Document Sample
                             THE EARLY YEARS

                                    PRESENTED BY

Donald Shattuck, AQCS Program Manager              Donald.Shattuck@ue-corp.com
Ken Campbell, Project Mechanical Engineer          Ken.Campbell@ue-corp.com
Michael Czuchna, Process Engineer                  Michael.R.Czuchna@ue-corp.com
Mary Graham, Project Manager                       Mary.Graham@ue-corp.com
Andrea Hyatt, Mechanical Engineer                  Andrea.R.Hyatt@ue-corp.com

I. Introduction

The history presented in this paper reflects the experiences of the authors with flue
gas desulfurization (FGD) systems in the initial years of development for the utility

This paper will document some of those experiences to show how the industry has
evolved, and so that young engineers and those new to the industry can better
understand why things are done the way they are today.

II. Drivers

According to the American Meteorological Society, air pollution has been a
recognized problem since the early 14th century. King Edward I of England banned
the burning of sea coal in London because of the smoke it produced. In this country,
both Chicago and Cincinnati enacted clean air legislation in the early 1880s.
In 1952, London, England experienced a smog event that killed over 4,000 people
between December 1st and 15th. On the 7th and the 8th, approximately 900 people
died each day. The smog was so severe that, during its peak, people had to walk in
front of the buses to show the way for the bus drivers. It is probably not a
coincidence that approximately two and one-half years later, the US Congress
enacted the first Clean Air Act1 (CAA). Congress amended the CAA again on
December 17, 1963; October 25, 1965; November 21, 1967; and December 31,
1970. Additional amendments were made to the CAA on November 18, 1971; August
7, 1977; and November 15, 1990.

This paper discusses the electric utility FGD system developments that arose as a
result of amendments to the CAA made between December 31, 1970 and August 7,
1977. The December 31, 1970 CAA amendments caused the US EPA to promulgate
Subpart D – Standards of Performance for Fossil-Fuel-Fired Steam Generators for
Which Construction is Commenced After August 17, 1971 (40 CFR §§ 60.40 - 46).
This is the subpart of the Part 40 – Standards of Performance for New Stationary
Sources (NSPS) that established the emission requirements that all new electric
utility plants had to satisfy. This subpart was applicable to any fossil-fuel or wood-
residue-fired unit of more than 73MW heat input rate that commenced construction
or modification after August 17, 1971. Signing a contract that included a penalty
clause for termination of the contract was sufficient to trigger this regulation.

    PL 80-159 July 14, 1955
PAGE 2 OF 34

For sulfur dioxide, the regulation limited emissions to 0.80 lbs per million Btu (MMB)
for derived liquid fossil fuel or liquid fossil fuel and wood residue. A limit of 1.2
lb/MMB was established for derived solid fossil fuel or solid fossil fuel and wood
residue. In short, new or modified coal-fired steam generators could not emit more
than 1.2 lb SO2/MMB. The regulation did not require the installation of FGD systems;
it merely limited the amount of sulfur dioxide that could be emitted.

One provision of this new regulation became an important consideration in the
design of FGD systems during this period. The reliability of absorbers during the
early years was very low. In order for a steam generator to operate at full load with
one or more absorbers out of service for maintenance, a bypass around the FGD
system was necessary and the US EPA allowed a bypass only when there was a
spare absorber. In an effort to exploit this allowance, early FGD systems commonly
consisted of three 50% absorbers, four 33% absorbers, or five 25% absorbers. One
system, the dual-alkali system at the Schaeffer Station of Northern Indiana Public
Service Co., did not have a bypass and thus avoided the cost of a spare absorber
module but, because they were scrubbing with a solution and not a slurry, their
absorbers were more reliable.

Three things happened as a result of this regulation. Development of FGD systems
began because the technology did not previously exist. Development of the Powder
River Basin (PRB) Region began because the lower-sulfur PRB coal could be fired in
boilers, thus eliminating the FGD system requirement. Finally, states and local
governments, primarily in the west, began developing emission regulations that were
stricter than the federal standard.

Since new generating units were being built in the west, many western states and
local governments felt that emission standards stricter than the federal standards
were necessary. California enacted some of the strictest emission regulations, thus
California utilities began to look to other states for locations in which to build new
generation capacity. Neighboring states and counties noticed what was happening
and promulgated regulations to minimize the influx of new generating stations. Clark
County (Las Vegas), Nevada, for example, did not ban new generating capacity from
coming into the county, but simply required the new generating capacity to meet the
emission requirements of the state where the electricity was to be sold or where the
parent utility company was headquartered.

However, not all states west of the Mississippi River jumped on the bandwagon to
impose emission limits stricter than the federal requirements. One western state
kept the federal requirements, and incurred a lawsuit by the Sierra Club2 that
ultimately resulted in the 1997 amendments to the Clean Air Act. A utility was
building a new coal-fired power plant in the state. The plant was planning on using
PRB coal, which eliminated the need for FGD systems, and planned to use hot-side
electrostatic precipitators, the best in particulate control technology at that time.
Arguably, the air in the area could have been considered pristine. The Sierra Club
filed a citizen suit against the newly-formed US EPA for allowing a plant to be built
that had the potential of contaminating the local air to the level the air in many
eastern cities. The courts agreed with the arguments that the Sierra Club put forth;
however, nothing could be done to require stricter emission controls for this

 Sierra Club v. Ruckelshaus, 344 F. Supp. 253 (D.D.C. 9172), aff’ds per curiam without opinion (D.C. Cir.
1972), aff’d by an equally divided Court, 412 U.S. 541 (1973)
PAGE 3 OF 34

particular plant, since there were no provisions in the CAA allowing EPA to make
regulations addressing the issue.

To meet the requirements that arose from the Sierra Club’s lawsuit, Congress had to
amend the CAA in 1977, adding Part C – Prevention of Significant Deterioration of Air
Quality. Part C has given the US EPA the authority to promulgate the New Source
Review and Visibility regulations. Other changes were also made to the law in the
1977 amendments, and the authors hope to discuss those changes in a future paper.

III. Technologies
A.     Introduction

Chemical and mechanical engineers had never dealt with the challenges they faced in
developing FGD systems for utility plants during this period. Chemical engineers had
never designed process equipment as large as was required, nor had they dealt with
the complex chemistry that occurred in the early FGD systems. Mechanical engineers
were faced with similar challenges. While they had designed equipment for either
acid service or slurry service, they typically had not designed for a combination of
the two. Generally, equipment was larger than what they normally dealt with in
chemical plants and refineries.

It is an understatement to say that the new source performance standards
promulgated by the EPA were technology-forcing. Electric utilities went from having
no scrubbers on their generating units to incorporating very complex chemical
processes. Chemical plants and refineries had scrubbing systems that were a few
feet in diameter, but not the 30- to 40-foot diameters required by the utility
industry. Utilities had dealt with hot flue gases but not with saturated flue gases that
contained all sorts of contaminants. Industry, and the US EPA, has always looked
upon new source performance standards as technology-forcing, because they force
the development of new technologies in order to satisfy emission requirements.
A list of the electric utility FGD systems installed during the early years can be found
in Table 1: FGD Systems Installed in the US from 1970 - 1978. This list is based on
start-up dates rather than contract award dates. Some of the facilities started up
after Subpart D was promulgated by the US EPA, but they were probably designed to
meet Subpart D requirements.

Table 2 presents operating conditions offered by sulfur dioxide removal system
suppliers. Note how low the liquid-to-gas ratios are when compared to today’s
systems. The stoichiometric ratios shown are higher than what is expected today,
while the solids content of the slurry and the recycle tank hold times are lower than
in today’s systems.

B.     Lime/limestone

In 1971, at the time of the new source performance standards for electric utilities,
the capital and operating cost differential between lime and limestone was not as
significant as it is today. Limestone delivered to the site cost between $5 and
$25/ton, depending on the transportation distance. Lime could be obtained for as
little at $30/ton, delivered. The choice between the use of lime or limestone in the
FGD system was based on the availability of the reagent and/or the total evaluated
cost of the system. The cost of preparing lime, exclusive of the reagent cost, is
always less than the cost of limestone since the less expensive slaker is used instead
of a ball mill. Arguments were made during the early years that lime was more
PAGE 4 OF 34

reactive and would do a better job of sulfur dioxide removal than limestone.
Proponents of limestone argued that the increased reactivity was not necessary, and
that limestone would do just as good a job as lime in removing the sulfur dioxide.
During these early years, the chemistry was not well understood and there were
significant operating problems within the absorbers. The removal of sulfur dioxide
was not the problem – scaling was. During the early years, natural oxidation of the
calcium-sulfur compounds was the normal operating mode. The result was a co-
precipitate of calcium sulfate/calcium sulfite that formed rosette structures. Figure
10 shows a rosette structure along with photomicrographs of calcium sulfate and
calcium sulfite crystals. This material would form on anything and everything;
absorber internals, mist eliminators, inside pumps, inside piping. It did not matter
whether the FGD system was lime-based or limestone-based - all were subject to
this scaling.

To respond to the scaling problem, Dravo Lime developed their high-magnesium
Thiosorbic® lime. Dravo built one facility to produce the Thiosorbic® lime near
Maysville, Kentucky. Columbus and Southern Ohio used Thiosorbic® lime at their
Conesville Station scrubber with great success. The material did not add to the scale
already present, and it reduced the amount of scale that was present before its
introduction. Consequently, several other FGD systems along the Ohio River began
using the Thiosorbic® lime.

C.      Double alkali

The double- or dual-alkali scrubbing process was developed to overcome some of the
early difficulties with lime/limestone scrubbing. The double-alkali process had two
distinct advantages over the lime or limestone scrubbing process. First, scaling
problems were eliminated because the double-alkali process used a sodium solution
instead of a lime or limestone slurry in the absorber. Second, the sodium sulfite
(Na2SO3) component in the scrubbing liquor is more reactive than either the calcium
hydroxide (Ca(OH)2) in the lime scrubber or the calcium carbonate (CaCO3) in the
limestone scrubber, resulting in much lower liquid-to-gas ratios in the double-alkali
scrubbers than in either the lime or limestone scrubbers. In early FGD systems, the
liquid-to-gas ratios (L/G) for lime or limestone scrubbers ranged from 40 to 90
gallons of scrubbing liquor per 1,000 ACFM of hot gas entering the scrubber. The
double-alkali scrubbers could achieve high sulfur dioxide removal efficiencies with
L/G ratios of 10 to 20.

As the name implies, there are two alkalis involved in the process. One alkali is used
to create the scrubbing chemical, and the other alkali is used to regenerate the
scrubbing chemical and create an acceptable disposal product.
The sulfur dioxide in the flue gas reacts with the sodium sulfite to form sodium
bisulfite (NaHSO3) according to the following reaction:

Na 2 SO3 + SO2 → 2NaHSO3
The sodium bisulfite-rich liquor is then sent to lime regeneration according to the
following reaction:

2 NaHSO3 + Ca (OH ) 2 → Na 2 SO3 + CaSO3 • 1 / 2 H 2 O ↓ +3 / 2 H 2 0
The right-hand side of the above equation is treated the same as the waste from a
lime or limestone FGD system in that it goes to primary and secondary dewatering
systems where the recovered water is returned to the scrubbing process.
PAGE 5 OF 34

Since there is oxygen in the flue gas, there is a competing reaction which is:

2 Na 2 SO3 + O2 → 2 Na 2 SO4

The sodium sulfate cannot be regenerated and goes out with the other waste
products. Approximately 1 – 5% of the sodium sulfite is oxidized to sodium sulfate.
Sodium is also lost from the system in the form of sodium sulfite that remains in the
entrained water that is in the final waste product. The sodium alkali (either as
sodium hydroxide, sodium bicarbonate, Trona or sodium carbonate) is added to the
system to make up for the sodium lost in the waste stream.

Because the double-alkali scrubber uses a non-abrasive solution (not a slurry) for the
scrubbing liquor, the system could be designed with metal pumps and piping (or
fiberglass piping) instead of the lower-efficiency rubber-lined, open impeller pumps.
While the process had some very distinct advantages, the economics were not
favorable since two expensive alkalis (lime and sodium carbonate or sodium
hydroxide) were used. One double-alkali process is still in operation at the AB Brown
Station in West Franklin, Indiana.

D.     Wellman-Lord

The Wellman-Lord (W-L) process, which was popular in acid plants and refineries,
was installed on only one power plant in the United States as a full-scale, non-
demonstration system. The scrubbing chemistry is similar to that of the double-alkali
process in that it uses a sodium sulfite solution for removing the sulfur dioxide from
the flue gas to produce a sodium bisulfite-rich solution that is sent to a regeneration
process. The sodium sulfite is regenerated in a steam-heated evaporator, a device
that uses a significant amount of low-pressure steam from the steam generator. The
steam-heated evaporator produces a sulfur dioxide-rich gas stream that goes to a
Claus unit that generates elemental sulfur, or to an acid plant that produces sulfuric
acid. A schematic of the system is shown in Figure 1: Wellman-Lord FGD System.
Fly ash and sulfur trioxide are problems in the Wellman-Lord process. The sulfur
trioxide reacts with the sodium sulfite to form sodium sulfate, which has to be
purged from the process as a waste stream. The Wellman-Lord process was
originally designed for use in an acid plant or a refinery, both of which have gas
streams free of the fly ash which can cause significant problems with the
downstream regeneration processes. To minimize these two constituents, a high-
efficiency venturi scrubber, which removed nearly all of the fly ash and some of the
sulfur trioxide, was added upstream of the absorber.

The Wellman-Lord process was originally installed at the Public Service Company of
New Mexico San Juan Station, Units 1 and 2, in Farmington, New Mexico. The
regeneration system for these two units was designed to produce elemental sulfur.
When Units 3 and 4 were built, a regeneration system was installed that produced
sulfuric acid from the four units. Recently, these scrubbers were converted to a
limestone forced oxidation FGD system.

E.     Berbau Forschung

The Berbau Forschung FGD system was a three-step process that produced a
saleable product - either gypsum or sulfuric acid - and was marketed in the United
States by Foster Wheeler. The first step was a dry absorber containing activated
PAGE 6 OF 34

coke that captured the particulate, NOx, SO2, and mercury. The second step was
regenerative - the activated coke was heated which converted the NOx back to N2,
driving off the SO2 as a rich stream, and enabling the mercury to remain with the
coke. The third step took the off-gas from the regenerator and produced the desired
saleable product. A schematic of the process is shown in Figure 2: Berbau-Forschung
FGD Process.

The dry absorber was truly a dry absorber in that no water was added to the flue gas
either before it entered or while it was inside the absorber. The absorber was a
device with a large entering and exiting face area but without much depth. On the
inlet and exit to the absorber were a series of horizontal angled slates much like a
partially-open venetian blind. The slates distributed the gas vertically and
horizontally across the face of the absorber so that it flowed evenly across the bed of
activated coke. The activated coke filled the space between the entering and exiting
slates, and from top to bottom of the absorber. The coke was added to the top of the
absorber and withdrawn from the bottom. Coke fines were purged from the system
and additional coke was added.

The activated coke from the absorber was added to the top of the regenerator, a
vertical, cylindrical vessel. Hot sand from a fired heater was added at various levels
of the regenerator. The hot sand drove the SOx and NOx out of the coke, and the
NOx was converted back to N2. The hot coke was drawn off the bottom of the
regenerator and returned to the absorber.

It was reported that the process could achieve the following removal efficiencies:
SOx: >98%
NOx: 10 to 50%
Particulate: 0.01 lb/MMB
Hg:    >90%

Only one facility was built in the United States, a 20MW demonstration unit at Plant
Scholz in Florida. The process has been modified and is now marketed under the
name ReACT by J-Power Group.

Other processes tested during this period were the Chiyoda Thoroughbred 101, the
magnesium oxide process, the cat-ox process and the citrate process. The authors of
this paper, however, have limited discussion to those processes of which they have
personal knowledge and experience.

IV. Equipment
A.    Absorbers

It wasn’t until the end of the early years, 1977 to 1978, that FGD systems began
using vertical open spray towers for the absorber. Prior to the spray tower, FGD
systems used venturi scrubbers, mobile bed scrubbers, packed bed scrubbers or
variations thereof as the absorber vessel. All had their own unique problems which
will be discussed later in this paper.

In the early years, absorber designs of the size needed for utility use did not exist,
and the absorber designs in use were different than those being used today. Most of
the time the absorber had to remove sulfur dioxide as well as particulate matter.
When state regulations required the retrofit of scrubbers onto existing plants for
sulfur dioxide, the state also required improved particulate removal. The most
PAGE 7 OF 34

prevalent particulate control technology was the cold side electrostatic precipitator.
The hot side electrostatic precipitator and the fabric filter were later developments.
Many of the electrostatic precipitators were not designed to give a clear stack, but
rather to reduce the particulate emissions so that more efficient induced draft fans
could be used.
Following are detailed descriptions of the different type of absorbers used during the
early years.
1.     Mobile Bed Scrubbers

Universal Oil Products (UOP) offered the mobile bed scrubber shown in Figure 3: UOP
Turbulent Contact Absorber (TCA). The mobile bed consisted of thousands of Ping-
Pong-ball-sized hollow hard plastic spheres that were enclosed in cages made from
stainless steel rods. Above the mobile bed was a single level of sprays that
distributed the scrubbing slurry over the bed. Above the spray header was a mist
eliminator. The base of the absorber consisted of a series of pyramidal hoppers, and
the discharge from each hopper was the suction to an absorber recycle pump. The
hot gas inlet was between the hopper and the bottom of the mobile bed.
To maintain the balls in suspension, gas velocity in the absorber was critical. If the
gas velocity was too high, the mobile packing pressed against the top of the cage,
causing the scrubbing slurry to accumulate on top of the packing until there was
enough head to break through. A gas velocity that was too low caused a similar
situation on the bottom of the cage.

Abrasion of the mobile packing was a significant problem. The original, blow-molded
hard plastic spheres had a seam which was a weak point, and was subject to wear
from the fly ash in the slurry and from rubbing against other spheres. Eventually, the
abrasion was sufficient for the spheres to separate at the seam. The half-spheres
dropped through the cage rods and into the slurry, where their sharp edges cut the
lining of the piping and pumps. UOP then developed a soft plastic sphere the same
size as the original spheres. As these balls heated up to the adiabatic saturation
temperature of the scrubber, the pores in the spheres would expand letting some of
the air contained in the sphere out. This was not a problem until the system was
shut down and the spheres cooled to the ambient temperature. No longer
maintaining a spherical shape because the air inside the sphere had escaped, they
would collapse and drop through the cage rods and get into the scrubbing slurry.
These soft plastic spheres would be found all over the plant site! Some were able to
work their way through the mist eliminator blades and then exit the chimney. Others
would enter the blowdown and be found floating on the thickener or in ponds.
Eventually, UOP developed a design that was satisfactory.

Today, it is hard to believe that this technology, in its day, was state-of-the-art.
Liquid-to-gas ratios were low, generally 40 or less. Residence times in the absorber
recycle tanks were short, generally 5 minutes or less. The design of the hoppers was
such that agitators were not required, which is an indication of how short the
residence time in the recycle tanks really was. The mist eliminators were in a flared-
out section above the spray header where the angle of the flare was greater than
what gases would normally expand without eddies forming. About 90% SO2 removal
was achieved with this design as well, as a 90% removal efficiency for particulate.

2.     Packed Bed Scrubbers

Combustion Engineering (CE), Research-Cottrell (R-C) and Mitsubishi Heavy
Industries (MHI) each offered packed bed scrubbers for sulfur dioxide removal, all of
PAGE 8 OF 34

which differed in design. The CE packed bed used a dumped packing while the R-C
packing was rigid. The MHI design had the flue gas flowing down through the rigid

The CE packed bed scrubber was one of the earlier scrubbers offered, and was
effective for both particulate and sulfur dioxide removal. The packing consisted of a
bed of glass marbles supported on a metal grid. While the bed provided good contact
for sulfur dioxide, the marbles were subject to fracturing because of the vibration in
the bed. As a result, grinders had to be located ahead of each absorber recycle pump
to reduce the glass fragments so they would not damage the rubber lining on the
pumps and piping.

The R-C absorber went through several revisions. R-C used a flooded disc scrubber
(FDS) for particulate removal, followed by a mist eliminator vessel. When increased
sulfur dioxide efficiencies were required, they installed a section of rigid packing
between the inlet and the mist eliminators in the top of the vessel. This required an
increase in the height of the mist eliminator vessel so that a spray header could be
installed above the packing. There were a couple of problems with this design. Since
the gases from the FDS entered the mist eliminator tangentially, they were swirling
when they entered the packed bed area, which caused maldistribution across the
bed. Also, the scrubber slurry for the packed bed was contaminated with the fly ash
from the particulate scrubber, which was thought to adversely affect the chemistry in
the absorber area. A schematic of the R-C FGD system, including the absorber, is
shown in Figure 4: Research-Cottrell FGD Process.

To overcome these two problems, R-C installed a “dentist bowl” below the packed
section. Above the dentist bowl was a donut-shaped plate that collected the
scrubbing liquor from the packed section and directed it to the dentist bowl located
below the hole in the donut plate. A pipe was connected to the bottom of the dentist
bowl that directed the packed section scrubbing liquor to an absorber recycle liquor
feed tank. Limestone added to the system for sulfur dioxide removal was added to
this tank. Between the dentist bowl and the donut plate were straightening vanes
that removed the swirl from the gas so that it could flow straight through the
packing. Above the packing was a single level of sprays with a mist eliminator
overhead. This design worked well and several are still in operation. On some of the
installations, in later years, the packing was removed and additional sprays were
installed, creating a vertical, open spray tower design.

The MHI scrubber was the only packed bed scrubber that had a down-flow design.
The flue gas entered the top of the absorber, was saturated, and then flowed down
across the rigid packing. The flue gas then turned 90o below the packing and flowed
across the absorber recycle reservoir before exiting through horizontal flow mist
eliminators. Two systems using this design were installed in the United States in
Indiana, one at the Merom Station in Merom, Indiana, and the other at the Bailey
Station in Gary, Indiana. The design has since been changed and is now marketed
under the name Advatech.

3.     Tray Scrubbers

The absorbers provided by Babcock & Wilcox (B&W), FMC and Wellman-Lord all
contained trays or tray-like contactors. B&W used and continues to use sieve trays.
FMC used a disc and donut contactor in their double-alkali scrubber. The Wellman-
Lord absorber utilized valve trays.
PAGE 9 OF 34

In the early years, B&W used a sieve tray in their absorber tower, and a rectangular-
throat venturi scrubber ahead of the absorber tower. The venturi served two
purposes: to adiabatically saturate the flue gas and to remove particulate matter. In
later systems, when fly ash removal was no longer necessary in the scrubber, they
were able to eliminate the venturi scrubber from their system. B&W’s current design
uses a sieve tray as the primary contactor for sulfur dioxide removal.
The FMC disc and donut design was ideally suited for the double-alkali system in that
it provided a good contact between the flue gas and the solution, yet had a low
pressure drop. The upper contactor was a tray that looked like a disc which fed a
donut beneath it. The scrubbing liquor flowed onto the donut and then down through
the hole in the contactor on the disc. The flue gas had to pass through the curtain
(waterfall) of scrubbing liquor flowing from the donut onto the disk.
4.      Venturi Scrubbers

Chemical Construction Co. (Chemico) offered a venturi scrubber for sulfur dioxide
removal, and R-C offered an orifice scrubber, which is similar to a venturi scrubber
but does not have a pressure recovery section downstream of the venturi throat.
The Chemico venturi scrubber was a variable-throat, vertical-downflow scrubber
which was initially used for particulate collection and later modified for sulfur dioxide
removal. Above the throat of the venturi was a plumb bob with a top-mounted
actuator that moved the plumb bob up or down to maintain the pressure drop across
the venturi necessary to achieve the required particulate removal. Downstream of
the pressure section of the venturi, the scrubbed flue gases made a 180o turn and
exited the vessel through a side outlet. There were chevron mist eliminators between
the place where the flue gases made the turn and the outlet. The base of the vessel
was the recycle reservoir. A cut-away drawing of the Chemico venturi scrubber is
shown in Figure 5: Chemico Single-Stage Venturi Scrubber.

Later, when SO2 removal became important, spray headers were installed in the
vessel upstream of the mist eliminators. Lime or limestone was added to the recycle
reservoir in the base of the vessel.

The R-C orifice scrubber, also called a flooded disc scrubber (FDS), was similar to a
venturi scrubber except that it did not have a pressure recovery section downstream
of the throat of the venturi. Instead, the vessel was a right circular cylinder with a
wetted elbow at the base. The flue gases made a 90o turn into a separate mist
eliminator vessel. The plumb bob in the orifice scrubber was below the throat of the
venturi, and its actuator entered the vessel through the wetted elbow.

5.     Vertical Open Spray Tower

The Utah Power & Light Huntington Station in Huntington, Utah was the first utility
plant to put a vertical, open spray tower into operation for the removal of SO2. The
first such tower had originally been scheduled to go on the Monticello Station in
Texas but, due to design delays, the tower on Unit 1 at the Huntington Station went
into operation first in May, 1978.

The 400MW system at Huntington consisted of four 33% absorbers with a common
inlet and outlet. Each tower contained four levels of sprays, and each level was
supplied by a dedicated pump. The absorber recycle reservoir was located in the
base of the tower. Vertical flow mist eliminators were located above the spray levels.
PAGE 10 OF 34

6.     Horizontal Scrubber

Dr. Alex Wier of Southern California Edison developed and patented the horizontal
scrubber. Initial tests of the 160MW scrubber test module at the Mohave Station in
Laughlin, Nevada, showed sulfur dioxide removal rates so high that the accuracy of
the sulfur dioxide monitors was questioned. After the tests at Mohave were
completed, the module was shipped to the Four Corners Station and tested. The
patent was later sold to Kellogg and several were installed. An elevation of the
horizontal scrubber is shown in Figure 6: Schematic of a Horizontal Scrubber. As
shown in the figure, flue gas entered from the left and discharged to the right. From
left to right, above the hopper, there were four spray zones and one horizontal-flow
mist eliminator. The water collected on the mist eliminator, and fresh scrubbing
liquor was added to the fifth hopper, under the mist eliminator. Scrubbing liquor
from the fifth hopper was pumped to the fourth-stage sprays, above the fourth
hopper. This fourth stage of sprays had the highest concentration of fresh alkali and
the lowest concentration of sulfur dioxide in the flue gas. The scrubbing liquor
collected in the fourth hopper was pumped to the third-stage sprays. This liquor
contained slightly lower concentrations of fresh alkali and the flue gas had slightly
higher concentrations of sulfur dioxide. This process continued step-wise to the first
stage, where the concentration of fresh alkali was the lowest and the concentration
of sulfur dioxide in the flue gas, the highest. A bleed stream from the first hopper
was sent to the dewatering area.

B.     Pumps

As stated earlier, NSPS regulations are technology-forcing, and this was especially
true for slurry pumps. Prior to the installation of FGD systems, pumps were not used
in any service that contained a slurry with an acidic pH and high TDS. There were
pumps in acid service, in slurry service, and in high TDS service, but there were no
pumps working in situations where all three service conditions existed

Denver Equipment Company (DECO) had a rubber-lined, open impeller pump that
was used in mining service. DECO also had pumps in the size range necessary for
FGD service, but the efficiencies were very low because of the large open passages
required for mining. The sulfur dioxide removal efficiencies were not as high as those
required today, so the L/G ratios were lower and therefore the pumps were not as
large. When increased pump sizes were required, the DECO pumps showed a
tendency to vibrate. The DECO pumps, like those of other pump manufacturers,
required seal water to flush the seals.

Galagher Pumps offered a pump that had expeller vanes on the back side of the
impeller, to help draw seal water through the seals and decreases slurry buildup.
This was a very good design on smaller pumps, but at the size necessary for the
spray pumps, the impeller backing plate was not strong enough to avoid vibration.
The vibrations were such that the expeller vanes rubbed against and tore the rubber

Despite their initial higher cost, Allen-Sherman-Hoff (ASH) pumps, which were still
rubber-lined and required seal water, became more popular because they had
eliminated the vibration and thus the vibration-related problems. The ASH pumps
were very successful, and were the primary pumps in use for many years, until the
PAGE 11 OF 34

Warman Pump began to successfully compete against the ASH pump during the
middle years (1977 to 1990) of FGD systems.

C.     Sludge Dewatering

There were four approaches to dealing with the FGD waste product. One was to pond
the material. A second approach was to dewater the material using primary and
secondary dewatering equipment, and landfill the ‘solid’ waste. The third approach
was similar to the second, but with a washing step added to create a saleable
product. A fourth approach was to perform some level of dewatering and then fixate
(stabilize) the product in a pozzolanic reaction. Forced oxidation and/or inhibited
oxidation of the waste material were not performed in the early years, and the
natural oxidation product was a thixotropic material (one that fluidizes when it is
subjected to shear) unless the solids content was high, in some cases greater than
80% suspended solids.

Figures 10, 11 and 12 show how difficult the material was to handle. Figure 10
shows the dewatered material after it was removed from a dewatering pond and
placed in a dump trailer. Note how the material is in well-defined clumps. Figure 11
shows the same material in the same truck after it has been driven a short distance
– the thixotropic nature of the material is very evident. Figure 12 shows how the
material had to be removed from the dump trailer. When the material fluidizes, it
flows into every nook and cranny in the truck bed. When the trailer is raised the
material will not flow out, and has to be assisted in its exit.

1.     Ponding

Ponding of the waste product, as an ultimate storage solution, required a large
amount of suitable land, such as a valley or canyon that could be dammed up,
located near to the utility plant. Some utility plants used a pond only for
intermediate storage, later dredging out the pond and taking the waste to a more
remote final disposal site. This required the use of three ponds, generally each sized
to accommodate 30 days of scrubber sludge. One pond was filling while a second
pond was drying (draining) and the third pond was being dredged out. This still
required a large amount of land, though not as much as when the pond was the
ultimate storage site. The three-pond option was also used when a canyon or valley
was not available, and the ponds had to be dug out of flat land.

2.     Primary Dewatering (Thickeners)

Nearly all of the FGD systems in the early years included primary and secondary
dewatering. Technology was borrowed from other industries, and thickeners were
borrowed from the water treatment industry for the primary dewatering of FGD
system wastes. As stated previously, neither forced oxidation nor inhibited oxidation
was used with FGD systems, which used natural oxidation instead. As a result, the
solids produced in the FGD systems were a co-precipitate of calcium sulfite and
calcium sulfate prone to the formation of rosette structures as pictured in Figure 7.
The material was very difficult to dewater, and if a thickener underflow of 30%
suspended solids was achieved it was considered successful. A photomicrograph of
calcium sulfite crystals is presented in Figure 8, and calcium sulfate crystals are
shown in Figure 9. During the early years of FGD systems, the use of hydrocyclones
for primary dewatering was not considered.
PAGE 12 OF 34

3.     Secondary dewatering

In early FGD systems, most secondary dewatering was done with rotary drum
vacuum filters. The drum filters were generally able to achieve between 50% to 60%
suspended solids in the filter cake. Horizontal belt filters were considered at some
installations, but because of their size and added cost they were not the best
economic choice.

One installation, at Units 1 and 2 at the Craig Station, in Craig, Colorado, used
centrifuges. Craig Station is located at a 6300 ft elevation where the standard
atmospheric pressure is 23.71 in Hg. Because of concern that there would not be
sufficient differential pressure across a vacuum drum filter to get the necessary
driving force to dewater the filter cake, they installed horizontal bowl centrifuges.
Unfortunately, the centrifuges never operated as desired. Recently, their FGD system
design was changed to a forced oxidation system, and the residence time in the
absorber recycle tanks was increased to give better solids formation. The centrifuges
were removed and replaced with rotary vacuum drum filters.

D.     Chimneys

Most of the FGD systems installed during the early years were retrofits, which meant
that system designs were often dictated by existing equipment. Most existing
chimneys were designed for natural drafting, but when the saturated flue gas from
an FGD system was introduced, there was not sufficient differential pressure to
achieve natural drafting. As a result, the chimneys became pressurized.
The chimneys during these early years either had steel or acid-resistant brick and
mortar liners. The use of FRP chimney liners was still several years from commercial
implementation. The steel liners were fabricated from carbon steel, not alloy steel,
and therefore were subject to severe corrosion by the gases containing acidic
droplets coming from the FGD systems. To counter the corrosion, two different
coatings for the steel liner were used. One such coating was gunite, which could be
sprayed onto the steel. The gunite worked very well for corrosion protection but it
had a different coefficient of thermal expansion than the steel liner so cracks
developed in the gunite, which allowed the corrosive material to migrate to the steel
liner and effect localized corrosion.

The other liner coating was a flakeglass polyester which had to be troweled onto the
steel. The flakeglass polyester provided a very good corrosion barrier but it, too, had
problems. It was known that the substrate had to be sandblasted to “white metal”,
but it was not known, at the time, that there had to be a specific profile on the steel
liner to maximize the bond between the flakeglass polyester and the steel.
Chimneys with acid-resistant brick and mortar also had problems. As the process
changed from hot flue gas (greater than 250 oF) to saturated flue gas
(approximately 130 oF) in the chimney, the chimney operation also changed from a
natural to a forced draft. This changed the inlet flue gas pressure from negative to
positive. As the liner aged, cracks developed in the brick and mortar which allowed
the pressurized saturated gases to migrate through the liner into the annular space.
The flue gas exiting the FGD system was saturated and also contained entrained
moisture which was at or near the pH of the scrubbing slurry – acidic. When the flue
gas entered the annular space it cooled, which caused more moisture to condense.
This condensed moisture collected on the steel bands surrounding the liner, on the
steel platforms in the annular space and on other equipment. This acidic moisture
PAGE 13 OF 34

caused corrosion of the steel surfaces. To counter this corrosion, the annular spaces
were pressurized so that the ambient air in the annular space was at a greater
pressure than the flue gas at the entrance to the chimney.

E.     Reheat

The effect of the saturated flue gas from the FGD systems on downstream equipment
and on the environment was a serious concern in the early years. Because the flue
gas was acidic, corrosion of the downstream ductwork and the chimney would occur.
The effect of the plume on the surrounding environment was also a concern. The
temperature of the flue gas was being reduced between 120 Fo to 200+ Fo in the
FGD system, and no one knew whether the plume would continue to rise or come
crashing back down to the ground. To help dry out the gas and to improve the plume
buoyancy, most FGD systems provided some type of system to reheat the flue gas.
Four different types of reheat systems were considered: in-line reheat, duct burners,
dilution air reheat systems and bypass reheat systems.

The in-line reheat system was one of the first reheat systems to be used, and the
first to be discarded because of operating problems. As the name indicates, the in-
line reheat system consisted of some type of device to directly heat the flue gas
downstream of the FGD system. Most used steam coils in the duct directly above the
mist eliminators. A typical design used bare tubes in the first few rows followed by
extended surface tubes. The assumption was that the bare tubes were hot enough to
sufficiently heat the entrained moisture to a vapor state before it contacted the
extended surface tubes. The reheater was intended to heat the flue gas so that its
buoyancy would be sufficient to avoid problematic ground-level sulfur dioxide
concentrations. In-line reheat systems were generally designed to provide between
10 Fo and 50 Fo of reheat to the flue gas.

The physics of reheating were not well understood during these early years. The
entrained moisture did evaporate, but the suspended and dissolved solids in the
droplets stayed behind on the tubes! The in-line reheat systems became a
maintenance nightmare. The dissolved and suspended solids would plate-out on the
steam tubes and, after a while, the scale would be thick enough to block gas flow
through the duct. The solids not only scaled the bare tubes, they scaled the extended
surface tubes as well.

One of the authors remembers seeing Haliburton use 5,000 PSI water to clean the
scale off of a set of reheat tubes. After the scale was removed they noticed there
was a two-foot-long split in one of the tubes. The scale had been so strong and thick
that none of the 300 PSI steam in the tube was escaping!

Duct burners were oil or gas-fired burners installed in the ductwork downstream of a
FGD system. The burners fired directly into the saturated flue gas. The authors are
not aware of any duct burner systems that were actually installed in the United
States, although many were discussed.

The dilution air reheat system was developed to overcome the operating and
maintenance problems of the in-line reheat systems. A dilution air reheat system
took ambient air, passed it over a set of steam coils, and then injected the heated air
into the saturated flue gas downstream of the absorber. The dilution air reheat
system has several advantages over the in-line reheat system: carbon steel could be
used in all system construction, there was no potential for scaling of the reheat
PAGE 14 OF 34

tubes, and it not only heated but also dried the saturated flue gas by diluting it with
drier air. The system also had many disadvantages. One was the significantly higher
steam usage required, because the ambient air to be heated was cooler than the
saturated flue gas. Another disadvantage was the amount of additional fans and
ductwork required.

Bypass reheat systems took hot flue gas from ahead of the FGD system, bypassed it
around the FGD system, and then injected the hot flue gas into the scrubbed and
saturated gas. The amount of reheat obtained was dependent on the amount of flue
gas that could be bypassed around the FGD system, which was, in turn, dependent
on the overall sulfur dioxide removal efficiency required to meet the regulations, and
the capability of the FGD system to remove sulfur dioxide. If a FGD system was
required to remove 90% of the sulfur dioxide and the absorbers were designed to
remove only 90% of the sulfur dioxide, then a bypass reheat system was not
possible. However, if an overall removal efficiency of, say, 72% was required and the
absorbers were capable of 90% removal, then a bypass reheat system was viable.
It was learned that care had to be taken in orienting the bypass duct relative to the
scrubbed gas duct. At an installation in the Midwest, the brick chimney liner started
to lean and eventually had to be replaced. The saturated flue gas entered the
chimney several feet above the hot gas bypass duct. Moisture from the scrubbed flue
gas ran down the brick liner between the scrubbed gas duct and the hot gas duct.
The hot gases, doing what they were supposed to do, evaporated the water and
warmed the flue gas. Unfortunately, the dissolved solids in the entrained moisture
reacted with the mortar and formed crystals. The crystals caused each mortar joint
to grow by a fraction of an inch. With the large number of mortar joints between the
hot gas and scrubbed gas inlets, the liner grew enough, on one side, to create a
structural failure in the lining.

F.     Mist eliminators

Most of the mist eliminators supplied with the original FGD systems were not
efficient. Either the blades did not have sharp enough angles or the spacing between
the blades was not correct. The result was a significant amount of mist carryover into
the outlet duct and the chimney. A couple of the authors of this paper have
personally seen sludge six or more feet deep in the outlet duct. Several of the
original FGD system suppliers provided their own “proprietary” mist eliminator
designs, none of which worked very well. Fortunately, original equipment
manufacturers, such as Munters, eventually came up with designs that were very

G.     Linings

The choice of construction materials for early FGD systems was very limited. There
was carbon steel, 316 and 316L stainless steel, flakeglass polyester and rubber
lining. Hastelloy C-276 had not yet been developed. 317 an 317LM were not
introduced until toward the end of the early years. Most of the early FGD systems
consisted of carbon steel absorbers with either a flakeglass polyester or rubber

Lining a vessel, such as the absorber, with rubber was a long and arduous process.
The vessel had to be erected completely except for the internals such as the spray
headers and trays. All of the internal surfaces requiring rubber coating had to be
sand-blasted to “white metal.” The rubber was then glued to the surfaces, a
PAGE 15 OF 34

dangerous process because the chemicals used were flammable and potentially
explosive in the right concentrations. Then the vessel was sealed and temporarily
insulated for the next step, which was the curing (vulcanization) of the rubber.
Curing was done by introducing low-pressure steam into the vessel to raise the
temperature of the rubber. The moist heat cured the rubber. Chemical curing was
possible, but was generally felt to be less effective than steam curing. After the
rubber had properly cured, the entire coated surface was spark-tested to ensure
there were no pin holes that could permit corrosion of the substrate steel.
A flakeglass polyester lining required a similar number of steps for installation. The
surface preparation process was the same as for the rubber lining. The flakeglass
polyester was troweled on in several thin layers. After it had been allowed to cure it
was spark-tested to determine if there were any pin-hole leaks.


Hopefully the reader has gained an appreciation of the successes and failures
experienced during the early years of FGD system development. The Standards of
Performance for New Sources are technology-forcing, and for the utility industry they
forced the development of a technology that had never been installed on facilities the
size of utility plants. That technology had to be developed, and a number of
installations completed in a short period of time. The US EPA continued to force
technology through the promulgation of successive regulations. The development of
this equipment was not an easy process. What may have appeared to be the simple
application of an equipment item from one industry to another often turned out to be
fraught with unforeseen challenges. Those challenges continue today.
Table1: FGD Systems Installed in the US from 1970-19783

Utility                      Plant             New/        Start-up    Absorber        Rem.   Supplier              Alkali       % S in   Coal
                                               Retrofit    Date        Type            Eff.                                      coal     type
                             Tombigbee                                                        Peabody Process
Alabama Elect. Coop          #2                New         Sep-78      Spray Tower            Systems               Limestone    1.15     Bitum
Arizona Electric                                                       Packed
Power Coop                   Apache #2         New         Aug-78      Tower           85%    Research-Cottrell     Limestone    0.50     Bitum
Arizona Public                                                         Packed
Service                      Cholla #1         Retrofit    Oct-73      Tower           92%    Research-Cottrell     Limestone    0.50     Bitum
Arizona Public                                                         Packed
Service                      Cholla #2         New         Apr-78      Tower           75%    Research-Cottrell     Limestone    0.50     Bitum
                             Duck Creek                                                       Riley Stoker
Central Illinois Light       #1                New         Jul-76      Rod Bed         85%    Environeering                      3.66     Bitum
Columbus &
Southern Ohio                                                                                 Air Correction Div,   Thiosorbic
Electric Company             Conesville #5     New         Jan-77      TCA Module      90%    UOP                   Lime         4.67     Bitum
Columbus &
Southern Ohio                                                                                 Air Correction Div,   Thiosorbic
Electric Company             Conesville #6     New         Jun-78      TCA Module      90%    UOP                   Lime         5.67     Bitum
                                                                       Venturi                Chemico Div,
Duquesne Light               Elrama #1-4       Retrofit    Oct-75      Scrubber        83%    Envirotech            Lime         2.20     Bitum
                                                                       Venturi                Chemico Div,
Duquesne Light               Phillips #1-6     Retrofit    Jul-73      Scrubber        83%    Envirotech            Lime         1.92     Bitum
Indianapolis Power &         Petersburg                                                       Air Correction Div,
Light                        #3                New         Dec-77      TCA Module      85%    UOP                   Lime         3.25     Bitum
Kansas City Power &                                                                           Combustion
Light                        Hawthorn #3       Retrofit    Nov-72      Marble Bed      70%    Engineering           Lime         0.60     Bitum
Kansas City Power &                                                                           Combustion
Light                        Hawthorn #4       Retrofit    Aug-72      Marble Bed      70%    Engineering           Lime         0.60     Bitum
Kansas City Power &                                                    Venturi/Sieve
Light                        La Cygne #1       New         Feb-73      Tray            80%    Babcock & Wilcox      Limestone    5.39     Subbit
Kansas City Power &                                                                           Combustion
Light                        Jeffery #1        New         Aug-78      Spray Tower     50%    Engineering           Limestone    0.32     Subbit

         EPA Utility FGD Survey: October – December 1980, EPA-600/7-81-012a, 1981
      PAGE 17 OF 34

Kansas City Power & Light    Lawrence #4        Retrofit   Jan-77     Rod Bed            73%       Combustion Engineering     Limestone         0.55          Bitum
Kansas City Power & Light    Lawrence #5        Retrofit   Apr-78     Rod Bed            73%       Combustion Engineering     Limestone         0.55          Bitum
                             Green River #1-                          Venturi/Mobile
Kentucky Utilities           3                  Retrofit   Sep-75     Bed                80%       American Air Filter        Lime              4.00          Bitum
Louisville Gas & Electric    Cane Run #4        Retrofit   Aug-76     Mobile Bed         85%       American Air Filter        Carbide Lime      3.75          Bitum
Louisville Gas & Electric    Cane Run #5        Retrofit   Dec-77     Spray Tower        85%       Combustion Engineering     Carbide Lime      3.75          Bitum
Louisville Gas & Electric    Mill Creek #3      New        Aug-78     Mobile Bed         85%       American Air Filter        Lime              3.75          Bitum
Louisville Gas & Electric    Paddy's Run #6     Retrofit   Apr-73     Marble Bed                   Combustion Engineering     Carbide Lime      2.50          Bitum
Minnkota Power               Milton R. Young                                                                                  Lime/Alkaline
Cooperative                  #2                 New        Sep-77     Spray Tower        75%       Combustion Equip. Assoc.   FA                0.70          Lignite

Utility                      Plant              New/       Start-up   Absorber Type    Rem. Eff.   Supplier                   Alkali                   %S     Coal type
                                                Retrofit   Date                        (%)                                                             in

Montana Power                Colstrip #1        New        Sep-75     Tower                        Combustion Equip. Assoc.   Lime/Alkaline FA         0.77   Subbit
Montana Power                Colstrip #2        New        May-76     Tower                        Combustion Equip. Assoc.   Lime/Alkaline FA         0.77   Subbit
Nevada Power                 Reid Gardner #1    Retrofit   Mar-74     Tray             90%         Combustion Equip. Assoc.   Sodium Carbonate         0.50   Bitum
Nevada Power                 Reid Gardner #2    Retrofit   Apr-74     Tray             90%         Combustion Equip. Assoc.   Sodium Carbonate         0.50   Bitum
Nevada Power                 Reid Gardner #3    New        Jul-76     Tray             90%         Combustion Equip. Assoc.   Sodium Carbonate         0.50   Bitum
Northern Indiana Public      Dean H. Mitchell                         Venturi/Tray
Service                      #11                Retrofit   Jul-76     Tower            90%         Davy McKee                 Wellman Lord             3.50   Bitum
                                                                      Venturi/Marble                                          Limestone/Alkaline
Northern States Power        Sherburne #1       New        Mar-76     Bed              50%         Combustion Engineering     FA                       0.80   Subbit
                                                                      Venturi/Marble                                          Limestone/Alkaline
Northern States Power        Sherburne #2       New        Mar-77     Bed              50%         Combustion Engineering     FA                       0.80   Subbit
                             Bruce Mansfield
Pennsylvania Power           #1                 New        Dec-75     Venturi          92%         Chemico Div, Envirotech    Thiosorbic Lime          3.00   Bitum
                             Bruce Mansfield
Pennsylvania Power           #2                 New        Jul-77     Venturi          92%         Chemico Div, Envirotech    Thiosorbic Lime          3.00   Bitum
Public Service of New                                                 Venturi/Tray
Mexico                       San Juan #1        Retrofit   Apr-78     Tower            85%         Davy McKee                 Wellman-Lord             0.80   Subbit
Public Service of New                                                 Venturi/Tray
Mexico                       San Juan #2        Retrofit   Aug-78     Tower            85%         Davy McKee                 Wellman-Lord             1.80   Subbit
South Carolina Public                                                 Venturi/Sieve
Service                      Winyah #3          New        Jul-77     Tray             45%         Babcock & Wilcox           Limestone                1.7    Bitum
South Mississippi Electric   R.D. Morrow, SR.                                                      Riley Stoker
Power`                       #1                 New        Aug-78     Venturi/Rod                  Environeering              Limestone                1.3    Bitum
Springfield City Utilities   Southwest #1       New        Jul-77     TCA Module       80%         Air Correction Div, UOP    Limestone                3.5    Bitum
     PAGE 18 OF 34

Tennessee Valley
Authority              Shawnee #10A       Retrofit   Apr-72   TCA Module      Prototype   Air Correction Div, UOP   Lime/Limestone   2.9    Bitum
Tennessee Valley                                              Venturi/Spray
Authority              Shawnee #10B       Retrofit   Apr-72   Tower           Prototype   Chemico Div, Envirotech   Lime/Limestone   2.9    Bitum
Tennessee Valley                                              Venturi/Spray
Authority              Windows Creek      Retrofit   May-77   Tower           80%         TVA                       Limestone        3.7    Bitum
Texas Utilities        Martin Lake #1     New        Apr-77   Packed Tower    75%         Research-Cottrell         Limestone        0.9    Lignite
Texas Utilities        Martin Lake #2     New        May-78   Packed Tower    75%         Research-Cottrell         Limestone        0.9    Lignite
Texas Utilities        Monticello #3      New        May-78   Spray Tower     74%         Chemico Div, Envirotech   Limestone        1.5    Lignite
Utah Power & Light     Huntington #1      New        May-78   Spray Tower     80%         Chemico Div, Envirotech   Limestone        0.55   Bitum
PAGE 19 OF 34

Table 2: Operating Conditions Offered by SO2 Removal System Suppliers4

Supplier                  Scrubber             Alkali            L/G          %S        Stoich Ratio   % Solids   Hold Tk Res.   SO2 Removal
                                                                                                                  Time, Min      Guarantee
Chemical Const            Venturi 2 stage      Lime              40-80        0.5 – 4   0.8-1.3        12         3              90
                          spray column         Limestone         80                     1.3-1.5        10-12      3              75
Combustion                1 or 2 marble beds   Lime              25-30        0.5-4     1.0            5-8        4-6            90
Engineering                                    Limestone         25-30                  1.2-1.3                                  80
                                               Carbide sludge    25-30                  1.1                                      90
B&W                       Low pressure         Lime              40-50        0.5-4     1.1            5-8        4-6            90
                          quencher plus tray   Limestone                                1.2-1.3                                  80
                          absorber             Carbide sludge                           1.1                                      90
Peabody Process           Venturi plus spray   Lime              50-60        0.5-4     1.1            15         8-10           90
Systems                   column               Limestone         80                     1.2-1.3                                  85
UOP                       TCA (3 stages)       Lime              40           0.5-4     1.25           10         5              90
                                               Limestone                                1.5                                      85
Combustion Equipment      Venturi plus spray   Lime              60           0.5-4     1.1-1.3        8-12       10-12          80-85
Assoc                     column               Limestone         80                     1.2-1.4
Research-Cottrell         Multi-contact        Limestone         1st stage    0.5-2     1.1-1.3        10-15      5-10           90
                          absorber             Lime              – 50-60      3-4
                                                                 2nd stage
                                                                 – 15-30

    Proceedings: Flue Gas Desulfurization Symposium – 1973, EPA-650/2-73-038, p 117
PAGE 21 OF 34

Figure 1: Wellman-Lord FGD System
PAGE 22 OF 34
PAGE 23 OF 34

Figure 2: Berbau-Forschung FGD Process
PAGE 24 OF 34
PAGE 25 OF 34

Figure 3: UOP Turbulent Contact Absorber (TCA)
PAGE 26 OF 34
PAGE 27 OF 34

Figure 4: Research-Cottrell FGD Process
PAGE 28 OF 34
PAGE 29 OF 34

Figure 5: Chemico Single-Stage Venturi Scrubber
PAGE 30 OF 34
PAGE 31 OF 34

Figure 6: Schematic of a Horizontal Scrubber
PAGE 32 OF 34

Figure 7: Calcium sulfate/calcium sulfite rosette

Figure 8: Calcium sulfite crystals
PAGE 33 OF 34

Figure 9: Calcium sulfate crystal from the Chiyoda process
PAGE 34 OF 34

Figure 10: Dewatered sludge loaded into a haul truck

Figure 11: Dewatered sludge in truck after traveling a short distance

Figure 12: Method for removing the thixotropic sludge from haul truck