
17Sugarek, R. L, and T. G. Sipes.
Controlling SO2 Emissions from
Coal-Fired Steam-Electric
Generators: Water Pollution Impact.
EPA-600/7-78-045b. Mar. 1 978.
18McGlamery, G. G., R. L. Torstrick,
W. J. Broadfoot, J. P. Simpson,
L. J. Henson, S. V. Tomlinson,
and J. F. Young. Detailed Cost
Estimates for Advanced Effluent
Desulfurization Processes.
EPA-600/2-75-006. Jan. 1 975.
19Princiotta, F. T. "Status of Flue Gas
Desulfurization Technology." In
Symposium on Environmental
Aspects of Fuel Conversion Tech-
nology, St. Louis, Mo. EPA-650/
2-74-118. 1974.
21
-------
This summary report was prepared jointly by the Radian Corporation of
Austin TX and the Centec Corporation of Reston VA. Elizabeth D. Gibson,
Teresa S. Hurley, and Julia C. Lacy of Radian are the principal investigators.
John Williams is the EPA Project Officer. All photographs were taken at
Louisville Gas and Electric Company's Cane Run Plant in Louisville KY.
Comments on or questions about this report or requests for information
regarding flue gas desulfurization programs should be addressed to:
Process Technology Branch
Utilities and Industrial Power Division
Industrial Environmental Research Laboratory
U.S. Environmental Protection Agency (MD 61)
Research Triangle Park NC 27711
This report has been reviewed by the Industrial Environmental Research
Laboratory, U.S. Environmental Protection Agency, Research Triangle Park NC,
and approved for publication. Approval does not signify that the contents
necessarily reflect the views and policies of the U.S. Environmental
Protection Agency, nor does mention of trade names or commercial products
constitute endorsement or recommendation for use.
COVER PHOTOGRAPH: Dual alkali S02 scrubber system, Louisville Gas
and Electric Company
22
------- |