United States
 Environmental Protection
 Agency
Industrial Environmental Research
Laboratory
Research Triangle Park NC 27711
 Research and Development
EPA-600/S7-81-159  August 1982
 Project Summary
Dual Alkali Acceptance Test at
Louisville  Gas and  Electric
Company
D. A. Watson, T. M. Martin, J. K. Donnelly, N. Zoueshtiagh, R. P. Van Ness, C.
R. La Mantia, and L. R. Woodland
  This report discusses results of the re-
cent acceptance test on the dual alkali
system serving Louisville Gas and Electric
Company's Cane Run Unit 6 boiler. The
test was conducted to measure system
performance with respect to the guaran-
tees offered Louisville Gas and Electric
by Combustion Equipment Associates.
Testing results were:
  • SO2 removal averaged 94% and
    143 ppm outlet concentration.
  • Lime consumption averaged 1.04
    mole CaO/mo4e S02 removed.
  • Power consumption averaged 1.05%
    of generation.
  • Filter cake solids averaged 52.2 wt
    % Insoluble solids.
  • There was no net paniculate matter
    addition.
Various problems  attributable to the
boiler, the FGD system, and the quality
and quantity of the carbide Urne supplied
to the system delayed acceptance tests
until July 1980. The year-long demon-
stration officially started in May 1980.
The problems experienced and their
solutions are discussed.

  This Project Summary wes developed
by iPA'3 Industrie! Environmental Re-
seerch Leboretory, Reseerch  Triengje
Perk, NC, to announce key findings of
the reseerch project thet Is fuiy docu-
mented In e seperete report of the seme
title (see Project Report ordering Infor-
mation et bock).
Introduction

  The Dual Alkali Demonstration Project
is a joint effort by a number of organiza-
tions under the sponsorship of the U.S.
EPA. The process is a sodium-based
concentrated mode using carbide lime as
a regenerant. Louisville Gas and Electric
Company (LG&E)  owns/operates the
dual alkali system serving their Cane Run
Unit 6 boiler, a nominal 280 MW high-
sulfur-coal-fired boiler (3.5-4.0%  S).
The design was developed by Combus-
tion Equipment Associates (CEA) and
Arthur D. Little, Inc. (ADD. The system
was erected by LG&E under the guid-
ance of CEA/ADL at a cost of about $ 22
million (1976-1980 dollars)  or about
$79/kW installed generating capacity
(including waste disposal).
  Figure 1 is a  process flow schematic
of the dual alkali process at Cane Run 6.
Flue gas from the boiler passes through
electrostatic precipitators (ESPs) and is
fed to two absorbers. A recycling sodium
sulfite solution, flowing countercurrent
to the flue gas across two stainless steel
perforated plate trays,  absorbs S02
according to the reaction:
S03
              H20-»2HS03-  (1)
  In addition, due both to the absorption
of sulfur trioxide from the gas and to the
oxidation of sulfite ion in solution, sulfate
(S04 =) is formed in the absorbent liquor:

-------
 H2 + S03-»H2S04-»2H+ + S04= (2)
S03
           02-«-S04 =
(3)
 The scrubbed flue gas is reheated by
 combustion gases from a direct oil-fired
 reheater and is ducted to the stack.
   Sodium carbonate is added to either
• the thickener or the absorber to make up
 for losses of sodium in the system. Bleed
 streams of the spent absorbent solution
 from the absorbers are sent to the regen-
 erator reactor trains where carbide lime
 is added to convert the bisulfite (HSO3 - )
 in the spent absorbent to sulf ite (S03 = )
 in the regenerated absorbent, precipitat-
 ing a mixture of calcium sulfite and sul-
 fate solids:
  2HS03- + Ca(OH)2-
        CaS03*+S03 =
                          2H2O  (4)
  SO4= + 2HSO3- + Ca
-------
period officially started in  May 1980.
The problems and solutions are discussed
later.
  The acceptance test was conducted
from July 17 to July 28,  1980.  With
one minor exception (filter cake quality),
the system proved to be capable of suc-
cessfully meeting its performance  guar-
antees.

Acceptance Test Results
  The 12-day acceptance test was con-
ducted to measure the performance of
the dual alkali system with respect to the
guarantees provided to  Louisville Gas
and  Electric Company by  Combustion
Equipment Associates. The guarantees
concerned operation in  the following
areas:
  • SO 2 removal.
  • Carbide lime consumption.
  • Soda ash consumption.
  • Participate matter emissions.
  • Power consumption.
  • Filter cake quality.
  • Year-long system availability.
  Table 2 summarizes the guarantees
offered and corresponding results of the
acceptance test. A brief discussion of
each guarantee test performed during
the acceptance test follows.

SO 2 Removal
  The primary method  of  determining
SO2 removal relied on a continuous Lear
Siegler monitor in the stack. This analyzer
was certified in December  1979 by an
outside contractor according to the pro-
cedure specified in the Federal Register.
During the acceptance test, as a backup
to the continuous monitor and as an on-
going confirmation of the analyzer accu-
racy, wet-chemical EPA Method 6  tests
of the stack effluent were also performed
daily, in conjunction with the paniculate
tests.
  Preliminary wet-chemical test results
showed a discrepancy between these
measurements and the continuous  mon-
itor readout.  An extensive check of the
system revealed a burned ground wire in
the signal line of the Lear Siegler monitor.
Data on the calibration sequences of the
analyzer prior to, during, and after elimi-
nation of the grounding  problem, indi-
cated that the Signal from the analyzer
was  offset 30 ppm on the  low side by
the malfunction.
  Therefore the continuous stack S02
monitor readings for the first 7 days of
the tests were corrected by 30 ppm.
  'ith this correction applied to the  early
  tidings, and subsequent to the repairs
Tab/a 1.    Performance Conditions
                                    Design
                Observed
Coal (Dry Basis)
  Sulfur, %                          5.0
  Chloride, %                        0.04
  Heat Content, Btu, Ib               11,000
Inlet Gas:
  Flow Rate (Volumetric), acfm        1,065,000
  Temperature, °F                   300
  SO 2, ppm                         3471
  O2, %                            5.7
  Paniculate, Ib/106 Btu      - -     0.10
Outlet Gas:
  S02/ ppm                         <200
  Paniculate, lb/106 Btu              0.10
Boiler Operation:
  Generation, MW                   280
                3.7  (ave.)
                0.02  (ave.)
                10,650  (ave.)

                1,045,000  (max.)
                280  (max.)
                2323  (ave.)
                6.7  (ave.)
                0.84  (ave.)

                143  (ave.)
                0.10  (ave.)

                240  (max.)
Table 2.    Performance Guarantees and Acceptance Test Results
               Guarantee
            Test Results
   SO 2 Removal
     220 ppm dry basis (D.B.)
     without additional air dilution
   Calcium Consumption
     1.05 moles available CaO/mole
     SO 2 removed
   Soda Ash Consumption
     0.045 moles Na2CO3/mole
     SO 2 removed
   Net Particulate Addition
     No net paniculate addition
     by FGD system
   Power Consumption
     System will consume (excluding
     reheat) not more than 1.2% of
     power generated at peak capacity
   Filter Cake Properties
     Filter cake  will contain a minimum
     of 55 wt % insoluble solids
 143 ppm (D.B.) without additional
air dilution
 1.04 moles available CaO/mole
SO2 removed

0.042 moles Na2CO3/mole
SO 2 removed

Net paniculate removal averaging
88% efficiency

System consumed 1.05% of power
generated
Filter cake averaged 52.2 wt %
insoluble solids
to the ground in the analyzer, the two
techniques were in good agreement.

  Both measurements showed that the
system could meet the 200 ppm S02
outlet concentration guarantee. Table 3
summarizes the  24-hour average SO2
results for the 12-day acceptance test.
Table 4 summarizes  the simultaneous
wet  chemical and continuous monitor
measurements (the Method 6 tests were
conducted only for the first 10 days).

Lime Consumption Guentntee
  The lime consumption guarantee was
specified as  "not [to] exceed  1.05
moles of available CaO in the lime per
mole of S02 removed from the flue gas.''
 Lime consumption was determined by
 analyzing representative samples of fil-
 ter cake collected as the cake was dis-
 charged from the filters prior to fixation.
 The cake was analyzed for total calcium
 and total sulfur. The total calcium repre-
 sented the lime used, and the total sulfur
 represented SO2  removed from the flue
 gas. A portion of the calcium entering
 the system with the carbide lime is pre-
 sent as carbonate and therefore does not
 represent alkalinity available for regener-
 ation. Each time the lime day tank was
 filled, a sample of time was analyzed for
 available alkalinity and total  calcium.
 From these results, a correction factor.
 was developed to account for unreactive
 calcium in the carbide lime feed. During

-------
the 12-day acceptance test, the calcium
consumption, corrected for available al-
kalinity as described above,  averaged
1.04 moles  of  available CaO/mole of
S02 removed, thus meeting the guaran-
tee which required less than 1.05 moles/
mole of sulfur removed.  Table 5 sum-
marizes the analyses of the filter cake.
Soda Ash Consumption
  Soda ash consumption was determined
by analysis of total sodium and total sul-
fur in the filter cake. According to this
analysis, consumption of soda ash aver-
aged 0.042  moles of Na2C03/mole of
SOj  removed,  meeting the guaranteed
requirement of  0.045.
Table 3. Acceptance Test Continuous SO2 Analysis
Acceptance
Test
Day
1
2
3
4
5
6
7
8
9
10
11
12
Average
24 Hour

A Inlet
2444
2674
a
a
2265
2567
2113
2116
2395
2372
2292
2167
2340
Continuous SO^ Analyzer Results
(ppm.
B Inlet
2418
2570
2390
2290
2315
2515
2021
2088
2339
2315
2233
2166
2305
dry basis)
Stack
130
129
130
152
157
140
132
124
146
171
156
130
141

% Removal
94.7
95.0
94.6
93.4
93.1
94.5
93.6
94.1
93.8
92.7
93.1
94.0
93.9
UG
gal.
103 act
9.5
8.7
9.4
9.2
9.1
7.8
8.3
8.9
8.2
8.7
9.6
9.8
8.9
Act.
Na
(Mi
0.42
0.40
0.42
0.44
0.54
0.52
0.49
0.49
0.49
0.49
0.43
0.43
0.46
'Analyzer printout malfunction

Table 4.    Acceptance Test Continuous Monitor and EPA Method 6 Analysis

                     	      S02 Concentration, ppm, dry basis
                           A Inlet
      B Inlet
Stack
  Acceptance          Dupont           Dupont             LSI
     Day     Hours  Analyzer Method 6 Analyzer Method 6 Analyzer" Method 6
1
2
3
4
5
6
7
8
9
10
1400-
1700
1100-
1300
100O-
1300
1200-
1500
1000-
1300
1000-
1300
1600-
1900
1100-
1300
1100-
1400
0900-
1200
2434
2434
2592
2836
2656
2716
2337
2395
2864
2690
2330
2150
2210
2390
2330
2350
2O40
2120
2530
2410
2516
2423
2670
2674
2606
2418
2250
2330
2721
2624
2330
2180
2290
2480
2410
2480
2030
2100
2450
2360
119
122
117
155
136
184
b
113
122
160
124
163
154
159
137
212
137
130
133
137
• Lear Siegler analyzer readings for days 1-6 corrected for the effect of the burned out
  ground wire
b Analyzer out of service for repairs to ground wire
Particulate Emission
  The  system.was guaranteed  not
make any net addition of paniculate
the gas stream prior to discharge. EP
Method 5 particulate  tests were  coi
ducted at the inlets to the absorber men
ules and in the stack (downstream of n
heaters) during the acceptance test. Tr
results of 10 simultaneous tests showe
convincingly that there was no net add
tion of  particulate across  the systen
Actually, the absorber  performed  as
particulate removal device,  averagin
88% net removal of incoming particular*
Table 6 shows  results of particulat
tests during the test program.
  Although the FGD system met guarar
tee requirements, the test was not ver
stringent due to the low level of perfoi
mance by the ESPs during the acceptanc
test. The FGD system was originally de
signed  to process an incoming flue ga
stream containing the equivalent of 0.'
Ib of particulate/106 Btu or less.  Durinj
the acceptance test, however, the leve
of incoming  particulate matter was al
most an order of magnitude higher. Thus
it is not surprising that the  absorber;
removed particulate even at the relative
ly low pressure drop at which they oper
ated. The particulate emissions from the
stack, however, were on the order of 0.1
lb/106  Btu as required for the Cane Rur
Unit 6 FGD system under the appropriate
requirements to control particulate emis-
sions.

Power Consumption
  The  system, excluding  reheat,  was
guaranteed not to use more than 1.2%
of the total power generated  by the boiler/
turbine unit at  gross peak  load.  During
the acceptance test, peak generation
was 240 MW.  Correspondingly,  power
consumed during peak generation  was
2.5 MW, or 1.05%. The guarantee was
met, based on both peak generation and
average generation during  the test.
During the 12-day test, the average load
was 178  MW  and the average  power
consumption by the FGD system  was
2.05 MW, or 1.15%.

Waste Filter Cake Properties
  The system was guaranteed to produce
a waste filter cake containing a minimum
of 55 wt % insoluble solids. The filter
cake averaged 52.2 wt % insoluble solids
during the acceptance test. While this
fell slightly short of guarantee, the pro-
duct discharged  to the IUCS process
was uniform in moisture content a
was suitable for working into a s

-------
r«6fo 5.    Acceptance Test Daily A verage Filter Cake Analysis

               As Received Basis
Test Day
1
2
3
4
5
6
7
8
9
10
11
12
Na
wt%
0.55
0.58
0.70
0.48
0.45
1.11
0.58
0.50
1.07
0.45
0.77
0.62
Ca
wt%
14.88
14.60
15.64
15.72
15.15
15.20
14.35
14.44
14.05
14.43
14.85
13.74
Total Sulfur
as SO 4
wt%
31.35
31.58
32.60
32.68
31.92
33.20
31.80
32.32
32.58
33.07
33.48
31.64
Insoluble
Solids
wt%
52.65
52.20
52.60
53.72
53.92
51.90
50.70
51.40
51.00
52.43
52.82
50.58
Mole Na2CO3
Mole SO 2
0.037
0.038
0.045
0.031
0.029
0.070
0.038
0.032
0.068
0.028
0.047
0.041
Mole CaO»
Mole SO 2
1.139
1.109
1.151
1.154
1.139
1.099
1.083
1.072
1.035
1.047
1.064
1.042
 Average
52.16
0.042
1.095
^Calcium consumption corrected for available alkalinity (1.095 x 0.95 = 1.040).
Correction factor of 0.95 was developed from analysis of incoming carbide lime for
mole of available alkalinity/mole of total calcium.

Table 6.    Acceptance Test Paniculate Test Results'

                                      Paniculate, Ib/W^Btu
Acceptance Test Day
1
2
3
4
5
6
7
8
9
10
Average
A Inlet
0.5320
0.6590
0.9470
0.9440
1. 1 100
0.9900
0.5890
0.6250
0. 7890
0.9620
0.8147
B Inlet
0.7120
0.3620
1.0700
0.8060
0.9200
1.4900
0.8470
0.6490
1.2000
0.5930
0.8649
Stack
0.0895
0.0932
0.11 1O
0. 1030
0. 1020
0.1020
0. 102O
0.0893
0. 1 100
0. 1020
0. 1004
% Removal
85.6
81.7
89.0
88.2
90.0
91.8
85.6
86.0
88.9
86.9
88.0
and manageable product through the fix-
ative process. Optimization of filter cloth
selection and filter cycle will continue,
with the goal of showing that compliance
with this guarantee can be  met during
the demonstration year.

System Availability
  System availability, as defined by the
Edison Electric Institute (available hours
divided by the total hours in the period
under construction), was guaranteed to
be greater than 90% for the demonstra-
tion year. While it is too early to report
such a figure, through the first 4 months
of the demonstration year (May-August),
system availability has averaged 99.8%.

Operating and Maintenance
Problems
  Up to the time of acceptance testing, a
number of mechanical problems and a
few chemical problems affected system
performance and led to cumulative delays
in executing the program. None of the
problems were insurmountable; but their
solutions were time consuming. It is im-
portant to  report the nature of these
obstacles so that future installations of
this or similar technology can benefit
from the experience.
Recycle and Thickener Return
Pumps
  There were two major problems with
the high-capacity low-speed pumps for
recirculation of absorbent liquor to the
trays, and return of thickener overflow li-
quor to the absorbers. The first problem
was the mechanical shearing of the im-
pellers at the hub. The original pump im-
pellers were manufactured in two parts:
a body and a separate hub for attach-
ment to the shaft. The hub was welded
to the body. All of the impeller failures
were at this welded seam. This problem
was eliminated when the pump vendor
supplied  a one-piece  molded impeller
body.
  The second major problem involved
the rapid failure of the suction side of the
pump liner. As a result of close tolerances
between the casing liner and the impeller,
the two surfaces were rubbing; the re-
sulting abrasion destroyed the liner. After
completely dismantling  the  pumps,  it
was discovered that a  finishing step ap-
peared to have been omitted at the fac-
tory, leaving about %-in. excess length
on each shaft. Milling  each shaft to its
design size eliminated this problem.

Mist Eliminator Collapse
  Within a few months after start-up,
both absorber modules experienced high
pressure drop problems. Inspection of
the internal structure revealed that the
mist eliminator sections had sagged or
collapsed structurally.  The problem was
solved by replacing the  mist eliminator
sections with those of  a different manu-
facturer. Since the replacement, in Aug-
ust 1979, there has  been no further
problem with the mist  eliminators.
Tray Pluggage
  One of the most perplexing problems
was the pluggage of the absorber trays
due to deposition. At first the observed
deposit  was thought  to be  carbonate
scale resulting from pH upsets in the
modules. Careful analysis showed the
precipitate to be an aluminum-hydroxy-
silicate complex. The mechanism of dis-
solution and subsequent deposition was
traced to the operating pH of the reac-
tion train. Aluminum was  entering the
system  with  the carbide lime. At the
above-11.5 operating pH in the reactor,
the aluminum compound is soluble in the
liquor. When the thickener overflow re-
cycle combined with  the  recirculating
absorbent, the resultant  drop  in pH
caused the aluminum  to precipitate on
the absorber trays.
  Reducing the operating pH of the reac-
tor to between 10.0 and 10.8 reduced
the solubility of the aluminum with the
reactor and thickener. This change ahead
of the absorber minimized pluggage. At
the reduced pH set point, however, there
is less buffering and control  of reactor
pH is more difficult.

Water Balance
  The system initially experienced a
severe water imbalance. This was partly
due to a lack of familiarity with the sys-

-------
tern, and partly because  of low-solids
concentration in the carbide lime feed.
The other lime slurry systems at Cane
Run can tolerate  an occasional open-
loop excursion. However, the dual alkali
process must operate in a closed-loop at
all times, since the high concentration of
solubles in the scrubbing liquor makes
disposal unacceptable for both environ-
mental and economic reasons. The sys-
tem was designed to accommodate 70%
water (30%  solids) in the incoming car-
bide lime slurry. Initially the water con-
tent was consistently 82-85%. At this
concentration, the system was receiving
twice the design input water flow. After
only a few hours of operation, the vol-
ume of water in the system had accumu-
lated to the  point where the lime feed
had to be cut off. The absorbers con-
tinued to function as evaporators until
the water level dropped low enough to
resume normal operation.
  Strict control of the incoming  lime
concentration from the  supplier and the
addition of a ball-mill/hydroclone system
to remove oversize particles alleviated
the problem.

Soda Ash Silo Pluggage
  Soda ash is added to  the system by a
dry weigh feeder which feeds dry solids
from a storage silo to a mix tank where it
is mixed with absorbent liquor. Vented
moisture vapor from the  hot mix tank
backs up into the weigh  feeder screw
conveyor and causes the soda  ash to
form lumps which prevent the smooth
flow of feed to the system. The system's
small fan, used to blow the moisture-
laden air back into the mix tank, proved
to be under-designed. Although a larger
fan improved the situation, the soda ash
feed system still remains a relatively high
maintenance item.

Thickener Blockage
  In mid-January  1980,  the thickener
rake seized during a boiler outage for re-
pair and ultimately required a shutdown
and major overhaul of the thickener. This
did not  occur during normal operation,
but rather during the transient period in
which the boiler and FGD were being shut
down for maintenance. The stoppage
was postulated to have  resulted from an
overloading   of  the  thickener with
washings of accumulated solids  (in-
cluding fly ash) from the bottom of the
absorber. Lacking a bottom drawoff, the
absorber allowed fly ash to be trapped
and accumulated  in its lower portion.
The problem  could apparently have been
avoided if the solids from the bottom of
the absorber had been slowly pumped to
the thickener while the thickener and
filters continued in operation until the
absorber bottom was purged of solids.
  Correction of the problem took about
3 weeks, during which about 2 million
gal. of liquid and solids had to be removed
from the thickener (liquid was temporarily
stored, and solids were impounded off
site).  To accomplish this,  large holes
were cut in the  thickener sides to allow
entry by personnel and equipment to dig
out the compacted solids.
  Overloading of the thickener has not
recurred. The solids in the bottom of the
absorbers are still not subjected to me-
chanical  agitation, but  are no longer
washed into the thickener in large slugs.

SO* Monitoring
  SO2 jn  the inlet to and outlet from the
absorbers is measured by  continuous
DuPont UV Model 460 S02 analyzers. In
the stack,  S02 concentration  of the
scrubbed gas is measured by  a Lear
Siegler SO2 analyzer.
  Three problems  have occurred in the
measurement of S02 using the DuPont
analyzers supplied with  the dual alkali
system:
  • Plugging of the sample probe.
  • Maintaining a steady calibration of
    the instruments.
  • Stratification of scrubbed gas across
    the absorber exit duct.
  The first two problems were minimized
by daily inspections to determine if the
probes need to be calibrated or cleaned.
An attempt to alleviate the last problem
will be made by moving the SO2 probes
downstream of the  reheaters, which
should also help reduce the first two
problems.

Failure  ofFRP Piping
  The FRP (fiberglass reinforced plastic)
piping in slurry service  (i.e., thickener
underflow and filter feed) has been a ma-
jor maintenance item. Some failures have
been major, others minor. Late in the fall
of 1979, a flush connection on the under-
flow line snapped off. Slurry from  the
thickener flooded  the  access tunnel
below the  thickener before the break
could be isolated. Routinely, elbows in
the line from the thickener to the filter
have required repairs because of erosion
damage  and failure of the connection
bond. Gradually all underflow FRP piping
is being replaced with mild steel. While
mild steel has a limited life span in this
service, failures will be less catastrophic.
pH Control
  Reliable and accurate pH measurement
for pH control in the reactors and in thi
scrubber bleed stream have been panic
ularly bothersome. The pH related prob
lems are attributed to:
  • Inability to keep the probes clean.
  • Poor responsiveness of the probes
  • Pluggage of the sample lines.
  • Poor calibration techniques.
  Experiments with different instrumen
designs and sampling methods are grad
ually alleviating the first three problems
Detailed calibration  instructions anc
cross checking of the results have mini-
mized the last. On-line  pH readings art
compared daily with pH measurement:
taken with a portable pH meter by LG&E
scrubber laboratory  personnel. If these
readings are in disharmony by more than
0.3 pH  units, the on-line probes are
recalibrated.
  All  the original L&N  pH probes have
been replaced with Great Lakes models.
To measure the  pH of  the primary and
secondary reactors, a Great Lakes Model
60 submersible probe was placed in the
overflow chute from the primary to the
secondary reactor, and  in the secondary
reactor  below the liquid level near the
overflow. The pH of the bleed and thick-
ener return streams was measured  by
Great Lakes Model 60 flow-through pH
probes with ultrasonic cleaners.

Filter Operation
  There have been two major concerns
with the rotary  vacuum drum filters.
First,  the cake quality has varied be-
tween 45% and  55% solids. Second, it
has not always been possible to properly
wash the cake to meet sodium consump-
tion guarantee.
  Prior to the acceptance test, experi-
ments with different filter cloths led to
installation of a new filter cloth. The ori-
ginal cloth was  a polypropylene cloth
supplied by National  Filter  Media  of
Hamden, CN.  During  the acceptance
test this cloth was replaced with a multi-
filament nylon cloth supplied by Thoerner
Products Corp., of Pittsburgh, PA. The
new cloth produced a more consistent
quality cake but had a tendency to blind.
During the acceptance test, cake wash-
ing was sufficient to meet the soda ash
consumption guarantee, but the blinding
detrimentally affected the percent solids
of the filter cake. There is also some con-
cern that poorer quality solids may  be
produced in the reactors at the lower pH
levels required to control dissolution ofa
aluminum and silicon compounds.    •

-------
  Proper cake washing on the filters is
subject to a number of considerations.
The wash water rate (as limited by the
water balance), the quality of solids pro-
duced, the thickness of the cake (con-
trolled by drum  angular velocity), the
wash spray configuration, and the quality
of the filter cloth (blinding characteristics)
are all important. Therefore, experiments
with different filter cloths and varying
operating parameters are continuing.

Conclusions to Date
  As indicated by operation since March
1980, and the successful completion of
the acceptance test  in July,  the  dual
alkali process can achieve greater than
90% S02 removal with an availability of
more than 99% while processing a flue
gas generated in a high-sulfur (>3:5%)
coal-fired full-size (280 MW) utility
boiler.  Consumption  of raw materials
and  power  was  less  than expected
(guaranteed)  while  the SO2  removal
average was over 94% for the 12-day
acceptance test.
  Most of the problems initially encoun-
tered were mechanical and  have been
solved or greatly reduced in the opera-
tion at Louisville Gas & Electric's 280
MW Cane Run Unit 6.
  Further investigation of filter opera-
tion, reactor operation,  filter  cloths,
materials of construction, and major pro-
cess component characterization is
underway.

Units of Measure
  EPA policy is to express all measure-
ments  in Agency documents in metric
units. When implementing this practice
will result in undue cost or difficulty in
clarity, IERL-RTP provides  conversion
factors for the non-metric units. Gener-
ally, this  paper  uses British  units of
measure.
  The following equivalents can be used
for conversion to the metric system:
British
%(°F-32)
1 ft
1 ft2
1 ft3
igr
1 in.
1 in.2
1 in.3
1 Ib (avoir.)
1 ton (long)
1 ton (short)
1 gal.
1 Btu
Metric
°C
0.3048 m
0.0929 m2
0.0283 m3
0.0648 g
2.54cm
6.452 cm2
16.39cm3
0.4536 kg
1.0160 mtons
0.9072 m tons
3.7854 liters
252 calories
References
1.  Van Ness, R.P. et al., "Full-Scale
   Dual-Alkali Demonstration System at
   Louisville Gas and Electric Co.—Final
   Design and System Cost," EPA-600/
   7-79-221b, NTISNo. PB80-146715,
   September 1979.
2.  Van Ness, R.P., et al., "Project Man-
   ual for Full-Scale Dual-Alkali Demon-
   stration at Louisville Gas and Electric
   Co.—Preliminary Design and Cost Es-
   timate," EPA-600/7-78-010,  NTIS
   No. PB278722, January 1978.
3.  Kaplan, N., "Summary of Utility Dual
   Alkali  Systems," in  Proceedings:
   Symposium on Flue Gas Desulfuriza-
   tion—Las  Vegas, Nevada,  March
   1979,  Volume  II,  EPA-600/7-79-
   167b, NTIS No. PB80-133176 (pp
   888-958), July 1979.
   D. A. Watson, T. M. Martin, J. K. Donnelly, andN. Zouesht/agh are withBechtel
     national. Inc., San Francsico, CA 94119; R. P. Van Ness is with Louisville Gas
     and Electric Company, Louisville, KY 40201; C. R.  La Manila and L R.
     Woodland are with Arthur D. Little, Inc., Cambridge, MA 02140.
   Norman Kaplan is the EPA Project Officer (see below).
   The complete report consists of three volumes, entitled "Dual Alkali Acceptance
     Test at Louisville Gas and Electric Company,"
       "Volume I. Acceptance Test and Appendices A-C." (Order No. PB 82-231
       267; Cost: $ 12.00, subject to change)
       "Volume II. Appendices D-F," (Order No. PB 82-231 275; Cost: $30.00,
       subject to change)
       "Volume III. Appendices G-J," (Order No. PB 82-231 283; Cost: $27.0O,
       subject to change)
   The above reports wilt 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
                                                                                                  •US QPO:1M2-S5B-092-453

-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Postage and
Fees Paid
Environmental
Protection
Agency
EPA 335
Official Business
Penalty for Private Use $300

-------