&EPA
United States
Environmental Protection
Agency
Industrial Environmental Research EPA-600/7-79-098a
Laboratory April 1979
Research Triangle Park NC 27711
Proceedings: Symposium
on Coal Cleaning to
Achieve Energy and
Environmental Goals
(September 1978,
Hollywood, FL) -
Volume I
Interagency
Energy/Environment
R&D Program Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the
effort funded under the 17-agency Federal Energy/Environment Research and
Development Program. These studies relate to EPA's mission to protect the public
health and welfare from adverse effects of pollutants associated with energy sys-
tems. The goal of the Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology. Investigations include analy-
ses of the transport of energy-related pollutants and their health and ecological
effects; assessments of, and development of, control technologies for energy
systems; and integrated assessments of a wide range of energy-related environ-
mental issues.
EPA REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved
for publication. Approval does not signify that the contents necessarily reflect
the views and policies of the Government, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document is available to the public tnrough the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/7-79-098a
April 1979
Symposium on Coal Cleaning
to Achieve Energy and Environmental Goals
(September 1978, Hollywood, FL) -
Volume I
by
S.E. Rogers and A.W. Lemmon, Jr. (Editors)
Battelle Columbus Laboratories
505 King Avenue
Columbus, Ohio 43201
Contract No. 68-02-2163
Task No. 861
Program Element No. EHE624A
EPA Project Officer: James D. Kilgroe
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
ABSTRACT
The Symposium on Coal Cleaning to Achieve Energy and Environmental Goals
was sponsored by the U.S. EFA's Industrial Environmental Research Laboratory
under Contract No. 68-02-2163, Task No. 861. The Symposium was held September
11-15, 1978, in Hollywood, Florida. The program provided an opportunity
for mutual review and discussion of the physical and chemical coal cleaning
programs of EPA, DoE, the Electric Power Research Institute, those of
numerous industrial organizations, and European and Soviet plans for the
future, as well as the problems of ongoing operations.
The Proceedings contain the contributions of the participating speakers
and include the following topics:
(a) Coal Characteristics
(b) Coal Cleaning Overview
(c) Physical Coal Cleaning Technology
(d) Environmental Assessment and Pollution Control Technology
(e) Chemical Coal Cleaning Technology.
ii
-------
FOREWORD
Man and his environment must be protected from the adverse effects of
pesticides, radiation, noise, industrial effluents, and other forms of
pollution, as well as the unwise management of solid waste. Efforts to
protect the environment require a focus that recognizes the interplay among
the components of our physical and biological environment—air, water, land,
plants, and animals. The industrial Environmental Research Laboratory (IERL/
RTF) of the U.S. Environmental Protection Agency (EPA) located at Research
Triangle Park, North Carolina, contributes to -this multidisciplinary focus
through programs engaged in:
• studies on the effects of environmental contaminants on
the biosphere, and
• a search for ways to prevent contamination and to recycle
valuable resources.
This Symposium Proceedings deals with the subject matter of concern to
an IERL/RTP program designed to focus on the effectiveness and efficiency of
coal cleaning processes as a means of reducing the total environmental impact
of energy production through coal utilization. The Symposium itself provided
a most vital communication link between the researcher and engineer on the
one hand and the user community on the other. To enhance future communica-
tion processes and encourage future applications of coal cleaning technology,
this Symposium Proceedings documents the results of the meeting held.
iii
-------
ACKNOWLEDGMENT
No symposium can be a success without the support and participation of
the attendees. For those participants listed elsewhere in this document,
then, we are grateful for their contributions. Thanks in full measure is
due also to the Session Chairmen and Cochairmen who patiently labored to
formulate an informative and timely program. And, of course, none of this
would have been possible without the long hours spent by the authors indivi-
dually and collectively, in the preparation of their papers.
Thanks are also due for the handling of the mechanical details of the
Symposium. These necessary functions were performed ably by a number of
people. Mr. Jack H. Greene (IERL/RTP) was responsible for the overall
arrangements with the hotel, and Ms. Susan R. Armstrong, Conference Coor-
dinator at Battelle's Columbus Laboratories (BCL), managed the day-to-day
activities. She was assisted in the many details of the necessary operations
by Ms. Joyce B. Fowler (IERL/RTP), Mrs. Rebecca S. Miller (BCL), and Mrs. Lucy
6. Fierson (BCL).
Special thanks are expressed to Mrs. Alexis W. Lemmon, Jr., and
Mrs. L. David Tamny for their efforts in making the week more pleasant
for the distaff accompaniers of the Symposium participants.
In the preparation of the printed Symposium Proceedings, Ms. Sharron
E. Rogers performed excellently aa Technical Editor. Mrs. Miller and
Mrs. Pierson organized, formatted, and provided the necessary typing. We
are grateful for their assistance.
iv
-------
TABLE OF CONTENTS
Page
Abstract ±±
Foreword Hi
Acknowledgment iv
Final Program 1
Conference Report and Activities 5
Session 0; Coal Characteristics
PETROGRAPHY OF COAL 9
Ronald W. Stanton and Robert B. Finkelman
U.S. Geological Survey
MINERALOGIC. AFFINITIES OF TRACE ELEMENTS IN COAL 29
F. L. Fiene , J. K. Kuhn , and H. J. Gluskoter
Illinois State Geological Survey
Exxon Production Research Company
EFFECTS OF COAL CLEANING ON ELEMENTAL DISTRIBUTIONS . 59
Charles T. Ford and James F. Boyer
Bituminous Coal Research, Inc.
PARTICLE SIZE DISTRIBUTION IN THE LIBERATION OF PYRITE IN COAL ... 91
Harold L. Lovell
The Pennsylvania State University
GEOLOGIC CONTROLS ON MINERAL MATTER IN THE
UPPER FREEPORT COAL BED 110
C. B. Cecil, R. W. Stanton, S. D. Allshouse, and R. B. Finkelman
U.S. Geological Survey
INTERPRETING STATISTICAL VARIABILITY 126
Ralph E. Thomas
Battelle's Columbus Laboratories
Session 1; Coal Cleaning Overview
AN OVERVIEW OF EPA COAL CLEANING PROGRAMS 149
J. D. Kilgroe and D. A. Kirchgesaner
U.S. EPA, IERL-RTP
OVERVIEW OF DOE COAL CLEANING PROGRAM 171
Cyril W. Draffin
U.S. Department of Energy
-------
TABLE OF CONTENTS
(Continued)
Page
OVERVIEW OF EFRI COAL CLEANING PROGRAMS 194
Kenneth Clifford and Shelton Ehrllch
Electric Power Research Institute
AN INTEGRATED ASSESSMENT OF COAL TECHNOLOGIES 195
Richard S. Davidson
Battelle's Columbus Laboratories
THE MAIN TRENDS OF WORKS ON ENVIRONMENTAL PROTECTION
AGAINST THE INFLUENCE OF COAL-PREPARATION PLANTS IN THE USSR .... 207
I. S. Blagov, G. G. Vosnyuk, V. V. Kochetov,
I. Ch. Nekhoroshy, and I. E. Cherevko
USSR Ministry of Coal Industry
THE CLEAN FUEL SUPPLY: FACTORS AFFECTING U.S. AND EUROPEAN
S02 EMISSIONS IN THE MID-1980'a 228
Anthony Bromley and Gary J. Foley
Organization of Economic Cooperation and Development
A TECHNICAL AND ECONOMIC OVERVIEW OF COAL CLEANING 256
Horst Huettenhaln, Jackson Yu, and Samuel Wong
Bechtel National, Inc. and Argonne National Laboratory
OVERCOMING THE BARRIERS TO INVESTMENT IN PHYSICAL COAL
CLEANING WITH REVISED NSPS FOR UTILITY BOILERS 298
Karel Fisher and Peter Cukor
Teknekr on, Inc *
ECONOMICS OF COAL CLEANING AND FLUE GAS DESULFURIZATION FOR
COMPLIANCE WITH REVISED NSPS FOR UTILITY BOILERS 324
Randy M. Cole
Energy Research-Combustion Systems
Tennessee Valley Authority
THE ECONOMICS OF BENEFICIATING AND MARKETING
HIGH-SULFUR IOWA COAL 360
C. Phillip Baumel, John J. Miller and Thomas P. Drinka
Department of Economics, Iowa State University
Session 2; Physical Coal Cleaning Technology
AN EVALUATION OF THE DESULFURIZATION POTENTIAL OF U.S. COALS 387
Jane H. McCreery and Frederick K. Goodman
Battelle's Columbus Laboratories
THE USE OF COAL CLEANING FOR COMPLYING WITH SO,
EMISSION REGULATIONS 416
Elton H. Hall and Gilbert E. Raines
Battelle's Columbus Laboratories and
Raines Consulting, Incorporated
vl
-------
TABLE OF CONTENTS
(Continued)
Page
STATISTICAL CORRELATIONS ON COAL DESULFURIZATION BY
CRUSHING AND SPECIFIC GRAVITY SEPARATION 448
Ralph E. Thomas
Battelle's Columbus Laboratories
DEWATERING AND DRYING OF FINE COAL: EQUIPMENT
PERFORMANCE AND COSTS 464
Donald H. Sargent, Bill H. Cheng, and G. Yeghyazarian Contos
Versar, Inc.
HOMER CITY COAL CLEANING DEMONSTRATION, TEST, AND
TECHNOLOGY EVALUATION PROGRAM . 488
James H. Tice
Pennsylvania Electric Company
COMPUTER CONTROL OF COAL PREPARATION PLANTS 503
Gerry Norton, George Hambleton, and Clive Longden
Norton-Hambleton Associates, Inc.
PHYSICAL AND PHYSIOCHEMICAL REMOVAL OF SULFUR FROM COAL ...... 519
David H. Birlingmair and Ray W. Fisher
Ames Laboratory, Iowa State University
CLEANING OF EASTERN BITUMINOUS COALS BY FINE GRINDING,
FROTH FLOTATION AND HIGH-GRADIENT MAGNETIC SEPARATION 534
W. L. Freyberger, J. W. Keck, D. W. Spottiswood,
N. D. Solem and Virginia L. Doane
Michigan Technological University
THE POTENTIAL OF MAGNETIC SEPARATION IN COAL CLEANING 568
Frederick V. Karlson, Kenneth L. Clifford, William W. Slaughter,
and Horst Huettenhaln
Bechtel Corporation, Electric Power Research Institute,
and Bechtel National, Inc.
TESTING OF COMMERCIAL COAL PREPARATION PLANTS
WITH A MOBILE LABORATORY 599
William Higgins and Thomas Plouf
Joy Manufacturing Company
CHEMICAL COMMINUTION—AN IMPROVED ROUTE TO CLEAN COAL ....... 600
Victor C. Quackenbush, Robert R. Maddocks, and George W. Higginson
Catalytic, Inc.
COAL CLEANING BY THE OTISCA PROCESS 623
C. D. Smith
Otisca Industries, Ltd.
vii
-------
TABLE OF CONTENTS
(Continued)
Page
Session 3; Environmental Assessment and
Pollution Control Technology
THE COAL CLEANING PROGRAM OF THE FUEL
PROCESS BRANCH OF EPA'8 IERL-RTP 637
T. K. Janes
U.S* Environmental Protection Agency
ENVIRONMENTAL ASSESSMENT METHODOLOGIES FOR FOSSIL
ENERGY PROCESSES: AN UPDATE . 646
Robert P. Hangebrauck
U.S. Environmental Protection Agency
REVIEW OF REGULATIONS AND STANDARDS INFLUENCING COAL CLEANING ... 683
P. Van Voris, R. A. Ewing, and J. W. Harrison
Battelle's Columbus Laboratories and Research Triangle Institute
DEVELOPMENT OF ENVIRONMENTAL ASSESSMENT CRITERIA FOR COAL
CLEANING PROCESSES 711
R. A. Ewing, P. Van Voris, B. Cornaby, and G. E. Raines
Battelle's Columbus Laboratories and Raines Consulting, Incorporated
APPLICATION OF ENVIRONMENTAL ASSESSMENT METHODOLOGY TO HOMER
CITY POWER COMPLEX BACKGROUND DATA: COMPARISON WITH MEG VALUES . . 753
D. A. Tolle, D. P. Brown, R. Clark, D. Sharp,
J. M. Stllwell, and B. W. Vigon
Battelle's Columbus Laboratories
AN OVERVIEW OF CONTROL TECHNOLOGY 793
A. W. Lemuon, Jr., G. L. Robinson, and D. A. Sharp
Battelle's Columbus Laboratories
CHARACTERIZATION OF PREPARATION PLANT WASTEWATERS . . 824
K. B. Randolph, L. B. Kay, and R. C. Smith, Jr.
Versar, Inc.
CONTROL OF TRACE ELEMENT LEACHING FROM COAL PREPARATION WASTES ... 856
E. M. Wewerka, J. M. Williams, P. Wagner,
L. E. Wangen, and J. P. Bertino
Los Alamos Scientific Laboratory
STABILIZATION OF COAL PREPARATION PLANT SLUDGES 875
David C. Hoffman
Dravo Lime Company
viii
-------
TABLE OF CONTENTS
(Continued)
Page
CHEMICAL AND BIOLOGICAL CHARACTERIZATION OF LEACHATE
FROM COAL CLEANING WASTES 898
R. M. Schuller, R. A. Griffin, and J. J. Suloway
Illinois State Geological Survey and Illinois State
Natural History Survey
Session 4; Chemical Coal Cleaning
INTRODUCTION TO CHEMICAL COAL CLEANING 923
R. A. Meyers
TRW, Inc.
CURRENT STATUS OF CHEMICAL COAL CLEANING PROCESSES -
AN OVERVIEW 934
L. C. McCandless and Mrs. G. Y. Contos
Versar, Inc.
STATUS OF THE REACTOR TEST PROJECT FOR CHEMICAL
REMOVAL OF PYRITIC SULFUR FROM COAL 960
M. J. Santy and L. J. Van Nice
TRW, Inc.
STATUS OF HYDROTHERMAL PROCESSING FOR CHEMICAL
DESULFURIZATION OF COAL 991
E. P. Stambaugh, H. N. Conkle, J. F. Miller, E. J. Mezey and B. C. Kim
Battelle's Columbus Laboratories
SURVEY OF COALS TREATED BY OXYDESULFURIZATION . . 1016
R. P. Warzinski, J. A. Ruether, S. Friedman, and F. W. Staffgen
U.S. Department of Energy, Pittsburgh Energy Technology Center
COAL DESULFURIZATION BY LEACHING WITH ALKALINE SOLUTIONS
CONTAINING OXYGEN 1039
R. Markuszewski, K. C. Chuang, and T. D. Wheelock
Ames Laboratory, Iowa State University
THE POTENTIAL FOR CHEMICAL COAL CLEANING: RESERVES,
TECHNOLOGY, AND ECONOMICS 1064
R. A. Giberti, R. S. Opalanko, and J. R. Sinek
Kennecott Copper Corporation and Resource Engineering, Inc.
JPL COAL DESULFURIZATION PROCESS BY
LOW TEMPERATURE CHLORINOLYSIS • • 1096
John J. Kalvinskas and George C. Hsu
California Institute of Technology
OXIDATIVE COAL DESULFURIZATION USING NITROGEN
OXIDES - THE KVB PROCESS 1141
E. D. Guth
KVB, Inc.
ix
-------
TABLE OF CONTENTS
(Continued)
Page
THE DRY REMOVAL OF PYRITE AND ASH FROM COAL BY THE
MAGNEX PROCESS COAL PROPERTIES AND PROCESS VARIABLES 1165
James K. Kindig and Duane N. Goens
Hazen Research, Inc.
PANEL DISCUSSION ON PROSPECTS FOR CHARACTERIZATION
AND REMOVAL OR ORGANIC SULFUR FROM COAL 1197
Chairman: Robin R. Oder; Panelists: Sidney Freidman,
Amir Attar, Douglas M. Jewell, and Thomas G. Squires
Gulf Research and Development Co.; Pittsburgh Energy Research
Center - DOE; University of Houston; Gulf Research and
Development Co.; and Iowa State University
List of Participants 1208
-------
4>EPA Symposium on
Coal Cleaning to
Achieve Energy and
Environmental Goals
FINAL PROGRAM
-------
f/EPA Symposium on
Coal Cleaning to
Achieve Energy and
Environmental Goals
Symposium Chairman
James D. Kilgroe
U.S. EPA, IERL-RTP
Symposium Cochairman
Alexis W. Lemmon, Jr.
Battelle's Columbus Laboratories
Monday, September 11.1978
1:00—Registration—Mezzanine Lounge
Session 0: Coal Characteristics
2-6 p.m.—Regency West
Chairman: David A. Kirchgessner
U.S. EPA
Industrial Environmental Research Laboratory (IERL-RTP)
Cochairman: Harold L. Lovell
Pennsylvania State University
2:00 Petrography of Coal
Ron W. Stanton
U.S. Geological Survey
2:20 Mineralogic Affinities of Trace Element* in Coal
Faith Fiene
Illinois State Geological Survey
2:60 Effects of Coal Cleaning on Elemental Distributions
Charles T, Ford and James F. Soyer, Jr.
Bituminous Coal Research
3:20 Coffee Break
3:30 Particle Site Distribution in Liberation of Pyrite
Harold L. Lovell
Pennsylvania State University
4:00 Contaminants In Coal: Geology
and Size-Gravity Separations
C. Blaine Cecil
U.S. Geological Survey
4:30 Interpreting Statistical Variability
Ralph E. Thomas
Battelle's Columbus Laboratories
7-9 Welcome Reception-Regency North
Tuesday, September 12, 1978
8:00 a.m.—Registration—Mezzanine Lounge
Session 1: Coal Cleaning Overview
Morning Program—9 a.m.-12 Noon
Chairman: James D. Kilgroe
U.S. EPA. IERL-RTP
9:00 Welcome
Norbert Jaworski, Deputy Director
U.S.EPA, IERL-RTP
9:15 Introductory Remarks
James D. Kilgroe, Symposium Chairman
U.S. EPA, IERL-RTP
9:30 Overview of EPA Coal Cleaning Programs
David A. Kirchgessner and James D. Kilgroe
U.S, EPA, IERL-RTP
10:00 Overview of DOE Coal Cleaning Programs
Cyril W. Draffm
Department of Energy (DOE)
10:30 CoHee Break
10:45 Overview of EPRI Coal Cleaning Programs
Kenneth Clifford and Shelton Ehrlich
Electric Power Research Institute
11:15 An Integrated Assessment of Coal Technologies
Roger Hansen
U.S. EPA, IERL-RTP
Richard Davidson
Battelle's Columbus Laboratories
12:00 Luncheon—Les Ambassadeurs Room
Noon impacts of the 1977 Clean Air Act Amendment
Luncheon Speaker: Frank Princiotta
Office of Energy, Minerals and Industry
U.S. Environmental Protection Agency
-------
Afternoon Program — 2-5 p.m.
Chairman: Kenneth Clifford
Electric Power Research Institute
2:00 Environmental Protection Against the Influence of
Coal-Preparation Planta in the USSR
I. S. Blagov, G. G. Vosnyuk, V. V. Kochetov,
I. Ch. Nekhoroshy, and I. E. Cherevko
USSR, Ministry of Coal
2:30 Clean Fuel Supply Requirements lor the
OECD Countries
Gary J. Foley
Organization of Economic Cooperation and Development
Richard Livingston
U.S. Environmental Protection Agency
3:00 A Technical and Economic Overview of Coal Cleaning
Horst Huettenhain Samuel Wong
Bechtel Corporation Argonne National Laboratory
3:30 Coffee Break
3:46 An Evaluation of Institutional, Economic. Regulatory and
Legislative Barriers to Investment in Physical Coal Cleaning as
an SOj Emission Control Strategy
Karel Fisher and Peter M. Cukor
Teknekron, Inc.
4:16 Economics of Coal Cleaning and Flue Qas Desulfurization for
Compliance with Revised NSPS for Utility Boiler
R. M. Cole
Tennessee Valley Authority
4:46 Impact of Transportation and Beneficiation on the Utilization
of High Sulfur Coal
C. Phillip Baumel, Thomas P. Drinka and John J. Miller
Ames Laboratory, Iowa State University
Wednesday, September 13, 1978
Session 2: Physical Coal Cleaning
Technology
Morning Program — 9 a.m.-12 Noon
Chairman: Richard E. Hucko
Coal Preparation and Analysis Laboratory
U.S. Department of Energy
9:00 An Evaluation of the Desulfurization Potential of U.S. Coals
Jane H. McCreery and Fredrick K. Goodman
Battelle's Columbus Laboratories
9:30 The Use of Coal Cleaning for Complying with S02 Emission
Regulations
Elton Hall Gilbert E. Raines
Battelle's Columbus Laboratories Resource Dynamics, Inc.
10:00 Statistical Correlations on Coal Desulfurization by Crushing
and Specific Gravity Separation
Ralph E. Thomas
Battelle's Columbus Laboratories
10:30 Coffee Break
10:46 Dewatering and Drying of Fine Coal: Performance and Costs
Donald Sargent and William Cheng
Versar Inc.
11:16 Homer City Coal Cleaning Demonstration.
Test, and Technology Evaluation Program
James H. Tice
Pennsylvania Electric Company
11:46 Computer Control of Coal Preparation
Gerry Norton, Clive Longden, and George Hambleton
Norton, Hambleton Associates
Afternoon Program — 2-5 p.m.
Chairman: Kenneth Harrison
Heyl & Patterson, Inc.
2:00 Physical and Physicochemical Removal of Sulfur from Coal
Ray W. Fisher and David Birlingmair
Ames Laboratory
Iowa State University
2:30 Cleaning of Eastern Bituminous Coals by Fine Grinding,
Froth Flotation and High Gradient Magnetic Separation
W. L Freyberger. J. W. Keck. D. W. Spottiswood,
N. D. Solem and Virginia L. Doane
Michigan Technological University
3:00 The Potential of Magnetic Separation in
Coal Preparation
Frederick V. Karlson, Horst Huettenhain. William W. Slaughter
Bechtel National Inc.
Kenneth L. Clifford
EPRI
3:30 Coffee Break
3:46 Testing of Commercial Coal Preparation Plants with a
Mobile Laboratory
William Higgins and Thomas Plouf
Joy Manufacturing Co.
4:16 Chemical Comminution—an Improved Route to Clean
Coal
V. C. Quackenbush, R. R. Maddocks, and G. W. Higginson
Catalytic, Inc.
4:46 Coal Cleaning by the Otisca Proceas
Speaker to be announced
6-9 Social Hour and Banquet— Lea Ambassadeurs Room
p.m. Tomorrow's Energy Supplies
Banquet Speaker: Richard J. Anderson
Formerly Associate Director, Battelle's
Energy Program; Currently Consultant to
Battelle Memorial Institute
Thursday, September 14,1978
Session 3: Environmental
Assessment and Pollution
Control Technology
Morning Program—9 a.m.-12 Noon
Chairman: G Ray Smithson, Jr.
Battelle's Columbus Laboratories
9:00 Introduction to the EPA Program
T. K.Janes
EPA, IERL-RTP
9:20 Environmental Assessment Methodologies
for Fossil Energy Processes
R. P. Hangebrauck
EPA, IERL-RTP
J. L. Warren
Research Triangle Institute
-------
9:60 Review of Regulations and Standard*
Influencing Coal Cleaning
Peter Van Vons
Battelle's Columbus Laboratories
J. W. Harrison
Research Triangle Institute
10:20 Coffee Break
10:30 Environmental
Processes
Impact Assessment of Coal Cleaning
• Establishing Goals
Barney W. Cornaby
Battelle's Columbus Laboratories
• Overall Methodology
Robert A. Ewing
Battelle's Columbus Laboratories
• Biological Transport
Peter Van Voris
Battelle's Columbus Laboratories
• Physical Transport and Partition Functions
Gilbert E. Raines
Raines Consulting, Inc.
11:30 Methodology Application to Homer City Background
Data: Comparison with MEG Values
0, A, Tolle
Battelle's Columbus Laboratories
12:00 Luncheon—Les Ambassadeurs Room
Noon Luncheon Host: Edward W. Ungar, Director
Battelle's Columbus Laboratories
Afternoon Program—2 p.m.-5 p.m.
Chairman: C. Grua
U.S. Department of Energy
2:00 An Overview of Control Technology
Alexis W. Lemmon, Jr., Gerald L. Robinson,
and David Sharp
Battelle's Columbus Laboratories
2:30 Effluents from Coal Preparation
K. Randolph
Versar, Inc.
3:00 Control of Trace Element Leaching from Coal Prepare-
tion Wastes
E M Wewerka
Los Alamos Scientific Laboratory
3:20 Coffee Break
3:30 Stabilization of Coal Preparation Plant Sludges
David Hoffman
Dravo Lime Company
4:00 Chemical and Biological Characterization of Leachate
from Coal Cleaning Wastes
R. A. Griffin, et al,
Illinois State Geological Survey
Friday, September 16.1978
Session 4: Chemical Coal Cleaning
Morning Program — 9 a.m.-12 Noon
Chairman: Thomas D. Wheelock
Iowa State University
0:00 Introduction to Chemical Cleaning
R. A. Meyers
TRW Inc.
9:20 Current Status of Chemical Coal Cleaning Processes-
An Overview
Lee C. McCandless and G. Y. Contos
Versar. Inc.
9:50 Status of the Reactor Teat Project for Chemical
Removal of Pyritlc Sulfur from Coal
L. J. Van Nice and M. J. Santy
TRW Inc.
E. Bobalek and L. 0. Tamny
U.S. EPA. IERL-RTP
10:20CorYMSr*a*
10:30 Status of Hydrothermal Processing for Chemicel DesuHuriie-
tlon of Coal
E. P. Stambaugh, J. f. Miller. H. N. Cookie. B. C. Kim. and
E. J. Mezey
Battelle's Columbus Laboratories
11:00 Survey of Coals Treated by OxydoeuH uritetlon
R. P. Warzinski, S. Friedman, and F. W. Steffgen
Pittsburgh Energy Research Center—DOE
11:30 Coal Desutfurizatlon by Leaching with Alkaline
Solution Containing Oxygen
Richard Markuszewski, K. C. Chuang, and Thomas D. Wheelock
Ames Laboratory, Iowa State University
Afternoon Program — 2-5 p.m.
Chairman: Robin R. Oder
Gulf Research and Development Co.
2:00 Potential for Chemical Coal Cleaning: Reserves, Technology.
and Economica
R. A. Giberti. R. S. Opalanko. and Joachim R. Sinek
Kennecott Copper Corp.
2:20 JPL Coal Dosumirization Process by Low Temperature
Chlorinolysis
John J. Kalvinskas and George Hsu
Jet Propulaion Laboratory
2:40 OxWative Coal Desulfurization Uaing Nitrogen OxMae-
the KVB Process
E. D. Guth
KVB. Inc.
3:00 Coffee Break
3:20 The Dry Removal of Pyrite and Ash from Coal by the Magnex
Process-Process Variablea and Clean Coal Properties
James K. Kindig and Duane N. Goens
Hazen Research, Inc.
3:40 Panel Discussion on Prospects for Characterization and
Remove! of Organic Sulfur from Coal
Chairman: Robin R. Oder
Gulf Research and Development Co.
Panelists: Sidney Friedman
Pittsburgh Energy Research Canter—DOE
Amir After
University of Houston
Douglas M. Jewell
Gulf Research and Development Co.
Thomas G. Squires
Iowa State University
-------
CONFERENCE REPORT AND ACTIVITIES
The Symposium on Coal Cleaning to Achieve Energy and Environmental Goals
provided a major forum for technical interchange among engineers and scien-
tists concerned with the development and use of coal cleaning technology.
The conference, sponsored by the U.S. Environmental Protection Agency's
Industrial Environmental Research Laboratory of Research Triangle Park,
North Carolina, was held in Hollywood, Florida, on September 11-15, 1978.
Approximately 225 engineers, environmental scientists, geologists, and
managers from the coal industry, R&D organizations, coal users, planning
agencies, and government attended the five-day conference at which' papers
were given on coal characteristics, coal cleaning overview, physical coal
cleaning technology, environmental assessment and pollution control tech-
nology, and chemical coal cleaning.
Mr. Frank Princiotta, Director of the Energy Processes Division, Office
of Energy Minerals and Industry, U.S. EPA, addressed the first Symposium
luncheon on the "Impacts of the 1977 Clean Air Act Amendment". The audience
showed particular interest in his review of the status and substance of the
draft New Source Performance Standards, which were published that week.
The Symposium banquet was highlighted by the presentation, "Tomorrow's
Energy Supplies", by Mr. Richard J. Anderson, Consultant to Battelle Memorial
Institute, and a brief address by Mr. Gennadiy G. Voznyuk, Chief of Nature
Protection, Directorate, U.S.S.R. Ministry of Coal Industry of the Soviet
Union. The Soviet representatives to the conference were honored and seated
at the head table at the banquet. They included Mr. Viktor Kochetov, General
Director, Donetskugleobogashcheniye, U.S.S.R. Ministry of Coal Industry;
Mr. Ivan Nekhoroshiy, Chief of Laboratory of IOTT, U.S.S.R. Ministry of
Coal Industry; and Mr. Voznyuk. Simultaneous translation was provided
during all technical sessions as well as social functions of the Symposium.
-------
A second Symposium luncheon was hosted by Dr. Edward Ungar, Director of
Battelle's Columbus Laboratories (BCL), at which conference organizers, Mr.
James Kilgroe of EPA-RTP and Mr. Alexis W. Lemmon, Jr., of BCL, were
recognized.
Some technical highlights of the Symposium included the paper presented
by Mr. Nekhoroshiy of the Soviet delegation and several first time reports
from major ongoing coal cleaning research programs. "Primary Trends of
Works on Environmental Protection Against the Influence of Coal-Preparation
Plants in the U.S.S.R.", the subject of Mr. Nekhoroshiy's presentation, drew
much audience interest. K. Randolph of Versar, Inc., reported in his paper
on "Effluents from Coal Preparation" that proof has been obtained for the
existence of priority pollutants in effluents from coal cleaning. J. McCreery
of Battelle reported that the amount of low-sulfur coals which can be made
available in the United States to meet the 1.2 Ib S02/106 Btu NSPS is approxi-
mately 41 percent of total reserves as opposed to an earlier figure reported
in the literature of 14 percent. Her presentation was "An Evaluation of the
Desulfurization Potential of U.S. Coals".
The entire program on Thursday, September 14, provided a detailed
overview of the plans and progress of the environmental assessment of coal
cleaning. The methodological approaches shared will be of use to many
current and future coal cleaning developments. Perhaps the most useful result
of the program was the mutual opportunity to review and discuss the physical
and chemical coal cleaning programs of EPA, DoE, Electric Power Research
Institute, those of numerous industrial organizations, and European and
Soviet plans for the future, as well as the problems of ongoing operations.
-------
PETROGRAPHY OF COAL
Ronald W. Stanton and Robert B. Finkelman
U.S. Geological Survey
956 National Center
Reston, Virginia 22092
ABSTRACT
Coal is a sedimentary rock composed of microscopically recognizable
organic constituents (macerals) and inorganic constituents (minerals). By
American Society for Testing and Materials (ASTM) definition, pure coal
contains less than 25 percent ash by weight, and impure coal contains 25-50
percent ash.
Macerals are products of coalified remains of plants; they have been
divided into groups: vitrinite, exinite, and inertinite. Vitrinite macerals
represent partially decomposed woody-plant tissues. If cell structure is
retained, the variety of vitrinite is termed telinite; if cell structure is
not visible, the variety is termed collinite. The reflectance of vitrinite
may be used as a,measure of the degree of metamorphism of the coal. Exinite
macerals are remnant parts of plants such as spores (sporinite), cuticles
(cutinite), resins (resinite), and algae (alginite); they differ chemically
and morphologically from vitrinite. Inertinite macerals result from the
oxidation or alteration of other macerals and include semifusinite, fusinite,
micrinite, macrinlte, and sclerotinite.
Major minerals in coal include quartz, marcasite, pyrite, siderite,
calcite, and dolomite. The more common accessory minerals in coal are
rufcile (Ti02), sphalerite (ZnS), chalcopyrite (CuFeS2), zircon (ZrSiO^,
and monazite ((Ce,LA)PO,). These minerals are either allogenic (minerals
transported into peat-forming environments) or authigenic (minerals that
formed in the peat).
Maceral analysis and vitrinite reflectance can be used to predict coking,
gasification, and liquefaction potentials of coals and also can be used to
determine certain geologic conditions prevailing during peat accumulation
and coalification.
-------
LIST OF FIGURES
Figure 1.— ASTM definition of coal on the basis of ash content. (Am.
Soc. Testing Materials, 1977).
Figure 2.— Comparison of the composition between granite and coal.
Figure 3 - 8.—Photomicrographs of:
3.-- Vitrinite in (a) transmitted light and (b) incident light. V,
vitrinite.
4.— Exinite in (a) transmitted light and (b) incident light. C,
cutinite; S, sporinite.
5.— Sporinite (macrospore) in (a) transmitted light and (b) incident
light, blue irradiation.
6.— Alginite (A) in (a) incident light and (b) incident light,
blue irradiation.
7.— Inertinites in (a) transmitted light and (b) incident light. I,
inertinite; S, sporinite; v, vitrinite.
8.— Fusinite (F) in incident light showing open cell structure.
Figure 9.— Parts of peat-forming plants that form macerals.
Figure 10.—Photomicrograph of submicron crystals of pyrite (brightly
reflecting grains) in vitrinite.
Figure 11.—Photomicrograph of large pyrite grain showing replacement of cell-
wall material.
Figures 12 -16.—Scanning electron photomicrographs of:
12.—Mineral-rich zone in polished section of coal. Q, quartz; C, clay
mineral; R, rutile.
13.—Mineral-rich zone in polished section of coal. Q, quartz; P,
framboidal pyrite; Z, zircon; M, tnaceral.
14.—Sphalerite in organic material. S, sphalerite; M, maceral.
15.—A Mineral grain in coal. C, chalcopyrlte; CL, clausthallte? (PbSe);
S, sphalerite; M, maceral.
16.—Cell filling in fusinite. CR, crandalite; K, kaolinite.
Figure 17.—Diagrams showing the composition of a bituminous coal.
10
-------
INTRODUCTION
For some thousands of years, coal has been extracted from the earth and
burned. In most applications, upon combustion, this homogeneous-appearing
substance yields heat, smoke, and ash. Because of the Increased awareness of
environmental quality, the need has arisen for a coal product that behaves as
a clean fuel and is economically acceptable.
Effective coal cleaning (physical or chemical) depends on the basic
properties of coal, which is not a homogeneous substance but a complex mixture
of microscopic components, each having different chemical and physical
properties.
COMPOSITION OF COAL
Coal, in the geologic sense, is a sedimentary rock that is combustible.
By convention, coal is distinguished from shale by its ash content, which is
the residual product of coal upon combustion (fig. 1).
Coal is similar to other rocks (fig. 2) in that it contains minerals;
however, it differs because it is predominantly composed of macerals (organic
matter). Macerals have textural characteristics inherited from the original
plant material, have variable chemical compositions, are nonerystalllnei and
have distinct reflectance in polished sections (Stach and others, 1975).
Minerals, on the other hand, have defined chemical compositions and crystall-
ographic properties.
Macerals are sensitive to minor increases in metamorphism. The measured
reflectance of certain macerals indicates the degree of coal metamorphism
from lignite through anthracite (Stach, and others 1975). The molecular
structure of macerals is generally aliphatic but when metamorphism (coalification)
increases this structure gradually takes on a more aromatic character (Stach
and others, 1975). Some minerals, for example the clays, may be also affected
by changes in metamorphism.
11
-------
PURE COAL
IMPURE COAL
CARBONACEOUS SHALE
SHALE
25 50
% ASH
(BY WEIGHT)
100
Figure 1. — ASTM definition of coal on the basis of ash content (Am. Soc.
Testing Materials, 1977).
ROCK
Granite
COMPOSITION
100 % Minerals
erals
Coal
Carbonates
10 % Minerals
Mirieral
erals
Figure 2.
Accessory
Vitrinite
Exinite
• Inert inite
Comparison of the composition between granite and coal.
90 % Macerals
12
-------
MACERALS
The major groups of macerals are vitrlnlte, exlnlte, and inertinite, each
of which can be subdivided (Internat. Comm. Coal Petrology, 1963, 1971; Stach
and others, 1975). Table 1 lists the microscopic characteristics, probable
maceral origin and chemical composition, the approximate density, and the
relative abundance of each maceral group in most bituminous coals.
Figures 3-8 are photomicrographs of various macerals as seen in incident
and transmitted light (SOX oil immersion). Depending on sample preparation,
combinations of transmitted light, incident light, and incident light', blue -
excitation can be used to identify macerals.
Woody tissues, leaves, spores, and algal mats are among the probable
sources of material that, through alteration and subsequent compaction, produce
macerals. Figure 9 schematically illustrates the source of the precursors of
macerals and where they may originate in a given swamp.
MINERALS
Minerals occur in coal in a wide variety of forms, including concretions,
fracture and cleat fillings, partings, and lenses; other minerals are finely
disseminated in the macerals (Mackowsky, 1968). All these minerals plus
inorganic elements that are organically bound in coal have been referred to as
"mineral matter" (Rao and Gluskoter, 1973). Although the normative mineralogy
can be inferred from bulk chemical analysis of coal, the actual minerals must
be identified microscopically and their identification confirmed by X'-ray
diffraction. The major minerals found in most coals are as follows:
QUARTZ Si02 ILLITE KAl2(Si3Al)010(OH)2
CALCITE CaC03 KAOLINITE Si2Al 05(OH)4
SIDERITE FeC03 MONTMDRILLONITE Ca-Fe-Mg-Al - silicate
ANKERITE CaMgFe(C03)4 PYRITE FeS2
DOLOMITE CaMg(C03)4 MARCASITE FeS2
13
-------
MACERAL
GROUP
VITRINITE
EXIH1TE
IHERTINITE
MICROSCOPIC CHARACTOUSTICS OK
MACERAL VARIETIES
-GREY IN REFLECTED LIGHT; RED IN
TRANSMITTED LIGHT -
- COLLIN1TE - NO VISIBLE CELL
STRUCTURE
- TEUNITE - VISIBLE CELL STRUCTURE
- VHRODEIUNITE - SMALL FRAGMENTS
OF VITRINITE
—DARK GREY IN REFLECTED LIGHT; YELLOW
IN TRANSMITTED LIGHT;YELLOW TO ORANGE
IN BLUE OR ULTRAVIOLET LIGHT
(FLUORESCENCE)
- SPORINITE -OULTINE OF COMPACTED
SPORE
- CUTINITE - OUTLINE OF COMPACTED
CUTICLE
- RESINITE - DARK CELL AND VOID
FILLINGS
- ALGINITE - DARK GLOBULAR MASSES
- UPTODETRINITE - SMVLL FRAGMENTS
OF EXINITE
-WHITISH GREY TO WHITE IN REFLECTED
LIGHT; OPAQUE IN TRANSMITTED LIGHT
- FUSINITE - HIGHLY REFLECTING;
CELL STRUCTURE DISTINCT
- SEfOFUSINITE- REFLECTANCE BETWEEN
VITRINITE AND
FUSINITE; SOME CELL
STRUCTURE
- HACRINITE - NO CELL STRUCTURE;
REFLECTANCE HIGHER
THAN VITRINITE
- MICRINITE - SMALL GRANULAR GROUND-
MASS; HIGHLY REFLEC-
TING
- SCLEROTDUTE - ROUND BODIES WITH
CAVITIES; HIGHLY
REFLECTING
- INERTOOETRI1JITE - SMALL FRAGMENTS
OF INERTINITE
ORIGIN
HUMIC ACID FRACTION OF HUMIC
SUBSTANCES DERIVED THROUGH
MOULDERING AND PEATIFICATION
OF PLANT CELL WALLS COMPOSED OF
LIGNIN AND CELLULOSE
HYDROGEN - RICH PLANT PARTS SUCH
AS SPORINE, CUTINE, RESINS, WAXES
FATS AND OILS OF VEGETABLE MATTE
MOST VARITIES ARE VERY RESISTANT
TO COALIFICATION
OXIDATION OF VARIOUS PLANT
PARTS PRIMARILY CELL WALLS;
SCLEROTINITE MAY ORIGINATE FROM
FUNGAL REHAINS.
CENEKAL COMPOSITION
HUMINS WITH AROMATIC NUCLEUS
SURROUNDED BY PERIPHERAL
ALIPHATIC CROUPS
SPORINITE— ALIPHATIC-AROMATIC
SKELETON WITHOUT FATTY ACID
ANHYDRIDES, SOME STEROID ZONES
CUTINE— GLYCERINE ESTERS OF
FATTY ACIDS.
RESINS — HIGHER IN HYDROGEN
CONTENT
HIGH CARBON AND LOW HYDROGEN
CONTENTS; GREATER DEGREE OF
AROMATIZATION AND CONDENSATION
APPROXIMATE
DENSITY
(30% volatile)
1.2 C/CM3
I.I C/CM3
1.4 - 1.5 G/CM-
U'l*OXIMATh
HKLA7 1 VE
ABUNDANCE
(volume %)
70
0 - 7
10 -20
Table 1.—•- General characteristics of bituminous coal macerals (Information from Stach and others, 1975;
Internat. Comm. Coal Petrology.
-------
Figure 3. — Photomicrographs of vitrinite In (a) transmitted light and
(b) incident light. V, vitrinite; S, sporinite.
Figure 4. — Photomicrographs of exinite in (a) transmitted light and
(b) incident light. C, cutinite; S, sporinite.
15
-------
Figure 5. - Photomicrographs of sporinite (macrospore) in (a) transmitted
light and (b) incident light. S,sporinite; V, vitrinite.
Figure 6. - Photomicrographs of alginite (A)in (a) incident light and
(b) incident light, blue irradiation.
16
-------
Figure 7. - Photomicrographs of inertinites in (a) transmitted light
ani (b) incident light. I, inertinite; S, sporinite; v, vitrinite.
Figure 8, - Photomicrograph of fusinite (F) in incident light showing open
cell structure.
17
-------
PRE - CUT1NITE
(FDOH LEAVES)
PRE-VITRINITE
PRE-FUSINITE
PRE-SEHIFUSINITE
(KJCBGSCOP1C
VIEW OF
FROND CROSS
SKIIOH)
PRE-RtSlNITE
PRE-SPORINITE
PRE-ALGINITE
Figure 9.—Parts of Peat-formitig plants that form macerals
18
-------
Both the relative and absolute abundances of these major minerals differ
widely within and among coal beds. Examples of major minerals In polished
blocks of coal are shown In figures 10-12. The quartz shown in figure 12 is
a common form of the mineral in many coals, whereas the morphology, maceral
association, and particle size of pyrite can differ both vertically and laterally
»
within any given coal bed (fig. 10 and 11). Of the above-listed minerals,clay,
quartz, and pyrite are common in coal. In addition to the above minerals,
appreciable concentrations of the iron oxide and sulfate minerals may be present
in weathered (oxidized) coal; sulfate minerals may also occur in low-rank coal
such as lignite but are rare in fresh, higher rank coal.
About 100 accessory minerals some of which may control the distribution
of many trace elements (Finkelman, 1978) have been found in coal. Recent studies
strongly indicate that much of the Se, Pb, Cu, 2»n, and Cd in Appalachian coal
may be in the accessory minerals (Finkelman and others, 1979). The following
is a list of the common accessory minerals.
APATITE Ca (POA) (OH) HYPERSTHENE (Mg,Fe)2Si06
BARITE Baio4 * 3 ILMENtTE FeT103
BIOTITE K-Mg-Fe-Al silicate MAGNETITE FeoO^
CHALCOPYRITE CuFeS, MONAZITE CePO.
CHLORITE Fe-Al silicate MUSCOVITE KAl2(AlSi3)0,0(OH)2
CLAUSTHALITE PbSe PLAGIOCLASE (Na.Ca)(Al.Siy^Og
CRANDALLITE (Ca,Ba,Sr)Al,(PO,),(OH)c'H20 RUTILE Ti02
DIASPORE AlO(oH) * SPHALERITE ZnS
GALENA PbS SPHENE CaT105
GOETHITE Fe,Oo'H00 SYLVITE KC1
GOLD Au TALC Mg Si,Ol0(OH)2
GYPSUM CaSOAH 0 TOURMALINE complex boro-silicates
HALITE NaCl XENOTIME YPO.
HEMATITE Fe20 ZIRCON ZrSiO^
The study of minerals in coal was long hampered by the difficulty in
removing extremely fine-grained minerals from the organic constituents until
Gluskoter (1965) used an electronic low-temperature asher to oxidize the organic
fraction. By means of this instrument, the macerals were oxidized at relatively
low temperatures (^200°C), leaving a residue of virtually unaltered minerals.
19
-------
Figure 10. — Photomicrograph of submicron crystals of pyrlte
(brightly reflecting grains) in vltrinite.
Figure 11. —- Photomicrograph of large pyrite grain showing replacement of
cell wall material.
20
-------
'igure 12. — Scanning electron photomicrograph of mineral-rich zone
in polished section of coal. Q, quartz; C, clay mineral;
R, rutlie.
21
-------
Another technique that has shown great promise for the study of minerals in
unashed coal is the scanning electron microscope (SEM) equipped with an energy
dispersive X-ray detector, which can chemically analyze grains as small as
O.Sym (Dutcher and others 1973; Finkelman and Stanton, 1978).
Minerals in coal are 1) allogenic (transported into the peat swamp as wind
or waterborne detritus) and 2) authigenic (formed in place). Authigenic
minerals may precipitate from solution, result from reactions involving bacteria,
or result from the oxidation, reduction or leaching of existing minerals.
A systematic relationship between certain minerals and macerals exists in
many coals (Finkelman and others, 1976; Finkelman and Stanton, 1978). Dull
bands in coal may consist of maceral fragments, illite, and quartz and trace
amounts of rutile, zircon, rare-earth phosphates, and feldspar (fig. 13). Most
fusinite and semifusinite pores are generally filled with kaolinite, but some
are filled with carbonates and pyrite. The pores may also contain sphalerite
(fig. 14), chalcopyrite, galena, clausthalite (fig. 15), apatite, rare-earth
phosphates, and crandallite (fig. 16). Pyrite, kaolinite, micron-sized grains
of sphalerite, chalcopyrite, and clausthalite may be dispersed throughout
vitrinite. Barite is commonly associated with resin bodies and spores.
COAL PETROGRAPHY IN APPLIED RESEARCH
When considering the use of coal for any particular purpose, the concept
of coal as a rock must be accepted. Petrographic analysis of coal has been
used widely over the past two decades in predicting coke stability (Schapiro
and Gray, 1964; Thompson and others, 1966) and to evaluate gasification and
liquefaction potentials of coal (Given and others, 1975a).
In figure 17, several diagrams are shown for describing coal. The most
conventional approach is to define the coal in terms of ash, sulfur, and
combustibles on a weight-percent basis (fig. 17a and b). Through coal-
petrographic analysis, the pie can be divided further into maceral groups and
22
-------
Figure 13. — Scanning electron photomicrograph of mineral-rich zone In
polished section of coal. Q, quartz; P, framboidal pyrite;
Z, zircon; M, maceral.
Figure 14. -^ Scanning electron photomicrograph of Sphalerite in organic
material. S, sphalerite; M, maceral.
-------
Figure 15. — Scanning electron photomicrograph of a mineral grain In coal.
C, chalcopyrite; CL, clausthalite? (PbSe); S, sphalerite;
M, maceral.
Figure 16. — Scanning electron photomicrograph of a cell filling in fusinite.
CR, crandalite; K, kaolinite.
24
-------
a.
( X HEIGHT)
MINERAL
MATTER
EX1NITE-1 H.71
WHOLE COAL COMPOSITION GWOLUKE)
c.
b.
( Z WEIGHT)
( 1.51 TOTAL SULFUR)
MINERAL
MATTER
EXINITE
WHOLE COAL COMPOSITION (Z WEIGHT)
d.
Figure 17. — Diagrams showing the composition of a bituminous coal.
25
-------
mineral matter (fig. 17d). Most petrographic data is reported and used on a
volume-percent basis (fig. 17c). Weight percent is calculated by using
approximate maceral densities (Table 1, fig, 17d), which are not accurately
known for various ranks of coal. Further subdivision of the mineral matter
into specific minerals and maceral groups into maceral varieties can
petrographically characterize the coal to a greater detail.
Coal cleaning is yet another area of applied research that can benefit
from petrographic data. Such data can help evaluate and perhaps even predict
the behavior of coal in physical coal cleaning in the following ways:
- use of maceral analyses to predict the size-gravity concentrations of
the bulk of the combustible coal;
- use of microscopic mineral determinations (pyrite in particular) to
predict whether and how certain coal may clean;
- use of data on specific trace elements, for example, the zinc in
sphalerite, to determine whether such elements can be separated from
the combustible matter during precombustion treatment.
Further understanding and applied knowledge of mineral and maceral relations
in coal can help produce a fuel that can be burned and that will produce
minimum pollution to the environment.
26
-------
REFERENCES
American Society for Testing and Materials, 1977, Standard definitions of terms
used for megascopic description of coal and coal beds and for micro-
scopical description and analysis of coal, Ija 1977 Annual Book of ASTM
Standards: Philadelphia, Pa., Am. Soc. Testing and Materials, pt. 26,
p. 346-349.
Dutcher, R. R., White, E. W., Lebiedzik, J., and Hoover, M. R., 1973, Quantitative
SEM Image Analysis - application to coal and coke microscopy [abs.]: Geol.
Soc. America, Abstracts with Programs, v. 5, no. 2, p. 157-158.
Finkelman, R. B., 1978, Determination of trace element sites in the Waynesburg
coal by SEM analysis of accessory minerals, In Scanning Electron
Microscopy/1978/Vol. 1, (ed. 0. Johari), p. 143-1*8, 52.
Finkelman, R. B., and Stanton, R. W., 1978, Identification and significance of
accessory minerals from a bituminous coal: Fuel (in press).
Finkelman, R. B. Stanton, R. W., and Breger, I. A., 1976, Energy dispersive
X-ray analysis of in situ minerals In coal: Geol. Soc. America,
Abstracts with Programs, v. 8, no. 6, p. 865-866.
Finkelman, R. B., Stanton, R. W. , Cecil, C. B., and Mlnkln, J. A., 1979, Modes
of occurrence of selected trace elements in the Upper Freeport coal [abs.]:
Am. Ch-m. Soc. Fuel Chemistry D4v., (in press).
Given, P. H., Cronauer, D. C., Spackman, W., Lovell, H. L., Davis, A., Biswas, B.,
1975a, Dependence of coal liquefaction behavior on coal characteristics;
1. Vitrinite-rich samples: Fuel, v. 54, no. 1, p. 34-39.
, 1975b, Dependence of coal liquefaction behavior on coal characteristics;
2. Role of petrographic composition: Fuel, 54, no. 1, p. 40-49.
Gluskoter, H. J., 1965, Electronic low-temperature ashing of bituminous coal:
Fuel, v. 44, no. 4, p. 285-291.
> 1975, Mineral matter and trace elements in coal in Trace Elements in
Fuel (ed. S. P. Babu) , Advances In Chemistry Series 141, ACS, p. 1-22.
International Committee for Coal Petrology, 1963, International handbook of
coal petrology: 2d ed., Paris, Centre National Recherche Scientifique.
_, 1971, International handbook of coal petrology: Supplement to 2d ed.,
Paris, Centre National Recherche Scientifique.
Mackowsky, M. Th., 1968, Mineral matter in coal, In Coal and Coal-bearing Strata
(D. G. Murchison and T. S. Westall, eds.): Oliver and Boyd, Edinburgh
and London, p. 309-321.
Rao, C. P., Gluskoter, H. J., 1973, Occurrence and distribution of minerals in
Illinois coals: Illinois State Geological Survey Circ. 476, 56 p.
27
-------
Schapiro, N., and Gray, R. J., 1964, The use of coal petrography in coke
making: Jour. Inst. Fuel, v. 38, p. 234-242.
Stach, E. and others, 1975, Stach's textbook of coal petrology: Gebruder-
Borntraeger, Berlin, 428 p.
Thompson, R. R., Shigo, J. J., Benedict, R. P., and Aikman, R. P., 1966, The
use of coal petrography at Bethelem Steel Corporation: Blast Furnace
and Steel Plant Steel Publications, Inc., Osceola Mills, Pa., p. 817-
824.
28
-------
MINERALOGIC AFFINITIES OF TRACE ELEMENTS IN COAL
11 2
F. L. Fiene , J. K. Kuhn , and H. J. Gluskoter
Illinois State Geological Survey
Urbana, Illinois
2
Exxon Production Research Company
P. 0. Box 2189
Houston, Texas
ABSTRACT
Data obtained from analyses of washed coals may be used to estimate the
affinity of ah element for the organic and inorganic portions of the coal.
An organic affinity value is determined by normalizing the washability curves,
removing from them a component that represents the contribution from inseparable
mineral matter, and then calculating the area under the washability curve.
The resulting value can be related to the ease with which an element can be
removed from the coal by currently practical procedures for cleaning coal.
Elements can be categorized as organic, intermediate-organic, intermediate-
inorganic, and inorganic. Elements grouped as inorganic are those identified
in discrete mineral phases: Fe, As, Zn, Cd, and Pb as sulfides; Ca, Mg, Fe,
and Mn as carbonates; Al, Si, K, and Mg as clays and silicates; Ca and P
as a phosphate; and Ba as a sulfate. A considerable portion of these elements
may be removed by cleaning. A number of metals, including Co, Ni, Cu, Cr,
and Se have affinities in the intermediate categories, suggesting that they
are present In coals either ±n organic association or with finely disseminated
and occluded minerals within the coal. Beryllium, B, Ge, and U are present
in organic association, presumably as chelated metals, and are not removed
by normal cleaning procedures.
The amount of an element that is organically combined in a coal can be
estimated by extrapolating the washability curve to zero percent recovery on
an adjusted washability plot. Such values generally agree with those directly
determined from acid-demineralized products of whole coals. When differences
between those values occur, they can generally be attributed to the presence
of exchangeable ions and chelated metals in the coal.
29
-------
INTRODUCTION
A large portion of many potentially hazardous trace elements
contained in coal may be removed by physical cleaning; however,
in order to assess effectively the value of desulfurization and
beneficiation procedures, it is desirable to determine baseline
levels of coal contaminants and to understand their modes of
occurrence.
The type of association or combination in which an element
occurs in coal can significantly influence its reactivity in
combustion and conversion processes. Major portions of many trace
and minor elements are associated with the inorganic fraction of
the coal as discrete mineral phases. Some species may be present
largely or partly as Ion-exchangeable cations associated with
either the organics or clay minerals. To a certain extent, some
elements may also be chelated or associated with stable organic
complexes. The proportions of these components differ with
maturation and geochemical conditions of coal formation, and, as
a result, cleaning characteristics vary from one coal to another.
METHODS AND PROCEDURES
Analytical Procedures
Comprehensive trace element and mineralogical studies have
been conducted to evaluate the modes of occurrence, distribution,
and potential for physical removal of elements in four coals from
several coal-producing areas of the United States. The following
30
-------
samples were selected to Illustrate the wide variations in chemical
and mineralogical composition encountered:
1. A high-sulfur, high-volatile "C" bituminous
Illinois coal from the Herrin (No. 6) Coal Member
2. A high-sulfur, high-volatile "A" bituminous
West Virginia coal from the Pittsburgh (No. 8) seam
3. A low-sulfur, medium-volatile bituminous Alabama
coal from the Blue Creek seam
4. A low-sulfur, subbituminous Montana coal from
the Rosebud seam
Gravity separations were made on a 3/8 inch by 28 mesh size
fraction obtained by stage crushing the coal and screening. The
sized coal was separated into five or six specific-gravity fractions
ranging from 1.28 float to 1.60 sink in mixtures of perchloroethylene
and naphtha. Representative portions of the float-sink fractions,
raw coals, and 3/8 inch by 28 mesh material were stage ground
to -60 mesh for low-temperature ashing and chemical analysis
and to -100 mesh for trace element determinations. Representative
portions of the raw coal were further ground to pass a -325
mesh screen for use in the demineralization process.
Analytical determinations of over 70 major, minor, and trace
elements and other normal coal parameters were made on the whole
coals, 3/8 by 28 mesh fractions, float-sink fractions, and acid-
demineralized products from each coal. Procedures used to determine
the elemental concentrations are described in detail in Ruch et al.
(1974) and Gluskoter et al. (1977). All of the usual coal
parameters were determined using the standard methods outlined in
the ASTM Book of Standards, v61. 26 (1977). Trace element values
are usually obtained using up to five different analytical methods
on the same sample to ensure a high degree of accuracy.
31
-------
Detailed mineralogic studies of single samples of the four
raw coals and of the various specific gravity fractions obtained
from them were conducted in conjunction with chemical analysis
to identify the mineral phases present and to correlate them with
the elemental content of the coals. The samples were characterized
by X-ray diffraction analysis and microscopic examination of low-
temperature ash residues prepared from the coals. The original
minerals contained in the coal are retained by this radio-frequency
.plasma ashing technique, and because temperatures are sufficiently
low (<150°C), the mineral phases are not significantly altered by
oxidation, dehydration, or decomposition (Gluskoter, 1965). Semi-
quantitative mineralogic analysis of the major nonclay minerals
using an internal standard and prepared calibration curves was
carried out by methods adapted from Ward (1977). Mineral phases
in quantities of less than one percent were generally not detectable
above background intensities. The total clay percentage was
obtained by difference. Clay analysis of the <2ym fraction was
conducted using the preparation and analysis methods of Stepusin
(1978).
Determination of Elemental Affinities
Whether an element is concentrated in the organic or
mineral matter portions of coal was determined from washability
curves, which are cumulative curves based on analytical data from
float-sink fractions of the coal. A washability curve with a
positive slope, such as that for arsenic in figure la or for copper
in figure 2a, indicates that the element is concentrated in the
inorganic portion of the coal. The steeper the slope, the more
strongly associated the element Is with the mineral matter and
32
-------
the more readily the element can be removed from the coal with
fairly high recovery. An element uniformly distributed in the
various gravity fractions of the coal has a relatively flat
washability curve, such as that for vanadium in figure 3a.
Washing such a coal would have little effect on the concentration
of the element .
Elements concentrated in the lightest specific gravity
fraction may not necessarily be organically associated, however.
The hypothetical elemental concentrations of the "cleaned coals"
represented by washability curves extrapolated to zero percent
recovery are often high because of the possible presence of
finely disseminated mineral matter encapsulated within coal
particles. By adjusting the curve for this component (P) using
a method developed by Qluskoter et al., (1977) and described
here, a more meaningful estimation of the element's concentration
in the organic material can be obtained. The adjusted washability
curve is constructed after determining the value for F, as in the
following example for Zn, and subtracting it from each data point
that was used to construct the washability curve:
where LTA%,-. . .* is the percentage of low-temperature ash in
v-Lignt; the llghtest float fraction (in this case,
6.10 percent),
LTA%,~ (•-. •, is the percentage of low-temperature ash in
u.ou s) the 1.60 sink fraction (in this example,
77-80 percent), and
Zn(l 60 s) is tne elemental concentration in the 1.60
^ sink fraction. This sample contained 250 ppm
Zn in the 1.60 sink fraction.
33
-------
If the value of a datum point was negative after f was subtracted
from the reported concentration, the value for that point was
taken to be zero.
The area under the curve on the adjusted washability diagram
is defined as the organic affinity of the element. This provides
a mechanism for quantifying the data and suggests the ease with
which the element can be removed from coal by gravity coal cleaning
procedures. The organic affinity index was determined by calculating
the area beneath a curve that had been normalized to a predetermined
and constant scale and from which the component F representing the
inseparable mineral matter had been subtracted. The entire
normalized area of the graph is defined to be 1.00. An element
that can be removed to any dgree by washing the coal has an
organic affinity index less than 1.00. Extremely low values indicate
that the element is present almost entirely in the mineral matter.
The higher the organic affinity index, the more probability the
specific element has for occurrence in organic association with
the coal and the less effect washing will have on the reduction
of that element in the cleaned coal. Examples of both standard
and adjusted washability curves and their organic affinity indexes
for the Herrin (No. 6) Coal are given in figures l, 2, and 3 for
comparison.
The affinity of an element for either the organic fraction
of the coal or the mineral matter can be confirmed by utilizing a
combination of physical and chemical methods to obtain an almost
entirely mineral-matter-free organic fraction for direct analysis.
To accomplish this, mineral matter was removed from cleaned coal
by means of selective acid dissolution while the coal organic
34
-------
50
i38
1
225
00
(A)
ARSENIC
1
32
£24
S 1.6-
08-
20
40 60
Percent recovery
Herrin (No. 6} Cool
80
(B)
ARSENIC
Organic affinity 0.04
~20 40 60 80
Percent recovery-adjusted
Herrin (No 6) Coal
100
Figure 1—Arsenic in specific gravity fractions of a sample from the
Herrin (No. 6) Coal Member from Illinois. (A) standard
washability curve. (B) adjusted washability curve.
i
I :
L48
.
(A)
COPPER
20 40 60 80
Percent recovery
Herrin (No 6) Coal
100
COPPER
Organic affinity « 0.66
20 40 60 80 100
Percent recovery-adjusted
Herrin (No 6) Coal
'igure 2—Copper in specific gravity fractions of a sample from the
Herrin (No. 6) Coal Member from Illinois. (A) standard
washability curve. (B) adjusted washability curve.
38.6
309
S23.I'
E
3
f 154
5
771-
0.0;
(A)
VANADIUM
20 40 60 80 10
1
(B)
oo
Percent recovery
Herrin (No 6) Coal
20 40 60 80
Percent recovery-adjusted
Herrin (No. 6) Coal
100
Figure 3—Vanadium in specific gravity fractions' of a sample from the
Herrin (No. 6) Coal Member from Illinois. (A) standard
washability curve. (B) adjusted washability curve.
35
-------
fraction was maintained in a relatively unaltered state. The
demineralization procedure is summarized as follows:
1. Raw coal floated at 1.40 specific gravity.
2. 2-hr reflux with 10 percent HNOV
3. 2-hr digestion with 48 percent HP at 70 C.
4. 1-hr digestion with 25 percent HC1 at 70 C.
5. Thorough washing with distilled water and vacuum
drying.
This procedure may oxidize some of the organic matter; however,
any major effect should be indicated by a reduction in the organic
sulfur content, which in these samples is not apparent.
In order to further differentiate the way elements are
held in the coal, ion-exchange studies were performed on the
whole coals. Coals used for exchangeable ion studies were reduced
to -325 mesh, and a 10-gram increment was placed in a 300-ml
polyethylene flask. Fifty ml of ammonium acetate (IN) was added
to the flask, and the mixture was then stirred at approximately
70°C for 20 hours. At the end of the dissolution period, the
material was vacuum filtered while being flushed with sufficient
NHjjAC solution to bring the volume up to 450 ml. A final flush
with 50 ml of ethyl alcohol was performed and the sample was
vacuum dried. Elemental determinations were then made on both
the residual coal material and on the extract.
RESULTS
The compilations of data'in tables 1, 2, 3, and 4 show the
calculated organic affinity values of specific elements in the
four coals studied. Also included are the elemental concentrations
of raw coals as well as their organic fractions as estimated from
adjusted washabllity curves (P/S EXT) and from direct analysis of
acid demineralized (MMF) coals by the previously described
36
-------
TABLE 1
Elemental concentrations and organic affinity of elements
in the Herrin (No. 6) Coal Member from Illinois (C18560)
Organic fractions
Element
Al
Ca
Fe
K
Mg
Na
Ti
Si
LTA
HTA
organic S
pyritic S
sulfate S
total S
Ag
As
B
Ba
Be
Br
Cd
Ce
Co
Cr
Cs
Cu
Dy
Eu
Ga
Ge
Hf
Hg
I
La
Lu
Mn
Ni
P
Pb
Rb
Sb
Sc
Se
Sra
Sn
Sr
Ta
Th
U
V
W
Yb
Zn
Zr
Organic
affinity
.30
.06
.06
.56
.27
.64
.29
.45
.16
.16
1.11
.05
.17
.45
.54
.04
.77
.15
.87
.92
.07
.07
.74
.77
.44
.66
.89
.67
.15
1.76
.48
1.02
_
.04
.59
.06
.75
.03
.32
.45
.90
.57
.28
.39
.16
.07
.44
.55
1.29
.97
_
.52
.04
.49
Raw
%
1.40
0.51
2.60
0.13
0.06
0.04
0.06
3.20
20.37
16.46
1.87
4.56
0.02
6.45
coal
ppm
0.03
3.4
200
54
1.4
13.4
<0.10
25
7:2
21
2.0
13
1.2
0.26
2.4
14
1.1
0.23
1.2
6.1
0.1
60
24
50
<1.0
23
0.49
4.1
4.3
0.86
0.40
33
0.25
3.6
1.9'
36
0.59
0.84
57
32
F/S Ext*
% ppm
0.10
0
0
0.04
0.003
0.01
0.13
0.017
0
0
2.33
0
0
1.1
0.11
0
57
2
0.64
12
0
0
1.7
20
0.25
3.3
0.75
0.11
0.30
1.7
0.10
0.17
—
0
0.03
0
5.9
0
0.32
0.3
0.37
0.83
0
0.21
0
0.8
0.05
0.8
2.7
35
—
0.18
0
9
MMFb
% ppm
41
25
66
<1
21
6
41
20
—
—
1.81
—
—
1.81
__
<.7
6.6
0.2
0.03
3.3
<0.1
0.1
0.36
6.8
0.1
2.1
0.5
0.1
0.73
—
0.11
—
<0.8
0.72
0.03
0.3
<1
<1
<1
<1
0.09
0.65
0.26
0.41
<0.10
1.5
0.09
1.0
0.09
3.5
0.06
0.23
1
.
Float-sink extrapolation
Acid-demineralized coal
37
-------
TABLE 2
Elemental concentration and organic affinity of elements
in the Pittsburgh (No. 8) seam from West Virginia (C19824)
Organic Fractions
Element
Al
Ca
Fe
K
Mg
Na
Si
Ti
HTA
LTA
organic S
pyritic S
sulfate S
total S
As
B
Ba
Be
Br
Cd
Ce
Co
Cr
Ca
Cu
Dy
Eu
Ga
Ge
Hf
Hg
I
In
La
Lu
Mn
Mo
Ni
P
Pb
Rb
Sb
Sc
Se
Sm
Sn
Sr
Ta
Tb
Th
Tl
U
V
W
Yb
Zn
Zr
Organic
affinity
.62
.04
.17
.10
.04
,71
.39
.58
.15
.13
1.15
.18
.12
.81
.11
1.14
.90
.77
1.02
.09
.68
.79
.58
.28
.49
.67
.67
.79
.41
.40
.47
.67
—
.68
.62
.06
.04
.62
.68
.04
.18
.37
.67
.53
.72
.08
.94
.51
.90
.62
.06
.74
.57
.67
.74
.31
.27
Raw coal
X ppm
1.02
1.61
1.12
0.102
0.16
0.068
1.95
0.06
12.87
13.09
1.10
1.08
0.05
2.23
3.9
82
130
0.45
12
0.24
.16
2.2
14
0.76
8.6
0.82
0.18
2.6
0.80
1.0
0.13
0.3
0.02
5.7
0.08
35
1.66
9.0
103
25
9.5
1.6
2.3
1.6
0.94
7.8
143
0.17
0.13
2.1
0.54
0.60
17
0.26
0.30
10.3
30
F/S Exta
Z ppm
0.43
0
0
0
0
0.036
0.33
0.023
0
0
1.67
0
0
1.65
0
86
100
0.26
12
0
6.2
1.5
5.0
0.07
2.1
0.41
0.11
2.1
0.29
0.15
0.03
0.14
—
3.0
0.03
0
0
2.6
42
0
0
0.17
1.1
0.7
0.52
0.02
0.11.
0.04
0.11
0.66
0
0.74
6.6
0.15
0.23
0.85
1.2
MMFb
X ppm
41
30
80
2.5
<20
8.8
40
11
—
—
—
—
—
—
0.09
_
27
0.07
12
<.05
2.5
0.25
2.0
0.03
1.5
0.48
0.06
1.4
0.27
0.6
0.03
0.14
<.l
2.6
0.02
0.68
<0.2
1.2
<5
<1
<1
0.85
0.08
0.5
0.47
1.0
24
0.08
0.04
1.1
<0.1
0.11
2.7
<0.9
0.13
<1
<1
a Float-sink extrapolation
Acld-demineralized coal
38
-------
TABLE 3
Elemental concentrations and organic affinity of elements
in the Blu'e Creek seam from Alabama (C18848)
Element
Al
Ca
Fe
K
Mg
Na
Si
Ti
HTA
LTA
organic S
pyritic S
sulfate S
total S
Ag
As
B
Ba
Be
Br
Cd
Ce
Co
Cr
Cs
Cu
Dy
Eu
Ga
Ge
Hf
Hg
I
In
La
Lu
Mn
Ni
P
Pb
Rb
Sb
Sc
Se
Sm
Sn
Sr
Ta
Tb
Th
U
V
W
Yb
Zn
Zr
Organic
affinity
.40
.34
.44
.12
.07
.20
.17
.54
.17
.19
1.08
.63
1.05
1.08
.73
.05
.37
.62
.76
1.20
.45
.64
1.08
.60
.10
.78
.78
.78
.64
1.10
.44
1.02
—
_
.74
.69
.05
1.01
.60
.68
.10
.64
.53
.58
.66
.68
.80
.34
.66
.43
.71
.75
.70
.56
.21
.60
Raw coal
% ppm
'l.90
0.35
0.70
0.28
0.05
0.030
2.80
0.15
11.65
12.67
0.50
0.04
0.01
0.55
0.01
1.8
15
230
0.68
2.5
<0.10
30
9.4
21
2.3
12
2.1
0.44
6.3
0.60
1.2
0.39
1.3
0.32
18
0.13
13
11
190
12
18
0.82
4.3
3.0
2.8
0.50
130
1.1
0.22
5.4
0.92
54
0.36
0.92
2.0
57
Organic
F/S Exta
X ppm
0.25
0.037
0.14
0
0
0
0
0.04
0
0
0.53
0.02
0.012
0.56
0.007
0
0.76
76
0.36
2.5
0.05
14
7.9
7.1
0
8.0
1.5
0.20
2.6
0.18
0.33
0.04
—
—
9.6
0.04
0
9.9
90
1.9
0
0.23
1.2
0.9
1.0
0,19
54
0.04
0.20
0.61
0.77
29
0.30
0.18
0
21
fractions
MMFb
% ppm
240
48
54
2.3
<20
<3
64
28
_
_
0.33
—
__
0.36
_,
<0.1
5.1
20
0.05
1.7
<0.1
3.5
10
14
0.05
4.1
0.90
0.13
0.06
—
0.33
—
0.5
<0.1
2.8
0.07
<1
1"
<4
<1
<1
0.73
2.5
0.56
0.06
<0.2
40
<0.07
0.29
0.02
0.3
<5
0.1
0.60
-------
TABLE 4
Elemental concentrations and organic affinity of elements
in the Rosebud seam from Montana
-------
procedures. Comparison of concentrations between the F/S EXT
and MMF shows that the majority are in close agreement and are
within the precision of the analytical methods.
-------
The following summarizes the mineral phases detected in
low-temperature ashes from the four coals studied:
1. Herrin (No. 6) Coal Member, Illinois—kaolinite, lllite,
mixed-layer clay, pyrite, quartz, calcite; traces of
apatite, marcasite, szomolnokite, gypsum.
2. Pittsburgh (No. 8) coal, West Virginia—kaolinite, illite,
mixed-layer clay, pyrite, quartz, calcitej traces of
dolomite, orthoclase, szomolnokite, sphalerite, rutile,
galena.
3. Blue Creek seam, Alabama—major kaolinite; muscovite, quartz,
siderite, calcite; traces of pyrite, apatite.
4. Rosebud seam, Montana—kaolinite, illite, mixed-layer
clay, quartz, pyrite, calcite, bassanite; traces of barite,
chlorite.
The results of semi-quantitative mineralogic analysis are
presented in table 5- Percentages of pyrite, quartz, calcite,
total clays, and other minor minerals present In the low-temperature
ashes are given for the whole coals and their various specific
gravity fractions. Results of additional clay analysis of the
West Virginia and Montana coals are given in table 6. Due to the
Inherent problems involved with clay mineral preparation'and
analysis, these data are given to indicate the general trends and
are semi-quantitative at best.
The values given In tables 1 through 6 are not meant to represent
the regional trace element and mineral contents for the given coal
seams because the actual amounts present vary from sample to
sample. It should be noted that amounts also vary with the
separatory methods and analytical techniques used.
Data are presented in table 7 indicating the concentrations of
major exchangeable ions found in the original coal, the residue after
ammonium acetate extraction, and the extracted solution found in
three of the coals studied. Data for the Illinois coal studied are
42
-------
TABLE 5
Results of mineralogical analysis
Sanple 1
* C18560
• • C18562
5 ° C18563
3 C18564
•H ^ C18565
J{ C18566
K C18567
C18848
j- C18889
• §C18878
£ « C18879
. .30.8880
a < C18881
« C18882
. C19854
•5 JSC19848
•g -C19849
« gC19850
* S C19851
C19852
£• C19853
o ,,019824
~ -g C19827
•" 'H C19828
g4jC19829
•g > C19830
*J « C19831
jj ^ C19832
Fraction
Raw .coal
28M x 0
1.29F
1.33FS
1.40FS
1.60FS
1.605
Raw coal
28M x 0
1.30F
1.32FS
140FS
1.60FS
1.60S
Raw coal
1.301F
1.32FS
1.35FS
1.40FS
1.60FS
1.60S
Raw coal
1.275F
1.292F
1.32FS
1.40FS
1.60FS
1.605
Recovery
_
—
34.3
25.9
18.6
12.5
8.7
__
—
25.3
20.5
36.0
11.8
6.4
_
36.8
24.4
13.1
12.3
10.4
3.0
—
27.8
26.5
19.7
13.3
5.5
7.2
LTA
(Z)
20.37
25.17
6.10
9.81
17.62
26.48
77.80
12.67
11.38
3.76
6.15
9.71
19.76
59.75
13.09
7.41
9.46
6.91
11.35
20.53
62.88
14.49
5.11
6.42
9.28
14.01
24.14
80.04
Pyrite
(Z)
39
28
19
22
23
31
49
-------
TABLE 6
Results of clay mineral analysis <<2 vm fraction of LTA) of selected coals
-e-
js
Pittsburgh CNo.
West Virginia
Rosebud ,
Montana
Sample #
8),
C1982A
C19827
C19828
C19829
C19830
C19831
C19832
C19854
C19848
C19849
C19850
C19851
C19852
C19853
Fraction
Raw coal
1.275F
1.292FS
1.32FS
1.40FS
1.60FS
1.60S
Raw coal
1.301F
1.32FS
1.35FS
1.40FS
1.60FS
1.60S
Illite (%)
26
20
21
21
21
25
33
17
17
12
16
17
20
28
Kaolinite (%)
47
60
65
62
55
38
19
64
68
71
77
71
67
56
Mixed layer clays (%) Chlorite (%)
27
20
14
17
24
37
48
11
5
9
7
12
13
16
|;i
^^^
^_
8
10
8
^_
—
-------
TABLE 7
Comparison of elemental concentration
in ammonium acetate extracted (lon>-exchanged) samples
Ul
C-19824 Pittsburgh (No. 8)
Element
Si
Al
Mg
Na
K
Ca
Fe
Ti
P
V
Cl
Li
Be
Sr
Ba
B
Original
coal
(%)
2.12
1.10
.16
.07
.10
1.21
1.07
.057
.012
(Ppm)
17
1400
6.7
.28
129
121
82
Residue
(Z)
2.17
1.11
.14
.04
.10
1.06
1.12
.054
.011
(ppm)
17
676
6.9
.30
130
109
21
Removed3
in NH4AC
(ppm)
_
—
40
300
<2
989
—
—
—
(ppm)
._
778
< .3
< .8
<4
10
C-19854 Rosebud
Original
coal
(Z)
2.41
1.15
.44
.019
.08
.97
.49
.06
.012
(ppm)
10
75
14.4
47
103
808
100
Residue
(Z)
2.42
1.16
.11
.00
.08
.28
.59
.07
.012
(ppm)
12
24
14
45
50
700
10
Removed3
in NH4AC
_
—
.27Z
200 ppm
12 ppm
.53%
—
—
—
,^_
50
< .3
< .8
67
86
C-18848
Original
coal
1.72Z
1.44%
.02%
170 ppm
1000 ppm
.09%
.36%
.13%
190 ppm
52 ppm
.02%
—
122 ppm
170 ppm
Blue Creek
Residue
1.71Z
1.45%
.02%
140 ppm
1000 ppm
.06%
.36%
.13%
204 ppm
53 ppm
.01%
—
120 ppm
155 ppm
Removed3
in NH4AC
(ppm)
_
—
<1
<8
.1
134
—
—
—
(Ppm)
,^_
—
—
<1
16
values actually found in the extract.
-------
not available at this time. Because different amounts of specific
ions may be exchanged by various media, the data presented here
can be Interpreted to indicate only the relative potential for
removal.
DISCUSSION
Dozens of minerals have been reported in coals. Sulfldes,
sulfates, carbonates, quartz and clay minerals, together with
many trace minerals,, form a multi-component system with complex
origins and variable chemical compositions. The chemical elements
present in the mineral matter occur not only as major components
of minerals, but also to a limited extent as isomorphic replacements,
in solid solution or as exchangeable cations on clays. These
types of sites in the mineral matter are presumably the position
of many of the trace elements found in coals.
Table 8 surveys the principal minerals commonly found In
coals and some of the trace elements potentially associated with
them. Specific associations have been compiled from the results
of trace element investigations of coals (Gluskoter et al., 1977;
0'Gorman and Walker, 1972; Miller and Given, 1978) as well as from
reviews of basic geochemlcal and mineral research (Deere et al.,
1966; Weaver and Pollard, 1973; and Grim, 1968). This partial
listing does not preclude the probability of additional mineral
matter-trace element associations.
Sulfides. In addition to the sulfides listed in table 8 ,
trace amounts of millerite (NiS), cinnabar (HgS), and galena (PbS)
have be§n reported in coals, and undoubtedly more sulfide phases
will be documented in the future with the use of advanced electron
microscope methods. Iron is the dominant'element in the sulfide
46
-------
TABLE 8
Elements potentially associated with
minerals commonly found in coals
Mineral phases
Sulfides
Pyrite, marcasite
Sphalerite
Chalcopyrite
Galena
Sulfatea
Barite
Gypsum
Carbonates
Calcite
Siderite
Ankerite
Dolomite
Phosphates
Apatite
Major constituents
Fe,
Zn,
Cu,
Pb,
Ba,
Ca,
Ca
Fe
Ca,
Ca,
Ca,
S
S
Fe, S
S
S
S
Fe
Mg
P, F
f As,
\ Fe,
1 Ni,
Sr,
(Ba,
\Fe,
^ •
Mn .
Trace constituents
Cd, Hg, Ag, Pb,
Zn, Cu, Co, Sn,
Mo, Se, Ga
Pb, Ca
Sr, Pb, Mn, Ca
Mg
HI . Mff . f!A . nt-h^r t*arA »nv
Silicates
Quartz
Zircon
Tourmaline
Plagioclase
Orthoclase
Muscovite
Clay minerals
Kaolinite
Illite
Montmorillonite
Mixed layer clays
Chlorite
Si
Si, Zr
Ca, Mg, Fe, B, Al, Si
Ca, Ka, Al, Si
K, Al, Si
K, Al, Si
Al, Si
Al, Si, K
Al, Si, Mg, Fe
Al, Si, K, Mg, Fe
Al, Si, Fe, Mn, Mg
Hf, Th, ?
Li, F
Ba, Sr, Mn, Ti, Fe, Mg
Rb, Ba, Sr, Fe, Mg, Ti, Li
F, Rb, Ba, Sr
Ti
Fe, Mg, Ca, Na, K, Ti,
Li, V, B, Mn, Cr, Cu, Ni,
Rb, Cs, Ga.Zn, Se, F,
La, Ba, Sr, Co, and others
NOTE: This partial listing does not preclude the probability of additional mineral-
trace element associations.
47
-------
group, usually occurring with Zn and Cu In solid solution; the
associated chalcophlle elements listed can also be present In small
amounts In sulflde minerals. For example, analyses of sphalerite
from coals in northwestern Illinois have revealed 0.15 to 0.93
percent Pe, 0.30 to 0.56 percent Cd, 360-470 ppm Cu, <10 to 440 ppm
Ni, 33 to 110 ppm Ga, and lesser amounts of other elements (Cobb
et al., 1978).
Carbonates. Calcite and siderite are the most common
carbonate minerals present in U.S. coals, although ankerite and
dolomite are more frequently reported worldwide. Compositions
of carbonate minerals vary appreciably because of the extensive
solid solution series of calcium, iron, and magnesium. In
addition, small amounts of Mn, Sr, and Ba may replace the major
cations in the crystal lattice and are commonly reported in
chemical analyses of carbonate minerals.
Clay minerals. Kaolinite, illite, and mixed-layered illite-
montmorillonite clays commonly make up a major portion of the
mineral matter of most coals. Cation absorption and exchange are
important properties of these minerals and the minor and trace
alkali and alkaline earth elements are favored for the exchangeable
sites in clays. Due to inherent higher cation-exchange capacities,
illites, montmorillonltes, and mixed-layered clays tend to absorb
a greater variety of ions than kaolinite. A number of elements
are also known to substitute for Al, Si, and other major constituent^
bound into the crystal lattice. Trace element analyses of partings
and shale strata associated with coal seams indicate higher
concentrations of many minor and trace elements in these components
but because of the complex combinations of clays and other
48
-------
Incorporated minerals, specific mineral-trace element associations
are not conclusive.
With the exception of quartz, many of the other minerals
listed occur only in trace amounts in most coals. Most of their
trace element associations are due to isomorphic substitution
of preferred ions in the crystal lattice.
Mineral analyses show that certain mineral phases are
ubiquitous in these coals, although the amount and proportions
of these minerals vary because of individual variations related
to rank, geological conditions, and the geochemical environment.
In general, the percentages of pyrite, calcite, quartz, and most
of the minor minerals increase in the heavier gravity fractions
as the relative percentage of total clays in the mineral matter
decreases. The exception, bassanite, is not a naturally occurring
mineral in coals and forms during low-temperature ashing preparation
by the fixation of organic sulfur with exchangeable Ca cations
derived from organic carboxyl groups in low-rank coals (Miller
and Given, 1978).
The low-temperature ash of whole coals shows significant
variation in total clay content, ranging from 42 percent to 85
percent. Increased percentages of the mineral matter in the
lighter gravity fractions of the coal is composed of clays finely
dispersed within the macerals. Compositional trends of the two
coals selected for additional clay analysis show higher proportions
of kaolinite in the mineral matter portions of the lighter fractions
and increased amounts of illite and mixed-layer clays in the 1.60
sink fraction. Such variations have a practical importance for
utilization. The composition of"the clay minerals in coals affects
49
-------
the fusion temperature of the resulting ash. White (1964) has
indicated that increased concentrations of illite, especially in
conjunction with higher amounts of carbonate, lowers the melting
point and viscosity and changes the glass-forming characteristics
of the ash. If a portion of the clays is removed during cleaning,
the resultant clay composition may substantially change, altering
trace element contents and adsorption properties as well as the
fusion and sintering characteristics of the ash.
Washabllity curves prepared from the mineral data are
shown in figures 4, 5, 6, and 7'. They indicate that a large
portion of the pyrite and calcite in the four coals studied can
be concentrated and removed through physical cleaning methods;
removal of quartz and clay minerals is not as effective as removal
of the heavier minerals.
Prom the information presented in the preceding tables
some conclusions can be drawn regarding the distributions of a
number of elements and their potential for removal during coal
cleaning. An inspection of the data in table 9 indicates the
wide variation in the organic association of trace elements in
coals. For the coals selected in this study, only Br, Qe, and
organic S are consistently in the most organic category. This
Indicates that these elements cannot be removed to any degree by
specific-gravity cleaning procedures. Only Mn and As show great
inorganic affinity in all coals and therefore can be most easily
removed. A number of other elements—Sn, K, Pb, P, Zr, Se, Ti,
Li, Al, Hf, Cs, Si, Cd, Pe, Zn, and pyritic S—all of which can
be related to discrete mineral phases—show a consistently high
degree of cleaning potential, whereas Hg, sulfate S, Sr, Ba, Qa, Na,
SO
-------
0 20 40 60 80 100
PERCENT RECOVERY
0 20 40 60 80 100
PERCENT RECOVERY
2.8-
2.4-
1.9-
1.4-
0.9-
0.5-
0.0
QUARTZ
' 20 ' 40' 80' 80
00
8.1
6.8-
6.4-
«-1-
2.7-
1.4-
0.0
TOTAL
CLAYS
PERCENT RECOVERY
I20406080 100
PERCENT RECOVERY
HERRIN (No. 6) COAL SEAM, ILLINOIS
Iftil HH
Figure 4—Mineral distributions in a single sample of the Herrin
(No. 6) Coal Member, Illinois.
2.21
0 20 ' 40 60 80 100
P.ERCENT RECOVERY
0 20 40 60 86 100
PERCENT RECOVERY
20'40 60 80 160
PERCENT RECOVERY
TOTAL
CLAYS
*0 20 40 60 80 100
PERCENT RECOVERY
PITTSBURGH (No. 8) SEAM, WEST VIRGINIA
Figure 5—Mineral distributions in a single sample of the Pittsburgh
(No. 8) seam, West Virginia.
51
-------
.10-
.09-
.07-
.05-
.04-
.02-
nn-
PYRITE 1
1
• • -J
PERCENT RECOVERY
6 20 40 60 80 100
PERCENT RECOVERY
2.9
2.4-
2.0-
1.5-
1.0-
0.5-
0.0
QUARTZ
II 20 40 60 80 100
PERCENT RECOVERY
7.6i
6.3
5.1-
3.9-
2.5-
1.3-
0.0
TOTAL
CLAYS
0 20 40 60 80 100
PERCENT RECOVERY
0.6
0.5-
0 20 40 60 80 100
PERCENT RECOVERY
BLUE CREEK SEAM, ALABAMA
Figure 6—Mineral distributions in a single sample of the Blue Creek seam,
• Alabama.
0.5-
0.4-
0.3-
0.2-
01-
n ft-
PYRITE
0.6-1
0 20 40 60 80 100
PERCENT RECOVERY
20 40 60 80 100
PERCENT RECOVERY
1.6-
1.3-
1.1-
% 0.8-
0.5-
0.3-
0.0
QUARTZ
0 20 40 60 80 100
PERCENT RECOVERY
7.0-
5.8-
4.6-
X 3.5-
2:3-
1.2-
nn.
TOTAL CLAYS^
•
0 20 40 60 80 100
PERCENT RECOVERY
ROSEBUD SEAM, MONTANA
Figure 7—Mineral distributions in a single sample of the Rosebud seam,
Montana.
52
-------
TABLE 9
Ranking of elemental organic affinities determined
from prepared washed coals
Herrin (No. 6) seam,
Illinois
ORGANIC
Ge 1.76
U
ORS
Hg
V
Br
Sb
Dy
Be .87
INTERMEDIATE ORGANIC
B, Cr .77
Ni
Co
Eu
Cu
Na
Lu
Sc
K
Th
Ag
Yb .52
INTERMEDIATE INORGANIC
Zr .49
Hf
TOS, Rb, Ti
Cs, Ta
Sm
Pb
Al
Si
Se
Mg .27
INORGANIC
SUS .17
Sn, HTA, LTA
Ba, Ga
Cd, Ce, Sr
Mn, Ca, Fe
PYS
As, La, Zn
P .03
Blue Creek seam,
Alabama
Br 1.20
Ge
Co
ORS, TOS
SUS
Hg
Ni .91
Sr .80
Cu, Dy
Be
V
La
AS
U
w
Lu
Pb, Sn
Sm, Tb
Eu, Ga, Sb, Ce
PYS
Ba
Cr, P, Zr
Se
Yb
Ti
Sc -S3
Cd .45
Hf, Fe
Th
Al
B
Ta, Ca .34
Zn .21
Na
LTA
HTA, SI
K
Cs, Rb
Mg
Aa, Mn .05
Pittsburgh (No. 8) seam,
West Virginia
ORS 1.15
B
Br
Sr
Ba
TOS
Co, Ga
Be .71
U
Sm, Na
Yb
Ce, La, P .69
Dy, Eu, I, Sc
W, Li
Lu, NI, Th, Al
Cr, Ti, V
Se, Ta .51
Du .50
Hg
Hf, Si
Sb, Zn
Ca
Zr .27
Rb, PYS .19
Fe
HTA
SUS
Aa
K
Cd
Sn
Mn, TI, Mg
Mo, Pb, Ca .05
Rosebud seam,
Montana
B, Na
W, Mo
ORS
Br, Ge, Yb
Sr, TOS
Mg, Sm
Sb, La
Ce, Lu
Ca
Co
Tb
Sc
Dy
F, Ga
Be
Sb
Lu
Ni
Ta
V
U
Th
Cu
Hf
Al
Ti
Li
LTA
TI
Cr
HTA
Cd, Si
Se
Mn, Pb, Sn, Zr
As, Cs, Hg, Rb
PYS, Ba, I, Zn,
Fe, K, SUS
1.24
.72
.68
.54
.44
.54
.18
.02
NOTE: Elements are arbitrarily grouped according to calculated organic affinity index
for individual coala.
53
-------
Mo, La, Mg, Cd, and Ca exhibit a high degree of variability in
cleaning potential depending upon the coal studied. Most other
elements such as U, V, Dy, Be, Cr, Co, Ni, Yb, W, Lu, Tb, So, Sb,
S, B, Sm, Eu, Ag, Cu, and Th are consistently in the intermediate
grouping where partial cleaning by physical methods can be
accomplished without losing a major portion of the coal. It is
further apparent that low-rank coals contain the largest number
of elements not readily removed. Since these elements in many
cases are at rather low concentrations, they may not present a
significant problem.
The potential cleaning of elements and minerals is dependent
on other parameters in addition to the organic-inorganic associations.
The particle size of the minerals plays a significant role in the
ease of cleaning. For example, scanning electron microscope studies
indicate that 95 percent of the pyrite occurring in this Pittsburgh
(No. 8) coal is encapsulated within the coal particles and has
an average particle size of 8ym. The broadness of the pyrite
washability curve in figure 5 reflects this relationship. Furthermore,
if an element such as Mn occurs in association with calcite, it is
easier to remove than if it occurs only in clay minerals.
The organic affinity can therefore be used only as an indicator
of cleaning potential; it is not absolute. Neither does the
calculated organic affinity value necessarily bear any relation
to the percentage of that element associated with the organic
fraction of the coal.
The agreement between the concentrations of elements as
extrapolated from float-sink data and the values determined from
demineralized material(tables 1, 2, 3, and 4) is within analytical
54
-------
and sampling errors in most cases. Variations greater than
expected from these causes are exhibited by some elements, however.
Although a number of factors could influence this difference, the
presence of exchangeable cations has been shown to be the primary
cause. The values given in table 7 indicate the exchanges that
can occur and show the levels of removal that may be effected.
An example of this exchangeability is shown in figure 8, where
it can be observed that the exchangeable Ca in the three coals
is approximately equal to the difference between the P/S EXT
and the MMF values in tables 2, 3, and 4. The Rosebud coal has
by far the greatest amount of exchangeable calcium. This pattern
holds true in a general way for most of the differences between
the values obtained by those procedures. Although chelated or
organically associated elements may be stripped off the organic
molecule during demineralization, especially in low-rank coals,
this factor is not often significant.
Although many elements have at least a limited organic
association, it should be noted that the total ash content of
acid-demineralized coals seldom exceeds 250 to 600 ppm. Addition
of the exchangeable, soluble, and chelated elements still results
in the conclusion that most of the trace and minor elements in
coal are in a mineral form and subject to significant reduction
by physical cleaning procedures.
SUMMARY
Data from analyses of whole coals and their specific gravity
fractions may be used to predict the organic-inorganic, associations
of trace elements in coal and to yield information about the
potential for cleaning these fuels by specific gravity procedures.
55
-------
cS
1.4 -
In
.2 •
1.0 -
0.8 -
0.6 -
0.4 -
0.2 -
0.0 -
1.21
1.06
.10
mm
!!•_
.97
«^™«^™«
.28
1
^
Pittsburgh (No. 8) seam, Rosebud seam,
West Virginia Montana
Whole coal
Residue
Extract
Blue Creek seam,
Alabama
Elemental Concentration in Ammonium Acetate
(Ion-exchanged) Samples
Figure 8—Elemental concentration of calcium in ammonium acetate
(ion-exchanged) samples.
56
-------
The acld-demineralized data can be used to estimate the total
cleaning that can be achieved by a combination of physical and
chemical separation methods.
The elements with low organic affinity are concentrated
in mineral phases that have been identified in the heaviest
washed coal fractions: Pe, Zn, Cd, Pb, As, and pyritic S occur
as sulfide minerals; Ca, Mg, Pe, and Mn occur as carbonate phases;
Al, Si, K, and Mg occur as clays and silicates; Ca and P occur
as a phosphate; and Ba has been Identified as a sulfate.
Although some elements with high organic affinities cannot be
beneficiated by float-sink procedures, they do not constitute a
major portion of the trace and minor element content of most coals.
This information was gathered with partial support from
U,S. EPA Qrant R80M03, U.S. EPA Contract no. 68-02-2130, and
U.S. ERDA (DOE) Contract no. EY77-X-21-2155. All data and
conclusions were compiled by members of the Coal Section and
Analytical Section of the Illinois State Geological Survey.
57
-------
REFERENCES
Cobb, J. C., J. D. Steele, C. G. Treworgy, J. P. Ashby, S. J.
Russell. 1978. The geology of zinc in coals of the Illinois
Basin, Final report submitted to the U.S. Geological Survey,
Branch of Eastern Mineral Resources, U.S. Department of the
Interior, Grant No. 14-08-0001-G-249, 69 p.
Deere, W. A., R. A. Howie, J. Zussman. 1966. An introduction
to the rock forming minerals. Longmans, Green and Co., Ltd.,
London, 528 p.
Gluskoter, H. J. 1965. Electronic low-temperature ashing of
bituminous coal. Fuel, 44:285-291.
Gluskoter, H. J., R. R. Ruch, W. G. Miller, R. A. Cahill, G. B.
Dreher, and J. K. Kuhn. 1977. Trace elements in coal:
Occurrence and distribution. Illinois State Geological Survey
Circular 499, 154 p.
Grim, R. A. 1968. Clay mineralogy. McGraw-Hill, New York, 596 p.
Miller, R. N., and P. H. Given. 1978. A geochemical study of
the inorganic constituents in some low rank coals. Technical
Report, Pennsylvania State University; Report FE-2494-TR-1
submitted to U.S. Department of Energy under contract No.
EX-76-C-01-2494, 314 p.
O'Gorman, J. V., and P. L. Walker, Jr. 1972. Mineral matter
and trace elements in U.S. Coals. Office of Coal Research,
U.S. Department of the Interior, Research and Development
Report No. 6l, Interim Report No. 2, 184 p.
Ruch, R. R., H. J. Gluskoter, and N. F. Shimp. 1974. Occurrence
and distribution of potentially volatile trace elements in
coal: A final report. Illinois State Geological Survey
Environmental Geology Note 72, 96 p.
Stepusin, S. M. R. 1978. Vertical variations in the mineralogical
and chemical composition of the underclay of the Herrin (No. 6)
coal in southwestern Illinois, M.S. thesis, University of
Illinois, Urbana. 68 p.
Ward, C. 1976. Mineral matter in the Springfield-Harrisburg (No. 5)
Coal Member in the Illinois Basin. Illinois State Geological
Survey Circular 498, 35 p.
Weaver, C. E., and L. D. Pollard. 1973. The chemistry of clay
minerals. Elsevier Scientific Publishing Co., Amsterdam,
London, New York, 213 p.
White, W. A., andN.R. O'Brien. 1964. Illinois clay resources
for lightweight ceramic block. Illinois State Geological
Survey Circular 371, 15 p.
58
-------
EFFECTS OF COAL CLEANING ON ELEMENTAL DISTRIBUTIONS
Charles T. Ford and James F. Boyer
Bituminous Coal Research, Inc.
350 Hochberg Road
Monroeville, Pennsylvania 15146
ABSTRACT
The coal industry, through its research affiliate Bituminous Coal Research,
Inc., has established a program for obtaining valid information concerning
potential trace element problems as they relate to coal. One means of con-
trolling trace element emissions is to remove these trace constituents prior
to combustion during the coal cleaning process. In the most recent BCR study
funded by the Department of Energy, twenty run-of-mine coals will be subjected
to conventional cleaning processes as well as to more extensive non-conventional
processes to evaluate the effect of such cleaning on fugitive elements—those
elements which might be released to the surroundings and cause environmental
problems. Based on the initial results in cleaning the first few coals, coal
cleaning represents a method for controlling potentially harmful constituents
of coal such as arsenic, cadmium, lead, mercury, and selenium.
59
-------
INTRODUCTION
In the early nineteen seventies, public concern was aroused by the
appearance of a number of newspaper and magazine articles concerning mercury
pollution. Although most of these dealt with mercury in natural waterways
and its effect on aquatic life, a few suggested that the burning of fossil
fuels might contribute significant amounts of mercury to the environment.
At that time, a program was established by the coal industry through its
research organization, Bituminous Coal Research, Inc., to investigate and
evaluate such potential problems concerning mercury and other trace elements
in coal and coal-related materials. An additional motive for the program
was the increasing concern with trace elements on the part of the government
regulatory agencies and the need to temper any future judgments against the
coal industry with the presentation of factual information.
The first order of business In the new BCR program was the development
or adaptation of analytical procedures for the precise, accurate determina-
tion of trace elements in coal and coal-related materials. Some of this in-
formation developed on the program was described in two reports (Bituminous
Coal Research, Inc., 1974 and 1975). Later, these analytical capabilities
were tested and applied to solving potential coal Industry problems in evalu-
ating the effect of a rough coal-cleaning process on the distribution and
removal of potentially harmful trace elements in coal (Ford, Care, and
Bosshart, 1976). In that study, involving coal samples from only one step
of an extensive cleaning process specifically designed for pyrite removal,
some beneficiation with respect to trace elements was demonstrated.even with
only the initial rough cleaning.
Based on those promising results, a more extensive coal cleaning scheme
was outlined, specifically designed to evaluate the effect of such cleaning
on trace or fugitive elements, those constituents of coal which have the
potential to be released and cause environmental problems, This paper des-
cribes initial progress on the fugitive element studies. The work was orig-
inally funded jointly by the U.S. Bureau of Mines, by the U.S. Energy Research
and Development Administration (ERDA), and by Bituminous Coal Research, Inc.
Funding is presently continuing with ERDA which is now part of the U.S.
Department of Energy.
Background and Justification
Coal has been found to contain nearly every naturally-occurring element.
With the exceptions of carbon, hydrogen, oxygen, nitrogen, and sulfur, which
are the principal constituents of coal, as well as iron, aluminum, silicon,
and alkaline earth metals In the ash, most elements are present in minor or
trace amounts.
60
-------
During coal mining, preparation, and utilization, these elements may be
released to the environment. Even though these fugitive elements are normally
present at extremely low levels, it may not be possible to ignore them as some
are known to be harmful to plant and animal life at relatively low concentra-
tions. The problem may, in fact, be compounded due to the large coal tonnage
consumed in electric power generation and the anticipated consumption for coal
conversion.
Trace elements, specifically mercury, cadmium, lead, arsenic, and sele-
nium, have been pointed to with increasing frequency over the past several
years as potential environmental pollutants. One means of controlling trace-
element emissions is to remove these trace constituents prior to combustion.
Those trace elements associated with the ash might be removed during the coal
preparation process. A preliminary study conducted by Bituminous Coal Research
search, Inc., demonstrated that coal cleaning is a possible technology for
controlling many of these potentially harmful constituents of coal which might
be released during combustion. Based on the results of the study, more ex-
tensive coal cleaning was recommended to determine which steps in the'coal-
cleaning process are effective in removing fugitive elements, both the trace
elements as well as the more prevalent constituents such as sulfur.
The need to better understand the fate of fugitive elements during the
coal mining, preparation, and combustion processes is imperative; however,
to date, few systematic studies of this nature have been undertaken. This
project is an initial step in that direction.
Objectives and Scope
The overall objectives of the project are: (a) to prepare a comprehensive
state-of-the-art report on the effect of coal mining, preparation, transporta-
tion, and utilization on the trace elements found in coal; (b) to determine
the effect of coal cleaning on fugitive elements; and (c) to provide a descrip-
tion of accurate analytical methods that can be used by coal industry and com-
mercial laboratories for determining the concentration of selected trace ele-
ments in a variety of coals. The portion of the study described here involves
only the effect of coal cleaning on fugitive elements; further, only the first
few of twenty coals to be cleaned for the overall study will be discussed.
The other objectives are being pursued but are in an intermediate stage of
development.
The objectives of this portion of the study were approached as follows:
The effect of coal cleaning on fugitive elements vas evaluated by exhaus-
tively studying samples of run-of-mine (ROM) coal obtained from various geo-
graphical locations throughout the United States. All ROM samples were
subjected to controlled coal cleaning at the BCR laboratories. Each coal was
crushed; the coarser fractions were subjected to heavy media separations, and
the fine fraction was hydraulically classified. This cleaning is representa-
tive of that which presently exists in the industry.
61
-------
Some of the coals were subjected to additional, more extensive, processing
and evaluation. Representative samples of these were crushed, screened, and
the resulting narrow particle size range fractions were hydraulically classi-
fied. This type of cleaning represents optimum coal cleaning attainable,
although not necessarily possible, using presently available technology.
Each sample produced by the coal-cleaning task was characterized by gen-
eral chemical and trace element analysis, as well as petrographic analyses
where appropriate. Those fugitive elements chosen for study include the
trace elements antimony, arsenic, beryllium, cadmium, chromium, copper, fluo-
rine, lead, manganese, mercury, nickel, selenium, vanadium, and zinc, along
with the more prevalent constituents such as sulfur.
EXPERIMENTAL PROCEDURE
A major portion of the overall study is to determine the effect of coal
cleaning on fugitive elements. This portion of the study can be divided into
three segments—coal cleaning, coal analysis, including petrographic analysis,
and data evaluation. The experimental approaches to each of the three seg-
ments are as follows:
Coal Cleaning
Approximately 450 kg (1,000 Ib) of run-of-mine coal will be collected at
each of twenty mines, placed in polyethylene bags to prevent trace-element
contamination, and returned to BCR in 55-gallon drums. In each case, the
coals will be representative of the seam or seams being mined at that site
and have been selected on the basis of present and near-future importance in
utilization.
The two basic laboratory-controlled cleaning procedures utilized at BCR
for this project are diagrammed in Figure 1, Preparation Flow Schematic.
Procedure A, represented by the left portion of the schematic diagram, simu-
lates coal cleaning techniques currently being utilized by the coal industry.
This procedure involves basic crushing and screening to reduce the coal to
three fractions: (a) 31.8 x 6.35 mm (1-1/4 ,x 1/4 inch), (b) 6.35 mm
(1/4 inch) x 30 mesh, and (c) 30 mesh x 0. The 31.8 x 6.35-mm (1-1/4 x 1/4-
inch) fractions and the 6.35-mm (1/4-inch) x 30-mesh fractions were then
cleaned by heavy media gravity separations (Leonard and Mitchell, 1968)
utilizing magnetite and the BCR double cone heavy media device pictured in
Figure 2. Three separating gravities (1.35, 1.55, and 1.80) were used, pro-
ducing four gravity fractions for each of the two coarse-size fractions.
The minus 30-mesh fraction was cleaned by utilizing the hydraulic clas-
sifier shown in Figure 3 after removal of the minus 270-mesh fines. Four
fractions were produced from this cleaning: a low-velocity overflow,
medium-velocity overflow, high velocity overflow, and a sink fraction.
62
-------
Run-of-Mlne
1000 Ib
I
Dry .Hi-Bay
PROCEDURE A
Jeffrey
Hammer Hill
Denver Jaw
Crusher
1-1/4
1.35 float«-
1.35 x 1.55*-
1.55 x 1.80-*-
1.80 sink«-
LVO
MVO
HVO
1.35 float*-
1.35 x 1.55*-
1.55 x 1.80<
1.80 sink4
Hydraulic
Classifier
Hydraulic
Classifier
PROCEDURE B
fl-1/4" x 0-«j
„Jeffrey
Hammer Mill
1/4"
_*_
(K
Pulvasizer
>*t Hammer Mill
0.039"
screen
Split
Pulvasizer
Hammer Mill
-30 mesh
I screen
-100 mesh
Wet-Vac
•' +30 -
30 x 40— »
40 x 50-*
50 x 70-
h70 x 100-*
100 x 140+
140 x 200*
•200 x 270*
-270
Retained Sample
Gilson
Wet-Vac
• +100 —+
100 x 140*
•140 x 200*
200 x 270*
-270
Each Fraction Individually
BCR 608201R1
Figure 1. Preparation Flow Schematic
63
-------
60I2P21
Figure 2. Double Cone Heavy Media
Separatory Unit
-------
60§2P7
Figure 3. Hydraulic Classifier Column
65
-------
Procedure B, represented by the right side of the schematic diagram, was
utilized to simulate an "optimum" coal cleaning process. Basically, as was
shown in Figure 1, the 31.8-mm (1-1/4-inch) x 0 ROM coal was reduced to a
series of fine coal fractions using the screening apparatus shown in Figure 4.
The fractions were sized from 30 mesh to minus 270 mesh and then cleaned in
the hydraulic classifier. A schematic of the components of the hydraulic
classifier system is shown in Figure 5. All but the minus 270-mesh fractions
were subjected to hydraulic classification to produce an ultra-clean coal
fraction, a clean coal fraction, a shale fraction, and a pyrite fraction for
each of the size fractions. This "optimum" coal cleaning may not presently
be feasible at commercial levels but was utilized in this project to identify
trace elements present in fine clean coal, shale, and pyrite fractions, and
to allow comparison with those found in larger-sized fractions.
By utilizing both Procedure A and B, as many as 81 samples could be
generated for trace element analysis.
Analytical Methods
For a thorough evaluation of the effect of cleaning on the coals, many
general chemical determinations were made as well as the petrographic and
trace determinations. These included proximate analysis (moisture, ash,
volatile matter, and fixed carbon), ultimate analysis (carbon, hydrogen,
nitrogen, total sulfur, chlorine, and oxygen by difference), calorific value,
and sulfur forms. ASTM methods (ASTM, 1977) were used whenever possible and
appropriate.
The major constituents of the coal ash—including silicon, aluminum,
iron, magnesium, calcium, and titanium—were determined by optical emission
spectrographic (OES) techniques. Sodium and potassium were determined by
atomic absorption. Phosphorus was determined either by a wet-colorlmetric
procedure or by atomic emission techniques utilizing apparatus purchased for
this project. The atomic emission spectrometer, a multi-element, direct-
reading analytical tool with a plasma excitation source, provided initial
accuracy checks for phosphorus and for those elements determined by atomic
absorption.
A major portion of the BCR methods for determining trace constituents in
coal is based on atomic absorption for the following reasons. First, and most
important, proper use of this technique has been demonstrated to result in
accurate, precise concentration values for trace constituents of coal when
precautions are exercised in preparing the sample for these measurements. A
second reason for selecting atomic absorption is its potential use for these
types of analyses by coal industry and other laboratories. While the corres-
ponding instrumentation is by no means inexpensive, it is certainly less ex-
pensive than other more exotic, equally accurate methods. Furthermore, a
skilled technician can conduct the analyses once the procedures are in an
advanced state of development.
66
-------
6 012 PI 6
Figure 4. Wet-Vac Screening Apparatus
67
-------
01
00
City Water
Supply
Classifier Column
325 Mesh Screen
To Drain
H
Water
Tank
300
Gallon
Differential
Pressure Gage
Water U-Tube
Nanom eter
BCR 6082G3
Figure 5. Hydraulic Classifier System Components
-------
The following elements were determined by atomic absorption using a
flame technique: arsenic, beryllium, chromium, copper, manganese, nickel,
selenium, vanadium, zinc, antimony, sodium, and potassium. The cadmium and
lead concentrations were too low in each of these coals to be detected by
this flame technique; therefore, these elements were then determined by a
flameless graphite furnace technique utilizing apparatus specifically pur-
chased for this determination. The apparatus will also be useful for other
low-concentration level determinations.
The solutions for flame and furnace atomic absorption as well as for
atomic emission were prepared by dissolving the ash in teflon beakers with
aqua regia/hydrofluoric acid (1:1) by boiling to dryness, adding nitric acid,
and diluting to volume.
Arsenic, selenium, and antimony were determined by a unique procedure
developed at BCR based on Eschka fusion, hydride generation, and atomic
absorption using a hydrogen-argon flame. This procedure is presently being
considered for round-robin testing by ASTM Committee DOS on Coal and Coke.
Mercury was determined by a combustion bomb method developed at BCR and
by a double-gold amalgamation system constructed at BCR. Both techniques
involve flameless atomic absorption and both are being considered for round-
robin testing by ASTM Committee DOS on Coal and Coke. The first method has
been tested with a great deal of success. Testing of the double-gold amal-
gamation system has been hampered by so few of the systems being available.
Fluorine was determined by a combustion bomb-selective ion electrode
method. As the work-progresses, the analytical techniques used, particularly
for the trace-element determinations, including the most recent refinements,
are being written, tested, and modified where needed. Later they will be
incorporated into an analytical procedures manual for submission with the
final report for this sponsored program.
The results of analyses are displayed on data sheets developed to show
the results of each analytical determination for each of the various fractions
obtained during the cleaning. One example of these is shown as Figure 6,
which contains the ash content for all of the fractions obtained during
cleaning of coal 1R, the first coal. From the data in this figure and from
the other similar data sheets, obvious analytical errors can be immediately
spotted. In this case, they could be observed readily as interruptions in
the trend of low to high ash content from clean coal to refuse fractions.
With the voluminous data being obtained during the course of this project, it
would be easy to miss such simple errors caused, perhaps, by an incorrectly
recorded sample number. In one case in a previous study, each fraction ob-
tained had a higher concentration value than the feed from which it came.
This was traced to a dilution error during analysis. Displaying the data as
in Figure 6 minimizes the chances for these kinds of errors.
The analytical results for the individual samples are displayed as in
Figure 7, The individual samples were also divided into groups for a clearer
69
-------
Analysis: Ash,
1-1/4"
x 1/4"
1
26.8
1
6.22
19.2
41.6
i
79.4
-270M
1
31.4
percent
1/4"
x 30M
1
15.8
1
5.86
i
22.8
i
55.8
i
78.6
200
x 270M
1
21.2
1
4.73
1
14.4
1
47.6
1
58.4
19.1 Coal: 1R
screened
30 x 270M
1
13.0
1
6.57
i
8.47
i
19.4
i
69.2
140
x 200M
1
19.9
1
4.76
1
12.4
1
50.2
1
64.6
1 split
•
-270M
1
21.1
100
x 140M
-270M
i
21.8
70
x 100M
1 1
17.2 16.9
1 1
4.16 3.57
1 1
11.4 9.64
1 1
45.8 44.8
66.8 70.0
-30M
22.3
i
200
x 270M
i
18.2
1
6.14
i
17.2
i
51.2
i
60.2
50
x 70M
1
15.6
1
3.22
1
8.75
1
39.0
76.0
1 split
140
x 200M
1
19.8
1
5.38
i
16.2
i
52.8
i
65.2
40
x 50M
1
17.2
1
3.28
1
8.68
1
45.5
79.8
100
x 140M
l
21.4
1
4.46
i
12.8
i
49.0
i
69.8
30
x 40M
1
22.8
1
4.22
1
12.5
1
58.2
82.3
-100M
i
21.9
i
-flOOM
1
27.6
1
4.61
,
12.0
i
43.9
i
73.9
+30M
1
36.5
1
6.14
1
32.0
1
83.4
82.0
BCR 6082G4
Figure 6. Ash Content of Fractions of Cleaned Coal 1R
-------
DATA SHEET, COAL PREPARATION-FUGITIVE ELEMENT STUDY
Analytical Lab No; 77-1071 Sample Code No: 1R-3M-2HC
PROXIMATE. %
Moisture
Ash
Volatile Matter
Fixed Carbon
ULTIMATE. Z
Carbon
Hydrogen
Nitrogen
Sulfur
Chlorine
Oxygen
SULFUR FORMS. Z
Sulfate
Pyrltic
Organic
MISCELLANEOUS
Calorific Value
Btu/lb
MAJORS IN ASH. Z
S102 45.4
A1203 25.7
Fe203 16.0
MgO 0.86
CaO 5.82
T102 1.38
MnO
Na,0
K20
S03 1.52
F*0s 0.56
0.64
8.47
32.8
58.7
78.4
5.20
1.44
1.48
0.20
4.81
0.00
0.91
0.57
13,934
ELEMENT, ppm
Arsenic
Beryllium
Cadml urn
Chromium
Copper
Fluorine
Lead
Manganese
Mercury (ppb)
Nickel
Selenium
Vanadium
Zinc
Antimony
Silicon
Aluminum
Iron
Magnesium
Calcium
Titanium
Manganese
Sodium
Potassium
Phosphorus
15.8
1.34
0.045
16.3
13.7
65.9
6.64
21.1
384
14.1
3.08
30.4
19.9
0.94
18,200
11,700
9,620
450
3,580
711
135
1,490
210
BCR FORM P82
Figure 7. Data Sheet- Coal Preparation
Fugitive Elements Study
71
-------
accounting of progress both in prepartion and in analyses. The relationship
of these groups to the cleaning scheme is shown in Figure 8. Groups I, II,
III, and IV involve cleaning Procedure A; the remainder concern Procedure B.
Details of each of the analytical procedures will be presented after the
twenty coals have been cleaned and analyzed. At that time, one discrete sec-
tion of the final report will describe the analytical procedures and their
development in sufficient detail for use by coal industry and other labora-
tories involved in similar work.
Petrographic Methods
Petrographic analyses were employed to supplement the chemical tests
designed to characterize coal separates evolved from the gravity cleaning
and sizing techniques. These analyses involve the use of a light microscope
to identify and measure coal constituents and included impurities which make
up each separate. The following analyses were used to optically characterize
each sample:
Coal composition: To identify and quantify the organic constitu-
ents characterizing the subject separate.
Pyrite mode of occurence; To determine the relative association
of the pyrite impurity with the coal. (Free, surface, or encased)
Pyrite size: To measure and quantify the mean projected area diam-
eter of the pyrite associated with the coal.
Results of the petrographic analyses will not be discussed in this
brief paper.
Data Evaluation
Further evaluation of the data was aided by use of a computer. Mass
balances were calculated and concentrations of the products were determined
(a) based on the ultra-clean and clean coal fractions No. 1 and 2; (b) based
on the ultra-clean and clean coal as well as the non-pyrite refuse fractions
No. 1, 2, and 3; and (c) based on the middlings fractions No. 2 and 3 which
might, in a real situation, be subjected to additional cleaning. A sample
of the computer printout for the program written for the mass balance cal-
culations is presented in Table 1. A differential value between the feed
and composite is also calculated and displayed in order that judgments can
be made on the entire operation including the cleaning and analyses.
Most of these calculated values might not be needed to evaluate each
coal and each group; however, they will be available and are easier to obtain
while the data are first being entered into the computer rather than having
to reenter all of the data at some later time. For one coal, the material
balances for 37 constituents and 18 different groups of feeds and separated
fractions could result in a total of almost 700 sets of individual calcula-
tions, such as that displayed in Table 1, available for evaluation.
72
-------
Analysis: Groupg
1-1/4"
x 1/4" x 30M
1
|
•
II
•
i
-270M x
1
1
|
i
III
i
i
200
270M
1
1
1
XII
1
1
ROM
1 split
(screened
30 x 270M -270M
I I
I»-
._ . . .
|
•
IV
1
1
140 100
x 200M x 140M
-270M
1
70
x 100M
1 1 1
1 1 1
1 1 1
XIII XIV XV
1 1 1
1 1 1
Coal :
(split
-30M
i
i
200
x 270M
i
1
i
VII
i
i
50
x 70M
1
1
1
XVI
1
1
1
V-
140
x 200M
1
VT
1
1
VIII
i
t
40
x 50M
1
1
1
XVII
1
1
100
x 140M
-100M
i
l
+100M
i i
1
•
IX
1
1
30
x 40M
1
1
1
XVIII
1
1
1
•
X
1
1
+30M
1
1
1
XIX
1
1
SCR 608265
Figure 6. Arrangement of Coal Fractions by Groups
-------
TABLE 1. SAMPLE PRINTOUT FROM MASS
BALANCE CALCULATION PROGRAM
FRACTION
NUMBER
1
2
3
4
WEIGHT
FRACTION
0.574
0.155
0.055
0.216
1.000
SULFUR
CONCENTRATION
1.09
1.76
3.80
3.58
[C] IN FEED - 1.76
[C] IN CLEANED COAL - 1.232
*[C] IN CLEANED COAL - 1.413
[C] IN MIDDLINGS - 2.294
[C] IN REFUSE - 3.625
WEIGHTED
CONCENTRATION
0.626
0.273
0.209
0.773
1.881
(COMP.-FEED) DIFFERENTIAL - 6.9%
% REMOVED
*% REMOVED
*% YIELD
% YIELD
52.228
41.116
78.400
72.900
74
-------
Linear correlation coefficients and F-test ratios were also calculated
to demonstrate the relationships between the percent removed of each constit-
uent with the percent removed of every other constituent. This was done as
an additional means of comparing the effectiveness of the fugitive element
removal of the two coals. At this time, the percent removed was selected
as the parameter to be evaluated. Later, when data for more coals are avail-
able, relationships between concentrations will be examined more thoroughly.
RESULTS AND DISCUSSION
Cleaning, analyses, and data evaluation of two coals are complete. One
coal is a mixture of Upper and Lower Freeport seams and the other an Illinois
No. 6 seam. Work on additional coals is in progress. The data were used in
evaluating the effect of such cleaning on fugitive elements.
Coal Sampling
A list of the twenty coal seams to be sampled is shown in Table 2. Most
of these have already been acquired and are available to the project.
Coal Cleaning
The first three coals were extensively cleaned using both Procedure A
and Procedure B; the fourth coal involved only the shorter Procedure A. From
the weight percent, ash, total sulfur, and calorific value data for the first
two coals, the cleaning strategies employed were particularly effective and
the separations needed to attain the objectives of the project were being
achieved.
Washability studies were utilized primarily to determine how much coal
could be produced at a given specific gravity and to define the ash and sul-
fur characteristics of the coal at that gravity. A washability study was
made by testing coal samples at carefully controlled specific gravities, then
analyzing the specific gravity fractions for ash and sulfur contents. A
table was developed detailing the weight percentages and the ash and sulfur
analysis for each fraction. The data were then mathematically combined into
"cumulative float" and "cumulative sink" for both ash and sulfur and were
utilized to develop "washability curves" which helped to characterize the
coal. One example of the washability tables developed for this study, this
one for the first coal, is presented in Table 3. The tables detail the sep-
aration gravities, the corresponding weight percentages, and the ash and
sulfur analyses for each of the three size fractions.
The washability curves for the 31.8 x 6.35-mm (1-1/4 x 1/4-inch) frac-
tion of coal 1R which was cleaned utilizing the heavy media unit are shown
in Figure 9. Five curves are plotted on the chart: cumulative float ash
(CFA), cumulative float sulfur (CFS), cumulative sink ash (CSA), cumulative
sink sulfur (CSS), and the yield curve. As stated previously, these curves
can be used to determine how much coal may be produced at a given specific
75
-------
TABLE 2. COALS TO BE USED IN BCR FUGITIVE ELEMENT STUDY
Sample
1
2
3
A
5
6
7
8
9
10
Seam
Upper/Lower
Freeport
Illinois
No. 6
Rosebud
Beulah-Zap
Hannah
No. 60
Adaville
No. 1
Castle Gate
B
Lower Kit-
tan ing
Pittsburgh
Meigs Creek
Sample
11
12
13
14
15
16
17
18
19
20
Sean
Pocahontas
No. 3
Stockton
Sewell
Mary Lee
Kentucky
No. 9
Illinois
No. 5
Imboden
Upper Elk-
horn
Pittsburgh
Lower Kit-
taning
76
-------
TABLE 3. WASHABILITY ANALYSIS OF COAL 1R
Specific Gravity
Sink
1.35
1.55
1.80
1.35
1.55
1.80
Float
1.35
1.55
1.80
1.35
1.55
1.80
Separates, percent Cumulative Recovery, percent
Average
1.27
1.31
1.42
2.46
Yield
Ash
Group 11
57.4
15.5
5.5
21.6
6.2
19.2
41.6
79.4
Group III
79.6
6.8
5.2
8.4
5.9
22.8
55.8
78.6
Group IV
31.9
47.6
14.9
5.6
6.6
8.5
19.4
69.2
Sulfur
1-1/4 x
1.09
1.76
3.80
3.58
Yield Ash Sulfur
1/4 Inch Heavy Media - coarse
57.4 6.2 1.09
72.9 9.0 1.23
78.4 11.3 1.41
100.0 26.0 1.90
1/4 Inch x 30 Mesh Heavy Media - fine
1.06
2.64
5.67
11.0
30 x 270
1.06
1.48
2.90
14.5
79.6 5.9 1.06
86.4 7.2 1.19
91,6 10.0 1.44
100.0 15.7 2.24
Mesh Hydraulic Classifier
31.9 6.6 1.06
79.5 7.7 1.31
94.4 9.6 1.56
100.0 12.9 2.29
Cumulative Rejects, percent
Yield Ash Sulfur
100.0
42.6
27.1
21.6
26.0
52.6
71.7
79.4
1.9
2.9
3.6
3.6
100.0
20.4
13.6
8.4
15.7
54.2
69.9
78.6
2.24
6.86
8.96
11.0
100.0 12.9 2.29
68.1 15.9 2.86
20.5 33.0 6.07
5.6 69.2 14.5
-------
40
50-
H-
z
Ul
LU
0.
*-
<
O 7Q_
i 72.9-
Ul
>
P
_: »o
3
$
3
U
Oft _
100 -
CFS
\l
I
CSA CSS
-CF>
\
\>
\
1
1
\ \
4V 14.
|\ S
L
2
i
i
i
I
i
i
L. ,
i
i
•
*r
\
1.85 1.7J
4.0
1
A
\
5
sN
\
P^
^
V
l^
sX
r^
^
ss
N
I
i
i
i
I
i
1
i
I
i
S
^
i
1
1.65 1.55 1.45
SPECIFIC GRAVIJTY
/
^
TIE
/>
s
^
1.35
LD '
1.2
FIOAT 0 4! i' l'2 ,6 >'o ' ' ! ' ' ' '
I 71.7,
50 60 70 , 80
1-23 CUMULATIVE ASH, PERCENT
FLOAT 1 1.2 1.4 1.6 1.8 2
— 5O
h-
z
Ul
U
Ul
a.
" ti
z
•" 30
• 27.1 j*j
»-
<
20 ^
*
D
in
- 0
5
SINK
— |- | — j ! .. — | j
3-6 4 SINK
CUMULATIVE SULFUR, PERCENT BCR ^^
Figure. 9. Washability Curves for the 1% x
Fractions of Coal 1R
Inch
78
-------
gravity and to define the ash and sulfur characteristics of the coal at that
gravity.
By referring to Figure 9, it can be seen that if a 1.55 specific grav-
ity were chosen for the cleaning gravity of the 31.8 x 6.35-mm (1-1/4 x 1/4-
inch) fraction of coal 1R, the following results would be expected: the
yield in the float fraction would be 72.9 percent of the total feed coal,
the float fraction would contain 9 percent ash and 1.23 percent sulfur, the
sink fraction would contain 71.7 percent ash and 3.6 percent sulfur. It
should also be noted that the slope of the cumulative float curves gives a
quick estimate of the difficulty of cleaning a coal. The greater the slope,
the more near-gravity material and the more difficult the coal is to clean.
By interpreting the data presented and by utilizing the washability
curves, general statements can be made about the characteristics of the two
coals and the response of the coals to cleaning.
For example, it can be inferred that coal 2R is more resistant to pul-
verization than 1R, Coal 2R contains a higher percentage of material in the
coarse size fraction, 31.8 x 6.35-mm (1-1/4 x 1/4-inch), and a lower per-
centage in the fines, 30 x 270-mesh, than coal 1R. The higher percentage in
the coarse screen fraction of coal 2R indicates that leas degradation oc-
curred during the screening, implying that coal 2R is harder than 1R.
When compared at a 1.45 specific gravity, the 6.35-mm (1/4-inch) x 30-
mesh fraction showed the best ash reduction potential for both coals. The
31.8 x 6.35-mm (1-1/4 x 1/4-inch) fraction of coal 1R had the highest per-
centage ash reduction (72 percent), but the feed ash was much higher than
in the 6.35-mm (1/4-inch) x 30-mesh fraction, which yielded a higher float
ash even with the 72 percent reduction. The ash reduction potential at the
1.45 gravity was poor for both coals at the 30 x 270-mesh size range. From
the washability curves, a lower specific gravity separation for the 30 x
270-mesh fractions would yield a better cleaning response.
Coal 1R showed better total sulfur reduction potentials than 2R for all
size ranges at the 1.45 specific gravity. Coal 1R inherently has a lower
sulfur content at the run-of-mlne level and also proved, even for the 30 x
270-mesh fraction, to be easier to clean to an acceptable sulfur level.
The changes in the Btu levels between the raw coals and the coals washed
at 1,45 gravity were not significant. Only the 31.8 x 6.35-mm (1-1/4 x 1/4-
inch) fraction of coal 1R showed a reasonable increase, 22.9 percent. The
fine fractions, 30 x 270-mesh, of both coals showed a large reduction in
Btu's when cleaned at 1.45 gravity. Again, a specific gravity lower than
1.45 would show a better response for the fines of both coals.
79
-------
Analytical Data
The concentration values for the first four run-of-inine coals for this
study are presented in Table.4. Differences in concentrations for each of
the four are apparent. For example, the average concentration for 14 of
the trace elements for coal 1R is 27 ppm; for coal 2R, it is 25 ppm; for
coal 3R, it is 15 ppm. For the lignite, coal 4R, the average concentration
for the same 14 elements is only 8 ppm.
Thirty-eight constituents of samples of each of the first four cleaned
coals, including the run-of-mine sample, were determined. Contamination of
samples was a continuing problem throughout the study. Instances of mercury,
chromium, copper, lead, nickel, and zinc contamination were identified and
documented. These will be described thoroughly both in a forthcoming paper
and in a report on analytical procedures. Contamination was held to a mini-
mum by routinely analyzing reagent blanks and other material used on the
project such as the magnetite used in the heavy media separations.
Every determination was done in duplicate. Lack of duplication of values
immediately precipitated an additional determination. Whenever possible and
as time allowed, more than one procedure was used for the trace element de-
terminations as a check on the accuracy of the analytical values obtained.
Finally, as a continual check on the analytical procedures during the
project, National Bureau of Standards (NBS) certified reference materials,
SRM-1632 Coal and SRM-1633 Fly Ash, were analyzed along with each new run-
of-mine coal and its cleaned fractions. One example of the type of agree-
ment normally obtained between the determined values and the NBS certified
values is presented in Table 5.
A more thorough description of the analytical procedures and results of
precision and accuracy studies is programmed at the end of the presently
funded portion of this program, when the twenty coals have been cleaned and
analyzed and the effect of the cleaning on fugitive elements has been deter-
mined.
Mass Balance Calculations
The analytical data for each coal were used in calculating mass balances
for the various individual cleaning schemes for each coal. The mass balances
for each of the determined constituents were calculated: (a) to check the
integrity of the cleaned fractions and the validity of the analytical data
by comparing the summation of the materials in each fraction with the feed
material, and (b) to evaluate the reduction or enrichment of each constitu-
ent as a result of the cleaning process.
For each constituent, the concentration determined in a particular sepa-
rated fraction was multiplied by the weight fraction of the feed coal which
reported to that zone. The weighted concentrations thus obtained were
80
-------
TABLE 4. COMPARISON OF CONCENTRATIONS OF RUN-OF-MINE
COALS 1R, 2R, 3R, AND 4R (dry coal basis)
Coal 1R Coal 2R Coal 3R Coal 4R
Ash, percent 22.3 16.0 15.8 11.0
Total Sulfur, percent 2.30 2.73 1.18 0.89
Chlorine, percent 0.12 0.33 0.01 0.07
Pyritlc Sulfur, percent 1.78 2.01 1.02 0.73
Organic Sulfur, percent 0.46 0.50 0.14 0.10
Arsenic, ppm 40.2 11.0 8.54 10.4
Beryllium, ppm 1.84 1.10 0.87 0.70
Cadmium, ppm 0.08 0.47 0.12 0.06
Chromium, ppm 29.7 21.1 8.21 5.16
Copper, ppm 22.1 11.0 13.2 6.98
Fluorine, ppm 115 113 63.6 32.0
Lead, ppm 15.4 21.8 12.8 1.86
Manganese, ppm 50.0 36.8 65.2 33.4
Mercury, ppb 662 154 154 110
Nickel, ppm 20.9 18.4 4.98 4.60
Selenium, ppm 6.20 1.83 1.61 0.96
Vanadium, ppm 44.4 26.2 14.0 10.8
Zinc, ppm 35.5 91.8 12.8 3.48
Antimony, ppm 1.09 0.40 1.78 0.55
Sodium, ppm 448 1,230 248 6,630
Potassium, ppm 5,460 2,980 1,110 574
Silicon, ppm 57,800 37,900 39,800 14,500
Aluminum, ppm 30,800 17,700 20,200 6,470
Iron, ppm 21,100 19,100 8,560 7,840
Magnesium, ppm 1,360 870 3,150 4,060
Calcium, ppm 2,090 3,110 6,790 16,100
Titanium, ppm 1,670 1,010 1.090 370
Phosphorus, ppm 260 160 85 140
Calorific Value, Btu/lb 11,640 11,902 10,990 10,558
81
-------
TABLE 5. COMPARISON OF TRACE ELEMENT CONCENTRATIONS
IN SRM-1632 COAL WITH NBS CERTIFIED VALUES
(All values in ppm on a dry coal basis)
Element
Antimony
Arsenic
Beryllium
Cadmium
Chromium
Copper
Lead
Manganese
Nickel
Potassium
Selenium
Sodium
Vanadium
Zinc
Average ppm
Determined
3.47*
6.22
1.63
0.17
20.5
17,
31,
45.6
14.1
2,680*
2.88
362*
36.4
41.6
NBS Value,
Ppm
not certified
5.9 ± 0.6
1.5*
0.19 ± 0.03
± 0,5
20.2
18
30
40
15
i 2
i 9
± 3
± 1
not certified
2.9 ± 0.3
not certified
35 i 3
37 ± 4
* In agreement with other published values,
t Informational value.
82
-------
8ummed» and this resultant composite concentration value was compared with
the concentration in the feed.
The differential between composite and feed was obtained from each mass
balance calculation as follows:
Differential - Composite Value - Feed Value -nn
Feed Value * 10U
These values, since they indicate the extent of agreement between feed and
product constituents, can offer much insight into both the cleaning and the
analytical study. Judgments are possible on the cleaning process based on
the agreement of many of the constituents within any one cleaning operation.
Poor agreement generally might indicate problems with the integrity of the
samples obtained by cleaning; good agreement with the exception of one or
two constituents, might indicate problems in the analytical determination
of the constituents. A consistent positive or negative bias would .indicate
gain or loss of a constituent during cleaning or analysis, or analytical
problems with certain of the fractions, perhaps the clean coal or (more
likely) the refuse.
The average percent differentials for each of the first four coals are
summarized in Table 6. The values presented in the table are the averages
(absolute value) from as many as 18 values from the individual cleaning op-
erations. Low values indicate good agreement between composite and feed.
Obviously, there is a great deal of satisfaction with these values. For the
most part, the agreement was better for coals 1R, 2R, and 3R than it was for
coal 4R.
Some of the lack of agreement can be attributed to the low concentra-
tions of trace constituents associated with this fourth coal. The concen-
trations in solution were often at or near the detection limits of the
analytical methods. Additionally, slight contamination of trace constituents
routinely experienced and tolerated in previous analyses exerted a greater
influence on results at the low concentration levels. Such low concentration
levels will probably be experienced with other coals for this project.
Effect of Cleaning on Distribution and Removal
The analytical data were used in evaluating the effect of cleaning the
run-of-mine coals on fugitive element removal. An example of the type of
data available, a summary of the percent of each constituent removed as a
result of cleaning coal 1R, is presented in Table 7.
This calculation assumes the first two fractions as the cleaned coal and
the last two as the refuse. The corresponding product recovery for each
group is presented at the top of each column in this table. Most of the con-
stituents which were concentrated, not removed, by the cleaning were not In-
cluded in this table. For the most part, these were inherent parts of the
83
-------
TABLE 6. COMPARISON OF COMPOSITE AND FEED VALUES,
AVERAGE PERCENT DIFFERENTIAL
COALS 1R, 2R, 3R, and 4R
Ash
Volatile Matter
Fixed Carbon
Carbon
Hydrogen
Nitrogen
Total Sulfur
Chlorine
Pyritic Sulfur
Organic Sulfur
Calorific Value
Arsenic
Beryllium
Cadmium
Chromiun
Copper
Fluorine
Lead
Manganese
Mercury
Nickel
Selenium
Vanadium
Zinc
Antimony
Sodium
Potassium
Silicon
Aluminum
Iron
Magnesium
Calcium
Titanium
Phosphorus
Average
Coal 1R
2
2
1
1
2
2
4
11
5
5
15
4
14
3
14
9
14
3
13
9
6
5
6
8
10
8
4
4
5
7
7
3
6
Coal 2R
3
1
1
1
3
2
2
6
3
3
8
3
21
8
17
7
14
7
61
8
15
5
21
12
11
4
4
4
5
4
7
4
11
Coal 3R
5
1
1
1
3
5
5
19
7
11
11
5
9
7
8
6
9
4
18
25
7
3
23
11
10
4
4
3
9
11
12
5
4
8
Coal 4R
15
6
5
3
5
12
7
36
9
7
17
11
23
16
11
22
8
13
14
41
36
6
39
19
19
31
31
20
12
12
8
14
20
16
84
-------
TABLE 7. PERCENT REMOVED BY CLEANING,* COAL 1R
E
a
GROUP
II
III
IV
XII
XIII
XIV
XV
XVI
XVII
g Product Recovery.
5_
A«h
ST
Cl
spyr
Sorg
At
Be
Cd
Cr
Cu
F
Pb
Mn
Hg
Ki
Se
V
Zn
Sb
Na
K
Si
Al
Ft
Mg
Ca
Ti
P
21
75
52
8
65
10
82
45
63
64
61
61
71
89
55
53
44
63
74
29
73
80
77
69
73
85
53
66
46
86
60
54
4
69
6
79
22
53
41
50
42
62
77
67
42
65
36
58
33
46
67
61
52
72
67
42
49
24
12
52
54
15
66
15
71
26
52
34
48
42
59
58
69
42
67
31
45
26
45
55
51
46
67
52
38
42
34
21
73
78
14
89
18
92
38
64
51
66
55
71
85
73
60
82
47
71
36
66
73
72
65
91
72
77
59
51
26
69
71
11
83
14
88
37
68
48
63
56
70
83
72
58
78
39
66
42
61
69
68
60
85
69
76
56
44
29
67
65
11
79
13
88
28
63
45
57
53
72
82
71
64
73
41
61
30
56
53
67
59
84
70
69
52
38
80
70
66
7
82
12
91
29
66
50
61
56
73
85
65
55
75
42
66
35
61
72
69
61
86
73
67
55
43
11
66
60
5
78
7
86
26
63
46
55
43
73
81
71
45
74
39
62
22
56
69
68
58
83
72
58
52
37
81
70
62
6
80
8
89
29
78
51
56
51
75
85
70
50
72
44
65
32
61
74
71
62
84
77
58
58
31
XVIII
XIX VII
VIII
IX X Av.
percent
2i
73
66
8
81
8
88
38
67
56
64
55
81
87
76
61
79
55
71
28
65
76
74
68
87
84
59
64
43
69
71
76
8
86
10
93
44
69
60
64
61
65
82
82
64
83
58
65
52
64
72
71
64
87
75
62
62
53
82
52
68
6
80
11
87
23
58
35
49
30
60
69
68
46
74
30
51
28
42
49
49
42
82
48
59
43
33
82
55
63
7
78
10
86
25
56
39
48
31
66
75
64
49
72
32
51
29
46
54
54
47
78
55
59
42
35
75 63
71 81
72 82
10 13
85 91
13 26
93 97
37 50
66 78
55 65
64 76
49 67
82 68
87 93
59 92
60 72
80 88
52 66
67 78
41 60
62 73
73 83
71 80
63 73
90 94
74 84
68 76
60 71
45 61
21
67
66
9
79
12
87
33
64
49
59
50
71
81
70
55
74
45
63
35
58
68
67
59
83
70
61
56
41
60 50 47 63 61 58 60 56 59
*Cl«aned Coal • Fraction* 1 and 2
63
64 49 50 62 74 58
85
-------
coal molecule. The exceptions were chlorine and organic sulfur. An initial
observation of the data for these two seems to show a slight amount removed—
9 and 12 percent, respectively. However, since the mean or average percent
of the material reporting to the refuse (for disposal) was 23 percent (100
percent minus the average product recovery of 77 percent), the overall re-
sult was a slight concentration of these two in the cleaned coal over the
amounts in the feed.
The averages by group and by constituent reflect generally good removal
of most of the constituents. Arsenic, iron, manganese, pyritic sulfur, sele-
nium, lead, mercury, and magnesium were removed most effectively. Beryllium,
antimony, phosphorus, and vanadium were removed least effectively. Chlorine
and organic sulfur were actually concentrated slightly in the cleaned coal.
Similar data are available for each coal, assuming either the first two
or first three fractions from the heavy media or the hydraulic classifier
separations as the cleaned coal. The data are further summarized, in Table 8,
by constituent for the first four coals. The values presented in Table 8
are the averages from the individual values for each of the cleaning opera-
tions, for example, from the last column on the right in Table 7.
The differences in the effectiveness of cleaning each coal can readily
be seen from the data in Table 8. Most constituents were removed more ef-
fectively from coal 1R than from the other coals, but the product recovery
was also the lowest for this coal. Cleaning was least effective with coal
4R, the lignite. However, the principal reason for this is simply that the
concentrations of the impurities in this coal, as previously displayed in
Table 4, were relatively low prior to cleaning.
It must be pointed out at the same time that the calorific value of this
coal was also the lowest of the four. Perhaps the concept of a calculated
expression relating the concentration of impurities per calorific value, sim-
ilar to the "pounds of sulfur per million Btu's," is needed to compare the
trace impurities in a wide variety of coals on an equal basis.
Differences in the relative rates of removal for each coal can also be
observed from the data in Table 8. Cadmium and zinc were removed more effec-
tively from coal 2R than from the others.. They were also present in higher
concentrations in this coal from the Illinois basin than in the other three.
Linear correlation coefficients and F-test ratios were calculated for
the constituents to demonstrate the relationships between the percent removed
of each constituent with the percent removed of every other constituent. The
relationships were evaluated among "percent" of the constituents removed by
group rather than among "concentration" of each constituent removed by coal.
This was done simply because two coals are not a statistically significant
representation to permit drawing any specific conclusions concerning coal in
general and the effect of cleaning them on fugitive element removal. Concen-
tration data will prove more meaningful later in the study when more coals
are cleaned.
86
-------
TABLE 8. COMPARISON OF EFFECTIVENESS OF CLEANING COALS 1R,
2R, 3R, AND 4R BY CONSTITUENT. AVERAGE PERCENT REMOVED
FOR ALL GROUPS
Coal 1R Coal 2R Coal 3R Coal 4R
Product Recovery 77 87 87 88
Btu Recovery 91 97 92 94
Ash 67 48 41 23
Total Sulfur 66 46 56 46
Chlorine 9 8 16 14
Pyritic Sulfur 79 55 61 50
Organic Sulfur 12 9 16 14
Arsenic 87 53 70 50
Beryllium 33 20 21 14
Cadmium 64 77 34 15
Chromium 49 28 29 24
Copper 59 37 28 18
Fluorine 50 33 26 12
Lead 71 61 31 22
Manganese 81 65 20 33
Mercury 70 42 41 31
Nickel 55 25 27 19
Selenium 74 34 45 24
Vanadium 45 23 25 16
Zinc 63 71 30 22
Antimony 35 19 22 19
Sodium 58 25 28 9
Potassium 68 39 53 19
Silicon 67 43 43 32
Aluminum 59 42 37 19
Iron 83 65 76 51
Magnesium 70 42 13 12
Calcium 61 72 14 16
Titanium 56 34 33 27
Phosphorus 41 66 21 14
Average 58 42 34 24
87
-------
The F-statistic was then utilized in evaluating the relationship at the
selected confidence level. The relationships are summarized in Table 9 for
the first two coals. Only those having the highest degree of correlation
are listed. Even with the imperfect means of evaluating, the distinct dif-
ferences between the two coals, particularly in response to cleaning, are
again evident. The associations for 2R are essentially those described pre-
viously and proficiently for Illinois coals (Ruch, Gluskoter, and Shimp,
1974, Gluskoter, 1977). Many of these associations are quite different with
coal 1R from the Appalachian region. Not only did the two coals respond
quite differently to the cleaning as described, but also the relationships
between the constituents as removed by the cleaning were different.
These studies with the first few coals, as well as previous limited
rough cleaning studies at BCR with eight coals have confirmed that removal
of some potentially harmful trace elements can be effected by coal cleaning.
The studies thus far with the coals reported here revealed differing respon-
ses to the various cleaning processes. Generalizations concerning the effect
of cleaning on fugitive elements are not possible at this time. Specific con-
clusions concerning elemental distributions are not possible either at this
time. Some trends have already been noted and these will be reexatnined when
a statistically significant number of coals have been cleaned and analyzed.
Studies with additional coals are planned and already under way.
The program is flexibly designed to add or delete constituents of coal
by mutual agreement between the sponsor and BCR, as a result of new proposed
legislation, or by some new evidence of environmental contamination. .The
extensive cleaning scheme will also be similarly modified when warranted by
the initial results of the studies. Milestones in reporting for this study
include a report in preparation on the state-of-the-art of fugitive element
emissions as they relate to the coal industry, as well as two final reports,
one describing the results of the studies, the other including a complete
description of the analytical procedures used in the study.
88
-------
TABLE 9. PERCENT REMOVAL RELATIONSHIPS, COAL 1R AND 2R
Highest Degree of Correlation
P.R.*
Ash
Spyr
S0rg
As
Be
Cd
Cr
Cu
F
Pb
Mn
Hg
Ni
Se
V
Zn
Sb
Na
K
Si
Al
Fe
Mg
Ca
Ti
Coal 1R
Be, P
Cr, Cu, F, Mn, Zn, Na,
Si, Al, Mg, Ti
Spyr
Sji As
Spyr, Fe
P.R.*, Cr, V, Na, Ti, P
Ash, Be, V, Zn, Na, Si,
Al, Mg, Ti
Ash, Zn, Na, Al, Ti
Ash, Na, Al, Ti
Ash, Zn, Si, Al
Be, Cr, Na, Al, Ti
Ash, Cr, Cu, Mn, Na, K,
Si, Al, Mg, Ti
Ash, Be, Cr, Cu, F, V, Zn,
Si, Al, Mg, Ti
Zn, Si, Al, Mg, Ti
Ash, Cr, Mn, Zn, Na, K,
Al, Mg, Ti
Ash, Cr, Cu, F, Mn, V, Zn,
Na, K, Si, Mg, Ti
Spyr* As
Ash, Cr, Mn, Zn, Na, K
Si, Al, Ti
Ash, Be, Cr, Cu, F, V,
Coal 2R
org
K, Si, Al, Mg
Spyr* Fe
C
&org
Si, Pb, Fe
P.R.*, Cl
V, Ti
Zn
V
K, Si, Mg
Spyr. Fe
Ca
Be, Cr, Ti
Cd
Ash, Si, Mg, P
Ash, F, K, Al, Mg
Ash, Si
ST» Spyr* Pb
Ash, F, K, Si
Mn
Be, V
Zn, Na, K, Si, Al, Mg
P.R.*, Be K
*P.R. » Product Recovery
89
-------
REFERENCES
1. The development of analytical methods for determining trace elements in
coal and coal-related products, Bituminous Coal Research, Inc., Pitts-
burgh, Pa., July, 1974. 91 pp.
2. Analytical methods for determining mercury in coal and coal mine water,
Bituminous Coal Research, Inc., Pittsburgh, Pa., July 1975. 23 pp.
3. Ford, C. T., Care, R. R., and Bosshart, R. E., Preliminary evaluation of
the effect of coal cleaning on trace element removal. Trace element pro-
gram, Report No. 3, Bituminous Coal Research, Inc., Pittsburgh, Fa.,
July, 1976. 116 pp.
4. Leonard, J. W. and Mitchell, D. R., Coal Preparation, 3rd ed., New York:
AIME, 1968.
5. 1977 Annual book of ASTM standards, part 26, gaseous fuels; coal and coke;
atmospheric analysis, American Society for Testing and Materials, Phila-
delphia, Pa., 1977. 840 pp.
6. Ruch, R. R., Gluskoter, H. J., and Shimp, N. F., Occurrence and distribu-
tion of potentially volatile trace elements in coal: a final report,
111. Geol. Surv., Environ. Geology Notes No. 72 (1974). 96 pp.
7. Gluskoter, H. J., et al, Trace elements in coal: occurrence and distri-
bution 111. Geol. Surv., Circ. 499 (1977). 154 pp.
90
-------
PARTICLE SIZE DISTRIBUTION IN
THE LIBERATION OF PYRITE IN COAL
Harold L. Lovell
Professor of Mineral Engineering
The Pennsylvania State University
121 Mineral Science Building
University Park, Pennsylvania 16802
ABSTRACT
The potential for sulfur reduction by physical coal preparation is
primarily dependent upon liberation of pyrite particles within the plant
feed. Secondarily, the principles of separation, as applied in the various
plant unit operations, determine the efficiency of rejection of the liberated
pyrite. Unfortunately, the procedures currently utilized to establish the
degree of pyrite liberation are crude, inadequate, and tedious. It follows
that lack of data may preclude optimizing preparation plant design and
operation preventing maximum pyrite rejection. It even appears that improper
pyrite liberation and pre-separation processing may inhibit maximum rejection
of liberated pyrite from a raw coal.
Pyrite liberation is basically dependent upon pyrite particle size, the
nature of the pyrite particle occurrence, and the associations of the pyrite
particles with other coal components. This paper describes studies designed
to provide additional information on these fundamental questions. Laboratory
procedures involving selective stage communication, particle sizing, gravity
separations, and microscopic examinations are employed to better define the
problems and seek pragmatic solutions.
Similar concerns apply to the liberation and separation of other coal
components (mineral and phytogenic) of concern to coal preparation whether
to minimize environmental problems or prepare feedstocks for coal conversion
processes. The same relationships are operative in water mineralization from
pyrite which creates coal mine drainage problems* The studies are viewed as
fundamental to enhanced coal beneficiation and coal industry environmental
approaches.
91
-------
The occurrence of sulfur in coals as three forms (pyrite,
inorganic sulfates, and organically-bound structures) explains
the primary limitation to sulfur removal by physical coal bene-
ficiation. Assuming none of the organic sulfur will be rejected,
the organic sulfur concentration becomes the theoretical reduc-
tion limit. This limit is less than rigorous since not all of
the pyritic or sulfate sulfur will be rejected, but some organic
sulfur-containing coal components will accompany the refuse.
The pragmatic question becomes: How much pyritic sulfur in a
given coal can be rejected by physical beneficiation procedures?
The extent of pyritic sulfur rejection achievable is
dependent upon: 1.) Pyrite particulate liberation, 2.) The
development of an appropriate feed and its characteristics for
the separational unit operation(s), and 3.) The characteristics
of the separational process(es), which determine the efficiency
of liberated pyrite rejection. Similar factors are usually
evaluated to serve several coal processing objectives including
product particle size and the rejection of other mineral matter
in addition to pyrite. Consequently, the factors are typically
not optimized solely to attain maximum pyrite rejection. Only
92
-------
the first of these factors -pyrite liberation - is considered
in this discussion although the others are interrelated.
The Liberation of Pyrite
Pyrite liberation is not a YES or NO accomplishment. The
extent of pyrite particle purity attainable varies from particles
which are predominantly coal and/or mineral matter to those
predominantly pyrite. The implications for rejection extends to
the separational process and especially the density of separation.
The more pyrite freed from lower density particles, the greater
is the probability of lowering the sulfur content of the clean
coal product. The concept that liberated "pyrite" may be readily
separated from low mineral matter coal particles because of its
high density can be misleading.
Pyrite liberation, as liberation of any mineral during pro-
cessing, is achieved by particle size reduction and is controlled
by the mode(s) of comminution. In many ores, there is a
relatively narrow natural grain size distribution whose maximum
frequency may range from 10 mm to roughly 40 microns. By
reducing the total particle size of the feed to this natural
prevalent grain size, the probability of freeing associated
particles of different composition is maximized. Breakage across
constant composition particles is not precluded. The cementitious
material between adjacent particles usually is a significant
parameter. This level of comminution must be approached (despite
cost) to achieve acceptable quality and yield of concentrate
following separation.
93
-------
With the extreme heterogeneity of coals, the particle size
of any given relatively homogeneous component may range from
30 mm to less than one micron, with a broad spectrum of com-
ponent size distribution. Pyrite itself occurs in such variable
modes while the desired organic coal components tend to have more
consistent and smaller sizes (one to 50 microns) with the
notable exception of vitrinite which may exist in massive formats
of relatively high purity having much larger dimensions. Existing
coal preparation procedures do not seek to liberate the organic
components of coal, although this may become more critical as
our coal conversion program matures. The shapes and associations
of coal components are highly variable due to their combined
phytogenic and geologic origin.
Consequently, coal beneficiation design for pyrite liberation
must follow a philosophy divergent from ore processing in its
comminution-liberation objectives based on economics, separational
process requirements, dewatering considerations, and clean coal
product handling.
Coal comminution, which begins with the mining system/ must
provide for as extensive a liberation as feasible, but need not
be constrained by the minimum top particle size (3/8 to 4-inches)
of the process feed. The friability of coals complicates the
comminution control and produces levels of coal fines greater
than desired but it also emphasizes the hardness variations of
coal components which is critical in liberation considerations.
94
-------
The potential for liberating pyrite particles in run-of-
mine coals is variable and dependent upon: 1.) Range of pyrite
particle size occurrence and their dissemination modes (from
several inches to less than one micron). 2.) Coal component
hardness varies from resilient exinites and pliable, fine grained
clays to hard, brittle quartz and pyrite, 3.) The associations
of coal components is highly heterogenous from isolated pyrite
lenses (of millimeter thickness) to the localized single grain
of pyrite (micron dimensions) completely encased in vitrinite.
Some examples of this association variability are shown:
A surface textured-euhedral pyrite particle of millimeter
dimension with intergrown coaly material. A dendritic pyrite
particle of millimeter dimensions. Pyrite crystals of micron
dimensions nucleated inside a plant fiber shown in transmitted
and in reflected light. Framboidal clusters of about three
micron diameter in a vitrinite. The highly textured, large
surface area of a pyrite framboid as seen through the scanning
electron microscope. Small opaque particles of pyrite encased
in vitrinite as seen by thin section.
Inadequacy of Available Procedures to Establish Degree
of Pyrite Liberation
The extent of liberation potential of any locked mineral
particle is typically established by chemical analysis on a
size-density designated set of fractions, often associated with
95
-------
comminution procedures. In contrast with metal ores where a
high concentration of the desired component mineral is essential,
in coals pyrite rejection is sought and seldom is there signi-
ficant levels of high purity pyrite particles achieved. Coal-
pyrite liberation evaluations seek measures of pyrite rejection
into acceptably small refuse fractions whose pyrite concentrations
may be no more than double those found in situ in the coal seam.
In fact, carefully sized coal fractions which are fractionated
by density (as at 2.80 gm/cc - where only iron sulfides and
alkaline earth carbonates could be expected to sink) seldom exceed
80% FeS2 purity. Extending beyond coal preparation, there have
been studies made (Lovell, 1967) to concentrate pyrites from coal
to be used for sulfuric acid manufacture and iron ores. This
historical approach may be misleading in achieving either maximum
pyrite rejection or desired coal recovery.
Approaches to Pyrite Liberation Evaluations
The conventional approach to express pyrite liberation from
coal may be illustrated for the Clarion seam using data from
Zeilinger and Deurbrouck, 1968. By procedural convention, the
samples were crushed to 1 1/2-inches top size (feed 32-58% +
1 1/2-inches) and 'a portion further crushed to 3/8-inch top size.
The minus 100 mesh (150 microns)(0.7 to 4.0 Wt. %) material was
removed prior to particle density fractionation between 1.30
and 1.58. By summary from Table 1, when utilizing the most ideal,
conditions for sulfur rejection (1.30 gm/cc separation), there
96
-------
Table 1
Pyrite Liberation Studies of the
Clarion Seams in Clarion County, PA*
%, S
Feed 3.48
5.90
4.03
4.74
1.30F 1.71
1.68
2.06
1.94
1.58F 2.43
vo 2.58
^ 2.64
2.84
1.58S 9.16
15.49
18.07
19.22
1 I/
Wt.,
43.9
29.3
63.5
52.4
84.4
74.3
91.0
88.4
15.6
25.7
9.0
11.6
2-inch x 100M
% S Recovery, % S Rejection, %
21.6
8.3
32.5
21.4
58.9
32.5
59.6
53.0
41.1
67.5
40.4
47.0
%, S
3.63
5.42
4.23
5.24
1.66
1.64
1.94
1.87
2.25
2.41
2.41
2.51
10.43
15.55
21.09
22.72
3/8-inch
Wt., % S
48.5
32.3
66.1
58.8
83.2
77.1
90.3
86.5
16.8
22.9
9.7
13.5
x 100M
Recovery, % S Rejection, %
22.2
9.8
30.3
21.0
51.6
34.3
51.4
41.4
48.3
65.7
48.4
58.5
Zeilinger, J. and A. Deurbrouck. Preparation Characteristics of Coals From Clarion County, Pa.
Report of Investigation 7174. U.S. Bureau of Mines. 1968.
-------
was only slight reduction in sulfur recovery for two sample
locations when contrasting the two stage 3/8-inch top size with
the single stage 1 1/2-inch product. A sulfur recovery reduc-
tion of 2.2% from 32.5% and of 0.4% from 21.4%. The sulfur
concentrations were lower, 0.04 and 0.12%, respectively, and
yields higher. The liberation effect appears more pronounced
when separating at 1.58 gm/cc where the sulfur rejection
increased between 0.8 and 11.5%, although the sulfur concentra-
tions in the clean coal are much higher under the latter
condition. These data reflect the difficulty in liberating "pure"
pyrite particles, but may be interpreted as indicating a coal with
limited sulfur removal potential. Actually, this coal is subject
to significant pyrite removal.
This type of data may be readily summarized to compare seam
characteristics expressing liberation responses by crushing as
shown in Zeilinger and Deurbrouck, 1968 - Table 36, and in
Deurbrouck and Palowitch, 1966 - Table 1. Liberation by more
extensive comminution to 14 mesh (1.2 mm) has subsequently been
detailed by the same research group CDeurbrouck, 1972; Cavallaro,
Johnston, and Deurbrouck, 1976), although this size is much
smaller than currently processed industrially.
Another approach to express pyrite liberation was utilized
at The Pennsylvania State University by Kestner, Confer, and
Charmbury, 1962, in evaluating responses of different crusher
types. The crushed product was screened and each size fraction
98
-------
separated at a density of 2.96 gm/cc to attain a measure of
relatively pure pyrite particles liberated.
Since many pyrite particles extend to the sub-sieve size range
other dimensional measurement techniques must be used. Micro-
scopic particle size measurement, distributions and particle
associations were reported from polished briquettes of minus
14 mesh coal samples by McCartney, O'Donnell, and Ergun, 1969.
The samples studied included heavy liquid sink fractions. These
workers concluded that "The probable efficiency of any process
for pyrite removal from a given coal can be better estimated if
the size distribution therein is known."
The use of coal particles mounted in polished epoxy resins
for reflectance microscopy for pyrite, organic, and mineral coal
components has been highly developed by Spackman, Davis, and
Vastola, 1977, and Reyes-Navarro and Davis, 1976, at The Pennsylr-
vania State University, including automated scanning and computer
data handling. Pyrite is especially responsive to this technique
due to its high reflectivity.
These microscopic techniques have been used to study coal
pyrite occurrences in relationship to coal utilization and
environmental impacts as coal mine drainage. Although microscopic
measurements can be very effective in dealing with the sub-sieve
pyrite particles, and especially with their component associations,
they are more limited in dealing with coal preparation sulfur
removal design studies since the particle sizes have been altered
during sample preparation, the particles are less subject to
99
-------
chemical analyses, and the size distribution data are expressed
on a volume percentage basis.
An Alternative Approach to Provide More Information on
Pyrite Liberation
A procedure designed to be more helpful in pyrite liberation
studies for physical coal beneficiation is being developed by
Richardson and Lovell at The Pennsylvania State University. We
also seek to further develop the concept as relates to the liber-
ation and beneficiation of coal components other than pyrite as
well as to environmental problems. The present format is tedious
and complex, but simpler procedures for survey purposes are
envisioned.
The coal sample was subject to the laboratory flowsheet
shown in Figure 1. The data reported resulted from a 214 pound,
hand-rmined, channel of Lower Clarion Seam coal. The air dried,
raw coal was screened at one-inch and 16 mesh (1.0 mm). The
sized portions were fractionated with Certigrav at 2.85 gm/cc
producing concentrates of essentially liberated pyrite with
associated nonliberated coal and .mineral components. These
fractions give a reasonably direct measure of pyrite liberation
as a result of the mining system employed. The fractions are
sized by testing sieves and each size chemically analyzed for
iron, sulfur, and other components. Should it be desired, a
split of the 2.85 float sizes could be further separated at some
lower density (as 1.60) to evaluate the pyrite distribution and
concentration in such intermediate fractions. This approach
100
-------
I—,
ROM
1 1" -
16M
b v
^^* ^^ j^
^ B*« OQC ^ O Ml 1
m^m r ^ z« oD ^b t^y 1
F < — 2.85 — ^
— \
;
<
i
\/0 1/2"
Jf \J
>
^
>
J
KX 10M
1 J^lOM
F < 2.80 > S©
E -
1
Storage^ Sp
^
^ r ^ .t.oj
s© 1*1"
^ 1/2"
t
* V
' H/
-* F / 0 fiS
<
^ 10M
r
.it
^. i AM
16M
.2.80.
.-16M
60M Centrifuge
= —>2.80
'«•
Storage< Split
I
S(3J
1/4"
-16M
Figure 1. Pyrite Liberation Flow Sheet.
101
-------
t:
•^^ 80H"
2.80
Centrifuge
4,
F
fcj^
Centrifuge
> o on
Jj
1
-16
J
*^
N'-8fl-,,,%F v
-J ^ "1 r /
S
— >s
V
(j
1 ' .-*
*•*
T
R x ?
^ Centi
*
>M
k
^. 30M
^
>
i) 80M
i. SOM
.— *• 9HOM
^^
Storage^L-_ Split
@ 200M
^ 200M
'^ZrbQQVL
2.80 -.
Centrifuge
2.
Centrifuge
ti.
^
Centrifuge
400M
-35y coal
Analyze for
v pyrite sulfur
"^ microscopic examination
S
E
FJigure 1, Pyrite Liberation Flow Sheet (Continued)
102
-------
eliminates unwanted, excessive breakage of the higher quality
float particles.
The plus one-inch 2.85 float fraction was selectively and
controllably crushed with the product separated at 2.85 gm/cc
to give a pyrite liberation measure attainable in this size
range (Sink Fraction 4). As detailed in Figure 1, individual
stage crushing, sizing, and 2.85 sink fractions removal is
developed at 1/2-inch, 1/4-inch, 1QM, 16M, 30M, 80M, 200M, and
400M. After reaching the minus 10 mesh particle size, only
aliquots of the residual sample were processed to reduce the
time required.
The pyrite analyses of the several 2.85 sink fractions are
shown in Table 2 with all having more than 72.0% pyrite, excepting
Sample No. 12 (minus 35 micron) which did not respond adequately
to the gravity separation.
The minus 35 micron 2.85 float material (representing 93.2%
of the feed} contained 0.98% pyrite in contrast to 6.23% in the
head sample. This represents a rejection of 84% of the pyrite
at 2.85 gm/cc in only 6.8% of the feedl The pyrite rejection at
the several stages of liberation studied is detailed in Table 3,
with the sample numbers relatable to the flowsheet in Figure 1.
Obviously, a separation at lower (.and more conventional) densities
would give substantially higher pyrite rejections at larger
particle sizes - which were not considered for the purposes for
which this study was designed.
103
-------
Table 2
Pyrite Content in Clarion Seam
2.85 Sink Fractions
Sample No.
1
2
3
4
5
6
7
8
9
10
11
12
13
Head (Measured)
Head (Calculated)
FeS0%2 FeS,%- Range by Size
79.9
78.0
78.4
72.4
75.5
74.8
77.0
77.1
79.5
76.0
77.8
44.4
0.98
6.23
6.02
1 Size and Chemical Analyses
2 Chemical Analysis based on
76.2-83.4
60.1-83.5
77.5-86.9
65.6-84.0
66.7-79.8
63.9-80.2
65.4-84.7
75.3-80.9
77.2-83.2
74.5-81.6
74.9-82.4
44.4
on each size.
iron.
104
-------
Table 3
Pyrite Particle Liberation by
Comminution Size in Clarion Seam Coal
Cumulative %
of Total Pyrite
3.9
13.0
20.1
21.0
24.2
29.1
35.3
39.9
54.1
67.3
79.3
83.9
100.0
Sample
1
2
3
4
5
6
7
8
9
10
11
12
13
% of Total Pyrite
3.9
9.1
7.1
0.9
3.2
4.9
6.2
4.6
14.2
13.2
12.0
4.6
16.1
105
-------
Photographs of the pyrite concentrate fractions above 1 nun
show the occulsion of other components with the pyrite. Other
visual observations show the same types of associations with
smaller sized particles extending to the minus 35 micron fraction.
These associations suggest explanations for the particulate
behavior and poor performance in certain beneficiation operations
- as froth flotation.
Another application, and perhaps the most significant data
developed from this procedure, is a more realistic pyrite particle
size distribution representing the pyrite as it existed in the
coal seam. This compilation is shown in the Rosin-Rammler plot
in Figure 2.
It is not suggested that some pyrite particles were not
crushed, but the procedure does give a superior approximation
on a weight basis to the existing distribution. Detailed review
of the data permits broad application of the information for
process design.
In summary, the reduction of sulfur in coals by physical
beneficiation relates to the nature of pyrite liberation. This
liberation is highly variable and relates to pyrite particle
size occurrence-disseminations, coal component hardness variations,
and the associations between the heterogenous coal components.
Although various techniques exist, for expressing these relation-
ships and utilizing them for coal beneficiation process design for
sulfur reduction, none are fully adequate.
106
-------
9 a
2 J 4 5 6 T g J «
f i i i f lu iltm*M" 'iii^ijJr^f it'll- •>•-'*•-.•
Clarion Seam
Liberated Pyrite
SCHECM OPENING
-• 10
4OO3ZS270 ZOO MO IOC »O I SO S3 40 SO ZO _ _ _
U.S STANDARD SIEVE DESIGNATION
111 i I I I I ' I I I I I I I I I
400 323 270 700 ISO 1OO 65 48 » 28 JO 16 14 12 IO 9 t
TYLER SIEVE DESIGNATION
35 -O4
-_,,
|
1 :
>r j— -
__,
' H
r
i
! ' I
1 i
1
- »-4-
i
II
<3C
!
1
1
;
.: I
~f
,
,>
HI
\
1
jj
r j _
i •
1
.h.
J!
II
-
1
:
I
<
i
I
•
r
r —
••-
j
e
-
.
.
-
.
1
^>**
F
i
_-
r
t
•
i
-
^'
--•
—
— -
•,
~
y
—
;
-
-C
? i
|
1
l
!
r
• i
t
t
s
1
• f
'
hj_
-
£-4
IT .
I-
: !
i '
• '
t {
! 1
i
1
i j
ji i •
i1 1
i. :
'.
. j
i;
i
-t- — *• —
9 |
!
i
O"
!
1
1
' 1
1
i !
,
i^C
1 1
i i
! :
•
-
1
I
'
•
J
1
i
*p
t
p"j
!
i
1 r
J_
! 1
i i
jj
I
• i
i
!
1
i
:•
:
O 40
-sAn
. -O
1
jj !{
T •
^ '
1 ,
il
! i
Q
Ii il
H
1!
h
i •
i i
!
!
i
I ll;
!
:j
; '
Ii;
ao
•
l
^
f
r
i
i
|
M
f
l :
» ;
t ;
't.
•
;
• •
•0 ' 9,0|
h li
iu.""
"^ R
••i
• UL
' n
: ot
[T,
1 ! ::i
i
.-
: ; i
t :i:
>• t :•!
•• I M
I'M:
ilt it
^i ;: i1!
-. i . :
] i iii
. '
1 • ';i
1 i _.'.
1 • '"
:i * 'i;
iTjj !
V
n
i iii
200
1
1
p
1
l
,
:
1
1
1
1
i
i
i
•
..
:
!
!
-
l
:
i
-
si
»-•»
•so
45
SO
75
1
.
.
.
iZ
93
97
985
•• " * Irt
SCREEN OPENING WCHES
c t«r >_
«r f>M
GRAPHICAL FORM FOR REPRESENTING DISTRIBUTION OF SIZE OF BROKEN COAL
Figure 2
Sample No.
_Test No_
Sieve Analysis by -
Date
Plotted
Date.
Weight of grosi sample:
Weight of test sample .
Sieve Analysis:
Powit SIC«F
No Opfflinfl
Percrntoae j
Retamrd Cumalotrx
1
i
i
.
1
i '
OaveMmtci
Rjint j Percentage X, niches j Y.inchss '
A
| !
B I |
Difference '
|
!
i
Distribution coeff kitnt,n .
Absolute sue constant 1
-------
A concept of evaluating coal pyrite liberation is being
developed and was illustrated for a Clarion seam coal. The
procedure involves stage crushing, size separation, and removal
of the relatively pure, liberated pyrite particles. These data
may be evaluated to provide detailed pyrite particle size dis-
tribution directly by weight, to relate the extent of pyrite
liberation with comminution stage, and to provide a realistic
approach to the extent of coal pyrite rejection which is possible
and that which is commercially feasible. It is proposed such
data from many coals will show that higher levels of pyrite
rejection are feasible than are currently believed.
Modifications of the procedure seek to develop a more
simplified approach for coal seam survey purposes and to apply
the same principles to the liberation of other coal components.
108
-------
References
Cavallaro, J., M. Johnston, and A. Deurbrouck. 1976. Sulfur
reduction potential of the coals of the United States.
A revision of Kept, of Inv. 7633. Kept. Inv. 8118. U.S.
Bureau of Mines.
Davis, A. and F. Vastola. 1977. Developments in automated
reflectance microscopy of coal. J. Microscopy 109, 3-12.
Deurbrouck, A. and E. Palowitch. 1966. Survey of sulfur re-
duction in Appalachian Region coals by stage crushing.
Inf. Circ. 8282. U.S. Bureau of Mines.
Deurbrouck, A. 1972. Sulfur reduction potential of the coals
of the United States. Rept. Inv. 7633. U.S. Bureau of
Mines.
Kestner, D., D. Confer, and H. Charmbury. 1962. The effect of
crusher type on the liberation of sulfur in bituminous
coal. Coal Research Section. The Pa. State Univ., Univ.
Park, PA. Special Report 32.
Lovell, H. 1967. Recovery of pyrites from coal refuse - A
pilot plant study. Preprints of Am. Inst. Mng. Met. Engr.
Joint Session: Coal, Ind. Min., and Min. Ben. Div. Las
Vegas, Nev.
McCartney, J., H. O'Connell, and S. Ergun. 1969. Pyrite size
distribution and coal-pyrite association in steam coals.
Rept. Inv. 7231. U.S. Bureau of Mines.
Reyes-Navarro, J. and A. Davis. 1976. Pyrite in coal - Its
forms and distribution in relation to environments of coal
deposition in three selected coals from Western Pennsylvania.
Coal Research Section. The Pa. State Univ., Univ. Park, PA.
Special Report 110.
Zeilinger, J. and A. Deurbrouck. 1968. Preparation character-
istics of coal from Clarion County, Pa. Rept. Inv. 7174.
U.S. Bureau of Mines.
109
-------
GEOLOGIC CONTROLS ON MINERAL MATTER IN THE
UPPER FREEPORT COAL BED
C. B. Cecil, R. W. Stanton, S. D. Allshouse, and R. B. Finkelman
U.S. Geological Survey
Reston, Virginia 22092
ABSTRACT
Mineral matter in coal originates during various stages of coal forma-
tion. Many elements are known to be incorporated in plants and may be
retained during peat formation and coalification. Major, minor, and trace
elements may also enter swamps by sedimentological and geochemical processes
where they may be retained during the peat-forming stage. After fixation
in the peat, all elements, whether original or transported, may be mobilized
in varying amounts and precipitated as authigenic minerals (e.g., carbonates,
silicates, and sulfldes) during and after peat accumulation. The various
sedimentological and geochemical processes tend to concentrate suites of
elements and minerals within coal zones.
Elemental, mineralogic, and maceral (i.e., vitrinite, exinite, and
inertinite) associations form zones in the Upper Freeport coal of west-central
Pennsylvania. Elements that tend to be concentrated in the top and/or bottom
zones of the bed include As, Cd, Cl, Cu, Fe, Hg, Mn, Pb, S, Se, and Zn. Many
trace elements such as B, Be, Cr, Cu, F, Li, Mn, Ni, Pb, Se, V, and Zn statis-
tically correlate with Al, K, Mg, Na, and Si (elements that form the common
clay minerals in the coal) and the ash. These trace elements (B, Be, Cr,
etc.) probably entered the ancestral Upper Freeport swamp in association
with clay particles as adsorbed and absorbed ions and/or they accumulated
from degradation of plants. Some of these elements (e.g., B, Cr, F, Li, Ni,
and V) may now be associated with the clay whereas others (Cu, Mn, Pb, Cd,
Se, and Zn) were mobilized and precipitated as authigenic nonsilicate minerals.
As, Ca, Fe, and Hg do not statistically correlate with ash content because
their distribution was controlled by authigenic processes.
110
-------
INTRODUCTION
Mineral matter refers to the inorganic constituents of coal (Rao and
Gluskoter, 1973) . Mineral phases and elements other than organically bound
hydrogen, oxygen, nitrogen and sulfur are included in this definition. The
variation of mineral-matter content of coal is governed by complex geologic,
paleobotanical and geochemical processes. These processes control vertical
and lateral variations in a given coal bed as well as the size and morphology
of minerals. Vertical and lateral variation of mineral contents as well as
size, morphology, and maceral association directly affect the preparation
characteristics of the coal. For example, chemical analysis of bench-channel
samples can demonstrate that pyrite is commonly concentrated in the top and
bottom of coal beds; however, it does not provide information on pyrite
size and form or indicate how the coal and associated pyrite will respond
to coal preparation.
Fundamental to the problems of physical coal cleaning are the answers
to questions such as, what is the mineral size, degree of crystallinity,
and association with macerals, and'do these factors vary systematically
vertically and laterally in a given coal bed? The answers to these and
related questions regarding mineral matter in coal can be obtained from
geologic, geochemical and petrographic research. The data from such Investi-
gations coupled with float-sink testing can provide useful criteria for mine
planning as well as for determining the preparation characteristics.
Ill
-------
Origin of Mineral Matter in Coal
Mineral matter (i.e. contaminants) in coal can result from interrelated
processes which are operable during the peat stage of coal formation prior
to burial and from processes which are operable after burial. The preburial
processes include (but are not limited to) 1) retention of mineral matter
of plant origin in the peat; 2) deposition of detrital minerals; 3) sorption
and/or precipitation of dissolved elements carried into the peat environment
by water (either surface run off or brackish or marine waters from tides or
storms); and 4) chemical and biological activity. Fostburlal processes
Include the formation of authigenic minerals in coal and on cleat surfaces
from elements present in the peat and/or from elements derived from formation
waters and surrounding sediments.
The pH of the peat-forming environment may be of critical importance
in controlling not only the biological activity during peat formation but
fixation and/or leaching of incipient coal mineral matter. Modern peat-forming
environments have pH values ranging from 3.5 (e.g., the Okefenokee swamp of
Georgia) to values greater than 7 (e.g., the Florida Everglades) (values
measured by the writers). Feats which form In separate environments in which
the waters are chemically dissimilar probably should also have dissimilar
suites of elements; however, a great deal of research Is needed to test this
assumption. At low pH (pH 4), sulfate reducing bacteria are inactive whereas
at higher pH values these bacteria are active resulting in the production
of H2S (Zajlc, 1969, p. 8) if sulfate ions are available. This may lead to
fixation of organic sulfur or iron sulfides depending upon the availability
of sulfate and ferrous ions (Neavel, 1966).
After burial, many elements may be mobilized and preclpated as authigenic
minerals in the coal and on cleats. Common authigenic mineral types are
112
-------
silicates, carbonates, and sulfldes. The node of occurrence of these minerals
is critical to coal preparation because removal is in part dependent upon
mineral size and maceral association.
Objectives
The primary objective of this study is to evaluate the geologic controls
on mineral matter content in coal with emphasis on pyritic sulfur in coal.
A major part of this investigation is to determine relationships between
vertical and lateral variation in mineral matter content of a coal bed
reserve and how these variations effect physical preparation. The dedicated
reserves of the Upper Freeport coal bed for the Homer City, Fa., generating
station (jointly owned by New York State Electric and Gas and the Pennsylvania
Electric Company) were selected for evaluation (fig. 1).
Methods of Study
The geologic controls on the mineral matter in the dedicated reserves
are being evaluated by 1) detailed sampling and coal description of the two
deep mines in the reserves (fig. 2); 2) analysis of core logs; 3) investiga-
tion of surface exposures of the Upper Freeport in the Homer City region;
and 4) review previous work. At 21 locations (fig. 2), a total of 21
complete-channel and 75 bench-channel samples were collected from the two
mines in the reserves. These samples are being analyzed for the following:
1) ultimate-proximate analysis and sulfur forms; 2) major, minor, and trace
elements; 3) low-temperature ash mineralogy; 4) maceral composition; 5) pyrite
forms, size and maceral association; 6) qualitative determination of accessory
minerals by scanning electron microscope; and 7) ion-probe and electron-
microprobe analysis of selected major, minor, and trace element mineral
and maceral associations.
113
-------
ALLEGHENY-CONEMAUGH
CONTACT
5O
1OO MILES
Figure 1. - Index Map of Homer City, Pa. Study Area and
Allegeny - Conemaugh Contact.
114
-------
\
r\ »
>a
*•* E>>>-^ ijv /' i
x -"rjF
N\" 2 'v-
J.S{ " 4-
• *J
""•-•» '
'"*•—. v'
»
\V-
.X
-
yO
CW*
v^J
1.1
1.2
1.3
^
_JS
jfPjr \
^V 0)
flft 1
"C "2 ..--"
S.3( _ „ - - "
" x* >
*v ^
A
C^ "i-
J0-
^ "it
%>'
^
rfn
u t
— I2
1.2 }3
L _ ^^ 1
1.3' 1 LJ<
r^'V
i U-
\ " /
^
t
\
^
n ^"'
•:J L-'
/
x'"a
* a
i }•
1 5
1 /
1 B
X
—
XEf
i!
2l
t
^ ^ ^
jjO-'
1
r'
J:!
i
3
f*
£ in?
flf
j31
SEE
r ~ >
\
\
Bv^
fc o.^
. J^T.
i
* %
|
LUC-W
i
C*
^
T ^
1
J,,-
y
FT
_
i|
M
4
2.1
\
|
2.2>
• l";i
2.3 ^,
f
1
_.s
v* ,
/
/
'
,
i
y
\
\
i
V
s
—™
(•t,
,-•
t
J
-1.
__ ..
'
/*
/
^.
H- 1
N.UC
rr
i
i^i
J*
^
fp
f
/i
i
,*j
-'/"v)
r" /
. ; /
^' v
''"" \
, /
/
' "~*'-~ •*"*'*•*» ^ /
/
t/
/
6.1 f
6.2 /'
6.3 .., /
' \ /
64 / \/
/ X/
/
_...J
/*"
2-WM-6 "v
•WM2 ,' s
' »* S
1 > .'
( .<-•'
/ '•-,
<•'
->
^ J ff"
I
s (
•f
\
i
1
t
— \
-
\
\
S
\
J
•\-,
FIGURE 2
SAMPLE NUMBERS AND MAP
LOCATIONS OF COMPLETE-AND
BENCH-CHANNEL SAMPLES FROM
• r, LUCERNES AND HOMER CITY"! MINES.
vN ( \
.^ Vv_.
P \ " ,«$AMFll NUMIII 4 MAP LOCATION
N 11 /" IDINOTIS IASI OF MAJOI PAITINOI
/ « *~i i • HOAI SIN" * u*°* IAMHII N
I'V 5 "~ '•' IINCM.CMANNIl IAMPII INTIIVAL '
\ y
* — " Ul ®
""1 SHAH t IONY 9HALI
j S ^**^
/ < O-L-o
/ «»
/-4
' f 100 lOjOOMITIM
i ' *" | fj I ' p^ 1
/ 0 1000 2000 3000 4000 ft «T
/ MAP SCALE
115
-------
The relationship between the geology of the coal and its preparation
characteristics is being investigated using float-sink tests. At each of
ten of the above locations (fig. 2), channel samples of about 100 Ibs each
through the total thickness (48 inches to 83 inches) of the coal bed were
collected. At one location, 100 Ibs of coal were collected from each of
three distinct zones which comprised the total coal thickness. The 100 Ib
samples were subjected to a 21-part size-gravity study (sized to 1/4",
1/4" x 8 mesh, 8 x 100 mesh and -100 mesh with float-sink testing at 1.275,
1.30, 1.325, 1.40, and 1.80 gravities on the three larger size fractions).
All of the various size-gravity fractions were analyzed for ash, Btu, and
sulfur forms. Five selected suites of the 21-part size-gravity samples are
being analyzed in the same manner as in the complete-channel and bench-
channel samples. Mineral matter distribution in the float-sink samples can
then be related to the geologic controls on mineral matter content.
Results
From field investigations and core-log analysis the following has been
determined. The ancestral Upper Freeport peat swamp of the Homer City study
area formed on a broad flat alluvial plain. Immediately preceeding the for-
mation of the ancestral peat environment, sedimentation consisted of mixed
carbonate (the discontinuous limestone Upper Freeport and fine-grained
clastic sediment. This suggests that waters moving through the area carried
abundant calcium carbonate which may have affected pH conditions in the
i
ancestral swamp during peat accumulation. The presence of calcite (CaCO?)
in coal macerals indicates partial neutralization of acid waters of the
swamp by dissolved CaC03 species. Partial neutralization of acid waters of
the swamp is also suggested by the presence of framboidal pyrite in the
bottom bench of the coal. Certain forms of pyrite including framboids may
116
-------
result from bacterial activity during the preburial peat stage of coal
formation (A. D. Cohen, personal commun,.), Thus, the pH of the peat is
important in biological activity as well as in the fixation of mineral
matter during the peat stage of coal development.
The Upper Freeport coal bed of the Homer City study area consists of
five distinct zones (fig. 3). These zones are genetically related to con-
ditions that existed in the ancestral swamp during peat accumulation. The
mineral matter variation of these zones is the result of the processes that
were active during peat formation and of authigenic processes which were
operable before and after burial. The C zone of the coal, in general,
contains the highest ash content and the lowest total sulfur content.
Major-, minor-, and trace-element analysis of the bench-channel samples
Indicates definite genetic relationships for specific suites of minerals
and associated elements. Forty-five elements show a positive statistical
correlation with the ash content at the 95% confidence level (table 1).
The data are indicative of a common source for the elements which make up
the coal ash (excluding pyrite and CaCOj. Pyritic sulfur does not cor-
relate with the ash but correlates positively with arsenic and mercury
(table 2). Elements that positively correlate with calcium (reported as
CaO) are shown in table 3. The correlations Indicate separate genetic
relationships for CaC03, pyrite, and the ash correlative elements.
Preliminary statistical analysis indicates that there is a positive
correlation between ash content and the sum of fusinite plus semifusinte
(correlation coefficient - 0.73). The content of the inertinite group
macerals as well as the ash is greatest in the C zone.
Preliminary analysis of pyrite forms and size of the bench-channel
samples show three major distinctive types of pyrite occurrences. Zone B
tends to contain massive replacement pyrite (some of which is arsenic
117
-------
PLANT SITf
FIGURE 3
FENCE DIAGRAM OF
UPPER FREEPORTCOAL
NEAR HOMER CITY, PA
L ZONE 1
I ZONE C
L ZONE D
BONE O»
ANNEL COAL
| ] SMALf
SAMPLE NUMtEt
AND MAP LOCATIOI
.-?,' MINE PKOPEKTV
IOUNOAIY
1000 1000 1000
118
-------
bearing) which is of sufficient size and form as to be easily removed by
physical preparation. Fyrite concentrations are lowest in zone C. Zone D
commonly contains framboidal pyrite which is difficult to remove by physical
preparation because of the small size of the framboids and their association
with vitrinite.
Scanning-electron microscope and electron-microprobe analysis of the
Upper Freeport coal were used to determine maceral and mineral associations
for certain elements. These analyses show that: 1) arsenic tends to be
associated with pyrite in the B zone of the Upper Freeport coal; 2) Zn and
Cu are present as sulfide minerals; 3) much, if not all, of the Se is in a
lead selenide and lead is also present in a lead sulfide; 4) chlorine is
bound organically in vitrinite. The chlorine content of the Upper Freeporrt
is lowest in the D zone and increases through the B zone where it attains
concentration levels of approximately 0.2 weight percent; 5) the rare-earth
elements predominantly occur as phosphates and to a lesser extent are associ-
ated with silicates; and 6) chemical analyses indicate that organic sulfur
content of the Upper Freeport coal of the study area averages 0.6 weight
percent. Preliminary electron-microprobe data tend to substantiate this
value with organic sulfur in vitrinite xexinite> inertinlte (Minkin et al.,
1979). Most of the organic sulfur is in vitrinite because this is the moat
abundant maceral.
Preliminary data from the float-sink studies indicate the following:
1) much of the type of pyrite which commonly occurs in the B zone of the
coal will be eliminated by coal preparation. This will include most of the
arsenic in the coal, and 2) below specific gravities of 1.40 on a wHole coal
basis, particle size (+1/4", 1/4" x 8 m, and 8m x 100m) reduction does not
liberate those elements which correlate with ash content. At specific
gravities 1.40 and greater (float 1.6, float 1.8 and sink 1.8 Sp.G.) particle-
119
-------
size reduction liberates increasing amounts of the mineral matter. There
are known exceptions to this preliminary generalization because Li and F
tend to be concentrated in the 1/4" sink 1.8 Sp.G. fraction rather than in
the 8m x 100m sink 1.8 Sp.G. fraction. The cause of such anomolies is currently
being investigated.
Conclusions
The mineral matter content of the Upper Freeport coal of the Homer City,
Pa. study area is genetically related to a complex set of geologic, geo-
chemical and paleobotanical variables. From statistical analysis, the genesis
of the mineral matter content has been divided into 1) ash related, (excluding
the pyrite and calcite components), 2) pyrite related, and 3) calcium related.
Those elements which positively correlate with the ash (table 1) probably
accumulated contemporaneously with the peat; the most probable sources
include 1) mineral matter incorporated by plants, 2) detrital minerals and
dissolved elements which were incorporated during the peat stage of coal
formation, and 3) a combination of plant, detrital and geochemical origin.
Some of the ash-correlative elements were later mobilized and precipitated
as authigenic mineral phases (i.e., PbSe, PbS, ZnS, CuFeS2). On the basis
of preliminary statistical analysis, a genetic relationship between the coal
ash and the fusinite-semifusinite content of the coal is apparent. Fusinite
and semifusinite are generally believed to be derived from oxidized plant
material. Oxidation of organic material during the peat stage of coal for-
mation would concentrate mineral-matter constituents. Therefore, as the
fuslnite and semifusinite of the coal increases, the ash content increases.
From preliminary studies, the pyrite in the Upper Freeport is usually
concentrated in the top and/or bottom zones of the bed throughout the Homer
City dedicated reserves. Genetically, pyrite content is apparently unrelated
to the bulk of the coal ash. This conclusion is based on statistical analysis
120
-------
°J* si AI . M| «* K TL > •• !• Cd C» Co Ct C» Cu Ku » C« Cd Hf La LI Lu Mn Ho Nb Md Ml fb fr U> Sb Be S* SB Sn Sr Tb Th 0 » « tfc t» tr
IO
Off S _
" !
K. *
Tl t
*-r ,
c- .
Cd .
t-A «
Lu .
MU *
Sc »
Tb *
Tb •
Table 1, Correlations among ash related elements at the 95% confidence level,
C+) indicates a positive correlation; (-) indicates a negative correlation;
G.) indicates there is no correlation
-------
§ 3
W ft
H H«
O ^
ca 2 n0
e" w isj ff 10
M C O O O
% $ E? As Co Ge Hg Mn Sr
M ^
Oxygen - ---+ . .--- +
Total Sulfur - +++. +. + + +
Eyritic Sulfur -+ ++_ + + + + +_
MnO + + + . + + . +
P2°5'' + • - H
•
AQ 4- 4- 4. -I- 4-
na .TTT... «T^T^« i
Co ..+++. . + + +,
Ge - + +++. ++ + + -
Hg _ + ++.. 4. + + ..
vf L ^ «t 4, J. A 4. «!•
nn *r T T T . , T T , ,
Sr +--..+ ..--.
Table 2. — Correlations among pyritic sulfur related elements at the 95%
confidence level. (+) indicates a positive correlation;
(-) indicates a negative correlation; (.) indicates no
correlation. .
122
-------
CaO Fe203 MnO P^ S03 Mn Mo Sr Th Tl
CaO
Fe2°3
MnO
P2°5
S03
Mn
Mo
Sr
Th
Ti
Table 3. — Correlations among calcium (reported as CaO) related elements at
the 95% confidence level. (+) indicates a positive correlation;
(-) indicates a negative correlation; (.) indicates no correlation.
123
-------
as well as vertical and lateral variation of the pyrite and ash throughout
the coal reserve. The pyrite which formed during the preburial peat stage
of coal formation is probably limited to framboids which may be difficult
to remove by physical coal preparation. The pyrite in the top benches
(A and B zones) probably formed after burial but before compaction was
complete. Much of the pyrite in the A and B zones is relatively large and
can be removed by physical preparation. These conclusions on preparation
characteristics of the various types of pyrite are currently being evaluated
on float-sink samples.
The calcium in the coal bed is primarily in authigenic calcite in the
coal and in cleats. Much of the calcium may have been fixed during the
peat stage by ion exchange and/or as calcium salts of humic acids. Libera-
tion of CC>2 and organically bound calcium during coallfication may have
resulted in the formation of calcite.
The origin of the calcium in the coal may be highly significant. The
buffering effect of calcium-carbonate species supplied to the ancestral
peat may have aided in the retention of much of the ash-correlative mineral
matter as well as in permitting anerobic bacterial generation of sulfide
species. Low pH may cause leaching of mineral matter whereas higher pH
values may favor retention.
Preparation characteristics for coal from a given coal-bed reserve are
predetermined by the geologic history of the bed. From the data obtained
thus far in the present investigation, it seems assured that the geologic
controls on coal quality and preparation characteristics can be determined.
If this is true, then basic geologic research will rapidly become an inte-
gral part of resource assessment, and coal petrology will be extensively
used in preparation plant-quality control.
124
-------
REFERENCES CITED
Mlnkin, J. A., Chao, E. C. T., Thompson, C. L., 1979, Distribution of elements
in coal macerals and minerals: Determination by electron microprobe
(abs.): (submitted to the American Chemical Society National Meeting
Program, April 1979).
Neavel, R. C., 1966, Sulfur in coal: its distribution in the seam and in
mine products: Pennsylvania State University, University Park, Pa.,
unpub. Ph. D. dissertation, 351 p.
Rao, C. P., and Gluskoter, H. J., 1973, Occurrence and distribution of
minerals in Illinois coals: Illinois Geol. Survey Circ. 476, 56 p.
Zajic, J. E., 1969, Microbial Biogeochemistry: New York and London, Academic
Press, 345 p.
125
-------
INTERPRETING STATISTICAL VARIABILITY
Ralph E. Thomas
Battelle
Columbus Laboratories
505 King Avenue
Columbus, Ohio 43201
ABSTRACT
The variability of averages of coal characteristics as a function of
sample size is considered in this paper. An exploratory application of the
methods of geostatistics is made to 1970 data on Ibs S02/million Btu for
the Helen mine. The application is made subject to several important quali-
fications including, for example, the requirement that successive mining days
are equivalent to sampling at uniformly spaced locations along a straight
line in a coal seam. The resulting empirical variogram is found to be well-
fitted by the standard Matheron spherical scheme. A formula is derived for
using either the empirical or fitted variogram to compute the predicted
variance of an average based on n successive daily measurements. For the
data examined the predicted variances as a function of sample size show
good agreement with observed variabilities. These results suggest that
the methods of geostatistics may have important applications to a wide
variety of coal sampling problems.
126
-------
INTRODUCTION
In order to assess the quality of coal it is necessary to
draw samples and make a variety of laboratory measurements. Such
measurements frequently include determinations of the weight per-
centages of sulfur, ash, moisture/ and Btu's per pound. These
measured characteristics of coal are often subject to consider-
able statistical variability in repeated samples. In addition to
the inherent variability due to inhomogenieties within a coal
seam, other possible sources of variability include the location
of the sample, the size, shape, and orientation of the sample,
handling and processing methods, laboratory and analytical
procedures, etc.
127
-------
In general, statistical methods are required to treat data
showing large amounts of statistical variability. Such methods
typically treat the data in terms of averages (as measures of
"signals") and variances or standard deviations (as measures of
"noise"). Of special importance in sampling procedures is the
2
variance of a mean based on n observations o_ and its relation
2 Y(n)
to the variance of the measurements o of the individual measure-
ments. In its simplest form this relation is given by the
expression:
2 2
o_ = o /n, (1)
y(n) y
where n denotes the number of observations in the sample mean y(n).
This relation indicates that the sample mean is statistically better
behaved (less noisy) than the measurements of which it is composed.
In fact, if the sample size is increased by a factor of 2, then
Equation (1) shows that the variance of the mean of the n measure-
ments is decreased by a factor of 1/2. For classical statistics
this relation is fundamental to virtually all problems involving
sample size. Once the variance of the individual measurements is
known (or estimated), and the desired precision of the average
is specified, then Equation (1) is frequently used to determine the
sample size to yield the required precision.
128
-------
It is well known that the fundamental relation given by
Equation (1) frequently does not hold for coal data. In general,
increasing a sample size by a factor of 2 will not yield averages
that are twice as precise. Sometimes it is found that the pre-
cision of an average is hardly increased at all by doubling the
sample size.
Figure 1 shows a plot of unpublished data for pounds of sulfur
per million Btu for successive mining days for the Helen mine,
January, 1970, through December, 1975. In general, the plot shows
considerable statistical variability superimposed on more slowly
varying drifts over time.
Table 1 shows measures of statistical variability for averages
based on different sample sizes for data shown in Figure 1. The
second row of the table pertains to averages of pounds of sulfur
per million Btu averaged over 5 successive mining days. Column 3
shows that a total of 245 such averages were computed over this
6-year time period. The standard deviation of these 5-day averages
is shown to be 0.480 pounds of sulfur per million Btu. The co-
efficient of variation in percent (relative standard deviation)
is given in Column 5 by the ratio of entries in Column 4 to those
in Column 2. The coefficient of variation for means based on an
aggregation interval of 5 days is seen to be 21.1 percent. In
other words, the variability associated with such an average is
approximately 21 percent of the average value itself.
129
-------
o
o
00
O
o:
LU
Q_
CO
CO
Q
O
Q_
0.000
2.000
4.000 6-000 8.000 10-000 12-000
MINING DAYS (X) (X10 2)
14.000
16-000
FIGURE 1 POUNDS OF SULFUR PER MILLION BTU FOR SUCCESSIVE MINING
DAYS. FOR HELEN MINE, JANUARY 1970 THROUGH DECEMBER 1975
-------
TABLE 1. MEASURES OF VARIABILITY FOR POUNDS OF SULFUR PER MILLION BTU
AVERAGED OVER SELECTED INTERVALS OF SUCCESSIVE MINING DAYS
Aggregation
Interval, Days
Overall Number of
Mean Aggregated Means
Standard Deviation
of Aggregated
Means
Coefficient
of Variation,
Percent
Helen Mine: January 1970—December 1975
1
5
20
113
226
2.28
2.28
2.28
2.29
2.32
1225
245
62
11
6
0.507
0.480
0.459
0.418
0.406
22.2
21.1
20.1
18.3
17.5
-------
An examination of the results shown in Table 1 indicates that
a small reduction in the standard deviation of the aggregated means
occurs as the aggregation interval increases approximately through
daily, weekly/ monthly, 6-months, and yearly time periods. These
increases in the time interval of aggregation are represented
approximately by factors 5, 4, 6, and 2 for the successive rows
of the table. In spite of these relatively large increases in the
size of the aggregation interval the reductions in the standard
deviations are seen to be relatively small.
In general, the results shown in Table 1 show that the standard
deviation of a mean does not decrease in accord with Equation (1).
The fourfold increase in sample size from 5 to 20 days, for example,
would be expected to decrease the 5-day standard deviations from
0.480 to 0.480//4*= 0.240 Ibs S/million Btu. Instead of a 50
percent decrease in the standard deviation, the decrease is seen
to be only 4.4 percent.
It is not difficult to propose possible explanations for the
wide discrepencies between actual data and results predicted by
Equation (1). The principle explanations include the fact that
coal is an inhomogeneous material. Thus, averaging over an increased
sample size may, in fact, include coal from different statistical
"populations". The standard deviation of the resulting average
may not decrease because the sample would then consist of a mix-
ture of different populations of coal, rather than a single popula-
tion as required by Equation (1). In addition, it is possible that
coal samples taken on successive days are highly correlated, so
that an unusually high value for sulfur on one day may frequently
132
-------
be followed by an unusually high value on the next day. Equation
(1) does not account for such correlations.
A number of investigators have attempted to sort out the
complexities of the variability associated with coal measurements.
In every instance known to the author an empirical approach has been
taken that, in effect, requires the generation of a curve that relates
the precision of an average to the size of the sample (weight).
Once such a curve is generated, the sample size required to give
a specified precision can then be estimated by interpolation.
Deeper explanations of empirically attractive procedures frequently
result in controversy (Visman, 1969. Duncan, 1971. Visman, Duncan,
Lerner, 1974.).
A more fundamental approach to statistical variability has
been developed by Matheron (1963,1965,1967) under the name
"geostatistics". These techniques have been developed for mining
engineers and have been primarily applied to obtain estimates of
ore reserves in a variety of foreign countries. To date applica-
tions of the methods of geostatistics in the United States are
relatively limited.
Because of the possibility that the method of geostatistics
would permit a better treatment of statistical variability of coal
measurements, an exploratory application of basic geostatistical
concepts has been made to coal data. Before presenting these
results a brief overview of a few of the central geostatistical
concepts is given below. More detailed descriptions can be
found in the references listed in a recent Bibliography on geo-
statistics (Alldredge and Alldredge, 1978); the publications of
Royle (1977) are especially recommended.
133
-------
A Geostatistical Approach
The primary distinction between the methods of geostatistics
and conventional statistics is the fact that a sample is char-
acterized both by its measured value and by its sampling location,
especially its location with respect to all other sample locations.
The sample may be further characterized by its size (volume),
shape, orientation, etc. Such information is used to interpret
a graphical plot called a variogram that/ in turn, is frequently
used to partition the overall variability of measured values into
two components.
One geostatistical component of the variance is called the
"nugget" variance. This variance, symbolized by C
-------
variance among the measurements frequently fluctuates about a
maximum level symbolized by C0+C.
The first objective of a geostatistical analysis consists of
generating the variogram. By applying fitting techniques to the
variogram it is then frequently possible to determine quantitative
estimates of C0,C, and a. The variogram is constructed from a set
of measured values: y1,y2,...,yn, ideally taken at uniformly spaced
coordinates: x1,...,xn, along a straight line. The symbol h is
used to denote the uniform distance between adjacent measurements.
With uniform spacing, sampling points may be separated by distances:
h,2h,...,kh,etc. The variogram function represented by the plot
is usually symbolized by y(k). The value of y(l) is first
computed for all pairs of measured values taken at adjacent sampling
locations separated by a distance equal to h; y(2) is then computed
for all pairs of measured values separated by a distance equal to
2h, etc., in accord with the following formula:
n-k _ _ 2
1
Y
-------
• the successive daily samples taken from run-of-
mine coal can be treated as though they were
obtained along a straight line at uniformly
spaced locations in the coal seam.
• All samples have the same weight, volume, orienta-
tion, etc. within the coal seam.
Both of these assumptions are likely to be incorrect. To the
author's knowledge, the daily samples may well represent different
mixtures of coal taken from different mining faces, different
weights, volumes, orientations, etc. In spite of these diffi-
culties, it is found that many of the computed variograms for
coal data yield rather well-defined values for C0,C, and a.
Figure 2 shows an example of a variogram for pounds of SO,
per million Btu for daily samples of coal obtained from the Helen
mine for 1970. The variogram extends over pairs of mining days
separated by lags of 1 mining day, 2 mining days, etc., up to
and including lags of 73 mining days. In other words, a mining
day is taken to represent the average distance mined in a coal
seam in one day. The variogram is seen to increase in a roughly
linear manner from a small nugget value of approximately 0.15.
The variogram then levels off at a value around 1.20. The
leveling-off appears to occur for sampled values separated by
approximately 36 mining days. By such a cursory inspection of
the variogram. It is seen that the values of C0,C, and a are
approximately 0.15,1.20, and 36, respectively.
A variety of models have been used to fit variogram data.
When the graphically estimated values of C0,C, and a are rather
136
-------
o
o
u>
O
§
n
o
o
o
*+jfc*-*r>
Y(x) = 0.15+1.20 [(3/2)(x/54)-(l/2)(x/54)3]
IBS S02/MILLION BTU
0.000 9.000 18-000
27.000 36.000 45-000 54.000
MINING DAYS (X)
i I I
63.000 72-000 81.000
FIGURE 2. VARIOGRAM FOR HELEN MINE, 1970
-------
well-defined as in Figure 2, the "spherical" scheme of Matheron
(1963) often yields a good fit. When a model is fit to the vario-
gram, the distance x between sampling locations can assume all
values in a continuum/ instead of being restricted to an integral
number of lags. Thus, the variogram function fitted to the data
may be symbolized by Y(X) instead of v(k). For the variogram of
Figure 2 the fitted spherical model is given approximately by the
equation:
Y(x) = C0+C [(3/2) (x/a)-(1/2) (x/a) 3] , (3)
with C0=0.15/ C=1.20, a=54, for x/a less than 1.0; when x/a exceeds
1.0, Y (x) is taken to be equal to C0+C. The range value of 54 is
obtained from the previous graphical estimate of 36 by multiplying
36 by (3/2) in accord with standard estimation procedures associated
with fitting a spherical model (Royle, 1977). The solid curve
shown in Figure 2 represents the fitted variogram associated with
Equation (3). For most applications it is important to have a good
fit for small lag values; for the larger lags, especially those
beyond the range, the general level is important, but usually not
the detailed fit. On this basis it is seen that a good fit is
provided for these data using Matheron's spherical scheme.
The variogram is the basic geostatistical tool for answering
virtually every question regarding the precision of averages obtained
from statistical sampling. Questions involving the required number
of samples, required sample volumes, sample shapes, orientations,
etc., must all be referred to an appropriate variogram. For the
purposes of this paper, however, we confine our attention to the
138
-------
problem discussed earlier: How does the variability of an average
of measured coal characteristics depend on the number of measure-
ments included in the average? In particular, can the methods of
geostatistios yield a better relationship than that given by
Equation (1)?
A Formula for the Variability of an
Average of n Successive Measurements
Based on geostatistical concepts the variability of an aver-
age of 2 measurements cannot be determined until the distance
between the two sampling locations is specified. In geostatistics
the distance x between a pair of sampling locations is usually measured
in units of the range as indicated/ for example, by the (x/a) terms
shown in Equation (3). The variability of a sample mean based on
two coal measurements taken 1 mining day apart may generally be
expected to be different from that which would be obtained if the
measurements were separated by 100 mining days. In general,
variabilities of averages are expected to be larger whenever the
sample locations lie within their mutual zones of influence,
smaller when they do not. The variogram is required to identify
and quantify the various possibilities, especially when some
sampled locations lie within mutual zones of influence and others
do not.
The formula proposed below for computing the variance of an
average coal characteristic is based on geostatistical methods.
However, the proposed formula has not been identified in the
limited geostatistical literature available to the author. Conse-
quently, the reader is cautioned to be critical of the following
development, especially with regard to the treatment of the nugget
variance. 139
-------
The proposed formula is first illustrated for an average of
2 measurement y, and y0 taken at sampling locations separated by
X *
k lags. A lag matrix is given as follows:
Xl X2
» i
0 k
k 0
• i
where the entries in the matrix indicate that a lag of k exists
between x1 and x2 and lags of zero exist between each point and
itself. The Y-values corresponding to these lags are then obtained
from the variogram and the following corresponding matrix of Y-values
is formed:
"Y (0) Y(k)
Y«0 Y(0)
• i
where y(0) is set equal to the nugget variance C0, and yOOis the
value obtained from the plot of the variogram at a lag equal to k.
The variability of an average of two successive measurements y,
and y2 separated by a lag of k is then given by the following
formula:
_2
y(2)
- 2c0+c-r0,
where
If the two sample locations are separated by a distance at least
equal to the range, then y(k) * Y(») * C0+C, and with y(0) " C0,
o
it is found that o • C0+(C/2). If the two samples are arbi-
y(2) 2
trarily close together, then y(k) = y (0 ) = C0 and o = C0+C.
y(2)
All other cases lie between these two extremes, so that:
C0+(C/2)«J2
-------
The same procedure is next applied to obtain the variance
of an average of n measurements yi'v2"*"vn assumed to be taken
on successive days at sampling locations x,,...,x uniformly
spaced on a line. The variance of the mean is given by the
formula:
2
= 2cn+c-r
y(n)
n'
(4)
where rn denotes the arithmetic average of the variogram
values shown as entries in the following matrix:
Y(0)
Y(D
Y(2)
Y(n-l)
Yd)
Y(0)
Yd)
Y(n-2)
Y(2)
Yd)
Y(0)
Y(n-3)
Y(n-l)
Y(n-2)
Y(n-3)
... Y(0)
n
If the uniform separation between adjacent sample locations exceeds
the range, then the entries down the diagonal of the T matrix
are given by Y(0)»C0 and all off-diagonal terms are given by
Y(»)=C0+C. The arithmetic average of these Y-values is then
found to be given by F-Co+d- (1/n) )C, and from Equation (4) the
variance of the sample average is seen to be
2
y(n)
- 2C0+C-C0-(l-d/n))C
with the result that
y(n)
C0+(C/n)
Again, if all sample locations are arbitrarily.close together/
then all entries in the r-matrix are equal to C0 and it follows
2
that a - C0+C. Thus, in every case the variance of the mean is
x(n)
bounded as follows:
C0+(C/n)
-------
Except for the nugget variance C0 the upper and lower limits
would correspond to the usual statistical case provided C is set
2
equal to the variance of the individual measurements 0 . In
geostatistical terms, however, it would appear to be more appro-
priate to take the variance of individual measurements to be
increased by the nugget variance:
For sample locations within their mutual zones of influence,
the average value of the variogram values in the r-matrix will
decrease, and the variance of the average will then increase in
accord with Equation (4) .
A more formal derivation of Equation (4) is given in Appendix
A; an efficient algorithm for computing r is given in Appendix B.
The derivation of Equation (4) makes use of a basic result in
classical statistics for the variance of a linear combination of
correlated random variables (Hald, 1952) . The derivation shows
a close formal relationship between methods of geostatistics and
those of classical statistics for correlated variables. In this
way it becomes evident that the methods of geostatistics augment
those of classical statistics for correlated variables, primarily
by introducing C0 to measure local inhomogeneities and by using
the variogram to measure the effects of correlations at specified
sampling distances.
142
-------
An Application to Coal Data
Figure 3 shows the results obtained by applying Equation (4)
to the variogram for pounds of S02 per million Btu obtained for
the Helen mine for 1970-1975 as shown in Figure 2, Figure 3 shows
the variance of of. relative to 0^ in accord with the following
y(n) y
expression:
R = 02 / 2 m (2c+C0-r )/(c0+C) (5)
y(n) X
for n varying between 1 and 20 successive measurements. The lower
curve shows the theoretical decrease in the relative variance of
a mean if the n observations are statistically independent. The
uppermost curve shows the predicted decrease in the relative
variance of the mean value of Ibs S02 per million Btu using the
fitted variogram, shown in Equation (3), for the 1970 sulfur data
for the Helen mine. The broken curve shows the relation obtained
when the empirical variogram values are used in Equation (4)
instead of those obtained from the fitted smoothed curve shown
in Figure 2. The data points shown in Figure 3 correspond to the
relative variances computed for Helen data 1970-1975.
In general, the agreement between the actual and predicted
values appear to be rather good for the two curves based on the
variogram. For small values of n the results suggest that the
actual values of the empirical variogram may give better results
than the smoothed variogram; for large values of n, there is
little difference between the results obtained using the empirical
and smoothed variograms. In either case the variogram provides
much improved predictions for the variance of a mean as a function
143
-------
Relative Variance,
R =
-------
of sample size than that provided by the assumption of statistical
independence.
APPENDIX A
Let y-,/...,y denote n successive measurements taken at
uniformly spaced sampling locations x,/...,x along a straight
line. If the variogram of the measurements can be represented
by a Matheron scheme:
Y(x) = C0+C [ (3/2) (x/a)-(l/2) (x/a) 3
where x denotes the distance between a pair of sampling loca-
tions, and C0,C, and a denote fitted parameters, then the vari-
ance of the arithmetic mean of the y-values is given by the
following formula:
a^_ = 2C0+C-T",
y(n)
where 7 denotes the arithmetic mean of the Matheron Y-values
2
associated with the n distances between sampling locations for
all possible pairs of y-values.
To obtain this formula we first note that the expected value
2
of the variogram function is given by y CO " o (1-p(k))+C0, where
C0 has been added to provide a residual nugget variance. The
average value of y(k) over all lags associated with all pairs of
_ 2
sampling locations is then given by r = cr (l-p)+C0. From the
—•2 2
statistics of correlated variables p = a /a , and this sub-
2 2 - y(n)
stitution yields a = C0+o -r. The desired result is then
y(n)2 y
obtained by putting a » C0+C.
145
-------
APPENDIX B
Let rn represent the sum of the entries in the following
matrix of y-values:
Y(0)
Yd)
t t t
_ Y(n-l)
Yd)
Y(0)
Y(n-2)
Y(n-l)
Y(n-2)
Y(0)
By considering the sub-matrix associated with r , it is seen that
where
,
n-i
n - -Y (0) ,n=2, 3, . . . ,
is defined to be equal to Y(0). This relation then permits
rn-l+2
the sums r2,r3/... to be computed recursively. The required aver-
ages are then given by
n
rn/n ,n=l,2,... .
REFERENCES.
Alldredge, J. R. and N. G. Alldredge. 1978. Geostatistics: a
bibliography. International Statistical Review, 46:77.
Duncan, A. J. 1971. Comment on general theory of sampling.
Materials Research and Standards, MTRSA, 11:25.
Hald, A. 1952. Statistical theory with engineering applications.
John Wiley & Sons, Inc. New York, 771.
Matheron, G. 1963. Principles of geostatistics. Economic
Geology, 58:1246.
Matheron, G. 1965. Les variables regionalisees et leur estima-
tion. Paris, Masson & Cie. 305.
Matheron, G. 1967. Elements pour une theorie des milieux poreux,
Paris. Masson & Cie. 164.
146
-------
Royle, A. G. 1977. Practical introduction to geostatistics.
Department of Mining and Mineral Sciences, University of Leeds,
103.
Royle, A. G. 1977. Exercises in geostatistics. Department of
Mining and Mineral Sciences, University of Leeds, 115.
Royle, A. G. 1977. Geostatistical tables. Department of Mining
and Mineral Sciences, University of Leeds, 70.
Visman, J. 1969. A general sampling theory. Materials Research
and Standards, MTRSA, 9:9.
Visman, J., A. J. Duncan, and M. Lerner. 1974. Further discussion:
a general theory of sampling. Materials Research and Standards,
MTRSA, 11:32.
147
-------
AN OVERVIEW OF EPA COAL CLEANING PROGRAMS
J. D. Kilgroe and D. A. Kirchgessner
U.S. Environmental Protection Agency
Office of Energy, Minerals and Industry
Industrial Environmental Research Laboratory
Research Triangle Park, North Carolina 27711
ABSTRACT
Recent environmental regulations have significantly altered the rules
under which pollution control technologies must compete. Potential applica-
tions of coal cleaning and the resulting research and development goals have
been affected.
Although the more stringent S02 emission standards being placed on some
sources effectively preclude coal cleaning as a sole means of compliance,
recent studies indicate that coal cleaning in conjunction with flue gas
desulfurization can provide substantial cost benefits in certain applications.
Coal preparation has the additional advantage of removing some harmful trace
elements from the coal prior to combustion.
In order to evaluate and enhance the position of coal preparation in
the pollution control technology mix of today and the future, EPA maintains
a three-part program. Its goals are: (1) to assess and develop coal cleaning
processes; (2) to assess the environmental impact of coal cleaning; and
(3) to develop pollution control technology for coal cleaning processes.
149
-------
INTRODUCTION
Two major goals of the National Energy Policy are the expanded
use of domestic coal supplies to replace imported oil, and abate-
ment of the adverse environmental impacts which result from coal
use. The legislation which gives substance to the second goal is
contained in the Clean Air Act Amendments of 1977, Federal Water
Pollution Control Act Amendments of 1977, The Resource Conservation
and Recovery Act of 1976, and The Toxic Substances Control Act of
1976.
In recent years coal cleaning, once used solely to remove
mineral matter from coal, has been recognized as a viable method
of removing sulfur from coal prior to combustion. The technique
is so successful that it is gaining recognition as an efficient
and relatively inexpensive method of making additional supplies
of environmentally acceptable coal available. The remainder of
this paper will develop a historical perspective to EPA's coal cleaning
program, an outline of the regulations to which it is responsive,
objectives of the program, and some significant results which have
been achieved.
150
-------
BACKGROUND
Although a Federal program of research in air pollution was
initially authorized in 1955, no substantial effort was put forth
in the area of coal cleaning until the passage of The Clean Air
Act in 1963. This Act called for an expanded Federal research
and development program with special emphasis on the investigation
of sulfur oxide pollution from the burning of coal and fuel oil.
In response to this mandate, the Department of Health, Education
and Welfare, the predecessor of EPA, began to study coal cleaning
as a means of sulfur oxide control in 1964.
The Air Quality Act of 1967 gave an additional impetus to
the effort by directing HEW to establish the National Center for
Air Pollution Control, formerly known as the Division of Air
Pollution of the Public Health Service. At the Center coal
cleaning studies continued and began to assume many of the aspects
which characterize the present program (U. S. Congress, 1968).
In 1970 legislation was passed which established the Environmental
Protection Agency. Today the coal cleaning research program which
was initiated in 1964 by HEW is being carried on by the Fuel Process
Branch of EPA's Industrial Environmental Research Laboratory at
Research Triangle Park, North Carolina.
151
-------
REGULATIONS
The following portions of the more important environmental
regulations are discussed briefly, either because they refer to
a pollution problem which coal cleaning can ameliorate, or because
they place restrictions on pollution from the coal cleaning tech-
nology itself. EPA's research and development activities under
the coal cleaning program are periodically modified to make them
responsive to changing regulatory requirements and energy goals.
Air
In accordance with the Clean Air Act of 1970, EPA has set
primary and secondary ambient air quality standards to protect
public health and welfare, respectively. Regulated pollutants
directly related to the use of coal include sulfur oxides, nitro-
gen oxides and particulate matter.
Section 111 of the Clean Air Act requires EPA to promulgate
emission standards for new stationary sources. As originally
promulgated, these New Source Performance Standards were emission
limitations and could be met with any control device or system
(40 CFR 60, 1976). The Clean Air Act Amendments of 1977 sub-
stantially altered the format of these standards. New stationary
152
-------
sources must now: 1) use best available control technology;
2) use a method of continuous pollution control; and 3) achieve
a specified percentage reduction of regulated pollutants from
fossil fuel-fired units. While these new regulations by themselves
do not affect the applicability of coal cleaning, the definition
of best available control technology and the percentage reductions
which are assigned to the various pollutants could significantly
restrict the use of cleaned coals as a sole means of compliance.
To comply with the intentions of the new regulations, EPA
has proposed revised New Source Performance Standards for utility
boilers in the future. The revisions will require: 1) an 85
percent reduction in sulfur between extraction and stack
emissions; 2) that sulfur oxide emissions not exceed 1.2 Ib*
SO2/10 BTU of heat input; 3) no further control if emissions
are 0.2 Ib SO2/106 BTU.
Background studies in support of industrial boiler NSPS are
currently underway. The format of the regulations, if not the
emission and reduction levels, will be similar to those for utility
boilers. The extent of coal cleaning's role as a means of compliance for
utility and industrial boilers is not presently known.
New Source Performance Standards for air pollutants also
apply to coal preparation plants and coal handling facilities.
Present regulations restrict particulate emissions from thermal
driers and handling facilities to 20 percent opacity. Pneumatic
cleaning equipment may not exceed 10 percent opacity (40 CFR 60,
1976).
In addition to the more conventional pollutants generally
discussed, the Clean Air Act of 1970 requires EPA to
Conversion factors for metric equivalents of non-metric units used
in this paper appear near the end of the paper «"«» usea
153
-------
establish a list of "hazardous pollutants" and to propose emission
standards for them. Both mercury and beryllium are emitted from
coal-fired boilers and are subject to these standards.
The Amendments of 1977 specify four unregulated pollutants
which EPA must investigate and, if necessary, regulate. These
are arsenic, cadmium, polycyclic organic matter and radioactive
pollutants. No determination has yet been made in this area. It
is known, however, that the coal cleaning process partitions
trace elements among the various size-gravity fractions. There
is a high probability that coal cleaning can serve as a pre-com-
bustion removal method for some of these elements.
Water
Federal control of industrial water pollution sources is
achieved through the issuance of discharge permits under the
authority of the Federal Water Pollution Control Act of 1972.
These permits stipulate the concentrations of various pollutants
allowed in the effluents. Effluent limitations are presently
based on the Best Practicable Control Technology currently avail-
able (BPT), and must be based on the Best Available Technology
Economically Achievable (BATEA) by 1985.
Mew Source Performance Standards on effluents from categories
of industrial sources are required by Section 306 of the Act
and intended to be the most stringent of the three levels of limi-
tations. Coal preparation plants are in a category for which NSPS
have been proposed (40 CFR 434, 1977). Tables 1 and 2 show
the proposed limitations xor facilities having acid and alkaline
effluents prior to treatment. Plants which do not recycle waste
water are subject to a "no discharge of pollutants" limitation.
154
-------
Effluent Limitations
Average of Daily
Value for 30
Effluent Maximum for Consecutive Days
Characteristic Any 1 Day, mg/1 Shall Not Exceed/ mg/1
Iron, Total 3.5 3.0
Manganese, Total 4.0 2.0
TSS 70.0 35.0
PH Within the range 6.0 to 9.0
Table 1. Limitations for acid effluents.
Effluent Limitations
Average of Daily
Value for 30
Effluent Maximum for Consecutive Days
Characteristic Any 1 Day, mg/1 Shall Not Exceed, mg/1
Iron, Total 3.5 3.0
TSS 70.0 35.0
pH Within the range 6.0 to 9.0
Table 2. Limitations for alkaline effluents.
155
-------
These regulations, of course, take a fairly simplistic view
of what constitutes a pollutant in industrial effluents. This
fact was recognized by several environmental groups and in 1975
EPA faced court action by these groups for not having made a thorough
assessment of the pollutants discharged into surface waters
by industry. The courts concurred with the environmentalists
and as a result, EPA was ordered to reassess the BAT effluent
limitations in view of a list of 129 specific "priority
pollutants."
In the course of this review, 16 coal preparation plants
were sampled for concentrations of priority pollutants. Tenta-
tively, 24 of the 129 pollutants have been identified in effluent
streams from preparation facilities and ancillary areas. Those
pollutants which consistently appear in significant quantities
can be expected to be regulated eventually.
Solid Waste
Solid wastes such as those generated by coal preparation
have not traditionally fallen under Federal control with respect
to the quantities disposed of or the specific means of disposal.
In October of 1976, however, Congress passed the Resource Conser-
vation and Recovery Act (RCRA). Although the ramifications for
coal combustion and coal cleaning wastes'are not yet known, a
determination that,these products are hazardous wastes would sub-
ject them to the most stringent applications of the law.
PROGRAM OBJECTIVES
With this historical perspective and regulatory framework
in mind, EPA has structured a program with three objectives:
1) to assess and develop coal cleaning processes; 2) to assess
the environmental impact of coal cleaning; and 3) to develop
pollution control technology for coal cleaning processes.
156
-------
Table 3 outlines the present interagency program which EPA
is funding. The budget for fiscal 1978 was approximately
eight million dollars.
PROGRAM OUTPUTS
Technology Assessment
Improved techniques for the preparation of fine coal are
needed to enhance sulfur removal and coal energy recovery. The
primary objectives of the technology assessment and development
activities are to evaluate the potential cleanability of U.S:
coals, and the performance and costs of commercial equipment
which can be used for the beneficiation of fine coal. The develop-
ment of chemical coal cleaning processes is supported, as is the
applied research necessary to characterize the basic mechanisms
which govern beneficiation processes.
Assessment of Coal Cleaning for SO, Emission Control
The 1977 Clean Air Act Amendments provide a new impetus for
evaluating coal cleaning as an S02 emission control technique.
The regulations mandated by this legislation have significantly
altered the positions of competing technologies. Studies are
in progress to assess the applicability of coal cleaning as a
means of S02 control with respect to other control technologies
for: 1) existing boilers regulated under SIPs; 2) NSPS for
coal-fired steam generators; 3) revised NSPS for utility boilers;
and 4) NSPS for industrial boilers. Preliminary results from
these studies suggest that:
157
-------
•oaf. 3. Active Interagency Goal Cleaning Projects (1977-1978)
Project Title (Contract, Grant, or
Interaeency Agreement)
Organization
Directing
Work
Organization
Performing
Work
Objectives
TECHNOLOGY ASSESSMENT AMD DEVELOPMENT
Coal Cleanablllty (IAG-D6-E685)
Coal Cleaning Technology As
and Development (68-C2-2199)
Interim Support for Homer City
Test Program (68-02-2639)
01 Dense Media Cyclone Pilot Plant (IAC-D6-E685)
oo
Demonstration of Coal-Pyrite Flotation
(IAG-D6-E685)
Adsorption-Desorptlon Reactions in
Pyrlte Flotation (IAG-D6-E685)
High Gradient Magnetic Separation
(IAG-D5-E68S)
Surface Phenomena in Dewatering of
Fine Coal (IAG-D6-E68S)
Coal Cleaning Test Facility
(IAG-D6-E685)
DOE
EPA
.(a)
(b)
EPA/DOE
DOK(a)/EPU/
PEHELEC/EPA
DOE
(a)
DOE
(a)
DOE
(a)
DOE
(a)
DOE
(a)
(a)
DOE
Versar, Inc.
Chen Systems/Penn-
sylvania Electric Co.
(a)
DOE
Heyl and Patter-
son Co. /Barnes and
Tucker Co.
university of Utah
General Electric Co.
Syracuse Univ.
Blrtley Engineering
Corp. Williams, Tre-
bilcock and Hhitehead
Determine desulfurization potential of U.S.
coals by size reduction and specific gravity
separation.
Evaluate performance and costs of equipment
for removing sulfur from coal.
Provide teat planning and Initial test
support for the Homer City Coal Cleaning
Demonstration Program.
Evaluate effects of cyclone design and oper-
ation variables on separation of fine coal and
pyrlte.
Commercial testing and operation of two stage
coal-pyrite flotation process developed by
DOE1".
Evaluate the adsorptlon-desorption mechanisms
ufalch control performance in the DOE two
stage coal-pyrite flotation process.
Evaluate technical feasibility of high gradi-
ent magnetic separation for removing pyrlte
from coal.
Evaluate phenomena governing the effectiveness
of surfactants in reducing the final moisture
content of coal vacuum filter cakes.
Design a DOE coal cleaning test facility. Pro-
vide architectural and engineering plans.
continued
-------
TAO£ 3. (continued)
Project Title (Contract, Grant, or
Interagency Agreement)
Coal Preparation Plant Computer Model
(IAG-D6-E68S)
Engineering/Economic Analysis of Coal
Preparation Operation and Cost
(IAG-D6-E685)
Organization
Directing
Hork
EPA/DOE***
DOE
Organization
Perfonlng
Work
DOE(a), U. of
Pittsburgh, and
Battelle
Boffman-Muntner
Corp.
Objectives
Develop computer model capable of predicting
performance of coal preparation plants.
the
Determine the costs of cleaning for eight repre-
sentative coal preparation plants - from
jig plants to complex heavy media plants.
In
vo
Reactor Test Project for Chemical Removal
of Pyrltlc Sulfur from Coal (68-02-1880)
Microwave Desulfurization of Coal
(68-02-2172)
Battelle Hydrothermal Process Improvement
Studies (68-02-2187)
Evaluation of Chfmiral Coal Cleaning
Processes (IAG-DS-E685)
Hydrodesulfurlzation of Coal (68-02-2126)
Environmental Studies on Coal Cleaning
Processes (IAG-D5-K721)
Cost Evaluations of Coal Cleaning and
Scrubbing (IAC-D5-E721)
EPA
EPA
EPA
DOE
EPA
EPA
EPA
(c)
TKU Defense and
Space Systems Croup
General Electric
Battelle Columbus
Laboratories
Bechtel
Institute of Gas
Technology
Tennessee Valley
Authority (TVA)
TVA
Evaluation of the Meyers chemical coal cleaning
process in a 1/3 tph reactor test unit.
Evaluate the feasibility of coal desulfurization
by microwave treatment.
Evaluate methods for liquid/solid separation and
leachant regeneration.
Evaluate relative costs and performances of
selected chemical coal cleaning processes.
Evaluate desulfurization of coal by mild oxl-
dative treatment followed by devolatilizatlon.
Evaluate technology for controlling pollution
at TVA coal preparation plants.
Evaluate relative costs of coal cleaning and
scrubbing in complying with various SO2 emission
standards.
EHVIKOtMENTAL ASSESSMENT
Environmental Assessment of Coal Cleaning
Processes (68-02-2163)
EPA
Battelle Columbus
Laboratories
Evaluate pollution resulting from coal cleaning,
transportation and storage. Evaluate coal
cleaning as an SO,, emission control technique.
continued
-------
TABLE 3. (continued)
Project Title (Centrist, Grant, or
Interagency Agreement)
Organization
Directing
Vtork
Organization
Performing
Work
Objectives
Trace Elements and Hjneral Matter in
U.S. Coals (R804403)
Geology of Contaminants in Coal
(IAC-D6-E685)
Trace Element Characterization of Coal
Preparation Wastes (IAG-D5-E681)
(-•
A Hasbability and Analytical Evaluation
° of Potential Pollution from Trace
Elements (IAC-D6-E685)
Evaluation of the Effects of Coal Cleaning
on Fugitive Elements (IAG-D6-E685)
EPA
EPA
EPA/DOE
DOE
(b)
DOE
(a)
Illinois State
Geological Survey
U.S. Geological
Survey
Los Alamos Scienti-
fic Laboratory (LASL)
(a)
DOE
Bituminous Coal
Research Inc.
Characterize the elemental constituents and
mineralogy of U.S. coals.
Characterize coal resources west of the
Mississippi as to their elemental and mlner-
aloglc composition. Evaluate the geologic
factors which affect or control coal cleana-
blllty.
Characterize trace element and mineralogic
associations in coal preparation wastes.
Evaluate partitioning of trace elements in 10
U.S. coals during specific gravity separation.
Evaluate partitioning of trace elements during
preparation and use.
DEVELOPMENT OF POLLUTION CONTROL TECHNOLOGY
Control of Trace Element Leaching from
Coal Preparation Wastes (IAC-D5-E681)
Control of Blackwater in Coal Preparation
Plant Recycle and Discharge (IAG-D5-E685)
Stabilization of Coal Preparation Waste
Sludges (IAG-D5-E685)
EPA/DOE
DOE*
(b)
DOE
(a)
LASL
Pennsylvania State
University
Dravo Lime
Determine leachabllity of trace elements from
coal preparation wastes and evaluate pollution
control methods.
Characterize blackwater generated by coal
preparation plants and assess potential con-
trol methods.
Collect coal preparation plant 'sludges and
perform laboratory stabilization tests.
a - Department of Energy. Coal Preparation and Analysis Laboratory, Pittsburgh, Pennsylvania
b - Department of Energy. Office of Environment, Washington, D.C.
c - Department of Energy, Office of Energy Technology, Washington, D.C.
-------
1) where technically feasible, cost savings from the
use of physical coal cleaning (PCC) can be realized for
both utility and industrial boilers, particularly
those boilers with low capacity factors;
2) PCC appears to be unable to meet revised NSPS for
utility boilers unless combined with FGD;
3) there apparently are cases under current Federal
and State standards where the combined use of PCC
and FGD is more cost effective than the use of FGD
alone;
4) chemical coal cleaning may be competitive with FGD
for small industrial boilers having low capacity
factors;
5) the most probable use of chemical coal cleaning is
in combination with PCC to provide lower sulfur
levels than can be achieved by PCC alone.
Coal Cleanability
The DOE Coal Preparation and Analysis Group at Bruceton,
Pennsylvania is conducting laboratory experiments to determine the
liberation potential of pyrite from the principal U.S. coal beds.
Over 600 samples have been analyzed thus far. The results of
these studies will be used to assess the impact of coal cleaning
on SO2 emissions by determining the feasibility of cleaning various
coals.
Technology Assessment
In January of 1977, Versar, Inc. began a project to evaluate
the performance of coal cleaning equipment. The emphasis of the
161
-------
program was to be on fine coal cleaning and drying.
Thus far, seven coal preparation plants have been visited
by the.mobile test laboratory to determine the capabilities of
their froth flotation units, heavy media and hydrocyclones in
reducing pyrite content in fine coals.
In addition, a study has been completed on coal preparation
requirements for synthetic fuel conversion processes-. An evalu-
ation of the 11 most promising chemical coal cleaning processes
has also recently been completed and published (Contos et al., 1978).
Estimated costs for the processes range from $38.50 to
$65.72/ton including coal.
Also under the general heading of technology assessment are
several projects which are funded by EPA but managed by DOE. In-
cluded in this group are two high gradient magnetic separation
studies. One project is conducted by General Electric and is an
investigation of the removal of pyrite from dry coal powders.
The second study is being done by MIT and involves the recovery
of magnetite using high gradient magnetic separation techniques.
Results of both of these studies are being prepared for publication.
Other projects included under this agreement involve characterization
of blackwater constituents, a study of the adsorption/desorption
reactions in pyrite flotation, and a study of the surface pheno-
mena involved in the dewatering of fine coal.
Homer City Coal Cleaning Demonstration
Construction has been completed on a pilot, multi-stream
coal cleaning plant at the Homer City Generating Station Power
Complex, Homer City, Pennsylvania. The 1200 ton per hour plant
is designed to yield two products; a middling stream containing
162
-------
4.0 lb SO2/10 BTU, and a deep-cleaned stream which contains
1.2 lb S02/106 BTU. Extensive testing of the plant over the
next three years will be conducted under the cooperative efforts
of EPA, DOE, PENELEC and EPRI.
Meyers Chemical Coal Cleaning Process
A 1/3 ton per hour pilot scale test reactor has been con-
structed at Capistrano, California for the purpose of evaluating
the performance of the TRW Meyers process in chemically removing
sulfur from coal. The process operates on the principle of
aqueous ferric sulfate leaching and apparently is capable of re-
moving up to 95 percent of the pyritic sulfur. No organic sulfur is
removed. The reactor, after 254 hours of test operations on
50,000 lb of coal, has been shut down due to metal corrosion
in the primary reactor.
Microwave Desulfurization
Laboratory experiments by the General Electric Company have
demonstrated the technical feasibility of coal desulfurization by
microwave energy. Pyrite is preferentially excited by the micro-
waves to produce volatile and water soluble sulfur compounds which
can then be removed from the coal. The experiments have also
shown that irradiation of mixtures of coal, water and NaOH appears
to convert both pyritic and organic sulfur to water soluble sul-
fides. Present cost estimates suggest that microwave desulfuriza-
tion should be competitive with other chemical desulfurization
techniques.
163
-------
Environmental Assessment
The overall objectives of the environmental assessment activi-
ties have been to characterize coal contaminants and to identify
the fate of these contaminants during coal processing and coal use.
Initial studies have focused on sulfur and potentially hazardous
minor and trace elements contained in coal. Recent studies have
been concerned with a wider range of pollutants - those which may
be considered hazardous or toxic under the provisions of the
Water Pollution Control Act (priority pollutants), the Resource
Conservation and Recovery Act (hazardous wastes), the 1977 Clean
Air Act Amendments (hazardous air pollutants), or the Toxic Sub-
stance Control Act. The basic intent of the environmental assess-
ment activities is to identify pollutants which pose health or
ecological threats and devise cost effective strategies for dealing
with the pollutants.
Environmental Assessment Project
A three year project to assess the environmental impacts of
coal preparation, coal transportation and coal storage is being
conducted for IERL-RTP by Battelle Columbus Laboratories. Major
project activities include:
1) the development of a technology overview containing
a description of all current coal cleaning processes
and their associated pollution control problems;
2) the development and performance of an environmental
test program to obtain improved data on pollutants
from commercial coal cleaning plants;
164
-------
3) the development of criteria to be used in assessing
the potential health and ecological impacts of pollu-
tants from coal cleaning processes;
4) the performance of studies to determine the relative
environmental impacts of coal cleaning/ FGD, and other
SO,, emission control techniques.
Studies to develop criteria for assessing the relative envi-
ronmental hazards associated with pollutants resulting from coal
preparation, coal transportation and coal storage are nearing
completion. The approach has been to characterize the physical
and chemical toxicity of pollutant or effluent streams sampled
at their respective sources. This differs from the approach
taken in environmental impact assessments - the characterization
of air, water, and biological quality in the facility under study.
The source assessment criteria incorporate methodologies being
developed by IERL-RTP and adapt them to coal cleaning processes
(Hangebrauck, 1978).
Concurrent with development of source assessment criteria,
studies are in progress to select coal cleaning plant sites for
environmental testing. A master test plan is being developed to
ensure a comprehensive test program and to facilitate the planning
and preparation of site-specific test plans for the designated
facilities.
Thus far, background studies of the air, water and biological
quality in the vicinity of the Homer City Generating Station
have been completed in preparation for an environmental assessment
165
-------
of the cleaning plant which has been built at the site. Testing
will begin when the plant normalizes operations.
Coal Contaminants
Three distinct programs are directed to the identification
and characterization of contaminants in coal. Specifically the
research attempts to demonstrate the occurrence, association and
distribution of trace element and mineral phases in the coal seam.
One portion of this research, led by the Illinois State
Geological Survey, concentrates on coals of the Illinois Basin.
This work has three principal goals: 1) to determine the mode
of occurrence and distribution of trace elements and minerals in
coal seams; 2) to study the mineralogy and genesis of sulfide
minerals in coal; and 3) to evaluate the potential for removal
of minerals from coal by various preparation techniques.
A second area of investigation is being conducted by the
U.S. Geological Survey in Reston, Virginia. This project has dual
objectives. One is to determine the geologic factors which
affect or control the physical cleanability of coal, and to develop
geologic models which can be used to help maximize efficiency
and minimize environmental impact from coal mining, cleaning and
burning. The second objective is to provide the necessary chemical,
physical and mineralogic data on the Nation's coal resources to
permit evaluation of the environmental impact resulting from coal
preparation and utilization.
The third study in this area, being conducted at the Los
Alamos Scientific Laboratory (LASL), deals with the contaminant
potential of coal preparation wastes. The research has three
166
-------
distinct phases: 1) to characterize the minerals, trace elements,
and their association in coal preparation wastes; 2) to study
the effects of weathering and leaching on trace elements in coal
wastes; and 3) to identify and evaluate techniques for controlling
or preventing trace element contamination from coal waste materials.
Phases one and two have been completed and the results published
(Wewerka et al., 1976, 1978a, 1978b).
Pollution Control Technology
The subprogram to develop coal cleaning pollution control is
in a formative stage. A wide variety of techniques exist for
controlling conventional pollutants such as total suspended
solids, total particulate emissions and pH. However, as coal
cleaning processes evolve and as pollution regulations become
more stringent, improvements will be required in pollution control
techniques. This portion of the program, therefore, addresses
projected as well as current pollution control techniques.
Control of Trace Element Leaching from Coal Preparation Wastes
The Los Alamos Scientific Laboratory is conducting studies
to assess the potential environmental contamination from
trace elements in coal preparation leachates, and to determine
suitable methods for control. Trace element and mineralogic
characterization of refuse samples has been completed. Analyses
of the leachates from these samples have also been completed.
The control technology assessment includes methods to pre-
vent the leaching of trace elements from coal cleaning wastes
and to remove trace elements from the leachate once they have
been entrained. Methods evaluated thus far include the addition
of lime, limestone, lye and other naturally occurring alkaline
167
-------
materials to the waste material. Calcining is presently being
studied. Techniques under consideration for water treatment in-
clude reverse osmosis, chelation, alkaline neutralization/ per-
manganate oxidation and others.
Control of Blackwater
Blackwater (process waste water) from coal preparation plants
consists of mixtures of fine coal, clay minerals, quartz, calcite,
pyrite and other fine-grained mineral particles suspended in
water. The composition of these effluents must be known before
they can be adequately treated for reuse or discharge. Pennsyl-
vania State University has completed an investigation to charac-
terize the solid constituents in blackwater and to determine
the best procedures for treating it (Apian et al., 1979). The
final report is being prepared for publication.
Stabilization of Coal Preparation Waste Slurries
Reject ponds are becoming increasingly undesirable because
of safety, environmental and land use considerations. An alter-
native approach to slurry disposal is the treatment of these
wastes to create stable solids.
EPA is funding a project at Dravo Lime Co. to characterize
the engineering, physical and chemical properties that affect
stabilization of fine wastes from coal preparation plants. The
requirements and conditions for stabilizing these wastes with
and without additives are being determined.
CONCLUSIONS
The past year has been one of change. Potential applications
of coal cleaning, and therefore our research and development goals,
168
-------
have been altered by new environmental legislation and impending
energy legislation. Studies are in progress to identify the
technical capability and costs of various coal cleaning technologies
for removing sulfur and other contaminants in coal. Progress has
been made in the development of physical coal cleaning techniques
for improved pyrite removal and coal energy recovery. Progress
is continuing in the development of chemical coal cleaning pro-
cesses, but impending environmental standards are causing uncer-
tainties regarding future market applications.
Methodologies have been developed for the environmental assess-
ment of coal preparation processes, and tests are scheduled to
begin shortly. The conditions under which trace elements are
leached from coal preparation wastes have been identified, and
preliminary studies have identified the effectiveness of several
Pollution control techniques.
CONVERSION FACTORS
ton = 0.907 metric ton
Ib - 0.436 kg
BTU = 1055.6 joules
BTU/lb » 2326 joules/kg
REFERENCES
Apian, F. F. and Hogg, R., 1979. Characterization of solid
constituents in blackwater effluents from coal preparation
plants. EPA-600/7-79-006 (in press).
Contos, G. Y., Frankel, I. F. and McCandless, L. C., 1978. Assess-
ment of coal cleaning technology: an evaluation of chemical
coal cleaning processes. EPA-600/7-78-173a (in press).
Hangebrauck, R. P. Environmental assessment methodology for
fossil fuel energy processes, in Symposium Proceedings:
Environmental Aspects of Fuel Conversion Technology, III
(September 1977, Hollywood, Florida). EPA-600/7-78-063,
PB 282429, April 1978.
169
-------
U.S. Congress. First report of the Secretary of Health, Education
and Welfare to the U.S. Congress, Progress in the prevention
and control of air pollution. June 28, 1968.
U.S. Environmental Protection Agency, Code of Federal Regulations,
40 Protection of Environment, revised as of July 1, 1976.
Part 60, Subpart D, Standards of performance for fossil fuel-
fired steam generators, pp. 23-28.
U.S. Environmental Protection Agency, Code of Federal Regulations,
40 Protection of Environment/ Revised as of July 1, 1976,
Part 60, Subpart Y, Standards of performance for coal pre-
paration plants, pp. 56-57.
U.S. Environmental Protection Agency, Code of Federal Regulations,
40 Protection of Environment, September 19, 1977, Part 434,
Subpart B, Standards of performance for new sources - coal
preparation plants and associated areas, p. 46937.
Wewerka, E. M., Williams, J. M., Wanek, P. L. and Olsen, J. D.,
1976. Environmental contamination from trace elements in
coal preparation wastes: a literature review and assessment.
EPA-600/7-76-007, PB 267339. August 1976.
Wewerka, E. M. and Williams, J. M., 1978a. Trace element char-
acterization of coal wastes - first annual report.
EPA-600/7-78-028, LA-6835-PR, March 1978.
Wewerka, E. M., Williams, J.M., et al. 1978b. Trace element
characterization of coal wastes - second annual progress
report. EPA-600/7-78-028a, PB 284450, July 1978.
170
-------
OVERVIEW OF DOE COAL CLEANING PROGRAM
Cyril W. Draffin
Fossil Energy Planning and Evaluation
Assistant Secretary for Energy Technology
U.S. Department of Energy
20 Massachusetts Avenue, NW
Washington, DC 20545
ABSTRACT
The U.S. Department of Energy has an active program in coal preparation.
The program includes coal characterization, physical and chemical cleaning,
and economic and market studies to assess the ability of coal preparation
to facilitate coal usage in the United States.
The DOE fiscal year 1978 budget in coal preparation is approximately
$9.0 million. The DOE fiscal year 1979 coal preparation budget will be
$8-14 million, depending on Congressional action.
This paper initially examines why coal preparation is important and
how it compares to competitors. After describing the major elements of the
DOE coal preparation program, the DOE organizations Involved are discussed
along with their current and planned projects. Finally, the issues facing
DOE and industry in the coal preparation arena are explored.
171
-------
Introduction
Coal cleaning has recently received attention as a method of
meeting environmental regulations. Using advanced coal cleaning
techniques to help meet environmental standards is a definite change
from conventional coal washing to remove heavy impurities.
This paper discusses the primary environmental standards that will
affect coal usage in the United States and how coal preparation stacks
up against its competition. Then it describes the U.S. Department of
Energy's (DOE) program in coal preparation, including its organization,
projects, people, and funding. Finally we hope to stimulate some
discussion by exploring the issues facing DOE and Industry in coal
preparation*
Projected Coal Usage and Applicable Environmental Standards
The general objective of the National Energy Flan (NEP) is to
reduce United States' dependence on scarce fuels (oil and gas) while
continuing to achieve economic growth and reduction of environmental
pollution. The programs outlined in current legislative proposals that
will encourage or require increased coal use In the utility and indus-
trial sector include (1) the mandatory coal conversion program, (2)
172
-------
taxation of oil and gas use, while providing tax rebates on coal
investments, and (3) gradual deregulation of natural gas.
Environmental goals impact the coal demands generated under the
NEP in two ways* First, pollution controls which increase costs and/or
unreliability can discourage individuals from responding to the NEP's
financial incentives, and/or exempt them from mandatory coal conversion!
Second, EPA or state regulations can block the use of coal in regions
where national environmental goals cannot be met (i.e., control systems
for coal use are not adequate)*
The Clean Air Act established two major mechanisms to control air
pollutant emissions from stationary sources. New Source Performance
Standards (NSPS) are national standards limiting emissions from speci-
fied sources. State Implementation Plans (SIP's) establish compliance
schedules and emission limitations for all types of sources so as to
ensure attainment/maintenance of air quality standards in each state*
All boilers are, at a minimum, subject to any applicable SIP require-
ments, and if a SIP limitation is more stringent than the corre-
sponding NSPS, the SIP is the binding regulation.
In addition to the SIP emission standard or NSPS,- new emission
sources may be subject to even more stringent levels of control on a
173
-------
case-by-case basis. All major new sources must be individually
reviewed and their impacts modeled to ensure that their emissions
will not cause more than the allowed degradation of air quality,
or prevent attainment of Ambient Air Quality Standards.
For purposes of this analysis, air emission limitations have been
divided into five general categories: SIP's applicable to utilities,
SIP's for smaller industrial boilers, the current utility NSPS, and
anticipated NSPS for both utility and industrial boilers* The basic
emission limits for each category for sulfur oxides, nitrogen oxides,
and particulates are shown in table It
Although air emissions are the primary environmental concern in
the firing of coal, associated water discharges and solid waste disposal
practices come under Federal regulation. They are important because
waste streams are produced in coal cleaning, disposal of coal combus-
tion residues, and processing of exhaust streams. The four Federal
laws influencing the disposal of these wastes are Resource Conservation
and Recovery Act of 1976, Federal Water Pollution Control Act as
amended by Clean Water Act of 1977, Toxic Substances Control Act of
1976, and Safe Drinking Water Act of 1974.
Projections of utility and industrial coal demand for 1985, 1990,
and 2000 are presented in figure 1. In general, the 1985 and 1990
174
-------
Regulation fi Source
Sulfur Oxides
Particulates
Nitrogen Oxides
Utility
Ln
State limits
Current NSPS2
NSPS Revision3
Industrial
State limits1
NSPS Revision3
1.0 - 6.0 (2.0) f/MMBtu
1.2 f/MMBtu
85* reduction,
.2 f/MMBtu floor,
1.2 f/MMBtu ceiling
.15-9.0 (2.3) f/MMBtu
Unknown, perhaps
70-85* with size
cut off
0.1 - 0.6 (.3) f/MMBtu
0.1 f/MMBtu
99% reduction
0.03 f/MMBtu
0.1-0.6 (.5) f/MMBtu
Unknown, perhaps
99% with size
cut off
None or 0.7 f/MMBtu
0.7 f/MMBtu
65% reduction, and
0.5-0.8 f/MMBtu,
based on coal and
furnace type
Most unregulated
Unknown
1 Variablility reflects differences by location, and size. Figure in ( ) used as representative for this study.
2 Applies to all boilers built after 1971 and over 250 MMBtu/hr .
3 No revisions have been formally proposed. Figures reflect current anticipated range. Utility standards
likely in effect in 1978 and industrial in 1980.
TABLE 1: AIR POLLUTION CONTROL REQUIREMENTS
-------
COAL DEMAND
IN QUADS
METALLURGICAL
COAL
INDUSTRIAL
UTILITY
1975
1980
1985
1990
1995
2000 YEAR
FIGURE 1: PROJECTED COAL DEMAND BY SECTOR AND APPLICABLE AIR
EMMISSION LIMITS
176
-------
projections are based on the "most probable" form of the National
Energy Plan expected to be ratified by Congress. The shaded area
represents portion of project demand falling under the existing air
pollution emission limits.
The split of coal use between utilities and Industry is expected
to be about 80/20 in 1985 and 1990, and 75/25 in 2000. Although the
upcoming revisions of utility and Industrial NSPS have clear long term
importance, affecting more than 50 percent of total projected coal
demand in 2000, the revision will affect only 15 percent of total
demand in 1985. State limitations are still Important in 1990,
affecting 46 percent of total non-metaalurgical coal demand. By 2000,
the impact of current state limitations declines to about 25 percent of
total demand.
How Coal Preparation Compares to Alternative Cleanup Technologies in
the Utility and Industrial Sector.
Because coals and environmental standards vary, different approaches
to coal cleanup will be required. This is particularly important because
the electrical utilities are expected to spend $49 to 73 billion on
pollution control devices in the next decade.1 A significant cost
^•Letter, John F. O'Leary, Deputy Secretary, DOE, to Douglas C.
Costie, Administrator, EPA, July 6, 1978. Present value cost through
1990 depends on availability of FGD systems (90 to 100 percent) and
which NSPS utility standards are finally agreed upon.
177
-------
saving to consumers can result from choice of the cheapest, most
reliable system, expecially if it saves high-cost oil which is passed
through under fuel adjustments*
FGD will be the primary compliance method for meeting proposed
sulfur New Source Performance Standards (NSPS) for new units built
after 1983. Because 85 percent removal is expected to be required,
physical coal preparation will be unable to meet sulfur standards
alone, and must be used in combination with an FGD system.
Coal preparation and low-sulfur coal may play major roles in
allowing existing coal-capable utility boilers, and units coming
on-line before 1983, to meet existing SIPS and NSPS. These units,
which are expected to represent 70 percent of the coal-fired utility
electrical generation in 1990 and will continue to operate under the
existing 1.2 #/MM Btu standard,are probably the most important area
where imported oil can be saved in the next decade. In these applica-
tions, low-sulfur coal, physical coal preparation, or a combined coal
preparation/FGD system is expected to be the cheapest option. An
example of a combined coal cleaning/FGD system is TVA's recently
announced Paradise coal cleaning plant.
Table 2 presents some results of current internal DOE study of
clean-up technologies applicable for the utility and industrial sectors.
178
-------
VO
APPLICATION
Environmental
Standard
1985 Coal Usage
(in Quads)
1990 Coal Usage
(in Quads)
Preferred
Technology
Acceptable
Technology
Specialized
Applications
UTILITY
Existing
SIP's
9.9
9.9
Physical Cleaning
Low-Sulfur Coal
Medium- Sulfur Coal
Non-Regenerable
FGD; Regenerable
FGD; SRC I; Chem-
ical Cleaning
Current
NSPS
4.8
4.8
Physical Cleaning
Low-Sulfur Coal
Non - Regene rab 1 e
FGD; Regenerable
FGD
SRC I
Chemical Cleaning
Upcoming
NSPS
1.8
6.6
Non-Regenerable
FGD; Physical
Cleaning § FGD;
Regenerable FGD
SRC I (for a
few coals)
INDUSTRIAL
Existing
SIP's
2.2
3.0
Physical Clean-
ing; Low-Sulfur
Coal
Chemical Clean-
ing; SRC I; Non-Re-
generable FGD;
Regenerable FGD
SRC I
Chemical
Cleaning
Upcoming
NSPS
1.0
2.5
Chemical Clean-
ing; SRC I; Non-
Regenerable FGD;
Regenerable FGD
SRC I
Chemical
Cleaning
TABLE 2: RANKING OF CLEANUP TECHNOLOGIES BY MARKET SEGMENT
(excluding fluidized bed combustion and low and medium BTU gas)
-------
From the table it can be seen that with existing regulations where
there is no percent sulfur emissions limitation, low-sulfur coal or
physical coal cleaning will be the preferred technology. Percentage
reduction requirements tend to favor flue gas desulfurization, particu-
larly in the utility sector.
Coal consumption is expected to grow fastest in the United States
industrial sector, with a tripling in coal usage from 1975 to 1990.
By 1990, new industrial coal usage is expected to be approximately 100
million tons which is equivalent to saving 1.0 million bbl oil per
day.
Because of the large number of users, environmental standards and
types of applications (e.g., boilers, process heaters, direct heat, and
metallurgical applications), evaluation of compliance techniques is
complex. Existing standards, which will be in effect for all industry
until 1980, are currently set by states or local governments. These
state standards can usually be met most cheaply by low-sulfur or
cleaned coal. In some industrial applications, however, available
waste streams make FGD more economical.
In industry, reliability and cost is even more Important than In
utilities because large energy users that would be likely to use coal
180
-------
(steel, refineries, chemicals, paper) run continuously and cannot shift
loads among plants as easily as an electric utility grid. With produc-
tion dependent on reliable process steam, heat, and electricity, they
expect a higher (95 to 98 percent) reliability. Because energy produc-
tion is only a small increment of total product cost, yet could jepardize
the entire industrial operation, reliability is the most important
factor. Other important factors are being competitive and having
flexibility in choosing to burn alternative fuels.
Coal preparation is particularly attractive because it produces a
more uniform coal feed which may make boilers and combustors more
reliable (approaching that of oil and gas, which are primary competi-
tors). In many applications, coal preparation allows environmental
standards to be met without FGD. This is particularly important
because sludge disposal is likely to be more of a problem for urban-
based industrial applications.
Because Federal industrial emission standards have not been set,
it is premature to Judge the value of chemical cleaning or SRC I in
meeting standards. However, preliminary indications suggest very large
industrial boilers (such as for cogeneration) are likely to need
scrubbers and small units will probably require only physical cleaning.
181
-------
DOE Organization and Funding
For the past 4 years the Federal efforts in coal preparation have
been scattered. In 1975, ERDA had a number of divisions working on
coal preparation, the Bureau of Mines was being encouraged and.funded
by EPA, and EPA was actively funding coal preparation work. In March
1976, a ERDA/Fossil Energy task force headed by C. W. Draffin reported
on the status of Federal coal preparation efforts, and on June 24, 1976
X
the task force submitted statements of work for a proposed Fossil
Energy Coal Preparation Program, the implementation of which was
assigned to R. A. Corey, in what is now the Division of Fossil Fuel
Processing.
When DOE was formed on October 1, 1977, the Bureau of Mines
coal preparation work ($5MM) was placed in the Division of Solid Fuel
Mining and Preparation, and ERDA's Advanced Research and Supporting
Technology's coal preparation program ($1.2MM) was placed in the
Division of Power Systems. Other coal prepration'work was continued
by the Division of Coal Conversion (in Fossil Energy under the Assistant
Secretary for Energy Technology) and the Division of Environmental
Control Technology (under the Assistant Secretary for the Environment).
FY 1978 funding is shown on table 3.
182
-------
Table 3
FY 1978 Federal Coal Preparation Program
DOE/Fossil Energy $5.1 Million
(From Bureau of Mines)
DOE/Fossil Energy 1.2 Million
(From ERDA)
DOE/Environmental Control 0.7 Million
(From ERDA)
Environmental Protection Agency 2.0 Million
(To Come to DOE in FY 1979?)
To coordinate DOE efforts, a DOE Fossil Energy discussion of
coal preparation was held on November 15, 1977 and George Fumich, the
Program Director for Fossil Energy, requested that a program plan for
coal preparation be prepared under the direction of C. W. Draffin* A
draft program plan for coal preparation was submitted on January 9,
1978, but never publicly released for comment* In the recent reorgan-
ization, coal preparation, coal mining, oil, gas, and oil shale were
assigned to Dick Hertzberg, Director of Fossil Fuel Extraction.
Currently, the combined Fossil Energy FY 1978 budget for coal
preparation is about $7MM. The FY 1979 budget is about $7MM, with
possible additional funding of $2 MM in EPA transfer funds and $3.5MM
in additional Congressional authorizations. Current projects are
listed in figure 2. The most significant recent decision made affecting
183
-------
$1.1 MM SOLID FUEL MINING & PREPARATION - FIELD
Petrographic Studies $0.1 MM
Froth Floatation .2
HGMS Tests on Coal Samples .2
Equipment Studies .1
Instrumentation
Advanced Gravity Separation .1
Washability Studies .3
Lignite Up-grading .1
Facility Planning .05
$4.1 MM SOLID FUEL MINING & PREPARATION - HEADQUARTERS
Bruceton Facility 2.2
Chloronalysis (Jet Propulsion Lab) .4
Oil Agglomeration (Jones&Laughlin) .75
Fine Wastes (Dravo) .1
Organic Sulfur (U.of Houston) .08
Lignite Drying (Grand Forks ETC) .1
Economic Assessment .05
Sulfur Functional Groups (Aerospace) .05
Waste Impoundment Assessment .2
(U. of Alabama)
Dry Fluidized Bed HGMS (Oak Ridge NL) .17
Coal Analysis (Warner Labs)(.25 USGS $) -
$1.1 MM POWER SYSTEMS - HEADQUARTERS
Oxydesulfurization (Pittsburgh ETC) .4
Beneficiation (Ames Laboratory) .275
Microwave Desulf.(via EPA) (GE) .18
Model for Predictions (U. of West Va.) .037
Effect of Minerals (U. of West Va.) .14
Recovery of Ultrafine Coal (Ohio St.) .02
Optomization of Bacterial Leaching .02
(State Univ. of NY at Binghamton)
Precombustion Desulfurization(U. Minn.).077
$0.7 MM ENVIRONMENTAL CONTROL TECHNOLOGY - HEADQUARTERS
(under Assistant Secretary for Environment)
Trace Element Analysis (Los Alamos NL) .225
Trace Metals & Radioactivity (BCR) .25
Coal Prep for Electric Util. (Homer .10
City)
State of Art for Coal Prep & Econ. .125
Assessment (Argonne National Lab)
$7.0 MM
FIGURE 2: FISCAL YEAR 1978 DOE COAL PREPARATION PROGRAM
184
-------
coal preparation was the funding of a $10 million Coal Preparation Test
Facility in Bruceton, Pennsylvania.
Scope of Fossil Energy Coal Preparation Program
The coal preparation program is being redirected to be a driving
force in the Fossil Energy Program that facilitates coal combustion,
gasification, liquefaction and MHD by (1) preparing coals to reduce
costs, to reduce environmental impacts, and to impove reliability of
coal use and (2) helping to establish which coal characteristics are
most suitable for specific applications.
The primary elements of the program include:
o Coal characterization and determination of feed requirements;
o Physical and chemical coal cleaning;
o Coal processing and handling (e.g. coal grinding, dewataring
and feeding);
o Economic and market assessments.
Coal characterization focuses on the coal constituents and
characteristics (ash, sulfur, Btu content, minerals, friability, caking
characteristics, caking properties, etc.) that determine the best use
185
-------
for different coals. The program will be closely integrated with the
ongoing DOE Coal Science Research Program.
The coal feed characteristics necessary for coal combustion,
gasification, liquefaction, and MHD would be determined in conjuction
with the appropriate program offices. The economic trade offs between
extensive coal preparation and over-designed coal utilization or
processing facilities would be made by the Process Design and Economics
Group in the Fossil Energy Division of Systems Engineering. Particular
emphasis would be paid to current combustion technology in the indus-
trial sector, so technical problems impeding the switch to coal could
be solved.
The second element, physical and chemical coal cleaning, would be
a continuation of physical cleaning activities of the Bureau of Mines
that have been transferred to the Fossil Energy Division of Fossil Fuel
Extraction. It would also include chemical cleaning work previously
funded by the Power Systems Division. Technologies to 'be stressed
Include oil agglomeration, froth flotation, advanced fine coal cleaning
and advanced chemical cleaning, including oxydesulfurizatlon. Particular
attention would be paid:
o to recovering fine coal both to eliminate environment
impacts of coal wastes and to increase Btu recovery;
186
-------
o to developing the most cost effective way of meeting
industrial and utility environmental standards (especially
sulfur).
Coal processing and handling would Investigate ways of handling,
storing, blending, grinding, dewatering, pelletizlng, transporting, and
feeding coal. Primary attention would be placed on advanced grinding,
dewatering, and palletizing processes. This element of the coal
preparation program would focus on coal handling problems that impede
coal usage and would be done in conjunction with the Division of Fossil
Fuel Processing, DOE's Assistant Secretary for Resource Applications
and the Electric Power Research Institute. Thir activity may aid DOE's
Economic Regulatory Administration by allowing fossil energy to provide
technical judgement on appropriate Industrial coal cleaning and handling
facilities.
The economics, environmental impacts and market applications for each
of these elements would be made an integral part of development and
assessment. Other efforts would focus on EPA's setting of industrial
environmental standards and cooperation with industry, EPA, and EPRI.
New research directions that are being pursued include the following:
Chemical cleaning - offers the potential for removal
of organic sulfur (up to 40 percent) as well as
187
-------
increasing the removal of pyritic sulfur (up to 95
percent) frequently using high temperatures, high
pressures, and chemical reagents or oxygen. DOE
effort being placed on understanding coal character-
istics (especially types of organic sulfur) and
process conditions for removing impurities•
Oil Agglomeration - allows fine-sized coal to be
recovered from water slurries and dewatered. Intro-
duction of oil eliminates surface moisture on coal
particles; occluded water between coal particles in
floe can be removed by centrifuges* The product has
10 to 12 percent water (mostly inherent water) which
can be direct-fired in a boiler or pelletized for
shipment.
Palletization (or Briquetting) - allows fine-sized
coal to be melded into forms which are more easily
transported and handled. This is part of the DOE
effort to facilitate fine coal transport and usage.
High Gradient Magnetic Separation (HGMS) - Although
technology has not been proven to date, use of large
magnets offer potential to increase both pyritic
sulfur removal and Btu recovery, at a price comparable
to conventional heavy media plants, and without
creating fine coal dewatering problem (oil agglomera-
tion would be used). Dry HGMS allows sulfur to be
removed in areas where water is limited.
Lignite and Subbituminous Drying and Pelletizing -
improve potential for using these coals by increasing
Btu content per pound, decreasing spontaneous combus-
tion and taking out impurities such as sodium.
Trace Element Removal - coal cleaning offers potential
of removing some heavy metals before combustion.
Automation - development of operational ash, sulfur,
and moisture meters that could be installed to
improve Btu recovery and product quality.
In DOE headquarters the primary Fossil Energy people involved in
coal preparation are Bill Warnke, Cyril Draffin, and Wayne McCurdy.
188
-------
DOE's field organizations include the 50 member coal preparation group
headed by Al Deurbrouck, which is divided into two groups. The Coal
Preparation Laboratory in Bruceton includes 25 full time and eight
part-time people while the Coal Preparation Analysis Group in Pittsburgh
has 15 full-time and two part-time people. In addition, parts of the
Pittsburgh Energy Technology Center (PETC) are actively involved in
chemical coal cleaning (under Jim Gray) and environmental work (under
Bill Peters). Ames Laboratory is actively involved in physical and
chemical cleaning and Argonne National Laboratory is involved in
environmental assessments.
.Scope of Coal Preparation Program under Assistant Secretary for Environment
The Assistant Secretary for Environment (ASEV) has the
responsibility to access the environmental, health, and safety aspects
of DOE energy technologies. To meet this responsibility, the Assistant
Secretary reviews the environmental, health, and safety aspects of
energy technology RD&D in the context of environmental policies, and
conducts research and development to meet the needs of DOE programs for
new information and data in the environmental, health, and safety area.
The primary people in headquarters are Myron Gottleib and Charles
Grua.
189
-------
Programs currently underway are directed at mitigating the impact
of waste material from coal preparation plants. An example is a
project to determine the reclamation techniques for waste banks in a
cooperative effort with state and local governments. Ecological
effects and long-term stability of reclaimed land will be assessed over
a multiyear period. Other studies address the potential environmental
impact of leachates from both unreclaimed waste banks and coal storage
piles, and the control technology requirements to mitigate these impacts.
The fate of trace and minor elements in the overall coal utilization
cycle is being assessed in a laboratory study of physical coal cleaning
utilizing coals from various regions.
A major study directed at the environmental implications of
generating electricity from coal is being conducted at Argonne National
laboratory* This study is expected to provide an assessment of the
current state-of-the-art of coal preparation and the tradeoffs through
the fuel cycle of benefits and costs. Also, through a cooperative
effort with EPA, an assessment of the control technology status is
being made of the Homer City Coal Cleaning Demonstration Plant.
Issues Facing DOE and Industry
Two primary issues face DOE and industry: (1) whether coal
preparation will be of significant value in allowing technologies to
190
-------
use coal in an environmental acceptable way, and; (2) whether private
industry will develop advanced coal cleanup technologies in a timely
manner•
Because the only real alternatives to clean coal combustion are
burning oil or gas or using nuclear power, it is necessary to have
reliable, reasonably economic, commercial cleanup systems available*
The general strategic options available to DOE in the coal cleaning
area include any or all of the following:
A. Industrial and Utility Tradeoff Studies
- To facilitate enforcement of Federal and state legislation
- To determine best way of meeting different environmental
standards
- To identify R&D needs
- To assess validity of requested cost and environmental
exemptions under the coal conversion provisions of the
National Energy Act
- To independently comment on EPA regulations
- To recommend new legislation
191
-------
B. Applied Research and Development
- To support reliable compliance with existing and anticipated
environmental standards
- To support EPA by developing best available control
technologies, especially for nonattainment areas
- To make up for lack of industrial R&D caused by potential
sudden obsolescene with changing environmental regulations
- To reduce costs of using coal
- To overcome industry's hesitency to develop improved
control technologies that they will be forced to install
at additional cost
C. Commercial Demonstration
- Decrease technical uncertainty to ascertain need for
exemptions
- Develop data to allow EPA to set more realistic standards
Further examination of industry's intentions and likely performance
appears needed in planning both R&D and demonstration projects because
we do not yet have a clear picture as to why vendors and coal users
will not develop technology themselves.
192
-------
Joint planning by industry and Government will be essential to
develop coal cleanup technologies that protect the environment while
assuring reliability and cost-competitiveness of using coal*
The Department of Energy is interested in your factual appraisal
of specific areas in (1) cleaning, dewatering, or handling coal or
lignite, (2) disposing of coal preparation wastes, or (3) operating
coal preparation facilities where there are significant process Improve-
ments needed. In making those suggestions, a documented discussion is
needed of what private industry is doing and how (If at all) the
Federal Government can be of specific assistance.
193
-------
OVERVIEW OF EPRI COAL CLEANING PROGRAMS
Kenneth Clifford and Shelton Ehrlich
Electric Power Research Institute
Palo Alto, California
No abstract or paper available.
194
-------
AN INTEGRATED ASSESSMENT OF COAL TECHNOLOGIES
Richard S. Davidson
Battelle
Columbus Laboratories
505 King Avenue
Columbus, Ohio 43201
ABSTRACT
The Coal Technology Assessment (CTA) is a part of EPA's overall Integrated
Technology Assessment (ITA) program. ITA was launched to "identify environ-
mentally, socially, ^and economically acceptable (energy development) alter-
natives". The CTA .study will try to anticipate what a given coal-based
energy-technology mix might mean to our society and to outline policy options
that can prepare us for that projected future. Two basic questions concern
the level of coal-derived energy desirable in the U.S. over the next half
century and the social, economic, environmental, and institutional impacts
of different coal technologies that might be employed to meet this level.
The product of the study is intended for use by policymakers at all levels
of government and in the private sector, not just EPA.
The study process involves six major modules: issue identification—
projected problems and conflicts; scenario development—descriptions of
different possible states of future society; projected technology mixes
within these scenarios; measurement of impacts of coal-based energy develop-
ment; development of policy options; and communication of the findings. The
bottom line of the study is policy options and the key element is public
participation. The policy options will be intended to avoid or mitigate
undesirable impacts—environmental, social, or economic—of coal-based energy
technology or to take advantage of opportunities that may be discovered.
Policy options will fall into two major categories—legal/institutional and
technological. Public participation throughout the process by which these
policy options are developed is critical. This is being accomplished through
such devices as interested party forums in different parts of the country,
a newsletter encouraging feedback, and involvement of a National Technical
Advisory Council.
195
-------
Traditionally, energy/environment research programs in the U.S.
have been concerned primarily with the health and ecological impact of
new energy plants and technology. In December, 1973, Dr. Dixie Lee Ray,
then Chairman of the U.S. Atomic Energy Commission submitted a report to
President Nixon titled, The Nation's Energy Future. This propelled in-
terest in a broader-based look at the effects of energy technology and
provided the impetus for a Government interagency program on energy and
the environment. As a result, two interagency panels — one dealing with
control technology development, the other with environmental effects re-
search — developed .Dr, Ray's suggested program in greater detail.
From the reports of these panels, often referred to as the
"Gage Report" and the "King-Muir Report"j the Office of Management and
Budget established an interagency task force on the "Health and Environ-
mental Effects of Energy Use". This task force was to (a) examine ongoing
federal research in the energy/environment field and (b) recommend an al-
location of research funds for a more effective research program.
196
-------
A major conclusion of the energy/environment task force was that
the social and economic consequences of alternative energy and environmental
policies had to be considered along with the more traditional health and
environmental impacts. The King-Muir report recommended the formation of a
research program to identify "environmentally, socially, and economically
acceptable (energy development) alternatives". As a result, the Environ-
mental Protection Agency (EPA) launched the Integrated Technology Assessment
(ITA) program.
Traditional EPA environmental research programs had followed the
same trend noted in the introduction. They had been "waste stream" oriented,
confining the environmental analysis to direct effects of pollutant emissions
and discharges from industrial facilities. Adequate attention frequently
had not been given to "nonpollutant" effects such as noise, land use, em-
ployment, community services, and esthetics. The Gage report reoriented
this program by calling for a series of "environmental assessments" (EA's)
designed to go much further in utilizing chemical and biological analysis,
as well as existing health/ecological effects data, to assess the impacts of
Industrial discharges on air, water, and land. The report also recommended
attention be given to nonpollutant effects, i.e., social and economic. Six
environmental assessments (EA's) are now under way for high- and low-Btu
gasification, coal liquefaction, fluidized-bed combustion, conventional
combustion, coal cleaning, the subject of this Conference.
197
-------
Further, EPA is currently sponsoring three technology assessments
on energy: the Western Energy Study under the direction of the University
of Oklahoma's Science and Public Policy Program, the Ohio River Basin Energy
Study (ORBES) being undertaken by a group of seven midwest universities,
and the Coal Technology Assessment study being reviewed here and being
conducted by Battelle's Columbus Laboratories and the University of Michigan's
Program in Technology Assessment.
Technology Assessment (TA) is a class of policy studies directed
to examine the broadest social implications of the introduction of a new
technology or the expansion or extension of an existing technology. It 1s
intended to provide the decision-maker with useful advice and guidance on
policies, programs, plans, and alternative actions. Since there exists no
"science" of TA, the specific techniques employed for any given technology
assessment are subject to considerable debate and variation. However, with
some variation, it is generally agreed that most technology assessments In-
clude at least these generic elements:
o Definition of the Problem
o Description of Alternative Technologies
o Identification of Biophysical and Socloeconomic Impacts
o Evaluation of Impacts
o Characterization of the Decision-Making Process
o Identification and Evaluation of Policy Alternatives
o Involvement of Interested Parties, I.e., those who have
a direct stake in possible impacts
o Utilization of Scenarios
o Communication of Results.
198
-------
The Coal Technology Assessment will involve each of these elements.
The general framework in which are they being included and the major objec-
tives of the CTA are as follows:
o Assess environmental, social, economic, and energy
impacts of coal-based energy technologies, supply
systems, and end use.
o Identify, analyze, and compare technological and
institutional methods of avoiding or mitigating
undesirable consequences of coal-based energy
development.
o Identify, analyze, and compare alternative policies
and implementation strategies for coal-based energy
development.
In short, the study will try to anticipate what a given coal-
based energy technology mix might mean to our society and outline options
that can prepare use for that future.
To provide "settings" for the study three "scenarios" representing
divergent, but plausible, concepts of our society to the year 2030 have been
created. These scenarios are described in narrative form and structured like
mini-dramas. They contain a number of assumptions about the future covering
the state of technology, government institutions, population distribution,
economic needs, and many other factors. Included are assumptions about
emerging social value systems that can affect energy — and therefore coal —
demand.
199
-------
Based on these assumptions about future states of society, we can
project various levels of demand for coal-based energy. The next step is
to determine what levels, mixes, and timing of the development and deploy-.
ment of coal-based energy technologies will meet these various demand levels.
A number of plausible combinations are being examined. The study will iden-
tify and evaluate the social, economic, environmental, and energy conse-
quences (impacts) resulting from the various technology mixes in order to
compare the "trade-offs" between them.
The approach being taken in this assessment might be described as
a series of snapshots in various time frames. We begin with 1978, which is
really the data base used for information. Then we look at a snapshot in
three different time periods or "slices": 1985, 2000, and 2030, simply
because there is not the time or resources to look at all of the years be-
tween 1978 and 2030.
Who are the potential users of this study? At this point we
don't really know what user groups are going to use the product of the study
most effectively. It might be an environmentalist group; it might be a
local government; it might be an agency or some other federal department
that is not so obvious as the EPA and the Department of Energy. It might
be a state government like Massachusetts rather than one in the Rocky
Mountain region. Therefore, the targets, for the products of this study are
viewed as multiple targets, meaning that we do not look upon EPA as the only
user of this study.
200
-------
The major products envisioned for the Coal Technology Assessment
Program include a comparative assessment of the social, economic, and en-
vironmental consequences of technology in two time frames — the present
to 2000, and 2000 to 2030. As mentioned earlier, a bottom line of the
study is policy options, meaning what actions among alternatives can be
taken to avoid or mitigate undesirable consequences of coal development.
As in all coal development research efforts, we must identify in a long
time frame, future research and development needs. There are special fea-
tures* of the CTA program that are not present in all studies conducted by
DPA, or by other government agencies. For example, we have a National'
Technical Advisory Committee composed of eight people representing a variety
of agencies, institutions/and disciplines. The study is sponsoring inter-
ested party forums such as one held recently at Keystone, Colorado. The
initial interested party forum was held at Airlie House, Virginia, in early
March. In June we assembled a forum of experts in Washington, D.C. to iden-
tify and prioritize issues from an initial list of over 200 issues. We feel
that communications and public Involvement are critical to this study.
There are a number of modules in the TA process such as the issues
module, policies module, Impacts module, scenario module, technology module,
and communication module. By putting these together the flow in the process
of the Coal Technology Assessment is characterized. First we must Identify
Important specific Issues because we do not have the resources to examine the
whole Universe. The Issues determine the tools to be used In the study
analyses. There are several different types such as broad Issues which really
provide the context for the study. There are reqlonal Issues and environmentally
specific Issues. For example, the CTA is Intended to provide a basis for
201
-------
answering two broad questions: (1) What level of coal-derived enerqv use
is most desirable in the U.S. over the next half-century (1978-2030)? and
(2) Does it matter, i.e., from the standpoint of the comparative costs and
benefits of social, economic, environmental, and institutional impacts,
which coal technologies are deployed in meeting this leve'l of energy use?
The conclusions of this study are intended for a primary audience of public
and private decision-makers who must act regarding questions such as these
over, perhaps, the next ten years. It is clear that these decisions will
have major implications for the U.S. as it moves through the remainder of
this century and into the 21st.
Examples of major issues would certainly include the long-term
build-up of C02 in the atmosphere as a global issue; the prevention of
significant deterioration of air quality is a very significant issue in the
Rocky Mountain §gion. How we control toxic substances, including the
hazardous substances that may be discharged from coal is a critical issue.
Water supply will certainly be addressed as an issue. Another example of
an issue is how do we resolve, or can we, the jurisdictional conflicts be-
tween federal, state, and local governments, and in the West we must include
Indian tribes.
The next module in the assessment process is the structuring of
scenarios. We have so far in the study structured only three national
scenarios, presently identified as A, B, and C. As an example, we will make
assumptions on GNP, on quads of energy demand and quads of coal demand for
each of the scenarios.
202
-------
The next module deals with the selection of technology mixes;
the level of coal development that will utilize these technologies are
predetermined by the assumptions in the scenarios. If we have a business-
as-usual scenario, the level of coal development will obviously be higher
than if we have a conservation scenario. Having determined the level of
development and the rate within the time frames that the technologies on
come on-stream, we can also look at the deployment of technologies on a
regional basis. Thus we will also be examining development according to
the assumptions in the scenario, the rate the technologies are commercialized,
and the regions in which they are deployed. A coal trajectory is composed of
various modules; an extraction module, processing module, transportation
module, and so forth. When we combine these modules together, there is an
almost infinite variety of combinations identified as a coal trajectory. A
combination of trajectories identifies a technology mix.
The next assessment module is the characterization of technologies.
We have structured the scenarios, and out of the scenarios we have selected a
technology mix to meet certain coal demands. We must characterize the selected
technologies in terms of what actions are taking place that have an effect on
the environment. A few examples of the manner in which technologies can be
characterized are: How many acres of land are required? How many feet of
water will be needed? How many employees are needed in the labor pool? How
much capital for construction in the plants? and so forth. There are some
obvious outputs in terms of the water and solid waste, health and disease
Implications," For example, possible long-range effects of cancer in the
general population; toxic discharges into the air. There are numerous such
factors which can be studied numerically. However, the CTA must consider many
characterizations such as aesthetics or the quality of life to which we cannot
conveniently put numbers.
203
-------
Having characterized the technologies, we estimate the impacts
in a number of ways. What is their magnitude? Are we characterizing a
large impact or a very small one of limited duration? How is the impact
distributed? Is it specifically regional, statewide, or national, or per-
haps international? What is the political perspective? Is the impact con-
sidered by identifiable groups of people to be important to them? Will it
last only during the construction of the project and then be gone? Or will
it be a long-term effect like discharges of radiation? And when will it
occur? Today, next month, 5 years from now, or, possibly like cancer,
20 to 30 years in the future?
The next step in the TA methodology is to evaluate the impacts.
The evaluation criteria serve as a screen for selecting the Impacts that
we are going to evaluate. There are hundreds of impacts that the study
could examine. We can't look at the whole Universe, so we have to have some
criteria for selecting the impacts that we feel are of most critical importance
both politically and scientifically to whomever we are dealing with. What
is the geographic perspective for the CTA? Are we dealing with a unique
resource? Do we have the ability to assess the impact? These are some of
the key criteria used for impact evaluation.
The next module of the process is the identification of policy op-
tions. The purpose of policy analysis 1s to identify the bases for policies
that will avoid or mitigate undesirable environmental, social, or economic
consequences of the deployment of coal-based energy technologies. Or,
hopefully, to take advantage of opportunities that may be discovered. Be-
cause Impacts can be both positive and negative, two basic types of policies
must be Identified: legal and institutional. This really goes to the questions*
do we need a new law? A state law or a federal law? Do we need a new government
204
-------
organization such as a regional council? Or a new type of authority to deal
with particular problems? The technology mix policies are concerned with
such considerations as scrubbers, predpltators, boiler modification, re-
search and development on new coal-cleaning processes, and similar engineering-
oriented applications. We also must look at policies that are both techno-
logical and legal/instutional. The policies are evaluated in two different
perspectives: their scientific perspective — that is, policies that address
Issues that have scientific data bases, and policies that address issues that
are sociological and political in nature. Policy makers must address prob-
lems and issues as they are perceived by the public 1n the political process.
Those problems and issues may or may not have a good scientific data base.
Then we review and critique. And 1n our review and critique we use our tech-
nical advisory committee, those who attend the public forums, panels of experts,
and, of course, the Program Core Team and EPA staff. The review process is
very critical to the study. Finally, the findings must be communicated. We
consider communications extremely Important and must use a variety of media:
newsletters, forums, written reports, slide or film presentations.
Having reviewed the process, to use an example, let's take the issue
°f acid rain. There are many dimensions of tfvis issue. How it affects the
productivity of the land: does it mainly degrade fishery or wild life habitat
or other parts of the ecosystem? It might result In climatic changes, for
example. It is not possible to look at all dimensions; however, we can select
key ones. Using the example of acid rain, the issue must be examined within
the context of a scenario. Scenario C, or what we call buslness-as-usual will
be used. The region Identified 1s the Intermountain West. The year 1s 2000,
We're looking at one characterization: the emission of S02- Under Scenario C,
205
-------
1n the intermountain West, in the year 2000, we project an energy demand
and determine how much energy of that demand will be provided by other
sources of energy. A projection is then developed for coal-based energy
in that region and the technical mixes selected. The question then becomes,
What technologies will be required to provide eight quads of energy in the
upper midwest region in the year 2000? Can we come up with this particular
type of mix? We characterize these technologies that we have selected as
being needed. In this case, we will characterize it onlv for SOg and our
characterization is that the total emissions to the reaion. or as a total
burden, will be 9-1/2 million tons of S0«. Then we evaluate the imoacts.
We've onlv selected one tvoe of impact: the land imoact of acid rain, and
one cateaorv of imoact. Of course, there are manv imoacts on acid rain on
land; those listed are a few of the more imoortant ones.
206
-------
THE MAIN TRENDS OF WORKS ON ENVIRONMENTAL
PROTECTION AGAINST THE INFLUENCE
OF COAL-PREPARATION PLANTS IN THE USSR
I. S. Blagov, G. G. Vosnyuk, V. V. Kochetov,
I. Ch. Nekhoroshy, and I. E. Cherevko
USSR Ministry of Coal Industry
Soviet Union
Coal industry is the basis of the Soviet power fuel industry. In
1977 coal production in the Soviet Union amounted to 715.700 tons, by 1980
it will be 790.000.000 - 810.000.000 tons.
It is envisaged that further development of not only Donetsk coal
basin but such important coal basins as Kusbass,Ekibastuz, Kansk-Achinsk and
South Yakutsk coal fields.
Under the existing conditions of intensive scientific and technical
progress and rapid growth of industry the problem of rational use and reproduction
of natural resources in fuel-power complex of the country has become one of the
most important state problems, solving of which is closely connected with health
protection of present and future generations of people and their wellbeing.
High rate of coal industry development and demands for high quality
fuel have caused a considerable growth of coal preparation output in the last
few years.
In 1977 nearly 345 800 000 tons of coal were cleaned.
The development of coal preparation is based on recent achievements
of science and industry, on building of new large-scale plants and reconstruction
of existing plants.
At preparation plants use is made of modern methods and high-capacity
equipment which provides simple operation and technological efficiency of coal
preparation. Progressive methods of cleaning,heavy media separation , jigging
and flotation are widely used at coal preparation plants.Great attention is
paid to the use of closed water-slurry circuits and purification of stack gases
exhausted from dryers and fans to sanitory standards. At a number of preparation
plants wastes are used as raw materials for the construction industry. At a
preparation plant in Moscow region coal basin non-refuse technology is used,
207
-------
In accordance with coal preparation development main trends in
environmental protection against the harmful effect of coal preparation plants
are:
- protection of water resources from impurities;
- preventing of airborn pollution;
- complex utilization of mineral matter and utilization
of wastes;
- reduction of technological losses in the process of
coal preparation.
Closed water-slurry circuits,
dewatering of slurry and flotation tailings.
. Coal preparation plants of the Soviet Union are processing a wide
range of coals at various stage of metamorphism and it characterises the
features of technology and equipment and at the same time it defines physico-
chemical content of industrial water.
The improvement of technological circuits envisages the creation of
non-refuse technology. For this purpose it is necessary to solve a number of
problems:
- the use of flotation for recovery of reclaimed water;
- the use of equipment and methods for dewatering of
flotation tailings;
- the use of closed water circuits at preparation plants.
The experience of a great number of coal praparation plants using
one-stage water-slurry circuit with the recovery of washing water showed the
3 3
possibility of reduction water-slurry volumes from 10-l'2m /h to 3-Am /h per ton
of cleaned coal and drain of maximum quantity of slimy particles.
In technological circuits of coal preparation plants flotation is
the initial operation for water recovery previous to dewatering. The
improvement of dewatering in the USSR is gained' by means of increasing filtering
2
area from 80 to 250-300m by improvement of filtration technology, the use of
thickened pulp and by the use of physicochemical means of intensification. A
new disc vacuum-filter DU-250-3,75 (Fig,I) came into operation at a Kusbass coal
208
-------
O
VO
Overflow
Air
Fig.l. Disc vacuum filter DU-250-3,75
l.trap; 2.receiver; 3.disc filter; 4,pulp drain;
5,receiver; 6.receiver; 7,conveyer; 8,pump; 9. sump;
10,filtrate tank; 11,moisture trap; 12,pump.
-------
preparation plant.It has the following specification
2
- filtering area,m 250
- disc diameter, m 3,75
- number of discs 14
- number of sections in a disc 16
- frequency of rotation,rpm
discs 0,32-1,2
stirrer 30;50
- capacity, kvt
drive of discs 8,5
drive of stirrer ..».. 7,5
- vacuum, mm 500-600
2
- blowing pressure, Vgs/cm to 0,6
-dimentions, mm 9 200x4 380x4540
-mass, kg 35 000
•s
Moisture of cake is 28,9%, loss of solids is 20kg/m .
The use of such filters helps to increase the specific capacity of
filters in 2-2,5 times and to reduce moisture content of cake by 2-3%,
The use of hydrophobic reagents Improves the capacity of filters
and specifications of the process.
For intensification of thickening and dewatering synthetic polymeric
flocculants are employed. They are employed in bowl centrifuge NOGSh-1350 with
3 3
the capacity up to 250m /h and in thickeners with the capacity up to 300m /h,
which are used prior to vacuum filters, etc.
In recent years polymeric flocculant polyethylenemin has found a wide
use in coal preparation. It is highly efficient in combination with polyaery1amide.
The new trend in this field is the use of granular flocculants.
As a rule closed water circuits of coal preparation plants include
sewage treatment structures.
For compensation of technological water losses coal preparation plants
'use mine water. In 1975 the volume of reclaimed waters at coal preparation plants
3
was accounted to 847 500 000m , and in I960 this figure will increase up to
I 120 000 000m3.
210
-------
1
notation
Tailing*
Cunoeiitrele
Overfbov
i r
(Disc vacua*, filters)
Underflow nitrate
Waahii« of filter cloth
Besldne
I
9o the circuit
terms
I (filter-presses)
nitrete
Sludge
To the circuit
1
Fig.2 Process flow diagram of flotation tailings dewatering
-------
Special filtration sections equlped with cell filter-presses providing
the closing of water-slurry circuit inside of plant building are used for
improving the flotation tailings transportability characteristics. (Fig,2),
Here are the most prominant characteristics of new water-slurry
circuits:
It Flotation is used as the most efficient and simple method
of water purification;
2. Conditioning of pulp provides its constant density without
water dilution,
3, Reducing of operations number and the volume of slime water,
elimination of the escess slime circulation and the
shortening of the time of slime stay in water, help to
reduce slime formation by 30-45%,
Introduction of new water-slurry circuits provides the reduction of
solids content in recycling water (to30-80g/l) and to Increase the apparatus
efficiency.
For the purpose of reducing industrial recycling water impurities
the Investigation of reagent regimes and the use of reagents in the process of
flotation is carried out with regard to their efficiency and residue concentration,
Frothing and foam removing alcohol reagents were usually doped into
flotation machine chambers. Now at some plants they are used at the initial stage
of the process and it helps to minimize the consumption of appolar reagent in
1,5 times. . It also improves froth structure and provides normal operation of
flotation machine.
In recent years great attention is payed to the collection and
purification of rain and thawing waters. Investigations were carried out to
determine contaminants of spontaneous surface water run-offs from theterritory
of coal preparation plants to water reservoirs, On the example of a number of
coal preparation plants it had been setteled that surface rain water run-offs
are contaminated with suspended matter which contain rock and magnetite particles,
oils and petroleum products. For utilization of water run-offs it is necessary
212
-------
to use mechanical and physicochemical purification. At present there is a number
of preparation plants using purified surface water run-offs in their techno-
logical circuits. It is supposed that this experience will be shared within the
industry. It is necessary to emphasize that investigation of methods for
surfuce water run-offs purification were carried out with regard to climatic
conditions and location of coal preparation plants.
Control for the operation of industrial water treatment plants is
carried out in accordance with the existing departmental regulations. Commissioning
of these plants is exercised by State Commission Including the representatives
of Sanitary Inspection and Water Utilization and Protection Agencies.
Purification of stack gases exhausted from
dryers of aspiration systems and other sources.
Operationof coal preparation plants results in atmosphere contamin-
ation. The sources of contamination are: emissions of dryers, emissions of
boilers, emissions of industry blowers, haulage and transportation stations,
stockpiling, dried mud-settling ponds, etc.
Industrial emissions of coal preparation plants contaminate the atmos-
phere by the following ingredients: coal dust, sulphur dioxide, carbon monoxide
and nitrogen oxide. In this case emissions industrial blowers contaminate the
atmosphere only by coal dust. The main trend of scientific investigatiobs in
the field of atmosphere protection from harmful emissions of coal preparation
plants is the development and Introduction of efficient ways for reducing the
absolute quantity of harmful emissions with concentration stipulated by
legislation on environment protection(Table I).
Alongside with blowers and boilers, dryers are the source of air
3
pollution. Dryers blow out into the atmosphere up to 53 000 000 000m of stack
gases per year which contain residual dust, nitrogen oxide, etc.
With a view to a wide use of mechanized coal winning the use of new
technology for coal getting, the development of new methods for fine coal
flotation and location of coal preparation plants In the districts with
severe climatic conditions the volume of dryed coal will increase from
45 700 000 tons by 1977 to 69 300 OOOtons by 1980.
213
-------
TABLE I,
Limit concentrations of contaminants in atmosphere
of populated localities stipulated by legislation
on environment protection,
SUBSTANCES
Coal dust
Sulphur dioxide
Hydrogen sulphide
Carbon sulphide
Nitrogen oxide
Carbon monoxide
Limit concentrations,
Maximum single
concentration
0,5
0,5
0,008
0,03
0,085
3,0
2
mg/m
: Aver age concentration
: per 24 hours
-
0,05
0,008
0,005
0,085
1,0
At present investigations aimed at the search of the most effective
flowsheets and methods of dryers dust trapping have been completed.The
following measures are provided for achieving the legislation standards of gas
purification:
- introduction of three-stage system for gas stack purification;
- use of highly efficieny unloaders for the complete separation
of dried coal from gasses;
- installation of battery cyclones with the efficiency up to
98-99% and apparatus for wet trapping with the efficiency to
99,5-99,9%;
- utilization of recovered products and conducting of a conti-
nuous control for the intensity of gas emissions;
- effective control for technology with regard to stack gases
dustiness;
- search for more effective methods to obtain drying agent or
utilization of fuel with low ash content. Three-stage flowsheets
214
-------
comprising effective devices for dry and wet purification of stack gases
exhaused from dryers are now introduced at coal preparation plants.Flowsheets
for purification of stack gases is shown in fig.3
As a rule, MPR wet dusters (Fig.4) for purification of stack gases
exhausted from drying sections are placed at the third stage. Specifications
for dusters are shown in table 2.
TABLE 2.
Type
o
Capacity, m /h
0
Water consumption, g/m
Cyclone diameter, mm
2
Aerodynamic drag, kg/tn
Dimentions, mm
length
height
width
Efficiency, %
Mass, kg
: MPR-75
75 000
not less
than 50
2500
ISO
5015
2900
10380
99
4950
•
•
100 000
t
•
•
•
: 3000
•
•
: 6500
: 3160
: 12950
: to 99
6500
MPR-IOO
- 125 000
-
M
In the nearest future all drying sections will be equiped with such
devices.
Aspiration blowers at all coal preparation plants blow out nearly
3
88 000 000 000 m per year. One or two stages of dusters are used for cleaning
of suction air. When dustlness of cleaning air achieves 3 g/m two-stage
floasheets for dust trapping are used.
One more source of air pollution is coal unloading junction at
car dumpers. In recent years system of suction and dust trapping is introduced
at a number of car dampers. (Fig. 5).
215
-------
Cl«*ua«d gas
Coal
Fig.3. Scheme of stock gases cleaning
1.stock gases
2.discharge gravitation chamber
3.battery cyclone
4.wet dust trap
5.duct
216
-------
Fig.4 MPR-100 wet duster
217
-------
10
Fig. 5 Suction scheme of car dumper.
1. Air receivers. 2. Air ducts
3. Car. 4. Car dumper. 5. Coal stockpile.
6. Battery, cyclones. 7. Blower
-------
At a number of new plants closed bins of silo type are built in
order to prevent coal dust blowing out. It gives the opportunity to
minimize land area for coal storages.
Monitoring of dust concentrations in stack gases is gained by
taking stack gas samples by means of suction tube with filter inside (Fig. 6).
Except direct monitoring a method of predicting solids concentration
in ground layer of air is used at coal preparation plants.
o
The predicted maximum dust concentration (Cm,mg/m ) with regard for
emissions is calculated according to the formula:
Cm- --- s« --- where,
H V T
A - coefficient depending on temperature stratafication of
2/3 1/3
atmosphere, sec grad ;
M - quantity of impurities exhausted into the atmosphere, g/sec;
F - dimensionless coefficient accounting for sedimentation rate
of impurities in atmosphere;
m - dimensionless coefficient accounting for conditions of
emissions dlcharge;
H - hight of exhaust stack over earth level, m;
q
V - volume of stack gases, m /sec;
T - difference of stack gas temperature and the temperature of
atmospheric air, C-.
The results of calculations made according to the above mentioned
formula are similar to the results of direct dust monitoring in ground layer
of air. Besides some coefficients should be specified more accurate with regard
for various climatic areas.
Complex utilization of
mineral matter and
utilization of wastes.
Coal preparation wastes are the source of atmosphere pollution.
Besides they occupy considerable areas good for agriculture and construction.
The amount of wastes at coal preparation plants sponsored by Ministry of the
219
-------
S3
Ni
O
Fig.6. Installation for stock gas sampling.
a) Installation
b) Dust collecting tube
1. dust collecting tube; 2. micromanometer; 3. T-piece joint;
4. rubber rube; 5. fast; 6. fastener; 7. casing; 8. cover;
9. suction tube; 10. cartridge; 11. aperture for measuring static
pressure; 12. sleeve; 13. sleeve; 14. grid; 15. filtering paper;
16. cotton wool.
-------
Coal Industry of the USSR reached 46 000 000 tons. The share of flotation wastes
with high ash content accounted for 7 200 000 tons,(70%).
It is known that wastes of coal preparation plants are stable in
their elementary and granulometric composition. They can be successfully
utilized in various branches of industry, The accomplished research and
development investigations proved that the major quantity of wastes had been
found suitable as an effective raw material.
Due to possible utilization wastes can be classified as follows:
- for production of construction materials (inert concrete
aggragate, effective construction ceramics, cementing
material, etc.);
- for construction of roads, earthwork (crushed stone,
ballasting, etc.);
- for production of sulphur compounds.
At present 10 000 000 tons of wastes per year are used in construction
and chemical industry, in tracklaying and recultivation of land,
o
The designing of a new pilot plant with the capacity up to 100 000 m
of aggloporite per year has been started. The pilop plant will treat wastes
from a coal preparation plant.
The investigations for the utilization of wastes from Donetsk and Pechora
coal basins showed that one of the main factors for mass quality adjusting is
the change of raw material size that causes the change of production quality,
The top size is less than Imm. So flotation wastes can be efficiently used
for the production of ceramic construction materials.
The research Institutes have developed the recommendations for trans-
portation of flotation wastes from coal preparation plants of Donetsk coal
basin to brick plants. Wastes should be used as an admixture in brick production,
In 1977 160 ooo tons of flotation wastes were utilized at brick plants.
Flotation wastes can be used for the production of inert concrete aggregates.
The preliminary work for the construction of industrial plant in
Donetsk coal basin is successfully completed. The technology for granulation
and roasting of wastes is promissing,
221
-------
The work Is started on the utilization of coal preparation wastes as
crushed stone, filled-up ground, etc. Coal preparation wastes were used for
the making an experimental track. The utilization of shale wastes is of
particular importance. Shale wastes are widely used in highway engineering
at the North-West districts of the USSR.
Wastes from coal preparation plants of Donetsk, Pechora, Kizel coal
basins are of high sulphur content which can be the source of atmosphere
pollution (Table 3).
TABLE 3
Sulphur content in Donbass coal
preparation plants wastes
Total sulphur
Pyritic sulphur
Sulphate
Organic sulphur
The range of
index changing
%
0,7-7,1
0,63-6,80
0 - 0,20
0 - 0,77
Indexes of the most
coal preparation
plants (80%)
1,5-6,6
1,3-6,5
0,03-0,10
0,03-0,40
average
4,10
3,82
0, 07
0,21
Investigations showed that 12% of sulphur containing in wastes turns
into sulphur dioxide and 9% turns into hydrogen sulphide.
Monitoring of noxious gases concentrations resulted In defining
the regularity of sulphur compounds scattering in the atmosphere at various
distances from refuse pilling.
The investigations showed the possibility of changing the existing
system of refuse pilling. At present measures are taken to prevent spontaneous
ignition of refuse pilling. It is, for example, the construction of flat piles.
Keeping strictly to the established measures of preventive treatment it is
possible to avoid self-ignition and emission of noxious compounds into the
atmosphere.
222
-------
Great attention is paid to the stockpilling of wastes in worked out
space. As investigations showed stockpilling of flotation refuse in worked out
space of coal mines will allow to avoid a construction of ponds.
However allocation of flotation wastes in worked out space causes the
contamination of aquifer levels with flotation reagents. Investigations of
various methods of refuse cleaning showed that nitriging is the most effective
method of cleaning.
Testing of wastes cleaning and stockpilling complex is at the stage
of control. Utilization of flotation wastes for preventing self-ignition of
stockpilling is of great interest and Soviet engineers have developed the
research programme for investigation of this trend.
Sulphur Is a noxious ingredient so the development and introduction
of efficient methods for reducing sulphur content in cleaned coals and util-
ization of wastes with high sulphur content is of great importance for environment
protection.
Usually standards for ash and sulphur content in coals (concentrate )
have been fixed according to specifications of each plant.
The top ash content of steam coals depends on their consumption, that
is, burning of steam coal in stationary boilers, brick roasting, municipal
needs, etc. For providing sulphur reduction of steam coal and utilization of
wastes research and development institutes have carried out investigations on
the utilization of sulphur containing wastes.
At steam electric stations of the USSR use Is made of coals with low
sulphur content (0,2-2%). Only small part of coals with high sulphur content
is used at power stations.
Every year the volume of coals with high sulphur content becomes
smaller. The volume of mining and marketing of steam coals with low sulphur
content (0,2-1%) is increasing on account of Eastern coal basins development.
Steam coals of Moscow region coal,basin have the highest sulphur
content.
223
-------
In accordance with growing demands to the quality of commercial coals
great attention is paid to the cleaning of Moscow region coals. Coals of
Moscow region coal basin are characterized by high pyrite content.
Investigations of Moscow region coals showed that in the process of
mining and crushing of coarse coal before coal preparation concretions of
pyrite were opened and they were not bind with organic matter of coal. It
gives the opportunity to reduce sulphur content of commercial coals and simul-
taneously to separate pyrite for chemical industry, to minimize atmosphere
pollution with noxious sulphur compounds and to improve the utilization of
natural recources.
The developed classification of Moscow region coals according to
sulphur content of cleaned coal helped to solve a range of problems on complex
utilization of coal.
It is necessary to emphasize that sulphur content in pyrite of
Moscow region basin coals is high and it accounts to 42-43%.
Investigations proved the possibility of complex coal preparation
methodes development. It will help to achieve the reduction of ash and
sulphur content.
The investigations also showed that concomitant rock consisted of a
variety of clay which could be used as raw material for construction industry.
Non-refuse technology is now used at coal preparation plant of the
Kimovsky open-cast mine , the Moscow region coal basin. Clean coal blended
with fines is delivered to power stations, sulphur pyrite is used for sulphuric
acid production at chemical plants and clay is sent to construction material
works (Fig.7).
All annual output of wastes (clay) of Kimovsky coal preparation plant
is delivered to brick plant. The use of wastes helped to improve brick quality.
Clay specification correspond to specifications of the brick yard.
They are:
alumina content -30,6%
iron oxide -9,7%
losses in the
process of roasting -29,5%
moisture -23%
size -0-300mm.
224
-------
K>
10
Ul
jftiin nf m\n* coal
reparation
Jift:
l«5-
25-300—
Omrflo* Itadarflo»
(olaanad ooal) (pyrlt* and oltjr)
J 1
Dnatarlttc and separation Hvavr aadia separation
14-100B 0.75-100» I ^^daaai
Onrtlou
«v^_ °U' C25-30Q-)
Grinding - tOOaa .
I
«ga 0^25» ^ I
VS. Conoaatemt* O-IOQa* |
^S. 1 1
Underflow
pjrrlta (Z5-30QM)
*JU
w J _ ^
<
-------
For pyrite separation from Moscow region coals various types of flowsheets
providing dry and wet methods of coal preparation are developed. Problems of
development and introduction of efficient technology for pyrite separation in
the process of dust preparation and burning of high sulfur coals at power stations
are of great interest.
Reduction of coal losses in the
process of mining and cleaning.
One of the main trends for coal industry development is the increase
of open-cast mining output based on the use of advanced technology and modern
techniques. As a result, the use of existing techniques and technology In
complicated geological conditions causes the great loss of coal from thin
seams. The reduction of coal losses by the improvement of mining methods has
not been always justified. For reduction of coal losses in stockpills of
open-cast mines it is advisible to clean the diluted raw coal.
For reduction of coal losses in the process of raw coal preparation
and for improvement of cleaned coal quality simple technology and a number of
counter-flow separators have been developed.
The,advantages of these separators are: the simplicity of construction,
easy maintenance and repair and high technological coefficients providing the
reduction of coal losses in wastes.
Such flowsheets are introduced at 10 coal preparation plants. Four
types of high-slope separators KNS with the capacity to 400 tons per hour are
used for separation of diluted steam coals.
Environment control regulations for coal preparation plants are
developed and adopted by the Ministry of Coal Industry of the USSR for five
years with the indication of annual work.
USSR Ministry of Coal Industry controls the fulfilment of environment
protection regulations. Besides, control for environment protection and for
utilization of natural resources at preparation plants of the USSR Is carried
out by medical, land reclamation and water management agencies, by State Committee
of Hydrometeorology and Environment Control of the USSR, etc.
226
-------
In conclusion, it is necessary to emphasize that research and
development cooperation between plants and institutes of the USSR Ministry of
coal industry and USA environment protection departments in the field of
development and introduction of efficient environment protection methods
can be useful for both countries.
227
-------
THE CLEAN FUEL SUPPLY: FACTORS AFFECTING
U.S. AND EUROPEAN S02 EMISSIONS IN THE MID-1980*s
Anthony Bromley and Gary J. Foley
Organization of Economic Cooperation and Development
2, rue Andre-Pascal
75775 Paris, France
ABSTRACT
This paper analyzes the factors affecting the availability of low-sulphur
fuels and the introduction of fuel desulphurization technologies in the OECD
region up to the mid-1980*s. The analysis examines energy scenarios for
North America, Western Europe, Japan, am Oceania, developed by the International
Energy Agency in the context of the various S02 emissions reduction policies
now being contemplated by the OECD member countries. The most probable
forecast of 1985 S02 emission for OECD as a whole is 57 million metrictons,
a 23 percent increase over the 1974 levels.
228
-------
I. INTRODUCTION
This paper summarizes a report published in 1978(1) by
the Environment Directorate of the OECD on the problem of
limited supplies of clean fuel for the OECD Member countries.
With increasing energy consumption, the supply of clean fuel
may not be sufficient to meet future levels of demand created
by more countries seeking to reduce emissions of sulphur oxides,
The term "clean fuel" is used here to refer to fuels from which
there are low emissions of sulphur oxides, principally sulphur
dioxide (S02). These can either be naturally low in sulphur,
or be desulphurized prior to combustion. If the factors which
affects this supply of clean fuel can be identified, it may be
possible for governments to take action where required,
individually and collectively, to increase the supply of clean
fuels and fuel cleaning techniques. It will also be possible
to determine when it makes more sense to desulphurize the
gaseous combustion products thus allowing high sulphur fuels
to meet low emissions standards for S02-
All OECD countries have recognized the need to attain
acceptable ground level concentrations of SOp and many have
Clean Fuel Supply - Factors Affecting SOP Emissions in the
Mid-1980's, OECD, Paris, 1978.
229
-------
implemented ambient air quality standards and/or emission
standards for it. The transport of S02 across frontiers
has also concerned many countries. The overall solution
to the problem will only be found through international
cooperation in which national policies are implemented to
attain acceptable ambient air quality, while at the same
time minimizing transport across frontiers.
This paper analyses the situation in the three
geographical regions of OECD, namely, Europe, North America
and Japan. (Australia and New Zealand have been excluded.
However preliminary calculations show that total SOg emissions
from fuel combustion in these countries are relatively low,
although these may be concentrated in certain areas).
The time frame for this report is 1985. This recognizes
the fact that any decisions taken on clean fuel supply or fuel
cleaning techniques would not have any major impact until the
mid I980's because of the long investment lead times involved.
Only those technologies that have already been corameercially
developed are considered as being available for wide scale
application for 1985. Therefore, the report does not include
such technologies as liquefaction, fluidized bad combustion
and chemical cleaning of coal.
II. METHOD
The approach used by the OECD Environment and Energy
Group is based on the study of the recent estimates for the
years 1974 and 1985 of energy consumption and supply prepared
by the International Energy Agency for the OECD for 1974 and
1985 (World Energy Outlook, IEA, 1976). Since the sources
of supply and the distribution of sulphur in fuel sources over
that time period are already fairly much determined, it is also
possible to make projections of sulphur oxide emissions. These
projections are presented in Table I.
230
-------
(10° metric tons of S02)
to
OJ
1985
1968 1974 Reference Case
OBCD
Europe 16.67 19.7 22.1-25.4
North
America 26.6 24.2 24.9-28.1
Japan 4.0 2.4 2.6-3.1
TOTAL 47.3 46.3 49.6-57.1
1985
Accelerated Policy Case
19.6-21.4
22.4-25.7
1.7-3.3.
43.7-50.4
TABLE I
SIMMAPY OF ESTIMATED AND FORECAST SO EMISSIONS IN THE OECD, 1968 to 1985
-------
The study shows that, for OECD as a whole, total
emissions from fuel combustion in 1974 were about the same as
in 1968, despite a 26 percent increase in fuel consumption.
The SOp emissions from fuel combustion were forecast for 1985
assumed no increase of desulphurization capacity over that
already installed and planned. In the "worst" case there
could be an increase of about 23 percent over the 1974 levels.
In the "best" case, assuming that countries strive towards their
energy independence objectives, (accelerated development of
indigenous resources, increased conservation and increased use
of low sulphur oil) there could be a decrease of the order of
6 percent or some 2.7 million metric tons of S02. The effect
of energy policy can be seen to be critical.
If the OECD countries were to achieve the goals of
energy conservation and development of indigenous energy
resources set out in the Accelerated Policy Case, the sulphur
emission standstill trend which has existed in OECD from 1968
to 1974 would probably continue up to 1985. However, the
Accelerated Policy Case has been Judged unrealistic for most
OECD countries and the Reference Case, which assumes a continu-
ation of present policies governing energy supply and conservation
by OECD countries, is now considered to be the better forecast.
There are also fears that even certain expectations in the
Reference Case may not be realized. Perhaps the most important
factor will be the share of total energy that nuclear power is
forecast to provide. If nuclear output falls short of projection*
there will be a tendency to shift to fossil fuels and a consequent
increase in S02 emissions. In this situation oil will be the
balancing fuel and will of necessity be made up of the medium to
high sulphur Saudi Arabian crude oils.
Given this general energy situation, the next step was
to investigate the potential reductions in S02 emissions that
couUbe achieved in the various OECD regions. Table II gives
the distribution of S02 emission between coal and oil use in
232
-------
Region & Fuel
% of Total SO
^W^_»«l^^_»«V«^—_^«B«B_^
Emissions 1985
Max. Potential for
Further Reduction *
(Percent of total SO emissions)
I. OECD Europe
Oil
Coal
67
33
100
HDS « 33%, FGD,= 36%
FGD • 25%, Coal washing
5%
II. Japan
Oil
Coal
90
10
loo™
HDS « 18%, FGD 22%
FGD, Coal Washing * Neg.
III. USA
Oil
Coal
20
80
IfflT
HDS, FGD » Neg.
Coal washing - 20%, PGD - 20%
IV. Canada
Oil
Coal
74
26
HDS » ? PGD = 6%
Coal washing » ?, PGD » 10%
* Over currently projected control measures for SO in 1985
TABLE 2. DISTRIBUTION OF SO EMISSIONS
A
233
-------
1935 in OECD and the possible improvements that could be
achieved by the available sulphur reduction technologies.
For Japan, where oil is the dominant source of emissions,
coal cleaning can have no real effect. Therefore, the
emphasis must be on oil desulphurization prior to combustion,
or flue gas desulphurization afterwards.
For OECD, Europe, despite the fact that coal contributes
one-third of the S02 emissions, the potential reduction by
coal washing is only about 1%. This is due to particular
factors in the U.K. and Germany, the two principle coal consumers
in this region. In the U.K. approximately 80% of the coal
combusted now receives some degree of washing primarily to
reduce ash content and upgrade heating value. . In Germany,
also, about 35% of the hard coal produced is washed. Furthermore,
increases in German output are expected to be in the form of
lignite, in v/hich the sulphur content is essentially 100%
organic, and thus not susceptible to removal by physical coal
cleaning methods. Consequently, for Europe, the emphasis for
S02 reduption will have to be FGD and oil desulphurization,
although some improvement is also possible through increased
coal washing, and careful matching of washed coal products with
combustion clean up technology.
In OECD North America, both Canada and the U.S. show
considerable potential for reduction in emissions from coal.
In U.S. the maximum estimated reduction would be 20% from coal
washing. This is due to the overall high pyritic sulphur
content of U.S. coal, the relatively high proportion of energy
produced from coal; and the relatively low proportion (under
25/0 assumed to be washed. Canada shows the same pattern,
although Canadian total S02 emissions amount to less than 10%
of the U.S. The next stop of the study was to examine in
detail the different sulphur reduction options in each region;
and the associated costs, in order to arrive at an estimate of
the most feasible strategy for emissions reduction.
234
-------
III. COAL CLEANING FOR OECD EUROPE
As the physical properties of coal and coal cleaning
practices and costs vary considerably from one coal producing
country to another within the OECD European region, and
economic analysis of coal cleaning in the region as a whole
would not be sufficiently representative for the coals of
any one country. In addition, coal properties and, therefore,
cleaning potential are best documented for the two ma^or
producers the Federal Republic of Germany and the United Kingdom.
By 1985 Germany plans to clean all coal to produce a clean
product having an average sulphur content of about 1.0 percent
down from about 1.39 in its natural state. The present coal
cleaning practice in the United Kingdom is not oriented
specifically towards sulphur reduction but reaches towards
producing a coal of uniform heating value and low ash content.
Taking the United Kingdom as an example for Europe, costs have
been analysed for desulphurization by coal cleaning.
In considering the current U.K. coal cleaning practice
there is potential for reorienting this practice to increase
sulphur removal. Three levels of cleaning can be examined:
(a) cleaning to produce a single product coal of uniform
heating value and ash content,
(b) cleaning at two specific gravity separations to
redistribute sulphur into a clean low-pyritic sulphur
coal product and a middlings product with higher
sulphur, and
(c) redistribution as in (b), cleaning of the middlings
to reduce sulphur and blending with clean coal product.
The redistribution of sulphur into the two product streams
does not in itself reduce the overall sulphur content of the
coal. It is through further processing of the middlings, for
example by regrinding, more washing and froth flotation, that
the S02 emissions reduction can be achieved. Alternatively,
the middlings may be used where FGD is applied to the post-
combustion gases. The costs of these two alternatives may be
compared.
235
-------
The costs for a new coal cleaning plant for each of
these levels of cleaning have been estimated by a group of
experts, as shown in Table III. The operating costs in
this table include only operation and maintenance. To
complete the analysis, it is necessary to add the annualized
capital charges which are estimated at 20 percent of the
capital investment per annual ton of capacity. For 3,000
hours per year of operation, this amounts to 81.70 per
metric ton. The 1976 total cleaning cost would be $2.80
per metric ton raw coal ($4.70 per metric ton washed coal).
In 1974, the U.K. coal consumption v/as 94 x 10 metric
tons of which about 60 x 10 metric tons were cleaned. In
1985 the coal consumption is forecast to be slightly higher.
The annual operating cost in 1985 to produce coal of uniform
heating value and ash content is $6.10 per metric ton of
washed coal (escalated to a 1980 dollars basis). To redistribute
the sulphur in the coal (level b) or reduce sulphur further
(level c) would increment the annual operating cost of $6.10
per metric ton by the costs shown in Table IV. The total cost
of $7.50 - 8.25 per metric ton would represent the sulphur
premium for desulphurized coal (level c) over washed coal
and the cost of $1.40 - 2.15 would represent the premium over
washed coal (level a).
It has been estimated that washing U.K. coal at specific
gravities of 1.3 and 1.8 would produce about one-third middlings
at 2.5 percent to 3.0 percent sulphur and about two-thirds
clean prdduct at 1.0 percent sulphur. As the organic sulphur
content of the coal (0.8 to 1.0 percent) would be almost the
same for the clean product and the middlings, the pyritic
sulphur content of the middlings would be in the range of 1.5
percent to 2.0 percent sulphur. It is expected that further
processing could reduce the middlings sulphur content from 2.0
to 1.8 percent with a minimum loss of coal to the waste. The
236
-------
(Currency at December 1976 values)*
Capital Costs
(currency/metric ton
of raw coal/hr)
Operating Costs
(currency/metric ton
of raw coal)
FRG
UK
Base cost to
produce coal
of uniform
heating value
and ash content
Incremental cost
over base
(redistribute
pyritic sulphur
into product and
middlings)
Incremental cost
over base
(redistribution
and reduction of
sulphur by
secondary process-
ing of middlings)
* $1.00 » 2.50 DM,0.60£
30-40,000 DM' .£15,000
($12-16,000) ($25,000)
+ 12%
(assumes dense medium
cyclones at
SG - 1.30-1.32)
FRG
2.5-3 DM
($1.00-1.20)
5-16%
UK
70 p
($1.17)
25-32%
20-40
TABLE 3: COSTS FOR NEW COAL CLEANING PLANTS
237
-------
ultimate coal sulphur content, if clean coal and middlings
were blended after all processing, would be about 0.25 percent
sulphur lower.
From Table 4 the incremental cost of sulphur removal
(level c) over washed coal (level a) is in the range $560 - 860
per metric ton of sulphur. However, if. existing cleaning
practice is such that only some small percentage of coal is
cleaned, then to the above incremental cost it would be
necessary to add 'a share of the cost for coal washing, that
is the cost of increasing the percentage of coal cleaned,
For example, at present about 50 percent of U.K. coals for
electricity sector combustion are washed. To increase the
sulphur removal on these 50 percent would cost .';i>560 - 860
per metric ton sulphur removed. If the other 50 percent of
the coal were to be treated for sulphur removal a 50 percent
share of the base cost would be included, making the average
cost for all 100 percent of coal equal to $1,800 - 2,100 per
metric ton of sulphur removed.
In conclusion, for the countries which wash a large
percentage of combustion coal production, a cleaner low sulphur
product can be obtained at an incremental cost of $1.40 - 2.15
per metric ton of coal or at a sulphur removal cost of
$560 - 860 per metric ton of sulphur. Sulphur removal costs
increase if a lower percentage is generally washed. For
Europe in 1935, if all combustion coal could be reduced by
0.25 percent sulphur content.on average by this method, the
quantity of sulphur removed would be 0.5 - 0.6 x 10 metric
tons. The potential removal is limited, but coal cleaning
may be the most economic method for coal desulphurization.
IV. FLUE GAS DESULPHURIZATION
The basis used for flue gas desulphurization costs for
coal-fired pov/er plants assumes particulate control to 0.1 lb/
238
-------
Incremental
Operating Cost
($/metric ton
washed coal)*
Increment of
Sulphur
(% wt. S)
Incremental Cost
of Sulphur Removal
($/taetric ton of sulphur}*
Redistribute
sulphur into
product and
middlings
0.55-0.85
U>
\O
Redistribution
and secondary
processing of
middlings
1.40-2.15
0.25
560-860
$ are 1980 US $
TABLE 4: SULPHUR REMOVAL AND INCREMENTAL OPERATING COSTS OVER BASE COST FOR REDISTRIBUTING
PYRITIC SULPHUR AND REPROCESSING MIDDLINGS, UK
-------
million Btu.** The "basis and the cost equations are
presented in Table 5. The costs shown assume sulphur
contents for each solid fuel as follows:
hard coal: the European average of 1.35/o S
in the 1985 Reference Case,
middlings: 2.75/5 S, within the range expected
for washing U.K. coals at specific
gravities of 1.3 and 1.8 to produce
clean product and middlings,
lignite: the European average of 1.15/5 S in
the 1935" Reference Case.
In determining the sulphur reduction, the sulphur
normally retained in the ash has not been included as part
of the reduction.
As expected, the costs are much higher for FGD on
hard coal-fired power pQants because of the relatively low
sulphur content of European hard coals. The FGD costs for
new plants ranging from $1,039 - 1,783 per metric ton of
sulphur removed are much higher than coal cleaning costs,
$560 - 860 per metric ton of sulphur removed. However, in
the case of U.K. coals for electricity sector combustion
where only 50 percent are washed, the cost of coal cleaning,
$1,800 -.2,100 per metric ton of sulphur removed is higher.
It is also clear that retrofit of FGD on existing coal-fired
plants would rarely be economical.
For lignite combustion, the costs of FGD are much more
attractive. Since lignites are often very low in pyritic
sulphur content and cannot be desulphurized by washing, FGD
is the only option for sulphur removal. Since over 80
percent of European lignites will be consumed in power plants
*** metric equivalent is 0.18 kg/kcal
240
-------
AVERAGE ANNUAL OPERATING COST (Size in MW; H in hours/yr operation)
Basis: Limestone Scrubbing/ 1980 Cost Basis ($/metric ton S removed)
90% S removal, with sludge fixation and
disposal, with flue gas reheat, and,
particulate control to 0.1 Ib/million Btu of heat
input.
!• FGD on New Coal-Fired Power Plant (particulate control)
Cost fsizel 0.7 0.7 fsizel
[6708 + 676(%S) ] -H7OT) [0.664 + 0.1019 (%S)](H)
Tsite *
(metric tons/yr) « 1.457(%S)QH)(500
Sulphur removed (metric tons/yr) « 1.457(%S)QH)(500 jj
II. PGQ on Existing Coal-Fired Power Plant (particulate control)
Cost fsize"lo.7 0.7
5 J (R) [6708 + 676(%S) J
I [0.859 •(• 0.1984(%S)] (H)
a) R is 1.2 for easy retrofit and 1.4 for difficult retrofit
Size
b) Sulphur removed (metric tons/yr) - 1.490(13) (H)(555~)
TABLE 5: SUMMARY OP FGD COST EQUATIONS
241
-------
- 9 -
in 1985, the potential exists to remove 1.2 x 10 metric
tons SOp from new lignite-fired power plants and 0.7 x 10
metric tons S02 from.retrofit of the newer and larger
existing lignite power plants.
V. COAL CLEANING vs FLUE GAS DESULPHURIZATION FOR OECD EUROPE
For hard coal, it may be more economical to use a
combination of coal cleaning and FGD where the coal cleaning
is used to segregate a middlings product for use in power
plants with FGD.
From Table 4 the incremental cost of segregation of
washed coal into clean coal and middlings is $0.55 - 0.65
per metric ton of coal washed. Similarly the incremental
cost of segregation of an unwashed coal is $6.65 - 6.95 per
metric ton of coal washed. The net segregation cost in
1985 would be proportioned to the mix of washed and unwashed
coal in a given country.
After segregation, two options exist for the middlings.
The first is to further the process middlings at the cost of
$0.85 - 1.30 per metric ton of washed coal. The second is
to use the middlings in a power plant with FGD at operating
costs ranging from $598 - 988 per metric ton of sulphur
removed for new power plants, or $14 - 24 per metric ton of
coal combusted.
Table 6 summarizes the total cost for these two options
for two cases of different washing practices. For the option
of segregation with further processing of middlings, the cost
per ton of coal in the table are simply the total cost of
segregation and further process of middlings distributed over
only the cleaned coal product (approximately two-thirds of
the products).
242
-------
Option 1
Coal Segregation and Processing
of Middlings (no POD)
Option 2
Coal Segregation with
Middlings to PGD
Normal Practice for
Combustion Coal
($/tetric t.
clean coal
product)*
($/netric
(tons S)*
$/metric
ton coal)*
(^/metric
(ton S)*
100% washed coal
03 50% washed, 50%
unwashed coal
1.25-1.95
5.80-6.55
560-860
1800-2100
6.00-10.20
9.10-13.20
770-1300
1160-1680
* 1985 operating costs in 1980 US $
TABLE 6* INCREMENTAL COST FOR SULPHUR REMOVED BY COMBINED COAL WASHING AND PQ>
-------
For the other option of segregating middlings for
use with FGD, it is necessary to have a price differential
between the cleaned coal and the middlings which would be
sufficient to offset the cost of FGD. In this case, the
cleaned coal product must also bear the cost of segregation.
The cost per ton of coal in the table represents the price
margin above the base cost for washed coal necessary to
establish the price differential. For the example in Table
6 this price differential would be set at $14 - 24 per ton
of coal.
It can be seen from Table 6 that, if the existing
practice is to wash a high proportion of power plant coal,
then cost of optional sulphur removal at the coal cleaning
plant in .terms of incremental cost per ton of sulphur
removed for improved sulphur removal in the coal cleaning
plant is much less than the cost of producing high sulphur
middlings for use with FGD (option 2). If the existing
practice is to wash a lower proportion of power plant coal
then the reverse may be the case. However, it should be
remembered that, in all cases, FGD has the potential to
remove much larger amounts of sulphur than does the coal
cleaning.
If, in Europe in 1985, all power plants greater than
100MW capacity constructed since 1975 were to use segregated
middlings as in option 2, then the sulphur removal would be
1.4 x 10 metric tons from the combustion of about 60 x 10
metric tons middlings. The total annual operating cost for
such a strategy is in the range $1,100 - 2,400 x 10 . In
comparison, for sulphur removal from segregation and further
processing of middlings for all combustion coals would be only
0.5 - 0.6 x 10 metric tons.
Another cost analysis of coal cleaning with scrubbing
for sulphur control was carried out by the U.S. EPA*. The
*Coal Cleaning with Scrubbing for Sulphur Control: An
Engineering/Economic Summary, EPA-600/9-77-017, USEPA,
August 1977.
-------
report examined a number of case studies combining some
physical coal cleaning of some of the combustion gases to
meet the U.S. Federal stankrd of 1.2 Ibs/MBTU for S02
emissions. It was concluded that, in many cases, the net
cost of physical coal cleaning followed by scrubbing of
part of the flue gas to meet standards is substantially less
than that associated with using only a full scale scrubbing
system. However, this conclusion depends upon cost benefits
from using clean coal, such as increased heat content,
transportation savings, ash disposal savings and pulverising
savings. Whether this approach would have the same cost
benefit in Europe would therefore be dependent on whether
the benefits of using cleaned coal are already being realized.
In countries where a significant proportion of coal is not
already cleaned the combined approach may have a cost benefit.
In conclusion, it is apparent that for most European
hard coals, coal cleaning to reduce sulphur content will be
the most economical approach either because of lowest incremental
cost or because of other benefits derived from the use of clean
coal. However, if further sulphur reduction is needed beyond
that which is possible with coal cleaning, a high sulphur
middlings should be segregated for use in a power plant with
FGD.
North America
Performing a similar analysis for N rth America,
begins with projections of FGD in 1985. For the eleven
year period from 1974 to 1985, the North American fossil
fuel fired electric generation capacity is expected to grow
as shown in Table 7. During this period, it is estimated
that there will be 48 G¥ of expansion capacity and about 20
G¥ of new capacity to replace retired capacity.
A group of experts on flue gas desulphurization was
asked by the OECD to forecast the amount of FGD capacity in
245
-------
to
**
o\
!
1974
1980
1985 Ref .
1985 AP
Fossil
Fuel
Capacity
GW
375
404
425
423
Average
Load
Factor
0.51
0.51
0.56
0.54
Sulphur Containing Fuels
Coal
TWh
993
1081
1415
1362
Lignite
OWh
19
29
42
41
Oil
IWh
326
348
334
216
Total
Fossil
TWh
1721
18Q5
2086
2017
TABLE 7 THE NORTH AMERICAN POWER PLANT FORECAST FOR 1985
-------
North America in 1985. Their estimate was as follows:
1976 (G\0 1985 (GW)
United States 6.5 46.4-80.0
Canada - 3.0-5.0
OECD 6..5 49.4-85.0
The United States estimate depends on the number of utilities
that must retrofit with flue gas desulphurization and the
ability to meet the NSPS with other technologies by 1985.
In the United States, it is expected that the flue gas
desulphurization capacity will be on coal-fired units only
burning coal with a 2.0 percent sulphur average, with a
boiler load factor of 0.65 average and an FGD removal
efficiency of 90 percent and reliability of 100 percent.
Then the emissions reduction can be calculated to be 4.3 - 7.4
million metric tons SOg. It is alos expected that the design,
operation and maintenance of FGD units by 1985 will have
reached a state of expertise such that 100 percent reliability
during the period of boiler operation will be realizable. An
average sulphur content of 2 percent for coal was used on the
basis that the U.S. average coal sulphur content will be about
1.6 percent S in 1985 but that the low sulphur fraction will
be used for plants without FGD, leaving a higher sulphur
fraction for use with FGD.
In Canada, it is expected that half of the flue gas
desulphurization capacity will be solely coal-fired units
only in the eastern provinces from Ontario eastward, burning
high sulphur, coals of 3.0 percent S, with a boiler load factor
of 0.55 and an FGD removal efficiency of 90 percent and
reliability of 100 percent. Then the emissions reduction
from coal can be calculated to be 0.18 - 0.29 million metric
tons S02« The other half would be on oil-fired boilers
with 2.8 percent S oil. This would produce emissions
reduction from oil of 0.09 - 0.15 million metric tons of S02-
247
-------
If the forecasts of FGD capacity in 1985 are met,
the North American electric: power producers will have
installed the maximum technologically feasible flue gas
desulphurization capacity.
U.S. Coal Cleaning Potential
The present coal preparation practices separate the
raw coal into a clean coal and a waste. In operating the
units, there is a trade-off between the sulphur removal
and the loss of coal into the waste. For the purpose of
analysis those levels of coal preparation are defined:
Level A - Coal is crushed to 1-1/2 inch top size
and beneficiated with 90 percent BTU
recovery;
Level B - Coal is crushed to 1-1/2 inch top size
and beneficiated with 80 percent BTU
recovery;
Level C - Coal is crushed to 3/8-inch top size and
beneficiated with 80 percent BTU recovery.
It should be noted that the resulting estimated sulphur
contents of cleaned coal Table 8 are those obtained by float-
sink analysis at the given Btu recoveries, and they may differ
from the actual values obtained in commercial coal cleaning
plants. There are two major factors which affect the sulphur
removal by coal cleaning:
(a) Inefficiency of coal cleaning equipment: The sulphur
content of commercially cleaned coal is generally
higher than that obtained float-sink analysis if the
Btu recoveries are the same.
(b) Btu recovery: In general, for any given coal, the
lower the Btu recovery, the lower the sulphur content
of the cleaned coal. The average Btu recovery in
commercial coal cleaning plants is approximately 80
percent.
A forecast of the quantity of U.S. coal to be cleaned
and estimates of the quantity of sulphur which could be
removed in 1985 are summarized in Table 8 which shows that
the total coal to be produced in that year to be 1040 x 10
248
-------
K>
•e-
v£>
Region
Northern
Appl.
Southern
Appl.
Eastern
Midwest
Western
Midwest
Western
Total
! •
Total
??8an.(106
short t.)
183
322
156
9
370
1040
1
Estimated
Mech.
Cleaning
(106 short t.
Cleaned Coal)
88.3
133.2
97.6
0.9
53.2
373,2
Total Sulphur Content
(wt.per cent)
Raw
Coal
3.01
1.08
3.92
5.25
0.68
2.20
Level A
Cleaned
2.06
0.97
2.73
3.91
0.54
1.63
Level B
Cleaned
1.83
0.96
2.57
3.76
0.51
1.53
Level C
Cleaned
1.61
0.94
2.47
3.45
0.53
1.45
Estd. Sulphur Removal
(106 short t.)
Level A
Cleaned
1.13
0.31
1.59
0.02
0.11
3.16
Level B
Cleaned
1.71
0.52
2.27
0.03
0.18
4.71
Level C
Cleaned
1.90
0.55
2.37
0.03
0.17
5.02
TABLE 8:
MECHANICAL CLEANING OF BITUMINOUS COAL AND LIQJITE IN 1985
-------
short tons of which an estimated 373.2 x 10 short tons
will be mechanically cleaned. The quantity of sulphur
removed is estimated to be in the range 3.2 x 10 short
tons.
The coal to be used for coking and export in 1985
is estimated to be 182.8 x 10 short tons for the Reference
Case (averaging 1.0 percent sulphur). Hence, mechanically
clean coal used for combustion is estimated to be 190.4 x 10
short tons with the average sulphur content for all combustion
coals in. the range 1.57 to 1.64 percent.
For the Accelerated Policy Case, the average sulphur
content of coals for combustion domestically would be in the
range 1.57 to 1.59 percent.
Although the above estimates of the quantity of coal
to be cleaned in 1985 for combustion purposes are likely to
be conservative, the most conservative Level A cleaning was
assumed in forecasting the 1985 emissions for the U.S.
In order to determine the maximum possible sulphur
removal, it is useful to examine how much mechanical cleaning
of coal is feasible. The quantity of sulphur removed would
be augmented considerably if 100 percent of the high sulphur
Northern Appalachian, Eastern Midwest and Western Midwest
V
coals were to be cleaned. Since little benefit would result
by increased cleaning of the low sulphur Southern Appalachian
and Western coals, the present percent of cleaning is not
changed. At Level A cleaning, the estimated sulphur removal
would be increased to 5.56 x 10 short tons from the 3.16 x 10
short tons shown in Table 8. For Levels B and C, the estimated
6 6
removal would be 8.12 x 10 and 8.73 x 10 short tons sulphur
respectively. Hence the total increased from 190.4 x 10
short tons in the Reference Case to 351.6 x 10 short tons,
and the average sulphur content for all combustion coals would
be in the range 1.30 to 1.45 percent. Similarly for the
Accelerated Policy Case, the average for all combustion coals
would range from 1.30 to 1.44 percent sulphur.
250
-------
An alternative strategy would be to clean the coal
at two specific gravity separations to redistribute sulphur
into a clean low sulphur coal product and a middlings
product with higher sulphur. The advantage in doing this
is that, by either further cleaning the middlings or applying
post-combustion emission control to their combustion, the
SOg emissions reduction is maximized without the need to
progress to the cleaning of the total coal production to
lower overall sulphur levels.
Table 9 summarizes the estimated sulphur content of
coals in 1985 if all coals were processed through a 1.3
specific gravity separation to produce a clean product,
followed by a 1.6 specific gravity separation to produce
a middlings product. It shows that the clean product
(1.3 s.g. float) from each region would contain much lower
sulphur content than the middlings (1.3 s.g. sink and 1.6
s.g. float).
In Table 10, the strategies are combined in the best
manner to produce two coal products one for combustion
without flue gas desulphurization (FGD) and the other requir-
ing FGD. This would result in 667.7 x 10 short tons of
combustion coals in the 1985 Reference Case having a 1.07
sulphur content and the remaining 189.5 x 10 short tons
having 80 percent of the sulphur removed by FGD giving them
an effective average sulphur content of 0.60 percent.
However, the constraints on achieving these levels
of cleaning by 1985 are many. N^t only is a greatly increased
cleaning plant capacity required but also the quantity of
raw coal mined, would need to be greater to compensate for
the heating value loss during cleaning.
251
-------
K>
Ui
to
Region
Northern
Appalachian
Southern
Appalachian
Eastern
Midwest
Western
Midwest
Western
Total
Total Coal
Production
(106
short tons)
183
322
156
9
370
1040
Clean Product
(1.3 SG Float)
(106
short tons)
69.8
185.1
84.5
4.2
217.1
560.7
Middlings
(1.3 SG Sink
1.6 SG Float)
(106
short tons)
113.2
136.9
71.5
4.8
152.9
479.3
Total Sulphur Content
(wt. %)
Raw Coal
3.01
1.08
3.92
5.25
0.68
2.20
Product
1.41
0.86
2.35
2.93
0.55
1.05
Middlings
2.56
1.10
3.53
4.74
0.60
1.68
TABLE 9: POTENTIAL SULPHUR REDUCTION OF US COALS IN 1985 AT 1.3 AND 1.6
SPECIFIC GRAVITY SEPARATION
-------
VI. CONCLUSION
A. OECD North America
The study shows that the potential SOp emissions
in North America in 1985 will come primarily from coal
combustion in power stations. However, since the installed
and planned flue gas desulphurization capacity in North
America is large (50-80 GW), much of these potential
emissions from power stations will be controlled. North
America must achieve its planned FGD capacity to avoid an
increase of emissions of 20-40 percent above the forecast.
As a result of this control technology and the fact
that coal is domestically produced, the 1985 emissions in
North America will be much less sensitive to the energy and
fuel import policies than will be the emissions in Europe.
The range shown in the summary Table 1 reflects only the
uncertainty in the amount of flue gas desulphurization to
be installed.
North America is the one OECD region where coal clean-
ing represents a method for substantial SOp emission reduction
by 1985. This is due in part to the expected increase in
coal use by 1985 and also to the fact that currently only a
relatively small proportion of production for combustion is
washed.
The segregation of coal in the washing plant into
high and low sulphur fractions is feasible. This would
permit higher sulphur coals and middlings from the washing
plant to be used in the large number of power plants with
flue gas desulphurization. The low sulphur coal million
metric tons of coal averaging 1.0 percent sulphur content
could be segregated out of the total 857 million metric tons
of combustion coal.
253
-------
B. OECD Europe
In order to achieve the maximum reduction of total
SOp emissions, following European strategy would have the
lowest cost of sulphur removal (1980 dollars):
segregate high sulphur and medium sulphur
residual fuel oils,
. install flue gas desulphurization in all oil-
fired power plants over 200 MW,
desulphurize by direct residue desulphurization
the remaining high and medium sulphur residual
fuel oils to a level of 0.5% S ($630-810/metric
ton of S removed),
physically wash all hard coals to minimize
sulphur content ($560-2,100/metric ton of S
removed),
. install PGD on all lignite-fired boilers over
100 MW and constructed since 1967 ($520-890/
metric ton of S removed),
require that all imported coals be washed to
minimize sulphur content,
. use naturally low sulphur or cleaned fuel in
the domestic, commercial and small industrial
sector where FGD is not practical.
This strategy would remove approximately 13 x 10
tons of S02 from residual fuel oil combustion in 1985 at
a total operating cost in 1985 of approximately $4 billion.
It would also remove approximately 1 x 10 tons of S02
from hard coal combustion at an operating cost of approximately
$0.35 billion and approximately 2 x 10 tons S02 from lignite
combustion at $0.6 billion. In the 1985 "worst" case, S02
emissions would be reduced from 25.4 x 10 metric tons of
SOg to approximately 9-10 x 10 metric tons at an annual
operating cost of $5 billion.
If it is desired to maintain emissions at their
present level of 20 million metric tons of S02 rather than
to obtain the maximum emission reduction, a standstill
254
-------
strategy could be put into practice by 1985. In the
"worst" 1985 case this would mean a reduction of 6 x 10
metric tons S02. The washing of all hard coals could
reduce S02 emissions by 1 x 10 metric tons at an annual
1985 operating cost of approximately $0.35 billion.
Installation of FGD on all new (post 1974) lignite-fired
boilers over 100 MW would reduce S00 emissions by another
£ C.
1 x 10 metric tons at $0.3 billion in 1985. The remaining
4x10 metric tons reduction could ,be accomplished by
segregation of 4.096 S fuel oil to new power plants (post 1980)
with FGD or by direct desulphurization of high sulphur
residual oil to 0.596 S. (The cost of low sulphur fuel oil
would be incremented by $7/metric ton). The 1985 operating
cost would range from $1.0-1.25 billion for these two options.
The S02 emissions could b« reduced by only 3 x 10 metric
tons by .the purchase of additional cost of about $1.75 billion,
A standstill strategy which would require the removal
of about 6 x 10 metric tons of S02 in 1985 would result in
a 1985 annual operating cost of $1.55-1.80 billion.
255
-------
A TECHNICAL AND ECONOMIC OVERVIEW OF COAL CLEANING
11 2
Horst Huettenhain , Jackson Yu , and Samuel Wong
Bechtel National, Inc.
San Francisco, California 94119
2
Argonne National Laboratory
Argonne, Illinois 60439
ABSTRACT
This paper is based on work performed by Bechtel National, Inc., San
Francisco, California, for Argonne National Laboratory, under an ongoing
program on Environmental Control Implications of Generating Electric Power
from Coal, sponsored by the Assistant Secretary for Environment, DOE, Division
of Environmental Control Technology. From the broad subject of coal cleaning,
the paper is limited to a technical and economical overview of physical coal
preparation and cleaning technologies (CPC).
CPC processes can be divided into comminution, classification, separation,
dewatering, and drying. A run-of-mine coal passes through all or any combi-
nation of these processes on its way to the consumer. Each process is performed
in unit operations-which may utilize various types of equipment and process
principles. The unit operations are combined in CPC plants, in sequences
depending on the raw coal characteristics and the extent of coal preparation
and cleaning effort required to produce a marketable product or products.
The objective is to strike an economic balance between high Btu recovery and
reduced ash or sulfur,levels. Five levels of CPC efforts have been defined,
each of which (when used with the appropriate raw coal) can produce marketable
products, i.e., at least one product which would release not more than 1.2
pounds of sulfur per million Btu when combusted.
The add-on costs to run-of-mine coal have been calculated based on the
processing of five different coals, one for each level of effort. These costs
include: capital cost, operating costs, and the cost of refuse disposal, as
well as the cost of environmental controls. An equation has been developed
which allows calculation of the total cost of upgraded coal at the CPC plant
boundaries, including the raw coal cost.
256
-------
INTRODUCTION
This paper discusses coal preparation and cleaning (CPC)
and is limited to a technical and economic overview. The paper
resulted from work on an ongoing Argonne National Laboratory
(ANL) program, "Environmental Control Implications of Generat-
ing Electric Power from Coal." The program is directed at
evaluating those environmental control technologies applicable
to coal utilization systems. As part of the ANL effort, Bechtel
National, Inc., performed a "Coal Preparation and Cleaning
"k
Assessment Study" for which a final report has been prepared
detailing the applicability, techniques, and economics of coal
cleaning. The program is sponsored by the Assistant Secretary
for Environment, DOE, Division of Environmental Control
Technology.
Coal is found in seams embedded between sedimentary geo-
logical formations and is contaminated with varying degree by
mineral matter. This contamination usually increases during
mining due to the recovery of top and bottom material. After
mining, the ROM coal consists of bulky material that is dif-
ficult to handle and to transport. CPC reduces the top size
and upgrades the quality of ROM coal to produce a product that
X
Bechtel National, Inc., Environmental Control Implications of
Generating Electric Power from Coalf 1977 Technology Status Rsport,
Argonne National Laboratory Report ANL/ECT-3, Appendix A,
Parts 1 and 2 (Dec. 1977).
257
-------
is cleaner and more convenient to handle. The top-size reduc-
tion requires relatively simple technology. The top size of
all the 600 x 10 TPY of ROM coal produced in the United States
is reduced. About one-third of this production undergoes addi-
tional cleaning including 120 x 10 TYP steam coal. The
primary objective of steam coal cleaning has been to remove ash
in the form of mine dilutions to save transportation costs.
Recently, however, another coal cleaning objective is to remove
sulfur to help utilities meet sulfur-oxide emission regulations.
Sulfur appears in coal mainly in two forms — pyritic
sulfur often of fine particle size and organic sulfur, which
is part of the coal matrix. Processes are under development
with the objective to remove all sulfur. However, this paper
will address only commercially available, physical CPC tech-
nologies that are limited to separate pyrite from coal or at
least that portion of pyrite that is liberated or can be
economically liberated by crushing.
PHYSICAL CPC TECHNOLOGIES
Coal preparation and cleaning involves a number of unit
operations that can be generally classified by the following
categories: comminution, classification, separation, and de-
watering and drying. The particular combination of unit opera-
tions depends on the raw coal characteristics and the quality
criteria of the marketable products (particle size, Btu^content,
258
-------
ash content, sulfur content, etc.). Some coals such as Eastern
bituminous coals may be subjected to all of these unit opera-
tions before they are marketed, while low sulfur Western sub-
bituminous coals only require crushing.
Figure 1 is a simplified schematic of the major unit
operations in CPC plants. Rotary breakers or roll crushers
are used for comminution of the ROM coal and screens for size
classification before the coal enters equipment to separate
the impurities. After the separation, screens, centrifuges,
and filters are used to recover the coal and refuse from the
separating medium, which sometimes includes thermal drying for
the coal product. The removal of mine dilutions from steam
coal has required mainly low cost coal cleaning with minimal
comminution. However, as the demand for the removal of pyrite
increases, the trend is toward more crushing and increased
processing of fine material.
Comminution
The crushing of coal can have two objectives:
• Top size reduction
• Liberation of mineral matter such as pyrite
Various types of crushers have been developed to pursue
these objectives. Figure 2 shows applications of the various
types of crushing equipment in use today with rotary breakers
and roll crushers for top size reduction with minimum fines
259
-------
ROM
COAI.
ro
o\
o
ROTARY
BREAKER
ROLLCRUSHER
COARSE FRACTION
SCREENS
DESLIMING
SCREENS
CYCLONES
JIG OR HEAVY
MEDIUM VESSEL
REFUSE
MEDIUM
FRACTION
FINE COAL JIG.
HEAVY MEDIUM
CYCLONES,
TABLES
REFUSE
FINE FRACTION
FROTH
FLOTATION
HYDFTOCYCLONES
REFUSE
SCREENS
SCREENS. SIEVE
BENDS.
CENTRIFUGES.
FILTERS
COMMINUTION CLASSIFICATION
SEPARATION
SCREENS
CENTRIFUGES
FILTERS
CENTRIFUGES
CLEAN
PRODUCT.
REFUSE
DEWATERING AND DRYING
Figure 1. A simplified schematic of the major unit operations in a CPC plant.
-------
a
ID
o
a
o
oc
a.
v>
HI
z
u.
UL
O
ai
LU
cc
o
z
ROTARY BREAKER
PRIMARY SINGLE
ROLL CRUSHER
PRIMARY DOUBLE
ROLL CRUSHER
SECONDARY SINGLE
I ROLL CRUSHER I
SECONDARY DOUBLE
ROLL CRUSHER
IMPACT CRUSHER
CAGE MILL
J I
0 1/4" 1/2"
2"
4"
3"
TOP SIZE OF PRODUCT
Figure 2. Range of application for crushing equipment in coal preparation and cleaning.
261
-------
production and impact crushers and cage mills for mineral
matter liberation. Figure 3 shows the relation of the particle
size distribution when different kinds of crushers are used for
medium hard coal.
Classification
Classification, performed predominantly with screens, is
the process of separating particles of different sizes. It is
accomplished on a screening surface with apertures of a given
size and/or shape. The screening surface may be either vibrat-
ing or stationary, the latter being used for scalping or wet
fine sizing. Vibrating screens include circular motion screens
with a sloped surface, to assist transport of the material, and
straight motion screens. Coarse screening requires a high
amplitude and low frequency, whereas fine screening is performed
with a low amplitude and high frequericy.
The screening can be performed dry or wet. The ranges of
wet and dry application for vibrating screens is shown in
Figure 4. With the increasing moisture of raw coal, dry screen-
ing is limited today to sizes above 1/4 inch (6 mm). Wet screen-
ing dominates, which has increased the use of stationary steeply
sloped screens, such as sieve bends, for fine sizing and dewater-
ing. The finer the screen opening, the lower is the capacity
for a given surface area; therefore, where a sharp classifica-
tion of fine material is not required, classifying cyclones that
262
-------
PARTICLE SIZE.d »
Figure 3. The relative production of fines using different
types of crushers for medium-hard coal.
263
-------
SCREEN
TYPE
ro
o\
•n
«5~
•*
to
3)
tu
3
f-+
10
v>
O
O)
rt>
n
O
o>
3
<->
o'
3
oo
ro
oo
j£ p
m 01
CO
X
> ^i
m
33
CO
-» o>
N
m o
^1 en
: 3
3
LOW AMPLITUDE
HIGH FREQUENCY
STRAIGHT MOTION
HORIZONTAL
SCREEN
O
i
m
3>
Z
O
LOW AMPLITUDE
HIGH FREQUENCY
CIRCULAR MOTION
INCLINED SCREEN
TO
33
m
O
m
^
j>
-J
m
33
0
33
rr.
rr
2
m
HIGH AMPLITUDE
LOW FREQUENCY
STRAIGHT MOTION
HORIZONTAL
SCREEN
O
m
^
^
H
m
33
O
CO
o
33
m
m
Z
Z
m
-H
33
33m
Z
C3
HIGH AMPLITUDE
LOW FREQUENCY
CIRCULAR MOTION
INCLINED SCREEN
CO
m
I51
CO
o
33
m
Z
O
m
H
CO
O
33
m
m
Z
Z
o
o
33
-------
have a high capacity and relatively low space requirement ate
preferred.
Separation
Separation divides raw coal into clean coal and refuse, and
sometimes additionally middlings, by utilizing differences in
physical properties between coal and mineral matter, Specific
gravity is the property most commonly used followed by surface
wettability for the separation of fine coal. The specific
gravity of separation is determined by a sink and float analysis
of a given coal. This analysis describes the amount and qua-
lity of material between the extremely low and extremely high
specific gravity fractions. The separation is considered easy
if the amount of material is low in the fraction within +.10
q
g/cm of the specific gravity of separation. The separation
becomes more difficult as the material to be separated becomes
finer. The specific gravity separation technologies commer-
cially available use air, water, or a heavy medium consisting
of fine ground magnetite in water. Froth flotation is the only
nonspecific gravity separation method for fine coal applied
commercially today.
Table 1 lists the application statistics in the United
States in the year 1971 and 1973 for various separation methods,
indicating that the major portion of the coal is clean in
jigs or heavy medium vessels. The trend towards more fine coal
265
-------
to
a*
METHOD OF
SEPARATION
JIGS
TABLES
LAUNDERS
HEAVY MEDIUM
PNEUMATIC
FLOTATION
PERCENTAGE OF COAL PROCESSED
1971
43
13
2
33
5
3
1973
48
12
3
32
0
5
Table 1. —Distribution of the coal cleaned in the
United States by method of separation used.
-------
cleaning with froth flotation is already visible and it is
interesting to note that dry separation methods for coal are
no longer competitive. The ranges of application of separation
equipment for wet processes is shown in Figure 5. The ordinate
is divided into three size fractions representing common size
splits in commercial coal cleaning, and the abscissa shows the
difficulty of cleaning expressed by the percentage of "near
o
gravity" material iii the +.10 g/cm fraction. Figure 5 is
supplemented by lists of characteristics for various separation
equipment for coarse, medium, and fine coal (Tables 2 and 3)
as well as performance curves (Figures 6 and 7)• Table 2 shows
that jigs and heavy-medium vessels are the most commonly used
devices for cleaning of coarse coal. The advantages of heavy-
medium separators are their ability to accomplish sharp separa-
tions even for coals with a high amount of near gravity material
and to follow accurate control of the cut point of separation
by adjustment for the specific gravity of the medium. Jigs are
more sensitive than heavy-medium separators to changes in the
feed rate and coal-to-refuse ratio.
Of the remaining equipment listed in Table 2, only hydro-
cyclones may be considered of value for coarse coal cleaning.
Although their performance ranks lower than that of jigs and
much below the performance of heavy-medium separators, the use
of hydrocyclones as a primary scalping device in conjunction
with other equipment, such as heavy-medium vessels, can enhance
267
-------
t/J
01
O
<
cc
u.
o
uu
N
CO
I
u
E
E
o
in
to
S
to
E
in
o
£
in
o
6
HEAVY MEDIUM VESSEL
JIGS
HEAVY MEDIUM VESSEL
JIGS
TABLES
HYDROCYCLONES
FLOTATION
MULTI STAGE FLOTATION
TABLES
HYDROCYCLONES
PYRITE REMOVAL ONLY
10 20
PERCENT NEAR GRAVITY MATERIAL
30
Figure 5. Ranges of application of separation equipment in coal beneficiation.
268
-------
0\
VO
EQUIPMENT
CHARACTERISTICS
Frequency of usage
Range of application based
on percent of near-gravity
material in feed
Common range of specific
gravity of separation
Relative sharpness of
separation
Relative ojkiratiny cost
Space requirements
Custom linilt
Sctisitivily of performance
to chaiiijcs in:
Feed rate
Feed st/e distribution
Amount of refuse
Control of cutpoint of
separation
Relative maintenance
costs
General comments
JIGS
Common
Up to 10%
Above 1.50
Good
Low
High
Yes
Yes
Yel
Yes
Difficult
Low
Relatively high
capacity per
single unit
HEAVY MEDIUM
VESSELS
Common
Up to 25%
Up to 1.80
Very good
Very high
High
Yes
No
No
No
Easy
High
Require desltmcd feed
and medium recovery
system
TABLES
Rare
Up to 15%
1 60 to 1.80
Poor
Low
Very High
No
Yes
Yes
Yes
Difficult
Low
Not recommended
fcr use with
coarse coal
HYDRO-
CYCLONES
Rare
Up to 7%
1.40 to 1.80
Fair
High
Low
No
Yes
Yes
Yes
Difficult
Fair
Preferred for
primary
scalping
LAUNDERS
Rare
Up to 7%
1.60 to 1.80
Poor
Low
High
Yes
Yes
Yes
Yes
Difficult
Low
Becoming
obsolete
PNEUMATIC
SEPARATORS
Rare
Up to 5%
1.40 to 1.80
Poor
Hujli
High
Yes
Yes
Yes
Yes
Difficult
Low
Renewed developmental
attention is being
focused on dry
separation methods.
but pneumatic methods
are not now widely
used commercially
'Not including auxiliary equipment such as medium-treatment systems.
Table 2. —Characteristics of equipment used for the separation of coarse coal (+%").
-------
K>
~J
O
CHARACTERISTICS
Frequency of usage
Range of application based
on percent of near-gravity
material in feed
Common range of specific
gravity cif separation
Relative sharpness of
separation
Relative |iyrite removal
capability
Relative operating cost
Spncc rui|iiircincnts
Custom built
Sensitivity of peiformaitce
to ch;iinjus"in:
Feed rale
Feed si/c distribution
Amount of refuse
Contiol of ciitpoint
of separation
Relative maintenance costs
Genet at comments
JIGS
Rare
Up to 10%
Above 1.60
Fair
Good
Fair
High
Yes
High
Hiyli
High
Difficult
Low
Require fcldspan
bed
HEAVY
MEDIUM
SEPARATORS
Common
Up to 25%
Up to 1.80
Very good
Good
Very high
Low
No
Low
Low
Low
Easy
Hicjb
Require heavy medium
recovery system;
relatively high capacity
per single unit
TABLES
Common
Up to 10%
1.60 to 1.80
Fair
Good
Low
High
No
High
High
High
Difficult
Low
Allow isolation
of pyrita
HYDRO-
CYCLONES
Rare
Up to 5%
1.40 to 1.80
Poor
Good
High
Low
No
Fair
High
High
Difficult
Moderate
Require multiple
Stages
FROTH
FLOTATION
Common
(Not applicable)
(Not applicable)
Poor
Poor
Low
High
Yes
Low
Low
Low
(Nut applicable)
Low
Poor selectivity of
coal from pyrite;
require reagents
'Not including auxiliary equipment such as medium-treatment systems
Table 3. —Characteristics cf the equipment used for
the separation of fine coal i nd ultrafine coul {—%"}.
-------
1.4
1.5 1.6
SPECIFIC GRAVITY-
1.7
1.8
Figure 6. Performance curve of a heavy medium vessel and jig processing +'
-------
100
1.3
1.4
1.5
1.6 1.7
SPECIFIC GRAVITY
1.9
2.0
Figure 7. Performance curve of a heavy medium vessel, fine coal jig,
table and hydrocyctone treating V*" (6mm) by 28 mesh (,5mm) fine coal.
272
-------
the overall economy of the separation system. Figure 6 shows
performance curves indicating the difference in performance
between a heavy-medium vessel and a jig for coarse coal.
Table 3 lists the characteristics of equipment used for
the separation of medium-sized coal and fine coal. Heavy-medium
separators, concentrating tables, and froth flotation units are
the equipment most commonly in use.
Heavy-medium separators are also the most efficient devices
for medium-size coal cleaning, and their application for fine
coal cleaning is under investigation. Tables clean medium-size
coal at relatively low cost; however, this cleaning has inferior
performance characteristics.
Baum-type jigs for medium-size coal have been replaced by
higher capacity automatically controlled Batac jigs.
Froth flotation when used under controlled conditions for
fine coal amiable to flotation does show good selectivity and
good economics for ash rejection. However, since pyrites show
similar surface characteristics as coal, froth flotation is not
very effective for the removal of pyrites unless practiced in
more expensive multiple stages. Reasonable pyrite removal from
medium-size and fine coal is possible with hydrocyclones, espe-
cially when used in multiple stages or supplemented by other
coal cleaning equipment. The space requirements are low, but
water pumping adds to the operating cost.
273
-------
Figure 7 shows the performance curves of the discussed
separation equipment for medium-size coal. The characteristics
and ranges reported in Tables 2 and 3 suggest a multitude of
choices of equipment for the cleaning of coal, which is an indi-
cation of the complexity, and sometimes the difficulty, of find-
ing an optimum equipment selection for a given coal.
Dewatering and Drying
With the increasing amount of finer materials, dewatering
and drying of the separation products have become a major effort
in CPC as indicated by the variety of equipment that has been
developed, the latest being pressure filters and centrifuges
with high-g forces. Mechanical dewatering is preferred over
thermal drying, which is only used where the mechanically de-
watered product cannot meet a given moisture specification.
Figure 8 shows the various kinds of dewatering equipment, both
mechanical and thermal, used with respect to the coal size and
the end product moisture desired. The characteristics of this
equipment are also reported in Table 4.
Dewatering equipment for the coarse 1/4-inch (6 mm) coal
are vibrating screens and sometimes basket-type centrifuges.
For 1/4-inch x 28 mesh (6 mm x 0.5 mm) coal, vibrating screens
and vibrating basket-type centrifuges are used. For below 28
mesh (0.5 mm) coal, dewatering is accomplsihed by a combination
of static thickeners or cyclones with filters or bowl-type
centrifuges.
274
-------
ou-
O
Q
OC
OL
0
on _
SLZ
% MOISTURE OF DEWATEF
-» E
no e
2.5
STATIC THICKNER*
VACUUM FILTER-
SOLID BOWL CENTRIFUGES*
VACUUM FILTER j f™fi.
PRESSURE FILTER JLO™A° p|
HY
H
ENCY
LITUDE
\ [SCREENS^
SCREEN BOWL CENTRIFUGES
CRAULIC CLASSIFIER
SIEVE BEND
LOW FREQUENCY
HIGH AMPLITUDE
SCREENS
>v VIBRATING BASKET CENTRIFUGES
THERMAL DRYING
1 1 1 1 1 1 1
t ,
LIQUID HANDLING
SOLID HANDLING
i '
•TAILINGS WITH USE
OF FLOCCULANT
LOW FREQUENCY |
HIGH AMPLITUDE SCREENS i
i
BASKET CENTRIFUGES J
1 1
325m 200m 100m 48m 28m 14m
8m
1/4"
1/2"
1"
PARTICLE SIZE
Figure 8. Common ranges of application for coal dewatering equipment
-------
EUrilPMlNf
CH AH ACT £11 IS HCS
Fioiueucy ol uw>jc
Cl*an coal dewalermg
licluie dnwaiciiiMj
Common |»at lt«J«: u/tf
Otltalliallle llnMUMTt ItUMtlUre
content (wei.jlll %)
Omtnmom ufi«rj|MMi
Scuwitlwlly Ilk v«ryni<| l*:nl
«*—••«»«•«
CtnliHii nVsiijnril
S|>jce ict|iMieiiHiiu
Hocculaitit. m«J (in dc-
w-,len,H, lailuiut
Melanin* inanttauauce cutlv
Ge,«Jt.mm«nls
TIIICKLNbllS
SIAIIC
ClHlHIIIIII
Har.
Cutiuiioii
2!) llMUl * 0
/-,
Yd
low
Y«
Vciy luylt
Yo
Linw
Low
C.i|iJilc ul tiuriiMj
dutiy; al^o usvtl
lu< WUltrl dailll-
calfm
CYCLONES
Common
Common
Common
K mcli x 0
70
Y,s
1 ow
Nu
l«*
Nu
Low
Low
(hi not produce
a cleat ^wef lluw;
low uiluls
leciHery
CENTRIFUGES
BASKET
Low
Common
Rant
I'/rx Kindt
3to4
Yes
Low
No
Low
No
Low
Low
Relatively (null
batket wear con-
siderirM| low de-
wee ol inotstuf e
reduction
VIBRATING
BASKET
Common
Commoit
Rare
K inch x 28 r.wsli
7 to 9
Yes
Low
No
Low
No
Low
Low
Product moisture
content ikuei^ds
on llw! amoun' of
slimes in llw 1. ed
SOLID BOWL
Low
Hare
Moderate
28 mesh x 0
3Sto40
Yes
Hiyh
No
Low
Yes
ItHjIl
Very high
Can iModuce clear
cliluenl under con-
uoUed conditions
FILTEHS
SCREEN BOWL
Low
Moderate
Rare
14 inesli x 0
lit to 20
Yes
Moderate
No
Low
No
HMjh
Iligli
Can be used lor
dcslimuMi Irotli
conceuliale
VACUUM
Common
Common
Rare
28 mesh x 0
20 to 30
Yes
Moderate
No
Moderate
Yes
Low
Low
Sensitive to
bunding
PRtSSUKE
Rare
Rare
Rare
28 mesh x 0
Under 20
Nu
Low
Yes
lligTi
No
High
Low
Hei|inie ex-
leiiune nuniiat
labor
IIILKMAt
UllYtHi (CUt
DOMINANTLY
FLUIUUtU BtO
Hare
CiHiimuii
(Niil a|>|4icat*lt:l
IX Kiel, x 0
Under S
Y«
Low
Y«
Idyll
•Nut a|Boli£dii*itffl
Very hiijh
Iliyti
l
-------
CPC PLANTS
The unit operations can be combined in various manners in
coal preparation plants to strike a balance between quality
improvement and product recovery. Since raw coal exists in a
variety of qualities, CPC plants range from simple one-stage
crushing to complex combinations of all unit operations. For
simplication, the combinations have- been divided into five
levels of CPC effort with the lowest as Level 1 and the most
complex as Level 5. The definitions of these five levels of
CPC are indicated in Table 5 and presented below:
• Level 1 involves no cleaning, but merely the
preparation of ROM coal to size specifications.
• Levels 2 and 3 use low efficiency separation
devices to process easy-to-clean coal to reject
mine dilutions and to free pyrites if present.
Level 2 plants clean only the coarse size frac-
tion, whereas Level 3 plants include the clean-
ing of the medium-size raw coal.
• Levels 4 and 5 make use of high efficiency sepa-
ration methods to clean all size fractions.
Level 5 combines the most sophisticated unit
operations to produce a clean coal and a middlings
product.
277
-------
LEVEL
DEGREE OF CLEANING
to
^.J
00
NO CLEANING
MINIMUM CLEANING
MODERATE CLEANING
EXTENSIVE CLEANING
ULTIMATE CLEANING
COMMENT
SIZING OF EXCELLENT QUALITY RAW
COAL
SEPARATION OF MINE DILUTIONS FROM
COARSE SIZE FRACTIONS OF GOOD
QUALITY RAW COAL
SEPARATION OF MINE DILUTIONS FROM
COARSE AND MEDIUM SIZE FRACTIONS
OF MEDIUM QUALITY RAW COAL
CLEANING OF ALL SIZE FRACTIONS OF
INFERIOR QUALITY RAW COAL
CLEANING OF ALL SIZE FRACTIONS AND
PRODUCTION OF MIDDLINGS FROM LOW
QUALITY RAW COAL
1 able 5. -The five levels of CPC.
-------
Figures 9 through 13 show block diagrams of CPC plants
corresponding to the five levels of CPC effort. Figure 9, a
Level 1 effort, shows two process diagrams; one, which can be
considered for typical Western subbituminous coal using roll
crushers, and the other for a typical Eastern coal using a
rotary breaker to reduce the raw coal top size, the latter with
a secondary benefit of mine-rock removal. Figure 10 represents
a Level 2 effort using a jig to clean the coarse coal after
removing medium size and fine coal from the raw coal by dry
screening. This operation can rarely fulfill today's coal
cleaning requirements. More frequently in use are plants fol-
lowing the flow diagram in Figure 11 (Level 3) where the coarse
and medium size coal is cleaned (for example) in jigs, and the
fine coal is recovered uncleaned. Figure 12 identifies cleaning
systems for all size fractions — a flow diagram that is typical
for most CPC plants producing metallurgical grade coal, but that
is becoming more and more acceptable for the production of steam
coal as well (Level 4).
Finally, a Level 5 CPC effort.as diagrammed in Figure 13
uses multiple stages of cleaning to produce a premium clean coal
and middlings product. This level of effort is only feasible
if the market for middlings can be developed.
Despite the simplified division of CPC into five levels,
there are no defined standards for selection of a CPC process
279
-------
WESTERN COAL
ROM
SCREENING
T
r
PRIMARY
CRUSHING
1
' 6" X 0
SCREENING
^ SECONDARY
CRUSHING
2"XO
DRYING
(OPTIONAL)
STORAGE
I
UNIT TRAIN
LOAD OUT
T
SHIPPING
1
BLENDING
(BY USER)
MEDIUM SULPHUR COAL
ROM
CRUSHING
SCREENING
3"XO
STORAGE
UNIT TRAIN
LOAD OUT
MINE ROCK
Figure 9. Block diagram for Level 1 CPC.
280
-------
10
Qo
MINE ROCK
ROM COAL
I
CRUSHING AND
SCREENING
4"{6")X 0
RAW COAL
PRODUCT
REFUSE
WATER
STORAGE
1
DRY SCREENING
DEWATER1NG
3/8" XO
4"(6") X 3/8"
CLEANING
->r*
U-
THICKENING
DEWATERING
REFUSE
DISPOSAL
OEWATERING
CRUSHING
CLEAN COAL
STORAGE
1 1/2" XO
UNIT TRAIN
LOAD OUT
Figure 10. Block diagram for Level 2 CPC.
-------
ro
CO
K»
MINE HOCK
ROM COAL
__t
CRUSHING AND
SCREENING
I
6"(4")XO
RAW COAL
PRODUCT
STORAGE
WATER
3/8" X 0
CLEANING
I ...*_
«-^M DEWATEHING
t
DEWATERING
28 MESH XO
___
H
THICKENING
S\
:NING
• — -— I at I
1
DESLIMING
CRUSHING
28 MESH XO
CLEANING
DEWATERING
DEWATERING
3/8" X 2!
MESH
\
1
DEWATEKING
t t
28 MESH X 0 f*^
t_
I
DISPOSAL
I
THERMAL
DRYING
CLEAN COAL
STORAGE
1 1/2" X 0
UNIT TRAIN
LOAD OUT
Figure 11. Block diagram for Level 3 CPC.
-------
ROM COAL
MINE ROCK
1 ^ +3/8"
1
1
i
1 4 CLEANING
M' "
DEWATERING I
1
y^
| r^-| CLEANING
1 Y
•<— ^p- j DEWATERING j 1
U A
1 _
! r
L.J. __^h TIIICKI
\
1 U
1^ ' 1 ' DEWAT
1 !
REFUSE DISPOSAL j L J
CRUSHING
— — __ . _ PAW POAI
AND ilCHhhNING KAVW t-UAU
± PRODUCT
5"(4"»XO
STORAGE WATER
>|
^^ Wtl aUHbtlMINti
ijr 3/8" XO fc
t t
1
>- . ^ npti iMiMd
T I
/ \ /" \ "*3/8 /• \ ^.^
JbWATERING I " " *3r
T ^| CRUSHING
PM , ^ _
3/8" X 28 MESH
D~*1
* 3/8" X 28
/• \ MESH ^^
)E WATER ING ~" ' ' ^
JT 28 MESH X 0 V
f
rAMiMG fc>
^ •_ - - • _/T\_ 28 MESH X 0
ENING DEWATERING ftJ
4
cHINu 1 lltltlVIAL Ull I IIMu ^^^ I'l
CLEAN
. STOR
0
X
CM
r-
f
COAL
AGE
i
UNIT TRAIN
LOAD OUT
-------
ROM COAL
4 MINE ROCK
CRUSHING AND
SCREENING
**
J, .2" X
STORAGE
T/n" 1 oolll-L-
3/8" XO.
STORAGE
' RAW COAL
PRODUCT
1 MIDDLING
REFUSE
WATER
i
-i—|DEWATERJNG|<—["'CLEANING \ »{DEWATERING|
200 MESH X 0
REFUSE
DISPOSAL
isi I
SAL j
-•" • 28 MESH X 200 MESH ,
* - *¥
"J-u-, 1
U-j CLEANING j ^JDEWATERING}-^-
t J
1 *
L " I THICKENING ^~ ~
^ * *>
r THERMAL MIDDLING 3/8" X 0
s
4 „ ,. f
\_J
\
\
f^
f
i
1 1/2"
N 200M
r
1 "- — [DEWATbHING
XO
CLEAN COAL
r
TJ
STORAGE
UNIT TRAIN
LOAD OUT
STORAGE
THERMAL
DRYING
UNIT TRAIN
LOAD OUT
Figure 13. Block diagram for Levti 5 CPC.
-------
and no standard solution for producing clean coal. Very few
CPC plants in the United States are identical even though their
block diagrams showing general unit operations may suggest
identical treatment of the coals. The equipment selection to
perform these unit operations can vary depending on a variety
of factors as discussed previously. By substitution and/or
addition of equipment, a plant may be converted from one level
of CPC to another for reasons such as increased coal desulfur-
ization, improved Btu recovery, or changes in feed
characteristics.
For optimal coal cleaning, the raw coal characteristics
must be known. Consequently, each coal deposit should be tho-
roughly explored and characterized, and a mining plan developed
that is compatible with effective CPC plant operation. The
implementation of such a mining plan can, for example, achieve
a more uniform feed to the CPC plant by blending the ROM coals of
known characteristics from different mining sections.
ECONOMIC OVERVIEW
A process flow diagram was prepared for each of the five
levels of CPC based on specific coals. The coals were selected
based on their reserves and on their sulfur reduction potential
to meet a product specification that would allow the burning
of the coal without releasing more than 1.2 Ib S02/10 Btu.
The flow diagrams and the equipment selection were prepared on
285
-------
the basis of 1,000 TPH or 3.25 x 10 TPY of marketable products.
Two coals — one with high sulfur and one with low sulfur coal
so that the two coals can be blended to meet product specifica-
tions - were in the Level 1 effort. The coal selected in the
order of levels of CPC are: Rosebud coal and Illinois No. 6
for Level 1; Cedar Grove coal for Level 2; Montrose coal for
Level 3; lower Kittaning coal for Level 4; and upper Freeport
coal for Level 5,
The flow diagrams served as the basis to estimate capital
and operating costs. Table 6 shows the summary of. the capital
cost data in 1977 dollars, and this table is supplemented by a
detailed cost breakdown of the direct field costs in Table 7.
It is interesting to note that a substantial part of the direct
field costs is attributable to environmental equipment and mate-
rials as shown in Table 7a. For the different levels of CPC,
the environmental-related direct capital field costs range from
15.9 to 7.6 percent. The operating requirements and operating
costs for the corresponding levels of CPC are shown in Table 8,
with a summary of the processing costs given in Table 9.
The processing costs are only part of the cost entering into
the calculation of final fuel cost. The cost of the refuse,
which is a function of the product recovery, must be recognized.
Figure 14 expresses graphically the composition of the final
fuel cost at the CPC plant boundary.
286
-------
to
00
Coal cleaning facility
Ancillary facilities
Environmental equipment and materials
Direct field cost
Indirect field cost
Total field cost
Engineering services
Construction cost
Allowance for uncertainty
Tot -I Construction Cost
Client engineering6
Environmental impact statement"
Working capital
Construction uiterest
Landb
Total Capital Cost
CAPITAL COSTS (IN THOUSANDS OF DOLLARS)
CASE 1.1"
1.460
2.610
770
4.840
200
5.040
600
S.640
1.160
6.800
204
102
122
1,479
150
8.8S7
CASE 1.2
1.110
2.510
540
4.160
160
4.320
520
4.840
960
5.800
174
87
100
1.253
150
7.5G4
CASE 2
3.640
4.150
890
8.680
290
8.970
1.080
10.050
2.050
12.100
363
182
146
1.960
225
14.976
CASE 3
8.490
4.150
1.270
13.910
1.170
15.080
1.810
16.890
3.410
20.300
609
304
187
4.385
330
26.115
CASE 4
10.260
4.260
2.170
16.690
1.250
17.940
2.150
20.090
4.010
24.100
723
362
272
5.205
900
31.562
CASES
16.590
7.290
1.980
25.860
2.250
28.110
3.370
31.480
6.320
37.800
1.134
567
385
8.16S
600
48.651
aAn additional construction cost of $6.9 million for an optional thermal dryer is not included in these figures for Case 1.1.
^A land cost of $3.000 per acre was used in calculating Land Cost.
C3% of Construction Cost.
dlJ5% of Construction Cost.
Table 6. —Conceptual estimate capital cost summary.
-------
00
00
COST ITEM
Coal Cleaning Facility
Site preparation
Concrete
Steel and building services
Mechanical
Conveyors
Piping and Instrumentation
Electrical
Ancilltary Facilities
Yard facility
Transportation and mobile equipment
Storaye and loadout
Service buildings
Enviromental Equipment and Materials
Oust collection
Noise abatement
Refuse disposal area
Total Direct Field Cost
CAPITAL COSTS (IN THOUSANDS OF DOLLARS)
CASE 1.1»
1.460
100
40
20
390
660
90
160
2.610
200
730
1,640
40
770
520
30
220
4.840
CASE 1J2
1.110
100
40
20
440
330
70
110
2.510
200
730
1.540
40
540
290
30
220
4,160
CASE 2
3.640
120
170
320
1.730
720
220
360
4.150
260
730
3.120
40
890
550
70
270
8,680
CASE 3
8.490
120
170
320
1.220
5.190
430
670
4.150
260
730
3,120
40
1.270
445
175
650
13.910
CASE 4
10.260
220
350
2.380
5.210
810
500
790
4.260
370
730
3.120
40
2.170
590
270
1.310
16.690
CASES
16.590
220
400
3.130
9.200
1.420
870
1.350
7.290
370
730
6,110
80
1.980
740
270
970
25.860
•An additional direct field cost of 4.17 million dollars for an optional thermal dryer is not included in these figures for Case 1.1.
Table 7. —Details of conceptual cost estimates.
-------
K>
00
CASE
1.1a
1.2
2
3
4
S
ENVI RONMENTAL-RELATED
DIRECT CAPITAL
FIELD COSTS (SIOOO's)
770
540
890
1.270
2.17011
1.980
DIRECT FIELD COSTS RELATED
TO ENVIRONMENTAL
CONTROLS
15.9%
13.0%
10.3%
9.1%
13.0%
7.6%
i .
•^without optional thermal drying
reject disposal
Table 7a. —Fraction of direct field costs devoted to
environmental control equipment in the CPC plants.
-------
C-.1.I
rum Cuttiliucikm Cml SG.800.000
ImbUcd ttiuHMMi.1 Cut! 1.850.000
|*i octfmua dt^tf 0 584
lltfliHM HM iMvciiiiMMl. 30*w bclui c foxvfc 0.245
li>4.il i^/ltMitky |iiutlucO 0.829
ftukctuuij Cum 0.02G
HriiM« IM hwmHMnil. 3UX licluic !<«» 0 Ol 1
lulrfllSMMUiul O.O37
Ciwl.2
$5.800.000
1.270.000
0.518
0.209
0.727
0.019
O.OO8-
0.027
CM 2
$12.100.000
3.G50.000
0.91G
0408
1.324
0.03G
O01G
0002
Ci«:3
$20.300.000
7.UKI.OOO
1.39G
0.723
2.119
O.053
0028
0081
&IM4
$24.100.000
8.180.000
1834
0.874
2.708
O.OG4
0.031
0095
CucS
$37800.000
13.85O.OOO
2825
1347
4.172
O.O97
004G
O 141
K>
\O
O
(Hirvl Cu»ti
IdtMU
O|Kiuluu| 1 ^MM MH| Su|iifinua«. S87/iiu«-ilull
tXjMibrii.nu.i- |.^|MM *". b87/«Mii klwll
lulal U«»i jtmt &i|iEioiMun
lt|«:ialu»i Utility J>«l CiMluiutjIilet
lyU^iicliic. Sett/tun
1 bicculjiil. V/U»
llo..«. S002ISMM>
Wj.i. iU Ib M IH!
O|«rl*lH<^ Su|l|4ll
Mxinicfunu-. M^iciul. S«M4y. »I (>!1r:CI »l Imui juce. !.!>% ul I1*u Coil
lutJ Irahiri:! IJnu
Oiliitjl llcbml Cmit
IM.I I«KI«I I *|«»u! I ItK. *• "JTX.I
CIHHIHM.IMMI luteictl 1 «|>«««. ITS. ul liitemt UtuiiHj CUIUIIIICIHM
W>»lu»9 l.>mJ Ink-fol. !l£
HC|MCCIU|MHI. 20 yeMk (bCl*«fhl IMM)
lutj (^t«l^ llrl^lnl Covl»
Ul 1 At rilUCCS^INC COST
^ 1. 1_. . II,
240.!>5!>
.222
33C.777
-
-
!>2.080
-
33.078
12'J.!iOO
48.7M)
204.008
tiOO.780
1U8.388
33.G78
102.000
304.0GG
428.400
38.UM)
91.517
435.3SO
aa'j.yii7
1.898.618
240.!>!<5
.!>00
48.VM)
428.7G7
SOI.7U7
2IG.50O
43.300
18I.500
441.300
7U2.300
J.2.U2S
121.748
73/.!iWJ
1.U74.&23
2.'J//.tiUO
JAL COST 7.420
4.200
131.313
I.7&&
•H..710
532.0OO
48. /bO
703.818
1.220.918
228.ti!iO
4!>.710
304.UOO
b78.7UO
1.215.900
84.41U
14G.2(i8
1.289.278
2.73b.8C4
4.535.542
5O5.122
240.5!>5
745.G77
97.240
2*1,030
VJ4.BHO
7.459
74.5G8
572.UH1
0
1.024.527
1.770.204
372.839
74.5G8
3G1.500
8O8.907
1.518.300
105.413
223.461.
1.533.100
3.380.27C
S.959.387
553.320
3GO./89
914.109
343.72O
62.77U
340.704
•J.5HG
91.411
M.9.5OO
48.7U1
1.8W..JG/
2.780.4/0
457.054
91.411
5G7.4G5
1.1 15.930
2.38 1.400
192.893
308,310
2.402.525
S.285J28
9.181.534
N»lr. HANI Cvuoiy. IOUO IfH; 3.25 nuMiui IOMI per \en. O|wi«lilU| U»U: 2&0 days |ici y«M. 13 |MMU> |ie< iliff.
Table 8. —Analysts o*f annual coal processing.
-------
\O
CASE
COAL
Plant Construction Cost
Pre-coiistruciion and Owners Costs
Total Depreciable Capital Cost
Land Cost
Total Capital Costs
Annual Costs
Operating and Maintenance
Capital-Related12
Tout Annual Costs
Return on Investment1)
Coal Processing Costs*
S/ton dry product
$/10<> Utu tliy product
Return on Investment
$/toii dry product
S/JOG Utu dry product
Total Costs
$/ton dry product
$/106 Btu dry product
Lib
MONTANA
(ROSEBUD)
6.800.000
1.907.000
8.707.000
150.000
8.857.000
904.851
993.767
1.898.618
797.130
0.584
0.026
0.245
0.011
0.829
0.037
1.2
ILLINOIS
(NO. 6)
5.800.000
1.614.000
7.414.000
150,000
7.5G4.000
824.471
857.669
1.682.140
680.760
0.518
0.019
0.209
0.008
0.727
0.027
2
W. VIRGINIA
fCEDAR GROVE)
12,100.000
2,651.000
14.751.000
225.000
14.976.000
1.303.067
1.674,523
2.977.596
1.327.590
0.916
0.036
0.408
0.016
1.324
0.052
3
COLORADO
(MONTROSE CITY)
20.300.000
5.485.559
25.785.559
330.000
26,1 15.559
1.799.678
2.735.864
4.535.542
2.350.400
1.396
0.053
0.723
0.028
2.119
0.081
4
PENNSYLVANIA
(LOWER
KITTANNING)
24.100.000
6.562.000
30.662.000
900.000
31.562.000
2.579.111
3.380.276
5.959.387
2.840.580
1.834
0.064
0.874
0.031
2.708
0.095
5
PENNSYLVANIA
(UPPER
FREEPORT)
37,800.000
I0.21i0.500
48.0S0.500
600.000
48.650.500
3.896.406
5,285.128
9.181.534
4.378.545
2.825
0.097
1.347
0.046
4.172
0.143
a Cortespoitds to 3.25 million tons per year (dry) at 250 thirteen hour annual operating days.
b Excluding thermal drying option which would add $0.45 per ton dry product to the processing cost, and $6,900,000 to the capital investment.
c Calculated at a 7:3 debt/equity ratio repaying debt with 9% 20-year bonds.
d Calculated on equity with a before tax return of 30% with no discounting.
e All costs for Case 5 relied the combined middlings and clean coal products. The separation of these costs will depend on market conditions.
Table 9. —Operating cost summary.
-------
C
OF
EQUIVALENT
RAW COAL
COST, X0
COST, p
COST OF RAW COAL, X
(dollars per ton raw coal)
Figure 14. The cost components of the total
cost of product coal at the CPC plant boundary.
292
-------
As can be seen, the total cost is very sensitive to the
amount of refuse, especially when the raw coal cost is high.
This is clearly demonstrated in Figure 15, which shows that
even though the processing cost for Case 4 is lower than for
Case 5, the final fuel cost becomes higher as the cost of the
raw coal increases. The reason for this is that the coal chosen
for Level 4 shows a low weight recovery of 68.6 percent versus
the 76.1 percent weight recovery of marketable product for the
Level 5 operation. The weight recovery, Btu recovery, as well
as other performance data for the five levels of CPC are shown
in Table 10.
293
-------
30
COST OF RAW COAL (S/TON)
Figure 15. Clean coal effective cost on a weight basis
compared to raw coal costs for cases 4 and 5,
294
-------
NJ
VO
01
CPC LEVEL
COAL
Raw Coal
Heating Value
% Sulfur {Pyritic}
Total
Ash%
Aveiitije Moisture %
Ib SO2/10& Ulu
Product Coal
Mealing Value
% Sulfur (Pyritic)
Total
AJi%
Moisture %
ll>SO2/10<>Btu
Performance
% Weight Recovery
% Blu Recovery
% Sulfur Reduction
Refuse
Ash%
% Sulfur (Pyritic)
Total
Btu
CASE 1.1
MONTANA
11,709
0.23
0.70
8.8
24.1
1.20
11.709
0.23
0.70
8.8
24.1 (13.6)
1.20
100 (97.6)
100 (97.6)
_
_
—
—
—
CASE 1.2
ILLINOIS
13.314
0.92
2.69
75
4.1
4.04
13.314
0.92
2.69
7.5
4.1
4.04
100
100
—
— -
—
—
—
CASE 2
WEST VI 3GINIA
11.810
0.11
0.81
22.7
5.0
137
12.655
0.03
d.75
17.0
6.5
1.19
90.9
97.4
7.4.
79.6
0.67
1.40
3.370
CASE 3
COLORADO
11.790
0.25
0.80
19.4
6.0
1.36
13,120
0.19
0.73
105
4.0
1.11
84.7
94.3
8.3
70.6
0.64
1.18
4.140
CASE 4
PENNSYLVANIA
12.830
2.19
2.77
13.0
1.3
4.32
14.250
0.22
0.80
4.7
6.0
1.12
68£
76^
71.1
31^
652
7.1
9.837
CASES
PENNSYLVANIA
11.486
2.79
3.40
23.4
5.0
5^2
14.608
12.342
0.22
156
0.83
2.17
3.3
175
6.0
6.0
1.14
352
48.1
28.0
61^
30.1
75.6
36.2
72.1
958
10.19
3.996
Table 10. —Performance summary for tlie six CPC systems designed.
-------
SUMMARY
Coal preparation and cleaning (CPC) processes include
comminution, classification, separation, dewatering, and drying.
Run-of-mine (ROM) coal is processed by all or any combination
of these processes on its way to the consumer. Each process
is performed in unit operations that may use various types of
equipment and process principles. The unit operations are com-
bined in CPC plants, in sequences depending on the raw coal
characteristics and the extent of CPC effort required to produce
a marketable product or products. The objective is to strike
an economic balance between high-Btu recovery and reduced ash
or sulfur levels. Five levels of CPC efforts have been defined,
each of which (when used with the appropriate raw coal) can
produce marketable products, i.e., at least one product that
would release not more than 1.2 pounds of sulfur per million Btu
when combusted.
The add-on costs to ROM coal have been calculated on the
basis of the processing of different coals, at least one for
each level of effort. These costs include capital cost, the
cost of refuse disposal, and the cost of environmental controls.
An equation has been developed that allows calculation of the
total cost of upgraded coal at the CPC plant boundaries, includ-
ing the raw coal cost.
296
-------
This paper shews that a variety of physical CPC technolo-
gies are available to upgrade ROM coal. Since the objectives
and degree of CFC vary from case to case and with changing
market requirements, standardization of CPC is very difficult.
The emphasis on pyrite removal, which necessitates the treat-
ment of finer material, has made CPC application more complex.
The economic evaluations show the sensitivity of the final
product cost to the product weight recovery. This important
factor is often neglected in the search for low-cost methods
to improve coal quality. With rising coal costs, product
recovery must be a dominant factor in any CPC effort selection.
Likewise, emphasis should be made to operate the CPC plants
with maximum efficiency.
ACKNOWLEDGMENTS
This paper is based on work performed by the staff of the
Research and Engineering Division of Bechtel National, Inc.
The sponsorship of the work by the Division of Environmental
Control Technology, Assistant Secretary for Environment, United
States Department of Energy, and the cooperation of Argonne
National Laboratory.(ANL) are gratefully acknowledged.
297
-------
OVERCOMING THE BARRIERS TO INVESTMENT IN PHYSICAL COAL
CLEANING FOR S02 EMISSIONS CONTROL
Karel Fisher and Peter Cukor
Teknekron, Inc.
2118 Milvia Street
Berkeley, California 94704
ABSTRACT
Teknekron, like others, has found that physical coal cleaning in many
cases could be the least-cost strategy for achieving dramatic reductions in
sulfur dioxide emissions. But right now physical coal cleaning for S02
emission control is little more than an intriguing idea. It has yet to be
implemented in any substantial way.
So how do we get from here to there? How do we encourage the use of
physical coal cleaning (PCC) in those situations where its economic and
environmental payoff can be predicted with reasonable confidence?
The key is to take a real-world approach, by looking at coal cleaning
from the point of view of the potential investor. In other words, getting
from here to there demands more than theoretical studies of the engineering
and environmental benefits of PCC. It means recognizing that the people
providing the risk capital for PCC face a number of formidable investment
barriers.
These investment barriers—and the strategies for overcoming them—
are the subject of this Teknekron paper. We begin by summarizing the barriers
Teknekron has identified. Then we turn briefly to a few of our engineering
calculations showing that PCC in certain circumstances can be the least-cost
S02 and particulate compliance option. Finally, we proceed to a fuller
discussion of the barrier analysis and explore the strategies that Teknekron
believes would lead to wider investment in PCC.
298
-------
The Barriers; A Summary
After interviewing bankers and a variety of potential coal^cleaning investors-**
large and small coal companies, electric utilities, and other current and
would-be industrial coal users-^and after considering a wide range of possible
investment barriers (institutional, economic, social, regulatory, legislative,
financial, legal, and contractual], Teknekron Has concluded that eight factors
in particular are currently inhibiting the widespread adoption of PCC, Some
of these barriers constrain only certain types of investors. Others -appear
relevant to all of them.
The first three barriers have to do with information or data deficiency:
• First, in general, none of the potential investors are aware of the
wide range of benefits and costs that might accrue from different ap-
plications of PCC.
• Second, electric utilities and the various public utility commissions
generally know relatively little about developments in PCC technology
and about the possible applications of these developments.
• Third, some potential investors with whom Teknekron has spoken say that
the engineering and economic studies performed to date are inadequate.
In other words, these studies are not "investment grade" analyses. They
lack data from full-scale PCC demonstrations, and they fail to examine
the sensitivity of PCC economics to such things as the price of coal,
the fraction of pyritic sulfur in the coal, transport costs, boiler or
contract lives, different SOg standards, and type of investor.
The next two barriers are essentially institutional in nature:
299
-------
• Whereas relatively small coal companies may control significant quan-
tities of cleanable reserves, these companies may be unable to bring
their coals into the marketplace without help in establishing insti-
tutional arrangements for the sharing of PCC plants.
• Furthermore, potential nonutility coal consumers see the environmental,
logistical, and Institutional costs of burning coal as prohibitive un-
der most circumstances.
The next barrier involves two regulatory constraints:
• Despite provisions in the 1977 amendments to the Clean Air Act that
permit credit for precombustion sulfur removal, coal cleaning (unlike
flue gas desuIfurizatton) is~ncrt considered a pollution control invest-
ment by the Internal Revenue Service. Nor does it qualify for invest-
ment tax credits. What this means Is that PCC does not enjoy certain
tax and financing advantages afforded FGD Investment.
The last two barriers involve uncertainties about government action:
• First, potential investors are unsure about the ultimate costs of the
air and water quality regulations that will be applied to PCC plants
themselves.
• Second, potential investors are inhibited by enormous uncertainty sur-
rounding several issues related to EPA activity—the enforcement and
enforceablHty of State Implementation Plans, the feasibility of con-
tinued noncompliance, decisions regarding quadrennial review of stand-
ards, definitions of Best Available Control Technology for PSD appli-
cations, decisions on offset policies, the rate at which state standards
300
-------
may become more stringent, and the ultimate stringency of S0« stand-
ards for coal-fired boilers.
Teknekron examined many other possible barriers to coal cleaning investment,
including barriers that have been suggested by others from time to time. For
example, we considered the concern that existing coal supply contracts lack
price adjustment clauses, and we considered the claim that certain types of
investors will reject coal cleaning out of hand no matter what the state of
th.e PCC investment environment may be. No evidence emerged, however, to sug-
gest that any of these other factors were significant.
Examples of PCC as the Least-Cost Option
Before discussing the investment barriers in more detail, It seems appropriate
to look at a few cases where Teknekron's engineering calculations show PCC to
be the least-cost strategy for complying with S0« and partlculate emission
limitations.
As a first example, consider a hypothetical 300 MW coal-fired power boiler
located in the vicinity of Harrisburg, Pennsylvania, where th.ere is a rela-
tively strict state S02 limit of 2.0 pounds per million Btu and a particulate
limit of 0.2 pound per million Btu. If we assume current f.o.b. mine prices,
current average rail tariffs, current cost data for new scrubbers, and a coal-
cleaning charge of $2.00 per raw ton with 88 percent Btu recovery, we get the
results shown 1n table 1.
The options shown 1n the table are ranked in order of increasing total cost of
operating the boiler in compliance with emission constraints. By total cost
we mean the sum of the delivered cost of the coal plus the annual1 zed pollu-
tion control Investment and operating costs. It's this "as burned," or total
301
-------
Table 1
Costs of Alternative Strategies for Meeting Emission Limits of
2.0 lb/106 Btu S02, 0.2 lb/106 Btu Particulates
(Southeast Pennsylvania Location)
Coal Source, Delivered Coal Cost. Cost of ESP Cost of FGD Total Cost
Type U/106 Btu) U/106 Btu) U/106 Btu) U/106 Btu)
Pennsylvania, 129 20.9 - 150
cleaned
West Virginia, 132 23.6 - 156
uncleaned
Pennsylvania, 108 18.1 45 171
uncleaned
302
-------
effective cost, that 1s the relevant measure to use when comparing different
compliance strategies from tPie utility's point of view.
In this case we see that, given our assumptions, the cleaned-coal option is
slightly cheaper than the option of buying higher quality, more distant, low-
sulfur West Virginia coal and considerably cheaper than the option of par-
tially scrubbing the more local coal.
As a second example, consider a new 5QQ MW coal-fired unit located 1n the
central part of Ohio and subject to the new source standards of 1,2 pounds
of SOg and 0.1 pound of particulates per million Btu of boiler heat input.
Using the same assumptions as in the previous example, we compare three dif-
ferent compliance strategies, as shown in table 2. Here, Importing high
quality, cleaned Central Appalachian coal is cheaper than either importing
low-sulfur, complying western coal or scrubbing "local" coal. Note, by the
way, the importance of considering the costs of particulate controls in these
types of calculations, especially wh.en western coals with relatively low ash
conductivity are being compared with eastern and midwestern competitors.
These examples illustrate the types of engineering calculations we have car-
ried out for many different situations. While several of the assumptions
behind the calculations are subject to argument, of course, the relevant
point in the context of this discussion is simply that there do exist a
variety of situations in which reasonable sets of assumptions lead to the
conclusion that the use of cleaned coals 1s the least-cost option for meet-
ing SCL emission limits. This conclusion holds especially for cases where
the emission limit is 1.2 pounds per million Btu or greater, and it includes
some cases where PCC is used in conjunction with flue gas desulfurization.
303
-------
Table 2
Costs of Alternative Strategies for Meeting Emission Limits
of 1.2 lb/106 Btu S02, 0.1 lb/106 Btu Participates
(Central Ohio Location)
Coal Source* Delivered .Coal Cost
Type U/106 Btu)
West Virginia,
cleaned
Wyoml ng ,
uncleaned
Ohio,
149
156
111
Cost of ESP
U/10& Btu)
15.6
25.5
13.0
Cost of FGD Total .Cost
U/106 Btu). U/106 Btu)
165
181
65.0 189
uncleaned
304
-------
With these encouraging results in mind, let's look at the barriers that are
impeding their realization. Given the constraints on the length of this
paper, we will highlight only the key points of th.e Teknekron analysis..
Investment Barriers and Strategies for Overcoming Them
Teknekron began with the recognition that PCC alone or in combination with
other strategies had previously been judged by several engineering studies-
including some of our own—to be the least expensive way for electric util-
ities and other coal-burning industries to comply with the SOg emission
standards, especially standards of around 1.2 pounds per million Btu or
greater. But we also recognized another critical point—namely, that least-
cost solutions from the customary engineering perspective do not always prove
feasible or even least cost whert certain nonquantifiable or hard-to-quant1fy
factors are taken into account.
Accordingly, we embarked on first-hand interviews with a variety of potential
PCC investors to identify the factors inhibiting the use of PCC for S02 emis-
sion control. We spoke with bankers in New York, Chicago, and Pittsburgh who
have been intimately involved in the financing of cleaning plants. We also
spoke with representatives from a variety of coal companies, utilities, and
industries that are now planning to burn coal.
Before examining our analysis of these interviews, let's look at the Invest-
ment environment in general.
All Investment decisions, whether related to PCC or not, are made under un-
certainty. Moreover, each investment decision is part of a web of other de-
cisions—part of a portfolio that covers many different types of Investment,
each with particular risk factor and an appropriate return.
305
-------
Even if investments were not made in this interdependent manner, each would
require some combination of assurances. The investor would want some assur-
ance of a market. He would want some assurance of rapid payback or of a high
discount rate applied to anticipated cash flows. He would want a relatively
high rate of return on his investment. And he would want reasonable options
to pursue in case the market for his service or product disappeared.
With all these assurances there would still be risk; and even with investments
of relatively little risk, there might be institutional or legal constraints.
Circumventing such constraints could be costly—for example, it would be costly
to rely on internal funds when borrowed funds would be cheaper. Generally,
such decisions are not made without extremely compelling reasons—for example,
to stay in business.
The investment environment fn general, then, is complicated and filled with
uncertainty. For the potential coal-cleaning investor, as we shall find, the
complexity and uncertainty are magnified. Getting from here to there—turning
a contemplated coal-cleaning investment tnto a real one—means overcoming a
range of barriers: barriers of data deficiency, institutional barriers, and
economic and regulatory barriers.
Barriers of Data Deficiency
To begin with, the potential applications,.or markets, for coal cleaning are
not well understood. The key factor here is uncertainty regarding future S0«
emission standards. On the one hand, there is movement toward greater strin-
gency, through a tightening of New Source Performance Standards and State Im-
plementation Plans. On the other hand, the current standards are often not
being enforced (at least one-third of all coal burned in utility boilers is
306
-------
technically in noncompliance), and in some cases SIPs are being relaxed [which
of course tends to increase the proportion of coals in compliance).
In a very real sense, the issue for the potential PCC investor is not what the
standards are but what they will be. Here are the questions the investor is
asking now:
• What specific standards will be set, will be enforced, for what period
of time, in what state or region, for what particular site, and for
which plant categories?
• And, assuming reasonably reliable answers to these questions, what
design and "investment alternative for meeting future SO* limitations
will prove to be the least-cost or otherwise "best" option?
The answers, which are very difficult to come by, are particularly vital for
coal consumers, And they are of great concern also to coal companies, equip-
ment manufacturers, and regulatory authorities—in other words, to the commer-
cial parties who need the answers for marketing and contract negotiations, and
to the public authorities who need them in order to respond with appropriate
regulatory, legislative, and R & D policies.
As matters now stand, the vast number of circumstances and control options to
consider can be overwhelming. Without the benefit of some very sophisticated
modeling—such as the Utility Simulation and Coal Assignment models, which we
have developed over three years at a sizable investment of EPA research funds—
the current or would-be coal consumer may be so defeated by the plethora of
Possibilities that he will fall back on those options that seem most familiar
or most risk-free. Among the most familiar options, for example, would be the
purchase of low-sulfur western coal. Among the most risk-free, though this is
307
-------
certainly a debatable point, might be the use of scrubbers. But these options
might not be the least costly ones in the long run to the ultimate consumer-*
that is, to the buyer of electricity.
Thus, while PCC either alone or combined with blending or scrubbing may well
be the "best" strategy, it may be ignored.
But what if the potential investor doesn't ignore PCC? At the heart of his
problem then is the matter of matching potential markets with appropriate PCC
technologies. Consider the potential markets. There are five:
• Utility boilers subject to SIP limits
• Utility boilers subject to the current NSPS
• Utility boilers subject to future NSPS
• Industrial boilers subject to SIP limits
0 Industrial boilers subject to forthcoming NSPS
Only one of the five is well-defined at present: the second market, utility
boilers subject to the current NSPS of 1.2 pounds of SOg per million Btu.
Since the NSPS for industrial boilers and the NSPS revisions for utility boil-
ers will soon be forthcoming, educated guesses can be made about the nature of
these potential markets.
Boilers subject to SIPs are something els.e again. SIPs, theoretically tied to
ambient air quality standards, are bound—even designed—-to change. What an
SIP for a particular plant might be two or three years from now (the time it
takes to build a PCC facility) is highly problematic. It could be strongly af-
fected by such developments as th.e use of a greater number of ambient air moni-
toring sites or the use of more sophisticated diffusion modeling techniques.
308
-------
Furthermore* SIP standards vary from state to state, and from region to region
within a state. In addition, some states set standards on a unit-by-unit or
stack-specific basis. All this variation frustrates the planner attempting to
generalize with regard to PCC's role 1n the SIP-affected market. Of course,
a given coal burner knows well what the current standards for his boilers are
—but future standards (the ones of Interest to the potential PCC Investor)
are another matter.
There 1s still another factor that cannot be Ignored: with so many boilers
now being fired 1n noncompllance, the potential PCC Investor wonders whether
noncompHance will continue to be a viable strategy. And what 1f SIP stand-
ards are 1n fact enforced? By what means w111 they be enforced? By length-
ening the averaging period or otherwise making the standards less stringent?
Recall now that we targeted as the heart of the problem the matching of mar-
kets with appropriate PCC technologies. We have just looked at the markets,
which turn out to be a most uncertain matter. Now, looking at technology, we
find that there are three complicating factors.
First, potential Investments are thwarted by the Image PCC has as an ash reduc-
ing, not a sulfur reducing, process.
Second, for those potential Investors who do connect PCC with sulfur removal,
there 1s great concern about the reliability and site-specific applicability
of engineering studies that project processing costs. The Investor motivated
by a preliminary study to undertake a rigorous Investment analysis will find
scores of PCC configurations and scores of cleaning levels from which to choose.
Chances are that he will not be comfortable with the numerous options and range
of costs presented him. Unless we begin very soon to develop a systematic and
309
-------
simplified method for presenting the investor with, realistic options, it is
likely that complexity alone will defeat the widespread adoption of PCC.
The third factor complicating the technical side of the market-technology equa-
tion 1s the startling array of non-PCC and with-PCC combinations. These combi-
nations, listed below, range from F6D with different types of coals to the fir-
Ing of high-sulfur coal with no control technology.
t Use of low sulfur compliance coal.
• Blending of low sulfur and higher sulfur coals.
• FGD with high, medium, or low sulfur coals.
• FGD + PCC with high or medium sulfur coals.
• PCC + blending.
• Chemical coal cleaning with or without FGD and/or blending.
• Use of medium or high sulfur coals without sulfur controls.
The first six options imply stringent emission limitations. The last option
implies relatively lax limitations, or the seeking of variances, or the viola-
tion of standards. The first six could apply to any new or existing source;
and the last option, presumably, only to existing sources.
Now, let's go back for a moment to the first complicating factor—the point
about PCC's image as an ash reducing, not a sulfur reducing, process. Most
of the potential investors we spoke with were surprised that PCC could be con-
sidered seriously as a sulfur removal strategy for most boilers subject to
NSPS revisions. Some were aware that the EPA Administrator has authority
under the Clean Air Act revisions to permit credit toward the percentage removal
requirement for precombustion cleaning, but they were skeptical that PCC com-
bined with FGD could prove to be the least-cost compliance option.
310
-------
The benefits of PCC that our Interviewees did generally know about are listed
in the first column of table 3. Other potential benefits, shown in the second
column, were hardly appreciated at all. Thus, of all benefits identified by
Teknekron, most potential investors are aware of only about one half.
Even knowing of these benefits, the Investor considering the market for boilers
subject to mandatory use of FGD, as in the case of the revised NSPS, might un-
derstandably ignore the PCC technologies 1f their payoffs were not proved to
be dramatically great. Further, the same investor might find 1t easier to fo-
cus on one technology, FGD, with one particular type of coal, than-to consider
a range of technologies applied to a range of coals.
Can we give that investor any rules of tKumb to use in weighing the benefits
of combining PCC with FGD to comply with the NSPS? Certainly Hoffman-Munter
may have made a contribution by estimating that the PCC/FGD combination makes
sense if less than 50 percent of the flue gas need be scrubbed to achieve
compliance. But rules of thumb don't really work for real-world, site-
specific situations. At most, they may motivate the would-be Investor to
undertake further investigation by at least pointing him in the general
direction of a least-cost solution.
Now, what about the potential Investor considering PCC with or without FGD
to meet the current NSPS and SIPs? Here there seems to be a somewhat better
appreciation of PCC's beneficial role. But even here the spate of options
involving different process combinations and coal types tends to complicate
the issue and cloud the financial analyst's understanding. Obviously, too
few benefits of PCC are known to too few people. There's been no great rush
to Invest in PCC for sulfur removal pruposes. Greater dissemination of in-
formation would certainly seem to be in order.
311
-------
Table 3
Benefits of PCC
Known Benefits
Other Benefits
Reduced transportation costs
Reduced ash disposal costs
Reduced FGD investment
Reduced FGD operating costs
Reduced pulverizing costs
Reduced sulfur variability
Reduced boiler maintenance
Reduced ESP costs
Expanded coal supply options
May eliminate need for reheat
after FGD
Increased FGD reliability
312
-------
An uncertain grasp of benefits, an uncertain future for S02 regulations, an
uncertain market, and an uncertain combination of technologies—all these
add up to a serious problem of data deficiency for the potential investor
in PCC.
There are some obvious strategies for reducing the uncertainty: EPA can
continue to disseminate information on the benefits of PCC through periodic
conferences such as this one or through "technology transfer" seminars.
Other strategies should also be pursued. One, briefly mentioned above, is
to develop a systematic and simplified method of analyzing PCC technologies
and applications. The .other is to expand the program carrying out commer-
cial-scale PCC demonstrations.
These initiatives should go far toward breaking down the barriers of data
deficiency. They will not, however, address a second set of factors we have
uncovered. These we group under the heading of "institutional barriers."
Institutional Barriers
Institutional barriers are best discussed in terms of particular types of
would-be investors. Let's look first at coal companies. Teknekron found
no significant institutional barriers in the case of large coal companies
but uncovered real problems for the small firms.
Currently, about 15 percent of the nation's coal is produced in about 80
percent of our mines. Generally, the coal produced at these mines (about
4,000 in number) is sold without any preparation and shipped at single-car
rates, which means this coal costs $2 to $3 more per ton to ship than coal
produced in the same area but from large mines.
313
-------
The difficulty for the small coal company hinges on capital expenditure. If
any coal company, large or small, is to take advantage of unit-train rates,
it must be able to guarantee high-volume movement between a single source and
a single market, and it must have a means for loading the unit-trains quickly
and efficiently. That means getting a coal transloader, and that involves a
major capital expenditure. Small coal producers cannot afford that kind of
investment. Nor can they afford to build preparation plants.
But it is precisely these small companies that may control a significant frac-
tion of the nation's cleanable reserves—especially in Central Appalachia.
How, then, can the small company's resources be tapped for PCC? Teknekron
has considered several possible strategies and has concluded that the most
promising would be for small coal companies to form cooperatives. Through
these cooperatives, several small companies together could finance the kind
of capital investment needed for a PCC plant. There is certainly precedent
for this: co-ops have been successfully organized and operated in several
fields of commerce, most notably in agriculture.
Consider now the barriers faced by electric utilities. Before a utility can
approve a PCC investment, it must be reasonably certain about the answers to
a number of questions. For example:
• How speculative is the PCC investment?
• How does PCC fit into the company's compliance strategies?
§ Will PCC accommodate future needs?
• Are boiler modifications necessary?
• Is it better to contract for cleaned coal than to clean at the power
plant?
314
-------
• Should there be a joint venture with a coal company?
• Should there be a multistream operation for several power plants or
with industrial coal consumers?
• Should the PCC plant be funded through an unregulated subsidiary?
• Should PCC be developed with captive mines?
• Will the PUC approve cost pass through.?
t Will the PUC approve inclusion of PCC tn the rate base?
We cannot discuss all th.e questions in this brief paper but would like to de-
vote a few pages at least to potential problems associated with captive mines
and actions by the regulating public utility commission. PUCs, like utilities,
are under constant and conflicting pressures to ensure that electricity wfll
be supplied on a reliable basis at lowest reasonable cost. Maintaining an ac-
ceptable balance between ensuring reliability, on the one hand, and least
cost, on the other, can prove very difficult. This is especially so since
reliability and cost determinations must be made on a prospective basis, re-
quiring the use of forecasting techniques.
More stringent environmental regulations tend to have adverse effects on both
anticipated reliability and anticipated costs. In this regard, options for
controlling SO* emissions from existing and new coal-fired plants simply add
another set of factors for PUCs to take into account.
One of the ways utilities have been responding to concerns about cost and
about reliability of both coal and electricity supply is to rely more
heavily on long-term contracts. At the same time, growing numbers of
315
-------
utilities are entering the coal-mining business to produce coal for their
own needs (and sometimes for other utilities). In 1976, captive coal repre-
sented about 11 percent of all coal delivered to electric utilities. By
1985, it is expected, this figure will have risen to about 19 percent.
Utilities now control about 10 percent of the nation's recoverable coal
reserves, two-thirds of which are in the western states.
Although utilities entering the coal business are generally large—for ex-
ample, Texas Utilities, TVA, AEP, Pennsylvania P & L, and Duke Power-
smaller utilities are considering cooperative ventures to meet their own
needs. The Western Coal Association is the first example of a co-op being
formed (by ten utilities) for both fuel procurement (contracts negotiated
with independent producers) and production purposes.
The significance of these developments for PCC is that utilities could be
in a position to invest in their own cleaning facilities at the mine—be
it mine-mouth generation or not. Moreover, with the apparent willingness
of at least some PUCs to permit such vertical integration, utilities may
gain valuable bargaining strength in negotiating with independent coal
companies for coal, including cleaned coal.
However, it should be noted that not all PUCs view the benefits (quantifi-
able and nonquantifiable) of vertical integration as greater in all cases
than the costs. This does not mean that such PUCs would necessarily dis-
approve of vertical integration; instead, it might mean that they would
make the possible financial advantages to the utility in the arrangement
less apparent.
316
-------
Three advantages to utilities 1n captive operations are:
• Greater reliability of supply
• Potential for lower costs of fuel
• Leverage In negotiating contracts with Independent coal companies
Other possible benefits include tax advantages and, in the case of unregulated
subsidiaries, the potential for greater return on equity than may be likely
for regulated aspects of the business. The potential for greater return,
however, will not always be realized. There are business risks to consider,
although with an assured market these risks are presumably reduced. PUCs
may see such risks reduced to the extent that the return to the captive
supplier is not "allowed" to be greater than the return to the parent. (In
fact, a PUC cannot really control the return on equity; this Is determined
1n the business and financial marketplace.) The PUC's choice of transfer
prices between supplier and purchaser may prove to be a disincentive to
vertical integration and hence an Institutional barrier to PCC investment.
There are also definite disadvantages to utilities moving into captive oper-
ations. Probably the most significant of these 1s that utility management
is relatively inexperienced in-dealing with mining companies and miners.
Electric utilities are capital Intensive, not labor intensive. A different
set of skills may be required for captive operations. Further, management
resources might be better applied at the very complex generation, transmis-
sion, and distribution end, rather than spread across another area.
The issue of management skills 1s important, too, when considering PCC
operations. Since coal companies have traditionally managed the efforts
of preparation engineers, they have an advantage in considering a PCC
317
-------
Investment for sulfur removal purposes. Power engineers definitely speak
a different language from that of preparation and mining engineers.
Both PUCs and the utilities will be weighing these advantages and disadvan-
tages of the move to captive coal and of coal washing for sulfur removal.
Whether PUCs will prove an Institutional barrier to PCC Investment 1s stm
uncertain. Presumably, 1f PCC 1n captive operations could be demonstrated
clearly to be the least-cost way to produce environmentally acceptable coal,
PUCs would approve the plan.
The final set of Institutional barriers are those faced by Industrial coal
consumers. Here, the most important point to understand is that no industry
will adopt PCC for sulfur removal unless it 1s first convinced that the use
of coal Itself is economically wise.
We know that many firms are now considering the use of coal. Among them are
firms located near a mine or capable of being so located, firms that expect
to have a need for relatively large amounts of coal, and firms that have
space available for coal storage, rail facilities, and PCC waste disposal.
Another Important point to keep in mind is that the new source standards
applicable to industrial boilers will probably be less significant than those
for utility boilers. Obviously, we do not yet know what standard these non-
utility boilers will have to meet; and, any event, recognizing that PSD and
nonattainment policies must be reckoned with, we can expect that the Issue
will be resolved on site-specific, boiler-specific, and perhaps firm-specific
bases. Despite these complications, we believe it reasonable to assume that
the great majority of Industrial boilers built 1n the near future will be
subject to emission limitations that can be met by means other than flue gas
scrubbing.
318
-------
This 1s where coal cleaning enters the scene. In many parts of the country*
new Industrial boilers will be able to meet the applicable S02 emission lim-
itations through the use of cleaned coals. This will happen where the level-
1zed cost of using cleaned coals 1s perceived to be lower than the cost of
using naturally occurring low-sulfur coals or blending low-sulfur with me-
dium-sulfur coals.
But, again, cleaned coals won't have a role to play 1n this market unless
Industry uses coal. And Industry, 1n many cases, won't use coal unless 1t
does one of two things:
• Forms consortla to acquire reserves, finance prep plants and trans-
loaders, obtain unit-train or trainload loads, and acquire land and
facilities for coal storage and handling
• Enters into joint ventures with coal producers to finance the con-
struction of such capital items as PCC plants
There are problems involved in establishing and maintaining these institu-
tional arrangements. One of the basic problems is that, compared with pub-
lic utilities, nonutllity firms face a higher degree of business risk. This
1s manifest 1n utility versus nonutHity "long term" fuel-supply contracts.
The Industrial firm may see a "long-term" contract as spanning three to
five years, whereas utilities generally take it to mean ten to thirty years.
Their monopoly status affords utilities the luxury of such contracts. Non-
utility firms, by necessity, have shorter planning horizons—at least inso-
far as fuel purchases are concerned.
Because their fuel-supply contracts are shorter, or at least have been so
traditionally, nonutllity firms must see to it that the sizable capital
319
-------
Investments made in coal-cleaning plants are protected by extraordinary
financial assurances. This is true whether such investments are made by
coal companies or by the industrial coal consumers themselves.
To provide such assurances, a consortium must be able to demonstrate with
reasonable certainty that it will honor its commitments to the coal company
and absorb any losses or higher costs that may come about should one member
of the consortium be unable to continue to receive cleaned coal. Such as-
surances are perhaps more easily provided through joint ventures between
coal consumers and producers. But there may be a significant problem in
matching the asset backing and reliability of the parties to the venture.
Another possible arrangement for nonutility firms to consider is a joint
venture with a utility—perhaps in a multistream coal-cleaning facility
located at the plant or at the mine. Significant savings could result in
financing, in economies of scale in cleaning plants, and in transport. If
the cleaning facility were located at the plant, a particularly good arrange-
ment would work where cogeneration projects are contemplated.
However, joint ventures pose many potential institutional problems that would
need to be resolved. For example, nonutility firms would be particularly con-
cerned lest joint ventures involving the regulated portions of electric util-
ities lead regulatory agencies to believe they have authority to scrutinize
the accounts of the unregulated firms. Furthermore, if PUCs were to take in-
terest in the contractual arrangements established between utilities and in-
dustrial firms, they could require as protection for utility investment that
these contracts be of the long-term variety that might make nonutility firms
uncomfortable.
320
-------
We proceed now to the final barrier uncovered by Teknekron. This harrier
involves two constraints which can be seen as either economic or regulatory.
Economic/Regulatory Barrier
As indicated earlier, it is only recently that the focus on coal cleaning
has broadened to include sulfur as well as ash removal. S'ince emission
standards in many cases are moving toward greater stringency, utilities
and some other coal-burning firms are paying more and more attention to
coal cleaning as an alternative to the expensive compliance options of
buying low-sulfur western coal or installing scrubbers.
But a major factor that may inhibit the adoption of coal cleaning as a sul-
fur removal strategy is that, unlike flue gas desulfurization, cleaning
plant investments do not now qualify for investment tax credits or as pol-
lution control investments. These determinations, made by the IRS, have put
coal cleaning on an unequal footing with scrubbers as well as with naturally
occurring low-sulfur coals (which of course do not require capital investment
of the kind discussed here).
The effect of a PCC plant's failure to qualify for the investment tax credit
is obvious: higher taxes payable to both states and the federal government.
The ruling in this regard, having to do with whether the plant is in fact a.
single entity or simply an assemblage of parts, could certainly be reviewed
in depth with an eye toward urging the IRS to reconsider its posture.
The effect of PCC's failure to qualify as a pollution control investment is
somewhat more complex. The ruling on this matter is based on th.e IRS's be-
lief that PCC investments are made principally for purposes of gaining a
321
-------
commercial advantage over firms that do not clean the coal they sell. The
fact that substances which may pollute are removed in the cleaning process
is seen as incidental to the principal purpose of the investment.
Section III (a)(l) of the 1977 amendments to the Clean Air Act may provide
the rationale for suggesting that the IRS reconsider this ruling. This sec-
tion authorizes the EPA Administrator to permit "credit" toward the percen-
tage removal requirement in NSPS revisions for precombustion cleaning of
coal. Since the only other way to receive "credit" 1s by using flue gas
desulfurization, and since FGD now qualifies as a pollution control invest-
ment, this would appear to offer prima facie evidence that precombustion
cleaning is placed on unequal footing with postcombustion cleaning as a
pollution control option.
If PCC were granted status as a pollution control investment, PCC investors
might reap as many as four financial and economic benefits:
• Accelerated depreciation (over five years) could be taken for tax
purposes. At present, a PCC plant is depreciable over its anticipated
useful life.
• A five percent investment tax credit might be made available, although
a parallel ruling on the separate but related issue of investment tax
credit qualification would be required.
• Tax-exempt pollution control bonds might be made available to finance
the facility. The economic benefit of such financing would vary from
firm to firm.
• PCC operations in some states might be granted exemption from state
sales and use taxes.
322
-------
We believe that steps should be taken to urge the IRS to reconsider its
rulings. It is not possible at this time to estimate with any certainty
what the combined effect of these initiatives would be on per-ton cleaning
charges. These matters are quite firm-specific, plant-specific, and state-
specific. However, we believe the initiatives would strongly enhance the
market prospects for coal cleaning as an S02 removal process.
323
-------
ECONOMICS OF CQAL CLEANING AND FLUE GAS DESULFURIZATION FOR
COMPLIANCE WITH REVISED NSPS FOR UTILITY BOILERS
Randy M. Cole
Energy Research-Combustion Systems
Tennessee Valley Authority
Chattanooga, Tennessee
ABSTRACT
Coal quality is declining and the effects of ash composition, ash, and
sulfur have increased the frequency of unscheduled outages and deratings and
increased operating and maintenance costs. Removing the mineral matter and
pyritic sulfur from the coal prior to combustion will improve power plant
performance and reduce sulfur dioxide emissions as shown in this paper.
In this study coal cleaning, combined with flue gas desulfurization,
was more cost effective when compared with flue gas desulfurization alone.
Although the capital investment and preparation costs are higher, the credits
gained by providing a superior product offset these costs. Removal of mineral
matter and pyritic sulfur by coal cleaning reduces the investment cost for
flue gas desulfurization systems. These savings exceed the investment cost
for coal cleaning.
Case 1 assumes run-of-mine coal is fired in two 1000-MW units followed
by flue gas desulfurization. Case 2 assumes 50 percent of the mineral matter
and 60 percent of the pyritic sulfur is removed before combustion, followed
by flue gas desulfurization.
324
-------
The author wishes to acknowledge the following TVA personnel for
their contribution to the paper:
Donald Anson
John G. Holmes, Jr.
Marvin N. Jarrett
Neal D. Moore
Stephen R. Smith
Charles D. Stephenson
William A. Thomas, Jr.
In addition, the following publications were helpful:
Gibbs and Hill, Inc., Coal Preparation for Combustion and Conversion,
Electric Power Research Institute, EPRI AF-791 Final Report, May 1978.
Joseph W. Leonard and David R. Mitchell, Coal Preparation. The
American Institute of Mining, Metallurgical, and Petroleum Engineers,
Inc., New York, 1968.
DISCLAIMER
This speech was prepared by Randy M. Cole, an employee of the
Tennessee Valley Authority (TVA). The contents do not necessarily
reflect the views and policies of TVA, nor does mention of trade names
or commercial reports constitute endorsement or recommendation for
use.
325
-------
I. Introduction
The 1977 Clean Air Act Amendments specify that new stationary coal-
fired sources regulated by EPA must use the best available control tech-
nology, use a method of continuous pollution control, and achieve a
percentage reduction of the regulated pollutants. Any reduction of a
pollutant by post extraction fuel processing may be credited to the
percentage reduction requirement.
EPA is currently preparing to propose revised New Source Performance
Standards (NSPS) for fossil-fueled boilers for power generation. The
regulations under consideration require a minimum 85 percent reduction
(24-hour average) in sulfur between the point of coal extraction and the
point of discharge of combustion products to the atmosphere and limit
sulfur emissions to no more than 1.2 Ib SCL/10 Btu (24-hour average).
A separate provision of the standards will permit a 75 percent minimum
reduction in sulfur and an exemption of the 1.2 Ib S02/10 Btu emission
limit three days per month. This provision is to allow for variations
in fuel sulfur levels and pollution control equipment performance.
While promulgation of these regulations would effectively preclude the
use of coal cleaning as a sole method for complying with SCL standards
in new electric utility boilers, coal cleaning with scrubbing may be
used in some cases in a cost-effective manner to meet the revised NSPS.
In the following exercise some of the important factors that can
affect the cost of producing electricity from coal while meeting air
quality standards are illustrated.
326
-------
Cases 1, 3, and 5 assume uncleaned coal is fired in the two 1,000-MW
units followed by flue gas desulfurization. Cases-2, 4, and 6 assume
50 percent of the mineral matter and 60 percent of the pyritic sulfur
are removed before combustion followed by flue gas desulfurization. In
all of the coals the study assumes no changes in the basic power plant
design, even though the electrical output and other parameters (such
as coal consumption) may change slightly. Coal cleaning and FGD costs
were estimated from the necessary requirements to fulfill the needs of
the basic power plant design common to all cases.
The basic power plant design consists of two 1,000-MW pulverized
coal units designed with a heat rate of 9,000 Btu/kWh and a capacity
factor of 65 percent. The operating life of the plant is 30 years.
The yearly design generation is 11 billion kilowatthours, utilizing 5
million tons of coal a year. The assumed raw coal analyses are:
Case 1 & 2 Case 3 & 4 Case 5 & 6
Moisture, %
Ash, %
Total Sulfur, %
Pyrite, %
Sulfate, %
Organic, %
Heating Value
(as rec'd)
Cost, $/ton 22 20 15
327
W. Ky. 11
6.7
20.55
4.5
2.39
0.43
1.64
10,416
W. Ky. 9
6.55
14.85
4.5
2.35
0.43
1.72
11,363
W. Ky. 12
5.75
26.05
4.7
3.08
0.21
1.40
9,538
-------
In all cases 85 percent removal of sulfur is required between the
mine and the stack discharge. All the sulfur input to the furnace
totally evolves as sulfur dioxide. In all cases assume the ash over-
head to be 80 percent requiring 99.5 percent fly ash removal within the
scrubber.
Penalty
Coal properties that have the most effect on boiler operation are
ash composition, ash content, and sulfur content. Ash composition affects
and influences the slagging of furnace walls and fouling of convection
passes. Fouling decreases heat transfer and promotes wastage by external
erosion in the convection passes, the induced-draft fans, and plugs the
air preheaters. Excessive ash content overloads the electrostatic
precipitators and bottom ash handling equipment. Sulfur content
influences the operation and maintenance of feeders, pulverizers,
furnaces, soot blowers, air preheaters, dust collectors, and induced-
draft fans. Pyrites cause excessive wear of the pulverizer internals.
Ash and sulfur contents directly affect the heating value of the coal
and along with moisture, limit the capacity of the combustion system.
Coal quality has declined in recent years and has contributed to
an increase in unscheduled outages, unit deratings, and operating and
maintenance costs.
Figure 1 shows the general trend of the relationship between coal
quality and coal quality related costs. The trend is essentially linear
below a point where the boiler is designed to handle coal of a certain
quality. As coal quality declines further, the costs rise exponentially
and, at some point the unit cannot be operated without an excessive
forced outage rate or an unacceptable loss in generating capacity. The
328
-------
01
o
o
to
VO
o
o
o
Cost Factors
• transportation
• maintenance
• ash disposal
• coal handling
• soot blowing
COST
Additional Cost Factors
•peaking capacity loss
•rated capacity loss
•availability loss
12.5
175
Ash plus Sulfur content of coal
FIGURE 1
-------
value of 12.5 percent for ash plus sulfur was assumed on the basis that
this is the best quality coal obtainable through purchases in the Tennessee
Valley region or that can be obtained by beneficiating available poor
quality coals. The value of 17.5 percent for the ash plus sulfur was
assumed on the basis that this is the worst quality coal that the boilers
were designed to utilize. In addition, experience has shown that coals
of poorer quality cause some or all of the additional cost factors. The
penalty for operating a unit on coal where the sum of the ash and sulfur
content totals more than 12.5 percent includes:
Transportation Costs—Assume a freight penalty of $2.50 per ton
per percent above 12.5 percent and an additional charge for excessive
moisture content above 10 percent.
Cost = 2.50 (ash + sulfur - 12.5 + moisture - 10)(0.01)
Maintenance Costs — Assume labor costs to be 50 percent of power
plant maintenance costs. The coal contract price adjustment since 1953
has been 0.1 cent per 10 Btu for ash plus sulfur above 12.5 percent,
Trades and labor rates have increased since 1953 from $5,112 to $15,100
per year, or by a factor of 2.95. Overtime amounts to 15 percent of
labor costs.
Maintenance Cost Factor =
(0.1) 2.95 + (0.1) 2.95 (1.15)
per
Cost = 0.635 (ash + sulfur - 12.5) () (2,000
Ash Disposal Costs—Assume $2.22 per ton of ash. A coal with ash
plus sulfur of 12.5 percent contains 10 percent ash.
330
-------
r _ f(20 lb ash/ton coal)1 , $2.22 ^ e I ton ash x
°sc ~ l 1% ash over 10% J ^ton ashj ^2000 lb ash'
$0.0222
1% ash over 10% per ton coal
Cost = (ash - 10) 0.0222 = $—T
v ton coal
The following additional costs are incurred for ash plus sulfur
content greater than 17.5 percent (exponential portion of curve in
Figure 1).
Peaking Capacity Loss—Assume wide open valve peaking capacity is
required 2 percent of the time or 172.5 hours per year. The unit capacity
factor is 45 percent. Peaking capacity is replaced with gas turbines at
an incremental cost of 32 mils per kilowatthour. Peaking capacity is
10 percent of rated capacity. One-third peaking capacity is lost for
each percentage point that the ash plus sulfur content exceeds 17.5 percent.
«w n i • i (ash + sulfur) - 17.5 .nA 0/
% Peaking loss = 20 5 - 17 5 x !0° = %
Cost =
capacity) (0.10) (% Peaking loss) (year ) (Incremental cost) (Heating value)
(Heat rate) (Gross Generation)
Rated Capacity Loss--Assume rated capacity is lost at the rate of 3
percent for each percentage point that ash plus sulfur exceeds 17.5 per-
cent. Rated capacity loss is replaced 50 percent of the time at an
incremental cost of 10 mils per kilowatthour. The unit capacity factor
is 56 percent.
Rated capacity loss = (ash + sulfur - 17.5) (0.03) (rated capacity)
HR
Generation loss = (8760 ^) (capacity factor) (rated capacity loss)
r t (generation loss) (incremental cost) (.5)
sv
year
331
-------
Availability Loss--Assume 1 percent loss in availability for each
percentage point that ash plus sulfur exceeds 17.5 percent. The unit
capacity factor of 56 percent corresponds to an availability of 72 percent.
Availability loss = (ash + sulfur) - 17.5
Cost =
(lost generation) (incremental cost) (.5) (heating value) (2000 —-')
(normal generation - lost generation) (heat rate)
The penalties are summarized as follows.
Case
Transportation 0
Maintenance 1
Ash Disposal 0
Peaking Capacity Loss 0
Rated Capacity Loss 3
Availability Loss 1
_$_
ton 6
Flue Gas Desulfurization
1 (WK11)
.26
.66
.23
.29
'.15
.22
.81
Case 3 (WK9)
0.085
0.989
0.108
0.196
0.48
0.30
21258
Case 5 (WK12)
0.35
2.21
0.36
0.27
5.86
2.17
11.22
Flue gas desulfurization (FGD) is required to meet the New Source
Performance Standard. Cases 1, 3, and 5 assume FGD is used alone to
meet NSPS. Cases 2, 4, and 6 assume coal cleaning followed by FGD.
A Turbulent Contact Absorber (TCA) was selected for the FGD system,
with limestone as the absorbent. The design premise for the FGD is
shown in Table I. The FGD capital investment and the annual revenue
requirements are shown in Figures 2 through 7.
332
-------
Table I
TCA SCRUBBER SYSTEM FOR 1,000-MW UNIT
Scrubber Description
Number of Operating Scrubbing Trains 8
Number of Redundant Scrubbing Trains 1
Number Beds Per Train 3
Height of Spheres Per Bed (inches) 5
Liquid to Gas Ratio (gal/1,000 ACF) 55
Scrubber Gas Velocity (ft/sec) 12.5
TCA Pressure Drop Across Three Beds (in. H20) 8.6
Total System Pressure Drop (in. H20) 14.8
„ . . . r Mole CaCoa * i c
Stoichiometry Ratio (Mole SQ2 Absorbed) 1<5
Entrainment Level (wt. %) 0.10
EHT Residence Time (min) 12
S02 Oxidized in System % 30
Solids in Recirculated Slurry (wt. %) 15
Scrubber Inlet pH 5.89
Steam Reheater (In-line)
Superficial Gas Velocity (ft/sec) 25
Saturated Steam Temp (°F) 470
Steam Consumption (Ib/hr) 217,700
Inlet Flue Gas Temp (°F) 126
Outlet Flue Gas Temp (°F) 175
Solids Disposal System
Solids in System Sludge Discharge (% wt.) 40
Available Pond Area (Acres) 9,999
Maximum Pond Depth 25
Distance to Pond (mile) 1
Absorbent
CaC03 in Limestone (% wt.) 97
333
-------
- 9
CO
WK II
7.56 GRAINS/SCF
50 60 7O
S02 REMOVAL (%)
100
Capitol Investment ond Operating Costs for Limestone Scrubbing vs. SOo Removal
FIGURE 2
-------
I4O
IO
Ui
130
120
no
100
90
80
70
60
SO
4O
30
20
IO
2O
WK II
3.78 GRAINS/SCF
_L
JL
30
40
50 60 70
S02 REMOVAL (%}
80
Capitol investment and Operating Costs for Limestone Scrubbing vs.
FIGURE 3
90
Removal
8
6
5
4 ^
£
je
>^
^
100
-------
I4O
I3O
I EC
NO
Oa
fa*
Ok
90
8O
70
GO
SO
4O
30
EO
10
O
IU
2O
WK9
4.95 GRAINS/SCF
I
I
3O
40
50 60 70
SOfc REMOVAL (%)
80
90
8
6
5
3 ^
'E
100
Copitoi Investment and Operating Costs for Limestone Scrubbing vs. SOo Removal
FIGURE 4
-------
**>
•W-
140
130
120 -
110
ICO
90
8O
70
6O
SO
4O
30
20
10
20
Capital Investment and Operating Costs for Limestone Scrubbing vs.
FIGURE 7
Removal
WK9
2.47 GRAINS/SCF
I
30
40
50 60 70
S02 REMOVAL (%)
60
90
Capitol Investment and Operating Costs for Limestone Scrubbing vs. SOg Removal
FIGURE 5
10
9
8
5
4
3 -=
£
100
-------
I4O
bl
2O
40
5O 60 7O
S02 REMOVAL (%)
8O
90
- 9
100
Capitol Investment and Operating Costs for Limestone Scrubbing vs.
FIGURE 6
Removal
-------
Ul
VO
140
I3O ~
120 ~
no -
100 -
WK 12
5.2 6RAINS/SCF
30
4O
50 6O 7O
S02 REMOVAL (%)
8O
90
100
-------
Coal-Cleaning Plant
The coal-cleaning plant is assumed to be a conventional heavy media
system costing about $35 million. The input capacity is 1,800 TPH,
consisting of three-600 TPH circuits with two of the three circuits
operating three shifts per day for a minimum of 230 days per year. The
annual output of clean coal is about 5.63 million tons. The yield, Btu
recovery, and operating costs are summarized in Tables II through VII.
Discussion
Cases 1, 3, and 5 assume run-of-mine coal is fired in the two 1,000-MW
units followed by flue gas desulfurization. The pulverized coal units were
designed with a heat rate of 9,000 Btu/kWh and a capacity factor of 65
percent. A heat rate of 10,000 Btu/kWh and capacity factor of 45 percent
was assumed for the cases where uncleaned coal was used. The gross annual
generation decreased from a design output of 11.39 billion kWh to 7.88
billion kWh for these cases.
Cases 2, 4, and 6 assume 50 percent of the mineral matter and 60
percent of the pyritic sulfur was removed before combustion followed by
flue gas desulfurization. The heat rate was held constant at 10,000 Btu/kWh.
It was assumed that by burning a better quality fuel the capacity factor
increased from 45 to 56 percent. The gross annual generation increased
from 7.88 to 9.81 billion kWh.
In case 1 the power plant burned West Kentucky No. 11 with 20.5
percent ash, 4.5 sulfur, and 10,416 Btu per pound. The particulate
emission to the TCA scrubber was 7.56 grains per standard cubic foot
(SCF). The particulate emission leaving the scrubber was 0.04 grains
per SCF, resulting in the required 99.5 percent removal. The sulfur
dioxide emission to the scrubber was 8.6 pounds per million Btu requiring
340
-------
TABLE II
SlIWRY
CASE "2" (WK U)
YIELD FACTOR/ % WEIGHT/ Y
RECOVERY FACTOR/ % BTU/ R 96%
Bru CONTENT ROM COAL/ BTU/& BR 10,416
BTU CONTENT CLEAN COAL/ Bru/#/ Bc = BR(R) 11,764
Y
HOURLY INPUT/ RDM COAL/ TPH 1,200
HOURLY OUTPUT/ CLEAN COAL/ TPH 1,020
TOTAL CAPITAL INVESTMENT/ $ 35 x
COST PER TPH ROM COAL CAP, $/TPH 29/200
COST PER ANNUAL TON CLEAN COAL/ $/n?H $6,22
FIXED CHARGES ON CAPITAL $/TON c,c, 1,33
08M COST INCLUDING REFUSE DISPOSAL/ $/TON c,c, 0,97
COST OF PREPARATION/ VToN c,c, (P/C) 2,30
COST OF BTU Loss/ $/TON c,c, 1,12
TOTAL COST OF PREPARATION/ $/ToN c,c, 3,42
COST OF ROM COAL/ $/[ON c,c, 22,08
COST OF CLEAN COAL/ $/TON 25,1 50
COST OF PREPARATION/ $/]D5 BTU/ (Cp) 0,0978
COST OF BTU Loss/ $/JjfT BTU, (CJ O.M77
TOTAL COST OF PREP/ $/]06 BTU/ (O OJ.A55
COST OF ROM COAL/ $/106 BTU/ (CJ LOB
COST OF CLEAN COAL/ S/JD6 BTU (Cc) 1,2055
COST OF CLEAN COAL/ $/TON 28,36
CREDIT FOR IMPROVED HEATING VALUE/ $/FoN -2,05
COST OF CLEAN COAL/ $/TON 25,50
TOTAL COST OF PREPARATION/ MILS/KWH L'I6
341
-------
TABLE III
ESTIMATE) ANNUAL REVENUE REQUIRSW #/TON
LABOR COSTS CLEAN COAL
SUPERVISION ,0500
OPERATION ,0327
MAINTENANCE .1338
,2665
UTILITIES 8 PROCESS RELATED COSTS
POWER 9500 HP 0,020 $Mln ,139 $/ToN c,c,
WATER 7206PM 0,020 $/K GAL, ,00085
MAGNETIC Loss 1,5 tf/ToN 0,0325 $/# ,0574
FLOCCULANT 6, *'/HR 1,50 W ,0088
LUBRICANT ,005
THERMAL DRYER G.7TPH 1,19 V106 BTU ,184
REFUSE DISPOSAL 180 TPH 1,00-VToN ,176
BTU Loss 3000Bru/# 1,06 $/10^ BTU 1,123
MAINT, MATS, 50/50 WITH MAINT, LABOR .1338
1.8279
FIXED COSTS
GEN. AD. EXP. (60%) & M LABOR) ,1299
PROPERTY TAXES & INS, (2,5% x PL, INV.) ,1554
PLANT DEPRECIATION STR, LINE (30 YR.) ,2072
CAPITAL CHARGE (11%) ,715
CONTINGENCY .1208
1,3283
TOTAL COST OF PREPARATION/ S/TON c,c, 3,4227
ANNUAL PREP, COST 342 $19/269/801
-------
TABLE IV
SlliWRY
(UK 9)
YIELD FACTOR/ % WEIGHT/ Y 85
RECOVERY FACTOR/ % BTU, R 94
BTU CONTENT ROM COAL/ BTU/#, BR 11,363
BTU CONTENT CLEAN COAL/ BTU/#/ Bc • BR (R) 12/566
Y
HOURLY INPUT/ ROM COAL/ TPH ]200
HOURLY OUTPUT/ CLEAN COAL/ TPH XRO
TOTAL CAPITAL INVESTMENT, $ 35 x
COST PER TPH RDM COAL CAP, $/TPH $29,200
COST PER ANNUAL TON CLEAN COAL, S/TPH 6,22
FIXED CHARGES ON CAPITAL $/ToN c,c, 1,33
08M COST INCLUDING. REFUSE DISPOSAL, $/TON c,c, 0.94
COST OF PREPARATION/ $/TON c,c, (P/C) 2,27
COST OF BTU Loss, $/TON c,c, L*I3
TOTAL COST OF PREPARATION, $/ToN c.c, 3,70
COST OF RDM COAL, $/TON c,c. 2QiQQ,
COST OF CLEAN COAL, $/!ON 23,70
COST OF PREPARATION, $/106 BTU, (Cp) 0,0903
COST OF BTU Loss/ $/L06 BTU/ (CJ Q.IBffi
TOTAL COST OF PREP/ S/IO6 BTU, (Cj.) O.W72
COST OF ROH COAL, $/]Jr BTU, (CJ JLSS
COST OF CLEAN COAL, $/10fi BTU (Cc) 1.0272
COST OF CLEAN COAL, $/TON 25,82
CREDIT FOR IMPROVED HEATING VALUE, $/ToN — 2jJ2
COST OF CLEAN COAL, $/TON 23,70
TOTAL COST OF PREPARATION, Mius/kNH l-^
343
-------
TABLE V
WK9
ESTIMATED ANNUAL REVENUE REQUIREMENT
LABOR COSTS
SUPERVISION
OPERATION
MAINTENANCE
UTILITIES & PROCESS RELATED COSTS
POWER
WATER
MAGNETITE LOSS 1,5 |^ 0,032
FLOCCULANT 6 j^ 1,50
LUBRICANT
THERMAL DRYER 6,3 TPH 0,97
REFUSE DISPOSAL 180 TPH 1,00 i
Bru LOSS
4609 0,
106BTU
MAINTENANCE MATERIAL 50/50 WITH MAINT, LABOR
FIXED COSTS
GENERAL ADMINISTRATIVE EXPENSE (60% 08M LABOR)
PROPERTY TAXES & INSURANCE (2,5% x PLT INV.)
PLANT DEPRECIATION (STR, LINE 3D YR.)
CAPITAL CHARGE
CONTINGENCY
TON c.c.
,0500
,0827
.3338
,2665
,3390
,00085
,0574
,0050
,1501
,176
1,4287
2,09965
,J299
,2072
.3208
1,3283
344
3.GOV6 20,799,75/1
-------
TABLE VI
STORY
6 OIK 12)
YIELD FACTOR, I WEIGHT, Y 35
RECOVERY FACTOR/ % BTU, R 94
BTU CONTENT RO^ COAL, BTU/#, Rn 9,538
BTU CONTENT CLEAN COAL, BTU/#, BC = BR(R) 10,548
Y
HOURLY INPUT, RfW COAL, TPH 1,290
HOURLY OUTPUT, CLEAN COAL, TPH 1,020
TOTAL CAPITAL INVESTMENT, $ 35 x 10°
COST PER TPH ROM COAL CAP, $/TPH 29,2.00
COST PER ANNUAL TON CLEAN COAL, $/TPH 6,22
FIXED CHARGES ON CAPITAL $/TON c,c, 1,33
ORM COST INCLUDING REFUSE DISPOSAL, $/!"ON c,c, .9?
COST OF PREPARATION, $/ToN c,c, (P/C) 2,25
COST OF BTU Loss, $/TOM c.c, 1.05
TOTAL COST OF PREPARATION, $/ToM c,c, 3,30
COST OF ROM COAL, $/ToN c,c, 15.00
COST OF CLEAN COAL, $/TON 18,30
COST OF PREPARATION, $/106 BTU, (Cp) 0,1065
COST OF BTU Loss, $/106 BTU, (CJ 0,0493
i6
TOTAL COST OF PREP, $/JOD BTU, (CT) 0,1564
COST OF RDM COAL, V10U'BTU, (CJ 0.786
COST OF CLEAN COAL, $/105 BTU (CJ f
c
COST OF CLEAN COAL, $/TON 19,89
CREDIT FOR IMPROVED HEATING VALUE, $/TON .1.59
COST OF CLEAN COAL, VToN 18.30
TOTAL COST OF PREPARATION, MiLs/Kl/fn 1,56
345
-------
TABLE VII
VfK 12
ESTIIWET1 ANNIJAI RFVFN1IF RFQUIRRfNT
LABOR COSTS
SUPERVISION
OPERATION
MAINTENANCE
UTILITIES ft PRQCFSS RELATED COSTS
POWER (9500 HP)
WATER 720 GPM
MAGNETITC 1,5
FLOCCULANT
LUBRICANT
THERMAL DRYER 7,'fe TPH
REFUSE DISPOSAL 180 TPH
BTU LOSS 3800
MAINTENANCE MATERIAL 5/50 f^inr, LABOR
FIXED COSTS
GENERAL ADMINISTRATIVE EXPENSE (60% Q8M LABOR)
PROPERTY TAXES & INSURANCE (2,S£ x PLT, INV.)
PLT, DEPRECIATION (STR, LINE 30 YR)
CAPITAL CHARGE (3JX)
CONTINGENCY
0,010
0,032
0,906 *
^fe
0,786_f
TON C,C,
0,0500
0,0827
0.1358
0,2665
0,139
0,00085
0,057/1
0,0088
0,0050
O.LW)
0,176
1,05
1,70185
,1299
,1554
,2072
,715
1,3283
3,29965
18,609,121
346
-------
an 86 percent removal efficiency to achieve the 1.2 pound standard. The
TCA scrubber system cost $244.5 million. The power plant production
cost was $98.19 million per year. A penalty of $6.81 per ton was
assessed for the use of this coal resulting in an additional annual
cost of $25.74 million. The scrubber operating cost was $58.3 million
per year. The total generation cost was $182.2 million per year or
23.1 mils per kWh.
In case 2 the West Kentucky No. 11 coal was beneficiated in a
heavy media coal-cleaning plant. The ash and sulfur content were
reduced to 10.3 and 2.7 percent, respectively. The yield from the coal-
cleaning plant was 85 percent. The cleaning plant was assumed to improve
the heating value of the coal from 10,416 to 11,764 Btu per pound with a
Btu recovery of 96 percent.
When fired by the power plant the particulate emission to the TCA
scrubber was 3.78 grains per SCF. The particulate leaving the scrubber
was 0.02 grains per SCF (99.5 percent removal). The sulfur dioxide
emission to the scrubber was 4.59 pounds per million Btu requiring a
74 percent efficiency to achieve the 1.2 pound standard. The cost of
the coal-cleaning plant was $35 million and the processing cost per ton
of cleaned coal was $3.42. The scrubber cost was $192.6 million. The
power plant production cost was $125.1 million per year. A penalty was
not assessed against the washed coal. The scrubber operating cost was
$55.92 million per year. The total generation cost was $177.1 million
per year or 18.1 mils per kWh, compared to 23.1 for case 1.
In case 3 the power plant burned West Kentucky No. 9 with 14.85
percent ash, 4.5 percent sulfur, and 11,363 Btu per pound. The particulate
emission to the TCA scrubber was 4.95 grains per standard cubic foot (SCF).
347
-------
The particulate emission leaving the scrubber was 0.02 grains per SCF
(99.5 percent removal). The sulfur dioxide emission to the scrubber
was 7.9 pounds per million Btu. An 85 percent removal efficiency
would achieve the 1.2 pound standard. The TCA scrubber system cost
$231.5 million. The power plant production cost was $81.65 million per
year. A penalty of $2.26 per ton was assessed for the use of this coal
resulting in an additional annual cost of $7.5 million. The scrubber
operating cost was $57.6 million per year. The total generation cost
was $146.75 million per year or 18.5 mils per kWh.
In case 4 the West Kentucky No. 9 coal was beneficiated in a heavy
media coal-cleaning plant. The ash and sulfur contents were reduced to
7.5 and 2.7 percent, respectively. The yield from the coal-cleaning
plant was 85 percent. The cleaning plant was assumed to improve tne
heating value of the coal from 11,363 to 12,566 Btu per pound with a
Btu/recovery of 94 percent. When fired by the power plant the particulate
emission to the TCA scrubber was 0.01 grains per SCF (99.5 percent
removal). The sulfur dioxide emission to the scrubber was 4.3 pounds
per million Btu requiring a 72 percent efficiency to achieve the 1.2
pound standard. The cost of the coal-cleaning plant was $35 million and
the processing cost per ton of cleaned coal was $3.70. The scrubber
cost was $203.7 million. The power plant production cost was $108.7
million per year. A penalty was not assessed against the washed coal.
The scrubber operating cost was $60.8 million per year. The total
generation cost was $169.5 million per year or 17.3 mils per kWh,
compared to 18.5 for case 3.
In case 5 the power plant burned West Kentucky No. 12 coal with 26.05
percent ash, 4.7 percent sulfur, and 9,538 Btu per pound. The particulate
348
-------
emission to the TCA scrubber was 10.43 grains per SCF. The particulate
emission leaving the scrubber was 0.03 grains per SCF (99.5 percent
removal). The sulfur dioxide emission to the scrubber was 9.84 pounds
per million Btu requiring an 88 percent removal efficiency to achieve
the 1.2 pound standard. The TCA scrubber system cost $274.1 million.
The power plant production cost was $72.88 million per year. A penalty
of $11.22 per ton was assessed for the use of this coal resulting in an
additional annual cost of $46.3 million. The scrubber operating cost
was $67.8 million per year. The total generation cost was $187.0 million
per year or 23.7 mils per kWh.
In case 6 the West Kentucky No. 12 coal was beneficiated in a heavy
media cleaning plant. The ash and sulfur content were reduced to 13 and
2.6 percent, .respectively. The yield from the coal-cleaning plant was
85 percent. The cleaning plant was assumed to improve the heating value
of the coal from 9,538 to 10,548 Btu per pound with a Btu recovery of 94
percent. When fired by the power plant the particulate emission to the
TCA scrubber was 5.2 grains per SCF. The particulate leaving the scrubber
was 0.03 grains per SCF (99.5 percent removal). The sulfur dioxide
emission to the scrubber was 4.9 pounds per million Btu requiring a 77
percent efficiency to achieve the 1.2 pound standard. The cost of the
coal-cleaning plant was $35 million and the processing cost per ton of
cleaned coal was $3.30. The scrubber cost was $203.7 million. The power
plant production cost was $100 million per year. A penalty was not
assessed against the washed coal. The scrubber operating cost was $56.9
million per year. The total generation cost was $156.9 million per year
or 16 mils per kWh, compared to 23.7 mils per kWh, for case 5. These
cases are summarized in Tables VIII through XVI.
349
-------
TABLE VIII
CASE .1
ROM ^2 - 307) W
COAL UNITS
20,555 ASH
10,416 Em
IT 7.5P SSMiS. nrw GRAINS
" SCF ""H SCF
CAPACITY FACTOR
GENERATION 7,88 x 109 YR
HEAT RATE 10,000 "
FUEL CONSUMPTION 3,78 x
YR
FUEL COST 83,46 x 106
PRODUCTION COST 98,19 x 106
PENALTY (A + S) > J7,5% 25,74 x ID5 ±
FGD INVESTMENT COST 122,24 Ir?
• 2^,5 x .in6 $
FGD REVENUE REQUIREMENT 0,0074
• 58,3 x 106 $
TOTAL GENERATION COST 182,23 x 105 |r
YR
a** i A
= 23,1
350
-------
TABLE IX
CASE 2
*_HOM >- PREP *- POWER *~ FGD >
COAL PLT PLT
#SO?
4,5% S 2,7% S 4,59 -^-
20,55% ASH 10,3% ASH lobBTU
10,436 tai 11,764 Bni g^
# * 3'78 SCF U|1""SCT
PREP PLT INVESTMENT 35 x 3D6 $
PREP COST ± 3,42 $/!ON CLEAN COAL
PREP COST 39,3xl06|p
CAPACITY FACTOR 56%
GENERATION 9,81.x
HEAT RATE 10,000
FUEL CONSUMPTION 4,17 x
FUEL COST 106,34 x 106
in
PRODUCTION COST (8S75 FUEL COST) 325,1 x 3D6 fa
FGD INVESTMENT COST 96,30
- 192,6 x 106 $
FGD REVENUE REQUIRE^NT 0,0057 -^
- 55,92 x 106 fa
TOTAL GENERATION COST 177,1 x Kr fa
MTI O
= 18,1
351
-------
TABLE X
COMPARISON
FGD ONLY WITH COAL CLEWING a FGD
WEST KENTUCKY NO, 11
857 $(>> WAL
CASE 1 ROM COAL ~2ti 1000 MW UNITS
CASE 2 ROM COAL —PREP PLT *-PfVER PLT - FGD
PREP PLT INVESTMENT, 10P $ 35,0
PREP COST a 3,42 =|rr CLEAN COAL
TON
PREP COST, 106 y| 19,3
GENERATION, 109 % 7,88 9,81
YR
PRODUCTION COST, 106 J 98,19 .125,1
YR
PENALTY (A + S > 17,5^), if n| 25,7/1
YR
, 106 $ 2/^,5 192,6
FGD REVENUE .REQUIREMENT/ ID —• 58,3 55,9
TOTAL GENERATION COST, 106 y| 1R2.2 177,1
23,1 18,1
352
-------
TABLE XI
QA$E 3
ROM
5- 2 - 1000 W
COAL UNITS
/i r"°7 c 7 Q
TiJ/o O 'i" _^r_ -UJ-j r
14,852 ASH 106BTU 106BTU
^'Y W5SBAm 0,02
# SCF SCF
CAPACITY FACTOR
GENERATION 7,88 x 109 YR
HEAT RATE 10,000 "
FUEL CONSUMPTION 3,47 x 10B
ir\
FUEL COST 69," x 105 ^
PRODUCTION COST 31,65 x 10G —
PENALTY (A + S) > 17,5? 7,5 x JO6
INVESTMENT COST IB. 75 ^
731,5 x 106 $
FGD REVENUE REQUIREMENT 0,0073
57,6 x 10fi
TOTAL GENERATION COST 146,75 x 106 j-
YR
353
-------
TABLE XII
RDM
COAL
4,5% S
14,85% ASH
11,363 Bm
PREP PLT INVESTMENT
PLT
2,7% S
7,4% ASH
12,566 Bni
2 it?
DRAINS
SCF
PREP COST a 3,70 J- CLEAN COAL
TON
PREP COST
CAPACITY FACTOR
GENERATION
HEAT RATE
FUEL CONSUMPTION
FUEL COST
PRODUCTION COST
FGD INVESTMENT COST
FGD REVENUE REQUIREMENT
TOTAL GENERATION COST
35 x 106
20,8 x 106
56%
9,81 x ID9
3,9 x 106
11
92,43 x 1#
108,7 x 106
GRAINS
SCF
YR
101,86
- 203,7 x 106
0,0062 |g
- 60,8 x 106 t
169,5 x 106 §
17,3
354
-------
TABLE XIII
OTARISON
FGD ONLY WITH GOAL CLEANING ft FGD
WEST KENTUCKY NO, 9
so? RFWAL
CASE 3 RDM COAL >- 28 1000 W UNITS
CASE 4 ROM COAL »- PRF.P PLT *- PNER PLT *- FGD
PREP PLT INVESTNENT, 10 $ 35
PREP COST a 3,70 ^ CLEAN COAL
PREP COST, 105 £ 20,8
GENERATION, 109 ^ 7,W 9,8]
f* d* «•
PRODUCTION COST, 10 ^ yp ^-p5
PENALTY (A + S > 17,570, 106^ 7,5
FGD INVESWENT, 106 $ 251'5
FGD REVENUE
111
TOTAL GENERATION COST, 106 ^|' 116.75 169,5
18,5 17,3
355
-------
TABLE XIV
CAFF 5
. *- 2 - BIO m ?- FGD
COAL
4,7% S 9,84
26,05% ASH
9/538 0 0^
' "
SCF SCF
(88% REMOVAL)
CAPACITY FACTOR L67.
GENERATION 7,88 x Jfr
HEAT RATE 10,000 gm
l\i 17.5S) 15,3 x 10R ^
YR
FGD INVESTMENT 137,05 ^
s 27'l,]D x 106
FGD REVENUE REQUIREMENT ,0086
= 67,8 x 106
TOTAL GENERATION COST 187,0 x 106 —
YR
MTI «
= 23,7 •;
356
-------
TABLE XV
CASES
ROM *~
CQAL PLT PLT
4,7% S 2,6% S -r,. inp_
26,051 ASH 13% ASH wm
9538 Bm 30,548 ton 5 2 SfiAm QQ-SGBAINS.
# # 'SCF "SCF
PREP PLT INVEST^€^rr ^ x
PREP CosT5)3,3QL7C,c,
PREP COST 18,5
1,2
^6
CAPACITY FACTOR
GENERATION y,»i x .ur ^
HEAT RATE 10/om M
r /i PR v irt6 iQNa
FUEL CONSUMPTION !iro x 1J YR
P _ pc i v inB. |_
FUEL COST ra'J- x JJ YR
PRODUCTION COST 10° x 10' YR
INVESTMENT
203,7 x Iff ^
*
REVENUE REQUIREMENT 0|0068 HfB
56,9 x 106
r
TOTAL GENERATION COST 156,9x10
357
-------
TABLE XVI
COMPARISON
FGD ONLY WITH COAL CLEANING ft
WEST KENTIICW NO, .12
857 $(>> RFWAL
CASE 5 RDM COAL *-2* 1000
CASE 6 ROM COAL — -*- PREP. PLT - >- PNER PLT - 5- FGD -
PREP PLT INVESTMENT/ 10 $ 35
PREP COST a 3,30 CLEAN COAL
PREP COST, 106
GENERATION, 109 ^ 7,88 9,81
PRODUCTION COST, 10fi ^ 72,88 100
PENALTY (A + S>37,5&), 10R ^ '16,?
FGD INVESTOR 106 $ 27'M 203,7
FGD REVENUE REQUIREMEMTv W P7,8 5P.P
TOTAL GENERATION COST/ 10" y| W,0 356,9
23,7 16
358
-------
Conclusions
The combination of coal cleaning and flue gas desulfurization is a
better choice than scrubbers alone for the three coals studied. Pre-
combustion cleaning provides some flexibility in operating the scrubber
and reduces the capital investment and operating costs for the absorbent
handling and preparation system, and for sludge disposal. Generally,
these savings will cover the cost of the coal preparation plant. Moreover,
if EPA requires chemical fixation of scrubber sludge, the economics will
tilt more in favor of the combination of precombustion cleaning and flue
gas desulfurization. The greatest potential savings attributable to coal
cleaning are reduced maintenance costs and increased rated capacity. Pre-
combustion cleaning would allow an increase in availability and capacity,
and reduced operating costs. Therefore, coal cleaning is an attractive
approach to achieve energy and environmental goals.
359
-------
THE ECONOMICS OF BENEFICIATING AND MARKETING
HIGH-SULFUR IOWA COAL
C. Phillip Baumel, John J. Miller, and Thomas F. Drinka
Department of Economics
Iowa State University
Ames, Iowa 50011
ABSTRACT
A mathematical programming model is used to evaluate alternative coal
transportation, coal beneficiation, and coal handling systems. The objective
of the analysis is to find the minimum-cost method of transporting and
distributing coal to supply Iowa's projected 1980 coal consumption and, at
the same time, meet the sulfur dioxide emission standards and constraints
on Iowa mining capacity. The model includes 33 potential origins of coal
for the identified 46 major coal users in Iowa. The 1980 projected coal
requirements are specified in Btu's rather than tons to account for the
differences in heating value of coal from different origins. The projected
Btu requirements can be satisfied by obtaining coal directly from seven
existing out-of-state coal sources or from two existing underground mines
in Iowa. Coal from 24 potential Iowa strip mine locations can be used only
if it is cleaned at one of eight potential coal beneficiation plant locations.
Each coal user can blend two or more coals to meet its sulfur dioxide
emission standard.
The model Includes six possible modes of coal transport. These include
truck, barge, single-car rail, 15-car rail, 50-car rail, and 100-car unit
train.
The model minimizes the delivered cost of coal to the user subject to
Btu requirements and S02 constraints. The delivered cost Includes the
FOB mine price of the coal, beneficiation costs if the coal is cleaned,
all transportation and variable receiving costs, and any additional invest-
ments in capacity required to receive larger size rail shipments.
Eight computer solutions obtained in the analysis were based on alter-
native sets of coal prices, rail rates, and truck weight limits.
360
-------
The U.S. Environmental Protection Agency has adopted a national
standard which limits sulfur dioxide emissions to 1.2 pounds of
S02 per million Btu of heat at coal-fired stationary boilers
with a heat input of > 250 million Btu constructed after August 17,
1971 (U.S. Environmental Protection Agency, 1971). Assuming 10,000
Btu per pound, only coal with <_ 0.6 percent sulfur could be burned
in these boilers under this emission standard.
Individual states, counties or cities may establish S02 emission
standards for smaller boilers and boilers constructed before
August 17, 1971. The current S02 emission standards for these
boilers in Iowa .are 5, 6, or 8 pounds of S02 per million Btu, depending
upon the location of the boiler. Assuming 10,000 Btu per pound,
only those coals with sulfur contents <_ 2.5, 3.0, or 4.0 percent,
respectively, could be burned in these boilers under these
emission standards.
Strippable coal reserves in Iowa typically average between
3.1 and 5.8 percent sulfur (Avcin, 1976). This, in part, explains
why Iowa coal production declined from over 1 million tons in
1971 to 540,000 tons in 1976. Only 259,000 tons of Iowa coal were
strip mined in 1975 (U.S. Department of the Interior, 1971, 1976, 1977).
One method of improving the competitive position of this
high sulfur coal may be to reduce the content of sulfur and other
impurities. An experimental coal beneficiation plant operated by
Iowa State University—the only such plant in Iowa—has shown that
the sulfur content of Iowa coal ran be reduced on the average about
35 percent (Grieve and Fisher, 1978). If coal beneficiation will
improve the competitive market position of the high sulfur coal,
361
-------
the optimal number and location of beneficiation plants must be
determined.
Another alternative for improving the competitive market
position of high sulfur coal is to reduce the cost of trans-
porting coal to users. Possible improvements in coal transportation
include larger size rail shipments such as 15- and 50-car units,
alternative types of trucks, and inter-modal truck-rail
combinations.
Method of Analysis
The purpose of this paper is to present estimates of the
impact of alternative transportation and coal beneficiation
systems and coal prices on the marketability of high sulfur Iowa
coal. A mixed integer-linear programming model is specified to
evaluate the feasibility of mining and beneficiating Iowa coal
for use by utility and industrial coal users in Iowa under
alternative combinations of coal prices and rail rates. The
model selects the least cost number and location of beneficiation
plants from 8 possible sites. The objective of the analysis is
to find the minimum cost method of supplying Iowa's coal needs,
subject to constraints on mining capacity, rail receiving capacity
of Iowa coal users, beneficiation plant capacity, sulfur dioxide
emission standards, and projected 1980 coal consumption in Iowa.
The model includes 33 potential sources of coal for the 46 major
utility and industrial coal users in Iowa, The projected 1980
coal consumption by each user is specified in heating units, rather
than tons, to account for differences in the heating value of coal
from different sources. User Btu requirements can be satisfied by
362
-------
obtaining coal directly from the two existing Iowa underground
mines or from the seven out-of-state sources of coal. Because of
its high sulfur content, coal from 24 potential strip mine loca-
tions (Avcin, 1976 and Lemish and Sendlein, 1977) in a- 3%-county
area in Iowa (Figure 1) can be used, only if it is beneficiated
at one of the 8 possible beneficiation plant sites. The average
sulfur and Btu content of the coal at these 24 potential strip
mine locations is presented in Table 1.
In addition to meeting its projected 1980 Btu requirement,
each user must satisfy the limits on sulfur dioxide emissions at
each user location. Each user, however, can blend coal from
two or more sources to meet its emission standard.
The supply of coal at the Iowa and Northern Missouri sources
is constrained by assumed annual mining capacities and estimated
coal reserves. Because Iowa consumes only a small percentage
of the total production of the 6 remaining out-of-state coal
origins, the supply capacity of these 6 sources is not constrained
in the model.
The model includes 6 possible modes of transport from sources
to users in 5 alternative computer solutions. The possible modes
are barge, truck, single-car rail, 15-car rail, 50-car rail, and
100-car unit train. Each .user has the option of receiving coal by
the least-cost mode or combination of modes, subject to its existing
rail receiving capacity. All users are given access to estimated
truck rates from Iowa and Northern Missouri coalmines. Barge trans-
portation is available only to users with existing barge receiving
capabilities. The 4 possible modes of transport from Iowa coal
363
-------
HMNL
K
AREA OF STUDY
AREA OF STRIPPABLE COAL
Figure 1. The selected Iowa coal-producing area,
-------
to
<*
Ul
Origin
Sheridan, Wyoming
Gillette, Wyoming
Canton, Illinois
Sparta, Illinois
West Harrisburg, Illinois
Nortonville, Kentucky,
Unionville, Missouri
T t* m •!• ¥T«I 1 II M II M II •! 1 Vffff I
lowa underground »""«'«
Mine I
Mine II
Potential Iowa Strip Mines
9 sites
3 sites
4 sites
1 site
1 site
2 sites
2 sites
2 sites
Btu per
pound
9.300
8,100
8,100
8,100
11,000
11,400
12,455
11,400
10,500
9,600
10,225
9,794
9,851
10,348
10,900
10,181
10,798
10,294
11,549
Percent
sulfur
content
0.70
0.48
0.48
0.48
3.25
2.90
1.97
2.50
2.62
2.75
4.60
5.25
5.33
5.83
5.60
3.24
3.11
5.49
4.27
Estimated 1977 FOB
mine prices
$12.65
7.65fc c
A
6.40d
24.70
22.20
23.35
22.33
20.16
15.72
13.53
14.95
14.89
14.55
14.70
16.91
17.09
14.78
15.75
FOB mine prices
based on
average Iowa
mining costs
$12.65.
7.65?
7.15j c
6.40d
24.70
22.20
23.35
22.33
21.35
15.72
13.53
16.87
16.81
16.47
16.62
18.83
19.01
16.70
17.67
FOB mine prices
based on high
Iowa mining
costs
$12.65.
7.65?
7.155'c
6.40d
24.70
22.20
23.35
22.33
24.90
15.72
13.53
19.60
19.54
19.20
19.35
21.56
21.74
19.43
20.40
estimated per ton reclamation costs Included in the strip mine prices are as follows: $0.15 for
Wyoming, $0.70 for Illinois, $0.83 for Kentucky, and $1.93 for both Missouri and Iowa; these reclamation costs
were weighted by the percentage of total coal production that is strip mined.
"Required annual volume of 500,000 - 1,500,000 tons.
^Shipments in 50- or 100-car trains.
"Required annual volume greater than 1,500,000 tons shipped in 100-car trains.
Cleaned coal.
Table 1. Estimated FOB mine coal prices based on coal bids and on estimated Iowa mining and reclamation costs,
by coal origin in dollars per ton, 1977.
-------
beneficiation plants to users included in the model are truck,
single-car rail, 15 -car rail, and 50-car rail. Each user is
restricted to its existing rail receiving capacity, unless it
incurs an additional annual fixed cost for expanding to the next
larger rail receiving capacity. If the projected 1980 coal consump
tion would provide less than one shipment per month at the next
larger shipment size or the user has historically received all
of its coal by truck and (or) barge, the user was not given the
opportunity to increase its rail receiving capacity.
The delivered cost of beneficiated Iowa coal includes the
FOB mine price of raw coal, the total annual cost of constructing
a beneficiation plant, variable operating and maintenance costs
of beneficiating the coal, the cost of transporting raw Iowa coal
from the mine to the beneficiation plant, the cost of transporting
the refuse from the beneficiation plant to the mine, and the cost
of transporting the cleaned coal from the beneficiation plant to
the user location.
The model uses continuous variables for the mining, trans-
portation, and beneficiation activities and zero-one integer
variables for the construction of beneficiation plants and the
expansion of rail receiving capacity at users. The model can be
summarized as follows:
(1) Minimize Z - ZP4M.
+ « -
+ hZZZItf.,., + ZZZZ
ijkm ijkin
366
-------
where
Z - total cost of coal unloaded at user locations,
P. - price per unit of coal at origin i,
M^ • volume of coal supplied by origin i,
aikm * transportation plus variable receiving cost per unit
of coal shipped from origin i directly to user k by
mode m,
Uikm " volume shipped from origin i directly to user k by
mode m,
Y - inverse of the fractional weight recovery at benefi-
ciation plants,
bji • transportation cost per unit of coal shipped from
origin i to beneficiation plant site j,
Vi1km " vo^ume °^ clean coal equivalent shipped from origin
i through beneficiation plant site j to user k by
mode m,
c. - transportation cost per unit of refuse and fines shipped
from beneficiation plant site j to mine i,
h - variable beneficiation cost per unit of clean coal,
d1km " transportation plus variable receiving cost per unit
of clean coal shipped from beneficiation plant site j
to user k by mode m,
FCi - annual fixed cost of establishing a beneficiation
plant at site j,
YJ - (0, 1), a binary variable; if site j is used, Y, - 1,
otherwise Y. - 0.
367
-------
EC, * annual fixed cost of expanding the rail receiving
capacity of user k to the next largest size,
Xfc * (0, 1) a binary variable; if user k expands its rail
receiving capacity, X^ » 1, otherwise X^ « 0.
The following constraints were imposed on the model.
The annual volume of coal shipped from an origin cannot
exceed the total annual production capacity of that origin.
(2) II \}1-tm + ms V,,. - M, < MC,
km ikm jkm ijkm i " i
where MC. - total annual production capacity of origin i.
The annual volume of coal processed at a beneficiation plant
site cannot exceed the annual beneficiation plant capacity,
(3) "ZV11, < BC for all j
ikm J
where BC - annual beneficiation plant capacity in units of
clean coal.
The projected consumption of coal at each user must be satisfied.
This projected consumption was specified in heating units rather
than tons to account for differences in the heating value of
coals from different origins,
where 0^ - heating value per unit of raw coal from origin i,
Y£ • heating value per unit of clean coal from origin i, and
D^ - exogenously determined consumption at user k.
Each user was required to meet an aggregate limit on sulfur
dioxide emissions. However, each user could blend coal from
2 or more origins to meet its sulfur dioxide emission standard.
368
-------
U + ™ V S - «
iikm l™i ijkm 1 k
where ^ - units of sulfur dioxide contained in one unit
of raw coal from origin i, 0^ - units of sulfur dioxide con-
tained in one unit of clean coal from origin i, S^ « maximum
allowable sulfur dioxide emissions at user k, and * ^ »
maximum allowable emission standard for user k measured as
units of sulfur dioxide per unit of heating value. Additional
nonnegativity and integer restrictions were:
<6> Mi' Uikm» Vijkm' 1 0; YJ - 0 or YJ - 1; and, Xfe - 0
or Xk -1.
Data
Data on 1975 and projected 1980 coal consumption in Iowa were
obtained from electric generating utilities and industrial firms
using coal-fired boilers. Nearly 131 trillion Btu's from coal
were consumed in 1975. The projected 1980 coal consumption
is 299 trillion Btu's. Sulfur dioxide emission standards applicable
to each coal user location were obtained from federal, state and
county agencies with pollution control authority (Linn County,
1975, Polk County Board of Health, 1972, State of Iowa, 1976,
and U.S. Environmental Protection Agency, 1971).
Data on the sources of coal consumed in Iowa in 1976 and
discussions with an advisory committee of executives from electric
utility companies and coal brokerage firms were the basis for
selecting out-of-state coal supply origins to be included in this
study. The seven out-of-state origins include Gillette and
Sheridan, Wyoming; Sparta, Canton, and West Harrisburg, Illinois;
Nortonville, Kentucky; and Unionville, Missouri.
369
-------
FOB coal prices and sulfur and Btu content for coal from
these out-of-state origins were obtained from bonded coal bids
submitted by coal brokers to electric generating plants from
mid-1976 to early-1977 and from discussions with the advisory
committee.
Price and quality data for 2 underground mines and for 5
strip mines currently operating in Iowa were obtained from
municipal electric utilities. Based upon the data on these
2 underground and 5 strip mines in Iowa, FOB prices for the
24 potential mine sites were estimated by the following equation
(Libbin and Boehlje, 1977 and Nagarvala, Ferrell and Oliver, 1976):
(7) P = aS8
where P - estimated price, S = sulfur content in percent of
weight, a = constant, and g - regression coefficient.
The resulting price-sulfur relationship for Iowa strip mine coal
is:
(8) P = $21.12S"°'29, R2 = 0.63.
These 1977 FOB coal prices do not include additional mining
costs resulting from The Surface Mining Control and Reclamation Act
of 1977 (U.S. Congress, 1977). Estimates of additional mining costs
resulting from this act were obtained from executives of coal
mining companies. These estimates, added to the estimated 1977
FOB mine prices in Table 1.
Two additional sets of FOB mine coal prices were generated
for the 24 potential strip mine sites in Iowa and for Northern
Missouri strip mine coal. Both additional sets of prices are
based on the assumption held by Iowa coal mine and utility execu-
tives, that 1977 FOB Iowa mine prices would not allow for the
370
-------
recovery of the total cost of opening and operating new mines.
The first additional set of FOB strip mine prices presented in
Table 3 is based on the estimated average 1977 cost of opening,
operating, and reclaiming a new 70,000 ton-per-year mine with an
average 50-foot highwall and a 30-inch seam. This cost was
estimated to be $17.33 per ton (Baumel, Drinka, and Miller, 1978).
The difference between $17.33 and $13.48--the estimated average
Iowa 1977 FOB mine price--was added to each estimated price obtained
from Equation 8 to approximate the 1977 FOB mine price to open,
operate and reclaim a new Iowa strip mine under average mining
conditions at each of the 24 potential strip mine sites in Iowa.
The second additional set of FOB strip mine prices presented in
Table 3 is based on the estimated cost of opening, operating and
reclaiming a new Iowa strip mine under high cost mining conditions.
This cost was estimated to be $20.06 per ton (Baumel, Drinka and
Miller, 1978). The difference between $20.06 and $13.48 was added
to each estimated price obtained from Equation 8 to approximate
the 1977 FOB mine price to open, operate and reclaim a new Iowa
strip mine under a high-cost mining operation at each of the
24 potential strip mine sites in Iowa.
These two levels of higher FOB prices were also applied to
Missouri strip mine coal, because the characteristics of Northern
Missouri coal are similar to those of Iowa coal. Because the
scale of operations is larger at Missouri mines than at Iowa
mines, an estimated $1.00 per ton cost savings was subtracted from
the Iowa price adjustments.
Coal beneficiation plant performance data and investment and
operating costs were obtained from a "package" beneficiation plant
371
-------
proposed for construction in Iowa and on performance data and costs
from the experimental coal beneficiation plant at Iowa State
University (Grieve, Chu and Fisher, 1976).
The "package" beneficiation plant would process 840,000 tons
of raw coal per year. The beneficiation process is estimated to
yield 77 percent clean coal and 23 percent refuse resulting in
646,800 tons of beneficiated coal per year. The process removes
about 35 percent of the sulfur and increases the Btu content of
the beneficiated coal by about 12 percent (Grieve and Fisher, 1978).
The total investment cost in 1977 dollars is estimated to be
$2,588,000 (Eldridge, 1977). At a 10 percent interest rate, the
annual interest and capital recovery is estimated to be $326,413.
Other fixed annual costs including management, insurance, taxes,
etc., are estimated to be $350,444 per year. The variable cost of
operating the plant is estimated to be $0.819 per ton (Eldridge. 1977)
The 23 percent refuse from the beneficiation process must be
returned to the mines for disposal. To minimize the distances
that refuse must be hauled to the mines, the 8 potential bene-
ficiation sites were restricted to the 3%-county producing area.
It was assumed that each site located on rail lines would need
5,800 feet of rail siding. The potential sites currently have
from 0 to 3,360 feet of rail siding. The annualized cost of
the additional siding was added to the annual investment cost
at each location.
Two sets of rail rates were used in the analysis. The first
set includes the actual rates on which coal moved from each
out-of-atate origin selected in this study to each Iowa coal
user during the period from January 7 to November 30, 1977. The
372
-------
rail rates in effect during this period of time are referred to
as the Ex Parte 336 rate level and were primarily for single-car
rail shipments. Only a few coal users had access to multiple-
car or unit-train shipments from the selected coal origins during
this period.
The second set of rates—referred to as estimated multiple-car
rates—includes all the Ex Parte 336 rates as well as estimated
L5-car, 50-car and 100-car rates for users who did not have access
to these shipment sizes in 1977. The estimated 15-car, 50-car, and
100-car rates were obtained from a computer program designed to
estimate variable rail costs. These estimated variable costs
were converted to estimated rates by multiplying the estimated
variable cost by a ratio consisting of published Ex Parte 336
rates for the same size shipments to different destinations divided
by the estimated variable costs to those destinations.
Trucks perform three coal-hauling functions in this analysis.
First, coal is hauled from strip mines to coal beneficiation plants
in tandem-axle dump trucks. Second, coal beneficiation refuse is
hauled for disposal from beneficiation plants to mines in tandem-
axle trucks. Third, trucks compete with rail and barge in hauling
beneficiated coal to utility and industrial users. Tandem-axle
dump trucks pulling pup trailers currently haul most of the coal
from Iowa mines to coal users. The costs of hauling coal for
each of these movements was estimated using mid-1977 cost levels
(Eldridge, 1977). The cost function for hauling coal from the
mines to beneficiation plants and for hauling refuse from the
plants to the mines was Ct - $0.1743 + $0.0578 m where Ct - cost
373
-------
per ton and m - loaded miles. Refuse was not permitted to be
a backhaul because of the difficulty of cleaning the refuse sludge
from the truck after each load. The cost functions for hauling
coal from the mine to users or from the beneficiation plant to
users are:
Loaded Miles Cost Function
0-20 Ct » $0.3668 + $0.0414 m
20.1 - 75 Ct - 0.3711 + 0.0411 m
75.1 - 200 Ct - 0.7439 + 0.0360 m
Assuming a 15 percent profit margin, trucking rate functions were
estimated from the trucking cost estimates by multiplying each
trucking cost function by 1.15.
Data on the cost of combined rail-barge movements from Sparta
and West Harrisburg, Illinois and from Nortonville, Kentucky
to Iowa destinations on the Mississippi River were obtained from
coal mining and barge companies.
Data on the 1977 rail receiving capacity were obtained from
each utility and industrial user. Estimates were made of the cost
of upgrading the rail receiving capacity of each coal user to the
next larger size of shipment. If the facility could receive
100-car unit trains, no additional investment in rail receiving
capacity was permitted. If the projected number of tons of coal
to be used in 1980 would provide less than one shipment per month
at the next size shipment, or if the user historically received
all of its coal by truck or barge, the user was not given the
opportunity to increase its rail receiving capacity. The variable
costs of receiving, unloading and transferring the coal to a live
storage area by mode and size of shipment were obtained
37A
-------
from utility company executives.
Findings.
Five computer solutions are presented in Table 2. Under
Ex Parte 336 rail rates, the amount of raw strip mine coal
produced in Iowa in 1980 would vary from a high 3,290,000 tons
at 1977 FOB mine prices to 600,000 tons at a price of $20.06
per ton FOB Iowa strip mine price. The number of coal bene-
ficiation plants would vary from one to five depending on the
FOB strip mine coal price and the type of rail rates. The
analysis of coal mining costs suggests that the most realistic
average FOB Iowa strip mine coal price is $17,33 per ton.
Solutions I and II are based on this price.
The estimated tons of coal consumed in Iowa in 1980 by coal
origin under Solution I and II are presented in Table 3. Under
Solution I, nearly 60 percent of the coal would be supplied from
Wyoming, up from-about 40 percent of Iowa's 1976 coal consumption;
Illinois would supply 30 percent, down from about 36 percent in
1976 (U.S. Department of the Interior, 1976). The remainder--
about 1.6 million tons—would come from Iowa sources. Under
Solution I, slightly over 300,000 tons would move directly from
underground Iowa mines to users. Two beneficiation plants would
require 1,680,000 tons of raw strip mine coal to produce 1,293,600
tons of cleaned coal. In 1975, total Iowa strip mine coal
production was only 259,000 tons (U.S. Department of the Interior,
1977). Thus, under the assumptions of this solution, strip mine
coal production would increase about 550 percent over* 1975
production.
375
-------
01
Solution
I
II
III
IV
vb
Iowa strip mine
Price per ton
$17.33
$17.33
$20.06
$20.06
1977 FOB Mine Prices
Rail rates
Ex Parte 336
Estimated multiple-car
Ex Parte 336
Estimated multiple-car
Ex Parte 336
Tons of raw
Iowa strip mine
coal produced
1,680,000
840,000
600,000
0
3,290,000
Number of coal
beneficlation
Plants
2
1
1
0
5
Estimated total
cost of 1980
coal consumption
$335,675,000
328,000,000
338,830,000
329,725,000
329,475,000
.Iowa Underground mines produce 307,290 tons of coal in all solutions.
"Source: (Eldridge, 1977)
Table 2. Summary of five computer solutions.
-------
Solution I Solution II
Source of
coal
Wyoming
Illinois
Kentucky
Missouri
Iowa
Underground mine
Beneficiated strip mine
Total
Tons of
coal
9,477,160
4,856,980
0
21,000
307,290
l,293,600a
15,956,030
Percentage
of total
59.4
30.5
0
0.1
1.9
8.1
100.0
Tons of
coal
11,096,220
4,419,560
0
0
307,290
646,800b
16,469,870
Percentage
of total
67.4
26.8
0
0
1.9
3.9
100.0
al,680,000 tons of raw coal are required to yield 1,293,600 tons of
beneficiated coal.
840,000 tons of raw coal are required to yield 646,800 tons of beneficiated coal.
Table 3. Estimated quantitites of coal consumed in Iowa by source of coal under
solutions I and II, 1980.
-------
Given that Iowa coal production has continued to decline
during a period of increased coal consumption, and given the
assertion of electric utility and mining executives that raw
Iowa strip mine coal consumption will continue to decline, one
can conclude that the estimated 550 percent increase in Iowa
strip mine coal production under Solution I over 1975 production
levels can be attributed to coal beneficiation plants.
Solution II is based on the assumption that multiple-car
rates will be available for both Iowa and out-of-state coal by 1980
to all but 13 Iowa coal users. The multiple-car rates were not
made available to these 13 users because of relatively low pro-
jected 1980 coal consumption, or because the coal user has his-
torically received coal only by barge. Therefore, Solutions I
and II provide an evaluation of the impact of improved trans-
portation on the production and marketability of Iowa coal.
Wyoming would supply 67 percent of the 1980 coal consumed
in Iowa under Solution II; this compares with almost 60 percent
under Solution I and 40 percent in 1976. Illinois would supply
about 27 percent in Solution II compared with 30 percent in
Solution I and 36 percent in 1976. No coal would be received
from either Kentucky or Missouri. Iowa mines would supply nearly
6 percent of total coal consumption under Solution II compared
with 10 percent under Solution I. This comparison shows that
the introduction of multiple-car rail rates would increase the
amount of Wyoming coal consumed in Iowa and at the same time de-
crease the consumption of Missouri and Iowa coal. The reason for
the large reduction of Iowa strip mine production under the mul-
tiple-car rate solution is that the estimated multiple-car rate
378
-------
reductions from single-car rates are much greater for Wyoming
and Illinois coal than the estimated rate reductions for the
short hauls from Iowa coal beneficiation plants to Iowa users.
The estimated rate reductions for 50-car trains from single-car
shipments from Wyoming and Illinois range up to $6.13 and $4.34
per ton, respectively, but only up to $2.46 per ton from Iowa
(Baumel, Drinka and Miller, 1978). Most of the Iowa rate
reductions are less than $1.50 per ton.
The amount of coal consumed by sulfur emission standard
under Solution I is shown in Table 4. All underground Iowa coal
would be consumed by boilers with the 8-pounds of S02 per
million Btu emission standard. About 50 percent of the bene-
ficiated coal would be consumed by boilers with the 8-pound
standard; about 11 percent would be consumed by users with the
6-pound standard, and about 39 percent by users with the 5-pound
standard. Boilers with the 1.2-pound emission standard would
consume Wyoming coal exclusively.
Table 5 shows the amount of Iowa and out-of-state coal
shipped by mode under Solutions I and II. Under Solution I,
nearly 89 percent of"Iowa coal would be transported by truck and
11 percent by rail; in 1976, 89 percent of the Iowa coal shipments
were transported by truck (U.S. Dept. of the Interior, 1977).
Under Solution I, more than 62 percent of the out-of-state coal
consumed in Iowa would be shipped by unit trains, 14 percent by
rail-barge, and less than 1 percent by truck. Under Solution
II almost 32 percent of the Iowa coal would be shipped in 15-car
shipments, and 14 percent would move in single-car rail shipments.
379
-------
u>
00
0
Assumed maximum SO.
emission standard in
pounds per million Btu
1.2
5
6
8
Total
Iowa coal
Beneficiated
Tons
0
498,710
148,090
646,800
1,293,600
Percentage
0.0
38.6
11.4
50.0
100.0
Underground
Tons
0
0
0
307,290
307,290
Percentage
0.0
0.0
0.0
100.0
100.0
Out-of-state coal
Tons
9,219,050
953,220
3,626,350
556,520
14,355,140
Percentage
64.2
6.6
25.3
3.9
100.0
Table 4. Estimated coal consumption by Iowa users by S0? emission standard and coal origin under
solution I, 1980.
-------
CO
oo
Mode of
transport
Truck
Rail
Single-car
15-car
50-car
100-car
Barge
Total
Iowa coal
1,419,490
181,400
0
0
0
0
1,600,890
Solution I
Out-of -state coal
21,000
3,340,730
0
376,540
8,545,580
2,071,290
14,355,140
Iowa coal
519,730
133,280
301,080
0
0
0
954,090
Solution II
Out-of-state coal
0
57,950
1,920,500
2,543,920
8,922,120
2,071,290
15,515,780
Table 5. Estimated amount of coal transported to Iowa users by mode and coal origin
under solutions I and II, 1980, In tons.
-------
The introduction of multiple-car rail rates would shift some
Iowa coal from truck to rail. The greatest impact, however,
would be to increase the amount of Wyoming coal consumed
in Iowa and to reduce the amount of coal consumed from all
other sources.
Most of the coal trucked from Iowa strip mines to benefi-
ciation plants would come from mines within 10 miles of the
plant. The maximum distance that coal would be trucked from a
mine to a cleaning plant in this solution is about 25 miles.
The estimated total cost (Table 2) of the delivered 1980
projected coal consumption is $335,675,000 under Solution I.
The estimated total cost would fall to approximately $328,000,000
under Solution II. Thus, while the multiple-car rates would
reduce the Iowa coal share of total 1980 coal consumption, the
multiple-car rates would reduce the total cost of this consump-
tion by about $7,675,000.
If Iowa and Missouri coal prices were to increase to an
average of about $20 per ton, the level of Iowa high sulfur
strip mine production would decline to 600,000 tons at the
Ex Parte 336 rate level and to zero tons under the multiple-car
rates (Table 2). At these FOB mine prices, the total delivered
cost of the coal consumed in Iowa would decline $9.1 million
per year under the multiple-car rates compared to the Ex Parte
336 rate levels.
Conclusions
1. Coal beneficiation plants would reverse the downward
trend in the production of Iowa strip mine coal except when very
high FOB Iowa coal prices are combined with multiple-car rail
382
-------
rates for most Iowa coal users, The estimated amount of strip
mine Iowa coal to be produced in 1980 would vary between 0 and
3,290,000 tons depending upon the assumed level of FOB Iowa mine
coal prices, and the level of rail rates.
2. The analysis of Iowa coal mining costs ($17.33 per ton
for typical mining operations) suggests that the most likely
range of 1980 raw strip mine coal production in Iowa--assuming
Ex Parte 336 rail rates--would be about 1,680,000 tons of coal
per year by 1980. This would make Iowa strip mine coal production
about 550 percent greater in 1980 than in 1975. This level of
strip mine coal production would require two coal beneficiation
plants.
3. The largest market for beneficiated Iowa coal is at
coal users with the 8-pound per million Btu S02 emission
standard. However, up to about 50 percent of the cleaned coal would
be consumed by users with 5- and 6-pound SO^ emission standards.
4. Almost all the Iowa coal would be transported to the
central and east-central Iowa coal users by truck. Typically,
trucks have a cost advantage over single-car rail rates up to
approximately 140 miles. Beyond that point, single-car rail rates
are cheaper than estimated truck rates.
5. Reduced transportation rates on multiple-car rail shipments
would reduce, rather than increase, Iowa coal production. The
amount of strip mine coal produced in Iowa in 1980 would vary from
zero tons to 840,000 tons under the estimated multiple-car rates
solutions, depending upon the assumed level of Iowa coal prices.
The reason for the reduction of Iowa strip mine production under
the estimated multiple-car rates is that the estimated Wyoming
383
-------
and Illinois multiple-car rate reductions are much larger than
the rate reductions for the short hauls from Iowa coal beneficiation
plants to Iowa users.
6. The estimated multiple-car rates would reduce the total
cost of supplying the 1980 coal requirements in Iowa by $7.7 to
$9.1 million depending upon the assumed level of Iowa coal prices.
This creates the following policy dilemma: Should lower-cost
multiple-car and unit-train rail rates from out-of-state coal
origins be discouraged to increase Iowa coal production, or should
multiple-car and unit-train rail rates be encouraged to reduce
the total cost of supplying Iowa's coal requirements?
7. Coal beneficiation plants would increase the marketability
of Iowa strip mine coal beyond the 1976 production level under all
solutions except the multiple-car rate solution with an average
Iowa FOB strip mine price of $20.06; If the Iowa coal industry
can produce strip mine coal for around $17.33 per ton, and if
coal users do not begin to jointly obtain large volumes of
Wyoming coal at low FOB prices shipped in unit trains, up to two
coal beneficiation plants would significantly increase the
marketability of Iowa strip mine coal.
384
-------
References
Avcin, J. Estimated quantity and quality of Iowa coal reserves
by county. Unpublished research, Iowa Geological Survey,
Iowa City, Iowa. 1976.
Baumel, C. P., T. P. Drinka, and J. J. Miller. Economics
of alternative coal transportation and distribution systems
in Iowa. Iowa State University Agricultural Experiment
Station,Special Report No. 81, Ames, Iowa. 1978 (in press).
Eldridge, C. L. The potential for improved transportation of raw and
beneficiated coal in Iowa. Unpublished M.S. thesis, Iowa
State University, Ames, Iowa. November, 1977.
Grieve, R. A., H. Chu, and R. W. Fisher. Iowa coal project-
preliminary coal beneficiation cost study progress report.
Unpublished report, Iowa State University Coal Refining
Plant, Ames, Iowa. September 23, 1976.
Grieve, R. A. and R. W. Fisher. Full scale coal preparation research
on high sulfur Iowa coal. IS-ICP-53, Iowa State University,
Ames, Iowa. February 1978.
Lemish, J. and'L. V. A. Sendlein. Personal communication: Information
on potential number of coal strip mines in townships of a
3%-county area in southeast Iowa. Department of Earth
Science, Iowa State University, Ames, Iowa. January, 1977.
Libbin, J. D. and M. D. Boehlje. Interregional structure of the
U.S. coal economy. American Journal of Agricultural
Economics, Vol. 59, No. 3, August 1977.
Linn County. Regulation number 1-72, Air pollution. Cedar Rapids,
Iowa. Effective January 1, 1975.
385
-------
Nagarvala, P. J., G. C. Ferrell and L. A. Oliver. Regional
energy system for the planning and optimization of national
scenarios; final report, clean coal energy: source-to-use
economics project. Prepared for the U.S. Energy Research and
Development Administration, Washington, D.C. by Bechtel
Corporation. June 1976.
Polk County Board of Health. Rules and regulations, chapter 5,
Air pollution control, article 9, division 2, section 5-27(a).
Des Moines, Iowa. Effective November 3, 1972.
State of Iowa. Iowa administrative code, section 400-4.3(3)a
(Sections 1, 2, 3, and 4). Des Moines, Iowa. Effective
July 19, 1976.
U.S. Congress, Public law 95-87, 95th Congress, 91 Stat. 445.
August 3, 1977.
U.S. Department of the Interior, Bureau of Mines. Bituminous
coal and lignite distribution, calendar year 1971. Washington,
D.C.
U.S. Department of the Interior, Bureau of Mines. Bituminous
coal and lignite distribution, calendar year 1976. Washington,
D.C.
U.S. Department of the Interior, Bureau of Mines. Coal—bituminous
and lignite in 1975. Washington, D.C. February 10, 1977.
U.S. Environmental Protection Agency. Standards of performance
for fossil-fuel fired steam generators. Federal Register,
Subport D, Vol. 36, No. 247. Washington, D.C. December 23,
1971.
386
-------
AN EVALUATION OF THE DESULFURIZATION POTENTIAL OF U.S. COALS
Jane H. McCreery and Frederick K. Goodman
Battelle
Columbus Laboratories
505 King Avenue
Columbus, Ohio 43201
ABSTRACT
We have derived a generalized approach to the evaluation of the desul-
furization potential of coals. It is applied here to various cleaning processes
for the coal reserves in each of the six major coal-producing regions in the
U.S. and for the coal reserve base for the U.S. as a whole. The methodology
characterizes the entire U.S. reserve base via 36,000 composite coal analyses
showing total weight, percent ash, percent sulfur, and Btu content. In
addition, each reserve record is associated with one float-sink analysis
as reported in RI 8118. The mathematical approach adopted allows the charac-
teristics of the cleaned coal to be obtained from those of the raw coal by
scaling the raw coal characteristics by factors dependent on the cleaning
process involved and the washability analysis of the raw coal. By reducing
all cleaning processes to this same general form of multiplying factors,
the data manipulation for all the cleaning processes under consideration
can be carried out simultaneously, thereby substantially reducing computer
costs. The approach is valid for general cleaning processes and is independent
of the specific performance measures to be used for the processes.
387
-------
1.0 INTRODUCTION
The purpose of the work described in this paper was to develop a com-
puter based methodology for the evaluation of the desulfurization potential
of U.S. coal reserves. The technique used had to allow for the subclassi-
fication of coal reserves by type, deep or strip, and by geographic sub-
area, region or state. The types of desulfurization processes to be
considered included single and multi-stream physical cleaning, chemical
cleaning, and combined physical and chemical cleaning. The actual
evaluation involved measuring values such as the following:
• The weight and Btu recovery percents which could be
achieved as a function of the required level of Ibs
of S02/MM Btu.
• The actual tons of coal or Btus which could be obtained
as a function of the required level of Ibs of SO-/MM Btu.
• The percent weights and Btus available versus the level
of flue gas desulfurization required to meet some proposed
removal NSPS.
• The actual tons of coal or Btus which could be obtained
as a function of the level of the percent removal standard
to be required.
Finally, the technique selected was to take into account the large vari-
ability associated with the characteristics of different coals.
The large number of coal reserve resources and the large number of
sample analyses associated with each resource that had to be processed
in order to take variability into account introduced into the problem
388
-------
the difficulties associated with large-scale data handling. In particular,
computer costs tend to escalate rapidly when large amounts of data pro-
cessing have to be done. An additional constraint here was that the final
computer programs produced were to be easily transferable from computer
to computer; thus, the data processing techniques used had to be restricted
to those very simple ones which are universally available.
The method used for computing the desulfurization potential of the
reserve base meets all of the above criteria. It significantly reduces
the computational effort required by other approaches. It also has the
additional advantage that much of the actual computation performed is
independent of the specific performance measures to be used for the
evaluation. Thus, multiple performance measures can be computed without
repeating the entire calculation. In essence, the approach taken allows
the composition of coal cleaned by a variety of cleaning processes to be
computed directly from the composition of the raw coal simply by scaling
by appropriate factors which are dependent on the washability analysis
of the coal and the specifications of the cleaning processes. These
scale factors, whose computation can be lengthy, need be computed only
once for each cleaning process and washability analysis pair. Once
calculated, however, each factor is used many times via a simple multi-
plication for every sample analysis corresponding to the same reserve
resource and for every reserve corresponding to the same washability
analysis. This effects considerable saving in the computing time re-
quired .
389
-------
2.0 OVERLAYING THE DATA
The data available for the evaluation was as follows:
• 587 sets of washability analyses for coal from sample
mines in the U.S. as reported in Cavallaro, Johnston,
and Deurbrouck (1976)—i.e. RI8118.
• The reserve base of U.S. coal, consisting of 3167 re-
cords specifying the weight of the resource for both
strip and underground coal, together with the maximum,
minimum, and mean levels of the major constituents of
the coal in that resource. This data is consistent
with that summarized in Thomson and York (1975) and Ham-
ilton, White and Matson (1975)—i.e. IC8680 and IC8693.
• Approximately 50,000 detailed sample coal analyses taken
from the coal data base of the U.S. Bureau of Mines in
Denver, Colorado. This data includes the composition of
each sample in terms of its ash, sulfur, and heat content.
Given these three sets of data as a starting point, the first step in
the analysis was to overlay them into a single data base which contained
36,000 coal resource records and which had the following information for
each:
• the location in terms of its region, state, county, and bed
• The weight in tons of both strip and underground coal
• The mean percent by weight of ash, organic sulfur, and pyritic
sulfur
390
-------
• The mean heat content expressed in Btu/lb
• The float-sink distribution of the coal characteristics
The coal reserve resources and the washability data of RI8118 are each
specified by state, bed and county; however there is not' an exact corre-
spondence between reserves and washability data since for many of the
reserves there are no washability data. In order to be able to deter-
mine the desulfurization by physical'cleaning processes of coal resources
having no washability data, the reserve resources were assigned wash-
ability data in the following manner:
1. If one or more state, bed and county matches are found
between a given reserve and the washability data, the
reserve is assigned that washability data which has coal
composition closest (in the least squares sense) to the
composition of the reserve. If no composition data is
given for that reserve resource, the resource is sub-
divided into as many parts as there are matching
washability data and each is assigned one of the
washability analyses.
2. If there are no state, bed and county matches between
a given reserve resource and the washability data,
look for state, bed and region matches. Assign the
reserve the matching washability data as in 1.
3. If no matches occur in either 1 or 2, look for state
and county matches. Assign the reserve the matching
washability data as in 1.
391
-------
4. If no matches occur in 1, 2, or 3 assign the reserve
the washability data from other beds in the same state
and region as in 1.
5. For some states there are no washability analyses at
all; reserve resources in those states are assigned
washability data from other states in the same region
as follows:
North Carolina is assigned washability data from Virginia
Michigan is assigned washability data from all states in the
Eastern Midwest region
Texas is assigned washability data from Oklahoma
South Dakota is assigned washability data from North Dakota
Idaho \
Oregon
Washington
are assigned washability data from
Montana and Wyoming
The washability data of the relevant state or states is assigned
to the resource as in 1.
Thus all the coal reserve resources were assigned washability data.
The analytical data file consists of approximately 50,000 records each
of which gives coal composition data for a reserve resource sample. This
sample analysis data was overlaid with the reserve base to obtain coal
composition data for each reserve resource. Each resource has several
sample analyses corresponding to it and, in the absence of any method
of assigning weights to the different analyses for the same resource, all
were weighted equally. The variation in the samples for a given resource
was taken into account by dividing all the coal in that reserve resource
392
-------
into as many parts as there are corresponding sample analyses and each part
was assigned the composition of one of the samples. For those reserves
that have composition data given on the reserves file and on the analysis
file it was assumed that the mean of all the sample analyses should be
equal to the composition data given on the reserves tape; if necessary
the sample analysis data was scaled to make this so. Reserves having no
composition data given on the reserves file were assigned the coal composi-
tion given by the RI8118 washability data. Reserves having composition
data given on the reserves file but no sample analysis used the coal
composition given on the reserves file.
By overlaying the coal reserves file and the analysis file in this
manner an expanded reserves file of approximately 36,000 records was ob-
tained, each record consisting of resource identification (by state, bed
and county), weight of coal for both strip and underground, and the compo-
sition of the coal. 36,000 records are obtained and not 50,000 as on the
original analysis file becaus'e a number of the sample analyses either do
not correspond to any of the reserve resources or correspond to a given
resource which shows no coal available in both strip and underground re-
serve. For a given state, bed and county group there will be several re-
cords on the file each having the same weight of reserves (such that the
total adds up to the actual weight in the respurce) but having possibly
different composition data corresponding to the different sample analyses
for that resource. The sulfur content of the coal is given in the coal
reserves file and in the analysis file only as total sulfur content; this
was divided into pyritic and organic sulfur in the ratio in which these
393
-------
two occur in the washability data that corresponds to that resource.
The extensive data manipulation that is necessary to obtain the over-
layed reserves and analytical file is independent of the cleaning process
to be considered and is dependent only on the original three data bases
of coal washability, reserves and sample analyses. The overlaid file has
therefore been created only once but has been used in many subsequent anal-
yses.
3.0 IMPLEMENTING THE CLEANING PROCESSES
To implement the effect of the cleaning processes on the reserve
resources, use has been made of the fact that a single washability anal-
ysis corresponds to many records on the overlaid reserves data file. By
doing so the computational time required to perform this part of the anal-
ysis has been reduced by a factor of approximately sixty. The methodology
developed can treat any cleaning process that is of one of the following
specific types.
1. A physical cleaning process.
2. A chemical cleaning process that removes specified
percentages of the characteristics of the raw coal
(ash, pyritic sulfur, organic sulfur).
3. A chemical cleaning process that reduces the levels of
the characteristics to given threshold values.
4. Combinations of 1 and 3 or combinations of 2 and 3.
5. A blend of the product coal from two of the above
processes.
394
-------
6. One of processes 1-4 on the coal product of another
of processes 1-4.
Reductions In the weight and Btu/lb of the coal by given percentages can
be specified directly for processes of types 2 and 3 and for processes of
type 1 as operating penalties over and above the reductions caused by the
physical separation process. Physical cleaning processes are restricted
by the RI8118 washability data to size fractions of 1-1/2 inches, 3/8 inch
or 14 Mesh, and to specific gravity fractions of Float -1.3, 1.3-1.4, 1.4-
1.6 or the sink from 1.6.
A physical cleaning process can be completely specified by the size
fraction to which the coal is crushed before separation plus the following
quantities for each of the four specific gravity fractions:
• The percent ash removed from the specific gravity fraction.
• The percent pyritic sulfur removed.
• The percent organic sulfur removed.
• The percent Btu/lb recovery for the specific gravity fraction.
• The percent weight recovery (=0.0 if this specific gravity
fraction is discarded).
These quantities are in addition to the amount of each characteristic that
is removed by the physical separation process. A cleaning process of type
2 can be expressed in terms of the above five quantities alone. A cleaning
process of type 3 can be expressed in terms of the above quantities together
with threshold values for those characteristics that are reduced to threshold
levels,
395
-------
Given such a specification of a cleaning process of type 1 or 2 and
the file of the RI8118 washability data, it is possible to construct an
array T(i,j,k) which fully characterizes the cleaning of coal from a partic-
ular state, bed and county group by the cleaning process. Here i corresponds
to the index of the washability data (determined from the state, bed and
county group), j corresponds to the cleaning process under consideration,
and k corresponds to the characteristics of the coal that are subject to
change by cleaning (weight, ash, pyritic sulfur, organic sulfur, and Btu/
Ib). On cleaning by process j a sample of raw coal having state, bed and
county group corresponding to washability index i and characteristics R(k),
one obtains cleaned coal having characteristics
C(k)=R(k). T(i,j,k).
Thus the effect of a cleaning process on coal of a given washability is ob-
tained simply by scaling the characteristics of the coal by the relevant
factors from the T array. Chemical cleaning which reduces characteristics
^i
to threshold values (type 3 processes) can be simulated by reducing the
relevant characteristics after scaling by the T factors.
The array T(i,j,k) is computed as follows. For a type 2 cleaning
process j the specification of the process described above completely
determines the T matrix. The process specification gives the proportion
D(k) of characteristic k of the feed coal that appears in the cleaned coal.
If
k=l corresponds to weight
k=2 corresponds to ash content
k=3 corresponds to pyritic sulfur content
k»4 corresponds to organic sulfur content
k»5 corresponds to Btu/lb for the coal
396
-------
then
T(i,J,k)-D(k)/D(l), k-2,3,4,5.
This is independent of the washability index i.
For a type 1 process the proportion P(fc,k) of the feed coal in specific
gravity fraction A and having characteristic k that appears in the cleaned
coal is given by the washability data for the feed coal. Any additional
reduction in the levels of the characteristics is given by the process spec-
ification and can be expressed as D(£,k). Combining these two, the propor-
tion of the feed coal appearing in the product is
E P(£,k), DU,k)
a
where the summation is over the four specific gravity fractions of the RI8118
washability data. Then
and
T(i,J,k>S P(A,k). DU,k)/T(i,j,l)> k-2,3,4,5.
A
Having constructed this T matrix from the specifications of the clean-
ing processes and the washability data, it is combined with the overlaid
reserves and analytical data file. The characteristics of the raw coal from
each of the 36,000 reserve resource records on the file are scaled by the
appropriate factors from the T matrix to obtain the characteristics of that
coal after cleaning by each of the processes. Any reduction in character-
istic values to threshold values for a type 3 process is done at this stage.
397
-------
A new file Is created consisting of 36,000 records as before but now each
record contains not just the reserve levels and characteristics of the
raw coal but those values also for the processed coal for each cleaning
process. This file is then used to assess the desulfurization potential
of the coal reserves.
4.0 AN EXAMPLE APPLICATION
As an example application of this methodology, Figures 1-3 show for
Northern Appalachian coal, Eastern Midwest coal and Western coal, the
percentage of the total regional resource Btus (from strip and underground
coals combined) that by cleaning can be made to meet a given percent sulfur
removal New Source Performance Standard. The abscissa of the curves is
the New Source Performance Standard, while the ordinate is the percent
Btus obtainable. An upper limit to the emissions level of the resultant
coal of 1.2 Ibs S02/10 Btu was assumed; above this level the processed
coal is unacceptable regardless of the percent sulfur removed in the process,
A lower limit of 0.5 Ibs SCL/IO Btu was used; raw coal below this level
need not be cleaned at all; if the processed coal lies below this level then
the amount of coal that has to be cleaned is just that amount so that the
combined raw and cleaned will reach 0.5 Ibs S02/m Btu. The cleaning pro-
cesses used here are as follows:
• Physical coal cleaning using 1-1/2 inch mesh at 1.6
specific gravity of separation.
• Physical cleaning using 3/8 inch mesh and separation
at 1.3 specific gravity. An operating energy penalty
of 1% was assumed in addition to the energy lost
directly through the separation process.
398
-------
• Meyers process: this is a chemical cleaning process
with a threshold of 0,2 percent pyritic sulfur. A
5% energy loss was assumed in the process together
with an operating penalty of 2% energy loss and a
weight loss of 10%.
• Gravichem process: crush coal to 14 mesh topsize and
separate at 1.3 specific gravity. Treat the sink with
Meyers process as above; combine float and processed
sink.
• A process with 95% pyritic sulfur removal and 20%
organic sulfur removal with a 10% energy loss and
an operating energy penalty of 2% and a weight loss
of 15%.
In all three of the regions shown the process that removes 95% of the
pyritic sulfur and 20% of the organic sulfur is the most effective. If
no percent sulfur removal standard is imposed but simply the 1.2 Ibs SO-/
10 Btu emission standard then using this cleaning process, 35% of the
Northern Appalachian coal meets the emission standard, 9% of the Eastern
Midwest coal and 75% of the Western coal. In the Eastern Midwest region
none of the other processes can clean more than 4% of the coal to meet
the emission standard. In the Northern Appalachian region the Meyers
chemical cleaning process and the Gravichem combined physical and chemical
process clean more of the coal to meet the emission standard than do either
of the purely physical cleaning processes. This ordering is reversed for
the Western region coal.
399
-------
Figures 4-6 show, for the same three regions the percentage of the
total regional resource Btus (from strip and underground coals combined)
that will meet a 90 percent sulfur removal standard if flue gas desulfuri-
zation (FGD) at specified levels is applied to the coal cleaned by the
above cleaning processes. The abscissa of the curves is the level of FGD
that is necessary to meet the NSFS with the cleaned coal. The ordinate
gives the percent of Btus obtained. The relative ordering, of effective-
ness of the cleaning process is the same as in the Figures 1-3 until
FGD cleaning at about 82% is applied. At this stage it is the energy
lost by the cleaning process that determines their order of effective-
ness. Beyond 90% FGD cleaning the percent Btus obtained is simply that
obtained by cleaning all of the coal in the region by the given process.
Figure 7 shows the percent weight of coal reserves (both strip and
underground) in the entire United States that can be cleaned to meet
specified emission standards. The dashed line indicates an emission
level of 1.2 Ibs S02/10 Btu. It is seen that approximately 41% of
the raw coal reserves by weight will meet an emission standard of 1.2 Ibs
S02/10 Btu without cleaning,
5.0 COMPARING THE RESULTS
The results produced using the techniques described here differ
dramatically from those produced by others — in particular RI8118 and
Foster Associates (1977). A value which is easy to obtain from all three
400
-------
sources, and one which reflects the difference very effectively, is that
percent of U.S. coal reserves which meet the 1.2 Ibs of SO./MM Btu emission
standard. As can be seen from Figure 7, our approach estimates it at 41%.
RI8118 on the other hand in their Figure 16 give an estimate of 12%. Foster
Associates (1977) do not give a value explicitly; however, it can be com-
puted directly from their schedule 25 as being equal to 27%. Our results
then for this summary value are significantly higher than those produced
by others. It should be observed that both of the other sources use parts
of the same data that we use; therefore, the differences cannot simply be
attributed to differences in the data.
The data used by RI8118 are of course the washability analyses which
are used by us as well. RI8118 makes no use, however, of any weight figures
associated with the reserves. Western coals are heavily under represented in
RI8118 in that only 40 of the 455 samples come from the Western region and
yet almost half of the U.S. coal reserves are in that region. If the values
for the individual regions as reported by RI8118 are weighted by the tons of
reserves in those regions when the overall U.S. average is computed, then the
RI8118 results predict a value of 38% for the percentage of U.S. coal reserves
which meet the 1.2 standard. -This calculation is shown in Table 1. Thus, the
bulk of the difference between our results and those of RI8118 can be accounted
for by the difference in the weighting assumptions.
The approach taken by Foster Associates (1977) is based on the IC8680-93
reserve data augmented by some additional data on heat content. They esti-
mate the distribution of Ibs of SO^/MM Btu for each state and coal type from
the distribution of sulfur for that state and coal and from the appropriate
401
-------
TABLE 1. U.S. COAL RESERVES WHICH MEET THE 1,2 EMISSION STANDARD AS REPORTED
IN RI8118 AND AS WEIGHTED BY LEVEL OF AVAILABLE RESERVES
Region
Northern Appalachian
Southern Appalachian
Alabama
Eastern Midwest
o
Western Midwest
Western
Total
(l>
No. of Samples
227
35
10
95
44
44
455
(2)
% Meeting
1.2 Standard
4
35
30
1
2.5
70
12
(3)
Samples Meeting
1.2 Standard
9
12
3
1
1
30
56
MM Tons
of Coal
68,274
34,907
2,982
89,029
18,992
203,776
417,959
Weight Meeting
1.2 Standard
(million tons)
2,731
12,217
895
890
475
142,643
159,851
vs>
Weighted Z
Meeting Standard
4
35
30
1
2.5
70
38
(l) Source RI8118.
(2) Number of samples times percent meeting standard divided by 100.
(3) Source the reserve base of U.S. coals.
(it) Percent meeting standard times MM tons of coal.
(5) MM tons meeting standard divided by MM tons of coal times iOO.
-------
average Btu content for that reserve. The particular value for the 1,2
standard is then taken from this derived distribution via interpolation.
As the authors, themselves, recognize this estimation of one distribution
from another is highly subject to error. They say the following in their
conclusions on page 89.
"Perhaps the most important conclusion to be drawn from this
exercise is that there are inherent difficulties in any at-
tempt to manipulate these reserve estimate distributions.
Even a slight change in assumptions can yield significantly
different and tentative results. As such, it appears that
any defensible statements or conclusions with respect to
the ability of coal reserves to comply with S09 emission
regulations must come from a redistribution of the individ-
ual coal analyses used in constructing the Bureau of Mines
distributions in IC8690 and IC8693."
From the standpoint of this presentation the critical simplification
in the Foster Associates approach is that all coal in a given state and of
a given type, regardless of its sulfur content, has the same Btu content.
This assumption is the primary cause of the lower value reported by them.
Btu content and percent sulfur tend to be negatively correlated. Based on
the 36,000 samples used in this study, the overall correlation coefficient
for the U.S. is -0.39. For a given coal, when the sulfur content goes down,
the Btu content tends to go up. For low sulfur coals, the Btu content is
generally higher than the mean; therefore, the use of the mean in the calcu-
lation of Ib S02/MM Btu tends to give a high estimate for low sulfur coals.
This point can be seen very clearly in Table 2 which shows four dis-
tributions of the total U.S. coal reserve base as classified by percent sul-
fur and by a compatible classification of Ib SO /MM Btu. The first two are
403
-------
TABLE 2. PERCENTAGE DISTRIBUTIONS OF TOTAL U.S, COAL
RESERVES CLASSIFIED BY PERCENT SULFUR AND
LB SO-/MM BTU RANGES
% Sulfur
range
3.0
Ib SO /
MM BCu
range
<0.75
0.76-1.08
1.09-1.42
1.43-1.75
1.76-2.42
2.43-3.08
3.09-3.75
3.76-4.42
4.43-5.00
>5.00
Foster Associates
% Sulfur SO /Btu
distribution distribution
19.4
13.0
11.9
7.1
7.2
4.9
4.2
4.2
3.3
24.0
6.4
15.5
15.6
9.1
7.8
6.8
5.9
3.5
5.0
24.4
Overlay File
% Sulfur S02/Btu
distribution distribution
21.4
15.2
8.6
6.6
7.4
3.6
3.4
3.2
2.9
27.8
22.4
15.8
7.7
6.2
6.4
5,1
3.7
3.7
2.7
26.3
-------
taken directly from Foster Associates (1977), while the second two have been
calculated from the overlaid reserve file whose construction was described
earlier in this paper. The percent of coal meeting the 1.2 standard can be
taken directly from the S02/Btu distributions by summing the first two values
plus one-third of the third value.
6.4 + 15.5 + (15.6/3) - 27.1
22.4 + 15.8 + (7.7/3) - 40.8
Note that the overlay file S02/Btu distribution is not calculated from the
overlay sulfur distribution; but rather, it is calculated by redistributing
the values for Ib SO-/MM Btu for the individual coal analyses as suggested
by the quote given earlier.
Now the two sulfur distributions are very similar. This is as expected
since ultimately the same data source was used for each. The S02/Btu distri-
butions are quite different, however, especially in the lower sulfur ranges.
This underestimation in the Foster Associates distribution of material in
the lower ranges is predictable from the negative correlation between sulfur
content and Btu content. Notice that from the overlay file analysis the
two distributions are quite similar. If one assumed that the Foster Assoc-
iates sulfur distribution was a good measure of the SC^/Btu distribution,
then the Foster Associates data would predict that 36 percent of U.S. coal
reserves meet the 1.2 standard. This calculation is as follows.
19.4 + 13.0 + (11.9/3) - 36.3
This value is much closer to ours.
In conclusion, it does appear that our estimate of 41 percent of the
405
-------
U.S. reserves meeting a standard of 1.2 Ibs SO./MM Btu is the best one to
date for this quantity based on the data currently available.
6.0 CONCLUSION
In conclusion, we have developed a methodology for the assessment of
the desulfurization potential of the entire U.S. coal reserve base. It is
independent of the specific assessment criterion used so that assessments
may be made using a variety of criteria without it being necessary to re-
peat much of the computations. As described here the coal reserves are
reported on a regional basis; it is a relatively simply procedure to adapt
the programs to consider the coal reserves on a state by state basis and
this work is currently in progress. The reliability of the results pro-
duced using this methodology depends entirely on the reliability of the
representation of the coal reserves by the analysis data and the wash-
ability data. Since there are several analysis records for each reserve
resource the variability of the coal composition within a given resource
is likely to be fairly well represented. However the distribution of the
sulfur content of the coal into organic and pyritic sulfur was taken from
the washability analyses and is therefore much more subject to error. Thus
the reliability of the results depends largely on the representation of the
reserve resources by the very limited number of washability analyses that
are available. Work is currently being undertaken to estimate the effect
that variation in the washability data might have on the desulfurization
potential of the reserve base.
406
-------
7.0 CONVERSIONS TO SI UNITS
Ton 907.18 kilograms
Btu 1054.35 joules
lb/10 Btu 430 nanograms/Joule
Inch 0.0254 meters
407
-------
8.0 REFERENCES
Cavallaro, J.A., M.T. Johnston and A.W. Deurbrouck. 1976. Sulfur
reduction potential of the coals of the United States. Bureau
of Mines Report of Investigation 8118.
Foster Associates, Inc. 1977. United States low sulfur coal reserves:
an assessment of alternative estimates, Volumes I and II.
Hamilton, P.A., D.H. White, Jr. and T.K. Matson. 1975. The reserve
base of U.S. coals by sulfur content, Part II: The Western
states. Bureau of Mines Information Circular 8693.
Thomson, R.D. and H.F. York. 1975. The reserve base of U.S. coals
by sulfur content, Part I: The Eastern states. Bureau of
Mines Information Circular 8680.
408
-------
Total Btus of raw coal = 1728 quack
100
80
60
4->
c
0)
o
11
en
u
40
20
X
X X X X
X X X X X X
4-
PCC 1-1/2 INCH, 1.6 S-G-
PCC 3/8 INCH, 1.3 S-G.
MEYERS PROCESS
GRAVICHEM PROCESS
.95 P.S-. 2 O.S- REMOVED
' + +
,
+ ^
,
,
X X
!
!
x
S a * I
', :
0
10
20 30 40 50 60 70
Percent Sulfur Removal Standard
80
90
100
Figure 1. Percent of Northern Appalachian coal that meets percent sulfur removal
standards with an emission limit of 1.2 Ibs S02/10 Btu.
-------
100
Total Btus of raw coal = 1999 quads
80
+ PCC 1-1/2 INCH. 1-6 S.G.
X PCC 3/8 INCH. 1.3 S.G.
O MEYERS PROCESS
* GRAVICHEM PROCESS
X .95 P.S., .2 O.S. REMOVED
4-1
s
o
60
40
20
XXXXXXXXXXXXXXx
x x x x
M M M
10
20 30 40 50 60
Percent Sulfur Removal Standard
70
8Q
90
100
Figure 2. Percent of Eastern 'Midwest Region coal that meets percent sulfur
removal standards with an emission limit of 1.2 Ibs SO^/IO Btu.
-------
Total Btus of raw coal = 3662 quads
c
a)
o
CO
4-»
ffl
100
80
60
40
20
0
+ PCC 1-112 INCH, 1.6 S.G.
X PCC 3/8 INCH, 1.3 S.G.
O BEYERS PROCESS
X X * X xCvEAVICHEM PROCESS
X .95 P.S-, 20.S REMOVED
ux x
10
XXX
-
XXX
x x x x x x x
X X X
x
'
0
xxxxxxxxxxxxxxxxxxx*
SS'fff^'t' + 'ff/f.^
00000000 l^*******^*****
J 1 1 I L
J I I I I
J L
20
30 40 50 60
Percent Sulfur Removal Standard
70
80
90
100
Figure 3. Percent of Western region coal that meets percent sulfur
removal standards with an emission limit of 1.2 Ibs SO-/10 Btu.
-------
100
Total Btus of raw coal = 1728 quads
80
+ FCC 1-1/2 INCH. 1.6 S.G.
X PCC 3/8 INCH, i -3 S-G.
O MEYERS PROCESS
.95 P.S
2 0.
REMOVED
X
x x
60
§
o
en
p
03
X
+
x x x x x >:
20
x +
X
X
X
x x
o i-»-
10
20
30
70
80
90
40 50 60
Percent FGD Removal Required
Figure 4. Percent of Northern Appalachian coal that meets a 90 percent
sulfur removal standard if FGD at a given level is applied to the cleaned coal.
100
-------
100
Total Btus of raw coal = 1999 quads
80
+ PCC 1-1/2 INCH, 1.6 S.G.
X PCC 3/8 INCH, i .3 S-G-
O MEYERS PROCESS
* GRAVICHEM PROCESS
X - 95 P.S- , 2 O.S- REMOVED
X
t t $
O O
60
c
0)
u
0)
ex
OT
3
40
I
X X X X X
X
X
20
I
X
« 1 I • • • A • • . * , A , » , *
O X
X
10
20
70
80
90
30 40 50 60
Percent FGD Removal Required
Figure 5. Percent of Eastern Midwest region coal that meets a 90 percent sulfur removal
standard if FGD at a given level is applied to the cleaned coal.
100
-------
100
Total Btus of raw coal = 3662 quads
80
+ PCC 1-1/2 INCH. 1.6 S-G.
X PCC 3/8 INCH, 1.3 S-G-
$ MEYERS PROCESS
+ GRAVICHEM PROCESS
X .95 P.S-, 2 0.S- REMOVED
+ o o o o
x x x x x ::
60
4J
c
0)
x
x x x x
3
40
20
x
•AiAiAMtitttitigiiiitititto*
j_
j_
20
30
70
80
9.0
40 50 60
Percent FGD Removal Required
Figure 6. Percent of Western region coal that meets a 90 percent sulfur removal
standard if FGD at a given level is applied to the cleaned coal.
100
-------
Total weight of raw coal = 417 billion tons
100
s
o
t-l
-H
I
RAW COAL
PCC I - )/ 2 INCH, I . t S6
PCC 3/81WCH, 1.4 OB 1 . 3 b 6
MEYERS PROCESS
BRAVJCHEd PROCESS
0.95 PV.S, 0.20 OR6.S
'BEST' FOR EACH RESERVE
0 .
0. 0
1.0 7.0 3.0 1.0 5.0 t.O T.O S.O
Emission Standard (Ib 50/10 Btu)
Figure 7. Percent of reserves in the entire U.S.A. that can be cleaned to
meet given emission standards.
-------
THE USE OF COAL CLEANING FOR
COMPLYING WITH S02 EMISSION REGULATIONS
Elton H. Hall1 and Gilbert E. Raines2
Battelle
Columbus Laboratories
505 King Avenue
Columbus, Ohio 43201
n
Raines Consulting, Incorporated
1016 Amberly Place
Columbus, Ohio 43220
ABSTRACT
Coal cleaning is an effective technique for reducing the sulfur content
of coal so that it can be burned in compliance with S02 emission regulations.
The potential role for coal cleaning in an overall S02 control strategy
depends upon the emission regulations and the cleanability of the coal.
The several types of S02 regulations are reviewed and compared with the
ranges of sulfur removal achievable through coal cleaning. . The quantities
of coal which could be cleaned to various sulfur levels, obtained by combining
coal reserve data with coal washability data, are presented. A comparison
is made of these quantities with the actual fuel requirements of existing
facilities classified according to the regulation each must observe. The
results indicate that coal cleaning could make a significant contribution
to S02 emissions control for boilers which are regulated by SIP's, and to
a lesser extent for boilers under current NSPS. The economic factors which
will determine the actual role of coal cleaning are discussed. The combined
use of coal cleaning and some other control technique, such as flue gas
desulfurization (FGD), has potential value in meeting existing NSPS, or
in meeting suggested revisions to the utility NSPS. In the latter case
credit can be taken for precombustion removal of sulfur by coal cleaning
which would reduce the percentage removal requirement for the FGD system.
416
-------
INTRODUCTION
Coal cleaning can be an effective technique for reducing
the sulfur content of coal, thereby reducing the emissions of
sulfur oxides when the coal is burned. Various technical and
associated environmental aspects of coal cleaning processes are
described in many other papers presented at this symposium. The
purpose of this paper is to explore how coal cleaning can be used
in complying with S02 emission regulations.
The potential of coal cleaning in an overall S02
emission control strategy depends upon the level and form of
the emission regulations, and on the cleanability of the coal.
This paper presents a review of the several types of SOj regu-
lations, a summary of the ranges of sulfur removal achievable
417
-------
through coal cleaning, a comparison of the quantities of clean-
able coal with actual fuel requirements of existing facilities
classified according to the regulation each must observe, and
an evaluation of how coal cleaning can best be used to comply
with SO2 emission regulations.
SO2 EMISSION REGULATIONS
The S02 emission standards for coal-fired steam generators
vary according to the size, age, and location of the facility.
Existing boilers are regulated by the State Implementation Plans
(SIP's). The SIP's vary from state to state with most states
using two or more levels, a low emission limit for plants in
metropolitan areas and higher emission limits for plants in non-
metropolitan areas. These regulations vary from 0.2 to 8.0 Ib
S02/10 Btu of boiler heat input. In many states the emission
limits apply only to boilers larger than a specified size.
Examples of SIP emission limits are given in Table 1.
New Source Performance Standards (NSPS) were promulgated
by EPA as required by the 1970 Clean Air Act Amendments. Those
standards apply to boilers, whether utility or industrial, larger
than 250 x 10 Btu/hr of boiler heat input, and constructed
after the date of promulgation. The emission limit for coal-
fired boilers under NSPS is 1.2 Ib S02/106 Btu. The 1977 Clean
Air Act Amendments significantly modified previous clean air
legislation. These Amendments require the use of best available
technology, a method of continuous pollution control, and
418
-------
achievement of a percentage reduction in the uncontrolled emis-
sions. EPA is now considering, and will soon propose, revised
NSPS for electric utility boilers. The revised NSPS probably
will retain a maximum emission limit of 1.2 Ib S02/10 Btu, but
will add a requirement of 80-90 percent reduction in uncontrolled
SO2 emissions. Uncontrolled emissions less than a minimum level
(0.2 to 0.5 Ib S02/10 Btu) would be exempt from the percentage
reduction requirement Credit for pre-combustion sulfur removal
will be given. The revised NSPS probably will apply to utility
boilers larger than 250 x 10 Btu/hr.
EPA also is considering tevised NSPS for industrial
boilers. At t-his time, the form of these revised regulations,
e.g. the maximum emission allowable, the percentage reduction
requirement/ and the size range of boilers to be regulated, is
unknown. A summary of these various regulations is given in
Table 2.
Variability of Sulfur in Coal
The fact that the composition and properties of coal can
vary widely, even within a given coal seam, is an important
consideration with respect to emission regulations. Because the
sulfur content varies, the average value for sulfur in coal can
be used to determine compliance with a given standard only if
long-term averaging of the resultant S02 emission is permitted.
If, however, the emission limit includes a "never to be exceeded"
419
-------
statement, a coal with average sulfur and heat content values
which are equivalent to the stated emission limit will be out of
compliance approximately half of the time. The net effect of
an emission regulation which calls for Anything other than long-
term averaging is to require the use of coal with a lower
average sulfur consent so that when upward deviations from the
average occur the unit will still be in compliance.
The impact of these considerations is shown in Table 3
in which the average emission level required by different aver-
aging times is listed for various emission limits.
It is apparent that short-term averaging requirements
will greatly reduce the quantities of raw coal which could be
burned in compliance with any given emission limit, because the
average sulfur content required for a 24-hour averaging period
is less than one-half of the value required for long-term aver-
aging. There is evidence to suggest that coal cleaning reduces
the sulfur variability. If this is the case, coal cleaning would
be even more effective in meeting emission standards with short-
term averaging periods than would be indicated by the reduction
in sulfur content achieved.
EFFECTIVENESS OF COAL CLEANING
The value of coal cleaning as a control technology for
meeting the emission regulations described, depends on the amount
of sulfur reduction which can be achieved. Cleanability data
420
-------
obtained by the U.S. Bureau of Nines suggest that common
commercial physical coal cleaning practice can remove from 28-55
percent of the pyritic sulfur as shown in Table 4. The result-
ing cleaned coals would, on combustion, produce emissions rang-
ing from 1.1 to 9.0 Ib S02/10 Btu. The best current technology
can achieve pyritic sulfur reductions of 43-80 percent, and
average emission levels of 0.9 - 4.4 Ib S02/10 Btu. Research
on advanced physical cleaning techniques indicates the potential
for removing 90 percent of the pyritic sulfur to yield a product
with S02 emissions in the range of 0.8 to 3.5 Ib S02/10 Btu.
Examination of the ranges in the effectiveness of
physical cleaning shows that the technology cannot be employed as
the sole control method to meet an 80-90 percent reduction
standard. However, the effectiveness of sulfur removal is such
that most of the SIP's can be met by burning physically cleaned
coal.
Similar data are shown in Table 5 for chemical coal clean-
ing. Greater reductions in pyritic sulfur can be achieved by
chemical cleaning. In addition, some of the processes under
development are capable of removing a portion of the organic
sulfur. Again, none of the chemical coal cleaning processes
could be used alone to meet an 80-90 percent reduction standard,
but chemically cleaned coal would be used to meet many of the
SIP regulations.
421
-------
QUANTITIES OF CLEJANABLE COAL
In order to evaluate the potential £or coal cleaning
in meeting S02 standards, it is necessary to estimate the
amounts of coal which could be cleaned to various levels by
using different cleaning processes. Such estimates were de-
veloped through an overlay of coal reserve data (U.S. Bureau
of Mines 1975), coal cleanability data (.U.S. Bureau of Mines
1976), and a third data set consisting of approximately 50,000
records of coal sample analyses. The computer programs which
were developed to carry out these calculations are described
in another paper presented at this symposium CMcCreery and Good«
man, 1978) .
Several different coal cleaning processes were examined
to determine their effects on coal from each of the six major
coal producing regions and from the U.S. as a whole The coal
cleaning processes considered were:
A. Physical coal cleaning using 1-1/2 inch mesh
at 1.6 specific gravity (s.g.)
B. Physical coal cleaning using 3/8-inch mesh
at 1.6 s.g. if this produced coal to meet
the standard being considered, otherwise 1.3
s.g. was used. An operating penalty of 1 per-
cent energy loss was assumed.
422
-------
C. Meyers process: for raw coal with greater than 0.2
percent pyritic sulfur, the level of pyritic sulfur
is reduced to 0.2 percent. No sulfur reduction
takes place if the raw coal pyritic sulfur level
is less than 0.2 percent, A 5 percent energy loss
was assumed plus an operating penalty of 2 percent
energy loss and a weight loss of 10 percent.
D. Gravichem process: crush coal to 14-mesh topsize,
separate at 1.3 s.g.; treat sink with Meyers
process Cwith the same energy and weight losses
as in C); combine the float and the processed sink.
E. 95 percent pyritic sulfur and 20 percent organic
sulfur removed with 10 percent energy loss plus
operating penalty of 2 percent energy loss and a
weight loss of 15 percent,
F. Physical cleaning using 3/8-inch mesh at 1.3 s.g.;
no operating penalty.
G. 95 percent pyritic sulfur and 40 percent organic
sulfur removed with 30 percent energy loss.
H. 70 percent organic sulfur removed.
Processes A-E are considered to be realistic processes while F-H
are considered to be hypothetical,
An example of the results obtained from these calculations
is shown in Northern Applichian coals in Figure 1 in which the
Percentage of the recoverable reserves of the region, expressed
423
-------
in terms of energy content, is plotted against the S02 emission
which would result on combustion. Curves are shown for raw
coal and for four of the cleaning processes. The interpretation
of these curves may be illustrated by considering an emission
standard of 1.2 Ib S02/10 Btu, From the raw coal curve, 5.3
percent of the reserve, or 91 x 10 Btu, could be burned in
compliance. The corresponding values taken from the curves for
the various cleaning processes are given in Table 6. The effec-
tiveness of the cleaning processes is apparent as the quantity
of compliance coal which could be made available through coal
cleaning can be as much as six or seven times that for raw coal.
Separate curves of this type were developed for the deep,
strip, and total reserves for each of the six coal producing
regions, and for the entire United States. The total United
States curves for raw coal and for four of the cleaning processes
are shown in Figure 2. The indicated percentages of the reserve
and the associated energy contents which could be burned in
compliance with a 1.2 Ib S02/10 Btu standard are given in Table
7,
An additional series of calculations was made from the
reserve/cleanability model to determine the percentage of sulfur
in cleaned coal which must be removed by scrubbing to achieve
90 percent overall suflur reduction by the combined control
methods. As an example of this type of output, the results for
three cleaning processes for the entire United States are shown
in Figure 3. The curves show that for a scrubber operating at
424
-------
70 percent sulfur removal, for example, the following quantities
of coal could be used in compliance with a 9Q percent reduction
standard:
Treatment Method Percent of Reserve Energy Content,10 Btu
PCC, 1-1/2 inch, 5.0 438
1.6 s.g.
Meyers 14.2 1251
95 percent Pyritic S, 17.3 1532
20 percent Organic
S Removed
These results indicate that, while coal cleaning alone cannot
satisfy a 80-90 percent reduction standard, the combined use of
coal cleaning and scrubbing could be an effective approach in
that it allows the scrubber to operate at lower and more readily
achieved level of sulfur reduction.
COMPARISON OF CLEANABLE COAL
QUANTITIES WITH COAL REQUIREMENTS
A procedure was developed to relate industrial energy
requirements to the quantities of raw coal and of coal that could
be made available by application of various cleaning processes
which could be used to meet prescribed SO,, emission standards.
Calculations have been completed for the Northern Appalachian
Region.
Quantities of raw coal and cleaned coal meeting various
standards were obtained from the reserve/cleanability -model as
discussed above. Industrial demand was determined from a
425
-------
characterization of- existing industrial fuel burning facilities
according to state, SIP requirements within the state, capacity,
and fuel. SIP requirements were simplified by using at most
two SIP standards (basically metropolitan and non-metropolitan)
in any one state. The major source of data was the FEA survey
of "Major Fuel Burning Installations CMFBI)", which gave data
as of 1974. Survey data included itemization of details for
each combustor above 100 x 10 Btu/hour.
The MFBI survey required listing of total capacity at each
installation but did not require an itemized breakdown for "small'
combustors, defined as those below 10.0 x 10 Btu/hour. Thus,
it was necessary to synthesize these small-combustors constitute
approximately 40 percent of the total industrial fuel burning
capacity.
The procedure for this synthesis consisted of applying data
from an EPA boiler survey CPaddock and McMann., 1975). on capacity
distribution to estimate the allocation among coal, residual,
distrillate, gas, and other for "small" boilers. The MFBI total
of "small" and "large" boilers and "large" non-boilers was sub-
tracted from the entire capacity in each region thus determining
the total of "small" nonOboilers. There are no data on capacity
distributions among fuels for "small" non-boilers. These
capacities were distributed assuming that the ratios of "small
coal non-boiler capacity" to "small gas non-boiler capacity" to
residual, etc. were the same as similar ratios for large boilers.
Thus, it was possible to determine "small" non-boiler capacities
by different.
-------
The total current industrial demand in each SIP region is
obtained by adding "large" boiler, "small" boiler, "large" non-
boiler, and "small" non-boiler capacities for each fuel. The
potential for coal utilization consists of the total of coal,
residual, distillate, and gas combustor capacities. Facilities
utilizing "other" fuels were not considered convertible to coal
because the "other" fuel is hog fuel, refinery off-gas, and
other waste ot by-product materials for which there is little
other demand. Table 8 illustrates these results for the coal-
producing states in the Northern Appalachian Region.
A "coal use" model was developed which assumes that a demand
area will use coal with the highest sulfur level possible to
meet its S02 emission standard. As these reserves are depleted,
the area then uses coal with lower and lower sulfur levels. This
procedure would be approximate in reality Cover a long time
period) if the cost of coal increases with a lowering sulfur
content. Thus, a large capacity user with a fairly non-restrictive
SIP, e.g., non-metropolitan Ohio, will ultimately be using the
same cserves as a smaller capacity user with a less restrictive
SIP, e.g., non-metropolitan Maryland. The model is run until
all of the coal in the region is consumed with the various users
running out of coal (.depending on supply, SIP, and demand) at
various times. Table 9 illustrates results for the case if only
the coal producing states are users, A load factor of 1,0 was
used in these runs. Load factors vary widely but typically are
no more than 0.5. It should be noted that the Northern Appalachian
427
-------
coal reserves considered do not include the very high, quality
coal in Southern West Virginia (a part of the Southern Appalachian
coal producing region). The processes designated in Table 9
are the same as listed previously.
The results shown in Table 9 are valuable in comparing the
usefulness of various coal cleaning processes in preparing coal
to meet existing SIP's for existing facilities. For example,
the results indicate that coal cleaning processes may be used
to increase the supply of coal to satisfy S02 emission standards
of 1.6 Ibs S02/10 Btu by a factor of up to 3 or 4 over raw coal.
The absolute magnitudes of the years of available coal are not
meaningful at this stage of the analysis since other consuming
states and other coal uses are not included. Analyses extended
to all regions and to the utility sector are being completed and
the results will be published in an EPA report,
CONTROL TECHNOLOGY COSTS
The results which have been presented show that large
quantities of coal can be cleaned to meet various S02 emission
limits. The extent to which coal cleaning will be actually
w
employed as an S02 control technique will be determined in part
on the basis of comparative costs. For the purposes of this.
paper only a generalized overview of costs is presented.
Coal cleaning costs depend upon a number of factors which
include:
o Plant complexity (level of cleaning)
428
-------
o Plant Size and Operating factor
o Coal Replacement Costs
o Pollution Control Costs
o Reliability and Product Control
o User Cost Benefits
o Finance Considerations
Ranges of annualized costs for both, physical and chemical coal
cleaning are given in Table 10. Chemical cleaning costs are
more than a factor of two higher than those for physical coal
cleaning.
The costs of flue gas desulfurization (FGD) also depend on
a number of factors which include:
o Type of FGD System
o Boiler Size and Operating Factor
o Sulfur Removal Requirements
o Pollution Control Costs
o Reliability and Control
o Finance Considerations
Estimated annualized FGD costs, based on lime-limestone scrubbing,
are shown in Figure 4. The costs range from about $1.00/10 Btu
for smaller boilers using high-sulfur coal at a low operating
factor to about $0.30/10 Btu for large boilers burning low-
sulfur coal at a high operating factor,
429
-------
A comparison of these cost ranges is given in Figure 5.
General conclusions from the comparison are as follows:
o Physical coal cleaning offers cost sayings over FGD
especially for small boilers with low capacity factors,
o Coal cleaning used in conjunction with. FGD may be
cost effective over FGD alone, in some cases, since
the cleaned coal requires less sulfur removal than
uncleaned coal.
o Chemical coal cleaning may be cost effective as
compared with FGD in some smaller boilers with.
low capacity factors.
Because the costs of coal cleaning and FGD are sensitive to a
number of different factors, site-specific analysis of costs is
required to determine the most cost effective' control technique
for each site.
CONCLUSIONS
The greatest role for physical coal cleaning appears to be
in meeting state S02 emission regulations on existing boilers.
Cleaning methods exist for preparing coal to meet many of the
various SIP levels and the estimated quantities of cleanable
coal are substantial.
430
-------
Since a number of coals can be cleaned sufficiently to
comply with a standard of 1.2 Ib S02/10 Btu, coal cleaning,
also can fill a role in meeting current NSPS for coal-fired
steam generators.
Coal cleaning cannot be used alone to meet an 80-90 per-
cent reduction standard, now under consideration. However, the
use of cleaned coal would allow a scrubber or other control
system to be operated at a lower level of sulfur .reduction. The
combination of techniques may offer cost savings in some cases.
Further, given the current uncertainty over the ability of
scrubbers to operate consistently at high levels at high levels
of sulfur reduction, the use of cleaned coal might be the only
means of complying with an 80-90 percent reduction requirement
over the near term.
The nature of revised NSPS for industrial boilers is unknown
at this time. However, the size range of regulated boilers .may
include smaller than is the case for utility boilers which would
increase the potential for coal cleaning in view of the cost
advantages for smaller boilers.
The current cost projections for chemical coal cleaning
indicate that the most probable applications for such processes
will be to provide lower sulfur levels than can be achieved by
physical cleaning where required by a particularly stringent
regulation.
431
-------
CONVERSION FACTORS
lb H 0,454 kg
Btu ~ 1055.6 joule
106 Btu = 1.056 GJ
lb/106 Btu * 0.430 kg/GJ
REFERENCES
The Reserve Base of U.S. Coals by Sulfur Content, 1C 8680 and
1C 8693, U.S. Bureau of Mines, 1975.
Sulfur Reduction Potential of U.S. Coals, RI 8118, U.S. Bureau
of Mines, April, 1976.
"An Evaluation of the Desulfurization Potential of U.S. Coals",
J. H. McCreery and F. K. Goodman, Symposium on Coal Cleaning
to Achieve Energy and Environmental Goals, Hollywood, Florida,
September 1978.
Paddock, R. E. and McMann, D.C., "Distributions of Industrial
and Commercial-Institutional External Combustion Boilers", EPA
650/2-75-021 CFebruary, 1975).
432
-------
Table 1. Typical State Emission Limits for Coal-
Fired Boilers (Ib S02/106 Btu)
METROPOLITAN NON-METROPOLITAN
STATE AREAS AREAS
^ ALABAMA
COLORADO
ILLINOIS
IOWA
KENTUCKY
OHIO
PENNSYLVANIA
WEST VIRGINIA
1.8
0.2
ijB
5.0
1.2
1.4
0.7
2JS
4.0
0.2
6.0
5.0
5.7
4.5
4.0
2.8
-------
Table 2. S02 Emission Standards for Coal-Fired Steam Generators
SULFUR EMISSION
REDUCTION. LIMITS,
PERCENT LB SO2/106 BTU
EXISTING BOILERS (SIP'S) - 0.2-8.0
CURRENT NSPS FOR STEAM - 1.2
GENERATORS
REVISED NSPS FOR UTILITY BOILERS^ 80-90 1.2 MAX.
0.2 FLOOR
NSPS FOR INDUSTRIAL BOILERS UNKNOWN UNKNOWN
(a) Values under consideration.
-------
-p-
10
Table 3. Effect of Averaging Time on Average Coal
Sulfur Level Required for Compliance
AVERAGE COAL SULFUR VALUE REQUIRED,
LB SO2/106 BTU
EMISSION LIMIT,
LB SO2/106 BTU
1.2
0.8
0.4
LONG-TERM
AVERAGE
1.2
0.8
0.4
30-DAY
AVERAGED
0.92
0.62
0.31
24-HOUR
AVERAGED
0.58
0.39
0.19
(a) Assumes relative standard deviation of 10 percent.
(b) Assumes relative standard deviation of 36 percent.
-------
Table 4, Physical Desulfuriz^tion
(a)
03
CLEANING
TECHNIQUE
UNCLEANED COAL
COMMON COMMERCIAL
PRACTICE
BEST CURRENT
COMMERCIAL
TECHNOLOGY
BEST POTENTIAL
TECHNOLOGY
AVERAGE PYRITE
REMOVED,
PERCENT
28-55
43-80
90
AVERAGE REDUCTION
IN LB S02/106 BTU,
PERCENT
11-40
16-55
30-60
AVERAGE EMISSION
LEVEL ON COMBUSTION,
LBSO2/106BTU
1.1-9.0
0.9-6.5
0.9-4.4
0.8-3.5
(a) Based on data from Bureau of Mines Rl 8118 for averages from each of six coal regions.
-------
Table 5. Chemical Desulfurization
Ca)
CLEANING
TECHNIQUE
AVERAGE PYRITE
SULFUR REMOVED,
PERCENT
AVERAGE ORGANIC
SULFUR REMOVED,
PERCENT
AVERAGE REDUCTION
IN LB SO2/106 BTU
-PERCENT
AVERAGE EMISSION
LEVEL ON
COMBUSTION,
LB SO2/106 BTU
UNCLEANED
COAL
1.2-9.0
LO
PYRIT6
LEACHING
BEST AVAILABLE
TECHNOLOGY
95
95
25
32-65
48-73
0.8-3.2
0.6-2.5
BEST PRACTICAL
TECHNOLOGY
95
40
58-77
0.5-2.0
(a) Based on data from Bureau of Mines Rl 8118 for averages from each of six coal regions.
-------
Table 6. Northern Appalachian Coals
AMOUNT OF COAL IN COMPLIANCE
WITH 1.2 LB S02/108 BTU
PERCENT OF ENERGY CONTENT,
TREATMENT METHOD RESERVE 1 pi5 BTU
RAW COAL 5.3 91
PCC, 1-1/2 INCH, 1.6 S.G. 12.8 221
PCC, 3/8 INCH, 1.6 OR 1.3 S.G. 24.7 427
MEYERS PROCESS 29.5 509
95% PYRITIC S, 20% ORQ. S REMOVED 36.8 636
"BEST" FOR RESERVE 40.9 707
438
-------
Table 7. United States Coals
AMOUNT OF COAL IN COMPLIANCE
WITH 1.2 LB SO2/106 BTU
PERCENT OF ENERGY CONTENT,
TREATMENT METHOD RESERVE 1015 BTU
RAW COAL 36.8 3252
PCC, 1-1/2 INCH. 1.6 S.Q. 45.1 3987
PCC, 3/8 INCH. 1.6 OR 1.3 S.Q. 49.2 4347
MEYERS PROCESS 51.3 4534
95% PYRITIC S, 20% ORQ. S REMOVED 57.2 5050
"BEST" FOR RESERVES 57.2 5050
439
-------
Table 8. Industrial Combustor Capacities (Excluding
"Other" Fuel) For Coal-Producing Areas of
the Northern Appalachian Region
State
Maryland, Metro
Maryland/ Non-Metro
Ohio, Metro
Ohio, Non-Metro
Pennsylvania, Metro
Pennsylvania, Non-
Metro
West Virainia
Coal
3
1
27
24
26
29
19
.310
.105
,447
.545
.483
.838
..187
Residual
3
1
5
5
36
3
4
.295
.750
,847
.222
.970
.885
.961
Distillate
9
0
3
0
2
0
0
.135
.000
.702
.164
.641
.907
.173
Gas
10
0
38
19
29
7
4
.229
.727
.512
.040
.939
.171
.565
Total
25,
3,
75.
48.
96.
41.
28.
968
583
508
972
033
801
886
Northern Appalachian 131.916 61.930
16.721
110.183 320.751
440
-------
Table 9,
Years of Available Coal in the Northern Appalachian
Region Using Various Cleaning Processes to Satisfy
the Coal Producing States' Current Aggregate of
Industrial Coal, Residual, Distillate, and Gas
Combustor Caoacities
State or Section
of State
Pennsylvania
(Metropolitan)
Ohio
(Metropolitan)
Maryland
(Metropolitan)
Vest Virginia
Pennsylvania
(Non-Metropolitan)
Maryland
(Ron-Metropolitan)
Ohio
f «r«^u.*._«_ * .* *. %
SIP ,
Ibs S02/10° Btu
0.8
•1.4
1.6
2.6
3.0
3.5
4.5
Tears
Raw
23
173
173
586
586
1.013
1,013
of Available Coal
A
82
311
311
880
880
880
880
B
267
447
447
816
816
816
816
Using Raw
C
123
679
684
832
832
832
832
Coal
D
205
641
641
816
816
816
816
or Each
E
370
641
641
720
720
720
720
of the Designated
F
205
428
428
800
800
800
800
G
308
603
603
768
768
768
768
Processes
H
164
233
233
683
1.038
1,088
1,088
-------
Table 10. Annualized S02 and Particulate Control
Costs ($/106 Btu)(a)
TECHNIQUE MINIMUM(b) MAXIMUM*0*
PHYSICAL COAL CLEANING 0.26 0.46
CHEMICAL COAL CLEANING 0.60 1.13
(a) Includes $0.10/106 Btu for participate control.
(b) Minimum coits correspond to 40,000 million Btu/hr plant capacity and
$0.05/10s Btu coal replacement costs.
(c) Maximum costs correspond to 10,000 million Btu/hr plant capacity and
$0.18/106 Btu coal replacement costs.
442
-------
ESTIMATED CLEANING POTENTIAL OF
NORTHERN APPALACHIAN COALS
LLJ
CJ
CC
Q_
^
CD
CJ
O
O
O
CD
CD
CC
TREATMENT METHOD
RAW COAL
PCC 1-1/2 INCH, 1.6 SPECIFIC GRAVITY
PCC 3/8 INCH, 1.6 OR U SPECIFIC GRAV.
MEYERS PROCESS
0.95 PYRITE S, 0.20 ORG. S REMOVED
'BEST* FOR RESERVE
/// / / '
//// / =
I /
If / /
ENERGY CONTENT OF
RECOVERABLE RESERVES:
Figure 1. Coal Sulfur Emission on Combustion, Ib S02/10 Btu
-------
DC
U-l
0_
GQ
s
O
LJU
O
CD
DC
80-
40H
ESTIMATED CLEANING
POTENTIAL OF U.S. COALS
TREATMENT METHOD
RAW COAL
PCC 1-1/2 INCH, 1.6 SPECIFIC GRAVITY
PCC 3/8 INCH, 1.6 OR 1.3 SPECIFIC GRAVITY
MEYERS PROCESS
0.95 PYRITE SULFUR, 0.20 ORG. S REMOVED
'BEST' FOR RESERVES
ENERGY CONTENT OF RECOVERABLE
RESERVES:
8834 X 1015 BTU
Figure 2. Coal Sulfur on Combustion, Ib S02/1Q6 Btu
-------
Figure 3.
Estimated Compliance Coals Available Through Cleaning
and Scrubbing to Meet Requirement of 90 Percent
Sulfur Reduction
100
CJ>
DC
LU
Q_
GO"
-------
ON
_
&
CO
8
3
1
3.00
2.00
1.50
1.20
1.00
0.20
HIGH SULFUR COAL,
LOW OPERATING FACTOR
LOW SULFUR COAL,
HIGH OPERATING FACTOR
50
100
Figure 4,
200 500 1,000 2,000
BOILER CAPACITY, 106 Btu/Hr
5,000 10,000
Estimated Annualized FGD Cost Ranges for
Sulfur and Particulate Emission Control
(Based on Lime-Limestone Scrubbing)
-------
Figure 5.
3.00
2.00
GO
CD
CD
CD
CC
1.00
CD
CO
0.50
0.40
0.30
Annualized S02 and Particulate Control Costs
—\ 1 1
MM
T
T
T
» •
1
1
100
200
500 1,000 2,000
BOILER CAPACITY, 10 6BTU/hr
5,000
10,000
-------
STATISTICAL CORRELATIONS ON COAL DESULFURIZATION
BY CRUSHING AND SPECIFIC GRAVITY SEPARATION
Ralph E. Thomas
Battelle
Columbus Laboratories
505 King Avenue
Columbus, Ohio 43201
ABSTRACT
The weight fractions for washability data for Homer City coal (Feed
No. 1) are represented by Rosin-Rammler distributions. Excellent fits are
obtained for the distribution of weight, according to size, for each of 12
specific gravity fractions. The characteristic size parameters and dispersion
parameters show some non-uniform behavior across the various levels of
specific gravity. This behavior is currently under, study in an effort
to obtain a surface that can be used as a general basis for interpolating
among coarse washability data to obtain more refined washability data.
448
-------
INTRODUCTION
Washability data for coal is costly to generate, especially
when it is desired to obtain precise estimations of weight,
pyritic sulfur, total sulfur, and ash according to, say, 12
specific gravity fractions for each of 9 size fractions. Such
detailed analyses appear to be required in order to properly
determine which (specific gravity/size)-fractions are most
affected by the various coal cleaning processes, by related
coal cleaning equipment, and by the various design parameters
associated with such equipment. Good predictions of the output
flow rates for clean/ middling, and refuse coal requires refined
washability data. Even the precisions of computer simulations
of coal cleaning processes and equipment are likewise con-
strained by the available level of refinement of coal washability
449
-------
data. The results reported below are the initial results of
an effort to identify a quantitative method for interpolation
among limited washability data. By interpolation such a method
would permit the refinement of coarse washability data, and
would also be expected to identify the minimum number of wash-
ability measurements required to achieve a specified precision.
METHOD
In this effort several interpolation methods have been
briefly examined. These methods include empirical curve fitting
with splines and polynomials, and a more traditional procedure
based on the Rosin-Rammler distribution.
In general, it is expected that whenever refined washability
data are available, several different interpolation methods are
likely to be acceptable. However, when the washability data are
coarse, as in R8118 with 3 or 4 specific gravity fractions and
3 size fractions, it is expected that some assumed distributional
form, such as the Rosin-Rammler distribution, will be essential
450
-------
in order to compensate for the severely limited washability
information.
It is exceedingly important to assume a correct distribu-
tional form for interpolating washability data. In effect/ with
severely limited data/ there is no satisfactory way to test the
correctness of the assumed distributional form. In statistical
terms/ some of the limited degrees of freedom are lost in
estimating the parameters of the distribution. As a consequence,
except for the most extreme cases/ the assumed distributional
form can neither be conclusively accepted nor rejected by the
data.
Because the Rosin-Rammler distribution has a long history
of successfully representing size distribution data for coal,
(Leonard and Mitchell/ 1968) this distribution is being
considered as the primary candidate distribution. Previous
efforts known to the author have fitted the Rosin-Rammler
distribution to size data without regard to specific gravity.
In contrast/ the strategy used in this effort consists of first
determining whether a separate Rosin-Rammler distribution gives
a good fit to the size-fraction data within each specific-
gravity fraction. That is/ if 12 specific-gravity fractions
are available, then 12 different Rosin-Rammler curves are fitted
to the data. If suitably well-behaved, these curves/ in turn,
are then regarded as parallel slices from a mathematical surface
representing all of the data. Finally the resulting
451
-------
Rosin-Rammler surface would then be used to interpolate among
the actual data values, and thereby "refine" the washability
data to the extent required.
Because the Rosin-Rammler distribution has been success-
fully used to fit size-distribution data, as a composite over
all specific gravities, it appears to be a good candidate for
fitting the size-distribution data for each separate specific-
gravity fraction. However, it should be noted that washability
data are usually presented as distributions of percent weight,
sulfur, ash, etc., according to specific gravity, for each
size fraction. For the present effort the data are first
re-cast to express the-.distributions, according to size frac-
tion, for each specific gravity fraction.
Estimation Procedure
The cumulative form of the Rosin-Rammler distribution is
given as follows-
F(x) - 1 - exp t-(x/d)n], o 1 x,
where d denotes the characteristic diameter and n denotes the
dispersion parameter. Thus, F(x) gives the fraction of the
distribution associated with a size less than, or equal to, x.
It is seen from the above expression that if x is set equal to
the characteristic size, then F(d) * 1 - (1/e), so that approx-
imately 63 percent of a Rosin-Rammler size distribution consists
of sizes less than the characteristic size.
452
-------
Limited examinations have been made of several different
methods for estimating the parameters d and n. In addition to
graphical methods these methods include the method of maximum
likelihood with correction for bias (Fishman, 1973) and least
squares regression (Hald, 1952). To date the most satisfactory
results are obtained by applying least squares regression to the
following logarithnmic form of the Rosin-Rammler distribution:
In [Ind-F(x))"1] = n[lnx] - nflnd].
This expression is seen to correspond to a linear regression
form:
Y - AX + B,
where X and Y correspond to In x and ln[ln(l-F(x))-1], respectively;
and the regression parameters A and B are equated to n and -n[ln d] ,
respectively. The regression estimate A is taken to be a direct
estimate of.n; the regression estimate § is set equal to
-n[lnd] = -A[lnd] so that d is indirectly estimated by
d = exp(-B/A). In contrast to the ideal situation, the estimators
of n and d are thus seen to be interrelated. However, this
deficiency does not appear to be important for the data examined
to date.
It must also be noted that the Rosin-Rammler distribution
accomodates sizes that are distributed over the entire positive
range. Arbitrarily large sizes are theoretically-permitted pro-
vided the associated level of probability is sufficiently small.
453
-------
Actual size distributions are truncated in that no sizes larger
than a truncation size, say 5/4 inch, will occur in the real
coal sample. This means that the actual empirical distribution
will have 100 percent of the size distribution smaller than the
truncation size, 5/4 inch for example. The theoretical Rosin-
Rammler distribution can never have 100 percent of the size
distribution smaller than any fixed finite truncation size.
This means that it is frequently necessary to estimate how much
truncation has occurred, and make a suitable correction for
such truncation, before fitting the empirical size distributions
to a theoretical Rosin-Rammler distribution. An iterative pro-
cedure is used to correct for truncation in the results reported
below.
Results
Figure 1 shows the empirical data (corrected for trun-
cation) and the fitted Rosin-Rammler distributions for the
percent weight distributions for Homer City coal, Feed No. 1.,
for the specific gravities shown on each caption. In general,
the results show that separate Rosin-Rammler distributions
provide excellent representations of the size-distribution data
associated with each specific gravity fraction.
Table 1 shows the numerical estimates of the characteristic
size d and the dispersion parameter n for each specific gravity
fraction for the individual fitted Rosin-Rammler distributions.
454
-------
s".
!
NATURAL LOO OF SIZE
SPECIFIC GRAVITY
-1.300 -1.400 -.MO .400 1.100
NATURAL LOG OF SIZE
SPECIFIC GRAVITY - •«.•»«
NATURAL LOO OF SIZE
SPECIFIC GRAVITY
I.Z7
NATURAL LOG OF SIZE
SPECIFIC GRAVITY
1.00*
FIGURE 1. Rosin-Rammler Representations of Weight Percent,
According to Size Fraction, For Each Specific Gravity Fraction
For Homer City, Feed No. 1, Washability Data
455
-------
S •••
s
M
I
H
HJ
•1.4M -I.W -I.M* -.'•»
' NATURAL LOO Of SIZE
SPECIFIC GRAVITY
t.tf
-••M* -4.ro* -1.004) -I. MO -I.Mt
NATURAL LOO W 81 It
SPECIFIC GRAVITY
•I.MO -4.MO
•1.404 'I.IM -I
NATURAL LOO OF SIZE
SPECIFIC GRAVITY
•MO -.700
HJ
-I.*M -4.IM .1.1(0 -I.IM .'I.«M
' NATURAL LOO OP SIZE
SPECIFIC GRAVITY
FIGURE 1. (Continued)
456
-------
a
M
8 T>
•I.FM -4.«o*
-l.M* -I.M*, •!.!•
NATURAL LOO Of 1IZE
SPECIFIC GRAVITY
>4.M« -4.«M -I.IN -I.IM -I.M
dATWAL LOO OF Illf
SPECIFIC GRAVITY' • /.7«
S!
•*.MO -I. MO
-4.
-------
TABLE 1,
Rosin-Rammler Parameters for Fitted
Size Distributions According to Specific
Gravity Fraction
Homer City, Feed No. 1
Specific Characteristic Dispersion
Gravity Size, d(mm) Parameter, n
FLOAT
1.25
1.27
1.30
1.32
1.35
1.40
1.45
1.50
1.60
1.70
1.80
- 1.25
- 1.27
- 1.30
- 1.32
- 1.35
- 1.40
- 1.45
- 1.50
- 1.60
- 1.70
- 1.80
- SINK
0.3
2.2
3,3
4.5
10.6
38.0
26.5
21.1
44.1
46.1
68.6
78.8
0.52
0.77
0.81
0.79
0.72
0.56
0.71
0.71
0.64
0.73
0.72
0.61
(1)
Rosin-Rammler distribution:
F(x) = 1 - exp (-(x/d)n),
where F (x) denotes the proportion of the distribution
less than size x, with d and n denoting the characteristic
size and dispersion parameter, respectively.
458
-------
An examination of this table shows an- irregular increase in
the characteristic size associated with increasing specific-
gravity fractions. Smooth behavior across the levels of specific
gravity would appear to require a characteristic size somewhat
smaller than the 38.0 mm size shown at a specific gravity of 1.4.
Characteristic sizes somewhat larger than 26.5 and 21.Lr™im would
also be indicated for specific gravities of 1.45 and 1.50.
The dispersion parameters are somewhat uniform over most
specific gravities. The lowest values of the dispersion para-
meter are seen to be associated with the smallest characteristic
size 0.3 mm, and with the possibly aberrant 38 mm size at a
specific gravity of 1.40. The arithmetic mean and standard
deviation of the values of dispersion parameters are found to
be 0.69 and 0.09, respectively, with a coefficient of varia-
tion of 13 percent.
Figure 2 shows an overlay of all 12 Rosin-Rammler dis-
tributions represented on In-In vs. In scales. Uniform behavior
from one specific gravity to the next would be indicated on
such a plot if, for example, all lines were coincident, or if
a uniform rotation occurred about the point of concurrence
at (0,0). In fact, however, no such uniform behavior is
exhibited. This is due, in part at least, to the non-uniformities
previously mentioned for the n and d parameters associated with
specific gravities 1.40, 1.45, and 1.50 as shown in Table 1.
Figure 3 shows a plot of log dn versus specific gravity.
This plot suggests that a discontinunity has occurred that
459
-------
o
< «
o
• <
o
CM .
O
O
. 3
o
00
o
o
o
£ §
O o
o 7
o
O
o
. .
o
o
< i
I
o
o
1*3
i
O
o
o
o
o
f-.l
10
i
I
-7.000
-5.000
T"
-3.000
-1.000
T"
1 .000
T"
3.000
5-000
FIGURE 2. Superimposed Rosin-Rammler Representations of Weight
Percent, According to Size Fraction, for 12 Specific Gravity
Fractions for Homer City, Feed No. 1, Washability Data
460
-------
1.4 j.
1.2 .
1-0 ..
0.6
o.o
1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8
FIGURE 3. Parameter Variation with Specific Gravity
461
-------
separates the last 4 specific gravity fractions from the earlier
fractions. Other examinations of these results are also being
made to account, if possible, for the observed non-uniform
behavior.
In summary, these results indicate that the weight
fractions for washability data for Homer City coal (Feed Mo. 1)
are well-represented by Rosin-Rammler distributions, with a
separate distribution used to represent each of the 12 specific
gravity fractions. The dispersion parameter is found to lie in
the 2 standard deviation interval 0.69 ± 0.18 across all specific
gravity fractions. The characteristic size parameter shows some
irregularity across the specific gravity fractions. This be-
havior is currently under study as part of the general effort to
obtain a Rosin-Rammler surface that can be used as a general
basis for interpolating among qoarse washability data to obtain
more refined washability data.
462
-------
REFERENCES
Fishman, G.S. 1973. Concepts and methods in discrete event
digital simulation, John Wiley & Sons, Inc.
Hald, A. 1952. Statistical theory with engineering applications,
John Wiley & Sons, Inc.
Leonard, J.W. and D.R. Mitchell (ed.). 1968. Coal preparation,
third edition, AIME, New York.
463
-------
DEWATERING AND DRYING OF FINE COAL:
EQUIPMENT PERFORMANCE AND COSTS
Donald H. Sargent, Bill H. Cheng, and G. Yeghyazarian Contos
Versar, Inc.
Springfield, Virginia
ABSTRACT
The physical cleaning of fine coal for pyrite removal results in product
and refuse streams with high moisture contents. The added costs for dewatering,
for transportation, and for environmental controls have been generally recog-
nized, but they have not previously been systematically documented such that
they may be compared with the benefits achieved by coal cleaning.
The Environmental Protection Agency therefore directed Versar, Inc.,
to fully define the costs of fine coal dewatering, handling and transportation;
with the objective of enabling the cost evaluation under any reasonable combi-
nation of fine coal product stream size consist and initial water content,
fine coal dewatering and drying unit process alternatives, and fine coal
handling and transportation alternatives.
First, the unit processes and equipment for fine coal dewatering were
systematically studied. For each, the performance in terms of water removal
capability; the useful range in terms of feed size consist and feed moisture
content; the required size as a function of throughput; and the equipment
costs as a function of size; were all defined. Examples are presented in
this paper.
Also presented are the results of an early case study in which the costs
to a preparation plant operator for alternative dewatering and drying schemes
are compared to the economic benefits achieved by shipping drier coal to an
electric utility coal user.
464
-------
Introduction
Historically, physical coal cleaning for ash removal was
directed at coarse coal fractions. Fine coal, 3/8-inch x 28 mesh
material, was not generally beneficiated but was directly blended
with the cleaned coarse coal. The very fine fraction, 28 mesh x 0
material, was discarded as refuse.
In contrast, physical coal cleaning for sulfur removal
necessitates crushing to sufficient fineness to liberate the pyrite.
The higher fractions of fine coal resulting from continuous mining
techniques and from crushing for pyrite liberation are processed
465
-------
by a variety of separation techniques, and constitute a major portion
of the clean coal product. Both the higher surface moisture content
of fine coal fractions and the higher percentages of fine coal in the
clean coal product result in added costs for dewatering and drying,
for transportation, and for environmental controls.
As part of a Coal Cleaning Technology Development program
being conducted for the Environmental Protection Agency under Contract
68-02-2199; EPA directed Versar, Inc., to perform a cost evaluation for
fine coal dewatering and drying under any reasonable combination of fine
coal product stream size consist and initial water content, fine coal
dewatering and drying unit process alternatives, and fine coal handling
and transportation alternatives. Although these added costs have been
generally recognized, they have not previously been systematically
documented such that they may be compared with the environmental and
economic benefits achieved by coal cleaning.
Dewatering and Drying Equipment Performance
Table 1 lists several major categories of equipment useful in
the dewatering and drying of fine coal. The ranges of moisture content
reductions shown in Table 1 dictate the place each type of device has
in an overall dewatering and drying process. The first stages of
dewatering a fine coal slurry with 80 to 95 per cent water might be
466
-------
followed by devices which can further dry the coal. The selection of
equipment is further guided by the size range of• the coal -a centrifuge
might be chosen for a 3/8-inch top size coal, whereas a vacuum filter
would be selected if the top size were 28 mesh. Several categories
of equipment result in comparatively low solids recoveries, and gen-
erally should be followed by some effluent treatment scheme.
Based upon extensive data gathered for specific items of
equipment, dewatering and drying performance curves were generated
to cover more generalized equipment types. Figure 1, for Screens, and
Figure 2, for Centrifuges, are two examples. The performance curves
are useful in predicting the moisture content of the product from
each device, while the use of the device is constrained by feed mois-
ture content and feed size consist requirements discussed above.
Also based upon the specific-equipment data base were
generalized sizing curves. An example, for hydrocyclones, is shown
in Figure 3. This family of curves is useful in specifying the nominal
equipment size required for a given throughput.
Dewatering and Drying Equipment Costs
Generalized purchase cost curves for fine coal dewatering and
drying equipment were also prepared from the specific-equipment data
base. The purchase cost is displayed in Figure 4, as an example (for
hydrocyclones), as a function of the equipment size. The combined
467
-------
use of the sizing curves and the purchase cost curves enables a
purchase cost to be estimated for a given device and for a given
throughput.
Plant capital costs (which include equipment, buildings
and structures, piping, electrical, erection, engineering, and
contractors fees) were estimated as 3.2 times the dewatering
equipment capital costs. Although individual items of equipment
might have snorter useful, lifetimes, the annual capital recovery
costs were calculated on the basis of a 15-year plant amortization,
and on the basis of a 10 per cent interest rate.
Operating and Maintenance Costs for Dewatering and Drying
Direct operating costs (including direct labor, electrical
power, heat, materials, and supplies) for each type of device
are listed in Table 2. These are costs per ton of dry coal throughput,
except for refuse ponding and disposal costs, which are per ton of
wet refuse, and except for the direct thermal dryer operating costs,
which are per ton of water evaporated. Also in Table 2 are annual
maintenance costs and annual indirect operating'costs (taxes,
insurance, supervisory, and administrative costs), which are estimated
as a percentage of the total plant capital costs.
468
-------
Additional Cost: Elements
Transportation costs are based upon the total weight shipped,
including moisture. In addition, the coal cleaning plant operator pays
approximately 74 cents for the Pension and Benefit Trust Fund for each
ton (including moisture) of coal shipped.
The coal user, e.g., an electric utility, is assumed to con-
tract for net heating value. For the purposes of analyzing dewatering1
and drying operations only, a constant heating value (of 13,650 Btu
per pound) of dry coal is assumed, since no appreciable change in coal
composition results from these operations. However, any associated
moisture in the coal received is, for the purposes of this study, pena-
lized by the requirement for sufficient additional coal to vaporize this
moisture. This additional coal penalty is the total cost of such
coal through mining and the entire cleaning plant benef iciation process
including separation, dewatering and drying, and refuse disposal; and
is assumed to be $20 per ton (dry basis). In addition, a power-plant
pulverization cost of 60 cents per wet ton is assessed to the additional
coal requirement.
Outline of Case Study
The first case study conducted to compare the costs of alterna-
tive dewatering and drying schemes with the economic benefits of
shipping dried coal is presented in this paper. A 580 MW electric
469
-------
utility in Montgomery County, Maryland has a net coal heating value
requirement of 35.5 x .1012 Btu per year. It is supplied with cleaned
coal from the Pittsburgh coal bed in Marion County, West Virginia, with
a heating value (dry basis) of 13,650 Btu per pound. Rail transporta-
tion for the 245-mile distance is at a cost of $6.23 per ton (includ-
ing moisture).
The coal preparation plant has a nominal raw coal throughput
of 500 tons per hour (tph), and operates 13 hours per day for 254
days per year. The 6-inch top size coal is screened, yielding 250
tph of 6 x 3/8 coal which is washed and dewatered, and 200 tph cleaned
coarse coal are recovered at 3.4 per cent surface moisture (equivalent
to 207.2 wet tons per hour). The fine coal circuit has a feed of 250
tph of 3/8 x 0 coal. The heavy media cyclone and froth flotation cir-
cuit has two clean coal streams: 160 tph of 3/8 x 28 M coal at 12
per cent surface moisture, and a slurry of 35 tph of 28 M x 0 with 69.9
per cent water. Each of these two fine coal streams is then dewatered
and dried in alternate ways, yielding a fine coal product with different
moisture contents. The fine coal product is then blended with the
cleaned coarse coal for shipment to the consumer.
470
-------
Based upon the capital and operating costs presented earlier
for fine coal dewatering and drying, the annual costs for the several
alternatives were evaluated for the nominal 500 tph plant. These
annual costs were then adjusted to account for the differing quan-
tities of coal which must be shipped (depending upon the final moisture
content) to supply the utility's net heat requirement.
Dewatering and Drying Alternatives
Seven alternative schemes for fine coal dewatering and dry-
ing were defined in this case study. Case O is the baseline case
in which no dewatering or drying is performed on the fine coal.
By definition, the cleaning plant capital, operating, and mainten-
ance costs attributable to fine coal dewatering and drying are
zero for Case O. However, the transportation and power-plant penalty
costs would be highest for Case 0, since the largest quantity of water
is also shipped. For Cases A through F, where water is removed
from the fine coal, the capital, operating, and maintenance costs
of water removal are greater than zero. For these cases, the
transportation and power-plant penalties are less than for Case 0,
and the economic benefits of water removal are evaluated as the
differential from Case 0 penalties.
471
-------
Case A is straightforward mechanical dewatering (which is
also performed in Cases B-E): the 3/8 x 28 M stream is centrifuged,
and the 28 M x 0 stream is filtered. More moisture removal is per-
formed in Cases B and C by drying the filter cake of 28 M x 0 coal:
in Case B, with an indirect heat exchanger; and in Case C, this
filter cake is dried in a direct thermal dryer. The objective in
Cases D and E is not further moisture removal than Case A, but it
is the recovery of fine coal from the centrate in the 3/8 x 28 M
circuit. In Case D, the centrate is processed in a hydrocyclone
for slimes removal, with the fine coal fraction then filtered,
and with the filter cake added to the fine coal product. In Case E,
slimes removal prior to vacuum filtration is performed by a flotation
cell rather than by a hydrocyclone.
Case F is intended to perform additional moisture removal
without solids loss from a centrifuge. In this case, the 3/8 x 28 M
stream is not dewatered in a centrifuge, but instead is directly com-
bined with the filter cake from the dewatered (in a vacuum filter)
28 M x 0 coal. The combined 3/8 x O coal is then dried in a
direct thermal dryer.
472
-------
The nominal coal and moisture quantities for each of the
seven cases are listed in Table 3. These quantities are consistent
with the 500 tph feed rate to the coal cleaning plant, and are the
basis of the dewatering and drying cost calculations. These quantities
(and subsequently, the costs) are adjusted for the power-plant penalty
of vaporizing moisture, on the basis of 3.2 x 10s Btu needed per ton
of water; and for the solids losses in the dewatering and drying
operations. The adjustment calculations are shown in Table 4.
Posts and Benefits for Dewatering and Drying Alternatives
Using the unit process capital and operating cost data,
annual costs for fane coal dewatering and drying, for each of the
seven alternatives, were calculated. These are listed in Table 5.
Table 5 also shows the total annual costs associated with
delivering the clean coal product to the utility consumer. In addition
to the trust fund and transportation costs, the utility pulverization
costs and the utility penalty for contained moisture are included.
These costs include the economic penalties for not dewatering and
drying the fine coal. Conversely, the total benefits from fine coal
dewatering and drying are the cost savings compared to the baseline
473
-------
case (Case 0), where no moisture removal occurred. The net annual
benefit is defined as the total benefit less the total processing
costs.
Table 6 is a summary of the costs and benefits of the fine
coal dewatering and drying alternatives, expressed per dry ton of
fine coal product. It must be emphasized that the operations listed
for the alternatives were performed on partial fine coal streams, so
that the costs do not represent cumulative costs for sequential
operations on the same stream.
Conclusions for this Case Study
For this case study, limited to one user/producer scenario,
all of the fine coal dewatering and drying alternatives show significant
benefits compared to the baseline case of no dewatering. It is instruc-
tive to compare the net benefits to those of Case A ($3.10 per ton),
which is limited to mechanical dewatering processes. Case B, in which
the filter cake is dried in an indirect heat exchanger, is only margin-
ally more attractive. Case C, where a direct thermal dryer is used,
is significantly less attractive than Case A. In Cases D and E, the
recovery of solids from the centrate appears attractive, reflecting
lower refuse disposal costs as well as recovered product values. The
use of a thermal dryer in Case F to avoid centrate solids losses is
apparently competitive with Cases D and E.
474
-------
Conclusions for Other Case Studies
The conclusions reached for other case studies may well
be quite different from those above. It is apparent from Table 5
that of all the additional cost elements from which the benefits are
derived, transportation costs are dominant. To assess the impact
of other user/producer distances, two additional scenarios were
examined. In one extreme scenario, the utility is adjacent to the
mine and to the coal preparation plant, so that transportation costs
are zero. In the second extreme scenario representing much longer-
distance hauling, transportation costs were assumed to be three times
those of Table 5.
The results of these tsao scenarios are summarized in Table 7,
along with those of the first case study. With much higher trans-
portation costs, alternatives B and F (which utilize thermal drying)
begin to appear more attractive. Conversely, with no transportation
costs, the thermal drying alternatives show negative benefits: e.g.,
these processing costs are excessive.
Based upon the few scenarios shown in Table 7, several pre-
liminary overall conclusions may be drawn. First, the net benefits
achieved in Case A, by straightforward mechanical dewatering, are not
lower by large amounts than those achieved by more complex (and more
capital-intensive) dewatering and drying schemes. Second, the net
benefits achieved in Cases D and E, by recovery of centrate solids,
appear attractive and are achieved with only moderate increases in
475
-------
capital investment as compared to Case A.
The early results presented in this paper were achieved
without testing other parameters for their effects. One such para-
meter which could change the conclusions is the amortization time
for capital investment: a significantly snorter period than the
15 years assumed in the early case study needs evaluation. Another
parameter not yet evaluated is a change in the basic coal prepara-
tion plant flow sheet, with different coarse coal/fine coal ratios
and with different moisture contents of the wet fine coal streams
prior to dewatering.
476
-------
TABLE 1
FINE COAL DEWATERING AND DRYING EQUIPMENT
EQUIPMENT TYPE
VIBRATING SCREEN
STATIONARY SCREEN
CENTRIFUGE
HYDROCYCLONE*
FLOTATION CELL*
VACUUM FILTER
STATIC THICKENER
DIRECT THERMAL DRIER
INDIRECT THERMAL DRIER
FEED
MOISTURE, I
40-90
60-90
60-80
85-90
95
65-75
85-95
12-15
20-25
„ PRODUCT „
MOISTURE, %
12-16
40-50
12-20
40-60
70
20-30
60-70
6-7,5
8-14
TYPICAL
SOLIDS „
RECOVERY, %
80
80
90
50
90
99+
99+
99+
99+
*USEFUL FOR SLIMES REMOVAL PRIOR TO FILTRATION,
477
-------
TABLE 2
OPERATING AND MAINTENANCE COSTS
DIRECT OPERATING COSTS:
DISC VACUUM FILTER $0,70/TON
DRUM VACUUM FILTER 0,80/TON
STATIC THICKENER 0,35/TON
SCREEN & SOLID BOWL CENTRIFUGE 0,22/TON
HYDRAULIC CYCLONE O.OVTON
FLOTATION CELL 0,20/TON
INDIRECT HEAT EXCHANGER o'28/TON
DIRECT THERMAL DRIER 4,00/TON
REFUSE PONDING & DISPOSAL LOO/TON
DRY COAL
DRY COAL
DRY COAL
DRY COAL
DRY COAL
DRY COAL
DRY COAL
DRY COAL
WATER
TOTAL REFUSE
INDIECT OPERATING COSTS AND MAINTENANCE (Per, OF INSTALLED CAPITAL COSTS)
TAXES 2,0%
INSURANCE
SUPERVISORY AND ADMINISTRATIVE COSTS
MAINTENANCE COSTS
1,5%
5,0%
TOTAL
9,5%
478
-------
TABl£3
COflL AFP fPISIURE QuMTITIES
BASIS: 500 TPH RAW COAL FEED
CASE
CLEAN COARSE COAL (6 x 3/8) :
DRY TPH
MOISTURE, GPM
MOISTURE, per.
WET TPH
DRY TPH
MOISTURE, GPM
MOISTURE, per,
WET TPH
TOTAL CLEAN COAL RHDUCT (6 x 0):
DRY TPH
MOISTURE, GPM
MOISTURE, per.
WET TPH
JtFUSE_ (AFTER THICI^NING):
DRY TPH
WET TPH (a 3K SOLIDS)
SOLIDS Loss TO ATMOSPHERE:
DRY TPH
0
200
28.5
3.4
207.2
195
412
34.6
298.1
395
¥0.5
21.8
505.3
0
0
0
A
200
28.5
3.4
207.2
187
91.6
10.95
210
387
120.1
7.2
417.2
8
26.7
0
B
200
28.5
3.4
207.2
136.6
60.8
7.5
201.8
386.6
89.3
5.5
109.0
8
26.7
0.4
C
200
28.5
3.4
207.2
186.6
54.4
6.8
200.2
386.6
82.9
5.1
407.4
8
26.7
0.4
D
200
28.5
3.4
207.2
191.3
94.8
11.0
215.1
391.3
123.3
7.3
422.3
3.7
12.3
0
E
200
28.5
3.4
207.2
194.2
101.2
11.5
219.6
394.2
129.7
7.6
126.8
0.8
2.7
0
F
200
28.5
3,4
207,2
193,0
19.2
6.0
205,3
393,0
77,7
4,7
412.5
0
0
2.0
-------
TABLED
OF COAL AND MOISTURE QUANTITIES
BASIS: 3,302 OPERATING HOURS PER YEAR
13,650 Bru PER POUND (DRY BASIS) HEATING VALUE
3,2 x ID6 Bru PER TON WATER POWER PLANT PENALTY
35,5 x ID12 Bru PER YEAR NET POWER PLANT REQUIREMENT
CASE
UNADJUSTED QUANTITIES:
WETTPH OF PRODUCT
WET IB6 TPY OF PRODUCT
DRYTPH OF PRODUCT
DRY ]D6 TPY OF PRODUCT
GROSS 1012 BTU/YEAR
MOISTURE, TPH
MOISTURE, 10s TPY
MOISTURE PENALTY, ID12 BTU/YEAR
NET ID12 BTU/YEAR
ADJUSTMENT FACTOR = REQ'D BTU/NET Bru
ADJUSTED QUANTITIES:
WET 3D6 TPY OF PRODUCT
DRY 10s TPY OF PRODUCT
ADDITIONAL REQUIRED QUANTITIES:
WET ID3 TPY OF PRODUCT
DRY 3D3 TPY OF PRODUCT
0
505,3
1,6685
395
1,3043
35,61
110.3
364.2
1.17
34,44
1,0308
1,7199
1,3445
51,4
40,2
A
417,1
1,3776
387
1.2779
34.89
30.2
99.7
0,32
34,57
1,0269
1,4147
1,3123
37,1
34,4
B
409.0
1.3505
386,6
1.2766
34,85
22.4
74.0
0.24
34.61
1.0257
1.38S
1.3094
34.7
32.8
C
W.4
1,3452
386.6
1.2766
34.85
20,8
68,7
0,22
34,63
1.0251
1.3790
1.3086
33,8
2,0
D
422.3
1.3944
391.3
1.2921
35.27
31.0
102.4
0.33
34.94
1.0159
1,4166
1.3126
22.2
20.5
E
426.8
1.4093
394,2
1.3016
35.53
32.6
107,6
0,34
35,39
1,0088
1,4217
1,3331
32,4
11,5
F
432.5
1.3621
393.0
1.2977
35.43
19.5
64.4
0.2L
35.22
1.0080
1.3730
1.3081
10.9
10.4
00
o
-------
TABLES
COSTS AND
(THOUSAND DOLLARS)
CASE
EQUIPMENT PURCHASE COSTS
PLANT INVESTMENT COST
AWUAL CAPITAL RECOVERY COST
ANNUAL DIRECT OPERATING COST
/torn. INDIRECT 08M COST
&NUAL REFUSE DISPOSAL COST
TOTAL ANNUAL PROCESSING COST
AWUAL TRUST FUND COST
ANNUAL TRANSPORTATION COST
AWUAL UTILITY PULVERIZ, COST
ANNUAL UTILITY MOISTURE COST
TOTAL ANNUAL ADDITIONAL COST
TOTAL ANNUAL BENEFIT*
NET ANNUAL BENEFIT
0
0
0
0
0
0
0
0
3273
10715
J032
804
13824
0
0
A
316
Ml
130
197
96
88
511
m
8814
849
688
11398
2426
1915
B
616
1971
254
229
137
88
758
1025
8630
831
656
11142
2682
1924
C
1116
3571
460
320
339
88
1207
1020
8591
827
640
11D78
2746
1539
D
425
1361
176
210
129
41
556
1048
8825
850
410
11133
2691
2135
E
392
1253
162
221
119
9
511
1052
8857
853
230
10992
2832
2321
F
992
3174
409
357
302
0
1068
1016
8554
824
208
10602
3222
2154
JS
CO
"BENEFIT is DEFINED AS COST SAVING COMPARED TO BASE CASE 0,
-------
TABL£6
COSTS AND BENEFITS PER DRY TON OF FINE COAL PRODUCT
NOTE: OPERATIONS LISTED WERE PERFORMED ON PARTIAL STREAMS
CASE
0
A
B
C
D
E
F
FINE COAL
DEWATERING & DRYING OPERATIONS
NONE
CENTRIFUGATION, FILTRATION
CENT,., FILL, IND, Hr, EXCH,
CENT,,, FILT,, DIRECT THERM, DRYER
CENT,, FILT,, HYDROCYCLONE, FILT,
CENT,, FILT,, FLOTATION, FILT,
FILT,, DIRECT THERMAL DRYER
(XJSI
PER TON
$0,00
0,83
1,23
1,96
0,88
0,80
1,68
PER TON
$0,00
3,93
4,35
4.46
4,26
4,42
5,06
PERlON
$0,00
3,10
3,12
2,50
3,38
3,62
3,38
482
-------
TABLE;
SENSITIVITY OF BENEFITS TO TRANSPORTATION COSTS
CASE
0
A
B
C
D
E
F
FINE COAL
DEWATERING & DRYING OPERATIONS
NONE
CENTRIFUGATION, FILTRATION
CENT,, FILT,, IND, Hr, EXCH,
CENT,,, FILT,, DIRECT THERM, DRYER
CENT,, FILT,, HYDROCYCLONE, FILT,
CENT,, FILT,, FLOTATION, FILT,
FILT,, DIRECT THERMAL DRYER
NET BENEFIT PER TON
OF FINE COAL PRODUCT
$6,23/ToN
$0,00
3,10
3,12
2,50
3,38
3,62
3,38
$18,69/ToN
$0,00
9,27
9,89
9,39
9,36
9,42
10,17
$0,00/ToN
$0,00
0,02
-0,26
-0,95
0,39
0,72
-0,01
483
-------
0,25
0,20
0,15
CO
0,10
0,05
0
FIGUE1
FDR FINE COAL PFV/ATFRINr.
0
ID 20 30
MOISTURE CONTENT, PER CENT
484
-------
FIGURE 2
CENTRIFUGE PERFORMANCE FOR FINE COAL DEWA1ERING
0,10
0,08
0,06
00
0
O.CXi
0,02
0
10 15 20
MOISTURE CONTENT, PER CENT
25
30
-------
FIGURE 3
SIZING OF HYDRAULIC CYCLONES
,.—
i
DIAMETER, INCHES
60 80 100
-------
FIGURED
PURCHASE COST OF HYDRAULIC CYCLONES
10,000
8,000
6,000
4,000
? mn
ouuu
1 000
X
-------
HOMER CITY COAL CLEANING DEMONSTRATION,
TEST, AND TECHNOLOGY EVALUATION PROGRAM
James H. Tice
Pennsylvania Electric Company
1001 Broad Street
Johnstown, Pennsylvania 15907
ABSTRACT
A 1,200-ton-per-hour coal cleaning facility at Homer City Station is
being intensively tested as a means of evaluating the impacts of coal cleaning
environmentally, economically, and operationally. The full-scale beneficiation
of coal through this heavy media process will afford a ready comparison between
front-end cleaning and on-line gas scrubbing as competing means of emission
control.
The primary objective of the Homer City testing program is that of
assessing the performance of the cleaning systems and comparing the actual
performance with that originally projected. Further evaluations are not
meaningful unless proper plant performance is assured. Several other high
priority objectives are those of assessing the impacts of the integrated
generating plant on the environment, boiler operation, and cost areas.
Secondary tests will assess the energy efficiency, total environmental
effluent discharge, and availability of coal cleaning in comparison with flue
gas desulfurization systems.
Preliminary calculations based on design data for the Homer City Cleaning
Plant and available operational data from existing flue gas desulfurization
systems will be tabulated for comparison.
The objectives of the Homer City evaluation are as follows.
A. Demonstrate the application of coal cleaning as a means of
emission control.
Homer City power generation complex is composed of three fossil-fueled units
located near the town of Indiana in Indiana County, Pennsylvania. Unit No. 3
is required to meet New Source Performance Standards while Units No. 1 and
No. 2 must comply with the State Implementation Plan. A coal cleaning approach
has been taken to clean captively mixed coal with 2.8 percent sulfur to meet
both standards while wasting only 5 to 6 percent of the available coal energy.
The Homer City facility is owned jointly by the New York State Electric & Gas
Company and the Pennsylvania Electric Company.
488
-------
B. Assess the performance and operating characteristics of several
types of heavy media equipment now available and in operation
at Homer City with suggestions for improved design or operation.
C. Evaluate the full costs of coal cleaning as an emission control—
environmentally, economically, and in terms of total system
energy consumption.
D. Introduce those segments of present coal cleaning applications
which require additional research or development work.
E. Assess the "secondary" effects of coal cleaning in boiler operation,
electrostatic precipitator performance, solid and liquid effluent
control, and total energy use.
F. Employ advance systems of control and measurement to monitor the
output of the coal cleaning process and permit a high degree of
quality control.
G. Utilize developing systems of mathematical process modeling to aid
in coal procurement, process control, and operation diagnosis.
489
-------
The Homer City Coal Cleaning Plant is the first demonstration of coal
cleaning to meet Federal New Source Performance Standards for a full-sized
fossil fuel unit. The adaptation of coal cleaning at Homer City has at-
tracted a growing interest from both regulatory and research-oriented concerns.
Following is a brief description of the Homer City facility, and of the
series of interactive tests and evaluations which have been planned to
satisfy the industry's need for information.
The Homer City Generating Complex was originally constructed as two, 600
megawatt coal fired units in 19^9» operating primarily on fuel which is
mined at two captive mines on site. This generating station, located in
Indiana County, Pennsylvania, is equally share-owned by the Pennsylvania
Electric Company (Fenelec), and the New York State Electric and Gas Corpora-
tion (NYSEG). The station as originally operated could not achieve compli-
ance with the Pennsylvania SOp regulations which were imposed in March of 1972,
using run-of-mine coal. when a new 650 megawatt addition was proposed for the
site, an assessment of available sulfur oxide control techniques was begun
with all three units in mind.
The new 650 megawatt Unit #3 proposed for the station at Homer City,
as well as the existing units, was in need of some method of sulfur removal
in order to comply with the Federal New Source Performance Standards
490
-------
(NSPS) and the State Implementation Plan (SIP) respectively. These
factors weighed heavily in favor of the development of an integrated,
multi-level coal cleaning system for the site. Several systems
were evaluated after a detailed washability of the captive reserves was
developed. The design which was selected utilized a broad spectrum of
conventionally applied coal cleaning equipment, working to its best
advantage on a preprocessed feedstock. This integrated system of coal
processing came to be known as the Multi-Stream Coal Cleaning System or
MCCS.
Coal cleaning was chosen as the only means of SCL control for the
Homer City site after a comparative evaluation of flue gas scrubbing and
MCCS coal cleaning projected some very tangible economic benefits which
could result from the successful operation of the highly specialized MCCS.
The cost comparison for both alternatives at Homer City is shown as
Figure 1.
The Heyl and Patterson Company (H&P) proposed to construct the
developed MCCS system on the Homer City site as two independently operating
600 ton per hour (raw coal) circuits. Their final design recognized all of
the limitations and capabilities of existing coal processing equipment,
while extending those capabilities through further refinement of conven-
tionally used systems. The key element in producing one-third of the coal
output at a quality sufficient to meet Federal NSPS is the H&P heavy media
cyclone operating at an effective separating gravity of 1.30 grams per
milliliter on specially sized and classified feedstock from the run-of-mine
coal. The balance of plant output will be recovered at a quality to meet
the Pennsylvania State Implementation Flan levels of ii.O Ibs. of SOg per
491
-------
million BTU's of "boiler heat input. The recovery efficiency of this inte-
grated approach to ooal cleaning has been projected to "be almost 95 percent.
That is, approximately 95 percent of the heating value of the raw coal will
be recovered as fuel for the Homer City units.
Heyl and Patterson has recently completed construction of one, 600
TPH circuit to provide these specialized coals for the operation of all
three Homer City generating units. Problems have been experienced in
operating the new plant, complicated by simultaneous construction work
on the second 600 TPH (raw coal) circuit, and the recent United Mine
Workers strike. Start-up is proceeding in an orderly manner, with in-
creased levels of plant operating capacity being met each week. Some
ooal has been produced by the plant which meets the NSPS, but quantities
have been limited to date.
Simultaneously with the construction of the Homer City Coal Cleaning
Plant, a multi-faceted plan was drawn up by the Homer City Owners to test
and evaluate the concept of emission control by coal cleaning in general
and the performance of the Homer City cleaning circuits in particular. The
original objectives for the test series were amended when both the United
States Environmental Protection Agency and the Electric Power Research
Institute expressed strong interest in the conceptual evaluation. The
test plan which has evolved will meet the program objectives of all of the
participating agencies, while moving to develop improvements in ooal clean-
ing technology applicable to the Homer City and most conventionally
constructed cleaning plants.
The moat complex aeries of tests will break each one of the eight
size/gravity circuits out of the plant and evaluate each against its design
492
-------
criteria in terms of feedstock characteristics, product characteristics,
circuit throughput, and cost of operation. This series of tests serves
as a foundation for many of the subsequent determinations by assuring the
participants that the plant is operating near design conditions and pro-
ducing clean coals of acceptable quality for further work. These first
test series will also help to optimize the performance of the individual
circuits in the plant and further characterize equipment performance when
each type operates with a closely controlled size and quality classified
feed as designed. The MCCS will also be tested to insure compliance with
equipment guarantees from the manufacturer and steady state of operation
at full plant capacity.
In the course of these tests, ancillary plant equipment will also be
tested to assure that environmental safeguards are operating correctly.
The Homer City MCCS was designed as a zero discharge facility with respect
to liquid effluents. This zero discharge concept, and the performance of
low head particulate scrubbers which control emissions from the four ther-
mal coal dryers, will be thoroughly evaluated to gauge applicability on
future coal cleaning installations. There is, of course, a strong inter-
action between the operation of the plant circuits at full capacity and
the successful use of these control devices.
In drawing the program for detailed circuit testing, it was found that
systems for accurately measuring the flow rates and for obtaining a repre-
sentative sample of a three-phase coal, magnitite, and water slurry
flowing in an unsteady state were not well defined. The need for accuracy
in the circuit tests led to the adaptation of two initial test series to
try to select the best system of flow measurement and flow sampling. In
493
-------
our test plan, these axe called "methodology development" and "slurry
sampling systems.11 Some evaluations of slurry samples have been done by
Versar, Incorporated, under a plan of their development and a further
series of trials will be made when large size slurry samplers are installed
in the operating plant.
Inquiries by the test committees have revealed that several systems
of flow measurement using either sonic, magnetic, nuclear absorption, or
some combination of these methods are available and have been applied to
slurry systems. Several of the most promising types of devices are being
purchased for comparative evaluation in the Homer City circuits. A
specialized flow loop is now envisioned to evaluate a number of these de-
vices side by side; with the ability to divert the stream over a weir
arrangement for a control flow indication. Some of these flow measuring
devices are portable, to easily monitor a number of streams in succession
to aid in balancing flows to parallel equipment. Early use of these flow
sensors has indicated that the division of slurry flows in an equipment
circuit is not an easy job, but should be approached with some specialized
design work for a successful split.
Because of the NCOS's strong reliance on the heavy media cyclone to
perform coal beneficiation, a program is now in process at the U. S.
Bureau of Mines facility in Bruceton, Pennsylvania, to optimize heavy
media cyclone performance by finding the best combination of operating
parameters. The operating criteria being assessed are media to coal
ratio, cyclone operating pressure, inlet and outlet orifice sizes, and
magnetite size characteristics. A series of eighty-one separate runs are
494
-------
being made with a 6-inch pilot scale cyclone at the Bureau, which will
gauge the effect of varying operation on the gravimetric separation of a
tightly controlled feedstock. Data developed at the Bureau will be com-
pared with similar data collected during the MCCS tests to find the
correct scaling factor for the 6-inch Bureau cyclone installation.
Heavy media cyclones used at the Homer City coal cleaning installation
are designed to perform a reasonably sharp gravimetric separation on coals
down to 100 mesh in size. The cyclone loop at the Bureau of Mines will be
instrumental in defining the performance of fine coal separation and in
assessing the detrimental effects of quantities of misplaced fine coal on
the sharpness of separation. This pilot scale operating loop affords to
the Homer City Owners a means of quickly determining the causes of operating
effects observed in the course of operating-the full-size plant cyclones.
This program also stands alone to help optimize- cyclone design and operation
in future coal cleaning systems on an industry-wide basis.
In designing the Homer City MCCS, questions arose concerning the
instruments that measure and control the many streams within the coal wash-
ing plant. A program has been initiated to assess and suggest improvements
to existing specific gravity controls on the heavy media circuits to limit
the variation in media'gravity during operation. This control becomes
extremely important when operating at low media gravities where a small
variation can drastically affect the quantity or quality recovery of deep
cleaned coal products. This task is tied to the assessment of test Instru-
mentation discussed earlier, because an accurate determination of specific
gravity is an important component of the flow calculation.
495
-------
The design of the MCCS will "be fully successful only when tight
operational control over the circuits allows the plant to be tailored to
the varying feed coals quickly. Control instrument work at the U. S.
Bureau of Mines has led Peneleo to believe that tight control is only
possible if care is used in operating and maintaining the monitoring
equipment. These early tests have indicated that proper measurement ac-
curacy is available to maintain the slurry within acceptable limits if
calibrations are made frequently and in the proper manner.
An additional instrument development program will demonstrate and
test a non-destructive means of instantaneous coal analysis to enable the
MCCS plant performance to be closely monitored without analytical delays.
The method of analysis employed is related to neutron activation in the
mineral matter constituents of the coal. A-summing of these elemental
quantities is the basis for a determination of total ash, for which a BTtf
value can be assigned if the BTU/ash relationship for the mine has been
previously defined. For the Homer City coals, the BTU/ash characteristics
are well established by almost ten years of operation. Moisture content
may also be determined by using this nuclear analysis, but specialty chute
designs are necessary to optimize the irradiation while eliminating back-
ground scatter from other nearby materials. Successful use of several
on-line instruments will enable much better control of the complex circuits
by making plant operators aware when various coals are introduced into the
process.
The original decision of the Owners to use coal cleaning in lieu of
FGD was made based on an economic analysis. Because of the projected
496
-------
nature of data use, both for evolving FGD and for MCCS cleaning, a subse-
quent and more thorough economic analysis must be made to compare these
alternative systems using actual operating data. This comparison will
use information concerning the cost of sulfur removal for both alternative
methods, and will ease the burden of future comparisons within the industry.
Further analytical developments may pave the way for designing a combination
of the two methods to improve availability and cost effectiveness over
singly employed systems. An economic design of the future could use coal
cleaning to do an 80 or so percent removal job at low cost, while using
partial scrubbing with by-pass gas reheat to achieve an overall 90 percent
to 95 percent removal with good availability and less disposal problems.
A possible economic benefit which was not considered in the adaptation
of coal cleaning technology at Homer City is that the use of extremely low
ash and sulfur fuel may improve the performance of the downstream combustion
cycle to provide additional economies. The effects of burning extremely low
sulfur coal in the #3 boiler will be tested by the boiler manufacturer,
Babcock and Vilcox, using a run-of-mine grade of coal initially, and compar-
ing this to a similar series of tests made using deep cleaned coal later
on. Longer torn economic analysis will be made by comparing operating and
maintenance costs of Homer City #3 against those of other Feneleo operated
units. Theoretically, the low ash fuel should permit a high boiler effi-
ciency by reducing heat losses in the ash. Combustion should be more
complete, and slag concentration on the heat transfer surfaces should be
minimized. Ash erosion and slag fall damage in the steam generator should
be noticeably reduced.
497
-------
Future coal burning facilities using compliant low sulfur and ash
cleaned coal will have to struggle with the particulate control problem
as Homer City has. The Homer City Owners selected an extremely overdesigned
electrostatic precipitator to clean the low resistivity dust which is ex-
pected to be produced. One facet of the test work here is the characteriza-
tion of emitted dust from the control ESP's, including a particle size
analysis. A second trial will assess the magnitude of precipitator rapping
losses currently thought to be a major cause of dust carry-over. Testing
here will fully characterize ESP performance on deep cleaned coal ash, and
should enable a less costly design to be used for future units.
Environmental evaluations will be made within the area of influence
of the Homer City site to assess the total impact of coal cleaning on the
background. Several initial tests have been made by Battelle Memorial
Institute, and these will be compared with data taken after the cleaning
system is in full operation. Data from the long-term operation of Penelec's
ambient air monitoring system will also be factored Into this evaluation.
Homer City's total environmental impact should be less detrimental than that
of a non-regenerative scrubbing unit due to the more limited quantities of
solid and liquid effluents discharged.
Coal cleaning at Homer City could be improved by pretreating the plant
feedstock to separate the mineral matter from the coal by selective comminu-
tion or crushing. Certain methods of coal crushing may preferentially
liberate the minerally rich veins from the bulk of the carbonaceous material.
Once broken free they can be easily separated by the heavy media cyclones,
or other gravimetric separating systems.
498
-------
Manufacturers of various types of crushing systems have been contacted
and a test series has been devised to measure the effect of coal crushing
methods in the liberation of mineral matter. This test series will also
assess the developing technique of chemical treatment as a means of comminu-
tion to free coal's mineral matter constituents. A system which works best
on the captive Homer City coals may not be the optimum method for coals
from other geographic locations.
Detailed in-seam sampling by the U. S. Geological Survey within the
two Homer City reserves has shown that mineral matter tends to be very
concentrated in layers within the coal seams. True optimization of the
coal cleaning operation should thus begin in the mine at the working face
and proceed to work on segregated coal fractions from that point. This
approach is impractical today, but further developments in mining equipment
and techniques could someday afford the opportunity to take advantage of
this important first step in coal cleaning.
A cost consideration in the operation of a heavy media cleaning plant
is the loss of magnitite to the product coals and refuse. At Homer City,
the Owners have done preliminary studies which indicate that media-grade
magnitite should be available from fly ash collected in Units #1 and #2
electrostatic preoipitators as a product from combusting iron pyrite in
coal. The recovery and use of this resource at the cleaning plant could
result in a substantial savings in cleaning plant operating costs. Further
improvement in the separation efficiency of the heavy media systems could
also be a benefit if the reprocessed magnitite were specially sized, or if
larger quantities of low cost material were available to make-up the
otherwise uneconomical losses resulting from the use of an extremely fine
499
-------
grade magnetite. The experiments mentioned earlier at the Bureau of Mines
should provide a guide to the most efficient magnetite grade for separating
at low gravity.
The final program which comprises part of the Homer City development
effort is the further improvement of a mathematical model to describe
operations of the coal cleaning plant circuits. Battelle Memorial Institute
has recently issued an advanced coal cleaning system computer model under
the name, "CPSM-U," which was extensively tested on the Homer City plant
configuration. Use of this and other modeling techniques can permit more
accurate optimization of a coal cleaning plant design if the characteristics
of the feedstock are known. By replicating runs of the simulator and chang-
ing the proposed equipment configuration mathematically, a much improved
facility design can result which will take maximum advantage of the feed
coal characteristics to produce the most optimum recovery rates. This
work had previously been done as a laborious process which restricted the
number of possible trials and thus the degree of optimization.
At Homer City, CPSM-U will have an important place in frequently pre-
dicting the plant performance to give readouts of existing product quality
around eaoh of the circuit cleaning equipment types. This modeling concept,
when tied to a series of on-line analyzers, can afford a new dimension of
control in the operation of a coal cleaning facility. As the feedstock
undergoes a short term change, either in source or size consist, the appro-
priate data inputs to an advanced model could alert the plant operator to
control changes necessary to meet recovery objectives in one or more of the
product streams. A fully automated preparation plant of the future will use
500
-------
modeling as an instantaneous control device, just as the downstream power
plants have integrated and automated the control and monitoring functions.
The next step in improving Battelle's mathematical model will be the addi-
tion of cost information to the matrix, which allows the user to "both
qualitatively and economically evaluate the coal cleaning concept for any
projected use at any given time. Other phases of the Homer City test
program will gather information for input and development of a cost re-
fined version of Battelle's "CPSM-U," which can find industry-wide
application.
This total proposed program of testing and technology evaluation at
the Homer City Generating Complex will provide operational, economic,
environmental, end product end use data which will demonstrate the appli-
cability of coal cleaning to the utility industry, and highlight further
the necessary Improvements in that technology for use in future, advanced
cleaning facilities.
501
-------
FIGURE 1
HOMER CITY GENERATING STATION
ALTERNATIVE SO? CONTROL STRATEGIES
COST COMPARISON - COAL PREPARATION vs. FGD
Capital Investment for
SOg Control,- S Millions FGD MCOS
Coal Preparation Facilities
Original Plant (for use with FGD) 18 18
MCCS Addition 0 32
Sub-Total 77 50
Annual Revenue Requirements
for SO? Control - & Millions
Fixed Charges 11.6 7.5
Operating and Maintenance Expenses
FGD 10.6 0
Coal Preparation 3.2 7,6
Sub-Total 25.U 15.1
NOTES;
1. Capital Costs include provision for AFDC, escalation and
contingency (MCCS Contingency—
2. MCCS Capital includes provision for separate plant coal handling
system (6.0) and R&D support (1.0) plus associated AFDC, escalation
and contingency.
502
-------
COMPUTER CONTROL OF COAL PREPARATION PLANTS
Gerry Norton, George Hambleton, and Clive Longden
Norton-Hambleton Associates, Inc.
Ann Arbor, Michigan 48104
ABSTRACT
This paper discusses previous unsuccessful attempts by computer companies
to control coal preparation plant operations, while outlining the problems
involved. Questions as to the purpose and advantages of computerized plants
are posed and answered. What can be done at present in computer applications
in coal preparation plants? What will be possible in the future? Past,
present and future concepts are illustrated with reference to plants in
which Concol has played the major role. Three aspects of computerization
of coal preparation plants are discussed, namely process control, operations
control, and management. The required combination of hardware and software
are outlined in general with some reference to economics and process design.
503
-------
1. INTRODUCTION
Raw coal from mining operations requires upgrading prior
to utilization, to ensure a cleaner environment and for
the twin economic reasons of profitability and energy
conservation. To satisfy these requirements, a coal
preparation plant must operate efficiently, safely, with
a minimum of "downtime" and produce the maximum yield
of clean coal at a consistent specified quality. Experience
with manually operated plants has shown an average
utilization factor of only 80%, a safety record that has
only improved with legislation, and significant variations
in yield and quality of clean coal products with
consequential losses of energy.
What can the computer do to improve operation efficiency,
plant utilization, maintenance and overall profitability
thereby improving energy recovery?
2- THE OBJECTIVES AND SCOPE OF COMPUTER CONTROL
The computer can keep watch, or stand guard over the process
continually, and it can self-diagnose faults within it-
self and diagnose faults on items of plant contained
within the process system. In conventional plants about
85% of the downtime is used in fault diagnosis and only
15% in actually replacing faulty units. The computer
will ensure that the plant remains in an operable
condition, being controlled to a preferred set of
parameters. When an unacceptable deviation from these
parameters occurs the computer will either,
(a) correct the deviation checking it for
"normality", and/or
(b) inform the central control room operator, and/or
(c) shut down the necessary plant items as the
particular case requires, and control the
consequential effects on the process, or
(d) shut down the plant completely.
Computer control of coal preparation plants covers three
activities, the major activity being plant contro. The
504
-------
other two activities are ancillary to plant control and
depending upon economic, and/or market considerations
and can be included or excluded, as required. For a
complete system all three types are necessary, these
are:
(a) plant control,
(b) process evaluation,
(c) management.
2.1 Plant Control
The plant control computer has evolved as a means of
controlling all functions necessary for the efficient
and safe operation of the plant. The first consideration
is maximum reliability. Since computer reliability
decreases as the number of peripheral devices, for
example keyboards, display screens, magnetic disc drives,
etc., increase, the system must minimize the number
of such devices on the plant control machine. The main
areas of plant control can be identified as follows:
(a) sequence control of motors,
(b) sequence control of automatic valves,
(c) analogue control of plant process variables.
Operation of the control system should be basically simple
and logical, and require a minimum of specialized know-
ledge. Priority is given to the ease of maintenance
and fault diagnosis. The basic system should operate on
the principle that the plant, whatever its function,
should always be in an operable state. In order to
minimize plant downtime, the following attributes are
required in the system:
(a) maximum reliability and minimum physical
components,
(b) rapid tracing of the causes of failures and
subsequent replacement of faulty units,
(c) minimum training for operating personnel,
(d) flexibility to accommodate changes in, or
extensions to the existing plant.
A major aspect of computer control of coal preparation
plants is the provision of the necessary measuring and
sensing devices required for data input and the accuracy
and reliability of the measurements and signals. This is
important, since a computer is only as good as the data
it is given.
505
-------
The sensing carried out on all drives would be as follows:
(a) Starter Ready - this would include any desired
hard sequence, any automatic or manual switching,
main power available at the starter, and control
power available.
(b) Electrical Fault - this would include earth leak-
age^overload and main fuse failure.
(c) Auxiliary Contact - this would be a physical
electrical contact mounted directly to the motor
main, or control contactor.
(d) Proving Switch - this would be a sensing implant
that the drive., i.e., pump, cell, etc. is
carrying out its required duty.
As an example, a pump would have a flow switch as far
towards the end of its effect as possible,—say, the over-
flow from the cyclone would be a check that the main
cyclone feed pump is operational.
This function is most important, for example, it is not
merely sufficient to monitor that a pump is in fact
rotating—the line may be blocked, a valve may have failed
to open, a pipe may have in fact burst. It is, therefore,
of particular importance that the chain of events stemming
from pump operation be monitored as far down the line as
pssible.
(e) Motor Current - this is an analogue signal
derived directly from the motor starter current
transformer.
The above items are generally the minimum data requirements
for each drive, but others are sometimes either necessary
or requested by the customer, such as:
(f) Speed Switches - would normally be a two-level
speed sensing device. This device when mounted
on a belt conveyor for example, would be on a
non-driven member of the belt, such as the tail-
drum, and would be set up such that if the tail-
drum ran, say at 40% or less of normal operating
speed, the device would detect this and notify
the computer to stop feed to the cdnveyor. A
second speed level, at say, 90% of full speed,
could warn the operator of a possible fault, giving
time and opportunity for correcting action.
(Such as the reduction of plant feed and rate.)
506
-------
(g) Insulation Testing - is an independent routine
for the computer. When the plant is in a
stationary condition and the main drive power
removed from the motor control centers, the
computer would in turn test each drive for its
insulation value. Set points of the minimum
allowable insulation value would be retained
in the memory and any error would be recorded
and an alarm activated.
(h) Analog Measurements - such as density of separation,
tank levels, feed rate, etc. for the successful
control of coal preparation plant rely on the
accuracy and the repeatability of these measure-
ments. These anlog measurements are now well
proven, but some field measurements still
require improvement or development, e.g., ash
and sulfur coal, solids concentration in froth
flotation feed, pulp etc.
Sensing of items such as valve open/close status,
and blocked chute probes, will be required by
the computer for plant control.
2.2 Process Evaluation
Test work performed on samples of i aw coal provides
washability information which is used to predict set
points in a particular coal washing process, for example,
densities of separation. The plant will maintain a fixed
operating condition, unless a change is requested. The
comparison of washability data with actual plant performance
assumes accurate data until excessive deviations occur.
When these deviations are noticeable, renewed test work
will be performed for more current washability data and
the prediction of new set points. The coal wash plant
flowsheet model is set up from the washability data, so
that calculated outputs are fed to the plant control
function. The latter function also receives information
on anomalies, such that the model can be altered when
necessary.
The reliability of the plant algorithm is determined by
the compatability of input washability data in relation
to the actual washability of the coal flowing through
the plant at a given time. When dense media processes
aPply* an ongoing measure of predicted error can be made
through online determinations of ash contents and specific
gravities of media. There is no need to carry out
specific gravities of media. There is no need to carry out
507
-------
performance tests after those associated with commission-
ing since performance can be continually monitored.
Process evaluation by computer is not new, but its
application as an integral part of the plant control is
unique in coal preparation.
The washability input consists simply of the theoretical
ash: separating gravity (Dp) relationship for a given coal
being cleaned. Online ash monitors determine clean coal
quality, and the theoretical cutpoint Dpt corresponding
to that ash content, is compared with the actual cut-
point Dpa. EP's as calculated from prediction programs
for dense medium processes are related to the actual and
theoretical outpoints by the function.
Ep = f(Dpt-Dpa)
Where:
Ep = probable error
Dpt = theoretical cutpoint for actual ash
obtained
Dpa = actual cutpoint
Serious deviations from expected EP's would mean either:
(a) inefficient operation of the plant, or
(b) a significant change in washability characteristics
of the coal
2.3 Management Data
The management data function of the modern advanced control
system can be the basis for all planned and breakdown
maintenance functions for a plant. This can obviously be
greatly beneficial for the organization of the preventative
maintenance that will be required on a plant throughout
its operating life. A brief listing of some management
functions available on modern control systems are:
1. Maintenance Data Accumulation, i.e.:
(a) number of operational hours for each device
(b) number of operations of each device
(c) prediction of device or component failure
from historical data
2. Maintenance Instructions.
3. Maintenance Schedules and Inventory Control.
4. Mimic Displays to Facilitate Plant Operation.
5. Process Optimization by:
(a) memorizing and using washability data for
all seams being washed using Mayer curves,
and
(b) maximizing feed rate by ensuring that the
current load limitation factor is identified
508
-------
and fully loaded.
6. Stockpile Blending to Control the Consistency
of Ash and Sulfur Contents of the Final Blend.
7. Shift Logs and Reports of All Aspects of Plant
Operation That Have Been Monitored.
The management computer used to carry out the above listed
functions would differ from the plant control computer
only by the addition of bulk storage devices such as
magnetic discs, extra color visual display units and hard
copy printers. The interrelationship of these units is
illustrated in the flow diagram.
3. COMPUTER CONTROL DEVELOPMENT
Early ventures made into the automation and control of
coal preparation plants were on,ly partially successful,
and the conclusion from these early experiments showed a
need for:
(a) Robust and reliable equipment, both for control
and monitoring functions. Control equipment
available at the time was fragile and needed
to be protected from shock loadings, dust,
moisture, and extremes of temperature, i.e.,
all of those environmental factors that may
be found on almost every coal preparation
plant.
(b) Cooperation between design, computer, and
electrical engineers in the application of these
advanced techniques to coal preparation. It
became apparent that for successful application
of advanced control techniques, the selection,
design, and construction of the plant must be
carried out before the event rather than after-
ward.
The three projects described below incorporated changes in
control systems which allowed development to the present
level of capability.
3«1 The Fording Project
The first successful application of some of these techniques
to coal preparation was on a plant in Canada in the period
1968-1971. This plant was designed to produce metallurgical
coal for export and to process this coal at 1,000 tons per
hour, through a twin-stream plant; each stream having large
509
-------
coal heavy medium, small coal heavy medium, and fine
coal sections. Alternate methods of achieving desired
objectives in control were evaluated and costed.
Computers for control purposes were examined, but were
found to be expensive, fragile, and extremely sensitive
to environmental changes. As a result of this the .basic
control system choice was between pneumatics and elec-
tronics .
Pneumatics were cheap, established, and readily avail-
able, with a considerable amount of technical backup.
Electronics were found to be in the region of four
times as expensive as pneumatics and not in general use.
However, when specifications and operating life were
considered, the electronic units were far superior.
Consequently electronic elements with an orthodox but
comprehensive central control desk was the chosen
control system.
Two major areas of investigation involved in the design
exercise were:
(a) transmission elements
(b) final control elements
Transmission problems arose when such equipment was
applied to coal preparation plants, the elements being
fastened to equipment for which they were never designed.
Nuclear density gauges were used and once interfacing
had been completed between in-plant monitors and
processing devices, they proved most reliable. Final
process control elements gave rise to problems in two
main areas. One was the control of diverter head
boxes, known as "Elephant's Trunk" (ETS) systems. The
actuation of the ETS systems was finally carred out by
means of pneumatic cylinders operating against a spring.
The second main problem area was that of valve actuation,
which was difficult to solve at the time, because actuated
valves were normally used only for remote manual control,
and not automatic control involving the use of feed-back.
It became apparent that even though this level of control
was successful, more work had to be carried out on ways
of:
(a) reducing the amount of control equipment,
(b) finding quicker methods of isolating
problems within the equipment, and
(c) obtaining more reliable equipment
Achieving these objectives would then reduce the number of
service personnel required, and also lessen the necessity
510
-------
for such personnel to have specialized knowledge hence
an indication towards later philosophy in that it is
desirable to achieve self-diagnosing, high reliability
systems capable of being operated and maintained by
unqualified personnel.
The project was however, successful and the client was
satisfied with the result in terms of efficiency, con-
sistency of products, safety, minimum downtime and
running costs.
3.2 The Thurcroft Project
In 1973,'the British coal industry was expressing
considerable interest in the use of computers for
proposed new coal preparation plants.
During September 1974, the authors became involved in
the design and construction of the Thurcroft coal
preparation plant as chief executives of the main plant
contractor and of instrumentation and process control
subcontractor.
The plant comprises coarse coal heavy medium bath
(3 product), small coal heavy medium cyclones
(primary and secondary) and froth flotation for the
production of very high quality metallurgical coal
with the intermediate S.G. range material as steam coal.
Control hardware for the plant comprises Modicon I/O
(input/output) cabinets, three Modicon processors
(programable logic controllers [P.L.]) and a variety
of measuring and monitoring apparatus. Digital aspects
of control are handled by the P.L.C. units, analog
Items such as S.G.'s and levels are handled by
Foxboro. Spec. 2000 equipment interfaced with the Modicon
units as necessary for alarms etc.
The system utilizes basic programing techniques that can
be loaded via paper or magnetic tape into the main
processor memory.
A visual display unit is provided with the control system
for fault finding diagnosis.
Reprograming with this system via a standard terminal key-
board is simple and quick.
511
-------
For control purposes, the plant (including raw coal
handling and product outloading) is split into ten
groups on the control desk.
Each group is complete with "start", "hold", and
"stop", buttons, and each drive group within the group
is represented by its description, control reference
number and three Light Emitting Diodes (L.E.D.'s) as
below:
Green - drive ready to start
Amber - electrical stoppage—external, i.e.,
stop botton, blocked chute probe,
pull wire
Red - electrical fault—internal, i.e., overload
fuses or thermistor
When a group start is initiated, the pre-start alarm
times out, and then thr group hold light pulses at one
second intervals, starting each drive sequentially within
the group. When the group is running, the flashing
hold light is extinguished and the group stop light
illuminated.
Run-up time for the total' plant is 6-8 minutes, and the
plant may be started or stopped (excluding Group 1) by
the use of three push bottons:
Auto Start
Auto Hold, and
Auto Stop
The desk is complete with an L.E.D. illuminated "dark"
type mimic diagram, i.e., the mimic is dark both when
the plant is stationary and operating. Illumination
only occurring when a drive is in a starting condition
or when the process has determined a drive to be the
cause of a stoppage.
The drive L.E.D.'s maintained above are red. Green
L.E.D.'s are continuously lit to give valve position
indication.
A closed circuit TV system is installed, comprising two
fixed .cameras and four remotely operated units with zoom,
pan, and tilt facility to observe when required, important
transfer points, feed launders, etc.
512
-------
Analog measurements include suspension gravities, tank
levels, feed and product tonnages, filter bowl levels,
filter bowl levels, filter vacuum, and bunker levels.
Digital measurements include speed switches, tilt switchees,
float switches, blocked chute probes, capacitance probes,
door positions, and all automatic valve limits.
The plant was brought "on-stream" in November 1976,
and the comissioning and achievement of commercial
operation required only three (3) daysl Due to the
nature and mode of control employed, efficiency of
separation and consistency of products are excellent
and the plant has, after two years of operation, a
first-class safety record.
Although the plant does not encompass any management data
functions, it is fair to say that due to the various
warning "nodes incorporated in the control system, some
potentially serious incidents to proprietary equipment
have been avoided. In terms of running costs, the plant
has upturned much conventional thinking, and has to
date proved to be cheaper to operate than a conventional
jig plant. This in itself being something of a
revelation.
The user has been extremely satisfied by the operation
of this plant at all levels, and indeed many people
world-wide have visited the plant and as a result of
these visits many favorable reports have been received.
The plant requires five men for operation, including
the control room operator. Without the control system,
ten men would have been, required to give what would
be without a doubt, an inferior level of control.
This plant has a control system that is a measure of
design, process, and instrument capability at the time
it was built. It has also provided an indication of the
way forward in terms of development in design,
instrumentation and process control.
3.3. The South Kirkby Project
This is a 1,000 ton per hour twin-stream plant comprising
large coal heavy medium, small coal heavy medium, and
a common fines recovery circuit.
513
-------
The client's requirements were for a comprehensive
management data function in addition to the normal
plant control and process evaluation modes.
Programable logic controllers were again considered for
this project because of their proven reliability. How-
ever, because of the comparatively large size of the
plant, this would have required thirteen control
processors, plus three communication processors, to give
the same degree of control as given by the three units
employed at Thurcroft. From a reliability standpoint,
therefore, it can be appreciated that even though, the
previous application gave very high reliability, the
overall reliability of sixteen units needed on the new
project would have been lower. After further investigation,
it was found that high reliability could still be
maintained by using a much larger single computer, if
the configuration was kept in its simplest form. The
plant control machine then consisted of computer,
memory, color VDU for operator output, a standard
typewriter keyboard for the operator to input to the
computer, and finally a printer, for hard copies,
alarms and logs, etc. It should be noted, therefore,
that the machine used had no magnetic disc drives and
other such high density information storage devices
which are generally the cause of computer failure.
Because one computer carries out the work that would
have been performed by sixteen programable logic
controllers, and because, this single machine has no
bulk storage memory system, the software has to be
capable of running in the computer's own, comparatively
small, high speed, random access memory.
It was necessary to handle three thousand five hundred
items of digital information, and seven hundred and
fifty items of analogue information, and in addition
provide a comprehensive management data system. The
final configuration, therefore', was the selection of
two PDF11/34 computers, one operating as a plant control
machine, the other as a management data processor as
shown on the control system flow diagram.
To provide the comprehensive management data required,
it was necessary to equip the standard machine with
visual display units, disc drives, etc., to give the
machine the large memory capacity required for this
duty.
However, if the main plant control machine fails then a
mechanical disconnection of the main processor and a
514
-------
reconnection of the management data machine, disconnected
from its peripheral devices, will allow normal running
of the process plant. Management data services would be
suspended until the failed machine could be serviced.
A failure of the management data machine, although
suspending management functions, would in no way interfere
with plant operation.
It will be noted that the two machines are not in fact
electronically connected together as installed, in order
to prevent the possibility of interference or argument
between the two machines in the event of failure of
either unit. Such electrical connection could in fact
lead to serious consequences on the plant in the event
of machine failure. The connection between a given
machine and plant control is thus effected by a multi-pin
plug.
4. PROBLEM AREAS
We have attempted in this paper to show why advanced
control systems are necessary, in modern coal preparation,
and how we translated this need into an operating concept.
We hope we have shown with specific reference to three
plants, with which we have been involved, how such
systems have evolved, and been applied in a practical and
effective manner, compatible with our overall objective
in the optimization of benefits from the preparation
of coal.
It is however, also necessary to mention some problem
areas encountered with the application of computer
techniques to coal preparation. A plant operator is
capable of value judgments or in computer terms, analog.
The computer itself is incapable of making such decisions
and relies solely on the interpretation of digital or
"yes"/"no" information for its actions. Consequently,
instead of one analog measurement, several digital
signals may be required and a problem commonly encountered
is that of transducer reliability.
Solid state technology produces reliable sensing elements—
though if and when these elements fail, we are left
with something of a problem, i.e., does one merely tell
the computer to ignore a faulty transducer, or does one
keep several million dollars worth of machinery idle,
while awaiting the procurement and installation of a
transducer costing a few dollars? If the first policy is
adopted and plant management is ineffective, the situation
515
-------
can arise where/ over a period of time, many transducers
are rendered useless and yet the plant will continue to
run. This state is unsafe from an operator's view, and
no warning can be given of certain conditions arising on
a plant which may lead to expensive damage to equipment,
with consequential loss of revenue and excessive down-
time. The second policy is unattractive financially,
in terms of lost revenue and downtime.
We have given this problem considerable thought and have
arrived at a conclusion which, answers the problem
effectively, provided that adequate control is maintained
by plant management. Memory access is achieved, either
by (a) direct access to the memory units via the visual
display units, or (b) altering hardwired sequence in the
I/O cabinets.
Thus, particular transducers can be disabled directly
from the keyboard. It is however, of the utmost
importance that accurate records of such temporary
alterations are fully documented and presented immediately
to plant management, so that the necessary rectification/
replacement of the unit can be done at the earliest opportunity,
and transducer monitoring levels maintained on all items
throughout the plant. This system has in fact worked
well. The management data system will perform this
recording function as part of its duties, though the
responsibility for implementation of the work will rest
with the plant management. The management computer will
produce a record of all disablements on every shift log
until restoration takes place.
5. FUTURE SITUATION
Developments in solid state technology are outstripping
those in almost every other field to such an extent that
by the time a control system has been conceived, designed
and installed, it is in terms of hardware, virtually out-
dated .
Future developments in mining will see the complete mine
control system, and a central processor being responsible
for both the extraction, handling, preparation, and out-
loading of coal from a particular mine. This total concept
mining should result in increasing cost effectiveness of
the unit and the ability of that unit to be efficient,
and closely tied to market trends and fluctuations.
Technology is now available to produce control systems
considerably in advance of those discussed in this paper.
However, it is important that economics be closely
evaluated when considering the implementation of these
516
-------
systems. In an industrial environment it is generally
uneconomical to achieve a particular level of technical
excellence without justification.
Control systems and their practical implementation must
at all times be viewed objectively. We must not be led
by people simply wishing to market computers, or conversely,
not be hindered by stilted or conventional thinking. It
is a question of keeping in touch with developments,
technological trends, and costs, and having the right
system in terms of plant and control to offer a client
to optimize his requirements.
517
-------
00
F/ELD SENSING
-DIGITAL
FIELD CONTROL
-DIGITAL
MOTOR CONTROL
CENTER
FIELD SENSING
-ANALOG
FIELD CONTROL
-ANALOG
SPECIAL
SENSORS
MODICON
INTERFACE
POWER
SUPPLIES
MODICON
INTERFACE
POWER
SUPPLIES
RECORDER
P.D.P II -34
PROCESSOR -
CONTROL
COMPUTER
DESK CONTROL
KEYBOARD
&
COLOR
MONITOR
P.D.P 11-34
PROCESSOR
MANAGEMENT
&
STANDBY
CONTROL
COMPUTER
REMOTE
KEYBOARD &
PRINTER
— s
REMOTE COLOR
DISPLAY MONITOR
SYSTEM
CONSOLE
TWO DESK COLOR
MONITORS
DESK PRINTER
DESK
KEYBOARD
FLOW DIAGRAM OF A COMPLETE COMPUTER CONTROL SYSTEM
-------
PHYSICAL AND PHYSICOCHEMICAL REMOVAL
OF SULFUR FROM COAL
David H. Birlingmair and Ray W. Fisher
Ames Laboratory
Iowa State University
Ames, Iowa 50011
ABSTRACT
A coal preparation plant for evaluation of existing state of the art
processes and for developing new processes has been constructed on the Iowa
State University campus. The conventional portion of the plant can be operated
in the 20 to 70 TPH range while the advanced processes are in the hundreds of
pounds per hour range. The installation was funded by the State of Iowa and
the Department of Energy and is operated through the Energy and Mineral
Resources Research Institute.
Processing equipment in the plant circuit includes a heavy media separator,
concentration tables, hydrocyclones, froth flotation cells, oil agglomeration
equipment, dewatering screens, cyclones and filters, crushers, ball mills,
pelletizers, conveyors, and thickeners to provide a closed circuit.
Coal samples ranging from 500 to 8000 tons have been processed using
coals with sulfur contents of 2.5 to 8.75 percent. Using the conventional
coarse coal processing circuit, sulfur reductions average 35 percent with
ash reduction averaging 45 percent. The equipment for advanced fine coal
beneficiation has been installed and a heavy media cyclone circuit and an
extruder are being added.
A description of the processes with data obtained to date is presented
as well as an overview of related coal research projects.
519
-------
I. INTRODUCTION
The development of improved coal cleaning methods has been underway at
Iowa State University since 197^ when the Iowa Coal Project was established
with funds provided by the Iowa Legislature and administered by the Energy
and Mineral Resources Research Institute (EMRRI). The primary goals of the
effort center on:
1) Demonstration of existing coal cleaning techniques in optimum cir-
cuit configurations so as to permit reasonable estimates of their
cost and effectiveness.
2) Development and demonstration of new coal cleaning techniques to be
utilized in support of existing technologies and to minimize the en-
vironmental impact of coal cleaning and utilization.
3) Development of supporting programs in coal characterization and anal-
ytical technology for major, minor and trace element determinations
and to better understand the nature of coal and its associated min-
erals such that rational cleaning processes can be developed.
When this research effort was launched, a dual attack was made on high
sulfur coal. One approach was the construction of a coal preparation plant
capable of demonstrating the performance of selected existing methods for
cleaning high sulfur coal which would allow for an assessment of the economics
of the processes. A logical selection process chose as the first methods se-
lected for demonstration and evaluation those commercial methods which seemed
to have the highest benefit/cost ratio. However, these were gravity separa-
tion methods which only removed coarse refuse. Since most high sulfur coal
also contains finely disseminated pyrites, a second approach was to screen a
number of promising but largely undeveloped methods for removing impurities,
to select several methods for further development, and to proceed in develop-
ing these methods. The separation methods selected for development include
those based on froth flotation, oil agglomeration, hydrocyclones, heavy media
cyclones, and high gradient magnetic separation. All of these methods are
designed for cleaning fine size coal which is the size that must be cleaned
if finely disseminated pyrites are to be liberated and removed and physical
coal cleaning is to become effective in meeting some of the various existing
and proposed air quality standards. The development of these methods has
been uneven since funds were limited. While methods based on froth flotation
520
-------
and oil agglomeration have received extensive bench scale testing and labora-
tory development leading to the recent installation of a 500-1000 Ib/per hour
process demonstration unit at the Ames coal research preparation plant, the
other methods have received limited attention. However, since hydrocyclones
(installed June, 1978) and heavy media cyclones (to be installed Spring 1979)
do not lend themselves to small scale systems, these devices have been or are
being incorporated Into the Ames coal preparation plant and will be evaluated
on a demonstration or semi-industrial plant scale.
To improve the separation of coal and pyrites achieved by either froth
flotation or oil agglomeration, research has focused on chemical pretreatment
of coal fines to enhance the difference in surface properties of the two com-
ponents. In addition, various combinations of gravity separation, froth flo-
tation, oil agglomeration, and comminution methods have been tested on a bench
or demonstration scale to determine what extent these methods complement each
other.
II. IOWA STATE UNIVERSITY DEMONSTRATION MINE NO. 1
Before describing our preparation plant, I would like to mention another
aspect of our project - the source of some of the nations finest 8% sulfur
coal.
A demonstration mine was established on a 40 acre site to determine the
potential for reclaiming surfaced mined land for row crop production. Top-
soil, nonacid overburden and shale were stockpiled separately during the
mining operation and returned In their original layers and re-contoured Into
bench terraces. (Figure 1)
Originally the site was suitable only for pasture but now contains ap-
proximately 25 acres of land suitable for row-crop farming. Studies are
being continued to produce optimal conditions for agricultural production.
III. AMES RESEARCH COAL PREPARATION PLANT
A coal preparation plant has been built on the campus of Iowa State Uni-
versity to demonstrate various methods of cleaning coal on a larger scale.
521
-------
A building to house the plant (Figure 2) and the first section of the plant
to clean coarse and medium size coal were completed in 1976. This section
included a primary variable speed crusher, heavy media separator, wet con-
centration table, size separation and dewatering screens, and materials
handling equipment. Hydrocyclones for cleaning fine size coal and disc
filters for dewatering fines were added in June, 1978; heavy media cy-
clones for cleaning medium size coal will be added In 1979. All of this
equipment Is of a semi-industrial scale (approximately 70 TPH) and was funded
by the State of Iowa. Pilot plant circuitry to demonstrate our modified
froth flotation and oil agglomeration methods of benefIciating fine size coal
was installed In 1978 using funds provided by the Fossil Energy Division of
the U. S. Department of Energy. A palletizing circuit was also included at
that time. (Figure 3)
The present plant utilizes three processes to separate coal from Its Im-
purities. The coal is received as mined and is sized in an Impact mill. A
vibrating grizzly allows the minus ijr Inch to bypass the crusher. The crushed
coal Is held in a surge hopper and metered onto the raw coal conveyor. A
separating screen pre-wets the coal and separates the plus 3/8" material from
the minus 3/8" material.
The l£" x 3/8" coal is then fed to a cone-shaped heavy media vessel.
The clean coal is swept around the surface of the cone and flows over a weir
to a scalping screen where the media is returned to the vessel. The coal pro-
ceeds across a vibrating screen where any remaining media is washed from the
coal and is delivered to the clean coal conveyor and onto a stockpile.
The refuse which has sunk to the bottom of the cone is pumped to a scalp-
ing screen to remove the media and then across a vibrating screen where it Is
rinsed and ultimately delivered to a refuse stockpile.
The 3/8" x 0 coal Is delivered from the separating screen to a double-
deck concentration table. The coal is washed across the table to a clean
coal launder while the refuse passes over the end of the table to the
refuse launder. Both streams then pass over sieve bends and onto a parallel
bar vibrating screen for dewatering.
The plus 48 mesh cleaned coal and refuse pass onto their respective con-
veyors while the minus 48 mesh material, along with the process water flows
522
-------
to a fine coal circuit.
The first hydrocyclone is adjusted to permit only clean coal in the over-
flow while the underflow contains all of the refuse and some coal. The second
hydrocyclone increases the accuracy of separation by discharging only refuse
through the underflow while permitting only a minimal amount of refuse to be
discharged with the coal in the overflow.
Overflow streams, containing clean coal from both hydrocyclones, Is thick-
ened in dewatering cyclones and proceeds to a vacuum disc filter for further
water removal.
Underflow streams from the hydrocyclones are fed directly to a separate
vacuum disc filter for water removal. Dewatered clean coal and refuse are de-
livered to the output conveyors.
All process water streams eventually flow through a mechanical clarifler
where solids are removed with the aid of a flocculant. The water is then com-
pletely recycled with only enough makeup to replace the water lost with the
plant product.
A heavy media cyclone circuit will be added In 1979 and will be used as
an alternative to the wet concentrating table for cleaning medium size coal.
The addition of the circuit will not affect other principal features of the
plant, but will increase our evaluation capability.
The main plant has been used to demonstrate cleaning of large samples
(1000 T) of coal from seven Ipwa mines on an Industrial scale and to process
40,000 tons of coal from the Iowa State University Demonstration Mine. The
samples from the different mines contained from 2.5 to 8.7% total sulfur and
from 11.6 to 20.0% ash. (Figure 4) As a result of processing In the plant,
the total sulfur content was reduced an average of 35% with a range of 2k to
45% and the ash content was reduced an average of 45% with a range of 34 to
57% for the series of coals. Moreover the pyrltlc sulfur content was reduced
an average of 52% with a range of 37% to 70%. The average weight yield was
75% with a range of 66 to 80% and the average calorific yield was 84% with a
range of 74 to 96%. Since these results were obtained before they hydrocy-
clones and filters were installed, none of the -48 mesh coal was recovered
and therefore the yields were lower than would be obtained with the present
equipment.
523
-------
IV. FROTH FLOTATION AND OIL AGGLOMERATION DEMONSTRATION UNIT
A unit for demonstrating the froth flotation and oil agglomeration
methods of cleaning fine-size coal has been installed In the Ames coal prepa-
ration facility. (Figure 5 S- 6) This unit includes equipment for grinding
and chemically pretreating 1000 Ib. batches of coal and for continuously bene-
ficlatlng the pretreated coal by froth flotation or oil agglomeration at a
rate of 100-200 Ib/hour. The circuit also includes a means for pelletizlng
the beneficiated coal.
Coal fines from the Ames coal preparation plant are placed as an aqueous
slurry In either of two agitated tanks which serve for both storage and chem-
ical pretreatment. For the pretreatment step, an alkali is added to the coal
slurry which is then heated to the required temperature. Air is introduced
next to oxidize the surface of the pyrite particles after which the slurry Is
cooled to a set temperature for the subsequent separation steps. If a finer
particle size is desired, the coal is ground with a ball mill before apply-
ing the chemical treatment. The ball mill circuit includes cyclones for both
thickening the pulp supplied to the mill and classifying the particles accord-
ing to size. Consequently, only the coarser particles enter the ball mill.
After the feed has been adequately ground and/or pretreated, it is pump-
ed to either a bank of froth flotation cells or the first stage of the oil
agglomeration system. If the feed Is directed to the bank of flotation cells,
a frothing agent is added and the coal is floated and removed in the froth
while the refuse is removed in the underflow. The float product is either
filtered to recover the coal or placed In a storage tank to await further
treatment.
Either coal fines cleaned by froth flotation or coal fines which have
only been chemically pretreated can be oil agglomerated. A slurry of these
fines is delivered to the first stage of a two stage agglomeration system.
Fuel oil is added and microagglomerates are produced by high shear mixing.
The suspension of microagglomerates is conducted to a vibrating screen for
dewatering and desliming. The microagglomerates are resuspended in fresh
water in the second stage where less vigorous agitation promotes the coal-
escence and growth of large agglomerates. The suspension is then dewatered
524
-------
on another vibrating screen. The agglomerated coal can either be recovered
at this point or conveyed to an Inclined rotating disc petletizer for further
size enlargement.
Since construction of this unit was completed only recently, a program
of demonstration runs is just getting underway. The results will be evalu-
ated in terms of the recovery of combustible matter, the sulfur, ash, and
moisture content of the product, and its physical properties. The effects
of important parameters such as residence time, slurry concentration, reagent
concentration, and temperature will be applied to several different represen-
tative coals from various regions of the country.
V. HIGH GRADIENT MAGNETIC SEPARATION
"High gradient magnetic separation" process Is being investigated
to determine its feasibility in cleaning high sulfur coal. In this process
fine coal In a water slurry is passed through a magnetic field in which a
mesh of ferromagnetic material (stainless steel wool) has been placed. The
refuse, which has different magnetic properties than the coal, is attracted
to the mesh thereby performing the separation. When the mesh is loaded with
refuse either the magnet Is turned off or the mesh is removed from the field
allowing the refuse to be flushed away. This process can be enhanced by
seeding the raw coal to increase the magnetic susceptibility of the refuse.
This project is presently in the planning stage.
Presently, laboratory studies are underway to determine the relative
merits of induction heating, microwave heating, and chemical treatment to
alter magnetic properties.
VI. SLAGGING AND FOULING CHARACTERISTICS OF RAW VS CLEANED COALS
To determine the relative effect cleaning coal has on boiler operation,
the ashing properties of four coals were determined before and after clean-
Ing.
Although the number of coals was limited, the results show a significant
benefit through cleaning. (Grieve, 1978)
525
-------
A complete chemical analysis was performed on the ash from four coals
which had widely varying sulfur content to determine the base content, acid
content, base/acid ratio, silica/alumina ratio, silica value, iron value and
dolomite percentage. The fouling index was computed by multiplying the base/
acid ratio with the percent dry sodium and the slagging index by multiplying
the base/acid ratio with the percent dry sulfur. These indexes were compared
to indicate the effect cleaning coal has on boiler operation. (Attig and
Dunzy 69)
Of the four coals selected, the reduction of the slagging index was
least in the lowest sulfur coal and dramatically reduced in the other three.
(Figure 7) A lesser reduction in the fouling index was also indicated.
The performance observed while burning these coals in the power plant
correlates to the calculated values of fouling and slagging indexes.
VII. RELATED PROJECTS
Several projects are also being pursued which relate either directly or
Indirectly with the utilization of coal as an energy source.
An on-line nuclear sulfur and ash monitoring device is being developed
to continuously measure the total ash and the total sulfur In a moving coal
stream while at the same time being insensitive to the moisture level.
An on-line x-ray diffraction technique has shown that monitoring of all
forms of Inorganic sulfur can be continuously monitored along with other
selected minerals. This device is also insensltve to moisture levels in the
measured stream.
Both the nuclear and the x-ray diffraction techniques give almost in-
stantaneous measurements (approximately 30 seconds).
A process for pelleting fine coal is being investigated with present
results showing that much less pressureis required than for conventional
briquetting and that higher moisture levels can be accommodated to a degree
which may eliminate the customary expensive step of thermal drying.
Methods for separating fly ash into components which may be commercially
utilized for the production of aluminum, iron and other metals while re-
moving undesirable elements from the remaining rejects are being investigated.
526
-------
Nuclear magnetic resonance studies of coal are being pursued to aid in
characterization of the basic structure of coal.
Analysis of major, minor and trace elements in high BID coal gas and
effluents resulting from refuse derived fuels are being done using plasma
fluorescence spectroscopy and other methods.
VIII. CONCLUSION
Using conventional equipment, physical coal cleaning has demonstrated
the removal of 2k to 45% of the sulfur In several high sulfur coals. More
intensive physical cleaning, with a smaller size consist, indicates the pos-
sibility of increasing the inorganic sulfur removal up to'88%. Research
plans include beneficiation tests of various medium and high sulfur coals not
previously tested with present equipment and alternate methods.
Related investigations should lead to a better understanding of the
basic nature of coal, reclamation of post-combustion products, and Improved
analysis of process streams both in physical coal cleaning plants and in coal
conversion plants.
REFERENCES
Attig, R. C., and Duzy, A. F., "Coal Ash Deposition Studies and Applications
to Boiler Design," The Babcock g. Wilson Company, Alliance, Ohio, Proceedings
of the American Power Conference, Vol. 31, 1969, Pages 290 to 299.
Cavallaro, R. A., and Deurbrouck, A. W., "U.S. Bureau of Mines Report of In-
vestigations Rl 8118", 1976, Pages 83 to 91, United States Department of the
Interior.
Grieve, Richard A., Chu, Henry, and Fisher, Ray W., "Iowa Coal Project Pre-
liminary Coal Beneficiation Cost Study Progress Report," Energy and Minerals
Resources Research Institute, Iowa State University, Ames, Iowa (September,
1976).
527
-------
Figure 1. Aerial View of Restored ICP Demonstration Mine No. 1 Site and
Reclaimed Childers Research Site
Laboratory Research Areas
Childers Reclamation
Research Site
528
-------
Ui
ro
Figure 2. Aerial View of Iowa State University Coal Preparation Plant (white roof) showing
close proximity to University Power Plant (Upper R.H. corner of photo)
-------
Run of Mine Coal
Drain 6 Rinse
Screen
Hydro Cydon
Centrifuge
To Sump
Crushing & Separating
L media Tl Magnetic Separalo
CCool
'SV--Refuse
Cone Type
Heavy Media Circuit
Heavy Media Cyclone Circuit <,*,,„.,»
Refuse
C Cool
Fines w/Process Water
Concentration Table Circuit
Heal Exchanger
Improved HGMS Ci
Pretreatment Circuit
_4__ Water from
Thickener
..Clarified
Woter to Emulsifier
Plant
Briquetter/Extrudet
Thermal Drier
—* Compacted
Coal
Sump
Hydro Cyclone Circuit
w/ Thickener & Filters
C Cool lo Filler
Froth Flotation Circuit
Chemical Desulfurization Circuit
(planned)
Cone Type Heavy Media
Heavy Media Cyclone
Concentration Table
Compacted Cool
Filter Cake
Cone Type Heavy Media
Heavy Media Cyclone
Concentration Table
Fitter Coke
Settling Basin Oil Agglomeration Circuit Briquetting Circuit lplm.d, Clean Coal Conveyors Refuse Conveyor
ISU/DOE Coal Preparation Research Facility — Configuration Diagram ©•*»«•'«*
Con
S&
Configurolion Diagram
33 CRP
Figure 3. Configuration Diagram
-------
Ln
CO
Coal
Processed
ISU ?!
ICO
Lovil ia1
Mich
Shinn
Big1
Ben
Jude
Average of
all
Coals Tested
Specific
Gravity
1.5
1.35
1.50
1.*0
1.35
1.60
1.*5
1.1*0
Size
l£"X
*8H
1" X
*8M
ii»x
*8M
ii"X
*8M
ii"X
*8M
li"X
*8H
ii»x
*8M
l£"X
*8M
Raw
Clean
Raw
Clean
Raw
Clean
Raw
Clean
Raw
Clean
Raw
Clean
Raw
Clean
Raw
Clean
BTU/ 1 b
10,572
11,312
10,690
11,72*
9,839
11.868
10,222
1 1 ,*3*
10,558
12,058
9,368
10.511
8,070
10,709
9,903
1 1 ,37*
Total
Sulfur
(%)
6.95
*.91
5.*8
3.37
2.51
2.0
8.7*
5.*6
*.56
3.*0
4.76
2.50
7.8*
*.22
5.83
3.69
Pyritic
Sulfur
(%>•
1.
1.
6.
2.
3.
1.
3.
1.
6.
1.
*.
1.
77
16
66
89
25
86
90
58
11
77
3*
85
Ash
ft)
16.21
10.73
11.55
7.22
17.16
9.57
2**. 51
1*.22
15.99
8.77
21.3*
9.8*
29.96
12.58
19.53
10. *2
Moisture
(%)
9.2*
9.39
12.9
10.2
11.7*
8.0
3.26
5.12
8.65
6.*9
18.32
15.57
9.0*
11.39
10. *5
9.*5
Lbs
S02/
MMBTU
13.15
8.68
10.25
5.75
5.10
3.37
17.10
9.55
8.6*
5.63
10.16
*.76
19. *3
7.88
11.98
6.52
B.O.M. Comparative Data (1)
li" X 100H & 1.40 S.G.
Tonnage
Yield
(%)
80.0
73.6
7*. 9
72.7
66.3
78.5
70.8
73.8
,••
72.3
BTU
Yield
(%)
85.7
78.3
86.7
82.9
7*.0
86.9
96.*
8*.*
78.8
Ash
Reduction
(%)
33.7
39.*
*6.5
*0.8
46.1*
48.8
56.9
44.6
*6.5
Pyritic
Sulfur
Reduction
(%)
37.0
55-7
**.l
51.3
70.2
51.7
59.5
Total
Sulfur
Reduct ion
(%)
29.2
*0.*
23.6
36.3
27.0
*2.6
**.8
3*.8
38.0
so2
Reduction
(%)
3*.0
*3.9
33.9
**.2
3*.8
*8.2
59.*
*2.6
*3.3
2Losses in yield computations include fines (-*8M) losses.
All proximate analysis-- Air dried basis.
All reduction factors - Moisture free basis.
Figure k. Iowa Coals Processed at Iowa Coal Project Preparation Plant
Data Represents Best Run Made on Each Coal
-------
Figure 5- The recovery of fine coal will be demonstrated by this
unit consisting of reagent feeder, conditioning tank,
and bank of four froth flotation cells.
Figure 6.
The oil agglomeration and recovery of fine coal will be
demonstrated by this system of agitated tanks and dewaterlng
screens.
532
-------
Ut
ISU
Coal Proximate Analysis
% Sulfur
% Ash
BTU
Specific Gravity
Ash Analysis
2Ash Fusion Temp (F°)
Base Content
Acid Content
Base/ Ac id Ratio
Silica/Alumina Ratio
Silica Value
Iron Ratio
Dolomite Percentage
^Fouling Index
'Slagging Index
Foul ing Type
Low Less than
Medium
High
Severe Greater than
1 . Air dried basis
(strip
RAW
11.52
17.86
11,384
2,087
52.63
34.41
1.53
1.83
30.39
2.49
27.61
.20
17.63
0.2
0.2 - 0.5
0.5 - 1 .0
1.0
mined)
CLEANED
5.42
11.84
12,484
2,110
35.01
60.61 _
.58(62%) 3
1.67
50.78
4.51
17.15
.09
3.14
Low
Medium
High
Severe
Lovilia
(deep, mined)
RAW
2.23
12.06
12,534
2,107
40.39
50.24
.80
2.03
47.20
1.48
37-39
.29
1.78
Slagging
Less
Greater
CLEANED
2.17
10.40
.12,901
2.150
39.35
49.74
.79(1:03
2.34
48.66
1.12
43.76
.33
1.71
Type
than 0.6
0.6 - 2
2.0 - 2
than 2.6
2. Average of Initial Deformation, Fusing, and Fluid Temperature under
MICH
(strip
RAW
8.74
24.51
10,222
2,037
53.99
33.17
1.63
2.09
29.59
3.07
24.15
.09
14.72
.0
.6
mined)
CLEANED
5.46
14.22
11,434
2,073
46.78
36.98 ,
1.27 (22%) •>
1.95
35.14
1.51
38.26
.02
7.32
ICC
(strip
RAW
7.10
18.28
10,896
2,147
76.10
14.82
5.13
1.35
10.11
4.90
16.77
.13
38.17
mined)
CLEANED
4.94
8.94
11,794
2,153
65.01
3T.16
2.09(59^)
2.:i%
24.6=
8.35
9-77
.03
11.24
reducing 'atmosphere conditions
3. Percent reduction from raw coal.
4. (Base/acid ratio) X
5. (Base/acid ratio) X
(%Na)
(Dry %s)
Figure /• Comparison of Ashing Properties
of Raw vs Beneficiated Coals
-------
CLEANING OF EASTERN BITUMINOUS COALS BY FINE GRINDING,
FROTH FLOTATION AND HIGH-GRADIENT MAGNETIC SEPARATION
W. L. Freyberger, J. W. Keck, D. W, Spottiswood,
N. D. Solem and Virginia L. Doane
Michigan Technological University
Houghton, Michigan 49931
ABSTRACT
Mineralogical and bench-scale beneficiation studies were conducted with
five Eastern bituminous coals to develop processes so as to recover 85 percent
of the Btu value of the coal while rejecting 85 percent of the pyritic sulfur
and the ash. The coals studied were from the Illinois, Pittsburgh, Middle
Kittaning, Hartshorne and No. 12 Coal Bed seams. The desired results were
obtained reasonably well by fine grinding of the raw coal, followed by froth
flotation and treatment of flotation concentrates or middlings by high
gradient magnetic separation. Alternative flowsheets involving froth flota-
tion alone or regrinding and retreatment of middlings fractions also gave
reasonably good results.
Several different treatment processes have been demonstrated in a pilot
plant treating 400 pounds of raw coal per hour, using either Illinois or
Pittsburgh seam coal as feed.
The processes described have not been optimized and a number of potentially
useful variations suggest themselves. Preliminary economic analyses have been
made with the available results. The overall program has demonstrated that raw
coal can be cleaned by employing the general approach commonly employed in
treatment of metallic ores.
534
-------
INTRODUCTION
Coal cleaning by physical means 1s presently receiving much attention as
a means of reducing environmental problems associated with coal combustion.
These problems arise from the ash and sulfur contents of the coal and from the
presence of trace amounts of hazardous elements such as heavy metals, beryllium,
selenium and arsenic. Cleaning by physical means can reduce the ash and pyH-
t1c sulfur contents of the coal. Hazardous trace elements will be removed to
the extent that these materials follow the waste rather than the coal.
Coal cleaning as presently practiced 1n the U.S. generally employs gravity
concentration of coarsely crushed coal and treatment of the fines produced dur-
ing crushing by froth flotation, cyclonlng or screening. For many eastern bi-
tuminous coals, liberation of pyrlte and ash from the coal 1s largely Incomplete
at the sizes employed 1n the washing operations (Cavallaro, 1976). No coal
cleaning plants deliberately grind coal to a very fine size to achieve sub-
stantial liberation of ash and pyrlte.
On the basis of these considerations a research contract was awarded to
Michigan Technological University by the U.S. Bureau of Mines (subsequently be-
ing transferred to the U.S. Department of Energy) to Investigate coal cleaning
by means of mineral processing techniques commonly employed for treatment of
base metal sulflde ores. Processes of particular Interest Included fine grind-
Ing, regrlndlng of middling products and application of separation processes
535
-------
such as froth flotation and high gradient magnetic separation. Targets set
forth in the contract were to recover 85% of the BTU value of the raw coal
while recovering 15% or less of the ash and of the pyritic sulfur.
Project Scope - Coal Samples
The contract called for laboratory investigation of several coals to be
selected by DOE. Samples of five eastern bituminous coals were obtained, as
listed in Table 1.
Laboratory studies were conducted with the five coals to develop processes
capable of achieving the desired targets. The resulting processes were then
tested further in pilot plant campaigns with the Illinois No. 6 Seam and Pitts-
burgh No. 8 Seam coals. These campaigns totalled 15 days of three-shift opera-
tion and 10 days of single shift operation.
This paper will be concerned primarily with the laboratory development of
the treatment processes and subsequent demonstration of these processes in the
pilot plant. It should be emphasized that the processes as described and the
results obtained are not necessarily optimum. The scope of the project did not
allow for such detailed investigation.
The paper will concentrate on the work done with the Illinois and Pitts-
burgh Seam coals. Results obtained with the other three coals were generally
similar.
Chemical Analyses
ASTM analytical procedures were used for determining ash, sulfur in all
forms, moisture and calorific value. To minimize the need for calorlmetric
analyses on a large number of test products, an equation for calculating the
coal analysis of a test product was developed as follows:
%(Cdal) = 100 - X(Ash) - 1.6(%PyS).
536
-------
Estimated coal analyses obtained by this relation were then used to determine
the coal recovery in a given laboratory or pilot plant test. These recoveries
were compared with BTU recovery values determined from calorimetric measure-
ments for a number of tests as listed in Table 2. In general the comparisons
were reasonable and the estimated coal recoveries tended to be a little low,
thus making analysis of the test results conservative. The results described
in this paper are in terms of coal recoveries.
FLOWSHEET DEVELOPMENT.
Size of Grind
All of the coal samples contained pyrite and ash grains from several hun-
dred microns in diameter down to only a few microns. Examples of very fine
pyrite and ash are shown'in Figures 1 and 2.
These observations clearly demonstrated the impracticality of grinding
coal to such a size that the pyrite and ash would be completely liberated.
Instead, the coal would have to be processed at some size where a part of the
ash and pyrite remained locked with coal in middling grains. The desired size
of grinding for each coal was established on the basis of meeting the desired
processing targets and ranged from -150 mesh to -325 mesh.
Application of Individual Treatment Processes
Three separation processes were studied individually in the laboratory-
froth flotation, high gradient magnetic separation and selective agglomeration
(or bulk oil separation). Samples of raw coal were stage crushed through 10-
mesh and stored in tightly closed containers to provide feed material for
laboratory tests.
Froth flotation. Laboratory flotation tests were made with 300g charges
537
-------
of -10 mesh raw coal. The coal was ground wet at 50% solids 1n a laboratory
rod mill. Flotation was done in a laboratory Fagergren machine. Distilled
water was used throughout.
Two stages of flotation were employed—a rougher stage, at about 12%
solids, followed by refloating the rougher froth in a cleaner stage at about
9% solids. Coal was floated with MIBC (methyl isobutyl carbinol) alone or
with No. 1 fuel oil. The flotation pH was adjusted to 8.5 with sodium hy-
droxide.
Variations in coal, ash and pyrite recoveries were controlled primarily
by varying the size of grind and the collector additions. The coarser the
coal, the more collector was required to float it.
Figures 3 and 4 summarize flotation results obtained in the laboratory
with the Illinois Seam coal. Two sizes of grind were studied, -150 mesh and
-325 mesh. (As used in this paper, the term "-150 mesh grind", for example,
means the ground coal was 90% to 95% -150 mesh in size.) The flotation collec-
tor used was a mixture of MIBC and No. 1 fuel oil.
The curves in Figure 3 demonstrate that pyrite rejection at a given coal
recovery improved substantially as the coal was ground finer. Grinds finer
than about 90% -325 mesh were not tested because it was felt that such fine
grinding would prove impractical. No significant improvement in ash rejection
resulted from grinding finer than about 90% -150 mesh with the Illinois coal.
With a -325 mesh grind for Illinois Seam coal, the pyrite recovery was
about 30% and the ash recovery about 17% at 85% coal recovery. Reagent addi-
tions were 1.2 Ib of NaOH; 1.05 Ib of MIBC and 1.55 Ib of fuel -oil per ton of
raw coal processed.
Somewhat better results were obtained with the Pittsburgh No. 8 Seam coal
with -325 mesh flotation feed. At 85% coal recovery, pyrite and ash recoveries
538
-------
were 24$ and 11$ respectively. Reagent additions were 0.20 Ib of NaOH, and
.50 Ib of MIBC per ton of coal.
High gradient magnetic separation (HGMS). Laboratory HGMS tests were con-
ducted with a mobile unit leased from Aquafine Corporation. This machine is a
cyclic separator using a cylindrical canister filled with magnetic matrix as
the separating chamber. Tests were made with a canister 1" diameter by 20"
long, packed with pads of magnetic stainless steel wool. The canister was
placed in a magnetic field of 20 kilogauss (as measured in.open space).
The magnet operating cycle consisted of pumping feed slurry upward through
the canister with the magnet on. Then, with the magnet still on, the feed
slurry was replaced with a flow of fresh water to purge non-magnetics entrained
in the steel wool matrix. Finally, with the magnet off, magnetic material was
flushed from the canister with a down flow of high velocity water.
Variables of the magnet operation included the feed rate of coal slurry
and purge water (expressed in terms of the equivalent retention time of slurry
in the canister), the duration of each of the three parts of .the total cycle
and the solids content of the feed slurry. Many sets of conditions were tested.
Those ultimately considered best were a feed of 15$ solids at a retention time
of 0.75 minutes and a cycle program of 3 minutes of feed on, 1.5 minutes of
purge and 1 minute of flushing. These conditions were maintained throughout
the pilot plants runs. Longer retention time Improved pyHte removal, but coal
losses in the magnetic product also increased and the capacity of the machine
decreased. Matrix packing was 6% by volume in the canister.
The results obtained with HGMS from Illinois No. 6 Seam coal are shown in
Figure 5. In these tests charges of -10 mesh coal were slurried with water and
deslimed prior to grinding to 9056 -325 mesh. This Initial desllming removed
over half of the ash. Such treatment was found essential to obtaining reasonable
539
-------
results by HGMS. If raw coal was used as the magnet feed, ash material was re-
moved in preference to pyrite. As there is considerably more ash than pyrite
in raw coal, pyrite removal from raw coal was generally poor.
The results presented in Figure 5 are about the same as those obtained by
flotation. For a coal recovery of 85%, the ash and pyrite recoveries were 25-
30%. Furthermore, the curves are very steep in this region of coal recovery.
Control of plant operations would be difficult to maintain under these condi-
tions .
Selective agglomeration. A limited study was made of the application of
selective agglomeration by treatment with bulk oil to the cleaning of raw coal.
In these tests dilute slurries of coal in water were mixed with varying amounts
of No. 2 fuel oil. Mixing was done at high shear in a blender. The coal parti-
cles combined with the oil to form agglomerates. These were screened away from
the water phase which contained the unagglomerated ash and pyrite fines.
Results obtained with Hartshorne Seam coal are shown in Figure 6. At 85%
coal recovery, pyrite recovery was about 35% and ash rejection about 28%. These
separations were not efficient and did not appear to be better than could be ob-
tained more conveniently by froth flotation. Therefore, the use of selective
agglomeration as a primary separation process was abandoned early in the project
and emphasis was placed on froth flotation and HGMS.
Processing by Froth Flotation and HGMS
None of the three separation processes studied was successful by itself in
obtaining the desired results on treating raw coal. However, the results suggested
that treatment by both froth flotation and HGMS would be effective. Flotation
was effective in removing ash and a substantial part of the pyrite. The flota-
tion concentrate would then be a desirable feed to HGMS to reduce the pyrite and
ash contents to the target levels.
540
-------
On thts basts laboratory tests were run according to the flowsheet shown
schematically tn Figure 7. The coal was ground to about 90% -325 mesh and
treated by two stages of flotation as before. The flotation concentrate was
then treated by HGMS and the non-magnetic product was the finished clean coal.
Typical results as obtained with the Illinois No. 6 Seam coal are pre-
sented tn Figure 8. For an overall coal recovery of 84%, the pyrlte recovery
was 15% and the ash recovery was 9%. The clean coal analysis was 5.8% ash and
0,36% pyritlc sulfur. Corresponding results obtained with Pittsburgh Seam coal
were 16% pyrlte recovery and 11% ash recovery for 85% coal recovery.
A key to the success of this process was to recover 92% to 95% of the coal
tn the flotation circuit to Insure an overall coal recovery,of 85%. Coal lost
tn the flotation tailings was Irretrievable 1n the process sequence employed.
Flotation reagent consumptions were about 1.05 Ib of MIBC and 1.50 Ib of fuel
otl per ton for the Illinois Seam coal and 0.51 Ib of MIBC per ton for the
Pittsburgh coal. HGMS operation was the same as already described.
PILOT PLANT OPERATION
Flowsheet A -Illinois Seam Coal
After study tn the laboratory"the combined flotation-HGMS process was tested
tn a conttnuous pilot plant operation treating 400 Ib of raw coal per hour. The
ftrst flowsheet employed, termed Flowsheet A, 1s shown 1n Figure 9,) .The first
pilot plant operation was with Illinois No. 6 Seam coal.
There were two principal differences between this flowsheet and that followed
tn the laboratory. First the pilot plant grinding circuit employed two stages
of grinding—an open circuit rod mill followed by a ball mill operating in closed
ctrcuit wtth a cyclone. The cyclone overflow was the flotation feed. The second
difference was the Incorporation of a ^00-mesh square mesh screen to screen the
541
-------
the flotation concentrate prior to treatment with HGMS. The screen was Intro-
duced Initially to protect the magnet from any +200 mesh material 1n the flo-
tation concentrate, which would tend to clog the steel wool magnet matrix. Both
of these flowsheet differences proved Important 1n the pilot plant operation.
A 5-Inch diameter canister was used 1n the HGMS during pilot plant opera-
tions. Slurry and purge water feed rates were scaled up from laboratory condi-
tions according to the ratios of cross-sectional areas of the pilot plant and
laboratory canisters.
Grinding. The cyclone overflow obtained at the start of the pilot plant
contained about 27% +325 mesh material, compared with a desired value of 1056.
Furthermore, the size distribution of the cyclone overflow was substantially
different from that obtained 1n a laboratory grind, as seen 1n Figure 10. The
coal ground 1n the pilot plant was deficient 1n material between 15 to 40 microns,
but contained an excessive amount of material coarser than 50 microns (or say
270 mesh). A number of changes were made 1n the grinding circuit to grind this
coarse fraction further. The best that could be done without reducing the feed
rate was to produce a cyclone overflow containing 1956 +325 mesh material.
Screening. Samples of the +270 mesh fraction of the cyclone overflow ob-
tatned with the Illinois Seam coal assayed about 0.5% pyrltlc sulfur and about
5.556 ash. These analyses were as good or better than could be obtained In the
non-magnetic product from HGMS 1n the pilot plant with this coal. Therefore,
tt was decided to remove this coarse material as finished coal by screening.
This was accomplished with the 200-mesh screen treating the flotation concen-
trate ahead of magnetic separation. The screen oversize was combined with the
non-magnetics from HGMS as the final concentrate.
Results. The pilot plant was operated with Flowsheet A for 15 consecutive
8-hour shifts with Illinois No. 6 Seam coal as feed. Samples were collected
during each shift.
542
-------
Several changes 1n operation were made during this period. These Included
varying the grind between 19% and 27% +325 mesh and varying the additions of
flotation reagents. The overall results from all of these shifts are summarized
1n Figure 11 1n terms of the recoveries of coal, pyrlte and ash 1n the finished
cleaned coal. The figure also Includes two points representing typical pyHte
and ash recoveries from a laboratory test.
The data 1n Figure 11 cluster about two curves, one for pyrlte recovery
and the other for ash, over a wide range of coal recovery. Results reasonably
close to the desired targets were obtained. In the pilot plant, at 85% coal re-
covery, the pyrlte recovery was about 24% and'ash recovery about 16%. Pyr1t1c
sulfur and ash analyses were 0.40% and 8.3% respectively.
However, the selectivity of separation was not as good 1n the pilot plant
as In the laboratory. The principal difference was 1n the flotation circuit.
Pyrlte and ash recoveries were 5% or more higher for a given coal recovery 1n
the pilot plant flotation circuit with the Illinois No. 6 coal.
Flowsheet A - Pittsburgh No. 8 Seam Coal
The second pilot plant campaign was conducted with Pittsburgh No. 8 Seam
coal, again with Flowsheet A. Again the cyclone overflow contained about 27%
+325 material as compared with 10% 1n the laboratory grind. The pilot plant
was operated for 15 shifts.
Several changes were made 1n an effort to Improve the flotation results.
The solIds content of the flotation pulps 1n the rougher and cleaner stages
were varied between 10% and 15% and the flotation times 1n the rougher and
cleaner stages were varied.
The results obtained during this operation are presented 1n Figure 12.
Again the results cluster about two lines and again, while results reasonably
close to the target values were obtained, the performance was not as good as.
543
-------
had been obtained in the laboratory. Failure to obtain proper selectivity in
the pilot plant flotation stm remained the problem.
For a total coal recovery of about 85%, the pyritic sulfur and ash analyses
of the cleaned coal were typically 0.63% and 5.8% respectively. The correspond-
ing pyritic sulfur and ash recoveries were 22% and 17%.
Flowsheets B and C - Pittsburgh No. 8 Seam Coal
In the third and last week of pilot plant operation, run with Pittsburgh
No. 8 Seam coal, two major changes were made consecutively in an effort to Im-
prove flotation and thereby improve the overall results. The total run was for
15 shifts.
Flowsheet B. The first change, called Flowsheet B, is shown in Figure 13
by the solid lines. The cyclone overflow was screened on a DSM screen and the
oversize and undersize fractions were floated separately. The two flotation
concentrates were then combined and treated as in Flowsheet A. .(In Figure 13
the primary rod mill-ball mill-cyclone circuit has been omitted for simplicity.)
This change was made because there was reason to believe that the presence
of coarse coal along with fine coal in the flotation circuit was the cause of
the decrease in flotation selectivity. Laboratory results had established that
more flotation collector was required, the coarser the coal being floated. In
the first two pilot plant runs, required collector additions had tended to be
higher than in the laboratory to attain desired coal recovery values. With in-
sufficient collector additions in the pilot plant, coarse coal was readily ob-
served in the flotation tailings.
These factors raised the possibility that excessive collector was being
required in the pilot plant to float the large amount of +325 mesh coal. This
could lead to decreased flotation selectivity, particularly for the fine sizes.
Flowsheet C. The second change, made during the last five shifts of the
544
-------
15 shift run, was to regrind the DSM oversize material before flotation. This
flowsheet is shown by the dotted lines in Figure 13. The DSM oversize was
passed through a ball mill-cyclone closed circuit and was ground to about 13%
+325 mesh, the same size as the DSM undersize product. During this period of
operation, then, the size distribution of the flotation feed was close to that
used in the laboratory tests which was about 10% +325 mesh.
The reground DSM oversize and the DSM undersize were floated in separate
circuits as in Flowsheet B. The rest of the circuit was also the same as for
Flowsheet B.
Results. The results obtained with Flowsheets B and C are presented in
Figure 14. Two lines are presented for pyrite recovery versus coal recovery,
one for Flowsheet B and one for Flowsheet C. However, the difference between
these lines is small and may not even be real. There was no difference in the
ash recovery obtained with the two circuits.
Comparison of the results in Figures 12 and 14 demonstrates that a small
reduction in pyrite recovery was obtained in going from Flowsheet A to Flow-
sheets B and C. Ash recovery appeared to be unaffected. At 85% coal recovery,
pyrite recovery was 20-21% and ash recovery was 17%.
Thickening and Filtration
Some of the pilot plant concentrates were thickened and filtered to gain
some information about these problems. The coal was flocculated with Superfloc
127 prior to thickening. Thickener feed contained about 10% solids and the
underflow about 35% solids.
Filtration was done on a vacuum drum filter. The coal readily formed a
cake about 0.75 inches thick which discharged easily. Filter cake moisture
was about 35%.
545
-------
DISCUSSION
Pilot Plant Operation
Mechanically, the pilot plant operations went very smoothly. The only
difficulty encountered in handling the coal was that the Pittsburgh Seam coal
tended to be sticky and hung in the feed bin. The flotation machines were easy
to control and the froth was handled readily.
HGMS operation was completely automated. Except for periodic recording
of operating conditions, the machine was left unattended. Constant flow rates
of feed coal slurry and purge water were maintained with a Moyno pump. Flush
water was fed directly from the pilot plant main. Pulsing of this flow by a
fast-acting solenoid valve provided better removal of magnetics than did a con-
tinuous flow. The magnet feed slurry had to be free of +200 mesh material to
prevent plugging of the steel wool matrix. The 5-1nch canister processed 80-
90 Ib of sol Ids per hour.
Efficiency of Separation Processes
The laboratory and pilot plant results clearly demonstrated that ash and
pyrlte can be effectively removed from bituminous coal, while maintaining high
coal recovery, in a circuit employing fine grinding, froth flotation and high
gradient magnetic separation of the flotation concentrate. Fine screening of
the cyclone overflow or flotation concentrate may be useful 1f the coarse frac-
tion 1s clean coal.
Typical laboratory and pilot plant results obtained with Illinois No. 6
Seam and Pittsburgh No. 8 Seam coals are summarized in Table 3. Figure 15 pre-
sents a complete materials balance for one shift of pilot plant operation with
Illinois Seam coal.
The Initial targets of 85% coal recovery and 15% recovery for ash and for
pyrlte were closely attained in the pilot plant and often exceeded in the
546
-------
laboratory. There remains an unexplained difference in the flotation results
obtained in pilot plant as compared with laboratory. Recent data suggest that
the problem was in the pilot plant flotation circuit itself, rather than due
the manner of preparing flotation feed. However, the source of the problem
remains unidentified.
Cost Analysis
A detailed cost analysis of the overall process as illustrated in Figure 9,
Flowsheet A, has not yet been made. However, the two principal cost centers will
undoubtedly be grinding and HGMS. The energy required for grinding in the pilot
plant totalled 39 KWH/ton of raw coal.
Capital and power costs for these operations have been estimated for a plant
treating two million tons per year of raw coal. These estimates are presented
in Table 4. Operating time for the plant was assumed to be 90% of 365 days per
year. Also, plant results were assumed to be about 85% coal recovery and 15-20%
recovery of each of the ash and pyrite. If more pyrite and ash can be tolerated
In the clean coal, coal recovery would be higher and capital and operating costs
might be lower (due to decreased costs for either or both of grinding and HGMS).
The costs presented in Table 4 represent only a part of the total costs re-
quired for the coal cleaning operation. However, they may be compared with
present coal preparation costs. A recent analysis of operating jig and heavy
media plants showed total costs in the range of $2.00 to $3.50 per ton of raw
coal processed (Hoffman-Munter Corp., 1978).
Further Work
There are at least three major areas of study which will be required before
treatment of coal by processes like those described can be applied commercially.
The first is to optimize the process for a given coal to reduce costs. Un-
doubtedly, the overall efficiency can be Improved over the results described
547
-------
in this paper. In particular, improvement in the flotation efficiency should
be possible. Also, any increase in coal recovery, as by retreatment of middlings,
will be very important. Finally, the trade-off between cost on the one hand and
separation efficiency on the other must be studied.
A second problem will be to transform the fine coal concentrate to a form
suitable for handling, transportation and use by the consumer. This most likely
will mean dewatering to an acceptable level and production of agglomerates such
as briquettes or pellets.
Finally, acceptable methods of disposal must be developed for the fine tail-
ings.
Acknowledgements
The authors wish to express their appreciation to the U.S. Department of
Energy and the U.S. Bureau of Mines for their support of this study. We also
wish to express our appreciation to Mr. Donald H. Rose and the other members of
the analytical chemistry laboratory at Michigan Technological University for per-
forming the necessary chemical analyses.
548
-------
REFERENCES
Cavallaro, J.A., M. T. Johnson and A.M. Deurbrouck. 1976. Sulfur reduction
potential of the coals in the United States. U.S. Bureau of Mines
Report of Investigation No. 8118.
Hoffman-Munter Corporation. 1978. An engineering/economic analysis of coal
preparation plant operation and cost. Report to U.S. Department of Energy.
Contract No. ET-75-C-01-9025.
549
-------
TABLE 1
SELECTED COALS
% SULFUR
1,
2,
3,
4,
5,
COAL SEAM
ILLINOIS NO, 6
COAL BED NO, 12
HARTSHORNE
MIDDLE KITTANING
PITTSBURGH NO, 8
TOT,
2
2
1
1
3
,02
,80
,05
,22
,25
PY,
1
2
0
0
2
,39
,14
,51
,69
,29
OX,
0,06
0,04
0,004
0,07
—
% ASH
26,
31,
16,
27,
33,
0
0
3
1
3
550
-------
TABLE 2
CALCULATION FOR ESTIMATING
COAL ANALYSIS OF TEST PRODUCTS
%(COAL) = 100 - Z(ASH) - l,6%(PvS)
ft*************
% COAL RECOVERY AS
CALCULATED FROM
EST, COAL
ANALYSES
62,0
76,4
77,2
80,4
81,5
85,7
87,3
CALORIMETRIC
BTU DET'NS,
63,0
80,3
79,2
82,6
80,9
89,3
90,7
551
-------
TABLE 3. COMPARISON OF LABORATORY
AND PILOT PLANT RESULTS
A, ILLINOIS NO, 6 SEAM COAL
CLEANED WT,
COAL FROM %
LABORATORY 55,6
PILOT PLANT 64,1
TOT.S
1,32
1,36
ASSAY
PvS
0,365
0,420
I
ASH
5,78
7,73
LB S09/MM
Ln
tn
to
CLEANED WT,
COAL FROM %
LABORATORY 64,0
PILOT PLANT 65,7
RAW COAL
CLEANED
TOT.S
2,00
2,00
COAL
B,
ASSAY
PvS
0,544
0,665
LB
BTU/LB
13,300
13,200
BTU (APPROX,)
4
2
COAL
83.5
85,0
DIST. %
TOT.S PvS
36,3 14,6
43.2 25,8
ASH
9,1
16.9
PITTSBURGH NO, 8 SEAM COAL
%
ASH
4,96
5.66
BTU/LB
14,300
13,900
COAL
83.1
83.1
DIST. I
TOT.S PvS
37.1 14.4
38.1 20,9
ASH
13,4
16,2
S02/MM BTU (APPROX,)
RAW COAL 6
CLEANED COAL 3
-------
TABLE 4
ESTIMATED CAPITAL
AND POWER COSTS
FOR GRINDING AND HGMS
BASIS; 2 MILLION TONS RAW COAL PER YEAR
250 TONS PER HOUR
POWER AT 2,5$/KWH
CAPITAL COST POWER COST TOTAL
TOTAL $/TQN $/TON $/TON
GRINDING $3,3MM 0,25 0,95 1,20
HGMS $21 MM 1,90 0,50 2,10
553
-------
FIGURE 1, PHOTOMICROGRAPH OF PITTSBURGH NO, 8
SEAM COAL,
FIGURE 2, PHOTOMICROGRAPH OF ILLINOIS NO, 6
SEAM COAL,
554
-------
FIGURE 3
50
in
l/i
In
30
car
LU
£20
10
0
LABORATORY FLOTATION RESULTS - ILLINOIS NO, 6 COAL
© - 325 MESH (DESLIMED)
A- 325 MESH (NOT DESLIMED)
x- 150 MESH (DESLIMED)
10
20
30
50 60
% COAL RECOVERY
70
80
90
100
-------
FIGURE
50
Ul
DC:
or
I
CO
«=c
30
20
10
0
LABORATORY FLOTATION RESULTS - ILLINOIS NO, 6 SEAM COAL
A - 100 MESH
o - 150 MESH
D - 325 MESH
O
10
20
30
40 50 60
% COAL RECOVERY
70
80
90 100
-------
o
30
o- PYRITE
x- ASH
CK:
LLJ
>
CD
t_>
LU
20
10
FIGURE 5. LABORATORY
HGMS RESULTS
ILLINOIS NO. 6 SEAM -325 MESH
o
o
0
% COAL RECOVERY
J L
50
60
70
80
90
100
-------
FIGURE 6
SELECTIVE AGGLOMERATION
HARTSHORNE SEAM COAL
-200 MESH
lOOi
90
80J
70
60-1
o
50-
30
20-
10-
0
5 m is zo
% OIL (Wi, OIL/WT, COAL)
558
-------
FIGURE 7
GENERALIZED
FLOWSHEET
RAW
GR
COAL
ND
FLOTATION
CONC,
HIGH GRADIENT
MAGNETIC SEPARATION
CLEAN COAL
(NON-MAG,)
-TAILS
-TAILS
(MAG,)
559
-------
Ul
o\
o
_
LU If
Di -LI
C/3
«=c
UJ
h—
ce
FIGURE 8, LABORATORY RESULTS
COMBINED FLOTATION - HGMS
ILLINOIS NO, 6 SEAM
-325 MESH
o - PYRITE
x- ASH
70
80 I COAL RECOVERY
90
-------
FIGURE 9
PILOT PLANT
FLOWSHEET A
U'FLOW
MIBC_
CAUSTIC
•O'SIZE-
— NON-MAGS-
CLEANED COAL
RAW COAL, -1/2"
ROD MILL
BALLMILL
CYCLONE
J*
O'FLOW
25% +325 MESH
.ROUGHER FLOT.
FROTH
CL
lANERFLQL
FROTH
SWECO SCREEN (200 MESH)
U'SIZE
HIGH GRADIENT
MAGNETIC SEPARATOR
L
561
-TAILS
-TAILS
TO WASTE
-------
FIGURE Iff
SIZE DISTRIBUTIONS OF FINELY GROUND COAL
N3
12
10
~ 8
UJ
o - BATCH LABORATORY GRIND
x - CLOSED CIRCUIT PILOT PLANT GRIND
5 10 20
PARTICLE SIZE - MICRONS
50
-------
FIGURE 11
35
30-
£.25
LU
UJ
Oi
in
0*
UJ
20
QCL
O_
15
10
50
PILOT PLANT RESULTS ILLINOIS NO, 6 SEAM
FLOWSHEET A
o - PYRITE
x- ASH
®- PYRITE, ASH REC., LAB RESULTS
60
70
% COAL RECOVERY
80
90
-------
FIGURE 12
35
30
Ul
>-25
ce^
LU
>
o
§20
cc:
o_
6-5
15
10
50
PILOT PLANT RESULTS
PITTSBURGH NO, 8 SEAM
FLOWSHEET A
o - PYRITE
x - ASH
®- PYRITE, ASH REC., LAB RESULTS
1
60
70
I COAL RECOVERY
80
90
-------
FIGURE 13
I
BALL MILL •'
r;
U'FLOW !
' t
CYCLONE
PILOT PLANT
FLOWSHEETS B&C
CYCLONE O'FLOW
27% +325MESH
DSM SCREEN. 0.002" SLO
'O'FLOW
13% +325MESH
r-TAILS—
TO WASTE
60% +325MESH
U'SIZE
13% +325MESH
ROUGHER PLOT.
MIBC-
NAOH
ROUGHER PLOT.
CLEANER PLOT.
-TAILS—i
CLEANER PLOT.
-TAILS—
J
SWECO SCREEN (200 MESH)
1 T
O'SIZE U'SIZE
HIGH GRADIENT
MAGNETIC SEPARATOR
-NON-MAGS-
•MAGS-
CLEANED COAL
TO WASTE
565
-------
FIGURE
35
30
25
UJ
0£.
% 20
-------
FIGURE 15
EXAMPLE OF PILOT PLANT
MATERIALS BALANCE-
ILLINOIS NO, 6 SEAM COAL
CYCLONE O'FLOW
ASSAY %
COAL 69,1
PYS 1,04
ASH 29,3
FLOT/
1
CONCENTRATE
ASSAY".* DIST, %
COAL 89,6 94,5
PYS 0,63 44,5
ASH 9,4 23,4
*
200 MESH
Q'S*IZE
ASSAY % DIST, %
COAL 91,4 17,9
PvS 0,37 4,8
ASH 8,0 3,7
< NDN MAGNETICS
^ liwll llnuiiuiiuo
ASSAY % DIST, %
DIST,
100
100
100
TION
I
COAL
PvS
ASH
SCREEN
COAL
PYS
ASH
COAL 91,7 67,1 COAL
PYS 0,43 21,1 PYS
ASH 7,64 13,2 ASH
i
CLEAN 567
COAL
%
TATI TMT —
1 Ml LINO
ASSAY % DIST, %
14,0 5,5
2,12 55,5
82,8 76,6
U'S'IZE
ASSAY % DIST, %
89,2 76,6
0,70 39,7
9.7 , 19,7
HGMS^
MAPMCTTTO _
NAUNL 1 1 Co »
ASSAY % DIST. %
74,3 9,5
2.20 18,6
21,6 6,5
TO WASTE
-------
THE POTENTIAL OF MAGNETIC SEPARATION IN COAL CLEANING
Frederick V. Karlson1, Kenneth L. Clifford2,
William W. Slaughter2, and Horst Huettenhain^
Bechtel Corporation
P. 0. Box 3965
San Francisco, California 94119
2
Electric Power Research Institute
P. 0. Box 10412
Palo Alto, California 94303
3
Bechtel National, Inc.
P. 0. Box 3965
San Francisco, California 94119
ABSTRACT
This paper presents estimated costs for application of wet, high-gradient
magnetic separation (HGMS) technology to fine coal cleaning. Coal pyritic
sulfur and, commonly, an appreciable amount of other coal mineral matter is
weakly magnetic. If pyritic sulfur and other magnetic mineral matter can be
liberated by size reduction (minus 100 mesh or less), HGMS is a possible
method for subsequent, efficient separation of coal from liberated, weakly
magnetic impurities.
Presented in the paper is a possible flow scheme for use of wet, HGMS as
a fine coal cleaning circuit in a plant cleaning coarser coal by conventional
methods. Operations included in the flow scheme, in addition to HGMS, are
wet coal grinding, clean coal mechanical dewatering, and coal refuse thickening.
Estimated order-of-magnitude capital and operating costs are presented for wet,
HGMS circuits designed to clean 75 to 310 tons/hour of coal.
Also briefly discussed are several types of HG magnetic separators which
might be used for coal cleaning and current HGMS coal cleaning research.
568
-------
High Gradient Magnetic Separation (HGMS) is a relatively
new solids separation technology. It can be used to separate
weakly magnetic, as well as strongly magnetic, solids down to
extremely fine particle sizes — on the order of microns. Coal
pyritic sulfur and, commonly, appreciable amounts of other
coal mineral impurities are magnetically distinguishable from
pure coal. (Coal pyrite and some other mineral impurities are
paramagnetic while pure, organic coal is diamagnetic [Oder,
1976].) In many coals, both pyritic sulfur and other minerals
are present as very fine, widely disseminated particles. Pro-
vided these impurities can be liberated through size reduction,
HGMS offers a potential method for their subsequent, efficient
separation from coal.
This paper presents one possible flow scheme for applica-
tion of HGMS to fine coal cleaning. The scheme is strictly
conceptual; but is based on coal HGMS research sponsored by
the Electric Power Research Institute (EPRI).
HGMS BACKGROUND
HGMS is currently used commercially to beneficiate
kaolin clay. High quality kaolin is used principally as fill-
ing and coating agents in the manufacture of paper products.
In these applications its brightness is of major importance.
To achieve required product quality, naturally occurring kaolin
must be cleaned to remove mineral impurities called colorbodies.
In many kaolin deposits, the colorbodies are very fine (some
569
-------
colorbodies are less than a micron in size), weakly magnet
minerals (anatase, rutile, iron pyrite, mica, etc.) which
present in dilute concentrations.
Kaolin can be cleaned by a variety of wet chemical leach-
ing and physical cleaning processes, However, thefie proce*mas
are sufficiently complicated and expensive that the kaolin
industry has expended considerable research effort to develop
less expensive beneficiation methods. One relatively recent
result of this research has been successful development of
high-capacity, high gradient magnetic separators. These wet
(water slurry) separators can remove colorbodies from kaolin
typically classified to a size consist of 80 to 90 percent
minus two microns with high kaolin recovery (Oder, 1976).
The kaolin separators can be used alone to achieve modest color
improvement or in combination with chemical leaching to pro-
duce high-grade kaolin which can compete with more expensive
paper pigment materials.
The success of HGMS as a method for kaolin beneficiation
and the high gradient magnetic separator designs which are respon-
sible for this success are major reasons for emergence of HGMS as
a new solids separation technology. Coal cleaning, with emphasis
on desulfurization, is only one area where the feasibility of
HGMS is currently being evaluated^ Other areas include indus-
trial and municipal wastewater treatment, beneficiation of
570
-------
oxidized taconite iron ores, removal of participates from gas
streams, uranium ore beneficiation, and biological separations
Recent U.S. HGMS coal cleaning research has primarily
been sponsored by the U.S. Bureau of Mines (USBM) before crea-
tion of the Department of Energy (DOE), DOE, and EPRI. USBM-
DOE contract research has included (Hucko, 1978):
• A laboratory scale assessment by General
Electric Company in conjunction with the
Massachusetts Institute of Technology (MIT)
and Eastern Associated Coal Corporation of
the HGMS, wet and dry cleanability of a
Pennsylvania Upper Freeport coal. (For
separations performed on coal-water slurries,
pyritic sulfur removals of 60 to 80 percent
and ash removals of approximately 50 per-
cent were achieved [ Luborsky, 19771.)
• Laboratory assessment by the Naval Ordnance
Station, Indian Head, Maryland of removal
of pyritic sulfur by HGMS treatment of
pulverized coal-oil slurries.
• Evaluation by MIT of the technical feasi^
bility of using HGMS for recovery of magnetite
from conventional, dense-medium coal cleaning
circuits.
571
-------
Currently, DOE is sponsoring evaluation of dry, HGMS,
pulverized coal cleaning at Oak Ridge National Laboratory
(ORNL) and wet, HGMS coal cleaning at Michigan Technological
University. An initial phase of the ORNL evaluation included
benchscale testing of a high gradient magnetic separator con-
cept in which a fluidized bed is created within a separator.
This testing was performed for ORNL by Auburn University. The
work being performed at Michigan Technological University
involves pilot-scale evaluation of combined froth flotation
and HGMS for fine coal cleaning (Freyberger, 1978). DOE is,
also, at its Bruceton, Pennsylvania Research Center performing
a comparison of wet HGMS and two stage coal—pyrite flotation
for cleaning of minus 35 mesh coal (Hucko, 1978).
EPRI's magnetic coal cleaning evaluation program involves
processes which use HGMS and conventional, high intensity mag-
netic separation. It consists of the following five contracts
« Sala Magnetics Incorporated - Wet, HGMS
benchscale and pilot-scale testing of five
Appalachia coals using iron enclosed, high
gradient magnetic separators
• Magnetic Corporation of America (MCA) —
Wet and dry HGMS benchscale testing of a
Pennsylvania Upper Freeport seam coal and
an Ohio Illinois No. 6 seam coal using a
superconducting, high gradient magnetic
separator
572
-------
Colorado School of Mines Research Institute
(CSMRI) — Benchscale research on the technical
feasibility of using aqueous and organic
base magnetic fluids to preferentially in-
crease the magnetic susceptability of either
coal mineral matter or pure coal. High
gradient magnetic separators were used to
perform coal-mineral matter separations
following magnetic enhancement of either
coal or mineral matter.
Bechtel National, Inc. - Development of com-
parable, commercial conceptual designs for
HGMS coal .cleaning plants based on the
coal HGMS tests performed for EPRI by Sala
Magnetics and MCA. The Bechtel contract,
also, includes preparation of capital and
operating cost estimates for these two HGMS
coal cleaning plants and comparison of
these estimated costs with those for a con-
ceptual, conventional, commercial coal cleaning
plant. (The HGMS fine coal cleaning circuit
conceptual design presented in this paper is
based on the HGMS, coal cleaning plant, concep-
tual designs being developed for EPRI.)
573
-------
• Nedlog Technology Group — Evaluation of
conventional, induced roll, high intensity
magnetic separators for use in the Magnexx_7
Process. The Magnex^? Process, invented at
Hazen Research Incorporated and being
developed by Nedlog Technology (Porter,
et al., 1978), treats coal with iron
carbonyl to significantly enhance the mag-
netic susceptability of both coal pyrite
and other mineral matter relative to pure
coal.
All five EPRI magnetic coal cleaning contracts are cur-
rently nearing completion. Final reports should be available
at the end of 1978 or early in 1979.
HIGH GRADIENT MAGNETIC SEPARATOR DESIGNS
A large number of different magnetic separator designs,
built and proposed, are referred to as high gradient magnetic
separators (lannicelli, 1976). However, only three high
gradient designs are briefly described in this paper. These
designs represent those used commercially for kaolin beneficia-
tion and by Sala Magnetics and MCA in their coal cleaning
evaluations. All three designs generate high gradient magnetic
fields through use of magnetic stainless steel matrices. The
magnetic gradients actually produced in these separators are
normally well in excess of 1,000 kilogauss per centimeter.
574
-------
Actual gradient magnitudes depend on both separator magnet
design and matrix design.
Figure 1 shows the main components for the type of wet,
HG magnetic separator currently used in the kaolin industry
and used by Sala Magnetics in their benchscale coal tests. It
consists of a solenoid electromagnet completely enclosed in
an iron box. The box provides a return circuit for the mag-
netic flux generated by the electromagnetic coils. Small pipe
penetrations in the box top and bottom permit transport of
slurry in and out of the solenoid cavity. This cavity is lined
with a thin wall canister inside of which is packed fine, fer-
romagnetic stainless steel wool. When the electromagnet is
energized, high magnetic gradients are induced on the individual
wool matrix filaments. Strong and weakly-magnetic particles
passing through the matrix are, due to the magnetic gradients
around the filaments, attracted to and captured on the
filament surfaces. Actual volume of the separator canister
occupied by matrix filaments is quite small; usually less than
CLEANED
J ([EFFLUENT
BED
WOOL STRANDS
NON MAGNETIC
PARTICLE
FEED
MAGNETIC PARTICLES •
Figure 1. Vertical Section through the Center of a
Batch (Cyclic), Wet High Gradient Magnetic Separator
575
-------
5 percent. Because of this, comparatively high feed rates at
low-pressure drops can be achieved with this separator design.
Captured magnetic particles are removed from the separator by
deenergizing the magnet and flushing. This cleaning method works
well for applications, such as kaolin beneficiation, where the
magnetic solids in the separator feed are present in low con-
centrations. For processing of solids in which the fraction of
magnetics present is appreciable, such as in coal, a continuous
rather than a batch (cyclic) separator is required. Cycle
times for batch separation of coal would be so short that numer-
ous batch, HGMS separators would be required.
Figure 2 shows a continuous operation, commercial, wet,
high gradient magnetic separator manufactured by Sala
Magnetics. (For a discussion of the development of this sep-
arator, see Arvidson, et.al., 1976.) This is the type of
separator used by Sala Magnetics in their EPRI pilot-scale,
coal cleaning tests. And, it is similar to the separator used
in the HGMS fine coal cleaning circuit presented later in this
paper. The basic operating principles of this separator are
identical to those described for the batch HGMS separator.
However, the capture matrix, the ring structure in Figure 2,
revolves through an electromagnet rather than remaining
stationary inside of a solenoid electromagnet. When a section
of the matrix is within a magnetic head (an individual electro-
magnet assembly), high magnetic gradients are induced around
its surfaces and magnetic particles are captured. When
576
-------
the matrix moves out of a magnetic head, the induced magnet
gradients rapidly decay and captured magnetics are flushed from
the matrix. In order to remove mechanically entrained non-
magnetic particles, a rinse is applied before the matrix exits
a magnetic head, i.e., while the magnetic capture forces are
still present. Clear water is, usually, used for both matrix
flushing and rinsing. Matrix revolution rate is on the order
of 1 rpm.
Figure 2. Sala Magnetics Incorporated Carousel,
Continuous High Gradient Magnetic Separator
577
-------
The carousel separator design shown in Figure 2 is built
with multiple magnetic heads when required separation rates
exceed the capability of a single head. A wide variety of
matrix designs can be used with this separator. Usually, these
are fabricated using various sizes of ferromagnetic stainless
steel wool or some form of crimped or otherwise formed ferrogmag-
netic, stainless steel, expanded metal. Actual selection of a
matrix material and its configuration depends on separation
application. Electromagnets used in the carousel separator
consist of double, saddle wedge shaped, water cooled, copper
coils. Except at the ends through which the matrix passes,
these coils are encased by iron. Special seals at the magnetic
head open ends are used to control slurry leakage. They also
allow use of pressurized water for matrix rinsing and flushing.
A third high gradient magnetic separator design that uses
a capture matrix is one which employs a superconducting magnet
for generation of high intensity magnetic fields. A bench-
scale, cyclic version of such a separator was used by MCA in
the HGMS coal cleaning tests they performed for EPRI. The
matrix used in these tests was magnetic, stainless steel wool.
Currently, MCA is developing a commercial scale, super-
conducting, HG magnetic separator. The collection matrix for
this separator consists of multiple, linearly connected sections,
578
-------
These sections reciprocate in and out of a cylindrical super-
conducting magnet. When a matrix section within the magnet
becomes fully loaded, it is pushed or pulled out and another
takes its place. After leaving the magnet, fully loaded matrix
sections move into an iron magnetic shield which results in
reduction of the magnetic field gradients in the matrix to
nearly zero values. The sections are then water flushed, after
which, they are ready for reinsertion into the magnet.
To achieve and maintain the temperature (4.2°K) required
for operation of a superconducting magnet, liquid helium is
required for coil cooling. Liquid nitrogen may, also, be re-
quired for operation of the helium refrigeration system. Handl-
ing and storage of both liquid helium and nitrogen would be new
to the coal cleaning industry.
Potential advantages of a superconducting, HG magnetic
separator compared with those which use iron enclosed electro-
magnets are use of considerably higher magnetic fields and
reduced power consumption. The higher magnetic fields (on the
order of 50 kilogauss) which can be generated by superconducting
magnets can be used to (Stekly, 1975):
• Separate more weakly magnetic or finer
particles
• Obtain performance comparable to lower
field strengths but at increased feed
rates
• Increase matrix loading
579
-------
Power consumption may be lower for superconducting magnetics
because essentially zero power is required to generate the
magnetic field. Significant amounts are, however, required for
operation of the magnet refrigeration system. The real advan-
tages of large superconducting, HG magnetic separators will
only become fully apparent after one is actually constructed
and operated.
WET, HIGH GRADIENT MAGNETIC SEPARATION FINE COAL CLEANING
CIRCUIT DESCRIPTION
Figure 3 presents a conceptual flow scheme and a design
material balance for application of wet HGMS to fine coal
cleaning. (Numbers in parenthesis after the equipment titles
are the number of actual pieces required of a specific equip-
ment type.) For actual use, this fine coal cleaning circuit
would be combined with other cleaning circuits and various
coal handling-processing operations (crushing, conveying, thick-
ening, dewatering, drying, storage, etc.) to form a complete
coal cleaning plant. Coal cleaned in the HGMS circuit could
come from a coarse cleaning circuit, a conventional fine clean-
ing circuit, or, possible, from a size reduction operation.
The circuit could be used to replace or supplement such conven-
tional fine coal cleaning methods as fine heavy media cyclones,
water cyclones, and froth flotation.
580
-------
BYPASS
MAGNETIC SEPERATOR COIL
COOLING HEAT EXCHANGER (2)
m
in
z
m
O
O
o
o
WATER FROM \
TREATMENT >
PLANT /
FOUR HEAD HIGH GRADIENT
MAGNETIC SEPERATOR (2)
FLOCCULANT
SOLUTION
SODIUM
SILICATE ,
SOLUTION /
ULASblMtU PULVER
BALL MILL
CLASSIFYING
CYCLONE
(88)
COAL DISPERSION
TANK 12)
CLEAN COAL
RECEIVING
TANK (21
CLEAN COAL
VACUUM FILTER 110)
BALL MILL
DISCHARGE
TANK (2)
COAL THICKENER
(II
COAL REFUSE
THICKENER (
TO WATER
TREATMENT
COAL
THICKENER
OVERFLOW
TANKdl
COAL REFUSE THICKENER
OVERFLOW TANK(I)
ULVERI2ED
COAL
TO REFUSE\
DISPOSAL )
STREAM NO.
SOLIDS FLOW
-TONS/HOUR
TOTAL FLOW
-TONS/HOUR
TOTAL FLOW
-GALLONS/MINUTE
SOLIDS CONCENTRATION
-WT. PERCENT
<£>
310
365
-
85
<2>
_
666.75
2,667
-
^
_
2,973.25
11,893
-
<6>
465
1,162
4,098
40
^
775
5,167
19,751
15
<§>
465
1,162
4,098
40
<2>
310
4,005
15,653
7.7
<5>
-
US
5
-
<$>
310
1031
3,765
30
<%>
-
T\
10
-
O
310
1 01S5
3,775
29.94
<$>
_
708.25
2,833
-
<§>
_
708.25
2,833
-
<$>
46.5
1,226.5
4,823
3.79
<^
263.5
1,225.5
4.618
21.5
^6>
-
1
4
-
4^
263.5
366
1,180
72
4>
—
860.5
3,442
-
<$>
—
0.2
0.8
-
<$>
46.5
132
449
35
€>
—
1,034.7
4,375
-
<£>
—
1.9&B.2
7,817
-
<8>
-
2,083.25
8,333
—
O
8
m
m
-------
In the HGMS circuit, coal is initially ground wet to a
fine size consist to liberate mineral matter. It is then
cleaned in HG magnetic separators. Clean coal is dewatered
by vacuum filtration. Clean, dewatered coal from an HGMS
circuit would likely be combined with other cleaning plant
clean coal streams and thermally dried. Magnetic cleaning
refuse is thickened to recover separator rinse and flush water,
This refuse could be further dewatered if required. The mag-
netic refuse would likely be combined with other cleaning
plant refuse before ultimate disposal.
Design operating requirements established for the wet,
HGMS fine coal cleaning circuit are presented in Table 1.
Table 1
WET HGMS FINE COAL CLEANING CIRCUIT
DESIGN REQUIREMENTS
Circuit Design Capacity 310 tons/hour of coal
on a dry basis
Feed Coal Conditions
Source Appalachia
Size Consist 14 mesh x 0
Total Moisture 15 wt percent
HG Magnetic Separator
Feed Conditions
Coal Size Consist 70 percent minus 200 mesh
Coal Slurry
Concentration 30 wt percent coal
Clean Coal Requirements
Coal Size Consist 70 percent minus 200 mesh
Coal Moisture 28-35 wt percent
582
-------
The design, dry coal circuit feed rate, 310 tons per
hour, is identical to that for complete, HGMS, conceptual,
coal cleaning plant designs being developed for EPRI. The
HG magnetic separator feed conditions result from the HGMS
coal cleaning tests performed by Sala Magnetics and MCA.
Fine Coal Grinding
Two basic approaches were considered for grinding coal
to 70 percent minus 200 mesh, an unprecedented size consist
for coal cleaning: dry and wet grinding. Dry coal grinding
(pulverizing) is a well established technology. All pulverized
coal required for utility power plant steam generators, large
industrial steam generators, and many other coal-fired pro-
cesses is pulverized in dry coal mills. As far as known, wet
coal pulverizing has never been used in any large scale, com-
mercial operation. There simply has never been any require-
ment for large scale, wet coal pulverizing. Large scale, wet
grinding, however, is widely used in mineral ore beneficiation
operations.
Despite the fact that dry coal grinding is a well devel-
oped technology compared with wet grinding, wet grinding was
selected for the conceptual, wet, HGMS fine coal cleaning cir-
cuit. The main reason for this selection is that for a wet
process, wet grinding is believed to be considerably less
expensive than dry grinding. This conclusion is based on
583
-------
considerable experience in design of large scale, dry, coal
grinding facilities for conceptual, commercial coal conversion
plants (Bechtel, 1977 and Karlson, et al., 1978).
Coal fed to the HGMS fine coal cleaning circuit is first
slurried with water in two grinding mill discharge tanks; one
tank per grinding train (see Figure 3). It is then pumped to
a series of parallel, 10-inch-diameter cyclones for size class-
ification. These cyclones are designed to product an overflow
containing 70 percent minus 200 mesh coal. Oversize coal from
the cyclones, the underflow stream, flows by gravity to two
ball mills for wet grinding. Coal slurry exiting the ball
mills returns by gravity to the mill discharge tanks where it
mixes with the feed coal. From here it is pumped to the class-
ifying cyclones. Each of the two wet grinding mills selected
for the HGMS cleaning circuit has a nominal diameter of 16 feet
and is 27 feet long. Each mill is equipped with a 4,500-hp drive
In order to produce coal sized to 70 percent minus 200
mesh, the classifying cyclones must be fed a slurry containing
on the order of 15 wt percent coal. If these cyclones are
fed slurry containing a significantly higher coal concentra-
tion, classification will deteriorate and large quantities of
oversize coal will be misplaced into the overflow. Also,
excessive minus 200 mesh coal will be misplaced to the under-
flow and be overground. The 15 wt percent cyclone coal feed
concentration limit is only an estimate. As far as could be
584
-------
determined, coal ground to 70 percent minus 200 mesh has
never been classified in commercial scale cyclones. Through
development work it may be possible to increase the limiting
cyclone feed concentration somewhat.
For a feed coal concentration of 15 wt percent, the over-
flow (sized coal) from the classifying cyclones will have an
estimated coal concentration of 8 wt percent coal. This con-
centration is considerably lower than those proposed by both
Sala Magnetics and MCA for wet, HGMS coal cleaning; 30-35 wt
percent coal. To achieve coal slurry concentrations in this
range, thickening is required.
As for efficient classification of coal at 70 percent
minus 200 mesh, no performance data was found for thickening
of coal this fine. Therefore, as part of the EPRI HGMS coal
cleaning evaluation program, a series of coal thickening
tests were performed by Envirotech Corporation. Pulverized
coal used in these thickening tests was an Eastern bituminous
coal provided by TVA. It was collected from a power plant
steam generator coal mill.
The HGMS fine coal cleaning circuit thickener design is
a direct result of the Envirotech thickening tests. Design
thickener loading is 3 square feet/ton-day. This results in
a single 175-foot-diameter thickener. To achieve the design
loading, use of an anionic flocculant is required. Design
flocculant addition rate is 0.02 pounds/ton of dry coal.
585
-------
Wet HGMS Coal Cleaning
Before pulverized coal is fed to HG magnetic separators,
it is treated with a dispersant. The dispersant used in the
Figure 3 flow scheme is sodium silicate. Design addition
rate is 1.8 pounds/ton of dry coal. Sodium silicate is
received in rail cars as anhydrous briquettes. These briquettes
are dissolved in a pressure dissolver using steam. Dissolved
sodium silicate is fed to the coal dispersion tanks as a
10 wt percent solution.
The conceptual, HGMS, fine coal cleaning circuit
uses a new, continuous HG magnetic separator being developed by
Sala Magnetics. This separator is similar to the one shown
in Figure 2. In the new design, however, the capture matrix
is a narrow width, segmented belt rather than a ring. This
continuous belt matrix is driven by sprockets or rollers
located at its ends. In side view the arrangement is similar
to that of a metal tractor tread. Magnetic heads, similar in
design to the head shown in Figure 2, are mounted along the
top and bottom of the matrix belt. Matrix belt width is
approximately 5.25 feet.
Two of the new, Sala Magnetics1 linear type, HG magnetic
separators are required to clean, with some spare capacity,
310 tons/hour of dry coal as a 30 wt percent slurry. Each
separator has four magnetic heads, is approximately 25 feet
high, and 43 feet long. Power requirement of each magnetic
head is 250 kW.
586
-------
As shown in the Figure 3 material balance, substantial
quantities of water are required for both matrix rinsing and
flushing; 2,833 gallons per minute for each operation. Approxi-
mately one-third of the rinse water reports to the clean coal
product. The remainder is combined with the flush water.
For the design material balance, a clean coal weight
recovery of 85 percent was assumed. This is consistent with
results obtained by Sala Magnetics in pilot tests using coal
that had been cleaned at a specific gravity of 1.6. Recoveries
obtained by Sala Magnetics, MCA, and CSMRI on uncleaned, Eastern
bituminous coals were, however, usually less than 85 percent.
Coal weight and Btu recoveries, as well as, coal sulfur and
ash removals, were found to be very dependent on the coal,
coal size consist, separator matrix design, and separator
operating conditions. A considerable amount of additional
test work is required to establish for even selected coals
the limits to which they can be cleaned by HGMS and at what
recoveries.
Clean Coal Dewatering
After HGMS, clean coal is dewatered with vacuum filters.
Vacuum filtration of 70 percent minus 200 mesh coal, like its
classification and thickening, is an area where only limited
design information is available. As a result, a cursory evalu-
ation of pulverized coal filtration was performed for EPRI
by Envirotech.
587
-------
In the filtration tests, the coal sample used for the
thickening tests was thickened using a flocculant, treated
with sodium silicate dispersant, agitated, and then filtered
using standard leaf filtration test procedures. Figure 4
summarizes the results of these filtration tests. A minimum
surface moisture of 25 wt percent was achieved at a filtration
rate of 20 pounds/hour-square foot. Further, the curve pre-
sented in Figure 4 shows that cake surface moisture concentra-
tion is relatively insensitive to filtration rate. A doubling
of the filtration rate from 20 to 40 pounds/hour-square foot
causes surface moisture to increase to only 27.5 wt percent.
Table 2 presents design data established for the wet,
HGMS fine coal cleaning circuit filtration operation. Appli-
cation of this design data results in a filtration system
which consists of 10 disc filters. Each unit has 15 discs
which are 12.5 feet in diameter. Total filtration area is
33,000 square feet.
Filter cake is collected on two belt conveyors. The dis-
charge ends of these conveyors is the clean coal outlet bound-
ary of the HGMS fine coal cleaning circuit. From here clean
coal would likely be dried or agglomerated and then dried.
588
-------
30
21
S"
§
24
20 40 10 00
FULL SCALE FILTNATION HATE
100 120
Figure 4. Effect of Filtration Rate on Coal
Surface Moisture
589
-------
Table 2
WET HGMS FINE COAL CLEANING CIRCUIT
FILTRATION OPERATION DESIGN DATA
Feed Coal Concentration 20 wt percent
Minimum Surface Moisture 25 wt percent
o
Design Filtration Rate 20 Ib/hr-ft
Estimated Filtrate Solids Less than 0.5 wt percent
Vacuum 22 in. HG
Filter Type Disc
Filter Cloth Polypropylene
Polymer Type Anionic
Polymer Dose 0.02 Ib/ton
WET, HIGH GRADIENT MAGNETIC SEPARATION FINE COAL CLEANING CIRCUIT
ESTIMATED ORDER-OF-MAGNITUDE CAPITAL AND OPERATING COSTS
Order-of-magnitude capital and mean annual operating
costs have been estimated for the flow scheme presented in
Figure 3. These estimates are derived from estimates devel-
oped for EPRI for complete, conceptual, HGMS coal cleaning
plants. The EPRI estimates were developed to determine general
cost levels for application of HGMS to pulverized coal cleaning
and to identify cost sensitive HGMS cleaning operations. The
estimates presented in this paper, as well as those developed
for EPRI represent an attempt to project possible economics
for wet, HGMS coal cleaning for the purpose of planning future
development programs. In using these cost estimates it should
590
-------
be recognized that application of HGMS to coal cleaning is new.
Indeed, consideration of cleaning coal purposely ground to
200 mesh or finer is new.
Order-of-Magnitude Capital Cost Estimate
Table 3 presents an estimated order-of-magnitude capital
cost. Estimated total capital cost including indirect costs
and an allowance for owners' costs is $44.95 million. The
estimate is based on:
o The conceptual process flow scheme presented
in Figure 3
o Order-of-magnitude quotations for major
equipment from manufacturers
o Actual cost information for relevant,
completed Bechtel projects
• A major equipment list
• Major motor and HG magnetic separator
power requirements
• Estimated building volumes
At the direct cost level ($30.62 million), 25.7 percent is
associated with coal grinding and 74.3 percent with HGMS coal
cleaning, clean coal dewatering, and refuse thickening. Esti-
mated cost of the HG magnetic separators was provided by Sala
Magnetics. Their estimated cost for a complete, HG magnetic
separator facility is $11.0 million. This cost includes equip-
ment (two 4-head HG magnetic separators, separator drives,
591
-------
Table 3
ESTIMATED ORDER-OF-MAGNITUDE CAPITAL COST
WET HIGH GRADIENT MAGNETIC SEPARATION FINE COAL CLEANING CIRCUIT
(All Costs Are For Mid-1978 In Thousands Of Dollars)
DIRECT COSTS
Mechanical Equipment
Piping and
Instrumentation
Electrical
Civil and
Architectural
Direct Cost Subtotal
Percent of Direct Cost
Coal Coal HGMS and
Grinding 'Filtration
Operation Operation
Total
4,520
14,850
19,370
Percent of
Total
43.1%
660
1,850
850
7,880
25.7%
1,880
4,490
1,520
22,740
74.3%
2,540
6,340
2,370
30,620
100.0%
5.6%
14.1%
5.3%
68.1%
INDIRECT COSTS
Engineering, Project
Services, and Fees
Spare Parts
Sales Taxes
Contingency
CONSTRUCTED COST SUBTOTAL
Allowance for Owners' Costs
TOTAL CAPITAL COST
5,760
330
730
5,370
42,810
2,140
12.8%
0.7%
1.6%
12.0%
95.2%
4.8%
44,950
100.0%
592
-------
a d.c. power supply system, a closed-circuit magnet cooling
system, controls, and some separator piping), equipment installa-
tion, and startup services. The cost for the complete magnetic
separator facility was treated as a subcontract. It is part of
$14.85 million estimated for mechanical equipment required by
the coal HGMS and filtration operation.
Design coal feed capacity for the HGMS cleaning circuit
is 310 tons of dry coal/hour. Therefore, on an hourly feed
capacity basis, estimated total capital cost is $145,000 per
ton per hour of feed coal. This is considerably higher than
the unit cost for any conventional coal cleaning circuit or
complete plant. However, most conventional circuits clean
much coarser coal. Because of this difference, any comparison
of costs for conventional circuits and circuits which clean
pulverized coal should consider the final form in which the
clean coal is used. If the final form is pulverized coal, a
cost credit should be assigned to a pulverized coal cleaning
circuit to account for the fact that no or only limited addi-
tional coal pulverizing is required before coal use.
To determine the approximate effect of circuit design
feed rate on estimated capital cost, the Table 3 estimate was
used to factor very approximate capital cost estimates for
circuits having design feed capacities of 155 and 75 tons of
dry coal per hour. These estimates were used to calculate an
exponential design capacity-capital cost scale factor. The
estimated value of this exponential scale factor is 0.8.
593
-------
Order-of-Magnitude Operating Cost Estimate
Estimated order-of-magnitude mean annual operating costs
for the wet, HGMS fine coal cleaning circuit are presented in
Table 4. Total estimated mean annual operating cost is $6.37
million including depreciation and $4.23 million excluding
depreciation. On an annual dry, coal feed rate basis (1.008
million tons per year), estimated total annual operating cost
is $6.32 per design annual ton including depreciation and $4.20
per design annual ton excluding depreciation.
The term mean annual operating cost is used to imply that
the operating costs are mean annual estimates for a 20-year
coal cleaning plant life. Ideally, multiplication of these
estimates by 20 results in an estimate of all operating costs,
excluding working capital, which would be incurred over the
entire 20-year plant life. Actual operating costs for a
specific year could differ significantly from the mean annual
estimate. Some operating costs components, such as major
maintenance subcontracts, are not incurred every year of
plant operation.
The mean annual operating cost estimate is based on:
• Cleaning circuit operation of 13 hours per
day (two shift operation), nominally five
days per week
594
-------
Table 4
ESTIMATED ORDER-OF-MAGNITUDE MEAN ANNUAL OPERATING COSTS
WET HIGH GRADIENT MAGNETIC SEPARATION FINE COAL CLEANING CIRCUIT
Direct Operating Costs
Annual
Unit Costs
Annual Cost
Percent
49.8xl06kWh $0.02/kWb
1.52x109 gal $0.15/1,000 gal
Major Commodities
Steel Balls, Filter Cloth
Flocculant, Sodium Silicate
Labor
Supervision
Operating
Maintenance
Subtotal, Labor 46,592 hr $8.39/hr
Utilities
Electricity
Water Treatment
Subtotal, Utilities
Materials and Supplies
Operating Supplies
Chemicals
Maintenance Materials and
Contracts
Subtotal, Materials
and Supplies
Subtotal, Direct Operating
Costs
Indirect Operating Costs
Depreciation — Straight Line 20
Year of Constructed Cost
Direct Payroll Overhead-35% of
Direct Labor
Administration and General
Overhead-60% of Direct Labor
Insurance and Local Taxes -
2.75% of Constructed Cost
Subtotal, Indirect Operating Cost
Total Annual Operating Cost
Notes: (1) Actual annual operation is 3,250 hours.
rvequ-L J-CLiicii i_»
5,824 hr
31,616 hr
9,152 hr
$9.87/hr
8.17/hr
8.22/hr
u
$ 57,500
258,400
75,200
r iocai
0.9%
4.0%
1.2%
$ 391,100
$ 996,000
228.000
$1,224,000
136,900
234,700
6.1%
15.6%
3.6%
19.2%
$ 71,000
236,000
761,000
$1,068,000
$2,683,100
1.1%
3.7%
12.0%
16.8%
42.1%
$2,140,500 33.6%
2.1%
3.7%
1.177.300 18.5%
$3.689.400 57.9%
$6,372,500 100.0%
595
-------
• An annual operating time of 250 days
• Mid-1978 United Mine Workers of America labor
rates
u Operation of the HGMS cleaning circuit at its
hourly design coal feed rate
• For estimation of indirect operating costs,
that the owners of the complete coal cleaning
plant are a private corporation as defined by
United States federal tax law
Estimated constructed capital cost, rather than total
capital cost, was used to estimate both annual depreciation
and insurance and local taxes. The reason for this is that
the allowance for owners' costs portion of the total capital
cost might be expensed, rather than capitalized, during clean-
ing plant engineering-construction. The allowance for owners'
costs also includes land costs and possible other costs which
are not depreciable under current tax law. If all or a portion
of owners' costs are capitalized, the depreciation cost should
be adjusted.
596
-------
REFERENCES
, Coal Mine-Coal Conversion Plant Interface Evaluation
Design, prepared by Bechtel Corporation for the U.S. Energy
Research and Development Administration, November 1977, FE-2370-
16, Volumes I and II.
Arvidson, B., Oberteuffer, J. A., and Wechsler, I., "Continuous
High Gradient Magnetic Separation Pilot Plant: Machine Descrip-
tion and Mineral Processing Results," presented at the MMIJ-
AIME Third Joint Meeting, Denver, Colorado, September 1-3, 1976.
Freyberger, W. L., Keek, J. W., Spottiswood, D. W., Solem, N. D.,
and Doane, V. L., "Cleaning of Eastern Bituminous Coals by Fine
Grinding, Floth Flotation, and High Gradient Magnetic Separa-
tion," presented at the EPA Symposium on Coal Cleaning to Achieve
Energy and Environmental Goals, Hollywood, Florida, September
11-15, 1978.
Hucko, R, E., "DOE Research in High Gradient Magnetic Separation
Applied to Coal Beneficiation, presented at the International
Conference on Industrial Applications of Magnetic Separation,"
sponsored by The Engineering Foundation, Auburn University and
the IEEE Magnetics Society, Rindge, New Hampshire, July 30-
August 4, 1978.
lannicelli, J., "New Developments in Magnetic Separation," IEEE
Transactions on Magnetics, September 1976, Vol. MAG-12 No ~~5
pp 436-443.
Karlson, F. V., and Slaughter, W. W., "Process Design Considera-
tions for Wet Pulverized Coal High Gradient Separation," pre-
sented at the International Conference on Industrial Applications
of Magnetic Separation, sponsored by The Engineering Foundation,
Auburn University, and the IEEE Magnetics Society, Rindge, New
Hampshire, July 30-August 4, 1978.
Luborsky, F. E., High Gradient Magnetic Separation for Removal
of Sulfur from Coal, Final Report and Supplement to Final Report,
prepared by General Electric Company for the U.S. Bureau of
Mines, Contract No. HO 366008, February 1977 and September 1977,
SRD-77-041 and SRD-77-147.
Oder, R. R., "High Gradient Magnetic Separation Theory and Ap-
plications," IEEE Transactions on Magnetics, September 1976,
Vol. MAG-12, No. 5, pp 428-435.
597
-------
Porter, C. R., and Goens, D. N. , "Magnex Process: A Dry Process
for Removal of Pyrite and Ash from Coal," presented at the Coal
Preparation Workshop, Ohio University, Athens, Ohio, January 19-
20, 1978.
Stekly, Z. J. J., "A Superconducting High Intensity Magnetic
Separator," presented at the Thirteenth International Conference
on Magnetics, London, England, April 14-17, 1975.
598
-------
TESTING OF COMMERCIAL COAL PREPARATION
PLANTS WITH A MOBILE LABORATORY
William Higgins and Thomas Plouf
Joy Manufacturing Company
Denver, Colorado
No abstract or paper available.
599
-------
CHEMICAL COMMINUTION—AN IMPROVED ROUTE TO CLEAN COAL
Victor C. Quackenbush, Robert R. Maddocks,
and George W. Higglnson
Catalytic, Inc.
Center Square West
1500 Market Street
Philadelphia, Pennsylvania 19102
ABSTRACT
Coal destined as fuel for electric power generation is often cleaned.
A typical coal cleaning plant involves mechanical crushing followed by
heavy medium separation facilities to remove pyritic sulfur and reduce
ash content of the feed coal.
A new process to fracture coal by chemical comminution is presented
which permits efficient removal of pyritic sulfur and ash-forming components.
Raw coal is contacted with a low molecular-weight chemical such as ammonia
vapor or liquid in a reactor system at moderate pressures and temperatures.
The chemical disrupts the natural bonding forces acting across the internal
boundaries of the coal structure where the pyritic sulfur and ash deposits
are located. Because of the manner in which the pyrite is liberated,
higher yields of clean coal for the same level of sulfur removal, or lower
sulfur level at the same yield can be achieved using chemical fracturing
as compared to mechanical crushing or grinding. Less fines are generated
by chemically comminuting coal than by mechanical size reduction, thus
permitting a significant increase in the usable coal fraction.
The process is in the development stage and indicates that production
of a high-grade clean coal at considerable savings is possible.
600
-------
Chemical comminution is a new technology for fracturing coal.
Compared to mechanical grinding, in most cases it liberates more
ash and pyritic sulfur from various grades of coal without creat-
ing substantial amounts of fines that complicate follow-on separation
plants. Chemical comminution involves the exposure of coal to low-
cost, recoverable chemical agents that disrupt the natural binding
forces holding a lump of coal together. The concept, patented by
the Syracuse Research Corporation, is currently at the bench-scale
level of development, and a pilot plant is presently being considered.
This paper reviews the chemical comminution concept, describes
probable pilot plant approaches, and provides results of two de-
tailed economic studies. The economic studies compare chemical
comminution to mechanical fracturing from the standpoints of equal
coal feed and equal coal product yield. In each case chemical com-
minution provides favorable economic benefits. In Catalytic.1 s view
the values presented in terms of cost per million BTU's for the two
comparisons represent the range of those likely to be realized.
601
-------
FRAGMENTING COAL
Coal cleaning typically has two major steps: fragmenting
followed by a separation process.
The fragmentation of coal releases impurities such as ash
and pyrites. These impurities can be later separated from the coal,
usually by methods that depend on density differences, such as
heavy medium separation.
Traditionally, mechanical crushing has served as the frag-
menting mechanism. Mechanical crushing usually takes the coal
down to a 3/8-in. top size. Crushing fragments the coal and im-
purities in a random fashion, producing a wide spectrum of coal
particle sizes. A significant amount of coal ground to 3/8 in.
top size will result in fines below 100 mesh. Recovering these
fines is frequently uneconomical, so they are.often rejected with
the tailings.
Finer mechanical grinding releases more coal impurities, but
produces greater quantities of fines, complicating the separation
process and raising cleaning costs. The added separation costs for
handling fines limit the degree to which coal can be mechanically
ground to liberate ash and sulfur.
Chemical comminution offers an attractive alternative to
mechanical crushing for fragmenting coal. In this process the
coal is exposed to low molecular weight chemicals, such as gaseous
602
-------
or liquid ammonia. The chemicals do not react with the coal, but
disrupt the bonding forces holding coal particles together.
These natural boundaries among coal particles are often com-
posed of impurities such as ash and pyrites. So the chemical
mechanism, which is not completely understood, selectively attacks
and releases the coal impurities.
The size distribution of the chemically comminuted coal
depends on the natural fault planes within the coal, so the size
distribution depends on coal petrology. Bituminous coals are
most susceptible to chemical fracture, with susceptibility de-
creasing with increasing coal rank.
For the same amount of released impurities, chemically
comminuted coal samples generally consist of larger coal particles
compared to mechanically ground coal. Figure 1 compares particle
size distributions of mechanically crushed and chemically com-
minuted Illinois No. 6 coal. Notice that for any given sieve
designation, except for run of mine coal, chemical comminution
produces a greater percentage of oversize particle coal. These
larger particle sizes tend to reduce the cost of follow-on pro-
cessing for separating ash and impurities, compared to mechanical
crushing.
Chemical comminution not only produces larger particle sizes,
but also tends to release greater amounts of impurities of compar-
603
-------
able particle size distributions. Figure 2, for example, gives
the results of float-sink tests on samples of coal fractured by
chemical comminution and mechanical grinding. Taking, for example,
a cumulative sulfur level of 1.4 percent, mechanically ground samples
yield recovery levels of about 73 percent (above 100 mesh), while
chemically comminuted samples yield over 95 percent recovery
levels. So chemical comminution produces greater recovery yields
for a given sulfur level, or cleaner coal for a given recovery level,
PROCESS DESCRIPTION
The chemical comminution process consists of five basic steps,
as shown in Figure 3;
o exposure of coal to chemical agent
o washing coal to remove residual chemical
o dewatering the resulting coal slurry
o recovery of comminuting chemical
o clarification and recycle of wash water
The wet, fragmented coal then would be sent to a separation process
to remove released impurities such as ash and pyrites.
During the first step the degree of fragmentation depends
primarily on chemical exposure time, temperature, and pressure.
The degree of fragmentation increases with increasing exposure
time, leveling off after a period that depends on the nature of the
coal. Probably the chemical agent diffuses through the larger coal
boundaries first, but eventually reaches smaller connecting faults.
604
-------
So penetration speeds depend on fault sizes. Maximum fragmentation
for a low volatility bituminous coal, for example, occurs during
about one and a half hours exposure to gaseous ammonia, as indicated
in Figure 4.
Temperature and pressure also affect fragmentation because
they affect such variables as diffusivity rates, surface tension,
and similar properties that influence the chemical's rate of diffu-
sion through the coal's capillary fault channels.
Chemical agents could be exposed to coal using either a batch
or continuous approach. In a typical batch cycle, a vessel is
charged with coal, sealed, and evacuated to remove air. The vessel
is then pressurized in two steps with a chemical agent such as
ammonia. The first step takes the vessel up to 60 psia by pressure
equalization with a second vessel that has completed an exposure
cycle. The second step raises the pressure to 120 psia, using an
ammonia compresser or ammonia stored under regulated pressure.
During exposure, the vessel temperature spontaneously rises to the
range of 120° to 150° F due to the heat of solution as ammonia is
absorbed in the coal moisture.
Following exposure of about 90 minutes, the vessel is de-
pressurized first by equalization with a newly charged vessel and
is then evacuated to 2 psia to minimize the retention of ammonia
in the coal. Finally the vessel returns to ambient pressure, and
the coal drops to a slurry mix tank.
605
-------
A continuous reaction approach to chemical comminution is also
possible and perhaps more economical. In this case the coal could
be continuously charged and discharged through pressure-locking
containers or devices. The use of a liquid chemical agent such as
aqueous ammonia may simplify continuous charging, and provide some
mechanical agitation to minimize exposure time.
The comminuting chemical does not react in any way with the
coal; it's recovered by washing and stripping operations. Following
exposure to the comminuting chemical, the coal would be mixed as a
35 percent slurry with water and fed to the mid-point of one or
more washing towers. The coal drops through the lower tower section
counter-current to hot water that extracts the remaining chemical.
The coal from the bottom of the washing towers would be de-
watered by vibrating screens or centrifuges. Hot water with re-
covered chemical leaves the top of the tower and goes to a steam
stripping operation to separate the chemical from the water.
ECONOMIC CONSIDERATIONS
Two detailed studies of the economics of chemical comminution
versus mechanical fraturing are reported here. One, the more
conservative study, compares mechanical and chemical coal cleaning
plants having equal coal feeds that are designed for the require-
ments of an actual coal-fired utility station. This study leans
heavily on the experience of Roberts & Schaefer Co. for estimates
606
-------
relating to modern mechanical crushing and separation plants. It
assumes identical separation plants for the two plant types, just
different fracture mechanisms. The specific gravities of the heavy
medium solutions have been chosen to maximize premium coal pro-
duced for each. case. Credit is given for excess premium yield
above the amount required for the specific power station.
The second, more representative study, compares chemical and
mechanical cleaning plants having equal coal product yields of
larger than 100 mesh material. In this case the coal feed rate
of the chemical comminution plant is substantially less than that
for the mechanical crushing. As stated earlier, for a given sulfur
level output chemical comminution provides greater clean coal yields
than mechanical fracture. Also in this study chemical comminution
is given a capital credit for producing fewer fines.
Equal Coal Feed. Catalytic Inc. and Roberts & Schaefer Co. have
made a rigorous, conservative economic comparison of two coal
cleaning plants: one using chemical comminution and the other
using mechanical crushing for coal fracture. The study was devised
in cooperation with representatives of both the Electric Power
Research Institute and the Tennessee Valley Authority. Both plants
are rated at 1200 TPH coal feed, and conform to the specific require-
ments of the Homer City Station owned by the GPU and New York State
Electric & Gas.
607
-------
This station uses a 68:32 blend of Helvetia and Helen Upper
and Lower Freeport coals. These coals are less amenable to chemical
comminution for the release of ash and pyrites than a coal like
Illinois No. 6, a factor that contributes to the conservative
nature of this study. Laboratory work for the study was performed
by Syracuse Research Corp.
The Homer City Station consists of three units. Two of these
units must comply with the emissions requirement of 2 Ibs of sulfur
per million BTU's fired. The third unit must meet Pennsylvania New
Source Performance Standards (NSPS) of 0.6 Ibs of sulfur per million
BTU's fired. For comparison purposes, Catalytic and Roberts &
Schaefer prepared designs that would meet or exceed NSPS fuel require-
ments.
A key factor in this study is the credit for yields of NSPS
fuel in excess of the basic quantity required for Homer City's Unit
3. The credit, suggested by EPRI, amounts to $0.65 per million BTU.
This credit is sufficient to tip the economic balance in favor of
chemical comminution despite higher capital investment and operating
costs than mechanical crushing.
The two conceptual plant designs used in this study are identical
except for the method of coal fracturing for liberation of pyrites
and ash forming components. Roberts & Schaefer designed the mechani-
cal crushing system, along with the separation system for both plants.
608
-------
Catalytic designed the chemical comminution system for coal frac-
ture. Both plants employ state-of-the-art technology for coal
preparation.
The mechanical crushing plant has three parallel trains of
600 TPH feed capacity. It operates two shifts a day, resulting in
an average daily output of 1200 TPH. The chemical comminution plant
operates around-the-clock with two of three 600 TPH recovery
trains on stream at all times. The idle train undergoes scheduled
maintenance to increase plant reliability under conditions of con-
tinuous operation.
The washing sections for both plants are identical, consisting
of two stage heavy medium separation including deep cleaning. The
middling circuit gravity in the washing plant for the mechanically
crushed coal is set at 2.03, which is relatively high and favors
the mechanically crushed case.
Table 1 gives the estimated yields for both plants under the
conditions set forth above. Premium yields for the chemical com-
minution plant are significantly higher in terms of both weight
and BTU content. While the chemical comminution plant recovers
14.2 percent more BTU's as premium fuel, it loses 2.9 percent more
BTU's as refuse.
Capital costs developed for the two plants, also given in
Table 1, total $41.5 million for the mechanical crushing plant and
609
-------
$62.6 million for the chemical comminution plant. Catalytic
estimated the cost of the chemical comminution section to be $22
million; Table 1 shows, however, $20 million which accounts for a
$2 million savings when the estimate for the chemical comminution
and washing plants are made as one project rather than separate
projects.
Annual operating costs given in Table 1 total $19.6 million
for the mechanical crushing plant and $26.3 million for the chemical
comminution plant, including the costs for deep cleaning using heavy
medium separation. In terms of costs per million BTU's the figures
are 9.3$ and 12.9$ respectively.
The average revenue required for production was developed for
both mechanical crushing and chemical comminution using the Catalytic
Clean Energy computer program. Plant life was assumed to be 20
years. Net results given in Table 1 indicate that the average re-
venue requirements are 4C/MM BTU higher for the chemical comminution
plant using regulated utility economics, and 4.8$/MM BTU higher
.using a typical industrial discounted cash flow method.
The cost benefits favoring chemical comminution over mechani-
cal crushing show up in the increased yield of premium low-sulfur
product. Recalling the 65
-------
duct. This gives chemical comminution a 9.2C/MM BTU credit over
mechanical crushing. Penalties noted in Table 1 for decreased BTU's
recovered, increased ash disposal, and increased revenue requirements
subtract 7.4C/MM BTU from this figure. So the net economic bene-
fit for chemical comminution is !•. 8C/MM BTU, or about $3.6 million
a year under utility economics.
Equal Product Yield. Another way to look at the economics of
chemical comminution versus mechanical crushing is to compare plants
having equal product yield rather than equal coal feed. Chemical
comminution yields greater recovery values for the same sulfur con-
tent, requiring less coal feed capacity than mechanical crushing.
This leads to decreased capital and operating costs for chemical
comminution compared to the equal coal feed case discussed earlier.
This equal product yield study conducted by Catalytic assumes
that the raw input coal (Illinois No. 6) is sized less than 1h
inches, and contains 2.1 percent sulfur. 12.5 percent ash; and
12,500 BTU/lb. For purposes of calculating the unequal input
coal costs for the two plants, the, coal's value is set at $1/MM BTU.
Product is considered to be only the yield over 100 mesh. The
washing plant receives coal fractured by chemical comminution or
mechanical crushing to 3/8 in. top size. The study assumes that
the washing plant operates 330 days/year.
611
-------
For a cumulative sulfur value of 1.4 percent in the output
of the washing plant, Figure 2 gives recovery yields over 100 mesh
of 73 and 95 percent by weight for mechanical crushing and chemical
comminution respectively. As indicated in Table 2, these recovery
values lead to coal feed rates of 12,000 TPD for mechanical crushing
and 9,000 TPD for chemicailcomminution to produce equal 8,000 TPD
coal product yields.
Using Roberts & Schaefer figures, Catalytic estimates the
total capital investment for a 12,000 TPD mechanical crushing and
washing plant would come to $36.5 million, as shown in Table 2.
These figures include land, interest during construction, startup,
working capital requirements, and contingency capital.
Capital investment figured similarly for the 9,000 TPD chemical
comminution plant would total $42.6 million, but this plant is
given a $5 million credit that takes the total down to $37.6 million.
The credit is for 5.5 percent fewer 28 x 100 mesh fines produced
by chemical comminution, with an estimated worth of $11,000 in
capital cost savings per daily ton of reduced fines production.
Total annual operating costs for the mechanical crushed and
chemical comminuted cases come to $109.5 million and $82.7 million
respectively. The cost of coal dwarfs other operating costs for
chemicals, labor, maintenance and physical separation. Since the
chemical comminution plant uses 25 percent less coal than the
612
-------
mechanical fracturing plant for the same product yield, operating
costs for chemical comminution are substantially lower.
Mechanical crushing generates significantly more refuse coal
than chemical comminution. The resale value of this refuse could
represent a credit to operational costs rather than a debit for
disposal. Calculating operational costs with this credit included
brings the total for mechanical crushing down to $95.8 million,
shown in Table 2 in parentheses. The corresponding total for the
chemical comminution case is $80.8 million, representing only a small
change since it generates relatively little refuse.
Table 2 also gives average revenue requirements for the
chemical comminution and mechanical crushing cases. Using regulated
utility economics, and including the resale value of refuse, chemical
comminution costs 17C/MM BTU's less than the mechanical crushing
plant. This calculation assumes 12 percent return on utility invest-
ment, 10 percent interest with 65 percent debt financing, and income
tax at a 50 percent rate.
Under commercial economics, based upon 15 percent discounted
cash flow and 65 percent debt financing, chemical comminution costs
are estimated to be 16«/MM BTU less than mechanical crushing.
SUMMARY
Chemical comminution appears to cost less than mechanical
fracture in both the equal coal feed and equal coal product yield
613
-------
cases. In Catalytic's view these two comparisons probably repre-
sent the range of conservative and liberal viewpoints. Most likely
the actual economic benefit of coal comminution lies somewhere
between the 1.8C and 17C/MM BTU values presented here.
To sum up, chemical comminution, in combination with con-
ventional coal separation processes, offers a promising alternative
to mechanical crushing for coal cleaning. Its special economic
value lies in its ability to improve the yield of cleaned coal
product without simultaneously producing a greater volume of fines.
This new technology warrants further development as a cost-
effective way to meet increasingly stringent pollution standards
using cheaper, lower grade coals. Coal cleaning alone may not meet
future New Source Performance Standards as it appears that the
new NSPS levels will be based on a percentage sulfur reduction
between mine and stack mouth. In the case of many coals, the com-
bination of coal cleaning and flue gas desulfurization may well
be the most economical approach. In some cases where coal sulfur
contents are high coal cleaning may be mandatory due to scrubber
efficiency limits. The shorter the averaging period for sulfur
measurement, the more difficult the job becomes.
Catalytic is presently seeking a pilot plant host site/
preferably in a utility industry coal cleaning test facility.
It is felt that a batch reaction, gas phase comminution plant in the
614
-------
size range of 2 tons per hour can provide meaningful engineering
and optimization data. Such a plant is currently under considera-
tion.
615
-------
• 5 b 7 I 9 10
ill, .liniliii liiiilmiiiMili
SCREEN OPENING1"
1 10
:
1
1
I
7
"
|
PERCENTAGE BY WEK
« BS a 3
1 "
k" »'
%
MS
3 04 05 06 H7 09 1
i ' •
r ;
— .r1~H~rT
— ' ' i 1 — |— j-
1 : ' i,
; I * : LI
-ULn
j !
-U-
i
- +f - ! ^
1 '
; j
ii
i
1
j "
! i
, i
2 3 « i * .7
I
.
j
1
,.
~¥
• i
i
t
•- 11 r
i
\/
^f
{
y
/T
i ! 1
i ; , i ii • i M
:M !
i ' i
t
/
i 1 IV
1
i ;
|
1
!i
i!
!
I
0 >
i
i
Qr
/
?
I >
^
t
s
/
i^
f
/<
/
T
J
/
j
T
^ i
^
i
,11
!
/
/
i
/
\
/
i
r
\
\ \
'
[
1
i
(•*"
i
y
v'
1 3
/
/
?
t
/
4
r
s~
/
i
^
\ S
\jf j'
J\\
/
x£
/*'
«0 3» 270 MO 140 100 80 MM 40 30 2O 18 16 14 12 10 § ^
US. STANDARD SIEVE DESIGNATION
i ; !
4
|
V
/*
^
^
s
1
i fc 7 i 9
"^
JJ
/fl
1
'
2
f
'
V
1
>
i
*
-
o
4
i.
20 30 40 SO 60 JO
! 1 :
.
s
'
1
/
/
<
1 j
;
_j_ j_
i 1
i
90
200 X
\
,
|
H1
i
i
~
R^
! !! •'! '
11/2 in., R.O.M. Sample
3/8 in.. Mechanically Crushed
14 Mesh, Mechanically Crushed
1 1/2 in.. Chemically Fragmented
Gaseous Ammonia, 120psig,
75° F, Exposure Time: 180 min.
! !i 1 I! i .1
4 >*J •*• -i"
s
i
!
1
-^ J,
|
f
-w. n -*• v i
fm «n>
-
-H
-\
' n
1
i
i , •
' i
' BO
400 325 210 200 190 IOC M to « 35 a
TYLER SIEVE DESIGNATION
20 16 14 1Z 10 » 8 S
SCREEN OPENING. INCHES1"
M«,»>> «._<
Figure 1. Coal particles fragmented by chemical comminution tend to
be larger than those mechanically crushed.
-------
100
11/2%in., R.O.M. Sample,
2.21 % (100 Mesh
3/8in.. Mechanically Crushed,
8.74% < 100 Mesh
14 Mesh, Mechanically Crushed.
21.9%( 100 Mesh
11/2in.. Chemically Fragmented
Gaseous Ammonia, 120psig,
75° F, Exposure Time: 180min.,
4.53% (100 Mesh
1.6
Cumulative % Sulfur
Figure 2. Chemical comminution produces greater coal recovery yields
over 100 mesh for a given sulfur level than mechanical crushing.
-------
CHEMICAL COMMINUTION PROCESS
COAL NH3STORAGE
COMMINUTION
STEP
WASHING
H20
Aq. NH3
H20
DEWATERING
H20
NH*
NH
«"3
RECOVERY
WATER
CLARIFICATION
FRESH H20
CHEMICALLY COMMINUTED COAL
TO SEPARATION
Figure 3. The five basic processing steps for chemical comminution.
618
-------
4000
Low Volatile Bituminous Coal
Comminution Condition :
Gaseous Ammonia
135psi.75° F
400
Exposure Timefmin.)
Figure 4. In chemical comminution processing nearly all coal frag-
menting takes place during the first 90 minutes exposure to gaseous
ammonia.
-------
TABLE 1: ECONOMIC COMPARISON -- EQUAL COAL FEED
Mechanical
Crushing
Chemical
Comminution
Coal Feed, TPH
Yields
Premium Yield
TPH (% of feed)
BTU,(% of feed)
Middlings yield
TPH (% of feed)
BTU (% of feed)
Refuse yield
TPH (% of feed)
BTU (% of feed)
Capital Investment, $ Millions
Fracture
Washing Plant
Other
Total
Annual Operating Cost, $ Millions (/MM
Materials (excluding coal)
Labor
Utilities
Fixed
Total
Average revenue required
Utility economics
$ Millions per year
Cents per MM BTU
Commercial economics
$ Millions per year
Cents per MM BTU
Economic benefit, chemical comminution,
1200
412 (34.3)
45.1
608 (50.7)
51.7
180 (15.0)
3.2
2.
33.
6.5
41.5
BTU)
12.77 ( 6.1)
1.29 (0.6)
2.5 ( 1.2)
3.04 (-1.4)
19.6 ( 9.3)
22.7
11
24.1
11.4
C/MM BTU
Excess premium yield
Decreased BTU's recovered
Increased ash for disposal
Increased production costs (utility economics)
Net economic benefit
1200
552 (46.0)
59.3
403 (33.6)
34.6
245 (20.4)
6.1
20.
33.
9.6
62.6
17.91 ( 8.8)
1.84 ( 0.8)
1.97 ( 1.0)
4.60 ( 2.3)
26.33 (12.9)
30.9
15
33.1
16.2
+ 9.2
- 2.9
- 0.5
- 4.0
+ 1.8
620
-------
TABLE 2: ECONOMIC COMPARISON -- EQUAL PRODUCT YIELD
Mechanical Chemical
Crushing Comminuted
Feed Coal, TPD 12,000 9,000
Yield
Product (+100 Mesh)
TPD 8,000 8,000
Ash, wt.% 5.2 8.8
BTU value, per Ib 13,500 13,000
Fines, TPD 1,000 400
Rejects, TPD 3,000 600
Capital Investment. $ millions
Chemical Treatment11.0
Washing Plant 20.0 16.0
Credit less Fines (5.0)
Contingency 4.0 5.0
Plant Facility Investment 24.0 27.0
Total Capital Investment 36.5 37.6
Annual Operating Cost. $ millions
CHIT * 99.0 72.8
Production Costs 6.9 7.0
Refuse Disposal (Resale) 1.3 (12.4) 0.3 ( 1.6)
Fixed Costs 2.3 2.6
Total Operating Cost 109.5 (95.8) BT77 (80.8)
Average Revenue Required
Utility Economics
$ millions per year 112.8 ( 99.1) 85.9 (84.0)
Cents per MM BTU 1.58( 1.39) 1.25( 1.22)
Commercial Economics
$ millions per year 115.4 (101.7) 88.1 (86.2)
Cents per MM BTU 1.60< 1.41) 1.27( 1.25)
621
-------
CONVERSION TABLE
English to International System of Units (SI)
BTU x 1.0550559
BTU/lb x 2.32600
degree Fahrenheit, °F: {tp + 459.67}/1.8
inch x .0254
pound x .45359237
psi x 6.8947573
ton x 907.18474
= kilojoule
= joule/gram
= degree kelvin,°K
= meter
= kilogram
= kilopascal
= kilogram
622
-------
COAL CLEANING BY THE OTISCA PROCESS
C. D. Smith
Otisca Industries, Ltd.
P. 0. Box 186
LaFayette, New York 13084
ABSTRACT
A brief overview of the Otisca Process will survey bench scale and
continuous pilot plant work to date, current projects, economics, and
environmental impact.
623
-------
Otisca Industries, Ltd. was founded in August, 1972 by
Dr. D. V. Keller, Jr., a chemist and C. D. Smith, a mechanical
engineer. Through a private venture capital stock issue,
approximately $250,000 was raised to support efforts focused
on the commercialization of techniques that had demonstrated
bench scale results that projected potential commercial effi-
ciency and economics in the areas of fossil-fuel extraction
and upgrading.
Six years of development and growth resulted in a
company that, in 1978, organized its approximately 30 employees
into four groups: basic and applied laboratory, engineering,
construction, and operation.
Techniques that have been or are being.studied are chemi-
cal comminution, the Otisca media or heavy liquid coal sepa-
ration technique, tar sand-hydrocarbon solvent extraction, a
coal beneficiation process called the T-Process, and a coal
beneficiation process called the B-Process. This paper will
focus on the development to-date of the Otisca heavy liquid or
media coal beneficiation process.
624
-------
The Otisca media coal beneficiation process uses an
organic heavy liquid - trichlorofluoromethane (CC1~F) rather
than water and magnetite. Gravimetric separation of product
coal from reject material takes place in a bath of media at
ambient conditions. Table 1 compares some of the properties
of Otisca media with those of magnetite and water.
Table 1
Otisca Mag. &
Media HO
1. Boiling Point (°F) 75 212
2. Latent Heat (Vap) (BTU/lb) 80 1000
3. Surface Tension Dynes/Centimeters 20 80
4. Viscosity (20°C - Centipoise) 0.4 14
The toxicity of the Otisca media is very low and is rated
in category 5A, by Underwriters Laboratories. It is non-flam-
mable, non-explosive and virtually odorless. It's chemical
stability is very high and in a coal processing situation, the
fact that it is a non-polar solvent, has the effect of reducing
corrosion potential virtually to zero especially as compared to
water with pH difficulties that develop when it is associated
with the coal and its various sulfur forms.
Another property which supports media's unique ability to
separate ultrafine particles is that with a volume percent
A
solids in a media slurry of 11.4X, yield stress is 15 x 10
dynes per square centimeter as compared to a magnetite water
slurry with a 37 volume percent solids, yield stress would be
625
-------
26 dynes per square centimeter. (Keller, 1978)
Typically, raw coal arrives at a preparation plant with
2 to 5% surface moisture, which is usually the result of dust
control techniques and exposure to the elements. Depending on
the site conditions and raw coal handling techniques, opportu-
nities generally exist for clays to be dispersed in this water
film. Otisca, where appropriate, can add compounds to the
media which enhance a transfer of raw coal surface moisture to
the refuse, in a low shear mixer, which we call a conditioner.
The conditioner's basic duty is to prepare the coal for
separation in the separator by adjusting its temperature, and
by removing slimes from the surface of the product coal and
transferring them to the refuse material,. This phenomenon
apparently is the result of the hydrophobic characteristics
of the media, the hydrophobic surface of the product coal and
by the hydrophilic characteristics of the majority of the
refuse material.
In 1972, Otisca started bench scale and batch pilot plant
activity in a small structure in LaFayette, New York. A 600
pound batch pilot plant was constructed and operation of this
plant commenced in 1973. Table 2 demonstrates the effects of
bench scale activity and batch pilot plant work to a run of
mine analysis.
626
-------
Table 2
UPPER FREEPORT
Run of
Mine
26.09
37.34
36.57
-
1.55
—
1.22
-
9,128
-
TT
-
6.5
(PENNA.)
Bench
Scale
36.75
55.36
7.89
85.9
0.98
58.6
0.53
71.4
13,911
52.8
80.5
1.50
-
1 Ib.
Pilot
Plant
36.62
56.08
7.30
87.0
0.88
63
0.56
70
14,009
54.8
84.1
1.50
2.07
600 Ib
Batch
Volatile Matter
Fixed Carbon
Ash
% Reduction
Total Sulfur
% Reduction
Pyritic Sulfur
% Reduction
BTU/lb
Weight Yield
BTU Yield
Specific Gravity
Moisture
MASS
It is important to note that in scale up from bench scale
to batch pilot plant, ash and sulfur reduction values were the
same or better due to the reduction of edge effects. Through
the operation of this batch plant, it was also observed that
the Otisca technique was capable of reducing surface moisture,
as evidenced by total moisture in the run of mine of 6.5X being
reduced to pilot plant product coal of 2.07X.
627
-------
Studies at this facility in LaFayette, New York from 1972
until 1975 focused on approximately 30 United States' bituminous
coals. The batch plant was decommissioned in 1976.
As a result of these laboratory and batch pilot plant
studies. Island Creek Coal Company supported the construction
and operation of a continuous pilot plant sited at their North
Branch Mine near Bayard, West Virginia. The continuous plant
constructed under a very tight budget was designed to treat a
1/4 x 0 slack coal. It operated at the mine site from approxi-
mately June, 1976 until January, 1977. Representative raw
coal and product coal analyses of the plant operation are
shown in Table 3. During the operation of the plant, sub-
stantial physical changes were made to enhance the ability of
the installed hardware to separate coal, recover the media
and stably and continuously convey the solids. (Keller, Smith
& Burch, 1977)
Figure 1 - Otisca Coal Beneficiation Plant - provides a
block diagram which provides an overview of the major unit
operations involved in the process. Raw coal, comminuted as
required, to-.meet a balance between maximum top size and product
coal specifications, is introduced to a conditioner where its
temperature is neutralized relative to the bath operating
temperature, and, where appropriate, surface moisture is
transferred to refuse. The conditioned raw coal is transferred
to a separating bath where conveyors skim the surface and
628
-------
Table 3
Representative Analyses Of The North
Branch Upper Freeport Raw Slack Coal
and
Product Coal Separated From This
Volatile Matter
Fixed Carbon
Ash
Ibs/MM BTU
Total Sulfur
Ibs/MM BTU
Pyritic Sulfur
Ibs/MM BTU
Organic Sulfur
Ibs/MM BTU
BTU/lb
BTU/lb (MAF)
Weight Yield
BTU Yield
Specific Gravity
Raw Coal
17.70
60.65
21.65
18.33
2.63
2.23
1.54
1.30
1.05
0.89
11,812
15,076
Product Coal
19.12
71.22
9.66
6.93
1.53
1.10
0.62
0.44
0.87
0.62
13,933
15,423
74.9
88.3
1.50
629
-------
RAW COAL
MEEHAUIC&L
COHMIUUT«€XJ
AS REQJ)
FIGURE 1
TOP SIZE. HAY
RAKIQC. rnori e*«
10 325 MESH.
1
SEFARATIOU BY
GRAVITY IU STATIC
I1EU1IA.
SIK1KS
MEDIA
MEDIA
-uou -
pcpusc
FIGURE*! OTISCA COAL BENEFICIATION PLANT
PATENT PENDING
JULY 1975
C.CLS.
P.O. BOX 211. LAmVCTTC. NCW VOMK 1»O«4
•ff
-------
remove float for drip drying and convey them to a product coal
evaporator. Sinks or rejects are conveyed from the bottom of
the bath to the rejects evaporators. The evaporators are
indirect fired, conductive evaporators, wherein the product
coal or rejects are heated to approximately 100°F, which results
in complete evaporation of the separating liquid.
Vapors are collected from the evaporators, carried in
ducts to recovery equipment, which generally consists of
compressor-condensing stages followed by polishing of non-
condensibles by either carbon adsorption or liquid absorption.
Product coal from the plant is discharged from a rotary
valve, virtually free of surface moisture and it can be con-
veyed and stored in conventional equipment. Rejects are
discharged through a rotary valve to conventional material
handling equipment, in a physical condition that is not
difficult to handle. Typically, the ease in handling the
refuse stems primarily from the fact that it contains -enough
moisture not to be dusty, but not so much moisture that it is
runny or extremely adhesive.
In 1977, the continuous pilot plant equipment at Bayard
was relocated to Florence, Pa. with the objective of continued
operation to demonstrate at relatively large scale, the ability
of the process to separate raw coal, and to provide bulk samples
for testing relative to the commercial use of the product coal.
An additional objective was to continue process hardware
631
-------
development through the installation and operation of modified
or completely different unit operations. Demonstration and
unit test work continue at Florence.
In 1977, demonstration work at the Florence Development
Center led to the signing of a contract with American Electric
Power for Otisca to construct and operate a 125 TPH Otisca
Demonstration Plant at American Electric Power's Muskingum
Mine near Beverly, Ohio. The facility, which is presently under
construction, is essentially a complete coal preparation faci-
lity in that it has its own site services, raw coal storage,
raw coal recovery, crushing station, crushed coal storage, the
Otisca separation facility, refuse storage and handling bin
and a product coal stacking conveyor. The total budget for
the construction, start-up and operation of the plant is 6.7
million dollars.
Plant start-up is projected for the third quarter of
1979, with Otisca responsible for the design, construction and
operation of the plant. It is important to note that the 6.7
'' s
million:dollar budget covers the installation and operating
cost of a significant amount of instrumentation and/or equipment
that would not be found in a commercial cost beneficiation plant,
whose objectives did not include the monitoring and generating
of a significant amount of detailed operating data.
Table 4 provides :an overview of separating characteristics
and by-product generation of untreated coal. Conventional
hydroprocess (water-magnetite) and coal cleaned by the Otisca
media process. The comparison represented in the table relates
632
-------
to the electricity consumed annually by an all electric house
in Ohio (25,000 kilowatt hours per year). Boiler fuel consump-
tion in the comparison is based on an overall combustion
generation and distribution efficiency of 35%. Values for
sludge from scrubbers assume the generation of 10 pounds of
sludge per 1 pound of sulfur in the input coal.
Table 4
% Ash
% Sulfur
BTU/lb
Ibs Coal/Year to
Produce 25,000
KWH
Ibs Ash/Year
Ibs Sulfur/Year
Ibs Sludge/Year
During Clean-Up
of Sulfur
Emissions of CFM
Ibs/Year
Untreated
Coal
24.8
6.2
10,700
2.3 x 104
5,704
1,426
14,260
Existing
Hydro-Process
18.0
5.8
11,500
2.12 x 104
3,819
2,131
12,310
Coal Cleaned By
Otisca Process
11.5
4.2
12,700
1.92 x 104
2,208
806
8,060
0.1 - 0.5
There are 14,260 pounds of sludge generated per year from
stack gas scrubbing of untreated coal; 8,060 pounds of sludge
are generated per year for coal cleaned by the Otisca Process,
giving a net reduction of 6,200 pounds of sludge per year,
which trade off against an emission of trichlorofluoromethane
of 0.1 to 0.5 pounds per year. (Otisca Industries, Ltd., 1978)
Comparative economics for a 400 TPH with a 1/2" x 0 feed
are listed below. The comparison intent is to fairly relate
the relative capital and operating cost of a conventional
633
-------
preparation plant assuming that that plant is equipped with
separating techniques capable of cleaning to ultrafine sizes,,
and that the product coal will be mechanically or thermally
dewatered to an equivalent level with the Otisca Process. The
comparison assumes a relatively typical case study for a coal
preparation plant involving annual operation of approximately
3,200 hours per year.
Table 5
Comparative Economics
400 Tons/Hour
1/2" x 0 Feed
Conventional Otisca
Preparation (Media)
First Cost:
Unit Cost
Total Cost
$/Ton-Hr.
Direct Operating Cost:
(Labor, Power, Fuel, Mat'l,
Maint.) $/Raw Ton
Total Cost:
(At 15% Return on Investment
$/Raw Ton)
25,000 15,000
10,000,000 6,000,000
2.00
1.45
5.47
3.65
634
-------
The word Otisca is an Onondaga Indian word meaning waters
much gone away or water much dried away. It is projected that
the demonstrated economics and efficiencies of the Otisca
Process which result in higher BTU yield per ton of raw coal
at projected lower operating cost will continue to motivate
additional development work and the near-term commercial
acceptance of the process as a practical and useful technique
to upgrade raw coal for downstream use for power generation,
coke manufacture, solvent extraction, gasification and
liquefaction processes.
635
-------
BIBLIOGRAPHY
1. Otisca Industries, Ltd., "Information on the Use of Chloro-
fluorocarbons for the Beneficiation of Carbonaceous Minerals
Such as Coal", presented to the Environmental Protection
Agency, February 22, 1978.
2. D. V. Keller, "The Otisca Process - The Physical Separation
of Coal Using a Dense Liquid", Presentation to Coal
Preparation Workshop, Ohio University (1978).
3. D. V. Keller, Jr., C. D. Smith & E. F. Burch, "Demonstration
Plant Test Results of the Otisca Process Heavy Liquid Bene-
ficiation of Coal", presented to Annual SME-AIME Conference,
Atlanta, Georgia (March 1977).
636
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing}
1. REPORT NO.
EPA-600/7-79-098a
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Proceedings: Symposium on Coal Cleaning to Achieve
Energy and Environmental Goals (September 1978,
Hollywood. FL)--Volume I
6. REPORT DATE
April 1979
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
S.E.Rogers and A.W.Lemmon, Jr. (Editors)
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Battelle Columbus Laboratories
505 King Avenue
Columbus, Ohio 43201
10. PROGRAM ELEMENT NO.
EHE624A
11. CONTRACT/GRANT NO.
8-02-2163, Task 861
12, SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
TasT
JL OB It
P
pVERED
14. SPONSORING AGENCY CODE
EPA/600/13
IB. SUPPLEMENTARY NOTES ffiRL-RTP project officer Is James D. Kilgroe, MD-61, 919/541-
2851.
16. ABSTRACT
The proceedings document presentations made at the Symposium on Coal Cleaning to
Achieve Energy and Environmental Goals, September 11-15, 1978, in Hollywood,
Florida. The symposium provided an opportunity for mutual review and discussion
of: the physical and chemical coal cleaning programs of EPA, DoE, the Electric
Power Research Institute, and numerous industrial organizations; European and
Soviet plans for the future; and problems of ongoing operations. The proceedings
include the following topics: coal characteristics, coal cleaning overview, physical
coal cleaning technology, environmental assessment and pollution control technology,
and chemical coal cleaning technology. The first three topics are covered in Volume
I; the last two, in Volume II.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS C. COS AT I Field/Group
Pollution
Coal
Physical Properties
Chemical Properties
Assessments
Pollution Control
Stationary Sources
Coal Cleaning
Environmental Assess-
ment
13B
21D,08G
14B
07D
IB. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (This Rtportj
Unclassified
21. NO. OF PAGES
6A7
20. SECURITY CLASS (TMtpagt)
Unclassified
29. PRICE
•PA Form 2220-1 (••73)
637
fcU.S. GOVERNMENT PRINTING OFFICE: 1979 -640 • 0131 3903 REGION NO. 4
------- |