United States Industrial Environmental Research EPA-600/7-79-167b
Environmental Protection Laboratory July 1979
Agency Research Triangle Park NC 27711
Proceedings: Symposium
on Flue Gas
Desulfurization -
Las Vegas, Nevada,
March 1979;
Volume II
Interagency
Energy/Environment
R&D Program Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the
effort funded under the 17-agency Federal Energy/Environment Research and
Development Program. These studies relate to EPA's mission to protect the public
health and welfare from adverse effects of pollutants associated with energy sys-
tems. The goal of the Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology. Investigations include analy-
ses of the transport of energy-related pollutants and their health and ecological
effects; assessments of, and development of, control technologies for energy
systems; and integrated assessments of a wide range of energy-related environ-
mental issues.
EPA REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved
for publication. Approval does not signify that the contents necessarily reflect
the views and policies of the Government, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/7-79-167b
July 1979
Proceedings: Symposium
on Flue Gas Desulfurization -
Las Vegas, Nevada, March 1979;
Volume II
Franklin A. Ayer, Compiler
Research Triangle Institute
PO. Box 12194
Research Triangle Park, North Carolina 27709
Contract No. 68-02-2612
Task Nos. 55 and 99
Program Element No. EHE624A
EPA Project Officer: Charles J. Chatlynne
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
ABSTRACT
This publication contains the text of all papers presented at EPA's
5th FGD Symposium held in Las Vegas, Nevada on March 5-8, 1979. Papers
cover such subjects as health effects of sulfur oxides, impact of FGD on
the economy and the energy problem, energy and economics of FGD pro-
cesses, actual operating experience, waste disposal and byproduct market-
ing, and industrial boiler applications.
ii
-------
VOLUME
Table of Contents
Page
Session 1: Energy and the Environment 1
Richard D. Stern, Chairman
Overview of Control Technology: The Bridge Between
Energy Utilization and Environmental Goals 2
Frank T. Princiotta and Clinton W. Hall
Remarks , .... 14
Leon Ring
Health Effects of SO2 and Sulfates 21
Bertram W. Carnow and Edward Bouchard
Energy, Environmental, and Economic Impacts of Flue Gas Desulfurization
Under Alternative New Source Performance Standards 48
Andrew J. Van Horn, Richard A. Chapman,
Peter M. Cukor, David B. Large
Session 2: Impact of Recent Legislation 87
Walter C. Barber, Chairman
Session 3: Economics and Options 88
Walter C. Barber, Chairman
Status of Development, Energy and Economic Aspects
of Alternative Technologies 89
P. S. Farber, C. D. Livengood,
K. E. Wilzbach, W. L. Buck, H. Huang
Economics and Energy Requirements
of Sulfur Oxides Control Processes 137
G. G. McGlamery, T. W. Tarkington,
S. V. Tomlinson
Combined Coal Cleaning and FGD 215
James D. Kilgroe
The Interagency Flue Gas Desulfurization
Evaluation Study 258
James C. Dickerman and Richard D. Stern
iii
-------
Session 4: Utility Applications 285
Michael A. Maxwell and Julian W. Jones, Co-chairmen
Status of Flue Gas Desulfurization in the United States 286
Bernard A. Laseke and Timothy W. Devitt
Recent Results from EPA's Lime/Limestone
Scrubbing Programs 342
H. N. Head, S. C. Wang, D. T. Rabb,
R. H. Borgwardt, J. E. Williams, M. A. Maxwell
TVA Compliance Programs for SO2 Emission 386
G. A. Hollinden and C. L. Massey
SO2 and NOx Removal Technology in Japan 418
Jumpei Ando
EPRI's FGD Program: From Problem Identification
to Development of Solutions 450
G. T. Preston
Cholla Station Unit 1 FGD System: 5 Years
of Operating Experience 469
Stephen R. Travis and Frank A. Heacock, Jr.
La Cygne Station Unit No. 1: Wet Scrubbing
Operating Experience 486
Terry J. Eaton
Dry FGD Systems for the Electric Utility Industry 508
Stephen J. Lutz and C. J. Chatlynne
Plan, Design and Operating Experience of FGD
for Coal Fired Boilers Owned by EPDC 526
Yasuyuki Nakabayashi
Current Alternatives for Flue Gas Desulfurization
(FGD) Waste Disposal—An Assessment 561
Chakra J. Santhanam, Richard R. Lunt, Charles B. Cooper
Marketing Alternatives for FGD Byproducts: An Update 595
W. E. O'Brien and W. L. Anders
Limestone FGD Operation at Martin Lake
Steam Electric Generator 613
Mark Richman
VOLUME II
Basin Electric's Involvement with Dry
Flue Gas Desulfurization 629
Kent E. Janssen and Robert L. Eriksen
iv
-------
Utility Conventional Combustion Comparative Environmental
Assessment—Coal and Oil 654
Charles A. Leavitt, C. Shih, Rocco Orsini,
Kenneth Arledge, Alexandra Saur, Warren D. Peters
Operating and Status Report: Wellman-Lord
S02 Removal/Allied Chemical SO2
Reduction—Flue Gas Desulfurization Systems
at Northern Indiana Public Service Company and
Public Service Company of New Mexico 702
D. W. Ross, James Petrie, F. W. Link
Citrate Process Demonstration Plant:
Construction and Testing 761
Richard S. Madenburg, Laird Crocker, John M. Cigan
Laurance L. Oden, R. Dean Delleney
Design and Commercial Operation of Lime/Limestone
FGD Sludge Stabilization Systems 792
Ronald J. Bacskai and Lee C. Cleveland
Power Plant Flue Gas Desulfurization Using
Alkaline Fly Ash from Western Coals 809
Harry M. Ness, Philip Richmond,
Glenn Eurick, Rick Kruger
Environmental Effects of FGD Disposal:
A Laboratory/Field Landfill Demonstration 835
N. C. Mohn, A. L. Plumley,
A. L. Tyler, R. P. Van Ness
Physical Properties of FGC Waste Deposits at the
Cane Run Plant of Louisville Gas and Electric Company 858
C. R. Ullrich, D. J. Hagerty, R. P Van Ness
Summary of Utility Dual Alkali Systems 888
Norman Kaplan
The FGD Reagent Dilemma:
Lime, Limestone, or Thiosorbic Lime 959
Donald H. Stowe, David S. Henzel, David C. Hoffman
Session 5: FGD Current Status and Future
Prospects—Vendor Perspectives 989
Frank T. Princiotta, Chairman
Session 6: Industrial Applications 990
Richard D. Stern, Chairman
The Status of Industrial Boiler FGD Applications
in the United States 991
John Tuttle, Avinash Patkar, R. Michael McAdams
v
-------
Environmental Assessment of the Dual Alkali FGD System
Applied to an Industrial Boiler Firing Coal and Oil 1023
Wm. H. Fischer, Wade A. Ponder,
Roman Zaharchuk
Operating History and Present Status of the
General Motors Double Alkali So2 Control System 1067
Thomas 0. Mason, Edward R. Bangel,
Edmund J. Piasecki, Robert J. Phillips
R-C/Bahco for Combined S02 and Particulate Control 1082
Nicholas J. Stevens
Status of the Project to Develop Standards of Performance
for Industrial Fossil-Fuel-Fired Boilers 1115
L. D. Broz, G. R. Often, D. D. Anderson,
J. D. Mobley, C. B. Sedman
Flue Gas Desulfurization Applications
to Industrial Boilers 1140
James C. Dickerman
Unpresented Papers 1160
Stack Gas Reheat—Energy and Environmental Aspects 1161
Charles A. Muela, William R. Menzies,
Theodore G. Brna
Minimizing Operating Costs of Lime/Limestone
FGD Systems 11 79
Carlton Johnson
By-Product-Utilization/Ultimate-Disposal of Gas Cleaning Wastes
from Coal-Fired Power Generation 1187
William Ellison and Edward Shapiro
Flue Gas Desulfurization and Fertilizer Manufacturing:
Pircon-Peck Process 1 204
R. B. Boyda
Dry FGD and Particulate Control Systems 1 222
K. A. Moore, R. D. Oldenkamp,
M. P- Schreyer, D. W. Belcher
Attendees 1 235
VI
-------
BASIN ELECTRIC'S INVOLVEMENT
WITH DRY FLUE GAS DESULFURIZATION
Kent E. Janssen
Manager of Production
and
Robert L. Eriksen
Environmental Control Supervisor
Basin Electric Power Cooperative
1717 East Interstate Avenue
Bismarck, ND 58501
ABSTRACT
This report will introduce a relatively new technique of flue gas
desulfurization using the dry scrubbing process. The dry scrubbing
system introduces a reagent slurry, via an atomizer into the flue gas
where the SO is absorbed by the reagent particles. The heat from the
flue gas dries the particles in the reaction chamber. The particles are
then collected in an electrostatic precipitator or a baghouse and
disposed of dry, eliminating the sludge dewatering process needed for
the wet scrubbing systems. Several reagents have been tested with soda
ash and lime yielding the most favorable results. Atomization can be
accomplished by either a rotary atomizer or spray nozzle. Baghouses and
precipitators are known equipment in utility plants and spray dryers are
operated in thousands of chemical and industrial plants around the
world.
Basin Electric Power Cooperative will use the rotary atomizer and
baghouse collection at the Antelope Valley Station Unit 1 (440 MW) near
Beulah, ND. The "Y" jet spray nozzle and precipitator collection will be
used on the third unit (500 MW) of the Laramie River Station near Wheatland,
WY. Reagents for both systems will be lime, due to its favorable economics
over soda ash. It is estimated that the dry scrubbing system will save
approximately $23 million at the Laramie River Station and $47 million
at the Antelope Valley Station over the life of the plants compared to a
"wet scrubbing" system for these site specific installations.
We believe the utility industry will eventually accept the concept
of dry scrubbing, not so much due to the economic factor, but because of
the system's simplicity and potential high availablity.
629
-------
BASIN ELECTRIC'S INVOLVEMENT
WITH DRY FLUE GAS DESULFURIZATION
INTRODUCTION
Numerous processes are available for removing SO^ from the flue gas
produced in fossil-fired power plants. Lime or limestone "wet scrubbing"
processes are the most common used today. The wet scrubbing systems
have generally been promoted by government agencies but have found few
proponents among the electric utility industry. The wet scrubbers are
complex and cumbersome. Added to this, wet scrubbers cause a heavy
capital burden, have a historic low availability record, are costly to
operate, difficult to maintain and consume a sizeable part of the energy
generated by the plants they serve. The byproduct, sludge, produces
additional environmental problems by taking up otherwise useful land for
settling ponds.
We question the rationale which dictates the current and proposed
air quality standards requiring high removal rates of SO , particularly
for the low sulfur coals. We question that sufficient research has been
done to determine the need for such strict standards and whether the
regulations will improve our air quality sufficiently to justify the
high cost of removal. Yet, in order to meet the standards in a manner
which is more operationally acceptable, we have worked with several
manufacturers to develop a process and hardware which is less complex
and, for our situation, more economical as well.
We believe the development of "dry scrubbers" during the past two
years may be a major improvement in the field of flue gas desulfurization.
HISTORY OF DRY FLUE GAS DESULFURIZATION
Basin Electric's involvement with dry scrubbing began with a pilot
baghouse at our Leland Olds Station near Stanton, North Dakota, in 1976.
The program was developed for the Coyote Project, a consoritum of five
utilities headed by Otter Tail Power Company. Unit 2 of the Leland Olds
Station was selected because it is a 440 MW cyclone fired boiler which
burns North Dakota lignite, therefore having conditions very similar to
the proposed Coyote Station. Bechtel Power Corporation, the architect-
engineer for Coyote, coordinated the program. The Wheelabrator-Frye,
Inc. (WFI) baghouse used a process of injecting dry powdered nahcolite
into the flue gas stream and onto fabric filter bags to remove SO . This
process worked quite well, but problems developed in the availability of
nahcolite. Major nahcolite reserves are in Colorado, but federal regulations
make mining of it very difficult. As a result, the long range outlook
for nahcolite was not good because of supply uncertainties. This lead
to the investigation of other processes. In the spring of 1977, prelim-
inary testing began on the open loop portion of the Rockwell Inter-
national, Atomics International Division Aqueous Carbonate Process,
utilizing a Bowen spray dryer with a rotary disc atomizer, followed by a
WFI baghouse.
630
-------
The first reagents used at the pilot plant were soda ash and trona.
These produced excellent SO removals and reagent utilization in the
dry scrubber. 2
PILOT PLANT TESTING
We, at Basin Electric, were interested in the pilot work as we were
involved in a large construction program with all units requring scrubbers-
the Antelope Valley Station (two 440 MW units) near Beulah, North Dakota
and the Laramie River Station (three 500 MW units) near Wheatland,
Wyoming; with plans for future power plants. Because of the potential
disposal problems and the high cost of sodium reagents, we decided to
investigate the use of other alkaline reagents.
Atomics International/Wheelabrator-Frye
Under agreements with Atomics International, several other reagents
were used with varying degrees of success. Reagents tested were potash,
fly ash from several sources, limestone, slaked lime, hydrated lime,
dolomitic lime, and ammonia. Initial test results indicated that slaked
lime had promise as the most economical reagent.
In addition to the pilot testing done by Atomics International/
Wheelabrator Frye — other companies, primarily Joy Manufacturing
Co./Niro Atomizer, Carborundum and Babcock & Wilcox pursued dry scrubber
piloting.
Joy/Niro
The Joy Manufacturing Company and Niro Atomizer invested several
million dollars to pilot a system in order to obtain data, optimize the
process, and determine the economic feasiblity of the dry scrubbing
system. Niro had good experience in their Copenhagen test labs and in
their pilot plants in Europe using both lime and sodium as reagents.
From this, Niro made an appraisal and decision to proceed with a large
scale pilot plant.
The Joy/Niro pilot was located at the Hoot Lake Station Unit 2
owned by Otter Tail Power Company at Fergus Falls, MN. Joy/Niro developed
a recirculating system in which lime slurry is mixed with recirculated
fly ash and spent reagent, for reinjection into the reactor, (patent
pending).
Babcock & Wilcox
Babcock & Wilcox installed a semi-wet reactor designed on the basis
of Japan's Hitachi process, followed by parallel streams to a precipitator
and a baghouse at our William J. Neal Station at Velva, ND. Babcock &
Wilcox later redesigned the system to a horizontal reactor using a "Y"
jet dual fluid atomizer followed directly by an electrostatic precipitator.
The flue gas and the slurry spray entered the reactor at the front wall,
the geometry being very similar to that of a circular burner.
631
-------
Carborundum
The Carborundum Company, using a De Laval spray dryer in combina-
tion with their baghouse, had a pilot test program on Unit 1 of our
Leland Olds Station. They tested both dual fluid spray nozzles and
rotary atomizers.
Test Parameters
All four pilot plants conducted tests which were based on the
design conditions for our Antelope Valley Station. These parameters
ranged from 400-2000 ppm SO at the inlet, at approximately 310 F inlet
flue gas temperatures, while burning North Dakota lignite.
Wyoming (Powder River Basin) sub-bituminous coal was later burned
in the boilers used for the Joy/Niro and Babcock & Wilcox pilots. These
test conditions were typical of the flue gas characteristics which will
be encountered at the Laramie River Station, i.e. 400-800 ppm S02.
Reagents tested at the four pilot plants included soda ash, fly
ash, lime, limestone, magnesium oxide and ammonia.
All pilots conducted one hundred hour endurance tests which demonstrated
that the emission requirements were fulfilled with a comfortable safety
margin.
With each week of testing, more and more was learned about the
process. This included such parameters as lime slaking variables, feed
slurry conditions, slurry feed temperatures, spray down temperatures,
outlet gas temperatures, outlet humidity, atomizer design, atomizer
speeds, spent reagent/ash recirculation and baghouse/precipitor SO-
removal. Stoichiometric ratios improved to yield nearly 100% utilization
of the lime feed.
DRY VERSUS WET SCRUBBERS
Although the process of dry scrubbing is new, the equipment is not.
Spray drying technology is used extensively in mining applications and
the food industry. Both fabric filter baghouses and electrostatic
precipitators have been used for quite some time.
Some of the reasons for Basin Electric's enthusiasm for the dry
scrubbers are as follows:
Waste Handling. The dry system has no sludge handling equipment,
which is usually troublesome and has a record of high maintenance. Wet
scrubbers require thickeners, centrifuges or vacuum filters and sludge-
flyash blenders in order to obtain a dry product. The product from a
dry scrubber can be handled with conventional dry handling systems used
for flyash. For the coals tested, the dry product appears to handle as
well as the flyash.
632
-------
Wet/Dry Interface. Scaling and plugging is common in wet scrubbers
at wet/dry interfaces and on scrubber packing materials and demisters.
In the dry system, the interface point occurs in suspension; only dry
powder makes contact with the walls. There are no packed beds or demisters
in a dry scrubber.
Materials of Construction. Wet scrubbers require expensive alloy
materials or coatings for protection from corrosion and erosion. The
dry system can use low carbon steel for vessels and ductwork. The I.D.
fans can be safely located just ahead of the stack without fear of fan
corrosion and imbalance.
Operations. It is estimated that considerably fewer operations
personnel will be required for the dry system. Wet scrubbers have
proven to take considerable manpower for operations and maintenance.
The dry scrubber offers flexiblity of operation. Feed rates can be
immediately adjusted with little concern for pH control. Turndown
capability for a dry scrubber is in the order of a 10:1 ratio. Wet
scrubber modules usually must be left in service at low loads to recir-
culate slurry. In the dry system, modules and/or atomizers can be
removed from service quickly and easily as the load varies.
Maintenance. Wet systems have inherent high maintenance costs with
slurry handling equipment recirculating abrasive materials at high
pressures and volumes. The dry system operates with low pressures and
low material volumes. Liquid to gas ratios are about 0.2 to 0.3 gallons
per 1000 actual cubic feet compared to about 40-100 for a wet scrubber.
The atomizer, which is probably the highest wear item in the system, can
be removed and replaced quickly. The elimination of dewatering equipment
reduces the maintenance expense considerably from that for a wet scrubber.
Energy Requirements. The dry system requires approximately 25% to
50% of the energy required for a wet system.
Particulate Collection. The gas volume to the particulate collector
is reduced to below that leaving the air heater as a result of the spray
down in the dry scrubber. The gas temperature to the particulate
collection device remains constant at all loads. The spray dryer conditions
the flyash with the added moisture resulting in a lower resistivity ash
in the precipitator. Although the scrubber reactants produce additional
particulates, the increased humidity and lower temperatures have a
positive effect on preciptitator operation. The pilot test results
indicated that the increased humidity did not adversely affect baghouse
operations, and the baghouse benefits from the reduced gas volume.
Water Consumption. The water requirements for a dry system are
much less than for a wet system. The dry scrubber at the Laramie River
Station Unit 3 will use about 50% of the amount required for the wet
scrubber on Unit 1 or Unit 2. Low quality water such as cooling tower
blowdown or ash water may be used in the spray dryer. Only a small
quanitity of treated water, about 20% of the total requirement, is
needed for lime slaking.
633
-------
ANTELOPE VALLEY STATION DRY SCRUBBER
After analyzing and evaluating the data from the four pilot plants
with the help of our architectural engineer, Stearns-Roger, we took bids
for a dry system utilizing lime for our Antelope Valley Station Unit 1.
After receiving bids from four vendors, a contract for the Antelope
Valley Station Unit 1 SO- and particulate removal system was awarded to
Joy Manufacturing Company, Western Precipitation Division with Niro
Atomizer Company as prime subcontractor. This plant is scheduled to
begin commercial operation in April, 1982.
The system will have five vertical Niro spray dryers (four in
operation, one as a spare) followed by a Western Precipitation baghouse
(See Figure 1). Each spray dryer will utilize one rotary atomizer with a
direct drive motor (See Figure 2). Quick lime will be slaked in a ball
mill slaker. Lime slurry, sludge from the primary water treatment
process, and a portion of the recycled ash product will be mixed to
create the feed slurry which goes to the atomizers (See Figure 3). A
dispersed mist of fine droplets is sprayed into the flue gas as it
enters the spray dryer at the top and center. A very rapid chemical
reaction occurs removing a substantial amount of SO- from the flue gas.
At the same time, the thermal energy of the flue gas evaporates the
water in the droplets to form a dry powder consisting of calcium sulfite,
calcium sulfate, unreacted lime, and flyash.
Some of the product is removed in the conical bottom of the dryer.
The remainder is suspended in the flue gas which goes to the baghouse
(See Figure 4) . Additional SO,, is removed by the powder on the fabric
filter bags, thus increasing the system efficiency.
The baghouse has 28 compartments having a total of approximately
8,000 fiberglass filter bags with fluoro-carbon coating. The bags are
12 inches in diameter and 35 feet high. Reverse air cleaning will be
used. The gas to cloth ratio will be 2.19 cubic feet of gas per minute
per square foot of cloth (gross) under maximum operating conditions.
The cleaned flue gas exits the baghouse to the I.D. fans and is
exhausted to the stack.
The SO- and Particulate Removal System is designed to operate at
62% SO- removal for performance lignite (average sulfur) and 78% S0_
removal for design lignite (maximum sulfur) in order to meet the emission
limitation of 0.78 Ib. SO- per million Btu as required by the North
Dakota State Department of Health. The coal analysis, is given in Table
1, and the operating conditions and flue gas characteristics are given
in Table 2.
The lime feed rate is essentially stoichiometric. A low stoichiometric
ratio is possible because of the utilization of available alkalinity in
the flyash.
634
-------
Antelope Valley Station
Gas Cleaning System
m
en
so
Figure 1
-------
NIRO Atomizer
Figure 2
636
-------
Antelope Valley Station Flow Diagram
Boiler Flue Gas
Partial Recycle
Figure 3
Clean
Exhaust
From
Stack
Dust Collector
Spray
Absorber
Reagent
Feed
Powder & Fly Ash
SO? Absorbent
Disposal
< ~~
CO
vO
-------
Antelope Valley Station Baghouse
Figure 4
638
-------
Table 1. ANTELOPE VALLEY STATION
FUEL CHARACTERISTICS
ULTIMATE ANALYSIS
Average
Range
UO
Carbon
Hydrogen
Nitrogen
Sulfur
Moisture
Oxygen
Chlorine
Ash
HHV, Btu/lb
39
2
0
0
37
11
0
7
.69
.70
.60
.68
.07
.49
.01
.76
6600
36.
2.
0.
0.
30.
10.
0.
5.
55
38
45
36
03
15
00
24
6093
- 41.92
- 2.96
- 0.81
- 1.22
- 42.43
- 12.86
- 0.02
- 13.83
- 7350
-------
Table 2. ANTELOPE VALLEY STATION
OPERATING CONDITIONS
&
FLUE GAS CHARACTERISTICS
OPERATING CONDITIONS
Generation Net/Gross, MW
Fuel Input
Coal Burn Rate
FLUE GAS CHARACTERISTICS
Gas Flow, acfm
scfm
o
Gas Temperature F
Mass Gas Flow, Ibs/hr
Moisture, % Volume
SO- ppm by Volume
Ibs/hr
Ibs/MKB
Particulate Ibs/hr
gr/acf
gr/scf
385/440
4930xl06 Btu/hr
375 tons/hr
INLET
2,055,000
1,248,900
310
5,690,000
15.6
800
10,120
2.07
88,070
5.0
7.7
OUTLET
1,894,380
1,369,000
185
6,097,000
19.4
304
3,845
0.78
210
0.012
0.018
Based on anticipated average fuel analysis and 100% plant load.
-------
LARAMIE RIVER STATION DRY SCRUBBER
We had Burns & McDonnell prepare specifications and take bids for a
dry system for the Laramie River Station Unit 3. These bids were evaluated
and the economics compared to that of an option available for a precipitator
and a wet scrubber identical to Units 1 & 2. A contract for the Gas
Cleaning System for Unit 3 of the Laramie River Station was awarded to
Babcock & Wilcox. Unit 3 is scheduled to begin commercial operation in
April, 1982. This scrubber will be a variation of the concept used at
the Antelope Valley Station. The system consists of four reactors
(three in operation, one as a spare) each followed by an electrostatic
precipitator (See Figure 5).
Each Babcock and Wilcox reactor is equipped with 12 "Y" jet nozzles.
The dual fluid atomizers will use a concentric pipe feeder for the
atomizing fluid, which is steam, and the reagent, which is lime slurry.
(See Figure 6). The atomizer is horizontally fitted into the center of
a circular throat on the front wall of the reactor (See Figure 7). The
throat is fitted with vanes which control the shape of the gas envelope.
The angle of the atomizer spray cone is matched to this envelope to
obtain the maximum mixing of the atomizer slurry and the flue gas. This
principle has been used in the mixing of fuel and air since the 1930's.
Under each reaction chamber are three hoppers to collect a portion
of the product which will drop out prior to the precipitator (See Figure
8).
A portion of flue gas (approximately 3%) bypasses the air heater
and reactor and enters the reactor discharge plenum where it is mixed
with the reactor discharge for reheat. The reheated flue gas enters the
B & W Rothemule electrostatic precipitator, which is the same design and
size as that on Unit 1 and Unit 2.
The gas cleaning system is designed to operate at 85% SO removal
with a 0.54% sulfur coal. The system is also designed for 90% S0_
removal with a 0.81% sulfur coal. The coal analysis is given in Table
3, and the operating conditions and flue gas characteristics are given
in Table 4.
WASTE PRODUCT DISPOSAL
The waste product material has an appearance of being totally dry.
There is a relatively small amount of moisture present, most of which is
chemically bound water of hydration. The waste product is a very fine
grain, powdery material, very similar in particulate size distribution
as the flyash normally removed from the flue gas stream of a coal-fired
boiler. A summary of physical characteristics from a waste product
sample, selected from a pilot plant burning North Dakota lignite and
using lime as the reagent, is given in Table 5. Of particular significance
is the relatively impermeable nature of the material.
The amount of water soluble material found in the spray dryer
product is dependent on the reagent material—the relative solubility of
641
-------
Basin Electric Power Cooperative
Laramie River Station - Unit No. 3
Start
Figure 5
-------
OUTfR
rust
INNER
TUBf
SPRAYE R
P1ATE
Y-Jet Slurry Atomizer
Figure 6
-------
Distance Piece
I
7
Slurry/Atomizing Air
Gas Flow
Gas Flow
B&W Atomizer & Throat Configuration
/els i P ot
Figure 7
644
-------
Laramie River Station Flow Diagram
J--
\J\
ToDi«po»«
Mill Product
Tank
Figure 8
-------
Table 3. LARAMIE RIVER STATION
FUEL CHARACTERISTICS
ULTIMATE ANALYSIS
Average Range
Of "/
%, /o
Carbon 49.07 45.32 - 51.32
Hydrogen 3.45 3.03-3.91
Nitrogen 0.92 0.4 - 1.0
5 Sulfur 0.54 0.15 - 0.54
ON
Moisture 28.92 22.2-34.65
Oxygen 9.17 8.51 - 12.07
Chlorine 0.04 0.00-0.04
Ash 7.89 4.0 - 13.0
HHV, Btu/lb 8139 7906 -- 8244
-------
Table 4. LARAMIE RIVER STATION
OPERATING CONDITIONS
&
FLUE GAS CHARACTERISTICS
OPERATING CONDITIONS
Generation Net/Gross, MW
Fuel Input
Coal Burn Rate
500/575
5513.625x10 Btu/hr
340 tons/hr
FLUE GAS CHARACTERISTICS
Gas flow, acfm
scfm
Gas Temperature F
Mass Gas Flow, Ibs/hr
Moisture % Volume
S0_ ppm by volume
Ibs/hr
Ibs/MKB
Particulate Ibs/hr
gr/acf
gr/scf
INLET
2,300,000
1,320,000
286
6,188,000
11
530
7320
1.33
78,860
4.0
7.0
OUTLET
1,946,000
1,405,000
157
6,461,000
17
80
1100
0.2
300
0.015
0.02
Based on anticipated average fuel analysis and 100% plant load.
-------
Table 5. PHYSICAL CHARACTERISTICS OF WASTE PRODUCT
Specific Gravity 2.8 - 2.9
Maximum Compacted Dry Density 62 pcf
Optimum Moisture Content 55%
Permeability 1x10 to 1x10 cm/sec
Atterberg Limits
Liquid Limit (%) 63
Plastic Index (%) 12
648
-------
calcium sulfite and calcium sulfate as compared to sodium sulfite and
sodium sulfate. In this respect the solubility of a sodium based
product has been found to be 50-60% soluble, whereas the solubility of a
calcium based product has been measured at 3-7% solubility. The ratio of
sulfite to sulfate in the waste product is quite variable. In sodium
based systems, sulfate is by far the most predominate specie. In lime
based systems, the ratio is more balanced. (1:1 to 2:1 favoring sulfite).
In addition to sulfite and sulfate as spray dryer reactants, there is
also some unutilized reagent present in the product, and also there is a
small amount of carbon dioxide absorption from the flue gas which forms
calcium carbonate in the product. Table 6 gives the expected chemical
characteristic of the waste product from the Antelope Valley Station.
The method of waste disposal planned for the Antelope Valley Station
is to transport the waste product to depleted areas of the mine for
landfill. The method of waste disposal planned for Laramie River
Station is to transport the material to a landfill where it will be
disposed of along with conventional limestone scrubber sludge.
The material is fairly cementitious and impermeable, however,
potential problems in the field might be weathering, erosion, dust
suppresion and structural stability. Disposal procedures will be determined
during field tests when the product first becomes available.
It is germane to note that there are no apparent disadvantages in
disposing of a calcium based scrubber waste product compared to that of
flyash. A waste product from a sodium based system may require special
handling, however, because of the increased solubility of the sodium
sulfite and sodium sulfate.
We believe there may be a potential market for the sale of the
waste product.
ECONOMICS
The economics of any FGD or particulate control technology for
full-scale utility boilers are unquestionably site-specific. Economics
are subject to wide variations depending on the type of fuel burned,
applicable environmental standards and geographical location of the
power plant. However, as a result of extensive engineering studies, the
operational experience with full-scale spray dryers, baghouses and
precipitators, and the many pilots tests performed to date; the general
economics can be derived with a relatively high degree of confidence.
The economic analysis of wet versus dry systems for both the Antelope
Valley Station and the Laramie River Station gave the dry system the
lowest evaluated cost over the lifetime of the plants.
Evaluated costs for the dry scrubber system at the Antelope Valley
Station produced approximately $47 million cost savings over a wet
scrubber system for a plant life of 35 years based on present worth and
a 75% annual plant factor (See Table 7). This is attributed to a savings
in operation and maintenance expenses as evaluated by Basin Electric.
649
-------
Table 6. EXPECTED WASTE PRODUCT CHEMICAL CHARACTERISTICS
FROM THE ANTELOPE VALLEY STATION
SO,, Reactants
Fly Ash
CaSO.
4
CaS03
CaCO
Lime Inert s
H00
2
Total
8.9
14.6
2.1
1.6
1.3
28.5
SiO.
2
P2°3
Fe203
A1000
2 3
T100
2
CaO
MgO
Na2°
K20
so3
Undetermined
Total
21.1
0.4
6.8
8.5
0.5
18.1
5.8
4.7
0.5
3.4
1.7
71.5
650
-------
Ui
ECONOMIC FACTORS
Capital Cost
Reagent Cost
Power Cost
Manpower Cost
Replacement
Materials
Table 7. ANTELOPE VALLEY STATION
WET VS. DRY SCRUBBER
ECONOMIC EVALUATION
DRY SCRUBBER
WET SCRUBBER
35 Year
Total Cost
$1981
$49,665,100
$38,587,500
$ 7,466,700
$15,925,000
Basis of
Evaluation
actual
18,000 tons/yr
lime
5726 KW
6 operators
35 Year
Total Cost
$1981
$55,927,400
$42,505,000
$13,040,000
$36,750,000
Basis of
Evaluation
estimate
41,848 tons/yr
limestone
10,000 KW
11 operators
$17,500,000
(est.)
7 maintenance
(est.)
filter bags,
atomizer wheel,
bearings, etc.
(est.)
$28,000,000
(est.)
19 maintenance
(est.)
pH control, sludge
fixation chemicals,
pump misc., ESP
bushings, & TR's
(est,)
Total
$129,144,300
$176,222,400
Evaluation based on 35 year life, annual plant factor of 75% and present worth of operation and maintenance
cost as of the unit's commercial date (1982).
-------
There are intangibles with both the wet system and the dry system
that are difficult to place a dollar value on. If a wet system had been
selected for the Laramie River Station Unit 3, such intangibles as
reduced spare parts and maintenance know-how of similar systems would
have been realized. Yet, in the dry system, there are potential im-
provements that may be obtained in stoichiometry as refinements are made
to the atomization, mixing techniques, and the slaking process. A substantial
improvement can be made in the economics if lime becomes locally available.
Limestone is locally available, but the closest source of lime is approximately
200 miles.
Even though these intangibles are not evaluated, the total evaluated
cost of the wet system is still higher than the dry system. This is
true even after considering the cost of transporting lime to the site
versus using locally available limestone. At the Laramie River Station
Unit 3, the evaluated costs of the dry system are approximately $23
million less than the wet system based on a plant life of 35 years and a
75% annual plant factor (See Table 8).
CONCLUSION
In the long run, availability will probably be the most important
advantage of the dry system. Attempts will not be made here to develop
numbers for this, since it will be three to four years before operating
data will be available. However, it should be apparent that the dry
system; when compared to wet scrubbers, has less equipment which can
cause trouble. The dry system has fewer pieces of complex equipment, is
simpler to control, and is not particularly sensitive to changes in
operating conditions.
We believe the utility industry will eventually accept the concept
of dry scrubbing, not so much due to the economic factor, but because of
the system's simplicity and potential availability. It is only fair to
say that the figures quoted comparing the wet and dry systems are site
specific for our applications. Changes which would tend to favor' the
economics of the wet system would be an increase in the differential
between the cost of lime and limestone and an increase in the coal
sulfur content. Although it is beyond the scope of this presentation,
results from continuing developmental work are increasingly favorable,
appearing to make the dry system competitive on higher sulfur coals.
Changes which would favor the dry system would be: (a) improvement in
system stoichiometry, (b) improvement in availability of the dry scrubber
compared to the wet scrubber, (c) future escalation in material and
labor costs, (d) higher fixed charges and (e) increased energy costs.
All of these appear to be most likely probabilities.
While the dry scrubber is new, we believe that it will be the
scrubber of the future, particularly for low sulfur western coals.
652
-------
Table 8. LARAMIE RIVER STATION
WET VS. DRY SCRUBBER
ECONOMIC EVALUATION
ECONOMIC FACTORS
DRY SCRUBBER
WET SCRUBBER
as
Ui
Capital Cost
Reagent Cost
Power Cost
Manpower Cost
Replacement
Pressure Loss
Total
35 Year
Total Cost
$1981
$49,807,000
$48,880,000
$ 4,806,000
$15,925,000
Basis of
Evaluation
actual cost +
ancillaries
20,920 tons/yr
lime
2451 KW
6 operators
35 Year
Total Cost
$1981
$60,632,100
$ 9,367,000
$19,610,000
$36,750,000
Basis of
Evaluation
estimate
35,400 tons/yr
limestone
10,000 KW
11 operators
$15,750,000
$ 4,623,800
$139,791,800
(est.)
7 maintenance
(est.)
spray nozzles,
pump misc., ESP
bushings & TR's
(est.)
6.5" w.g., air
heater outlet to
chimney inlet
$28,000,000
$ 8,678,600
(est.)
19 maintenance
(est.)
pH control, sludge
fixation chemicals,
pump misc., ESP
bushings, & TR's
(est.)
12.2" w.g., air
heater outlet to
chimney inlet
Evaluation based on 35 year life, annual plant factor of 75% and
maintenance cost as of the unit's commercial date (1982).
$163,037,700
present worth of operation and
-------
UTILITY CONVENTIONAL COMBUSTION COMPARATIVE
ENVIRONMENTAL ASSESSMENT - COAL AND OIL
Charles A. Leavitt, C. Shih, Rocco Orsini, Kenneth Arledge, and Alexandra Saur
TRW, Inc., Redondo Beach, Cal.
Warren D. Peters
U.S. Environmental Protection Agency
Industrial Environmental Research Laboratory
Research Triangle Park, N.C.
ABSTRACT
This paper summarizes the results of a comparative multimedia assessment of
two utility boilers, one firing coal, the other oil, to determine relative
environmental impacts. Comprehensive sampling and analyses of multimedia
emissions from the boilers and control equipment were conducted to identify
criteria pollutants and other species. The results indicate that:
(1) particle emissions from coal firing are 10 times those from oil firing
(scrubbing reduced the mean size of the particles from coal firing and
increased the emissions of particles below 3 ym size); (2) S02 emissions
from coal firing are 7 times those from oil fixring; (3) NO emissions
from coal firing are 6 times those from oil firing; (4) for" coal firing,
arsenic, cadmium, chromium, iron, nickel, lead, and zinc emissions are of
potential environmental concern -- for oil firing, only chromium and nickel
are of potential concern; (5) organic and polycyclic organic emissions
from either fuel are not of environmental concern; (6) liquid emissions
from coal firing contain many inorganic elements that are of environmental
concern although the organic content is innocuous; and (7) solid emissions
from coal firing also contain many trace elements at concentrations of
concern -- there are no significant solid wastes from oil firing.
654
-------
INTRODUCTION
Objective
Conventional methods of converting fossil fuels to usable forms of energy
have impacts on the air, land, and water; i.e., multimedia impacts,
These impacts are not separate and distinct; rather, they are all inter-
related and involve delicate balances and trade-offs.
The Environmental Protection Agency (EPA), with primary responsibility for
controlling adverse environmental impacts of pollutant emissions, has been
active since its inception in determining the identities and quantities of
•potential pollutants released to the environment when fossil fuels are
burned. Information from EPA R&D efforts is being used for three principal
purposes: to assess the health and environmental impacts caused by the
release of combustion pollutants to the environment; to define the needs
for technology to control the release of these pollutants; and to develop
standards to limit emissions.
CCEA Program
In response to the need for a comprehensive environmental assessment of
conventional combustion systems, EPA's Industrial Environmental Research
Laboratory at Research Triangle Park (EPA/IERL - RTP), North Carolina, has
established a unified Conventional Combustion Environmental Assessment (CCEA)
program.1 It is a major new program aimed at the comprehensive assessment
of environmental, economic, and energy impacts of multimedia pollutant
emissions from stationary industrial, utility, residential, and commercial
conventional combustion processes. The primary objective of the CCEA program
is to identify and evaluate information from all relevant sources in order to:
determine the extent to which this information can be utilized to assess the total
environmental, economic, and energy impacts of conventional combustion processes;
identify and acquire additional information needed for such assessment; define
the requirements for modifications or additional development of control technology;
and define the requirements for modified or new standards to regulate pollutant
emissions.
The CCEA program will coordinate and integrate ongoing and future studies
into a unified environmental assessment structure and serve as a centralized
base of information on the environmental impacts of conventional combustion
processes. Coordination and information exchange between CCEA-related studies
should minimize duplication and maximize the return from available resources.
The environmental assessment (EA) methodology employed in the CCEA program
draws heavily on the philosophies of the existing EPA/IERL-RTP EA methodology,
but has been expanded and modified to be more responsive to the assessment
of conventional combustion processes.
655
-------
In the most elementary description, an environmental assessment consists
of three basic iterative steps (Figure 1):
1. Characterization of the combustion process (including any associated
pollution control devices) and its effluents.
2. Assessment of the health and ecological impacts of the combustion
process and its effluents on the environment.
3. Evaluation of alternative control strategies to reduce pollution impacts
to acceptable levels.
The EA procedure used in the CCEA program is shown in the generalized methodology
diagram (Figure 2).
It is the goal of the CCEA program to integrate ongoing projects and recommend
new efforts to address all practical combinations of information. It is
expected that EPA/IERL-RTP, with the assistance of contractors with
experience and expertise in the various areas associated with the comprehensive
environmental assessment of conventional combustion processes, will
implement and expand the CCEA program as needs dictate and as resources permit.
Multimedia Assessment Method
A major goal of the CCEA program is to support the implementation of the
National Energy Plan, which aims at increasing the use of coal. Since fuel
switching from oil to coal is an important facet of the NEP, the CCEA program
initiated a study to compare the environmental impacts of oil and coal
combustion.
The objectives of this study were to conduct multimedia environmental assessments
of oil and coal firing in a controlled utility boiler in order to compare
environmental, energy, and societal impacts of firing coal vs. firing oil.
In order to conduct the comparative assessment, it was necessary to fully
characterize feed streams, emissions, and effluents from the utility
boilers selected for study and all associated pollution control equipment.
Plant Description and Fuel Analyses
Two electrical utilities were chosen for this study, with their agreement
and cooperation. A midwestern coal fired utility with flue gas desulfurization
was one selection. A west coast oil fired utility with staged combustion and
flue gas recirculation was the other selection.
656
-------
CHARACTERIZATION OF COMBUSTION
PROCESS AND ITS EFFLUENTS
ASSESSMENT OF HEALTH AND
ECOLOGICAL IMPACTS
• Identification of environmental
impacts
t Development of environmental
goals and objectives
• Comparison of impacts with
goals and objectives
• Quantification of pollution
impacts
EVALUATION OF ALTERNATIVE
CONTROL STRATEGIES
Ui
ENVIRONMENTAL ASSESSMENT
AND OUTPUTS
• Technology transfer documents
• Standards development
recommendations
t Control technology development
recommendations
FIGURE 1
GENERALIZED ENVIRONMENTAL ASSESSMENT METHODOLOGY
-------
COMBUSTION PROCESSES
AND EFFLUENTS
CHARACTERIZATION
HEALTH AND
ECOLOGICAL IMPACTS
IDENTIFICATION
ENVIRONMENTAL
GOALS AND OBJECTIVES
DEVELOPMENT
REDEFINE
DATA BASE
00
CONTROL TECHNOLOGY
DEVELOPMENT
RECOMMENDATIONS
ARE
GOALS
MET?
ALTERNATIVE CONTROL
STRATEGY
EVALUATION
POLLUTANTS
IMPACTS
QUANTIFICATION
STANDARDS
DEVELOPMENT
RECOMMENDATIONS
FIGURE 2
GENERAL METHODOLOGY
-------
The coal fired utility boiler tested burns local coal and is equipped with an
emission control system composed of eight two-stage venturi/absorber scrubber
modules using a slurry of local limestone, see Figure 3. The boiler is a cyclone
fired, supercritical, once through, balanced draft B&W unit rated at
2.81 x 100 kg steam per hour at 538°C, 26.13 MPa (6.2 x 106 Ib/hr at 1000°F,
325 psig). The coal is a low grade sub-bituminous class with a typical as-fired
heating value of 20.9 to 22.6 MJ/kg (9000 to 9700 Btu/lb), with an ash content
of 25 percent and containing 5 to 6 percent sulfur, see Tables 1 and Z.
TABLE 1.
COAL FIRED BOILER DESIGN DATA
Type Crushed coal, cyclone fired
Manufacturer Babcock and Wilcox
Type of Burner Cyclone
Number of Burners 18
Air Preheaters Yes
Fuel Sub-bituminous from Pittsburgh and
Midway Coal Mining Company
Design Steam Rate 2.81 x 106 kg/hr (6.2 x 106 Ib/hr)
26.13AMPa (325 psig), at 538°C
(1000°F)
The oil fired utility boiler burns low sulfur fuel oil and utilizes sub-
stoichiometric firing with flue gas recirculation to control NO emissions.
No other air quality control equipment is installed, see Figure 4. Boiler
No. 5 was tested. It was designed to burn either natural gas or oil. Oil
analyses are given in Table 2. It is a double reheat, supercritical pressure,
once through, front and rear fired B&W Universal Pressure boiler. See Table 3.
659
-------
AIR
HEATER
RECIRCULATION TANK
TO ASH DISPOSAL
S)= SAMPLING POINT
TO SETTLING POND
AND ONTO COOLING
LAKE
FIGURE 3
COAL FIRED BOILER SCHEMATIC
-------
AIR
HEATER (2)
FORCED
FUEL OIL
BOILER
UNIT
NO. 5
FLUE GAS
STEAM
RECIRC.
FAN (2)
TURBINE
BOILER
FEEDWATER
CONDENSATE
POLISHING
CONDENSATE
GENERATOR
AIR PREHEATER &
BOILER WASH
WASTE
WATER
STORAGE
WATER
TREATMENT
SAMPLING POINT
RAW WATER
STORAGE
* CITY WATER
TO RIVER
FIGURE 4
OIL FIRED BOILER SCHEMATIC
-------
TABLE 2.
SUMMARY OF ULTIMATE FUEL ANALYSES
Component
Moisture
Carbon
Hydrogen
Nitrogen
Chlorine
Sulfur
Ash
Oxygen
Coala
Weight %
1.34
57.28
3.84
1.01
0.04
5.45
26.56
4.48
c
a
0.07
1.28
0.09
0.12
0.01
0.64
1.03
0.66
011b
Weight %
~ 0
86.54
12.39
0.39
Not Analyzed
0.81
0.008
0.658
c
a
—
0.31
0.04
0.33
--
0.02
0.005
0.41
Heating Value
(kj/kg)
24,027
384
44,025
a Coal analysis based on average of 5 tests.
b Oil analysis based on average of 4 tests.
c a = standard deviation.
103
Control Devices
The coal fired emission control system was designed by B&W as an integral part of
the steam generation plant. It was designed to treat the boiler flue gas flow
of 78,000 acmm (2,760,000 acfm) (9,800 acmm or 345,000 acfm per module at
140°C or 285°F). The ductwork design does not provide for flue gas bypass of
the system. Also, the plant does not have an alternate or secondary fuel
supply. Each module can be isolated for maintenance by individual dampers.
On site limestone grinding and slurry storage facilities provide up to 907 Mg
(1,000 tons) of slurry per hour. The unit has a balanced draft system with three
7,000 hp forced draft fans and six 7,000 hp induced draft fans located between
the emission control system and the 213 m (700 foot) stack. There is a common
plenum at both the scrubber inlet and outlet. Spent slurry and fly ash are
removed from the module recirculation tank through rubber lined pipes to a
settling pond at the rate of 3,175 Mg (3,500 tons) of solids per day. Clear
make-up water is pumped from the pond and the loop is closed by recycling ban
mill and module make-up water back into the system.2
662
-------
TABLE 3.
BOILER NO. 5 DESIGN DATA
Type Oil/Gas
Manufacturer Babcock and Mil cox
Number of Burners 24
Burner Arrangement Front and Rear Firing
Air Preheater Yes
Combustion Air Temperature 299°C (570°F)
Combustion Air Volume 20,200 Nm3/min/27°C (766,000 scfnj/80°F)
Recirculation Gas Yes
Volume 10,900 Nm3/min (414,000 scfm)
Temperature 374°C (705°F)
Reheat Two Stage
Design Steam Rate
Super Heat 970,000 kg/hr, 25 MPa, 538°C (2,135,000 Ib/hr, 3650 psig at 1000°F)
First Reheat 821,000 kg/hr, 7.3 MPa, 562°C (1,807,000 Ib/hr, 1065 psig at 1025°F)
Second Reheat 726,000 kg/hr, 2.6 MPa, 566°C (1,598,000 Ib/hr, 365 psig at 1050°F)
663
-------
In abbreviated terms, as the hot flue gas enters the venturi, Figure 5, it
is sprayed with slurry from 48 spray and 32 wall wash nozzles, resulting in
up to 99 percent of the particles being agglomerated to the sump below.
The gas continues through the sump making a 180 degree turn up through the
absorber section. In the reaction chamber, the sulfur dioxide is removed as the gas
is forced through a limestone slurry solution sprayed on stainless steel
sieve trays. The chemical reaction in part combines the calcium carbonate,
water and sulfur dioxide to form two relatively insoluble calcium salts
in the sump, calcium sulfate and calcium sulfite. The cleaned gas passes
through mist eliminators to remove moisture and then is reheated to avoid
deposits on the fans and provide buoyancy from the stack.2
The oil fired utility boiler is designed to control the emission of sulfur
dioxide by burning low sulfur content fuels. The emissions of particles
and hydrocarbons are controlled by furnace design, operation, and maintenance
to ensure complete combustion. Sub-stoichiometric combustion and flue
gas recirculation are used to control the emission of nitrogen oxides.
Sub-stoichiometric combustion involves two stages. The first stage is carried
out in a fuel rich, air lean environment. The second stage uses air injected
into the combustion zone to ensure complete combustion of the fuel. These
techniques lower the flame temperature which results in reduced formation of
nitrogen oxides.
Test Description and Conditions
Multimedia emission tests were conducted at the coal burning utility from
April 8 through 19, 1978, and at the oil burning utility from August 30, 1978,
through September 11, 1978. Gaseous, .liquid, and solid emissions were sampled
during coal and oil firing to obtain data for the comparative environmental
impact assessment. Flue gas sampling was conducted before and after the
scrubber at the coal fired utility to determine which pollutants are removed
or modified by the control device.
Emissions were characterized using EPA's phased approach. This approach
utilizes two levels of sampling and analysis (Level 1 and Level 2). Level 1
screening procedures are accurate within a factor of 2 to 3. They provide
preliminary assessment data and identify problem areas and information gaps.
Based on these data, a site specific Level 2 sampling and analysis plan is
developed. Level 2 provides more accurate and detailed information to
confirm and expand on the information gathered in Level 1. The methods and
procedures used for Level 1 are documented in the manual, "Combustion
Source Assessment Methods and Procedures Manual for Sampling and Analysis," September
1977. The Level 2 methods and procedures include "state-of-the art" techniques
required for this particular site.3
Normally all Level 1 samples are analyzed and evaluated before moving to Level 2.
Because of the program time constraints, the Level 1 and Level 2 samples were
obtained during the same test period.
664
-------
FLUE
GAS
WALL
WASH
(32)
VENTURI
SPRAY
(48)
SPENT SLURRY
TO POND
REHEAT
STEAM
750° F
r+
REHEAT COILS
400°C 4
x
X
X
X
X
TO
>I.D.
FAN
AND
STACK
DEMISTERS
CONTINUOUS
WASH
7\ A 7\ A 7\ 7\ A A
PREDEMISTER
ABSORBER
SPRAY
SIEVE
TRAYS
RECIRCULATION TANK
VENTURI
RECIRC. PUMP
ABSORB.
RECIRC.
PUMP
FIGURE 5
COAL FIRED BOILER FGD MODULE
665
-------
Gaseous Effluents
The boiler flue gas at the coal burning utility was sampled at the FGD inlet
gas manifold before the gas flow was divided to the eight scrubber modules.
Comparative scrubber outlet data were obtained by sampling at the stack. In
addition, some special S02, S03 and $04 measurements were made at the inlet
and outlet of the last scrubber module.
Bag samples were taken for on-site analysis for 02, N2, CO, C02, S02, NO,
and C-j-Cs hydrocarbons. Continuous monitoring was not employed during coal
fired testing.
Three sampling trains were used: (1) Source Assessment Sampling System, (2)
modified EPA Method 5 train, and (3) modified Goksoyr-Ross controlled condensa-
tion train. Isokinetic sampling conditions were maintained during particulate
sampling. The Source Assessment Sampling System (SASS) shown in Figure 6 was
used for Levels 1 and 2 organic sampling and for total particulate sampling.
The train consists of a heated probe, three cyclones, and a filter (the cyclones
and filter are in a heated oven). The cyclones were used only during the coal
inlet tests. During the other tests, the particle loadings were too low for
the cyclones to work effectively. The remainder of the system consists of a
gas conditioning system, an XAD-2 polymer absorbent trap, and a series of
impingers. The polymer traps gaseous organics and some inorganics, and the
impingers collect the remaining inorganics. All sample contact surfaces are
Type 316 stainless steel, Teflon, or glass. The train was run for 6 to 8 hours
so that a minimum of 30 cubic meters of gas was collected.
Previous sampling and analysis experience had indicated that the SASS train mate-
rials might contaminate certain organic and inorganic samples. The contamina-
tion is of concern only when the pollutant is present at a concentration that
is near the detection limit of the Level 2 methods. To avoid that possibility,
all-glass sampling trains were used to collect Level 2 samples. Method 5
sampling trains were modified as shown in Figure 7 for total particulate sam-
pling. This train sampled approximately 10 cubic meters of flue gas during
a 6 to 9 hour test run.
A modified Goksoyr-Ross controlled condensation train as shown in Figure 8
was used at each location to sample $03, 504, F", Cl", HF, HC1, and S0? (impin-
ger).
During Level 2 test runs, an MRI impactor (Figure 9) was used to obtain outlet
particle samples by particle size fraction. Polarized light microscopy (PLM)
was used on the particles collected by Method 5 train to obtain the
inlet particle size distribution.
Since the oil burning utility had no control device, all gas was sampled
at the stack. Bag samples were taken for N2, C02, and GI-CS hydrocarbons.
Continuous monitoring was done by gas chromatography for 02 and CO, by chemi-
luminescence for NO and NOX, and by pulsed fluorescence spectography for S02.
The SASS train was used to sample Level 1 and 2 organics. The Method 5 train
was employed for particulate collection. No inorganic analyses were jDer-r
formed. The modified Goksoyr-Ross train provided samples for $03, S0|, F", CT,
and S02 (impinger) analyses.
666
-------
STACK T.C.
HEATER
CON-
TROLLER
F LTER , GAS COOLER
riON
a
M 1
i
I
1
i
GAS
TEMPERATURE
T.C.
DRY GAS METER/ORIFICE METER
CENTRALIZED TEMPERATURE
AND PRESSURE READOUT
CONTROL MODULE
A A
OVEN
T.C.
XAD-2
CARTRIDGE
7
IMP/COOLER
TRACE ELEMENT
COLLECTOR
CONDENSATE
COLLECTOR
IMPINGER
T.C.
17 CMH (10 CFM) VACUUM PUMPS
FIGURE 6
SOURCE ASSESSMENT SAMPLING SYSTEM (SASS) SCHEMATIC
-------
00
FILTER
HOLDER
HEATED
CONTAINER
93°C (200°f)
CYCLONE
STANDARD
IMPINGERS
THERMOMETER
CHECK
VALVE
ORIFICE
H
THERMOMETERS
BY-PASS
VALVE
DRIERITE
0.2M(NH4)2S2Og
+0.02MAgNO3
VACUUM
GAUGE
MAIN VALVE
AIR-TIGHT
PUMP
FIGURE 7
MODIFIED METHOD 5 SAMPLING SYSTEM SCHEMATIC
VACUUM
LINE
-------
STACK
ON
VO
ADAPTER FOR
CONNECTING
HOSE
T.CWELL
ASBESTOS CLOTH
INSULATION
GLASS-COL
HEATING
MANTLE
RUBBER VACUUM
HOSE
DRY TEST
METER
THREE WAY
VALVE
3%
SILICA GEL
EMPTY
Na2C03
3% H0O
RECIRCULATOR
THERMOMETER
2^2
FIGURE 8
CONTROLLED CONDENSATION SAMPLING SYSTEM SCHEMATIC
-------
LONG RADIUS PROBE
12.5mm (% in) STAINLESS STEEL
VINTERNA
^CASCADE
IMPACTOR
FLEXIBLE
VACUUM
HOSE
CONNECTED
DIRECTLY TO
FIRST IMPINGER
- IMPINGER BOX OF
_ METHOD 5 TRAIN
CONTROL
BOX
FIGURE 9
MRI IMPACTOR SCHEMATIC
-------
Liquid Effluents
The liquid effluent from the scrubbing system installed at the coal burning
utility is a side stream of the slurry recycling system. Approximately
1,500 1/min (400 gpm) of spent slurry is discharged from each scrubber
module (a total of 12,100 1/min or 3,200 gpm for the entire scrubbing system)
and piped to a settling pond on site. Table 4 gives the approximate analysis
of the spent slurry solution.2 Upon entering the settling pond, the slurry
is diluted and the suspended solids settle. The clarified solution is
recycled to the scrubber or discharged to a cooling lake.
TABLE 4.
COMPOSITION OF SPENT SCRUBBER SLURRY
CaC03
CaS03
CaS04
Flyash
Total Solids
PH
46 g/1
50 g/1
16 g/1
30 g/1
14%
5.6
There is no liquid effluent connected with air quality control from the oil
burning utility.
Solid Effluent
At the coal burning plant, the scrubber solids are discharged to a settling
pond at a rate of approximately 132 Mg/hr. The majority of this waste material
is disposed of in a company owned landfill.
Dredged solids from the settling basins at the oil fired utility are hauled to
municipal landfills. .No solids are discharged directly to the environment.
Laboratory Analyses
The samples from the various sampling trains were returned to the laboratory
for analysis. Detailed analysis procedures can be found in the manual
"Combustion Source Assessment Methods and Procedures Manual for Sampling and
Analysis," September 1977.
671
-------
Level 1 analyses for particles and gases were made for inorganics by SSMS
and for selected anions and organics by liquid chromatography, infrared, and
mass spectroscopy. Solids, "slurries, and liquids were similarly analyzed, although
the work-up procedures were different.
More detailed and more quantitative Level 2 analyses were performed to identify
and quantitate specific compounds indicated by the Level 1 analyses.
Test Conditions
Five tests were performed with the coal burning utility boiler. Limited
supplementary coal firing tests were performed using the controlled condensation
train only. Four tests were performed with the oil burning utility boiler.
Test conditions are summarized in Table 5.
o
The flue gas entered the scrubber at the coal burning utility at about 163 C
(325 F) and exited at about 93QC (200 F). The stack gas at the oil burning
utility was about 121°C (250 F).
The coal unit operated at 71 to 87percent of design; the oil unit, at 62 to
94 percent.
MULTIMEDIA EMISSION RATES
Gaseous Emissions
As discussed previously, flue gas generated by the furnaces at the coal fired
power plant were passed through a two stage venturi absorber limestone slurry
scrubbing system prior to release to the atmosphere through a tall stack.
The oil fired boiler tested, on the other hand, operated with flue gas
recirculation for NOX control but was not equipped with an SC>2 removal system.
Emissions information presented in the following sections for the coal fired
plant normally includes both scrubber inlet and scrubber outlet (stack) data,
whereas information presented for the oil-fired plant is for stack emissions.
Criteria Pollutants
Federal New Source Performance Standards (NSPS) currently in effect define
allowable emission rates of NO (as NO^). S02 , and total particles from
fossil fuel fired utility boilers having 25 MW output or greater. More
stringent limitations have been proposed by EPA for NO , S02, and total particles
emissions. Federal NSPS do not currently address either CO or total hydrocarbon
emissions. Existing NSPS and corresponding proposed or potential emission
standards for coal- and oil-fired utility boilers are summarized in Table 6.
672
-------
01
^J
OJ
TABLE 5
SUMMARY OF TEST CONDITIONS
Test No.
132
133
134
135
136
141
142
143
144
Electrical
Output
(Gross-MW)
620
640
690
760
760
330
218
325
310
% of Maximum Nominal Fuel
Boiler Load3 Feed Rate
kg/hr
COAL FIRING
71
73
79
87
87
OIL FIRING
94.3
62.3
92.9
85.7
254,000
254,000
295,000
318,000
300,000
69,008
54,934
66,284
61,744
% 02 in
Flue Gas
6.0
6.0
6.0
6.0
6.0
6.35
5.76
5.66
5.98
Excess Air
at Furnace'-1
-b 15
^ 15
^ 15
-v 15
* 15
•x. 20
* 20
•b 20
* 20
Based on gross megawatts produced.
•'Full load excess air level. At lower loads, slightly higher levels are expected. No
allowance made for inleakage.
-------
TABLE 6.
EXISTING AND PROPOSED FEDERAL EMISSION STANDARDS
ng/J (lb/106 Btu)
Coal Fired Utilities
NSPS
NOV (as NOJ 300 (0.70)
A C.
S02 520 (1.20)
Proposed Standard
260 (0.60)
520 (1.20)
max. with 85%
reduction to
85 (0.20)
Oil Fired Utilities
NSPS Proposed Standard
130 (0.30) 130 (0.30)
344 (0.80) 520 (1.20)
max. with 85%
reduction to
85 (0.20)
Total
Particles
43 (0.10)
13 (0.03)
max. with 99%
reduction
43 (0.10) 13 (0.03)
Five tests were performed to determine the emissions from the coal fired
boiler. A summary of the criteria pollutant emissions data for the five
test series is presented in Table 7. For the oil fired boiler, four tests
were performed with oil firing. Criteria pollutant emissions data for
these tests are summarized in Table 8. In Table 9, average emissions
data from the coal fired and oil fired boilers are presented. The emissions
data are discussed by specific pollutant in the ensuing subsections.
Particles
Total
Average emissions of total particles were 1090 ng/J for coal firing prior to
scrubbing and 7.5 ng/J for oil firing. Total particles emissions from the
coal fired boiler after scrubbing were 80 ng/J. This corresponds to 93 percent
particle removal efficiency for the scrubber. Controlled particle emissions
from the coal fired boiler tested are still above the NSPS limit of 43 ng/J
for utility boilers. Particle emissions from the oil fired boiler tested,
however, are well below this limit. When comparing stack emissions, release
of particles to the atmosphere from coal firing is 10 times greater than
from oil firing on an emission factor basis.
Size Distribution
For the coal fired boiler, size distributions of particles at the scrubber
inlet and outlet were determined by two methods. Due to the high particle
674
-------
TABLE 7. SUMMARY OF CRITERIA POLLUTANT EMISSIONS - COAL FIRING
o\
>»»
Us
Tpci" MA
132 Inlet
132 Outlet
133 Inlet
133 Outlet
134 Inlet
134 Outlet
135 Inlet
135 Outlet
136 Inlet
136 Outlet
Average Inlet
Average Outlet
Emission Factor (ng/Jl
NOX
(as N02)
460
300
250
190
480
380
700
370
7*
400
520
330
CO*
<500
5500
5500
5500
S500
5500
*550
5550
5550
5550
5520
5520
so b
L
3210
640
3440
820
2970d
240
3650
ND
2560
770
3170
740
Crc6
Organics
NDC
ND
ND
ND
1-2.3
0.85-2.0
ND
ND
ND
ND
1-2.3
0.85-2.0
C7~C16
Organics
ND
ND
ND
ND
0.45
0.12
ND
ND
ND
ND
0.45
0.12
>C16
Organics
ND
ND
ND
ND
1.32
0.48
ND
ND
NDT
ND
1.32
0.48
Total
Organics
ND
ND
ND
ND
2.77-4.07
1.45-2.60
ND
ND
ND
ND
2.77-4.07
1.45-2.60
Total
Parti culates
ND
ND
ND
ND
1280
ND
900
80
ND
ND
1090
80
CO emission factor was based on the detection limit of 1000 ppm.
S02 emissions were determined from grab bag sampling for Test Nos. 132-134, and from the impinger
solution of the controlled condensation system for Test Nos. 135 and 136.
ND - data not available.
Determination of S02 emissions at the scrubber outlet for Test No. 134 appears to be in.error. This
data point was Judged to be an outlier at 90% probability level by the method of Dixon, and discarded
in the computation of average S02 emissions.
-------
TABLE 8. SUMMARY OF CRITERIA POLLUTANT EMISSIONS - OIL FIRING
Emission Factor (ns/J)
Test No.
141
142
143
144
Average
N0x
(as N02)
114
91
117
98
105
CO
11.3
5.2
5.6
4.5
6.6
so2
105
95
105
103
102
P -C
Ll L6
Organics
0.35-0.56
0.29-0.43
0.11-0.22
ND
0.25-0.40
r ~r
L7 L16
Organics
NDa
0.02
0.02
ND
0.02
>C16
Organics
ND
0.16
0.14
ND
0.15
Total
Organics
0.51-0.74b
0.48-0.61
0.27-0.38
ND
0.42-0.58
Total
Participates
ND
8.3
6.6
ND
7.5
ND - data not available.
'Total organics for Test No. 141 were estimated using C7-C,fi and >C,fi organics emissions from
Test Nos. 142 and 143.
-------
TABLE 9. CRITERIA POLLUTANT EMISSIONS FROM UTILITY BOILERS
Emission Factor (ng/J)
Pollutant
NO (as N02 near
x full load)
eo
so2
Total Organlcs
Total Particulates
Coal
Before Scrubber
715a
<520
3170
2.77-4.07
1090
Fi rl ng
After Scrubber
385*
<520
740
1.45-2.60
80
Oil Firing
116
6.6
102
0.42-0.58
7.5
Computed using average NO,, emissions from Test Nos. 135 and 136. These two tests
were conducted at 87% boiler load.
-------
loading at the scrubber inlet during coal firing, polarized light microscope
(PLM) analyses were utilized to obtain a size distribution in terms of optical
diameter and number of particles per size range. All other particle size
distribution determinations involved streams with substantially lower solids
loadings; therefore, an MRI cascade impactor was used. The cascade impac-
tor data differs from PLM analyses in that size distributions are determined
in terms of aerodynamic diameter and weight percent in each size range. Thus,
data from the two methods cannot be directly compared. For this reason, the
PLM data have been converted to the same basis as the impactor data by assum-
ing that particle density is independent of particle diameter. This is a
reasonable assumption because the major components of the particles gener-
ated from coal combustion, the aluminosilicates and iron oxides, are known to
partition equally between small and large particles. With the constant den-
sity assumption, the weight distribution in each size range would be propor-
tional to the product of the number distribution and the particle volume
representing the size range. The particle volume was calculated based on
the geometric mean diameter for the size range.
Particle size distribution data from coal firing tests are summarized in
Table 10. These data show a significant change in particle size distribu-
tion before and after scrubbing. The increase in the fraction of fine par-
ticles across the scrubber indicates that coarse particles were removed
more efficiently than fine particles. Particles larger than 3 ym were
removed with efficiencies of greater than 99 percent while particles smaller than
3 ym actually showed a net increase in emission rates across the scrubber.
This net increase indicates that the venturi scrubber at the coal fired plant
is probably not effective in removing fine particles present in the flue
gas. Additionally, the net increase also raises the possibility that fine
particles may be generated within the scrubber, or that the particle size
distribution may be modified during the high energy scrubbing process.
Particle size distribution data are not available from the oil firing tests.
However, data available from the literature have indicated that emitted par-
ticles are generally 90 wt percent less than 7 ym for oil fired boilers.5
In a recent paper from EPA's Health Effects Research Laboratory,6
it is stated that larger particles (from 3 to 15 ym) deposited in the upper
respiratory system (in the nasopharynx and conducting airways) can be
associated with health problems. This is in contrast to the past belief
that particles of health consequence were those less than 3 ym size and
deep lung penetrable. The area of hazard now is particles which are 15 ym
and less, which have been designated as "inhaled particles" (IP).
For oil firing, it can be reasonably assumed that almost all the particles
emitted are 15 ym or less. Emissions of inhalable particles at the coal
fired plant after scrubbing were approximately 75 ng/J, as compared to 7.5 ng/J
from the oil fired plant. Again, emissions of inhalable particles from coal
firing are 10 times greater than from oil firing, even after control.
678
-------
TABLE 10, SCRUBBER INLET AND OUTLET PARTICULATE SIZE
DISTRIBUTION - COAL FIRING
VO
Aerodynamic
Diameter
Size Range,
urn
< 1
1 - 3
3-10
> 10
Total
Weight %
Scrubber
Inlet
<0.01
<0.3
13
87
100
Scrubber
Outlet
82
11
1
6
100
Emission Factor (ng/J)
Scrubber
Inlet
<0.1
<3
141
946
1090
Scrubber
Outlet
65.6
8.8
0.8
4.8
80
Removal
Efficiency,
%
<0
<0
99.4
99.5
92.7
-------
Sulfur Compounds
For the coal fired boiler, SC>2 emissions were determined from grab bag
sampling by pulsed fluorescent analyzer and from the impinger solution of
the controlled condensation system by titration. For the oil fired boiler,
S02 emissions were monitored continuously by pulsed fluorescent analyzer.
Average S02 emissions were 3170 ng/J for coal firing prior to scrubbing and
102 ng/J for oil firing. Average S02 emission rates from the coal fired
boiler were 740 ng/J after scrubbing. This represents a mean scrubber effi-
ciency of 77 percent for SO? removal, within the design range of 75-80 per-
cent. For the oil fired boiler tested, the measured S02 emissions of 102 ng/J
are well below the NSPS limit of 344 ng/J for oil fired utility boilers.
In addition to the grab bag sampling at the coal fired plant and continuous
monitoring at the oil fired plant to determine S02 in flue gas emissions, the
modified Goksoyr-Ross controlled condensation system was also used at both
plant locations to determine S02, SOs, and SOzj emissions. A summary of these
sulfur compound determinations is presented in Table 11. Based on the total
sulfur compounds present in the flue gas, an average of 98.5 percent of the input
sulfur is emitted as S02 in coal firing when emissions are uncontrolled, and
97.7 percent of the input sulfur is emitted as S02 in oil firing. The removal
efficiency for S0£ by the scrubber system at the coal fired plant averaged 77
percent as previously discussed. About 80 percent of the $03, based on average
inlet and outlet concentrations, was also removed by the scrubber. Removal
efficiency for S0| averaged 83 percent, again based on averages, which is lower
than the total particle removal efficiency. This is an indication that a large
fraction of S0| may be associated with the fine particles in the flue gas stream
from the coal fired boiler, which are less efficiently scrubbed.
A comparison of the sulfur oxide emissions is best made by comparing emission
factors which are expressed as mole percent of total sulfur compounds (S02»
$03, and S04) present in the flue gas. The percentages of total sulfur com-
pounds present as S02 are almost identical for coal and oil firing. For S03,
the values were an average of 1.1 percent for coal firing and 1.2 percent for
oil firing. The SO^ emissions during oil firing were lower than typical values
reported in the literature , probably because of the lower vanadium and nickel
content of the fuel oil burned, and the use of flue gas recirculation which
reduces the oxygen concentration available for S03 formation. Conversion of
sulfur compounds to SO^ was an average of 0.4 percent for coal firing and 1.1
percent for oil firing. However, actual sulfate emissions were higher for
coal firing due to the higher fuel sulfur content of coal.
Nitrogen Oxides
NOX emissions from the coal fired boiler were determined by chemiluminescence
from grab bag sampling. For the oil fired boiler, NOX emissions were monitored
continuously by chemiluminescent instrumentation. The NOX emission factors
680
-------
TABLE ,11. S02, S03, AND S04= EMISSIONS FROM COAL AND OIL FIRING
oo
Sulfur
Compound
so2
SO.,
o
so/
1
Test
No.
135
136
I
II
III
IV
V
135
136
I
II
III
IV
V
135
136
I
II
III
IV
V
Scrubber
Inlet
ng/J
3650
2560
3260
3450
3980
3640
3610
99
54
49
39
24
35
39
15.5
18.2
41.2
22.2
24.2
18.6
15.3
Coal
Scrubber
Outlet
ng/J
NDa
770
ND
ND
ND
ND
ND
17
2
ND
ND
ND
ND
ND
5.1
2.4
ND
ND
ND
ND
ND
Fi rl ng
Mole % of
Total Sulfur
Compounds In
Flue Gas
97.6
97.8
98.0
98.7
99.1
98.9
98.8
2.1
1.7
1.2
0.9
0.5
0.8
0.9
0.3
0.5
0.8
0.4
0.4
0.3
0.3
Oil Firing
Mole % of
Removal Test Flue Total Sulfur
Efficiency No. Gas Compounds in
% ng/J Flue Gas
143 75.5 97.7
70.0
_-
__
—
--
--
82.8 143 1.1 1.2
96.3
__
—
--
--
— —
67.1 143 1.3 1.1
86.8
__
__
--
--
ND - data not available.
-------
near full load conditions were 715 ng/J for coal firing prior to scrubbing
and 116 ng/J for oil firing. Examination of the published data on NOX emis-
sions from coal fired cyclone boilers indicates an average emission factor of
662 ng/J.8 Thus, the NOX emission factor for the coal fired boiler tested
is well within the range typical of coal fired cyclone boilers. For oil
firing, the lower NOX emission factor was due to the use of flue gas recir-
culation for NOX control.
NOX data from the coal fired boiler tested generally indicate a significant
reduction of NOX across the scrubber. However, these NOX data were deter-
mined by chemiluminescence from grab bag sampling, and it has been recently
established that NOX decay inside these grab bags is rapid in the presence
of air. NOX removal by the FGD system at the coal fired power plant was
therefore probably not a real phenomenon. Additional data from on line
monitoring are needed to confirm any NOX reduction across the scrubber.
%
The measured NOX emissions from the coal fired boiler tested at both the
scrubber inlet and scrubber outlet exceed the NSPS limit of 300 ng/J for
coal fired utility boilers. On the other hand, the NOX emissions from the
oil fired boiler tested at the stack are slightly below the NSPS limit of
130 ng/J for oil fired utility boilers. NOX emissions from the coal fired
boiler (at the scrubber inlet) are approximately 6 times those from the
oil fired boiler. The differences in NO^ emissions are attributed to three
factors: (1) higher thermal NOX generation in cyclone furnaces because of
extremely high heat release rates and the resulting high furnace gas temper-
atures; (2) higher fuel nitrogen content in coal as compared to oil, leading
to higher fuel NOX generation in the coal fired boiler; and (3) the use of
flue gas recirculation for NOX control at the oil fired boiler tested.
Carbon Monoxide
Average CO emissions from the oil fired boiler tested were 6.6 ng/J. No
effort was made for accurate determination of CO from the coal fired boiler.
The reported CO emission factor of less than 520 ng/J for the coal fired
boiler was based on detection limits of the instrumentation. However, CO emis
sions from coal fired and oil fired utility boilers should be comparable in
magnitude.
Organics
In the determination of organic emissions, gas chromatographic analyses were
performed on grab bag samples of flue gas and catches from the Level 1 sam-
pling (SASS train). Additionally, gravimetric analyses were performed on
Level 1 samples to quantify high molecular weight organics. Each bag sample
was collected over an interval of 30 to 45 minutes, with a single sample
being collected per test. These samples were utilized to measure C-j to
hydrocarbons. The SASS train collects approximately 30 cubic meters of
flue gas isokinetically during the test. Samples from the SASS train
682
-------
were analyzed to determine organics higher than Cg. The Cy to C-jg fraction
was determined by gas chroma tograph while organics higher than C-jg were
determined gravimetrically.
Average organic emissions were 2.8-4.1 ng/J for coal firing prior to scrub-
bing and 0.4-0.6 ng/J for oil firing. Comparison of the data presented
indicates that emissions from coal firing are greater than those from oil
firing in Cj-C* organics, Cy-Cig organics, and high molecular weight (>C]5)
organics. At the coal fired plant, emissions of organics after scrubbing
were 0.85-2.0 ng/J for the C-j-Cg fraction (14 percent removal), 0.12 ng/J
for the Cy-C]5 fraction (72 percent removal), and 0.48 ng/J for the high
molecular weight fraction ( 63 percent removal). However, the organic emis-
sions from coal firing after scrubbing are still greater than those from oil
firing in all fractions.
Selected samples from the coal firing and oil firing tests were analyzed by
combined gas c h roma tog raphy /mass spectrometry (GC/MS) for the identification
of organic compounds present. For the coal fired boiler, the organic com-
pounds identified include aliphatic hydrocarbons, substituted benzene,
ethyl benzaldehyde, dimethyl benzaldehyde, 2,6-pereriden-dione-4, and 2,6 di-
methyl-2,5-heptadion-4-one, and the methyl ester of a long chain acid, at
concentration levels ranging from 0.2 to 20 yg/nr in the flue gas prior to
scrubbing. With the exception of ethyl benzaldehyde, substituted benzenes,
and aliphatic hydrocarbons, none of the other organic compounds were identi-
fied at the scrubber outlet. For the oil fired boiler, the organic compounds
identified include aliphatic hydrocarbons, benzaldehyde, trimethyl cyclo-
hexane-one, G£ substituted acetophenone, the methyl esters of benzoic acid
and substituted benzoic acid, diethylphthalate, and ethyl benzaldehyde, at
concentration levels between 0.02 and 2 yg/m^. Thus, the organic compounds
from coal firing and oil firing are somewhat similar.
Emissions of polycyclic organic matter (POM) determined by GC/MS for coal
firing and oil firing are summarized in Tables 12 and 13. For coal firing,
most of the POM compounds identified are naphthalene, substituted naphtha-
lenes, biphenyl, and substituted biphenyls. No POM compounds were identified
at the scrubber outlet of the coal fired plant. POM compounds found at the
scrubber inlet are at levels several orders of magnitude below their respec-
tive MATE values. The Minimum Acute Toxicity Effluent (MATE) values are
emission level goals developed under direction of EPA, and can be considered
as concentrations of pollutants in undiluted emission streams that will not
adversely affect persons or ecological systems exposed for short periods
of time. ' 0
The total POM emissions from oil firing are lower than those from coal firing.
Again, naphthalene is the principal component of the POM emissions. With the
exception of the possible presence of benzo(a)pyrene, all POM compounds from
oil firing are at levels too low to be of environmental concern.
Inorganics
For the coal fired boiler, trace elements present in the flue gas were deter
mined using atomic absorption spectroscopy (AAS). For the oil fired boiler,
683
-------
TABLE 12. POM EMISSIONS FROM COAL FIRING
PRIOR TO SCRUBBING
Emission
Compound Concentration
yg/m3
Decahydronaphthal enea
Ditert-butyl naphthalene
Dimethyl isopropyl naphthalene
Hexamethyl biphenyl
Hexamethyl, hexahydro indacene
Dihydronaphthalene
CIQ substituted naphthalene
CIQ substituted decahydro-
naphthalene3
Methyl naphthalene
Anthracene/phenanathrene
1-1 '-biphenyl
9,10-dihydronaphthalene /
1-1 '-diphenylethene
1 ,l-bis(p-ethyl phenyl )-ethane/
tetramethyl biphenyl a
5-methyl-benz-c-acridine
2,3 dimethyl decahydro-
nanh^hal Ana
0.1
0.3
0.3
0.6
1.0
0.03
0.06
1.0
1.6
0.3
4.0
0.2
9.0
0.2
<0.03
MATE
Value
ug/m3
130,000
230,000
230,000
1,000
No data
130,000
230,000
130,000
130,000
1,600
1,000
130,000
1,000
10,500
130,000
Potential
Degree of
Hazard®
<0.0001
<0.0001
<0.0001
0.0006
<0.0001
<0.0001
<0.0001
< 0.0001
0.0002
0.004
<0.0001
0.009
<0.0001
< 0.000.1
Total
18.7
The MATE values for decahydronaphthalene, dihydronaphthalene and any
substituted decahydronaphthalene are assumed to be the same as that for
tetrahydronaphthalene.
The MATE values for alky! naphthalenes are assumed to be the same as
that for methyl naphthalene.
c The MATE value for hexamethyl biphenyl is assumed to be the same as
that for biphenyl.
The MATE value for 5-methyl-benz-c-acridine is assumed to be the same
as that for benz(c)acridine.
e Calculation carried to four significant figures.
684
-------
TABLETS. POM EMISSIONS FROM OIL FIRING
Compound
Naphthalene
Phenanthridine9
Dibenzothiophene b
Anthracene/
phenanthrene
Fluoranthene
Pyrene
Chrysene/
benz(a)anthracene
Benzopyrenec and
perylenes
Tetramethyl - ,.
nhpnanthvono
Emission 3
Concentration, pg/m
Test 142 Test 143
7 3
0.3
0.6
1 0.2
1
1
0.1
0.04
0.6
MATE
Value
pg/m 3
50,000
No data
23,000
1,600
90,000
230,000
45
0.02
1 ,600
potential Degree
-, of Hazard
Test
142
0.0001
ND
<0.0001
0.0006
<0.0001
<0.0001
0.0022
2
• «
Test
143
< 0.0001
__
--
0.0001
--
--
—
--
0.0004
Total
11.0
3.8
The presence of phenanthridlne has not been positively identified.
The presence of dibenzothiophene has not been positively identified.
Also, the MATE value for dibenzothiophene is assumed to be the same
as that for benzothiophene.
c The MATE value for benzo(a)pyrene is used in the computation of the
potential degree of hazard.
The MATE value for tetramethyl phenanthrene is assumed to be the same
as that for phenanthrene.
685
-------
trace element concentrations in the flue gas were computed by assuming that
all trace elements present in the fuel oil are emitted in the stack. Trace
elements in the fuel oil were determined using spark source mass spectrometry
(SSMS).
Concentrations of 18 major trace elements present in the flue gas during coal
and oil firing are presented in Tables 14 and 15. To assess the hazard poten-
tial of these emissions, the emission concentrations are compared with the
Minimum Acute Toxicity Effluent (MATE) values. The MATE values tabulated
here represent air concentrations which were derived from human health con-
siderations. Analysis of the flue gas generated during coal firing indicates
that 16 elements exceeded their respective MATE values at the scrubber inlet
and 7 exceeded their MATE values at the scrubber outlet. These seven elements
which are of potential hazard are arsenic, cadmium, chromium, nickel, lead,
iron, and zinc. During oil firing, only chromium and nickel exceeded their
MATE values at the stack. The MATE values for nickel and chromium are extremely
low due to considerations for potential human carcinogenicity. If Threshold
Limit Values (TLV's) are used as the basis for comparison, then emissions of
chromium and nickel from oil firing are respectively below and at their TLV's
which are each 0.5 mg/nr.
Emission factors for the 18 trace elements analyzed are presented in Tables
16 and 17. Comparison of these emission factors shows that, with the excep-
tion of cobalt and nickel, emissions of trace elements from coal firing after
scrubbing are considerably greater than corresponding uncontrolled emissions
from oil firing. Also presented in Table 16 is the scrubber removal efficiency
for each element during coal firing. An overall removal efficiency of 94 per-
cent was obtained for these trace elements. However, some elements were removed
with less than the average removal efficiency.
To better understand the removal efficiency of the individual trace elements,
the enrichment factor has been computed for each trace element across the
scrubber during coal firing. The enrichment factor is defined here as the ratio
of the concentration of a trace element to that of aluminum in the scrubber
outlet, divided by the corresponding ratio in the scrubber inlet. Aluminum
is selected as the reference material because it has been known to partition
equally among particles of different size.* The enrichment factors presented
in Table 16 show that all the other 17 trace elements are enriched across the
scrubber. The enrichment observed is due primarily to the partitioning of
trace elements as a function of particle size, and the greater collection
efficiency of the scrubber for the large size particles. Also note
that many of the trace elements' that show an enrichment trend, such as
mercury, selenium and arsenic, either occur as element vapors or form volatile
compounds at furnace temperatures. Condensation and surface adsorption of the
more volatile elements or their oxides and halides downstream of the furnace
could, therefore, result in higher concentrations of these trace elements on
smaller particles.
* Silicon, iron, and scandium have also been used by other investigators as
the reference element in the computation of enrichment factors. Notice
that iron has an enrichment factor of 1.4 in this study while silicon and
scandium were not measured.
686
-------
TABLE 14. EMISSION CONCENTRATIONS OF TRACE ELEMENTS
DURING COAL FIRING
El ement
Al
As
Be
Ca
Cd
Co
Cr
Cu
Fe
Hg
Mn
Ni
Pb
Sb
Se
Sr
V
Zn
Concentration
, mg/m
scrubber Scrubber
Inlet Outlet
132
0.98
0.021
49
5.1
0.19
1.3
1.2
401
0.095
0.70
2.0
11
0.78
0.37
0.46
0.78
105
3.0
0.94
0.0018
2.0
0.58
0.013
0.12
0.19
13
0.0057
0.15
0.054
2.9
0.27
0.088
0.038
0.083
21
MATE For Air
(Health Basis),
mg/m 3
5.2
0.002
0.002
16
0.010
0.050
0.001
0.20
1.0
0.050
5.0
0.015
0.15
0.50
0.20
3.0
0.50
4.0
Potential
. of Haz
Scrubber
Inlet
25
490
11
3.1
520
3.7
1300
6.0
400
1.9
0.14
130
73
1.6
1.8
0.15
1.6
26
Degree *»
ard
Scrubber
Outlet
0.58
470
0.90
0.13
58
0.26
120
0.95
13
0.11
0.03
3.6
19
0.54
0.44
0.013
0.17
5.2
Potential degree of hazard is defined as the ratio of the discharge
concentration to the MATE value.
687
-------
TABLE 15. EMISSION CONCENTRATIONS OF TRACE ELEMENTS
OWING OIL FIRING
Element
Alb
As
Be
Cab
Cd
Co
Cr
Cu
Fe
Hgc
Mn
Ni
Pb
Sb
Se
Sr
V
Zn
Flue Gas
Concentration
mg/m3
0.084
<0.0007
<0.0007
1.66
<0.0007
0.03
0.006
0.007
0.1
<0.005
0.003
0.5
0.003
<0.0007
0.001
0.006
0.07
0.01
MATE for Air
(Health Basis)
5.2
0.002
0.002
16
0.010
0.050
0.001
0.20
1.0
0.050
5.0
0.015
0.15
0.50
0.20
3.0
0.50
4.0
Potential
Degree of
Hazards
0.016
<0.33
<0.33
0.10
<0.07
0.6
6
0.035
0.1
<0.1
0.0006
33.3
0.02
<0.0014
0.005
0.002
0.14
0.003
Potential degree of hazard is defined as the ratio of the discharge
concentration to the MATE value.
Estimates based on average trace element content of fuel oil reported
in literature. SSMS analysis performed for the fuel oil samples
reports Al and Ca as major components with no numerical values
given.
By elemental sparging.
688
-------
TABLE 16. EMISSION FACTORS FOR TRACE ELEMENTS
DURING COAL FIRING
El ement
AT
As
Be
Ca
Cd
Co
Cr
Cu
Fe
Hg
Mn
Ni
Pb
Sb
Se
Sr
V
Zn
Emission
Scrubber
Inlet
49
0.37
0.0079
,18
1.9
0.069
0.48
0.45
150
0.035
0.26
0.73
4.1
0.28
0.14
0.17
0.29
39
Factor, ng/J
Scrubber
Outlet
1.1
0.35
0.00067
0.73
0.21
0.0047
0.046
0.072
4.9
0.0021
0.054
0.020
1.1
0.099
0.033
0.014
0.030
7.7
Removal
Efficiency,
%
98
4
91
96
89
93
90
84
97
94
79
97
74
66
76
92
89
80
Enrichment
Factor
1.0
41
3.7
1.7
4.8
2.9
4.1
6.9
1.4
2.6
8.8
1,2
11
15
10
3.5
4.4
8.4
Total
265
17
94
689
-------
TABLE 17. EMISSION FACTORS FOR TRACE ELEMENTS
DURING OIL FIRING
El ement
Ala
As
Be
Caa
Cd
Co
Cr
Cu
Fe
Hg
Mn
N1
Pb
Sb
Se
Sr
V
Zn
Emission Factor
ng/J
0.029
<0.0002
<0.0002
0.567
<0.0002
o.on
0.002
0.002
0.05
<0.002
0.0009
0.182
0.0009
<0.0002
0.0005
0.002
0.023
0.005
Estimated based on average trace element
content of fuel oil reported in literature.
SSMS analysis performed for the fuel oil
samples reports Al and Ca as major components
with no numerical values given.
690
-------
Liquid Waste
There are no significant wastewater streams associated with the oil fired
boiler tested. For the coal fired boiler, the two major wastewater streams
are: (1) wastewater discharge from the slag tank to the ash pond; and (2)
overflow from the settling pond for spent scrubber slurry. The flowrates
for these two wastewater streams are approximately 2.89 Gg/hr and 0.77 Gq/hr,
respectively.
Inorganics
Analytical results for major inorganic cations in the wastewater stream from
the slag tank to the ash pond and the scrubber slurry settling pond overflow
are presented in Table 18. Also presented are the analytical results for the
scrubber make-up water obtained from the settling pond, and for the scrubber
discharge liquid (filtrate from the spent slurry). Of the 18 elements ana-
lyzed, iron exceeds its health MATE value and iron, calcium, aluminum, cad-
mium, vanadium, and zinc exceed their respective ecological MATE values for
the wastewater stream to the ash pond. For the scrubber slurry settling pond
overflow, calcium, cadmium, manganese, nickel, and lead exceed both their
health and ecological MATE values; additionally aluminum, iron, and zinc
exceed their respective ecological MATE values. Comparison of the inorganic
data for the scrubber slurry pond overflow and the scrubber make-up water
(from the settling pond) indicates that the trace element concentrations in
these two streams are almost identical. This agreement supports the relia-
bility and accuracy of sampling and analysis of trace elements for the waste-
water streams.
Organics
Concentrations of Cy to C-jg organics and high molecular weight (>Cis) organics
measured in the wastewater streams from coal firing are summarized in Table
19. The total organics detected are low, ranging from 0.06 mg/liter in the
wastewater to the ash pond to 0.57 mg/liter in the scrubber slurry settling
pond overflow.
GC/MS analyses were performed to identify the organic compounds present in
the wastewater streams from coal firing. In the extraction of the aqueous
samples, the samples were first acidified to pH 2 and extracted with methylene
chloride. The samples were then adjusted to pH 7 and reextracted with methyl-
ene chloride. A final extraction with methylene chloride was made at pH 11.
The results of the GC/MS analyses are presented in Table 20. In general, the
detected compounds consist of oxygenates such as ketones, alcohols, ethers,
and cyclic ethers. Some of these are lightly halogenated. Typical MATE
values for these classes of compounds are greater than 1000 yg/liter. Thus,
the levels of organics present in the wastewater streams from coal firing do
not appear to warrant any environmental concern.
691
-------
TABLE 18.
TRACE ELEMENT CONCENTRATIONS IN WASTEWATER DISCHARGES FROM COAL FIRING - TEST 135
MATE Value, mg/1
Element
Al
As
Se
Ca
Cd
Co
Cr
Cu
Fe
Hg
Mg
Mn
Si
Pb
Sb
Sr
V
Zn
Health
80
0.250
0.030
240
0.050
0.75
0.25
5.0
1.5
0.010
90
0.25
0,23
0.25
7.5
46
- 2.5
25
Ecolpgv
1.0
0.050
0.055
16
0.001
.0.25
0.25
0.050
0.25
0.250
86
0.10
0.010
0.05
0.20
No MATE
0.15
0.10
Water to Ash
mg/1
3.5
0.012
0.0003
146
0.020
0.002
0.012
0.008
3.0
•£0.0002
13.8
•=•0.38
0.01
0.030
0.002
0.836
0.25
0.13
Pond
Potential Degree
of Hazard
Health
0.044
0.048
0.010
0.61
0.40
0.0027
0.048
0.0016
2.0
-------
TABLE 19. OR6ANICS IN WASTEWATER DISCHARGES FROM COAL-FIRING
Concentration, mg/1
C7
C8
C9
C10
cn
C12
C13
C14
C15
C16
>C1C
Water to
Ash Pond
0
0.02
0
0
0.02
0
0
0
<0.01
0.01
0
Settling
Pond
Overflow
0
0
0
0
0.04
0.01
0
0
<0.01
0.01
0.5
Scrubber
Make-up
Water
0
0
0
0
0.04
0
0
0
<0.01
0.01
0.3
Scrubber
Discharge
Liquid
0
0
0
0
0
0
x
0
0
0
0
0.1
Total
0.06
0.57
0.36
0.1
693
-------
TABLE 20. GC/MS ANALYSES OF ORGANICS
IN WASTEWATER STREAMS FROM COAL FIRING
Compound
Water to Ash Pond Settling Pond
Acid Neutral Basic Overflow
Extract3 Extract Extract Acid Extract
Scrubber Make-up Mater
NeutralBasic
Extract Extract
Scrubber
Discharge Liquid
Basic Extract
vO
Olefin or ketone; Cg - C,,
Tetrachloropropane (possible)
e-chloro-N-ethyl-N'-O-methyl ethyl)-! ,3,5-
tr1az1ne-2,4-d1am1ne
8-methyl-3a-d1hydronaphthalene-one
3a,7a-d1hydro-5-methyl-1ndene-l ,7(4h)-d1one
Qulnoline
Butyl naphthalene(4) (plus a possible alkyl
substituted naphthalene)
1-chloro-2,4-hexadiene
Cg nltrlle or C5 alcohol
Di-2-ethyl-hexyl ester of nonane dioic add
2,2,5,5-tetramethyl hexane
Dlphenylheptane (possible)
4 ug/1
0.5 yg/1
0.3 vg/1
2 yg/1
Identified compounds are present in this extract at concentrations below 15 vg/1.
-------
Solid Waste
There are no significant solid wastes generated from the oil fired boiler
tested. For the coal fired boiler, the two major solid waste streams gener-
ated are: (1) bottom slag/fly ash from the slag tank; and (2) scrubber sludge
from the FGD operation. These two solid wastes are generated at the rates
of 110 Mg/hr and 130 Mg/hr on dry basis, respectively.
Inorganics
The concentration of trace elements present in the combined fly ash/bottom
slag and in the scrubber sludge are presented in Tables 21 and 22. In about
two-thirds of the cases, the trace element concentrations in the combined
ash and the scrubber sludge have exceeded the health or the ecological
MATE value for solids and in about haYf of the cases, have exceeded both
values. The potential degree of hazard for the trace elements in the
combined ash and the scrubber sludge is sufficiently high to warrant the
disposal of these solid wastes in specially designed landfills.
Organics
Concentrations of C? to C,fi organics and high molecular weight (>C-ig) organics
measured in the solid wastes from coal firing are summarized in Table 23. The
total organics amount to 86.2 mg/kg for the combined bottom ash/fly ash and
6.6 mg/kg for the scrubber discharge solids. High molecular weight organics
were not detected for either solid waste.
Organics present in the bottom ash/fly ash are probably the result of incom-
plete combustion, or the adsorption of organics by fly ash particulates.
Organics are present in the scrubber discharge solids because of the partial
removal of these compounds from the flue gas stream in the FOG system.
Although no specific organic compound identification information is available,
the high trace element content of these solid wastes far outweighs the concern
for the organic content. Disposal in specially designed landfills should be
satisfactory to handle the potential degree of hazard.
695
-------
TABLE 21. TRACE ELEMENT CONTENT OF BOTTOM AND
FLY ASH FROM COAL FIRING
Element
Al
As
Be
Ca
Cd
Co
Cr
Cu
Fe
Hg
Mg
Mn
Ni
Pb
Sb
Sr
V
Zn
Concentration
yg/g
82,000
10.6
8.2
66,000
1.2
44
208
822
174,000
< 1
ND3
698
328
160
7.0
282
86
1,508
MATE val
Health
16,000
50
6
48,000
10
150
50
1,000
300
2
18,000
50
45
50
1,500
9,200
500
500
ue, yg/g
Ecology
200
10
11
3,200
0.2
50
50
10
50
50
17,400
20
2
10
40
ND
30
20
Potential Degree
of Hazard
Health
5
0.21
1.40
1.40
0.12
0.29
4.1
0.82
580
< 0.5
—
14 ,
7.3
3.2
0.0046
0.03
0.17
3.02
Ecology
410
1
0.75
21
6
0.88
4.1
82
3,480
< 0.02
—
35
164
16
0.18
—
2.87
75
ND - data not available.
696
-------
TABLE 22. TRACE ELEMENT CONTENT OF SCRUBBER DISCHARGE SOLIDS
Element
Al
As
Be
Ca
Cd
Co
Cr
Cu
Fe
Hg
Mg
Mn
Ni
Pb
Sb
Sr
V
Zn
Concentration
yg/g
24,000
111.4
2.48
51 ,000
36
22
52
188
50,000
< 1.0
0.487
564
96
1,080
38
994
188
6,492
MATE val
Health
16,000
50
6
48,000
10
150
50
1,000
300
2
18,000
50
45
50
1,500
9,200
500
500
ue, yg/g
Ecology
200
10
11
3,200
0.2
50
50
10
50
50
17,400
20
2
10
40
NDb
30
20
Potential Degree
of Hazard a
Health
1.5
2.23
0.41
1.06
3.6
0.15
1.04
0.19
167
< 0.5
< 0.0001
11.28
2.13
21.6
0.03
0.11
0.38
12.98
Ecology
120
11.14
0.23
16
180
0.44
1.04
18.80
1,000
< 0.02
< 0.0001
28.2
48
108
0.95
6.27
325
a Calculation carried to four significant figures.
ND - data not available.
697
-------
TABLE 23. ORGANICS IN SOLID WASTE STREAMS
FROM COAL FIRING
Carbon
Number
C7
C8
C9
C10
cn
C12
C13
C14
C15
C16
>clfi
Concentration
Bottom Ash/
Fly Ash
0
33.0
28.8
0
0
0
0
0
0
24.4
0
, mq/kg
Scrubber
Discharge Solids
0
0
4.7
1.9
0
0
0
0
0
0
0
Total
86.2
6.6
698
-------
RESULTS
(1) Average emissions of total particles were 1090 ng/J for the coal fired
boiler prior to scrubbing and 7.5 ng/J for the oil fired boiler. Con-
trolled particle emissions from the coal fired boiler tested were
80 ng/J, which were 10 times the particle emissions from the oil
fired boiler.
(2) For the coal fired boiler, over 99 percent of the total particle was
greater than 3 ym before scrubbing. The particle size distribution
data indicate that approximately 93 percent of the total particles
were less than 3 ym after scrubbing.
(3) For the coal fired boiler, there appeared to. be a net increase in
emission rates across the scrubber for particles less than 3 ym in
size. This net increase can be attributed to the poor removal effi-
ciency of the scrubber for fine particles, the possibility that
fine particles may be generated within the scrubber, or that the
particle size distribution may be modified during the high energy
scrubbing process.
(4) Stack emissions of inhalable particles (<15 ym) were 75 ng/J for the
coal fired boiler and 7.5 ng/J for the oil fired boiler.
(5) Uncontrolled S02 emissions from the coal fired and oil fired
boilers were 3170 ng/J and 102 ng/J, respectively. Controlled S02
emissions from the coal fired boiler were 740 ng/J, which corresponds
to a mean scrubber efficiency of 77 percent. Stack emissions of S02
from the coal fired boiler were 7 times the S02 emissions from the
oil fired boiler.
(6) For the coal fired boiler prior to scrubbing, approximately 1.1 percent
of th§ sulfur compounds in the flue gas were present as 503 and 0.4 percent
as S04. For the oil fired boiler, approximately 1.2 percent of the sulfur
compounds in the flue gas were present as SO- and 1.1 percent as S0|.
(7) Stack emissions of SOo were 9.5 ng/J for the coal fired boiler and
1.1 ng/J for the oil fired boiler. Stack emissions of $04 for the coal
fired boiler, at 3.8 ng/J, were also greater than SO^ emissions of 1.3
ng/J for the oil fired boiler.
(8) NOX emissions from the coal fired boiler tested were approximately 6
times the NOX emissions from the oil fired boiler tested. The differ-
ences in NOX emissions are mainly due to the high heat release rates of
the cyclone furnaces for coal firing and the use of N0x controls for
oil firing.
(9) Total organic emissions from the coal fired boiler were 2.8-4.1 ng/J
before scrubbing and 1.4-2.6 ng/J after scrubbing. These were greater
than the total organic emissions of 0.4-0.6 ng/J from the oil fired
boiler.
699
-------
(10) For coal firing, most of the POM compounds found at the scrubber inlet
were naphthalene, substituted naphthalenes, biphenyl, and substituted
biphenyls. Concentration levels of these POM compounds were several
orders of magnitude below their respective MATE values. No POM com-
pounds were identified at the scrubber outlet.
(11) Total POM emissions from oil firing were lower than those from coal fir-
ing before scrubbing. Naphthalene was the principal component of the
POM emissions from oil firing. With the exception of the possible pres-
ence of benzo(a)pyrene, concentrations of POM emissions from oil firing
were also several orders of magnitude below their respective MATE values.
(12) Of th.e 18 major trace elements analyzed in the flue gas stream for the
coal fired boiler, 16 exceeded their health MATE values, at the scrubber
inlet and 7 at the scrubber outlet. The seven elements which are of
environmental concern from coal firing are arsenic, cadmium, chromium,
iron, nickel, lead, and zinc. For oil firing, only chromium and nickel
in the flue gas exceed their health MATE values.
(13) Two major wastewater streams from coal firing were not associated
with oil firing. These were the wastewater discharge to the ash pond
and the overflow from the scrubber slurry settling pond. Analysis of
these wastewater streams indicates that the concentration levels of some
of the trace elements present exceeded their respective health and eco-
logical MATE values. However, the concentration levels of organics
present in these wastewater streams do not indicate any major environmental
concerns currently.
(14) No significant solid wastes were generated from the oil fired
boiler. For the coal fired boiler, the two major solid waste streams
were the combined bottom slag/fly ash from the slag tank and the scrubber
sludge. For these two solid wastes, almost all the 18 trace elements
analyzed have exceeded both their health and ecological MATE values.
Because th2 trace elements may leach from the disposed ash and
scrubby sludge, these solid wastes must be disposed of in specially
designed landfills.
(15) In summary, the controlled multimedia emissions from the coal fired
boiler tested were of greater environmental concern than the multimedia
emissions from the oil fired boiler tested. Stack emissions of pollut-
ants from coal firing could be further reduced by proper NO control and
higher pressure drop across the venturi scrubber for more efficient S0?
and particulate removal. However, the reduction in gaseous emissions
from coal firing could also result in higher pollutant levels in the
solid wastes generated at the coal fired power plant. The consequence
will be the transferring of a high volume, low concentration pollution
stream to a low volume, high concentration stream which can be more
readily contained.
700
-------
ACKNOWLEDGEMENT
We wish to thank the owners and operators of the two utility boilers for their
cooperation throughout the test program. The work was conducted under EPA
contract No. 68-02-2613, work assignment No. 8. Dr. Wm. H. Fischer of Gilbert
Associates, Inc. was of material assistance in the preparation of this paper.
REFERENCES
1. Ponder, W.H. and D.C. Kenkeremath, Conventional Combustion Environmental
Assessment Program. MS 78-44, Rev. 1, Mitre Corp., McLean, VA., September
1978.
2. Personal communication from the utility plant Superintendent of Air
Quality Control, April 17, 1978.
3. Harris, J.C. and P.L. Lewis, "EPA/IERL-RTP Procedures for Level 2 Sampling
and Analysis of Organic Material," Arthur D. Little, Inc., EPA-600/7-78-016
(NTIS No. 279212), February 1978.
4. Dixon, W.J., Ann. Math. Stat. 22 (1951) 68.
5. Weast, R.E., L.J. Shannon, P.G. Gorman and C.M. Guenther, "Fine Particu-
late Emission Inventory and Control Survey," Midwest Research Institute,
EPA-450/3-74-040, (NTIS No. PB 234156), January 1974.
6. Miller, S.S., "Inhaled Particulate," Environ. Sci. and Tech.. 12 #13
(1978) 1353-1355.
7- Homolya, J.B. and J.L. Cheney, "Assessment of Sulfuric Acid and Sulfate
Emissions from the Combustion of Fossil Fuels," in Workshop Proceedings
on Primary Sulfate Emissions from Combustion Sources, Vol. II, pp 3-11,
EPA-600/9-78-020a (NTIS No. PB287437), August 1978.
8. Ctvrtnicek, T.E. and S.J. Rusek, "Applicability of NOX Combustion Modifi-
cations to Cyclone Boilers (Furnaces)," Monsanto Research Corp.,
EPA-600/7-77-006 (NTIS No. PB263960), January 1977.
9. TRW, Inc., "Emissions Assessment of Conventional Combustion Systems,"
Progress Report No. 1, EPA Contract No. 68-02-2197, June 1978.
10. Cleland, J.G. and G.L. Kingsbury, "Multimedia Environmental Goals for
Environmental Assessment," Vols. I and II, Research Triangle Institute,
EPA-600/7-77-136a/b (NTIS No. PB276919/20), November 1977.
701
-------
TITLE
OPERATING AND STATUS REPORT
WELLMAN-LORD SO REMOVAL/ALLIED CHEMICAL SO REDUCTION
FLUE GAS DESULFURIZATION SYSTEMS
AT
NORTHERN INDIANA PUBLIC SERVICE COMPANY
AND
PUBLIC SERVICE COMPANY OF NEW MEXICO
by
D. W. Ross
Director of Technical Services
Davy Powergas Inc.
Lakeland, Florida
James Petrie
SO Chemical Plant Supervisor
Public Service Company of New Mexico
San Juan Station
Waterflow, New Mexico
F. W. Link
Supervisor Plant Engineering
Northern Indiana Public Service Company
Michigan City, Indiana
A paper presented at the Fifth Symposium on Flue Gas Desulfurization
for the Environmental Protection Agency
March 5-8, 1979
Las Vegas, Nevada
702
-------
ABSTRACT
This paper has been written in two Sections.
Section I contains an up-to-date report on the FGD plant at NIPSCO's
Dean H. Mitchell Station at Gary, Indiana, and the most recent data
pertaining to plant operations, problems encountered with corrective
actions and anticipated plans for future plant operations. The FGD
plant consists of the Davy/Wellman-Lord SO- Recovery Process and Allied
Chemical Corporation's S0_ Reduction Process. The plant information
starts after the successful completion of the performance test on
September 16, 1977, and continues to January 1, 1979.
Section II has a report on Public Service Company of New Mexico, at
Waterflow, New Mexico. This plant also consists of the Davy/Wellman-Lord
SO^ Recovery Process and Allied Chemical Corporation's SO- Reduction
Process. The Wellman-Lord SO- Recovery Process is attached to power
generation Units No. 1 and 2, and is composed of three operating sections:
(1) Separate S0_ Absorption for each power generation unit, (2) a dual
train Chemical Regeneration Plant and single Purge Treatment Unit, and
(3) a dual train Allied Chemical SO Reduction Unit. This part of the
report contains information beginning with the initial plant start-up in
April 1978 and follows to January 1, 1979. It too will outline the
plant operations, problems encountered with corrective action taken, and
plans for continued plant operations.
703
-------
INTRODUCTION
The Wellman-Lord S0« Recovery Process was developed in the late 1960's.
The process is being used throughout the world, Exhibits 1 and 2, pages
53 and 54. At every installation, the Process has proven itself as
operationally reliable and has met or surpassed all governmental
regulations regarding SO- emissions.
The Process may be applied to any flue gas containing SO-. It may be
applied to the flue gas from all fossil fuel fired boilers, nonferrous
smelters, sulfuric acid plants, and Glaus plants. The Process uses a
Wellman-Lord recycle system and yields an SO- gas suitable for conversion
to elemental sulfur as demonstrated in the Northern Indiana Public
Service Company's D. H. Mitchell Plant and Public Service Company of New
Mexico's San Juan Station using Allied Chemical technology for the
reduction process.
704
-------
PROCESS DESCRIPTION
Wellman-Lord SO Recovery^
The Wellman-Lord process consists of three major operating sections -
S02 absorption, purge treatment and SO- regeneration.
In the SO^ absorption section, residual fly ash in the flue gas is
removed by water scrubbing. SO is then removed from the flue gas by
scrubbing with a solution of soaium sulfite. The chemicals contained in
this solution remain completely dissolved throughout the absorber. Flue
gas scrubbing with a clear solution, free from suspended solids, plugging
and scaling, is a fundamental reason underlying the exceptional on-stream
reliability experienced in the commercial operations of the Wellman-Lord
process. A general schematic flow diagram of the process is shown in
Figure 1, page 4.
The purge treatment section selectively removes inactive oxidized sodium
compounds from a sidestream of the absorbing solution and converts this
material into a dry granular product which is marketed.
The third section of the Wellman-Lord process involves thermal regeneration
of the absorbing solution to release the absorbed SO- as a concentrated
gas stream and return of the reconstituted solution to the absorber.
The concentrated SO- gas may be converted to liquid SO-, sulfuric acid
or elemental sulfur. NIPSCO and PNM elected to use the Allied Chemical's
SO- Reduction Process. A general schematic flow diagram of the process
is shown in Figure 1, page 4.
Allied Chemical S0? Reduction to Sulfur
Sulfur is recovered by Allied Chemical's SO- reduction process which
consists of two principal operating sections.
In the primary reduction section, more than one-half of the entering SO-
is converted to elemental sulfur. A key feature of this section is the
effective control of chemical reactions between SO- and natural gas over
a catalyst developed by Allied Chemical for this purpose. Heat generated
by these chemical reactions is recovered and utilized to preheat the S02
gas stream entering this section.
Packed bed regenerative heaters provide a rugged and efficient means for
achieving this heat exchanger function. The process gas flow through
the regenerators is periodically reversed to alternately store and
remove heat from the packing; hence, the overall section is thermally
self-sustaining.
705
-------
Automatic control of the flow reversing cycles and other process conditions
achieves optimum performance in the system, with high sulfur recovery
efficiency and reductant utilization at all operating rates.
The gas leaving the primary reactor system is cooled in a sulfur condenser,
for condensation and recovery of sulfur product. The remaining gas,
containing proper proportions of S0~ and H^S is processed through a
Glaus conversion system for recovery of additional sulfur product. The
Glaus system off-gas is incinerated and recycled to the Wellman-Lord S0_
absorber.
706
-------
PROCESS CHEMISTRY
The Wellman-Lord Process is based on the chemistry of the sodium
sulfite/bisulfite system: flue gas containing SO- is scrubbed with a
sodium sulfite solution which absorbs S0~, converting sodium sulfite to
sodium bisulfite:
(a)
The sodium bisulfite solution is regenerated by thermal decomposition.
Application of heat simply reverses equation (a):
(b) 2 NaHS03
The S0? is recovered in a concentrated stream.
The concentrated stream of SO- gas is then reduced to high purity
elemental sulfur in the Allied Chemical Process. This conversion is
carried out in two steps. In the first step, a portion of the SO- in
the feed gas reacts with natural gas, yielding a mixture of elemental
sulfur, hydrogen sulfide, carbon dioxide and water vapor:
(c) 2CH, + 3SO- ** »» S + 2H-S + 2CO- + 2H-0
4 2 2. 2. 2.
In the second step, H-S formed in the first step reacts with the
remaining S09 yielding additional elemental sulfur and water vapor:
(d) 2H2S + S02 * ^ 3S + 2H20
The tail gas from the sulfur plant is incinerated and recycled to the
Wellman-Lord absorber.
707
-------
MOTHER
LIQUOR
EPARATOR
_
NA2C03
T Ifr.m FR
DISSOLVING TANK SO
CO
x
.STEAM
STRIPPER
'2
COMPRESSOR
TAIL GAS RECYCLE
t3 -fc
r
i
t
PRIMARY
.
FIGURE I,
WELIMAN-LORD RECOVERYMLLIED CHEMICAL S0? REDUCTION
GENERAL SCHEMATIC FLOW DIAGRAM
-------
NIPSCO
SECTION I
The first part of this paper, Section I, deals with the Unit No. 11
Boiler FGD System Northern Indiana Public Service Company's Dean H.
Mitchell Station located in Gary, Indiana.
This FGD system will be referred to as the NIPSCO FGD Plant in this
paper. By definition, the NIPSCO FGD Plant includes the flue gas
booster fan, flue gas isolation damper and the flue gas louver bypass
damper, all of which are outside battery limits as well as the primary
battery limits portion of the FGD plant consisting of the pre-scrubber,
the Wellman-Lord absorption, regeneration and purge treatment units and
the Allied Chemical S09 Reduction Unit. Byproduct storage and loading
facilities are included within battery limits.
709
-------
I
. - •
c
NoMh*m Public 8«rv>c« Co 4
U 6 Environment*' Prolvctton Ag.nc
Figure 2
GENERAL VIEW OF NIPSCO PLANT
-------
1< NIPSCO Project Background
NIPSCO and the U.S. Environmental Protection Agency entered into a cost
shared contract in June of 1972 for the design, construction and operation
of a regenerable flue gas desulfurization (FGD) demonstration plant.
The system selected for the project was a combination of the Wellman-Lord
SO Recovery Process and the Allied Chemical SO- Reduction Process. The
FGD plant was to be retrofitted to NIPSCO's 115 MW pulverized coal-fired
Unit No. 11 at the Dean H. Mitchell Station in Gary, Indiana. NIPSCO
entered into contracts with Davy Powergas Inc. for the design and con-
struction of the FGD plant and with Allied Chemical Corporation for
operation of the plant.
A successful performance test was run from August 29, 1977, through
September 16, 1977.
A one-year demonstration test period began on September 16, 1977, and
has now been extended to March 15, 1979. A decision is to be made prior
to March,15, 1979, regarding further extension of the demonstration test
period/ '
TRW under contract to the U.S. Environmental Protection Agency, has
continued to monitor and report the performance of the boiler and the
FGD plant during the demonstration test period.
This section of the paper will discuss the FGD plant operating experience
and the operating and maintenance costs from the beginning of the demon-
stration test period on September 16, 1977 through December 31, 1978.
NIPSCO is continuing to assess the Wellman^Lord/Allied Chemical option
for SO- emission control.
NIPSCO is still optimistic that longer periods of continuous operation
will be achieved during the extended demonstration test period. Through
continued operations at NIPSCO it will be possible to make a more accurate
evaluation of the economics and plant reliability.
2. NIPSCO FGD Performance Design Criteria & Results
During the plant acceptance test, from August 29, 1977, through
September 14, 1977, the FGD system performance criteria and obtained
results were:
92 MW Equivalent Test
S02 Removal
1. Required: Minimum SO- removal of 90%, measured continuously
and averaged every 2 hours for a period of 83
hours.
711
-------
2. Results:
Particulate Removal
Required:
Results:
Soda Ash Consumption
1. Required:
2. Results:
S0? removal averaged 91% over the 12-day test
period. In only two 2-hour periods (out of 144)
was the SO- removal less than 90%, and for those
periods, it averaged 88% and 89%.
Particulate emission measured once daily will not
exceed the Federal NSPS for fossil fuel fired
steam generators of 0.1 Ib/million Btu heat input.
Particulate emission averaged 0.04 Ib/million Btu,
or 40% of the maximum allowable. Of the 12 days,
tests could not be run on four days due to inclement
weather. On one day, the test data was not valid.
Average over the 12-day test period not to exceed
6.6 STPD.
Soda ash consumption determined by daily inventory
obtained from storage bin measurement (official
result) averaged 6.2 STPD, or 94% of the maximum
allowable. Consumption determined by manual
weighing the feeder output every two hours
throughout the 12-day test period averaged 5.7
STPD, or 86% of the maximum allowable.
Aggregate Cost of Steam, Electricity and Natural Gas
1. Required:
2. Results:
Sulfur Purity
1. Required:
2. Results:
Aggregate cost not to exceed $56 per hour based on
predetermined unit cost.
Hourly cost averaged $43 per hour over the 12-day
test Period, or 77% of the maximum allowable.
Minimum sulfur purity 99.5%, suitable for conversion
to quality sulfuric acid by standard production
practice.
Sulfur purity determined from a composite sample
collected over the 12-day test period was 99.9%,
easily exceeding the required purity.
712
-------
110 MW Equivalent Test
Removal
1. Required: Minimum SO- removal of 90%, measured continuously
and averaged every 2 hours for a period of 83
hours .
2. Results: S02 removal averaged 91% over the 3-1/2 day test
period. In only one 2-hour period (out of 42) was
the SO removal less than 90%, and for that period,
it averaged 89%.
Particulate Removal
1. Required: Particulate emission measured once daily will not
exceed the Federal NSPS for fossil fuel fired
steam generators of 0.1 Ib/million Btu heat input.
2. Results: Particulate emission averaged 0.04 Ib/million Btu,
or 40% of the maximum allowable. Of the 3-1/2
days , a test could not be run on one day due to
inclement weather.
Viability of the NIPSCO FGD System^ ^
During the 12-day performance period at the 92 MW equivalent rate, there
was a total of 26 hours in interruptions in the fully integrated operation
of the FGD system. Of the 26 hours, 18 were related to boiler problems
and 8 were related to problems in the FGD system. In addition, there
was a 4-hour period in which the SO- removal averaged "only" 88.5%; this
4-hour period was added to the performance test at the end of the 12-day
test period. It should be mentioned that outages in the FGD system did
not interrupt SO. removal. Furthermore, an SO- removal of 88-5% at the
NIPSCO site results in an emission well below the NSPS of 1.2 Ib/million
Btu heat input. During the acceptance test the parameters used by EPA
for judging the viability of a FGD system were: availability, 94%;
reliability, 100%; operability, 100%; utilization 'factor, 94%.
3. Operating Experience
The monthly operating hours as shown on charts l"a" through "p" on pages
15 through 30 illustrate the operating periods of NIPSCO 's Unit No. 11
and of the FGD plant. In connection with the bar graphs the following
definitions were used:
a) A solid line indicates that Unit No. 11 or the FGD plant was in
operation.
b) The definition of Unit No. 11 being in operation is: Unit synchronized
on line regardless of the megawatt load.
713
-------
c) The definition of the FGD plant being in operation is:
Receiving flue gas from Unit No. 11.
No SO- is being returned from the evaporator to the existing
Unit No. 11 stack.
d) The Unit No. 11 operating conditions required to permit FGD plant
operation are:
Unit No. 11 operating on high sulfur coal at 46 GMWE minimum
load.
Sufficient main steam available (530 psig minimum).
Sufficient demineralized make-up water available.
Unit No. 11 supplied utilities available (electricity, boiler
feed water).
4. Problems Causing FGD Outages
Chart 2 on page 31 illustrates the FGD Plant Operating Factor based on
hours of FGD operation divided by hours of generating unit operation.
Table 1 on page 32 through 34 summarizes the FGD outages as shown on the
monthly bar graphs and indicates whether the outage was attributable to
the boiler, the FGD plant or a combination of both. At the bottom of
Table 1 are listed four categories of major FGD outages. These outages
are discussed here.
5. Booster Fan
Problem;
Imbalance of the air foil flue gas booster fan has been a continuous
problem since early in the demonstration period. Operating conditions
at or below the acid dew point of the flue gas (and below the specified
design temperature of 300°F) plus an ineffective guillotine isolation
damper upstream of the booster fan contributed to an accumulation of fly
ash, water and ice in the fan housing and on the fan blades. Over a
period of several months during which the FGD system (and the booster
fan) was idle for substantial periods because of high silica problems in
Unit No. 11 make-up water, these conditions resulted in corrosion and
erosion of the air foil blades which finally required a complete reblading
of the fan. Turbine governor malfunctions have apparently been caused
by exposure of the governor to outdoor weather and dust conditions.
Solution:
The following revisions and additions are in progress or have been made
to resolve the booster fan problems:
a) The intermediate layer of Unit No. 11 air preheater elements was
removed to raise the flue gas exit temperatures.
714
-------
b) The booster fan inlet and outlet duct and the booster fan housing
were insulated.
c) The booster fan was rebladed and Inconel shields were installed on
the blade leading edges.
d) A steam soot blower was installed in the booster fan for on line
blade cleaning.
e) A booster fan drive turbine enclosure is presently under construction.
Results;
Preliminary results are very encouraging. The air preheater element
removal has resulted in an increase of approximately 30°F for the exit
flue gas temperature with no apparent adverse effects on the boiler
other than a slight decrease in boiler efficiency. The flue gas
temperature is now consistently at or above 300°F and condensation has
not occurred in the ductwork and booster fan as it did in the past. A
complete assessment of the results of the above on long term fan perfor-
mance will be made after the plant has operated throughout this winter.
6. High Silica Levels in Boiler
Problem:
In October 1977, routine Unit No. 11 boiler water chemistry tests indicated
high levels of silica. The first attempts to solve the problem focused
on inspection of Unit No. 11 boiler and condenser for leaks. Frequent
or continuous blowdown of the boiler and reduced steam pressures made
the steam supply to the FGD plant so unreliable that it precluded
integrated FGD operation. After extensive investigation including
ultrasonic condenser tube testing, the condenser was found not to be the
cause of the contamination. The causes of the silica build-up appear to
have been a boiler chemistry upset due to high condensate make-up require-
ments imposed on the boiler by the FGD plant requirements. This was
compounded by the continuous boiler blowdown and lack of continuous
silica monitoring on a portable demineralizer being used to supplement
the station demineralizer to supply Unit No. 11 and FGD condensate
make-up demand,.
The need for using a portable unit resulted from failure of a reverse
osmosis unit, installed by NIPSCO, to operate successfully.
The design of the FGD system was based on the use of filtered lake water
for use as flushing water to the packing gland on the process pumps.
During plant start up it was found that the quantity of fine silt in the
lake water made it impossible to adequately filter the water for packing
gland flushing. As an expedient the use of condensate as a flushing
fluid was implemented and the resulting consumption of condensate in the
FGD system became substantially greater than initial design.
715
-------
Since the water treating facilities (station demineralizer and reverse
osmosis unit) available at the NIPSCO plant were expected to have adequate
capacity, this increase in condensate consumption was not seen as a
problem at the time except from the cost standpoint. The unit cost for
treating water with the portable demineralizer equipment is very high.
Solution:
A continuous silica analyzer was installed on the portable demineralizer
outlet and the silica content of the treated make-up water was limited
to 10 parts per billion. Condensate from the FGD system is diverted to
waste at a low pH reading of 6.5 or a high reading of 9.5 and at a
conductivity reading of 5.5 micromohs/cm and higher. The FGD plant
condensate return diverter system was changed from manual to an automatic
reset mode. If the condensate is diverted to waste because of high
conductivity, low or high pH, it will automatically return to the Unit
No. 11 condenser when the pH and conductivity readings are again within
acceptable limits. As part of the solution to this problem, a compre-
hensive program was undertaken in the FGD plant to reduce condensate
consumption and losses.
Results:
The above modifications have been effective in controlling Unit No. 11
boiler silica. Combined Unit No. 11 and FGD condensate make-up is now
generally in the 40 to 50 gpm range with an occasional excursion to
higher volumes if the return condensate has been diverted to waste.
7- Guillotine Isolation Damper
Problem:
A top entry guillotine damper was installed between the outlet of the
existing Unit No. 11 boiler induced draft fans and the FGD plant booster
fan inlet. During operation with the damper open a build-up of fly ash
occurred in the bottom and lower side channels of the damper frame.
When the damper was closed it would strike the fly ash build-up and the
bottom seals of the damper sustained mechanical damage. Considerable
corrosion damage has also been experienced with the damper.
Solution:
Considerable work was done on maintenance and modification of the damper.
A bottom channel purge air system was installed and modified. Damper
seals were replaced with seals made of a more corrosion resistant alloy.
716
-------
Results:
The fly ash build-up continued to occur at the bottom of the damper. A
decision was made not to invest any additional money in the damper.
Manual slide gate dampers were installed in each of the booster fan
inlet housings and will be used to isolate the fan when necessary for
maintenance.
N°te: During the recent period, October through December 1978, the
guillotine isolation damper has operated successfully but has not
provided total shut-off capability, probably caused by fly ash damaged
seals. The recently completed modifications to the Unit No. 11 air
preheaters and installation of thermal insulation on the flue gas duct
work have allowed the flue gas temperatures to be maintained above the
dew point and alleviated some of the problems with the guillotine
isolation damper as well as the flue gas booster fan.
8. Coal Quality
Problem;
Wet, poor quality coal caused erratic Unit No. 11 boiler operation. As
a result, steam pressure and flow to the FGD plant fluctuated and prevented
sustained integrated FGD operation.
Solution:
NIPSCO's coal procurement department negotiated with suppliers to insure
a consistent supply of a more acceptable quality coal.
Results:
The coal quality has improved and is no longer considered a problem.
9. Problems Not Causing FGD Outages
There were some problems experienced during the early plant operations
which are mentioned here. These situations did not cause FGD Plant
outages.
Absorber Tray Leakage
Sodium balances and analyses of the fly ash sump discharge indicated
that there were solution losses occurring in the absorber. Inspection
of the absorber lower collector tray showed that solution was splashing
over the chimney free-boards and leakage was also occurring between the
tray perimeter and the absorber wall. The chimney free-board height has
now been raised and the tray perimeter recaulked and sealed with rubber.
717
-------
Evaporator Circulating Pump Driver
The evaporator slurry circulating pump was originally equipped with a
steam turbine drive, utilizing main steam supplied by Unit No. 11 boiler
to the FDG plant, at 550 psig. Exhaust steam from the drive turbine was
utilized as part of the heating steam requirements for the evaporator.
During the demonstration period the main steam supply to the FGD plant
was partially or totally interrupted on many occasions resulting in slow
down or stopping of the circulating pump. To avoid settling of solids
and plugging of the evaporator it was necessary to dilute the solution
and dump the evaporator slurry into a holding tank. After the steam
supply was re-established, two days would be required to refill and
re-establish solids content before evaporator operation could be
resumed.
To avoid these lengthy interruptions to the FGD operation an electric
motor drive was installed on the pump in September 1978.
Purge Dryer Capacity
System sulfate inventories and dry sodium salt production have recently
indicated that the purge dryer possibly was not operating at rated
capacity. A test program has been undertaken by Davy Powergas.
Mechanical modifications are in progress aimed at increasing the purge
dryer capacity.
718
-------
Chart 1 ( a )
NIPSCO FGD PLANT
OPERATING HOURS
SEPTEMBER 1977
DATE
FGD PLANT
Acceptance Test in
Progress.
UNIT NO, 11
DATE
10
10
11
11
12
13
14
15
360 HRS.
Variable Main steam
12
13
14
15
16
17
Pressure From Unit 11.
16
17
18
19
20
21
22
67 HRS.
Below 46 GMW Operation
On Unit 11 Due to Wet
OFF 7:42P
ON 8;13P
Wet Coal -
Mills Plugged.
18
19
20
21
22
23
OFF 9:44P
ON10i.29P
OPE12:50P
ON 3;19P
Control Valve Problems.
23
24
S
OFF11:IIP
ON 6;43P
24
25
25
26
OFF 6:54P
ON 7..03P
26
27
OFF 7:37P H Blr. Feed Pump Trip Du<
ON 9:30P ^ to water From 5O2
27
28
28
29
FeedwtrRel. Vlv.
29
30
30
719
-------
Chart 1 ( b )
NIPSCO FGD PLANT
OPERATING HOURS
OCTOBER 1977
DATE
.FGD PLANT
UNIT NO, 11
DATE
Kept Shutdown Through 0 :t
5 To Run Heat Balance
Flue Gas Flow Rate Test
9n Unit 11
start Up or Absorber
/Evap. System
instrument Problems in
SO3 Reduction System.
Incinerator Repairs, Tupe
Loako in No. 1 Sulfur
Condenser.
10
11
10
11
12
12
13
14
15
16
17
18
132 HRS.
Bearing failure on
steam Turbine Drive
13
14
15
16
17
18
19
19
20
for Evaporator circul-
ating pump
20
21
22
Evaporator Started and
Held In Stand-By
21
22
23
Unit 11 Boiler Tube
Repair
23
24
24
25
25
26
26
27
27
28
Absorber and Evaporator
Operating.
Boiler Tube Leak &
Check for Condenser
Leaks.
28
29
31
8:24 PM
7:46 AM
29
30
31
720
-------
Chart l(c)
NIPSCO FGD PLANT
OPERATING HOURS
NOVEMBER 1977
DATE
FGD PLANT
8 HRS.
Booster Fan Cleaning
and Balancing.
22 HRS.
UNIT NO, 11
DATE
10
11
12
13
14
15
16
17
18
10
11
12
13
14
15
16
17
18
19
20
21
22
428 HRS
Side
19
20
21
22
23
8:14 PM
Inspect Steam
of Condenser
23
24
24
25
25
26
Guillotine Damper
Jammed In Part
7:54 AM
26
27
Open Position.
Gasket, in Solution —
Line from Evaporator
to Circ,
Out.
Pump Blew
27
28
29
-30
30
721
-------
Chart 1 (d )
NIPSCO FGD PLANT
OPERATING HOURS
DATE
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
DECEMBE
FGD PLANT
FGD Plant Held In
Standby Pending
Resolution of High
, Silica Problem on
Unit 11 Boiler.
NOTE : "After ' extensive
investigation, it was
determined the high
Silica Problem on the
Unit 11 Boiler was
caused by high
condensate make-up
requirements imposed on
Boiler by the FGD
System. Normal Boiler
make-up without FGD is
approx 15 GPM. Boiler
operation is normally
can increase to 150 GPM
during FGD start up or
upset operating modes.
Hiah silica in the make-
up water to Unit 11
and subsequently main--
tained under control
bv NIPSCO through;
a. installation of a
' • cT>irtxrro<:>ttS""sil±ca
analyzer on the
treated make-up
wa1-f»r l-o permit
adequate control of
1 1-, -I .,, ? , T1-J.1 T TJL1
operation .
b. Installation of
additional controls
on the return conde:
o U-LVfci. I — my m4UjL
ment to avoid unnec
, essary and prolonge
di ve>T-cHr>n nf r-ppHet1
to waste.
c. A comprehensive
program was under-
taken to reduce the
and losses.
* 1977
UNIT NO, 11
1
1
1
12-58 AM B SO2 Guillotine Damper
Repair and condenser
Inspection and clinker
1 ' 10 AM mjjl
10 • 51 PM iB RePair Air Circuit
3:22 AM mm BreaKer on ffll
Disconnects.
1
1
10:50 PM • Repair #3 Precip.
Field and Check
Condenser for Leaks.
12:48 AM m
7 : 58 m ii L°SS "°f Contro1 Air
m
m
m
|||3.D. For Coal Millrepair
1
I
1
]^o jj
j|
1
1
•
DATE
i
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
722
-------
Chart Me)
NIPSCO FGD PLANT
OPERATING HOURS
DAIE
i
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
JANUAF
FGD PLANT
FGD Plant held in
olution of High Silica
Problem on Unit 11
Boiler
FGD Plant isolated
Supply, Boiler Feed-
water Supply, and
Condensate Return.
FGD Plant Steam
to Unit 11 Main
• • Steam. SO2 comp-
ressor gasket leak.
Cleaning ice and
booster ,fan.
*Y 1978
UNIT NO, 11
1
1
1
1
1
8:45 PM m Repair Precip. Fields
and check condenser
for leaks i
8:40 AM ^
11
1
1
1
H Condenser Tested for
1
1!
jji
ji i
1
1
9:30 PM iH Repair Precip. Field
and 11-3 Mill Hot
Air Damper
3:31 AM m
m
1
ii
1
1
1
1
1
DATE
i
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
723
-------
Chart 1 ( f )
NIPSCO FGD PLANT
OPERATING HOURS
DATE
i
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
FEBRUAf
FGD PLANT
Booster Fan Cleaning
Complete Additional
Drains Installed in
Fan Housing.
FGD Start-UP Held
"Dcar^ -i t-»rr TT-n i' +• it 1 1
Boiler Tube Repair
Clean Duct Downstream
Damper Push Rods broke
when attempt was made
to open Damper.
Repair started on
Guillotine Damper.
Damper Repair Complete
at 2:28 A.M.
Booster Fan started
on slow Roll. Shaft
Air Line to Governor
f r"7»n . ^g" vpha 1 anced .
Flue Gas admitted to
Absorber. Booster Fan
Oil Cooler frozen -
lost Bearinq Oil.
- • Refilling Oil System
Booster Fan and Absorber
in Operation .
Evaporator Circulating
Pump down for Repacking .
Frozen SO2 Superheater
Line. i
1
Evaporator Circulating
Pump Packina Problem.
i r
?Y 1978
UNIT NO, 11
8-.4b AM ™» Trip on High Furnace
9:14 PM m T»T-M«,I™ **p»ir
Boiler Tube Leak
2:50 AM •
1
1
10:40 PM B Stop Valve Repair
Stop Valve Work Comp-
Finish SO2 Damper Repair
10:14 AM „ .
11:28 PM 11!
^:47 AM HH Turbine Pre— emereiencv
3:08 AM Hi „ m ^ __..__ j
3:24 AM HH . . .
||| Unit to Trip.
Ill
11
1
1 Limited to 55 GMW
Bearing Problem.
s|! Switched to Low
111 Sulfur Coal Due
Ito Feeding Problems
W1L11 ri-Lgll SUllUJC Coal .
I
DATE
i
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
724
-------
Chart 1( g )
NIPSCO FGD PLANT
OPERATING HOURS
MARCH 1978
DATE
15
FGD PLANT
Booster Fan Vibration
Booster Fan and Absorbed
Started Up. Main ate
pressure to FGD fluct-
uating down to
Waiting for SOo
Feed Problems
Load below 46 GMWE
to Build-UP so Reductio i
Precipitator Reduced
Section can he
Efficiency
Two coal mills
Started Sulfur
Production.
Main Steam Pressure to
own to 300
Coal Mill problem in AM
'Unit shutdown due to
ad below 46 GMWE -
10:34 PM
low main, steam pressure
Guillotine Isolation
damper did not close
UNIT NO, 11
®0ne coal mill in operati
Precipitator wire repair.
Repair ash hopper.
DATE
J.5
16
Q
ir
16
17
completely
Balance Booster Fan
10:04 PM
Trouble Shoot Turbine
Pre-emergency governor
17
18
Start Up Booster Fan
and Absorber
18
19
20
Guillotine isolation
- jammed. Opened
after manual assist.
19
20
21
21
22
2:35 PM
6:05 PM
Guillotine Damper
would not close.
22
23
24
25
Booster Fan Balanced
Again. Absorber Started
Up.
Unit shutdown to
•balance booster
fan.
23
24
25
26
Instrument Problems
On Low Sulfur Coal
.ue to Feeding prob-
26
27
27
28
29
30
31
Booster Fan Vibration
Fan on Slow roll
loal.
pending arrival of
• service engineer.
an Vibration
iblem. Fan Shutdown
pending Unit 11 outage.
Clean Guillotine
damper bottc
8:51 PM
& Balance
West I.D. Fan.
ipair
28
29
30
31
725
-------
Chart 1 ( h )
NIPSCO FGD PLANT
OPERATING HOURS
APRIL 1978
DATE
i
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
FGD PLANT
Service Engineer to
advise Re: Booster fan
balancing.
Booster fan blades re-
placed due to corrosion
• -and erosion
blading.
or existing^
Absorber lower collector
tray vas recaulked arid
loose rubber was re-
moved, t
Booster fan reblading
complete Fan ready for
balancing.
Balanced booster fan.
Guillotine isolation
damper inoperable .
UNIT NO, 11
Repair & balance West
T.r>_ Fsn Tnspt»r!l-
guillotine isolation
A.n awi ass* damper and duc-L.
*"livl SSSw
1
11
jj
1
ll
jj
1
I
ll
10:25 PM iH Repair I.D. fan
blading and repair
No. 3 precipitator
field.
v
1:30 AM m
._. Malfunction of
11:45 AM m automatic controls.
i
o i
• w|
•
DATE
i
2
3
-4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
726
-------
Chart 1(1)
NIPSCO FGD PLANT
OPERATING HOURS
DATE
1
•— — ^— ^_ .
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
MAY
FGD PLANT
Plant on stand-by awaiti
— openina and
repair of guillotine
ibulati-on damper
Guillotine isolation
damper blocked in fulloj
position
Booster tan and absorber
starfprl up.
Evaporation started but
SO-> flow rock too low
for operation of SO2
''-
7'-30 PM started!"0^1011
1
|1| Boiler teedwater booste:
McumD failure.' No BFW
T ?n PM llPupply avallatle f°r
U the FGD Svstem.
SO 2 levels too low
for operation of reduc-
tion sections
Alternate boiler feed-
water tie-in complete .
12 : 00 N , BFW supply now at
— fl|f' temperature substan—
11! tially below required
HHp ^^o^r .
11
J!
1
Jl
jj|
11:00 PM HReduct:'-on area
shutdown due to
hicth pressure drop
through coalescer.
1978
UNIT NO, 11
^ 1
•
fl Shutdown for repair
8:37 PM H0f SO2 pi »„+ T,4n«»
tine isolation damper.
Two coal mills out of
service.
10:35 AM ^
m
It
• Changed to low sultur
^^ coal because of wet H.S.
Ill coal .
ill Resumed high sulfur coa.
Ill feed to bunkers .
5
n
ill
K
»
•
II
K
3
fe
i
DATE
1
2
3
4
5
6
7
8
9
10
'11
12
13
14
15
16
1.7
18
19
20
21
22
23
24
25
26
27
28
29
30
31
727
-------
Chart 1 ( J )
NIPSCO FGD PLANT
OPERATING HOURS
DATE
i
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
JUNE
FGD PLANT
Continuing to repair
and clean coalescer
Main steam to FGD shut c
.... ,jjue (>o pressure reducing
valve problem.
Erratic pressure on mair
steam supply to FGD plai
Booster ran turoine
bearing failure.
Booster fan turbine
repair in proaress.
Started removal of
catalyst in A&B claus
converters .
Replacement complete
on 6-23 in PM.
Booster fan turbine
repair complete, ran-
out of balance when
started up. Cleaning
proceed until Unit 11
is shutdown because
quillotine isolation
damper is inoperable.
1978
UNIT NO, 11
1
own
11
t •
j|
1
1
1
j
1
1
I
1
III
1
1
II
1
I
1
1
Jj
!
!
I
I
i
1
ii
DATE
i
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
728
-------
Chart 1 ( k )
NIPSCaFGD PLANT
OPERATING HOURS
DAlb
i
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
JULY
FGD PLANT
Waiting for Unit 11
— Outacre to clean and
balance booster fan.
Balanced Booster Fan.
Booster fan and absorbei
started up. 'Evaporator
starting up.
Booster fan at minimum
speed.- Accessing lube
oil leak on fan bearing
Boaster fan neia at
minimum speed awaitina
delivery of oil seals.
Repaired oil leak on
booster fan outboard
bearing.
Repair main steam press
reducing valve.
Booster fan trip.
Pressure of main 'Steam
supply to FGD fluctuatii
down to 280PS1G
Repair main steam
pressure reducing yalve.
Booster fan turbine '
_. governor repairs .
11: 00AM Integrated operation
started..
1978
UNIT NO, 11
H
m
m
m
1
11 • 08 AM Induced draft fan
repair, booster fan
cleaning, guillotine
damper repair , pre—
cipitator repair.
1 : 40AM ip
1
j
n
II
1
1
I
1
1
1
ij
jj
1
jj
q • .
1
1
1
jj
1
DATE
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
729
-------
Chart 1(1)
NIPSCO FGD PLANT
OPERATING HOURS
AUGUST 1978
DATE
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
FGD PLANT
Integrated Operation.
7:20 PM
9:25 PM
luction Area
Due to power
failure caused by
Electrical- Storm.-
UNIT NO, 11
730
-------
Chart i(m)
NIPSCO FGD PLANT
OPERATING HOURS
DAIE
i
-a
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
SEPTEMBER 1978
FGD PLANT
I
•
fi
1
H
iii
— i - wySw
I
] 0 : 00 AM 5! Reduction area shut-
12:15 PM |p down due to SO_ comores-
lU sor malfunction.
8:55 AM ®*» FGD Plant shutdown
because of boiler
shutdown. Major
tasks are:
Duct insulation.
Booster fan steam
soot blower.
Manual Booster Fan
Dampers .
Electric Drive on
Evaporator Circulatincr
Pump. Booster Fan
Turbine Enclosure.
Rubber Repairs on
Absorber. Sulfur
Condenser Tiibe Maint-
enance.
UNIT NO, 11
1
u
1
1
1
1
1
j
1
1
10 : 41 PM iBi Scheduled shutdown
for general maint-
enance .
Included in the work
.to be performed is
overhaul of the Unit 11
Ail PiehedLer seals to
reduce leakage and re-
moval of intermediate
baskets to effect an
increase in the outlet
New pumps for supply of
BFW to the FGD svstem
could not be procured
on this turnaround.
DATE
l
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
731
-------
Chart 1 ( n )
NIPSCO FGD PLANT
OPERATING HOURS
OCTOBER 1978
DATE
i
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
FGD PLANT
Annual scheduled
boiler maintenance
FGD kept down for
boiler baseline test
by TRW.
1
Evaporator start-up
absorber in operation
when booster fan was
10: 30AM Booster fan balanced.
Jill Louver bypass dampers
H lammed.
|g Booster fan oil
H cooler problem.
HH SO2 reduction
start— up
I
1
1
1
111
10: 45AM ™ Booster tan governor
3 : 00PM ill problem.
jj
1
1
1
•
UNIT NO, 11
Annual scheduled
boiler maintenance
continued from last
5:13 PM m
HH LeaK at weld on
1 : 16 PM TO .
valve
9:57 AM Between steam drum
HI and pressure trans-
mitter
Jj
1
B
1
1
1
•
H
I
1
1
1
jj
I
11
1
1
ill
1
1
1
1
•
DATE
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
732
-------
Chart l(o)
NIPSCO FGD PLANT
OPERATING HOURS
NOVEMBER 1978
DATE
10
11
12
13
14
15
16
17
18
19
20
FGD PLANT
7:00 AM
11:30 AM
oster blower snut-
for repair of
leaking flue gas duct.
UNIT NO, 11
DATE
10
11
12
13
14
15
16
17
18
19
20
21
22
21
22
23
23
24
2:15 PM
3:20 PM
Repair drain valve in
Reduction area
24
25
25
26
26
27
27
lownmReplace modulating serve
10:25 PMJiion turbine intercept vallve'
28
reduction area snut
10:25PM HI due to turbine outage.
•SL
1:25 AM
12:35PM
iReduction area down due
steam pressi
.1:04 AM
29
30
733
-------
Chart 1 ( P)
NIPSCO FGD PLANT
OPERATING HOURS
DECEMBER 1978
DATE
FGD PLANT
1:30 PM
3:45 PM
Repair evaporator
UNIT NO, 11
DATE
heat exchanger tube
5
5:40 AM
6:00 AM
Reduction section down
to low steam pressure.
10:59 PM
Unit 11 shutdown.
10:59 PM
Replace quick closing
10
11
12
13
14
2:00 AM
10:50 AM
2:40 PM
12:49 AM
8:35 AM
11:30 AM
9:12 AM
leduction section down
to low steam pressure.
me
Servo on turbine
it valve.
Reduction section down,
Evaporator fouling due
scaling. Cleaning heat
bo
exchanger tubes.
8:20 PM
Clean Air Preheaters.
10
11
12
13
14
15
15
16
16
17
18
19
20.
21
22
23
24
25
26
27
28
29
30
31
8:05 AM
11:45 PM
1:06 PM
6:54 AM
Leak in 1st sulfur
condenser.
Tube leaks in evaporate
heater. •
Absorber on at 8:00 AM
17
18
19
20
21
22
23
24
25
26
27
28
_29_
30
31
734
-------
O.F.
MRS. FGD INTEGRATED OPERATION
MRS. GENERATOR ON LINE
XIOO
S 0 N
16*30
1977
MONTHS
CHART 2
FGD PLANT OPERATING FACTOR (O.F.)
NIPSCO
735
-------
TABLE 1
Cause
SUMMARY OF FGD OUTAGE CAUSES
Attributable to Boiler (B),
FGD plant (FGD) or combina-
tion of both (C)
1. Booster fan cleaning
balancing and reblading
2. High boiler silica levels
3. Booster fan guillotine isolation
damper.
4. Booster fan bearing oil leak,
booster fan turbine governor
repairs, and turbine bearing
repair.
5. Scheduled boiler maintenance
6. Wet, poor quality coal -
erratic boiler operation.
7- FGD plant main steam pressure
reducing valve malfunction.
8. Unit No. 11 boiler tube leaks
9. Unit No. 11 electrostatic
precipitator malfunction.
10. Evaporator circulating
pump repacking.
11. Boiler baseline test
12. Evaporator heat exchanger tube
sealing and tube leaks
13. Heat balance and flue gas flow
rate tests.
14. Bearing failure - evaporator
circulating pump turbine drive.
15. Induced draft fan imbalance
FGD
(*")
C1 }
FGD
FGD
Total Number
of Days in
Outage
67
53
32
26
B
B
B
B
B
FGD
C
FGD
B
FGD
B
22
20
13
12
7
7
6
6
6
5
5
736
-------
Cause
Attributable to Boiler (B),
FGD plant (FGD) or combina-
Total Number
of Days in
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
Coalescer pluggage.
Gasket failure - evaporator
solution line.
Turbine-generator stop valve
repair.
Air preheater cleaning
Various instrument problems
Tailgas incinerator malfunction.
Sulfur condenser tube leaks.
S02 compressor gasket leak.
Process water booster pump failure
Steam leak at drum pressure
transmitter line.
\**s
FGD
FGD
B
B
FGD
FGD
FGD
FGD
FGD
,B
5
4
4
3
3
2
2
2
2
1
RECAPITULATION
Total days in period September 16, 1977 through
December 31, 1978.
Total days integrated FGD operation.
Total days FGD outage.
Total days FGD attributable outage.
Total days boiler-turbine attributable outage.
Total days combination attributable outage.
Total days boiler and turbine operated (on line)
Total days boiler outage (off line)
Longest continuous integrated FGD operation period.
472
157
315
163
80
59
424
48
43 days (disregarding
2-2 hour interruptions
in reduction section).
737
-------
MAJOR FGD OUTAGES
Booster fan related
High silica levels in boiler
Guillotine damper related
Coal quality related
Percent of total outage
Days % of Total Outage
93 30
53 17
32 10
20 6
63
(*) The cause of high silica in the boiler feed water was a silica
breakthrough in the portable demineralizer unit that was being used
to supplement the increased water demand. The additional usage was
compounded by increased boiler blowdown. This was corrected by
setting more stringent silica limits and closer monitoring equipment
of the portable demineralizer unit. It was also followed up by a
program to reduce condensate consumption and losses including
installation of automatic reset controls on the return condensate
diversion system. (See pages 11 and 12, item number 6.)
738
-------
10• FGD System Imports and Exports
Table 2 on page 740 lists the monthly Imports and Exports of the FGD
system covering the time period of September 1977, through December
1978.
739
-------
SEPT. .
I6th-30th
IMPORTS 1977
ELECTRIC POWER 11000 KWH's) 223
STEAM (1000 IDS.) C) 21301
NATURAL GAS (1000 SCF) 2487
CONDENSATE MAKE UP (1000) 4918
POTABLE WATER (1000 Ibs. I 2054
BOILER FEEDWATER (1000 Ibs. I 584
SODA ASH (NET TONS 1 0
FLUEGAS (AVE. PPMV S02I 2020
COOLING WATER
EXPORTS
SULFUR (LONG TONS 1 0
SODIUM SULFATE PURGE 0
(NET TONS 1
CONDENSATE RETURN (1000 Ibs. ) 10200
FLUE GAS IAVE PPW S02) 188
aY ASH PURGE
*- OTHER DATA
FGD PLANT INTEGRATED OPERATION 67
(UDC \
UNIT 1 1 OPERATION-ON LINE 371
IHRS)
AVERAGE S02 REMOVAL EFFI- 90.
CIENCY (%)
UNIT II NET GENERATION lOOOKWH's 18931
AVE. NET MWE 51
OCT.
417
29133
4074
14067
3568
2288
67
2287
NOV.
452
34302
5526
16443
2384
1749
171
2373
DEC.
434
28428
3333
16533
4791
1730
97
-
JAN.
1978
412
15544
3307
12495
1299
1415
0
FEB.
368
17142
2707
6561
778
955
12
MAR.
511
34420
3732
14661
2639
1750
235
2016
APRIL
253
12809
1885
262
1252
1529
15
-
DESIGN FLOW RATE
102
14
30888
230
285
56
18157
228
0
12
9787
-
0
29
4486
0
28
764
135
41
24642
230
0
14
10554
MAY
472
35952
3916
8422
5111
2404
251
1823
4, 545 GPM
191
44
26126
205
JUNE
412
21551
1807
6084
4525
2288
53
2243
0
45
14337
288
JULY
426
28422
2052
7240
4488
2957
108
1797
18
35
24140
198
AUG.
536
47683
7863
14094
5665
2539
282
2182
526
65
39323
225
SEPT.
- 377
17513
3250
6633
5191
2484
103
2221
171
33
17073
229
OCT.
385
25145
3671
6944
5554
2417
105
2283
227
36
35042
250
NOV.
517
40888
6822
10907
6055
2392
83
2190
513
68
43256
217
DEC.
497
37511
5106
14409
3178
2555
83
2044
305
9
36145
220
!6to50GPM, 1% SOLI OS OR LESS
131
709
7 89.9
49549
70
428
660
90.4
44534
67
0
665
89.8
42101
63
0
654
50498
77
0
478
40208
84
215
690
88.5
43270
63
0
544
-
37168
68
263
682
88.3
.48076
70
3
720
87.1
49420
69
30
658
88.8
39891
61
742
774
89.7
51880
70
271
287
89.6
19719
69
324
634
89.0
50878
80
713
717
90.1
49879
70
479
682
89.1
48199
71
('} STEAM/CON DENS ATE IMBALANCES ARE DUE TO- I METERING INACCURACIES.
2. EMERGENCY STEAM NOT METERED UNTIL 11-78.
3. PORTION OF CONDENSATE MAKE UP TO UNIT II BOILER IS ESTIMATED.
TABLE 11
NORTHERN INDIANA PUBLIC SERVICE COMPANY
DEAN H. MITCHELL STATION - UNIT NUMBER II
FGD PLANT IMPORTS AND EXPORTS
-------
!!• FGD System Operating and Maintenance Costs
Table 3 on page 742 is a summary of FGD Operating, Maintenance and Improve-
ment Costs for September 16, 1977, through December 31, 1978.
741
-------
TABLE 3
OPERATING & MAINTENANCE COSTS
NIPSCO UNIT NO. 11 FGD PLANT
Operating, Maintenance and Improvement Costs are listed from the period of
September 16, 1977, through December 31, 1978.
Operation and Maintenance - Offsites Facilities $ 520,700 (a)
(including booster blower, .flue gas ductwork
and dampers,utilities system)
Operation and Maintenance - FGD Process 3,309,200 (b)
(Includes by-products storage and loading,
raw materials unloading and storage)
UTILITIES
Steam @ $2.00/1000 Ib 895,500
Demineralized water 531,900 (c)
September 1977 through February 1978 ($0.03/gallon)
March 1978 through December 1978 ($0.0125/gallon)
Electric power @ $0.024/kWh 160,600
Natural gas @ $1.9812/million Btu 121,900
Total Utilities $1,709,900
Total FGD Costs before by-product credit $5,539,800
Credit for sulfur and sodium sulfate (97,000)
Total FGD Operating, Maintenance and Improvement $5,442,800
Costs after by-product credit
(a) Includes $20,000 for installation of manual slide gates at booster
fan inlets.
(b) Includes approximately $150,000 for one time mechanical system
modifications.
(c) Water costs were abnormally high due to use of demineralized water
as condensate for flushing pump seals at 0.03 per gallon versus use
of their own process water at an estimated cost of $.12 per one
thousand gallons. Approximately 75% of the condensate consumption
in the FGD system is attributable to pump seal flushing.
742
-------
NIPSCO
STATION
MANAGER
TECHNICAL
SUPERVISOR
ALLIED
PLANT
MANAGER
OPERATING
SUPERVISOR
PLANT
CONTROLLER
MAINTENANCE
SUPERVISOR
TECHNICAL
ANALYSTS
OPERATING
FOREMAN
MAINTENANCE
MEN
OPERATORS
FIGURE 3.
NIPSCO EMISSION CONTROL SYSTEM
FGD PLANT OPERATING ORGANIZATION
-------
PNM
Section II
The second part of this paper, Section II, pertains to the FGD facilities
of the Public Service Company of New Mexico's San Juan Station located
in Waterflow, New Mexico (near Farmington, New Mexico). The FGD facilities
will be referred to as the PNM plant.
744
-------
Figure 4
GENERAL VIEW OF PNM PLANT
-------
1. PNM Project Background
Public Service Company of New Mexico (PNM) selected the Wellman-Lord/Allied
Chemical FGD processes early in 1974 from four different systems that
were being considered. The process compared favorably both economically
and technically with the lime-limestone, double alkali, and with a dry
char adsorber. The Wellman-Lord/Allied Chemical process was also considered
advantageous because the end product, elemental sulfur, could be marketed.
The initial concept was to regenerate the purge salt, sodium sulfate.
In the meantime, a contract had been obtained to sell this material.
The elimination of the sulfate regeneration process reduced the capital
requirements for Units No. 1 and 2 by about $4,000,000.
A major advantage of the Wellman-Lord process is using a clear scrubbing
solution that prevents absorber pluggage . A more soluble substance is
produced in the scrubbing liquor as SO^ is absorbed. The materials
handling is also much simpler since relatively small volumes of materials
enter and exit the system compared to the calcium based system.
The main disadvantage of the Wellman-Lord process, is that it is somewhat
more complex than the lime-limestone systems because the solutions are
regenerated.
2. PNM Design Criteria
The Wellman-Lord system for Units No. 1 and No. 2 was designed to remove
90% of the SO- from the flue gas when firing coal ranging in sulfur from
0.59% to 1.3% by weight with an average of 0.8%. Relatively high pressure
drop prescrubbers were specified for the system because of New Mexico's
stringent regulation of a fine particulate emission and to provide some
back up for the electrostatic precipitators. The purge from the pre-
scrubber is sent to the plant waste water system where it is treated and
then recycled.
At PNM's request four scrubber absorber modules were installed on each
power plant unit each sized to handle one-third of the total gas flow.
Therefore, the plant has one complete spare scrubber-absorber available.
This permits a maintenance program to be established whereby absorbers
can be rotated in and out of service for routine and preventative
maintenance purposes.
The chemical plant has two double effect evaporators which provide steam
conservation (the overhead vapors from the first effect are utilized as
the heat source in the second effect). For reliability each evaporator
is connected to steam and offgas compressor manifolds so any one evaporator
can be taken out of service for maintenance without affecting the operation
of the three absorber units.
746
-------
The purge treatment plant consists of three low temperature crystal-
lizers where sodium sulfate is precipitated in the decahydrate form.
The crystallizers are followed by a melt tank and evaporator. The
evaporator is similar to, but smaller than, the main evaporators. Water
is driven off and the sulfate purge is centrifuged, then dried in a
flash dryer. The entire purge treatment plant was specified to obtain a
concentration of 70% sodium sulfate and 30% sodium sulfite in the dried
purge salt. The actual sulfate content has been very high; purities
have been achieved in the area of 90% sodium sulfate. Residual moisture
in the dried purge salt has been less than 1.0%.
Two identical Allied Chemical SCL reduction trains were installed as
part of Units 1 and 2 FGD systems. Each of the trains has a design
capacity of more than 50% of the total FGD system capacity based on the
use of low grade coal (1.3% sulfur) in the boilers. A requirement for
two (2) S02 reduction trains was specified by PNM with the objective of
achieving an FGD system with essentially a 100% on-stream reliability.
The system being designed for the Unit Number 3 and 4 power plants will
be somewhat similar with the following exceptions. The prescrubbers on
Unit Number 3 and 4 system will have a lower pressure drop for energy
considerations. The five stage tray absorbers also function quite well
for residual particulate removal; sulfate purge quality will not be
degraded since the fly ash is filtered out of the solution.
In the interest of economy the absorbers have been designed for 4 units
in operation per boiler when burning low grade coal; i.e., this removes
the one module spare as on Units 1 and 2. The coal being used at
San Juan Station has rarely exceeded 0.95% sulfur for long durations of
time hence this should not decrease the plant operability and maintenance
capability.
A sulfuric acid unit will be installed in the FGD system for Units 3 and
4. The sulfuric acid plant design capacity will be based on low grade
coal (470 tons per day as 100% sulfuric acid). The rationale for selection
of sulfuric acid rather than sulfur as the by-product is presented in a
later portion of this paper.
3. Plant Operations
General - The plant scrubber operation is controlled by two separate PNM
groups; the power generation plant personnel are responsible for the
booster blowers, scrubbers, absorbers, fly ash filters, and waste water
treatment. The chemical plant operations personnel are responsible for
supplying solutions for the absorbers, regeneration of SCL from the
absorbing solutions, operation of the sulfur dioxide reduction unit,
purge treatment, and loading and shipping of the products. Maintenance
people at San Juan are organized by zones. Figures 5 and 6 illustrate
the organization structure for operations and maintenance.
747
-------
SAN JUAN
AREA MANAGER
CHEMICAL PLANT
SUPERINTENDENT
I
-e-
00
OPERATIONS
SUPERINTENDENT
1
TECHNICAL
SUPERINTENDENT
OPERATIONS SUPERINTENDENT
GENERAL PLANT
SHIFT
SUPERVISORS
SHIFT
SUPERVISORS
OPERATORS
FIGURE 5
OPERATIONS
PNM OPERATIONS ORGANIZATION
-------
SAN JUAN
AREA MANAGER
MAINTENANCE
SUPERINTENDENT
I
UNIT NUMBER #1
1
INCLUDES
SCRUBBERS
AND ABSORBERS
UNIT NUMBER #2
INCLUDES
SCRUBBERS
AND ABSORBERS
CHEMICAL PLANT
AND WASTE WATER
FIGURE 6-
PNM MAINTENANCE ORGANIZATION
-------
Personnel - There are approximately 100 personnel involved in support of
the FGD process.
Scrubber operators are in their fourth year apprenticeship and are near
the end of their formal training as Journeymen Operators for the power
generation plant. All operators are rotated through this position as
part of their training. First and second year apprentices are assigned
the field activities for the scrubbers and the fly ash filters areas.
All of these operators are responsible to the power generation plant
supervision for direction of their activities. The formal training
given to the operators included lectures by Davy. In-plant training
during start-up was directed by Davy start-up engineers, while some of
the "on the job training" was by PNM supervision. The chemical plant
operators were given more extensive training initially because of the
chemical process involved. This included training during initial
equipment testing for acceptance by the engineering contractor, as well
as the plant commissioning and subsequent plant operations.
Approximately one-third of the FGD personnel were selected from the
power plant operations group and the remainder from the chemical industry.
The creation of an operationally experienced crew for the chemical
plant, with the necessary levels of experience and ability presented
numerous problems of finding and relocating personnel. Recruiting
outside of the local plant area was necessary because of local
competition for experienced people. Past experience of operators and
supervisors included ammonia plants, power generation plants, sulfur
plants, chemical plants, refineries, and PNM's own power plant.
4. How The Plant Operates
Scrubber-absorber operations does not affect the power plant operation
because the FGD system can be by-passed. During a normal unit start-up,
scrubbers and absorbers are put on line after the electrostatic precipi-
tators are functioning. Power for the electrostatic precipitators,
booster blowers, and absorber circulating pumps is supplied from PNM's
power plant.
Each scrubber-absorber module is operated independently (except the
control data transmission to the control panel board), and put on line
separately. Each unit has a reheat system that is needed to protect the
stack from corrosive products resulting from condensation in the flue
gas when three modules are in operation simultaneously.
Two 750,000 gallon tanks for absorber product and feed solution provide
surge in the system to prevent the chemical plant operation from being
affected by normal operating fluctuations in the scrubber-absorber area.
The ideal operating situation is to have the feed solution tank full and
the product tank level very low.
750
-------
The chemical plants ability to operate depends on the power generation
unit operations. To date this has been the major operating problem.
Steam and water to the chemical plant originates in the power plant
area; clean condensate and process offgas are returned from the chemical
plant. The sumps and waste water streams are collected and neutralized
in a water treatment plant.
SC>2 is recovered from the absorber product solution in a double effect
evaporator. Precise control of the rate of SO- recovery from the slurry
in the evaporators is possible. The oxygen level in the recovered SO.
stream is controlled at less than 0.35%, as per the design which facili-
tates operation of the SO- reduction units at maximum natural gas utili-
zation. Evaporator slurry concentrations up to 70% by volume have been
achieved with little or no solids accumulation in any of the vessels.
Condensed water from the overhead SO--H-0 stream is recycled and used to
dissolve the solids from the evaporators. Both evaporator trains have
operated for many consecutive days without interruptions (other than
steam availability) over a wide range of operating conditions. Critical
parameters for SCL recovery rate are: steaming rates, slurry solids
percent, and chemical composition. Enough versatility has been provided
so that all four evaporators may be operated simultaneously or independently.
High speed, dry compressors are used to lower the absolute pressure in
the evaporators and pressurize the SO- gas stream through the reduction
process. Offgas blowers downstream 01 the SO- reduction units serve to
return incinerated tail gas and vessel vent gases back to the scrubber-
absorbers.
Each of the Allied Chemical SO- Reduction units includes a primary
reactor system where a portion of the SO- is reduced to sulfur and H»S
using natural gas as a reductant. Exothermic heat of reaction is stored
in combination reactor generator vessels for subsequent use in preheating
the feed gas. The flow through the primary reactor system is reversed
on a periodic basis. The sulfur formed is condensed and the cooled gas,
containing proper proportions of SO- and H-S, is processed through a
Claus conversion system for recovery of additional sulfur. The Claus
system off-gas is incinerated and recycled to the Wellman-Lord absorbers.
The raw natural gas available at the San Juan Station contains quantities
of C and higher hydrocarbons which make it unacceptable for use as a
reductant in the S02 reduction process. The heavy ends are removed in a
gas treating unit and utilized as fuel in the tail gas incinerators.
Sulfate formed by the the oxidation of sulfite in the absorber solution
must be purged from the solution since sulfate is inert in the system
and will not absorb or release SO-. Sulfate is separated from the
absorber product solution by low temperature crystallization.
The sulfate is recovered as a decahydrate, melted and sent to an evaporator
for removal of the water of hydration, then separated and dried to less
than 1% water for ease of storage and shipment.
751
-------
Separate plant utilities are dedicated to the chemical plant for cooling
towers, air compressors, steam reducing stations, water supply and
recovery systems. Steam, high quality water, and electricity as stated
previously are provided from the PNM generating plant. Steam condensate
is returned to the generation plant from the Chemical Plant.
5. Plant Operations
While considerable time will be spent on discussing operational problems,
it is worth mentioning those units that have operated as designed and
with little or no start-up problems.
Computer control of the scrubbers from the remote control room has
worked well from the beginning. As with most computer-controlled
systems, the success or failure of the control system depends on the
sensor elements in the field.
Absorbers do a good job of removing S0_ from the flue gases. With
nearly design solutions and normal amounts of soda ash being added to
the system, Unit Number 2 has made compliance with State and Federal
regulations as demonstrated on November 29, 1978. Unit Number 1
compliance will be demonstrated as soon as possible.
Evaporators and S0? compressors have been operated for long periods of
time. The only problem has been some erosion of the SO- compressor
impellers probably caused by a condensing condition at the inlet of the
compressors. Corrections and revisions have been made to prevent this
condensation. These units have been operated, since the corrections,
over a wide range of flow rates with no difficulty.
Soda ash addition to the process has been relatively easy and very
dependable. Soda ash is stored dry and dissolved when needed for use at
each absorber or at the evaporator slurry dissolving tank.
6. Plant Operating Problems
As in most new plants there were various mechanical problems experienced
during the initial start-up. At times these problems created operating
difficulties. Most of these problems have been resolved with a few to
be completely cleared. The following is a brief summary of the major
problems that have occurred.
Steam and Water Availability:
Problem;
The steam and water supply to the FGD plant has frequently been inter-
rupted or curtailed thereby greatly reducing or stopping the S0? chemical
plant operations. This resulted from operating difficulties in power
generation units.
752
-------
Solution;
The original design parameter was that relatively unlimited steam and
water supplies would be available from the power plant. This has not
been true. Studies and evaluations of the steam and water systems are
underway to assure more continuous supply, or to provide standby
quantities of steam and boiler quality water solely dedicated to the
chemical plant.
Results:
Not enough time has elapsed for results to be evaluated pertaining to
changes or modifications that have been made or completed.
Solution Losses;
Problem;
There have been solution losses occurring which have not been readily
identifiable, especially in the winter months. This solution loss
appears as though it may be coming from the bottom tray of the absorber
into the scrubber sump when less than minimum design gas flow rate is
being processed.
The design turndown is 50% for each absorber. The solution losses occur
when the absorbers are operated below the design turndown in order to
limit the amount of incoming flyash during periods when the electrostatic
precipitators malfunctioned. Additional solution losses have occurred
at the fly ash filter whenever the automatic cycling fails.
Solution:
a. Improved operator attention.
b. Correction of the electrostatic precipitator problems
c. Operate the absorbers at or above the minimum design gas flow rate.
Results:
Not all of the planned corrections have been completed or performed
because of scheduling and operation of the plant. Plant turnaround in
the future during which these changes are to be accomplished.
Electrostatic Precipitator:
Problem:
The electrostatic precipitators have failed or malfunctioned leading a
condition which overloads the scrubber solution with fly ash. Design
specifications are that the scrubber should be capable of accepting a
complete precipitator failure for a period of two (2) hours. The
753
-------
scrubbers have demonstrated the capability of accepting a complete
precipitator failure for time periods up to 36 hours. Furthermore, the
scrubbers have operated for extended periods of time (several days) when
the precipitators operated at less than 60% efficiency. Solution density
measurement devices have proved inadequate to warn of pending recirculation
problems. At one time significant amounts of fly ash were collected in
the recirculating solution to increase the density where recirculation
was reduced to allow hot flue gases to bypass. The hot gases contacted
the down stream heat sensitive Chevron mist eliminators that separate
the scrubber and absorber. Some of the Chevrons were warped and required
replacement in four modules.
Solution;
a. Correction of problems in the electrostatic precipitators.
b. Improved operator attention and manual sampling.
c. Establishment of operating techniques to handle sudden and unexpected
fly ash, such as setting a 15% by volume fly ash concentration
limit for shut down.
Results:
No real results will be available until all the planned improvements have
been executed. These corrections are to be performed during the next
planned plant turnaround scheduled in the near future.
Purge System:
Problem:
The purge system evaporation of crystallized and remelted sulfate to a
dry solid has not operated sufficiently to provide evaluation. This is
the result of power plant shutdowns and steam and water shortages.
There has been some indication that solids being formed in the purge
evaporator are too small to be separated by a screen type centrifuge. A
problem has also been experienced in the method of feeding the wet
centrifuge cake into the flash dryer.
Solution:
a. Improved utilities, such as steam, water and power.
b. Sieve sized changes in the centrifuge.
c. Improved operating techniques for improved crystal growth.
d. Improvement in conveying chutes, conveyor and ducts.
Results:
To date results have been difficult to evaluate because of insufficient
operating time.
754
-------
Glaus Catalyst Beds;
Problems:
Portion of catalyst in the SO. reduction area was found to have shifted
past the Claus converter catalyst support screens into the bottom of the
vessels. Investigation revealed that support screens had been improperly
installed during plant construction with substantial gaps between sections
of screen and adjacent to the vessel walls.
Solution;
Correct installation of the screens in the Claus converter has been
completed.
Results:
The problem has not reoccurred since correction was made.
7- Future Plans & Chemical Sales Program
The sales of FGD process products, the main concern in the immediate
future, has several effects on the operating and the administration
costs of the scrubber operations. Since beginning the chemical sales
contracts, PNM has received numerous contacts concerning contracts and
logistics involved in selling and shipping these materials. The follow-
ing is a brief discussion that attempts to answer the typical inquiries
received on this program. The sales of these commodities have been
executed through a broker.
The most obvious cost reduction effected by the sales program is the
elimination of disposal costs. On an average basis the four units (two
future units to be built) at San Juan Station will produce about 300
tons/day of S0~. This translates into 150 tons per day of elemental
sulfur or 1340 tons per day of 60% calcium sulfate and water sludge had
limestone scrubbers been used. Whereas the sulfur produced in a regener-
able system is a salable product which results in a significant monetary
credit to help offset the operating cost of the FGD system, use of a
limestone scrubber FGD system would necessitate the disposal of some 45
truck loads of sludge per day. Gypsum disposal costs for the plant
would have exceeded $1,000,000 per year for hauling, mine preparation
costs excluded.
PNM plans to install a sulfuric acid plant on the FGD Units Number 3 and
4, which is presently in the engineering phase. The advantage of producing
sulfuric acid instead of sulfur is the' elimination of natural gas consumption.
At $1.70/MCF Units Number 1 and 2 consume about $450,000 per year of
natural gas. With sulfur plants on Units Number 3 and 4, San Juan
Station would require $1,007,000 per year of natural gas. A sulfuric
acid plant costs less in capital than a comparable capacity S0_ reduction
unit. Although the natural gas is classed as chemical plant feed stock,
it is still subject to curtailment during the winter months, which is
another reason to eliminate this requirement.
755
-------
When the San Juan complex is complete the following materials will be
produced from the scrubber systems. (On an average coal basis):
Approximate
Item Short tons per Day Truck loads per Day
1. Salt Cake 60 2
2. Sulfur 60 2
3. Sulfuric Acid , 250 10
TABLE 4
Anticipated PNM Material Produced
While it would seem inappropriate to discuss unit prices here, it can be
stated that the revenues PNM will receive for these materials will
reduce the overall FGD system operating costs by about 10%, or roughly
$1,000,000 per year, in addition to the savings discussed earlier.
Revenues received by others for product chemicals depends on site location
and.the availability of transportation. The net return on chemicals is
freight sensitive and distances of 100 miles can easily double the
selling price if one expects to recover the freight. Easy access to
rail or barge networks would allow cheaper freight, hence the radius
encompassing final customers is extended somewhat if those options are
available. At times these commodities may have to be sold on a freight
equalized basis; that is, the seller absorbs part of the freight in
order to offer a competitive price for the material. The extent of
freight equalization depends upon market conditions, the proximity of
other production plants, and upon what modes of transportation are
available at the plant site. In any case, it is possible that freight
costs can reduce the possible return on the product sales.
Since the market prices for sulfur and sulfur products have varied
considerably over the years, the contract between a broker and a
producer should be structured to protect both parties. One form is the
standard "evergreen" contract with a floor price for the producer with
an additional increment for the broker to cover his costs. Anything
beyond the increment could be split on a percentage basis that is
mutually agreeable to the contracting parties.
One aspect of scrubber operations that is affected through sales of
products is that the quality of the end product must be considered in
daily operations. For salt cake this means color, particle size, and
composition. For sulfur, quality is based on color, purity, and the
absence of remelt sulfur. The sulfuric acid quality will be clear (free
of particulates) and have a low iron content. The Wellman-Lord scrubber
system is expected to produce electrolytic grade acid because the SO-
from the evaporators is very clean.
756
-------
COMPANY AND LOCATION
FEED GAS ORIGIN
SCFM GAS TREATED
DISPOSITION OF S02
Ul
—j
JNITS ON STREAM
01in Corp./Paulsboro, NJ (now shut down)
SOCAL /El Segundo, CA
Allied Chem./Calumet, IL
Olin Corp./Curtis Bay, MD
SOCAL /Richmond, CA
SOCAL /Richmond, CA
SOCAL /El Segundo, CA
NIPSCO/Gary, IN
PSCNM/Waterflow, NM
Sulfuric Acid Plant 45,000
Claus Plant 30,000
Sulfuric Acid Plant 30,000
Sulfuric Acid Plant 78,000
Claus Plant 30,000
Claus Plant 30,000
Claus Plant 30,000
115 MW Coal Fired Power Plant 310,000
700 MW Coal Fired Power'Plant 1,880,000
Recycle to Acid Plant
Recycle to Claus Plant
Recycle to Acid Plant
Recycle to Acid Plant
Recycle to Claus Plant
Recycle to Claus Plant
Recycle to Claus Plant
Elemental Sulfur Plant
Elemental Sulfur Plant
UNITS IN DESIGN OR CONSTRUCTION
PSCNM/Waterflow, NM
Getty/Delaware City, DE
1100 MW Coal Fired Power Plant 2,727,000
Three Coke Fired Boilers 500,000
Sulfuric Acid Plant
Sulfuric Acid Plant
EXHIBIT I
WELLMAN-LORD PLANT INSTALLATIONS IN THE UNITED STATES
-------
COMPANY AND LOCATION
FEED GAS ORIGIN
SCFM TREATED
DISPOSITION OF S09
Ul
c»
UNITS ON STREAM
Japan Syn. Rubber/Chiba
Toa Nenryo/Kawasaki
Chubu Electric/Nagoya
Japan Syn. Rubber/Yokkaichi
Sumitomo Chem./Sodegaura
Kashima Oil/Kashima
Mitsubishi Chem./Mitzushima
Toa Nenryo/Hatsushima
Toyo Rayon/Nagoya
Japan Nat. Railroad/Kawasaki
Kurashiki Rayon/Okayama
Fuji Film/Fujinomiya
Shin Daikyowa/Yokkaichi
Sumitomo Chem./Niihama
Mitsubishi Chem./Mizushima
Mitsubishi Chem./Kurosaki
Tohoku Electric/Niigata
Oil Fired Boiler
Claus Plant
220 MW Oil Fired Power Plant
Oil Fired Boiler
Oil Fired Boiler
Claus Plant
Oil Fired Boiler
Claus Plant
Oil Fired Boiler
200 MW Oil Fired Power Plant
Oil Fired Boiler
i
Oil Fired Boiler
Oil Fired Boiler
Oil Fired Boiler
Oil Fired Boiler
Oil Fired Boiler
100 MW Oil Fired Power Plant
124,000
41,000
390,000
280,000
225,000
20,000
373,000
10,000
218,000
435,000
248,000
89,000
253,000
91,000
390,000
i30,000
236,000
Sulfuric Acid Plant
Recycle to Claus Plant
Sulfuric Acid Plant
Sulfuric Acid Plant
Sulfuric Acid Plant
Recycle to Claus Plant
Sulfuric Acid Plant
Recycle to Claus Plant
Sulfuric Acid Plant
Sulfuric Acid Plant
Sulfuric Acid Plant
Liquid SO2
Sulfuric Acid Plant
Liquid SO2
Sulfuric Acid Plant
Sulfuric Acid Plant
Sulfuric Acid Plant
EXHIBIT 2
WELLMAN-LORD PLANT INSTALLATIONS OVERSEAS
-------
ACKNOWLEDGEMENTS
The authors would like to extend their appreciation to the following persons
that helped in preparing this paper. Much of the material and data in
this paper has been received from, or reviewed and corrected by, the
following without specific mention.
E. L. Mann
Executive Manager of Electric Production
Northern Indiana Public Service Company
5625 Hohman Ave.
Hammond, Indiana 46325
L. C. McGrath
Manager of Plant Engineering
Northern Indiana Public Service Company
591 Marquette Mall Office Building
490 St. John Road
Michigan City, Indiana 46360
D. W. Christian
S02 Chemical Engineer
PuBlic Service Company of New Mexico
Albuquerque, New Mexico 87103
Robert Bierbower
Project Manager
Allied Chemical
Industrial Chemicals Division
P.O. Box 1139R
Morristown, New Jersey 07960
A. Giovanetti
Supervising Process Engineer
Davy Powergas Inc.
P.O. Drawer 5000
Lakeland, Florida 33803
R. I. Pedroso
Principal Process Engineer
Davy Powergas Inc.
P.O. Drawer 5000
Lakeland, Florida 33803
759
-------
REFERENCES
1. The Operating History and Performance Experience of Unit Number 11
and the FGD System During the Pre-Acceptance Test and Acceptance
Test Periods.
F. W. William Link of NIPSCO and Wade Ponder of EPA, presented
in November 1977 at the FGD Symposium, Hollywood, Florida
2. EPA Demonstration of Wellman-Lord/Allied Chemical FGD Technology
Acceptance Test Results
Interagency Energy/Environmental R&D Program Report
EPA-600/7-79-Ol4a
January 1979
3. Sulphur Recovered From SC- Emissions at NIPSCO1s Dean H. Mitchell
Station
Howard A. Boyer
Allied Chemical Corporation
Morristown, New Jersey
and
Roberto I. Pedroso
Davy Powergas Inc.
Lakeland, Florida
4. "Sulfur Recovered From Flue Gas at Large Coal Fired Power Plants"
Roberto I. Pedroso
Davy Powergas Inc.
Lakeland, Florida
Paper presented to the Third Symposium on Sulphur and other Airborne
Emissions, Salford, England - April 1979
760
-------
4Q
CITRATE PROCESS DEMONSTRATION PLANT
- CONSTRUCTION AND TESTING -
Authors:
Richard S. MadeTiburg
Morrison-Knudsen Company, Inc.
Boise, Idaho
Laird Crocker
U.S. Bureau of Mines
Citrate Process Demonstration Plant
Monaca, Pennsylvania
John M. Cigan
St. Joe Zinc Company
Monaca, Pennsylvania
Laurance L. Oden
U.S. Bureau of Mines
Albany Metallurgy Research Center
Albany, Oregon
R. Dean Delleney
Radian Corporation
Austin, Texas
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
Flue Gas Desulfurization Symposium
Regenerable Processes Session
March, 1979
Abstract
Construction of the Citrate Process Demonstration Plant represents a major milestone in the achievement
of the Federal Bureau of Mines goal to minimize the undesirable environmental impact associated with
S02 emissions from industrial sources. Baseline performance testing of the boiler prior to retrofit of
the Citrate Process was conducted during November, 1978. Mechanical testing, start-up and operational
commissioning of the plant are now in progress. The demonstration of the Citrate FGD process at a 60
MWe coal-fired power generating station will confirm the design basis, the technical merits and the pro-
cess's capital and operational economics at a commercial scale installation.
This paper discusses the construction of the demonstration plant, the mechanical and pre-start-up testing,
and the test and evaluation program. A brief description of the process, recent material evaluations and an
assessment of application of citrate technology to other industrial SC"2 emitting sources are presented.
761
-------
CITRATE PROCESS DEMONSTRATION PLANT
- CONSTRUCTION AND TESTING -
INTRODUCTION/PERSPECTIVE:
Construction and testing of the Citrate Process Demonstration Plant as shown in Figure I, represents an
interim milestone in the successful development and demonstration of the Citrate Flue Gas Desulfuriza-
tion (FGD) Process. The Citrate Process was developed by the U.S. Department of the Interior's Bureau of
Mines to meet the goal of minimizing the undesirable environmental impact of industrial plants emitting
SC>2 bearing gas. For almost a dozen years, the Federal Bureau of Mines researched, pilot tested and
evaluated this technique of aqueous scrubbing with hydrogen sulfide regeneration. Successful pilot testing
of the Citrate Process by both government and industry at both metallurgical and coal-fired industrial
applications confirmed the technical merits of the Citrate Process over a wide range of differing flue gas
characteristics.
Consistent with the national objectives for energy independence and to derive "Clean Power from Coal",
(1 )1 it was decided by a joint government-industry team to demonstrate the Citrate Process at a 60 MWe
coal-fired electric power generating plant owned and operated by the St. Joe Zinc Company (St. Joe) of
Monaca, Pennsylvania. A cost sharing cooperative agreement between the Bureau of Mines, the U.S.
Environmental Protection Agency (EPA) and St. Joe provided the necessary impetus for the engineering,
design, construction, and one-year demonstration testing of the Citrate Process at a commercial scale
power plant installation. The engineering, construction management, construction and operational testing
responsibility are being provided to the project principals by the Morrison-Knudsen Company, Inc., of
Boise, Idaho.
The primary objective of the Citrate Process Demonstration Plant project is to demonstrate that the
Citrate Process can reliably and efficiently remove sulfur oxides from the flue gases at a commercial scale
power plant installation. To achieve this objective, the Citrate Process was installed to treat a nominal
156,000 SCFM (180,000 SCFM, maximum) of SC>2 bearing flue gases from St. Joe's G. F. Weaton power
station, as shown in Figure II. The existing flue ducting configuration allows for transfer of additional
untreated flue gases from an adjoining twin boiler to the Citrate FGD unit without physical modification
of gas ducting. The St. Joe power plant operates essentially base loaded and provides electrical energy to
St. Joe's adjoining zinc smelter and to the local power grid. The power plant's main steam flow is 450,000
Ib/hr at 1000°F, 1850 psig and is provided with integral reheat. The station is operated on a continuous
basis and load changes are typical of an electric utility. The overall station heat rate is approximately
10,200 Btu/kWh.
Underlined numbers in parentheses refer to the list of references at the end of this report.
762
-------
ON
UJ
Overview Photograph of Constructed Demonstration Plant - Figure I
-------
nm pm rrm
nxD II""1 i^-L-iu—
1. Flue Gas Monitoring Platforms/Stack
2. S02 Absorber
3. Electrostatic Precipitator
4. Sodium Sulfate Crystallizer
5. Sulfur Precipitation Reactor No. 1
6. Sulfur Precipitation Reactor No. 2
7. Digestor Vessel
8. Sulfur Flotation Tank
9. Sulfur Slurry Tank (Not Visible)
LEGEND
10. Sulfur Melter
11. Citrate Solution Storage Tank
12. Hydrogen Sulfide Generator
13. Fuel Oil Storage Tank
14. CO/CH4 Compressor House
15. Molten Sulfur Storage Tank
16. Lime Neutralization Tank
17. G. F. Weaton Power Station
18. Existing 275 Foot Stack
-------
Photo of G. F. Weaton Power Station
Figure II
765
-------
PROGRESS:
A network diagram of the engineering, procurement and construction progress is shown in Figure 111. This
figure is a simplified critical path analysis and approximate representation of the various engineering and
construction activities and work tasks that comprise the effort to design, build, and test the Citrate
Process Demonstration Plant.
A summary of activity milestones associated with the project's four separate phases follows:
Phase 1 — Preliminary engineering and design development necessary to
establish a definite construction cost estimate was completed in Nov-
ember, 1976.
Phase II — Final Engineering, detailed design, and equipment procurement
of most major equipment and engineered items was completed by March,
1978. Major equipment including columns, vessels, tanks, reactors, pumps
and exchangers was received on site by September 1978. Installation of
process and utility piping, instrumentation, electrical work, insulation,
and flake glass lining of process vessels was completed in early 1979.
Preliminary mechanical testing of the demonstration plant and completion
of most construction activities was achieved by March, 1979.
Phase III — Plant testing was initiated in March, 1979, and is expected to
be completed by June, 1979.
Phase IV — A one year demonstration testing and performance evaluation
program will be conducted by the Radian Corporation (Radian) of Austin,
Texas, which has been retained by the Bureau of Mines as an independent
testing and evaluation contractor. The one year demonstration test is
scheduled to commence upon completion of performance testing and
plant acceptance, after which time St. Joe fully expects to continue
operation of the Demonstration Plant to achieve continued compliance
with applicable environmental regulations.
766
-------
31 41516
21 22 23 24
PIPING DESIGN & SPECS.
PIPING DESIGN & SPECS
-IN PERFORM I
UTILITY TIE-INS
ABOVE GROUND PIPE INSTALLATION
PROCURE PIPING MATERIALS
COMPLETE PIPING
I MAJOR STEEL ERECTION
RECEIVE/^STRUCT. STEEL
FABRICATE STRUCT.
TEEL DESIGN & SPECS.
MPLETE STEEL_ERE£TION
STEEL ERECTION
COMPLETE UNDERGROUND
I PIPING ,DESIGNI
UNDERGROUND PIPING f PROCURE UNDERGROUND
DESIGN & SPEC
MOBILIZE CON
FORCES I
INSTALL UNO
PIPE
TE GRADING & ipf,
EXCAVATION & BACKFILL
& POUR CONCRETE
FORM & POUR CON
COMPLETED ABSORBER
CIVIL ! r )
>ESIGN & SPECS -
CIVIL DESIGN & SPECS/ ,
EQUIPMENT
COMPLETE RECEIPT OF
I /-^ EQUIPMENT1
EQUI
NT DESIGN A
MOBILI
DESIGN
FORCE
SYSTEM
OPERATIONAI
TEST
FLAKE GLASS LINING
& REFRACTORY BRICK
COMPLETE
PIPING
SPECIFICATIONS
EQUIPMENT FABRICATION
ELECTRICAL SINGLE
LINE DIAGRAMS
SUBCONTRACT BID &
AWARD OF
CONTRACT
& SPECS SUBCONTHAWI BIU
^ "O * O AWARD ELECTRI9AL
COMPLETE VALVE
INSTALL VALVES
(^INSTALLATION
RECE1VE UNDERGROUND
ELECTRICAL MATERIAL
UNDERGROUND ELECTRICAL
ELECTRICAL MATERIALS.
DESIGN
& SPECS
INSTALLTUNDERGROUND
ELECTRICAL POW
& LIGHTING DESI
COMPLETE ELECTRICAL
COMPLETE PANEL
BOARDS
INSTRUMENT DWGS. & SPECS
FAB. PANEL BOARD
INSTRUMENTATION
COMPL. PROC.
OF REL VALVES
PROCURE RELIEF VALVES
CQMPL. PROC.
CONT'L VALVES
PROCURE CONTROL VALVES
INSTRUMENTATION
\ PROCURE MISC. INSTRUMENTS
RECEIVE PANEll
QMOUNTED INSTRUMENTS
v PROCURE PANEL MOUNTED INSTRUMENTS
-------
THE CITRATE PROCESS:
Citric acid has been demonstrated as an effective buffer for the aqueous absorption of sulfur dioxide.
The absorption/regeneration system using this organic acid is known as the Citrate Process. Citrate chem-
istry as well as a detailed description of the process has been previously reported and is cited in the
references. (2-5)
The Citrate Process for sulfur dioxide emission control comprises five basic steps:
1. Preconditioning the entering flue gas. Depending upon the application,
the entering flue gases may require reductions in the levels of chlorides,
sulfuric acid mist, and particulates, as well as quenching to adiabatic
saturation temperatures.
2. Absorption of SC<2 in an aqueous buffered solution of organic acid.
Conditioned flue gas enters the S02 absorber where it flows counter-
current to descending citrate solution.
3. Reaction of the SC*2 loaded solution with H2S in a closed vessel to
precipitate elemental sulfur. Sulfur dioxide absorbed in the citrate
solution is reduced to elemental sulfur by reaction with h^S, thus,
regenerating the absorbent solution.
4. Separation of the sulfur from regenerated solution. The elemental
sulfur precipitate is concentrated by air flotation into a sulfur slurry
which is separated from the regenerated solution. The sulfur slurry
is heated to form liquid sulfur to enable decantation from the retained
citrate solution.
5. H2S generation. The H2S required for use in regeneration is either
obtained as a by-product of petroleum refining or produced on-site
by reaction of recovered sulfur product with a reducing gas and steam.
The basic processing unit operations associated with the Citrate Process are diagrammatically presented
in Figure IV.
768
-------
H2S GENERATION
LIME
NEUTRALIZER
CRYSTALLIZER
SULFUR
FLOTATION
SULFUR
PRODUCT
PRECONDITIONING &
SO2 ABSORPTION
SULFUR PRECIPITATION
& RECOVERY
Citrate Process Flow Diagram - Figure IV
769
-------
DEMONSTRATION FACILITY:
The Citrate Process Demonstration Plant is designed as a single process train consisting of flue gas pre-
treatment, S02 absorption, sulfur precipitation, sodium sulfate removal, and H2S generation. Mechan-
ical and process equipment within the demonstration plant is capable of being scaled upward so that
once demonstrated in the plant, the process can be applied with confidence to larger coal-burning utility,
refinery, or smelter installations. Typically, equipment selected for the demonstration plant has been
a successfully tested in similar commercial applications. Equipment sizing is representative of com-
ponents for process trains associated with a 500 MWe power plant installation.
As a demonstration plant, a conservative design approach was followed emphasizing the use of specific
knowledge acquired through operation of two previous Bureau of Mines' citrate pilot plants in antici-
pating and preventing problems. Certain optional process equipment and technical features were
provided to enhance process flexibility, optimization, and to avoid potential problem areas. Design
features included in the constructed plant which emphasize this philosophy follows:
• A low energy venturi scrubber was provided upstream of the ab-
sorber column to remove a substantial percentage of chlorides, 863
and NC>2 which could interfere with operation of the process. The
venturi scrubber also functions to cool incoming gases to achieve
efficient absorption.
• To prevent corrosion problems, the ductwork and the flue gas fan
were located upstream of the venturi scrubber thus being exposed
to only hot, dry gas above the acid dew point.
• The demonstration plant can be tested to ultimate capacity. The
flue ducting configuration allows for transfer of additional untreated
flue gases from the adjoining twin boiler to the Citrate FGD unit
without physical modification of gas ducting. Approximately
18 percent excess fan capacity is provided that will allow for gas
treatment during overload conditions. Similarly, the SC>2 absorber,
H2S generator, and other process limiting unit operations are de-
signed for overload conditions.
• Both indirect hot air reheat with steam and an external combustion
oil-fired reheater are provided to enhance system reliability and to
permit evaluation of these alternative reheat methods.
• Rich absorbent solution is pumped from the absorber to the highest
of two stirred precipitation reactors and the solution then flows
through the other reactor, digester, and flotation tanks by gravity.
The reactors feature a turbine and gas sparger design which will not
be subject to plugging. Agitation is maximized to provide good
reaction with h^S gas; the degree of reactor agitation can be adjusted
by gear modification.
770
-------
• Experience has shown that excess h^S tended to cause difficulty
in the flotation of sulfur. Therefore, the stirred digester tank was
provided with excess capacity to allow time for mixing and reaction
of any excess H2S with a small flow of rich citrate solution which
by-passes the reactors. This gives better flotation and makes sub-
sequent handling of the regenerated solution safer.
• Separation of sulfur from the absorbent solution and subsequent
melting was successfully demonstrated during pilot plant testing.
The method, based on air flotation of sulfur, is similar to existing
technology used in the Stretford process. All lines which contain
molten sulfur are steam jacketed and insulated to prevent plugging.
• The oxidation of absorbed S02 is expected to be less than 2%.
Excess sulfate in the system will be removed as Glauber's Salt in
a vacuum crystallizer, which will keep the sulfate level sufficiently
low so that crystallization will not occur in the process lines.
• The H2S generator was designed to minimize carry-over of sulfur
vapor into the H2S product lines. An alternating steam heated/water
cooled heat exchanger was provided to condense out sulfur carry-
over. In the event of the sulfur carry-over the H2S gas lines are
steam jacketed so that any sulfur will be transported to the precipi-
tation reactors for recovery.
• Due to the corrosive nature of the absorbent solution and the possi-
bility of chlorides being present, construction materials for the
demonstration plant were carefully chosen. Hastelloy C-276
or Inconel 625 were used for agitators, pumps and load bearing
clips. Piping is FRP, polypropylene lined steel or rubber lined
steel. Tanks are lined with rubber or vinylester flake glass. Some
smaller tanks are FRP.
• The two sulfur precipitation reactors are provided with by-pass
capability to enable continuous process operation with only one
reactor in operation; this feature will permit optimization of H2S
utilization.
• The H2S generator is designed to operate on carbon monoxide as
an alternate reducing gas. However, operation with a CO feedstock
will require the installation of a gas transport pipeline from the
adjoining smelter, where CO is an available by-product from the
electrothermic reduction of zinc.
771
-------
CONSTRUCTION:
The decision to proceed with the detailed design and construction of the Demonstration Plant was
made in March, 1977. Materials and special components such as Hastelloy clad vessels and Inconel-
625 agitators were promptly ordered and the fabrication of special process equipment and major ves-
sels was started.
Shop fabricated vessels including the S02 absorber, sulfur precipitation reactors, digester, sulfur slurry
tank, and citrate storage tank, and H2S generator reactors and vessels were received at the project
site in June, 1978.
26-foot diameter by 98-foot long SC>2 absorber and the 102-foot long by 10-foot diameter top mounted
stack were both shop-fabricated and delivered to the site as single units. However, the absorber internals
including the chevron mist eliminator, liquid distributor channels, absorber packing, and reheater were all
field installed.
Field erection of the 862 absorber, stack assembly, and major vessels required considerable planning
and site preparation. The vessels were shipped by barge approximately 1200 miles, off-loaded, and
transported two miles on special vehicles before being field erected and set into position on foundations.
The absorber section required significant rigging to enable field erection of this 150-ton vessel with
a single lift. Erection of the absorber vessel is shown in the accompanying photograph, Figure V.
Erection of the S02 absorber was the critical path in the construction schedule because numerous related
work activities such as installation of flue gas fan, ducting, and scrubber recycle pumps were restrained
until the absorber was erected. The project critical path was further complicated by the late delivery
of the absorber and other major vessels, fabricated lined pipe, and by limited availability of skilled
craftsmen (pipefitters). Although the major process vessels and columns were promised for delivery in
December 1977. the actual equipment did not arrive on site until mid-June, 1978. Late material de-
liveries were also experienced with the factors coupled with a high demand for skilled craftsmen in the
local labor area resulted in significant loss of potential production and contributed directly to delaying
the overall completion of the project by approximately six months.
All wetted process surfaces inside the steel vessels were lined with a flake glass vinylester to protect the
carbon steel from corrosion attack by an allowable two weight percent chlorides contained in the citrate
solution. To assure that the protective lining would withstand the corrosive environment for the design
plant life of approximately 20 years, the metal surfaces to be coated were sandblasted to white metal and
immediately primed; a layer of vinylester glass, approximately 35 mils, was then troweled and rolled,
and a second layer of vinylester glass, approximately 30 mils, was spray applied. Field quality control
consisted of continuous visual inspection while lining installation was in progress, random magnetic
thickness measurement, and a 100 volt/mil spark test for the entire lined surface.
772
-------
Erection of Absorber Vessel - Figure V
773
-------
A total of nine vessels and tanks were flake glass lined representing some 17,000 square feet of lined
surface. In addition to flake glass lining, the venturi scrubber and SC>2 absorber sump were lined with
acid brick over a neoprene rubber membrane liner.
An internal fiberglass sleeve was installed within the carbon steel stack to prevent acid corrosion. In
addition, a Pennguard 2 foamed borosilicate glass block liner was installed in the reheat section of the
absorber and tower stack area to protect the carbon steel shell from flame impingement when operating
the oil-fired reheat and general corrosion from the treated flue gases.
Site preparation work, which consisted of demolishing some minor concrete structures, relocation of
a power plant sump and utility piping, and identifying underground interference within the demon-
stration project's battery limits, was completed by October, 1977, at which time installation of under-
ground and yard piping was started, and shortly thereafter construction of equipment foundations was
begun. Sixty-nine pour-in-place concrete piles were installed, two process sumps and a total of 100
equipment foundations were built. Approximately 1500 cubic yards of concrete and 126,000 Ibs. of
structural rebar were used. About three and one-half months were required to complete the concrete
and civil work activities which are shown in Figure VI. Installation of the structural steel supports
for the sulfur precipitation reactors, digester, and air flotation vessel was completed in February 1978.
The process and utility pipe-rack, a pipe bridge, structure for the sulfate crystallizer system, and h^S
generator, and ducting support structure, as shown in Figure VII, were installed intermittently due to
scheduling restraints. Total time for the structural activities was approximately seven months; in all some
550,000 Ibs of structural steel, anchor bolts, and miscellaneous steel were consumed by the project.
Approximately 15,000 feet of utility and underground piping were installed at the project site; addition-
ally, some 8,500 feet of process piping, mostly carbon steel polypropylene lined pipe and steam jacketed
pipe, were installed. All process and utility piping 2-inch diameter and less were field fabricated and all
larger diameter piping, lined pipe, and related fittings, hangers, and supports was completed over a period
of eight months.
Utility, steam, flue gas, and process off-gas tie-ins to the G. F. Weaton power station were scheduled to
coincide with the power plant annual shutdown.
Electrical work consisted of installation of 22,000 feet of conduit, 1,300 feet of cable tray, 112,000 feet
of power cable, equipment grounding, area lighting and power service, and the installation of five 2300
volt motors and sixty-four 600-volt or smaller motors. In addition, switchgears, power distributor
centers, transformers, and an all-electronic instrumentation package consisting of some 200 sensing
elements, transmitters, and recorders were installed. The installation of the electric work and instrumen-
tation occurred over a five month period.
Reference to specific trade names is made for identification only and does not imply endorsement by the Bureau
of Mines.
774
-------
Concrete and Civil Work Activities - Figure VI
-------
,=3*--
H2S Generator and Ducting Support Structure - Figure VII
-------
All field construction was performed by journeymen and apprentice craft labor. Some twelve different
labor skills were used; these included pipefitters, iron workers, carpenters, masons, millrights, electri-
cians, boilermakers, insulators, painters, and laborers. The peak project manloading for all craft and
subcontractors was 102 men. A total of approximately 1.50,000 manhours of craft labor was expended
in, the construction of the demonstration plant.
Concurrent with the completion of construction of different process systems, mechanical testing of
each component, vessel and related pipe network was made to assure completeness and operational
intonritw
integrity.
777
-------
MECHANICAL TESTING:
Mechanical testing of the various mechanical and process components, piping systems, and auxiliaries
was initiated during February, 1979.
Mechanical testing implemented to avoid plant start-up problems, consisted of visual inspection of all
equipment and piping systems to assure completeness, spark testing of all lined vessels, and hydrostatic
leak testing of pressure vessels, solution tanks, and pipe networks. Hydrostatic testing of jacketed, lined
and specialty piping and stress relieved vessels was conducted in the fabrication shop prior to release for
field installation. Once installed, entire systems were then systematically pneumatically pressure tested
and/or hydraulically tested, depending upon service. In the case of the refractory lined H2S generator
vessels and piping, a halogen leak test was used to avoid wetting the refractory surfaces.
All electric motors, switchgear, and distribution apparatus were functionally tested to assure proper
motor rotation, and to determine workability of motor controls and safety interlocks. Similarly, all
electronic instrumentation was field calibrated and tested over the intended operating range. Control
valves, regulators, and safety relief valves were all "stroke" tested and adjusted to performance specifi-
cations.
Process lines were water washed to remove debris and foreign matter; steam and condensate lines were
degreased and steamed-out. Unlined carbon steel tanks for caustic and molten sulfur storage were
cleaned of loose mill scale and corrosion products.
778
-------
PERFORMANCE TEST PROGRAM:
Testing of the demonstration plant will establish Citrate FGD technology in mechanical and process
equipment of sufficient size so that scale up to 500 and 1000 MWe power plants is feasible, and that
such applications can be built and the performance, capital, and operational costs predicted with con-
fidence. To insure the objectivity of the results, the Bureau of Mines solicited proposals for independent
testing and evaluation of the Demonstration Plant. Radian Corporation of Austin, Texas, was the
successful bidder and was selected to perform the evaluation.
Testing will be done by operating with the demonstration plant in a variety of operating modes of
varying loads, while burning various coal fuel mixtures with sulfur contents between 2.5 to 4.5 weight
percent. The test program consists of baseline testing, acceptance testing, and a one-year demonstration
program.
• Baseline Test — Baseline performance testing of Unit No. 1 of the
G. F. Weaton power station with selected coal fuel was conducted to
characterize the flue gas in the untreated mode prior to start-up of
the demonstration plant. The baseline test was completed during
November, 1978. These tests characterized the boiler, particularly
with respect to flue gas emissions, sulfur balance, and combustion
efficiency. Results of the baseline test serve as a basis for evaluating
the effectiveness of the Citrate Process and for determining what
effect, if any, retrofit of the demonstration plant has on boiler
operation. Preliminary results indicating the flue gas emissions and
the coal sulfur content and heating value are presented in Table I.
Information gathered during baseline testing includes:
— Definition of the characteristic operating boundaries of the steam
generating facility. Historical operating data on normal load fluc-
tuations, excess air requirements and fuel analysis were studied,
along with the boiler reliability of the George F. Weaton Power
Station.
— Determinations of the relationship between generating facility
control settings and operating conditions of the boiler and the
resulting emissions.
— Collection of quantitative data on pollutants for the purpose of
establishing realistic emission parameters for varied operating con-
ditions.
— Determination of realistic baseline operating parameters as required
for the accurate and economical operation of the Citrate FGD
Process.
779
-------
TABLE 1. Preliminary Results of Baseline Testing
Unit No. 1 - G. F. Weaton Power Station
GAS ANALYSIS (to stack) COAL ANALYSIS (dry basis)
S02 1990 PPM Ash 13.1%
SOs 8 PPM Sulfur 3.04%
NOX 350 PPM
Cl 70 PPM
Particulates 0.024 Gr/DSCF
• Acceptance Test - For the purpose of fulfilling the contractual
agreements relative to system performance and operation, a ten
consecutive day test of the Citrate Process Demonstration Plant
will be made. During the acceptance test, the G. F. Weaton sta-
tion will be fired with coal containing 3.0% percent sulfur. Perform-
ance and operating parameters to be measured during the acceptance
test include:
— Sulfur content of coal fuel.
— Average boiler load.
— SC"2 removal efficiency, percent.
- Sulfur product output, STPD.
- S02 in outlet flue gas, PPM and Ibs/MM BTU input.
— Sulfur assay.
— Electrical power requirements.
— Steam requirements.
— Particulate in absorber inlet flue gas, Gr/SCF.
— Particulate in absorber outlet flue gas, Ibs/MM BTU input.
— Process feedstock requirements.
— Sulfate concentration in lean absorber liquor.
— Reheat capability
• One-Year Demonstration Test — Following the completion of plant
performance testing and final acceptance, a one year demonstration
test will be initiated. The objectives of the demonstration test are:
— To characterize completely the Citrate FGD system with respect to
the various system operating parameters.
— To determine the system's optimum operating conditions.
— To establish the long-term system reliability.
— To assess the environmental impact of the Citrate FGD system.
— To define the technical and economic feasibility of the Citrate FGD
system.
- To document the results of the test program such that comparisons
of the Citrate system with other flue gas desiilfurization systems can
be performed.
To accomplish these objectives, the one year demonstration program is designed to determine what
adjustments or process refinements are required to minimize the absorber liquid/gas ratio; system pres-
sure drop; feedstock consumption; power consumption; reducing gas consumption; and process capital
and operating costs; and to maximize sulfur oxide removal; particulate removal; system reliability; and
system availability.
780
-------
The one year demonstration test will be divided into three periods of approximately four months each.
Each period will consist of about two months of optimization testing followed by about two months
of steady-state testing.
During optimization testing, parameters will be changed frequently and the response of the system to
these changes will be noted. The removal of S02 will not be constrained during these tests. The more
promising operating conditions identified will be examined further in the steady-state runs.
The steady-state runs will be divided into two one-month test periods, each of which will involve opera-
tion at different steady-state conditions. During these tests the removal of 862 will be set at 90 percent
and other operating conditions will be varied. Parameters which can only be measured during longer
term testing (e.g., sulfate formation) will be investigated during steady-state testing.
A summary of the operational and performance tests to be conducted during the one year program
follows:
• First Optimization Test Series — During the first optimization
test series, emphasis will be placed on characterizing the absorber for
SC>2 removal and the prescrubber (venturi) for particulate removal.
A preliminary examination of the regeneration reactor section will
also be carried out. Because the S02 absorber and the reactor
section are closely coupled through the absorber feed rate and the
amount of S02 in the citrate liquor leaving the absorber, a signifi-
cant number of reactor section tests will occur naturally. Items of
lesser importance that will also be evaluated include the magnitude
of the rich citrate bypass, flotation air requirements, and turndown.
• First Steady-State Test Series — Two long-term tests will be con-
ducted during the first steady-state test series. Selection of the
operating conditions will be based on the results of the first opti-
mization test series. However, the variables to be emphasized are
the absorber L/G and the H2S/S02 ratios.
• Second Optimization Test Series — The second optimization test
series will stress the regeneration reaction section, the ability of the
system to follow load, the importance of temperature in the absorp-
tion and regeneration steps, and the effect of citrate concentration.
• Second Steady-State Series - This test campaign is very similar to
the first steady-state test series. Again, the variables emphasized
will be the absorber L/G and the H2S/S02 ratios. However, the tests
will be run with a reduced citrate concentration and the effect of
reheat will be studied. Also, the effect of operating the absorber and
regeneration reactors at an off-design temperature for longer time
periods may be examined.
781
-------
Third Optimization Test Series - The third set of optimization runs
will examine variations in reactor configuration, power plant load
following, operation on higher sulfur coal fuel, and the effect of
kerosene addition to the flotation tank.
Third Steady-State Test Series - Conditions for the final steady-
state run will be selected based on the cumulative experience gained
testing the unit. The SC>2 removal will be fixed at 90 percent during
this period and the citrate plant will be required to follow the load
on the boiler. This final steady-state run will characterize the citrate
process at the optimum conditions that can be achieved at the
demonstration plant. Chemical and utility consumption, unit avail-
ability and reliability, and sulfate formation rate at a given SC>2
removal rate will be studied during this period.
Special Studies — Some aspects of the demonstration testing of the
citrate process require special study. These aspects are discussed
separately because they do not directly influence the absorption or
regeneration steps and, therefore, are not readily examined with the
approach used for routine demonstration testing. The areas of
special concern during the demonstration tests are sulfate formation,
corrosion, and reheat requirements.
— Sulfate Formation — One of the advantages of the Citrate
Process is the suppression of sulfate formation. The total sulfate
being introduced into the Citrate Process will be monitored.
The sources of sulfate include:
— Sulfuric acid mist in the inlet flue gas.
— Air oxidation of sulfite during absorption.
— Disproportionation during regeneration.
— Air oxidation during flotation.
— Formation in the melting/decanting stage.
— Corrosion — The materials used in the FGD system are import-
ant factors in determining both the reliability of operation and
the capital costs. In the demonstration plant, emphasis is on
reliability, and therefore, higher quality materials than may be
necessary were specified. Operation of the test facility will
provide an opportunity to test and evaluate other materials
which may ultimately lower system costs or provide a backup in
case of failure of' present materials. The corrosion topic is
further discussed in the material evaluation section of this
paper.
782
-------
Reheat Requirements - The effectiveness of flue gas reheat in
the Citrate system by air injection from a steam/air heat ex-
changer will be assessed. An oil-fired burner type reheat system
was also installed to allow a technical and cost-effective com-
parison of the indirect steam reheat versus oil reheat.
783
-------
MATERIALS EVALUATION:
Corrosion testing was begun at the Albany Metallurgy Research Center of the Bureau of Mines following
observation of corrosion within the unit operations associated with the Citrate Process pilot plant.
Apparatus was assembled in the laboratory to simulate the following environments of the demonstration
plant:
• Sulfur vapor at 1,150°F to 1,450° and 1 atmosphere pressure. An
alternate atmosphere in this category is sulfur vapor plus steam in the
same temperature interval.
• Gaseous atmosphere H2S-C02-H20 at 200° F to 1,150°F and 1
atmosphere pressure.
• Rich citrate solution in the absorber containing 5 to 15 g/l SC>2 at
120°F and ambient pressure.
• Lean citrate solution in the sulfur melter and decanter containing
no S02 at 260° F and 4 atmospheres pressure.
A summary follows of test results obtained to-date. The investigation is not complete except for eval-
uating refractory materials in sulfur vapor and sulfur vapor containing steam. All other results must be
considered tentative pending replication in future tests. Final reports will be issued as Bureau of Mines
Report of Investigations.
Laboratory Tests in Sulfur Vapor and Sulfur Vapor Containing Steam
Three castable refractories were evaluated for application in the h^S generator. Cylindrical specimens
(3/4 in. dia. x 1-1/4 in. long), after exposure for 30 days at 1,450°F to sulfur vapor, steam and sulfur
vapor containing approximately 50 percent steam, were tested for compressive strength and compared
with similar specimens heated in air under the same conditions of time and temperature. The data,
as shown in Table 2, indicates significant loss of strength for High Strength Brik Cast G in sulfur plus
steam. A. P. Green Co. KS-4 and Gunning Refractories Co. HYDRA-SEH4 were satisfactory under all
test conditions.
TABLE 2. Compressive Strength (psi) of Castable
Refractories After 30 Days at 1450° F
Test Atmosphere
Material Air Sulfur Sulfur 50 pet. Steam Steam
A.P. Green Co. 2250 2100 5500 3450
KS-4
General Refractories Co. 3650 5000 550 4450
High Strength Brik Cast G
Gunning Refractories Co. 3850 3250 2200 2600
HYDRA-SEH4
784
-------
Most metals were severely attacked by sulfur vapor at high temperature, as shown by the preliminary
test results listed in Table 3. Chromium is the principal element deterring sulfidation of iron-base
alloys, and nickel provides additional resistance at 1,150° and 1,250° F. The austenitic stainless steels
were attacked more than ferritic alloys at 1,350° F, probably because of the tendency for nickel to
lower the melting temperature of sulfide scales.
Alonizing, a diffusional process which enriches a metal surface in aluminum, was very effective for
decreasing sulfidation of iron-base a|loys.
Titanium was found to be very resistant to sulfur vapor in a single test at 1 ,250° F.
TABLE 3. Corrosion of alloys in 1 atm. Sulfur Vapor
mils per year (mpy)
Alloy Temperature, ° F
1150 1250 1350
Carbon Steel >14004 - -
Carbon Steel, Alonized - 4 19
CRM-4 (Fe-6AI)3 - >2504
Type 406, Alonized — 5 —
Type 446 120 140 80
E-Brite26-1 - 115 80
Type 316 43 110 160
Type 31 6, Alonized - 9 12
Type 310 43 115 205
Type 310, Alonized - 7 27
Hastelloy C-276 42
Titanium. — 27 —
Laboratory Tests in ^S - 16 pet. CC>2 - 16 pet. H2O at 200° to 1,150° F
Remarkably similar corrosion rates for iron and nickel-base alloys containing greater than 12 percent
chromium were measured in the subject atmosphere with and without intermittent exposure to steam
and thermal cycling to room temperature. Corrosion rates for Incoloy 800; Inconel 601; types 304,
31 6L, 31 7L, 310, and 446 stainless steels; E Brite 26-1; Uniloy 18-2; and MP-35N lie within the fol-
lowing ranges at the temperatures indicated; 600° F< 2 (mpy); 800° F, 5±4 mpy; 950° F, 20±15 mpy;
1 ,075° F, 50±30 mpy; and 1 ,1 50° F, 100±50 mpy.
Aluminum as a binary additive to iron was not beneficial in the test atmosphere, but it reduced sulfi-
dation considerably in combination with chromium. For example, Armco 18SR (18Cr-2AI), type 406
(13Cr-4AI), and Hoskins 815 (26Cr-4 6AI) formed adherent sulfide scales and corroded at 37, 10, and 6
mpy, respectively, at 1 ,1 50° F.
3 Experimental iron base alloy of Chrysler Corp.
4 Specimens were penetrated in the test.
785
-------
Laboratory Tests in Citrate Solution
Stainless steels and Fe-Cr-Ni base alloys including Incoloys, Inconels and Hastelloysaswell asTi and Zr
were exposed for 30 days in static salty citrate solutions containing either 10 g/l S02 (rich Citrate) or
less than 0.2 g/l S02 (lean Citrate). The nominal composition of the test solutions and other conditions
are listed in Table 4. Three coupons 1-1/2 in. x 3/4 in x 1/16 in. were exposed in 400 ml of rich citrate
solution through which flue gas was passed via a bubble pump to effect stirring. One of the coupons
was half immersed, the second fully immersed, and the third was fitted with a nylon bolt, nut, and
washer to invite crevice corrosion. Glass vessels containing 500 ml of lean citrate solution and two
coupons were pressurized with air. Both coupons were fully immersed, and one was provided with a
crevice. General and localized corrosion (crevice corrosion and pitting) were determined by weight
loss, by visual examination at low magnification, and by metallography.
TABLE 4. Test Conditions and Nominal Composition of
Citrate Solutions
Component
Citric Acid
NaOH
Na2s2°3
N32S04
NaCI
S02
Rich Citrate Solution
96 g/l
40
63
71
83
5 to 10
Lean Citrate Solution
96 g/l
40
63
71
83
<0.2
Flue Gas Bubbled
Atmosphere Through Solution Pressurized with Air
pH 4.2 to 4.3 4.4 to 4.5
Temperature, ° F 120 260
Pressure, atm 1 4
Time, Days 30 30
Unmodified ferritic stainless steels and ferritic stainless steels containing 2 to 4 percent Al or 1 to 2
percent Mo corroded excessively (> 100 mpy) and are not suitable for use in either lean or rich citrate
solutions. In general, the austenitic alloys corroded with rates less than one mpy and differed primarily
in their resistance to localized corrosion. As shown in Table 5, the resistance of Fe-Cr-Ni base alloys
to localized corrosion increased with Mo + W content. In rich citrate solutions at 120°F, alloys con-
taining 6 percent (Mo + W) were resistant. In lean citrate solutions at 260° F, Fe-Cr-Ni base alloys
containing less than 6 percent (Mo + W) were still susceptible to localized corrosion, and only alloys
containing 9 percent or more (Mo + W) were completely resistant. Ti and Zr were resistant to corrosion
in both rich and lean citrate solutions.
786
-------
TABLE 5. Resistance of Alloys to Localized Corrosion
(Crevice Corrosion and Pitting) in Citrate Solutions
Alloy
Hastelloy C-276
Hastelloy C-4
MP-35N
Inconel 617
Inconel 625
Hastelloy G
Allegheny Ludlum 6X
RA333
N-155
Titanium
Zirconium 705
Sandvik2RK65
Incoloy 825
Type 317 L
Type 316L
Type 316
Carpenter 20Cb-3
RA330
Hastelloy B5
Inconel 600
Incoloy 800
Type 304
Mo
16
15
10
9
9
6.5
6.5
3
3
0
0
4.5
3
2-3
2-3
2-3
0
28
0
0
0
W
0
0
0
0
0
0
0
3
3
0
0
0
0
0
0
0
0
0
0
0
0
0
Tendency for Localized Corrosion
Rich Citrate Lean Citrate
Solution, 120° F Solution, 260° F
None
None
None
None
None
None
None
None
None
None
None
Slight /
Slight
Slight
Slight
Slight
Moderate-Severe
Moderate-Severe
Moderate Severe
Moderate-Severe
Moderate-Severe
Moderate-Severe
None
None
None
None
None
None
None
Slight
Slight
None
None
Moderate-Severe
Moderate-Severe
Moderate-Severe
Moderate-Severe
Moderate Severe
Moderate-Severe
It should be noted that static conditions might encourage localized corrosion and minimize general
corrosion, whereas flowing solutions might increase general corrosion and decrease the tendency for
localized corrosion. Consequently, the laboratory test results may disagree with the results of in-plant
testing.
Contains only 1 percent Cr compared to > 12 percent for most other alloys.
787
-------
APPLICATIONS AND FEATURES:
In addition to coal and oil-fired utility and industrial boilers, the Citrate Process can be applied to
sulfur dioxide emission discharges from the refinery and metallurgical industries. (6) Engineering
and economic designs for Citrate Process installations can reflect the site specific requirements of
smelting processes, and the combustion products associated with petroleum and chemical operations.
At petroleum refineries the H2S derived from sour crude can be used for citrate regeneration. Thus,
the Citrate Process permits refiners to improve the economic balance between use of high sulfur fuels
for boiler plant operations and the necessary abatement of S02 discharges. A Citrate facility installation
would also permit integration of other stack gas desulfurization requirements, such as required for
process headers, Claus and acid plant tail gas, and fluidized catalytic cracker emissions. Typically, a
centralized Citrate regeneration facility would be provided to serve a variety of point sources located
throughout the facility.
Sulfur dioxide emissions from steel mill iron ore roasters can be reduced to elemental sulfur with H2S
derived from coke oven gas (COG). More than half the hydrocyanic acid present in COG is stripped
in the wet removal of hydrogen sulfide. HCN must be separated from the H2S prior to its use in citrate
regeneration. Various methods are available for separating HCN including aqueous scrubbing and
catalytic cracking. Economic design would include the balancing of H2S derived for Citrate regeneration
versus sulfur dioxide formed in the incineration of the COG.
The economic limits of industrial application for the Citrate Process include treatment of flue gas dis-
chargers ranging from five volume percent S02 to as low as 500 PPMV. Greater than 90 percent removal
of S02 can be achieved and cleaned gas can contain less than 25 PPMV S02- The specific features
of the process which permit this range of performance and versatility include:
• S02 removal efficiencies exceeding 99.0 percent have been docu-
mented.
• The Citrate Process has a high capacity for short-term S02 overloads
and can accomodate rapid^ load fluctuations and variable S02 con-
centrations.
• Marketing the elemental sulfur product is less costly in storage and
transportation than sulfuric acid. The high purity sulfur product is
of feedstock quality for many industrial processes.
• Flue gas conditioning and S02 absorption are free of scaling and
plugging problems; precipitation of sulfur takes place in the regen-
eration section outside of the absorber.
788
-------
• Low rate of oxidation to sulfate. Less than 2 percent of the entering
S02 is converted to sulfate which can be selectively purged by cry-
stallization; the sulfate is recovered as Glauber's salt,
• Low process energy demand in terms of both process steam and
electrical power. A utility application requires approximately
three percent of station energy output.
• Minimal environmental impact. There is no significant sludge dis-
posal requirements necessitating ponding or voluminous lagoons.
Depending upon the character of the entering waste gas, disposal. of
particulate and acids is required. The citric acid reagent is a non-
toxic biodegradable organic which eliminates risk of further environ-
mental deterioration.
• The citrate solution regeneration facility can be physically separated
from the gas quench and absorption operation allowing each to
operate independently. This feature is more advantageous at space-
limited sites or sites where regeneration is more economically per-
formed at a central location other than the emission source.
.• The process can use other organic acids such as glycolic acid in lieu
of citric acid, should other organics be shown to have improved
performance.
• The citric acid reagent serves as a pH buffer and does not directly
enter the absorption chemistry.
• A low absorber liquid-to-gas ratio results in reduced pumping and
fan head losses.
• Use of the Citrate Process in applications where h^S is available
without on-site generation will recover the sulfur value of the
gas as well as the absorbed SC>2.
789
-------
SUMMARY:
Fulfillment of the objectives of the Citrate Process Demonstration Program will provide an attractive
alternative for the abatement of S02 discharges to industry. Most importantly, the demonstration test
will establish credible basis for informed decisions by industrial and utility managers for application of
Citrate technology. Within the next year, the system performance will be documented; the reliability and
availability of the Citrate system will be established; and the actual costs associated with plant operations,
maintenance, and utility/feedstock consumptions will be reported.
790
-------
REFERENCES:
1. CLEAN POWER FROM COAL: THE BUREAU OF MINES CITRATE PROCESS,
U.S. Department of the Interior, 1978; pp 1 -11.
2. Korosy, L, et al., CHEMISTRY OF S02 ABSORPTION AND CONVERSION TO
SULFUR BY THE CITRATE PROCESS. Presented at 167th American Chemical
Society Meeting, Los Angeles, California, April 5, 1974; 32 pp.
3. Rosenbaum, J.B., et al., SULFUR DIOXIDE EMISSION CONTROL BY HYDROGEN
SULFIDE REACTION IN AQUEOUS SOLUTION - THE CITRATE SYSTEM. Bu
Mines Rl 7774,1973; 31 pp.
4. Madenburg, R. S., et al., CITRATE PROCESS DEMONSTRATION PLANT - A PRO-
GRESS REPORT, Presented at U.S. Environmental Protection Agency, FGD Sym-
posium, Miami, Florida, November, 1977; pp 4-10.
5. Paulsrude, D. M., et al., COMMERCIAL APPLICATION OF THE CITRATE FGD
PROCESS, Power Magazine — Energy Management Guidebook, 1977 Edition; pp 19-24.
6. Turpin, F. G., et al., INDUSTRIAL APPLICATION OF CITRATE FGD TECH-
NOLOGY, Air Pollution Control Association, Houston, Texas, June, 1978; pp 1-14.
791
-------
DESIGN AND COMMERCIAL OPERATION OF LIME/LIMESTONE
FGD SLUDGE STABILIZATION SYSTEMS
Ronald J. Bacskai
Lee C. Cleveland.
IU Conversion Systems, Inc.
115 Gibralter Road
Horsham, PA 19044
Presented at
ENVIRONMENTAL PROTECTION AGENCY'S
FGD SYMPOSIUM
Las Vegas, Nevada
March 7, 1979
792
-------
ABSTRACT
Installation and operation of flue gas desulfuriz-
ation (FGD) systems in coal-fired utility boilers are
becoming essential to combat air pollution and
meet the Clean Air Act S02 emission standards as
promulgated by government regulatory agencies.
All non-regenerative FGD systems produce large
volumes of low percent solids sludges, which create
a waste disposal problem for the utilities. These
sludges, unless properly treated and disposed, can
cause water pollution and damage the environ-
ment.
The Resource Conservation and Recovery Act
will soon mandate standards for SC>2 scrubber
sludge and ash waste disposal. It is expected that
the FGD wastes will have to be chemically stabiliz-
ed to minimize environmental degradation and to
produce a structural product to minimize land use.
Stabilization of FGD S02 sludge and fly ash
waste has, therefore, become an important factor
in our country's ability to meet environmental
standards with corresponding power generation
growth.
Although there are many SC>2 scrubbers current-
ly operating with large volumes of sludge being
produced annually, there' is limited operating
experience on full scale dry stabilization systems.
That is particularly true of high sulfur eastern
coals. This paper discusses the only four commer-
cially operating plants on eastern high sulfur coal.
The stabilization process is the IU Conversion
Systems Poz-0-Tec System which involves de-
watering of the scrubber sludge followed by
mixing with waste coal ash and chemical additives
to produce a structurally stable and environ-
mentally compatible material for landfill, land
reclamation, roadbase, liners ahd other useful
applications.
A discussion is provided on the process and
plant design considerations which must be evalu-
ated to ensure successful plant operations. Pro-
blems encountered in designing and initially oper-
ating SC-2 Waste Treatment Systems for waste
materials that were not available for evaluation
and testing prior to the completion of the multi-
million dollar facilities will be reviewed. The de-
sign and operation of these systems involves ex-
pertise in material handling, chemical processing,
geotechnical engineering and regulatory require-
ments. Actual "hands-on" operating and main-
tenance experience is reviewed.
Emphasis will be placed on: 1) plant design pro-
blems and solutions for them; 2) variances'in S02
scrubber sludge and coal ash characteristics and
their effect on equipment design, plant operation
and chemical stabilization; 3) the handling of the
waste materials including the dewatering charac-
teristics of sludges and the controlled feeding of
fly ash at higher rates than used previously in any
industry; and 4) flexibility of choosing disposal
sites with the use of the stabilized material pro-
duced in a "dry" system.
Five years in this field and the experience
derived from the operation of four major waste
treatment systems and from design of eleven addi-
tional treatment systems has resulted in a sub-
stantial knowledge of all facets of such waste treat-
ment systems and the development of more effi-
cient and- more economical processing plants to
serve industry.
INTRODUCTION
With the uncertainty of natural gas and pe-
troleum imports, and the growing public concern
over nuclear power plants, more attention has been
focused on the utilization of coal, the most abun-
dant energy resource in the United States, for gen-
eration of electric power. With this expansion of
coal usage and the increased necessity for environ-
mental controls, it has become necessary for
utilities to install and operate a growing number of
flue gas desulfurization (FGD) systems for S02
removal. It is estimated that nearly 60,000 MW of
FGD capacity will be installed by 1980.*
*PEDCO Environmental - February, 1979
793
-------
FIGURE 1
7\
c=
— ^
DUST COLLECTOR
/
EMERGENCY POND
r
=3—
]
-=D
n
FLY ASH
SILO
y
[A A 7
DUST COLLECTOR
SCREW FEEDER
Figure 1 Process flow
Among the various FGD scrubber systems avail-
able, wet lime/limestone scrubbing and double
alkali (indirect lime/limestone) scrubbing have
gained the most industry acceptance. In excess of
90% of all scrubbers installed or committed are of
this type. These scrubbing operations produce an
enormous volume of low solids content sludge,
which must be properly disposed so that ground
water and surface water is not polluted by un-
acceptable concentrations of heavy metals and dis-
solved solids.
A solution to this massive sludge disposal pro-
blem is chemical stabilization of scrubber sludge by
the Poz-0-Tec process to prevent significant en-
vironmental damage and minimize land disposal
requirements. Essentially, the process involves
treatment of S02 sludge with fly ash and one or
more additive(s) to produce, via pozzolanic reac-
tions, a stabilized material, suitable for landfill
794
disposal. A simplified process flow diagram is
show in Figure 1.
There has been extensive literature written on
the Poz-0-Tec process regarding its physical and
environmental properties, and this paper will not
attempt to address these issues. In summary, the
process has been tested on pilot plant applications
as early as 1973 at Southern California's Edison
Mohave Power Plant; it has been thoroughly evalu-
ated by the EPA at TVA Shawnee with extensive
test reports circulated throughout the industry,
and it has been tested and evaluated by Louisville
Gas & Electric at the Cane Run Plant under an
EPA sponsored program. In addition to extensive
lab, pilot tests and demonstrations, the Poz-0-Tec
system has received large scale commercial accep-
tance on a wide variety of scrubbers as demon-
strated by the Utilities committed to the full scale
process applications as shown in Figure 2.
-------
FIGURE 2
IU CONVERSION SYSTEMS, INC.
SO2 SLUDGE STABILIZATION CONTRACTS
UTILITY
Big Rivers Electric
Corp.
Central Illinois
Public Service Co.
Cincinnati Gas &
Electric Co.
Columbus & Southern
Ohio Electric
Commonwealth Edison
Duquesne Light
Company
Duquesne Light
Company
East Kentucky
Power Cooperative
Indianapolis Power &
Light
Indianapolis Power &
Light
City of Lakeland
Southwestern Electric
Power Company
Texas Municipal
Power Agency
STATION
Reid #1 & 2
Newton #1
East Bend #2
Conesville #5 & 6
Powerton #5
Phillips
Elrama
Spurlock #2
Petersburg #3
Petersburg #4
C.D. Mclntosh Jr.
#3
Henry W. Pirkey
Gibbons Creek
SCRUBBER
Lime
Double Alkali
Lime
Lime
Limestone
Lime
Lime
Lime
Limestone
Limestone
Limestone
Limestone
Limestone
MW
480
575
600
820
450
400
500
500
515
515
350
720
400
Waste
Materials T/Yr.
855,000
989,000
1,106,000
995,000
608,000
456,000
629,000
761,000
620,000
886,000
470,000
1,146,000
788,000
795
-------
Figure 3 Initial processed material
The Poz-0-Tec process is a complete waste man-
agement system for coal fired power plants. It
blends fly ash, bottom ash (if desired), scrubber
sludge, lime and other additives. Concentrated
streams from the evaporator and cooling tower
Figure 4 Material after final placement
sludge can also be incorporated. The stabilized
material is a cementitious material, and with pro-
per placement and compaction, exhibits low per-
meability and superior structural properties. The
material in its initial processed form and after final
placement is shown in Figure 3 and 4.
796
-------
PROCESS CONSIDERATIONS
Sludge disposal has become a major considera-
tion in the design of a power plant and therefore,
should not be an afterthought. It should not be
left until the rest of the plant has been designed
because good sludge disposal is dependent on many
aspects of the power plant design including the
boilers, coal, type of scrubber, reagent, thickener
design, etc.
The stabilization of power plant wastes involves
more than just combining the wastes by them-
selves or with an additive. Each of the waste ma-
terials contributes chemically and physically to
the process, and variations in those materials must
be considered in developing the specific process
design.
Fly Ash
Fly ash is utilized in the Poz-0-Tec process for
several reasons. It is a waste material which must
be disposed of and is usually available at the same
source as the sludge waste. It is a fine particle ma-
terial and provides the alumina and silica which
are necessary for the pozzolanic reactions to tie up
the sulfur compounds of the sludge.
As the Poz-0-Tec system is a dry stabilization
method, the quantity of fly ash will also contribute
to the final solids content of the product and
effect its handleability. Generally, ash to sludge
ratios of 1:1 or higher will result in an immediately
placeable material; those below that ratio will
usually necessitate interim stockpiling prior to
final placement.
Particle size of the ash also contributes to the
process chemistry. The finer the particles, usually
resulting from modern precipitators, the more reac-
tive the ash. Fly ash from both mechanical col-
lectors and electrostatic precipitators will have a
large range of particle size which may reduce its
reactivity. Cyclone boilers generate large quanti-
ties of bottom ash rather than fly ash. That ash is
reactive but is of large particle size and usually
collected wet. The ash can be ground to provide
adequate particle size for reactivity, but the cost
of grinding and the additional moisture make
stabilization more expensive and more difficult.
Additives used in the precipitator to improve fly
ash collection can cause chemical and safety pro-
blems in operating stabilization systems. Ammonia-
based additives, for example, can cause a signifi-
cant reduction in strength characteristics of land-
filled material. Further, ammonia gas is released
when the ash is in contact with an alkaline ma-
terial, such as lime. The quantities of ammonia
released are considerably higher than permitted by
OSHA and such processing facilities would require
extensive air cleaning equipment. Even then, the
ammonia odor is present from the product stock-
pile and from the landfill.
Scrubber Sludge
Sludge generated at power plants presents a
major disposal problem, from both the physical
and environmental aspects. The characteristics of
sludges from FGD systems vary greatly, depending
on the coal burned, boiler, scrubber, reagent used
and other factors.
The chemical composition of a sludge is one of
the most important considerations in designing a
stabilization system, because it can vary greatly,
even during standard power plant operation.
One difficulty in developing the stabilization
process for a specific power plant is that the scrub-
ber sludge is usually not available prior to the start
of actual scrubber operation. As a result, know-
ledge and experience in stabilizing and handling
various types of sludges becomes quite important
in establishing the proper process design para-
meters.
All FGD sludges can be stabilized, but it is im-
portant to understand those characteristics of
sludge which have the greatest potential effect on
stabilization systems.
Sulfite/sulfate proportions primarily affect de-
watering. The larger size of the sulfate particles
affords easier dewatering. However, given the same
ash to sludge ratio, sulfate based sludges require
a numerically higher solids content of the final
product to be equally handleable than do sulfite
based sludges.
Other materials, such as silica, iron, etc. which
exist as trace elements in prototype scrubber
testing, usually show up as grit in the sludge.
Some scrubbers actually produce a cement type
material in quantities of 1,000 pounds per day.
While that does not affect the chemical stabiliza-
tion of the sludge, it does affect plant operations
and is addressed later in the paper.
The lime or limestone used in scrubbers also
have their effects on process design. Poor quality
reagent will require that more be added in the
scrubber to achieve the required S02 removal and
the high proportion of non-lime materials in-
creases loads on the dewatering equipment. When
in the form of grit, it causes extensive wear on
piping and process equipment.
Chemical additives used by the utility to en-
hance dewatering in the thickener can have signifi-
cant affect on dewatering. The flocculents which
are formed in' the sludge to provide increased
settling in the thickener are broken down by the
shearing action of centrifugal thickener under-
797
-------
flow pumps and piping systems and cause the
vacuum filter medium to become partially blinded.
There are several additives available which will
provide the necessary thickener settling but mini-
mize filtration problems. These are recommended
to the utility where Poz-0-Tec systems are opera-
tional.
Solids variations from a thickener are often
the result of an upset in scrubber operation or
inattention to thickener operation. As thickener
solids have a direct relationship to filter cake
solids and ultimately final product solids, reduced
solids underflow could result in increased disposal
costs due to the need for additional handling of the
product and the additional quantity of product.
Process Additives
Most stabilization processes require that some
sort of additive be used to initiate chemical reac-
tions. Although this activator may already be pre-
sent in some coals, such as lignite, it must be added
separately for most conditions.
For pozzolanic stabilization, the additive most
used is lime and it is available as pebble lime re-
quiring crushing, pulverized quicklime, hydrate or
lime slurry. The important considerations for the
additive lime are (1) CaO content, as that is the
necessary activator for the chemical reaction, and
(2) lime particle size distribution.
PLANT DESIGN CONSIDERATIONS
Concurrent with the evaluation of process
variables to achieve the chemical stabilization,
the physical processing systems must also be
planned. And, as is true of the power plant and
scrubber design, large quantities of materials are
involved.
A dry stabilization plant is a materials handling
system which processes liquids, sludges, damp
solids and dry solids in one facility. Figures 5 and 6
show the stabilization facilities at Columbus &
Southern Ohio Electric Conesville Station and
Indianapolis Power and Light Petersburg Station.
Siting
The location of the stabilization facility will
depend on land availability at the power plant,
location of the disposal area, and consideration of
other factors, both physical and economic.
A location adjacent to the thickeners will mini-
mize pumping and piping costs for underflow and
filtrate. Additional savings could be realized if the
facility were located next to the power plant ash
silos to permit direct feeding of the ash to the
facility and eliminate pneumatic conveying be-
tween the main silo and surge silo.
Conversely, locating the facility at the disposal
area will necessitate extensive transportation
systems for liquids and solids. There are also
Figure 5 Columbus & Southern Ohio Electric Conesville Station - SC>2 stabilization facility
798
-------
limitations for pneumatic ash conveying systems
with respect to distance and economics. Pneu-
matic trucks could be used to transport ash for
long distances but that cost would be prohibitively
high.
Advantages to locating the facility at the dis-
posal area include reduced transportation costs for
the processed material. In some cases, the product
could be transported directly by conveyor.
A 600-800 MW plant producing 1,000,000 tons
per year of stabilized material will require about
150 acres of disposal area 100 feet high over a 20-
year period. Regardless of the plant location and
method of product transport to the landfill, the
daily delivery of material will normally involve
trucks or other hauling equipment.
As can be derived from the above discussion,
the most advantageous situation is to have the
landfill located near the power plant to benefit
from reduced transportation system costs for both
the waste materials and the stabilized product.
Figure 7 shows the site layout of the I DCS fa-
cility at Indianapolis Power & Light Petersburg
Station.
Dewatering
Dewatering of the sludge is important in dry
stabilization systems because the more liquid
removed, the less ash that is necessary to improve
handleability and the higher the final product
solids. Some scrubber systems, such as a double
alkali system, include thickening and filtration as
part of the scrubber operation to reclaim the
sodium liquor for reuse. Most FGD systems, how-
ever, only thicken the sludge to 25-30% solids.
Dewatering in the stabilization facility is usually
accomplished by vacuum filtration, although
centrifuges, hydrocyclones and other methods have
been considered and used in some limited applica-
tions. Figure 8 shows the vacuum filters installed
at Columbus & Southern Ohio Electric.
Scrubber sludge can be vacuum filtered at rates
of 50-100 Ib/sf/hr. and sometimes higher, depend-
ing on the chemical composition of the sludge, the
filter medium, i.e. cloth, screen, etc. and filter
aids. Sulfate sludges usually yield higher filtration
rates and solids than sulfites. Excess lime concen-
trations and sheared flocculents in the sludge may
cause cloth blinding.
Filtration will usually dewater a lime based
sludge from 30% solids to 42-60% solids and lime-
stone based sludge to 55-60% solids. Oxidized
sludges are reported to achieve 80-85% solids,
which when mixed with fly ash and additive,
would result in a high solids final product. This
product may have to have liquid added to achieve
optimum placement density. Figure 9 shows effec-
tive dewatering ranges for various scrubber sludge
compositions.
Figure 6 Indianapolis Power & Light Petersburg Station — SC>2 stabilization facility
799
-------
POWER PLANT
3500 FT. NORTH
BOTTOM ASH
DECANT BINS
"ELECTRICAL
BUILDING
ADDITIVE SILO
PROCESSING BUILDING
LANDFILL
110 ACRES ^
INDIANAPOLIS POWER AND LIGHT
PETERSBURG STATION UNIT #3 - 515 MW
SO2 SLUDGE AND ASH WASTE TREATMENT SYSTEM LAYOUT JT
FIGURE 7
Site layout
Centrifuges have been used for dewatering but
are more susceptible than filters to variations in
sludge composition, solids content and grit.
Hydrocyclones have been offered as a substitute
for thickeners and filters but actual field experi-
ence is limited. A major drawback to hydrocy-
clones is the 5% plus solid in the return liquid,
which normally exceeds the minimum for the
scrubber liquid. Also, hydrocyclones have a critical
liquid solids separation phase for which particle
sizes below 10 microns are not removable. Most
FGD sludges have particle sizes below the 10 mi-
cron size, thereby rendering the hydrocyclone in-
effective.
Vacuum filter aides to improve yield and solids
can be both mechanical and chemical. Vibrators,
compression rolls and flappers can provide signifi-
cant improvement in filtration. Lime, fly ash and
chemical polymers can also provide improvement
in filtration when added in the proper proportion.
However, too much fly ash in the sludge, a condi-
tion which occurs when ash collection equipment
is in poor repair, causes blinding of the cloth and
actually reduces filtration rates.
800
-------
Materials Feeding
Feed systems for fly ash and lime involve more
than just adding these materials to the sludge. For
situations where there is limited ash available,
controlled feed is important to conserve ash and
ensure that all sludge has ash available for the
chemical reaction.
Fly ash is similar to water in that it acts like a
liquid, and therefore requires positive control when
being fed. Most utilities are experienced in handl-
ing ash using collection hoppers, hydrovac systems,
and pneumatic conveyors. However, their primary
interest is in just moving it, not feeding it at 20-70
tons per hour with a 2-5% accuracy.
IUCS has used several types of ash feeding
systems and the two which work best are: 1) a
mass flow system; and 2) a gravimetric feeder with
a live bin bottom silo. This equipment not only
feeds accurately but controls flooding. Those who
have witnessed a fly ash flood involving hundreds
of tons of fluid ash will appreciate the importance
of having a quality fly ash feed system.
As lime constitutes a small percentage of a mix
on a dry weight basis and is the activator of the
chemical stabilization reactions, accuracy and dis-
persion of the lime in the product is absolutely
necessary. There are a number of feed systems
which have the capability to feed the 2-5 tons per
hour of lime within the 5% accuracy considered
necessary. Dispersion within the mix, however,
depends upon accuracy of feed, location of lime
feed into the system, particle size, and quality of
mixing.
Mixing
Mixing is the physical combining of the waste
and additive material to provide a homogeneous
blend to permit the fly ash and additive to contact
all sludge particles so complete chemical action
can take place.
The mixer must be able to provide the required
blending even though the ratio of wet and dry ma-
terials may vary over any given period and the
mixer designed for 200 TPH may only be operating
at 100 TPH due to reduced station load.
The specific combination of waste materials to
be mixed at a facility must be evaluated for ma-
terial ratios, solids content, particle size, retention
time, type additive, etc. to ascertain the proper
mixer design.
The most difficult material to handle in a
mixer is filter cake. It clings to the mixer housing
and paddles; it may not blend with the other ma-
terials if not mixed for the proper duration. If
mixed too much, it will become thixotropic and
even with the ash mixed in, a semi-fluid material
will exit the mixer.
Figure 8 Vacuum filters — Columbus & Southern Ohio Electric - Conesville plant
801
-------
FIGURE 9
SLUDGE DEWATERING CHARACTERISTICS
EFFECTIVE
SLUDGE DEWATERING
SCRUBBER COMPOSITION RANGE
Magnesium
Modified Lime CaS03/CaS04/MgSOx 42% • 50%
Lime CaS03/CaS04 50% - 60%
Limestone CaS03/CaS04/CaC03 55% - 60%
Dual Alkali CaS03/CaS04/Na2SO4 50% - 60%
Magnesium
Modified Lime
with partial fly CaS03/CaS04/MgSOx + 48% - 52%
ash removal f|V ash
With the I DCS system approach, the reactive
chemistry is such that variation in product consis-
tency can be tolerated and a satisfactory final
product still achieved.
Much of the equipment in the stabilization
facility is process related, such as dewatering,
feeding and blending. Other equipment is suppor-
tive, but of equal importance in designing a system.
Two of these, dust collectors and conveyors,
require special attention and will be discussed
briefly.
Dust Collectors
I DCS has a number of dust collector systems on
its plants, and experience has shown that these
systems pass some dust. Such things as particle
size, grain loadings, humidity, dew point, air to
cloth ratio, and quality and quantity of purge air
have a significant effect on dust collector per-
formance.
Dust collectors for mixers where wet and dry
materials are combined also warrant special consid-
eration. Depending on the additive used, the
mixing operation can produce a steam which
hinders dry dust collection. In some cases, a wet
scrubber is necessary.
Conveying
Belt conveyors are used in stabilization plants to
transport filter cake to the mixer, damp fly ash or
bottom ash to the mixer, final product to a stacker
and for the stacker itself.
Dry or slightly damp materials are not much of
a conveying problem. Wet materials, especially
those having thixotropic properties, are another
matter. That type material clings to belts, scrapers,
transfer points, and has to be physically removed
if the conveyor system is not properly designed.
It is important that all components of a belt
conveyor system be designed together. To install a
"super" belt cleaner without providing adequate
support against which the cleaner can act, will
result in a non-cleaned belt. To design a transfer
point without an adequate hopper will result in a
plugged transfer point or a pile of material on the
ground. This may seem basic, but the knowledge
comes from actual experience, not from the manu-
facturer of the equipment. Figure 10 shows a typi-
cal belt scraping system.
Landfill considerations must also be taken into
account when designing a stabilization facility.
Radial stacker capacity depends on hourly produc-
tion rates, consistency of the product, whether
interim conditioning is required before placement,
the disposal schedule and the equipment planned
for the transport of the product to the landfill.
Figure 11 shows the radial stacker installed at
Indianapolis Power and Light Petersburg Station.
PROCESS OPERATION AND MAINTENANCE -
SOME EXPERIENCES
IUCS has four stabilization facilities opera-
tional which involve a representative selection of
scrubber sludges, ash, additives, landfills and other
operational aspects. Scale-up of both process and
design considerations has been a learning process
and sometimes a difficult and expensive one. One
thing is certain: Scale-up from a 100 pound bench
scale mix to a 200 ton per hour, 24 hours per day,
7 days per week operation is a significant step and
IUCS has gained considerable experience by oper-
ating four plants for periods up to four years.
Three of these plants are owned and operated by
IUCS and this has become an excellent vehicle
by which to increase our knowledge of plant de-
sign requirements.
Much of what IUCS has learned in operating
plants is used to upgrade existing plants and those
presently under design or construction. These
improvements are also available to utilities who
are operating IUCS stabilization facilities.
Start-up of stabilization plants, although not as
complicated as a scrubber start-up, is still important
since the stabilization facility must produce a pro-
duct that will satisfy environmental regulations
and allow continuous scrubber operations.
Start-up consists of (1) equipment checkout,
(2) determination of equipment capability now
that the actual sludge and ash are produced and
(3) confirmation of the expected chemical reac-
tion. Operating personnel must be trained on the
equipment and quality control procedures.
Quality control is necessary on a continuous
basis as a tool to ascertain what is being received,
what should be expected when wastes are pro-
802
-------
cessed, what is actually produced and placed and
what process modifications are necessary when
variations occur. Therefore, the initiation of the
quality control program concurrent with the
equipment startup is a necessity.
Figure 10 Belt scraping system
Actual operational experiences can best be
described by reviewing some instances following
the flow of materials through a stabilization
facility starting with the scrubber sludge. Earlier,
a discussion on grit in the sludge was presented.
Figure 11 Radial stacker — Indianapolis Power & Light Petersburg Station
803
-------
This grit causes pipes to wear out, especially un-
lined pipes. Grit wears heavily on rubber lining of
pumps and valves causing replacement in less than
a year when 5 years life was expected. It settles
in pipe joints, tees, and elbows plugging them as
the "hybrid cement" hardens. It builds up in
surge tanks causing agitators, which were designed
for 30% suspended solids, to seize up. This pro-
blem is compounded in thickeners. Several utili-
ties, as well as one of our own plants, have ex-
perienced this problem. Removal of four feet of
settled and possibly reacted sludge/cement in a
thickener is time consuming and expensive.
For the most part, Utilities operate thickeners
with a primary concern for overflow clarity, since
this liquid is used for scrubber makeup. Thickener
underflow usually receives less attention. Under-
flow is usually specified contractually at a certain
percent solids, however in actual operations, solids
concentration vary widely. For example, at one
installation 30% solids at 300 GPM was specified,
although the stabilization plant received 20%
solids at 450 GPM. The lower solids produce lower
filtration yield and, at the higher than expected
flow rate, the system surge tank level increases to
a point where a back-up filter needs to be oper-
ated. When that occurs, the filtrate load exceeds
the sump and sump pump capacity and the cake
may exceed the conveyor and mixer capacity.
Excess lime in the scrubber may be necessary to
achieve compliance with emission requirements or
to minimize scaling. However, that excess lime
passes through the thickeners and affects the fil-
tration operation. The lime is usually fine particle
material, blinds the cloth, and significantly reduces
the filtration rate. Acid washing of the cloth has
been tried, and although somewhat effective, it
has met with resistance from the operating per-
sonnel who actually have to handle the acid. Older
boilers may use oil during combustion and that
also blinds filter cloth.
One answer to these problems is frequent filter
cloth changing. Given experienced peronnel, a
complete cloth change on a 12' diameter x 20'
length vacuum filter can be accomplished by 3 men
in eight hours. Labor and material costs per change
is $800-31,000 depending on labor rates and cloth
type. Changing cloths weekly, as has been done in
one of the stabilization plants, cost $50,000 per
year. The life cycle for filter cloth was expected
to be three months per the filter manufacturers
experience.
Temperature is a factor in vacuum filtration.
Sludge of higher temperatures will yield signifi-
cantly higher filtration rates than at lower tem-
peratures. When a utility installs several thousand
804
feet of uninsulated sludge pipe line four feet off
the ground in areas where the temperature and
wind chill factor are frequently below zero degrees
F, sludge temperature will drop considerably from
the utility's thickener to the processing facility.
Hence, consideration should be given to reheating
sludge prior to filtration as an alternative to addi-
tional filters to achieve the necessary production.
Filtration can be aided by the addition of lime,
at certain particle sizes, or fly ash in the sludge.
Improvements of 50% in filtration rates have been
obtained in pilot tests and full scale testing is
being conducted.
Previously, the importance of fly ash feed was
discussed. At C&SOE's Conesville Station, IUCS
installed vibrating pan feeders and live bin bottoms
on a 500 ton capacity silo. The vibration caused
cracking of the vibrating motor bracket attached
to the bin bottom without providing controlled
ash feed. It did cause considerable ash flooding
through a 150 foot screw conveyor into and out
of the mixer, into the processing building and onto
the discharge conveyors — several hundred tons,
at times. Experience has shown the.importance of
good ash feed equipment to prevent floods. Mass
flow feed systems have now been installed and
consistent accurate feed has been obtained.
Exterior belt conveyors handling damp ma-
terials have also presented challenges. Visualize a
24" belt conveyor on 30° troughing idlers having
several transfer point hoppers with 45° sides, sim-
ple rubber scrapers and discharging onto a stacker.
Approximately 80% of the material will drop off
the belt at the head pulley by gravity. The balance
is carried to the belt scraper which removes ano-
ther 18%, leaving 2% on the belt. For a 150 ton per
hour plant that 2% is equivalent to 74 tons of
material per day which could plug pulleys, idlers,
rollers, and then have to be picked up by shovel
and wheelbarrow. IUCS is utilizing several types
of belt cleaners, sometimes in tandem, to keep
the belts clean. A good belt cleaning system has
to be 100% effective to be good.
The 98% of the material removed from the belt
is discharged into the transfer hopper and, because
the material is wet, it adheres to the 45° steel
sides usually resulting in bridging and spillage.
Transfer hoppers with 80°-90° sides are required
to prevent this problem. At the radial stacker hop-
per, spillage has caused additional problems. Some
of the material drops at the pivot point and unless
cleaned regularly, it will set up or freeze. If the
latter occurs, and the stacker is rotated, it rides up
the hard material and separates from the pivot
point. Result: an inoperable radial stacker system
and, if no backup is available, an inoperative plant.
-------
To prevent such problems, plant design, equip-
ment redundancy, operating philosophy, and
maintenance programs must be planned, coordin-
ated, and evaluated.
For plants which IUCS operates, maintenance
capability is provided on all operating shifts to
minimize backup equipment for certain process
systems. For turnkey plants, experience has shown
that utilities prefer to have a high degree of redun-
dancy and backup equipment, and limit mainten-
ance to day shift operations. Since the stabiliza-
tion systems are often located remote to the im-
mediate power plant area and since problems may
occur simultaneously with turbine, boiler, or scrub-
ber equipment, stabilization systems have charac-
teristically received a low priority of maintenance.
Maintenance requirements vs. backup equip-
ment is a compromise and must be addressed by
plant operations as well as engineering. This can
be accomplished by good planning up front with
due consideration to capital expenditure require-
ments and power plant operation philosophy.
FINAL PRODUCT HANDLING
The achievement of a structurally stable and
environmentally compatible landfill requires a
detailed material handling and placing program,
landfill preparation, and quality control proce-
dures. In many respects, the disposal and place-
ment procedures are as important to the overall
stabilization system as the processing facility itself.
The overall treatment system involves two areas of
consideration: the processing plant and the landfill.
The operating and design parameters of the pro-
cessing facility have been discussed at length. How-
ever, it is important to understand that the me-
thods and techniques used to handle the processed
material are also of key importance to the overall
system design.
Once the processed sludge leaves the facility it
is normally placed in a surge pile. Time in the surge
pile can be for a couple of hours to 4-5 days de-
pending on the physical and chemical character-
istics of the sludge, availability of fly ash, final
product solids, and disposal schedule.
Normally, a dryer consistency final product
will require shorter time in the surge pile prior to
handling. For example, final product solids in the
70-75% range can be moved almost immediately,
whereas product solids range of 50-58% will re-
quire initial conditioning of several days before
movement. Figure 12 gives a range of final product
solids and required conditioning times prior to
movement.
FIGURE 12
Solids
Range
(%)
50-58
58-63
63-70
70-80
Stockpile
Conditioning
(Days)
4+
3-4
1-3
0-1
Stockpile
Capacity
(Days)
8
7-8
5-7
4-5
If the material is moved too soon after produc-
tion, it tends to exhibit thixotropic tendencies
requiring greater handling care. The material is
difficult to place in the landfill as it becomes
"oozy" and tends to stick to and inhibit movement
of dozers and compaction equipment. The material
can be quite slick requiring special material handl-
ing equipment to affect efficient placement.
Temperature can also affect the material cure,
handling and placement operations. During winter
months, with temperatures below 40° F, the chemi-
cal reactions in the material are slowed — similar to
cement chemistry. As a result, greater curing times
may be required for the material in the surge pile
prior to movement and placement in the landfill.
Adequate storage capacity in the surge pile area
must be accomplished during system design to plan
for this requirement.
The processed material can be moved to its
final disposal area by (1) front end loading directly
onto trucks; (2) by direct conveyor to the landfill;
(3) by unit train to landfill area, etc. In all in-
stances grading, placing and compaction is required
for final placement. The optimum mode of trans-
portation is based on site-specific parameters, i.e.,
distance to landfill, landfill access, product charac-
teristics, and Utility preferences. The attached
Figure 13 shows typical processed material being
placed in a landfill.
Material is usually placed in 12 to 24" lifts to
ensure adequate placement density and curing.
The disposal operation should be maintained such
that a minimum surface area of fresh material is
exposed to the elements. The working face should
have a slight grade such that any rainfall will tend
to runoff rather than collect in pockets. Should
rain water pockets occur, especially on fresh ma-
terial, the material stabilization will be adversely
affected or stopped completely creating soft spots
in the landfill. It is very difficult for material
handling equipment to move on or through this
material.
805
-------
In the landfill plan, the material can be built-up
to elevations up to 100' and greater. The landfill
is developed in approximately 25' lifts and bench-
ed at the outer surface to provide haul roads and
prevent erosion. Side slopes can be 2:1 horizontal
to vertical with 50' benches. The finished surfaces
should have at least an 18" layer of top soil and be
revegetated to retard erosion.
At several of I DCS' installations, long range
plans call for material to be built into mountains
in excess of 100' in height thus minimizing land
area requirements. Normally, these landfill areas
are on or adjacent to utility property so that a
minimum of material hauling is required.
The biggest potential environmental impact
could be caused by water runoff from the landfill.
For this reason, exposed surface area of freshly
placed material should be kept to a minimum.
I DCS has used sedimentation ponds to collect the
runoff discharge from the landfill area. Suspended
solids tend to settle out with the supernatant suit-
able for discharge to the terrain or evaporation.
A good landfill plan and operation will require
the use of monitoring wells to sample ground
water. These wells should be installed well in ad-
vance of the beginning of operations to obtain
appropriate background data.
REGULATORY OVERVIEW
Regulatory approval has not presented a pro-
blem to date. Since most disposal sites are on
utility property, the utilities must apply for the
necessary permits. While IUCS stands ready to
support all its clients in the permitting procedure,
most utilities prefer to handle the matter them-
selves. However, IUCS takes positive steps to
smooth the way for permit approval. IUCS at-
tempts to obtain conceptual approval and to pro-
vide an understanding of our process in all states
in which we anticipate business before bid spe-
cifications are issued. The regulatory agency is
therefore familiar with the process before the
permit application is submitted. To date, all permit
applications have been approved including two
submitted by IUCS for landfills that we operate.
The system has been accepted in the following
states: Indiana, Illinois, Kentucky, Ohio, Pennsyl-
vania, Texas, Florida, West Virginia and New
Mexico.
Since no one knows, at this time, what the
RCRA regulations will be for fly ash or scrubber
sludge -- will they be classified as hazardous or
non-hazardous — IUCS has been working with the
Federal Office of Solid Waste to demonstrate the
benefits of a structural landfill constructed of
materials with good engineering properties. The
Figure 13 Typical stabilized material placement
806
-------
outcome of our efforts will not be known until
the requirements for special wastes have been
defined later this year. IUCS believes that it has
been able to demonstrate that a properly designed
and managed structural landfill of stabilized
sludge/ash provides equal or better environmental
protection than procedures permitted in the pro-
posed regulations. If the EPA agrees with our cal-
culations, there would be no increase in cost to
the utilities for disposal even though the sludge or
ash is classified hazardous.
OPERATIONAL DATA
The operation of the four dry stabilization
systems discussed in this paper has produced a
total of 1,627,000 tons of stabilized material. In
1983, when the additional eleven stabilization
facilities are operational, 22,491,060 tons per year
of stabilization capacity will be available to assist
in minimizing environmental pollution.
Design, production and landfill data for the four
ID Conversion .Systems commercially operating
stabilization systems is shown on Figure 14.
FIGURE 14
SPECIFICATION DATA- OPERATIONAL STABILIZATION SYSTEMS
Utility
Duquesne
Light Co.
Duquesne
Light Co.
Plant
Phillips
Elrama
Indianapolis Petersburg
Power and #3
Light Co.
Columbus
& Southern
Ohio
Electric Co.
Conesville
#5
MW Scrubber
400 Magnesium
modified
lime
500 Magnesium
modified
lime
515 Limestone
860 Magnesium
modified
lime
Design
Capacity
Tons/Hr.
70
100
130
170
Annual
Production
100% L. F. Operational
Tons/Hr.
613,000
876,000
1,138,000
1,490,000
Date
11/77
11/75
9/78
1/77
Total
Production
To Date
Tons/Hr.
489,000
833,000
37,000
268,000
Disposal Services
Total disposal services
including permitted site
Total disposal services
including permitted site
Design and technical
consultation. Utility site
adjacent to plant
Design & management.
Utility site adjacent to
plant
807
-------
SUMMARY
SC>2 scrubber sludge and ash disposal has become
an important consideration in the design of power
plants. In the stabilization process by IUCS, fly ash
is blended with dewatered sludge and lime additive
to form a product which is suitable for landfill
disposal. Chemical and physical characteristics of
the waste products must be analyzed and their
affects known to optimize the process design.
The stabilization system design must include an
analysis of the optimum siting location. The availa-
bility of land at the power plant, location of dis-
posal area, and material transportation analysis
must be accomplished.
Dewatering of sludge is important in a dry sta-
bilization system as the material handleability and
chemical reactions will be affected. Sludge de-
watering is dependent on the chemical composition
of the sludge, i.e. sulfate to sulfite ratio, excess
lime concentrations, magnesium concentration,
etc. and the physical characteristics, i.e. particle
size and sheared flocculants.
Material feeding and blending must be con-
trolled to insure optimum final product charac-
teristics. This is especially true when limited fly
ash is available to the process. Lime feed accuracy
is important so that additive costs are minimized.
Blending of the material requires special techni-
ques such that the system can accept a wide range
of wet and dry material. Over mixing will create
thioxotropic final material and under mixing
creates poor dispersion of individual materials
resulting in degradation of structural and environ-
mental final product characteristics. Dust collec-
tion equipment should be used which takes into
consideration blending and conveying of wet and
dry materials. Conveyor and radial stacking re-
quirements must include an analysis of material
quantity, quality, handleability, tendency to flui-
dize, and rate of chemical stabilization reaction
time.
IUCS has had considerable process operation
and maintenance "hands-on" experience based on
four stabilization systems in operation. All of the
systems have been operating since startup with-
out a major outage, however, several problems
have been encountered from time to time. These
problems have included sludge pipe pluggage and
wearout because of grit in the sludge, controlled
fly ash feeding problems, sludge dewatering chal-
lenges, vacuum filter cloth blinding, conveyor belt
cleaning difficulties, material handling challenges
etc. With the state-of-the-art development in this
field, and the many problems that can be en-
countered, IUCS has recognized the need for an
intensive quality control program to ensure smooth
and continuous operations.
The regulatory outlook for power plant wastes
is still in question at this time. IUCS believes,
however, that it has been able to demonstrate
that a properly designed stabilization system with
a properly managed landfill will yield a final pro-
duct and long term solution to power plant waste
disposal which is environmentally sound and
economically feasible.
808
-------
POWER PLANT FLUE GAS DESULFURIZATION USING
ALKALINE FLY ASH FROM WESTERN COALS
by
1 9
Harvey M. Ness,1 Philip Richmond,
Glenn Eurick,3 and Rick Kruger4
ABSTRACT
A characteristic of Western coals is that they contain high levels
of alkali such as calcium, magnesium, and sodium. These alkali species
can be leached from power plant fly ash for use in flue gas desulfurization
(FGD) wet scrubbers in lieu of lime or limestone. At present, there are
nearly 2,600 MW's of generating capacity in the Western United States
that utilize either fly ash or fly ash supplemented with lime or limestone.
An additional 3,500 MW's of Western generating capacity are being planned
or are in various stages of construction which will use the alkaline fly
ash. This report describes the Western alkali ash FGD systems.
1 U.S. DOE, Grand Forks Energy Technology Center, Grand Forks, N. Dak.
2 Square Butte Electric Cooperative, Center, N. Dak.
3 Minnesota Power and Light Company, Duluth, Minn.
4 Northern States Power Company, Becker, Minn.
809
-------
POWER PLANT FLUE GAS DESULFURIZATION USING
ALKALINE FLY ASH FROM WESTERN COALS
INTRODUCTION
Two important characteristics of most Western coals is that they contain
significantly less sulfur than most Interior and Eastern region coals, and
that the coal ash contains high levels of alkali such as calcium, mangesium,
and sodium. The significance of the generally low sulfur content, which
averages about 0.7 pet, is that a sulfur dioxide removal efficiency of only
30 to 40 pet is required to meet the existing Federal emission standard of
0.52 g/MJ (1.2 Ib S02/MM Btu), although higher removals are required to meet
some state and local standards. The lower removal requirements, however,
have allowed novel developments in stack gas cleanup technology.
The alkaline constituents in Western coals tend to be the highest in the
low-rank lignites, and progressively less prevalent in the subbituminous and
bituminous coals. The alkaline oxides tn the ash can vary widely, ranging
from under 10 pet to over 50 pet, with significant variations from mine to
mine in a specific coal region, and even between locations within a single
mine. Characteristic differences in coal ash compositions are illustrated in
Table 1.
Table 1. ASH COMPOSITION RANGES FOR WESTERN
BITUMINOUS AND LIGNITE COALS1
Bituminous,
pet
Lignite, pet
North Dakota
Texas
Silica, Si02
Aluminum oxide, A1203
Ferric oxide, Fe20s
Calcium oxide, CaO
Magnesium oxide, MgO
'Sodium oxide, Na20
Potassium oxide, KpO
Titanium oxide, Ti&2
Phosphorous pentoxide, P205-.
Sulfur trioxide, S03
20 to 60
10 to 35
5 to 35
to 20
to 4
1
0.3
0.2
0.2
to 3
to 4
0.5 to 2.5
0.0 to 3
0.1 to 12
11 to 28
8 to 14
2 to 16
18 to 31
2 to 9
1.4 to 6.5
0.2 to 0.6
0.2 to 0.6
0.0 to 0.6
12 to 27
38 to 66
14 to 23
1.8 to 11.8
3.8 to 20.5
0.9 to 6.1
0.2 to 4.5
0.1 to 1
0.5 to 1
.8
.6
0.0 to 0.2
2.6 to 19
810
-------
potential usefulness of the alkali in Western coal ash for flue gas
desulfurization can be illustrated by comparing the mole ratio of coal ash
calcium oxide to coal sulfur content. For a coal containing 7.5 pet ash,
with 25 pet calcium oxide in the ash, the calcium oxide is chemically equi-
valent to slightly more than 150 pet of a 0.7 pet coal sulfur content. The
alkali in some lignites can have an alkali-to-sulfur mole ratio of several
hundred percent. The amount of fly ash alkali that would be available for
use in a wet scrubber would depend on the coal ash content, method of firing
(pc versus cyclone), and scrubber operating conditions; however, there is
often ample fly ash alkali available to react with sulfur dioxide in a wet
scrubber.
Laboratory studies have shown that the fly ash alkali can be solubilized
into an aqueous media, and that solubilization is primarily a function of pH
and, to a lesser extent, mix time. Figure 1 illustrates the calcium oxide
availability as a function of pH for three North Dakota lignite fly ashes.
The data were generated using batch leach procedures^ and, in general,
indicate an increase in the amount of available calcium oxide as the pH of
the solution decreases. Similar trends are evident for most Western coal
fly ashes. There are, however, differences in leach characteristics for fly
ashes from different mines and power plants. The calcium oxide content of
the three fly ashes shown in Figure 1 are nearly identical; however, the
amount of available calcium oxide at pH 7 varies from about 10 pet to about
40 pet. As the solution pH decreases, however, the differences in solubility
become less significant and tend to approach similar availability values.
The variations 1n alkali solubilities can be attributed to differences in
chemical compositions of the original coals, and also to differences in the
boilers from which the fly ash was derived. Other major fly ash species,
such as magnesium and sodium, can also be leached. Magnesium exhibits a pH
profile similar to calcium, while the amount of available sodium is not
significantly affected by pH. The alkali availability data indicate that
significant amounts of calcium alkali can be leached from fly ash.
FULL-SCALE F6D UNITS USING ASH ALKALI
As noted previously, the relatively low sulfur content combined with
the high levels of soluble alkali in the Western coal fly ash have resulted
in some Innovative engineering for flue gas desulfurization (FGD) wet
scrubber systems on Western power plants. Three extensive pilot plant
programs investigating fly ash alkali wet scrubbing have been conducted,
and have resulted in the construction of full-scale scrubber systems
utilizing fly ash alkali. A fourth pilot plant test program is currently
being conducted. Additionally, the Department of Energy's Grand Forks
Energy Technology Center (GFETC) has been investigating the use of alkaline
fly ash in a pilot plant wet scrubber since 1970.
At present there are nearly 2,600 MW's of generating capacity that
utilize fly ash, or fly ash supplemented with lime or limestone. An addi-
tional 3,500 MW's of generating capacity that will use all or part of the
alkaline fly ash are currently being planned or are in various stages of
construction. The FGD units that are presently in-service vary in design
from venturi-marble bed, to venturi-spray tower, to only a spray tower
811
-------
70
60
50
UJ 40
30
O
20
10
STATION
O-LELANO OLDS
O - MILTON R. YOUNG
- HOOT LAKE
5 6
PH
8
Figure 1. Effect of pH on calcium availability for three Western
coal fly ashes.
812
-------
preceded by an electrostatic precipitator. A summary of the current
utility pilot plant programs, along with planned and operating fly ash alkali
FGD wet scrubbers, are presented in Table 2. A brief description of each
installation is presented below.
NORTHERN STATES POWER COMPANY (NSP)
NSP is currently operating two 700-MW boilers which were put into
service in May 1976 (Unit 1), and May 1977 (Unit 2) at the Sherburne County
Generating Station. A third unit is planned to be in operation in 1984.
The generating units are required to meet a permit requirement of 0.037 g/MJ
(0.087 Ib/MM Btu) for particulate (99 pet removal), and 0.41 g/MJ (0.96
Ib/MM Btu) for sulfur dioxide (about 50 pet removal). The primary fuel is a
subbituminous coal from the Col strip area of Montana and has the following
typical analysis: 25 pet moisture, 9 pet ash, 0.8 pet sulfur, and a HHV of
19,733 J/g (8,500 Btu/lb). A typical fly ash would contain 17 pet calcium
oxide.
A pilot plant test program investigating the utilization of fly ash
alkali was conducted by NSP, Combustion Engineering, and Black and Veatch
Consulting Engineers. A conclusion of the test program was that fly ash
alkali could provide 70 to 80 pet of the alkali requirements, with the
remainder added as limestone.3 Based on pilot plant data, two-stage
scrubber systems were installed on Units 1 and 2.
A schematic of the flue gas cleaning system is shown in Figure 2. The
system was designed to remove particulate in a venturi, and sulfur dioxide
in a marble bed. The L/G at the venturi is about 17, and the L/G at the
marble bed is about 10. Each system consists of 12 scrubber modules; 11
modules are required for full-load boiler operation. Each module consists
of a venturi section, a marble bed, a mist eliminator section, and a
reheater section.
The section containing the marble bed is constructed of carbon steel
coated with flakeglass lining for corrosion protection. The slurry reaction
tank is not coated with flake fiberglass lining, and provides a liquor
holdup time of about 12 minutes. A single pump with Ni-Hard internals
provides slurry to the venturi and the marble bed. As solids accumulate in
the module, part of the slurry is bled to a thickener, and the underflow is
pumped to an ash pond. Cooling tower blowdown water is added to the
thickener overflow and ash pond liquor, and is used for: 1) mist eliminator
wash water; 2) flush water for slurry lines; and 3) makeup water to the
scrubber modules. The scrubber systems are operated in a closed-loop mode;
no liquor is discharged to surface waters.
The flue gas particulate is removed in the venturi section using
slurry from the reaction tank. The fly ash trapped in the water droplets
drains to the slurry mix tank when the calcium oxide leaches and becomes
available for reaction wtth sulfur dioxide. The calcium available in the fly
ash (5.7-6.9 g/m3 [2.5-3 gr/scf], 17 pet CaO) represents approximately 60 to
70 pet of the total alkali put into the scrubber system. Supplemental
calcium is added as limestone at a rate of about 2.7 metric tons/hr (3
tons/hr). The limestone is ground in wet ball mills and delivered to the
813
-------
Table 2. F6D WET SCRUBBERS UTILIZING ALKALINE WESTERN FLY ASH
00
Utility/Station
Northern States Power Co.
Sherburne County
Minnesota Power & Light Co.
Clay Bo swell
Montana Power Co.
Col strip
Square Butte Electric Coop.
Milton R. Young
United Power Association-
Cooperative Power Association
Coal Creek
Capacity Status FGD Design Alkali Source
Unit 1 - 700 MW In-service Venturi -marble Subbituminous
Unit 2 - 700 MW In-service bed fly ash -
Unit 3 - 700 MW Planned ~ 1984 Venturi -spray limestone
tower
Unit 3 - 1 MW Pilot plant
Unit 4 - 500 MW Under construction Venturi-spray Subbituminous
tower fly ash - lime
Unit 1 - 500 MW Planned -- Undetermined
mid-1 980 's
Unit 1 - 360 MW In-service
Unit 2 - 360 MW In-service Venturi-spray Subbituminous
Unit 3 - 700 MW Planned tower fly ash - lime
Unit 4 - 700 MW Planned
Unit 2 - 440 MW In-service Spray tower Lignite fly
ash - lime
Unit 1 - 550 MW Under construction Qnrav i-nwpr Lignite fly
Unit 2 - 550 MW Under construction p y ash - lime
-------
00
I—*
I/I
FLUE GAS
TO STACK
LIMESTONE
FLUE GAS 1'***""*""""
» SLOWDOWN
Figure 2. Flow schematic for Sherburne County FGD system, Northern States Power Co.
-------
slurry reaction tanks as a 4 pet slurry. The limestone represents about 30
to 40 pet of the total calcium added to the system and is used to maintain
the slurry pH at a range of 5.0 to 5.5. The average sulfur dioxide removal
efficiency is about 60 pet (inlet concentration range of 500-650 ppm). The
fly ash calcium reacting with the absorbed sulfur dioxide represents about 50
pet of the total amount available. A summary of the sulfur dioxide removal
efficiencies, fly ash alkali utilizations, and operating conditions is
presented in Table 3.
Table 3. NSP FGD OPERATING CONDITIONS
S02:
Inlet ppm.. 500 - 650
Outlet ppm.. 200 - 300
Removal pet.. 60 (average)
L/G:
Venturi 17
Absorber tower 10
Fly ash utilization pet.. 40 - 60
Slurry pH 5.0 - 5.5
Limestone metric ton/hr.. 2.7
Calcium sulfate scale formation is controlled by circulating 10 pet
suspended solids in the scrubber slurry. The amount of calcium sulfate
crystals present in the slurry is about 2 to 3 pet. The formation of
calcium sulfate is enhanced by forcing about 350 pet stoichiometric air
into the reaction tank, and the sulfur species in the system are essentially
100 pet sulfate. Without forced-air oxidation, pilot plant results indi-
cated that about 40 pet of the sulfur in solution was present as sulfite, and
the remainder as sulfate. However, sulfite would oxidize to sulfate, and
since the slurry contained few calcium sulfate seed crystals to provide
nucleation sites, sulfate scale would form on the marble bed, which would
result in the module being removed from service for cleaning. With forced
air oxidation in the slurry reaction tank, which provides adequate time for
desupersaturation, scale formation on the marble bed is not a problem. Bed
pluggage, however, resulting from plugged spray nozzles or uneven flow
distribution, requires regular maintenance.
Past problems associated with the chemistry of the scrubber system
which have been most difficult to control were the levels of recirculated
solids and slurry pH. Other problems have been excessive wear of spray
pumps, failure of rubber-lined spray piping, pluggage of strainers on the
spray pumps, and deterioration of the marble bed support plates. Detailed
discussions of these problems have been previously published;4,5,6 however,
816
-------
most of the problems have been resolved by cooperative efforts of the
scrubber vendor and the NSP engineering staff. At present, NSP has converted
one marble bed module to a spray tower in an on-going investigation to
reduce maintenance costs.
MINNESOTA POWER AND LIGHT COMPANY (MP&L)
MP&L is currently constructing a 500-MW pc-fired boiler at the Clay
Boswell Station at Cohasset, Minnesota. The boiler, called Clay Boswell
Unit 4, will burn subbituminous coal from the Rosebud seam of the Big Sky
mine at Colstrip, Montana, and has the following typical analysis: 24.5 pet
moisture, 9.4 pet ash, 0.54 pet sulfur, and a HHV of 19,239 J/g (8,287 Btu/lb),
An analysis of the coal ash is presented in Table 4.
Table 4. MP&L COAL ASH ANALYSIS
Pet
Silicon dioxide, Si02 43.3
Aluminum oxide, A1203 17.0
Ferric oxide, Fe203 5.7
Titanium oxide, Ti02 0.8
Phosphorous pentoxide, P205 0.1
Calcium oxide, CaO 12.0
Magnesium oxide, MgO 6.7
Sodium oxide, Na20 0.2
Potassium oxide, K20 0.9
Sulfur trioxide, SOs 13.3
Clay Boswell Unit 4 will be required to meet a particulate emission
standard of 0.04 g/MJ (0.1 Ib/MM Btu), and a sulfur dioxide emission standard
of 0.52 g/MJ (1.2 Ib/MM Btu). A flue gas cleaning system, designed and
constructed by Peabody Process Systems, will provide both particulate and
sulfur dioxide control. A schematic of the FGD system is shown in Figure 3.
The flue gas cleaning system consists of four parallel modules; three
modules are designed to handle flue gas from the boiler at full load, and
the fourth module serves as a spare. Each module consists of one radial
flow venturi for particulate removal, and one spray tower for sulfur dioxide
removal. Reheat for the cleaned gas will be provided by mixing with flue
gas taken upstream of the air preheater and cleaned of particulate by two
electrostatic precipitators. The system is designed to operate using lime;
however, it is planned to use the alkalinity present in the fly ash. If
the fly ash present in the flue gas from Unit 4 is not sufficient to meet
the sulfur dioxide emission limit, fly ash collected from baghouses on Units
1 and 2 will also be used. The unit will have a backup lime system available
for use during periods of high levels of sulfur dioxide.
817
-------
00
I—'
00
MAKE-UP
WATER
LIME SLURRY
TANK
FILTER WASH
TANK
Figure 3. Flow schematic for Clay Boswell FGD pilot plant, Minnesota Power & Light Co.
-------
Control of calcium sulfate scaling will be accomplished primarily by
solids circulation and liquor retention in the slurry mix tank. The level
of suspended solidsxin the scrubber solution will be maintained at about
12 pet to provide seed crystals for precipitation of calcium sulfate.
The recirculated scrubbing liquor will have a 10-minute retention time in
the reaction tank to allow desaturation of calcium sulfate. Pilot plant
data indicate that the state of oxidation of the sulfur species is
essentially 100 pet sulfate. Sludge solids will be disposed of in a pond.
The design and operating data were obtained on a 1-MW pilot plant wet
scrubber using a sidestream of flue gas from Unit 3, which has a boiler
similar in design to Unit 4 and uses the same coal. Tests at the pilot
plant scrubber are continuing; however, preliminary results indicate that
sufficient alkali can be dissolved from the fly ash to meet emission require-
ments without supplemental lime. No unusual chemistry-related operating
problems have been encountered. Some selected results are presented in
Table 5.
Table 5. SELECTED RESULTS OF MP&L PILOT PLANT PROGRAM
Test Date: 4-19-78 5-3-78 5-4-78
S02:
Inlet
Outlet
Removal ...
ppm.
ppm.
• •••••• M v \f •
L/6
Fly ash utilization pet.
Slurry pH
Lime requirement, full-
scale unit...metric ton/hr.
1005
430
57.2
58
0
5.4
2.1
684
219
68
62
I00
3.2
0
823
305
62.9
64
VI00
3
^ ratio represents the sum of the L/G values to the venturi and
the spray^tpwer. The results indicate that nearly 100 pet of the fly ash
calcium can be utilized and that the sulfur dioxide removal efficiency is
sufficient to meet the required emission standard. If similar results are
obtained on the full-scale FGD unit, the lime reagent savings would be
substantial .
MONTANA POWER COMPANY (MPC)
MPC operated a two-year pilot plant test program at the J. E. Corette
Station in Billings, Montana. The pilot plant program was designed to
investigate particulate removal and sulfur dioxide removal using the cap-
tured alkaline fly ash. Based on results of the pilot plant test program,
819
-------
the scrubber vendor, Combustion Equipment Associates (CEA) constructed full-
scale flue gas cleaning systems on two 360-MW units located at Col strip,
Montana. Colstrip Unit 1 was put into commercial operation in May 1976, and
Unit 2 was put in commercial operation in August 1976. The information
presented in this report was obtained from earlier publications'»8,9 and
from Dr. Carl ton Grimm, Montana Power Company.
The boilers burn coal from the Colstrip mine, which has 8.2 pet ash,
0.78 pet sulfur, and a HHV of about 20,082 J/g (8,650 Btu/lb). The composi-
tion of the fly ash is shown in Table 6.
Table 6. FLY ASH COMPOSITION AT THE COLSTRIP
STATION, MONTANA POWER COMPANY*
Pet
Silicon dioxide, Si02
Al umi num oxide, Al 203 ,
Titanium oxide, Ti02
Ferric oxide, Fe203 ,
Calcium oxide, CaO ,
Magnesium oxide, MgO
Sodium oxide, Na20 ,
Potassium oxide, K20
Phosphorous pentoxide, P?®5
41.6
22.4
0.8
5.4
21.9
5.0
0.3
0.1
0.4
* Sulfur trioxide-free basis.
The flue gas cleaning system consists of a downflow venturi located in
an upflow spray tower contactor, as shown in Figure 4. The venturi is
equipped with a variable throat to maintain a constant pressure drop at
variable loads. Each flue gas cleaning system on the two units are com-
prised of three modules, with each module capable of handling 40 pet of full
load. Slurry from the reaction tank is pumped to the venturi (L/G 15) for
particulate control, andJ^o the spray tower (L/G 18) for sulfur dioxide
absorption. The flue gas leaving the spray section passes through a wash
tray which is designed to remove entrained solids which could foul the
mist eliminator. The saturated flue gas is then reheated using steam coils
and discharged to the atmosphere.
The scrubber system is operated in a closed-loop mode. Recycle slurry
from the absorber tower recycle tank is bled to an intermediate pond to
settle the solids, and the clarified liquor returned to the scrubber system.
Part of the pond return liquor is diluted with makeup water and used for
bottom wash of the mist eliminator. The remaining portion of the pond water
is used to wash the bottom of the wash tray. Makeup water is added to the
pond return water to minimize calcium sulfate saturation levels in the mist
eliminator washwater. To date there have not been any significant scaling
problems associated with the demisters.
820
-------
oo
NJ
?
cl,
FLUE (
J
1— -
OEMISTER
1
fV
If)
^T
^-€? v J
\
FLY -ASH
3AS — >
D
__•
_ !W 1 ^
fc A
\/\/
\t3>L
-------
The scrubber system is operated so that maximum fly ash calcium is
utilized while achieving the desired sulfur dioxide control and minimizing
calcium sulfate scaling. A summary of the sulfur dioxide removal effi-
ciencies and utilizations are presented in Table 7.
Table 7- MPC FGD OPERATING CONDITIONS
S02:
Inlet ppm.. 600-750
Outlet ppm.. 50-200
Removal pet.. 80-90
L/G:
Venturi 15
Absorber tower 18
Fly ash utilization pet.. 70-90
Slurry pH 4.5-5.5
Lime requirement..metric ton/hr.. 0.19-0.27
The sulfur dioxide removal efficiencies obtained by the FGD system
range from 80 to 90 pet when the boiler is burning coal with a sulfur
content varying from about 0.6 pet to about 0.9 pet. The contribution of
the fly ash calcium to the removal efficiency varies from 65 to 90 pet,
and the corresponding approximate utilization values based on fly ash
containing 20 pet calcium oxide ranges from 70 to 90 pet. Slaked quicklime
is added to the system as a slurry if it is necessary to augment the fly
ash alkali and maintain the scrubber liquor at the desired pH. At average
coal sulfur conditions, the amount of lime added to the system represents
about 10 pet of the total alkali. The annual lime savings is estimated to
be 10,000 metric tons/year (11,000 tons/year) for each FGD unit.
Since the operating pH range of the slurry is moderately low to leach
calcium from the fly ash, other species will also be leached, and high
levels of magnesium and sulfate have been reported. The major sulfur
species present in the slurry is sulfate, which presents over 90 pet of the
total sulfur.
To date, no significant calcium sulfate scaling problems have occurred.
The scaling is controlled by recirculation of suspended solids (about 12
pet), adequate retention time in the slurry reaction tank, and control of
the slurry pH at a constant value. Other operational problems have been
previously reported.7,8,9
822
-------
SQUARE BUTTE ELECTRIC COOPERATIVE (SBEC)
SBEC is currently operating a 238-MW (Unit 1} and a 450-MW (Unit 2)
cyclone-fired generating unit at the Milton R. Young Station at Center,
North Dakota. The primary fuel is a North Dakota lignite which has the
following typical analysis: 35 pet moisture, 8 pet coal ash, 0.8 pet sulfur,
and a HHV of 15,555 J/g (6,700 Btu/lb). An analysis of a typical fly ash is
presented in Table 8.
Table 8. TYPICAL ANALYSIS OF LIGNITE FLY ASH PRODUCED BY CYCLONE-
FIRED CENTER UNIT 1 AT THE MILTON R. YOUNG STATION
Pet
Loss on ignition at 800° C
Silica, Si02
Aluminum oxide, AlgOs
Ferric oxide, FezOs
Titanium oxide, TiOz
Phosphorous pentoxide, P205
Calcium oxide, CaO
Magnesium oxide, MgO
Sodium oxide, Na£0
Potassium oxide, K20
Sulfur trioxide, SOs
2.2
29.8
12.7
10.6
0.5
0.3
25.7
4.5
2.2
2.0
6.4
The 450-MW unit is required to meet the Federal emission standard of
0.52 g/MJ (1.2 Ib S02/MM Btu). Particulate control for both units is pro-
vided by electrostatic precipitators, and sulfur dioxide control on Unit 2 is
provided by a wet scrubber using fly ash alkali.
A pilot plant test program was conducted under a cooperative agreement
by SBEC, Minnesota Power and Light Company, Sanderson and Porter Consulting
Engineers, the Department of Energy's Grand Forks Energy Technology Center,
and Combustion Equipment Associates-Arthur D. Little (QEA;;ADL). The pilot
plant results have been reported previously.10''1»I2,73,14,15 system design
and operating criteria were generated for construction and operation of a
full-scale spray tower system on Unit 2 which would use fly ash alkali with
supplemental lime.
A schematic of the full-scale scrubber system is shown in Figure 5.
The system consists of two modules, with each module designed to handle 60
pet of the flue gas at full load. Each module contains a reaction tank,
six spray nozzle banks, a wash tray, and demistors. Approximately 15 pet of
the inlet flue gas bypasses each absorber tower and is used for reheat. The
spray tower, reaction tanks, and mix tanks are coated with flakeglass lining
for corrosion control; all pumps are rubber-lined for erosion/corrosion
control. Solids are controlled by bleeding to a thickener. Wfien the
823
-------
REHEAT BY-PASS
BOOSTER
FAN
00
r
S THICKENER r-
<
L
/
J
THICKENER
OVERFLOW
TANK
FILTER CAKE
Figure 5. Flow schematic for Milton R. Young FGD system,
Square Butte Electric Cooperative
-------
thickener underflow contains 35 to 50 pet solids, they are pumped to a
rotary vacuum filter and dewatered to approximately 60 to 80 pet solids.
Dewatered sludge from the vacuum filter is disposed of in the strip mine.
The state of oxidation of the sulfur species is essentially 100 pet sulfate.
The sulfur dioxide removal efficiency for flue gas passing through the
spray towers when burning coal with average sulfur content has been 90 pet
or better, using only fly ash alkali. Some selected results are presented
in Table 9. The removal efficiency data represent only the spray tower
efficiency; removal efficiency for the FGD system would be less since
approximately 15 pet of the flue gas is bypassed and used for reheat.
Table 9. SBEC FGD OPERATING CONDITIONS
Absorber Tower S02:
Inl et ppm.. 450-900
Outlet ppm.. 10-50
Removal pet.. 90-99
L/G 80
Fly ash utilization pet.. 75-100
Slurry pH 4-5
Lime requirement metric ton/hr.. 0
Calcium sulfate scaling in the absorber tower and slurry lines is con-
trolled by circulation of suspended solids, by sufficient liquor retention
time in the recycle slurry tank, and maintaining the slurry at a pH range
of 4 to 5. The recycle tank has a liquor retention time of about 8 minutes
to allow calcium sulfate desupersaturation. The level of suspended solids
in the recycle slurry is maintained at about 12 pet to provide seed crystals
for precipitation of calcium sulfate. Scaling in the fly ash wetting tank
has been a problem and requires cleaning every two to three months. Scaling
has also occurred in the lime slurry feed line. Neither calcium sulfate
scaling nor deposits have been observed on the wash tray or mist eliminator.
The recycle slurry is maintained at a pH range of 4.0 to 5.0 by the
addition of only fly ash alkali. Fly ash from both Unit 1 and Unit 2 is
available for use, and is stored in a single storage silo. Operation of the
FGD system at this pH range minimizes significant problems in the thickener
and leaches 75 to 100 pet of the calcium alkali from the fly ash. Pilot
plant data indicate that significant amounts of other fly ash species such as
sodium, magnesium, and trace elements, can also be solubilized, resulting in
high levels of dissolved solids. The level of dissolved-solIds reached an
equilibrium value of about 10 pet in the pilot plant operation. The level of
dissolved solids accumulated during the limited operation of the full-scale
system is only 2 to 3 pet.
825
-------
The system availability has been poor, averaging only 46.1 pet. The
scrubber system problems have been primarily mechanical. However, one
problem relating to the chemistry of the fly ash alkali which is responsible
for a significant amount of downtime is associated with the thickener tank
and the slurry pH.
The slurry is maintained at a pH range of 4.0 to 5.0 by the addition of
fly ash alkali. If the slurry rises above pH 5, the settled solids in the
bottom of the thickener begin to exhibit pozzolanic activity,- which even-
tually results in a hardened mass requiring shutdown of the thickener (and
the FGD system) and manual removal of the sludge. Similar problems have
occurred in the vacuum filter dump bin. Additionally, the thickener is
operated so that the thickener underflow liquor is maintained in a pH range
of 4.5 to 6.0. If the thickener underflow liquor pH rises above 7.0, the
sludge material becomes slimy, resulting in poor dewatering and caking
properties.
If the pH of the recycle slurry is too low, the suspended solids in
the thickener overflow increases. Data generated during the pilot plant
program indicated that the solids settling properties and filter charac-
teristics degraded as the pH of recycle slurry decreased. Similar obser-
vations have been observed during operation of the full-scale system. SBEC
is currently considering the use of a centrifuge for solids separation.
The poor settling characteristics of the suspended solids can be
attributed to a decrease in the particle size distribution. Figure 6
illustrates the particle size distribution of sludge collected during the
pilot plant program. In general, as the pH of the scrubber slurry decreases,
the particle size of the recirculated suspended solids decreases. Prelimi-
nary results on the sludge particles using scanning electron microscope
techniques (SEM) indicate that calcium sulfate crystals are not as large at
low pH conditions as they are at high pH conditions. The SEM investigation
also indicates that some calcium sulfate is precipitating on the surface of
fly ash particles.
Additional mechanical or electrical problems resulting in signifi-
cant downtime are due to the booster fans, liquid level monitors, flue
gas bypass dampers and chains (freeze-up during cold weather operation),
slurry spray pumps, erosive failure of flakeglass lining in one spray
tower, and the fly ash feed system.
The problems encountered with the slurry spray pumps are caused by an
accumulation of 1.25 cm (1/2-inch) or greater solids in a strainer immediately
upstream of the pump. As the slurry flow became restricted, the rubber
lining would be pulled from the pump casing, resulting in a significant
number of plugged spray nozzles. At the present time, pressure drop across
the strainers is monitored by the operator and manually cleaned when
necessary. The solids, formed in the slurry recycle tank, are caused
primarily by fluctuations in the liquor level. When the liquor level
decreases, slurry containing suspended solids dried on the tank^walls. As
the deposit thickness increases due to continuous fluctuations in the level
of recycle slurry liquor, the deposits eventually flake off and are caught
in the strainers.
826
-------
100
80
60
40
g 20
LJ
M
CO
LJ
O 10
£
RECYCLE SLURRY pH
A 2.9
D 3.8
O 4.9
I 2
10 20 30
50
70 80 90
Figure 6. Particle size distribution of fly ash alkali sludge from
Square Butte Electric Cooperative pilot plant wet scrubber.
827
-------
The fly ash delivery system was designed to add fly ash to a wetting
tank using a conveyor weigh belt. However, the delivery system was not
adequate since the fly ash would flow off the sides of the conveyor belt
(similar to water). At the present time, the belt has been replaced by a
star feeder; however, accurate control of the fly ash feed has not yet been
achieved.
COOPERATIVE POWER ASSOCIATION-UNITED POWER ASSOCIATION (UPA-CPA)
UPA-CPA is currently constructing two 550-MW generating units near
Underwood, North Dakota. The two generating units, called the Coal Creek
Station, will burn lignite from the Falkirk Mine, which has the following
average as-received composition: 39.8 pet moisture, 7-1 pet ash, 0.63 pet
sulfur, and a HHV of 14,530 J/g (6,258 Btu/lb). The two units are expected
to be put in operation in March 1979 and late 1979. The following informa-
tion was obtained from previous publications16'17 and from Mr. Wayne Hickok,
Cooperative Power Association.
The Coal Creek Station is required to meet the current Federal
standards of 0.04 g/MJ (0.1 Ib/MM Btu) for particulates, and 0.52 g/MJ
(1.2 Ib S02/MM Btu) for sulfur dioxide emissions. Particulate control will
be provided by electrostatic precipitators and sulfur dioxide control by wet
scrubbers using fly ash alkali. The FGD system was designed and is being
constructed by Combustion Engineering, Inc.
A schematic of the FGD system is shown in Figure 7. The system will
contain four pressurized spray towers for each unit. The four spray towers
will be capable of operating independently as the boiler load varies. When
lignite containing average sulfur is being used, one or more of the spray
towers will be on standby. The four spray towers are designed to remove 90
pet of the sulfur dioxide from 60 pet of the flue gas at a L/G ratio of 60.
The remaining flue gas (40 pet) will bypass the spray towers and be used
for reheat. The spray towers and outlet ductwork are constructed of stain-
less steel containing a minimum of 2.5 pet molybdenum to resist corrosion
and erosion.
Each FGD system will contain two unlined carbon steel reaction tanks,
with one reaction tank providing slurry to two spray towers. The reaction
tanks will provide a slurry retention time of about ten minutes (two spray
towers in operation). The slurry will be maintained at a suspended solids
level of about 15 pet and at a pH of about 7.0 by addition of fly ash. Fly
ash will be added to a wetting tank using gravimetric feeders; overflow
from the wetting tank will feed to the reaction tank. Supplemental lime
can be added when required. UPA-CPA calculations indicate that capital
costs for equipment, ash silos, and building space for the fly ash system
can be recovered in about four years of operation if the fly ash alkali
reduces the lime requirements by 50 pet.
828
-------
oo
N3
BYPASS
11.0.
FAN
FROM
PRECIPITATOR
TO STACK
OEMISTER
DEMISTER WASH
A /1\
A A A/l\
-SPRAY TOWE8
MAKE UP
ao
REACTION TANK
IN-TANK
STRAINER
FLYASH SILO
FLYASH FEEDER
LIME SILO
MAKE UP
1
[»l ' .
f^ DISPOSAL POND
1
C
r1
0
LIME
FEEDER
LIME
SLAKER
LIME
FEED TANK
Figure 7. Flow schematic for Coal Creek FGD system, United Power Association-
Cooperative Power Association.
-------
Scaling in the FGD system will be controlled by maintaining the
suspended solids concentration at about 15 pet, providing adequate slurry
retention time in the reaction tank (about ten minutes), and maintaining
the pH at a constant value (about pH 7). Makeup water will be provided by
cooling water blowdown. Sludge generated from the scrubber system will be
pumped to an ash pond for disposal.
DISCUSSION
If the alklai present in Western coal fly ash is to be utilized, in
part or completely, as a substitute for lime/limestone, it is necessary to
operate at a moderately low pH. The FGD units currently in operation vary
in operating pH value from a low of 4.0 to a high of about 5.5. The actual
pH operating range for each FGD system varies, depending on site-specific
factors and the supplemental alkali (lime or limestone) added to the system.
One consequence of using fly ash alkali at a moderately low pH value
is that other species in the fly ash will be leached and accumulated in
the scrubber liquor. Analyses of several scrubber liquors are presented in
Table 10.
Table 10. LIQUOR ANALYSES FOR FLY ASH ALKALI FGD SYSTEM*
Calcium
Magnesium
Sodium
Potassium
Aluminum
Iron
Chloride
Fluoride
Sulfite
Sulfate
Dissolved
solids
NSP
520-570
2,700-3,900
275-350
40-65
70-275
1.8-5.9
480-765
2.3-6.0
0
14,500-36,500
2.2-4.6
MP&L
Pilot Plant
630-670
635-1,300
5-15
1.5-24
180-380
25-63
72-127
<1.0
0
12,000-15,700
1.8-2.3
SBEC
Full-scale
430-550
500-1 ,600
400-880
130-500
1.3-42
0.5-2.3
72-370
2.7-7.3
0
13,300-19,000
2-4
Pilot
Plant
530
13,000
13,000
1,010
329
808
200
<1.0
0
65,500
10
Concentration units are mg/1.
The data presented in Table 10 were generated using a series of liquor
samples collected during conditions believed to be representative of the
NSP and MP&L plants. The data presented for the SBEC full-scale scrubber
are not considered to be representative of steady-state operation since
the samples were collected after only about 30 days of continuous operation,
830
-------
and pilot plant data indicate that two to three months are required for
ionic equilibrium. For comparative purposes, an analysis of the pilot
plant liquor is also included in Table 10. The pilot plant liquor was
collected after about six months of operation, and the dissolved solids had
reached a steady-state level of about 10 pet. As can be seen, the levels
of magnesium, sodium, and sulfate are present in appreciable concentra-
tions; similar concentrations are expected in the full-scale unit, but will
be dependent on the degree of closed-loop operation.
The high values of dissolved solids are beneficial to the ash alkali
chemistry since the system can be operated at a lower slurry pH compared to
open-loop operation, and still maintain efficient sulfur dioxide removal.
Recent results at the MP&L plant plant!8 indicate that a 63 pet sulfur
dioxide removal efficiency can be achieved during closed-loop operation
(L/G 62, high dissolved solids) at a slurry pH of about 3.0. Similar tests
conducted during open-loop operation (L/G 62, low dissolved solids)
indicate that a pH in the range of 4.8 to 5.0 would be required to obtain a
similar sulfur dioxide removal.
The sludge generated from the fly ash FGD systems generally contain
significant amounts of sulfate as compared to sulfite. The SBEC FGD unit
and the MP&L pilot plant wet scrubber contain essentially 100 pet sulfate
in the sludge waste. The sludge produced at NSP contains essentially all
sulfate, since air oxidation is employed. Without forced air oxidation,
the scrubber liquid would contain about 25 to 40 pet of the sulfur species
as sulfite, and the physical properties of the sludge would be modified.
Some physical characteristics of the various ash alkali sludges are presented
in Table 11.
Table 11. PHYSICAL CHARACTERISTICS OF ASH ALKALI FGD SLUDGE
SBEC MP&L NSP
Moisture
Permeabil i ty
Unconfined compressive
strength
Secant modulus of
elasticity
Surface area
• • • • LJ w w • •
.cm/ sec. .
. kg/cm2. .
. kg/cm^. .
. . .m2/g. .
21.6
0.1-4 x TO'5
1.27
39.1
19.0
17.1
5 x 10'5
0.81
19.4
7.36
19'7 s
2.8 x 10'5
1.4
22.2
7.05
The compressive strength and modulus of elasticity values were obtained
from the stress-strain profiles shown in Figure 8. The profiles are indi-
cative of the unreacted fly ash alkali in the sludge. In general, the MP&L
sludge had the highest calcium alkali utilization and exhibits the lowest
stress value. The.NSP sludge had the lowest calcium alkali utilization,
831
-------
00
co
3.0
2.5
2.0
CO
CO
I*,*
CO
1.0
1.5
MSP
MRS L
SBEC
1.467
1.2225
0.978
0.7335
0.489
0.2445
e
o>
co"
LJ
tr
to
10 20 30 40 50
UNIT STRAIN, cm/cm IO'3
60 70
80
Figure 8. Stress-strain profiles for three fly ash alkali sludges.
-------
in addition to a reported low limestone utilization, and exhibits the highest
compressive strength. In general, however, all three sludges are structurally
suitable for landfill disposal without additional fixation agents. At
present, the SBEC F6D sludge is being disposed of 1n the striptnine, and
studies on structural stability and potential groundwater contamination are
being conducted by the University of North Dakota under EPA funding.
In conclusion, the utilization of Western alkaline fly ash, either
partially or completely, as a substitute for lime or limestone in a wet
scrubber FGD system has been successful. The present report does not attempt
to detail economic savings. Actual savings will be site-specific, depending
on plant location, scrubber system design, and mode of operation; however,
substantial savings can be realized because of reduced lime or limestone
requirements.
REFERENCES
1. Westerstrom, Leonard. Minerals Yearbook: Metals, Minerals, and
Fuels, Vol. 1, p. 1383, 1975.
2. Thunem, C. B. Laboratory Studies on the Applicability of Western
Fly Ashes to Wet Scrubbing for Sulfur Dioxide Removal. Masters Thesis,
University of North Dakota, May 1975.
3. Noer, J. A., et al. Results of a Prototype Scrubber Program for the
Sherburne County Generating Plant. IEEE-ASME-ASCE Joint Power
Generation Conference, Miami Beach, Florida, Sept. 15-19, 1974.
4. Kruger, R. J. and M. F. Dinville. Northern States Power Company
Sherburne County Plant Limestone Scrubber Experience. Utility
Representative Conference on Wet Scrubbing, Las Vegas, Nevada,
February 23-25, 1977.
5. Dinville, M. F. Experience with Limestone Scrubbing Sherburne
County Generating Plant. Conference on Wet Scrubbing, Chattanooga,
Tenn., October 18-20, 1978.
6. Kruger, R.J. Experience with Limestone Scrubbing Sherburne County
Generating Plant, Northern States Power Company. Presented at the
Environmental Protection Agency Symposium on Flue Gas Desulfurization,
Hollywood, Florida, November 1977.
7. Grimm, C., et al. The Col strip Flue Gas Cleaning System, Chemical
Engineering Progress, February 1978.
8. Grimm, C., et al. Particulate and S02 Removal at the Colstrip
Station of the Montana Power Company. Presented at the National
Meeting, AIChE, Denver, Colorado, 1978.
833
-------
9. Berube, D. T. and C. D. Grimm. Status and Performance of the Montana
Power Company's Flue Gas Desulfurization System. Presented at the
Environmental Protection Agency Symposium of Flue Gas Desulfurization,
Hollywood, Florida, November 1977.
10. Ness, H. M., et al. Status of Flue Gas Desulfurization Using Alkaline
Fly Ash from Western Coals. Presented at the Environmental Protection
Agency Symposium on Flue Gas Desulfurization, New Orleans, Louisiana,
March 1976.
11. Ness. H. M., et al. Pilot Plant Scrubbing of S02 with Fly Ash Alkali
fV'om North Dakota Lignite. Presented at the Lignite Symposium,
Grand Forks, North Dakota, May 18-19, 1977.
12. Ness, H. M., et al. Flue Gas Desulfurization Using Fly Ash Alkali
from Western Coals. EPA-600/7-77-075.
13. Ness, H. M., et al. Physical and Chemical Characteristics of Fly Ash
and Scrubber Sludge from Some Low-Rank Western Coals. Ash Management
Conference, Texas A&M University, College Station, Texas, Sept. 25-27, 1978.
14. Ness, H. M., et al. Mass and Trace Element Balance for a Pilot Plant
Ash Alkali Scrubber System. Manuscript in preparation. GFETC/RI-77/3.
15. Honea, F. I., et al. Summary of Test Results for a 5000-acfm Pilot
Plant Ash Alkali Scrubber System. Manuscript in preparation,
GFETC/RI-77/3.
16. Moen, D. A., et al. Coal Creek Station Air Quality Control System.
29th Annual Conference of the Association of Rural Electric Generating
Cooperatives, Vail, Colorado, June 11-14, 1978.
17. Kettner, D. C. Design of a Spray Tower Scrubber for the Coal Creek
Station. Presented at the Second Pacific Area Chemical Engineering
Conference, Denver, Colorado, August 28-31, 1977.
18. MP&L Pilot Plant Progress Report. Carlton Johnson, Peabody Process
Systems.
834
-------
Environmental Effects of FGD Disposal
A Laboratory/Field Landfill Demonstration
by
N. C. Mohn, A. L. Plumley, A. L. Tyler
Kreisinger Development Laboratory
Combustion Engineering, Inc.
Windsor, Connecticut
and
R. P. Van Ness
Louisville Gas & Electric
Louisville, Kentucky
Presented At
EPA SYMPOSIUM ON FLUE GAS DESULFURIZATION
March 5-8, 1979
Las Vegas, Nevada
835
TIS-6126
-------
ABSTRACT
Results of an EPA-funded demonstration program for the conversion of
fly ash and chemically treated sludge from sulfur removal scrubbers
(AQCS) into a stabilized landfill having structural or recreational
applications are discussed. An earlier paper (TIS-5485) described the
laboratory studies and the preparation of field sites for the disposal
of this waste. This paper covers efforts by Combustion Engineering, Inc.
and Louisville Gas & Electric Co. to monitor the disposal sites over a
two-year period.
Properly prepared landfill from FGD sludge/fly ash mixtures will not
contaminate the surrounding groundwater. Results obtained from analysis
of leachates from the series of landfill impoundments in this study show
that trace elements on the RCRA list of contaminants were found in
concentrations below those established to characterize hazardous or toxic
waste.
836
-------
ENVIRONMENTAL EFFECTS OF FLUE GAS OESULFURIZATION WASTE DISPOSAL
A LABORATORY/FIELD LANDFILL DEMONSTRATION
INTRODUCTION
The most extensive commercial experience in flue gas desulfurization
has been with lime/limestone wet-scrubbers, and it is anticipated that
they will account for the majority of sulfur or S02 removal systems at
electric power stations for the next 10 to 15 years; One of the major
challenges associated with the commercial development of these systems'
is the large amount of by-product sludge that must be disposed of within
the constraints of land- and water-quality regulations. It has been
estimated that within five years, air quality control regulations will
require the desulfurization of flue gas from 60 million kW of electricity
generating capacity per year. If this estimate is realized, over 30
million tons of ash-free by-product sludge (50% solids) will be produced
per year. {*•'
Combustion Engineering has been working on means of disposal and possible
uses of the by-products from flue gas desulfurization (FGD) systems since
1969 with the primary objective of providing environmentally safe methods
for the disposal and/or utilization of both the solid and liquid wastes
from lime/limestone FGD systems. This work has continued with a major
laboratory/field demonstration in cooperation with Louisville Gas &
Electric Co. of the techniques for landfill disposal of by-product sludges.
The work was performed under an EPA contract with the Industrial
Environmental Reserach Laboratory at Research Triangle Park, North
Carolina. This paper gives the results of the demonstration.
HISTORICAL BACKGROUND
During the past ten years, nearly fifty different procedures for direct
disposal and utilization with varying degrees of beneficiation have been
investigated. The technically more promising of these sludge utilization
processes are indicated in Table 1. Previous papers have discussed these
potential uses and special disposal methods.(2,3j Based on our studies,
however, we have concluded that utilization will not be able to eliminate
the problem of disposal any more than the utilization of fly ash (10-15
percent of the annual production) has solved the problem of fly ash disposal
Consequently, most of the waste by-products from flue gas desulfurization
will be disposed of in ponds or used as landfill. The choice of disposal
methods and amount of treatment required will depend on the geographical
requirements, economic considerations and the particular preferences of
the operating company.
Prior laboratory work by Combustion Engineering and others had indicated
the environmental advantages of the disposal of stabilized FGD by-product
sludges over untreated sludges. Haas and Ladd(4) showed that waste solids
from a limestone scrubbing system could be stabilized by dewatering and,
subsequent mixing with clay soil or a Western type fly ash having a high
alkali content. Further studies^5'6) showed that the addition of fly ash
837
-------
Table 1 POTENTIAL USES FOR FGD BY-PRODUCT SLUDGE
1. Cement manufacture
2. Concrete admixture for construction
3. Fill material for land recovery
4. Manufacture of sinter bricks
5. Manufacture of gypsum board
&_ 6. Manufacture of wall panels
•° 7. Manufacture of light weight aggregate
2: 8. Manufacture of mineral aggregate
9. Production of mineral wool
10. Recovery of calcium oxide
11. Recovery .of sulfur
12. Road base in highway construction
13. Soil stabilizer for embankment and water retaining
structures
14. Airport pavement mixture
15. Asphaltic filter and wear-course aggregate
16. Grouting agent in wells
17. Filler for glass
18. Filler for fertilizer
19. Filler for paint
20. Filler for plastic
21. Filler for rubber
22. Fill material in abandoned mines, for fire control
o> 23. Fill material in abandoned mines, for subsidence control
+3 24. Neutralization of acid mine drainage
0 25. Manufacture of autoclaved bricks
26. Manufacture of porous pipes
27. Production of cenospheres for lightened building materials
28. Reclamation of polluted lakes
29. Recovery of aluminum
30. Recovery of magnesium oxide
31. Sand blasting grit
32. Soil amendment
33. Filter aid for sewage sludge dewatering
838
-------
and/or lime to F6D solids resulted in the formation of a number of
mineral compounds of high strength and low permeability.
Since the completion of the earlier work, it has become apparent that
FGD sludges contain varying concentrations of trace elements and dissolved
salts which, in an unstabilized state, could contaminate surface or ground
water. 'Although some soils will absorb many of the trace elements in FGD
sludge, major ions, e.g., calcium, slufate and chloride may not be readily
absorbed. Therefore, the impoundment of sludge must be accompanied by
various stabilization procedures that allow a sufficient margin of safety
to control or prevent seepage. Consequently, leachate analyses were added
to the'unconfined compressive strength tests and permeability studies that
were already a part of the sludge-landfill stabilization program.
LG&E/EPA WASTE DISPOSAL PROGRAM
The laboratory/field landfill demonstration was designed to show the
feasibility of producing structural landfill from mixtures of by-product
sludge and fly ash. (Last year Louisville Gas & Electric Co. and Combustion
Engineering, Inc. reported on the progress of this demonstration. Details
of the program setup and procedures can be found in Ref. 7.) To be considered
acceptable, a landfill must provide a material of sufficient structural
integrity to meet minimum standards of a compressive strength >1 ton/ft^
(0.1 MPa) and permeability <5 x 10'5 cm/s.W In addition, a landfill must
not contaminate groundwater by leachate or surface water by runoff or
erosion. The standards used were the levels established for defining
leachates from hazardous wastes under Resource Conservation and Recovery
Act (RCRA), section 250.13(d).l9>
The program was divided into two phases: laboratory and field demonstration.
Although laboratory tests were run for the previous sludge-landfill
studies,,(4,6) new laboratory tests were necessary to determine the
optimum composition of materials and additives for the field tests,
because the sludge for this demonstration had a somewhat different
composition and the landfill would be deposited in a part of the country
with a different climate from the sludge in the earlier studies. These
tests provided the baseline values for the strength, permeability, and
leachate quality of each mixture evaluated. The field demonstration was
designed to provide similar information on the behavior of stabilized
materials under natural environmental conditions including precipitation,
and freeze/thaw. Included in the field phase was evaluation of the
handling, transportation, and placement of the various sludge mixtures.
The by-products used in this demonstration were obtained from the wet
scrubbing of flue gas from combustion of 3% sulfur West Kentucky coal at
the 65 Mw steam generator (No. 6) at the Paddy's Run Station of LG&E.
The FGD system at Paddy's Run was placed in operation in 1973 and has used
carbide lime, a locally available by-product from an acetylene manufacturer
as the absorbent. Phase I of the waste disposal program was designed to
provide a demonstration of impoundment of mixtures of fly ash and chemically
treated sludge from carbide lime scrubbing.
From the standpoint of general usage, commercial lime is more likely to be
used as the S02 absorbent. Phase II of this program was conducted, there-
fore, using sludge obtained from scrubber operation with a commercial lime
absorbent.
839
-------
The flue gas desulfurization system consists of two scrubber modules
which operate in parallel at full load. Figure 1 shows the overall
arrangement of the scrubbing system during the collection of the by-
product used during this study. Inlet SO? concentrations were about
2000 ppm at a gas flow rate of 180,000 Nm3/hr (175,000 ACFM), with
the boiler at half load. A liquid/gas ratio (L/G) of 7.5 L/Nm3
(28 gal/1000 CFM) was maintained during the test program. For
Phase I, Carbide Lime Operation, S02 removal ranged between 75 and
83 percent. A slurry inlet pH of 8 was controlled over the six week
period required to collect the process sufficient by-product to fill
six impoundments.
During Phase II, Commercial Lime Operation, about 2000 ppm magnesium
was added to allow assessment of its effect on system operation. The
slurry inlet pH of 8 was maintained and S02 removal exceeded 90%. The
sludge by-produce was processed and all ten remaining impoundments
were filled within one month.
Laboratory Tests
The ability to stabilize waste solids from AQC systems is a strong function
of the moisture content of the solids. As the solids content of a sludge
increases, the void ratio decreases producing a material with a higher dry
density. Smaller quantities of stabilizing additives are required to harden
sludges with low void ratios because the particles are in closer proximity
to react with each other and the hardening agent. In addition, drier mixtures
that are close to their optimum moisture content for compaction can be placed
in the landfill at much higher densities than wet sludges.
The laboratory program evaluated a series of stabilized sludges, with different
degrees of dewatering. Process I involved mixing thickener underflow with
fly ash and a fixative to attempt to form a pumpable self-hardening mixture.
Process II mixed a partially dewatered sludge with fly ash and a fixative
to form a stable, compactable landfill. Process III combined a more highly
dewatered sludge with fixative and/or fly ash.
Sludges from scrubbing processes using both carbide and commercial lime
additives were used in the laboratory test. Sixty-two mixtures were prepared
for the laboratory screening. The sludges were mixed with fly ash in ratios
ranging from 0:1 to 1.5:1 parts by weight fly ash to dry scrubber solids.
Varying percentages of fixative (lime, hydrated lime, carbide lime, or
Portland cement) were added to aid the (cementing) reaction,
840
-------
00
INLET
GAS
AAAAAAAA
uiu ,
BED REJECT
BOTTOM TANK
TO STACK
SCRUBBER BOTTOM
UPPER DOWNCOMER
SPRAY WATER
LOWER DOWNCOMER
SPRAY WATER
UNDER FLOW
MAKEUP WATER
I ADDITIVE
FILTER LIQUID
FILTER SOLIDS
BOTTOMLESS
ANNULUS
Figure 1 Flow arrangement for carbide and commercial lime testing
-------
To predict the landfill behavior of stabilized sludges, the following series
of tests were performed:
1. Unconfined compressive strength
2. Permeability
3. Leachate analysis
The results of the laboratory testing were reported in depth in the previous
publication^). Briefly, from the lab tests, a total of ten mixtures were
chosen for field evaluation. The mixtures were chosen to give a comparison
between sludges with different degrees of dewatering and/or fixation
additives. The mixes chosen for the field demonstration are shown in
Table 2 with the results of laboratory strenoth and permeability tests.
Table 3 presents the analyses of leacnates collected from column leaching
tests of three of the mixes.
Table 2 MIXTURES FOR FIELD EVALUATION
60-Day
Compressive
Field Sludge Composition Fly Ash Fixative Permeability Strength
Mix No. Percent Solids/Type Ratio F/S Percentage/Type cm/s Tons/ft^
1 24% Carbide Lime Sludge 1:1 5% Carbide 7.6 x 10"5 too soft
2 42% Carbide Lime Sludge 1:1 5% Carbide 2.9 x 10"6 8.2
4 55% Carbide Lime Sludge 0:1 5% CaO 4.5 x 10"7 9.2
6 55% Carbide Lime Sludge 1:1 3% Carbide 2.1 x 10"7 25.1
7 65% Commercial Lime 1:1 0 7.0 x 10~6 3.0
8 50% Commercial Lime 0.5:1 3% CaO 4.1 x 10"6 7.3
9 50% Commercial Lime 1.5:1 3% CaO 5.7 x 10"7 12.5
10 50% Commercial Lime 1:1 3% P.C. 5 x 10"5 5.4
11 50% Commercial Lime 1:1 3% CaO 2.94 x 10"6 8.5
12 50% Commercial Lime 1:1 3% Ca(OH)2 9.2 x 10"6 7.1
Note: Originally 12 mixes were chosen for field evaluation.
Mix numbers 3 and 5 were omitted due to time and weather
constraints, hence the numbering sequence.
A quantity of the selected mixtures was prepared in the field with a process
train designed for this purpose. The sludge/fly ash/treated mixtures were
impounded in specially prepared sites designed to allow determination of
leachate quality.
842
-------
Table 3 LEACHATE ANALYSES
oo
Pore Volume No,
Conductivity (S/cm2)
pH
TDS (ppm)
Cl'(ppm)
S03= (ppm)
Cd (ppm)
Cu (ppm)
Pb (ppm)
Hg (ppm)
As (ppm)
S04= (ppm)
Ca (ppm)
Mg (ppm)
Se (ppm)
Mix 1
24% Carbide Lime
1:1 Fly Ash Sludge
5% Carbide Lime
1 & 2 5 & 6
3550 2350
11.4 11.1
115 <5
65
<0.01 <0.01
0.02 0.02
0.1 0.1
0.003 0.002
0.03
1580 1470
300 310
0.05 0.05
0.019 0.006
Mix 9
50% Comm. Lime
1.5:1 Fly Ash to Sludge
3% CaO
1 & 2
1800
9.3
1100
15
230
0.01
0.02
0.1
0.001
0.01
440
6.7
0.02
0.008
5 & 6*
Mix 7
65% Comm. Lime
1:1 Fly Ash to Sludge
0
1 & 2 5 & 6
2850 2000
7.8 7.7
1700 1200
10 15
40
0.02 <0.01
0.02 0.02
0.1 0.1
0.018 0.001
0.03 0.03
1580 1870
320 300
0.18 0.16
0.003
*Due to low permeability of samples, pore volumes 5 & 6
were not available after 60 days of collection.
-------
Field Demonstration
A series of ten commercial above ground swimming pools and five larger
impoundments (Figure 2) were used as monitored disposal sites for
the sludge-fly ash mixtures.
Small Scale Impoundments
One type of disposal site consisted of small scale impoundments (Figure 3).
The primary purpose of these test sites was to provide a means of deter-
mining the quality of the leachate and runoff from the test mixtures under
field conditions. The impoundments were lined, above ground and have a
capacity of about 25 cubic yards. A total of ten commercial above-ground
swimming pools have been used as the small scale impoundments. Four of
these were used for sludge mixtures made from carbide lime and six for
mixtures using commercial lime.
The bottom six inches of each pool contained non-reactive graded gravel
where the leachate, which permeated the test mixtures, was collected. The
leachate was drained to a collection tank. The amount of leachate collected
was estimated and, together with the National Weather Service Rainfall Data,
was used to determine the average rate of leachate generation. The leachate
from the impoundments was analyzed for dissolved ions. "The first leachate
samples were collected for analysis one week after filling each impoundment.
Thereafter the leachate was collected and analyzed at two or three month
intervals.
The runoff was collected from the small scale impoundment areas through a
gravel filter held in place by a coarse screen. This procedure insured that
drainage was always provided regardless of the level to which the sludge
consolidated. The runoff was analyzed for dissolved species at the same
intervals as that indicated for the leachate.
Large Scale Impoundments
The small scale impoundments provided a convenient means of determining
maximum leaching rates and leachate quality without any interference from
local surroundings. In the actual field site, the landfill will either
absorb or release moisture to the surrounding soil. The large scale
impoundment areas provided a means of assessing the impact of the disposal
material in terms of its effect on local soil moisture and the quality
(dissolved ions) of the moisture in the soil and of the water in the
aquifer beneath the disposal sites.
-------
FILTER CAKE
AVG. % SOLIDS
F:S
FIXATIVE
MIXTURE % SOLIDS
RANGE
24%
1:1
5%
C.L
36-50
oo
55%
0:1
5%
C.L
56-65
42%
1:1
5%
C.L
49-60
C.L. = CARBIDE LIME
P. C. = PORTLAND CEMENT
55%
1:1
3%
C.L
50%
1.5:1
3%
CaO
56-59
50%
0.5:1
3%
CaO
45-70
50%
1:1
3%
CaO
55-78
PIT 4
MIX No. 12
50%
1:1
3%
Ca
-------
VINYL LINER
SCREEN AND GRAVEL
'FILTER FOR RUN OFF COLLECTION
NON-REACTIVE GRAVEL
PRIMARY
LEACHATE
COLLECTION
RESERVOIR
SECONDARY RUN OFF
LEACHATE COLLECTION
COLLECTION RESERVOIR
RESERVOIR
Figure 3 Small scale impoundment
846
-------
Five large scale impoundment areas were excavated, each with a capacity
of about 50 cubic yards. The disposal sites located in natural soil are of
two styles: 10 foot x 10 foot x 8 foot deep pits and 30 foot x 8 foot x
4 foot deep pits. Two contained carbide sludge mixtures with the remainder
containing mixtures of commercial lime sludge (Figure 4). Soil moisture
was monitored by suction lysimeters located 6, 24 and 72 inches beneath
the bottom of the test site. Particular attention was paid to the water
collected in the lysimeters to determine if the soil is absorbing any dis-
solved ions from the sludge leachate. Leachate and ground water were
analyzed perfodtcally.
A schematic flow diagram of the waste material handling system used to process
the sludge during the field demonstration phase is shown in Figure 5.
The entire thickener underflow was pumped around an 800 ft. circulation
loop. A slip stream taken from the loop was used to fill a 10 ft. dia. x
10 ft. high slurry surge tank. The remaining slurry was then returned to
the vacuum filter which is normally used to dewater the solids prior to
disposal. Three separate processes were used to prepare the mixtures for
disposal:
- For Process I, the sludge was pumped from the surge tank directly to the
mixer into which additive and fly ash was being metered.
- In Process II, the sludge was dewatered in the filter press to produce a
filter cake of the same solids content as the filter cake from the commercial
rotary vacuum filter. When removed from the filter press, the filter cake
fell into a surge bin from which it was metered into the mixer by means of
a 6-inch variable speed screw (VSS) conveyor. Fly ash and additive were
simultaneously metered into the mixer by a 4-inch VSS conveyor.
- Process III was planned to evaluate the use of a filter press operating at
high pressure to dewater the sludge. The filter press provides a means of
obtaining a much drier cake than can be obtained with a vacuum filter. A
drier filter cake requires less fly ash and additive for stabilization.
A 65% solids filter cake was expected, however, the maximum cake solids
produced was 5Q%. The filter cake was stabilized with fly ash. All mixtures
were discharged into trucks, transported seven miles and then placed as
landfill.
Leachate Chemical Analyses
Due to freezing weather shortly after placement of the first series of
lime mixes, very few liquid samples were collected from the small
impoundments until the spring thaw, approximately four months later.
Analyses of total dissolved solids (JDS) and selected trace elements
in samples obtained from the inner ring or primary leachate reaservoir
(See Figure 3) and the overflow or runoff reservoir are depicted in
Figure 6. The small impoundments represent the worst case for leachate
generation, no attenuation by local soil is provided nor is any vegetation
cover provided to minimize runoff. In addition, precipitation is held
on the impoundments.
847
-------
ORIGINAL VERSION
-16 FEET•-
0.5 FEET
10 FEET
6 FEET
_L
L—5 FEET -J
2 FEET
N.
-LYSIMETER
LOCATIONS
\
\
\
\ >»_6FEET
\r_
o o o
X
/
/
f
/
/
f
/
/
*
\/
"R
1 6 FEET
\
\
\
\
N
16 FEET
SIDE VIEW
PLAN VIEW
AS REVISED
7 FT. 6 IN.
LYSIMETER LOCATIONS
PLAN VIEW
r
9FT.-
36FT.-
-9 FT;
j-0.5 FT. i
4 FT.
fl.
-L.
SIDEVIEW Q_
6 FT.
Figure 4 Sludge disposal pits
848
-------
TO VACUUM FILTER
HIGH VOLUME
PUMPS
HIGH PRESSURE VARIABLE SPEED
PUMP SCREW CONVEYOR
u,
VARIABLE SPEED
SCREW CONVEYOR
TRUCKS TO DISPOSAL
Figure 5 Equipment schematic
849
-------
TOTAL DISSOLVED SOLIDS
BARIUM IHCRA LIMIT 10 PPMI
MOO
3000
PPM
1000
0
MOO
3000
PPM
1000
0
5000
3000
pm
1000
0
00
en
o °
0
PPM°
0.
0.
0
0
m.0
0.
0.
0.3
a
«•»•
01
oa
SMALL IMPOUNDMENT \
IMIX 1 (FLY ASH/SLUDGE 1 1.5% CARBIDE LIME FIXATIVE)
36 80% SOLIDS MIXTURE
005
11 n n I I "*"
NS NS NS Ml 1 I NS NS 002
» 60 80 ItO 270 360 460 640 630 720 001
DAYS
SMALL IMPOUNDMENT « 0
MIX • IFLV ASH/SLUDGE o 5.1,3% c»o FIXATIVEI o.os
45 70* SOLIDS MIXTURE
0.04
10.03
PPM
1002
001
NS NS NS NS
30 60 M ISO 270 360 460 540 N30 720 ku (
OAVS 005
SMALL IMPOUNDMENT 7 OQ4
MIX 11 (FLY ASH/SLUOOE1 5 1,3% C.O FIXATIVEI
ss 7n SOLIDS MIXTURE o 03
PPM
002
001
_
SMALL IMPOUNDMENT 1
MIX t n °
0
n II ^
1 fl M M NS . 11 1 1 1 * II . „ -s . !
30 60 90 160 270 360 450 540 630 720
DAYS
SMALL IMPOUNDMENT - 6
MIXS (
Pf
0
10
fl NS NS NS NS
.i
.0
0
.6
M
4
.2
•°
.2
0
.0
M
.0
.4
7
30 60 " 90 " HV-2M J60 ~4SO sV 630 7^0
DAYS 0.0
SMALL IMPOUNDMENT 7 '-2
MIX 11 ^
0.8
0.6
NS NS NS NS NS NS f! NS NS NS PPM
II .... 0.4
NS NS NS NS NS NS IS NS 0 30 60 90 180 270^ 360 450 640 630 720 ^
30 GO 90 1H 270 360 450 640 830 720
DAYS
KEY- L~~l - RUNOFF
U& • PRIMARY LEACHATE
NS • NO SAMPLE IN RESERVOIR
IS - INSUFFICIENT SAMPLE
CADMIUM [HCHA LIMIT 0.1 PPMt
D r 0 026 r
SMALL IMPOUNDMENT • 1
15 MIX 1 0020
0 . 0.015 .
PPM
» _ 0 0.010
_ NS NSNS [I H NS NS
] iH n . . n n II H . B , . n , nnfte
0 30 60 90 180 270 360 450 540 630 720
0
» f 0 -025
SMALL IMPOUNDMENT 6 - „
PPM015
e - fin w"°
JSNS H NS NS NS NS
MB , n . . . * . . . i 000,
KEY f~~J • RUNOFF g
^ • PRIMARY LEACHATE
NS - NO SAMPLE IN RESERVOIR
LEAD (HCHA LIMIT 0.5 PPM| Q ^
SMALL IMPOUNDMENT -1
"""
2
•°
SMALL IMPOUNDMENT 1
MIX 1
PNSNSNS nn(IBNSB II*8
» ' ' JL-H-JL-"-*fer fl •*& cis J*ir—
DAYS
SMALL IMPOUNDMENT 6
MIXB
•JS NS NS NS [| | NS - NS NS NS
] 30 60 90 180 270 360 460 540 630
DAYS
SMALL IMPOUNDMENT 7
MIX 11
MS NS NS NS NS NS NS NS NS
30 60 90 ISO 270 360 4SO 540 360
DAYS
KEY- Q • RUNOFF
E3 -PRIMARY LEACHATE
NS - NO SAMPLE IN RESERVOIR
MERCURY (RCRA LIMIT Q 02 PPM)
SMALL IMPOUNDMENT 1
MIX 1
INS NS NS NS NS NS H NS
_n a n B • B ,
MS NS NS NS | NS 0 30 60 90 180 270 360 450 &40 630 720
n . . «n«fl,B. . n , DAYS
30 60 90 180 270 360 450 540 630 720 0.0015
.jAYS
SMALL IMPOUNDMENT 6 Q M,0
MIX 8
PPM
0.0005
0 30 60 90 ISO 270 360 4SO 540 630 720 "" NS NS NS NS « NS NS NS NS
0 r 0025
SMALL IMPOUNDMENT 7
5 M.X1I °-020
0 . 0.016
PPM
5 . 0010 '
NSNS NSNS NSNS NS NS NS
30 60 90 180 270 360 450 540 630 720 uo
DAYS
SMALL IMPOUNDMENT 7 0.0015
MIX 11
0.0010
n PPM
0 30 60 90 180 270 360 450 540 630 720 NS NS NS NS NS NS If NS NS NS
0 30 60 90 100 270 360 450 540 630 720
KEY Q c RUNOFF DAYS
C3 = PRIMARY LEACHATE
NS -NO SAMPLE IN RESERVOIR KEY-CU = RUNOFF 0.0
ESJ • PRIMARY LEACHATE
SMALL IMPOUNDMENT 6
MIX 8
NS NS NS NS n R NS NS n NS NS NS
30 60 90 180 270 360 450 540 630 720
DAYS
SMALL IMPOUNDMENT 7
MIX 11
NS NS NS NS NS NS NS
n
30 60 90 1 BO 270 360 460 S40 630 720
DAYS
KEY CU - RUNOFF
^ - PRIMARY LEACHATE
NS - NO SAMPLE IN RESERVOIR
Figure 6 Leachate and runoff samples from small impoundments
-------
These samples were collected from sites with different degrees of
stabilization. Mixl(small impoundment No. 1 and large impoundment No. 1)
was placed as a wet slurry, with a solids content of approximately
42%. No compaction was possible and after several days, the mix had
settled leaving a foot of free standing water in the impoundments.
The materials remained soft, with little load bearing capacity,
throughout the test program.
Mix 8.(small impoundment Mo. 6) had an initial solids concentration of
approximately 57% and represented a higher degree of dewatering of the
sludge than Mix 1. This material developed a bearing capacity of
>1.5 TSF, as measured by the University of Lousivilie.
Mix 11 (small impoundment No. 7 and large impoundment Mo. 5) contained twice
as much fly ash as Mix 8 and was placed at an initial solids concentration
of approximately 68%. This mixture had a bearing capacity of »1.5 TSF.
In^all cases, Mixture 1 in impoundment 1, a non-compacted mixture formulated
using non-dewatered sludge, provided the highest concentrations of trace
elements in the leachate. It should be noted, however, that all analytical
results were well below the recommended RCRA (Resource Conservation and
Recovery Act) limits for trace elements (Federal Register December 18, 1978,
section 250.13d).
Note that the leachate and runoff samples from the uncompacted mixture
tend to contain high initial concentrations of contaminents which decrease
gradually with time (the lack of sample at 60 and 90 days resulted from
early freezing as noted earlier). Corresponding samples in the compacted
small impoundments 6 and 7 (mixtures 8 and 11) show a much more gradual
buildup of trace elements, a lower maximum and a tapering off to a minimum
level or no sample after 1-1/2 years.
Sampling of the large impoundment was not affected as severely by weather
as in the small above-ground impoundments. The lysimeters were located
well below the frost line. '
The analysis of a series of leachate samples collected from large impoundments
No. 1, 3 and 5 are shown in Figures 7-10 for several major and trace elements.
Several points should be noted:
1) The quality of leachate improved and the quantity decreased with
increasing depth beneath the disposal sites. This is an indication
thSt the filtering action of the soil may aid in decreasing the
concentration of contaminants reaching the ground water supply.
In fact, no samples were available for collection from the 2-foot
or 6-foot sample depths below, large impoundment No. 5. This mixture
had a permeability of less, than 4 x 10~6 cm/s within 60 days after
placement.
2) The leachate quantity decreased and the quality improved with
increased initial dryness of the mix. Mix No. 1, which was placed
as a "soup-like" slurry, produced the highest concentration of
contaminants measured in the leachates during the program.
3) The concentration of contaminants generally decreased with time
from those impoundments where a sufficient number of samples were
collected to assess the long term trend.
851
-------
24* CARBIDE LHK SLUDGE • 1 PART FLT ASH: 1 PART ORV SOLIDS * SX CARBIDE LIME
MIX Ho. 1
•" SAMPLE DEPTH
I. i
a. . .11
100 200 300 «0 60) tu 700
TIME. DAVS
LARGE IMPOUNDMENT No. 3
«0» COMMERCIAL LIME SLUDGE 1 OS PARTS FIV ASM: 1 PART DRY SOLIDS 13% ClO
MIX Ha. >
••'SAMPLE DEPTH
1L
in a» 3» «» MO an TOO
TIME. DAVS
LARGE IMPOUNDMENT No. 5
50% COMMERCIAL LIME SLUDGE * 1 PART FLY ASM: 1 PABT DRV SOLIDS * 3% C.O
MIX No. 11
6" SAMPLE DEPTH
NSNSNS IS NS IS
TOO 200 300 400 BOO 600
TIME. DAVE
00
l/l
N>
SAMPLE DEPTH
. *. ". I . . .11 .
200 300 400 BOB 800 700
TIME. DAYS
2' SAMPLE DEPTH
WO 2W 300 *» K» «» 700
TIME. DAYS
2' SAMPLE DEPTH
NS NSNSNS NS NS
100 200 3UO 400 MO
TIME. DAVS
2000
1000
V SAMPLE DEPTH
MOTE:
NS-MO SAMPLE
IIH KS NS NS NS
J— J 1—
280 300 400 500 COO 700
TIME. DAYS
SAMPLE DEPTH
NS • NO SAMPLE
300 400 BOO 600 700
TIME, DAVS
6' SAMPLE DEPTH
NS 4SNSNS NS NS
NS = HO SAMPLE
IS = INSU=FICl€HT S»fL£
100 200 300 400 500 600
TIME. DAYS
Figure 7 Total dissolved solids (TDS)--large impoundments
-------
1000
BOO
600
N
400
200
CALCIUM
MIX No. 1
6" SAMPLE DEPTH
100 200 300 400 600 EDO
TIME. DAVS
CALCIUM
LARGE IMPOUNDMENT No. 3
MIX No. 8
6" SAMPLE DEPTH
100 200 300 400 500 COO >00
TIME, DAYS
CALCIUM
LARGE IMPOUNDMENT No. S
MIX No. 11
600 '
400 •
H
200 -
NS VSNSNS
«" SAMPLE DEPTH
100 200 300 400 500 600
TIME. DAYS
SAMPLE DEPTH
00
400
200
BOO
600
ZOO
NS
-I—U_l 1 Lll
200 300 400 600 600
TIME. DAYS
6' SAMPLE DEPTH
NS - NO SAMPLE
NS NS NS NS
100 TOO 300 400 500 600
TIME. DAYS
BOO
600
Ci
PPM 400
r SAMPLE DEPTH
BOO
600
100 200 300 400 500 600 700
TIME. DAYS
6' SAMPLE DEPTH
NS = NO SAMPLE
-n
100 200 300 400 SOO 600 700
TIME, DAYS
2' SAMPLE DEPTH
N! US NS NS NS NS NS
100 200 300 400
TIME. DAYS
6' SAMPLE DEPTH
NS" NO SAMPLE
NSNSNSNS MS NS NS
100 200 300 400 500
TIME, DAYS
Figure 8 Calcium--large impoundments
-------
oo
O.100
0.075
0.050
0.025
NS
a too
0.075
OJBO
0.02S
NS
aioa
O.O75
agso
0,025
NS
LARGE IMPOUNDMENT No. 1 ^
MIX No. 1
8" SAMPLE DEPTH °-100
O.O75
**•* 0.060
tn ao»
fl 0 ^
n, II. II .
100 200 300 400 500 600 700
TIME. DAYS
0.100
2' SAMPLE DEPTH *'075
W>M 0.050
4X175
NS
lit " "
100 200 300 400 500 600 700
TIME. DAYS ,
0 100
6* SAMPLE DEPTH
0.075
!S.
HS'NOSAMPLE PM OQ50
„
I) 0.025
||MS NS NS NS NS ^
11 . . . , . t
100 ZOO 300 4OO BOQ 600 700
TIME. DAYS
MIX No. 8
0.100
S" SAMPLE DEPTH
0.075
S*
"** 0.050
HS NS 0025
»t? ....,,. NS
100 200 300 400 600 600 700
TIME. DAYS
r SAMPLE DEPTH
0.100
.
0.075
S«
PPM
NS NS NS °a»
100 2OO 300 400 &OQ 600 700 0.025
TIME. DAYS NS
r SAMPLE DEPTH 0.100
0,075
NS • NO SAMPLE S»
PPM
0.050
US 0.025
~ *" ~*~ ~" ^^ (^ ^^
TIME. DAYS
MIX No. 11
6" SAMPLE DEPTH
NSNSNS -,.001 NS
100 200 300 400 500 600
TIME. DAYS
2' SAMPLE DEPTH
NS NS NS NS NS
100 200 300 400 500 600
TIME. DAYS
6' SAMPLE DEPTH
NS - NO SAMPLE
NS NS NS NS Ni.
100 200 300 400 500 600
Figure 9 Selenium--large impoundments
-------
CHLORIDE
LARGE IMPOUNDMENT No 1
MIX No. 1
CHLORIDE
LARGE IMPOUNDMENT No. 3
MIX No. 8
CHLORIDE
LARGE IMPOUNDMENT No. 5
MIX No. 1 1
•"SAMPLE DEPTH
100 200 300 400 500 600
TIME. DAYS
«- SAMPLE DEPTH
300 300 400 SOD
TIME. DAYS
Cl 400
PPM
6" SAMPLE DEPTH
00
tn
ui
2' SAMPLE DEPTH
MS UllNS NS HS
7 SAMPLE DEPTH
2* SAMPLE DEPTH
NSNSMS NS NS MS
SAMPLE DEPTH
NS-NO SAMPLE
NS NS NS NS
100 200 300 400 500 GOO 700
TIME. DAYS
r SAMPLE DEPTH
NS - NO SAMPLE
100 200 300 4OO
500 600 700
TIME, DAYS
6'SAMPLE DEPTH
NSNSNS NS NS NS
100 200 300 400 500 600
TIME. DAYS
Figure 10 Chloride—large impoundments
-------
4) With only three exceptions, the concentrations of trace
elements in1 all of the leachate samples collected during
the program were below the levels established for defining
leachates from hazardous wastes under RCRA 250.13(d). The
only leachates exceeding these limits were from Mix No. i
(Selenium and Cadmium) and Mix No. 9 (Cadmium).
Ground water background levels of many priority pollutants approached
RCRA limits prior to the start of this demonstration program. Samples
obtained from existing wells on the site during the test program have
shown no detectable increase in any species currently being monitored.
It should also be noted that rain water samples collected during this
test program approached primary drinking water standards for several
substances.
SUMMARY
Properly prepared landfill from FGD sludge/fly ash mixtures will not
contaminate the surrounding groundwater. Results obtained from analysis
of leachates from the series of landfill impoundments in this study show
that trace elements on the RCRA list of contaminants were found in
concentrations below those established to characterize hazardous or toxic
waste.
A trend toward decreasing concentrations, with time, of trace contaminants
was observed in both leachate and runoff samples obtained from the
stabilized sludge mixtures. The small impoundments have provided higher
concentrations since no attenuation by local soil is provided and
vegetation that might minimize runoff was not established on these sites.
Most sites developed compressive strengths significantly (up to tenfold)
greater than the minimum required for recreational or light structural
landfill. Water samples obtained from 4?eneath the large impoundments
indicate that the filtering action of the soil aids in decreasing the
concentration of contaminants reaching the ground water supply. Certain
mixtures have undergone a, fixation reaction thus minimizing the release,
of moisture and/or contaminants to the surrounding soil.
856
-------
REFERENCES
1. Letter, Julian Jones of EPA to A. L. Plumley, Jan. 8, 1979.
1 •
2. Taylor, W. C., "Experience in the Disposal and Utilization of
Sludge from Lime/Limestone Scrubbing Processes," paper presented
at the Flue Gas Desulfurization Symposium, New Orleans, Louisiana,
May 14-17, 1973.
3. Taylor, W. C. and Haas, J. C., "Potential Uses of the By-Product
from the Lime/Limestone Scrubbing of S02 from Flue Gases," paper
presented at the American Institute of Mining, Metallurgical, and
Petroleum Engineers 1974 Annual Meeting, Dallas, Texas, Feb. 23-28,
1974; Combustion Engineering publication TIS-3774A.
4. Haas, J. C. and Ladd, K., "Environmentally Acceptable Landfill
from Air Quality Control Systems Sludge," paper presented at
Frontiers of Power Technology Conference, Oklahoma State
University, Stillwater, Oklahoma, Oct. 1974; Combustion
Engineering publication TIS-4216.
5. Klym, T. W. and Dodd, D. J., "Landfill Disposal of Scrubber Sludge,"
ASCE Annual and National Environmental Engineering Convention,
Kansas City, Missouri, Oct. 1974.
6. Haas, J. C. and Lombardi, W. J., "Landfill Disposal of Flue Gas
Desulfurization Sludge," paper presented at NCA/BCR Coal Conference
and Expo III, Louisville, Kentucky, Oct. 19-21, 1976; Combustion
Engineering publication TIS-4926.
7. Van Ness, R. P., Plumley, A. L., Mohn, N. C. and Stengel, M. P.,
"Field Studies in Disposal of Air Quality Control System Wastes,"
paper presented at the Third Annual Conference on Treatment and
Disposal of Industrial Wastewaters and Residues, Houston, Texas,
Apr. 1978; Combustion Engineering publication TIS-5485.
8. Prior to this field demonstration, it was agreed among Louisville
Gas & Electric Co., the Environmental Protection Agency, Combustion
Engineering, Inc., and Aerospace Industries (a major contractor to
EPA on EPA's landfill program) on the compressive strength and
permeability standards for landfill.
9. Resource Conservation and Recovery Act of 1976, Publication L94-580,
Oct. 21, 1976; Federal Register Dec. 18, 1978, section 250.13(d).
857
-------
PHYSICAL PROPERTIES OF FGC WASTE
DEPOSITS AT THE CANE RUN PLANT
OF LOUISVILLE GAS AND ELECTRIC COMPANY
C. R. Ullrich and D. J. Hagerty
Civil Engineering Department
University of Louisville
Louisville, Kentucky 40208
and
R. P. Van Ness
Louisville Gas & Electric Company
Louisville, Kentucky 40201
Abstract
Physical tests performed on FGC wastes from the Cane Run plant of LG&E
included in situ shear strength determinations and in situ plate loading tests,
as well as shear strength and permeability tests performed in the laboratory on
samples taken from waste deposits at the plant. In situ testing and sampling
were done at nominal intervals of 0 days, 30 days, 90 days, 180 days, and 360
days after field placement of wastes. Various mixtures of lime process FGD
sludge, flyash and lime (or cement) were tested. EPA-sponsored testing of field
deposits disclosed non-homogeneous and discontinuous conditions created by
weather effects ^freezing and thawing), interruptions in filling and incomplete
mixing.
Sample disturbance of the brittle cured wastes was severe; thus, lab
strength values were low, about one-half the value of corresponding in situ
test strengths. Lab permeability tests on field samples yielded values as much
as two orders of magnitude higher than values obtained on lab-cured undisturbed
samples. Permeability of field samples varied between 3 x 10~5 cm/sec and
3 x 10~6 cm/sec. In situ shear strength values 30 days after placement varied
from about 250 psf to more than 3,000 psf. Strength increases were noted with
age for all but one deposit (which was disrupted by freezing). In situ strengths
were much lower than strengths of samples cured and tested in the~laboratory.
Plate loading tests on selected deposits of wastes showed stiff fixed materials
with bearing capacities greater than 15,000 psf.
Freezing created layers in waste deposits and retarded or impaired stabili-
zation reactions between lime and wastes. Some dewatering of exposed surfaces
(to depths of six inches or so) was caused by freezing and thawing.
Commercial lime was more effective than carbide lime (a waste product avail-
able near the Cane Run plant) in stabilizing wastes and performed as well as
Portland cement. oco
ojo
-------
PHYSICAL PROPERTIES OF FGC WASTE DEPOSITS AT THE
CANE RUN PLANT OF LOUISVILLE GAS AND ELECTRIC COMPANY
OVERVIEW
To characterize various mixtures of FGC wastes and additives and to in-
vestigate their behavior during land disposal, a series of impoundments was
constructed at Louisville Gas & Electric Company's Cane Run Plant. After field
placement, in situ strength tests were conducted, and samples were taken and
were subjected to laboratory permeability tests and unconfined compression tests.
Additionally, plate load tests were performed on materials in two test pits.
These physical tests indicated that behavior of mixes of FGC wastes were
highly dependent on mix composition. Total solids content fis important; in-
creasing solids content from 50% to 67% may increase strength 10-fold while
giving a 10-fold decrease in permeability. However, strength gain is highly
dependent on the fixative material added to the mix: in these tests, Portland
cement was far more effective than commercial lime. Variation of flyash-to-
sludge ratio changed the mix behavior but not as drastically as did changes in
total solids content and type of additive. High flyash content in a mix changed
the material from a somewhat plastic material to a more pervious, friable and
brittle material (for these mixes with calcium sulfite sludge from a lime scrubber).
Other factors of great influence were quality of mixing, continuity of place-
ment, weather conditions at placement and weathering effects (freeze-thaw
especially) after placement. Poor mixing caused heterogeneity in waste deposits.
Interruptions in placement caused the formation of distinct layers (wetter,
softer zones between denser, harder layers). Freezing accentuated layering,
with consequent changes in solids content, strength and permeability. Hot
weather led to rapid reactions and early strength gain; in some cases, some
of this strength disappeared during a complete cycle of weathering.
Laboratory tests prior to field placement, conducted on very well-mixed,
carefully cured samples, did not forecast accurately the behavior of not so
well-mixed materials placed in layers in all extremes of weather. Moreover,
extreme care was required to obtain samples from field deposits without
seriously disturbing the brittle but very stiff wastes mixes.
These tests showed that mixes of FGC wastes with 3-5% lime or cement may
exhibit strengths as .high as 5,000 psf (undrained shear) with permeability as
low as 10 cm/sec, if total solids content is high and placement conditions
are favorable. Details of these findings are given in the following pages.
859
-------
INTRODUCTION
This paper presents results of field and laboratory testing done on various
mixtures of scrubber sludge, flyash, and lime. The sludge mixtures were contained
in impoundments at Louisville Gas and Electric Company's Cane Run Plant, and
testing was done by the University of Louisville from November 1976 to October
1978, under contract to LG&E.
Sludge mixtures impounded at the Cane Run disposal site were contained in
two basic styles of impoundments: 25-^cuyd capacity backyard swimming pools and
50-cuyd excavated pits. Ten swimming pools were erected: four of these were
used for sludge mixtures utilizing carbide lime and six for mixtures utilizing
commercial lime. Excavated pits were of two styles: 10-foot x 10-foot x 8-foot
deep pits and 20- foot x 10-foot x 4-foot-deep pits. Two deep pits were used
for sludge mixtures utilizing carbide lime and the other three pits were used
for commercial lime-sludge mixtures. Details of the impoundment layout, mix
designs, and placement procedures are given in a companion paper by Mohn, e_t
DESCRIPTION OF WORK PERFORMED
As requested by Louisville Gas & Electric Company, work done by the University
consisted of three basic tasks:
1) The local soil /hydrology at the disposal site was studied for purposes
of locating tests impoundments.
2) In-situ vane shear strength tests were to be performed at nominal
intervals of 0 days, 30 days, 90 days, 180 days, and 365 days after placement.
3) Laboratory unconfined compression tests and permeability tests were
to be performed on undisturbed samples taken at the same time that the in situ
vane shear strength tests were performed.
Sampling and testing was done as closely as possible to scheduled times,
although inclement weather and other circumstances sometimes interf erred with
strict adherence to the schedule. Laboratory testing was done usually within
one week of sample procurement if at all possible. Because of the expiration
of the first phase subcontract, it was not possible to do field and laboratory
testing on sludge mixtures containing carbide lime 365 days after placement.
Also, testing at 90 days, 180 days, and 365" days after placement was not done
on sludge mixtures containing commercial lime.
To satisfy the deficiencies in the initial work program in 1978 University
of Louisville investigators conducted a program of additional physical testing.
860
-------
The additional physical tests included field vane shear strength measurements
and sampling of sludge mixtures contained in Pits 1, 2, 3, 4, and 5 and Pools
2, 5, 9, and 10. Laboratory unconfined compression tests and permeability tests
were performed on obtained samples whenever possible. Pools 1, 3, 6, 7 and 8
contained mixes identical to those in Pits 1, 2, 3, 5 and 4, respectively, and
were not re-tested. The mix in Pool 4 had emerged from the mixing truck as
spheres at placement, had soon frozen and disintegrated upon thawing; conse-
quently, tests on Pool 4 were not relevant.
Vane shear strength tests were made at one-half foot intervals in all
impoundments where it was possible to insert the field vane shear strength
measuring device. Continuous sampling was done on the sludge mixtures contained
in the selected impoundments. Samples were obtained by use of 30-inch long by
2-inch diameter Shelby tubes. A drilling rig was utilized in obtaining samples
in the pools and pits where the stiffness of the mixture precluded pushing the
sampling tubes into the material manually. From each 2-foot length of sample,
two unconfined compression test samples and one laboratory permeability sample
were prepared. The laboratory unconfined compression test samples were prepared
from 8-inch lengths of sample from the ends of each Shelby tube sample. The
unconfined compression tests were performed in accordance with ASTM Standard
Method of Test D2166. Laboratory permeability tests were performed on five-
to six-inch lengths of sample cut from the middle eight inches of each Shelby
tube sample. Testing was done on the sawed length of Shelby tube to avoid
sample disturbance.
TEST RESULTS FOR EACH IMPOUNDMENT
Pit No. 1;
Mix 1 was placed in Pit No. 1 and consisted nominally of sludge (thickener
underflow at 24% solids content), with an equal dry weight of flyash, and 5%
by dry weight of carbide lime added, to give a nominal final solids content of
39.3%.; The pit was filled in truckloads with overnight interruptions in the
filling operation, on November 8-11, 1976. After each sequence of filling, the
solids mixture settled and supernatant water appeared on top of the deposit.
After final fill ings the depth of supernatant water reached approximately 10-
12 inches before evaporation removed all standing water. After this water
evaporated during November and early December, the exposed top of the deposit
froze in late December. The frozen crust of the deposit reached thicknesses
of at least 12-18 inches during the winter of 1977 and possibly greater thick-
ness during the severe winter of 1978.
861
-------
Until the most recent sampling effort in September 1978, only one tube
sample had been obtained from Pit No. 1, on February 8, 1977 when the sludge
had frozen to a depth of several feet. Prior to that time, the extremely soft
consistency of the mixture in the pit precluded sampling. On September 28, 1978
the investigators obtained continuous samples from Pit No. 1 for the first time,
and a virtually complete record of field vane shear strength with depth also
was obtained. During summer 1977, a desiccated crust was found in the upper
few inches of material in the pit, above material of varying consistency. When
the pit was sampled in September 1978 a crust again was found in the top 12 to
16 inches of the deposit. A stiff'layer was found at 22 inches below the top
of the deposit, but this stiff layer was underlain by softer materials.
The results of the 1978 vane shear tests in Pit No. 1 are shown in
Figure 1, together with results of field vane shear tests performed during 1976
and 1977- Although the nominal solids content of the material in Pit No. 1
was approximately 39% at the beginning of the testing period, evaporation of
supernatant water increased the solids content to between 45 and 55%. At the
time samples were obtained in September 1978, the average solids content of
the materials at depth in the pit was about 58%; in the top 8-12 inches of
material, a dried-out crust hajd formed with a solids content of approximately 75%.
The tube sample which had been obtained from Pit No. 1 in February 1977
melted when it was returned to the laboratory. The samples secured during the
latest sampling operation showed unconfined compression strengths between 250
and 830 psf, with corresponding undrained shear strengths between 125 and 415
psf. The sample obtained in the dry dessicated crust crumbled when it was re-
nioved from the sampling tube. The strain at failure in these unconfined com-
pression tests varied from 6.5% to 8% indicating that the material behaved in
a more plastic fashion than did much of the material in other pits. Laboratory
tests on this mix had shown 60-day unconfined compression strengths inadequate
for sample integrity (too soft to test).1
Permeability tests were performed on two samples obtained from Pit No. 1.
The sample obtained from a depth of 6-12 inches in the deposit exhibited a
-5
permeability coefficient of 2.3 x 10 cm/sec at a solids content of approxi-
mately 62%, while the sample obtained between 27 inches and 33 inches below .
the surface of the deposit showed a coefficient of permeability of 2.9 x 10
cm/sec at a solids content of approximately 56%. Permeability tests on this mix
conducted prior to field placement had shown a coefficient of permeability of
7.6 x 10-5 cm/sec.1
862
-------
~ 4
•)j
O.
8
8
0 Days, Negligible strength
0
500
Pit No. 1 (Mix 1)
A 29 Days
a 89 Days
V 197 Days
• 686 Days
I
1000 1500 2000
Shear Strength (psf)
2500
3000
Figure 1 In-Situ Shear Strength Values for Pit No. 1
863
-------
Pit No. 2
Pit No. 2 was filled, in intermittent fashion by truckloads, from
December 2 to December 6, 1976, with Mix 2, which consisted of sludge (42%
solids content, by weight), an equal dry weight of flyash, and carbide lime
(5% of the dryweight of the sludge) and yielded a mix with a solids content
nominally of 59.75%. On December 8, 1976, the resistance to insertion of the
vane shear device varied considerably with depth, indicating the presence of
layers with a stiff consistency interbedded with layers of soft consistency, in
the upper 3 to 4 feet of the pit deposit. The undrained shear strength values
obtained during the first investigation in December 1976 varied from about 300 psf
near the top of the sludge deposit to about 700 psf near the bottom.
By mid-January, the sludge mixture had frozen downward to about 30 inches
below the top of the deposit. The values of undrained vane shear strength varied
erratically in the upper 4 feet of the deposit from 750 psf to more than 3,000
psf. A very high value was obtained in the wholly frozen upper zone and another
high value (greater than 3,000 psf) was obtained at a depth of 3 feet. Below a
depth of 4 feet, however, the strength varied much less dramatically and averaged
•about 7000 psf. When in situ tests were done in March 1977, the effects of
freezing again were apparent to a depth of about 4 feet although some surficial
thawing had occurred. Below a depth of 4 feet, the vane shear strength values
varied from about 500 psf to about 1,000 psf. Again, a stiff layer was indi-
cated at a depth of about 3 feet. In late June 1977, rn situ strength tests
again were performed. Below a depth of 4 feet the shear strength varied from
600 psf to 1,200 psf, slightly greater than the in situ values measured 100
days after placement. In the upper 4 feet of the deposit, strength values in
excess of 3,000 psf were measured at depths of 2 feet and 3.5 feet, with much
lower values measured between these depths.
When the site was revisited on September 28, 1978, the jji situ shear
strength in the upper 4 feet of the deposit again varied erratically from a
value of about 1,900 psf near the surface to about 800 psf at a. depth of 18
inches. Hard layers were found at 2 feet and 3 feet, with a strength value in
excess of 3,000 psf for the hard layer at a depth of 3 feet. Below a depth of
4 feet, the in situ strength values varied in a pattern similar to the varia-
tion noted during previous sampling. In general, the strength values ranged
from about 550 psf to a high of about 1,200 psf, in a rather regular fashion.
The strength values between 4 feet and 5 feet depth were somewhat lower than
the strength values measured in 1977, but below 5 feet the strength values
864
-------
measured in 1978 generally were somewhat higher than the strength values measured
previously. The vane shear test results for Pit No. 2 are shown in Figure 2.
In comparison, laboratory tests on Mix 2, conducted prior to field placement,
had shown an undrained shear strength of about 4,100 psf.1
When the investigators visited Pit No. 2 on December 8, 1976, the sludge
mixture was too soft to be retained in the tube sampler. In mid-January 1977,
samples obtained were completely frozen and required thawing in the laboratory
before testing. The strength values obtained in unconfined compression tests
on the thawed samples varied from about 250 psf to about 500 psf, and the
moisture contents of the thawed samples were considerably lower (38-64%) than
the moisture content nominally set for the mixture (67%). On March 15, 1977,
when samples were obtained, 99 days after placement, the sludge mixture appeared
to be wholly or partially frozen to a depth of about 4 feet. Samples obtained
at this time were allowed to thaw slowly in the lab and were tested in June,
177 days after placement. Strength values from these tests varied from 400
psf to 750 psf and moisture contents for these samples varied from about 36%
to 49%.
Samples again obtained from Pit No. 2 on June 28, 1977, showed undrained
shear strengths between 500 psf and 2,100 psf, with a high value found for a
sample taken from a depth of 2 feet. This testing date, July 2, was 208 days
after placement. Moisture content values varied from 40% to 59%.
On September 28, 1978, samples again were obtained. When these samples
were tested in the laboratory 672 days after placement, the undrained shear
strength values varied from 184 psf to 1085 psf. Solids contents ranged from a
high of 79% to a low of 65%. These last shear strength test results indicated
some deterioration of the near-surface layers of the mixture but at depth the
strength appears to have increased slightly with time compared to initial values
obtained. The very hard layer found at a depth of 2% feet and represented by
a very strong sample tested 208 days after placement, was not represented in
the laboratory shear strength values found in 1978. The values of laboratory
shear strength determined for samples from Pit No. 2 are shown in Figure 3.
Whenever samples were obtained from Pit No. 2, laboratory permeability
tests were performed on the obtained samples. The first group of samples ob-
tained from Pit No. 2 were tested from 87-113 days after placement of the
material. In general, these samples showed permeability coefficients between
4 x 10"6 cm/sec and 2 x 10"5 cm/sec. Samples tested between 193 and 195 days
after placement showed permeability values between 4.5 x 10"6 and 6.5 x 10
cm/sec except for one disturbed sample taken from a very shallow depth which
865
-------
Pit No. 2 (Mix 2)
2 Days
38 Days
o 99 Days
y205 Days
661 Days
500
1000 1500 2000
Shear Strength (psf)
Figure 2 In-Situ Shear Strength Values for Pit No. 2
2500
3000
866
-------
+J
Q.
d)
O
Pit No. 2 (Mix 2)
A 51 Days
g 177 Days
V 208 Days
• 672 Days
8
I
500
1000 1500 2000
Shear Strength (psf)
Figure 3 Laboratory Shear Strength Values for Pit No. 2
2500
3000
867
-------
.c
showed a permeability value in excess of 2 x W~ cm/sec. Samples obtained later
during 1977 and tested between 270 and 284 days after placement were considerably
more disturbed than previous samples; these samples showed low moisture contents
and relatively high permeability values of 2-3 x 10~5 cm/sec. Four samples
taken in fall 1978 showed moisture contents between 39% and 58%, with correspond-
ing solids contents between 63% and 72% and permeability values generally less
than 6 x 10 cm/sec except for a sample taken within the top 12 inches of the
deposit which showed a permeability of 4.5 x 10-5 cm/sec. In general, except
for the upper 18 inches of the deposit where disturbance of the material has
occurred through weathering, the permeability of the material in Pit No. 2
appears to be approximately 6 x 10~6 cm/sec or less. Laboratory tests on Mix 2
performed prior to field studies had shown a coefficient of permeability of
2.9 x 10-6 cm/sec.1
Pool No. 2
This pool was filled between November 22 and November 24, 1976 with Mix 4,
consisting of thickened sludge (nominally 55% solids content) and carbide lime
(5% by dry weight). The final solids content of Mix 4 was 56.2%, nominally.
Several weeks after placement the mixture was compacted by blows and pressure
from the bucket of a rubber-tired tractor excavator. Prior to compaction, the
mixture had a very irregular surface in the pool. Before compaction the sludge
was sampled and tested in situ on September 1, 1976, seven days after placement.
Samples obtained on that date were tested in the laboratory on January 5, 1977,
42 days after placement. The sludge had partially frozen during early December.
The compacted, partially frozen sludge was sampled and tested again on
December 24, 1976, 30 days after placement. Samples obtained at this time
were tested on January 14, 1977 after they had been allowed to slowly thaw in
the laboratory. At seven days after placement, the in situ strength was 500
to 700 psf while the thawed samples taken seven days after placement showed
strengths of only 200-250 psf. The difference between the values apparently
was due to freezing which gave a false high strength reading . Thirty-day in_
situ tests on partially frozen sludge gave higher results than the values
obtained on thawed samples tested 51 days after placement (750 psf in situ
vs 250 psf on lab samples). The entire thickness of sludge in the pool appeared
to be frozen by the end of February 1977. At that time the frozen sludge had
an apparent strength of from 1,200 psf to 2,100 psf, while the badly disturbed
samples had negligible strength after they were thawed in the laboratory 125
868
-------
days after placement. Testing on June 28, 1977 showed an in situ strength of
800 psf to 1,200 psf with the higher value recorded near the surface of the
deposit. A sample tested in the laboratory on July 2, 1977 showed a strength
slightly less than 500 psf. This material was quite sensitive to sampling
disturbance. Laboratory tests on Mix 4 prior to field placement had shown shear
strength of 4,600 psf.1 When the pool was sampled on September 19, 1978, the
In situ strength values varied from 223 psf to 186 psf with the higher value
recorded closer to the surface of the deposit. This is a drastic reduction in
strength 664 days after placement compared to the" strength values obtained at
earlier times and compared to the initial pre-placement laboratory test results.
The material had been frozen at the time tests were performed 90 days after
placement and the high values obtained at that time are not relevant. However,
when the pool was tested 216 days after placement, the material had completely
thawed and the strength values obtained at that time served as a basis, for
comparison for the values obtained in September 1978. It appears that a drastic
reduction in strength of the material has occurred as a result of weathering
during the intervening year's time.
Samples obtained from the pool on September 19, 1978 were tested on
September 24, 1978 in the laboratory in unconfined compression. These samples
showed shear strengths slightly higher than the in situ strength values obtained
at the site: from 196 psf to 336 psf. This is the only instance in the testing
program in which the laboratory shear strength values were higher than the _in_
situ vane shear test values. In all other cases, the laboratory values were
lower than the in situ values, primarily because of disturbance of the material
during the sampling operation, during the transport of samples to the laboratory
and during the preparation of the samples in the testing devices. These test
results appear to indicate that the material has deteriorated in the field to
the point where weathering .effects have severely disturbed the mixture and
greatly reduced the strength of the material. The moisture content and cor-
responding solids content of the materials taken from Pool No. 2 indicate that
the materials have not changed appreciably in moisture content since they were
placed in November 1976.
Permeability tests were performed on samples obtained from Pool No. 2 as
soon as those samples could be thawed and prepared for the tests. The values
of coefficient of permeability for test samples taken nominally at 30 days and
at 90 days after placement varied from about 0.7 to 1.6 x 10-5 cm/sec. A test
was run on a sample obtained on September 19, 1978 and the corresponding perme-
ability value obtained for that sample was 5.2 x 10"6 cm/sec. Laboratory tests
869
-------
on Mix 4 prior to field placement had indicated a coefficient of permeability
of 4.5 x 10~7 cm/sec.1
Pit No. 3
Pit No. 3 was filled between July 13 and July 16, 1977, intermittently
with truckloads of Mix 8 consisting of sludge (50% solids content), an amount
of flyash equal in weight to one-half the dry weight to the sludge solids, and
commercial lime in an amount equal to 3% of the dry weight of the sludge solids.
This mix had a nominal solids content of 60.5%. On July 19, three days after
finish of placement, samples were obtained and in situ strength measurements
were made. The in situ strengths varied from about 100 psf to more than 3,000
psf. The very stiff zone at a^depth of 4.5 to 5 feet corresponded approximately
to the surface of the first day of filling of the deposit. Exposure at the top
of this deposit to sun and open air during July before the remainder of the mix
was placed caused drying with subsequent lime-flyash reactions and hardening.
The samples taken on July 19 were tested on July 29 and August 1 in the laboratory.
The sample taken from a depth of 12 inches crumbled when it was extruded.
Samples from 6 and 6% feet depth were cracked and disturbed and could not be
tested. Tests on the remaining samples showed strengths varying from about
1,300 psf to 1,800 psf at moisture contents of 40% to 53%. The results of
these tests, 15 days after placement, are shown in Figure 4. Samples again
were obtained on August 24, 39 days after placement, but only with great diffi-
culty. It was not possible to insert the vane shear strength device into the
deposit at that time. Samples were obtained only by driving the sample tubes
into the mixture with a 16-lb sledge hammer. The samples so obtained were severely
disturbed. When these samples were tested on September 2-3, 48 days after
placement, shear strength values ranged from 200 psf to 2,200 psf. The high
value was obtained on a sample from a depth of about 18 inches with a moisture
content of about 37%. The remainder of the samples had apparent strengths of
750 psf or less. These samples contained cracks and voids from the disturbance
created while they were being obtained. Laboratory tests on Mix 8, prior to
field placement had shown a 60-day undrained shear strength of 3,650 psf.1
On October 6, 1978 an attempt was made to obtain in situ vane shear
strength values from this deposit but it was impossible to insert the vane
shear device to a significant depth in the deposit. The surface of the deposit
was disturbed, apparently from the effects of weathering, and the dry material
was very crumbly. The samples obtained at this time were tested in the laboratory
870
-------
461 days after placement in the field. The shear strength values for these
samples ranged from 127 psf to 3,034 psf. The very low value was obtained on
a sample which split and crumbled as it was being tested in the preliminary
stages of the test. This sample is not considered to be truly representative
of the material in the deposit.
The upper 3 to 4 feet of this deposit appear to have deteriorated somewhat
through the effects of weathering since the values of strength of the samples
obtained in October 1978 are lower than values of strength obtained at earlier
dates at the same depths. However, below a depth of about 4 feet the material
appears to be intact and the shear strength in the most recent test varied from
slightly more than 1,500 psf to more than 3,000 psf. These samples did not
display the disturbance which was apparent in the samples tested' earlier because
a drilling device was utilized in obtaining the later samples. This device
eliminated much of the disturbance associated with hammering a sampling tube
into the sludge deposit. Wherever possible, sampling tubes were rotated into
the sludge deposit, or they were pushed with a steady smooth push into the
material rather than being hammered. The results of the strength tests are
shown in Figure 4 where laboratory shear strength values are combined with
field shear strength values.
Two samples obtained on July 19, 1977 from Pit No. 3 were subjected to
permeability tests on September 3, 1977, 49 days after placement and on
September 14, 1977, 60 days after placement. These two tests gave indicated
—fi —fi
values of coefficient of permeability of 7 x 10" cm/sec and 2 x 10" cm/sec
at moisture contents of 26% and 37%, respectively. Three samples taken on
October 6, 1978 also were subjected to permeability tests. These samples had
moisture contents between 41 and 45%, or solids contents between 69% and 71%.
These solids contents are appreciably higher than the solids content nominally
established at the time of placement (60%). The samples were somewhat dry and
contained some very small cracks. These samples yielded values of coefficient
of permeability of 1.4 x 10" cm/sec, 5.9 x 10" cm/sec, and 0.9 x 10" cm/sec.
The lower value of permeability coefficient obtained on the sample from a depth
of 65-70 inches in the deposit may reflect a lower permeability for the
material in this pit below the depth of material affected by weathering.
Laboratory tests on Mix 8 prior to field placement had shown a coefficient of
permeability of 4.1 x 10"6 cm/sec.1
871
-------
+J
If-
Q.
O)
O
8
3 (Mix 8)
O
A
a
15 Days (lab)
48 Days (lab)
461 Days (lab)
3 Days (in-situ)
500 1000 1500 2000
Shear Strength (psf)
Figure 4 Shear Strength Values for Pit No. 3
2500
3000
872
-------
Pool No. 5
Pool No. 5 was filled from July 8 to July 11, 1977 with Mix 9, a mixture
of sludge (50% solids content), commercial lime (3% by weight of the dry weight
of the sludge solids), and flyash (150% of the dry weight of sludge solids).
The mixture had a nominal solids content of 71.7%. On July 11, samples were
obtained and in situ shear strength measurements were made. The strength values
varied from 400 psf to about 650 psf. A sample tested in the laboratory next
day showed only slight disturbance and yeilded a strength value of about 550 psf.
Laboratory tests prior to field placement had shown a 60-day shear strength of
6,250 psf.1
On August 11, 1977, 31 days after placement, the sludge had hardened
sufficiently that the vane shear strength device could not be inserted into the
deposit. Samples were obtained by driving thin-walled sampling tubes into the
sludge. When these samples were extruded from the tubes in the laboratory on
August 22, it was seen that the sajnples had cracked and split badly. The high
flyash content of this mix made the material extremely brittle and susceptible
to disturbance. On September 19, 1978, it was impossible to obtain a reading
of shear strength in situ because the vane shear device could not be inserted.
This high flyash mixture which had been exposed to the elements was disturbed
only on the uppermost surface. Several samples were obtained and returned to
the laboratory for testing. Prior testing had indicated a shear strength value
of 550 psf on a sample taken on the day of placement. Samples taken on
September 19, 1978 were tested 445 days after placement; two of these samples
were severely disturbed. One sample yielded a strength value of 2,770 psf;
the solids content for this sample was 69%, very close to the initial solids
content established for the pool.
A sample taken on July 11 was tested on September 9, 1977, 59 days after
placement, for coefficient of permeability. The indicated coefficient of
permeability was 4.5 x 10"6 cm/sec, compared to a coefficient of 5.7 x 1Q~7
cm/sec obtained in the laboratory prior to field placement on undisturbed
material.1
o
The samples taken from Pool No. 5 during September 1978 were so friable
and brittle it was not possible to prepare a sample for permeability testing
in the laboratory.
Pit No. 4
Pit No. 4, only 4 feet deep compared to the 8-foot deep earlier pits,
was filled during the period of July 25 to July 31, 1977 with Mix 12 consisting
of sludge (50% solids content), an equal dry weight of flyash, and calcium
873
-------
hydroxide equal to 3% of the dry weight of the sludge solids. The nominal
solids content of the mix was 67%, with a moisture content of 50%. Samples
were obtained and in situ shear strength values were measured immediately
after the pit was filled. The in situ strength values varied from 700 psf to
2700 psf as is shown in Figure 5. A significant variation in resistance to
penetration of the vane shear device at the time of the testing indicated
definite layering in the deposit. Samples taken on July 31, 1977 were subjected
to unconfined compression tests on August 1 and 2. Some sample disturbance was
noted, but it was not severe. Shear strength values obtained from the unconfined
compression strength tests ranged between about 500 psf and 1,000 psf, from
one-half to one-third of the in situ strength for the same depths. The lower
strength values resulted from the inevitable sample disturbance that was created
when sampling tubes were hammered into the brittle but stiff mix in Pit No. 4.
Samples selected for permeability tests also proved to be unsuitable because
of the presence of cracks and voids in the samples.
On August 31, 31 days after placement, it was impossible to insert the
vane shear device into the hardened sludge. Samples were obtained only with
great difficulty. Laboratory strength tests on September 3 and September 6,
1977 showed shear strength values of from 3,000 psf to 1,300 psf for samples
with moisture contents of 36% and 37%, respectively. This moisture content
was significantly below the placement moisture content of 50%. The lower
strength was obtained on a sample from 30-34 inches while the higher strength
was obtained for a sample from a depth of 9 to 13 inches. Cracks throughout
the samples precluded permeability testing.
This pit was sampled and tested again on October 6, 1978. At that time,
the surface of the deposit was too hard to permit inserting the vane shear
test device. Samples were obtained by use of the drilling rig. The sample
tubes were filled by pushing the tubes into the deposit or by rotating the
tubes into the deposit with the drilling rig. These samples were subjected
to unconfined compression strength tests 452 days after placement. The moisture
contents for the unconfined compression test samples varied from 41% to 51%.
Many of the samples contained crevices and lateral cracks. The shear strength
values ranged from a low of 492 psf for a sample taken between 2 inches and
6 inches depth in the deposit, to a high of 2,034 psf for a sample from 23
to _27 inches depth. The corresponding solids content for these samples was
68%. Laboratory tests on samples of Mix 12 prior to field placement had
shown a 60-day shear strength of 6,850 psf.1
874
-------
Q.
(U
Q
Pit No. 4 (Mix 12)
O 0 Days
500 1000 1500 2000 25000
Shear Strength (psf)
Figure 5 In-Situ Shear Strength Values for Pit No. 4
3000
875
-------
Because of disturbance during sampling in the program carried out during
1977, it had not been possible to perform permeability tests on samples taken
from Pit No1. 4. The samples taken from Pit No. 4 on October 6, 1978 did not
exhibit the high degree of disturbance associated with the samples taken earlier,
although some of the samples had small cracks and crevices. These samples
were subjected to laboratory permeability tests 473 days after placement of
the material in the field. The resultant coefficients of permeability for
samples from Pit No. 4 were 3.0 x 10~6 cm/sec, 3.3 x 10 cm/sec, and 1.8 x
10 cm/sec, for samples with corresponding solids contents of 68%, 71%, and
68%. Laboratory tests on Mix 12 prior to field placement had shown a co-
efficient of permeability of 9.2 x 10 cm/sec, at 67% solids.1
Pit No. 5
Pit No. 5 was filled on July 19 to 21, 1977 with the same mixture, Mix 11,
which had been placed earlier in Pool No. 7: sludge (50% solids content), an
equal dry weight of flyash, and commercial lime (3% of the dry weight of the
sludge). The mix was placed in layers and compacted with a shovel device on
a tractor excavator. On July 31, 10 days after placement, the surface of the
mix could not be penetrated with the vane shear device. Samples were taken
but only with great difficulty. These samples were significantly disturbed.
Tests on these samples, on August 2, 1977, showed shear strengths of from
750 psf to 2,000 psf with the higher value obtained from a sample taken near
the surface of the deposit. The moisture contents of the samples were about
41%, slightly lower than the nominal mix placement moisture content of 50%.
On August 31, 1977, additional samples were obtained. Laboratory tests
on these samples showed strengths of 750 psf and 1,750 psf at moisture contents
of 39% and 34%, respectively. These samples had been disturbed significantly
when they were obtained.
On October 12, 1978 samples were obtained through the use of the drill
rig, without a significant degree of disturbance. When these samples were
tested in the laboratory 466 days after the material had been placed, one sample
showed a very low value of strength of 152 psf, but this sample had been taken
between a depth of 1 inch and 5 inches in the deposit and it contained very
significant lateral cracks. Another sample taken between depths of 28 and
34 inches split when it was being extruded from the sampling tube. The re-
maining samples were intact and showed good test results in that no abrupt
splitting failure occurred; the test results could be considered representative
of the material in the field. The shear strength for these materials varied
876
-------
from 1,100 psf to more than 1,500 psf at sol Ids contents of approximately 70%
(moisture contents between 34 and 43$). Laboratory tests on samples of Mix 11
performed prior to field placement had shown an undrained shear strength of
4,250 psf, at a solids content of 67%.!
Permeability tests were performed on samples taken on July 31, 1977, 10
days after placement of the material in the pit. These tests were performed
57 days after placement and 90 days after placement. The obtained coefficients
of permeability for these samples were 4.5 x 10"6 cm/sec for the 57-day-old
sample, and 16.5 x 10~ cm/sec for the 90-day-old sample. No permeability
tests were possible on the samples taken in August 1977 because of severe
sample disturbance. The samples obtained on October 12, 1978 from Pit No. 5
were subjected to laboratory permeability tests 484 days after placement of
the material in the field. The moisture content of these samples was 45%,
with a corresponding solids contents of 69%. These tests showed coefficients
of permeability of 1.2 x 10"6 cm/sec for the sample taken from a depth of
12-18 inches and 1.4 x 10"6 cm/sec for the sample taken from a depth of 34-
40 inches. Laboratory tests on Mix 11 prior to field studies had shown a
coefficient of permeability of 2.9 x 10"6 cm/sec.1
Pool No. 9
Pool No. 9 was filled on August 1, 1977 with Mix 10, consisting of sludge
(50% solids content), an amount of flyash equal in weight to the dry weight
of the sludge solids, and Portland cement (3% of the dry weight of the sludge
solids). The mix had a nominal solids content of 67% with a corresponding
moisture content of 50%.
On August 11, 10 days after placement, samples were obtained and in situ
shear strength measurements were made. The strength of the material at all
depths exceeded the capacity of the vane shear device so that it was possible
only to say the strength was greater than 3,000 psf. Samples were obtained
only with great difficulty. One of the samples was tested in the laboratory
even though it was cracked slightly. This sample showed a shear strength of
3,250 psf and had a moisture content of 40%, 22 days after placement. All
of the samples contained cracks and were disturbed to such a degree that
permeability testing was not feasible.
Samples were obtained again on September 21, 1977, 51 days after place-
ment. The material could not be penetrated with the vane shear device.
Also, it was very difficult to insert tube samplers into the hardened mix and
the samples obtained were disturbed severely. One of the disturbed samples
877
-------
yielded a shear strength value of 7,445 psf at a moisture content of 40%, when
it was tested on October 29, 1977, 79 days after placement. The samples
that were tested were cracked but strong; the cracks could not be closed by
imposing external pressure. Thus, permeability testing was not appropriate.
When Pool No. 9 was visited on September 19, 1978 again it was not possible
to insert the vane shear device into the deposit. However, relatively undis-
turbed samples were obtained with the drill rig, and these samples were re-
turned to the laboratory and were tested 417 days after placement. These two
samples with moisture contents of 37% and 40% showed shear strength values of
5,106 psf and 4,900 psf, at solids contents of 73% and 71%, respectively.
Laboratory tests on undisturbed samples of Mix 10 prior to field studies had
indicated an undrained shear strength of 2,725 psf, 60-days after sample
preparation, with a solids content of 67%.1
One of the samples obtained on September 19, 1978 from Pool No. 9 was
sufficiently undisturbed to permit a permeability test to be performed 427
days after placement. This sample had a moisture content of 43% with a cor-
responding solids content of 70% and it yielded a permeability value of 3.2
-5
x 10 cm/sec. Lab tests on Mix 10 prior to field studies had indicated a
-5
coefficient of permeability of 5 x 10 cm/sec.
Pool No. 10
Pool No. 10 was filled on August 1 and 2, 1977 with Mix 7 consisting of
sludge (65% solids content) and flyash in an amount equal in weight to the dry
sludge solids; no lime was added to this mix. The nominal solids content of
the mix was 78.8%, or 27% moisture content. On August 11, 1977 nine days after
placement, samples were obtained and in situ strength measurements were made.
The strength values varied from about 650 psf to about 1,300 psf. Samples
taken at this time crumbled and broke apart when they were extruded from the
sampling tube, on August 23 and 25, 1977. Moisture contents for these samples
were 35% and 37%, values higher than the nominal initial mix value of 27%.
One of the samples, though cracked and full of voids, supported more than 1,000
psf before it failed abruptly. No permeability tests could be performed on
these cracked samples. Sample disturbance was severe; this disturbance was
not caused by intense hammering during sampling. Sampling tubes had been in-
serted without hammering. The high flyash content and lack of cementing agent
in this mix made it very crumbly and susceptible to disturbance.
878
-------
On September 21, the surface of the deposit had hardened to a condition
so that the vane shear device could not be inserted. Samples were obtained
but only with considerable effort. When these samples were tested on October 3,
1977 and on October 19, 1977, they yielded shear strength values of 1,650 psf
and 850 psf, respectively. The samples crumbled somewhat during removal from
the sampling tubes and cracked further during test set-up so that no permea-
bility tests were possible.
On September 19, 1978 samples were obtained and it was also possible at
that time to insert the vane shear device into the deposit. Field vane shear
strengths varied from 2,100 psf to 2,775 psf. Samples obtained at this time
were tested in the laboratory, but the very friable and crumbly nature of the
material led to significant sample disturbance. One sample with a solids content
of 73% yielded a shear strength of 771 psf. The other sample with a solids
content of 71% crumbled before it could be tested in unconfined compression.
Laboratory tests on Mix 7 prior to field testing had shown a 60-day undrained
shear strength of 1,500 psf.1
Because of the extremely friable and crumbly nature of the samples ob-
tained from Pool No. 10 all during the testing program it was not possible
to perform permeability tests on samples from that pool. Laboratory tests
prior to field studies had indicated a coefficient of permeability of 7 x
10~ cm/sec for undisturbed samples of Mix 7.1
ADDITIONAL TESTING PROGRAM
Because of the high strength results which were obtained from Pit No. 4
and Pit No. 5 during the vane shear testing and during the unconfined compression
testing of samples taken from those pits, it was decided that it would be
appropriate to conduct plate load tests on those deposits. Consequently, on
October 12, 1978, the University of Louisville investigators conducted plate
load tests on those two pits. In the plate load tests, a steel loading plate
12 inches by 12 inches was placed on the surface of the mix material in each
of the pits after the surface had been swept free of loose debris and had been
leveled. Dial gauges were positioned at the corners of the steel plate to
measure the downward movement of plate corners under load. An hydraulic jack
was utilized to exert downward force on the steel plates used in the loading
tests. The results of these loading tests are shown in Figures 6 and 7. In
the test on Pit No. 4, a total load of 28,000 Ibs was placed on the material
to cause a settlement of 1.2 inches. At a settlement of 1.0 inch the pressure
exerted by the plate was slightly more than 25,000 Ibs/sq ft. This loading
879
-------
Pit No. 4
Plate Load Test Results
October 12, 1978
5,000 10,000 15,000 20,000
Load on One-Foot Square Plate (Ibs.)
Figure 6 Plate Load Test Results for Pit No. 4
25,000
30,000
880
-------
O)
OJ
oo
Pit No. 5
Plate Load Test Results
October 12, 1978
5,000
Figure 7
10,000 15,000 20,000
Load on One-Foot Square Plate (Ibs.)
Plate Load Test Results for Pit No. 5
25,000
30,000
881
-------
intensity with the accompanying 1-inch settlement indicates that the material
in Pit No. 4 would certainly be competent to bear significant foundation loads.
A similar loading test was performed on the material in Pit No. 5, with
removal of the load and reloading for two cycles, to investigate the rebound
characteristics of the material. The results of this test are shown in Figure 7.
It is obvious from the figure that during the loading to more than 16,000 IDS/
sq ft with an accompanying settlement of approximately one-half inch, much of
the settlement was recoverable; when the load was removed, the plate rose
producing a net settlement of only about 0.13 inches. When the load was re-
applied, the net settlement under the 16,000-psf pressure again was only
slightly in excess of 0.5 inch, but when the load was removed after the second
cycle of loading a permanent net settlement of almost 0.4 inches was noted.
A third cycle of loading extended to pressures greater than 27,500 psf, and
produced a settlement of 1.4 inches. When the load was removed, a net settle-
ment of 1.1 inches had occurred. The behavior of the material in Pit No. 5
under this load test indicates that the material would be competent to bear
very high foundation loads.
SYNOPSIS
The material presented in preceding sections of this paper indicates that
many factors have influenced the sampling and testing of the sludge-flyash-
additive mixtures at the Cane Run plant. These factors also have influenced
the interpretation of test results. For example, sample disturbance was severe
whenever frozen wastes were sampled; this was particularly true with respect
to carbide lime sludge mixtures with very low solids contents. Sampling dis-
turbance also was significant for material that had frozen and subsequently
thawed, even in the summer, because the mixture was brittle and sensitive to
disturbance in most cases. The use of a drilling rig to secure samples in the
1978 sampling program alleviated this difficulty to a certain extent but did
not entirely eliminate sampling disturbance. Additionally, significant layer-
ing was noted in pits and pools whenever the filling operations had been inter-
rupted for a period of 12 hours or more; this layering effect was especially
severe in the more fluid mixtures and those mixtures placed during extremely
cold weather. However, layering effects were also very significant for those
materials which were placed in truckloads intermittently during extremely hot
weather; in these instances, pozzolanic reactions were accelerated near the
surface of the sludge mixtures which were exposed to sunlight and high tempera-
tures, and very stiff layers were produced in some deposits.
882
-------
Because of the sampling disturbance mentioned above and because of the
layering in the deposits, the test data should be viewed with some skepticism.
The in situ vane shear strength values are the most reliable data obtained
in this investigation. Sampling disturbance reduced the strength of the
materials tested in laboratory unconfined compression tests as shown by com-
paring lab test results for undisturbed samples obtained prior to field studies
with those obtained on samples from the field. The effects of sampling distur-
bance were even more significant in terms of the measured coefficients of
permeability obtained in the laboratory. These coefficients of permeability
should be considered the maximum values of vertical permeability for the
materials. Mass values of vertical permeability in situ may be as much as an
order of magnitude lower than the measured laboratory values, because of cracks
and fissures in the samples tested in the laboratory. Because of the layering
effects mentioned above, mass horizontal permeability is likely to be much
higher than mass vertical permeability.
To illustrate the effects of layering and weathering, behavior in Pits
1-5 can be reviewed. Mix compositions for pits/pools are given in Table 1.
With respect to the materials in Pit No. 1, layering, especially in the
upper 2 to 3 feet of this deposit, was particularly pronounced. This mix
was placed in truckloads with long interruptions between deposits of material.
In the upper 2 to 3 feet, the layering may have been caused by poor mixing of
the waste materials or by the inclusion of a truckload of much higher solids
content material. The effects of freezing may have been accentuated by the
presence of layers of more fluid material, created by the intermittent deposi-
tion of the wastes in the pit. Ice lenses may have formed when the material
froze to a depth of 3 to 4 feet during the winter of 1977. The effects of
weathering do-not-appear-to have extended below a depth of about 3 feet. The
strength of the material in place today could be taken at- no higher than 500
Ibs/sq ft for purposes of design. The material is not brittle.but is rather
plastic and is undergoing very slow consolidation. The strength tests and the
permeability tests which were performed on these materials indicate that virtu-
ally no cementation has occurred in this deposit.2
In Pit No. 2, the upper 4 feet of the material exhibits severe layering
effects. This material froze within the month after it was placed. As in
Pit No. 1, it is likely that somewhat wetter areas were formed near the top
of each truckload of material that was placed in the pit. This semi-segregation
of materials in layers near the top of the pit may have led to the forming of
883
-------
Table I PIT/POOL MIX COMPOSITIONS
Test Site
Pit No. 1
Pool No. 1
Pool No. 2
Pit No. 2
Pool No. 3
Pool No. 4
Pool No. 5
Pit No. 3
Pool No. 6
Pit No. 4
Pool No. 8
Pit No. 5
Pool No. 7
Pool No. 9
Pool No. 10
Sludge Used
Carbide lime
24% solids
Carbide lime
55% solids
Carbide lime
42% solids
Carbide lime
55% solids
Cornm. lime
50% solids
Comm. lime
50% solids
Comm. lime
50% solids
Comm: lime
50% solids
Comm. 1 ime
50% solids
Comm. lime
50% solids
Flyash:
Sludge
1:1
(dry)
0:1
(dry)
1:1
(dry)
1:1
(dry)
1.5:1
(dry)
0.5:1
(dry)
1:1
(dry)
1:1
(dry)
1:1
(dry)
1:1
(dry)
Fixative
5% carb.
lime
5% carb.
lime
5% carb.
lime
3% carb.
lime
3% comrn.
lime
3% comm.
lime
3% Ca(OH):2
3% comm.
lime
3% Port.
cement
None
Date
Placed
11/11/76
11/24/76
12/06/76
12/09/76
7/11/77
7/16/77
7/31/77
7/21/77
8/01/77
8/2/77
Nominal
Solids Content
39.3% \
56.2%
59.8%
71,0%
71.7%
60.5%
67.0%
67.0%
67.0%
78.8%
884
-------
ice lenses when the materials froze to a depth of 3 to 4 feet below the surface.
Ice lensing would have pre-compressed layers of material to yield higher
strengths and lower moisture contents. The strength and moisture content values
obtained for the upper 4 feet of Pit No. 2 agree with this hypothesis. The
materials below a depth of about 4 feet do not appear to have been affected
by weathering since the time of placement. The strength of the materials below
that depth appear to have increased somewhat in the first two months after
placement but they do not appear to have increased significantly since then.
The materials in this pit reflect rather poor mixing. The strength of the
material would still be insufficient to support sizable structures. The values
of coefficients of permeability obtained for the materials in Pit No. 2 varied
—fi fi
from about 4 x 10~ cm/sec to 10 x 10 cm/sec, except for values in the upper
18 inches of the deposit where the materials have apparently been severely
affected by freeze-and-thaw cycles and other weather effects.
In Pit No. 3, the placement of material verged on random dumping. Large
voids were discovered between lumps of the mixture during subsequent testing
operations. The material in Pit No. 3 is' irregular, contains numerous voids,
and could not be considered adequate to support even light structures. Two
hard layers were found in the pit between depths of 12 inches and 24 inches
and again between depths of 48 inches and 60 inches; these hard layers were
apparently caused by the method of placement of the material in the pit. The
strength of the material at best could be taken at 1,000 Ibs/sq ft for the
intact portions of the deposit. Likewise, the intact lumps of material in
the pit showed permeabilities of about 10" cm/sec, but the mass permeability
of the deposit would be much higher than that value because of the numerous
voids contained in the pit.
The material in Pit No. 4 showed strength values from a high of 3,000
Ibs/sq ft near the surface of the deposit to a low of about 1,000 Ibs/sq ft
near the bottom of the pit, within one day after the material was placed in
mid-summer 1977. The material is subject to sampling disturbance as shown by
the fact that the unconfined compression strength values of shear strength,
obtained on samples tested in the laboratory, amounted to about one-half the
value of undrained shear strength obtained with a vane shear device in the
field. When the pit was visited in the fall of 1978, a powdery cracked crust
was found in the upper few inches in the deposit. Below a depth of 18 inches,
this weathering effect was not noted. The undrained shear strength of the
material below 18 inches could be taken at a value of 2,000 Ibs/sq ft minimum.
885
-------
The samples taken at 0 days and 31 days after placement were severely disturbed
because they were taken by hammering the tubes into the deposit. The samples
taken in 1978, approximately 450 days after placement, were taken with a drill
rig and were not severely disturbed. The fact that these samples yielded shear
strength values somewhat lower than earlier test values indicates that the
material in the pit may have weakened somewhat through weathering effects.
The values of coefficient of permeability, 2-3 x 10" cm/sec, were obtained
on samples which exhibited very little cracking and disturbance. It is likely
that the permeability of the material in the pit is rather close to 10" cm/sec.
In Pit No. 5, the surface of the deposit has deteriorated'with weathering
effects evident to a depth of 12 to 18 inches. The samples that were taken
12 and 41 days after placement and later tested in the laboratory were signifi-
cantly disturbed by the sampling operation, but the samples obtained during
fall 1978 were not disturbed in that way. The undrained shear strength of the
material in Pit No. 5 could be taken at a minimum value of 1,500 Ibs/sq ft,
on the basis of the strength tests performed.
Plate load tests indicate that the materials in Pit 4 and in Pit 5 are
brittle and tend to gradually yield with a collapsing and compression of voids
in the material under load. These materials supported pressures in excess of
25,000 Ibs/sq ft during the loa'd tests with measured settlements between 1
and 1.5 inches. The load test results would indicate that a safe bearing
value for the materials in Pit No. 4 and in Pit No. 5 would be at least 5,000
Ibs/sq ft. These values are considerably in excess of the values obtained
through interpretation of the laboratory and field strength tests. The
differences between the laboratory and field strength tests, and the results
of the plate load tests, reflect the brittle nature of the material. In an
undrained shear test, the brittle material tends to be strain-weakening; a
failure occurs at a very low value of strain with a loss of strength with
subsequent strain. In a compression test such as a plate loading test, the
collapse of voids in the material would tend to make the wastes less com-
pressible with increasing strain. In other words, the behavior of the materials
in the two tests is different and the results of the two tests are not incon-
sistent. These materials behave somewhat akin to very weakly cemented aggre-
gates or very weakly cemented natural rock materials, rather than cohesive
soils.
Finally, it is important to note the behavior in Pools 9 and 10, since the
mix in Pool 9 contains a very effective cementing agent, while the high-solids
sludge-flyash mix in Pool 10 contains no fixation additive.
886
-------
The material in Pool No. 9 is extremely hard at the present time. Very
high strength values were obtained in this pool soon after the material was
placed. These high strength values indicate the great effectiveness of
Portland cement as a binding and stabilizing agent in sludge and flyash mix-
tures. Samples obtained from Pool No. 9 were disturbed significantly, but
even so, the minimum shear strength value for the samples obtained in 1978
was 5,000 Ibs/sq ft. A disturbed sample taken late in 1977 gave a value of
shear strength in excess of 7,000 Ibs/sq ft. The material in Pool No. 9
also showed very little evidence of deterioration from weathering effects; the
surface of the deposit consists of a very hard crust. The permeability of
the materials in Pool No. 9 should not be greater'than the value of 3 x 10
cm/sec obtained on a disturbed sample from that pool. It is likely that the
in situ permeability of this pool is considerably lower than that value.
The material in Pool No. 10 is extremely friable, exhibiting very little
permanent cementation. The shear strength values of the material in Pool No.
10 varied from 700 to 1200 Ibs/sq ft nine days after placement, and 30 days
after placement the material was too hard"to penetrate with the vane shear
strength apparatus. Samples, obtained in 1978, and tested more than 400 days
after placement, showed shear strength values (in laboratory tests) in excess
of 2,000 Ibs/sq ft. The undrained shear strength of the material could be
taken conservatively at a minimum value of 2,000 Ibs/sq ft. It appears that
the surface of the deposit, in the upper 12 to 18 inches of the material, has
deteriorated somewhat through weathering effects. The material is extremely
sensitive to disturbance such as occurred when tubes were hammered into the
deposit in order to obtain samples. All the samples obtained from Pool No. 10
contained cracks and fissures in such a degree that no permeability tests could
be run on samples from that pool.
REFERENCES
1. Mohn, N.C., A.L. Plumley, A.L. Tyler, and R.P. Van Ness, "Environmental
Effects of FGD Disposal: A Laboratory/Field Landfill Demonstration,"
Paper presented at EPA Symposium on Flue Gas Desulfurization, March 5-8,
1979, Las Vegas, Nevada.
2. Hagerty, D.J., and C.R. Ullrich, "Scrubber Testing and Waste Disposal
Studies," Final report prepared for Louisville Gas & Electric Company under
EPA Contract No. 68-02-2143, December 15, 1978.
887
-------
SUMMARY OF UTILITY DUAL ALKALI SYSTEMS
Norman Kaplan
Industrial Environmental Research Laboratory
Office of Research and Development
Environmental Protection Agency
Research Triangle Park, North Carolina 27711
EPA Flue Gas Desulfurization Symposium
Las Vegas, Nevada
March 5-8, 1979
888
-------
ABSTRACT
The Environmental Protection Agency (EPA) has been actively involved in the
progressive development of dual alkali flue gas desulfurization (FGD) tech-
nology in the U.S. from bench-scale to pilot plant to prototype to full-scale
application. This logical progression has allowed a better understanding of
the chemistry and made possible the identification of optimum levels of design
and process parameters at a moderate cost. EPA continues to fund research and
development activity aimed at identifying new variants of the process and efforts
toward improving the existing processes, including the use of limestone for re-
generation.
This year is a major milestone in the application of dual alkali technology
to full-scale utility systems in the U.S. Three full-scale systems ranging in
size from 250 to 575 MW are scheduled to come onstream during 1979. The owners
of these systems, their size, and the vendors providing them are: (1) Central
Illinois Public Service Co., Newton No. 1, 575 MW unit, Envirotech; (2) Louis-
ville Gas & Electric Co., Cane Run No. 6, 277 MW unit, Combustion Equipment
Associates/Arthur D. Little, Inc.; and (3) Southern Indiana Gas & Electric Co.,
A.B. Brown No. 1, 250 MW unit, FMC. All of these systems are of the concentrated,
sodium-based dual alkali type and they service boilers firing high-sulfur coal.
This paper presents a brief description, the design bases, and a summary of ven-
dor guarantees and reported costs for these systems. In addition, it presents
the fundamentals of dual alkali technology, reviews the status of commercial
dual alkali systems, and briefly describes some of the non-sodium sulfite based
dual alkali processes.
889
-------
NOTES
Company Names and Products.
The mention of company names or products is not to be considered an
endorsement or recommendation for use by the U.S. Environmental Pro-
tection Agency.
Consistency of Information.
The information presented was obtained from a variety of sources (some-
times by telephone conversation) including system vendors, users, EPA
trip reports, and other technical reports. As such, consistency of
information on a particular system and between the several systems dis-
cussed may be lacking. The information presented is basically that which
was voluntarily submitted by developers and users with some interpretation
by the author. The order of presentation of information or the amount of
information presented for any one system should not be construed to favor
or disfavor that particular system.
Units, of Measure
EPA policy is to express all measurements 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. Generally, this paper uses British units of measure.
The following equivalents can be used for conversion to the Metric system:
British
Metric
5/9 (°F-32)
-I
ft3
grain
•n'2
m.^
in.3
Ib ('avoir.)
ton (long)
ton (short)
gal.
°C
0.3048 m
0.0929 m2
0.0283 m3
0.0648 gram
2.54 cm .
6.452 cm2
16.39 cm-3
0.4536 kg
1.0160 m tons
0.9072 m tons
3.7853 liters
890
-------
ACKNOWLEDGEMENT
The author is indebted to Messrs A.H. Abdulsattar and D.A. Burbank of Bechtel
National, Inc. for writing the first draft of this paper, making arrangements
for its publication, and for preparing the visual aids for the talk at the
Fifth EPA FGD Symposium in Las Vegas.
Also, the help of several other individuals, without whose cooperation this
paper would not have been possible, is sincerely appreciated. Those deserving
special thanks are:
Individual
Bloss, H.E.
Grant, R.J.
Kawahara, D.Y.
LaMantia, C.R,
Lee, G.C.Y.
Ramirez, A.A.
Van Ness, R.P.
Wagner, N.P.
Company
Envirotech
Central Illinois Public Service Co.
Bechtel National, Inc.
Arthur D. Little, Inc.
Bechtel National, Inc.
Food Machinery Corporation (FMC)
Louisville Gas & Electric Co.
Southern Indiana Gas & Electric Co.
891
-------
SUMMARY OF UTILITY DUAL ALKALI SYSTEMS
Section 1
INTRODUCTION AND BACKGROUND
1.1 INTRODUCTION
At the 1976 EPA Symposium on Flue Gas Desulfurization (FGD), held in New Orleans,
it was reported that the EPA planned to co-fund a full-scale utility boiler dual
alkali (D/A) demonstration program and that three viable proposals for instal-
lation of a full-scale, dual alkali, FGD system on coal-fired utility boilers had
been received. The successful utility bidder was the Louisville Gas & Electric Co.
(LG&E). LG&E had contracted to install the Combustion Equipment Associates/A.D.
Little (CEA/ADL) designed D/A FGD unit at LG&E's Cane Run No. 6, 277 MW boiler.
The demonstration monitoring program on the Cane Run FGD system is scheduled
to begin in April 1979.
This is a memorable year for application of dual alkali technology to utility
boilers in the U.S. because three full-scale systems, ranging in size from 250
to 575 MW, are scheduled for startup this year. The owners of these systems,
their size, and the vendors providing them are: (1) Central Illinois Public
Service Co., Newton No. 1, 575 MW unit, Envirotech; (2) Louisville Gas & Elec-
tric Co., Cane Run No. 6, 277 MW unit, CEA/ADL; and (3) Southern Indiana Gas &
Electric Co., A.B. Brown No. 1,,250 MW unit, FMC. All of these systems are of the
concentrated, sodium based dual alkali type and they service boilers firing high-
sulfur coal.
This paper presents a brief description, the design bases, and a summary of ven-
dor guarantees and reported costs for these systems. In addition it presents
the fundamentals of dual alkali technology, reviews the status of commercial
dual alkali systems, and briefly describes some of the non-sodium sulfite based
dual alkali processes.
Before describing the three full-scale utility systems, a review of the techno-
logical background is warranted. Like any specialized technology, a host of
terms for and variations of the sodium-sulfite based dual alkali processes have
evolved. A detailed discussion of terms and the significant process, design and
cost considerations is included in Appendices A and B. However, a brief qualita-
tive discussion is given below.
1.2 BACKGROUND
"Double alkali" or "dual alkali" processes are characterized as non-recovery
S02 abatement processes which involve aqueous alkali scrubbing of sulfur oxides
from the flue gas, followed by regeneration of the scrubbing solution with
lime or limestone to precipitate the sulfite/sulfate reaction waste. These
processes are also referred to as "indirect" lime/limestone processes.
Continued developmental activity has resulted in several distinct process
variations. However, common to each process is the separation of the absorp-
892
-------
tion and the regeneration cycles. This separation permits the use of clear
absorption solutions (i.e., no slurry in the scrubber) and results in the
following advantages relative to the direct lime/limestone processes.
t The scaling, plugging, and erosion potentials within the scrubbing
loop (i.e., absorption step) are greatly reduced.
• The S02 absorption efficiency is increased since the important rate
limiting lime/limestone dissolution is not required in the scrubber.
t The higher S02 removal efficiency results in reduced absorbent
liquor circulation and simpler scrubber requirement as compared to
slurry systems.
• The production of waste solids is reduced by the higher utilization
of lime or limestone in the regeneration process.
The typical dual alkali process is illustrated in Figure 1-1. It comprises
the following basic operations:
t Absorption: reaction of sulfur oxides with an aqueous alkaline
absorbent to form soluble sulfite and bisulfite ions in solution
in the scrubber.
• Regeneration: treatment of a slip stream of the recirculating
absorbent solution with lime or limestone to precipitate the
insoluble calcium sulfite/sulfate waste product and increase the
alkalinity of the absorbent. The regeneration process is spatially
separated from the scrubbing operation.
• Dewatering: separation of the calcium-based precipitate from
the absorbent liquor and recovery of alkali liquor.
t Softening: lowering the dissolved calcium ion concentration in the
regenerated solution (i.e., subsaturating it with respect to gypsum)
to reduce scaling potential in the scrubber. Sodium carbonate is
generally used to precipitate calcium ion as calcium carbonate in
dilute dual alkali systems.
Each dual alkali process is characterized by a specific cation (Na+,
Mg , Al ), associated with the absorbent base, an alkaline earth regenerant
(lime, limestone), and a solution strength at which the active absorbent can be
regenerated (a discussion of dilute mode vs. concentrated mode D/A process is
given in Appendix A). Regeneration is accomplished via the calcium sulfite/
sulfate precipitation step.
With the exception of NH 4, all of the above are non-volatile cations, thus en-
abling the use of simpler scrubbing systems. The loss of the volatile alkali,
emission of visible plume of ammonia salts from the stack, and deposits of
893
-------
SCRUBBER
GAS FEED
SODA ASH
REHEAT
(OPTIONAL)
-On
I I
_T I SOFTENING
I | .(OPTIONAL)
LIME OR
LIMESTONE
i
STACK
SOLIDS
DISPOSAL
S02 ABSORPTION
REGENERATION
DEWATERING
Figure 1-1 TYPICAL DUAL ALKALI PROCESS
-------
ammonium compounds in the ducts and stacks have been reported as problems
with ammonia scrubbers. The cost involved in the solution of theSe problems
has retarded the development of the ammonia-based dual alkali scrubbing
technology.
A status summary of operating and planned full-scale dual alkali systems in
U.S. and Japan is given in Tables 1-1 and 1-2, respectively. The tables show
a total of approximately 7,250 MW equivalent operating and planned dual alkali
systems in industrial and utility applications in the U.S. and Japan represent-
ing 73 applications of this technology. Of these, approximately 5,775 MW equ'iya-»
lent, representing 65 applications are operational. In the U.S. eight bpe'raticinal
units approximately equivalent to 252 MW are listed. As a standard for compari-
son, the latest PEDCo surveyv28) Of utility FGD systems in the U.S. indicates a
total of 59,469 MW representing 139 units of operating and planned FGD systems.
Of these, 14,480 MW representing 40 units are operational.
895
-------
00
vO
Table 1-1
SUMMARY OF SIGNIFICANT OPERATING AND PLANNED
FULL-SCALE DUAL ALKALI SYSTEMS IN THE U.&H3, 28, 31)
PROCESS DEVELOPER
Food Machinery
Corporation (FMC)
General Motors Corp.
Zurn Industries
FMC
Combustion Equipment
Associates (CEA)/
ABSORBENT,
PRECIPITANT
Na,SO,, Ca(OH),
£. O £.
NaOH/Na,SO,,
Ca(OH)2 J
NaOH/Na,SO.,,
Ca(OH)zd 3
NazS03, Ca(OH}2
Na,SO,, Ca(OH),
C 3 £
ACTIVE ALKALI
Concentrated
USER
FMC
PLANT SITE HW
TYPE OF PLANT(q'
Modesto, California 10 (Gas Rate) Reduction ki1n'R'
YEAR OF , )
COMPLETION1 '
1971
30 (Regen.)
Dilute
Dilute
Concentrated
Concentrated
General Motors
Corporation
Caterpillar
Tractor Co.
Firestone Tire
& Rubber Co.
Gulf Power Co.
Parma, Ohio 32 (Gas Rate) Industrial boiler'"'
40 (Regen.)
Joliet, Illinois 20-30
Pottstown, Pa. 3
Sneads, Florida 20
Industrial boiler'"'
Demonstration^'
Utility boiler^
1974
1974
1975
1975
Arthur D. Little (ADL)
FMC
Zurn Industries
FMC
FMC
Envirotech
CEA/ ADL
FMC
FMC
RIC
(b)
(b)
Na.SO,, Ca(OH)2
NaOH/Na,SO,,
CafOH}^ J
Na2S03, Ca(OH)2
Ha.SO,, Ca(OH)2
Na2S03, Ca(OH)2
NaoSO,, Carbide
Lime J
Na,SO,, Ca(OH),
to f.
Na?S03, Ca(OH)2
Na2SO,, Ca(OH)2
n.a.
n.a.
Concentrated
Dilute
Concentrated
Concentrated
Concentrated
Concentrated
Concentrated
Concentrated
Concentrated
n.a.
n.a.
Caterpillar
Tractor Co.
Caterpi 1 1 ar
Tractor Co.
Caterpillar
Tractor Co.
Caterpillar
Tractor Co.
Central Illi-
nois Public
Service Co.
Louisville Gas
& Electric Co.
Southern Indi-
ana Gas & Elec-
tric Co.
Arco/Polymers,
Inc.
Chanslor West-
ern Oil & Dev.
Co.
Dupont Inc.
Grissom Air
Force Base
Mossville, 111. 50
Morton, Illinois 12
East Peoria, 111. 100
Hapleton, 111. 140
Newton, Illinois 575
Louisville, KY 277
Evansville, Ind. 250
Monaca, Pa. 100
Bakersfield, Ca 25
Georgia, Al. 100
Bunker Hill, Ind, 12
2 Industrial boilers(1N>1R) 1975
Industrial boiler^
1978
4 Industrial boilers(2N>2R) 1978
Industrial boiler^'
Utility boiler'"'
Utility boiler'1*'
Utility boiler"1'
Industrial boiler'R'
Industrial boiler"1'
Industrial boiler"1'
(1979)
(1979)
(1979)
(1979)
(1980)
(1979)
(1987)
3 Industrial boilers(1N>2R' (1979)
a) Dates in parentheses are projected start-up dates.
b) Vendor not selected.
c) N = New; R = Retrofit
-------
Table 1-2
SUMMARY OF SIGNIFICANT OPERATING
FULL-SCALE DUAL ALKALI SYSTEMS IN JAPAN (1)
00
VO
Process Absorbei
Developer Percipi
Showa Denko Na2S03,
Showa Denko-Ebara
it,
tant User
CaC03 Showa Denko
Kanegafuchi
Showa Pet. Chem.
Nippon Mining
Yokohama Rubber
Nisshin Oil
Poly Plastics
Ajinomoto
Kyowa Pet. Chem.
Japan Food
Yokohama Rubber
Asia Oil
Nippon Kokan (NH4)2S03, CaO Nippon
Tsukishima Na2$03, CaO Kinuura Utility
Kurabo Eng. (NH4)2S
Dai showa Paper
D4, CaO Kuraray
Dai eel
Bridgestone Tire
Bridgestone Tire
Jujo Paper
Dowa Mining Al2(S04)3, CaCOs Taenaka Mining
Dowa Mining
Naikai Engyo
Yahagi Iron
Nihon Seiko
Kowa Seiko
Mitsubishi Metal
Kureha Chemical CHsCOONa Kureha Chemical
Kobe Steel CaCl2,CaO Kobe Steel
Kobe Steel
Nakayama Steel
Kobe Steel
Kawasaki H.I. MgO, CaC03 Unitika
MgO, CaO . Nippon Exlan
Kureha-Kawasaki Na2S03>
CaC03 Tohoku Electric
Shikoku Electric
Shikoku Electric
Kyushu Electric
Tohoku Electric
Plant site
Chiba
Takasago
Kawasaki
Saganoseki
Hiratsuka
Isogo
Fuji
Yokkaichi
Yokkaichi
Yokkaichi
Hie
Yokohama
Keihin
Nagoya
Fuji
Tamashima
Aboshi
Tosu
Tochigi
Ishinomaki
Mobara
Okayama
Okayama
Nagoya
Nakase
Tobata
Nishiki
Amagasaki
Kobe
Osaka
Kakogawa
Okazaki
Saidaiji
Shinsendai
Sakaide
Anan
Buzen
Akita
Capacity
(1,000
Nm3/hr) Mil
580 178
300 107
200 71
120
105 38
100 36
212 76
82 29
150 54
100 36
100 36
243 87
150
185 66
264 94
100 36
163 58
60 21
80 28
200 71
150 x 2
70 25
50
30
72
140
5 2
175 x 2
350 x 2
375
1,000
200 71
300 107
420 150
1,260 450
1,260 450
730 261
1,050 375
Source of Gas
Industrial boiler
Industrial boiler
Industrial boiler
H2S04 plant
Industrial boiler
Industrial boiler
Industrial boiler
Industrial boiler
Industrial boiler
Industrial boiler
Industrial boiler
Industrial boiler
Sintering plant
Industrial boiler
Industrial boiler
Industrial boiler
Industrial boiler
Industrial boiler
Industrial boiler
Industrial boiler
Kiln
H2S04 plant
Industrial boiler
Sintering plant
Sintering plant
H2S04 plant
Smelting furnace
Industrial boiler
Sintering plant
Sintering plant
Sintering plant
Sintering plant
Industrial boiler
Industrial boiler
Utility boiler
Utility boiler
Utility boiler
Utility boiler
Utility boiler
Inlet
S02 (ppm)
1,500
1,500
1,400
400
1,500
1,300
1,200
7,500
650
1,500
2,500
5,000
750
4,000
1,500
500
500
500
500
1,600
1,400
420
1,500
1,500
1,500
1300
Year of
Completion
1973
1974
1974
1973
1974
1974
1974
1974
1974
1975
1975
1975
1972
1974
1975
1974
1975
1975
1975
1976
1972
1974
1976
1976
1976
1978
1978
1975
1976
1976
1976
1978
1975
1975
1974
1975
19.75
1977
1977
-------
Section 2
DESIGN CRITERIA AND CURRENTLY OFFERED PROCESS GUARANTEES
Based on the performance and reliability demonstrated in various dual alkali
pilot and prototype plants in the U.S. and Japan, and in an increasing number
of full-scale applications in Japan, several general design and performance
criteria are of interest. Principal among them are: (1) S02 removal performance,
(2) particulate matter removal performance, (3) sodium consumption, (4) calcium
consumption, (5) energy consumption, (6) waste solids quality, and (7) system
reliability.
2.1 DESIGN CRITERIA
SQ2 Removal
A commercial dual alkali system must remove the desired quantity of S02 to allow
compliance with the applicable Federal and/or local standards. Uith the present
state-of-the-art, it is reasonable to expect long-term average S02 removal capa-
bility on the order of 95% with moderate, 10-15 gal./103acf, liquid-to-gas ratios,
and high (up to 4,000 ppm) S02 inlet concentration.
Particulate Matter Removal
The scrubbed gas from the dual alkali flue gas desulfurization (FGD) unit should
not contain particulate matter in excess of applicable standards. In some cases,
this could require particulate matter removal by the FGD unit; in others it could
merely imply no net addition of particulate matter to the gas stream by the FGD
unit.
Sodium Consumption
Sodium consumption is an important performance criterion for a dual alkali
system not so much from the viewpoint of economics but rather from its potential
for secondary pollution. Sodium consumption is only a minor factor in the
operating cost of a system, representing about 2% of the annual operating cost.
Thus, even if soda makeup to a system were to increase by 100% over its design
value, operating cost will be increased by a factor of less than 1.02. On the
other hand, the environmental consequences of higher sodium consumption,may be
significant if the sodium is leached from the waste product to the environment.
A logical way to measure sodium consumption is in moles of Na consumed per mole
of sulfur removed by the system. A value of 0.05 moles of Na makeup per mole
of sulfur removed (equivalent to 0.025 moles of Na2C03 make-up per mole of sul-
fur removed) appears to be a reasonable design target based on present U.S. tech-
nology. This target is achievable in concentrated dual alkali systems burning
relatively high sulfur coal (over .3% sulfur) and having a relatively low oxygen
content in flue gas. In Japan, sodium makeup is reportedly as low as 0.02 moles
Na/mole sulfur removed for some systems.(13)
898
-------
Calcium Consumption
A logical way to specify calcium consumption is as calcium stoichiornetry; i.e.,
moles of calcium added per mole of sulfur removed (or collected). A calcium con-
sumption of 0.98 to 1.0 appears to be a reasonable design target for concentrated
dual alkali systems (values less than 1.0 are possible due to the alkali added
with sodium mak~eup to the system).
Energy Consumption
Design targets for energy consumption can be in the range of 1-2% of power
plant output without reheat, or under 4% even with 50 F° of reheat. These
figures are based on a system which has a scrubber pressure drop in the range
of 6-8 in. H20 and a scrubber system L/G ratio of about 10 gal./103acf. This
assumes that the power plant is equipped with some means of efficient parti-
culate collection upstream of the FGD unit (e.g., an electrostatic precipitator).
Waste Solids Quality
The dual alkali process must produce easily handled, transportable, environmen-
tally acceptable waste. Present state of technology indicates that a filter cake
containing a minimum of 65 wt. % solids and a low level of soluble salts, such
as Na2S03 and NaCl, is a reasonable expectation. Systems currently being built in
the U.S. (concentrated mode) will produce a waste solid containing primarily cal-
cium sulfite with some amount of sulfate in the crystal. The Japanese systems
are primarily designed to produce saleable gypsum.
System Reliability
The FGD system should have a high availability. Based on existing vendor guar-
antees availability of 90% for I year, as defined by the Edison Electric Insti-
tute for power plant equipment, may be a reasonable target.
2.2 CURRENTLY OFFERED PROCESS GUARANTEES
As examples of the currently offered process guarantees, those for the three,
full-scale, utility dual alkali systems in the U.S. are summarized in Table 2-1.
Because of site-specific factors, the reader is cautioned against any interfacil-
ity comparison or evaluation in light of the above-noted design criteria.
899
-------
Table 2-1
FGD SYSTEM GUARANTEES (3)(22)(32)
o
o
Availability
S02 Removal
Particulate Matter
Removal lb/106 Btu
HC1 Removal
Sodium Consumption
moles Na^CO-/mole SO-
Calcium Consumption
moles Ca/mole SCL re
Energy Consumption
Solids Quality
Vendor: Envirotech
LG&E
Vendor: CEA/ADL
SIEGECO
Vendor: FMC
90% for 70% load factor for
30 year life span
90% or outlet SO- less than
200 ppm whichever is greater
(b)
<0.10
0.024 (per mole SOg) and
0.023 (per mole of acid
gas)
1.10
None
None
90% for a 1-year operating
period
200 ppm in scrubber outlet
or 95% removal if sulfur
content of the coal is
greater than 5 wt.%
<0.10 and no net addition
of particulate
Not applicable
0.045 when maximum coal
chloride level is 0.06%
0.5 moles additional for
each mole of chloride in
the coal above the 0.06%
level
Maximum 1.05
1.2% of peak operating
rate (300 MW)
Minimum 55 wt.% insoluble
solids
95% for 30- and 60-day
consecutive test runs
1.2 lb/106 Btu (85%
for 4.5 wt.% sulfur coaV
< 0.10 and no net addi-
tion of particulate
Not applicable ;
0.03 plus
Na lost with
. chloride
Approximately 1.0
Less than 1% of
operating rate
Minimum 55 wt.% solids
a) Figures are not guarantees but rather expected design values.
b) FGD system designed for possible ESP upsets.
-------
Section 3
FULL-SCALE UTILITY DUAL ALKALI SYSTEMS IN THE U.S.
This year (1979) three U.S. full-scale, utility boiler dual alkali systems,
ranging in size from 250 to 575 MW, are scheduled for startup. The owners of
these systems, their size, and the vendors providing them are: (1) Central
Illinois Public Service Co., Newton No. 1, 575 MW unit, Envirotech; (2) Louis-
ville Gas & Electric Co., Cane Run No. 6, 277 MW unit, CEA/ADL; and (3) Southern
Indiana Gas & Electric Co., A.B. Brown No. 1, 250 MW unit, FMC. All of these
systems are of the concentrated sodium based dual alkali type and they service
boilers firing high sulfur coal.
Brief descriptions of the these full-scale dual alkali facilities, their design
bases, and reported costs are presented in this section.
3.1 FACILITY DESCRIPTIONS
Central Illinois Public Service Co.
Boiler
Newton Station Unit No. 1, Newton, Illinois, utilizes a pulverized coal-fired
boiler manufactured by Combustion Engineering Inc. It is designed for a maximum
steam generating capacity of 4,158,619 Ib/hr at 2620 psig and 1005°F at the steam
outlet connection. Design excess air in the boiler is 22% with flue gas flow
to the scrubbing system totalling 6,615,000 Ib/hr. The boiler design includes
two parallel induced draft fans with discharge into a common plenum which can
feed flue gas to the FGD system or to the stack. The boiler is tangentially
fired with coal from bowl-type pulverizing mills utilizing high-sulfur Illinois
bituminous coals from several local sources.
FGD Facility
Process equipment of interest, in the "duplex-type" dual alkali FGD system,
built to serve Newton No. 1, includes:
• Four booster fans.
• Four precooler spray towers, including mist eliminators.
• Four mobile-ball gas scrubbers, including mist eliminators.
t Three spent absorbent causticizers (reactors).
• One precooler effluent neutral izer.
901
-------
• One reactor/clarifier.
• Two 100-ft diameter thickeners, for the dual alkali solids concentra-
tion system, with concrete bottoms and access to bottom discharge cone
through a tunnel.
• One 50-ft diameter thickener, for the precooler loop, constructed of
coated steel.
• Three horizontal Eimco-Extractor filters for dewatering and multi-stage
washing of thickener underflow.
Figures 3-1 and 3-2 are the process flow diagram and a general view of the FGD
system, respectively. An overview of the absorption, regeneration, and dewatering
sections is given below.
Absorption Section
Precooler/HCI Absorber: Because of design criteria requiring the use of
local Illinois coal, the system is designed to accommodate these high-chloride
coals. The presence of chlorides in the coal may require a preconditioning step
before the flue gas is treated for S02 removal in the main scrubbing loop.
This preconditioning of the flue gas in the precooler spray towers also mini-
mizes soda ash consumption.
The precooler spray towers treat and cool the flue gas. The towers are de-
signed to remove 90% of the chloride in the form of HC1. To minimize any
carryover of mist or particulate matter, vertical mist eliminators are installed
in the horizontal precooler tower outlet ducts (liquid drains perpendicular to
the gas flow). These mist eliminators control the acid mist within the precooler
loop and the level of dissolved solids (chlorides) in the scrubbing liquor to re-
duce sodium consumption and minimize operational problems.
The spent liquor from the precooler is sent to the fly ash thickener for clari-
fication and prevention of significant concentrations of abrasive slurry solids
in the recirculating liquor. Also, to maintain steady sodium level, pH, and
solids level in the precooler loop, the clarifier underflow is dewatered and
treated with lime and sodium carbonate. The regenerated liquor is returned to
the precooler loop.
S02 Scrubber: S02 emissions are reduced to less than 200 ppm utilizing a
countercurrent, two-tray-stage mobile-ball bed scrubber operating at 8.3 ft/sec
design gas velocity with a L/G of less than 10 gal./lO^acf). A mist eliminator
mounted vertically in the horizontal duct (liquid drains perpendicular to the
gas flow) controls sodium liquor carryover. The eliminator vanes are intermit-
tently washed with service water, which is collected and returned to the precooler
loop to maintain a favorable water balance.
In the initial operation of the unit, two types of reheat systems are used (half
the gas volume for each):
902
-------
O
oo
1 BOOSTER FAN
2 CHLORIDE/FLYASH PRECOOLER
3 FLYASH THICKENER
4 ABSORBER
5 NEUTRALIZER
6 REACTOR-CLARIFIER
0) | 7 CAUSTICIZERS
8 THICKENER
9 HORIZONTAL BELT FILTER
10 REHEAT FACILITY (EXPERIMENTAL)
Figure 3-1 ENVIROTECH FLUE GAS DESULFURIZATION SYSTEM ON CIPSCO'S NEWTON NO. 1 UNIT
-------
1C
o
Figure 3-2 GENERAL VIEW OF CIPSCO NEWTON NO.1 FGD UNIT
-------
1. Recycle Reheat
2. Bypass Reheat
With the recycle reheat mode of operation, the reheat is accomplished by
blending a side stream of steam-heated flue gas with the saturated flue gas
exiting the scrubber trains. The blending occurs across a high-alloy perfor-
ated-plate diffuser section in the main duct. The steam heated flue gas is
maintained above the saturation point during normal operation by recycling and
preheating gas on the downstream side of the diffuser. The gas entering the
reheater will be above the saturation point to reduce corrosion potential. The
gas in the recycle reheat section during normal operation will range between
156°F and 300°F, as compared to the saturated gas temperature of 131°F. The
system is designed to achieve an exit gas temperature of 156°F to the stack.
In the Bypass Reheat method, a portion of the unscrubbed flue gas is bypassed
around the scrubber and mixed with the cleaned, saturated, scrubbed gases exit-
ing the scrubber. A diffuser section, as described above, is used to disperse
and mix the two gas streams. This system is designed to achieve a maximum of
43°F reheat because of the limitation on allowed S02 and particulate emission
levels. The quantity of bypass gas is controlled by damper control to realize
the desired degree of reheat. In addition, the gas is also monitored in regard
to opacity and S02 level as an additional means of control.
The overall design philosophy in regard to the reheat systems was to provide
sufficient flexibility to allow either type of reheat system to be installed
later if one of the systems proved to be more effective. Space has been re-
served to achieve up to 50 F° of reheat, based on space requirements for the
larger system.
Regeneration Section
A portion of the spent liquor from the absorber is pumped to the first stage
of a three-stage causticizer (reactor) system, where slaked lime slurry is
mixed with spent liquor.
The primary causticizer overflows by gravity into the secondary causticizer
which in turn overflows into the tertiary causticizer. The slurry from the
tertiary causticizer is pumped to the dewatering section. It is not certain
whether one, two, or three reactor stages will be operated. If one is needed,
there will be two spares. Likewise, if two are needed one will act as a spare.
Dewatering Section
After thickening of the sulfur oxides solids collected in the mobile-ball scrub-
ber, the solids are dewatered and fresh-water washed in a top-loading horizontal-
belt-type extractor filter. The unique geometry of this equipment permits counter-
current multi-stage washing of the raw cake with limited quantities of service
water. Over 80% of the entrained sodium in the vacuum filter cake is recovered.
This sodium is recycled to the absorption loop for reuse. In addition, the liquid
purge stream from the precooler loop, containing collected chlorides, residual fly
905
-------
ash, and trace elements, flows to a lime-neutralization tank and then is uti-
lized as the wash medium for a final "cake-impregnation wash." Thus, high-chlo-
ride, low-pH liquor bypasses the S02 absorber and is discarded as surface moisture
in the final waste cake.
Louisville Gas & Electric Co. (LG&E)
boiler
Cane Run Station Unit No. 6, Louisville, Kentucky-, utilizes a pulverized coal-
fired boiler manufactured by Combustion Engineering Inc. It is designed for a
maximum steam generating capacity of 1,854,217 Ib/hr at 2600 psig and 1005°F at
the steam outlet connection. Design flue gas flow to the scrubbing system totals
3,372,000 Ib/hr. The boiler is tangentially fired with coal from bowl-type pulver-
izing mills utilizing, high-sulfur, midwest bituminous coals.
FGD Facility
Process equipment of interest, in the dual alkali FGD system for Cane Run No. 6,
includes:
• Two booster fans.
• Two dual-tray absorbers.
• Two pairs of spent absorbent regeneration reactors.
• One 125-ft diameter thickener with concrete bottom and access to bottom
discharge cone through a tunnel.
• Three rotary-drum vacuum filters with water wash.
• Two external combustion, scrubbed gas reheaters.
Figure 3-3 is a process flow diagram of the FGD System. Figures 3-4, 3-5 and 3-6
are photographs showing a general view of Cane Run Nos. 4, 5 and 6 FGD units,
Unit 6 regeneration and dewatering areas, and Unit 6 vacuum filter, respectively.
An overview of the absorption, regeneration, and dewatering sections is given
below.
Absorption Section
The flue gas from the existing electrostatic precipitator (ESP) induced draft fan
is forced by a booster fan into an absorber. There are two absorber modules,
each equipped with a booster fan. A common duct connects the two inlet ducts to
the booster fans.
The hot flue gas is adiabatically cooled and saturated by sprays of absorber solu-
tion directed at the underside of the bottom tray. These sprays keep the under-
side of the tray and the bottom of the absorber free of buildup of fly ash solids.
The cooled gas then passes through two sieve trays, where S02 is removed, and
906
-------
STACK
vo
O
FLUE GAS
^
S3
FORCED DRAFT FAN
SPRAY/TRAY ABSORBER
MIST ELIMINATOR
PRIMARY REACTOR
SECONDARY REACTOR
THICKENER
FEED FORWARD TANK
VACUUM FILTER
FILTRATE SUMP
10 REHEAT FACILITY
11 MIX PLANT
r
WASH WATER
CaS03/CaS04
SOLIDS
Figure 3-3 CEA/ADL FLUE GAS DESULFURIZATION SYSTEM ON LG & E'S CANE RUN NO.6 UNIT
-------
o
:
Figure 3-4 GENERAL VIEW OF LG&E CANE RUN NOS. 4, 5, AND 6 FGD UNITS
-------
k
Figure 3-5 CANE RUN NO. 6 FGD UNIT REGENERATION AND DEWATERING AREAS
-------
Figure 3-6 CANE RUN NO. 6 FGD UNIT VACUUM FILTER
-------
leaves the absorber through a chevron type mist eliminator. Prior to entering
the stack, the saturated gas from the mist eliminator is heated 50°F, to 175°F,
by hot combustion gas from a grade-mounted reheater fired with No. 2 oil.
Each absorber is designed for 9.0 ft/sec gas velocity, a rate consistent with
good mass transfer, low pressure drop, and minimal entrainment. Each absorber
is sized to handle 60% of the design gas flow rate and the overall system can
be turned down to 20% of the design flow rate by shutting down one absorber.
At levels less than 50% of design capacity, the system can be operated with
one absorber module by use of the common inlet duct.
For control of tray feed liquor pH, regenerated scrubbing liquor from the thick-
ener hold tank is mixed in-line with absorber recycle liquor for each unit; the
mixture is then fed to the top tray in each absorber. The absorber recycle liquor
is used in the spray section below the trays. A bleedstream of the absorber re-
cycle liquor is withdrawn and sent to the reactor system for regeneration. The
bleed rate is controlled by the liquid level in the absorber. The feed forward
rate of regenerated liquor from the thickener hold tank to the scrubber trays
corresponds to an L/G of 4.0 gal./103acf (saturated) at design conditions. The
total recirculation rate for each absorber (sprays plus trays but excluding the
feed forward regenerated liquor) corresponds to an L/G of 5.7 gal./103acf. Thus,
the total L/G is about 10 gal./103acf.
Regeneration Section
The spent liquor (absorber bleed) is fed to the primary reactor of the two-stage
reactor system along with slurried carbide lime. The primary reactor has a nomina"
liquor holdup time of 4.5 minutes at design flow. The primary reactor overflows
by gravity into the secondary reactor, which has a nominal holdup of 40 minutes
at design flow, where the reaction between lime and sodium salts is completed.
The reaction product is a slurry containing 2-4% insoluble calcium salts and
the regenerated sodium salt solution. The reactor slurry is pumped to the solid/
liquid separation section. The pumping rate is controlled by the liquid level
in the reactor.
Two reactor trains are provided, each train consisting of a primary reactor,
a secondary reactor, and a reactor pump. At design conditions, both of the
reactor trains would normally be in service, with each reactor train regenerating
the spent liquor from the corresponding absorber. The reactor trains are iden-
tical, and each can be operated on liquor from either absorber or combined liquor
from both the absorbers. For short-term, a reactor train may handle the total
liquor from both the absorbers operating at design conditions. Thus, maintenance
can be performed on one reactor train while the other is operating.
Dewatering Section
The reactor effluent streams are fed to the thickener. Clarified liquor over-
flows to the thickener hold tank from which the regenerated solution is pumped
automatically to the absorbers to maintain the pH of the absorber liquor. The
total volume in the system is maintained by controlling the liquid level in the
thickener hold tank using process makeup water.
911
-------
The thickener underflow slurry, controlled at about 25 wt.% solids, is pumped
to the filter system where solids separation is completed. The filter cake is
washed with fresh water to recover the sodium salts in the liquor. Combined
filtrate and wash water are returned to the thickener.
There are three rotary drum vacuum filters, each rated to handle 50% of the total
solids produced at the design conditions. Each filter can be operated independent-
ly. For optimum performance (to obtain cake containing high solids content and low
soluble salts content) it is desirable to operate the filters at fixed conditions
(constant drum speed, submergence, wash ratio, etc.). Therefore, the cake rate is
controlled by changing the number of filters in operation. The number of filters
in operation is determined by the amount of solids accumulated in the thicken-
er, which is reflected in the solids concentration in the underflow slurry. The
density of the underflow slurry is measured and thickener hold tank liquor is
added as required to maintain the percent solids in the underflow slurry at
about 25%. The number of filters in operation is changed if the concentration
of solids in the underflow slurry cannot be controlled using the dilution liquor.
Southern Indiana Gas & Electric Co. (SIGECO)
(22)
FGD Facility
Process equipment items of interest, in dual alkali system for Unit 1, include:
• Two booster fans.
t Two, three-stage disc contactors.
t One lime reactor.
• One, 100-ft diameter, thickener for dual alkali solids concentration.
• Three, rotary-drum, vacuum filters with water wash.
Figures 3-7 and 3-8 are the process flow diagram and a view of the FGD system
absorber, respectively. A description of the absorption, regeneration, and de-
watering sections is given below.
Absorption Sectioji
The FGD system receives flue gas from two points on the discharge side of the
electrostatic precipitator. The flue gas is drawn into two fans to increase
the static pressure to the level required to force the gas through the absorbers
and to the stack.
The S02 absorbers are low pressure drop, three-stage, counter-current disc con-
tactors. Liquid/gas contacting is accomplished with fixed position flow diverters.
The units are 30 ft-5 in. in diameter and have an overall height to the discharge
of 70 ft. The top of the scrubber contains a chevron type mist eliminator to pre-
vent liquid entrainment. This mist eliminator is washed approximately once per
shift to remove any accumulation of salts and/or particulate matter.
912
-------
STACK
VO
I-"
co
(MAKE-UPI
1 BOOSTER FAN
2 DISC CONTACTOR
3 REACTOR
4 SURGE TANK
5 THICKENER
6 DRUM FILTER
Figure 3-7 FMC FLUE GAS DESULFURIZATION SYSTEM ON SIGECO'S A.B. BROWN NO.1 UNIT
-------
Figure 3-8
VIEW OF SIGECO A.B. BROWN NO. 1 FGD UNIT ABSORBER
914
-------
Four manways are provided on each scrubber to give access to the mist eliminator,
scrubber internals, and liquid sump. The bottom of the vessel is used as a reser-
voir for the scrubbing solution. During normal operation, the flue gas provides
adequate heat to prevent freezing of the scrubbing solution. During system shut-
down and subsequent startup, a steam sparger is provided to heat the solution.
The bottom 8 ft of the contactor is insulated with fiberglass and covered with an
aluminum jacket.
Two rubber-lined recirculation pumps circulate scrubbing liquor to the top stage
of each scrubber. Each pump is capable of supplying 5200 gal./min. of solution.
One pump is the primary operating pump; the second pump provides 100% spare capa-
city. The pumps are piped into the system with check valves to permit remote se-
lection and startup without manual valve changes. A crossover starting system
automatically starts the spare pump when flow fails. The flow of recirculated
solution, not modulated for scrubbing load, is set for maximum gas flow.
Regeneration Section
Scrubbing solution to be regenerated is pumped to the lime reactor where it is
neutralized with the lime slurry. The pH of the lime reactor is controlled by
two separate pH control loops, one in operation and one spare.
The system has high and low pH alarms. A selector switch allows the operator to
convert from one pH analyzer element to the other. The pH analyzer probes in this
service are cleaned periodically. The regenerated slurry overflows from the lime
reactor to the thickener where the calcium sulfite solids settle out of the regen-
erated solution. The rate of soda ash solution makeup to the scrubber modules
controls the regenerated solution density. Density is an indication of the total
solution concentration which must be maintained for optimum operation of the dual
alkali process. The thickener underflow is pumped by diaphragm pumps to three ro-
tary drum vacuum filters. The slurry flow rate to each filter is controlled by a
low level switch. The regenerated scrubbing solution is collected in the regenera-
tion surge tank and pumped back to the scrubbers based on pH demand.
Liquid level in the regeneration surge tank is controlled by addition of makeup
water. The temperature of the surge tank is monitored and, in the event of low
temperature, the tank is heated with steam supplied through a sparger.
Dewatering System
The underflow from the thickener is pumped to three rotary drum vacuum filters
where the calcium sulfite precipitate is separated from the sodium sulfite solu-
tion. Cake removal is facilitated by low pressure air blown through the filter
cloth just before the scrapers at the discharge. The filtrate is drawn into
a filtrate receiver and is pumped back to the thickener. The vacuum is pro-
duced by a mechanical vacuum pump using water flushing. The water for this
service is used only once.
915
-------
3.2 DESIGN BASES
The design bases and design features for the three, utility dual alkali FGD sys-
tem are presented in Tables 3-1 and 3-2, respectively. Typical analyses of the
coals fired are summarized in Table 3-3.
Information in these tables leads to several general observations:
• All installations service boilers firing high-sulfur coal. .Design coal
sulfur range is 4.0 to 5.0 wt.%. Inlet S02 entering the absorbers
ranges between 2600 and 3500 ppm.
• Design coal chloride content varies widely from 0.04 to 0.20 wt.%.
• Excess air including preheater leakage varies considerably as indicated
by the 3.2 to 7.6% flue gas 03 content range.
• Flue gas flow rate per scrubber module ranges between 395,000 and
522,000 acfm at 300°F.
• The number of regeneration reactors varies between designs, indicating
differing design philosophies.
• All facilities use primary (thickener) and secondary (vacuum filters)
dewatering devices. Sodium is recovered by wash schemes ranging from
single- to multi-stage.
• Planned ultimate disposal of filter cake is to be accomplished, in all
cases, by landfill. At the Newton No. 1 and Cane Run No. 6 plants,
stabilization efforts may range from simple mixing with fly ash to mixing
with fly ash and lime.
3.3 REPORTED COSTS
Before presenting the reported costs, a word of caution is in order. Because of
many site-specific factors and/or different bases, quantitative evaluation of
capital and operating costs for the various dual alkali processes for comparison
with each other and with lime/limestone scrubbing is inadvisable.
Site-specific parameters that influence the capital cost of FGD systems include
location and size of plant, new versus retrofit application, presence or absence
of upstream fly ash removal facilities, fuel sulfur content, fuel chlorine con-
tent, degree Of spare capacity, solid waste disposal method, and cost of real
estate. The different cost factors will not apply equally to all processes. Each
application' requires separate assessment at conditions specified for a given in-
stallation to evaluate the cost ranking of the proposed dual alkali systems.
.ikewise, for operating costs, the impact of raw material (e.g., quick lime @ $40/
on vs. carbide lime @ $13.29/ton)s utilities, and maintenance costs varies accord-
Li I
ton
ing to the type of process selected and applicable site-sensitive parameters. In
addition, financial parameters, such as taxation and capital charges, vary with
the location and accounting practice of the utility.
916
-------
Table 3-1
FGD SYSTEM DESIGN BASES <3M22)(32)
CIPSCO
NEWTON NO. 1
LG&E
CANE RUN NO. 6
SIEGCO
A.B. BROWN NO. 1
Coal (Dry Basis):
Sulfur, vt.%
Chloride, wt.X
Heat Content, Btu/lb
4.0
0.20
10.900
5.0
0.04
11,000
4.5
0.05
13,010
Inlet Gas:
Flow Rate (volumetric), acfm
(weight), Ib/hr
Temperature, °F
°2>
Particulate, lb/106 Btu
6,615,000
327
2590(a)
7.7
not available
1,065,000
3,372,000
300
3471
5.7
0.10
790,036
2,415,764
290
281
3.2(-
0.10
Outlet Gas:
S02, ppm
Particulate, lb/106 Btu
200
o.io(b)
200
0.10
520
0.10
(a.) Calculated from other data.
(b) System designed to accommodate ESP upsets.
-------
Tabto3-2
FGD SYSTEM DESIGN FEATURES (3) <22)(32)
CIPSCO
VENDOR: ENVIROTECH
LG&E
VENDOR: CEA/ADL
SIGECO
VENDOR: FMC
oo
FGD Unit Rating, MW
Process
Fly Ash Collection
No. of Modules
Module Design
Regeneration Reactor
Reactor Residence Time
Dewateri ng
Filter Cake Disposal
Additional Equipment
575
Sodium/Calcium Hydroxide
Dual Alkali/Concentrated
ESP
Four
High velocity cocurrent spray
tower„ countercurrent two-
tray-stage polysphere absorber
Three reactors in series for
four modules ^aJ
No information
Three horizontal belt filters
Fly ash stabilization/onsite
landfill
Chloride/fly ash precooler;
fly ash thickener; neutralizer;
•reactor clarifier; and two ver-
sions of experimental reheat
facilities
277
Sodium/Carbide Lime
Dual Alkali/Concentrated
ESP
Two
Two-stage tray tower
absorber
Two reactors in series
per module
Primary - 4.5 minutes
Secondary - 40 minutes
Three rotary drum vacuum
filters
Fly ash stabilization/onsite
landfill
250
Sodium/Calcium Hydroxide
Dual Alkali/Concentrated
ESP
Two
Three-stage disc contactor
One reactor per two modules
No information
Three rotary drum vacuum
fi1ters
Transported to landfill
without stabilization
Scrubbed gas reheat facility; None
mix plant for sludge stabili-
zation
(a) In actual operation, one, two or three reactors may be used.
-------
Table 3-3
"TYPICAL COAL ANALYSES (3) (22) (32)
(Dry fiatls)
Species
ULTIMATE
Carbon
Hydrogen
Nitrogen
Chlorine
Sulfur
Ash.
Oxygen
PROXIMATE
Moisture
CIPSCO
NEWTON NO. 1
Value, wt.%
66.70
4.92
1.68
0.22
4.47
14.19
7.82
100.00
10.50
Heat Content, Btu/lb 10,900
LG&E
CANE RUN NO. 6
Value, wt.%
67.15
4.72
1.28
0.04
4.81
17.06
4.94
100.00
8.95
11,000
SIGECO
A.B. B.ROWN NO. 1
Value, wt.%
72.19
5.01
1.56
0.05
3.97
9.88
7,34
100.00
11.34
13,010
919
-------
Actual capital and operating costs for full-scale, utility, dual alkali sys-
tems in the U.S. are as yet not available; construction of the first three,
commercial-size systems is scheduled for completion in 1979 (see Table 1-1).
Estimated costs for these applications are therefore based only on costs for
prototype and small-scale commercial units, or on vendor and engineering esti-
mates. Published cost data for the small-scale systems are fragmentary and have
not been determined on equivalent bases. Generally, estimates by vendors tend
to be low since they reflect vendor supplied portions of the project; a signifi-
cant number of owner-incurred cost items are commonly omitted. Engineering
estimates are usually based on specific applications and cannot be readily
extrapolated for different conditions.
However, for informational purposes only, the capital and operating cost data
are included herein. Tables 3-4 and 3-5 summarize those for the LG&E instal-
lation; Table 3-6 contains those for the SIGECO facility.
920
-------
TaMt3-4
ESTIMATED CAPITAL COSTS FOR THE DUAL ALKALI
FGD SYSTEM ON LG&E CANE RUN NO. 6 UNIT (32)
Dollars
Total Materials Costs:
7,037,000
Major Equipment Cost 2,525,000
Other Materials Cost3 900,000
Sludge Disposal Equipment 700.000
Additive Supply System° 11,162,000
Total Materials Cost
^Erection: d
Direct Labor (252,800 hrs. @ $12/hr) c 3,034,000-
Field Supervision 273.000
Total Erection Cost 3,307,000
Engineering Costs:
System Supplier Engineering 1,323,000
Owner's Engineering Expenses 303,000
Owner's Consulting Engineer 852.OQQ
Total Engineering Cost 2,478,000
Spare Parts, 2% of the Total Materials Cost 223,000
Working Capital 200.000
Total Capital Investment e 17,370,000
$/kW (Based on 300 MW gross peak load) 57.9
(Based on 277 MVJ gross net load) 62.7
Basis: 300 Mw gross peak load existing coal fired boiler (277 MW net peak load)
S in coal, 5.0%
S02 removal efficiency in scrubber, 94.2%
Stack gas reheat, 50% - direct oil fired reheater
Disposal of sludge after treatment to onsite pond
Project beginning mid 1976, ending late 1978, Avg. 1977 dollars
Necessary parts in storage and reasonable spare capacity
aSludge disposal equipment cost ($900,000) is shown separately. The $900,000
is the 3/7 portion of the total sludge disposal cost for Cane Run 4, 5 and 6.
bAdditive supply system cost ($700,000) is shown separately. The $700,000 is
the 3/7 portion of the total additive supply system for SOx removal systems
for Cane Run 4, 5 and 6.
cIncludes plant overhead.
dErection equipment cost is included in plant overhead.
eThe capital investment for the above system in 1976 dollars is equivalent to
$15.95 million or 53.15 $/kW (based on gross peak capacity).
921
-------
3-6
KM THE fJUAL ALKALI FOB fWBU AT LOME G»
IO
N>
Direct Costs:
Materials
Carbide Lime3
Soda Ash
Fuel Oil6
Electricity
Water
Sludge Removal
Maintenance Materials
Labor
Operation
Maintenance
Analysis
Supervision
Total Direct Costs
Indirect Costs:
Overhead
Interest
Depreciation
Total Indirect Costs
Total Annual Operating Cost, $
Hills/kWhr
4/106 Btu
$/Ton of S Removed
Quantrty Unit Cost, $
58,728 tons 13.29/ton
1,912 tons 100.00/ton
1,802,808 gals. 0.43/gal
16,188,480 kWhr 0.01/kUhr
126,100,000 gals. 0.05/1,000 gals.
185,280 tons 2.01/wet ton
2.5% of total materials cost
from Table 3-4.
26,280 hrs 8.18/hr
26,280 hrs 8.28/hr
2,080 hrs 10.00/hr
2,080 hrs Various
No Reheatb
59.4% of 493,368 (Total Labor)
6.125% of 17,379,000
4.17% of 17,379,000
Hith Reheat
5,142,600
3.27
32.8
219.7
Annual Cost $/yr
780,500
191,200
775,200
161,900
6,300
372,400
279,000
215,000
217,600
20,800
40,000
3,059,900
2,384,700
293,100
1,064,500
724,700
2,082,300
bNo Reheat
4,367,400
2.77
27.9
186.4
Basis: (a) If quick lime at $40/ton is used as would be the case in most installations, the annual operating cost would increase from
about $4.4 to $5.8 million. The corresponding increase in unit operating cost will be from 2.77 to 3.65 mills/kWhr.
(b) Operating cost for the no-reheat option was calculated without the cost of fuel oil.
-------
ESTIMATED AVERAGE ANNUAL OPERATING COSTS
FOR THE DUAL ALKALI FGDSYSTEM AT SIGECO
Unit Cost. $
Annual Cost,$/Year
Reagent Costs
Pebble Lime
Soda Ash
Water
$40/ton
$100/ton
$0.50/1,000 gals
1,210,000
162,000
41,000
Power Costs
Demand Charge
Energy Charge
$700/kW
$0.015/kWhr
210,000
158,000
Capital Charge (@15% of Capital Cost)
Maintenance
Operating Labor (2/shift)^
Disposal Charge $2.00/ton
1,500,000
180,000
240,000
232,000
TOTAL
3,933,000/year
3.0 Mills/kWhr
Basis: 265 MW (Gross peak capacity) existing coal-fired plant, 60%
load factor, 4 wt.% S in coal, 80 to 95% SO^ removal, on-site
disposal without treatment.
(a) For actual operation 3 operators per shift will be required.
(b) Total capital expended on unit estimated to be $12,500,000 or
$50/kW,
923
-------
Section 4
SUMMARY OF STATUS OF DUAL ALKALI TECHNOLOGY IN
4.1 GENERAL BACKGROUND
About 42 major dual alkali FGD plants, having a combined capacity of about
4,000 MW equivalent, were operational in Japan at the beginning of 1978.
Table 4-1 summarizes these systems by process developer, constructor, process
type, and capacity. Approximately 45% of the total capacity represents utility
boiler application (primarily oil-fired) while the remainder includes industrial
boilers, sintering plants, smelters, and sulfuric acid plants. A summary of
those applications has previously been presented in Table 1-2.
An interesting facet of emerging dual alkali technology in Japan is processes
having unlimited oxidation tolerance and utilizing limestone as a regenerant.
This is generally accomplished by circulating absorbents other than sodium sul-
fite. Examples include Dowa's basic aluminum sulfate process, Kawasaki's magne-
sium-gypsum process, and Kureha's sodium acetate-gypsum process. Some of these
processes do not use a clear solution as the absorbent. The current EPA funded
program at Gulf Power's Scholz Station is also presently investigating the
feasibility of limestone regeneration with sodium sulfite systems. A brief
discussion of the Dowa, Kawasaki, and Kureha processes follows.
4.2 DOWA BASIC ALUMINUM SULFATE FGD PROCESS^33)
This process has beeri developed by Dowa Mining Co. of Japan. In the U.S. this
process is licensed by Universal Oil Products (UOP) and is presently being tested
at the TVA Shawnee Test Facility to evaluate its applicability to coal-fired boilers.
Process Description
As shown in Figure 4-1 this process consists of three operations: absorption,
oxidation, and neutralization.
Absorption
S02 is absorbed in a clear solution of basic aluminum sulfate, Alo(S04)o.
A1203 of pH 3 to 4 to form A12(S04)3 .A12(S03)3;
A12(S04)3. A1203 + 3S02 = A12(S04)3. A12(S03)3 (1)
Oxidation
The aluminum sulfite in the spent liquor is oxidized by air to aluminum sulfate:
A12(S04)3 -A12(S03)3 + 3/2 02 = A12 (S04)3- A12(S04)3 (2)
924
-------
Table 4-1
NUMBERS AND CAPACITIES OF DUAL ALKALI FGD PLANTS
BY MAJOR PROCESS DEVELOPERS/CONSTRUCTORS IN JAPAN (1)
PROCESS DEVELOPER/CONSTRUCTOR
N
Dowa Engineering
Kawasaki Heavy Industries
Kobe Steel
Kurabo
Kureha
Kureha
Nippon
Engineering
Kawasaki
Chemical
Kokan (NKK)
Tsukishima Kikai (TSK)
Showa
Showa
Denko
Denko-Ebara
TOTAL
umber Capacity
MW
8 200
2 180
6 745
5 215
5 1685
1 2
1 45
2 160
3 360
9 370
42 3,962
925
-------
STACK
vo
10
o\
FLUE GAS
1 BOOSTER FAN
2 ABSORBER
3 MIST ELIMINATOR
4 OXIDIZER
5 THICKENER
6 CENTRIFUGE
7 BASIC ALUMINUM SULFATE
REGENERATION TANKS
8 REHEAT FACILITY
L
WASH
WATER
r
GYPSUM
Figure 4-1 DOWA BASIC ALUMINUM SULFATE - GYPSUM FLUE GAS DESULFURIZATION PROCESS
-------
Neutralization
The oxidized liquor is treated with powdered limestone to precipitate gypsum and
to regenerate the basic aluminum sulfate solution:
A12(S04)3 • A12(S04)3 +. 3CaC03 + 6H20 = A12(S04)3 . A1203
+ 3 (CaS04 • 2H£0) + SCO (3)
Commercial Applications
One of the earlier applications ("1 MW) was at the Mobara Works of Taenaka
Mining. This plant started operation in'October 1972 to treat waste gas from
a molybdenum sulfide roaster containing 7500 ppm S02 at 100°C. No problems
have been reported to date. Two commercial units, each with a capacity of
treating 150,000 Nm3/hr (50 MW) of tail gas from a sulfuric acid plant were
built at Okayama Works of Dowa. For these systems, inlet S02 averages 600 ppm.
They are designed to remove at least 95% S02 at a liquid-to-gas ratio of 15-20
gal./l(Pacf. These plants have operated successfully since startup in July 1974,
with an average S02 removal efficiency of 99% during 20 months of testing. The
only installation on a boiler, a small one at Naikai, also continues to perform
well.
Other recent applications include two on iron ore sintering plants, one on a sul-
furic acid plant, and one on a smelter.
Evaluation
The attractive features of this process, relative to the concentrated sodium-
based processes, are unlimited oxidation tolerance and the use of limestone as
a regenerant. However, this process operates at a higher liquid-to-gas ratio
than the sodium-based processes.
4.3 KAWASAKI MAGNESIUM - GYPSUM FGD PROCESSED')
Kawasaki Heavy Industries markets two versions of the magnesium-gypsum flue gas
desulfurization process. In their "standard" process only a portion of the
absorber 'bleed stream is oxidized and the mole ratio of calcium to magnesium in
the absorbent liquor is maintained between 3 and 4. In the "new" process all
the bleed stream is oxidized and the Ca/Mg mole ratio is held below 1.0.
Even though both the known commercial applications use the "standard" process,
the description presented in this paper relates to'the new process because
of its greater potential. Figure 4-2 is a simplified process flow diagram of
this "new" process.
Process Description
Unlike the Dowa process, this process does not employ clear liquor scrubbing.
927
-------
VO
N3
oo
STACK
1 FORCED DRAFT FAN
2 ABSORBER
3 MIST ELIMINATOR
4 OXIDIZER
5 THICKENER
6 CENTRIFUGE
7 MAGNESIUM HYDROXIDE
REGENERATION TANK
8 REHEAT FACILITY
Figure 4-2 KAWASAKI MAGNESIUM - GYPSUM FLUE GAS DESULFURIZATION PROCESS (NEW)
-------
Its three unit operations (absorption, oxidation, and gypsum recovery, and
magnesium hydroxide regeneration) are described below.
Absorption
Flue gas containing S02 is contacted with a mixed slurry of calcium (as CaS04)
and magnesium compounds. SO? is absorbed and removed according to reactions
(4) and (5).
Mg(OH)2 + S02 = MgS03 + HgO (4)
MgS03 + S02 + H20 = Mg(HS03)2 (5)
Oxidation and Gypsum Recovery
The spent liquor from the absorption section is oxidized by air to convert the
magnesium sulfite to sulfate by the following reactions:
MgSOs + 1/2 02 = MgS04 (6)
Mg(HS03)2 + 1/2 02 = MgS04 + S02 + H20 (7)
Since magnesium sulfate has a high solubility in water relative to calcium
sulfate (calcium sulfate is recycled for seeding), separation of the calcium
and magnesium sulfates is easy. Gypsum slurry from the oxidizer is dewatered
to less than 10 wt.% moisture with a centrifuge. Magnesium sulfate solution is
forwarded to the magnesium hydroxide regeneration section.
Magnesium Hydroxide Regeneration
Since magnesium sulfate has no capacity to absorb S02, it is converted
back to magnesium hydroxide. The regeneration is accomplished by adding lime
or limestone. The following reaction occurs with lime.
MgS04 + Ca(OH)2 + 2 H20 = Mg(OH)2 + CaS04 • 2 H20 (8)
Magnesium hydroxide thus regenerated is returned to the absorber along with the
precipitated gypsum.
Commercial Applications
Kawasaki Heavy Industries has built two commercial plants which have been oper-
ational since the beginning of 1976. Both plants use the partial oxidation
mode (standard process). The first commercial plant utilizing this process
is the Okazaki Works of the Unitika Co.; the second is the Saidaji Works of
929
-------
Japan Exlan Co. For the latter plant, inlet S02 averages 1400 ppm. For 95%
S02 removal, a liquid-to-gas ratio of 35-45 gal./103acf is required. The mole
ratio of calcium to magnesium is maintained between 3.5 and 4.
Lime is used in the plant for Japan Exlan Co., whereas limestone is used as
the main absorbent (with a little lime) for the Unitika Co. For the Unitika
application, lime is added to the reactor, and limestone is added to the
absorber; the ratio of limestone to lime is about 5 to l.U) The use of lime-
stone is a Unitika Co. refinement. Gypsum is recovered as a by-product in
both plants.
Evaluation
Like the Dowa process, the major attractions of this process are unlimited
oxidation tolerance and the option of limestone as a regenerant. However,
the reported liquid-to-gas ratio for the "standard" version is significantly
higher than that of the sodium-based processes. However, for the "new"
version the liquid-to-gas ratio is comparable to that of the sodium-based
process because most of the scrubbing is done by magnesium.
4.4 KUREHA SODIUM ACETATE-GYPSUM FGD PROCESS^25)
Two limitations of the sodium sulfite dual alkali process marketed byKureha
were its inability to accommodate high oxidation levels and its difficulty in
using limestone as a regenerant. In response to these shortcomings, Kureha
developed the sodium acetate process described below.
Process Description
The process is composed of three unit operations: absorption, oxidation, and
gypsum recovery.
Absorption
As illustrated in Figure 4-3, the absorption tower consists of two sections:
S02 absorption and an acetic acid recovery section linked in series in the
same absorption tower.
In the S0£ absorption section, S02 is removed by circulating sodium acetate
solution. The following reaction takes place:
2 CH3COONa + S02 + H20 = Na2S03 + 2 CH3COOH (9)
930
-------
^^
4
VO
co
FLU EG AS
f
\
STACK
(MAKE UP)
1 FORCED DRAFT FAN
2a ABSORPTION SECTION
2b ACETIC ACID RECOVERY
SECTION
3 MIST ELIMINATOR
4 OXIDIZER
5 ABSORBENT SURGE TANK
6 CENTRIFUGE
7 CENTRATE HOLD TANK
8 REHEAT FACILITY
9 ACETIC ACID RECOVERY TANK
10 GYPSUM RECOVERY REACTOR
-1 /^~\
-ir
A
~t i
®
i
(MAKE-UP)
Figure 4-3 KUREHA SODIUM ACETATE - GYPSUM FLUE GAS DESULFURIZATION PROCESS
-------
Part.of the acetic acid thus formed volatilizes in the scrubbed flue gas, and
is recovered in the upper section. In this recovery section, fresh limestone
slurry is added to the top chamber, from where it flows down countercurrently
from chamber to chamber to the bottom to completely remove the acetic acid vapor.
This step also completes the removal of remaining S02 in the flue gas from the
absorption section.
Oxidation
Sodium sulfite formed in the absorber is oxidized to sodium sulfate in the
oxidation tower in which perforated plates facilitate fine dispersion of air
bubbles and promote oxidation. Sulfite oxidation to sulfate takes place as
follows:
Na2S03 + 1/2 02 = Na2S04 (10)
After oxidation, the liquor is sent to the gypsum recovery section, where gypsum
is produced by the addition of limestone slurry.
Gypsum Recovery
In this operation, calcium carbonate reacts very rapidly with sodium sulfate
in the presence of acetic acid; the reaction of calcium carbonate with sodium
sulfate is very slow in the absence of acetic acid.
It is believed that calcium carbonate reacts first with acetic acid present in
the liquor to form calcium acetate, which reacts further with sodium sulfate
to form calcium sulfate and sodium acetate by a double-decomposition reaction.
The regenerated sodium acetate is recirculated to the scrubber after separation
of gypsum.
The reaction mechanism can be expressed as in (11) and (12):
CaC03 + 2 CH3COOH = (CH3COO)2 Ca + H20 + C02 (11)
(CH3COO)2Ca + Na2S04 = CaS04 + 2 CH3COONa, (12)
Evaluation
Like the Dowa and Kawasaki processes, unlimited oxidation tolerance and the use
of limestone as a regenerant are attractive. The claimed liquid-to-gas ratio is
on the order of 10 gal./lO^acf. Kureha also offers the "lime-regenerant version"
which is claimed to be simple and cheaper due to elimination of the acetic acid
recovery section, a smaller effluent hold tank, and the consolidation of the oxi-
dation and the regeneration operations in one vessel.
Reportedly, the process can also be modified by adding catalysts to the scrubbing
liquor for simultaneous removal of S02 and NOX.
932
-------
Commercial Applications
Unlike the Dowa and Kawasaki processes, there are no commercial applications
of this process. However, this process has been tested in a bench-scale plant
(100 MITT/hr) with flue gas from an oil-fired boiler for 2 years since 1973.
Furthermore, on the basis of the results in the bench-scale plant, a 5,000
Nm3/hr pilot plant was constructed in 1975. This pilot plant has operated
smoothly and data necessary for scale-up have been obtained.
933
-------
Section 5
THE EPA DUAL AUCALI RESEARCH, DEVELOPMENT, AND DEMONSTRATION PROGRAM
5.1 BACKGROUND
The EPA has been actively involved in the development of dual alkali technology
since the Second International Lime/Limestone Wet-Scrubbing Symposium held in
New Orleans in November 1971. Some of the incentive to develop this technology
stemmed from a paper presented by R.J. Phillips'20) Of General Motors (GM) con-
cerning GM's laboratory and pilot plant work with a dilute mode dual alkali sys-
tem. The results of the GM effort appeared very encouraging at that time in
light of the difficulties being experienced with lime/limestone systems. The
development of dual alkali technology by the EPA has followed an orderly
progression of scale from laboratory to pilot plant to prototype to a full-scale
utility demonstration of the process. In addition, EPA also funded a program to
evaluate the full-scale dilute mode dual alkali system in operation at a GM
industrial boiler system (32 MW equivalent).
Some initial laboratory work on regeneration chemistry was done in the EPA/IERL-
RTP laboratories at Research Triangle Park, in addition to an initial feasibility
studyl23) which indicated that dual alkali systems might be somewhat lower in
capital and operating costs than lime or limestone systems under certain cir-
cumstances.
After these initial studies, the EPA contracted with ADL to conduct a study of
the dual alkali process. The scope of work included in the initial laboratory
and pilot plant program was subsequently expanded to include prototype testing
at the Scholz Plant of Gulf Power Company where a 20 MW FGD prototype system was
constructed by CEA/ADL for Southern Services. The Southern Company (parent of
Southern Services) provided the funds for the project.
Goals of the initial EPA dual alkali program were to:
• Demonstrate reliable system operation.
t Demonstrate high SOg removal, 95% desirable, with high-sulfur coal.
t Demonstrate environmentally acceptable sulfate removal schemes.
t Minimize soluble materials in disposable waste.
• Minimize moisture in disposable waste.
t Demonstrate closed-loop operation.
• Minimize costs.
• Minimize Ca concentration in the scrubber.
934
-------
These goals have been successfully achieved up to the prototype level,
5.2 LABORATORY PROGRAM
Areas of investigation in the laboratory program include:
• Regeneration of simulated scrubber effluents with lime and limestone.
• Sulfate removal by precipitation of a mixed crystal containing CaSOo
and CaS04 with water of hydration.
• Sulfate removal by reaction of Na2S04, CaS03, and sulfuric acid.
• Feasible ranges of sulfate, chloride, magnesium, and iron in solution.
• Settling characteristics of the product solids.
• Fixation of dual alkali product solids.
• Density, comparability, Teachability, and permeability of fixed and
unfixed solids.
5.3 PILOT PLANT PROGRAM
In the pilot plant both short- and long-term runs (5 weeks/run) have been
conducted to examine various modes of dual alkali operation. These modes of
operation include:
• Concentrated alkali, sulfuric acid treatment, lime.
t Concentrated alkali, two-stage reactor system, lime.
t Concentrated alkali, single reactor, lime.
• Dilute alkali with sulfite oxidation, lime.
• Concentrated alkali, multi-stage reactor, limestone regeneration.
• Extra high concentrated alkali, two-stage reactor system, lime.
LaMantia, et al.(15) summarizes the above noted laboratory and pilot plant work.
5.4 PROTOTYPE PROGRAM
The results of 17 months of testing at the 20 MW prototype facility of Gulf
Power are contained in the report entitled "Final Report: Dual Alkali Test and
Evaluation Program, Vol. III. Prototype Test Program - Plant Scholz," by LaMantia,
et al.U6) Current activity at this facility involves'preparation for testing
935
-------
of limestone as a regenerant.
5.5 GM INDUSTRIAL BOILER FGD SYSTEM EVALUATION
The GM test program at the 32 MW equivalent system at the Chevrolet Transmission
plant in Parma, Ohio, was conducted under an agreement between the EPA and GM.
The test program design and some chemical analyses were perforned for GM and
the EPA, under contract, by ADL. The FGD system consists of four, Koch tray,
stainless steel scrubbers, 32 MW equivalent, and a 40 MW equivalent regenera-
tion system consisting basically-of two tanks and two reactor-clarifiers.
The boiler system consists of two 60,000 Ib/hr (of steam) and two 100,000
Ib/hr boilers. The GM system is a dilute mode dual alkali system.
The system was operated intermittently since startup in February 1974. Avail-
ability data for the system were not easily defined since there are four boilers
and four scrubbers while the steam demand was frequently less than the capacity
of one or two boilers. Thus, when there was a problem with one of the systems
it was taken out of service and replaced by one of the stand-by units.
A good account of this system was presented by Dingo and Piasecki(?) at the
1974 EPA FGD Symposium in Atlanta. In addition, the results of a
evaluation of this system by Arthur D. Little are also available.
5.6 FULL-SCALE UTILITY FGD'SYSTEM DEMONSTRATION
As indicated earlier, Louisville Gas and Electric's Cane Run No. 6 unit was
chosen as the site for a full-scale, EPA co-funded, dual alkali demonstration
effort. The vendors providing the system and the monitoring program director
are CEA/ADL and Bechtel National, Inc., respectively.
The program, to be conducted on this 277 MW unit, consists of four phases:
(I) Process design and cost estimate.
(II) Engineering design, construction, and mechanical testing.
(Ill) Startup and acceptance testing.
(IV) One year operation and long-term testing.
As of this date, phases I and II have been completed. Phase III is presently
in progress. Results of phase I have been reported by Van Ness, et al.132)
Summary design criteria for this demonstration unit include:
t Unit will meet all applicable pollution control regulations when
th.e pulverized coal-fired boiler burns 2.5 - 4.5 wt.% sulfur coal.
• Instrumentation will allow accurate material and energy balances.
936
-------
• S02 outlet will be held below 200 ppm for coal sulfur less than
5 wt.%. For higher sulfur coals, SO? removal will be at least
95%.
• Stack gas will be reheated.
• Chemical makeup levels:
Na = 0.0495 mole Na or monovalant Cation/mole S
Ca = 1.01 mole Ca/mole S
An extensive monitoring program will be conducted at this facility. The
program will characterize effluents, document chemical consumption, evaluate
sludge disposal, and evaluate system reliability and process economics.
5.7 FUTURE PLANS
Based on the level of future funding, the EPA dual alkali program may be up-
graded and supplemented by expanding the currently planned monitoring program
on the Louisville Gas and Electric's 280 MW facility by 1 year to allow testing
of: (1) limestone as an alternative to the more expensive, energy-intensive
lime reactant; (2) methods for upgrading the quality of sludge produced and
to compare disposal options; and (3) strategies to control the multimedia en-
vironmental impact of all effluents and emissions.
In addition, process evaluation programs may be initiated on the other two full-
scale dual alkali facilties presently in the final stages of construction, if
appropriate contractual arrangements can be made with the respective utility
company and process supplier. These facilities are the 575 MW Central Illinois
Public Service and 250 MW Southern Indiana Gas and Electric systems being en-
gineered by Envirotech and FMC, respectively. Parallel test programs may be
conducted by an independent EPA subcontractor to evaluate, characterize, and
compare these full-scale facilities, and allow evaluation of promising dual
alkali process variations.
937
-------
Section 6
REFERENCES
1. Ando, J., et al., "S02 Abatement for Stationary Sources in Japan,"
EPA Report No. 600/7-78-210, (NTIS PB 290198), November 1978.
2. Ando, J., et al., "S02 Abatement for Stationary Sources in Japan,"
EPA Report No. 600/2-76-013a, (NTIS PB 250585), January 1976.
3. Bloss, E.H., Private Communication, Envirotech.
4. Cornell, C.F. and D.A. Dahlstrom, "Performance Results on a 2500 ACFM
Double Alkali Plant for S02 Removal." Presented at the 66th Annual
Meeting of A I Ch E, Philadelphia, Pennsylvania, November 11-15, 1973.
Condensed version of the paper appeared in December 1973 CEP.
5. Cornell, C.F-, "Liquid-Solids Separation in Air Pollution Removal Sys-
tems." Presented at the ASCE Annual and National Environmental Engineer-
ing Convention, Kansas City, Missouri, October 21-25, 1974.
6. Devitt, T., et al., "Flue Gas Desulfurization System Capabilities for
Coal-Fired Steam Generators," Volume II, Technical Report, EPA Report
No. 600/7-78-032b, (NTIS PB 279417), March 1978.
7. Dingo, T. and E. Piasecki, "General Motor's Operating Experience with
a Full-Scale Double Alkali Process." Presented at the EPA Symposium on
Flue Gas Desulfurization, Atlanta, Georgia, November 4-7, 1974.
8. Ellison, W., et al., "System Reliability and Environmental Impact of S0£
Scrubbing Processes." Presented at Coal and the Environment, Technical
Conference, Louisville, Kentucky, October 22-24, 1974.
9. Epstein, M., R. Borgwardt, et al., "Preliminary Report of Test Results
from the EPA Alkali Scrubbing Test Facilities at the TVA Shawnee Power
Plant and at Research Triangle Park." Presented at Public Briefing-,
Research Triangle Park, North Carolina, December 19, 1973.
10. Hoi hut, W.J., et al., "Zero-Effluent Throwaway S02 System Design for
High-Chloride, High-Sulfur Coal," Proceedings of the American Power
Conference, Vol. 38, 1976.
11. Interess E., "Evaluation of the General Motors' Double Alkali SOo Control
System." EPA Report No. 600/7-77-005, (NTIS PB 263469), January 1977.
12. Johnstone, H.F., et al., "Recovery of Sulfur Dioxide from Waste Gases."
Ind. & Eng. Chem., Vol. 30, No. 1, January 1938, pp. 101-109.
938
-------
13. Kaplan, N., "Introduction to Double Alkali Flue Gas Desulfurization
Technology" Presented at the EPA Flue Gas Desulfurization Symposium, New
Orleans, Louisiana, March 8-11, 1976.
14. LaMantia, C.R., et al., "EPA/ADL Dual Alkali Program Interim Results."
Presented at EPA Symposium on Flue Gas Desulfurization, Atlanta,
Georgia, November 4-7, 1974.
15. LaMantia, C.R., et al., "Final Report: Dual Alkali Test and Evaluation
Program, Volume II. Laboratory and Pilot Plant Programs." EPA Report
No. 600/7-77-050b, (NTIS PB 272770), May 1977-
16. LaMantia, C.R., et al., "Final Report: Dual Alkali Test and Evaluation
Program, Volume III. Prototype Test Program - Plant Scholz." EPA Report
No. 600/7-77-050C, (NTIS PB 272109), May 1977.
17. Maxwell, M.A., et al., "Sulfur Oxides Control Technology in Japan,"
Interagency Task Force Report, June 30, 1978.
18. McGlamery, G.G. and R.L. Torstrick, "Cost Comparisons of Flue Gas De-
sulfurization Systems." Presented at the EPA Symposium on Flue Gas
Desulfurization, Atlanta, Georgia, November 4-7, 1974.
19. Phillips, R.J., "Operating Experiences with a Commercial Dual-Alkali
S02 Removal System." Presented at the 67th Annual Meeting of the Air
Pollution Control Association, Denver, Colorado, June 9-13, 1974.
20. Phillips, R.J., "Sulfur Dioxide Emission Control for Industrial Power
Plants." Presented at the Second International Lime/Limestone Wet-
Scrubbing Symposium, New Orleans, Louisiana, November 8-12, 1971.
21. Ponder, W.H., "Status of Flue Gas Desulfurization Technology For Power
Plant Pollution Control." Presented at Thermal Power Conference, Wash-
ington State University, Pullman, Washington, October 4, 1974.
22. Ramirez, A.A., Private Communication, FMC.
23. Rochelle, G.T., "Economics of Flue Gas Desulfurization." Presented
at EPA Flue Gas Desulfurization Symposium, New Orlean, Louisiana, May
14-17, 1973.
24. Seamans, T., Private Communication, Ionics Incorporated, Watertown,
Massachusetts.
25. Saito, S., et al., "Kureha Flue Gas Desulfurization-Sodium Acetate-
Gypsum Process," Kureha Chemistry Industry Co., Ltd., Tokyo, Japan.
26. Selmeczi, J.G. and R.G. Knight, "Properties of Power Plant Waste Sludges."
Presented at the Third International Ash Utilization Symposium, Pitts-
burgh, Pennsylvania, March 13-14, 1973.
27. "Sulfur Dioxide and Flyash Control," FMC Corporation Technical Bulletin.
FMC Corporation, Air Pollution Control Operation, 751 Roosevelt Road,
Suite 305, Glen Ellyn, Illinois 60137.
939
-------
28. Summary, Report. Flue Gas Desulfurization Systems. June-July 1978.
Prepared by PEDCo Environmental Inc. EPA Report No. 600/7-78-051d,
(NTIS PB 288299), November 1978.
29. Torstrick, R.L., et al., "Economics Evaluation Techniques, Results, and
Computer Modeling for Flue Gas Desulfurization," Proceedings: Symposium
on Flue Gas Desulfurization Symposium, Hollywood, Florida, November 1977
EPA Report No. 600/7-78-058a, (NTIS PB 282090), March 1978.
30. Tsugeno, H., et aT., "Operati-ng Experiences with Kawasaki Magnesium -
Gypsum Flue Gas Desulfurization Process," Proceedings: Symposium on
Flue Gas Desulfurization Symposium, Hollywood, Florida, November 1977,
EPA Report No. 600/7-78-058b, (NTIS PB 282091), March 1978.
31. Tuttle, J., et al., "EPA Industrial Boiler FGD Survey: Third Quarter
1978." EPA Report No. &00/7-78-052c, November 1978. Prepared by
PEDCo Environmental Inc.
32. Van Ness, R.P., et al.. "Project Manual for Full-Scale Dual Alkali
Demonstration at Louisville Gas and Electric Co. - Preliminary Design
and Cost Estimate." EPA Report No. 600/7-78-010, (NTIS PB 278722),
January 1978.
33. Yamamichi Y- and Nagao J., "The Dowa's Basic Aluminum Sulfate-Gypsum Flue
Gas Desulfurization Process," The Dowa Mining Co., Ltd., Okayama Research
Laboratory, Okayama, Japan.
940
-------
Appendix A
DEFINITION AND DISCUSSION OF
As with any specialized technology, a discussion of flue gas desulfurization
in general, and dual alkali technology more specifically, involves the use
of special terminology which has evolved with the technology. While terms are
understandable to those dealing with the subject on a daily basis, they can be
somewhat ambiguous to others. To clarify some of these ambiguities, and to
define terms used here and by others describing dual alkali technology devel-
opment, a number of terms and concepts are defined and discussed in a general
sense.
A.I
ABSORPTION/REGENERATION CHEMISTRY
The main chemical reactions that take place in dual alkali systems can be
divided functionally into the absorption and regeneration reactions. A number
of secondary reactions which have very important effects on the overall func-
tioning of the system also take place. These include oxidation, softening, and
sulfate removal reactions which are discussed under the appropriate headings.
The regeneration reactions and in some cases the absorption reactions will
depend upon which calcium supplying regenerant is used—lime or limestone.
With lime, the system can be operated over a wider pH range than with limestone.
This wider pH range allows lime systems to operate over the complete range of
active alkali hydroxide/sulfite/bisulfite, whereas limestone systems can only
operate in the sulfite/bisulfite range.
The main overall absorption reactions are described by the following equations:
ZNaOH + S02 = Na2S03 + H20 (1)
Na2S03 + S02 + H20 = 2NaHS03 (2)
The main overall regeneration reactions are described by the following equation:
for lime and limestone, respectively:
Lime
Ca(OH)2 + 2NaHS03 =
CaSOg • 1/2
3/2 H20
Ca(OH)2 + Na2S03 + 1/2 H20 = 2NaOH + CaS03 • 1/2 H20
Ca(OH)2 + Na2S04 + 2H20 = CaS04 . 2H20 + 2NaOH
(3)
(4)
(5)
941
-------
Limestone
CaC03 + 2NaHS03 + 1/2 H20 = Na2S03 + CaS03« 1/2 H20 + C02 + H20 (6)
D/A systems are frequently and erroneously referred to as sodium ion scrubbing
systems. It should be stated that, from a "pure" chemistry viewpoint, the
reactions presented in equations (1-11) and (15-18) do not involve the sodium
ion (Na+); however, the presentation is made using compounds of sodium because
sodium systems are prevalent in D/A applications and because this allows showing
the reactions using electrically neutral reactants and products rather than
charged ions. For example, the absorption reactions jnvolve reaction of S02
with an aqueous base such as OH", S03~, HC03", or C03=, rather than with Na+
which does not take part in the reaction, but is only present to maintain
electrical neutrality. Thus equations (1) and (2) for example could have been
written as equations (la) and (2a), respectively:
2 OH" + S02 = S03= + H20 (la)
S03= + S02 + H20 = 2HS03~ (2a)
A.2 ACTIVE ALXALI
This term refers to the concentrations of NaOH, Na2C03, NaHC03, Na2S03, and
NaHS03 in the scrubbing solutions. Sodium bisulfite is included in this defi-
nition although it is not technically an alkali (i.e., it cannot react with S02
in these systems); however, it can be converted to an alkali by reaction with
lime or limestone. It should also be noted that the, molar capacity of each of
these species for absorption of S02 is different, and can vary from zero to 2
moles of S02 per rnole of active alkali. This difference in molar capacity for
absorption of S02 is illustrated by the following reaction equations:
Na2C03 + 2S02 + H20 = 2NaHS03 + C02 (7)
(sodium carbonate molar capacity: 2 moles S02/mole)
NaHC03 + S02 = NaHS03 + C02 (8)
(sodium bicarbonate molar capacity: 1 mole S02/mole)
NaOH + S02 = NaHS03 (9)
(sodium hydroxide molar capacity: 1 mole S02/mole)
Na2S03 + S02 + H20 = 2NaHS03 (10)
(sodium sulfite molar capacity: 1 mole S02/mole)
942
-------
NaHS03 + S02 = No reaction (11)
(sodium bisulfite molar capacity: zero mole S
Molar capacity is simply the number of rnoles of S02 needed to convert 1 mole of
the absorbent alkali completely to sodium bisulfite. Since there is a difference
in the molar capacity of different alkali components to absorb S02, active
alkali is a descriptive rather than a quantitative term. If the concentration
of each active alkali component (moles/liter) is known, the capacity of the
scrubbing liquor to absorb S02 (moles of S02/liter of solution) can be calculated
as the sum of each of the active alkali component concentrations multiplied by their
respective molar capacities as follows:
Scrubber liquor S02 capacity (moles/liter) =
2 [Na2C03] + [NaOH] + [NaHC03] + [Na2S03]
A.3 TOS
Total oxidizable sulfur (TOS) denotes the concentration of sulfur compounds in
solution in which the sulfur is in the +4 oxidation state. Simply, this is
the total concentration of sulfite plus bisulfite.
TOS (moles/liter) = [S03=] + [HS03~]
Sulfate is not part of TOS, since the sulfur is in the +6 oxidation state in
this species. Sulfur dioxide dissolving in scrubbing solutions increases the TOS
in solution.
A.4 ACTIVE SODIUM
This is the concentration of sodium in solution which is associated with the
active alkali.
[Na+]act = [NaOH] + 2 [Na2C03] + [NaHC03] + [NaHS03] + 2 [Na2S03]
If NaOH, NaHC03, Na2C03, NaHS03, or Na2S03 solids are added to dual alkali
solutions, the increase in sodium ion in solution is "active sodium." If
Na2S04 or NaCl, for example, is added, the increase in sodium ion is "inactive
sodium." Active sodium is not increased by the dissolution of S02 in scrubber
solutions. Note that the term "active sodium" can be misleading in that the
sodium ion doesn't participate in any of the process reactions.
943
-------
A.5 OXIDATION
Oxidation in a dual alkali system refers to the conversion of TOS to sulfate
by one of the following equations:
HS03" + 1/2 02 = S04= + H+ (12)
S03= + 1/2 02 = S04= (13)
Simple oxidation of S02 to $03 in the flue gas is also considered oxidation in
the dual alkali system:
S02 + 1/2 02 = SOs (14)
Oxidation in the system has the effect of changing active sodium to inactive
sodium, or active alkali to inactive alkali.
Oxidation may occur in any part of the system: in the scrubber, the reaction
vessels, or in the solids separation equipment. In general, the rate of oxida-
tion in the system is thought to be a function of the rate of dissolution of
oxygen, pH of the scrubbing solution, impurities present in solution, and con-
centration of reactant's. Oxidation rate is thus affected by composition of the
scrubbing liquor (scrubbing liquors containing high concentrations of dissolved
salts may absorb oxygen more slowly), oxygen content of the flue gas, impurities
in the coal and lime or limestone, and the design of the equipment (the regener-
ation and solids separation sections of the system in particular can be designed
to limit dissolution of oxygen and the number of scrubber contact stages is
extremely important).
Oxidation rate is expressed as a percentage and is calculated from an overall
material balance on the system:
Oxidation rate (%) = [S0a= leaving the system (moles)] x 100
[Total sulfur collected (moles)]
Sulfate leaving the system is total moles of sulfate in the solid waste plus any
sulfate in the associated liquor.
A.6 SULFATE REGENERATION
This term is a misnomer. What is really meant is sulfate removal from the system
with regeneration of active alkali from inactive sodium sulfate. (See Sulfate
Removal , bel ow. )
A. 7 SULFATE REMOVAL
Sulfate is removed from the system with regeneration of active alkali from in-
active sodium sulfate. Examples of these sulfate removal reactions are:
944
-------
Na2S04 + Ca(OH)2 + 2H20 = 2NaOH + CaS04 • 2H20 (15)
Na2S04+ 2CaS03 • 1/2 H20 + H2S04 + 3/2H20 = 2NaHS03 + 2CaS04 • 2H20 (16)
(gypsum)
y Na2S04 + x NaHSOs + (x+y) Ca(OH)2 + (z-x) HgO = (17)
(x+2y) NaOH + x CaS03-y CaS04 • z H20
(mixed crystal or solid solution)
Electrolytic cell
2NaOH + H2S04 + H2 + 1/2 02 (18)
Sulfate should be removed in an environmentally acceptable manner; a simple
purge of soluble Na2S04 from the system to land or waterway disposal is not
acceptable.
A.8 SOFTENING
This term is used to describe various methods used to lower the dissolved calcium
ion concentration in regenerated solutions. The purpose of softening the scrub-
bing liquor before recycling to the scrubber is to ensure that it is subsaturated
with respect to gypsum. This reduces the gypsum scaling potential in the scrubber.
Examples of softening reactions are:
Ca++ + Na2C03 = 2Na+ + CaC03 (19)
Ca++ + Na2S03 + 1/2 H20 = 2 Na+ + CaS03 • 1/2 H20 (20)
Ca++ + C02 + H20 = 2H+ + CaC03 (21)
In each of the above reactions, calcium ions are removed from solution as part
of an insoluble material, outside the scrubber system. Reactions (19) and (21)
are referred to as carbonate softening. Reaction (20) is considered sulfite
softening. Generally, dilute systems employ carbonate softening; concentrated
system inhibit scaling due to the high sulfite concentrations which prevent
high Ca ion concentrations in the scrubber liquor.
A.9 DILUTE VS. CONCENTRATED SYSTEMS
Dilute or concentrated refers to the active alkali concentration in a particular
system. This differentiation is made because, in theory at least, based on their
solubility products in water, both CaS03 and CaS04 should not precipitate from
a solution of sulfite and sulfate simultaneously when using relatively small
quantities of lime slurry for regeneration, unless the concentrations of sul-
fite and sulfate are present in a certain ratio. This can be shown by dividing
one solubility product equation by the other:
945
-------
[Ca++] [S04"] = Ksp
LCa-1"1-] [S0s=] = Ksp
from this, cancelling Ca"1"1" ion concentration,
[S04=] = constant
The ratio of sulfate to sulfite for simultaneous precipitation of CaSO^ and
CaSQjis shown to be a constant. It is on the order of lO^-lO^ in ideal, ex-
tremely dilute solutions.
The constant in the above equation is the ratio of solubility product constants
of calcium sulfate and sulfite. Thus, in theory, if the ratio of sulfate to
sulfite is higher than this constant, only calcium sulfate should precipitate;
and if the ratio is lower than the constant, only calcium sulfite should pre-
cipitate.
This very simplified consideration of the chemistry given above is clouded in
the "real world" by factors that contribute to non-ideal behavior of these
systems. These factors include changes in ionic activities in solutions con-
taining high electrolyte concentrations, and evidence of coprecipitation of cal-
cium sulfite and sulfate in the form of a "mixed crystal" PI\!ls9Jid solution"
in a manner which is not completely understood at present. (9)(14)
With due consideration to the non-ideal behavior of these systems, however,
under given conditions, a ratio of sulfate to sulfite in solution can be
determined at which the previously cited examples hold. The ratio establishes
a definition for "dilute" or "concentrated" dual alkali systems. When the
ratio is such that gypsum alone or both gypsum and calcium sulfite will pre-
cipitate from the solution with the addition of slaked lime, the system is
"dilute."
A.10 LIME OR LIMESTONE STOICHIOMETRY
Lime or limestone stoichiometry can be expressed as a percentage based on an
overall material balance around the system:
Lime Stoichiometry = moles CaO added x 100
mole sulfur collected
Limestone Stoichiometry = moles CaCOa added x 100
mole sulfur collected
Lime or limestone stoichiometry is an indication of the efficiency of utiliza-
tion of lime or limestone used. Ignoring alkali components in the fly ash
collected and the alkalinity added with sodium makeup to the system, 100%
stoichiometry is complete utilization; stoichiometries over 100% represent less
946
-------
efficient utilization of lime or limestone,
alkalinity from other sources.
Stoichibmetries under 100% indicate
A.11 FEED STOICHIOMETRY
This is calculated by a material balance around the scrubber. It is usually
expressed as the ratio:
Feed Stoich = Liquor S02 Capacity (moles/liter) x Flow (liter/min)
(mole/min)
This ratio is evaluated for the gas and liquid streams entering the
scrubber.
Feed stoichiometry is a measure of the ability of the incoming liquor to react
with or absorb all of the incoming S02 in the scrubber, assuming ideal contact
of gas and liquor. Feed stoichiometry above 1.0 is required for high S02 re-
moval capability. At feed stoichiometry at or below 1.0, assuming ideal contact
between the gas and liquor, there will be significant equilibrium SOg partial
pressure above the liquor, and thus S02 removal is theoretically limited to the
value calculated on the basis of this S02 partial pressure in the exiting flue gas.
947
-------
Appendix B
DESIGN, PROCESS AND COST CONSIDERATIONS^13)
A commercial dual alkali system must be designed to remove the desired quantity
of sulfur oxides from a given flue gas stream, while operating in a reliable
manner and discharging environmentally acceptable §olid waste product. In ful-
filling these design objectives, cost is also important.
B.I S02 REMOVAL
For some time it has been known that small quantities of sulfur dioxide can be
removed from large amounts of relatively inert gas by cyclic processes involving
absorption into aqueous solutions of sodium sulfite/bisulfite. Johnstone et al.
published a paper in 1938 giving data on the vapor pressure of S02 over solutions
of sulfite/bisulfite and methods of calculating these equilibrium values under
various conditions. The equilibrium partial pressure of S02 over sulfite/bisulfite
solutions, the theoretical limit which a/practical design can approach, is generally
a function of solution temperature, pH, concentration of sulfite/bisulfite, and
total ionic strength. Since Johnstone's work, a number of organizations have
pursued this technology with laboratory, pilot plant, and full-scale applications
for flue gas desulfurization, and many have demonstrated its ability for high
removal efficiencies. (Note that, although Johnstone's work was aimed at cyclic
processes with thermal regeneration, such as the Wellman-Lord system, the vapor
pressure data are also applicable to dual alkali systems which use chemical
regeneration.)
Once methods have been established to determine equilibrium S02 vapor pressure
over scrubbing solutions of the various concentrations to be encountered in an
operating system, it becomes a matter of standard chemical engineering practice
to design adequate gas absorption equipment to accomplish the desired S02 re-
moval in a system. For comparison, note that the design of lime/limestone slurry
absorption equipment is further complicated by the kinetics of dissolution of
the lime or limestone, the particle size of the suspended material, and the crystal
morphology of the lime or limestone.
B.2 RELIABLE OPERATION
System reliability can be adversely affected by two classes of problems: mechan-
ical and chemical.
Mechanical problems include malfunction of instrumentation and mechanical and
electrical equipment such as pumps, filters, centrifuges, and valves. These
problems in a commercial FGD system can be minimized by careful selection of
materials of construction and equipment and by providing spares for equipment
items such as pumps and motors which are expected to be in continuous operation
and are prone to failure after a relatively short period of operation. Another
948
-------
important consideration in minimizing mechanical problems is the institution
of a good preventive maintenance program.
Chemical (or physical/chemical) problems which may be associated with a dual
alkali system include scaling, production of poor-settling solid waste product,
excessive sulfate buildup, water balance, and buildup of non-sulfur solubles
which enter the system as impurities in the coaf or lime. Each factor is
associated with reliable system operation, or production of an environmentally
acceptable solid waste.
Scaling - One of the primary reasons, and probably the most important, for
development of dual alkali processes was to circumvent the scaling problems
associated with lime/limestone wet scrubbing systems. Therefore, a dual
alkali system should be designed to operate in a non-scaling manner.
Scaling is caused by precipitation of calcium compounds (from process liquors)
on the surfaces of various system components. In the scrubber, this is parti-
cularly troublesome since the flue gas path through the scrubber, if affected,
could shut down the boiler/scrubber system and lower reliability.
Since scrubbing in dual alkali systems employs a clear solution rather than a
slurry, there is a tendency to ignore potential scaling problems. Testing ex-
perience with dual alkali systems has indicated, however, that scaling can
occur and indeed the problem should be a legitimate concern in the design of
any system. Both gypsum and carbonate scale buildup has been recognized in
these systems. Gypsum scaling is caused by the reaction of soluble calcium ion
with sulfate ion formed in the system through oxidation of the absorbed S03
or from absorbed $03 according to the reaction:
Ca++ + $04= + 2H20 = CaS04 -2H20 (1)
In dilute systems, gypsum scaling is controlled by softening the1 regenerated
liquor prior to recycling to the scrubber. In concentrated systems, gypsum
is not a problem since the high sulfite concentration keeps the Ca ion concen-
trations low. Softening ensures that the liquor recycled to the scrubber system
is unsaturated with respect to gypsum; therefore, with proper softening even if
some sulfate is formed in the scrubber, the liquor will not be saturated with
gypsum and cause scaling on the inside surfaces of the scrubber. In concentrated
active alkali systems, a special softening step is not necessary since high sulfite
concentration is maintained throughout the system. This sulfite maintains a low
Ca++ ion concentration (sulfite softening), and thus maintains the scrubbing
solution unsaturated with respect to gypsum.
Based on experience gained in lime/limestone scrubbing testing, a certain factor
of safety in the prevention of gypsum scaling probably exists in dual alkali
systems. Gypsum has been found to supersaturate easily to about 130% saturation.
Thus, scaling will occur only if supersaturation is excessive.
Carbonate scaling usually occurs as a result of localized high pH scrubbing li-
quor in the scrubber where C02 can be absorbed from the flue gas to produce COo
949
-------
ions. These ions subsequently react with dissolved calcium to precipitate cal
cium carbonate scale according to the following series of reactions:
Carbon dioxide absorption by high pH liquor:
C02 + 2 OH~ = C03= + H20 (2)
Calcium carbonate scaling:
+ Ca++ = CaC03 (3)
Based on experience with the General Motors, full-scale, dual alkali system,
carbonate scaling could occur with scrubber liquor pH above 9. At lower pH,
the carbonate/bicarbonate equilibrium system tends to limit the free carbonate
ion and thus prevent precipitation of calcium carbonate:
Hf + CO = = HCO - (4)
Thus, carbonate scaling can be eliminated by control of pH in the scrubber.
Solids Quality - Under certain conditions, the waste solids produced in the
regeneration sections of various dual alkali systems have a tendency not to
settle from the scrubber liquors. This creates problems in the operation of
settlers, clarifiers, reactor clarifiers, filters, and centrifuges. Although
observed in the laboratory testing conducted by the EPA on dilute systems and
in the laboratory and pilot plant work conducted by ADL on dilute and concen-
trated systems, this phenomenon is not completely understood, but is thought
to be a function of reactor kinetics. (14)
Some of the factors thus far identified which appear to affect the solids set-
tling properties are reactor configuration, concentration of soluble sulfate,
concentration of soluble magnesium and iron in the liquor, concentration of sus-
pended solids in the reaction zones, and use of lime vs. limestone for reaction.
Based on laboratory work in dilute systems (about 0.1 M active sodium) using
limestone, it appears that solids settling characteristics degraded significantly
at soluble sulfate levels above 0.5 M. Based on laboratory work with concentra-
ted systems (about 0.45 M TOS, 5.4 pH, 0.6 M sulfate) using limestone, marked
degradation of solids settling properties occurred at a magnesium level of 120
ppm and virtually no settling of solids occurred at the 2000 ppm magnesium level.
Equal degradation of solids settling properties also occurred in concentrated
systems when the sulfate level was raised to 1.0 M while maintaining low magne-
sium level (about 20' ppm) and keeping other variables constant. (14)
Envirotech'4) advocates the recycle of precipitated solids from the thickener
underflow to the reaction zones in an effort to grow crystals which settle
faster and are more easily filtered.
950
-------
ADL cites reactor configuration as being important in the production of solids
with good settling and filtration characteristics. (14) Their basis for this is
comparative tests of a simple continuous flow stirred tank reactor (CFSTR)
with the ADL/Combustion Equipment Associates (ADL/CEA) designed reactor system
under similar conditions. The ADL/CEA reactor system appeared to give better
settling solids over a greater range of conditions than a simple CFSTR. The
ADL/CEA reactor system consists of a low residence unit followed by a high re-
sidence unit.
Sulfate Removal - In dual alkali systems, some of the sulfur removed from
the flue gas takes the form of soluble sodium sulfate due to oxidation in the
system, thus changing some of the active sodium to the inactive variety. When
sodium in the system is converted to the inactive form (^SCty), it is relatively
difficult to convert back to active sodium. To convert inactive sodium to active
sodium, sulfate ion must be removed from the system in some manner, while leaving
the sodium in solution. The alternative to this is to remove the sodium sulfate
from the system at the rate it is being formed in the system. This alternative
is not desirable since it wastes sodium and generally is carried out by allow-
ing the sodium sulfate to be purged from the system in the liquor which is oc-
cluded in the wet solid waste product. (14) The solid waste product can then po-
tentially contribute to water pollution by leaching. Water runoff can contaminate
surface water, while leaching and percolation of the leachate into the soil can
contaminate ground water near the the disposal site. Failure to allow for sulfate
removal from dual alkali systems will ultimately result in-a) precipitation of
sodium sulfate somewhere in the system if active sodium is made up to the system,
or b) in the absence of makeup, eventual deterioration of the SOg removal capabil-
ity due to the loss of active sodium from the system.
Equations (15), (16), (17),and (18), shown previously under the definition of
"sulfate removal" in Appendix A, describe several sulfate removal techniques
which have been used in FGD system pilot tests.
The first equation depicts the sulfate removal technique used in dilute active
alkali systems:
N32S04 + Ca(OH)2 + 2^0 = 2NaOH + CaSC>4 -2^0 (15)
(gypsum)
Concerning the full-scale dilute alkali system installed and operating at the
Parma, Ohio, transmission plant of General Motors, and dilute systems in general,
Phillips(19) stated:
"The presence of Na^Stty in the scrubber effluent is the prime factor
influencing the design of the regeneration system. NaoSO/j is not easily
regenerable into NaOH using lime, the reason being that the product,
gypsum, is relatively soluble . . . Na2S04 cannot be causticized in the
presence of appreciable amounts of $03" or OH- because Ca++ levels are
held below the CaS04 solubility product. To provide for sulfate causti-
cization, the system must be operated at dilute OH" concentrations below
951
-------
0.14 molar. At the same time, SO/5 levels must be maintained in the
system at sufficient levels to effect gygsum precipitation ... We
selected 0.1 molar OH~ and 0.5 molar 504 as design criteria."
In a previous paper, Phillips^) showed a plot of equilibrium caustic formation
in Ca(OH)2-Na2S04 solutions at 120°F which is the basis for selection of the
design criteria. The essence of this discussion is that, if the active sodium
concentration is sufficiently dilute, sulfate can be removed from the system
by simple precipitation as gypsum by reaction of lime with sodium sulfate.
Since, as explained above, this reaction will not proceed to a great extent
in concentrated active alkali systems, other .techniques must be employed to
effect sulfate removal in these systems.
The second equation depicts a technique which is used in the full-scale dual
alkali systems in Japan, and which has been pi lot -tested by ADL under contract
with EPA:
Na2S04 + 2Ca SOs -1/2 H20 + ^4 + 3H20 = 2NaHSOs + 2CaS04'2H20 (16)
(gypsum)
This technique is used to precipitate gypsum by dissolving calcium sulfite in
acidic solution, thus increasing the Ca*+ in solution enough to exceed the solu-
bility product of gypsum. Ideally according to equation (16) 2 moles of gypsum
should be precipitated for each mole of sulfuric acid added. In practice, how-
ever, this is not the case since any material which functions as a base can
consume sulfuric acid and reduce the efficiency of this reaction for its intended
purpose. (14) Unreacted lime or limestone, sulfite ion, and even sulfate ion can
consume sulfuric acid, thus lowering sulfate removal from the system.
Conceivably, this method of sulfate removal may be economically unattractive in
applications with very high oxidation rates, and where the gypsum produced must
be discarded. The economic picture is considerably changed where this system
is used merely as a slipstream treatment to supplement other sulfate removal
methods and/or where the solid product gypsum is saleable as is the case in
Japan.
The third equation describes a phenomenon which has been referred to as mixed
crystal or solid solution formation:
x NaHS03 + y Na2S04 + (x+y) Ca(OH)2 + (z-x) H20 = (17)
(x+2y) NaOH + x CaS03 -y CaS04 - z H20
(mixed crystal or solid solution)
This phenomenon is described by EPA/IERL-RTP's R.H. Borgwardt(^) as it applies
to lime/limestone wet scrubbing based on pilot plant investigations. A similar
phenomenon has been observed by ADL in some of their early pilot testing of dual
alkali systems in conjunction with CEA, and later in the EPA/ADL dual alkali
test program.
952
-------
Under certain conditions the solids precipitated in lime/limestone and dual
alkali systems contain sulfate, sulfite, and calcium; however, the liquor from
which these solids precipitate appears to be subsaturated with respect to gypsum.
This is based on the fact that pure gypsum crystal could be dissolved in the
mother liquor from which the mixed crystal/solid solution was precipitated. In
addition, the solid material, examined by X-ray diffraction contained no gypsum;
infrared analysis confirmed the presence of sulfate.
Borgwardt found that the molar ratio of sulfate to sulfite in these solids was
primarily a direct function of sulfate ion activity in the mother liquor. In
pilot test work with lime/limestone scrubbing, with little or no chlorides
present and normal magnesium level (below 1000 ppm) in solution, the sulfate to
sulfite molar ratio in the mixed crystal solids reached a maximum level of 0.23.
This is equivalent to a [SO^/total [SOX] ratio in the solids of 0.19.
In pilot test work with concentrated dual alkali systems, ADL observed the
simultaneous precipitation of sulfate and sulfite with calcium in lime and
limestone treatment of concentrated dual alkali scrubbing liquors. This
phenomenon was surprising at first, in light of the reasoning which led to the
development of dilute dual alkali systems; i.e., gypsum cannot be precipitated
from solutions containing high active alkali concentrations. It was a simple
technique for sulfate removal in concentrated systems. The [SO^]/ total [SOX]
ratio observed in pilot dual alkali work was as high as 0.20.(l?j Coincidentally,
this was the same value observed by Borgwardt in lime/limestone testing. This
led to the belief that the same phenomenon was occurring in both processes.
The mother liquor from which these solids were precipitated was also found to
be subsaturated in gypsum, and when the solids were examined, pure gypsum was
not found.
Based on the observed data, it appears reasonable to design a concentrated
active alkali system for a particular situation in which the system oxidation
rate is below about 20%.U4) in this case, sulfate can be removed at the desired
rate, without purging Na2S04 or supplementing the system with other complex
methods of sulfate removal.
The fourth equation shows sulfate removal as sulfuric acid in an electrolytic
cell:
Electrolytic cell
Na2S04 + 3H20 > 2NaOH + H2S04 + H2 + 1/2 02 (18)
This method is the basis for operation of the Stone and Webster/Ionics process
sulfate removal technique. In Japan, Kureha/Kawasaki has pi lot-tested the
Yuasa/Ionics electrolytic process for sulfate removal in conjunction with their
dual alkali process. They claim that this process is less expensive overall
than the presently used sulfuric acid addition method. In addition, they claim
that sodium losses from the system can be cut in half through the use of this
method: from 0.018 moles Na loss/mole S02 absorbed to 0.009 moles Na loss/mole
S02 absorbed.(24)
953
-------
Another approach to sulfate control is to limit oxidation. With sufficient
limitation of oxidation, by process and equipment design, it may be possible
to control sulfate by a small unavoidable purge of Na2S04 with the solid waste
product. To design for minimum oxidation, there should be minimum residence
times in equipment where the scrubber liquor is in contact with oxygen-contain-
ing flue gas; all reactors, mixers, and solids separation equipment should
be designed to minimize absorption of oxygen from air. In addition, it has been
reported(14) that oxidation of scrubber liquors can be minimized by maintain-
ing very high ionic strength. One possible explanation for this is that high
ionic strength liquors are poor oxygen absorbers and that oxidation in these
systems is oxygen absorption rate limited.
Water Balance and Waste Product (Cake) Washing - To operate a closed system
to avoid potential water pollution problems, system water balance is a primary
concern. Water cannot be added to the system at a rate greater than normal
water losses from the system.
Generally fresh water is added to a D/A system to serve many purposes, including:
t Saturation of flue gas.
• Pump seal.
r Mist eliminator washing.
• Slurry makeup.
t Waste product washing.
• Tank evaporation.
On the other hand, water should only leave the system through:
• Evaporation by the hot flue gas.
• Water occluded with solid waste product.
• Water of crystallization in solid waste product.
Careful water management, part of which is the use of recycled rather than
fresh water wherever possible, is necessary in order to operate a closed system.
As previously indicated, disposal of wet solid waste containing soluble salts
is ecologically undesirable. In addition, allowing active alkali or sodium
salts to escape from the system is an operating cost factor. Sodium is usually
made up to dual alkali systems by adding soda ash (recently quoted at $60 per ton
f.o.b. source) at some point in the system. Thus, both ecological and economic
considerations dictate that waste product washing is desirable.
954
-------
A rotary drum filter, belt filter, or centrifuge is usually where the final
solids separation is made. This equipment can be designed for solids washing
with fresh water.
One concern in waste product washing is the extent to which the cake should be
washed, because of the effect it has on the concentration of solubles in the
waste. Solubles in the waste of any FGD system (i.e., lime/limestone and dual
alkali) have the potential to contaminate ground water.
One of the major goals of the Resource Conservation and Recovery Act (RCRA)~
which requires Federal regulations for disposal of hazardous waste and guidelines
(for use by the state) for disposal of non-hazardous waste—is to prevent ground
water contamination.
Efforts are currently underway to develop design standards for disposal of utility
wastes (i.e., coal ash and scrubber waste) which can be incorporated into RCRA
regulations and guidelines.
Although it will be sometime before these standards are developed, an FGD system
operator must take this potential for ground water contamination into account in
designing his disposal site.
One obvious consideration in waste product washing is system water balance.
Unlimited waste product washing is not possible if a closed system operation
with no liquid stream discharge is a goal. Another more subtle reason for limit-
ing waste product washing is the potential problem of non-sulfur/calcium solubles
buildup in the system. These non-sulfur/calcium solubles enter the system with
the fly ash, flue gas, and lime and/or limestone and the makeup water. Of these,
probably the soluble material in highest concentration would be sodium chloride
which results from the absorption of HC1 from the flue gas by the scrubber solu-
tion. A material balance around the system at steady state necessitates that
solubles leave the system at the rate they enter. Thus, depending upon how well
the waste product is washed, a certain level of non-sulfur solubles will be es-
tablished in the system. Since the only mechanism for these solids to leave the
system is as part of the wet solid waste, a certain purge is necessary. This purge
also necessitates the loss of some sodium from the system. Practical limitations
in filter design and water balance probably would limit a system to two or three^.
"displacement washes" of the waste product (one displacement wash means washing
with an amount of fresh water equivalent to the amount >of water contained in the
finalwet waste product per unit of waste). Depending on the characteristics of
the waste product and the design of the washing system, one displacement wash
can reduce the solubles content of the waste product by as much as 80%.
B.3 ENVIRONMENTALLY ACCEPTABLE SOLID WASTE
Dual alkali systems should be designed to produce an environmentally accept-
able end product. Desirable solid waste product properties include:
• Non-toxic.
• Low soluble solids content.
955
-------
t Low moisture content.
• Non-thixotropic.
• High compressive or bearing strength.
Gypsum vs. Calcium Sulfite - One of the options available to some dual
alkali processes is whether or not to oxidize the solid waste. Advantages
accruing from oxidation include: gypsum has better handling properties than
calcium sulfite because sludges containing a high ratio of gypsum to calcium
sulfite are less thixotropic, faster settling, more easily filtered, and can
be more completely dewatered than sludges containing a high proportion of cal-
cium sulfite. Another important characteristic which has been attributed to
high gypsum (as opposed to calcium sulfite) sludges is their higher compressive
or bearing strength.
An explanation for the behavior of high sulfite sludges is given by Selmeczi
and Knight.'2°) Although filter cakes appear dry, they still contain a consid-
erable amount of water and thus, upon vibration or application of stress, have
a tendency to again become fluid. This thixotropic property and high moisture
content are both explained by the morphology of calcium sulfite clusters. Be-
cause of the highly open, porous, or sponge-like nature of these clusters, a
considerable amount of water is retained in the clusters. The calcium sulfite
crystals are rather fragile and break under pressure, releasing some of the
water, which result in fluid sludges.
In Japan, where by-product gypsum is saleable, the calcium sulfite solids pro-
duced are oxidized completely to gypsum in a separate oxidation process tacked
on to the tail end of the system. In applications where high excess combustion
air is present, where low-sulfur coal is burned, or a combination of these
conditions, the oxidation rate in the system tends to be high (possibly about
90%) and the proportion of gypsum in the sludge tends to be high. In some dilute
systems, the proportion of gypsum in the sludge can be increased by augmental
aeration of the scrubbing liquor.v8/ Crystal seeding techniques used in conjunc-
tion with augmental aeration can produce relatively coarse grained gypsum crystals
with good dewatering and structural properties in the final waste product.
Sludge Fixation Technology - Chemical or physical fixation of the sludge
produced in a dual alkali system is another potentially important means of
producing an environmentally acceptable solid waste product. This technology
is commercially offered by I.U. Conversion Systems, Inc., Dravo Corporation,
and Chemfix Corporation. Most of their efforts are concentrated on sludge
produced from the more prevalent lime/limestone systems; however, there has
been some evaluation of dual alkali sludges. The objective of sludge fixation
technology is the production of a non-toxic, unleachable solid waste product which
has reasonably high load bearing strength. If dual alkali sludges are amenable
to this type of treatment, the need to reduce soluble sulfates in the solid waste
product is mitigated. Some sodium sulfate has been found to be physically or
chemically tied up in the solid calcium sulfate/sulfite crystal lattice;U4)
however, the extent of this phenomenon is not generally considered to be adequate
to remove all of the sodium sulfate produced by oxidation. Sludge fixation
technology may be a mechanism by which additional sodium sulfate can be removed
956
-------
from the system without adverse environmental effects. There is some concern
that this is not viable, however, since sludge fixation chemistry involves
pozzolanic reactions between calcium compounds and fly ash components in the sludge
which may only involve rnultivalent ions rather than monovalent sodium. In other
words, monovalent ions such as sodium may either a) not take part in the pozzolanic
reactions, or b) inhibit or limit such reactions. Further investigation is needed
in this area.
B.4 ECONOMICS
Dilute vs. Concentrated System - The selection of a dilute or concentrated
dual alkali system is an important consideration in any application. In gen-
eral, concentrated systems are desirable for applications where oxidation is ex-
pected to be relatively low. Conversely, dilute systems are suitable for appli-
cations where oxidation rates are high. For high-sulfur Eastern coal applica-
tions on utility boilers (where excess air is controlled carefully and maintained
at the lowest value consistent with complete combustion), concentrated systems
are favored. On the other hand, in utility or industrial boiler applications
(where Western low-sulfur coal is burned, and/or where control of oxidation is
difficult due to high excess air), dilute systems may be more suitable.
Oxidation rate is promoted when low-sulfur coal is burned, since the ratio of
oxygen to S02 in the flue gas is higher than in high-sulfur coal applications.
Since oxidation is a strong function of the rate of absorption of oxygen, liquor
which is dilute in TOS is subject to having a greater proportion of these species
oxidized by a given amount of absorbed oxygen than one in which the TOS is more
concentrated.
Under a given set of conditions, without consideration of waste disposal, a con-
centrated system can be installed at lower capital cost than a dilute system
as previously discussed; however, the desirability to produce a manageable
solid waste (dilute systems can be designed to produce high gypsum sludges)
could, in some cases, override the capital cost issue.
Cost Data - Based on available information, dual alkali systems are economically
competitive with the "first generation" wet lime/limestone slurry scrubbing
systems. This is particularly true in cases where the lime or limestone system
would be required to be equipped with solids separation equipment (i.e., thickener
and filter).
Factors that allow a dual alkali system to be less expensive than a lime/limestone
system, both in initial cost and annualized operating cost, include:
• Lower scrubber liquid/gas ratio (L/G).
• Lower scrubber pressure drop (AP).
• Simpler scrubber design:
- fewer stages.
957
-------
- no slurry in scrubber.
t Less exotic materials of construction.
• Solid waste with better handling properties.
It is estimated'29' that a new (or simple retrofit) dual alkali system could be
installed at a 500 MW plant for a cost around $100/kW and operated at 4.19
mills/kWhr excluding sludge disposal (which would vary with the particular
application). These estimates are based on the following:
• 500 MW or larger system.
• 3.5 wt.% sulfur coal.
• 80% load factor.
• Capital charges @ 15-16% of capital investment, annually.
• Maintenance @ 3-4% of capital investment, annually.
• Operation with two operators/shift.
• Power @ $0.03/kWhr.
• Soda ash @ $90/ton (delivered).
• Lime @ $42/ton (delivered).
958
-------
4W
THE FGD REAGENT DILEMMA
LIME, LIMESTONE, OR THIOSORBIC LIME
by
(,
Donald H. Stowe
David S. Henzel
David C. Hoffman
Dravo Lime Company
Pittsburgh, Pa.
959
-------
Introduction
The Clean Air Act of 1970 forced the utility industry into
the chemical processing field because of its restrictions on
sulfur dioxide (SO2> emissions. Chemical reactions are required
to remove the S02 from the combustion of coal. In the early
seventies, SO_ compliance could also be achieved by burning
"compliance" coal. This alternative has rapidly disappeared as
compliance coal is becoming scarce and expensive. In 1977, the
Clean Air Act Amendments were proposed and, if enacted, will make
scrubbing an integral part of all new coal-fired electrical
power plants.
Since SO- removal is a major expense, utilities will attempt
to obtain SO- compliance at the lowest possible cost; including
reliability and maintenance. In order to achieve this, one must
realize that these interrelated unit processes—boiler operation,
scrubbing, and waste disposal—effect SO2 removal. Central to
the optimization of these three processes is the selection of an
alkaline reagent for S02 removal. The most prevalent reagents
used today are limestone, lime, and Thiosorbic® lime. This
discussion will briefly review the basic chemistry involved with
each reagent use, outline major advantages and disadvantages of
each, and present a realistic example of the cost differences
between limestone, lime, and Thiosorbic lime scrubbing.
- Registered trademark of Dravo Corporation.
960
-------
Discussion
Generally, scrubbing is usually referred to as the removal
of sulfur dioxide CS02) from the flue gas of a utility boiler.
The source of the S02 is the sulfur content in the combustion
fuel, most commonly bituminous coal. Current combustion coals
may be ranked as low sulfur (<1%), medium sulfur (1-3%), and
high sulfur (>3%) . Presently, the most generally referred to
guideline for S02 control is the Federal New Source Performance
Standard (NSPS) calling for a maximum 1.2 Ib. of S02 emission per
million BTU's of heat input. Frequently, state and local regu-
lations are even more stringent.
As near term alternatives to FGD, a station may possibly
meet regulations by the following:
1. Switching to low sulfur fuel (compliance coal).
2. "Deep" cleaning of the coal prior to combustion.
3. Solvent refining of the coal prior to combustion.
4. Gasification or liquefaction of coal and combustion
of resultant gas.
5. Fluidized bed combustion of coal (new boiler design).
Items 3 through 5 have not been demonstrated on a commercial
scale although extensive development work is being performed.
Of the remaining two alternatives to FGD, fuel switching is the
predominant alternative. However, the use of "compliance" coal
is being restricted by the following:
1. Availability of product - current sources are
the West, Northwest, and sections of Appalachia.
961
-------
2. Costs associated with production, transportation,
and boiler modifications.
3. State regulations prohibiting use of imported
coal (i.e., State of Ohio).
In summary, if all the factors are just right, burning of
compliance coal may be cheaper than scrubbing. However, this
is generally not the case. It has been estimated that compliance
coal will be available until 1985 and after that, only high
sulfur coal will be obtainable. For the scrubbing industry with
a construction lead time of about 3 years, and longer times for
new mine-prep plant development; the effect of new, low sulfur,
coal sources on the utility market is nearly exhausted if a
utility does not already have commitments for the coal. Deep
cleaning of coal may be the only near term alternate to FGD.
However, the amenability of coals to this process is limited, due
primarily to the organic sulfur content of the coal. Consequently,
FGD, as we currently know it, will probably be the most widely
used technology for the next 10 years.
The non-regenerative lime/limestone processes are the most
commonly applied commercial FGD systems in the United States.
Primarily, this is due to the lower costs of the system, avail-
ability of reagents, ease of operation, system reliability, and
experience. Table 1 presents a summary of the total operational
commercial FGD systems, as of July, 1978. Of the total 14,420
megawatts, 13,496 megawatts (94%) represent wet calcium-based
lime/limestone processes.
962
-------
Table 1
VO
O\
oo
Operational FGD Systems
By Chemical Process - July, 1978
Process
Lime/Limestone
Thiosorbic Lime
Lime
Carbide Lime
Lime/Alkaline Fly Ash
Limestone
Limestone/Alkaline Fly Ash
Subtotal
Other
Magnesium Oxide
Sodium Carbonate
Wellman Lord
Subtotal
Total
No. of
Units
6
4
4
3
13
2
32
1
3
2
a
38
Total Months
MW Experience
3,370
679
851
1,170
6,006
1,420
13,496
120
375
429
924
14,420
151
175
93
65
383
43
910
34
126
23
183
1,093
Source:
B. Laseke, et.al.,EPA Utility FGD Survey: June-July 1978
EPA 600/7-78-051 d
-------
Of the total lime/limestone units currently operating,
only seven have design removal efficiencies of greater than or
equal to 85 percent; the proposed new federal regulation. A
breakdown of these units by process is shown in Table 2. Based
on this 85% removal criterion, the greatest operational experience
lies with Thiosorbic lime.
Process Chemistry
As previously mentioned one of the primary reasons for the
dominance of lime/limestone scrubbers is the relative simplicity
of the chemical reactions involved. The aqueous sulfur dioxide
reacts with the more basic anionic species to produce bisulfite.
The predominant, cation is calcium. Those basic reactions which
occur in these types of systems are listed in Table 3.
Although the lime/limestone processes appear relatively
simple, their capabilities are limited by the solubility of the
basic anions involved. As a result of this, these processes
exhibit two disadvantages: 1) the need for high mechanical energy,
and 2) their high potential for gypsum scaling. These systems
are basically liquid film limited; that is insufficient dissolved
species are present to neutralize the absorbed S0_. Therefore,
the amount of S02 removed is a function of the rate at which the
basic species go into solution. Since sufficient dissolved alka-
linity is not present, the mechanical energy must be increased to
bring more liquid into contact with the gas stream.
964
-------
Table 2
Operational Lime/Limestone FGD Systems
Designed for ^ 85% Removal - July 1978
Design Design Months
SO2 Removal MW Coal % S Experience
Thiosorbic
Mansfield 1
Mansfield 2
Conesville 5
Conesville 6
Carbide Lime
Cane Run 4
Cane Run 5
Limestone
Duck Creek
92.1
92.1
89.5
89.5
85.0
85.0
85.0
825
825
400
400
2,450
178
183
361
400
400
4.7
4.7
4.7
4.7"
3.75
3.75
2.75
27
12
18
1
58
23
7
30
0
0
Source:
B. Laseke, et.al.,EPA Utility FGD Survey: June-July 1978
EPA 600/7-78-051 d
-------
Table 3
Basic Reactions
Lime/Limestone Systems
ON
1. OH" + SO2^HSO3
2. SO3= + SO2 + H2O
3. HCCV + SO2^HSO3" + CO2
4. CO= + 2SO, + H,O^2HSO " + CO
-------
Calculations have been performed to determine the
theoretically required dissolved alkalinity to achieve 85
percent removal as a function of liquid to gas ratio and inlet
sulfur dioxide concentration. The results are tabulated in
Table 4. It should be emphasized that these values do not
include a factor for the mass transfer capabilities of the
scrubber. Typical dissolved alkalinity values obtainable with
conventional lime/limestone systems are in the order of 100 to
300 mg/1, as CaCO_. Referring to Table 4, it can be seen that
these low alkalinities are adequate only at low S0_ concentrations
and/or high liquid to gas ratios.
The introduction of magnesium in the scrubbing loop produces
some dramatic changes in the dissolved liquor composition. The
solubility of magnesium sulfite is more than two orders of
magnitude greater than that of calcium sulfite. As a result of
this, an increase in dissolved magnesium produces an increase in,
sulfite concentration, thus dissolved alkalinity. The final result,
of course, is higher removal efficiency. Dissolved sulfite con-
centrations are generally increased to greater than 2000 mg/1 as
compared to the typical range of 100 to 200 mg/1 obtainable in a
conventional lime/limestone system. Alkalinities increase from less
than 200 mg/1 to greater than 1000 mg/1. Tests have shown that the
Thiosorbic process is doing the work of 2.0 to 2.5 identical calcium-
based scrubbers. Because of this, it is not only practical to
obtain high removal efficiencies (>90%) on high sulfur coal, but
it also reduces the overall mass transfer characteristics required
of the system. Figure 1 exemplifies the net effect of dissolved
magnesium upon SG>2 removal.
967
-------
Table 4
Theoretically Required Dissolved Alkalinity
To Achieve 85 Percent Removal
Inlet SOz (ppm)
oo
L/G
30
40
50
60
70
80
90
100
1,000
473
354
284
236
203
178
158
142
2,000
946
710
568
473
405
354
315
284
3,000
1,420
1,064
852
710
609
532
473
426
4,000
1,892
1,420
1,136
946
811
710
631
568
Legend:
L/G - Gallons of Scrubbing Liquor per 1,000 SCF
-------
Figure 1
Effect of Magnesium
Upon Removal Efficiency
vd
II
CO
O
(O
Mg
SO
Alkalinity
-------
The second key characteristic of magnesium-enhanced
scrubbing is the ability to operate subsaturated with respect
to gypsum. All lime/limestone processes are based upon operation
in a mode saturated for calcium sulfite. With increasing sulfite
concentration, the calcium concentration is thereby suppressed.
Typical calcium values of greater than 1000 mg/1 are readily de-
creased to less than 100 mg/1. This, of course, reduces the
relative saturation with respect to gypsum even though dissolved
sulfate concentration increases. This overall effect of magnesium
upon gypsum saturation is shown graphically in Figure 2.
In addition to the work performed by Dravo Lime, magnesium-
enhanced scrubbing has been extensively studied by the EPA at the
Shawnee test facility. Empirical equations have been generated to
calculate removal efficiency in a..spray tower and relative gypsum
saturation as a function of liquor composition. Their equations,
along with experimental data have been used to generate the data
presented in Table 5. From the table, two trends are evident:
1) as magnesium increases, relative saturation decreases, and
2) as magnesium increases, removal efficiency increases.
Pilot Plant Demonstrations
Several direct comparisons have been made between lime, lime-
stone, and Thiosorbic lime in pilot plant facilities. Results of
one series of tests are shown in Table 6. The 50 cfm pilot,
a low energy venturi followed by a spray tower, had a mix tank
residence time of 1% minutes during the lime and Thiosorbic lime
970
-------
Figure 2
Effect of Magnesium
Upon Gypsum Saturation
10
II
CO
o
Mg
CO
O
c
o
SO
Ca
-------
Table 5
The Effect of Magnesium Concentration Upon Relative Gypsum
Saturation and SO 2 Removal Efficiency (Low Energy
Spray Tower)
B
Liquor Composition
PH
C1~
Mg^
SO3=
L°^
% Saturation
7.0
1,000
660
91
4,704
920
154
7.0
1,000
1,180
134
5,701
500
84
7.0 7.0 7.0 7.0
1,000 1,000 1,000 1,000
1,940 3,200 3,960 4,760
285 1,604 2,010 2,908
7,925 10,867 13,383 15,443
250 60 60 50
42 10 9 8
Removai Efficiency
L/G = 40
50
60
70
80
90
100
50.2
58.2
64.9
70.5
75.2
79.2
82.5
55.4
63.5
70.2
75.7
80.1
83.7
86.7
63.2
71.3
77.6
82.6
86.4
89.4
91.8
75.9
83.1
88.1
91.7
94.2
95.9
97.1
82.8
88.9
92.8
95.4
97.0
98.1
98.8
88.9
93.6
96.3
97.9
98.8
99.3
99.6
-------
Table 6
50 CFM Venturi, Spray Tower Comparison of
Limestone, Lime and Thiosorbic Lime
Test No.
CO
Reagent SO2 Inlet L/G Total Removal Eff.
1
2
3
4
5
6
7
8
9
Limestone
Limestone
Limestone
Lime
Lime
Thiosorbic
Thiosorbic
Thiosorbic
Thiosorbic
1,500
1,500
1,500
3,000
3,000
3,000
3,000
3,000
3,000
50
60
90
60
80
36
60
82
90
59%
65%
78%
53%
67%
57%
82%
90%
93%
-------
tests and 16 minutes for the limestone test. It should be
noted the removal efficiencies obtained are somewhat lower
than typical values obtained today; however, the trend of in-
creased removal efficiency and lower L/G requirement is very
evident.
A two stage 1500 cfm venturi was later operated with an
inlet S02 concentration of approximately 3000 ppm SO . Tests
were conducted with both high calcium lime and magnesium-enriched
lime. Results in Table 7 show the increased alkalinity and higher
removal efficiency. If we compare theoretical transfer units, it
is evident the Thiosrbic lime was doing the equivalent work of
approximately two scrubbers in comparison to the high calcium
lime scrubber.
Table 8 lists results of tests conducted on a 750 cfm
turbulent contact absorber at 3000 ppm inlet. SO_ concentration.
Once again, the superiority of the magnesium-enriched system is
very evident.
Extensive testing was also performed on a full scale FGD
system consisting of single stage Venturis. On high calcium lime,
typical removal efficiencies ranged from 50 to 65 percent with
severe internal scaling, so much in fact that the throat dampers
could not be operated. Thiosorbic lime operation not only
increased the removal efficiency to greater than 80 percent; but
also eliminated the scaling problems; so much in fact that most
scale present prior to the Thiosorbic operation actually dissolved,
974
-------
Table 7
1500 CFM Venturi Comparison of
Lime and Thiosorbic Lime
L/G
Per Alkalinity
Srage High Ca
1 40/40100-140
240/40 -
3 50/40 100-140
450/50100-140
5 40/--
6 50/--
PPM
% Removed
Thio. High Ca
600
1,100
600
600
800
800
70
-
82
88
-
-
Thio.
92
96
97.2
99.6
80
87
Transfer
High Ca
1.20
-
1.71
2.12
-
-
Units ^
Thio.
2.53
3.22
3.58
5.52
1.61
2.04
it
Tft*iM /KJ
Thio./Nt£8
2.1
-
2.1
2.6
-
-
Note:
Transfer Units = ISL =1n [inlet SO2 ]
[outlet SO2]
-------
Table 8
750 CFM TCA Comparison of Lime and
Thiosorbic Lime
Test
Condi-
tion Stage
A
B
C
D
1
1
1
1
1
1
1
2
2
2
2
3
3
3
3
L/G
20
40
60
30
40
50
60
30
40
50
60
20
30
40
50
SO2
Inlet
1,315
1,351
1,340
3,150
3,100
3,100
3,050
2,850
2,850
2,900
2,800
3,050
3,100
3,100
3,100
Removal
Efficiency %
High Ca
60
86
94
45
59
72
60
70
79
90
60
76
85
90
Note:
Transfer Units— Nj = 1n
Thio.
81
84
86
90
94
97
i [inlet
Number of N,
Transfer Units Thio./N,
High Ca
.92
1.97
2.81
.60
.89
1.27
.92
1.20
1.56
2.30
0.92
1.43
1.90
2.30
S02]
Thio. *Ca
1.66 2.8
1.83
1.97 2.2
2.30 2.5
2.81 2.3
3.51 2.3
[outlet SO2]
-------
Perhaps the most exemplary comparison of magnesium-
enriched scrubbing versus high calcium has been performed by EPA
at the Shawnee test facility. Some average liquor compositions
are listed in Table 9. The liquor composition of run 619-1A,
with high calcium lime, contains a very low sulfite concentration
of 40 ppm. A sulfite level of this nature will result in two
characteristics, low removal efficiency and gypsum saturated
operation. On the other hand, run 611-1A with magnesium-promoted
lime and approximately 3200 ppm dissolved magnesium; a sulfite
concentration of about 500 ppm results. The higher sulfite
concentration not only increased dissolved alkalinity and removal
efficiency but at the same time, suppressed the calcium concen-
tration from 1860 ppm to 320 ppm. Tests results show the addition
of magnesium not only decreased the relative gypsum saturation
from 125 to 45 percent; but at the same time increased removal
efficiency from 84 to 95 percent. Figures 3, 4 and 5 are super-
imposed test results for the two runs. Advantages of the magnesium
are very evident.
Full Scale Operations
Six full scale units are currently operating with the Thiosorbic
lime process. These units include Mansfield 1 and 2, (1650MW);
Conesville5 and 6, (800MW); Elrama Station CSlOMWl; and Phillips
Station (410MW). Sulfur dioxide emission limitsareO.6 Ibs/MBTU
with the exception of the two units at Conesville which have a
limit of 1.2 Ibs/MBTU. S'ulfur content of the coals range from
2.0 to 4.7 percent, by weight. All of the units currently operating
have passed compliance tests, with the exception of Conesville #6.
One additional unit, Pleasants #1 (625MW), is in the start-up phase.
977
-------
Table 9
Shawnee Test Results
Average Liquor Composition
ID
~J
oo
Run
pH
Ca"""
Mg
SO3=
so4-
cr
% Sulfate Saturation
Lime
61 9-1 A
7.8
1,860
330
40
1,830
2,920
125
Lime + MgO
611-1A
7.0
320
3,200
510
11,000
2,500
45
-------
Figure 3
Comparison of High-Calcium Lime
and Magnesium-Containing Lime
(0
Q 1,500
J 1,000
> 500
I 0
I I I I I I T I 1 I 1
0 40 80 120 160 200 240 280 320 360 400 440
Time, Hr.
- High-Calcium Lime
- Magnesium-Containing Lime
979
-------
Figure 4
Comparison of High-Calcium Lime and
Magnesium-Containing Lime
CM
vo
oo
o
is
Id
UJ"
<
1.0
0.8
0.6
0.4
Time, Hr. 0 40
I i i I i i I i i
80 120 160 200 240 280 320 360 400
-e-
High-Calcium Lime
Magnesium-Containing Lime
-------
Figure 5
Comparison of High-Calcium Lime and
Magnesium-Containing Lime
VO
oo
Q.
8
7
6
5
4
Time, Hr. 0
i I I i I r I I i i
40 80 120 160 200 240 280 320 360 400
High-Calcium Lime
-e— Magnesium-Containing Lime
-e-
-------
System Economics
It is extremely difficult to make a direct economic
comparison of Thiosorbic lime, lime and limestone scrubbing
systems. Perhaps the best available comparison is the work
performed by PEDCo under EPA contract. Their work consists
of a survey of currently operating FGD systems with adjustments
made, thus enabling a comparison on a common basis.
The only systems designed specifically to date for the
Thiosorbic scrubbing process have been Pleasants 1 and 2.
Unfortunately, since the units are not yet operational, operating
costs are not known. All other units currently using our process
were designed for high calcium lime. This discussion will there-
fore compare the lime and limestone systems, keeping in mind that
that the bulk of the lime systems operate with Thiosorbic lime.
The work performed by PEDCo compares capital costs on a $/kW basis
and operating costs on a mills/kWh basis. The fallacy of this
approach is that significant contributing factors such as the
design sulfur content of the coal and the required removal
efficiency have been excluded. This is extremely important since
the capital and operating costs are a function of the amount of S02
removed and removal efficiency requirements. Systems removing less
SO- will undoubtedly have lower costs.
PEDCo's work encompases 17 scrubber installations currently
operating. Calculations have been performed, based upon design
criteria, to evaluate capital and operating costs in terms of
dollars per ton of SO- removed. Data for these stations
4*
982
-------
are listed in Table 10. From this table it is evident that
even on a dollar per ton basis wide fluctuations still exist.
Two of the important variables contributing to the wide
fluctuations in costs are start-up date and the difference
between new and retrofit units.
If we compare only those new units burning high sulfur
coal with start-up during the years 1976-1977 impressive
correlations exist. Figure 6 exemplifies the capital require-
ments in terms of dollars per ton of S02 removed for the six
units falling into this category; three of which are lime systems
using Thiosorbic and three limestone. Values for the lime systems
range from $12.96 to $18,19 per ton of S02 removed. The limestone
systems range from $29.46 to $32.38. Average values for the two
processes are $16.45 and $30.55, respectively; for lime and
limestone.
Annual operating costs are compared on the same basis in
Figure 7. Once again the lime process is less expensive. Values
for lime scrubbing range from $232.10 to $233.14 per ton of S02
removed. On the other hand, limestone costs range from $234.71
to $331.98. Average values for the annual operating costs are
$232.70 and $285.54; respectively, for lime and limestone.
Capital and operating costs_for a system designed specifically
for Thiosorbic'scrubbing would yield even.higher savings.
983
-------
Table 10
Operational FGD Systems Economics
to
oo
Alkali Unit
Thiosorbic Lime
Conesville 5
Elrama 1-4
Phillips 1-6
Mansfield 1
Mansfield 2
Lime
Hawthorn 3
Hawthorn 4
Green River 1-3
Cane Run 4
Cane Run 5
Paddys Run 6
Limestone
Cholla 1
Petersburg 3
LaCygne 1
Winah 2
Southwest 1
Widows Creek 8
FGD New or
MW Retrofit Startup
411
510
410
917
917
110
110
64
190
200
70
126
532
874
280
194
550
N
R
R'
N
N
R
R
R
R
R
R
R
N
N
N
N
R
1/77
10/75
7/73
4/76
7/77
11/72
8/72
9/75
8/76
12/77
4/73
10/73
10/77
2/73
Tin
4/77
5/77
Coal
%
Sulfur
4.7
2.0
2.0
4.7
4.7
2.0
2.0
3.8
3.75
3.75
3.75
.55
3.25
5.0
1.0
3.5
3.7
Design
SO,
Removal
89.5
83.0
83.0
92.1
92.1
70.0
70.0
80.0
85.0
85.0
80.0
58.5
80.0
76.0
69.0
80.0
80.0
Annual Tons
Capital Operating SO2
Cost Cost Removed
($/kw) (mills/Kwh) (Tons/Yr)
70.8
134.5
149.7
102.2
102.2
87.3
87.3
77.6
80.6
67.5
76.5
56.0
100.6
68.0
66.5
117.7
113.2
7.42
8.59
9.54
7.67
7.67
4.09
4.09
5.24
8.64
5.56
6.51
2.58
6.56
3.78
2.92
6.17
5.28
74,816
36,636
29,453
171,774
171,774
6,665
6,665
8,419
26,208
27,587
9,088
1,754
59,857
143,723
8,360
23,507
70,451
Capital
Cost
($/Ton SO
Removed)
12.96
62.41
69.46
18.19
18.19
48.03
48.03
19.66
19.48
16.31
19.64
139.04
29.80
13.78
74.24
32.38
29.46
Annual
Operating
Cost
2 ($/Ton SO,
Removed)
232.10
680.88
756.18
233.14
233.14
384.36
384.36
226.80
356.66
229.51
285.53
1,055.09
331.98
130.89
556.87
289.94
234.71
Source:
B. Laseke, et.al.,EPA Utility FGD Survey: June-July 1978
EPA 600/7-78-051 d
-------
Figure 6
Capital Costs - Dollars per Ton SO2
Removed New Units Burning >3%
Sulfur Start-Up in 1976-1977
35.00 r
•o
30.00 -
0)
cc
CM
25.00 -
o
20.00 -
o
O
15.00 -
32.38
10.00
Lime
Limestone
985
-------
Figure 7
Annual Operating Costs - Dollars
per Ton SO2 Removed New Units
Burning >3% Sulfur
Start-Up In 1976-1977
350
331.98
100
Lime
Limestone
986
-------
Conclusion
The discussion today has been directed at reagent
decision-making from an operational point of view for those
faced with burning high sulfur coal and meeting the proposed
S02 emission standards. This information is supported by the
largest base of experience to date, the Thiosorbic lime process.
Over the last several years, the state of the art within
the FGD industry has moved forward at a rapid pace. This
success must be attributed to organizations such as EPA, TVA,
EPRI, the utilities and private enterprise. Even with the
great advances made to date, the magnesium-enriched lime scrubbing
system still remains a very favorable option, both technically
and economically.
987
-------
REFERENCES
1. J. G. Selmeczi, "Process for Wet Scrubbing of Sulfur
Dioxide From Flue Gases", U.S. Patent #3,914,378,
October 21, 1975.
2. J. G. Selmeczi, "The Thiosorbic Scrubbing Process",
presented at the Annual Fall Meeting of Society of
Mining Engineers, AIME, Acapulco,-Mexico, Septem-
ber, 1974.
3. Harlan Head, 'EPA Alkali Scrubbing Test Facility:
Advanced Program", Second Progress Report. EPA
600/7-76-008, September, 1976.
4. Harlan Head, "EPA Alkali Scrubbing Test Facility:
Advanced Program", Third Progress Report. EPA
600/7-77-105, September, 1977.
5. B. Laseke, et.al., "EPA Utility FGD Survey",
EPA 600/7-78-051d, June-July, 1978.
6. Federal Power Commission, "The Status of Flue Gas
Desulfurization Applications in the United States:
A Technological Assessment", NTIS PB-271-362, July, 1977,
988
-------
SESSION 5
FGD CURRENT STATUS AND FUTURE PROSPECTS:
VENDOR PERSPECTIVES
FRANK T. PRINCIOTTA, CHAIRMAN
Panel: Brief discussion of significant issues
followed by questions from the audience.
Members: Abdus Saleem
Irwin A. Raben
James R. Martin
Henry M. Majdeski
Robert J. Gleason
Vincent B. Birkner
No papers or discussions are included for
this session.
989
-------
SESSION 6
INDUSTRIAL APPLICATIONS
RICHARD D. STERN, CHAIRMAN
990
-------
6A
THE STATUS OF INDUSTRIAL BOILER
FGD APPLICATIONS IN THE
UNITED STATES
Prepared by:
John Tuttle
Avinash Patkar
PEDCo Environmental, Inc.
and
R. Michael McAdams
U.S. Environmental Protection Agency
Industrial Environmental Research Laboratory
Research Triangle Park, North Carolina
For Presentation at the U.S. EPA
Las Vegas FGD Symposium
March 5-8, 1979
991
-------
ABSTRACT
PEDCo Environmental, Inc., under contract to the U.S. EPA's
Industrial Environmental Research Laboratory at Research Triangle
Park, has been monitoring the status of industrial boiler flue
gas desulfurization (FGD) applications since 1976. The informa-
tion provided in this survey program has been obtained by visits
to these industrial FGD sites, and through regular contact with
company representatives, process designers, equipment suppliers,
and government agencies.
This paper summarizes the EPA Industrial Boiler FGD Survey
report, including:
0 The current (fourth quarter 1978) status,.of industrial
FGD applications in the U.S., identifying the number of
systems operating, under construction, or in a planning
phase.
0 A summary of system suppliers, including the number of
systems served, processes offered, and total gas flow
treated.
0 A summary of FGD process types currently being imple-
mented, with a discussion of some major installations,
including system design and operating experience.
0 A summary of currently reported costs for the indus-
trial FGD applications.
992
-------
INTRODUCTION
This paper addresses industrial boiler flue gas desul-
furization (FGD) applications, a topic about which little was
widely known until rather recently. Under contract to U.S. EPA's
Industrial Environmental Research Laboratory at Research Triangle
Park in North Carolina, PEDCo Environmental has been monitoring
FGD technology as applied to industrial boilers since the summer
of 1976. The product of this program is the quarterly updated
report titled, "EPA Industrial Boiler FGD Survey." Anyone de-
siring to receive th£s report, at no cost, may request the addi-
tion of his' or her name to the mailing list by contacting one of
the authors of this paper.
Heretofore much of the focus of FGD has been on utility
applications. This has been due to the high levels of emissions
from large utility plaffts, the tremendous capital and operating
costs associated with utility applications, the need for rapid
FGD development in the utility sector, and the good communication
among the various utility companies. During the past year or so,
however, increased attention has been given to the problem of
sulfur dioxide (S02) emissions from industrial boilers.
Therefore, a survey was instituted to summarize the status
of FGD systems in the industrial sector. In this survey effort
no claim is made that all systems operating, under construction,
or planned are identified. Indeed, several known FGD systems are
993
-------
not included in the report, at the request of the FGD plant
representatives. However, we do maintain close contact with
these plants and monitor their FGD activities. In addition,
there are other known FGD systems whose company representatives
choose to maintain confidentiality concerning their activities.
Despite these deficiencies, the survey report presents a fairly
representative picture of FGD systems on industrial boilers.
In matters concerning utility companies, there is a well-
developed communication network. Items such as rates, fuels,
load forecasts, boiler sales, and pollution abatement strategies
are well publicized and addressed in many arenas. The activities
of the private corporations, though not necessarily secret, are
simply not as highly publicized in such a coordinated manner.
Basic differences exist between utility and industrial
applications. Some of these differences are discussed in the
following paragraphs.
The most obvious difference is size. Whereas the average
utility FGD installation is on the order of 350 MW (about 700,000
scfm)*, the average industrial application is on the order of
40,000 scfm (about 20 MW equivalent). However, in comparing
numbers of systems, there are 46 operating utility systems and
2
132 operating industrial systems.
Another basic difference is the type of technology being
used to treat the flue gas in the two different applications. In
*
The British measurement system is used in this report, despite
EPA's metric policy, for convenience to the readers. The
British to metric conversion factors are given at the end of
this paper.
994
-------
utility applications, about 90 percent are treated by calcium-
based technology (lime or limestone). In industrial applica-
tions, about 90 percent are treated by sodium-based technology
(sodium hydroxide, sodium carbonate, dual alkali, or caustic
waste stream). This affects the equipment required, types of
problems encountered, control level attainable, and the types of
end materials generated.
Yet another difference is that, in general, the smaller
industrial FGD installations incorporate less redundancy, less
ancillary equipment, and less instrumentation. However, oper-
ating histories have been generally successful. The most recent
survey report, for the fourth quarter of 1978, listed no fewer
than 18 sites reporting 90 percent or better reliability. These
sites represent 90 FGD systems controlling 109 boilers or steam
generators. Nine of these sites reported a 100 percent reli-
ability index value for their 14 FGD systems (controlling 32
bbilers or steam generators). Three other sites did not report
the relevant operating hours although they did report that no
problems occurred. Of the remaining 18 operating sites, oper-
ating hours were made available by only 2 sites; GM's Parma FGD
system demonstrated a 46 percent reliability for October and
November, and GM's Dayton (Delco Moraine) system was down for the
whole period.
The fourth quarter of 1978 was not unique with respect to
these high system reliabilities. It has become clear that the
995
-------
application of FGD technology in the industrial sector has been
generally successful.
Table 1 summarizes the number and gas flow capacity of FGD
systems operating on industrial boilers in the United States. It
can be seen that the number of operating industrial FGD systems
(132 as of December 1978) far exceeds operating utility FGD
systems (46 as of January 1979). However, utility FGD applica-
tions are far exceeding industrial applications with respect to
the amount of flue gas being treated. Utility applications are
treating about 16,000 MW whereas industrial applications are
treating 2700 MW equivalent, using a conversion of 2000 scfm per
MW equivalent.
Figure 1 shows the steady increase in the number of FGD
applications to industrial boilers beginning in 1972. The growth
trend is shown to drop off in the 1980's because companies are
not making firm plans that far ahead.
Table 2, a breakdown by FGD process and gas flow capacity
being controlled, shows that about 86 percent are using a sodi-
um-based S02 absorption mechanism. The advantage of sodium-based
systems is that absorption is effected by soluble salts. This
eliminates the scaling and plugging problems that accompany
calcium-based systems.
Tables 3 and 4 summarize the vendors who are serving the
industrial FGD market.
996
-------
TABLE 1. NUMBER AND CAPACITY OF
U.S. INDUSTRIAL BOILER FGD SYSTEMS*
Status
Active
Operational
Under construction
Planned
Total
Number of
FGD units
(sites)
132 (39)
27 (11)
14 (11)
173 (61)
Capacity,
scfm
5,487,000
1,236,000
2,386,000
9,109,000
*
There are probably several systems in various planning phases
which have not yet been located.
997
-------
10
IX)
o 9
x
J 8
u
in
o
CD
g 5
in
§ 4
LU
>
H-l _
5
i 2
i i i i i i
I I I I I I I I
72 74 76 78 80 82
YEAR OF STARTUP
84
86
88
Figure 1. Use of industrial FGD as a function
of startup date.
998
-------
TABLE 2. INDUSTRIAL FGD PROCESS APPLICATIONS
Process
Nonregenerable sodium
(sodium hydroxide or
sodium carbonate)
Dual alkali
Lime
Limestone
Caustic waste stream
Ammonia waste stream
Citrate
SULF-X
Process not selected
Capacity, scfm x 10
Operating
3,135
591
84
55
1,009
626
0
0
Under
construction
656
437
0
0
0
0
142
0
Planned
286
417
30
320
0
0
0
10
1,324
999
-------
TABLE 3. VENDORS WITH MORE THAN ONE INDUSTRIAL FGD APPLICATION
Vendor
Airpol
C-E Natco
FMC Environmental
Equipment
GM Environmental
Heater Technology
Koch
Zurn
No. of
installations
7
3
12
2
2
3
2
Process
Sodium hydroxide
Sodium carbonate
Dual alkali
Caustic waste stream
Sodium hydroxide
Sodium hydroxide
Sodium carbonate
Dual alkali
Caustic waste stream
Sodium hydroxide
Dual alkali
Sodium hydroxide
Sodium carbonate
Sodium carbonate
Dual alkali
Total size,
scfm
1,274,000
24,000
1,585,000
235,000
92,000
478,000
105,000
Status*
0,C,P
0,C
0,C,P
0
0,C
0,P
0
o
o
o
Operating, Construction, Planned are 0, C, P, respectively.
-------
TABLE 4. VENDORS WITH ONE INDUSTRIAL FGD APPLICATION
Vendor
Process
Size, scfm
Status
o
o
A.D. Little
Bureau of Mines
Carborundum
Ceilcote
Combustion Equipment Assoc,
Ducon
Entoleter
Flakt
Pittsburgh Environmental
and Energy Systems
Research-Cottrell/Bahco
Swemco
Thermotics
W.W. Sly Manufacturing
Wheelabrator-Frye/Rockwell
International
Sodium hydroxide
Citrate
Lime
Sodium hydroxide
Sodium carbonate
Sodium carbonate
Sodium hydroxide
Sodium carbonate
SULF-X
Limestone
Sodium carbonate
Sodium hydroxide
Caustic waste
stream
Dry sodium
carbonate
64,000
142,000
30,000
380,000
490,000
117,000
36,000
39,000
10,000
55,000
140,000
12,000
18,500
52,300
0
C
P
O
O
C
O
C
P
o
o
o
o
0 = operating, C = construction, and P = planned.
-------
NONREGENERABLE SODIUM FGD TECHNOLOGY
The predominant technology currently in use in industrial
FGD applications employs either sodium hydroxide or sodium car-
bonate as the makeup reagent. Sizes range from 12,000 to 245,000
scfm.
There are 93 operating FGD systems involved at 18 plant
sites. Table 5 summarizes these plants. A brief discussion of
FGD sites follows.
Alyeska Pipeline Service Company, Valdez, Alaska
Background—
The SO- control system at this installation is not specifi-
cally an environmental control device, as it is'not required to
comply with an emissions regulation. The fuel being burned in
this three-unit boiler plant is low sulfur oil (0.03 to 0.1
percent). Alyeska needed an inert, noncorrosive blanket gas for
its oil storage tanks. It was determined that boiler flue gas
would have the required low oxygen levels. However, the SO,,
concentration in the boiler flue gas caused corrosion problems.
FMC Environmental Equipment Division solved the corrosion problem
by providing a twin-module FGD. system incorporating FMC's disc-
and-donut tray design (four trays per module). The system is
designed with 100 percent redundancy in that only one module is
used at a time. The spare module assures Alyeska that it will
have an uninterrupted supply of inert gas for use in the oil
storage tanks. Scrubber bleed-off (10 gpm) is aerated to sodium
1002
-------
TABLE 5. INDUSTRIAL SITES USING NONREGENERABLE SODIUM FGD TECHNOLOGY
Company
Alyeska Pipeline Service Co.
Bel ridge 011 Co.
Bel ridge 011 Co.
Chevron U.S.A., Inc.
FMC (Soda Ash Plant)
General Motors Corp.
General Motors Corp.
General Motors Corp.
General Motors Corp.
Getty 011 Co.
Getty 011 Co.
ITT Rayonler, Inc.
Kerr-McGee Chemical Corp.
Mead Paper-board Co.
Mobil 011 Co.
Northern Ohio Sugar Co.
Texaco, Inc.
Texasgul f
Location
Valdez, AK
MqKlttMck, CA
McK1ttr1ck, CA
Bakersfleld, CA
Green River, WY
Dayton, OH
Pontlac, MI
St. Louis, MO
Tonowanda, NY
Bakersfleld, CA
Bakersfleld, CA
Fernandlna Beach, FL
Trona, CA
Stephenson, AL
San Ardo, CA
Freemont, OH
San Ardo, CA
Granger, WY
Fuel
011
011
011
011
Coal
Coal
Coal
Coal
Coal
011
011
Bark & 011
Coke, Coal ,
& 011
011
011
Coal
011
Coal
Flow rate,
scfm
50,000
12,000
12,000
248,000
446,000
36,000
107,300
64,000
92,000
72,000
95,000
176,000
490,000
100,000
175,000
40,000
380,000 -
140,000
Reagent
Sodium hydroxide
Sodium hydroxide
Sodium hydroxide
Sodium Carbonate
Sodium carbonate
Sodium hydroxide
Sodium hydroxide
Sodium hydroxide
Sodium hydroxide
Sodium carbonate
Sodium carbonate
Sodium hydroxide
Sodium carbonate
Sodium carbonate
Sodium hydroxide
Sodium hydroxide
Sodium hydroxide
Sodium hydroxide
Vendor
FMC Env. Equipment
Heater Technology
Thermotics, Inc.
Koch Engineering
FMC Env. Equipment
Entoleter, Inc.
GM Environmental
A. D. Little
FMC Env. Equipment
FMC Env. Equipment
In-house Design
Airpol Industries
Combustion Equipment
Assoc.
Airpol Industries
In-house design
In-house design
Ceil cote
Swemco, Inc.
* SO-
removal
96+
90
90
90
95
86
N.A.
90+
90-95
90+
96
80-85
98+
95
90
N.A.
73
90
o
o
(JO
N.A. = Not available.
-------
sulfate, and mixed with ballast water (10,000 gpm) from tankers
in Port Valdez, and is discharged into the bay.
Operating Experience—
Because the scrubber plant is required for successful opera-
tions elsewhere on-site, Alyeska spares no expense in operating
the FGD system. Although the design control pH was 6.5, control
pH is now 8.0, for additional SO control. Another unique ap-
J\,
proach has been the installation of cyclohexylamine sprays down-
stream of the absorber to reduce corrosion. An emissions test
run in late 1977 showed that, while inlet S02 concentrations were
ranging from 150 to 160 ppm, outlet SO (S02/S03/S04) levels were
about 5 ppm giving approximately 97 percent S02 removal effi-
ciency.
Alyeska has released no capital or operating cost data for
this control system.
FMC Chemical Corporation Soda Ash Plant, Green River, Wyoming
Background—
This soda ash plant operates two coal-fired boilers firing a
1.0 percent sulfur coal. Each boiler produces 330,000 acfm at
320°F. Particulate control occurs in a hotside ESP, which is
followed in sequence by the economizer, a forced draft (with
respect to the FGD system) fan, and an FMC Environmental Equip-
ment Division S0_ control system.
Each boiler has its own FGD system consisting of two disc-
and-donut absorber modules. The absorber shells and discs are
1004
-------
carbon steel with a Ceilcote liner; the outer donut portions are
Inconel 825. Recirculation pumps and piping are rubber-lined.
To meet applicable codes, approximately 50 percent of the
boiler flue gas is scrubbed; the remainder is bypassed and used
for reheat. A bleed off stream is taken to a holding pond for
evaporation.
FMC reported that the overall air pollution control system
cost was $10 million in 1975 dollars.
Operating Experience—
Tests in 1978 showed that SO- removal efficiency varied from
87 to 94 percent depending on L/G ratio and pH. Problems have
included: a module liner failure, inadequate cold weather pro-
tection, inadequate pH control, faulty damper operation, and a
broken shaft on a recirculation pump. These problems have been
rectified and the FGD reliability index for the fourth quarter of
1978 was reported at close to 100 percent.
General Motors Corporation's Delco Moraine plant, Dayton, Ohio
Background--
This plant has two coal-fired (0.7-2.0 percent sulfur)
boilers each of which generates 34,000 acfm of flue gas at 500°F.
Initial particulate collection occurs in internal multiclones; an
Entoleter, Inc. emission control system provides secondary par-
ticulate and primary S02 control. Entoleter's "vane-cage" system
sets up a vortex of scrubbing liquor mist., A bleed-off from the
recirculation line is pumped to the onsite w.astewater treatment
1005
-------
facility, where it is clarified and pH adjusted, then discharged
to city sewers.
The design includes 316L stainless steel for absorber shells
and internals, mist eliminators, stacks, and recycle tanks. The
scrubber forced-draft fans, caustic tank, caustic pump, and inlet
ductwork are carbon steel. The recirculation pumps and piping
are rubber-lined.
The total installed capital cost was $668,000 in 1974 dol-
lars.
Operating Experience—
Plant personnel are currently carrying out extensive modifi-
cations necessitated by serious corrosion of the primary mist
eliminator. The corrosion was caused by acid rain fallout from
the secondary mist eliminator (no wash was provided). The new
configuration will consist of Hastelloy-G primary and secondary
radial vane mist eliminators and an intermediate Hastelloy-G
convex impingement plate to assist in rapid drainage of the mist
eliminator runoff. New flue gas flow sensors and a new pH moni-
toring system are also being installed.
Kerr-McGee Chemical Corporation, Trona, California
Background—
In June 1978, production began at Kerr-McGee's new soda ash
plant, the largest yet built to yield soda ash by direct carbona-
tion of brine. Annual production will be 1.3 million tons per
year. Processing innovations include carbonation under 13.5 psig
pressure and the recovery of CO^ from fossil-fuel-fired boilers.
1006
-------
The plant has two boilers, each producing 600,000 Ib/h steam
at 1500 psig. The steam is used initially to drive two 32-MW
non-condensing steam turbines to generate electricity, and is
then used as process steam. The boilers burn a mixture of west-
ern coal (0.7 percent sulfur) and petroleum coke (5.5 percent
sulfur). The flue gas flow rate from each boiler is 363,000 acfm
at 320°F and may contain SO2 concentrations ranging from 335 to
5985 ppm depending on the fuel mix used.
The flue gases are scrubbed by the end liquor from the soda
ash plant in flaked-glass lined mild steel vessels, each with
three 317L stainless steel sieve trays. Combustion Equipment
Associates supplied the FGD system. Spent liquor from the scrub-
ber bottom is recirculated by rubber-lined pumps; fresh end
liquor is added directly to the recirculation line. A bleed
stream returns absorber reaction products to the salt ponds.
The scrubbed flue gases are processed through two mono-
ethanolamine plants for CO2 extraction. Ambient air is heated
externally by a steam-tube bank and mixed with exit flue gases
for 50°F reheat.
The capital cost of this system was reported to be $6 mil-
lion in 1978 dollars.
Operating Experience—
The currently operating 317L stainless steel sieve trays
replaced the original polypropylene trays. The holes in the
original trays were cut unevenly and were not large enough. No
1007
-------
other problems have been reported. Both scrubbers demonstrated
100 percent reliability for the fourth quarter of 1978.
Mead Paperboard Company, Stevenson, Alabama
Background—
This plant is a 500 ton per day neutral sulfite pulp mill.
Two oil-fired (1.5 to 3.0 percent sulfur) boilers produce 175,000
Ib/h of steam each, at 600 psig. Airpol Industries supplied the
FGD system which treats 173,000 acfm at 450°F. The flue gas
passes through a stainless steel venturi/quench section, and then
into the 316L stainless steel absorber which includes three
bubble-cap trays for SO- removal. Sodium carbonate solution is
added directly to the recirculation line. Scrubber effluent is
sent to the onsite pulping operation by a continuous bleed-off.
Operating Experience—
Early operations were marked by the failure of a flaked-
glass liner (stress cracking), and a Carpenter 20 liner (seepage
between the liner and the shell). In 1977, a rubber liner was
installed throughout the entire scrubber, including the spin
vanes. As of the fourth quarter of 1978, the rubber lining was
still serving well, and the FGD system demonstrated 100 percent
reliability.
Texasgulf, Inc., Granger, Wyoming
Background—
At this 1 million ton per year soda ash plant, the boiler
plant consists of two coal-fired (0.75 percent sulfur) boilers
1008
-------
which generate a total of 600,000 Ib/h of steam at 300 psig.
Primary particulate control occurs in a hot-side ESP S0_ control
occurring in a Swemco, Inc. FGD system which consists of a
quencher, a two-stage sieve tray absorber, and a mesh-type mist
eliminator. The system is provided with bypass.
The quench section is Inconel 625, the absorber is carbon
steel with a flaked-glass liner, the sieve trays are Inconel 625,
and the mist eliminator is Teflon. The stack is insulated steel
with a flaked-glass liner.
Each absorber is divided vertically into two sections one of
which has a damper at the top. The damper allows for turndown to
25 percent of full boiler load. A bleed-off stream goes to a
holding pond for evaporation.
Operating Experience—
Only minor problems have been reported since the September
1976 startup. Ductwork at the outlet to the absorber was re-
placed by 304 stainless steel, some minor nozzle plugging has
occurred, and some piping was replaced. Most replacements have
been made during scheduled boiler outages. The fourth quarter
1978 reliability index was 100 percent.
California Enhanced Oil Recovery (EOR) Sites
Of the more than 170 FGD units currently covered by the
survey, more than half are in use or scheduled for use in EOR
sites, in California. Since these units predominate in the sur-
vey, a brief discussion of the EOR industry follows.
1009
-------
In 1976, the San Joaquin Valley Air Basin produced about 45
percent of California's crude oil. EOR operations accounted for
about 50 percent of that production. Additionally, EOR opera-
tions in the basin accounted for approximately 80 percent of
California's total EOR operations.
__The_jno_st common means of producing crude oil relies on
natural underground pressure to push the oil to the surface of a
well. If natural pressure is lacking, pumps are used to pull the
oil out of its reservoir. These methods of producing crude oil
are termed "primary recovery." If water is pumped into an oil
reservoir (secondary recovery), crude recovery can be increased
or "enhanced." When the capabilities of the primary and second-
ary methods are exhausted, additional enhanced recovery (tertiary
recovery) methods are considered. Tertiary oil recovery employs
thermal means to make heavy, viscous crude oils move through sand
and rock strata more easily.
Typically steam is injected continuously to flood the under-
ground oil reservoir with steam and hot water. There are cur-
rently about 800 steam generators (typical steam quality is 80
percent) in the San Joaquin Basin, most of which have capacities
in the range of 15 to 65 million Btu/h. Steam generators in this
size range vary from about 5000 to 30,000 acfm at approximately
600°F.
The California Air Resources Board (GARB) has proposed
S02/N0x regulations which would require the following for oil-
fired steam generators:
1010
-------
S02 NOX
emission limit, emission limit,
Source ppm ppm
New generators 60 100
Existing generators 200 150
To date, the following EOR production companies have given
permission to discuss FGD operations at their oil sites:
Belridge Oil Company
Chanslor Western Oil and Development Company
Chevron U.S.A., Inc.
Getty Oil Company
Mobil oil Company
Shell Oil Company
Sun Production Company
Texaco, Inc.
These companies are firing oil (1.0 to 1.7 percent sulfur) re-
covered from their own onsite wells. Furthermore, all are using
sodium-based chemistry for SO2 removal, in scrubber/absorbers
supplied by companies such as Thermotics, Heater Technology,
Ducon, and C-E Natco, in addition to such firms as FMC Environ-
mental Equipment Division and Koch. Getty Oil and Mobil Oil are
designing their own equipment.
At the Chanslor Western site near Bakersfield, California,
dual alkali technology will be used to regenerate sodium hydrox-
ide scrubbing liquor and produce a calcium sulfite/sulfate sludge.
All of these systems are designed for approximately 90 percent
S02 removal.
1011
-------
Belridge Oil currently has two operating FGD systems sup-
plied by different vendors, each employing an eductor-type ven-
turi for S0_ removal. The venturi has an adjustable disc, which
forms the contact stage for the gas and the recirculating liquor.
L/G ratio is 40 gal./lOOO acf. Because gas and liquid move in
the same direction and only one stage of absorption exists, a
higher pH (about 8.0) is maintained in the recirculating loop.
Problems have included a pump failure, pump vibration, and faulty
electric hookups. Nevertheless, both units demonstrated better
than 90 percent reliability in the fourth quarter of 1978.
Chevron U.S.A., Inc. employs three Koch Engineering FGD
systems to control S02 emissions from 18 steam generators (six
generators per FGD system). The three systems are controlling a
total of 450,000 acfm at 500°F (150,000 acfm per system). The
scrubber modules consist of a quench section followed by three
Koch Flexitrays and a Fleximesh mist eliminator. The scrubbers
and trays are 316L stainless steel, the Fleximesh is Incoloy 825,
and the stub stack is 316 stainless steel. Recent operations
have been completely problem-free.
Getty Oil Company operates an oil field near Bakersfield
which occupies 15 square miles. The EOR operations use 136 steam
generators, 100 with a heat rate of 50 million Btu/h and 36 with
a heat rate of 20 million Btu/h, all firing a 1.1 percent sulfur
oil.
1012
-------
Currently six FGD systems are operating to control S02 emis-
sions from 51 steam generators. FMC Environmental Equipment
supplied a three-stage disc-and-donut (316L stainless steel)
absorber followed by an Inconel wire mesh mist eliminator. This
system controls emissions from six steam generators each rated at
50 million Btu/h. These systems are briefly discussed as fol-
lows.
The FMC system has required some fairly extensive rework
including the addition of a fourth absorption tray and a switch-
over from a mesh to a chevron mist eliminator. Getty is using an
in-house-design which will consist of a system controlling emis-
sions from nine steam generators, each rated at 50 million Btu/h.
Each absorber will consist of three Koch Flexitrays and a Flexi-
chevron mist eliminator. These systems are designed to remove 96
percent of the inlet 600 ppm S0_. As of January 1979, five of
these systems were on-line with four more scheduled to start up
in mid-February 1979.
The Getty in-house-designed systems have experienced some
scaling, vibrating fans, and some poor mist elimination charac-
teristics. Getty will evaluate the costs of running sodium
hydroxide as compared to sodium carbonate.
DUAL ALKALI FGD TECHNOLOGY
In industrial FGD applications, dual alkali technology ranks
second with respect to total gas flow being controlled by FGD
systems now operating, under construction, or planned (see Table
2). The process offers the advantage of SO- removal by soluble
1013
-------
alkali salts (usually sodium salts), with the concomitant elimi-
nation of scaling and plugging problems in the absorber. The
soluble scrubber effluent is then treated with lime (or lime-
stone) to regenerate the scrubbing liquor and produce a solid
product, calcium sulfite/sulfate, for disposal.
This paper will not discuss dual alkali applications in any
depth, because of other papers being presented at this symposium
on the Firestone, Pottstown and the General Motors, Parma instal-
lations. However, a brief summary of other dual alkali systems
follows. Table 6 summarizes these plants.
Caterpillar Tractor Company, Illinois
Caterpillar has purchased dual alkali FGD systems from Zurn
Industries and from FMC Environmental Equipment Division. The
first system to begin operation was a Zurn dilute-mode system at
the Joliet plant in September 1974. Since then the following
Caterpillar plants have initiated FGD operations: Mossville (FMC
concentrated-mode, October 1975), Morton (Zurn dilute-mode,
January 1978), and East Peoria (FMC concentrated-mode, April
1978) . The Mapleton plant was scheduled to begin operating an
FMC system in January 1979.
All the Caterpillar heat plants fire a high sulfur (2.5-3.2
percent) Illinois coal. In general, recurrent problems with the
systems in the Caterpillar plants have yet to be solved. Al-
though all of these facilities rely on the FGD system for some
particulate removal, the systems have yet to yield fully satis-
1014
-------
TABLE 6. INDUSTRIAL SITES USING DUAL ALKALI FGD TECHNOLOGY
o
»—i
01
Company
Caterpillar Tractor Co.
Caterpillar Tractor Co.
Caterpillar Tractor Co.
Caterpillar Tractor Co.
Firestone Tire & Rubber Co.
General Motors Corp.
Location
E. Peorla, 11
Jollet, IL
Morton, IL
MossvUle, IL
Pottstown, PA
Parma, OH
Fuel
Coal
Coal
Coal
Coal
Coal
Coal
Flow fate
scfm
210,000
67,000
38,000
140,000
8,070
128,400
Reagent
Dual alkali
(Concentrated)
Dual alkali
(Dilute)
Dual alkali
(Dilute)
Dual alkali
(Concentrated)
Dual alkali
(Concentrated)
Dual alkali
(Dilute)
Vendor
FMC Env. Equipment
Zurn Industries
Zurn Industries
FMC Env. Equipment
FMC Env. Equipment
6M Environmental
% so2
removal
90
90
90
90+
90.5
90
-------
factory performance in this area. An exception is the Morton
installation where some in-house modifications have improved
overall operations.
Caterpillar has not released cost data for these systems.
Other Systems
The remaining dual alkali FGD systems covered in the EPA
Industrial Boiler FGD Survey are either under construction or
planned.
As mentioned earlier, Chanslor Western Oil and Development
Co. in Bakersfield, California has purchased a system from FMC;
system startup is scheduled for the spring of 1979. This system
is designed for 96 percent S0_ removal (710 ppm at the inlet),
and will control 133,400 acfm of flue gas at 550°F from eight
steam generators totaling 310 million Btu/h.
Arco/Polymers, Inc. in Monaca, Pennsylvania is planning on a
June 1980 startup for its FMC dual alkali system. The boiler
plant at Monaca consists of three coal-fired units (firing 3.0
percent sulfur coal) which generate 900,000 Ib/h of steam at 700
psig and 700°F. This system is designed for 90 percent S09
removal (1800 ppm at the inlet).
The U.S. Air Force has purchased an Airpol Industries dual
alkali system for its Grissom Air Force Base near Bunker Hill,
Indiana. This system is scheduled for startup in September 1979.
It will control emissions from three coal-fired (0.7 to 4.6
percent sulfur) boilers.
1016
-------
Dupont, Inc. is planning a coal-fired (1.5 percent sulfur)
installation at an as yet unspecified site in Georgia. The
boiler plant will consist of three units generating a total of
1.2 million Ib/h of steam. Regulations are expected to require
90 percent SO_ removal. Startup is scheduled for 1987.
ALKALI WASTE STREAM FGD TECHNOLOGY
Pulp mills, textile mills, and beet sugar plants may use an
onsite waste stream as the scrubbing liquor for S02 removal.
This would hold true for any industrial operation where a high-pH
waste stream is available in sufficient quantity. Industrial
sites using alkali waste streams for S02 control are:
0 American Thread, Marion, NC — caustic waste stream;
coal (1.0-1.5 percent sulfur); 18,500 scfm; vendor:
W.W. Sly Manufacturing.
0 Canton Textiles, Canton, GA — caustic waste stream;
coal (0.8 percent sulfur); 25,000 scfm; vendor: FMC.
0 Georgia-Pacific Paper Co., Crossett, AR — caustic
waste stream; bark, coal, and oil (1.5-2.0 percent
sulfur); 220,000 scfm; vendor: Airpol.
0 Great Southern Paper Co., Cedar Springs, GA — caustic
waste stream; bark, coal, and oil (1.0-2.0 percent
sulfur); 420,000 scfm; vendor: Airpol.
0 Great Western Sugar, several locations -- ammoniacal
waste stream; coal (0.6-2.0 percent sulfur); 462,000
scfm; in-house design.
0 Minn-Dak Farmers' Co-op, Wahpeton, ND — ammoniacal
waste stream; lignite (1.0 percent sulfur); 164,000
scfm; same design as Great Western Sugar.
0 Nekoosa Papers, Inc., Ashdown, AR — caustic waste
stream; coal (1.0-1.5 percent sulfur); 211,000 scfm;
vendor: Airpol.
1017
-------
St. Regis Paper Co., Cantonment, FL — caustic waste
stream; bark, oil, and gas (less than 1.0 percent
sulfur); 115,000 scfm; vendor: Airpol.
OTHER FGD TECHNOLOGIES
Table 7 summarizes plants using other FGD technologies.
Lime/Limestone—
This is not a predominant technology in industrial FGD
applications. The two operating systems are at Rickenbacker Air
Force Base near Columbus, Ohio, and at Armco near Middletown,
Ohio.
Bureau of Mines Citrate—
St. Joe Zinc Co., in Monaca, Pennsylvania will soon be
initiating the startup phase of Citrate FGD system operations.
This FGD system, discussed in another paper at this symposium,
will be the first full-scale (60-MW) demonstration for this
promising regenerable process.
SULF-X—
This regenerable acid mine waste process has been piloted on
a 1-MW slipstream at Fort Benjamin Harrison. A larger scale
demonstration (10,000 scfm or about 5 MW equivalent) will begin
this year at the Western Correctional Institute in Pittsburgh.
Startup of the absorber is scheduled for May 1979 and sulfur
production will begin in late fall 1979. This system has demon-
strated the potential for simultaneous SO9 and NO removal. The
^ X
process is offered by Pittsburgh Environmental and Energy Sys-
tems.
1018
-------
TABLE 7. INDUSTRIAL SITES USING OTHER FGD TECHNOLOGIES
Company
Armco Steel
Rickenbacker Air Force Base
St. Joe Zinc Co.a
Western Correctional
Institute5
Location
Middletown, OH
Columbus, OH
Monaca, PA
Pittsburgh, PA
Fuel
Coal
Coal
Coal
Coal
Flow rate,
scfm
84,000
55,000
142,000
10,000
Reagent
Lime
Limestone
Citrate
Ferric sulfide
Vendor
Koch Engineering
Research-Cottrel 1 /Banco
Bureau of Mines
Pittsburgh Env.
and Energy
% SO-
removal
---
90+
90+
90
Under construction.
Planned, contract awarded.
-------
CAPITAL AND ANNUAL COSTS
Table 8 summarizes available data on capital and annual
costs for nonregenerable sodium and alkaline waste stream FGD
systems operating on industrial boilers. Sufficient data was not
available for the other technologies to report meaningful aver-
ages.
The cost figures reported by the plants were adjusted by
PEDCo for inflation to 1978 dollars. In some cases adjustments
were made for equipment, installation, instrumentation, and
engineering. These costs are not to be taken as representing the
product of a complete cost analysis. They are presented to
provide a guideline for the costs associated with these two
representative FGD processes. The capital costs are expressed in
terms of dollars/ scfm because the amount of flue gas to be
treated is the most critical design parameter in sizing an FGD
system. The operating costs, expressed as <=/10 Btu, are based
on 340 days operation at 100 percent load. Realistic cost fig-
ures will be higher since, in actual operation, a yearly capacity
factor could be as low as 20 to 30 percent.
TABLE 8. ADJUSTED CAPITAL AND ANNUAL COSTS
Process
$/scfm
(number of plants)
C/106 Btu
(number of plants)
Nonregenerable sodium
Alkaline waste streams
15.50
(13)
9.68
(4)
15.49
(5)
2.42
(2)
1020
-------
SUMMARY/CONCLUSIONS
Existing FGD applications in the industrial sector have
generally shown that high reliabilities of system operation can
be attained along with S0_ removal efficiencies as high as 98
percent. There has been a steady increase in the application of
industrial FGD, with an average of about seven new systems start-
ing up each year since 1972. Flue gas capacities of known opera-
tional and planned FGD systems totals over 9 million scfm (4600
MW equivalent). Regenerable as well as the more commonly used
nonregenerable technologies are being developed by a number of
FGD system vendors with varying degrees of success. The new more
stringent regulations for SO« emissions may be expected to in-
crease interest in these technologies, and to stimulate perform-
ance of future FGD systems.
1021
-------
REFERENCES
1. Melia. M., et al., EPA Utility FGD Survey: October-November
1978, EPA-600/7-79-022b, PEDCo Environmental, Inc., Cin-
cinnati, Ohio, January 1979.
2. Tuttle, J.D., et al., EPA Industrial Boiler FGD Survey:
Fourth Quarter 1978, EPA-600/7-79-067a, PEDCo Environmental,
Inc., Cincinnati, Ohio, February 1979.
3. Goodley, A., Consideration of a Proposed Model Rule for
Control of Emissions of Sulfur Oxides and Oxides of Nitrogen
from Steam Generators in the San Joaquin Valley Air Basin,
California Air Resources Board, Sacramento, California,
March 24, 1978.
SUPPLEMENTARY INFORMATION
The British system of measurement is used in this report. Some
of the conversion factors between the British and metric systems
are shown below:
British Unit
1 t (short ton)
1 Ib (pound)
1 gal. (gallon)
1 scfm
1 gal./lOOO acf
1 Btu
Metric Unit
0.91 metric ton
0.45 kilogram
3.79 liter
1.58 normal cubic meter/hour
0.134 liters/actual cubic meter
1.05 x 103 joules
In addition, certain engineering terms used in this paper are
defined below:
acf
acfm
Availability
Reliability
Actual cubic feet, unit of gas volume measured at
its actual temperature and pressure
Actual cubic feet per minute, unit of gas flow
rate measured at its actual temperature and pres-
sure
Hours the FGD system was available (whether oper-
ated or not) divided by hours in period, expressed
as percentage
Hours the FGD system was operated divided by hours
it was called upon to operate, expressed as per-
centage
1022
-------
ENVIRONMENTAL ASSESSMENT OF THE DUAL ALKALI FGD SYSTEM
APPLIED TO AN INDUSTRIAL BOILER FIRING COAL AND OIL
Wm. H. Fischer Wade H. Ponder Roman Zaharchuk i
Gilbert Associates U.S. Environmental Firestone Tire and Rubber Co.
Reading, Pa. Protection Agency Pottstown, Pa.
Industrial Environmental
Research Laboratory
Research Triangle Park, N.C.
ABSTRACT
This paper summarizes the results of a comparative multimedia assessment of
a dual alkali flue gas desulfurization system on a coal/oil industrial
boiler to determine relative environmental, energy, economic, and societal
impacts. Comprehensive sampling and analyses of multimedia emissions from
the boiler and its control equipment were conducted to identify criteria
pollutants and other species. The results indicate that: (1) while the
quantity of uncontrolled particulate matter from oil-firing is considerably
less than from coal-firing, the oil-fired particles are generally smaller
and the concentration of particles in the treated flue gas from oil-firing
is approximately the same as from coal-firing; (2) uncontrolled emissions of
NO and CO during coal-firing are about triple those during oil-firing; (3)
while sulfate emissions from the boiler during coal-firing are about triple
those during oil-firing, sulfate emissions after the control equipment are
essentially identical; (4) emissions of cadmium, calcium, magnesium, nickel
and vanadium are higher during oil firing; (5) oil-firing may produce
cadmium concentrations in vegetation approaching levels injurious to
humans; coal-firing may produce molybdenum levels injurious to cattle.
1023
-------
INTRODUCTION
Objective
Conventional methods of converting fossil fuels to usable forms of energy
have impacts on the air, land and water, i.e., "multimedia impacts." These
impacts are not separate and distinct; rather, they are interrelated and
involve delicate balances and trade-offs.
The Environmental Protection Agency (EPA), with primary responsibility
for controlling adverse environmental impacts of pollutant emissions, has been
active since its inception in determining the identities and quantities of
potential pollutants released to the environment when fossil fuels are burned.
Information from EPA R&D efforts is being used for three principal purposes:
to assess the health and environmental impacts caused by the release of
combustion pollutants to the environment; to define the needs for technology
to control the release of these pollutants; and to develop standards to limit
emissions.
CCEA Program
In response to the need for a comprehensive environmental assessment of
conventional combustion systems, EPA's Industrial Environmental Research
Laboratory at Research Triangle Park (EPA/IERL-RTP), North Carolina, has
established a unified Conventional Combustion Environmental Assessment (CCEA)
program. It is a major new program aimed at the comprehensive assessment of
environmental, economic, and energy impacts of multimedia pollutant emissions
from stationary industrial, utility, residential, and commercial conventional
combustion processes. The primary objective of the CCEA program is to
identify and evaluate information from all relevant sources in order to:
determine the extent to which this information can be utilized to assess the
total environmental, economic, and energy impacts of conventional combustion
processes; identify and acquire additional information needed for such
assessment; define the requirements for modifications or additional
development of control technology; and define the requirements for modified or
new standards to regulate pollutant emissions.
The CCEA program will coordinate and integrate ongoing and future studies
into a unified environmental assessment structure and serve as a centralized
base of information on the environmental impacts of conventional combustion
processes. Coordination and information exchange between CCEA-related studies
should minimize duplication and maximize the return from available resources.
The environmental assessment (EA) methodology employed in the CCEA
program fundamentally consists of three iterative steps (Figure 1):
1. Characterization of the combustion process (including any associated
pollution control devices) and its effluents.
1024
-------
CHARACTERIZATION OF COMBUSTION
PROCESS AND ITS EFFLUENTS
ASSESSMENT OF HEALTH AND
ECOLOGICAL IMPACTS
• Identification of environmental
impacts
• Development of environmental
goals and objectives
• Comparison of impacts with
goals and objectives
• Quantification of-pollution
impacts
EVALUATION OF ALTERNATIVE
CONTROL STRATEGIES
ENVIRONMENTAL ASSESSMENT
AND OUTPUTS
• Technology transfer documents
t Standards development
recommendations
• Control technology development
recommendations
FIGURE 1
GENERALIZED ENVIRONMENTAL ASSESSMENT METHODOLOGY
-------
2. Assessment of the health and ecological impacts of the combustion process
and its effluents on the environment:
Identification of environmental (health and ecological) impacts.
Development of environmental goals and objectives.
Comparison of impacts with environmental goals and objectives.
Assessment of the magnitude of pollution impacts.
3. Evaluation of alternative control strategies to reduce pollution impacts
to acceptable levels.
The EA procedure used in the CCEA program and in this specific study is
shown in the generalized methodology diagram (Figure 2).
It is the goal of the CCEA program to integrate ongoing projects and
recommend new efforts to address all practical combinations of information.
It is expected that EPA/IERL-RTP, with the assistance of contractors with
experience and expertise in the various areas associated with the
comprehensive environmental assessment of conventional combustion processes,
will implement and expand the CCEA program as needs dictate and as resources
permit.
Multimedia Effects of Coal Conversion
A major goal of the CCEA program is to evaluate the effects of
implementation of the National Energy Plan, which calls for the increased use
of coal to meet the Nation's energy requirements. Since fuel switching from
oil to coal is an important facet of the NEP, the CCEA program initiated a
study to evaluate the environmental effects of oil and coal combustion in a
controlled industrial boiler in order to compare environmental, energy, and
societal impacts of firing coal vs. firing oil. In order to conduct the
comparative assessment, it was necessary to fully characterize feed streams,
emissions, and effluents from the industrial boiler selected for study and all
associated pollution control equipment.
Plant Description
The site chosen was the Pottstown, Pennsylvania, plant of the Firestone
Tire and Rubber Company, with Firestone's agreement and cooperation. Boiler
No. 4, one of four boilers which supply process and heating steam to the
plant, was used in the assessment. The boiler burns either coal or oil and
has a pilot FMC dual alkali flue gas desulfurization system designed to treat
approximately one-third of the boiler flue gas.
Boiler No. 4 is a dry, bottom, once-through integral furnace, Babcock and
Wilcox (Type FH-18) unit. (See Table 1 for boiler specification data and
Figure 3 for a schematic of the boiler and associated equipment.) When it was
installed in 1958, the boiler was designed as a coal-fired unit but was
converted to fire either coal or oil in 1967. The changeover from one fuel to
the other can be accomplished in less than 30 minutes.
1026
-------
COMBUSTION PROCESSES
AND EFFLUENTS
CHARACTERIZATION
HEALTH AND
ECOLOGICAL IMPACTS
IDENTIFICATION
ENVIRONMENTAL
GOALS AND OBJECTIVES
DEVELOPMENT
REDEFINE
DATA BASE
o
Ni
-vl
CONTROL TECHNOLOGY
DEVELOPMENT
RECOMMENDATIONS
ARE
GOALS
MET?
ALTERNATIVE CONTROL
STRATEGY
EVALUATION
POLLUTANTS
IMPACTS
QUANTIFICATION
STANDARDS
DEVELOPMENT
RECOMMENDATIONS
FIGURE 2
GENERAL METHODOLOGY
-------
EXHAUST
GAS TO STACK
SCRUBBER
FEED SOLIDS EXHAUST GAS
TO STACK
EXHAUST
GAS
l\
FGD
SCRUBBER
Cf
\
\
KEI
LANDFILL
MAKEUP WATER
IMULTI CLONES
EXHAUST
GAS
FLY ASH
FLY ASH
STORAGE
FEEDWATER FROM
PRETREATMENT UNIT
STEAM
DRUM
COAL
IN
STEAM
= SAMPLING POINT
MUD
DRUM
TO MUNICIPAL
SEWAGE
TREATMENT
FIGURE 3
BOILER SYSTEM SCHEMATIC
1028
-------
For test purposes, Firestone agreed to fire one fuel and then the other
as long as required to conduct the appropriate sampling.
TABLE I BOILER NUMBER 4 DESIGN DATA
Boiler Type: Oil/Pulverized coal;
face fired;
integral furnace;
dry bottom
Manufacturer: Babcock and Wilcox, Type FH-18
Type of Burner: Circular conical
Number of Burners: 3
Burner Arrangement Triangular, one face
Air Preheater: Yes
Fuel: No. 6 fuel oil;
High volatile Pennsylvania bituminous
coal;
Class II, Group 2, of ASTM D388;
from Island Creek Coal Co.
Design Steam Rate: 45,000 kg/hr (100,000 Ib/hr);
1.4 MPa (190 psi);
at approximately 193°C (380°F)
Use: Process steam
The two fuels are usually not burned simultaneously except when
converting from oil to coal firing. The coal is ignited by continuing oil
firing until a stable coal flame is obtained. Oil is fired simultaneously
with coal to maintain^ acceptable steam generation rates when coal with a low
heat content is burned. Fuel analyses are given in Table 2.
Control Devices
The flue gases are treated by an air pollution system which consists of
multiclone units and a pilot FGD unit. The multiclones are the primary
particle control device. All of the flue gas passes through the multiclones
after which the stream is split: two-thirds of the flue gas is ducted to the
stack; and the other one-third is ducted to the pilot FGD system which removes
S09 and additional particles. There are no NO controls on the system.
^ X
1029
-------
TABLE 2 SUMMARY OF AVERAGE ULTIMATE COAL AND OIL ANALYSES
Weight %
Species Coal cr^ Oil of
Moisture 7.15 0.86
Carbon 72.10 1.07 86.28 0.39
Hydrogen 4.28 0.06 10.92 0.03
Nitrogen 0.92 0.07 0.36 0.06
Chlorine 0.12 0.02
Sulfur 1.64 0.23 1.96 0.08
Ash 9.90 0.85 0.02 0.004
Oxygen 3.89 0.23 0.46 0.40
kJ/kg (Btu/lb) 29,485 (12,686) 459 40,741 (17,528) --
cra One standard deviation.
The collection efficiency of the multiclone varies as a function of the
particle size distribution and grain loading. Typically, raulticlones remove
90 percent of those particles with diameters of lOpm and greater, and 50 to 80
percent of those particles with diameters of 3|Jm and greater. The collection
efficiency of multiclones drops off rapidly for particles less than 3(Jm
diameter.
The flue gas desulfurization (FGD) system was designed and manufactured
by FMC Corporation. The FGD system is a pilot unit designed to handle 280
acm/min (10,000 acfm) of flue gas, which is approximately one-third of the
volume of the flue gas from the boiler. The pilot plant was placed on-line
in January 1975. Figure 4 is the basic flow diagram of the FMC FGD system, as
applied at this site.
The flue gas (stream 1) is withdrawn downstream of the boiler on the exit
side of the multiclone dust collectors. Fly-ash loading at the scrubber inlet
is substantially higher during co'al-firing than during oil-firing. To
accommodate the wide variation in fly-ash loading, the FGD system was designed
to operate with or without fly-ash, and can be operated on either fuel.
1030
-------
CYCLONIC MIST
ELIMINATOR
MESH MIST ELIMINATOR
VENTURI
SCRUBBER
SLURRY
THICKENER
ROTARY DRUM
FILTER
LEGEND:
1. BOILER FLUE GAS TO SCRUBBER
2. SCRUBBER OUTLET TO ATMOSPHERE
3. SOLID WASTE TO LANDFILL
4. ABSORBENT SOLUTION TO SCRUBBER
&. ABSORBENT SOLUTION TO REGENERATION
G. SODIUM CARBONATE MAKEUP
7. REGENERATION SOLUTION
8. REGENERATED SCRUBBER SOLUTION
9. CONCENTRATED SLURRY
10. RETURNED SCRUBBER SOLUTION
FIGURE 4
FMC UNIT AT THE INDUSTRIAL FACILITY
-------
Upon entering the FGD unit, the flue gases are contacted with a slightly
acidic scrubbing solution (stream 4) which removes SCL and particles. The SCL
and particles are removed at the scrubber throat and carried away in the
scrubbing solution. The process utilizes a sodium sulfite/sodium bisulfite
solution as the absorbent. The basic reaction for SC- removal is:
Na2S03 + S02 + H20 -> 2NaHS03< (1)
A bleed stream (stream 5) of the scrubbing solution is removed from the
system at a rate which keeps the pH of the solution in an acceptable range.
The bleed stream is reacted with calcium hydroxide in a short retention time,
agitated vessel to regenerate the sodium sulfite. The basic chemistry of
sodium sulfite regeneration is:
2NaHS03 + Ca(OH)2 •> CaSCy ^0 + ^0 + Na^Og. (2)
The slurry of precipitated sulfur compounds (stream 8) is concentrated
and pumped to a rotary drum filter where the essentially clear liquid is
separated from the solid waste products. The clear liquid (stream 10) is
returned to the system for further utilization. The solid wastes, in the form
of filter cake containing 40 weight percent water (stream 3), are removed from
the rotary drum filter and conveyed to a storage bin to await transportation
to the dump site. Mainly due to the heavier particle loading, more filter
cake is produced during coal firing than during oil firing.
The on-site landfill, which is the final disposal facility for flyash and
scrubber cake generated at the facility, has several test wells from which
samples are collected every 3 months and sent to an independent laboratory for
analysis. Monthly tests are conducted by plant personnel to monitor Na and
specific conductivity. With permission from the Pennsylvania Department of
Environmental Resources, this site is being used as an experimental disposal
area for the filter cake from the FMC unit.
Test Description and Conditions
Multimedia emission tests were conducted on Boiler No. 4 of the Firestone
Plant from September 27 through October 8, 1977- Gaseous and solid emissions
were sampled during coal and oil firing to obtain data for the assessment.
Flue gas was sampled before and after the scrubber to determine which
pollutants are removed or modified by the control device. Sampling points
used are indicated on the process diagram, Figure 3.
Emissions were characterized using EPA's phased approach. This approach
utilizes two levels of sampling and analysis (Level 1 and Level 2). Level 1
screening procedures are accurate within a factor of 2 to 3. They provide
preliminary assessment data and identify problem areas and information gaps.
Based on these data, a site specific Level 2 sampling and analysis plan is
developed. Level 2 provides more accurate and detailed information to confirm
and expand on the information gathered in Level 1. The methods and procedures
used for Level 1 are documented in the manual, "Combustion Source Assessment
Methods and Procedures Manual for Sampling and Analysis", September 1977, (in
press). This Level 2 methods and procedures include "state-of-the-art" techniques
required for this particular site.
1032
-------
Normally all Level 1 samples are analyzed and evaluated before moving to
Level 2. Because of the program time constraints, the Level 1 and Level 2
samples were obtained during the same test period; however, analysis of the
samples did proceed in a phased manner except where sample degradation was of
concern. In that case, Level 2 analysis was performed on the sample prior to
Level 1 completion.
Gaseous Effluents
The boiler flue gas was sampled at the inlet and the outlet of the pilot
flue gas desulfurization unit. Integrated bag samples were taken at both
points during each test. On-site analyses of CO., 0_, N~ and C. - C, organics
were conducted on the bag samples. Continuous monitors were used to analyze
CO, NO, N02, NO , SO and total hydrocarbons (as CH,). Figure 5 is a
schematic of the continuous monitor setup. A Thermal Electron Corporation
(TECO) gas conditioner was used to remove condensate and particulate matter
from the gas sample. Gaseous streams were isokinetically sampled at each
location during all tests using four different sampling trains.
The Source Assessment Sampling System (SASS) was used to collect Level 1
gaseous and particulate emission samples at the scrubber inlet and outlet.
The SASS train is illustrated in Figure 6. The train consists of a heated
probe, three cyclones and a filter in a heated oven. The cyclones were used
only during the coal inlet tests. During the other tests, the particle
loadings were too low for the cyclones to work effectively. The remainder of
the system consists of a gas conditioning system, an XAD-2 polymer absorbent
trap and a series of impingers. The polymer traps 'gaseous organics and some
inorganics and the impingers collect the remaining inorganics. All sample
contact surfaces are Type 316 stainless steel, Teflon, or glass. The train
was run for 6 to 8 hours so that a minimum of 30 cubic meters of gas was
collected.
Previous sampling and analysis experience had indicated that SASS train
materials may contaminate certain organic and inorganic samples. The
contamination is of concern only when the pollutant is present at a
concentration that is near the detection limit of the Level 2 methods. To
avoid that possibility, all glass sampling trains were used to collect Level 2
samples. Method 5 sampling trains were modified as shown in Figure 7 for
organics and Figure 8 for inorganics. Both trains sampled approximately 10
cubic meters of flue gas during a 6 to 8 hour test run.
A controlled condensate train (Goksoyr-Ross), shown in Figure 9, was used
at each location to obtain samples for S0_, S0_ (as BLSO,), particulate
sulfate, HC1, and HF.
During Level 2 test runs, Andersen cascade impactors were used to obtain
particle samples by particle size fraction. A pre-separating 10pm cyclone was
used up-stream of the impactor on the inlet side.
1033
-------
EXHAUST
GAS INLET
?
FGD UNIT
HEAT TRACED SAMPLE LINES
t
EXHAUST
GAS OUTLET
GAS CONDITIONER
FIGURE 5
FLUE GAS CONTINUOUS MONITOR SETUP
NGER
10 CFM VACUUM PUMPS
FIGURE 6
SOURCE ASSESSMENT SAMPLING SYSTEM (SASS) SCHEMATIC
1034
-------
SORBENT MODULE
NOZZLE
N,
X
GAS FLOW
FIGURE 7
ORGANIC SAMPLING TRAIN
NOZZLE
\
s
s
s
\
X
\
s
s
\
V,
\
\
s
s
s
FILTER
GREENBURG-SMITH
IMPINGERS
CYCLONE
GLASS LINED SS PROBE
-.^-.-.-.-.;::?::--'--..--*=S
GAS FLOW
PITOT TUBE
UMBILICAL CORD
FIGURE 8
INORGANIC SAMPLING TRAIN
1035
-------
ADAPTER FOR CONNECTING HOSE
TCWELL
ASBESTOS CLOTH
INSULATION
STACK
GLASS-COL
HEATING
MANTLE
X
RUBBER VACUUM
HOSE
DRY TEST
METER
THREE WAY
VALVE
SILICA GEL
8% Na2CO3
RECIRCULATOR
THERMOMETER
STYROFOAM
ICE CHEST
FIGURE 9
CONTROLLED CONDENSATION TRAIN
Solid Effluents
Composite samples of the flyash and scrubber filter cake were collected
according to Level 1 procedures and returned to the laboratory for analysis.
Grab samples of the scrubber feed solids were also obtained for laboratory
analyses.
Laboratory Analyses
The samples from the various sampling trains were returned to the
laboratory for analysis. Detailed analysis procedures can be found in the
manual "Combustion Source Assessment Methods and Procedures Manual for
Sampling and Analysis", September 1977, (in press).
Level 1 analyses for particles and gases were made for inorganics by SSMS
and for selected anions and organics by LC, IR, and MS. Solids, slurries and
liquids were similarly analyzed, although the work-up procedures were
different.
More detailed and more quantitative Level 2 analyses were performed to
identify and quantitate specific compounds indicated by the Level 1 analyses.
1036
-------
TABLE 3 SUMMARY OF TEST CONDITIONS
Test
No.
200
201-1
201-2
201-3
201-4
§ 202-1
202-2
202-3
202-4
203
Steam Production Rate
kg steam/hr
39,700
44,200
43,100
34,000
40,800
45,000
45,400
44,200
42,200
31,800
Ib steam/hr
87,500
97,500
95,000
75,000
90,000
100,000
100,000
97,500
93,000
70,000
% of Nominal Fuel
Maximum Feed Rate,
Boiler
87.
97.
95.
75.
90.
100
100
97.
93.
70.
Load kg/hr gal/hr
COAL FIRING
5 3629
5 3629
0 3629
0 3175
0 3629
OIL FIRING
900
900
5 880
0 805
0 600
I 02 at
Scrubber
Inlet3
7.8
8.2
8.4
8.3
6.7
5.8
6.3
6.1
4.0
Not
Measured
Estimated %
Excess Air,
to Furnace
20
20
20
20
20
21
21
21
21
21
Due to air leaks in ducting upstream of the scrubber inlet, tabulated 0~ values are not
representative of combustion zone 0~ concentrations, which normally range from 3 to 4%
for this unit.
excess air is estimated to be 100 x
02 - CO/2
0.264 N2 - (02/2)
, where 02 was assumed to be
3.5% and other species concentrations were computed from fuel analysis.
-------
Test Conditions
Ten tests were performed with the industrial boiler, five each with coal
and oil. Unit loadings ranged from 31,800 to 45,400 kg steam per hour (70,000
to 100,000 Ib per hour), which corresponds to between 70 and 100 percent of
full load operation. Specific test conditions are summarized in Table 3.
Test data relating to scrubber throughput and loading and total flue gas
generation rates are presented in Table 4. The scrubber is a pilot unit which
does not process the entire flue gas output of the furnace. From 11 to 32
percent of the total flue gas was processed through the scrubber during the
tests. Typical inlet and outlet gas temperatures for the scrubber unit were
149°C and 52°C (300°F and 125°F). Only 51 to 57 percent of design loading,
rather than full loading, was maintained during coal-fired testing because
failure of the multiclone particle removal system upstream of the scrubber
resulted in high solids loading at the scrubber and unacceptably high scrubber
filter-cake production rates. During oil-fired testing, 88 to 100 percent of
full design flows were maintained.
Analytical results were used to estimate total boiler emission on the
basis of treatment of 100 percent of the flue gas from the boiler. That is,
it was assumed that additional scrubber modules could be added in parallel to
the system such that the total flue gas output would be processed with a mean
scrubbing efficiency equal to that of the pilot scrubber.
All stack emissions data are based on this assumption.
TABLE 4 FRACTION OF FLUE GAS PROCESSED BY SCRUBBER
Flow Rate
at Scrubber
Inlet
dscm/min
Average 96
% of Design
Load
COAL FIRING
54
Total
Flue Gas
Flow Rate
dscm/min
741
Fraction of
Total Flue Gas
Processed by
Scrubber
0.13
OIL FIRING
Average 180 102 737 0.25
o
Dry standard cubic meters per minute
1038
-------
MULTIMEDIA EMISSION RATES
Gaseous Emissions
Particulate Matter
Particle concentration during coal firing at the inlet to the scrubber
averaged 2951 ng/J (6.86 lb/10 Btu). Since the multiclone unit, upstream of
the scrubber, failed during the test period and thus removed little or no
particulate material, this may be taken to be representative of uncontrolled
emissions. The outlet concentration after scrubbing was 18.6 ng/J (0.04
lb/10 Btu), which corresponds to 99.4 percent particle removal efficiency.
The uncontrolled particle emission rate during oil firing was 113 ng/J
(0.26 lb/10 Btu). After scrubbing it was 17.6 ng/J (0.04 lb/10 Btu) at the
outlet, for a removal efficiency of 84.4 percent. These data are summarized
in Table 5.
Particle emissions after scrubbing are well below the existing NSPS of 43
ng/J (0.10 lb/10, Btu) but slightly higher than the proposed limitation of 12
ng/J (0.03 lb/10° Btu).
TABLE 5 PARTICLE CONCENTRATION
ng/J (lb/10 Btu)
Coal Oil
Scrubber Inlet 2951 (6.86) 113 (0.26)
Scrubber Outlet 18.6 (0.04) 176 (0.04)
NSPS 43 (0.10)
Proposed Limit 12 (0.03)
Size Distribution
The particle size distribution at the scrubber inlet during coal firing
was determined with a polarizing light microscope on a filter catch, due to
the high loading. These number percent results were converted to aerodynamic
diameter and weight percent by assuming that all particles had the same
density. This is a reasonable assumption because the major components of the
particle generated from coal combustion, aluminosilicates and iron oxides, are
known to partition equally among small and large sizes. With the constant
density assumption, the weight distribution in each size range would be
proportional to the product of the number distribution and the particle volume
representing the size range. The particle volume was calculated based on the
geometric mean diameter for the size range.
1039
-------
Size distributions of particles from coal firing at the scrubber outlet
and of particles from oil firing at both scrubber inlet and outlet were
determined with an Andersen cascade impactor. The impactor calibration gives
aerodynamic diameter and weight distribution directly.
Tables 6 and 7 show a significant change in the particle size
distribution before and after scrubbing. Large particles are removed with
greater efficiency than are small particles. Table 7 indicates that the mass
emission rate of the smallest particles increases after scrubbing, suggesting
that fine particles are generated within the scrubber.
TABLE 6 SCRUBBER INLET AND OUTLET PARTICLE SIZE DISTRIBUTION (Weight %)
Coal
Aerodynamic Diameter
Size Range (pm)
<1
1-3
3-10
>10
Inlet
0.0017
0.041
2.24
97.7
Outlet
62
30
7
1
Oil
Inlet
20
1
74
5
Outlet
83
12
5
0
TABLE 7 PARTICLE EMISSION RATES AND REMOVAL EFFICIENCY
Aerodynamic
Size Range (pm)
10
Coal
-kg/hr-
Inlet Outlet
0.0055 1.30
0.13 0.63
7.3 0.15
316.5 0.021
% Removal
Efficiency
<0
<0
97.9
>99.9
Oil
-kg/hr-
Inlet Outlet
4.48
0.22
16.6
1.12
2.27
0.33
0.14
.0.00
% Removal
Efficiency
49.2
<0
97.4
100
Total 324.0 2.10 99.3 22.4 2.74 87.8
1040
-------
Sulfur Compounds
The average SO- emission rate from coal firing ahead of the scrubber,.was
1112 ng/J (2.59 lb/10 Btu) and after scrubbing was 36.3 ng/J (0.08 lb/10°),
for a mean scrubber efficiency of 96.7 percent.
During oil firing, the SO- emission rates were 993 ng/J (2.3 lb/10 )
ahead of the scrubber and 26.8 ng/J (0.06 lb/10 Btu) after the scrubber, for
a mean scrubber efficiency of 97 percent.
In both cases, the SO- emissions after the FGD system were substantially
below existing and proposed NSPS emission limit, see Table 8.
TABLE 8 SULFUR DIOXIDE EMISSIONS
ng/J (lb/10b Btu)
Scrubber Inlet
Scrubber Outlet
NSPS
Efficiency, %
Coal
1112 (2.59)
36.3 (0.08)
520 (1.2)
96.7
Oil
993 (2.31)
26.8 (0.06)
344 (0.80)
97
The scrubber removed 95-97 percent of the SO- during coal firing and
97-98 percent during oil firing. Only 32-33 percent of the SO- was removed
during coal firing and 28-29 percent during oil firing. This relatively poor
removal efficiency for S0_ is an indication that SO- is either present as very
fine aerosols in the scrubber inlet gas or is converted to very fine aerosols
as the flue gas is rapidly cooled in the scrubber.
The removal rate for SOT was 88 percent during coal firing and 60 percent
during oil firing, indicating that most of the SOT in the scrubber inlet is
associated with larger particulates. However, combustion generated sulfates
may not be simply passing through the scrubber. Because of the possibility
that the SO? species from coal combustion may be. changed by the scrubbing
process, an analysis effort to determine the SO, species was initiated. Both
the Fourier Transform IR (FTIR) analysis and the X-Ray Diffraction (XRD)
analysis have confirmed the presence of sodium bisulfate (NaHSO,) in the
scrubber outlet, but not in the scrubber inlet. This is positive proof that
sulfates are generated within the scrubber as the result of oxidation of
sodium bisulfite (NaHSO-) and sodium sulfite (Na_SO ) and emitted in the
scrubber effluent gas. Also, tests on boilers with flue gas concentrations of
400 to 8,000 ppm SO- have shown that there is no correlation between initial
1041
-------
4
SCL concentration and the net sulfate formation rate . This implies that the
scrubber has a minimum sulfate emission rate that is virtually unaffected by
inlet SO- concentration. These data must, however, be evaluated in the
context of the potential for significantly increased sulfate loadings to the
environment which would result from SCL emissions if the boiler flue gases
were not controlled.
Based on the analysis of SCL and SO, emission data, it has been estimated
that up to 40 percent of the fine particle emissions at the scrubber outlet
could be attributed to scrubber generated NaHSO,. The remaining portion of
the net increase in fine particles across the scrubber may be attributable to
the uncertainties associated with the assumptions used in converting the
polarizing light microscope number size distribution data to weight size
distribution, and to calcium sulfite hemihydrate (CaSO *l/2 H_0) particles
generated by the scrubber. This should be confirmed by further study since
the unknown or unexplained portion amounts to more than half of the net
increase.
Nitrogen Oxides
Mean NO emissions during coal firing tests were 421 ng/J (0.98 lb/10
Btu) into the scrubber and 372 ng/J (0.87 lb/10 Btu) out of the scrubber.
The comparable data during oil firing were 168 ng/J (0.39 lb/10 Btu) and 161
ng/J (0.37 lb/10 Btu). As expected, little or no NO was removed.
X
Emissions of NO varied ±5 percent around these figures. An air leak in
the sampling line from the scrubber outlet intermittently allowed sample
dilution. Hence, any indicated removal of NO by the scrubber is actually a
sampling phenomenon, not a real reduction. Tfiis was confirmed by later
testing.
The NSPS limit is 300 ng/J (0.70 lb/106 Btu) for coal firing and 127 ng/J
(0.30 lb/10 Btu) for oil firing, neither of which was met, except at reduced
furnace loads.
Carbon Monoxide
Uncontrolled CO emissions averaged 15,9 ng/J (0.04 lb/10 Btu) during
coal fired tests and 5.47 ng/J (0.01 lb/10 Btu) during oil firing.
Emissions after scrubbing varied ±5 percent around these figures, with no
clear trends depending on test parameters. The analytical sensitivity for CO
is about 15 percent of the measured value at these concentrations, hence the
small changes across the scrubber are of no significance. Furthermore, an air
leak in the sampling line from the scrubber outlet intermittently allowed
sample dilution. Thus, any indicated removal of CO by the scrubber is
actually a sampling and analytical phenomenon, not a real reduction.
Organics
Uncontrolled hydrocarbon emissions, measured as methane, averaged 5.79
ng/J (0.01 lb/10 Btu) for coal firing and 2.49 ng/J (0.01 lb/10 Btu) for oil
firing.
1042
-------
Analytical results from the scrubber outlet are not available due to
sample handling problems. Indications from FID, GC, and gravimetric methods
for C.-C.., C -C ,, and >C,, organics, respectively, are that, very
approximately, 75 percent of organics over C, are removed, either in the
scrubber or in the gas conditioner in the analytical train.
Polycyclic organic material (POM) was not found in the scrubber inlet or
outlet samples from either coal or oil firing at the detection limit of 0.3
mg/m . Since two POM's (benzo-a-pyrene~and dibenzo(a,h) anthracene) have been
assigned MATE values below the 0.3 mg/m detection limit, additional testing
with increased analytical sensitivity is indicated. Also, a more accurate
determination of oxygen in the flue gas at the furnace outlet could be
important since POM levels decrease as excess air increases at constant
temperature.
Inorganics
The emission concentrations during coal firing for 22 major trace
elements at the scrubber inlet and outlet are presented in Table 9. To assess
the potential degree of hazard of these emissions, the emission concentrations
are divided by the Minimum Acute Toxicity Effluent (MATE) values. The MATE
values are emission level goals developed under the direction of EPA, and can
be considered as concentrations of pollutants in undiluted emission streams
that will not adversely affect persons or ecological systems exposed for short
periods of time (less than 8 hours) . MATE values for air derived from human
health considerations are used as the basis for comparison here.
As shown in Table 9 for coal firing, of the 22 trace elements presented,
17 exceed their MATE values at the scrubber inlet and 5 at the scrubber
outlet. The four trace elements in the scrubber outlet that pose a potential
hazard are arsenic, chromium, iron, and nickel. Additionally, it may be noted
that the emission concentration of beryllium at the scrubber outlet is equal
to its MATE value. At this emission concentration, the total beryllium
emissions from boilers greater than 50 MW in capacity would amount to more
than 10 grams per day and exceed the National Emission Standard for Hazardous
Air Pollutants.
In Table 10, the emission factors and the mass emission rates for the 22
major trace elements during coal firing at the scrubber inlet and outlet are
presented. The mass emission rates were used £° calculate the removal
efficiency for these trace elements by the scrubber. The overall removal
efficiency for these trace elements is approximately 99.5 percent. As
indicated in Table 10, however, some of the trace elements were not removed
as effectively as others.
To better understand the removal efficiency of the individual trace
elements, the enrichment factor has been computed for each trace element
across the scrubber. The enrichment factor is defined here as the ratio of
the concentrations of trace element to aluminum in the scrubber outlet,
divided by the corresponding ratio in the scrubber inlet. Aluminum is
selected as the reference material because it has been shown to partition
1043
-------
TABLE 9 EMISSION CONCENTRATIONS OF TRACE ELEMENTS DURING COAL FIRING
Trace
Element
Beb
HgC
Ca
Mg
Sb
As
Bb
Cd
Cr
Co
Cu
Fe
Pb
Mn
Mo
Ni
V
Zn
Se
Sr
Al
Zr
Total
Scrubber
Inlet
mg/m
0.1
0.011
74
19
3.7
7.8
0.2
0.47
2.6
3.6
9.6
450
8.5
0.78
10
1.4
3.1
2.3
3.2
11
480
1.6
1100
Scrubber
Outlet
/ 3
mg/m
0.002
0.005
0.036
0.011
0.025
0.22
0.03
0.0010
0.13
0.012
0.020
2.4
0.021
0.015
0.027
0.063
0.058
0.048
0.099
0.058
2.6
0.018
6.2
MATE
Value
mg/m
0.002
0.05
16
6.0
0.050
0.002
3.1
0.010
0.001
0.050
0.20
1.0
0.15
5.0
5.0
0.015
0.50
4.0
0.200
3.1
5.2
5.0
Potential Degree
Scrubber
Inlet
50
0.22
4.6
3.2
74
3900
0.07
47
2600
72
48
450
57
0.16
2.0
93
6.2
0.58
16
3.5
92
0.32
•a
of Hazard
Scrubber
Outlet
1.0
0.10
0.002
0.002
0.5
110
0.01
0.1
130
0.24
0.10
2.4
0.14
0.003
0.005
4.2
0.12
0.012
0.50
0.019
0.5
0.004
-
Potential degree of hazard is defined as the ratio of the discharge concentration
to the MATE value.
Approximate values as determined by Spark Source Mass Spectrometry (SSMS).
The other values presented are determined by Inductively Coupled Plasma
Optical Emission Spectroscopy (ICPOES).
Mercury was determined by cold vapor analysis of SASS train samples.
1044
-------
TABLE 10 EMISSION FACTORS AND MASS EMISSION RATES OF
TRACE ELEMENTS DURING COAL FIRING
Trace
Element
Be"
Hg"
Ca
Mg
Sb
As
Ba
Cd
Cr
Co
Cu
Fe
Pb
Mn
Mo
Ni
V
Zn
Se
Sr
Al
Zr
Total
Emission Factor, ng/J
Scrubber
Inlet
0.04
0.08
32
«.2
1.6
3.4
0.1
0.20
1.1
1.6
4.1
190
3.7
0.34
4.3
0.60
1.3
0.99
1.4
4.7
210
0.69
470
Scrubber
Outlet
0.001
0.037
0.015
0.0046
0.010
0.092
0.01
0.00042
0.054
0.0050
0.0084
1.0
0.0088
0.0063
0.026
0.026
0.024
0.020
0.041
0.024
1.1
0.0075
2.6
Emission Rate, g/hr
Scrubber
Inlet
5
0.50
3300
860
170
350
10
21
120
160
430
20,000
380
35
450
61
140
100
140
500
22,000
72
50,000
Scrubber
Outlet
0.09
0.23
1.6
0.48
1.1
9.7
1.2
0.044
5.7
0.53
0.88
110
0.92
0.68
1.2
2.8
2.5
2.1
4.3
2.5
110
0.79
270
Removal
Efficiency
98
55
99
99
99
97
88
99
95
99
99
99
99
98
99
95
98
98
97
99
99
99
99
Enrich-
ment
Factor
3.7
84
0.09
0.11
1.2
5.3
2.1
0.4
9.5
0.6
0.4
0.99
0.5
3.4
0.5
8.6
3.6
3.9
5.8
0.9
1.0
2.1
Approximate values as determined by SSMS. The other values were determined
by ICPOES analysis.
Mercury was determined by cold vapor analysis of SASS train samples.
1045
-------
equally among particles of different size*. The enrichment factors presented
in Table 10 show that beryllium, mercury, antimony, arsenic, boron, chromium,
manganese, nickel, vanadium, zinc, selenium, and zirconium are enriched across
the scrubber. The enrichment observed is due primarily to the partitioning of
trace elements as a function of particle size, and the greater collection
efficiency of the scrubber for the large size particles. It may also be noted
that many of the trace elements that show an enrichment trend, such as
mercury, selenium and arsenic, either occur as element vapors or form volatile
oxides and halides at furnace temperatures. Condensation and surface
absorption of the more volatile elements or their oxides and halides
downstream of the furnace could, therefore, result in higher concentrations of
these trace elements on smaller particles.
Concentrations of 22 major trace elements present in the flue gas during
oil firing at the scrubber inlet and outlet are presented in Table 11. MATE
values for these elements are also presented for comparison.
Of the 22 elements analyzed, 11 exceed their respective MATE values at
the scrubber inlet and 5 exceed their MATE values at the scrubber outlet. The
five elements exceeding their MATE values at the scrubber outlet are arsenic,
cadmium, chromium, nickel, and vanadium.
3
Beryllium emissions were measured to be 0.001 mg/m after scrubbing, •
corresponding to half the MATE value for this element. At this emission
concentration, the National Standard for Hazardous Air Pollutants limitation
of 10 grams beryllium per day would only be exceeded by boilers of 100 MW
capacity or greater.
Emission factors and mass emission rates during oil firing for the 22
elements analyzed are presented in Table 12. Also presented in Table 12 is
the scrubber removal efficiency for each element. An overall removal
efficiency of 87 percent was obtained for these elements, although several
elements were removed with less efficiency, i.e., calcium, arsenic, cadmium,
nickel, and vanadium. Note that, with the exception of chromium, all elements
that exceeded their MATE values at the scrubber outlet were removed with lower
than average efficiency during scrubbing.
The enrichment factor across the scrubber 'has been computed for each
element and is presented in the last column of Table 12.
^Silicon, iron, and scandium have also been used by other investigators as
the reference element in the computation of enrichment factors. Notice
that iron has no enrichment in this study while silicon and scandium were
not measured.
1046
-------
TABLE 11 EMISSION CONCENTRATIONS OF TRACE ELEMENTS DURING OIL FIRING
Element
Beb
HgC
Ca
Mg
Sb
As
Bb
Cd
Cr
Co
Cu
Fe
Pb
Mn
Mo
Ni
V
Zn
Se
Sr
Al
Zr
Total
Scrubber
Inlet
mg/m
<0.001
0.0016
0.41
0.31
0.062
0.15
0.53
0.28
0.17
0.10
0.54
4.8
0.20
0.03
0.22
1.1
2.7
0.61
0.050
0.043
5.7
0.015
18
Scrubber
Outlet
mg/m
0.001
0.0002
0.070
0.030
0.006
0.030
0.039
0 . 066
0.018
0.012
0.007
0.28
0.013
0 . 004
0.025
0.20
0.82
0.065
0.006
0.001
0.48
0.001
2.5
MATE
Value
mg/m
0.002
0.05
16
6.0
0.50
0.002
3.1
0.010 •
0.001
0.050
0.20
1.0
0.15
5.0
5.0
0.015
0.50
4.0
0.200
3.1
5.2
5.0
Potential Degree
Scrubber
Inlet
<0.50
0.032
0.026
0.052
0.124
75.0
0.171
28.0
170
2.0
2.70
4.8
1.333
0.006
0.044
73.3
5.40
0.153
0.25
0.014
1.096
0.003
»a
of Hazard
Scrubber
Outlet
0.50
0.004
0.004
0.005
0.012
15.0
0.013
6.60
18.0
0.24
0.035
0.28
0.087
0.001
0.005
13.33
1.640
0.016
0.03
0.0003
0.092
0.0002
Potential degree of hazard is defined as the ratio of the discharge concentration
to the MATE value.
Beryllium was determined by Spark Source Mass Spectrometry (SSMS). The
other values, with the exception of mercury, are determined by Inductively
Coupled Plasma Optical Emission Spectroscopy (ICPOES) analysis.
f*
Mercury tfas determined by cold vapor of SASS train samples.
1047
-------
TABLE 12 EMISSION FACTORS AND MASS EMISSION RATES OF
TRACE ELEMENTS DURING OIL FIRING
Element
Be3
Hgb
Ca
Mg
Sb
As
B
Cd
Cr
Co
Cu
Fe
Pb
Mn
Mo
Ni
V
Zn
Se
Sr
Al
Zr
Total
Emission Factor,
ng/J
Scrubber
Inlet
<0.0003
0.0006
0.13
0.10
0.02
0.049
0.17
0.091
0.055
0.033
0.18
1.6
0.065
0.010
0.072
0.36
0.88
0.20
0.016
0,014
1.9
0.0049
6.0
Scrubber
Outlet
0.0003
0.0001
0.022
0.0094
0.0019
0.0094
0.012
0.021
0.0057
0.0038
0.002
0.088
0.0041
0.0013
0.0079
0.063
0.26
0.02
0.002
0.0003
0.15
0.0003
0.78
Emission
g/hr
Scrubber
Inlet
<0.04
0.05
16
12
2.5
5.9
21
11
6.7
3.9
21
190
7.9
1.2
8.7
43
110
24
2.0
1.7
220
0.59
710'
Rate,
Scrubber
Outlet
0.04
0.006
2.7
1.1
0.23
1.1
1.5
2.5
0.69
0.46
0.27
11
0.50
0.15
0.95
7.7
31
2.5
0.23
0.038
18
0.038
96
Removal
Efficiency
Unknown
87
83
91
91
81
93
77
90
89
99
95
94
87
89
83
71
90
87
98
92
94
87
Enrichment
Factor
>11.9
1.48
2.03
1.15
1.15
2.37
0.87
2.80
1.26
1.43
0.15
0.69
0.77
1.58
1.35
2.16
3.61
1.27
1.43
0.28
1.0
0.79
Beryllium was determined by SSMS. The other elements, except mercury, were
determined by ICPOES.
Mercury was determined by cold vapor analysis of SASS train samples.
1048
-------
Chloride, Fluoride, Nitrate
The results of anion analyses on extracts of particulate matter collected
at the inlet and outlet of the scrubber are presented in Table 13. Fluoride
is removed with reasonable efficiency, as is to be expected from the overall
high removal efficiencies of the trace element cations with which fluoride may
be associated. The lower removal efficiency for nitrate suggests that it may
be preferentially associated with the fine particles which are not efficiently
removed by the scrubber. The chloride removal is much higher in the coal
fired tests than in the oil fired, which may be anomalous or may reflect the
occurrence of some coal chloride in larger mineral inclusions as contrasted to
organic chlorides in oil which would form smaller particles during combustion.
TABLE 13 CHLORIDE, FLUORIDE, AND NITRATE
EMISSIONS FROM COAL AND OIL FIRING
Fuel Cl F N03
Inlet Outlet Removal Inlet Outlet Removal Inlet Outlet Removal
mg/J mg/J Efficiency mg/J mg/J Efficiency mg/J mg/J Efficiency
Coal 4.7 <0.004 >99 0.22 <0.03 >86 <0.48 <0.25 >52
Oil 0.15 0.072-0.075 52 0.017 0.002-0.003 89 0.076 0.033 57
Solid Emissions
Three solid waste streams are produced by the system:
o Bottom ash
o Fly ash
o Scrubber cake.
Table 14 shows the approximate quantities of bottom ash and scrubber cake that
were produced. Only small quantities of fly ash were collected during the
test because the multiclone malfunctioned.
The scrubber cake produced afeter filtration has the characteristics of
silty soils, but its behavior closely resembles a clay in many respects. As
obtained from the vacuum filter, the scrubber cake consists of small lumps and
appears to be relatively dry; however, the water content generally ranges from
about 30 to 50 percent.
1049
-------
TABLE 14 GENERATION RATE OF SOLID WASTE FROM
10 MW CONTROLLED INDUSTRIAL BOILER
Waste
Bottom ash
Fly ash
Scrubber cake
Rate of Production,
Coal Firing
80
240a
700
kg/hr
Oil Firing
1
5a
400
Q
This is the amount of fly ash recovered by the cyclone collector.
Approximately 25% of the fly ash is recovered in the scrubber
and removed with the scrubber cake.
If it is assumed that calcium sulfite hemihydrate (CaSO«*l/2 H«0) is
formed as a result of the SO- scrubbing and Na_SO_ regeneration processes,
then the mass balance of coal firing in Table 15 shows that the scrubber cake
is composed of 28.5 percent coal fly ash and 23.8 percent CaSO -1/2 H»0.
However, if the multiclone had not malfunctioned during the test, more fly ash
would have been removed upstream of the scrubber and the fly ash content of
the scrubber cake would have been lowered proportionately. The amount of
scrubber cake produced could be reduced to 600-750 kg/hr on wet basis,
assuming approximately 60 to 80 percent multiclone efficiency.
Table 15 also shows the estimated composition of scrubber cake produced
during oil firing. The cake is composed of 44 to 50 percent unbound water and
at least 47 percent calcium sulfite hemihydrate. These data reflect the low
particle emissions which are characteristic of oil firing. Only 1 percent of
the scrubber cake during oil firing is estimated to be primary particle
emissions, due to their small size and low removal efficiency.
Although the scrubber cake material is composed predominantly of
relatively insoluble solids (calcium sulfite, calcium sulfate, and some
calcium carbonate), the interstitial water does contain soluble residues of
lime, sulfate, sulfite, and chloride salts. Trace elements in the fly ash may
also contribute to the leachate from the disposed scrubber cake and are of
special concern. The concentrations of 20 trace elements in the scrubber cake
are presented in Tables 16 and 17. Note that, except for boron, the trace
element concentrations in the scrubber cake from coal firing far exceed the
MATE values for solids. Except for antimony, boron, molybdenum, and zinc, all
trace elements in oil fired scrubber cake were found to exceed human health
based MATE values for solids. Similarly, except for boron, all trace elements
were found to exceed ecology based MATE values for solids. These results are a
consequence of reducing a high volume of low concentration wastes to a low
volume of concentrated wastes. The high potential degree of hazard for most
elements appears to warrant disposal of these solid wastes in specially designed
disposal areas.
1050
-------
TABLE 15 ESTIMATED SCRUBBER CAKE MASS BALANCE
Contribution to Scrubber Cake
Component kg/hr Weight %
Coal Oil Coal Oil
Fly ash removed by scrubber 324 5 29 1
CaS03*l/2 H20 formed from SO 262 210 24 47
scrubbing and Na.SO regeneration
CaSO,, CaCO , Na SO , Ca(OH) 10-85 6-35 1-8 1-8
NaHS04, and Na^SO^ losses testimated)
Water 429-504 193-222 39-46 44-50
Average 1,100 443 100 100
The concentrations of 20 trace elements present in fly ash are shown in
Table 18. Again, in almost every case, the trace element concentration in the
fly ash far exceeded its MATE value for solids. Trace element concentrations
in the bottom ash would be similar to those of the fly ash, except that the
more volatile elements and the elements that form volatile compounds would be
more enriched in the fly ash. Thus, the concentrations of arsenic, antimony,
boron, chromium, manganese, nickel, vanadium, zinc, selenium, and zirconium
would all be lower in the bottom ash.
The overall mass balances for the 20 trace elements have been performed
and the results are summarized in Tables 19 and 20. The percentage of trace
element in the feeds that could be located in the effluent streams (scrubber
cake, scrubber effluent gas, bottom ash, and fly ash) is used as a measure of
mass balance closure. Except for boron, copper, strontium, and zirconium, the
closure of mass balance for the trace elements in coal has been found to be
good.
Good mass balance closure for the trace elements in oil was obtained for
arsenic, boron, chromium, cobalt, copper, molybdenum, nickel, vanadium, zinc,
and selenium. However, as expected due to these extremely low elemental
concentrations, mass balance closure for some elements is poor. Instances in
which the effluent flow rate of an element substantially exceeded the input
feed rates, such as with iron and aluminum, may be the result of the extremely
high elemental concentrations attained during coal firing and subsequent
contamination of the recycle scrubber solution.
1051
-------
TABLE 16 INORGANIC CONTENT OF SCRUBBER CAKE FROM COAL FIRING (DRY BASIS)
Element
Ca
Mg
Sb
As
B
Cd
Cr
Co
Cu
Fe
Pb
Mn
Mo
Ni
V
Zn
Se
Sr
Al
Ar
Total
Concentration
Hg/g
60,715
1,458
315
532
88
13
141
424
112
47,241
297
51
1,117
114
195
282
256
642
45,310
106
159,409
MATE Value, Mg/g
Health
480
180
15
0.5
93
0.1
0.5
1.5
10
3.0
0.5
0.5
150
0.45
5.0
50
0.10
92
160
15
Ecology
32
174
0.4
0.1
50
0.002
0.5
0.5
0.1
0.5
0.1
0.2
14
0.02
0.3
0.2
0.05
—
2.0
Potential Degree
of Hazard
Health
126
8.1
21
1,064
0.9
130
282
283
11
15,738
594
102
7.4
,253
39
5.6
2,560
7.0
283
7.1
Ecology
1,897
8.4
788
5,320
1.8
6,500
282
848
1,120
94,482
2,970
255
80
5,700
650
1,410
5,120
—
22,655
Potential degree of hazard is defined as the ratio of the discharge concentration
to the MATE value.
1052
-------
TABLE 17 INORGANIC CONTENT OF SCRUBBER CAKE FROM OIL FIRING (DRY BASIS)
Element
Concentration
MATE Value, |Jg/g
Potential Degree
of Hazard
Ca
Mg
Sb
As
B
Cd
Cr
Co
Cu
Fe
Pb
Mn
Mo
Ni
V
Zn
Se
Sr
Al
Zr
Total
M8/8
200,000
3,799
3b
15b
40
lb
15
19b
16
2,164
6b
6
I4b
132
203
• 36
9b
239
1,684
37
208,450
Health
480
180
15
0.5
93
0.1
0.5
1.5
10
3.0
0.5
0.5
150
0.45
5.0
50
0.10
92
160
15
Ecology
32
174
0.4
0.1
50
0.002
0.5
0.5
0.1
0.5
0.1
0.2
14
0.02
0.3
0.2
0.05
—
2.0
--
Health
417
21
0.2
30
0.4
10
30
13
2
721
12
32
0.1
293
41
0.7
90
2.6
11
2.5
Ecology
6,250
22
7.5
150
0.8
500
30
38
160
4,328
60
80
1
6,600
677
180
180
--
842
--
Potential degree of hazard is defined as the ratio of the discharge concentration
to the MATE value.
SSMS analyses were utilized where ICPOES analysis provided upper limit data
only.
1053
-------
TABLE 18 INORGANIC CONTENT OF FLY ASH FROM COAL FIRING
Element
Concentration
MATE Value, |Jg/g
Potential Degree
of Hazard
Ca
Mg
Sb
As
B
Cd
Cr
Co
Cu
Fe
Pb
Mn
Mo
Ni
, V
Zn
Se
Sr
Al
Zr
Total
Mg/8
378
2,478
438
1,015
20
18
434
408
320
129,330
438
121
1,288
165
376
179
378
728
109,450
187
248,149
Health
480
180
15
0.5
93
0.1
0.5
1.5
10
3.0
0.5
0.5
150
0.45
5.0
5.0
0.10
92
160
15
Ecology
32
174
0.4
0.1
50
0.002
0.5
0.5
0.1
0.5
0.1
0.2
14
0.02
0.3
0.2
0.05
--
2.0
—
Health
0.8
14
29
2,030
0.2
180
868
272
32
43,110
876
242
9
367
75
36 ,
3,780
8
684
12
Ecology
12
14
1,095
10,150
0.4
9,000
868
816
3,200
258,660
4,380
605
92
8,250
1,253
895
7,560
--
54,725
--
Potential degree of hazard is defined as the ratio of the discharge concentration
to the MATE value.
1054,
-------
TABLE 19 MASS BALANCE ON TRACE ELEMENTS FROM COAL FIRING
Element
Ca
Mg
Sb
As
B
Cd
Cr
Co
Cu
Fe
Pb
Mn
Mo
Ni
V
Zn
Se
Sr
Al
Zr
Coal Feed
g/hr
2,794
1,270
308
497
8.7
12.7
174
461
26
44,455
308
44
1,063
134
171
203
265
247
50,806
980
Scrubber
Cake
g/hr
40,072
962
208
351
58
8.6
93
280
74
31,179
196
34
737
75
151
186
169
424
29,905
70
Scrubber
Effluent Gas
g/hr
1.6
0.5
1.1
9.7
1.2
0.04
5.7
0.53
0.88
110
0.92
0.68
1.2
2.8
2.5
2.1
4.3
2.5
110
0.79
Bottom and
Fly Asha
g/hr
30
198
35
81
1.6
1.4
35
33
26
10,346
35
9.7
103
13
30
14
30
58
8,756
15
Percent,
Recovery
c
91
79
89
700
79
77
68
39
94
75
100
79
\ 68
107
100
77
196
76
9
For mass balance calculations, bottom ash has been assumed to have the same
trace element concentrations as fly ash. This is an approximate assumption,
as some trace elements are enriched in the fly ash.
Percent recovery is 100 times the ratio of the sum of the emissions for a
trace element to the trace element in the coal feed.
Percent recovery is not calculated because most of the calcium in the
scrubber cake is from the lime slurry.
1055
-------
TABLE 20 MASS BALANCE OF TRACE ELEMENTS FROM OIL FIRING
Element
Ca
Mg
Sb
As
B
Cd
Cr
Co
Cu
Fe
Pb
Mn
Mo
Ni
V
Zn
Se
Sr
Al
Zr
Oil Feed
g/min
16.4
(12.2)C
( 2.4)
( 5.9)
(20.9)
(n.o)
3.6
( 3.9)
4.2
36.7
( 7.9)
( 1.2)
( 8.7)
47.7
108.8
8.9
( 2.0.)
0.7
10.4
( 0.6)
Scrubber
Cake
g/min
50,000
950
0.8d
3.9d
10.0
0.2d
3.9
4.7d
4.1
541
1.5d
4.0
3.5d
33.0
50.7
9.1
2.4d
59.8
421
9.2
Scrubber
Outlet
g/min
2.8
1.2
0.2
1.2
1.5
2.6
0.7
0.5
0.3
11.0
0.5
0.2
1.0
7.9
32.3
2.7
0.2
0.04
18.9
0.04
Q
Percent
Recovery
b
>1,000
42
68
55
25
125
133
105
>1,000
25
350
52
91
76
133
136
>1,000
>1,000
>1,000
Percent recovery is 100 times the ratio of its total emission rate (scrubber
cake plus scrubber outlet) to its feed rate.
Percent recovery is not calculated because most of the calcium in the
scrubber cake is from the lime slurry.
ICPOES data from the analysis of scrubber inlet particles were utilized
when fuel analysis provided upper limit data only.
SSMS data were utilized where ICPOES analysis provided upper limit data only.
1056
-------
TABLE 21 ANNUAL EMISSIONS
O
Cn
-vl
Pollutant
Gaseous: HO (as NO.)
so2
so3
S°4
CO
Organics (as CH.)
C Cb
Cl C6
C7 - C16
>clfi
Total Particulates
10um
kg /year
Scrubber Inlet
Coal Firing
500,810
1,127,300
6,184
67,214
16,119
5,870
<5,606
345
2,311
2,991,700
—
—
—
—
Oil Firing
182,478
1,006,891
8,054
23,216
5,546
2,524
<4,627
172
2,646
59,813
—
—
—
—
Coal/Oil
2.75
1.12
0.77
2.90
2.91
2.32
—
2.00
0.87
50.0
—
—
—
—
Coal Firing
442,520
36,800
4,157
8,110
14,497
6,377
<5,606
274
335
18,856
11,691
5,657
1,320
188
Scrubber Outlet
Oil Firing
174,878
27,170
5,759
9,226
5,383
2,778
<4,627
20
436
15,207
12,621
1,824
704
Oc
Coal/Oil
2.53
1.36
0.72
0.88
2.69
2.30
—
13.7
0.77
1.24
0.93
3.11
1.74
—
Solid: Bottom Ash
Fly Ash
Scrubber Cake
~ 778,600
~1, 800, 000
0
~ 8,444
~16,667
0
~ 93
~108
—
~ 778,600
~1, 800, 000
8,054,100
~ 8,444
~16,667
3,345,556
~ 93
~108
2.40
Assuming 1002 load, 45 weeks per year (7,560 hrs/year).
Represents the detection limit of. the instrument used.
CRepresents oil firing particulate with a minimum of coal ash contamination.
-------
The scrubber cakes were also analyzed for organics but none was detected.
This is expected since concentrations of organics in the flue gas stream were
very low.
ANNUAL MULTIMEDIA EMISSIONS
Table 21 presents estimates of the annual emissions of the major
pollutants for the controlled and uncontrolled case. It was assumed that the
boiler operates at 100 percent load, 87 percent of the year (7560 hours/year).
SCRUBBER EFFICIENCY
Flue gas analyses indicate that the scrubber removes a significant
percentage of input sulfur oxides (SC- , SO. and particulate SO, ), total
particulates, and organics of the C7 class and higher. Scrubber removal
efficiency data for these flue gas components are presented in Table 22. As
discussed previously, the significance of data indicating NO and CO removal
appears questionable.. Therefore, these components are not included in this
discussion. There is no NO control equipment.
Average removal efficiencies have been discussed; however, the C? and
higher hydrocarbons are removed with 77 percent efficiency for coal firing and
85 percent efficiency for oil firing. These fractions comprise 32 to 96
percent of the total generated organics. Hence, based on the total generated
organics, a removal efficiency of 25 to 53 percent was obtained for coal
firing and 32 to 84 percent for oil firing.
TABLE 22 SCRUBBER EFFICIENCY
Coal
Oil
Total
Particles
99
75
so2
97
97
soa
32
29
so'3
>88
>60
Organics
>25
>32
Trace
Elements
(overall)
99
87
This removal rate, based on only one data point for each fuel, is
actually a net change rate. The scrubber both removes and generates
sulfates.
AIR QUALITY
Simplified air quality models, were used to estimate the relative ground
level air quality resulting from uncontrolled and controlled emissions. Worst
case and typical (not average) weather conditions were considered. The worst
1058
-------
case was assumed to be plume trapping, which was assumed to persist for as
long as 3 hours. Typical conditions can reasonably be expected to occur
almost anywhere in the country. It was further assumed that all species were
inert and that no photochemical reactions occurred. Models for particles,
SO.,, NO and CO were made. Keeping in mind the assumptions mentioned above,
several observations can be made:
o Controlled emissions of particles for all cases are less than all
particle emission standards.
o For controlled emissions of SO- during both coal and oil firing, no
standards are exceeded.
o The NO standard is exceeded under both weather conditions during coal
firing. During oil firing, the NO standard was exceeded under worst case
weather conditions but not under typical weather conditions. Since the
scrubber does not remove significant amounts of NO , there is no substantial
difference between the air quality resulting from inlet and outlet emissions.
(The boiler has no NO controls.)
A
o CO standards are not exceeded under any conditions. As with NO there
is no substantial difference between the inlet and outlet concentrations.
COMPARATIVE ASSESSMENT
Scrubbing removed 99 percent of the particles from coal firing and 75
percent of the particles from oil firing. The lower removal efficiency
obtained during oil firing is attributed to the increased fraction of
particles smaller than 3|Jm; at least 21 percent of the uncontrolled particles
from oil firing are less than 3 Mm i-n diameter while substantially less than 1
percent of uncontrolled particles from coal firing are under 3 |Jm.
There appears to be a net increase across the scrubber for particles less
than 3(Jm in diameter from coal firing. This net increase can be attributed to
the poor removal efficiency of the scrubber for fine particles^ and to the
sodium bisulfate (NaHSO,) and calcium sulfite hemihydrate (CaSO '1/2 H20)
particles generated by the scrubber. Both NaHSO^ and CaSO -1/2 HO have been
identified at the scrubber outlet but not at the inlet. Although a very
slight increase in particles from oil firing in the 1-3 |Jm range was observed,
a net decrease in particles less than 3 pm was observed during oil firing.
Based on the results of coal firing tests, it appears reasonable that scrubber
generated particles were present in the scrubber outlet stream during oil
firing but that the high fine particle loading associated with oil firing
masked detection of these particles. Particle emissions after scrubbing are
below either existing or proposed NSPS limits.
Controlled SO- emissions for coal and oil firing are lower than either
existing or proposed NSPS limitations. The overall uncontrolled sulfur
balance indicates that over 92 percent of the fuel sulfur is emitted as S02,
less than 1 percent as SO-, and approximately 3 percent as SO, from coal
burning and 1.5 percent from oil firing. The remaining input is in the bottom
1059
-------
ash or is unaccounted for. Sulfates are more efficiently removed than SO^ (60
percent removal for. oil firing and 88 percent for coal firing). This
indicates that SO ~ is probably associated with the larger particles, which
are more efficiently removed than smaller particles. The higher sulfate
removal from the coal flue gases is explained by the higher particulate
loading during coal firing.
NO emissions increased with increasing load for both coal and oil
firing,Xas expected. Available data indicate that for boiler loadings between
90 and 100 percent, NO emissions from coal firing are approximately 3
times greater than from oil firing. Observed reductions of NO emissions for
coal firing and early oil firing tests appear to be due, at least in part, to
air leakage into the scrubber outlet sampling line. Data from later oil
firing tests, not known to be subject to leakage problems, indicate that
maximum NO removal across the scrubber is on the order of 2 percent. Without
controls, £he emissions of NO from clustered coal-fired furnaces will cause
x
the applicable NAAQS to be exceeded.
Uncontrolled CO emissions from coal firing were 3 times those from oil
firing, although both emissions are very small. Apparent reductions in CO
emissions across the scrubber are not considered significant due to air
leakage in the sampling train and the low sensitivity of analysis at the
measured CO concentrations.
Organic emissions for coal and oil firing were very similar, and appear
to be primarily C.. to C, hydrocarbons and organics heavier than C ,. While
uncontrolled emission rates for both coal and oil firing are low, emissions of
these organics were further reduced by about 85 percent in the scrubber unit.
The organic compounds identified in the gas sample from both coal and oil
firing were generally not representative of combustion generated organic
materials, but were compounds associated with materials used in the sampling
equipment and in various analytical procedures. This again confirms the low
level of organic emissions. Polycyclic organic material (POM) was not found
in the scrubber inlet or outlet at detection limits of 0.3 (Jg/m f°r either
coal or oil firing. MATE values for most POM's are greater than this
detection limit. However, since the MATE values for at least two POM
compounds -- benzo(a)pyrene and dibenz(a,h)anthracene -- are less than 0.3
(Jg/m , additional GC/MS analyses at higher sensitivity would be required to
conclusively determine the presence of all POM's at MATE levels. Also a more
accurate determination of oxygen in the flue gas at the furnace outlet could
be important since POM levels decrease as excess air increases at constant
temperature.
The air concentration of trace elements from plant clusters is expected
to be approximately 4 orders of magnitude below the "allowed exposure levels"
proposed for hazardous waste management facilities. They are also below
typical urban ambient background, except for cobalt and selenium, which
approach or slightly exceed endogenous levels. The concentrations are similar
from coal and oil firing, except for cadmium, which is 40 times larger from
oil firing than from coal firing.
1060
-------
Trace element concentrations in runoff water, which arise from deposition of
emissions on soil and foliage, may be about 10 times the standards for livestock
drinking and potable water. Concentrations due to oil firing are slightly
lower than those due to coal firing; however, selenium and molybdenum
concentrations in water are predicted to exceed their background levels. Mass
closure for most trace elements from coal-firing has been found to be in the
75 to 107 percent range. Mass closure for half of the trace elements from
oil-firing is in the 50 to 136 percent range; closure for the remainder of
oil-firing trace elements is poorer due to the extremely low elemental
concentrations measured and/or contamination of the recycle scrubber solution
during coal firing tests. These good closures instill confidence in the
validity of the sampling and analysis data. Beryllium emissions after
scrubbing were less than or equal to the beryllium MATE value during coal and
oil firing. At the measured emission concentrations, the National Standard
for Hazardous Air Pollutants limitation of 10 grams beryllium per day would
only be exceeded by boilers of 50 MW capacity for coal firing and 100 MW
capacity for oil firing.
Chlorides were removed with greater than 99 percent efficiency from coal
flue gases and with about 51 percent efficiency from oil flue gases. This
difference was attributed to the higher removal efficiency for the larger coal
particles. Fluorides were removed with greater than 86 percent and about 87
percent efficiency for coal and oil firing, respectively. Nitrate emissions
were removed from coal flue gases with at least 52 percent efficiency and from
oil flue gases with 57 percent efficiency.
Scrubber cake production during coal firing was 3.3 times greater than
during oil firing. If the multiclone had not malfunctioned, this ratio would
have been reduced to 2.7, assuming 60 percent multiclone efficiency.
Available data indicate that the principal difference between scrubber cake
production rates from coal and oil firing is the particle loading and
associated unbound moisture.
The scrubber cake produced from coal firing contains a significant amount
of fly ash. Except for boron, trace element concentrations in the scrubber
cake have exceeded their MATE values. This is the result of transferring an
air pollution problem by scrubbing to an easier solid waste disposal problem.
Because the trace elements may contribute to the leachate from the disposed
scrubber cake, these solid wastes must be disposed of in specially designed
landfills. In such landfills, leachate impact on ground water is expected to
be insignificant.
CONCLUSIONS
General
Several major conclusions have evolved from the environmental analysis.
The difference in environmental insult expected to result between coal
and oil combustion emissions from a single controlled 10 MW industrial boiler
is insignificant. This is because: 1) there are only slight differences in
the emissions levels of the pollutants, 2) the absolute impact of either fuel
use is insignificant, and (3) the effectiveness of the control equipment makes
environmental impacts small.
1061
-------
The environmental impacts of emissions from a cluster of five 10 MW
industrial boilers at 200-meter intervals aligned with the prevailing wind are
potentially significant. The impacts include health effects, material damage,
and ecological effects from high ambient levels of SO-, NO , and suspended
particulate matter; health effects and ecological damage due to trace metal
accumulation in soils and plants; and degradation due to visibility reduction
and the presence of waste disposal sites.
The environmental acceptability of a cluster of controlled industrial
boiler emissions depends more on site specific factors (e.g., background
pollution levels, location and number of other sources) than on the type of
fuel used. Careful control of the site-specific factors can avert potential
environmental damage and generally compensate for any differential effects
arising between the use of coal or oil.
With the possible exception of ambient levels of NO , the risk of
violating the NAAQS due to operation of clusters of controlled industrial
boilers is essentially the same whether the fuel is coal or oil. Based on
tests of the reference 10 MW boiler (which was not controlled for NO
emissions), localized NO concentrations produced by coal firing are estimated
to be twice those resulting from oil firing, and greater than those permitted
by the NAAQS for 24 hour and 1 year averaging periods.
Coal firing appears to produce a greater enrichment of trace elements in
the flue gas desulfurization filter cake than does oil firing; however, the
scrubber cake resulting from either coal or oil firing contains sufficient
amounts of heavy metals and toxic substances to require specially designed
disposal areas.
Health Effects
Regional emission levels of suspended sulfates from controlled oil or
coal fired industrial boilers would not be expected to cause a significant
impact on regional health.
Sulfate emissions from clusters of controlled industrial boilers might be
expected to cause significant adverse health effects in a localized area near
the plant cluster. Oil firing would be expected to result in localized health
effects about one-third less severe than those resulting from coal firing
since oil firing produces only one-third the particle emission of coal firing.
The impact of solid waste generation on health is essentially the same
for controlled coal firing and oil firing, if suitable land disposal
techniques are employed to ensure sufficiently low leaching rates and
migration of trace elements to groundwater and the terrestrial environment.
The concentration of metals in runoff waters due to controlled oil firing
is predicted to be slightly less than that occurring from controlled coal
firing; in either case, hazard to human health by drinking water is remote.
1062
-------
Trace element emissions from clusters of controlled industrial boilers
may significantly increase local background levels in drinking water, plant
tissue, soil, and the atmosphere; however, the expected increases in the
levels of such elements are generally several orders of magnitude less than
allowable exposure levels. Oil firing is estimated to cause cadmium burdens
in plants approaching levels injurious to man. Because cigarettes contain
significant cadmium levels, smokers are more apt to achieve thresholds of
observable symptoms for cadmium exposure if they also consume additional
cadmium via the food chain. Coal firing may produce plant concentrations of
molybdenum which are injurious to cattle.
Ecological Effects
The potential for crop damage from either controlled coal firing or oil
firing depends greatly on ambient levels of NO , SO-, or trace element soil
concentrations. If such levels are currently nigh, localized plant damage
would be expected to occur within 1 to 2 km of a controlled boiler cluster.
Leaf destruction from SO- exposure would be expected to be slightly more
severe in the vicinity or a cluster of controlled boilers which are coal fired
as opposed to oil fired. Plant damage may possibly occur even at levels below
ambient air standards.* For uncontrolled NO emissions, plant damage would be
expected to be significantly greater in the vicinity of the coal fired
cluster, because of the higher levels of ambient NO produced. Emissions of
CO and hydrocarbons will have negligible impacts on plants. The likelihood of
damage occurring in plants due to emissions of trace elements from either
controlled oil or coal firing is remote, with the possible exception of injury
due to elevated levels of molybdenum and cadmium in plant tissue resulting
from coal firing and oil firing, respectively.
The impact of fossil fuel combustion in controlled oil or coal fired
boilers on plant damage via acid precipitation would be relatively
insignificant. The levels of suspended sulfate (the precursor of acid rain)
would be essentially the same whether the controlled boilers are coal or oil
fired.
Measurements and analyses of leaching rates at experimental solid waste
disposal sites indicate that landfills of untreated flue gas desulfurization
system scrubber cake can be constructed without significant adverse impacts.
Societal Effects
The impact of boiler emissions on corrosion in the local area near a
cluster of controlled industrial boilers will be significant . The corrosion
rate will be slightly greater when the boilers are coal fired; however, the
extent of this overall impact (oil or coal) is minor compared to that which
occurs when industrial boilers are uncontrolled.
* See pp. 5-29 to 5-32 of volume 2 of reference 3.
1063
-------
The increase in annual total suspended particulate matter and the
resulting soiling damage in the vicinity of a cluster of controlled industrial
boilers results in additional cleaning and maintenance costs about 107to 15
percent greater than that already experienced in a typical urban area . The
cleaning costs may be slightly greater when the boilers are coal fired.
Emissions of particulate matter from controlled industrial boilers will
result in visibility reduction. This form of environmental degradation will
occur in a localized area near the boiler cluster, and occurs to essentially
the same extent whether the controlled boilers are oil or coal fired.
Total land disposal requirements for scrubber waste generated by
controlled coal firing are 3 times greater than those for controlled oil
firing. Disposal of the scrubber wastes may result in significant
depreciation of property value and environmental degradation in the area of
the disposal site. These impacts would be more severe if boilers use coal
rather than oil.
Economic Effects
The direct economic impacts associated with residuals of fuel combustion
involve the costs of damages (or benefits) sustained when the residuals enter
the environment. Second order economic impacts associated with the residuals
involve the alterations that occur in employment, the tax base, energy prices,
income, and land values due to the damages (or benefits) resulting from
combustion residuals. The quantification of direct economic impacts involves
the difficult task of ascribing economic values to environmental changes.
Quantification of second order economic effects is yet more difficult because
of gaps in knowledge which make it impossible to determine the complex
relationships between cost and the numerous socioeconomic factors involved.
The scope and data of this program did not permit such quantification.
Because the significant effects of direct economic impacts occurring from
controlled oil fired and coal fired boiler emissions are limited to a
relatively small area near the source, the total costs of the incremental
environmental damages are apt to be insignificant on the regional basis.
Consequently, significant incremental second order economic impacts (such as
changes in hospital employment, alteration of tax bases, or changes in income)
would be unnoticeable between controlled oil and coal fired industrial
boilers.
Energy Effects
Obviously, our abundance of coal and the uncertainty of our oil supplies are
the driving forces for studies such as the coal versus oil comparative
assessment study. Whether coal assumes a more significant role as an energy
source by national choice or because we no longer have a choice, it is
essential to be aware of and to be prepared to deal effectively with any
environmental problems which result. Thus, the comparative impact of coal
versus oil firing is indeed complex and involves consideration of all aspects
of energy supply and use, including emissions characterization, multimedia
environmental impacts identification, comparison of projected impacts with
accepted levels of impacts, and evaluation of techniques for mitigating
1064
-------
unacceptable levels of impacts. This study, which is one of several projects
in the CCEA Program, is intended as a prototype of future projects dealing
with the impacts and control of stationary conventional combustion processes.
While this study identified some potentially significant differences between
coal and oil firing in clusters of boilers, the fuel choice of oil or coal may
be a relatively minor factor in determining the environmental acceptability of
controlled industrial boilers. Other site specific and plant design factors
may exert greater environmental effects than fuel choice. As concern for
environmental protection increases, the issue may become whether the
increasing use of fossil fuels can be continued at the present levels of
control technology without potential long term dangers. If it is found that
long term effects of pollution (e.g., trace metals accumulation, lake acidity
from acid rains) from fossil fuels combustion and other sources are
environmentally unacceptable, it is clear that energy use may be affected.
Increasing control requirements could result in energy cost increases to the
level where the combustion of fossil fuels loses its economic advantage over
other, cleaner sources of energy production.
ACKNOWLEDGEMENT
We wish to acknowledge the assistance of the Environmental
Engineering Division of TRW, Inc., Redondo Beach, California, in
preparing this paper. The data presented were obtained by TRW
during a program to conduct a comparative assessment of coal and
oil firing in an industrial boiler (EPA Contract No. 68-02-2613,
Work Assignment No. 8). The following TRW personnel were instrumental
in performing the assessment: C. Leavitt, K. Arledge, C. Shih,
R. Orsini, A. Saur, W. Hamersma, R. Maddalone, R. Beimer, G. Richard,
and M. Yamada.
REFERENCES
1. Ponder, W. H. and D. C. Kenkeremath, Conventional Combustion Environmental
Assessment Program, MS78-44 Rev. 1, Mitre Corp., McLean, VA, September 1978.
2. Cleland, J. and G. Kingsbury, Multimedia Environmental Goals for
Environmental Assessment, 2 vols., EPA-600/7-77-136 a/b (NTIS
PB276919/276920), November 1977.
3. Leavitt, C., et al., Environmental Assessment of Coal- and Oil-Firing
In A Controlled Industrial Boiler, 3 vols., EPA-600/7-78-164 a/b/c
(NTIS PB289942/289941/291236), August 1978.
4. Capabilities Statement, Sulfur Dioxide Control Systems, Tech. Report 100,
FMC Corp., Environmental Equipment Div., Itasca, Illinois, November 1977.
1065
-------
5. "Standards Applicable to Owners and Operators of Hazardous Waste Treatment,
Storage and Disposal Facilities", draft of proposed rules obtained from
EPA, Office of Solid Waste, March 1978.
6. Upham, J., "Atmospheric Corrosion Studies in Two Metropolitan Areas,"
J. Air Poll. Control Ass., June 1967.
7- Michelson, I., "The Household Cost of Living in Polluted Air in the
Washington B.C. Metropolitan Area," a report to the U.S. Public Health
Service.
1066
-------
OPERATING HISTORY AND PRESENT STATUS
of the
GENERAL MOTORS DOUBLE ALKALI S02 CONTROL SYSTEM
Thomas O. Mason, Chevrolet Div., GMC, Parma, Ohio
(Speaker)
Edward R. BangeI, Chevrolet Div., C.O., GMC, Warren, Michigan
Edmund J. Piasecki, EAS, GMC, Warren, Michigan
Robert J. Phillips, EAS. GMC, Warren, Michigan
1067
-------
I. INTRODUCTION
The General Motors version of the double-alkali wet scrubbing
process went on line in March of 197*+ at the Chevrolet Motor
Division facility in Parma, Ohio. This installation, was at
that time, the latest development in a program started by
General Motors engineers in 1968 to find an environmentally
acceptable means of burning coal in an industriaI-size power
plant. Most of the boilers in the Corporation range from
steaming capacities of 50,000 to 150,000 pounds per hour,
with an average plant steaming capacity of 250,000 pounds per
hour. Generally, the pressure ranges below 200 psi, and the
steam is used in process operations and building heat.
There are a number of operating characteristics associated
with industrial boilers which have to be considered in the
scrubber system design. These include boiler type, load
fluctuation, high excess air, and dust load. In addition to
these basic design characteristics, the system had to be
simple to operate, reliable, and economically competitive
with other control alternatives.
With this in mind, the Chevrolet Parma, Ohio, plant was
selected in 1972 for a prototype wet scrubbing system. The
history and status of the installation follow.
1068
-------
II. DESCRIPTIVE LAYOUT OF THE FACILITY AND EQUIPMENT
The Chevrolet Parma facility includes a sheet metal stamping
plant, an automatic transmission manufacturing plant, and a
propeller drive shaft manufacturing plant, steam is supplied
for process parts washers and building heat from four
boilers, centrally located in a separated facility. Two,
60,000 pound-per-hour boi.lers were installed in 19^8 and two,
100,000 pound-per-hour boilers were installed in 1966 and
1967. The boilers are all travel-ing grate, continuous-ash-
dump type with stoker feeders. They are also equipped with
mechanical dust collectors. The larger boilers have
economizers, and stack discharge temperatures are
approximately 350 degrees Fahrenheit. The stack discharge
temperature of the smaller boilers is approximately 600
degrees Fahrenheit.
The double-alkali type system was chosen because of its
potential reliability over Iime/Iimestone systems. The
system is closed-loop with scrubbing being accomplished by
use of a dilute caustic mixture and regeneration with lime.
Each boiler has its own scrubber and caustic control. The.
lime regeneration system is common to all units.
Additional draft fans were necessary to overcome the seven to
eight inch pressure drop of the scrubbers. These fans are
1069
-------
installed upstream of the scrubbers to avoid corrosion
problems.
The scrubbers have three bubble cap absorption trays and a
mesh-type mist eliminator. This type of scrubber was chosen
to obtain the highest sulfur dioxide removal along with
maximum turndown potential. All wetted surfaces are
constructed of 316L stainless steel.
Caustic make-up to the scrubbers is determined by the pH of
the liquor as it is recircuIated. Liquid bIowdown is
diverted to two in-line reactor tanks where slaked lime
slurry is added. These tanks overflow to a reactor clarifier
where solids settling occurs and the lime reaction continues.
The overflow from this clarifier goes to a second clarifier,
where soda ash is added for softening and sodium make-up, and
additional solids settling occurs. Overflow from the second
clarifier is the make-up liquor used to replace blowdown at
the scrubber.
1070
-------
III. ASSIGNMENT OF MANPOWER
It was determined that scrubber operation was to be placed
under the supervision of the powerhouse personnel. This was
done in order to coordinate the scrubber operations with
steam generation.
A rotating continuous operation was established with boiler
operators operating the scrubber. Union agreements limit us
to the use of powerhouse personnel to run the scrubber
facility. Although these people normally run water tests,
set valves, monitor controls, and adjust set points in their
boiler duties, they have a limited chemical processing
background. This has created some operational difficulties.
There are a total of eight hourly people permanently assigned
to the operation. This is compared to sixteen hourly people
already employed for powerhouse operations. Of these eight,
four are assigned as shift operators, two to vacuum
filtration and sludge handling, and two to maintenance
activities. No additional supervision is assigned at the
scrubber faciIity at this time.
A set of operational guidelines and instruction books were
published. Personnel were indoctrinated in classes and
1071
-------
assigned duties and shifts. With no other system to study,
we decided to go into operation without further delay and
work through our problems.
1072
-------
IV. EARLY DEVELOPMENTS
The scrubbers were started up for the first time on February
28, 1974. A number of initial problems were encountered, many
of these were due to lack of experience with the system. For
example, just prior to an open house, we experienced an upset
when a maintenance employe started cleaning the building
walls with a soap solution. Since all spills or leaks drain
back into the system via sump pumps, the soap entered-the
process and caused all the solids in the clarifiers to go
into suspension. During our open house, we were scrubbing
with slurry. This eventually shut down the system.
We also encountered a number of functional problems with the
system. First, we had a problem with reliable pH control.
This was due to initial location of the pH sensor. Since the
probe was in a pressure line, it would frequently break. In
addition, because the response time was so short, pH readings
were unreliable. We tried to solve the problem by placing
the probe in a gravity sample line off the top tray. This
stopped the breakage, but we soon coated the entire probe
with sol ids.
These experiences led to the present arrangement of sampling
pH in a gravity sample line off the bottom tray. This lower
pH location has eliminated the scaling problem and allowed us
1073
-------
to maintain reliable pH control, which is maintained at a
I eve I of pH six.
Eventually we experienced a problem due to the lack of
intermixing of recycle liquor and scrubber feed which caused
the top tray to plug. This was the result of a high pH
condition on the top tray, with calcium carbonate scaling
shutting down operations within eight hours. An early fix
was to have both lines feed into a mixing box above the top
tray of the scrubber. This proved ineffective because
sufficient time was not allowed for mixing. The next step
was to put caustic feed into the recycle tank. This
effectively reduced top tray pH and eliminated the carbonate
scaling problem. Unfortunate 1y, solids build-up of calcium
sulfite occurred at a fantastic rate in the recycle tanks,
causing them to fill with solids. This was primarily due to
the long hold-up time in the tank at a lower pH. Final I y, we
determined that in order to minimize scaling, the proper
place to introduce caustic make-up was in the recycle line
just prior to entering into the top tray of the scrubber.
This has greatly reduced the scrubber scaling problem.
An additional problem was encountered with settling calcium
salts in the first clarifier. The problem was caused by
hindered settling due to excess sludge recycling and trouble
maintaining a constant sludge blanket in the bottom of the
1074
-------
clarifier. This was resolved by eliminating sludge recycle
and by not operating with a sludge blanket.
We have had a very long learning curve on the operation of
vacuum fiIters. The first cloth was a polypropylene material
that had good wear characteristics but poor cake release.
This was replaced with nylon and improvements have resulted.
Also contributing to the problem of cake release was the high
percentage of calcium sulfite in the cake. Cake
characteristics were greatly improved by adding an oxidation
step prior to lime addition. This produces a higher
percentage of calcium sulfate. We are now obtaining approx-
imately 60 percent solids in the cake.
Overdesign of the system has caused problems with control.
The system was originally designed to handle ^00,000 Ibs./hr.
of steaming capacity burning three percent sulfur coal.
However, the maximum actual operations have been only fifty
percent of design capacity. Since the controls were designed
for maximum levels, there have been problems measuring flows
and controlling the operation at these reduced conditions.
This made it necessary to reduce line sizes in specific areas
of the process in order to more accurately measure and
control liquid flows.
1075
-------
We also found a serious problem with closed gravity lines
carrying solids to the clarifiers. This caused numerous
outages of the system due to plugging of these lines. We
finally resolved the problem by installing removable covered
troughs of a rectangular cross section which increased the
flow rate in these lines and made them accessible for
mai ntenance.
1076
-------
V. CURRENT OPERATING PROBLEMS
We are currently in the process of modifying our lime feed
systems. The original system has never been able to
accurately control the amount of lime added to the system.
This was due to the number of variables in the control loop
which made this lime addition so complex it could not be
maintained. We are simplifying the system by installing
linear feed pumps and proportioning that feed to caustic flow
to the scrubbers.
Mist carryover continues to be one of our most serious
problems. Initial scrubber design did not allow for
utilization of the entire mesh pad. This was corrected by
modifying the transition from the scrubber to the exhaust
stack and by relocating the mesh mi st 'e I iminator. However,
this has not totally resolved the problem. We are continuing
to investigate potential solutions.
We are also currently modifying the vacuum filter agitators,
slurry inlet, and cake support roll. We hope to have better
solids distribution over cloth and improve cake filtering
character!sties.
-------
VI. SYSTEM PERFORMANCE VERSUS TIME AND OPERAS ILITY
The scrubbers have held up very well with little or no
corrosion discernible. The rubber-lined recycle pumps have
proved to be very maintenance free. The recycle tanks are
showing corrosion and will require relining. They are mild
steel painted with a bitumastic coating. We are presently
patching them and relining with plastic. Corrosion of the
(
stacks occurred almost immediately after startup. A
stainless liner was added within the first year, and that
has also failed. We have since replaced them with FRP
stacks.
The sludge handling operation receives the most abuse and,
therefore, requires the most service. The present sludge
pumps, however, have provided good service. The original
sump pumps failed early and were replaced with air operated
diaphragm pumps.
Piping in the system has been mostly trouble free. All
piping in areas of the system where corrosion could occur is
hand-laid, fiber-cast plastic.
The ratio of scrubber hours versus boiler hours has been
increasing to a level now near 70 percent. One of the factors
that must not be overlooked in this evaluation is that these
1078
-------
hours include many scheduled shut-down periods for planned
maintenance, evaluation programs, or system modifications.
Typically, the system operates for extended periods of time
without interruption.
1079
-------
VII. RELATIVE COST
The latest costs of constructing and operating our system are
as follows:
1978 $/ton of coal
Capital Charge 6.85
Chemicals 1.26
Utilities 0.93
Sol id Waste 0.70
Maintenance 6.21
Labor k.2k
Total $20.19
The construction cost to General Motors was 3.2 million
dollars, based upon 1973 dollars. This figure includes all
design and construction-related costs performed by other
corporate divisions.
1080
-------
VIM. FUTURE OUTLOOK
The system is running now. It may require more modifications
to reach a reasonable reliability level. We are experiencing
carryover problems, and the mist eIiminator wi I I have to be
modified. This is a major concern, especially in the
appearance of the surrounding area.
General Motors continues to believe that until the system's
reliability has been improved to 90 percent, this double-
alkali process should be considered developmental. With the
implementation of the programs outlined, we hope to obtain
this goal.
Ladies and gentlemen, it's very difficult to squeeze five
years into a short talk, but I thank you for the opportunity
to try.
1081
-------
R-C/BAHCO FOR COMBINED SO2 AND PARTICULATE CONTROL
NICHOLAS J. STEVENS
RESEARCH-COTTRELL, INC., SOMERVILLE, NEW JERSEY
ABSTRACT
The R-C/Bahco system controls 803 and particulate
emissions in industrial applications where high performance
is required. Research-Cottrell started up the system suc-
cessfully on its first commercial application at Rickenbacker
Air Force Base, Columbus, Ohio in March, 1976. The scrubbing
system has been operating at high availability (>90%) since
the Air Force took beneficial occupancy in September, 1976.
An EPA-sponsored test program was conducted to evaluate the
viability of the R-C/Bahco technology for air pollution
control. Effective S02 removal is accomplished using
either a lime or limestone reagent. The Air Force switched
from lime to limestone in May, 1977 because of the substantial
reagent cost savings. High particulate collection efficiencies
(98+%) are obtained for fly ash particle sizes above 1 to 2
microns. This paper describes the R-C/Bahco system installed
a.t RAFB, evaluates its performance and operation and indicates
costs and range of R-G/Bahco application for coal-fired
industrial boilers.
1082
-------
R-C/BAHCO FOR COMBINED S02 AND PARTICULATE CONTROL
INTRODUCTION
In September 1974, the United States Air Force took a
significant step 'to demonstrate the applicability of flue
gas scrubbing technology to industrial coal-fired plants.
Research-Cottrell was awarded a contract for an S02 and
particulate emissions control system for the Central Heat
Plant at Rickenbacker Air Force Base (RAFB) near Columbus,
Ohio. An R-C/Bahco scrubber system was selected for this
project. The scrubber, which simultaneously accomplishes
SO2 and particulate removal, is based on technology devel-
oped by A.B. Bahco in Sweden and tested worldwide on oil
fired boilers, incinerators and other applications. The
system installed at RAFB, represents the first application
of the Bahco system in the U.S., and the first application
worldwide on a coal-fired industrial boiler.
The U.S. Air Force awarded a contract to Research-
Cottrell in April, 1975, by means of an interagency grant,
to evaluate the viability of the R-C/Bahco air pollution
control technology. The objective of the EPA sponsored
program was to characterize the scrubbing system at RAFB
in terms of .its performance, reliability and economics for
SO2 and particulate control on industrial-scale, coal-fired
boilers.
The R-C/Bahco system is especially effective in those
industrial applications where high performance for S02 and
particulate removal is required. However, the process also
has been applied to a wide range of industrial pollution
control problems at 18 installations in Japan and Sweden
(1). This paper describes the R-C/Bahco system installed
at RAFB, evaluates its performance and operation and
indicates the costs and range of R-C/Bahco application for
coal-fired industrial boilers.
RICKENBACKER AFB HEAT PLANT
The Central Heat Plant at Rickenbacker Air Force Base
consists of a bank of six coal-fired hot water generators.
The total installed capacity of the stoker-fired boilers is
about 330 million BTU/hr. firing rate and they typically
burn 11,300 BTU/lb. coal containing 2,5-3.5% sulfur.
1083
-------
The generators emit approximately 6.2 Ibs. of SO2 and
5.6 Ibs. of particulate per million BTU of coal burned. The
Ohio Air Pollution Control Regulations in force limited emis-
sions to 2.2 Ibs. sulfur dioxide and 0.16 Ibs. particulate
per million BTU of coal burned. Thus, in order to comply
with the Ohio Regulations the S02 and particulate emissions
had to be reduced by 65% and 97.5%, respectively.
The R-C/Bahco system (Figure 1) was designed to treat
up to 108,000 ACFM of flue gas generated at the peak winter
firing rate of approximately 200 million BTU/hr. The system,
which has essentially unlimited turndown capabilities, handles
seasonal load variations from 20 to 200 million BTU/hr., S02
concentrations up to 2000 ppm and particulate loadings of
up to 2 gr/SCFD. In addition, the scrubbing system copes
with 100% load increases occurring in as short a time as one
hour.
PROCESS DESCRIPTION (Figure 2)
Hot flue gas from each of the Heat Plant Generators is
passed into a common flue which contains a by-pass stack.
The by-pass stack allows makeup air to be drawn into the
system at low load to maintain efficient operation of the
mechanical collector and scrubber.
The booster fan forces the flue gas into the first stage
of the scrubber where it is vigorously mixed with the scrub-
bing slurry in an inverted venturi. In this stage, the flue
gas is cooled to its adiabatic saturation temperature and S02
and particulate are scrubbed from the gas. The partially
scrubbed gas rises to the second stage and is contacted with
slurry containing fresh reagent to complete the required SO2
and particulate removal. Gas from the second scrubber stage
slurry is separated by centrifugal force to produce an es-
sentially droplet-free effluent.
At Rickenbacker pebble lime is slaked and added directly
to the spent slurry recycled to the dissolver tank. The re-
sulting fresh lime mixture is pumped to the second stage (top)
venturi to treat the flue gas stream. The slurry flows by
gravity from the top stage to the lower stage where it contacts
the entering hot flue gas. This countercurrent flow arrange-
ment in the scrubber produces high S02 removal combined with
efficient reagent usage.
Spent slurry flows by gravity from the lower stage of
the scrubber to the dissolver tank. Part of the spent strea,m
leaving the lower stage is diverted to the thickener where
the slurry is concentrated to approximately 40% solids. The
1084
-------
o
00
FIGURE 1: The R-C/Bahco Scrubbing System at RAFB
-------
o
CO
REAGENT SYSTEM
MODULE
UNLOADING'
STATION
LIME
STORAGE
THICKENER
OVERFLOW
LIME
FEEDER
& SLAKER
STACK
MAKEUP
WATER
t
t
THICKENER
OVERFLOW
TO
LIME
DISSOLVING
TANK
-R-C/BAHCO
SCRUBBER
SLUDGE
TO POND
BOOSTER
FAN
MECHANICAL
COLLECTOR
BY-PASS
MAKE UP STACK
FLUE GAS
FROM HEAT
^—PLANT
-LIME
DISSOLVING TANK
LEVEL
TANK
2nd STAGE PUMP
TO
FLY ASH
DISPOSAL
MILL PUMP
Figure 2 R-C/Bahco Scrubber System Flow Diagram
-------
thickener overflow is recycled to the dissolver tank and the
thickened sludge is pumped to a Hypalon-lined pond adjacent
to the Heat Plant.
MAJOR EQUIPMENT
The major equipment installed in the R-C/Bahco system at
RAFB, shown in Figure 2, is described below.
Flue System
The flue system includes individual tie-ins to each of
eight boilers. Manual diversion dampers allow gas flow into
the flue system or bypassing through individual stacks. In
addition, a stack in the main flue upstream from the mechan-
ical collector permits the addition of makeup air to the gas
or by-passing of the scrubber.
Mechanical Collector
A Flex-Kleen mechanical collector handles 108,000 ACFM
of flue gas at 475°F at dust loads up to 2 gr. per SCF. This
collector is located in the main flue upstream from the booster
fan. The collector operates at approximately 5 in. W.C. pres-
sure 'drop at full load and 1.5 in. W.C. at minimum load.
Booster Fan
The booster fan draws gas and air mixtures through the
mechanical collector and forces them into the R-C/Bahco
scrubber. The fan was oversized by 200 H.P., for a total of
700 H.P., to allow for high gas flow rates at pressures up
to 30 in. W.C. for the EPA test program (2). The scrubber
normally operates at 15 to 18 in. W.C.
Reagent System
The reagent system consists of a reagent unloading system,
storage bin, feeder, and lime slaker. The equipment can
handle pebble lime as well as limestone as reagent for SO2
removal.
1087
-------
The pneumatic unloading system handles 3/4" pebble lime
at a maximum rate of 1 ton per hour. The 120-ton capacity
storage bin is equipped with a motor driven live bottom which
is activated by the lime slaker.
The weigh belt feeder is fitted with manual and automatic
controls and a totalizer. Either the SC>2 mass flow rate or
dissolving tank pH can be used to control the reagent feed
rate. The lime slaker includes a water totalizer, grit re-
moval circuit, high temperature alarm, and dust and vapor
venting system.
Lime Dissolving Tank
The lime dissolving tank, which serves as the surge tank
for the entire system, is made of 316L SS. In this tank,
slaked lime or limestone is blended with spent slurry for
recirculation to the scrubber.
R-C/Bahco Scrubber
The scrubber is an R-C/Bahco size 50 module fabricated
from 316L SS. This module is approximately twelve feet in
diameter by sixty feet high and has a.nominal gas handling
capacity of 50,000 SCFM. A fiberglass reinforced polyester
(FRP) stack, 5.5 feet in diameter and 20 feet high, is mounted
on top of the scrubber.
The scrubber module incorporates two inverted fixed-
diameter Venturis. Each has a level tank located outside
the shell of the scrubber with a manually adjustable weir.
A fluid mill, which grinds coarse reagent particles, is
located in the bottom of the scrubber module.
Second Stage Slurry Recycle Pump
The second stage recycle pump circulates slurry through
the entire scrubbing system. The pump is rubber-lined and is
rated at 2600 gpm at 20 psig. A 316L SS shaft sleeve and a
water purge in the stuffing box are used to minimize wear and
corrosion.
Mill Pump
The mill pump is identical to the second stage slurry
pump but it operates at 2000 gpm at 25 psig. This pump is the
prime mover for the fluid mill at the base of the scrubber module,
1088
-------
Thickener
The thickener, which is 25 feet in diameter and 8 feet
in height, is used for solids surge capacity, slurry density
control, and thickening sludge for disposal. The tank is
Douglas Fir and the rake mechanism is rubber-covered carbon
steel. The rake mechanism has a lifting device and torque
sensor with a high torque alarm and cutoff for its protection.
Sludge Transfer System
The sludge transfer system includes two thickener under-
flow pumps and two transfer lines to provide 100% standby
capacity. The two underflow pumps are air operated and fitted
with replaceable neoprene diaphragms and 316 SS wetted parts.
The pumps are capable of pumping up to 40 gpm of sludge at 75
psig. The lines to the sludge pond run underground inside a
sleeve to permit easy removal for cleaning or replacement if
it is necessary.
Sludge Pond
The sludge disposal pond, shown in Figure 3, is located
approximately 700 feet from the scrubbing system. The pond
is lined with Hypalon (chloro-sulfonated polyethylene), and
is approximately 450 feet long, 250 feet wide and 12 feet
deep. An underdrain system allows ground water to be removed
from beneath the liner and also serves as a means of detecting
any leaks which may occur. The pond was designed to hold
sludge produced by scrubbing flue gas from the combustion of
200,000 tons of 5% sulfur coal.
PROCESS CONTROL
A combination of manual and automatic controls are used
to adjust and regulate the scrubber system variables.
Manual Controls
The manual mode is used to control variables which do not
have to be adjusted frequently or, once set, are essentially
constant over the entire operating range of the system. Gas
flow rate, slurry circulation rate and first and second stage
pressure drops are controlled manually.
1089
-------
o
FIG U R E 3: Lined Storage Pond at RAFB
-------
Gas Flow Rate At RAFB the gas rate to the scrubber is
set manually by adjusting the booster fan inlet dampers.
Merely positioning the damper to obtain the desired total gas
flow rate is sufficient for control purposes. Variations in
flue gas rates are accommodated by mixing makeup air with the
flue gas to maintain the desired total gas flow rate. Control
is maintained as long as the flue gas volume is less than the
total flow that the booster fan is able to accommodate for the
damper position selected. If the flue gas flow exceeds this
total rate, gas bypasses through the makeup air stack and ac-
tivates a temperature alarm to alert the Heat Plant operators.
Slurry Circulation Rates All slurry circulation rates
are manually adjusted and are set to maintain line velocities
between 4 and 8 ft/sec. The system is designed to accommodate
all loads and load changes without adjusting these slurry
circulation rates:
The following loops use this constant flow principle:
• Scrubber mill or first stage slurry recycle loop
• Second stage slurry feed
• Thickener feed
• Thickener underflow (sludge disposal)
Pressure Drop Pressure drop in either venturi stage is
manually adjusted by raising or lowering the weir in a level
tank outside the scrubber. Each stage can be adjusted in-
dependently to produce a pressure drop from 5 to 15 in. W.C.
Raising or lowering the weir causes the slurry level in the
scrubber near the lower edge of the venturi to rise or fall.
The venturi pressure drop is linearly related to the slurry
level in the vicinity of the venturi so that weir adjustments
produce proportional changes in pressure drop.
The pressure drop, once set, is virtually unaffected by
changes in gas flow rate. This insensitivity of pressure
drop to gas flow results from the self-adjusting action of the
slurry level in each venturi. When an increase in gas flow
tends to increase the pressure drop, the slurry level tends
to drop because of increased pickup of slurry by the gas
stream. The drop in slurry level causes a decrease in pres-
sure drop. When gas flow rate decreases, the slurry level
rises increasing the pressure drop. Thus, the pressure drop
is self-compensating as the gas flow varies and tends to
stabilize at a value near the initial setting.
Automatic Controls
There are three essential automatic controls in the R-C/
Bahco scrubber: reagent feed, slurry density, and makeup
water or system level.
1091
-------
Reagent Feed The reagent feed system at RAFB is designed
to maintain a preselected reagent-SC>2 stoichjLpmetry for any-
load condition and any coal sulfur content within prescribed
limits. Both the gas rate and the S02 concentration to the
scrubber are measured continuously. These measurements are
combined in a ratio controller which can regulate the reagent
feed rate over a range of 20 to 1 to maintain the desired
reagent-S02 stoichiometry.
Slurry Density Controlling the variations in slurry
density is important for proper operation of the sludge de-
watering system and to minimize scale formation. The slurry
density control system at RAFB operates between set points of
10 and 12% solids. A sensor monitors slurry density and a
controller is activated to allow thickened sludge at 40%
solids to flow to the pond when the density reaches approx-
imately 12% solids. Sludge flows continuously to the pond
until the density in the system drops to 10% solids. When
this point is reached, sludge is recycled to the scrubber
and the line to the pond is flushed with water. This switch-
ing process is repeated as necessary to maintain slurry
density in the desired range.
Makeup Water The total water requirement for the system
varies almost directly with load since evaporative cooling of
the flue gas consumes the bulk of the water used. Water is
added to the system at several locations including the lime
slaker and slurry pump seals. The balance of the makeup water
is added through six spray manifolds located inside the scrub-
ber module. The amount of water added through these sprays
is regulated by a level sensor located in the lime dissolving
tank. This sensor activates a programmed controller which
adds water in a preselected sequence. Water losses from
evaporation and sludge removal cause the level in the dis-
solver to drop initiating the water addition cycle to main-
tain the desired system operating level.
SYSTEM OPERATION
Start-up
The March, 1976 start-up of the the R-C/Bahco scrubber
system proved to be a straightforward, one-day activity as a
result of a careful, detailed check-out phase. The major
equipment was checked out and operated with air and water for
a few days prior to stdrt-up. Operation on manual control
for a few weeks after start-up proved simple and required
very little attention. Switching the three basic control
loops (lime feed, slurry density and tank level) over to auto-
matic was accomplished smoothly and with a minimum of effort.
1092
-------
Research-Cottrell maintained around-the-clock shift
coverage for one week after start-up. After the first week
of operation, RAFB Heating Plant personnel took over oper-
ation of the scrubber. R-C continued to provide one engineer
on stand-by for an additional month.
Problems Encountered
The problems encountered during the early operation of
the R-C/Bahco system were mechanical and operational rather
than process related. The scrubber operation itself has been
excellent since start-up. No outages have occurred as a
result of scrubber problems such as plugging or scaling.
High lime reagent utilization and continuous slurry particle
size reduction in the scrubber mill section appear to have
contributed significantly to trouble-free scrubber operation.
System downtime has been related almost completely to
auxiliary equipment failures and has occurred in two major
areas. First, problems centered on the booster fan resulted
in several periods of downtime for diagnosis and repair,
expecially during the early months of operation. Fan-related
problems, particularly excessive vibration caused by bearing
and support difficulties, were by far the major contributors
to downtime.
The second category consisted of problems in a number of
different areas. A thickener rake fabrication error discovered
when the assembly was being installed contributed to downtime.
Minor difficulties caused by pump, lime slaker and instru-
mentation installation errors or manufacturing defects also
were encountered. Each of the problems has been solved and
the system continues to operate very well.
The problems contributing to downtime occurred, for the
most part, during the first months after start-up when the
last few equipment and control items were installed and shakedown
completed. Except for a several week shutdown caused by a
fan malfunction, the R-C/Bahco scrubber availability has
been considerably greater than 90% since the Air Force took
beneficial occupancy in September, 1976, Availability has
been greater than 98% since April, 1978.
Scrubber Inspections
During the first fifteen months of operation, a number
of inspections were made of the scrubber internals when down-
time permitted. The inspections were used to monitor the
effectiveness of the water makeup system to keep areas clean
that were potential sources of solids accumulation. The
1093
-------
initial inspection, a month after start-up, revealed an ac-
cumulation of solids in the second stage venturi slurry pan.
It posed the only problem to smooth scrubber operation. The
pan was emptied, solids were dislodged from the wet/dry zone
and removed from the system. Subsequent investigation in-
dicated that the lime slaker grit removal circuit in the un-
slaked lime was entering the scrubber and accumulating in
the second stage slurry pan. The grit removal circuit was
adjusted to remove this material and eliminate the problem.
Two additional precautions were found, as a result of
the inspections, to be effective in improving scrubber oper-
ation. Increasing the blow-down frequency prevented accum-
ulation of solids in the slurry outlets. Also, the slow
accumulation of solids in the slurry outlets. Also, the
slow accumulation of solids in the stack and gas straighten-
ing vanes at the top of the scrubber was further reduced by
operating the second stage slurry at a lower pressure drop.
In this way, the possibility of slurry droplet carryover was
minimized.
No significant problems related to the accumulation of
solids were found in the R-C/Bahco scrubber. The scrubber
has the ability to tolerate substantial solids accumulation
before scrubbing performance is adversely affected. Deter-
ioration in performance, if it occurs, is gradual and usually
can be rectified conveniently during a scheduled shutdown.
S02 REMOVAL PERFORMANCE
The R-C/Bahco scrubber is a high performance S02 removal
device. The two-stage venturi unit at Rickenbacker very ef-
fectively scrubs S02 to low levels using a lime or limestone
reagent.
Lime Reagent
Figure 4 shows that the S02 emission rate from the scrub-
ber using lime is much lower than either the R-C/Bahco contract
guarantee emissions rate of 1.0 Ibs. S02/MM BTU (at 1.4 stoich-
iometry) or the state of Ohio limit for Rickenbacker Air Force
Base of 2.2 Ibs. S02/MM BTU. A stoichiometry of only 1.1
moles lime/moles S02 was required at Rickenbacker to almost
completely eliminate SO2 from the scrubber stack gas.
Of course, complete S02 removal is not required by the
state of Ohio but stringent S02 emissions levels have been
set for certain other sites. For example, emission rates as
low as 0.3-0.4 Ibs. S02/MM BTU have been promulgated for Wright
Patterson AFB boilers at Dayton, Ohio. As Figure 4 shows,
1094
-------
o
IO
in
2.5
CD
CO
CO
m
cc
1.5
O
g 1.0
2
UJ
CM
O
CO
0.5
STATE OF OHIO RAFB LIMIT
h- — — — GUARANTEE EMISSION RATE
I
I
90% LIME USAGE
I I
0.7 0.8 0.9 1.0 1.1 1.2 1.3
LIME STOICHIOMETRY, MOLES LIME/MOLE SO2
FIGURE 4: The relationship between
emission rates and lime
stoichiometry.
1.4 1.5
-------
these low SO^ emission rates were attained at Rickenbacker
using a lime/SC>2 stoichiometry of 1.0 or less when the boiler
was burning 2.5-3.5% sulfur coal. The R-C/Bahco scrubber exper-
ience at RAFB indicates that, if necessary, the unit can be
designed to remove SC>2 to easily comply with even the most
stringent emissions requirements.
Over the range of conditions studied, the R-C/Bahco
scrubber S02 removal efficiency was found to be a function
only of the lime/SC>2 stoichiometric ratio. As Figure 5
shows 862 removal efficiency approaches 100% as the lime/
S02 stoichiometry exceeds 1.0. These results were obtained
with inlet S02 concentrations in the range of 500-2000 ppm
and a total system pressure drop of 15-20 inches W.C. In
this pressure drop range, the degree of mixing is sufficiently
great to permit rapid reaction between the SC>2 and lime. Lime
utilizations in the range of 90-100% were achieved in all tests
performed at Rickenbacker even those at very high SO2 removal
efficiencies.
Limestone Reagent
Limestone reagent can also be used to meet the require-
ments for SC>2 removal at industrial coal-fired installations.
Tests at RAFB indicated that SC>2 removal of 80-90% was achieved
at limestone stoichiometries of 1.0 or greater. Limestone
stoichiometry and slurry feed rate to the scrubber signif-
icantly affect SC>2 removal, as Figure 6 shows. Increased
stoichiometry and slurry feed rate result in higher S02 re-
moval. Limestone utilizations range from about 60% at con-
ditions where higher SC>2 removal percentages are achieved to
90% at lower SO- removal values.
The effect of. stoichiometry and slurry feed rate on S02
removal in the R-C/Bahco scrubber using a limestone reagent
were correlated by the following empirical model:
% S02 Removal = (St)°*52 (L)0'55
where:
St = Stoichiometry, moles CaC03/mole S02 in the inlet
gas and,
L = Slurry feed rate to the second stage, gpm.
Figure 7 shows a comparison between the S02 removal pre-
dicted by Equation (1) and observed S02 removal. Actual S02
removal performance and the model prediction fall within +
15% of each other indicating that a satisfactory relationship
was obtained from the field tests.
1096
-------
o
VO
100
o
s
HI
cc
CVJ
O
80
RAFB CODE REQUIREMENT
60
40
20
LIME
UTILIZATION '/
10096/ /
^90%
//
I
I
O SCREENING TESTS
A VERIFICATION TESTS
I
0.2 0.4 0.6 0.8 1.0
LIMESTOICHIOMETRY, MOLES, LIME/SO2
1.2
FIGURE 5 SO2 removal efficiency as a function of lime
stoichiometry.
-------
100
90
80
170
UJ
C 60
LL
UJ
50
40
30
20
10
7
// '/, v
I I
.2
A .6
.8
1.0 1.2 1.4 1.6
LIMESTONE/SO2 STOICHIOMETRY
(LB-MOLE CaCOa/LB MdLE SO2)
FIGURE 6: 862 removal efficiency as a function
of limestone/SOa stoichiometry and
slurry pumping rate.
1098
-------
o
VO
VO
o
2
UJ
DC
CM
O
03
Q
UJ
O
Q
UJ
DC
Q.
100
LIMESTONE MODEL
% S02 Removal = Sta52L°'55
where
St = moles limestone/mole SO2
L-GPM slurry feed
I
I
30
40
50 60 70 80
OBSERVED SO2 REMOVAL %
90
100
FIGURE 7: A comparison of predicted and observed SC>2
removal efficiencies for limestone.
-------
Scrubber Slurry Product
The slurry effluent from the scrubber contained a much
greater amount of oxidized sulfur product with a limestone
reagent than with lime (3). Table 1 shows the highly oxidized
slurry product when using limestone with 62.3 wt. % calcium
sulfate (CaSO4«2H20) and 16.1 wt. % calcium sulfite (CaSOs
•*SH20) in the solids. In contrast, the lime generated slurry
product contained appreciable percentages of both calcium sulfate
(33.4 wt. %) and calcium sulfite (54.5 wt. %). As a result of
the higher calcium sulfate content, limestone slurries tended
to dewater at a faster rate than lime slurries and produced
higher solids concentrations in the dewatered product.
A comparison of average lime and limestone slurry anal-
yses made during similar boiler load periods listed in Table 1
indicates the oxidation trend was directly attributable to
the reagent type since all other operating conditions were
essentially the same. The much greater extent of oxidation
using a limestone reagent may likely be a result of the lower
scrubber slurry pH coupled with high flue gas 02 content (_> 14%) .
Figure 8 shows slurry pH measurements taken in the reagent
dissolver tank during operation with lime and limestone * The
pH of the slurry recycle loop is highest in the dissolver tank
since fresh basic reagent is added at this point. Dissolver pH
values are significantly lower for limestone operation (4.9-
6.2) than for lime (4.3-9.6). The high level of oxidation
experienced with limestone in the present work contrasts with
results obtained in other FGD studies.
PARTICULATE REMOVAL PERFORMANCE
The R-C/Bahco system utilizes a mechanical collector in
series with two venturi stages located in the scrubber to re-
move particulate. This series combination results in up to 99%
fly ash particulate removal. The mechanical collector protects
the forced draft fan against excessive wear by removing 70-80%
of the total fly ash, particularly the larger particles (>5-10
microns). The scrubber removes essentially all of the remaining
particles above 1-2 microns.
Andersen impactor tests were conducted at Rickenbacker to
determine particulate collection efficiency. AS shown in
Figure 9, the results indicate an aerodynamic cut-off particle
size (i«e., 50% collection efficiency) of 0.6-0,7 microns.
These high collection efficiencies were obtained at an inlet
fly ash loading of 0.2-0.3 grains/SCF using an L/G of 15-25
gpm/1000 ACFM for each venturi stage.
1100
-------
TABLE 1
LIME AND LIMESTONE SLURRY ANALYSES
Lime Slurry Limestone Slurry
Slurry Solids Weight % Weight %
CaSO4 • 2H2O
CaSO3 -1/2H2O
CaCO3
MgC03
Acid Insolubles &
Others
TOTAL
33.4
54.5
3.7
—
8.4
100.0
62.3
16.1
14.2
0.8
6.6
100.0
-------
O
N>
9.0
8.0
I
a.
oc
UJ
CO
O
7.0
6.0
5.0
O LI ME TESTS
OLIMESTONE TESTS
O O
4.0
I
I
I
I
0
0.2 0.4 0.6 0.8 1.0 1.2 1.4
STOICHIOMETRY
FIGURE 8: The effect of lime and limestone stoichiometry on
dissolver pH.
1.6
-------
100
90
80
o
z
UJ
o
u_
u_
UJ
O
O
UJ
d
o
70
60
50
40
30
10
0
I
I
3 4
PARTICLE SIZE, MICRONS
FIGURE 9: R-C/Bahco System Collection Efficiency.
-------
The R-C/Bahco scrubber exhibited the typical high per-
formance expected from a device that removes particulate
primarily via an inertial impaction mechanism. However,
significant amounts (0.1-0.2 gr/SCF) of sooty, fine partic-
ulate ( <1 micron) were encountered in the Rickenbacker
flue gas. This type of particulate is generated when com-
bustion is incomplete. Under these conditions, the col-
lection efficiency dropped significantly for this fraction,
as Figure 9 indicates.
Under conditions where fine particulate was generated at
RAFB, considerably higher pressure drop was required to meet
guarantee emissions rate than anticipated for a fly ash
particulate. At least 18 inches W.C. total pressure drop was
employed in the two R-C/Bahco scrubber venturi sections to
meet guarantee, as Figure 10 shows.
The amount of soot found in the flue gas at RAFB is
substantially higher than normal for stoker-fired generators.
As a result, the USAF undertook an extensive program to up-
grade the Heat Plant at RAFB. Information Research-Cottrell
obtained during the EPA test program contributed significantly
to the specific tasks of the upgrading program. This in-
cident illustrates how pollution control findings may be
translated into substantial fuel cost savings in the boiler
operation.
R-C/Bahco Fractional Efficiency Results
Fractional particulate removal efficiencies by the R-C/
Bahco scrubber were obtained over a wide range of operating
conditions. Key process variables were related to 90% and
50% collection efficiencies.
The particle cut-off diameter at which 90% collection
efficiency was obtained varied from 0.7 to 1.25 microns. In-
creasing the total scrubber pressure drop increased collection
efficiency and decreased the particle diameter at which 90%
collection was observed. Figure 11 indicates the effect of
the combined pressure drop of the first and second scrubber
stages on the particle size at which 90% collection was achiev-
ed. Particulate collection efficiency in the 90% range was
only a function of pressure drop and appears unaffected by
other system variables.
For 50% particulate collection efficiency, the cut-off
diameter ranged from 0.4 to 0.7 microns and is related to
scrubber pressure drop and gas flow rate. Figure 12 shows
1104
-------
CO
0.60
0.50
I
5 0.40
Q
CO
CO
0.30
0.20
__. A
0.10
I
I
GUARANTEE EMISSION RATE
I
I
10 14 18 22 26 30
TOTAL SCRUBBER PRESSURE DROP - IN. W.C.
FIGURE 10: The effect of scrubber pressure
drop on particulate emission rates.
1105
-------
1.2
1.1
CO
o 1.0
o
Ol
a.
Q
0.9
0.8
0.7
GAS FLOW
O = 50,000 SCFM
O =^42,000 SCFM
£ = 35,000 SCFM
10 12 14 16 18 20
TOTAL SCRUBBER PRESSURE DROP IN. W.C.
22
24
FIGURE 11: The effect of total scrubber pressure drop on the
cut off diameter for 90% collection efficiency.
-------
Q
o
in
Q.
Q
0.9
0.8
0.7
0.6
0.5
0.4
= 50,000 SCFM
=42,000 SCFM
=35,000 SCFM
0.3
I
I
I
I
10
12 14 16 18 20
TOTAL SCRUBBER PRESSURE DROP — IN.W.C.
22
24
FIGURE 12: The effect of total scrubber pressure
drop and gas rate on the cut off
diameter for 50% collection efficiency.
-------
that increasing the pressure drop and the gas rate enhances
particulate collection as indicated by the decrease in the
diameter observed for 50% collection efficiency.
Further analysis of the fractional efficiency information
using a penetration model led to several interesting conclusions
regarding particulate collection in the two stage R-C/Bahco
scrubber:
• Larger particles (>1 micron) are collected in
the first (lower) scrubber stage. Ninety-eight +%
of the particles in the 2.0-5.0 size range are
collected in this stage.
o Fine particles are collected primarily in the
second scrubber stage. About 40-50% of the 0.3-0.5
micron particles and 60-65% of the 0.5-1.0 micron
particles are collected in the scrubber with most
of the collection accomplished in the second stage.
Since large size particles are collected efficiently in
the first stage, very few large particles remain to be col-
lected in the second stage. In addition to collecting large
particles, first stage gas cooling and humidification may
condition the fine particles via nucleation and growth
mechanisms to enhance collection in the second stage* Thus,
each scrubber verituri stage functions to collect a different
portion of the total particulate size distribution.
R-C/BAHCO CAPITAL COSTS
Nine standard R-C/Bahco module sizes are available for
gas treatment. Installed capital costs for the R-C/Bahco
scrubber system vary from about $35/SCFM for a size 100
module to $50/SCFM for a size 20 module. Based on 2000 SCFM/
MW/ the equivalent costs are $70/KW for size 100 and $100/KW
for a size 20 module. For a typical project scope, Figure 13
presents R-C/Bahco system turnkey costs for the different
size modules. Site specific conditions including SO2 and
particulate cleaning requirements, available space, sludge
disposal requirements and ductwork arrangements, of course,
will dictate the scope and influence the cost.
Size Selection
The R-C/Bahco scrubber system is particularly suitable
for small and medium size industrial installations with a
single module treating a gas stream in the range of 10,000
to 90,000 SCFM (4). This range is roughly equivalent to a
1108
-------
CO
O
O
2
LU
CO
CO
20
30
40
50 60
(THOUSANDS)
FLUE GAS FLOW RATE, SCFM
70
80
90 100
FIGURE 13: R-C/Bahco Capital Costs.
1109
-------
plant generating 30,000 to 300,000 Ibs. steam/hour. Figure
14 presents a chart to help determine the proper size R-C/
Bahco scrubber module for a particular application. The
maximum, minimum and average flue gas rates encountered at
Rickenbacker AFB are depicted in the chart along with the
most efficient operating range for each size module. During
low load operation, makeup air is added to the scrubber in
order to maintain high pollutant removal efficiencies.
A size 50 R-C/Bahco scrubber, which handles a gas volume
corresponding to 200 million BTU/hr. firing rate, was selected
for the RAFB installation. Combinations of the boilers at
RAFB are operated from 58 to 208 million BTU/hr. of fuel fired,
resulting in a maximum winter rate nearly four times the summer
rate. However, in 1973 when the project was first considered
the 200 MM BTU/hr. firing rate was exceeded on only one day
(by 5%). In view of the rare need for extra capacity, the
additional cost and the increased turn-down ratio required
for the next larger size R-C/Bahco scrubber, the size 50 module
was chosen.
The R-C/Bahco scrubber system at Rickenbacker, shown in
Figure 1, occupies about 2,500 square feet and the scrubber
plus stack is eighty-one feet high. Besides the scrubber,
the lime silo, thickener, fan and control house take up most of
the space required. The remaining space is used for pumps,
piping, ductwork and access. A 2.5 acre sludge disposal pond
at Rickenbacker is located about 750 feet from the scrubber.
Sludge handling development work carried out by Research-
Cottrell since the design of the Rickenbacker Air Force Base
system has produced more efficient sludge dewatering tech-
niques. Hydroclones and centrifuges can now be utilized in-
stead of a thickener to permit a more compact design while
producing a more concentrated, truckable sludge which may be
suitable for disposal in landfill areas. As a result, signif-
icantly less space than that used at RAFB may be required for
the system.
ANNUAL OPERATING COSTS
R-C/Bahco system annual operating costs based on Ricken-
backer AFB operation are presented in Table 2. A total annual
operating cost of about $235,000 is estimated for lime and
$212,000 for limestone. The estimates are based on operation
in a size 50 scrubber module treating flue gas generated by
40,000 toms/yr. of 3.5% sulfur coal. The costs are equiv-
alent to about $5.90/ton of coal burned for lime and $5.30/ton
for limestone. As Table 2 shows, power and reagent consumption
1110
-------
RAFB FLOW RATE
MINIMUM
AVERAGE
MAXIMUM
UJ
N
CO
cc
UJ
g
3
cc
to
0
O
O
cr
RAFB SCRUBBER SELECTED
MOST EFFICIENT OPERATING RANGE
1 1 1
1 1 1 1
1 1
1 1
0 10 20 30 40 50 60 70
(THOUSANDS)
FLUE GAS FLOW RATE, SCFM
80
90
100
110
120
FIGURE 14: R-C/Bahco Scrubber Sizing.
-------
UTILITY
TABLE 2
R-C/BAHCO ANNUAL OPERATING COSTS
(Based on 40,000 tons/yr. of 3.5%
S coal consumption, 70% SO2 removal)
REQUIREMENT
UNIT COST
($/Unit)
ANNUAL COST
($)
Lime Limestone
Power
Water
CHEMICALS
Pebble Lime
Limestone
OTHER EXPENSES
Operator
Supervision (25% of labor)
Maintenance M&L
Direct Overhead (75% of
op. & Maint. Labor)
Taxes and Insurance
350 KW
10GPM
0.2 tons/hr.
0.41 tons/hr.
0.25 man/shift
0.027/KWH
0.37/1000 gal.
40.35/ton
12.72/ton
8/hr.
10/hr.
78,250
1,950
66,820 —
43,390
16,560
5,170
12,600
19,170
35,000
TOTAL ANNUAL OPERATING COST 235,520 212,090
OPERATING COST/TON COAL 5.90 5.30
-------
are the two principal contributors to expenses, comprising
more than half of the annual operating cost. During periods
of inefficient boiler operation at Rickenbacker, where very
high excess air (120-250%) was utilized or copious amounts
of soot were generated, up to 35% additional power was employed
to cope with the situation.
At RAFB, limestone is more economical to use than lime
despite the fact that more than twice as much limestone is
needed to obtain the same S02 removal. The price of lime-
stone delivered to RAFB during 1977 was $12.72/ton compared
to $40.35/ton for lime. This large price diffenential means
that for every $100 spent for limestone, greater than $140 must
be spent for lime for equivalent performance to meet code.
Limestone has other advantages over lime because it is
not hygroscopic and need not be slaked thereby eliminating the
need for a slaker. Also, limestone is less likely to cause
injuries to operating or maintenance personnel since it does
not exhibit the caustic properties inherent to lime. For
these reasons, RAFB switched from lime to limestone reagent
in May, 1977, and has since experienced smooth, effective
scrubber operation as well as considerable cost savings.
The R-C/Bahco system requires only a minimum amount of
operator attention. At RAFB, a scrubber technician operates
the system on day shift Monday through Friday and also handles
routine maintenance. Rickenbacker Heat Plant operating per-
sonnel monitor the scrubber operation during off shifts and
on weekends. The cost for operating labor, including super-
vision and overhead, plus maintenance material and labor is
about $55,000/yr.
Sludge disposal operating costs are.minimal at RAFB since
the sludge is pumped to a disposal pond. The installed pond
cost amounts to $0.45/ton of coal.
ACKNOWLEDGEMENTS
We wish to thank Mr. James Rasor, Base Civil Engineer,
Rickenbacker Air Force Base, Lockbourne, Ohio for his help
and cooperation in enabling Research-Cottrell to conduct the
R-C/Bahco test program at Rickenbacker. The direction and
aid of Mr. John Williams, EPA Project Officer, is also
greatly appreciated. R-C personnel who contributed heavily
to the commercial job and the EPA test program are Robert
Ferb - Project Manager, Edward Biedell, David Ruff and Gary
Malamud.
1113
-------
REFERENCES
1. R. S. Atkins, Advances in Chemistry Series No. 127,
Pollution Control and Energy Needs.
2. EPA Evaluation on Bahco Industrial Boiler Scrubber
System at Rickenbacker AFB, EPA-600/7-78-115, June, 1978
3. RAFB R-C/Bahco Capsule Report,EPA Contract No. 68-03-
1333, In Press.
4. N.. J. Stevens and R. J. Ferb, R-C/Bahco Scrubber for
Industrial Pollution Control, 16th Annual Purdue Air
Quality Conference, November, 1977.
1114
-------
Status of the Project to Develop Standards of Performance for
Industrial Fossil-Fuel-Fired Boilers
L. D. Broz, G. R. Offen, P. D. Anderson,
Acurex Corporation
Raleigh, NC
and
J. D; Mobley, C. B. Sedman,
U.S. Environmental Protection Agency
Research Triangle Park, NC
ABSTRACT
The Environmental ^Protection Agency has undertaken a study of indus-
trial boilers with the intent of proposing standards of performance based
on information gathered during the project. The study is being directed by
EPA's Office of Air Quality Planning and Standards, and technical support
is being provided by the Agency's Industrial Environmental Research Labora-
tory at Research Triangle Park, N.C. Acurex is the systems integration
contractor for both groups on this project. Other EPA offices and support
contractors are also involved.
Through a series of tasks, background information is being collected
on industrial coal-, oil-, and gas-fired boilers and the technologies that
have demonstrated the ability to reduce nitrogen oxides, sulfur oxides, and
particulate emissions from these sources. This information collection
activity is nearing completion and will be followed by extensive analyses
of the potential economic, energy, and environmental impacts of alternative
regulatory options. The data and analyses will be documented in a Back-
ground Information Document and used to develop and support a New Source
Performance Standard for the industrial boiler source category.
1115
-------
INTRODUCTION
The Clean Air Act, as amended in 1977, provides authority for the U.S.
Environmental Protection Agency (EPA) to control the discharge of air
pollutants into the atmosphere. The Act contains several regulatory and
enforcement options for control of emissions from stationary sources.
Options include (1) National Ambient Air Quality Standards (NAAQS) on the
national level and State Implementation Plans (SIP's) on the state level,
(2) new source performance standards (NSPS) on the federal level, and (3)
national emission standards for hazardous air pollutants (NESHAPS).
Section 111 of the Act calls for issuance of emission standards for
emissions from new and modified sources which may contribute significantly
to air pollution, and which could endanger public health or welfare. The
standards must reflect the best degree of control, taking cost, energy, and
nonair environmental quality impacts into account. No new plant can be
built, however, if it will violate a NAAQS, even if it meets the NSPS.
Amendments to the Act in 1977 specifically mention the need to develop
standards for fossil-fuel-fired boilers. Further, in a prioritized list of
sources for which standards of performance should be developed, industrial
boilers rate 13th of 66 source categories. Accordingly, EPA has undertaken
a study of industrial boilers with the intent to propose standards of
.performance, or NSPS, based on information gathered during the project.
Further, to accommodate the accelerated timetable for the development of
NSPS, as mandated by the 1977 amendments, EPA has greatly accelerated its
use of contractors to develop these standards. Acurex Corporation, one of
several contractors serving EPA in this capacity, has been assigned the
industrial boilers source category.
1116
-------
Much of the background information concerning the control technologies
for this source category has been or is being developed by EPA's Industrial
Environmental Research Laboratory—Research Triangle Park (IERL-RTP) and
the Office of Air Quality Planning and Standards (OAQPS). OAQPS is lead
office on the project and ultimately responsible for development of the
standard. In its support role to OAQPS, IERL-RTP is providing direct
assistance by developing information concerning control technologies.
OAQPS will use this and other information to prepare a Background Infor-
mation Document (BID), which will form the basis for the standards to be
proposed. The input from IERL-RTP will be in the form of Technology Assess-
ment Reports (TAR's), which will be referenced in the BID. These reports
will also document IERL-RTP's assessment of the state of the art in control
technology for industrial boilers. In addition, they will be used within
EPA's Office of Research and Development (ORD) to evaluate the proposed
emission standards.
Acquisition of background information by IERL-RTP, OAQPS, and contrac-
tors is directed by OAQPS and will culminate in preparation of the BID. To
avoid duplication of effort in preparing the TAR's and the BID, IERL-RTP-
and OAQPS are utilizing the same integration contractor, Acurex, directed
by an OAQPS Task Manager.
Current Regulations for Industrial Boilers
One measure of the impact of an NSPS on emissions is an evaluation
relative to typical state regulations. Hence, these limits are summarized
briefly here. Most states restrict particulate matter (PM), sulfur oxides
l»
(SO ) and nitrogen oxides (NOV) emitted from industrial boilers. Typical
X X
emission limits range from 43 to 344 ng/J (0.1 to 0.8 lb/106 Btu)
1117
-------
heat input for participates and 86 to 3870 ng/J (0.2 to 9.0 lb/106 Btu)
heat input for SO . Typical NO emission limits range from 86 to 387 ng/J
S\ f\
(0.2 to 0.9 lb/106 Btu) heat input; however, 24 states have no regulations
for NO emissions.
A federal standard of performance currently exists for large boilers.
This standard applies to all fossil-fuel- and wood-fired steam generating
units capable of firing at a heat input of more than 73 MW (250 x 106 Btu
per hour) and constructed or modified after August 17, 1971. EPA recently
proposed a revision to this standard which will apply only to electric
utility steam generating units. The existing NSPS for fossil-fuel-fired
steam generators will continue to apply to industrial boilers with heat
input greater than 73 MW (250 x 106 Btu per hour), until revised as a
result of this study.
Scope of Industrial Boiler Source Category
The industrial boiler source category currently under consideration
for regulation by an NSPS includes all non-residential and non-utility
boilers. Therefore, all institutional, commercial, and industrial boilers
are included, although this study is commonly referred to as the industrial
boiler NSPS activity. The population of industrial boilers consists of
many different types of units, which use various fuels and methods of heat
transfer. This study, however, deals only with fossil-fuel-fired boilers;
separate EPA studies are evaluating the feasibility of standards on waste-
fired boilers and incinerators. On the other hand, technologies such as
fluidized bed combustors (FBC), synthetic-fuel-fired boilers, and fuel
cleaning are being investigated.
1118
-------
Pollutants of Concern
Emphasis is being placed on the criteria pollutants emitted by combus-
tion sources--PM, SO , and NO . Growth projections for industrial fuel
X X
usage vary over a wide range, but a value of 3.7 percent compounded annual-
ly appears reasonable. Based on this nominal growth rate projection,
emissions from industrial boilers would increase to nearly 250 percent of
their current level by the year 2000 if not controlled beyond present
regulations.
Total particulate emissions depend primarily on the quantity of coal
used and the ash content. Hence, as fuel conversion strategies are imple-
mented, particulate levels will increase. Total SO emissions will change
A
in direct relation to changes in fuel sulfur content of the fuels burned;
thus, they will also increase somewhat beyond the levels that would be
otherwise projected to the extent that coal replaces expected natural gas
and distillate oil consumption. For the same reasons—growth arid emphasis
on coal utilization—NO emissions will change considerably from current
J\
levels (NO emissions from coal-fired units are 200 to 300 percent greater
/\
than from distillate oil or natural gas boilers).
GENERAL APPROACH
The general approach being taken to develop the NSPS for industrial
boilers follows:
1. Characterize the source category and current emission rates,
either controlled or uncontrolled as applicable, to meet SIP's.
2. Collect process information and performance data on boilers with
emission control systems:
- from available literature
1119
-------
- from emission tests at sites with candidate best demonstrated
controls.
3. Identify potential changes to existing sources that could be
deemed modifications or reconstructions, as defined by EPA regu-
lations, and which thus would require compliance with the NSPS.
4. Select model plants (model boilers) for use in site specific
economic, environmental and energy impact analysis. Model boilers
are combinations of representative boilers and control systems.
5. Utilize a computer model to predict the potential economic,
environmental, and energy impacts due to alternative regulatory
options.
6. Determine the potential economic, environmental (all media), and
energy impact due to each of the alternative regulatory options,
using both the computerized analysis and model boilers.
7. Recommend a standard based on the best demonstrated technological
system of continuous emissions reduction considering economic,
environmental, and energy factors.
Formal proposal of the NSPS is preceded by circulation of a draft BID,
which discusses the emission sources and emission control alternatives and
assesses the performance, cost, energy requirements, and overall environ-
mental impact of each alternative control system. Various environmental
groups, industry, and other interested parties participate in formal review
of the document which culminates in a public meeting before the National
Air Pollution Control Techniques Advisory Committee (NAPCTAC). Following
NAPCTAC review, the document and recommended standards are subjected to
additional intragovernmental review before the final standards package is
1120
-------
approved for proposal by the Administrator. The proposed NSPS are then
published, in the Federal Register, inviting public comment. Once all
comments have been received and resolved to the Administrator's satisfac-
tion, final standards of performance are promulgated by publication in the
Federal Register.
•:V
Project Organization
*
Figure 1 illustrates the organization of responsibilities. Basically
it .shows the following:
• The Emission Standards and Engineering Division (ESED) of OAQPS
is the lead organization with ultimate responsibility for develop-
ment of the standard. As such, it directs the identification and
evaluation of regulatory options, considers public comments, and
oversees the preparation of the documents needed to propose and
promulgate a standard. In this study, ESED is:
Supported by Acurex, as systems contractor, to: (1) integrate
the effort which leads to the BID and recommended standard;
(2) provide support during the review/proposal/promulgation
cycle; and (3) ensure that all technology contractors use a
consistent approach.
Further supported by additional contractors with specific
data collection and analysis assignments.
ESED is also responsible for:
Source testing, in conjunction with IERL-RTP.
Coordinating dispersion calculations with the Source-Recep-
tor Analysis Branch, Monitoring and Data Analysis Division,
OAQPS.
1121
-------
OAQPS/SASD
Impact Analysis
Project Qfficer
OAQPS/ESED
Lead Engineers
Contractor
Systems
Contractor
IERL-RTP
Technology
Coordinator
I
IERL-RTP
Project Officers
N>
S3
Emission
Measurement
Project Officer
Dispersion
Calculations
Project Officer
ESED
Project Officers
Technology
Contractors (9)
Source Test
Contractors
Support
Contractors
Figure 1. Project organization,
-------
• IERL-RTP actively supports OAQPS on this project. In this role
IERL-RTP:
Provides technical input on boiler characteristics and
emissions and on available control technologies, including
data on performance, costs, energy requirements, and other
environmental impacts of the controls.
Assists in selecting alternative control options and model
boilers.
• The Strategies and Air Standards Division (SASD) of OAQPS also
supports the project actively by:
Providing analyses of the nationwide economic, environmental,
and energy impacts associated with each alternative control.
Considering relationships between possible standards and fuel
usage patterns.
Support by IERL-RTP was planned and arranged from the outset of the
industrial boiler project. This approach utilizes the Laboratory's avail-
able expertise from ongoing R&D projects. Through its in-place contractual
mechanism for obtaining outside support from contractors with control
technology expertise, IERL-RTP is able to provide information to OAQPS in a
timely fashion. This capability enables EPA to meet its stringent schedule
for setting NSPS.
The effort to collect process and control technology information and
prepare the BID is divided into the 15 tasks listed in Table 1. Figure 2
illustrates the interrelationships between these tasks by showing both the
1123
-------
Table 1. INDUSTRIAL BOILER TASKS
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
Task
Boiler Classification
Emission Data
S IP's /Emission Estimates
Modifications
Fuel Switching
Cogeneration
Waste Firing
Model Boilers
Dispersion Analyses
Performance Tests
Control Technologies
Emerging Technologies
Fuel Use/Energy Impacts
Environmental Impacts
Economic Analysis
Responsibility
IERL-RTP
IERL-RTP/OAQPS
OAQPS
OAQPS
OAQPS
OAQPS
OAQPS
IERL-RTP/OAQPS
OAQPS
IERL-RTP/OAQPS
IERL-RTP
IERL-RTP
OAQPS
OAQPS
OAQPS
Due Date
6/78
6/78
9/78
9/78
9/78
9/78
9/78
2/79
6/79
7/79
8/79
8/79
8/79
10/79
12/79
-------
BOILER
CLASSI-
FICATION
IERL-RTP
BID
CH. 3
MODIFI-
CATIONS
OAQPS
BID
CH.,5
OAQPS/IERL-RTP
FUEL
SWITCHING,
COGENERATION
OAQPS
f
i
•10DEL
3fiT 1 FD
3U1 LC.K
INDIVIDUAL TECHNOLOGY
ASSESSMENT REPORT ClTAR)
IERL-RTP
BID
CH. 4
EMISSION
TESTING
OAQPS/IERL-RTP
ENVIRONMENTAL IMPACTS
OAQPS
ENERGY IMPACTS
OAQPS
COMPREHENSIVE TECHNOLOGY
ASSESSMENT REPORT (CTAR)
IERL-RTP
COST ANALYSIS
OAQPS
ECONOMIC IMPACTS
OAQPS
ALL BID CHAPTERS PREPARED BY ACUREX
_L
M
A S
1978
N
M
M
J J
1979
N
Figure 2. IERL-RTP/OAQPS integrated tasks.
-------
timetable for completion of each task and the BID chapter which it supports
*
The key topics addressed in each chapter are listed in Table 2.
Tasks 1-7 were completed in 1978 and define and scope the problem.
Task 8 establishes the model boilers for the economic, environmental, and
energy analysis of alternative standards. Most of the technical, environ-
mental, and cost data that will be used to support a standard are being
collected under Tasks 9-12. Descriptions and performance data on the
demonstrated controls are being documented in eight Individual Technology
Assessment Reports (ITAR's). As explained in a later section of this
paper, each ITAR covers one control technology. Fluidized bed combustors,
synthetic fuels, and/or cleaned fuels are considered potential controls for
the purpose of this study. All eight ITAR's will be summarized by Acurex
in a document entitled a Comprehensive Technology Assessment Report (CTAR>.
Tasks 13-15 provide for the analysis of the potential impacts due to each
alternative regulatory option.
ISSUES ANALYSES
During the early stages of this project, several studies were con-
ducted to scope the project and investigate issues. The tasks for these
projects are listed in Table 1 as Tasks 1-7 and are described briefly here.
The boiler classification study of Task 1 characterized industrial
boilers, provided installed capacity and projected growth data, developed
baseline emissions data, and identified the effects of boiler design,
operation, and fuel characteristics on emission levels.
X
Figure 2 shows only the technology and impact analyses chapters that
follow directly from the 15 tasks in Table 1. Supporting appendices have
been omitted from Figure 2 for clarity. All of these are used to prepare
Chapters 1, 2, and 9, thereby completing the BID.
1126
-------
Table 2. CHAPTERS IN A BACKGROUND INFORMATION DOCUMENT
10
-vl
1. Summary
2. Introduction
3. Industrial Boilers
4. Control Technologies
5. Modifications and Reconstructions
6. Best Systems of Emission Reductions
7. Environmental Impacts
8. Economic Impacts
9. Rationale
Appendices
Summarizes the recommended standard, with
rationale, and estimates associated impacts.
States the legislative basis and administra-
tive process for this regulatory action.
Provides a description of the source, its
uses, current and projected population/emis-
sions, and applicable SIP's.
Describes the demonstrated and available
controls and gives data on their emission
reduction capability.
Describes possible changes to existing
boilers to help decide, later, which
should be considered modifications or re-
constructions .
Develops model boilers with candidate
best systems and alternative regula-
tory options to be analyzed in the
following chapters.
Identifies and discusses all environmental
and energy impacts (beneficial and adverse)
due to each alternative regulatory option.
Establishes and evaluates the expected eco-
nomic impacts of each alternative regula-
tory option.
Provides the rationale for the recommended
standard.
Contain test data and information on the
history of the development of the standard,
possible test methods, and enforcement issues,
-------
Task 2 was a coordinated activity, involving both OAQPS and IERL-RTP,
that determined the availability of sufficient data to support standards.
As will be discussed later, EPA identified data gaps during this task and
proposed a plan for conducting the necessary tests.
To evaluate the impact of emissions under each regulatory option
relative to those that would occur if the boiler were required to comply
only with a typical SIP level, Task 3 (SIP's/Emission Estimates) was in-
cluded in this project. The object of this task was to define an average,
or typical, SIP. This SIP level was also used to compute an average emis-
sion estimate for the U.S. population of boilers.
Since any existing boilers which are modified and reconstructed after
the date of proposal of an NSPS must comply with the standard, Task 4
(Modification) was conducted to define the terms modifications and recon-
structions for use in the standard writing procedure. In this task, the
contractor identified likely changes to existing boilers that would be
considered modifications or reconstructions according to the definition of
those terms in the Code of Federal Regulations.
The effects on fuel usage patterns are an important consideration in
this study. Therefore, Task 5 (Fuel-Switching) was issued to describe the
technical factors involved in fuel switching and identify the liklihood of
such changes among existing installations. To make this prediction, the
task also estimated the cost and environmental impacts which result when
industrial boilers currently burning gas or oil switch to oil, coal, or
coal/oil mixtures.
Task 6 (Degeneration) reports information on cogeneration, including
an explanation of the concept of cogeneration, identification of those
1128
-------
industries most likely to employ cogeneration, emission levels from boilers
used in (regeneration systems, applicable emission control techniques, and
projected growth of capacity.
Task 7 (Waste-Firing) made a determination on the status of boilers
firing industrial wastes. The result of this study was to establish waste-
fired boilers as a separate NSPS activity.
CONTROL TECHNOLOGY ASSESSMENT
A key element in the process of developing an NSPS is the assessment
of relevant demonstrated-control technologies. Out of this assessment
comes a list of candidate best systems, based on performance, commercial
availability, adaptability to the process, etc. These candidate systems
are then compared, and one or several may be selected as the basis for the
standard, taking into account emission reduction capability, cost, energy
penalty, and environmental side effects.
Standard Boilers
In this program the detailed assessment began with a survey of the
industrial boiler population to better define and scope the study (see Task
1, above). Analysis of the distribution of installed boiler capacity and
projected sales according to the boiler type, size, and fuel led to a
classification of the members of this diverse category. A "standard"
boiler was then selected to represent each class. Use of a classification
scheme and "standard" boilers provided a common basis for all control
technology contractors when assessing the overall performance of their
assigned technology.
The following criteria were used to select the representative boilers:
1129
-------
• Extent of usage.
• Potential for uncontrolled emissions of particulate matter, SOX,
and NOX.
t Representation of a cross section of the population.
• Projected sales.
Seven "standard" boilers (Table 3) were selected as representative of
the industrial boiler population for use in the model boiler analysis. In
addition, three coals were designated for the coal-fired boilers to repre-
sent the range of control requirements and associated impacts that could be
expected from standards of performance. These are:
Higher Heating Value
Coal Sulfur (%) Ash (%) (kJ/kg [Btu/lb])
High sulfur eastern 3.5 10.6 27,450 (11,800)
Low sulfur eastern 0.9 6.9 32,100 (13,800)
Low sulfur western 0.6 5.4 22,300 (9,600)
Additional boilers and coals are being considered for use in an associated,
more detailed computerized impact analysis.
Design parameters were also identified for each boiler to further
ensure a common basis for reporting control technology capabilities. These
parameters are: fuel rate; flue gas temperature and composition; load
factor; excess air; and flue gas rate.
Individual Control Technology Assessments
For the technical and cost evaluations of the control techniques,
IERL-RTP, with OAQPS agreement, divided the spectrum of possible technolo-
gies into eight generic types and assigned each one to a contractor with
specific expertise in that technology. Table 4 lists these eight technoTo-
1130
-------
Table 3. STANDARD BOILERS
Type
Fuel
Capacity
MW (106 Btu/hr)
Package, firetube
Package, firetube
Package, underfeed stoker,
watertube
Field erected, chain grate
stoker, watertube
Package, watertube
Field erected, spreader stoker,
watertube
Field erected, pulverized coal,
watertube
Oil (distillate)
Natural gas
Coal
Coal
Oil (residual)
Coal
Coal
4.4 (15)
4.4 (15)
8.8 (30)
22 (75)
44 (150)
44 (150)
59 (200)
Table 4. TECHNOLOGY ASSESSMENTS
Project Title
1. Characterization of the Industrial Boiler Population
2. Clean Oil and Oil Treatment Technology
3. Coal Cleaning and Existing Clean Coal
4. Synthetic Fuels
5. Combustion Modification for NO Control
J\
6. Fluidized Bed Combustion
7. Particulate Collection Technology
8. Flue Gas Desulfurizatton Technology
9. NO Flue Gas Treatment Technology
^
10. Comprehensive Technology Assessment Report (CTAR) and General Support
1131
-------
gies. It also shows the boiler population characterization study mentioned
earlier and the task given to Acurex to help integrate these eight studies
and compile their results into a single comprehensive report.
As noted earlier, the product of each control technology study will be
an Individual Technology Assessment Report (ITAR). Within each ITAR the
control systems that belong to that generic type are discussed separately.
For example, the FGD report contains subsections dealing with lime/lime-
stone, sodium scrubbing, Wellman-Lord systems, etc. To provide a complete
evaluation of the capabilities of each control technology, the topics
listed in Table 5 were addressed by every contractor.
Since control technologies operate over a wide range of emission
reductions, three levels of control were identified for each technology and
boiler/ fuel combinations at each level of control. In effect, the report
for each technology documented the performance and impacts associated with
the control of a single pollutant.
Information on each control technology is presented in the ITAR's
according to the outline shown in Table 6. These ITAR's are currently
being completed and will then be integrated into a Comprehensive Technology
Assessment Report (CTAR) by Acurex. This overview document is intended to
provide, in a single report, EPA's assessment of the state of the art in
the control of NO , SO , and particulate emissions from fossil-fuel-fired
/\ J\
industrial boilers. It will contain detailed discussions of individual
technologies and of the systems that could be assembled to control all
three pollutants simultaneously. The CTAR will discuss technologies that
are in various stages of development and commercialization as well as those
which have been demonstrated and are widely available.
1132
-------
Table 5. INDIVIDUAL TECHNOLOGY ASSESSMENT REPORT ISSUES
Status of development.
Applicability of control system to different boiler sizes and types.
Estimated capital and operating cost of the control system as a function
of boiler size and type.
Control system cost as a function of removal capability.
Energy requirements of the control system.
Reliability of control system.
Environmental impact of waste streams.
Vendor availability.
Compatibility with other control systems.
Performance and operating data.
Table 6. OUTLINE OF THE INDIVIDUAL TECHNOLOGY ASSESSMENT REPORTS
I. Executive Summary
II. Emission Control Techniques
III. Candidates for Best System of Emission Reduction
IV. Cost Analysis of Candidates for Best System of Emission Reduction
V. Energy Impact of Candidates for Best System of Emission Reduction
VI. Environmental Impact of Candidates for Best System of Emission
Reduction
VII. Emission Source Test Data
The I"(AR's have been used as the basis for selecting model boilers for
X
the BID. These model boilers are a combination of "standard" boilers and
candidate best control systems jointly accepted by OAQPS and IERL-RTP. The
1133
-------
"model boilers" are those which can be used by a boiler owner or operator
to meet the prescribed emission level (i.e., one of the three discussed
earlier), given a specified fuel.
The ITAR's and the CTAR form a partial basis for the control techno-
logy chapters of the Background Information Document (BID) prepared under
the direction of OAQPS to support the proposed standard. The ITAR's and
the CTAR contain much of the technology information that goes into the BID
and provide in-depth coverage of all available and emerging control techno-
logies. The BID, then, summarizes the key performance and operational
characteristics of the demonstrated systems, stressing those which can be
used by boiler owners under alternative regulatory options.
DATA GATHERING
Where data gaps exist, further performance testing is required. Test-
ing is a joint activity of OAQPS and IERL-RTP. Testing requirements for
data gathering were determined by summarizing the available test data and
then assessing their quality. Criteria of acceptability included the
presence of a complete process characterization of the boiler during the
test, proper and representative operation of the boiler during the test,
the use of Federal Reference Methods (FRM) to measure emissions, and a test
of sufficient duration to assess the impact of coal variability. A matrix
of available good quality data was established and compared to the data
needed to support any probable standard at all levels, from moderate to
stringent. That is, comparisons were used to determine, for example, if
sufficient quality data existed to propose and defend an intermediate S02
standard for a medium sized stoker-fired boiler.
1134
-------
Priorities were set for filling in the data gaps on the basis of the
expected need for that data, the extent of the data inadequacy, the cost of
obtaining the data, and the availability of test sites. Tests were initiated
in October 1978. Key tests in the series are: (1) continuous monitoring to
obtain 30 days of data at boilers with flue gas desulfurization (FGD) and
NO combustion modifications; (2) tests using FRM 5 at sites for particulate
j\
control of small oil- and coal-fired boilers; and (3) additional data on
low Btu gasification, and fluidized bed combustion.
IMPACT ANALYSIS
Impacts which can be expected to result from the implementation of
controls to meet alternative regulatory options are studied and discussed
in the BID. These include potential environmental, energy, and economic
consequences from these alternative actions. In the environmental analysis
both primary impacts (e.g., those directly attributable to each control
system, such as reduced levels of specific pollutants) and secondary im-
pacts (e.g., indirect or induced impacts, such as exacerbation of another
pollutant problem through utilization of a control system) will be identi-
fied and discussed. Primary emphasis will be placed on changes to the
ambient air surrounding a typical source (NO , SO , and PM) and to changes
X X
in nationwide emissions, but attention will also be paid to other media
(e.g., FGD sludge). Both beneficial and adverse effects wil-1 be assessed.
The major emphasis will be on providing the reader with an accurate assess-
ment of the incremental impact of the regulatory alternatives compared to
sources which are uncontrolled or controlled to meet state regulations.
In addition, irreversible and irretrievable commitment of resources
will be considered. As appropriate, this section will include a discussion
1135
-------
of the extent to which the alternative emission control systems may involve
a trade-off between short-term environmental gains at the expense of long-
term environmental losses, or vice versa, and the extent to which the
alternative systems may result in irreversible and irretrievable commitment
of resources.
The potential impact of an NSPS on energy consumption will be assessed
by determining the amount and type (electric, fossil fuel, etc.) of energy
required to operate each alternative emission control system. Where possi-
ble, these values will be compared to the quantities and types of energy
required by a typical facility, both without emission control as well as
controlled to comply with state regulations. Boiler-specific energy penal-
ties will be extrapolated to estimate the impact of each regulatory option
on increased national energy demands and on industry growth projections.
The economic impact of alternative control options will be evaluated
during the preparation of the BID. To provide information for use in
assessing the economic impact on a boiler owner of the cost associated with
control options, a "business/economic" profile of the industry will be
presented. Included in this discussion will be such factors as: industry
structure, industry statistics, markets, prices, production and capacity,
competition from imports, and other background information and data, as
appropriate.
Following this industry overview, cost data will be presented for each
emission control system. These costs will be given for new boilers and,
separately, for existing units to cover modifications and reconstructions.
Based on these results, the incremental costs associated with the alterna-
tive regulatory options will be identified. If the costs associated with
1136
-------
possible emission monitoring or compliance testing could be significant,
these costs will be included in the overall costs. The costs will also
consider thoSe incurred to dispose of, in an environmentally acceptable
manner, any liquid and solid wastes generated by air pollution controls.
The cost data described above will be used to evaluate the economic
impacts associated with the incremental costs imposed on the source(s) to
meet alternative regulatory options. The prime objective of the analysis
and discussion will be to identify the incremental economic impact asso-
ciated with alternative regulatory options.
This analysis will include the potential socio-economic and infla-
tionary impacts that might result from the application of the alternative
regulatory options.
In a typical BID the analysis is based on model plants (i.e., model
boilers for this source category). Due to the complexity of the industrial
boiler source category, a major portion of the analysis in this case will
be performed by using a computer model. Included in these analyses of
potential nationwide impact of various alternative regulatory options are
the effects on fuel usage patterns—for example, the extent to which a
standard on coal boilers might induce users to buy gas- or oil-fired boilers
instead of coal-fired units. This analysis uses the Industrial Fuel Choice
Analysis Model (IFCAM), a model originally developed for the Federal Energy
Administration (now part of the Department of Energy—DOE) and now being
modified for this project. This combination of approaches (i.e., models
and computers) yields local impacts due to controls on representative
samples of boilers plus nationwide impacts due to the application of con-
trols to all affected boilers.
1137
-------
BID PREPARATION
The descriptions, analyses, and results described up to this point
will be documented by Acurex for OAQPS in a Background Information Document
(BID). This report will present a description of industrial boilers and
the control technologies which have demonstrated their ability to control
NO , SO , and PM from such boilers. It will also document the analyses of
r\ f^
the potential economic, energy, and environmental impacts of the various
alternative regulatory options that can be developed for these boilers.
These analyses will assist EPA in arriving at a recommended standard, which
is summarized in the BID and supported by a rationale.
SCHEDULE
The industrial boiler study was initiated in April 1978. It is scheduled
for completion in August 1981, with promulgation of the standard by publi-
cation in the Federal Register. To date, draft ITAR's have been prepared,
preliminary model boilers have been identified for internal review, and
emission testing has begun. Presently targeted dates for the key remaining
milestones are:
• BID completion, March 1980.
• National Air Pollution Control Techniques Advisory Committee
meeting to be held, June 1980.
• Proposed standard published in Federal Register. October 1980.
• Promulgation of standard, August 1981.
Not included explicitly in this list are the numerous informal inter-
actions planned with members from industry, DOE, environmental groups, and
other government agencies.
1138
-------
SUMMARY
This paper has described a multifaceted EPA study that is being con-
ducted with the intent of developing NSPS for industrial boilers. Two
major organizations within EPA — OAQPS and IERL-RTP — are involved, as
well as other EPA offices and support contractors. Through a series of
tasks, background information is being collected on industrial coal-, oil-,
and gas-fired boilers and on the technologies that have demonstrated the
ability to reduce NO , SO , and PM emissions from these sources. This
f\ J\
information collection activity is nearing completion and will be followed
by extensive analyses of the potential economic, energy, and environmental
impacts of alternative regulatory options. The data and analyses will be
documented in a Background Information Document and used to develop and
support an NSPS for the industrial boiler source category.
1139
-------
6F
FLUE GAS DESULFURIZATION APPLICATIONS
TO INDUSTRIAL BOILERS
James C. Dickerman
Radian Corporation
Durham, North Carolina
INTRODUCTION
The Clean Air Act Amendments of 1977 require the Environmental Protection
Agency to coordinate and lead the development and implementation of regulations on
air pollution. These include standards of performance for new and modified sources
of pollution. Specifically mentioned in the Act are fossil fired steam generators.
Accordingly, EPA has undertaken a study of industrial boilers with intent to pro-
pose standards of performance based upon the results of this and other studies.
This paper presents a summary of a study conducted to evaluate the applicability
of various flue gas desulfurization (FGD) technologies for treating S02 emissions
produced from industrial boilers. Results of this evaluation will be used by the
EPA in preparing a NSPS for industrial boilers. Factors that were considered in
evaluating the applicability of FGD technologies to industrial boilers included
development status, capital and operating costs, energy requirements, environmental
impacts, and performance and operating data.
APPROACH
In order to satisfy the objective of this study, a multiphased project approach
was used. First, a comprehensive list of FGD processes was reviewed. This list
included processes in commercial use, processes under development, and processes for
which development efforts have ceased. Process status reports were prepared for
eleven of the processes: those which are currently commercially used or are under-
going major demonstration efforts. Status reports for each candidate process con-
tained detailed process descriptions, and discussions of design considerations and
process performance characteristics.
The second phase involved selecting from the list of eleven candidate processes
those that appeared to be best suited to industrial boiler applications.
1140
-------
Finally, detailed process analyses were prepared for the selected processes.
Material and energy balance calculations were performed for each process to assess
energy and environmental impacts as a function of boiler size, fuel sulfur content,
and percent S02 removal. Also, a series of capital and operating cost estimates
were made to assess the cost impact of each selected process as a function of
boiler size, fuel sulfur content, and percent SO2 removal.
SELECTION OF CANDIDATE PROCESSES
There are currently some 100 FGD processes that are in various stages of
development, including processes in early developmental stages and those for which
development efforts have ceased. Of these processes, there are five that are in
commercial use today in the United States. In addition, there are six that are
currently at the demonstration stage. It is felt that these eleven processes will
be used for the majority of near-term FGD applications for both utility and indus-
trial boilers. Table 1 presents a summaryx of the development status and applica-
bility of these eleven processes to industrial boilers.
Ill order to select the candidate .processes that appeared to be best suited for
industrial boiler applications, a set of evaluation or screening criteria were
established to provide an objective and consistent means of comparing the processes
and to ensure that the same factors were considered for each process. The screen-
ing criteria were then applied to each process and the results were compared and
used to select the processes that appeared to be best suited for near-term indus-
trial boiler applications. The criteria used for this screening are listed in
Table 2.
As a result of this screening step, four FGD processes were selected as can-
didate systems for application to industrial boilers. These processes were:
• Lime/Limestone
Double Alkali
• Sodium Scrubbing
• Wellman-Lord
1141
-------
Each of these processes is commercially available and has been demonstrated to be
an effective emission control technique. The first three techniques are currently
being used to control SOa emissions from industrial boilers in the U.S. The
Wellman-Lord process has been used for several other types of applications in the
U.S., and on industrial boilers in Japan. A summary of the major characteristics
of these processes is presented in Table 3.
The criteria having the greatest influence on this selection was status of
development. The processes selected are better developed and have more acceptable
operating histories than the remaining processes. It was felt that in order to
provide support for new standards, FGD systems would have to be judged on their
proven performance and not on what might be possible. The Citrate/Phosphate,
Bergbau-Forschung/Foster Wheeler, Atomics International, and Shell/UOP processes
were not selected because of their relatively undeveloped status. The magnesium
oxide process, while operational on a full-scale unit, was not selected for evalua-
tion because it has not yet operated continuously for longer than eight days.
In addition to the systems selected for further study, the Chiyoda 121 and
the spray drying system have the potential for widespread application to both
industrial and utility boilers. However, the development status of these two
processes is such that there is insufficient data to permit the detailed analysis
that is required to support an environmental standard. Because of the current
interest in spray drying as an SO2 control alternative, though, an independent
analysis of that technology is being conducted as a continuation of this study.
ENERGY IMPACTS OF CANDIDATE CONTROL PROCESSES
Process energy requirements were evaluated as a function of process size, fuel
sulfur content, and level of S02 control. Results of these calculations, shown
below, indicate that the process energy penalties range from about 2 to 3% percent
of the gross heat input to the boiler for the three throwaway processes and from
about 3 to 8 percent for the Wellman-Lord process. The larger energy consumption
for the Wellman-Lord process is due to the steam and methane requirements for the
regeneration and S02 reduction portions of the process.
1142
-------
RANGE OF FGD PROCESS ENERGY REQUIREMENTS
Energy requirement
SO2 control method (Percent of net heat input to boiler)
Lime/Limestone 2.6 - 3.7
Double-Alkali 2.0 - 2.3
Sodium Scrubbing 2.0 - 2.5
Wellman-Lord 3.2 - 8.0
The variations in energy requirements for these processes are due to different
levels of sulfur in the coal, different levels of SOa control, and to a smaller
extent, plant size.
A summary of the relative percentage of the energy requirements of each pro-
cess area as compared to the overall energy requirement for the throwaway FGD pro-
cesses is presented in Table 4. This table shows that stack gas reheat is the
largest energy consumer for each of the throwaway processes. Methods of reducing
the amount of energy required for flue gas reheat are currently under investigation
by EPA. The primary method under consideration is bypassing a portion of the flue
gas around the scrubber to heat the exit gas. However, the applicability of this
method for installations requiring a high SO 2 control level is questionable.
The major energy consuming areas of the Wellman-Lord process vary depending
upon the sulfur content of the coal being burned as shown in Table 5. For the
eastern 3.5 percent sulfur coal case; which processes almost five times as much
S02 as the western 0.6 percent sulfur coal case, the regeneration and sulfur pro-
duction areas are the major energy users. For the low sulfur western coal case,
stack gas reheat is the major energy consumer. It is doubtful that the energy
requirements of the regeneration processing area can be significantly reduced since
double effect evaporators were assumed for these calculations. Double effect evap-
orators are some 45 percent more energy efficient than single effect evaporators.
As mentioned previously, methods of reducing stack gas reheat energy requirements
are currently under investigation by EPA.
1143
-------
ENVIRONMENTAL IMPACTS OF CANDIDATE CONTROL PROCESSES
The air, liquid, and solid waste impacts of the candidate processes were con-
sidered as functions of size, S02 removal level, and fuel sulfur content. With
regard to air pollution, each of the candidate FGD processes has the capability of
both particulate and S02 removal, but only S02 removal was considered. The impact
of all the candidate systems as far as SOa emissions is the same since each of the
processes can be designed to achieve the same degree of S02 control.
Liquid Waste Impacts
With regard to water pollution, only the sodium throwaway process should pro-
duce a significant environmental impact. The other three systems can produce waste-
water streams, but good design and operating practices can minimize any impact from
these streams.
The aqueous waste stream from a sodium throwaway system will contain about five
percent dissolved solids. In these systems, the absorbed SOa reacts to form Na2SOs
and Na2SOit which are removed from the system as dissolved solids in an aqueous waste.
Consequently, the amount of aqueous emissions is directly related to both the S02
control level and the coal's sulfur concentration. Discharge rates and average
stream compositions for the cases considered in this study are given in Table 6.
Several existing water treatment technologies are potentially applicable for
treatment of sodium throwaway FGD aqueous wastes. The other FGD processes should
require no water treatment if good design practices are used. Technologies that
may be used to reduce the level of total dissolved solids (TDS) in the sodium scrub-
bing system waste stream are: reverse osmosis, vapor compression distillation, and
multistage flash evaporation.
Although these technologies may be technologically applicable for treating
sodium throwaway process wastes, their overall compilexity and cost may prohibit
them from being used solely to treat aqueous wastes from industrial boiler FGD
systems. It is likely that due to the small size of the discharge streams (Table
6), the small industrial boiler operators will be able to discharge this stream
1144
-------
into existing treatment facilities. The existing treatment facilities, however,
will probably consist of the processes discussed above.
The major waste stream from a lime/limestone or double-alkali process is the
sludge, which can contain significant quantities of supernatant liquid. However,
good design and operating practices for the limestone and double-alkali processes
include dewatering the sludge and recycling the supernatant liquid. Consequently,
there should be essentially no water emissions from these systems except for times
of severe rainfall or process upsets.
The aqueous waste stream from the prescrubber of the Wellman-Lord process will
be characterized by a low pH which results from the chlorides that are removed from
the gas stream. However, except for the high chloride concentrations and low pH,
the quality of the prescrubber discharge will be very similar to that of the boiler
ash sluice water. Since this stream has been estimated to be approximately one
percent of the ash sluicing requirements for a power plant, it can be used for ash
sluicing where it will become diluted and neutralized with the other ash sluice
water. -Consequently, water emissions from the Wellman-Lord prescrubber stream should
be limited to intermittent discharges from the ash pond.
Solid Waste Impact
Of the four candidate processes only the limestone and double-alkali processes,
both of which produce a sludge waste stream, would result in a major solid waste
impact. A solids purge stream of Na2SOi» is produced in the Wellman-Lord process,
but the stream is relatively small and should not constitute a major solid waste
impact, especially for the size applications under consideration in this study.
Both the limestone and double-alkali sludges are composed primarily of calcium
sulfite and sulfate salts. Significant amounts of, fly ash may be present also,
depending on the method of particulate control in use. The sludges are relatively
inert and with proper site selection and proper disposal procedures, can be disposed
of in an environmentally acceptable manner. The disposal methods currently in use
are lined and unlined ponding and landfilling of treated and untreated materials.
Potential adverse impacts of sludge disposal lie in the areas of disposal acreage
requirements, water contamination through leaching and percolation of soluble
1145
-------
components of the solid waste into the groundwater system, and land use impacts
due to poor structural properties. Treatment techniques to minimize adverse
impacts may involve dewatering, addition of alkaline ash, and/or application of
commercial stabilization technology. These techniques are used to decrease the
sludge volume, decrease its permeability, and improve its structural properties.
As with the sodium throwaway system, the S02 absorbed from the flue gas by a
double-alkali or limestone system must leave the process in a waste stream, in
these cases as a waste sludge. Consequently, the amount of sludge produced is a
function of unit size, fuel sulfur content, and the SOa removal level. Table 7
presents the results of the limestone process material balance calculations and
shows the variation in sludge production with coal sulfur content and SOa removal.
The volume of sludge produced is also important as the sludge volume will
determine the size of the holding pond or landfill area. Figure 1 illustrates the
results of the sludge volume calculations graphically and shows the variation in
sludge production with coal sulfur content, boiler size, and level of removal.
Sludge volumes are presented in units of cc/sec, and acre-feet/15 years. The last
category, acre-feet/15 years gives an indication of the total volume of sludge to
be handled over the life of the plant assuming a 15 year life and an onstream
factor of 60 percent.
COST IMPACT OF CANDIDATE PROCESSES
Process costs were evaluated as a function of process size, fuel sulfur content,
and SO2 removal for the four candidate FGD processes. The general approach used in
developing the process costs consisted of four main steps. First, a series of
material and energy balance calculations were performed for each process to define
process stream rates and energy requirements as functions of unit size, SOa removal,
and the amount of sulfur in the coal. Second, each of the FGD processes was divided
into a number of process areas or modules, and economic scaling factors were developed
for each process area based upon the type of equipment used. Third, detailed engi-
neering design-studies were used to obtain costs for each of the process areas. All
of the FGD process costs were based on detailed design and economic estimates pre-
pared by TVA except for the wastewater processing area used by the sodium scrubbing
process. These costs were obtained from a report assessing wastewater control
1146
-------
technologies. Finally, the costs were individually scaled for each process area
using the stream rates calculated in the material balances and the economic scaling
factors developed for each process area.
Direct capital investment costs for the FGD processes ranged from a low of
$277,000 for a sodium throwaway process (8.8 MW, 75 percent removal, 0.6 percent
sulfur) to a high of $2,628,000 for a Wellman-Lord process (58.6 MW, 90 percent
removal, 3.5 percent sulfur coal). When indirect capital expenses are added, the
total capital investment costs for these two cases become $691,000 and $4,321,000
respectively. From a capital cost standpoint, for all cases considered the sodium
throwaway process appears to be the least costly process, and the Wellman-Lord the
most costly process.
With regard to annualized costs, the relative ranking of the various processes
change with variations in unit size, SOa removal, and coal sulfur content. For the
small unit size treating flue gas from a low sulfur coal (8.8 MW, 0.6% S) the
sodium throwaway process appears to have the lowest annualized costs. However,
for larger FGD system sizes treating flue gas from higher sulfur coals, the sodium
process annualized costs increase faster than do the costs of the other processes.
This is due to the relatively high cost of the sodium sorbent and costs associated
with wastewater treatment. In fact, for the largest size considered (58.6 MW,
3.5% S coal, 90% removal) the sodium process had the highest annualized costs
although it had the lowest capital investment costs. Figures 2 and 3 illustrate
the relative capital and annualized costs of these processes graphically.
The cost effectiveness of the various FGD processes was also determined as
part of this study. Cost effectiveness was defined as dollars per kilogram of
removed S02 ($/kg SOa) and was calculated "by dividing the annualized process costs
by the kilograms of S02 removed in a year assuming a 60 percent load factor. Results
of these calculations show that both coal sulfur content and process size signifi-
cantly affect the cost effectiveness of an FGD process. For a given size system,
cost effectiveness increases with an increasing coal sulfur content. For a fixed
coal sulfur content, cost effectiveness increases with increasing process size.
Consequently, the most cost effective systems are those designed for the 58.6 MW
(200xl06 Btu/hr) boiler burning 3.5 percent sulfur coal, and the least cost effec-
tive systems are those designed for the 8.8 MW (30xl06 Btu/hr). boiler burning 0.6
1147
-------
percent sulfur coal. Figure 4 illustrates these effects for the lime/limestone
processes. Curves developed for the other process showed similar effects.
Major findings of the cost analysis were:
1) An FGD system will add 10-50 percent to the uncontrolled boiler Total
Capital Investment, depending upon the FGD system, boiler size, fuel
sulfur content, and SOa control level. FGD annualized costs, including
capital charges (15 yrs FGD life), will add 15-50 percent to uncon-
trolled boiler costs depending upon the same variables listed above.
2) It is less cost effective to remove SOa from flue gas with low SOa
concentrations than from flue gas with high concentrations. Using an
8.8 MW sodium throwaway process as an example, it costs about 3% times
as much per unit of SOa removed ($4.58/kg vs. $1.33/kg) to remove 90
percent of the SOa from a 0.6 percent sulfur-based flue gas as from
a 3.5 percent sulfur based gas.
3) The sodium throwaway process is estimated to have the lowest capital
costs at all levels of operation. For small boiler sizes, and for low
sulfur coal operations, the annualized costs of the sodium throwaway
process are also lowest. However, for large high sulfur applications,
the sodium throwaway process becomes the most costly alternative.
This cost swing is due to costs associated with the sodium sorbent
and the wastewater treating facility used for this study. For the
large, high sulfur applications, either the lime/limestone or double-
alkali process appear most attractive.
1148
-------
TABLE 1. FGD SYSTEM SUMMARY
Process
Development status
Ho. of operational plants
Industrial Utility
Applicability to industrial boilers
Lime/Limes tone
Double Alkali
Commercial industrial and
utility applications.
Commercial industrial appli-
cations - Three utility
applications are planned.
28 Generally applicable. Possible
limitations due to solids disposal
land requirements.
Generally applicable. Possible
limitations due to solids disposal
land requirements.
VO
Wellman-Lord
Magnesium Oxide
Sodium Scrubbing
Spray Drying
Citrate/Phosphate
Commercial applications for
tail gas treating. A 115 MW
utility demonstration test
has been completed.
Commercial utility appli-
cations. No planned
industrial applications.
Commercial industrial and
utility applications.
Pilot-scale. A 410 MW utility
application is planned
1 MW pilot-scale. A 64 MW
industrial boiler applica-
tion is planned.
22
Generally applicable. Process costs
and complexity may limit applications
to only large boilers.
Process complexity will limit applica-
tions for industrial boilers. Has not
been operated continuously for longer
than eight days.
Generally applicable. Possible limitations
due to sorbent availability and cost, and
requirements for wastewater treatment.
SOz removal may be limited for lime
based high sulfur coal applications.
System is generally applicable except
for land requirements for solids dis-
posal. High reliability is claimed
but undemonstrated.
Applicability to small boilers will
be limited by overall complexity and
the need for a reducing gas to
produce HzS.
-------
TABLE 1. (Continued)
Process
Development status
Mo. of operational plants
Industrial Utility
Applicability to industrial boilers
Bergbau-Forschung/
Foster Wheeler
Atomics International/
Aqueous Carbonate
Process
Shell/UOP
Chiyoda 121
20 MW demonstration in U.S.
and a 45 MW demonstration
in Germany.
1.25 MW nonintegrated pilot
plant. A 100 MW utility
demonstration is planned.
0.6 MW pilot plant in U.S.
on coal-fired boiler.
40 MW in Japan on oil-
fired boiler.
Small-scale pilot plant.
A 20 MW utility demon-
stration is planned.
Applicability will be limited by
overall complexity and the require-
ment for extensive solids handling
equipment.
Applicability will be limited by
overall complexity for small boiler
applications. Use of unfamiliar
technology in the reducing reactor may
hinder process acceptability.
Applicability will be limited by
overall complexity and the require-
ment for hydrogen for regeneration.
Generally applicable. Possible
limitations due to solids disposal
land requirements in cases where
by-product gypsum marketability is
not feasible.
-------
TABLE 2. FLUE GAS DESULFURIZATION SCREENING CRITERIA
1. Status of Development
• Overall Process Development
• Availability of Data
2. Performance
SO2 Removal
• Reliability
• Response to Flue Gas Composition Changes
3. Applicability
• Simplicity
• Flexibility
• Controllability
• By-Product Marketability
4. Economic Considerations
Capital Investment Costs
• Operating Costs
5. Energy Considerations
• Liquid Pumping Requirements
• System Pressure Drop
• Regeneration Energy
• Requirement for Reducing Gas
6. Environmental Considerations
• Multipollutant Control
• Secondary Pollutant Emissions
1151
-------
TABLE 3. MAJOR CHARACTERISTICS OF CANDIDATE SYSTEMS
N>
Lime/Limestone
Double-Alkali
Sodium Scrubbing
Wellman-Lord
SO 2 Removal
Efficiency
(Percent)
70 to >90
>90
70 to >90
>90
Environmental
Impacts
Calcium based
sludge disposal
Calcium based
sludge disposal
High TDS waste-
water
Na2SOit solids
Relative
Energy
Impact s
Medium
Low-
Medium
Low-
Medium
Medium-
High
Reliability
Utility systems had
early problems, newer
systems have shown
improved reliability.
High reliabilities
reported, clear liquor
system.
High reliabilities
reported, clear liquor
system with few process
steps.
High reliability
during utility demon-
Development
Commercial
Commercial
Commercial
Commercial
stration test. Clear
liquor system although
it has many process
steps.
-------
TABLE 4. PERCENTAGE ENERGY CONSUMPTIONS FOR NONREGENERABLE PROCESSES
(58.6 MW Boiler, 90% Removal, Eastern Coal)
Source of
energy
consumption
Raw materials
handling and preparation
Pumps
Fans
Reheat steam
Disposal
Utilities and services
Total
Limestone Double Alkali Sodium TA
Percent Percent Percent
kW of total kW of total kW of total
114.2
317.4
471.7
884
38.2
8.6
1834.1
6 14.5 1
17 46.3 4
26 222.0 18
48 884 73
2 28.7 2
<1 9.3 <1
1204.8
14.5
40.7
222
884
160.4
9.3
1330.9
1
3
17
66
12
TABLE 5. PERCENTAGE ENERGY CONSUMPTION FOR WELLMAN-LORD PROCESS
(58.6 MW, 90% S02 Removal)
Source of
energy
consumption
Raw materials
handling and preparation
Pumps
Fans
Reheat steam
Process steam
Methane
Utilities and services
Total
kW
1.9
42.2
208.6
900
469
219
10.2
1851
Western coal
(0.6%S)
Percent of total
<1
2
11
49
25
12
<1
kW
9.3
82.4
205.0
884
2221
1048
9.9
4460
Eastern coal
(3.5ZS)
Percent of total
<1
2
5
20
50
23
<1
1153
-------
TABLE 6. WATER POLLUTION IMPACTS FOR THE SODIUM THROWAWAY SYSTEM*
Control
Boiler size and type level
8.8 kW (30 106 Btu/hr)
Underfeed Stoker
22 MW (75 106 Btu/hr)
Chaingrate Stoker
44 MW (150 106 Btu/hr)
Spreader Stoker
58.6 MW (200 10s Btu/hr)
Pulverized coal
Avg. Dissolved Solid Compositions
90
85
75
56
90
75
56
90
75
56
90
85
75
56
3.5%
S eastern coal
0.
6% S western coal
fc/m3 (gpro) fc/ra3
50.
47.
42.
31.
130.
106.
75.
262.
211.
146.
348.
315.
282.
200.
7
3
4
2
9
4
3
7
6
8
2
6
7
2
13
12
11
8
34
28
19
69
55
38
92
83
74
52
.4
.5
.2
.4
.6
.1
.9
.4
.9
.8
.0
.4
.7
.9
NaaSOs 70 percent
NazSOi, 18 percent
Na2C03 12 percent
Avg. TDS Concentration (wt. %)
5.
1
10
9
9
27
22
55
44
73
66
59
Na2SO3
NaaSOi,
Na2CO3
5
.2
.8
.1
-
.6
.3
—
.3
.7
—
.8
.6
.4
—
.2
(gP"0
2
2
2
7
5
14
11
19
17
15
52 percent
35 percent
13 percent
.7
.6
.4
—
.3
.9
—
.6
.8
—
.5
.6
.7
—
*Based on material balance calculations provided in Appendix A.
-------
TABLE 7. SOLID WASTE IMPACT FOR THE LIMESTONE FGD PROCESS
(Ash-Free Basis)
en
en
Boiler size and type
8.8 MW (30 10s Btu/yr)
Underfeed Stoker
22 MW (75 10s Btu/hr)
Chaingrate Stoker
44 MW (150 106 Btu/hr)
Spreader Stoker
58.6 MW (200 Btu/hr)
Pulverized Coal
Percent
removal
90
85
75
90
75
90
75
90
85
75
3.5% S eastern coal
g/s
128.4
121.2
106.9
323.7
271.6
652.9
543.5
866.9
823.0
724.2
(Ib/hr)
(1018)
( 961)
( 848)
(2567)
(2154)
(5177)
(4310)
(6874)
(6526)
(5743)
S,/min
5.0
4.7
4.2
12.9
10.6
25.3
21.2
33.7
32.2
28.4
(gal/min)
(1.33)
(1.25)
(1.10)
(3.4)
(2.8)
(6.7)
(5.6)
(8.9)
(8.5)
(7.5)
g/s
29.9
27.7
24.8
74.6
62.2
149.4
124.6
199.3
188.4
165.8
0.6% western coal
(Ib/hr)
( 237)
( 220)
( 197)
( 592)
( 493)
(1185)
( 988)
(1580)
(1494)
(1315)
fc/min
1.1
0.9
0.9
3.0
2.3
5.7
4.9
7.6
7.2
6.4
(gal/min)
(0.30)
(0.25)
(0.25)
(0.80)
(0.60)
(1.5)
(1.3)
(2.0)
(1.9)
(1.7)
-------
600(133) -
500(115) -
3
•a
s
a,
400(92) •
300(69) -
200(46) -
100(23) •
Eastern Coal
90 Percent
Eastern Coal
75 Percent Removal
Western Coal
75 Percent Removal
8.8
22
44
Boiler Size (MW)
58.6
Figure 1. Sludge production rates for the limestone FGD process.
1156
-------
4000
(0
M
at
o
•a
en
o
o
•H
a.
3000
2000
1000
Wellman-
ord
imestone
uble
Alkali
odium
Throwaway
14.6 29.2 43.8 58.6
(50) (100) (150) (200)
Size in MW (106 Btu/hr)
Figure 2. Capital costs versus unit size.
(3.5% S coal, 90% removal)
1157
-------
1400
w
a 1150
o
•o
900
N
•H
H
ctf
650
400-
Sodium
Throwaway
Wellman-
Lord
Double Alkali
Limestone
14.6 29.3 43.8
(50) (100) (150)
Size in MW (10s Btu/hr)
58.6
(200)
Figure 3. FGD annualized costs versus unit size.
(3.5% S coal, 90% removal)
1158
-------
5 •
T3
tU
(A
N 4
ca
CO
QJ *5
3 3
u
0)
10.6% Sulfur Coal
>3.5% Sulfur Coal
14.6 2973 44
(50) (100) (150)
Size in MW (10s Btu/hr)
(200)
Figure 4. Limestone process cost effectiveness.
1159
-------
UNPRESENTED PAPERS
1160
-------
STACK GAS REHEAT - ENERGY AND ENVIRONMENTAL ASPECTS
Charles A. Muela
William R. Menzies
Radian Corporation
Austin, Texas
Theodore G. Brna
Industrial Environmental Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina
ABSTRACT
Radian Corporation has completed a technical and economic assess-
ment of stack gas reheat for the Environmental Protection Agency (EPA).
The objectives of this study were to: 1) identify and analyze the
potential problems associated with the generation of saturated flue
gases by wet flue gas desulfurization processes, 2) determine current
flue gas reheating practices in the utility industry, 3) determine how
much reheat is actually required as well as its cost, and 4) compare the
costs of the various techniques that can be used to reheat flue gases.
Background data for this study were obtained from literature sources,
from responses to questionnaires that were distributed to various users
and suppliers of reheat equipment, and from visits to several facili-
ties. In this paper, the potential benefits and corresponding energy
requirements associated with the use of stack gas reheat are discussed.
Flue gas desulfurization (FGD) processes that cool and saturate
flue gases with water may cause: 1) corrosion of downstream equipment,
2) a visible plume, 3) acid rainout in the vicinity of a plant's stack,
and 4) increased ground-level concentrations of pollutants. Electric
utilities, which are currently the primary users of FGD scrubbers, have
cited equipment protection as the principal reason for the use of stack
gas reheat. According to these users, the flue gas must normally be
reheated by at least 16.7°C (30°F) to adequately protect equipment
downstream of a wet FGD process.
It is extremely difficult to quantify the amount of reheat actually
required in any given application because each of the four problems just
cited is resolved by different levels of reheat. In general, however,
the results of the analyses discussed in this paper indicate that reheat-
ing stack gases is an effective method of preventing the corrosion of
downstream equipment, but it is less effective in significantly reducing
the other potential problems.
1161
-------
INTRODUCTION
An EPA-sponsored study to assess the current state-of-the-art for
stack gas reheat has been completed. Specific objectives of this study
were to:
1. Identify and analyze the potential problems resulting from the
generation of saturated flue gases by wet flue gas desulfurization
processes.
2. Determine current practices in the utility industry regarding
stack gas reheat.
3. Determine how much reheat is required to significantly reduce
the potential problems caused by saturated flue gas.
4. Compare the costs of various reheat techniques.
The complete results of the analyses conducted to achieve these objectives
are discussed in a draft report that is currently being reviewed by the
EPA. However, the scope of this paper is limited to an analysis of
potential reheat-related problems and a determination of the energy
requirements associated with significantly reducing those problems.
Background information for the analyses presented in this paper and in
the comprehensive study was obtained from the literature, questionnaires
that were sent to companies in the utility industry and to vendors of
reheat equipment, and visits to several power plants.
SO emissions from utility plant boilers have been commonly controlled
by flue gas desulfurization (FGD) processes. One type of FGD process,
wet scrubbing, adds moisture to the flue gas and cools it to its adiabatic
saturation temperature. A saturated flue gas may cause the following
problems: 1) corrosion of equipment downstream of the FGD process, 2) a
visible plume when the gas exits the stack, 3) acid rainout in the
vicinity of the stack, and 4) increased ground-level concentrations of
pollutants downwind from the stack.
1162
-------
The impacts of these problems can be reduced or eliminated by
heating the flue gas above its saturation temperature. This can be
achieved by several reheat configurations. The most commonly used are:
1. Inline Reheat (Figure la), in which the flue gas is heated by
being passed through a heat exchanger downstream of the scrubber.
2. Indirect Hot Air Reheat (Figure Ib), in which heated air is
mixed with the scrubbed flue gas.
3. Direct Combustion Reheat (Figure Ic), in which hot combustion
gases generated by firing fuel oil or natural gas are mixed
with the scrubbed flue gas.
4. Bypass Reheat (Figure Id), in which a portion of the boiler
flue gas is routed around the scrubber and mixed with the
scrubbed flue gas.
The viability of bypass reheat is highly dependent on S0« standards.
Because current and proposed SO standards may restrict the use of
bypass reheat in many applications, this reheat method is not considered
in this paper. It should be noted, however, that bypass reheat has been
used effectively at several facilities.
CURRENT PRACTICES IN UTILITY INDUSTRY
FGD process users and vendors, reheat users, and architect/engineer
(A/E) firms were surveyed in order to identify current reheat practices
in the utility industry as well as to determine the reliability of
various reheat configurations. A profile of the different types of
reheat configurations that have been or will be used in the utility
industry was developed from data gathered in the survey and is presented
in Table 1. These data indicate that inline reheat is the configuration
utilities most prefer. The A/E's responding to the survey recommended
indirect hot air be used when reheat is needed. However, most responding
A/E's and vendors indicated that reheat is not necessary and, generally,
do not recommend the use of reheat.
1163
-------
a) Inline Reheat
Flue Gas
Scrubber
To Stack
Reheat Exchanger
Steam or
Hot Water
b) Indirect Hot Air Reheat
Scrubber
Flue Gas
To Stack
Ambient
Air
Air Heater.
Auxiliary
Air Fan
Steam or
Hot Water
c) Direct Combustion Reheat
Scrubber
Flue Gas
To Stack
Fuel Oil/Natural Gas
Ambient Air
d) Bypass Reheat
Flue Gas
Combustion
Chamber
T
Auxiliary Air Fan
* v
To Stack
Scrubber
Figure 1. Simplified schematics of various reheat configurations.
1164
-------
TABLE 1. PROFILE OF VARIOUS REHEAT CONFIGURATIONS
USED WITH EXISTING OR PROPOSED UTILITY FGD SYSTEMS
Startup
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
Planned
TOTAL '
Indirect
Inline Hot Air
1
2
4
4
3
1 1
2 1
3 4
3 2
4 1
3 2
2
1
1
31 14
Direct
Combustion Bypass
2
1
1
3
1 3
7
1 8
3 7
1
13 25
Wet Stack1
1
2
1
3
4
4
3
1
1
20
Wet stack—no reheat. Saturated flue gas is discharged directly to
the atmosphere.
1165
-------
Although the need for reheat is site specific, utility companies
responding to the survey indicated that reheat is generally used to
prevent condensation and the corrosion of equipment. The degree of
reheat used by these utilities varies from 0°C (0°F), corresponding to
wet stack operation, to more than 55.6°C (100°F). A typical utility
could be expected to reheat the flue gas by 22-33°C (40-60°F). The
heating requirements corresponding to 27.8°C (50°F) of reheat in a 500-
MW power plant were calculated for the inline, indirect hot air, and
direct combustion reheat configurations. The results of these calculations
are presented in Table 2. As the results in Table 2 indicate, the
energy requirements associated with achieving this degree of reheat can
be substantial, ranging from 1.6 to 2.8 percent of the total boiler heat
input.
POTENTIAL BENEFITS AND ENERGY REQUIREMENTS ASSOCIATED WITH STACK GAS
REHEAT APPLICATIONS
Each of the four problems that may be caused by saturated flue
gases can be resolved with a different level of reheat. The impacts of
stack gas reheat on these prablems are examined below in terms of the:
"benefits resulting from various levels of reheat, and
"energy required to achieve these levels of reheat
with several different reheat configurations.
The Use of Reheat for Equipment Protection
Scrubbed flue gases normally contain S0?, S0_, C0_, chlorides, and
sulfuric acid mist. Consequently, any water droplets in the flue gas
will absorb these gases and become highly acidic. These acidic droplets
can subsequently cause severe corrosion of equipment downstream of a wet
FGD process. Moisture will be present when scrubbing liquor is entrained
in the flue gas or when the flue gas temperature falls below its adiabatic
saturation temperature and condensate forms. This temperature drop can
1166
-------
TABLE 2. HEAT INPUT REQUIRED BY VARIOUS REHEAT CONFIGURATIONS TO RAISE FLUE GAS TEMPERATURE 28°C (50°F)d
Inline
Direct
Combustion
Indirect
Hot Air
Teaperature 0 Scrubber Exit, "C (°F)
Temperature 8 Stack Exit, °C (°F)
Reheat Air Characteristics
Ambient Temperature, °C (°F)
Relative Humidity, Z
Heated Air Temperature, "C (°F)
Flow Rate, 106 kg/hr (10& Ib/hr)
Entrained Liquid*, 10~3 kg/std m3(gr/scf)
Heat Required for Vaporizing Entralnment
Mi (106 Btu/hr)
Heat Required to Hake Up Heat Losses
MW (106 Btu/hr)
Sensible Heat Required for Reheat Medium0
MW (106 Btu/hr)
Sensible Heat Required to Raise Flue Gas
Temperature 28°C (SO°F), MU (10& Btu/hr)
Total Heat Required, MW (106 Btu/hr)
(Z of boiler input)
53.9 (129)
81.7 (179)
53.9 (129)
81.7 (179)
15.6
53.9 (129)
81.7 (179)
0.027 (0.012) 1.842 (0.805)
0.038 (0.13) 2.64 (9.00)
1-.93 (6.60) 1.93 (6.60)
19.6 (66.8) 19.6 (66.8)
21.5 (73.5) 24.2 (82.4)
1.63 1.83
0.027 (0.012) 1.842 (0.805)
0.038 (0.13) 2.64 (9.00)
1.93 (6.60) 1.93 (6.60)
0.056 (0.19) 0.13 (0.44)
19.6 (66.8) 19.6 (66.8)
21.6 (73.7)
1.64
24.3 (82.8)
1.84
(60) 15.6 (60)
50 50
204 (400) 204 (400)
0.60 (1.33) 0.68 (1.49)
0.027 (0.012) 1.842 (0.805)
0.038 (0.13) 2.64 (9.00)
1.93 (6.60) 1.93 (6.60)
11.6 (39.6) 13.0 (44.4)
19.6 (66.8) 19.6 (66.8)
33.1 (113.1) 37.2 (126.8)
2.51 2.82
*From the literature survey, the typical entrainment value for bottom wash
mist eliminators Has 0.012 gr/scf; for top wash mist eliminators,
0.805 gr/scf. (Ref. 2)
bAssumed heat losses correspond to a 2.8°C (5°F) temperature drop which
utilities have indicated Is reasonable for duct work and stack.
"This sensible heat equals the heat required to raise the reheat medium
from its ambient temperature to the stack exit temperature.
ases: 1) 500-MW Unit
2) 9503 kJ/kWh (9,000 Btu/kWh) Heat Rate
3) Forced Draft Fan Configuration
4) Flue Gas Saturation Temperature - 53.9°C (129°F)
5) Flue Gas Flow Rate - 2.33xl06 kg/hr (5.14x10° Ib/hr)
6) Flue Gas Heat Capacity - 1.09 U/kg-C" (0.26 Btu/lb-°F)
7) Direct combustion natural gas usage and an ambient
air temperature of 15.6°C (60°F)
-------
result from heat losses through the walls of the duct work and stack. A
flue gas will typically cool about 2.8-5.6°C (5-10°F) in the duct work
and stack following the scrubber.
Complete protection from corrosion of equipment downstream of a
scrubber would involve the prevention of moisture and vaporization of
sulfuric acid mist. The quantity of heat required to prevent the condensation
is dependent on the total heat loss experienced by the flue gas, the
amount of entrained scrubber liquor, and the type of reheat configuration
used. The theoretical heat requirements of various reheat configurations
to prevent the occurrence of moisture downstream of the scrubber were
calculated and are compared in Table 3. In these calculations, only the
forced draft fan arrangement of the various configurations was considered;
consequently, no credit was taken for the work on compression by the fan
(a fan downstream of a scrubber will raise the temperature of the flue
gas). The data in Table 3 show that, although the indirect hot air
configuration reduces the flue gas dew point, it requires about 23
percent more heat than inline reheat to eliminate moisture downstream of
the scrubber. This increased heat requirement is a result of the energy
required to heat ambient air to the flue gas dew point. The data in
Table 3 also show that entrainment has a significant impact on the
reheat requirement. Revaporization of sulfuric acid mist requires a
considerably greater heat input than is needed to prevent condensation
because the flue gas must be heated to the sulfuric acid dew-point
temperature. In most systems, this temperature will be in the range of
93-149°C (200-300°F).
A comparison of Tables 2 and 3 indicates that the heat input
required to raise the flue gas temperature 27.8°C (50°F) is significantly
greater than the heat input needed to prevent moisture downstream of the
scrubber. Consequently, the level of reheat generally used by the
utility industry seems to be more than adequate to effectively protect
downstream equipment from corrosion caused by components other than
sulfuric acid.
1168
-------
TABLE 3. HEAT INPUT REQUIRED TO PREVENT MOISTURE DOWNSTREAM FROM SCRUBBER0
VO
Inline
Entrained Liquid,
10-3 kg/std m3 (gr/scf) 0.027 (0.012) 1.842 (0.805)
kg/hr (Ib/hr) 59.0 (130) 4010 (8840)
Dew Point Temperature at
Stack Exit, °C (°F) 53.9 (129) 53.9 (129)
Ambient Air Characteristics
Temperature, °C (°F)
Relative Humidity, % -
Heated Air Temperature,
"C (°F) - -
Flow Rate, 10s kg/hr
(10* Ib/hr)
Assumed Heat Loss3, HW
(106 Btu/hr) 1.93 (6.60) 1.93 (6.60)
Heat Required for Vapor-
izing Entrainment
MW (106 Btu/hr) 0.038 (0.13) 2.64 (9.00)
Sensible Heat Required for
Reheat Medium
MW (106 Btu/hr)
Theoretical Heat Required
to Prevent
Moisture, HU
(106 Btu/hr) 2.0 (6.7) 4.57 (15.6)
(% of boiler input) 0.15 0.35
Reheat Configurations
Direct Combustion
0.027 (0.012) 1.842 (0.805)
59.0 (130) 4010 (8840)
53.9 (129) 53.9 (129)
-
1.93 (6.60) 1.93 (6.60)
0.038 (0.13) 2.64 (9.00)
0.003 (0.01) 0.006 (0.02)
2.0 (6.7) 4.57 (15.6)
0.15 0.35
Indirect Hot Air
0.027 (0.012) 1.842
59.0 (130) 4010
53.7 (128.6) 53.1
15.6 (.60) 15.6
50
204 (400) 204
4.10 (9.25) 4.68
1.93 (6.60) 1.93
0.038 (0.13) 2.64
C.47 (1.6) 1.1
2.4 (8.3) 5.63
0.18
(0.805)
(8840)
(127.5)
(60)
50
(400)
(21.25)
(6.60)
(9.00)
(3.6)
(19.2)
0.43
Total heat losses are assumed to be equivalent for all configurations
and correspond to a 2.8°C (5°F) drop in flue gas temperature.
This sensible heat equals the heat required to raise the reheat
medium from its ambient temperature to the stack exit temperature.
°Bases: 1) 500-MW Unit
2) 9503 kJ/kHh (9,000 Btu/kwh) Heat Rate
3) Forced Draft Fan Configuration
4) Flue Gas Saturation Temperature - 53,9°C (129°F)
5) Flue Gas Flow Rate = 2.33x106 kg/hr (5.14x106 Ib/hr)
6) Flue Gas Heat Capacity - 1.09 kJ/kg-C" (0.26 Btu/lb-°F)
7) For direct combustion, natural gas ambient temperature
is assumed to be 15.6°C (60°F)
-------
Suppression of Visible Plume
The use of a wet scrubber can result in a visible plume. While a
visible plume does not have a direct negative impact on the environment,
it can be aesthetically displeasing, potentially hazardous to ground and
air traffic, and in violation of air pollution control ordinances restricting
opacity. The mechanics of visible plume formation are illustrated in
Figure 2. Ambient air conditions (temperature, relative humidity) are
represented by point 1 in this figure. Point 2 corresponds to the
conditions of a hot flue gas as it exits the boiler. The flue gas is
saturated and cooled during the scrubbing operation and is represented
by point 3. As a saturated flue gas leaves the stack, it mixes with
ambient air according to line 3-1. A visible plume is formed when the
ambient-air-saturated flue gas mixture intersects the saturation curve
3 4
and crosses into the fogged field area of the chart. ' The persistence
of the visible plume increases with increasing length of the portion of
line 1-3 in the fogged field.
SATURATION
CURVE AT BARO-
METRIC PRESSURE*!
f COMBUSTION GAS
(LEAVING SCRUBBER
AMBIENT AIR
ENTERING
FURNACE
"lNDIRECT\
COMBUSTION GAS
LEAVING FURNACE-J
DRY BULB TEMPERATURE
Figure 2. Psychrometric chart showing state points of flue-gas/
air mixture with scrubbing and reheat. (Ref. 4,5)
1170
-------
Prevention of a visible plume with reheat involves the clockwise
rotation of the saturated flue-gas/ambient-air mixing line (line 3-1,
Figure 2) until it is tangent (line 1-5) to the saturation curve. The
temperature to which the flue gas must be heated in order to prevent a
visible plume is represented by point 5 when inline reheat is used and
by point 4 when indirect hot air reheat is used. Note that the degree
of reheat for inline reheating (the difference in dry bulb temperatures
at points 5 and 3) is greater than for indirect hot air reheat (dry bulb
temperature at point 4 less that at point 3) shown in Figure 2. Due to
the shape of the saturation curve, the temperature of the heated flue
gas and the corresponding reheat requirements are dependent on the
ambient air's temperature and relative humidity. The reheat temperature
(temperature of the scrubbed flue gas after heating) and the theoretical
heat input requirements of the inline and indirect hot air configurations
at various ambient air conditions are shown in Table 4. (The flue gas
saturation temperature in Table 4 corresponds to point 3, while point 6
of Figure 2 represents the dry bulb temperature of hot air after indirect
reheat.) The data in Table 4 show that the required reheat temperature
(point 5 or 4) and heat input are highly dependent on the ambient air
condition (point 1). These data also show that indirect hot air reheat,
via dilution of the flue gas, significantly decreases the reheat temperature
required by inline reheat. Figure 2 also illustrates this: the tem-
perature at point 4 is less than at point 5. However, the addition of
air increases the total mass of gas which must be heated. As a result,
the theoretical heating requirements for the two reheat configurations
are approximately equal.
When a forced draft, indirect hot air configuration is used (Fig-
ure Ib), the work of compression from the auxiliary air fan ultimately
helps to raise the temperature of the flue gas; this reduces the quan-
tity of heat that must be added to the flue gas to achieve the required
reheat temperature. These reduced heat inputs are shown in Table 4. As
the data in Table 4 indicate, the prevention of a visible plume could be
very costly in terms of the required heat input.
1171
-------
TABLE 4. ENERGY REQUIREMENTS ASSOCIATED WITH PREVENTION OF A VISIBLE PLUME3
-J
to
Reheat Configuration Inline
Ambient Air Temperature, °C (°F) 15.6 (60) 15.6 (60) 0 (32)
Ambient Air Relative humidity, Z 50 100 100
Flue Gas Saturation Temperature,
°C (°F) 53.9 (129) 53.9 (129) 53.9 (129)
Heated Air Temperature, °C (°F)
Quantity of Heated Air Required,
106 kg/hr (106 Ib/hr) -
Flue Gas Keheat Temperature Required
to Prevent Visible Plume, °C (°F) 83.9 (183) 115.6 (240) 226.1 (439)
Reheat Required to Prevent Visible
Plume Formation, MW (106 BCu/hr) 20.8 (71.0) 43.7 (149.0) 121.9 (416.0)
(% Boiler Input) 1.58 3.31 9.24
Indirect Hot Air
15.6 (60)
50
53.9 (129)
204 (400)
0.400 (0.881)
74.4 (166)
20.8 (71.0)
20. 6b (70.3)
1.58
1.56
15.6 (60)
100
53.9 (129)
204 (400)
0.798 (1.76)
91.1 (196)
43.7 (149.0)
43.2 (147.7)
3.31
3.28
0 (32)
100
53.9 (129)
204 (400)
2.03 (4.47)
122.8 (253)
121.9 (416.0)
121.0 (412.7)
9.24
9.17
Bases and Comments: 1) Flue gas is representative of a 500-MW plant.
2) Heat losses in duct work and stack are assumed to be negligible.
3) Entrainment is assumed to be zero.
4) Forced draft fan arrangement.
Underlined reheat requirements were developed for indirect hot air by taking credit for work of compression produced by the
auxiliary fan (see Figure Ib). The pressure drop was assumed to be 6 in. Hy) and an 85 percent fan efficiency was also assumed.
-------
If the flue gas saturation temperature in Table 4 is raised by
27.8°C (50°F), the resulting reheat and exit temperature would be 81.7°C
(179°F). Comparing this temperature with those flue gas temperatures
that are required to prevent a visible plume indicates that the amount
of reheat typically used by industry will not prevent visible plumes at
many meteorological conditions.
Using Reheat to Prevent Acid Rain
When a wet FGD process is used, SCL, SO,, and water in the flue gas
can react to form H^SCL and H^SO,. Because the dew point of sulfuric
acid is normally higher than the adiabatic saturation temperature of the
flue gas, the H SO, vapor which is formed can condense even when the
flue gas is above its saturation temperature. The resulting acid mist
may, in turn, provide the nuclei for additional condensation, and rainout
may occur when the flue gas leaves the stack. Acid rain may also be
formed when residual SO- is oxidized and absorbed by moisture on the
stack wall and the resulting H S0_ and or H SO, droplets are entrained
in the flue gas leaving the stack.
Reheat can suppress acid rain by:
°preventing condensation of water vapor in the system,
"vaporizing any entrained liquid leaving the mist
eliminator, and/or
"revaporizing sulfuric acid mist that is present in
the system.
As mentioned previously, preventing the condensation of water vapor and
eliminating any entrainment that may be present require the input of
only enough heat to keep the scrubbed flue gas above its dew point
(about 52-60°C (125-140°F) in most systems), while revaporizing any
sulfuric acid that is present requires the flue gas to be heated to
1173
-------
above the sulfuric acid dew point (approximately 93-l49°C (200-300°F)
in most systems, although this temperature is dependent on HO and SO.,
concentrations in the flue gas). Note that, even though all the sulfuric
acid in the system can be revaporized with a substantial heat input, the
condensation and subsequent rainout of sulfuric and/or sulfurous acids
could still occur when the scrubbed flue gas mixes with cooler ambient
air.
The impact of reheat on the potential for rainout from a wet plume
was simulated to determine the impact that reheat could have on the
water concentration in a plume and the time period over which this
O f
concentration was between 0.5 and 1 g/m . Blum suggested this as the
critical concentration range that would result in rainout from a plume.
This analysis indicated that reheat can have a significant impact on the
potential for rainout from a plume at conditions of mild temperatures
and low relative humidities. However, since the mechanism causing
rainout is not fully understood, this conclusion must be considered to
be somewhat speculative.
Impacts of Scrubbing and Reheating on Ground-Level Pollutant Concentration
Although wet FGD processes remove SO from the flue gas, they will
increase the ground-level concentrations of other pollutants (such as
NO ) by reducing plume buoyancy and rise. Reheating the flue gas increases
X
buoyancy of the gas and thus reduces the ground-level pollutant concentrations
exhibited by the scrubbed flue gas. The effect of scrubbing and reheating
a flue gas on the short term (3-hour) SO- and NO ground-level concentrations
was analyzed. Two atmospheric stabilities, unstable and neutral, and
three levels of reheat were considered in the analysis. An unstable
atmosphere, although occurring infrequently, is known to produce high
ground-level pollutant concentration. A neutral atmosphere does not
produce as high a ground-level concentration as the unstable atmosphere,
but it occurs more frequently.
1174
-------
The results of the unstable atmosphere analysis are presented for
SO and NO in Figure 3, which shows that:
£* X
°The highest S0« and lowest NO ground-level concentrations
are exhibited by the unscrubbed flue gas. These
results reflect the ability of the scrubbing process
to remove SQn but not NO .
2 x
"Scrubbing the flue gas reduces the maximum unscrubbed
ground-level S02 concentration by approximately 48
percent. However, scrubbing increases the maximum
unscrubbed ground-level NO concentration by about
160 percent. It should be noted that for both SO-
and NO the predicted concentrations are below
applicable ambient air quality standards.
"Reheating the scrubbed flue gas 27.8°C (50°F) reduces
the maximum ground-level SO- and NO concentrations
by about 33 percent compared with scrubbing with no
reheat.
°The addition of 55.6°C (100°F) of reheat to the
scrubbed flue gas reduces the maximum ground-level
SO^ and.NO concentrations by 47 percent over a
scrubbed but unreheated flue gas.
The use of reheat under neutral atmospheric stability conditions produces
similar trends in S09 and NO concentrations as those predicted for an.
£• X
unstable atmospheric condition. From this analysis, it is concluded
that:
"Reheating a scrubbed flue gas significantly reduces
the ground-level pollutant concentrations attributed
to the unreheated scrubbed gas.
°A substantial degree of reheat (greater than 55.6°C
(100°F)) is required to reduce the ground-level
concentrations of NO to its original (unscrubbed)
concentration.
1175
-------
0.5
SYMBOL REHEAT LKVEL
900
. 800"
700
600-
400
UNSCKUBliED
SCRUBBED, NO REHEAT
SCRUBBED, 27.8°C (50°F)
REHEAT
SCRUBBED, 55.6°C (100°F)
REHEAT
SYMBOL
(miles)
(kilometers)
REHEAT LEVEL
UNSCRUBBED
SCRUBBED, NO REHEAT
SCRUBBED, 27.8°C (50°F)
REHEAT
SCRUBBED, 55.6"C (100°F)
REHEAT
(miles)
1 (kilometers)
DOWNWIND DISTANCE FROM STACK
DOWNWIND DISTANCE FROM STACK
Bases: 1) 3% sulfur coal
2) Ambient air temperature, 15.6°C (60°F)
3) 80% SO removal
4) Vertical temperature gradient, -19.5°C/km
(-10.7°F/1000 ft)
5) Flue gas exit temperature and velocity:
Temperature
°C (°F)
Exit Velocity
m/s (ft/sec)
Unscrubbed
Scrubbed, No Reheat
Scrubbed, 27.8°C (50°F) Reheat
Scrubbed, 55.6°C (100°F) Reheat
149 (300) 10.7 (35)
53.9 (129) 8.75 (28.7)
81.7 (179) 9.50 (31.1)
109 (229) 10.2 (33.5)
Figure 3. Three-hour, ground-level SO and NO concentrations downwind of stack
(unstable atmosphere, wind speed = 2.24 m/s (5 raph)).
-------
The impact of the inline and indirect reheat configurations on
ground-level pollutant concentrations was compared. This comparison
showed that for the same heat input the two reheat configurations produced
similar maximum ground-level pollutant concentrations. When the scrubbed
gas is heated to the same temperature with the two configurations, the
indirect hot air configuration produces the lower maximum ground-level
concentration.
SUMMARY
The major points of the information presented in this paper are:
"Currently, the utility industry utilizes stack gas
reheat primarily to protect equipment from corrosion.
The degree of reheat used varies from 0-55.6°C (0-
100°F), with the average being about 27.8°C (50°F).
The heat input required to reheat the scrubbed flue
gas from a 500-MW power plant through 27.8°C (50°F)
can vary from approximately 1.6 percent of the
boiler input for the inline reheat to about 2.8
percent for the indirect hot air configuration.
"Moisture in the flue gas enhances the corrosion of
equipment downstream of the scrubber. For the cases
and reheat configurations analyzed in this paper,
the heat input required to prevent moisture downstream
of the scrubber varies from 0.2 to 0.4 percent of
the boiler input. The inline reheat configuration
requires the least heat input in preventing moisture
downstream of the scrubber.
"The heat required to prevent a visible pluma is
about the same regardless of the reheat configuration
used. Prevention of a visible plume is very dependent
on meteorological conditions, and reheat cannot
prevent a visible plume with a reasonable degree of
reheat at many ambient air conditions.
1177
-------
°At present, knowledge about the mechanisms of acid
rainout is limited. Consequently, the impact of
reheat on acid rainout cannot be fully evaluated.
However, reheat can certainly lessen the potential
for acid rain by preventing moisture downstream of
the scrubber.
"Stack gas reheat can significantly reduce the ground-
level pollutant concentrations from a scrubbed flue
gas. Although the lowest pollutant concentrations
are produced by high levels of reheat, an incremental
increase in the reheat level does not produce a
proportional decrease in pollutant concentration.
REFERENCES
1. Menzies, William R. and Charles A. Muela. Stack Gas Reheat Evaluation.
Draft Report prepared by the Radian Corporation under U.S. Environ-
mental Protection Agency (IERL-RTP) contract number 68-02-2642.
2. Choi, P.S.K., S.G. Bloom, H.S. Rosenberg, and S.T. DiNovo. Stack
Gas Reheat for Wet Flue Gas Desulfurization Systems. Final Report,
EPRI-FP-361, Battelle Columbus Laboratories, Columbus, OH, February
1977.
3. Rohr, F.W. Suppression of the Steam Plume From Incinerator Stacks.
In: National Incinerator Conference, New York, May 1968, Proceedings.
New York, ASME, 1968, pp. 216-24.
4. Rohr, Fred W. Suppressing Scrubber Steam Plume. Pollution Eng. 1
(1), 1969, pp. 20-22.
5. Leivo, C.C. Flue Gas Desulfurization Systems: .Design and Operating
Considerations, Volume II. Technical Report, EPA-600/7-78-030b
(NTIS PB-28025), Bechtel, San Francisco, CA, March 1978.
6. Blum, A. Drizzle Precipitation from Water Cooling Tower. Engineer,
August 6, 1948.
1178
-------
Minimizing Operating Costs
of Lime/Limestone FGD Systems
Carlton Johnson
Manager of Process Engineering
Peabody Process Systems
Flue gas desulfurization systems all too fre-
quently are put in the category of black magic.
Statements are made that sophisticated con-
trols are required and very specific conditions
set to make SO2 scrubbing systems perform as
required. Nothing could be further from the
truth. It is Peabody's experience gained in de-
sign, construction and operating results over a
period of years that such exact considerations
are not required and that the system, if properly
designed originally, can operate under a wide
variety of conditions. We must emphasize "if
properly designed originally".
Selection of Design Criteria
When selecting the design criteria for its FGD
System, a utility must allow for a wide variety of
conditions. Consideration must be given to the
maximum percent sulfur in the coal that is to be
burned. Allowance must be made for the in-
creased leakage of air into the system as the
power plant ages. Contingency may also be al-
lowed for the fact that it may be possible to de-
bottleneck the generating system beyond the
rated capacity. The combination of the above
factors will result in design conditions for the
FGD System in excess of what would be re-
quired to meet normal operating conditions. In
fact, such design conditions may never be
reached during the life of the generating unit,
but are required to insure meeting emission
standards under all conditions. The margin be-
tween the design and normal operating condi-
tions will ensure that environmental regulations
can be met; however if a higher operating cost
is then necessary, while it is necessary to meet
the environmental regulations, to exceed them
provides no benefit to the utility.
Make Up Water V
Hydroclone
What Comprises Operating Costs?
When establishing design criteria, the owner
and engineer must be aware of the major factors
which contribute to operating costs of the FGD
system. These are:
1. Alkali Consumption
a) Quantity of SO2 removed
b) Stoichiometry
2. Sludge
a) Quantity of SO2 removed
b) Alkali Stoichiometry
3. Power
a) L/G ratio (GPM/1000 ACFM)
b) Gas System Pressure Drop
Alkali consumption and waste sludge produc-
tion comprises the largest part of the operating
cost. Each of these costs is dependent upon the
amount of S02 removed and the Stoichiometry
or efficiency of alkali utilization at which the SO2
is removed. A poor utilization of alkali means
that not only is the alkali consumption higher
but also that unreacted alkali is contained in the
waste sludge and it is disposed of with an in-
creased quantity of waste sludge.
Figure 1
1179
-------
A smaller but still significant cost factor is the
power consumption which is a function of both
the liquid to gas ratio required to remove SO2
from the flue gas as well as the pressure re-
quired to make the gas flow through the system.
The design of the FGD system should provide
for minimizing these costs under actual operat-
ing conditions. Prior to discussion of how these
costs can be minimized it would be appropriate
to consider the details of the Peabody FGD
System shown in Fig. 1.
System Description
The Peabody FGD System is based upon the
use of a patented, high-velocity spray tower as
an absorber. This system offers the advantages
of an open gas flow path which provides for high
reliability because surfaces which scale or plug
are minimized, low gas pressure drop which re-
duces system power requirements and adapta-
bility to a wide variety of load conditions.
Fig. 1 shows the basic components of the Pea-
body FGD System. Flue gas enters the bottom
of the absorber and is contacted countercur-
rently with a slurry containing a mixture of cal-
cium sulfite, calcium sulfate and unreacted
alkali. The slurry enters the tower via multiple
spray headers. The number of spray headers is
a function of both design and normal operating
conditions. After leaving the absorption zone in
the tower, where the gas has been contacted
with the slurry, the scrubbed gas contains a
significant amount of entrained slurry. The gas
then flows upward through an interface tray and
mist eliminator for removal of the entrained
slurry prior to being discharged to the stack.
The slurry leaving the absorber flows by gravity
to a slurry recycle tank where the reaction be-
tween SOj and the alkali goes to completion.
The bulk of the slurry is then recycled to the
absorber and excess waste slurry leaves the
system as overflow from the slurry recycle tank.
A critical area in the design of the absorber is
the interface tray. Entrainment of the slurry in
the gas leaving the absorption zone is very sig-
nificant. Operating data from the Detroit Edison
St. Clair installation has shown, for example,
that a 30 foot diameter absorber would have
more than 100 gpm of slurry entrained in the
gas stream leaving the absorption zone. The
alkali contained in the entrained slurry will re-
act with the remaining SO» in the flue gas. Un-
less proper precautions are taken, a severe
plugging problem at the interface tray due to
calcium sulfite precipitation will occur. The
method used to overcome this problem is to
deluge the weeping tray with a washing medium
and thus present a liquid barrier to the en-
trained slurry.
In many current designs the wash medium used
is water, either fresh water, reclaimed water
from the sludge dewatering system termed su-
pernate, or a mixture of both; however, the
quantity of wash water available is limited by
the close loop requirement and maintenance of
proper slurry concentrations within the absorb-
er slurry system. Thus the quantity of wash
water may be inadequate to eliminate plugging
particularly at less than design load and per-
cent sulfur in coal. Conversely, to adequately
wash the tray at less than design conditions can
result in an open loop water balance which is
considered unacceptable. This is particularly
true when scrubbing low sulfur coal flue gases.
A development made by Peabody, which pres-
ently is being patented, has been to utilize a
hydroclone to provide the interface tray wash
medium. Operating data has shown that the
hydroclone, which is a liquid cyclonic classifier,
can be used to classify the solids in the recycle
slurry such that calcium carbonates can be
separated from the calcium sulfite and calcium
sulfate. Without calcium carbonate the slurry
cannot react with SO2 to create a plugging
problem. The decarbonated slurry can then be
used in whatever quantities are necessary for
the interface tray washing without upsetting the
water balance. Thus a closed loop water bal-
ance can be maintained under all operating
conditions.
The hydroclone not only furnishes a means of
providing a wash medium but can also be used
to achieve a 100% utilization of alkali or a
stoichiometry of 1.0. In the schematic shown the
hydroclone would be sized to meet both the
washing and stoichiometry improvement re-
quirements. Overflow from the wash tank is
routed to a baffled section near the overflow
nozzle from the slurry recycle tank such that
only decarbonated slurry leaves the system as
1180
-------
waste slurry. Carbonates removed from the
slurry as hydroclone underflow would be routed
back to the main stream of the slurry recycle
tank where it becomes part of the slurry re-
cycled to the absorber.
The waste slurry from the slurry recycle tank
overflows by gravity to a waste slurry sump
where it is pumped to a solids recovery system
which could be a thickener and filter or a pond.
In transporting slurry, maintaining a minimum
velocity in the transfer line at all times, is man-
datory to prevent settling of solids and elimi-
nate plugging. Since the quantity of the waste
slurry can vary widely with load conditions, it is
not possible to maintain a non-plugging flow
velocity in the transfer line unless other provi-
sions are made. The slurry transfer system is
operated on a constant velocity loop between
the waste slurry sump and the solids recovery
system. Supernate from the solids recovery sys-
tem is added to the waste slurry sump in a quan-
tity equal to the difference between the maxi-
mum and actual waste slurry quantity. This
maintains a constant velocity in the slurry trans-
fer line and thus eliminates plugging.
The basic design philosophy of Peabody's FGD
System is that each component of the system
should have its own function and that perfor-
mance of one component should not depend on
the other. Thus the absorber's only function is
to remove SOj under varying operating condi-
tions. The wash system does not depend on the
absorber operating conditions. The hydroclone
insures that alkali utilization does not change at
different load conditions. The quantity of waste
slurry produced does not affect the slurry trans-
fer system.
Control Concept
System controls can be complex or they can be
simple. The complexity of the control system
can be adapted to reflect the philosophy of the
owner as well as the requirement of the system.
Several yearjs of experience with operating
plants has shown that complex controls are not
a requirement of the Peabody FGD System.
The same concept of designing an uncompli-
cated mechanical system also carries over into
the control philosophy. Every attempt has been
made to minimize controls wherever possible.
Using overflow from the wash tank and slurry
recycle tank instead of a level controller is an
example of this approach. The fewer the num-
ber of control loops, the lower is the risk of op-
erating problems due to instrument malfunction.
Fig. 1 shows the basic controls of a Peabody
FGD System.
Because the spray tower does not depend on
gas velocity as a criteria of performance, gas
flows from 0 to 100% of design can be accom-
modated. This means that flow control of the
flue gas to each spray absorber, even in multi-
ple absorber systems, is not necessary because
of the adaptability of the spray tower to various
gas flows.
SOj analyzers on the inlet and outlet gas
streams provide the basis for establishing the
required system performance. However, neither
of these analyzers serve to control system vari-
ables. They serve only a monitoring purpose.
The quality of the coal being burned establishes
the allowable outlet ppm of SO, in the flue gas.
This then becomes the guide to the operator
for adjusting system operating variables to
achieve desired performance and minimized
operating costs.
pH control of the recycle slurry establishes the
alkali makeup to the FGD System. As the
pounds per hour of SOj to the system changes
as a result of boiler load or sulfur content of
the coal the pH controller correspondingly in-
creases or decreases the amount of alkali to
the system while maintaining the pH at the de-
sired set point. Generally when multiple taps
are made from an alkali slurry loop hydraulic
conditions can vary. To compensate for these
hydraulic variations, the pH controller resets a
flow controller and thus insures a stable control
condition.
Various ways can be used to control the quan-
tity of fresh makeup water to maintain a closed
loop water balance. A boiler load signal is but
one of these ways.
To insure that the proper percent solids in the
recycle slurry is maintained, to avoid scaling
conditions a density control is used.
Supernate from the waste solids recovery sys-
tem is returned to the slurry recycle tank by
1181
-------
FlgunS
means of a density controller. As the quantity
of waste solids produced varies with operating
conditions, the density controller modulates the
quantity of supernate returned to the slurry re-
cycle tank to maintain the proper percent solids.
During our Detroit Edison contractual perfor-
mance run, high and low sulfur coals were
burned interchangeably, with inlet SO2 concen-
tration cycling from 300 to 2700 ppm and load
condition fluctuating between 30% to 100% of
design. The control concept for Detroit Edison
is the same as outlined above and has proved
itself extremely adaptable to rapid changing op-
erating conditions.
Operating Variables
Having discussed the components of the FGD
System, their inter-relationships and the con-
trols required to maintain desired perfor-
mances, let us look into the operation of the
system at other than design conditions.
Design conditions establish the quantity of flue
gas to be scrubbed, the inlet S02 concentration
as well as the pounds of SO2 to be removed and
the efficiency at which it has to be removed.
Normal operating conditions generally result in
a reduced quantity of flue gas, lower inlet S02
concentration, less pounds of SO2 to be re-
moved and most significantly a lower SO2 re-
moval efficiency.
The operator of the FGD System has basically
two options open to him. The system can be
operated as if design conditions prevailed at all
times and no changes made to the system or the
system variables can be adjusted to reflect ac-
200
150
100
50
SO, Removal Power Requirements
20 40 60
Percent SO> Removal
100
Effect of L/C on SO, Removal Efficiency
Liquid pH = 6.3
60 80
L/G
Inlet SO, ppm 1870
Data: U.S. Steel, Barlow
140
100 120
tual operating conditions in order to minimize Flguns
operating costs. The operating variables avail-
able in order to reduce operating costs are:
a. Quantity of gas scrubbed
b. L/G ratio
c. pH of recycle slurry
A. Quantity of Gas Scrubbed
As SOj removal requirements decrease the ten-
dency is to think that the most economical way
to reduce operating cost is to by-pass as much
of the gas as possible and scrub only a mini-
mum portion of the gas. This is a fallacy. As
more and more gas is by-passed around the
system, the efficiency of removal of SO2 in the
portion left to be scrubbed increases signifi-
cantly. The energy to remove SO2 at a higher
efficiency correspondingly increases the power
requirements markedly. Fig. 2 shows the effect
of removing SO2 under different efficiency con-
ditions. This curve is based upon an inlet SO2
concentration of 2900 ppm using limestone as
the alkali. As the S02 concentration decreases
the power required to remove 1000 pounds of
SOj will increase. The curve for other alkalis
will have different horsepower values but the
shape of the curve would be the same. From
this curve it is seen that to remove a thousand
pounds per hour of SO2 at 90% efficiency would
require 160 H.P.; however, if the same thousand
pounds of S02 were to be removed at a 50%
efficiency, the horsepower requirement is less
than half or 75 HP. The pounds per hour of SO»
to be removed is the same regardless of how
the FGD System is operated. Thus the most
economical way of removing S02 is to scrub
as much of the gas as possible which means
that the S02 is removed at the lowest efficiency
possible.
1182
-------
As a general rule, for FGD Systems consisting
of multiple absorbers, an absorber should be
shut down only when the remaining units have
sufficient capacity to scrub all of the gas.
B. L/G Ratio
Fig. 3 shows plant operating data indicating the
effect of the L/G ratio on S02 removal effi-
ciency. As can be seen from these curves, as
the required SO2 removal efficiency decreases
from the design point there is a marked reduc-
tion in the required L/G. For example, at a lime-
stone slurry pH of 6.3 decreasing the required
SO2 removal efficiency from 90% to 70% de-
creases the required L/G ratio from 85 to 40.
The net effect of this fact is that a power saving
is possible when the L/G ratio is reduced to
reflect the actual SOZ removal efficiency re-
Flgure4 quirements.
Effect of pH on SO, Removal Efficiency
Inlet SO, ppm 1870
Data: U.S. Steel, Bartow
5.2
5.4
5.6 5.8
PH
6.0
6.2
6.4
C. pH of Recycle Slurry
Control of the pH of the recycle slurry becomes
another important factor in the control of SOj
removal efficiency. As can be seen on Fig. 4,
varying the pH of a limestone recycle slurry
from 5.2 to 6.3 for a given L/G ratio causes a
change in S02 removal efficiency from 86 to
96%.
Application of Operating Variables
The FGD Systems must be mechanically de-
signed so that the previously described operat-
ing variables can be applied to minimize operat-
ing costs. This is achieved by several means:
a) Multiple Absorbers per FGD System '
The number of absorbers per FGD System, Fig.
5, is dependent upon the design gas flow as well
as the required flexibility. For example, a 600
MW system would require mechanically a mini-
mum of 3 absorbers. However, if the power plant Figure 5
is to be operated at 75% load for long periods
of time four smaller absorbers may be consid-
ered. The power saving resolution from main-
taining one absorber off stream may justify the
added capital expenses.
b) Multiple Slurry Recycle Pumps per Absorber
The minimum number of slurry recycle pumps
per absorber, Fig. 6, is established by the total
Multiple
recycle
pumps
gallons of slurry being pumped. However, a
greater number of smaller capacity pumps can
be provided if additional flexibility is required.
A typical system might have three recycle
pumps each feeding a set of spray headers. No
attempt is made to manifold one pumping sys-
tem to the other. This improves reliability and
allows each of the pumps to work independently
of each other. If S02 removal efficiency has de-
creased such that L/G can be reduced, step
wise control of the L/G is achieved by shutting
off one or more of these pumps. Reducing the
Figure 6
1183
-------
L/G ratio also achieves a fan horsepower sav-
ing in addition to pumping horsepower. At de-
sign gas flow each slurry spray header contrib-
utes a Vs" w.c. pressure drop to the flue gas flow
path. Thus where two spray headers are mani-
folded to one pump, reducing the L/G ratio by
shutting off one pump per module reduces the
system pressure drop by 1" w.c.
pH control
FlgunJ c) pH of Recycle Slurry
Step wise control of the L/G ratio produces a
rough approach to the desired SOa removal effi-
ciency. Control of the recycle slurry pH, Fig. 7,
becomes the means of fine tuning the system.
An operator need only to adjust the set point of
the pH controller until the S0a analyzer meets
the desired SO2 concentration in the scrubbed
Flgunt gas.
Typical Boiler Design Condition
Boiler Capacity — 600 MW
Fuel — 3.5% S Design
1.4% S Normal
ACFM to Scrubber —1,851,000
Gas Inlet Temp. — 300°F
Percent of Gas Scrubbed —100%
Gas Inlet SO, Concentration — Design 2900 ppm
Normal 1170 ppm
Allowable SO, Concentration in Stack — 450 ppm max.
Required SO, Removal Efficiency — 89%
Alkali — Limestone
Design L/G — 90 Gal/1000 CFM
No. of Spray Tower Modules — 3
No. of Slurry Recycle Pumps per Module — 3
Example
Fig. 8 shows typical boiler design conditions
for a 600 MW boiler illustrating some the prin-
ciples discussed. Fig. 9 shows that, at the de-
sign sulfur coal, an 89% efficiency would be
required but as the sulfur in the coal decreases,
the required removal efficiency drops to slightly
over 60% for the 1.4% sulfur coal. As a result of
this lower efficiency requirement the number of
pumping stages per module could correspond-
ingly be decreased. Operating experience has
shown (Fig. 10) that each of the three pumps
operates as a stage with an equal SOa removal
efficiency. When one pump per absorber oper-
ates, an SOi removal efficiency of 52% is
achieved. With two pumps per absorber operat-
ing, a 78% removal efficiency is achieved. Three
pumps in operation provides the design SOa
removal efficiency of 89%. Flgunt
Required SO, Removal Efficiency
at Less Than Design Sulfur Coal
Design
Sulfur
Coal
"Normal
Sulfur
Coal
1000 2000
Inlet S0> Concentration
3000
0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
% Sulfur In Coal
For the normal 1.4% sulfur coal condition where
a 60% removal efficiency is required, operating
with one pump per absorber at a 52% efficiency
would be inadequate, two pumps per absorber
produces a 78% efficiency which is too high.
The two pumps per absorber method of opera-
tion would be selected and the pH of the recycle
slurry reduced until the required 60% efficiency,
as indicated by the 450 ppm SO* concentration
in the stack gas, is reached. At this condition
the amount of SO2 removed and the amount of
sludge produced is no more, nor no less than
required to meet emission standards.
1184
-------
IMh
tf.
Effect of Numbw of Pumping Stagts
on Efficwncy
IPump
SPump.
3Pump*
Figure 10
1 2
No. ol Recycle Pumps Operating Per Module
Fig. 11 shows the potential power saving pos-
sibilities under different load and inlet SO2 con-
centrations. Zero to 33% of design gas flow
requires one module on stream. 33 to 67% of de-
sign gas flow requires two modules on stream.
67 to 100% of design gas flow requires 3 mod-
ules on stream. With inlet SO2 concentrations of
450 to 850 ppm a minimum of one recycle pump
per module is required. Inlet concentrations of
850 to 1900 ppm requires a minimum of two re-
cycle pumps per module. 1900 to 2900 ppm of
SOj requires all three recycle pumps per mod-
ule operating. For each of these variations in
operating conditions, the absorber pressure
drop (D.P.) and power savings are shown (P.S.).
At design gas flow with an inlet SO2 concentra-
tion of 1170 which corresponds to the normal
1.4% sulfur coal condition a power saving of
2466 H.P. is achieved. This is a 33% saving in
horsepower.
However, the potential horsepower savings and
benefits do not end here. For illustrative pur-
poses a gas flow 80% of the design value is
used. From the systems which Peabody has de-
signed and quoted this flow rate realistically
approaches an actual full load condition. The
balance 20% usually represent design safety
factors.
Fig. 12 shows the alternate methods of opera-
tion possible at a gas flow 80% of design under
different SO2 concentration conditions.
For the case where three absorbers are on
stream, with the gas flow to the absorbers de-
creased the actual L/G ratio increases, in this
case by 25%. The result is a system which has
a SO2 removal efficiency capability greater than
the design 89%. This permits shutting off a re-
cycle pump at a higher inlet SO2 concentration
than indicated earlier, in this case 2150 ppm.
As the inlet SO2 concentration decreases, by-
passing some of the gas and shutting down ab-
sorbers while maintaining 450 ppm SO2 in the
stack gas is a possible method of operation. At
approximately 1800 ppm inlet SO2 concentra-
tion it is possible to shut down one absorber,
thus making two alternate methods of operation
possible. At 750 ppm inlet SO2 concentration it
Power Savings Under Reduced Load Conditions
(Basis: 100% Gas Scrubbed)
^\lnlet So, Cone
% of ^v.
Design ^-^
Gas Flow ^\
0-33
33-67
67-100
450-850
ppm
1 Module
1 Pump/Module
D.P. = 1V2" W.C.
P.S. = 6258 H.P.
2 Modules
1 Pump/Module
D.P. = 1V2" W.C.
P.S. = 5595 H.P.
3 Modules
1 Pump/Module
D.P. = 1" W.C.
P.S. = 4932 H.P.
850-1900
ppm
1 Module
2 Pumps/Module
D.P. = 21/2" W.C.
P.S. = 5118 H.P.
2 Modules
2 Pumps/Module
D.P. = 2V2" W.C.
P.S. =3792 H.P.
3 Modules
2 Pumps/Module
D.P. = 21/2" W.C.
P.S. = 2466 H.P.
1900-2900
ppm
1 Module
3 Pumps/ Module
D.P. = 3Va" W.C.
P.S. = 3978 H.P.
2 Modules
3 Pumps/Module
D.P. = 31/2" W.C.
P.S. = 1989 H.P.
3 Modules
3 Pumps/Module
D.P. = 3V2" W.C.
P.S. = 0
Flgun11
1185
-------
Operation at Reduced SO, Conditions
(basis 80% of design gas flow)
No. of Operating Modules
I
3000
2000
8"
1000
500
3 pumps/module
2.2 InchAP
P.S. = 742 hp
2 pumps/module
1.6 InchAP
P.S. - 2780 hp
1 pump/module
1.0 InchAP
P.S. • 4932 hp
3 pumps/module
3.5 inchAP
P.S. = 1752 hp
2 pumps/module
2.5 inchAP
P.S. = 3491 hp
1 pump/module
1.5 inchAP
P.S. = 5230 hp
P.S. = 3504 hp
P.S. = 4659 hp
P.S. = 5814hp
Flgun 12 is possible to shutdown two absorbers making
three methods of operation possible.
The power saving achieved for the alternate
methods of operation resulting from shutting
down absorbers and recycle pumps are shown
in Fig. 12.
At the expected normal inlet SOa concentration
of 1170 ppm corresponding to the 1.4% sulfur
coal, it appears that the greatest power savings
is achieved with two modules on stream and by
passing some of the gas. The earlier generaliza-
tion that scrubbing all of the. gas minimizes
power consumption still holds true. However,
because the L/G ratio is controlled in a stepwise
manner exceptions do occur. This is one of
them.
By looking at this chart another significant fac-
tor emerges. In designing the system for design
conditions no consideration had been given to
a spare module. However, for inlet SO2 concen-
trations of 1800 ppm equivalent to 2.2% sulfur
coal or less where two or more alternate meth-
ods of operation are possible a spare module
does exist. Thus the owner of this system would
have the capability of handling peak operating
condition should they occur and yet for no addi-
tional capital cost have the added security of a
spare module for a wide range of normal oper-
ating conditions.
Had no attempt been made to apply the princi-
ples used for 80% of design gas flow and the
1.4% sulfur coal, SOj removal would have ex-
ceeded the recommended requirements by 5200
Ibs./hr. of SOa. At $8 a ton for limestone and $10
a dry ton for waste sludge this would represent
an added operating cost of $850,000 per year.
This cost saving in addition to the ability to save
approximately 50% of the power requirement or
3491 H.P., highlights the significance of apply-
ing the principles outlined.
Summary
a) Alkali consumption and sludge handling
costs are minimized by adjusting the L/G ratio
and slurry pH to insure that no more than the
required amount of SOa is removed.
b) The use of the hydroclone provides the
means for achieving 100% alkali utilization and
thus stoichiometry is no longer an operating
cost factor.
c) Pumping power can be reduced by shutting
off absorbers and reducing the L/G ratio to re-
flect load and SOj removal efficiency require-
ments.
d) Operating the absorbers at reduced L/G ratio
by shutting off one or more recycle pumps per
absorber decreases the system pressure drop
with corresponding fan horsepower saving.
In summary it can be stated that the Peabody
FGD System is extremely adaptable to a wide
variety of operating conditions. What becomes
significant in designing these systems is not
only an understanding of the design conditions
but also the normal operating conditions. This
permits designing a system which produces the
lowest operating cost to the owner.
1186
-------
BY-PRODUCT-UTILIZATION/ULTIMATE-
DISPOSAL OF GAS CLEANING WASTES FROM
COAL-FIRED POWER GENERATION
by
William Ellison^ and Edward Shapiro^
ABSTRACT
A review is given of pollution control rules in solids disposal called
for by Public Law 94-580, The Resource Conservation and Recovery Act
of 1976, and now formally proposed by EPA. The text of these require-
ments indicates that the disposal of solid collected and formed in clean-
ing of flue gas from coal-fired boilers can be expected to be subject to
comprehensive environmental regulations governing design, permitting and
operation of storage and disposal sites. Levels of concentration of heavy
trace elements in fired coal will directly influence hazardousness in
disposal of both fly ash and scrubber sludge, as well as the complexity of
controls required for groundwater protection. A review of recently published
details of EPA stir-testing procedures and toxicity data on mixtures of wet-
collected fly ash and scrubber sludge indicates that fly ash and SO2
solids may be expected to regularly test out as non-hazardous. At the
same time, isolated disposal and broad availability of current and future
fly ash production for by-product utilization may be significantly limited.
This is the result of anticipated increasing use of dry fly ash catch as a
fixation reagent to stabilize SO2 sludge for improved ultimate disposal.
Major site-related hydrological and geological factors affecting selection,
design and operation of sites for permanent disposal of raw and stabilized
solid wastes are reviewed along with techniques for preliminary investi-
gation of acceptability of existing facilities and available sites. Utili-
zation of fly ash and pyrites wastes in flue gas desulfurization (FGD),
and emerging of commercial design technology for broadened application
of coal-based regenerative FGD technology is described along with
interrelated progress in management of coal cleaning waste.
I/Assistant General Manager, Air Pollution Control, NUS Corporation,
Rockville, Maryland
2/Vice President, Engineering, Pittsburgh Environmental and Energy
Systems Incorporated, Pleasant Hills, Pennsylvania
1187
-------
INTRODUCTION
FGD sludge and fly ash from cleaning of coal-fired boiler flue gas
are a major waste product of the utility industry. These materials are
of concern because of the large quantities generated as well as the
possible adverse pollution effects due to the waste properties. Because
of the hydrological and geological complexity of typical solid disposal
storage sites such as illustrated in Figure 1, many environmental
questions, including that of potential long-term effects on groundwater
quality, can arise at a major facility. Thus, fly ash and scrubber sludge
management currently poses a significant technological, environmental
and transportation problem.
In recognition of potential effects on human health and the environ-
ment resulting from the improper disposal of solid wastes, the Solid
Waste Disposal Act, as amended by the Resource Conservation and
Recovery Act of 197-6 (Public Law 94-580^/), has resulted in comprehensive
performance standard si/ and, in the case of wastes determined to be
hazardous, has established a management control system requiring
"cradle-to-grave" cognizance, including appropriate monitoring, record-
keeping and reporting throughout the system^/. Planning and design
of waste disposal facilities is thereby a major engineering undertaking
requiring effective integration in design and operation of pollution
control and ultimate disposal systems after thorough geotechnical and
hydrological study of the site^/.
3/Public Law 94-580, 94th Congress, Resource Conservation and
Recovery Act of IS76, October 21, 1976.
4/U.S. Environmental Protection Agency, "Solid Waste Disposal
Facilities, Proposed Classification Criteria", 40 CFR-257,
Federal Register, Vol 43, No. 25, pp 4952-4955, February 6, 1978.
5_/U0S0 Environmental Protection Agency, "Hazardous Waste, Proposed
Guidelines and Regulations and Proposal on Identification and
Listing" 40 CFR-250, Federal Register, Vol. 43, No. 243, pp 58946-
59028, December 18, 1978.
6/Ellison, W. and R.S. Kaufmann, "Toward Safe Scrubber-Sludge Disposal",
Power, pp 54-57, July, 1978.
1188
-------
Solid Waste
Disposal Site
Power Plant
Municipal,
Agricultural,
Industrial,
Water Use
t t
rfcnT;;;
Leachate
Watertable
WW..V..V<~"-V
Direction of Movement
S Recharge
Recharge Zone
-r- Recharge I
r t r : '
Key
Zone of Aeration
Unconfirmed Aquifer
(Zone of Saturation)
Confining Layer
Confined Aquifer
35 Lake
FIGURE 1
Interaction of Solid Waste Disposal
and Groundwater Systems
1189
-------
SOLID WASTE MANAGEMENT
TO MEET RCRA/EPA REGULATIONS
A review of pollution control guidelines for ultimate disposal of
solid waste called for by Public Law 94-580, The Resource Conservation
and Recovery Act of 1976, and now issued by EPA as proposed rules
Indicates that the disposal of boiler flue-gas-scrubber solid wastes
from flue gas desulfurization (FGD) as well as collected fly ash can be
expected to be subject to comprehensive controls and standards under
new governmental regulations.
RCRA/EPA Criteria for Disposal of Solids
While these general provisions, which are made under Subtitle D,
do not contain as stringent operating and monitoring requirements as
apply to hazardous waste disposal facilities under Subtitle C, they
impose a comprehensive new discipline for all large-scale storage or
disposal of solids. Major aspects of the criteria are as follows: .
1. Broad provisions to prevent siting of disposal facilities in environ-
mentally sensitive areas including wetlands, floodplains, so.le
source aquifers, etc.
2. Protection of surface water bodies through extension of the NPDES
permitting requirement for any point source discharge and by control
of non-point sources to prevent or minimize pollutant discharges
3. Stringent provisions as follows protecting groundwater in usable
aquifers, i.e. those containing less than 10,000 mg/1 total
dissolved solids, (except those that have received State designation
fora use other than as a drinking water supply for human consumption):
(a) Quality of groundwater beyond the disposal-facility property-
boundary shall not be endangered by the facility, i.e. degraded
such that more extensive treatment is thereby required to prepare
the water for drinking water purposes.
(b) Prevention of endangerment shall be assured either by the collection
and proper treatment and disposal of leachate produced by the
facility or alternatively, by site selection and facility design to
adequately control migration of leachate from the disposal structure.
(c) Where appropriate, prevention of endangerment shall be verified by
a suitable groundwater monitoring program.
Evaluation of Solids Disposal Sites and Facilities
1190
-------
EPA Inventorying of Deficient Sites
Subtitle D also requires that EPA immediately proceed to compile a
nationwide inventory of objectionable solid waste disposal sites. Using
readily available geologic and hydrologic data, all existing facilities
are to be assessed by the government for potential contamination of
groundwater and surface water so that necessary in-depth site evaluations
including field collection of detailed data can be selectively carried out
where necessary. -EPA's new SIA (surface impoundment assessment)
evaluation system-^ which yields a first-round numerical approximation
of the relative environmental impact of waste impoundments, provides
means for consistency in this preliminary prioritizing assessment of
existing facilities.
Applicability of SIA System
The SIA system consists of a step-by-step procedure for collecting
available site information from which the water-pollution potential of a
specific facility may be estimated. SIA can also be used to calculate
the comparative pollution potential of each of several alternative sites
being considered for future use in solids storage or disposal so as to
provide an initial ranking of site acceptability.
Parameters Influencing SIA Score
The SIA system makes a best available estimate of the following
principal parameters bearing on the two major concerns in control of
water pollution from disposal of solids:
Groundwater Contamination
; The potential for groundwater contamination is most strongly influenced
by:
(1) The permeability of the existing earth material and the thickness of
the unsaturated zone above the groundwater, which tends to protect
the underlying aquifer
(2) The permeability of the existing earth material and the thickness of
the saturated zone (the aquifer itself) which represents the quantity
of groundwater availability from the aquifer
(3) The usefulness of this groundwater as determined by whether the
aquifer is currently in use as a source of drinking water, or if not,
the total dissolved solids (TDS) content of the groundwater
7/U.S. Environmental Protection Agency, "A Manual for Evaluating
Contamination Potential of Surface Impoundments", Report No.
570/9-78-003, June, 1978.
1191
-------
(4) The potential for presence of critical contaminants in the waste
as judged by the waste-material source or type and the corresponding
EPA hazard potential rating .
Endangerment of Current Water Supplies
The potential endangerment of water supplies currently in use is
most strongly influenced by:
(1) The degree of vulnerability to contamination by the impoundment
as judged by the type of existing water use, i.e. water well or
surface supply
(2) The degree of anticipated attenuation of contaminant flow as
related to the distance of existing water uses from the impoundment
(3) The proportion of contamination flow that may reach use-points as
determined by the location of existing water uses with respect to
the anticipated flow direction of the contaminated groundwater .
Interpretation of SIA Scores
EPA's SIA manual includes charts tabulating selected values of
each of these parameters and displaying corresponding standardized
numerical scores. The estimate of contamination potential is obtained
by summing individual score components for its parameters. This
total groundwater-contamination-potential score can range from 1 to
30. In comparing sites and making a preliminary site evaluation, a
score of 10 or less may mean a potentially good site for disposal while
a score of 20 or more indicates that there may be significant problems.
The supply-endangerment score will range between 0 and 9, 4 or less
being favorable. Use of the SIA system makes it possible to identify
specific low-scoring sites and thereby justify more detailed study of
them for the purpose of verifying future use for disposal; or, in the case
of existing facilities, to identify high-scoring sites and thereby justify
further investigation of them to ascertain possible need for remedial
measures.
Utilization of SIA System by Owners of Facilities
The SIA system and the EPA manual in which it is presented affords
the owner's staff the opportunity to gain valuable technical understanding
•of water pollution control effects of solid waste disposal and appropriate
design and operating practices. Its timely use can help guide assembly
of site information in a standardized format that can be understood by all
concerned individuals while at the same time providing an early indication
of possible actions required to achieve compliance with RCRA-EPA rules.
1192
-------
It should be noted however that the SIA evaluation 'will be made most
effectively and lead to most meaningful scoring if it is carried out by
a professional geologist who is capable of adequately characterizing
geologic environments and of selecting suitably conservative values for
the applicable parameters.
rRequirements for Design and Operation of Hazardous-Waste Disposal Sites
Section 3004 of RCRA Subtitle C addresses standards applicable to
owners and operators of hazardous-waste storage, treatment, and dis-
posal facilities. These regulations, now issued by EPA as proposed
rules, define the levels of environmental protection to be achieved and
provide the criteria against which EPA will assess applications for
permits to design and operate disposal and storage sites for hazardous
waste.
Of particular significance are RCRA-EPA-pre scribed alternative
leachate-containment methods applicable to hazardous waste facilities
located above usable aquifers, (containing less than 10,000 mg/liter
total dissolved solids). See Figure 2-A, B, and C:
landfill Containment
landfills over usable aquifers must be either:
(a) Arranged as per Figure 2-A, when natural geologic and climate
conditions allow, to provide a containment in the form of a
natural liner of soil meeting specified criteria and at least the
equivalent of a 10-foot thickness with a permeability no greater
than 1x10" cm/sec (but with thickness no less than 5 feet), or
Constructed and operated such that leachate formed can be
contained and removed from the landfill site when natural conditions
do not permit the above containment. Under this option, per
Rgure 2-B, a limited containment shall be provided consisting of
a minimum 5-foot thickness with a permeability no greater than
1 x 10 cm/sec. This liner shall bo sloped at a 1% minimum
grade so that the leachate is drained directly by gravity to a
collection sump for removal and be overlain with an implaced
permeable layer of material such as gravel or sand so that any
generated leachate can move rapidly to the sump. (Alternatively,
the regulation allows a double liner and leachate-collection
Installation, the upper liner being a soil liner at least 3 feet
thick with permeability no greater than 1 x 10~7 cm/sec.)
1193
-------
A. Natural Containment
Landfill or
Disposal Pond
B. Leachate Removal (Landfill)
FIGURE 2 RCRA-EPA-PRESCRIBED ALTERNATIVE
GROUNDWATER PROTECTION METHODS
1194
-------
C. Leachate Removal (Pond) Leachate
Well
Fig. ZC.
RCRA-EPA-Prescribed
Alternative Groundwater
Protection Methods
1195
-------
. Pond Containment
Disposal-pond type impoundments over usable aquifers must be either:
(a) Arranged as per Figure 2-A, when the conditions allow, to
provide a containment along the bottom and sides of the pond
in the form of a natural liner of soil meeting specified criteria
and at least the equivalent of a 10-foot thickness with a
permeability no greater than 1 x 10~7 cm/sec, (but no less than 5
feet thick and no more than 1 x 10~7 cm/sec permeability.), or
(b) Constructed and operated such that leachate formed can be
contained and detected between a top liner and a bottom liner and
removed when the conditions do not permit the containment above.
Under this option, per Figure 2-C, the top liner shall be
constructed of specific reconstituted clays or of artificial
materials meeting prescribed criteria with permeability no
greater than 1 x 10~7 cm/sec and of sufficient thickness to
ensure mechanical integrity. The bottom soil liner shall be
of natural in-place soil meeting specified criteria and at least
5-foot thickness with a permeability no greater than 1 x 10""'
cm/sec. (An artificial bottom liner may be used only for
temporary disposal sites.) The leachate detection and removal
system is to be a gravity flow drainage system installed between
the top and bottom liners.
Assessment of Criteria for Solids Hazardousness. Identification
Section 3001 of RCRA Subtitle C includes criteria for ascertaining
hazardousness" based on testing of a representative sample of the solids
source. In view of the possible presence of significant trace quantities
of heavy metals originating from the fired coal, toxicity evaluation is
a key factor in determining if FGD waste and fly ash from a specific
source is hazardous.
Solids Evaluation by Toxicant Extraction Procedure (TEP)
Provisions for site-specific examination for toxicity anticipate testing
the potency of the leachate (extract) yielded by the proposed EPA Toxicant
Extraction Procedure (TEP) applied to a sample of the actual solids. The
proposed TEP calls for dewatering the representative solid waste sample/
(crushing it if necessary to pass a 3/8" standard sieve), and stirring it
in contact with 16 times its weight of deionized water while controlling
1196,
-------
pH at 5.0 by addition of acetic acid during a 24-hour extraction period.
A minor adjustment in the volume of the separated liquid extract, (by
addition of deionized water), and the combining of this liquid mixture
with the liquid-phase material formed in the dewatering of the sample
results in a final liquid product, referred to as the TEP extract.
EPA's proposed rules indicate that in application of the TEP a solid
waste will be deemed to be toxic if the TEP extract contains a concentration
of any of the trace elements listed in EPA's National Interim Primary
Drinking Water Regulation, (NIPDWR), greater than or equal to ten times
the concentration allowed in drinking water by NIPDWR. (See Figure 3
which identifies the potentially toxic constituents listed in NIPDWR as
well as the NIPDWR-allowed concentrations of these substances in
drinking water and the ten-fold higher critical/threshold concentrations
for TEP extract corresponding to a critical toxicity level in. the tested
solids.) This factor of ten reflects EPA's judgment that underground
strata will attenuate inorganic concentrations in leachates ten-fold.
Thus a leachate potency corresponding to the critical extract strength,
i.e. containing one or more of the trace elements at a concentration of
10 X NIPDWR, is expected to be gradually mitigated, the leachate
ultimately entering the groundwater system with concentrations of trace
elements no greater than NIPDWR.
Generic Assessment of Typical Fly-Ash-Containing FGD Sludge
The TEP results in a twenty-fold mixing/dilution effect (weight
ratio of water feed to sludge sample feed) by the use of extraneous
water in forming the TEP extract. Therefore if it were assumed that:
(a) all inorganic substances in the extract originate solely from
liquid-phase material in the dewatered sludge sample and from
the liquid-phase material formed in the dewatering of the sludge
sample and
(b) the free-moisture (raw-occluded-liquor) content of sludge is
50% by weight as in typical plants
then the TEP extract concentrations would correspond to occluded-raw-
liquor concentrations that are 20 -^ 0.50 = forty-fold higher than those
in the extract. Since a significant amount of the trace element content
of the extract may originate from the solid-phase material in the sample,
the proportionality between the liquor and extract concentrations could
be expected to be substantially lower than 40/1 in actual sludge-sample
testing, e.g. in the range of 10/1 to 20/1. This reduction of concen-
tration in the extract as compared to that in the waste, reflected in TEP,
corresponds to the diluting effect of percolating rainwater that causes
1197
-------
Contaminant
FIGURE 3
CRITICAL CONTAMINANT CONCENTRATIONS
IN TEP EXTRACT FOR DESIGNATION
OF WASTES AS HAZARDOUS
Drinking Water
NIPDWR
Level, (EPA Primary Std)
Milligrams per liter
Arsenic 0.05
Barium 1.
Cadmium 0.010
Chromium 0.05
Lead 0.05
Mercury 0.002
Selenium 0.01
Silver 0.05
Endrin (1, 2,3,4,10,10-hexacloro~6 .0002
7-epoxy-l,4,4a,5,6,7,8,8a-octahydro-l,
4-endo, endo-5, 8-dimethano naphthalene).
Lindane (1,2,3,.4,5,6-hexachlorocyclohexane 0.004
gamma isomer).
Methoxychlor 1,1,1-Trichloroethane) 0.1
2,2-bis (p-methoxyphenyl).
Toxaphene (CioH10cl8~techn^-cal 0.005
chlorinated .camphene,
67-69 percent chlorine).
2,4-D, (2,4-Dichlorophenoxyacetic acid). 0.1
2,4,5-TP Silvex (2,4,5-Trichloro-
phenoxypropionic acid).
0.01
Critical Extract
Level,
Milligrams per Liter
(10 X NIPDWR)
0.50
10.
0.10
0.50
0.50
0.02
0.10
0.50
0.002
0.040
1.0
0.050
1.0
0.10
1198
-------
the leaching phenomenon. Thus sludge liquor concentrations that are
capable of yielding critical extract concentrations will be typically
greater than ten times ten X NIPDWR or 100 X NIPDWR.
8/
On the other hand, recent publication of results of sludge
management study activity by EPA centered on disposal tests at TVA's
Shawnee Station, Paducah, Kentucky, includes extensive field data
on trace-element chemical-composition of this surface moisture in
fly-ash-containing limestone type FGD systems at bituminous and
sub-bituminous coal fired utility plants, (with wide-ranging propor-
tions of fly ash), which indicate comparatively low concentrations.
Ratioed to NIPDWR this data indicates that inorganic constituents
affecting drinking water quality are present typically at concentrations
only five times greater than NIPDWR and at levels no greater than
twenty times NIPDWR. Thus fly-ash and scrubber sludge can be •
expected to be typically found to be non-hazardous as a result of
TEP measurement by a margin of over 5/1 and to contain significantly
less than the EPA-defined critical/threshold concentration of leachable
toxic inorganic contaminants.
ASH/COAL UTILIZATION IN FGD SYSTEM DESIGN
Coal-based energy generation and attendant production of fly ash
provide significant opportunity to gain maximum benefit of coal and
fly ash use.
Fly Ash as Fixation Agent in FGD Sludge Stabilization
Means of Stabilizing Gas Cleaning Waste
Sludge stabilization is chemical processing to fix scrubber sludge so
as to facilitate improved handling, transportation, placement, and con-
solidation at the ultimate disposal site. A common method now used at
new coal-fired units includes pug mill blending of mixtures of dry fly
ash with FGD filter cake and other dry additives followed by several
days curing before compaction at a landfill disposal site.
8/Rossoff, Jerome, etal., The Aerospace Corporation, "Landfill and
Ponding Concepts for FGD Sludge Disposal", U.S. Environmental
Protection Agency, Industry Briefing Conference on Technology for
Lime/Limestone Wet Scrubbing, Research Triangle Park, North
Carolina, p. 4, August 29, 1978.
1199
-------
Benefits of Solids Stabilization
Extensive test program work by EPA to date verifies that this method
of chemical treatment significantly improves the load bearing characteristics
of FGD sludge, decreasing the solubility of the major chemical species by
vi factor of two to four, and reducing sludge permeability by an order of
magnitude^/. Thus stabilization of FGD sludge, made possible by the
presence of substantial quantities of fly ash available at coal-fired
plants can help to insure effective ultimate disposal at a landfill site.
In some instances, land may be improved and be reclaimed for beneficial
use. At the same time, stabilization acts to seal in the surface moisture
occluded in the sludge (which serves as a vital purge of soluble solids)
reducing leachate formation and recycle to the FGD system.
Dry Scrubbing Waste
Dry-scrubbing type FGD systems have been extensively demonstrated
during 1978 in lignite-fired boiler service and a number have been ordered
during 1978 for full-scale application in bituminous, sub-bituminous and
lignite service. This new FGD method uses slaked lime in a spray-type
gas absorber comparable to a spray dryer to collect SG^ and convert it
to a dry reaction product. The desulfurized flue gas flows to a fabric
filter or electrostatic precipitator to collect dry SO2-reaction-product
and fly ash. The waste product from dry scrubbing is a dry mixture of
fly ash, unused lime and calcium sulfite and sulfate comparable to
stabilization mixtures from wet-scrubber type FGD systems and thus may
be expected to be a self-stabilizer in landfill disposal.
Fly Ash as FGD Reagent
Background
Extensive testing and application of wet scrubbers on lignite nnd sub-
bituminous service has demonstrated the effectiveness of alkaline fly ash in
absorbing flue-gas SO2. In conjunction with lime or limestone addition, SOo
removal efficiency ranges as high as 65 to 90%. Commercial scrubber
installations designed to utilize fly ash to achieve FGD include Montana
Power Company's Colstrip Station Unit Nos. 1 and 2, Northern States Power
Company's Sherburne County Station Unit No. 1, Minnkota Power Cooperative's
M. R. Young Station Unit No. 2, and Minnesota Power and Light Company's
Clay Boswell Station Unit No. 4.
9/U.S. .Environmental Protection Agency, "Control of Waste and Water
Pollution from Power Plant Flue Gas Cleaning Systems; First Annual
R and D Report", Report No. EPA-600/7-7G-018, p. 2, October 1976.
1200
-------
SO- Absorption and Gypsum Scale Control
i
In recirculating scrubbing slurries containing a substantial level of
alkaline western fly ash, the dissolving of alkali metal components of the
fly ash results in a buildup of soluble sulfites, principally sodium and
magnesium, in the liquid phase. This, dissolved active alkali helps achieve
a high SC>2 removal rate, forming bisulfite ions, which react with lime, the
predominant fly-ash alkali component, to precipitate the SO£ catch as
calcium sulfite and sulfate. The elevated sulfite ion concentration may
depress the dissolved calcium content, (because of the very low solubility
of calcium sulfite), to a level sufficient to reduce the calcium and sulfate
to an unsaturated level. In this gypsum-scale-controlled mode, sulfate
precipitates from the system by attaching itself as calcium sulfate to the
principal calcium-sulfite precipitate-crystal, forming a co-precipitate.
Recent disclosure of tests on full-scale scrubbers at Arizona public
Service Company's Four Corners Station Unit Nos. 1 and 2,-^/ supplemented
by electron micrographs indicates that at fly-ash slurry solids concentration
in excess of 5%, calcium sulfate precipitate from the unsaturated liquid
phase nucleates on the fly ash particles. Due to absence of abrasive
effects at elevated fly-ash slurry solids concentrations, APS has concluded
that precipitation of SO2 catch on the fly ash particles significantly reduces
the abrasiveness of the suspended fly ash particles.
Regenerative FGD Selection
General
Appropriate selection of by-product FGD technology to suit site
conditions offers a major degree of flexibility in meeting environmental
pollution control requirements. With such systems collected fly ash
emission can be fully isolated from other waste materials for immediate or
future use in meeting growing market demands for this by-product material.
Regenerative-type SO2 removal systems provide a basis for meeting the
sludge disposal problem by converting the bulk of the SC>2 catch to saleable
by-product: gypsum, sulfuric acid, elemental sulfur or other marketable
products. Although limited by the size and locality of gypsum markets,
by-product gypsum manufacture can offer important economies in view of
the simplicity of FGD system design in this regenerative process mode,
Significant advances have been made in West Germany in the last few
years in application of regenerative FGD systems to yield commercial
gypsum by-product.
10/NeIms, W. M., and C. F. Turton, "Sulfur Removal Testing of Particulate
Scrubbers at Four Corners". Arizona Section of American Society of
Mechanical Engineers, Phoenix, Arizona, pp 13-15, January 12, 1978.
1201
-------
Waste Management
Fly-ash collection is carried out upstream of FGD, preferably by dry
collectors, thus permitting fly-ash to be stored in a landfill-type facility
without generation of process liquid effluents other than leachates produced
and controlled at the disposal site. The FGD system, generally of the
wet-process type, will generate a process effluent in the form of a purge
of scrubbing liquid containing non-precipitating components including
chlorides originating in the fired coal. Minimum requirements for liquid
effluent treatment prior to discharge typically include ^eaction with
limestone or slaked lime at elevated pH to precipitate heavy metal com-
ponents .
Utilization of Pyrites Waste and Coal for Reagent/Energy Needs
The State of Pennsylvania has recently announced current full-scale
installation for 1979 startup at a small State-owned coal-fired power plant
in Pittsburgh of an FGD system that can utilize ferrous sulfide from pyrites
waste as the chemical absorption reagent. This gas cleaning technology
of Pennsylvania Environmental and Energy Systems, Incorporated, (PENSYS),
Pittsburgh, Pennsylvania, called Sulf-X Process, is based on use of waste
Iron feedstocks including concentrated pyrites waste from bituminous
coal cleaning or non-ferrous mining and milling to make up system iron
losses. The unique chemical characteristics of ferrous sulfide maintains
an absorption-reduction process mode resulting in efficient simultaneous
removal of SC>2 and NO (nitric oxide). Common coal may be combusted
to calcine the pyrites makeup and the spent ferrous/sulfur chemical
thereby producing elemental sulfur by-product in vapor form and regenerating
the ferrous sulfide reagent for reuse in the absorption-reduction gas cleaning
step. Development of this pyrites-fed regenerative FGD process for
commercial service has been under way since 1976 through coal-based
experimental and demonstration programs including:
(a) Wet scrubber pilot plant gas cleaning operation and testing
for the Department of Army, the Pennsylvania'Science and
Engineering Foundation of the Pennsylvania Department of
Commerce, and the Appalachian Regional Commission.
(b) Testing and assessment of technology for gas cleaning, sulfide
regeneration, and sulfur by-prbduct manufacture for the
Appalachian Regional Commission
1202
-------
(c) Process design evaluation through laboratory, bench scale
and pilot/demonstration stage testing for the U.S. Department
of Energy
CONCLUSIONS
Current energy planning based on increased use of coal-fired boiler
plants in the United States requires support through continued development
and availability of improved techniques for controlling environmental impact.
Commercial by-product utilization and/or acceptable ultimate disposal of
by-product solids generated by stack-gas cleaning systems at existing
and new utility and industrial plant? is of major significance in limiting
environmental pollution.
Moreover, compliance with new RCRA/EPA solid waste regulations
applicable to storage and/or disposal of scrubber sludge and collected
fly ash will be most economically achieved by integration, both in design
and operation, of pollution control and waste handling facilities, and by
adequate evaluation of site conditions so as to achieve advantageous
production and utilization of marketable by-product forms where applicable
and to assure effective containment of contaminants in solids which are
to bs discarded or stored.
1203
-------
FLUE GAS DESULFURIZATION AND FERTILIZER MANUFACTURING:
PIRCON-PECK PROCESS
by
R. B. Boyda
Arthur G. McKee & Company
10 S. Riverside Plaza
Chicago, Illinois 60606
1204
-------
FOREWARD
Information presented in this paper is intended to introduce
the Pircon-rPeck flue gas desulfurizatioh process and highlight
some of its attractive features. Because process developments,
are still generating patent applications, in addition to those
already issued or still pending and allowed," many key details
are not included. We invite anyone desiring additional information
to contact Mr. Gene Barber in McKee's Chicago office at telephone
number (312) 454-3899.
1205
-------
INTRODUCTION
The Pircon-Peck flue gas desulfurization (FGD) process presented in
this paper was developed by Mr. Ladd Pircon and Dr. Ralph Peck. The
process utilizes patented heterogeneous reactor technology developed by
Mr. Pircon as cooling and absorber towers in conjunction with chemistry
developed and demonstrated by Mr. Pircon and Dr. Peck at the Illinois
Institute of Technology. The process is unique in that in addition to
providing a means of controlling S02 emissions, it provides the owner
with an opportunity to earn a profit on the invested capital.
PROCESS DESCRIPTION
The Pircon-Peck FGD process, as shown in Figure I, utilizes "activated"
phosphate rock, ammonia, and flue gas as raw materials to produce ammoniated
phosphate fertilizers. Laboratory and pilot plant testing indicate that
in the process of producing fertilizer, sufficient S02 can be removed
from the flue gas to provide compliance with all .EPA regulations, both
existing and proposed.
The chemistry utilized to produce the product material can be
summarized by three chemical reactions. These are:
(1) Sulfur Diammonium Ammonium Monoammonium
Dioxide Phosphate Bisulfite Phosphate
S02 (NH4)2HP04 NH4HS03 NH4H2P04
1206
-------
AMMONIA-
ACT1VA1
PHOSPHATE ROCK
FIGURE I
CLEAN FLUE GAS
i
I
SQS RICH FLUE GAS
AMMONIATED
PHOSPHATE
FERTILIZER
(2) Phosphate
Rock
Ammonium
Bisulfite
Calcium
Sulfite
Diammonium,
Phosphate
Ca3(P04)2
3NH4HS03
3CaS03 (NH4)2HP04
4- Monoammonium
Phosphate
NH4H2P04
1207
-------
(3) Monoammonium Diaminoniuin
Phosphate Ammonia Phosphate
NH4H2P04 NH3 (NH4) 2HP04
An important feature of this chemistry is the pretreatment of the
phosphate rock which facilitates the reaction of the rock with the weak
sulfurous acid produced in the scrubber. Standard phosphate fertilizer
chemistry requires concentrated sulfuric acid to acidulate the rock and
liberate the desired phosphate molecule.
In many ways, the Pircon-Peck process is similar to standard double
alkali technology. In each system, a soluble alkali is utilized in the
scrubber portion of the process and in each system this alkali is regen-
erated by a calcium source. In the Pircon-Peck process, the soluble
alkali is diammonium phosphate and the calcium source is phosphate rock.
However, where alkali regeneration with lime produces water as a by-
product, diammonium phosphate regeneration with phosphate rock produces
additional phosphate. Neutralization of this by-product with ammonia
produces the final diammonium phosphate product.
In addition to producing diammonium phosphate, the regeneration
reaction also produces calcium sulfite and sulfate. This material can
be included in the fertilizer product or separated for disposal. If
included in the fertilizer product, the material would be similar to
superphosphate fertilizers presently produced by the fertilizer industry.
1208
-------
These fertilizers contain the calcium sulfate solids produced by the
acidulation of phosphate rock with sulfuric acid. While the final
product form will be dictated by the local market conditions, the
advantage of including the calcium sulfite/sulfate solids in the product
is obvious.
Figure II provides a view of the scrubber portion of the process.
Flue gas enters the system and. is first passed through a high efficiency
particulate control device. The use of a particulate removal system
minimizes the possibility of product contamination from fly ash. After
fly ash removal, the gas is sent to a cooling tower where it is quenched
and saturated to its adiabatic saturation temperature. The cooled and
saturated flue gas is next processed through the heterogeneous reactor
tower (see Appendix I) where the flue gas is contacted with a saturated
solution of diammonium phosphate and the SC>2 is removed. After passing
through a demister, the S02 free flue gas is sent to the stack. If
necessary, reheat can be added to aid in plume dispersion.
The SC>2 rich absorption liquor is taken from the heterogeneous
reactor system and sent to an agitated attack tank where it is contacted
with ammonia and phosphate rock. The slurry generated in the attack t
tank is next sent to a product separator (clarifier) where solid diam-
monium phosphate and calcium sulfite/sulfate crystals are withdrawn as a
slurry. The product slurry is sent to a fertilizer plant where the
product is converted to its final form. The overflow from the product
separator is a saturated solution of diammonium phosphate and is sent
back to the heterogeneous reactor tower for further reaction with
1209
-------
PIRCON~PECK
FGD PROCESS
COOLING HETEROGENEOUS
TOWER
REACTOR
WATER
FLY ASH
FLUE GASnREMOVAL
FROM BOILER
FLY ASH
TO DISPOSAL!
FLUE GAS TO STACK
ACTIVATED PHOSPHATE ROCK
AMMONIA
WATER
TO
FERTILIZED
PLANT
BLOW DOWN
FIGURE 1C
-------
The material from the product separator, on. a dry basis, has an
approximate analysis of 7-20-0 (N, PO^S' ^9^' TMS analysis can be
adjusted as desired by the addition of ammonia, phosphoric acid^ or
other N-P-K materials. The product form chosen by McKee for evaluation
is a granular material with an analysis of 9-30-0. This analysis was
chosen because posted prices are available to facilitate an economic
evaluation of the process. Since fertilizer granulation plants are
standard technology, no details are included.
As indicated, it is possible to produce product in various forms.
Some alternates to granules with N and ?2^s va-*-ues are:
1. Pure diammonium phosphate
U
ff 2. Granules with N, P, and potassium values
/n 3. Suspensions
it 4. Flakes
1211
-------
PROCESS HISTORY
The Pircon-Peck process evolved from Mr. Pircon's knowledge of
phosphate chemistry, which he developed during his employment in the
fertilizer industry, and his invention of the heterogeneous reactor
technology. Mr. Pircon invented the heterogeneous reactor technology
in the early 1970fs and commercialized it as a low cost, high efficiency
particulate control device. This technology was marketed commercially
and over a dozen installations are presently in operation.
The major event in the development of the technology occurred
when funding was obtained from the Illinois Institute for Environmental
Quality for the design, construction and operation of an 800 acfm pilot
plant. This pilot plant was installed at the Illinois Institute of
Technology (IIT) and operated by graduate students for more than two
years. This work was supervised by Dr. Ralph Peck. Work performed by
students resulted in IIT granting two doctorate and six masters degrees
in chemical engineering.
The operation of this pilot plant provided confirmation of the
process principles and information for design of a commercial sized
demonstration plant. ,In addition, sufficient fertilizer product was
produced so that test work could be performed by Argonne National
Laboratories. This test work was also funded by the Illinois Institute
for Environmental Quality.
-------
At the conclusion of the pilot plant work, the following items were
demonstrated:
1. More than 95% of the sulfur oxides present in flue gases
generated by a 6.2% sulfur coal, could be removed by the
pilot plant scrubbing system.
2. Particulates generated by the underfeed stoker fired boiler
were removed so that no particulates were visible in the
stack plume.
3. The process produced a satisfactory fertilizer product.
4. The required liquid/gas ratios for the system were con-
siderably less than those for conventional scrubbers.
5. The overall pressure drop across the pilot plant system,
including the boiler, cooling tower and scrubber, was
about two inches of water.
1213
-------
PROCESS EVALUATION
McKee's evaluation of the Pircon-Peck process was performed by
reviewing laboratory data, pilot plant records and pilot plant operating
experience. In addition, extensive discussions were held with the
process inventors to develop additional background information. This
investigation has shown that:
1. The process can be carried out in a plant of very simple
design.
2. Few moving parts are involved in the equipment.
3. There is considerable flexibility in the process.
4. The process can be run with clear liquids or slurries
in the heterogeneous reactor tower.
5. The product fertilizer analysis can be adj'usted td fit
the needs of the local market.
6. Scaling in the absorber is not a problem due to the
process chemistry.
7. For the same removal rates, equipment is simpler and
smaller than for other scrubbing processes.
The economics of the process were evaluated by preparing a
mechanical design and a heat and material balance for the 100 mw sized
facility. Once this was complete, capital and operating cost estimates
were made and an economic analysis performed. The design basis used for
this work was:
1214
-------
Fuel: 3.5% Illinois High Volatile-C Coal @ 11,500 Btu/lb.
Heat Rate: 11,500 Btu/kwh
Flue Gas: 2.12 scfm/kw @ 75° F
SOX Removal: 95%
Phosphate Rock Conversion: 95%
Service Factor: 85% (7450 Hr./Yr.):
On this basis, it was estimated that the plant, including the
scrubber and fertilizer granulation plant, would cost approximately
$10.7 million and produce 128,600 tons per year of a 9-30-0 analysis
granular fertilizer product. It was estimated that the margin between
the fertilizer manufacturing cost and selling price would be $50.00 per
ton.
Using these economic parameters,-a project feasibility analysis was
performed. This analysis was based on:
Discounted Cash Flow Techniques
Twenty Year Project Life
Income Tax 0 50%
Sum of the Years Digits Depreciation
Working Capital of 13% of Sales
This study indicated that the project would have an Investors Rate of
Return (ROI) of 22% over the life of the project.
Because of the preliminary nature of the estimate and the inherent
uncertainty of prices in the fertilizer industry, McKee believes that
the significance of the indicated ROI is not its absolute value, but the
fact that it is significantly positive for even a small sized facility.
1215
-------
The economic viability of the process was further confirmed by studies
of the sensitivity of ROI to changes in key economic parameters. Para-
meters studied included capital cost, production volume, variable costs
and margin. As an example of this work, Figure III shows the sensi-
tivity of ROI to variations in the margin between selling price and
manufacturing cost. These analyses show that this margin is the most
sensitive parameter. A summary of the projected operating costs is
given in Figure IV.
McKee has also looked at the market for the process product. At
the present time the market for fertilizer, on a ?205 basis, exceeds 10
million tons per year and is expanding at a rate of approximately 6%
per year. On the design basis presented in this paper, a 500 mw plant
would produce 190,000 tons per year of ?205 equivalent product which
would fit easily into the growing fertilizer market. This is particu-
larly true in the midwest and farm belt regions.
1216
-------
n
o
4-1
(U
CJ
100
so
80
70
60
40
30
20
10
— i • ' • ••'
H
p — ,^. —
M ..-..i
. ,-T . 1 , ~
j
1
1
U
1
~ ' ' ' ' ' '
. 4
— s
— 1
j
u— t
fc=*=
b=ir=
1 1
1— ' '
\ ^ :
.
Figu
ens it
LOI vs
i ,
i
i j
— J
=*=
1
. r~
\. — 4,... . .
re III
£vity
\=t=
^=
of
in
— : — ,
=*=*
— 1
M
*=
— -
1 , , .
1 ^ — _
1
!
=i
=tt=
— f —
It i
^=t=n
'4 i 1
1 1
1=
r_ ... . — .
1 *
1 J . .
1
^ ^^ -
11 '
1 — • — i —
•8
8
16
ROI (%)
32
48
1217
-------
FIGURE IV
CAPITAL & OPERATING COST SUMMARY
INVESTMENT $ 10,726,000
PRODUCTION VOLUME 128,635 Tons/Yr.
SALES MARGIN $ 50 Ton
SEMI VARIABLE COSTS
MAINTENANCE $ 845,000
LABORATORY 96,000
INSURANCE & PROPERTY TAX 370,000
PLANT OVERHEADS 534,000
ADMINISTRATION & MARKETING 294,000
$ 2,139,000
VARIABLE COSTS
RAW MATERIAL $ 8,225,000
LABOR & SERVICES 1,009,000
UTILITIES & FUEL 924,000
$ 10,158,000
TOTAL OPERATING COSTS $ 12,297,000
1218
-------
CONCLUSION
As a result of our process studies McKee believes that the Pircon-
Peck process offers significant advantages over existing commercial
processes. Not only does the process provide an opportunity to make
pollution control profitable, it actually provides an economic incentive
to utilize high sulfur coals. This incentive results from the fact that
the quantity of fertilizer product producted is directly proportional to
the quantity of sulfur captured. As a result, the plant operator will
want to operate the system at S02 capture rates in excess, of EPA standards,
McKee has recently entered into an agreement with the process
owner, Mr. Ladd Pircon, to commercialize the technology. It is our
present co-objective to accomplish this by the design, construction, and
operation of a commercial sized demonstration plant with a minimum
rating of 100 mw. At the present time we are working to put together
the four items necessary to achieve our objective. These four items
are; funding, a site for the projec.t, an operator with fertilizer produc-
tion experience and a commitment from a fertilizer company to market the
fertilizer product. Significant progress has been made in this effort
at two separate locations.
1219
-------
APPENDIX I
HETEROGENEOUS REACTOR
TOWER
The heterogeneous reactor tower included in this design utilizes
differential velocities between liquid, gas, and solid particles to
obtain high mass transfer coefficients. The tower internals (see Figure
V) consist of a series of two conical restrictions preceded by spray
nozzles and followed by impinger plates. Each plate is washed by a
dedicated spray nozzle.
The differential velocities are achieved at a low absolute velocity
as the gas enters the conical restrictions. Because the gas phase is
compressible it accelerates more rapidly in the cones than the solid or
liquid particles which are accelerated by drag forces from the gas. When
the gas exits the conical restriction and impinges on the plate, the
streamlines of the gas cross those of the droplets and the solid
particles and again produce a high contact efficiency.
As indicated in the process history section this technology has been
commercialized as a low cost high efficiency control device. This success
with large scale units, plus field tests of the process chemistry, provides
strong indications that the demonstrated success of the pilot plant
scrubber can be duplicated in commercial units.
1220
-------
HETEROGENEOUS REACTOR
\
FIGURE
1221
-------
DRY FGD AND PARTICULATE CONTROL
SYSTEMS
By
K. A. Moore
R. D. Oldenkamp
Rockwell International
Energy Systems Group
M. P. Schreyer
Wheelabrator-Frye, Inc.
D. W. Belcher
Stork Bowen Engineering, Inc.
1222
-------
DRY FGD AND PARTICIPATE CONTROL SYSTEMS
INTRODUCTION
For the past 8 years, Rockwell International has been developing advanced
systems for flue gas desulfurization. The regenerative Aqueous Carbonate Process
for Flue Gas Desulfurization* uses a spray dryer as a flue gas contactor and
generates dry reaction products which are collected in a particulate removal
device, regenerated chemically and reused. A simplification of the process,
wherein the regeneration system is not used, has also found acceptance as the
Two Stage "dry scrubbing" Process.** This process was developed jointly by
Rockwell and Wheelabrator-Frye and tests have shown the dry, two-stage system's
capability to remove both S02 and particulates to levels which meet existing and
proposed environmental standards. Recent tests have shown that a variety of
alkali compounds can be used in solution or slurry form in a spray dryer contac-
tor to remove SOg from boiler flue gases. The dry particulates that leave the
spray dryer, such as fly ash, alkali sulfates, sulfites, and unreacted alkali,
are removed from the flue gas by a fabric filter and disposed of as a dry powder.
Tests have shown the dry, two-stage system's capability to remove both SOp
and particulates to levels which meet existing and proposed environmental stan-
dards.
This past year, three "dry scrubbing" systems, totaling 1450 MW of capacity,
have been contracted for. Each system features different alkali injection and
product collection concepts to achieve dry scrubbing. This paper proposes to
review the basic concepts and design features of the three different systems.
Major emphasis will be given to the Rockwell/Wheelabrator Two Stage Process.
*U.S. Patent 3,932,587
**Patent Pending
1223
-------
DRY FGD PROCESS DESCRIPTION
The process of SCL removal utilizing spray dryers is inherently simple; see
Figure 1. The process features the wet contact of S02 in flue gas with a fine
mist of water containing a solution or slurry of alkali, typically either lime or
soda ash. The SCL is absorbed in the mist water droplet and neutralized by the
alkali. With boiler flue gas nominal temperatures of 300°F, the small quantity
of water (about 0.3 gal per 1000 acf) in the alkali mist is completely evaporated.
Thus, the flue gas is not saturated and remains warm. The exit gas temperature
from the spray dryer is 150° to 200°F, a temperature safely above the water dew
point. This technique results in particles of dry, spent alkali which have
captured the SO^. Essentially all of the particulate, including the fly ash, is
then removed in product collection equipment (fabric filter or electrostatic
precipitator). S02 removal efficiencies of 85% or more are achievable with
particle emissions to the stack of less than 0.03 #/mBtu (-0.01 grain/scfS).
There are many benefits to keeping the flue gas warm and dry. The materials
of construction can be of carbon steel with minimum risk of erosion, corrosion,
or scaling. Spray dryer designs prevent wetting the walls of the vessel, and the
dry alkali powder provides a renewable protective dust coating on chamber and
duct surfaces. The flue gas is warm and dry at the exit of the product collec-
tion device; thus, reheat is not normally required to achieve plume dispersal out
of the stack. The ID fan, downstream of the product collector, operates in a
clean, low-temperature environment.
Water chemistry and PH controls typical of wet scrubbing systems are elim-
inated. The spray dryer exit temperature is the preferred primary signal for
control of total water feed to the system. The alkali feed concentration is
determined by the SOp emission requirements.
Additionally, the resulting dry product eliminates the disposal problem of
sludge ponds. The product is a free-flowing material that can be returned to the
coal mine or another pit for final disposal.
1224
-------
SYSTEM DESCRIPTION
t-o
r-o
TWO STAGE DRY FGD SYSTEM
Flow Diagram Key:
Scrubbing Solution
or Slurry
Spent Dry Salts
Cleaned Air
Scrubbing
Solution or Slurry
Fabric Filter
or
Electrostatic
Precipitator
Combustion Air
gt vJ!5p«nt. Dry Sifts;'.j,•»
Alkali Storage
Tank
Dry Product
Disposal
ID.1 Fan
78 M30 54-7 A
Figure 1. Two Stage Dry FGD System
-------
DRY SCRUBBING SYSTEMS SUMMARY
In 1978, three large contracts totaling 1450 MW were awarded to suppliers
of dry-type FGD systems. These contracts were all for North Dakota lignite coal
applications. A comparison of the three different scrubbing concepts selected
for these plants provides an interesting review of the development of spray
drying technology as a means of SCL removal and also illustrates the trends in
particulate control technology.
1) Coyote Unit 1, 410 MW (1,890,000 acfm) features four spray dryer
chambers arranged in parallel, each sized for about one-half million
acfm. Each spray dryer chamber incorporates 3 centrifugal atomizers,
consisting of 150 hp motors driving step-up gear boxes turning a disk
or wheel. The high speed of the rotating disk, 18,000 rpm, centrifu-
gal ly atomizes the solution of water and alkali. This resulting mist
is injected in cross-current flow to the flue gas entering from the
top of the chamber and distributed by vane rings around each atomizer
(see Photo Figure 2.) Each chamber is furnished with a standby
atomizer. A multicompartment fabric filter, utilizing a combination
mechanical and pneumatic cleaning cycle, collects the dry product and
fly ash. The Coyote owners elected to use soda ash as the alkali.
Soda ash will be stored as monohydrate crystals in a system with a
minimum of equipment and manpower requirements. This plant is pres-
ently under construction with preoperational testing scheduled for
July 1980, and acceptance tests to be completed June 1981.
2) Antelope Valley Unit 1, 440 MW (2,200,000 acfm) features 5 chambers (4
with a ducted spare) arranged in parallel, each sized for 25% of the
volume. In this design, the spray dryers use one large high-
horsepower rotary atomizer per chamber. For maintenance, the flow is
switched to the spare chamber. A multicompartment fabric filter,
utilizing reverse air cleaning, collects the product. The alkali will
be slaked lime.
1226
-------
Figure 2. Spray Dryer With Three Atomizers
1227
-------
3) Laramie Station Unit 3, 600 MW (2,810,000 acfm) features yet another
design. This "spray dryer" utilizes a large number of two-fluid
nozzles that inject an alkali of slaked lime into a multicompartmented
section of enlarged ducting. The product collection is by electro-
static precipitator.
APPLICATION OF SPRAY DRYER TECHNOLOGY TO FGD
The variation of spray dryer atomization technology demonstrated by these
large systems leads to an examination of the state of the art of the commercial
spray dryer industry and the application of this technology to flue gas
desulfurization.*
Three types of atomizers have been used successfully in the commercial
spray dryer industry. They are:
1) Pressure nozzles (single fluid) in which the feed is atomized through
small orifice nozzles with a high-pressure pump.
2) Two-fluid nozzles in which the feed is atomized through relatively
larger orifice nozzles by the action of a second pressurized fluid. \
3) Centrifugal atomizers in which the feed is atomized by the mechanical
action of a high-speed rotating atomizing device.
Large spray dryers equipped with pressure nozzles usually contain a multi-
plicity of nozzles in a cluster or other array that is designed to cover the
entire cross-sectional area of the chamber with sprays. The incoming gas is
usually distributed over the chamber cross section by one perforated plate.
Liquid pumping pressures may be well above 1000 psi in order to achieve fine
atomization. Tests conducted by Rockwell with pressure nozzles for flue gas
desulfurization indicated that SOg removal performance could not be maintained
when the gas flow was reduced. The implication of this result was that each
atomizing device should be mated with its own gas distribution system if
accceptable performance was to be maintained over a wide range of gas flows.
* The design features of the Coyote FGD System of Rockwell/Wheelabrator are
discussed below.
1228
-------
Since pressure nozzles typically have a small capacity and limited turndown, the
gas distribution problem could become quite complex. Also, pressure nozzles are
not suitable for abrasive slurry service. Thus, this type of atomization was
rejected by Rockwell for the general FGD application.
Large spray dryers equipped with two-fluid nozzles are designed in the same
fashion as described above for pressure nozzles. However, two-fluid nozzles
have inherently greater capacity than pressure nozzles, and the relatively
larger orifice is somewhat less susceptible to abrasion. The energy required
for equivalent atomization is typically 50 to 100% greater than that for pres-
sure nozzle or centrifugal atomization. Tests conducted by Rockwell with two-
fluid nozzles for the FGD application indicated that it could be an operable
technique if gas distribution could be optimized and maintained over the
expected range of boiler loads, and if performance deterioration from nozzle
erosion could be minimized. Even with solutions to these problems, the energy
penalty tends to make two-fluid nozzles less desirable than the centrifugal
concepts for the general FGD application.
Large spray dryers equipped with centrifugal atomizers have been used in
the spray dryer industry for most slurry services. A single, high-capacity
atomizer is matched to a "vane ring" gas distribution device. The trend has been
to develop increased capacity for single atomizers rather than to build dryers
with multiple atomizers. This trend is specific to the normal spray drying
applications where gas flow is minimized and feed rate is maximized. Also, many
applications require specific product characteristics that could not be assured
if multiple atomizers were employed. There is no technical reason that multiple
centrifugal atomizers cannot be used in a single large chamber for the FGD
application. In fact, Masters (of Niro Atomizer) in his book Spray Drying
states that, "Although one atomizer unit is used per drying chamber, there is no
reason why multi-atomizer units cannot be applied in very large drying chambers.
Where multi-atomizer units are used, drying air is supplied around each atomizer
wheel*"
1229
-------
One of the largest spray drying chambers ever built (~50-ft diameter)
contained multi-atomizers of the centrifugal type. It was built in 1930 by
Bowen Engineering for a potato drying application where product particle size
was not a key criteria. It operated successfully. More recently, Bowen (now
Stork-Bowen) and Rockwell have tested three centrifugal atomizers in a single,
7-ft-diameter pilot spray dryer. This provided a severe test of the concept
since maximum spray pattern overlap was to be expected in the small chamber.
The essential experimental results were:
1) For fixed L/G ratios, three atomizers with three vane rings gave
equivalent S02 removal (~90%) as that observed when a single, larger
atomizer was tested in the same chamber.
\
2) When one^f the three atomizers was shut off and the L/G was preserved
by increasing fluid flow to the two remaining in operation, acceptable
S$2 removal was maintained (-89%).
3) Even when two of the three atomizers were shut off (again preserving
the L/G), the S02 removal dropped to about 82%. In all cases, the gas
continued to flow through the three separate vane rings.
It would be expected that SOp removal performance would be affected some-
what mo-re if three larger atomizers were tested in the same manner in a large
chamber. However, if the gas were diverted from a nonoperational atomizer to
those which were in operation, no serious performance deterioration should be
observed. In fact, agglomeration of the droplets will result in larger product
particle size which will be a positive factor in its collectability.
The key for FGD applications is to provide adequate gas distribution to
each of the operating atomizers. If this is accomplished, there should be no
concern about spray pattern overlap. Of course, adequate residence time must be
provided to ensure that the agglomerated droplets are fully dried.
Recent flow model tests (1/16 scale) of the Coyote FGD system have provided
excellent indications of the operability of the gas distribution system designed
for four 46-ft-diameter chambers with three atomizers per chamber. Under all
1230
-------
gas flow conditions from 25 to 100% of design flow, the gas distributed equally
to each of the four chambers. Specifically, the individual chamber flows were
25 ±2% of the total system flow over the entire flow range. The distribution of
gas to each of the three atomizer vane rings in the four chambers was found to
be 33 ±2% of the flow to that chamber. This again was tested over the full flow
range. The results of this testing should not be surprising, since it is
relatively easy to divide the rectangular gas inlet duct of a Stork-Bowen spray
dryer into three equal cross-sectional areas.
It should also be noted that, although standard spray dryers have been
built with centrifugal atomizers of 400 hp and above, no dryers are currently in
service which handle gas volumes of greater than about 200,000 acfm (at dryer
outlet conditions). Thus, the use of a single centrifugal atomizer in an F6D
application with 400,000 acfm (again dryer outlet conditions) or greater gas
flows constitutes a significant departure from proven gas distribution (atomizer-
vane ring) spray dryer technology. Conversely, the use of three centrifugal
atomizers, each handling 1/3 of the gas flow, is within the experience level of
atomizer-vane ring design.
Considering all of the above factors, tests, concepts, and experience, the
Rockwell-Wheelabrator Joint Venture and Stork-Bowen have designed the Coyote FGD
system with the following features:
1), Each of the four 46-ft chambers will contain three 150-hp atomizers
with matching vane ring gas distributors. In the event of an atomizer
malfunction, the feed can be diverted to the two operating atomizers
to sustain SOp removal performance. The atomizer motors are specif-
ically oversized to allow this option.
2) All four spray dryer chambers are connected in parallel for total gas
flow. Separate ducts and dampers are provided to each atomizer.
Therefore, gas flow control is available to assure high S02 removal
levels for the range of boiler loads.
1231
-------
3) A spare atomizer ts provided for each chamber to permit rapid replace-
ment in the event of malfunction or atomizer removal for maintenance
purposes. Therefore, no spare chamber is required to ensure that the
FGD system is fully available for service.
4) The chambers have been sized with at least a 50% greater residence
time than that proven by 9 years of pilot test experience. This will
protect against inadequate drying because of possible agglomeration.
5) Atomizer position in the large chambers was specifically designed to
preclude impingement of wet droplets on the chamber walls. This
prevents chamber wall buildup.
PARTICIPATE CONTROL
The proposed new source performance standards for particulates are being
significantly tightened. Figure 3 reveals the impact the requirement will have
on equipment suppliers who must design the particulate control equipment to
exceed typical NSPS requirements by a comfortable margin. For many fuels, high
resistivity fly ash precludes the economic use of an electrostatic precipitator.
For the Coyote plant, a multicompartment fabric filter will provide reliable
particulate control. Figure 3 indicates the results of typical fabric filter
performance during pilot tests with lignite for the Coyote program.
As fabric filters for power boilers are quite large, a number of factors
need to be considered to achieve economical design and operation. The effect of
the spray dryer upstream of the fabric filter is to reduce the gas volume
(reduced temperature) but increase the dust load.
The increased dust load requires a positive approach to bag cleaning. For
Coyote, the fabric filter will use a combination mechanical shake and reverse
air cleaning cycle to remove the heavy filter cake and operate at a minimum
system pressure drop. In pilot tests conducted by WFI, the combination cleaning
proved superior to reverse air only.
1232
-------
0.040
-5 0.030
3
w
i
UJ
LU 0.020
I-
U
CC
0.010
0.000
I
EPA PRESENT LIMIT
(0.1 lb/106 Btu)
,EPA PROPOSED LIMIT
(0.03lb/106Btu)
WHEELABRATOR FRYE ROCKWELL
AVERAGE EMISSION
FROM PILOT UNIT
Btu/lb COAL
I
| 6,000
TYPICAL
LIGNITE
8,000 i
TYPICAL
WESTERN COAL
10,000 f12,000
TYPICAL
BITUMINOUS COAL
78-JU7-81-6B
Figure 3. Particulate Emission Control Requirements
1233
-------
The lowered flue gas temperature provided by the upstream spray dryer
allows the fabric filter to be smaller and to utilize a long-lived acrylic or
polyester bag fabric instead of fiberglass. These features of fabric filter
design result in fewer filter compartments, and fewer, less expensive bags.
SUMMARY
A reliable FGD system must combine simplicity with redundancy. The use of
a multihead spray dryer in combination with a fabric filter fits this descrip-
tion. System availability is insured because of the ability to maintain the
equipment "on-line" without affecting the boiler operation. The result is an
SOp and particulate control system that provides the simplest, most reliable
solution to emission control on the market today.
1234
-------
ATTENDEES
Flue Gas Desulfurization Symposium
Las Vegas, Nevada
March 5-8, 1979
Ln
Abdulsattar
Ahrams
Acbenbach
Achtner
Adams
Adams
Adams
Agee
Ahman
Aikawa
Albrandt
Alfredsson
Alger
All
Almand
Ando
Andresen
Andrew
Angelovich
Ansari
Ardell
Arras
Ashley
Auger
Avinash
Ayer
Uabcock
Bachman
Bacskai
Badger
Baer
Bailey
Bailey
Bains
Balakrishnan
Ball
Bambrough
Banchik
Barber
-Barber
Barnes
Bartlett
Basciani
Bass, Jr.
Batra
Bauman
Beachler
Beals
Beckman
Beeler
Abdul Hameed
Jack
D.
S.
R. L.
Radford C.
Robert A.
James L.
Stefan
H.
Alan
Per-Olo
John R. M.
Sy
Charles F.
Jumpei
R. L.
Bob
John M.
Armjad H.
Marilyn
Karlheinz
Michael J.
Robert
P.
Franklin A.
John E.
Gerald G.
Ronald J.
Bill
Richard E.
Albert E.
Ann
C. S.
N. S.
Myron L.
H. A.
N.
G. D.
Walter C.
Bruce
Terry W.
Kenneth A.
Loren 0.
Sushil K.
Robert D.
David S.
Joseph L.
Eugene B.
Linda
50 Beale Street
50 Beale Street
One Penn Plaza
One Penn Plaza
600 Grant Street
201 N. Roxboro Street
77 Havemeyer Lane
1200 6th Avenue
S-35187 VASJO
1300 Park Place Building
5900 E. 37th Avenue
AB Svensha Flaktfabriken
Building 2-447
1000 E. Main Street
6303 Barfield Road, 3te. 219
1-13 Kasuga, Bunkyo-Ku
1545 Wilshire
300 Civic Drive
920 SW 6th Avenue
209 E. Washington Avenue
2970 Maria Avenue
Gervinusstrasse 17/19 6000
George Street Parade
125 Jamison Lane
11499 Chester Road
P. 0. Box 12194
P. 0. Box 220
1623 Harney Street
115 Gibraltar Road
300 Civic Drive
600 N. Highland Avenue
33 Sproul Road
35 South Jefferson Road
400 East Sibley Blvd.
4233 N. United Parkway
400 Rouser Road
351 Boulevard
3701 Kirby Drive
10 South Riverside Plaza
Mail Drop-10
Avon Refinery
10 UOP Plaza
200 S. Michigan
20 Turnpike Road
MD-12
MD-12
2540 S. 27th Avenue
180 Knoll Road
P. 0. Box 87
San Francisco
San Francisco
New York
New York
Pittsburgh
Durham
Stamford
Seattle
Seattle
Denver
Stockholm
Schenectady
Plainfield
Altanta
Tokyo
Los Angeles
Pleasant Hill
Portland
Jackson
Northbrook
Frankfurt/Main
Birmingham, B31QQ
Monroeville
Cincinnati
Res. Tri. Park
Austin
Omaha
Horsham
Pleasant Hill
Aurora
Frazer
Wliippany
Harvey
Schiller Park
Pittsburgh
St. Joseph, Hull
Houston
Chicago
Res. Tri. Park
Martinez
Atlanta
Des Plaines
Chicago
Westbo rough
Res. Tri. Park
Res. Tri. Park
Broadview
Boonton
Knoxville
CA
CA
NY
NY
PA
NC
CT
WA
SWEDEN
WA
CO
SWEDEN
NY
IN
GA
JAPAN
CA
CA
OR
MI
IT,
GERMANY
ENGLAND
PA
OH
NC
TX
NB
PA
CA
IL
PA
NJ
IL
IL
PA
Quebec
TX
IL
NC
CA
GA
IL
IL
MA
NC
NC
IL
NJ
TN
94119
94105
10001
10001
15219
27701
06904
98101
98101
80207
12345
46168
30328
112
90017
94524
97204
49201
60062
15146
45246
27709
78767
68104
19044
94524
60507
19355
07981
60426
60176
15108
CANADA
77098
60606
27711
94553
30247
60016
60604
01581
27711
27709
60153
07005
37901
Bechtel National Inc.
Bechtel National, Inc.
Chemico Air Poll. Control
Chemico Air Poll. Control
Wheelabrator-Frye Inc., APCD
TRW Inc.
Dorr-Oliver, Inc.
W. States Reg. Coordinator, 208 Program
Flakt Industrial Division
.Chiyoda International Corporation
Public Service Company
General Electric Company
Public Service Indiana
Evans L. Shuff & Associates
Chuo University
Babcock & Wilcox
Industrial Clean Air
Pacific Power & Light Company
GiIbert/Commonwealth
Mcllvaine Company
Lurgi Umwelt und Chemotechnik GmbH
Lodge Cottrell Ltd.
U.S. Steel Corporation
PEDCo Environmental Inc.
Research Triangle Institute
Lower Colorado River Authority
Omaha Public Power Distri.
IU Conversion Systems, Inc.
Industrial Clean Air
Barber-Greene Company
A-S-H PUMP Div. Envirotech Corp.
Apollo Chemical Corp.
ARCO Petroleum Products Co.
Environeering, Inc.
Envirotech Corp.
Federal Gov't.-Dept. of Environment
King Wilkinson Inc.
Arthur G. McKee & Company
U.S. EPA
TOSCO
The CADRE Corporation
UOP - Air Correction Division
Pullman Power Products
New England Power Service Co.
U.S. EPA
Northrop Services, Inc./APTI
Wallace & Tiernan Di.v./Pennwalt Corp.
Foster Wheeler Energy Corporation
Carborundum Company
-------
LO
Behie
Belcher
Berst
Bhatti
Biedell
Bielawski
Bierbower
Binkley
Biondo
Birkner
Bitsko
Blair
Blau
Blinckmann
Bloss
Blythe
Bond
Bordrick
Borgwardt
Borio
Borsare
Bowen
Bowling
Boyd
Boy da
Braden
Brandt
Breunig
Briggs
Broaddus
Brown
Brown
Brown
Browne
Broz
Bruce
Bryson
Bryson
Buckingham
Budin
Bumpass
Burbank
Burford
Burgess
Burns
Busko
Bussell
Buzak
Byrne
Cameron
R.
S. W.
Donald W.
A. H.
Mohammed
Edward L.
G. T.
R. G.
M. J.
Samuel J.
V. B.
Ralph
Gary T.
Ted
R.
H. Edward
William B.
George
Duane
Robert H.
D. C.
David
George
Chester H.
D.
Robert B.
Herbert H.
Robert C.
Bill
H.
W. R. "Pat"
Charles S.
Ralph
W. S.
N. B.
Larry D.
R. B.
George M.
Kevin R.
Paul A.
Michael
Tom
Dewey A.
David P.
Robert J.
Eugene A.
William D.
Harvey
Jan
John M.
Sidney M.
900-One Palliser Square
P. 0. Box 598
P. 0. Box 2206
1300 Greenway Plaza E.
P. 0. Box 1500
20 S. Van Buren Avenue
P. 0. Box 1139R
P. 0. Box 226226
1100 Independence Avenue
P. 0. Drawer 5000
2301 Market Street
P. 0. Box 932
Black Horse Lane
One Penn Plaza
700 N. 6th Street
576 Azalea Road, Ste 124
420 Rouser
9420 Telstar Avenue
,MD-65
1000 Prospect Hill Road
1000 Prospect Hill
85 Research Road
515 S. Flower St., Ste. 863
P. 0. Box 87
6200 Oak Tree Boulevard
P. 0. Box 1500
P. 0. Box 47837
2000 Market Street
1 River Road
P. 0. Box 1603
161 E. 42nd Street
1129 Bellwood Avenue
650 Smithfield Street
One Oliver Plaza
3203 Womens Club Drive, Suite 220
Mississauga
P. 0. Box 908
3 Executive Campus
3333 Michelson Drive
Brook Road
P. 0. Box 7444
50 Beale Street
800 Shades Creek Parkway
P. 0. Box 529100
P. 0. Box 1620
P. 0. Box 1975
Bldg. 23, Rm. 352, 1 River Road
1111 No. 19-th Street
200 West Monroe Street
P. 0. Box 2389
Calgary, Alberta
Somerville
Birmingham
Houston
Somerville
Barberton
Horristown
Dallas
Washington
Lakeland
Philadelphia
Elkhart
North Brunswick
New York
Levanon
Mobile
Coraopolis
El Monte
Res. Tri. Park
Windsor
Windsor
Hingham
Los Angeles
Knoxville
Independence
Somerville
Atlanta
Philadelphia
Schenectady
Dalton
New York
Bellwood
Pittsburgh
Pittsburgh
Raleigh
Ontario
Bloomington
Cherry Hill
Irvine
Conshohocken
Longview
San Francisco
Birmingham
Miami
La Jolla
Baltimore
Schenectady
Arlington
Chicago
Knoxville
CANADA
NJ
AL
TX
NJ
OH
NJ
TX
DC
FL
PA
IN
NJ
NY
PA
AL
PA
CA
NC
CT
CT
MA
CA
TN
OH
NJ
GA
PA
NY
GA
NY
IL
PA
PA
NC
CANADA
IN
NJ
CA
PA
TX
CA
AL
FL
CA
MD
NY
VA
IL
TN
T2G086 Montreal Engineering Company, Ltd.
08876 Stork Bowen Engineering Inc.
35201 Zurn Industries
77046 Pullman Kellogg
.08876 Research-Cottrell
44203 Babcock & Wilcox Company
07960 Allied Chemical
75266 Glitsch, Inc.
20545 Department of Energy
33803 Davy Powergas, Inc.
19101 Philadelphia Electric Company
46515 Miles Laboratories, Inc.
08902 Ted Blau Associates
10001 Chemico Air Poll. Control
17042 Envirotech
36609 Envirotech Corporation
15108 Envirotech Corporation
91731 South Coast Air Qual. Mgmt. Distr.
27711 U.S. EPA, - IERL-RTP
06095 Combustion Engineering Inc.
06095 Combustion Engineering
02043 Martek Inc.
90071 Atlantic Richfield
37901 Carborundum Company
44130 Arthur G. McKee & Company
08876 Research-Cottrell, Inc.
30362 The Cadre Corporation
19103 FMC-ALKALI
12345 General Electric
30720 Dalton Rock Products Company
10017 Koch Engineering Co., Inc.
60104 Faville-LeVally Corporation
15222 Dravo Lime Company
15222 Dravo Corporation
27612 Acurex Corporation
L5K 1B3 Ontario Research Foundation
47401 Hoosier Energy Division
08034 Stone & Webster Eng. Corp.
92715 Fluor ESC
19428 Trane Thermal Company
75602 Texas Eastman Company
94119 Bechtel National Inc.
35209 Southern Company Services
33169 Florida Power & Light
92038 Systems, Science & Software
21203 Eastern Stainless Steel Company
12345 General Electric Company
22209 Energy and Environmental Analysis, Inc.
60606 Fluor Power Services, Inc.
37901 American Limestone Company
-------
Campbell
Canterbury
Capp
Carden
Ca rnow
Carpenter
Casada
Casey
Cavanagh
Caylor
Cesped
Chapman
Chatham
Chatlynne
Cheng
Cherry
Chopra
Christman
Chu
Clark
Clasen
Claussen
Cleveland
Cline
Coe, Jr.
Collier
Colyn
Conklin
Cook
Cooley
Couppis
Cowen
Crandall
Cranston
Craton
Crocker
Crowe
Crucq
Cubisino
Cukor
Culhane
Cwik
Daggett
Dahiem
Dalton
Davis
Davis
Davis
De Masi
De Priest
Ivor.E. 150 E. Broad Street
John A. 1800 FMC Drive, West
Joseph A. 555 Madison Avenue
W. 420 Rouser Road
Bertram W. 2121 W. Taylor
John K. P. 0. Box 149
Donald A. ORNL Box Y, 9104-1
John W. P. 0. Box 1380
J. Gordon P. 0. Drawer 5000
F. 434 Allegheny River Blvd.
Ric 300 Lakeside Drive
Richard A. 2118 Milvia Street
Russell L. P. 0. Box 35000
C. J. MD-61
Gregory H. 4233 N. United Parkway
Millard W. 100 W. Walnut Street
P. One Penn Plaza
Roger C. 4701 Sangamore Road
Richard R. P. 0. Box 3
Louise B. P. 0. Box 10087
Donald D. 1300 Park Place Building
Robert W. 2410 East Busch Boulevard
Lee 115 Gibraltar Road
John R. 7903 Westpark Drive
E. L. P. 0. Box 2744
Clark W. P. 0. Box 173
Clarence A. 12076 Grant Street
Edwin R. 666 Fifth Avenue
Judith E. MD-60
William J. 309 W. Washington Street
Evis C. 400 Prudential Plaza
Robert 85 Research Road
W. A. 260 Cherry Hill Road
George 835 Hope Street
Robert E. 6161 Savoy
Laird St. Joe Zinc Co. Power Station
Robert 26 W. St. Clair Street
C. Vlissingen
Art 433 Hackensack Avenue
Peter 2118 Milvia Street
F. R. 600 Grant Street
R. A. 5618 W. Montrose Avenue
Harry 23480 Park Sorrento
F. E. 215 Central Avenue
Stuart M. 3412 Hillview Avenue
Richard L. 2701 Koppers Building
Robert 555 Madison Avenue
S. A. 600 Grant Street
Cosimo 220 W. Sixth Street
William 20 S. Van Buren Avenue
Columbus
Itasca
New York
Coraopolis
Chicago
St. Louis
Oak Ridge
Houston
Lakeland
Oakmont
Oakland
Berkeley
Houston
Res. Tri. Park
Schiller Park
Pasadena
New York
Washington
Houston
Palo Alto
Seattle
Tampa
Uorsham
McLean
Los Angeles
Kansas City
Thornton
New York
Res. Tri. Park
Chicago
Denver
Hingham
Parsippany
Stamford
Houston
Monaca
Cincinnati
NETHERLANDS
Hackensack
Berkeley
Pittsburgh
Chicago
Calabasas
Louisville
Palo Alto
Pittsburgh
New York
Pittsburgh
Tucson
Barberton
on
II,
NY
PA
IL
MO
TN
TX
FL
PA
CA
CA
TX
NC
IL
CA
NY
DC
TX
CA
WA
FL
PA
VA
CA
MO
CO
NY
NC
IL
CO
MA
NJ
CT
TX
PA
Oil
NJ
CA
PA
II,
CA
KY
CA
PA
NY
PA
AR
OH
43215 Clyde Williams & Company
60143 FMC Corporation
10022 Combustion Equipment Associates, Inc.
15108 Envirotech Corporation
60612 University of Illinois Medical Center
63166 Union Electric Company
37830 Oak Ridge National Labs.
77001 Shell Dev. Company
38303 Davy Powergas, Inc.
15139
94612 Kaiser Engineers Power Corporation
94704 Teknekron
77035 Fluor E & C Inc.
27711 US EPA, IERL-RTP
60176 Environeering, Inc.
91006 Ralph M. Parsons
10001 Envirotech/Chemico
20016 Teknekron, Inc.
77001 Brown & Root, Inc.
94303 Research-Cottrell, Inc.
98101 Chiyoda International Corporation
33612 Seminole Electric Cooperative, Inc.
19044 IU Conversion Systems, Inc.
22102 Engineering-Science
90051 Joy Manufacturing Company
64141 Burns & McDonnell Engineering Co.
80241 Tri-State G&T
10019 Pullman Power Products
27711 U.S. EPA-IERL
60657 111. State-Inst.Natural Resources
80265 R. W. Beck and Associates
02043 Martek Inc.
07054 GPU Service Corporation
06909 Peabody Process Systems Inc.
77036 Davy Powergas Inc.
15061 U.S. Bureau of Mines
45268 U.S. EPA
Royal Schelde
07950 Pullman Kellogg Company
94704 Teknekron
15219 Wheelabrator-Frye Inc., APCD
60634 Linden Equipment Company
91304 Neptune AirPOL
40277 American Air Filter Company
94303 Electric Power Research Institute
15219 Koppers Company, Inc.
10022 Combustion Equipment Associates Inc.
15230 U.S. Steel Corporation
85702 Tucson Gas & Electric
44203 Babcock & Wilcox Company
-------
N>
00
00
Deacon J. 420 Rouser Road
Dempsey J. Herbert P. 0. Box 12796
Dene Charles E. P. 0. Box 2000
Devitt Timothy W. 11499 Chester Road
Dharmarajan N. N. 200 N. 7th Street
Di Gioia, Jr. A. M. 570 Beatty Road
Dickerman Jim 8500 Shoal Creek Blvd., Box 9948
Dietrich David C. 100 N.E. Adams Street
Dietrich Gary N. WH-562, 401 M Street, S.W.
DiPol Chester V. 8900 DeSoto
Dixon Kathleen E. 14000 Georgia Avenue
Dollmeyer Clarence H. 1200 N.W. 63rd, Box 10400
Donahue Bernard A. P. 0. Box 4005
Donoghoe Jim 33 Industry Avenue
Doody Calvin N. P. 0. Box 36444
Doty Robert W. 600 Grant St., 42nd Floor
Doty William S. 1501 Alcoa Building
Dowdy David A. 799 N. Main Street
Downs William 1562 Beeson Street
Doyle William J. Speed Scientific School
Dragos John Consol Plaza
Duffy Ron 835 Hope Street
Dunkle David 115 Gibraltar Road
Durkee Kenneth R. MD-13
Dutton Roger W. P. 0. Box 8405
Earl C. B. P. 0. Drawer 5000
Eaton Terry J. P. 0. Box 211
Ebrey John M. 601 Jefferson, 27th Floor
Ebzery Joan 1231 25th N.W., Bur.Nat'l.Affairs
Efimenkp Alex 1740 W. Adams St.-Bur.Air Quality
Eisenlohr, Sr. D. H. Two Rector
Elder Henry W. OSWHA
Elliott David 480 University Avenue
Enck T. W. 10 South Riverside Plaza
Engdahl Richard 505 King Avenue
Eriksen Robert L. 1717 E. Interstate Avenue
Erskine George 1820 Dolley Madison Avenue
Esche Michael Sultbachstr. 22, 6600
Fackler Robert H. 818 Kansas Avenue
Fail Albert B. P. 0. Box 476
Farber Paul S. 9700 South Cass Avenue
Farrington James 1500 E. Putnam Avenue
Favell M. A. Box 12121, 555 West Hastings^Street
Featherman Donald P. 0. Box 1500
Feuerborn Dale F. 1330 Baltimore Street
Fife C. H. 29 S. LaSalle Street
Fike Waldo E. 799 N. Main Street
Finer J. 420 Rouser Road
Finkel Lawrence H. 3 Girard Plaza
Fish William H. P. 0. Box 499
Coraopolis
Res. Tri. Park
Paducah
Cincinnati
Lebanon
Monroeville
Austin
Peoria
Washington
Canoga Park
Silver Spring
Oklahoma City
Champaign
Auburn
Houston
Pittsburgh
Pittsburgh
Akron
Alliance
Louisville
Pittsburgh
Stamford
Horsham
Res. Tri. Park
Kansas City
Lakeland
LaCygne
Houston
Washington
Phoenix
New York
Muscle Shoals
Toronto, Ontario
Chicago
Columbus
Bismarck
McLean
Saarbrucfcen
Topeka
Beaver
Argonne
Old Greenwich
Vancouver
Somerville
Kansas City
Chicago
Akron
Coraopolis
Philadelphia
Columbus
PA
NC
KY
OH
PA
PA
TX
IL
DC
CA
HD
OK
IL
ME
TX
PA
PA
OH
OH
KY
PA
CT
PA
NC
MO
FL
KA
TX
DC
AZ
NY
AL
CANADA
IL
OH
ND
VA
GERMANY
KS
PA
IL
CN
BC
NJ
MO
IL
OH
PA
PA
NE
15108
27709
42001
45242
17042
15146
78766
61629
20460
91364
20910
73156
61820
04210
77036
15219
15227
44310
44601
40222
15241
06909
19044
27711
64114
33803
66040
77005
20037
85007
10006
35660
M5G IV2
60606
43201
58501
22102
66612
15009
60439
06870
V6B 4T6
08876
64141
60603
44310
15108
19102
68601
Envirotech Corporation
Acurex Corporation
EPRI - Shawnee Test Facility
PEDCo Environmental, Inc.
Buell ECD-Envirotech
GAI Consultants Inc.
Radian Corporation
Caterpillar Tractor Company
U.S EPA/Office of Solid Waste
Rockwell International-Energy Systems Gp
Automation Industries, Inc.
Benham-Holway Power Group
U.S. Army, CERL
Hamilton & Son
Davy Powergas, Inc.
Eckert, Seamans, Cherin & Mellott
Aluminum Company of America
Brad Associates
Babcock & Wilcox Company r
University of Louisville
Consolidation Coal Company
Peabody Process Systems, Inc.
IU Conversion Systems, Inc.
U.S. EPA-ESED
Black & Veatch, Consulting Engineers
Davy Powergas, Inc.
Kansas City Power & Light
Lodge-Cottrell Operations Div. Dresser Ind
Environment Reporter
Arizona State Health Department
Ebasco Services Inc.
Tennessee Valley Authority
Acres Davy Limited
Arthur G. McKee & Company
Battelle
Basin Electric Power Cooperative
The MITRE Corporation, Metrek Div.
Saarberg-Holter Umwelttechnik GmbH
Kansas Power and Light Company
M-K National Corporation
Argonne National Laboratory
Flakt
British Col. Hydro & Power Authority
Research Cottrell
Kansas City Power & Light Company
Babcock & Wilcox Company
Brad Associates
Envirotech
Suntech, Inc.
Nebraska Public Power District
-------
N>
CO
vo
Fisher Ray W. Energy & Min. Res. Rsch. Inst.
Fitch Robert E. P. 0. Box 3707, Mail Stop 9A-01
Fletcher J. C. 9041 Executive Park Drive
Fling Richard B. 2805 Vists Mesa Drive
Foley Gerry F. 185 Crossways Park Drive
Forck B. Klinkestr. 27-31, D4300
Forrest J. A. 283 Rt. 17 South
Fowler Carolyn P. MD-61
Fox Harvey Box 750
Fox Landon D. 400 Commerce Avenue W10B104
Francis Daniel V. Frankfort Road
Frank Roger J. P. 0. Box 14219
Franklin Alan Lynn P. 0. Box 999
Frees Robert C. 650 Smithfield
Freire Joseph L. 393 Seventh Avenue
Freshcorn Donald E. Box A
Friedrichs G. E. 5001 W. 80th Street
Fuchs Michael J. 576 Standard Avenue
Fuller David A. P. 0. Box 12194
Furlong D. 200 N. 7th Street
Futryk Robert The American Road, Room 626
Gaines Joseph L. 9165 Rumsey Road
Gallagher T. One Penn Plaza
Galloway Edward E. P.O. Box 960
Galloway Marsh J. 140 Sheldon Road
Gambarani P. One Penn Plaza
Garcia Alberto 74 Inverness Drive East
Garner Jim R. P.O. Box 2624
Garvey Bernie 433 Hackensack Avenue
Garvin W. D. P. 0. Box 30013
Gatehouse C. 420 Rouser Road
Gaudette Paul R. 6701 W. 64th Street
Gault John 5916 Fannin Street
Gaumer Lee S. P. 0. Box 538
Gehri Dennis C. 8900 De Soto Avenue
Giberti Richard A. 128 Spring Street
Gilbert Carl A. 650 Smithfield Street
Gilmore Donald B. P. 0. Box 15027
Giovanetti Albert P. 0. Drawer 5000
Giovanni Dan V. P. 0. Box 10412
Glamser John 6161 Sandy Drive
Gleason Robert J. P. 0. Box 1500
Gleason Thomas 470 Park Avenue, South
Glenn Roland D. 50 East 41st Street
Gluck Ted 7760 W. Devon Avenue
Goenner Roger K. 576 Standard Avenue
Gogineni M. R. 1000 Prospect HilJ Road
Golden Dean 3412 Hillview Drive
Goldschmidt Klaus Bismarckstr. 54
Goodwin R. W. P. 0. Box 10087
Ames
Seattle
Knoxville
Rancho PalosVerdes
Woodbury
Essen
Paramus
Res. Tri. Park
Somerville
Knoxville
Monaca
Phoenix
Richland
Pittsburgh
New York
Monaca
Minneapolis
Richmond
Res. Tri. Park
Lebanon
Dearborn
Columbia
New York
Cincinnati
Berea
New York
Englewood
Birmingham
Hackensack
Raleigh
Coraopolis
Shawnee Mission
Houston
Allentown
Canoga Park
Lexington
Pittsburgh
Las Vegas
Lakeland
Palo Alto
Houston
Somerville
New York
New York
Chi cago
Richmond
Windsor
Palo Alto
Essen
Palo Alto
TA
WA
TO
CA
NY
GERMANY
NJ
NC
NJ
TN
PA
AZ
WA
PA
NY
PA
MN
CA
NC
PA
HI
MD
NY
OH
OH
NY
CA
AL
NJ
NC
PA
KS
TX
PA
CA
MA
PA
NV
FL
CA
TX
NJ
NY
NY
IL
VA
CT
CA
GERMANY
CA
50011
98124
37919
90274
11797
07652
27711
08805
37902
15061
85063
99352
15222
10001
15061
55437
94802
27709
17042
48121
21045
10001
45201
44017
10001
80112
35202
07950
27612
15108
66202
77004
18105
91304
02173
15222
89114
33803
94303
77036
08876
10016
10017
60631
94804
06095
94303
D-4300
94303
Iowa State University
Boeing Eng'g. and Construction Co.
United Engineers & Constructors
Til? Aerospace Corporation
Burns & Roe, Inc.
VGB Tech Ver.der Grosskraftwerks.E.V.
Burns & Roe, Inc.
U.S. EPA-IERL
Research-Cottrell, Inc.
Tennessee Valley Authority
ARCO/Polymers
Hogan Manufacturing, Inc.
Battelle Northwest
Dravo Lime Company
Gibbs & Hill, Inc.
St. Joe Minerals Corporation
Combustion Engineering, Inc.
Chevron Research Company
Research Triangle Institute
Envirotech/Buell
Ford Motor Company
Niro Atomizer, Inc.
Chemico Air Poll. Control
Cincinnati Gas & Electric Company
The Ceilcote Company
Envirotech/Chemico
Lear Siegler, Inc. Environ.Tech.Div.
Southern Company Services
Pullman Kellog
Martin Marietta Aggregates
Env'irotech Corporation
Research-Cottrell
Bovay Engineers Inc.
Air Products & Chemicals Inc.
Energy Systems Group-Rockwell Internation
Kennecott Copper Corporation
Dravo Lime Company
EMSL-LV, MSA
Davy Powergas Inc.
Electric Power Research Institute
Davy Powergas
Research-Cottrell, Inc.
Swemco Inc.
Combustion Processes, Inc.
EPCO
Chevron Research Company
Combustion Engineering
Electric Power Research Institute
STEAG Aktiengesellschaft
Research-Cottrell, Inc.
-------
Gosik
Gottlieb
Gottschalk
Granda
Grant
Grassel
Green
Green
Greene
Grimm
Grimm
Groves
Grubb
Grumbrecht
Guay
Gude
Guernsey
Gnllett
Gupta
Gylfe
Hall
Hall
flalpern
Hamersma
Hamm
Hansen
Hargrove
Harris, Jr.
Harrison
Hartman
Hartman
Havlik
Hayashida
Hayes, Jr.
Heacock, Jr.
Head
Headley
Healey
Hein
Heinz
Heinze
Henderson
Hendron
Hennico
Hentges
Herlihy
Hess
Hettler
Hickok
Hill
Robert Denver Federal Center
Myron ECT Division
Chris P. 0. Box 2900, Shawnee Test Fac.
M. R. P. 0. Box 3
Richard J. 607 E. Adams Street, Rm. 721
Eugene E. P. 0. Box 1299
Clois L. P. 0. Box 472
Robert M. P. 0. Box 880
Jack II. MD-60
Carlton D. 40 East Broadway
Richard P. 4500 Cherry Creek Drive
Kenneth 0. 2020 Bldg., Abbott Road
Theron 525 S. Hayden Road
Volker P.O. Box 1949/1960 - 5270
Bill 44 Briar Ridgeroad
Klaus 9165 Riimsey Road
Edwin 0.
David E. 100 Summer Street
Pat 2 Country View Road
J. Donald 890 DeSoto Avenue
John 175 E- 5th Street
William G. Walden Avenue
Mark 1900 Pennsylvania Avenue
J. Warren 1 Space Park
Jim 1000 S. Fremont Avenue
Svend Kris 9165 Rumsey Road
Buddy 8500 Shoal Creek Blvd., Box 9948
William B. Bldg. 2, Rm. 132B1, River Road
Doug 800 Kipling Avenue
John R. 640 S. Main Street
Scott 11499 Chester Road
Frank A. P. 0. Box K-7
Paul K. P. 0. Box 400
Richard D. 1501 Alcoa Building
Frank A. P. 0. Box 2166
Harlan N. 50 Beale Street
Larry C. P. 0. Box 880
Joseph R. 835 Hope Street
Glen 300 W. Washington
Elwood 433 Hackensack Avenue
Ka.y 420 Rouser Road
W. B. 200 N. 7th Street
David M. 55 E. Monroe.Street
A. "Boite.Postale 311
Robert A. 6105 Center Hill Road
Jim 401 M Street, S.W.
H. F. 1800 FMC Drive, West
Robert F. 2800 Mitchell Drive
Wayne 3316 W. 66th Street
David Lee 44 Briar Ridge Road
Denver
Washington
Paducah
Houston
Springfield
Minneapolis
Rockdale
Addison
Res. Tri. Park
Butte
Denver
Midland
Tempe
Gummersbach 1
Danbury
Columbia
Paulsboro
Boston
Malvern
Canoga Park
St. Paul
Buffalo
Washington
Redondo Beach
Alhambra
Columbia
Austin
Schenectady
Toronto, Ont.
West Covina
Cincinnati
Richmond
Denver
Pittsburgh
Phoenix
San Francisco
Morgantown
Stamford
Chicago
Hackensack
Coroapolis
Lebanon
Chicago
Rueil -Ma Imaison
Cincinnati
Washington
Itasca
Walnut Creek
Minneapolis
Danbury
CO
DC
KY
TX
IL
MN
TX
TX
NC
MT
CO
MI
AZ
W.GERMANY
CT
MD
NJ
MA
PA
CA
MN
NY
DC
CA
CA
MD
TX
NY
CANADA
CA
OH
VA
CO
PA
AZ
CA
WV
CT
IL
NJ
PA
PA
IL
FRANCE
OH
DC
IL
CA
MN
CT
80005
20545
42001
77001
62701
55440
76567
75001
27711
59701
80217
48640
85282
06810
21045
08066
02110
19355
91304
55101
20068
90278
91802
21045
78766
12345
M8Z 5S4
91790
45246
23288
80201
15219
85036
94119
26505
06909
60606
07645
15108
17042
60044
92506
45224
20460
60143
94598
55435
06810
U.S. EPA
Department of Energy
Tennessee Valley Authority
Brown & Root Inc.
Central 111. Public Service Company
Donaldson Company, Inc.
Alcoa
Metal Components
U.S. EPA
Montana Power Company
Stearns-Roger, Inc.
Dow Chemical Company
Fabric Filters
L & C Steinmuller
Newmont Exploration
Niro Atomizer Inc.
Mobil Research & Development Corp.
United Engineers and Constructors
Ecolaire Systems, Inc.
Rockwell International
Burlington Northern Inc.
ANDCO
Potomac Electric Power Company
TRW Systems and Energy
CF Braun
Niro Atomizer
Radian Corporation
General Electric Company
Ontario Hydro
Wallace & Tiernan
PEDCo Environmental, Inc.
Infilco Degremont, Inc.
Mitsubishi International Corporation
Alcoa
Arizona Public Service Company
Bechtel National Inc.
U.S. Department of Energy
Peabody Process Systems, Inc.
Marblehead Lime Company
Pullman Kellogg
Envirotech
Envirotech/Buell
Woodward-Clyde Consultants
Institut Francais du Petrole
Procter & Gamble Company
U.S. EPA
FMC Corporation
Dow Chemical Company
Cooperative Power Association
Newmont Exploration Limited
-------
N>
Hill
Hilton
Hintz
Hoffman
Hojnicki
Hollett, Jr.
Hoilinden
Hong
Hooker
Hormand
Horn
Horn
Horowitz
Horwitz
Hosemann
Howell
Hpyne
Hsieh
Hunt
Hunter
Hurst
Hurwitz
Hushen
Hutta
lannelli
Ingram
Ireland
Ivask
Jahnig
Jalbert
Janssen
Jashnani
Jensen
Jensen
Jewell
Johannes
Johnson
Johnson
Johnson
Johnson
Jones
Jones
Jones
Jones
Jones
Judersleben
Jurgensen
Juris
Kameoka
Kamrath
Russell Nard
Robert
Charles F.
Frank W.
Tom
Grant T.
G. A.
Sun-nan
Neal
Svend
Nelson E.
Richard J.
A. L.
II.
Robert P.
R. D.
T.
C. K.
Joseph M.
W. D.
T. B.
Howard
Thomas M.
Paul J.
Eric A.
T. J.
Paul
S.
Charles E.
Paul E.
Kent E.
I.
Alan
Robert M.
Robert
Richard
Carlton
Howard
Robert A.
Russell R.
J. R.
Jerry L.
Julian W.
Marion
Robert D.
Peter
H. D.
Kenneth
Y.
Jim
G.
D.
P. 0. Box 908
115 Gibraltar Road
P. 0. Box 2825
P. 0. Box 269
9700 Argonne National Labs.
7N015 York Road
470 CUBE
P. 0. Box 538
4282
Rumsey Road
2100 Clearwater Drive
Two Countryview Road
6105 Center Hill Road
One Penn Plaza
77 Beale Street
100 W. Walnut Street
420 Rouser Road
P. 0. Box 800
Research Center
P. 0. Box 1139R
20 S. Van Buren Avenue
One Penn Plaza
1880 Republic Avenue
Somerton Road
100 Holland Avenue
P. 0. Box 1498
P. 0. Box 5888
One Penn Plaza
P. 0. Box 8
591 Poquonnock Road
1717 E. Interstate Avenue
1450 S. Rolling Road
105 W. Adams
50 Beale Street
1700 West Loop South, Suite 1145
P. 0. Box 173
835 Hope Street
361 East Broad Street
1000 Fremont Avenue
P. 0. Box 101
P. 0. Box 47127
333 Ravenswood Avenue
MD-61
P. 0. Box 908
P. 0. Box 538
P. 0. Box 1600
400 Rouser Road
3890 Carman Road
1300 Park Place Building
1300 Larkspur
Bloomington IN 47401 Hoosier Energy Division
Horsham PA 19044 IU Conversion Systems, Inc.
Allentown PA Mbsser/Ecolaire
Springfield MA 01101 Calfran Industries
Argonne IL 60439 Argonne National Labs.
Bensenville IL 60106 Flick-Reedy Corporation
Chattanooga TN 37401 Tennessee Valley Authority
Allentown PA 18105 Air Products & Chemicals, Inc.
N. Canton OH 44720 Babcock & Wilcox
Columbia MD 21045 Niro Atomizer
Oak Brook IL 60540 Combustion Engineering Inc.
Malvern PA 19355 Ecolaire Systems Inc.
Cincinnati OH 45224 Procter & Gamble
New York NY 10001 Envirotech/Chemico
San Francisco CA 94106 Pacific Gas & Electric Co.
Pasadena CA 91124 Ralph M. Parsons Company
Coraopolis PA 15108 Envirotech Corporation
Rosemead CA 91770
Brackenridge PA 15014 Allegheny Ludlum Steel Corporation
Morristown NJ 07960 Allied Chemical Corporation
Barberton OH 44203 Babcock & Wilcox Company
New York NY 10001 Chemico Air Pollution Control Corp.
San Leandro CA 94577 Andco Power Industry Products, Inc.
Trevose PA 19047 Betz Labs. Inc.
Peapack NJ 07977 Komline-Sanderson
Reading PA 19603 Gilbert Associates, Inc.
Denver CO 80217 Stearns Roger
New York NY 10001 Chemico Air Poll. Control
Linden NJ 07036 Exxon Research & Engineering Co.
Groton CT 06340 Proto-Power Management Corporation
Bismarck ND 58501 Basin Electric Power Cooperative
Baltimore MD 21227 Martin Marietta Corp.
Chicago IL 60603 Foster Wheeler Energy Corporation
San Francisco CA 94105 Bechtel Power Corporation
Houston TX 77027 Pullman Power Products
Kansas City MO 64141 Burns & McDonnell
Stamford CT 06909 Peabody Process Systems, Inc.
Columbus OH 43212 Ohio EPA
Alhambra CA 91802 C. F. Braun & Company
Florham Park NJ 07932 Exxon Research & Engineering
Dallas TX 75247 Gifford-Hill & Company, Inc.
Henlo Park CA 94025 SRI International
Res. Tri. Park NC 27711 U.S. EPA-IERL
Bloomington IN 47401 Hoosier Energy Division
Allentown PA 18105 Air Products & Chemicals Inc.
Somerville NJ 08876 Research-Cottrell, Inc.
Coraopolis PA 15108 Envirotech
Schenectady NY 12303 New York Power Pool
Seattle WA 98101 Chiyoda International Corporation
Austin TX 78758 Texas Air Control
-------
to
•c-
to
Rang
Kaplan
Kaplan
Kaplan
Kasik
Kaye
Keen
Kendle
Kent
Kesler
Kesstner
Kettner
Kiff
Kilgroe
Kincaid
Kindig
Klemovich
Knefelkamp
Knight
Knowles
Koba
Kodras
Kolbe
Kondo
Konig
Korman
Koshkin
Kovach
Koval
Krause
Krekels
Kreuziger
Kronenberger
Krumme
Kuo
Kuzela
Lagarias
Lamb
Lambert
Langeland
Lanois
Laseke
Lasky, Jr.
Laslo
Laughlin
Lawson
Lawson
Layman
LeBoeuf
Lee
Cecilia C. 450 East Edsall Boulevard
Marilyn 2970 Maria Avenue
Norman MD-61
Steven M. 9165 Rumsey Road
Lawrence A. 2000 Second Avenue
Ron 835 Hope Street
R. T. P. 0. Box 140232
James R. 1800 FMC Drive, West
Raymond P. 0. Box 1500
Rick P. 0. Box 16067
Mark 35 South Jefferson Road
Don Power Production
John 4565 Coco Boulevard
James D. MD-61
Jim 6100 Center Hill Road
James K. 4601 Indiana Street
Ronald M. P. 0. Box 173
Robert A. 3800 Race Street
R. Gordon 4301 Dutch Ridge Road
W. D. P. 0. Box 3105
S. 1300 Park Place Building
Frank D. 1500 Market Street
John L. 1450 S. Rolling Road
Ken 925 S. Niagara Street
Robert J. P. 0. Box 300
Samuel 632 W. 125 Street
M. One Penn Plaza
John J. Collins Ferry Road
T. P. 0. Box 87
Robert E. 955 Mearns Road
Y. T. 57 Engelen Kampstraat
Wolfgang Stauffenbergstrase 26,
Robert W. 20 S. Van Buren
J. Lee 2555 Cumberland Parkway, N.W.
Wen L. 12400 E. Imperial Highway
Ed 24 Perimeter Center East
John S. P. 0. Box 23210
Jack F. P. 0. Box 189
Steve L. 920 S. W. 6yh Avenue
Wes 85 Research Road
G. D. 1500 East Putnam Avenue
Bernie A. 11499 Chester Road, Chester Tower
Bernard A. 11499 Chester Road
Dennis P. 0. Box 1107
James H. P. 0. Box 12796
Clifton P. 0. Box 9538
P. 700 University Avenue FC M5G1X6
George 0. P. 0. Box 1151
Gene 3913 Algoma Road
George C.Y. 50 Beale Street
Palisades Park
Northbrook
Res. Tri. Park
Columbia
Detroit
Stamford
Charlotte
Itasca
Somerville
Denver
Whippany
Elk River
Los Angeles
Res. Tri. Park
Cincinnati
Golden
Kansas City
Denver
Beaver
Houston
Seattle
Philadelphia
Baltimore
Denve r
Tulsa
New York
New York
Morgantown
Knoxville
Warminster
Sittard 6131 J.E.
Berlin W 30
Barberton
Atlanta
Norwalk
Atlanta
Oakland
Addison
Portland
Hingham
Old Greenwich
Cincinnati
Cincinnati
Darien
Res. Tri. Park
Madison
Toronto
Pensacola
Green Bay
San Francisco
NJ
IL
NC
MD
MI
CT
NC
IL
NJ
CO
NJ
MN
CA
NC
OH
CO
MO
CO
PA
TX
WA
PA
MD
CO
OK
NY
NY
WV
TN
PA
07650
60062
27711
21045
48226
06909
28224
60143
08876
80216
07981
5330
90039
27711
45231
80401
64141
80216
15009
77001
98101
19102
21227
80224
74102
10027
1000 1
16505
37901
18974
NETHERLANDS
W.GERMANY
OH
GA
CA
GA
CA
TX
OR
MA
CT
OH
OH
CT
NC
WI
Ontaria
FL
WI
CA
44203
30339
90650
30348
94623
76001
97204
02043
06870
45246
45246
06840
27709
53579
CANADA
32520
54301
94119
Catalysis Research Corporation
Mcllvaine Company
U.S. EPA, IERL-RTP
Niro Atomizer Inc.
Detroit Edison Company
Peabody Process Systems Inc.
Catalytic, Inc.
FMC Corporation
Research-Cottrell
Mine and Smelter
Apollo Chemical Corp.
United Power Association
Joy Manufacturing (WP Div.)
U.S. EPA-IERL
Procter & Gamble
Hazen Research, Inc.
Burns & McDonnell
Mine & Smelter Corporation
Michael Baker, Jr., Inc.
Shell Oil Company
Chiyoda International Corporation
Catalytic, Inc.
Martin Marietta Labs.
Mitsubishi International Corporation
Cities Service Company
Columbia University
Chemico Air Poll. Control
Department of Energy
Carborundum Company
Pennwalt Corp.-Sharpies-Stokes Div.
NEOM B.V.
Bewag-Berliner Kraft-und Licht AG
Babcock & Wilcox Company
Vinings Chemical Company
Bechtel
UOP Air Correction Division
Kaiser Engineers, Inc.
Forney Engineering Company
Pacific Power & Light Company
Martek Inc.
Flakt, Inc.
PEDCo Environmental, Inc.
PEDCo Environmental, Inc.
Air Correction Div., UOP
Acurex Corporation
Warzyn Engineering, Inc.
Ontario Hydro
Gulf Power Company
Feeco International Inc.
Bechtel National Inc.
-------
ro
*-
(-0
Lee John R. P. 0. Box 1010
Leffmann Warren W. 6120 Westover Drive
Legatski L. Karl FHC Drive, West
Legier William B. Prudential Center
Lehtola Philip R. 29801 Euclid Avenue
Leitman J. D. P. 0. Box 87
Leivo Charles P. 0. Box 19566
Levis James J. 2200 Churchill Street
Levy, Jr. Newton 2 Executive Plaza
Lewis C. J. P. 0. Box 15453
Liegois William A. Stanley Building
Lillestolen Tom 1500 E- Putnam
Lingle Jim 231 W. Michigan Street
Link F. William 591 Harquette Mall
Lisiewski Alfred A. Two Country View Road
Lisk Ian 1301 South Grove
Lloyd Charles G. 1415 Rollins Road
Long M. E. 600 Grant Street
Longfellow R. L. 257 Geneva Drive
Loquercio Peter 309 West Washington
Lorfing Rick 51 Corporate Woods ,
Lowell Philip S. 4107 Medical Parkway, #214
Luce Ken 4565 Colorado Blvd.
Lucy Thomas H. P. 0. Box 1500
Lundy Terry P. 0. Box 230
Lunt Richard R. 20 Acorn Park
Lutz Stephen J. 201 N. Roxboro
Macaskill D. 55 Avenida De Orinda
MacDonald Bruce 85 Research Road
MacRae Terry Two Country View Road
Maddalone Ray One Space Park
Madenburg Richard S. P. 0. Box 7808
Madonia John J. 55 E. Monroe
Magtoto Artemio R. P. 0. Box 1139R
Majdeski H. M. 20 So. Van Buren Avenue
Makar John E. 1800 FMC Drive, .West
Maley L. B. 4550 W. 109th Street
Malki Kal 1000 Prospect Hill
Mann Earl L. 5265 Hohroan Avenue
Marbry R. F. 10 South Riverside Plaza
Harder Sidney 2058 Huntleigh Road
Mardirossian Aris P. 0. Box 5691
Martin C. E. P. 0. Box 1595B
Martin C. K. P. 0. Box 173
Martin James R. 1000 Prospect Hill Road
Mason Louis 145 Cedar Lane
Mason Thomas 0. General Motors Tech. Ctr.
Matoi H. J. P. 0. Box 217
Maxwell Michael A. MD-61
Mayfield John R. 2005 Wai den Avenue
Tuscaloosa AL 35401 B. F. Goodrich Engineered Systems Co.
Oakland CA 94611 Bechtel Power Corporation
Itasca IL 60143 FMC Corporation
Boston MA 02199 Charles T. Main, Inc.
Wickliffe OH 44121 Bailey Controls Company
Knoxville TN 37901 Carborundum Company
Irvine CA 92713 Dresser Industries
Springfield IL 62706 Illinois Environ.Prot.Agency
Hunt Valley MD 21030 Martin Marietta Chemical
Lakewood CO 80215 National Lime Association
Muscatine IA 52761 Stanley Consultants
Old Greenwich CN 06870 Flakt Inc.
Milwaukee WI 53201 Wisconsin Electric Power Company
Michigan City IN 46360 Northern Indiana Public Service Co.
Malvern PA 19355 Ecolaire Systems Inc.
Barrington IL 60025 Technical Publishing Company
Burlingame CA 94010 Sharpies-Stokes Div., Pennwalt Corp.
Pittsburgh PA 15219 Rockwell International/Joint Venture
Aliquippa PA 15001 Dravo Lime Company
Chicago IL 60606 111.Institute Natural Resources
Overland Park KS 66210 Olin Water Services
Austin TX 78756 P. S. Lowell & Co., Inc.
Los Angeles CA 91006 Joy Manufacturing Company
Somerville NJ 08876 Research-Cottrell
Las Vegas NV 89151 Nevada Power Company
Cambridge MA 02140 Arthur D. Little, Inc.
Durham NC 27701 TRW Environmental Engineering
Orinda CA 94563 Macaskill Associates
Hingham MA 02043 Martek Inc.
Malvern PA 19355 Ecolaire Systems, Inc.
Redondo Beach CA 90278 TRW Inc.
Boise ID 83729 Morrison-Knudsen Company, Inc.
Chicago IL 60603 Sargent & Lundy
Morristown NJ 07960 Allied Chemical
Barberton OH 44203 Babcock & Wilcox Company
Itasca IL 60143 FMC Corporation
Overland Park KS 66210 Babcock & Wilcox
Windsor CT 06095 Combustion Engineering
Hammond IN 46325 Northern Indiana Public Service Co.
Chicago II. 60606 Arthur G. McKee & Company
Springfield IL 62704 Harder and Associates
Derwood MD 20855
Indianapolis IN 46206 Indianapolis Power & Light Company
Kansas City MO 04141 Burns & McDonnell
Windsor CT 06095 Combustion Engineering, Inc.
Englewood NJ 07631 Neptune Airpol Inc.
Warren MI 48090 General Motors Corporation
Fontana CA 92335 Kaiser Steel Corporation
Res. Tri. Park NC 27711 U.S. EPA, IERL-RTP
Buffalo NY 14240 Abdco Power Industry Pro,
-------
N3
McCarthy
McCormick
McCurdy
McDowell III
McFarlane
McGlamery
Mcllvaine
Meadows
Mehta
Hehta
Meinig
Merdes
Merlet
Merrill
Messing
Meyers
Meyler
Michels
Michener
Miller
Miller
Miller
Miller
Minnella
Mirchandani
Mobley
Mohn
Moody
Moody
Morasky
Morelli
Moser
Mullen
Murad
Muren
Murphy
Murray
Musgrove
Mutsakis
Naeve
Nakabayashi
Naumann
Nelson
Ness
Newby
Newhams
Newman
Nguyen
Nguyen
Nicholas
Carol L. 8500 Shoal Creek Blvd.
C. J. 650 Smithfield Street
Wayne 200 Massachusetts Avenue, N.W.
Robert V. 800 King Street
Lloyd R. 12631E Imperial Highway, Suite 107B
Gerald G. OSWHA
Robert 2970 Maria
Michael .L. P. 0. Box 8405
Dhirendra C. 555 Madison Avenue
R. E. 555 Madison Avenue
John P. 0. Box 10087
R. P. 0. Box 87
Heinz Gervinusstrasse 17/19 6000
Richard S. 485 Clyde Avenue
Aubrey F. 1271 Avenue of the Americas
James E. 2000 Second Avenue
J. A. 4565 Colorado Blvd.
Harold T. 1 New York Plaza
A. W. Columbia Road
Bruce A. 1 Broadway
Dick 115 Gibraltar Road
E. Mack 380 Civic Drive
Michael W. P. 0. Box 24407
Thomas J. 2964 LBJ Freeway
T. M. 300 Lakeside Drive
J. David MD-61
Nancy C. 1000 Prospect Hill Road
Harvey C.
Ron 111 Windsor Drive
Thomas M. 34)2 Hillview Avenue
Mark H. 105 South Meridian
Robert P. 0. Box 3822
Hugh 115 Gibraltar Road
Fred Y. 555 Madison Avenue
E. J. 1211 W. 22nd Street
Kenneth R. Bldg. 2, Rm. 710, 1 River Road
Daniel N. P. 0. Box 1440
John 520 South Post Oak
Michael 161 East 42nd Street
Steve P. 0. Box 1700
Yasuyuki Thermal Power Department
C. E. 300 W.. Washington Street
T. R. 607 East Adams Street
Harvey M. P. 0. Box 8213, University Station
Richard A. 1310 Beulah Road
Thomas 835 Hope Street
Carl L. Two Country View Road
Thuyet Due P. 0. Box 3
Xuan T. Trans. Can. Highway
George W. 1550 Northwest Highway
Austin
Pittsburgh
Washington
Wilmington
Santa Fe Springs
Muscle Shoals
Northbrook
Kansas City
New York
New York
Palo Alto
Knoxville
Frankfurt/Ma in
Mountain View
New York
Detroit
Los Angeles
New York
Morristown
Cambridge
Horsham
Pleasant Hill
Ft. Lauderdale
Dallas
Oakland
Res. Tri. Pk.
Windsor
Port Edwards
Oak Brook
Palo Alto
Indianapolis
San Francisco
Horsham
New York
Oak Brook
Scbenectady
Erie
Houston
New Yor
Houston
Tokyo
Chicago
Springfield
Grand Forks
Pittsburgh
Stamford
Malverne
Houston
Senneville, Quebec
Park Ridge
TX
PA
DC
DE
CA
AL
IL
MO
NY
NY
CA
TN
GERMANY
CA
NY
MI
CA
NY
NJ
MA
PA
CA
FL
TX
CA
NC
CT
WI
IL
CA
IN
CA
PA
NY
IL
NY
PA
TX
NY
TX
JAPAN
IL
IL
ND
PA
CT
PA
TX
CANADA
IL
78766
15222
20545
19899
90670
35660
60062
64114
10022
10022
94303
37901
94042
10020
48226
90039
10004
07960
02142
19044
94523
33307
75234
94623
27711
06095
54469
60067
94303
46225
94119
19044
10022
60521
12345
16533
77027
10017
77001
60606
62701
58202
15235
06907
19355
77001
HQX 3L7
60068
Radian Corporation
Dravo Lime Company
Department of Energy
Delmarva Power & Light Company
Ecolaire Inc.
Tennessee Valley Authority
Mcllvaine Company
Black & Veatch
Combustion Equipment Associates
Combustion Equipment Associates, Inc.
Research-Cottrell, Inc.
Carborundum Company
Lurgi Umwelt und Chemotechnik GmbH
Acurex Corporation
Empire State Electric Energy Res. Corp.
The Detroit Edison Company
Joy Manufacturing Company
International Nickel Co., Inc.
Allied Chemical Corporation
Badger America Inc.
IU Conversion Systems, Inc.
Industrial Clean Air
Parkson Corporation
UOP/Air Correction Division
Kaiser Engineers Power Corporation
U.S. EPA, IERL-RTP
Combustion Engineering Corp.
Nekoosa Papers Inc.
Sharpies Centrifuges
Electric Power Research Institute
Amax Coal Company
Brown & Root, Inc.
ID Conversion Systems, Inc.
Combustion Equipment Associates, Inc.
Research-Cottrell, Inc.
General Electric Company
Hammermill Paper Company
Bechtel Power Corporation
Koch Engineering Company
Houston Lighting & Power Company
Electric Power Development Co., Ltd.
Marblehead Lime Company
Central 111. Public Service Company
Department of Energy
Westinghouse Electric Corporation
Peabody Process Systems, Inc.
Ecolaire Systems Inc.
Brown & Root, Inc.
Domtar Inc. Research Centre
Dames & Moore
-------
to
**
in
Nissen
Nixon
Nixon
Nodiff
Norton
Novack
Novak"~
O'Brien
O'Brien
O'Brien, Jr.
Oestemeyer
Oguchi
Okazawa
Oldenkarap
O'Leary
Oliver
Olsson
Omohnndro
Ongemach
Onnen
Orem
Osborne
Ostroff
Oven, Jr.
Ozol
Parikh
Parikh
Parish
Parke
Parker
Pannley
Parr
Parsons
Patkar
Patten
Patton
Pendergraft
Pepper
Peters
Peters
Petrie
Petrou
Petti
Phelan
Phillips
Piasecki
Pickering
Pierce
Pircon
Pitcher
William I. 1600 E. First Street
David C. 30 W. Superior Street
W. Robert 77 Beale Street
Marvin J. 325 W. Adams Street
Richard D. P. 0. Box 173
Robert 85 Research Road
R. J. 5509 Tarrywood Court
~JoeT C" 415 Power Building
William E. OSWKA
A. W. P. 0. Box 15W~S
Kenneth W. 4809 Todd Avenue
Tomoyoshi 700 S. Flower Street
K. 1 Houston Center Ste. #1810
Richard D. 8900 De Soto Avenue
Norman F. 115 Gibraltar Road
Earl D. 333 Ravenswood Avenue
Lars 1500 E. Putnam Avenue
George A. 1800 FMC Drive, West
Ken P. 0. Box 111
James H. 6702 Hollow Tree Road
Sidney R. 700 N. Fairfax Street, Suite 304
Michael C. MD-61
Norman 835 Hope Street
Hamilton S. 2600 Blair Stone Road
Michael A. 1450 S. Rolling Road
K. N. P. 0. Box 1500
Lax 2845 Clearview Place
Helen R. 4614.Ramsey
John 115 Gibraltar Road
C. E. 420 Rouser Road
Randy D. 4614 Ramsey
Steve 12076 Grant Street
Lloyd J. P. 0. Box 2825
Avi 11499 Chester Road
Thomas W. 125 Baker Street
Richard W. 115 Gibraltar Road
Lynn K. MD-61
Wesley W. P. 0. Box 111
H. J. 600 Grant Street
Warren D. MD-61
Jim r P. 0. Box 227
Gus P- 0. Box 880
V. 600 Grant Street
John H. 600 Grant Street
James B. P. 0. Box 2180, Dresser Tower
E. J. General Motors Tech. Ctr.
H. C. 110 S. Orange Avenue
Gary G. CNO. 27
L. J. 10 South Riverside Plaza
Norman D. 2 North 9th Street
Salt Lake City
Duluth
San Francisco
Springfield
Kansas City
Hingham
Raleigh
Chattanooga
Muscle Shoals
SomeWille
E. Chicago
Los Angeles
Houston
Canoga Park
Ilorsham
Menlo Park
Greenwich
Itasca
Tampa
Louisville
Alexandria
Res. Tri. Park
Stamford
Tallahassee
Baltimore
Some rvi lie
Atlanta
Austin
Horsham
Coraopolis
Austin
Thornton
Bethlehem
Cincinnati
Costa Mesa
Ilorsham
Res. Tri. Park
Los Angeles
Pittsburgh
Res. Tri. Park
Waterflow
Addison
Pittsburgh
Pittsburgh
Houston
Warren
Livingston
Trenton
Chicago
Allentown
UT
MN
CA
IL
MO
MA
NC
TN
AL
NJ
IN
CA
TX
CA
PA
CA
CT
IL
FL
KY
VA
NC
CT
FL
MD
NJ
GA
TX
PA
PA
TX
CO
PA
OH
CA
PA
NC
CA
PA
NC
NM
TX
PA
PA
TX
MI
NJ
NJ
IL
PA
84112 U.S. Bureau of Nines
55802 Minnesota Power & Light Company
94106 Pacific Gas & Electric Co.
62706 111.Institute of Natural Resources
64141 Burns & tfcDonnel
02043 Hartek Inc.
27609 Energy Resources
37401 Tennessee Valley Authority
35660 Tennessee Valley Authority
08876 Research-Cottrell
46312 Graver Energy Systems, Inc.
90017 Sumitomo Metal America Inc.
77002" Japan Trade Center
91304 Rockwell International
19044 IU Conversion Systems, Inc.
94025 SRI International
06830 Flakt Inc.
60143 FMC Corporation
33601 Tampa Electric Company
40228 "American Air Filter Company
22314 Industrial Gas Cleaning Institute
27711 U.S. EPA, IERL-RTP
06907 Peabody Process Systems Inc.
32301 Florida Dept. of Env. Regulation
21227 Martin Marietta Laboratories'
08,876 Research Cottrell, Inc.
33092 The CADRE Corporation
78750 University of Texas
19044 IU Conversion Systems, Inc.
15108 Envirotech Corporation
78756 University of Texas
80241 Tri-State GST
18001 Mosser/Ecolaire
45246 PEDCo Environmental, Inc.
92626 M. C. Patten & Co., Inc.
19044 IU Conversion Systems, Inc.
27711 U.S. EPA-IERL
90051 Los Angeles Dept. of Water and Power
15219 Wheelabrator-Frye Inc., APCD
27711 U.S. EPA, IERL-RTP
87421 Public Service Company of Hew Mexico
75001 Metal Components
15219 Wheelabrator-Frye Inc., APCD
15219 Wheelabrator-Frye
77001 The Carter Oil Company
48090 General Motors Corporation
07039 Foster Wheeler
08625 NJ Depart, of .Environ. Protection
60606 Arthur G. McKee & Company
18101 Pennsylvania Power & Light
-------
N)
-P-
CT\
Pless
Plumley
Plyler
Plys
Ponder
Ponder
Pope
Potterton
Powell
Preston
Princiotta
Prodesky
Provol
Pullen
Pullman
Pursell
Quackenbush
Rabb
Raben
Ramirez
Rao
Rautzen
Ray
Reichard
Reid
Reilly
Reinauer
Reisinger
Remillieux
Renberg
Renko
Retz
Reynolds
Rhodes
Rhudy
Ricci
Richardson
Richman
Richmond
Rieland
Riggs
Ring
Ritchie
Robbins
Rochelle
Roe
Rogers
Rogers
Rohlik
Rollins
L.
Don
A. L.
Everett
A. G.
Thomas
Wade H.
Kenneth S.
S. T.
James M.
George T.
Frank T.
E. J.
Steve J.
T. K.
D.
L. A.
V. C.
Dave
Irwin A.
Agustin A.
Richard
Robert
William G.
Herman
John C.
John B.
Thomas V.
A. A.
Jean
W.
Ronald
John A.
Karen Anthony
Robin B.
Richard
Hugh
Phillip M.
Mark
Philip F.
William G.
Keith A.
Leon E.
Charles I.
Stephen M.
Gary
S. F.
Kenneth J.
Wyatt
Ron
K. B.
P. 0. Box 111
1000 Prospect Hill Road
MD-61
400 East Sibley Blvd.
11499 Chester Road
MD-61
520 S. Post Oak
P. 0. Box 2423
Stanley Building
P. 0. Box 10412
401 M Street, S.W.
20 S. Van Buren Avenue
P. 0. Box 1380
110 Sutler Street
200 N. 7th Street
101 S. Wacker Drive
Centre Square West-1500 Market Street
P. 0. Box 2900, Shawnee Steam Plant
44 Montgomery St., Suite 4220
1800 FMC Drive, West
P. 0. Box 1500
P. 0. Box 1123
235 East 42nd Street
1 Moritime Plaza
P. 0. Box 8405
555 Madison Avenue
10 Chatham Road
600 Grant Street
19 Avenue Dubonnet
420 Rouser Road
P. 0. Box 372
P. 0. Box 1975
2600 Blair Stone Road
Andover Road
3412 Hillview
201 S. Fall Street, Capitol Complex
1007 Market Street
P. 0. Box 1500
P. 0. Box 127
1800 Washington Road
12301 Kurland
400 Commerce Avenue
2475 E. 22nd Street, Suite 510
P. 0. Box 3
Dept. of Chemical Engineering
P. 0. &ox 6428
555 Madison Avenue
7100 Broadway #3D
P. 0. Box 197X, Rt. 1
P. 0. Box 47320
Tampa
Windsor
Res. Tri. Park
Harvey
Cincinnati
Res. Tri. Park
Houston
N. Canton
Muscatine
Palo Alto
Washington
Barberton
Houston
San Francisco
Lebanon
Chicago
Philadelphia
Paducah
San Francisco
Itasca
Somerville
Dayton
New York
San Francisco
Kansas City
New York
Summit
Pittsburgh
Courbevoie 92400
Coraopolis
Wellsville
Baltimore
Tallahassee
Wellsville
Palo Alto
Carson City
Wilmington
Somerville
Center
Pittsburgh
Houston
Knoxville
Cleveland
Houston
Austin
Ft. Myers
New York
Denver
Bakersfield
Dallas
FL
CT
NC
II,
OH
NC
TX
OH
IA
CA
DC
OH
TX
CA
PA
IL
PA
KY
CA
IL
NJ
Oil
NY
CA
HO
NY
NJ
PA
FRANCE
PA
NY
MD
FL
NY
CA
NV
DE
NJ
ND
PA
TX
TN
OH
TX
TX
FL
NY
CO
CA
TX
33601
06095
27711
60426
45246
27711
77001
44720
52761
94303
20460
44203
77063
94104
17042
60606
19102
42001
94104
60143
08876
45401
10017
94111
64114
10022
07901
15219
15108
14895
21203
32301
14895
94303
89710
19898
08876
58530
15241
77034
37902
44115
77001
78712
33904
10022
80221
93308
75247
Tampa Electric Company
Combustion Engineering, Inc.
U.S. EPA, IERL-RTP
ARCO Petroleum Products Co.
PEDCo Environmental, Inc.
U.S. EPA, IERL-RTP
Bechtel Power Corporation
Babcock & Wilcox Company
Stanley Consultants
Electric Power Research Institute
U.S. EPA-OEMT
Babcock & Wilcox Company
Shell Development Company
Ecolaire, Inc.
Envirotech/Bue] 1
United States Gypsum Company
Catalytic, Inc.
Bechtel National Inc.
Combustion Equipment Associates, Inc.
FMC Corporation
Research-Cottrell
Chemineer Inc.
Pfizer Inc.
Combustion Engineering, Inc.
Black & Veatch Consulting Engineers
Combusion Equipment
Mikropul Corporation
Wheelabrator-Frye Inc., APCD
Air Industrie
Envirotech Corp.
C-E Air Preheater Co.
Eastern Stainless Steel Company
Florida Dept. of Env. Regulation
CE Air Preheater
Electric Power Research Institute
Nevada Div. of Environmental Protectioi
E.I. du Pont de Nemours & Co., Inc.
Research-Cottrell Inc.
Minnkota Poer Coop
Consolidation Coal Company
Houston Lighting & Power
Tennessee Valley Authority
Chemico Air Pollution Control Co.
Brown & Root, Inc.
University of Texas
The Hunters Corporation
Combustion Equipment
York Research
Getty Oil Company
Celanese Chemical Co., Inc.
-------
NJ
*^
-J
Rosenberg Harvey S. 505 King Avenue
Ross Dennis R. 76 S. Main Street
Ross Donald W. P. 0. Drawer 5000
Ross R. W. P. 0. Box 1958
Rossoff Jerry 2350 E. F.I Segundo Blvd.
Rubin Edward S. Schenley Park
Ruggiano Lou 115 Gibraltar Road
Rukovena Frank P. 0. Box 350
Sainz Darwin E. 201 S. Broadway
Saleem Abdus One Penn Plaza
Sannes Carl 414 Nicollet Mall
Santhanam C. J. 20 Acorn Park
Santy Myrrl One Space Park
Sargent Donald H. 6621 Electronic Drive
Schaffer John M. 2 Country View Road
Schendel Ronald L. 3333 Michelson Drive
Scher J. P- 0. Box 87
Schorsch Henry 4233 N. United Parkway
Schreyer M. P. 14920 S. Main Street
Schroeder Rick P.- 0. Box 1980
Schwartz Richard 161 E. 42nd Street
Scott W. P. 0. Box 87
Scotti Louis 433 Hackenr.ack Avenue
Scroggins James Edwin P. 0. Box 100
Seabrook, Jr. B. Lawrence 115 Gibraltar Road
Scale William C. Box 220
Selle J. B. 1629 Bonnie Brae
Semrau Konrad 333 Ravenswood Avenue
Senatore Peter J. 235 East 42nd Street
Serdoz Dick Capitol Complex
Sevcik Vaclav J. 2700 S. Cass Avenue
Shafer K. 0. 600 Grant Street
Shah N. D. 3412 Hillview Avenue
Shanks Alfred T. 581 Creamery Road
Sharp John A. Chem.Res.Lab., McMasterville
Sharpe Patricia MD-60
Sherwin R. M. 50 Beale Street
Shimtzu Taku MCEC 118 Tomihisa-Cho
Shimoda Elwyn P. 0. Box 1267
Simko Alexander P. 10025 North 21st Avenue
Slack- A. V. Route 1, Box 69
Slaughter Dale M. Empire State Plaza
Sliger A. Glenn R&D Center, Park 10
Slingsby Donald K. 591.Poquonnock Road
Slocum, Jr. Ernest F. 433 Hackensack Avenue
Smith Craig 324 E. llth Street
Smith Dan B. 999 Touhy Avenue
Smith Earl 0. 1500 Meadowlake
Smith G. H. 33 City Center Drive #358
Smith Gordon L. 1930 Bishop Lane
Columbus
Akron
Lakeland
JHuntington
El Segundo
Pittsburgh
Ho r sham
Akron
Orciitt
New York
Minneapolis
Cambridge
Redondo Beach
Springfield
Malverne
Irvine
Knoxville
Schiller Park
Gardena
Phoenix
New York
Knoxville
Hackensack
Granger
Ilorsham
Austin
.Houston
Menlo Park
New York
Carson City
Argonne
Pittsburgh
Palo Alto
Tel ford
Quebec
Res. Tri. Park
San Francisco
OH
OH
FL
WV
CA
PA
PA
OH
CA
NY
MN
MA
CA
VA
PA
CA
TN
IL
CA
AZ
NY
TN
NJ
WY
PA
TX
TX
CA
NY
NV
IL
PA
CA
PA
CANADA
NC
CA
Ichigaya, Shinjuku-Ku
Ponca City
Phoenix
Sheffield
Albany
Houston
Groton
Hackensack
Kansas City
Des Plaines
Kansas City
Mississauga
Louisville
OK
AZ
AL
NY
TX
CT
NJ
MO
IL
MO
Ontario
KY
43201
44308
33803
25720
90245
15213
19044
44309
93454
10001
55401
02140
90278
22151
19355
92730
37901
60276
90248
85001
10017
37901
07061
82934
19044
78767
77006
94025
10017
89710
60515
15219
94303
18969
J3G1T9
27711
94119
TOKYO
74601
85021
35660
12054
77084
06340
07601
64106
60018
64114
CANADA
40277
Battelle Columbus Division
Ohio Edison Company
Davy Powergas, Inc.
Huntington Alloys, Inc.
Aerospace Corporation
Carnegie-Mellon University
IU Conversion Systems, Inc.
Norton Company/Stowe
Union Oil Company of California
Chemico Air Pollution Control Corp.
Arthur D. Little, Inc.
TRW incorporated
Versar, Inc.
Ecolaire Systems, Inc.
Fluor Engineers and Constructors, Inc.
Carborundum Company
Environeering, Inc.
Wheelabrator-Frye Inc., APCD
Salt River Project
Koch Engineering Company
Carborundum Company
Pullman Kellogg
Texasgulf Inc.
ID Conversion Systems, Inc.
Lower Colorado River Authority
Selle Alloys & Equipment Co.
SRI International
Pfizer Inc.
Nevada^Div. Environmental Protection
Argonne National Laboratory
Wheelabrator-Frye Inc. APCD
Electric Power Research Institute
Fischer E. Porter Company
Canadian Industries Limited
U.S. EPA, IERL-RTP
Bechtel National Inc.
Mitsubishi Heavy Industries, Ltd.
Continental Oil Company
Arizona Public Service Company
SAS Corporation
NY State Energy R&D Authority
Pullman Kellogg
Proto-Power Management Corporation
Pullman Kellogg
U.S. EPA, Region 7
Air Correction Division, UOP Inc.
BJ.ack & Veatch Constr. Engr.
Ecolaire Canada Ltd.
American Air Filter
-------
CO
Smith
Smith
Smith
Smith
Smith
Smith
Smith
Smith
Smithson, Jr.
Snider
Snyder
Sommer
Sood
Spellman
Sperry
Stachura
Stalter
Stanbro
Stange
Staszechy
Statnick
Steeves
Steiner
Stenby
Stengle
Stern
Stevens
Stewart
Stewart
Stewart
Stone
Stout
Stowe
Strakey,
Strauss '
Straw
Strong
Stuparich
Sturtevant
Su
Swahlstedt
Swartz
Swenson
Syler
Takvoryan
Tanner
Tao
Tennyson
Thaxton
Theodore
Jr.
Mark D.
Norman B.
Peter V.
Phil
Roger
Russell K.
Scott
Sidney T.
G. Ray
A. J.
R. Bruce
George A.
A jay
James P.
Larry J.
Stan
Harold C.
William D.
John
F. M.
Robert M.
H. D.
Peter
Edward W.
William F.
Richard D.
Nicholas J.
Dorothy A.
Gerald W.
Merrill J.
Robert E.
Norman D.
Donald H.
Jos.eph P.
Jerome
Harry A.
Erwin R.
Joseph J.
Robert L.
Y. P.
Kim
Russell L.
Donald 0.
Donald E.
Nurhan
George E.
John C.
R. P.
L. A.
Louis
421 South 500 East
Stanley Building
P. 0. Box 1500
P. 0. Box 1500
525 South Hayden Road
505 W. King
P. 0. Box 880
P. 0. Box 173
505 King Avenue
1000 Prospect Hill Road
121 SW Salmon
1800 FMC Drive, West
P. 0. Drawer 2038
Ten UOP Plaza
1029 Corporation Way
8900 De Soto
P. 0. Box 932
Appl.Physics Lab.,Johns Hopkins Rd
393 7th Avenue
50 Beale Street
401 M Street, S. W.
P. 0. Box 299
1.2 Beach Tree Road
P. 0. Box 5888
One Penn Plaza
MD-61
P. 0. Box 1500
3412 Hillview
P. 0. Box 880
709 Cedar Way
P. 0. Box 13138
4500 Cherry Creek
650 Smithfield Street
4800 Forbes Avenue
6621 Electronic Drive
1007 Market Street
P. 0. Box 400
P. 0. Box 1107, Tokeneke Road
P. 0. Box 6,
P. 0. Box Three
300 Liberty Street
P. 0. Box 1318
1500 Meadow Lake
1945 W. Parnall Road
1500 E. Putnam
P. 0. Box 1764
P. 0. Box 538
P. 0. Box 6428
420 Rouser Road
Manhattan College Campus
Salt Lake City
Huscatine
Somerville
Somerville
Tempe
Columbus
Addison
Kansas City
Columbus
Windsor
Portland
Itasca
Pittsburgh
Des Plaines
Palo Alto
Canoga Park
Elkhart
Laurel
New York
San Francisco
Washington
Mahone Bay
Livingston
Denver
New York
Res. Tri. Park
Somerville
Palo Alto
Morgantown
Oakmont
Atlanta
Denver
Pittsburgh
Pittsburgh
Springfield
Wilmington
Naperville
Darien
Solvay
Houston
Peoria
Baltimore
Kansas City
Jackson
Old Greenwich
Houston
Allentown
Fort Myers
Coraopolis
Riverdale
UT
IA
NJ
NJ
AZ
OH
TX
MO
OH
CT
OR
IL
PA
II,
CA
CA
IN
MD
NY
CA
DC
N SCOTIA
NJ
CO
NY
NC
NJ
CA
WV
PA
GA
CO
PA
PA
VA
DE
IL
CT
NY
TX
IL
MD
HO
MI
CN
TX
PA
FL
PA
NY
84102
52761
08876
08876
85281
43201
75001
64141
43201
06095
97204
60143
15230
60016
94303
91304
46515
20810
10001
94119
20460
BOJ2EO
07039
80217
10001
27711
08876
94303
26505
15139
30324
80217
15222
15213
22151
19898
60540
06820
13209
77001
61602
21203
64114
49201
06870
77001
18104
33901
15708
10471
Brock, Easley, Inc.
Stanley Consultants Inc.
Research Cottrell
Research-Cottrell
Fabric Filters
Battelle Columbus Laboratories
Metal Components
Burns & McDonnell
Battelle Columbus Laboratories
Combustion Engineering, Inc.
Portland General Electric Company
FMC Corporation
Gulf Science & Technology Company
Air Correction Div., UOP Inc.
Research-Cottrell, Inc.
Rockwell-Energy Systems Group
Miles Laboratories, Inc.
Johns Hopkins University
Gibbs & Hill Inc.
Bechtel Power Corporation
U.S. EPA - OEMI
Atlantic Bridge Company Limited
Foster Wheeler Dev. Corp.
Stearns-Roger Inc.
Chemico Air Pollution Control
U.S. EPA, IERL-RTP
Research-Cottrell
Electric Power Research Institute
Department of Energy
Chemsteel Construction Co., Inc.
Boiler Equipment
Dravo Lime Company
U.S. Department of Energy
Versar, Inc.
E.I. du Pont de Nemours & Co., Inc.
Amoco Oil Company
Air Correction Division, UOP Inc.
Allied Chemical Corporation
Brown & Root, Inc.
Central Illinois Light Company
Environmental Elements Corporation
Black & Veatch
Consumers Power Company
Flakt Inc.
Gravon Tank & Mfg. Company
Air Products & Chemicals
The Munters Corporation
Envirotech Corporation
Manhattan College
-------
fO
Thompson
Tieman
Toedtman
Tokerud
Toler
Toor
Townsend
Traum
Trautner
Treshler
Tsutsui
Turpin
Tuttle
Twombly
Uehara
Ullrich
Urbanik
Vail
Van der Brugghen
Van Horn
Van Ness
Vasan
Vaughn
Veesaert
Vogelsang, Jr.
Von Bergen
Voss
Walker, Jr.
Walters, Jr.
Wang
Wang
Waters
Webb, Jr.
Webber
Weber
Weber
Wedig ,
Weeter
Weiner .. ,
"WeiE.^ Jr.
-Wfeiss •.
.Wei's*,--*"
,;¥«ilS.
$fes-t, / .
West
White
;WtJ'ifce
.Wfchterman
Wiik
Wicks
Charles C.
John W.
John R.
Arvid
Helen G.
Herbert L.
Joanne
Steven B.
Richard P.
Joseph R.
Koyo
Frank G.
John D.
Robert
Shigeru
Charles R.
Bernard
David L.
F. W.
Andrew J.
Robert P.
Srini
C. F.
Marlin J.
C. W.
Mick
Stepehn C.
Hamilton G.
Richard V.
David S.
Shih-Chung
R. E.
Clifford A.
Emlyn
Gerhard W.
Henry C.
.Christopher P
Dennis W.
Paul E.
Alexander
Kenneth N.
.Xawrence H.
Murray
Brian
Stephen L.
Charles S.
Lowell D.
Rodger
W. D.
Dale A.
9041 Executive Park Drive
350 Hochberg Road
P. 0. Box 652
Ole Degigsv 10
P, 0. Box 12194
Schenley Park
1111 Bonanza
P. 0. Box 22317
2625 Townsgate Road, Suite 360
139 S. Linden
P. 0. Box 400
P. 0. Box 7808
11499 Chester Road
115 Gibraltar Road
118 Tomihisa-Cho Ichigaya
Civil Engineering Department
175 E. 5th Street
100 Summer Street
Utrechtseweg 310
2118 Milvia Street
P. 0. Box 32010
835 Hope Street
P. 0. Box 230
2500 Drilling Service Drive
Engineering Department
2701 Stoughton Road
6320 Augusta Drive
P. 0. Box 1318
P. 0. Box 7808
1102 Q Street
50 Beale Street
700 University Avenue
P. 0. Box 1589
115 Gibraltar Road
Bldg. 3, Empire State Plaza
415 E. 52nd St, Ste. ID
225 Franklin Street
73 73 Perkins Hall-UT
21 West Street
P. 0. Box 800
6601 W. Broad Street
747 Third Avenue
8500 Shoal Creek Blvd., Box 9948
4014 Long Beach Blvd.
700 West State, Div. of Environment
555 Madison Avenue
3422 South 700 West
115 Gibralter Road
29 S. La Salle Street
Stanley Building
Knoxville
Monroeville
Princeton
Oslo 6
Res. Tri. Park
Pittsburgh
Las Vegas
Tampa
Westlake Village
Palatine
Denver
Boise
Cincinnati
Horsham
Shinjuku-Ku
Louisville
St. Paul
Boston
Arnhem
Berkeley
Louisvil le
Stamford
Las Vegas
Maryland Heights
Wilmington
Ma d i s on
Springfield
Baltimore
Boise
Sacramento
San Francisco
Toronto, Ontario
Hattiesburg
Horsham
Albany
New York
Boston
Knoxville
New York
Rosemond
Richmond
New York
Austin
Long Beach
Boise
New York
Salt Lake City
Horsham
Chicago
Muscatine
TN
PA
NJ
NORWAY
NC
PA
NV
FL
CA
IL
CO
ID
OH
PA
TOKYO
KY
MN
MA
37919
15146
08540
27709
15213
89102
33622
91361
60667
80201
83729
45246
19044
40208
55101
02110
NETHERLANDS
CA
KY
CT
NV
MO
DE
Wl
VA
MD
ID
CA
CA
CANADA
MS
PA
NY
NY
MA
TN
NY
CA
VA
NY
TX
CA
ID
NY
UT
PA
IL
IA
94704
40232
06907
89151
63043
19898
53716
22150
21203
83729
95812
94105
FCM5G1X6
39401
19044
12223
10022
02110
37916
10006
91770
23261
10017
78766
90807
83720
10022
84119
19044
60603
52761
United Engineers & Constructors
Bituminous Coal Research, Inc.
Princeton Chemical Research, Inc.
Flakt Norsk Viftefabrikk A/S
Research Triangle Institute
Carnegie-Mellon University
Las Vegas Review-Journal
Badger America, Inc.
Environ. Res. & Technology, Inc.
Air Correction Division, UOP Inc.
Mitsubishi International Corporation
Morrison-Knudsen Company, Inc.
PEDCo Environmental, Inc.
IU Conversion Systems, Inc.
Mitsubishi Heavy Industries, Ltd.
University of Louisville
Burlington Northern Inc.
United Engineers & Constructors
N.V. KEMA
Teknekron
Louisville Gas & Electric Company
Peabody Process Systems
Nevada Power Company
General Aggregate Corporation
E. I. du Pont de Nemours and Co., Inc.
Warman International Inc.
Kulak, Voss & Co., Incorporated
Environmental Elements Corp.
Morrison-Knudsen Company, Inc.
California Air Resources Board
Bechtel National Inc.
Ontario Hydro
S. Mississippi Electric Power Assoc.
IU Conversion Systems, Inc.
NU State Public Service Commission
H & W Management Science
Stone & Webster Engineering Corp.
University of Tennessee
Ebasco Services Inc.
Southern California Edison Company
Reynolds Metals Company
Chem Systems Inc.
Radian Corporation
SCS Engineers Inc.
Bureau of Air Quality
Combustion Equipment Associates
ASARCO Incorporated
IU Conversion Systems, Inc.
Babcock & Wilcox Company
Stanley Consultants, Inc.
-------
Wiedersum George 2301 Market Street
Wilbur John C. P. 0. Box 101
Willett Howard 835 Hope Street
Williams Angela J. 2020 Bldg., Abbott Road
Wilson D. D. 607 East Adams Street
Wolfe Brian A. P. 0. Box 835
Woodyard John 4014 Long Beach Boulevard
Worthington Sidney 401 M Street, S.W.
Wright Charles 525 South Hayden Road
Wunder Pat P. 0. Box 15027
Yagi S. 1300 Park Place Building
Yanagioka Hiroshi 3-13 Moriya-cho Kanagawaku
Yavorsky John 901 Oak Tree Road
Yosick Paul 1930 Bishop Lane
Young Darrell T. 1515 Mineral Square, Box 11299
Zaharchuk Roman P. 0. Box 699
Zarchy Andrew S. 1 River Road
Ziegenhorn Geroge J. 400 East Sibley Blvd.
Ziminski Richard W. 10 Chatham Road
Zuchowski Leon A. 670 Winters Avenue
Zuckerman I. Environ. Engineering Div.
Philadelphia PA 19101 Philadelphia Electric Company
Florham Park NJ 07932 Exxon Research & Engineering Co.
Stamford CT 06909 Peabody Process Systems Inc.
Midland MI 48640 Dow Chemical Company
Springfield IL 62701 Central 111. Public Service Co.
Alliance OH 44601 Babcock & Wilcox
Long Beach CA 90807 SCS Engineers
Washington DC 20460 U.S. EPA
Tempe AZ 85281 Fabric Filters
Las Vegas NV 89114 EMSL-Las Vegas, US. EPA
Seattle WA 98101 Chiyoda International Corporation
Yokohama JAPAN 221 Chiyoda Chemical Eng. & Con.Co.,Ltd.
S. Plainfield- NJ 07080 ASARCO
Louisville KY 40277 American Air Filter
Salt Lake City UT 84147 Kennecott Copper Corporation
Pottstown PA 19464 Firestone Tire Company
Schenectady NY 12301 General Electric
Harvey IL 60426 Atlantic Richfield Company
Summit NJ 07901 MikroPul Corporation
Paramus NJ 07652 Burns and Roe, Inc.
Redondo Beach CA 90178 TRJtf
Cn
o
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/7-79-167b
2.
4. TITLE AND SUBTITLE Proceedings i Symposium on Flue Gas
Desulfurization—Las Vegas, Nevada, March 1979;
Volume IE
3. RECIPIENT'S ACCESSION-NO.
5. REPORT DATE
July 1979
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Franklin A. Ayer, Compiler
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Research Triangle Institute
P.O. Box 12194
Research Triangle Park, North Carolina 27709
10. PROGRAM ELEMENT NO.
EHE624A
11. CONTRACT/GRANT NO.
68-02-2612, Tasks 55 and 99
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Proceedings; 3/5-8/79
14. SPONSORING AGENCY CODE
EPA/600/13
is. SUPPLEMENTARY NOTES ffiRL-RTP project officer is Charles J. Chatlynne, Mail Drop 61,
919/541-2915.
is. ABSTRACT The publication, in two volumes, contains the text of all papers presented
at EPA's fifth flue gas desulfurization (FGD) symposium, March 5-8, 1979, at Las
Vegas, Nevada. Papers cover such subjects as health effects of sulfur oxides,
impact of.FGD on the economy and the energy problem, energy and economics of
FGD processes, actual operating experience, waste disposal and byproduct market-
ing , and industrial boiler applications.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. cos AT I Field/Group
Pollution Energy
Flue Gases Waste Disposal
Desulfurization Byproducts
Sulfur Oxides Marketing
Environmental Biology
Economics Boilers
Pollution Control
Stationary Sources
Health Effects
Industrial Boilers
13B 14B
2 IB
07A,07D
07B
06F
05C 13A
8. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
628
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (»-73)
I250a
*U.S. GOVERNMENT PRINTING OFFICE: 1979 -640-01 ^ 39 3 8 REGION NO. 4
------- |