EPA
ERDA
USBM
U.S. Environmental Protection Agency Industrial Environmental Research EPA~600/7"7o™073
Office of Research and Development Laboratory _ *| H f\~rQ
Research Triangle Park, North Carolina 27711 ApTll 1 9/O
U.S. Energy Research
and Development Administration
Fossil Energy
Washington, D.C. 20545
U.S. Bureau of
Mines
Coal Preparation and Analysis Laboratory
Pittsburgh, Pennsylvania 15213
PROCEEDINGS
OF THE ENGINEERING
FOUNDATION CONFERENCE
ON CLEAN COMBUSTION OF COAL
Interagency
Energy-Environment
Research and Development
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-600/7-78-073
April 1978
PROCEEDINGS OF THE ENGINEERING
FOUNDATION CONFERENCE ON
CLEAN COMBUSTION OF COAL
Victor S. Engleman
Conference Chairman
Science Applications, Inc.
1200 Prospect Street
La Jolla, California 92038
EPA Purchase Order DA-7-03662B
Program Element No. EHE624A
ERDA Grant No. EF-77-G-01-6003
USBM Purchase Order No. PO172317
EPA Project Officer: G. Blair Martin
ERDA Project Officer: A. Macek
USBM Project Officer: A.W. Deurbrouck
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, N.C. 27711
Prepared for
Coal Preparation and Analysis Laboratory Office of Research and Development
U.S. Bureau of Mines U.S. Environmental Protection Agency
Pittsburgh, PA 15213 Washington, D.C. 20460
Fossil Energy
U.S. Energy Research and Development Administration
Washington, D.C. 20545
-------
CONTENTS
SESSION
INTRODUCTION 1
Conference Chairman: V.S. Engleman,
Science Applications, Inc.
IA TECHNICAL, ECONOMIC AND ENVIRONMENTAL PROBLEMS
IN CLEAN COMBUSTION OF COAL 3
Session Chairman: Andrej Macek, U.S. ERDA
TECHNICAL AND ECONOMIC PROBLEMS FOR CLEAN
COMBUSTION OF COAL 5
S.I. Freedman, U.S. ERDA
COAL COMBUSTION AND FUTURE EMISSION REGULATIONS 21
Thomas Schrader, U.S. EPA
RECENT DEVELOPMENTS IN COAL COMBUSTION TECHNOLOGY .... 27
James I. Joiibert, Pittsburgh Energy Research
Center
IB STRATEGY AND APPROACH TO SPONSORED R&D U3
Session Chairman: Andrej Macek, U.S. ERDA
STRATEGY IN COAL PREPARATION RESEARCH PLANNING 1*5
W.E. Warnke, U.S. Bureau of Mines
STRATEGY AND APPROACH TO ERDA RESEARCH AND
DEVELOPMENT ON CLEAN COMBUSTION OF COAL ^9
S. William Gouse, U.S. ERDA
EPA R&D PROGRAM RELATING TO CONVENTIONAL COAL
COMBUSTION 63
Frank T. Princiotta, U.S. EPA
II PRECOMBUSTION PROCESSES 77
Session Chairman: P. Stanley Jacobsen, Colorado
School of Mines Research Institute
COAL PREPARATION HISTORY AND DIRECTION 79
Robert L. Llewellyn, Roberts & Schaefer Company
THE INFLUENCES OF MINING PRACTICES ON COAL
PREPARATION 83
C.A. Goode, Bureau of Mines
CURRENT COAL PREPARATION RESEARCH AND DEVELOPMENT .... 87
Richard P. Killmeyer, Jr., Bureau of Mines
COAL TRANSPORTATION IN 1985 91
David J. Hoexter, U.S. Department of Commerce
-------
ii CLEAN COMBUSTION OP COAL
CONTENTS (Continued)
SESSION FAGE_
II PRECOMBUSTION PROCESSES (Continued)
COAL DESULFURIZATION TEST PLANT STATUS - JULY 19TT- ... 97
L.J. Van Nice, M.J. Santy, E.P. Koutsoukos,
R.A. Orsini and R.A. Meyers, TRW Systems and
Energy
STATUS AND PROBLEMS IN THE DEVELOPMENT OF HIGH
GRADIENT MAGNETIC SEPARATION (HGMS) PROCESSES
APPLIED TO COAL BENEFICIATION ...... , ........ 109
Y.A. Liu and C.J. Lin, Auburn University
A THEORETICAL APPROACH TO WASHABILITY CURVES IS
COMPARED TO THE OTISCA PROCESS SEPARATION OF
FINE COAL ........................ 131
D.V. Keller, Jr., Otisca Industries, Ltd.
Ill COMBUSTION TECHNOLOGY ..................
Session Chairman: G. Blair Martin, U.S. EPA
SOME CHARACTERISTICS OF COAL COMBUSTION SYSTEMS .....
Janos M. Beer, Massachusetts Institute of
Technology
POLLUTANT FORMATION DURING COAL COMBUSTION ........ 171
M.P. Heap and R. Gershman, Energy and
Environmental Research Corporation
THE DUAL REGISTER PULVERIZED COAL BURNER: FIELD
TEST RESULTS ....................... 185
E.J. Campobenedetto , Babcock & Wilcox Company
COAL-OIL MIXTURE COMBUSTION IN BOILERS - AN UPDATE. . . . 193
Sushil K. Batra, New England Power Service Company
STOKERS FOR INDUSTRIAL BOILERS: ASSESSMENT OF
TECHNICAL, ECONOMIC, AND ENVIRONMENTAL FACTORS ...... 205
Robert D. Giammar, Battelle, Columbus Laboratories
INITIAL OPERATION OF THE 30 MWe RIVESVILLE MULTICELL
FLUIDIZED BED STEAM GENERATION SYSTEM .......... 219
Robert L. Gamble and Newton G. Watt is,
Foster Wheeler Energy Corporation
DEVELOPMENT OF AN EFFICIENT SOLIDS FUEL BURNER ...... 227
Norman A. Lyshkow, Combustion Equipment Associates,
Inc.
-------
iii
CONTENTS (Continued)
SESSION PAGE
IV POSTCOMBUSTION CLEANUP 235
Session Chairman: Sidney R. Orem,
Industrial Gas Cleaning Institute
ELECTROSTATIC PRECIPITATION STATE OF THE ART 237
R.S. Atkins and D.V. Bubenick, Research-Cottrell
STATE OF THE ART — FABRIC FILTRATION 255
Richard L. Adams, Wheelabrator-Frye, Inc.
STATUS OF FLUE GAS DESULFURIZATION, THE FEDERAL
RESEARCH, DEVELOPMENT AND DEMONSTRATION PROGRAM 273
Julian W. Jones and Michael A. Maxwell, U.S. EPA
STATUS OF FLUE GAS TREATMENT TECHNOLOGY FOR CONTROL
OF NOX AND SIMULTANEOUS CONTROL OF SOX AND NOX 293
J. David Mobley and Richard D. Stern, U.S. EPA
V WHERE DO WE GO FROM HERE? 311
Session Chairman: Victor S. Engleman,
Science Applications, Inc.
THE NATIONAL ENERGY PLAN 313
C. William Fischer, Department of Energy
THE OUTLOOK FOR COAL THROUGH 1985 323
Robert L. Major, AMAX Coal Company
WHERE DO WE GO FROM HERE IN R&D? 335
S. William Gouse, U.S. ERDA
-------
CLEAN COMBUSTION OF COAL
-------
INTRODUCTION
Coal is a major energy resource for the United States. The use of
coal as an energy source will increase in importance over the intermediate
term. While synthetic fuels from coal will be needed for specific appli-
cations, direct combustion of coal will probably continue to be the most
efficient and most economical use of coal. However, the environmental
problems associated with coal combustion must be solved economically and
efficiently at the same time for the continued growth of utilization of
this vital energy resource.
Clean combustion of coal must address issues of air quality, surface-
and groundwater quality and solid waste management. Control technology
for clean combustion of coal must include precombustion, combustion, and
postcombustion processes. Air quality impacts from coal combustion
include gaseous and particulate combustion product emissions as well as
potential fugitive emissions from handling and storage. Potential
pollutant emissions include NOX, SOX, particulates and trace elements.
Generally hydrocarbon and carbon monoxide emissions are low from large
coal combustors but new combustion- technologies should be checked for
emissions of these pollutants. Water quality impacts would result
primarily from leaching of chemical compounds from the solid residues of
the combustion process; this problem therefore goes hand in hand with
the solid waste management problem. As a minimum, the ash in the coal
must be managed. In technologies which produce additional solid waste,
(precombustion, combustion, or postcombustion) an additional solid waste
management problem is created.
Energy efficiency, availability, and economics must be considered
in establishing acceptable methods for clean combustion of coal. One
could postulate the emissions limit going to zero while paying the price
of the energy efficiency going to zero as well. Therefore, it is not
only a matter of technical, economic, and environmental considerations,
but also policy considerations. This is recognized in the program for
this conference, and while the engineering aspects of coal combustion
are emphasized, the vital importance of energy policy and its impact on
clean combustion of coal is highlighted.
Special thanks to the Organizing Committee for their role in
establishing the program and helping to arrange for speakers.
Andrej Macek - ERDA .
Albert W. Deurbrouck - Bureau of Mines
G. Blair Martin - EPA
Sidney R. Orem - Industrial Gas Cleaning Institute
Joseph W. Mullan - National Coal Association
-------
2 CLEAN COMBUSTION OF COAL
The suggestions provided "by the Advisory Committee were invaluable
in selecting the highly qualified slate of speakers and the equally
qualified group of conference participants. Their efforts are grate-
fully acknowledged.
E.K. Bastress - ERDA
R. Beck - ERDA
J.O. Berga - National Research Council
R.A. Carpenter - National Research Council
D.E. Gushee - Congressional Research Service
R. Hangebrauck - EPA
G.R. Hill - EPRI
G.A. Mills - ERDA
E. Plyler - EPA
W.E. Warnke - Bureau of Mines
The experienced staff of the Engineering Foundation Conferences
with guidance and assistance both before and during the conference were
instrumental in its success. Special notes of thanks to
Sandford S. Cole, Conferences Director
Harold Comerer
Dean Benson.
In addition, the staff at the conference site at Franklin Pierce College
provided invaluable assistance during the course of the conference.
The publication of the proceedings are due in large part to the
patience and perseverance of Ms. Billie St. Pierre who saw that the manu-
scripts were put in their proper final form.
-------
SESSION IA - TECHNICAL, ECONOMIC AND ENVIROMMEHTAL
PROBLEMS IN CLEM COMBUSTION OF COAL
SESSION CHAIRMAN: ANDREJ MACEK, U.S. ERDA
In broad terms, we will have to learn to burn coal under very
restrictive conditions. The restraints will be imposed by national
policy considerations and environmental requirements. This opening
session will include a general description of coal combustion technol-
ogies which are, or might become, candidates for clean combustion, and
of the problems associated with these technologies. Policy considera-
tions, which are also a key element in clean combustion of coal, will
be included in the closing session of the conference. Two of the papers
in this session describe four candidate technologies: fluidized bed
combustion of coal, combustion of solvent-refined coal, combustion of
coal/oil mixtures, and combustion of coal for direct-powered magneto-
hydrodynamic generators. In addition, there is a discussion of the
potential environmental constraints in the forseeable future.
-------
CLEAN COMBUSTION OF COAL
-------
TECHNICAL AND ECONOMIC PROBLEMS FOR CLEAN COMBUSTION OF COAL
by
S. I. Freedman
Assistant Director, Combustion and Advanced Power Development
Division of Coal Conversion and Utilization
Fossil Energy
U.S. Energy Research and Development Administration
Washington, D.C.
INTRODUCTION
Ever escalating demand for limited and dwindling supplies of con-
ventional fuel resources has been recognized as a key factor in pre-
cipitating the worldwide financial and energy crises. The gravity of
these crises has forced nations around the world to look for ways and
means to improve their energy supply base to meet the ever increasing
demands. To meet the challenge, President Carter declared the "moral
equivalent of a war" for energy independence.
Although nuclear technologies might satisfy a substantial portion
of world energy needs in a distant future, there exists an immediate
necessity for the development of energy technologies to conserve and
efficiently use a wide base of natural resources.
Current estimates indicate the U.S. coal reserves, which are econ-
omically mineable with conventional technology, at about 600 billion
tons and the overall coal resource at 3.2 trillion tons. Depending on
the efficiency of mining, population growth, and growth in energy use
per capita, and efficiency in end use, we have enough coal to supply
our energy needs for several hundred years; sufficient, it is hoped, to
carry us through a transition to a non-fossil fuel energy source. Ad-
vances in coal mining technologies and increased coal price can increase
the estimate for coal which can be economically produced considerably
beyond the 600 billion ton figure. Recognizing this abundance of coal,
the U.S. Energy Research and Development Administration (ERDA) is greatly
emphasizing the development of technologies to utilize efficiently a
wide variety of coals in an environmentally acceptable manner.
The ongoing ERDA efforts in the development of fossil energy tech-
nologies are a continuation of the programs of its predecessor organiza-
tions, the U.S. Department of Interior's Office of Coal Research and
Bureau of Mines. These technologies include: (1) Coal gasification,
(2) Coal liquefaction, (3) Direct combustion of coal, and (4) Advanced
power systems using coal and coal-based fuels.
ERDA is supporting the development of these technologies by funding
various projects ranging in size from exploratory research to pilot and
demonstration projects.
-------
6 CLEAN COMBUSTION OF COAL
ERDA's main objective is to bring up the technologies to technical
maturity for commercial implementation. Although coal gasification and
liquefaction are of considerable importance, this paper emphasizes the
ERDA efforts in direct combustion technologies.
Direct combustion of coal in boilers and furnaces is a well estab-
lished practice for heat and power generation. Only a certain fraction
of the U.S. coal reserves, with low-sulfur content, termed compliance
coal, can be burned directly in conventional furnaces to meet the envir-
onmental regulations if burned in conventional burners. As Dr. James
Schlesinger aptly points out, "Americans want a clean environment and do
not want to turn the country into something equivalent to Pittsburgh in
the 1930's." Of course, all modern coal fired equipment is built for
high combustion efficiency and includes a bag house or precipitator to
reduce the emission of particulates, thereby avoiding the two most obvious
sources of pollution from coal fired equipment, soot and fly ash. To
achieve increased coal utilization while meeting stringent environmental
regulations, the Government is directing the development of promising
coal utilization technologies.
These technologies can be categorized as follows:
• Precombustion treatment of coal for ash and sulfur removal.
• Post combustion treatment for S02 removal.
• In-situ S02 removal at the source during coal combustion.
A brief description of these technologies follows.
Precombustion Treatment of Coal
Precombustion treatment of coal includes a number of physical and
chemical approaches such as:
• Physical cleaning.
• Chemical cleaning.
• Solvent refining.
Precombustion treatment, essentially aimed at removing ash (mineral
matter) and other obnoxious constituents such as sulfur in its pyritic
form, increases the specific heating value (Btu/lb) and reduces the sulfur
content of some coals to an environmentally acceptable level.
Physical and chemical approaches for cleaning coal are in different
stages of development: A full-scale plant to demonstrate a multi-stream
physical cleaning concept is being built at Homer City, Pennsylvania, and
is scheduled to begin operation this year (1977). A pilot plant to test
sulfur removal through chemical leaching is being built in California.
Commercial implementation of this technology, depending on favorable
pilot plant results, can be anticipated in the early 1980's. The impact
of these technologies is to increase the availability of compliance coals
by a maximum of approximately 175 million tons per year. This still
leaves the majority of American coal being mined as non-compliance coal.
Physical cleaning is estimated to cost about $1.50 to $4.50 per ton while
chemical cleaning would cost about $10-$15 per ton.
-------
TECHNICAL AND ECONOMIC PROBLEMS
Some variable amount of our existing coal reserves can be considered
as compliance coal, based on the S02 emissions criteria that apply locally,
and can be used in existing equipment. The Federal EPA rules on SC>2 emis-
sions for new sources are quite definite, 1.2 Ibs of S02 per million Btu.
Low-sulfur coal may be defined as coal that occurs naturally with
a sulfur content low enough that the coal can be burned without violating
applicable S02 emission standards. As S02 emission standards vary, the
same coal may be considered compliance for one application but not for
another. The extent to which low-sulfur coal is used to meet national
ambient air quality standards (NAAQS) and New Source Performance Standards
(NSPS) will depend primarily on the availability. It should also be noted
that S0-j in combustion gases, which accompanies S02> enhances the col-
lection efficiency of electrostatic precipitators, so the use of low-
sulfur coal at existing facilities necessitates modifications of electro-
static precipitators to control particle emissions. Low-sulfur coals
with low ash fusion temperatures can cause fouling in boilers, reducing
capacity and increasing maintenance costs. This has happened in several
instances when western coals were used in boilers designed for eastern
coals.
The availability of low-sulfur coal depends on the expeditious dev-
elopment of our domestic resources and of the transportation network re-
quired to get the fuel to the markets. In 1974, almost 390 million tons
of coal were consumed by electric power plants; about half would comply
with S02 emission regulations in effect on July 1, 1975. In 1980, utility
coal demand is projected to be about 620 million tons with an 8% annual
growth rate. In an October 1974 EPA report, low sulfur coal production
in 1980 was projected to supply less than 44% of the 620 million ton
demand.
The production capacity for low-sulfur coal and its use will depend
on many factors, including:
(1) The relative economics of low-sulfur coal in comparison with
other control approaches.
(2) The resolution of disputes over compliance schedules and
state sulfur regulations.
(3) Political/economic decisions concerning low-sulfur coal
recovery and distribution.
(4) The compatability of low-sulfur coal usage with existing
boilers and electrostatic precipitators.
The costs of complying with SOn standards by using low-sulfur coal
include: (1) The increased cost of the low-sulfur coal, (2) Transporta-
tion cost, and (3) The cost of power plant modifications which may be
necessary. While low-sulfur coal is generally more expensive than high-
sulfur coal, the price differential currently is highly variable.
Transportation cost often causes significant increases in low-sulfur
coal costs. Another factor to consider is that about one and one-half
tons of the western coal are needed to equal the heating value of one
ton of eastern coal.
-------
CLEAN COMBUSTION OF COAL
Post Combustion Treatment for SO? Removal
Flue gas desulfurization (FGD) is a post combustion method of re-
moving S02 from combustion gases. FGD processes contact the combustion
gases with a sorbent to react with the S02.
The numerous FGD processes can be categorized as nonregenerable or
regenerable processes. In nonregenerable processes, the sorbent reacts
with absorbed S02 and is not regenerated or reused. These processes
produce a sludge which consists of a mixture of fly ash and water plus
calcium sulfite and calcium sulfate which result from the S02/sorbent
reactions. Several alternatives for disposition of the sludge are avail-
able. The alternatives include ponding of untreated sludge, landfilling
of untreated and treated sludge, and commercial utilization.
Regenerable processes recover SOo from the combustion gases and
convert it into marketable by-products such as elemental sulfur or
sulfuric acid.
The best developed FGD processes are the nonregenerable lime and
limestone scrubbing processes. In these processes, a lime or limestone
slurry absorbs S02 from the flue gases, and subsequent reactions produce
calcium sulfite and calcium sulfate. A variation of lime/limestone scrub-
bing, known as the "double alkali" process, uses a clear solution of a
soluble salt rather than a slurry to absorb S02. Subsequent reactions
to produce calcium sulfite and calcium sulfate sludge occur outside the
absorber.
The most significant of the regenerable processes are magnesium
oxide scrubbing and sodium sulfite scrubbing (the Wellman-Lord process).
The magnesium oxide process utilizes an MgO slurry to absorb S02 and
produce magnesium sulfite and magnesium sulfate. The Wellman-Lord pro-
cess employs a clear solution of sodium sulfite to absorb S02 and react
with it to form sodium bisulfite.
The first full-scale U.S. installations of the lime and limestone
scrubbing processes were operated in the late 1960's, and the processes
that evolved have controversially been considered to be commercially
available since 1974.
The economics of FGD are the most critical factor affecting its
widespread use and also the factor that arouses the most debate. Problems
of inflation, extent of redundancy, site-specific requirements, vendor
optimism/user skepticism, variations in process options, variations in
the kinds or quantities of by-products, differences in operating condi-
tions, and other factors have contributed to the complexity of the
situation. The cost estimates for scrubbing equipment have escalated
from about $18/kw in 1968 to about $89/kw in 1977. On top of the equip-
ment costs have to be added the cost of heat and power to operate the
scrubbers, the additional site labor costs, interest during construction
costs and escalation costs. The difference in capital costs between
two new plants which are identical except that one is equipped with a
scrubber is presently $150/kw. This 400% increase is not wholly at-
tributable to inflation. The figure does underline the difficulty of
accurately projecting control costs over a period of a few years for
an emerging new technology.
-------
TECHNICAL AND ECONOMIC PROBLEMS 9
A member of my staff recently had the opportunity to visit a scrubber
installation and closely examine the operation. The plant visited was
representative of the new large coal-fired plants.
The plant consists of two 825 MW units. Steam condition is 3500 psia
1000°F/1000° and steam is supplied by a supercritical boiler to
turbo-generators. The units burn pulverized coal. Operational control
is by digital computer. Natural Draft cooling towers are used.
Flue gas desulfurization is accomplished by Wet Lime Scrubbers.
Capital costs for each of the two units were:
Plant and Equipment $510 x 106
Scrubbers plus Cooling
Tower $240 x
$750 x 106
The scrubbers are of the two vessel venturi type. The first vessel
is for particulate control and the second does the sulfur absorbing. A
demister follows the second vessel, then a reheater and the stack. There
are six equal size parallel trains for each unit while each train handles
approximately 150 MW. All six units must be worked to handle the full
output. The vessels and piping are all rubber lined. The limestone is
calcined separately at an installation 300 yards away, slaked, doped
with MgO and piped to the units. Wet sludge is disposed of in a man-
made lake with an estimated time to fill up of twenty years.
Pittsburgh seam coal, approximately 3 percent sulfur is used, most
deep mined, some Ohio surface mined.
The design operation calls for a sulfur removal guarantee of 92%.
43 MW of electrical power is 'used to drive the scrubber and 1200 gal/hr
of #2 distillate fuel oil is burned to reheat the 125°F air from the
demister to the 165°F stack.
Nominal crew for the two units is 136 men. This does not include
the extra men required to maintain and modify the scrubbers.
Commercial operation started December 1975 and the scrubbers were
not yet running satisfactorily in June 1977.
Performance achieved to date has been very good considering the
circumstances. The turbine-generator heat rate is 8881 Btu/kWh (which
is probably due to the impluse turbines) and 9981 Btu/kWh for the whole
plant. The achieved capacity factor is 71%.
The turbine, generator, and boilers work well. The scrubbers do not
work well when they are operating. With six trains in parallel on each
825 MW unit, they can get some power out through the use of at least one
train 91% of the time.
-------
10 CLEM COMBUSTION OF COAL
Sulfur removal runs ca 60% in lieu of the guarantee of
Cost of electricity at the busbar is:
Plant 9 Mil/kWh
Scrubber 3 Mil/kWh
Total Unit 12 Mil/kWh
The lime requirement is 7.8 percent by weight of the coal used. For
a 3 percent sulfur coal, this gives a Ca/S mole ratio of approximately
1.5.
They have had endless operating troubles, and the only apparent
way to minimize the problem is very, very careful control of the pH of
the solution in the particulate and S02 absorption vessels. The pH can
vary from 2 to 10 in lieu of the desired 7.
The rubber lining in the vessel and piping suffers mechanical damage
from the displaced deposits. Due to the acid condition, this has eaten
through the 3/8 carbon steel plate in as little as three days.
The demister chevrons and cascades plug in three days to a month.
The original demister structure failed as did the succeeding two genera-
tions of replacements.
The sludge carry-over frequently plugs the lines, requiring shutdown
to clear them.
Additional manpower over and above the 136 normal, runs twenty to
fifty men. These are required to keep the scrubbers maintained and are
a mix of power plant, vendors, and contractor personnel since the scrub-
bers are not accepted yet.
In summary, the new coal plant is working well except for the wet
lime scrubbers. They do not work well, and add greatly to the cost at
this time.
In-Situ S02 Removal at the Source During Coal Combustion
One ERDA approach is to develop new alternative technologies which
control emissions during the combustion process or regulate input of
potential pollutants. This is being pursued by fluidized-bed combustion
(FBC) for heat, steam and power generation in the presence of sulfur
dioxide sorbent.
The increased use of coal by this technology for heat and power
will benefit the public. Coal is appropriate for use in moderate and
large-sized fuel burning installations where it is economical to have
equipment necessary to handle large amounts of solids (i.e., coal, lime-
stone, ash). Increased coal use can reduce oil and natural gas use by
the utility and industrial sector and make domestic supplies of these
fuels available for the unique requirements of the residential, commer-
cial and transportation sectors of the economy.
-------
TECHNICAL AND ECONOMIC PROBLEMS 11
The user sectors which these programs will supply with technology
are:
• Industrial and institutional heat and steam.
• Utility steam-electricity generation.
Even with a low-energy growth future, the need for new and replacement
fossil-fueled equipment facilities for these user sectors represent a
major market.
Emission control for fluidized-bed boilers centers in the combustion
zone. Sized coal is burned at atmospheric pressure or at elevated pres-
sures in a fluidized bed of inert ash and limestone or dolomite. The
limestone or dolomite is calcined and the calcium oxide reacts with S02
to form a solid sulfate material which can be disposed of with the ash.
Sorbent regeneration could produce a highly concentrated SC>2 gas from
which elemental sulfur could be readily removed using existing processes.
Fluidized-bed boilers are operated at combustion temperatures of 1500 F
which are lower than conventional boilers which have flames well above
3000°F, thus inhibiting the formation of nitrogen oxides from combustion
air nitrogen. This lower combustion temperature greatly reduces the
tendency of slag to build-up on boiler tubes. Western coals, with high
alkali content, and which foul conventional boilers rapidly, have been
burned in fluidized bed research combustors without fouling. An atmos-
pheric pressure coal-fired fluidized-bed boiler having a capacity of
5,000 pounds of steam per hour has been successfully operated for over
12,000 hours at furnace temperatures of approximately 1,600°F. It demon-
strated that all types of coal, char and coal wastes can be burned in
an environmentally acceptable manner.
The ERDA strategy to accomplish the development of FBC of coal is
to have the majority of the technology development done in industry,
to work on the technologies promising the earliest significant payoff,
and to utilize experience gained on early development of small-scale
equipment to benefit the development of the following larger-scale units.
This reduces cost and accelerates commercial implementation. An example
of the relative costs for FBC vs. conventional systems is shown in
Table I for large utility applications.
The minimum risk and the best schedule of progress are achieved by
performing parallel engineering development activities in several modest
size and cost systems and by testing various applications at the same
time. The Federal Government will expedite industrial development by
providing the flexible, costly test facilities for common use that industry
finds to be too expensive. By having development work performed by in-
dustrial firms with the use of these test facilities, the program can
proceed rapidly to commercial implementation when the development is
completed.
The user's requirements are recognized and included from the start.
The industrial-institutional scale units are of of a size large enough
to realize the economies of using coal, sorbent and ash handling systems,
which are not economical in very small systems. Utility systems require
much larger plants with very firm technical specifications, and longer
construction times. Improvements which come from developments in the
intermediate size industrial units will benefit the utility size systems.
-------
TABLE I
CHARACTERISTICS OF COAL COMBUSTION
PERFORMANCE AND
COST CHARACTERISTICS
Capital Cost ($/kW)*
O&M Cost (Mills/kWh)
Efficiency (%)
(Heat Rate - Btu/kWh)
Max. Capacity Factor
Average Capacity Factor
Initial Capacity Factor
Years to Reach Max.
Capacity Factor
(End of Year)
Construction Time (Years)
Earliest Commercialization
Date
Steam with
AFB (3500 psi;)
1000°F/1000°F
450
(350)**
2.2
36
(9,481)
.70
.60
.50
3
5
1984
Steam with
PFB (3500 psi;)
1000°F/1000°F
514
(400)**
2.3
39
(8,751)
.70
.60
.50
3
5
1988
Conventional
Steam with FGD
578
(450)**
2.8
35
(9,751)
.70
.60
.50
3
5
o
o
o
CO
i-3
H
O
o
I
* 1975 Dollars with Allowance for Funds During Construction.
** 1975 Dollars without Allowance for Funds During Construction.
-------
TECHNICAL AND ECONOMIC PROBLEMS 13
The development of modular, transportable systems and components
capable of factory production is to be emphasized where economically
competitive with field constructed systems. This strategy will afford
the program the benefits of rapid production, standardization and quality
control, but of more importance from the viewpoint of national objectives,
will provide an industrial production system capable of rapid expansion
in the event of a political curtailment of alternative fuels.
DIRECT COMBUSTION PROGRAM
Direct combustion is the most efficient and economic method of
utilizing coal.
Fluidized-bed combustion of coal has been a topic of research since
the early 1960's, both in the U.S. and in Europe; in the U.S., ERDA is
the manager of the Federal RD&D program for FBC. Consequently, a large
data base exists to identify the technology gaps and to point out areas
which require additional research. To address the relevant technical
issues through the commercial implementation phase, ERDA has developed
a comprehensive RD&D program for fluidized-bed combustion.
An operational pilot plant, 30-megawatt in size and designed for
atmospheric pressure, represents both the largest size of fluidized-bed
combustion industrial boilers and is of sufficient size to be representa-
tive of modules for electric utility boiler applications. It has super-
heater tubes, delivering the steam at 1325 psig and 930°F; consequently,
it is more typical of a utility application. Because the unit is located
in a private power company's boiler house, it will be evaluated for its
capability for integrated operation in an actual operational environment.
This pilot plant is the largest fluidized-bed boiler in the world
and will address scale-up problems never previously explored on a scale
this large. The unit consists of four cells, three primary cells and
a carbon burn-up cell. A 13 MWe pilot plant is being designed to explore
the pressurized fluidized-bed concept on a large scale.
For potential industrial applications, ERDA has developed a multi-
project program to evaluate fluidized-bed combustion for both boiler and
process heat applications. Several units are being built to demonstrate
the technological, operational and economic viability of using fluidized-
bed combustion in industrial applications. Demonstrating the viability
of such systems is essential for the commercial implementation by
industry.
An additional and important element in the fluidized-bed combustion
program is the design and construction of component test and integration
units (CTIU) for both atmospheric pressure and elevated pressure combustor
operation. The CTIU's will be flexible facilities, available for use by
all of industry and the research community, that will permit optimization
of the combustion process/steam or heat regeneration cycle, testing of
components in an integrated operating mode, and evaluation of concepts
as yet untested such as the vertically stacked bed concept.
The following are descriptions of several projects being conducted
by ERDA which are directly applicable to the industrial sector, where
-------
14 CLEM COMBUSTION OF COAL
RD&D efforts can be expected to produce near-term implementation of the
technology, which will be in compliance with existing and projected emis-
sion criteria.
PROJECT: 30 MWg AFBC BOILER
Description; Represents large-scale industrial boiler. Three
parallel primary cells, each 10' x 12'. Cells operate up to 12 fps
fluidizing velocity with a nominal 24" static bed in each cell (one
steam generating cell, two superheat cells, plus one carbon burn-up
cell). Installed in a utility generating plant. Steam conditions are
1325 psig and 930°F. Coal feed of 15 tons per hour with sorbent feed
dependent on sulfur in coal.
Contractor: Pope, Evans and Robbins, Inc.
Site: Rivesville, WV, Monongahela Power Company Station
Objective; Early demonstration of FBC technology in utility environ-
ment using high-sulfur coal. Evaluate scale-up of heat transfer, com-
bustion efficiency, emission control. Demonstrate automatic combustion
control in central load dispatch utility system.
Status: Boiler installation completed. Successful ignition of
coal in the carbon burn-up cell 7 December 1976. Successful start-up
of superheater cell 20 April 1977.
Plans: Full load steam generation expected to begin by the end of
CY 1977. A full spectrum test program will then be initiated to meet
the unit/program objectives.
PROJECT: INSTITUTIONAL SIZE INDUSTRIAL FLUIDIZED-BED BOILER
Description; This fluidized-bed boiler will generate 100,000 Ib
per hr of saturated steam at 650 psig. The boiler will furnish this
steam to a university and hospital for space heating. The boiler plant
will burn high-sulfur coal while attaining low levels of pollutant
emissions. The steam will be used to drive a small turbine for demon-
stration as a co-generation system.
Contractor: Georgetown Hospital
Site; Georgetown University, Washington, D.C.
Objective: To demonstrate an institutional saturated steam boiler
capable of complete shop assembly can be built in capacities up to
150,000 Ib per hr and that materials handling systems for this type of
application can be optimized.
Status: Letter of intent to negotiate contract issued.
Plans: To be operational sometime in 1979. Three years operational
test program and data dissemination plan under final development.
-------
TECHNICAL AND ECONOMIC PROBLEMS 15
PROJECT: INDUSTRIAL APPLICATION FLUIDIZED-BED COMBUSTION PROCESS - TWO
UNITS
Description; The work plan is divided into two phases, subscale
testing of natural circulation design capability and design and testing
of a demonstration natural circulation fluidized bed boiler. This demon-
stration unit will generate 50,000 Ibs of steam per hour at 500 psig
and 600°F, or about 120°F of superheat.
Contractor; Combustion Engineering, Inc.
Site; Great Lakes Naval Training Center
Objective; Confirm the subscale test data design parameters and
procedures subsequently developed. Provide endurance test data on
materials. Provide design information for possible improvements to
demonstration and subsequent units.
Status; Letter of intent to negotiate contract issued.
Plans; Subscale design through testing for the natural circulation
unit will take 1-3/4 years from contract initiation. Demonstration design
through testing will take two years.
PROJECT: INDUSTRIAL BOILER
Description; Develop a demonstration multi-solids fluidized-bed
boiler. Basically, a multi-solids FBC can be regarded as a recirculating
or entrained fluidized-bed superimposed on a conventional dense fluidized
bed region using fluidizing velocities as high as 35-40 fps. Heat trans-
fer advantages are expected. The demonstration unit for this process is
to generate 25,000 Ib/hr of saturated steam at 100 psig.
Contractor; Battelle Memorial Institute's Columbus Laboratories
Site; Columbus, Ohio
Objective; Development of the multi-solids FBC system to the point
where industry will accept and use it for commercial industrial steam
generation.
Status; Contract signed.
Plans; Phase I - R&D in support of demonstration plant (25 months
duration); Phase II - Construction and start-up of demonstration plant
(14 months duration); Phase III - Demonstration plant operation (36
months duration).
PROJECT: FBC PROCESS HEATER/TUBE STILL
Description: Evaluate the FBC concept for fluid process heaters
such as those used in refineries. A Process Development Unit (PDU)
will be designed and constructed to test a fired heater in the 10-15MM
Btu per hour range for application to a tube still.
-------
16 CLEAN COMBUSTION OF COAL
Contractor: EXXON Research and Engineering Company
Site: Linden, New Jersey (Engineering)
PDU Location at Refinery
Objective: Conduct R&D to obtain engineering design data for coal
fired fluid bed process heaters. Demonstrate operation of the PDU.
Prepare design specifications and cost estimate for a commercial FBC
process heater.
Status; Contract signed.
Plans: Phase I - R&D test work (18 months duration) ; Phase II -
PDU design (6 months duration); Phase III - PDU evaluation and commercial
design (12 months duration starting two years after contract initiation).
PROJECT: FBC PROCESS HEATER/AIR HEATER
Description; The FBC heater will supply 39 x 10 Btu per hour to
heat clean air to 900°F in the heat exchanger. The total load will be
served by three beds with a thermal storage unit in the system to pro-
vide heat during intermittent operation of each bed to achieve load
matching. Bituminous coal, lignite, refinery coke, peat as primary
fuels and such things as wood and paper waste plus cornstalks as secondary
fuels will be burned.
Contractor; Fluidyne Engineering Corporation
Site; Owatonna, Minnesota
Objective: Demonstrate FBC heating of clean, low-pressure air to
900°F. Provide long-term data on life of air tubes in and above the bed
at different temperatures. Prove application to manufacturing processes
and space heating.
Status; Contract signed.
Plans: Construct by late 1978 or early 1979.
Another facet of the ERDA direct combustion program that is accel-
erating in importance is the use of western coals in FBC units. The
future of the western coals as compliance coals could heavily depend
on FBC.
The ash of lignites and western subbituminous coals contains a high
percentage of the alkali constituents, calcium, magnesium, and sodium.
These coals, when burned in a conventional pc-fired furnace, will retain
from 10 to 40 percent of the sulfur on the ash. These same coals, when
burned in a fluidized-bed combustor, should retain a greater percentage
of sulfur on the ash because of: (1) The increased contact time between
the sulfur dioxide produced during the combustion process and the inherent
ash alkali, and (2) The lower combustion temperatures for the fluidized
combustion process. The longer contact time allows for the reaction
between coal ash alkali, oxygen, and S02 to be carried further towards
completion. The lower temperatures allow for greater alkali utilization
-------
TECHNICAL AND ECONOMIC PROBLEMS 17
because the temperatures are below the disassociation temperature of the
sulfates and below the ash fusion temperatures so that the alkali is not
tied up in the glassy agglomerate. The sulfur retention percentages ob-
tained in preliminary tests on one North Dakota lignite at the Grand
Forks Energy Research Center range from 38 to 58 percent without ash
reinjection.
The role of the Grand Forks Energy Research Center of ERDA in the
fluidized-bed combustion program is to develop data on sulfur retention
on the alkaline coal ash from western United States coals, when burned
in a fluidized-bed combustor without a sorbent. FBC will result in
substantially lower S02 emissions than from conventional combustion sys-
tems, and may therefore meet the Federal New Source Performance Standard
of 1.2 Ib S02/10 Btu or even new lower federal or local emission
level. The effects of operating conditions and coal ash composition on
the retention of sulfur dioxide during fluidized-bed combustion of western
coals is being thoroughly evaluated to provide a design base for the use
of these coals.
Six coals from the western United States other than the lignite
previously mentioned have been tested in the fluidized-bed combustor at
Grand Forks. Preliminary results are presented in Table II. Recycle
was not used for these tests. The sulfur retention would be expected
to improve significantly with recycle.
It must be emphasized that this work with western coals was done
without a sorbent addition or a sorbent bed. The importance here is
certainly both in improving the economics of FBC and in making the
auxiliaries less complicated. The bottom line, however, is in the
significant impact of making currently non-compliance western coals
compliance coals in FBC without the use of a sorbent. As a bonus,
the low temperatures used with FBC will minimize or eliminate the foul-
ing/slagging experienced in conventional units with these fuels.
Coal Conversion Processes
The mandatory increase in the use of coal over the next quarter to
half of a century mandates the widespread application of control tech-
nologies, especially for S02, singly and in combination, to prevent
extensive degradation of the environment. An SOo control program which
permits timely and economic attainment of the regulatory requirements
is essential. Such a program will be comprised of various control options
including naturally occurring low-sulfur coal, coal cleaning and flue
gas desulfurization separately or in combination, coal gasification,
coal liquefaction, fluidized-bed combustion of coal, and others. Coal
gasification and liquefaction are viewed as SOo control technologies
for the overall program.
The choice of appropriate 862 control strategy for a specific site
is often complex due to the interaction of such factors as commercial
availability of competing systems, reliability and operability of the
systems in other applications, capital costs and annual revenue require-
ments, fluctuation in the availability and cost of low-sulfur fuels,
and uncertainty about pollutant emission regulations.
-------
TABLE II
oo
PRELIMINARY SULFUR DIOXIDE AND NITRIC OXIDE
EMISSION DATA FOR VARIOUS WESTERN COALS AND LIGNITES
Rank
I±!
S.B.2/
L
S.B.
B^
S.B.
S.B.
Coal Tested
Beulah, N. Dak.
Naughton, Wyo.
Velva, N. Dak.
Cols trip, Mont.
Browning, Utah
Sarpy Creek, Mont.
Decker, Mont.
Calcium-to-
Sulfur Mole
Ratio
1.27
0.67
5.09
1.29
0.81
1.56
1.38
Total Alkali-
to-Sulfur
Mole Ratio
1.99
1.09
7.54
1.87
1.23
2.18
2.48
Percent
Coal
Sulfur
1.01
0.34
0.23
0.62
0.88
0.71
0.33
Ash
7.49
5.04
6.36
8.16
8.07
8.98
4.25
Coal
Nitrogen
0.45
1.1*'
0.7*'
0.6*7
1.28^
0.68*/
1.0*'
Emission, ppm @
45% excess air
Sulfur NQ
Dioxide x
480 340
307 400
10 390
.377 360
588 400
412 339
121 350
o
IT"
o
Q
C
CO
H3
H
O
£=!
O
o
- Lignite.
2/
— Subbituminous.
3/
— Bituminous.
— Percent of nitrogen not available for specific sample burned in FBC.
sample of same coal.
Data taken from earlier
-------
TECHNICAL AND ECONOMIC PROBLEMS 19
Depending on application and end product, specific gasification
processes can be described by one or more of the following terms: low-
Btu, intermediate Btu, high Btu, fixed bed, entrained bed, fluidized bed,
slagging, air blown, oxygen blown, atmospheric and pressurized.
Gasifier systems like Lurgi, Winkler, Wellman-Galusha, Koppers-Totzek,
HyGas, Bi-Gas, Synthane, C02-Acceptor, Texaco, Riley-Morgan, U-Gas, and
COGAS are familiar names to those who have followed the history of coal
gasification. Sulfur clean-up systems such as Stretford, iron oxide,
Benfield, fluidized bed dolomite, Trencor, Selexol, Sulfinol, and Rectisol
are partner systems well known to the industry.
Considerable research, development, and engineering have already
taken place in coal gasification and some systems are commercially avail-
able and are used around the world. Installations of the small sized
Lurgi, Winkler, Koppers-Totzek, and Wellman-Galusha gasifiers are too
numerous for mention here and these are definitely considered commercial.
New combinations of or scale-ups of these gasifiers, along with I^S and
particulate matter removal systems need to be coupled in practice.
Economics of the leading coal gasification systems are available
from a number of studies completed during recent years. TVA has conducted
considerable work in evaluating the economics of coal gasification and
H2S removal systems. Economic analyses of 500 MW power units with sulfur
production of about 218 tons per day, indicated the following direct
capital cost (in 1975 dollars) and annual revenue requirements for 7,000
hours/year of operation:
Direct Capital Investment
Annual Revenue
Process Without Glaus With Glaus Requirements
Stretford $18,000,000 Not needed $3,400,000
Benfield 5,000,000 $10,000,000 3,600,000
Trencor 10,000,000 3,800,000
Selexol 12,000,000 17,000,000 3,500,000
Total project cost would be somewhat higher than the direct costs indicated.
Economics for new combined cycle plants depend a great deal on the
state-of-the-art of gas turbine operation. For existing gas turbine
technology where the maximum temperature is around 1950 F, power genera-
tion via a coal gasification unit integrated into a combined cycle plant
is more expensive than power generation with conventional boilers with
FGD supplying steam to a conventional steam turbine plant. If the advance-
ment of the gas turbine technology permits turbine temperatures up to
2500°F, the advantage could possibly switch to the gasification-combined
cycle units.
Liquid fuels can be produced from coal by four general methods.
First, the coal can be pyrolyzed to break it down into a synthetic crude
oil, a by-product gas, and a char residue. The char residue must be
used in some way since it represents a significant fraction of the total
heating value of the coal. Second, the coal can be dissolved in an
organic solvent, treated with hydrogen, and filtered to remove the non-
combustible minerals. The combustible fraction filtered out with the
-------
20 CLEAN COMBUSTION OF COAL
non-combustibles must be reclaimed and used if the process is to be
economically competitive with alternative technologies. Third, the
coal can be gasified to produce carbon monoxide and hydrogen which can
subsequently be converted to methanol or other liquid hydrocarbon fuels.
Fourth, pulverized coal can be contacted with hydrogen in the presence
of catalysts to form fractions of liquid and gaseous hydrocarbons.
The economic aspects of coal liquefaction are not well defined
since the technology is still being developed. Some studies were made
earlier in the development of the technology, but these estimates were
optimistic and are generally not regarded as accurate. The preparation
of additional estimates is now underway on many of the systems. In
general, the cost of coal liquefaction is comparatively high; base load
power generation using fuels from these processes is probably not econ-
omically competitive with low-Btu coal gasification, fluidized-bed com-
bustion, or conventional boilers with FGD. Coal liquefaction will un-
doubtedly find application for intermediate duty & peaking service, where
it is much more competitive with other technologies.
Conclusion
The accelerated use of coal in our national energy picture brings
along with it some interesting points. It is going to be expensive.
It is a definite technology challenge; a surmountable one, but definitely
complex. The environment can be protected. Currently, the direct com-
bustion of coal in a fluidized-bed would appear to have the overall edge
when the technical and economic impacts are evaluated against some base-
line environmental compliance criteria; however, it would appear that
we will have a continuing need for both liquid and gaseous conversion
fuels to meet our diverse utility/industrial energy consumer fuel needs.
-------
21
COAL COMBUSTION AND FUTURE EMISSION REGULATIONS
by
Thomas Schrader
Office of Planning and Evaluation
U.S. Environmental Protection Agency
It is a pleasure for me to be here this morning and have the oppor-
tunity to talk with a group of professionals who are involved with many
of the same problems that are important to my work at the Environmental
Protection Agency.
As Dr. Freedman highlighted earlier this morning, the concern for
clean air has had and will l:.ave a considerable impact on the pattern and
techniques of burning coal. My intent this morning is to talk primarily
about environmental policy and future emission regulations for coal
burning sources. I will discuss three major areas which have consider-
able effect on the implementation of clean coal technologies. These are:
- emissions standards which are to be met by new facilities,
- regulations to prevent the significant deterioration of air
quality, and
- limitations necessary to attain air quality standards and protect
public health.
As you appreciate, each of these areas will have an effect on the
demand, or market, for clean coal combustion technologies. Moreover, as
you might expect the technical capability of clean technologies will
greatly influence the emission regulations in each of these cases.
Over the next 15 years coal will become an increasing energy
resource. The shift to coal is already occurring due to the existing
scarcity and price of oil and natural gas. The President's proposed
National Energy Plan is to encourage the substitution of coal for gas
and oil through taxes on oil and gas and through investment tax credits
for the purchase of coal handling and burning equipment, including
pollution control.
With increased coal use comes increases in atmosphere emissions of
a number of pollutants—sulfur oxides, particulate matter, nitrogen
oxides, trace metals, radioactive materials, and organic compounds. The
potential damages are very real; they include:
- Impaired health and premature death from respiratory ailments;
- Stunted growth of crops and other plants;
- Reduced resistance of plants and animals to disease.
-------
22 CLEAN COMBUSTION OF COAL
Air quality standards and pollution emission regulations are intended to
protect against these damages.
New Source Performance Standards
The application of clean combustion technologies at new facilities
will TDe important to minimize these effects. In this regard, the Envi-
ronmental Protection Agency prescribes nationally applicable standards
of performance for large new sources of air pollution. The 1970 amend-
ments to the Clean Air Act specified that the new source performance
standard, NSPS for short, was to reflect the degree of emission limita-
tion achievable through the application of the best, adequately demon-
strated system of emission reduction taking the costs of achieving the
reduction into account.
Under this authority, EPA promulgated new source performance stand-
ards for large coal-fired boilers in 19T1- Considering the regulation
for sulfur dioxide presently in effect, the specific regulation of
1.2 pounds of sulfur dioxide per million Btu of heat input can be met by
burning low sulfur coals or by burning high sulfur coal with pollution
control equipment. Although such a regulation may have been appropriate
for the early 1970's when commercial use of pollution control equipment
and clean combustion technologies was just emerging, Congress has stated
that it considers a standard which permits the use of untreated fuels to
be inadequate and in direct conflict with the purposes of new source
standards. Accordingly, in the Clean Air Act Amendments considered in
1976 and proposed again this year, the House of Representatives and the
Conference Committee have adopted legislative language which would
require that the new source standard for new fuel-burning stationary
sources require a specified reduction in emissions which is achievable
when applying the best control technology. In establishing the standard,
EPA would be required to determine which technologies were adequately
demonstrated and to consider cost impacts and the energy and non-air
quality impacts before setting a regulation.
Anticipating the passage of the amendments this year and considering
recent advances in control technology development, EPA has already
initiated a review of the new source standards for electric power plants.
In what is one of the most rigorous analyses performed by the Agency,
EPA is evaluating the technical performance (primarily reliability and
removal efficiency) of recent flue gas desulfurization systems. In
addition, the Agency is assessing potential coal production shifts and
estimating the economic impact on the electric utility industry and its
customers. While several alternative emission regulations are being
considered, one guess at the new standards would be a requirement for
90 percent removal of sulfur dioxide with an emission ceiling somewhere
between 0.5 and 1.2 pounds of sulfur dioxide per million Btu, a regula-
tion of 0.05 pounds per million Btu (one-half the present standards) for
suspended particulates, and an incremental tightening to 0.6 pounds per
million Btu for NOX. The new standards for power plants are expected to
be proposed in early 1978.
The implications of these standards, especially for sulfur dioxide,
for developing clean combustion technologies are significant. They set
both a performance and a cost target with which other technologies must
-------
FUTURE EMISSION REGULATIONS 23
compete. For example, to "be allowed as a substitute for a flue gas
desulfurization (FGD) system, a sulfur removal technology would have to
achieve comparable or better sulfur removal efficiency. Use of coal
washing alone would not be suitable unless it were combined with another
technique to achieve the required overall reduction in emissions. In
effect, this could mean that those technologies which are not capable of
high removal and are not economically competitive when used in combina-
tion with other technologies would be foreclosed from use on new coal-
fired power plants.
While requiring the maximum available degree of emission reduction
from new sources, it is hoped that development of innovative means of
achieving equal or better degrees of emission control would also be
encouraged. In this regard, it has been EPA policy to allow special
consideration for developing technologies. To encourage innovative
technology, Congress has also proposed a formal variance for new tech-
nologies for up to 10 years from the federal new source performance
standards. Only those technologies with promise of achieving equivalent
or greater emission control than that required under the NSPS would be
eligible for a variance.
Continued progress in control technology development for each
pollutant is expected to lead to competition in both performance and
costs. At some date in the future, with the demonstration of greater
emission removal efficiency at a reasonable cost, EPA would again revise
the new source performance standards to reflect new technical capabili-
ties.
Prevention of Significant Deterioration
At this point I would like to discuss the regulations to prevent
the significant deterioration (PSD) of air quality in areas of the
country which are presently enjoying clean air. As you all probably
know, the Supreme Court upheld a lower court decision that required EPA
to promulgate regulations to prevent the significant deterioration of
air quality. These regulations were promulgated in December 197^ after
extensive technical analyses and public participation and provide air
quality designations which States may use to manage air resources.
Briefly, the EPA regulations establish three classifications of air
quality levels based on how much increase will be permitted in ambient
concentrations of particulate matter and sulfur dioxide.
Class I - Pristine areas where deterioration in any air quality
deterioration would be considered significant.
Class II - Areas where deterioration in air quality that would
normally accompany moderate growth would not be considered significant.
Class III - Areas where intensive and concentrated major industrial
growth is desired, thus permitting levels of air quality to rise to the
national ambient air quality standards.
EPA's regulations initially designated all areas of the country as
Class II, allowing states to redesignate areas to either Class I
-------
24 CLEAN COMBUSTION OF COAL
(pristine) or Class III (intensive economic growth). The regulations
passed, by the Senate and House are similar in general approach, although
numeric limits, Class I designations,-and sources covered vary.
Briefly, the regulations are intimately tied to the capability to
control emissions of pollutants. Limits on size of facility or possible
constraints on siting a large facility within the requirements of each
air quality designation depends upon the ambient ground-level concentra-
tions produced by the facility. Since ground-level concentrations are
roughly proportional to emission rates, any constraints on a facility
translate into a need for greater emission control.
For example, consider a national park or wilderness area which is
protected as a Class I, or pristine, area. In order to preserve the
clean air values of this area, a 1000 Mw coal-fired power plant meeting
the present NSPS for sulfur dioxide would have to locate approximately
60 miles from the Class I boundary. Depending upon the emission rate
and the specific terrain in the Class I area, a 1000 Mw plant could
locate less than 10 miles from the park or wilderness area. A signifi-
cant portion of this distance reduction would be accomplished by the
planned revision in the NSPS for sulfur dioxide. And, in reality, the
demand for additional emission control due to PSD requirements may be
small. It will, however, be one option considered by large power plants,
synthetic fuel plants, and other facilities which wish to locate near
Class I areas or near mountainous terrain.
Attainment of Air Quality Standards
The last topic I would like to address concerns the measures neces-
sary to protect public health in areas which presently have poor air
quality.
The 1970 Clean Air Act required each State to develop and implement
a State Implementation Plan (SIP) to attain national, health-related air
quality standards. The Act provided for attainment of the air quality
standards by mid-1975 • 1975 has come and gone and there remain a number
of areas, primarily metropolitan areas, which are not attaining these
health standards. Further, there are other areas which may have diffi-
culty maintaining air quality standards as industry and population grow.
These problems are cause for considerable concern from the stand-
point of public health. If we are to progress toward attaining our
health standards we are faced with either requiring greater emission
control on existing sources or limiting the construction of new sources
in these problem areas.
Although the nonattainment problem is most widespread with respect
to photochemical oxidants and particulates, NOX and sulfur dioxide are
also problems in specific cases. To manage new growth in these areas,
EPA adopted an interpretative ruling in December 1976. Referred to as
an "emission offset policy," the ruling allows new sources to locate in
areas not attaining health standards if the new facility meets the
lowest achievable emission rate (comparable to best controls on fuels
of moderate or low pollution potential) and a decrease in emissions from
existing sources more than offsetting the emission increase from the new
-------
FUTURE EMISSION REGULATIONS 25
facility. The emission decreases can be from other resources operated
by the owner of the new facility or from existing facilities operated
by others.
The EPA policy is one approach which will lead to attainment of air
quality standards. In effect, the policy creates an incentive for
States and local communities to adopt and enforce emission regulations
which will attain the air quality standards.
EPA has issued calls to a number of States to revise their SIPs to
assure attainment of air quality standards as soon as possible. In
addition, the Agency has designated at a number of other areas as air
quality maintenance areas'. These areas may require some adjustment in
emission regulations to assure maintenance of air quality due to antici-
pated growth in population and industrial activity between now and 1985.
In formulating SIP emission limits or in meeting the requirements
of the emission offset policy a central question is the technical feasi-
bility and cost of different levels of control on different types of
sources. Thus, here, as with NSPS, the technical feasibility and afford-
ability of clean combustion technologies will determine the nature of
emission limits. The best technologies for new sources would be relied
upon to achieve the lowest possible emissions and thereby minimize the
air quality effects of new facilities. Moreover, the available technol-
ogies will establish the bounds for the emission regulations which could
apply to existing and smaller facilities. In this latter case, the pre-
combustion treatment of fuels and those technologies that are economic
to retrofit will be particularly important.
Conclusion
In closing, I would like to read a paragraph from the National
Academy of Science publication on Man, Materials, and Environment,
(pp. 8-9):
Many foreseeable problems cannot now be solved by available
technology. Even if we control 99-5 percent of some pollut-
ants, the remaining one-half of 1 percent, because of the
large absolute amounts projected by the year 2000, can create
environmental problems for which a workable remedy has not
yet emerged from the laboratory.
-------
26 CLEAN COMBUSTION OF COAL
-------
27
RECENT DEVELOPMENTS IN COAL COMBUSTION TECHNOLOGY
by
James I. Joubert
Pittsburgh Energy Research Center
U. S. Energy Research and Development Administration
ABSTRACT
This paper considers three distinct areas of research and develop-
ment directed toward increasing the utilization of coal by industry and
electrical utilities: coal-oil slurry combustion; solvent-refined-coal
combustion; and coal-based magnetohydrodynamic pover generation. Current
research programs at the Pittsburgh Energy Research Center are discussed,
and an assessment of technical problems that require solution is presented.
Finally, an attempt is made to predict the potential impact of each of
these technologies on the U. S. energy supply situation.
I. INTRODUCTION
The increasing dependence of the U. S. on foreign supplies of
petroleum and the growing scarcity of domestic natural gas have placed
increased emphasis on developing technology to utilize our vast coal
resources in an environmentally acceptable manner. Three approaches to
accomplishing this objective are discussed in this paper.
The first approach considered involves retrofitting existing
utility and industrial boilers to burn coal-oil slurry. The use of
coal mixed with heavy fuel oil could conserve 25 to 35 percent of the
oil now used for steam raising, and could be implemented almost immediately.
The advent of low-sulfur, low-ash fuels derived from coal, such as
solvent refined coal (SRC), could make an impact on our energy supply
situtation in the mid-term (1985-2000). Use of such fuels would be
preferable to the use of fuel oil or natural gas for power generation,
and may provide an attractive alternative to flue gas scrubbing.
The development of coal-based magnetohydrodynamic (MHD) power
generation holds great promise for the period 1990 and beyond. This
approach to utilizing coal offers a highly efficient means for gen-
erating electricity with minimal degradation of the environment.
The combustion-related aspects of each of these approaches to coal
utilization will be considered. Current research programs in coal-oil
slurry combustion, SRC combustion, and MHD combustor development at the
Pittsburgh Energy Research Center (PERC) are reviewed, and the status
of each of these technologies is assessed.
-------
28 CLEAN COMBUSTION OF COAL
II. COAL-OIL SLURRY
The concept of burning coal and oil together as a slurry is not
novel. Numerous reports dealing with the handling and combustion
characteristics of slurries have been published, the earliest appearing
nearly 100 years ago. Research on coal-oil slurries was particularly
active during the period between World War I and World War II*.
The main incentive for utilizing coal-oil slurries in lieu of fuel
oil alone is simply that a significant reduction in oil usage can be
achieved by substituting coal for oil. Systems designed for feeding
and burning liquid fuels can still be employed, with some modifications.
For a slurry containing 40 weight percent coal, a fuel oil savings
of 26 to 34 percent is possible for coals ranging in heating value from
10,000 to 14,000 Btu/lb. Based on earlier studies by the U. S. Bureau
of Mines and Kansas State College, the maximum allowable^qncentration
of coal appears to fall in the range of 40 to 50 percent ' .
For a 1 percent sulfur coal, a slurry containing about 50 percent
coal could be used without violating EPA emission standards for SO^,
assuming a 13,000 Btu/lb coal. With a 2 percent sulfur coal, however,
stack gas scrubbing would be necessary if the slurry coal-content exceeded
about 17 percent.
A number of potential problems may arise in attempting to convert an
existing oil-fired facility to coal-oil slurry firing. These include: 1)
erosion and/or corrosion of piping, valves, flow meters, pumps, and burners;
2) erosion or fouling of boiler tubes; 3) deposition of solids in various
system components; and 4) particulate and gaseous emissions. An additional
concern results from the tendency of the coal to settle out of the slurry
when stored; efforts to develop an effective and inexpensive "stabilizer"
have been only partially successful ' . While a fair amount of success
was achieved in early short-duration combustion tests with slurries (see
Reference 1), not all of the problem areas listed above have been addressed.
A coal-oil slurry combustion research program was initiated at PERC
in June 1975 . The initial objectives of the program were to delineate
problems associated with firing coal-oil slurry in a package boiler designed
to burn oil or gas, and to determine the effect of coal ash on the boiler
components and feed system over an extended period (1000 hours). The
boiler used was a 100 HP Cleaver-Brooks firetube type with four passes.
The fuel used throughout the test period consisted of a 20 percent by
weight concentration of Pittsburgh seam coal in No. 6 fuel oil. The coal
size consist was nominally 90 percent minus 200 mesh.
The major difficulty that persisted throughout the test period was
formation of carbonaceous deposits in the refractory-lined burner zone in
the combustion chamber, resulting from impingement of the flame on the
refractory. It was necessary to shut the boiler down frequently to remove
the deposits.
*An extensive bibliography of literature relating to coal-oil slurries
is given in Reference 1.
-------
COAL COMBUSTION TECHNOLOGY 29
Increasing atomizing air pressure and excess air level was only
partially successful in limiting the rate of deposition. Reduction of
load to about 80 percent of capacity also increased operating time.
Further improvement was achieved by modifying the nozzle to provide a
narrower flame pattern which reduced impingement. However, deterioration
of the spray pattern occurred due to erosion of the orifices in the original
brass nozzle, and in some instances the nozzle had to be replaced after
100 hours of operating time.
After completion of the 1000 hour test, the boiler was opened and the
internal tube surfaces were inspected. Only a light coating was evident,
and there was no indication of slag deposition on any of the surfaces.
The test tubes installed in the second, third, and fourth passes were
removed from the boiler and sent to an independent testing laboratory for
metallographic examinination. The analysis indicated that there was no
evidence of intergranular or subscale corrosion attack. Estimated cor-
rosion rates were 0.94, 0.68, and 1.2 mils per year in the second, third,
and fourth passes, respectively; these are considered to be acceptable
rates.
Subsequent to the 1000-hour test, the 100 HP test facility was
upgraded with respect to instrumentation and controls. Problems due to
formation of carbonaceous deposits in the burner zone have now been
completely overcome as a result of modifications to the burner diffuser.
The unit was recently operated for 65 continuous hours burning a slurry
containing 30 percent coal. The test was considered satisfactory in all
respects, and an efficiency of 79.9 percent was achieved at full boiler
load.
A comprehensive research program has now been initiated at PERC
directed toward evaluation of the flame characteristics of coal-oil
slurries. Combustion aerodynamics and heat transfer in the combustion
zone of the 100 HP unit (first pass) will be examined in detail and
existing mathematical models will be modified and applied to the system to
aid in interpretation of data. Coal concentration, coal size-consist, and
boiler load will be varied systematically and an attempt will be made to
develop fundamental design criteria for slurry-fired systems. In addition,
laboratory studies of slurry rheological properties will be conducted, and
various slurry stabilizers will be tested.
Much of the experience gained to date with the 100 HP boiler has
aided in the design of a larger coal-oil slurry test facility now under
construction at PERC. This facility will include a 700 HP watertube
boiler which is more representative of the type of boiler employed in
industry. The 700 HP combustion test facility will be highly instrumented
and equipped with a data acquisition system. Shakedown tests are expected
to begin in March 1978.
In addition to the program at PERC, ERDA is sponsoring several
industrial organizations in large-scale testing of coal-oil slurry .
Studies have also been conducted by a consortium of companies headed by
General Motors, with support from ERDA .
-------
30 CLEM COMBUSTION OF COAL
III. SOLVENT REFINED COAL
Solvent Refined Coal (SRC) is a Ipwjrsulfur, low-ash solid with a
solidification point of 300°F to 400°F . The technology for producing
SRC has been successfully piloted at a 6 ton/day plant in Wilsonville,
Alabama, and at a 50 ton/day plant in Fort Lewis (Tacoma) Washington.
The properties of SRC are remarkably unaffected by the quality of the
original feedstock. Typically, the final product contains 0.5 to 0.8
percent sulfur, less than 0.25 percent ash, and has a heating value of
15,500 to 16,000 Btu/lb. A fuel with these specifications would meet
federal standards for SO emissions from new stationary sources (1.2 Ib
SO-/10 Btu for a coal-fired station). Utilization of SRC in industrial
or utility boilers would eliminate, or at least minimize, the need for
elaborate flue-gas cleanup systems.
Early studies (1964-1965) of the handling and combustion charac-
teristics of SRC by Babcock and Wilcox, Combustion Engineering, and the
U. S. Bureau of Mines were inconclusive ~ . At that time, only limited
quantities of SRC were available and the testing programs were too brief
to resolve difficulties encountered. However, all investigators reported
problems with respect to pulverizing and conveying the fuel, and burner
.fouling, when attempting to burn SRC as a solid. To fire SRC as a liquid,
it was necessary to heat the fuel to 700-800 F, which resulted in evolu-
tion of vapors at about 350 F and severe foaming at about 650 F.
Research on the combustion and handling of characteristics of SRC
were initiated at PERC in 1974. Analysis of the earlier studies of SRC
combustion indicated that firing of the fuel as a solid would present
fewer difficulties than firing as a liquid. With solid SRC, no external
heating of pumps, transport lines, or fittings is required, and problems
with respect to devolatilization or foaming of the liquid are avoided.
The SRC used in this investigation was obtained from the Southern
Services, Inc. pilot plant in Wilsonville, Alabama and from the Fort
Lewis, Washington, pilot plant operated by Pittsburg and Midway Coal Co.
for ERDA.
In contrast to earlier investigations of SRC properties, no problems
were encountered in pulverizing or transporting the fuels used in this
study. An impact mill was used rather than a hammer mill or ball-and-race
mill; attempts to use the latter two types had resulted in mill plugging
due to softening and agglomeration of the SRC. The grind obtained was
about 90% through 200 mesh, somewhat finer than that usually obtained when
pulverizing coal in the same mill.
Initial attempts to burn SRC with the burner usually used in coal
combustion studies resulted in clogging of burner passages due to SRC
melting upon contact with hot burner surfaces. Water cooling was provided
to all surfaces which contacted the fuel prior to entry into the furnace;
a water-cooled deflector was added to protect the oncoming fuel stream
from flame radiation; and port velocity was increased to 100 ft/sec.
These modifications permitted trouble-free operation.
-------
COAL COMBUSTION TECHNOLOGY 31
Most of the SRC combustion tests* were conducted at fuel rates of 425
Ib/hr (=6.5 million Btu/hr) at stoichiometric air levels ranging from 115
to 140%, and with air preheat temperatures of 540°F. In almost all tests,
carbon combustion efficiencies in excess of 99% were achieved (conversion
was 98^5% at 115% S.A.). NO emissions ranged from about 0.45 to 0.72 Ib
NO^/IO Btu, increasing with increasing stoichiometric air levels. Com-
pliance with EPA emission standards of 0.7 Ib N0_/10 Btu could easily be
achieved by operation at less than 130 percent of stoichiometric air.
Recent studies of SRC combustion have also been carried out by
Babcock and Wilcox , and Combustion Engineering . B&W reported that SRC
could be pulverized easily with an impact mill but that extreme diffi-
culties were encountered with a ball-and-race mill; combustion tests were
successfully conducted using a water-cooled burner. Combustion Engineering
carried out a 100 hour combustion test with SRC with no major problems.
It was concluded that SRC could be pulverized readily using a C-E bowl
mill, and that it could be fired using conventional fuel admission assem-
blies provided with water cooling.
A large-scale combustion test**using SRC has recently been completed
at Plant Mitchell owned by the Georgia Power Co. The 22.5 MW B&W boiler
used was equipped with modified B&W dual-register burners. The test was
considered a success in all respects with no problems encountered in
pulverizing, transporting, or burning of the SRC. SO and NO emissions
were substantially less than EPA standards for solid fuel firing.
IV. COAL COMBUSTORS FOR MHD POWER GENERATION
A promising technique for improving the efficiency in converting
thermal energy into electricity involves the principles of magnetohydro-
dynamics (MHD). In an MHD power plant, electricity is generated by
passing a high-temperature, electrically conducting fluid (or plasma)
through a magnetic field. The interaction of the conducting fluid with
the magnetic field results in a flow of electrons which can be collected
by electrodes on the walls of the MHD generator.
To enhance the electrical conductivity of the plasma, a "seed" mate-
rial, such as potassium carbonate, is injected into the high-temperature
gases produced in the combustion system. In coal-fired systems, it
has been shown experimentally that the potassium reacts with SO- to
produce K^SO, which can be collected as a solid in downstream components.
The K9SO, can be regenerated to K-CO,,, thereby recovering the seed and pro-
ducing elemental sulfur. The economic requirement for recycling seed also
results in a built-in pollution control technique. Nearly quantitative
removal of sulfur oxides can be achieved.
The combustion products leaving the MHD generator are at a suffi-
ciently high temperature to generate additional power in a "bottoming"
steam-turbine plant. -The recently completed Energy Conversion Alter-
natives Study (ECAS) concluded that a combined open-cycle MHD-steam
it
Detailed experimental results are reported in References 13 and lU.
f4&
Sponsored by ERDA in cooperation with Southern Company Services, EPRI,
Pittsburg and Midway Coal Company, and Babcock and Wilcox.
-------
32 CLEAN COMBUSTION OF COAL
power plant offers overall efficiencies greater than 50 percent as compared
to the predicted performance of advanced "conventional" steam plants of
about 40 percent.
There-is now an active ERDA-sponsored national MHD development
program ' . The goal of the program is to place in operation a com-
mercial scale demonstration plant, fueled by coal, by the 1990's.
The discussion here will be confined to MHDRcoal combustion systems.
The reader is referred to recent publications ' ' for a description of
the overall national MHD program and development activities involving
other MHD system components.
The development of high-temperature combustion systems for MHD appli-
cations was begun in the early 1960's . Most of the progress to date has
resulted from programs in Great Britain, the USSR, and the USA. The work
in Great Britain, discontinued in 1968, was geared to the use of coal and,
to a lesser extent, oil. Efforts in the USSR have been directed primarily
toward development of natural gas combustors, although some research in
coal combustion has also been conducted. In the United States, current
research is oriented toward development of coal-fired systems exclusively.
A. Combustor Design Considerations
Combustors for open-cycle MHD power plants are actually plasma
generators. Although similar to conventional combustion systems in some
respects, there are certain aspects of MHD combustors that are unique. An
MHD combustor must be able to meet the following design requirements:
1. Burn natural gas, oil, or coal with high efficiency (ap-
proaching 100 percent).
2. Produce a working fluid with a temperature of 2700-3100 K
(4400-5100°F).
3. Operate at pressures ranging from 5 to 10 atmospheres.
4. Accommodate highly preheated oxidizer at a temperature of
1800-2000 K (_2780-3140°F) , and possibly up to 2300 K
(3680°F) in future plants .
5. Exhibit low heat losses (preferably 5 percent or less of
the fuel thermal input).
6. Incur low pressure losses (preferably less than 10 percent
of the oxidizer inlet pressure).
7. Accommodate injection of seed in dry or aqueous form
(preferably dry).
8. Provide a high degree of seed ionization (approaching 100
percent).
9. Produce a uniform plasma.
-------
COAL COMBUSTION TECHNOLOGY 33
10. Operate with low levels of instabilities.
11. Have a long operating life (many thousands of hours).
12. Operate under conditions conducive to control of nitrogen
oxides; the formation of nitrogen oxides must be suppressed
sufficiently to allow their decomposition as the gases pass
through the MHD plant.
In addition to meeting these general requirements, combustors oper-
ating on coal must be designed to cope with liquid slag. Materials of
construction must be able to withstand slag attack. If slag is to be
rejected from the combustor proper, a suitable means for tapping the
molten material must be incorporated into the design. Also, the inter-
action of seed and slag must be minimized to prevent excessive losses of
seed.
Due to the electrical generating nature of the MHD channel, the
combustor will have a high potential, which in the case of commercial-
scale units may be 25,000 to 35,000 volts. This results in the necessity
of electrically insulating the combustor from all supports and feed lines
(fuel, oxidizer, seed, cooling water), as well as from the slag rejection
system.
Combustors designed for MHD applications may consist of one or more
stages. When burning natural gas or fuel oil, there is no incentive to
use other than a single-stage configuration. With these fuels, satis-
factory performance has been obtained in experimental single-stage com-
bustors operating under predominantly plug flow conditions. When burning
coal, the selection of a combustor configuration will be dictated by
technical and economic considerations with respect to the level of slag
carryover that is permissible in downstream portions of the power plant.
The presence of slag in an MHD plasma has a negative effect on
plasma conductivity, results in loss of seed material, and may seriously
affect the operation and integrity of downstream components. While avail-
able evidence suggests that low slag carryover from the combustor is
desirable, it is not possible at present to say what constitutes a toler-
able level of carryover. Depending on the combustor design chosen, as
much as 100 percent or as little as 10 percent of the mineral matter in
the coal will pass through the MHD generator and be deposited or collected
in downstream equipment. The fundamental differences in combustor design
that most strongly affect the degree of slag rejection achieved are dis-
cussed below.
Single-Stage Coal Combustors
Single-stage combustors offer the advantage of simplicity of design
and operation, and lower cost relative to multistage combustors; also,
heat losses will probably be lower in single-stage units. They may oper-
ate with 100 percent slag carryover, or with partial removal of molten
ash from the combustor proper.
-------
34 CLEM COMBUSTION OP COAL
If no slag is rejected from the combustor, a method must be developed
to remove slag from the combustion gases downstream of the generator to
avoid damage to other plant components. To achieve this without signifi-
cant seed loss represents a major technical problem. An advantage of this
mode of operation, however, is that the difficulty of tapping slag from a
combustor at high pressure, temperature, and voltage is avoided.
In a single-stage combustor operating at temperatures in excess of
2700 K (4400°F), a major portion of the coal mineral matter will appear as
vapor in the plasma*. Hence, the amount of liquid slag that can be tapped
from the combustor proper will be limited. Also, there will be some
degree of seed loss to the tapped slag in addition to that lost by reac-
tion with slag carried over from the combustor. Contact of seed and slag
in the combustor could be avoided by injecting the seed in a separate
chamber following the combustor. However, this would lead to additional
system heat losses.
Two-Stage Coal Combustors
In a two-stage coal combustor, slag rejection levels approaching 90%
are possible. In such a system, the first stage is operated as a coal
gasifier at temperatures ranging from 2200 to 2300 K (3500-3700°F). At
these temperatures, very little of the coal slag is vaporized, and the
bulk of the liquid slag can be separated from the gas using a first stage
designed similar to a conventional cyclone combustor. The gas produced in
the first stage is burned in the second state where seed is injected.
Operation of a two-stage combustor is more complex than in the case
of a single stage unit. Additional high-temperature air lines, controls,
and instrumentation are required. It is also likely that thermal losses
will be somewhat higher in a two-stage system because of added system
surface area.
Multistage Coal Combustors
Many conceptual studies of-multistage MHD combustion systems have
been reported in the literature . Several of these systems offer
potential slag rejection levels in excess of 95 percent. In general, the
systems are highly complex, and represent novel technology that must be
further developed before their feasibility is ascertained.
B. Current MHD Coal Combustor Development Programs
Pittsburgh Energy Research Center
Combustor development work at PERC is presently focusing on a two-
stage design. Studies are being conducted with an atmospheric-pressure
vertical cyclone that represents the first stage of a two-stage combustion
system. Major objectives of the program include: 1) evaluation of
cyclone combustors operating under fuel rich conditions with coals of
*For example, in burning Pittsburgh seam coal (10% ash) with 100% of
stoichiometric air at 8 atm, thermodynamic calculations indicate that
about 40% of the slag is vaporized at 2800 K, and 75% is vaporized at
3000 K.
-------
COAL COMBUSTION TECHNOLOGY 35
different ranks; 2) testing of high-temperature plasma diagnostic tech-
niques; 31 studies of seed-slag interactions; 4) evaluation of materials;
and 5) development of techniques for controlling nitrogen oxides emissions.
The size of the facility, and its atmospheric-pressure operation, allow
rapid and inexpensive changes to be made in hardware design thereby per-
mitting several design options to be tested during short periods.
During the past year, three different cyclone configurations were
evaluated. Tests were carried out at coal feed rates ranging from 0.01 to
0.03 kg/s, oxidizer/fuel ratios ranging from 55 to 110 percent of stoi-
chiometric, and (vitiated) oxidizer temperatures of 1100-1360 K (1500-
2000 F). Primary emphasis has been placed on optimizing cyclone perform-
ance with respect to slag rejection and carbon conversion. A Montana
Rosebud seam coal has been used in most of the tests.
The tests conducted thus far indicate that coal size consist plays a
major role in determining slag rejection and carbon conversion. Slag
rejection rates as high as 90 percent and carbon conversions greater than
98 percent have been achieved under substoichiometric conditions.
A separate facility has been constructed at PERC to test MHD com-
bustors (nominal 5 MW thermal input) at pressures up to 8 atmospheres.
The facility is capable of continuous operation and includes equipment
for coal pulverizing, drying, conveying, and feeding under pressure.
A two-stage combustor has been installed in the facility. The
first stage, a vertical cyclone, will be operated as a coal gasifier.
The second stage is a gaseous fuel combustor.
The maximum total flow rate attainable in the facility is 1.7
kg/s. An air preheat temperature of 1200 K (1700 F) will be possible,
with provisions available for vitiated air (or 0«-enriched air) at
temperatures up to 1866 K (2900°F). Initial tests will be conducted
at 6 atm pressure, a coal thermal input of 2-3 MW, and a total mass
flow of 1.2 kg/s.
The objectives of the test program are similar to those indicated
for the atmospheric-pressure facility. However, the operation is more
complex due to the high pressures involved and the two-stage configura-
tion. Also, an instrumented test section, designed and supplied by
Avco Everett Research Laboratory, will be installed downstream of the
second-stage combustor to permit measurements of plasma properties in
time and space including temperature, pressure, velocity, and electrical
conductivity.
Component Development and Integration Facility (CDIF)
The Energy Research and Development Administration is presently
constructing a major MHD test facility in Butte, Montana. The Component
Development and Integration Facility (CDIF) is designed to permit long
duration testing of major components required in commercial MHD plants
of the future. The test program will allow resolution of many technical
problems associated with direct use of coal, and will provide valuable
scale-up data for the design of a complete prototype MHD power plant.
Initial testing at the CDIF is expected to begin in late 1978.
-------
36 CLEM COMBUSTION OF COAL
The facility will have a thermal rating of 50 MW; the fuel thermal
input will be 38 MW and the balance will be due to sensible heat of
the oxidizer. Total mass flow rate will be 9.6 kg/s. Initial tests
will be conducted using vitiated air at 1866 K (2900°F); high-temperature
regenerative air heaters will be added at a later date. The test
train will be operated at pressures ranging from 3 to 10 atmospheres.
The first coal combustor tested will be supplied by the Pittsburgh
Energy Research Center. The combustor will be a two-stage design
similar to the unit presently being tested at PERC.
University of Tennessee Space Institute (UTSI)
A new MHD test facility is being constructed at UTSI in Tullahoma,
Tennessee, and is expected to be operational in 1979. The purpose of
the facility, in addition to generator development, is to provide
design data for single-stage coal-fired combustors. Problems associated
with coal handling, downstream slag removal, and seed-slag separation
will be investigated. A major objective is to determine the effects of
long-duration operation on component performance.
Two plug flow combustors will be tested. One unit, with a mass
throughput of 3.6 kg/s, will be capable of continuous operation. Total
thermal input will be 23.7 MW (21.1 MW coal). Oxygen and vitiated air
will be used as oxidizers. The combustor will operate at a pressure of
8 atm and a temperature of approximately 3000 K.
Short duration tests (several hours) will be conducted with a
larger combustor having a mass throughput of 13.6 kg/s. Coal will
provide 56.2 MW of the total thermal input of 66.8 MW. This combustor
will also operate at a pressure of 8 atm and will generate a plasma at
about 2800 K.
V. DISCUSSION AND CONCLUSIONS
Three different approaches to utilizing coal to reduce our depend-
ence on natural gas and foreign oil have been discussed. An attempt
will now be made to assess the status of each of these technologies, and
to determine their potential impact on our national energy supply
situation.
Coal-Oil Slurry
There appears to be no technical reason why coal-oil slurry could
not be utilized now in existing oil and/or gas fired boilers. Consid-
erable success has been achieved in firing slurries in industrial units
for short periods. Tests at PERC indicate that erosion and corrosion
would not appear to be a problem in retrofitting industrial gas/oil
boilers to fire a slurry.
Some additional research is required to better define the effects
of coal rank, coal concentration, and coal particle size consist on com-
bustion characteristics of slurries. The development of an effective,
-------
COAL COMBUSTION TECHNOLOGY 37
inexpensive stabilizer would greatly enhance the attractiveness of slur-
ries; coal-oil mixtures could then be prepared in one location and
transported as other liquid fuels are, thereby obviating on-site prepara-
tion of coal by the user.
In general, utilization of slurries in lieu of low-sulfur oil or
natural gas will produce air pollution concerns. Particulate emissions
will have to be dealt with, as will SO emissions unless a low-sulfur
coal is used. However, adequate particulate removal equipment is avail-
able, and effective flue gas desulfurization technology is being developed.
Wide spread utilization of coal-oil slurry by utilities could
potentially produce a savings of about 500,000 bbl/day of fuel oil, based
on 1975 oil consumption data* . Considering only the fuel oil used by
industry to generate steam in boilers , an additional 400,000 bbl/day
could be conserved. If one also considers direct heating operations, it
is likely that a total savings of fuel oil of well over one million
bbl/day could be achieved.
It appears that the incentive for a particular industrial user to
convert to coal-oil slurry will simply be to conserve oil supplies, or
to have fuel available in the event of natural gas curtailments. At
present the estimated cost of slurry per million Btu is nearly equiva-
lent to that for fuel oil. However, should costs for foreign oil
escalate (and assuming coal costs remain stable), use of coal-oil slurry
could also provide a significant savings in fuel costs.
Solvent Refined Coal
SRC appears to be a very attractive fuel for utility and industrial
applications. Problems in handling and burning SRC have been overcome,
and the technology for producing SRC is close to being demonstrated on a
commercial scale.
The use of SRC provides an alternative to flue-gas desulfurization
which can be a significant consideration with respect to capital costs
for a new coal fired power plant. Since the properties of SRC appear to
be uniform regardless of the coal feedstock, economies in design and
construction of new power plants can be achieved.
SRC could be used directly in many existing power plants which were
designed for coal but are now burning low-sulfur oil to avoid the use of
flue-gas scrubbers However, the major impact of SRC will be through its
use in future (1985 and beyond) power plants which will undoubtedly be
prevented from using natural gas or low-sulfur oil.
While the cost for producing SRC, or any coal-derived fuel is
uncertain, it may well be competitive with fuel oil . Costs, however,
will likely become a secondary concern in future years when the choice
is between an assured domestic supply of fuel on the one hand, and an
uncertain foreign supply on the other.
*Assuming 40% coal/60% oil slurry.
-------
38 CLEAN COMBUSTION OF COAL
Coal Fired MHD Combustors
The attractiveness of MHD as an efficient, pollution-free process
for converting coal to electricity provides a strong incentive for
further development of this technology. Although some degree of success
has been attained in the development of MHD combustors, a considerable
amount of additional development work must be carried out before com-
mercial-size units are a reality. Many uncertainties exist with respect
to scale-up from present experimental information.
A commercial MHD power plant of, say, 2000 MW (thermal) capacity
will have a combustion products mass flow rate of about 700 kg/s, and
will operate at a peak pressure of about 8-10 atm. However, experi-
mental combustors have operated at mass flow rates in the range of 0.2
to 3 kg/s. In general, experiments have been conducted at pressures
considerably below that required in commercial applications.
In a base-load MHD power plant, a single combustor (with some
amount of turndown capability) would probably be most desirable. Addi-
tional data must therefore be acquired to formulate reliable scaling
laws; this can only be accomplished by construction and operation of
combustors considerably larger than those tested to date, yet still much
smaller than a commercial unit. Key questions that must be answered
include the effect of scale-up on combustion stability, efficiency, and
heat losses.
It may be feasible to limit the magnitude of scaling required if
one considers the use of multiple combustors in a commercial plant.
Such an arrangement provides an added degree of flexibility in terms of
turndown capability. However, it is likely that the use of multiple
combustors will result in greater heat losses than those associated with
a single large combustor. In addition, the complexity of the total
combustion system is increased substantially due to the need for addi-
tional controls, hot-air piping, and fuel feed systems; and problems
related to combustion stability may be magnified. Even if the use of a
multiple-combustor system were practicable, the size of the individual
units would be substantially greater than combustors that have been
operated to date, and the need for further testing on a larger scale
would still be necessary.
Aside from concerns directly related to the combustion process,
there are additional issues that must be addressed in the development of
commercial MHD combustion systems:
1) reliable, long-term operation of MHD combustors must be demon-
strated. Typical continuous tests conducted thus far have been of short
duration (several hours or less);
2) reliable and durable systems for feeding large quantities of
high-temperature air, fuel, and seed into pressurized combustors must be
developed. The uniform injection of coal at rates of about 65 kg/s
(2000 MW thermal) represents a particularly challenging problem;
-------
COAL COMBUSTION TECHNOLOGY 39
3) techniques for electrically isolating a high-temperature com-
bustor from feed lines, cooling water lines, slag rejection system, and
supports must be developed;
4) optimum coal-fired combustor configurations must be established;
this will be governed primarily by the percentage of ash deemed desirable
to be carried over through the generator and the downstream portions of
the system. Single-stage combustors are likely to provide relatively
poor slag rejection, but have lower heat losses than multistage com-
bustors, whereas two- and three-stage combustors can probably operate
with slag rejection rates approaching 90%;
5) methods for minimizing heat losses from large-scale combustors
should be explored, such as air-cooling of combustor walls or transpira-
tion cooling of combustor refractories.
Many of the items discussed above are being addressed in programs
currently underway in the United States and the Soviet Union, or will be
as new test facilities become operational in the near future.
-------
40 CLEAN COMBUSTION OF COAL
REFERENCES
1. Demeter, J. J., C. R. McCann, G. T. Bellas, J. M. Ekmann, and
D. Bienstock, "Combustion of Coal-Oil Slurry in a 100 HP Fire-
tube Boiler", Pittsburgh Energy Research Center, PERC/RI-77/8,
May, 1977.
2. Barkley, J. F., A. B. Hersberger, and L. R. Burdick, "Laboratory
and Field Tests on Coal-in-Oil Fuels", Trans. ASME, 66, 185 (1944)
3. Jonnard, A., "Colloidal Fuel Development for Industrial Use",
Bulletin No. 48, Kansas State College, Engineering Experiment
Station, Manhattan, Kansas (1946).
4. Anon., "Coal-Oil Slurry Combustion", Energy Research and Develop-
ment Administration, FE-COSC-1, April 1976.
5. Cook, T. D., "Problems with Burning Coal and Oil Slurry in a
Packaged Oil-Fired Boiler", Fall Meeting, Amer. Flame Res. Comm.,
Philadelphia, November 17, 1976.
6. Harrison, W. B., "The Solvent Refined Coal Process: Potential-
ities and Problems", Short Course on Coal Characteristics and
Coal Conversion Processes, Pennsylvania State University,
May 23, 1974.
7. Schmid, B. K., "The Solvent Refined Coal Process", Symp. on Coal
Gasification and Liquefaction, University of Pittsburgh, August
6-8, 1974.
8. Anon., "Fossil Energy Program Report", ERDA 76-10, Energy Research
and Development Administration (1976).
9. Anon., "Coal Liquefaction", Quarterly Report July-September 1976,
Energy Research and Development Administration (1976).
10. Sage, W. L., "Combustion Tests on a Specially Processed Low-Ash
Low-Sulphur Coal", Babcock and Wilcox Report No. 4439, Prepared
for Office of Coal Research, July 1964.
11. Frey, D. J., "De-Ashed Coal Study", Interim Report to Office of
Coal Research, Contract No. 14-01-001, Prepared by Combustion
Engineering, Inc., September, 1964.
12. McCann, C. R., J. J. Pfeiffer, A. A. Orning, and W. H. Oppelt,
"Combustion Trials Spencer Low-Ash Coal", Pittsburgh Coal
Research Center, January, 1965.
13. McCann, C. R., J. J. Demeter, and D. Bienstock, "Combustion of
Pulverized, Solvent-Refined Coal", Presentation at Spring
Meeting of Combustion Institute Central States Section, Battelle
Columbus Laboratories, April 5-6, 1976.
-------
COAL COMBUSTION TECHNOLOGY 41
14. McCann, C. R., J. J. Demeter, and D. Bienstock, "Combustion of
Pulverized Solvent-Refined Coal", ASME Paper No. 76-WA/Fu-6
(1976).
15. Wagoner, C. L., et al., "Investigating the Storage, Handling,
and Combustion Characteristics of Solvent Refined Coal", Babcock
and Wilcox Monthly Reports to EPRI, RP-1235-1, January 6, 1975 -
June 16, 1975.
16. Borio, R. W., Z. J. Fink, G. J. Goetz, and J. C. Haas, "Solvent
Refined Coal Evaluation", Technical Report 2, Prepared for EPRI
by Combustion Engineering, Inc., June, 1976.
17. "Comparative Evaluation of Phase 1 Results from the Energy Conver-
sion Alternatives Study (EGAS)" Prepared by NASA for ERDA and NSF,
NASA TM X-71855, February 1976; "Evaluation of Phase 2 Conceptual
Designs and Implementation Assessment Resulting from the Energy
Conversion Alternatives Study (EGAS)", Prepared by NASA for ERDA
and NSF, NASA TM X-73515, April 1977.
18. Anon., "Fossil Energy Research Program of the Energy Research
and Development Administration, FY 1978", ERDA 77-33, April 1977.
19. Proceedings of Symposia, Engineering Aspects of Magnetohydrody-
namics, available from Dr. John Fox, Dept. of Mech. Eng.,
University of Mississippi, University, Mississippi 38677.
20. "Fuel and Combustion", Chapter 10 in Joint US-USSR Status Report
on Open Cycle MHD Power Generation, Energy Research and Develop-
ment Administration (in press).
21. Carrasse, J., "Chemical Recovery of Energy in a Combined MHD-Steam
Power Station", Proc. Int. Symp. on MHD Electrical Power Genera-
tion Salzburg, Vol. Ill, p 883 (1966).
22. Way, S., "Char Burning MHD Systems", Trans. ASME, J. Eng. Power,
p 345, July 1971.
23. Lacey, J. J., J. J. Demeter, and D. Bienstock, "Production of a
Clean Working Fluid for Coal-Burning, Open-Cycle MHD Power Genera-
tion", Proc. 12th Symp. on Eng. Aspects of MHD, Argonne, Illinois,
p VI.2.1 (1972).
24. Gannon, R. E., D. B. Stickler, and H. Kobayashi, "Coal Processing
Employing Rapid Devolatilization of Reactions in an MHD Power
Cycle" Proc. 14th Symp. on Eng. Aspects of MHD, Tullahoma,
Tennessee, p II.2.1 (1974).
25. Zinko, H., S. Linder, and J. Raunsborg, "A New Scheme for a Coal
Gasification-MHD Power Plant", Proc. 6th Int. Conf. on MHD
Electrical Power Generation, Washington, Vol. 1, p 105 (1975).
26. Brzozowski, W. S., J. Dul, and W. Pudlik, "New Concepts of Coal
Burning MHD Plants", Ibid., p 137.
-------
42 CLEM COMBUSTION OF COAL
27. Hoy, H. R., A. G. Roberts, and D. M. Wilkins, Chapter 4 in "Open-
cycle MHD Power Generation", J. B. Heywood and G. J. Womack,
Eds., Pergamon Press, Oxford (1969).
28. Zelinski, J. J., J. Teno, and L. F. Westra, "A Coal Combustion
System for MHD Generators", Proceedings 5th Intersociety Energy
Conversion Engineering Conference, Las Vegas, Vol. 1, p. 7-41
(1970).
29. Shanklin, R. V., L. W. Crawford, J. F. Martin, J. B. Dicks, W. D.
Jackson, C. R. Gamblin, and C. H. Tsai, "The UTSI Coal Burning
MHD Program", Proceedings 13th Symp. on Eng. Aspects of MHD,
Stanford University, p.II.8.1 (1973).
30. Dicks, J. B., L. W. Crawford, J. W. Muehlhauser, J. F. Martin,
N. L. Loeffler, and B. S. Arora, "The Direct-Coal-Fired MHD
Generator System", Proceedings 14th Symp. on Eng. Aspects of
MHD, University of Tennessee Space Inst., p.II.1.1 (1974).
31. Tager, S. A., E. V. Samiulov, I. B. Rozhestvensky, R. U. Talumaa,
and F. M. lakilevich, "Development and Investigation of High-
Temperature Combustor to be Used for a Solid Fuel MHD Generator
and Thermodynamic Analysis of Combustion Conditions", Proc.
Fifth Int. Conf. on MHD Power Generation, Munich, Vol. 1, p. 471
(1971).
32. Bienstock, D., R. C. Kurtzrock, R. J. Demski, and J. H. Field,
"Experimental Unit for Study of High-Temperature Combustion of
Coal for MHD Power Generation", Paper 62-WA-147, ASME Winter
Annual Meeting, New York (1962).
33. Bienstock, D. , R. J. Demski, and R. C. Kurtzrock, "High-Temperature
Combustion of Coal Seeded with Potassium Carbonate in the MHD
Generation of Electric Power", Bureau of Mines Rep. of Investiga-
tions 7361 (1970).
34. Bienstock, D., P. D. Bergman, J. M. Henry, R. J. Demski, J. J.
Demeter, and K. D. Plants, "Air Pollution Aspects of MHD Power
Generation", Proceedings 13th Symp. on Eng. Aspects of MHD,
Stanford University, p.VII.1.1 (1973).
35. Anon., "Quarterly Report of Foreign and Domestic Developments
Affecting Energy", Energy Research and Development Administration,
Planning and Analysis, May 21, 1976.
36. Locklin, D. W. , H. H. Krause, A. A. Putnam, E. L. Kropp, W. T. Reid,
and M. A. Duffy, "Design Trends and Operating Problems in Com-
bustion Modifications of Industrial Boilers", EPA-650/2-74-032,
April 1974.
37. Montgomery, J., "De-Polluting of Coal Before It Is Burned Is
Tested as an Alternative to Scrubbers", Wall Street Journal,
June 14, 1977.
-------
43
SESSION IB - STRATEGY AMD APPROACH TO SPONSORED R&D
SESSION CHAIRMAN: ANDREJ MACEK, U.S. ERDA
While it is realized that the massive projected increase in utili-
zation of coal (more than 50% between 1976 and 19&5) cannot be accom-
•plished merely by infusion of federal funds, federally sponsored R&D
will clearly be instrumental for this attainment. This session provides
information from top officials of the three principal federal agencies
sponsoring R&D in this area: ERDA, Bureau of Mines, and EPA. These
papers cover what the Government is doing, and what it intends to be
doing in the future, to implement clean combustion of coal.
A significant amount of research is also being sponsored by the
Electric Power Research Institute, but they were not able to send a
representative to this conference.
-------
44 CLEAN COMBUSTION OF COAL
-------
45
STRATEGY IN COAL PREPARATION RESEARCH PLANNING
W. E. Warnke
Coal Preparation Research Manager
U.S. Bureau of Mines
Good evening to all of you. Because we've just eaten dinner there
is some danger of being lulled to sleep. As it happens, I don't have
any visual aids or funny stories to tell as an aid in maintaining your
attention but I will cite some percentage numbers that may pique your
interest. Those of you who are familiar with RI 8118 on sulfur removal
potential may be pleasantly surprised by these numbers tonight.
Let's not belabor the issue of increasing the utilization of coal.
The necessity of replacing gas and petroleum with coal has been
eloquently discussed by earlier speakers. So, let's talk about the
four options as I perceive them in burning more coal. Then, I'll out-
line the Bureau of Mines strategies in planning a five year research
program for coal preparation.
The electrical utility industry and other industries requiring
large amounts of fossil fuel have four options for burning coal in
compliance with sulfur dioxide emission standards. The first is the
use of flue gas scrubbers. Costs of scrubbing sulfur oxides from flue
gases range from $10 per ton of coal burned to $19 per ton depending on
how the capital costs are amortized and who does the cost evaluating.
For lack of a better number, let's assume these costs are $15. Flue
gas scrubbing also is an energy user because limestone must be mined
and transported to the utility plant and perhaps as much as 5 percent
of the energy derived from burning coal is used to drive the scrubbing
system.
Transporting low sulfur subbituminous coal from Wyoming and
Montana to Eastern and Southern utilities is the second option. Unit
train costs are about one cent per ton mile so a haul of 1200 miles
adds $12 to the costs of a low grade fuel averaging less than 20,000,000
Btu's per ton. Unfortunately, much of the subbituminous and lignite
coals contain one percent or more sodium which causes the ash minerals
to slag at temperatures of about 2000 degrees F. Boilers designed to
burn the bituminous coals of the East need retrofitting to avoid slag
buildup on heat transfer surfaces. The retrofitting is expensive and
decreases boiler efficiencies. I've not considered coal slurry pipe-
lining as a means of moving coal from mine to utility for the simple
reason that pipelines may not be used for reasons other than economics.
If costs of installing a coal slurry pipeline escalate like the Alaskan
oil line, the pipelines may not be competitive with unit trains.
-------
46
CLEM COMBUSTION OF COAL
The third option is cleaning high sulfur Northern Appalachian and
Midwestern coals to acceptable sulfur levels. This option is restrict-
ed to cleaning coals from these two regions because the Mississippi
Basin and Northeastern regions of this country represent areas of high
demand for steam coal. According to a study conducted by the Bureau of
Mines on the sulfur removal potential of 227 samples taken from major
bituminous seams in the states of Pennsylvania, Northern West Virginia,
Ohio, and Maryland, 100 samples could be beneficiated or burned without
any beneficiation as compliance coal. The mean Btu recovery was 75
percent. Coals from the Midwestern region are generally higher in both
pyritic and organic sulfur than Eastern or Southern coals. Therefore,
it was not at all surprising to learn that only 11 of the coals sampled
in the Midwestern region could be beneficiated to compliance levels of
sulfur with a mean recovery of 64.5 percent. The results of sink/float
tests on 455 samples of coal from throughout the United States show
that 42 percent can be burned as run-of-mine coal or can be benefici-
ated to meet compliance sulfur levels. Mean Btu recovery of these 192
coals was 83.6 percent. These washability tests simulated current
washing practices in which only pyritic sulfur is removed.
Costs of cleaning coal by some form of physical separation to
remove the pyrite and ash varies from about $2 per ton for the more
simple flowsheets to $6 to $7 per ton for the more complex flowsheets.
The latter includes equipment for fine grinding, flotation cells,
thickeners, filters and dryers.
If we plan to use the 58 percent of the coals that are not amen-
able to conventional cleaning technology, then we must develop new
physical cleaning technology and in some cases, combine it with chemi-
cal desulfurization. Bechtel Engineering Corporation recently complet-
ed an evaluation of six chemical cleaning processes. Although several
of the six removed some organic sulfur, none of the six appeared
promising. Costs for the entire process including crushing, grinding,
chemical processing and compaction of the fine clean coal, ranged from
$18 to $20 per ton. The Bureau of Mines has reservations about these
processes for the simple reason that the costs equal or exceed flue gas
scrubbing costs. We are investigating several other chemical processes
but Richard Killmeyer of the Bureau's Coal Preparation Laboratory will
discuss our plans later on another day.
The fourth option mentioned earlier is a combination of coal
cleaning and scrubbing. If a coal can be partially desulfurized by a
physical process costing $3 per ton or so and the coal burned with
about 1/3 of the flue gases going to a scrubber, the combined costs may
be less than scrubbing alone. There is no hard and fast rule on the
amount of sulfur in the run-of-mine coal, the cost of partially desul-
furizing by physical cleaning and the ratio of unscrubbed flue gas to
scrubbed flue gas. Each coal would require testing to determine the
amount of sulfur removed by cleaning and the amount of the gases
requiring scrubbing to meet the 1.2 pounds of sulfur dioxide standard.
Hoffman-Muntner Associates prepared a study on combined cleaning-
scrubbing costs for the 3ureau. This study which is being published
for public dissemination, outlines 12 combinations of coals, flowsheets
and scrubbing systems. The costs of combined cleaning/scrubbing are
compared to scrubbing alone.
-------
COAL PREPARATION RESEARCH FLAMING 47
We in the Bureau of Mines believe that cleaning the coal prior to
combustion is the most attractive of the four options both economically
and environmentally. When coal is cleaned to remove pyrite, a major
portion of the more objectionable trace elements such as cadmium, lead,
arsenic, and mercury are also removed. These elements are generally
not removed by present day scrubbing processes which means they are
redistributed as fallout from the emitted flue gases. Furthermore,
wastes from washing plants can be dewatered and used as stable landfill
material in an environmentally acceptable manner, whereas sludge from
scrubbing plants is more difficult to impound or store without jeopard-
izing the environment. And, from an economic standpoint, if you'll
excuse reference to the obvious, a $6 or $7 per ton cost for sophisti-
cated cleaning processes is considerably less than the $15 scrubbing
costs.
The Bureau recognizes that current coal washing practices are
inadequate in removing finely disseminated pyrite and remove no organic
sulfur. Consequently, the Bureau's program is heavily oriented towards
fine grinding to obtain optimum liberation of pyrite and with emphasis
on flotation or high gradient magnetic separation to remove most of the
pyrite without sacrificing Btu recovery. The Bureau is also concerned
with chemical desulfurization of coal as a means of removing organic
sulfur. When coal was selling for 10-15 cents per million Btu's, the
type of projects the Bureau authorized was constrained by the funds
available for research and by the costs of the processes under
investigation.
Our coal preparation research program in the past consisted of
small projects designed to achieve engineering advancements to the
state-of-the-art. We didn't have funds for large projects and couldn't
afford much risk. Virtually overnight the game changed from penny ante
to a sky-is-the-limit high roller game. This is not to imply that
we're squandering funds on worthless ideas because the projects we fund
are based on bench scale data and evaluated for technical feasibility.
But the stakes are measured in billions of dollars annually to the
consumers of electrical energy and so because of the stakes we can
afford greater risks in the hope of developing a process for cleaning
Eastern coals.
Perhaps I'm too optimistic but I believe that physical coal
cleaning technology will be developed so a much larger percentage of
the coals in the Eastern half of the United States can be cleaned to
compliance levels. I'm also optimistic about increasing significantly
the Btu recovery in the clean coal product. That, in essence, is the
Bureau's strategy. We are convinced that both physical and chemical
coal cleaning deserve more consideration by policy and decisionmakers
in industry and government as very attractive alternatives to scrubbing
or transporting coal long distances!
-------
48 CLEAN COMBUSTION OF COAL
-------
49
STRATEGY AND APPROACH TO ERDA
RESEARCH AND DEVELOPMENT
ON CLEAN COMBUSTION OF COAL
Dr. S. William Gouse
Deputy Assistant Administrator for Fossil Energy
U.S. Energy Research and Development Administration
As the population and economy of the United States continue to grow,
the need for energy will increase. In 1970 the U.S. used about 67
quads; in 1976 about 74, and projections for the year 2000 range
from less than 100 to over 150 quads, or more than twice the 1976
level.
Although conservation and new end-use technologies can limit the
amount of energy we will need, society will still need large amounts
of energy to support our growing and productive economy. This energy
can come from a variety of forms, including solar, nuclear, and new
and conventional solid, liquid and gaseous fossil fuels.
The main issue for RD&D planning is not how many quads we will need
in any future year, but how we will choose to provide them.
Today about 75 percent of our energy comes from petroleum and natural
gas, and we import some 40 percent of our liquid fuels. There is a
great deal of uncertainty regarding the future availability and cost
of those fuels. Known reserves are being depleted rapidly and new
discoveries have not kept pace with rising demand. Indeed, there is
some question as to how long the earth's total recoverable resources
could support historically rising demand for oil and gas. All studies
indicate that synfuels in various forms will be required in the future.
We need appropriate action now to insure adequate supplies of energy
at reasonable prices to preserve our national security and provide
reasonable lifestyles.
There is no magic solution. Any specific approach to the problem has
its own advantages and disadvantages and these are often perceived
differently by different groups. There are many complicating factors,
each with its own degree of uncertainty and each interacting with all
the others, often in unpredictable and counter-intuitive ways.
As an example of this interaction, consider new home heating systems.
A perceived supply problem in natural gas could push electric heating,
even though that would require considerably more primary energy.
-------
50
CLEAN COMBUSTION OF COAL
The complexity of the problem makes it difficult to determine the
"best" solutions. The first part of the fossil energy strategy is to
define alternative ways to meet future energy service requirements in
terms of social, institutional and economic consequences. The second
is to provide the information to stimulate a productive national
debate, which we hope will lead to a consensus on acceptable alterna-
tives. Finally, we seek to conduct an R,R&D program that will clarify
issues and insure that an effective number of viable alternatives is
available.
There are several dimensions to the energy picture, and that fact
accounts for much of the complication. Two of the most basic dimen-
sions are the energy resources and the activities required to use them.
These are effectively related by the Reference Energy System.
Reference Energy System, Year 1985 — With the NEP
Resource
Nuclear Fuels
U235 LJ238
Hydropower
G isothermal
Solar
Fossil Fuels
Coal
Crude Oil
Natural Gas
Total Resource
Consumption:
Refining &
Conversion
Transport &
Storage
Central Sta.
Conversion
Transmission,
Distribution &
Storage
Decentralized
Conversion
7.62 ( ( ^
Enrich & Fabricate Truck
*" Enrich & Fabricate Truck
3.11
Dam Hydroelectric 765 KVAC
Dpl8 Long Diitance
"0.06 *" *" *'
0.1 2 Thermal — — — _ _^ Q.18
0.03 B'mB
F^ion
-*•'" LMFBR
'-'1.34.
LWR '
HTGR *""
1.34) ^m
(.34)
f
L34)
^
"^%80
iQ^^b.
UG. Dili
°A if*.
• ?i fr*"
w! v
v/45 ¥ ^
i.9as>~ ~ -. -«.
92.67 * 1015 BTU
The major energy sources are listed down the left-hand side and include
nuclear, hydro, geothermal, solar, and the fossil fuels. The activi-
ties listed across the top include extraction, refining and conversion,
transport, distribution and use. Conversion can be both centralized
and decentralized.
The Reference Energy System, by identifying both service demand and
utilizing device, permits the implicit consideration of conservation
and efficiency improvements. This system can help us consider alterna-
tive pathways. Once the potential demands are estimated and various
sets of resources and use technologies identified, we can determine
-------
ERDA RESEARCH AND DEVELOPMENT
51
various resource mix requirements. This forms a standardized basis for
comparison in these two dimensions.
Fossil Pathways
Transport
Transport
Process ||
'^^•LTconveyorU D^Lf *" T_ ;*"» "P-'-H.,
K'Mi"8 J I Truck J I Crush J Lp^meJ ' dn-SSK. B,u,
Convenion Convertjon
I || Distribution
Liquefy
Gasify) Low Btul
Natural.
Gas
G*w«" T . _ , J" „ ^ ]- Gas Pipeline -,
Gas Pipeline -i- Pressurize J V
Oil Well J I rLNG Tanker J
,»i,«»4 r:«l *- Liquefy —*
(Associated Gas!
•-Underground Mi
L s,
Oil Shale 1
tin-Situ -i
In-Situ J
f Rail -i
irground Mine -i I Aboveground
.rfaceMine H" ^ T~ Htl<"1 1
L> Conveyor -* ^> Pipeline
i_ 0:... 1 tn-Situ Retort-l
(Mine- Assisted)
Oil Well
Primary Recovery
4rnmarv Hecovery -i
Oil Well J
Secondary Recovery I
nu Waii -J
Oil
Oil Well
Tertiary Recovery
• Option points Figure 2
Within fossil energy there are four major resources—oil, gas, coal and
shale. Those resources can, via alternative pathways, meet end-use
needs for all solid, liquid and gaseous fuels.
I should point out that exploration, although not shown here, is a
major activity, particularly for oil and gas.
Coal Resources
Legend
Bituminous Coal
Subbituminous Coal
L'9nite
Anthracite
Regions
1. Northern Appalachia
'•i- 2. Central Appalachia
K conn. 3 Southern Appalachia
4. Midwest
5. Central West
6. Gulf
7. Eastern Northern
Great Plains
8. Western Northern
Great Plains
9. Rockies
10. Southwest
11. Northwest
12. Alaska (not shown)
Figure 3
-------
52
CLEAN COMBUSTION OF COAL
Within each major category, there are wide variations as to the nature,
characteristics and location of the resources. This figure shows where
the nation's major coal deposits are found, as well as the location of
coal by rank from peat and lignite and peat through bituminous and
anthracite.
An additional complication concerns the fact that coal is not a single
material; rather, it is a family of related materials, differing widely
in heat content, contaminates (such as sulfur and ash) and physical
properties (such as caking and hardness). These characteristics
greatly affect their use. Processes for coal conversion do not work
equally well on all coals. It is not useful to consider liquefaction
of anthracite.
Shale Resources
Green River Shales
1.8 trillion bbl oil equivalent
in shale over 15 gal/ton
Explanation
Tertiary deposits
Green River Formation
in Colorado, Utah, and
Wyoming; Monterey
Formation, California
middle Tertiary deposit
in Montana. Black areas
are known high-grade de-
posits
Mesozoic deposits
Marine shale in Alaska
Permian deposits
Phosphoria Formation,
Montana
Devonian and Mississippian
deposits (resource estimates
included for hachured areas
only). Boundary dashed
where concealed or where
location is uncertain
Figure 4
This figure on shale resources shows the great variability in the
location and characteristics of the shale resource. Western shales
have a high potential for liquids production but are far from prime
markets. Eastern shales have promise for both gas and liquids produc-
tion and are near demand centers. A similar situation exists for oil
and gas.
-------
ERDA RESEARCH AND DEVELOPMENT
53
Heavy Oils And Tar Sands
Legend
Oil field
3 Area of heavy oil accumulations
100 0 100 200 300
I I I I I
Scale, miles
Figure 5
Low Permeability Sandstone Areas/
Geopressured Zones
Geopressured
Aquifers
(Zones)
Figure 6
1. Greater Green
River Basin
2. Northern Great
Plains Province
3. Piceance Basin
4. Uinta Basin
5. Anadarko Basin
6. Arkoma Basin
7. Big Horn Basin
8. Cotton Valley Trend
9. Denver-Julesburg
Basin
10. Douglas Creek Arch
11. Ft. Worth Basin
12. Ovachita Mountains
Province
13. Raton Basin
14. San Juan Basin
15. Snake River
Downwarp
16. Sonora Basin
17. Wasatch Plateau
18. Western Gulf Basin
19. Williston Basin
20. Wind River Basin
-------
54 CLEAN COMBUSTION OF COAL
Figure 1 can be examined again to explain much of our program content.
For example, the oil and gas program focuses on enhanced recovery tech-
niques; shale focuses on in situ conversion; and coal conversion on
producing liquids and gases.
In the industrial sector, for example, we are concerned with the need
for process steam, feedstocks, electrical and mechanical energy,
indirect heat and direct heat. Each of these represents a demand for
liquids, gaseous and solid fuels.
Transportation, with its very large requirement for liquids, is an
excellent sector to illustrate the available programs and pathways.
The obvious way to satisfy these needs is through petroleum—off shore,
enhanced recovery, etc. But coal and shale may also produce liquid
fuels. However, coal liquids are chemically different than shale
liquids. Coal could make gasoline while shale could produce jet fuels.
Boiler fuels from all four resources can meet electricity utility
demands, but the emphasis is on coal.
End use devices can offer compromises with supply technologies. For
example, we can heat homes with coal through the following paths:
o Coal to SNG, conventional gas furnaces
o Coal to SNG, gas heat pump
o Coal to synthetic liquids, conventional oil furnaces
o Coal to electricity, scrubber, resistence heat or heat pump
o Coal to SNG to electricity, resistence or heat pump
/
o Coal to synthetic liquids to electricity, resistence or heat
pump
Because the energy system is so pervasive and complex, the two-
dimentional matrix is not sufficient. There is another important
dimension to the problem. Another dimension involves constraints to
applying each supply and end-use technology, I have listed eight of the
most obvious ones here, representing the areas in which there must be
compromises in selecting R&D activities for Federal support.
Each of these must be understood in terms of its consequences and the
manner in which it interacts with all others. For example, any
decision on environmental controls will certainly have social and
economic consequences.
The fourth dimension is time. We are facing the transition from a
normal oil and gas-based economy to one based on renewable or non-
depletable resources. We aim to use coal, and other conventional
fossil resources to get us through this transition with minimum
disturbance to society.
-------
ERDA RESEARCH AND DEVELOPMENT
55
Constraints
Constraints
//
/&&&$3f$?m8t$vSW
Constraints
1.
2.
3.
4.
5.
6.
7.
8.
Scientific knowledge
Applied technology
Geographical
Legal & legislative
Political & social
Economics (capital, taxes)
Environment
Manpower & Materials
1
PlilSPi!
Coals
// Oils
/ Gases
Shales
Activities
Figure 7
In our decisions, we must recognize that we are being pushed by the
decreasing availability and increasing cost of oil and gas. We must
also recognize that this situation will become more urgent with time,
especially, if we do not act now.
Time is important because all technologies are not in the same state
of readiness. Because of several fossil energy options for each end-
use requirement plus other ERDA programs, all technologies may not have
an implementation window. Careful study of potential market penetra-
tion must be used as an aid to setting R&D priorities as well as other
things mentioned thus far.
With this background and overview let us concentrate on the Fossil
Energy program, which in addition to clarifying social issues, seeks to
develop the quantities we need at acceptable economic, social and
environmental costs.
We want to increase supplies in the near term for all markets using
liquids and gaseous fuels, as well as for electric power. The appro-
priations and activities for coal, oil, gas and oil shale differ. The
reason concerns the fact that because of the unequal levels of private
R&D investment for the different fossil fuels, the state of development
of conversion and extraction technologies varies considerably. You
will note that the requested FY 78 Budget for Fossil Energy is $656.9
million and coal receives almost 82 percent of that amount.
-------
56
CLEAN COMBUSTION OF COAL
Fossil Energy Budget Estimates-Distribution of Funds
Petroleum and Advanced Research and
Natural Gas Supporting Technology
Oil Shale
and In Situ
Technology
X \ I l./7a 1 /
Coal
Utilization
15.4%
12.0%
Modifications
atERC'S
_
1.5%
Coal Conversion
Coal Utilization
Advanced Research
and Supporting
Technology
Demonstration Plants
Magnetohydrody-
namics (MHO)
Petroleum and Natural
Gas
Oil Shale and In Situ
Technology
Modifications at ERC'S
Total
Percentage Distribution of Fossil Energy Budget
Estimates in FY 1977 and FY 1978 Shown as Follows:
FY 1977%
FY 1978%
Budget Authority
(Dollars in Millions)
Increase
FY77 FY78 Decrease
$150.3 $233.3 $+83.0
74.4 79.1 +4.7
37.1
100.3
40.3 +3.2
125.9 +25.6
40.0 50.5 +10.5
43.2 76.7 +33.5
31.0 41.5 +10.5
6.9 9.6 +2.7
$483.2 $656.9 $+173.7
Figure 8
Thus our commitment to Fossil Energy, especially coal is great, because
the resources are abundant. In total, they can supply all of the
Nation's additional needs for over a century—more than enough time to
develop technologies to exploit "inexhaustible" resources economically.
Our objective is to develop technologies that will have widespread
applicaton by private industry. How do we go about this? Is ERDA/FE
the main customer for the technology it helps create? The answer is
no. At ERDA we feel that as the main producer or consumer of energy,
the private sector is primary and the role of the Federal Government
should be the supplementary one of sharing risks with private industry.
Thus, the national needs we are charged to satify are felt and satis-
fied almost entirely in the private sector—by industry, commerce and
individuals acting through the marketplace.
So in order to achieve our missions and goals, ERDA/FE seeks to (1)
establish the appropriate policy and technical climate (with appro-
priate incentives, if needed) for private sector action; (2) share
risks with private industry; and (3) support a complementary RD&D pro-
gram to obtain necessary and timely information and to help stimulate
the private sector. Thus, ERBA does not expect to be in the energy
business. We have nothing to "sell." Our approach is to utilize
federal funds so that the private sector will participate right from
the start of a project with its own know how and financial resources.
Indeed, we aim for industry to be involved in our programs every step
of the way—as the major contributors of technical ideas and approaches
and as the major beneficiaries of operating experience. Therefore,
industry will be completely familiar with the problems as well as with
-------
ERDA RESEARCH AND DEVELOPMENT
57
the results and data. Such involvement will put our industrial
partners as well as the industries themselves in an ideal position to
decide whether to implement a new technology. In turn, we will be in
an ideal position to know the obstacles—financial and otherwise—to
overcome in order for the technology to be implemented.
This figure shows that industry receives the overwhelming percentage
of the Fossil Energy Budget.
Distribution of Fossil Energy Funds
Universitii
3.8%
435%
National Labs
7.3%
5.2%
Percentage Distribution of Fossil Energy Budget
Estimates in FY 1977 and FY 1978 Shown as Follows
FY 1977%
Figure 9
FY 1978%
Fossil Energy Budget Estimates Breakdown
of Funds by R£rD Agency Budget Authority
(Dollars in Millions)
Energy Research Centers
National Laboratories
Universities
Industry
General Plant and Equipment,
Construction, OSHA and
Environment at Energy
Research Centers
Total
FY 1977
$47.0
35.2
18.2
375.9
(% of
Total)
(9.7)
(7.3)
(3.8)
(77.8)
FY 1978
$60.9
34.0
26.1
526.3
(% of
Total)
(9.3)
(5.2)
(4.0)
(80.1)
(1.4)
(1.4)
-------
58
CLEAN COMBUSTION OF COAL
Industry is involved in the typical development sequence shown below.
Typical Development Sequence
II
1
1
Private Industry
II 1
i
L 11 Jt
Exploratory
Research
1-4 Years
Process
Development
Unit (PDU)
4-6 Years
Pilot
Plant
5-8 Years
/ Demonstra- ' '
tion
Plant
/ I Commercial /
Plant /
8-12 Years I I
rr IT IT IT "\i
Government
15 to 20 Years
Figure 11
Each phase in the sequence involves scaling up to larger units, until
in demonstration plants, the scale is large enough to provide firm
data for cost estimates and design of commercial scale plants. Cost-
sharing, while concentrated in the pilot plant and later phases, can
also take place earlier. The time involved with each phase of develop-
ment varies, depending upon the complexity of the process, project,
scope, and resources applied.
Because in our accelerated program the phases overlap, the total
development time is less than the sum of all phases. From laboratory
to completion of demonstration plant operation is typically 15 to 20
years, and we evaluate technical feasibility of the concepts in each
phase. Tentative economic evaluations start early in process develop-
ment and continue through pilot and demonstration phases. We make more
extensive economic evaluations with demonstration plants, as well as
tentative environmental acceptability evaluations, water resources
availability assessments, and environmental impact statements, as
required.
There has been much talk about the possible need to start up whole new
industries to implement new technologies. For some ERDA programs, such
as solar energy, where the existing industrial base is small or non-
existent, this idea undoubtedly has some validity. Such is not the
case for Fossil Energy where for years, a large base of existing com-
panies has supplied the Nation's oil, natural gas, petrochemical and
electricity needs. And, I might add, they have done that job well.
For FE we need, not new industries, but technologies that either make
sense in the marketplace, or which can be made to make sense by govern-
ment initiatives, if there is sufficient public benefit from developing
such technologies.
-------
ERDA RESEARCH AND DEVELOPMENT 59
An example would be the synthetic liquids we will need in the future,
derived from coal or from oil shale. There are barriers to their
commercialization. For that reason, ERDA/FE has a commercial demon-
stration program, which we hope will resolve the uncertainties related
to economic/environmental feasibility, socioeconomic impacts, resource
requiremments, capital cost, financing and regulations. We need to
resolve them in order to bring about the kind of broad-scale plant
investment in the mid 1980's which we will need in order to achieve
significant production in the 1990's.
At the same time we are developing Federal initiatives to build a
synfuels industry, we are also working in the FE program to overcome
the technical deficiencies in the current state-of-the-art. We do this
through heavy investments in RD&D per se. We now support research in
universities and industrial labs, and process development in corpora-
tions and laboratories. Pilot plant scale operations are underway in
companies, in government and in independent centers. And ERDA/FE has
an active and ambitious program to demonstrate at near commercial
scale, promising second generation coal and shale processes developed
by both industry and government.
Such activity pushes forward the frontiers of knowledge and advances
the state-of-the-art. Our premises are: (1) Because of the regulatory
and economic uncertainties I mentioned before, the private sector won't
push as fast or as extensively as the national interest warrants.
Thus, government provides incentives and the means for faster and more
extensive progress; (2) As the state-of-the-art advances, a company can
finish developing a commercially attractive process more easily. For
patent reasons, a company may want to put its own stamp on a process,
and thus, not adopt any we have sponsored in its entirety. But a com-
pany will have to contend with far less uncertainty than it does now
(or did a year ago) and should have in hand the ingredients for suc-
cess; (3) Many "systems" problems faced by new technologies can be
understood and resolved only by actual trial (and some, by not too
much error). Environmental impacts, effects of regulations, process
bottlenecks, etc., are still most clearly identified and best dealt
with in actual practice. Hence, our demonstration program aims to put
the pieces together at a near-commercial scale to show potential users
of the technology what applying it will involve; (4) The existence and
scale of the program highlights an important technological area and
focuses attention on it. Companies not funded by ERDA thus may also be
induced to invest in a particular technology. Companies having undis-
closed processes may have to advance them in contention. Researchers
may be stimulated to work on related topics, once the area, as a whole,
becomes "fashionable." For such spill-over benefits, the government
need not pay directly; (5) The people involved in the program, in
ERDA, in our Energy Research Centers, in industry, and in the universi-
ties, acquire knowledge and skills indispensible to the eventual
largescale commercialization of new technology.
In our program, we also address, in several ways, the obstacle of
economic risk of the R&D.
-------
60
CLEAN COMBUSTION OF COAL
Targeted Cost-Sharing Participation
PDU's
Pilot Plants
Demo Plants
Alternative Fuels
* Some Plants May Be Cost-Shared
Figure 12
We pay 100 percent of the costs of basic and applied research and all
the costs for the early stages of process development. We pay only
about 2/3 of the costs for pilot plants, and about 1/2 of the costs for
demonstrations. We do this partly because the costs rise as one moves
along the development sequence. But more importantly, we do it to
begin commercialization. Cost-sharing ensures that projects have merit
by private standards and that industry will make a later commitment
of other resources to protect its investment.
Because our industrial partners must put up a significant share of the
funds, we expect them to make a significant share of the decisions
about what and how to test. Thus, they bring their technical and
marketing judgements to bear early. What will they learn from the
project? Will the product sell? Will the process work? What will
the product cost? We issue flexible Program Opportunity Notices or
specific requests for proposals in areas that appear to warrant that
scale of development. Companies then respond with detailed proposals
specifying what they prefer to do. Because their stockholders' money
is on the line, they must apply their best commercial insights and
expertise.
As a result, the process development sequence ceases to be a straight
path. Each stage, merges, modifies or drops components and adds new
ones. We would be surprised if many proposed demonstration projects
were direct lineal? descendents of our process development units or
pilot plants.
More generally, ERDA could find other ways to reduce economic risk
besides cost-sharing, including loan guarantees, price supports and
special tax credits for energy related investment.
-------
ERDA RESEARCH AND DEVELOPMENT 61
No one fully knows the long-run impacts of these different approaches.
We are studying some of them now and we need to look at others. We
are studying one of them now and we need to look at others. We are
studying, for example, projected market demands for the several forms
of fossil energy and costs of advanced technologies. We are also
looking at contingency plans and their impact on overall strategy
and choice of federal role. We consider too, in selecting the Fossil
Energy project "portfolio," how to balance the need for short-term,
relatively sure "blue chips" against the need for near-to-mid term
improved processes, moderate risks and potential breakthrough "high
flyers."
In conclusion, I am optimistic that our Program can supply the fossil
energy bridge to the next century if we can develop a real government
private industry partnership. Such a partnership will clear away the
roadblocks--environmental, price, regulatory and legislative--to
bringing promising technologies into the marketplace. Conferences
such as these help to find the path to innovation because we need new
technology to solve our energy problems. We need entrepreneurship
because whatever is developed needs to be brought to market, not
bought by the government.
-------
62 CLEAN COMBUSTION OF COAL
-------
63
EPA R&D PROGRAM RELATING TO
CONVENTIONAL COAL COMBUSTION
by
FrankT. Princiotta
Director, Energy Processes Division
Office of Energy, Minerals and Industry
U.S. Environmental Protection Agency
The Clean Air Act Amendments of 1970 are the major driving force for
control of air pollution from both new and existing combustion sources.
The Act has a statutory requirement to achieve acceptable ambient air
quality for the so-called criteria pollutants (Figure l). Among these
pollutants are sulfur dioxide and total suspended particulates which are
essentially stationary source pollutants and nitrogen oxides which are
generated by both stationary and mobile sources in roughly equal quanti-
ties. In addition, the Clean Air Act calls for the promulgation of New
Source Performance Standards (NSPS) for a variety of polluting industries
including coal-fired steam generators. Presently standards are on the
books for control of sulfur dioxide, total suspended particulates and
nitrogen oxides from coal-fired, oil-fired and gas-fired steam genera-
tors. The present NSPS for coal units, as well as typical uncontrolled
emissions are included in Figure 1. As you can see, the present NSPS for
coal-fired steam generators calls for approximately 70-80% control of
sulfur oxide; for a high degree of particulate control (approximately
98$); and a moderate 30!? control for nitrogen dioxide.
I must point out that there are certain changes which may occur both
in terms of the Clean Air Act as well as some revised standards under
the present Act which could have important impact on control technology
requirements for fossil-fuel combustion units. First, the House and
Senate have recently passed legislation which has gone to joint committee,
calling for a best available control technology approach (BACT) for new
coal-fired power plants insofar as sulfur dioxide and particulate pollu-
tion is concerned. Although the implementation details of this approach
have not been worked out, and the final version of the Act has not yet
passed both Houses, it appears that this change would require best
available control technology for all new sources on both low- and high-
sulfur coal applications eliminating the low-sulfur control option to
meet sulfur oxide standards.
The second change might involve more stringent NSPS for coal-fired
power plants than the existing standards (Figure l). Revised standards
are under consideration'which could lead to more stringent control for
sulfur dioxide, nitrogen oxides and total suspended particulates.
-------
6A CLEAN COMBUSTION OF COAL
STATUTORY REQUIREMENT TO ACHIEVE ACCEPTABLE AMBIENT AIR QUALITY FOR:
S02 _
~ STATIONARY SOURCES
TSP
NOX
HYDROCARBONS
\ MOBILE SOURCES
CARBON MONOXIDE
PHOTOCHEMICAL OXIDANTS
STATUTORY REQUIREMENT TO MEET NSPS FOR COAL-FIRED STEAM GENERATORS
STANDARD UNCONTROLLED
SO
'2'
1.2 LB/106 BTU 5 LB/106 BTU
TSP: 0.1 LB/106 BTU 6-10 LB/106 BTU
N02: 0.7 LB/106 BTU 1 LB/106 BTU
Figure 1. Clean Air Act—Driving Force for Flue Gas Cleaning.
-------
EPA R&D PROGRAM 65
Third, the Agency is considering the possibility of NSPS for indus-
trial boilers which would be defined as being less than about 25 Mwe
(equivalent) for nitrogen oxide, total suspended particulate and possibly
sulfur oxides.
The President's recent energy message highlights the need for
effective control technology on coal combustors. The message calls for
the expansion of the annual coal production rate from the TOO million
tons presently produced to over 1 billion tons by 1985. His policy also
calls for massive conversion of existing utility and industrial power
facilities from oil and gas to coal, and for essentially no new oil or
gas industrial or utility boilers. Future options for these applications
would generally be coal or nuclear, or one of the emerging energy tech-
nologies. Although the conservation aspects of the President's plan,
and the assumption of strict environmental control trends to minimize
environmental degradation, total emissions of nitrogen oxides and sulfur
oxides, are projected to rise above present levels by 1985. The Presi-
dent's energy message also calls for accelerated research, development
and demonstration for the so-called clean coal technologies: coal
cleaning, flue gas desulfurization, particulate control, fluidized bed
combustion, gasification, liquefaction and coal mining. In fact, we, in
the EPA energy-environmental program, have been working with the Office
of Management and Budget at planning an accelerated program for certain
coal combustion control technologies. We are hoping to get a relatively
large funding spike in fiscal year 1978 to accelerate our present
development/demonstration efforts in the following technology areas:
flue gas desulfurization, nitrogen oxide control, particulate control,
coal cleaning, and coal processing.
So what we basically have is the possibility of more stringent
emission regulations, in light of possible Clean Air Act revisions and
upgrading of present NSPS, superimposed on an energy plan which calls
for increased burning of coal. Clearly, effective and low-cost control
technology for utility and industrial sources are needed in the near
term.
Figure 2 summarizes the critical problems associated with utility
and industrial conventional combustion. This table attempts to briefly
describe: the primary problems associated with such combustion; whether
or not there are existing standards; major near-term control technologies
available; the present status of these technologies; the secondary resid-
uals produced by these technologies; and finally the needed control
technology R&D in order to resolve some of the remaining problems. As
you can see, the primary problems associated with industrial conventional
combustion include the primary air pollutants, nitrogen oxide, sulfur
oxide, particulates and potentially hazardous materials, as well as some
of the waste and water effluents associated with these control technology
and the power plant itself.
Now that I have completed my background discussion, I'd like to
present some of the highlights of our ongoing research, development and
demonstration program.
-------
Figure 2
DESCRIPTION OF
PROBLEM
PRIMARY
POLLUTANTS
so0
2
NO-_
X
PARTICIPATES
POTENTIALLY
HAZARDOUS
MATERIALS
STANDARD TYPE OF FGC
PRESENTLY CONTROL
ESTABLISHED TECHNOLOGY
YES
NSPS &
AAQS
YES
NSPS &
AAQS
YES
NSPS &
AAQS
NO
COAL CLEANING
FGD
COMBUSTION
MODIFICATION
FLUE GAS TREAT-
MENT
ELECTROSTATIC
PRECIPITATORS
BAG HOUSES
WET SCRUBBERS
NOVEL DEVICES
UNDEFINED
PRESENT STATUS
1ST GENERATION
DEMO PLANNED
1ST GENERATION IN
FULL SCALE DEMO
2ND GENERATION IN
BENCH AND/OR
PILOT SCALE
COMMERCIAL FOR SOME
NEW UNITS
PILOT SCALE AND
DEMO IN JAPAN
COMMERCIAL
1ST GENERATION DEMO
1ST GEN. COMMERCIAL
2ND GEN. FULL SCALE
DEMO
BENCH OR PILOT SCALE
UNDEFINED
SECONDARY
RESIDUALS
HIGH-S REFUSE
SLUDGE,
PURGE STREAMS
POSSIBLY MORE
PARTIC. AND CO
VARIES WITH
PROCESS
FLY ASH
UNDEFINED
NEEDED CONTROL
TECHNOLOGY R&D
(INCLUDING ASSESSMENTS)
A) ELIMINATION OF SECONDARY
POLLUTANTS
B) DEMONSTRATE PRACTICABILITY
C) BROADEN APPLICABILITY
A) ELIMINATION OF SECONDARY
POLLUTANTS
B) IMPROVE RELIABILITY
C) BROADEN APPLICABILITY
D) IMPROVE ENERGY EFFICIENCY
A) BROADEN SOURCE
APPLICABILITY
B) IMPROVE ENERGY EFFICIENCY
C) IMPROVE NOX CONTROL
EFFICIENCY
D) MINIMIZE IMPACT OF
RESIDUAL POLLUTION
A) IMPROVE COST EFFECTIVE
FINE PARTICULATE CONTROL
B) BROADEN APPLICABILITY
C) DEVELOP NOVEL DEVICES
WITH IMPROVED CAPABILITY
PROBLEM REQUIRES
DEFINITION
o
o
o
c!
CO
i-3
H
O
o
g
IT"
-------
EPA R&D PROGRAM 67
Flue Gas Desulfurization
I would first like to discuss our flue gas desulfurization (FGD)
program. One can characterize these systems as producing either throw-
away (disposable) or saleable products. In the "throwaway" FGD area,
one of our most important single projects is the Shawnee lime/limestone
prototype program which we've conducted with the aim of improving lime
and limestone scrubbing processes. As many of you know, such processes
have been selected for many utility applications; approximately ^0,000 Mwe
or approximately 3 billion dollars worth of these systems are presently
in operation or on order at this time.
The Shawnee program has been a cooperative effort of the EPA, the
TVA and the Bechtel Corporation. Two 10 Mwe scrubbers have been operated
since 1972, and a 0.1 Mwe pilot scrubber has been operated in support of
the two larger facilities since about 1973. This program has demonstrated
long-term reliable operation of both lime and limestone processes. The
particularly troublesome mist eliminator plugging problem has been solved
by using a combination of careful operating conditions as well as care-
fully selected mist eliminator washing configurations. During the course
of our program, we discovered the potential of unsaturated operation in
order to avoid gypsum scaling which had plagued earlier commercial sys-
tems. Basically, this approach involves selecting operating parameters
so that scrubbing liquors never get saturated or super-saturated in cal-
cium sulfate thereby avoiding potential scaling problems. Also, during
the course of the program, we've learned how to achieve high-alkali
utilization. For example, limestone utilizations of over 90% have been
achieved. This leads to lower alkali requirements and lower sludge pro-
duction rates, both of which yield lower operating costs. We have also
achieved high sulfur oxide removal efficiencies. Typically, efficiencies
for both lime and limestone systems in excess of 95% can be achieved
without an excessive economic penalty. One of our most recent findings
is that minor process modification can allow for sludge oxidation by
gypsum producing a material capable of 90% dewatering by filtration.
Since last year, we've developed a design/cost computer model which acts
as a data base for all the information the EPA, TVA, Bechtel team knows
about lime/limestone scrubbing and allows a given utility or other FGD
user to input application parameters and output a conceptual design
along with some pretty good cost estimates.
We have been working with the Air Force in operating and testing a
lime scrubber system on a 21 Mwe coal-fired industrial boiler at the
Rickenbacker Air Force Base. This unit started up in March 1976, and,
although there were some boiler control and fan problems, the efficiency
and reliability of the scrubber have been good. Also, it appears that
the economics of such scrubbing systems on industrial boilers of this
size do not appear prohibitive.
As an alternative to lime and limestone nonregenerable FGD systems,
our program has been actively developing the double alkali scrubbing
process. Although the double alkali system has basically the same
chemicals entering the process and leaving the process, they do have
several potential advantages over lime/limestone scrubbing processes.
These include: less energy consumption; higher sulfur oxide removal
efficiency; lower maintenance; and, lower capital and operating costs.
-------
00 CLEAN COMBUSTION OF COAL
The EPA program has actively developed double alkali technology at the
"bench and pilot levels. We've worked with Southern Services at 20 Mwe
electric prototype system and, most recently, we've worked with the
General Motors Company at applying a double alkali process variation to
one of GM's industrial boilers. In the past, this double alkali scrubber
had generally good operability on the 32 Mwe industrial boiler, although
some problems were encountered. It should be noted that,as for the
previously mentioned Rickenbacker lime unit, the approximate capital
costs, on this size industrial boiler, of $100/Kwe do not appear exces-
sive.
Recently, the EPA announced its plans to demonstrate the double
alkali process on a high-sulfur coal utility boiler. The Louisville Gas
and Electric Company and its 270 Mwe electric Cane Run No. 6 unit was
selected for this demonstration. Combustion Equipment Associates/Arthur
D. Little comprise the process supplier team. I think it is noteworthy
that the cost of this unit, and these costs are fairly firm, are esti-
mated to be only $55 per Kwe capital costs; and 2.5-2.9 mills/Kwe hour
for total annual revenue requirements. Also, only 1.1$ of the power
plant's energy output is required to run the fans and pumps for this
process. This is roughly one-third to one-half the energy requirement
of similar lime or limestone systems.
As an alternative to lime and limestone systems and their inherent
sludge production, we have been active at developing and demonstrating
regenerable (or saleable) product FGD systems. For example, we have
demonstrated the promising magnesium oxide scrubbing -process at Boston
Edison's Mystic Station. This scrubbing facility was tested on a 155 Mwe
oil-fired boiler and produced saleable sulfuric acid. The test program was
initiated in April 1972 and completed in June 197^- Although many early
problems were identified, particularly those associated with the various
solid handling operations, operability improved substantially toward the
end of the test program. Unfortunately, no meaningful demonstration of
this process has been performed on the critical coal-fired combustion
units.
Our major ongoing project in the regenerable area is the Wellman
Lord demonstration test program. Wellman Lord technology involves scrub-
bing with a soluble reactant followed by thermal regeneration producing
concentrated sulfur dioxide which can yield either sulfuric acid or
elemental sulfur. This process has been successfully demonstrated in
Japan on a variety of oil-fired facilities. The EPA demonstration pro-
gram is on a 115 Mwe electric coal-fired facility at the Northern
Indiana Public Service Company and produces elemental sulfur. This
facility started up in April 1976 but due to a boiler explosion we are
now in a restart-up mode. Hopefully, definitive findings on this process
will be available at our next conference.
Recently, the Atomic International's Aqueous Carbonate Process has
been selected for demonstration. The process will be demonstrated at
Niagara Mohawk's Huntley Station and we are hoping that the system will
start up in early 1979- It offers cost and other advantages over alter-
native regenerable processes, but must be considered a relatively high-
risk venture due to the relatively small scale of previous test experi-
ence. Also note that the EPA is working with the U.S. Bureau of Mines in
-------
EPA R&D PROGRAM 69
applying the regenerable Citrate Process to a large industrial facility.
Figure 3 summarizes the major EPA-sponsored demonstration programs for
both throwaway and saleable project FGD processes.
In addition to our demonstration activities, TVA working for the EPA
has conducted a series of very relevant by-product marketability studies
which help put the sulfur, sulfuric acid, and other sulfur by-product sale
situation in perspective. Also, studies have been performed at evaluating
alternatives to scarce natural gas, as the reductant materials to produce
elemental sulfur from the concentrated sulfur dioxide associated with many
regenerable FGD processes.
Let me conclude my discussion of our FGD program by briefly mention-
ing three other important programs. First, the TVA has prepared a number
of FGD process economic studies which in my mind are the most reliable
cost estimates for both throwaway and saleable FGD processes. Secondly,
the EPA working with the Council on Environmental Quality is working on
a scrubber availability study which attempts to compare scrubber avail-
ability to that of other power plant components. Preliminarily, it
appears that there is insufficient information to do a statistically
convincing comparative study. However, it has become apparent that
several major power plant components appear to have reliabilities lower
than some of the more recent FGD systems.
Finally, I'd like to mention the very active technology transfer
program in the FGD area. Some of the outputs of this program include
the PEDCo Status Report -(published every other month) which summarizes
what's going on in the scrubber field both in terms of operating and
planned systems. Also, we now publish quarterly RD&D status reports
which summarize the results of our ongoing research program. We are
preparing lime and limestone data books and cost and reliability hand-
books which present potential users with important information on commer-
cial and noncommercial technologies.
Nitrogen Oxide Control
Now, let us discuss our NOX control program. I think a little back-
ground might be in order so I've added some background information that
I think is relevant. As many of you know, there was an unreliability
problem discovered in the ambient air quality measurement technique for
NOv, back in 1972. Prior to discovery of the problem, it was believed
that kj Air Quality Control Regions (AQCR) out of the total of 2kl AQCRs
for the country had an N02 ambient air problem. What we've found out is
that due to an inherent measurement error, ambient levels of W02 were
measured too high. Using more accurate techniques, only four areas
(AQCRs) for the country really seemed to have an NC>2 problem.
However, since 1972 there hasn't been too much progress in
control from either stationary or mobile sources. As a result, they are
now finding new AQCRs exceeding the TTO2 standard and, quite frankly, the
trend seems to be for further NC>2 ambient quality problems. Therefore,
our present N02 control strategy does not appear very effective. Our
present control strategy, basically, includes control of both mobile
(automobile) sources as well as stationary sources.
-------
Figure 3
EPA-SPONSORED STACK GAS DESULFURIZATION DEMONSTRATION SYSTEMS
EPA-SPONSORED PROCESS
(BY PRODUCT)
NON-REGENERABLE
LIMESTONE SLURRY SCRUBBING
(SLUDGE)
LIME SLURRY SCRUBBING
(SLUDGE)
DOUBLE ALKALI SCRUBBING
PARTICIPATING
UTILITY
TVA
TVA
LOUISVILLE
G&E
PROCESS
DEVELOPER
BECHTEL AND
OTHERS
CHEMICO,
BECHTEL,
AND OTHERS
CEA/ADL
LOCATION
SHAWNEE
UNIT 10
PADUCAH, KY
SHAWNEE
UNIT 10
PADUCAH, KY
CANE RUN 6
UNIT SIZE EXPECTED
AND TYPE START UP
10 MW UNDER WAY
COAL
10 MW UNDER WAY
COAL
-
270 MW EARLY - 1979
CJ
i
o
Q
a
CO
t-3
H
O
^
REGENERABLE
MAGNESIA SLURRY
SCRUBBING - REGENERATION
(98% SULFURIC ACID)
SODIUM SCRUBBING
REGENERATION (SULFUR)
AQUEDOUS CARBONATE
BOSTON EDISON
CHEMICO-BASIC
NORTHERN INDIANA DAVY
PUBLIC SERVICE CO POWERGAS
ALLIED CHEMICAL
NIAGRA MOHAWK
ATOMICS
INTERNATIONAL
MYSTIC
STATION 6
BOSTON, MASS
D.H. MITCHELL
STATION 11
GARY, IND
HUNTLEY
STATION
150 MW
OIL
115 MW
COAL
100 M
COAL
COMPLETED
LATE - 1975
EARLY - 1979
o
o
-------
EPA R&D PROGRAM 71
I should point out that almost one-half of the NOX emissions are
from mobile sources; a little more than one-half are associated with
stationary sources
air quality goals.
stationary sources. So both are of interest if ve are to achieve NOX
We have set so-called interim standards for automobiles. They are
not very stringent standards, and are in fact substantially less strin-
gent than the Clean Air Act goal. Due to economic, technological and
political considerations, it appears that in the near term relatively
moderate NOX emission standards will prevail for autos.
Also, the new source performance standards philosophy is based on
best available control technology. Since control technology for strin-
gent control is not available, the NSPS for utility boilers is not very
restrictive and only requires, as I indicated earlier, on the order of
30 or 35 percent control. There's also an NSPS for nitric acid plants
which is a significant contributor to the NOX problem.
So, our present strategy doesn't seem to be sufficient to get these
AQCRs back into line, and for that matter to turn around the trend toward
additional AQCRs getting out of standard. Therefore, additional control
of stationary sources beyond the present level may be necessary. Of
course, the NOX problem is more complex than just NOg; NO is a major
precursor for photochemical oxidant and nitrate production.
There are three major categories of control for NOX control for
stationary sources. The only near-term technology is combustion modifi-
cation, which the present EPA program is emphasizing. This technology is
the basis for the present new source performance standard on large fossil
fuel generators. Flue gas cleaning is the second technology. It's
similar to flue gas desulfurization, and probably should be called flue
gas de-nitrification. We have several embryonic programs in that area
which I will briefly describe.
This combustion modification program aims at developing technology
capable of controlling emissions from the two major sources of NOX from
stationary sources; namely, "thermal" NOX and "fuel" NOX. One hopes to
control "thermal" WOX by lowering the combustion temperature since the
equilibria relationships are such that lower temperatures retard NOX
formulation, i.e., higher temperatures favor the oxidation of nitrogen.
And, one can control "fuel" NOX problems by lowering oxygen concentration.
One might note that the control approaches are generally common to both
thermal and fuel NOX, with the exception of flue gas recirculation, which
is really oriented primarily toward minimizing thermal NOX-
One has to be careful implementing these techniques since it is
possible to aggravate other pollutant emission problems. We could, for
example, increase emissions of carbon monoxide or particulate matter, if
we drastically lower excess air. We could also lower thermal efficiency
with approaches such as flue gas recirculation. These techniques can
also lead to operating problems since boilers will not be operated in
some cases under the design conditions that they are designed for. How-
ever, the results of our program to date indicate that each of these
potential problems are controllable if one is careful about the particular
control technology and the design parameters utilized.
-------
72 CLEAN COMBUSTION OF COAL
Let me now discuss the ongoing nitrogen oxide combustion modifica-
tion program. I'll discuss this program by source area: utility and
large industrial boilers; small industrial, commercial and residual
systems; stationary engines; and industrial processes and after-turner
equipment.
In the utility and large industrial boiler area, the program's major
effort has focused on staged combustion approaches. By utilization of
this combustion modification technology which involves adding combustion
air sequentially in more than one location to minimize total oxygen
requirements and combustion temperature, control levels of 0.^5 lb/10 Btu
appear achievable. ¥e are presently planning corrosion tests to ascertain
whether or not boiler tube corrosion is an inherent problem. It has been
postulated by some that the reducing environment associated with staged
combustion can remove the protective oxidative coatings from the tubes,
thereby accelerating corrosion. Therefore, corrosion tests are considered
quite important.
Perhaps our most encouraging activity in the NOX program is our work
in the low-NOx burner area. By redesigning pulverized coal burners to
more carefully combust coal, experimental results have indicated that NOX
levels of 0.2 to 0.3 lb/10° Btu are achievable. This represents overall
control of an uncontrolled coal-fired boiler of over 10%. Since such
burner technology appears inherently inexpensive, this could be the
answer to low-cost nitrogen oxide control from both industrial and utility
pulverized coal boilers. If we receive the incremental fiscal year 1978
funding that I mentioned earlier, we will be able to proceed on an orderly
demonstration program aimed at evaluating this approach at sizes culminat-
ing in an integrated full-scale demonstration program.
In the small industrial, commercial and residual system area our
program's emphasis has focused on residual oil package boilers, residen-
tial furnaces and coal-stoker boilers. We have demonstrated that staged
combustion is an effective control approach on oil-package boilers. Pre-
liminary information indicates that optimum burners can achieve 10% NOX
control over conventional burners for residential distillate oil furnaces.
Our work in the coal stoker areas has been limited so far to the develop-
ment of emission factors for existing designs. A combustion modification
technology program is presently under development for this increasingly
important class of coal boilers.
Staionary internal combustion engines are a major source of nitrogen
oxide emissions, primarily because of their widespread use as the prime
mover for pipelines and gaslines around the country. Preliminary work
indicates that it may be possible to lower nitrogen oxide concentrations
in combustion gases to 50-100 ppm. This would represent an overall NOX
control of 75$> and a substantial improvement both in NOX efficiency and
overall energy conversion efficiency to the presently available control
technology, which involves introduction of water into the combustion zone.
In the industrial processes and after-burner equipment area, we have
a relatively low level of activity due primarily to limitations in funds.
Our near-term goal in this area is to characterize NOX and other emissions
for these facilities.
-------
EPA R&D PROGRAM 73
Let me "briefly describe the NOX flue gas treatment research and
development program. One might ask, in light of the comprehensive com-
bustion modification program I have just described, "Who needs flue gas
treatment?" and "What is the advantage of this technology over combustion
modification?" Well, the only identifiable advantage in my mind is that
this approach has the potential for very high NOX removal efficiencies
which could conceivably be required if standards were to tighten in the
years ahead.
However, you pay the price for high removal efficiency in terms of
high capital and operating costs and system complexity relative to com-
bustion modification technology. These approaches are under active
development in Japan due to the stringent ambient and emission standards
for nitrogen oxides there. Two basic approaches are being developed, dry
and wet processes. The dry processes are the simpler of the two and
generally involve a chemical reduction reaction. Typically, ammonia is
used as the reductant and a selective catalyst is needed to reduce
nitrogen oxide to elemental nitrogen and oxygen. Problems with this
approach include uncertainty regarding the viability of this process for
coal-flue gas, the possibility of secondary emissions (such as ammonium
sulfate), and potentially high costs associated with capital expenses
and reagent needs. Wet processes are generally more complex and can
involve oxidation to nitrogen dioxide followed by a scrubbing step;
others reduce the nitrogen oxides in solution. For the oxidative/reduc-
tive approach, ozone is a typical oxidant. The main advantage of this
class of processes is the potential for combined sulfur oxide and nitrogen
oxide removal. Problems include large ozone and energy needs for the
oxidative/reductive process and the production of secondary wastes and
high costs for all versions of the wet processes. The EPA program in
this area involves piloting promising processes treating flue gas from
coal-combustors, borrowing heavily from Japanese technology on oil-fired
boilers there. Two pilot programs are in the final contractor selection
phase, one will involve nitrogen oxide removal only and the other will
investigate concurrent removal of nitrogen oxide and sulfur oxide.
Fine Farticulate Control
First a little information in the way of background. Fine particu-
lates are health hazards because they are airborne for extended time
periods, can penetrate deeply into the lung, and can act as transport
agents for other pollutants. Our research, development and demonstration
program includes the following major areas:
• Improvement of characterization of present technology, e.g.,
electrostatic precipitators, scrubbers, fabric filters
• New ideas/novel devices
• High temperature/high pressure control technology
• Collectability of dust
• Control from low sulfur coal combustors
Our program is actively evaluating the potential for electrostatic
precipitators as high efficiency fine particulate control devices. We
have characterized precipitators for seven particulate sources, and
-------
74 CLEAN COMBUSTION OF COAL
developed a math model which characterizes precipitator performance as a
function of particulate characteristics. We have concluded that precip-
itators offer the possibility of high fine particulate control when there
is no ash resistivity problem. Precipitators are very sensitive to the
chemistry of the ash they must collect. For example, for low sulfur coal
combustion facilities, the ash generally is of a low resistivity since
there is not sufficient sulfur trioxide to raise the ash conductivity
thereby leading to collectability problems. Our program is actively
considering ways of upgrading precipitator performance when the ash has
a less-than-optimum resistivity. We are actively evaluating fly ash con-
ditioning agents such as sulfur trioxide, ammonia, etc. We are also
actively evaluating the possibility of precharging the flue gas upstream
of the precipitator in order to improve the collectability of the ash.
We are presently planning a pilot demonstration of an attractive pre-
charging approach. We are carefully coordinating our program with the
Electric Power Research Institute who is also active in this area.
We are also evaluating and developing various scrubber devices for
the removal of fine particulates from combustion facilities. We have
evaluated ten devices on a variety of particulate sources, and currently
find a consistent pattern where fine particulate removal efficiency is
highly dependent on pressure drop (and, therefore, energy requirements)
of the scrubber device. Generally, the higher the pressure drop the
better the fine particulate removal efficiency. However, at least one
scrubber type has appeared uncharacteristically efficient in fine partic-
ulate removal, namely, the turbulent contact absorber (TCA). Our studies
have indicated that a condensation mechanism is responsible for this good
performance. We are evaluating this mechanism more carefully and hope to
be able to apply it to other scrubbers in other facilities. We are also
piloting a flux force/condensation scrubber which uses a condensation
mechanism for efficient fine particulate removal. Recent data indicate
that scrubbers may be limited in their fine particulate performance by
mist eliminators. Mist eliminators are designed to avoid the entrainment
and carryover of scrubbing liquors from the scrubber into the existing
flue gas. It appears that inefficient mist eliminators on commercial
units have allowed such entrainment thereby leading to carryover of par-
ticulates with subsequent degradation of fine particulate removal per-
formance. Studies continue in this area.
Fabric filtration is being evaluated as a fine particulate control
scheme. We have tested filters applied commercially to three sources,
and find that fabric filtration is quite efficient down to 0.3 microns.
These devices have energy requirements between the low energy usage
precipitators and the high energy usage scrubbers. Our present program
is aimed at increasing the applicability, operability and economic desir-
ability of these very promising devices. We are presently planning to be
involved in a 350-megawatt demonstration program applying fabric filtra-
tion to a low-sulfur coal utility boiler. This will be the first commer-
cial operation of a low-sulfur coal fabric filtration combination. Our
program has also been active at evaluating new ideas and novel devices
for the removal of fine particulates from various sources. Many of these
devices have been tested at bench scale and the most attractive devices
are scheduled for pilot scale testing pending availability of funds.
-------
EPA R&D PROGRAM 75
We have also actively pursued mobile particulate control devices
that enable us to collect data from a variety of particulate sources.
We presently have constructed and tested mobile units of electrostatic
precipitators, scrubbers, and fabric filters which can allow us to
determine what the best control device might be for a given application
and its associate dust problem.
A final element of our program is the control of particulates from
low-sulfur coal combustors. This is really not a new or separate part
of our program, but actually incorporates our ongoing work in other areas.
However, we find it helpful to focus in on this major problem afflicting
those utility and industrial sources who must meet particulate standards
and burn low sulfur coal. Our present emphasis at this time is the
development of fabric filtration and modified electrostatic precipitators
for cost-effective particulate control from these sources.
In conclusion, I believe the results of the conventional fossil fuel
control technology program I have discussed this evening will go a long
way in helping our nation achieve two of its most important goals com-
patibly: energy availability and environmental protection.
-------
76 CLEAN COMBUSTION OF COAL
-------
77
SESSION II - PRECOMBUSTION PROCESSES
SESSION CHAIRMAN: P. STANLEY JACOBSEN, COLORADO SCHOOL OF MINES
RESEARCH INSTITUTE
The Air Quality Act of 1963 initiated at all levels of government
an effort to preserve the nation's air quality. Because of the large
role that coal shares in providing the total fuel energy consumption of
the nation, the effects of coal combustion in electric power generation
were given special emphasis. Coal consumption by utilities was
kOh million tons for 1975 and may be as high as 700 million tons by
1985, and as recommended by the President could be over 1200 million
tons in the year 2000.
Control of sulfur oxide emissions from fossil-fuel-fired combustors
has for years been considered desirable, but with the promulgation of
air quality standards the control has become mandatory. Typically, the
removal of sulfur has been accomplished by postcombustion processes such
as flue gas desulfurization. But the advantage of utilizing precombus-
tion processes as well offers very definite advantages, such as:
1. Using high sulfur coals with little to medium physical cleaning.
2. Availability of a more uniform coal.
3. Lower effective fuel transportation costs.
h. Reduced maintenance costs.
5. Lower coal pulverizing and ash disposal costs.
The overall advantage of combining precombustion processing with
postcombustion processing can be substantial up to a savings of lUo$
over using postcombustion processing alone.
To bring into perspective the complete gamut of precombustion pro-
cesses, the session commences with papers covering the following four
areas: an overview of coal preparation; the influence of mining prac-
tices on the quality of raw coal; research aspects of coal preparation;
and transportation of coal. Three papers in the evening session cover
rather innovative approaches to coal preparation, namely: chemical coal
cleaning, magnetic desulfurization, and a waterless coal cleaning method,
called The Otisca Process.
The entire combustion process including precombustion, combustion,
and postcombustion surely must be considered as an integrated whole. It
is anticipated that the papers on precombustion processes will serve to
refocus attention on the important role this sometimes-neglected segment
plays in the overall combustion process.
-------
78 CLEAN COMBUSTION OF COAL
-------
79
COAL PREPARATION
HISTORY AND DIRECTION
by
Robert L. Llewellyn
Vice President
Roberts & Schaefer Company
Historically, coal has "been prepared in some fashion since the
beginning of mining over 150 years ago in the continent alone, and double
that figure in Europe. Slotted hand shovels were used to leave the
untreated fines inside the mine; men, boys and even women sorted the
coal from impurities on picking tables. Then, many ingenious men, too
numerous to mention, invented jigs, tables, launders and vessels; using
water and air, and the combination of both to separate the coal and
refuse.
Later, sand, minerals (such as magnetite) and heavy liquids were
added to these various washers to increase their efficiencies of separa-
tion. Fifty-five years ago oil agglomeration was patented here in the
States to clean ultra fines, but the use was considered uneconomical
because these fines (at the time) could be wasted. Then, over 30 years
ago electrostatic coal cleaning was demonstrated in a test plant on
lUm x 0, and this system is still resurrected by some about every 10
years without success.
Some confused Convertol with agglomeration, which was a process
using oil with the fines to dewater and improve the bulk density.
Agglomeration reduces the ash. Electrostatic was a unique idea to dry
clean fines by surface separation; it worked in the laboratory, but not
in practice because the coal had to be practically bone dry and drying
coated the particles.
We in the coal preparation field are very fortunate to have excel-
lent cleaning devices available for obtaining efficient physical sepa-
ration of the raw feed from the mines throughout our nation. Chemical
processes are also being considered to make coal practically ash- and
sulphur-free and the activity of all this science is very exciting to
every man and woman connected with coal, which will be affecting the
energy field for the next decade. Huge sums of money will be expended
by government and industry and it must be spent wisely; it must not be
wasted.
If electric power is to be most economical, the fuel must be obtain-
able at a reasonable cost. All of us have been made aware of this when
receiving our monthly or quarterly bills, showing the fuel cost adjust-
ment. Therefore, it behooves all of us to investigate the lowest cost
mining and cleaning systems. We have "mine mouth" operations, which were
-------
80 CLEAN COMBUSTION OF COAL
recognized during the past 15 years as an excellent method to provide
cheap fuel "by locating the power plants at the coal mines. Pipelines
from Cadiz to Cleveland transported coal in slurry form from the mine
to the power station, which was a pioneer to other pipelines in the
west for the same purpose. Railroads obtained unit train load rates
to greatly stimulate coal transportation from mines to power plants by
rail. Mining equipment for surface and deep mines have been manufactured
to improve the cost of coal to the ultimate user.
Never before have there been so many people involved with the art
of coal preparation. This will continue into the next century. And
with such people whose backgrounds vary in education and talent, the
possibility for developing new cleaning techniques are bound to be fan-
tastic. The interest in coal preparation has attracted numerous compan-
ies whose expertise were in different fields of engineering and construc-
tion.
Many consultants have recognized the tremendous growth in the
immediate future in coal preparation and now there are industries outside
of coal mining who are engaged in a tremendous effort and with financial
aid in the coal preparation field. Utility companies with their vast
networks in the U.S.A. have sought and secured large coal reserves for
energy resources. Oil companies are making huge investments in coal
properties, which will be mined and the product beneficiated to suit the
use in the making of energy or steel. Those people involved have learned
that many railroads and steel corporations are already owners of coal
lands since the turn of the century.
It is predicted that great activity will be observed as we look
toward the year 2000. As coal production doubles, or triples, the type
of coal will have to be prepared by the highest efficiency equipment
available. These coals left unmined in the east and midwest by our
grandfathers or great-grandfathers will require the best cleaning methods
to obtain the ash, sulphur and Btu requirements. Our ancestors had to
mine the cleanest seams in the area, because coal preparation was in its
infancy and the price of coal was relatively cheap.
For your information, the first million dollar coal preparation
plant was built near Benton, Illinois, for U.S. Steel in 1918; the book,
Coal Washing by Prochaska, described the plant to some extent. I
visited the plant in 1938 and personally saw the Plato Concentrating
Tables and Feldspar Jigs for fines which were intended to reduce the
sulphur content for metallurgical purposes.
Coal cleaning systems can be utilized economically, but only if
applied properly. Coal Washers differ significantly; magnetic recovery
circuits have wide variations, and unless these factors are not dili-
gently scrutinized, preparation plants will be built without the best
coal beneficiation equipment or systems.
Anyone who has studied flowsheets for years can readily observe the
various systems to prepare coal. First, I believe that one should
thoroughly acquaint themselves with what is going on underground. That
is where the coal and refuse are formed. Just lately I have learned how
-------
COAL PREPARATION 81
much "bottom and roof gets into the R.O.M. The plastic bottles used for
water inside has "become a problem. We all have known about wood, metal,
aluminum cans, and now plastic bottles have to be removed before any
cleaning system.
Preparing the coal ahead of cleaning is most important; recognize
that breaking and/or crushing can help or hurt the raw product. Often
times reducing the size improves the quality and recovery on some coals.
Necessity breeds invention which brought on devices to prepare coal
in the past and some of us here will perhaps be involved in some way in
the future. However, there is nothing wrong to apply proven equipment.
We do not have to invent the Edison light bulb again; the "bulb" has
been improved in our lifetime by many companies.
We must learn by personal experiences and from experiences of
others. It is very difficult to "weed" out good and bad information.
Some integrity has been lost in our life style and I believe that
presently you cannot believe everything you read or hear as you could
some years ago.
Conferences like this one attract people who are very eager to
learn, otherwise they would not be attending. The program input should
be rewarding to all who are interested in coal and its growth.
-------
82 CLEAN COMBUSTION OF COAL
-------
83
THE INFLUENCES OF MINING PRACTICES ON COAL PREPARATION
C.A. Goode
Bureau of Mines
United States Department of the Interior
There is no doubt that the coal reserves in the United States are
more than adequate to meet the energy needs of the country. The reserve
base in the East is over 200 billion tons. Of that total, it is esti-
mated that 160 billion tons will have to be extracted by underground
mining methods owing to the depth of cover, which averages over 600 feet
in the seams that are currently being exploited. The maximum cover in
the East in active mines is about 2,300 feet. Extraction by surface
methods seldom exceeds 100 feet in total depth, but this is more a func-
tion of seam thickness and quality than of equipment capability. Recov-
erability averages from 50 to 60 percent for underground and from 80 to
90 percent for surface methods. Again, coal quality and environmental
aspects tend to constrain the systems.
Coal mining is a high-volume, high-production effort; it is highly
developed art. Preparation of the product has its inception in explora-
tory stages prior to earthmoving. The area to be mined is drilled on
5 mile centers or more, and based on an analysis of the cores, a decision
is made whether or not to proceed. Favorable analyses lead to a second-
phase program whereby the drilling centers are tightened to anywhere from
1 to 3 miles, and again a decision point is reached before finally going
to 1/2" to 1/4-mile core hole centers. Intelligence gathered on the de-
posit characteristics includes superjacent strata information as well as
coal character. Selection of mining and preparation equipment, layout of
the mine, location of points of entry, areas to be mined and mining se-
quence, drainage of the workings, method of mining, and other mining and
marketing considerations are based largely on the analysis of the cores
and any outcrops. High-resolution seismic reflection may also be used to
give definition to the deposit in terms of geologic anomalies (faults,
channel sands, washouts, etc.) that may influence extraction.
Information gathering relative to deposit characteristics continues
throughout the life of the mine. Isopach maps are prepared showing the
deposit properties of interest, be they sulfur, volatile matter, BTU, or
ash. These are useful in establishing blends of coal from various sections
of the mine to produce a more marketable product. Proper input in the
early stages of mine planning is of incalculable importance when consider-
ing the life of the property, which may be in excess of 25 years, and the
total planning effort which may take 5 years before production is realized.
Once a commitment to layout, equipment, and method has been made, any
changes could be economically disastrous.
-------
84 CLEAN COMBUSTION OF COAL
The decision whether to surface mine or to attack the deposit by
underground methods is based on the extent, quality, and thickness of the
deposit and topography. Basically, a ratio of 20 feet of overburden per
foot of coal is the maximum stripping that is currently undertaken. The
size of the equipment is also based on the stripping ratio, the concern
being to match coal extraction rates and overburden removal rates over
the life of the property. Fragmentation by blasting or by ripping equip-
ment and loading practices affect the size and quality of the coal ex-
tracted. Proper choice of equipment for removing overburden, for clean-
ing the top of the exposed coal bench, and for loading the coal can en-
hance the marketability and/or the beneficiation character of the product.
Augering mining is a technique that can be used to extend production
capability after surface mining methods are no longer economical. It is
less efficient in terms of recovery. Usually, the augers are sized
6 inches or more under the seam thickness to be taken in order to leave
coal on the top and/or on the bottom, thus minimizing rock contamination.
Research is continuing in the development of coal-rock interface detectors
which allow for vertical control of the auger to maintain its position
in the seam. Auger coal is normally coarse and dry, but product size
decreases with depth of penetration. Webs of coal are left between holes
to insure the integrity of support during mining. Too thin a web may
lead to seizure of the unit, as floor and roof converge under stress.
Attempts to develop low-cost reliable sensing equipment to maintain
consistent and substantial web thicknesses are being made.
Selective mining can be readily accomplished underground by produc-
ing coal from various sectionsof the mine and blending the product by
controlling section output. Of the two generally used methods of mining,
room-and-pillar mining is more amenable to selectivity than longwall min-
ing. Flexibility of the former in terms of the number of operating units
and relative ease of moving production units from one area of a mine to
another leads itself to planned development and scheduling. Room-and-
pillar development can be accomplished by either conventional methods
(cutting, drilling, blasting, loading, and hauling) or continuous methods
where a single machine mechanically rips the coal from the face, thus
eliminating drilling and blasting. Studies have been conducted on both
these systems in terms of quality and quantity of product, and they show
significant differences.
While size of product is a function of the seam characteristics and
is governed by the hardness of the coal and fracture system, continuous
mining units generally produce finer sizes and introduce more extraneous
materials into the product than do conventional units. The difference
is due more to operator judgment than to anything else. Lower rotational
cutting speed and deeper bit penetration could reduce the size differen-
tial between the two systems. The increased use of continuous miners at
the expense of conventional units attests to the fact that productivity
is the thrust of today's mining. The burden is on the cleaning plants
to upgrade the product.
Relief to overtaxed cleaning facilities can be accomplished by ex-
ercising face preparation techniques, and its importance should not be
overlooked or minimized. Ash-forming or sulfur-bearing impurities can
be avoided in the mining process if they occur, as is often the case, in
-------
INFLUENCES OF MINING PRACTICES 85
association with the bottom and top portions of the seam. The coal it-
self is often a better room or floor material than the associated shales
and clays, and in some instances a 6-inch layer left in place provides
a better bearing surface or can give better protection against falls of
roof as well as yielding a better product. Vertical distribution of im-
purities as well as seam thickness dictates the acceptability of this
practice. Cutting and blasting in conventional mining can be used to
advantage to control banded impurities. Soft bands can actually be cut
out, or hard bands can be fragmented for size control and eventual ease
of physical elimination in a screening operation.
Longwall mining is touted as a safer, higher production system than
room-and-pillar. There are approximately 80 active units in the United
States today, and they account for about 5 percent of the country's under-
ground production; in foreign countries, longwall mining is almost uni-
versally used. There are opportunities in this system for both high
production and quality control. The miner at the working face is in a
protective envelope of hydraulic supports (chocks or shields) and is in
better control of the environment. Dust and methane are more easily
diluted by directed quantities of air. Additionally, the size of the
panels, i.e. up to 600 feet wide and a mile long, allows for more con-
tinuous operation. The mining unit (shearer or plow) is generally under
better control in terms of its attitude in the coalbed. More importantly,
especially in the case of the shearing machine, there is an opportunity
for guidance control and remote operation with the potential for better
quality control and higher production. The mining machine can be kept
within the confines of the seam, and a mining strategy to leave room and/
or floor coal can be readily established. Probes, sensors, and controls
leading to full face control are being developed, and some are already
available. The economic practicability of the control systems has yet
to be established.
The preceding discussion is by no means all inclusive and is fraught
with generalities. The intent here is only to give the nonminer a
glimpse of the mining processes. The high degree of variability of coal
seams and associated strata must be recognized. It is this fact that
makes each mine unique and product control difficult.
BIBLIOGRAPHY
Leonard, J. W. and D. R. Mitchell, ed., Coal Preparation, Third ed.,
AIME, New York, 1968.
Stefanko, R., R. V. Ramani, and I. K. Chopra. The Influence of Mining
Techniques on Size Consist and Washability Characteristics of
Coal Office of Coal Research and Development, Report 61, 1973,
85 pp.
Thomson, P. D. and H. F. York. The Reserve Base of U. S. Coals by
Sulfur Content, BuMines 1C 8680, 1975, 537 pp.
-------
86 CLEAN COMBUSTION OF COAL
-------
87
CURRENT COAL PREPARATION RESEARCH AND DEVELOPMENT
Richard P. Killmeyer, Jr.
Chemical Engineer
Coal Preparation and Analysis Laboratory
Bureau of Mines
U.S. Department of the Interior
Pittsburgh, Pennsylvania
The renewed interest in coal preparation and its rising importance
are evidenced by the many new and innovative research projects ongoing
around the country. The purpose of this paper is to point out and de-
scribe some of this research. Some of the Bureau of Mines new projects
and updates of the major ones will be covered, followed by a few others
from industry and other Government agencies.
The Bureau of Mines has recently awarded three research contracts
to study chemical coal desulfurization, dewatering, and high-gradient
magnetic separation (HGMS). Under one contract Jet Propulsion Labora-
tory will perform bench-scale desulfurization experiments with the aim
of optimizing their low-temperature chlorinolysis process. This
process looks promising as an organic sulfur removal step following
physical removal of pyrite, because it reduces both pyritic and organic
sulfur (up to 80 and 50 percent, respectively). At present, there are
several other chemical desulfurization processes being researched with
only one or two ready for demonstration units.
The dewatering contract is with Dravo Lime Co. and is titled
"Management of Coal Preparation Fine Wastes Without Disposal Ponds."
Fine wastes will be sampled at 10 selected preparation plants and
analyzed for their physical, chemical, and engineering properties.
Bench-scale stabilization tests will be conducted using Calcilox and
other stabilizing agents. The results will be evaluated for effec-
tiveness in hardening fine refuse.
The third contract was awarded to the Naval Ordnance Station at
Indian Head, Maryland. They will investigate the feasibility of mag-
netically separating pyrite from finely crushed coal dispersed in fuel
oil. A Frantz Ferro Filter containing a stack of closely spaced grids
will be used for the experiments. The Bureau is also in the process
of establishing in-house capability in high-gradient magnetic separa-
tion (HGMS). A high-gradient magnetic separator has been purchased
from Sala Magnetics to treat finely ground coal. The results will be
compared with representative fractions treated by the Bureau's two-
stage coal-pyrite flotation process.
-------
88 CLEAN COMBUSTION OF COAL
A continuing HGMS contract proposes an application different from
removing pyrite. Researchers at MIT are studying the problem of re-
covering fine magnetite from dense-medium circuits. Recent advances
in the cleaning of finer sizes of coal require that the entire coal-
magnetite streams be fed directly to magnetic separators. Because of
their high capacity, fever HGMS units may be needed to handle a given
flovrate than conventional drum-type separators, and at a higher re-
covery of fine magnetite particles.
The Bureau's lignite development and utilization program is com-
posed of several projects. In-house work is being done on the wash-
ability of Texas and Western lignites to assess ash and sulfur reduc-
tion potentials. Also, work on an ion exchange process for sodium
reduction is in progress. In conjunction with this work, the
Salt Lake City Metallurgy Research Center, through an interagency
agreement, is developing scale-up data for the design of a continuous
ion exchange pilot test facility. Finally, a contract for the develop-
ment and demonstration of a lignite-pelletizing and pellet-drying
process was awarded to Holley, Kenney, Schott, Inc. The initial
research and development phase of the contract will be carried out at
Michigan Technological University to determine such things as optimum
size consist of the lignite, type of binder, and pellet size.
Two of the Bureau's fine-coal-cleaning projects are entering into
new phases, the dense-medium cyclone and the two-stage coal-pyrite
flotation. A closed-loop dense-medium cyclone test circuit has been
installed with the objective of detailing and optimizing the perform-
ance of cyclones for fine-coal cleaning at lower-than-normal specific
gravities of separation. The program is in cooperation with the owners
of the Homer City Coal Preparation Plant, the Environmental Protection
Agency, and the Electric Power Research Institute.
A coal-pyrite flotation demonstration unit has been constructed
at Barnes and Tucker Lancashire No. 25 mine. It will process 12 tons
per hour in a pyrite flotation circuit to show the improvement in
sulfur reduction over conventional flotation. Also, an instrumentation
circuit which will control reagent addition according to the mass flow
of the feed pulp will be added at Barnes and Tucker, after testing at
the Bureau.
The Ames Laboratory—ERDA, located at the Iowa State University,
is doing work on the magnetic fraction of coal fly ash. The work
covers three areas: (l) the recovery of the magnetic iron oxide from
the fly ash by magnetic separation, (2) the evaluation of the physical
and chemical properties of the two products of separation, and (3) the
utilization of the magnetic material. Coal fly ash is a product of
combustion and is being produced in increasing quantities as the use
of coal increases. It represents a considerable source of iron and
aluminum oxide.
Roberts and Schaeffer Co. has been experimenting with the
upside-down, or inverted, cyclone. They recently started up the first
commercial installation that uses dense-medium inverted cyclones. It
is a 100-tph plant owned by the Wise Coal Co. in Virginia, and it is
-------
COAL PREPARATION RESEARCH 89
cleaning the full 1-1/2-inch by 0 size fraction in the cyclones. Early
reports indicate good performance, although the plant operation is
still "being optimized.
A major piece of washing equipment which has remained essentially
unchanged for many years is the Baum jig. Now, however, McNally-
Pittsburg is developing an innovation which should reduce the amount
of misplaced float material in the refuse. The float mechanism for
sensing the bed depth and the refuse gate for rejection usually work
the width of the jig, but McNally is dividing these units into three
separately working sections across the jig. This accounts for varia-
tions in the bed. They have been running tests at some preparation
plants with the revamped jigs, and results seem promising.
Finally, FMC Corp. is researching two pieces of cleaning equip-
ment . One, the dry table, has been in development for a few years.
It is a vibrating, wedge-shaped separator which is being touted for
cleaning coal in the West where water is scarce. FMC is currently
testing a 12-foot pilot unit on bituminous and subbituminous coals for
ash and pyrite reduction and claim the cleaning performance is somewhat
like that of a Baum jig.
The other FMC washer is a wet vibrating table, about the size and
shape of the conventional concentrating table. Work on it has not been
as extensive as on the dry table, but FMC is beginning additional re-
search on it. The company claims it has over twice the capacity of a
concentrating table, which would greatly increase the efficacy of wet
tables in preparation plants.
These examples of current R & D projects show the wide variety of
ideas in washing, dewatering, and desulfurization for Btu recovery and
environmental protection. Many of these ideas have become feasible
because of the high price of coal and the need to meet environmental
regulations.
-------
90 CLEAN COMBUSTION OF COAL
-------
91
COAL TRANSPORTATION IN 1985
by
David J. Hoexter
Office of Energy Programs
U.S. Department of Commerce
Introduction
There are a number of basic questions concerning the transportation of
coal. Some of these questions include: What are the factors
affecting the modes by which coal will be shipped? Can we supply the
required facilities and equipment? And, will special incentives or
legislation be required to insure that the demand for transportation
services will be met? A great deal of analysis, and a fairly large
literature, is part of the public record on this topic. By presenting
and comparing some of these analyses, we can get a good handle on the
magnitude of the transportation services which will be required.
It is the stated intention of the current Administration, as it was
the goal of the previous Administration, that production and consump-
tion of coal in this country in 1985 should be slightly in excess of
one billion tons; the exact quantity is given in various places as
anywhere from 1.1 billion tons to a "doubling" of our use based on
1975 or 1976 consumption. This same range of estimated use has been
suggested by private sector sources, both within and outside of the
coal and transportation industries.
Factors Affecting Coal Delivery Mode
What are the factors affecting the modes by which coal will be
shipped? First, there is the basic cost factor, expressed as dollars
per ton or cents per ton-mile. Given the more rapid relative
development of large open-pit mining operations in the West and, to
a much lesser extent, underground western mining operations, a premium
is arising on the ability to move large quantities of coal long
distances. The first candidates which come to mind when considering
very long distance hauling of coal are dedicated, or unit, trains and
slurry pipelines. Campbell and Katell, in their 1975 report for the
Bureau of Mines, do provide some estimates of the comparable modal
costs. They point out that, for similar volumes of coal, the cost per
ton mile varies between 3 and 7 mills for slurry pipelines, and 4 and
9 mills for unit trains. This comparison only serves to tell us that
there will probably have to be features or conditions unique to each
case before a decision on which mode to select can be made. Both
modes require long-term commitments on the part of coal users, yet
both have features which recommend them. A slurry pipeline, once
installed, will prove to be highly resistant to inflation in operating
costs, simply because its operating costs are minimal. The railroads
cannot make this claim. Reliability of service provides another point
of distinction between the two modes. The two existing slurry pipe-
lines — in Ohio and Arizona — have been available for operation over
-------
92 CLEM COMBUSTION OF COAL
98% of the time. No railroad has provided this degree of reliability.
Where flexibility of route and scheduling is desired, the railroad is
the mode of choice.
Not all coal will be shipped very long distances. In fact, over
three-quarters of our coal will still be coming from eastern surface
and underground mines in 1985. There is likely to be a greater
variety in the size of these mining operations, as well as a greater
variety in the distances between the mine and the final consumer. In
this situation, the cost factors favor trucks, conveyors, non-unit
trains and barges. In the same study by Campbell and Katell, the
costs for truck and conveyor belt for very short distances (less than
15 miles in the case of conveyors, and, not specified but quite likely
less than, 100 miles for trucks) were about equal to the costs for unit
trains and slurries on a cents per ton-mile basis. The average cost
per ton-mile for barges was the lowest for any mode examined. Data
obtained from the National Energy Transportation Study recently
completed by the Congressional Research Service indicated a modal cost
of just over 3 mills for an average trip of 480 miles. Obviously, the
limiting factor in the use of barges is the availability of usable
rivers and canals. The bulk of the coal, however, in 1985 will still
be carried by the railroads. They are second only to trucks in terms
of flexibility, and they provide the best unit costs for the medium
tonnages and distances which will be encountered in bringing eastern
coal to eastern users..
There are other factors which affect the choice of coal transportation
modes.
There are political factors which affect the choice of mode. I
believe the issue of eminent domain will be resolved at the national
level in favor of the slurry pipeline interests. The question of
economics is not a stumbling block; the railroads have even agreed
that building a slurry pipeline is cheaper than laying a new rail
line.
The problem of water is at the heart of the controversy. The use of
the West's limited water supply will be the last political issue to
be resolved in the matter of slurry pipelines. A smaller parallel
line could carry the water back from Arkansas to Wyoming to be re-used.
This would minimize the amount of water used in the slurry, and solve
the problem of what to do with 20,000 acre feet of almost unreclaimable
water.
Another political factor concerns such environmental constraints as
sulfur emission standards and surface mine reclamation standards.
According to a number of studies, the impact of the surface mine
legislation will, among other things, reduce coal output in Appalachia
by tens of millions of tons below what it otherwise would have been.
The likely effect of this would be to increase the modal shares
accounted for by unit trains and slurry pipelines in the west. To
the extent that demand for low-sulfur coal increases relative to the
demand for all coal, this, too,' bodes well for the owners and
operators of unit trains. The impact on truck hauling and short-haul
-------
COAL TRANSPORTATION 93
rail operations of a drop in demand for medium to high sulfur coals
would not be insignificant.
There are also structural factors involved in determining the mode of
shipment. One structural factor which should be mentioned is owner-
ship of coal reserves. The railroads have the largest coal reserve
base in private ownership. The oil and gas companies are estimated to
hold the next largest share, and coal-consuming industries such as the
steel, electric utility and chemical companies are next. The implica-
tions of this pattern of ownership would tend to confirm an already
sound view that the railroads will continue to dominate the coal
hauling business.
Transportation Equipment Requirements and Cost
Can we supply the facilities and equipment required to move the coal?
For estimates of the equipment requirements, we have several recent
reports to draw on. The first report is "Coal Transportation Practices
and Equipment Requirements to 1985," written by Gary Larwood and David
Benson of the Bureau of Mines. The authors assume that 1.2 billion
tons of coal will need to be moved in 1985. According to their study,
the railroads will be required to carry between 894 million and 945
million tons of coal. To do this would require between 126,000 and
142,000 hopper cars or about 1,260 to 1,420 unit trains in 1985. A
report prepared by Beehtel Research & Engineering for the Office of
Energy Programs projected a requirement for the equivalent of about
700 unit trains in 1985.
The third report is the 'toal Transportation Capability of the Existing
Rail and Barge Network, 1985 and Beyond" prepared by Manalytics, Inc.,
for the Electric Power Research Institute in September of 1976. Two
scenarios were prepared. In the first, it was assumed that 1.5 billion
tons of coal would be moved by rail in 1985; in the second, it was
assumed that 1.24 billion tons would be shipped by rail. The rolling
stock required for these amounts is projected to be the equivalent of
1,947 unit trains and 2,129 unit trains for the first and second
scenarios, respectively. Manalytics estimated the number of locomo-
tives required to be 9,735 and 10,645 in the first and second scenarios,
respectively. Larwood and Benson estimated the number of locomotives
required to be between 6,900 and 7,800.
Considering the number of barges which will be needed in 1985, Larwood
and Benson projected that between 1,800 and 3,400 barges of 1,400 ton
equivalent will be required to move the barge share of coal traffic.
The Office of Energy Programs report, converted to 1,400 ton equiva-
lent units, puts this figure at just over 3,000 barges.
How much is all of this equipment going to cost? We have the estimates
of two companies. The first is the Beehtel Corporation and the second
is the Bankers Trust Company. Both estimates cover the years 1976-
1985, inclusive, and are expressed in 1974 dollars.
-------
94 CLEM COMBUSTION OF COAL
The Bechtel Corporation estimates that approximately $8.8 billion
will be required to design, construct and start up transportation
facilities. Bankers Trust Company projects that about $10.25 billion
will be required to provide the same services. Both estimates show
expenditures beginning at around three quarters of a billion dollars
in the late 1970's, and peaking in 1985 at $1.1 billion and $1.2
billion, respectively.
Modal Share Estimates
It is clear that there will not be any very large shifts in the share
of coal hauled by a particular mode. Rail traffic, which in the first
quarter of 1977 moved half of the coal shipped, will likely increase
its share of the volume as western mining comes into its own and the
increased use of unit trains thus lowers unit shipping costs. The
Federal Power Commission, in January of this year, released a .staff
study which estimated the modal shares for coal delivered to new coal-
fired electric utility units coming on stream between now and 1985.
The railroads' share of this traffic is estimated by the FPC to be
about 66% in 1980 and 62% in 1985.
River traffic, which accounted for 15% of coal moved in the first
quarter of 1977, is projected by the FPC to carry 9% of the coal
required by new coal-fired utilities in 1980, and just over 6% of
the coal required in 1985.
The FPC projects a large increase in the share of truck traffic
bringing coal to new generating facilities relative to its current
level. In the first quarter of 1977, trucks hauled over 13% of all
coal; in 1980 and 1985, trucks are projected to account for about
22% of all shipments to new coal-fired facilities. The single
greatest cause of this rise in modal share is probably the increase
in the number of mine-mouth electric generating plants which are
expected to come on stream.
The remainder of the coal-carrying modes — conveyors, pipelines,
etc. — will have their shares affected little, if at all, although
the actual absolute tonnages they carry may increase dramatically.
-------
COAL TRANSPORTATION 95
REFERENCES
Battelle Columbus Laboratories. "A Report to the Interagency Coal
Task Force, Project Independence Blueprint, on the Manpower
Requirements of Coal Transportation." Washington, D. C.: June,
1974.
Battelle Columbus Laboratories. "A Report to the Interagency Coal
Task Force, Project Independence Blueprint, on the Modal Transpor-
tation Costs for Coal in the United States." Washington, D. C.:
May, 1974.
Bechtel Corporation. "Capital, Manpower, Materials and Equipment
Requirements for a Department of Commerce Energy Projection." San
Francisco, California and Washington, D. C.: December, 1976.
Campbell, T.C., and Katell, Sidney. "Long-Distance Coal Transport:
Unit Trains or Slurry Pipelines." Washington, D. C.: Department
of the Interior, 1975.
Congressional Research Service and U.S. Geological Survey. "National
Energy Transportation." Committee Print of the U.S. Senate #95-15.
Washington, D. C.: May, 1977.
Electric Power Research Institute. "Coal Transportation Capability of
the Existing Rail and Barge Network, 1985 and Beyond." Palo Alto,
California. September, 1976.
Federal Power Commission. "Status of Coal Supply Contracts for New
Electric Generating Units 1976-1985." Washington, D. C.: January,
1977.
Larwood, Gary M. and Benson, David C. "Coal Transportation Practices
and Equipment Requirements to 1985." Washington, D. C.: Department
of the Interior, 1976.
Mutschler, P.H., et. al. "Comparative Transportation Costs of
Supplying Low-Sulfur Fuels to Midwestern and Eastern Domestic
Energy Markets." Washington, D. C.: Department of the Interior,
1973.
Peat, Marwick, Mitchell and Co. "Railroad Freight Car Requirements
for Transporting Energy, 1974-1985." Washington, D. C.: Federal
Energy Administration, November, 1974
-------
96 CLEAN COMBUSTION OF COAL
-------
97
COAL DESULFURIZATION TEST PLANT STATUS - JULY 1977
L.J. Van Nice, M.J. Santy, E.P. Koutsoukos, R.A. Orsini and
R.A. Meyers
TRW Systems and Energy
Redondo Beach, CA 90278
I. INTRODUCTION
An 8 metric ton/day process test plant for chemical desulfurization
of coal has just been built at TRW's Capistrano Test Site in California.
The plant, shown in Figures 1 and 2, was constructed under an Environ-
mental Protection Agency sponsored project for the development of the
Meyers Process. Current plans call for plant shakedown followed by
processing of 100-200 tons of American Electric Power Service Corpor-
ation's Martinka Mine coal.
The Meyers Process removes up to 80 percent of the total sulfur
content of coal through chemical leaching of 90 to 95 percent of the
pyritic sulfur contained in the coal matrix with aqueous sulfate
solution at temperatures of 90° to 130°C. The ferric sulfate content
of the leach solution is regenerated at similar temperatures using air
or oxygen, and elemental sulfur and iron sulfates are recovered as
reaction products or alternatively gypsum can replace all or a portion
of the iron sulfates as a product. The physical form of the coal re-
mains unchanged; only pyrite and some inorganic materials are removed.
Low and medium organic sulfur coal can be desulfurized prior to
combustion using the Meyers Process (1,2) to meet governmental require-
ments for sulfur oxide emissions.
The Environmental Protection Agency estimates that 90 x 109 tons
(80 x 109 metric tons) of coal reserves in the U.S. Appalachian Coal
Basin can be reduced in sulfur content by the Meyers Process to levels
which will meet New Source Performance Standards. Successful bench-
scale testing (3,4) and promising engineering analyses (3,5-7) together
with applicability testing (8,9), have led the Environmental Protection
Agency to sponsor the construction and operation of a test plant.
Present estimated processing costs using utility financed depreci-
ation of capital, and including coal grinding and compaction of the
product (where necessary), are $8-12/ton (3,6). The price varies with-
in this range according to factors such as offsite, coal pre-cleaning,
reaction rate of the particular coal and coal top-size utilized, i.e.,
3/8"processing and pre-cleaning of run-of-mine coal contribute to lower
cost. Recent advances, which have not yet been fully designed, promise
to reduce the above costs.
Process chemistry, and test plant design and operation will be
described below.
-------
CO
§
g
Figure 1. Test Plant - Front View
-------
I
I
M
i
Figure 2. Test Plant - View Through Tank Farm
-------
100 CLEAN COMBUSTION OF COAL
II. PROCESS CHEMISTRY, KINETICS AND SCHEME
The process is based on the oxidation of coal pyrite with ferric
sulfate (Equation 1). The leaching reaction is highly selective to
pyrite with 60 percent of the pyritic sulfur converted to sulfate
sulfur and 40 percent to elemental sulfur. The reduced ferric ion Is
regenerated by oxygen or air according to Equations 2 or 3.
FeS2 + 4,6 Fe2(S04)3 + 4.8 H20 -v 1}
10.2 FeS04 + 4.8 H2S04 + 0.8S
2.4 02 + 9.6 FeS04 + 4.8 H2S04 + 2j
4.8 Fe2(S04)3 + 4.8 H20
2.3 02 + 9-2 FeS04 + 4.6 H2S04 -»• 3j
4.6 Fe2(S04)3 + 4.6 H20
Regeneration can be performed either concurrently with coal pyrite
leaching in a single operation or separately. The net effect of the
process is the oxidation of pyrite with oxygen to yield recoverable iron,
sulfate sulfur, and elemental sulfur. The form of process products
varies to some extent with the degree of regeneration performed. Thus,
Equations 1 and 2 lead to the overall process chemistry indicated by
Equation 4 with products being a mixture of iron sulfates and elemental
sulfur. Equations 1 and 3 yield ferrous sulfate, sulfuric acid, and
elemental sulfur as indicated by Equation 5.
FeS2 + 2.4 02 * 0.6 FeS04 + 0.2 Fe2(S04)3 + 0.8S 4)
FeS2 + 2.3 02 + 0.2 H20 -»• FeS04 + 0.2 H2$04 + 0.8S 5)
Several options exist in product recovery. Iron sulfates may be recov-
ered as pure solids by stepwise evaporation of a spent reagent slip-
stream with ferrous sulfate being recovered first because of its lower
solubility. Alternately, ferrous sulfate may be recovered by crystal-
lization, and ferric sulfate or sulfuric acid removed by liming spent
reagent or spent wash water slipstreams. Iron sulfates may be stored
as solids for sale or may easily be converted to highly insoluble basic
iron sulfates (by air oxidation) or calcium sulfate (by low-temperature
solid phase reaction) for disposal. Elemental sulfur may be recovered
from coal by vaporization with steam or by vacuum, or it can be leached
out with organic solvents such as acetone. Product marketability and
product recovery economics will dictate the choice. Recovery economics
may be influenced by quantity and concentration of product in the
process effluent streams which in turn are influenced by the pyrite
concentration in the coal and the desired extent of desulfurization.
The process has been extensively studied at bench-scale. Parameters
investigated included coal top-size, reagent composition, slurry concen-
tration, reaction temperature and pressure, and reaction time.
-------
COAL DESULFURIZATION 101
Additional investigations completed or underway include concurrent coal
Teaching-reagent regeneration, product recovery, product stability,
and the effect of coal physical cleaning on process performance and
economics. The process scheme depicted in Figure 3 is based on the
bench-scale testing. Coal is a) crushed to the desired size for pro-
cessing, b) contacted with hot recycled reagent in the Mixer (90-100°C),
c) leached of pyrite in the Reactor(s) with simultaneous or separate
reagent regeneration, d) washed with hot water, and e) stripped of
elemental sulfur, dried and finally cooled. The iron and sulfate sulfur
are recovered from spent reagent slipstreams prior to reagent recycle.
Figure 4 shows typical data on pyrite removal rates from Appalachian
coal as a function of temperature. Removal of 10-20 percent of the
pyrite is obtained during slurry mixing and heat-up.
Bench-scale data indicated that the pyrite leaching rate from coal
can be adequately represented by the empirical rate expression (Equation
6).
'L ' -
where
KL = AL exp (-EL/RT),
Wp = wt percent pyrite in coal,
Y = ferric ion-to-total iron ratio in the reactor reagent, and
A. and E. are constants for each coal and particle size at least
over most of the reaction range.
The leach rate is a function of coal type. Pyrite extraction rates
vary considerably, as detailed in a study of the Meyers Process as
applied to U.S. coal (8), e.g., there was more than one order of magni-
tude difference between the fastest and slowest reacting coal. The
reagent regeneration rate is governed by the rate expression (Equation
' ) •
_ dFe = K p (Fe+2)2 7)
K P Ue ' n
_
R --- d R 02
where
KR = AR exp (-ER/RT),
Pn = oxygen partial pressure,
°2
Fe+2 = ferrous ion concentration in the reagent solution, and
A and E are constants.
-------
102
CLEAN COMBUSTION OF COAL
COAL
Figure 3. Process Flow Schematic
90
80
70
60
3?
I"
UJ
Of.
m 40
30
20
10
130"C
120»C
i
110°C
0.5
1.0 1.5 2.0
REACTION TIME, HOURS
2.5
3.0
Figure 4. Temperature Effect on Processing of 14 Mesh
Top-Size Lower Kittanning Coal (33% w/w Slurries)
-------
COAL DESULFURIZATION 103
Engineering evaluation of available data shows that it is prefer-
able to process fine coal (<2 mm top-size) under simultaneous Teaching-
regeneration conditions in the temperature range of 110-130°C until the
majority of the pyrite 1s leached out. Ambient pressure processing
(approximately 100°C) is indicated for the removal of the last few
tenths percent of pyrite since the low Wp value substantially reduces
the rate of ferric ion consumption and, therefore, the need for
simultaneous reagent regeneration. Ambient pressure processing appears
to be indicated also for coarse coal (e.g., 10 millimeter top-size) for
several reasons. It is difficult to continuously feed a non-siurryable
coal into and remove it from a pressure vessel. It is much easier and
less costly to drain leach solution from the coal and pump it into a
small pressure vessel for regeneration. Also the slower reaction rate
with coarse coal would require much longer residence times and unreason-
ably large total volume for pressure vessels. These engineering evalu-
ations were part of the data used to design the test plant.
III. TEST PLANT DESIGN AND OPERATION
A test plant sized to process up to 8 metric tons per day of coal
has been built, under the sponsorship of the Environmental Protection
Agency, at TRW's Capistrano Test Site. A plant flow diagram is shown
in Figure 5. The facility is capable of on-line evaluation of the
following critical process operations:
• Pressure leaching of pyritic sulfur from 150 micron to 2 mm top-size
coal at pressures up to 100 psig,
• Regeneration of ferric sulfate both separately, for processing larger
top-size coal or low pyrite coal, and in a single vessel with the
leaching step for processing of suspendable coal,
• Filtration of leach solution from reacted coal,
• Washing of residual iron sulfate from the coal.
Iron sulfate crystallization, elemental sulfur recovery and coal-
drying unit operations will be evaluated in an off-line mode in equip-
ment vendor pilot units. Leaching of 10 mm top-size coal can be
evaluated in an off-line mode in an atmospheric pressure vessel installed
in the test plant. Coarse coal processing (5-10 mm top-size) has been
very promising in laboratory tests (3). If this approach proves out
in bench-scale evaluations, more extensive and on-line coal leaching
units can be readily added to the present test plant. Processing fine
coal allows the highest rate of pyritic sulfur removal, while processing
coarse coal, although slower, allows lower cost coal dewatering units
and the direct shipping of desulfurized coal product without need for
pelletizing.
The test plant constructed at the Capistrano Test Site is a highly
flexible facility capable of testing the numerous alternate processing
modes of potential interest in the Meyers Process. The flow diagram
shown in Figure 3 presents an equipment train for continuous process
testing of slurried coal. Fine coal ground to the desired size is
-------
104
CLEAN COMBUSTION OF COAL
ATMOS.
WATER
1 I
Y}™
1
ATMOS.
TO TRUCK FOR DISPOSAL
Figure 5. Test Plant Flow Diagram
-------
COAL DESULFURIZATION 105
stored under nitrogen gas in 1.8 metric ton sealed bins. As required,
bins are emptied into the feed tank (T-l). Dry coal is continuously
fed by a live bottom feeder to a weigh belt which discharges through
a rotary valve to the three stage mixer (Stream 1). The aqueous iron
sulfate leach solution (Stream 2) enters the mixer after first passing
through a foam breaker (T-2). Steam is added (Stream 3) to raise the
slurry to its boiling point. Foaming will occur in the early stages of
mixing, but will cease when particle wetting is complete. It is
believed that the mixing time and conditions necessary to complete the
wetting and defoaming of the slurry will depend on the coal type and
size and on the residual moisture in the feed coal. To allow study of
the mixing parameters, the mixer stages have variable volume with
variable speed agitators, and the feed flow rates for coal, leach
solution and steam can be varied over wide ranges.
The defoamed slurry (Stream 4) is pumped to a five stage pressure
vessel (Reactor 1) in which most of the pyrite removal reaction occurs.
Some of the pyrite reaction occurs during mixing, but in the mixer the
reaction rate slows rapidly because the remaining pyrite (Wp) decreases
and because the ferric iron is rapidly being converted to ferrous iron
(Y decreases). The pressure reactor overcomes the decreased rate in
two ways. First, it increases the temperature (and pressure) to
increase the reaction rate constant. Second, oxygen is introduced
under pressure to regenerate ferric iron and maintain a high solution
Y. The flow diagram shows that steam and oxygen can be added to any
or all of the five stages and that cooling can be provided for any
stage if necessary to remove the excess heat of reaction. The unused
oxygen saturated with steam (Stream 7) is contacted in a small pressure
vessel (T-3) with the feed leach solution (Stream 5) to provide heated
leach solution for the mixer (Stream 2) and cooled vent gas. The vent
gas from both T-2 and T-3 are scrubbed in T-4 to remove any traces of
acid mist. The reaction parameters of importance have already been
well studied at laboratory and bench-scale in batch mode. The test
plant reactor will accommodate the necessary studies of key parameters
in a continuous reactor at coal throughputs between 2 and 8 metric tons
per day. Parameters which will be studied include temperature,
pressure, oxygen purity, slurry concentration, iron sulfate concentra-
tion, iron sulfate concentration, acid concentration, residence time
per stage, number of stages, mixing energy, type of mixing, coal size
and type. The reactor can also be used to study leach solution regen-
eration in the absence of coal.
Reacted coal slurry (Stream 8) at elevated temperature and
pressure is flashed into a gas-liquid separator vessel (T-5). The
steam generated (Stream 9) is condensed in T-4 and the condensate plus
any entrained acid mist is removed with the water. The residual slurry
(Stream 10) is fed to a belt filter. The filtrate, which is regener-
ated leach solution, is removed from the coal slurry through a vacuum
receiver (T-9) and pumped (Stream 12) to a large leach solution storage
tank (T-6). The coal on the filter belt is washed with water (Stream
11) and discharged from the filter belt. The wash water is removed
through a vacuum receiver (T-10) and sent to a large liquid-waste
holding tank (T-8) for subsequent disposal. The filter is a highly
versatile unit which should provide the data necessary for scale-up.
-------
106 CLEM COMBUSTION OF COAL
It has variable belt speed, variable belt areas assigned to washing,.
variable cake washing rates, belt sprays if needed to control blinding
of the pores in the belt, and steam nozzles to provide for partial cake
drying.
As an alternate process step, the slurry from the flash tank (T-5)
can be passed into a secondary reaction vessel (Reactor 2). At typical
coal feed rates, this vessel can be filled in about two hours and then
closed off, stirred and heated for any desired period of time before
being pumped to the filter. Residence times up to about 10 hours are
available in the primary reactor, Reactor 1. This secondary reactor
can be used to extend residence times to much longer times for examining
the removal of final traces of pyrite or examining any other long term
behavior. The stirred vessel also can serve to repulp the filter cake
for additional coal washing studies.
The final item of major equipment in the test plant is the coarse
coal contact vessel (Reactor 3). This insulated and heated tank will
hold a full bin (about 1.8 metric tons) of coarse coal (5 to 10 milli-
meter top-size). The principal use for this vessel is to convert the
regenerated leach solution in storage tank T-6 to a more depleted solu-
tion in the process feed tank, T-7. In general, the iron sulfate leach
solution in the filtrate going to tank T-6 will have a high Y because
no secondary reactor was in use. For some test conditions, the feed
to the process must be at a lower Y to simulate recycle leach solution
from a secondary reactor. Passing all or some portion of the solution
through coal will lower the Y of the solution to the desired value.
This vessel is basically a coarse coal reactor and if appropriate
sampling ports and possibly some flow distribution internals were
added, it could be used to obtain design data for coarse coal processing.
Solution tanks are sized at about 50,000 liters to provide for
about a week of continuous operation on the same feed without recycle
or change. It also provides for uniform leach solution and coal samples
of a large enough size for product recovery studies performed by equip-
ment vendors. Operation at the scale of the test plant will provide
experience and data expected to be adequate for the design of a demon-
stration-size commercial plant.
REFERENCES
1. R. A. Meyers, J. W. Hamersma, J. S. Land and M. L. Kraft, Science.
177. 1187 (1972).
2. R. A. Meyers, "Removal of Pyritic Sulfur from Coal Using Solutions
Containing Ferric Ions," U.S. Patent 3768988 (1973).
3. E. P. Koutsoukos, M. L. Kraft, R. A. Orsini, R. A. Meyers, M. J.
Santy and L. J. Van Nice (TRW Inc.), "Final Report Program for
Bench-Scale Development of Processes for the Chemical Extraction of
Sulfur from Coal," Environmental Protection Agency Series
EPA-600/2-76-143a (May 1976)."
-------
COAL DESULFURIZATION 107
4. J. W. Hamersma, E. P. Koutsoukos, M. L. Kraft, R. A. M.eyers, G. J.
Ogle, and L. J. Van Nice (TRW Inc.), "Chemical Desulfurization of
Coal: Report of Bench-Scale Developments, Volumes 1 and 2,
"Environmental Protection Agency Series, EPA-R2-173a (February 1973).
5. E. M. Magee (Exxon Research and Engineering Co.), "Evaluation of
Pollution Control in Fossil Fuel Conversion Processes, Coal Treat-
ment: Section 1. Meyers Process," Environmental Protection
Technology Series, EPA-650/2-74-009-k (September 1975).
6. W. F. Nekervis and E. F. Hensley (Dow Chemical, U.S.A.), "Conceptual
Design of a Commercial Scale Plant for Chemical Desulfurization of
Coal," Environmental Protection Technology Series. EPA-600/2-75-051
(September 1975).
7. M. Rasin Tek (U. of Michigan), "Coal Beneficiation," in Evaluation
of Coal Conversion Processes, PB-234202 (1974).
8. J. W. Hamersma and M. L. Kraft (TRW Inc.), "Applicability of the
Meyers Process for Chemical Desulfurization of Coal: Survey of 35
Coal Mines," Environmental Protection Technology Series,
EPA-650/2-74-025-a (September 1975).
9. U.S. Environmental Protection Agency, Office of Research and Devel-
opment, Washington, DC, "Applicability of the Meyers Process for
Chemical Desulfurization of Coal: Initial Survey of Fifteen Coals,"
by J. W. Hamersma, et al., Systems Group of TRW, Inc., Redondo
Beach, CA, Report No. EPA-650/2-74-025 (April 1974) Contract No.
68-02-0647.
-------
108 CLEAN COMBUSTION OF COAL
-------
109
STATUS AND PROBLEMS IN THE DEVELOPMENT OF
HIGH GRADIENT MAGNETIC SEPARATION (HGMS) PROCESSES
APPLIED TO COAL BENEFICIATION*
Y. A. Liu and C. J. Lin
Department of Chemical Engineering, Auburn University
Auburn, Alabama 36830
ABSTRACT
An overall discussion of the status and problems in the develop-
ment of HGMS processes applied to coal beneficiation is presented. In-
cluded in the discussion are such topics as: bench-scale and pilot-
scale experimental studies; quantitative modeling of experimental
results; conceptual process design and cost estimation; and comparison
with alternate technologies for coal beneficiation. The needs and
opportunities in the future research and development work related to
magnetic beneficiation of coal are also suggested.
INTRODUCTION
High gradient magnetic separation (HGMS) is a new technology which
provides a practical means for separating micron-size, feebly magnetic
materials on a large scale and at much faster flow rates than are pos-
sible in ordinary filtration. The technology is also applicable to
separating nonmagnetic materials which can be made to associate with
magnetic seeding materials. It was developed in 1969 for the wet sepa-
ration of weakly magnetic contaminants from kaolin clayl~6. A typical
HGMS unit for this wet application is shown schematically as Figure 1.
The electromagnet structure consists of energizing coils and a sur-
rounding iron enclosure. The coils in turn enclose a cylindrical work-
ing volume packed with fine strands of strongly ferromagnetic packing
materials such as ferritic stainless steel wool. With this design,
a strong field intensity up to 20 kOe can be generated and distributed
uniformly throughout the working volume. Furthermore, by placing in
the uniform field the ferromagnetic packing materials which increase
and distort the field in their vicinity, large field gradients of the
order of 1-10 kOe/ym can be produced. In the wet beneficiation of
kaolin clay, the HGMS unit is used in a batch or cyclically operated
process like a filter. The kaolin feed containing the feebly magnetic
*This work was supported fay the National Science Foundation (grant no,
GI-38701), by the Energy Research and Development Administration
(contract no. W-7450-eng-26 ORNL/Sub-7315), by the Gulf Oil Foundation
and by Auburn University. The authors wish to thank Messrs, A. W.
Deurbrouck and R. E. Hucko of the Coal Preparation and Analysis Labor-
atory of Bureau, of Mines for their continued assistance in providing
valuable technical information and coal samples for the work reported
herein.
-------
110
CLEAN COMBUSTION 07 COAL
CLEANED
"EFFLUENT
IRON
ENCLOSURE
ELECTROMAGNET
BED
STAIN LESS STEE
WOOL STRANDS
NON MAGNETIC
PARTICLES
FEED
MAGNETIC PARTICLES
Figure 1. Cyclic High Gradient Magnetic Separator^
contaminants of low concentrations is pumped through the stainless
steel wool packing or matrix of the separator from the bottom while the
magnet is on. The magnetic materials (mags) are captured and retained
inside the separator matrix, and the nonmagnetic components (tails)
pass through the separator matrix and are collected as the beneficiated
products from the top of the magnet. After some time of operation, the
separator matrix is filled to its loading capacity. The feed is then
stopped, and the separator matrix is rinsed with water. Finally, the
magnet is turned off, and the mags retained inside the separator matrix
are backwashed with water and collected. The whole procedure is re-
peated in a cyclic fashion. The significance of HGMS, and its latest
engineering and commercial development have been adequately discussed
elsewhere*-"".
Because of its very low costs and outstanding technical perform-
ance demonstrated in the kaolin application, HGMS was recently adapted
to solving many separation problems related to minerals and chemical
processing industries. An important and promising application of HGMS
is the magnetic removal of inorganic sulfur and ash from coal. Pre-
vious experimental investigations have indicated clearly that most of
the mineral impurities in coal which contribute to its pyritic sulfur,
the sulfate sulfur and the ash content are paramagnetic. These sulfur-
bearing and ash-forming minerals, if sufficiently liberated as discrete
particles, can be separated normally from the pulverized diamagnetic
coal by magnetic means?"**, indeed, the technical feasibility of apply-
ing HGMS to the beneficiation of pulverized coal suspended in water has
been demonstrated in a number of recent studies, with substantial
amounts of sulfur and ash removal reported""-^. Recently, a bench-
scale feasibility study of utilizing a novel air-fluidized separator
matrix in HGMS applied to the beneficiation of the dry pulverized coal
was conducted in the author^' laboratory in cooperation with Oak Ridge
National Laboratory. The available results from this study were quite
encouraging, indicating that a dry HGMS process for coal beneficiation
-------
HIGH GRADIENT MAGNETIC SEPARATION
111
with performance comparable to the wet magnetic approach could be devel-
oped at much cheaper cost.-"
In this paper, an overall discussion of the recent development and
current status of HGMS processes applied to coal beneficiation is pre-
sented. A quantitative assessment of the technical and economical fea-
sibility of applying HGMS to remove sulfur and ash from both dry and
wet coals is given. In particular, because of its more advanced state
of development, the magnetic beneficiation of water slurries of pul-
verized coals is discussed in some detail. The needs and opportuni-
ties in the future research and development work related to magnetic
beneficiation of coal are also suggested.
MAGNETIC BENEFICIATION OF COAL/WATER SLURRY
"D&>CAJJp£Lon. ^ A conceptual process for the magnetic
beneficiation of pulverized coal suspended in water by HGMS is shown
schematically in Figure 2.
Water Supply
1.653 million gal/day
Wash Water
1440 gal/cycle
HGMS 7'D, 20" L.
2.61 cm/sec Velocity
480 gal. Canister
Volume
0.96 Tons S.S. Wool
94% Void Volume
38.46 GPM/Ft2
Cycle Time 6.lmin.
Duty Cycle 67.4%
Water
Mags
Refuse 9.92 TPH-
Toils
Water
•Vacuum Filter
Thermal Dryer
Clean Dry Coal
56.21 TPH
Figure 2. Magnetic Beneficiation of Coal/Water Slurry by HGMS
A coal slurry of a fixed concentration is prepared by mixing known
amounts of pulverized coal, water and a dispersant (wetting agent) like
Alconox. The HGMS unit used here is the largest commercial unit now
in use for producing high-quality paper coating clays. It is operated
at a fixed field intensity of 20 kOe generated in an open volume 7 ft
in diameter and 20 in. long. A stainless steel wool separator matrix
having 94% voids is placed in the open volume. The coal slurry is
pumped through the energized separator matrix at a fixed retention
time (flow velocity) until the matrix reaches its loading capacity.
After rinsing with water, the mags are sent to a settling pond or a
clarifier to recover water for re-use. The tails are collected, dewa-
tered, and dried. The typical specifications of process streams are
also included in Figure 2, and the detailed operating conditions for
such a case are given in the Appendix.
Exp&ujne.nta£. Studio oft ?n.o
-------
112 CLEM COMBUSTION OF COAL
adapted in a bench-scale exploratory study to remove sulfur and ash
from a finely pulverized Brazilian coal suspended in water.-'--'- Since
then, other investigators have utilized pilot-scale HGMS units to
desulfurize and deash water slurries of some pulverized Eastern U.S.
coals. For example, results of pilot-scale studies that demonstrated
the technical feasibility of the magnetic separation of sulfur and ash
from water slurries of pulverized Illinois No. 6, Indiana No. 5 and No.
6, Kentucky No. 9/14, and Pennsylvania Upper Freeport and Lower Kittan-
ning coals have been reported.9'10'12'13*15 Depending upon the types
of coal and the separation conditions used, the existing bench- and
pilot-scale results have shown that the use of single-pass HGMS was
effective in reducing the total sulfur by 40-55%, the ash by 30-45%,
and the pyritic sulfur by 75-90%, while achieving a maximum recovery of
the beneficiated coal of about 95%.8>16 These available results have
indicated also that both the grade and recovery of the separation can
be enhanced generally with the use of a larger separator matrix or by
the recycle of the tail products. Although the existing data have not
yet established total deashing by magnetic means, there is some indica-
tion that by enhancing the magnetism of ash-forming minerals and by op-
timizing the separation conditions, etc., the effectiveness of magnetic
separation of ash from coal can be improved. Further discussion of
the reported experimental results of the magnetic beneficiation of
water slurries of pulverized coals can be found in the literature.
Also, the published Proceedings of Magnetic Desulfurization of Coal
Symposium held at Auburn University in March, 1976 can be used for
ready reference on HGMS and its experimental application to coal bene-
ficiation.
Quantitative, ttod&ting and Pizdiction o^ Se.pa/iation PeA^o^once. An
important and latest breakthrough in applying HGMS to the beneficiation
of coal/water slurry is the successful development of the conceptual
understanding and the quantitative model required for predicting the
technical performance of pilot-scale separation. In particular, the use
of an experimentally verified mathematical model developed in the authors'
laboratory ' ^ has now allowed one to quantitatively identify the trade-
off of separation variables so as to optimize the magnetic removal of
sulfur and ash from coal. Also, the model can be used to assess the
technical and economical feasibility of the magnetic beneficiation of
coal without extensive trial-and-error testing. In order to stimulate
a more systematic approach to the future development of HGMS processes
applied to coal beneficiation, some new insights on the quantitative
aspects of magnetic beneficiation of pulverized coal in water are
briefly discussed below.
The technical performance of a HGMS is generally characterized
by the grade and recovery of the separated product, and by the capacity
of the separator. When applying HGMS to beneficiating a fixed amount
of pulverized feed coal of known sulfur and ash contents, the grade
may be represented by the sulfur and ash contents in the beneficiated
coal; while the recovery may be specified by the amount of the benefi-
ciated coal. The capacity of a separator in the magnetic beneficiation
of coal may be characterized by the amount of coal processed from the
start of the separation until the instantaneous sulfur and ash contents
of the beneficiated coal reach a prespecified percentage of the known
sulfur and ash contents of the feed. In particular, the instantaneous
-------
HIGH GRADIENT MAGNETIC SEPARATION 113
change in effluent sulfur and ash contents with time, the so-called
breakthrough curve, can be used to determine the processing time after
which a separator is loaded. Here, a separator is said to be loaded
if the effluent sulfur and ash contents of the coal to be beneficiated
remain practically unchanged.
There are a number of important considerations involved in the
quantitative modeling and prediction of the technical performance of
HGMS applied to coal beneficiation. The first one is related to the
particle trajectory and buildup. Specifically, the technical perform-
ance of an HGMS depends on how efficiently it captures magnetic parti-
cles, and how much of the captured particles can be retained in the
separator matrix. The capture of magnetic particles by HGMS has recent-
ly been studied theoretically by using the equations of motion of
magnetic particles flowing around a single magnetized, ferromagnetic
collecting wire, and the performance of an ideal, unloaded HGMS com-
posed of many such wires in the separator matrix is then related to the
particle trajectories computed from the specified separation condi-
tions. 2»17 Preliminary efforts have also been made to examine both
theoretically and experimentally the effects of particle buildup or
matrix loading on the performance of a nonideal, partially loaded HGMS.
However, none of the existing models based on the recent analyses of
particle trajectory and buildup has been shown to be applicable to
quantitatively predicting the effects of separation variables on the
technical performance observed in pilot-scale experimental studies of
HGMS. The next important consideration is the characteristic of the
feed stream to be magnetically beneficiated. In the literature, most
of the reported modeling and experimental studies of HGMS have been
limited only to the feed streams containing either pure magnetic par-
ticles or simple mixtures of magnetic and nonmagnetic particles of
approximately monodispersed or narrowly distributed sizes. However,
little attention has been devoted to relating the technical perform-
ance of HGMS to the characteristics of the feed stream containing
particles of a wide range of sizes, densities and magnetic susceptibil-
ities as found in the magnetic beneficiation of coal. The third, but
relatively less important, consideration is related to the mechanical
entrapment or filtration of particles at low or zero field intensity.
For separations at high field intensity, however, the effects of me-
chanical entrapment on the technical performance of HGMS are often con-
sidered to be negligible. In the recent work from the authors' labora-
tory, 6,15 the technical performance of HGMS has been quantitatively
examined with reference to the above major considerations. Two practi-
cal observations of importance in the modeling of HGMS applied to
coal beneficiation resulted from this work can be summarized as follows.
(1) The technical performance of HGMS is quantitatively related
to the particle trajectory (or capture) and buildup. Under the pres-
ently used or proposed conditions for wet HGMS processes, however, the
capability of the wire matrix to capture magnetic particles remains
practically unchanged and essentially all magnetic particles are cap-
tured before the matrix is saturated or loaded with the buildup of
particles. As a result, there is practically no need to calculate the
trajectories of particles. The main factor in determining the techni-
cal performance of a pilot-scale or an industrial HGMS appears to be
the particle buildup, but not the particle trajectory (or capture).
(2) Both the particle trajectory (or capture) and buildup are
-------
CLEAN COMBUSTION OF COAL
highly dependent upon the ratio of the so-called magnetic velocity Vm
of the particle to the free stream fluid velocity V^. In particular,
the maximum amount of particle buildup on the wire increases with in-
creasing value of VHI/VK,; and there exists a minimum value of Vm/V^, for
the captured particles to remain sticking to the wire matrix. Here,
the magnetic velocity Vm is essentially a terminal velocity of a par-
ticle in a magnetic field, and it is defined by:
V = 2y° M H° * R2 (1)
m 9n a
Except for the length L and packing void volume e of the separator ma-
trix, the ratio V^/V^ contains almost all the major independent vari-
ables in HGMS, namely: (a) particle properties - radius R and magnetic
susceptibility x> (b) flow field - fluid viscosity n and free stream
fluid velocity V^; (c) magnetic field - field intensity Ho and magne-
tization of wire M; and (d) separator matrix packing characteristics -
wire radius a (and magnetization M) . Under conditions of constant mag-
netic and flow fields as well as fixed separator matrix packing charac-
teristics, the magnetic velocity is mainly a function of particle prop-
erties. In order to quantitatively relate the technical performance
of HGMS to the characteristics of the feed stream containing particles
of a wide range of sizes, densities and magnetic susceptibilities, it
is necessary to determine the magnetic velocities of particles in the
feed stream.
Based on the above findings, a simple particle buildup model in-
corporating the feed characteristics for predicting the technical per-
formance of HGMS under the presently used or proposed conditions for
wet HGMS processes has been developed by first calculating the minimum
magnetic velocity for particle buildup, denoted by Vm m^n The latter
is uniquely determined from the knowledge of the following process
variables in HGMS: (a) (LOAD) - the total weight of captured magnetic
particles per unit cross section of the separator matrix, (b) L - the
length of the separator matrix, (c) F - the packed fraction of the
separator matrix (F = 1-e) , (d) d - the apparent density of the par-
ticle buildup, and (e) V^ - the free stream fluid velocity. The
specific details for calculating VmjIQin from these variables can be
found in the literature "»15. Next, a so-called magnetic velocity
distribution function for characterizing the feed stream, denoted by
F(Vm) , can be introduced. Specifically, F(Vm) represents the cumula-
tive weight fraction of magnetic particles in the feed stream having
a magnetic velocity less than Vm. Thus, if Vm = Vm)min, F(Vm) cor-
responds to the cumulative weight fraction of magnetic particles in
the feed stream which will not be captured inside the separator matrix.
This result suggests that a simple particle buildup model for HGMS can
be written as
COUt _ -C./TT \ TT
•— - Vm =
Here, COut and Cin refer to the concentrations of the specific magnetic
particles of interest, such as pyritic sulfur content in the magnetic
beneficiation of coal, at the outlet and inlet of the separator matrix,
respectively. Furthermore, a simple material balance of particles sug-
gests that (LOAD) can be calculated from the total weight of feed
-------
HIGH GRADIENT MAGNETIC SEPARATION 115
particles per unit cross section of the. separator matrix, (FEED), by
integrating the following equation
d(LOAD)
(4)
The magnetic velocity distribution, F(Vm) is to be quantitatively de-
termined by the experimental characterization of the feed stream. In
the magnetic beneficiation of coal, for example, such a characteriza-
tion requires the determination of particle size distribution, a stan-
dard float-and-sink separation, and measurements of the magnetic sus-
ceptibility along with the ash, sulfur and pyritic sulfur contents of
each separated fraction. Specifically, suppose that x(W) and Ai(W)
(i=l,2, and 3) are the magnetic susceptibility, and the ash, sulfur,
and pyritic sulfur contents, respectively, of the pulverized feed coal
at a cumulative weight percent W; and U(R) is the cumulative weight
percent of particles having a radius smaller than R. Under conditions
of constant magnetic and flow fields as well as fixed separator matrix
packing characteristics, R can be related to the magnetic velocity of
the feed stream Vm according to (1). Thus, in order for particles to
have a magnetic velocity less than Vm, the radius of particles with a
magnetic susceptibility X should be smaller than
9 Vm n a h
R = ( - )
2 y0MHo*
The magnetic velocity distribution of the feed stream, F(V ) , can then
be obtained by
100 l
F(Vm) = 1 { / U{(9VmTiav )*} dW + Wo} (5)
100 ^o 2 y0 M H0X u '
Here, W is the weight percent of particles in the feed stream having
negative magnetic susceptibilities. Note that implicitly included in
(5) is the practical approximation that except by mechanical entrapment
or filtration, no diamagnetic particles will be attracted magnetically
by the wire matrix. Based on (5) and the measurements of ash, sulfur
and pyritic sulfur contents of the pulverized feed coal, the specific
magnetic velocity distributions of ash, sulfur and pyritic sulfur
(i=l,2, and 3, respectively), denoted by F.j_(Vm) , can be approximated
by:
, 100 A,(w) 9 Vm 1 a *
F,(Vm) = J_ { / ( lV ) U{ (— - ) } dW + A. W }
T± W0 100 2 y0 MH0 X(W) ±o °
(6)
In (6), T± (i=l,2, and 3) are the average ash, sulfur and pyritic sul-
fur contents, respectively, of the pulverized feed coal; and Aio
(i=l,2, and 3) are the average ash, sulfur and pyritic sulfur contents,
respectively, of the fraction WQ with negative magnetic susceptibili-
ties.
(1) - (6) represent the new mathematical model developed in the
authors' laboratory for predicting the technical performance of HGMS
-------
116
CLEAN COMBUSTION OF COAL
applied to coal beneficiation &>1^. The model has been successfully
applied to predict satisfactorily the grade, recovery, concentration
breakthrough, etc., observed in pilot-scale experimental studies of
HGMS applied to the beneficiation of water slurries of several pul-
verized coals such as Illinois No. 6, and Pennsylvania Upper Freeport
and Lower Kittanning coals, etc. For example, Figure 3 illustrates
the magnetic velocity distributions for coal, ash, sulfur and pyritic
sulfur for characterizing the pulverized Illinois No. 6 coal under
conditions of HQ = 107/2u ampere-turn/meter (20 kOe), M = 45 x 106 am-
pere-turn/meter (1.4 kOe) , y0 = 4ir x 107 henry/meter, a = 45 x 10~6 m
and n = 10~3 kg/m-sec. These conditions correspond to the typical
separation conditions for the experimental results reported in Ref. 9.
Figure 3. Magnetic Velocity Distributions of
Pulverized Illinois No. 6 Coal
Figure 4 shows the typical comparison between the theoretical
and experimental concentration ratios, cou^/^±n> °f pulverized Illinois
No. 6 coal particles in tail products obtained at a field intensity
HQ = 20 kOe, a superficial slurry velocity Vo = 1.86 cm/sec, and dif-
ferent total feed rates per unit cross section of a pilot-scale HGMS.
The comparison indicates that concentration ratios predicted by
the model are in a reasonable agreement with experimental results. In
Figure 5, a typical comparison between theoretical and experimental
concentration breakthrough curves is illustrated. The curves in the
figure are obtained by numerically integrating (3) based on the same
feed characteristics and separation conditions as in Figure 4. It can
be seen that there is also a reasonable agreement between theoretical
and experimental concentration breakthrough curves. Note that there
is no free parameter included in the new model given by (1) to (6) .
Also, the distributions of magnetic susceptibility, particle size and
density along with ash, sulfur and pyritic sulfur contents of each sep-
arated fraction of the pulverized feed coal after the standard float-
and-sink separation can all be easily measured experimentally prior to
-------
HIGH GRADIENT MAGNETIC SEPARATION
117
100
Model Prediction:
Experimental Data: O
VQ 1.86 cm/sec
H0 20 We
I 1 i i I I 1 1
0 40 80 120 160
20
TOTAL FEED ( g coal/cm2 )
Figure 4. Theoretical and Experimental Concentrations of Coal
Particles in Tails as a Function of Total Feed
Model Prediction:
Experimental Data: ago
VB 1.86 on/sec
FEED TOTAL SULFUR
PRODUCT TOTAL SULFUR
40 BO 120
TOTAL FEED ( g coal/cm2 )
Figure 5. Theoretical and Experimental Ash and Sulfur
Concentrations of Coal in Tails as a Funtion
of Total Feed
actual magnetic separation testing. Thus, based on the above compar-
ison between model predictions and experimental results for Illinois
No. 6 coal, and the successful application of the model to several other
Eastern U.S. coals described elsewhere6»ij>li), it can be suggested
that thesie. U now an e.x.peAvne-ntMy veAlfced, &
-------
118 CLEAN COMBUSTION OF COAL
developed, it is possible to quantitatively determine the optimum sep-
aration conditions so as to optimize the magnetic removal of sulfur and
ash, while achieving an economically acceptable recovery of the bene-
ficiated coal. The new model can also be used to provide the needed
information such as the proper separation duty cycle under the selected
process conditions for the engineering design and cost estimation of
wet HGMS processes applied to coal beneficiation . Further discussion
on the practical applications of the modeling results can be found in
the literature^.
P-toce/6.6 PotuntiaJt and EconomJicA . By removing 75-90% of the
pyritic sulfur magnetically and achieving 85-90% recovery as was demon-
strated by the results of reported studies of magnetic beneficiation
of pulverized coals in water slurries 8-15 the process illustrated in
Figure 1 may be used to clean about one-fifth of the recoverable U.S.
coals with a low organic sulfur content of 0.7-0.9 wt % to produce an
environmentally acceptable fuel. A detailed documentation of the re-
serve and production of U.S. coals which may be magnetically cleanable
to 1 wt% total sulfur according to the seam, district, and county in
each state, along with the total and organic sulfur contents can be
found in the literature*°. A technical survey of such magnetically
cleanable coals from Pennsylvania is currently being conducted in the
authors' group in cooperation with the Coal Preparation and Analysis
Laboratory of the U.S. Bureau of Mines.
Here, a reasonable range of add-on costs (excluding those for
grinding, dewatering, drying and refuse disposal) can be estimated for
the magnetic beneficiation of water slurries of pulverized coals which
have desulfurization characteristics similar to those reported in the
recent studies °~l->. -phe method used to estimate the costs of magnetic
desulfurization is based on the technique used by the Federal Power
Commission Synthetic Gas-Coal Force to estimate the cost of synthesis
gas*"' . The investor capitalization method used in this approach is
the discount cash flow (DCF) financing method with assumed DCF rates of
return such as 15% after taxes. This method essentially determines the
annual revenue during the plant life which will generate a DCF equal to
the total capital investment for the plant. Several assumptions are
included in the methodl9,20:
(1) The plant life is assumed to be 20 years with no cash value
at the end of life.
(2) A straight-line method is used to calculate the annual depre-
ciation.
(3) Operating costs and working capital requirements are assumed
to be constant during the construction period, and 100% equi-
ty capital is assumed.
(4) Total plant investment, return on investment during the plant
life, and working capital are treated as capital costs in
year zero (the year ending with the completion of start-up
operations).
(5) Start-up costs are treated as an expense in year zero.
(6) 48% federal income tax is assumed.
Based on these assumptions, equations for calculating the unit costs
($ per ton of coal processed) are suggested by the published docu-
ments-^ »20. They are summarized in the Appendix, in which the de-
tailed operating conditions and estimated costs for a typical example
shown in Figure 1 are also given. The costs of major installed equip-
ment and the unit costs listed in the Appendix are based on values of
-------
HIGH GRADIENT MAGNETIC SEPARATION 119
J-jne 1976. For instance, the costs of pump and tank used were estima^
ted first according to Ref . 21 and then brought up to date by multiply-
ing by the CE plant cost index ratio of (205/113.6), while the cost of
the installed HGMS unit with a separator matrix of 7 ft in diameter
and 20 in. long was estimated to be 1.936 million23.
The estimated capital investment and unit costs for four typical
cases, designated as A-D, are summarized in Table I. Slurry veloci-
ties of 2.61 and 4.0 cm/sec, slurry concentrations of 15, 25 and 35
wt%, as well as separation duty cycles of 59.0-77.9% are considered.
These separation conditions illustrate clearly the effects of slurry
velocity and concentration as well as separation duty cycle. For in-
stance, comparison of cases A to C shows that at the same slurry veloc-
ity and similar magnetic desulfurization characteristics, the higher
the slurry concentration, the cheaper will be the investment and unit
costs. While this observation is to be expected, there have been pilot-
scale tests which indicate that increasing the slurry concentration of
pulverized Illinois No. 6 coal, from 2.57 to 28.4 wt% did not appreci-
ably change the grade and recovery of the separation9. Further effects
of processing conditions, and other cost factors on the unit costs, are
illustrated in Table II. It is seen from the table that by doubling
the amount of coal processed per cycle relative to a fixed amount of
stainless steel wool packed in the separator matrix, the unit cost can
be reduced by about 15%. This result shows the importance of the sep-
arator matrix loading characteristics on the costs of magnetic desul-
furization. Another factor which affects the unit costs considerably
is the washing time required in a complete separation cycle. This is
seen by comparing items 4 and 6 in Table II. In particular, the com-
puted results indicated that doubling the amount of wash water required
only leads to a negligible increase (0.27 - 0.60%) in unit costs.
However, if both the amount of wash water and the washing time are
doubled, the unit costs are increased by about 15%. The above obser-
vations clearly suggest an important economic incentive for further
pilot-scale investigation of the separator matrix loading and washing
characteristics in the magnetic beneficiation of coal/water slurry.
Finally, item 7 of Table II shows that labor cost seems to be a signif-
icant fraction of the unit cost. Fortunately, it is not expected that
the labor requirement will be doubled in actual commercial practice
from the nominal case in Table I. This follows because existing ex-
perience in the commercial cleaning of kaolin clays by HGMS indicates
that the labor requirements for both operation and maintenance are
A comparison of approximate estimated capital and unit costs of
different pyritic sulfur removal processes currently under develop-
ment30'31 is given in Table III. With the exception of the MAGNEX pro-
cess,30 all the processes listed in Table III are wet cleaning methods
and require grinding of feed coal, thus requiring relatively comparable
grinding, dewatering, and drying costs. This table indicates that the
costs of coal beneficiation by HGMS are attractive when compared with
those of other approaches, even after adding the necessary costs of
grinding, dewatering and drying. However, the above comparison is only
an approximate one, because of the difference in the methods used to
estimate the costs and in the desulfurization characteristics reported,
etc. Based on the available cost information for these pyritic sulfur
removal processes, it is not yet possible to carry out a rigorous com-
parison.
-------
120
CLEAN COMBUSTION OF COAL
Table I. Cost of Desulfurization of Coal/Water Slurry
by HGMS Using Separator Matrix of 7-Ft
Diameter and 20-In Length
Case A
Case B
Case C Case D
1.
2.
3.
4.
5.
6.
7.
Slurry velocity (cm/sec)
Slurry concentration (wt %)
Coal feed rate (ton/hr)
Cycle time (min)
Duty cycle (%)
Tons of coal processed
per cycle
Unit costs ($ per ton
coal processed)
2.61
15
44.77
9.00
77.9
6.7
2.61
25
66.13
6.10
67.4
6.7
2.61
35
83.07
4.85
59.0
6.7
4.0
25
89.61
4.50
59.6
6.7
Mo
8.
Capital investment per
ton coal processed
per year ($)
2.083 2.369
1.802 1.063
3.676 2.412
6.93
4.69
1.109 1.067
0.858 0.829
1.967 1.880
3.73
3.53
Table II. Sensitivity Analysis of Unit Costs ($ Per Ton Coal Processed)
of Desulfurization of Coal/Water Slurry by HGMS
1. Basis: 2.61 cm/sec, 25 wt %
slurry, and other conditions 1.0628 0.00 2.4117 0.00
in the Appendix and Table I
2. Amount of coal processed
per cycle doubled 0.9004 -15.23 2.0341 -15.66
3. 25% reduction in capital
investment 0.9389 -11.28 2.0341 -19.12
4. Amount of wash water
required doubled 1.0691 +0.60 2.4181 +0.27
(washing time unchanged)
5. Coast of water increased
5/3 times (5C/1000 gal) 1.0835 +1.95 2.4324 +0.86
6. Both amounts of wash
water and washing time 1.2256 +15.32 2.7883 +15.62
doubled
7. Labor requirement doubled 1.3587 +27.82 2.7077 +19.12
-------
HIGH GRADIENT MAGNETIC SEPARATION 121
Table III. Comparison of Estimated Approximate Capital and
Unit Costs of Different Pyrite Removal Processes
Process U a * U * Capital Investmentb
1.
2.
3.
4.
MAGNEX-Hazen ..
£Q
Research, Inc.
Froth flotation-
Bureau of Mines 24
Meyers-TRW
Systems and Energy 25
Ledgemont oxygen
leaching-Kennecott
5.83 7.05
2.77 4.47
6.0-14.0
comparable
to Meyers
4.17
5.71
13.80
(leaching only)
11.30
(leaching only)
Copper Corp.2
5. HGMS-This work, 0.83-1.06 1.88-2.41 3.53-4.69
see Table I
°- Unit costs expressed in $ per ton coal processed.
" Capital investment expressed in $ per ton coal processed per year.
MAGNETIC BENEFICIATION OF DRY PULVERIZED COAL
The application of HGMS to remove sulfur and ash from dry pulver-
ized coal has just been initiated recently. Much of the work reported
thus far has been limited to investigating the technical feasibility
of magnetic beneficiation of dry pulverized coal via either gravity
feeding or air-entrained fluidization. Although some degree of mag-
netic removal of sulfur and ash was observed in using these approaches,
the experimental results reported had not yet achieved separation
performance comparable to those observed in magnetic beneficiation of
coal/water slurry. In fact, the weight percent of sulfur removed _from
dry pulverized coal was rarely more than 10 to 20% even with multiple-
pass separation. There was also no definitive indication in the
existing reports10'12 as to why dry beneficiation with either gravity
feeding or air-entrained fluidization was less effective compared to
wet separation. Quite recently, a joint experimental program on the
high gradient magnetic beneficiation of dry pulverized coal has been
initiated at Auburn University and Oak Ridge National Laboratory. Its
objectives are to develop a novel recirculating air-fluidized separa-
tor matrix for use in HGMS applied to dry pulverized coal, and to com-
pare the performance of both dry and wet HGMS processes for coal bene-
studies are briefly discussed below. '
-------
122 CLEAN COMBUSTION OF COAL
Fe.zdU.ng Approach., The simplest way for feeding dry pul-
verized coal to a magnetic separator is by gravity. Preliminary re-
sults on the desulfurization of dry pulverized Indiana Nos, 5 and 6
coals by HGMS via gravity feeding reported in 1976 are summarized in
Table IV.
Table IV. Magnetic Desulfurization of Dry and Wet Pulverized
Indiana Coals via Gravity Feeding Approach^
Sulfur Removal, % Sulfur Removal, %
No. 5 Coal Ho. 6. Coal
Drya Wet
1
2
3
a
b
. l-Passc
. 2-Pass<1
. 3-PassC
99% below 200 mesh,
99% below 200 mesh,
void, 30 wt% solids
9.
17.
18.
3
9
4
20 kOe,
20 kOe,
slurry.
18
33
35
.8
.0
.2
Frantz screens
stainless steel
13
18
20
wool
.7
.5
.6
packing ,
24
42
44
94%
.5
.4
.8
retention time per pass: 30 seconds
Also included in the table are their comparison with the results ob-
tained by wet magnetic beneficiation.12 There was no explanation given
in Ref. 12 as to why wet beneficiation was more effective in sulfur re-
moval compared to dry separation with gravity feeding as shown in the
table. Some insights to answering the latter question can be provided
by the recent experimental results obtained in the authors' labora-
tory.12 Thus, by studying the magnetic desulfurization characteristics
of dry pulverized coal of different particle size ranges, it was found
that the use of feed coal of small particle sizes would often lead to
the agglomeration of the pulverized coal and its mineral impurities. In
order to minimize the occurrence of fine coal agglomeration which
could impede magnetic beneficiation, it would not be desirable to use
the feed coal ground to, for example, less than 100 to 200 mesh. This
point may be illustrated by the experimental results given in Table V,
in which the effect of particle size on magnetic desulfurization of
Pennsylvania Lower Kittanning coal via gravity feeding is shown.^
Essentially no sulfur was removed from the pulverized Lower Kittanning
coal of particle sizes between 100 and 200 mesh; and actual agglomer-
ation of coal and its mineral impurities was visually observed from the
separated samples. In contrast, over 15% of the sulfur could be
separated magnetically from the coal of particle sizes between 60 and
100 mesh. A major problem in using a moderately ground coal in dry
separation, however, is that its sulfur-bearing and ash-forming miner-
als may not be sufficiently liberated; and there is a trade-off be-
tween avoiding fine coal agglomeration and grinding to liberate the
mineral impurities in coal. In order to effectively apply HGMS for
beneficiation of dry pulverized coal, it is necessary to closely exam-
ine the proper grinding level for a specific coal chosen at given
separation conditions, while taking into account the desired libera-
tion characteristics of its sulfur-bearing and ash-forming minerals.
-------
HIGH GRADIENT MAGNETIC SEPARATION 123
Table V. Effect of Particle Size (Fine Coal Agglomeration) on Magnetic
Desulfurization of Dry Pulverized Lower Kittanning Coal,
Jefferson County, Pennsylvania via Gravity Feeding
Approach: 100-gm Feed Coal, 5-minute Feeding Time
Case A Case B Case C
1. Particle size, mesh -60 + 100a -100 + 200fa
2. Field intensity, kOe 10 10
-60 + 100a
20
-100 + 200b
20
3. Mags
weight, gm
sulfur, wt%
4. Sulfur removal, %
5.3
10.54
8.89
7.4
6.34
3.70
8.14
11.42
16.05
9.24
8.43
3.90
a Feed coal contains 1-2% moisture, 4.05% total sulfur, 3.73% pyritic sulfur and 18.8%
ash.
° Feed coal contains 1% moisture, 4.335% total sulfur, 3.354% pyritic sulfur and 13.70%
The second problem which seems to contribute to the reported poor
performance of magnetic beneficiation of dry pulverized coal is the
limited separator matrix loading capacity in using such packing mate-
rials as stainless steel screen, expanded metal, etc. Because of the
possible excessive pressure drop, the low viscosity of the carrier gas,
etc. in dry separation, the latter packing materials are generally more
appropriate compared to stainless steel wool. Unfortunately, the in-
sufficient sharp edges and surface areas associated with these packing
materials limit their capabilities for producing large field gradient
and for capturing magnetic particles. For example, an approximate in-
dication of such a limited matrix loading capacity in dry separation
with stainless steel screens is illustrated in Table VI.
Tatle VI. Illustration of the Limited Matrix Loading Capacity
in Dry Magnetic Separation with Stainless Steel
Screens13
Pyrite , -60 + 200 mesh, 24 gm feed, 50 screens, gravity feeding
1.
2.
Field intensity,
kOe
Mags
weight , gm
wt%
0 5 15 20
3.8 6.3 7.0 8.6
15.7 26.3 29.3 35.7
For the experiments shown in the table, fifty Frantz screens of 1-in. I.D.
and 0.3125-in. thickness constructed from stainless steel strips were
stacked together along with forty-nine spacers of 0.1875-in. thickness
as separator matrix packing materials. By calculating the active
surface area of a screen and by assuming a monolayer particle capture
and buildup on both sides of a screen, the maximum amount of spherical
pyrite particles which can be collected as mags can be found to be
about 8.0 gm for 60-mesh particles, 5.68 gm for 100-mesh particles,
and 2.80 gm for 200-mesh particles. These estimated amounts are fairly
consistent with those collected experimentally as shown in the table.
-------
124 CLEAN COMBUSTION OF COAL
Despite the problems of fine coal agglomeration and limited sep-
arator matrix loading capacity in dry separation via gravity feeding as
as illustrated in the above, two important findings resulted from
the recent work with Pennsylvania Lower Kittanning and Upper Freeport
coals in the authors' laboratoryM First, it was found that the coal
feeding rate had essentially no effect on the performance of magnetic
beneficiation of dry pulverized coal via gravity feeding. Secondly,
under comparable experimental conditions, the sulfur and ash contents
of the mags resulted from dry separation via gravity feeding were found
to be almost the same as those obtained from wet beneficiation. This
point may be illustrated by the approximate comparison between dry and
wet magnetic beneficiation of Lower Kittanning coal shown in Table Vli.
Table VII. An Approximate Comparison Between Dry and Wet Magneti
Beneficiation of Lower Kittanning Coal, Jefferson
County, Pennsylvania^
p
Grams Ash (Z)
6 g
Sulfur (%)
—*~
Pyrite (%)
1. Dry, gravity feeding, 100 9.40 13.07 30.12 4.49 12.96 2.19 10.23
5 minute feeding time,
-100 + 200 mesh, stainless
steel screens, 19 kOe
2. Dry, air-entrained 100 5.29 14.68 30.60 4.42 12.96 2.59 12.92
f luidization, upward flow,
28.3 cm/sec, stainless steel
screens, 19 kOe
3. Wet, 2.5 vtZ solids slurry, 100 18.02 10.66 29.28 3.93 11.22 2.52 9.11
2.3 cm/ sec, stainless steel
wool packing, -200 + 325
raesh, 19 kOe
It can be seen from item nos. 1 and 3 of the table that the weight per-
cents of ash, sulfur and pyrite in the mags resulted from dry separation
via gravity feeding and wet beneficiation are of the same order of
magnitude, having the ranges of 29.28-30.60, 11.22-12.96, and 9.11-
12.92%, respectively. The only difference between dry and wet benefi-
ciation is the amount of mags collected due to their different separator
matrix loading capacities. This difference has thus led to the generally
poor performance of dry separation via gravity feeding compared to wet
beneficiation based on the ash, sulfur and pyrite contents of the sep-
arated product coal. Further implications of this difference between
dry and wet beneficiation are discussed below along with some results
obtained on the magnetic beneficiation of dry pulverized coal in a
fluidized-bed separator matrix.
xL-. ftuAdization Approach. Prior to the work in the au-
thors' laboratory}3 there was only one reported study on the magnetic
beneficiation of dry pulverized coal via an air-entrained fluidization
approach.10 In that study, a screw or vibratory feeder was used to inject
dry pulverized Delmont coal from Westmoreland County, Pennsylvania, into
an air stream which would carry it doWW)OJid through the separator matrix
of 1-in. I.D. and 6 in. long. The preliminary tests with a screw feeder
and an expanded metal matrix showed that the presence of fines in the
typically 60-mesh feed coal would make the feeder inoperative; and no
significant reduction in the sulfur and ash of the product coal was ob-
served. Subsequent tests conducted using a vibratory feeder with a
field intensity up to 50 kOe and air velocities of 59-1019 cm/sec indi-
cated that the weight percent of sulfur removed from the feed coal was
very small and irregular, rarely more than 10%. In contrast, experiments
-------
HIGH GRADIENT MAGNETIC SEPARATION 125
carried out with the same pulverized coal in a water slurry showed that
HGMS was effective in reducing the total sulfur of the feed coal by 50%,
the ash by 50%, and the pyrite by 60-80%. No explanation regarding this
significant difference in the performance of dry and wet beneficiation,
however, was given in Ref. 10.
An extensive, pilot-scale experimental study of magnetic beneficia-
tion of dry pulverized Pennsylvania Lower Kittanning and Upper Freeport
coals via an air-entrained fluidization approach was conducted recently
in the authors' laboratory, and the results have been described else-
where. One major difference of this study compared to that described
in Ref. 10 was that an upuXVid. air-entrained fluidization of the pulver-
ized feed coal through a separator matrix of 5-in. I.D. and 20 in. long
was examined. The experiments revealed the same problems of fine coal
agglomeration and limited separator matrix loading capacity as observed
in dry separation via the gravity feeding approach. Further, it was
found that under comparable experimental conditions, the sulfur and ash
contents of the mags resulted from air-entrained fluidization were of
the same orders of magnitude as those found from dry separation via
gravity feeding or wet beneficiation. This point has already been il-
lustrated earlier in Table VII as item no. 2. In addition, the exper-
iments indicated that due to the high velocity of the air stream re-
quired to achieve a good entrained fluidization of the pulverized feed
coal, the retention time of magnetic particles in coal in the separator
matrix was generally very short. As a result, the efficiency of the sep-
arator matrix in capturing and retaining magnetic particles in coal was
very low. For example, item no. 2 in Table VII shows that for 100 gm of
pulverized Lower Kittanning coal entrained in an air stream of velocity
28.3 cm/sec, the amount of mags collected in the separator matrix was
only 5.29 gm. The latter was also less than that collected in dry sepa-
ration via gravity feeding or wet beneficiation as shown in the table.
Recx>LCu£a£oi<2 AsiA-P&LU.dU.za£ion Approach.. Based on the preceding
discussion of experimental results and their implications, it is evi-
dent that an effective HGMS process to be developed for the beneficia-
tion of dry pulverized coal in a fluidized-bed separator matrix must
include at least two desirable features. Thus, it must have a simple
means to reduce the presence of fines in the fluidized coal stream
and to avoid their possible agglomeration. It must also provide a su-
ficient retention time for the fluidized coal stream to promote the
contact between the magnetic particles in coal and the active surface
areas available on the separator matrix packing materials. The latter
is of importance in increasing the capacity of the separator matrix for
capturing and retaining magnetic particles in coal. As a result of a
joint research program initiated at Auburn University and Oak Ridge
National Laboratory on magnetic beneficiation of dry pulverized coal,
a recirculating air-fluidization approach possessing the above desir-
able features has been developed recently; and several novel fluidized-
bed separator matrices for use in HGMS applied to dry pulverized coal
are currently being tested experimentally.^ In fact, by using one of
the recirculating fluidized-bed separator matrices developed, the per-
formance of magnetic separation of sulfur and ash from dry pulverized
Pennsylvania Upper Freeport coal was repeatedly found to be be^tte/i than
that from coal/water slurry. This separator matrix was made of three
primary sections of increasing inside diameters arranged from the
-------
126
CLEAN COMBUSTION OP COAL
bottom to the top of the matrix. These primary sections are of 5-in. I.D.
and 5 in. long (section A), 3.5-in. I.D. and 10 in. long (section B) ,
and 0.75-in. I.D. and 7 in. long (section C) , respectively. Connecting
sections A and B, and sections B and C are two expanded sections of^3 in.
and 4 in. long, respectively. In the experiments conducted with this
matrix, packing materials such as stainless steel screens were placed in
section B, and the pulverized feed coal placed in an auxiliary fluidized
bed outside of the separator was air-fluidized through the separator.
By properly controlling the flow velocity of the air stream during the
whole separation period, it was possible to collect most of the fines
in the pulverized coal stream as top tail product of low sulfur and ash
contents. At the same time, because of the unique expanded sections from
the bottom to the top of the separator matrix and the resulting gradual
decrease in the upward fluidization velocity of the pulverized coal
stream, the majority of the pulverized coal particles of medium and large
sizes would tend to recirculate inside the expanded sections. As a re-
sult, a sufficient retention time inside the separator matrix could be
provided to the bulk of the fluidized coal stream without the presence
of fines, thus allowing the magnetic particles in coal to be captured
and retained by the matrix. Toward the end of the desired separation
period, the flow velocity of the air stream was reduced, and the magnet-
ically beneficiated coal of low sulfur and ash content was collected as
bottom tail products.
Table VIII shows the typical experimental results obtained with
Pennsylvania Upper Freeport coal of particle sizes between 100 and 200
mesh.
Table VIII. Magnetic Beneficiation of Dry Pulverized Upper Feeport
Coal, Jefferson County, Pennsylvania via Recirculating
Air Fluidization: Air Velocity - 17.7 cm/sec13
Grams
Sulfur (%)
Pyrite (%)
Ash (%)
1. Feed 100.00 2.123
2. Mags
(20 kOe)
1-pass 9.05 12.80
2-pass 5.99 4.68
3-pass 5.12 2.25
total 20.20 7.71
3. Total tails
(20 kOe) 79.80 0.68
4. Feed separated
as mags 20.20 68.16 a
5. Mags
(0 kOe) 5.03 1.35
1.519
11.79
4.18
1.93
7.02
0.130
86.80a
0.85
6.320
24.87
13.26
10.00
17.65
3.45
52.43a
4.66
a Weight percent of sulfur, pyrite or ash separated from feed as mags.
It can be seen that by using the recirculating air-fluidization approach
and three-pass separation, HGMS was able to reduce the total sulfur of
the Upper Freeport coal by 68.16%, the pyrite by 86.8%, and the ash by
-------
HIGH GRADIENT MAGNETIC SEPARATION 127
52.43%. Also, the total sulfur content of the tail product was suffi-
ciently low (0.68 wt%) that the magnetically beneficiated Upper Free-
port coal could be used immediately as an environmentally acceptable,
low sulfur fuel.
The experimental results illustrated in Table VIII have clearly
suggested that HGMS with recirculating air-fluidization appears to hold
much promise as an effective physical method for cleaning coal. In par-
ticular, these results along with those obtained from wet beneficiation
described elsewhere13,15 have shown that the performance of magnetic
separation of sulfur and ash from dry pulverized coal via recirculating
air-fluidization can be even better than that from coal/water slurry.
Obviously, further research and development work related to the quanti-
tative modeling and prediction of separation performance, equipment and
process design, etc., in the magnetic beneficiation of dry pulverized
is justified.2'
APPENDIX: BASIS FOR ESTIMATING THE UNIT COSTS OF MAGNETIC BENEFICIA-
TION OF COAL/WATER SLURRY
The detailed operating conditions and estimated unit costs for a
typical conceptual process for the magnetic beneficiation of coal/
water slurry (case B in Table I) are illustrated as follows.
Operating Conditions
(1) Concentration of coal/water slurry = 25 wt%
(2) Superficial flow velocity = 2.61 cm/sec
(3) Stainless steel wool packing density = 6 wt%
(4) Amount of coal processed per cycle = 7 times weight of stainless
steel wool
(5) Amount of washing water required per cycle = 3 times volume of
separator matrix
(6) Amount of rinse water required per cycle =1.5 times volume of
separator matrix
(7) Flow velocity of rinse water = flow velocity of coal slurry
(8) Washing time per cycle = 1 min
(9) Time of energizing the magnet per cycle =0.5 min
(10) Labor required = 2 men per shift
(11) Amount of dispersant = 10 ppm
Investment Costs
(1) Costs of major installed equipment ($)
one HGMS unit $1,936,000
pump 38,480
tank 24.370
1,998,850
(2) Add 20% Contingency 399.770
Total Investment (!_•$) $2,398,670
Operating Costs ($ per year)
(1) Dispersant (57c/lb) $ 24,120
(2) Electric power (2c/KW, 650KW) 102,960
-------
128 CLEAN COMBUSTION OF COAL
(3) Water (30/1000 gal) $ 16,440
(4) Operating labor
(2 men/shift X 8304 man-hr/yr X 6.5$/man-hr) 107,960
(5) Maintenance labor (1.5% of operating investment cost) 35,980
(6) Supervision (15% of operating and maintenance labor
costs) 21,590
(7) Operating supplied (30% of operating labor costs) 32,390
(8) Maintenance supplied (1.5% of investment cost) 35,980
(9) Local taxes and insurance (2.7% of investment cost) 64.760
Annual Net Operating Cost N $ $ 442,180
Coal Processed Annually Q tons 528,900
Unit Costs ($ per ton coal processed)
See Ref. 19 and 20 for the cost equations used below.
(1) Based on 0% DCF rate of return:
TJ0 = (N + 0.05 JO/G = 1.063 $/ton
(2) Based on 15% DCF rate of return:
U15 = (N + 0.34749 I)/G = 2.412 $/ton
(3) Based on capital amortization over 20 years at 1Q% interest
rate:
U =
-------
HIGH GRADIENT MAGNETIC SEPARATION 129
11. Trindade, S.C., "Studies on the Magnetic Demineralization of
Coal", Ph.D. Thesis, Department of Chemical Engineering, Massa-
chusetts Institute of Technology, Cambridge, MA (1973).
12. Murray, H.H., "High Intensity Magnetic Cleaning of Bituminous
Coals", National Coal Association and Bituminous Coal Research,
Inc., Coal Conference and Expo III, Lexington, KY (1976).
13. Liu, Y.A., "A Feasibilty Study of High Gradient Magnetic Benefi-
ciation of Coal in a Fluidized Bed", Progress Report, issued by
Auburn University to Oak Ridge National Laboratory, Energy Re-
search and Development Administration, under contract no.
W-7450-eng-26 ORNL/sub-7315, September (1977).
14. Lin, C.J., and Liu, Y.A. , "Desulfurization of Coals by High-Inten-
sity, High-Gradient Magnetic Separation: Conceptual Process De-
sign and Cost Estimation", Paper presented at ACS National Meet-
ing, New Orleans, LA, March, 1977; accepted for publication in
ACS Symp. Ser., "Coal Desulfurization", December (1977).
15. Liu, Y.A., Oak, M.J., and Lin, C.J., "Modeling and Experimental
Study of High Gradient Magnetic Separation Applied to Coal Bene-
ficiation", Symposium on novel separation technique, AIChE Annual
Meeting, New York, NY, November (1977).
16. Liu, Y.A., Editor, "Proceedings of Magnetic Desulfurization of
Coal Symposium: A Symposium on the Theory and Applications of
Magnetic Separation", IEEE Trans. Magn. MAG-12 (5), 423-551
(1977).
17. Watson, J.H.P., "Magnetic Filtration", J. App. Phys., 44. 4209
(1973).
18. Hoffman, L., "The Physical Desulfurization of Coal: Major Con-
sideration of S02 Emission Control", Special Report, Mitre Corp.,
McLean, VA, November (1970).
19. Federal Power Commission, "Final Report: The Supply-Technical
Advisory Task Force on Synthetic Gas-Coal", April (1973).
20. Batchelor,J.D., and Shih, C., "Solid-Liquid Separation in'Coal
Liquefaction Processes", AIChE National Meeting, Los Angeles, CA,
November (1975).
21. Guthrie, D.M., "Capital Cost Estimating", Chem. Eng., 76. 114-
142, March 24 (1969).
22. lannicelli, J., "Assessment of High Extraction Magnetic Filtra-
tion", Special Report to the National Science Foundation, avail-
able as document No. Pb240-880/5 from the National Technical
Information Service, Springfield, VA (1976)
23. lannicelli, J., personal communication, Aquafine Corp., Bruns-
wick, GA, October (1976).
24. Kindig, J.K., Turner, R.L., "Dry Chemical Process to Magnetize
Pyrite and Ash for Removal from Coal," Preprint No. 76-F-366,
SME-AIME Fall Meeting, Denver, September (1976).
25. Van Nice, L.J., Santy, M.J., Meyers, R.A., "Meyers Process:
Plant Design, Economics and Energy Balance", National Coal
Association and Bituminuous Coal Research, Inc., Coal Conference
and Expo, III, Lexington, KY, October (1976).
26. Agarwal, J.C., Gilberti, R.A., Irminger, P.F., Sareen, S.S.,
"Chemical Desulfurization of Coal", Min. Congr. J., 70 (3),
40-43 (1975).
27. Liu, Y.A., and Lin, C.J., "Research Needs and Opportunities in
High Gradient Magnetic Separation of Particulate-Gas Systems",
-------
13° CLEAN COMBUSTION OF COAL
invited paper, to appear in Proceedings of National Science
Foundation-Environmental Protection Agency Research Workshop on
Novel Concepts, Methods and Advanced Technology in Particulate-
Gas Separation, University of Notre Dame Press, Notre Dame,
IN (1977).
-------
131
A THEORETICAL APPROACH TO WASHABILITY CURVES
IS COMPARED TO THE OTISCA PROCESS
SEPARATION OF PINE COAL
by
D. V. Keller, Jr.
Otisca Industries, Ltd.
(On Leave Prom Syracuse University)
ABSTRACT
An approach to the theoretical washability curve was developed
based on the variables of raw coal size distribution, ash and pyritic
sulfur concentration. The size distribution of iron pyrite and mineral
matter in that coal seam is also required. The analytical treatment
was compared to the washability curves developed independently in an-
other laboratory on a pulverized coal sample and on the data developed
from the separation of the same coal by the Otisca Process. Although
numerous assumptions were required for a complete solution at various
gravities, the theoretical curve for pyritic sulfur was positioned
about 10 percent higher than the observed washability or Otisca Process
data.
INTRODUCTION
Coal preparation technologists have long recognized that their
immediate task is to desulfurize and deash coal with a maximum yield of
coal product; and this task is complicated by extreme variations in
coal seam chemistry and morphology which include the coal itself as well
as iron pyrites, mineral matter, and moisture. Furthermore, they also
face severe variations in the coal preparation procedure such as the de-
gree and techniques of size reduction, separation specific gravity and
procedure, and other significant factors. Some generalities have
emerged from this complexity over the years. For example, it is gener-
ally recognized that increasing size reduction permits the release of
more mineral matter and pyritic sulfur for separation. Verification of\
this is made evident in numerous reports from U. S. Bureau of Mines*1"^"'
as is the observation that at a constant size distribution, a reduction
in separation specific gravity will produce a coal product also reduced
in ash and sulfur. The extreme complexity arising from the presence of
these interdependent variables, together with the additional problem
that each coal seam is quite different and that there are also varia-
tions within one seam of coal from mine to mine, have presented a coal
preparation task which is usually solved by experience of the engineer
in charge or good old-fashioned intuition.
-------
132 CLEAN COMBUSTION OF COAL
This approach has served the industry well for many years. However,
with the demand for large quantities of cheap coal with an absolute min-
imum of sulfur and mineral matter we are now faced with the problem of
separating these variables in order to achieve the maximum possible
reduction of mineral matter. The purpose of this paper is to examine a
theoretical approach to the determination of the ultimate limits of ash
and pyrite removal from coal, given the specific gravity of separation
together with the concentration and size distribution of the raw coal,
mineral matter, and pyrite. Some of the experimental data used to test
this theoretical analysis were obtained from several other sources and
were accumulated for other purposes. As a consequence, the precision
suffers. However, it was presumed that with the establishment of this
theoretical approach, more accurate experimental data will be developed,
leading to an improvement in the correlation. The application of this
approach to mineral matter reduction other than pyrite was also devel-
oped but could not be tested as the necessary size distribution data
were not available.
THE EFFECT OF PARTICLE SPECIFIC GRAVITY
The following analysis is based on the rather straightforward as-
sumption that in the ideal gravity separation of a coal particle from
refuse, a very small mineral matter particle buried in the center of a
relatively massive coal particle will be recovered as coal product.
That is, a theoretical lower limit of retained ash and pyritic sulfur
in the coal product is calculated for an ideal separation of the raw
coal in an ideal bath. In the simplest case, consider a two-phase sys-
tem of coal with an unique specific gravity of 1.30 and iron pyrite
with a specific gravity of 5.0 where the iron pyrite is embedded in the
coal
TABLE I
Variation Of Sulfur Concentration In An Ideal
Coal~fi'onPyrite System As A Function Of
Volume Fraction Of Iron Pyrite
Volume Fraction Combined Weight Percent
Iron Pyrite Specific Gravity Pyritic Sulfur
0.1 1.6? 15.99
0.0? 1.559 11.49
0.05 1.485 8.99
0.04 1.448 7.38
0.03 l.ioi 5.68
0.02 1.3?4 3.89
0.01 1.337 2.00
such that the apparent specific gravity of the two-phase particle in-
creases with the volume fraction of pyrite as would, the concentration
of the pyritic sulfur in the particle. This relationship is important
because in an ideal gravity separation of coal and refuse, free pyrite
with a specific gravity of five will obviously separate as a reject
instantaneously; however, it is now evident that coal product with a
-------
WASHABILITY CURVES 133
substantial amount of pyrite could also be recovered from the separation
bath with the coal product at gravities from 1.60 to 1.30. One should
observe that there is no regard for particle size in this relationship
since only the volume fraction and specific gravity are involved. This
relationship will be used later to control the effect of separation
bath specific gravity on the concentration of ash or iron pyrite in the
coal product. A similar argument can be developed for the three-phase
system coal, mineral matter and iron pyrite if we assume an average
specific gravity for the coal and the mineral matter.
ANALYTICAL APPROACH
The analysis used in this discussion is based on the assumption
that in a given narrow size range of raw coal, if that system were sep-
arated in an ideal parting liquid of a particular specific gravity,
the coal product floated from the bath will also include some fraction
of mineral matter and iron pyrite which was included in the coal parti-
cles as a three-phase mixture.
The reason that these high density mineral matter particles were
included in the coal product was the fact that they had a particle
volume fraction small enough that the apparent specific gravity of the
particle was less than that of the separation bath, and that volume
fraction of mineral matter was embedded in the coal matrix. The rest
of the mineral matter (including iron pyrite) is rejected due to the
fact that its specific gravity exceeds that of the separation bath. It
should be recognized that this line of argument also implies that a
small volume fraction of coal is included in the mineral matter such
that the combined specific gravity never becomes less than that of the
bath.
Consider, as an example, a raw coal crushed to 1 cm x 0 from a
coal seam which had a six inch (15 cm) slate parting. If we examine
only that fraction of raw coal in the size range of 1 cm x 0.5 cm, we
can assume that all of the mineral matter from the mineral parting not
in the interfacial zone in this size range has a specific gravity which
exceeds the separation bath specific gravity of, say, 1.50; and thus,
will be rejected. In fact, we can also assume that all mineral matter
particles larger than that raw coal size range will be rejected. The
only recoverable mineral matter will be that which is small enough so
that its volume fraction does not cause the specific gravity of the
multiphase particle to exceed that of the separation bath. The re-
quired information for this solution is the concentration of the raw
coal and mineral matter in the given size range and the bath specific
gravity, if we can assume an average specific gravity of coal and min-
eral matter. After examining all size ranges and accumulating the re-
sults, we can theoretically ascertain limits of the coal product yield,
ash and pyritic sulfur concentration and heat content, if we are pro-
vided with the BTU/lb-ash relationship for that particular coal. Fur-
thermore, we can also examine how these functions vary with various
methods and limits of raw coal size reduction and separation bath grav-
ity.
-------
134 CLEAN COMBUSTION OF COAL
RAW COAL SIZE DISTRIBUTION
According to Evans, et al., coal fractures into a particle size
distribution which can be represented by a Rosin-Rammler distribution
function of the form
R = e-b (x)t (1)
where R is the weight fraction of coal retained on a sieve of opening
size (x) in millimeters. The constants b and t. are a function of the
fracture characteristics of the coal and fracturing technique. For ex-
ample, different top sizes of raw coal prepared by different comminu-
tion techniques result in different slope t and intercept b values for
one particular coal when the data are presented on Log-Log R versus
Log x coordinates. Mechanically crushed coals usually follow this rela-
tionship (linear) for raw coal top sizes well above one inch to sizes
well below 0.037/^m (400 mesh). Chemically comminuted coal^6-', on the
other hand, shows a distinct break in the curve in the range of 595_//m
(28 mesh), giving two lines of different slope and intercept. The
Rosin-Rammler distribution was accepted as a reasonable representation
of the raw coal size distribution in the following analysis.
In order to examine a narrow size range distribution of raw coal,
say from 30 mesh to 60 mesh, the weight fraction of the raw coal within
this range with a particular set of parameters (b, t) is given by
(R! - R2) (2)
where R,, R? are the fractions retained and are given by equation 1.
MINERAL MATTER SIZE DISTRIBUTION
To the author's knowledge, the particle size distribution of min-
eral matter in raw coal, disregarding iron pyrite, has not as yet been
quantified. However, we are all aware that mineral matter occurs in
various seams not only as partings 6 to 18 inches (15 to 4-5 cm) thick
and larger, but also in thicknesses in the millimeter range as well as
individual sand and/or clay particles in the micrometer and submicro-
meter size range. Let us presume for the moment that we can describe
the particle size distribution of mineral matter in the raw coal seam
using a Rosin-Rammler function described above, even though any other
suitable function representing that size distribution would also be
applicable as long as the key variables were fraction retained (R) on
a sieve of opening size (x). The Rosin-Rammler parameters for this
mineral matter are taken as R1, b1, t', and x'. Consider a. raw coal
sample from this particular seam which is crushed to some top size
where the raw coal size distribution is given by R, b, t, and x in
equation 1. In a narrow size range, x + dx, the fraction of raw coal
in that range is given by (RI - R2) and we can assume that the average
particle diameter (x + ^r) aiso represents an average particle diameter
on the mineral matter size distribution curve. All of the mineral mat-
ter larger than this average size can be regarded as having been re-
duced in size by the crushing procedure to reduce the raw coal sample
-------
WASHABILITY CURVES 135
to its size consist; and as such, has the specific gravity of mineral
matter, say, 2.2 as an average specific gravity. A fraction of the
remainder of the mineral matter, i.e., smaller than the average particle
size, will be recovered with the coal product. That fraction is deter-
mined by the volume fraction of mineral matter with an average specific
gravity of 2.2 which will just cause the coal product to float and be
recovered as coal product. All of the mineral matter less than the
average particle diameter (x ) is given by
cL
1 - Ba (3)
where the fraction with diameters larger than xa is given by Ra which is
defined by V , t1, and xa (=x + =£). The fraction of two-phase, ash-
coal particles which meet the specific gravity requirements is provided
by a factor (m); that is, equation 3 becomes
1- exp (-b'CmxJ*') (4)
CL
where m is the ratio of the diameter of the mineral matter particle (da)
embedded in the coal matrix as a two-phase system to the diameter of the
whole system (dm). The specific gravity of the mineral matter-coal com-
bined particle is given by (G) which is equivalent to the specific grav-
ity of the separation bath. More specifically, the density of the two-
phase particle is given by the volume fraction of mineral matter /^a.\
^
times its average specific gravity (fa) plus the volume fraction of coal
(y~) times its average specific gravity (f^) such that
v P + (-' (v -v ) - G v (5)
a a s c v m a' m N^'
if we assume a two-component system where Va + Vo = Vm and Vra is the
volume of the two-phase particle. Since we are examining the density of
any two-phase particle which will separate from the bath if its density
is greater than that of the bath and relating this to the apparent diam-
eter of the mineral matter (da) and the two-phase particle (dra) we can
then define (m) as r
|G -
m = da/dm
where G is the specific gravity of the separation bath. The effect of
m on the exponential equation 4 is to cause more mineral matter to re-
port to the rejects as m becomes a smaller fraction, i.e., as the grav-
ity of the bath is increased, increasing m, more mineral matter will
report to the product coal as we all recognize from practical experience.
The weight percent concentration of mineral matter (M) in the coal
product for any raw coal is then given by
n=o
where the first terra Gm is the concentration of mineral matter which is
-------
136 CLEM COMBUSTION OF COAL
approximately 1.1 times the weight percent ash concentration (dry basis)
in the raw coal(3). The second term represents that fraction of raw
coal in a particular size range under examination and the third term
represents that fraction of mineral matter which can be recovered based
on the limitations of specific gravity. The summation is taken over
all size ranges of the raw coal size consist such that M represents all
of the ash reporting to the coal product as a two-phase system. In
achieving this end we have made (inferred) some rather significant as-
sumptions which, in fact, are not necessarily true but may not seriously
effect the results. Firstly, one has to assume that the mineral matter
is homogeneously distributed which is probably true only for very small
sizes, I ^, less than 0,1 SOL, but is quite unlikely for particles greater
than 1 cm. This suggests that the smallest error will be incurred in
the analysis of raw coals with a top size below 0.5 cm. The specific
gravity of the pure coal was taken as an unique value which is, of
course, untrue due not only to the coal itself, but also due to the
mineral matter dispersed therein. The latter problem was accounted for
in the analysis of iron pyrite. However, there must be a cross term
which is complex due to the large number of exponential terms. In fact,
an ultimate analysis should contain each component of mineral matter
with its respective size and density accounted for and this is clearly
a complex mathematical problem.
IRON PYRITE (PYRITIC SULFUR) SIZE DISTRIBUTION
With the determination of the ideal mineral matter concentration
in the coal product, the concentration of the iron pyrite (specific
gravity (pp) of five), or the concentration pyritic sulfur (cp), can
be calculated following a similar procedure, recognizing that the
specific gravity of coal should be modified to account for the ash con-
tent.
The relationship between the iron pyrite concentration, Cip, and
that of the pyritic sulfur (Cp) is
0.53^ Gip= Gp (8)
and the density of coal with M weight percent mineral matter is given
by
- c
~ /k + 0.01 M (Pc-tO)
where * a is the average density of mineral matter (assumed 2.2) and Pc
is the average density of coal (assumed 1.30).
Due to the lack of extensive data describing the size distribution
of iron pyrite in coal, a distribution function has not been estab-
lished; however, it was interesting to observe that data recently accu-
mulated could be roughly represented by a.Rosin-Rammler plot. Figure 1
shows data points for the observed values^?' of the iron pyrite parti-
cles as plotted on Rosin-Rammler coordinates for the upper, middle and
-------
WASHABILITY CURVES
137
5
10
20
30
40
50
§70'
S 80
Lu
DC
oc
LU
90--
95--
97--
98--
99
LOWER BENCH
MIDDLE BENCH
UPPER BENCH
AVERAGE CURVE
345 10 20 40 60 80 100 200
PARTICLE SIZE mm x I03
Figure 1. Particle Size Distributions for Iron Pyrite
-------
138 CLEM COMBUSTION OF COAL
lower benches of one particular coal seam. For a precise Rosin-Raramler
correlation, all of the points should lie on one straight line which is
evidently not the case for these particular pyrite particles. The fol-
lowing analysis will incorporate an approximate Rosin-Rammler fit to the
olsserved data for mathematical simplicity. The distribution used here
could be replaced with almost any other function which more closely
represents the data when it is developed as long as that function in-
volves the variables applicable to the Rosin-Rammler relationship. The
error of the approximation used in this case should not be significant.
RETAINED FYRITIC SULFUR
The concentration of pyritic sulfur retained (Sp) in a coal product
after separation of raw coal is given by
, t*
_ / . / _n \ i / -.n^"j_ \ \ / •« t jt '\cun/
G (exp -b(a2 ) - exp -b(a2 ) ,) (1 -exp -b* i-r-(
F ~Q P (^
n = 0,1,2
(10)
where 0 is the concentration of pyritic sulfur in the raw coal system;
the second term represents the weight fraction retained between succes-
sive screen openings, with the size distribution having Rosin-Rammler
parameters b and £; and the last term represents that fraction of pyri-
tic sulfur which will be recovered with that coal. The series estab-
lished by (a2n) allows a continuous examination of each size range of
coal beginning at some minimum value (a) in mm and progressing to the
coal top size limit where we must set (n+l) =ODas there is no coal
beyond this size. The term | (2n+2n 1) in the third term establishes
the average size of the iron pyrite particle in the raw coal size range
under investigation.
The iron pyrite size distribution is established with new distribu-
tion parameters b* and t* and m based on the ash modified density of
coal (^k). Due to the lack of an exact distribution function for the
pyrite under investigation, it was presumed that the dashed lines shown
in Figure 1 represented a Rosin-Rammler fit for the data, giving a solu-
tion which must be considered only a first approximation.
RESULTS
A raw coal sample pulverized commercially to a nominal 200 mesh
) x 0 size consist from a coal mine in which the iron pyrite size
distribution was recently measured was used to test equation 10. Raw
coal samples were carefully separated at different specific gravities
by the Otisca Process and also separated at a different laboratory by
means of the accepted washability procedures. The complete sulfur form
analysis of all of the products of separation and the raw coal were con-
ducted by all parties as a cross-check.
-------
WASHABILITY CURVES 139
The observed size distribution of the iron pyrite in the three
benches of that coal seam were given in Figure 1. In order to simplify
the mathematics, each curve was approximated by a straight line, or
Rosin-Rammler equivalent, to represent all of the observed data points.
The approximation is given by the dashed line in Figure 1. The total
pyritic sulfur concentration in the raw coal was about 2 percent (dry
basis) which was distributed through the three benches in approximately
the following order t
BENCH
Percent Mass
Pyritic Sulfur Fraction Pyritic Sulfur (C-p)
Top Bench 3.02 0.339 1.022
Mid-Bench 1.81 0.39? O.?l8
Bottom Bench 1.21 0.264 0.31?
Due to different size distributions and concentrations of pyritic sulfur
in the three benches, equation 10 was solved for each bench separately
and then the data accumulated to ascertain the total pyritic sulfur that
would be recovered. The variables necessary for the solution of equation
10 for various separation gravities are shown in Table 2 where it was
assumed that the coal retained 8 percent ash as indicated by the wash-
ability data. The raw coal size ranges tested, i.e., for n=o, n=l, n=2,
etc., when (a=5xlO~-3mm) were 5/(m-10/'m, 10/'m-20/tm, 20An-40./iii, and
so on; and there was no coal larger than 0.32 mm.
Identical samples of the pulverized coal sample were separated by
an outside laboratory experienced in the standard washability technique
and independently on a riffled quarter of that sample of raw coal by the
Otisca Process at our laboratory. The products of the Otisca Process
separation were then sent to the former laboratory for complete proxi-
mate and sulfur forms analysis. The results for the pyritic sulfur in
the raw coal cited before and those for the coal products separated at
various gravities are shown in Figure 2. A comparison of the results
from these two tests for coal product yield, total sulfur and ash are
shown in Figures 3 and ^ respectively.
DISCUSSION OF RESDLTS
The experimental separation data of the pulverized coal at various
gravities shown in Figures 3 and b can be considered quite reliable due
to the fact that the products and raw coal were analyzed independently
by several different groups. The agreement of the theoretical analysis
to within 10 percent of the experimental values for the retention of
pyritic sulfur in the product coal was considered somewhat fortuitous
considering the large number of assumptions that were required to com-
plete the analysis and the fact that the analytical data necessary for
the solution of the equations was received by chance rather than by a
directed effort. Clearly, the next step is to test this analytical
procedure further for other coals and coal seams in order to prove
whether or not it is indeed a direction to a theoretical determination
of the washability curves.
-------
140
CLEAN COMBUSTION OF COAL
TABLE 2
SOLUTIONS FOR THE DETERMINATION OF PYRITIG SULFUR
RETAINED IN GOAL PRODUCT
BENCH
Upper
Mid
Lower
Variables For Equation 10
b
t
b*
t*
CP
41
1.30
14.2
1.33
1.033
41
1.30
17.6
1.1
0.?18
41
1.30
11.2
0.72
0.317
Solutions For m=0.345 (1.50 Sp.Gr.) a=5xlO"^)
0.0003 o.ooi o.o°3
0.002 0.0047 0.0089
0.0079 0.0187 0.0259
0.026 0.0505 0.0505
0.0368 0.0567 0.0412
0.0095 0.0112 0.0062
0.0825 0.1428 0.1357
3
4
5
TOTAL
Total Summation 0.361 Percent Pyritic Sulfur Recovery
At 1.50 Specific Gravity
Solutions For m=0.300 (1.46 Sp.Gr.) a=5xlO"3
n = o
1
2
3
4
5
TOTAL
0.0
0.001
0.00?
0.022
0.031
0.008
"67559
0.001
0.004
0.016
0.044
0.050
0.011
0.126
Total Summation 0.321 Percent Pyritic Sulfur Recovery
At 1.45 Specific Gravity
Solutions For m=0.239 (1.40 Sp.Gr.) a=5xlO"3
n = o
1
2
3
• 4
5
TOTAL.
0.0001
0.001
0.005
0.016
0.024
0.006
0.052
0.0006
0.003
0.013
0.035
0.041
0.009
0.107
Total Summation 0.272 Percent Pyritic Sulfur Recovery
At 1.40 Specific Gravity
0.002
0.007
0.021
0.042
0.035
0.006
0.113
-------
WASHABILITY CURVES
141
(ft
or
>-
a.
Q
UJ
UJ
QC
UJ
o
cc
LU
CL
0,40
2 0.30
0.20
THEORETICAL
EQUATION 10
INDEPENDENT
WASHABILITY
OTISCA PROCESS
0.15-
1.50
i i
1.40
\
V
1
1.30
SEPARATION SPECIFIC GRAVITY -
Figure 2. Percent Retained Pyritic Sulfur as a Function
of Separation Gravity for 200 Mesh X 0 Coal
and Observed Data
-------
142
2.0 -
Q
CL
CL
D
ID
CO
O
OL
LJJ
£L
0.8
1.30
CLEAN COMBUSTION OF COAL
PULVERIZED COAL
200m x 0
x OTISCA PROCESS
• WASHABILITY
1.40 1.60 1.80
SEPARATION SPECIFIC GRAVITY
Figure 3. Washability Results for Pulverized Coal
'•80
•-70
--60
--50
40 u
D
O
s
Q.
--30
o
u
--20
•• 10
-------
WASHABILITY CURVES
143
14 +
12 +
D
O
2
0.
O
u
10 I
^ 8 +
x
CO
26 +
LU
O
QC
LU
CL
4 +
WASHABILITY
OTISCA PROCESS
PULVERIZED COAL
200m x 0
+
1.30 1.40 1.60 1.80
SEPARATION SPECIFIC GRAVITY
1.90
Figure 4. Observed Washabllity Results for Pulverized Coal
-------
144 CLEAN COMBUSTION OF COAL
Since there is virtually no mineral matter size distributions
available, let us turn immediately to the pyritic sulfur analysis cited.
If we assume that current coal preparation techniques have little effect
on the organic sulfur in coal and the concentration of sulfate sulfur
in eastern coals is small enough to be inconsequential, then the key to
desulfurization of eastern coals lies in our ability to eliminate iron
pyrite from the raw coal matrix. The variables required to express a
quantitative analysis of iron pyrite extraction from raw coal include
raw coal size distribution, pyrite size distribution and the specific
gravity of the separation bath. Only the pyrite size distribution is
outside the control of the coal preparation engineer unless he has a
choice of coal seams that have significant differences in pyrite size
distribution from which he can extract raw coal. In this case, the
choice, if the size distribution is Rosin-Rammler would be that seam
with the smallest positive slope (t*) and the smallest intercept value
(b*). This would present the greatest release of pyrite with the least
amount of raw coal size reduction.
Whether or not various techniques of raw coal size reduction, i.e.,
Hardinge mill versus hammer mill versus other impact methods or even
chemical techniques, promote greater iron pyrite release in a given
situation, i.e., fixed iron pyrite distribution, has not been demon-
strated unequivocally. Quantitatively what is implied is that for a
given iron pyrite size distribution, a given raw coal particle size
distribution generated by the two different comminution techniques and
a given bath separation specific gravity, one form of comminution might
produce coal product with less retained iron pyrite. Such might well be
the case, if one of our initial assumptions, which was purely arbitrary,
was incorrect and that the very small iron pyrite was not homogeneously
dispersed throughout the coal system, but lay on preferred sites which
during fracture were preferentially freed from the coal particle.
Clearly, pyrite embedded within the coal particle will remain there al-
ways. Only very careful experimentation with strict control over the
variables will provide the final answers.
The analytical expression given in equation 10 is only a first ap-
proximation as the effect of certain variables known to exist in the raw
coal system have not been considered in their entirety and their effect
can only be surmised at this time. Firstly, it was assumed in the util-
ization of the term (m) which accounted for the variation in bath spe-
cific gravity that coal was of only one specific gravity (1.30), and, of
course, this is just not true. If, however, this value were an appro-
priate average specific gravity for all of the coal in that particular
seam, one might suspect that the coal of lower specific gravity and its
recovery of iron pyrite would just about cancel the lost iron pyrite
with the coal fraction of higher specific gravity. The average value
of the specific gravity of the coal might be obtained from a standard
experimental washability curve which was developed to include the calo-
rific value of that particular coal. After accounting for the BTU
units in the pure rejects, i.e., 1.60 gravity or higher, then find that
gravity where there are equivalent coal heat units in the rejects and
products.
The mathematical simplification of first separating the mineral
-------
WASHABILITY CURVES 145
matter and then the pyritic sulfur neglected the fact that in the real
case a three-phase system is involved, i.e., the limiting specific
gravity (m) for mineral matter retention should also include the effect
of the^retained iron pyrite. Although this correction is rather small,
i.e., Tk *" 1»35Q versus a correct value of Cfc« = 1.353t it would tend
to depress the theoretical values shown in Figure 2 into the range of
the observed data. Most encouraging, however, is the general slope
change of the theoretical curve with the variation of specific gravity
and its correlation with the experimental curves. That is, the retained
pyrite versus specific gravity curve tends towards negative infinity as
the separation bath specific gravity approaches the specific gravity of
the coal, cf. equation 6.
The development of a washability curve based on the data determined
from the above equations is relatively straightforward, that is, the ash
in the coal product versus separation specific gravity curve is obtained
from solutions for 0.9 M in equation 7i as the G value in equation 6 is
varied. The total sulfur in the coal product is determined through
equation 10 which provides the pyritic sulfur content plus the sulfur
contributions due to organic sulfur as most sulfate sulfur is lost in
the rejects. Again equation 10 is varied over the entire specific grav-
ity ranges.
Since each of the above calculations represents that fraction of
the mineral matter and pyrite recovered in the coal product, the residue
of the concentration in the raw coal identifies that quantity of mater-
ial rejected in the separation and that difference from one represents
the coal recovered, i.e., yield, in each separation specific gravity
case. A theoretical washability curve has been generated provided our
assumptions are not too severe.
The difference between the standard washability curve and the
Otisca Process data was due to the different procedures that were used.
The standard washability tests used a standard commercial solvent system
with a separation time of about twenty-four hours. The Otisca Process
used a proprietary liquid and an additive package which allowed a sep-
aration time of not more than twenty minutes for size consists of less
than 100 mesh (0.1 mm) x 0. Figures 3 and k represent the first head-
on tests using the Otisca Process and standard washability on a common
sample. The results were rather encouraging.
CONCLUSIONS
A first approximation for the theoretical determination of wash-
ability curves was presented and partially tested using a nominal 200
mesh x 0 raw coal separated by classical washability techniques and the
Otisca Process. The results indicated that the pyritic sulfur retained
in the product coal could be determined within 10 percent for raw coal
separations in an ideal bath at specific gravities of 1.50t 1.^5. and
1.40. The key variables required for this analysis include for the raw
coal the size distribution of the iron pyrite, mineral matter, and the
crushed raw coal and the concentration of the ash and iron pyrite. The
specific gravity of the coal, mineral matter, iron pyrite, and separation
-------
146 CLEM COMBUSTION OP COAL
bath are also required. One of the key utilizations of this analytical
approach is that for a given seam, i.e., fixed mineral matter and iron
pyrite size distribution, the optimum comminution schedule can be de-
termined to permit the maximum reduction of ash and pyritic sulfur in
the product coal. Furthermore, once the theoretical limits for this
seam are available, then one has a fixed reference point for comparison
with practical coal preparation procedures.
ACKNOWLEDGMENT
The author would like to acknowledge the valuable comments and in-
sights provided by Dr. Andrew Rainis during his review of this paper.
REFERENCES
1. Cavallaro, J.A., Johnson, M.T., and Deurbrouck, A.W., "Sulfur Re-
duction Potential of the Coals of the United States", U. S. Bureau
of Mines, Report of Investigation 8118, 1976.
2. Deurbrouck, A.W., "Sulfur Reduction Potential of the Coals of the
United States", U.S. Bureau of Mines, Report of Investigation 7633,
1972.
3. Lowry, H.H., «i, "Chemistry of Coal Utilization", John Wiley and
Sons, Inc., New York, Chapter 8, 1963.
Ij-. Leonard, J.W., Mitchell, D.R. ed, "Coal Preparation", Amer. Inst. of
Mining, Met. and Pet. Eng., Inc., Chapter**, 1968.
5. Evans, I. and Pomeroy, C.D., "Strength, Fracture and Workability of
Coal", Pergamon Press, New York, Chapter 7_» 1966.
6. Keller, Jr., D.V. and Smith, C.D., "Spontaneous Fracture of Coal",
Fuel, &, 273-80, 1976.
7. Private Communication.
-------
147
SESSION III - COMBUSTION TECHNOLOGY
SESSION CHAIRMAN: G. BLAIR MARTIN, U.S. EPA
In achieving the projected increased use of coal in the industrial
and utility sectors, a major emphasis must be placed on the environmental
impacts of the combustion systems. The pollutant species to "be consid-
ered include not only the criteria pollutants (SOX, NOX, CO and particu-
late) but also trace species (POM, organics and trace metals). To control
emissions in the most energy-efficient and cost-effective manner, the
combustion system should be considered from an overall point of view and
may include provisions for precleaning of the fuel, proper design of the
combustion device and postcleaning of the flue gas. The combustor design
is determined by the choice of process for control of SOX emissions. The
sulfur may be removed from the fuel (coal cleaning or gasification),
removed during the fuel combustion process (fluid bed combustion), or
scrubbed from the flue gas (pulverized coal boiler).
Coal cleaning and flue gas treatment are covered in other sections
of the report; therefore, this session concentrates on the combustion
processes or equipment that may be used in conjunction with the various
SOX and particulate removal schemes. The pollutants that may be con-
trolled by proper design of the combustion process include NOX, CO, POM,
organics and carbon particulate. Modification of the combustion process
has significant potential for NOX control while maintaining low levels
of the other pollutants. Since the currently available technology is
the conventional pulverized or stoker coal-fired boiler with low sulfur
fuel or flue gas desulfurization, emphasis is placed on this type of
equipment. Some potential near-term technologies are also being devel-
oped and two of these are also represented.
The overall session is split up into three types of papers. There
are two papers covering general aspects of combustion technology:
(l) the basic design and operating principles of combustion systems,
and (2) the basic aspects of pollutant formation and control in the fuel
combustion process. There are three papers dealing with current tech-
nologies for combustion of coal: (l) low emission burner design for
pulverized coal-fired utility boilers, (2) coal oil mixture combustion
in industrial and utility boilers, and (3) stoker coal-fired industrial
boilers. Finally, there are two presentations on potential near-term
alternatives: (l) fluid bed combustion of coal for industrial steam
generation, and (2) pulverized coal burners for industrial process
furnaces.
-------
148 CLEAN COMBUSTION OF COAL
-------
149
SOME CHARACTERISTICS OF COAL COMBUSTION SYSTEMS
Janos M. Beer
Massachusetts Institute of Technology
Cambridge, Massachusetts
INTRODUCTION
Coal is burned in a large number of industrial processes: in
furnaces, kilns and boilers. The recent development of coal combustion
systems is, however, closely tied to the development of steam raising
plants. Stoker firing has reached a high level of sophistication but
could not follow the steady increases in unit capacity demanded by the
economy of scale in power station development and gave way to pulver-
ized coal combustion which in its various forms is the presently pre-
vailing mode of industrial coal combustion. As a result of environmen-
tal concerns and also due to the plans for significant expansion in the
use of coal, interest turned toward a new combustion system, the fluid-
ized combustion of coal, which holds out promise for sulphur retention
without flue gas desulfurization, and low NOX emission at low capital
and operating costs. The scope of this paper does not permit detailed
discussion of these processes; it is intended instead to draw attention
to some characteristics of these combustion systems, seen through the
physical-chemical processes that coal undergoes during combustion.
In coal combustion processes the engineering task is to insure
• stable ignition of the coal
• complete combustion of both the volatiles and the residue of
char
• low pollutant emission
• good availability of plant (free of excessive deposit formation
and corrosion)
• all these achieved with a minimum of excess air at an acceptable
cost in pressure energy.
It is of special advantage if a combustion system can be easily inte-
grated into energy conversion cycles of high thermodynamic efficiency.
In the following we shall look at combustion systems from the point of
view of how far they satisfy the above criteria.
-------
150 CLEAN COMBUSTION OF COAL
PYROLYSIS, IGNITION AND COMBUSTION OF COAL
When coal is heated, the moisture is expelled first, followed by
the thermal decomposition of the coal. This latter process is usually
called devolatilization. The amount of the volatile matter driven off
depends upon the type of coal and also upon the rate of heating and the
final temperature of pyrolysis. The higher the final temperature and
the rate at which the coal is heated, the larger the proportion of the
mass lost during devolatilization. At sufficiently high temperature
the total solid combustible mass of the coal can be volatilized. The
process of volatile evolution can be quantitatively described in terms
of a large number of simultaneous irreversible, first-order reactions .
Ignition usually occurs during the thermal decomposition of the
coal. It is likely that some of the lower ignition-temperature high
molecular weight hydrocarbons ignite first. Their ignition may be
catalyzed by the solid surface which may also be ignited during this
process2. In large particles, where the volatile evolution and burning
goes on concurrently with the combustion of the solid matter, it is
difficult to separate these processes in time. (This is reflected in
the need to provide sufficient space for volatile-combustion in stoker-
fired combustors almost above the whole length of the grate.)
The combustion process of the residual char can be considered to
consist of steps such as (a) the diffusion of the oxidant to the particle
surface in counter-diffusion with the products of combustion away from
the surface, (b) the diffusion into the pores of the particle, and
(c) the chemisorption reaction at the surface, consisting of the acti-
vated adsorption of the oxidant and the desorption of the compounds
formed at the surface. The relevance of these individual steps to the
overall rate of burning of the particle depends upon parameters such as
particle size and reactivity of the char, temperature, and the nature
of the oxidant (COg, H20, OH or 02). For example, when the particles
are large and the temperature is high (stoker firing), the rate deter-
mining step is external diffusion; when the particles are large and the
temperature is low, diffusion and/or chemical reaction can be the rate
limiting steps (fluidized combustion); and when the particles are small
and the temperature is high (pulverized coal combustion), diffusion and/
or chemical reaction at the surface determine the overall rate of the
oxidation reaction .
STOKER FIRING
Ignition on the Traveling Grate
The top layer of coal on the grate is heated by radiation from the
flame, and by convection from hot combustion products. The combustion
air preheat temperature is usually low (<200°C) in order that the air
can cool the grate. When coal is ignited from below by air preheated
to higher temperatures, high relative velocities between coal and gas
will reduce ignition delay. Figure 1 illustrates how ignition delay-
times depend upon the air velocity for the case of ignition from the top
by radiation and/or from below by air preheated to H25°C; it can be seen
that there is an "ignition gap" in the range of gas velocities of 0.11-
0.27 Nm/s in which neither of the above-mentioned ignition modes can
-------
COAL COMBUSTION SYSTEMS
151
produce ignition by themselves. Ignition can, however, be obtained by
the combination of the two methods .
Figure 2 represents an optimized form of the refractory arches
determined by modeling radiative transfer between the burning coal on
the grate and a refractory arch on one hand, and between this refractory
arch and the green coal on the other5. Care had to be taken to ensure
that the ignition arches serve their purpose without causing slagging
at the high temperature end of the grate.
Another engineering solution to improve the ignition of a high
moisture coal is shown in Figure 36. The lower part of the coal hopper
is altered to enable hot combustion products to be drawn through the
coal layer before it reaches the grate. In this arrangement only the
fixed grate, at the bottom of the hopper, has to stand up to high gas
temperatures and the beneficial effect of combined ignition can be
obtained without the flow of highly preheated air through the traveling
grate.
20
min
is
K
ft
IGNITION
C3
From Above
1170
From Below
Coal Layer
Ignition
— -- First Inflam.
0
-------
152
CLEM COMBUSTION OF COAL
is shown for three temperatures7. While the combustion rates can be
seen to level off at low air velocities (laminar flow), they are increas-
ing again, once the flow becomes turbulent at higher air velocities.
Another significant finding of these researchers is that the composition
of the combustion products shifts toward higher CO/C02 ratios as the
blast velocity increases. This has been confirmed to hold also for a
fixed bed of particles8 as shown in Figure 5-
On the traveling grate steady state conditions prevail: the events
that happen sequentially on the fixed grate can be seen spatially sepa-
rated, but occurring concurrently on the traveling grate. Figure 6
illustrates this point: the composition of gaseous species above a
traveling grate are shown on the top, and their respective volume flow
rates through the burning layer of coal, on the lower part of Figure 69.
The operation of practical fuel beds is hampered by nonuniform
pressure drop across the bed due mainly to the nonuniform distribution
of the fine particles on the grate. Fine particles tend to segregate
during their passage through storage bunkers and in the hopper above the
grate and will cover the grate in patches.
25
2.0
1.5
1.0
0.5
\fltXf
X
10 20 30 40 50 60 l/min
Fig. U. Variation of Combustion
Rate of a Carbon Channel
with Air Velocity (After
Tsukhanova7).
Fig. 3. Ignition by Hot Recirculated
Gas.
Figure 7 represents results of experimental studies10 showing the
effect of the proportion of fines (<5mm) in the feed upon boiler per-
formance and efficiency.
-------
COAL COMBUSTION SYSTEMS
153
10 XI JO IO 10
Distance into layer, mm
II
t*
ISOO
u " t JJfojo ta40
Distance into layer, mm
**r "I
e a 10 la to so
Distance Into layer, nun
m
Fig. 5- Gas Formation in a Layer Fig. 6. Species Concentration and
of Particles of Electrode
Carbon: I) Pate of Blast
0.11 m/sec; II) Rate of
Blast O.U9 m/sec; III) Rate
of Blast 1.50 m/sec (After
Khitrin8).
Flow Rate Distribution Above
a Travelling Grate (After
Werkmeister ).
u
SO (O TO tto
fCRCCNTMC MCATER 7 HAN 5mm.
Fig. 7- Effect on Efficiency and Performance
of Removing Fines from a Fuel Bed
(After Zagon in Ref. 6).
-------
154
CLEAN COMBUSTION OF COAL
One of the engineering solutions of this problem was the develop-
ment of spreader stokers (Figure 8): the coal, unclassified, is thrown
or blown into the combustion chamber and due to the inertial and drag
forces acting o'n the particles in flight, the smaller particles land on
top of the large particles on the grate. The finest particles burn in
suspension above the grate. This in turn makes it necessary to raise
the height of the combustion chamber and then to cool walls by screens
of steam generating tubes to avoid slagging of the refractory surfaces
by the fly-ash. Because of the high dust loading of the flue gas,
spreader stokers normally require mechanical particle precipitators.
Another solution of the problem caused by fine particles, a combi-
nation of coal classification, stoker and pulverized coal combustion
is shown schematically in Figure 911- The coal is fed into a flash
Fig. 8.
Traveling Grate Stoker. Section Through Rotograte Stoker In-
stallation.
1.—Raw Ceil.
2.—Feeder.
3.—Classifier Tube.
4.—Fine Coal.
5.—Coarse Coal.
6.—-Vessel.
7.—Rotary Seal.
8.—Hopper.
9.—Combustion Chamber.
10.—Bypass.
11.—Transport Duet.
12.—Pulverizer and Fan.
13.—P.F. Duet.
14.—Branch tor Temp. Control.
Fig. 9- Diagrammatic Arrangement of the Combined Firing System.
(Beer11)
-------
COAL COMBUSTION SYSTEMS 155
drier tube and drops in counter-flow with hot combustion products drawn
from the combustion chamber. The fines are lifted from the surface of
large particles, carried by the gas stream into a pulverizer, and blown
through burners into the combustion chamber above the grate. Due to
the interaction of the combustion of classified coal on the grate and
the pulverized coal flame above the grate, high combustion efficiencies
with low excess air can be obtained even when burning low grade coal.
When existing stoker-fired boilers are fitted with this system, the
boiler performance can also be significantly raised11.
PULVERIZED COAL COMBUSTION
As the unit capacity of boilers rose significantly, during the
period between the two World Wars, and there was a requirement for
improved efficiency and better automatic control, pulverized combustion
of coal has steadily replaced stoker firing in units larger than
250,000 Ib/hr steam, and therefore in utility application almost entirely.
The combustion efficiency in pulverized coal combustion is high
because particles dried and ground to below 200 ym can be intimately
mixed with the combustion air, and burned completely with a minimum of
excess air. The combustion air can be preheated to high temperatures
unlike in stokers where the combustion air has to cool the grate. The
use of higher air-preheat means that the flue gas leaving the boiler
can be cooled by the much larger air preheater and hence advantage can
be taken of regenerative feed water preheat with steam bled from the
turbine without any deterioration in boiler efficiency.
Due to the relatively short residence time of the burning coal in
the combustion chamber (<2 sec) load-following is only limited by the
rate at which the coal grinding-feeding system can respond to command.
This varies with different coal preparation systems but the response is
generally fast, the time constant being about one minute. The fast
response characteristics are compatible with requirements for flexible
automatic control and good load following of the boiler plant.
Last but not least, the nature of the flame: capable of filling
the combustion chamber with a strongly and uniformly radiating medium
made it possible to adapt to this combustion system the water walled,
"radiant" combustion chamber design, and this has ensured the scaling
up of pulverized coal combustion to the sizes of the largest utility
units (>1000 MWe).
Ignition of Pulverized Coal
When a cloud of pulverized coal particles is injected into a
furnace, it is heated up as it approaches the flame front partly by
radiation and partly by mixing with hot recirculated combustion products.
Nusselt12in a classifical paper developed the theory of unidimen-
sional laminar flame propagation in a coal dust cloud and solved his
equations for the flame speed as a function of particle size, dust
loading of the gas and temperature of the flame, assuming solely
radiative heat transfer. Essenhigh and Csaba13 have used Nusselt's
-------
156 CLEM COMBUSTION OF COAL
theory for the interpretation of Csaba's experiments. Figure 10 shows
results calculated from their study for "ignition temperatures" of
650°K and 900°K, respectively. The prediction is that particles of
30 um diameter will have ignition delays of about 100 ms and flame
speeds lower than 1 m/s for the above ignition temperatures of the dust
cloud and assuming a maximum radiance of the flame at 100 kcal/m^sec.
Rates of flame propagation in practical pulverized coal flames are,
however, about an order of magnitude higher than those shown in
Figure 10, due mainly to the effect of hot gas entrainment into the
fuel-jet as it approaches the flame front (Figure ISa)1"*. The respective
contributions of radiation and hot gas entrainment to the ignition of a
pulverized anthracite as a function of distance from the burner and for
different coal finenesses are shown in Figure 12, and ignition distances
for the same anthracites and for two burner types in Figure II15.
The heat required for raising the combustion air temperature to
ignition is several times that necessary for heating up the coal parti-
cles in the coal-air mixture. This is one of the reasons for injecting
the coal dust with only a fraction of the total combustion air, the so-
called primary air. The rest of the combustion air is normally mixed
in downstream of the flame front.
The primary air fraction is usually proportional to the volatile
matter of the coal so that sufficient air is available for the combus-
tion of the volatiles as they evolve. Figure 13a and b show the effects
of the primary air fraction and the volatile and ash content of the coal
upon the flame propagation rate.
Results of Badzioch's experiments on isothermal decomposition of
coal16 are shown in Figure Ik. The family of curves illustrates the
point that volatile yield is a strong function of pyrolysis temperature.
In recent studies carried out by Anthony and Howard17 and by
Kobayashi, et al. ,18 it was shown that an increasing proportion of the
combustible solid mass can be volatilized as the pyrolysis temperature
increases and that the volatile yield so produced can far exceed the
volatile matter determined by ASTM standard test (Figure 15). In staged
combustion processes it will be of advantage to volatilize a large pro-
portion of the coal and thus provide conditions for the conversion of
the maximum amounts of fuel-nitrogen to N2 in the first, fuel-rich
stage of the process19'20.
The solid residue burns out in the second stage or the tail-end of
the flame. The combustion of "large" particles (>100 ym) is normally
controlled by external diffusion and hence their burning times are
proportional to the square of the initial particle diameter** (Figure l6).
For smaller particles the burning time becomes longer than that predicted
from diffusional transport because the rate determining step in their
reaction mechanism is the activated adsorption of the oxidant on active
sites at the surface or in the pores of the solid. Since the rate of
mass transfer varies inversely with particle diameter and the rate of
chemical reaction is independent of particle size, the temperature at
which the transition from mass transfer to chemical rate control occurs
must depend on the particle size. The nature of this dependence is
-------
COAL COMBUSTION SYSTEMS
157
O-2 0-4 O-6 0-8 1-0
Input velocityJt/0 i m/s
0-1 0-2 O-3 0-4 O-5 0-6 O-7
Input velocity, Ua : m/s
Particle diameter
Inlet temperature = 350"K
If = 100kcal/mzs
Fig. 10. Theoretical Variation of Ignition Time with Input Velocity,
Coal Concentration and Ignition Temperature in Plug Flow.
4-Qt 1 1 1 1
3-0
S.
"1-0
\
X S
Low-volatile coal
O
• \
X Burner A
O Burner B
Fig. 11.
20OO 4OOO 600O
Specific surface of fuel ,SW : cm2/g
The Variation of Ignition Distance in Anthracite Flames
as a Function of Fineness15.
-------
158
CLEM COMBUSTION OF COAL
200
100 200
distance from burner
Fig. 12. Heat Transferred to the Pulver-
ized-Fuel Jet Upstream the Flame Front by
Recirculation and Radiation.
ft
It
It
10
\>
s,
f
i
Quantify of ttr — *•
*
1
1
/
/
1
1
1
>,
\
•rf.
N
^
%
\
x<
'
R
1
sf-
^
^
\
"*+
^
t -a 20 x w so so 70 so so 100 tin so oar. Kt
frimary air-coal ratio for 'a' *
Bituminous coal 'a'l
Votatilm 30%
Ash : SY,
Moisture: 3V..
H, : 7530 Heat/kg
LU-- i.25Noi/Kg
Fig. 13a. Relation Between Pri
mary-Air Fraction and Ignition
Velocity of Pulverised-Coal
Mixture.
Wl
^
f"
«e ^
-<.
V
V
A
X
» ^ 30 V. W
Votatilf .
^"^ matter -
Ash -
Influence of Volatile
Fig. 13b.
Matter Content and Ash Content on
Ignition Velocity of Pulverised-
Coal Mixture (After Dolezal1*).
20 4O 60 60
Isothermal decomposition time,/: ms
100
120
Fig.
Variation of Weight Loss with Time at Various Temperatures
For a Coal of Low Rank (After Badzioch16).
-------
COAL COMBUSTION SYSTEMS
159
O2100 K
01940
O174O
0151O
V126O
A 1000
SO 100 150
RESIDENCE TIME (ms)
200
Figure 15.
(After Kobayashi-Howard and Sarofim18).
iv i
1.4
1.2
1.0
0.8
<
\
\
V
\
1
\
\
V
<^
**•«
JO 1.2 1.4 1.6
excess air
b
,j
It
If
rr
^
n
In
c
n
*s
1
Ol
•)i
/t
//
'T
JS
r
/
/
al
t
f
/
/
/
/
I,
j
' f
/ 1
S
/
/
J
i
/
' S
**t
I
/
i
f
f
yS
/
J
I
'
/
f
f
f
t,
cc
r6vy
/
/
/
i
I
f/
t
f
>K
rn
/
/
/
J
^
/
e
cc
1
'
i
f
f
j
,
f
a
/
i
j
}
)
) 0.1 0.2 0.
Particle Diameter
sec
2.0
1.5
u
E
en
1.0 <=
c
X!
0.5
O
m.m
Fig. 16. Burning times of Coal Particles.
(After Gumz1*).
-------
160
CLEM COMBUSTION OF COAL
illustrated by the logarithmic plot of burning time against particle
diameter given in Figure IT21- It follows that the rate of burning of
small particles (<50 ym) depends more on the temperature and the partial
pressure of the oxygen, than on the particle size.
In practice there are many methods for aiding ignition or combus-
tion of difficult-to-burn coals. The environmental constraints of low
NOX emission, however, make their application more difficult. Two such
methods are mentioned in the following with examples of application to
low grade coal combustion in power station boilers: (a) improving
ignition and combustion by means of changes in the coal preparation
system, and (b) the use of pulverized coal pilot flames to ensure stable
ignition when burning low grade coals.
K>
10 10'
Initial particle diameter 6»m)
10
A
B
C
D
E
F
G
H
T OK) E (kcal mole-i) P (atm.)
1000
2000
1000
2000
1000
2000
2000
2000
20
20
10
10
5
5
10
10
10
100
Solid lines indicate mass transfer control, all pres-
sures; a, 1000-K; b. 1600"K; c, 2000'K
Fig. 17. Theoretical Particle Burning Times Showing Effects of
Particle Diameter, Temperature and Pressure. (Ref. 21)
-------
COAL COMBUSTION SYSTEMS
161
Unlike in other coal burning systems, in pulverized coal combustion
the coal preparation (drying, grinding, transport of fuel to the burners)
is an integral part of the combustion system. The coal preparation can
be "direct" or "bin and feeder." Direct coal preparation (Figures l8a,
b) means that the drying medium—preheated air and/or flue gas—is
moving the fuel through the grinding mill, and the particle size classi-
fier, transports the ground coal to the burners and serves also to
provide the immediate environment for the coal during pyrolysis, igni-
tion and the initial stages of burning. Direct coal preparation plants
are relatively simple, but the task to satisfy the many, often conflict-
ing requirements for drying, classification, transport and primary air
during ignition and combustion is difficult, particularly for low grade
coals22.
Burner.
Feeder.
Fig..I8a.
Direct-system
with i/ie pu/vemer
mill under pressure.
Fig. l8b.
Direct-system
with ihe pulveriser
mill under suction.
Mill-Fan.
Mill.
Preheated air and/or
Furnace Coses.
Figure 18.
The "bin and feeder" systems (Figures 19a, b) permit the separation
of the drying-grinding circuit from that of transporting the fuel to the
burner, so that the coal can be injected into the flame by a prescribed
proportion of the combustion air. "Bin and feeder" systems are more
complex in their operation, require more precaution particularly when
the dried pulverized coal is stored.
An example of how the suitable choice of the coal preparation
system can be used for improving plant efficiency is shown in Figure 2023.
Here the total amount of flue gas is recirculated and used as drying
medium for a high moisture lignite. The flue gas can, in this way, be
cooled down to below 100°C instead of the usual 200°C when expensive
heat exchangers in the boiler have to be protected from low temperature
corrosion.
-------
162
CLEAN COMBUSTION OF COAL
Fig. 19a.
Open-circuit
with storage.
Furnace
gas.
Tottie
atmosphere
Exhausf '
tan._
-RF.precipitator
T0rfhe
PFburner
Fig. 19b.
C/osed-c/rcuit
with storage.
Cyclone
Mill.
Furnace
Coses
Prtmoru
Air Fon
VFan.
Effluent
Burner
\f. Burner
Preheated
Air
Fig. 19. "Bin and Feeder" Systems.
A/ff
COMBUSTION CAS.
Fig. 20. Applied Open-Circuit System with Storage.
(After P.N. Kendys, World Pover Conference,
1956. Report No. 2^862/lJ.)
-------
COAL COMBUSTION SYSTEMS 163
In the other example the combination of a "bin and feeder" coal
preparation system and a regenerative-type burner has been applied to
solve the load following problem when burning a low grade coal2".
Figure 21 shows the arrangement of the pilot burner in relation to the
main burners in a corner-fired combustion chamber and Figure 22 illus-
trates the schematic of the recuperative-type burner. The high air
preheat (650°C), the specially fine grinding (90%
-------
164
CLEM COMBUSTION OF COAL
nir
coal
dust
. .. .^T"
^ ji.t.
*v-
^c-^v
t'*_V!.\J V
- j_a .
i secondary
mm
Fig. 22. Schematic of Pilot Burner for Low Grade Pulverized Coal (Ref.
Ash 4
removal
Grit
retiring
Economiser
PRATT
-DANIEL
Precipitator
Circulating
Air pumps
Fig. 23, BCSL Fluidized Combustion Water-Tube Boiler.
-------
COAL COMBUSTION SYSTEMS 155
fluidizing air will rise through the bed in the form of bubbles27. The
coal particles are larger than in pulverized combustion (mm size range)
and their combustion is, normally, external-diffusion limited. At the
low temperature limit of operation, however, at around 650°C, the slow-
est step in the overall reaction mechanism is most likely the desorption
step in the chemisorption reaction at the coal surface. The oxygen
concentration in the dense phase is low (l-3%), and an important factor
in the rate of burning of the coal particle is the rate at which oxygen
will diffuse from the bubble to the dense phase. The basic considera-
tions and methods of calculation of combustion of coal particles have
been established by Avedesian and Davidson28. Extensive theoretical and
experimental research to develop further this theory is in progress at
MIT.
Figure 2\ shows the reduction of sulphur emission that can be
achieved in fluidized combustion by the use of sorbents, limestone of
dolomite in the bed. The NO emission determined in a number of pilot
plant experiments and plotted as a function of bed temperature is given
in Figure 25- Recent studies of the formation and destruction of WO in
fluidized combustion processes have shown significant potential for the
further reduction of WO emission29.
There are several fluidized combustors in industrial operation and
their present development can be considered to have reached the stage
for application in industrial process heat or power generation systems.
The scaling up of fluidized combustion to utility size boilers requires
some further research and development: an improved coal feed system
would have to be developed which does not require one feed point for
every MW thermal input, and further studies would have to be carried out
to ensure that superheater tubes immersed in the fluidized bed have com-
parable life time to steam generating tubes in the bed. Figure 26 shows
the variation of the enthalpy composition of steam (i.e., liquid heat,
latent heat and superheat) as a function of steam pressure. It can be
seen that as the steam pressure rises, the latent heat proportion of the
total enthalpy decreases. As the steam pressure rises beyond a value
(>1500 psi), it becomes difficult to maintain a balance between the heat
that has to be extracted from the bed to maintain its temperature below
900°C and the latent heat proportion of the steam generated in the bed;
additional bed cooling by superheater tubes becomes necessary. This
explains the need for further research on the corrosion-erosion of
superheater tubes immersed in the bed, before fluidized combustion can
be safely applied to large utility boilers.
CONCLUSIONS
An attempt has been made to present coal combustion systems from
the point of view of the physical-chemical process that the coal and
accompanying mineral matter undergo during combustion and by considering
the requirements of plant and power generating cycle efficiency. It was
considered that the development of coal combustion systems was closely
tied to that of steam raising plant, that the transition from stoker
firing to pulverized coal combustion occurred because the latter could
be scaled up more easily to larger units, produced higher combustion
efficiencies at lower excess air and could make use of higher air preheat
which in turn was thermodynamically advantageous in power generating cycles.
-------
166
CLEM COMBUSTION OF COAL
IUO
9O
00
7O
o5 6O
•^-*
.1 5O
5 4O
61 30
i; 100
H- 9O
•3
in
BO
7O
An
1 — • i ~^5^
o /b^V-"-
/S*-x
X /Pressurised -
O Unpressunsed
/
' / 0 DOLOMITE'
I i i i
LIMESTONE ^--
^x
Unpressurised''
^X/ rf
/ "reSSUr^/^-
,/ , ^JL .
IUU
9O
^ 8O
^100
.0
S 90
L.
^ 8O
_, i-><^
00
70
•______' ..^i__ — Q-
Pressurised
Unpressurised
• ""
-------
COAL COMBUSTION SYSTEMS
167
\
a a §
1" I 1
8 e
k cc
i
il/kg
800
600
400
200
ft w — —
Superheating
•~—
/
x
.
•.
Eva
*^*
.
—~.
oratio
--
•^
,-•
Preheating
• — ..
•^"*
N
y
-
C.R
50
100 ISO
200
Fig. 26. Enthalpy Pressure Diagram of Water.
Fluidized coal combustion, a new advanced system, has the added
advantage of being capable of retaining sulphur in the bed and it prom-
ises also significantly lower emission of NOX. Due to its lower operat-
ing temperature, fewer or no submicron particulates are emitted from
fluidized combustors. This system which is ready for application for
smaller industrial combustion systems requires further concerted research
for its application to utility-size power generators and for combined gas
turbine-steam cycle operation.
REFERENCES
1. Anthony, D.B., J.B. Howard, H.C. Hottel, and H.P. Meissner: "Rapid
Devolatilization of Pulverized Coal," Fifteenth Symposium (Int'l)
on Combustion, p. 1303, Combustion Institute, Pittsburgh, PA (1975).
2. Howard, J.B., and R.H. Essenhigh: "Pyrolysis of Coal Particles in
Pulverized Fuel Flames," Ind. Eng. Chem. Process Design Dev. , 6_,
7^ (1967).
3. Beer, J.M., and R.H. Essenhigh: Nature, 187, 1106 (i960).
k. Gumz, W. : Kurzes Handbuch der Brennstoff und Feuerungstechnik
Springer - Berlin, p. U92 (1962).
5. Beer, J.M.: "Combustion Chamber Design for Stoker Firing and Low
Grade Coals Based on Both Laboratory Light Model Experiments and
Large Scale Trials," (in Hungarian), Magyar Energiagazdasag, No. 9,
pp. 306-315 (1951).
6. Beer, J.M.: "Combustion Research for Industrial and Power Station
Boilers," (in Hungarian), Magyar Technika. No. U, pp. 210-216 (195M
-------
16g CLEAN COMBUSTION OF COAL
7. Tzukhanova, O.A.: J. Techn. Physics 9 (1939) Vol. U, pp. 295-30**,
Leningrad, USSR.
8. Chukhanov, Z.F., and M.K. Grodzovskii, as cited in L.N. Khitrin:
Physics of Combustion and Explosion Israel Program for Scientific
Translations, Jerusalem (1962).
9. Wertaneister, H. : VDI Berichtsheft , Trier (193*0, pp. 7^-79-
10. Zagon, P. , as cited by Beer, in Ref. 6, this paper.
11. Beer, J.M. : "Some Current Trends in Combustion Research in
Hungary," University of Sheffield Fuel Soc . , J. pp. 1-12 (1958).
12. Nusselt, W. : Die Ferbrennung in der Kohlenstaubfeverung Z. V.D.I.
68, No. 6, pp. 12*1-128 (192*0.
13. Essenhigh, R.H. , and J. Csaba: Ninth Symposium (int'l) on Combus-
tion, Combustion Institute, Pittsburgh, PA, pp. 111-125 (1963).
1U. Dolezal, R. : Large Boiler Furnaces, Fuel and Energy Sci. Series,
Ed. J.M. Beer, Elsevier Publishing Company, England (1968).
15. Beer, J.M.: "The Effect of Fineness and Recirculation on the Com-
bustion of Low-Volatile Pulverized Coal," J. Inst. F. , pp. 286-313
(July
16. Badzioch, S. : "Thermal Decomposition," Chapter U in M.A. Field,
et al. : Combustion of Pulverized Coal, BCURA, England (1967).
17. Anthony, D.B., and J.B. Howard: "Coal Devolatilization and Hydro-
gasification," AIChE Journal, 22_, 625 (1976).
18. Kobayashi , H. , J.B. Howard, and A.F. Sarofim: "Coal Devolatiliza-
tion at High Temperatures," Sixteenth Symposium (int'l) on Combus-
tion, The Combustion Institute, Pittsburgh, PA, pp. U11-U25 (1977).
19. Pohl, J.H. , and A.F. Sarofim: "Devolatilization and Oxidation of
Coal Nitrogen," ibid, pp. ^91-501.
20. Song, Y.H., J.M. Beer, and A.F. Sarofim: Fate of Fuel Nitrogen Dur-
ing Pyrolysis and Oxidation, Second Symposium on Stationary Source
Combustion, EPA, New Orleans (August 1977).
21. Mulcahy, M.F.R., and I.W. Smith: Kinetics of Combustion of Pulver-
ized Fuel: A Review of Theory and Experiment, Rev. Pure and Appl.
Chem., 19_, 8l (1969).
22. Beer, J.M. : "Choice of Coal Preparation Systems for Pulverized
Fuel Firing," Engineering and Boiler House Review (September 1-6,
1959).
23. Beer, J.M. : "Some Current Trends in Combustion Research in
Hungary," University of Sheffield Fuel Soc., Journal, pp. 1-12 (1958).
-------
COAL COMBUSTION SYSTEMS 169
2U. Beer, J.M., T. Csorba, and J. Csaba: "Pulverized Fuel Burner for
Burning Low Grade Coal," (in Hungarian), Magyar Energiagazdasag,
No. 2, pp. 61-68 (1956). Translated into English by the Ministry
of Pover.
25- Skinner, D.J.: Fluidized Combustion of Coal, Mills and Bonn Mono-
graph CE/3 (1971).
26. Robinson, E.B., R.D. Glenn, S. Ehrlich, J.W. Bishop, and J.S.
Gordon: EPA Contract CPA 70-10, PB210 828 (February 1972).
27. Davidson, J.F., and D. Harrison: Fluidized Particles, Cambridge
University Press (1973).
28. Avedesian, M.M., and J.F. Davidson: Trans. Inst. Chem. Eng.,
£L, 121 (1973) London.
29. Beer, J.M.: "Fluidized Combustion of Coal, A Review" (invited Paper)
Ibid., Paper No. 33.
-------
170 CLEAN COMBUSTION OF COAL
-------
171
POLLUTANT FORMATION DURING COAL COMBUSTION
M. P. Heap and R. Gershman
Energy and Environmental Research Corporation
Irvine, California
1.0 INTRODUCTION
Increased coal utilization is undoubtedly the key to decreased
reliance upon imported crude oil and energy independence. However, the
methods which will be used in the next three decades to increase con-
siderably the amount of energy being supplied from coal are open to
question. Economic and environmental considerations will be factors in
the choice of several available energy conversion alternatives since
there is a need -to utilize coal in an efficient, but environmentally
acceptable manner. Coal contains constituents which have the potential
to form pollutants which might be emitted into the atmosphere during the
energy conversion process. These constituents are:
• Sulfur which forms SO and SO., under oxidizing conditions.
• Mineral matter which results from the biological, chemical
and environmental conditions existing during the formation
of coal and is vaporized or converted to ash or slag during
combustion. In a study of 65 Illinois coals Gluskoter(l)
found that the mineral content ranged from 9.4 to 22.3 per-
cent. Coal contains almost all the elements in the periodic
table.
• Nitrogen which can also be oxidized to give nitrogen oxides
(NOX).
The above compounds, together with the products of incomplete combustion
(CO, hydrocarbons, carbonaceous particulates and large molecular weight
organic molecules), are potential pollutants resulting from coal utili-
zation. Economics of pollution control are extremely important with
regard to the increased utilization of coal as a fuel.
For one of the major pollutants produced from coal, sulfur, the
choice is one of removal from the fuel or from the products of combus-
tion. Removal from the fuel involves the production of a liquid or
gaseous fuel or a clean, solid fuel. Coal gasification converts the
majority of the sulfur to hydrogen sulfide which is then scrubbed from
the fuel gas. Neglecting absorbant efficiency, the reduced gas volume
ensures that the task of removing ^S from the fuel gas is easier than
the removal of S02 from the combustion products. Fluidized bed combus-
tion of coal offers the possibility to burn coal as mined and yet con-
trol sulfur emissions by in-bed absorption of S02. Although coal gasi-
fication is not a new technology, its application to today's needs is
-------
172 CLEM COMBUSTION OF COAL
far from straightforward and there are no operational fluidized bed
combustors on a commercial scale. It appears that direct coal combus-
tion in conventional plants provides the greatest opportunity for a
rapid increase in coal utilization. The major problem associated with
direct coal combustion is the control of three atmospheric pollutants —
SOX, NOX and particulate matter.
This paper is primarily concerned with the formation of and control
of nitrogen oxides since this is one pollutant which can be controlled
by modifying the combustion process. Coal-fired utility and industrial
boilers account for 37 percent of the stationary source emissions of NOjj,
and gas- and oil-fired boilers contribute a further 16 percent to the
total. The existing New Source Performance Standard (NSPS) for large
coal-fired boilers is 0.7 Ibs NOX/106 Btu which is 3.5 and 2.3 times as
high as the NSPS for gas- and oil-firing respectively. The increased
emphasis on use of coal in large boilers will increase total NOX emis-
sions unless appropriate control technology is applied.
This paper does not provide a complete literature survey on the
formation of nitrogen oxides in coal-fired systems. Only those refer-
ences have been given which are necessary to illustrate a point or which
contain controversial data. Before the fundamental aspects of NOX for-
mation are discussed, the state-of-the-art of NOX control for coal-
fired boilers will be reviewed. Finally, the prospects for advances in
NOX control technology for pulverized coal-fired boilers are discussed.
2.0 REVIEW OF NOX CONTROL TECHNOLOGY
The reader is referred to several reports in the literature(2>3,4,5)
which discuss the application of NOX control technology to coal-fired
boilers. The existing state-of-the-art is reviewed below:
• Firing Type. Coal can be burned crushed in cyclone furnaces,
pulverized in tangential or wall-fired boilers, or spread as
lump coal in stoker-fired boilers. NOX emissions are normally
highest for cyclone-firing, and wall-fired boilers characteris-
tically emit more NOX than tangential firing. Spreader
stokers, which might have 50 percent of the heat released in
the suspension phase, generally give higher emissions than
underfeed stokers.
• Operational Parameters. Reduced load or reduced excess air
operation tends to reduce emissions from all boiler types.
• Flue Gas Recirculation. The addition of cooled combustion
products to the combustion air is not an effective method of
controlling NOX emissions from coal-fed boilers. This can
be attributed to the predominance of fuel NOX in the total
emission from coal-fired units.
• Staged Heat Release. Staged heat release accomplished either
by the use of overfire air ports or removing burners from
service is an effective method of NOX control.
-------
POLLUTANT FORMATION AND CONTROL 173
• Burner Redesign. Redesign of the burner to decease the rate
of fuel/air mixing for wall-fired burners reduces NOX forma-
tion in pulverized coal flames.
3.0 NOX FORMATION DURING COAL COMBUSTION
The nitrogen oxides emitted by coal-fired boilers are formed by the
oxidation of molecular nitrogen (thermal NOX) and nitrogen which is
chemically bound in the fuel (fuel NOX) . Small-scale studies carried
out by Pershing and Wendt(^) indicate that for pulverized coal-fired
combustors the major portion of the emission can be attributed to fuel
NOX. However, under controlled conditions it might be expected that
the thermal NOX fraction will become more significant. The major por-
tion of this discussion will be restricted to NOX formation in pulver-
ized coal combustion since this is the predominant method of coal
utilization by direct combustion.
3. 1 Coal Chemistry
In the restricted sense of this review coal chemistry refers to
those changes which occur during the period of particle heating. Parti-
cles in pulverized coal flames are subjected to heating rates between
10,000 K/sec and 100,000 K/sec causing volatiles to be driven from the
coal. The particles might swell and off gas as a jet, they might explode
or the volatilization process might occur with little physical change
in the particles. Blair et al(?) and Pohl and Sarofim(S) have shown
that the total quantity of volatiles produced is a strong function of
pyrolysis temperature.
Blair et al(?) carried out a series of controlled pyrolysis experi-
ments with single particles and found that:
• Total nitrogen volatilized is a more sensitive function of
pyrolysis temperature than is total mass pyrolyzed.
• Only 20 percent of the volatile fuel nitrogen appears to be
released from the coal as light gases (HCN,
This latter conclusion may have less significance because even though
the initial nitrogen fragments may be heavy molecular weight compounds
it is probable that these compounds undergo pyrolysis and that the
nitrogen is converted to either HCN or NO depending upon the availa-
bility of oxygen. Of much more significance is the fact that the
initial volatile fractions are low in nitrogen content. Thus these
volatile gases have the opportunity to deplete the available oxygen
before the major portion of the nitrogen-containing volatiles are
evolved .
Studies (8) in a laminar flow furnace indicate that residence times
of approximately one second are necessary at temperatures in excess of
1500K to reduce the nitrogen content of the char to less than 40 per-
cent of the value of the original coal. Oxidation experiments in the
same furnace clearly indicated that the conversion of char nitrogen to
NO occurs with a lower efficiency than the conversion of coal nitrogen
under comparable conditions of temperature and equivalence ratio.
-------
174 CLEAN COMBUSTION OF COAL
In summary it appears that:
• At flame temperatures pyrolysis kinetics are sufficiently
fast to suggest that in the absence of oxygen HCN would be
the primary fuel nitrogen product.
• Increased temperatures decrease char nitrogen content.
• Initial volatile fractions are low in nitrogen species.
• The efficiency of char nitrogen conversion to NO is lower than
that of the parent coal nitrogen to NO.
3.2 Homogeneous Gas Phase Reactions
The details of fuel NO kinetics are outside the scope of this
paper other than to qualitatively describe those conditions which
minimize fuel NO formation. Reaction zone stoichiometry appears to
have the strongest influence on fuel NO formation, the conversion of
fuel nitrogen to NO decreases rapidly as the mixture becomes fuel-rich.
This can be accounted for by considering two competing reaction paths:
XN + Ox -> NO + . . . . (1)
which predominates in oxygen-rich mixtures and
XN + YN -*• N2 + .... (2)
which is faster in fuel-rich mixtures. There is also considerable
evidence to suggest that NO is a necessary nitrogen specie for N£ pro-
duction. Another type of reaction which may be very significant in
pulverized coal flames is the reduction of NO by hydrocarbon specie,
e.g.,
NO + CH -> XN + . . . . (3)
thus allowing the production of nitrogen via reaction (2).
Premixed gaseous studies indicate that reaction zone temperature is
not a significant factor in fuel NO production. However, the conversion
of fuel nitrogen to NO appears to be greater in hydrogen flames than in
hydrocarbon flames. The presence of sulfur appears to enhance the pro-
duction of fuel NO in rich, premixed gaseous flames and two-phase
turbulent diffusion flames suggesting that sulfur specie might inter-
fere in the competing paths represented by reactions (1) and (2).
-------
POLLUTANT FORMATION AND CONTROL 175
3.3 Coal Char Reactions
Char is that solid which remains after the volatile fractions have
been driven from the coal particles. Char contains nitrogen; however,
as noted earlier, the efficiency of char nitrogen conversion is less
than that of the parent coal nitrogen. This can be attributed to the
fact that the stagnant boundary layer surrounding the char particles is
probably reducing. Thus even if the char nitrogen were to produce
nitric oxide at the particle surface there is the possibility that it
could be reduced to N2 as it diffuses through the boundary layer. Sur-
prisingly little concrete information is available concerning the pro-
duction of NO from char nitrogen. Theoretical calculations carried out
by Wendt and Shulze(9) suggest that the production of NO from char
nitrogen is most sensitive to free stream oxygen concentration and
temperature.
In a series of experiments designed to define probe conditions
necessary for the measurement of NO in pulverized coal flames, Heap
et al'lO) showed that NO in nitrogen could be reduced by coal particles
at temperatures above 200°C. Beer and co-workers (H) have shown that
under fluidized bed conditions NO can also be reduced by coal char.
Sarofim(12) has recently reported that coal char suspensions will also
reduce NO to N2 in the absence of oxygen. Consequently, it appears that
NO produced in the early stages of heat release could be reduced by char
in the burnout regions. Further research effort is necessary to ascer-
tain whether heterogeneous NO reduction by char is an important phenom-
enon in pulverized coal flames.
3.4 Fuel/Air Contacting
Pulverized coal is burned in an air suspension. Fuel/air mixing,
entrainment of recirculated combustion products, particle heating,
volatile evolution, volatile combustion and char oxidation occur almost
simultaneously. The major factor controlling the emission of nitrogen
oxides is the fuel/air contacting process since this will affect:
the rate of particle heating and the final particle tempera-
ture which will determine the proportion of the bound nitrogen
remaining in the char;
the atmosphere under which the volatile compounds react. If
this is oxygen-rich then it would be expected to maximize
NO production;
the residence time in any rich zone. The degree of conversion
of fuel nitrogen fragments to N2 or the reduction of NO to N£
will depend upon residence time in the rich zone; and
the rate of heat release and reaction zone quench rate which
will affect the production of thermal NO.
Heap et al(' carried out a series of subscale experiments to esta-
blish the influence of fuel/air contacting on NOX production in pulver-
ized coal flames by varying burner parameters. These investigations
demonstrated that NOX emissions were very sensitive to the fuel/air
-------
176 CLEAN COMBUSTION OF COAL
mixing rates. Combinations of fuel injection method and axial and
tangential velocity distributions produced the two major flame types
sketched in Figure 1. The top sketch represents the conditions in a
near-field dominated flame in which the coal and air are rapidly mixed.
Maximum particle heating rates and stable ignition are provided by the
entrainment of hot combustion products from an axial recirculation zone.
These conditions are typical of flames in uncontrolled wall-fired boilers
and maximize fuel NOX formation since ample oxygen is available in the
regions of coal volatilization. With the same coal, excess air level
and air preheat temperature NO levels are reduced from approximately
800 ppm to 200 ppm by producing a high fineness ratio diffusion flame
with a minimum oxygen level associated with the fuel jet. Provided
there is a stable ignition front around the fuel jet the volatile frac-
tions will remain in an oxygen-deficient atmosphere providing the oppor-
tunity for fuel nitrogen fragments to form N2- The type of fuel/air
contacting producing the high fineness ratio diffusion flame has many
similarities with the processes occurring in corner-fired boilers whose
NOX emission levels are interestingly lower than uncontrolled emissions
from wall-fired units.
Figure 2 taken from the work of Pershing and Wendt^") provides
experimental evidence to support the hypothesis presented above. These
workers substituted argon for molecular'nitrogen in the combustion "air,"
and thus the total emission can be attributed to fuel nitrogen conver-
sion. Two different injectors were used which produced a near-field
dominated flame (divergent injector) and a high fineness ratio diffusion
flame (axial injector) and it can be seen that the fuel NO production is
much less in the case of the high fineness ratio flame.
4.0 ADVANCED CONTROL TECHNOLOGY DEVELOPMENT
The stated goal of the U.S. Environmental Protection Agency is the
development and demonstration of maximum NOX control for stationary
sources. This strategy is based upon the premise that stringent con-
trols cannot be applied to mobile sources without incurring serious
economic penalties. Consequently, if air quality is not to be degraded,
more restrictive control must be imposed upon stationary sources. Con-
version to coal or coal-derived liquids will tend to emphasize the
significance of stationary source emissions, thus providing impetus for
the development of advanced NOX control technology.
The use of off-stoichiometric firing in combination with reduced
load, low excess air operation and a modified burner design reduced
NOX levels to 200 ppm in an opposed wall-fired 270 MW boiler(14).
Gershman et al^5' have reported on the development of a low NOX proto-
type coal burner which is being sponsored by the EPA. This burner con-
cept was based upon the subscale work of Heap et al(16) and relies upon
distributed air addition which is explained schematically in the sketch
shown in Figure 3. The air supply is divided into three streams. The
distribution of axial and tangential velocity at the burner throat
creates a high residence time, hot recirculation zone into which the
coal is injected. The third air stream is injected at the periphery of
the burner exit to provide an oxidizing envelope to prevent impingement
of corrosive reducing gases on the furnace walls. The distributed air
-------
POLLUTANT FORMATION AND CONTROL
177
Axial Recircul
Zone
Hot Products - 1950 K
~~^9
ation^/\
Near Field Dominated
Air + Recirculated Products
Early Volatiles Deplete CU
High Fineness Ratio Diffusion Flame
Figure 1. Major Flame Types
-------
—I
CO
1200
1000
800
o
4J
o 600
400
200
I I I I I I
Divergent Injector
Thermal NO
Fuel NO
O Air
A Ar/02/C02
I I I I
I I
\ \ I I I I
Axial Injector
Thermal NO -^ _
Fuel NO
I I I I I I I
.0 1.1 1.2 1.3 1.0 1.1
Stoichiometric Ratio
1.2
1.3
o
tr1
£
O
i
en
O
23
O
g
f
Figure 2. Comparison of Thermal and Fuel NO Production for
Two Basic Flame Types (After Pershing and Wendt^/)
-------
Tertiary Air
Divided Secondary
Air Stream
Coal + Primary Air
Progressive
Leaning Out
Burnout Zone
O
tr>
f
o
I
(-3
M
O
O
O
Figure 3. Distributed Fuel/Air Mixing Concept
-------
180 CLEAN COMBUSTION OF COAL
addition ensures that the coal first reacts .in a rich region (
-------
POLLUTANT FORMATION AND CONTROL
181
800
600
CVJ
o
o
4->
to
400
Q.
O
200
Small-scale Hot Tunnel
5 x 106 Btu/hr
O Low Swirl
O Medium Swirl
A High Swirl
Water-Cooled Simulator
50 x 106 Btu/hr
Medium Swirl
— 35% Excess Air
55% Excess Air
100
I
0.8 1.0 1.2 1.4 1.6
Primary Zone Equivalence Ratio
1.8
Figure 4. Test Results
-------
182 CLEAN COMBUSTION OF COAL
REFERENCES
1. Gluskoter, H.J., Occurrence of Mineral Matter in Coal: An Intro-
duction, Paper presented at The International Conference on Ash
Deposits and Corrosion from Impurities in Combustion Gases,
Henniker, New Hampshire, July 1977.
2. Martin, G.B. and Bowen, J.S., Development of Combustion Modifica-
tion Technology for Stationary Source NOX Control: Health,
Environmental Effects, and Control Technology of Energy Use, EPA
Report 600/7-76-002.
3. Crawford, A.R., et al, Field Testing: Application of Combustion
Modifications to Control NOX Emissions from Utility Boilers, EPA
Report 650/2-74-066, June 1974.
4. The Proceedings of the NOX Control Technology Seminar, EPRI Special
Report, February 1976.
5. Pershing, D.W., Ph.D. Dissertation, Department of Chemical Engi-
neering University of Arizona, Tucson, 1976.
6. Pershing, D.W. and Wendt, J.O.L., Pulverized Coal Combustion: The
Influence of Flame Temperature and Fuel Composition on Thermal and
Fuel HOX, presented to the 16th Symposium on Combustion, Cambridge,
Massachusetts, August, 1976.
7. Blair, D.W., Wendt, J.O.L. and Bartok, W., Evolution of Nitrogen
and Other Species During Controlled Pyrolysis of Coal. Presented
at 16th Symposium (International) on Combustion, Cambridge, Mass.,
August 1976 (to be published).
8. Pohl, J.H. and Sarofim, A.F., Devolatilization and Oxidation of
Coal Nitrogen. Presented at 16th Symposium (International) on
Combustion, Cambridge, Mass., August 1976 (to be published).
9. Wendt, J.O.L. and Schulze, O.E., On the Fate of Fuel Nitrogen
During Coal Char Combustion, AIChE Journal, 22, 102 (1976).
10. Heap, M.-P.,et al, The Development of Combustion System Design
Criteria for the Control of Nitrogen Oxide Emissions from Heavy
Oil and Coal-Fired Furnaces, Vol. II, EPA Report 600/2-76-061B.
11. Gibbs, B.M., Pereira, F.J. and Beer, J.M., The Influence of Air
Staging on NO Emission from a Fluidized Bed Coal Combustor, pre-
sented to the 16th Symposium (International) on Combustion,
Cambridge, Mass., August 1976.
-------
POLLUTANT FORMATION AND CONTROL 183
13. Heap, M.P., et al, Burner Criteria for NOX Control, Vol. I, EPA
Report 600/2-76-061A, March 1976.
14. Barsin, J.A., Dual Register Burner as NOX Control Device, Proceed-
ings of the EPRI NOX Control Technology Seminar, San Francisco,
February 1976.
15. Gershman, R., Heap, M.P. and Tyson, T.J., Design and Scale-Up of
Low Emission Burners for Industrial and Utility Boilers, Pro-
ceedings of the Second Stationary Source Combustion Symposium,
Vol.,V, EPA Report 66/7-77-073E.
16. Heap, M.P., et al, The Optimization of Burner Design Parameters to
Control NOX Formation in Pulverized Coal Flames, Proceedings of
the Stationary Source Combustion Symposium, Vol. II, EPA Report
600/1-76-152B, June 1976.
17. Brown, R.A., et al, Investigation of Staging Parameter for NOX
Control in Both Wall and Tangentially Coal-Fired Boilers, Pro-
ceedings of the Second Stationary Source Combustion Sumposium,
Vol. Ill, EPA Report 600/7-77-073C, July 1977.
-------
184 CLEAN COMBUSTION OF COAL
-------
185
THE DUAL REGISTER PULVERIZED COAL BURNER: FIELD TEST RESULTS
by
E. J. Campobenedetto
Development Engineer
Combustion Systems
Fossil Power Generation Division
Babcock & Wilcox Company
Power Generation Group
Barberton, Ohio
ABSTRACT
The Federal Environmental Protection Agency's "Standards of
Performance for New Stationary Sources" limits the emissions of oxides
of nitrogen for pulverized coals (other than lignite) to a maximum
level of 0.7 Ibs N0£ per million Btu heat input to the boiler. To
comply with the above standard, it was necessary to develop a new pul-
verized coal burner because the existing Babcock & Wilcox (B&W) stan-
dard high turbulence burner was not capable of producing NOX emissions
within the above limit.
B&W's judgmental criteria for the success of the new burner
development was that the burner/furnace system be capable of reducing
NOX generation to levels below the EPA NOX limit while maintaining
carbon utilization equal to or higher than the former standard burner.
After extensive laboratory and field testing, the B&W Dual
Register Burner was developed. The data obtained to date has shown
that this burner is capable of operating below the existing NOX limit
without increasing unburned carbon and excess air levels over those
experienced with the previous burner.
This paper discusses the NOX results obtained to date utilizing
the Dual Register Burner.
THE DUAL REGISTER BURNER
The Dual Register Pulverized Coal Fired Burner (Figure 1) is a
limited turbulence, controlled diffusion flame burner designed to mini-
mize the amount of fuel and air mixed at the burner to that required
to obtain ignition and sustain stable combustion of the coal. A ven-
turi mixing device is located in the coal nozzle to provide a homo-
geneous coal/primary air mixture at the burner. The remainder of the
combustion air (secondary air) is introduced through two concentric
air zones which surround the coal nozzle. The air flow to each air
zone is independently controlled through inner and outer air zone
registers. Adjustable spin vanes are located in the inner air zone to
provide varying degrees of swirl to the inner air to control coal/air
mixing during the combustion process.
-------
186
CLEAN COMBUSTION OP COAL
Primary
air & coal
Adjustable air vane
and register drives
«^-*
1
1 1
t=
Hr-
~"~"~— — ~ ~ — ___ — , i
s
>fk
econdary a
/ i
nj
L
ir i
*r
- Windbox -
Dual register burner
Figure 1
By controlling the mixing of the coal and air, the combustion
process is initiated at the burner throat and the zone of completion
can be varied in the furnace chamber. This method of delayed combus-
tion reduces the high temperature zones at each burner. Thus the
peak temperature in the furnace is lowered, minimizing the thermal
conversion of combustion air nitrogen to HOX. Also, through controlled
fuel and air mixing, the oxygen availability is minimized during the
process reducing fuel nitrogen conversion.
The Dual Register Burner lowers NOX by delaying combustion and is
not a staging device. Previous work on pulverized coal firing has
shown that two-stage combustion is the most effective method of NOX
reduction. However, in reviewing overall unit performance, the Dual
Register Burner has the following benefits over staging techniques:
l) The furnace is maintained in an oxidizing environment to
minimize slagging and reduce the potential for furnace wall corrosion
when burning high sulfur bituminous coal.
2) More complete carbon utilization through better air/coal
mixing in the furnace.
3) Lower oxygen levels required with total combustion air ad
mitted through the burners rather than above the burner zone.
-------
PULVERIZED COAL BURNER
187
In addition to the Dual Register Burner, the burner windbox
design was changed to provide air flow control on a per pulverizer basis
(Figure 2). This compartmented windbox permits careful control of fuel
Compartmented
windbox
Furnace
observation doors
Burner secondary
air control dampers
Burner secondary
air foils
Pulverizer-burner system
Figure 2
and air flows to each burner group. This permits operation with lower
total excess air while maintaining an oxidizing atmosphere around each
burner. The results are lower UOX emissions and increased flexibility
in the introduction of excess air to the burner zone for both slagging
and NOX control.
FIELD TEST DATA
To date, three (3) Babcock & Wilcox pulverized coal-fired boilers
equipped with the Dual Register Burner have been tested to determine
TTOX emissions. The program for each unit consisted of a series of WOX
tests performed at various operating conditions over the load range.
When operating the units at normal excess air required for combustion,
the NOX emissions were well below the present EPA limit of 0.7 Ibs/lOo
Btu. Figure 3 is a summary of the full load NOX emissions under normal
-------
188
CLEAN COMBUSTION OF COAL
NOx-lbN02/106 Btu
1.0 |-
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
Circular burner
250
Unit capacity, MW
EPA NOx
emission unit
700
Comparison of NOx emissions
circular burner vs dual register burner
Figure 3
-------
PULVERIZED COAL BURNER 189
conditions (15 - 17$ excess air to the burners). The emissions range
from 0.33 to O.H7 lias/million Btu for the three (3) units, which vary
in capacity from 90MW to TOO MW. Also, depicted in Figure 3 are the
NOX emissions obtained from similar units equipped with the high tur-
bulence Circular Burner previously supplied by Babcock & Wilcox.
The direct data comparison indicates that a k^ to 6d% reduction
in NOX is achievable through burner design. The majority of the reduc-
tion is attributable to controlled air/coal mixing in the furnace cham-
ber. Through controlled mixing at the burner, the peak flame tempera-
ture is reduced thus minimizing the thermal NOX formation at the burner.
In addition, the controlled mixing decreases the oxygen available to
react with the nitrogen chemically bound in the fuel.
The type of coal utilized also affects KOX formation. The 90 and
TOO MW units burn a sub-bituminous type coal while the 250MW unit burns
a bituminous coal. A comparison of typical analysis for these two (2)
coals (Figure k] indicates why WOX emissions are lower when firing
Bituminous Sub-Bituminous
Total moisture, % 8.5 30.6
Proximate analysis, % (dry)
Volatile matter 30.8 45.5
Fixed carbon 54.8 47.5
Ash 14.4 7.0
Ultimate analysis, % (dry)
Carbon 71.8 69.2
Hydrogen 4.9 5.0
Nitrogen 1.0 0.8
Sulfur 1.6 0.7
Ash 14.4 7.0
Oxygen 6.3 17.3
HHV, Btu/lb (dry) 13,060 11,950
Nitrogen, lb/106 Btu 0.77 0.67
Typical coal analyses
Figure 4
sub-bituminous coal. The higher moisture content of the sub-bituminous
fuel (30.6$ versus 8.5$) provides for lower thermal NOX generation be-
cause of reduced furnace zone temperatures due to the greater heat of
vaporization required. The nitrogen content of the coal also affects
NOX generation. Comparing the nitrogen content on a Ibs. nitrogen/
million Btu heat input basis (thereby eliminating variations in heating
value), the fuel-contained nitrogen introduced to the furnace is 13%
lower with sub-bituminous coal (0.67 versus 0.77 Ibs N/million Btu).
-------
190
CLEAN COMBUSTION OF COAL
With less nitrogen available, the contribution of fuel bound nitrogen
conversion is reduced.
In addition to variations in coal supply, the two principal param-
eters affecting NOX formation are combustion zone temperature and oxy-
gen availability. NOX versus heat release rate in the burner zone of
the furnace (HA/SC) for various excess air levels is plotted in Figure
5. The correlations were based on actual field data from the 250MW
NOx - ppm v at 3% 02
800
700
600
500
400
300
200
100
Circular burner
base line data
at 115% air'
to burners
I
I
I
High excess
air, 45%
Moderate excess
air, 30%
Normal excess
air, 15%
Low excess
air, 7%
100 200 300 400
HA/Sc - MBtu/hr-ft2
500
600
Field test results
Figure 5
-------
PULVERIZED COAL BURNER 191
bituminous fired unit. Statistical analysis was used to develop the
linear best fit curves from the test data.
The effect of furnace zone temperature is illustrated by the in-
crease in NOx for higher heat release rates. The WOX increases 80 ppm
per 100 MBtu/hr-ft2- For a given unit, the larger the HA/SC, the higher
the temperature will be in the burner zone of the furnace. NOX genera-
tion attributable to thermal fixation will increase per the relationship.
Since the thermal WOx formation is extremely temperature sensitive at
high temperatures, the shaded areas on Figure 5 represent the degree
of uncertainty at the higher heat release rates. As the burner zone
temperature approaches the threshold temperature, the rate of thermal
WOX fixation rapidly increases.
The formation of HOX also requires oxygen availability and the
greater the excess oxygen, the higher the KOX generated. NOX emissions
versus excess air quantities ranging from 7% to ^5% is plotted in
Figure 5- Actual field data indicates NOX emissions increase by 20 -
25 ppm per 5% increase in excess air. Utilization of the Dual Register
Burner coupled with the compartmented windbox has provided proper con-
trol of air and fuel mixing which minimizes air requirements thus
limiting NOX formation.
SUMMARY
Results from the three (3) pulverized coal units tested to date
show the Dual Register Burner to be an effective tool for NOX emission
control. NOX levels ^0 to 50% lower than those achievable with the
high turbulence Circular Burner are attained through limited turbulence
Combustion. At the same time, carbon utilization has been maintained
at levels comparable to those obtained utilizing a high turbulence
Circular Burner. The Dual Register Burner provides low NOX emission
without sacrificing unit efficiency.
The data acquisition and analysis program as reported herein will
continue to provide data to update and improve future units. It is
expected that by the end of 1977 KOX and performance data will be
available from an additional four (U) units now in the final stages of
start-up. These tests will include units firing both bituminous and
sub-bituminous grade coals.
REFERENCE
1. Barsin, J.A. and Brackett, C.E., i:The Dual Register Pulverized
Coal Burner", presented at the Electric Power Research Institute NOX
Control Technology Seminar, San Francisco, California, February, 1976.
-------
192 CLEAN COMBUSTION OF COAL
-------
193
COAL-OIL MIXTURE COMBUSTION IN BOILERS - AN UPDATE
By
Sushil K. Batra
Manager of Energy Conversion Research, New England Power Service Co.
1. INTRODUCTION
As a result of unavailability and increased cost of fuel oil for
power plants, the nation faces large capital outlays in converting the
existing oil-fired steam generating capacities to coal burning. Also,
power plants that have been converted from coal to oil firing would be
converted "back to coal burning. At some plants, however, it may not be
practical to convert to coal burning due to lack of coal storage/coal
handling facilities or high capital cost and short remaining life of
the units. An alternative to the above could be to burn pulverized
coal-oil mixture (COM). This will be a near-term alternative to mini-
mize the expense and complexities of total conversion, while at the
same time reducing the dependency on foreign oil.
In general, by making use of coal-oil mixture, a utility will be
able to take advantage of relatively low prices and more stable supply
of this country's coal reserves. Coal-oil mixture will provide an in-
terim solution for short-term coal utilization, before new technologies
such as coal gasification, coal liquefaction, and solvent refined coal
become commercially available. Another extremely attractive feature
of the coal-oil mixture combustion concept is that an extensive new
technology doesn't need to be developed before application of coal-oil
mixture. The fuel could be made commercially available within two to
four years, with a minimum requirement for new equipment and technology.
The paper presented is devoted to: l) discussion of past studies
on coal-oil mixture stability; 2) a brief discussion on experimental
programs of combustion of coal-oil mixture at Pittsburgh Energy Re-
search Center (PERC) and Institute of Gas Technology (IGT); and 3} a
review of coal-oil mixture development programs at General Motors and
New England Power Service Company.
2'. HISTORICAL BACKGROUND
Since 18791, several attempts have been made to burn coal-oil
mixture . Combustion of coal-oil slurry was first tried in ocean-
going vessels and locomotives to save bunker space and minimize the
risk of fire in the storage of coal. In 19322, the Cunard liner
Scythia steamed from Liverpool to New York and back with one boiler
fired on a coal-oil mixture. However, the principal disadvantage of
such a fuel was the difficulty of maintaining the coal in suspension.
During World War II, the research in coal-oil mixture was again re-
sumed in Japan, Germany, and in this country, and burning of coal-oil
mixture was tried on a small scale. However, there has been no large-
scale application of this technology. More recently, several programs
-------
194
CLEM COMBUSTION OF COAL
have been initiated in this country funded by ERDA or by private funds
for development of coal-oil mixture technology.
3. COAL-OIL MIXTURE PREPARATION
Coal-oil mixture can be prepared by simply mixing finely ground
coal in No. 6 residual oil and providing continuous agitation to pre-
vent settling. Alternatively, a stabilized coal-oil mixture may be
prepared by adding a chemical additive to a mixture of pulverized coal
and No. 6 residual fuel oil with or without water. A stable mixture is
essential for good flame stability and also to keep the coal from set-
tling in tanks, heat exchangers, valves and pipelines. Evidently, both
of these systems have their own merits. Whereas it may be possible to
eliminate continuous agitation by use of chemical additives, current
cost of these additives is very high. Further research and development
needs to be done to investigate possibilities of less expensive addi-
tives that can be used for stabilization of coal-oil mixture.
4. STABILITY AND VISCOSITY OF COAL-OIL MIXTURE
The stability of coal-oil mixture has been studied by various re-
searchers in the past. In the laboratory work reported in 1944 by
Barkley^, et al., it was observed that the stability of the mixture in-
creased considerably with increase in percentage of coal in coal-oil
mixture. Also, the effect of temperature, as well as particle size on
stability of the mixture, was studied. Figure 1 shows the effect of
percentage of coal in mixture on stability at 77°F. It was observed
80
100
6>O 80
AGE, DAYS
too
120
Figure 1. Effect of percentage of coal in mixture on stability
at 77F, Ref. 3.
-------
COAL-OIL MIXTURE COMBUSTION
195
that the mixtures are relatively stable when the maximum particle size
does not exceed 230 mesh. Furthermore, the distribution of the size
of coal affected the stability of the mixture considerably. An addi-
tion of even small quantities of coarser materials in the coal-oil mix-
ture adversely affected the stability of the mixture.
The viscosity of the mixture increases markedly in the region of
40 to 45 percent of coal by weight (Figure 2). Also, the rapid rise
20 30 40 50
COAL IN SUSPENSION
Figure 2. Effect of percentage of coal in mix-
ture on viscosity at 77F, Ref. 3-
in viscosity in this range was found to be relatively independent of
the viscosity of the oil and temperature of the mixture. A rapid in-
crease in viscosity was also observed when the coal size was reduced
from 230 mesh to 230/325 mesh.
In view of the results reported above, from the point of view of
stability and viscosity of the mixture, the amount of coal in the mix-
ture should be limited to 40 percent, and the coal size to 98 percent
passing through 230 mesh screen. For practical purposes and to reduce
grinding and pumping costs, however, it is desirable to limit the per-
centage of coal in the coal-oil mixture to 30 percent, and coal size
-------
196
CLEAN COMBUSTION OF COAL
to 80 percent passing through 200 mesh.
5. COAL-OIL MIXTURE COMBUSTION DEMONSTRATIONS
The combustion of coal-oil mixture has recently been investigated
at Pittsburgh Energy Research Center and at General Motors facility in
Saginaw, Michigan. The later research was sponsored by ERDA, EPRI^,
General Motors, and more than 15 companies including New England Power
Service Company. Besides these investigations for application of coal-
oil mixture to package boilers, ERDA is partially funding the develop-
ment and demonstration of coal-oil mixture combustion at two utility
plants: namely, New England Power Company's 80 MW unit at Salem Harbor
Station; and City Public Service of San Antonio's 69 MW unit.
6. EXPERIMENTAL PROGRAM AT PERC
Coal-oil mixture combustion studies were conducted at Pittsburgh
Energy Research Center5 of ERDA in June 1975. The principal objectives
of these studies were to determine major problem areas while firing 20%
coal-oil mixture in a package boiler designed to burn oil or gas and to
determine corrosion and deposits after 1,000 hours of operation.
A new burner designed for No. 6 fuel oil was installed in a sur-
plus 100 hp 4 pass Cleaver-Brooks fire tube boiler and a coal-oil mix-
ture preparation and feed system was assembled. Figure 3 shows a sim-
STEAM x*-V*
T^}
/^v
SLURRY
TRANSFER PUMP
FEED WATER PUMP
Figure 3. Simplified flow diagram of the 100 H.P. coal-oil mix-
ture combustion test facility, Ref. 5-
plified flow diagram of the 100 H.P. unit. This system was intention-
ally assembled utilizing existing or readily available equipment to as-
-------
COAL-OIL MIXTURE COMBUSTION 197
certain the minimum equipment changes necessary to adapt an oil-fired
boiler to coal-oil mixture combustion. Modifications and/or improve-
ments were made in control and instrumentation as required during the
experimental program.
In general, the program concluded that no slagging was experienced
in the "boiler, and carbon combustion efficiency was more than 99%.
Boiler efficiency tests indicated that the efficiency obtained with
coal-oil mixture combustion was essentially the same as that obtained
with fuel oil. Some char deposition was experienced when firing at a
high rate due to an excessively wide spray pattern that produced flame
impingement on the relatively narrow combustion chamber. This became
progressively worse with time, as the orifices in the brass nozzle be-
came enlarged due to erosion and the atomization pattern was changed.
This problem was eliminated by modifying the combustion air diffuser to
obtain the desired flow pattern and fabricating nozzles of stainless
steel with a narrower spray pattern. Inspection of the boiler tubes
after the completion of the 1,000-hour test revealed only a light de-
posit on the internal tube surfaces. The carbon monoxide concentration
in the flue gas was slightly higher than that obtained with oil (260 vs.
160 ppm) but was acceptably low. Similarly, the NOX emissions were
also slightly higher reflecting the higher fuel bound nitrogen content
of the coal-oil slurry. Sections of tubes were taken from within the
passes of the boiler to determine the effect of corrosion/erosion. No
evidence of severe corrosion attack was observed. Also, there was no
evidence of inter-granular or subscale corrosion attack on any of the
test specimens.
7. COMBUSTION TESTS AT GENERAL MOTORS FACILITY
In 197-4, General Motors had a three-day, 10,000-gallon pilot test
using a mixture of No. 2 fuel oil and pulverized coal delivered to
their plant in Saginaw, Michigan. Some of the technical problems en-
countered in that demonstration were settling of the coal out of the
mixture, increased viscosity of the mixture, and problems associated
with coal-oil mixture preparation techniques. A development program
was initiated to solve these problems, and to burn coal-oil mixture
with No. 6 oil in existing oil-fired package boilers. In March of
1975, a consortium of companies and organizations was established.
This included U. S. Energy Research and Development Administration,
Electric Power Research Institute, several utilities including New
England Power Service Company, coal suppliers, and chemical manufactur-
ers. Preliminary work was conducted to gain specific information in
regard to fuel-oil and coal selection, fuel blending, fuel characteri-
zation, fuel atomization, and optimization of combustion before pro-
ceeding to extensive tests. The results in general indicated that
coal-oil mixture should perform as well as residual fuel oil using
conventional equipment.
8. COMBUSTION TESTS AT IGT
Detailed combustion tests were carried out at the Institute of Gas
Technology, Chicago, to identify possible operational problem areas and
define combustion characteristics for coal-oil mixture. These tests
involved the use of steam and air atomized burners on coal-oil mixture
-------
198 CLEM COMBUSTION OF COAL
in a rectangular furnace to investigate flame stability, geometry,
emissivity and NOX emissions. The results of these tests indicated
that the combustion flame produced from coal-oil mixture (30-50 weight
percent coal) using conventional atomizing nozzles was stable with good
geometry. The flame was also very similar to the flame produced by
straight No. 6 fuel oil. Good carbon burnout within the test chamber
was noted. As expected, the particulate emissions were higher than
for oil only. The coal-oil mixture's flame emissivities were the same
as or higher than the emissivity values obtained from residual fuel oil.
Furthermore, it was observed that the presence of 5$ water in the fuel
did not degrade combustion, and in fact appeared to enhance the atomi-
zation.
9. DEMONSTRATION AT GENERAL MOTORS FACILITY
Tests were carried on combustion of coal-oil mixture in a packaged
oil-fired boiler rated at 120,000 pounds per hour steam capacity
installed at the Chevrolet Nodular Iron Plant Power House in Saginaw,
Michigan. A summary of the boiler specifications is given in Table 1.
TABLE 1
Test Boiler Specifications
Plant Boiler No. 5
Year Installed 1966
Boiler Manufacturer CE-Wickes
Type A
Pressure Standard 250
Rated Capacity (PPH) 120,000
No. of Drums 3
Heating Surface (Ft.2) 8,897
Furnace Heating Surface (Ft.2) 1,073
Furnace Volume (Cu. Ft.) 2,180
Furnace Width 7' 8-1/2"
Furnace Length 30' 8-3/4"
Gas Passes Horiz.
Temp. Flue Gas Leaving Boiler (°F) 505
Number of Burners 2
Air Preheater Yes
Gas Temp. Entering (°F) 505
Gas Temp. Leaving (°F) 353
Air Temp. Entering (°F) 134.
Air Temp. Leaving (°F) 309
Heating Surface (Ft.2) 1,400
Stack Control Devices None
This work has been carried out in two phases: Phase I, March to June
1976; Phase II, March to May 1977. The test boiler has a natural draft
with no flue gas cleaning equipment. Also, the boiler has no provision
for ash removal. A lance-type traveling soot-blower was mounted in
the boiler floor to cover the entire length of the boiler, for blowing
steam to either re-entrain deposited ash or push ash to the rear of the
boiler for removal. The ash collected in the ash hopper at the rear
-------
COAL-OIL MIXTURE COMBUSTION 199
of the boiler could be sluiced out. (This system was, however, not ,
used in actual operation, since there was very little ash deposition
in the radiative section of the boiler.)
9.1 Phase I
During the period May to June 1976, the boiler was operated for
250 hours, burning a mixture of 35 weight percent coal, 59 weight per-
cent oil, and 6 weight percent emulsifier solution. (3/4 of 1% emulsi-
fier and 5 1/4$ water). The steam load ranged from 45,000 to 70,000
Ibs. per hour, with occasional loads of 96,000 Ibs. per hour. Two
types of burners were tested during this period: namely steam atomized
burner and air atomized burner. With both types of burners, the flame
envelop was as good or better than with No. 6 oil. Although the fire
appeared less brilliant with coal-oil mixture, no sparklers were no-
ticeable. In general, the air atomized burners gave a better perfor-
mance than the steam atomized burners, and both burners could be
switched from coal-oil mixture to No. 6 oil, with minor air register
adjustments. The only major problem associated with the steam atomized
burners was the deposition of carbonacous materials on the tip of the
burners. The burners had, therefore, greater tendency to plug than the
air atomized burners and had to be cleaned more frequently.
Stack testing was conducted to measure NOX, CO, and opacity over
a wide range of oxygen levels. The stack had a gray smoke with 32-35$
average opacity during the test. The opacity increased considerably
during soot blowing periods and was unacceptable. Fly ash participates
samples were drawn and analyzed. The fly ash size was very small and
almost all of the material was less than 20 microns in size. The NOX
levels increased approximately 100-150 ppm for coal-oil mixture com-
bustion over burning No. 6 oil. This can be somewhat attributed to
the higher nitrogen level in coal. Carbon monoxide levels were, how-
ever, almost the same in both cases (Table 2).
TABLE 2
Air Atomized Burner Performance
Coal-Oil
No. 6 Oil Mixture
Steam flow rate (ib/h) 40-45,000 40-45,000
Fuel pressure (psig) 116 140
Oxygen (%} 5.2-8.0 4-1-7.6
NO (ppm) 155-198 265-328
CO (ppm) 6-33 4-25
Particulate emissions (gr/acf) .0153-.0188 .373-.472
Opacity (%} 0-8 32-35
9.2 Phase II
During this phase, which ended April 30, 1977, a mixture con-
taining 50 weight percent coal, 43.3 weight percent oil, 6.5 weight
percent water, and 0.2 weight percent additive was prepared and burned.
Tests were carried out to determine emissions, combustion conditions,
and coal-oil mixture stability characteristics. The boiler fired coal-
-------
200
CLEAN COMBUSTION OF COAL
oil mixture for 494 hours at a range of loads, almost twenty percent
of this time was at steaming rates over 70,000 Ibs. per hour, with the
highest being 100,000 Ibs. per hour. As in the case with Phase I, it
was observed that coal-oil mixtures burn in a manner very similar to
No. 6 fuel oil. Also, it verified many of the observations that had
been made earlier in regard to flame characterization, boiler perfor-
mance, ash build-up, slagging and trouble-free operation of the air
atomizing burners.
>In this phase of the program, a new additive was used to improve
the stability of coal-oil mixtures. A high pressure homogenizer was
used to make an oil-additive-water emulsion. The oil-water emulsion
thus obtained was mixed with pulverized coal in a Marion mixer.
During the period of the test, there was some settling of coal
from the coal-oil mixture in the storage tank. This was caused mainly
by high temperature of the coal-oil mixture. This affected the vis-
cosity as well as the stability of the coal-oil mixture. This problem
can, however, be alleviated by controlling the temperature of the
mixture in the storage tank to below 150°F.
As in Phase I of the program, the steam atomized nozzle was
installed initially. This system performed satisfactorily as the per-
centage of coal in the mixture was increased gradually from 10$ to 40$
coal. However, at 50$ coal in the mixture, the nozzle plugged, and
the remaining tests were carried out with air-atomized burners manu-
factured by Forney Engineering Company. A schematic of the burner head
is given in Figure 4. In general, the air-atomized burner performed
satisfactorily.
SECOND-STAGE
OUTER CONE. SWIRL CHAMBER
INNER CONE
FIRST-STAGE
SWIRL CHAMBER
AIRFLOW
SPIRAL ATOMIZER
CONTROL SLEEVE
AIRFLOW
SWIRL CONTROL
MECHANISM
Figure 4. Burner head of air atomized Verloop burner.
-------
COAL-OIL MIXTURE COMBUSTION
201
10. DEMONSTRATION AT SALEM HARBOR STATION
New England Power Service Company has been awarded a cost-
sharing contract by the Energy Research and Development Administration
(ERDA) to develop and demonstrate a coal-oil mixture as a primary fuel
for generating electricity. In this program, the overall feasibility
of coal-oil mixture combustion will be demonstrated in an 80 MW unit
at Salem Harbor Station in Salem, Massachusetts. This boiler was ori-
ginally designed for burning pulverized coal, but has been converted
to and is presently burning residual oil. The specific objectives of
the project are to achieve a stable combustion with coal-oil mixture
R
111
SECONDARY
SUPERHEATER
Figure 5. Sectional view of 80 MW Babcoek and Wilcox boiler.
-------
202
GLEAM COMBUSTION OF COAL
fuel and to demonstrate satisfactory performance of all systems and
sub-systems, while meeting state and federal environmental regulations.
The target concentration' of coal in coal-oil mixture is 30 weight per-
cent. Initial tests in the program are planned to be conducted with
straight coal-oil mixture without the use of a chemical additive for
stabilization. Concurrently, however, research and development is
being conducted for development of an effective and economical stabi-
lizing additive. Tests will be carried out with the use of this ad-
ditive to determine its effect on coal-oil mixture combustion.
The test boiler (Figure 5) is a Babcock and Wilcox front-fired
radiant boiler rated at a steam flow of 625,000 Ibs./hour with a super-
heat/reheat temperature of 1,OOOF and a design pressure of 1,675 psi.
Presently, there are 12 mechanical atomizing high pressure burners that
supply a full load heat input of 880 million Btu/hr. Particulate re-
moval from the flue gases is accomplished by means of a Research Cot-
trell electrostatic precipitator. The design collection efficiency of
the precipitator is 97% when burning coal.
Several modifications and additions are planned for this demon-
stration program. Figure 6 shows a simplified flow diagram for the
NO. 6 OIL
t
COAL
UNLOADING
HOPPER*
FEEDER
»
PULVERIZER
•
PREWETTING
4
WATER
«••
V
BLENDING
H EMULSIFIER |
COM FEED
COOLING AIR FAN A
Figure 6. Simplified flow diagram of coal-oil mixture combustion
demonstration.
various systems and their interfaces. Coal will be delivered to the
dock by barge and unloaded into a pile adjacent to an existing conveyor
belt. Coal will be pushed into hoppers which feed the enclosed conveyor
to carry coal into storage bins, inside the power plant. The bins will
gradually be emptied into coal pulverizing equipment and the finely
ground coal will be delivered to a storage silo to be constructed ad-
jacent to the south wall of the power plant. The pulverized coal will
then be blended with oiland pumped to an existing fuel storage tank
for distribution to the burners. The existing high pressure burners
and control system will be replaced with low pressure air atomized
Verloop burners and the associated burner control and logic system.
-------
COAL-OIL MIXTURE COMBUSTION 203
As indicated in the schematic diagram, provision will be made for
prewetting and addition of emulsifier additive to improve the stability
of the coal-oil mixture, if necessary.
The test procedure is divided into two phases, each containing
incremental test steps, in order to debug the various systems and to
determine the effects of coal-oil mixture combustion on the boiler,
precipitator, and ash system, as well as on stack emissions.
10.1 Phase I - Feasibility
The purpose of this phase is to develop combustion stability and
evaluate the effects of coal-oil mixture combustion on the boiler, pre-
cipitator, auxiliary systems and stack emissions.
The initial operation will involve baseline testing of the Ver-
loop burners on No. 6 oil. This testing will be used to debug the new
system and to obtain baseline data for later comparison with the coal-
oil mixture combustion results. Initially, coal-oil mixture will be
burned in a single row (three burners) with remaining nine burners
burning No. 6 oil. Tests will be done with mixtures of 10$, 20$, and
30$ by weight coal concentration and will evaluate flame stability,
emissivity, opacity, and burner performance, etc. Once the combustion
stability is demonstrated, the next step in this test sequence will be
a six-burner test, supplying half of the boiler heat input from coal-
oil mixture fuel. At this time the effects of coal-oil mixture on the
equipment, pumps, piping systems, instrumentation and burner system
will be closely monitored. Potential problem areas that will be moni-
tored are slagging in the furnace and the convection pass, plugging of
the air preheater, effects of larger quantities of coal ash on the ash
water recycle system and stack emissions.
10.2 Phase II - Demonstration
In this phase of the program, all twelve burners will burn coal-
oil mixture, thus 100$ of the boiler heat input will be supplied by
the coal-oil mixture. Throughout the test period, close attention will
be given to environmental aspects of burning coal-oil mixture as well
as the effects of coal-oil mixture combustion on erosion, corrosion,
and boiler performance will be monitored. Tests will be carried out to
optimize performance of various systems and subsystems as well as to
determine the efficiency of boiler and precipitators, while burning
coal-oil mixture.
10.3 Schedule
It is presently anticipated to complete the installation and
construction of the coal-oil mixture preparation, blending and com-
bustion facility in the middle of 1978. The Phase I of the program is
scheduled for six months duration. After the Phase I results have
been evaluated, a decision will be made to proceed into Phase II of the
program. It is anticipated that this phase will begin in early 1979
and run through the conclusion of the project scheduled for the end of
1979-
-------
204 CLEM COMBUSTION OF COAL
11. POTENTIAL BENEFITS
There are a number of potential benefits to the United States if
coal-oil mixture combustion feasibility can be demonstrated. The most
obvious advantage is a reduced dependency on imported oil if 30-40$ of
the fuel burned becomes domestically produced in the coal-oil mixture.
Combustion of coal-oil mixture will, evidently, have less impact on
air quality than 100% coal burning. It has potentially a wide area of
application on boilers originally designed to burn coal or oil but can-
not burn coal in view of changed circumstances. Although some further
technical development is desirable, coal-oil mixture as a fuel has a
potential for a near-term alternative for utilities and industrial boi-
ler applications.
Furthermore, a central coal-oil mixture preparation plant can
be visualized that prepares and ships a stabilized coal-oil mixture to
outlying power plants. Small users could buy such fuel and presumably
interchange coal-oil mixture and oil with a minimum of turn-around and
(•rpense. Large users of coal could probably justify their own prepara-
tion facilities.
ACKNOWLEDGEMENT
The author gratefully acknowledges the assistance of Mr. Andrew
Brown, Jr., Project Manager, General Motors Corporation, for providing
the results of the General Motors demonstration tests; and Mr. Richard
M. Dunn, for his contribution related to the coal-oil mixture project
at Salem Harbor Station.
REFERENCES
1. Smith, H.R. and H.M. Hunsell, Liquid Fuel, U.S. Patent 219,181
(February 24, 1879)
2. Manning, A.B. and R.A.A. Taylor, "Colloidal Fuel", Trans. Inst.
Chem. Engrg., 14, 45 (1936).
3. Barkley, J.F., et al., "Laboratory and Field Tests on Coal-in-Oil
Fuels", Trans. ASME, 66, 185 (1944).
4. Giiman,H.H., "Coal-oil Emulsions for Boiler Fuel", EPRI Journal 2,
56, April 1977. ~
5. Demeter, J.J., et al., "Combustion of Coal-oil Slurry in a 100 H.P.
Fire Tube Boiler" Pittsburgh Energy Research Center. Report
PERC/RI-77/8, May 1977.
6. Brown, A. Jr., "General Motors Powdered Coal-oil Project", pre-
sented at Fuel Switching Forum U.S. ERDA, June 6-7, 1977.
7. Annon, "Proceedings of the Coal-oil Mixture Combustion Technology
Exchange Workshop", U.S. ERDA Washington, D.C., CONF. 767019 -
M77-8, February 1977.
-------
205
STOKERS FOR INDUSTRIAL BOILERS:
ASSESSMENT OF TECHNICAL, ECONOMIC,
AND ENVIRONMENTAL FACTORS
Robert D. Giammar*
Battelle,
Columbus Laboratories
Columbus, Ohio ^3201
INTRODUCTION
The firing of boilers for process steam, space heating, and on-
site power generation accounts for about half of all the fuel used by
American industry—nearly as much as that consumed in the generation of
electricity by utilities. Coal once was the dominant fuel for such in-
dustrial boilers; it now provides only about one-quarter of their fuel.
Economic and environmental factors over the past 30 years have
made coal relatively unattractive for industrial boilers. Both the capital
and operating costs of coal-fired boiler installations are inherently
higher than those of equivalent facilities designed for gas and/or oil
firing. Thus, with the widespread availability of oil and gas in the
late 1940's and early 1950's, there was a drastic decline in the demand
for the industrial stoker. Only in some of the largest boiler installa-
tions could the savings resulting from the somewhat lower cost of coal
offset the greater capital and operating costs. As a consequence, there
has been little incentive for manufacturers to improve stoker technology
through research and development. However, a resurgence of demand can
be forecast for the near term in view of the fuel supply situation and
possible legislative action.
Stoker-fired boilers are now at a further disadvantage with
respect to environmental considerations, as their emissions of particulate
and SO- are high in comparison to gas- and oil-fired boilers. The estab-
lishment of stringent air-pollution regulations in recent years (with
national standards for large new installations and local standards for
many other installations) has accentuated the economic disadvantages of
stoker boilers. Costs of installing and operating the downstream con-
trol equipment that is needed to control emissions of fly ash and S02
for industrial boilers are relatively higher on a Btu basis than for
larger utility power plants.
The potential market for stokers is dependent on many factors,
some of which are difficult to assess because they involve both public
* The author wishes to acknowledge the guidance and assistance given by
Battelle staff—R. E. Barrett, D. W. Locklin, and R. B. Engdahl.
-------
206 CLEAN COMBUSTION OF COAL
and industry acceptance of stoker firing. The industrial boiler
(typically with rates steam-generating capacity between 10,000 and 500,000
pph*) is primarily utilized to generate steam for process and space
heating and often is subjected to widely varying loads. Unlike the
utility boiler, the industrial boiler is usually a secondary consideration
in an industrial complex. The characteristics sought include:
Low first costs
Low maintenance
High reliability
Ease of operation
Minor auxiliary equipment requirements
Wide turndown
Quick response.
Most of these features are more easily obtained with the oil-
or gas-fired boiler. Recently, however, the availability of oil and gas
has become an increasingly important factor in the selection of boilers
and may outweigh all other factors if the shortage of these fuels becomes
as severe as expected. Thus, the stoker-fired boiler will become more
popular in the near term until coal conversion processes such as coal
desulfurization, gasification, or liquefaction are commonly available
or advanced coal-burning techniques such as fluidized-bed combustion
reach the economically competitive stage.
CHARACTERIZATION OF STOKER TYPES
Stokers of various types are designed to mechanically feed coal
uniformly onto a grate or tuyeres within a furnace, to supply combustion
air to the fuel bed, and to remove ash from the zone of combustion. The
development of the mechanical stoker has advanced through the years,
and the modern stoker is still considered an important element in the
industrial combustion of coal as well as other types of solid fuel. The
modern stoker-boiler system incorporates controls to coordinate air and
fuel supply with changing loads, dust-collecting equipment to minimize
emissions, and, in many cases, fly-carbon return systems to increase
efficiency.
Stokers are classified according to the method of feeding fuel
to the furnace, namely:
• Spreader
• Underfeed
• Chain grate or traveling grate
• Vibrating grate.
The type of stoker used in a given installation depends upon
* pph = pounds steam generated per hour; one Ib/hr - approximately
1000 Btu/hr.
-------
STOKERS FOR INDUSTRIAL BOILERS 207
the general system design, the capacity required, and the type of fuel
burned. In general, the spreader stoker is the most widely used in the
capacity range of 75,000 to 400,000 pph because it responds quickly to
load changes and can burn a wide range of coals. The underfeed stokers
are principally utilized with small industrial boilers of less than
30,000 pph. In the intermediate range the large underfeed units, as well
as the chain- and traveling-grate stokers, are being displaced by spreader
and vibrating-grate stokers. A brief description of each class of stoker
is given below. The major features of each are summarized in Table 1.
Detailed discussions of the various stoker types can be found in the
37th and 38th editions of Steam, Its Generation and Use, published by
the Bab cock & Wilcox Company, New
TRENDS IN COAL FIRING AND STOKER TYPES
Recent Battelle-Columbus studies directed to the characteriza-
tion of the industrial boiler population provide some insight as to trends
in coal-firing methods by boiler size range. Information on the indus-
trial boiler population and design trends was developed in a special
survey of boiler manufacturers conducted jointly by Battelle and the
American Boiler Manufacturer's Association (ABMA) and reported in
Reference (2). These results were later refined on the basis of an
analysis of recent sales data for industrial-size water-tube boilers'^).
Firing methods are boiler-size dependent. Thus, different size
categories were defined as follows for the presentation of trend information:
Size Category Size Range, 1000 pph
A 10-16
B 17 - 100
C 101 - 250
D 251 - 500
Smaller boilers are placed in the "commercial" class and larger boilers
are considered to be "utility boilers".
Trends in Coal-Firing Capability
Figure 1 shows trends in coal-firing capability estimated for
boilers in the four size categories installed in 1930, 1950, and 1970
and forecasted for 1990^'. The data apply to boilers designed to fire
coal or having a capability to fire coal as a secondary fuel. This dis-
tribution is shown as a percentage of all industrial boilers. The pro-
jection for 1990 was revised by Battelle from the earlier survey on the
basis of the following broad assumptions:
• Oil and gas supplies will be limited.
• Oil and gas will be utilized in smaller equipment and
for high-priority uses (but not fully by mandatory
allocation).
• Coal will dominate new installations in the larger sizes.
-------
TABLE 1. CHARACTERISTICS OF VARIOUS TYPES OF STOKERS
to
o
00
Stoker Type
and Subclass
Typical
Capacity Range, Maximum Burning Rate,
pph(a) Btu/hr-ft2
Characteristics
Spreader
Stationary and
dumping grate
Traveling grate
Vibrating grate
Underfeed
Single or double
retort
Multiple retort
Chain grate and
traveling grate
Vibrating grate
20, 000 to 80, 000
100,000 to 400, 000
20,000 to 100, 000
20,000 to 30, 000
30,000 to 500,000
20,000 to 100,000
30, 000 to 150,000
450,000
750,000
400,000
400,000
600,000
500,000
400, 000
Capable of burning a wide range of coals,
best ability to follow fluctuating loads,
high fly-ash carryover, low load smoke
Capable of burning caking coals and a wide
range of coals (including anthracite), high
maintenance, low fly-ash carryover,
suitable for continuous-load operation
Characteristics similar to vibrating-
grate stokers except these stokers
experience difficulty in burning strongly
caking coals
Low maintenance, low fly-ash carryover,
capable of burning wide variety of
weakly caking coals, smokeless
operation over entire range
o
o
O
CO
i-3
M
O
o
o
(a) pph = Ib steam per hr; 1 pph=*1000 Btu/hr.
-------
rj -Jr
Q. O W
o _- *r w
O o 2 i-
0*^0
I e ° c
O iT C —
X <•> T3
^ O if «
^ *- « o
o >
CD C
uu
80
60
40
?n
r\
—
/ '
/ /
//'
+ / 1
it
-/ /
f
/ /
/ /
>---|
E-drm
'30 '50 70
A
v \
^ \ s
t\ \
'90
10.000-16,000 pph
' /
f .
f/t
f / 1
; >
' ,
/ /
f /
'30
—
- « —
—
—
'50
i : ! i
'70
B
N X
^ \ '
„ \ ^
V •
^ \
. \ N
'90
/ ,
' / 1
' t
'//
/ /
/ 1
' / f
f
' /
/ /
'30
_ .
---
_ _r
'50
* i
i i
i i
i i
i i
70
C
17,000- 100.000 pph
101,000
r ~i
1
1 t
;\\
V \ ^
s \ N
^ \ ^
S \ N
S \
/ /
' / i
'/,
/ / 1
f/'
' t
' 1
f /
:"-":
—
—
—
—
—
T
'90 '30 '50 70
D
1 1
\ "^
^
\ s
v \ N
x\X
\ \
s N S
^ \
CQ
8
Q
O
H
a
CO
»
M
'90 P
td
O
-250.000 pph 251. 000-500.000 pph y
Figure 1. Estimated Percentage of all Boilers in Four
in 1930, 1950, and 1970 and Forecasted for
Fire
Coal*
Size Categories Installed
1990
Having a Capability to
N)
O
IO
-------
210 CLEAN COMBUSTION OF COAL
• Clean liquid and gaseous fuels from coal conversion
processes will not be available in large quantities
by 1990.
• Firing refuse as a supplementary fuel will increase,
but will not become significant in terms of percentages.
The coal projection includes the firing of chemically refined solid fuels,
but does not include future synthetic liquid or gaseous fuels derived
from coal.
The trends in Figure 1 indicate that coal was the dominant fuel
for industrial boilers in the pre-World War II era and that most boilers
had a coal-firing capability. By 1950, the number of units with a coal-
firing capability dropped dramatically for boilers being installed in the
smaller size categories; in the largest size category, 65 percent of the
boilers had the capability. By 1970, coal firing nearly disappeared from
the smaller boiler size categories, but 40 percent of the largest boilers
were being installed with a capability to fire coal.
For the future, a return to the capability for direct firing
of coal was forecast for new installations in the next 15 years, especially
in larger sizes, where about 70 percent of the new boilers are expected
to have coal-fire capability. In the event of new legislation and/or
fuel priority allocations that prohibit firing gas and oil in large
boilers, all new industrial boilers may be required to have a coal
capability. Some large industrial companies have already made the
decision to provide for a coal capability in new installations as a
measure of insurance to keep manufacturing plants operating in the face
of an uncertain fuels situation.
Trends in Stoker Types
Figure 2 shows trends in coal-firing methods for industrial
boilers in each of the four size categories as developed in the studies
cited earlier(•*). Firing methods included in these data are:
• Stoker types
Spreader
Underfeed
Overfeed (chain grate, traveling grate, or vibrating grate)
• Other
Pulverized coal
Miscellaneous (cyclone, refuse, sawdust, wood, etc.).
The following overall observations can be made from Figure 2:
• Before the introduction and commercialization of the
spreader stoker in the 1930's, underfeed stokers
dominated the smaller size categories and had an
appreciable share of the market in the larger sizes,
The overfeed types were significant in all sizes,
especially larger sizes. Firing of pulverized coal
was significant only in the largest sizes.
-------
Overfeed
Size
30 '50 '70 '90
A
category 10.000 - 16.000 pph
Underfeed
Spreader
Pulverized
coal
'30 '50 '70 '90
B
17,000- 100.000 pph
'30 '50 '70 '90
C
101,000-250,000 pph
Other
'30 '50 '70 '90
D
251,000-500,000 pph
en
1-3
o
01
"d
o
8
CO
S
H
Figure 2. Estimated Breakdown by Firing Method of Solid Fuel Boilers
in Four Size Categories Installed in 1930, 1950, and 1970
and Forecasted for 1990 .
-------
212 CLEM COMBUSTION OF COAL
• By 1950, the spreader stoker had gained a significant
share of the market below 250,000 pph, mainly dis-
placing overfeed stokers below 17,000 pph and underfeed
stokers from 17 to 100,000 pph. Pulverized coal made
some inroads in the size category from 101 to 250 pph
during this period.
• In the 1950's and 1960's, the penetration of spreader
stokers into the share of the underfeed stokers continued
to the point that it was the dominant type for the range
of sizes encompassing 17,000 to 250,000 pph. Pulverized-
coal firing gained further, nearly dominating the market
above 250,000 pph by 1970.
• For the near future, spreader stokers can be expected to
continue to gain popularity in the smaller sizes, with
underfeed stokers holding the major share below 17,000 pph.
Spreader stokers can be expected to decline slightly in
the market share above 100,000 pph, where firing of pul-
verized coal is strongest.
It should be noted that introduction and successful commercialization of
the fluidized-bed combustion concept to industrial boilers could displace
the more conventional firing methods by 1990, and, thus, significantly
alter the distributions forecasted in Figure 2, especially above
100,000 pph.
EMISSIONS FROM STOKER-FIRED BOILERS
Table 2 summarizes the emission factors* established by EPA as
averages for fossil-fuel combustion'^). As noted, the particulate and
sulfur dioxide emissions are significantly higher for combustion of coal
than for fuel oil or natural gas, especially considering that coal can
have an ash content as high as 20 percent and a sulfur content as high as
4 percent. Also, whereas fuel oil can be desulfurized before firing, there
are currently no commercially viable methods for mechanically treating or
preparing coal to reduce emissions of S02 (other than washing, which
cannot remove organically combined sulfur or fine dispersions of pyritic
sulfur). Consequently, to reduce emissions from coal-fired combustion
sources necessitates the utilization of downstream emission-control
equipment.
Because available emission data for stokers are limited, there
are no computational procedures for establishing specific relations among
emission levels, stoker type and design, boiler load, coal characteristics,
and coal size. The emission factors in Table 2, however, are representative
of a broad grouping of equipment types and can serve as guidelines.
Generally, stokers that agitate the fuel bed and disturb the
ash increase the fly ash loading in the stack. Generally, this agitation
*
The emission factor as used here is a statistical average of the rate
at which a pollutant is released to the atmosphere per unit of fuel
consumed^'.
-------
TABLE 2. EMISSION FACTORS FOR FOSSIL-FUEL COMBUSTION WITHOUT CONTROL EQUIPMENT^
Emission Factor, lb/ 10" Btu of fuel consumed
Sulfur Carbon
Fuel and Type of Combustion Unit Particulates Oxides Monoxide
Coal
Power plant, pulverized
General 0. 64 A(a) 1.25S(b) 0.04
Cyclone burner 0. 08 A 1.52S 0.04
Commercial and industrial stoker
Spreader 0. 52 A 1.52S 0.08
Others 0.20 A 1.52S 0.08
Residual oil
Power plant (also includes 0.053 1.04S 0.02
distillate oil-fired plant
boilers
Commercial and industrial 0.153 1.04S 0.026
Distillate oil: commercial and 0.01 0.95 0.026
industrial
Natural gas
Power plant 0.015 C.0006 0.017
Commercial and industrial 0.18 0.0006 0.017
Nitrogen
Hydrocarbons Oxides
0.012 0.72
0.012 2.2
0.04 0.60
0.04 0.60
0.013 0.7
0. 02 0. 27 to 0. 54
0. 02 0.27 to 0. 54
0.001 0.600
0. 003 0. 120 to 0.230
(a) The letter "A" indicates that the weight percentage of ash in the coal should be multiplied by the value given. Example: If the factor
is 0.64 and the ash content is 10 percent, the paniculate emissions before the control equipment would be 10 times 0.64, or 6.4 lb/10
Btu (about 160 Ib/ton).
(b) "S" equals the sulfur content [see footnote (a) above] in weight percent.
03
(-3
O
CQ
§
H
£
trj
O
H
t->
O)
to
-------
214 CLEAN COMBUSTION OF COAL
is intermittent, as is the case for dumping-grate or vibrating-grate
stokers. The spreader stoker, however, burns up to 50 percent of the
coal in suspension and, thus, has the highest fly-ash loading of all
types of stokers.
Sulfur oxide emissions from stokers, as well as from all fossil-
fuel combustion equipment, are directly related to the sulfur content of
the fuel. Presently, there are no combustor designs or equipment opera-
tional modifications that can reduce these emissions except for the still
commercially unproven fluidized-bed combustors. In addition, unlike the
utility power plant, the economics and the geographical location of stoker-
fired commercial and industrial plants usually are unfavorable for utilizing
tall stacks to reduce ambient S02 to acceptable levels.
Emissions of carbon monoxide, unburned hydrocarbons and nitrogen
oxides from stoker boilers generally are within air-pollution standards,
provided that the units are operating properly.
COMPETITIVE SITUATION IN INDUSTRIAL FUELS
Figure 3 provides a comparison of steam-generation costs for
various fuels on the basis of a load factor of 0.30, a boiler efficiency
of 0.80, and an annualization rate of 16.7 percent. The estimated costs
include both capital and operating expenses. It is apparent from the
example shown that the stoker boiler is economically attractive when
the cost differential between stoker coal and other fuels,is greater
than $1/106 Btu. For example, if stoker coal costs $1/10 Btu (approxi-
mately $26/ton), steam-generation costs would be approximately $3/10 Ib
of steam. In an oil or gas boiler, the fuel costs would have to be no
higher than $2/10° Btu to generate steam at $3/103 Ib. In the past, the
cost of all industrial fuels was much less than $1/10° Btu and cost
differentials were about $0.50/10 Btu between fuels.
Figure 3 also indicates the steam costs from a stoker boiler
equipped with S02 scrubbing equipment. Because of the relatively high
costs estimated, it is unlikely that S0£ scrubbing will be utilized with
medium-size industrial boilers (unless dictated by Government regulation).
ECONOMICS OF INDUSTRIAL-STOKER OPERATION
The economics of industrial utilization of stoker coal are
difficult to assess for reasons that include:
• Uncertainty in the availability and price of all fuels
• Wide variation in the cost and quality of stoker-boiler systems
• Unavailability of extensive operational cost-breakdown
information
• Unavailability of proven S02~control technology.
Delineation of the economic situation vith some degree of confidence would
have required an effort that was beyond the scope of this paper. As a
consequence, the analysis that was developed and is given below should be
regarded only as a guideline and for making relative comparisons between
types of fuel.
-------
STOKERS FOR INDUSTRIAL BOILERS
215
7.00
6.00
5.00
4.00
"b
v^
•ifr
O
O
E
o
CD
V)
3.00
2.00
1.00
O.OQ.
Annualizaiion 16.7%
Load factor 0.30
Boiler efficiency 0.80
l i
0.00
l.OO
2.00
3.00
4.00 $/IO Btu
Equivale
o
o
O)
U_
0
i
0.0
1
0
2 4
1
0.5
i i
10 20
6 8
i
1.0
i
30
10 12
1 1
1.5
1
2.0
i
14
i
40 50 60
Fuel Cost
16 3
_) ,
2.5 $/l
$/ton
I/DDI Of Oil
0 ft of natural gas
of coal
Figure 3. Cost Comparison of Various Types of Boiler Systems.
-------
216 CLEM COMBUSTION OF COAL
Capital Costs
Because coal-fired stoker boilers are more complex, the capital
costs of a stoker-fired boiler are significantly greater than those of
either an oil- or gas-fired unit. In comparison to an oil- or gas-fired
boiler, the stoker boiler requires:
Ash-handling facilities
Fly-carbon reinjection system (for some spreader stokers)
Very-high-efficiency dust-collection equipment
Soot blowers
Larger flow passages and additional heat-transfer surfaces
More extensive combustion controls
More extensive fuel-handling and storage facilities
More extensive field erection work
More physical space for facilities.
These factors result in the capital cost of a stoker boiler being from two
to four times that of a packaged boiler for oil or gas firing. Capital-
cost differential will vary, depending on equipment design, manufacturer,
and type of installation.
As an example of the economic attractiveness of gas- or oil-fired
equipment, one manufacturer estimated that the delivered and erected cost
of a 75,000-pph boiler was $6/pph for an oil/gas packaged unit and $20/pph
for a stoker boiler (with dust control). The stoker alone can account for
up to 20 percent of this cost with another 20 percent attributed to dust-
collection equipment.
With the other costs associated with a heating plant, the capital
costs of placing a boiler on line are estimated to be approximately
$10/pph for oil/gas and $25/pph for stoker coal. (These numbers are based
upon information from a manufacturer and were derived from one set of
assumptions.) S02~cleanup equipment would add an additional $12/pph to
the cost of the stoker boiler'".
Operating Costs
The average operating costs of boilers are difficult to deter-
mine because of differing accounting procedures. However, in comparison
with oil- or gas-fired units, stoker boilers clearly have higher operating
costs. These can be attributed to:
More frequent and closer attention required by operator
Higher degree of boiler-operator skill
Additional mechanical equipment to service
Ash removal
Fuel handling.
In addition, the fuel costs must be considered. These costs vary from
region to region and have "become difficult to predict. A check on the
current fuel prices in Columbus, Ohio, indicates little difference among
the energy costs of stoker coal and natural gas, while fuel oil costs
about 50 percent more.
-------
STOKERS FOR INDUSTRIAL BOILERS 217
CONVERSION TO STOKER FIRING -
FACTORS TO CONSIDER
Stoker fuel can be a viable fuel for meeting industry's demand
for energy. However, as alluded to earlier, certain factors must be
considered in the design and operation of the overall system that include:
• Space
- boiler and auxiliary equipment
- coal handling and ash disposal systems
- pollution control equipment
• Costs
- capital
- operating
• Operation
- operator skill
- high maintenance.
Generally, a boiler that has not been designed for coal firing cannot be
converted. Additionally, if a new coal-fired boiler is installed, a
substantial amount of space is required for the new and larger boiler
and all the ancillary equipment associated with it. Also, even with the
higher cost of gas and oil, stoker firing presently in many instances
cannot be justified on economics alone.
BCL Stoker Conversion
During 1976, Battelle's Columbus Laboratories (BCL) converted
one of their 600-hp (25,000 pph steam) oil/gas boilers, originally designed
for coal firing, to stoker (spreader) firing. These boilers were originally
installed with oil or gas firing, but the units were designed and assembled
with conversion to coal firing in mind. The decision to convert to coal
was not primarily one of economics, but one rather to assure continued
operation of the steam plant. The overall cost of converting one of
the existing boilers was $800,000. About half the cost was associated
with the boiler modification and stoker installation while the remaining
half with the coal and ash handling equipment, pollution controls, and
engineering design.
The conversion essentially doubles the space of the steam plant
facility. Ash and coal silos, above and below ground bunkers, and
bucket elevators were located adjacent to the steam plant, while the
two-stage dust collector, forced and induced draft fans, and 18-ton
coal hopper were located within the building.
With the conversion, higher maintenance is required. Instead
of the fuel being supplied to the boiler through a pipe, coal is delivered
through a system of bucket elevators, screw feeds, and conveyors. As
expected, several small mechanical difficulties have occurred time to
time. Like-wise, operational problems, such as low-load smoke and exces-
sive slagging with low ash-fusion temperature coals, were experienced.
These and other problems have been minimized by proper coal selection
and stoker operation.
-------
218 CLEAN COMBUSTION OF COAL
CONCLUSIONS
It appears that the future market for stoker boilers and the
corresponding demand for stoker coal -will be dependent on factors that
include:
• National policy decisions that affect the availability of
fuels
• Local and national S02 regulations
* Development of suitable SOg control systems (either in-
furnace control systems or stack-gas scrubber systems)
• Availability of clean coal-derived fuels or commercialization
of fluidized-bed combustion systems with sulfur removal.
The stoker market appears promising for the next 5 to 10 years and per-
haps longer, particularly for the large stokers. The market for the
small- and medium-sized stoker is less definitive, but sales of medium-
sized industrial stokers will probably increase over the essentially neg-
ligible levels of recent years. Although stokers firing high-sulfur coal
emit S02 at levels above those specified by many state source standards,
it is anticipated that some relaxation of these standards may occur,
provided that the local primary ambient standards are being met. It is
also possible that economic and practical S02 stack-gas cleanup systems
will be developed, but these systems appear to have many problems.
REFERENCES
1. Steam, Its Generation and Use, 37th and 38th Editions, The Babcock
& Wilcox Co., New York (1963, 1973).
2. Barrett, R.E., Miller, S.E., and Locklin, D.W., "Field Investigation
of Emissions from Combustion Equipment for Space Heating", Final
Report on Contract 68-02-0251 from Battelle's Columbus Laboratories
to U.S. Environmental Protection Agency, EPA-R2-73-08Ha (June,1973).
3. Locklin, D.W., Kropp, E.L., "Design Trends and Operating Problems
Part A - Industrial Boiler Population and Design Trends", Final
Report Grant R-802H02 by Battelle's Columbus Laboratories to U.S.
Environmental Protection Agency, EPA-650/2-73-032 (April, 197M •
k. "Compilation of Air Pollutant Emission Factors", U.S. EPA, Office
of Air Programs, Publication No. AP-1+2 (April, 1973).
5. "Initial Operating Experiences with Dual-Alkali S02 Removal System",
Presented at the EPA Symposium on Flue-Gas Desulfurization, Atlanta,
November k, 197U.
-------
219
INITIAL OPERATION OF THE 30 MWe RIVESVILLE MULTICELL
FLUIDIZED BED STEAM GENERATION SYSTEM
by
Robert L. Gamble and Newton G. Wattis
Foster Wheeler Energy Corporation
ABSTRACT
The world's largest fluidized bed steam generator located at Rivesville,
West Virginia is currently undergoing initial operation. This unit is
designed to generate 37.8 kg/s (300,000 Ib/hr) equivalent to 30 MWe of
superheated steam while firing coal in a fluidized bed of limestone.
Coal was first fired in this unit in December, 1976 and start-up opera-
tion is progressing.
INTRODUCTION
Coal is presently our nation's most abundant energy source. Current
environmental regulations require that coal be mined and consumed in
an environmentally acceptable manner. The majority of the energy con-
sumption in this country is east of the Mississippi river and the ma-
jority of the coal resources east of the Mississippi contain quantities
of sulfur which prohibit direct combustion without means for control-
ling emissions within current limits. Direct combustion of coal in
fluidized beds of limestone has shown to be a near term practical meth-
od for energy conversion with high sulfur coal.
The fluidized bed steam generation system located at the Monongahela
Power Company, Rivesville, West Virginia plant has been sponsored by
the United States Energy Research and Development Administration (ERDA).
The Rivesville system has been designed by Pope, Evans and Robbins, Inc.
and includes a multicell fluidized bed steam generator designed and
erected by Foster Wheeler Energy Corporation. The entire system has
been retrofit into an existing power plant and steam produced from this
system will be used to drive an existing turbine-generator and generate
approximately 30 MWe of electric power. The system installed at Rives-
ville is designed to generate 37.8 kg/s (300,000 Ib/hr) of superheated
steam and is several times larger than any other fluidized bed steam
generation system in the world.
Following the initial start-up operation, de-bugging of the systems at
the Rivesville plant and installation of test instrumentation, a test
program of approximately one year duration will commence. Successful
operation of this unit will be a major step toward the commercializa-
tion of both industrial and utility fluidized bed steam generators
which will be capable of directly firing low grade and high sulfur
coals within present environmental emission limits.
-------
220
CLEAN COMBUSTION OF COAL
FLUIDIZED BED COMBUSTION
A fluidized bed is a bed of granular particles supported by a non-
sifting grid through which an upward flow of air or, gas is passed with
sufficient velocity to lift and float the granular particles. As the
velocity of fluidizing gas is increased, the bed will expand and bub-
bles will form similar to a pot of boiling water, and in this state
the fluidized bed exhibits the properties of a liquid. Figure 1 sche-
matically represents a fluidized bed steam generator. In the case of
a fluidized bed steam generator firing high sulfur coal, the fluidized
bed particles are normally crushed limestone. Crushed coal is injected
into the fluidized bed of limestone and burns, converting the fuel
bound sulfur to sulfur dioxide (802)• If the bed is designed and oper-
ated to maintain a bed temperature of 815-870C (1500-1600F), the S02
released from the coal combustion will be chemically absorbed by the
limestone bed material. Tests have indicated that S02 emissions can
be controlled for virtually all United States coals by maintaining ac-
tive limestone bed material within the fluidized bed. This is done by
feeding raw limestone with the coal and regulating the bed material in-
ventory by a gravity drain system which withdraws spent material and
large coal ash particles.
FLUE
FUEL
INJECTION
PIPES
1550*F
AIR
DISTRIBUTION
GRID
fN-211
Figure 1. Fluidized Bed Steam Generator.
Due to the relatively low operating temperature of the fluidized bed
NOX emissions are inherently low and no special systems are required'
for control of NOX emission to meet the current emissions limit. Par-
-------
Fluid!zed Bed Steam Generation
221
ticulate emissions can be controlled conventionally with either elec-
trostatic precipitators or baghouse filters.
Many thousands of hours have been logged in pilot unit operation at the
Pope, Evans and Robbins, Inc., Alexandria, Virginia facility, the
Foster Wheeler Energy Corporation test unit at Livingston, New Jersey
and other pilot units throughout the United States and overseas.
RIVESVILLE UNIT DESIGN
The Rivesville systems design can be divided into four (4) primary
areas which include (a) the air and gas systems, (b) the water and
steam systems, (c) the fuel and limestone injection systems and (d)
the bed material recycle system. A description of each of these sys-
tems follows:
a. Air and Gas Systems - The air and gas systems are typical of that
found in a utility power plant steam generating system. Air is
supplied by a forced draft fan through a regenerative air preheat-
er and the flow rate is regulated to each section of the boiler by
individual air regulating dampers. The air passes through the air
distribution grid and the fluidized bed coal combustion zone, and
from this point flue gas flows over convection type heat transfer
surface and to cyclone particle separators, through a hot (385C,
730F) electrostatic precipitator, and on to the gas side of the re-
generative air preheater and induced draft fan at the inlet of an
exhaust stack.
b. Water and Steam Circuit - A schematic of the steam generator cir-
cuit is indicated in Figure 2. Feedwater from the existing plant
feedwater system enters the economizer inlet and flows in series
Cell"A" Cell]*" CeH"C" Cell"D*
SERIES
PARALLEL
BOILER
CIRCUIT
flTENKMTN SHUT
300,000 LB/HR
1350 PSIG 925 °F
STEAM TO TURBINE
Figure 2. Demonstration Unit Circuit Schematic.
-------
222
CLEAN COMBUSTION OF COAL
through four (4) economizer sections in each of four (4) independent
boiler cells. Feedwater enters the drum and flows through the
downcomers to two (2) forced circulation pumps as seen in Figure 3.
The forced circulation pumps provide the head required to maintain
the proper mass flow rate in the five (5) parallel boiler circuits.
The boiler circuits consist of four (4) horizontal tube boiler
banks and vertical tubes which form the boiler enclosure walls and
cell partition walls. A steam water mixture exits these boiler
circuits to the steam drum, and dried saturated steam flows from
the drum through a primary superheater and subsequently through a
finishing superheater. An attemporator spray station is located
between the two (2) superheaters for final steam temperature trim.
Steam leaving the finishing superheater is routed to the existing
steam turbine where it enters at 9 MPa (1300 psig) and 495C (925F).
Figure 3. Multicell Fluidized Bed Demonstration Steam Generator In-
stalled at Monongahela Power Co., Rivesville, WV, Plant.
The arrangement of the heat exchange surface indicated in the circuit
schematic (Figure 2) is shown on Figure 4.
Cell's"
CellmA"
Figure 4. Demonstration Unit.
-------
FLUIDIZED BED STEAM GENERATION
223
c. Fuel and Limestone Injection - Crushed limestone is received at the
plant in pneumatic discharge trucks and blown into storage bunkers
above the steam generator as indicated in Figure 5. Coal is either
received at the plant in a crushed and dried state and loaded into
the storage bunkers with the plant conveyor systems or is received
as run of mine and is crushed and dried in the plant systems prior
to loading in the storage bunker. A coal/limestone mix is regulat-
ed to each fluidized bed cell with rotary flow control feeders at
the outlet of each of the bunkers. The coal and limestone mix in-
to a common pipe as they fall by gravity to a rotary air lock. The
rotary air lock prevents back flow of air from the slightly pres-
surized vibrating distributors, each of whcih distribute the coal/
limestone mixture to eight (8) outlets. The vibrating distributor
is pressurized to approximately 14 kPa (2 psig) with air which
pneumatically carries the coal/limestone mixture through downward
sloping injection pipes into the lower section of the fluidized
bed. The Rivesville unit contains seven (7) coal/limestone in-
jection systems, two (2) of which are indicated on Figure 5.
COAL6TOAAQE
Figure 5. Fluidized Bed Fuel Injection System.
Cells A, B and C (refer to Figure 4) each receive coal and lime-
stone from two (2) sides and Cell D, which is the smallest cell,
receives coal and limestone through the single end wall. The
Rivesville system uses limestone crushed to -3.2 mm (-1/8 inch).
and coal crushed to -12.7 mm (-1/2 inch).
Bed Material Recycle System - The bed material recycle system, is
used to control the bed material size and inventory in each Qfill*
Bed material is extracted from each cell through a gravity drain
and the flow is regulated onto a common vibrating conveyor as in-
dicated in Figure 6. Bed material from the common vibrating con-
veyor may be removed from the system by opening the bypass or may
flow through a classifier which moves oversize bed material and
maintains properly sized bed material for recycle back to each of
the cells. Bed material to be removed from the system passes
through an indirect solids cooler and is discharged to a pneumatic
system for disposal. The properly sized bed material is pneumati-
cally carried up to a bed material storage tank which contains
-------
224
CLEM COMBUSTION OF COAL
|STEAM_r-fc~lJ0-§k?CIROSTATIC
CYCLONE VACUUM PRECIPITATOR
SEPARATOR EJECTOR
COOLING WATER OUT
INDIRECT
SOLIDS
ILER
OVERSIZE TO
Dm DISPOSAL
Figure 6. Bed Material Recycle System.
multiple outlets so that bed material can be returned to each of the
four (4) steam generator cells. This system is designed to handle
hot bed material leaving the fluidized beds at up to 1093C (2000F).
INITIAL OPERATION
During the last three (3) months of 1976, checkout and initial cold mode
operation of most of the Rivesville facility systems took place.
Coal was first ignited in the D cell of the fluidized bed steam genera-
tor on December 7, 1976 with the fluidized bed temperature reaching a
maximum of approximately 790C (1450F) for a short period of time due to
insufficient fuel flow into the bed. Following several other short
coal firing periods at the early end of the operations learning curve,
coal fires were established and maintained for 33 hours on December 22
and 23, 1976. During this time bed temperature was controlled at an
average of 79C (1450F) and all systems performed well. The bed temp-
erature was easily controlled and responded well to changes in coal
feed rate. Coal firing rate was approximately 10% of the steam gene-
rator full load and a drum pressure of 8.275 MPa (1200 psig) was main-
tained while generating saturated steam at approximately 5 kg/s (40 000
Ib/hr). During this coal firing operation, firing was maintained with-
out any ignition assistance. Figure 7 indicates the operating condi-
tions during this thirty three (33) hour run at the end of which the
unit was shut down for a scheduled outage.
-------
FLUIDIZED BED STEAM GENERATION
225
300
D—CELL BED TEMPERATURE
First continuous run—Dec. 22-23, 1976
Coal Flow 2800-3000 Ib/hr
Saturated Steam Flow 40,000 Ib/hr
Drum Pressure 1200 psig
8 10 12 14 16 18 20 22 24 26 28 30 32 34 36
Time, hrs.
Figure 7. D-Cell Bed Temperature First Continuous Run
December 22-23, 1976.
The coal handling system is designed to handle a maximum of 5% moisture
coal. During January 1977 a shipment of wet coal caused operating dif-
ficulties and resulted in only intermittent coal firing operation.
Slide Gate Slide Gate Beds Equalized and
Closed Open C-Cell Ignited
Air
Distributor
C-Cell
400F
H
"
I — //// — '
D-Cell
1800F
Ly///-l
3 Ft.
I
C-Cell
400F-^
D-Cell
1800F
^
it * »
C-Cell
1550F
Jftl : .
D-Cell
1800F
in.
x s
Air Control
Dampers
(1) (2) (3)
Figure 8. Ignition of the C-Cell.
-------
226 CLEM COMBUSTION OF COAL
Ignition of the C cell is achieved by allowing hot bed material in the
D cell to flow through an opening in the partition wall between the D
and C cell as indicated on Figure 8. By starting with a deeper bed in
the D cell flow through the partition wall occurs rapidly and the bed
depth in each cell are equalized in approximately three (3) minutes.
Coal in the C cell is rapidly ignited by the hot material near the
slide gate opening and coal ignition propagates rapidly throughout the
C cell raising the bed temperature to its operating temperature.
The first successful ignition of the C cell occurred on February 10,
1977 and temperatures in the C cell were raised to approximately 870C
(1600F). Shortly after obtaining the 870C (1600F) bed temperature in
C cell, coal feed was lost, most likely due to the wet coal conditions.
Following this short coal firing operation in the C cell, the unit was
inspected and the boiler found to be in good condition. Problems with
the forced draft and induced draft fans, however, required repair and
no subsequent coal firing was attempted prior to this writing.
TEST PLAN
Following additional start-up efforts and upon achieving the ability to
operate the remaining boiler cells, test instrumentation will be in-
stalled at the Rivesville facility and a one (1) year test program will
be initiated. This test program will include determination of heat and
material balances at various operating loads and a 3,000 hour continuous
operation run. Installation of the test instrumentation should take
place during the third quarter of 1977.
SUMMARY
Initial operation has shown that stable combustion of coal can be main-
tained in large fluidized beds of limestone. As coal feed rate is con-
trolled, bed temperatures can be easily maintained and adjusted.
As of this writing, only preliminary operation has been achieved at the
Rivesville facility; however, results of this operation are encouraging
and indicate that upon coordinated operation of the systems the Rives-
ville unit should operate satisfactorily.
REFERENCES
Gamble, R. L. and Warshany, F. R. "Commercial Development of Atmospheric
Fluidized Bed Utility Steam Generators" September 1975, 1975 Joint Power
Generation Conference, Portland, Oregon
Mesko, J. E., "Multicell Fluidized Bed Boiler Design, Construction and
Test Program", Monthly Status Report #51, December 1976, ERDA Contract
No. E(49-18)-1237, Report FE-1237-62
-------
227
DEVELOPMENT OF AN EFFICIENT SOLIDS FUEL BURNER
by
Norman A. Lyshkow
Combustion Equipment Associates, Inc.
INTRODUCTION
Much interest has been directed to the development of solids
burners for use in various chemical processes. The compatibility of
the heat source with the particular requirements of the process neces-
sitates detailed study of the chemical system.
The emissions of the heat source must be analyzed in order to
predict their probable effect on the product. In certain cases, emission
effect is minimal; however, in the largest number of cases, solids
burners are sought to replace oil and gas in stringent processes.
In addition to melt applications, the requirements of these pro-
cesses vary from drying to supplying thermal energy for chemical reac-
tions in rotary kilns. Coal use has been limited due to product con-
tamination. The development of a combustor capable of providing a
"Clean Flame" for such applications has required a thorough study in both
the physical and chemical processes of solid fuel combustion. Removal
of fly ash within the combustor, while providing the necessary tempera-
ture for efficient combustion, directed development efforts to a slagging
type burner. As fly ash capture efficiency requirements generally vary
with the process, the slagging type burner allows for the most efficient
capture of fly ash as well as a practical method of ash removal. The
use of a wet wall combustor also simplifies the mathematical analysis
of the system.
Efficiencies in excess of ninety percent are required in many
processes; theoretical efficiencies greater than 99% are predicted with
pulverized coal. Extensions of the model combustor with the construc-
tion of new test combustors are planned. To date, particle efficiencies
have exceeded 95%» while the possibility exists for coal combustion with
particle emission levels below .1 Ib/MBTU.
The design also permits such efficiencies at combustor pressures
of less than two inches of water. Computer analyses have determined
dimensions of burners from 1 million to l60 million BTU, with larger
sizes practical.
Key to the use of solid fuels is an efficient combustor, capable
of utilizing low quality, high ash fuels. The use of such fuels requires
a method for qualifying them and predicting their combustion characteris-
tics in a real system.
-------
228 CLEAN COMBUSTION OF COAL
The need for these predictions and descriptions has promoted the
development of a computer model sufficiently complex to simulate the
combustion requirements and processes of fuel. The model input must be
the most complete fuel description, thus necessitating the development
of specialized instrumentation and procedures.
This information development must then be described in a series
of equations. These equations must compensate for the variable system
conditions. Initially, the equations were of the most simple type.
Further expansions were tested in a model combustor to predict their
accuracy.
Key to the prediction of accuracy was both carbon and particulate
emission. Emission testing by EPA methods and the use of an Anderson
impactor provided this correlation. The computer model devised has
expanded to approximately six hundred lines with many simultaneous iter-
ation loops. The pitfall of such a complex model is the necessity to
continually maintain relevance. The complexity of solids combustion, a
dynamic situation, cannot be simply described. The only constant
criteria of such a model can be its prediction of accurately sized
burners which meet the requirements of a successful system.
FUEL ANALYSIS
Elemental and proximate analyses are utilized in the model. In
addition, a special device was constructed to ignite material under in-
tense radiation and determine ignition time. The device was constructed
to eliminate convection and conduction effects during irradiation. Other
devices were constructed to determine emissivity and envelope specific
gravity. Particle size analysis was performed and a technique developed
to determine effective particle thickness. Ash fusion temperatures were
determined to fix the temperature requirements of the combustor.
PROGRAM DEVELOPMENT
The motion and travel of a mathematically generated system of
particles to predict their interactions was developed, and includes the
following procedures:
(l) Continuous analysis of changing characteristics during the ignition-
combustion cycle is determined for the particles. The burner is
also analyzed by sector to determine energy balance of each sector.
(2) Centrifugal Force—drag force determinations—are utilized to
predict particle impaction points.
Since a path of movement for each group of particles is simulated,
a mathematical technique is utilized to determine the radius of curvature
and centrifugal force which predicts the reasonable particle cut of the
combustor. Tests using an Anderson impactor have found the combustor to
be in close correlation with the mathematical simulation. The model was
then directed to determine a flow field optimal for a given system of
particles. The flow analysis on levels of predicted turbulence. In later
studies of spectral emission, this assumption was reasonably verified.
-------
SOLIDS FUEL BURNER 229
FIELD STUDIES
A number of combustors designed by the computer model have been
constructed and tested using a vide variety of solid fuels. Burners have
been constructed from 3 million BTU's to kO million BTU's. A schematic
of the combustor (Figure l) is presented. The unit is basically a
cyclone-type burner. The burner is sized by the appropriate input data
(Figure 2) which generates the dimensions, flow, and volume considera-
tions for a particulate fuel at the desired combustion rate.
The output data is developed from the fuel characterization
(Figure 3) by the computer model.
A burner is then constructed to the dimensions provided by the
model and operated at the test facility which consists of a large water-
filled chamber with a steam vent, droplet extractor, circulators and
measuring means to determine water addition. In addition to the perform-
ance of material and heat balances, the test burner allows gaseous and
particulate testing.
Thermocouples are located throughout the system and in the chamber
gas effluent. Particulate is collected in the chamber and at the stack
sampling point. Data is then analyzed to determine correlation with
mathematical representations (Figure h).
Chemical data on emissions and slag has been collected (Figure 5).
Further studies in emission are planned.
To date, low NOX values have been found to exist. In addition,
concentration of sulfur in the slag has been observed and will be
studied in more detail.
SUMMARY
Continuous development in mathematical representations, prototype
construction, and materials of construction is underway on this project.
The success of the burner design utilizing fuels as high as 35% ash and
BTU contents as low as 5,500 with resultant particulate emissions as low as
.3^ Ib/mm BTU illustrate the possibilities for utilization of the design
during the present energy situation.
-------
230
CLEAN COMBUSTION OF COAL
PRIMARY AIR VANES
SOLIDS ENTRY
SECONDARY AIR VANES
(IF REQUIRED)
DIRECTION
OF
FLOW
i *
TT
A
SUPPORTING FUEL NOZZLES
(IF REQUIRED)
Figure 1.
-------
SOLIDS FUEL BURNER 231
LINKING SYSTEM LIB
ENTER BTU/HR (IN MILLIONS)
73
ENTER BTU/LB
712260
ENTER TYPE OF SUPPORT FUEL
0 -- FOR #2 OIL
G -- FOR GAS
?G
ENTER % CARBON
770.72
ENTER % HYDROGEN
77.36
ENTER % OXYGEN
714.73
ENTER % FIXED CARBON
743.24
ENTER % ASH
78
ENTER ASH SPECIFIC GRAVITY
71.4
ENTER ASH PARTICLE SIZE IN MICRONS
7.2
ENTER ASH FUSION TERMPERATURE FLUID
72350
ENTER % MOISTURE
710
ENTER PARTICLE EMISSIVITY
7.95
ENTER MATERIAL SPECIFIC GRAVITY
71.2
TEST IGNITION TIME MATERIAL AS RECEIVED
7.162
ENTER IGNITION TIME MATERIAL DRY
7.162
ENTER LIFTING VELOCITY of 90% - TILE PARTICLE
765
ENTER 90% - TILE PARTICLE SCREEN SIZE IN MICRONS
7100
ENTER 50% - TILE PARTICLE SCREEN SIZE IN MICRONS
730
ENTER 10% - TILE PARTICLE SCREEN SIZE IN MICRONS
75
ENTER % FOSSIL FUEL
75
ENTER WALL THICKNESS
7.75
ENTER CONE ANGLE
7107
Figure 2. Program Input Data.
-------
232 CLEM COMBUSTION OF COAL
LBS OF SOLID FUEL/HR 232.4633
CUBIC FEET OF GAS/HR 149.5215
CFM AIR SOLIDS VANE 477.2646
CFM SUPPORT FUEL 24.1726
TOTAL CFM COMBUSTION AIR 501.4373
AERODYNAMIC .8271E-04
CYLINDER DIAM. INCHES 8.6467
CYLINDER WALL THICKNESS INCHES .7500
MAJOR DIAM. OF BURNER 18.1491
LENGTH OF IGNITION SECTION INCHES NONE
LENGTH OF COMBUSTION SECTION, INCHES 7.2972
WIDTH OF SUPPORTING FUEL VANE, INCHES NOT USED
ANGLE OF INJESTION IN DEGREES 69.3950
INJECTION VELOCITY FEET/MINUTE 3286.0000
WIDTH OF SOLIDS FUEL VANE, INCHES 2.5624 COMBINED
OPTIMAL NUMBER OF INJECTION PORTS 4
CONE PARTICLE CUT IN MICROS 1.6099
Figure 3- Computer Generated Sizing Data.
PARTICLE SIZE CORRELATION
ARITHMETIC MEAN SIZE OF COLLECTED % OXYGEN
FLY ASH IN MICRONS
2.5 3.5
3.6 0.0
2.8 1.3
3.2 1.0
PREDICTED 2.21 MICRONS
Figure U.
-------
SOLIDS FUEL BURNER 233
SUB BITUMINOUS COAL
FLY ASH AS PER CENT OF TOTAL ASH
4.2
FLY ASH CARBON
21.26%
TOTAL PARTICULATE EMISSIONS IN LB/MBTU
.40
S02 EMISSION
392.5 PPM
.66 LB/MBTU
THEORETICAL S02
1.64 LB/MBTU
40% OF AVAILABLE SULFUR EMITTED AS S02
NOX EMISSION
252 PPM
.30 LB/MBTU
THEORETICAL NO (BY FUEL NITROGEN)
.86 LB/MBTU
35% OF AVAILABLE FUEL NITROGEN EMITTED AS NOX
Figure 5- Particulate Emissions.
-------
234 CLEAN COMBUSTION OF COAL
-------
235
SESSION IV - POSTCOMBUSTION CLEANUP
SESSION CHAIRMAN: Sidney R. Orem, Industrial Gas Cleaning Institute
The previous sessions have dealt with preparing coal, cleaning it
and burning it in a variety of ways. The papers in this session high-
light the present state of technology on backend cleanup of products of
combustion. Topics include two methods for high efficiency particulate
control, flue gas desulfurization and flue gas treatment for NOX control.
Some of you no doubt heard that the Senate/House Conference
Committee reached a compromise report on the Clean Air Act Amendments
at 2:15 a.m. yesterday. Action is expected in both houses before ad-
journment tomorrow for the August recess.
-------
236 CLEAN COMBUSTION OF COAL
-------
237
"ELECTROSTATIC PRECIPITATION
STATE OF THE ART"
by
Dr. R. S. Atkins & D. V. Bubenlck
Research-Cottrell
It is extremely appropriate during this symposium on
the Clean Combustion of Coal to discuss electrostatic pre-
cipitation. Electrostatic precipitation has played a major
role in controlling coal-fired boiler emissions and will
continue to play a significant role in the future. In
fact, if it were not for this technology, environmental
pressures would not have permitted the coal-fired boiler
market to have grown to its present size (figure 1).
Likewise, many significant improvements in this and other
pollution control technologies developed as a result of
the "1970 Clean Air Act" which necessitated stricter pollu-
tion control requirements and performance standards. The
precipitator was first applied to fly ash control problems
in 1923 but many significant technological advances are
occurring now. With government and economic pressures for
a more coal-dependent economy, we will continue to see
many more improvements in electrostatic precipitation to
meet these needs.
The electrostatic precipitator is the major high ef-
ficiency particulate control device for coal-fired boilers.
Figure 2 illustrates that as late as the 1960's, the
average user required only about 97.5% particulate collec-
tion efficiency. With today's pollution control laws .
we are designing and supplying systems with an average re-
moval efficiency of 99.5% and in some instances a design
requirement of 99.9%.
The demand for higher efficiencies, the use of more
strip-mined low sulfur western fuels, the site-specific
situations and their resulting economics necessitate the
consideration of alternate pollution control devices.
Baghouses, scrubbers, electrostatic precipitators and com-
binations thereof are offered by the company I represent
depending upon customer needs, pollution codes and econom-
ics. Each situation requires the selection of an appro-
priate control device strategy. However, today I will
only discuss the electrostatic precipitator, where it is
and where it is going.
-------
to
CO
450
900
r
UJ
I
UJ
o
u.
LJ
Z
O
o
u
_J
_l
o
u
99.9
99.8
99.7
99.5
99.3
99
98
97
95
93
9O
MAXIMUM
AVERAGE _
o
o
§
CQ
1-3
a
o
1920 1930 I94O 1950 I960 1970 1980
YEAR
1920 1930 1940 1950 I960 1970 1980
YEAR
FIGURE 1. COAL CONSUMPTION AND INSTALLED
FLY ASH PRECIPITATOR CAPACITY
FOR PUBLIC UTILITIES IN THE
UNITED STATES, 1920-1975.
FIGURE 2. TRENDS IN THE AVERAGE AND
MAXIMUM COLLECTION EFFICIENCIES
OF FLY ASH PRECIPITATORS IN THE
UNITED STATES.
-------
ELECTROSTATIC PRECIPITATION 239
What is an Electrostatic Precipitator?
The conventional coal-fired boiler electrostatic pre-
cipitator as we know it today is a rectangular configura-
tion with the gas flowing between parallel grounded plates,
called collecting electrodes, interdispersed with regularly
spaced discharge electrodes. Usually the discharge elec-
trodes are flexible weighted wires, rigid masts or other
electrode geometries which are of negative polarity to
provide a source of electrons which produce ions for
charging the dust particles.
The charged dust is electrically attracted to the
collecting plates where it deposits and is removed from
the gas stream. Collected particulate is released from
the collecting and discharge electrodes during on-stream
operation by rapping and vibrating devices. The collected
dust falls by gravity into hoppers located beneath the
precipitator. The system is designed for continuous oper-
ation with multiple ducts and chambers in parallel to
handle the volume of gas and different lengths of treatment
to achieve desired collection efficiency levels
(Figure 3).
The precipitator can be located up or downstream of
the boiler air heater and respectively operated at 650 to
850°F (hot precipitator) and 250 to 350°F (cold precipi-
tator). The face velocity through the precipitator typi-
cally ranges from 3 to 6 ft/sec. Various discharge and
collecting electrode geometries are used depending on the
dust properties. Most units use 8 to 12 inch plate
spacing with various treatment lengths depending on the
degree of high tension sectionalization required. Dis-
charge electrodes are equally spaced on centers between
the plates. Full and half wave D.C. current is used to
negatively charge the discharge electrodes with respect to
the grounded plates. Typical levels of corona power
density range from 0.5 to 3.5 watts/ft of collecting sur-
face corresponding to resistivity levels of 10 to
10 ohm-cm.
Today's precipitators have a combined charging and
collecting function in the same treatment length. This
has Resulted in a compromise between charging and collec-
tion properties to obtain optimized single-staged perfor-
mance. Theoretically better performance may be obtained
by designing units with independent charging and collec-
ting sections as illustrated in Figure 4 to optimize their
individual needs. Two-staged precipitator research was
conducted many years ago but progressed no further than
the laboratory, possibly because the then moderate per-
formance requirements could not justify the increased
costs. However, work in this area is again being con-
-------
-P-
O
H.V. SYST.
SUPPT. INSUL
RAP INSUL.
BUS DUCT
INSUL.
CHAMBER
(TYP)/
o
o
CO
1-3
H
O
O
o
FIGURE 3, TYPICAL PRECIPITATOR CONFIGURATION,
-------
ELECTROSTATIC PRECIPITATION
241
CONFIGURATION
GAS
FLOW
GAS
FLOW
SINGLE-STAGED PRECIPITATOR
• •
• •
TWO-STAGED PRECIPITATOR
OPTIMUM PERFORMANCE REQUIREMENTS
CHARACTERISTICS
CURRENT
FIELD STRENGTH
MAX, FIELD
SPACE CHARGE
DUST
CHARGING
SECTION
HIGH
NON-UNIFORM
AT CATHODE
LOW
NONE ON ANODE
COLLECTING
SECTION
LOW
UNIFORM
AT ANODE
HIGH
COLLECTED
ON ANODE
FIGURE L\, COMPARISON OF SINGLE-STAGED AND TWO-STAGED
PRECIPITATORS,
-------
242 CLEAN COMBUSTION OF COAL
ducted by EPRI, EPA, research institutes and precipitator
manufacturers to meet the demand for more efficient, less
costly equipment. By the end of this decade, large-scale
two-staged precipitators with enhanced charging and collec-
ting sections will probably be commercially available.
These units may be dry systems or combined dry/wet units
with additional capabilities for fine particulate removal,
reduced reentrainment and gaseous pollutant control. The
two-staged precipitator will be less costly and less sensi-
tive to particle composition.
In conjunction with High Voltage Engineering we are
applying pulsed techniques to improve corona generation
and the resulting collection of fly ash. The results of
our studies continue to be extremely promising. It has
shown that this new electrostatic method greatly increases
particulate collection at reduced precipitator sizing.
Several other projects in our laboratories have been
aimed at control of particulates from the newer combustion
methods such as low and high Btu gasification units, atmos-
pheric and pressurized fluidized bed boilers, MHD genera-
tors and combined-cycled systems. Each of these processes
has its own distinct collection problems which are some-
what dissimilar to conventional boiler applications. For
example, in fluidized bed combustion it is desirable to
collect the particulate at elevated temperatures of 1000
to 1500°F and at pressures of 10 to 20 atmospheres. The
flue gas stream contains fly ash, high levels of car-
bonaeous materials and reacted and unreacted limestone.
Each of these materials will have a different effect on
the performance of the precipitator.
One of the projects that we are working on under an
EPA contract is to develop a precipitator for operation
up to 2000°F and 500 psia. Our bench-scale unit has
successfully operated generating higher than expected
corona current levels. We are now seeking a pilot plant
demonstration. Today, research is being geared to provide
the needed know-how and technology to serve the future
coal conversion market place.
What Affects the Design of an Electrostatic Precipitator?
In the design of an electrostatic precipitator, the
range of conditions over which the system will operate
must be specified. As indicated in Table 1, information
on gas flow rate, temperature, and flue gas analysis as
well as chemical composition and electrical characteristics
of the dust, particle size distribution, and mass loading
should be obtained. In many cases this information is not
available, extremely general or very variable. In these
instances, assumptions based upon fuel analyses, ash
-------
ELECTROSTATIC PRECIPITATION 243
TABLE 1. DESIGN PARAMETERS AND DESIGN CATEGORIES
FOR ELECTROSTATIC PRECIPITATORS.
PERFORMANCE-RELATED PARAMETERS
GAS FLOW
GAS TEMPERATURE
GAS (TREATMENT) VELOCITY
SCA
OVERALL MASS COLLECTION EFFICIENCY
FRACTIONAL MASS COLLECTION EFFICIENCY
INLET GRAIN LOADING
OUTLET GRAIN LOADING
GENERATED PLANT POWER OUTPUT
FUEL BURNING RATE
FIRING METHOD AND COAL CHARACTERISTICS
FIRING METHOD
% ASH
% SULFUR
% MOISTURE (AS RECEIVED)
BTU/LB (WET)
SAMPLE SOURCE
ASTM CLASS
MINE, STATE
MINE, COUNTY
MINE NAME
SEAM NAME
ASH CHEMICAL ANALYSIS
Si02 Na20
A1203 Li2O
Fe203 P205
TiO2 S03
CaO SAMPLE SOURCE
MgO MEAN
K90 DEVIATION
SAMPLE TYPE
-------
244 CLEAN COMBUSTION OF COAL
content, Btu rating, mine source, seam type, boiler con-
figuration, firing methods and efficiency levels must be
made prior to selecting precipitator design parameters.
Specifications which are too general, or averages of a
wide variety of proposed fuels lead to very conservative
equipment sizing. However, when many sets of data con-
sisting of ultimate, proximate and ash chemical analyses
are provided for each fuel and ash source, the supplier
can more adequately design a unit to meet the customer's
reliability and efficiency needs. Site-specific design
constraints such as space limitations, equipment location,
minimum velocity given the ductwork configuration, etc.,
must also be considered.
For many applications precipitator design parameters
are known from prior experience with the same coal. In
some cases, estimated parameters must be developed from
prior experience with similar fuels. However, problems
in meeting design efficiency levels do develop when the
fuel being fired differs significantly from that used for
the initial design. If better information is made avail-
able in the design stages, good performance can be ex-
pected with the operating unit.
Ash Resistivity
The electrical properties of the dust often analyzed
and reported as resistivity are extremely important in
designing a precipitator. For optimum precipitator per-
formance the ash resistivity should be in the range of 10
to 10 ohm-cm. Figure 5 indicates the typical effect of
sulfur and flue gas temperature on ash resistivity. As
can be observed, the ash resistivity increases with de-
creasing quantities of sulfur and varies significantly
with temperature. Also, factors such as flue gas moisture
and alkali, Na~0 and P2^5 contents as well as Fe20~ levels
in the ash have a major effect on particle conductivity
and collection.
With low sulfur fuels, the likelihood of back corona
occurring increases. Back corona causes a reduction in
the precipitator operating voltage and current levels
which in turn decreases its performance. Also, higher
resistivity ash requires more intense rapping to remove
it from the collecting electrodes and thereby increases
the possibility for dust reentrainment, structural plate
failure and equipment maintenance. Hot precipitators
limit back corona and ash rapping problems by operating
at higher temperatures and under more optimum resistivity
conditions.
-------
ELECTROSTATIC PRECIPITATION
245
10
12
10"
o
i
S
X
o
en
I09
/•--0,5-1,0 I S COAL
-1,5-2,0 % S
COAL
,-2,5-3,0 I S COAL
200 250 300 350 400
TEMPERATURE, °F
450
FIGURE 5, EFFECT OF SULFUR CONTENT IN COAL ON FLY
ASH RESISTIVITY,
-------
246 CLEAN COMBUSTION OF COAL
Another method to enhance particulate collection is
to modify the resistivity of the fly ash.4 This technique
has been successfully applied in more than 15,000 MW of
commercial installations by the addition of chemical ad-
ditives to the flue gas upstream of the precipitator. S03
addition has been used to reduce the resistivity of low
sulfur fuels. NH3 conditioning has been used to increase
the resistivity and cohesiveness of fly ash from high
sulfur fuels and thus decrease its reentrainment during
rapping. Sodium salts also have been successfully applied
on a hot precipitator application to reduce the resis-
tivity of the fly ash and improve electrical performance.
We have observed that sodium compounds at elevated tem-
peratures can react with other trace constituents of the
ash thereby modifying their electrical conduction. Each
of these conditioning techniques can increase the relia-
bility and flexibility of a precipitator to handle a wider
range of conditions. It is expected that during the next
several years, ash conditioning will be specified as part
of some new precipitator installations.
Migration Velocity
Whereas resistivity is a. rough indicator of precipi-
tator performance, effective migration velocity is used
to specify and determine precipitator size. We have
developed a good deal of know-how and experience to re-
late coal, fly ash and flue gas properties to migration
velocity.
The conventional Deutsch-Anderson equation was used
for many years to relate migration velocity with collection
efficiency
n - 1 - e-(A/V)w
where
w = Particle migration velocity, (ft/sec)
V = Gas flow rate, (actual ft /sec)
o
A = Effective collection surface,(ft )
n = Overall mass collection efficiency, (fractional).
The migration velocity w is a function of electrical ener-
gization and overall mass collection efficiency. The
variation in w within a given precipitator is caused by
changing particle size distribution as precipitation pro-
ceeds in the direction of gas flow. Since the requirement
of high collection efficiency corresponds to collection
of submicron particles, it is understandable why w de-
creases with increasing efficiency requirement.
-------
ELECTROSTATIC PRECIPITATION 247
The modified migration velocity, wk , as presented by
Matts and Ohnfeldt5 can be treated essentially as a con-
stant for any application. We use the modified migration
velocity as a more accurate means of predicting precipita
tor performance.
The following equations used in sizing prec ipitators
relates modified migration velocity with specific collec-
tion area (SCA) and efficiency.
SCA = 16'67 In2(l-n)
Wk
n =|~1 - X (H.V.) _ 100
L (Ash) (A.C.)J1UU
where
SCA = Specific collection area (ft2/1000 ACFM) =
16.67 (A/V)
H = Overall mass collection efficiency, (percent)
i
Mod
ified migration velocity, (ft/sec)
X = Emission standard, (lb/106 Btu)
H.V. = Heating value of the coal , (Btu/lb )
Ash = Ash in the coal, (fraction by weight)
A.C. = Ash carryover, (fraction by weight)
The required overall mass efficiency, therefore, is
a function of the coal heating value and ash content as
well as the fraction of ash carryover, which is a function
of boiler type. The modified migration velocity is a
function of electrical energization of the precipitator
and of gas properties. It is often conveniently linked
with resistivity level, such that for a moderate resis-
tivity of 10^ ohm-cm the value will be between 1.6 and
1.9 ft/sec whereas for a very resistive dust it may
approach 0.5 ft/sec.
Resistivity plays a significant role in the selection
of wk and power density. It is generally predicted from
correlation of ash components and knowledge of environmen-
tal conditions such as moisture content, gas phase com-
position, and temperature. Design exponents based on re-
gression analysis are often used in relating coal pro-
perties to w , SCA, and power density.
1C
-------
248 CLEM COMBUSTION OF COAL
Several Selected Case Illustrations
To illustrate some of the points made in the prior
discussion, several western low sulfur cold precipitafcor
sizing examples will be discussed. Western low sulfur •-.
coals are becoming more important to the U.S. energy
picture and have their attendant particulate collection
problems. The selected cases represent some of the
choices available to the supplier and user in developing
an appropriate control strategy.
Table 2 illustrates three types of western low sulfur
fuels each exhibiting different precipitator requirements.
Fuel A fly ash has a low silica content with a moder-
ately high level of Na20 present mainly in the form of
sulfates. This condition in lignite coal fly ash yields
a favorable conductivity effect; hence the ash has low
resistivity and is easily collected with a cold precipita-
tor. Both from technical feasibility and economic view-
points no other control device option need be considered.
Fuel B is a Colorado subbituminous coal. The fly ash
analysis indicates a moderate Na,,0 level and a low base
(< 20%) and low base/Na20 ratio (< 10). This suggests
that no special precaution must be taken to ensure good
operation. However, the high Si02 content tends to sup-
press the conditioning effect of the sodium present. As
a result, the resistivity is higher than moderate sug-
gesting that cold, hot, and cold SO., conditioned electro-
static precipitators are all technically workable options.
The hot Na»0 conditioned option would not be required be-
cause of the favorable base/Na.O i-atio which generally
suggests that hot precipitators will perform well.
Fly ash analysis of fuel C, a Wyoming subbituminous
coal, indicates a base/Na00 ratio in the marginal ranged
suggesting potential for electrode fouling with hot pre-
cipitator operation. Coupled with this is a high P-0
content which is responsible for corona quenching and
power-limited hot precipitator operation. The high re-
sistivity shown for fuel C in Table 3 means that a
large cold precipitator would be required. Improved pre-
cipitator operation and smaller precipitator size for this
type of ash can be obtained by increasing the sulfate
content using SO- conditioning. Sodium salt conditioning
can be used to condition the ash to reduce its base/NaoO
ratio and bring it into an improved hot precipitator
operating range. In the final analysis, given the tech-
nical feasibility of options to be considered, the optimum
control device must be determined on the basis of econom-
ics and reliability.
-------
ELECTROSTATIC PRECIPITATION
249
TABLE 2. TYPICAL SPECIFICATIONS ON SELECTED WESTERN
FUELS.
Fuel A
Fuel B
Fuel C
Fuel Characteristics
Rank
Location
As Fired
% Moisture
% Ash
% Sulfur
Heating Value
(Btu/lb)
Fly Ash Analysis, %
SiO,
Fe2°3
CaO
MgO
K2O
Na2°
SO o
P2°5
Base
Base/Na20
nite
th Dakota
23.9
6.7
.8
8135
20.6
15.6
.6
9.2
31.0
8.8
3.0
3.4
6.6
1.2
55.4
16.3
Subbit.
Colorado
14.8
5.1
.4
10730
45.1
23.2
.8
5.6
8.8
1.7
.6
2.5
10.7
1.0
19.2
7.7
Subbit.
Wyoming
28.0
8.5
.5
8200
33.0
15.1
1.0
6.1
21.6
4.0
.6
1.1
16.0
1.5
33.4
30.4
-------
250
CLEAN COMBUSTION OF COAL
TABLE 3. COLD PRECIPITATOR SPECIFICATIONS FOR
SELECTED WESTERN LOW SULFUR COALS.
FUEL
COLD
COLD PPRT.
SPECIFICATIONS
LIGNITE
NO. DAKOTA
B
SUBBIT
COLORADO
SUBBIT
WYOMING
SCA @ 99.5%
wk, (ft/sec)
p , (ohm-cm)
n @ .1 lb/106 Btu
n @ .05 lb/106 Btu
250
1.87
1 x 109
98.76
99.39
440
1.06
6 x 10
97.90
98.95
10
800
0.59
3 x 10
99.04
99.52
12
Figures 6 and 7 indicate the relative capital in-
vestment ($/KW) and annual cost (mills/KWH) for these
various options. From a strictly economic standpoint,
it can be observed that SO conditioning in conjunction
with cold precipitation, wnen feasible, is the lowest
capital cost approach. However, because of the relative
newness of this approach, users are concerned with its
long-term reliability. Hot precipitation can be an at-
tractive option and does offer fuel flexibility and relia-
bility. Care must be taken not to universally apply any
one solution since we have observed that ash properties
can affect even the performance of hot precipitators.
Before selecting a particulate control strategy, the
various fuels to be combusted in the boiler should be
reviewed and their ashes analyzed. A study should be made
to determine particulate equipment sizing and its flex-
ibility in handling changes in fuel and in boiler oper-
ation. Capital and operating costs should be calculated
and only then should a particulate control strategy be
select ed .
-------
ELECTROSTATIC PRECIPITATION
251
I-
UJ
en
UJ
Q.
<
O
80
70
60
50
40
30
25
20
15
10
9
8
7
I i i—I—I—|MI|
1—r
95.0
98.0 99.0.
EFFICIENCY, %
99.8 99.9
FIGURE 6, ECONOMIC COMPARISON OF CAPITAL INVESTMENTS
FOR VARIOUS ELECTROSTATIC PRECIPITATOR
OPTIONS FOR A 600 flW POWER PLANT,
-------
252
CLEM COMBUSTION OF COAL
95.0
98.0 99.0
EFFICIENCY, %
99B 99.9
FIGURE 7, ECONOMIC COMPARISON OF ANNUAL COSTS FOR
VARIOUS ELECTROSTATIC PRECIPITATOR OPTIONS
FOR A 600 RW POWER PLANT,
-------
ELECTROSTATIC PRECIPITATION 253
References
1. White, H. J. "Electrostatic Precipitation of Fly Ash:
Outlook for Future Growth" JAPCA Vol. 27, No. 1,
(January 1977) .
2. Ibid.
3. "Technology for Electrostatic Freeipitators," Indus-
trial Gas Cleaning Institute Inc., Publication .
No. E-P1.
4. Atkins, R. S. and D. H. Klipstein, "Improved Precipi-
tator Performance by S0~ Gas Conditioning," Proceedings
of the American Power Conference, Vol. 37 (1975),
pp. 693-700.
5. Matts, S. and P-0 0*hnfeldt, "Efficient Gas Cleaning
with SF Electrostatic Precipitators."
6. Walker, A. B. "Operating Experience with Hot Precipi-
tators on Western Low Sulfur Coals," presented at the
American Power Conference, Chicago, Illinois
(April 18-20, 1977) .
7. Bubenick, D. V. "Economic Comparison of Selected
Scenarios for Electrostatic Precipitators and Fabric
Filters," presented at the 70th annual meeting of
AP,CA, Toronto, Ontario, Canada (June 20-24, 1977).
-------
254 CLEAN COMBUSTION OF COAL
-------
255
STATE OF THE ART — FABRIC FILTRATION
Richard L. Adams
Vice President, Systems and Technology
Wheelabrator-Frye Inc.
Air Pollution Control Division
INTRODUCTION
If this conference had "been held five years ago, there would not
have been a paper entitled "State of the Art — Fabric Filtration."
Five years ago, electrostatic precipitators were the only particulate
removal device seriously considered for high efficiency removal of fly
ash generated by the burning of coal. Certainly, a few pilot fabric
filters had been in successful operation by this time but the first
full-scale units were still on the drawing board. Today, fabric
filters are being installed on both utility and industrial coal-fired
boilers. Since fabric filters are probably one of the oldest means of
removing solid particulate from the gas stream, what are the reasons
for the long delay in their consideration for use on coal-fired boilers
and their quick acceptance once the barrier was broken?
The promulgation of the laws pertaining to new source emissions by
the EPA in 1970 brought about requirements for emission control that
were in most cases substantially more stringent than those which had
been in existence up to that time. These requirements for increased
efficiency of the pollution control devices substantially increased
the size and thereby the cost of an electrostatic precipitator instal-
lation. In addition, new sulphur regulations have increased the usage
of low-sulphur, Western coal and this has also placed an increased
burden on an electrostatic precipitator. Finally, because of earlier
difficulties with precipitator performance, a greater degree of con-
servation was included in precipitator sizing and specifications. The
net result of all of these changes was to cause a marked increase in
the cost of an electrostatic installation and suddenly the fabric fil-
ter was more than competitive in certain areas. Figure No. 1 shows
the cost of both the fabric filter and an electrostatic precipitator
as a function of plant size and in the case of precipitators, specific
collecting area. It will be noted that where a cold electrostatic
side electrostatic precipitator has a requirement for an SCA above
approximately 500 the fabric filter will normally result in a less
expensive installation.
Today, there are several fabric filters in operation, some with
operating histories of over four years. Within the next year, many
large units (up to 575 megawatts) are scheduled to go into operation.
This paper will briefly trace the history of fabric filtration on
coal-fired boilers over the past five years and present our thoughts
as to the future.
-------
256
CLEAN COMBUSTION OF COAL
34-
28-
226-
S 24-
i 22-
UJ "
\ 20-
>
S 18-
2 16-
? 14-
c
-------
FABRIC FILTRATION 257
TYPES OF FABRIC FILTERS
For those somewhat unfamiliar with fabric filtration, it should be
pointed out that there are two types of fabric filters that may be con-
sidered for use on a coal-fired boiler. These are the so-called high-
energy fabric filter and the low-energy fabric filter. In a low-energy
fabric filter, as shown in Figure No. 2, filtration is accomplished by
building a filter cake using a woven filter fabric as a grid or matrix
to support this cake. The cake is intermittently removed from the fil-
ter fabric by use of mechanical agitation or by backwashing the cloth.
In a high-energy fabric filter, as shown in Figure No. 3, a felt
is utilized which acts as a depth filter as well as supporting a cake
on the surface of the felt. In order to remove the collected material
from this type of filter fabric, it is necessary to apply a great deal
more energy, thus the term high-energy fabric filter. Normally, this
energy is applied by utilizing compressed air which will both agitate
and backwash the filter fabric and remove the filter cake. Currently,
high-energy filters have been applied only to the smaller industrial
boilers. The filter fabric utilized has been felted Teflon™ or a
heavy woven fiberglass. We would expect to see, however, a greater
use of high-energy fabric filters in this service in the future.
To-date, most of the major installations in the utility industry
in this country have utilized low-energy filtration. Fiberglass is
the normal filter fabric employed in low-energy filters but there is
a great deal of discussions currently with regard to the proper finish
on the fiberglass fabric. There are also two very different fabric
cleaning systems used today on low-energy fabric filters. There is a
system called "deflate and shake" and a system utilizing "reverse
flow." Both these will be described later. It is interesting to note
that while fiberglass has been the normal fabric in this country,
synthetics such as acrylics and polyesters have been used with a great
deal of success 'in Australia.
To this day, fabric filtration is still more of an "art" than an
engineering science- There are approximately ^0 variables that can
affect the performance of a fabric filter and most of these cannot be
predicted at the time of initial design. We are thus in the position
of having to rely heavily on actual field operating experience to make
judgments as to the proper application of fabric filters. Just as
there is little or no correlation between resistivity and the precipi-
tator sizing factor "W" there is also little correlation between
laboratory values of "K" factor or filter drag and actual field per-
formance. It might be well, therefore, for us to review three fabric
filter installations on coal-fired boilers which have been in operation
for sufficient time to allow us to make judgments as to their perform-
ance.
OPERATING EXPERIENCE
In reviewing these installations, we will examine the operating
problems encountered with each. However, these problems have been
-------
258
CLEAN COMBUSTION OF COAL
Figure 2
Low Energy
Filter
-------
FABRIC FILTRATION
259
Figure 3
High Energy
Filter '
-------
260 CLEAN COMBUSTION OF COAL
minor and it should be noted that none of these units has caused the
loss of production on a single kilowatt of electrical power. When
compared to the records of other types of control devices being ap-
plied to utility boilers, we believe this is an enviable record.
The first full-scale fabric filter installed in the United States
on a coal-fired utility boiler is the installation made at Pennsylvania
Power & Light's Sunbury Station. There are four units, each handling
222,000 ACFM % 325°F from two 87.5 megawatt boilers. The first of
these units was started in February of 1973. The detail design data
for these units is given in Table Wo. 1.
The performance of the units at Sunbury since their initial start-
up has been excellent. The problems encountered have been related
primarily to design concepts developed for this job due to peculiar-
ities in space and plant layout, since it was a retrofit. The fabric
filter installed at Sunbury operates as a pressure unit, i.e., it is
on the discharge side of the I.D. fans, and the gases are forced through
the fabric under pressure and then exhausted to the atmosphere. Since
reverse gas cleaning is utilized, it is not necessary from a fabric
cleaning standpoint to have completely tight gas shutoff valves on each
compartment. However, the valves supplied on this installation should
also function to allow maintenance workers to enter the compartments
for inspection and bag replacement and since they are not gastight on
this installation, the only time it is possible to enter the compart-
ments without the use of protective equipment has been during boiler
shutdown. Adequate attention to this detail during the initial design
phase could have overcome the problem.
Since the Sunbury baghouses operate under pressure, they were also
designed such that the reverse air fan was located on the dirty gas
side of the unit and the reverse air fans handle fly ash laden gases.
Wear problems have been encountered with these fans and maintenance
in this area has been high. The fans have not, however, caused a
reduction in fabric filter availability.
The next units which we would like to examine were installed at
the Nucla Station of the Colorado-Ute Electric Association. In this
case, there were three units each ventilating a 13 megawatt stoker-
fired boiler. The detail design data for these units is given in
Table No. 2. The first of these units was placed into service in
December 1973 and the others followed in early 191^. These units were
suction-type with I.D. fans located downstream from the fabric filter
and none of the problems associated with the Pennsylvania Power & Light
units at Sunbury were encountered.
Shortly, after start-up, however, it was noted that there was ex-
cessive wear on the bottom of the fabric tubes. The type of tube
attachment used in these collectors is shown in Figure No. U. A study
of the problem was initiated and it was determined that the density of
the collected material was not 50 Ib. per cubic foot as specified by
the consulting engineer but rather approximately 25 Ib. per cubic
foot. This, coupled with the hopper emptying procedures, caused an
excessively high level of collected fly ash in the hoppers and thus
caused reentrainment and excessive wear on the bottom of the filter
-------
FABRIC FILTRATION
261
Table No. 1
Pennsylvania Power & Light
Sunbury, Pa.
Boiler Pulverized coal
Size 1*00,000 lb./hr. steam
222,000 acfm % 325°F
Coal Anthracite and coke
Design 2.07 to 1 air/cloth ratio
2 gr./acf, 2.5-3.5 inch w.g.
99-9$ plus
Boiler
Size
Coal
Design
Table No. 2
Colorado-Ute
Spreader stoker
Traveling grate
132,000 Ib./hr. steam
86,200 acfm/300°F
Ave. 0.7% sulfur (bituminous)
'35$ ash
3.35 to 1 air/cloth ratio
Ave. 1.5 grains
Ij-inch pressure loss, 99-9$ + design
-------
262
CLEAN COMBUSTION OF COAL
Figure 4
Tube Attachment
and Baffles
air side ^
baffle depth-8"
t
-------
FABRIC FILTRATION 263
bags. In order to correct the problem, the hoppers were lowered ap-
proximately two feet to provide added capacity and additional baffling
was installed. Finally, thimbles for each DUSTUBE which form pro-
tecting extensions were installed on the bottom of the tube sheets.
This arrangement is also shown in Figure No. IK
It should be noted that a fabric filter almost identical to those
at Colorado-Ute was also installed at this time on a coal-fired boiler
in the Nyssa, Oregon, plant of the Amalgamated Sugar Company. This
unit was put into service in September of 1973 and had a history very
similar to that described for the units at the Colorado-Ute Electric
Association. The operating data for the unit at Amalgamated Sugar is
given in Table No. 3.
The third unit to be examined is the Penna. Power & Light instal-
lation at Holtwood, Pennsylvania. This unit handles 200,000 CFM or
approximately 50% of the discharge from this 80 megawatt boiler. The
remaining 50% is handled in an existing wet scrubber. The unit became
operational in April 1975- The design and operating data is given in
Table No. k. Shortly, after start-up of the PP&L installation at
Holtwood, it was noted that there was a gradual increase in the oper-
ating pressure drop across the unit. This increase did not seem to
be related to any external operating conditions. An investigation into
the problem revealed that it was most likely related to the finish used
on the fiberglass filter fabric.
All of the early deflate-and-shake units, including the pilot
plants which had been operated over the prior ten years, had used
fiberglass filter fabric coated with a silicone graphite lubricant.
The filter bags in the PP&L installation at Sunbury, however, were
fiberglass-coated with Teflon B as a lubricant. The Sunbury bags
were equipped with rings and cleaned by reverse air only. It was
decided in view of the excellent operating history of Teflon-coated
fiberglass at Sunbury to utilize Teflon-coated fiberglass on the
Holtwood installation. It wasn't recognized that the finish might
not be compatible with the different cleaning method utilized at
Holtwood.
Investigation of the increasing pressure drop problem at Holtwood
indicated that it was most likely due to the Teflon B finish used on
the fiberglass bags. Photomicrographic studies showed that the Teflon
appeared to flow and to fill some of the crevices between individual
fibers. Figure No. 5 shows this situation. As a contrast, the
silicone graphite finish, Figure No. 6, appeared to leave the fibers
in a more open condition. It is unknown why the Teflon B finish
performed satisfactorily in conjunction with the ringed reverse air
cleaning at Sunbury and in a less satisfactory manner in conjunction
with the deflate-and-shake cleaning at Holtwood.
To confirm the results of the laboratory investigation, a complete
compartment in the Holtwood baghouse was fitted with silicone graphite
fiberglass fabric. Operating comparisons were made with respect to
volume and pressure drop between the compartment with the silicone
graphite finish and the compartments with Teflon B finish. The results
indicated that silicone graphite would operate at a substantially lower
-------
264
CLEAN COMBUSTION OF COAL
Table Ho. 3
The Amalgamated Sugar Co.
Boiler Spreader stoker
Traveling grate
Size 200,000 rb./hr. steam
91,800 acfm/300°F
Coal 0.5% sulfur (bituminous)
5$ ash
Design 3-56 to 1 air/cloth ratio
2.0 grains, 3-inch pressure loss
99.9% design
Table No. k
Pennsylvania Power & Light
Hpltvood, Pa.
Boiler Pulverized coal
Size 1/2 of 700,000 Ib./hr. steam
200,000 acfm 360°F
Coal 1.8% sulfur (anthracite)
Design 2.1+2 to 1 air/cloth ratio
J.h to 8 grains, 2.8-inch pressure loss
99-9$ + design
-------
FABRIC FILTRATION
265
Figure 5
Fibreglass With
Teflon®B Finish (10,000 X)
-------
266
CLEAN COMBUSTION OF COAL
Figure 6
Fibreglass With
Silicone Graphite Finish
(10,000 X)
-------
FABRIC FILTRATION 267
pressure drop when handling the same volume of gas. The Holtwood bag-
house was then changed so that 11 of the compartments, except for the
two control compartments which were left with the original Teflon B
finished bags, were converted to silicone graphite finish. This con-
version resulted in a substantial lowering of the operating pressure
drop of the unit.
CURRENT DESIGN PHILOSOPHY
It may be well to examine current design philosophies on fabric
filters to be applied to coal-fired boilers. It is far too early to
tell which of these design philosophies will prevail over the long
run. Each has advantages and disadvantages and only extended periods
of operation on many different types of boilers, fired with differing
coals, will give the ultimate answer as to which represents the best
approach to the problem.
While there is currently a lack of operating data, we believe
that pulse-jet type collectors, i.e., high energy equipped with heavy
woven fiberglass filter bags will be an attractive answer for some of
the industrial segments of this marketplace. It is possible that this
type of collector may extend itself into the large utility installa-
tions, but additional work history is needed before this type of unit
can be considered on a major boiler installation.
Among the low-energy fabric filters, there are two basic design
philosophies. One philosophy indicates the use of higher air-to-cloth
ratios and deflate-shake cleaning to maintain bag pressure drop. The
other philosophy uses lower air-to-cloth ratios and reverse air
cleaning of ringed bags. Generally speaking, the design air-to-cloth
ratio with deflate-shake cleaning would be approximately 3 to 1 and
the design air-to-cloth ratio for reverse air cleaning will be approxi-
mately 2.25 to 1 in order to achieve the same overall system pressure
drop.
The reason that shake-deflate cleaning can operate at somewhat
higher air-to-cloth ratios is because the bags are cleaned more uni-
formly. Figure No. 7 shows a ringed bag being cleaned by reverse air.
In the filtering mode, it is normal for the finer particles to end up
in the top of the bag and the heavier particles to be collected in the
lower areas of the bag. When reverse flow is applied, the filter cake
is removed from the lower surface of the bag first. This opens up the
filter fabric in this area and allows short-circuiting of the cleaning
airflow and thus interferes with the uniform cleaning of the filter
bag. No such phenomena exists in the mechanical agitation of a filter
bag where the mechanical energy is distributed evenly throughout the
length of the bag.
The Dorsey equation given in Figure No. 8 is the general formu-
lation for describing the performance of a fabric filter. In our
experience, we have found the Dorsey equation to be inaccurate but it
probably -still represents the best simple mathematical model for per-
formance of a filter. Unfortunately, the value "K" is extremely
difficult to define and in most cases can only be calculated after a
-------
268
CLEAN COMBUSTION OF COAL
Figure 7
Bag Cleaning
Methods
-------
FABRIC FILTRATION 269
FIGURE HO, 8
Dorsey Equation
operating pressure drop
AP=KCVXT
K is the specific dust-fabric filter resistance
coefficient, usually of order 1 to 10
(in WG/lb/ft2/ft/min),
C is the inlet dust concentration (Ib/ft3),
V is the average filtering velocity
(ft/min), and
T is the operating time between cleaning
of a particular compartment (minutes).
-------
270 CLEM COMBUSTION OF COAL
unit is in operation. Use of this equation, however, does allow us to
compare the relative performance of the two cleaning systems as they
are in operation at the Pennsylvania Power & Light Stations at Holtwood
and Sunbury. Both toilers burn similar coal. Table No. 5 shows the
comparative data between the two installations. We do not believe that
this comparison is quantitative but certainly in a qualitative way
indicates that the deflate-shake cleaning system provides substantially
lower "K" or better filter performance than does the reverse-air
cleaning system.
It has been suggested that bag life may be better with reverse-air
cleaning. The facts are not available to support this theory at the
present time, but if, indeed, it is true, then the relative merits of
the two methods will have to be judged on an economic basis. The
deflate-shake method will provide for a smaller and generally less
expensive initial installation. Should bag life be shorter, this is
partially offset by the increased cost of the bags in the reverse-air
installation due to the fact that there are a greater number of bags
and the bags have sewn-in anticollapse rings. We strongly urge poten-
tial users to make a complete economic evaluation between the two pro-
posed cleaning methods. Currently, this would appear to be the most
reasonable form of evaluation.
It should be pointed out that in light of the successful operation
of fabric filters on smaller boilers, there are currently several units
under construction on boilers in sizes ranging from 350 to 575 megawatts.
The first of these new large installations will go into operation in the
first quarter of 1978 at the Monticello Station of Texas Utilities. In
this case, there are two fabric filter systems each handling 80$ of the
discharge from a 575 megawatt boiler. The remaining 20% of the discharge
will be handled by existing electrostatic precipitators. The design
data on these units is given in Table Wo. 6. The utility industry is
watching with a great deal of interest these new, large installations
and once they are successfully on-stream, we believe that the fabric
filter will be considered a viable device for all size boilers.
SCL REMOVAL
No presentation regarding the state of the art on fabric filtration
with regard to coal-fired boilers will be complete without mentioning
that fabric filters are also being considered for the removal of SOp.
There have been successful pilot operations utilizing a fabric filter
coated with a naturally occurring sodium bicarbonate called Nahcolite.
These filters have been able to remove up to 90$ of the SOg entering the
system and cost analysis indicates that the total system will be more
than competitive with the existing wet scrubbing systems.
Finally, fabric filters are also proposed as particulate collectors
and chemical contactors following a spray dryer which can remove up to
90% of the S02 entering the system. We expect to see increased atten-
tion paid to both of these SOg removal processes in the near future.
-------
FABRIC FILTRATION
271
Table No. 5
Operating
Comparisons
K = AP/CT^T
Shake/Deflate
3.75" AP
8 gr./cu. ft. * TOGO C
2.8 fpm VX
UT.O min. T
8.9 K
Reverse Air
3.0"
2 gr./cu. ft. + TOGO
2.0 fpm
33.0 min.
79-5
Table Wo. 6
Boiler
Size
Coal
Design
Texas Utilities
Pulverized coal
2 boilers 575 MW each
1,8UO,000 acfmAOO°F each
1.2% sulfur (lignite)
13$ ash
2.9 to 1 air/cloth ratio
5.6 to 8.8 grains, h.5-inch pressure drop
99.9$ + design
-------
272 CLEAN COMBUSTION OF COAL
We believe that the almost-perfect on-stream reliability record of
fabric filters on coal-fired boilers combined with their very economical
first costs and operating costs will result in an ever-increasing
acceptance on both utility and industrial coal-fired boilers. In a
way, the use of fabric filters on boilers is in its infancy; however,
we have learned much over the past five years and we believe that the
time has now come where the fabric filter can take its place alongside
the electrostatic precipitator as a device that can economically meet
the most stringent emission requirements.
-------
273
STATUS OF FLUE GAS DESULFURIZATION
THE FEDERAL RESEARCH, DEVELOPMENT AND DEMONSTRATION PROGRAM
Julian W. Jones and Michael A. Maxwell
U.S. Environmental Protection Agency
Industrial Environmental Research Laboratory
Research Triangle Park, North Carolina
INTRODUCTION
One of the Nation's major energy-related environmental problems
concerns the need to control sulfur dioxide (S0_) emissions from
stationary fuel combustion sources. Flue gas desulfurization (FGD)
is the term used to denote processes for removal of S02 from flue
gas, usually by means of a gas scrubbing operation. FGD is the
major near-term technological approach to meeting new source per-
formance standards when utilizing high-sulfur coal supplies. In
addition, U.S. Environmental Protection Agency (EPA) studies indicate
that FGD is competitive in cost with advanced control methods, such
as chemical coal cleaning and fluidized bed combustion; therefore,
FGD should play an important role in controlling emissions even in
the 1990's.
The current Federal energy program calls for a policy of rapid
expansion in the use of coal. If this policy is to be successfully
implemented, it is essential that FGD technology be fully optimized
for application in the utility and industrial sectors.
In the Federal energy/environment research and development
program, FGD technology development has been given a high priority.
Several FGD studies, pilot plants, prototypes, and demonstration-
scale facilities have been funded by EPA. Although prior progress
had been achieved in FGD development, the overall pace of develop-
ment was increased by the initiation, in 1975, of the Federal inter-
agency effort.
The FGD program is being conducted through EPA's Industrial
Environmental Research Laboratory in Research Triangle Park, North
Carolina (IERL-RTP). The program has aimed at demonstrating reliable
and cost-effective FGD processes, including both nonregenerable
(throwaway) and regenerable (saleable by-product) systems. EPA's
key program in the nonregenerable area is the lime/limestone proto-
type test program at the Shawnee Steam Plant of the Tennessee Valley
Authority (TVA), near Paducah, Kentucky. Other major activities in
the area of nonregenerable technology include full-scale demonstra-
tions of double-alkali scrubbing systems, and a comprehensive program
in scrubber sludge disposal and utilization.
-------
274 CLEAN COMBUSTION OF COAL
In the regenerable FGD area, an aggressive program has also
been pursued. Full-scale demonstrations of the Cat-Ox (producing
80% sulfuric acid), Wellman-Lord (W-L)/Allied Chemical (producing
sulfur), magnesium oxide (producing sulfuric acid), aqueous carbonate
(producing sulfur), and the citrate* (producing sulfur) processes
have been undertaken.
A number of supporting studies have also been initiated for
both nonregenerable and regenerable processes. Included are design
and cost evaluations for advanced S02 removal technologies, byproduct
marketing studies, bench-scale research on key processing steps,
investigation of reductants for SO- to sulfur, energy optimization
studies, and a comparison of utility/industry equipment reliabil-
ity/availability.
To complement these efforts to develop technology, companion
technology transfer efforts are also underway. Through a series of
briefings, symposia, capsule reports, summary reports, and a survey
of FGD installations, the industry is being aided in its efforts to
stay abreast of the rapidly advancing FGD technology. The technology
transfer efforts include reporting, to the fullest extent possible,
the status of the many full-scale utility and industrial FGD systems
designed, constructed and operated under private funding.
The purpose of this paper is to summarize the results and
status of the efforts described above as well as the status of FGD
technology in the electric utility industry.
NONREGENERABLE SYSTEMS
Lime/Limestone Scrubbing
Lime/limestone processes are the most highly developed FGD
processes, having the greatest amount of operational experience.
Over 90% of the FGD systems, which have been built, are under con-
struction, or are being planned for service by the early 1980's, are
lime/limestone processes.
The Shawnee Program. An important part of the lime/limes tone
effort involves the operation of a prototype scrubbing test facility
at Shawnee Steam Plant. This versatile facility allows comprehensive
testing of up to three 10 MW scrubber types under a variety of oper-
ating conditions. Bechtel Corporation of San Francisco designed the
test facility and directs the test program, and TVA constructed and
operates the facility.
The major concerns regarding lime/limestone scrubbing have cen-
tered around potential operating difficulties caused by scaling and
*The citrate process is based on pilot plant work by the
U.S. Bureau of Mines
-------
FLUE GAS DESULFURIZATION 275
plugging (especially in the mist eliminator area), the large quanti-
ties of waste sludge generated, and the high costs (capital and
operating) of scrubbing. It is toward these areas of concern that
the Shawnee program has been directed.
The testing at Shawnee since 1972 has made major contributions
toward improvement of lime and limestone scrubbing technology. The
most significant results to date include: (1) demonstration, on a
10 Mw scale, that conventional lime/limestone systems can be operated
reliably (two separate reliability problems have been identified—
scaling and accumulation of soft mud-type solids—and methods to
control each have been demonstrated); (2) evidence that mud-type
solids deposition is a strong function of alkali utilization and at
high utilization (greater than about 85%) these solids are much more
easily removed; (3) demonstration of equipment or process variations
which individually improved alkali utilization (thereby reducing
operating costs), improved S02 removal efficiency, and favorably
influenced the system chemistry; and (4) development of useful
industrial tools, such as the design/economic study computer program
and the computerized Shawnee data base.
The EPA Pilot FGD Scrubber Program. The FGD pilot plant operated
by IERL-RTP consists of two scrubbers having a flue gas capacity of
about 0.1 MW each. They have been in operation since 1972 for the
principal purpose of providing in-house experimental support for
EPA's larger, prototype scrubber test facility at Shawnee Steam
Plant. In addition to supporting Shawnee, the pilot plant also
provides IERL-RTP with the capability to independently evaluate new
concepts in lime/limestone scrubbing technology.
Results from this pilot unit indicate that forced oxidation
limestone scrubbing can be achieved in a two-stage scrubbing system
promoting formation of gypsum, a more desirable form of calcium
solids, and improving limestone utilization. An extension of this
work has recently been undertaken at the Shawnee test facility.
Further tests at IERL-RTP have been directed toward achieving the
oxidation step in a single-stage scrubber, and information has been
developed toward application in larger commercial scrubbing systems.
Thus far, it has been demonstrated in the IERL-RTP scrubber that
single-stage forced oxidation can be achieved with no loss of S0»
removal efficiency.
A study of the formation of solid solutions in lime/limestone
scrubbers was completed which verified earlier findings made at
IERL-RTP that sulfate can be purged with the solids at low oxidation
levels while maintaining subsaturated liquor. Larger-scale studies
of this "subsaturated" operation mode are being made at Shawnee and
Louisville Gas and Electric's (LG&E) Paddy's Run lime scrubber.
Bahco Program. The AB Bahco Ventilation (Sweden) lime scrubbing
process has been installed on about 20 small industrial-size oil-
fired boilers outside of the United States. Research-Cottrell is
the licensee in the United States for this process. The Bahco
-------
276 CLEAN COMBUSTION OF COAL
system appears particularly suited for small industrial applications;
it is manufactured in standard sizes of about 5-50 MW equivalent.
The system is readily adaptable to a high degree of automation.
Although automation results in a somewhat higher capital investment
cost initially, labor costs are low because boiler operating personnel
can also handle operation of the FGD system.
The Air Force contracted with Research-Cottrell to install a
Bahco system for SOo and particulate control on up to seven small
coal-fired heating boilers (approximately 21 MW equivalent total) at
Rickenbacker Air Force Base near Columbus, Ohio. EPA is sponsoring
a 2-year test program on this system. Although numerous mechanical
problems have been encountered since startup in March 1976, results
generally have been promising. Despite the high particulate loading,
which at times has been predominately very small sooty particles,
design specifications for both the particulate and 862 removal
efficiency have been exceeded consistently.
LG&E Scrubber Test Program. In November 1974, results from
the IERL-RTP pilot-plant testing were reported which showed that
lime and limestone SC^ scrubbers could be operated subsaturated with
respect to dissolved CaSO,'2H?0 (gypsum). This mode of operation
avoids the problem of gypsum scaling on the scrubber internals.
Subsequent investigation indicated that at least two commercial
scrubber systems were operating subsaturated with respect to gypsum,
one at the Mitsui Aluminum Plant in Omuta, Japan, and the other at
Paddy's Run Station of LG&E.
Because of EPA's interest in studying the subsaturated mod.e of
operation on a full-scale system, a program was undertaken at LG&E
in the spring of 1976 to evaluate operational and chemical factors
(identified by scrubber testing at IERL-RTP and Shawnee) which
appear to have an effect on subsaturated-operation. The carbide
lime phase (baseline tests) of the test program was initiated in
October 1976 and was concluded in December 1976. No major scrubber
operational problems occurred during these tests. Waste sludge from
the system was collected for sludge treatment/disposal tests which
are being conducted in conjunction with the scrubber test program.
The commercial lime phase of the test program was initiated in
March 1977; shortly after start-up, scaling occurred in the scrubber.
The scaling problem was a result of higher oxidation (than with
carbide lime) and a lack of gypsum crystals, causing "locally"
excessive gypsum saturation levels. The marble bed scrubber used in
the tests is also more prone to difficulties from scaling. It has
been concluded that the carbide lime contains an oxidation inhibitor
in trace quantities. Currently, commercial lime testing has resumed
with the addition of small quantities of magnesium oxide to prevent
scaling. High S02 removals, along with no significant operational
problems, have been reported.
-------
FLUE GAS DESULFURIZATION 277
Double-Alkali Scrubbing
The double-alkali process provides an alternate wet scrubbing
"throwaway" system to the more prevalent lime/limestone slurry
scrubbing processes. Such systems employ a clear liquid absorbent
rather than the slurry used in lime/limestone processes. As a
result, the scrubber in a double-alkali unit is expected to be less
prone to fouling and plugging problems.
The double-alkali process is now offered commercially by several
companies for control of industrial and utility boilers. Process
capabilities include 90% or more SCL removal, less than 2% energy
consumption exclusive of reheat energy, close to 100% lime/SO-
stoichiometry, and soda ash consumption in the range of 5% of the
lime on a molecular basis. However, these processes may in certain
instances be more costly than lime/limestone systems.
Combustion Equipment Associates (CEA)/A.D. Little (ADL) Program.
After initial in-house engineering feasibility studies and laboratory
experiments in 1971 and 1972, EPA contracted with ADL in May 1973 to
conduct a laboratory and pilot-plant study of various double-alkali
modes of operation. These efforts included the study of "dilute"
and "concentrated" systems, lime and limestone regeneration, sulfuric
acid addition for sulfate removal, and solids characterization.
In early 1975 the project was expanded to include a prototype
test at the 20 MW facility installed at the Scholz Steam Plant of
Gulf Power Company by The Southern Company and constructed by CEA.
Testing at the Scholz Steam Plant lasted from February 1975 to July
1976, with the EPA-sponsored portion of the testing beginning in May
1975. As a whole, the prototype system performed very well and
indicated that a double-alkali system could be a viable FGD system
for coal-burning utilities. SO^ removal was generally in the range
of 90 to 99%.
General Motors (GM) Industrial Boilers Demonstration. GM and
EPA have participated in a cooperative program to demonstrate, test,
characterize, and evaluate GM's "dilute" mode double-alkali system
for control of SO- emissions from coal-fired industrial boilers.
The program was conducted at GM's industrial boiler complex in
Parma, Ohio. The system, consisting of four coal-fired boilers
having a steaming capacity of 320,000 Ib/hr (equivalent to 32 MW
electric generating capacity), was constructed and operated by GM.
ADL designed and conducted the test program to evaluate the system
with funding from EPA. The test program consisted of three 1-month
intensive test periods and longer term nonintensive testing. Each
of the intensive tests evaluated a slightly different flow scheme.
LG&E Double-Alkali Demonstration Program. In September 1976,
EPA contracted with LG&E for a cost-shared, full-scale coal-fired
utility demonstration of the double-alkali process at the 280 MW Cane
Run No. 6 boiler. The demonstration project consists of four phases:
-------
278
CLEAN COMBUSTION OF COAL
(1) design and cost estimation; (2) engineering design, construction,
and mechanical testing; (3) startup and performance testing; and (4)
1 year of operation and long-term testing. Construction is expected
to be complete by the end of 1978, and testing will begin in early
1979. A contract with Bechtel Corporation has been initiated to
design and conduct a test program for the LG&E facility and evaluate
the process technically and economically.
FGD Sludge Disposal. One of the major considerations involved
in the selection, design, construction and operation of nonregenerable
FGD systems is the disposition of large quantities of sludge. Since
1972, efforts have been underway in the EPA FGD research and develop-
ment program to evaluate and demonstrate sludge disposal techniques.
Currently twelve projects are being conducted in this area; these
are listed in Table 1. These projects address the two major environ-
mental concerns associated with FGD sludge disposal: (1) the water
pollution potential of soluble materials, and (2) the land degrada-
tion potential of physically unstable wastes.
Table 1: EPA Projects in FGD Waste Disposal
Project Title
FGD Waste Characterization,
Disposal Evaluation, and Transfer
of FGD Waste Disposal Technology
Lab and Field Evaluation of 1st
and 2nd Generation FGD Waste
Treatment Processes
Ash Characterization and Disposal
Studies of Attenuation of FGD
Waste Leachate by Soils
Establishment of Data Base for
FGC Waste Disposal Standards
Development
Shawnee FGD Waste Disposal Field
Evaluation
Louisville Gas and Electric
Evaluation of FGD Waste Disposal
Options
FGD Waste Leachate/Liner Compat-
ibility Studies
Lime/Limestone Wet Scrubbing
Waste Characterization
Dewatering Principles and
Equipment Design Studies
Conceptual Design/Cost Study
of Alternative Methods for Lime/
Limestone Scrubbing Waste Disposal
Evaluation of Alternative FGD
Waste Disposal Sites
Contractor/Agency
The Aerospace Corporation
U.S. Army Corps of Engineers
Waterways Experiment Station
Tennessee Valley Authority
U.S. Army Materiel Command
Dugway Proving Ground
Stearns, Conrad and Schmidt
Consulting Engineers, Inc.
(SCS Engineers)
Tennessee Valley Authority
The Aerospace Corporation
Louisville Gas & Electric
Company (Subcontractor:
Combustion Engineering)
U.S. Army Corps of Engineers
Waterways Experiment Station
Tennessee Valley Authority
Auburn University
Tennessee Valley Authority
Arthur D. Little, Inc.
-------
FLUE GAS DESULFURIZATION 279
Results in this program area have been substantial. FGD sludge
chemical characteristics, to a large degree, have been quantified.
Sludge liquors exceed drinking water standards for total dissolved
solids (IDS), with high concentrations of calcium, sulfate, and
chloride (and in some cases, magnesium and sodium). In addition,
excessive concentrations of several trace metals have been noted.
The chemical composition of FGD sludge solids consists of calcium
sulfite hemihydrate, calcium sulfate dihydrate (gypsum) and/or
hemihydrate, and calcium carbonate, plus any fly ash collected in
the scrubber. The percentage of each solid constituent is primarily
a function of the alkaline additive (e.g., lime, limestone), the
percent sulfur in the coal, and the manner in which the scrubber
system is operated (e.g., whether forced oxidation is applied,
whether fly ash is collected separately). Although the fly ash has
been shown to be a major contributor of trace elements to the sludge
solids and liquor, separate collection of fly ash does not necessarily
mean that concentrations of all these elements in the sludge liquor
will be insignificant. In summary, chemical characterization of FGD
sludge has shown the need for protection of drinking water supplies
from intrusion by sludge leachates.
The physical properties of FGD sludge vary considerably from
system to system; chemical composition is related to, but does not
adequately define, the sizes and types of the sludge solid crystals.
Many FGD sludges tend to liquefy easily, even after substantial
dewatering. Several approaches to improving physical stability con-
tinue to be studied, including stabilization using underdrainage and
compaction, production of gypsum, and chemical treatment ("fixation")
for landfill. Chemical treatment of FGD sludge has been shown to
result in significant structural improvement, a 50-75% reduction in
major solubles (e.g., chloride) in the leachate and an order of
magnitude (or more) reduction in permeability. Further testing of
these disposal methods, including revegetation (reclamation) of
disposal sites, is planned.
The costs of FGD sludge disposal vary considerably, depending
on the disposal system design, and site-specific factors such as
labor costs or the cost of a pond liner (if one is installed). Pre-
liminary cost estimates for a typical high-sulfur-coal-burning plant
are about $4-$9 per metric ton (dry basis, including ash) for
ponding, and about $8-$12.50 per metric ton (same basis) for chemical
treatment and landfill. The ponding costs do not include reclamation
costs. More detailed economics for these disposal methods have been
defined by TVA, and will be reported soon. The next phase of the
EPA-sponsored study at TVA will include gypsum disposal.
Costs of FGD sludge disposal represent a major part (up to 20%)
of the capital and operating costs of an FGD system. These costs
can be drastically reduced by improved absorbent (e.g., limestone)
utilization, controlled solids quality, and by improved sludge
dewatering equipment. One approach to control solids quality which
is currently under study is an attempt to develop a procedure to
obtain consistent, easily dewatered sulfite solids. An alternative
-------
280 CLEM COMBUSTION OF COAL
approach would be to use forced oxidation to produce only gypsum
crystals, which are normally much larger than calcium sulfite crystals,
A complementary approach is to improve the performance of dewatering
equipment. Separation of the clarification and thickening steps can
result in improved performance of gravity settlers, with a substantial
reduction in the equipment size. With the exception of controlling
sulfite solids quality, all of these improvements have been shown to
be feasible and are currently making their way into the process
supply market. However, further development/refinement of these
techniques is continuing; their full commercial use is expected in
the next 2-3 years.
Coal-mine disposal of FGD sludge has greatly interested engineers
in the flue gas desulfurization industry for many years, because of
established means of transportation between the coal mine and the
power plant, and the need for material to fill the void left by
mining of the coal. In addition, many plants may not have sufficient
land area for on-site disposal. Recent technical/economic assessments
indicate that active Midwestern surface mines and Eastern/Midwestern
room-and-pillar underground mines are the most promising candidates
for this disposal approach. One utility plans to begin disposal of
FGD sludge and ash in a surface mine this summer. Plans under the
EPA program are to conduct a 2-year monitoring/assessment effort at
this utility site. Successful demonstration of this disposal approach
could make conversion to coal quite feasible even in areas where
land for disposal is limited.
Ocean disposal of FGD sludge is also being assessed because
many plants in the Northeast may have difficulty switching to coal
for lack of disposal sites; however, many of these plants do have
access to the ocean. It was also recognized that the major soluble
chemical constituents in FGD sludge are found in relatively high
concentrations in seawater. Studies of ocean disposal of FGD sludge
by Arthur D. Little for EPA have identified several potential environ-
mental problems. It appears that these problems could be alleviated
by either chemical treatment to a "brick-like" form (possibly creating
an artificial reef) or oxidation to gypsum (followed by a widely
dispersed disposal operation). The costs of these approaches are
being defined and are expected to be somewhat higher than for chemical
treatment/landfill near the plant. Pilot disposal simulation studies
are underway to define the environmental effects of both untreated
and treated FGD sludge disposal in the ocean.
Currently there are no Federal regulations which specifically
address the disposal of FGD. sludge. However, the Resource Conserva-
tion and Recovery Act (RCRA) of 1976 calls for the eventual Federal
regulation of disposal of hazardous solid wastes and the issuance of
guidelines (to be used by the states) for disposal of non-hazardous
solid wastes. The RCRA specifically identifies solid wastes and
sludges, including those generated by air pollution control devices,
as being covered by the Act. Although no official designation'
(hazardous or non-hazardous) has been placed on FGD sludges, it is
currently assumed that they will be considered non-hazardous, with
-------
FLUE GAS DESULFURIZATTON
281
disposal guidelines to be issued in the next 2-3 years. An effort
has been underway since mid-1975 to prepare a preliminary technical
support document which could be potentially useful in setting FGD
waste disposal guidelines. A draft of the document is currently
under review.
Studies of the characteristics of coal ash and the effects of
coal ash disposal have been underway and are continuing. These
efforts are not as extensive as those for FGD sludge. However, they
are no less significant because of the increasing generation of coal
ash, and because many FGD sludges contain significant quantities of
fly ash, either collected in the scrubber or added to the FGD sludge
prior to disposal (e.g., in a chemical treatment process). A report
has been issued which summarizes and evaluates existing data on the
characteristics of coal ash from studies made by TVA and others.
FGD Sludge Utilization. This part of the FGD program has only
become active in the past 2 years; it currently consists of five
projects, listed in Table 2.
Table 2: EPA Projects in FGD Waste Utilization
Project Title Contractor/Agency
Gypsum By-product Marketing Studies
Lime/Limestone Scrubbing Waste
Conversion Pilot Studies
Fertilizer Production Using
Lime/Limestone Scrubbing Wastes
Use of FGD Gypsum in Portland
Cement Manufacture
FGD Waste/Fly Ash Beneficiation
Studies
Tennesssee Valley Authority
Pullman-Kellogg
Tennessee Valley Authority
Babcock & Wilcox
Portland Cement Association
C.E. Lovewell
TRW, Inc.
Since FGD sludge is a relatively new by-product, utilization in
the United States is not yet a commercial reality. However, conversion
of FGD sludge to gypsum (or direct production of gypsum) for use in
wallboard and portland cement manufacture is practiced extensively
in Japan. Although the Japanese experience has primarily been with
oil, gypsum-producing FGD experience with coal is increasing in
Japan and the United States. However, thus far there have been no
full-scale commitments by American utilities to produce FGD gypsum
for sale. This situation may change as the utilization of coal for
electric power expands and an energy and resource conservation ethic
begins to take shape.
Tools for the development of FGD gypsum market strategies have
been developed. Studies currently underway at TVA include a thorough
economic evaluation of several gypsum-producing FGD processes—e.g.,
-------
282 CLEM COMBUSTION OF COAL
limestone/gypsum, Chiyoda (H-SO./gypsum), and Dowa (aluminum-based
double alkali/gypsum)—and a detailed U.S. marketing study of FGD
gypsum for wallboard. A report on this study is expected later in
1977; indications are that processes less complex than those used in
Japan will be necessary for a profitable situation to occur. Wallboard
production using FGD gypsum from a Southeastern utility has been
successfully demonstrated. Feasibility demonstration of FGD gypsum
used in portland cement in cooperation with trade associations is
planned.
Development of FGD sludge utilization in fertilizer is continuing
at the pilot level at TVA. Spreading the material over a relatively
large land area in this manner would not only alleviate the disposal
problem, but would also minimize the potential localized environmental
impact of a highly concentrated waste; i.e., the leachate's chemical
constituents would be highly diluted by rainfall and interaction
with the soil. Further development of the fertilizer production
process is needed to establish its viability.
Conversion of FGD sludge to elemental sulfur with recovery of
calcium carbonate (using coal as the reductant) for recycle to the
scrubber has been studied on a pilot level by Pullman-Kellogg and
Ontario Hydro. Further studies are planned under a contract currently
being negotiated with Pullman-Kellogg. This effort should be under-
way later in 1977.
REGENERABLE SYSTEMS
Since its inception in 1970 EPA has assisted in the development
of several recovery processes capable of producing sulfuric acid,
elemental sulfur, or liquefied SO-. These processes have been
pursued in hopes of conserving a valuable natural resource and
reducing overall S02 control costs. Most of the EPA efforts have
been directed towara full-scale demonstrations of a number of leading
processes; however, support has also been given to bench-scale and
pilot plant efforts.
Magnesium Oxide Scrubbing Program. The Mag-Ox scrubbing
process—developed by Chemical Construction Company (Chemico) and
Basic Chemicals, and currently offered commercially by Chemico—is
one of the more promising regenerable FGD approaches. The process,
which produces sulfuric acid, is widely applicable to both existing
and new power plants. It is also amenable to the centralized pro-
cessing concept; i.e., spent sorbent can be regenerated at a central
plant capable of servicing a number of power or industrial plants.
In 1974, EPA and Boston Edison completed a co-funded demonstration
program of a 155 MW capacity scrubbing/regeneration system. Results
obtained during 2 years of operation indicated: (1) SO- removal
efficiencies in excess of 90% were obtained consistently, and (2) more
than 5,000 tons of saleable sulfuric acid of high quality was recovered
from the stack gas and sold commercially. A number of problems were
encountered that were primarily equipment, rather than process,
related; however, continuous, long-term, reliable operation was not
achieved.
-------
FLUE GAS DESULFUBIZATION 283
In 1973, Potomac Electric Power Company installed a 100 MW Mag-
Ox scrubbing system at its coal-fired Dickerson Station. EPA pro-
vided the Mag-Ox regeneration system for Potomac Electric's use in
processing spent scrubber sorbent. Results indicate S0~ removal
efficiencies greater than 90% are possible and particulate removal
of 99.6% was attained. Over 2000 tons of sulfuric acid was produced
and marketed. Unfortunately due to a shortage of funds this demon-
stration did not run long enough to completely answer all questions
regarding absorbent recycle, absorbent losses, and process reliability.
Philadelphia Electric will soon begin operating a 120 MW MgO
scrubbing facility at the Eddystone Unit 1. After initial startup
in September 1975, this unit was shut down when the regeneration-
sulfuric acid system at Olin Chemicals, Paulsboro, New Jersey, plant
was permanently closed. Regeneration will now take place at Essex
Chemical's acid plant in Newark, New Jersey. EPA plans to supply
consulting for startup, operation, and test program formulation.
Wellman-Lord/Allied Chemical Demonstration Program. EPA and
Northern Indiana Public Service Company (NIPSCO) have jointly funded
the design and construction of a flue gas cleaning demonstration
plant utilizing the Wellman-Lord S0~ recovery process and the Allied
Chemical SO reduction process to convert recovered S0_ to elemental
sulfur. The operational costs for the system will be paid by NIPSCO,
and a comprehensive test and evaluation program will be funded by
EPA. The demonstration system has been retrofitted to the 115 MW,
coal-fired Unit 11 at the D.H. Mitchell Station in Gary, Indiana.
Construction of the facility was completed in August 1976.
Startup activities and acceptance testing were delayed by
boiler problems which developed when Unit 11 was shut down for
annual maintenance. The boiler and the FGD plant were restarted
in June, and integrated operation and acceptance testing will follow
shortly thereafter. Long-term duration testing will begin after
acceptance.
Bureau of Mines (BOM) Citrate Demonstration Program. EPA and
BOM have entered into a cooperative agreement to pool funds and
technical talents to demonstrate the citrate process developed by
BOM. A concurrent development program, carried out by an industrial
consortium headed by Pfizer Chemical Company, also led to a pilot
operation of the process. Based on the results of these two pilot
programs, EPA and BOM have.initiated the demonstration of this tech-
nology on a 53 MW coal-fired boiler at St. Joe Minerals Corporation
in Monaca, Pennsylvania. Construction of the facility is expected
to be completed by mid-1978. After acceptance, a 1-year test and
evaluation program will be initiated.
Aqueous Carbonate Demonstration Program. EPA and Empire State
Electric Energy Research Corporation (ESEERCO), a research organiza-
tion sponsored by New York's eight major power suppliers, have
recently contracted to fund jointly the design and construction of a
demonstration of Atomics International's sulfur-producing aqueous
carbonate process. The demonstration system is being retrofitted to
-------
284 CLEM COMBUSTION OF COAL
Niagara Mohawk Power Company's 100 MW coal-fired Huntley Station in
Tonawanda, New York. Construction of the facility is expected to be
completed by mid-1979. After acceptance testing, a 1-year test and
e*\ra 1 imlH cm m-ncri-am is nlarmpd.
evaluation program is planned.
Catalytic Oxidation (Cat-Ox) Demonstration Program. The Cat-Ox
process is Monsanto Enviro-Chem Systems' adaptation of the contact
sulfuric acid process. EPA and Illinois Power Company attempted
to demonstrate the process on a 103 MW coal-fired boiler at Illinois
Power's Wood River Station between 1970 and 1976.
Detailed design, construction, shakedown testing, and performance
guarantee testing were completed in July 1973. The unit met all
guarantees and was subsequently accepted. Because of the shortage
of natural gas, the burners were modified to allow either oil- or
gas-firing. However, design and startup problems precluded suc-
cessful operation and initiation of the comprehensive 1-year test
program.
In view of the problems and long delays encountered, a thorough
technical and economic study was made of the costs and benefits of
continuing the Cat-Ox demonstration at the Wood River Station.
Results of this study led to the decision to end the project in
December 1976.
Ammonia Scrubbing with Bisulfate Regeneration Pilot-Plant Program.
In 1970, EPA and TVA jointly undertook the development of a completely
cyclic ammonia scrubbing/ammonium bisulfate regeneration process
which has as its major product a concentrated stream of SO- which
can then be used to produce sulfuric acid or elemental sulfur. This
process was evaluated at a 3,000 ft-Vmin (5,000 m^/hr) pilot unit
located at Colbert Steam Plant. While initial developmental efforts
at the pilot unit were concentrated on the absorber, later work
included investigation of all subunits of the system except the
electrical decomposer. It became apparent that the process had two
major problems: (1) the formation of a persistent fume which could
not be consistently controlled or eliminated by reasonable control
efforts, and (2) unfavorable economic projections due primarily to
energy consumption by the decomposer. As a result of these problems,
the development project was terminated during the summer of 1976.
FGD SUPPORT STUDIES
Key supporting studies in several problem areas of FGD technology
have been sponsored by EPA to further the advancement and application
of commercial systems. In many cases, the studies undertaken are
broad general assessments which are directed toward a wide variety
of potential users. Examples of results from these studies are
described below.
Comparative Economics of SO,, Control Processes
The purpose of this continuing EPA/TVA project is to study the
most promising S02 removal processes advancing toward commercialization.
-------
FLUE GAS DESULFURIZATION 285
It includes selection of those processes which have the greatest
degree of development and which are potentially attractive both
technically and economically. These evaluations include preparation
of flowsheets, material balances, and layouts; definition of process
equipment; preparation of capital investments and operating costs;
and analysis of design and economic variables for cost sensitivity
analysis. Currently, this is being done for the citrate and double-
alkali processes using the limestone system for comparison; results
should be available by the fall of 1977.
By-product Marketing Studies
TVA is conducting a program for EPA (1) to determine the quantities
of FGD by-product acid sulfur and gypsum that would be produced at
power plant sources, and (2) to analyze the markets for these by-
products. The computer model which has been developed considers
compliance of actual utility plants with SO- emission standards and
potential by-product market demand for each plant (based on current
markets and freight rates). Reports on the marketing of sulfur,
sulfuric acid, and gypsum (for wallboard) are expected in the fall
of 1977.
Reductant Gases
Conversion of SO to elemental sulfur requires the use of a
reductant; in the pas? major emphasis was placed on the use of
natural gas. Since this is now an impractical approach, other
sources of reductant gas must be sought, such as coal gasification.
A study conducted for EPA by Battelle-Columbus concludes that a
gasifier-based reduction system would significantly increase the
complexity of the overall FGD system. However, the gasifier could
be used to supply part of the energy for the FGD process, including
stack gas reheating.
Magnesium Oxide Scrubbing Support Studies
Two studies in support of magnesium oxide scrubbing have been
conducted for EPA by Radian Corporation. In the first study, Radian
evaluated the feasibility of producing elemental sulfur directly
from magnesium sulfite. This would expand the applicability of
current magnesium oxide processes which only produce sulfuric acid.
The second study is concerned with the mechanism of formation of
trihydrate and hexahydrate forms of magnesium sulfite (MgSO^-SH-O,
MgSO -6H-0). The hexahydrate crystals separate and handle easily;
the trihydrate crystals require less drying energy but are more
difficult to separate and handle. This study has attempted to
generate information on formation mechanisms and operating conditions
that can be used to control the type of crystal formed.
Comparison of Availability and Reliability of Equipment Utilized in
the Electric Utility Industry
For the past several years one of the major objections of the
utility industry to installing FGD systems has been that reliabili-
-------
286 CLEAN COMBUSTION OF COAL
ty/availability of FGD systems is much lower than for other major
utility equipment items, such as boilers, turbines, generators,
electrostatic precipitators, and gas turbines. It is desirable that
good information and data on this problem be gathered so that a
valid comparison of performance can be made.
To meet this need and to provide information as input for a
current National Academy of Sciences study of SO^ control technology,
a study by EPA was initiated (with Radian Corporation) as a jointly
sponsored project of EPA and the Council on Environmental Quality.
The study concluded that a statistically meaningful comparison of
reliability/availability of these components cannot now be made,
primarily because of the small number and short service time of FGD
system data (a meaningful comparison can probably be made in 1979) .
However, the study also concluded that the mechanical reliability of
some types of conventional equipment now being used by the electric
utility industry is not much different from that of similar items
used in FGD systems.
TECHNOLOGY TRANSFER
For several years, EPA has disseminated FGD technology data and
information through the traditional outlets of FGD symposia, industry
briefings, capsule reports, and project summary reports. Six symposia
have now been held, the last one in March 1976. The next symposium
is scheduled for November 8-11, 1977, in Hollywood, Florida. Progress
on lime/limestone technology has been reported through industry
briefings, the last one in October 1976. Thus far, three capsule
reports have been issued on the EPA/TVA/Bechtel Shawnee program.
Numerous project reports have also been issued.
In order to improve the effectiveness of the FGD technology
transfer effort, the survey of utility FGD installations and, more
recently, the Engineering Applications/Information Transfer (EA/IT)
program were initiated. These are described below.
Survey of Utility FGD Installations
To meet the continuing demand for technical and economic data
on operational, under construction, and planned future utility FGD
units, EPA has employed PEDCo-Environmental to monitor this field of
technology and prepare periodic reports for use by utilities, system
vendors and designers, and regulatory authorities. In addition to
detailed technical reportst PEDCo is providing bimonthly status
reports indicating the number of each type of S02 control system in
operation, under construction, or planned in the United States, and
the megawatt capacity controlled or to be controlled. The bimonthly
status report gives technical and economic information on all known
U.S. utility FGD systems categorized in 15 tables and 4 appendices
to promote ease of use. Some of the information from the latest
PEDCo report April-May 1977) is summarized in the final section of
this paper.
-------
FLUE GAS DESULFURIZATION 287
Engineering Applications/Information Transfer (EA/IT)
The EA/IT program is a comprehensive effort designed to augment
traditional EPA technology transfer efforts so that dissemination of
information will be more efficient and effective. The program
includes several new and innovative reporting activities, including:
- A series of quarterly reports on FGD research, development and
demonstration efforts sponsored by EPA.
- A series of Cost/Reliability Handbooks to assist potential
users in choosing the specific SOX control strategy, FGD system,
and FGD system design which best fit their needs.
- Lime and Limestone Scrubbing Data Books (cooperative EPA/Electric
Power Research Institute efforts).
- Non-utility (industrial) combustion source survey report assess-
ing applicability of various SO control strategies.
X
FULL-SCALE UTILITY FGD APPLICATIONS
Overview
According to the latest PEDCo survey, 119 utility boilers
representing over 50,000 MW of electric generating capacity will be
controlled by FGD by the mid-1980's. The current status of these
systems is given in Table 3.
Table 3: Status of Utility FGD Systems
No. of
Status Units MW
Operational 27 7319
Under Construction .29 12648
Subtotal 56 19,967
Planning
Contract Awarded 20 9797
Letter of Intent 5 1892
Requesting/Evaluating Bids 5 3565
Considering only FGD Systems ^3_ 14856
Total 119 50,077
Of the units which are operational or under construction, 91%
of the generating capacity or roughly 18,000 MW will be controlled
by lime/limestone scrubbing systems. Of the units for which con-
tracts have been awarded, 88% of the generating capacity or roughly
8000 MW will also be controlled by lime/limestone systems. However,
of those units in the remaining stages of planning shown above
(e.g., letter of intent) for which a process type has been selected,
the trend appears to be away from lime/limestone, with only 55% of the
-------
288 CLEAN COMBUSTION OF COAL
generating capacity being planned for control by these systems.
However, of the 20,467 MW in these remaining stages, over 13,500 MW,
or 66%, represent units for which the specific FGD system has not
yet been selected. Therefore, it is perhaps too soon to tell if
there will be a substantial move toward the regenerable FGD processes.
Nevertheless, as operating experience with regenerable systems
increases, particularly with the Wellman-Lord and magnesium oxide
systems, they will no doubt occupy a significant portion of the FGD
market in the 1980's.
Table 4 summarizes the FGD systems which were operational as of
May 1977.
CONCLUSIONS
The experience level of FGD technology is rapidly developing.
The development of lime/limestone scrubbing is far ahead of other
FGD systems primarily because of information gathered from operational
systems and through development sponsored by EPA. If future power
plant scrubber systems are designed to take advantage of data already
available, improvement in overall reliability, cost, and effective-
ness should transpire. Additional work is needed in this area of
technology, however, to improve performance, reduce costs, and
increase reliability.
Double-alkali technology is rapidly developing as a viable
means of SO™ control. The potential advantages of such systems to
improve SO- removal, sludge characteristics, and operating reliability
first need to be confirmed through the full-scale demonstrations,
then applied as an alternate to lime/limestone technology.
The area in which the most work remains to be done is in regen-
erable process technology. The presently planned demonstrations
need to be completed as soon as possible. In addition, further work
is needed to confirm the long-term effectiveness and economics of
magnesium oxide scrubbing.
-------
Table 4: Summary of Operational Utility FGD Systems*
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.
Company
Arizona Public Service
Columbus and Southern
Ohio Electric
Commonwealth Edison
Duquesne Light
Duquesne Light
Gulf Power Co.
Kansas City Power & Light
Kansas City Power 4 Light
Kansas City Power & Light
Kansas Power & Light
Kentucky Utilities
Louisville Gas & Electric
Louisville Gas & Electric
Montana Power Co.
•
Montana Power Co.
Nevada Power
Nevada Power
Nevada Power
Northern States Power Co.
Pennsylvania Power Co.
Philadelphia Electric Co.
Tennessee Valley Authority
Tennessee Valley Authority
Northern Indiana Public
Service
Northern States Power
Springfield City Utilities
Tennessee Valley Authority
Unit
Coal Percent
Unit Name Capacity, Hw FGD Process Type
Cholla No. 1
Conesville No. 5
Will County No. 1
Elrama Power Station
Phillips Power Station
Scholz Nos. IB & 2B
Hawthorn No. 3
Hawthorn No. 4
LaCygne No. 1
Lawrence No . 4
Green River Nos. 1 & 2
Cane Run No. 4
Paddy's Run No. 6
Colstrip No. 1
Colstrip No. 2
Reid Gardner No. 1
Reid Gardner No. 2
Reid Gardner No. 3
Sherburne County No. 1
Bruce Mansfield No. 1
Eddys tone No. 1A
Shaunee No. 10A
Shawnee No. 10B
D. H. Mitchell No. 11
Sherburne County No. 2
Southwest No. 1
Widows Creek No. 8
115
400
167
510
410
23
140
100
820
125
64
178
65
360
360
125
125
125
710
835
120
10
10
115
680
200
550
Limestone Scrubbing
Lime Scrubbing
Limestone Scrubbing
Lime Scrubbing
Lime Scrubbing
Chiyoda 101
Lime Scrubbing
Lime Scrubbing
Limestone Scrubbing
Limestone Scrubbing
Lime Scrubbing
Lime Scrubbing**
Lime Scrubbing**
Lime/Alkaline Ash
Scrubbing
Lime/ Alkaline Ash
Scrubbing
Sodium Carbonate
Scrubbing
Sodium Carbonate
Scrubbing
Sodium Carbonate
Scrubbing
Limestone Scrubbing
Lime Scrubbing
Magnesium Oxide
Scrubbing
Lime/Limestone
Scrubbing
Lime/ Limes tone
Scrubbing
Wellraan-Lord
Limestone
Limestone
Limestone
Sulfur
0.
4.
4.
1.
1.
5.
0.
0.
5.
0.
3.
3.
3.
0.
0.
0.
0.
0.
0.
4.
2.
2.
2.
3.
0.
3.
3.
44-1.0
5-4.9
0
0 -2.8
0 -2.8
0 (max)
5 -3.5
5 -3.5
0
5
8
5 -4.0
5 -4.0
8
8
5-1.0
5 -1.0
5 -1.0
8
7
5
9
9
5
8
5
7
Waste Disposal
Start-Up Date Method
10/73
2/77
2/72
10/75
7/73
3/75
11/72
8/72
2/73
12/68
9/75
8/76
4/73
10/75
7/76
4/74
4/74
7/76
3/76
4/76
9/75
4/72
4/72
6/77
4/77
4/77
4/77
(Evaporation) Ponding
Chemical Treatment/
Landfill
Chemical Treatment/
Landfill
Chemical Treatment/
Landfill
Chemical Treatment/
Landfill
(Lined) Ponding
(Gypsum)
Ponding
Ponding
Ponding
Ponding
Stabilization/Ponding
Ponding
Stabilization/Landfill
Ponding
Ponding
(Evaporation) Ponding
(Evaporation) Ponding
(Evaporation) Ponding
(Lined) Ponding
Chemical Treatment/
Ponding
Not Applicable —
By-product Acid Produced
Ponding
Ponding
Hot Applicable -
By-product sulfur produced
(Lined) Ponding
Stabilization/Ponding
Ponding
^
f
s
p
>
rr\
\JJ
a
M
Ul
t-i
^
S
H
G
y>
f_g
H
0
&
*Source: PEDCo Summary Report, April-May 1977, Contract 68-02-1321, Task No. 28
**Carbide Lime
NJ
00
VO
-------
290 CLEAR COMBUSTION OF COAL
BIBLIOGRAPHY
1. Ando, J. "Status of Flue Gas Besulfurization and Simultaneous
Removal of SO,, and NO in Japan." In Proceedings; Symposium
on Flue Gas, Desulfurization, New Orleans, March 1976, Vol. I,
EPA-600/2-76-136a (NTIS PB 255 317), May 1976, pp. 53-78.
2. Borgwardt, R. H. "EPA/RTP Pilot Studies Related to Unsaturated
Operation of Lime and Limestone Scrubbers." In Proceedings;
Symposium on Flue Gas Desulfurization, Atlanta, November 1974,
Vol. I. EPA-650/2-74-126a (NTIS PB 242-572), December 1974.
3. Borgwardt, R. H. "IERL-RTP Scrubber Studies Related to Forced
Oxidation." In Proceedings: Symposium on Flue Gas Desulfurization,
New Orleans. March 1976, Vol. I. EPA-600/2~76-136a (NTIS PB
255-317), May 1976, pp. 117-143.
4. Borgwardt, R. H. "Improving Limestone Utilization in FGD
Scrubbers." AIChE Symposium Series, "Air-1976," in press.
5. Borgwardt, R. H. Sludge Oxidation in Limestone FGD Scrubbers.
EPA-600/7-77-061 (NTIS PB 268-525), June 1977.
6. Bucy, J. I., J. L. Nevins, P. A. Corrigan, and A. G. Melicks.
"Potential Utilization of Controlled SO Emissions from Power
Plants in Eastern United States." In Proceedings: Symposium
on Flue Gas Desulfurization, New Orleans, March 1976, Vol. II.
EPA-600/2-76-136b (NTIS PB 262-722), May 1976, pp. 647-700.
7. Crowe, J. L., and H. W. Elder. "Status and Plans for Waste
Disposal from Utility Applications of Flue Gas Desulfurization
Systems." In Proceedings; Symposium on Flue Gas Desulfurization,
New Orleans, March 1976, Vol. II. EPA-600/2-76-136b (NTIS PB
262-722), May 1976", pp. 565-577.
8. Devitt, T. W., G. A. Isaacs, and B. A. Laseke. "Status of Flue
Gas Desulfurization Systems in £he United States." In Proceedings;
Symposium on Flue Gas Desulfurization, New Orleans, March 1976,
Vol. I. EPA-600/2-76-136a (NTIS P-B 255-317), May 1976, pp. 13-
51.
9. Epstein, M. EPA Alkali Scrubbing Test Facility: Advanced
Program - First Progress Report. EPA-600/2-75-050 (NTIS PB
245-279), September 1975.
10. Epstein, M. EPA Alkali Scrubbing Test Facility: Summary of
Testing Through October 1974. EPA-650/2-75-047 (NTIS PB
244-901), June 1975.
11. Epstein, M., H. N. Head, S. C. Wang, and D. A. Burbank.
"Results of Mist Elimination and Alkali Utilization Testing at
the EPA Alkali Scrubbing Test Facility." In Proceedings:
Symposium on Flue Gas Desulfurization, New Orleans, March 1976,
Vol. I. EPA-600/2-76-136a (NTIS PB 255-317), May 1976 pp
145-204.
-------
FLUE GAS DESULFURIZATION 291
12. Lowell, P. S., W. E. Corbett, G. D. Brown, and K. A. Wilde.
Feasibility of Producing Elemental Sulfur from Magnesium
Sulfite. EPA-600/7-76-030 (NTIS PB 262-857), October 1976.
13. Head, H. N. EPA Alkali Scrubbing Test Facility: Advanced
Program - Second Progress Report. EPA-600/7-76-008 (NTIS PB
258-783), September 1976.
14. Hissong, D. W., K. S. Murthy, and A. W. Lemmon, Jr. Reductant
Gases for Flue Gas Desulfurization Systems. EPA-600/2-76-130
(PB 254-168), May 1976.
15. Hollinden, G. A., R. F. Robards, N. D. Moore, T. M. Kelso, and
R. M. Cole. TVA's 1-MW Pilot Plant: Final Report on High
Velocity Scrubbing and Vertical Duct Mist Elimination. EPA-
600/7-77-019 (TVA PRS-19), March 1977.
16. Interess, E. Evaluation of the General Motors' Double Alkali
S0? Control System. EPA-600/7-77-005 (NTIS PB 263-469), January
1977.
17. Jones, B. F., P. S. Lowell, and F. B. Messerole. Experimental
and Theoretical Studies of Solid Solution Formation in Lime
and Limestone SO.-, Scrubbers, Vol. I—Final report, and Vol. II—
Appendices. EPA-600/2-76-273a and -273b (NTIS PB 264-953 and
264-954), October 1976.
18. Jones, J. W., "Disposal of Flue Gas Cleaning Wastes," CHEMICAL
ENGINEERING, Vol. 84, No. 4, pp. 79-85, February 14, 1977.
19. Jones, J. W., Brna, T. G., Crowe, J. L., Flora, H. B. and Ray, S. S.
"Environmental Management of Effluents and Solid Wastes from Steam
Electric Generating Plants." In Proceedings; Second National
Conference on the Interagency Energy/Environment R&D Program,
June 1977, in press.
20. Kaplan, N. "Introduction to Double Alkali Flue Gas Desulfuriza-
tion Technology." In Proceedings: Symposium on Flue Gas
Desulfurization, New Orleans. March 1976, Vol. I. EPA-600/2-
76-136a (NTIS PB 255-317), May 1976, pp. 387-422.
21. Koehler, G., and J. A. Burns. Magnesia Scrubbing Process as
Applied to an Oil-Fired Power Plant. EPA-600/2-75-057 (NTIS
PB 247-201), October 1975.
22. Koehler, G. Magnesia Scrubbing Applied to a Coal-Fired Power Plant.
EPA-600/7-77-018 (NTIS PB 266-228), March 1977.
23. LaMantia, C. R., R. R. Lunt, R. E. Rush, T. M. Frank, and N.
Kaplan. "Operating Experience—CEA/ADL Dual Alkali Prototype
System at Gulf Power/Southern Services, Inc. In Proceedings:
Symposium on Flue Gas Desulfurization, New Orleans, March 1976,
Vol. I. EPA-600/2-76-136a (NTIS PB 255-317), May 1976, pp. 423-
471.
-------
292 CLEAN COMBUSTION OF COAL
24. Leo, P. P. and J. Rossoff, Control of Waste and Water Pollution
from Power Plant Flue Gas Cleaning Systems: Second Annual R
and D Report. (To be published for EPA).
25. Lowell, P. S., F. B. Messerole, T. B. Parsons. Precipitation
Chemistry of Magnesium Sulfite Hydrate in Magnesium Oxide
Scrubbing. EPA report in press.
26. Lunt, R. R., C. B. Cooper, S. L. Johnson, J. E. Oberholtzer,
G. R. Schimke, and W. I. Watson. An Evaluation of the Disposal
of Flue Gas Desulfurization Wastes in Mines and the Ocean:
Initial Assessment, EPA-600/7-77-051, May 1977.
27. McGlamery, G. G., H. L. Faucett, R. L. Torstrick, and L. J.
Henson, "Flue Gas Desulfurization Economics." In Proceedings:
Symposium on Flue Gas Desulfurization, New Orleans, March 1976,
Vol. I. EPA-600/2-76-136a (NTIS PB 255-317), May 1976, pp. 79-99.
*
28. McGlamery, G. G., Stern, R. D. and Maxwell, M. A. "The Federal
Interagency Flue Gas Desulfurization Program." In Proceedings;
Second National Conference on the Interagency Energy/Environment
R&D Program, June 1977, in press.
29. PEDCo-Environmental, Inc. "Summary Report - Flue Gas Desulfuriza-
tion Systems." April-May 1977. EPA Contract 68-02-1231, Task
No. 28.
30. Rossoff, J., R. C. Rossi, R. B. Fling, W. M. Graven, and P. P. Leo,
Disposal of By-Products from Nonregenerable Flue Gas Desulfuriza-
tion Systems: Second Progress Report, EPA-600/7-77-052, May
1977. (In Print).
31. Tennessee Valley Authority. Pilot-Plant Study of an Ammonia
Absorption - Ammonium Bisulfate Regeneration Process, Topical
Report Phases I and II. EPA-650/2-74-049a (NTIS PB 237-171),
June 1974.
-------
293
STATUS OF FLUE GAS TREATMENT TECHNOLOGY FOR CONTROL OF NO
AND SIMULTANEOUS CONTROL OF SOY AND NO X
A X
J. David Mobley and Richard D. Stern
Industrial Environmental Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
ABSTRACT
The status of flue gas treatment technology for control of
NOX and simultaneous control of NOX and SOX applicable to stationary
combustion sources is presented. Dry processes and wet processes
are described and applications discussed with respect to performance,
operating experience, and economics.
As a result of a very stringent NOX ambient standard in
Japan, the Japanese NOX flue gas treatment technology appears to
be the most advanced in the world. For this reason, Japanese
technology is emphasized in the paper. EPA's past, current, and
planned flue gas treatment program is also discussed.
INTRODUCTION
Nitrogen oxides (NOX) in the atmosphere have been determined
to have adverse effects on human health and welfare. To aid in
preventing these adverse effects, the Industrial Environmental
Research Laboratory at Research Triangle Park, N. C. (IERL-RTP) is
leading the U.S. Environmental Protection Agency's (EPA) efforts
to develop and demonstrate NOX control technologies for stationary
combustion sources. There are two main technologies being developed:
combustion modification and flue gas treatment.
Combustion modification (CM) technology attempts
to minimize the formation of NOX during the combustion
process. CM techniques include staged combustion,
low excess air operation, flue gas recirculation,
water injection, and burner redesign. CM technology
should be able to reduce NOX emissions from stationary
combustion sources by 50% or more in a relatively
cost effective manner. CM technology will not be
discussed in this paper; however, additional information
is available from other sources.1
-------
294 CLEM .COMBUSTION OF COAL
Flue gas treatment (FGT) technology attempts to
remove NOX from the gaseous products of combustion.
FGT techniques include dry selective catalytic reduction
processes and wet scrubbing processes. FGT technology
should be able to reduce NOX emissions by 90% and has
the potential for 90% control of both NOX and SOX
emissions.
NO FGT research and development programs have received a
relatively low level of funding by EPA since it has not been determined
conclusively that high NOX removal efficiencies will be required to
achieve and maintain the current National Ambient Air Quality
Standards (NAAQS). However, there are significant uncertainties
which may affect the required level of NOX control. Due to these
uncertainties, EPA is proceeding with small scale NOX FGT experimental
projects in parallel with control strategy and technology assessment
studies. One phase of the assessment is an evaluation of Japanese
FGT technology which has progressed to the point of being commer-
cially applied to gas- and oil-fired sources. In addition, the
Japanese are developing processes for application to flue gas from
coal-fired sources. EPA is investigating the Japanese and other
worldwide technologies for potential application to the U.S. coal-
fired situation to save both development time and money. Through
these actions, the basic foundation will be established if the
technology is required in the United States and acceleration of
the development program becomes necessary.
FLUE GAS TREATMENT PROCESSES
There are two main categories of flue gas treatment (FGT)
processes for the control of NOX and the simultaneous control of
NOX and SOX emissions from stationary combustion sources: dry
processes and wet processes. A description of the most promising
processes in each category and their developmental status in Japan
is discussed below.
DRY PROCESSES
The following dry process types are being developed:
Selective catalytic reduction
Selective noncatalytic reduction
Adsorption
Nonselective catalytic reduction
Catalytic decomposition
Electron beam radiation
Of these, only selective catalytic reduction (SCR) has achieved
notable success in treating combustion flue gas for removal of NO
and has progressed to the point of being commercially applied. The
other process types are much less attractive at this time. Selective
noncatalytic reduction processes do not achieve high NO removal
X
-------
NO /SO FLUE GAS TREATMENT 295
efficiences and adsorption processes are not applicable to combustion
sources. Nonselective catalytic reduction, catalytic decomposition,
and electron beam radiation processes are at a very low level of
development. These process types will not be discussed in this
paper, but additional information is available from other sources.2»3»4,5
Dry NOx Processes
Selective catalytic reduction processes are based on the
preference of ammonia (N^) for NOX over other flue gas constituents.
Since the oxygen enhances the reduction, the reactions can best be
expressed as:
catalyst
4NH3 + 4NO + 02 *- 4N2 + 6H20 (1)
catalyst
4NH3 + 2N02 + 02 *- 3N2 + 6H20 (2)
Reaction (1) predominates since approximately 90-95% of the NO in
combustion flue gas is in the form of NO. Since 1 mole of NH3 is
required per mole of NO, most processes operate with an NH^/NO mole
ratio of 0.9 to 1.1. The reaction temperature is usually in the range
of 300 to 450°C, but space velocities vary considerably depending on the
process. Under these operating conditions, NOX removal efficiencies of
90% or greater are typical.^
The catalysts used in SCR processes vary with process developer.
However, there are some general traits known about the catalysts. The
catalyst carrier or substrate is usually alumina, silica, or titanium
dioxide. Alumina is satisfactory for application to flue gases without
SOX such as from natural gas firing. However, alumina tends to react
with SOX, particularly 803, to form aluminum sulfate. This "poisons"
the catalyst by decreasing the available surface area and the catalyst
activity. Titanium dioxide and silica are less susceptible to attack by
803 and are applicable to flue gas from heavy oil or coal firing. The
active metal on the substrate may include Co, Cr, Cu, Fe, Mn, Ni, Pt,
and V or combinations thereof, but the exact composition of the catalyst
is usually proprietary. These metals or their oxides can also react
with the SOX to form sulfates. Many of these sulfates are also cataly-
tic in the reduction of NO with NH3, and therefore, can be tolerated.
The catalysts are normally designed to have a life of at least 1
year.4,5
The formation of ammonium sulfate and bisulfate is a major concern
with SCR processes. Ammonium bisulfate will form downstream of the
reactor if NH3 and 803 are present in sufficient quantities and if the
gas temperature drops sufficiently. It is very difficult to avoid the
conditions for formation; for example, ammonium bisulfate will form if
the gas contains 10 ppm of NH3 and 10 ppm of 803 at a temperature of
210°C. Ammonium sulfate will form at lower temperatures. Fine partic-
ulate emissions of these compounds are a concern, but the major problem
-------
296 CLEAN COMBUSTION OP COAL
with ammonium sulfate and bisulfate is deposition on heat exchanger
surfaces. Since these compounds are very corrosive and interfere with
heat transfer, the heat exchanger must be made of corrosion resistant
material and must be cleaned periodically by soot blowing or water
washing. Approaches to preventing this formation entail use of an
ammonia decomposition catalyst or operation at lower NH3/NO ratios.4,5
Another concern with SCR processes is catalyst plugging. Signi-
ficant progress has been made in avoiding plugging problems through
reactor and catalyst design. Fixed bed reactors, such as parallel
passage, tube, and honeycomb, are being designed which can tolerate
particle loadings typical of coal firing. (Unless otherwise specified,
the fixed bed reactors referred to in this paper will be this open
passage type.) Moving bed reactors are also being developed which can
tolerate and remove moderate amounts of par tides. ^»^ However, the
particle concentrations acceptable to a moving bed reactor are approxi-
mately an order of magnitude less than those tolerated by a fixed bed
system. The space velocity through a moving bed reactor is expected to
be about double that of a fixed bed reactor, but the pressure drop
across the moving bed reactor should be less.
Published information on the cost of SCR processes is limited and
estimates available are based on different design premises. The
reported estimates of the required capital investment range from $10 to
$80/kW with an average of about $30/kW. The revenue requirements range
from 0.2 to 3.3 mills/kWh with an average of about 1.7 mills/kWh.2»5
A list of Japanese SCR process developers is given in Table I along
with information on their developmental status. There are 16 commercial
scale plants in operation treating 100,000 to 750,000 Nm3/hr of flue gas
(33 to 250 MW). In addition, there are 11 prototype plants treating
from 15,000 to 99,999 Nm3/hr of flue gas and numerous pilot and bench
scale plants in operation.6 The prototype and commercial scale plants
are achieving 90% control of the NOX from flue gas derived primarily
from gas- and oil-fired sources. The operating experience in Japan
qualifies NOX control by SCR as a viable control technique in the U.S.
when high NOX removal efficiencies are required from gas- and oil-fired
sources. SCR technology has not been demonstrated in Japan on coal-fired
sources although several of the pilot plants are currently evaluating
such an application and larger scale demonstrations are expected in the
near future. In fact, a northern utility in Japan in considering
installation of a 90 MW SCR system on a coal-fired boiler which will
begin operation in 1980.6
Dry Simultaneous NOX/SOV Processes
There are two noteworthy variations of SCR processes which
have the capability to simultaneously remove NOX and SOX: the
activated carbon process and the Shell copper oxide process.
The activated carbon process requires a special carbon bed
which acts as an adsorbent for SOX and as a catalyst in the
reduction of NOX with NH3. When the bed is saturated with SOX,
flue gas is switched to a fresh bed, the carbon is regenerated'
-------
TABLE I.
MAJOR NO SELECTIVE CATALYTIC REDUCTION PLANTS IN JAPAN
x
Process Developer
Asahi Glass
Hitachi Ltd.
Hitachi Ltd. -Mitsubishi
Hitachi Zosen
Hitachi Zosen
Hitachi Zosen
Hitachi Zosen
Japan Gasoline
Japan Gasoline
Kurabo
Mitsubishi H.I.
Mitsui S.B. -Mitsui P.C.
Mitsui S.B. -Mitsui P.C.
Mitsui Toatsu
Mitsui Toatsu
Mitsui Toatsu
MKK-Santetsu
Osaka Gas
Seitetsu Kagaku
Sumitomo Chemical
Sumitomo Chemical
Sumitomo Chemical
Sumitomo Chemical
Sumitomo Chemical
Sumitomo Chemical
Sumitomo Chemical
Sumitomo Chemical
Plant Owner
Asahi Glass
Kawasaki Steel
PC Mitsubishi PC
Toshin Steel
Idemitsu Kosan
Shindaikyowa P.C.
Kawasaki Steel
Kashima Oil
Fuji Oil
Kurabo
Fuji Oil
Mitsui Pet. Chem.
Ukishima Pet. Chem.
Mitsui Toatsu
Mitsui Toatsu
Osaka Pet. Chem.
Nippon Yakin
Osaka Gas
Seitetsu Kagaku
Sumitomo Chem.
Sumitomo Chem.
Higashi Nihon Met.
Sumitomo Chem.
Sumitomo Chem.
Nihon Ammonia
Sumitomo Chem.
Sumitomo Chem.
Plant Site
Keihin
Chiba
Yokkaichi
Hime j i
Chiba
Yokkaichi
Chiba
Kashima
Sodegaura
Hirakata
Chiba
Chiba
Chiba
Takaishi
Takaishi
Takaishi
Kawasaki
Sakai
Kakogawa
Sodegaura
Anegasaki
Sodegaura
Anegasaki
Niihama
Sodegaura
Sodegaura
Sodegaura
(Nm
70
350
150
70
350
440
750
50
70
30
200
200
240
170
87
90
15
63
15
30
100
200
200
200
250
250
300
„ Capacity
/hr) (-VMW)
,000
,000
,000
,900*
,000
,000
,000
,000
,000
,000
,000
,000
,000
,000*
,000*
,900*
,000
,800*
,000
,000
,000*
,000*
,000*
,000*
,000*
,000
,000
23.
116.
50.
23.
116.
146.
250.
16.
23.
10.
66.
66.
80.
56.
29.
30.
5.
21.
5.
10.
33.
66.
66.
66.
83.
83.
100.
3
6
0
6
6
6
0
6
3
0
6
6
0
6
0
3
0
3
0
0
3
6
6
6
3
3
0
Source of
Gas
Glass Furnace
Coke Oven
Oil-fired Boiler
Heating Furnace
CO-fired Boiler
Oil-fired Boiler
Sintering Machine
Heating Furnace
CO Boiler
Oil-fired Boiler
Oil-fired Boiler
Oil-fired Boiler
Oil-fired Boiler
Ammonia Plant
Ammonia Plant
Heating Furnace
Oil-fired Boiler
Gas Generator
Oil-fired Boiler
Oil-fired Boiler
Gas-fired Boiler
Heating Furnace
Gas-fired Boiler
Heating Furnace
Heating Furnace
Oil-fired Boiler
Oil-fired Boiler
Start-up
Apr.
Oct.
Dec.
June
Nov.
Dec.
Nov.
Nov.
Mar.
Aug.
Late
Sep.
Aug.
Oct.
Mar.
Sep.
June
July
June
July
Feb.
May
Feb.
Mar.
Mar.
Mar.
Oct.
1976
1976
1975
1976
1975
1975
1976
1975
1976
1975
1977
1975
1976
1976
1976
1976
1976
1976
1975
1973
1975
1974
1975
1975
1975
1976
1976
<^,
0
03
O
f
M
Q
Cfl
Kj
W
Kj
§
i-3
Flue gas contains minimum amounts of SO and particles
X
NJ
\O
-------
298 CLEM COMBUSTION OF COAL
and a concentrated S02 stream is produced which can be used to
generate a salable byproduct. The process has the potential
for removing 90% of both pollutants, but its application may be
limited to flue gas containing relatively equal concentrations
of NOX and SOX.^ The economic projections for the process are
about $65/kW for the capital investment and 6.3 mills/kWh in
revenue requirements. The activated carbon process is being
evaluated on a pilot plant scale in Japan by Takeda, Unitika,
and Sumitomo Heavy Industries.-*
In the Shell process, copper oxide reacts with 862 to
form copper sulfate. The copper sulfate and, to a lesser
extent, the copper oxide act as catalysts in the reduction of NOX
with NH3- As in the activated carbon process, a multiple bed
system is required so that a bed is available for acceptance
while regeneration takes place. In the regeneration cycle,
hydrogen is used to reduce the copper sulfate and a concentrated
S02 stream is produced which can be used to generate a salable
byproduct. 4 The economic projections for the process are $131/kW
for the capital investment and 5 mills /kWh for the revenue requirements.
The process has been installed at the Showa Yokkaichi Sekiyu
Company (SYS) plant in Yokkaichi, Japan on a commercial scale
(120,000 Nm3/hr) and has removed 90% of the SOX and 70% of the
WET PROCESSES
The following wet process types are being developed:
Oxidation-Absorption Processes
Absorption-Oxidation Processes
Oxidation- Absorption-Reduction Processes
Absorption-Reduction Processes
Wet N0_ Processes
..... - A.
The first two process types are generally for NOX control
only. In oxidation-absorption processes, the relatively insoluble
NO is oxidized in the gas phase to N02 which is absorbed into
the liquid phase. The typical oxidizing agents used are ozone
and chlorine dioxide. The absorbents vary with process developer.
The process seems more feasible for flue gas containing equimolar
mixtures of NO and N02 which is not typical of combustion flue
gas. Much of the absorbed N02 remains in the liquid phase in
the form of nitrate salts which are water pollutants.1^
In absorption-oxidation processes, NO is absorbed directly
into the liquid phase and then oxidized. Liquid oxidizing
agents such as sodium hypochlorate or hydrogen peroxide are
used to convert the NO to a nitrate salt. Due to the insolubility
of NO, relatively large absorbers are required. In addition,
the process is not applicable to flue gas containing S02 since
the more soluble SO^ would consume the liquid oxidizing agent
by converting the absorbed sulfite ion into sulfate. 4
-------
N(VS°x FLUE GAS TREATMENT 299
Due to their complexity, limited applicability, and water
pollution problems, wet processes cannot compete economically
with dry selective catalytic reduction processes for control of
NOX in combustion flue gas. Therefore, these process types will
not be addressed further in this paper, but additional information
is available from other sources.2,3,4,5
Wet Simultaneous NO^/SO^. Processes
11 " * •'••"• i •• inJv lA, i,,
The attractiveness of wet processes is their potential for
simultaneous removal of NOX and SOX. Oxidation-absorption-
reduction processes and absorption-reduction processes are
designed for this type of control.
The oxidation-absorption-reduction processes basically
evolved from flue gas desulfurization (FGD) systems. A gas
phase oxidant is injected before the scrubber to convert NO to the
more soluble N02- The N02 is then absorbed into an aqueous solution
with S02- The absorbent varies with the type of FGD system being
modified. The absorbed S02 forms a sulfite ion which reduces a
portion of the absorbed nitrogen oxides to molecular nitrogen. The
remaining nitrogen oxides are removed from the waste water as nitrate
salts. The remaining sulfite ions are oxidized into sulfate by air
and removed as gypsum. The percentage of nitrogen oxides going
either to the preferred molecular nitrogen or to the troublesome
nitrate salts is uncertain; however, it is estimated to be about 50-
50, but this can vary considerably.^
The oxidation-absorption-reduction processes have the potential
to remove 90% of both SOX and NOX from combustion flue gas.^
However, there are several drawbacks remaining to be overcome before
the processes can be widely applied. The process chemistry is
complex and use of a gas phase oxidant, such as ozone or chlorine
dioxide, is expensive. Chlorine dioxide, although cheaper than
ozone, adds to the waste water problems created by the nitrate
salts. Chlorine dioxide also causes concern due to the possibility
for chlorination of organics in the waste water to produce carcinogenic
compounds.
Despite these drawbacks, the potential of a simultaneous
removal process warrants further research and development. Table II
lists the process developers evaluating oxidation-absorption-reduction
technology. One small commercial scale plant (33 MW), four prototype,
and three pilot plants are currently being operated in Japan. As
with SCR processes, published economic data are limited, but the
average reported capital investment is about $110/kW and the average
revenue requirement is about 7.5 mills/kWh.
The absorption-reduction processes were seemingly developed to
avoid the use of a gas phase oxidant. A chelating compound, such as
ferrous-EDTA (ethylenediamine tetraacetic acid) which has an affinity
for the relatively insoluble NO, is added to the scrubbing solution.
The NO is absorbed into a complex with the ferrous ion and the S02
is absorbed as the sulfite ion. The NO complex is reduced to molecular
-------
TABLE II.
MAJOR WET NO /SO CONTROL PLANTS IN JAPAN
x x
o
o
Process Developer
Plant Owner
Plant Site
(Nm3
Capacity
(MW)
/hr) Approx.
Source of
Gas
Start-up
OXIDATION - ABSORPTION - REDUCTION PROCESSES
Chiyoda
Ishikawa j ima H.I.
Mitsubishi H.I.
Osaka Soda
Shirogane
Sumitomo Metal-Fujikasui
Sumitomo Metal-Fujikasui
Sumitomo Metal-Fuj ikasui
Chiyoda
Ishikawa j ima H.I.
Mitsubishi H.I.
Osaka Soda
Mitsui Sugar
Sumitomo Metal
Toshin Steel
Sumitomo Metal
Kawasaki
Yokohama
Hiroshima
Amagasaki
Kawasaki
Amagasaki
Fuji
Osaka
ABSORPTION - REDUCTION
Asahi Chemical
Chisso Corp.
Kureha Chemical
Mitsui S.B.
Asahi Chemical
Chisso P.C.
Kureha Chemical
Mitsui P.C.
Mizushima
Goi
Nishiki
Chiba
1
5
2
60
48
62
100
39
,000
,000
,000
,000
,000
,000
,000
,000
0
1
0
20
16
20
33
13
.3
.6
.6
.0
.0
.6
.3
.0
Oil-fired
Oil-fired
Oil-fired
Oil-fired
Oil-fired
Oil-fired
Boiler
Boiler
Boiler
Boiler
Boiler
Boiler
Heating Furnace
Oil-fired
Boiler
Aug.
Sep.
Dec.
Mar.
Aug.
Dec.
Dec.
Dec.
1973
1975
1974
1976
1974
1973
1974
1974
PROCESSES
5
600
300
,000
150
0
0
1
0
.2
.1
.6
.05
Oil-fired
Oil-fired
Oil-fired
Oil-fired
Boiler
Boiler
Boiler
Boiler
Apr.
Apr.
Apr.
Apr.
1974
1974
1975
1974
o
i
o
Q
M
O
o
o
-------
FLUE GAS TREATMENT 301
nitrogen by reaction with the sulfite ion. A regeneration step
recovers the ferrous chelating compound and oxidizes the sulfite ion
into sulfate which is removed as gypsum.-*
The absorption-reduction processes also have the potential to
remove 90% of both the NOX and SOX in combustion flue gas. 4 Although
the processes seem to have advantages over the oxidation-absorption-
reduction processes, there are obstacles to be overcome before the
processes can be widely applied. Even with the addition of the
chelating compounds, a large absorber is required to absorb the NO.
The replacement, recovery, and regeneration costs of the chelating
compounds, although potentially less than the gas phase oxidants,
are still significant. The process chemistry is complex and is
sensitive to the flue gas composition of SC>2, NOX, and oxygen. The
molar ratio of SC>2 to NOX must remain above approximately 2.5 and
the oxygen concentration must remain low.^
Table II also lists the process developers evaluating absorption
reduction technology in Japan. There are four pilot or bench scale
plants currently being operated. 6 The average reported capital
investment is about $96/kW and the revenue requirements are about
6.3 mills/kWh.5
Table III summarizes the status, cost, and performance of the
dry and wet processes for control of NOX and simultaneous control of
NOX and SOX.
CONTROL OF NOx, SOX, AND PARTICLES
Over 90% control of NOX, SOX, and particles may eventually be
required in the U.S. for stationary combustion sources, especially
for new, large, coal-fired sources. Schematics of some of the
alternatives for such an overall control system are shown in Figure
1.
Perhaps the ideal situatiori would be represented by one control
device that simultaneously removes NOX, SOX, and particles. Such a
control device is not yet available, but it is conceivable that a
wet scrubber system could remove all three pollutants. Such a
system is represented in Schematic A of Figure 1. Since the recovery
of heat from the flue gas is important, a heat exchanger is shown in
the schematic. A reheater is also shown since most wet scrubber
systems require reheat of the flue gas prior to its discharge from
the stack.
The most developed overall control system is NOX control by
selective catalytic reduction (SCR), SOX control by flue gas desulfur
ization (FGD) , and particle control by an electrostatic precipitator
(ESP) . The sequence of these control devices is variable as illus-
trated in Schematics B, C, and D. Schematic B shows particle control
first, followed by NOX control, and then S02 control. In this
configuration, the SCR system could be a moving bed type which would
require most of the particles to be removed first. To maintain high
-------
TABLE III. STATUS OF FLUE GAS TREATMENT PROCESSES IN JAPAN
Process Type
Dry Selective Catalytic Reduction
(with ammonia)
Dry Simultaneous NO /SO
} XX
Activated carbon
Shell Copper Oxide
Developmental
Plants
16 commercial
11 prototype
3 pilot
1 commercial
Removal Efficiency (%)
NO SO
x x
_> 90 N.A.
^ 90 ^90
^70 ^90
Approx.
Capital
Cost
($/kW)
30
65
110
Approx.
Revenue
Requirements
(mills /kWh)
1.7
6.3
5.0
0
f
o
o
1
H
O
!2|
Wet Simultaneous NO /SO
X X
Oxidation-absorption-reduction
1 commercial
4 prototype
3 pilot
< 90
> 90
131
7.5
o
o
Absorption-reduction
4 pilot or bench < 90
> 90
96
6.3
-------
NCL,/SOX FLUE GAS TREATMENT
303
Wet Scrubber
for KOX, SOX, &
Participates
SCR
(Moving Bed)
SCR
(Fixed Bed)
ESP
Dry MOX/SOX
(Activated
carbon)
ESP
Wet
NO /SO
x' x
ESP - Electrostatic Precipitator for Particle Control
FGD - Flue Gas Desulfurization for S02 Control
SCR - Selective Catalytic Reduction for NOX Control
FIGURE 1 - POTENTIAL EQUIPMENT CONFIGURATION FOR NO , SO ,
AND PARTICLE CONTROL
x' x'
-------
304 CLEAN COMBUSTION OF COAL
temperatures needed by the SCR system, a hot-ESP would be necessary.
The requirement for a hot-ESP could be avoided by use of a fixed bed
SCR system which could tolerate high particle concentrations. This
configuration is illustrated in Schematic C.
Since the sulfur compounds in the flue gas lower the resistivity
of the fly ash and improve the collection efficiency of the ESP, it
is not deemed advantageous to have the ESP follow the FGD system.
However, it is uncertain if ammonia, ammonium sulfate, or ammonium
bisulfate will leave the SCR system and adversely affect the performance
of the FGD system. To avoid this possibility, the SCR system could
follow the FGD system as shown in Schematic D. However, the dis-
advantage of this configuration is the extensive reheat that would
be required to raise the flue gas temperature to a level suitable
for the SCR system.
The simultaneous NOX/SOX systems present an apparent simplification
of the control systems as shown in Schematics E, F, and G of Figure 1.
Schematic E depicts the activated carbon process and Schematic F
illustrates the Shell process for dry simultaneous NOX/SOX control.
Two configurations are presented due to the different tolerance of
particle concentrations and to the different operating temperatures
of the processes. The wet simultaneous NOX/SO control systems are
shown in Schematic G and can be represented by either the oxidation-
absorption-reduction processes or by the absorption-reduction processes.
At this time it is mere speculation as to which configuration
will emerge as the optimum overall control system from a technical
and economic standpoint. In all probability, the optimum system
will be dependent on specific flue gas and site considerations as
well as user preference.
EPA'S NO FLUE GAS TREATMENT PROGRAM
A
High NOX removal efficiencies may not be required to achieve
and maintain the current National Ambient Air Quality Standards
(NAAQS). Further, the current New Source Performance Standards
(NSPS) for NOX can be achieved by implementation of combustion
modification techniques which are more economical than FGT processes.
However, there are significant factors which may cause the NAAQS and
the NSPS to become more stringent. These include: the alarming
increase in NOX emissions from stationary combustion sources pro-
jected for the next decade, the possibility of a short-term NO?
NAAQS, the relationship of NOX emissions to levels of photochemical
oxidants, the impact of increased use of coal resources, the impact
of relaxing mobile source emission standards, the role of NO
emissions as precursors to other pollutants of concern such as PAN
(peroxyacyl nitrates), other nitrates, nitrosamines, and nitric
acid, and the health effects of NOX and related pollutants.
If more stringent standards are promulgated, then NO FGT
technology may be required to meet the standards to protect human
health and welfare. Therefore, the NOX FGT program is proceeding
-------
N0x/S0x FLUE GAS TREATMENT 305
with experimental projects progressing toward full scale demonstration
of highly efficient NOX and simultaneous NOX/SO control technology
in parallel with control strategy and technology assessment studies.
The results of these studies will assist in determining the appropriate
scale of the experimental projects. EPA's past, current, and planned
activities in these areas are summarized below.
CONTROL STRATEGY AND TECHNOLOGY ASSESSMENT STUDIES
The control strategy and technology assessment studies are
mainly research projects to examine various aspects of NO control
technology and to determine if and when NO FGT technology will be
needed in the U.S. x
Assessment of Japanese Technology
Since Japanese technology in this field is more advanced than
any other country's, EPA has sponsored the publication of periodic
reports and papers to facilitate the transfer of information on NOX
and NOX/SOX abatement technology from Japan. These documents have
been mainly prepared by Dr. Jumpei Ando of Chuo University in Tokyo,
Japan.2,3,4,7,8 Dr. Ando is also assisting EPA in activities
associated with the Stationary Source Pollution Control Project of
the US/Japan Environmental Agreement which includes a subproject on
NOX and NOX/SOX FGT technology.
In addition to monitoring published information from Japanese
sources, EPA personnel have made periodic trips to Japan to observe
testing facilities and to discuss the technology with process
developers and operating personnel. The most recent trip was in
March 1977.
Ozone Oxidation of NO9
Gas phase oxidation of NO to N0£ is essential for wet oxidation-
absorption-reduction processes. Therefore, a task order was issued
by IERL-RTP to the Research Triangle Institute to analyze the supply,
demand, production economics, and energy consumption of this key FGT
process step. The results of the study indicate that only a stoichio-
metric amount of ozone is required to achieve essentially complete
conversion of NO to N02 which may be subsequently scrubbed from the
gas stream.
The energy requirements and the capital and operating costs
were examined for ozone generation with both air and oxygen as input
to the ozone generator. Approximately 13% more energy is required
for ozone generation from oxygen than from air. The capital invest-
ment for ozone generation from oxygen is about 3 times as large as
that required from air, and operating costs are about twice as
large. For a 500 MW plant with air as input to the ozone generator,
the estimates for oxidizing 200 ppm of NO were: energy requirement,
1.1 X 108 kWh/yr or 3.1% of station capacity; capital investment,
$17.60/kW; and operating costs, 2.0 mills/kWh. The 200 ppm concentra-
tion is representative of a coal-fired source with combustion
-------
306 CLEAN COMBUSTION OF COAL
modification techniques applied or an oil-fired source without
supplementary NOX control applied. The estimates are for oxidation
of NO only; the energy requirements and cost of control for N0£
would be additive.
The report indicates that unless there is a significant improve-
ment in ozone generation technology, wet processes using ozone for
oxidation of NO to N02 will be very expensive. However, since these
processes have potential for simultaneous NOX/SOX control, the
energy and cost impacts may be more acceptable.
N0y Control Strategy Assessment
IERL-RTP contracted with Radian Corporation to determine the
potential effectiveness of applying NOX controls to large stationary
combustion sources. The Chicago Air Quality Control Region (AQCR)
was selected for a modeling study of emissions from point, area, and
mobile sources to determine the relative impact of each category on
ambient NOX concentrations. The calibrated dispersion model pre-
dictions of annual average concentrations indicate that the major
point sources, which contributed nearly 40% of the total NOX
emissions in Chicago, accounted for less than 10% of the ambient N02
levels in 1974. Preliminary investigation of expected short-term
concentrations of total NOX shows that major point sources may
contribute as much as 80% of measured NOX levels. Therefore, it
appears that stringent NOX control for large point sources may be
required to meet a potential short-term N02 standard, but cannot be
justified currently on the basis of the existing annual average N02
standard.10 However, NOX emissions from stationary combustion
sources are expected to increase significantly in the next decade.
As a result of these findings, the Chicago AQCR modeling study was
expanded to determine more accurately the short-term ambient N02
levels, to project the annual and short-term N02 concentrations to
1985, and to assess the use of NOX emission control on stationary
combustion sources to attain or maintain compliance with possible
N02 ambient short-term and annual average standards. The results of
this study should be available by early 1978.
Another Radian Corporation study is seeking to determine the
key factors relating to "if" and "when" NOX FGT technology will be
needed in the U.S. Since research and development of a technology
should lead its application by several years, it is necessary to
monitor factors which could require implementation of NOX FGT
technology in the near future. By these efforts, the decision to
emphasize, maintain, or terminate the research, development, and
demonstration of NOX FGT technology can be made on the best available
information. This study should also be available in early 1978.
Economic Assessments of NOy FGT Processes
The Tennessee Valley Authority (TVA), through an interagency
agreement with EPA, is developing comparative economics of NOX and
NOX/SOX FGT emission control processes. This state-of-the-art
review will be conducted in two phases. Phase I will evaluate and
summarize the technical feasibility of all candidate NOX control
-------
NOX/SOX FLUE GAS TREATMENT 307
processes being offered in the U.S. and Japan. The Phase I report,
which includes descriptions of about 45 processes, should be published
in the summer of 1977.5 Phase II will concentrate on the most
promising processes identified in Phase I and will perform a preliminary
economic assessment of each, including development of material and
energy balances. In addition, a direct comparison of the economic
and technical feasibility of the dry and wet processes will be made
to determine the most effective method to remove NOX and SOX from
combustion flue gas. The project is cofunded by IERL-RTP and the
Electric Power Research Institute.
EPA is planning a third phase of the project to prepare detailed
economic projections of as many as four of the most promising process-
es. This activity should be complete in late 1978. Further, it is
contemplated that a study will be conducted during this phase to
determine the impact of ammonia utilization by SCR processes. The
cost and energy requirements of ammonia generation for a typical
utility application will be examined. In addition, the impact on
the supply, demand, and cost of ammonia worldwide will be analyzed.
This study may be available by the end of 1977.
EXPERIMENTAL PROJECTS
EPA's experimental projects have been directed toward enhancing
the evolution of FGT technology from bench scale research to full
scale demonstration on coal-fired sources by the mid-19801s. The
technology must be applicable to utility and large industrial combustion
sources and must achieve highly efficient NOX and simultaneous
NOX/SOX control in a relatively energy efficient and economical
manner.
Bench Scale Catalyst Research^-^-
In 1975, a research grant was awarded to the University of
California at Los Angeles (UCLA), School of Engineering and Applied
Sciences to further the development of promising catalysts. The
study extended the catalyst screening work performed earlier by UCLA
under an IERL-RTP contract with TRW, Inc.12 The objectives of the
grant were to optimize the compositions of vanadium and iron-chromium
catalysts for selective reduction of NOX with ammonia and to perform
long-term durability studies of the optimum catalyst compositions in
flue gas containing sulfur dioxide. The results of the study,
completed in mid-1976, indicated that a 15% loading of vanadium
oxide on alumina support material and a 10% loading of iron oxide-
chromium oxide on alumina support material with an iron/chronium
ratio of 9:1 were the optimum catalyst formulations.
Parametric tests showed that both catalysts were selective, in
that only the NOX was reduced. The tests also showed strong enchancement
of NOX conversion rates due to the presence of Q£ under typical
operating conditions. C02, H20, and S02 did not affect the NOX
reduction in the concentration ranges applicable to power plant
exhaust. Both catalysts were most active between 400°C and 425°C
and required excess NH3 for maximum activity. Long-term durability
tests of both catalysts in the presence of SOX indicated no degradation
-------
308 CLEAN COMBUSTION OF COAL
in catalyst performance. Typical conversion levels for the vanadium
and iron-chromium catalysts operating at 400°C in simulated flue gas
were about 90% and 80%, respectively, at 20,000 hr~l space velocity.
In addition, preliminary tests of iron-vanadium and iron-chromium-
vanadium catalysts indicated 99% removal of NOX from the simulated
flue gas.
Pilot Plant Evaluation of Gas and Oil Firing13
In 1973, a contract was awarded to Environics, Inc. to evaluate
the performance, reliability, and practicality of a SCR system with
ammonia and a platinum catalyst on alumina support material. A
pilot plant, treating a slipstream from an operating utility boiler,
was designed, installed, and tested on gas and oil firing. Laboratory
testing was conducted to supplement the pilot plant testing. Satisfactory
results were found on gas-fired operation with 85-90% NOX removal
achieved for over 4000 hrs at a space velocity of 50,000 hr~l.
Results on oil-fired operation indicate that the catalyst system was
not suitable for flue gas containing SOX. The maximum NOX removal
efficiency achieved was 65% with the average only 50%. Fluctuation
in flue gas temperature and catalyst plugging with soot and ammonium
sulfate caused problems on oil-fired operation.
Pilot Plant Evaluation on Coal Firing
The next phase of the experimental program is evaluation of FGT
processes on a coal-fired application. A request for proposal was
issued in September 1976, and best and final offers are currently
being evaluated. Contract award is anticipated by the end of
September 1977. It is contemplated that two contracts will result
from this procurement process. One will be for a pilot plant to
evaluate removal of NOX emissions, and the other will be to evaluate
simultaneous removal of NOX and SOX. However, budgetary constraints
and technical considerations may impact on the final decision in
this regard.
The pilot plants must treat a flue gas volume equivalent to 0.5
MW and achieve a NOX removal efficiency of 90%. For the simultaneous
control of NOX and SOX, 90% removal of both pollutants must be
achieved. The pilot plant projects will each consist of a 24 month
program which will be conducted in four phases. Phase I includes
the preparation of a detailed process design and an estimation of
capital and operating costs for the pilot plant. Following erection
of the pilot plant and mechanical acceptance testing in Phase II,
the contractor will perform system start-up and debugging, parametric
testing, and optimization testing over a wide range of flue gas
conditions during Phase III. Phase IV provides for testing and
evaluation of the plant during 90 days of continuous operation. It
is currently anticipated that final reports will be published on the
results of the pilot plant operations in early 1980. A project
manual conveying the total concept of the proposed plant is planned
for early 1978. In addition to these projects, there is the possibility
of a pilot plant project being initiated in 1977 with the Tennessee
Valley Authority.
-------
NOX/SOX FLUE GAS TREATMENT 309
The pilot plant projects will enable an assessment of the
technical, environmental, energy, and economic aspects of applying
N0x and NOX/SOX FGT technology to the U.S. coal-fired situation.
This information, in conjunction with the control strategy and
technology assessment studies, will provide technical and budgetary
direction and emphasis for EPA's NOX and NOX/SOX FGT program.
REFERENCES
1. Bowen, J. S., G. B. Martin, R. D. Stern, and J. D. Mobley.
"Stationary Source Control Technology for NOX." The Second
National Conference on the Interagency Energy/Environment R&D
Program, Washington, D.C., June 6 and 7, 1977.
2. Ando, J., R. D. Stern, and J. D. Mobley. "Status of Flue Gas
Treatment Technology for Control of NOX and Simultaneous Control
of SOX and NOX in the United States and Japan." American
Institute of Chemical Engineers, 69th Annual Meeting, Chicago,
Illinois, November 28 - December 2, 1976.
3. Ando, J., H. Tohata, and G. A. Isaacs. NO Abatement for
Stationary Sources in Japan. PEDCo-Environmental Specialists,
Inc. EPA-600/2-76-013b (NTIS No. PB 250 586/AS), January 1976.
U.S. Environmental Protection Agency, Research Triangle Park,
N.C.
4. Ando, J., H. Tohata, and K. Nagata. "NOX Abatement for Stationary
Sources in Japan - August 1976." PEDCo-Environmental Specialists,
Inc., (Draft Report, to be published Summer 1977 by EPA).
5. Faucett, H. L., J. D. Maxwell, and T. A. Burnett. "State-of-
the-Art Review of Processes for Removal of Nitrogen Oxides from
Power Plant Stack Gas." Tennessee Valley Authority (Draft
Report, to be published Summer 1977 by EPA).
6. Ando, J., personal communication with J. D. Mobley, 06/25/77.
7. Ando, J., and H. Tohata. Nitrogen Oxide Abatement Technology
Japan - 1973. Processes Research Inc., EPA-R2-73-284
(NTIS No. PB 222 143), June 1973. U. S. Environmental Protection
Agency, Research Triangle Park, N. C.
8. Ando, J. "Status of Flue Gas Desulfurization and Simultaneous
Removal of S02 and NOX in Japan." In Proceedings: Symposium on
Flue Gas Desulfurization, New Orleans, March, 1976, Vol. I,
EPA-600/2-76-136a (NTIS No. PB 255 317), May 1976, pp 53-78.
U.S. Environmental Protection Agency, Research Triangle Park,
N. C.
-------
310 CLEAN COMBUSTION OF COAL
9. Harrison, J. W. Technology and Economics of Flue Gas NOV
Oxidation by Ozone, Research Triangle Institute, EPA-600/7-76-
033 (NTIS No. 261 917/AS), December 1976. U.S. Environmental
Protection Agency, Research Triangle Park, N.C.
10. Eppright, B.R., personal communication with J. D. Mobley, 09/27/76.
11. Nobe, K., G. L. Bauerle, and S. C. Wu. Parametric Studies of
Catalysts for NOV Control from Stationary Power Plants, University
of California, Los Angeles, EPA-600/7-76-026 (NTIS No. PB 261
289/AS), October 1976. U.S. Environmental Protection Agency,
Research Triangle Park, N.C.
12. Koutsoukos, E. P., J. L. Blumenthal, M. Ghassemi, and G. L.
Bauerle. Assessment of Catalysts for Control of NOX from
Stationary Power Plants, Phase I, Volume I, Final Report,
TRW, Inc., EPA-650/2-75-001-a (NTIS No. PB 239 745/AS), January
1975. U.S. Environmental Protection Agency, Research Triangle
Park, N.C.
13. Kline, J. M., P. H. Owen, and Y. C. Lee. Catalytic Reduction
of Nitrogen Oxides with Ammonia: Utility Pilot Plant Operation,
Environics, Inc., EPA-600/7-76-031 (NTIS No. PB 261 265/AS),
October 1976. U.S. Environmental Protection Agency, Research
Triangle Park, N.C.
-------
311
SESSION V - WHERE DO WE GO FROM HERE?
SESSION CHAIRMAN: VICTOR S. ENGLEMAN, SCIENCE APPLICATIONS, INC.
The papers in previous sessions have presented the foundation on
which we stand and have supplied the building blocks to achieve clean
combustion of coal. This closing session takes a look into the future
in terms of policy, supply, and technology for direct coal utilization.
It is clear from the conference thus far that precombustion, combustion,
and postcombustion technologies will all "be important in achieving "Best
Available Control Technology" for coal utilization. The National Energy
Plan will have a major impact on the overall energy picture and the
incentives and requirements for coal combustion. Increased coal utili-
zation will require increased supplies which will require new mines and
increased capacity in the delivery system. New developments and improve-
ments in technology will be required to achieve the likely environmental
requirements in a technically and economically sound manner. The papers
in this session focus on these three key areas in the future of clean
combustion of coal.
-------
312 CLEAN COMBUSTION OF COAL
-------
313
Transcript of Talk on
THE NATIONAL ENERGY PLAN
)
by
C. William Fischer
Department of Energy
Washington, D.C.
August 5, 1977
I spoke to Jim Schlesinger a week ago today in connection with the
job of putting together a planning mechanism for the new Department of
Energy and he gave me some very succinct and terse guidance which I
believe is particularly relevant to this conference and which is why I
am more eager to come here and listen to you than to talk. He said,
"You know I have been around Washington a little bit." I should say
that Jim Schlesinger, now Secretary Schlesinger, is quite a remarkable
guy. His nomination and confirmation for the job of first Secretary of
Energy in less than one day must have set a speed record. Hearings on
the nomination were held yesterday morning from 8:00 to 11:30, and in
the afternoon the Senate voted to confirm. But there is another record
here: This is the fourth cabinet-level agency Jim Schlesinger has
headed in the last seven years. Prior to becoming the President's
chief energy advisor, Schlesinger was Chairman of the Atomic Energy
Commission, Director of the Central Intelligence Agency, and Secretary
of Defense. Prior to that he was Assistant Director of the Bureau of
the Budget which was his forst job in Washington. His directions to me
the other day were very clear. He said that in the defense business and
in some of the other places where he has worked, you make a decision on
a project, you make some estimates about what it ought to cost to imple-
ment, and then you summon the four-star generals and you say this is
what I want you to do, this is how much money you have to do it with,
and you may also call in the president of North American Rockwell and
say that if you do it for that amount within that time schedule this is
what the fee will be, and if you don't make the time schedule or you
don't make the performance standard, then the fee will be reduced accord-
ingly, and away you go from there. That is exactly the opposite of what
the Department of Energy must do. The Department, if it is going to do
its job for the Nation, should set a broad policy framework of overall
objectives and then determine, with the help of the people who are going
to have to work with these energy problems, how to get from here to
there. In setting the broad policy goals we should ask, "What is feas-
ible?" The output of this department is not going to be a set of things
or a set of widgets; it is going to be a set of decisions that have to
be made by millions of consumers and thousands of leaders in the corpo-
rate structures and the Federal, State and local public institutions of
this country. The Federal Government will not be able to design,
develop, direct ajid implement this program as though it is in control
-------
314 CLEAN COMBUSTION OF COAL
of all, or even the majority, of the tools necessary to solve our energy
problem. We are going to have to do an awful lot of listening, and we
are going to have to do a lot of cooperative work with the people who
have to make the programs work. I thought I would share these thoughts
with you. It is a perspective that is far from messianic. Just the
opposite; it is something that is, I think, a realistic recognition of
what it will take to move this country toward more plentiful energy
resources and away from the profligacy in the use of energy that we have
come to know, in part because of our pricing policy. This basic philos-
ophy of listening to the public, is particularly applicable to me. I am
not an engineer; I am not an expert; I have no specialized knowledge of
the technical parameters within which we must operate. That is why I
said last night that after this conversation I would like to hear your
recommendations on legislation to facilitate siting, both for nuclear
and nonnuclear power.
As we begin our discussion this morning, it might be worth it to
review the gravity of the situation that we face. The world community
is now using about 60 million barrels of oil a day. We are increasing
our demands for that scarce commodity each year by about 5 percent and
we are just now loading the first tanker out of Alaska. The entire
yearly Alaskan North Slope production, when at full yield, is going to
be roughly equivalent to the world's increase in consumption over a
period of nine months. The whole of Texas oil production would equal
the world's increase in oil consumption for one year; total Saudi Arabian
production is roughly equivalent to the projected net increase in world
consumption over the next three years. World production can probably
keep increasing for six to eight years. Sometime in the late 80's or
early 90's, however, it won't go up much more. If we could hold the
growth rate in world demand to five percent yearly, we would use all of
today's proven reserves—notice I said proven reserve, not inferred, not
potential—by the end of the next decade. There are lots of arguments
about how much resource there is in the ground. Oil companies are making
all kinds of noises that if you just gave them a little more incentive
they could turn all inferred and potential reserves into proven reserves.
In that regard, I would like to talk about the general structure of
the National Energy Plan, and, specifically, about this incentive ques-
tion that has to do with basic pricing policy. Later, I would like to
concentrate on the coal part of the program and give you an update on
what the specific provisions were in the President's Plan, what the Ways
and Means and the Commerce Committees have done to the regulatory and
tax provisions, and what are the major issues pending for House floor
resolution today, and I mean literally today, that could alter the Plan
significantly. Then we will open it up to questions and discussion.
Alaskan flow, at maximum production, will equal about two years'
growth in U.S. energy demands. If we don't act now, by 1985 we will be
using about 33 percent more energy than we are using today and because
we probably can't substantially increase domestic oil production, we
will have to. import twice as much oil. To give you a frame of reference,
in 1970 we were paying $2.7 billion a year for imported fuels. By 1976
the cost of imports was $36 billion. In 1977 we expect to be paying
$1*5 billion for oil imports by 1985- What are we doing about this?
Before the National Energy Plan we were continuing merrily on our way.
-------
NATIONAL ENERGY PLAN 315
We Charted a twelve-month moving average of gasoline consumption in the
United States. The chart shows that in August of 1973 we were consuming
6.6 million barrels of motor gasoline per day. By August of 1975 we
were still using about 6.6 MMB/D but this statistic reflected the period
of the 1973-7U embargo when there were physical restrictions on supply,
long lines for gasoline and so on. By June of 1977 this country was
consuming 7.1 million barrels of gasoline per day. Since June of 1975
we have been increasing consumption at the rate of 3.8 percent a year.
Production rates, on the other hand, have decreased. Another twelve-month
moving average for production shows that in August of 1973 this country
was producing 9.3 million barrels of crude per day. By June of 1977
production was down to 8 million barrels per day—the equivalent to a
decreasing rate of 3 percent per year over the past two years. In
January of 1973 we were importing U.8 million barrels a day; by January
of 1977, 7-5 million barrels a day—a rate of increase of 17.2 percent
per year since June of 1975- The Administration's program consists of
more than 150 specific provisions which are now moving through the United
States Congress faster than I have ever seen any bill of that size, on
any subject, move. Earlier this week I was engaged in a public television
discussion with Representative Jim Jones from Oklahoma and the Moderator
commented that in April when the President gave his message there was
much concern expressed, but that now things have quieted down and there
seems to be an energy gap in the Congress. Well, the first thing I did
was to deny that. I had never seen such a sustained level of intense
Congressional inquiry, debate, and activity. It may be that the coinci-
dence of the timing of the National Energy Plan hitting the Congress and
the Speaker's dedication to getting control of the House of Representa-
tives may have facilitated fast action on the bill. I don't agree with
this interpretation of the reasons for this speedy consideration of the
bill by the House. I think Congress is exhibiting genuine interest in
the energy problem. But the road is not smooth.
We have some of our greatest problems in the coal area—which perhaps
reflects the technical difficulties we encounter in this area and the
difficulties that have been discussed here in balancing our enhanced
concern for the environment with our consciousness that we have to move
at least in the intermediate term to more plentiful sources of energy.
In some cases these difficulties can be resolved together—their solutions
are not mutually exclusive nor inconsistent. I am thinking specifically
of automobile emissions. It now looks as though going to cleaner tail-
pipes may also get us more efficiency. Cutting down the size of automo-
biles, which our European friends have seen fit to do for quite some
time, actually accomplishes' both goals simultaneously. But there are
plenty of areas where we have very real and difficult technical problems
to solve. You are in a position to help find some of the solutions.
Beyond finding solutions to technical problems, some necessary compromises
will have to be made to get us where we need to go in terms of balancing
our need to burn coal, for instance, and our desire to protect the envi-
ronment .
The Administration's program is basically four-pronged. First, we
must have adequate price incentives to produce additional resources
without undue enrichment. That is the guts of the new pricing policy
which says that we are going to try to get the price of scarce fossil
fuels up to their replacement value without causing undeserved windfall
-------
316 CLEM COMBUSTION OF COAL
gains to producers. The mechanism for avoiding windfall gains to pro-
ducers will be a tax coupled with a rebate to consumers to help compen-
sate for higher fuel prices. There is room for disagreement on whether
or not feeding these taxes that are imposed on the higher prices at the
pump back to the consumer will in fact put a demand restraint in the
system. It remains to be seen whether prices together with the rebate
will have the full effect of reducing consumption. I believe that it
will, because people react to their immediate consumption patterns and
if they have a rebate in the future, they will not necessarily spend it
on fuel. The second prong of the National Energy Act is the strong
incentive to conserve without an unfair burden on the consumer. Adequate
price incentives for production and correspondingly higher prices are the
front part of this element of the plan. Prices will go up for both
natural gas and oil—especially for new production, at the margin. We
define "new" production in terms of where the well is drilled with regard
to an already-existing producing well—at which depth and at what distance
in relation to that last producing well. In case of gas, the Administra-
tion has agreed to a new reservoir concept, the details of which are yet
to be worked out. And prices for those "new" commodities will be essen-
tially uncontrolled. Theoretically, they would go to the world price.
Additionally, residential, commercial and industrial consumers would be
encouraged to conserve through adopting energy efficiency measures, such
as retrofitting homes and businesses. Tax credits are provided for this
purpose. Third are the proposals for strong incentives to use more
abundant fuels, that is, the conversion program and the development of
safe nuclear energy. Fourth is the vigorous research and development
program on new and renewable energy resources.
With regard to the incentives for oil production, one of the propos-
als , known as the Jones-Schroeder Amendment, would allow producers a tax
break against their well-head tax due. The credit would equal one-half
a percent a month for each month for which exploration expenditures
exceeded 25 percent of gross revenues. At the end of the first year the
credit would equal 6 percent of the well-head tax due. This amendment
which the Administration opposed was defeated yesterday on the floor of
the House. The reason I am discussing the amendment is because it goes
right to the heart of the major producer's case—that there are inadequate
incentives for exploration and that more incentives are needed to develop
new fields. First the plow back would be in addition to proposals that
are in the President's bill to put secondary and tertiary enhanced
recovery at the world price and all newly discovered oil at the world
price by 1980, plus adjustments to account for the domestic rate of in-
flation. It is also in addition to the Administration's provisions that
would allow domestic producers to receive a greater-per-barrel margin
than producers any place else in the world. We surveyed the different
world markets and found that with the world price being set on new oil
production, domestic producers would receive a greater-per-barrel margin
than they would any place else in the world including the North Sea and
other fields. The Ways and Means Committee estimated that the Jones-
Schroeder proposal would increase oil industry revenues by $89 million
in 1978; we estimate that the amendment would result in an oil industry
revenues increase of up to $100 million. If the tax break is made to be
cumulative, then, by the end of 1980, the plow back would increase the
oil industry's revenues by about $2 billion and $5.2 billion by the end
of 1985.
-------
NATIONAL ENERGY PLAN 317
Another problem with the Jones-Schroeder proposal is that the
25 percent threshold is so low the producers would be rewarded for doing
something that they are doing already. According to the Bureau of the
Census 1975 Annual Survey of Oil and Gas, independent producers were
already reinvesting 63 percent of their revenues, apart from royalty
revenues which account for 12 percent of their total revenues, for
exploration and development. If you make the adjustment for royalty oil
you are still above the 25 percent threshold. In 1975 the majors—the
top eighteen oil producers—were reinvesting 35 percent of their revenues.
Additionally, the plow back would cause considerable administrative
difficulties since it would be based on taxes and consequently the crude
oil equalization tax would have to be levied on producers rather than
first purchasers as was proposed in the Administration bill. Taxes
would have to be collected from over 1600 producers rather than refiners
and other first purchasers of which there are about 300. Furthermore,
the cash balances that the major producers will have without the plow back
after their investment in exploration and development is adequate. These
funds are, of course, not excess profits. They are funds for use in
other business purposes; they are essential to the business enterprise.
They can be resources for alternative investments, or can be paid out to
shareholders as dividends. But these are funds that could be used for
additional investment for exploration and development which are not being
so used. These producer cash balances amounted to $2 billion in 1972 and
went up to $U.9 billion in 1975.
Still another problem is that the credit is available only to pro-
ducers of flowing oil. A producer has to be established to have the
25 percent threshold ascertained. New entrants to the field would not
be able to earn the credit. Major oil companies—the top 18—own two-
thirds of all old domestic oil for which the plow back would be available,
and, therefore, the majors would benefit the most. The reported domestic
profits of the largest 18 oil companies increased from $3.5 billion in
1972 to $5-7 billion in 1975—a 60 percent increase in reported profits.
Profits for 1975 were second only to 197^ profits in the history of the
industry. Almost all of this increase in income comes from domestic
rather than foreign production.
The government is currently quite limited in its ability to examine
the industry profit structure. The Administration has put forward a
proposal, authorized under existing law and mandated under the Department
of Energy Organization Act, to improve the financial reporting by major
producers. The major producers are not averse to the proposed Petroleum
Company Financial Reporting System—they recognize that they have credi-
bility problems.
Capital investment in development and exploration has increased in
recent years. The average annual level in the early 1970Ts was about
$9 billion a year; the actual spending in 1975 was $21.7 billion. In
February of this year the Oil and Gas Journal projected that 1977 expen-
ditures for development and exploration would be up to $26.7 billion.
The sum total of this evidence leads us to believe that the profits are
up, the investments are up, and the physical activity is up with regard
to exploration and development without price decontrol. That gets us
back to the question of whether the degree of enrichment that would
-------
318 CLEAN COMBUSTION OF COAL
result for producers from decontrol and the rise in prices to consumers,
which the producers themselves are advocating, is necessary to stimulate
exploration and development activity. ¥e think there is reason to ques-
tion the producers' arguments. With regard to physical activity, the
number of rotary rigs in operation within the U.S., including Alaska, off-
shore, and the outer Continental Shelf, has increased significantly. In
1968, out of 2100 total rigs, 1500 were active and 600 were inactive. In
1972 the numbers dipped to about 1^00 active, and the total number of
rigs dropped to about 1760. By 1976 the total number of rigs was back
up to 2200. There are now almost 2000 active rigs. The number of exist-
ing but inactive rigs is down to 225- About 90 percent of the total gas
and oil rigs were active as of 1976. I've seen these figures in the
annual rotary rig census of the Reid Tool Company. Hughes Tool does a
similar census and they have some more recent numbers. Their figures
indicate that the number of rigs active in 1977 is also around 1900 to
2000. So exploration and development activity is up and the number of
inactive rigs is significantly down. And this is all prior to the
President's policy under which new oil goes to the world price and new
reservoirs go to the world price all of which makes for additional incen-
tives. By the way, 2% miles from a currently producing well is not a
great distance in determining what is so-called new gas and oil.
In spite of all this activity, however, additions to reserves have
declined, and proven reserves have dropped by 12 percent since 1972, to
a proven domestic reserve level of 30.0 billion barrels at the end of
1976. What has happened is that as the prices have risen, and they have
risen substantially even under the controlled situation, we have made
wells in fields that were earlier passed over. Instead of stepping out
into new reserves and new pools, producers are going back and opening up
previously uneconomic reserves that don't expand our reserve. As it
turns out there is also no significant effect on production—domestic
oil production is continuing to fall. So this raises the question whether
higher prices do yield significant additional production. For gas there
have been increases in intrastate prices from 1970 averaging about 25<£
per Tcf to $1.30, to $1.97 and in some intrastate cases to above $2.00
(in some new contracts in parts of Texas and Louisiana) and interstate
increases from 35<£ to an average of $1.21 per Tcf in 1976. Gas well
drilling is today more than twice the level it was in 1971, but again
proven reserves have shrunk by over 25$ from the 1967 high of 293 tril-
lion cu ft.
A few words about the oil and gas replacement program or the so-
called coal conversion program. The proposal has two pieces and affects
two major sectors. The pieces are the regulatory and the tax provisions.
They affect major fuel-burning industries as well as utilities. I will
discuss the regulatory proposals and what has happened to them in the
Congress. Our regulatory proposals would cover industries and utility
units capable of consuming fuel at a fuel heat input rate of 100 million
Btu per hour or greater or a combination of several such units which have
a total capability of consuming fuel at a fuel input rate of 250 million
Btu's per hour or greater. In industry it includes boilers, combustors,
combined cycle units and internal combustion engines. I am going to go
through these provisions and tell you what the principal committees in
the House, where the action is right now, have done to those provisions.
But before I begin, to give you some perspective, the total amount of oil
-------
NATIONAL ENERGY PLAN 319
imports that the President's program would save is kit million barrels a
day. Total U.S. consumption of oil in 1976 vas 17.1* million barrels a
day. Of that 17-U MMBD consumed in 1976, our imports were 7.3 million
barrels a day. If we don't do anything, by 1985 our oil imports will go
to 11.5 million barrels a day. Under the President's program, imports
would level off at 7-1 to 7-2 MMBD which means about a IK 5-million-
barrels-a-day savings. Of that, 2.2 million barrels a day or about
50 percent is to be achieved through the oil and gas replacement program.
The House Commerce Committee, which has jurisdiction over the
regulatory provisions has made a fundamental change in our proposal with
regard to industry. The Committee bill excludes combustors from the
proposed regulatory activity. If this provision is finally approved it
will result in an oil savings loss of somewhere between 250 thousand
barrels and half a million barrels a day for that provision alone. Under
the President's bill no oil and gas will be burned by new utilities.
Exceptions are made for environmental site limitations. Existing utili-
ties will not be allowed to burn gas after 1989. If a utility is coal-
capable it will be ordered to switch to coal. If it is not coal-capable
it can be ordered to shift from gas to oil. The Commerce Committee
version deletes this last provision. The Administration proposal makes
exceptions for existing utility units based on environmental grounds or
site limitations or the use of synthetic fuels. With regard to new
industrial units no oil or gas may be used for boiler fuel. Nonboiler
installations may also be prohibited by regulation or order from using
oil or gas. Once again, exceptions will be made based on site limita-
tions, environmental grounds, or technical impossibility. The House
Committee has made no change in these provisions and they are up for a
vote on the floor of the House today. Existing industrial units which
are coal-capable may be prohibited by regulation or order from using oil
or gas. If they are capable of using mixtures or combinations of fuels,
they may be ordered to use the minimum possible amounts of oil or gas.
Exceptions here also are based on site limitations, environmental grounds
or technical impossibility.
I should also discuss briefly the tax provisions. On the industry
side there would be a single tax rate effective in 1979 at a level of 900
a barrel for oil used starting in 1979 and would phase up to $3 a barrel
for oil used in 1985, adjusted for inflation. Effective in 1979 gas
used would be taxed at a rate equal to the difference in Btu equivalent
price between gas and distillate fuel minus $1.05 per million Btu of gas
used during that year. The $1.05 deduction phases out, so that by 1985
the tax will equal the difference in price between gas and Btu equivalent
distillate. The Ways and Means Committee went from a one-tier tax
provision and made a distinction between the tax rate for boilers,
turbines and other internal combustion engines, and other nonexempt
industrial users, mostly coal-capable nonboilers. The Committee also
basically reduced the initial tax from 900 (in the case of industrial
oil) to 300 a barrel, but phasing up to the same $3 a barrel by 1985-
These reductions in taxes are estimated to lose, in terms of oil saved,
another half a billion barrels of oil. Industries and firms using more
than 500 billion Btu's or 85,000 barrels of oil equivalent a year—account-
ing for less than 2 percent of the total number of industrial firms or
about lilOO out of 100,000 firms— will be taxed under these provisions.
-------
320 CLEM COMBUSTION OF COAL
The Ways and Means Committee included an exception from this tax
for process use of oil or natural gas if other fuels are not usable for
technical, economic or environmental reasons. The Administration's
proposal, as well as in the Ways and Means proposal, made exceptions for
transportation including aircraft, rail and water, farming, ammonia feed-
stocks, refining, automotive or special fuels, and for gas that is
reinjected for the extraction of additional gas. The Ways and Means
Committee added exemptions for residential facilities, commercial
facilities, and any facilities not an integral part of manufacturing,
processing or mining. Additionally, Congressman Steiger introduced an
amendment in the Ways and Means Committee that would extend exemptions
from the regulatory provisions for environmental and technical feasibility
reasons to the application of the tax. The Administration's bill provided
that, even though you might be exempt from the prohibition against using
oil and gas, you would still have the tax imposed. The Steiger amendment
would say that if you were exempt from the regulation you are also exempt
from the tax. The Steiger amendment adds another loss in savings of
barrels of oil a day. The Gorman amendment that is up on the floor today
would eliminate that additional exemption. Finally, the Ways and Means
Committee has added two other exemptions to the bill. Industries that
can show that they would be placed at a competitive disadvantage would
be exempt. That adds up to another 100,000 barrels of oil a day.
Another amendment provides a procedure whereby the Secretary would have
discretionary authority to reclassify industrial uses of oil and gas to
a lower tier tax rate or altogether exempt such uses from tax. That
would add another loss in savings of a quarter of a million barrels a
day.
With regard to the utility oil tax, the Administration proposed a
250 per million Btu or $1.50 a barrel flat tax beginning in 1983 but
with inflation adjustments beginning on the date of passage. The Ways
and Means Committee put a 250 a million Btu or $1.50 a barrel flat tax
beginning in 1983, but with inflation adjustments beginning in 198l.
The utilities gas use tax begins in 1983 and is keyed to the regional
distillate price as a target. In 1983 the tax would be 500 a million Btu
less than the target price. By 1988 it would be equal to the target
price, that is, the regional distillate price. Inflation adjustments are
made from 1975- Once again, the Committee would start the tax in 1983,
initially at 550 per million Btu and going to 750 per million Btu in
1985. The tax can't exceed the regional residential gas prices and it
is inflation-adjusted from a later date. This loses the additional one-
half a billion barrels that I mentioned previously. All those additional
amendments and exemptions reduce the 2.2 million barrels a day oil saving
that is supposed to result from the coal conversion program to somewhere
between .8 and 1.0 million barrels a day. The Administration's bill also
contains a rebate program which is basically a tax exemption from this
oil and gas use tax for investment in alternative energy properties,
including boilers not using oil or gas, coal utilization equipment, low
Btu gas control equipment, required pollution control equipment, but
excluding buildings and structural components. The Ways and Means
Committee added to that high-Btu-gas-from-coal equipment, up to the
turbine stage, and for geothermal and hydroelectric power.
For utilities the rebate is the same as the industrial rebate in
the Administration's bill except that in the case of the industry there
-------
NATIONAL ENERGY PLAN 321
is an alternative. They can either take the rebate or they can take an
additional 10 percent investment tax credit on top of their existing
lO^percent investment tax credit. This does not apply to utilities. The
utility rebate is available only for investments that replace or convert
existing oil- or gas-fired facilities. Industrial firms are not limited
in the Administration's bill to conversion or replacement investments.
The Ways and Means Committee provides that the rebate for utilities is
also available for phasing down instead of only being available for the
conversion and puts the incentive on phasing down and converting from
base load to peak load use of oil- or gas-fired facilities. That also
would mean a substantial loss in energy savings.
I would say that the major issue on the House side that is up for
action today is the Corman-Steiger debate on whether the regulatory
exemption should also apply to the taxes. The rationale here, of course,
for allowing the tax to be imposed even when the regulations are not, is
that you need a variable incentive to move people away from the use of
oil and gas or to allow them, in the case of large firms—and these
taxes only apply to large firms—to take advantage of being able to
convert where they can convert, and pay the taxes where they choose not
to convert or where they can't convert. Those taxes would then be recap-
tured for them in terms of their conversion investments on their units,
because they can trade off inside the firm the taxes they pay on one
unit with the rebates they get for the conversion of another unit.
I hope I have given you some idea of the importance of these pro-
visions. I appreciated very much getting from you your suggestions on
what the Federal government ought to do to facilitate this conversion
to more plentiful sources of energy both in terms of its regulatory
activities and the siting question and in terms of the technical activity
in research and development.
-------
322 CLEAN COMBUSTION OF COAL
-------
323
THE OUTLOOK FOR COAL THROUGH 1985
Robert L. Major
Manager - Market Research
AMAX Coal Company
Indianapolis, Indiana
Although the Carter Administration has indicated that it is
counting on coal to be the "swing fuel" which will allow us to reduce
our dependence on imported oil, it has refused to come to grips with
the real and potential barriers to the increased mining, transporta-
tion, and utilization of coal.
The future role of coal must be evaluated within the context of
the total energy picture. While it is obvious that the United States
needs a "comprehensive national energy policy," there has been little
agreement as to what specific policies should be included in such a
national energy policy. In general, industry had in mind a program
which relied on the market forces of supply, demand, and price to
allocate scarce energy resources among the various competing uses and
to achieve desired fuel switching. However, as Walter Mead stated
recently in an article in Science:
". . .to political Washington, the cry for a national energy
policy is interpreted as a demand for more government decision-
making and less reliance on the market."*
Mead further states that most professional energy economists have
argued for less government and more reliance on the market to solve
the "energy problem." This preference does not stem from a political
conservatism as much as from their awareness of the "poor record of
government interference in the energy market. That record is one of
massive and repeated resource misallocation" (emphasis addedT-* The
result of such an approach is that great emphasis is being placed on
the "stick" and little on the "carrot." A "stick" policy which seeks
to force certain actions which run counter to the internal economics
of the energy user is likely to be met with strong opposition, while
policies which seek to create economic incentives ("carrots") to
switch fuels or to use energy more efficiently are much more likely to
succeed. With these brief remarks as an introduction, I would now
like to discuss the general factors underlying the markets for coal.
*Mead, W. J., 22 July 1977, "An Economic Appraisal of President
Carter's Energy Program," Science, vol. 197, p. 340.
-------
324 CLEM COMBUSTION OF COAL
General Market Factors
Coal markets tend to be segmented by coal types. The primary
segmentation of the market is between metallurgical and steam coals.
Among the steam coals there is a further segmentation by coal types--
bituminous, subbituminous, lignite, and anthracite. Each of these
coals tends to serve special niches within the overall coal market.
Some of the markets are much more constrained than others. Metal-
lurgical coking coals are specialty coals which command premium
prices. The demand for these coals is determined by the output of pig
iron and steel, both here and abroad. The bulk of the coals exported
overseas from the United States (excluding Canada) is of the metal-
lurgical variety. Although there has been some movement of lower
grade (high-volatile) metallurgical coals into the "compliance" steam
coal* market in recent years, the markets for metallurgical and steam
coals are essentially separate. Because the main growth sector over
the next decade will be in the steam coal market, little further
discussion of metallurgical coal markets will be included in this
paper.
Steam coal is the largest segment of the coal industry, and
electric power plants are the largest consumers of steam coal, account-
ing for some 68 percent of total coal consumption in 1976. Lesser
amounts of steam coal are used by industry for process steam, space
heating, and self-generation of electricity. Other uses are minor and
are expected to essentially disappear in the future. Potentially, the
largest user of steam coal in the future will be coal-based synthetic
fuel plants. However, because of technical, economic, and financial
problems, it now appears that coal gasification and coal liquefaction
will not be a significant factor in the market until 1990 or later.
Traditionally, the utility and industrial coal markets have been
strongly regionalized because of the limitations of transportation
costs, and consumers have tended to buy coal from the nearest coal
supply areas. The "shape" of the market was determined by the de-
livered cost of various coals vis-a-vis the delivered cost of other
coals and other fuels. This regionalization of the markets has been
broken down to some extent by the new air pollution control regula-
tions for power plants, which have forced utilities to use more dis-
tant "compliance coals" rather than local high-sulfur coals. As a
result, large tonnages of Western compliance coals have been moving
east into the Upper Midwest markets at the expense of previously used
Illinois Basin coals. In an attempt to slow down this movement and to
restore the use of "Eastern" coals, the Carter Administration has
proposed a policy in which scrubbers would be mandated on all new
power plants regardless of the sulfur content of the coal burned
*"Compliance coal" is used here to designate any coal whose
sulfur content is sufficiently low so that it can be burned without a
scrubber and still be in compliance with the 1.2-pound of SOo/MM
Btu standard. *
-------
OUTLOOK FOR COAL 325
Anthracite coal production is mostly limited to the northeastern
part of the State of Pennsylvania. At one time, it was mined in
significant quantities and supplied a large share of the energy market
in the Northeast. However, high mining costs combined with competi-
tion from cheaper oil and gas have resulted in this industry's shrink-
ing to a mere shadow of its former self. At present, anthracite
accounts for less than 1.0 percent of the total U.S. coal production.
However, this coal does possess several advantages. It is well
situated to serve the populous Northeast energy markets and it is a
high-Btu, low-sulfur fuel. As a result, FEA has recently established
a task force to assess the feasibility of increasing the use of anthra-
cite in the future.
Among the steam coals, lignites play a special, more restricted
role. Although the United States has vast reserves of lignite which
can generally be mined at low cost, the low heat value (6,000-7,000
Btu per pound) and high moisture content of the lignites tend to
restrict their use to markets near the lignite deposits. With few
exceptions, lignite is utilized at mine-mouth locations for electric
power generation. At present, the bulk of the lignite produced in the
United States is mined in the states of North Dakota and Texas. While
the use of such local lignites reduces the demand for coals in more
distant supply areas, the converse is not true, because lignite cannot
compete in distant markets.
In the case of bituminous and subbituminous coals, the product is
sufficiently high in quality that long distance movements of such coal
are competitive under certain conditions. However, in most cases the
markets for these coals are strongly regionalized. It is useful to
discuss the markets for these coals in terms of the main coal supply
regions—Appalachia, the Illinois Basin, and the "West." The West, in
turn, contains a number of subregions, each of which serves its own
markets.
- The great bulk of the new expansion in coal production in the
United States through 1985 will occur in the West, especially in the
Powder River Basin of Wyoming. It is important to discuss this expan-
sion, because there is a great deal of confusion about this develop-
ment. The demand for this Western coal arises from three distinct
markets, each of which is controlled by different economic variables.
The first market demand for Western coal is from utilities and in-
dustries located in the Western coal supply areas to meet local
demands for energy.
The second market demand is for Western coal to serve utilities
and industries located in the West but outside of the coal-producing
areas. This market will be served both by coal used in mine-mouth
power plants whose electrical output is "wheeled" to distant market
centers and by coal which is transported to distant markets by train,
barge, and/or slurry pipelines. This new demand for Western coal is
primarily stimulated by the shortfall in natural gas supplies, which
is causing utilities and large industrial energy users in the West
(especially in the West South Central Region) to build new coal-fired
plants to replace their old gas-fired plants.
-------
326 CLEAN COMBUSTION OF COAL
The third market demand is the so-called "compliance coal" market.
This market is stimulated by the desire of utilities (mainly in the
Upper Midwest) to avoid the installation of expensive flue gas desul-
furization units ("scrubbers"), which would be required if they con-
tinued to use high- and medium-sulfur coals from the Midwest. These
utilities and industries, in some cases, are paying a considerable
premium over the cost of local coals in order to purchase low-sulfur
(compliance) coals from the West or from Appalachia. The impact of
this penetration has been most severe on the market for Illinois Basin
coals.
Future Coal Demand
Table 1 indicates our forecast of the total demand for coal in
1985 by major consuming sectors. You will note that we are more
optimistic than the Carter Administration regarding the demand for
utility coal, but much less sanguine in regard to industrial demand.
For the sake of comparison, we have also included the forecasts pre-
pared by Joel Price of Dean Witter and by Island Creek Coal Company.
The key point to note is the consensus that the Carter Administration
is too optimistic, and that it is unlikely that U.S. coal production
will reach 1.27 billion tons by 1985.
Table 2 indicates the trends in the demand for coal by electric
utilities through 1985. These data indicate estimated coal demands by
NERC region. The boundaries of the NERC regions are shown in Fig-
ure 1. Although the overall growth rate of 7.3 percent per year may
appear to be high, we feel that it is reasonable because the phase-out
of natural gas as boiler fuel, and the delays in bringing new nuclear
plants on stream will force utilities to rely heavily on coal for
their future fuel supplies. The very high growth in the Southwest
regions of ERGOT and SPP is mostly a response to the phasing out of
extensive gas-fired capacity.
Despite the rapid growth in utility coal use in the West, from
89 million tons in 1976 to 337 million tons in 1985, some 59 percent
(488 million tons) of the total coal used by utilities in 1985 will
still be consumed east of the Mississippi River. Secondly, although
there has been much talk about Western coal being used in the East,
only about 56 million tons, or 11.5 percent, of the 1985 coal supply
for Eastern utilities will be Western coal. Therefore, more than
85 percent of the coal expansion in the West is expected to serve
Western markets.
The future demand for coal by industry is very difficult to
predict because of the great uncertainty which exists regarding the
Federal Government's policies on coal use and environmental regula-
tions. However, if we assume that the long-term switch away from coal
can be reversed, and that part of the new energy demands by industry
can be captured by coal, then we estimate that the potential regional
demand for industrial coal in 1985 could be as shown in Table 3. The
boundaries of the regions used in Table 3 are indicated in Figure 2
Under either scenario, the bulk of the future industrial coal usage*is
expected to occur in only two regions: the Midwest and the Southwest
-------
Table 1
COMPARISON OF UNITED STATES COAL DEMAND PROJECTIONS FOR 1985
BY MAJOR CONSUMING SECTORS
CONSUMING SECTORS
Electric Utilities
AMAX
COAL (1)
824
no
107-154
92
-
(Million Tons)
ISLAND
CREEK (2)
744
TOO
90
90
-
DEAN
WITTER (3)
778
no
104-134
87
-
CARTER
ADMINISTRATION (4)
779
105
278
90
12
1
Metallurgical 110 100 110 105 §
1-3
Industrial 107-154 90 104-134 278 o
Q
Export
Synthetics
o
Other
TOTAL 1133-1180 1024 1080*-1110* 1265
(1) AMAX Coal Company, Market Research Department, July 1977.
(2) Barker, Stom'e, Jr., February 8, 1977, "Coal - Current Status and Future Prospects", speech
before the Washington Society of Investment Analysts, Washington, D.C.
(3) Price, Joel, June 15, 1977, "The Supply and Demand for Steam Coal: Implications and Challenges",
an address before the Richmond Society of Security Analysts, Richmond, Virginia.
(4) Executive Office of the President, Energy Policy and Planning, June 2, 1977, Replacing Oil and
^ With Coal ajid^ Other Fuels In Tne Industrial and Utility Sectors, page III-2.
* Does not add due to independent rounding.
-------
328
CLEAN COMBUSTION OF COAL
Table 2
UTILITY COAL DEMAND, BY NERC REGIONS*
(Million Tons)
Annual Percent
NERC Region
ECAR
ERCOT
MAAC
MAIN
MARCA
NPCC
SERC
SPP
WSCC
TOTAL
1976
142.2
12.0
32.8
57.3
28.4
7.8
104.3
9.1
39.9
433.8
1980
166.6
39.9
38.6
73.4
45.9
10.1
117.5
48.3
70.7
611.0
1985 Growth, 1976-85
207.4
69.3
40.3
86.1
66.9
13.8
139.9
101.4
98.9
824.0
+ 4.3
+21.5
+ 2.3
+ 4.6
+10.0
+ 6.5
+ 3.3
+30.7
+10.6
+ 7.3
* See Figure 1 for regional boundaries.
SOURCE: National Electric Reliability Council, 1977
Preliminary data.
-------
UNITED STATES
OF AMERICA
FIG.* I
SCALE OF MILES
0 100 200 300 400
o
o
3
o
o
U)
NO
FIG. * I - BOUNDARIES OF NATIONAL ELECTRIC RELIABILITY COUNCIL REGIONS
-------
330
CLEAN COMBUSTION OF COAL
Table 3
INDUSTRIAL COAL DEMAND IN 1985
BY REGIONS*
(Million Tons)
LOW GROWTH** HIGH GROWTH***
New England
New York/New Jersey
Mid-Atlantic
South Atlantic
Midwest
Central
North Central
Southwest
West
Northwest
Total
0.7
3.6
15.9
13.4
31.6
4.7
4.4
25.6
4.1
2.6
106.6
1.3
4.8
18.5
18.1
38.7
5.9
6.2
49.0
7.7
3.9
154.1
*See Figure 2 for FEA industrial energy regional boundaries.
**"Low Growth" scenario assumed that industry within each region
will continue to use coal at the same rate that it did in
1975 plus 10 percent of the incremental increase (1975-1985)
in energy demands will be met by coal.
***"High Growth" scenario is similar to low growth, but assumes
that 20 percent of the incremental demand (1975-1985) will
be met by coal.
-------
(NEW ENGLAND
NORTH CENTRAL
UNITED STATES
OF AMERICA
FIG. #2
SCALE OF MILES
0 100 200 300 400
T I I I I
Q
O
o
LO
CJ
FIG. #2 - BOUNDARIES OF FEA INDUSTRIAL ENERGY DEMAND REGIONS
-------
332 CLEAN COMBUSTION OF COAL
However, it should be noted that depending on the" regional availability
of natural gas and future pricing policies of the Federal Government,
the regional shares could be altered significantly. Despite this, it
is our opinion that the total industrial coal demand in 1985 will be
in the forecast range.
Regional Demand
Now that we have estimated the future demand for coal through
1985, let us examine the question as to which regions will supply the
coal to meet the forecast demand. Table 4 indicates our forecast of
the 1985 coal demand by market sector and by supply region. Despite
the large increase in Western coal production between now and 1985,
Eastern and Midwestern coals are expected to continue to play a
significant role in meeting our future demands for coal.
Supply/Demand Balances
Recent research by AMAX Coal has indicated that while the overall
coal industry is expected to have surplus capacity through the early
1980's, potential shortfalls* in steam coal capacity could develop in
the Illinois Basin by 1979 and in Appalachia by 1981. While part of
the potential shortfalls in the East could be filled by surplus
Western coal, this option is not likely to offer a total solution to
the supply problem. Lastly, because there are a number of environ-
mental, legal, and technical problems which might reduce the effective
supply of Western coal by as much as 50 million tons or more below the
currently forecast levels, the "surplus" capacity projected for the
West may never materialize.
Summary
If public policy conflicts regarding the mining, transportation,
and utilization of coal are not quickly resolved, coal's future role
will be diminished and our dependence on imported oil will continue to
increase. Many of the problems and uncertainties which face the coal
industry are the result of our failure to develop a consistent set of
policies for coal. Melvin Laird, in an excellent little pamphlet
entitled Energy - A Crisis In Public Policy makes the following
observation:
*Note: These potential shortfalls were derived by the difference
between the estimated demand and the projected supply
(which is equal to current capacity minus depletions plus
announced capacity). Because the data on new mine capacity
in the 1981-1985 period are less complete for the East,
there is a tendency to underestimate actual supply. The
projected shortfalls should be taken as indicators of the
regions in which there is a potential demand for additional
mine capacity.
-------
Table 4
COAL DEMAND IN 1985 BY MARKET AND SUPPLY REGION
(Million Tons)
SUPPLY REGION
SECTOR/MARKET
Electric Utilities
Industrial*
Metallurgical Coal
Exports**
— Metallurgical Coal
— Steam Coal
Total
APPALACHIA
285
52-65
97
76
16
526-539
ILLINOIS
BASIN
152
18-23
5
-
-
175-180
TEXAS
LIGNITE
56
13-26
-
-
-
69-82
NORTH
DAKOTA
LIGNITE
23
1-2
-
-
-
24-25
WESTERN
COAL
308
23-38
8
-
-
339-354
TOTAL
824
107-154
no
76
16
1133-1180
s
o
o
p
* Includes minor usage by commercial and retail customers.
** All exports allocated to Appalachia as exports from other regions have been small
and irregular.
U)
u>
to
-------
334 CLEAN COMBUSTION OF COAL
"The failure of the Federal Government to come to grips with the
crucial environmental/energy trade-offs is partially attributable
to the fact that, in the Congress as well as in the executive
branch, policy direction for energy and environment is treated
separately. With missionary zeal, those in government dealing
with each set of issues push their own programs toward conclu-
sions that are often contradictory and sometimes actually impos-
sible of resolution by the decisionmaker in the private sector,
at the end of the regulatory chain."*
We, in the coal industry, are ready and willing to do our part in
solving the energy problem. However, unless many of the policy road-
blocks are resolved quickly, it will be very difficult, if not impos-
sible, for coal to play as large a role in our future energy supply as
is currently projected. Even under the most optimistic assumptions,
it is questionable whether we will reach the Carter Administration's
goal of almost 1.3 billion tons by 1985. Coal can be a part of the
solution to the energy problem of the United States, but it is not a
panacea.
*Laird, M. R., 1977, Energy - A Crisis In Public Policyi
American Enterprise Institute for Public Policy Research, Washington,
u * i>>»9 p. y •
-------
335
WHERE DO WE GO FROM HERE IN R&D?
by
S. William Gouse
Deputy Assistant Administrator
Fossil Energy, ERDA
ERDA is going into the Department of Energy on October 1, 1977.
Figure 1 shovs the projection of the major Fossil Energy activities (not
including mining and coal preparation) for four years "beyond FY '78.
Once in the Department of Energy, we will have added to the program coal
mining research and development and coal preparation work. The numbers
for FY '78 are pretty firm, totaling a little over $600 million of new
contract authority. The numbers for FY '79 and "beyond are what the
program managers are projecting as their needs for carrying out the
program that has been initiated in the past and for the next fiscal year.
That also includes some new starts that they believe they will be able
to make as a result of research now underway. As you can see, it goes
over a billion dollars per year in FY '80.
The chart shows a general breakdown of activities. Coal conversion
includes gasification and liquefaction. Utilization means coal-oil
slurry combustion, atmospheric and pressurized fluid-bed combustion,
and direct combustion; as well as advanced high temperature turbines
for use of coal-derived fuels. Advanced research and supporting tech-
nology means, what it says with respect to research, but includes mate-
rials and component development work in support of the total program.
MHD is self-explanatory. Demonstration plants cover the large
commercial scale projects we have under contract or are about to enter
into contract. Petroleum, natural gas, and oil shale are not a subject
of this Conference, but they are in the Fossil Energy Program of ERDA
and they are likely to be in the program in the Department of Energy.
It is a large program.
At this point in time, there are about 800 work efforts under con-
tract in place around the world. I do not know how many subcontracts
there are. There are the order of 200 contracts in universities alone
on fairly basic studies. I believe our feeling, at the present time,
is that the level of support in new or exploratory R&D has been limited
by the number of ideas rather than by funds or other policy judgments.
We are probably funding more than we ought, rather than the other way
around.
The growth of the program has been very rapid. In FY '73, Fossil
Energy's budget was the order of $50 million. Such rapid growth cannot
take place without errors in judgment. In addition, some of our programs
are in place by Congressional direction. This rate of growth of an R&D
-------
FIGURE 1
U)
u>
FOSSIL ENERGY OVERVIEW
BUDGET AUTHORITY DISTRIBUTION
BY MAJOR PROGRAM AREA
FY 1977-1982
MODIFICATIONS AT ENERGY RESEARCH CENTERS
4001—
1977
1978
1979
1980
1981
1982
-------
COAL R&D 337
activity is faster than is prudent when one thinks in terms of making
wise investment decisions. However, the nature of the problem facing
the country caused Congress to launch an accelerated activity with
supplemental appropriations in FY '7^, with the understanding that such
an acceleration would bring with it more risk than had been the practice.
When such an increase in obligational authority takes place, with pres-
sure to obligate rapidly, you are forced to put under contract the best
of what is available at that particular time. Once something substan-
tial is under contract, jobs are at stake and it is difficult to stop.
Efforts are made to redirect and make projects more effective if
it is determined that the original thrust is not as useful as it might
have been. However, even though one finds that certain undertakings
could be judged to have been mistakes from a Federal point of view, it
may not be so. In the process, we train people; we improve the general
knowledge base in fossil fuels; and, of course, we learn considerable
about the behavior of components, materials, and chemistry of various
unit operations involved in coal conversion and utilization.
So far as Fossil Energy's research, development and demonstration
program in ERDA is concerned, we believe all of it supports the National
Energy Plan with respect to improving the prospects and potential for
increased use of coal, and making available as reserves, large resources
of oil and gas in tar sands, bitumens and various tight formations. We
do not see any breakthroughs coming and we do not see any results that
indicate that getting coal, shale, tight formations, etc., is going to
be inexpensive or simple. People fail to observe that we gave up exten-
sive use of coal in many applications for good reasons. The world used
considerable amounts of coal oil between whale oil and natural petroleum
and stopped because natural petroleum is much easier. In addition, many
parts of the world used coal-derived gas to operate cities—some until
the late 1950's. Again, that was slowly phased out for good reasons.
Natural gas is cleaner, easier, and less expensive. Now, we have a
great deal of coal and we may go back to it. But, I think we should
understand that the reason we go back is not because it is easy or inex-
pensive, but because it is the only choice we can see in front of us.
There are many people entering the research and development arena
today who make the mistake of net looking into the history of coal use
around the world. This, of course, has to be done carefully. If you
look too hard at what has been done, then you fall into the trap of
beginning to believe that there are no new ideas left to be found. That
is not true. But, if one approaches R&D without looking at what has
been done, then there is a great deal of time and effort wasted in re-
inventing the wheel. Coal is a very difficult material—every piece is
different, even inches apart in the same seam. It has many constituents,
including potential toxic substances which have not yet been regulated.
Some products of coal are carcinogenic. The manufacture of coal-derived
products has to be done with appropriate regard for the health and safety
aspects of employees. While one can refine coal-derived liquids into
specification-grade products that we are accustomed to dealing with,
there is movement now to examine the impact of petroleum-derived liquids
on health and safety. Regulations that would come from there would, of
course, apply to coal liquids.
-------
338 CLEAN COMBUSTION OF COAL
Today, one can purchase a license to operate coal liquefaction and
gasification plants. They are very expensive and not competitive with
alternative fuels. Atmospheric fluid-bed combustion appears to be
coming along well. You probably will be able to buy one with commercial
guaranties in the not-too-distant future, especially at the small indus-
trial scale. Advanced power systems are also under development. Higher
fuel costs will justify more complex technical systems to increase
efficiency. The ideas involved really are not new, but the technology
involved will require materials which do not now exist.
MHD is the Fossil Energy Program's long shot. It is an interesting
approach for several reasons. One is that it offers the highest ultimate
conversion efficiency when examined in terms of energy in coal to elec-
tricity off the busbar. The other reason has nothing to do with genera-
tion of electricity. To make MHD concepts work on coal, you must have
a. combustor up front that delivers products of combustion at the order
of U500°F. The present system seems to be favoring a one- or two-stage
slagging combustor. If such combustors are actually made to function
reliably, it will have a great deal of fallout in the development of
coal gasifiers. MHD actually requires more severe service than coal
gasification schemes under development. The MHD people are really out
at the frontier of all the technology they are concerned with. They
have had to think in terms of revolution in their concepts, rather than
evolution. Even if only partial success is achieved in the MHD objective,
the fallout will be substantial and beneficial to many applications. In
addition, one has to inject and recover seed material. The recovery must
take place at high temperature and, thus, extremely difficult hot gas
cleanup conditions have to be dealt with. Again, if this sort of problem
is solved in MHD, then it undoubtedly will have application to the other
combined cycle operations involving cleanup of coal-derived gases. The
ultimate configuration for coal-fired MHD that provides the maximum
potential efficiency would be a three-cycle type of plant; that is, an
MHD converter, some kind of gas turbine and a steam plant. This is a
complex engineering system involving extraordinary operating conditions
and being able to keep all parts operating at the same time for reasonable
periods of time. This probably is one of the greatest engineering chal-
lenges facing fossil energy R&D.
We are just beginning to look at three liability and operating
problems of such systems as advanced combined cycles. The Petroleum and
Natural Gas Program has several thrusts. In the Enhanced Oil Recovery
Program the objective is to be able to economically extract very large
amounts of oil, in place, left behind after primary recovery has been
completed. In the gas area, the principal thrust in either the Eastern
or Western formations is to develop technology to economically exploit
the large amounts of gas available in tight formations. What we mean
here is a conventional well will not produce at a rate sufficient to
make very useful recovery or investment possible. Fracturing or other
techniques may raise production rates to where exploitation becomes
attractive. In the gas area, we are essentially trying to convert a
resource to a reserve. In the Eastern Devonian fields, there are con-
siderable numbers of producing wells. These are usually in areas with
extensive natural fracture patterns and, thus, have sufficiently high
production rates. Another interesting characteristic of wells in
Devonian fields is that they produce for very long periods of time,
-------
COAL R&D 339
albeit at a lower rate. Many holes have been drilled in Devonian fields--
in tens of thousands. Most of them have been declared dry holes. I think
you are aware that, generally, if a hole is not a producer, it is written
off as drilling expense. If one produces from a hole, then the hole
must be capitalized. Many times it is not useful to produce from low-
production rate holes because the investment recovery is too slow. It
is conceivable that one could increase production in Devonian shales in
a significant way by just not requiring capitalization of low producers.
Within Fossil Energy, we are estimating that gas from tight forma-
tions may be less expensive than gas from coal. However, until enough
work is done, this will not be well understood.
Synthetic pipeline quality gas from coal will be the order of $3 or
more per million Btu. Our lowest estimates are for processes that are
in the development stage rather than for processes that are in the stage
of commercial availability. Some of this low-potential cost of the more
advanced processes may be real. On the other hand, as is often true,
development of economics are attractive early in the development process
because one has not yet fully developed their technology and come to
grips with all the problems. ¥e are estimating low-Btu gas to be the
order of $2.30 a million Btu or thereabouts with technology that is
currently available.
The major uncertainty is in the environmental area, but I do not
think that this is a serious problem. It is mainly uncertainty because
such units have not been operated under modern standards. There does
not appear to be any unusual process requirements for cleanup of low-Btu
gas processes. ¥e have eight industrial or commercial low-Btu demonstra-
tions under contract at the present time for various applications over a
range of sizes using a variety of coals. It will almost all be operating
in 1980. Assuming reasonable success, these ought to demonstrate the
capability of this technology for meeting certain potential demands. If
natural gas prices go in the direction people seem to think they are
going in, these industrial low-Btu applications ought to be commercially
attractive. Also, one can operate a low-Btu gasifier system to meet very
tight environmental standards—tighter than is possible in a direct
combustion application.
Liquefaction of coal seems to indicate a product at the order of
$30 a barrel; shale liquids, the order of $20 a barrel. While these
numbers sound high compared to world price of oil, they are not high in
a functional way. Gasoline at 75<£ a gallon without taxes is really
pretty interesting compared to walking. Of course, the economy is not
adjusted to that sort of cost and people are not accustomed to thinking
about it. People often ask why we have both a shale liquids program and
a coal liquids program. They are different raw materials and produce
products of different characteristics. Coal liquids are very aromatic.
Shale liquids are not nearly so. The route to gasoline is probably
through coal. The route to jet fuels and other distillates is probably
through shale. One can produce either product from either source, but
the economics are more attractive with a particular feedstock/end-product
combination such as shale oil to distillate rather than shale oil to
gasoline.
-------
340 CLEAN COMBUSTION OF COAL
In addition, there is the Fischer-Tropsch process currently in
operation in South Africa. If operated with high temperature entrained
gasifiers, it is potentially a very clean process from an environmental
point of view and, with development of more selective catalysts, capable
of producing a wide range of hydrocarbon liquids.
At the present time, this looks to be the most expensive process.
Again, perhaps we know most about it and, therefore, we have the most
realistic number.
QUESTION: WHAT ABOUT THE ENERGY EFFICIENCY OF THE FISCHER-TROPSCH
PROCESS?
ANSWER: It all depends on how you look at the process. Even in
Fischer-Tropsch, if you try to make all one product (such as gasoline),
the efficiency is probably hd percent. However, coal does not really
want to behave that way. Neither does shale. Most processes involving
coal or shale are somewhat similar to processes for finding oil: You
put something in, you heat it up and a range of products come out. If
one can arrange a process that produces a useful product mix, even the
Fischer-Tropsch process may be as much as 60 percent efficient.
Other liquefaction processes may be 70 percent efficient. While
it may be easier in the early stages of building synthetic fuels industry
to have a single-product plant, it is probably likely that large-scale
commercialization would have multi-product plants.
QUESTION: WHAT ABOUT FLASH HYDROPYROLYSIS?
ANSWER: We are moving ahead rapidly with very rapid hydropyrolysis.
The work on the West Coast with Rocketdyne is very encouraging—maybe
because we have so little data. At this time, this process route seems
to offer very high grade hydrocarbon products with potentially very
large capacity per unit of investment in steel.
In general, as we have learned more about any process, it is sur-
prising how the economics of all converge on something fairly high and
not too far from existing technology. Look at a coal liquefaction or
gasification plant of nearly any kind—the order of 30 percent to UO per-
cent of the product price is the coal feedstock. About 50 percent or so
is the cost of money. The rest is other costs. R&D can make its biggest
impact on decreasing the amount of capital required. Actually, if we
could figure out a way to use the interest rate with our R&D funds, we
would have a very high benefit/cost ratio. The capital charges become
very small if one could go to a forty-year write-off and 3 percent money.
However, we have been doing many process evaluations trying to
understand where the capital and operating costs are. The principal
objective is to identify targets of opportunity for Federal research
dollars.
CONCLUDING REMARKS:
I have tended to indicate the problems associated with coal conver-
sion are very difficult. They are. It is not going to be inexpensive,
-------
COAL R&D
341
tov ^straightforward. However, research and development will help
to make the plants more reliable, more easily operable, more environmen-
tally acceptable and safer from an occupational health and safety point
-------
342 CLEAN COMBUSTION OF COAL
-------
343
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO
EPA-600/7-78-073
2.
3. RECIPIENT'S ACCESSION NO.
, TITLE AND SUBT.TLE
proceedings rf ^ Engineering
Foundation Conference on Clean Combustion of Coal
5. REPORT DATE
April 1978
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Victor S. Engleman, Conference Chairman
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Science Applications, Inc.
L200 Prospect Street
La Jolla, California 92038
10. PROGRAM ELEMENT NO.
EHE624A
11. CONTRACT/GRANT NO.
EPA Purchase Order
DA-7-03662B
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 CC
Proceedings; 5/77-2/78
COVERED
14. SPONSORING AGENCY CODE
EPA/600/13
15.SUPPLEMENTARY NOTES IERL-RTP project officer is G. B. Martin (MD-65, 919/541-2235).
(*)Cosponsors! project officers are A. Macek of ERDA and A. W. Deurbrouck of USBM
(both now part of DOE).
16. ABSTRACT
The proceedings document the 27 presentations made during the Engineering Founda-
tion Conference on Clean Combustion of Coal, at Rindge, NH August 1-5, 1977.
Sponsored by the Environmental Protection Agency, the Energy Research and Devel-
opment Administration, and the Bureau of Mines (the last two now part of the Depart-
ment of Energy), the Conference dealt with the technical, economic, environmental,
and policy aspects of clean combustion of coal. The five Conference sessions dealt
with problem definition, precombustion processes, combustion processes, postcom-
bustion processes, and future prospects. The Conference was intended to provide an
assessment of the status and trends of clean combustion of coal.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Pollution
Combustion
Coal
Coal Preparation
Coal Gasification
Flue Gases
Combustion Control
Combustion Effi-
ciency
Pollution Control
Stationary Sources
Flue Gas Cleaning
Combustion Modification
13 B
2 IB
2 ID
081
13 H
3. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
347
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
-------
|