United States
 Environmental Protection
 Agency
 Industrial Environmental Research
 Laboratory
 Research Triangle Park NC 27711
 Research and Development
EPA-600/S7-83-062 Feb. 1984
 Project Summary

 The  Current  Status  of
 Commercial  Utility  Flue  Gas
 Desulfurization  Systems
G. P. Behrens and J. C. Dickerman
  This  summarizes a  report  on the
status of commercial flue gas desulfur-
ization (FGD) processes applied to coal-
fired utility  boilers in the U.S. Major
objectives of this work were to examine
the impacts of the 1979 New Source
Performance Standards on FGD system
design and  operation and to  identify
recent  improvements in the technol-
ogy. In the 4 years since the promulga-
tion of the  NSPS, the wet limestone
process has been selected by utilities
for 75  percent  of the  new  plant
capacity. In  this time period, 77 plants
representing  over 37,000  MW  of
capacity have selected FGD systems.
Several major trends in the design of
limestone systems have become fairly
standardized. Nearly all new systems
are being built with spare absorber
modules  to qualify for the  NSPS
emergency  bypass  provisions. The
predominant absorber  design  is the
open spray tower, due to minimal main-
tenance requirements.  Forced
oxidation  to produce gypsum  solids,
which can then be landfilled, is being
incorporated in  many  new  units.
Selection  of the spray  drying process
for 15 percent of the new sites has also
occurred  in the  last  4 years. The
remaining throwaway and regenerable
systems have not experienced any sig-
nificant  increases in  applications.
Finally, organic acid addition has been
successfully  demonstrated  on lime-
stone systems to improve SO2 removal
and  system  reliability.  Organic acid
addition is being used at two sites to
upgrade the performance of older
limestone systems.

  This Project Summary was developed
by  EPA's Industrial  Environmental
Research Laboratory,  Research  Tri-
angle Park.  NC, to announce  key
findings of the research project that is
fully documented in a separate report of
the same title (see Project Report order-
ing information at back).

Introduction
  This report examines the current status
of flue gas desulfurization (FGD) systems
applied  to  coal-fired  utility boilers to
control sulfur dioxide (SO2) emissions. It
documents changes in the design of FGD
systems which have occurred over the
years, specifically  in  response to  the
requirements  of the 1979 New Source
Performance Standards (NSPS) for utility
boilers. This information is useful to EPA
Regional Offices  in  evaluating  SO2
control  technology for new  utility
applications, to the Office of Air Quality
Planning and Standards (OAQPS) in their
review of the 1979 NSPS,  and to the
Office of Research and Development in
establishing  research  initiatives.  This
information is also useful to architects,
engineers, vendors, utility  companies,
and others interested in FGD technology.

-------
Objectives of this study were to examine
differences in  the designs and applica-
tions of FGD systems applied to plants
subject to the 1971 and 1979 NSPS. The
basic  difference  between  the  two
standards is inclusion of  a  minimum
percent reduction  requirement of  SO2
emissions in the new standard. The 1971
standard limited SO2 emissions to 1.2 Ib*
SO2/million Btu of heat input. The 1979
NSPS requires a minimum of 70 percent
removal for all controlled emissions less
than 0.6 Ib  SO2/million Btu. A sliding-
percentage  removal scale  is  used as
sulfur levels  increase. The  maximum
emission limit  under the new standard is
also  1.2 Ib  S02/million  Btu.  Figure 1
shows the  removal requirement of the
NSPS for different sulfur  levels.  The
major effect of the new standard  has
been to   require   sulfur   oxide
(SOx)emission  controls  for most  low
sulfur coals which,  under the 1971
standard, would not have required any
SO2 removal.  The  rationale behind the
decision to require SOx emission controls
for most coals is discussed in detail in the
Federal  Register.  Two major reasons
cited were to promote the use of the  best
available FGD technology according to
provisions of the Clean Air Act,  and to
maintain a competitive economic balance
between high- and low-sulfur coal  pro-
ducing regions in the U.S.
FGD Technology Status
  The  most  dominant  trend  in  FGD
technology is the overwhelm! ng selection
of the  limestone FGD process  for new
units. Table 1 shows the status of the
various   FGD technologies  as  of
September 1978, priorto promulgation of
the 1979 NSPS, and as of January 1983.
The equivalent scrubbed gas capacity (in
megawatts)  and number of units are
presented for  each   process.  The
percentage distribution (by megawatts) of
the various processes is also shown. The
totals presented are for units operating,
in construction , contracted, and planned
which have selected an  FGD technology.
  The processes in Table 1 are listed in
order of application over the last 3 years.
As can be seen, limestone systems have
been  specified for  75  percent of the
capacity  of  these  units. Spray drying
systems account for most of the remain-
ing  units. The other  throwaway and
•Although EPA policy is to use metric units, certain
 nonmetric units are used here for convenience.
 Readers mor familiar with the metric system may
 use the conversion factors at the back.
regenerable  technologies  show only
minor increases in application. The two
processes showing decreases represent
the shutdown of a magnesium oxide test
unit and a reduction in size of several
units which had planned (in 1978) to use
alkaline ash scrubbing. Of particular note
is the reduced use of the lime FGD
process: in September 1978, 25 percent
of the FGD capacity used  the wet lime
process;  however, in the  next 3 years,
lime processes were specified at only 5
percent of the new units. Sixteen times as
many limestone systems were selected
as lime units in this interval. In 1978, the
ratio was two limestone systems to each
lime unit.
  The decline in the use of lime systems is
due to: (1) the higher reagent cost of lime
       7.4-
       1.2-
 a
 $

-------
Table 1 . Comparison of FGD System Status in September 1978 and January 1983
September 1978 January 1983 Increase from 1978 to 1982
Process MW
Limestone 23,849
Spray Drying 400
Lime 12,065
Sodium Carbonate 884
Dual Alkali 1. 102
Wellman-Lord 1,855
Aqueous Carbonate 100
Citrate 60
Magnesium Oxide 846
Alkaline Ash-
Lime/Limestone 7,316
Total 48,477
for small size units or low sulfur fuels: in
both applications, the cost of lime is a
fairly small portion of the total revenue
requirement. However, the spray drying
process has recently attained
prominence in low sulfur applications,
due to anticipated favorable economics
and performance characteristics.
Therefore, the relative ranking of the wet
lime process will probably continue to
decline in future applications. Lime is the
predominant reagent used in the spray
drying systems, which may result in an
increase in industry's demand for this
reagent.
The regenerable systems (Wellman-
Lord, magnesium oxide (MgO), Citrate,
and Aqueous Carbonate) have not shown
any substantial increase in applications.
Part of the reason includes the develop-
mental nature of two of these processes,
as well as the higher costs associated
with regenerable systems in general.
Table 2 shows the distribution of pro-
cesses according to the emission require-
ments of each unit. Again, the totals are
for systems operating, in construction,
contracted, and planned.
Each process in Table 2, except MgO,
Aqueous Carbonate, and Citrate, will be
or have been used in commercial
applications to meet the 1 979 NSPS or a
more stringent state standard. However,
the selection of a process technology is
very site specific; not all processes are
applicable to every situation. The
following discussion outlines the range of
applications of each process and
Percent Percent Percent
Number MW MW Number MW MW Number MW
54 49.2 51,662 103 60.3 27.813 49 74.8
1 0.8 6.353 18 7.4 5.953 17 16.0
29 24.9 13.736 32 16.0 1.671 3 4.5
4 1.8 3,155 9 3.7 2,271 5 6.1
3 2.3 2.288 6 2.7 1.186 3 3.2
6 3.8 2.074 8 2.4 219 2 0.6
1 0.2 100 1 0.1
1 0.1 60 1 0.1
4 1.8 724 3 0.9 (122) (1) 1-0.3)
1± 15.1 5.493 13 6.4 (1.823) (1) (-4.9)
117 100.0 85.645 194 100.0 37,168 77 100.0
Table 2. FGD Systems and Regulatory Classifications - January 1983
Applicable Emission Standard. MW. (Number of Units)
More Stringent
Than 1971 NSPS.
More Stringent But Less Than Less Than
Process 1979 NSPS 1 971 NSPS Than 1979 NSPS 1979 NSPS 1971 NSPS
Limestone 27,461(45) 9.309(25) 7.239(14) 3.151(10) 4,502(9)
Lime 715(2) 4.551(10) 2.140(4) 5.486(12) 844(4)
Spray Drying 2.077(8) 1.180(3) 2.099(4) 887(2) 110(1)
Alkaline Ash-
Lime/ 'Limestone 475 (1) 1,374 (4) 1.979 (5) 1,665 (3)
Sodium Carbonate 1,900 (4) 375 (3) 330 (1) 550 (1)
Wellman-Lord — - -— 7,779 (4) 295 (4)
Dual Alkali 1.107 (3) 881 (2) -— 300 (1)
Magnesium Oxide — — — 724 (3) —
Aqueous Carbonate — — — - — WO (1)
Citrate -— 60 (1)
Total 33.735 (63) 17.73O (48) 15,566 (32) 13.058 (36) 5.556 (15)
comments generally on performance, requirements, environmental considera-
energy requirements, environmental tions, and reliability. Table 3 summarizes
effects, and reliability. these comments.
FGD Applications and Applications and Recent
Characteristics Process Improvements
Although most commercial FGD The limestone process is the
processes could be used at many types of predominant FGD system in use because
sites, generally the plants using the same of several factors. First, limestone
process have some similar characteris- deposits are, in many parts of the country,
tics which favor the use of the selected leading to low transportation and reagent
process. These trends are discussed, and costs. Second, limestone systems have
general comments are made on energy been in use longer than other types of

-------
Table 3.    Qualitative Evaluation of Commercial FGD Processes
Process
     Performance
                                                  Energy Requirements     Environmental Considerations     Reliability
Limestone
New systems are being de-
signed to meet the 1979
NSPS with high and low
sulfur coals. Additives
can be used to enhance
SO2 removal.
Moderate to high. Depends
on absorber pressure
drop, amount of grinding
required to prepare
limestone.
                                                                         Produces large volumes
                                                                         of sludge.
                         Historically the lowest
                         of commercial processes,
                         additives may improve
                         performance.
Lime
Traditionally lime sys-
tems are somewhat more
effective than limestone
units at equivalent operat-
ing parameters.
                                                 Moderate, since reagent
                                                 preparation does not
                                                 require grinding.
                        Produces large volumes
                        of sludge.
                        Better than limestone;
                        however, still can experi-
                        ence scaling and plugging.
Spray Drying
Currently only able to
meet NSPS for lower
sulfur fuels.
Very low because of small
pumping requirements
and no reheat require-
ments.
Dry waste product often
requires wetting for
disposal.
Insufficient data
base; however,
process simplicity
is an advantage.
Alkaline Ash
Limited to certain fuel
types, usually only low
sulfur coals.
Comparable to lime
systems.
                                                                         Produces large volume
                                                                         of fly ash/FGD sludge.
                         Comparable to lime
                         systems.
Sodium Carbonate
High SO2 removal
achievable.
                                                 Low due to high alka-
                                                 linity of reagent.
                        Large volume of soluble
                        waste produced. Requires
                        special disposal.
                         Very reliable due to
                         simplicity.
Dual Alkali
High SO2 removal
achievable on high
sulfur fuels.
Low due to high alkalinity
of reagent.
                                                                         Large volume of sludge
                                                                         produced.
                         Based on operating units,
                         most reliable commer-
                         cial system.
Wellman-Lord
Magnesium Oxide
High SO2 removal
achievable on high
sulfur coal.

High SO2 removal
achievable.
Fairly high because of
steam requirements and
gas for regeneration.

Moderate, coal can be
used in drying and
regeneration.	
Produces sulfur, sulfuric
acid, and sodium sulfate,
saleable by-products.

Produces sulfuric acid.
Fairly good, but may
suffer somewhat due to
process complexity.

No commercial systems.
Prototype plants had
many mechanical problems.
 processes;  consequently, vendors  and
 utilities are familiar with them. Perhaps
 most  importantly,  until  the advent  of
 spray  drying,  limestone  units were
 typically  the  lowest  cost  process
 available.  Commercial experience  has
 also shown that limestone systems can
 be used successfully on a wide variety of
 fuels.  Several  innovations  in process
 design have  become  commonplace  in
 limestone process designs. The absorber
 vessel is almost always  an open spray
 tower, which  requires less maintenance
 than packed towers. Forced oxidation of
 the  sludge  produces  better handling
 solids and reduces the scaling potential of
 the slurry. Additives such as magnesium
 and  organic  acids  are  being  used
 commercially to raise  the SO2 removal
 closer to levels typically reached by more
 reactive sodium and lime systems. This
 flexibility  has  resulted   in  the  large
 number of limestone  systems  in use.
                  However, because of site specific factors
                  and  utility preferences, other processes
                  have often been selected.
                    The  most  significant determinant of
                  process   selection  is  probably  fuel
                  composition,  specifically  sulfur  level,
                  although chloride and ash levels are also
                  important. The sodium based processes
                  (dual alkali,  Wellman-Lord)  are  good
                  candidates for high sulfur fuels because
                  of the  high sorption  capability of the
                  scrubbing  liquor.  A dual alkali system
                  vendor is now marketing a process using
                  limestone  (rather  than  the  more
                  expensive  lime)  as the regenerant to
                  improve process economics.
                    For  low  sulfur applications,  spray
                  drying  is  currently the  most  popular
                  process, due to  the  reduced  reagent
                  volumes  needed  and  reduced  capital
                  costs. Depending on the ash composition,
                  the alkaline ash process can also be used
                  in low  sulfur applications, although no
                                    new units are being planned for this pro-
                                    cess. High chloride coals can cause both
                                    process  chemistry  problems  and
                                    corrosion. A prescrubber before the SO2
                                    absorber is  often used  to  remove  the
                                    chloride, although it can cause serious
                                    corrosion.
                                      Land availability favors the regenerable
                                    over the throwaway processes such as
                                    the Wellman-Lord  and  MgO systems.
                                    These  processes produce  saleable  by-
                                    products rather than throwaway sludges.
                                    The local availability of  a  reagent also
                                    affects  process  selection.  All  sodium
                                    carbonate  systems  are  near  natural
                                    deposits of the material. If transportation
                                    costs are a major portion of the delivered
                                    reagent   price,   lime  can   have   an
                                    advantage over limestone because of its
                                    greater reactivity per pound.
                                      Water availability  is often cited as a
                                    factor in FGD process selection; however,
                                    all  processes  use  nearly  the  same
                                     4

-------
amount of water. Spray drying may use
slightly less water because the flue gas is
not totally saturated; however, the dry
waste   product  is  moistened  before
disposal. The water requirements of the
FGD system are typically less than  10
percent of the plant's total requirement. A
more important consideration is the type
of  water used  in  the  FGD  system.
Throwaway systems can generally tole-
rate  poorer   quality   water   than
regenerable systems.

Energy Requirements
  Energy requirements for FGD systems
vary significantly,  depending  on  the
basis for analysis. For example, electrical
and steam requirements for commercial
processes have been generally shown to
increase as follows (lowest to highest):
spray drying,  sodium carbonate, dual
alkali, alkaline ash, lime, limestone, MgO,
and Wellman-Lord.  The energy
requirement as a  percentage  of new
power plant output is often reported to be
less than 1 percent for sodium svstems,
normally about 3 percent for limestone
systems (excluding reheat), and even 10
percent for some regenerable units. The
problem with comparing the  energy
consumption required by the FGD system
alone is that it does not reflect the energy
required to process the reagents or the
sludge, nor does it include any by-product
energy  credit.  These  energy
requirements are reflected in the reagent
cost and by-product credits.
  A TVA study examined the ground-to-
ground  energy requirements of three
FGD systems. The design basis was a 500
MW unit burning 3.5 percent sulfur coal
meeting the  1971 NSPS. Results for the
three systems are shown in Table 4. The
lime system  had  the lowest  energy
requirement if only the  FGD system is
considered.  However,  the  energy
required to calcine limestone into lime
(which is reflected in the  price of lime) is
significant. When  all  of these energy
requirements  are considered,  the
limestone  process has the lowest energy
requirement. The credit for sulfuric acid
in the MgO process reflects the energy
required   to  mine and  convert  an
equivalent amount of elemental sulfur.
This  credit  makes  the  regenerable
process almost as efficient as  the lime
system. Since  spray drying and dual
alkali systems  use lime,  their overall
ground-to-ground  energy requirements
are much closer to the other systems than
would be indicated by just comparing the
FGD system energy requirement.

Environmental  Considerations
  The  six  commercial  throwaway FGD
systems produce large volumes  of waste
material requiring disposal. The calcium
based  sludge from the limestone, lime,
alkaline ash,  and dual alkali processes
are very similar in composition. Typically,
this sludge is disposed of in a clay-lined
pond. A thickened slurry is pumped to the
pond  where  the  solids settle out.
Supernatant water is usually returned to
the  FGD  system. Many of the newer
plants are using landfilling for disposal. In
this practice,  the  sludge  is  vacuum
filtered or  centrifuged to  a higher solids
content, and mixed with  lime and/or fly
ash  to form  a stable mass. In  either
installation, a liner material  is used to
prevent ground water contamination due
to leachate from the sludge.
  Solid waste from spray drying, if lime is
used as the reagent, is also similar to the
material described above, except that it is
a dry powder. However, to dispose of this
Table 4.    FGD Ground-to-Ground Energy Requirements
Btu/lb Sulfur Removed
Function
Mining
Absorbent Processing
Transportation
FGD
Waste Disposal
By-product Credit
Total
Btu/kWh
% Gross Power Unit Output
Limestone
438

176
J 4.042
22

14.678
291
3.2
Lime
356
6.198
143
13.165
15

19,877
395
4.4
Magnesium Oxide
25
161
33
26.387
	
(5,491)
21.115
420
4.7
waste, it is often wetted to increase its
compressive  strength  and to  reduce
dusting  problems.  Several  research
programs are being sponsored by the EPA
and EPRI to determine the best methods
of disposing  of spray drying wastes.
  The sodium  carbonate  process
produces a liquid  purge stream  high in
sodium sulfites and sulfates. The plants
which use the process, use evaporation
ponds to crystallize  the  salts. This
technique is  only  applicable  in arid
regions.  Pond liners may  also  be
necessary  to prevent ground  water
contamination.
  The regenerable  processes  have  a
much smaller secondary environmental
impact since they produce by-products
which are  sold  to  user industries.
However, lime and limestone processes
may be designed to produce gypsum wall
board or other by-products if desired.
Reliability
  The reliability of an FGD system is a
major concern to utilities operating plants
subject to the 1979 NSPS. Older plants
and regulations often allowed bypassing
flue gas around malfunctioning absorber
modules or the entire system. However,
the 1979 NSPS has significantly reduced
the possibility of bypassing the  FGD
system except in emergencies. The final
standards allow an owner or operator to
bypass  uncontrolled flue  gas around a
malfunctioning  FGD  system  if: (1) the
FGD system has a built-in spare FGD
module, (2) enough FGD modules are not
available to treat the entire quantity of
flue gas generated, and (3) all available
electric  generating  capacity  is being
utilized  in  a  power  pool or  network
consisting of the generating capacity of
the affected utility company (except  for
the  capacity  of  the largest single
generating  unit in the company), and the
amount of power that could be purchased
from neighboring interconnected utility
companies.  From  the  constraints
expressed in the third requirement it is
evident   that  utility  companies  are
expected to install reliable FGD equip-
ment. This requirement also places  an
economic incentive on the utilities to
obtain reliable operation from their FGD
systems, since the available electric gen-
erating   capacity  will  often   include
uneconomic gas-fired turbines and oil- or
gas-fired boilers.
  The few  operating  units regulated by
the 1979 NSPS have not experienced
enough  operating time to determine if
they are more reliable than older units.

-------
Several  studies  have  addressed  the
reliability of some of the older units. A
study by the National  Electric Reliability
Council (NERC) has categorized causes of
outages  for 400 MW and larger FGD
systems. Results of the study are given in
Table 5. As can be seen, two-thirds of the
outage time was caused by non-specific
FGD technology problems. This indicates
that FGD processes often suffer a poor
reliability image because of  peripheral
equipment failures rather than process
related problems.

Table 5.   Percent of Plant Outage  Time
          Caused by Specified FGD Compo -
          nent Failure (NERC Data)
            Table 6.    Dependabilities of FGD Systems (March 1981-March 1982)
     Component
Percent of
 Outage
  Time-
Dampers                         28
Duct system, baffles               19
Absorber tower                   16
Fans, ID. FD, blades, vibration        17
Mist eliminator                    9
Reaction tank                      4
Spray pumps                      3
External causes                    2
Thickener                         1
Test programs	1

  An FGD data base, maintained for the
EPA by PEDCo, contains information on
availability and operability  of individual
plants.  The  average  availability and
operability  of  the  commercial  FGD
processes are presented in Table 6, along
with definitions of the  two terms. The
availability index can be artificially high if
the boiler is not in operation although the
FGD   system  could  operate. The
operability index is a  more  accurate
estimates of the reliability of the various
processes  since  it  involves the  actual
time that the boiler is operated. It can be
lowered, however, if a decision is made
not to operate the FGD system.
  Table 6  indicates that limestone and
lime systems are  generally the most un-
reliable processes. However, it is difficult
to  compare  reliability data of various
processes because of differences in plant
designs and operating requirements.  No
data are available from plants regulated
by the 1979 NSPS. Very few operating
systems  in the EPA/PEDCo data base
have  spare  modules  which  increase
reliability. Additionally, the design differ-
ences  make  it  difficult  to  compare
reliability figures.  New  units  are
incorporating  design  improvements
obtained from experience with existing
Process
Limestone
Lime
Wellman-Lord
Alkaline Ash
Sodium Carbonate
Dual Alkali
Number of
Operating
Systems
39
24
7
9
5
3
Systems
Included
in Averages
24
23
7
6
4
3
Availability*
73.5
84.1
88.5
89.3
89.9
96.2
Operability**
73.8
75.4
70.0
78.0
87.4
79.7
             Percentage of hours the FGD system is available for operation (whether used or not) divided by
             the hours in the period.
             * Percentage of hours the FGD system was operated divided by the boiler operating hours during
             the period.
systems. Many of the older limestone and
lime systems with  poor designs  are
included in the averages. Their recurring
problems reduce the reliability figures for
the process. Consequently, the construc-
tion of new units should  produce a
gradual  increase  in  overall  process
reliability. The effects of a new technol-
ogy, such as the use of adipic acid with
limestone systems, can also significantly
improve  the  reliability  of  a  system.
However,  it is difficult to quantify in
advance the  magnitude  of reliability
increases due to process improvements.
  A recent study examined the effect of
forced  outages  of scrubber modules on
utility  system  production  costs.  Two
different utility fuel-mix  systems were
analyzed.  The results  indicate  that, in
many instances, the addition of a second
spare module is economically justified.
For the system based primarily on coal
capacity (49 peu'ent coal, 24 percent oil),
module outage  time greater than 14
percent favors using a  second spare. In
the oil-based system,(19 percent coal, 57
percent oil), two spare modules are justi-
fied above a 10 percent module forced
outage rate. The assumptions used in this
study are site specific; however, they do
provide an interesting perspective on reli-
ability. However, the most crucial piece of
data, the actual module outage rate for a
well designed system, is not given.

Economics
  The costs of FGD systems are very site
specific.  Coal  composition,  degree  of
control required, and economic bases all
have a  major influence on the reported or
estimated  costs.   Consequently,  the
actual  cost of a "standard" FGD system
can vary considerably,  leading  to
confusion as to what the actual costs are.
Several recent reports have estimated
the costs of various FGD processes for a
standard  size plant by examining the
various design and economic bases. TVA
has published the most recent economic
study of limestone and  lime  systems,
using a  mid-1982  time basis for capital
costs. The standard plant is a  500 MW
unit  burning  3.5 percent sulfur bitumi-
nous coal meeting the 1979  NSPS. A
bypass  duct  capable  of handling 50
percent  of the flue gas is also  included.
Although many site specific factors affect
the capital cost of an FGD system, four
parameters account for most of the cost
differences  between estimates  for  a
standard size plant and fuel: the estimate
date, use of a spare module, bypass ducts,
and the S02 removal requirement.
  Examining  the  various site-specific
cost factors used to estimate the costs of
several  FGD systems, and adjusting the
factors to a common basis, can produce a
fairly consistent set of cost estimates. The
results of this exercise are shown in Table
7, and in  Figures 2 and 3. As  shown in
Table  7   and  Figure   2,  the   raw
(unadjusted) cost estimates  reported for
six  commercial  FGD processes  vary
considerably.  However,   when  these
estimates are  adjusted  to  a common
basis (Figure 3)  the  variance  in the
estimate significantly decreases.
  Costs   were  adjusted,  using  the
Chemical  Engineering plant cost index to
adjust for inflation. A spare module and
associated equipment for a 500 MW unit
was assumed to increase capital costs 25
percent. Bypassducts increased the costs
5 and 10 percent, respectively, for 50 and
100 percent bypass. Increasing the S02
removal from the 1971  to the 1979 NSPS
was estimated as  a 5 percent increase.
  As shown in Table 7, the 1982 cost of
most of  the FGD  systems  is about
$200/kW. The average costs  show the
same relative rankings as in  individual
reports which estimate several processes.
Lime systems have the lowest capital cost,
followed  by  limestone and dual  alkali

-------
Table 7. Estimates of FGD
System Costs
and spray drying are usually the
expensive, followed by lime and
flange of
Number
Unadjusted
of Cost
Process
Limestone
Lime
Dual Alkali
Estimates






Spray Drying*
Wellman-Lord

Magnesium
Oxide
5
3
2
3
4
Costs.
$/kW
Average of
Unadjusted
Costs.
$/kW
98-206
90-192
101-166
102-200
126-267
162
144
134
165
176
Range of
least
dual
Average of alkali, and then, the regenerable systems.
Adjusted
Costs,
$/kW
169-240
155-193
174-214
176-248
181-345
Adjusted The EPA/PEDCo
utility
Cosrs, compiles reported cost data
$/kW systems and adjusts
the
FGD survey
on operating
costs
to a
203 common basis. Averages of the different
processes are shown in Table 8. The low
sulfur systems have the lowest capital
734 and operating costs.
The high sulfur lime-
stone, lime, and dual alkali units all cost
208 between $140 and $150/kW. Since
246 these are operating systems, they are
predominantly units designed for the


4
132- 193
159
195-227
1971 NSPS, with no spare modules or
^ kwnnnf rlim+n nn~J n !»...__ O/-» 	 1
*Costs include a baghouse for paniculate control. An ESP would be used for the other FGD systems requirement Using the previous cost
($38/kW) but was not included in the process cost estimates. multipliers results in a 1 982 cost of $200-
processes. The regenerable systems are
20-30 percent higher. The spray drying
systems include integral particulate
control, and have an average cost of
$208/kW. ESPs for the other processes
will add about 20 percent to the total air
pollution control equipment cost. Note
that the designs
of the spray drying units
have not been successfully demonstrated
for high-sulfur coal.
360
340
320
300
280
260
240
220
200
., 1 on
;*c / OU
w- 160

w
3 740
720
80
60


40
20
0
-
-
-
-
-
-
-
-


_

-
-
.
.


•
-









































































___






















































Limestone
Systems






















The annual costs of the different FGD 220/kW, which is very close to the
processes depend as much on capital estimated costs previously presented. As
charges as fixed and Variable operating can be seen- wnen tne costs are P|aced on
costs. There are too many variables a common basis, they agree very well.
(reagent costs, interest rates,
operating profile return on
labor rates, __
investment SO2 Removal Test
Results
(ROI), etc.) to use a simplistic approach to Before promulgation of the
1979
present standardized annual costs as was NSPS, the EPA performed emission tests
done for the capital costs. In individual on various types of FGD systems. These
reoorts using the same bases, limestone units were built to meet the 1971 NSPS














—



























Lime
Systems










^~"




















































Dual
Alkali
Systems


















































— i












































Spray Wellman-Lord




























































































j^
CM
IQ
CM
W
°9

Magnesium
Drying Systems
Systems
Oxide
Systems
Figure 2.    Raw estimates of FGD system costs.

-------
 o
o
360
340
320
300
2SO
260
240
220
200
180
160
140
120
100
80
60
40
20
0
-
-
-
-
-
-
-
-
-
-
_
-
	





—





	





































































	














































































	
















fx
10
(M
?
Limestone Lime Dual Spray Wellman-Lord Magnesium
Systems Systems Alkali Drying Systems Oxide
Systems Systems Systems
Figure 3.    FGD system cost estimates adjusted to consistent basis.
Table 8.   Average Costs of Operating FGD Systems (1981 Dollars)
Process
Limestone-Low Sulfur*
Lime-Low Sulfur
Sodium Carbonate
Alkaline Ash
Dual Alkali
Limestone-High Sulfur**
Lime-High Sulfur
Wellman-Lord
Number of
Units
16
5
4
7
3
7
15
3
Capital Cost
$/kW
79
72
111
118
147
149
141
272
Annual Cost
mills /kWh
4.8
5.1
6.4
6.7
8.7
9.2
10.5
18.1
* Low sulfur - < 1.5 percent.
**High sulfur -> 1.5 percent.

or state  standards, which were  not as
stringent as the current NSPS. However,
many of the units tested  meet the 1979
NSPS, or are very close to satisfying its
requirements. Based  on  these test re-
sults, EPA set the new standard at a level
attainable  by  current technology. The
sites tested and test results are shown in
Table 9. In addition, a comment column
indicates  the  SO2  removal level  that
would be required for each site to achieve
compliance with the 1979 NSPS. As can
be seen, 7 of the  13 units built under
more lenient standards meet the current
NSPS. Several  of the other units were
within  a few  percentage points  of
meeting  the  standard.  Consequently,
designing and operating an FGD system
to meet the 1979 NSPS required only a
moderate advance in the status of FGD
technology.
  Two readily apparent deficiencies in
the  continuous  emission  monitoring
results data base need to be recognized:
(1) the limestone systems tested were not
typical of those being designed and built,
predominantly using spray towers; and
(2) the  growing  'acceptance  of  spray
drying by the utility industry indicates
that a number of these units should also
be tested.


Conclusions
   The 1979  NSPS for utility  boilers
requires FGD systems on all new coal-
fired  boilers. Since initial promulgation
on September 18,1978, 73 FGD systems
representing over 35,000 MW of capacity
have  been  constructed, contracted, or
planned  by the utility  industry. In the
previous seven years, 48,000 MW (117
units) had been built with FGD systems.
                                   8

-------
Table 9.   FGD System Continuous Emission Monitoring Results
Process
    Unit
               Emission
    SO2 Removal   Rate
     Efficiency  Ib SO2//06
     percent      Btu   Comments
Limestone


Lime








Dual Alkali


Wellman-Lord
Magnesium Oxide
Spray Drying

Shawnee V/ST*
Lawrence 4
Southwest 1
Cane Run 4

Conesville 5

Bruce Mansfield 1

Shawnee TCA

Pleasants 1
A.B. Brown 1

Cane Run 6
D.H. Mitchell 1 1
Eddy stone 1A
Celanese Amcelle^

96.1
96.6
>90
83.8

89.2

85.3

88.6

>90
87.7

91.6
89.6
96.8
70.0

0.22
0.03
<0.64
0.88

0.80

0.78

0.64

<0.50
0.85

0.70
0.60
0.16
0.92

meets 1979 NSPS
meets 1979 NSPS
meets 1979 NSPS
88.8% removal needed
for 1979 NSPS
90.0% removal needed
for 1979 NSPS
88.5% removal needed
for 1979 NSPS
89.3% removal needed
for 1979 NSPS
meets 1979 NSPS
90.0% removal needed
for 1979 NSPS
meets 1979 NSPS
meets 1979 NSPS
meets 1979 NSPS
80.4% removal needed
for 1979 NSPS
*Units tested with adipic acid additive.
VndustrM boiler, utility SO2 NSPS not directly comparable.

Since  the  promulgation  of  the  1979
NSPS, the  wet limestone FGD process
has been selected for nearly two-thirds of
the future plants. The predominance of
this  process  is due to  its technical
flexibility   and  favorable   economics.
Improvements in reliability and efficiency
(due to the use of spray towers, forced
oxidation, and SO2  removal enhance-
ment additives}  enable  the limestone
process to be used  in many applications.
  The use of spray drying to  control SO2
and particulates from lower sulfur coal is
the only  other technology to be widely
favored in the industry. Many commercial
spray drying systems will soon be on-line
demonstrating this technology. The other
throwaway systems (lime, dual alkali,
alkaline  ash,  sodium  carbonate) will
probably  be used  infrequently  at  new
plants. Higher reagent costs are the
primary reason for the lack of new orders
for these  systems.  The regenerable
systems,  due to their higher costs, will
probably  be  used  only  in  specialized
applications;  e.g.,  land-restricted sites
and those with favorable S02 regenera-
tion opportunities.

Conversion Factors
Nonmetric	Times   Equals Metric
                           G. P. Behrens andJ. C. Dickerman are with Radian Corp., Austin, TX 78766.
                           J. David Mobley is the EPA Project Officer (see below).
                           The complete report, entitled "The Current Status of Commercial Utility Flue Gas
                             Desulfurization Systems," (Order No. PB 84-133 016; Cost: $20.50, subject to
                             change) will be available only from:
                                   National Technical Information Service
                                   5285 Port Royal Road
                                   Springfield, VA 22161
                                   Telephone: 703-487-4650
                           The EPA Project Officer can be contacted at:
                                   Industrial Environmental Research Laboratory
                                   U.S. Environmental Protection Agency
                                   Research Triangle Park, NC 27711
Btu
Btu/lb
lb/106 Btu
1055
2326
 430
  J
J/kg
ng/J

-------
                                                                                                                    (D


                                                                                                                    O
30   "^
r«i   o
m   z
                                                                                                                    g

                                                                                                                    5'
                                                                                                                    o

                                                                                                                    5'

                                                                                                                    0)
O
I

*>.
tn
NJ
o>
oo
                                                ft US. GOVERNMENT PRINTING OFFICE: 1984-759-102/87

-------