United States
Environmental Protection
Agency
Industrial Environmental Research
Laboratory
Research Triangle Park NC 27711
J*
_^
Research and Development
EPA-600/S7-83-039 Oct. 1983
Summary
Full-Scale Dual Alkali FGD
Demonstration at Louisville Gas
and Electric Company
R. P. VanNess, L R Woodland, and E. D. Gibson
This report summarizes the 1-year
demonstration (the demonstration was
extended to 13 months and is hereafter
referred to as "the demonstration") of
the full-scale dual-alkali flue gas de-
surfurization (FGD system, at Louisville
Gas & Electric Co.'s (LG&E) Cane Run
Unit6. System performance is described
in terms of performance guarantees
and other parameters that were moni-
tored throughout the demonstration.
The report gives a detailed history of
operation, including problems encoun-
tered in system operation and how
they were resolved. Capital and operat-
ing costs (estimated and incurred) are
also reviewed.
The overall demonstration consisted
of four phases: Phase I - preliminary
design and cost estimates; Phase II -
engineering design, construction, and
mechanical testing; Phase Ill-start-up
and acceptance testing; and Phase IV-
1-year operation and testing. This re-
port summarizes Phase IV. (Phases 1,11,
and III are summarized, respectively, in
reports EPA-600/7-78-010 (NTIS
PB278722), EPA-600/7-79-22lb(NT1S
PB80-146715), and EPA-600/7-81-
159a, b, and c( NTIS PB 82-231267, -
231275, and-231283).)
This Project Summary was developed
by EPA's Industrial environmental He-
search Laboratory, Research Triangle
Park. NC, to announce key findings of
the research project that is fully doc-
umented in a separate report of the
same title (see Project Report ordering
information at back).
Introduction
The dual-alkali process was developed by
Combustion Equipment Associates, Inc.
(CEA, now known as Thyssen-CEA En-
vironmental Systems, I no.) and Arthur D.
Little, Inc. (ADL) to overcome the dis-
advantages of lime and limestone scrubbing
(such as scaling) while retaining the cost
advantages of a throwaway system. The
main features that distinguish it from
conventional direct lime and limestone
scrubbing are: (1) using a clear solution
rather than a slurry to contact the flue gas
in the absorber and (2) reaction of the
solution in a separate absorbent regenera-
tion section to form the waste solids rather
than forming the waste solids as part of
the scrubbing operatioa
Using a scrubbing solution instead of a
slurry minimizes scale potential in the
scrubber circuit and permits high S02
removal efficiencies over a wide range of
inlet S02 concentrations. Precipitation of
the waste calcium sulfite sulfate solids
outside the scrubber facilitates formation
of waste solids with good dewatering
properties.
The U.S. EPA selected LG&E's Cane
Run Unit 6, an existing coal-fired boiler of
300-MW peak capacity, as the full-scale
demonstration plant for dual-alkali FGD
and participated in funding the operation,
testing, and reporting of the project
As a demonstration system, the purpose
of the installation and operation was to
establish:
• Overall performance- S02 removal,
lime utilization, sodium makeup, re-
generation of spent liquor, water
balance, scaling or solids buildup,
materials of construction, waste cake
properties, reliability, and availability.
• Economics- capital investment and
operating cost
-------
Process Description
The LG&E dual alkali system (Figure 1)
has five process sections: flue gas scrub-
bing, absorbent regeneration, solids de-
watering, raw materials preparation, and
waste disposal
The flue gas scrubbing section consists
of two identical scrubber modules, each
containing a booster fan, an absorber, an
oil-fired flue gas reheater, and two recircu-
lation pumps (one operating and one back-
up). As the flue gas flows through the
absorber, it is contacted by a recirculating
solution containing a mixture of sodium
salts-sodium sulfite(Na2SOs) and sodium
hydroxide (NaOH) are the primary active
species. S02 diffuses into this solution
and reacts with the sodium-based alkali to
form soluble sulfur oxide (SOJ salts, which
are drawn off in the scrubber effluent
Desulfurized flue gas leaves the absorber
and is reheated before passing through
the stack to the atmosphere
The SOx-rich scrubber effluent is routed
to the absorbent regeneration section to be
reacted with carbide lima The reaction
regenerates the sodium-based alkali for
recycle to the absorber and precipitates
the SOX as a mixed crystal of calcium
sulfite and calcium sutfate The absorbent
regeneration section of the LG&E system
consists of two parallel reactor trains. Each
contains a primary and secondary reactor.
The two-stage reactor system prolongs
the reactor residence time and produces
solids with improved settling and filtering
characteristics.
The precipitated SOX salts are separated
from the scrubbing liquor and concentrated
for disposal in the solids dewatering section.
The slurry from the secondary reactor is
fed to the thickener, where it settles to 15-
30 wt % solids. Clear liquor overflow from
the thickener is collected in the thickener
hold tank, which provides surge capacity
for the absorbent liquor fed to the scrubber
system and maintains overall control of
the volume of liquor in the system. The
solids are dewatered further in a vacuum
filter and are washed to recover sodium
salts before disposal There are three filters
at LG&E, each rated to handle 50% of the
total solids produced at design conditions.
For optimum performance, the filters should
run at fixed conditions. The number of
filters to be placed on-line is determined
by the quantity of solids accumulated in the
thickener.
The dual-alkali FGD process requires
two materials: lime for process liquor
regeneratioaand soda ash to compensate
for sodium losses. At Cane Run, carbide
lime byproduct is obtained from a local
acetylene plant It is routed through a
To Stack
Mist Eliminator
Solid Waste
to Landfill
f Makeup /
S. /VazCOa \
Fly Ash —, Quicklime Wafer
Vacuum Filter
Figure 1. Generalized process flow of the dual-alkali system at LG&E Cane Run No. 6.
classifying and grinding system, and the
fine fraction is stored and used to feed the
primary reactor. Dry, dense soda ash is
dissolved in thickener overflow liquor and
recycled to the thickener or absorbers.
The FGD waste solids from the filters
must be treated to produce a more physi-
cally stable and water-impermeable sludge
for disposal The treatment process used
at Cane Run was developed by Conversion
Systems, Inc. (CSI). Fitter cake from the
dual-alkali system is conveyed to the CSI
processing plant and combined with quick-
lime and fly ash from Unit 6. The fly ash
and limestone feeds are adjusted according
to the filter cake discharge rate and amount
to approximately 60% and 2-3% (weight
basis), respectively, of the dry filter cake
The waste product is transported to a dry
landfill on LG&E property south of the
plant
Process Guarantees
The CEA/ADL dual-alkali system was
designed to meet seven process perform-
ance guarantees. The system's ability to
comply with these guarantees was closely
monitored during a 12-day acceptance
test and throughout the demonstration
period. Table 1 summarizes the parameters
and the corresponding system performance
during the acceptance test As shown, the
FGD system did meet all performance
guarantees during the acceptance test
except filter cake properties. Table 2 gives
results from the 13-month demonstration.
All performance guarantees were met
except insoluble solids and sodium con-
sumption, which were not met primarily
because of problems related to operation of
the vacuum filter. Filter performance de-
pends primarily on the thickness of the
filter cake and the amount of cake washing.
Manual control of cake thickness requires
simultaneous manipulation of slurry level,
drum speed, and thickener underflow solids.
It has been possible to obtain the proper
filter cake thickness and wash rate under
controlled tests and reduce sodium losses
to meet the guarantee, but under long-
-------
Table 1. Performance Guarantees and Acceptance Test Results
Guarantee
Test Results
S02 Emissions Limit
200 ppm (dry basis)
without additional air dilution
Calcium Consumption
1.05 moles available CaO
per mole SO2 removed
Soda Ash Consumption
0.045 moles Na2CO3 per mole
SO2 removed
Net Paniculate Addition
No net paniculate addition by
FGD system
Power Consumption
System will consume (excluding reheat)
not more than 1.2% of power generated at
gross peak boiler capacity (300 MW)
Filter Cake Properties
Filter cake will contain a minimum of
55 wt% insoluble solids
System Availability
System will have an availability of at
least 90% for 1 year'
141 ppm (dry basis) without additional air
dilution (93.9% SO2 removal)
1.04 moles available CaO per mole SO2
removed
0.042 moles Na2CO3 per mole
SO2 removed
Net paniculate removal efficiency
averaged 88%
System consumed 1.06% of power
generated at highest actual boiler operating
load (240 MW)
Filter cake averaged 52.2 wt% insoluble
solids
Not applicable to acceptance test
'Later modified to require an operability of 90% or greater, coupled with a daily average
removal efficiency of 90% or greater.
term operation, these filter conditions were
not held.
Monthly Performance
Parameters
Operation of the FGD system was moni-
tored monthly on the basis of several
performance parameters, defined in Table
3. The performance for the demonstration
period is shown in Table 4. The low value
for FGD system utilization was primarily
due to limited boiler availability.
Commercial Lime Test
Airco carbide lime, a byproduct of acety-
lene production, was normally used to
regenerate the sodium-based scrubbing
liquor at Cane Run. Commercial-grade
lime tests were conducted from March 2 7
to April 30,1981 ,to determine the effects
of a high quality, pure lime feed on FGD
system operability. The overall conclusion
from these tests was that the FGD system
performance with commercial lime did not
differ significantly from its performance
with carbide lime. One potential benefit
expected with commercial lime was higher
filter cake solids content with lower resid-
ual sodium because of enhanced cake
washing. However, there was no signifi-
cant improvement in the filter cake solids
content even under controlled filter condi-
tions. In addition, the proper cake thickness
(V4- toH-ia or 0.635 to 0.952 cm) could
not be maintained adequately by the oper-
ators because of the higher density solids
generated by commercial lime.
Capital and Operating Costs
Installation of the dual-alkali system at
Cane Run Unit 6 required capital invest-
ment in three sub-systems: FGD, lime
slurry feed, and waste processing and
disposal Each subsystem involved inde-
pendent design and installation efforts,
and the associated costs are shown in
Table 5. The costs shown were reported
on an as-incurred basis and thus do not
represent a constant dollar value of the
capital investment The actual cost of
$19,535,200 (actual costs incurred during
1977-1979 and reported in 1978 dollars)
compares with the original estimate for
Table 2. Summary of System Performance in Terms of Performance Guarantees'
SO2 SO2 CaO Na2CO3
Emissions1' Removal Consumption0 Consumption^
ppm %
May 1980
June 1980
July 1980
August 1980
September 1980
October 1980
November 1980
December 1980
January 1981
February 1981
March 1981
April 1981
May 1981
Average
Process
Guarantees
130.0
115.5
143.0
131.0
157.5
150.0
312.0
J
205.0
129.8
155.7
143.8
204.9
164.9
200.0
93.2
95.0
94.4
93.6
92.6
89.8
82.6
J
91.2
93.7
92.0
93.1
88.2
91.6
-
1.23
1.20
1.O4
j
1.15
1.04
0.89
J
1.04
1.06
1.04
0.93
0.95
1.05
1.05
0.079
0.085
0.042
->>
0.058
0.076
0.132
J
0.066
0.065
0.069
0.045
0.068
0.071
0.045
Power
Consumption
%
1.10
1.15
1.05
-
1.26
1.43
1.57
J
1.85
1.3
1.7
1.18
—
1.36
1.20
Insoluble
Solids
wt%
55.0
47.8
52.2
-f
49.7
46.8
36.9
J
49.1
5O.3
50.5
52.1
46.1
48.8
55.0
System
Availability
%
99.9
99.9
99.9
99.3
94.0
99.9
85.1
60.0
84.0
99.6
99.8
98.6
96.8
93.2"
90.0
Paniculate matter removal efficiency was determined during acceptance test only (88%).
All values shown (except May 1980) are 24-hour average stack SO2 emissions, corrected for moisture andreheater. May 1980 results represent
the average of the two absorbers' outlet emissions.
Total moles available CaO consumed per mole SO2 removed.
Moles Na2C03 consumed per mole SO2 removed.
No filter cake samples were analyzed in August 1980.
FGD system operated only 14.4 hours in December. No system characterization tests were performed.
Average availability based on hours in total period.
3
-------
Table 3. Definitions of Performance Parameters?
Period Hours (A)
Boiler Hours (B)
FGD Hours (C)
Available Hours (D)
Called Upon Hours (E)
Boiler Utilization
([BA]x 100)
FGD Utilization
ffCAJx 100)
FGD Availability
ffDAJx 100)
FGD Reliability
([CE]x100)
FGD Operability
ffCBJx 100)
Hours in the period under consideration
Hours during the period that the boiler was actually operated
Hours during the period that the FGD system (one or both modules) was
actually operated
Hours during the period that the FGD system was capable of being
operated
Hours during the period that the FGD system was demanded (or
wanted) for operation
Hours that the boiler was actually operated divided by the period hours
multiplied by 100
Hours that the FGD system was actually operated divided by the period
hours multiplied by 100
The hours that the FGD system was available for operation divided by
the period hours multiplied by 100
The hours that the FGD system was actually operated divided by the
called upon hours multiplied by 100
The hours that the FGD system was actually operated divided by the
boiler hours multiplied by 100
"Source: Edison Electric Institute.
Table 4. Summary of Monthly Performance Parameters for FGD System
Utilization Availability
Reliability
Operability
May 1980
June 1980
July 1980
August 1980
September 1980
October 1980
November 1980
December 1980
January 1981
February 1981
March 1981
April 1981
May 1981
99.5
88.9
83.1
84.5
86.9
99.9
55.1
1.9
80.2
71.6
71.9
82.6
96.8
99.9
99.9
99.9
99.3
94.0
99.9
85.1
60.0
84.0
99.6
99.8
98.6
96.8
99.9
99.9
99.9
94.6
98.8
99.9
96.8
6.0
83.3
99.5
99.7
98.3
96.8
99.9
99.7
99.8
94.6
98.6
99.9
96.8
6.0
83.3
99.5
99.7
98.3
96.8
Demonstration Period9
77.4
93.2
94.1
94.0
capital costs of $17,379,000 estimated
in 1978 (1977 dollars). The material
costs for the FGD system and field super-
vision were slightly above the original
estimate, while erection labor overhead
and the sludge disposal system were
nearly double. (However, the original esti-
mate for sludge disposal was based on a
different system that was not used.)
Actual annual operating costs for the
dual-alkali system were much higher than
estimated, both in total annual cost and
mills per kWh. The total annual operating
cost estimated in 1979 was $5.1 million
(with reheat), compared with the $7.8
million actually incurred (Table 6). The
major reasons for this difference were
higher labor, maintenance, and soda ash
requirements, as well as soda ash and fuel
oil unit costs that were about twice those
estimated. In addition, the average boiler
load was less than originally predicted.
Since some operating requirements (such
as labor and power) remain constant de-
spite boiler load, operation at lower boiler
load increases the cost per kWh for these
components. Reduction of maintenance
costs, together with improved sodium
recovery and an increased unit operating
load factor of 60%, could lowerthe operat-
ing cost from 6.9 to 5.0 mills kWh. This
figure would be comparable to that reported
for the dual-alkali unit at Southern Indiana
Gas and Electric Company (5.0 mills kWh),
but much less than the 13.9 mills kWh
reported for the Central Illinois Public
Service dual-alkali installation.
"Demonstration period averages were calculated as shown in Table 3.
Table 5. Actual Capital Cost Breakdown by Subsystem for Cane Run Unit 6 (1977-79 Costs in 1978 $)
Erection Costs
Engineering Costs3
Subsystem
FGD
Lime Slurry0
Waste Disposal
Total
Material
Costs1'
10, 183,200
460,600
2.084,200
1Z728.000
Direct
Labor
Z386.400
1 18,300
387,400
2,892,100
Field Supervision
and Engineering
277,900
9,100
3Z500
319,500
Construction
Overhead
1,373,000
1 12.000
398,700
1,883,700
System
Supplier
833,500
208,400
1.041.900
LGBE
Consultant
670,000
_//
_
-------
'able 6. Actual Average Annual Operating Costs (with Reheat) lor Cane Run Unit 6 Dual-Alkali
FGD System* (May 1. 1980 - April 30, 1981)
Quantity Unit Cost, $ Total Costs, $
44,647 ton
5,438 ton
1,204,279 gal.
12,925,000 kWh
79,170,000 gal.
117,890 ton
Direct Costs:
Carbide Lime
Soda Ash
Fuel Oil
Electricity
Water
Sludge (Trucking 8- Landfill)
Maintenance Materials
Labor
Operation 83,670 hrs
Maintenance 71,820 hrs
Analysis Z080 hrs
Supervision 2,080 hrs
Total Direct Costs
Indirect Costs:
Overhead (Payroll ft Plant)
Interest
Depreciation
Total Indirect Costs
Total Annual Operating Cost $
Actual Operating Cost (at 44% load factor)
Mills kWh
15.32/ton 683,992
134.53/ton 731,574
1.02/gal. 1,228,365
0.012/kWh 155,100
0.05/1,000 gal. 3,959
2.50/wetton 294,725
506,280
8.45/hr
8.55/hr
12.00/hr
various
52.9% of1,398, 033 (Total Labor)
6.125% of 19,535,200
4.65% of 19,535,200
$ ton of S removed
Adjusted Operating Cost (projected for 60% load factor)
Mills kWh
( l&Btu
-f ton of S removed
707,012
614,061
24,960
52,000
5,002,028
739,560
1,196,531
908,387
2,844,478
7.846,506
6.9
69.2
492.5
5.8
58.1
411.3
3asis: 300 MW (gross peak capacity) existing coal-fired plant-9.960 Btu kWh, 3.5% S coal,
91.6% SO2 removal, average SO2 removed-2 79 Ib/min. at 300 MW (gross peak) load, on-
site solids disposal by trucking after treatment stack gas reheat SOP.
R. P. VanNess is with Louisville Gas & Electric Co., Louisville, KY 40201; L. R.
Woodland and E. D. Gibson are with Arthur D. Little, Inc., Cambridge, MA 02140.
Norman Kaplan is the EPA Project Officer (see below).
The complete report, entitled "Full-Scale Dual Alkali FGD Demonstration at
Louisville Gas and Electric Company," (Order No. PB 83-241 364; Cost: $ 11.50,
subject to change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield. VA22161
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
-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Official Business
Penalty for Private Use $300
t°SSiH2PKUTECTlUN
CHICAGO IL
•er U.S. GOVERNMENT PRINTING OFFICE: 1983-759-102/077
------- |