EPA-600/2-76-149
June 1976
Environmental Protection Technology Series
SYMPOSIUM PROCEEDINGS:
ENVIRONMENTAL ASPECTS OF
FUEL CONVERSION TECHNOLOGY, I
(December 1975, Hollywood, Florida]
Industrial Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, US Environmental
Protection Agency, have been grouped into five series. These five broad
categories were established to facilitate further development and application of
environmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4 Environmental Monitoring
5. Socioeccnomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL PROTECTION
TECHNOLOGY series This series describes research performed to develop and
demonstrate instrumentation, equipment, and methodology to repair or prevent
environmental degradation from point and non-point sources of pollution. This
work provides the new or improved technology required for the control and
treatment of pollution sources to meet environmental quality standards.
EPA REVIEW NOTICE
This report has been reviewed by the U.S. Environmental
Protection Agency, and approved for publication. Approval
does not signify that the contents necessarily reflect the
views and policy of the Agency , nor does mention of trade
names or commercial products constitute endorsement or
recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600j2-76-149
June 1976
SYMPOSIUM PROCEEDINGS:
ENVIRONMENTAL ASPECTS
OF FUEL CONVE RSION TE CHNOLOGY, II
(DECEMBER 1975 , HOLLYWOOD, FLORIDA)
Franklin A. Ayer (Compiler)
Research Triangle Institute
P. O. Box 12194
Research Triangle Park, NC 27709
Contract No. 68-02-1325, Task 57
Program Element No. EHB529
EPA Task Officer: William J. Rhodes
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
-------
FOREWORD
The proceedings for the symposium on "Environmental Aspects of Fuel Con-
version Technology, II," is the final report submitted to the Industrial Envi-
ronmental Research Laboratory for the Environmental Protection Agency
Contract No. 68-02-1325. The symposium was held at the Diplomat Hotel,
Hollywood, Florida, December 15-18, 1975.
The main objective of the symposium was to review and discuss environ-
mentally related information on coal conversion technology. Papers were
presented that covered a summarization of major environmental programs
and contaminants in coal, process technology, control technology, process
measurements, sampling and analytical information pertinent to coal gasifica-
tion and liquefaction, and product usage.
Mr. William J. Rhodes, Chemical Engineer, Industrial Environmental Research
Laboratory, Environmental Research Center, Environmental Protection
Agency, Research Triangle Park, North Carolina, was the Project Officer and
General Chairman of the Symposium.
Mr. Franklin A. Ayer, Manager, Technology and Resource Management
Department, Center for Technology Applications, Research Triangle
Institute, Research Triangle Park, North Carolina, was the Symposium Co-
ordinator and Compiler of the proceedings.
ii
-------
Table of Contents
(* indicates speaker)
15 December 1975
Opening Comments. . . . .
John K. Burchard, Ph.D.
. . . . .
. . . . . . . . .
Session I: ENVIRONMENTAL PROBLEM DEFINITION, PART A
Gary J. Foley, Ph.D., Session Chairman
EPA's Energy/Environmental Program
Stephen J. Gage, Ph.D.
. . . . . .
. . . . . . . .
The President's Synthetic Fuels Commercialization Program
Matthew J. Reilly, Ph.D.
. . . . . . . . . . . .
Environmental Controls and Synthetic Fuels
S. B. Alpert
. . . . . . . . . . . . . .
Federal Regulation of Atmospheric Emissions From Coal Gasification Plants. .
James F. Durham
Summary of Multimedia Standards
John G. Cleland
. . . . .
An Investigation of Trace Elements in Coal
H. J. Gluskoter, Ph.D.,*
R. A. Cahil,
W. G. Miller,
R. R. Ruch, and
N. F. Shimp
Geochemistry of Trace Elements in Coal
Peter Zubovic
. . . . . . . . . . . . . . . .
. . . . .
Session II: PROCESS TECHNOLOGY
E. C. Cavanaugh, Session Chairman
. . . . . . . . . . . . . . . . . . . . . . .
Trace Elements and Major Component Balances Around the Synthane PDU Gasifier
A. J. Forney,*
W. P Haynes,
S. J. Gasior,
R. M. Kornosky,
C. E. Schmidt, and
A. G. Sharkey
iii
-------
Table of Contents (con.)
Page
CO2 Acceptor Process.
Carl Fink,
George Curran,* and
John Sudbury
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
Effluent Considerations in Coal Gasification. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
Frank C. Schora and
Donald K. Fleming*
Bigas Process. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .105
V. D. Kliewer* and
V. L. Brant
16 December 1975
Session II: PROCESS TECHNOLOGY (con.)
E. C. Cavanaugh, Session Chairman
The Winkler Process, A Route to Clean Fuel From Coal. . . . . . . . . . . . . . . . . . . . . . . . .117
I. N. Banchik
Pressurized, Stirred, Fixed.Bed Gasification
D. W. Gillmore, Ph.D.,* and
A. J. Liberatore
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .125
The Westinghouse Fluidized Bed Combined Cycle Process:
Status of Technology and Environmental Considerations
L. A. Salvador,*
E. J. Vidt, and
J. D. Holmgren
. . . . . . . . . . . . . . . . . . . . . . . .133
Role of Gasifier Process Variables in Effluent and Product Gas Production
in the Synthane Process. . . . . . . . . . . . . . . . . . . . . . . . . .
M. J. Massey, PhD.,*
D. V. Nakles,
A.J. Forney, and
W. P. Haynes
. . . . . . . . . . . . . . .153
Environmental Aspects of Synthoil Process for Converting Coal to Liquid Fuels
Sayeed Akhtar, Ph.D.,
Sam Friedman, and
Paul M. Yavorsky
(presented by James W. Mulvihill)
. . . . . . . . . . . .179
Environmental Aspects of the SRC Process
Russell E. Perrussel, *
Walter Hubis, and
J. L. Reavis
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .183
iv
-------
Table of Contents (con.)
Page
Coalcon Liquefaction
Edward T. Coles
(Paper presented by John Hall, but not submitted for publication)
Session III: CONTROL TECHNOLOGY
Rene R. Bertrand, Ph.D., Session Chairman
. . . . . . . . . . . . . . . . . . . . . . . . . . . .19.1
Low and Intermediate Btu Fuel Gas Cleanup
C. B. Colton,*
M. S. Dandavati, and
V. B. May
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .193
The Benfield Activated Hot Potassium Carbonate Process:
Commercial Experience Applicable to Fuel Conversion Technology. . . . . . . . . . . . . . . . . . .217
D. H. McCrea
Coal Conversion Process Wastewater Control
W. A. Parsons, Ph.D., * and
R. A. Ashworth
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .225
17 December 1975
Session III: CONTROL TECHNOLOGY (con.)
Rene R. Bertrand, Ph.D., Session Chairman
A Tapered Fluidized-Bed Bioreactor for Treatment of Aqueous Effluents
From Coal Conversion Processes. .. """""""'"
Charles D. Scott, Ph.D.,*
Charles W. Hancher,
David W. Holladay, and
George B. Dinsmore
. . . . . . . . . . . . . . . .233
Climatic Effects on Wastewater Treatment
Stanley L. Klemetson, Ph.D.
. . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . .241
Waste Management of Fuels Processing Effluents
G. W. Grove
(Presented by Ronald V. Trense)
. . . . . . . . . . . . .
. . . . . . . .253
Control Technology R&D Needs
C. E. Jahnig,*
R. R. Bertrand, Ph.D., and
E. M. Magee, Ph.D.
. . . . . . . . . . . . . . . .
. . . . . . .
. . . . . . .259
v
-------
Table of Contents (con.)
Page
Session IV: PROCESS MEASUREMENTS
James A. Dorsey, Session Chairman
. . . . . .
. . . . . . . . . . . . . . . . . . . . .
. . .267
Measurement Programs for Environmental Assessment
R. M. Statnick, Ph.D.,* and
L. D. Johnson, Ph.D.
. . . . . . . . . . . . . . . . . . .
. . . . . .269
Sampling Procedures for Process Streams
C. A. Flegal, Ph.D.,* and
J. W. Hamersma
. . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . .275
Analytical Techniques for Sample Characterization in Environmental Assessment Programs
C. H. Lochmuller, Ph.D.
. . . . . .285
Evaluation of Particulate Characterization Techniques. . . . . . . . . . . . . . . . . . . . . . . . . .293
H. Mahar, Ph.D.,* and
N. Zimmerman, Ph.D.
18 December 1975
Session V ENVIRONMENTAL PROBLEM DEFINITION, PART B
Robert P. Hangebrauck, Session Chairman
. . . . . . . . . . . . . . . . . .305
Water Requirements for an Integrated SNG Plant and Mine Operation
D. J. Goldstein, Sc.D.,* and
R. F. Probstein
. . . . . . . . . . . . . . . . .307
Sulfur Emission Controls for a Coal Gasification Plant. . . . . . . . . . . . . . . . . . . . . . . . . .333
Milton R. Beychok
Low Energy Gas Retrofit to Industry
D.A. Ball
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .341
Combined-Cycle Power Systems
F. L. Robson, Ph.D.,
W. A. Blecher, and
A. J. Giramonti*
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . .359
NOx Considerations in Alternate Fuel Combustion
G. Blair Martin
. . . . . . . . . . . . . . . . . . . . . . . . . . .373
Environmental Impact and R&D Needs in Coal Conversion. . .
E. M. Magee, Ph.D.,*
R. R. Bertrand, Ph.D., and
C. E. Jahnig
. . . . . . . . . . . . . .395
vi
-------
15 December 1975
Opening Session
William J. Rhodes
General Chairman
-------
OPENING COMMENTS
John K. Burchard, Ph.D. *
It is a real and sincere pleasure to welcome you to
our second symposium on the Environmental Aspects of
Fuel Conversion Technology. As I indicated at our first
symposium in St. Louis in May of 1974, we really don't
have any need to justify the importance of establishing a
viable industry for the conversion of coal, liquid and gas,
as fuels. With the cooperation of American industry and
universities, the Federal Government has already initia-
ted many programs with large amounts of funds to ac-
celerate development of this technology. However, the
establishment of this technology without the protection
of the public's health and welfare would be a real mis-
take. It would brand our country as one which neither
cares for all its citizens nor possesses the technical
know-how all this would create.
Within the Federal Government, assessment of envi-
ronmental problems relating to this technology is one of
the primary responsibilities of the Environmental Protec-
tion Agency, and in particular our I ndustrial Environ-
mental Research Laboratory. We are currently negotia-
ting various contracts for the performance of studies
aimed at finding solutions for these problems. Up to this
time, our evaluations have necessarily been limited
mostly to existing data and engineering estimates. How-
ever, our new studies will lead to the acquisition of new
data that will be used to expand and update our environ-
mental assessments. These data have to be acquired on a
continu ing basis in order for us to keep abreast of the
technological changes that will emerge while progressing
from bench scale through demonstrations in commer-
cial status.
Why do we need this information? Primarily for
assurance that a given potential problem is either insig-
nificant or can be handled by existing technologies.
Where there are problems, controls will have to be devel-
oped and applied. This can be done most efficiently by
full cooperation on the part of all Federal agencies and
their contractors as well as by other private and public
organizations. Meetings such as this, we hope, can do
much to stimulate the necessary cooperation and pro-
vide the opportunity for those involved to assure for
themselves, as well as to assure each other, that their
efforts are complementary rather than dupl icative or
* 0 irector, Industrial Environmental Research Laboratory,
Environmental Protection Agency, Research Triangle Park,
North Carolina.
opposing in nature. It should be obvious that the devel-
opment of either conversion technology or environ-
mental control technology, without proper considera-
tion of the other, is both unacceptable and unworkable.
Fuel Conversion Technology in and of itself is suffi-
ciently complex. The additional consideration of the
environmental effects and control compounds that com-
plexity, particularly as the effective control has a decid-
edly significant economic impact. EPA is concerned
about both factors, control and cost. Our goal is to aid
and achieve the most environmentally sound fuel conver-
sion systems at the lowest possible overall cost. This
concern not only means developing efficient add-on con-
trols, but also investigating, developing and promoting
the use of more efficient operating procedures and opti-
mum processes. These efficiency improvements would
include, for example, process modifications and the
recovery and reuse of environmentally damaging constit-
utents that have economic value.
Because of a certain increase in coal utilization dur-
ing the next few years, we just don't know when all
environmental pollutants will be adequately identified
and modified, or when the effect of all pollutants will be
adequately measured. Obviously then, all control systems
can not be defined precisely at this time or in the fore-
seeable future. However, in order to permit commercial-
ization with minimum delay, existing controls must be
applied to known problems; potential problems must be
assessed; and controls for these potential problems must
be evaluated, utilized, modified or developed as neces-
sary and as soon as possible.
This symposium is expected to provide an exchange
of meaningfu I information between process developers,
process users and environmental groups. The first session
prov ides information on various overall coal programs,
potential pollutants in coal and possible environmental
standards. Subsequent sessions present some of the proc-
esses and control technologies that are being considered.
These are followed by a session dealing with the sam-
pling analysis of process and effluent streams. The final
session defines environmental problems in terms of
today's information.
I think that the value of free and open meetings
such as this is evident-not only because of the informa-
tion to be exchanged during the meeting, but because of
the stimulus it provides for continuing communications
between participants and attendees long after the
symposium adjourns. We in EPA, as well as in other
Federal, local and State government organizations, are in
3
-------
busi ness to serve you and the rest of our country. In
order to do the most effective job, we all need the most
valid information available when making decisions. Much
of this information can come from this group. The
better the information, the better will be the decision
for all concerned.
In conclusion, I hope you will enjoy your stay here
and get something meaningful and productive from this
meeting. I hope further that you will become involved
with each other's interests and activities both during and
after the formal sessions. Again, I'd like to extend my
most sincere welcome to all.
4
-------
15 December 1975
Session I:
ENVIRONMENTAL PROBLEM DEFINITION, PART A
Gary J. Foley, Ph.D.
Session Chairman
5
-------
EPA's ENERGY/ENVIRONMENTAL PROGRAM
Stephen J. Gage, Ph.D.*
It is a pleasure for me to be here this morning, Gary,
and thank you for the very kind introduction. I am
surprised that the three people who preceded me up here
to the platform have not acknowledged the physical
surroundings. I think this is a marvelous place for a
symposium. I do not know how you manage to keep
everyone sitting out here in the audience when there is
all that beautiful weather outside and such a lovely
location. I just wish I were going to be able to be here
longer and we could run the meeting for about 2 or 3
weeks so we could really enjoy the surroundings. I also
add my congratu lations to the Synthetic Fuel Group at
the I ndustrial Environmental Research Laboratory. It
looks like a very excellent session they have arranged
here and I am glad to be able to participate.
The 1973 Arab oil embargo brought home to
Americans the realization that we would have to depend
increasingly on our domestic energy resources. If fuels
were to be extracted, transported, processed, converted,
etc., as well as utilized, on American shores, then we
reasoned that the American environment would receive
all of the impact of increasing energy use.
Although the measures required during the Arab oil
embargo to mitigate the impact of the oil cutoff were
almost exclusively conservative-belt tightening is
probably the most appropriate term-we recognized at
that time that the problems of inadequate domestic
energy supplies, inefficient use of those supplies, and the
necessary environmental damage resulting from their
development and use, would probably be with us for a
good many years. If nothing else, the oil embargo served
as that very useful "slap-in-the-face" that you see in the
television commercial-"I think we needed that!" The
only thing that disappoints me is that we have not
awakened nearly as much as the television performers
seem to awaken with their slap-in-the-face. At least we
are not doing a great deal about it even if we have
awakened.
At the Federal level, there has been a considerable
expansion of research on both the energy technologies
and the associated environmental research. The program
and the capabilities within the environmental research
are scattered among a variety of departments and
agencies. This calls for and requires centralized
coordination of these programs so that our national
* Deputy Assistant Administrator, Office of Energy,
Minerals, and Industry, Environmental Protection Agency,
Washington, D.C.
energy goals can be matched with an effective R&D pro-
gram in that critical area where energy needs and envi-
ronmental protection goals overlap.
Major Federal Government involvement in energy
R&D goes back to the early 1950's with the major
historical emphasis on nuclear power. The Federal R&D
concentrated first on the lightwater nuclear reactor and,
later, on the liquid metal fast breeder reactor. The first
major comprehensive study of national energy R&D
needs and the role of the Federal Government in meet-
ing those needs was called for by President Kennedy in
the early 1960's. That study concluded in the so-called
"Cambel Report" that concluded we did not face insur-
mountable problems in energy or energy R&D. The
Federal Government's role at that time was seen as a
limited one. Of course the underlying assumptions in
that study were based on the then widely prevalent view
that we had nearly inexhaustible domestic oil and gas
resources.
In 1967, President Johnson established an Energy
Policy Office within the Office of Science and Tech-
nology to sponsor a thorough study of energy resources.
The first general results of those studies were incorpo-
rated in President Nixon's first energy message to
Congress in 1971. Energy/environmental R&D, which
had long been languished as a "poor stepchild," was
given its first major public visibility in that address.
As the danger signals became apparent in the Mid-
east during early 1973, Administration and Congression-
al consensus for increased Federal spending on energy
R&D developed: In the summer of 1973, the President
directed the Chairman of the Atomic Energy Com-
mission to prepare a comprehensive and integrated
national energy R&D plan. The result, entitled The
Nation's Energy Future and commonly referred to as the
Dixie Lee Ray Report, was released in December 1973.
Drawing upon the efforts of 36 Federal departments and
agencies, as well as the private sector, it recommended a
5-year, 10-billion-dollar energy R&D program to achieve
national energy self-sufficiency through the development
of five energy resource areas-energy conservation, oil
and gas, coal, nuclear, and advanced energy systems.
The Ray Report helped shape the Federal budget
requests for the fiscal year 1975. These requests con-
tained a substantial increase in energy R&D, including an
increase for the environmental aspects of the major
thrusts in energy resource development. As you can see
in table 1, the Administration, through EPA's budget,
requested 191 million dollars in FY 1975 for the
7
-------
Table 1. Background
Early 1960's
(Kennedy)
1967
(Johnson)
1971
(Nixon)
1973
(Nixon)
1974
1974
Late
1974
- First major comprehensive study of
energy R &0 needs. The "Cambel
Report" concluded that the prob-
lems were not insurmountable since
we had nearly inexhaustible domes-
tic oil and gas resources!
- Energy policy office established to
"sponsor a thorough study of ener-
gy resources"
- First general results of study, re-
ferred to above, incorporated into
the President's first energy message
to Congress
- "The Nation's Energy Future" (The
Dixie Lee Ray Report) prepared
and released. Recommendation:
A 5-year $10 billion energy re-
search and development program
- FY 75 Federal budget request of
$191 million. $134 million auth-
orized by Congress of which about
$50 million was for the interagency
program
- Two interagency task forces estab-
lished by OMB:
. Health and environmental
effects of energy use
. Environmental control tech-
nology for energy systems
- Task force reports issued - identi-
fied program gaps and located tar-
get areas for "pass through" funds
e ne rgy/environmental R&D program. The Congress
authorized 134 million dollars late in 1974. The inter-
agency portion of that program was about 40 percent of
the total, approximately 50 million dollars.
Following the release of the Ray Report and initial
formulation of the 1975 budget by OMB, two inter-
agency task forces were established by OMB and the
Council on Environmental Quality to examine ongoing
Federal research programs and to recommend allocations
of funds which would provide the most effective inte-
grated environmental/energy R&D program. Their
reports covered two areas-health and environmental
effects of energy use and environmental control tech-
nology for energy systems. Issued in November 1974,
the task force reports were developed by key representa-
tives from more than a dozen Federal agencies, depart-
ments, and laboratories, all involved in energy-related
environmental research. One of the major purposes of
the reports was to determine whether serious gaps
existed in the overall Federal energy/environmental
R&D program. By performing a crosscut review of the
entire program, it was possible to identify such gaps, to
determine areas where adequate support was available
for national energy goals, and to locate target areas for
supplemental EPA "pass through" funds. Thus, EPA, in
implementing the interagency program based on the two
task force reports, was in an excellent position to help
maintain a balanced and coordinated Federal energy/
environmental R&D program.
Also late in 1974 came the establishment of the
Ene rgy Research and Development Administration
(ERDA) through enabling legislation. The establishment
of ERDA created the centralized location within the
Federal government for the development of energy tech-
nology. It also supported a major program of environ-
mental research which had been taken over from the
Atomic Energy Commission. As we move ahead toward
the development of the new energy technologies, the
role of E RDA will undoubtedly increase. It will, I
believe, probably be concentrated on the environmental
research that is required to make sure that those energy
technologies do not degrade the environment.
Overall planning and coordination of the Federal
interagency energy/environmental R&D effort, con-
ducted under EPA auspices, is the responsibility of the
newly created Office of Energy, Minerals, and Industries
within EPA's Office of Research and Development
(ORO). The Office of Energy, Minerals, and Industry
(figure 1) is one of four major headquarters offices with-
in ORO. Ea~h of those headquarters offices has asso-
ciated with it two or more of the 15 EPA field labora-
tories specializing in various areas of environmental
research and development.
Within the Office of Energy, Minerals, and Industry,
there are two major divisions: one is the Industrial and
Extractive Processes Division which concentrates on
multimedia, long-range planning, and coordination in the
area of industrial and mineral processes; and the other,
the Energy Processes Division, has responsibility along
the same lines as they relate to energy activities.
8
-------
t-J~
Ztz1
-3
>t""
t-'H
tI1
'l;j~
:;Q>-
0:;1::1
>-'1>-<:
tzJ
Q
~
H
o
Z
>-
~
21
o
....
ASSISTANT ADMINISTRATOR
FOR RESEARCH AND DEVELOPMENT
OFFICE OF
MONITORING AND
TECHNICAL SUPPORT
OFFICE OF ENERGY
MINERALS AND
INDUSTRY
OFFICE OF AIR, LAND
AND WATER USE
-------
In addition,there is an Energy Coordination Staff.
This staff has the responsibility of coordinating the
health and environmental effects part of the interagency
program both within the Environmental Protection
Agency as well as with other Federal agencies.
For planning purposes, the control technology
development program has been divided into five program
areas as shown in figu re 2. The I ntegrated Technology
Assessment Program area is implemented and coordi-
nated by the headquarters staff because of its key role in
assuring the utility of research results from all of the
program areas. EPA's two Industrial Environmental
Research Laboratories, which report to the Office of
Energy, Minerals, and Industry, and six other Federal
agencies, are responsible for carrying out the other pro-
gram areas. Since the Fuel Processing area is of most
interest to this audience, I will restrict myself to that
area. The fuel processing area is promoting and partici-
pating in the development of advanced technologies for
fuel processing by providing environmental control tech-
nology development and characterizing environmental
emissions and effluents. It is especially important that
these efforts support the needed information flow
between the energy technology developers and the regu-
latory offices to insure that standards do not delay or
prevent energy development, yet establish adequate
regulations to protect the nation's environmental quali-
ty.
In fiscal year 1975, as I mentioned, the energy/
environmental R&D program was appropriated 134
million dollars of which approximately 21 million was
allocated to the Fuel Processing area. In fiscal year 1976,
the funding appropriated for the entire program is 100
million and, assuming that the overall budget will remain
at approximately 100 million dollars per year through
fiscal year 1980, funding for the Fuel Processing pro-
gram is projected to increase as shown in figure 3. Here
you see the allocation between the subprograms within
Fuel Processing and the projected growth in the syn-
thetic fuel area, which we feel is necessary in order to
complement the efforts in the Energy Research and
Development Administration and in the private sector.
I should interject a note of uncertainty at this
particular time with which you are probably quite
fam il iar; that is, the Conference report of the E R DA
authorization bill was defeated in the House last
Thursday. This raised some severe questions about how
fast synthetic fuels will be developed in this country. I
feel quite confident that some incentive program will
probably emerge from this session of Congress; but it is
almost anyone's guess, at this point, what the size of
that incentive program will be and what will be the
ultimate thrust.
Depending on the final disposition of this incentive
program, how many dollars, how fast, and whatever
stipulations are written in the appropriation, we may
have to reprogram dollars within our budget or seek
supplemental funding to provide EPA support.
I have tried to put the EPA synthetic fuel program
into perspective with respect to the Federal Interagency
R&D Program, and I would like to summarize in the
final minutes I have here some of the resu Its of the
synthetic fuel program that EPA has been conducting
over the past few years.
I am sure you will be hearing more about these
results during the rest of the symposium. As you can see,
much effort has gone into analyzing coals which are
likely to be converted into synthetic fuel. This is prob-
ably providing some of the best information on the
geochemistry of the coal and should be of considerable
benefit to energy technology development as well as to
environmental protection. Engineering studies to deter-
mine the potential pollutant releases have been made on
a number of processes, as can be seen in table 2:
Synthoil, Synthane, Lurgi, etc. This table also shows
some of the results of the control technology studies,
particularly on high-temperature, high-pressure processes
and the efforts directed at first generation coal conver-
sion technologies.
The Office of Energy, Minerals, and Industry has
established some very rigorous objectives for itself over
the next few years. These are in response not only to
internal planning, but in response to the reality that we
all face-that everyone is going to have to do a super job
in order to make sure that the fine rhetoric we have
evolved about bringing environmental and energy goals
together in this emerging area comes to pass. This is a
very, very tough challenge. I n the environmental assess-
ment area, the five objectives shown in table 3 have been
given very high priority by the Office of Energy,
Minerals, and Industry. If, in the future, legislation is
passed through Congress to provide incentives to the
synthetic fuels industry, these five objectives will assume
a very high priority because it will be on the basis of the
words referred to in table 2 that the next generation of
air and water standards will have to be established. I will
refer just briefly to the objectives of the control tech-
nology development program (table 4). As you are well
aware, most of the control technology work will be
carried out as part of the energy systems development.
There will be generic types of control technology
needs which are closely related to control technology
work being done in other areas by the Industrial Envi-
ronmental Research Laboratory at the Research Triangle
Park. I think that technology transfer from some of
these areas to the synthetic fuels area can be very valu-
able. The magnitude of these efforts will depend both on
the results of the environmental assessment program
10
-------
CONTROL
TECHNOLOGY
.....
.....
CONSERVATION
AND-
ADVANCED ENERGY
SYSTEMS
UTILITY
AND
INDUSTRIAL
COMBUSTION
EPA
IERl - RTP
JERl - CINC
EPA
IERl - RTP
IERL - CINC
TVA
ENERGY /ENVIRorJMENTAl
R&D
PROGRAM
INTEGRATED
TECHNOLOGY
ASSESSMENT
EPA
HDQTRS
TV A
ERDA
USDA
HUD
ENVIRONMENTAL
PROCESSES
AND EFFECTS
FUEL
PROCESSING
ENERGY RESOURCE
EXTRACTION
AND
HANDLING
EPA
IERl - RTP
IERl - CINC
ERDA
USBM
EPA
IERl - CINC
USDA
ERDA
Figure 2. Energy/environmental R&D program.
-------
I\J
30M
,...
NUCLEAR
- NUCLEAR SYNTHETIC FUELS
NUCLEAR
RESID. OIL GASIF. RESID. GASIF. OIL
NUCLEAR
SYNTHETIC FUELS
SYNTHETIC FUELS RESID.OIL:GASIF.
COAL CLEANING
- SYNTHETIC FUELS
COAL CLEANING
COAL CLEANING
COAL CLEANING
FBC FBC
FBC
FBC
--
20M
10M
o
1975
1976
-1977
1978-1980
ANNUAL AVERAGE
OF 3 YEARS
Figure 3. Funding summary for fuel processing program.
-------
Table 2. Results of synthetic fuels
program to date
. Analyzed over 1,500 coal samples from more than
100 eastern and midwestern coal sources and char-
acterized them for pollution potential.
. Identified potential pollutant releases by seven
conversion processes:
- Koppers-Totzek gasification
- Synthane gasification
- Lurgi gasification
- CO2 acceptor gasification
Bigas gasification
- COED liquefaction
- SRC liquefaction
. Developed and demonstrated highly effective desul-
furization process for high-temperature gas stream
(at bench scale)
. Completed an analysis of high-temperature versus
low-temperature cleanup of gas streams
. Completed an analysis of problems and opportunities
in retrofitting industrial processes to uti lize low-Btu
gas
. Examined commercial-scale gasification plants in
five foreign countries. Contracted for operational
data and pollutant-emission measurement program
on Lurgi un its in several countries
. Sponsored a symposium in 1974 that produced
a comprehensive report on the state-of-knowledge
on environmental effects on fuel conversion proc-
esses
. Published a survey of potentially hazardous emissions
from the extraction and processing of coal and oil
Table 3. Synthetic fuel program
environmental assessment
objectives
. Perform environmental testing of first generation
coal conversion processes as negotiations permit.
- Lurgi units in Yugoslavia, South Africa, and
Poland
- F ischer-Tropsch in South Africa
. Assess the effectiveness of available control tech-
nology for first generation coal conversion proc-
esses and prepare standards of practice manuals
for environmental control.
. Perform environmental testing of second genera-
tion coal conversion processes being developed by
ERDA.
- Hygas, Institute of Gas Technology, Chicago,
Illinois
- Solvent refined coal, Pittsburgh and Midway,
Fort Lewis, Washington
- CO2 acceptor, consol, Rapid City, South
Dakota
- Synthane, ERDA, Bruceton, Pennsylvania
. Determine requirements for control technology
development for second generation systems.
. Review proposed environmental standards from
the R&D viewpoint.
Table 4. Synthetic fuel program
control technology develop-
ment objectives
. Evaluate, develop, and/or demonstrate, as needed,
best practicable control technology for:
- Fuel storage, handling, and preparation
- Raw product purification
- Product and byproduct utilization
- Waste treatment and disposal
. Provide environmental support and guidance on
control technology development to ERDA and the
industrial sector in their coal conversion R&D
13
-------
AFF
Sector
Group
Issues/and
Priorities
Consideration
....
~
Recommendations
./
/'
./
./
./
./
./
./
'"
EPD
Development of
dynamic environ-
mental R&D
Program
IERL - RTP
Program Execution
OEM I
R&D
Pro ram Development
lAG Program
--.--
-"
I and EPD
Development of
dynamic environ-
mental R&D
Pro ram
IERL - RTP
Program Execution
.-.---
---------
Figure 4. Advanced fossil sector group relationship.
Pass through
Advisory
--
DOC
DOl
ERDA
FEA
FPC
HEW
HUD
NASA
NBS
TVA
USDA
-------
referred to in table 3 and on the relative magnitude of
efforts by E R DA and private industry to provide the
advanced control technology.
Before closing, there is one final activity in my
office that I would like to mention. In order to promote
communication and insure EPA's R&D program obtains
advice from the best sources, I have formed three sector
groups. One of these-the Advanced Fossil Fuels Sector
Group, which is chaired by Dr. Gary Foley-has the re-
sponsibility for the synthetic fuel program. On figure 4
you can see the relationships between the sector groups,
which is shown in the center le"ft, and the rest of the
office structure. The purpose is to bring together all of
those people in the energy development area-environ-
mental control technology development, environmental
effects studies, regulatory efforts, and industrial
activities. The membership of the sector groups is shown
in table 5 and, as you can see, there is a very wide
representation from the various interest groups. The
sector groups are overseen by an executive committee
which plans the agenda for a meeting, summarizes the
meeting resu Its and recommendations, and provides the
recommendations to the Office of Energy, Minerals, and
I ndustry for incorporation as appropriate in the planning
of our future research and development programs. To
date, we have had two successful meetings of the group
within the past 6 months, and we look forward to many
futu re successes.
That concludes my presentation this morning. I
wish you very well in this session and hope that this
serves as the point at which we really take off to do the
hard-nosed environmental assessment and control tech-
nology development which we are going to have to do.
We have done a great deal of talking over the last couple
Table 5. Advanced fossil fuel sector
group meeting composition.
. Executive Board - EPA representatives (responsible
for agenda)
. Sector group membership
- Government
EPA (R&D, program offices, regions)
ERDA (environment and safety, fossil energy)
DOl (USGS, USBM)
FEA
NIOSH-HEW
DOD (N.avy)
TVA
- Industry and research groups
Consultants from:
American Petro leu m Institute
Coal industry
Gas industry
Electric Power Research Institute
EPA R&D contractors
of years, and now we have to really begin with the im-
portant work that is at hand. I look forward to having
many of you represented in this room work with us and
with the other Federal agencies in making sure that all of
our fine rhetoric is not just so many hollow words.
Thank you very much.
15
-------
16
-------
THE PRESIDENT'S SYNTHETIC FUELS
COMMERCIALIZATION PROGRAM
Matthew J. Reilly*
Let me talk a little bit about the background to this
paper. This is an effort that began in January 1975. It
was initiated by the President's State of the Union mes-
sage in which he referred to the need to both develop
new technology for synthetic fuel, but at the same time
see what needed to be done to remove some of the
barriers and some of the constraints toward the intra-
structure to the environmental, social, economic, insti-
tutional, refinement, regulatory, mining, arid transpor-
tation aspect of synthetic fuels commercialization. And
so, the President has directed that an interagency task
force be formed with the objectives of seeing what could
be done to use existing technology and encourage the
private sector through a program of consensus, to build
and operate first off accounting commercial skills into a
plant, emphasizing domestic resources, coal, oil shale,
and solid waste.
Now, a variety of financial incentives was men-
tioned by the President in his message, including low
price guarantees and direct grants. And, incidentally, the
Congressional action on Thursday was one in which they
took exception-at least the House took exception-with
the loan guarantee portion of this program. ERDA has
existing authority to implement the other aspects of the
program, and while the House's view was a disappoint-
ment, by no means does that mean the program is dead.
The output of this interagency task force was a
two-phase program recommendation. Alternately aiming
at a million barrels per day, but with a first phase target
of 350 thousand barrels of synthetic fuel equivalent
before 1985, it was viewed that th is two-phase effort
would provide information that would allow one to
determine the best way to structure the second phase,
and to do this with minimum environmental and other
risks.
Now, I am going to talk principally about the envi-
ronmental assessment, their other aspects of this pro-
gram, and I can refer you to the task force's four-volume
report which will be coming out the end of this month.
The last volume to go to the printers was "Environ-
mental Assessment" which is in the hands of the printers
now. As soon as we get that volume back-and it runs
about 1,000 pages-all four volumes wi II be released to
*Dr. Reilly is acting Assistant Director for Environment and
Safety, ERDA, Washington, D.C.
the public as well as sent to the Council on Environ-
mental Quality and other Federal agencies. The other
volumes deal with an overview of the program with an
analysis of its costs and benefits and a discussion of the
various incentives that were considered and their break-
down and cost over time. Volume 4 deals with the envi-
ronmental assessment.
We begin by looking at five coal regions and the
principal oil shale regions that were viewed as being
those parts of the country most likely to receive signifi-
cant development within the time period. We selected
illustrative synthetic fuel technologies emphasizing those
that were believed to be commercially available today,
but also looking at some second-generation technologies
that were coming down the pike. We selected the
uniformed size (roughly equivalent to 250 million cubic
feet per day) by Btu gasification plant and analyzed the
impact from units in fuel plants in each part of the
country, using various types of coal and other regional
differences. Then we looked at the impact from the
conjunctive development such as mining, transportation,
community expansion, water supply development, and
so forth. This allowed us to assess what the impact of a
single central plant of these various types would be in
each of the geographical regions analyzed. In order to
get a handle on the total aggregate impact of the pro-
gram, we examined some alternative ways in which a
million- barrels-per-day syn-fuel industry might be com-
posed. We looked at several different production levels,
350 thousand barrels per day, one million, and 1.7. The
impacts for each of these alternative industry composi-
tions aggregated essentially as a matter of multiplying
together as a first order of approximation the impact of
a single plant by the number of plants that would go in
each part of the country. After looking at the unavoid-
able impact of the resource commitments, alternatives to
the syn-fuel program were analyzed; these are covered in
detail in the Environmental Impact Statement. And
finally, a unique aspect of this effort was the design then
in the program of an environmental protection strategy
to reduce and litigate the environmental effects. I will
talk more about this strategy later.
Let us consider the five regions by coal-producing
regions and the oil shale region that were analyzed.
Appalachian and Eastern interior offering principally a
high sulfur bituminous coal for union offering, lignite
out of rivers for a subbituminous coal, and so on. We
looked at the plans in terms of their size and the ener~,J
17
-------
value, for all of the cc'al and oil shale answers are sub-
stantially the same and bring only about 10 percent. The
biodegradable synthetic fuel from urban waste is smaller
and it strives to function according to the amount of
solid waste that would be available from a typical com-
munity of about 100 to 400 thousand population. Now
I will not go into the details of analyzing the impact
from a single plant. I think you can pretty well under-
stand the methodology to be followed. For example,
knowing what went into the plant in the way of coal,
sulfur content, etc., we could compute what the emis-
sions would be and then, doing some air quality model-
ing, we could determine whether the Federal air quality
standards would be satisfied for the major pollutant. In
this case, and under several atmospheric conditions, we
satisfied ourselves that a Fisher truck plant located in
the Appalachian region would, in fact, satisfy the three-
R sulfur oxide concentration. This type of calculation
was repeated then for all of the air pollutants and in the
water area as well. We also looked at the socioeconomic
impact, where we examined the buildup of the work
force of the time in the construction phase and in the
operational phase. For a high-Btu gasification plant, the
number of employees would peak at about 3,000 during
the construction phase and then level off to a steady
state of about 800 employees during operation. Based
upon the number of employees in using various socio-
economic multipliers, you ,could then determine the
number of police cars, hospital beds, family secondary
industries, etc., that would be required in the vicinity of
the plant.
Having thus determined what the impact of a single
plant would be, we looked at three different ways in
which the industry might be composed, reflecting three
different emphases on the fuel track. A pipeline gas
emphasis with a high-Btu gas product, roughly half of
the million-barrels-per-day would be high-Btu gas. Now
since the goal of the programs proceeds to be of infor-
mational nature, all of these emphases contain a mix of
all the different types of syn-fuel plants; so even in a
high-Btu emphasis, in one we have a syn-crude from
coal, syn-crude in oil shale, etc., included in the mill ion
barrels per day.
The two other emphases we examined were an
industrial utility low Btu-fuel emphasis and a premium-
liquid emphasis in which the syn-fuel production would
be larger.
We note that the water requirements in terms of
gallons per year are greatest in the Eastern interior and
in the Appalachian region in all of the cases. This reflects
in part the fact that the plant would be designed differ-
ently in the Eastern and Western regions of the country.
Some of the key findings for water quality were
concluded based upon also examining the availability of
water-would water be abundant in the Appalachian and
Eastern interior regions for the million barrels per day
industry. Sufficient water would also be available in the
Powder River and Fork Union region and in the four
corners of oil shale regions, but diversion from its pre-
sent use as two-cent fuel would present various prob-
lems, and these are discussed in more detail in the state-
ment.
We examined various impacts of the mining activi-
ties from downwater in quality as well as the effect of
removing water from the streams and thus reducing the
flow of those streams.
We find it noteworthy that the principle source of
emissions would be the boilers and other process heat
equipment, rather than the actual reactors and pieces of
process equipment itself. For the most part, air quality
standards can be met or somewhat constrained and one
would not be able to locate the essential plants near
major existing pollution forces or in locations with
severe air quality standards. The matter of carcinogens
received extensive treatment in the statement. This is
one that limited experience, as this case indicates, is an
industrial occupational health concern-the one that,
with prudent occupational health measures, can be con-
trolled. Our environmental protection strategy, to which
I referred previously, calls for a comprehensive monitor-
ing program to insure that this problem does not get out
of hand. With respect to mining, we will ask by the
findings here, particulates of nitrogen oxides represent
the main pollutant, but we received no significant area-
wide impact resulting.
Now about the resource requirements, I think the
note-worthy point is that in all cases, over the 20-year
life of a one-million-barrel-per-day syn-fuel industry with
all of the three different cases at which we looked, the
maximum drawdown of the available resources would be
less than 3 percent in any region of the country. The
finding here surprises us if there are adequate domestic
resources available to support an industry. We point to
the loss of water resources due to the damaging of aqui-
fers during the mining activity; the possible loss of agri-
cultural value in the land due to the mining activity; the
loss from scenic resources, and the fact that here in
terms of the land area affected, it is the community
expansion and the conjunctive development that will, in
fact, require more land area than the mining of the
plant. And insofar as the community expansion is direct-
ly associated with people, and these people are going to
need to be housed and provided services in any event, it
is somewhat uncerta in the extent to which that is a
18
-------
direct result of this program. There are some key find-
ings with regard to the biological environment: the
habitat disturbance leading to new climax communities
even after the land has been restored; the revegetation
and reclamation problems that are good fault in the
semiarid areas of the West; the fact that moving into
some of the remote areas and opening them up to
increased population would also take the stress on some
of those animal species that are least able to tolerate the
proximities of man; and the impairment of the habitat
due to the impact on waterfall being stream flow-
With respect to the increase in regional population,
a key increment of the socioeconomic impact, we note
the increase in the regional population as a percent due
to the one-million-barrel-per-day industry.
It is to be seen that the oil shale region presents the
largest population increase, approximately 55 percent
over the present population in the areas. This would take
place over the entire tenure or lifetime of the program
leading to an average increase of 6 percent per year. That
figure would be a manageable one if it were distributed
evenly in this region. If it were not, that potential prob-
lem would then be impacted upon the environmental
protection strategy that I will present shortly.
Other key findings in this socioeconomic area are in
addition to the problems of large and rapid population
increases. The changes in community lifestyles, the
increased demands upon community services, and facili-
ties there are beneficial impacts. Economic stimulus to
the community would result, and there would be an
increase in the tax base of new job opportunities. Final-
Iy, we note that the development will cause a permanent
change in the remote areas from rural lifestyles into an
urbanized industrial one. The environmental protection
strategy then brought into the program is the direct
result of the environmental assessments which were
formed, and at this point let me digress. We began the
environmental analysis at the same time that the other
task forces were looking to the cost and tt',e technol-
ogy-the environment was not a weak sister that came in
at the tail end of everything and an after-the-fact justifi-
cation-what was it that one sought to do.
It was integral coming in at the very beginning and
as a result inherent to the program design of some of
these key elements-for essentially, the Environmental
Advisory Board would represent the regions that would
be affected, would represent the technical and scientific
community, and would represent various government
agencies including the Energy Research and Develop-
ment Administration.
Environmental criteria would be essential in the
proposal evaluation process. There are procedures devel-
oped and will be published in the Federal Register for
State, local, and public involvement in the program like
specific development plans. Environmental impact state-
ments will accompany each major action under this
program. Federal approval of the site developments
would be a prerequisite to proceeding with the construc-
tion phase. Public hearings and opportunities for com-
ments would be provided. The regional impact assistance
is in the form of guaranteed loans to the impacted local
communities, so that in the event that a facility would
fail, the local community would not be left holding the
bag for its front-end investment in intrastructure costs. I
refer to the environmental mon itoring program, the
strict enforcement of the environmental standards.
Finally, the impact statement when it has identified
some of the uncertainties at the same time identifies the
type of research that would be needed to resolve those
uncerta inties.
There are some key areas here for environmental
research. First, the effluent from fuel sites Commercial
Processes Program would need to receive careful atten-
tion in our data-gathering activities. The essential pro-
gram would include within it research on improved con-
trol technology on the effect of the pollution waste
management technique for disposal of the solid waste,
particularly, resulting from the oil shale processes and
citing criteria needed to be a part, if you like, of the
handbook which could accompany these technologies
when they are presented to the Nation as commercially
feasible.
Second, the intrastructural requirements-how one
in fact does go about providing for community expan-
sion when placed under the stresses of rapid population
growth and do this in a prudent and economical fashion.
Third, optimum water usage. Until we actually put
the best engineering talents to work in designing these
full-scale commercial plants, we do not really know what
might be the best we can do in the way of conserving
water. This program will bring it forth.
Fourth, utility planning. Regular budgetation and
reclamation are two other areas' in which there are
research opportunities in this program.
Well, that concludes my presentation. Again let me
mention that the details of th is environmental assess-
ment are contained in our draft environmental impact
statement which is at the printers now. I would hope
that we can deliver it as a Christmas present to CQ.
Thank you.
19
-------
ENVIRONMENTAL CONTROL AND
SYNTHETIC FUELS
S. B. Alpert*
The Electric Power Research Institute (EP R I) is the
central research agency of the electricity industry. The
organization has been in full operation for over two
years and has launched a major program in coal conver-
sion and utilization that has as one of its major objec-
tives the economic utilization of the U.S. coal supply to
provide fuel for electricity manufacture in an environ-
mentally acceptable manner.
Since we are meeting in Florida, we must use the
local football fever related to the upcoming Super Bowl
game to draw an analogy to the status of synthetic fuels
and the upcoming papers in this session. The process
technology game is such that we are more or less just
starting. The team is assembled on the 20-yard line
having just received the kick-off. We have a game plan,
and the ball has been snapped to our quarterback. He
throws a pass down the field. As the receiver heads down
field, a guy runs alongside writing rules that mayor may
not declare the playas legal. That is the role of the State
and Federal EPA. In addition, as he turns to catch the
ball, he is being kicked in the shin. That represents the
environmentalist. Besides all this, and to perhaps push
the analogy a bit too far, at the end of the field the goal
posts are being moved from the 100-yard line and we are
not really sure where the goal line will be set up. That
represents the inflationary pressures that have escalated
the cost of synthetic fuel development from what were
considered $200 million problems, to what now are
apparently $1 billion investments that are required to
develop and demonstrate commercial power systems
that utilize coal gasification of coal liquefaction. The
risks with respect to private investment are rather large,
and it is doubtful that any rational industrial manage-
ment will have a reason for assuming such complex tasks
without substantial Federal support.
Table 1 shows the 1976 budget of the Fossil Fuel
Department. The electricity industry is prepared to pay
for support of the research and development necessary
to develop synthetic fuels. We anticipate that in future
years, our expenditures will increase and we will have a
substantial role to play in carrying out a directed pro-
gram in cooperation with industry and the Energy Re-
search and Development Administration (ERDA). The
Fossil Fuel Department budget is a significant part of
*The author is with the Electric Power Research Institute,
Palo Alto, California.
the overall Institute budget of about $130 million in
1976.
Table 2 shows the capacity of typical commercial
plants that daily handle many thousands of tons of
solids and liquids to produce the products required by
our industrial society. The engineer has learned to take
advantage of continuous processing systems that have
reduced the costs of products to the public. None of the
process technologies that are competitive alternates to
current methods of supplying fuel for electricity genera-
tion has achieved such a status wherein it can be applied
at such scales. We still have a great deal to learn before
such plants can be confidentally designed and operated
with components engineered for the service.
Table 1.
Fossil Fuel Department Proposed
1976 Research and Development
Plan Department Summary
Program
Funding Level ($106)
Gasification
12.0
10.9
3.7
12.2
Liquefaction
Direct Utilization
Environmental Control & Combustion
Supporting Research
TOTAL
....!:L
40.5
Table 2. Size of U.S. Industrial
Plants in Tons/Day
Coal Fired Power Stations
Oil Refineries
12,000
50,000
8,000
2,000
2,000
Cement Plants
Ethylene Plants
Ammonia Plants
21
-------
A major area being supported by EPR I is the re-
search and development required to convert coals to
clean liquid fuels. Such products range from solid fuels
cleaned of mineral matter and reduced in sulfur content
to highly refined clean liquids that represent turbine
fue.ls very much the same as fuels that derive from
petroleum. Our program is relatively broad, but our
major emphasis is in support of:
Solvent Refined Coal
H-Coal
Exxon Donor Solvent
We believe that the three approaches should be pursued.
Solvent Refined Coal (SRC) is a process technology that
results in a clean solid boiler fuel. EPR I supports a 6
ton/day pilot plant in Wilsonville, Alabama in associ-
ation with Southern Services, Inc. Over the last two
years, successful operation of that facility has indicated
that the process can be operated on selected U.S. coals
to produce products to specifications set by the EPA.
The noncatalytic SRC process technology is also being
explored at Tacoma, Washington in a 50 ton/day pilot
plant. Our SRC program is being coordinated with the
federal sponsors of the Tacoma plant. Currently, cost
estimates and pi ant designs are underway for large sized
plants.
It is likely that this technology is ready for demon-
stration at the 1000 ton/day scale although some further
development is still required that promises lower costs.
Before much too long this process should be proved out
in larger equipment. A number of utilities find the SRC
fuel attractive for their systems.
The H-Coal process which represents direct catalytic
coal liquefaction has been under development for a
decade. This technology is one that has been too long at
a small scale in the laboratory. EPR I has joined with
other organizations and ERDA to prove out the process
in a large pilot plant that will handle 250-600 tons/day
of coal.
Recently our Board of Directors approved participa-
tion in the Exxon Donor Solvent technology. A test in
large pilot plant equipment is scheduled for the late
1970's.
We believe that these three technologies represent
process technology that is furthest advanced and that
have strong teams able to complete the required develop-
ment work.
Table 3 summarizes the pilot plants liquefying coal.
When one realizes that about 40,000 barrels a day are
required to fuel a single 1000 MW power station, one
can see how far we still have to go before coalliquefac-
tion can be considered a reality by the power industry.
There are viable technologies, able teams, and estab-
lished organizations with excellent capabilities. The
major missing ingredient is the money to carry out the
required program.
The gasification of coal was practiced in this coun-
try primarily to produce gaseous fuel in historic fuel gas
producers, for illuminating gas and low-heating value
fuel gas for domestic use. That industry all but dis-
appeared after the second world war when low cost
natural gas transmitted to population centers priced
gaseous fuel from coal from the market.
An option that is of interest to the power industry
is the utilization of gaseous fuel from coal in combi-
nation with power generation equipment to produce
electricity economically in plants that can satisfy all en-
vironmental requirements. At this time, there are no
plants in the United States producing gaseous fuel from
coal being used to generate electricity.
Fuel gas from coal has a number of potential
markets in the power industry. These include:
retrofit of gas-fired boilers that have been curtailed
from their natural gas supply,
fuel supply for new boilers that require clean
gaseous fuel free of particulates and pollutants,
fuel supply for applications which use gas turbines
and steam turbines combined for efficient utiliza-
tion of energy, and
generation of raw materials for
methanol production useful for peaking fuel,
hydrogen production,
reducing oxides of sulfur to elemental forms in
advanced scrubber systems, and
fuel supply for fuel cells.
The ideal gaseous component from coal for power gener-
ation is carbon monoxide. In contrast, the ideal gaseous
component for the gas industry is methane. Thus, pro-
cess technology of interest to the power industry that
converts coal to gaseous fuel is a very different kind of
system than that required to supplement natural gas
supply. The emphasis in power systems is on process
technology that can convert coal to gaseous fuel at a
high capacity at modest pressure using relatively simple
systems that have high reliability in conjunction with
power generating equipment. A further requirement is
that such systems be competitive with conventional
pulverized fuel combustion with stack gas scrubbing.
Taking account of the escalated price of crude oil and
the likely shortage in the next ten to twenty years, the
base loaded power station will tend to use nuclear power
since those generating systems are high in capital cost
and low in fuel cost. Thus, gasification plants will be
used to satisfy what the power system engineer refers to
22
-------
Table 3. Operating Pilot Plants Liquefying Coal
(Note: Excludes Bench Scale Units)
Process Status
Solvent Refined Coal Operational
Solvent Refined Coal Operational
CaNSO l CSF Shut-Down
H-Coal Operational
Exxon EDS Operational
Gulf CCl Construction
SynthoiJ Construction
Note: EPRI .. Electric Power Research InstitUte
ERDA= Energy Research and Development Agency
Capacity, Tons/Day
Source of Funds
6.12
Southern Company
50
26
EPRI
ERDA
3
ERDA
ERDA, EPRI, Sun Oil,
Amoco, ARCO, Ashland Oil Co.
Exxon
8
Gulf
ERDA, Bethlehem S"teel
*Smaller scale units at various locations are investigating liquefaction alternates. lummus,
Universal Oil Products, and Continental Oil Co. as well as others have such efforts under-
way. In addition there are several coal liquefaction programs at the University of Utah,
Auburn University, University of North Dakota, etc.
as intermediate load demand. r'ower systems that gasify
coal to produce a fuel supply will therefore need to
follow a demand curve that translates into a capability
for starting up and shutting down under control.
Such power applications put a new requirement on
the application of process technology that is very differ-
ent from the conventional application of coal gasifica-
tion technology for petrochemical applications, such as
ammonia or methanol plants. There is considerable
doubt that coal gasification plants can be "swung" to
follow load demand of a power system.
Gasification using al ready existing technology that
is represented by fixed bed, fluid bed, or entrained
systems is uneconomic in competition with conventional
power industry practice. The purpose of the EPR I pro-
gram in gasification is to develop power systems that
are:
lower in capital cost,
lower in net power cost,
have high efficiency (Heat Rate). and
lower in cooling water consumption.
In order to accomplish these objectives, a key tech-
nological need is the development of gas turbines of high
reliability capable of operating in the range of inlet
temperatures of 2500° F to 3000° F. Such a program of
development is necessary to gain a large share of the
powe r ge neration market. Currently available gas
turbines operate at about 2000° F and are not consider-
ed of sufficient reliability.
Table 4 shows the gasifier characteristics that are
ideally desirable which, when combined with efficient
gas turbines, could produce a power generating system
that has many desirable features.
Table 5 shows the key development issues that are
being addressed in the program that is underway and
being funded by Federal and industry sources. In this
discussion we will not have sufficient time to deal in
depth with each of these issues. In developing a system
for power generation, all of these issues must be con-
sidered simultaneously for the particular gasification
system under development.
Table 6 shows one of the reasons why this simulta-
neous consideration of these issues is needed. The last
column shows the flame temperature of the gaseous fuel
produced from a number of new systems. Note that the
adiabatic flame temperature of some of the fuels is fairly
low. There would be little reason to develop 3000° F gas
turbines if the flame temperature is below the operating
temperature of the gas turbine. The entire system must
be considered in applying coal gasification for power
industry applications.
Table 7 shows a number of the unresolved problems
in developing power systems. These problems are
common to all such systems and underline the need for
development of power systems rather than gasifiers.
23
-------
Table 4. Desirable Gasifier Characteristics
Table 5. Power From Gasification
Combined Cycle Systems Key
Development Issues
. Simplicity
. Easy Maintenance
. Minimum Number of Units and Stages
Integral Control
. Large Gasifier Capacity
. Minimum Coal Processing Upstream, Gas Cleaning
Load Following Capability
- No Tars in Product Gas
- Capability for Feeding a Caking Coal
Turbine Reliability
- Minimum Amount of Fine Dust
Air Versus O2 Gasification
. Capability for Quick Shutdowns and Restarts
. Capability for Processing a Wide Variety of Coals
. No Gross Instabilities
Combustor Design
High Temperature Gas Turbine Development
. Capability for Load Following
. Capability for Operation at a Pressure Level
Suited to the Integrated Combined-Cycle
Installation
Availability of Funds
System Reliability
Environmental Impact
Table 6. Low Btu Gas Characteristics
Clean-up .-Gas Gas Gas HHV Adiabatic
Gasifier System Temp. of Pressure, Psia Btu/Scf Flame Temp. of
Lurgi Hot Pot 230 300 172 3350
BGC Siagger Benfield 230 300 183 3600
C.E. Stretford 105 16 105 2900
B&W Stretford 105 65 84 2500
B& W Benfield 230 65 68 2200
Foster Benfield 230 365 169 3500
Wheeler
Texaco Benfield 230 350 79 2600
Table 8 summarizes an assessment of gasification
technology. In order to resolve the fundamental ques-
tion as to whether a gas turbine can be directly connect-
ed to a coal gasification system, a power system based
on Lurgi technology is to be designed, constructed and
operated by Commonwealth Edison Company. The
system capacity is 25 MW and will utilize currently avail-
able gas turbines. It will also provide a test site for
advanced gas turbines and an opportunity to study the
control system required to permit following changes in
electricity demand while using a number of U.S. coals
having a range of caking properties.
24
-------
Table 7
Effects of System Integration on Design, Interconnection
and Operation of Each of the System Components:
. Influence of Gas Heating Value and Clean Up
System on Combustor Design.
. Dual Fuel Capabilities and Gas Turbine Mismatch.
. Effects of Gas Quality Variations on Gas Turbine Operation.
. Gas Storage Requirements for System Buffering.
. Control of Multicomponent Integrated System.
Table 8. Coal Gasification
Development Status
Currently Commercial
Lurgi
Koppers-Totzek
Winkler
Atmospheric Producers
- Fixed
- Entrained
- Fluid
- Fixed
Improved Processes of Near Term Potential (Pre- 1985)
British Gas Siagging
Babcock & Wilcox
Texaco
Combustion Engineering
Pressurized Koppers-Totzek
- Fixed
- Entrained
- Entrained
- Entrained
- Entrained
Second Generation Processes (Post- 1985)
Foster Wheeler
IGT Fluid Bed
Westinghouse
COGAS
Conoco
Union Carbide - Battelle
Advanced Concepts
Atomics International
Fast Fluidized Bed
- Entrained
- Fluid
- Fluid
- Fluid
- Fluid
- Fluid
- Molten Carbonate
The EPRI progrem also includes support of the
British Gas Council slagging gasifier. This development,
which is cofunded with 13 other organizations, is under-
way at Westfield, Scotland. It represents a significant
advance over Lurgi technology which will not be gone
into in detail. The slagging gasifier is in an advanced state
of development and if steady progress is realized could
represent an evolutionary step for a system generating
power.
EPRI has also joined with ERDA and the Combus-
tion Engineering Company in a pilot plant test of an
atmospheric pressure two-stage entrained gasifier. This
technology seems to represent a low-cost producer of
gaseous fuel suitable for power generation.
The fossil fuel department also has programs in:
direct utilization of coal, and
environment and combustion.
These su bjects are outside the interest of this conference
and will not be discussed.
At th is time, synthetic fuels from coal are at the
pilot plant level. The power industry, through EPRI, has
underway a program that will lead to power system
options in the next decade that are economic. Environ-
mental control of emissions from synthetic plants, while
of importance, could represent a counter productive
block in the development of new technology. There
would seem little benefit to setting emission standards
for promising new sources until the operating charac-
teristics, performance and costs are firmly established.
Through cooperation between industry and government,
sensible approaches to environmental control need to
evolve. The time span to commercialization is a long
one. Even for relatively advanced technology, ten to
fifteen years elapse. The rules are obsure. For example,
it is possible that after a process is completely demon-
strated, the fuel would not be acceptable and the billion
dollars invested would be wasted. Such an uncertainty
represents one of the factors that have discouraged
private investment and represents another risk of devel-
opment. It is a challenge for regulatory bodies to reduce
such risks by establishing realistic objectives and permit
the development community to develop new technology
without such complications.
CONCLUSION
It will require a great deal of joint industry and
federal cooperation to develop synthetic fuels tech-
nology to a status where the electricity industry will be
25
-------
able to commercialize new methods which are reliable,
economic, efficient and environmentally acceptable. The
accomplishment of these objectives is no mean task, but
one in which the Electric Power Research Institute is
playing a significant role.
ACKNOWLEDGMENT
The author wishes to acknowledge the contributions
of R. Wolk, N. Holt, and M. Gluckman to this paper.
The opinions expressed are those of the author and do
not necessarily represent those of the Electric Power
Research Institute.
26
-------
FEDERAL REGULATION OF ATMOSPHERIC EMISSIONS
FROM COAL GASIFICATION PLANTS
Abstract
James F. Durham*
The U.S. Environmental Protection Agency's (EPA)
intentions regarding the development of standards that
regulate the atmospheric emissions from advanced fossil
fuel conversion facilities will be outlined. As a further
indication of the possible course future Federal regula-
tions will take, the provisions of the statute likely to be
used to regulate these emerging industries along with a
discussion of the Congressional intent and pertinent
judicial decisions concerning the statute will be re-
viewed. In addition, the factors that must be considered
in the development of a standard and the criteria that
are used in the selection of sources and pollutants will be
discussed. The facilities and pollutants within coal gasifi-
cation plants that are currently under consideration for
standard development will be reviewed. Finally the
advantages and disadvantages of the possible standard
formats as related to technological capability, monitor-
ing of emissions and determination of compliance will be
presented. Hopefully the information presented here will
provide guidance to private industry and other govern-
ment agencies responsible for developing and installing
energy conversion processes.
INTRODUCTION
This presentation outlines the U.S. Environmental
Protection Agency's (EPA) intentions regarding the
development of standards that regulate the atmospheric
emissions from advanced fossil fuel conversion facilities
such as coal I iquefaction or coal gasification plants and
oil shale processing facilities. As a further indication of
the possible course future Federal regulations will take,
the provisions of the statute likely to be used to regu late
these emerging industries along with a discussion of the
Congressional intent and pertinent judicial decisions con-
cerning the statute are reviewed. The factors that must
be considered in the development of a standard, the cri-
teria that are used in the selection of sources and pollu-
tants and the facilitie£ within coal gasification plants
that' are currently under consideration for standard
development are outlined. In addition, the procedures
used to enable interested parties to review and comment
*The author is with the U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina.
on Agency findings prior to formal proposal of regula-
tions in the Federal Register are discussed. Hopefully,
the information presented will provide gu idance to pri-
vate industry and other government agencies responsible
for developing and installing energy conversion proc-
esses.
AGENCY INTENTIONS
EPA plans to use Section 111 of the Clean Air Act
of 1970, "Standards of Performance for New Stationary
Sources"-more commonly called New Source Perform-
ance Standards (NSPS), as the primary mechanism for
Federal regulation of emissions from emerging industries
such as advanced fossil fuel conversion facilities. This
decision is consistent with the legislative intent of Sec-
tion 111 as it is the least costly means of preventing the
occurrence of new emission problems. Additional regula-
tion will likely occur through Section 110 of the Act
which requires that emissions be maintained at levels
that are necessary to comply with the Federal ambient
air quality standards or to prevent significant deteriora-
tion of the ambient air. It is expected that New Source
Performance Standards will reduce futu re problems in
terms of nonuniform State and local regu lations. In
addition, the regulations and concepts being considered
to prevent significant deterioration rely to a significant
degree on NSPS to define the best available control tech-
nology that can be used to minimize degradation from
major new sources.
It is not likely that Section 112 of the Act, "Na-
tional Emission Standards for Hazardous Pollutants,"
will be used to regulate advanced fossil fuel conversion
facilities. Section 111 provides EPA with a great deal of
freedom regarding the type of pollutants that can be
regulated provided it is prudent to do so. This would
include pollutants that are or could be classified to be
hazardous under Section 112 or for which Federal
ambient air quality standards exist or are under consider-
ation. Section 111 is particularly attractive for regulating
emerging industries because new emission problems can
be minimized at a considerable savings' in time and
resources at the Federal, State, and local level. The sav-
ings will result since there will be few if any existing
sources to which the standards would be applicable.
Therefore, controlling certain pollutants from existing
27
-------
sources as is required under Section 111 (d) of the Act
will not be necessary. Nor will it be necessary to con-
sider all significant existing sources of hazardous pollut-
ants as would be required under Section 112 of the Act.
The timely development of NSPS could also result
in a considerable savings to organizations responsible for
developing and operating advanced fossil fuel conversion
facilities. Early direction regarding regulatory require-
ments will assist in infusing adequate consideration of
the environmental factors into the overall assessment of
the technical and economic viability of these processes.
The environmental problems associated with advanced
fossil fuel conversion processes should not be any more
difficult to resolve than those encountered in the devel-
opment of the basic energy conversion process. If such
problems do exist and cannot be resolved at a reasonable
cost, identification early in the development process will
lead to a more rapid assessment of the value of the
process and could result in a substantial savings in time
and resources. Moreover, new facilities not equipped
with best available emission control technology run the
risk of eventually having to retrofit the facility due to
possible changes in State and local regulations. Such
retrofit costs are in general significantly more expensive
than installing the control equipment at the time of con-
struction.
Considering the nation's current concern for the
environment, it is unlikely that advanced fossil fuel con-
version processes will reach commercialization without
being reasonably certain that significant environmental
problems will not be created and that all reasonable pre-
cautions have been taken to ensu re that emissions to the
environment have been minimized. A detailed EPA
assessment of what can be done to minimize emissions
could lead to a more rapid acquisition of the necessary
State construction and operating permits and Federal
agency approvals; will tend to satisfy environmental
groups; and will assist in obtaining the necessary financ-
ing.
There are many advanced fossil fuel conversion
processes currently under development-most in the
early stages of development. Thus, there is ample time to
adequately assess the environmental considerations with-
out delaying the commercialization of the most promis-
ing processes. It is possible that some processes currently
under development may be terminated due to environ-
mental considerations; however, such termination would
be in the best national interest and should not be con-
strued to mean that EPA is delaying the nation's goal of
achieving energy independence.
Therefore, as the regulatory body responsible for
insuring that new emission problems are prevented, the
Emission Standards and Engineering Division with EPA
will continue to review the environmental problems as-
sociated with advanced fossil fuel conversion facilities
and to develop standards (as sufficient information
becomes available) consistent with the legal provisions
and the Congressional intent of Section 111.
PROVISIONS AND- CONGRESSIONAL INTENT
Section 111 of the Act requires the establishment of
standards of performance for new stationary sources
that" . . . may contribute sign ificantly to air pollution
wh ich causes or contributes to the endangerment of
public health or welfare." The Act requires that stand-
ards of performance for such sources reflect" . . . the
degree of emission limitation achievable through the
application of the best system of emission reduction
which (taking into account the cost of achieving such
reduction) the Administrator determines has been ade-
quately demonstrated."
The overriding purpose of performance standards
for new stationary sources is to prevent the occurrence
of new air pollution problems. In establishing Section
111, Congress also sought to maintain existing high
quality air and to ensure uniform national standards for
new facilities within the various stationary source cate-
gories. As indicated below, the purpose of NSPS is clear-
ly stated in the legislative history of the Clean Air Act
(ref. 1).
u. . . We must also guard against new problems. Those
areas which have levels of air quality which are better
than the national standards should not find their air
quality degraded by the construction of new sources.
'These new source performance standards would require
industry to apply the latest available emission control
technology and processes wherever a new plant is locat-
ed. . . . The concept is that wherever we can afford or
require new construction, we should expect to pay the
cost of using the best available technology to prevent
pollution.
"Maintenance of existing high quality air is assured
through provision for maximum control of new major
pollution sources.
"Perhaps one of the most significant provisions in the
bill deals with the establishment of Federal emission
standards for new stationary sources. . . . The purpose of
this new authority is to prevent the occurrence any-
where in the United States significant new air pollution
28
-------
problems arising from such sources either because they
generate extra hazardous pollutants or because they are
large scale polluters. n
Congress intended that regulated stationary source
categories would be designed, constructed, equipped,
and maintained so as to reduce atmospheric emissions to
a minimum. It prescribed that such standards be ex-
pressed as emission limits, thus enabling the owner or
operator to select the most economical and acceptable
means of achieving them. These would include process
changes, operational changes, and direct emission con-
trols, as well as any other practical method.
Congress also intended that NSPS be applied to
modified sources when such modifications result in in-
creased air pollution emissions. Such modifications have
been generally defined as a process or raw material
change that results in an increase in emissions.
Standards of performance were not viewed as being
static. Congress foresaw the need for continued review
of existing standards, and the promulgation of revised
standards as new technology or operating methods
emerged.
Congress also provided that the Administrator in
establishing standards of performance would determine
the degree of control that has been or can be achieved
by means of technology that is or can be made available.
Congress prescribed that such a determination must be
reasonable in terms of the cost of such technology, and
in the time of its availability.
Within the framework of congressional intent, a
number of general concepts that serve to foster the
review of recommended standards of performance and
the support rationale have been derived. These are brief-
ly set forth below:
1, I ndividual standards are not intended to be pro-
tective of health or welfare effects; that is, they are not
designed to achieve any air quality goals. The standards
are designed to reflect best technology for each individ-
ual source. The long-range goal and overriding purpose
of the collective body of standards is to prevent new
pollution problems from developing. To achieve this
end, the standards must force technological change and
the justification for the standards must allow for tech-
nology transfer.
2. Because the standards attempt to define the
best systems of emission reduction, the data base upon
which the standards are justified will necessarily be limit-
ed. Test data on existing well-controlled plants are the
most desirable basis for setting emission limits for new
plants. Since the control of existing plants generally
represents retrofit technology, or is designed to meet
some existing State/local regu lation, even those levels are
subject to improvement. Therefore, the Administrator's
judgment is necessarily involved in every proposed stand-
ard, even when the data base is "adequate."
3. There are no waiver provisions for standards of
performance. Unl ike State/local agencies, EPA does not
have a mechanism to waive the standards when a source
has made a good faith effort to comply but misses the
mark by some amount. Short of not allowing the source
to operate, EPA can: (a) neglect to enforce the standard,
since use of enforcement authority is discretionary; or
(b) issue an order that would provide the source some
finite time to improve the control system to achieve the
standards. Both the options are subject to policy
changes. Therefore, the standards must be set so the
probability of compliance is high. This, of course, is con-
tradictory with the mandate to set standards that force
technology.
CLARIFYING JUDICIAL OPINIONS
The intent of Congress, and the resultant legislative
language, has been further clarified by court opinions
emanating from litigation over promulgated standards of
performance. The Federal Court of Appeals, District of
Columbia Circuit has found that: (1) EPA is required to
consider costs in terms of counter-productive environ-
mental effects as well as the economic impact upon the
industry. (2) EPA is not required to perform a cost-
benefit analysis comparing benefits derived from re-
duced emissions with those costs associated with con-
trol. (3) The court makes it clear that it is sufficient for
EPA to show that a standard can be achieved rather than
that it has been achieved. The essential consideration is
that there will be technology available for installation in
new plants (ref. 2) .
". . . Section 111 looks toward what may fairly be pro-
jected for the regulated future, rather than the state of
the art at present, since it is addressed to standards for
new plants. .. The essential question was rather whether
the technology would be available for installation in new
plants. . . . the question of availability is partially de-
pendent on 'lead time,' the time in which the technology
will have to be available. Since the standards here put
into effect will control new plants immediately, as op-
posed to one or two years in the future, the latitude of
projection is correspondingly narrowed. If actual tests
are not relied on, but instead a prediction is made, 'its
validity as applied to this case rests on the reliability of
[the] assumptions' . . ."
The extent to which EPA must demonstrate achiev-
ability was expanded upon in a second opinion when the
Court held that
29
-------
"It is the system which must be adequately demon-
strated and the standard which must be achievable. This
does not require that a sulfuric acid plant be currently in
operation which can at all times and under all circum-
stances meet the standards, nor, however, does it allow
the EPA to set the standards solely on the basis of its
subjective understanding of the problem or 'crystal ball'
inquiry. . . . An adequately demonstrated system is one
which has been shown to be reasonably reliable, reason-
ably efficient and which can reasonably be expected to
serve the interests of pollution control without be-
coming exorbitantly costly in an economic or environ-
mental way. An achievable standard is one which is with-
in the realm of the adequately demonstrated system's
efficiency and which, while not at a level that is purely
theoretical or experimental, need not necessarily be
routinely achieved within the industry prior to its adop-
tion (ref. 3). "
SELECTION OF SOURCES AND POLLUTANTS
Selection of sources and pollutants for standard
development is based to a significant degree on the anti-
cipated reduction in the mass rate of emission that can
be derived from the standard. This emission reduction
(benefit) is estimated from the emission rates that can be
achieved through the application of available control
techniques. Generally, two or three control techniques
or processing schemes that cover a range from average
control to best control will be evaluated. Existing State
and local regulations and "average" control levels are
used to determine the reduction in the rate of emissions
that would result from applying this technology. The
estimated number and size of new and replacement facil-
ities are then used to calculate the emission reductions
that can be expected in future years (ref. 4).
Other factors are also weighed in the selection of
sources and pollutants. For example, emphasis is cur-
rently on the development of standards for hydro-
carbons and nitrogen oxides. Although it appears that
most areas of the country are in compliance with the
Federal ambient ai r qual ity standard for NOz, many
cities are near the standard. Standards that require new
- .
sources to install best demonstrated technology will
minimize the effect of growth on ambient air quality.
A recent internal report shows that the ambient air
quality oxidant standard will not be maintained in many
urban areas without increased control of stationary
sources (ref. 5). It is becoming obvious that maximum
control of all hydrocarbon sources nationwide is the
only way to handle the oxidant problem.
Recent findings have additional implications for
control of the nonurban oxidants problem. Both the
more reactive and less reactive organics, under the con-
ditions of light winds associated with persistent high
pressure systems, may have sufficient time to generate
oxidants. Thus, reduction of total emissions and not just
reduction of the more reactive compounds would need
to be emphasized in nonurban areas. Control of organic
rather than nitrogen oxide emissions appears to be the
most effective route to reduce oxidant levels in rural
areas. However, to attain the standard, coordination of
both organic and nitrogen oxides emission control may
be necessary (ref. 6).
Additional benefits that can be derived from a
standard, such as heat or resource recovery or the reduc-
tion of toxic metals, carcinogens, or other hazardous
materials are also considered in the selection process.
GENERAL APPROACH FOR COAL
GASIFICATION PLANTS
Before standards are promulgated, EPA must
establish the magnitude of the emission problem,
identify the technology available to reduce the emis-
sions, and identify the emission levels and emission re-
ductions that would result from the application of that
technology. I n addition, the penalties associated with
the technology, such as investment and operating costs,
economic impact on the industry, consumption of
energy or other natural resources, and adverse environ-
mental problems, must be defined. The Agency must
then determine if these factors are reasonable. Informa-
tion obtained from foreign companies operating coal
gasification facilities, from EPA contractor studies and
from U.S. companies that are planning to install full-
scale commercial coal gasification facilities is being used
to assess these factors. The Agency is relying heavily on
the detailed engineering and thought that has gone into
the proposed U.S. installations.
Our initial standards for coal gasification plants will
likely focus on restricting the atmospheric emissions of
the gaseous sulfur compounds and non-methane hydro-
carbons formed in the gasifiers. The potential emissions
are significant and there appears to be adequate informa-
tion available to identify emission levels that can be
achieved with available technology.
The procedures that are followed during the course
of the standard development program provide ample
time to enable industry, environmental groups and other
interested parties to prepare su bstantive comments on
the Agency's initial conclusions well before proposed
regulations are placed in the Federal Register.
A Standard Support Environmental Impact State-
ment (SSE IS) will be prepared that contains the
Agency's findings concerning the technology that is con-
30
-------
sidered demonstrated, the emission levels that can be
achieved and the environmental and economic costs of
applying the technology. The document will be distrib-
uted to the National Air Pollution Control Techniques
Advisory Committee * (NAPCT AC), to industry trade
associations and other interested parties along with the
Agency's preliminary recommendations. Approximately
one month after distribution of the SSE IS, it will be
reviewed in an announced public meeting with the
National Air Pollution Control Techniques Advisory
Committee. Comments received at the NAPCTAC meet-
ing are given serious consideration before proposed
standards are prepared for publication in the Federal
Register.
SUMMARY
EPA intends to vigorously proceed with the devel-
opment of standards to regulate the atmospheric emis-
sions from advanced fossil fuel conversion facilities.
Regulation of these emerging industries will likely be
implemented through Section 111 of the Clean Air Act
of 1970, "Standards of Performance for New Stationary
Sources." The overriding purpose of standards of per-
formance is to prevent the occurrence of new air pollu-
tion problems. Standards will be developed as informa-
tion becomes available that satisfies the legal provisions
and Congressional intent of Section 111.
Early direction regarding the regulatory require-
ments for advanced fossil fuel conversion facilities will
provide incentive to adequately consider environmental
factors in the overall assessment of the technical and
econor"ic viabil ity of advanced fossil fuel conversion
facilities. Identification of serious environmental prob-
lems early in the development process is in the best na-
tional interest since it could result in a substantial sav-
ings in time and resources.
The majority of the fuel conversion processes are in
the early stages of development. Thus, there is ample
time to consider environmental factors without delaying
the nation's goal of achieving energy independence.
Review of the provisions of Section 111, the Con-
gressional intent of the statute, and the pertinent judicial
decisions indicates that EPA must demonstrate that a
new source performance standard can be achieved, that
* An advisory group established pursuant to sections
108(b)(1) and (2) and 1171f) of the Clean Air Act as amended.
The Committee advises on information documents regarding air
pollution control techniques and testing and monitoring method-
ology for categories of new sources and air pollutants subject to
the provisions of Sections 111 and 112 of the Clean Air Act as
amended.
the technology is available, and that the resulting costs,
economic impact, energy consumption, and adverse envI-
ronmental effects are reasonable when compared to the
reduction in atmospheric emissions that will occur.
Thus, EPA efforts will be directed toward identifying
and basing standards on processes or techniques that can
achieve the lowest emission levels consistent with reason-
able cost in terms of dollars, energy consumption, and
adverse environmental penalties.
Consideration of the Congressional intent and legal
provisions of Section 111 should provide substantive
guidance regarding future Federal regulatory require-
ments to those responsible for developing, designing, and
operating advanced fossil fuel conversion processes.
A major factor used to select sources and pollutants
for standard development is the resulting reduction in
the mass rate of emissions from new or replacement fa-
cilities. Pollutant selection is also based on the effect the
standards would have on maintaining or achieving the
Federal ambient air quality standards. Additional factors
considered include the prevention of significant deterior-
ation of air quality, the reduction of highly toxic mate-
rials, and the energy savings that would result from a
standard.
Since available data will be more limited for emerg-
ing industries, more engineering judgment and transla-
tion of technology from other application areas will be
required than has been used in the past on more estab-
lished industries. As a result, early standards will likely
contain more "cushion" to allow for the increased
uncertainty.
REFERENCES
1.
"A Legislative History of the Clean Air Amend-
ments of 1970," Serial No. 93-18 (GPO, January
1974). pp. 227, 260, 402, and 816.
Portland Cement Association vs. Ruckelshaus,
Medusa Portland Cement Co., et aI., 486 S.2d 375,
July 29, 1973.
Essex Chemical Corporation, et aI., vs. Ruckelshaus,
Appalachian Power Company, et aI., vs. EPA, 486
F .2d 427, September 10, 1973.
TRC, The Research Corporation of New England,
"I mpact of New Source Performance Standards on
1985 National Emissions from Stationary Sources."
Volume I, EPA Contract 68-02-1382, Task No.3,
October 24, 1975. t
2.
3.
4.
tThe procedures used by TRC were developed by Mr.
George W. Walsh, Assistant to the Director, Emission Standards
and Engineering Division, U.S. Environmental Protection
Agency, Research Triangle Park, N.C.
31
-------
5.
U.S. Environmental Protection Agency, Office of
Air and Waste Management, "Air Quality Impact of
Alternative Emission Standards for Light Vehicles,"
p. 1, March 4,1975, Revised March 12, 1975.
32
6.
U.S. Environmental Protection Agency, Office of
Air and Waste Management, Office of Air Quality
Planning and Standards, "Control of Photochemical
Oxidants-Technical Basis and Implications of
Recent Findings," EPA-450/2-75-005, July 1975.
-------
SUMMARY OF MULTIMEDIA STANDARDS
John G. Cleland*
Abstract
While sulfur control in the atmosphere is presently
of major concern to synthetic fuel plants, other pollut-
ants such as those affecting water quality, solid waste,
and other substances emitted into the air (especially
organic carcinogens and toxic trace metals) must also be
controlled. This discussion summarizes existing State
and Federal multimedia regulations, the rationale and
processes for the developing standards, and possible legal
trends applying to the synthetic fuel industry. EPA-RTP
has recently assumed responsibility for assessment and
control of effluents in all media for this industry. In the
absence of specific standards for fuel conversion, appli-
cation and extrapolation of standards for other indus-
tries are examined.
At this time, a discussion of environmental regula-
tions entitled "Multi-Media Standards" is a misrepresent-
ation of facts. While pollutants in all media-air, water,
land-have been addressed by standards, there is no lone
standard which limits any pollutant in every medium.
The regulations fostered by the former Atomic Energy
Commission (AEC) pertaining simultaneously to radio-
nuclides in both air and water may be closest to a multi-
media standard.
The question of whether such inclusive standards
should be formulated has been raised at various times.
Without resort to technical arguments, it can be simply
stated that many complications make the probability of
multimedia standards unlikely. For instance,
1. Addressing all media effluents from a particular
industry under a single standard is possible, but regula-
tions and methods for their implementation would have
to be restructured;
2. Controlling any pollutant in all media by a
single regulation is virtually impossible because of dis-
similar monitoring and analysis, different toxic effects
and synergistics, and different control technology for
substances in varying streams.
There has been a great deal of difficulty encountered in
attempting to correlate standards applied to the same
medium.
A recent report (ref. 1) analytically examined the
problem of cross-media interaction, e.g., a cooling
*The author is with the Energy and Environmental Research
Division of the Research Triangle I nstitute, Research Triangle
Park, North Carolina.
system for controlling thermal water pollution may
adversely affect local air quality. Table 1 illustrates an
included comparison of the importance or consideration
which should be given to each medium. The expert
opinion column was derived through a survey of EPA
personnel. Those opinions were coupled with selected
regional characteristics and ultimate weights assigned by
a statistical computer model.
The second column was obtained by further adjust-
ing by total damage or treatment costs so called "social
costs," i.e., dollars. I n the final analysis, "the proof is in
the pudding."
Table 1. Estimates of media weights
Expert opinion Social costs
Air 290 540
Water 390 290
Land 320 170
1000 1000
Using another yardstick, EPA says it took four
times as many enforcement actions in its second 2 years,
1972-1974, as in its first 2: about 8 percent of the
actions were for air emissions from stationary sources; 9
percent for stationary source water effluents; 20 percent
for oil spills; and 60 percent for pesticides. Solid waste
disposal is not enforced. Ford Motor Company picked
up $7 million of the total $8.7 million tab in fines
levied.
Some justification for concern about the environ-
mental impact of synthetic fuel plants is seen in table 2.
The tonnages are a very rough projection (ref. 2) for a
total of 100 plants. By the year 2000, enough waste
could have accumulated to cover all of Rhode Island a
foot deep in ash and sulfur.
Table 3 shows reasonably comprehensive standards
which may be of interest to coal conversion planners.
While certainly new source performance standards for
fossil fuel conversion facilities are forthcoming (ref. 3),
other regulations are only in embryonic stages.
State implementation plans are those designed to
meet EPA Air Quality standards under the Clean Air
33
-------
Table 2. EPA estimate of potential pollutants from synthetic fuel
Pollutant Tons/yrl Accumulated 1980-2000'
Sulfur 36,000,000 327,000,000
Nitrogen (fuel) 14,000,000 130,000,000
Ash 150,000,000 1,330,000,000
Ammonia 35,000,000 318,000,000
Tars, oil 21,000,000 191,000,000
Waste water 683,000,000 6,209,000,000
Chlorine 1,300,000 12,000,000
Vanadium 35,000 318,000
Flourine 6q,000 600,000
Bromine 17 ,000 155,000
Arsenic 15,000 136,000
I Basis: Year 2000, total tons of coal 1.1 X109 tons(yr
'10X109 tons of coal
Table 3. Pollutant regulations
Air
Water
Land
Ambient air quality standards
State and local standards
(Implementation plans)
Receiving water quality standards
State and local standards
State and local standards
National primary standards
National secondary standards
Drinking water standards
Incentives
Solid waste disposal act
Resource recovery act
N ondegradation standards
U.S. Public Health Service
EPA
AEC radioactive disposal
Sou rce em ission standards
Source and local standards
Effluent limitations
State and local standards
MEgA regulations
New source performance
standards
Hazardous pollutants
New source performance
standards
Best practicable technology
currently available
Workroom atmosphere
OSHA
TLV's-ACGIH
Best available technology
economically achievable
Pretreatment standards
Zero discharge
Hazardous substances
AEC radioactive
34
-------
Table 4. Excerpts-applicable national pollutant regulations
Effluent Drinking
Air quality standards water
standards NSPS Workroom Ib/1000 Ib standards
Substance mg/m3 Ib/106 Btu mg/m3 30 day avg. mg/liter
Ammonia 18
CO 10 55
Hydrocarbons 0.16 Nonmethane 360
H,S 15
NOx, nitrates 0.1 (NO.) 0.7 9 10 (nitrate)
Particulates 0.075 0.19 10
Phenols 19 0.2 0.001
SO" sulfates 0.08 (SO.) 1.2 13(50.) 250
Act. Many have been revised from the original national
standard attainment date of July 1975 to later dates,
such as 1977.
The States' Receiving Quality Standards comprise
implementation plans for water. But Federal water qual-
ity regulations are not as quantitative as those for air.
They specify control methods-"Best Practicable Tech-
nology" by 1977 and "Best Economically Achievable
Technology" by 1983. Pretreatment applies to dis-
charges to municipal sewers. Zero discharge criteria for
aqueous effluents are based on the stipulation that dis-
charge cannot contain higher pollutant levels than intake
streams. Some relaxation of "zero" discharge may be
available because of the industrial quality of intake
water.
There are no Federal statutory limitations on solid
waste disposal in the public sector. The incentive legisla-
tion listed provides only for Federal research, demon-
strations, train ing and technical assistance. Most state
standards are concerned primarily with municipal dis-
posal. The Mining Enforcement and Safety Administra-
tion has added more stringent requirements for coal
refuse banks and the disposal of mine spoils, especially
since the Buffalo Creek disaster.
There are Federal standards which are applicable to
synthetic fuel plants at present: The New Source Per-
formance Standards (NSPS) reflected in table 4 are for
coal-fired steam generators. The American Conference of
Governmental Industrial Hygenists (ACGIH) and the
Occupational Safety and Health Administration (OSHA)
publish enforceable regulations in the form of threshold
limit values (TLV's) for substances in industrial air. The
limits are usually for 8-hour exposure periods and along
with the National I nstitute of Occupational Safety and
Health (NIOSH) Toxic Substances List have been used in
various analyses for proper environmental assessment.
OSHA and ACGIH also have established limits for physi-
cal agents such as noise and have prohibited the presence
of determined carcinogens.
There is only one effluent limit shown-for
phenol-but the effluent standards also include such
water parameters as B005, COO, TSS and pH for new
sources. About 38 industries or categories are already
included under the EPA effluent standards, and synthe-
tic fuels processes are a possible addition.
Selected air and water pollutants (table 5) are
commonly addressed by the individual States. These
average standards are for all States having pertinent regu-
lations. Where the average is higher than the correspond-
ing Federal regu lation-such as for SOx and particu.
lates-Iegal exceptions have been made for certain equip-
ment types or locations.
Some State standards are much more stringent than
Federal-for instance, the only State with an emissions
plan for gasification plants, New Mexico, proposes less
than one-third the allowance of particulates given by the
NSPS for coal-fired generators.
It should be remembered that standards which are
not directly applicable to fuel conversion may have the
greatest impact on the industry-such as strip min ing
legislation.
More detailed summaries of then existing environ-
mental standards are given in the combination of several
fine papers (refs. 4,5,6,7) offered at the last Synthetic
35
-------
Table 5. Selected averaged State regulations
Pollutant
Mean standard for states
Particu lates
New fuel burning {10' Btu/hr .19 Ib/106 Btu
Industrial processes (>500 tph) 52.81b/hr
Sulfur oxides
New fuel burning {solid fuell
Sulfur recovery plants
1.251b/106 Btu
.08 Ib/lb S
Nitrogen oxides
Coal-fired combustion
.70 Ib/106 Btu
Hydrocarbons
85-90% reduction of all
emissions over 15 Ib/day
NOz
Ammonia, NH,
41.5 mg/l
2.2 mg/l
.08 mg/l
.37 mg/l
5.1°F
Chromium
Manganese
~ Temperature
(Maximum)
Maximum pH
8.86
Fuels Symposium. One of these, by Rubin and
McMichael (ref. 7), includes a brief analysis of some im-
plications of existing emission and effluent regulations.
Table 6 shows effluent standards adjusted to the
same units for two industries which are often compared
to coal gasification and liquefaction plants. It will be
noted that the limits approximate one another.
More relevantly, table 7 includes effluents for the
same refineries and coking plants and also for coal con-
version plants in a six-State region. Effluents are based
on estimates of 22 percent of U.S. crude oil flow and 63
percent of U.S. coke produced in this region. Applying
the effluent regulations for coke plants to coal conver-
sion plants, it was found that six 20,000 Ib coal/day
conversion plants would put out a comparable weight of
effluents.
While this sort of extrapolation may be helpful to
designers of U.S. synthetic fuel complexes, it is empha-
sized that planners should attempt to keep abreast of
new proposals and possible legislation in the environ-
mental field. There have been no environmental regula-
tory actions taken in the year and a half since the last
fuel conversion symposium which might significantly
change the course of the industry. Table 8, however,
does list a few events which could have effect.
EPA recently denied a relaxation of its S02 stand-
ard for those with the problem of sludge from lime
scrubbers in fossil fuel-fired steam generation plants.
However, a change in the particulate opacity measure-
ment methods was allowed.
The House Interstate and Foreign Commerce Sub-
committee submitted on September 5, 1975, its draft on
prevention of "significant deterioration" in clean air
{nondegradation} areas. These proposed rules are tough-
er than those discussed at the May 1974 symposium in
St. Louis and have been extended from just S02 and
particulates to all air pollutants covered by the Ambient
Air Quality Standards, i.e., CO, oxidants, hydrocarbons
and N02. For synthetic fuel plants contemplating loca-
ting in areas of the country where clean air quality
Table 6. Adjusted new source performance standards
(pounds of pollutant per 1012 Btu feedstock, 30-day manixum)*
Pollutant Petroleum refineries By-product coke making
BOD, 230-1015 477
TSS 143-646 242
Ammonia (as N) 46-400 242
Oil and grease 71-233 12
Phenols 1.5-7.1 5.8
Sulfide 1.2-5.8
* Assumes heating values of 6.5 MM Btu/bbl for crude oil and 12,000 Btu lib for coal, with a
coke yield of 0.69 Ib coke/lb coal.
36
-------
Table 7. Estimated total wastewater emissions of selected pollutants for six major eastern
coal states (pounds per day based on 3D-day average NSPS limits)
Pollutant Refineries* Coke plants** Coal conversion plants***
Phenols 30-138 39 33
Ammonia 900-7,800 1638 1378
8005 4,500-145,000 2910 2756
TSS 2,790-12,600 1638 1378
Oil and grease 1,380-6,300 1638 1378
* Assumes 22 percent of U.S. crude flow processed in the six-state region.
**Assumes 63 percent of U.S.by-product coke processed in the six-state region.
*** Assumes six 20,000 tonlday plants operating NSPS limits for coke plants over the six-state region.
already exceeds standards, the deterioration limits may
present the most formidable constraint.
1974 New Source Performance Standards produced
are:
1. A particulate standard of 0.070 g/dscm for coal
preparation plants;
2. Monitoring methods for all designated new and
modified plants; sources required to install continuous
Table 8. Recent regulatory activities
1975
Implementation plan - primary drinking water standards
No revision - NSPS - fossil fuel generators
House subcommittee draft - degradation
NIOS!:! recommendation - arsenic, .002 mglm'
1974
NSPS - coal preparation plants
NSPS - monitoring and testing
Proposed regulations - hazardous substances
Draft recommendation - effluents and NSPS,
railroad operations
monitoring instruments of proscribed performance and
report data to State agencies quarterly; and
3. Proposed air emission and water effluent limita-
tions for railroad operations such as cleaning, draining
and maintenance.
More than 350 chemical compounds may be desig-
nated as hazardous pollutants and regulations set. Civil
penalties would be imposed for discharges of substances
determined to be "not actually removable." The last
known "optimistic" date for finalization was published
as September 1975.
Finally, EPA program managers in synthetic fuels
recently made a very preliminary assessment of pertinent
effluents (table 9) which are likely to be controlled by
standards. No timetable is proffered. Preliminary Stand-
ards of Practice Manuals with a minimum of new data
may be issued by EPA within the next year. However,
standards setting-especially for water and land effluents
and for nonambient air quality emissions-is certainly
lagging technology. A reliable data base is required. Pro-
vision of this data base to environmental agencies by
industry and researchers in the field and a reciprocal
inclusion of industry in the standards setting process is
obviously the most expedient and satisfactory method
for insuring environmentally sound fuel conversion tech-
nology.
37
-------
Table 9. Prediction of new EPA standards for
synthetic fuels
Effluents
standards likely
Control technology
Available R & 0 required
Airborne
Fugitive dust Yes No
Sulfur, oxidized Yes No
Sulfur, reduced Yes Yes
Nitrogen, oxidized No Yes
Nitrogen, reduced Yes Yes
Carbon monoxide Yes No
Hydrocarbons, storage Yes No
Particu late ash Yes No
Noise Yes Yes
Aqueous
Storage pile runoff Yes Yes
Sour water/preparation Yes No
Sou r water /process Yes Yes
Siu rry transportation Yes No
Thermal Yes Yes
Solid & sludges
Water treatment sludge Yes Yes
REFERENCES
1. "Development of Cross-Media Evaluation Method-
ology," prepared for Environmental Quality Council
by Battelle Columbus Laboratories, NTIS No.
PB-232414, 15 January 1974.
2. Project Independence, FEA Final Task Force Re-
port, FE 1.18:G29, November 1974.
3. James F. Durham, EPA presentation on "Federal
Regulation of Atmospheric Emissions for Advanced
Fossil Fuel Conversion Facilities," Advanced Fossil
Fuel Sector Group Meeting, EPA-RTP, 13 Novem-
ber 1975.
4. Jack R. Farmer, "Environmental Quality and Stand-
ards for Air," Symposium Proceedings: Environ-
mental Aspects of Fuel Conversion Technology,
EPA-650/2-74-118, May 1974.
5. Kenneth M. Mackenthun, "Environmental Quality
and Standards for Water," Symposium Proceedings:
Environmental Aspects of Fuel Conversion Tech-
nology, EPA-650/2-74-118, May 1974.
6. A. Blakeman Early, "Coal Conversion Technology
and Solid Waste Disposal: Time to Take Stock,"
Symposium Proceedings: Environmental Aspects of
Fuel Conversion Technology, EPA-650/2-74-118,
May 1974.
7. E. S. Rubin and F. C. McM ichael, "Some I mplica-
tions of Environmental Regulatory Activities on
Coal Conversion Processes," Symposium Proceed-
ings: Environmental Aspects of Fuel Conversion
Technology, EPA-650/2-74-118, May 1974.
38
-------
AN INVESTIGATION OF TRACE ELEMENTS IN COAL
H. J. Gluskoter, R. A. Cahil, W. G. Miller,
R. R. Rueh, and N. F. Shimp*
Abstract
A variety of coal samples is currently being exten-
sively analyzed for constituents, including many trace
elements, at the Illinois State Geological Survey. The
samples include whole coals, washed coals, and bench
samples. Among the many determinations made on each
sample are analyses for approximately 60 elements,
almost twice the number of elements previously deter-
mined. The increase is in part the result of the addition
of instrumental neutron activation analysis (lNAA)
equipment to the laboratory.
Twenty-five samples of Herrin (No.6) Coal that had
been analyzed previously were subjected to INAA analy-
sis and were found to include Ba, Ce, Cs, Dy, Eu, Au,
Hf, I, In, La, Lu, Rb, Sm, Se, Ag, Sr, Ta, Tb, Th, W, V,
and Yb, none of which were reported by previous tech-
niques. These elements generally are present in very
small amounts and, with the exception of barium,
exhibit no wide range in concentration. The rare earth
elements are among those having the narrowest ranges.
Wide variations in element content have been ob-
served in bench sets of coals (samples of vertical seg-
ments of the coal seam). Many elements, notably germa-
nium, are concentrated at the top and/or bottom of the
seam, the high concentrations of Ge being found there in
all four bench sets analyzed to date.
INTRODUCTION
Research concerning the modes of occurrence and
distribution of trace elements in coals has been intensive-
ly conducted at the Illinois State Geological Survey for
the past six years and has resulted in the publication of a
number of reports. I neluded among the more substantive
of these are three that were published in the ! lIinois
State Geological Survey's Environmental Geology Notes
series (refs. 1-3). The most recent of these EGN's was
also distributed by the U.S. Environmental Protection
Agency.
In this short report, an attempt is made to update
this continuing research, summarize new or significantly
modified analytical techniques, and present some very
recent-although still preliminary, data on the distribu-
tion of elements that we have not previously determined
*The authors are with the Illinois State Geological Survey,
Urbana,llIinois 61801.
in coals and on the vertical distribution of trace elements
throughout the coal seam. A more complete discussion
of these data must await the completion of the chemical
analyses of the current group of samples, but we hope it
will be published in a series of papers within the next
year.
Obtaining data concerning the distribution and
mode of occurrence of trace elements in coals is an ob.
vious first step in assessing the environmental aspects of
fuel conversion, the subject of this symposium. It has
not, however, always been among the first steps taken.
Not only are trace element data essential for determining
potential environmental effects of coal conversion proc-
esses, but they are also needed for assessing which mate-
rials recoverable from coal conversion plants have pos-
sible economic importance, and for predicting the poten-
tial effects, both positive and negative, these elements
may have on catalytic reactions involved in the conver-
sion processes.
Research reported here and in the publications pre-
viously mentioned (refs. 1-3) was supported in part by
Contracts 68-02-0246 and 68-02-1472, and Grant
R800059 from the U.S. Environmental Protection
Agency, Energy Assessment and Control Division, Fuel
Process Branch, Research Triangle Park, North Carolina.
CURRENT INVESTIGATIONS OF
TRACE ELEMENTS IN COAL
Previous investigations of trace elements in coal at
the Illinois State Geological Survey have been reported
in several publications within the past year and a half.
Those reports presented data on chemical analyses of
101 whole coal samples and 32 samples of washed coals
(separated by specific gravity in the laboratory) and
summarized the geochemical significance of those data
(refs. 2-4). Thirty-three elements were determined for
each sample, in addition to normal coal parameters, in
the original investigations. Our current investigations are
expanding the previous study by the analyses of addi-
tional whole coal samples, washed coal samples, and
bench samples, and also by determining many more ele-
ments in a series of 25 samples of Herrin (No.6) Coal
from Illinois. The latter coal samples had previously
been analyzed for the original 33 elements.
Seventy additional whole coal samples are being
analyzed. Of these, 31 are from the Illinois Basin (Illi-
nois, southwestern Indiana, and western Kentucky), 14
39
-------
are from eastern states, 22 from western states, and 3 are
from Iowa. These samples are all face-channel samples or
composite face-channel samples made by combining
individual face-channel samples. Most of them were col-
lected by Illinois State Geological Survey personnel.
Several sets of washed coals also are being analyzed.
I n the laboratory they have been separated into fractions
based on specific gravities, and each fraction has been
chemically analyzed. At least five sets, comprising more
than 35 total samples, are to be tested: three coals from
Appalachian coal fields (Alabama and northern and
southern West Virginia), one from Illinois, and one from
Arizona.
Other coal samples have been collected by sampling
the coal seam in vertical segments or "benches." Four
sets, consisting of a total of 33 bench samples, have been
collected from the Herrin (No.6) Coal in Illinois. The
sets are from sites that have been judged to be signifi-
cantly different from each other with respect to their
geological setting in the Illinois Basin.
All the coal samples have now been analyzed for
approximately 60 major, minor, and trace elements, in
addition to carbon, hydrogen, oxygen, and nitrogen,
which are included in the standard "ultimate" analysis
of coa I.
Determination of such a large number of elements
in the current series of analyses is now possible because
within the past few years, we have significantly enlarged
our capabilities for determining trace and minor ele-
ments in coal. Table 1 shows the procedures now avail-
able and the elements they detect. The major addition
ha s bee n instrumental neutron activation analysis
(lNAA). in which a high-resolution Ge(Li) detector is
used. The technique can provide accurate, precise values
for some 38 elements. The method uses whole coal as
the sample without ashing or dissolution and thus avoids
volatilization losses. In the procedure, the whole coal
samples are irradiated in the University of Illinois
TRiGA MK II nuclear reactor facility, counted on the
10 percent Ge(Li) detector coupled to a 4096 multi-
channel analyzer, and compared to a multielement com-
posite standard, which has been treated identically. To
exploit adequately all the nuclear properties and param-
eters available for neutron activation, a combination of,
P rocedu re
Table 1. Analytical procedures used in coal analysis
Element
Type of sample
Instrumental neutron
activation analysis
(INAA)
Whole coal
Neutron Activation
Analysis-Radioch em ical
separation
(NAA(RC))
Na., K., Rb, Cs, Sr., Ba, Ga,
In, As, Sb, Se, CI", Br~, t,
Se, V., Cr*, Mn*, Co., Feil,
Ni., Zn~, Mo., Ag., Hf, Ta, W,
Au, La, Ce, Sm, Eu, Tb, Dy,
Yb,Lu,Th,U
Hg
Tet
Whole coal
Low-temperature ash
(150°C)
X-ray flourescence-
wave-Jength dispersion
(XRF)
Whole coal
Atomic absorption
(AA)
Na., K., Mg, Ca, AI, Si, S, P,
CI., Br., Ti, Ni., Zn., Fe.
Low-temperature ash
(150°C)
Ni" Cu* Zn. Cd. Pb. Tit
t' , ~ I , , ,
Li
Optical emission-
direct reader (OED)
Optical emission-
photographic (OEP)
High-temperature ash
(500°C)
High-temperature ash
(500°C)
Be, Sr., B, Ge., V., Cr., Co.,
Cd., Ni., Cu., Zn*, Zr., Mo.
Ge*, Pb', V., Cr., Co., Ni",
Cu., Mn., Zn., Zr., Mo., Ag.,
snt
lon-selective electrode (lSE)
Whole coal
F
"Elements for which two or more analytical procedures have been applied.
tElements for which further methods of analysis are being developed.
40
-------
different irradiation and counting periods is used. This
method is much superior to the previously used instru-
mental approach that employed the low-resolution
Nal(TI) detector. INAA also essentially eliminates some
previously required, tedious radiochemical separations
for Sb, Se, As, and Ga. Radiochemical separations-
NAA(RC)-are still required, however, to get adequate
analysis for Hg.
X-ray fluorescence wave-length dispersive (XR F),
atomic absorption (AA), optical emission-direct reading
(OED), and ion-selective electrode (lSE) techniques all
have been upgraded and refined during the present pro-
ject.
An energy-dispersive X-ray fluorescence technique,
which employs a high-resolution Si(Li) detector, isotopic
source e41 Am), and secondary metal foils for sample
excitation, is currently being developed at ISGS for
future used in determining up to 20 elements in whole
coal samples.
RECENT DATA
Two groups of data will be included in the brief
discussion to follow. The first group concerns 22 addi-
tional elements determined on samples previously an-
alyzed, and the second group concerns the trace-element
d i str ibution in bench samples. All of the values
presented here should be considered as preliminary and,
although we are confident that most of them will not be
changed prior to publication of the final values, a few
changes may be made. This disclaimer is necessary so
that we might present data obtained so recently that we
have not yet had time to make all the usual tests for
accu racy, both analytical and clerical.
Additional Elements Determined in Coals
Ruch et al. (ref. 3) reported in 1974 the analyses of
101 whole coal samples. Trace elements determined on
each of these samples were antimony, arsenic, beryllium,
boron, bromine, cadmium, chromium, cobalt, copper,
fluorine, gallium, germanium, lead, manganese, mercury,
molybdenum, nickel, phosphorus, selenium, tin, vanadi-
um, zinc, and zirconium. In addition, the following
major and minor elements were also determined:
aluminum, calcium, chlorine, iron, magnesium, potas-
sium, silicon, sodium, sulfur, and titanium. Standard
coal analyses-i.e., proximate, ultimate, heating value,
varieties of sulfur, and ash also were reported.
Twenty-five of the 101 coal samples, from the
Herrin (No.6) Coal in the Illinois Basin, have been an-
alyzed further, and the following trace elements have
been determined: barium, cerium, cesium, dysprosium,
europium, gold, hafnium, iodine, indium, lanthanum,
lutecium, rubidium, samarium, scandium, silver,
strontium, tantalum, terbium, thorium, tungsten,
uranium, and ytterbium. These elements were deter-
mined primarily by instrumental neutron activation an-
alyses, although corroborative analyses made by other
methods were obtained for some elements (table 1).
The data from the analyses of the 25 coals are
summarized in table 2. Arithmetic mean, standard devi-
ation, and range of elemental concentrations (minimum
and maximum) are given for each element. For all these
elements, the standard deviations from the mean are less
than or equal to the arithmetic mean, with the exception
of those for barium and silver. This is another way of
stating that the range in values for these elements is
rather narrow. Among those elements with narrower
ranges in values are rubidium, cesium, hafnium, tanta-
lum, and the lanthanides (La, Ce, Eu, Tb, Dy, Vb, and
Lu). In previous reports (refs. 3 and 4), we have noted
that those elements that occur in discrete mineral phases
in the mineral matter of the coal are those that exhibit
the widest ranges in concentration of the elements and
the largest standard deviations relative to their arith-
metic means. The mineral barite (8aS04) has been
observed in coals from several localities in Illinois, which
is a reasonable explanation for some very high barium
concentrations in coals from the Illinois Basin. Relative-
ly wide ranges also were observed for the chalcophile
elements, those elements that commonly occur in nature
as sulfides. Silver is a chalcophile element and may also
be found in very limited amounts in solid solution in
pyrite (FeS2) and/or in marcasite (FeS2).
Analyses of Bench Samples
In an attempt to determine the vertical variation in
trace elements within coal seams, we have collected, and
are still collecting, sets of samples called bench samples.
The vertical segments of the coal seam are the benches,
and from four to eight benches make up the sample set.
Normally, the rock unit immediately overlying the coal
also is sampled, as is the underclay (or other seat rock)
and any rock parting within the seam over three-eighths
of an inch thick. Each bench of coal is analyzed for the
full range of chemical elements, and several of the associ-
ated rock units also have been extensively analyzed
chemically. The data mentioned here are from the an-
alyses of four such bench sets, and all the analyses were
performed in the laboratories of the illinois State Geo-
logical Survey (lSGS). With the cooperation of the
United States Geological Survey (USGS), this work is
being extended in scope and, in the future additional
bench analyses will be made at both the ISGS and the
USGS.
The four sets of bench samples recently analyzed at
the ISGS are all from the Herrin (No.6) Coal in Illinois.
41
-------
Table 2. Summary of additional analytical values for 25 samples of Herrin
(no. 6) coal from the Illinois basin (ppm)
Arithmetic Standard Range of values
Element mean deviation Minimum Maximum
Ag 0.19 0.24 < 0.03 0.80
Au 0.Q1 0.01 < 0.0004 0.032
Sa 130. 150. 33. 750.
Ce 11. 4.3 4.4 24.
Cs 1.0 0.26 0.49 1.5
Dy 0.99 0.30 0.70 1.81
Eu 0.23 0.07 0.10 0.40
Hf 0.45 0.14 0.24 0.81
I 2.0 1.2 < 1.0 5.8
In 0.02 0.02 < 0.008 0.09
La 6.9 2.2 3.3 12.
Lu 0.07 0.02 0.041 0.11
Rb 14. 4.0 7.4 20.
Sc 2.4 0.57 1.4 3.6
Sm 1.1 0.62 0.4 3.8
Sr 37. 20. 1"9, 130.
Ta 0.15 0.05 0.10 0.30
Tb 0.15 0.06 0.04 0.24
Th 2.01 0.47 1.2 3.3
U 1.6 1.1 0.5 4.5
W 0.74 0.56 0.04 2.1
Yb 0.51 0.13 0.31 0.77
The sample sites were selected to provide a range of
geological and geochemical characteristics of the coals.
Samples were taken from areas of high-sulfur coal and
from areas of low-sulfur coal, from underground mines
and from strip mines, and from areas with marine roof
rocks and areas with nonmarine roof rocks. The four
samples are separated geographically by as much as 70
miles, and no two sites are closer than 25 miles. I n this
discussion, the four sites will be identified by the letters
J, M, P, and R.
As coal is a relatively heterogeneous material in all
respects, wide variations in trace-element contents of
individual benches are to be expected and, in general,
that is what we are finding. However, we also find that
in some bench sets, some elements occur uniformly
throughout the seam. Perhaps the most uniform distribu-
tion observed thus far is that of bromine in sample set J
(fig. 1). Figures 1-5 represent the benches of the total
coal seam, with the proportional thickness of each bench
plotted along the ordinate and the concentration of the
element along the abscissa. The top of the coal seam, or
the rock above the seam, is plotted at the top of each
figure.
o
20
40
"'
'"
.c
u
~
-------
.j::>.
tv
v;-
a>
.<::
u
'"
~ 40
Go>
t::
"""
U
:<::
I-
60
20
v;-
a>
.<::
u
.=
~ 40
a>
<::
.....
.~
.<::
I-
o
20
o
60
80
o
20
40
Vanadium (parts per million)
o
<
174
'"
a>
.<::
u
<::
~ 40
a>
<::
...
.~
.<::
I-
80
o
10
20
Molybdenum (parts per million)
30
40
60
80
100
20
60
80
o
2
3
Uranium (parts per million)
Figure 2. Distribution of vanadium, molybdenum, and uranium
in coals of bench set R.
<==J 28
4
-------
o
o
20 20
II) M R
Q)
.c II)
c.J Q)
c: .c
c.J 40
"; 40 c:
II)
Q) II)
c: II)
..00: Q)
c.J c:
.c ..00:
I- .~
.c 60
60 I-
- 40
II)
Q)
.c
c.J
c:
~ 60
Q)
c:
..00:
c.J
:c
I-
80
0.0
0.2 0.4 0.6
Antimony (parts per million)
o
--.
-
- J
-
-
-
r I T
20
80
100
120
o
4 8 12
Antimony (parts per million)
o
20
-
II)
Q)
.c
c.J
c:
"; 40
II)
Q)
c:
..00:
.~
.c
I-
60
80
80
0.0
0.6 1.2 1.8
Antimony (parts per million)
2.4
~6.6
p
o
123
Antimony (parts per million)
D Coal
~ Claystone
Figure 3. Distribution of antimony in coals of bench sets
M, R, J, and P.
44
-------
o
o
20 20
en en
Q) Q)
.J:: .J::
(,.) M (,.) R
.!: c
';; 40 ';; 40
III III
Q) Q)
c c
~ ~
(,.) (,.)
.J:: .J::
I- 60 1-60
en
Q)
.J::
(,.)
C
~60
Q)
c
~
(,.)
.J::
I-
80
80
o
6 12 18 24
Germanium (parts per million)
30
o
I
2
en
Q)
.J::
(,.)
C
- ~4
III
III
Q)
C
~
.~
J .J::
- 1-6
- 8
I I I I
20
40
100
120
o
4 8 12
Germanium (parts per million)
16
80
o
5 10 15
Germanium (parts per million)
20
o
o
o
p
o
o
o
2 4 6
Germanium (parts per million)
Dcoa'
~ Claystone
8
Figure 4. Distribution of germanium in coals of bench sets
M, R, J, and P.
45
-------
o
20
D Coal
~ Claystone
"'
..
~ 40
"'
"'
~
~ 60
80
o
0.5
1.0 1.5
Terbium (parts per million)
2.5
2.0
Figure 5.
Concentration of terbium in coals
and associated strata of bench
set P.
The expected variability in trace-element distribu-
tion is apparent in figure 2. The three elements U, Mo,
and V have a wide distribution range and in figure 2 are
all concentrated in the uppermost bench of this sample
set. Although we have not yet had the opportunity to
determine statistically from the analytical data which
part of the coal seam is most likely to contain concentra-
tions of the trace elements, the top and/or bottom
benches appear to be the most likely sites. The concen-
tration of antimony in the uppermost bench of each of
the four sample sets is represented in figure 3. The maxi-
mum concentration within the coal seam is in the top
bench of each sample set, but even higher amounts of
antimony were obtained from the rock units associated
with the coals.
Distribution of germanium in the bench sets is
shown in figure 4. The pattern is distinct and consistent
in the four coals. The germanium content of the top
bench and that of the bottom bench are both greater
than the germanium content of any other bench in all
four sample sets. Our earlier efforts, as well as those of
Zubovic (ref. 5), demonstrated that germanium is pri-
marily associated with the organic fraction of the coals
in Illinois and not in the mineral matter fraction (refs.
3-5). This and the observation that the germanium is
concentrated at the boundaries of the coal seam-the top
and the bottom-suggest that the germanium was intro-
duced into the coal seam after burial and is not related
to conditions in the swamps in which the coal formed.
The germanium was transported into the coal seam in
solution and assimilated by the coal when geochemical
conditions were favorable for the removal of the german-
ium from the solutions. The horizontal boundaries (top
and bottom) of the seam were necessarily in contact
with these solutions before the innermost parts of the
seam.
The roof shale, underclay, and a clay parting (blue
band) were analyzed along with the seven benches of
coal in bench set P. Many elements, including Ba, Cs, F,
Ga, Hf, In, La, Sc, Sr, Zr, and most of the rare earths
that were determined, occur in higher concentrations in
these rock units than in the coals. An example of the
concentration of the elements in the strata associated
with the coal is given in the illustration of terbium con-
centration (fig. 5) in bench set P.
All the data from the many analyses mentioned here
will soon be published, the data analyzed statistically in
detail, and the stratigraphic and geographic distribution
of the elements mapped. These investigations should
make available much information that will assist those
who need to make informed decisions about the use of
the energy resources of this country and about the envi-
ronmental effects that may result.
REFERENCES
1.
R. R. Ruch, H. J. Gluskoter, and E. Joyce Kennedy,
"Mercury Content of Illinois Coals," Illinois State
Geological Survey Environmental Geology Note 43,
15 p., February 1971.
R. R. Ruch, H. J. Gluskoter, and N. F. Shimp,
"Occurrence and Distribution of Potentially Vola-
tile Trace Elements in Coal: An Interim Report,"
III inois State Geological Survey Environmental
Geology Note 61, April 1973.
R. R. Ruch, H. J. Gluskoter, and N. F. Shimp,
"Occurrence and Distribution of Potentially Vola-
tile Trace Elements in Coal: A Final Report,"
III inois State Geological Survey Environmental
Geology Note 72, August 1974. Also distributed by
U.S. Environmental Protection Agency publication
650/2-74-054, July 1974, NTISTB 2308091/AS.
H. J. Gluskoter, "Mineral Matter and Trace Ele-
ments in Coal," in Babu, ed., Trace Elements in
Fuel, Advances in Chemistry Series 141, American
Chemical Society, 1975, pp. 1-22.
Peter Zubovic, "Physicochemical Properties of
Certain Minor Elements as Controlling Factors in
Their Distribution in Coal," in Coal Science, Ad-
vances in Chemistry Series 55, American Chemistry
Society, 1966, pp. 221-230.
2.
3.
4.
5.
46
-------
GEOCHEMISTRY OF TRACE ELEMENTS IN COAL
Peter Zubovic*
Abstract
The accumulation and distribution of inorganic
elements in coal result from the interactions of geologi-
cal, physico-chemical, and biochemical factors. Geologi-
cal factors such as (1) rate of subsidence of a deposition-
al basin; (2) rate of uplift of the drainage area; (3) types
of rocks being eroded in the source area and deposited in
a depositional basin; and (4) ratio of the size of drainage
area to the area of organic accumulation, all determine
the availability of the elements to the site of coal depo-
sition.
Physico-chemical and biochemical factors control
the accumulation and distribution of the elements at the
site of organic deposition. Elements such as Ge, Be, and
Ga have a high organic affinity, and are generally found
in higher concentrations in coals deposited adjacent to
source areas, whereas the transition metals are nearly
uniformly distributed throughout a basin. Metals such as
Zn, Cd, and Pb, when present in relatively high concen-
trations, have accumulated by post-depositional pro-
cesses. The occasional high concentrations of these
metals in coals appear to be related to known ore de-
posits. Some of the more toxic elements such as As, Hg,
Se, and Sb do not appear to have a discernible distribu-
tion pattern in coals within a basin.
Generally, coals of the western states contain sig-
nificantly lower concentrations of environmentally ob-
jectionable elements than do coals of the eastern or in-
terior provinces.
INTRODUCTION
The introduction and retention of trace elements in
coal results from a series of geologic and chemical pro-
cesses. The end result of these processes is that many
trace elements that are in low concentration and in a
dispersed state in the parent rock are concentrated in
coal. Further concentration and activation takes place
when coals are utilized. Thus, trace elements eroded and
transported from source areas of hundreds of thousands
of square miles are entrapped in a much smaller volume
of coal, which is then utilized at a point location.
Organisms adapted to relatively low availability of trace
elements in the general terrestrial environment may not
be adapted to the environment produced when coal is
utilized.
*The author is with the U. S. Geological Survey, Reston,
Virginia 22092.
This paper is concerned primarily with the processes
involved in the emplacement of the trace elements in
coal. Because the coals of the Interior Province have
been studied in greatest detail, most of the following
discussion will focus on the geochemistry of the trace
elements in these coals (fig. 1) (ref. 1). This discussion is,
however, applicable to all coals.
GEOLOGICAL PROCESSES
The availability of any particular element to
chemical emplacement processes in the coal-forming
swamp is regulated by several geological processes in the
source area. Rapid and extensive uplift of the source
areas would tend to reduce the length of time necessary
for chemical weathering and would result in a greater
influx of clastic materials and a smaller amount of
soluble inorganic matter into the depositional basin.
Under these conditions, a larger amount of the trace
elements would be associated with the nonauthigenic
mineral matter of coal and a lesser amount with the
organic fraction.
Tectonic and climatic activity in the depositional
basin must maintain conditions for the accumulation of
organic matter. If subsidence is too rapid, little or no
vegetation will be able to grow and clastic materials will
dominate the depositional sequence. If little or no subsi-
dence takes place, the biochemical cycle will be essenti-
ally completed, and most of the vegetative organic
matter will be oxidized. For organic matter to accumu-
late in significant quantities, an optimum balance of
tectonic and climatic conditions must exist.
The rate of accumulation of organic matter is im-
portant in determining the concentration of trace ele-
ments in the resulting coal. If the availability of trace
elements from a source area remains constant, then the
concentrations that will result in the organic matter will
be inversely proportional to the rate of accumulation of
the organic matter.
In addition, if the areal extent of organic deposition
is large, then low concentrations of trace elements will
result. This is why generally thick widespread coal beds
have low trace-element concentrations and, conversely,
why thin-bedded coals have high concentrations.
During the initial phase of organic accumulation in a
basin, the trace elements may be obtained by the plants
from the underlying soil profile. When the peat accumu-
lates to a point where the root systems are within the
peat, then the dominant source becomes the elemental
47
-------
42°
101$"
,
\ I
, I
NEBRASKA .\--t--u_-----
,
,
.
38>
\,
KANSAS i
I
t .
'&'.0 I :
r--I'---.--~
I .
~of.- / OKLIt,.:OMA
L' ~_/~;:;. :.
, ,1:10 ~ .
..
.
92°
90°
~"
MISSOURi
ARKANSAS
Figure 1. Distribution of coal samples in the Interior
Province. Solid line indicates extent of Pennsylvanian
sedimentary rocks containing coal beds; arrows indicate
transport directions of sediments that become basal
Pennsylvanian sandstones. Samples east of dashed line
in Eastern region not used in plots (ref. 1).
influx from the surrounding borderland. The mineral
matter that makes up the soil must probably also had its
origin in the same area.
Zubovic et al. showed (ref. 2) that the trace-element
content of coal can be related to the type of rocks being
eroded. Potter and Glass reported (ref. 3) that during
Pennsylvanian time, rocks of the source area contribut-
ing sediments to the Illinois basin progressively changed
from recycled sediments to more immature types, in-
cluding some contributions from metamorphic and
igneous rocks. Igneous rocks, particularly the mafic
types, are the principal original source of the dark min-
erals that contain most of the V, Cr, Co, Ni, and Cu in
these rocks. The dark minerals (amphiboles, pyroxenes,
sulfides) are less resistant to weathering and to abrasion
during transport than are the light-colored minerals, such
as quartz and feldspar. As a result, the metals that make
up these dark minerals have a higher seaward-migration
rate than those found in the more resistant minerals.
Consequently, most recycled sediments would be de-
pleted in metals such as mentioned above. As a result of
these geologic conditions, progressively larger amounts
of these metals should be available to the younger
Pennsylvanian coals. According to the data, this appears
to be the case.
In some coals of the Interior Province, high concen-
trations of zinc (>500 ppm) are frequently found. When
zinc is present in high enough concentrations, the
mineral sphalerite (ZnS) has been identified (refs. 4 and
5) as fracture fillings in the coal. It appears that all the
Cd found in coal is associated with the sphalerite (ref.
5). As the sphalerite is deposited in fractures, most of it
is removed from the coal when the coal is crushed and
washed. Obviously, the cadmium associated with the
zinc in sphalerite is also removed. Highest concentrations
of zinc are found in areas of the coal basin nearest
known areas of zinc mineralization such as the upper
Mississippi Valley District of Illinois and Wisconsin, the
Tristate District, and the Southern Illinois District. Coals
in the intermediate areas have far lower zinc contents.
48
-------
At present, the reason for this difference has not been
definitely establ ished.
Heyl et al. (ref. 6) suggested that mineralization of
the Wisconsin and northern Illinois lead-zinc ores took
place between late Paleozoic time and the close of
Mesozoic time. As the coals in these areas were deposit-
ed in late Paleozoic time (Pennsylvanian), the processes
that introduced the lead and zinc into the carbonates
and other sediments of northern Illinois and Wisconsin
may also have introduced these elements into the coals
of the Interior Region.
These minerals were deposited in fractures; there-
fo re, the coal must have undergone considerable
diagenetic transformation past the peat stage to be hard
enough for fractures to have formed. The mineralization
is probably post-Paleozoic. If the zinc had been intro-
duced into the coals during Mesozoic time, it could still
have been obtained from the same source as the ore
deposits north of the coal basin. If, however, the zinc
enrichment of coal was post-Mesozoic, then weathering
and erosion of the existing ore deposits could have been
the source of zinc in the coal.
Physico-Chemical Processes of Trace-Element Emplace-
ment in Coal
Interactions of elements with sulfide ions and organ-
ic ligands.
Trace elements in solution entering an area in which
organic matter is being deposited are subjected to reten-
tion according to the principles of coordination chemis-
try. It can be assumed that a large variety of organic
ligands is present within such an organic environment.
This environment would also be conducive to the growth
of sulfate-reducing bacteria and consequently to the pro-
duction of H2 S, which dissociates into HS- and S=.
There would then be competition between the organic
ligands and sulfide ions for the available metal ions, par-
ticularly the chalcophilic elements such as Cu, Pb, Zn,
Cd, Sb, Hg, and Fe.
The partition of these elements between sulfide and
organic phases is dependent upon the concentration of
organic ligands and of the sulfide ions. Because this is a
dominantly organic environment, the concentration of
organic ligands would remain relatively uniform. The
concentration of sulfide ions could vary considerably, as
their availability is dependent upon the influx of sulfate
ions as well as upon the development of sulfate-reducing
bacterial colonies which would produce H2 S. That the
availability of sulfide ions is erratic is best indicated by
the sporadic occurrence of sulfide in coal. Frequently,
massive pyrite is found adjacent to coal that is free of
this mineral. When the sulfide ions are uniformly avail-
able, then small disseminated crystals of pyrite are found
throughout the coal.
In addition to a discussion of the availability of the
reacting anionic species, consideration must be given to
the stability of the compounds that could be formed.
This stability can be compared by considering the disso-
ciation constant of the compounds. Because however, no
data are available on the dissociation constants of
metallo-organic compounds in a swamp environment,
consideration must be theoretical and based on concen-
trations of the dissociated metal ions. If the concentra-
tion of the dissociated metal ion is lower in the presence
of the sulfide of the metal than it is for any possible
metallo-organic compound it could form, the reaction
would be directed toward the formation of sulfides.
Conversely, if it is higher. then metallo-organic
complexes would be formed. A hypothetical phase dia-
gram of such conditions is shown as fig. 2. No units are
indicated. However, the relations are applicable to any
units which are comparable. An illustration of such com-
petitive reactions is shown by the reaction of copper-
amino acid complexes with H2 S. When H2 S is added to
a solution of the copper complex, CuS is ~recipitated.
Equilibrium constants for the copper amino acid com-
plexes are about 10-1 7, whereas that for CuS is about
10-44. Obviously, if H2S is present, CuS will form. The
fact that copper has some organic association in coal is
an indication of the absence of any sulfate-reducing
activity at the time organically associated copper was
introduced into that particular area of the swamp. The
preceding discussion is applicable to all the chalcophilic
elements.
Determination of Organic Association of Metals in
Coal.
The suggestion that organic complexes of various
metals exist in coal and other organic sediments is based
on indirect evidence. With the exception of V and Ni
porphyrins, no other organic complex has ever been iso-
lated or identified in organic sediments. The suggestions
by many investigators that metallo-organic complexes
exist in coal is based on observations that elements such
as Ge and Be were always found .f} large amounts in the
ash of clean coal and ash of pure coal macerals such as
vitrinite.
The first attempt to quantify the organic association
of metals in coal was made by Horton and Aubrey (ref.
7). Three vitrain samples were separated by sink-float
techniques, and the quantities of each element contained
in the different specific gravity fractions were deter-
mined. In their samples, Ge and V were 100 percent
associated with the organic fraction, whereas Sn was 100
percent associated with the mineral fraction of coal. The
other elements had intermediate associations (table 1).
Zubovic (ref. 81. in a study of 13 coal samples, in
which he used sink-float methods, reported a dominant
49
-------
Table 1. Average organic affinity of some metals
determined by float-sink methods
[N.D., not determined]
Percent organic Percent organic
Element affinity (ref. 8) association (ref. 7)
Germanium 87 100
Beryllium 82 75 - 100
Gallium 79 75 - 100
Titanium 78 75 - 100
Boron 77 75 - 100
Vanadium 76 100
Nickel 59 0 -75
Chromium 55 -0 - 100
Cobalt 53 26 - .50
Yttrium 53 N.D.
Molybdenum 40 50 - 75
Coppsr 34 26 - 50
Tin 27 0
Lanthanum 3 N.D.
Zinc 0 50
organic association for the same six elements reported in
ref. 7 as having a high organic association. I n this study,
zinc showed no organic association (fig. 3). Ruch et al.
(ref. 9) reported on the organic association of 21 ele-
ments in four coal samples. Ge, Be, and B were found to
be organically associated, whereas Hg, Zr, Zn, Cd, As,
Pb, Mo, and Mn were largely associated with the mineral
matter of the coal. The other elements had varying de-
grees of organic association. The Ruch et al. studies were
based on a physical separation of the organic fractions
from the mineral matter of coal and on an investigation
of the distribution of the elements in these components.
The studies indicate that most of the elements have a
partial organic and partial inorganic association. Some of
the inorganic associations, especially for the chalcophilic
elements are caused by the formation of sulfides, as pre-
viously discussed. Some of the inorganic associations
shown by such elements as Ge, Be, B, Zr, and others
result from their introduction into the coal-forming
swamp in resistant clastic minerals.
Additional evidence for the existence of organic
complexes (chelates) for some of the elements in coals is
shown by the relation of the organic affinity of the
elements in coal to the complexing ability of these
metals with organic iigands. The stability of chelates of
the metals is directly related to the ionic potential of
these elements. The organic affinity of the metals in coal
is also related to their ionic potential (fig. 3). This sug-
gests that those metals having high organic affinity in
coal are present as chelates.
Compilations of stability constants for a wide range
of organic ligands (ref. 10) give a stability order of
Ga>Y>La for these trivalent metals. This is the same
order as is their organic affinity (fig. 3). Other compila-
tions for the bivalent metals (refs. 10,11) suggest that
their stability order is Be>Cu>Ni>Co>Zn>Fe. In coal,
the organic affinity for some of these elements is
Be>Ni>Co>Co>Cu>Zn. Iron, of course, is always
dominantly present as the sulfide. With the exception of
copper, the two series are identical. The displacement of
50
-------
"'0
!::
IC
C'1
'r-
r-
U
'r-
!::
IC
C'1
S-
o
4-
o
Q)
U
!::
Q)
II)
Q)
S-
o..
!::
'r-
r--1
+
+
::E
~
Chelate field
a
r-
'0""
r-
o
[M++l argo = 10-10
[M++J S- = 10-5
,/
/
/
/
/
I'
/
./
/
,/
/
./
/
/
/
/
1 0""' 1
./
./
/
./
/
/
/'
./
/
Sulfide field
C)
[M++J argo = 10-5
[M++J S= = 10-10
I
10-10
[M++l in presence af S-
Figure 2. Hypothetical phase diagram showing relation
of the dissociation of divalent chalcophilic metals to
their sulfide and chelate phases.
51
-------
-;)00
Q)
+->
+->
co
:::E:
c.!J 75
a:::
o
>,
+-> ..s:::::
'r- +->
s:: 'r-
4: ::: 50
4- -0
c:( Q)
+->
u co
.~ .,....
~ 2525
01 V1
!- V1
o c:(
+->
s::
Q)
U
!-
Q)
c...
Ge Be Go Ti
B V Ni Cr Co Y Mo Cu Sn La Zn
10
15
r- ,-...
co V1
'r- :::s
+-> 'r-
s:: -0
2 & 5
o ........
c... Q)
01
U !-
'r- co
s:: ..s:::::
o u
........ .........
o
03+
o
3+ ---
o
Cte Be Go Ti B
V Ni Cr Co Y Mo Cu Sn La Zn
Figure 3. Relation of organic affinity and ionic potential
of the elements arranged in order of decreasing affinity
for organ ic matter.
52
-------
copper to a lower organic affinity than is indicated by its
stability series results from its occasional interaction
with sulfide ions, as discussed previously. The evidence
indicates that metals that have high organic affinity in
coal are present as chelates.
Breger et al. (ref. 12) made an intensive study of the
forms of uranium in a North Dakota lignite. By using ion
exchange and acid leaching of the lignite, they deter-
mined that more than 85 percent of the uranium was
bound as an metallo-organic compound and the 1.2 per-
cent was held in ion-exchange positions. The rest was
assumed to be inorganicalry associated.
Trace-Element Distributions in a Coal Basin.
The geographic distribution of the elements in a
coal basin appears to be controlled by their chemical
properties. In the following figures (figs. 4-8), the coal
samples are arranged to portray a cross section of a
basin. Samples nearest to the edges of the diagrams are
closest to the northern margins of the basin and thus to
I
I
31 0 i '
1 0 00 0 i ;
o i
-J I i
°
o
o
o
o
o 0
o
00
00 00 a 0
~.~
o
o 0
.... 0 0
o
o 0
ARKANSAS
IL.LlNOIS
Figure 4. Distribution of germanium and gallium. Tick
marks indicate medians for the elements in the samples.
Areas 1 and 4 are near elemental source area (ref. 1).
53
-------
AREAS
2 3
BERYLLIUM II
7
I 0
I
I
,
o
o 0
o
o
~
u
~
z
o
~
~
~ 2
0::
~
2
00
o
o
o 0
o
o
°00 00°0 000 000
en
~
~
'1 oQ
r 0000:1
~OWA, ~ISSOURI
NOGTH ul-.LAI-OMA
o
o
o
o
o
o 0
~ 0... 010
o 0000 oc.,
SOUTH OKLAHOMA
o
° °
o 00 OQOOO
o
o
o
00 0 0 0
o 0 0 0
o
°
I I
I ° '
1-0 c 00000°0 ~
. "-
':''iKANSAS
II
i
I
. .
.
° . °
°
°
00 0
1--,,- \:1:)
Figure 5. Distribution of beryllium and boron. Tick marks
indicate medians for the elements in the samples. Areas
1 and 4 are near elemental source area (ref. 1).
54
0°
o
r
r
I
t
ooj
o
o
-------
VANADIUM
o
o
o
o
o
o
o
00
000
o
o
o
000 °00
o 0
o
o
o
o
o
o
Ie o~o 0 l: 0 I
0000 0 0 0 0 ~
0°0 00 0 0::° !
---------I
IOWA, MISSCURI, SOUTH OKLAHOMA
fl()RTH OKLAl-WA
AREAS
1 I
o
o
o
o
o
o
o
o
0° 0° 011- 0 0 00
o l' 00 Co
o 0 0
o 0
o 000 --~
o
I
01
I
111
II I
o I Q 0 0 J
ro 0 Q
00 00" 0 ~ I
o 0 f
~------_J
'...L jf"CtS
00
o
o
°000 0 000
'-------
A.QKAN3AS
Figure 6. Distribution of vanadium and chromium. Tick
marks indicate medians for the elements in the samples.
Areas 1 and 4 are near elemental source area (ref. 1).
55
-------
COBALT
o
0 0
-J 0 0
~ 0 0
u
~ 0 0
~
-J
-J
i
cr 7 NICKEl
~
i ~ . 0 I
J.':' ~ }I'
J o.
IOWA. MISSOURI,
t-O!TH ()(LAH)MA
o
ARE>S
3
-
0
0
0
0
0 0
0 0
0
0 0
.0 0 a'
0 0
0 0 0 0
00 00
0 0 00 00 0 0 fJ
00 00 0 0 0
o 0 00 0
0 0
o
o
I I
10 000 0 Q f
I !OO OOOOOO!
IlliNOIS
I
I
I
o 0/ Loo
o 0 0 I 0
o °0° 0 0 -i 10
,0 0 00 0 0- 0: i
SOUTH OKLAHOMA
o
o
o 0
o
o
Figure 7. Distribution of cobalt and nickel. Tick marks
indicate medians for the elements in the samples. Areas
1 and 4 are near elemental source area (ref. 1).
o
AF
-------
ARrAS
4'1 I
YTTRIUM
~
2'J
I
201
I I
I !
I !
! !
i i
I :
I I
I !
, 100
00
1
I
I
I °
~ 0
1°
II
I I
I :
I I
I i 0
i !
;;J.I
o
u
?;I
z
o
:J
-J
~
~ " LANTHANUM I II
~3Oj I
25j i I
! ' ,
o
o
°
o
00
o
00
o
o
o 0 0
°0 0°0°00
o
o 0
o
o
o
°0°0
°
201
I~
i I 0
"
i
I~o
! I + 0
s.i ~ :
I t C Co :
O} 0"00 ~,,0""1 r "" 00°0°,,°000 ,,~
IOWA, MIS5I..'XJf;, SOUTH OKLAHOMA
NORTH C>t
-------
However, when the YILa ratios are plotted (fig. 9), a
significant separation of these elements takes place.
Yttrium, which has a higher organic affinity than la, is
significantly more enriched over La in the near-source
areas. Thus, the behavior of Y resembles that of the
other elements that have high organic affinities.
Vertical distributions of the elements also have sig-
nificant regional differences. The three samples from
Illinois bed 5 in figs. 10, 11, and 12 were collected at the
northern, central, and the southern parts, respectively,
of the coal basin. As the distance from the source area
increases, the elemental distribution patterns (except for
boron) become very similar. A pattern closely resemb-
ling that of the ash is evident. The heterogeneity of the
element distribution in the near-source sample (fig- 10)
can be explained by the fact that many of the elements
may be entering the swamp area as undissolved clastic
minerals. In addition, those elements in solution would
not be subjected to entrapment processes before enter-
ing the area of organic deposition. This explanation also
holds for coals deposited adjacent to the distributory
channels within the swamp. The solutions reaching the
deeper recesses of the swamp become depleted in soluble
elements. The elemental concentrations in these waters
assume relations to each other which are dependent
upon the dissociation constants of the most stable com-
pounds that could be formed in the marginal areas of the
swamp. Additional complexing or other reactions can
take place only if the equilibrium concentrations are
exceeded. This can take place only by evaporation of
water, which would result in a uniform concentration
increase for all the elements and give the type of distri-
butions depicted in figs. 11 and 12. The degree of vari-
ance found vertically throughout the bed is caused by
changes in the residence time of the solutions in the
swamp, which, in turn, is largely dependent upon clima-
tic conditions.
REGIONAL DISTRIBUTION OF THE ELEMENTS
The average trace-element contents of coal from
four areas in the United States are listed in table 2.
The Eastern Interior region is highest in six of these
elements, the Northern Great Plains, in five, the Appala-
chian region, in three, and Western I nterior region, in
one. The distribution of environmentally objectionable
elements in coals is shown in table 3 (refs. 9,13).
r- I I l I I I I
'~ '"''''" I I t
LANTHANUM I
I I I
I I 0
12j I " 0 r
I ° I I I
111 101 I I I f
9 I, I
0 I : !
cr 8! 0 " I I
C I I
...J I
'. Gj cj
"" I
410 ° .,
0 i I
i I 00000 ° Vo c c ! I
nOc I 0 I
210 ° I i r
I 0
°1 (J (' 00 °c i !OOooooooo(') C' I, C c"
l C
~-~ I -. '------- ---------'-
IOWA, MI:,sQUf"/: SOUTH 04\L.lu..;:;V:' ARKANSA~ -- \.':-
NORTH OKLAHOMA
Figure 9. Distribution of the yttrium!1anthanum
ratios. Areas 1 and 4 are near elemental source
area (ref. 1).
58
-------
...
CI)
CI)
- 2
c:
..
..
CI)
c:
-"
.!:2
£
01 -c
-------
~F ~fl
... ,~ ! ! ~
'"
'"
~ ! I I
.!: I !
'" I
'" I
'"
c: I,
.><
.2
en .c 1
...
0 "'C
'"
ID
o
o
I
Alh
II
o
2
81
c
100 7.00 0 400
8 TI
o 110
V
o
20 40
Cr
o z
Co
4 0 20 40
HI
o 10 20 0 2
CII Go
.. 0 20
GI
o
4
,",0
.
o
4
Y
.
% in coal
Note: Numbers for the elements are parts per million in coal.
Figure 11. Distribution of the elements in sample 111-F-5, taken from the central part of the
coal basin, some distance from the elemental source area.
-------
~
4
...
Q)
Q)
'>- 3
.s
0') '"
.... '"
Q)
c:
~
.~ 2
.<:
...
"'0
Q)
CC
II
o
o
11
4 0 10
T
o 10 20
L..
4
10 20
G.
o t
"0
o
4
AI.
o
o 20 40
8
o 40 10 0
TI
20 40 0
V
10 0 to 40 0
HI
20 40 0 20 40 0 2
C. Zfl Go
20
C,
o s
C.
80
% in coal
Note: Figures for the elements are parts per million in coal.
Figure 12. Vertical distribution of the elements in sample 111-H-5, taken from the southern part
of the coal basin, farthest from the elemental source area.
-------
Table 2. Trace-element content of American coals
[Parts per million]
Northern Western Interior Eastern Interior Appalachian
Element Great Plains region region region
Be 1.5 1.1 2.5 2.5
B 116 33 96 25
Ti 591 250 450 340
V 16 18 35 21
Cr 7 13 20 13
Co 2.7 4.6 3.8 5.1
Ni 7.2 14 15 14
Cu 15 11 11 15
Zn 59 108 44 7.6
Ga 5.5 2.0 4.1 4.9
Ge 1.6 5.9 13 5.8
Mo 1.:7 3.1 4.3 3.5
Sn .9 1.3 1.5 .4
Y 13 7.4 7.7 14
La 9.5 6.5 5.1 9.4
Table 3. Distribution of environmentally hazardous
trace elements
{Parts per million in coal]
Powder River Western Interior Eastern Interior Appalachian
Element Basin region region region
Sb 0.67 3.5 1.3 1.2
As 3 16 14 18
Be .7 2 1.8 2.0
Cd 2.1 20 2.3 .2
Hg .1 .13 .19 .16
P.b 7.2 34 12
Se .73 5.7 2.5 5.1
Zn 33 250 13
62
-------
The highest concentrations of Sb, As, Be, Cd, and
Se occur in the Western Interior region. Hg, Pb, and Zn
are highest in the Eastern Interior region. Be is high in
the Appalachian region, whereas Powder River Basin
coals are lowest in all three elements.
SUMMARY
Geological factors are principally responsible for the
availability of trace elements to the emplacement pro-
cesses of a coal-forming swamp. These same factors are
responsible for regional differences in the trace-element
content of coals. Physicochemical processes control the
type of associations that various metals will have in coal
as well as their distribution in the coal of a basin.
Some of the trace-element distributions, such as the
concentrations of elements having high organic affinities
and the vertical similarity or heterogeneity of the ele-
ments within a bed, can be used to ascertain the position
of the coal within the depositional basin. It can also be
used to predict the type of distribution that will be
found in the vicinity of the sampled area.
REFERENCES
1. Peter Zubovic, "Minor Element Distribution in Coal
Samples of the Interior Coal Province," Advances in
Chemistry Series 55: 231-247, 1966.
2 Peter Zubovic, T. M. Stadnichenko and Nola B.
Sheffey, "Relation of the Minor Element Content
of Coal to Possible Source Rocks,' 1&;. S. Geological
Survey Prof. Paper 400-B: p. 683, 1960.
3. P. E. Potter, and H. D. Glass, "Petrology and Sed-
i mentation of the Pennsylvanian Sediments in
Southern Illinois," Illinois State Geological Survey
Report of Inv. No. 24: 52-53, 1958.
4. Peter Zubovic, "Minor Element Content of Coal
from Illinois Beds 5 and 6 and their Correlatives in
Indiana and Western Kentucky," U. S. Geological
Survey, open-file report; 79 pp., 1960.
5. H. J. Gluskoter, and Peter C. Lindahl, Cadmium:
Mode of "Occurrence in Illinois Coal," Science,
181: 264-266, July 20, 1973.
6. Allen V. Heyl, Allen F. Agnew, Erwin J. Lyons, and
Charles H. Behre, Jr., "Geology of the Upper
Mississippi Valley Zinc-Lead Deposits," U. S. Geolo-
gical Survey Prof. Paper 309,1959.
7. L. Horton and K. V. Aubrey, "The Distribution of
Minor Elements in Vitrain: Three Vitrains from the
Barnsley Seam," J. Soc. Chem. Industry,50 (suppl.
issue 1): 541-548, London, 1950.
8. Pete r Zubovic, "Physicochemical Properties of
Certain Minor Elements as Controlling Factors of
their Distribution in Coal,' Advances in Chemistry
Ser. 55: 221-230, 1966.
9. R. R. Ruch, H. J. Gluskoter, and N. F. Shimp,
"Occurrence and Distribution of Potentially Vola-
tile Trace Elements in Coal," Illinois State Geologi-
cal Survey, Environ. Geol. Note 72: 96 p., 1974.
10. F. Basalo, and R. G. Pearson, "Mechanisms of In-
organic Reactions," John Wiley and Sons, Inc., New
York, 1958.
11. H. Irving, and H. Rossotti, Acta Chern. Scand., 10,
No.1: 72-93.
12. Irving A. Breger, Maurice Deul, and Samuel Ruben-
stein, "Geochemistry and Mineralogy of a Urani-
ferous Lignite," Econ. Geol.,50(2): 206.226, 1955.
13. Jack Medlin, Personal Communication, U. S. Geol.
Survey.
63
-------
15 December 1975
Session II:
PROCESS TECHNOLOGY
E. C. Cavanaugh
Session Chai'rman
65
-------
TRACE ELEMENTS AND MAJOR COMPONENT
BALANCES AROUND THE SYNTHANE PDU GASIFIER
A. J. Forney, W. P. Haynes, S. J. Gasior,
R. M. Kornosky, C. E. Schmidt,and A. G. Sharkey*
Abstract
A series of gasification tests were run in the
SYNTHANE PDU gasifier in an attempt to define the
distribution in the effluent streams of the trace elements
that enter the gasifier in the coal. The feed coal and
nongaseous products were analyzed for 65 trace ele-
ments. Results indicate most of the elements are found
in the char and dusts with some specifically in the tars
and waters. More accurate methods of analysis are need-
ed to make reasonable balances of the trace elements.
INTRODUCTION
Many coal-to-gas plants are expected to be con-
structed in the United States in the 1980's to supple-
ment the dwindling supplies of gas. Because of environ-
mental considerations, it is important to know the fate
of various trace elements fed into the gasifier in the coal
during coal gasification. This report is a first attempt to
analyze all feed and product streams of the small
SYNTHANE laboratory ga'sifier. This gasifier has the ad-
vantage that all feed and effluent streams are representa-
tive of those that will be obtained from a commercial
operation. I n these experiments, only one coal-Illinois
#8 (River King Mine, Monroe Co., Illinois) was used
since it is believed that the Illinois coals, (because of
their high sulfur content and enormous reserves) are the
prime candidates to be used in the first commercial
coal-to-gas plants east of the Mississippi.
Other authors have discussed trace elements, (refs.
1,2) but this we believe is a first attempt to make bal-
ances of trace elements around the gasifier.
THE GASIFIER
A schematic diagram of the laboratory gasifier is
.shown in figure 1. The gasification section is a 6-foot
long section of 4-inch diameter pipe (310 SS) inside a
*The authors are with the Energy Research and Develop-
ment Administration, Pittsburgh Energy Research Center,
Pittsburgh, Pennsylvania. Forney is a research Supervisor;
Haynes and Gasior are supervisory chemical engineers;
Kornosky, a chemical engineer; Schmidt, a research chemist; and
Sharkey, a supervisory research physicist.
1O-inch pipe (304 S5) acting as a pressure shell, with
heaters and refractories in the annulus. Above the gasifi-
cation section is another 6-foot section of 10-inch
diameter pipe. The 4-inch section contains the high
(1832° F) temperature fluid bed. Coal, steam, and
oxygen are the feeds to the gasifier (plus some nitrogen)
and the products are the char, filter fines, tar, water, and
gas. With a mildly caking coal like Illinois #8, it was
necessary to pretreat the coal. While commercially the
pretreating feed gases would be steam plus oxygen, in
the small unit, nitrogen plus oxygen is used to avoid
problems due to steam condensation. The products from
the pretreater enter the gasifier so there are no separate
byproducts. Other analyses from the gasifier are dis-
cussed in more detail in an earlier publication (ref. 3).
THE TESTS
A series of gasification tests were made in the gasi-
fier using Illinois #8 coal. Attempts were made to hold
conditions as constant as possible. The overall results of
3 of the tests are shown in table 1. In addition to the
regular analysis made on the feed coal and the product
streams, trace element analyses were made on the coal
feed, the feed water and the char, fines, tar, water, and
gases. The trace element analysis was done by spark
source mass spectrometric analysis (SSMS) on all except
the gas. Trace elements in the gas were determined by
neutron activation, but results were incomplete and not
conclusive. Determinations on the feed water were also
incomplete. Mercury analysis was done by flame less
atomic absorption procedures.
Samples were taken as follows:
1. Solids are collected from hoppers by a vacuum
connected to a cyclone separator tank. Samples are
then taken by randomly scooping some coal or char
out with the sample can.
2. Condensate and tar mixture is drained from the con.
denser into 5-gallon steel buckets, decanted and in-
dividual tar and condensate samples are taken ran-
domly by dipping glass jars into the buckets of de-
canted condensate and tar.
There were 3 separate sets of data obtained for each
separate test, # 162, 163,and 164.
67
-------
Cool feed
hopper
Steam <:"
generator <
<: ..
<:~
Woter
Oxygen
Water
Steam genero tor
gasifier
Oxygen
Oxyge n
Fluid bed pretrealer
Condenser
Dust filter
Refractory
Q
Oxygen
analyzer
Ash extrac tor
Figure 1. Bruceton 40-atmosphere gasifier.
68
BPR
Bog fi Iter
DD
C hromatogroph
-------
Table 1. Run conditions and results for all tests
Run Con diti ons
Time (Hrs)
Coal Feed Rate ,¥/hr.
Coal Feed #/hr'ft3
#Steam/#Coal Feed
#02/ #Coa 1 Fee d
Press ure (ATM)
Temperature Pretreater, °C
Max.
Avg.
Temperature Gasifier, °C
Max.
Avg.
Gasifier Sup. L. Vel. (ft/sec)
Pre treater Sup. L. Vel. (ft/sec)
Run Re s u 1 ts
SCF ProQl/Gas/#Coa1
SCF CH4 Eq/#Coa1
Steam Cecomp, ~b
Ca rbon Gas ifi ed, 7~
Carbon Converted, ~
#Tar/#Coal Feed
CH 4 (E Q U) % 0 f Pur if i e d Gas
Gas Analysis, Dry, N2 Fre~ (%)
H?
~~
CH4
C02
C2HS
H2S
1/H2 + CO + CH4 + C2HS
Gas An a 1 ys is,
CH 3SH
COS
S02
C4H 4S
CSH 8S
CSHS
C7H8
C 8H 1 0
HCN
C5HSS
Dry, 1'12 Free U~)
162
6.25
20.0
42.6
1. S8
0.37
40
408
363
985
917
0.29
0.79
12.3
3.08
17.8
67. 1
71. 4
0.03
24.6
35.1
0.0
12.0
12.8
37.4
1. 29
1. 43
0.002
0.014
0.001
0.004
0.001
0.039
0.010
0.002 -9
8.8 x 10
0.001
163
5.50
20.4
43.3
1. 75
0.37
40
409
364
1020
918
0.30
0.80
12.6
3.01
17.6
68.5
72.2
0.02
23.4
35.7
0.0
13.3
12.4
35.8
1. 30
1. 41
0.002
0.020
0.001
0.001
0.001
0.012
0.003
0.002 -8
3.2 x 10
0.001
164
6.67
23.7
50.4
1. 54
0.30
40
40S
349
990
925
0.29
O. 71
11.7
3. 15
19.0
66.5
71. 7
0.04
26.3
35.4
0.0
12.3
13.9
35.3
1. 56
1. 62
0.003
0.030
0.001
0.004
0.001
0.022
0.005
0.002 -9
9.1 x 10
0.001
69
-------
Table 2. Feed stream into the gasifier
(tests 162, 163, 164)
162 163 164
Coal Feed 56750 50848 71749
gms wt% wt% wt%
C 65.0 63.0 65.0
H 4.9 4.9 4.8
o 11.6 13.4 11. 1
N 1.2 1.2 1.2
S 3.6 3.6 3.7
Ash 13.7 13.9 14.2
Moisture 3.7 5.5 2.7
Fi xed
Ca rb on 44.5 43.0 45. 1
Volatile
Matter 38. 1 37.6 38.0
Water 95350 88757 11 0549
gms wt% wt% wt%
H 11. 1 11. 1 11. 1
o 88.9 88.9 88.9
Oxygenll 20884 1 861 4 21338
gms wt% wt% wt%
o 100.0 100.0 100.0
Nitrogen£! 84802 82567 89117
gms wt% wt% wt%
N 100.0 100.0 100.0
lINot analyzed
?JNot analyzed
70
-------
Table 3. Streams from the gasifier
(tests 162, 163, 164)
162 163 164
Condensate 81723 75894 90721
gms wt% wt% wt%
C 1.6 1.6 1.6
H 10.9 10.9 10.9
o 87.0 87.0 87.0
S 0.5 0.5 0.5
Filter Dust 684 269 871
gms wt% wt% wt%
C 75.0 72.8 74.9
H 1.7 1.7 1.9
o 2. 1 2.2 2.0
N 0.8 0.8 0.9
S 1.3 1.4 1.4
Extractor
Hopper 12939 11804 22019
gms wt% wt% wt%
C 50.9 61. 3 48.7
H 0.7 0.7 0.6
N 0.3 0.4 0.3
S 0.6 0.7 0.5
Tar 1844 1368 2920
gms wt% wt% wt%
C 83.0 82.8 83. 1
H 6.4 6.3 6.4
o 6.4 6.0 6.4
N 1.2 1.2 1.3
S 2.7 2.6 2.7
Gas 154682 143195 206444
gms wt% wt% wt%
C 15.2 14.2 15.1
H 2.6 2.5 2.6
o 26.5 24.8 23.4
N 54.8 57.7 43.2
S 0.9 0.9 1.0
Gasifier1J 3355 2740 590
gms wt% wt% wt%
C 69.3 62.6 60.8
H 0.8 1.0 1.2
o 0.7 1.5 0.9
N 0.4 0.6 0.9
S 0.8 1.1 1.4
lIChar left in gasifier at the end of run.
71
-------
DISCUSSION OF RESULTS
Table 1 shows the overall results of 3 tests. They
were roughly duplicates of each other; for example, the
percent steam decomposition and percent carbon gasi-
fied for each experiment are in good agreement with
each other.
Tables 2 and 3 show the feed plus product streams
of the gasifier. Table 4 shows the overall weight balances
of the five main components for each test. As can be
seen, they are good-ranging from 96 percent to 100
percent. Table 5 shows the analysis of 65 trace elements
made by spark source mass spectrometry. The results in
this table are reported as ppm wt. (I1g/g) while the water
is reported as pg/ml. It is recognized that higher accur-
acy can be achieved and will be done on selected ele-
ments in the next series of tests rather than getting so
many analyses. However, it is believed for this first
attempt that the coverage of elements analyzed should
be as wide as possible. Some of these values on the feed
coal do not aggree with Ruch (ref. 4).
Table 4. Overall mass balances of the major
components (tests 162, 163, 164)
L: In L: Out %
kg kg Recovery
162
C 36.9 35.8 97.0
H 13.4 13.3 99.5
o 112.2 112.2 100.0
N 85.5 84.9 99.3
S 2.0 2.0 100.0
163
C 32.0 31. 8 99.4
H 12.3 12.1 97.6
o 104.3 101. 7 97.4
N 83.2 82.6 99.4
S 1.8 1.8 100.0
164
C 46.6 46.7 100.2
H 15.7 15.4 98.2
o 127.6 127.4 99.9
N 89.9 89.2 99.2
S 2.7 2.6 96.3
Table 6 shows the percent recovery of 16 selected
elements which are considered the most important of
the 65 analyzed. While the elements Hg, U, As, Zn, Mn,
Cr, V, P, F, B, and Be indicate some reasonable degree of
recovery, Pb, Cd, Se, Ni, are too inaccurate. Other ele-
ments, while not listed in table 6 are as inaccurate as Sr,
Fe,and AI. It is recognized that SSMS values are usually
considered good to within a factor of 3.
These wide ranges, both plus and minus, would indi-
cate more precise analytical methods are needed to get
better yield data.
Tables 7, 8,and 9 show where these trace elements
are in the various streams for 3 experiments. The first
part of each table was made by totalling the yields from
the gasifier and then, calculating of the total, where the
percentage of the elements are in the water, tar, fines,
and char. The second part shows how much of the feed
quantities are found in each of the off streams. It is seen
(table 9) in both cases that the greatest percentage of the
16 elements (remembering the accuracy discussed with
table 6) is in the chars and dusts except Hg, B, Se, and F,
some of which are found in the water. Almost all of the
chlorine is in the water. The tars contain some As, Pb,
Cd, and Hg. While the determinations of elemental
analysis of the feed water and off-gas were incomplete
and not used, it is believed these are too small to affect
the balances. This belief is strengthened by the analysis
shown in TPR-6 (ref. 3) in the gas.
Table 10 shows the results of a comparison of trace
element concentrations in the condensate from the gasi-
fier compared to the same type analysis on a Mononga-
hela River sample. As can be seen, except for elements
such as B, Hg,and Se, in many cases the percentages of
the trace elements is actually higher in the river water.
CONCLUSIONS
These three gasifier tests made in the small PDU
gasifier at Bruceton, Pa., indicate that the trace elements
primarily remain in the chars and dusts emanating from
the gasifier. Some elements, such as boron, chlorine,
fluorine, and selenium are found in the water and some
such as arsenic, lead, 2nd cadmium are in the tars. Of
course, the high sulfur percentage in the tars was already
known from the major component analysis. Further
tests are needed to concentrate on better analyses of
fewer elements to obtain greater accuracy and better
balances. Unfortunately, the mercury analyses were
incomplete, but most of the mercury appears in the tar
and water, with little remaining in the char or dust.
72
-------
Table 5A. Trace element analysis of all streams (test 162)
Feed Coal Filter Fines Char Tar H20
162 ppm (~g/g) ppm (~9/g) ppm ("gig) ppm (,,9/9) ppm ("gig)
Ag 0.01 <0.01 <0.05
Al >0.5% 540 1800 29 0.007
As 0.87 3.7 6.5 0.71 0.001
8 86 64 380 12 43
8a 170 130 98 3.6 0.10
8e 1.5 7.2 4.6 0.03
8i <0.10 <1. 7 <0.44 0.20
Br 0.23 0.65 1.6 0.02 0.001
Ca >1% >0.5% >1% 450 2.4
Cd 0.097 0.88 1.6
Ce 47 25 54 0.29
C1 93 11 33 1.5 300
Co 14 17 95 0.09 0.002
Cr 170 47 240 7.1 0.043
Cs 0.26 1.2 0.65
Cu 39 70 40 0.74 0.003
Oy 1.4 3.9 1.6
Er 2.1 0.41 0.80
Eu 0.55 0.39 0.65
F 490 610 150 0.97 39
Fe >1% >0.5% >1% 240 (J.081
Ga 8.3 3.6 4.5 0.08
Gd 1.9 1.2 0.48
Ge 1.1 1.3 5.4 0.08
Hf 0.83 3.5 11
Hg 0.10 0.20 * 1.2 0.027
Ho 0.43 0.16 0.45
I 0.4 1.9 0.27 0.02
K >1% 190 5400 14 0.31
La 22 6.7 17 0.03
Li 0.8 34 67 0.51 0.001
Lu 0.085 <0.18 0.40
Mg 2800 4600 3500 240 0.57
t-'.n 160 48 240 2.2 0.20
Mo 15 21 14 0.31
Na 1900 >1% 4700 360 6.6
rib 4.7 7 13 0.08
Nd 23 19 11 0.06
Ni 43 12 25 1.2 0.018
P 130 460 460 14 0.04
Pb 0.55 2.2 21 0.22 0.003
Pr 7.3 7.5 4.2 0.02
Rb 180 36 27 0.10
5 >1% 7700 2100 120 1.6
Sb 0.18 0.04 1.9
Sc 5.3 6.4 17 0.02
Se 2.2 15 4.7 0.23 0.14
Si >1% >1% >1% 500 2.8
Sm 2.7 0.30 0.86 0.01
Sn 0.6 0.75 1.9 0.03
Sr 3.3 44 70 4.5 0.12
Ta 0.73 0.64 1.2
Tb 0.2 0.20 0.89
Te <0.29 <0.19 0.15
Th 3 5.8 4.3 0.06
Ti 880 1800 3300 8.4 0.003
Tl <0.12 <0.19 <0.25 0.11
Tm 0.24 0.10 0.20
IJ 1.4 5.6 5.4 0.01
V 100 44 190 0.21
W 0.08 2.2 4.8 0.09
Y 21 37 48 0.10
Yb 0.35 2.7 2.2
Zn 25 11 100 0.48 0.13
Zr 10 22 28 0.26
*Insuffi cient results.
73
-------
Table 58. Trace element analysis of all streams (test 163)
Feed Coa 1 Fi lter Fines Char Tar H20
163 ppm (ug/g) ppm (ug/g) ppm (ug/g) ppm (ug/g) ppm (Ug/g)
Ag <0.009 <0.054 <0.054
A1 2400 1000 790 30 0.006
As 1.3 6.5 3.3 1.6 <0.001
B 86 32 300 52 17
Ba 190 80 98 6.0 0.095
8e 1.3 4.6 2.1 0.07
Bi <0.10 3.1 1.2 0.17
Br 0.13 3.2 1.6 0.08
Ca >1% >0.5% >1% 2000 3.6
Cd 0.097 1.6 0.77 0.15
Ce 22 22 89 1.5
C1 220 33 14 1.2 170
Co 6 4.1 9.5 0.35
Cr 100 120 120 10 0.004
Cs 0.13 0.33 0.65 0.04
Cu 27 18 40 6.4 0.002
Oy 0.7 1. 21 3.4 0.10
Er 1.6 0.40 0.54
Eu <0.22 0.65 1.3 0.02
F 420 740 320 12 37
Fe >0.5% >1% > 1% 530 0.18
Ga 5.8 2.1 2.1 0.17
Gd 0.081 0.48 0.95 0.01
Ge 2.1 2.5 2.5 0.08
Hf 0.73 2.2 1.1 0.04
Hg 0.14 0.07 * 0.47 0.029
Ho 0.19 0.30 0.30
I 0.23 2.7 0.23 0.08
K >0.5% 2300 2300 130 0.46
La 11 7.7 7.7 0.16
Li 0.8 38 38 0.48 0.001
Lu <0.081 0.35 0.40
Mg 1600 3500 3500 820 1.0
Mn 220 140 500 4.7 0.01
Mo 11 14 28 0.32
Na 1900 >0.5% >1% 780 6.8
Nb 4.7 13 6.5 0.18
Nd 23 4.6 11 0.31
Ni 10 11 25 1.6 0.002
P 63 1100 2100 31 0.04
Pb 1.1 8.9 21 0.48 0.002
Pr 7.3 1.8 4.2 0.12
Rb 67 23 27 1.1
5 >1% 4300 2100 1000 0.74
Sb 0.10 0.95 0.95
Sc 7 8.3 8.3 O. lO
Se 1.2 4.7 4.7 0.24 0.002
Si >1% >1% >1% 4000 4.7
Sm 1.1 0.86 1.7 0.04
Sn 1.0 1.4 1.9 0.07
Sr 2.3 33 70 9.8 0.055
Ta 0.31 1.2 1.2 0.02
Tb 0.086 0.11 0.22
Te <0.29 <0.11 <0.11
Th 6.4 4.3 9.2 0.21
Ii 800 590 1700 67
T1 <0.12 <0.25 <0.25 0.24
Tm 0.24 0.12 0.13
U 3 2.7 14 0.06
V 43 41 95 2.6 0.001
W <0.024 4.8 9.6 0.21
Y 10 9.5 20 0.37
Yb 0.4 0.96 0.96 0.03
Zn 49 dB 100 3.3 0.007
Zr 4.8 9.3 28 0.57
*Insuffi cient results.
74
-------
Table 5C. Trace element analysis of all streams (test 164)
Feed Coa 1 Fi 1ter Fines Char Tar H20
164 ppm (jJg/g) ppm (jJg/g) ppm (jJg/g) ppm (jJg/g) ppm (ug/g)
Ag 0.013 <0.09 0.21
A1 2000 330 3200 20 0.009
As 1.5 6.1 2.9 1.1 0.001
B 86 40 160 12 82
Ba 83 80 140 1.5 0.058
Be 0.64 2.6 5.7 0.05
Bi <0.10 1.5 <0.99 <0.30
Br 0.13 0.75 1.9 0.01 0.001
Ca >1% 3500 >1% 240 2.0
Cd 0.087 0.73 1.4
Ce 16.5 16 15 0.20
C1 93 6.7 14 0.42 190
Co 3.5 11 10 0.24
Cr 100 29 220 3.0 0.006
Cs 0.13 0.73 0.69
Cu 39 23 22 0.85 0.003
Dy 1.1 1.8 3.8
Er 1.6 0.6 0.57
Eu <0.22 0.21 0.34
F 280 690 240 1.4 32
Fe >1% >1% >1% 200 0.11
Ga 1.9 2.3 4.6 0.02
Gd 1.4 0.37 0.35
Ge 0.91 1.9 7.6 0.02 0.001
Hf 0.66 1.4 0.45
Hg 0.08 0.06 0.04 0.21 0.030
Ho 0.13 0.16 0.15
I 0.4 1.4 <0.20 0.001
K >1% 190 5400 14 0.31
La >0.5% 2300 2300 130 0.46
li 0.4 21 40 0.33
Lu <0.081 0.14 0.12
Mg 800 4400 2700 360 0.32
Mn 220 40 160 1.6 0.014
Mo 6.3 13 25 0.11 0.001
Na 1900 >1% >1% 530 5.4
Nb 4. 7 8.7 4.1 0.03
Nd 9.9 12 11 0.09
Ni 10 18 25 0.95 0.001
P 130 >1% 630 9.1 0.022
Pb 0.55 2.8 5.7 0.16
Pr 7.3 3.5 4.4 0.04
Rb 25 4.9 54 0.06
5 >0.5% 4800 2300 42 0.57
5b 0.1 0.06 0.11
5c 7 4 11 0.01
5e 0.58 9.3 3.8 0.08 0.18
5i >1% >1% >1% 742 6.6
5m 1.1 0.43 0.61 0.10
5n 0.45 0.93 0.88 0.02
5r 15 59 55 2.4 0.034
Ta 0.73 0.8 0.16
Tb 0.086 0.12 0.12
Te <0.029 <0.12 0.16
Th 4.5 5.1 4.9 0.04
Ti 800 1100 1100 8.1 0.036
T1 <0.12 <0.12 <0.11 0.16
Tm 0.24 0.15 0.14
U 3 5 4.7 0.02
V 35 50 47 0.31 0.002
VI <0. 024 1.9 2.9 0.02
Y 10 23 22 0.07
Yb 0.28 1.7 1.6
Zn 49 28 66 1.4 0.096
Zr 10 31 13 0.08
75
-------
Table 6. Trace element balances of selected components
L: In L: Out % L: In L: Out %
mg mg Recovery mg mg Recovery
162 163 (con.)
As 49.4 109.9 222.5 Mn 11186.6 7317.0 65.4
B 4880.5 9771.8 200.2 Ni 508.5 369.0 72.6
Be 85.1 79.9 93.9 P 3203.4 30884.4 964.1
Cd 5.5 26.7 485.5 Pb 55.9 308.6 552. 1
C1 5277.7 25065.0 474.9 Se 61.0 70.6 115.7
Cr 9647.5 3959.4 41. 0 U 152.5 204.5 134.1
F 27807.5 6050.4 21. 8 V 2186.5 1 396. 4 63.9
Hg 5.7 4.6 80.7 Zn 24Q1. 6 1472.5 59. 1
Mn 9080.0 3963.9 43.7
Ni 2440.3 419.3 17.2 164
P 7377.5 7839.2 106.3 As 107.6 74.2 69.0
Pb 31. 2 344.4 1103.7 B 6170.4 11126.5 180.3
Se 124.9 788.5 631. 3 Be 45.9 131.3 286.1
U 79.5 100.0 125.8 Cd 6.2 32.3 521.0
V 5675.0 3126.4 55.1 C1 6672. 7 17560.7 263.2
Zn 1418.8 1648.5 116.2 Cr 7174.9 5008.6 69.8
F 20089.2 8934.3 44.5
163 Hg 5.7 4.3 75.4
As 66. 1 52.0 78.7 Mn 15784.8 3658.3 23.2
B 4372.9 5951. 9 136. 1 Ni 717.5 583.9 81. 4
Be 66.1 31. 9 48.3 P 932 7. 4 14280.0 153. 1
Cd 4.9 11.8 240.8 Pb 39.5 131.8 333.7
C1 11186.6 15302.9 136.8 Se 41.6 111.2 267.3
Cr 5084.8 1 791. 6 35.2 U 215.2 110.7 51.4
F 21356.2 8153.6 38.2 V 2511. 2 1107.2 44.1
Hg 7. 1 3.2 45.1 Zn 3515.7 1529.5 43.5
76
-------
Table 7. Percentage of selected components in major effluent streams
(test 162)
Out H20 Tar Char Fines
162 mg % % % %
As 109.9 O. 1 1.2 96.4 2.3
B 9771.8 36.0 0.2 63.4 0.4
Be 79.9 o. 1 93.8 6.1
Cd 26.7 97.7 2.3
C1 25065.0 97.8 O. 1 2. 1 0.0
Cr 3959.4 O. 1 0.3 98.8 0.8
F 6050.4 52.7 0.0 40.4 6.9
Hg*
Mn 3963.9 0.4 o. 1 98.7 0.8
Ni 419.3 0.4 0.5 97. 1 2.0
P 7839.2 0.0 0.3 95.7 1.3
Pb 344.4 O. 1 o. 1 99.4 0.4
Se 788.5 1.5 O. 1 97. 1 1.3
U 100.0 0.0 96.2 3.8
V 3126.4 0.0 0.0 99.0 1.0
Zn 1648.5 0.6 0.0 98.9 0.5
In H~O Tar Char Fines
mg % % %
As 49.4 0.2 2.7 214.4 5.2
B 4880.5 72.0 0.5 126.9 0.9
Be 85. 1 0.0 O. 1 88. 1 5.8
Cd 5.5 0.0 0.0 474.0 10.9
C1 5277. 7 464.5 0.05 10.2 0.14
Cr 9647.5 O. 1 o. 1 40.5 0.3
F 27807.5 11.5 0.0 8.8 1.5
Hg 5.7 38.9 39. 1 2.5
Mn 9080.0 0.2 O. 1 43. 1 0.4
Ni 2440.3 O. 1 O. 1 16.7 0.3
P 7377.5 O. 1 0.4 101.6 4.3
Pb 31. 2 0.7 1.3 1096.7 4.8
Se 124.9 9.6 0.3 613. 1 8.2
U 79.5 0.0 O. 1 120.9 4.8
V 5675.0 0.0 o. 1 54.6 0.5
Zn 1418.8 0.8 O. 1 114.8 0.5
*Ins uffi ci ent res ults
77
-------
Table 8. Percentage of selected components in major effluent streams
(test 163)
Out H~O Tar Char Fines
163 mg % % %
As 52.0 0.3 4.2 92.2 3.3
B 5951. 9 25.4 1.2 73.3 O. 1
Be 31. 9 0.3 95.9 3.8
Cd 11.8 1.7 94.8 3.5
Cl 15302.9 98.6 0.0 1.3 O. 1
Cr 1 791. 6 0.0 0.8 97.4 1.8
F 8153.6 40.3 0.2 57.1 2.4
Hg*
Mn 7317.0 0.0 O. 1 99.4 0.5
Ni 369.0 O. 1 0.6 98.5 0.8
P 30884.4 0.0 O. 1 98.9 1.0
Pb 308.6 O. 1 0.2 98.9 0.8
Se 70.6 1.0 0.4 96.8 1.8
U 204.5 0.0 99.6 0.4
V 1396. 4 0.0 0.3 98.9 0.8
Zn 1472.5 0.0 0.3 98.8 0.9
In H20 Tar Char Fines
mg % % % %
As 66. 1 0.2 3.3 72.6 2.7
B 4372.9 34.5 1.6 99.8 0.2
Be 66. 1 0.0 0.2 46.3 1.8
Cd 4.9 0.0 4.1 228.6 8.8
Cl 11186.6 134.9 0.01 1.8 0.08
Cr 5084.8 0.0 0.3 34.3 0.6
F 21356.2 15.4 O. 1 21. 8 0.9
Hg 7. 1 36. 1 9.0 0.3
Mn 111 86 . 6 0.0 O. 1 65.0 0.3
Ni 508.5 O. 1 0.4 71. 5 0.6
P 3203.4 O. 1 1.3 953.4 9.2
Pb 55.9 0.3 1.2 546.3 4.3
Se 61. 0 1.1 0.5 112.1 2. 1
U 152.5 0.0 O. 1 133.6 0.5
V 2186.5 0.0 0.2 63.2 0.5
Zn 2491. 6 O. 1 0.2 58.4 0.5
*Insufficient results
78
-------
Table 9. Percentage of selected components in major effluent streams
(test 164)
Out H~O Tar Char Fines
164 mg % % %
As 74.2 0.3 4.3 88.3 7.1
B 11126.5 66.9 0.3 32.5 0.3
Be 131.3 0.1 98.2 1.7
Cd 32.3 98.0 2.0
C1 17560.7 98.2 0.0 1.8 0.0
Cr 5008.6 0.0 0.2 99.3 0.5
F 8934.3 32.5 0.0 60.8 6.7
Hg 4.3 63.5 14.3 21.0 1.2
Mn 3658.3 0.0 o. 1 98.9 1.0
Ni 583.9 0.0 0.5 96.8 2.7
P 14280.0 0.0 0.2 99.7 o. 1
Pb 131.8 0.4 97.8 1.8
Se 111.2 15.3 0.2 77.2 7.3
U 110.7 o. 1 96.0 3.9
V 1107.2 0.0 o. 1 96.0 3.9
Zn 1529.5 0.6 0.3 97.5 1.6
In H~O Tar Char Fines
mg % % %
As 107.6 o. 1 3.0 60.2 4.9
B 6170.4 120.6 0.6 58.6 0.6
Be 45.9 0.0 0.2 280.8 5.0
Cd 6.2 0.0 0.0 509.7 11.5
C1 6672. 7 258.3 0.02 4.7 0.09
Cr 7174.9 0.0 O. 1 69.3 0.4
F 20089. 7 14.5 0.0 27.0 3.0
Hg 5.7 47.4 10.6 15.7 0.9
Mn 15784.8 0.0 0.0 22.9 0.2
Ni 717.5 o. 1 0.4 78.8 2.2
P 9 32 7. 4 O. 1 0.3 152.7 o. 1
Pb 39.5 0.0 1.2 326.3 6.2
Se 41.6 40.8 0.6 206.5 19.5
U 215.2 0.0 O. 1 49.4 2.0
V 2511.2 0.0 o. 1 42.3 1.7
Zn 3515.7 0.3 o. 1 42.4 0.7
79
-------
Table 10. Trace element analysis of Monongahela riverwater and an earlier
gasification condensate
Man River: 12th Synthane Man River: 12th Synthane
St. Bridge, Pgh CHPFI-120 St. Bri dge, Pgh CHPFI -120
ppm (vol) ppm (vol) ppm (vol) ppm (vol)
Ag 0.025 0.009 Na 16 0.22
A1 0.13 2.7 Nb 0.040
As 0.010 0.071 Nd 0.017
B 0.24 85 Ni 0.11
Ba 0.50 0.51 P 0.17 0.40
Br 0.041 0.017 Pb 0.021 O. 032
Ca 25 1.5 Rb 0.31 0.77
Cd S 21 NR*
Ce 0.003 0.005 Sb 0.009 O. 032
C1 3.8 0.41 Sc 0.012 0.025
Co 0.005 0.031 Se 1.1
Cr 0.004 0.029 Si 16 43
Cs 0.015 0.015 Sm 0.011
Cu 0.012 0.013 Sn 0.13
Fe 0.88 0.90 Sr 0.56 0.28
Ga 0.007 0.014 Ti 2.2 1.1
Ge 0.020 0.038 U 0.025 0.025
Hg <0.0001 0.015 V 0.008 0.008
I 0.005 0.033 W
K 1.8 <0. 1 Y 0.01 0.009
La 0.016 0.036 Zn 0.40 0.17
Li 0.01 Zr 0.043 0.028
Mg 6.5 0.39
Mn 0.19 0.008
Mo 4.9 1.8
*Not reported due to matrix difficulties
-Less than 0.001 mg/m1 (ppm)
80
-------
ACKNOWlE DGMENTS
We acknowledge with thanks the help of our analyt-
ical groups at Bruceton; especially the General Analysis
Group headed by H. Schultz, with special thanks to W.
McKinstry, F. E. Walker, J. F. Smith, M. J. Mima, and
M. F. Ferrer. Other in-house analyses were done by the
Spectro-Physics Group headed by R. A. Friedel. The
trace analyses were handled by A. G. Sharkey through
arrangements with Accu-lab Co., Wheaten, Colorado.
REFERENCES
1.
R. R. Bertrand and C. E. Juhrig, "A look at Envi-
ronmental Aspects of Coal Conversion," American
81
2.
Chemical Society, Division of Fuels Chemistry,
Philadelphia, Pa., April 7-11, 1975.
A. Attari, M. Mensinger, and J. C. Pare, "Fate of
Trace Constituents of Coal During Gasification"
(Part II), American Chemical Society Division of
Fuel Chemistry, Philadelphia, Pa.,ApriI7-11,1975.
A. J. Forney, W. P Haynes, S. J. Gasior, G. E.
Johnson, and J. P. Strakey, Jr., "Analysis of Tars,
Chars, Gases and Water Found in Effluents from the
Synthane Process," BuMines TPR 76, January 1974,
9pp.
R. R. Ruch, H. J. Gloskoter, and N. F. Shimp,
"Occurrence and Distribution of Potentially Vola-
tile Trace Elements in Coal: A Final Report," Envi-
ronmental Geological Notes, No. 72, Illinois Geo-
logical Survey. Urbana, III., August 1974.
3.
4.
-------
CO2 ACCEPTOR PROCESS
Carl Fink, George Curran, and John Sudbury*
Abstract
Recent operation of the CO2 Acceptor Process pilot
plant has demonstrated that the gasifier product gas con-
tains no detectable amounts of tars, oils, phenols, or
cyanides. Three key process advantages also have been
confirmed: (1) suitable synthesis gas is made without
the use of an oxygen plant, (2) no water-gas shift plant is
required, and (3) gasifier product gas cleanup is simpli-
fied because the gas contains less H2 S and CO2 than in
any other process. Significant progress has been made
toward improving reliability of plant operation.
INTRODUCTION
The CO2 Acceptor Process, a fluidized bed system
to convert lignite or subbituminous coal to pipeline gas,
is undergoing testing on a pilot plant scale at Rapid City,
South Dakota. The demonstration program is being
carried out by Conoco Coal Development Company
under contract with the Energy Research and Develo-
pment Administration and the American Gas Associ-
ation.
Description of the Process
A schematic diagram of the CO2 Acceptor Process is
shown in figure 1. There are two fluidized bed reactors,
a gasifier, and a regenerator which operate at a pressure
of 150 psig. Lignite or subbituminous coal is fed to the
bottom of the gasifier where, after rapid hydrodevolatil-
ization, gasification of fixed carbon with steam occurs.
The gasifier temperature is in the range of 1480° F to
1550° F. Heat for the gasification reactions is supplied
by a circulating stream of lime-bearing material called
acceptor. The acceptor, which can be either limestone or
dolomite, supplies heat needed for gasification princi-
pally by the exothermic CO2 acceptor reaction:
CaO + CO2 = CaC03
t:" H = -76,200 Btu/lb mole (77" F)
The CO2 acceptor reaction is reversed in the regen-
erator at about 1850° F where heat is supplied by burn-
ing the residual char from the gasifier with air. Ash is
removed from the regenerator by elutriation and collect-
ed via an external cyclone and lockhopper system. Seals
between the gasifier and regenerator are maintained by
purged standlegs of solids.
*The authors are with the Conoeo Coal Development
Company, Library, Pennsylvania.
C02 ACCEPTOR
PROCESS DIAGRAM
AIR
LIGNITE
STEAM
STEAM
Figure 1. CO2 acceptor process diagram.
Since the acceptor loses reactivity to the CO2 accep-
tor reaction as it circulates between the reactors, some
of the circulating inventory purposely is withdrawn from
the gasifier and replaced with fresh stone makeup. The
makeup is added to the acceptor which is returning to
the regenerator.
SUMMARY
Since October 1974,7 runs have been made in the
Rapid City pilot plant bringing the total to 28. The last
3 runs in this series were of sufficient length and
smoothness to provide process heat and material balance
data. The final run, Run 28, featured 12 days of integra-
ted operation at process temperatures, nearly 10 days in
which the acceptor supplied the entire gasifier heat duty,
and 3 days of steady-state plant operation. Major
accomplishments of these runs include:
1. Demonstration that the accumulation of inter-
mediate fines in the gasifier char bed can be success-
fully controlled (Runs 26,27, and 28).
83
-------
2. Demonstration of gasifier operation with no recycle
gas dilution ofthe inlet steam (Runs 27 and 28).
3. Accomplishment of a smooth, orderly plant shut-
down (Run 27).
4. The successful use of limestone quarried in Rapid
City as a CO2 acceptor (Run 28).
The problems and progress in plant operation since
February of 1974 are summarized in table 1. This table
lists the problems encountered and the accomplishments
of all the runs since Run 17. Problems resulting in run
termination are indicated as are the solutions of plant
problems. During the year, a major process problem-the
accumulation of intermediate fines in the gasifier char
bed-was solved by adding a lockhopper system to con-
tinuously remove a slip stream of solids from the char
bed. A major effort towards establishing plant reliability
was made following Run 25. The goal of this effort was
to make mechanical and procedural revisions where
necessary to eliminate plant upsets. The success of Runs
26, 27, and 28 shows that this goal is near at hand.
The results of Runs 26, 27, and 28 have reaffirmed
the advantages of the CO2 Acceptor Process:
1. The ability to produce a synthesis gas su itable for
conversion to pipeline gas without the use of oxy-
gen.
2. The absence of hydrocarbons other than CH4 in the
gasifier product gas.
3. Substantially total carbon utilization as shown by
the fact that the carbon contained in the ash re-
moved from the regenerator is less than 1 percent of
the carbon in the feed.
4. A ratio of H2 to CO in the product which exceeds
3: 1, such that all of the CO and part of the CO2 can
be methanated with no water-gas shift required.
5. Low concentrations of CO2 and H2 S in the product
gas which reduce the gas cleanup requirements.
PLANT OPERATION
Solu tion of the Intermediate Fines Accumulation
Problem
Last year it was reported that a major process prob-
lem had surfaced during Run 21. Particles too large to be
elutriated from the regenerator with the char ash and
too small to be removed with the reject acceptor, accu-
mulated in the gasifier char bed affecting the ability to
maintain a stable char-acceptor interface. After 10 days'
operation during Run 21, these particles accounted for
34 percent by weight of the gasifier char bed. Removal
of a sl ip stream of gasifier char bed material into lock-
hoppers has solved this problem entirely. The level of
intermediate fines in the gasifier char bed was controlled
throughout Run 28 which was 12 days in duration. The
char withdrawal rate was set such that the intermediate
fines level was essentially constant at 10 percent during
the final 6 days of the run. In a commercial plant, the
purged char bed material could be ground to a size
which permits the intermediate fines to be elutriated
from the regenerator bed with the char ash. This ground
material would be returned to the regenerator thus mini-
mizing the thermal penalty.
Mineral impurities existing in the coal seam are the
source of these intermediate fines. In the Velva lignite
and Husky char, these impurities are principally grains of
a: quartz. Figure 2 shows the percent grains by weight
plotted against the weight percent Si02 in the Velva
lignite and Husky char. The percent grains were deter-
mined by ashing the coal and screening the ash residue.
The portion of the residue that cannot be crushed and
forced through a 200 Tyler mesh screen is the grains
fraction. The figure illustrates that the fraction of the
coal which forms intermediate fines (grains) is directly
proportional to the silica content of the feedstock. The
quartz grains react with char ash and attrited acceptor
forming intermediate fines.
In the plant, the char withdrawal stream is removed
from the standleg through which the regenerator fuel
char flows. During Run 28, 200 Ibs/hr of material were
withdrawn to maintain the intermediate fines level at 10
percent. The char content of the withdrawn material was
about 52 percent by weight. The intermediate fines
. problem had been solved during Run 23. Early in that
run, the char withdrawal rate was too low and the inter-
mediate fines concentration in the gasifier char bed built
up to a potentially dangerous level. The level of inter-
mediate fines concentration was then reduced to an
operable level by minimizing the lignite feed rate and
maximizing the char withdrawal rate. During this opera-
tion, the weight of intermediate fines in the system was
decreased by 1100 Ibs in 26 hours. Thus, the ability to
recover from an accumulation of intermediate fines has
been demonstrated.
Although the silica content and, therefore, inter-
mediate fines potential may vary in different feedstocks,
the char removal approach offers a means of coping with
the problem. Five runs have been made since Run 23,
and no problems related to intermediate fines accumula-
tion have occurred.
Improvement of Plant Reliability
Even though the intermediate fines .problem had
been solved, the plant did not operate reliably during
Runs 24 and 25. Problems existed primarily in two areas
of the plant, the gasifier quench system and the fuel char
standleg which includes the char removal lockhoppers.
Following Run 25, a major effort was made to make
84
-------
Table 1. Problems and progress in CO2 acceptor process
Run No.
Date
Mechanical Problems
Carburi7atlon Attack of Gasifier Heater Coils
Weld Failure - r.asifier Internal Cyclone
Str~ss Corrosion Cracking - Cold Expansion Joints
Miscellaneous - Peed Preparation System
Lift Line Erosion (Misaligned Slip Joints)
Plugs in Gasifier Quench System
Char Withdrawal Lockhopper Pressure
Weld Failure Butterfly Valve
Utility Power Failllre
Upsets
Reject Acceptor Lockhopper Upsets
Problems With Solids Sample Station
Process Problems
ex:>
(J'1
Accumulation of Intermedi~te Fines
Process Accomplishments
Integrated Plant Operation
Six O",ys or More
Supplied All Process Heat by Circulating Acceptor
100 Percent Replacement of Dead-Burned Dolomite
Successfui r.asiflcation of Raw Lignite
Obt::dned Heat and Material Balance Data
JOO Percent Steam to Gasifier
Rapid City Limestone as Acceptor
Orderly Shutdown
Lignite, With Higher Sodium Content as Feedstock
Subbituminous Coal as Feedstock
Operation of Turbine Test Stand
M'thao<=J.tion
1,
Problem substantially solved
~ Caused termination of run
17 18 19 20 21 22 23 24 25 26 27 28
Feb 22 Apr 19 June 5 Aug 2 Aug 10 Sept 20 -Jan 7 Feb 21 Mar 31 May 17 June 27 Aug 26
Apr 2 Apr 29 July 17 Aug 19 Se p t 15 Dec 22 Feb 5 Mar 14 Apr 10 June 17 Aug 12 Sept 27
1974 1974 1974 1974 1974 1974 1975 ]975 1975 1975 1975 1975
*
c=r=J
c=r==r=::J
r
~~~~~
*
-------
ZoO
@ Va.VA I..IGNI"~ FEED
A ~USJ
-------
Table 2. Use of auxiliary fuel during run 26
Char Loss From Gasifier
Char Withdrawal
Char Loss Overhead
Total
Auxiliary Fuel Char
Lbs/hr
Heat of Combustion
MM Btu/hr
75
112
0.828
1.006
1.834
100
1.120
Net Heat Loss As Char = 1.834 - 1.120 = .714 MM Btu/Hr
Table 3. Log of operations
runs 26, 27, and 28
Run No.
Days
Integrated Plant Operation at Process TemperatUre
Lign ite Feed
Dolomite Makeup
Limestone Makeup
All Heat Supplied by Acceptor
All Dead-Burned Dolomite Replaced
Gasifier With 100 Percent Steam
Steady-State Plant Operation
26 27 28
6.0 10.1 12.0
5.3 9.8 11.3
5.5 9.0
11.2
2.7 2.7 9.6
.5 4.0 5.0
o 1.0 3.8
o 0 3.0
truly steady-state operation during Runs 26 and 27. In
Run 26 the acceptor activity increased slowly and, as a
resu It, the acceptor circulation had to be reduced. In
Run 27 minor upsets resu Ited in the use of a small
amount of air in the gasifier between the two air-free
periods. Truly steady-state operation was achieved dur-
ing Run 28. The plant operated for 12 days at process
temperatures, including nearly 10 days without air in the
gasifier and 3 days at steady-state.
Other major achievements of Runs 27 and 28 in-
clude the demonstration of gasifier operation in the
absence of recycle gas (Runs 27 and 28), the perform-
ance of a smooth, orderly shutdown (Run 27), and the
use of limestone as acceptor (Run 28). Also, a compre-
hensive set of solids and liquid samples were collected
during smooth periods of Runs 27 and 28 for detailed
analysis including trace elements.
In all runs prior to Run 27, some recycle gas was
mixed with the gasifier inlet steam. Early laboratory
work had shown that the acceptor will agglomerate at
steam partial pressures of 13 atmospheres or greater;
therefore, some caution has been exercised in demon-
strating the no recycle operation. The gasifier has now
been operated without recycle gas in the gasifier inlet gas
for 1 day during Run 27 and for nearly 4 days during
Run 28. There have been no problems attributable to
operation without recycle gas.
As shown in table 1, Run 27 was terminated when
the weld failed on the butterfly valve which controls the
rate of regenerator fuel char flow from the gasifier to the
regenerator. When this problem was recognized, the
plant was shut down in a smooth, orderly fashion. Sub-
sequent inspection of the gasifier and regenerator re-
actors showed that no deposits had formed after 10 days
of operation at temperature.
Since gasification of lignite was first begun during
87
-------
Run 19, Tymochtee dolomite from Ohio has been the
acceptor used. A principal purpose of Run 28 was to test
the Minnekahta limestone which is quarried at Rapid
City as the COz acceptor. The geographic location of the
Minnekahta stone is such that it can be considered as a
potentially economical acceptor in a commerical plant.
Also, since it is a limestone, it can be reconstituted using
. . - .
a process which has been demonstrated by Conoco on a
bench scale. The experience of Run 28 in which lime-
stone makeup was fed for 11 days proved this limestone
to be a satisfactory COz acceptor. There were no prob-
lems which could be attributed to the use of limestone.
The measured activity of the equilibrium limestone
acceptor inventory was quite similar to that observed
with the Tymochtee dolomite when expressed on a
weight of COz accepted per unit weight of stone basis.
Sets of samples representing all of the major solids
and liquid streams were obtained during a smooth period
at the end of Run 27 and also during the steady-state
period of Run 28. Each of these samples will be
thoroughly analyzed by the South Dakota School of
Mines. The concentrations of trace elements will be
determined for all samples. Although the analyses are
not yet complete, early results are encouraging. The
gasifier quench water was found to contain no detect-
able phenols or cyanides.
DETAILED RUN DATA (RUN 26)
Although a true steady-state condition was not
achieved during Run 26, the changes occurring were very
slow and unidirectional. The slow trend towards higher
acceptor activity and lower acceptor circulation has no
discernible effect on the I ignite feed rate, the regenerator
duty, or the yield and composition of the gasifier pro-
duct gas. Neither the detailed data from Run 27 nor the
steady-state operation data from Run 28 are presented
since analyses of the key solids streams have not yet
been completed.
Heat and Material Balances
A detailed heat and material balance representing a
smooth period at the end of Run 26 is shown in table 4.
Since there is no direct measurement of either the fuel
char rate or the acceptor circulation rate, detailed plant
results are calculated by the simultaneous solution of the
heat balance and elemental balances. The heat balance
for Run 26 is summarized in table 5.
During Run 26, 2,450 Ib/hr of Velva lignite were
fed to the gasifier and 56,400 SCFH of product gas were
produced. As shown in table 4, the heat of combustion
of the product gas was 76.9 percent of that of the lignite
feed. This is despite the inefficiencies inherent in pilot
plant operation which include:
1. Large plant heat losses.
2. Char losses from the gasifier in excess of those com-
pensated for by the use of auxiliary fuel.
3. Large sensible heat duty from the use of a large
amount of regenerator recycle gas.
The data from Run 26 confirm the excellent carbon
utilization of the COz Acceptor Process. The ash exiting
from the regenerator contains only 0.65 percent of the
carbon in the lignite feed. The gasifier cyclone loss and
the char withdrawal loss account for 7.5 percent of the
carbon in the lignite feed. However, the addition of the
carbon in the auxiliary fuel (to simulate partial char re-
covery) reduced the net carbon loss to 3 percent of the
lignite feed. High carbon utilization has been demon-
strated in all of the data producing runs made in the
pilot plant.
In Run 26, as in all of the runs made with no air
input to the gasifier, the ratio of Hz to CO in the gasifier
product gas was in excess of the 3: 1 requirement for
methanation. In Run 26, the ratio was 3.8: 1. As a result,
all of the CO and about 34 percent of the COz could be
converted to CH4. Gas cleanup would also be simplified
by the absence of hydrocarbons other than methane.
Only removal of NH3, HzS, and part of the COz are
required prior to methanation.
The data obtained from steady periods of plant
operation have been consistent with projections made
from the bench-scale data and those projected for use in
commercial designs.
Pressure Balance Relationships
Since Run 16, over 2300 hours of acceptor circula-
tion at process temperatures have been logged. The
ability to circulate acceptor and to feed fuel char from
the gasifier to the regenerator is dependent upon main-
taining a system pressure balance. Figure 3 shows the
static pressures at 15 key points in the reactor-standleg-
lift line system. These data correspond to the conditions
shown in table 4 for Run 26.
EXPERIMENTAL PROGRAM
The immediate experimental goals in the Rapid City
pilot plant are as follows:
1. Operation of the packed tube methanation unit.
2. Testing of possible turbine blade materials for use in
power recovery from the regenerator flue gas
stream.
3. Testing the use of two additional lignites. Feed-
stocks having higher sodium contents than Velva
lignite.
4. Testing the use of subbituminous coal as gasifier
feedstock.
88
-------
Basis:
Datum:
1 hour
1120 (I) at 600p
Input
Lignitc(l)
Acceptor(8)
Boot Gas
Recycle (9)
StCBIII.
Ring Gas
Steam
Purge
Recycle (9)
Inert(l1)
N2
Subtotal
lieats of Reaction
TO~:~+C02==CaC°3
106
43
79
14,558
Table 4.
Heat and material balance, run 26, 2000 hours, June 15, 1975,
system pressure - 150 PSIG
~
M Btu
~ l Sensible Heat Heat of Combustion
Lbs
2,450
8,795
466 14,077
1,573 33,082
1,046 22,000
3,218
540
1,070
30.07 Mois 0 76,200 Stu/Mol
00
CO
Mohture, Wt %
~Iois tun~ free Bas is. Wt \
II
C
N
°
5
As.h (as oxides)
CaS-CaD
C02
623
75
112
9,749
317
1,866
572
1,220
6
18
14,55a
56,369
17 ,295
25,664
69
398
LlGN ITE CIIAR
(1)
o
(2)
o
4.59 0 . 83 0 . 78
66.16 73.29 59.10
1.00 0.36 0.22
20.87 0.00 0.00
0.54 0.00 0.00
6.84 18.70 33.05
0.00 0.07 0.49
0.00 6.75 6.36
400
1840
1430
253.453
3810.364
421.693
2724.700
1505
1855.654
60
9065.863
2291. 273
11357.136
1495
325 . 154
39.143
55.418
1495
3522.511
114.626
1495
1777.457
545.359
2157.906
2.524
17 .100
8557.198
2511.269
45.373
243.295
11357.135
(3)
o
(4)
o
0.00 2.20
7.38 74.30
0.00 0.90
0.00 7.10
0.00 1.50
90.79 14.00
1.83 0.00
0.00 0.00
27825.979
27825.979
6881.688
828.440
1006.129
21404.000
43.978
173.013
30337.248
(5)
3.80
ACCEPTOR
Wt \ (6) (7) (8)
MgO 22.41 36.02 41. 23
CaD 27.01 43.42 49.70
Inert 4.93 7.93 9.08
CO 45.65 12.61 0.00
CaS -CaQ 0.00 0.03 0.00
Activity 0.370
REGENERATOR
M Btu
~ ~ l Sensible Heat ileat of C;ombustion
6881. 688
1119.690
623
100
9,749
600
5,733 75,340
21 440
4,386 50,857
14 297
553
248
I
7,000
2.873
17
Output
Char
To Regencrator(2)
Rej ect (2)
Char Loss From Cyclone(3)
Acccptor( 7)
To Regenerator
Reject
Gas
Product (9)
Recycle(9)
Steam
"25
NU2
Subtotal
Beats of Reaction
lIeat of Combustion (Output-Input ex Acceptor)
CaO+$==CaS+O 0.23 Mols @ 196,540 8tu/Mol
Heat Loss
Total
Input
Char(2)
Auxiliary Fuel(S)
Acceptor
From Gasifier(7)
Makeup (6)
Air(12)
Ac~~~tor Lift (10)
ch~l Lift Gas
Inort Gas (11)
Recycle (10)
"2°
Purge
Recycle(lO)
Inert(1t)
5~total
Heats of Reaction
Heat of Combustion (Output-Input ex Acceptor)
CaS+O==CaO+S 0,03 Mols @ ]96,540 Htu/Mo]
Total
Output
Char( 4)
Acceptor(B)
To Gasifier
Ove rhead
Gas
Flue Gas (0)
Recyc1e(10)
Steam
COS
~t
2
Subtotal
Heats of Reaction
MgC03"MRO+C02
"c~~~a;~ao+C02
Total
165
18
9
22,217
1,916
230
119
143 1840
1840
8,795
49
1840
3,330 96,559
4,799 55,647
97 2,045
I 7
I 6
2 14
22,217
Mol %
01
11/
CO
C02
N2
02
1495
60
325.154
3522.511
1096.142
29.587
1236.221
22 . 485
89.062
39.683
1.087
6361.933
7142.227
4.923
13509.083
65.980
3810.364
21.191
3972.083
2289.096
191. 242
0.269
0.421
0.592
10351. 239
150.040
2421.108
586.712
13509.099
GAS
(10) (II)
0.00 0.00
0.06 0.00
2.20 0.00
29.35 12.00
68.40 88.00
0.00 0.00
1495
60
830
1175
725
60
(9)
13.75
58.80
15.47
9.08
2.9]
0.00
8001.379
150.538
700.537
3.640
4.438
aSY.152
(12)
0.00
0.00
0.00
0.00
79.00
21.00
-------
Table 5. Summarized heat balance
ru n 26
In
Lignite Feed. Heat of Combustion
Out
Heats of Combustion
Gasifier Streams
Product Gas
Net Char (1)
NH3 & H2S
Regenerator Streams
Flue Gas + H2 S + cas
Overhead Ash
Net Heats of Reaction
MgC03 = MgO+C02
CaC03 = CaO+C02
CaO+S = CaS+O
Sensible Heats
Outlet Streams - Inlet Streams
Heat Losses
Gasifier
Regenerator
MM Btu/Hr
%
27.B26
100.00
21.404 76.92
.714 2.57
.217 0.78
.708 2.54
.151 0.54
.150 0.54
.130 0.47
.041 0.15
3.481 12.51
.243 0.87
.587 2.11
27.826 1 00.00
(1 )Net char = Heat of combustion of the "har losses not compensated for by the use of
auxiliary fuel (see table 2).
GASIFIER
313
317
Figure 3. Pressure balance diagram.
operated. The fixed blade turbine test stand is installed
and ready for operation. Testing will begin when a suita-
ble steady-state condition is achieved. Some caution will
be exercised as operation of this unit could easily upset
the gasifier-regenerator pressure balance.
Bench-scale runs have shown that lignite with a very
high sodium content (14 percent of the sulfur free ash)
may form deposits in the gasifier fludized bed and,
therefore, not be useful as feedstock. A run will be made
using Glenharold lignite containing about 7 percent
sodium on the sulfur free ash. This lignite, which is now
stockpiled in Rapid City, has a higher content of sodium
in the ash than most North Dakota lignites. Glenharold
lignite having an even higher sodium content is available
and will be purchased for use in future runs.
Subbituminous coal from Montana has been
purchased and is stockpiled in Rapid City. A run or
series of runs will be made to determine the performance
of this coal as a feedstock for the CO2 Acceptor Process.
90
-------
EFFLUENT CONSIDERATIONS IN COAL GASIFICATION
Frank C. Schara and Donald K. Fleming*
Abstrac t
Several processes are being considered for the con-
version of coal into high-Btu gas. The HIGAS Process,
now being developed at !he pilot-plant level under
ERDA/AGA funding, is the furthest advanced of these
processes and has operated for extended periods on U.S.
coals.
The HIGAS Process is based on the utilization of
high-pressure operation and close control of gas-solids
contacting in three countercurrent stages. These operat-
ing conditions promote the direct production of
methane within the gasifier, maximizing the process
efficiency by utilizing the heat of methane formation
directly within the gasifier, thereby minimizing heat re-
quirements from other sources. Under the conditions of
gasification, high yields of methane are achieved - near-
ly 70 percent of the feed product methane is produced
in the gasifier. The operating conditions of the HIGAS
reactor apparently also inhibit the production of tar and
heavy liquids that are made in moving-bed gasifiers.
Similarly, the conditions of operation tend to produce
simple sulfur compounds, apparently by hydrocracking
of the heavier sulfur complexes found in coal. Therefore,
the HIGAS reactor generally tends to minimize the for-
mation of materials that may be classed as pollutants.
Depending upon the gas processing train prior to final
methanation, these potential pollutants can be reason-
ably well-controlled, and emissions from the facility can
be held to reasonable limits.
The exact quantities of such materials, which would
be produced from various coals under consideration for
gasification, remain to be established in the program.
INTRODUCTION
This discussion has been prepared as an overview on
the environmental aspects of coal gasification. For
several years, the Institute of Gas Technology (IGT) has
been evaluating coal gasification from an environmental
viewpoint, .based both on the overall process concept
and our own specific processes. This discussion presents
some of our findings in this area and also indicates where
environmental data are lacking.
The primary IGT objective in gasification develop-
ment has been directed at the proof of the overall hydro-
gasification concept, with specific emphasis upon gasifier
design and operation. The IGT effort in the environ-
llJ.!!ntal area has necessarily been modest because it is
*The authors are with the I nstitute of Gas Technology,
Chicago, Illinois.
primarily funded in-house. Now that we have successful-
ly demonstrated, in a large pilot plant, the viability of
more efficient, second-generation coal gasification, we
anticipate that additional effort will be expended in de-
fining the possible discharges to the environment.
Magnitude of Energy Problem
Figure 1 presents a projection of possible energy
su pply-demand relationships for the United States
through the next 25 years. The discussion of the data
behind this graph is a long dissertation in itself; we will
only touch upon the highlights here:
The energy demand is based on a 3.7 percent
annual growth rate in the per capita energy re-
quirements, a rate that appears necessary to
sustain our economy.
The energy supply is based upon maximum
conventional utilization of our indigenous
energy resources including -
The maximum projected expansion of
nuclear power generation (without limita-
tions of capital availabil ity or construction
moratoria) .
Maximum increase in the drilling rate for
crude oil and natural gas (without the re-
duction in finding rate that has recently
been experienced).
Maximum increase in direct coal utilization
(assuming transportation and environ-
mental limitations are overcome).
The noncrosshatched area of figure 1 indicates
the energy shortfall that must be eliminated,
either by imports or development of non-
conventional energy resources.
Obviously, we must develop all of our energy resources,
including oil shale, coal conversion and other sources, if
we are to maintain our economic independence.
To make this energy shortfall more meaningful, con-
sider the energy output from one 250 million SCF/day
coal gasification facility. The coal feedstock requirement
is 6-8 times the amount required for a coal-based
ammonia plant of 1000 ton/day capacity. The energy
output is nearly identical to 3000 megawatts-thermal
power generation, but is equal to only about 40,000
bbl/day of oil - 25 such coal plants are required to
equal the million bbl/day goal that has been set in Proj-
ect Independence to reduce our oil imports.
Coal Utilization-Environmental Aspects
We Americans must develop all of our energy
resources and, by far, our largest resource is coal.
91
-------
240
220
200
180
160
~
>.
........
::J 140
+-
!D
10
"0 120
~
<.9
a: 100
w
z
w
80
60
40
20
o
1970
HYDROELECTRIC
AND GEOTHERMAL
SHORTFALL-TO BE
MET BY NEW ENERGY
SOURCES OR IMPORTS
PROJECTED
U.S. ENERGY
CONSUMPTION
COAL
ETROLEUM
~~~~~r~~~~~j~~
1975
1980
1985
YEAR
1990
1995
2000
A75102489
Figure 1. Possible energy supply-demand relationships in the
United States projected over the next 25 years.
Unfortunately, coal is not a "nice" fossil fuel. Signifi-
cant environmental problems must be overcome by an
energy use that is based upon coal. Some of the prob-
lems that must be faced, in one form or another, in any
coal utilization scheme include:
Fate of sulfur present in coal.
Fate of nitrogen in coal.
Fate of other non hydrocarbon material in coal,
including chlorine, trace elements, and, of
course, the large ash disposal problem.
Fate of possible products of incomplete or side
reactions in coal utilization, including carbon
monoxide, tars, oils, soot, or other materials.
Environmental problems related to large-scale
coal mining.
The aesthetic effect of the necessarily large coal
piles required to maintain production.
COMPARISON: ENVIRONMENTAL ASPECTS OF
COAL GASIFICATION AND DIRECT COMBUSTION
The gasification of coal appears to be an environ-
mentally and economically preferred method of coal
utilization, compared to direct combustion, as illustrated
by the estimated data in table 1. (Most of the data pre-
sented in table 1 will be discussed later.) First, we should
list a number of the qualifications that underlie these
data.
The comparison is based upon a net resu Itant pro-
duct of residential thermal energy. We recognize that
92
-------
~----
Table 1. SNG manufacture vs. electricity generation
==-==.==--==;=====-==--:::...~-=-=== -=- -= ==.===-
GAS PLANT
250 X 106SCF/day
(== 3000 MW thermal)
POWER PLANT
3000 MW electric
COAL FEEDSTOCK, tons/day
(12,500 Btu/lb, 4% sulfur)
30,000
Disposal
Ash, tons/day
Lime sludge from stack clean-up
(tons/day CaS03'2HP)
Other sludges
Anticipated emissions
Power plant
S02' tons/day
NO~, tons/day
Particulate, tons/day
Process
Claus tail gas
Sulfur species, tons/day as S
CO2-Vent gas
Sulfur species, tons/day as S
Hydrocarbon, tons/day
Wastewater
Cooling tower blowdown, gpm
Estimated Capital Cost *
'Cost of Product, $/106 Btu
At Plant gate
As residential heat
15,000
1,500
475
e:: 200
5
0-200
0-750
$1 billion
2.59
4.99
(hot air heat
at 75% eff.)
3,000
4,750
45
26
3.75
450
260
37.5
0.75
$2 billion
10.01
15.30
(resistance heat
at 100% eff.)
*Ear!y 1975 basis.
electricity is thermodynamically equivalent to work;
however, as long as a significant sector of energy plan-
ning recommends electrification of our energy base, de-
grading much of that electricity into residential heat, the
comparison is valid. The basis for production of electri-
city is 33 percent net thermal efficiency (including effi-
ciency losses necessary for stack-gas scrubbing). The
stack-gas scrubbing system is assumed to be a first-gener-
ation lime treatment technique, and is assumed to meet
Federal EPA New Source Performance Standards for
sax, NOx' and particulate discharges.
The basis for projection of emissions from high-Btu
gas production is less firm and the data presented in
table 1 are, therefore, only approximations. Not only do
we have significant data gaps in estimating potential
emissions, but the problem is further compounded by
the number of variables that must be considered:
Variety of Coal Feedstocks. Coal is not a
simple material and its composition, including
its potential for formation of possible pollut-
ants, will vary widely, depending on the coal
source.
Variety of Gasifier Configurations. A wide
variety of gasifiers is either available, under de-
velopment, or proposed. Each of these systems
has different operating conditions of tempera-
ture, pressure, number of stages, and gas-solid
contacting mode. Therefore, each of these gasi-
fiers has a different potential for generation of
potential pollutants from the constituents in
93
-------
the raw coal. This is the primary area where
gasification developers, such as IGT, can in-
fluence the overall emissions from the facility.
We believe that our concepts, although develop-
ed for maximum gasification efficiency, will
also result in systems that have reduced overall
emissions of potential pollutants.
Variety of Treatment Processes. In this area,
the engineering contractor integrates the gasi-
fier into an overall system; many options are
available for purification of the raw gas in the
overall facility. The engineering contractor
defines the design philosophy that will deter-
mine most of the overall emissions. Such design
philosophies might include:
Incineration of any hydrocarbon-conmin-
ing stream. We have seen this design philo-
sophy in several proposed systems, al-
though, as we discuss later, the energy cost
is excessive.
Waste water management We have seen de-
signs that recycle water to extinction and
others that propose a liquid discharge
stream from the facility. This point will
also be discussed later.
Minimum control systems with absolute
minimum facility capital cost. We have not
yet seen this design philosophy, but this is
the possible approach that concerns the
EPA in its studies of coal gasification
emissions.
With these qualifications in mind, we can consider the
data of table 1. First, the generation of electricity uses
twice the coal feedstock as the production of an equiva-
lent amount of thermal energy in the form of gas. There-
fore, the generation of electricity for residential heating
will deplete our reserves about twice as rapidly. With
twice the feedstock requirement, electrical power gener-
ation must also dispose of twice the quantity of ash.
By far the greatest source of sulfur emissions from a
typical gasification facility is the on-site boiler house,
even when it releases S02, NOx and particulates in
conformance with Federal New Source Performance
Standards. The quantities of these pollutants that may
reach the atmosphere are a factor of 10 lower in the coal
gasification facility than in the electric power plant.
Other advantages of the coal gasification facility in.
clude a significantly reduced energy cost, both at the
plant (or bus bar) and as actually consumed in the resi-
dence when the inefficiencies of transmission and utiliza-
tion are considered.
On the other side of the ledger, the gas plant has
environmental deficits in some areas compared to elec-
tric generation. For example, the process area of the
gasification plant will emit sulfur that does not have a
direct counterpart in the electric facil ity. We anticipate
that most of this sulfur will be carbonyl sulfide, an odor-
less sulfur species that is considered comparable to or
even environmentally preferred to sulfur dioxide accord-
ing to a recent study for the EPA. Even with the sulfur
emitted from the process area, the overall sulfur
emissions from the gasification plant are still a factor of
7 less than the electric power plant of comparable out-
put.
The gas facility also has a potential for hydrocarbon
emissions. In some combinations of gasifiers and purifi-
cation systems, the hydrocarbon emissions may be quite
low. In some process flow sheets, the expected hydro-
carbon emissions are primarily methane, a nonreactive
hydrocarbon. And, as mentioned before, some engineer-
ing contractors, to play safe, will incinerate any steam
containing hydrocarbons, even at high cost.
Most gasification processes will also have solid dis-
charges of sludges from raw-water and wastewater treat-
ment and the potential of a wastewater discharge from
the facility (although perhaps not to the local water-
shed), depending upon the design philosophy of the
engineering contractor.
On the balance, the production of high-Btu gas
appears to be an environmentally preferred alternative
for the manufacture of clean energy from coal.
TYPICAL COAL GASIFICATION EMISSIONS
Figure 2 presents a block flow diagram of a coal
gasification facility that might be constructed in the
future. Note that the gasifier developer is responsible for
just 1 or 2 of the 20 blocks on this diagram-the gasifier
and perhaps a pretreater. The engineering contractor will
select and integrate the remainder of the subsystems into
an overall facility. Fully half of the blocks on this dia-
gram are shaded, indicating purification processes that
are primarily responsible for the overall environmental
control of the facility. It is obvious that a large fraction
of the overall cost of the facility may be ascribed to
environl11.ental control.
On this block flow diagram, there are three primary
points for gaseous discharge from the plant - the boiler
hOIl"e stack, the Claus plant off-gas, and the CO2 -rich
ga~ .Iat must be vented from the process. Solids disposal
consists of ash, lime-scrubbing sludge, and water-treat-
ment sludges. Additionally, there is potential for an
aqueous liquid discharge stream from the process.
We will now consider how the gasifier configura-
tion-the province of developers such as IGT -can affect
potential emissions in the discharges.
94
-------
"
~h(1\
to
CJ1
FEED COU
..u;~ UP
-UER
D75061326
Figure 2.
Block-flow diagram of a coal gasification facility.
GAS PRODUCT
-------
Surprisingly, the maximum effect is found in the
boiler-house stack. Although the concentrations of pol-
lutants in this gas are regulated, the quantity of fuel
consumed in the boiler house is a direct function of the
efficiency of the process, and the gasifier configuration
has a pronounced effect upon the overall process effi-
ciency. The gasifier configuration also has an influence,
but a lesser effect, upon the constituents that are present
as impurities in the CO2 -vent gas, the composition of the
discharged ash, and contaminants in the possible waste-
water from the facil ity.
Process Efficiency
The primary influence on the overall process effi-
ciency is the configuration of the gasifier and its effect
in favoring certain chemical reactions. I n the overall
process flowsheet, the total chemical reaction is:
2C + 2H2 0 ~ CO2 + CH4 (1 )
Theoretically this overall reaction should be highly
efficient. However, when we mix coal with water,
nothing 'happens. As we increase the temperature to
about 1800° F, we find that the coal and water react,
endothermically, to form carbon monoxide and hydro-
gen (not methane):
C + H2 0 ~ CO + H2 (2)
The heat required to sustain this reaction must be
supplied from an outside source and the mechanism of
this heat supply is one of the primary variables in system
concept and gasifier design.
After the coal has been steam-gasified to carbon
monoxide and hydrogen, the ratio of these two species is
adjusted by converting some of the carbon monoxide
into additional hydrogen:
CO + H2 0 ~ CO2 + H2 (3)
Then, the final product, methane, can be manufactured
from the carbon monoxide and hydrogen:
CO + 3H2 ~ H20 + CH4 (4)
The reaction to 'form methane is exothermic and could,
thermodynamically, offset the heat requ ired for the
steam-carbon reaction. However, the methanation re-
action takes place at lower temperatures (700° -900° F)
and is not useful for supplying the heat required for the
steam-carbon reaction at 1800° F.
Generally, when we follow the chemical reaction
system outlined above, according to the older technolo-
gy that can presently be employed for coal gasification,
the heat required to sustain the steam-carbon gasi'fica-
tion must be added to the gasifier itself at 1800° F or
higher. The most common way to supply this heat is to
add oxygen directly into the gasifier and burn some of
the carbon from the original coal. With the direct utiliza-
tion of steam-oxygen-carbon gasification, the primary
chemical reactions for gasification can be represented in
the flowsheet of figure 3.
The oxygen for the process is separated from air by
power-usually steam power. The steam and oxygen
loads for the process are the primary energy consumers
'for the overall process, as indicated in figure 4. This
diagram was based upon data available for the energy
loads for an older technology process and indicates that
the majority of the steam raised in the boiler house is
used either for process steam or oxygen production. The
total energy demand for this boiler house (for a 250
million SCF/day facility) is equivalent to a 500 meg-
awatt power plant.
One goal of the developers of the newer processes
has been to reduce this energy demand. The technique
used for improvement of overall process efficiency is
called hydrogasification. In this system, the coal is re-
acted directly with hydrogen according to the reaction:
C + 2H2 ~ CH4 (5)
This direct methanation reaction is exothermic and the
heat of this reaction can be used to offset the heat re-
quired by the steam carbon reaction. The hydrogen re-
quired for the direct methanation reaction is produced
by steam-oxygen-carbon gasification. I n the flowsheet of
figure 5, the more reactive constituents of the coal are
hydrogasified and the less reactive carbon in the coal is
used to manufacture the required hydrogen. The result-
ing steam and oxygen requirements for the overall
facil ity are significantly reduced, perhaps to about 300
megawatt equivalent heat load for the boiler house.
Therefore, the boiler house emissions expected for the
older technology processes are perhaps 67 percent higher
than will be anticipated for the newer systems. Referring
back to table 1, the boiler house was the major source of
emissions from the overall process of coal gasification
and the emissions quoted in table 1 were based upon
newer technology, not the higher emissions expected
from older processes.
Claus Plant Off-Gas
The gasifier developer can have only a modest effect
upon the potential pollutants that are present in the
final off-gas from the sulfur plant in the facility. The
engineering contractor will determine the number of
stages for the Claus plant and the process that is used for
tail gas purification. Until specific standards are set for
coal re'fineries, the engineer will probably attempt to
meet those standards that have been set for petroleum
refineries - 300 ppm total sulfur in the undiluted gas.
We emphasize that this standard is based upon petro-
leum refineries where the sulfur concentration in the
initial feed to the sulfur plant will be relatively high. The
standard may be unrealistically low for coal refineries
where high concentrations of carbon dioxide will exist in
the sulfur plant feed. Indeed, for most viable systems of
96
-------
SNG VIA CONVENTIONAL GASIFICATION
COAL WATER-
STEAM- ACID- METHANE
STEAM GAS GAS METHANATION
OXYGEN SHIFT
OXYGEN GASIFICATION PURIFICATION
REACTION
STEAM WATER
co "C"+H20 - CO + Hz CO+H20 ~ C02 SULFUR CO+3H2~
"-I "("+ 02 --- C02 CO 2 + H 2
CH4 + H20
OVERALL REACTION: 2C + 2H20- C02+ CH4
Figure 3. Pipeline gas production using conventional gasification.
-------
POSSIBLE HIGH-PRESSURE STEAM
UTILIZATION IN COAL GASIFICATION
PLANT
ELECTR ICITY
REFRIGERATION
~ ~ RECYCLE COMPRESSION
l l PRODUCT COMPRESSION
MISC. LARGE PUMPS AND COMPRESSORS
PROCESS
STEAM
BOILER HOUSE
HIGH-PRESSURE
STEAM
500 MW
EQUIVALENT
HEAT LOAD
OXYGEN
PRODUCTION
Figure 4. Possible high-pressure steam
utilization in coal gasification.
98
-------
co
co
COAL WATER- ACID -
GAS METH
SHIFT GAS METHMjATION
STEAM REACTION PURIFICATION t
i
CO +H20 ~ t . WATE
r C02 + H 2 C02 SULFUR CO+3H2~
CH4 + H20
STEAJ04 HYDRO- "C"~2H2 ~ CH4
GASIFICATION "C"+ H20 -- CO + H2
OVERALL REACTION:
STEAJ04 2C + 2H20 --COZ+CH4
STEAM- C +- H20 ~ CO ~ H2
OXYGEN OXYGEN C+ 02 ~ C02
GASIFICATION
ANE
R
Figure 5. Pipeline gas production using hydrogasification.
-------
Claus plant tail-gas cleanup, this standard may be ther-
modynamically impossible under many conditions.
Carbon Dioxide Vent
As indicated earlier in the chemistry of coal
conversion, the quantity of carbon dioxide that must be
vented from the coal gasification complex must be even
greater than the quantity of methane that is produced.
For a second generation process, the CO2 -rich vent gas
will total 275 to 300 x 106 SCF/day for a 250 million
cu ft/day gas output. Because of the higher oxygen con-
sumption in processes utilizing the older technology, the
carbon dioxide formation increases and we have seen
designs with up to 400 million SCF/day of CO2 dis-
charqed.
fhis carbon dioxide cannot be pure. Present separa-
tion technology indicates that small concentrations of
other materials must be present in the gas. I n particular,
the separation between some species of sulfur com-
pounds and carbon dioxide cannot be perfect; a signifi-
cant fraction of the organic sulfur species present in the
raw off-gas will be discharged with the carbon dioxide.
Our inquiries indicate that most gasifier developers do
not have hard data on the concentrations of organic
sulfur species in the raw off-gas, nor do they know if it
will be converted to hydrogen sulfide (for easier re-
moval) during the processing prior to the sulfur and CO2
removal steps. Our "calculational" studies, based upon
inputs from acid-gas removal licensors, indicate that per-
haps 1-4 percent of all sulfur in the gasifier feed may
appear in the CO2 off-gas (based only upon theoretical
considerations). Practically, the losses might be higher.
However, we understand that a recent study for EPA
indicated that this sulfur, primarily in the form of car-
bonyl sulfide, is a relatively preferred sulfur species
because it is odorless and noncorrosive.
The raw product gas will also contain hydrogen,
carbon monoxide, as well as hydrocarbons - some of
these must also be lost with the carbon dioxide because,
again, the separation cannot be perfect. The expected
hydrogen and carbon monoxide concentrations will be
low (although the tonnages may be significant because
the total CO2 release is large), but the potential for
hydrocarbon emissions causes the greatest concern. It is
here that we pay some penalty for the improved efficien-
cies discussed earlier. If the process were not designed
for direct methane formation, no hydrocarbons would
exist in this gas stream and hydrocarbon emissions
would be a moot point. However, when we design the
gasifier for direct methane formation, some of that
methane will be lost with the carbon dioxide. Addition-
ally, when we design for direct methane formation,
higher hydrocarbons may be present in the gas stream
and some of these will also be lost. In particular, if C2
and C3 hydrocarbons are present, the processes that now
appear most economical for acid-gas removal will show a
relatively high loss of these hydrocarbons to the carbon
dioxide vent gas. We have seen flowsheets that indicate
up to 200 tons/day of methane or ethane will be present
in this gas stream, as indicated in table 1.
Of even greater importance is the potential for dis-
charge of nonsaturated hydrocarbons. Some processes
tend to generate olefins such as ethylene or propylene.
These species are active in the photochemical oxidation
cycle and they will be largely lost to the carbon dioxide
vent from the system. We at IGT know that the olefin
production in our gasifier is much less than has been
estimated for other systems, but we do not yet have
sufficient operating experience on thE! pilot plant to de-
fine the scale-up production rates for these species.
Some engineering companies have operated under
the philosophy of incineration of any vent stream that
contains hydrocarbons. We believe this philosophy is un-
realistic in the necessary compromise between energy
supply and environmental protection. In one plant de-
sign, the energy required for incineration of the
COrvent gas was 1000 million Btu/hr. If this quantity
of energy were taken from the product gas, 10 percent
of the plant output would be consumed for incineration.
This is enough energy to supply the average fuel re-
quirements for 65,000 homes-or the total residential
energy requirements of a city of one-quarter million. If
feedstock coal is used for the incineration, the inciner-
ation fuel requirement is 25-33 percent of the boiler
house fuel requirement, with a resulting similar increase
in emissions of S02, NOx, and particulates.
Defining a reasonable loss of hydrocarbons to this
vent gas is difficult. We believe that guidelines cannot be
realistically defined until the first of these facilities have
been demo nstrated and hard, steady-state data are
collected on quantities and composition of hydrocarbon
losses. Further, the quantities and types of emissions
from developing processes will be different from those
from older technology and each system must be care-
fully weighed on a cost-benefit analysis.
Solid Waste D!scharge
The flowsheet of figure 2 indicates a single solids
disposal system comprised of wastes from three major
areas - gasifier and boiler house ash, calcium sulfite-
sulfate sludge from stack-gas scrubbing, and sludges from
both raw water and wastewater treatment. Of these
various solids streams, the gasifier developer has a major
influence on only the gasifier ash and perhaps a minor
100
-------
effect upon the sludge from wastewater treatment. The
waste solids from sources other than the gasifier and the
wastewater have been handled elsewhere and no new
problems are anticipated. For that matter, the additional
gasifier ash may help eliminate the usual problems en-
countered with lime sludge disposal. The sludge from
wastewater treatment is a function of the gasifier type
and the treatment processing used; it will be discussed in
the next subsection.
The ash from the gasifier can exist in many forms,
depending upon the characteristics of the coal feedstock
and the operating conditions within the gasifier. In some
cases, this ash will resemble fly ash; in others, the ash
will be a slag. In one of our gasifier configurations, the
ash is a nearly white (or red if the coal contained signifi-
cant pyrite) powder; in another configuration, the ash
resembles small, round, hard pellets. We do not antici-
pate that significant environmental hazards will be found
with the ash because it is the solid material that remains
after extreme treatment. Nevertheless, we do not yet
have hard, long-term data to confirm or deny this
hypothesis.
The only published results that we have seen on this
subject are studies by Peabody Coal Company on ash
produced at the Westfield tests. The results of that work
indicated that gasifier ash could be satisfactorily dis-
posed of in worked-out sections of the original coal
mine.
Wastewater
A fou I water stream will be generated in most gasifi-
cation processes. This stream will result from the
quenching and water-washing of the raw gas to remove
water-soluble impurities such as ammonia, H2 S, and
phenols. The characteristics of this foul water will be
functions of the raw coal composition, the gasifier de-
sign, and the techniques that are used for quenching and
washing the raw gas.
The nitrogen content of the raw coal is expected to
be largely converted to ammonia in many gasifiers. Some
configurations will tend to produce small quantities of
cyanide as well. Much of this cyanide is further con-
verted to thiocyanates in the quench water. Coals with
high oxygen content tend to produce phenolic com-
pounds that may appear in the wash water. Most of the
chlorine in the coal will be gasified to chlorides and will
also appear in the water.
Some gasifier configurations tend to convert a signi-
ficant proportion of the coal into tars and heavy oils.
Depending upon the configuration of the quench
system, some of these higher organic compounds may be
emulsified into the wash water. I n our own high-Btu
gasification system, tars are not produced and heavier
aromatic compounds can be recycled to extinction with
little contact with quench water.
Figure 6 presents a possible water reuse plan for a
hypothetical coal gasification facility. For this 250
million SCF/day plant, we have assumed 6000 gpm of
raw water as typical of a water-conservation plant that
maximizes air cooling. Consumption of water into the
product is 1330 gpm and the cooling tower evaporation
is 3820 gpm. The total net foul water make is 1570 gpm.
If the quantity of phenols made in the gasifier were
high, recovery of the phenols from the foul water would
be economical. A liquid-extraction system would recover
most of the phenols and other organic impurities from
the foul water for byproduct credit. This box is shown
in dashed lines on figure 6 because this recovery is not
necessarily economic with all gasifier types and coals.
The next step in foul water treatment is stripping
for recovery of ammonia and elimination of H2 Sand
CO2 that dissolved with the ammonia. The ammonia is
recovered for byproduct value and the acid-gases would
be directed to the Claus plant for sulfur recovery.
The next step shown in figure 6 is biological oxida-
tion. if phenol recovery is not practiced, this bi-ox stage
is necessary for destruction of organic matter. If phenol
recovery were practiced, the bi-ox could perhaps be
shifted to the smaller, cooling tower blowdown stream.
The waste solids sludge from the biological oxidation
will be a function of the impurities present in the foul
liquor and the degre9 of pretreatment before the sludge
is produced. Note also that biological oxidation is not
the only alternative for this purification. Pressure ther-
mal oxidation may be considered and secondary treat-
ment, such as ozonation, chlorination, or oxygen treat-
ment, may be the better route.
The treated foul water now has sufficient purity
that it may be injected into the cooling water circuit,
together with treated raw water. We have assumed a net
cooling tower blowdown of 900 gpm for this facility.
This blowdown will contain the water-soluble com-
pounds that were not removed earlier. These include
chlorides, perhaps as ammonium chloride, and a portion
of the thiocyanates may still be present.
We have shown a dotted box for concentration as
the next step in the water management. If the design
philosophy of the facil ity is no water discharge, the engi,
neer may use excess low-pressure steam from the process
to concentrate this blowdown stream by a factor of
about 6. The concentrated blowdo\M1 would then be
used for ash quenching within the gasifier (with resultant
steam make into the gasifier) and the final discharged
ash would contain small quantities of water. Under this
philosophy, the ash will contain all of the water-soluble
material in the cooling tower blowdown, including the
101
-------
....
o
I\J
RAW )6000)
WATER
TO
POTABLE )30)
WATER
SYSTEM
CONDENSATE
RETURN
)2870)
BO/LER-
FEED
WATER
TREATMENT
RAW
WATER
TREATMENT
)3100)
)3820)
EVAPORATION
COOLING
TOWER
TURBINE
DRIVES
ETC
BOILER
BLOWDOWN
20
PROCESS
COOLING
;900)
~ BLOWDOWN
EVAPORATION i--- ----:
)740) I CONCEN-I
: TRATION I )160)
I I
L_____-...J
SIMPLIFIED WATER RE -USE PLAN
( HYPOTHETICAL)
~ DEAERATOR LOSS
~
~
ASH
QUENCH
)85)
CO
SHIFT
1740
(CONSUMED
INTO
PROCESS
GAS)
I::=J FLOW RATE IN gpm
WATER OR EQUIV STEAM
REACTION PRODUCT
)410)
GAS
SCRUB
)1570)
ACID
GAS
REMOVAL
1""--- ---..,
I I
I I
I PHENOL I
:RECOVERyl
I I
L___- ---....J
)1570)
-----------1
I
I
I
BIOLOGICAL
)1600) OXIDATION
)30)
FROM
POTABLE
WATER
SYSTEM
Figure 6. Simplified water re-use plan (hypothetical)
STRIP-
PING
METH-
ANATION
-------
chlorides of the coal and salts present in the raw water.
Therefore, under this philosophy, the ash would be sub-
ject to leaching.
The alternative solution is to either evaporate the
cooling tower blowdown to dryness, generating a small,
water-soluble solids stream, or develop disposal tech-
niques for a waste liquid.
At present, we can only estimate the characteristics
of this waste. We do not have hard data on the con-
taminants present in the foul water as a function of coal
feedstock, gasifier operating conditions, and method of
gas quenching and washing. Further, we do not have
quantitative data on the fate of potential pollutants as
this foul water passes through the various purification
stages. Data from a pilot plant are not necessarily appli- ,
cable because the engineering design of a commercial
facility may use different gas processing and water treat-
ment techniques. In short, the quantities and composi-
tions of water-soluble materials that may be present as a
final waste from a coal gasification facility cannot now
be predicted with accuracy.
SUMMARY
Coal gasification appears to be an environmentally
and economically preferred alternative for the gener-
ation of clean fuel from coal to meet residential de-
mands. The gasifier developer has the maximum impact
upon emissions from the overall plant in the areas of
boiler-house emissions and characteristics of the dis-
charged ash. The gasifier will have a lesser impact upon
the quantities of potential pollutants in the discharged
carbon dioxide stream and wastewater; the plant design
engineer will have a greater effect on these streams and
the maximum effect on other potential discharges.
At present, there are data gaps in several areas of
environmental concern; we anticipate that funding will
soon be available to enable us to attack these gaps. With
additional pilot plant development work, we believe that
the environmental effects for large-scale plants can be
estimated with better accuracy than is now possible. The
environmentally related purpose of pilot plant operation
is to develop data on the raw products of the gasifier as
functions of the operating variables and coal feedstocks,
acquiring basic data for the design engineers. However,
the short-term pilot-plant operations, even with engi-
neering evaluation, cannot provide a hard basis for de-
fining future environmental discharges - there are
simply too many unknown factors and variables to allow
precise predictions of large-scale performance on the
basis of pilot plant testing. Rather, the measurement of
the overall system discharges must await the construc-
tion and operation of large-scale, continuously-operated
plants, to evaluate the effectiveness of proposed engi-
neering solutions to environmental control problems.
103
-------
BIGAS PROCESS*
V. D. Kliewer and V. L. Brantt
Abstract
Construction of a plant to pilot the BIGAS Coal
Gasification Process is nearing completion. This plant is
a grass roots plant processing coal through a Coal Prepa-
ration Unit and Gasifier, and carrying the gas through
shift and methanation, producing a high Btu fuel gas
product. An advantage of this process is the flexibility of
the gasifier in that it could be made to handle many feed
stocks and the relatively high yield of methane produced
when directly gasifying coal. Where foreseeable in this
new process, facilities have been provided to handle
vapor emissions and liquid effluents, and to produce a
sweet gas. Actual operation will provide more knowledge
on solving environmental problems of BIGAS plants.
,
Several years ago, Bituminous Coal Research, Incor-
porated, located in Monroeville, Pennsylvania, under-
took the development of a new coal gasification process.
This new process is now referred to as the BIGAS Proc-
ess. Development work by BCR proceeded from batch
type experiments through continuous flow experiments
in a 5 Ib/hr externally heated reactor to operation of a
100 Ib/hr internally fired, Stage II, Process Equipment
Development Unit. Using the correlations obtained in
the Stage II test program, Stearns-R oger, Incorporated,
performed the detailed engineering and construction of a
pilot plant. That pilot plant is located in Homer City,
Pennsylvania and is almost ready to go onstream. Equip-
ment shakedown is now in progress. Management for the
BIGAS Research Program at the plant is under the direc-
tion of the Phillips Petroleum Company of Bartlesville,
Oklahoma. Stearns-Roger, Incorporated, will operate the
plant.
This plant is a part of the high Btu gas development
program supported by the Energy Research and Devel-
opment Administration and the American Gas Associa-
tion.
Well, so much for history. I recall from my school
days that I never did too well in history; therefore, I
must turn to the real subject of this presentation which
is the B IGAS Process itself. Many of you, I am sure, are
*This paper is based on work carried out by Bituminous
Coal Research, Inc. and Stearns-Roger, Inc., with support from
the Energy Research and Development Administration under
contract E(49-18)-1207 and the American Gas Association.
tV. D. Kliewer and V. L. Brant are with Stearns-Roger,
Incorporated, Box 5888, Denver, Colorado 80217.
familiar with the process as it has been the subject of
several presentations at various technical meetings and
references to it have been made in several publications.
On the other hand, I suppose that some of you may not
be familiar with the process. This afternoon, I would like
to briefly discuss the process, point out its current sta-
tus, and briefly review some environmental aspects and
advantages of the process.
Now those who are more professional at speaking
than I will undoubtedly say that it is too much subject
to cover in the brief period that I have. They may be
right. However, to cover the background of many of you
who are here, I feel that I must cover this broad a sub-
ject. Please bear with me if I get into too much detail in
one case or not enough in another. We will allow some
time later for questions.
The pilot plant is designed to process 5 tons/hr of
coal. Figure 1 is a Generalized Block Flow Diagram of
the plant. The plant is a grass roots plant handling coal
beginning with feed preparation, through gasification,
and ultimately to a methanation unit where pipeline
quality gas having a heating value of about 940
Btu/cubic foot is produced. The greatest quantity of
byproducts is slag and sulfur.
Figure 2 shows the Coal Preparation Section. Coal
which will be received by the pilot plant is 1Y:, inch by
zero size. It will be unloaded into large storage bins for
feed into the unit. However, about a 14-day supply will
be stockpiled outside. Briefly, the function of coal prep-
aration section is to prepare coal for feed to the gasifier
by pulverizing it to approximately 70"10 minus 200 mesh.
The coal is wet ground and concentrated into a cake of
approximately 50-65% solids. The cake is then slurried
with water to the consistency used in the process.
The High Pressure Slurry Feeding System is repre-
sented in figure 3. Looking at the lower left, slurry of
the proper consistency from the blend tank shown in the
previous figure is received by the high pressure slurry
pumps. These pumps discharge the slurry at the high
pressure required for the reaction. The slurry is fed to a
spray drier where all of the water is vaporized. Then the
stream is taken to a cyclone where the coal and vapors
are separated. The coal is fed by gravity to the gasifier.
The vapors from the cyclone are scrubbed, the water
condensed, and the vapors heated and returned to the
spray drier to vaporize the water in the slurry. There are
no environmental problems associated with this section
of the plant. The inert gas vapors carrying heat to the
105
-------
o
en
J"$At"'." "",' "", ","
f,Lj};', ", "F
';~TORAG({ "i~~~;TlON~Nf ,
'I , '>"j ,GAS!
t--w~) ""77:7'1' ,
--,--.1L.:..._--- , \--'-~----'.~~-- 1 " : ':,~;, \~~l~L ,', " ,
":) COAL U :!WAAY L eA!3lfl ER L CO UACJD-~!,
L~~EJ'...N~E~~_~~_~~ r--r' ~H !r:, E01R50\1AW:.,.,..,.
-_____n, ',; ~T ClAUG -V<
: '" i ~W ,',.1I'lANT 'ii',\,:",
rl;~.~,.~~, ',' ",'. ", ".,".',:,' :'>'i.:,,~:t~:~~n:}f~;I~
,', ""'" '., ".,' "., """pUkfUR'"
/:"'::> ,'&:t.I-"A' 11 5T"N P'£R UOUO 'p' ,'L'Q~,1b~"'A' N"'r',';J~:U:~:';,~":,,,;,(,,<,:P:\.
., .'" ~ V ~ . \J , "" T' ,r :, . ", f '" " , f .. ,I
;~;:; .-;(',,:;', , 'i', "'., ':,' . ',". ',' " "",' '.,<,', ,'" ",,' ;'~'U;";~";:,;\.~~:~!~(i,;,i;H'1f~~
~ii:::}: ii~r':~; ." '" '. '~,,;,~,,~~W,OI AG RA~,;,iJjN:[:,~ ,'!L;;}i;;'; i~"f:I~~;N~:Mi\:
,~
)
;J
" ~
, >/.,
',}
. -.)!
, '/)
-I
,~1
Figure 1. Bigas 5 ton per hour pilot plant flow diagram.
-------
.....
o
-...J
CAGE COAL \CLA55tFIEQ
MI~~ WATER \ I ' ~! ~HIC!K£NER
![] I FILTRATE \ " C=!:::;~
L__J ' " j - 1- .1 CENTRIfUGE
COAL (AL1:)---, \' '"
I ~
I .~
I ~.
I .j
I }
~ ,.",,"/
tJALL MILL-
FLUX
!- WATER
fJLEND '.' ' IFROM
TANK \ . o(H.~
c~Cy . PUMP5
Tt~~~~J ~
COAL PREPARATION
~
\
'\ PULP
, ,'TANK
. ,->,-J .
. \.-'
Figure 2. Coal preparation.
-------
RECOf\1PRE550R
GA~ WMHER
CYCLONE
-..- -----
.....:.;;:.~
~::::::~~
--......-
CLENO .
PREHEATER
TO 13LENDTANK
-'
o
00
TO 6A IFlfR
2ND ~TA6E-
I1l 5PRAY
'III DRYER
SLURRY
fROM
OlEND TANK
"'7,
1" H.~ 5LURRY
PUMP
Figure 3. High-pressure slurry feed system.
-------
spray drier are in a captive system and are not released.
The water from the gas washer is returned to the feed
preparation section to slurry the coal and is also not
released.
The heart of the BIGAS Process is the two-stage
gasifier shown schematically in figure 4. The gasifier has
three defined sections: Stage II at the top; Stage I in the
middle; and the Slag Quench Zone. Each section is sepa-
rated by formed restrictions. The gasifier processes pul-
verized coal in entrained flow. On the left side of the
figure, coal from the cyclone is fluidized by recycle gas
and conveyed into Stage II of the gasifier. Steam is also
introduced at the same point. There, they combine with
the rising hot synthesis gas from Stage I. The mixing
temperature of about 2200° F is rapidly attained and
the coal is converted to methane, synthesis gas, and char.
This raw gas and char are carried overhead from the
gasifier at about 1700° F. Char in the raw gas is sepa-
rated in a cyclone and returned to Stage I of the Gasi-
fier. Steam required for the reaction carries the char into
the Gasifier. Oxygen is fed through a separate injector
and combines with the char in Stage I. The reaction
temperature reaches 2700° F to 3000° F and the char is
gasified under slagging conditions. The synthesis gas
produced along with any unreacted char pass through
the throat into the second stage.
The Slag Removal System is shown in figure 5. Mol-
ten slag is deposited on the walls of Stage I and flows
out the slag tap opening in the bottom of the conical
section. The molten slag drop5 into a reservoir of water
in the bottom of the gasifier for rapid quenching where
the slag is solidified, and shatters into small pieces. The
slag falls through the water and exits the bottom of the
gasifier, assisted by the recirculating water. The slag falls
into lockhoppers and is ultimately taken as a slurry to a
settling pond. The recirculating water is cooled and re-
turned to the gasifier. A slip stream is taken from the
Recirculating Water System to the settling pond for fines
removal. Clarified water from the settling pond is reused
for filling the slag lock hoppers and for dust and fume
scrubbing as needed.
Figure 6 shows the Raw Gas Cleanup System. The
raw gas leaving the char cyclone passes through a water
wash column, where the gas if further cooled and dust is
removed. The recirculating water is cooled and returned
to the wash tower. Fresh water is added as needed to
remove solids. Purged water is sent to the disposal pond.
The remainder of the process steps are listed in
figure 7. The clean raw gas is partially shifted to provide
the proper H2/CO ratio. It is important to note that the
Shift Unit is also designed to convert all sulfur com-
pounds-such as carbonyl sulfide and carbon desulfide,
to hydrogen sulfide. Thus, no processing nor disposal
problem associated with these carbon sulfides exists.
Hydrogen sulfide and carbon dioxide are removed selec-
tively from the entire gas unit in a SELEXOL Unit. The
H2 S is then converted to sulfur product in a Claus Unit.
The shifted gas is sent to a methanation unit, then re-
turned to the SELEXOL Unit for final CO2 removal to
produce the high Btu pipeline gas.
Figure 8 is a picture of the plant. The plant site is
located approximately 2 miles north of Homer City,
Pennsylvania. The site is bounded by Two Lick Creek on
the west and north, a coal company on the east, and an
industrial part to the south. As you can see, the plant is
virtually complete with only some insulating and paint-
ing remaining. Equipment is being checked out and start-
up of the utility systems and feed preparation section is
anticipated within the next month.
SUMMARY AND CONCLUSIONS
The BIGAS Process, I think, is a good process. I
remember my boyhood days when I had to carry coal
from the coal room to a furnace and then carry the ashes
to the ash pit outside. If the BIGAS Process-and for
that matter any coal gasification process, can eliminate
that, its got to be a good process. Seriously, the Process
offers several advantages. A high yield of methane is
obtained directly from coal, thus reducing oxygen con-
sumption and downstream processing. All types of coal
can be gasified without prior treatment because the coal
is in an entrained system within the gasifier rather than a
fixed or fluid bed system. The rapid heating and reaction
conditions in Stage I! are such that no tars and oils are
formed, thus eliminating that problem. The two-stage
gasifier, being an integral unit, is relatively simple in de-
sign and can be easily scaled-up to larger sizes. The gasi-
fier is also extremely versatile and can accept many feed
stocks. Most any carbon containing feed could be proc-
essed in the gasifier to produce syngas.
Though heavy sludges are not expected to be
formed in this process, it is interesting to note that if
there were sludges, a system could be worked out to mix
that sludge with the coal slurry and charge it to the
gasifier for conversion to syngas.
In the design of the plant, all possible points of
emission have been considered and measures taken to
mitigate those emissions.
All of the waste water streams in this plant are col-
lected in a disposal pond. In this pond, coal fines from
the several fines containing streams wi II settle out and
can ultimately be removed. This process produces a rel-
atively small amount of sour water. However, those
waters wh ich are produced are flashed in a low pressu re
109
-------
.....
....
o
<3A5 + CHAg. ~
~~~~"'.""'"'~~':.........~",
COAL [:OO"F
FROM .~i' j"
CYCLONE P!',J 5TAGE II
, ' ;,,;.: l ~
STEAUM f, ':, " Lt, " ",is] i ?TEAM JETS
"~'~7J
./'/ :.is' :
\;/#i:::~}:~Jk:, Q-rA ~E I
/ ~;;, j",,'~, ROXYG£"'-
[ .. 'e.'
RECYCLE ~~''1--SLAG QUENCH
GA9 f'::~~
~J"---WATER
Figure 4. Gasifier ~ystem.
'''''~~'''''''*'I
.~,~
{, CYCLONE
1~~- : .
CHAR
5TEAtvi
I- ~
~;, "If
GABI FI ER
SYSTEM
-------
o LAG
f)REAKER
SLAG HEATING tJURNER
FRESH
WATER
....
-'
....
T.\I. CAMERA
ANt) Of)5!
PORT
FLOATING
CHAR
DRAWOFF
c.w.
WEIGHING
LOCK HOPPER
RBJtRCULATI NG
~:':"':;:, WATER
tj;~ ;~),"~J
TO L. R FLA51-t
AND D'SP05AL
'R(
TO 6ETTUNG',;J.
POND
SLAG REMOVAL
Figure 5. Slag removal.
-------
.....
.....
N
t3PEC'AL TREATMENT
'8ECTION
..
RAW GA'3
FROM CYOLONE
- ..
CLEAN <3A6 TO
' .Of";; 7''' ,'-' "... ",.- ~t~,.."".~ ~ '.f~-- r '-v--~
0HIFT gECTION
FRE?H WATER
COOLER
I
--GA5 WA'3HER
TO LF: FLASH
AND D'~P05AL
OIL AND FLOATING RECIRQULAT'NG
Cf.4AR DRAWOFr WATER
RAW GAS CLEANUP
Figure 6. Raw gas cleanup,
-------
f" '
Jt:':6UU=UR- ;
~IWrtOVALJ
,b=1
t ~O" "
f' ''-''I ~' ,'1
tREMQVA1, 'J
I he
I
GAS CONVfR510~GH ~TU 5NG ~
....
....
w
RAW GA ).!', 6HIFT' ,;;
FROM G1s COHValOObi
WASHING
TO THERMAL
OXIDIZER
t,.' ~
r,' ~U1 C, U' " ',;1
,P: .....' R ,~
h". 'm' "I'" ,:J
~ ,c,
C
SULFUR
I I
~~~
.=--....,"""""-,~
Figure 7 G
. as conversion.
-------
....
....
~
t'-
Figure 8. BIGAS plant.
~~
~
,,,,'.>1"....:::
'?'- :'
~
II
.
!!J
'i1
I III
"ill
~~
'I!'
~
~..
..,.
..~
-------
flash drum are disposed of in a thermal oxidizer. The
disposal pond itself is concrete, thus there will be no
seepage of the waste waters into the ground water sys-
tem. The waters from the disposal pond are processed in
a secondary treatment sewage disposal plant. The coal
storage pile itself is inside a curbed area. Rainwater run-
off is collected and pumped to the disposal pond.
In the coal preparation section, coal is pulverized in
a wet system minimizing dust emissions. Where dry coal
is handled, all conveyors are covered and processing
occurs in closed vessels or containers where dust can be
collected. At points of possible dusting, a dust
suppression liquid is sprayed on the coal. Where there
may be dust in vent gases, they are passed through bag
filters prior to discharge into the atmosphere.
One problem, in my opinion-especially from com-
mercially sized coal gasification units, whether it be the
BIGAS Process or other processes, is disposal of the large
amount of ash or slag formed. In the BIGAS Process,
however, this slag has a glass-like finish and is rather
impervious. It is essentially inert and should cause no
odor problems.
115
-------
THE WINKLER PROCESS
A ROUTE TO CLEAN FUEL FROM COAL
Abstract
I. N. Banchik*
The Winkler Process is a fluid bed, coal gasification
process and has been in successful commercial service
since its discovery by Dr. Fritz Winkler in the early
1920's.
Davy Powergas is currently updating the technology
to adapt it to the conditions existing in this country and
to make it more economically attractive. This paper de-
scribes the technical features of the current version of
the process and discusses many of its effects' on the en-
vironment. In addition, the economics of a large scale
fuel gas from coal facility are presented.
The first commercial plant utilizing the Winkler Pro-
cess was built in 1926 at Leuna, in what is now East
Germany and, since that time, 36 gasifiers have operated
successfully on a commercial scale. These facilities--all
designed, engineered, and commissioned by what is now
DPG GmbH, have been used to produce fuel gas and
synthesis gas for the production of ammonia, methanol,
and gasoline when coal was the only raw material avail-
able. To date, a number of these units are still in contin-
uous operation and may be visited, if desired. All of
these units have operated at such a pressure as to deliver
the gas at 1 psig after cooling and particulate removal;
historically, the required gasifier pressure was in the
order of 1.5 Atmospheres Absolute.
Several years ago, Davy Powergas undertook an
analysis of the applicability of this time-proven Winkler
technology to the U.S. market and found that although
it was technically sound, changes would have to be made
to improve the overall economics of the process. It be-
came rather obvious that the process' greatest deficit was
the low pressure of operation since the effects of low
pressure are threefold:
1. The large size equipment necessary to handle
the large volume of gas.
2. The high capital cost attributable to a gas com-
pression station.
3. The high daily operating cost for the gas com-
pression station.
Proceeding on a course to increase the pressure of oper-
ation, laboratory tests were performed to verify that the
*1. N. Banchik is with Davy Powergas, Inc., Lakeland, Florida.
same fluidization characteristics which are present at
atmospheric pressure could be maintained at elevated
pressure. In the opinion of the experts, it was expected
that at elevated pressures the thermodynamics and kinet-
ics of the process would be as good or better than at
atmospheric pressure; hence, an analysis of the mechan-
ics of operating a Winkler system under pressure was
begun. It was found that there are available several com-
mercially proven pressurized feed systems operating
under similar conditions, i.e., pressure blast fu rnaces,
Lurgi gasifiers, etc., and these form the basis for the
design of the pressurized Winkler system described be-
low.
A unit designed to produce fuel gas from coal is
shown in simplified form in figure 1. The plant may be
divided into the following operations: (1) coal prepara-
tion, (2) gasification and char handling, (3) compression,
(4) desulfurization, and (5) sulfur recovery.
Coal Preparation
The feed preparation section can include any or all
of the following: crushing to the required size, drying,
and transfer to the low pressure feed bunkers. Normally,
the drying step is only required if surface moisture exists
on the coal such that plugging of bins and conveyors will
occur; indications are that this level may exceed 300fo
with some lign ites. So long as the coal is surface dry, the
decision as to whether to dry will be made on a purely
economic basis.
Run of pile coal from the storage pile is conveyed to
the cage mill crushers where the particle size is reduced
to 3/8" x 0". If predrying of the coal is deemed neces-
sary, or desirable, fluid bed dryers would be utilized to
reduce the moisture content to the desired level. In this
unit, hot air-heated by the combustion of coal, is used to
fluidize the feed material and provide the heat necessary
for drying. Most of the dried coal is removed directly
from the fluid bed; however, a portion is entrained in
the gases leaving the dryer. This entrained coal, re-
covered in cyclones, is returned to the dried gasifier feed
and conveyed to the gasification section. The hot gases
from the cyclone are scrubbed with water for particulate
removal prior to disposal.
Gasification and Char Handling (figure 2)
Coal from the low pressure feed bunker passes by
means of lock hoppers into the high pressure bunker. At
117
-------
801LE" FEEO WATER
~ HI' Ill' STUll "OR
"lAIIT REOUIREIlEIIU
...
.....
en
I
, -
I
.......... .........
AS R£C:O
COAL
~--_._--
COAL
GR-
ANO
DRYING
NlTROnR
A'R OR
OXYGEN
----~-----.
Figure 1.
G...,.'CAT-
ANO
CtlAR HAHOlING
R"'" GAS
FIGURE 1
.,..,.,..
DOlT CHAR.
fUEL GAS
COW'R£SS'ON
I.'. ..........,
CONDENSATE
CARBOHTl
SULFIDE
HYDROLYSIS
PROOtICT ,.lIt:lOA5
AlUIIO
OESULFlJRllAT-
CHAR SlUOGE
co.\l SlUOGE
.11 011
OIITOEN
SUI.'UII
Fuel gas from coal simplified block flow diagram.
:..,~_...:
-------
......
......
to
AIIC(M'Rt:5S011S
---<
l' fRDllMf1
ASHCOffV('I'OfI
"'ASIt KA' Il[COV£RY J1A'~
cyo.Ot«S
ASH CONYfYf.';
ifII(T ~Clltlfll£l
lOC(fl}f>PUlS
t;ASITlflf
ASH "NUl
H" mOlll«£R
1IO'''''''locr;~
FttDSCIt(W';
Figure 2. Winkler coal gasifier process schematic.
-------
the bottom of the high pressure bunkers, variable speed
screw conveyors control the flow of coal into the
Winkler generator. The lock hoppers consist of two small
chambers with valves located at the top of each
chamber. Depending upon the coal feed rate, two or
three parallel lock hopper assemblies are provided: one
or two in service and one spare. A nitrogen atmosphere
is maintained in the system and the pressure in the high
pressure bunker is slightly above that in the generator to
prevent the backup of wet gases resulting in plugging of
the feed system and generator shutdown.
The bulk of the coal in the gasifier is maintained in
a fluid bed operating under pressure. A mixture of steam
and air or oxygen is injected at several points within the
bed to gasify the coal, while steam alone is injected at
the lowest level to fluidize the coal and cool the larger
ash particles discharging from the bottom of the gasifier.
Air is used as the oxidant when producing low Btu fuel
gas while for the production of medium Btu fuel gas, or
synthesis gas for ammonia or methanol, oxygen would
be required. The high bed temperatures, typically
1700° -2200° F, are obtained by partial combustion of
the coal's carbon and contained hydrocarbons. Due to
the relatively high gasification temperatures, the tars,
gaseous hydrocarbons, and carbon present in the coal are
converted to carbon monoxide, hydrogen, and carbon
dioxide. Only a small percentage of methane remains in
the raw product.
The primary coal gasification reactions are:
C + 02 = C02 (exothermic)
C + Yo02 = CO (exothermic)
C + C02 = 2CO (endothermic)
C + H20 = CO + H2 (endothermic)
C + 2H20 = C02 + 2H2 (endothermic)
At a constant coal feed rate to the gasifier, the ratio
of oxidant and steam to coal is controlled to maintain
the desired bed temperature. Optimum bed temperature
is a compromise between product gas calorific value, car-
bon efficiency, and overall thermal efficiency, but is
limited by the ash-softening temperature. If this temper-
ature is exceeded, the ash may fuse and agglomerate,
thus upsetting the fluidization characteristics of the bed
and possibly plugging the reactor.
As a result of the fluidization, the particles of ash
and their contained carbon are segregated according to
size and specific gravity, i.e., the heavier particles fall
down through the fluidized bed and pass into the ash
discharge unit at the bottom of the generator while the
lighter particles are carried up and out of the bed by the
raw product gas. Approximately 50-75% of the incoming
ash will be entrained in the hot product gases leaving the
top of the generator. The exact quantity entrained with
a given gas velocity is primarily dependent upon the
particle size distribution in the feed coal. Since the
height of the fluidized bed is relatively small compared
to the total height of the generator, the upper or major
portion of the generator is available to perform two
other functions: to further gasify any entrained carbon
particles; and to effect a separation of any heavier solid
material. To aid this further gasification, a portion of the
steam and oxidant is added to the generator near the
upper limit of the fluid bed. This secondary gasification
causes the maximum temperature to occur in the space
above the bed, therefore, any ash particles which may be
exposed to temperatures above their softening point are
those being carried up and out of the gasifier. To prevent
the molten particles from forming deposits on the refrac-
tory and possibly blocking the exit duct, a radiant boiler
is installed in the generator above the gasification zone.
This boiler serves to cool the gas 300°-400° F before it
leaves the generator and thus resolidifies any molten
particles. The higher gasification temperatures made pos-
sible by this feature result in higher gasification effici-
encies than were possible previously and also make possi-
ble the gasification of less active fuels.
Ash containing hot gases leaving the generator pass
through the waste heat recovery train where heat is re-
moved from the gas and particulate by generating the
superheating high pressure steam, and preheating boiler
feed water. An air preheater may be included to increase
the overall thermal efficiency of the process. High pres-
sure steam, thus generated, is normally far in excess of
that required by the process and is therefore available to
drive steam turbines or for export from the plant.
The heavier char particles leaving the bottom of the
gasifier pass through a crusher, exit screw conveyor, and
then through a transfer screw conveyor to the ash bunk-
er. The screw conveyors are jacketed for water cooling
of the char and a crusher has been included to assure a
maximum particle size which will be readily transported
in the pneumatic system. The balance of the char is
carried out of the gasifier in the overhead product gas.
The bulk of this char is removed in the waste heat re-
covery train and then passes to the ash bunker by means
of the transfer screw conveyor. Additional char is re-
moved in a cyclone or multiclones, and reaches the ash
bunker in a similar manner, but has its own transfer
screw. I n combination, the waste heat recovery train and
cyclones are designed to remove at least 85% of the
entrained solids. From the ash bunker, the char is
pneumatically conveyed to a remote location. This
material may be disposed of or used as supplemental fuel
in a solid fuel boiler.
The gas from the cyclone is scrubbed in a venturi
type scrubber to reduce the residual solids to a level of 1
grain/1000 SCF, and a settler is provided to separate the
120
-------
solids from the circulating water loop. Exiting the scrub-
ber, the gases pass through the compression section, if
required, and then to desulfurization.
Desulfurization
Removal of the contained sulfur may be
accomplished by scrubbing the gas with any of a number
of solutions such as amine or hot carbonate. These solu-
tions can be regenerated by stripping to yield a concen-
trated stream of H2S and cas, from which the sulfur
can be recovered by conventional Claus technology.
Alternatively, the H2S may be converted directly to
elemental sulfur in an absorption/oxidation type system
such as Stretford. Normally, however, these systems will
not remove sulfur compounds other than H2S, For this
particular application, the removal of H2S and cas to a
level of 100 ppm, a combination of carbonyl sulfide
hydrolysis and H2S absorption by Alkazid solution was
selected.
Raw product gas from either the scrubber or com-
pression section is heated to reaction temperature and
passed to the cas hydrolysis reactor, where the bulk of
the carbonyl sulfide reacts with the contained water over
a catalyst to form hydrogen sulfide and carbon dioxide.
Leaving the reactor, the gas is cooled and sent to the
Alkazid absorber.
The alkazid system has been selected for this appli-
cation as it exhibits a great selectivity for hydrogen
sulfide in the presence of carbon dioxide. Therefore, car-
bon dioxide is not unnecessarily removed, at great
energy cost, and a concentrated hydrogen sulfide stream
is delivered to the Claus plant for sulfur recovery.
H2S in the gas is absorbed by lean solution in the
absorber and the spent solution is regenerated by steam
stripping in a reboiled stripper. The overhead gases from
the absorber containing less than 100 ppm of total sulfur
are sent to battery limits as product fuel gas.
Sulfur Recovery
Acid gases from the Alkazid regenerator containing
approximately 21% by volume hydrogen sulfide are fed
to the Claus plant reaction furnace where a portion of
the contained sulfur is recovered. Two stages of catalytic
conversion are then employed to recover 92% of the
sulfur entering the plant as hydrogen sulfide. After re-
moval of the bulk of the elemental sulfur, the gas is sent
to an incinerator where all of the residual sulfur and
sulfur compounds are oxidized to sulfur dioxide. So that
the final effluent gas from this unit is in compliance with
air pollution regulations, a Wellman-Lord system has
been included to recover the sulfur dioxide and recycle
it to the Claus plant inlet. I n the event there are other
sulfur bearing streams in the plant, i.e., dryer stacks, coal
fired boi ler stacks, etc., these may be scrubbed in indi-
vidual absorbers, and all of the sulfur may be recovered
in the central chemical plant servicing the Claus plant
tailgas.
In the Wellman-Lord system, the gas is contacted
with a sulfite rich sodium sulfite-bisulfite solution in a
two-stage counterflow absorber. The absorption of
sulfur dioxide converts the sulfite in the solution to
bisulfite and then the bisulfite rich solution is pumped
to the chemical plant for sulfur dioxide recovery. The
bisulfite rich solution is fed to a vacuum evaporator
where sulfur dioxide and water are driven off resulting in
a sulfite rich slurry. The slurry is withdrawn, redissolved
in water condensed from the evaporator overhead, and
returned to the absorber. Concentrated sulfur dioxide
(95 wt.% after water removal) from the evaporator is
compressed and recycied to the Claus plant inlet.
An adequate amount of storage is provided for the
circulating solution before and after the chemical plant
so as to assure continuous operation of the tail gas treat-
ing unit even if it should be necessary to shut down the
chemical plant for minor repairs and clean outs.
Davy Powergas has recently completed a study to
determine the effect of gasifier pressure on the cost to
produce low Btu fuel gas from a western subbituminous
coal having a heating value of 8640 Btu/lb. The plant
being studied had a capacity of 6400 ST/D of coal and
delivered the product gas at 210 psig. Capitalization was
done on a utility basis, i.e., 65% debt, 35% equity, and
the results are shown in figure 3. It is interesting to note
that although product gas compression is required in all
cases, the optimum pressure for gasification occurs
below the desired product gas delivery pressure. It
should also be noted that several conservative assump-
tions were made throughout this study concerning the
effect of pressure on gasifier throughput, and it is our
opinion that results actually obtained when operating
under elevated pressures will be still better than those
shown in figure 3.
- With respect to the environmental effects of this or
any other coal gasification process, it becomes obvious
that those plants which had heretofore been built and
operated were never erected in what could be considered
an environmentally pristine area. I n fact, in almost all
cases, the effects on the environment are of no concern
at all; hence, the lack of detailed information on effluent
streams. We recently asked the owners of one of the still
operating Winkler facilities for permission to conduct an
effluent survey and analysis and permission was denied.
The reason for the denial was that if nobody knew what
was being emitted from the plant, there could be no
legislation against it.
121
-------
2.50
COAL COST
sin
15
12
:;)
..
.. 2.00
2: 9
2:
~
..:
OIl
0
U
OIl 6
C
0
..
III
:;) 1.S0
Ii.
3
1.00
FIXED CHARGES
(BASED ON UTILITY FINANCING)
o
50
100
GASIFIER
PRESSURE, PS I A
Figure 3. Low But fuel gas from coal by the Winkler process.
Delivery pressljre 210 PSIG
122
150
-------
Problems attributable to coal piles are common to
all processes and, consequently, will be omitted from
these discussions. Let it suffice to say that the pile re-
quired for this process is relatively small since the pro-
cess is extremely flexible in both capacity and coal
quality. It can be operated on the quality of coal receiv-
ed without the necessity for blending. As an example,
when the plant in Kutahya, Turkey began operation, the
coal feed contained approximately 20% ash; the current
coal supply contains in excess of 40%.
The 3/8" x 0" particle size requirements for Winkler
feed eliminates the noise and dust problems normally
associated with grinding and also eliminates the fines
rejection problem. I n some processes, the fines rejected
represent 30% of the total plant throughput.
The major effluents from the gasification and char
handling section is the char, wet and dry, as the system
is enclosed and will contain the dust and gas. Dry char
rejected from the system contains unconverted carbon
and it therefore would be desirable to recover the heat-
ing value by using it as supplemental fuel in a coal fired
boiler, as is done at the Winkler unit in Turkey. To do
this in the United States would probably require some
emission control equipment, but it should be noted that
the sulfur level in the char will be far lower than in the
raw coal. Although the char exhibits properties of activa-
ted carbon, it is relatively inert with respect to the at-
mosphere and, therefore, could be used as landfill if the
carbon loss could be economically tolerated. Dusting of
the dry char could be controlled by partial wetting.
Many contaminants normally expected in gases pro-
duced from coal such as ammonia, cyanides, etc. would
be destroyed, to a large extent, by the high temperature
of gasification; and the remainder would probably be
adsorbed on the active fly ash as the gas is cooled. Un-
fortunately, trace element analyses are not available and
this subject must be held in abeyance until more data is
collected.
Byproduct sulfur produced in the sulfur recovery
unit could most probably be sold and thus eliminate the
disposal problem; however, depending upon the sulfur
removal system selected, there may be small amounts of
chemical degradation products to be handled.
The other major effluent problems to be considerad
are cooling tower blowdown and water treatment plant
effluent. Siting of the facility may make the disposal of
these streams difficult or easy; but under any circum-
stances, they must be handled and will be present in
almost any chemical or petrochemical plant, irrespective
of the raw material.
123
-------
PRESSURIZED, STIRRED, FIXED-BED GASIFICATION
D. W. Gillmore and A. J. Liberatore*
Abstract
Morgantown Energy Research Center (MERC) of
the Energy Research and Development Administration
(E RDA) has been operating a 42-inch diameter, pressur-
ized, stirred, fixed-bed gasifier to define operating pa-
rameters for various coals and develop technology for
commercial scale. Recently, emphasis has been focused
on the detailed analysis of gaseous and liquid products
and the current program reflects this objective. Some
data are reported on the analysis of tar and gaseous con-
stituents.
INTRODUCTION
Morgantown Energy Research Center has been oper-
ating a pressurized, stirred, fixed-bed gasifier since 1968,
investigating various coals and sizes, and developing tech-
nology and hardware for scale-up to commercial size.
Although numerous reports have been issued regard-
ing operating parameters, yields, heat and material bal-
ances (refs. 1,2,3,4), little attention has been paid until
recently to the minor and trace constituents of the
liquid and gaseous products. Plans are laid to investigate
extensively the constituents in the raw, cleaned and
combusted gases in Fiscal Year 1976. Some data are
available now which are worthy of reporting in this envi-
ronmental symposium.
EXPERIMENTAL
Equipment
The MERC experimental fixed-bed gasifier is a
3.5-foot diameter pressure vessel with provision for stir-
ring deeply within the fuel bed. The fuel bed, 9.6 square
feet cross section and a depth of 6-y:' feet, is supported
on a grate that is rotated continuously. Coal is fed from
lock hoppers into the gasifier while ash and cyclone dust
are removed by lock hopper systems. Figure 1 shows the
entire system.
Air mixed with superheated steam is introduced
below the grate and flows upwards through the descend-
ing fuel. Gas is withdrawn through an offtake in the top
cover which also contains the support and seal for the
*The authors are with the U.S. Energy Research and Devel-
opment Administration, Morgantown Energy Research Center,
Morgantown, West Virginia.
stirrer. Pressure in the gasifier is developed by the resist-
ance to gas flow through orifices. Two fixed-diameter
orifices are used for controlling pressures to 300 psig. A
cyclone separator, located between the two orifices,
removes most of the dust entrained in the gas. The main
gas stream is vented through mufflers and burned in the
atmosphere. A small gas flow is diverted from the main
stream, cooled, metered, and cleaned of dust, tar, and
water before it is analyzed by gas chromatograph for
CO, H2' N2, CO2, H2S, CH4, C2H6, and O2.
The water-cooled stirrer is supported by a radial
thrust bushing, counterbalanced and sealed by a packing
gland. The two lower stirring arms are water-cooled, but
the top leveler arm is not cooled. Horizontal rotation
and vertical motion is controlled within limits by setting
of the variable-speed drive and limit switches. The stirrer
usually is rotated at one-half revolution per minute and
the vertical movement is limited to 2 feet. In practice,
the stirrer passes through the bed in 15 minutes, but this
rate can be slower or faster, depending upon coal proper-
ties. The lowest point reached by the stirrer is usually set
2 feet above the top of the grate, but may reach within 1
foot of the grate.
Nuclear density gages are used to indicate con-
ditions within the pressure vessel. Ash zone, bed level,
and voids in the bed are detected with these gages (ref.
5).
Procedures
Startup is made by igniting kindling wood and char-
coal covered with a layer of anthracite (pea size, 9/16 by
13/16 inch). The grate is protected from overheating
during startup by a covering of ash upon which the fuel
was placed. Anthracite was used during startup to avoid
depositing tar in the cold system. Admission of steam
was started and the air rate increased when combustion
was established, as indicated by rising gas temperatures.
More anthracite was added until full bed depth was
reached. Combustible gas was produced within 1-2 hours
from ignition. Coal feed was changed from anthracite to
bituminous when off-take temperature reached 1,000°
F Approximately 4 hours were required for the anthra-
cite to burn out and be completely replaced by a bitumi-
nous coal bed.
Air supply was set at a selected rate, and a constant
inventory of fuel was maintained by adding coal in small
increments of 250 to 300 pounds whenever a signal from
a nuclear density gage indicated fuel was low. Steam
admission was estimated initially based on the expected
125
-------
,.uc.
-,.
COOliNG _nit
1.\CII
TAC'
'tCLON[
AS.
.""".
J1$cun
,,'GH
PlATf0lt8
Figure 1. ME RC stirred-bed gasification system.
coal rate and known burning characteristics of the fuel.
Weight ratio of steam to coal was close to 0.5, but could
be more or less for a particular coal as adjustments were
made seeking optimum performance and steady opera-
tion.
A data period was started when steady conditions
were attained as shown by a constant coal rate over
several hours. Conditions were not changed during a data
period. Should a change in conditions be necessary, the
data period was terminated and a new period started
when steady conditions were reached again. Samples col-
lected during a data period included dust, tar, and water
entrained in the gas, cyclone dust, and ash. Dust, tar,
and water content of the product gas was calculated
based upon the weights collected from a sample (about
25 scfh) of measured volume taken over a timed interval.
Operating data and analyses for a variety of coals
are given in tables 1 and 2.
Gas Sampling and Analysis
The system in figure 2 was developed to remove and
recover all particulate matter, tar and water from a mea-
sured sample stream end deliver a clean gas to several
process analyzers.
The sample inlet tap is located in a muffler between
the flares and the pressure let-down orifice. The inlet tap
is located flush with the muffler wall with no provision
for isokinetic sampling. The sample offtake pipe is main-
tained higher than 8000 F. to prevent tar condensation
in the high-temperature dust filter. Tar is removed in a
40 kV, steam-traced, electrostatic precipitator. Water
and light oils are condensed with a small water-cooled
heat exchanger. The gas stream is then vacuum pumped
through a series of dry absorbents (silica gel and drierite)
to protect the dry gas meter and pump and to deliver a
clean gas to the process analyzers.
126
-------
Table 1. Analyses of coals, bottom ash, and tars from producer runs
WV
New Mexico Kentucky WV Illinois WV WV Middle
Coal Sub-Bit. 119 Upper Freeport 116 Arkwright Loveridge Kit tanning
Run-Period 62-2 64-1 61-5 39-7 71-3 52-4 45-3
A. Coal
FSI 2 4-1/2 8-1/2 4-1/2 8 8
H20 % 8.8 5.0 1.0 9.2 1.1 6.3 2.0
Ash 24.2 15.4 20.3 8.5 8.3 8.0 8.3
C 50.8 64.5 66.8 66.7 75.9 76.2 79.4
H 5.1 5.5 4.4 4.6 5.7 5.7 5.7
S 1.0 3.9 3.8 1.3 2.7 2.9 1.1
N 1.6 1.7 1.2 1.4
0* 8.5 4.0 3.5 4.9
~ Btu/1b 8900 11450 11580 12000 13675 13755 14000
I\,)
-..J
B. Bottom Ash,
%, (Dry Basis)
Ash 94.6 96.2 92.3 84.3 89.7 77.6 82.5
C 4.1 1.2 6.0 12.9 8.4 20.2 15.9
H .6 .3 1.5 1.2 .2 1.7 1.1
S .2 2.1 .5 1.2 .3 .8 1.1
C. Tar,%
(Dry Basis)
Ash 8.1 11.6 0.0 0.0 .2 5.0
C 77.6 70.7 84.5 86.1 83.2 88.0
H 6.6 7.2 6.9 5.4 6.0 4.0
S 1.3 2.4 2.0 1.9 1.1 1.0
N 1.6 1.7 1.3
Btu/1b 13565 14000 16000 15735 15896 16000
* By difference
-------
Table 2. Operating data and test results from producer runs
New Mexico Kentucky WV Illinois WV WV WV Middle
Coal Sub-Bit. . #9 Upper Freeport 116 Arkwright Loveridge Kittanning
Run-Period 62-2 64-1 61-5 39-7 71-3 52-4 45-3
Pressure (psig) 145 207 119 17 134 90 45
Input, 1b/hr
Coal 1800 1987 1196 1030 1376 1330 866
Air 4583 5708 4033 2350 4994 4419 3084
Stearn 763 987 501 322 635 641 400
Input Ratios, 1b/1b
Stearn:Coa1 .4 .5 .4 .3 .5 .5 .5
Air:Coa1 2.6 2.9 .34 2.3 3.6 3.3 3.6
Output, 1b/hr
Ash 330 243 135 69 101 132 110
Cyclone Dust 86 18 16 28 6 20 10
Gas 5784 8096 5375 3322 6647 5705 3906
II.) Tar 53 21 19 23 29* 52 41
co
Water 889 348 182 242 444 481 283
Gas Yield
MSCFH 77.7 101. 4 74.7 44.2 88.6 83.2 58.6
SCF/LB coal 43.4 54.4 64.7 47.0 65.4 61.1 64.2
Gas Analysis
CO 15.3 18.1 22.5 21.8 19.1 19.8 23.9
CO2 12.7 11.0 7.1 6.9 8.8 9.5 7.6
N2 59.3 54.4 53.6 51.5 55.4 52.0 51.5
H2 10.7 13.3 14.0 17.8 12.8 15.8 14.7
CH4 2.1 2.0 2.5 2.0 2.9 2.9 2.2
C2H6 0.0 .2 0.0 .2 .3 0.0 .2
H2S .2 .9 .3 .2 .5 .2 .1
O2 0.0 .1 0.0 .1
Heating Value
Btu/CF 104 128 142 153 138 144 151
* Average for several periods.
-------
HV POWER
i
STACK
S TE A M -----t
/VVII\I\IIN\A
J\/'oNVVVV\
N\II/V\
VVV'IN\
OUST FILTER
800' F
TAR
PRECIPITATOR
250' F
rco;-l :L
~i'-"J""
VENT
ICOA~
1
LAB
spoor SAMPLE
50' LEVEL
----1---
CONTROL ROOM
TO GAS CIiROMATOGRAPIl
Figure 2. Gas sample system for pressurized gas producer
CO and COz levels are monitored by infrared ana-
lyzers and Hz by a thermo-conductivity analyzer; fluctu-
ations in component concentrations will alert the oper-
ator to malfunctions such as stirrer breakdown, bed void
formation and improper coal feeding. An oxygen ana-
lyzer has been added to detect Oz levels above .1 per-
cent. The presence of oxygen indicates a vacuum leak
usually caused by a dust plug in the sample inlet.
The gas chromatograph is a dual column, thermal-
conductivity type analyzer which records on a 7-minute
cycle the major components, Nz, COz, CO, Hz, CH4.
CzH6' and HzS.
Following a typical test period of 24 hours, the dust
filter, tar trap, and condensed water plus absorbent traps
are weighed; the dust and tar were analyzed for ultimate,
proximate, and heating value analyses. Results of these
analyses are used to obtain overall producer material and
heat balances.
DISCUSSION
Direct measurement of gas production was impossi-
ble because metering devices rapidly clogged with tar
and dust and became inoperative. Gas production was
calculated by obtaining a preliminary value from a mass
balance in which gas output was taken as the difference
between the input (coal + air + steam) and the output
(bottom ash, dust, tar, and water). This preliminary gas
yield was used to calculate balances for carbon, hydro-
gen, nitrogen, oxygen, and sulfur. Values of the output/
input ratio falling between 0.9 and 1.1 were taken as
acceptable data for gas production. Carbon and nitrogen
balances generally gave ratios falling within the desired
range, hydrogen and oxygen balances were less frequent-
ly suitable, and sulfur balances seldom produced suitable
ratios. Gas productions given in the report met the out-
put-input test.
Operating runs for seven coals, given in tables 1 and
2, were selected further to be typical of the respective
fuels based on carbon in the bottom ash, heating value
of the product gas and length of the run. These values
are averages for the period and do not necessarily repre-
sent the optimum for any particular coal. It is recog-
nized that changes have occurred in operating mode and
data collection over the years during which the producer
was operated.
129
-------
Table 3. Elemental analysisll of tar from the gasification of
Western Kentucky no. 9 coal in the Morgantown Producer
(ref. 6)
Pct. of 2/ H:C atomic
total tar Carbon Hydrogen Sulfur Nitrogen o.xygen- Moisture Ash ratio
Whole tar 100..0. 82.1 7.6 1.0. 1.0. 8.3 0. 0..0.7 1.11
Tar fractions 31. 4~j
Soluble in n-pentanell 82.3 7.8 0..8 0..4 8.7 0. 0. 1.13
Insoluble in n-pentanel/ 50. . frY 82.2 6.6 1.6 1.5 8.1 0. trace 0..96
Insoluble in benzene l8.o.Y 71.0. 4.9 1.9 1.5 7.0. 2.2 11.5 0..83
1/ As received.
1/ By difference.
3/ Soluble in benzene.
i/ Moisture free.
For tar analyses, the dirty gas was taken from a
sidestream before the pressure let-down orifice, filtered
at 8000 1,0000 F. to remove dust, cooled to ambient
air temperatures, and a tar-water emulsion was collected
in a series of water-cooled traps. The emulsion was bro-
ken and water was separated by distillation.
Table 3 summarizes data on whole tar as collected
by the University of Kentucky from producer tests in
Morgantown, West Virginia, in 1972 (ref. 6). By infrared
analysis of the asphaltene fraction (benzene-soluble, pen-
tane insoluble) and ultraviolet spectra of a cyclohexane
extract of asphaltenes, 54 compounds were identified
including basic structures such as phenols, phenanthrene,
anthracene, fluoranthene, pyrene, chrysene, picene,
pentacene, dibenzofluorene, and dibenzopyrene. Infra-
red analysis of the residue fraction was weakened by the
presence of ash, but revealed highly aromatic structures
as well as phenolic oxygen compounds.
In 1973, a tar sample was taken from a producer
run with Navajo, New Mexico, subbituminous coal.
Light oil and water were distilled from the tar which was
extracted by a series of solvents. Cyclohexane extraction
gave 39.7 percent soluble (oils) and 60.3 percent insolu-
ble; the cyclohexane insoluble when extracted with ben-
zene showed 22.8 percent benzene-soluble (asphaltenes)
and 37.5 benzene-insoluble on an original tar basis. Suc-
cessive extraction of the cyclohexane-soluble portion
revealed only a trace of tar acids and no tar bases.
Recently a tar sample was collected trom a producer
run with Pittsburgh-seam coal. The tar was stripped of
water and fractionally distilled at 20 mm Hg. The light
oil fractions were combined to represent three boiling
point ranges, 20 to 2400 C, 240 to 3200 C, and 320 to
3400 C and analyzed by gas chromatography; the pitch,
representing 69 percent of the tar, is being analyzed.
Table 4 gives the analyses of the respective fractions
and shows again the wide assortment of chemical species
in the light portion of the tar.
Finally, grab samples of the raw producer gas from a
run with Pittsburgh-seam coal showed cas ranging from
315-350 ppm, NH3 from 529-1,028 ppm and HCN less
than 10 ppm with H2S analyses of 4,500-4,800 ppm.
Table 5 gives analyses of bottom ash from two
coals, showing the excellent agreement between x-ray
and wet chemical results.
SUMMARY
Although considerable data have been amassed on
producer runs with a variety of coals over a range of
operating conditions, providing yields and analyses suit-
able for calculating material and heat balances, little in.
formation has been published on the analyses of tar and
trace constituents in the gases.
This report presents some tar analyses for runs with
three different coals and indicates the quantity of cas
and NH3 in a producer gas from Pittsburgh-seam coal.
Elaborate plans are laid to obtain detailed analyses
on solids, water. tar. and gas from producer runs with
three different coals. This will include isokinetic sam-
pling and dust particle size distribution.
130
-------
Table 4. Analysis of light oil fraction of
producer tar from Pittsburgh-Seam coal
Distillate
Fraction~ °c
320 - 340
20 - 240
240 - 320
% of Tar
3
11
17
Consitituents in light oil,%
Phenol
O-Cresol
M-P-Cresol
Indan
Indene
Xylenols
Naphthalene
2-Methylnaphthalene
l-Methylnaphthalene
Dimethylnaphthalenes
Ethylnaphthalene
Acenaphthene
Dibenzofuran
Fluorene
4-Methyldibenzofuran
3-& 2-Methy1dibenzofuran
Methylf10renes
Phenanthrene
Methylanthrene
F1uoranthenes
Unknowns
4.88
4.18
10.68
4.77
1.89
18.64
20.63
9.28
4.66
1.63
1.23
0.48
0.68
0.14
0.47
3.36
4.30
6.29
5.23
18.06 1.32
5.19
8.88 1.89
6.38 1. 97
5.71 7.34
4.57 7.09
1. 93 7.73
2.84 23.90
0.44 15.94
8.81
0.28
25.05 23.73
17.53
PLANS
It is recognized that gasification of high-sulfur coals
with air and steam is meaningless unless economic
methods are developed to remove H2S and other con-
stituents prior to combustion of the low-Btu fuel gas in
boilers, furnaces, or turbines.
MERC is operating a 16-inch diameter atmospheric
producer (10,000 seth) to study removal of H2S from
hot gas by means of an iron oxide absorber system.
Also, a wet system is being installed, for operating
early in 1976, on a sidestream of the gasifier (30,000
scfh) to water scrub the raw producer gas and to study
combustion of the cleaned gas in a turbine simulator.
In both programs all gaseous, liquid, and solid prod-
ucts will be analyzed extensively for major, minor, and
trace constitutents.
ACKNOWLEDGMENTS
The authors express their thanks to Paul S. Lewis,
Robert V. Rahfuse, and Scott Marshall for their assist-
ance in assembling the data on producer runs and anal-
yses. Special recognition is given to J. R. Comberiati for
performing the tar analyses.
131
-------
Table 5. Analysis of bottom ash from producer runs
Western Kentucky
Coal Pit tsburgh Seam No. 9 (ref. 7)
Anal. Method Wet X-ray Wet
Constituent, %
Si02 44.9 45.0 45.9
A1203 27.3 23.3 18.2
Fe203 17.0 17.1 23.2
CaO 4.8 5.1 6.2
MgO 1.3 2.2 0.9
Na20 1.0 1.6 0.5
K20 1.5 1.3 2.2
P205 .3 .2 0.2
Ti02 1.3 1.3 1.0
S03 1.6
Total 99.4 97.1 99.9
REFERENCES
1. R. V. Rahfuse, A. J. Liberatore, and G. R. Friggens,
"Gasification of Caking-Type Bituminous Coal at 75
I to 150 psig in a Stirred-Bed Producer," U.S. ERDA,
MERCfTPR-75/3, July 1975,11 pp.
2. P. S. Lewis, A. J. Liberatore, R. V. Rahfuse, and G.
R. Friggens, "Bituminous Coal Gasified in a Stirred-
Bed Producer," U.S. ERDA, MERC/RI-75/1, June
1975,16pp.
3. R. V. Rahfuse, G. B. Goff, and A. J. Liberatore,
"Noncaking Coal Gasified in a Stirred-Bed Pro-
ducer," BuMines TRP 77, March 1974,8 pp.
4. P. S. Lewis, A. J. Liberatore, and J. P. McGee,
"Strongly Caking Coal Gasified in a Stirred-Bed Pro-
ducer," BuMines RI 7644, 1972, 11 pp.
5. G. R. Friggens and A. W. Hall, "Nuclear Gages for
Monitoring the Coal bed of Commercial-Scale Pres-
surized Gas Producer," BuMines RI 7793, 1973,9
pp.
6. Third Quarterly Progress Report: A Kentucky Coal
Utilization Research Program, Institute for Mining
and Minerals Research, University of Kentucky,
April 30, 1973, pp. 23-27.
7. Seventh Quarterly Progress Report: A Kentucky
Coal Utilization Research Program, I nstitute for
Mining and Minerals Research, University of
Kentucky, May 1974, pp. 15-19.
132
-------
THE WESTINGHOUSE FLUIDIZED BED COMBINED CYCLE PROCESS:
STATUS OF TECHNOLOGY AND ENVIRONMENTAL CONSIDERATIONS
Abstract
L. A. Salvador. E. J. Vidt, and J. D. Holmgren*
The conversion of coal to a clean, low Btu gas which
can fuel a combined-cycle power generation plant holds
promise of being one of the most economic, efficient,
and environmentally acceptable methods of utilizing our
coal resources to provide the nation's energy needs.
Westinghouse Electric Corporation as leader of a six-
member government/industry team is developing this
power generation system utilizing a multiple-fluidized
bed process for devolatilization, desulfurization and gasi-
fication of coal to produce low-Btu fuel for a combined-
cycle power generation package.
The Westinghouse team is implementing one of the
first totally integrated coal conversion programs, from
bench-scale to a complete coal to electric power energy
conversion package. Begun in 1972, the program is pres-
ently in the design and development phase. Bench-scale
development activities are being performed in conjunc-
tion with the operation of a 1200 pounds per hour proc-
ess development unit (PDU) to evaluate process feasibili-
ty and operability. In addition, a conceptual design for
full-scale generating plan t has been completed.
The lab and pilot scale testing has been directed
toward a number of technical and environmental consid-
erations including the behavior of high-sulfur, caking
coals, ash removal and disposal, sulfur removal and dis-
posal, control of combustion product effluents, turbine
protection, and product gas cleaning. The recent encour-
aging results of these studies indicate that operation of
the system to provide power in an efficient and environ-
mentally acceptable way is feasible.
INTRODUCTION
Putting aside all of the rhetoric of recent years rela-
tive to the existence or nonexistence of an energy crisis,
three facts remain clear:
(1) Energy needs will continue to grow as popula-
tion increases.
(2) The use of natural gas and petroleum, because
of diminishing supplies, will someday decline relative to
coal, nuclear, and other energy sources in supplying our
energy needs.
(3) We must find ways to economically utilize our
*The authors are with Westinghouse Research Laboratories,
Pittsburgh. Pennsylvania.
coal resources compatible with our desire for a clean,
high quality environment.
The work described in this paper grew out of a reali-
zation that the above factors would be important consid-
erations, particularly in the electrical utility industry,
and also that a new type of power plant was needed
which could meet the new restrictions on sulfur oxides.
nitrogen oxides, particulates, and other emissions; wh ich
could burn domestic high-sulfur, caking coals; which
could generate power more economically than present
generating plants at intermediate capacity factors of
30-50 percent; and which could make the most efficient
use of our coal resources. The system concept chosen to
fulfill these criteria is an advanced, multiple fluidized
bed coal gasifier which produces a clean, low-Btu fuel
gas to be burned in a pre-engineered, packaged, com-
bined cycle generating plant composed of both steam
and gas turbines and a waste heat recovery system.
The present goal of the program is the integration
and operation of a gasifier/power-generating plant on the
Public Service of I ndiana system at the D.resser Station
in Terre Haute, Indiana. To achieve this goal, a govern-
ment/industry team was assembled. The team is com-
posed of members with broad interests which function-
ally integrate the project: Amax Coal and Peabody Coal,
coal raw material producers; Bechtel I ncorporated, an
architect-engineer and plant constructor; Public Service
of Indiana, a public utility; Westinghouse Electric, a
producer of electrical generating and transmission equip-
ment; and the Energy Research and Development Ad-
ministration, the sponsoring government agency. Figure
1 illustrates the team membership and functional respon-
sibilities and relationships. Other associate members
sponsoring the program include Northern I ndiana Public
Service, Tennessee Valley Authority, Consumers Power,
Union Electric, Duke Power, New England Electric
System, Columbus and Southern Ohio Electric, Pennsyl-
vania Power and Light, Montana Power, Iowa Power and
Light, and Tampa Electric.
The Westinghouse program is totally integrated.
Tasks and goals are laid out to include a step-by-step
approach from bench-scale laboratory research, pilot
scale development and system design and evaluation to
the operation of the demonstration plant. The latter is
of sufficient capacity to represent a viable commercial
plant. Figure 2 illustrates the task orientation on a time
scale indicating the anticipated full-scale plant operation
in the 1980-1983 time period.
133
-------
....
(.)
~
ERDA - W HEAT TRANSFER DIVISION INDUSTRY PARTNERS
FOSSIL ENERGY Project Management Amax Coal Company
Systems Management Peabody Coal Co.
Overall Conmercialization Public Service Indiana
Bechtel. Inc.
I Westinghouse Electric
I I I / I I
W RESEARCH W GAS TURBINE BECHTEL, INC. PUBLIC SERVICE PEABODY COAL AMAX COAL
DIVISION INDIANA COMPANY COMPANY
Process Design Turbine Plant Design & Generating Coal Behavior Coal Supply &
& Development Development Construction Plant Studies Processing
Operation
I
I
LAB SUPPORT
PROGRAMS
PROCESS
I-- DEVELOPMENT
UNIT
SYSTEMS DESIGN
'--- & EVALUATION
Figure 1. Westinghouse program organization.
-------
ANALYTICAL AND LAB SUPPORT PROGRAMS
PROCESS DEVELOPMENT UNIT (0.6 T / H )
DESIGN CONSTRUCT
OPERATE
SYSTEM DESIGN AND EVALUATION
POWER GENERATION PILOT PLANT
(50T/H)
DESIGN CONSTRUCT OPERATE
...
w
U1
1972
1973 1974
1975 1976 1977 1978 1979 1980 1981
CALENDAR YEAR
1982
Figure 2. Program tasks and goals.
-------
THE PROCESS
Figure 3 is a simplified diagram which shows the
main features of the process. Coal and dolomite are
crushed, sized, and dried prior to being fed to the gasifi-
cation system. Hot fuel gases composed largely of car-
bon monoxide, hydrogen, and nitrogen are cleaned of
particulates and combusted with air in a combustor. The
hot combustion gases are expanded through a gas tur-
bine which produces about 95 MW of electricity. Ex-
haust gases from the turbine are then used to generate
steam in a heat recovery steam generator (H RSG), and
the steam is expanded through a steam turbine which
produces another 40 MW of electricity. Two such
gasifier modules can also be employed with two 95 MW
Westinghouse 501 D gas turbines and one 80 MW steam
turbine to produce 270 MW.
With the exception of the low-Btu gas combustor,
the power generating equipment is more or less standard
equipment of the kind currently fired by natural gas or
fuel oil. The gasifier reactors, on the other hand, are new
concepts currently under development. Figure 4 is a
closer look at the details of these subsystems. The fresh
coal is pneumatically transported and injected into the
devolatilizer reactor where it mixes with 1600° F char
and passes at 20 to 40 feet per second up a "draft tube"
on the center line of the reactor. The high velocity and
dilution with char prevent particles of caking coals from
agglomerating as they become plastic on the surface and
give up their volatile matter.
As the gases are evolved from the coal in the reac-
tor, they pass through a shower of dolomite particles
which remove the sulfur present as hydrogen sulfide.
These sulfided dolomite particles, heavier than the char
formed from the devolatilization of coal, migrate to the
bottom of the fluidized bed along with some char which
recirculates around the draft tube in the "downcomer."
Char and dolomite are discharged separately
through ports at different elevations in the reactor.
Spent dolomite is collected for oxidation to the sulfate
and disposal or alternatively for regeneration and recycle
to the devolatilizer. Char is fed to the gasifier reactor
where it is totally consumed by combustion with air and
by reaction with steam at temperatures from 1900° F to
2100° F. These reactions provideoa hot fluidizing gas
composed of carbon monoxide, nitrogen, and hydrogen
to provide heat to the devolatilizer which is mildly endo-
thermic.
In the process of combusting the char at 2100° F,
the ash reaches its softening point and begins to agglom-
erate. Larger, heavier ash particles become defluidized,
fall from the bed, and are continuously drawn off at the
bottom of the reactor.
Gases from the devolatilizer contain some particu-
lates, but little tar or oils which are cracked to lighter
hydrocarb9ns at 1600° F or above. The particulates are
removed via cyclones and a sand bed filter.
For the generating plant with "once through" utili-
zation of dolomite, an overall material balance is given in
figure 5. With dolomite regeneration and recycle, the
quantities are basically the same except that dolomite
feed is reduced to 3 tons per hour and waste dolomite is
reduced to 2 tons per hour.
CURRENT PROGRAM STATUS
While many of the portions of the Westinghous
process have been tested by others, the feasibility of this
particular combination of subsystems has not yet been
demonstrated. To do this, a well-balanced program of
analysis, lab bench-scale research, and pilot scale devel-
opment is being conducted. As one can see from figure
6, the lab-scale research projects cover the important
areas of coal fluidization and processing, systems
analysis, sulfur removal, gas cleaning, combustion and
turbine expansion. These will be discussed later.
The pilot development is being conducted in a 1200
pounds per hour process development unit (PDU). The
PDU was built in 1973 and 1974, and mechanical com-
pletion was achieved in September 1974. Figures 7 and 8
are two views of the plant and figure 9 shows the control
room from which essentially all operations and data
acquisition in the plant are controlled. Following
mechanical completion, the plant was commissioned in
January 1975 and all of the process and auxiliary sub-
systems were run at design conditions with the excep-
tion of the main reactors. During this period, a number
of design and operational problems were encountered
and solved. Typical of these problems-all of the ball
valves in the plant were removed and repaired to prevent
leaks; cyclone dip legs and fines pneumatic transport
systems were redesigned to affect adequate fine removal;
and cooling water systems were modified to operate
with fines loadings heavier than were expected. I n addi-
tion, two partial oxidation burners which utilize pro-
pane-air combustion with steam and carbon dioxide to
produce a working gas for the devolatil izer were signifi-
cantly redesigned, rebuilt, and successfully put online.
Figure 10 summarizes these early tests and accomplish-
ments with the PDU.
These efforts resulted in the first successful runs
with coal in October 1975. The devolatilizer reactor was
started up with Husky lignite char. Subbituminous coal
was then fed to the reactor for 16 hours and devolatil-
ized. The gas produced consisted of carbon monoxide,
hydrogen, methane, carbon dioxide, water, and nitrogen
136
-------
w
......
~
.- ..~
Figure 3. Generating pilot plant.
-------
Figure 4. Westinghouse fluidized bed gasification process.
and had a heating value of 75 to 135 Btu/SCF. The
exact composition and heating value is not being report-
ed here because of some uncertainties in gas analyzer
outputs which are currently being resolved.
As shown in figure 10, several important process
and operational features were demonstrated during these
initial tests with coal. Coal was continuously fed to the
reactor, fluidized, devolatilized, recirculated around the
draft tube, and discharged through the char removal
systems. Slight particle growth by agglomeration was
noted, and the char particles which formed appear to
have high attrition resistance. Figure 11 shows cross sec-
138
-------
FRESH
DOLOMITE EXHAUST
15 T/H GASES
COAL RAW
47 T/H PRODUCT COMBUSTION
DEVOLATlLlZATION GAS DESULFURIZATION CLEAN AND ELECTRIC
AIR FUEL COMBINED POWER
AND AND CYCLE 135 MW
136 T/H CHAR GAS
GASIFICATION GAS CLEANING 130 BTU/SCF POWER
FINES 190T/H GENERATION
STEAM
21 T/H
SULFIDED
DOLOMITE
AIR
AGGLOMERATED
ASH DOLOMITE
5 T /H OXIDIZER
SULFATED
DOLOMITE
14 T/H
Figure 5. Simplified material balance 50 T/H generating plant with
once-through dolomite utilization.
.
Coal Processing
Flu id ization
Coal Behavior
Sorbent Behavior
Reactor Analysis
Systems Analysis
tions of the char particles produced showing the tenden-
cy for "popcorning" as volatiles are released. Figures 12
and 13 summarize the test conditions and results.
The PDU will continue to be used to evaluate proc-
ess feasibility. In January. tests with dolomite injection
with high-sulfur caking coals will begin. These will be
followed by demonstrations of the adequacy of the
gasifier and, later, the integrated operation of both reac-
tors.
.
Gas Cleaning
Particulate Removal - Hot
Low Temperature Scrubbing
Hot Gas Cool ing
Alkali and Trace Chemical Removal
During this period of PDU testing, other activities
are being directed toward the definition and solution of
technical and environmental problems associated with
the process. Leading this effort is the conceptual design
of a 50 tons per hour plant which has recently been
completed. This is being augment~d by the lab and
analytical studies mentioned above. Because the Westing-
house process was designed to solve environmental prob-
lems, the technical problems and environmental consid-
erations are closely interrelated. I n fact, the criteria for
fuel cleanliness for the gas turbine may exceed the envi-
ronmental criteria. Whereas these studies are by no
means conclusive or complete at this time, the following
brief discussion summarizes the present status of the
process studies relative to several critical concerns.
TECHNICAL!ENVI RONMENTAL CONSIDERATIONS
.
Combustion
Low Btu Gas Characterization
Combustor Development
Emissions Control
.
Turbine Expansion
Tolerance to Chemical Species
Particulate Deposition
Erosion Tolerance
Figure 6. Analytical and lab support programs.
139
-------
-~~
,;
< ",; ~.:.J
, ;. :-1
n ~
~
'" ~
~,
.
Figure 7. PDU process structure & operating building.
140
-------
111
"'Y'
11 ~..
~
~~",:f' ""~ ;."t ~
'~f~
..,'10
'.... '-
~.. '~~'.
(
. .
. 1
:i:
,
-
.....
""'F= '.~
.h- j---------- ~
'-",r . " ".":\1"'..)" -...~ .
",i;. \ ~"-'~ "':'''~::~:~<~~'''''':)i',-.""" ..n-;'-'~'~"'\..":~~ . ~ .
"Ii,.....~...v"'\"\"';"\:':".\'.....v~\...,.,;,,'.t..... ...~...~....~..,..~. .~").\:.. ~ .::.....-
~ ~....H..'''.\.'''> \",-.'" Io."t,,' ~.\AA'~. ~ ...-..'" -:.-
Figure 8. PDU process structure & raw materials storage.
141
-------
Figure 9. PDU control room.
.
Plant construction complete September 1974.
.
All utility and auxiliary systems operated at design
capacities and plant commissioned January 1975.
.
Cold-flow solids feed and discharge and reactor
fluidization studies with char completed May 1975.
.
Synthesis gas generators testing completed Septem-
ber 1975. System redesigned, rebui It,and tested at
design conditions.
.
I nitial test of devolatilizer system with subbitumi-
nous coal successfully completed. Reactor fluidized
at 1600° F subbituminous coal devolatilized, bed
recirculation established, cyclone and char removal
systems, cooling system, recycle gas system, tar-
water waste disposal and thermal oxidation system
operated continuously during 3D-hour run October
1975.
Figure 10. Process development unit accomplishments.
142
-------
~
Figure 11. Devolatilized subbituminous coal.
143
-------
Transport Gas Flow Rate
Inlet Flow Rate to Draft Tube
Inlet Flow Rate to Downcomer
4001b/hr
12001b/hr
12001b/hr
Draft Tube Velocity
Downcomer Velocity
Freeboard Velocity
12-16 ft/sec
0.4-0.8 ft/sec
1.5 ft/sec
Transport Gas Temperature
Inlet Gas Temperature
Reactor Bed Temperature
Reactor Exit Temperature
2000 F
24000 F
16000 F
14000 F
Coal Feed Rate
Char Withdrawal Rate
Fines Removal Rate
Bed Height
3501b/hr
70-150Ib/hr
70-100Ib/hr
22 ft
Total Run Time at T ~ 16000
Total Sub-C Coal Run Time
Total Husky Char Run Time
32 hr
14 hr
18 hr
Figure 12. Summary of test conditions and
resu Its.
Sulfur Removal
The prime design for sulfur removal involves the use
of dolomite at 16000 F in the devolatil izer reactor or,
perhaps, in a second desulfurizer reactor where Hz S is
adsorbed and reacted with the half-calcined stone. Two
alternate schemes being studied are a lower temperature
(12000 F) removal on a fixed bed of iron oxide and
"cold," wet-scrubbing systems. Efforts on the prime
system have centered around defining which selection
criteria are important for dolomites and how well a
number of dolomites available for use in the Terre Haute
area conform to these criteria. Selection criteria include:
(1) Sulfur removal efficiency.
(2) Attrition resistance of the stone.
(3) Alkali metal release.
(4) Regeneration characteristics.
(5) Economic availability.
Figure 14 shows the economically available local
dolomites in and around Indiana. These have been
shown to have adequate sulfur removal characteristics.
With feed rates based on a 1.2-to-1 calcium-to-sulfur
ratio in the reactor and with a reactor volume to give a
1-second residence time, 90 percent sulfur removal is
Inputs Outputs
ib/hr Ib/hr
Coal 340
Downcomer Gas 1500
Draft Tube Gas 1500
Conveying Gas 400
Purges 80
Product Gas 3000
Char 70
Fines 70
Totals 3820 3140
Closure Error 18%
Figure 13. Overall material balance for devola-
tilizer test with subbituminous coal.
feasible. Fluidized bed studies show many stones to have
excellent attrition resistance, notably Geneva, Salo-
monie, and Bonne Terre stones from Indiana and Canaan
stone from Connecticut. Alkali release of potassium and
sodium which causes turbine blade erosion and corrosion
is smallest for Canaan, Geneva, and Salomonie stones.
Regeneration of these stones has been adequately
demonstrated in the laboratory, and it appears that 30
cycles of sulfidation and regeneration may be optimum.
In summary, sulfur removal and other stone criteria
required for the generating plant can be met by locally
available stones if the larger scale PDU tests corroborate
lab studies. Stone disposal and regeneration processes
continue to be a major effort of the program since in the
final analysis, it will be a combination of economic and
environmental values which prescribe the selection of
"once-through" versus "dolomite recycle" plant designs.
Gas Cleaning
Particulate removal is equally important from the
environmental viewpoint and from the viewpoint of tur-
bine protection. There are two cleanup systems under
consideration. The first or preferred method is hot gas
particulate removal at 16000 F by cyclones and sandbed
filters. A second is the use of wet scrubbers at low
temperatures. Lab evaluations of commercially available
cyclones indicate that collection efficiencies are inade-
quate in the 0 to 6 micron range for most conventional
cyclones. A special tangential jet device has on a lab
144
-------
Figure 14. Location 01 dolomite supplies.
145
100
MILES
-------
100
. .
~ 99
0
.
>-
u
z 98
w
-
u
-
LL. 97
LL.
w
z 96
0
I-
u 95
w
...J
...J
0 94
u
0
0
o
(II
o
m
o
t-
o
(()
;t 0
!., 10
it 0
w v-
o
rt)
o
N
o
o
o
132 SCFM Primary Dirty
15 SCFM Secondary Clean @ 33psiO
Overall Efficiency: 80%
2 3 4 5 .6 7 8 9
Dp (\I)
Figure 15. Tan-jet efficiency.
.
.
GAS VELOCITY 35 FPM
BED MATERIAL 20- 30 MESH OTTAWA SAND
BED DEPTH 3"
INLET LOADING 0.3 - 2.50 gml m3
OVERALL EFFICIENCY 97.5- 98.8 %
2
3 4 5 6
AVERAGE PARTICLE
7 8
SIZE ,um
9
10
Figure 16. Dust collection on a granular bed filter.
146
-------
CD
10
- CD
C\I
I
C
.
o
ILl &C)
a..
(J)
.
~
.-: ~
(J)
.
oct
..
0:::
ILl Ii)
m
2
:)
Z
ILl C\I
~
o
:E
(J)
co
FUEL: NO.2 OIL
~
SCH 1- F
o
o
1600
1000
1200
1400
1800
2000
COMBUSTOR EXIT TEMPERATURE, 0 F
Figure 17. The effect of heating value on CO emissions
for low Btu coal gas.
147
-------
o
~
o
CJ)
o
CO
(J) 0
- ~
en -
~
m ~
>- -
~
c 0
o
.
..J
0 ~
>
I 0
CO
.
M 0
o v
z
o
N
o
OIL FUEL
SCHEME I-F
BTU
COAL GAS, 200 SCF
SCHEME I
COAL GAS,130~
CSCHE~E I-F
1000 1200 1400 1600 1800
COMBUSTOR EXIT TEMPERATURE, of
2000
Figure 18. Comparison of NOx emissions from the combustion of
distillate oil, and low heating value coal gas of 130 and 200 Btu/Scf.
148
-------
cno
Ci50
-
a:::
C
..0
'0
..J~
o
>
..
::i
a..
a..
..0
LLlO
Crt)
-
X
o
Z
o
::i0
zO
0C\J
m
a:::
-------
scale, been shown to be capable of achieving good parti-
cle removal efficiencies down to 3 microns as shown in
figure 15. Since the turbines are intolerant of particles
greater than 1 to 3 microns, a secondary cleanup system
will probably be required. Sandbed filters operated hot
at 1600° F appear to have promise as shown by the
recent results of tests shown in figure 16. With further
development, an adequate hot particulate removal sys-
tem should be feasible, but may prove to be more costly
than wet scrubbing.
Combustion
The standard combustion basket for the gas turbine
fired on fuel oil is not adequate for the low-Btu fuel
produced in the gasification process. Current work on
combustor design and evaluation in both small- and full-
scale experiments is directed toward the development of
a design which provides stable combustion at partial and
full-load flow rates while minimizing emissions of nitro-
gen oxides, sulfur oxides, hydrocarbons, carbon mon-
oxide and smoke. Figures 17, 18,and 19 show the results
of recent tests with the full-scale combustor for the
Westinghouse 501 D turbine. NOx' CO, and smoke are
kept within present environmentally acceptable limits.
These tests will be continued with gases formulated to
simulate the low-Btu coal-derived fuel gas and later with
gas produced in the PDU. Of particular concern is the
amount of NH3 in the gas, not yet measured, which
could be converted to additional NOx'
Turbine Expansion
A fundamental concern to all programs utilizing gas
turbines with coal-derived fuel is the erosion, corrosion,
and chemical deposition of materials on turbine blades.
A substantial effort is underway to define turbine toler-
ance to contaminants and particulates. Alkal i metals,
halogens, and particulates greater in size than 1 to 3
microns are potentially damaging. The control methods
being studied include the use of low alkal i-releasing dolo-
mites, addition of chemicals to the fuel to prevent sul-
fate deposition, adequate particulate removal as dis-
cussed previously, and adsorbtion of harmful species
prior to introduction into the turbine. Whereas these
studies are in their early stages, it appears that adequate
turbine protection will be feasible.
Resource Utilization
The low-Btu gasification/combined-cycle plants con-
templated here have a marked advantage in terms of
resource use over conventional fossil plants if predictions
of operating efficiencies are born out in practice. I n ad-
dition, land use is relatively small because the system is
operated at a pressure of 225 psig. For example, the 50
ton5 per hour plant at Terre Haute is estimated to use
only about 15 acres for gasification, power generation,
and solids storage.
Water use is also smaller than other electrical gener-
ating plants and other high-Btu gasification plants as
shown in figure 20. Nuclear reactors typically reject
6500 Btu/kWh to cooling water circuits, with the loss of
nearly 7 pounds of water per kWh of power produced,
due to evaporation from the cooling towers. The best
large central station coal fired power plants reject 5000
Btu/kWh to cooling water circuits so that these plants
lose about 5 pounds of water per kWh of power pro-
duced due to evaporation losses.
The present day integrated gas turbine/steam tur-
bine plants using fuel from a low-Btu gasification system
such as that under development by Westinghouse would
reject only 2000 Btu/kWh to cooling water, resulting in
a water consumption for both gasification and cooling
tower evaporation of less than 3 Ib/kWh. The much
higher efficiency gas turbines of the 1980's with 2600°
F inlet temperatures could reduce the water require-
ments to as low as 2 Ib/kWh in the next decade.
In contrast, coal gasification systems designed to
produce synthetic methane or pipeline gas require on the
order of ten times the water makeup that a low-Btu
gasifier requires.
CONCLUSION
The Westinghouse process promises to be an effi-
cient, economical, and environmentally acceptable proc-
ess for generating electricity from coal. A number of
questions exist concerning technical problems and envi-
ronmental considerations. These are being attached in a
well-balanced program consisting of lab-scale experi-
ments, pilot scale (PDU) development tests, and analyti-
cal studies. The program is still in its early stages, but
results thus far are encouraging and give promise that the
technical, economic, and environmental requirements of
Nuclear electric
Conventional fossil
electric
low Btu gasification,
combined cycle
High Btu gasification
Pounds water Pounds water
per kWh per Ib/coal
7
5
2.3
0.44
5.0
Figure 20. Water utilization estimates.
150
-------
the system can be met in the design of the generating
plant to be built for Public Service of Indiana.
ACKNOWLEDGMENTS
The authors wish to acknowledge the contributions
of coworkers on this program who furnished material for
this paper: D. H. Archer, J. R. Hamm, B. W. Lancaster,
E. F. Sverdrup, R. M. Chamberlin, D. l. Keairns, J. l.
Chen, J. P. Morris, E. P. O'Neill, C. H. Peterson, W. C.
Yang, P. Cherish, l. H. Grob, P. J. Margaritis, l. K.
Rath, R. D. Shah, S. S. Kim, E. J. Chelen, M. J. Arthurs,
J. E. Macko, R. B. Mangold, W. R. Miller, D. E. Rudisill
and S. P. Tendulkar.
151
-------
ROLE OF GASIFIER PROCESS VARIABLES IN EFFLUENT
AND PRODUCT GAS PRODUCTION IN
THESYNTHANEPROCESS
M. J. Massey, D. V. Nakles, A. J. Forney, and W. P. Haynes*
Abstract
During steam-oxygen gasification of coal by the
Synthane Process and others, substantial quantities of
foul condensate and sulfur-bearing char and tar are pro-
duced. Little is known quantitatively either about steady
state rates of production of various gasifier effluents or
about the relative effects of process variables in coal
gasification on the types and rates of their production.
Utilizing the Pittsburgh Energy Research Center's
(PERC) 4-inch diameter Synthane gasifier, a series of 19
tests on North Dakota lignite were conducted to study
quantitatively the nature of effluent production and the
various gasifier process variables which affect it. Process
variables, e.g., fresh coal heatup rate, product gas resi-
dence time, reaction temperature, and the extent of gas-
solid contacting were varied by altering the fresh coal
injection position within the gasifier. Among the major
findings were: (1) effluent production rates, viz., quan-
tities of hydrocarbon condensate and water soluble con-
taminants, vary significantly with time from gasifier
startup to the attainment of steady state operation; (2)
steady state effluent composition and production rates
are affected strongly by changes in coal injection geome-
try; (3) at least two mechanisms appear to be responsible
for observed variations in the levels of production of
various effluents; and (4) modified gasifier design offers
at least a complement and quite possibly an alternative
to large-scale conventional treatment of gasifier efflu-
ents.
INTRODUCTION
During steam-oxygen gasification of coal by the
Synthane Process and others, substantial quantities of
foul condensate and sulfur-bearing char and tar are pro-
duced. Little is known quantitatively either about steady
state rates of production of various gasifier effluents or
about the relative effects of process variables in coal
*Michael J. Massey is Assistant Professor, Chemical Engi-
neering and Public Affairs, Carnegie-Mellon Universtiy, Pitts-
burgh, Pennsylvania. David V. Nakles is Chemical Engineer, Pitts-
burgh Energy Research Center and a graduate student at Carne-
gie-Mellon University. Albert J. Forney is Research Supervisor;
William P. Haynes is Supervisory Chemical Engineer, Pittsburgh
Energy Research Center.
gasification on the types and rates of their production.
Ut i I i zing the Pittsburgh Energy Research Center's
(PE RC) 4-inch diameter Synthane gasifier, the present
experimental program was initiated to study quantita-
tively the nature of effluent production and the various
gasifier process variables which affect it. Included in the
investigation were studies of:
(1) rates of production of various gasifier effluents as a
function of time from gasifier startup to shutdown;
(2) effects of selected process variables on the rate of
production of various effluents;
(3) associated effects of changes in these variables on
yields of total product gas and equivalent methane
(methane plus twice the ethane production); and
(4) impacts of changes in selected process variables on
the composition and physical properties of char and
tar produced. -
To minimize experimental complications, noncaking
North Dakota lignite coal was used exclusively in the 19
gasification tests conducted during this investigation.
Process variables, e.g., fresh coal heatup rate, product gas
residence time, reaction temperature, and the extent of
gas-solid contacting, were varied by altering the fresh
coal injection position, and their impact on effluent pro-
duction was monitored. During any given test, condens-
ible hydrocarbons and contaminated water were sepa-
rated from raw producer gas continuously; at 45-minyte
intervals, accumulated aqueous and hydrocarbon con-
densate were withdrawn from condensers, weighed, and
sampled for analysis. Noncondensible product gas pro-
duction and composition were monitored at 3D-minute
intervals throughout each run. Gasifier char was col-
lected in a batch reservoir during each run and sampled
once for analysis.
EXPERIMENTAL EQUIPMENT AND PROCEDURE
Gasifier Configuration and Effluent Sampling Apparatus
In support of its Synthane process development
program, the Pittsburgh Energy Research Center (PER C)
has since 1969 been conducting coal gasification ex-
periments in a 4-inch diameter fluid bed reactor. A flow
diagram illustrating the conventional equipment train for
these tests is presented in figure 1. Fresh coal is injected
into the gasifier under pressure (4D atm) at the rate of
about 25 Ibs/hr; reactant gases (steam and oxygen) are
introduced at the base of the gasifier. Unreacted solids
153
-------
Feed coal
Gasifier
40 atm
-'
~
Steam
Oxygen
Char
Particulate
trap
Tar 8
condensate
- - - - ...
.. - - .. --
- - - -
- - --
Gas
chromatograph
Product gas
to vent
Tar 8
-- --: condensate
Condenser No.1 Condenser No.2
Figure 1. Basic laboratory scale equipment train for the batchwise
collection of aqueous and liquid hydrocarbon effluents
from the Synthane gasifier.
-------
(char) are discharged from the base of the reactor to a
batch receiver while raw product gas is withdrawn from
the top of the reactor. Following coarse particulate
removal, this gas is cooled in a series of two water-
jacketed condensers where high molecular weight hydro-
carbons and contaminated water are condensed and col-
lected in large batch reservoirs. Prior to venting, the
composition and flowrate of noncondensible gases leav-
ing the second condenser are measured. At the end of
each test, aqueous and hydrocarbon condensates from
batch reservoirs are weighed and sampled. Results of
batch analyses of a large number of gasification tests
with caking and noncaking coals are presented in two
recent PERC reports (refs. 1,2).
As is illustrated in figure 2, the Synthane gasifier is
divided into two sections, an upper carbonization zone
("'6 ft) through which fresh coal free-falls and/or un-
reacted solids disengage from product gases, and a lower
fluidized bed zone ("'6 ft). In the present experimental
program, fresh lignite coal was injected into the gasifier
at one of three locations: (1) the top of the carbonizer
(free-fall injections); (2) approximately 1-1/3 ft into the
fluidized bed (shallow bed-injections); and (3) approxi-
mately 4% ft into the fluidized bed (deep bed-injec-
tions). Although some variation in gasifier temperature
profile with coal injection position occurs (see figure 3).
typical reactor temperatures range from an average of
4000 to 5000 C at the top of the carbonizer to 8000 to
9000 C at the base of the fluidized bed.
To permit periodic sampling of hydrocarbon and
water condensates during the present test program, the
batch condensate collection equipment shown in figure
1 was modified as shown in figure 4. Small reservoirs
were attached to the base of each condenser. At regular
intervals, these reservoirs were pressurized with nitrogen
and filled with accumulated condenser condensate. Sub-
sequently, each reservoir was depressurized and the con-
tents were drained, weighed, and sampled for analysis.
Gases vented during depressurization of these reservoirs
were collected during one run and analyzed to determine
whether condensate fractions, particularly light hydro-
carbons such as benzene were being vaporized and lost.
Mass spectrographic analysis of blowdown gases indicat-
ed the presence of largely N2 and CO2 and essentially no
light hydrocarbons.
Experimental and Analytical Procedure
With the exception of the coal feeding technique,
experimental procedure was standardized as much as
possible from run to run. At startup, N2 was employed
as the initial fluidizing agent and reactor heat was sup-
plied electrically. Gradually, steam and oxygen flows
were increased, nitrogen flow was decreased and reactor
neat requirements were met increasingly by exothermic
gasification reactions. Following an initial accumulation
of solids in the reactor's fluid bed, char extraction rates
were varied to maintain a desired bed height. Bed tem-
peratures were maintained largely by varying the flow-
rate of oxygen to the reactor. Note that both oxygen
flowrate and char extraction rate were varied manually.
A certain amount of the observed fluctuation in gasifier
performance during a run is believed to be related to this
manual control.
In each experiment, a measured quantity of cold
coal was discharged at a known rate from a pressurized
reservoir to the gasifier through a star feeder. In the
free-fall injection runs, coal was allowed to fall by
gravity from the top of the gasifier. In the shallow and
deep bed-injection runs, nitrogen was utilized as a pro-
pellant to transport the coal through a standpipe from
the top of the reactor to the appropriate point of injec-
tion within the fluidized bed.
From reactor startup to shutdown (typically about
6 hours), product gas samples were collected every 30
minutes. Condensates (hydrocarbon and aqueous) were
collected for 1% hours during reactor startup and at 45
minute intervals for the remainder of the run. Product
gas was analyzed routinely by three independent
techniques-on line gas chromatography, mass spectro-
scopy and laboratory gas chromatography-and the
results were averaged. Condensates first were separated
and weighed. Samples of water then were withdrawn and
routinely analyzed for phenol, cyanide, thiocyanate,
chemical oxygen demand, and total organic and inorga-
nic carbon. Ultimate analyses, true boiling point curves
and specific gravities were determined for a limited
number of hydrocarbon condensate samples to charac-
terize tar composition.
Unreacted solids (char) were discharged from the
gasifier through an extractor to a pressurized receiver
where they accumulated throughout a run. At the end of
each run, this receiver was emptied and a batch sample
of the char was collected for ultimate analysis.
Following each gasification test, hydrogen, carbon,
and oxygen mass balances were determined on the basis
of the known composition of the feed coal, measured
steam, oxygen, and nitrogen feedrates, and the weights
and compositions of collected product gas, gasifier char,
and hydrocarbon/water condensates.
EXPERIMENTAL RESULTS
Five free-fall, eight shallow, and six deep bed-injection
gasification tests were conducted on North Dakota
lignite. Relevant operating statistics for each of these
tests are presented in table 1. Time series plots. of
product gas and equivalent methane production rates as
155
-------
Cool feed
T
Carbonization zone
- 6ft high,IO Inches i. d.
Fluidized bed
-6ft high,4 inches Ld.
-900°C
Char
Char
removal system
Product gas and
condensible effluent
Free fall coal injection
point..... 5 ft above the
fluidized bed
3/4"0.0. dip tube
..... 7000C
Shallow bed injection
pOint I 1/3 ft below the
surface of the fluidized
bed
Deep bed -injection
point 4 1/2 ft below the
surface of the fluidized
bed
~ Stea m / oxygen
Figure 2. Basic configuration and coal feed locations of
Synthane gasifier.
156
-------
900
800
o
0
700
w
a::
::)
I- 600
<1:
a::
w
.... 0..
U1
..... ~
w
I-
500
400
300
o
. Free fall injections of North Dakota lignite
-o-Shallow bed- injections \I" \I \I
A Deep bed- injections \I II \I \I
r-- ~~e~ef~~lt~~ne --+-- Fluidized bed zone of reactor
2
3
4
5
6
THERMOCOUPLE
7
NUMBER
8
9
10
II
Figure 3. Variations in gasifier axial temperature profiles with
coal injection location.
-------
Condenser NO.1
No.2
Non ..condens1ble
product gas
( - 130°F )
Condenser
Raw product gas
from gasifier
at 40 atm
----
---
----
- --
=--.= Tar a
- - - - - condensate
=-== Tar 8
- - - - - condensate
- --
- ---
--
0'1
CD
High
pressure
nitrogen
High
pressure
nitrogen
Vent
Trap
No.1
Trap
No.2
Vent
Tar a condensate sample
Figure 4. Apparatus for intermittent sampling of tar and condensates
produced during the gasification of coal.
-------
Table 1. Operating and product gas statistics, series of steam-oxygen
gasification trials: free-fall, shallow and deep bed-injections
of North Dakota Lignite coal
Gasifier Reac: tent F1uld Bed Temp. 0c Re8.~f~~. 0t~~a tln8 t Reac tant Farti.l Coal SCmbP~~~H~~d
Feedra tea. {J Ihr Conversion Anal. vt t
~ Dato ~ Steam Oxygen ~ ~ Top Total S ~~:~~a 5 team ~ !!z!!. Ash ~b Produc t GasC
FREE FALL INJECTION OF LIGNITE COAL
CHPFL-1n 08-15-74 24.14 39.41 7.40 857 834 804 5.30 2.0 33.7 84.2 23.4 7.6 5.42:!; .36 34.94 :!; 1.60
CHPFL-ln 08-20-74 23.10 39.56 6.77 856 827 799 6.00 2.5 28.8 84.9 2.3.9 7.0 5.841: .29 n.97:!; .87
CHPFL-133 08-26-74 23.10 39.10 6.77 871 823 769 6.00 3.0 24.5 90.2 12.2 9.0 3.991: .24 25.11 1: .80
CHPFL-134 08-28-74 26.96 39.10 7.67 875 832 753 5.75 2.5 29.6 82.1 22.2 7.3 6.241: .09 30.62 :!; 1.70
CHPFL-135 08-30-74 25.70 39.10 7.22 873 823 717 5.00 3.0 24.4 70.8 19.7 7.5 5.33 :!; ..38 34.56 :!; 1.08
SUA!l.W BED-INJECTION OF LIGN In: COAL
CHPFL-123 07-02-74 27.02 36.10 6.33 839 801 742 4.97 3.5 33.3 12.2 12.7 8.5 6.45:!; .48 32.24 :!; 2.10
.... CHPFL-124 07-08-74 26.73 36.10 6.77 794 790 731 5.33 4.0 34.0 74.6 18.4 6.8 6.57:!; .56 32.701:3.10
C1I
CO CHPFL-125 07-17-74 26.29 40.46 6.95 844 784 718 5.50 3.5 30.0 76.1 14.4 7.5 6.20:t .14 3l.21:!; .91
CHPFL-126 07-19-74 26.38 39.71 6.95 831 784 729 5.50 3.5 28.5 77.7 15.9 7.5 6.08:!; .41 29.44 :t 1.80
CHPFL-127 07-23-74 24.40 39.41 6.60 836 774 703 6.00 3.5 24.2 81.2 12.2 8.2 5.781: .23 26.38 :!; 1.30
CUPFL-136 09-04-74 25.60 39.21 8.38 858 794 733 6.00 3.0 n.8 80.8 13.0 7.9 5.68:!; .17 31.37 :!; 1.80
CUPFL-137 09-06-74 23.80 39.51 7.75 877 796 714 6.00 3.0 23.4 87.3 22.7 7.3 5.12:!; .11 28.84:!; . 75
CHPFL-138 09-12-74 22.46 38.36 7.13 868 792 712 6.00 3.5 24.1 84.1 18.5 8.1 5.391: .75 26.30 1: 3.30
DEEP BED-INJECTION OF LIGNIn: COAL
CHPFL-143 10-20-74 23.40 39.11 7.67 784 759 716 3.42 1.5 15.6 68.9 23.1 7.4 6.72 1: 1.60 41.27 :!; 2.70
CHPFL-144 10-16-74 24.50 39.11 7.84 827 787 731 4.40 3.0 30.8 73.8 16.2 8.0 5.66:t .29 33.78 :!; 2.70
CHPFL-146 11-04-74 25.52 39.16 9.00 809 781 735 5.00 3.0 30.3 67.9 16.1 8.7 5.71:t .30 36.98 :!; 1.60
CHPFL-147 11-06-74 24.80 39.11 9.27 819 775 717 5.30 2.5 28.2 68.9 21.8 8.6 5.26:1: .54 36.56:t .98
CHPFL-148 11-08-74 25.70 39.11 9.45 804 775 729 5.00 3.0 29.4 64.3 18.7 8.4 5.93:1: .58 37.04 t 1.50
CHPFL-149 11-13-74 25.70 39.11 9.18 776 758 711 5.00 2.5 26.9 66.0 16.9 8.5 5.48:t .45 32.96 t .4'
(a) 'lbat periO
-------
a function of time for individual tests are presented in
figures 5 and 6, respectively. Carbon, hydrogen, and
oxygen mass balances for each test are summarized in
figure 7. As is apparent from this figure, material was
consistently lost during each experiment conducted. The
closest approximation to closure of a mass balance
occurred during the six deep bed-injection tests.
Aqueous Effluent Production
Time series measurements of phenol, chemical
oxygen demand, and total organic and inorganic carbon
contents of aqueous condensates are presented in figures
8-11, respectively. Similar measurements of thiocyanate
concentrations are presented in table 2; cyanide
measurements, though taken routinely, rarely exceeded
0.01 ppm and, therefore, were not tabulated. Data in
each figure are grouped according to coal injection
geometry, viz., free-fall, shallow, and deep bed-injection.
Note that with the exception of inorganic carbon
measurements, all the data reported indicate major re-
ductions in effluent production as a function of both
time and depth of coal injection. Consistent inorganic
carbon contents were anticipated; they reflect the pre-
sence of dissolved CO2 in basic (pH 9) condensate. At
steady state, measured phenol production rates range
from a high of 12lbs/ton coal, MAF (free fall injections)
to a low of < 0.5 Ibs/ton coal, MAF or equivalently less
than 30 ppm (deep bed-injections). Similarly, chemical
oxygen demands range from a high of ~80 Ibs/ton coal,
MAF to a low of ~2.5 Ibs/ton coal, MAF and total
organic carbon contents range from a high of ~21
Ibs/ton coal, MAF to a low of ~2 Ibs/ton coal, MAF.
Thiocyanate production rates range from a high of
~.045 Ibs/ton coal, MAF (free-fall injections) to a low
of less than 0.017 Ibs/ton coal, MAF (deep bed-
injections), the limiting sensitivity of the measurement
technique employed.
Recent analytical work carried out at PERC on
aqueous condensates from a range of Illinois No.6 gasifi-
cation tests (ref; 3) reflect the trends toward decreased
phenol production with increased depth of fresh coal
injection reported here in figure 8. Concentrations of
selected compounds measured in condensates from two
free-fall and three shallow bed-injection tests with
Illinois No.6 coal and reported in reference 3 are sum-
marized in table 3.
In contrast to the considerable scatter evident in
previously reported effluent production data (ref. 121.
note that in the present series of experiments, individual
data for runs at constant coal injection geometry are
substantially reproducible. Also, note that in the particu-
lar case of free-fall injection tests, steady state produc-
tion rates of all the effluents measured actually exceed
those at gasifier startup. This result was not anticipated;
an initial surge in effluent production followed by a de-
cline toward a reduced production rate at steady state
had been expected. The expected pattern of production
with time did occur in the shallow and deep bed-
injection tests.
Ammonia Production
Time series data on ammonia production rates were
not collected in the present experimental program. How-
ever, at least one measurement of ammonia content of
aqueous condensate was made during all but two of the
tests run. In contrast to the patterns of other major
constituents of gasifier condensate, ammonia production
was not influenced significantly by variations in coal in-
jection geometry. Within the accuracy of the data avail-
able, ammonia production averaged 15-20 Ibs/ton coal,
MAF. This yield is substantially greater than the approx-
imately 6-8 Ibs NH3/ton coal, MAF typically produced
in byproduct coking (ref. 4).
Char Production and Composition
Statistics on char production and composition are
summarized in table 4 for free-fall, shallow, and deep
bed-injections of coal.* From free-fall to shallow to deep
bed-injections, there is a distinct trend towards increased
volumes of char having progressively higher carbon,
hydrogen, oxygen, and nitrogen contents, lower ash con-
tents, and higher heating values. Char sulfur levels,
though stable and well below the Federal New Source
Performance Standard of 1.2 Ibs S02/MMBTU for
shallow and free-fall coal injection runs, increase sharply
for deep bed-injection runs. These data indicate clearly
that, with deepening injection, increasing amounts of
feed coal exit the reactor prematurely through the char
extractor, reducing overall levels of char conversion. This
is a direct result of the particular injection geometry
employed in these experiments (see figure 2), and could
be altered with proper modification of the reactor in-
jection system.
Hydrocarbon Condensate (Tar) Production
Tar production rates for each of the 19 experiments
conducted are presented as a function of time in figure
12. Grouped according to coal injection geometry, these
data indicate that: (1) with the exception of free-fall
coal injection runs, tar production declines with time
from an initially high rate to a reduced rate at steady
- - ---- - -----
state; and (2) steady state tar production levels fall
rapidly as coal is injected deeper into the fluidized bed
portion of the reactor. As with the previously cited
aqueous effluent production data, steady state tar pro-
duction levels in the free-fall injection tests actually ex-
ceed those observed during gasifier startup. Measured
*Detailed data for individual experiments are provided in
reference 2.
160
-------
40
38
36
34
32
30
28
26
24
~
Q)
;;: 40
I/)
0
0> 38
c:
0
.0 36
~
o
() 34
.0
.... 32
-
()
VI
30
(f)
<.:z-
, ..." ' I...
132 ---.,/' ~..-
../
..-- .
/"
134/
133,~
'.., /"""-.
'I h----\\-
Shallow bed - injection
124
123 \....
."
\ ...
.
,... ...' ~ -- /'" 148
-~ \ '-,t.' "'."
/-147"""'''\'' .;;. '.
j /'. \~::J-0.
146 j.,../ ,\.;' '. ~49
. ---- '---
/
144/
Deep bed - injection
2
OPERATING
3
4
5
TIME, hours
Figure 5. Product gas production versus time:
gasification of North Dakota lignite.
161
6
-------
8.0
7.0
6.0
5.0
4.0
~
.!?
-
'"
0 3.0
0'
C
o 8.0
.c>
~
0
0
~
...... "(.0
-
o
'"
Z
0 6.0
I-
o
:>
0
0
a: 5.0
0-
W
Z
«
:I: 4.0
I-
W
:E
I- 3.0
z
w
-' 8.0
~
:>
2
7.0
6.0
5.0
4.0
3.0
o
Free fall injection
132
I~~..~~:-.'_..-..'"
... \ ---............
'\ ' --..............-
".. '- ,---...../ ...'--
'-...._..>""131
133........
~..........
'---..... -'
'-...-'
Shollow bed - injection
Deep bed - injection
,'- --,
.. \
............ ~9.... \,
146/ /'.~~....V~" .,'\
/ -~....----'/\148
144 \>(~. ..147
2
3
4
5
6
OPERATING TIME. hours
Figure 6. Equivalent methane production versus time:
gasification of North Dakota lignite.
162
-------
Free fall Shallow bed - DeeR bed-
injections injections injections
+.20
z +10
w
C>
0 0
a::
CI
>-
- J: - 10
c
cu
0
"-
cu
Do -20
- +30
~
01
cu
~ +20
w
u
z +10
~ z
.J 0
~ CD
CD a:: 0
(/) ~
(/) u
~ -10
:!:
z -20
(/)
z
0 -30
~ +20
~
>
w +10
CI
z
w
C> 0
>-
x
0
-10
-20
131 133 135 123 125 127 137 143 146 148
132 133 124 126 136 138 144 147 149
EXPERIMENT NUMBER
Figure 7. Hydrogen, carbon, and oxygen mass balances:
gasification of North Dakota lignite.
163
-------
'0
E
0
0
u
c
2
...... 10
'"
~
z
0
I-
0
::>
0
0
0:::
0..
...J
0
Z
W
J:
0..
15
10
jF~';:->-~~-~~
/~,. .. ~ /' "~!!-
,;fi!l ~>,~,/~'-
:/1'
133 j«
131 :
135 .
5
Free fall injec lion
15
136
138.
Shollow bed - injection
5
x"
127" '-- ". / "'
x" .~ """""""" x/x ./~,
x .~.......-~~
, x_x
x...- x
15
144\
147', \.
"'" \
149\ "'. \
, ,
\"', \
148, \ .
146~, \ \
~"
',;,,,~,--, - .
':::::--:.' '.:..--. '..,.. -' ~-:---
123
5
Deep bed - injection
10
5
o
4
OPERATING TIME. hours
Figure 8. Phenol in byproduct water versus time:
gasification of North Dakota lignite.
164
6
-------
120
110
100
90
80
70
60
50
40
30
20
10
-
0
E 120
"0 110
o
u
c 100
~
..... 90
..
£
. 80
o
z 70
>-
x 30
o
.J 20
..'.1:...~_,/,/
,:1-.......'" ,~- \I'~.. r ""-
1/' 'Iv-t.."... /~-!" - --
/J/ /~~/y .." '~,-
134/ ../
133'//\
132 \
131 1"
135 .
Shallow bed-injection
--x---x-- .---
\-.'~"'-
127-:X_X" '''-'''-''
----x-x e_.~~.
Deep bed -injection
144
147\
149,,<\
148, "'.
" "~. ~
146---', '\:.
~'-.
,~~_.-
~......."
- - - -'
2
3
4
5
6
OPERATING
TI ME . hours
Figure 9. Chemical oxygen demand in byproduct water versus time
gasification of North Dakota li.gnite.
165
-------
-
0
E
.
0
0
u
c:
2
......
III
~ 20
z
0
CD
et:
ct
0
0
z
ct
C)
et:
0
-'
ct
t-
0
t-
30
20
"----. /'
/ 11_11- ,
"_,. ' //
, /' " -.... ,----;;- - - .../ \
" /. /' -,--- --,,-//
/ ~.. /'-"-~"-
. .' '- . .' -,-- . . . . -... \
~/l . "'''-"/''-'-'-/
;; /V\~ /..'-1 I
133.. V/."\
134 /1/
131 " /
132 '.
135
10
Free fall injection
30
Shallow bed - injection
10
.:::,.--.., x
"'-.e. ---;-x--x-- ...
~ x-x-x~
~. .- .
..:::::::-...~.=....---
30
. . . I
. . . I
Deep bed - injection
20
10
147
149\
144\....
'\.
148, '\:
146~ \.
"'~~.~...~...~
\ - - - ~.:::~:" ~ --.: ~ - - --.:~
o
2
:3
4
5
6
OPERATING
TI ME . hours
Figure 10. Total organic carbon in byproduct water versus time:
gasification of North Dakota lignite.
166
-------
- 30
o
E
c; 25
o
u
c:
2
'- 20
on
~
z 15
o
m
a:
«
(.') 10
o
z 5
«
C)
a:
0
z
30
25
20
15
10
5
o
30
25
Free fall injection
20
15
.. -"
...-
",,-
/'
. .' .""--,.-.= ,-,~.,~ "- "-
./ . ....~...~. ..~-'-I-'-'
134/" / ~. ~'? ~~~-;;-""-;i~:
~.. -- ,-
. .,.,..,
132 ,_I
133/,.-- 1--
1351
131
10
5
Shallow bed - injection
Deep bed - injection
/,
/ '.
:..""'""" / '.
..' .,,'- - - - - - - Z -- ~ '- '-''':' -'":,:,, :.:~:":" :. '- '-'-',:, ..: ',::
.. ,.," h ~- ==:-. --- - -- - -- ,- ~ - -- - ----
. "" -./ -- -- --- ---- --- -
147"_'- / -- -
148" _? ...---
149 --:/' ..........
144"'-Y
146
2
OPERATING
3
4
TIME. hours
Figure 11. Inorganic carbon in byproduct water versus time:
gasification of North Dakota lignite.
167
5
-------
Table 2.
Time series measurements: thiocyante content of aqueous
condensates collected during the steam-oxygen gasification
of North Dakota lignite: free-fall, shallow, and deep bed-
injections
Aqueous Thiocyanate Concentrations. pom
!.tl!l
1-1/4 hrs
2 hrs
2-3/4 hrs
3-1/2 hrs
4-1/4 hrs
5 hrs
5-3/4 hrs
Free Fall Iniection Tests
CHPFL-13l
CHPFL-l32
CHPFL-133
CHPFL-134
CRPFL-135
15
s5
11
20
25
16
s5
9
18
2S
1S
sS
12
20
30
12
sS
11
21
24
18
';:S
11
20
22
..
7 15
S5 sS
7 7
20 18
1S 4
..
......
Shallow Bed-Injection Tests
CRPFL-127
CHPFL-136
CHPFL-137
CHPFL-138
.....
Values consistently s5 ppm
Trap ~l (a) Trap #2(a) Trap #3 (a)
CHPFL-123 ';:S sS 12
CHPFL-124 :5:S sS 7
CHPFL-125 S5 :5:5 :5:S
CHPFL-126 :5:5
Deep Bed-Iniection Tests
CRPFL-143 17 '" ~ 7 t t
CHPFL-l44 t t
CHPFL-146
CHPFL-147 s5 consistently s5 ppm SS :5:S :5:S
CHPFL-148 + ~ ~ ~ ~ ~
CHPFL-149
Notes
(a):Batch sampling of condensate was c4rried out for four experiments. In these cases,
- ~rap #1 contains condensate. collected in condenser #1 (Figure 1) over the
course of the entire run.
- Trap #2 contains overflow condensate from condenser #1.
- Trap #3 contains condensate collected in condenser #2 (Figure 1) over the
course of the entire run.
Note that significant concentrations of thiocyanate are detectable only in Trap ~3.
This is consistent ~i~h the fact that water was condensed at a lower temperature in
condense r in.
168
-------
Table 3. Phenol compound distribution in typical synthane PDU gasifier
condensates as a function of coal injection geometry
Compound Present
in Condensate
Free Fall Injections, ppm
CHPFI-49 CHPFI-55
Shallow Bed-Injections, ppm
CHPFI-80 CHPFI-96 CHPFI-97
Phenol
Cresols
C2-phenols
Crphenols
Dlhydrics
Benzofuranols
Indanols }
Acetophenones
3,400
2,840
1,090
110
250
70
150
2,660
2,610
780
100
540
100
100
1,000
930
330
50
20
50
60
1,300
530
140
20
60
30
40
1,270
890
270
50
20
40
50
Table 4. Summary statistics on char and tar production as a function
of coal injection geometry
Item
Free Fall
Injection
Deep
Bed-Injection
Deep
Bed-Injection
Char Yield,
lbs/ton coal, MAF . . . .
Ultimate Analysis, wt %
C . . . . . . . . . .
H . . . . . . . . . .
S . . . . . . . . . .
o . . . . . . . . . .
N ..... . . . . .
Ash. . . . . . . . . .
Estimated HHV, BTU/lb . .
Combustion Equivalent
lbs S02/MMBTU . . . . . .
418 + 77 510 + 55 729 + 75
52.4 + 4.8 61.3 + 8.7 64.8 + 3.3
1.2 + 0.2 1.3 + 0.1 1.8 + 0.5
0.3 + 0.1 0.3+0.1 0.8 + 0.2
3.9 + 0.8 3.8 +0.7 5.5+1.9
0.5 + 0.1 0.4 +0.0 0.6 +" 0.1
41.9 +" 5.3 32.4 + 9.0 27.5 +" 6.3
8135 + 796 9513 + 1344 10,216 ~ 628
0.74 + 0.27 0.72+0.18 1.47 + 0.43
169
-------
110
100
90
80
70
60
50
40
30
20
10
- 110
o
E 100
.
o 90
o
u
c 80
2
..... 70
VI
.D 60
-
.
z 50
o
I- 40
()
:> 30
o
o
n:: 20
0-
n:: 10
./ "'\.... \", /\
135 "---,,'7-,'-...,'-... /'
'"
Shallow bed -injection
136
138...
137 -. \\ .x~
--- . '. -"",:X
125-......... 126.~... "'-.
-~~-A -e--8- --e-=::::::!--~._~~.:
--- A ~A--A_A_A-'
------- -x",--~--
Deep bed - injection
147
144~,.
143~~
145 - - -- 0-0_0 148
146- ~~;---..?=-o-:o-o .~-L,
---"~-'. ~~.. . ....... . ..
2
OPERATING
3
4
5
6
Figure 12. Tar production versus time: gasification in North Dakota lignite.
TIME ,hours
170
-------
Table 5. Sulfur content as a function of time, free-fall injection of coal
into the synthane PDU gasifier
wt % S in Equivalent
Trial Sample Time, hrs Tar 1 bs S02/MMBTU
131 0.63 1.25 1.87
4.63 0.82 1. 14
133 2.38 0.90 1. 17
4.13 0.79 1. 05
135 0.63 0.98 1.44
3.88 0.83 1.11
steady state tar production levels range from a high of
"'100 Ibs/ton coal, MAF (free-fall injections) to a low of
"'5 Ibs/ton coal, MAF (deep bed-injections). In contrast
to previously reported batch data on tar production dur-
ing gasification (ref. 2), present time series data-
particularly those for shallow and deep bed-injection
tests, are highly reproducible from run to run.
Tar Composition
Ultimate analyses (including sulfur content), esti-
mated high heating values (BTU/lb), and approximate
sulfur combustion emission rates (equivalent Ibs
S02/MMBTU) were determined for a limited number of
tar samples collected at different times during several
gasification tests. Tar sulfur contents consistently ex-
ceeded the Federal New Source Performance Standard
for liquid fuels of 0.8 Ibs S02/MMBTU (ref. 5). How-
ever, as is apparent from the free-fall time series data on
tar sulfur shown in table 5, tar composition approaches
steady state only slowly, with sulfur content declining
steadily from initially high levels. While a similar trend is
expected for shallow and deep bed-injection tars, data
have not been gathered to confirm this. A major reason
for this is a lack of sufficient sample to conduct a reli-
able analysis. As indicated earl ier, tar production rates
fall precipitously as coal is injected deeper into the gasi-
fieL
Tar/water emulsions initially complicated efforts to
determine densities and true boiling point (TBP) curves
for selected hydrocarbon samples. A satisfactory d isti 1-
lation procedure eventually was developed for the de-
hydration of tar samples. Subsequent analyses of de-
hydrated samples indicate densities consistently greater
than 1.0.
TBP distillation curves for a limited number of de-
hydrated tar samples are presented in figure 13.* Note
first that regardless of coal injection geometry, the initial
boiling point of each of the samples analyzed is sub-
stantially in excess of 100° C. This suggests the presence
of essentially no low molecular weight hydrocarbons, in
particular benzene, toluene and xylene (BTX). Secondly,
note that depending upon the sample considered, from
30 to 60 percent of the hydrocarbons present in the tar
boil at temperatures in excess of 500° C. No trend in
TBP curves with coal injection geometry is discernible
from the limited data available in figure 13. Curves for
two shallow bed-injection samples (138A and 1380) lie
above the curves for the free-fall injection samples
(1310 and 1320), and two (137A and 137B&C) lie be-
low; the curve for the one available deep bed-injection
sample lies below that of the free-fall sample. A trend in
the shape of TBP curves with reactor operating time is
suggested for individual gasification runs. Tars collected
late in a gasification run (samples 137B&C and 1380)
have higher TBP curves, i.e., contain a slightly larger
fraction of lower molecular weight hydrocarbons, than
tars collected early in a run (samples 137A and 138A).
However, interestingly, observed shifts toward lower
molecular weight tars do not give rise to increases in
BTX production, at least as measured by their presence
as constituents in the tar. The BTX content of the pro-
duct gas was not monitored.
* I nherently low tar production rates in shallow and deep
bed-injection tests limited the extent of meaningful sampling and
analysis. I n the specific case of deep bed-injection tests, only one
sample collected was large enough to yield a workable quantity
of tar following dehydration.
171
-------
100
Dehydrated Sample collection
90 sample t i me, h rs
Free fall injection of lignite
. CHPFL 131 D 3 1/8
80 II CHPFl 132 D 3 1/8
Shallow bed -injection of lignite
o ~ CHPFL 137 A 5/8
w 70 A CHPFL 137 BBC 2
~ 'V CHPFL 138 A 5/8
~ V CHPFL 138 D 3 1/8
-
r-
-------
DISCUSSION OF RESULTS
Time Series Analyses of Effluent Production and Com-
position
As had been expected, individual effluent produc-
tion rates were found to vary significantly with time
from reactor startup to shutdown. Available data indi-
cate that individual production rates are typically high
during reactor startup, and decline with time, typically
approaching steady state levels after about 2 to 3 hours
of reactor operation. Unexpectedly, effluent production
rates in the free-fall injection trials ran counter to this
trend; production rates consistently increased with time
following reactor startup. However, regardless of the
coal injection geometry involved, it is significant to note
that, for a given mode of gasifier operation, effluent
production rates are consistently reproducible, a result
not achieved previously where batch sampling tech-
niques were employed for effluent collection (ref. 2).
Interestingly, certain effluent compositions as well
as production rates vary significantly with time. For
example, the sulfur content of tar declines substantially
during the first 3 to 5 hours of gasifier operation. Al-
though not monitored in this experimental program, the
composition of char, in particular its sulfur content, is
expected to vary similarly with time.
On the basis of the extensive time series analyses
presented here, it is evident that not only steady state
production rates but, in certain cases, the composition
of individual effluents must be known to assess accurate-
ly the types and sizes of process equipment needed for
satisfactory effluent treatment. Effluent data obtained
from the analysis of batch samples collected over an
entire run, e.g., past PERC batch data (refs. 1,2). while
qualitatively helpful in characterizing the effluent po-
tential of a particular coal or reactor operating proce-
dure, are not sufficiently precise to form a basis for the
detailed design of an effluent treatment system.
Effect of Coal Injection Geometry on Gasifier Steady
State Effluent Production and Composition
(1) Effluent Production Rates
The 19 tests conducted in this experimental pro-
gram amply demonstrate that gasifier operating practice,
specifically the coal injection position within the gasifi-
er, strongly influences steady state rates of production
of the full range of reactor effluents. A statistical
summary of steady state values of the 7 indicators of
effluent production monitored throughout this study is
presented in table 6. For purposes of comparison, similar
data also are presented (where available) for weak
ammonia liquor typical of a steel industry coke plant
effluent. Note first that phenol production rates and
chemical oxygen demand (COD) are more than a factor
of 10 greater for the free-fall gasification of lignite than
for the manufacture of blast furnace coke. However,
cyanide and thiocyanate production are more than a
factor of 10 lower. Secondly, note that, although dra-
matic reductions in gasifier effl uent production are
achieved by shifting from free-fall to deep bed-injection
of lignite, residual phenol and COD levels are still rough-
ly equivalent to those of untreated coke plant weak
ammonia liquor. Tar production levels are greatly re-
duced from free-fall to deep bed-injection tests-by more
than a factor of 10. However, notice that the largest
percentage reduction in gasifier tar production, viz., 86.4
percent, resulted from the shift from free-fall to shallow
bed-injections of lignite. I ncreasing the depth of injec-
tion of lignite from 1-% to 4-% ft in the fluidized bed
portion of the gasifier (deep bed-injections) resulted in
an additional reduction of only 37.6 percent. Similar
trends in chemical oxygen demand (COD) and total
organic carbon (TOC) of aqueous effluents are apparent;
COD's are reduced by 84.2 and 69.5 percent, TOC's by
78.2 and 43.8 percent, respectively. Interestingly, the
above pattern does not hold for phenol production.
Shifting from free-fall to shallow bed-injections of lignite
results in a 70.4 percent reduction in phenol production;
however, shifting from shallow to deep bed-injections of
lignite results in a further reduction in phenol produc-
tion of 85.7 percent. Such evidence strongly suggests
that different mechanisms may be responsible for ob-
served reductions in various steady state effluent produc-
tion rates with changes in fresh coal injection geometry.
On the basis of data presented here, it appears that
with additional residence time, i.e., injection of coal into
a deeper fluidized bed, and/or increased reaction tem-
perature, essentially zero rates of production of the
major effluents-tar, phenol, and COD-could be
achieved. Conoco Coal Development Company (ref. 9)
experience with its CO2-acceptor pilot plant in Rapid
City, South Dakota, tends to support this supposition.
There, essentially no hydrocarbons heavier than methane
are detected in gas leaving a fluid bed reactor operated at
1400-1500° F with a gas residence time of about 20
seconds.
(2) Tar Composition
Although tar production rates drop precipitously
with the shift from free-fall to deep bed-injection of
lignite, the overall composition of condensible hydro-
carbons remains remarkably similar (see true boiling
point curves in figure 13). This is an unexpected result.
It has been presumed that reductions in the yield of
heavy tar would be paralleled by a greatly increased pro-
duct i on of light hydrocarbons including benzene,
toluene, and xylene (BTX). In fact, little, if any BTX
173
-------
Table 6. Comparative steady state effluent production rates:
byproduct coking, free-fall, shallow, and deep bed-
injections of North Dakota lignite
Gasification of North Dakota Lignite
Byproduct (a) Free-faTl Sha 11 ow Deep
Pollutant Coke Plant Injection Bed-Injection Bed-Injection
Tar, lbs/ton coal, MAF 93 74.1 + 27 10.1 + 5 6.3 + 2.2
Phenol, lbs/ton coal, MAF 0.86 - 0.97 11.9 + 1.3 3.5 + 1.9 0.5 + 0.6
Chemical Oxygen Demand,
lbs/ton coal, MAF 4.0 - 5.5 77.7 + 14.4 11.8+5.4 3.6 + 2.4
Total Organic Carbon,
lbs/ton coal, MAF 1. 60 - 1. 96 22.0 + 3.3 4.8 + 1.3 2.7 + 0.7
-- InorganicCarbon,(c)
.....,
~ N.D. 12.5 + 2.4 11.0 + 2.3 11.4 + 2.4
lbs/ton coal, MAF
Cyanide,
lbs/ton coal, MAF 0.018 - 0.054 Negligible Negligible Negligible
Thiocyanate, 0.045 + 0.083 <0.016 + 0.002(b) <0.017 + 004 (b)
lbs/ton coal, MAF 0.31 - 0.35
Notes
(a) Sources: References 4, 6, 7, and 8.
(b) Measurements at or below the limiting sensitivity of the analytical method employed.
(c) The relatively constant inorganic carbon production levels reported for different
lignite injection positions are expected and reflect the presence of a saturated
amount of C02 in basic (pH of 9) condensate.
-------
was found in free-fall tars, and no sign ificant percentage
increase in light hydrocarbons was observed as overall tar
production declined. This observation may be important
in view of various independent reports of high yields of
BTX in bench-scale experiments in which coal is heated
rapidly, exposed to high partial pressures of hydrogen
for a short time (typically less than 10 sec), and subse-
quently quenched (refs. 10,12). While coal heatup rates
were necessarily relatively low in all the free-fall injec-
tion tests (most of the length of the carbonizer was heat-
ed to only about 4000 C; see figure 3), they were quite
rapid in both the shallow and the deep bed-injection
tests. At the same time, gas residence times at the eleva-
ted temperatures in the fluid bed portion of the gasifier
were relatively short, less than 7 seconds even in the
deep bed-injection tests. The essentially complete lack of
BTX production (i.e., condensed BTX; quenched pro-
duct gas was not analyzed for BTX) under these condi-
tions as well as the consistently high molecular weight of
residual tar stand in sharp contrast to smaller-scale lab-
oratory results referenced above. While hydrogen partial
pressures were significantly lower in the present ex-
periments (approximately 180 versus 1000 psig or more)
and there was a significant partial pressure of water in
each of the present tests, the relevance of these differ-
ences is unknown.
Effect of Coal Injection Geometry on Total Product and
Methane Gas Yields
Summary statistics on gasifier temperature profiles,
total product gas, equivalent methane yields, and effec-
tive product gas residence times for each of the three
coal injection geometries investigated in the current pro-
gram are presented in table 7. Within the accuracy of the
data available and on the chosen basis of SCF /Ib of
carbon gasified, equivalent methane yields are essentially
unaffected by changes in coal injection geometry. Total
product gas production rates are similarly unaffected.
However, as indicated by the increasing carbon content
of residual char with increasing depth of coal injection,
apparently stable gas makes on a SCF/lb of carbon gasifi-
ed may be deceiving. The efficiency of char gasification
generally declines rapidly with carbon content, suggest-
ing that at equal percentage carbon conversions for free-
fall, shallow, and deep bed-injections of coal methane
and total product gas production rates may differ.
Additional experiments are needed to determine
whether or not methane and/or product gas yields are
influenced by coal injection geometry and, if so, how
much.
Possible Mechanisms to Explain Observed Results
It is startling to find that production rates of essen-
tially all undesirable "byproducts" of coal gasification
(except sulfur) can be dramatically reduced by minor
modifications in gasifier design with apparently negligi-
ble reduction in desirable products, viz., equivalent
methane, total product gas and ammonia. Such a result
clearly suggests that distinctly different mechanisms
must be at work in the production and/or destruction of
various gasifier constituents. I n varying the coal injection
geometry from free-fall to deep bed-injection, several
operating conditions change in the gasifier. Among these
are: (1) the heatup rate of the fresh coal is greatly in-
creased; (2) the mean reaction temperature is increased;
(3) gas-solid mixing is enhanced; (4) product gas resi-
dence time is increased; and (5) contact between fresh
product gas and potentially catalytic char particles is
increased. Depending upon the specific coal injection
geometry involved, changes in various subsets of these
operating conditions tend to dominate.
I n shifting from free-fall to shallow bed-injections of
coal, the major reactor conditions undergoing change are
fresh coal heatup rate and mean reaction gas tempera-
ture- Although a study is needed to verify it, a major
portion of the observed improvement in effluent pro-
duction might well be attributed to a reduced rate of
formation under the more severe conditions of shallow
bed-injection operation rather than to decomposition of
formed material. However, even if such a supposition is
shown to be correct, it cannot account for all of the
observed reduction in effluent production rates since
deep bed-injections of lignite produce additional reduc-
tions in effluent production. Further, since reaction
temperatures in the deep bed-injection tests were
actually lower than those in the shallow bed-injection
tests, temperature cannot explain the observed reduc-
tions. The only significant changes in operating condi-
tions appear to have been increased product gas resi-
dence time and additional gas-solid contacting. It would
appear that at least these two significant mechanisms
influence observed reductions in effluent production
with shifts from the shallow to the deep bed coal injec-
tion geometry. No attempt was made in the present ex-
perimental program to sort out the dominant mecha-
nisms involved. However, bench-scale work is now in
progress to determine the relative importance of various
reactor operating parameters in effluent production.
Applicability of Current Results to Other Coals and/or
Different Gasifiers
Lignite coal was selected for testing here specifically
because it is a noncaking coal. Although caking coals
(e.g., Illinois No.6 and Pittsburgh Seam) appear to pro-
duce most effluents in yields similar to those of lignite
(see data available in reference 2). they differ in one
important respect: tar production is substantially greater
for caking than for noncaking coals. In addition, due to
their agglomerating properties during heatup, caking
175
-------
Table 7. Summary statistics, gasification of North Dakota lignite:
free-fall, shallow, and deep bed-injection of the coal
Data in parentheses throughout this table indicate the number of experi-
ments considered in reported average results.
Gas production figures are stated relative to lbs of carbon gasified for
wide variations in the extent of gasification achieved.
Total gas production includes CH4' C2H6' H2' and CO.
Equivalent CH4 is the sum of SCF of CH4 and twice the SCF of C2H6'
Estimated effective residence time of freshly volatilized coal hydrocarbons
in the high temperature, fluidized bed portion of the gasification reactor
(see Figure 2). Calculations are based on an assumed pIng flm~ of gases
through the reactor. The effective residence time for all free fall runs
is assumed to be essentially zero.
Run ClIPFL-133 has heen omitted from this a',rerage.
Run ClIPFL-14J has been omitted from this average.
Parameter
Free Fall
In;ection
Fluid Bed Temp Profile,oC
Bottom. . . . . . . . . . . . . . . . . . . . .
866 :!:. 9.00
(5)
Average. . . . . . . . . . . . . . . . . . . .
828:!:. 5.00
(5)
768:!:. 36.00
Top....................... .
(5)
Total Product Gas
SCF/lb C Gasifiedb,c............
33.26 + 2.10
(:)f
Equiv. CH4 Production
SCF/lb C Gasifiedb,d.... ........
5.70 + .47
(:)f
Effective Gas Residence
Time, sec e . . . . . . . . . . . . . . . . . . . . . .
-0-
(a)
(b)
( c)
(d)
(e)
(f)
(g)
176
Coal Feed Configurationa
Shallow
Bed-In;ection
Deep
Bed-In;ection
843 :!:. 26.00 803:!:. 20.00
(8)
789:!:. 8.00 773 :!:. 12.00
(8) (6)
723 :t: 13.00 723 :!:. 10 .00
(8) (6)
29.76:!:. 3.10 35.50:t: 2.30
(8) (5)g
5.92:!:. 1.13 5.62:!:. 0.47
(8) (5)g
2.82 :!:. .ll 6.62 + .11
(8) (6)
-------
coals are difficult to feed to a gasifier in any geometry
other than free-fall. PERC efforts to date to inject pre-
treated caking coal directly into the fluidized bed por-
tion of the Synthane gasifier have been unsuccessful. If a
satisfactory means of injection is eventually found, re-
sults of the present study on I ign ite indicate that more
severe processing conditions will be required to treat
caking coal to reach the reduced levels of effluent pro-
duction achieved with lignite.
Without some quantitative understanding of the
mechanisms responsible for the reductions in effluent
production observed here, projections of possible effects
in other gasifiers are speculative at best. A qualitative
interpretation of the data here suggests that observed
effluent reductions are not specific to the Synthane
gasifier; this should be achievable in other reactors.
REFERENCES
1. Albert J. Forney et aI., "Analyses of Tars, Chars,
Gases, and Water Found in Effluents from the
Synthane Process," Bureau of Mines Technical Pro-
gress Report No. 76, U.S. Department of the In-
terior, January 1974.
2. D. V. Nakles, M. J. Massey, A. J. Forney, and W. P.
Haynes, "I nfluence of Synthane Gasifier Conditions
on Effluent and Product Gas Production,"
PERC/RI-75/6, Energy Research and Development
Administration, December 1975.
3. C. E. Schmidt, A. G. Sharkey, and R. A. Friedel,
"Mass Spectrometric Analysis of Product Water
from Coal Gasification," Bureau of Mines Technical
Progress Report No. 86, U.S. Department of the
Interior, December 1974.
4. H. H. Lowry, editor, Chemistry of Coal Utilization
Volumes I and II, John Wiley & Sons: New York,
1945.
5. "Standards of Performance for New Stationary
Sources," Federal Register 36 (247), pp.
24876-24895.
6. P. J. Wilson, and J. H. Welles, Coal, Coke and Coal
Chemicals, McGraw Hill: New York, 1950.
7. R. D. Hetner, G. R. Tallon, and C. W. Fischer,
"Coke Plant Effluent Treatment Investigations,"
Presentation at the Eastern States Blast-Furnace and
Coke Association Winter Meeting, Pittsburgh, Pa.,
February 12-13, 1970.
8. R. W. Dunlap r.Jnd F. C. McMichael, "Environ-
mental Impact of the Treatment and Disposal of
Coke Plant Wastewaters," Environmental Studies
Institute Report, Carnegie-Mellon University,
January 17, 1975.
9. Carl Fink, George Curran, and John Sudbury, "C02
Acceptor Process Pilot Plant 1975, Rapid City,
South Dakota," Presentation at the Seventh Syn-
thetic Pipeline Gas Symposium, Chicago, Illinois,
October 27,1975.
10. "Improved Techniques for Gasifying Coal," Second
Annual Briefing of the Clean Fuels Institute, De-
partment of Chemical Engineering, The City College
of New York, January 13-14,1975.
11. Meyer Steinberg and Peter Fallon, "Coal Liquefac-
tion by Rapid Gas Phase Hydrogenation," Presenta-
tion at the 169th National American Chemical
Society Meeting, Philadelphia, Pa., April 6-11,1975.
12. F. C. Schora, "A.G.A. Coal Gasification Research,"
Presentation at the Fifth Synthetic Pipeline Gas
Symposium, Chicago, III., October 29-31, 1973.
177
-------
ENVIRONMENTAL ASPECTS OF SYNTHOIL PROCESS
FOR CONVERTING COAL TO LIQUID FUELS
Sayeed Akhtar, Sam Friedman, and Paul M. Yavorsky'"
Abstract
SYNTHOIL process converts high-sulfur coals to
low-sulfur, low-ash liquid fuels. A slurry of coal in re-
cycle oil is reacted with hydrogen in the presence of
Co-MoISiOrAI203 catalyst in a turbulent-flow, packed-
-bed reactor at 45rf C and 2,000-4,000 psi. Thus, coal is
converted to liquid hydrocarbons and sulfur is elimi-
nated as H2S gas. The liquids and unreacted solids are
separated from the gases and centrifuged to obtain a
low-sulfur, low-ash liquid fuel. Primarily designed for the
conversion of coal to a utility fuel, the process is easily
adaptable to the production of distillate fuels if the
economics justify. The results of a study of the environ-
mental aspects of the SYNTHOIL process will be re-
viewed and a quantitative estimate of the effluents from
a 75,000 bbl/day SYNTHOIL plant operating on a high-
sulfur coal will be presented.
INTRODUCTION
Availability of environmentally acceptable utility
fuels prepared from U.S. coals will eliminate the need
for importing low-sulfur utility fuel oils for industrial
use. Consequently, conversion of coal to a low-su Ifur
utility fuel is a high priority objective of ERDA, and
several processes addressed to the objective are currently
under development. One such process in an advanced
stage of development is the SYNTHOI L process. A
bench-scale continuous unit processing 0.5 ton of coal
slurry per day has been in operation for over 2 years,
and a process development unit of 10 tons of coal per
day capacity is under construction. Concurrently,
engineering study and laboratory investigation of the
environmental problems of the process are also in prog-
ress. This presentation is a review of this work.
DESCRIPTION OF THE PROCESS
Figure 1 is a flow diagram of a SYNTHOI L plant
processing 1,000 tons of coal per hour. The yield of
product oil is 3,100 bbl per hour corresponding to a
nominal plant capacity of 75,000 bbl per day. The yield
of light oil and C1 -Cs gases also is shown in the figure.
*The authors are with the Pittsburgh Energy Research
Center, Energy Research and Development Administration,
Pittsburgh, Pennsylvania.
Hydrogen and a slurry of coal in recycle oil are
passed through a turbulent-flow, packed-bed reactor
charged with pellets of Co-Mo/Si02 -A12 03 catalyst. At
2,000-4,000 psi and 4500 C, coal is converted to liquid
hydrocarbons and the heteroatoms in coal are eliminated
as hydrides. The product stream from the reactor is led
to a gas disengager where the liquid and unreacted solids
are separated from the gases. The liqu id stream is passed
through a centrifuge to remove the unreacted solids con-
sisting of mineral matter and unconverted coal sub-
stance. The centrifuged liquid is a low-sulfur, low-ash
product oil, suitable for use as a utility fuel. A portion
of the centrifuged oil is recycled to convey more coal
into the plant while the rest of it is available as the net
product.
The solids from the centrifuge are pyrolyzed to
recover an additional quantity of nonpolluting fuel oil.
The residues from the pyrolyzer, consisting mostly of
mineral matter and some carbonaceous material, is
gasified to produce H2 for the process. If necessary,
some coal also may be gasified to produce sufficient H2
for process requirement.
The gases from the gas disengager are processed
through a purification train to remove ammonia and
ammonium salts by water wash, hydrocarbon gases and
vapors by oil wash, and acid gases by amine wash. The
purified hydrogen is combined with makeup hydrogen
and recycled to the reactor. The gas purification is con-
ducted at the plant pressure to minimize the cost of
recycling the purified hydrogen.
The yield data in figure 1 refer to the processing of
a Kentucky coal at 4,000 psi and 4500 C (ref. 1). Anal-
ysis of the high-sulfur, high-ash, run-of-mine Kentucky
coal is given in table 1.
THE PROCESS EFFLUENTS
The scope of this presentation will be limited to the
process effluents only. The environmental aspects of the
effluents from ancillary activities, e.g., mining and trans-
portation of coal, generation of steam and power for
plant use, and transportation of products, will not be
included in this paper.
Coal Preparation
SYNTHOI L process will operate on run-of-mine
coal with up to 20 percent ash. Mineral matter in run-
179
-------
H2
pia nt
Makeup
gas
Coal
1,000t/h
(XI
o
S I u rry
pump
Makeup
compressor
15xI06scfh
Reactor
gas
Crushing
and
grinding
Slurry feed
Recycle oil
Recycle gas
Recycle
compressor
Effluent
Leon amine
Reactor
Amine
absorber
Rich nmine (H2S)
H20
CI- Cs gases
3.0 x 106 scfh
Light oil 120.bbl/h
Residue to
H2 plant
Quench
oi I
Solids
separator
Oil
Pyrolyzer
Product oil
3,100 bbl/h
Figure 1. Flow diagram for a future commercial SYNTHOI L plant.
-------
Table ,. Analysis of feed coal*, as received
Proximate analysis, wt. %
Moisture
Ash
Volatile matter. .
Fixed carbon
. . . . . . . .
4.2
16.5
36.2
. . 43.1
Ultimate analysis, wt. %
Hydrogen
Carbon
Nitrogen
Oxygen.
Sulfur. .
Ash
. . . . . . . . .
.4.8
. 60.7
1.2
11.3
. . . .5.5
16.5
Forms of sulfur, wt. %
Su Ifate
Pyritic
Organic
. . . . . .
. . . 0.47
3.08
1.95
. . . . . . . . . . . . . . .
Calorific value, Btu/lb. . .
.11,020
Rank: h.v.A.b.
* A blend from Kentucky seams 9, 11, 12, and 13 which are
mined together; Ohio County, Western Kentucky.
of-mine coals, of course, varies with seam and mining
method, but currently most eastern coals, after prelimi-
nary mine-mouth cleaning to remove mine debris, con-
tain 12 to 20 percent ash. Such coals will be processed
with 0 ut a sh reduction. Coal preparation for the
SYNTHOI L process will therefore consist of (a) drying
to reduce moisture to about 3 percent for efficient pul-
verization, and (b) comminution to approximately 70
percent through 200 mesh, U.S. standard sieve, and 100
percent through 100 mesh. The effluents from such a
coal preparation section will be moisture and dust. A
quantitative estimate of these effluents from a 1,000
ton/hr plant is given in table 2. There will also be an
indeterminate quantity of rain runoff from the storage
and handling area for raw coal. Considerable information
on the management of contaminated water from coal
mines is now available that will apply to the manage-
ment of aqueous effluents from the SYNTHOI L coal
preparation section (refs. 2,3). The dust can be collected
efficiently with baghouse filters and added to the
powdered coal.
Table 2. Effluents from the coal preparation
section of a ',000 ton coal/hr SYNTHOI L plant
1. Water evolved in drying coal from an aver-
age of 8 pct to 3 pct moisture.
12,500 gph
2.
Dust from crushing, pulverization, and
transfer operations.
1 tonlhr
3.
Rain runoff from storage and handling
areas.
Indeterminate
Slurry Preparation and Pumping
Since the preparation and transfer of coal slurry will
be conducted in an enclosed system, the effluent of con-
cern in this section is the vapor leak from pump seals.
The concentration of the vapors will, of course, depend
on the size of the leak, slurry temperature, and air circu-
lation. An investigation of the air surrounding the slurry
preparation and pumping area is currently in progress
(ref. 4). The results of this investigation will be pre-
sented at a future symposium.
The High-pressure Section
This section comprises the reactor, the heat ex-
changers, the gas-liquid separator, and the gas purifica-
tion system. Aside from any accidental release, the efflu-
ents of environmental concern from this section will be
the following:
1. Sour water, consisting of process make-water
and the scrub water injected into the gas stream to dis-
solve ammonia and ammonium salts.
2. Spent scrub oil from the gaseous hydrocarbon
recovery unit.
3. S pe nt a mine solution from the acid gas
scrubber.
The contaminants in these effluents will, of course,
be determined by the gaseous and entrained impurities
in the reactor outlet gas. A mass-spectrometric analysis
of the gas is given in table 3. The gas was sampled after
the product stream had been cooled and the liquids
separated from the gas. The absence of NH3 from the
gas may be noted. Presumably, NH3 was eliminated
quantitatively from the gas phase as solid ammonium
sulfide/polysulfide by reaction with H2 S present in
excess. Conceivably, other trace components in the gas
might also have escaped detection. Nevertheless, they
181
-------
Table 3. Analysis of SNYTHOI L reactor gas
Component
Volume percent
Ho
CH.
Co H6
C3H8
C.H,o
Co H.
C3H6 .
C.H8 ..
HoS . .
CO
coo. . . .
No . . . . . . .
. . . . . .
. . . . . . . . . . . . . . .94.35
. 2.84
. 0.90
0.61
. 0.30
. . . . . . . . . . 0.03
. 0.14
0.12
. . . . . . . . . . .. 0.04
. . . . . . . . . . . . . . . 0.23
0.07
. . 0.31
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
will accumulate in the scrub liquids and must be con-
sidered while planning disposal of these liquids. We
believe substantive research with respect to the scrub
liquid effluents will be possible when the 10 ton/day
process development unit is operating. The unit will have
the gas purification system so that samples of the spent
scrub liquids will be available for analysis.
An estimate of the make-water, ammonia, and sul-
fur from a SYNTHOI L plant processing 1,000 ton/hr of
the Kentucky coal is given in table 4.
Solids Separation and Pyrolysis of Solids
As in the slurry preparation and pumping section,
the effluent of concern in the solids separation and
pyrolysis sections will be organic vapors. The work cur-
rently in progress on the investigation of the air sur-
rounding the slurry preparation and pumping area will
also be relevant to the problem of organic vapors in the
solids separation and pyrolysis sections. A rotary cal-
ciner and a fluidized-bed carbonizer are under construc-
tion. Quantitative data on effluents from the pyrolysis
section will be available when this equipment is in opera-
tion.
The carbonaceous residues from the pyrolyzer will
Table 4. Makewater and byproducts from a
1,000 ton coal/hr SYNTHOI L plant
1. Makewater:
.28,000 gph
2. Ammonia (NH3):
. . . . . .
. 9 tons/hr
3. Sulfur (5):
. . . . . . .
. 38 tons/hr
be gasified to generate H2 for the process, and the
mineral matter originally present in coal will be ultimate-
ly rejected as a solid effluent from the gasifier. This solid
will be used as mine fill and, therefore, the extent to
which it will contaminate rain runoff is of concern. We
plan to study the concentration of leachable salts in
solids derived from different coals (ref. 4).
REFERENCES
1.
Sayeed Akhtar, James J. Lacey, Murray Weintraub,
Alan A. Reznik, and Paul M. Yavorsky, "The
SYNTHOIL Process-Material Balance and Thermal
Efficiency," presented at the 67th Annual A.I.Ch.E.
Meeting, Washington, D.C., Dec. 1-5, 1974.
"Processes, Procedures, and Methods to Control Pol-
lution from Mining Activities," EPA-430/9-73-011,
Oct. 1973, U.S. Environmental Protection Agency,
Washington, D.C. 20460.
Z. V. Kosowski, "Control of Mine Drainage from
Coal Mine Mineral Wastes; Phase II: Pollution
Abatement and Monitoring," EPA-R2-73-230, May
1973, Environmental Protection Technology Series,
Office of Research and Monitoring, U.S. Environ-
mental Protection Agency, Washington, D.C. 20460.
Andrew G. Sharkey, Jr., and Coworkers, Pittsburgh
Energy Research Center.
2.
3.
4.
182
-------
ENVIRONMENTAL ASPECTS OF THE SRC PROCESS
Abstract
Russell E. Perrussel, Walter Hubis, and J. L. Reavis*
The prime purpose of the Solvent Refined Coal
(SRC) process is to produce a clean, low-ash, low-sulfur
solid fuel from high-ash and/or high-sulfur coals of var-
ious ranks for use in stationary combustion systems.
This paper presents an overview of the SRC process and
its associated environmental problems.
Some of the topics discussed are: The process de-
scription, plant site, baseline and proposed environ-
mental studies, fugitive particulates and vapors, process
waste disposal system, trace elements programs and
sulfur plant.
SRC pollution problems are similar to those faced
by other industries and can be handled adequately with
existing technology to meet environmental standards.
INTRODUCTION
Coal contains inorganic mineral matter and when
burned is converted to fly ash, which can go out as
particulate matter. Electrostatic precipitators are used to
remove most of this fly ash, and a problem exists in
removing and disposing of this alkaline fly ash. Coal also
contains sulfur compounds, both inorganic and organic.
When coal is burned, these sulfur compounds are
expelled as oxides of sulfur which are said to be a threat
to our ecology. Several methods have been devised to
remove these sulfur oxides and fly ash from stack gases.
There are different schools of thought as to how success-
ful these removal methods are. Another approach is to
remove the mineral matter and sulfur-containing com-
pounds from the coal before it is burned. Several new
processes have been developed using this principle. This
presentation discusses one such process called the
Solvent Refined Coal (SRC) process which is being
developed by The Pittsburg & Midway Coal Mining Co.
(P&M), under the sponsorship of the U.S. Energy Re-
search & Development Administration (ERDA).
The process provides a solid fuel (which is low in
sulfur and ash) to meet environmental requirements, has
uniform characteristics regardless of the raw coal proc-
essed, and has a high energy content. It is well suited for
". R. E. Perrussel is Chief Chemist, The Pittsburg & Midway
Coal Mining Co., SRC Pilot Plant, North Fort Lewis (Dupont),
WA; W. Hubis is Technical Coordinator, Gulf Mineral Resources
Company, Denver, CO; and J. L. Reavis is Project Engineer, The
Pittsburg & Midway Coal Mining Co., Research & Development
Department, Merriam, KS.
use as fuel in stationary combustion systems such as
used in the electric utility industry for power generation
and for a variety of other end uses.
PROCESS OVERVIEW
The Solvent Refined Coal (SRC) process has been
described in detail in the literature (refs. 1-6,8) and is
very well known. Basically, it is very simple. Figure 1
shows a general block flow diagram of the process with
water and gases treatment units. Feed to the preheating
section consists of pulverized coal as a slurry in the re-
cycle process solvent where it is contacted with hydro-
gen. The dissolution step is done at about 850°F and
1500 psig.
The gas phase of the reactor (dissolver) discharge is
fed to the acid-gas removal plant where carbon dioxide
and hydrogen sulfide are removed by amine scrubbing.
The hydrogen sulfide is converted to sulfur in a Stret-
ford su Ifur recovery unit. Carbon dioxide is vented to
the atmosphere.
The liquid phase from the gas-liquid separator goes
to the filtration area where the insoluble mineral phase is
separated using a rotary precoat filter. The filtrate
produced in the filtration operation is further separated
by vacuum-flashing into a distillate and residue. The
distillate is fractionated into a light oil product (coal tar
naphtha), wash solvent and recycle process solvent. The
residue from the vacuum flashing operation, a low-ash
sol id with a low-sulfur content, is the major product of
the process. This product can be remelted and burned as
a liquid fuel or pulverized and burned as a solid fuel.
This product will solidify at about 350° F Whether
liquid or solid, it is named Solvent Refined Coal (SRC).
The properties and chemical analysis of typical SRC and
feed coal (as received and MAF basis) are compared in
table 1. The SRC process removes essentially all of the
inorganic sulfur, 60-70 percent of the organic sulfur and
98 percent of the ash in the coal. The oxygen is reduced,
producing a higher heating value product than the MAF
coal.
PLANT SITE
Figure 2 is an aerial photo of the pilot plant located
about one mile inside the U.S. Army military reservation
at North Fort Lewis (near Dupont), Washington. The
plant site covers about 12 acres. Approximately two
183
-------
Waste Prace ss WASTE HZO Plant liqUid
..
.... DISPOSAL JII"'"
liquid Streams TREATMENT Effluent
Natural System
Gas
MAKEUP Pur ge
HZ 10
SYSTEM Flore
Steam
SULFUR SULFUR LIGHT
N OIL
:r REMOVAL RECOVERY
0. PRODUCT
=>
Q)
-'"'
a
Cool ::;: Re cycle HZ
COAL COAL SOLVENT WASH
GAS - LIQUID
STORAGE SLURRY DISSOLUTION SOLVENT
a SEPARATION RECOVERY PRODUCT
Pulverization PREHEATING
....
~ R ec }Ide Process Solvent
SRC
Vap or s From FILTER- SEPARATION
FLARE FILTR ATiON
Storage Tonks AID BY VACUUM
FLASHING
MINERAL
RESIDUE
DRYING a
STORAGE
SRC
SOLIDIFICATION
a
STORAGE
Figure 1. Block flow diagram, SRC pilot plant, Fort Lewis, Washington.
-------
Table 1. Comparison of feed coal and solvent refined coal
Kentucky Coal
(Hopkins County) Typical
As Received MAF* SRC
% Carbon 68.2 78.0 88.0
% Hydrogen 4.8** 5.6 5.9
% Nitrogen 1.3 1.4 2.2
% Sulfur 3.6 1.4* 0.7
% Chlorine 0.05
% Oxygen (by difference) 8.4** 12.7 3. 1
% Moisture 4.8
% Ash 8.7 0.2
Forms of Sulfur
% Sulfate 0.25
% Pyritic 2.2
% Organic 1. 15 1.4* 0.7
Btujlb 12,900 14,030 16,250
Fusion point (gradient bar) 3500 F
Density (gm/cc) 1.2
*
MAF or "moisture and ash free: in
sulfur
**
corrected for moisture
this cas~ includes only the organic
185
-------
Figure 2. Aerial view of solvent
refined coal pilot plant,
North Fort Lewis (Dupont),
Washington.
186
-------
years were required for construction at an estimated cost
of $20 million. The plant has now been processing coal
for over a year with down time being less than 30 per-
cent.
The plant facilities can be conveniently grouped
into several areas including coal preparation and han-
dling, coal liquefaction and filtration, gas cleaning and
acid-gas removal, product handling and storage, solvent
recovery, tank farm, hydrogen production, and finally
auxiliary facilities such as steam generation, inert gas
manufacture, water treating, and a sulfur plant.
ENVIRONMENTAL CONSIDERATIONS
Any process based on coal will inevitably encounter
many environmental problems. In all processes, regard-
less of nature, all materials coming into its plant site will
either become a product for removal or a waste for dis-
posal with their associated pollution problems. The SRC
pilot plant is no exception to this.
A large proportion of the total plant cost is devoted
to environmentally related matters and to byproduct dis-
posal. The environmental aspects of an SRC plant are
similar to those encountered in power plants (coal dust,
flue gases), coke plants (coal tar hydrocarbons), and
petroleum refineries (distillation, hydrogenation, and
handling of solvents, tars and pitches). No unique prob-
lems and no greater difficulty in meeting the environ-
mental standards of these industries are anticipated for a
commercial SRC plant.
Prior Studies
Some consideration has been given previously to
environmental aspects of the SRC liquefaction process in
several papers presented at the first EPA symposium on
fuel conversion technology. One of these (ref. 5), "Envi-
ronmental Aspects of SRC Process," was presented
about five months before the pilot plant started process-
ing coal.
Exxon Research and Engineering Co. recently com-
pleted a report for the EP A (ref. 6). The report gives
results of a review of the Solvent Refined Coal (SRC)
process of The Pittsburg & Midway Coal Mining Co.,
from the standpoint of its potential for affecting the
environment. It includes estimates of the quantities of
solids, liquid, and gaseous effluents, where possible, as
well as the thermal efficiency of the process. It proposes
a number of possible process modifications or alterna-
tives which cou Id facilitate pollution control or increase
thermal efficiency, and points out new technology
needs. A complete I isting of all plant flow streams and
reagents used is reviewed.
I n view of the year's operation of the pilot plant,
now a more precise definition and discussion of the envi-
ronmental problems can be presented.
Baseline and Proposed Ambient Air and Water Quality
Studies
Before construction on the plant began, an environ-
mental quality assessment was performed in the vicinity
of the proposed plant site at North Fort Lewis, Washing-
ton (ref. 4). Specifically, an air and water quality study
was conducted to establ ish existing background levels of
selected pollutants as a basis for determining any signifi-
cant environmental changes which may subsequently re-
sult from the operation of the new facility.
Overall, the baseline study showed that the air was
clean, the water was pure, and the soil-though rocky,
would support vegetation consistent to the area.
A new proposal has been drawn up to evaluate the
air, water, solid wastes, soils, vegetation, and aquatic
biology at the operating plant site and surrounding area.
This will be a comprehensive environmental assessment
conducted over an extended time. It will include a de-
tailed study of all potential emission sources (flare stack,
drier vent, dust collectors, solvent storage tanks, etc.).
These data will then be used to design an effective
ambient monitoring program.
This facility is one of the first of its kind. An ade-
quate assessment of the environmental problems as-
sociated with air and water, and their effects on humans
and the vegetation in and near the facility is essential to
determine the current environmental impact. This infor-
mation is necessary to implement process control tech-
niques as well as industrial hygiene safeguards, both at
the North Fort Lewis facility and future installations.
Fugitive Particulates and Vapors
Fugitive particulates (dusts) and vapors (fumes) can
originate in several areas throughout the plant. Dust
problems are imminent when the coal is pulverized, the
filter aid precoat slurry is prepared, the mineral residue
is dried and discharged, Solvent Refined Coal trans-
ferred, or when dry chemical containers are emptied.
These are controlled by using covered conveyors and,
induced draft vents with the vent gas being filtered to
remove particulates where possible.
As the coal is pulverized, a hot inert gas dries it and
carries it into a bag house. Here the dried, powdered coal
is collected for discharge into the slurry blend tank. This
operation is in a closed system. The water I iberated from
the coal is sent to the waste reservoir for treatment.
Diatomaceous earth, with or without asbestos, is
187
-------
used as a filter aid on the rotary filters. In this operation,
bags of filter aid are opened, dumped into a hopper for
later precoat slurry preparation. To minimize the inhala-
tion of these dusts, a large glove box was designed and
constructed. A bag of filter aid is inserted into the box
which is then closed. The hands and arms of the opera-
tor are then inserted into the "gloves" attached to the
box. The bag can then be opened in an enclosed chamb-
er. The box is then placed over the hopper and the re-
leased filter aid discharged without evolved dust.
In order to eliminate vapors from the solvent tanks
and other vessels, they are enclosed with an inert gas
blanket. As this blanket is purged out when a vessel is
filled, the gas and vapors are discharged into an inert gas
flow. This flow, as well as all of the gases purged from
the pilot plant, go into the flare piping system to be
carried to the flare stack. All safety relief valves are also
connected to the flare system. The gases are burned in
the flare stack, supplemented with natural gas if neces-
sary to maintain combustion.
The disposal of mineral residue poses several prob-
lems. In a commercial plant, utilization to useful prod-
ucts instead of the costly disposal would make the proc-
ess more economical. One method is to gasify the miner-
ai residue. A slagging gasifier would convert the mineral
residue into a relatively clean slag and synthesis gas (H2
+ CO). The sulfur in the mineral residue would be con-
verted into hydrogen sulfide which would appear in the
synthesis gas and could be removed by usual acid gas
scrubbing methods. Thus, the slag could be disposed of
as land fill. Also, part of the hydrogen needs of the plant
would be fulfilled.
Waste Water-Oil Disposal System
Liquid waste streams from the plant are combined
and treated in a waste treatment plant consisting of clari-
fication, biological treatment, sand filtration and finally,
carbon filtration. A schematic flow diagram of the waste
water-oil disposal system is presented in figure 3. No
unusual or unexpected problems have been encountered
in the operation of this equipment.
The analysis ranges of the water as it goes through-
out the treatment units are shown in table 2.
Trace Elements
In assessing all of the pollution aspects of the SRC
process, it is necessary to measure and follow the course
and fate of trace elements, particularly those of environ-
mental interest.
From previous data (ref. 1), it was shown that the
SRC product contains appreciable amounts of certain
trace elements, especially titanium, and in many cases
these constituents are in different ratios relative to their
content in the feed coal. Although titanium is not con-
sidered to be one of the more toxic elements, the signifi-
cance and impact of this element needs to be carefully
evaluated.
It is obvious that all trace elements in the coal feed
must leave the plant either in the products, or in other
gas, liquid, or solid effluents. It is not yet possible to
make complete balances due to the early stage of process
development.
The Washington State University's Nuclear Radia-
tion Center has contracted to develop a program for the
application of neutron activation analysis (both instru-
mental and chemical) and atomic absorption spectro-
scopy to the determination of trace elements in the SRC
process.
The work to date has been very preliminary. Sam-
pling techniques, sample points and shipping methods
have now been worked out. Instrumental neutron activa-
tion analyses procedures for all SRC products are now
complete. A program has now been set up to do a com-
plete balance on toxic and potentially toxic elements
throughout the pilot plant.
In some process designs for coal conversion opera-
tions, certain streams are recycled in the process. For the
SRC design, this is the case on the sour process water
from the liquefaction reactor and process solvent. It will
be apparent that if certain trace elements are collected
by a recycle stream, they will tend to build up due to
recycling since they may not be able to escape. This
could apply, for example, to volatile compounds of ar-
senic, lead, boron, and fluorine. More information is
needed to define the problem, and some provision for
separating and disposing of trace elements may have to
be added.
A review of occurrence and distribution of trace
elements in coal is given in ref. 7. Some of the problems
encountered in sulfur and nitrogen balances in the Sol-
vent Refined Coal process are discussed in an EPA report
(ref. 8).
Sulfur Plant
Some of the sulfur liberated from the coal by proc-
essing appears as hydrogen sulfide in various gas streams.
The hydrogen sulfide with carbon dioxide is separated
by an amine absorption unit. This concentrated gas is
sent to a low pressure Stretford Absorber where the
hydrogen sulfide is converted to elemental sulfur as a
byproduct or waste depending on the local market.
Problems in these conventional units have been
limited to those experienced during normal plant start-
ups. In a commercial plant, a Claus sulfur plant would
probably be used with an appropriate tail gas cleanup
unit. Proven technology with which to build a workable
188
-------
PROCESS WASTE SEWER
SULFURIC ACID
TO NEUTRALIZE
ALUM
POL YE LECTROL YTE
STEAM
AIR
SURGE
RESERVOIR
BACK WASH FROM FILTERS
RECYCLE PROCESS
WATER
WASTE
DISPOSAL
TREATER
FLOTTAZUR
WATER
QUENCH
AFTER
DISSOLVER
OXYCONTACT
BIO- UNIT
SLUDGE
[\ ISPOSAL
BY TRUCK
SAND
FILTER
CAR BON
FILTER
PROCESS WATER
INCINERATOR
(THERMAL -
OXIDIZER)
BACK WASH
SUMP TANK
TO SWAMP
Figure 3. Waste water-oil disposal system flow.
189
-------
Table 2. Average waste treatment analysis ranges
Waste
Surge Disposal Bio-Unit Plant
Reservoir Treater F10ttazur Effluent Effluent
pH 6.9-9.0 6.2-6.8 6.2-6.8 6.2-7.4 6.2-7.4
BOD, ppm 135-350 20- 11 0 4-23
COD, ppm 1000-9600 650-5000 500-4000 40-400 5-75
TSS, ppm 90-400 50-300 30-200 20-300 0-20
Phenol, ppm 30-1500 25-1100 10-1000 0.1-3.0 0.0-0.6
Extractable 10-250 6-150 4-30 0-4 0-3
Oil, ppm
sulfur recovery plant exists and will be utilized to effec-
tively remove the hydrogen sulfide from the process.
CLOSING REMARKS
Our nation will run out of oil and gas not to long
after the year 2000 and we get about three-fourths of
our energy from these dwindling sources. Therefore,
coal-derived fuels will be needed. The SRC process is
another step towards keeping the people of the world
from "freezing in the dark" without undue pollution
from the utilization of coal.
REFERENCES
1. D. L. Kloepper, T. F. Rogers, C. H. Wright, and W.
C. Bull, "Solvent Processing of Coal to Produce a
De-Ashed Product," Research and Development Re-
port No.9, Office of Coal Research, U.S. Depart-
ment of Interior, 1965.
2. "Development of a Process For Producing An Ash-
less, Low-Sulfur Fuel From Coal," Research and
Development Report No. 53, Interim Report No.2,
Volume III, Part 1, "Design of Pilot Plant," Office
of Coal Research, U.S. Department of Interior,
1969.
3. C. H. Wright and D. E. Severson, "Experimental
Evidence for Catalyst Activity of Coal Minerals,"
PREPRINT~_Di'!i!!~n of_Euel, Am. Chern. Soc.,
Vol. 16, No.2, pp. 68-91 (1972).
4. A. T. Rossano, "Air and Water Quality Assessment
Prior to Operation of The Solvent Coal Cleaning
Plant," Environment Resources Associates, Inc.,
Redmond, WA, March 1974.
5. C. R. Hinderliter, "Environmental Aspects of the
SRC Process," pp. 159-169, Report EPA-650/1-74-
118," Symposium Proceedings: Environmental
Aspects of Fuel Conversion Technology," Office of
Research and Development, U.S. Environmental
Protection Agency, October 1974.
6. C. E. Jahnig, "Evaluation of Pollution Control in
Fossil Fuel Conversion Processes-Liquefaction:
Section 2. SRC Process." Report EPA-650/2-
74-009-f, Office of Research and Development, U.S.
Environmental Protection Agency, March 1975.
7. R. R. Ruch, H. J. Gluskoter, and N. F. Shimp,
"Occurrence and Distribution of Potentially Vola-
tile Trace Elements in Coal: A Final Report," illi-
nois State Geological Survey Environmental Geolo-
gy Note No. 72, 1974.
8. Charles H. Wright, "Sulfur and Nitrogen Balances in
the Solvent Refined Coal Process," Report
EPA-650/2-75-011, Office of Research and Develop-
ment, U.S. Environmental Protection Agency, Janu-
ary 1975.
190
-------
16 December 1975
Session III:
CONTRo.l TECHNOLOGY
Rene R. Bertrand, Ph.D.
Session Chairman
191
-------
LOW AND INTERMEDIATE BTU FUEL GAS CLEANUP
C. B. Colton, M. S. Dandavati, and V. B. May*
Abstract
There is a variety of systems presently available and
under.. development to control the emissions from and
protect the turbine of a combined cycle power plant
burning fuel gas produced by gasifying coal. In the
broadest sense, these processes can be characterized as
high and low temperature processes. Many of these
processes have unique features governing their applica-
tion to specific combinations of coal gasifiers and
combined cycle power plants. This paper presents an
overview of high and low temperature cleanup processes,
and addresses the considerations involved in their selec-
tion.
INTRODUCTION
For many years, the conventional fossi I-fueled
steam electric power plant has been the mainstay of the
electric utilities. However, faced with oil and natural gas
supply problems as well as increasingly stringent environ-
mental regulations, the electric utilities are closely exam-
ining their alternatives. One such alternative is the low-
Btu gas-fired combined cycle power plant. This type of
electric generating facility offers a potential for efficien-
cies in the neighborhood of 50 percent, as well as for
significant reductions in emissions, effluents, solid
wastes, and water requirements. While it is true that the
efficiency quoted is not obtainable with today's gasifier
and combined cycle technology, a concentrated applica-
tion of engineering resources could make efficiencies in
this order of magnitude achievable within the next 10 to
20 years.
With regard to near-term implementation of this
low-Btu gasifier and combined cycle technology, it is
important to note that there are commercially available
gasifiers and low temperature fuel gas cleanup systems.
This type of electric power generation facility could be
built with today's technology and would probably be
competitive with conventional coal-fired power plants.
In addition-as was previously mentioned, this type of
power plant llliouid offer some significant advantages
over conventional coal-fired plants, particularly those
firing high sulfur coal.
*Colton and May are with Hittman Associates, Inc., Colum-
bia, Maryland 21045. Colton is Manager of Advanced Programs;
May is a Senior Engineer. Dandavati, previously with Hittman
Associates, is now with Gilbert Associates, Reading, Pennsyl-
vania.
Both literally and figuratively, the heart of the low-
Btu gas-fired combined cycle power plant is the portion
of the system that conditions the fuel gas produced by
the gasifier prior to its entry and subsequent combustion
in the combined cycle turbine. The cleanup system must
remove those constituents of the fuel gas that would be
harmful to the turbine or that would be harmful to the
environment should they be present in a turbine ex-
haust. In general, the cleanup system serves three func-
tions:
Fuel gas desulfurization
Particulate removal
NOx control
Some of these systems incorporate processes with princi-
ples of operation and design features that only allow
them to operate at low temperatures, while others may
operate at greatly elevated temperatures. Low tempera..
ture systems are arbitrarily defined as those requiring
cooling of the dirty fuel gas to 2500 F or below, whereas
high temperature systems require little or no cooling of
the dirty gas as it comes from the gasifier.
LOW TEMPERATURE DESULFURIZATION
PROCESSES
Low temperature processes can be subdivided into
four categories, according to the principle of operation.
These are:
(1) Chemical solvent processes
(2) Physical solvent processes
(3) Direct conversion processes
(4) Dry bed processes
Table 1 lists examples for each low temperature
cleanup system category, as well as data for each
example process.
Chemica! Solvent Processes
These processes employ aqueous solutions of organ-
ic and/or inorganic agents to scrub the "dirty" gas.
These agents are capable of forming "complexes" with
H2 S, CO2, and other acid gas components present in the
raw gas stream. The "complex" is then decomposed
during regeneration at elevated temperatures, thereby
releasing the acid gases for fu rther processing and recov-
ery. The regenerated solution is recycled for further
absorption. These processes may be subdivided into
those based on amine scrubbing solutions, and those
based on alkali scrubbing solutions. Most of these proc-
esses exhibit little or no selective absorption of H2S over
CO2,
193
-------
Table 1.
Low temperature cleanup processes
Basis: 8400 tons/day Illinois No.6 Coal fed to BCR Gasifiar, or 6700 ppm of Influant H1S
Process Alnorbent Type of Temp. PN.re Efficiency of S Remon Absorbent Form of Stltus
Alnorbent 0 F Chmcteristics Sulfur
Recovery
%H,S In. Effluent Life R"en- Setectivity Mike up
fluent H,S tion tOMrd rete
ppm
Chemicel
solwnt type
1. MEA Monoetha-- Aqueous 8Oto Insensitive Thermal Forms non.. 50 to As H,S
nolamine solution 120 to variation 99 -100 regen. camp" 100% gu Commercial
in pressure with COS,
CS,
2. DEA Diethanol Aqueous 100to Insensitive Thermal Absorbs CO" A. H,S
amine solution 130 to variation 99 -100 doe. not < 5% g.. Commercial
in preuure absorb
CDS, Cs,
3. TEA Trieth&- AQueous 100to Insensitive Thermal H,S As H,S
nolamine solution 150 to variation 99 -100 < 5% 91' Commercial
in pr&ISUr8
4. Alkazid Potassium Aqueous 70'0 Insensitive With H,S A. H,S
dimethyl solution 120 to variation 99 -100 steam 91' Commercial
amino in preaure
acetate 1-80atm
5. Benfield Activated Aqueous 150to /tJs ~nlim- With H,S A. H,S
potassiu m solution 250 99 ited steam il high gal Commercial
carbonate -100 No
solution dogr..
dation
6. Catacarb Activated Aqueous 150to I "sensitive H,S With H,S - p8r~ As H,S
potassium solution 250 to variation 99 + CDS steam tial al50 < 5% ga. CommerciaJ
carbonate in pressure -100 absorbs
solution generally CDS, CS,
> 300 p.i
Physicel
sol..nt type
7. Sultinol Sulfolane Organic 80to High Low H'2 S. and A.H,S
+ solvent 120 pressu re 99 H,S presaure also absorbs 981 Commercial
DillOpro- preferred +COS heati ng COS, CS,
panoami ne -100 or with and mer~
steam capta"s
8. Solexol Poly ethyl- Organic 20'0 +'Ms H1 Salsa As H,S
ene glycol solvent 80 99 absorbs 98' Commercial
ether -100 COS
9. Rectisol Methanol Organic <0 99 -100 H,S Commercial
solvent
Direct
convenion
10. St,et- N82 CO, + Alkaline 50 '0 Eisman..
ford anthrequin solution 99.9 -10 H,S 100% t.1 Commercial
one sui. sulfur
fonic acid
11. Town.. Triethylene Aqueoul 150to 99.9 -10 H,S Elemen-
28I1d glycol solution 250 tal
sulfur
Drybed type
12. Iron Hydrat.d Fixed 70 to H,S H2 Sand Eleman-
sponge Fe,0:! bed 100 \ 99 +COS also toward. tal Commercial
-100 CDS, CS, sulfur
and mer..
captanl
194
-------
Physical Solvent Processes
These processes use organic solvents to remove acid
gases by physical absorption, rather than by chemical
reaction. The extent of absorption is directly propor-
tional to the partial pressure of the acid gas components.
These processes are best-suited to high-pressure gas treat-
ing where appreciable quantities of acid gases are
present. The solvent is then regenerated by heat and/or
pressure reduction, thereby releasing a concentrated
stream of acid gases and a recyclable solvent. These
processes exhibit a selective absorption of H2 S over
CO2. I n addition to removing H2 S and CO2, these
processes are all capable of removing COS, CS2, and
mercaptans without solvent degradation.
Direct Conversion Processes
These consist of two types of processes:
(1) Those based on oxidation reduction reactions;
and
(2) Those based on the stoichiometric reaction of
H2S with S02 in the presence of a solvent.
I n the first type, H2 S is absorbed in an alkaline
solution containing oxidizing agents. The H2S is then
oxidized to elemental sulfur by air feeding to the regen-
erator and the sulfur product is separated from the
regenerated solution by froth flotation. Partial removal
of COS, CS2, and mercaptans is also possible.
The second group of direct conversion processes are
those in which H2 S is absorbed in a solvent and con-
verted to elemental sulfur by the Claus type reaction
with S02 .
2H2S + S02 --7 3S + 2H20
The solvents are usually aqueous solutions of organic or
inorganic agents.
Dry Bed Processes
These are based on absorption of acid gases by a
fixed bed of solid absorbent. Due to their low absorbent
loading, they are best-suited to removing small quantities
of acid gases. These processes can be subdivided into the
historical iron oxide processes and the various molecular
sieve processes.
Factors Affecting Low Temperature Cleanup System
Selection
The following factors emerge as being the most
important in selecting low temperature desulfurization
systems most appropriate for treating fuel gas:
(1) Su Ifur removal capabil ities, not on Iy with
respect to H2 S,. but also other sulfu r com-
pounds such as COS and CS2 .
(2) Selective absorption of sulfur compounds over
carbon dioxide. The latter need not be removed
from low-Btu fuel gas intended for use in
advanced power cycles, therefore, its absorp-
tion is undesirable since it represents an
increased operating load on the cleanup system.
(3) Type of absorbent insofar as the treated fuel
gas may contain entrained or volatilized solvent
which could be detrimental to downstream
system components such as turbine blades, etc.
(4) The system's tolerance to other contaminants
present in the raw fuel gas such as ammonia,
cyanides, phenols and tars.
(5) Overall energy requirements and operating
costs.
HIGH TEMPERATURE DESULFURIZATION
PROCESSES
(1 )
Several high temperature desulfurization processes
are currently under development. A list of some of the
more noteworthy processes is shown in table 2. None of
the processes has been commercialized. However, the
Bureau of Mines Sintered I ron-Oxide Process and the
Consolidation Coal Company's Half-Calcined Dolomite
Process are relatively advanced in their development and
may result in commercialization sooner than the others.
The principle underlying high temperature desulfuriza-
tion is the formation of metal sulfides by chemical reac-
tion of the absorbent with sulfur compounds in the gas
at high temperatures. The extent of sulfur removal
depends on the chemical equilibria for the particular
system. The following three high temperature desulfuri-
zation processes will be discussed in this presentation
since they are generally representative of various proc-
esses under "development:
(1) IGT-Meissner Process;
(2) Babcock and Wilcox I ron-Oxide Process; and
(3) Air Products Process
IGT-Meissner Process
This process is being developed by the I nstitute of
Gas Technology in conjuction with its U-Gas Process.
This process, still in the conceptual stage, utilizes a
splashing molten metal-gas contact to remove H2 S from
the gas. The contact results in the formation of a metal
sulfide which is then decomposed electrolytically to
release H2S and regenerate the molten metal for recycle.
The operating temperature is 900°F and a high sulfur
removal efficiency (98 percent) is achieved. The exact
composition of the molten metal absorbent is proprie-
tary information. The estimated energy requirements
195
-------
(0
(j)
,
Process Absorbent Type of Temp. Pressure Efficiency of S Absorbent Form of Energy Status
Bed 0 F Removal Characteristics Sulfur Required
- Recovery
%HzS In- Effluent life Regenera- Selec- Make up Elec. Oth-
fluent HzS ation tivity rate kw er
ppm toward stu
1. Bureau Sintered Fixed 1 000 to Insensitive -95 -350 >174 With air H.S, <5% As SO. Pilot
of Mines pellets of bed 1500 to variation cycles cas gas
Fe. 03 (25%) in pressure Wt loss
and fly ash <5%
2. Babcock Fe. 03 Fixed 800 to Insensitive -99 -75 As Experi-
and bed 1200 to variation 12-15% mental
Wilcox in pressu re SO. gas
3. Consolid. Half calcined Fluidized 1 500 to -200 psia -95 -350 1 0-13% H.S, 1% of As H. S 96.360 Pilot
Coal dolomite bed 1800 H. S removal with cas circula- gas to
is high at steam tion Claus
low pressure and CO. rate process
4. Air prod- Calcined Fixed 1600 to Insensitive mini- 80-90% H.S, As H. S Aban-
ucts dolomite bed 2000 to variation mum with cas gas to doned
in pressure 5-6 steam Claus
cycles and CO. process
5. Battelle Molten Solution 1100 to Atmospheric -95 -350 With H.S, As H. S Pilot
North- carbonates 1700 H. S removal steam cas, gas to
west (15% CaC03) is high and CO. fly ash Claus
at low process
pressu re,
5-6 psig
6. 1ST - Molten metal Splashing 900 -98 -150 Elec- H.S, 9830 Concep-
Meis- (proprietary) contact troly- cas tual
sner tic
Table 2. High temperature cleanup processes
Basis' 8400 tons/day Illinois No 6 Coal Fed to BCR Gasifier or 6700 ppm of Influent HzS
-------
given in table 3 are preliminary (ref. 1). Further develop-
ment is being directed toward establishing mass transfer
rates.
Babcock and Wilcox Process
The Babcock and Wilcox Process is similar to the
Bureau of Mines' process in that it utilizes iron oxide to
remove H2S from the gas at high temperatures. The
difference lies in the material used by the two processes.
While the Bureau of Mines' process uses a sintered mate-
rial made from iron oxide and fly ash, the Babcock and
Wilcox process starts out with carbon steel and generates
an iron oxide scale on the steel surface which is then
used as the desulfurization agent. Briefly, the process
chemistry is described by the following reactions:
Fe/FeOx + H2 S -+ Fe/FeSx + H2 °
At some point in time, all of the available iron oxide
scale is converted to the sulfide scale. At that point, the
system is regenerated with air as follows:
Fe/FeSx + Air -+ Fe/FeOx + XS02
The overall process accomplishes two things:
(1) It concentrates sulfur from the raw gas to 10 to
13 volume percent S02 in the regenerant gas.
(2) It provides S02 in the regenerant gas that is
either oxidized to sulfuric acid or reduced to
elemental sulfur.
A sulfur removal efficiency greater than 90 percent
is claimed by the developers. Absorption can be carried
out at temperatures as low as 6750 F; however, higher
temperatures are desirable for effective regeneration. If
regeneration is performed below 10000 F, the sulfide is
oxidized to FeS04 and not to FeOx' Also, higher
temperatures help activate the surface by developing a
thick iron oxide layer over which effective absorption
occurs; hence, operation is usually at temperatures in
excess of 10000 F.
The concentration of H2S in the desulfurized gas
increases as the volume of gas desulfurized on a given
iron-oxide scale increases. Therefore, the hardware
design for desulfurization and regeneration is one that
(ref. 2):
(1 )
(2)
Has a large number of compartments at various
stages of regeneration to give an average H2S
concentration in the fuel gas, relatively
independent of the regeneration cycle; and
(2) Gives a maximum S02 concentration in the
regenerant gas.
The hardware that has been designed uses a number
of compartments for sulfur removal and the so-called
cou nte r-current principle of air regeneration. The
desulfurizer uses a modified regenerative type air heater
and is referred to as the "regenerative desulfurizer," a
schematic of which is shown in figure 1. The cylindrical
unit is segmented into 16 compartments. Each compart-
ment is filled with carbon-steel plates oriented longi-
tudinally with the gas flow. The vessel itself will be
constructed from high alloy 5teel.
(3)
Table 3. Comparison of operating conditions for
fluidized bed combustion and gasification
Operating
Conditions
Fluidized Bed
Gasification
Fluidized Bed
Combustion
Pressu re (atm)
Temperature. F
Gas flow (lb gas/
Ib fuel)
Projected dust
loading prior to
gas cleanup
(gr /scf)
Projected particle
size
10 to 20
1 0 to 20
1600 to 1800
1500 to 1700
-12.5
-5.5
10 to 30
1 0 to 30
10 to 25% < 10J.L
5 to 15% < 5J.L
10 to 25% < 10J.L
5 to 15% < 5J.L
197
-------
RAW GAS
REGENERANT
GAS
AIR
1
PRODUCT GI\S
Figure 1. Regenerative desulfurizer.
(Source: N. J. Kertamus, "Removal of H2S on Oxidized Iron"
Babcock & Wilcox Research Center)
198
-------
Desulfurization
Raw gas is passed downward through 13 of the 16
compartments where desulfurization occurs on the
surface of the carbon-steel plates. The desulfurized gas
issues from the base of the unit and is routed to the
combustion device.
Regeneration
The sulfided iron surface is converted back to the
oxide in three of the 16 compartments. The regeneration
air passes in and upward in the first compartment to a
crossover, then downward for a second pass, and upward
for a third and final pass. At two revolutions per hour,
each of the 16 compartments is regenerated twice per
hour. Air enters the first regenerated surface accom-
plished in the second and third pass downstream. At the
end of the first pass, the O2 concentration is well below
21 percent. During the second pass, the O2 concentra-
tion is further reduced while S02 increases. Purging the
third (most FeS fouled) compartment with a gas con-
taining a minimum concentration of O2 and a maximum
concentration of S02 insures a maximum S02 concen-
tration of the final regenerant gas. The regenerant gas
should contain 10 to 13 volume percent S02'
The effect of temperature on the sulfur concentra-
tion in the desu Ifurized gas is shown in figure 2. This
may also be represented by the amount of H2S removed
as a function of the temperature. This is referred to as
"sulfur pickup" (figure 3).
The process concept has been demonstrated on
bench scale equipment and a hardware design has been
developed. The process has yet to be demonstrated on a
large scale.
Air Products Process
This process, now abandoned, employed a fixed-bed
of fully calcined dolomite to absorb H2S from the raw
gas. The sulfided dolomite was then regenerated with
steam and carbon dioxide before being recycled to the
absorber. Poor regenerability of the sulfided dolomite
led to the abandonment of this process by the Air
Products Company.
Factors Affecting High Temperature Cleanup System
Selection
Several factors affecting selection of high tempera-
ture cleanup systems emerge as the most important in
selecting high temperature desulfurization systems for
fuel gas cleanup:
(1) Operating temperature;
(2) Capability for removing sulfur compounds,
COS, CS2, as well as H2S; and
(3) The form in which the sulfur is removed; H2S,
S02, or elemental sulfur. Elemental sulfur is
the preferred form since it can be stored with-
out significant pollution problems.
(4) Regenerabil ity of the absorbent without sub-
stantialloss of activity.
(5) Overall energy requirements and operating
costs.
From a qualitative comparison based on the above
factors, the Bureau of Mines and the Babcock and
Wilcox Iron-Oxide processes appear well suited for use
with first-generation gasifiers. These processes are suited
for sulfur removal at temperatures below 1500°F,
preferably around 1000° F, which is the operating range
for first-generation fixed-bed gasifiers. Off-gas from a
high temperature second-generation gasifier would
require cooling to the operating temperature of the
iron-oxide process. This would result in a lower thermal
efficiency than for integrated systems using the dolomite
based processes, such as the Consol process. A disadvan-
tage of the iron-oxide process is the recovery of sulfur as
sulfur dioxide. I n order to convert this to elemental
sulfur, part of the sulfur dioxide must be reduced to
hydrogen sulfide, and this step consumes fuel. The IGT-
Meissner process, when developed to the point of
commercialization, should also be applicable to first-
generation gasifiers, since its operating temperature is
about 900°F. The efficiency of sulfur removal for the
IGT-Meissner process is estimated at 98 percent, while it
is selective toward both H2 Sand COS over CO2 .
As previously mentioned, the second-generation
gasifiers could employ the Consol dolomite process
which has an operating temperature of 1500°F and
above. The Battelle molten salt process also operates at
temperatures around 1500°F, but its sulfur removal
capability is questionable, particularly in the high pres-
sure range.
PARTICULATE REMOVAL SYSTEMS
Particulates of varied sizes, shapes, and composition
are: a major contributor to air pollution; a health
hazard; and the target of statutory limitations. In addi-
tion to their effect on human health, particulates
adversely affect pollution control efforts by fouling
catalysts used for S02 reduction, sulfur recovery, and
NH3 decomposition, etc. Particulates in fuel used for
firing gas turbines may cause erosion and/or corrosion of
turbine blades. The need for particulate removal from
gas streams where they are present in significant quanti-
ties cannot be overemphasized.
In order to understand the requirements that must
be met by particulate removal systems, an understanding
199
-------
.05 -
-
c
W
lL.I .04
LL.
:z:
.....
~
-
-
u
:z:
0
u .03
V)
N
::I:
lL.I
~
o::x:
n':
lL.I
:>
o::x:
.02
.01
600
SPACE VEL= 2000 - 2500 v/v-HR
END POINT= .10% SULFUR
--1
700
800
900
TENPERATURE or
lOGO
1100
1200
Figure 2. Sulfur concentration vs. temperature.
(Source: N. J. Kertamus, "Removal of H2S on Oxidized Iron"
Babcock & Wilcox Research Center)
200
-------
13.0
LLJ 12.0
u
c(
LL
~
::> 11.0
V)
LL
a
N 10.0
t-
1.1-
a 9.0
a
po-
.........
0
LLJ 8.0
:>
C>
:E:
UJ
0: 7.0
V)
N
:x:
r 1'\
-- V.V
L'-
U
V)
........ 5.0
.
...J
a
:> 4.0
3.0
2.0
600 700 800 900 1000
14.0
END POINT:
SPACE VEL:
0.10'; SULFUR
2000-2500 v/v Hr.
11 00
1200
TEMPERATURE OF
Figure 3. Sulfur pickup.
(Source: N. J. Kertamus, "Removal of H2S on Oxidized Iron"
Babcock & Wilcox Research Center)
201
-------
of what might be encountered in the way of particulate
size distributions and loadings from the various gasifiers
must be developed. I n this regard, particulates from
fixed-bed, fluidized bed, and entrained bed coal gasifica-
tion will be characterized from a qualitative standpoint
in this discussion.
The primary differences between the three gasifica-
tion methods lie in:
(1) The manner in which the coal feed is sup-
ported;
(2) The rate of gas flow (superficial velocity);
(3) Temperature; and
(4) Feed size distribution.
Gas flow rate is the least in a fixed-bed and greatest
in an entrained bed. The maximum temperature that can
be used in any particular bed depends on the caking
properties of the coal fed, the surface area of the coal
particles, etc. The surface area is much higher in flu-
idized and entrained beds than in fixed-beds, therefore,
higher temperatures can be employed. Temperatures can
generally be raised as the fluidizing/entraining velocity is
raised and, hence, are highest in entrained beds and
lowest in fixed-beds. Fixed-beds use coarser (heavier)
feeds, whereas fluidized and entrained beds use finer
feeds.
Considering the three types of beds individually,
and using temperature, gas flow rates, and feed size as
parameters, the following qualitative analysis can be
made of the particu late quantity and size distribution in
the fuel gas from each.
Fixed Beds
With fixed-bed gasifiers and their relatively low gas
flow rates, the coarser particles tend to settle down, so
that the particles carried away by the gas tend to be the
finer ones. These gasifiers operate at temperatures (up to
1500°F) lower than ash slagging temperatures. There-
fore, the ash does not slag and agglomerate. This leads to
a higher ash loading in the gas then if the ash were to
slag and agglomerate. The coarse feed (""'1/5"), on the
other hand, tends to reduce particle entrainment to
some extent. The net effect of the three factors, there-
fore, would be to yield a fairly high particulate loading
of finer particles comprised of ash and possibly un-
burned carbon.
Fluidized Beds
Gas flow rates of fluidized beds are higher than for
fixed-beds. Therefore, there is a greater tendency for the
bigger particles to be carried away by the gas. Particulate
loading in the product fuel gas is also abetted by the
finer feed size used in fluidized beds (minus 200 mesh).
However, temperature has an opposing effect. The
higher temperatures encountered cause the ash to
agglomerate. Therefore, the ash drops out of the bottom
of the gasifier as slag. This effect reduces the amount of
ash carried over with the gas. The net effect is that the
fluidized bed yields a slightly reduced particulate loading
of comparatively larger particles than off a fixed-bed
'gasifier. This is illustrated by the particle size distribu-
tion curves (figures 4-7) for fluidized bed pyrolysis used
in the COED Process. It should be noted that conditions
in the four pyrolysis stages, to which the curves cor-
respond, change gradually from fixed-bed type condi-
tions in the first stage to fluidized bed type conditions in
the fourth, with a corresponding increase in average
particle size going from the first stage to the fourth.
A quantitative approximation of particulate loading
and size distribution for fluidized bed gasification was
obtained by Westinghouse researchers using a theoretical
model developed by Kuniiand Levenspiel (ref. 3) and by
comparison with the corresponding figures for fluidized
bed combustion. Projected operating conditions for
fluidized bed combustion were deduced (by Westing-
house researchers) from experimental data obtained
from the National Coal Board (NCB) of England and
EPA contractors. The projected conditions for fluidized
bed combustion and fluidized bed gasification are given
in table 3.
The projected figures for fluidized bed gasification
were derived by Westinghouse from the corresponding
figures for fluidized bed combustion and the theoretical
model of Kunii and levenspiel taking into consideration
the following differences:
(1) The gasification process has less gas flow; less
than one-half the air flow.
(2) Fluidizing velocity in the coal gasification sys-
tem design is about 4 fps, which is one-fourth
to one-half that for fluidized-bed combustion.
(3) Ash will essentially be concentrated and
removed in the gasification process as agglom-
erates or slag, and not carried out of the system
in the fuel gas.
Based on a qualitative assessment of the differences
in the combustion and gasification systems noted above,
it is reasonable that a particulate removal system which
is designed to handle the dust loading from the fluidized
bed combustion process will be able to cope with that
from the fluidized bed gasification process.
Entrained Beds
Entrained bed gasifiers have high flow rates and
operate at high temperatures (3000° F) with fine feed
sizes (minus 200 mesh). In this flow regime the entire
feed is carried over; therefore, the particulate loadings
202
-------
100 I
I
90 l
~ 80 '
-'
N
(I)
0 1
~ 70~
I-
<:!
I-
(f). 60
w
>
0 \ 1
(0 50
G:
~-
:-~ ,
\
lU 1'1'" \
u -tv - ...,
c.: \ I
I
1:..J I
CL \
I- 30 -
\ I
-- I
, I
(? ,
w \ I
~ 20 -
t
I I
\
i !
10 -1
\ I
\
o I I I I
I 2 5 10 20 50
.
MI CRONS D!AMETER
Figure 4. Size distribution of fines from first stage char.
(Source: IIChar Oil Energy Development," R&D Report No. 73, OCR,DOI)
203
-------
100 1
80 - .J
l'J 80 -
!',! l
(/)
0
LLJ 70
1--
c~
L:
en
hJ 60 \
:>
0
c:,
<1: 50 \
I-
;::~ -'
w
u £10;-
f,: '
w \
0..
J- 30 \
.x
~
l.!..J
~ 20 -
-
10
.--1 - J. \.
o .
-
I 2 5 Ie. 20 50
MICRONS Ol,".~/IET[R
Figure 5. Particle size distribution of fines from second stage char.
(Source: "Char Oil Energy Development," R&D Report No. 73, OCR, 001)
204
-------
100
---~"-... £....__._--.-_.
I I
;~
I
~
~o
w eo
N
U)
a 70
llJ
..--
fi
U) 60
0,
I
-~ I I '
-
2 5 10 ~O
MICRONS DU\~J CT [f~
\
\ l
\ I
\ ~
, I
\ J
\ !
\ I
\j
.1
.
~O
I..!J
'..,.
6
01 C. 0
-------
1°°1
90r
l;.J 80~
~ I
~ 70~
.-
~
(/)
hJ 60 r
> I
~ 5°1'
Z
lLJ
~ 40r
.-- 30-
:L
o
lLJ
::: 20-
-,
I~"'=:\
\
10
o
I
I
2
1 J
:> 10
M I C RON SOl r~ ~..o E T!: R
20
Figure 7. Particle size distribution of fines from product char.
(Source: "Char Oil Energy Development," R&D Report N. 73, OCR,DOI)
206
---,
i
.J
I
I
I
i
_I
, .....
\ i
\ I
, i
\ ~
\ I
~ I
\1
\ I
~
1
:. r-
..IV
-------
are very high, with the particle size distribution propor-
tional to the feed size.
Other Factors Affecting Particulate Production and
Removal Systems Selection
In all three coal gasification systems discussed
above. the particle size distribution and particulate load-
ing are affected by several other factors such as the ash
content of the coal, the method of coal preparation, the
change in particle density due to chemical reactions,
attrition in the bed, and decripitation of particles due to
their history. The effects of these factors cannot easily
be quantified.
Allowable particulate loadings and size distributions
that consider gas turbine requirements as well as air
pollution emission regulations have to be considered for
the design of particulate removal equipment. Westing.
house (ref. 4) has indicated that particulate loading in
the fuel gas being fed to the current generation of gas
turbines be limited to no more than 0.01 grains/SCF for
particles larger than 211, with the total loading not to
exceed .15 grISCF. Particles less than 211 are expected to
pass through the turbine without impacting the blades.
Many low temperature particulate removal systems
are commercially available and descriptions of these are
abundant in the literature. Only high temperature sys.
tems have been considered for this discussion.
Several high temperature and high pressure particu-
late removal systems either available or currently under
development are listed in table 4. The table includes
different types of high temperature particulate removal
systems and shows the operating conditions and effi-
ciencies attainable with each.
New Devices
Particularly promising is the moving bed granular
filter being developed by the Combustion Power
Company (ref. 5). This filter uses a moving sand bed to
trap particulates. The filter developer claims a high effi.
ciency (>90 percent) for particle sizes down to 211, that
the associated pressure drop is relatively low (10 to 15 in
H20), and that it can be used at high temperatures
(1400°F). A current problem area is the disintegration
of the filter media. The use of metallic and ceramic
media instead of sand is being investigated by Combus-
tion Power Company. Figures 8 and 9 show schematics
of the granular filter.
Other promising devices are sonic agglomerators and
ceramic filters. These are still largely in the develop-
mental stage. Among the commercially proven devices,
cyclones and tornadoes are effective at high tempera-
tures, and are being used in coal gasification pilot plants
and process development units.
A literature search will reveal that data on particu-
late loading and particle size distributions in the off
gases from gasifiers are sorely lacking. Considerable
effort needs to be directed toward the acquisition of
such data if the details of particulate removal systems to
protect combined cycle turbines are to be developed.
NOxCONTROLSYSTEMS
Nitrogen oxides, collectively referred to as NOx' are
an important group of air pollutants. The term NOx
refers primarily to NO (nitric oxide). although similar
quantities of N02 (nitrogen dioxide) and N20 (nitrous
oxide) may also be formed. These oxides are intercon-
vertible, and the equilibrium between them depends on
the action of sunlight, the presence of oxidizing agents,
temperature, pressure and other factors. Gas turbines,
like other combustion engines, are potentially a major
source of NOx emissions. NOx is formed in the hot
combustion zones of engines. Two known primary
mechanisms are responsible for NOx formation in
combustion engines:
(1) Thermal NOx: Thermal NOx is formed by the
reaction of atmospheric N2 and O2 in the hot
combustion zone within the engine. This is the
dominant mechanism when relatively clean
fuels are burned in the engine. Removal of NOx
from flue gases is an extremely difficult prob-
lem. However, it is possible to control the
thermal NOx formation by several techniques,
some of which are:
(a) Primary zone leaning by modified combus.
tion chamber design;
(b) Water injection;
(c) Exhaust gas recirculation; and
(d) Reduced turbine inlet temperature.
Each of the above techniques results in a lower
peak temperature within the combustion zone,
thereby reducing thermal NOx formation. The
above techniques have their drawbacks, the
most important of which is reduced turbine
efficiency.
(2) NOx from fixed nitrogen in the fuel: This
source of NOx is important only when dirty
fuels such as coal and residual fuel oil are gasi.
fied. Gas produced from these fuels may
contain organic nitrogen compounds such as
ammonia, hydrogen cyanide, and pyridines.
Ammonia is the primary nitrogen compound,
207
-------
Table 4. High ~mperature particulate removal systems
I\J
o
en
Type of ramoval Manufac- Capacity CoDection Minimum Maximum Maximum Maximum Applicable Pressu re Status
system tu rer ACFM efficiency particle operat- operat- collection dust load- drop
% size with ing ing pres- efficiency ing range in.W.G.
efficiency temp. 0 F sure atm % grainslSCF
> 50%,/J.
Mechanical
Collectors
Cyclones Buell 50.000 80-90 5-10 1400 2 90-95 4-40 Commercial
Ducon 58,000 80-90 5-10 1500 10 90-95 4-40 Commercial
Tornado Aerodyne 30,000 93-97 0.5 1500 10 98 <30 30 Commercial
Bed Filters
Granular Combustion >90 2 1400 >90 10-15 Under
Power Co. Development
Ducon Under
Development
Panel C.U.N.Y. Under
Development
Rex Rexnord 20,000 95-99 > 900 1 > 99 <40 4-15 Commercial
Sonic
Agglomeration
Collection
Systems
Alternating
Velocity Under
Precip itator Braxton Development
Scrubbers
Fused salts Battelle Under
Development
Filters
-
Metal and Selas and > 99 <0.5 2000 1 >99 Commercial
ceramic others
Electrostatic
Precipitators >99 < 0.5 800 1 > 99 <1 Commercial
-------
i
RETURN
CLEAN
GAS
\t
::-".:
.0': :
"..,..
I . .
. .
.' 01'
. ,
. .
. .;, . .
. :.
6' ID -1 .~> ~ i>~4-,"S"CTJD" &
I .<:. :.;.:}LJ INSTRUHENTATION
27',3'f~DmTI }:'ll~;! ~"mCTORY
'.:. :'~-:YI I NSIJLATTON
",:" 12 IN THICK
:::'::J.' 1,:."Jr SAKD
I' '/:"I
. --:-:'!'.I
..' --- ',0.' I
":. :..:.::....:.1.r,
-Udi rty filCC '--1
; 73 ft/"dn. , ---.-.
I . . :;,~-I-U
. ' '.~': ,;.: . ',':~:!
. G". "
_T_. .=J' -:.:':.:-,
" r, '. ".
-' : ." {,: .: ::Jr:-- SUP P:~'. . .:' Ulouvre
,,;,,, :"/
:...:: . I
.:.}. !:1
: ',':' I' ,
,',..': A:'~"
L. :~;
D :;)'
;'::.{j ''1.'
:~::..~:,~.. .
_L~..-~i~
DRAIN ..-'
10'
Figure 8. Pilot plant granular filter concept,
(Source: Combustion Power Company, Inc.)
209
-------
SAND
DRAIN
Figure 9. Model
(Source: Comb' granular filter.
ustlOn Power C
ompany, Inc.)
210
-------
while the others are present in smaller concen-
trations. If retained in the gas, these com-
pounds are oxidized during combustion to
NOx' The nitrogen compounds are removed by
water scrubbing when a low temperature clean-
up system is used. To prevent this carry-
through to the turbine, under high temperature
cleanup-ammonia and other nitrogen com-
pounds also must be removed from the gas at
elevated temperatures. Methods to accomplish
this were investigated.
A potential method to remove ammonia is the
decomposition of ammonia into stable elemental nitro-
gen and hydrogen at elevated temperatures. The
decomposition of ammonia is governed by the following
reaction:
k1
2NH3 ~ N2 + 3H2
k2
The equil~rium constant k1 increases as the
temperature increases and the total pressure decreases.
Temperature has a greater effect than pressure as is veri-
fied by the fact that k1 is practically constant over the
pressure range, 1 to 50 atmospheres. Hence the higher
the temperature, the greater the decomposition of
ammonia into nitrogen and hydrogen. The equilibrium
percentage of ammonia in a 3:1 hydrogen gas mixture is
shown as a function of pressure and temperature in
figure 10.
The equilibrium constants for ammonia formation,
kp = ~ I are tabulated in table 5 (ref. 6) for some tem-
peratures and pressures.
The equilibrium constant for ammonia formation is
seen to be very low at high temperatures and low pres-
sures. Under the conditions present in the BCR gasifier
(1700°F, 475 psia), the equilibrium ammonia concentra-
tion is 200 ppm. However, the actual ammonia concen-
tration is 4260 ppm. This suggests that low equilibrium
concentrations, although thermodynamically favored,
are not kinetically feasible. Therefore, the decomposi-
tion kinetics must be aided by a catalyst.
A literature search will yield considerable recent
work in the development of catalysts for ammonia
decomposition. General Motors (ref. 7) and Ford Motors
(ref. 8) have addressed the ammonia decomposition
problem with the goal of ridding automotive exhausts of
poisonous nitrogen oxides by first reducing it to ammo-
nia and further decomposing the ammonia to elemental
nitrogen and hydrogen. Research in this direction led to
the development of several catalysts suitable for ammo-
nia decomposition. Among these are Ni, Pt, W, Mo, Re,
(4)
and Ru. Of these catalysts, a Cu-Ni-AI203 catalyst was
seen to have the highest activity for ammonia decompo-
sition. The extent of ammonia decomposition over
Cu-Ni-AI203 catalyst as a function of temperature is
shown in figure 11.
Unfortunately, the above catalysts have a serious
drawback. They are poisoned by even trace quantities of
sulfur compounds present in the feed gas (ref. 9). The
poisoning is due to the formation of a metal-sulfide
wh ich deactivates the catalyst. Generally the metal/
metal-sulfide equilibrium favors the formation of the
metal-sulfide at low temperatures and favors its
decomposition at high temperatures. As an example, the
poisoning by sulfur compounds of Ni catalyst used in
methanation reactions is shown in figure 12. The poison-
ing effect is seen to diminish only at high temperatures
(~2000° F). Effectively, this is the temperature above
which, for example, a Cu-Ni-AI203 catalyst could be
used to decompose ammonia. It is impractical, however,
to use the catalyst at this temperature for two reasons:
(1) At such high temperatures, sintering would
significantly reduce the activity of the catalyst.
(2) In all the gasification systems considered in the
study on which this paper is based, where fuel
gas ammonia is significant, fuel gas tempera-
tures approaching 2000°F are unattainable.
The only commercial catalyst potentially capable of
decomposing ammonia is an iron oxide catalyst com-
posed of five percent Fe203 mounted on high tempera-
ture fired inert alumina spheres. This catalyst could
simultaneously remove H2 S from the gas stream, thus
combining the ammonia and sulfur removal operations
in a single step. However, the operating conditions neces-
sary for this catalyst are not known and must be deter-
mined before the catalyst becomes acceptable (ref. 10).
SUMMARY
In conclusion, it is seen from a state-of-the-art
study, that low temperature cleanup systems (operating
temperature below 250°F) are commercially available
and highly efficient from a pollutant removal stand-
point. The fuel gas produced by a fixed bed (low tem-
perature) gasifier contains particulates, condensible tars,
phenols, ammonia and other organics. The most conven-
ient scheme for removing the condensibles and particu-
late involves a liquid quench followed by the separation
and recycle of the condensed tar to the gasifier in order
to improve thermal efficiency. If a high temperature
cleanup system were considered for such an application,
it would be faced with the currently unsolved problems
associated with high temperature removal of the tars and
211
-------
100
200.C
90
80
70
C"")
= 60
:z;
I-
Z
w 50
u
a:::
w
a..
40
N
-
N
30
20
10
0
PRESSURE A rl"osP HE RES
Figure 10. Equilibrium percentage of NH3 in a 3:1 hydrogen-nitrogen gas mixture.
(Source: G.S. Mitchell, "Ammonia Its Production From Natural Gas," Lion Oil Company)
-------
100
o
---
I
800'-
~ 60L
~
-
"'--
a
~
<:
c::::
~ 40
I\J -.
.... w
w u
--
a
u
l""')
:I:
-. 20
I
I J I I I I I '
. I
0 lQO 200 300 400 500 600 ?GO 800
CATALYST TH1P °C
Figure 11. Ammonia decomposition over Cu-N i-A12 03 catalyst (8.1 % Cu, 9.2% N i).
(Source: R.L. Klimisch and K.C. Taylor, "Ammonia Intermediacy as a Basis For
Catalyst Selection for Nitric Oxide Reduction," Envir. Sci. & Tech. Vol. 7, No.2
(Feb. 1973).
-------
V)
c:!"
~.
(!.J
M
~
o
o
r-
"-
0:::
::J
Lt..
....J
::J
V)
t.!'
1'00. 0
10.0
1.0
"
.01
--1
200
J
800
I
1000
I
400
-1-
600
TEHPE RA TU RE. °c
Figure 12, Poisoning of N i-catalysts used for adjustment of equilibrium:
CO + 3H2 ~ CH4 + H2 O.
(Source: H. Pichler, "Advances in Catalysis" Vol. 4, Academic Press)
214
1200
-------
Table 5. Equilibrium constants for
ammonia formation
TF
p = 10 atm. p = 30 atm. p = 50 atm.
kp kp kp
0.0266 0.0273 0.0278
0.0129 0.0129 0.0130
0.00659 0.00676 0.0069
660
750
840
ammonia in the fuel gas. Thus, a low temperature clean-
up system is a logical choice for the majority of the
gasifiers-specifically fixed-bed gasifiers, like Lurgi, being
considered for commercial application. However, a con-
siderable thermal efficiency penalty is associated with
the use of the low temperature cleanup systems in com-
bination with coal gasifiers and combined cycle power
plants. This penalty has spurred the recent development
of high-temperature cleanup systems which make it
possible to attain higher thermal efficiencies for the inte-
grated plants. High-temperature cleanup systems are not
yet commercially available, and those under develop-
ment have yet to prove their effectiveness and reliability
on a commercial scale.
ACKNOWLEDGMENT
The information on which this paper is based was
drawn from work carried out by Hittman Associates,
Inc., under contract to the United Technologies Re-
search Center in support of their efforts on EPA Con-
tract 68-02-1099. Some of the information presented in
this paper was adapted from previous efforts by the
Foster Wheeler Corporation, in their support of the first
phase of EPA Contract 68-02-1099. The authors
particularly wish to acknowledge the guidance and sup-
port provided by the United Technologies Research
Center project team headed by Dr. F. L. Robson.
REFERENCES
1. Letter communication, November 15, 1974. Dennis
Duncan, Institute of Gas Technology, to M. S.
Dandavati, Hittman Associates, Inc.
2. Babcock and Wilcox Research Center, "Removal of
H2S on Oxidized Iron," Kertamus, N.J.
3. D. Kunii and O. Levenspiel, Fluidization Engineer-
ing, John Wiley and Sons, Inc., 1969.
4. "Clean Power Generation From Coal," R&D Report
No. 84, OCR, DOl.
5. Letter communication, December 10, 1974. R. A.
Chapman, Combustion Power Company, Inc. to C.
B. Colton, Hittman Associ~es, Inc.
6. S. H. Maron and C. F. Prutton, Principles of Physi-
cal Chemistry, The Macmillan Company, New York,
1958.
7. R. L. Klimisch and K. C. Taylor, "Ammonia Inter-
mediacy as a Basis for Catalyst Selection for Nitric
o x ide Reduction," Environmental Science and
Technology, Vol. 7, No.2, February 1973.
8. M. Shelef and H. S. Gandhi, "Ammonia Formation
in Catalytic Reduction of Nitric Oxide by Molecular
Hydrogen," Industrial Engineering Chemistry, Vol.
11,No.1,1972.
9. H. Pichler, Advances in Catalysis, Vol. 4, Academic
Press.
10. Personal communication with Roy Wensink, The
Harshaw Chemical Company.
215
-------
THE BENFIELD ACTIVATED HOT POTASSIUM CARBONATE PROCESS:
COMMERCIAL EXPERIENCE APPLICABLE TO FUEL CONVERSION TECHNOLOGY
D. H. McCrea*
Abstract
Activated hot potassium carbonate absorption
systems designed and licensed by The Benfield Corpora-
tion are widely used for CO2 and sulfur compound
removal. More than 400 installations are operating or are
under construction worJq-wide. Many are purifying gas
streams similar to those that will be generated by the
new fuel conversion processes.
Benfield technology has been thoroughly evaluated
for use with these new processes. Detailed designs and
recommendations have been prepared for specific pro-
posals in cooperation with major engineering firm licens-
ees. Several research institutes have prepared more gener-
al technical and economic evaluations, with Benfield's
cooperation, for EPA, ERDA, TV A, and other govern-
mental and private organizations. Some of their reports
are available to the public. Benfield reports oqtlining
process performance and economics are also available.
While the applicability of Benfield technology to
fuel conversion processes has been established, each
application requires individual evaluation. Integration of
purification into the overall process and meeting envi-
ronmental standards require careful consideration.
BENFIELD ABSORPTION PROCESSES
Description
The Benfield Activated Hot Potassium Carbonate
Process is a liquid absorption system for the removal of
CO2 and sulfur compounds. Absorption is accomplished
in conventional packed or trayed towers, normally at an
elevated pressure, and at temperatures up to about 2700
F by contact with an aqueous solution containing 25 to
35 percent K2 C03 along with small quantities of an
inorganic corrosion inhibitor and an organic or inorganic
activator to catalyze the rates of absorption and desorp-
tion. Absorbed components are recovered in a second
tower by contact with a stripping gas, usually steam
supplied from a process gas or steam heated solution
reboiler. The arrangement is illustrated in figure 1. In
practice, however, other configurations are often used to
achieve higher degrees of purification, to satisfy specific
*The author is with The Benfield Corporation, Pittsburgh,
Pa.
process requirements, and to minimize capital and oper-
ating costs. Some of these modifications have been
discussed in earlier Benfield reports (refs. 1-5).
Removal and recovery of CO2, H2 S, mercaptans
and other acidic components occur due to reversible
chemical reactions with K2 C03 :
K2C03 + CO2 + H20 = 2KHC03,
K2 C03 + H2 S = KHC03 + KHS,
. K2 C03 + RSH = KHC03 + KRS.
The organic sulfur compounds, COS and CS2, do not
react directly with K2 C03, but are instead hydrolized to
H2S and CO2:
COS + H20 = CO2 + H2S,
CS2 + 2H2 0 = CO:?: + 2H2S.
Hydrolysis is catalyzed by the solution and proceeds
orders of magnitude faster than in water. Aside from
acidic gases and gases that can hydrolize, the compounds
found in raw gas are, with few exceptions, unreactive
with the solution.
Background
The hot potassium carbonate process was developed
by Benson, Field, and coworkers at the U. S. Bureau of
Mines (now a part of ERDA) in the early 1950's (refs.
6-9). Their work was prompted by the need for an
improved process to purify moderate to high pressure
coal-derived synthesis gas and demonstrated the techni-
cal feasibility of the process as well as its advantages over
other available technology. For many applications, the
hot carbonate process requires much less thermal energy
than amine scrubbing systems (ref. 10) and significantly
lower capital investment than processes using physical
solvents (ref. 11). Other advantages, compared to either
amine or solvent scrubbing, include use of nonvolatile,
lower cost, chemically stable solution, reduced solubility
of process gas, and higher temperature operation.
The economics of the hot carbonate process has
been improved and its range of applicability extended
through research conducted by The Benfield Corpora-
tion. Significant developments include:
1. Introduction of more effective corrosion inhib-
itors and control methods (ref. 12) which allow
mostly carbon steel construction.
2. Development of activators to improve kinetics
and alter equilibria relationships (ref. 13) so
that tower sizes can be reduced, higher gas puri-
ties obtained, and thermal efficiency increased.
3. A multi zone tower design (ref. 14) that reduces
217
-------
PURIFIED
GAS
'..
Q
REGENERATED
GAS
Figure 1. Simplified flow diagram of the Benfield activated hot
potassium carbonate process (single stage design),
4.
absorber diameter and lowers investment cost
for high pressure service.
Flowsheet modifications (ref. 15) to improve
thermal efficiency.
A process configuration (Benfield HiPure Proc-
ess) (ref. 4) by which purified gas containing
less than 1 ppm Hz Sand 10 ppm COz can r"
obtained.
6.
A process configuration permitting selective
absorption of sulfur compounds from an excess
of COz .
Characterization of performance parameters for
tower packings at commercial operating con-
ditions (ref. 16) allowing more accurate designs
and reduced tower sizes.
7.
5.
218
-------
Experience
Benfield technology is offered worldwide through
licensing agreements with most of the major engineer-
ing-construction firms. More than 400 Benfield designed
systems have been built or are under construction. Prin-
cipal applications have been purification in the produc-
tion of:
Town Gas
Ammonia
Hydrogen
SNG from naphtha
Ethylene oxide
Vinyl acetate
Benfield systems have also been used for the purification
of natural gas and of gases produced by partial oxidation
or coal gasification. Experience in these applications is
of particular interest in evaluating use of Benfield tech-
nology in the new fuel conversion processes since the gas
streams are, in many respects, similar.
The Benfield Process was used for over 13 years to
treat raw gas made from coal by the Lurgi Process at
Westfield, Scotland (ref. 17). Contaminants in the gas
were not harmfully reactive with the solution, there was
no significant corrosion, and efficient removal of both
H2 S and organic sulfur was obtained. Similar success has
been achieved in eight applications treating sulfur-
containing gas produced by partial oxidation of heavy
oil and in 11 applications where natural gas containing
both CO2 and H2 S is being purified. In one of these, the
H2 S content ofthe gas is being reduced from 12 percent
to less than 1 ppm.
Operating data for some of these units are summa-
rized in table 1. While the table illustrates conditions at
which Benfield technology has been demonstrated, it
does not indicate the degree of gas purity obtainable. In
most cases, the units were designed to most economi-
cally meet predetermined criteria rather than to achieve
maximum gas purity.
EVALUATION OF BENFIELD TECHNOLOGY
FOR USE IN FUEL CONVERSION PROCESSES
Scope of Studies
The integration of Benfield technology into fuel
conversion processes has been extensively studied in
cooperation with engineering firm licensees, for many
specific proposals. These have included:
Production of pipeline gas from coal or heavy oil
Production of liquids from coal
Production of low and medium Btu gas from coal or
heavy oil
I n situ coal gasification
Oil shale processing
Geothermal power generation
Fuel cell processes.
Studies conducted for licensees are, of course, confiden-
tial. Some of the results, however, have been included in
a Benfield report (ref. 5) that can be used to estimate
capital cost and utility consumption, thus serving as a
guide to process developers in making preliminary evalu-
ations of specific applications.
A number of independent evaluations of fuel con-
version processes including Benfield purification have
also been conducted. Benfield has been privileged to
cooperate in these studies which have been mainly spon-
sored by governmental research organizations. A partial
list of evaluations is shown in table 2. In at least some
cases, reports are publicly available.
Conclusions
Utilization of Benfield technology in the new fuel
conversion processes does not appear to present a major
engineering challenge at least in satisfying the process
needs of gas purity and reliability. All of the require-
ments have been demonstrated on an industrial scale.
1. Trace components expected in the raw gas have
been successfully handled without undo chem-
ical or mechanical problems.
2. Pressures expected are in the range of industrial
experience. Large system~ have operated at
1100 psig.
3. Ability to treat the large volumes of gas ex-
pected has been demonstrated. For example,
the gas produced in a 250 MM SCFD based
SNG plant can be treated in three Benfield
trains without exceeding commerciaHy demon-
strated equipment size. Scale-up to a two-train
plant should present no problems.
4. Achieving required gas purity from CO2 and
H2 S has been demonstrated. Organic sulfur re-
moval to the level needed for catalyst protec-
tion has not been achieved industrially because
there has been no need. However, commercial
results have confirmed the design parameters
for organic sulfur removal and have demonstrat-
ed that levels now required are obtainable.
In satisfying the requirements for specific applica-
tions, the principal design goals are usually to:
1. Minimize capital investment
2. Achieve maximum thermal efficiency by best
utilizing available low level heat
3. Assure stable operation and avoid the possibi-
lity of CO2 or sulfur breakthrough in the event
of plant upsets
4. Satisfy environmental emission limits
Even though process objectives can readily be
achieved and design goals have been reasonably well es-
tablished, specifying the purification system still requires
219
-------
Table 1. Partial list of Benfield units purifying natural gas or gas
produced by partial oxidation or coal gasification
Raw Gas Pure Gas
Appl ication Capacity Pressure Analysis Analysis
MMSCFD psig CO2 H2S CO2 H2S
Natural Gas 150 1100 7% 16% 1% .05%
Natural Gas 165 1000 12.3% .8% .5% 20 ppm
Natural Gas 330 1000 16% 50 ppm 1% <1 ppm
Natural Gas 31 600 42% 12% .1% <1 ppm
Natural Gas 18.5 600 7% 1% .2% 4ppm
Natural Gas 60.5 600 24% 20 ppm 1% nil
Partial Oxidation 21.9 370 33% .06% .4% <1 ppm
Partial Oxidation 6.4 140 26% 1% 110 ppm <1 ppm
Partial Oxidation 52.1 440 5.2 .93% .95% 45 ppm
Coal Gasification 23.9 325 28% .6% .8% 70 ppm
ingenuity and careful study. The many evaluations for
specific projects conduct~d by Benfield and by others
have demonstrated that:
1. Purification must always be considered as an
integral part of the overall process. Location
and characteristics of the purification step or
steps can significantly influence the selection,
cost, and reliability of both upstream and
downstream units.
2. Environmental criteria will influence the eco-
nomics of the purification system and may even
determine the process sequence specified.
Two examples, illustrating the effects of sulfur
emission standards on design, are given in the concluding
section of this report.
INFLUENCE OF EMISSION STANDARDS
ON SPECIFYING BENFIELD SYSTEMS
Production of Pipeline Gas
One design situation often encountered is the gasi-
fication of a high sulfur coal by a process producing a
relatively low quantity of CO2 so that the CO2 to sulfur
ratio in the raw gas is less than about 8 to 1. A process-
ing sequence as illustrated in figure 2 would be specified.
By removing sulfur and CO2 prior to shift conversion, a
gas suitable for Claus processing, which is more economi-
cal than systems suited to dilute streams, can be used for
final sulfur recovery. With this sequence, the most eco-
nomical approach is to design the first stage of Benfield
purification for nearly 100 percent H2 S removal but not
to attempt complete organic sulfur removal. The residual
organic sulfur is removed in the second stage of Benfield
purification and appears as H2 S, typica"lIy 20 to 50 ppm,
in the CO2 stream. If this H2S level is not permissible,
the increased investment cost of complete sulfur removal
in the first Benfield stage must be compared to recover-
ing sulfur from the CO2 vent stream or for oxidizing the
H2 S to S02 .
Production of Fuel Gas
The effect of sulfur em ission standards on the in-
vestment cost and thermal efficiency of Benfield purifi-
cation systems is more easily seen in cases where sulfur
removal is only an environmental requirement and is not
required for the process. One case evaluated recently was
selective desulfurization of fuel gas produced by partial
220
-------
Table 2. Partial list of independent studies of fuel conversion
process which include Benfield purification
Organization
Subject
Sponsor
Air Products and Chemicals
Babcox and Wilcox
Battelle Memorial Inst.
Battelle Memorial Inst.
Beard Engi neeri ng
Bituminous Coal Research
Braun
Chern Systems
City COllege of City
University of New York
Dravo
Exxon Research and Eng.
General Electric
Institute of Gas Tech.
Institute of Gas Tech.
Koppers
NASA
Procon
Swindell-Dressler
TVA
U.S. Bureau of Mines
U.S. Bureau of Mines
Westinghouse Electric
OCR
EPRI
EPA
TVA
Muni ci pa 1 ity
OCR
OCR-AGA
OCR
City College of City
University of New York
ERDA
ERDA
EPRI
OCR
EPA
OCR
NASA
OCR
OCR
EPRI
Bureau of Mines
Bureau of Mines
OCR
221
Pipeline gas
Low Btu gas
Pipeline gas
Low Btu gas
Combined power cycles
Pipeline gas
Pipeline gas
Pipeline gas
Combined power cycles
Pipeline and low Btu gas
Pipeline gas
Fuel Cells
Pipeline gas
Low Btu gas
Pipeline and low Btu gas
Combined power cycles
Pipeline gas
Low Btu gas
Low Btu gas
Low Btu gas
In situ gasification
Combined power cycles
-------
Coal
°2orH2
Gasilicalion
C02 + H2S Remova-l
H2S + C02 to
Claus Process
Shill Conversion
C02 Removal
C02
Mdhanalion
Drying
SNG
Figure 2. Benfield purification in an SNG
from coal process: acid gas removal
before and after shift conversion.
oxidation. Investment cost, energy requirement, and
dilution of the H2S stream for Claus processing are all
dependent on the degree of desulfurization required (see
table 3).
As emission restrictions become more stringent,
both the investment cost and energy needs of the Ben-
field system increase. In addition, the H2 S content of
the feed stream to elemental sulfur recovery decreases,
thus increasing the cost and reducing the efficiency of
this unit. The data for the 40 and 50 ppm designs show
the effect of a change in unit configuration trading off
investment for reduced energy consumption.
When producing a low Btu gas, perhaps 125
Btu/NCF, the energy consumed for purification is a sig-
nificant portion of the total available and is strongly
dependent on emission standards. The imposition of
standards more stringent than for an equivalent direct
coal-fired system with flue gas cleanup could affect the
economic feasibility of projects for fuel gas production.
REFERENCES
1. D. H. McCrea and H. E. Benson, "Benfield Processes
for SNG and Fuel Gas Purification," 165th Ameri-
can Chemical Society National Meeting, Dallas,
Texas, April 8-13,1973.
2. R. W. Parrish and J. H. Field, "The Benfield Process
in Coal Gasification," 24th Annual Gas Condition-
i ng Conference, The University of Oklahoma,
Norman, Oklahoma, March 14-15,1974.
3. R. W. Parrish and H. B. Neilson, "Synthesis Gas
Purification Including Removal of Trace Contami-
nants by the Benfield Process," 167th American
Chemical Society National Meeting, Los Angeles,
California, March 31-April 5, 1974.
4. H. E. Benson and R. W. Parrish, "HiPure Process
Removes CO2 /H2 S," Hydrocarbon Processing, 53,
4, pp 81-2.
5. D. H. McCrea and J. H. Field, "The Purification of
Coal Derived Gases: Applicability and Economics of
Benfield Processes," 78th American Institute of
Chemical Engineers National Meeting, Salt Lake
City, Utah, August 18-21, 1974.
6. H. E. Benson and J. H. Field, "Method of Separat-
ing CO2 and H2 S from Gas Mixtures," U. S. Patent
2,866,405, May 1959.
7. J. S. Tosh, J. H. Field, H. E. Benson, and W. P.
Haynes, "Equilibrium Study of the System Potas-
sium Carbonate, Potassium Bicarbonate, Carbon
Dioxide, and Water," Bureau of Mines Report of
Investigations 5484, 1959,23 pp.
8. J. S. Tosh, J. H. Field, H. E. Benson, and R. B.
Anderson, "Equilibrium Pressures of Hydrogen Sul-
fide and Carbon Dioxide over Solutions of Potas-
sium Carbonate," Bureau of Mines Report of Inves-
tigations 5622,1960,25 pp.
9. J. H. Field, H. E. Benson, G. E. Johnson, J. S. Tosh,
and A. J. Forney, "Pilot Plant Studies of the Hot
Carbonate Process for Removing Carbon Dioxide
and Hydrogen Sulfide," Bureau of Mines Bulletin
597, 1962, 44 pp.
10. Benfield Corporation, "The Way to Low Cost
Scrubbing of CO2 and H2S from Industrial Gases,"
1970,11 pp.
222
-------
Table 3. Influence of emission standards on purification
cost and utility requirement
Sulfur in Purified Gas
ppm
H2S in Claus Feed
%
Investment
$/1000 NCFD Feed
Energy Required
Btu/NCF
750 22.9 12.53 6
210 19.7 13.13 10
100 15.35 11
50 17.16 15
40 17.1 29. 16 13
11. Benfield Corporation, "Low Cost Reliable Systems
for the Removal of Acid Gases from Ammonia
Synthesis Gas Trains Based on Partial Oxidation of
Fuel Oil:' 1974, 13 pp.
12. Joseph H. Field and Donald H. McCrea, "Method of
Removing Carbon Dioxide From Gases:' U.S.
Patent 3,863,003, January 1975.
13. Joseph H. Field, "Separation of CO2 from Gas
Mixtures," U.S. Patent 3,907,969, September 1975.
14. H. E. Benson, "Separation of CO2 and H2S from
Gas Mixtures," U. S. Patent 3,563,695, Feb. 1971.
15. H. E. Benson, "Separation of CO2 and H2S from
Gas Mixtures:' U. S. Patent 3,823,222, July 1974.
16. L. J. P. Evans and O. M. Siddique, "C02 Removal
Studies and Their Application to SNG Plant Conver-
sions:' IGE Autumn Research Meeting, London,
England, Nov. 25-26, 1975.
17. T. S. Ricketts, "The Operation of the Westfield
Lurgi Plant and High Pressure Grid System:' The
Institution of Gas Engineers 100th Annual General
Meeting, London, England, May 14-17, 1963.
223
-------
COAL CONVERSION PROCESS
WASTEWATER CONTROL
w. A. Parsons and R. A. Ashworth*
Abstract
Wastewater contributions are generated from
various unit operations incorporated in the process
sequence of coal conversion plants. Such discharges are
potentially highly pollutional unless wastewater control
is engineered into the production facility. This paper
identifies the principal wastewater generating processes
and describes the magnitude and characteristics of the
effluents expected from various coal conversion proc-
esses. Process-oriented effluent control alternatives are
presented that exploit the prospects for process revision,
recovery, recycle and serial reuse. Treatmen~riented
effluent control concepts are analyzed for applicability
to wastewater disposal situations. Technical problems
associated with the implementation of wastewater con-
trol are discussed. The expected performance of the
wastewater control concepts are evaluated in terms of
contributions of pollutants and residuals.
Wastewater control is a factor of great significance
in the siting, engineering and operation of major coal
conversion plants. Maximum water conservation will be
mandated in the design of coal conversion plants located
in coal reserve areas with limited water supply. Exem-
plary wastewater control will be required to obtain
NPDES permits for plants discharging process effluent or
contaminated runoff to natural watercourses. Water
quality limited situations could result in site specific
effluent requirements. In addition, it is to be expected
that applications for discharge permits will receive close
scrutiny by fish and wildlife agencies, conservation
groups and the public, relative to possible environmental
impacts from the array of pollutants generated by coal
conversion processes. Protracted delays in project imple-
mentation have occurred as a result of these consider-
ations.
Stringent wastewater control requirements pose
challenges of legion proportions to design engineers
because of uncertainty as to: (a) the wastewater charac-
teristics, (b) the performance of wastewater control
technology and (c) the discharge permit requirements.
Wastewater characteristics will differ with the particular
coal conversion process selected for application and with
the characteristics of the co_al. f~~d. Few modern, high
*Dr. Parsons is Director, Corporate Environmental Control;
Mr. Ashworth is Assistant Director, Alternate Fuels Develop-
ment, Arthur G. McKee and Company, Cleveland, Ohio.
pressure, coal conversion plants are available to provide
definition of wastewater characteristics, but some infor-
mation is available from foreign installations and from
experimental units. In cases where a preview projection
must be made, expediency may require the speculative
synthesis of wastewater characteristics based upon some-
what analogous effluents from byproduct coke plants.
Information on the performance of process-oriented
and treatment-oriented wastewater control technology is
sparse - particularly at the stringent levels of abatement
associated with new point source discharge standards.
Information on the performance of wastewater control
technology at byproduct coke plants is available for con-
ceptual guidance in analogous situations, but presently
demonstrated control technology will rarely meet new
point source discharge standards without additional
process development. Since process development re-
quires identification of problem areas and screening of
alternatives, the object of this paper is the presentation
of a generalized overview of sources of. wastewater plus
process oriented and treatment-oriented alternatives for
wastewater control.
SOURCES AND CHARACTERISTICS
OF WASTEWATER
Storm-Generated Wastewater
Wastewater control involves consideration of
process wastes and contaminated runoff. A generalized
schematic illustrating wastewater sources from coal con-
version processes is given as figure 1, where it is evident
that wastewater control commences at the coal supply
and storage areas. Runoff from coal receiving, unloading
and from coal processing areas may require collection
and treatment prior to discharge - especially if receiving
waters are water quality limited. The expected charac-
teristics of the runoff are undefinable inasmuch as they
would vary with the site and storm situations. It can be
presumed that the runoff would contain traces of oils
and phenols characteristic of runoff from railroad yards,
plus trace levels of suspended solids and reduced sulfur,
characteristic of coal pile drainage. The runoff from the
process area would also qualify as contaminated runoff
inasmuch as it would be expected to contain trace con-
centrations of byproduct chemicals such as phenols,
ammonia and oils.
Drainage from the coal pile is another source of
storm-generated wastewater. A distinguishing feature of
225
-------
COAL SUPPL Y
So
STORAGE
RECEIVING AREA
RUNOFF
COAL PIl.E
DRAINAGE
I'J
I'J
0)
COAL
PREPARA TlON
COAL DRIER
SCRUBBER
PURGE
ACCESSORY PROCESSES
COOLING
TOWE RS
BLOWDOWN
STEAM
PLANT
1
...
B LOWDOWN
FLUE GAS
SCRUBBER
PURGE
COAL
CONVERSION
PROCESS
CONDENSATES
LIGHT OIL
PLANT WASTE
DESUlFURIZATION
UNIT PURGE
GAS PRODUCTS
UPGRADING
(PIPELINE
GAS)
C02
SCRUBBER
PURGE
PRODUCT
WATER
Figure 1. Mainstream process schematic.
PRODUCT
DEWATERING
CONDENSATES
LIQUID PRODUCTS
UPGRADING
(LIQUEFACTION)
,
PRODUCT
WATER
FUEL
-------
coal pile drainage is that it may be expected to contain
substantial concentrations of pollutants. The grab
sample analysis given in table 1 by C. W. Rice (ref. 1)
suggests that coal pile drainage differs from coal washery
waste in that concentrations of dissolved solids are high
relative to concentrations of suspended solids.
The wastewater control significance of contami-
nated runoff becomes very great when regulations are
based upon extreme value storm probabilities such as the
24-hour storm with a recurrence interval of 10 years. In
their analysis of a 14,250 TPD Synthane Process plant,
Kalfadelis and Magee (ref. 2) estimated design flows for
coal pile drainage, process area runoff, and yard ru noff as
5,000 gpm, 15,000 gpm,and 60,000 gpm respectively. If
5 inches of precipitation are taken as the representative
storm for the Eastern coal fields, it can be projected that
the volume requirement for containment of the runoff
from the subject plant would amount to about 42
million gallons. The areal requirement would be about
20 acres. Storage requirements for Western coal fields
would be less inasmuch as the 10-year, 24-hour storm
would be expected to be in the range of from 1 to 2.5
inches.
Process Wastewater
The process wastes from coal conversion plants as
illustrated by figure 2 are process specific in respect to
volume and characteristics. Thus, the wastewater charac-
teristics listed in table 1 should be considered as more
illustrative than applicable. However, the condensate
from the primary scrubber ca.n be recognized as being of
particular significance since it is a major flow that con-
tains substantial concentrations of ammonium salts of
chlorine and sulfur. The primary scrubber condensate
also contains free ammonia, phenol, and cyanide com-
pounds. Other condensates may contain volatile pollut-
ants such as ammonia, phenol, or sulfur compounds, but
are low in concentration of salts which improves the
prospects for practical reuse after treatment. Purge flows
from desulfurization, light oil refining, or product up-
grading may contain high salt concentrations and, there-
fore, tend to possess limited prospects for reuse after
treatment.
CONTROL ALTERNATIVES
Storm-Generated Wastewater
Wastewater control measures for stormwater com-
mence with positive exclusion of uncontaminated runoff
from contaminated runoff in order to minimize handling
and storage problems associated with the vast flows. If
minimum treatment is required, e.g., removal of traces
of oil film, full flow devices similar to hay filters may be
more feasible than storage concepts. If more sophisti-
cated treatment is required-such as removal of suspend-
ed solids or phenol, control concepts providing for
storage in conjunction with treatment will normally be
employed out of consideration of process kinetic fac-
tors. Units such as variable level aerated ponds can be
used for lowering trace levels of deoxygenating pollut-
ants and suspended solids. The maintenanc3 of culture
activity during dry periods is a potential problem. I n the
case of concentrated flows such as coal pile drainage,
sealed collection and storage facilities may be required
to limit percolation. Underdraining of sealed areas has
been advocated by some parties to verify the integrity of
sealing procedures.
Process Wastewater
Wastewater control for process flows illustrated in
Hgure 2 commences with at-the-source separation of oils
and tars for separate collection and handling. Segrega-
tion of process wastes for recovery of ammonia and sul-
fu r byproducts is common practice. Segregation of
wastes is indicated for recovery of phenol, but segrega-
tion is unnecessary for destruction of phenol by bio-
logical or other means.
Reuse of wastewater flows is a universal objective of
designers of modern process plants. I n coal conversion
processes, secondary condensates and other flows con-
taining low concentrations of salts are readily amenable
to incorporation into common reuse concepts. Paren-
thetically, primary scrubber condensate and other wastes
purged from processes because of accumulations of salts
or incompatible constituents are not readily amenable to
incorporation into common recycle concepts. Primary
scrubber condensate is the dominant salt bearing process
waste and controls the characteristics of the effluent
from the ammonia stripping operation. The effluent is
corrosive. Recycle systems at byproduct coke plants
have experienced pitting of 316 stainless steel; fiber re-
inforced plastic is a preferable material for low tempera-
ture service. Materials screening tests led to the use of
Inconal water chambers and titanium tubes in a demon-
stration plant evaporator (ref. 4). Simplistic assignment
of the effluent to recycle could result in significant re-
visions to materials specifications for high pressure
service. The effluent from a lime still stripping operation
will be saturated with calcium-sulfur compounds and
will possess scale-forming tendencies upon subsequent
chemical or biochemical oxidation of reduced sulfur to
sulfate. Hence recycle alternatives are unattractive, and
dilution below potential saturation is indicated in order
to maintain viable treatment operations. Decreased
biomass activity in the activated sludge process has been
attributed to precipitation of calcium sulfate upon
bio-oxidation of reduced sulfur.
Sui fur removal immediately following primary
227
-------
PRIMARY
SCRUBBER
COAL
C
REACTOR
GAS I FICA TlON
OR LIQUEFACTION.
I'.)
""
\XI
I' I
STEAM 102' H2 I
I' I
I 8 I
PHENOL
AMMONIA
C'" CONDENSATE
P = PURGE
SULFUR
REMOVAL
PHENOL
REMOVAL
AMMONIA
STRIPPING
AMMONIA
ABSORPTION
HYDROCARBON
RECOVERY
AFTER
COOLER
C
LIGHT
OIL
REFINERY
r - - - - - - - --,
I I
, I
I PHYSICAL I PROCESS
r"'- CHEMICAL ,---~
I I TREATMENT I EFFLUENT
I I I
, I I
: L______--_.J
I
I
I
I
I
I
,
BTX
1- - - - - - - - ~
, I
, I
I PRODUCT I
,- -, UPGRADING I
EFFLUENT I I :
I I I I
I : I ,
I CP I 1______--..1
1_____-- - - - - ------- - -_:"-J
BIOLOGICAL
TREATMENT
PROCESS
Figure 2. Coal conversion process effluents.
-------
Table 1. Illustrative coal conversion process wastewater characteristics
Coal
Pile
Drainage
(ref 1)
Synthane
Condensate
*(ref 3)
Flow gal/ton
200
pH
Suspended Solids, mg/l
Dissolved Solids, mg/l
COD, mg/l
Cyanide, mg/l
Pheno 1, mg!l
Ch 1 ori de, mg!l
Ammonia, mg/l
Sulfur, total, mg/l
Thiocyanate, mg/l
7.6
9
412
50
1413
1995
3.2
0.4
3000
500
8500
500
*
Adapted from reference 3.
90
scrubbing per figure 2, where feasible, presents a con-
ceptual opportunity for increased reuse of secondary
condensates upon provision of independent recovery and
treatment systems. Caustic could be used for ammonia
stripping since low doses would be required. The stripper
bottoms could be used for dilution of lime still bottoms
from the stripping of primary scrubber condensate to
enable biological treatment or for makeup to the flue gas
scrubber system. The Sulfammon process is an example
of a desulfurization operation that can be placed early in
an applicable process flow sequence. The process fea-
tures selective removal of hydrogen sulfide by controlled
contact of the gas flow and the ammonia absorption
solution. It is capable of effective hydrogen sulfide re-
moval and the purge is compatible with biological treat-
ment operations. I ndependent recovery and treatment of
secondary condensates offers potential for improved
wastewater control, but the practicality is limited by the
increased cost of independent systems and by the limit-
ed volume of secondary condensates available for reuse.
Treatment-Oriented Alternatives
Biological treatment, physical-chemical treatment,
and evaporation/incineration are alternatives for waste-
water control by terminal treatment. None of the alter-
natives offers demonstrated technology capable of meet-
ing new source standards for discharge from byproduct
coke plants or presumably from coal conversion plants.
Biological treatment is capable of effective removal ofa
broad spectrum of organic pollutants (particularly
phenols) and reduced sulfur compounds. Although some
removal of ammonia is effected by both volatilization
and metabolic synthesis, the indicated performance is
generally poor because of release of ammonia upon deg-
radation of thiocyanates and amines.
Physical-chemical treatment uses activated carbon
for adsorption of phenol. Additional unit processes are
229
-------
I';)
(..)
o
WASTES MAKEUP
~io- .j~
PRIMARY WASTEWATER STACK PU R G E
SCRUBBER TREATMENT SCRUBBER ,.
...... ...~
CONDENSATE CONDENSATE VAPORS
,
AMMONIA "'- EVAPORATION ... INCINERATION
~ ,.. I ,
STRIPPING I
I
I
I
I ... ASH
.., ,
LOW VOLUME I
PURGES I r------'.
I I
I I I
AMMONIA I I EVAPORATION I
RECOVERY ....~ POND I
I I
OTHER
COOLING TOWER
'?
I I
L______J
~~I\tlapted from reference If.
Figure 3. WAL process (modified)*
-------
required for the removal of cyanides, residual ammonia
and reduced sulfur compounds. A physical-chemical
system is being installed at a byproduct coke plant that
employs adsorption of phenol by activated carbon in
conjunction with pickle liquor coagulation (ref. 5). De-
velopment studies have been made relative to upgrading
the removal of cyanides and ammonia by ion exchange.
A wastewater control demonstration plant incor-
porating evaporation and incineration of ammonia still
waste has been installed (ref. 4). The process concept is
outlined by the modified flow schematic given as figure
3. The concept of evaporation of ammonia still waste
with disposal of dry ash facilitates the engineering of
increased water reuse into coal conversion plants. How-
ever, the demonstration plant has not to date compiled a
record of sustained performance that would qualify for
status as demonstrated technology.
Summary and Conclusions
Wastewater control is a factor of prime significance
in the siting, engineering, and operation of major coal
conversion plants because of water supply constraints
and stringent effluent standards. The sources of waste-
water can be classified as stormwater generated and pro-
cess generated. Pollutants associated with the wastes in-
clude phenols, ammonia, oils, cyanides, reduced sulfur
compounds, and suspended solids. The stormwater
wastes are indicated to consist of dilute, high volume
flows from contaminated yard areas and more concen-
trated coal storage drainage. The process wastes are in-
dicated to consist of wastes requiring blowdown such as
salt bearing primary condensate and purges from unit
processes plus wastes with low salt content that are more
amenable to reuse.
A potential exists for corrosion and scale problems
upon reuse of salt bearing primary condensates and
purges from unit processes. The internal control of such
problems in modern high-pressure coal conversion
systems is viewed as particularly challenging and possibly
beyond practical attainability. Therefore, segregation
and terminal treatment of the wastes is advocated. Sul-
fur removal immediately following primary scrubbing
offers a conceptual opportunity for increased recycle of
secondary condensates. Such procedure should increase
the practicality of reuse of low salt process wastes and
collected stormwater waste after such reconditioning as
required.
Alternatives for terminal treatment or recondition-
ing include biological, physical-chemical and evapora-
tion-incineration. Biological treatment is developed to
the extent that it has been employed at a number of coal
carbonization plants for treatment of similar wastes, but
the expected performance is inadequate to meet new
source standards. Additional development is required to
define the expected performance reliability and costs of
physical-chemical and evaporation-incineration proc-
esses.
REFERENCES
1. Cyrus Wm. Rice Division, NUS Corporation, "Draft
Development Document Iron and Steel Industry,
Phase II," Addendum (August 1974). U.S. Environ-
mental Protection Agency, Washington, D.C.
2. C. D. Kalfadelis and E. M. Magee, "Evaluation of
Pollution Control in Fossil Fuel Conversion Proc-
esses - Gasification, Section I: Synthane Process,"
EPA-650/2-74-009-b (June 1974). Office of
Research and Development, U.S. Environmental
Protection Agency, Washington, D.C.
3. A. J. Forney et aI., "Analyses of Tars, Chars, Gases
and Water found in Effluents from the Synthane
Process," U.S. Bureau of Mines TPR 76 (January
1974).
4. R. Jablin and G. P. Chanko, "A New Process for
Total Treatment of Coke Plant Waste Liquor,"
Water 1973, No. 136, Vol. 70 (19741. AIChE
Symposium Series.
5. C. A. Naso and J. W. Schroeder, "A New Method of
Treating Coke Plant Waste Waters," AIME Iron-
making Proceedings, Vol. 33 (1974).
231
-------
A TAPERED FLUIDIZED-BED BIOREACTOR
FOR TREATMENT OF AQUEOUS EFFLUENTS
FROM COAL CONVERSION PROCESSES
Charles D. Scott, Charles W. Hancher,
David W. Holladay, and George B. Dinsmore*
Abstract
A bioreactor system based on a tapered fluidized
bed is under development for the microbiological
degradation of hazardous organic compounds expected
in aqueous effluent waste streams from coal conversion
processes. The concept will be used with fluidized parti-
cles having adhering microorganisms. The use of a fluid.
ized reaction medium alleviates operational problems
associated with biomass buildup and allows easy removal
or addition of the active materials. The tapered reactor
tends to stabilize the fluidized bed while allowing a
much wider range of operating conditions. Preliminary
experimental results from a study made with a 3-in.
diameter tapered fluidized-bed reactor that contained
adhering Pseudomonas bacteria and operated aerobically
indicate that the phenol content of a feed stream can be
reduced to less than 25 ppb. This concept is compatible
with multistage operation, and scale-up is considered
practical.
INTRODUCTION
The waste process and condensate liquors from coal
gasification, liquefaction, or carbonization processes are
qualitatively similar on a chemical constituent basis.
They are expected to contain relatively large amounts of
dissolved hydrogen sulfide and ammonia, as well as
various phenols, thiocyanates, and lesser amounts of
other hydrocarbons (refs. 1-4).
Some of the components of the waste stream may
be present in sufficiently high concentrations to be
recoverable. These may include H2 S, ammonia, and
phenol. Both ammonia and H2 S could be recovered by
stripping processes, while phenol recovery could be
achieved by solvent extraction (ref. 5) (figure 1). How-
ever, phenol levels in the dephenolated liquors may still
be greater than 50 ppm, well above the acceptable
concentration for wastewater, and additional treatment
will be necessary.
*The authors are staff members of Oak Ridge National
Laboratory, which is operated by Union Carbide Corporation for
the U.S. Energy Research and Development Administration, and
is located at Oak Ridge, Tennessee 37830.
Studies of bacterial treatment of both raw and
dephenolated waste liquors from gasification and car-
bonization processes have been made (refs. 4,6). In
general, activated sludge systems utilizing large, stirred-
tank reactors were used. Although residence times of
many hours were required, very low phenol levels were
achieved. To a lesser degree, thiocyanate levels were also
reduced by the activated sludge system as were other
organic compounds. Thus the processing scheme for coal
conversion wastewater treatment will likely include a
step for the biological degradation of phenols and other
organic compounds. Since allowable phenol levels will
probably be the controlling factor, the process step must
be highly efficient in removing phenol. Further, al-
though large stirred-tank reactors have been shown to
operate in this application, efficient fixed- or fluidized-
bed systems may be more compatible with the process
needs. A tapered fluidized-bed bioreactor is being evalu-
ated specifically for this application.
BIOLOGICAL DEGRADATION
The biological degradation of phenolic-type
compounds has been extensively studied and can be
represented by the overall reaction
Substrate (phenol, etc.) +
O2 microorganis~ CO2 +
increased biomass + reduced substrate.
(1 )
In this respect, the microorganism can be represented as
a catalyst for the degradation. Actually, the degradation
process is a relatively complex chemical pathway requir-
ing a whole series of enzymes to catalyze the various
degradation steps. The proper microorganism will con-
tain all of the necessary enzymes to achieve a rather
complete breakdown. Various types of Pseudomonas
bacteria appear to be best adapted for phenol degrada-
tion. An excellent example of such a metabolic pathway
was determined for Pseudomonas putida (ref. 7) (figure
2).
Biological systems using microorganisms in activated
sludge processes have been used for treating aqueous
waste containing organic compounds similar to those
expected in coal conversion processes (refs. 4,6). When
233
-------
WAST E
WATER
WASTE WATER
SURGE TANK
N
~
PURIFIED
WATER
SED I MENTATION
TANK
SLUDGE
REMOVAL
H2S
TO SULFUR
RECOVERY
H2S
TOWER
BIOREACTOR
AIR OR
OXYGEN
NH3
NH3
FRACTIONATOR
PHENOL
ETC.
PHENOL
EXTRACTION
FEED
SURGE TANK
NUTRIENT
SOLUTION RESERVOIR
Figure 1. Possible simplified aqueous waste treatment flowsheet.
-------
o
II
ALDOLASE /002H
HO ~
OH
I
OH
I
~
HYDROXYLASE.
~
PHENOL
CATECHOL
CH3- CO- COOH
+
CH3-CHO
PYRUVIC ACID
+
ACETALDEHYDE
4
.....OH
OXYGENASE
OH
I
~
COOH
o
II
'H
..
~
2- HYDROXYMUCONIC
SEMIALDEHYDE
HYDROLASE
HYDRASE
o
II
CeOOH
2-0XOPENT-4- ENOIC
ACID
Figure 2. Simplified metabolic pathway used by Psuedomonas Putida to degrade phenol.
(From ref. 7).
used with mixed cultures of microorganisms, these
systems have been shown to reduce phenol levels from
greater than 1000 ppm to less than 60 ppb, although
residence times of many hours and very large reaction
tanks were required (ref. 6). A mutant strain of Pseudo-
monas bacteria has now been developed commercially
for the degradation of dissolved carbonaceous materials.
This material, called Phenobac,* is reported to be highly
efficient for phenol degradation. The biological process
requires oxygen, which is best supplied by air or by
simply adding O2.
TAPERED FLUIDIZED-BED REACTOR CONCEPT
The usual activated sludge bioreactor utilizes a very
large tank or pond in which most of the reactive micro-
.Worne Biochemical, Inc., Frontage Industrial Park, West-
ville, New Jersey.
organisms are maintained in dilute suspension by agita-
tors or gas sparging. Such systems operate as single-stage
batch reactors. Since the concentration of the pollutant
in the effluent of such a reactor is the same as that in the
reactor (both very low), the concentration driving force
for the degradation reaction is very low. Thus, extremely
large reactor volumes are required.
Many microorganisms can remain stable and viable
while being attached to solid surfaces. The immobilized
microorganisms can be used in columnar stagewise bio-
reactor configurations using fixed beds or fluidized beds
and thus provide a much more efficient system. In the
fluidized-bed concept, the fluidized particles will have
adhering microorganisms and the aqueous waste stream
will be the fluidizing medium.
The usual fluidized-bed has a relatively narrow range
of optimum operating conditions. It is difficult to main-
tain nonfluctuating operation at optimum conditions
since high bed expansion and low stability are frequently
experienced. A new tapered reactor design is being
235
-------
investigated to circumvent this problem (ref. 8). The
tapered reactor is in the form of a cone with an entrance
which has a small cross-sectional area expanding to one
several times that of the entry point. If the cross section
of the entry is sufficiently small and the expansion is
sufficiently gradual, the flow profile throughout the
reactor will have few large eddies and thus provide fluid-
flow patterns that have minimal backmixing, especially
at the feed entry point. The taper allows a wide range of
flow rates without loss of bed material since the fluid-
izing velocity decreases with reactor height, thereby
providing a bioreactor that can effectively operate under
a variety of feed conditions.
The fluidized medium can contain a relatively large
number of microorganisms per unit reactor volume, and
the material can be easily introduced into or removed
from the reactor. Excess biomass buildup will tend to
exit the reactor as a suspension after a steady-state level
is reached. The fluidized bed is also compatible with a
discontinuous gas phase such as that required by air or
oxygen sparging.
EXPERIMENTAL
Experimental Facility
The experimental facility includes a glass, tapered
fluidized bed, the diameter of which increases from 1 in.
at the bottom of the reactor to 3 in. at the top in a
42-in. length (figure 3). Other equipment includes a feed
reservoir, a feed metering pump, a settling chamber, and
a waste reservoir. A pump is also available for recycling
part of the waste stream for additional fluidization when
needed. Since this is an aerobic bioprocess, provisions
are available for introducing O2 or air to the bottom of
the column or to a feed stream oxygenator.
Materials
The solid particles used in the fluidized bed were
anthracite coal in the particle size range of 0.0149 to
0.0177 cm. Although various microorganisms were used,
including mixed cultures from activated sludge systems,
the best results (and the experimental results reported
here) were obtained from work with the mutant strain
of Pseudomonas bacteria available under the trade name
Phenobac.
Some scrub solutions from coal pyrolysis and
hydrocarbonization were used in this study. The most
definitive results, however, were obtained with synthetic
feed solutions made from deionized water and reagent
grades of phenol and other additives. It is anticipated
that the pH of the coal conversion scrub solution may
require adjustment to 6.5-8.0 via the addition of lime
(CaCI2 was added to the synthetic feed solution); also,
the addition of some phosphate (1 :70 phosphate:
phenol) will be required for microorganism metabolism.
The source of nitrogen for metabolism will be available
as dissolved ammonia (ammonium sulfate was added to
the synthetic feed solution), and various necessary trace
metals will be present from the process water (table 1).
Table 1. Trace metals added to synthetic
phenol feed solution
Dissolved Concentration,
trace metal ppm
B 0.02
Zn 0.Q1
Mo 0.Q1
Mn 0.Q1
Cu 0.Q1
Fe 0.Q1
Analytical Technique
Phenol analyses were carried out using previously
developed procedures (ref. 9). Phenol concentrations as
low as 1 ppm were determined by using the 4-aminoanti-
pyrine colorimetric method, while levels of phenol down
to 25 ppb were determined by steam distillation fol-
lowed by extraction with chloroform and colorimetric
analysis.
Operation
The reactor was operated with approximately 2
liters of an expanded fluidized bed of anthracite parti-
cles and approximately 0.5 liter of solution above the
bed prior to the column exit. A relatively narrow range
of inlet flow rates of 374 to 520 ml/min was used. The
reactor operation was completely stable over this range,
in fact, it was left unattended for a period of several
days. I n some tests, air or oxygen was introduced at the
bottom entry point through a porous metal frit, result-
ing in a three-phase system. However, the most efficient
operation was found when the feed solution was pre-
saturated with oxygen and a third phase was not main-
tained in the reactor.
The microorganisms were introduced into the reac-
tor by circulating a suspension of the live microorga-
nisms through the reactor for 8 to 12 hr. This period was
sufficient for establishing an initial "seed" of attached
microorganism that then tended to multiply rapidly,
236
-------
EFFLUENT
I'J
(,J
-.J
RECYCLE
PUMP
SETTLING
TANK
SLUDGE
REMOVAL
EXHAUST
GAS
',\ .;\....
-....' \ \, "
~':nf
,',,","...";"
TAPERED ~~:U
FLU IDIZED BED ::-~-,
42 in. LONG ;, :-;:'
1 in, DIAM. BOT. :'>';
3 in. DIAM. TOP ,'..::
2349 ml VOLUME '('/;
':,\
\J~
AIR OR ,
OXYGEN---...J
EXHAUST
GAS
OXYG ENATOR
.~.." OR 01 LUTOR
. 0'
DOO 0 ,/
°o.'i>
00 0 '
. . . 0
o 4) . ~
fI' ,.. Q
:.'06:
.0' 0
. . .
" 0" "
. o' .
. "
" ., .'0'
. .'
o . .
. ~..
FEED
TANK
AIR OR
OXYGEN
FEED
PUMP
REACTOR
PUMP
Figure 3. Experimental tapered fluidized bed bioreactor system.
-------
reaching a steady-state condition after about one week.
This is the condition at which additional biomass forma-
tion results in aqueous suspension rather than additional
attachment. At this point, the fluidized bed can be used
for the continuous degradation of a waste stream, or it
can be stored at 4°C almost indefinitely for later use. In
the latter case, the bed material can be reintroduced to
the bioreactor and the system will be ready for opera-
tion in a very short time.
Flow rates were measured by calibrated metering
pumps and/or rotameters, and aliquots of the inlet and
effluent streams were collected for determining the
concentrations of phenol and other constituents.
A stock solution with a specific phenol concentra-
tion was used in most cases, and the phenol inlet concen-
tration was varied by utilizing effluent recycle. This
recycle mode of operation could represent the actual
method to be used for very high phenol concentrations,
or the lower inlet concentrations (after recycle mixing)
could be indicative of the lower levels expected in
instances where a phenol recovery process was used.
RESULTS AND CONCLUSIONS
The tapered fluidized-bed bioreactor is still in the
bench-scale development phase. However, preliminary
results indicate that the system will be a useful tool for
degrading phenol to very low levels.
Experimental Results
Several scouting tests were made in which a prelimi-
nary feed stream containing 450 to 4,800 ppm phenol
was continuously mixed with a recycle stream of the
reactor effluent to make up the bioreactor feed stream.
Oxygen was added either as a gas sparge (air or oxygen)
to the bottom of the fluidized bed or by saturating the
feed stream with oxygen prior to its entry to the reactor.
No attempt was made to evaluate the effect of feed
stream flow rate. There was no evidence of fluid-phase
mass transport control of the reaction. Since a single bed
volume was used in these studies, the effect of bed size
has not yet been established.
The system was found to be capable of degrading
phenol to levels less than 25 ppb with reactor residence
times as low as a few minutes. As the phenol inlet con-
centration was increased in each series of test runs, a
breakthrough point was reached where additional inlet
phenol resulted in a significant increase in the effluent
phenol concentration. With air sparging or a feed stream
saturated with oxygen at ambient pressure, a specific
conversion rate of about 6.6 g/day'liter was observed
(table 2). When the oxygen content of the system was
increased by O2 saturation of the feed stream at a pres-
sure of 40 psig, the breakthrough conversion rate was
significantly higher, i.e., about 10 g/day'liter (table 2).
Typical conversion rates in stirred tank reactors are
much less than 1 g/day'liter (ref. 6). Thus, a significant
decrease in reactor volume could be expected with the
fluidized-bed system.
Based on this preliminary data, it would seem that
the conversion rate in the tapered fluidized bed is
primarily limited by available O2 rather than by an
inherent limitation in the biological system. Conse-
quently, at high phenol concentrations, additional O2
must be added to the system by gas sparging or over-
pressure, or the feed stream must be diluted by recycle
of some of the reactor effluent.
The microorganism loading in the reactor was very
high at about 0.1 g of dried organisms per milliliter of
bed volume, and biomass production was typically 0.6 g
of dry microorganisms per gram of phenol.
Scouting tests were also made in which potassium
thiocyanate was introduced in the feed. The bioreactor
was not as efficient for thiocyanate degradation. How-
ever, a typical feed stream containing 34 ppm of thio-
cyanate was reduced to 11 ppm. Additional work is
required in this area.
Conclusions
Preliminary results from the operation of a small,
tapered fluidized-bed bioreactor for the degradation of
dissolved phenolic compounds indicate that such a
concept is potentially useful for the treatment of the
aqueous waste from coal conversion processes. The
reactor uses a multistage concept, which is at least a
factor of 10 more efficient than the more conventional
stirred-tank reactors, allowing very low effluent values.
Additional studies will be required to determine the
effects of bed height and other operating parameters,
and long-term operation with an actual coal conversion
aqueous waste will be required to confirm its utility.
Scale-up by at least a factor of 10 in column diameter is
probably practical.
REFERENCES
1.
Johnson, C. A., M. C. Chervenak, E. S. Johnson, and
R. H. Wolk, "Scale-Up Factors in the H-Sol Proc-
ess," paper presented at the 65th Annual Meeting,
AIChE, New York, 1972.
Forney, A. J., W. P. Haynes, S. J. Gasior, G. E.
Johnson, and J. P. Strakey, Jr., "Analysis of Tars,
Chars, and Water Found in Effluents from the
Synthane Process," Bureau of Mines Technical
Progress Report TRP 76, January 1974.
2.
238
-------
TaWe 2. Phenol degradation rates in tapered
fluidized-bed bioreactora
Feed streanib Effluent Reactor
phenol phenol conversion Method
flow rate, concentration, concentration,c rate, d of
ml/min ppm ppm g/day'liter oxyge"nation
409 14 0.05 2.4 Air sparge in column
425 38 6.6 Air sparge in column
412 140 100 6.9 Air sparge in column
475 9 < 0.025 1.6 Feed stream saturated
with O2 at ambient
pressure
480 17 .050 3.0 Feed stream saturated
with 02 at ambient
pressure
482 20 .050 3.9 Feed stream satu rated
with 0:1 at ambient
pressure
500 35 6.6 Feed stream saturated
with 02 at ambient
pressure
505 31 10 6,6 Feed stream satu rated
with 02 at ambient
pressure
374 30 < 0.025 4.6 Feed stream saturated
with 02 at 40 psig
388 58 < 0.050 9.3 Feed stream saturated
with 02 at 40 psig
391 63 0.50 10.2 Feed stream saturated
with 02 at 40 psig
aAIl runs were made at ambient pressure, 25 1. 'rC, pH 7.0.-7.2.
bin all tests recycle effluent was used with a primary feed stream containing 450 to 4800 ppm of phenol.
cMost sensitive assays of less than 1 ppm had a sensitivity of 0.025 ppm. The less sensitive assay had a
sensitivity of 1 ppm.
dlncluded the volume of the fluidized bed, as well as the volume of solution above the bed and the volume
of the settling chamber.
~39
-------
3.
EI Paso Natural Gas Company, "Second Supplement
to Application of EI Paso Gas Company for a
Certificate of Public Convenience and Necessity,"
Docket No. EP73-131, October 1973.
Ashmore, A. G., J. R. Catchpole, and R. L. Cooper,
"The Biological Treatment of Carbonization Efflu-
ents - I. Investigation into Treatment by the
Activated Sludge Process," Water Research, 1 :605,
1967.
Lauer, F. C., E. J. Littlewood, and J. S. Butler,
"Solvent Extraction Process for Phenols Recovery
from Coke Plant Aqueous Waste," Iron and Steel
Engineer YearBook, p. 315,1972.
Davis, W. R., "Control of Stream Pollution at the
Bethlehem Plant," Iron and Steel Engineer Year
4.
5.
6.
Book,p. 785,1972.
7. Sala-Trepot, J. M., K. Murray, and P. A. Williams,
"The Metabolic Divergence in the Meta Cleavage of
Catechols by Pseudomonas Putida," Eur. J. Bio-
chern. 28, 347 (1972).
8. Scott, C. D., C. W. Hancher, and S. E. Shumate, II,
"A Tapered Fluidized Bed as a Bioreactor," pre-
sented at the 3rd International Conference on
Enzyme Engineering at Reed College, Portland,
Oregon, Aug. 3-8, 1975 (proceedings to be pub-
lished by Plenum Press).
9. Standard Methods for the Examination of Water
and Wastewater, 12th Ed., Am. Public Health
Assoc., New York, 1965.
240
-------
CLIMATIC EFFECTS ON WASTEWATER TREATMENT*
Stanley L. Klemetson, Ph.D.t
Abstract
The development of coal gasification process to pro-
duce synthetic natural gas (SNG) will create potential
pollution problems. The sources of pollutants at the site
and basic constituents of the waste streams are dis-
cussed. The wastewaters are generally treated in several
typical physical, chemical, and biological treatment
processes.
The principal factors of the climate-temperature,
solar radiation, heating days, wind roses, precipitation,
and snowfall, are discussed in their relationship to waste-
water treatment processes. The effects of temperature
variations on the biological reaction rates, BOD, re-
movals, and detention time are presented. It was
recommended that more concern be placed on the effect
of climatic conditions on wastewater treatment design.
INTRODUCTION
There is an energy crisis in the United States. One
solution to this problem is seen in the conversion of coal
to a clean fuel by the use of a coal gasification process.
A primary concern is that the treatment and/or conver-
sion process that generates the clean fuel does not itself
become a major pollution source. While the potential
pollutants can be expressed in any or all of the three
possible states of air emissions, solid wastes, and liquid
*This work was carried out as part of Grant No.
ENG7510251 from the National Science Foundation, and as a
Visiting Professor, I ndustrial Environmental Research Labora-
tory, EPA.
tAssistant Professor, Department of Civil Engineering
North Dakota State University, Fargo, North Dakota.
A I R POllUTI 011
COAL
GASIFICATION
PlAHT
SOLID
WASTES
LIQUID
WASTES
Figure 1. Potential states of pollutants.
effluents (figure 1), all of them ultimately contribute to
the wastewater effluents of the plant and its site.
Energy production utilizing a coal gasification
process to produce high-Btu synthetic natural gas (SNG)
with a heating value of about 970 Btu/SCF will become
a reality in the United States within the next few years.
The enactment of increasingly stringent environmental
legislation requires that information be made available
on the types of wastes to be generated, the environ-
mental effects of these wastes, and the best available
treatment technology. Since coal gasification plants are
net consumers of water, it is also important to assess the
wastes and the treatment methods from the standpoint
of water reuse potential. The proposed plants will be
operated under widely different climatic conditions, and
this could have a significant effect on the choice of
wastewater treatment processes. These aspects of control
technology are summarized in table 1.
A variety of processes has been developed to pro-
duce both high-Btu and low-Btu synthetic natural gas
(table 2). Each of the gasification processes has a differ-
ent configuration of unit operations, but the resulting
wastewater effluents are somewhat similar. Since the
Lurgi gasification process is currently being planned for
several areas in the Northern Great Plains, the discussion
will be limited to this process. However, much of the
information will be applicable to the other processes
also. The purposes of this paper are presented in table 3.
GASIFICATION PROCESS
The Lurgi coal gasification plants planned for con-
struction in the United States are being designed to pro-
duce 250 x 106 SCFD of medium to high Btu synthetic
Table 1. Control technology
Types of was tes
Environmental effects
Available treatment technology
Water reuse potentials
Climatic effects on wastewater
treatment
241
-------
Table 2. Typical coal gasification processes
under development
High-Btu
gasification
Low-Btu
gas i fi ca ti on
B i-gas
CoGas
C02 acceptor
HyGas
Lurgi
Molten salt
Synthane
Agglomerating gasifier
AtGas
Entrained gasifier
Koppers- Totzek
Lurgi
Stirred fixed bed
natural gas (SNG) that will yield about 970 Btu/SCF.
The average consumption of coal in these plants is about
1,000 to 1,500 tons per hour, and the average annual
usage of raw water is about 17,500 acre-feet (AF). The
plants will be required to meet strict air, solid waste, and
water pollution regulations, and will be expected to
practice conservation by a reasonable water reuse pro-
gram.
The complexity of the coal gasification process re-
quires that the major operations and sources of effluents
be described in terms of modules that will allow for the
analysis of process streams and effluents. A simplified
block diagram of the Lurgi gasification process and the
related facilities is presented in figure 2. The various
process and effluent streams have been left off since
Table 3. Scope of paper
Potential sources of wastewater
effl uents
Typical constituents of waste-
water streams
Typical wastewater treatment
processes
Potential climatic effects
Climatic effects on wastewater
treatment
they are beyond the scope of this paper. The modules
have been given a number and are described in more
detail in table 4.
Several of the modules are affected directly by the
climate. Rainfall runoff has a great potential as a source
of pollution. Poor housekeeping in some areas of the
plant in the winter can result in very high concentrations
in the runoff during the spring thaw. I mproper storage
of ash piles and feed coal can result in leachate and
runoff problems during spring thaw and rainstorms. All
of the wastewater recovery and treatment processes
produce some sludges, brines, or effluents that will cause
pollution if not disposed of adequately. Temperature
and other climatic effects will affect the quantities of
these materials to be disposed of.
WASTEWATER CONSTITUENTS
There is a wide variety of products, byproducts, and
waste effluents produced by a coal gasification plant.
Many of these wastes are complex organics, some are
very toxic and/or carcinogenic. The actual concentration
of each constituent is dependent on the constituents in
the feed coal and on the plant operation. Some of the
typical constituents in the wastewaters that are deter-
mined before a treatment facility is designed are shown
in table 5. This is not a complete list, but rather indi-
cates some of the difficulties that might be encountered
in the treatment process.
WASTEWATER TREATMENT PROCESSES
The complexity of the various waste streams from a
gasification plant has led to a variety of research projects
and contracts to determine the best and most econom-
ical processes for use. The enactment of stricter effluent
standards for many types of pollution sources has put
pressure on the designers of gasification plants to
adequately dispose of their wastes. Some of the typical
treatment processes suitable for use in a coal gasification
plant are presented in table 6. Again this list is far from
complete, but is representative of some of the treatment
methods. Most of the treatment processes require that
an adequate pretreatment has previously been used.
Some of these processes are also used to produce clean
water needed in various parts of the plant.
There are various paths that can be followed as far
as the selection of treatment units that will produce
comparable treatment results; however, the biological
processes are often selected as part of the process since
their overall costs are generally lower. For this reason,
most of the effects of climatic conditions will be dis-
cussed with regard to the biological processes. There are
242
-------
RAW STORe
COAL 1
GAs.
2
~
~
W
ASH
10
POND
18
UENC
3
COOL
SHIFT
4
T-O
PHENO
12
AMMON
13
1
WTP . B. FEEL "'- STEAM
19 20 21
5
. PUR I F
6
METH
7
COOL
15
I STOR I
23
Figure 2. Potential sources of wastewater effluents
at coal gasification plant.
WWT~
14
I om I
22
PUR IF
8
COMP
. SNG
9
SUlFUF ~ STAcK
16 17
I WA~:l
lLANO)
25
-------
Table 4. Effluent source modules of coal gasification plant
Module
Effl uent
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Description
Coal storage and preparation
Oxygen blown gasification
Quenching and cooling
Shift conversion
Gas cooling
Gas purification (Recti sol wash)
Methanation
Final gas purification (Recti sol wash)
Compression and dehydration
Ash handl ing
Tar-gas liquor separation
Phenol recovery (Phenosolvan process)
Ammoni a recovery
Wastewater treatment
Cooling tower
Sulfur recovery (Stretford process)
Stack gas cleaning
Ash ponds
Raw water treatment
Boiler feed preparation
Steam and power plants
Oxygen plant
Storage (product and byproduct)
Washroom and work area
Land surfaces
Rainwater runoff
Ash
Tar-oi 1
Catalyst
Gas liquor
Gas 1 i quor
Catalyst
Condensate
Wet ash
Tar & oi 1 water
Was tewater
Was tewater
Sludge & effluent
Blowdown
Was tewater
Was tewater
Leachate & runoff
Sludge
Sludge & brine
Ash & b lowdown
Condensa te
Runoff
Sewage
Runoff
& runoff
design of wastewater treatment facilities is temperature.
The historical data (ref. 1) for temperatures at Dunn
Center, North Dakota, are presented in figure 3. The
wintertime average temperatures fall to -100 C with
about one half of the year having temperatures below 00
C. The extremes range from -450 C to +450 C, and can
cause severe effects on the operation of the treatment
units or biological systems. The winter low temperatures
are sustained for extended periods of time and will effec-
also definite climatic effects in the physical and chemical
processes. Any treatment process that is operated in the
outside environment will be subjected to the effects of
the climate that will affect their operation and mainte-
nance.
CLIMATIC CONDITIONS
One of the principal climatic factors that affects the
244
-------
Table 5. Typical constituents of
wastewater streams
Monovalent phenols
Multivalent phenols
NH3 fixed (Cl)
NH3 fixed (fatty acids)
NH 3 free
Fatty aci ds
Cl
F
NAOH
C02
H2S
HCN
B005
tively draw the available heat out of almost any process.
The summertime highs combined with heat from the
plant can also cause a detrimental effect on the opera-
tion of the units.
The amount of sunshine and the net solar radiation
affects the rate heat transfer from cooling ponds and
biological ponds. It affects the amount of photosyn-
thesis in the ponds and, ultimately, the amount of
oxygen available for biochemical oxidation of waste
materials. A heat input of 0.2713 Btu/sf/day is produced
by one Langleys/day. The historical data are presented
in figure 4.
Atmospheric temperatures less than about 18° Care
considered heating days. While this concept is not com-
pletely applicable to wastewater treatment processes, the
. curves in figure 5 are indicative of the relative cold and
hot period in the area where the plant is to be built.
These climatic conditions could be considered favorable
from the standpoint of cooling of the process waste
streams and the cooling water. The rate of cooling will
affect the design of the holding ponds or units where the
hot waste streams are allowed to cool some before being
sent to the wastewater treatment processes.
The wind speed and direction at the site will affect
the rates of cooling and the rate of oxygen transfer in
ponds. The wind speed and atmospheric temperatures
are combined to produce a wind chill factor that reaches
very low temperatures. With an average wind speed of
about 15 mph, it could be an important factor in the
design of treatment methods. A typical wind rose is pre-
sented in figure 6.
One of the principal contributors to pollution is
rainfall itself. Runoff from land surfaces in work and
storage areas, as well as stockpiles, can contribute large
quantities of pollution. The contamination of the
aquifer from leachate is also a problem in some areas.
The control of the waters produced by runoff is an im-
portant part of the wastewater treatment process. As
shown in figure 7, the low precipitation in some parts of
Table 6. Typical wastewater treatment processes
Physical
Cherni ca 1
Biological
Sedimentati on
Flotation
Oil separati on
Stripping
Solvent extraction
Adsorption
Comb us ti on
Fi ltrati on
Neutralization
pH adj us trnen t
Coagulation
Preci pi tati on
Oxidation
Ion exchange
Activated sludge
Trickling filter
Aerated lagoons
Waste stabilization
pon ds
245
-------
100
20
10
.-.. 50 .-..
u. u
o 0
......... '-"
w 0 W
0:: 0::
:J :J
I- I-
-
-
w Z
...J :I:
\!J
Z en
:) 400 200 z
RADIATION :J
en
z u.
o 0
- en
I- 200 100
-------
1500
en
1000 >-
0:(
Q
(.!)
z
......
500 t-
o:(
W
:t:
o
500
JAH
I=EB
JUN
MAR
APR
r1AY
C,!)
z
~U)
J>-
00:(
OQ
U
JUL
AUG
SEP
tloy
DEe
OCT
Figure 5. Heating and cooling degree days,
Dunn Center, North Dakota (ref. 1).
NORTIJ
12.3
ANNUAL
10 % FREOUENCY
OF WIND
15 D I RECTI ON
Figure 6. Percent frequency of wind direction and mean wind speeds,
Dickinson, North Dakota, 1960-1971 (ref. 2).
247
-------
8
.........
U)
UJ 6
:J:
U
~
"'-
......
.........
:z
0 4
......
I-
«
I- AVERAGE
0...
...... 2
u
UJ
0:::
0...
o
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
DEC
Figure 7. Monthly average and extreme precipitation,
Dunn Center, North Dakota (ref. 1).
the country will determine the degree of treatment re-
quired to conserve water and provide for water reuse.
Coal gasification plants located in northern climates
are also confronted with snowfall and drifts. While
actual snowfalls may be considered small in many cases,
the drifts that build up may affect the operation of the
wastewater treatment units. Generally, the snow is more
of an inconvenience rather than a detriment to opera-
tion. The historical snowfall records are presented in
figure 8.
CLIMATIC EFFECTS
There is a wide variety of possible climatic effects
on wastewater treatment processes. Some of these are
important and some are in sign ificant. A few of these
effects on the biological treatment systems are presented
in table 7. The variation of the rate of biological activity
will be discussed first.
The biological reaction rate constant, k1 , varies with
temperature according to the following equation:
kt = k1 8 (t - 20)
where: kt = rate constant at
desired temperature
k1 = rate constant at 20° C
t = temperature,OC
Each type of biological or chemical process has a
different value for the temperature coefficient, 8. This
coefficient for several types of biological treatment proc-
esses is shown in table 8 (ref. 3). The equation is con-
sidered valid only within the indicated temperature
ranges. A plot of a typical reaction rate versus temper-
ature is shown in figure 9. This curve shows that the
organisms are slowed down by the low temperatures, but
are not always killed if maintained at that temperature
for short periods of time. At the other end, however,
temperatures above 45° C will usually kill the mesophilic
bacteria normally found in these waters. Similar equa-
tions exist for the chemical process; however, the tem-
perature coefficient and reaction rate constants will be
248
-------
24
33
........ 16
U> HIGH HIGH
IJJ
:J:
U
Z
..... 12
-....J
3:
0
Z
U>
8
20
4
o
JAN
APR
MAY
FEB
MAR
JUN
SEP
DEC
OCT
NOV
JUL
AUG
Figure 8. Monthly average and extreme snowfall,
Dunn Center, North Dakota (ref. 1).
Table 7. Climatic effects on
wastewater treatment
Variations in rate of biological
activity
Atmospheric oxygen transfer rate
Changes in oxygen sol ubi 1 i ty
Variations in photosynthetic oxygen
production
Variations in
Va ri ati ons in
detention time
treatment efficiencies
different. Likewise, physical processes are affected by
temperature variations.
One of the detrimental effects of low temperatures
is that treatment efficiencies are reduced or detention
times are increased. Using the following equation, the
detention time necessary to meet a given removal effi-
ciency at a known reaction rate constant can be deter-
mined.
t
E
2.3 k1 (100 - E)
where:
= detention ti me, days
t
E = removal efficiency, %
k1 = Biological rate constant (base 10)
249
-------
Table 8. Biological temperature coefficients, ()
Process
Temp range °C
e Ra n ge
Stabilization pond
Act i vated sludge
Ae ra ted 1 agoon
Trickling filter
Aerobic~facultative
An ae rob i c 1 agoon
Extended aeration
lagoon
1. 072 - 1. 085
1.0 - 1.041
1. 026 - 1. 058
1.035
1. 06 - 1. 1 8
1.08 - 1. 10
1. 037
3 - 35
4 - 45
2 - 30
10 - 35
4 - 30
5 - 30
1 0 - 30
":'- 2.0
~ ....
,
I- \
z: \
~
~ 1.0 \
\
IL \
IL
UJ LOG 9 \
o
u \
UJ \
I- 0,5 \
«
<>: 0 = 1.0116
:z
2 K20 = 0.6
I-
u
«
UJ
(Y-
0.2
0 10 20 30 40 50 60
TEMPERATURE (Oc)
Figure 9. Effect of temperature on
reaction rates.
The effect of temperature on the detention time
and percent BOD removal are presented in figure 10. As
the temperature drops in a treatment unit-such as a
stabilization pond, the required detention time to reach
the desired BOD removal is greatly increased. The net
effect is a large increase in size and cost of unit.
In addition to the effects on biological activity due
to variations in the reaction rate constant, k1, temper-
ature also effects the oxygen transfer rate and the solu-
bility of oxygen in the water. Wind velocities will affect
the rate of surface reaction, and solar radiation will
affect the photosynthesis reactions. Ultimately, all of
these factors affect the rate of biological activity.
100
50
'"
z
::;
q: 10
ffi
'"
g 5
'"
9 ~ 1.046
K20 ~ 0.6
o
o
1
------- 0 °c
------- liJoc--
2
3
It
5
8
7
9
10
6
DETENTION TIME, DAYS
Figure 10. Effect of temperature on
treatment efficiency.
CONCLUSIONS
Climatic effects are one of the items that should be
considered in the development of any control technol-
ogy program. While not all aspects of the climate will
adversely affect the treatment efficiency of a given unit
operation, the physical and electrical design aspects may
requ ire special attention to prevent malfu nctions of their
operations.
The climate of an area will affect the selection of
the treatment sequence to reach desired effluent limita-
tions. Too many environmental assessments are being
prepared that assume that a biological process will dis-
250
-------
pose of all the Wastes when this may not be true. Aware-
ness of the potential problems and a more thorough eval-
uation of the design problems may take care of the cli-
matic effects on wastewater treatment.
REFERENCES
1. J. M. Ramirez, "Meteorology/Air Quality Aspects of
a Coal Gasification Plant in Dunn County," Project
No. 2520-513-4886, Department of Soils, North
Dakota State University, to be completed June
1976.
2. J. M. Ramirez, "Implications of the Atmospheric
Environment on Coal Development in Western
North Dakota:' Project No. 16-418-GR, Subcon-
tract No. ND 2535-4692, Department of Soils,
North Dakota State University, to be completed
May 1978.
3. W. Wesley Eckenfelder, Jr., and A. J. Englande,
"Temperature Effects on Biological Waste Treat-
ment Processes," International Symposium on
Water Pollution Control in Cold Climates, p. 180,
Water Pollution Control Research Series, 16100
EXH 11/71, U.S. Environmental Protection Agency.
251
-------
WASTE MANAGEMENT OF FUELS PROCESSING EFFLUENTS
Abstract
G. W. Grove*
Refining of hydrocarbon fuels results in wastewater
streams containing hydrocarbons, solids, and water.
These contaminants must be removed from the water
prior to discharge or reuse. Whether resulting from
synthetic or conventional fuels refineries, the techniques
for concentration and disposal of waste hydrocarbons
and solids should be much the same. These techniques
involve a series of concentration steps resulting in a
reduced volume of sludge that can be treated in the
lowest cost environmentally acceptable manner.
This paper describes technology developed in the
last several years for refinery wastewater treatment
sludge concentration. The advantages compared to pre-
viously used techniques are:
1. Less energy consumption.
2. Lower investment and operation cost.
3. Higher sludge concentration which reduces final
sludge treatment or disposal cost.
4. Improved separation of waste and water.
Over the last several years Exxon Research and En-
gineering has improved the technology for concentrating
oily solid contaminants which result from treatment of
refinery wastewaters. The dewatering of these sludges is
essential to reduce the volume and therefore disposal
cost of the waste material. This paper compares waste-
water sludges resulting from conventional refinery and
synthetic fuels wastewater treatment plants. It will then
describe application of new sludge concentration tech-
nology for dewatering of these effluent sludges.
The exact composition of effluents from synthetic
fuels conversion plants has not been well defined and in
fact will vary widely depending on the source of synthe-
tic fuel and the conversion process used. However,
several general comparisons can be made with effluents
from conventional refinery wastewater treatment plants
(fig. 1).
1. Larger quantities of fine solids may be anticipated
due to coal cleaning, grinding, and drying opera-
tions. Portions of these solids will be present in the
wastewater.
2. Fines from stone, refuse, and slag are likely to be
more abrasive than solids found in conventional
refinery water effluents.
*The author is with Exxon Research and Engineering Co.,
Florham Park, New Jersey 07932.
WA S rEWA TER TREAtMENT SLUDGE
FUElS CONVERSION PLANT VS. CONVENTIONAL REF IN un
1. GREATER QUANTITIES OF FINE SOLIDS
2. FINES MO~E ABRASIV[
J. IIIGtlER AMOUNTS OF T~ACE MUALS
4. OIL REMOVAL AND OIOlOGICAl SOLIDS
Figure 1. Wastewater "treatment sludge.
3. Higher levels of trace metals in coal and shale will
require greater concern about leaching from waste
sludges at landfill sites. This is especially true since
acidic nature of the waste may iQcrease solubility of
these metals. Leaching problems can be minimized
by reducing the volume of sludge disposed on the
land.
4. The BOD and insoluble oil content of synthetic fuel
plant wastes will depend largely on the conversion
process and temperatures involved. However, for
liquefaction processes, it can be assumed that proc-
ess water will contact liquid hydrocarbons and oil
removal and biological treatment of the wastewater
will be required.
Based on these assumptions, wastewater from fuels
conversion plants will be similar to refinery wastewater
in that it will contain fine inert solids, biological solids,
and insoluble oil. There will likely be greater quantities
of abrasive solids with higher trace metal concentrations.
A typical process sequence for removal of waste-
water contaminants is shown across the top of figure 2.
The processes used for removal of these contaminants
include dissolved air flotation, granular media filtration,
biological treatment, and activated carbon. All of these
processes result in dilute slurries of 0.1-5.0 weight per-
cent oily solids. These slurries must be concentrated to
reduce the large volumes of water retained with the fine
solids.
The oily solids from these treatment processes are
concentrated in a series of steps shown within the dotted
line. In a typical refinery, the area within the dotted line
253
-------
PRETREAT-
MENT
DA FOR
FILTERS
r--- --------- ---------~---,
I I
I I
I FI RST STEP I
: I
I I
I I
I I
I SECOND STEP I
I I
~ I
: THIRD STEP I
I I
L______----------------~-~
WASTE-
WATER
ACTIVATED
SLUDGE
BIOX
POll SH I NG
FILTERS
EFFLUENT
ULTIMATE
DISPOSAL
Figure 2. Typical refinery or chemical plant wastewater treatment.
accounts for 40 percent of the investment and 60 per-
cent of the operating cost of the wastewater treatment
plant.
Figure 3 shows the concentration steps in more de-
tail. First the sludge is concentrated by gravity settling
or flotation. These are relatively low cost processes per
unit volume of sludge treated. Next, the sludge is further
concentrated mechanically. Traditionally centrifuges,
pressure, and vacuum filters have been used for mechani-
cal dewatering. These higher cost processes, along with
the newer technology of gravity belt filtration, will be
discussed later in more detail. Finally, incineration may
be used to reduce sludge volume to an absolute mini-
mum or the sludge may be disposed directly to the land.
In either case, it is important to reduce the volume of
sludge being disposed in order to minimize cost and envi-
ronmental problems.
Recent studies at Exxon Research and Engineering
have concentrated on mechanical dewatering because of
the critical technology needs of equipment used for this
purpose. We have found that gravity belt filtration has
significant advantages compared to previously used cen-
trifuges, pressure, and vacuum filters. It will also be
shown how gravity belt filtration enables greater flexibil-
ity in design of the entire wastewater treatment system.
This increased flexibility can result in considerable
investment and operating cost savings.
First Step Second Step Third Step
Gravity Settling Filtration Landfill
Vacuum
Flotation Pressure SI udge
Farming
cen~U!lation
roll Sand Lime
Disc
I mperforate Basket Incineration
Gravity Belt Filtration
Figure 3. Methods on concentration.
Figure 4 describes some of the problems associated
with conventional mechanical dewatering equipment.
Vacuum or pressure filters have not been used exten-
sively for refinery sludges because filter surfaces blind
rapidly. This necessitates adding additional inert solids
to increase porosity and results in increased solids for
disposal by factor of two to three. Centrifuges have been
used more extensively for concentrating petroleum and
254
-------
. VACUUM AND PRESSURE fiLTERS
~IGH 6P BLINDS CLOT~
ADDITION Of INEIIT SOLIDS INCREASES DISPOSAL COST
. CENTRifUGES
~IGH SPEED MACliiNES RESULT IN ABRASION AND
PLUGGING
tOw SERVICE fACTOR AND ~IG~ MAINTENANCE COST
Fi!:lure 4. Problems of mechanical dewatering.
petrochemical sludges. These are high-speed, precision
machines. The grit in the sludge can cause rapid erosion
and in some cases plugging. This results in frequent
downtime for maintenance and cleaning.
Now I would like to discuss gravity belt filtration
for concentration of refinery and petrochemical waste-
water sludges. These devices have been used in Europe
for 10 to 12 years on municipal sludges, but application
to oily refinery sludges is yet to be proven commercially.
Our studies were carried out on one gravity belt filter
press that is manufactured both in Europe and the
United States. It is similar to presses manufactured by
several U.S. companies.
Gravity belt filter presses dewater sludge in three
basic steps. First, polymers are added to flocculate the
fine sludge particles. Next, the flocced feed is poured
onto a coarse screen where the clear water readily drains
away. Third, the cake is pressed slowly to remove addi-
tional water. Figure 5 shows a schematic of the press.
Polymer is added to the sludge in a rotating drum where
the floe is formed. Water then drains from the sludge on
the top belt. At the end of the top belt, sludge falls to
the bottom belt and is pressed between the two screens.
The rate at which pressure is applied is very important.
If pressure is applied rapidly, the sludge goes through the
screen. However, if pressure is applied gradually, addi-
tional water drains from the sludge and solids concentra-
tion in the cake increases.
We have tested a full scale gravity belt filter press at
Exxon's Bayway and Benicia Refineries. In our tests, we
were interested in four parameters: (1) solids produc-
tion rate, (2) solids capture, (3) properties of the cake,
and (4) polymer dosage and costs.
Based on the full scale tests, it was determined that
production rate is a function of feed concentration.
255
Feed concentration was determined by drying a small
representative sample of the feed at 1030 C to determine
dry solids concentration. This "dry solids" therefore
includes true solids as well as higher boiling hydrocar-
bons (fig. 6).
At low feed concentrations there is a constant fil-
trate flow rate determined by resistance of the screen to
filtrate flow. Production rate is therefore proportional to
feed concentration. At higher feed concentrations, fil-
trate flow rate decreases due to increased resistance of
the thicker sludge blanket to filtrate flow. Production
rate continues to increase, but at a slower rate. Specific
production rates may vary depending on the specific
sludge and gravity belt press manufacturer. The
machines will operate at any rate up to these values and
rate does not depend on the ratio of oil to solids in
typical feed.
Solids capture has been found to be very high for
gravity belt filter presses compared to centrifuges. Fil-
trate plus wash water, from biological sludges has con-
tained as low as 100 ppm solids. This is a solids capture
of about 99.7 percent. Solids capture for oily sludges is
about 97-98 percent. These numbers compare to solids
capture of 86-92 percent encountered using imperforate
basket centrifuges.
Gravity belt filter presses use the normal water
treating high molecular weight anionic and cationic
polymers. The flocculation of sludge does not appear to
be very sensitive to the amount of polymer used. There
is an optimum dosage from a cost standpoint, but opera-
tion is not adversely affected by minor changes in sludge
composition and minimum operator attention is re-
quired.
Polymer costs derived from our studies are in the
range of $2 to $10 per ton of dry solids. High polymer
cost results from dilute feeds and can generally be re-
duced by gravity settling or dissolved air flotation ahead
of the press. Combining sludges has also resulted in
reducing polymer cost 10 to 25 percent compared to
handling the sludges separately. This is probably due to
neutralization of electrostatic charges in the feed. There-
fore, combining sludges may not always result in lower
cost but it should never result in increased cost.
Cake concentrations obtained with gravity belt filte,
presses have been two to three times that obtained with
imperforate basket centrifuges. This is the primary ad-
vantage of gravity belt filtration since it greatly reduces
incineration or final disposal cost.
Figure 7 gives a comparison of gravity belt filtration
and our previously preferred mechanical dewatering de-
vice, the imperforate basket centrifuge.
The use of gravity belt filtration indicates lower
investment compared to centrifuges for refineries
-------
SCHEMATIC OF A GRAVITY BELT FILTER PRESS
SLUDGE FEED
CHEMICAL
FLOCCULANT
ADDITION
Figure 5. Schematic of a gravity belt filter press.
we have studied. This is due to the lower installation
cost of these slow speed, low horsepower devices.
Operating costs are about even since the higher
polymer cost is offset by reduced power cost.
European experience with municipal sludge shows
lower maintenance and high service factor. About
95 percent service factor is estimated for gravity
presses as opposed to 85 percent for centrifuges.
Also, much of the maintenance can be done by the
operators whereas skilled machinists are required for
centrifuge maintenance.
As mentioned, a major advantage of gravity belt
filtration compared to imperforate basket centri-
fuges is that the cake produced is more concentrat-
ed. Note that a centrifuge will not concentrate DAF
float as well as filter backwash whereas the gravity
press concentrates both sludges equally well. This
enables use of DAF pretreatment rather than granu-
lar media filtration and may result in a considerable
investment savings.
The dryer cake also reduces the amount of water to
be evaporated in an incinerator resulting in a con-
siderable fuel savings. Many locations haul sludge to
offsite disposal. In that case, hauling cost savings
would be directly proportional to the volume re-
duction in the amount of sludge.
The above applications are for concentration of
wastewater treatment sludges. These sludges may be
oily, slimy and gelatinous, which will plug vacuum and
pressure filter media. Also, the specific gravity of the
solids removed is close to that of water so that centrifu-
gation is difficult. Part of our program has evaluated use
of gravity belt filtration to remove water from FLEXI-
COKER venturi scrubber fines. FLEXICOKING is a
fluid coking process followed by gasification of the cake.
The hard abrasive nature of these coke fines from the
scrubber are likely to be duplicated in synthetic fuels
conversion plants and would result in severe abrasion of
high speed centrifugal equipment.
Tests on a full scale gravity belt filter press showed
256
-------
PRODUCTION RATE VS. FEED CONCENTRATION
N
l-
LL
..........
a:::~
:J:w
iCa:::
-~
0:
eLL
o
(/')
():J
-1
Q
Q
W
I-
e
o
a:::
a-
t:)
2
(/')
-------
COMPARISON OF THE GRAVITY BELT FILTER PRESS
WITH IMPERFORATE BASKET CENTRIFUGES
... LOWER INVESTMENT AND OPERATING COST
. REDUCED MAINTENANCE AND HIGHER SERVICE FACTOR
. VERY DRY CAKE
610 SLUDGE NO PRETREATMENT
PRETREATMENT SLUDGE
DAF FLOAT
FILTER BACKWASH
~
9
0/0 DRY SOLI DS
GRAVIlY
.EB!.&.
18-20
11
20
30-35
30-35
. ENABLES USE OF LOWER INVESTMENT DAF
.. INCINERATOR FUEL SAVINGS OR REDUCED DISPOSAL
COST
Figure 7. Comparison of the gravity belt filter press
with imperforate basket centrifuges.
that production rates were six times higher than rates
obtained on wastewater treatment sludges. Production
rate was a function of feed concentration, but cake
solids concentration was independent of feed concen-
tration. The cake was 44 percent solids vs 37 percent
obtained by vacuum filter tests. Solids recovered were
98 percent.
In summary, our studies of mechanical dewatering
have shown that there are significant incentives for using
gravity belt filtration on refinery and chemical plant
sludges. Synthetic fuel plants are likely to have waste-
water sludges similar in some respects, but with more
abrasive fine solids. Removal of these solids can easily be
done by gravity belt filtration.
258
-------
CONTROL TECHNOLOGY R&D NEEDS*
C.E. Jahnig, R. R. Bertrand, E. M. Mageet
with steam and oxygen (ref. 1). The model provides
typical heat a~d material balances and flow rates to use
as an overall framework in environmental evaluations.
Although the various coal conversion processes are
distinctly different, they also employ a number of com-
mon operations, such as coal preparation, acid gas
removal, waste water treatment, and provision for utili-
ties such as electric power, steam, and water (table 1).
For all of these common areas, careful planning is re-
quired in order to assure that the operations are environ-
mentally satisfactory. We believe that the areas of
environmental concern can be properly taken care of,
but they need to be examined and evaluated early in the
development program so that the required work can be
done in an orderly manner and not under the pressure of
a crisis. A program and procedures for this have been
developed (ref. 2).
Abstract
This paper discusses research and development
needs in environmental control technology related to
fossil fuel conversion operations. In some cases, existing
control systems are inadequate or leave significant room
for improvement in pol/ution control or in energy con-
servation. In other cases, existing information may be
inadequate and additional work is needed to define and
evaluate environmental aspects. Such items are discussed
for the major areas of coal preparation, acid gas removal,
sulfur recovery, methanation, wastewater treatment, and
trace elements. These areas of concern are common, or
are at least similar, in the various conversion processes
being developed.
Technology needs developed as a result of extensive
studies include the fol/owing:
1. Indirectly heated coal dryer (e.g., fluid bed) which
contains dust and also recovers water from moisture
in coal.
2. Objectives are defined for improved acid gas re-
moval systems, including hot gas cleanup.
3. Methanation systems to optimize recovering the
heat of reaction, equal to 20 percent of the HHV of
reactants.
4. Techniques to minimize water consumption, and for
wastewater cleanup and reuse.
5. Handling of trace elements, especially the large
amount of volatile ones that leave the gasifier and
are ca"ied into the gas cleanup system.
INTRODUCTION
There are many different conversion processes for
making clean liquid or gas fuels from coal, but many of
the environmental concerns and problems are quite simi-
lar. This is not surprising since every process starts with
coal as the raw material and rejects sulfur and ash, while
generating a wastewater stream that contains a broad
spectrum of contaminants. In order to facilitate environ-
mental assessments, we have developed a generalized
basis which broadly represents the operation of gasifying
*This work was carried out under Contract No. 68-02-0629,
U.S. Environmental Protection Agency.
tThe authors are with the Government Research Labora-
tories of Exxon Research and Engineering Company, Linden,
New Jersey.
Table 1. Areas for improvement
Coal preparation
Acid gas removal
Methanation
Wastewater cleanup
Trace elements
Energy and water conservation
I n some areas of environmental concern, present
technology is not completely adequate, as in the cases of
acid gas removal and wastewater cleanup. In other areas
environmental control technology is available, but there
is considerable room for improvement-for example in
coal drying, methanation, and overall aspects of energy
conservation.
The area of trace elements deserves special mention
because available information is not sufficient at this
time to allow making a sound evaluation of potential
environmental problems. Some of the trace elements in
coal are appreciably volatile at gasification conditions
and will be carried into the gas cleanup system where
they may have to be removed and disposed of. Although
259
-------
trace elements are a major concern, the associated re-
quirements for control technology can hardly be defined
until more information is obtained on their amounts,
where they appear, and in what form.
Research and development needs for control tech-
nology will now be defined in more detail for some spe-
cific areas that are common to most of the coal conver-
sion processes. Starting with coal drying, it is a large
operation which consumes fuel equal to 5 to 10 percent
of.the total coal (table 2). A desirable objective is to save
this fuel by using low level waste heat for drying the
coal. There is more than enough waste heat available
from the process and the need is to devise practical
~ethods for its use. One possibility is to contact the coal
with warm flue gas from the utility boiler or a process
furnace. This may be particularly applicable if stack gas
cleanup is used subsequently, thereby controll ing dust
emissions. Another possible source of heat is warm air
from an air-fan cooling unit, some of which are large
enough to supply all of the required heat.
Table 2. Coal preparation
Fuel used may be 5-10% of total coal.
Consider fluid bed dryer using waste
heat.
On lignite, water from drying can
amount to half of plant makeup.
Refuse from coal cleaning may be 400-
800 acre ftjyr.
One type of drying system to consider is indirect
heating using coils in a fluidized bed of coal (ref. 3). This
greatly reduces the volume of gas that must be cleaned
up, although it will also be somewhat more difficult to
accomplish drying in the presence of an offgas having
higher moisture content. A potential advantage is that
water can be recovered from the drying operation by
using a simple air cooler on the vent gas. In plants using
lignite coal containing 30 percent or more moisture,
water recovered from the dryer can contribute 50 per-
cent of the total water makeup required for the plant.
Work is needed on disposing of solid wastes from
coal conversion, including waste from coal cleaning as
well as ash or char. This may not be as complex a techni-
cal problem as the gasification or liquefaction process,
but the volume of material to dispose ot is enormous
and can cause concern regarding dust, leaching, revegeta-
tion, and land use. Refuse from coal cleaning can
amount to 25 percent of the run of mine coal, equal to
400-800 acre feet/year for a large plant. I n addition
residual ash may be 10-20 wt. percent of the coal con-
verted. If gasification takes place without a change in
particle size, the volume of ash can be as much as the
volume of coal used. A slagging operation should
decrease the ash volume and also would minimize the
disposal problems due to dusting or leaching. Here is an
area that has opportun ities for land improvement.
Moving on to the subject of acid gas removal, it
appears that available technology can probably provide
acceptable environmental controls, but with consider-
able difficulty; and there is room for improvement with
regard to energy consumption. Research objectives can
be spelled out as follows for an ideal process:
1. Remove all forms of sulfur to a low level.
2. Provide a concentrated H2 S stream (50 vol. percent)
to sulfur plant.
3. Assure that the CO2 reject stream is clean and free
of sulfur, combustibles, etc.
4. Have low consumption of utilities and energy.
5. Generate no chemical wastes to be disposed of.
In some applications, it is also 'very desirable to make a
I
sharp separation between H2 S and CO2, beyond the ca-
pability of current technology.
Acid gas removal systems based on amines are not
effective on carbonyl sulfide, which may account for
perhaps 10 percent of the total sulfur in the raw gas
from gasification. Contaminants such as HCN interfere
with regeneration of amine. Hot carbonate can partially
remove carbonyl sulfide, but it is very difficult to pro-
vide a concentrated H2 S stream and at the same time
have a clean CO2 stream that can be vented directly to
the atmosphere. It may be that all the various sulfur
compounds could be hydrolyzed efficiently to H2 Saver
a bauxite catalyst at 500-700° F ahead of acid gas re-
moval to facilitate sulfur removal, or a hydrolysis cata-
lyst might be incorporated in the scrubbing system.
Scrubbing with refrigerated methanol is used in some
proposed plants, but the separated H2 S is shown to be
mixed in the rejected CO2 at very low concentration,
along with 1-2 vol. percent of combustibles (refs. 4,5).
Cleanup of this stream will be complicated.
A Claus plant is usually included to recover
byproduct sulfur. An alternative is absorption/oxidation
which combines absorption of H2 S with oxidation to
make byproduct sulfur via a liquid phase Claus type
reaction (ref. 6). This alternative has the advantage that
it can give high conversion to sulfur even at a very low
H2 S concentration. Unfortunately, it is not effective on
260
-------
carbonyl sulfide or other sulfur compounds, and lacks
the incineration of combustibles inherent in a Claus
plant.
On a Claus plant, the volume of tail gas increases as
the concentration of sulfur compounds in the feed gas is
decreased (table 3). Therefore a larger volume of tail gas
must then be cleaned up to a lower sulfur level in order
to maintain the same tons/day of sulfur emission. For
example, with a feed of 25 vol. percent H2 S the tail gas
volume is two times what it would be at 100 percent
H2 S. At 15 percent H2 S the volume is threefold, and at
10 percent H2S it is fourfold. The latter levels are
encountered in some gasification designs.
In a typical plant making SNG, the weight of CO2
rejected to the air from acid gas removal is roughly equal
to the total amount of coal gasified. Consequently, it is
very important that the CO2 stream be clean and free of
contaminants or odors. It is quite difficult to assure a
low sulfur level, and some study designs have shown up
to 3,000 ppm H2S, which is excessive. Other designs
based on scrubbing with refrigerated methanol show 1-2
vol. percent of combustibles in the CO2 vent stream,
including ethane and CO. Incineration is one way to
destroy combustibles, but the amount is not sufficient
to support combustion in a furnace, although it may
amount to 3,000 MM Btu/day or over 1 percent of the
SNG heating value. Using extraneous fuel for incinera-
tion could introduce a large debit, but it might be
avoided if catalytic combustion can be applied-possibly
with heat exchange for preheat and recovery. Recovery
of this heating value is a technological challenge, and an
environmental objective.
Table 3. Want to remove sulfur
first, then CO2
Vol. % H?S
to Clau5
Vol. Tail
Gas (Ratio)
100
50
15
10
CO? rejected to air = wt. of coal
ga5ified
1% combustible content = 1% of coal
HHV
1
2
3
4
An important objective is lower utilities and energy
consumption for acid gas removal. It usually represents.
one of the largest consumers of utilities in the plant,
comparable to the oxygen plant. In optimizing the util-
ity balance, heat pumps should be evaluated so that the
heat of condensation on the overhead can be used to
supply heat to the reboiler. The theoretical thermo-
dynamic work for the separation is quite small, so it is
possible that better techniques can be developed. Semi-
permeable membranes that are selective to CO2 or to
sulfur compounds represent one approach.
Removing sulfur at higher temperature may also
have advantages. Techniques are known for removing
small amounts of sulfur using metals or metal oxides
such as iron, zinc, copper, or nickel. The difficulty has
been that these cannot be conveniently regenerated as
required for removing large amounts of sulfur. It should
be possible to develop practical techniques for regenera-
tion so that the sulfur absorbent could be recycled con-
tinuously or used batchwise. Limestone and dolomite
have been tested for this purpose, and the results are
very encouraging (refs. 3,7). Incidentally, in applications
such as the combined cycle for power generation, it is
neither necessary nor desirable to remove CO2 along
with sulfur.
One final comment on acid gas removal-most sys-
tems need an appreciable amount of ma keup chemicals,
which can then show up as plant effluents. Often there is
a purge or chemical waste stream to reject products of
side reactions or stable compounds formed with contam-
inants that cannot be regenerated adequately. Chemical
waste streams can be a disposal problem and should be
minimized or eliminated.
The next area for discussion is methanation (table
4). Reaction of CO and H2 to make methane has a very
high heat release, equal to 20 percent of the heating
value of the reactants. This serves to emphasize that low
Btu fuel gas should be used where possible rather than
SNG. The loss of 20 percent in heating value is not an
entirely real loss when making SNG, since it is used to
make high pressure steam which would otherwise have
to be made in a boiler. Conventional methanation uses a
fixed bed of catalyst, making it difficult to remove the
large amount of heat and to provide good temperature
control. Gas is recycled from the reactor outlet through
waste heat boilers and back to the reactor inlet, with a
considerable power consumption on the recycle gas com-
pressors. Recent test work has shown that methanation
can be carried out in a fluidized bed or in a slurry reac-
tor (ref. 8). These look promising in that the reaction
heat should be recoverable at a higher temperature level
using exchanger surface in the reactor, while eliminating
261
-------
Table 4. Methanation
Heat release is 20% of HHV in reac-
tant.
Recover steam, but recycle compressor
is large.
Slurry or fluid bed reactor avoids
compressor.
Methanation and shift might be com-
bined.
the need for recycle gas compressors to control reactor
temperature. In the slurry operation, water gas shift has
been found to occur during methanation; thus the two
reactions might be carried out in one reactor so that the
water produced by methanation is available to provide
that consumed by the water gas shift reaction.
The ultimate extension of this would be to carry
out all methanation in a shift reactor ahead of acid gas
removal. In effect, CO would then be reacted with steam
to form methane and CO2. Hot sulfur removal would
probably be needed first. If practical, this arrangement
would mean that the gas only has to be cooled once and
not reheated, whereas with conventional methanation,
the gas is cooled for sulfur removal, reheated for meth-
anation, and cooled again. A limitation in the proposed
simplification is that methanation would have to be se-
lective in the presence of CO2, but the subsequent CO2
removal should be easier in view of the decreased gas
volume to be scrubbed.
One of the major environmental problems on coal
conversion is the cleanup and disposal or reuse of waste-
water (table 5). The sour water condensed from the raw
gas may contain a wide variety of sulfur, nitrogen, and
oxygenated compounds, together with dust and oil or
tar (ref. 9). In addition it is expected that various trace
elements will be present in the sour water, and may have
to be separated or deactivated. Extensive processing will
be needed to make the wastewater suitable for discharge
to the environment, in which case it will represent a very
desirable makeup water for the plant.
A probable sequence of processing steps to clean up
sour water is as follows:
1. Sour water stripper to remove H2S, NH3, CO2, light
gases.
Settler to remove solids and skim off oil.
Extraction of phenols (if required).
Biological oxidation (biox) to consume residuals.
Sand filter, activated carbon, etc.
2.
3.
4.
5.
The major R&D need on wastewater cleanup is to
demonstrate that the selected systems will give adequate
cleanup under realistic conditions. This will no doubt
require operating with actual streams from an operating
plant or pilot plant, since the array of possible contami-
nants is so complex that it may be impractical to prepare
a synthetic mixture, or to predict performance of treat-
ing operations.
Table 5. Wastewater - contaminants
Sulfur compounds:
H2S, sulfate, thiosulfate
Nitrogen compounds:
ammonia, cyanides
Oxygenated compounds:
phenols, acids
Other:
hydrocarbons, carbonates, chlorides
A possibility to consider (figure 1) is not to clean up
the sour water, but rather convert it to steam which
would be fed to the gasifier. Although conventional
steam generators may not be applicable due to fouling
and corrosion, it should be possible to vaporize the sour
water by injecting it into a bed of hot fluidized solids
which are heated indirectly (ref. 10).
Regarding trace elements, a large number may be
present in the various coal feeds, and many of them are
toxic. Many of the trace elements are volatile to some
extent at gasification conditions and can then be carried
into the gas cleanup system where they may cause com-
plications (table 6). Considerable information is available
on the amount of minor and trace elements in various
coals (ref. 11), and a limited amount of work has been
done to determine the percent volatile for specific ele-
ments at gasification conditions (ref. 12). To make the
picture more meaningful, the amount of certain trace
elements in a typical coal have been combined with the
estimated percent volatile to give the amount in pounds
per day that may be carried out of the gasifier with the
raw gas. Results are shown on table 6 for a coal feed rate
of 12,000 tons/day. Volatility of chloride has not been
specifically measured for gasification but it is known
that during combustion it is nearly all volatile. The com-
bined amount of trace elements that may be volatile and
262
-------
FLUIDIZED BED
OF INERT PARTICLES
SOUR HATER IN
~
STEAM OUT
------
"'-~
- TO GASIFIER, SHIFT
150 psig - 366~F
1000 psig - 545°F
,--.A-~
HEATING COILS
HASTE I-lEAT
OR lIOT OIL
- ---
:->-
AERATION
Figure 1. Make steam from sour water.
may have to be handled in the gas cleanup system is
quite formidable. Obviously this can pose serious prob-
lems if not adequately considered early enough in the
development of a coal conversion process. Much more
information is needed to define the situation and show
where trace elements appear and in what form. Systems
can then be worked out for deactivating or separating
them where necessary.
As a final point on R&D needs, mention should be
made of the need to conserve energy and water. A few
items in this category are shown in table 7. Without
going into details, these serve to show some approaches
that may lead to savings in energy and water consump-
tion, as well as stimulate thinking in this important area.
Other suggestions are given in a series of 14 reports
covering our environmental assessments made for the
Environmental Protection Agency (refs. 2,3,10, etc.).
263
-------
Table 6. Example of trace elements that may appear
in gas cleaning section
Possiblea % Volatile b In ga~
Element ppm in toal for, example 1 b/day
Cl 1500 >90+ 32,400
Hg 0.2 90+ 5
Se 2.2 74 39
As 31 65 484
Pb 7.7 63 116
Cd 0.14 62 2
Sb 0.15 33 1
V 35 30 252
Ni 14 24 81
Be 2 18 9
Zn 44 (10) 106
B 165 (10) 396
F 85 (10) 204
Ti 340 (10) 816
Cr 22 nil nil
aMainly based on Pittsburgh Seam Coal.
bMainly based on lower temp gasifier (ref.
10% for Zn, B, and F, in absence of data.
cFor 12,000 tons/day of coal feed.
12) and indicated at
264
-------
Table 7. Energy conservation
Heat pump--AGR and S.W. stripper
Furnace air preheat and low excess air
Process preheat by heat recovery
Recovery from high pressure gas/liquid
Burn high S coal using stack gas
cleanup
Water conservation
Air cooling
Hea t pump
Recover moisture in coal
Separate water from sales in waste-
water
REFERENCES
1. R. R. Bertrand and C. E. Jahnig, "Environmental
Aspects of Coal Gasification," presented at AIChE
Meeting, Sept. 8, 1975, Boston, Mass. (to be pub-
lished).
2. C. D. Kalfadelis, E. M. Magee, G. E. Milliman and T.
D. Searl, "Evaluation of Pollution Control in Fossil
Fuel Processes, Analytical Test Plan,"
EPA-650/2-74-009, October 1975.
3. C. E. Jahnig and E. M. Magee, "Evaluation of Pollu-
tion Control in Fossil Fuel Conversion Processes;
Gasification: CO2 Acceptor Process,"
EPA-650/2-74-009d, Dec. 1974 (NTIS PB-241 141).
4. EI Paso Coal Gasification Project Environmental
I mpact Statement 1974.
5. T. E. Berty and J. M. Moe, "Environmental Aspects
of the Wesco Coal Gasification Plant," Proceedings
of EPA Symposium, St. Louis, Mo.,
EPA-650/2-74-118, Oct. 1974.
6. J. E. Ludgerg, "Removal of Hydrogen Sulphide
from Coke Oven Gas by the Stretford Process,"
64th Annual Meeting of the Air Pollution Control
Association, June-July 1971.
7. R. C. Hoke and R. R. Bertrand, "Pressure Fluidized
Bed Combustion of Coal," Institute of Fuel Sym-
posium, Paper No. 18, Sept. 1975, London.
8. M. B. Sherwin, "Progress in Liquid Phase
Methanation," Seventh Synthetic Pipeline Gas
Symposium, Oct. 27-29,1975, Chicago, III.
9. A. J. Forney, W. P. Haynes, S. J. Gasior, G. E.
Johnson, and J. P. Strakey, "Analyses of Tars,
Chars, Gases and Water Found in Effluents from the
Synthane Process," Bureau of Mines, TPR 76,
January 1974.
10. C. E. Jahnig, "Evaluation of Pollution Control in
Fossil Fuel Conversion Processes; Gasification: BI-
GAS Process," EPA-650/2-74-009g May 1975
(NTIS PB-243 694).
11. R. R. Ruch, H. J. Gluskoter, and N. F. Shimp,
"Occurrence and Distribution of Potentially Vola-
tile Trace Elements in Coal," EPA-650/2-74-Q54,
July 1974.
12. A. Attari, "The Fate of Trace Constituents of Coal
During Gasification," EPA-650/2-73-004, August
1973.
265
-------
17 December 1975
Session IV:
PROCESS MEASUREMENTS
James A. Dorsey
Session Chairman
267
-------
MEASUREMENT PROGRAMS FOR ENVIRONMENTAL ASSESSMENT
R. M. Statnick, Ph.D. and L. D. Johnson, Ph.D. *
Abstract
The current energy shortage and the public's de-
mand for environmentally safe energy conversion proc-
esses have focused the attention of both the public and
the private sectors on the potential environmental prob-
lems which are associated with these technologies. An
environmental assessment test strategy has been develop-
ed by the Process Measurements Branch, Industrial
Environmental Research Laboratory, Research Triangle
Park (JERL/RTP).
The proposed environmental assessment sampling
and analytical test strategy consists of two distinctly
different levels of effort. Level 1 is a survey phase which
identifies the pollution potential of all process feed,
product, and waste streams through chemical and bio-
logical testing. The second level of effort, Level 2, is
characterized by quantative determination of potentially
hazardous substances in process streams and by ex-
panded bioassay of the process streams which were
sampled. This paper describes the features of this
environmental assessment sampling and analysis strategy
and its associated costs.
INTRODUCTION
The sampling and analytical effort which supports
an environmental assessment has its roots in the histori-
cal failure of both the public sector and the private
sector to evaluate the environmental and the health
effects of a given product or of a given technology. The
most publicized example of this is the case of DDT. The
chemical fulfiUed a need-control of pests which were
injurious to both humans and domesticated animals. As
the full market potential of this chemical was realized,
several scientists began to notice side effects arising from
the use of DDT. Increased levels of DDT were found in
humans, in drinking water, and in domesticated animals
and wildlife. The weight of evidence was such that EPA
banned the scale of DDT in the United States.
The energy shortage which grips the United States
has focused attention on the development and full scale
implementation of coal conversion technologies, on
expanded us!:! of subs~itute fuels for natural gas and oil,
and on a variety of new energy related technologies.
*Industrial Processes Division, Industrial Environmental
Research Laboratory, Research Triangle Park, North Carolina
(Environmental Protection Agency).
Each of these processes has associated with it a possible
environmental insult. EPA has initiated a program to
determine the magnitude of the potential environmental
problems associated with the emerging energy technol-
ogies. The technology assessment program is called an
integrated assessment and includes social, economic and
environmental effects. The environmental assessment is a
key input to an integrated assessment and provides
environmental effects information on all process feed
stocks, product streams, and waste streams. The type of
information which is developed is based on extensive
chem ical analysis and bioanalysis of the appropriate
samples. The details of these analyses are discussed later
in the paper. The information which is developed by an
environmental assessment program can be used for the
purposes identified in table 1.
Environmental Assessment Testing Strategies
The Process Measu rements Branch * has extensively
examined two field test options for the performance of
an environmental assessment. The analytical goals of
each option were identical: 1. identification of specific
inorganic and organic compounds, 2. cytotoxicity test-
ing, 3. mutagenicity testing, and 4. carcinogenicity
testing. Each option would acquire samples of each
process feed stock stream, product stream, and waste
stream in sufficient quantity to complete the rigorous
chemical and biological test matrix. The basic difference
between the two options is the mechanism used to
achieve the goals, and the resultant cost effectiveness of
each.
The first option considered is the direct approach.
The operation involves the planning and the execution
of a single comprehensive sampling and analysis effort.
Table 1. Selected uses of environmental
assessment field sampling data
.
.
I dentify problematic process streams.
Identify the physical characteristics (e.g. flow, temperature,
physical state, etc.) of problematic streams.
Assist in setting control technology RID priorities.
Provide data to health effects, monitoring, and integrated
assessment programs.
.
.
269
-------
The direct approach is philosophically attractive and has
been examined on the basis of cost to implement at a
constant level of information output.
The second option considered was a phased
approach which requires two separate, distinct levels of
sampling and analysis. Level 1 sampling and analysis has
as its goal the gross identification of pollution potential
of a source and to set sampling and analysis priorities for
level 2. Level 2 sampling and analysis has as its goal the
refined, accurate identification of specific pollutants in
specific streams from a given source.
Assuming a constant level of information output,
TRW Systems Inc. (ref. 1) has estimated the perform-
ance costs associated with the environmental assessment
of a limestone wet scrubber and a coal gasifier. The cost
comparisons are given in table 2. Based on this analysis,
the phased sampling and analysis option was selected as
the most cost effective mechanism for performance of
an environmental assessment. The phased sampling and
analysis option is detailed in the following text.
Phase 1
Environmental Assessment
The details of the phase 1 sampling and analysis
effort are presented in other papers in this volume. This
paper addresses the goals of phase 1 and the end uses of
phase 1 data.
Since the objective of an environmental assessment
is to identify any potential environmental pollution
from a process, all process feed stocks, product streams,
and waste streams will be sampled in level 1. For
purposes of this paper, there are four categories of
sampling (see table 3). The procedures which are used to
acquire samples from each of these categories are
described in the next paper (Lockmiiller).
The samples which are acquired will be subjected to
elemental analysis (inorganics), to liquid chromatog-
raphy (organics), and to bioassay (cytotoxicity and
mutagenicity). Elemental analysis are performed with a
spark source mass spectrometer. The hydrocarbons are
Table 3. Sampling categories
Particu late
All gas streams are assumed to contain particulate and
-------
Phase 1 Sampling
particulates
Gas Stream
gaseous inorganics
'"
.....
....
gaseous organics
T
bioassay
Phase 1 Output
no particulates collected
SSMS on portion of
Tenax, result no metals
detected
L.C. into 8 fractions;
three fractions contain.
all of the samplei
equal to 100 g/m~
no cytotoxicity
mutagenicity
Phase II Sample
& Analysis
delete
delete
Tenax absorber;
or other means;
GC/M.S. analysis
of the sample
fraction.
bioassay
Figure 1. Phased environmental assessment of gaseous sample containing
no particulate material.
~
Phase II Output
delete
delete
specific organics
identified; mass
emission rate of
selected organics.
no cytotoxicity
mutagenicity
carcinogenici ty
-------
NO
LEVEL 1
ENVIRONMENTAL
ASSESSMENT
(SCREENING
ANALYSIS)
FACILITY OR PROCESS
TO BE ENVIRONMENTALLY
ASSESSED
... CENTRALIZED
DATA BANK
"
I
I
I
I
I
I
I PRIORITIZE oC
I STREAMS
r------J ~
: LEVEL 2
I ENVIRONMENTAL
I ASSESSMENT
: (COMPREHENSIVE
I SEMI-QUANTITATIVE
I ANALYSIS)
! !
~ REPORT: EMISSION RATES OF SPECIFIC COMPOUNDS
r I
: I I I
I HEALTH EFFECTS MONITORING CONTROL ENVIRONMENTAL
I PROGRAMS PROGRAMS TECHNOLOGY IMPACT STATEMENT
I I I I - I
L------_-__L - -----..l- --------1- ---------1
""
-..J
II.)
Figure 2. Environmental assessment and other environmental programs.
-------
Table 5. Uses of outputs from the
phased environmental assessment
.
.
Input to supported EPA programs.
I nformation necessary to undertake requisite control
technology and health effects studies.
Can be used to prepare an environmental impact statement.
.
Since the sampling and analytical procedures are
optimized for each pollutant class and for each process,
specific approaches cannot be described in detail. How-
ever, several manuals are being issued by the Process
Measurements Branch which describe sampling and
analytical options for level 2 sampling and analysis (refs.
2,3,4). In addition to the optimized sampling and analyt-
ical procedures which are used, sufficient samples must
be obtained for bioassay (500 mg). To acquire such a
sample size, level 1 procedures may have to be used.
The specific data which are generated by the phased
environmental assessment can be used as shown in table
5. The relationship between environmental assessment
and other environmental programs is shown in figure 2.
Conclusions
1. The phased approach to testing is more cost
. effective than alternative schemes.
2. Each phase of the environmental assessment
yields information concerning the environmental insult
associated with an energy related technology. The
primary differences are cost, accuracy, specificity, and
completeness of the analysis, and representativeness of
the sample.
3. If the equivalent of level 1 information exists,
then the researcher can prioritize the process streams
and proceed directly to level 2.
REFERENCES
1.
J. W. Hamersma and S. L. Reynolds, "Environ-
mental Assessment Sampling and Analytical Strate-
gies and Test Sampling and Analytical Costs," draft
report, Oct. 1975.
Procedures for Process Measurements: Trace I n-
organic Materials, TRW Systems Inc., July 1975.
P. W. Jones et aI., "Analytical Methodology for
Organic I ndustrial Effluents," draft report,
September 1975.
W. Fairheller et aI., "Sampling Techniques for
Organic Industrial Effluents," draft report, August
1975.
2.
3.
4.
273
-------
SAMPLING PROCEDURES FOR PROCESS STREAMS
C. A. Flegal and J. W. Hamersma*
Abstract
This paper explores the application of two different
approaches to the sampling of processes for environ-
mental assessment The approaches are illustrated for a
specific type of coal gasification complex. These include
a direct sampling approach in which all sampling and
other necessary information is obtained in a single com-
prehensive effort and a phased approach where the same
level of information is obtained in a two-StEp effort of
varying aims and complexity. The requirements of these
approaches in tErms of sampling methodologies are dis-
cussed in detail with emphasis placed on two-step phased
approach.
The current energy shortage and the emphasis on
the use of energy with minimal environmental impact
will result in the development of a large-scale new
synthetic fuels industry based on coal conversion, oil
shale processing, and biomass conversion. These new
technologies will by nature of both their size and num-
bers have an impact on the environment. I n addition, the
environmental impact assessment of existing industries is
becoming increasingly important. EPA is actively im-
proving environmental assessment methodologies and
prioritizing both the technical and economic require-
ments for these efforts.
A broad range of components that may be of en-
vironmental concern are being considered for environ-
mental assessments of synthetic fuels processes such as a
coal gasification complex. These components range from
the primary gaseous products of carbon monoxide and
hydrogen to the extremely high boiling polynuclear
organic compounds (POM) as well as the entire range of
inorganic species (refs. 1,2). A partial group of items of
interest is shown in table 1. Biological response screening
tests are also of interest.
Two sampling and analytical approaches (the phased
and direct approaches) designed to give the same level of
information for decisionmaking have been evaluated
(ref. 3). The philosophy, information, benefits, and cost
implications of each strategy were presented in detail in
the previous paper.
The phased approach requires two separate levels of
sampling and analytical effort to perform a complete
environmental assessment. The first level of sampling
and analysis has as its goal the identification of the
*The authors are with TRW Systems, Redondo Beach,
California.
pollution potential of a sou rce in a qualitative and/or
semiquantitative manner. This is done because at the
initiation of an environmental assessment, little is known
about the specific sampling requ irements of a source
both practically and technically. Thus, the first level
utilizes qualitative and/or semiquantitative sampling and
analysis procedures, identifies problem areas and allows
the elimination of certain streams, components and
classes of materials from further consideration in the
overall assessment. Physical and biotesting is of a qualita-
tive and survey nature consistent with the characteristics
of the sample.
The second level sampling and analysis effort after
having been focused by level 1 is designed to provide
information that will resolve the questions or data gaps
identified in level 1. Level 2 sampling and analysis will
provide definitive data that will allow the environmental
assessment of a source to be established. In order to
perform this in a timely and cost-effective manner, the
basic questions to be answered and major problem areas
must already have been defined in leveI1.Consequently,
level 2 sampling and analysis is characterized byobtain-
ing representative samples, accurate stream flow rates
and by identification and quantification of specific
organic and inorganic chemical classes and individual
species. I n this effort, biotesting in selected areas is ex-
panded to include dose response data and also carcin-
ogenicity testing.
The level 1 and level 2 sampling and analysis efforts
are intimately linked in the overall environmental assess-
ment effort. Levell develops and focuses the questions
that must be answered by level 2. For example, if a level
1 test showed the presence of three to seven ring aroma-
tics (PNA), level 2 sampling and analysis would be de-
signed to' determine the exact quantities of organics, the
percentage of PNA, and the identity of as many specific
PNA as possible (ref. 3). The results of this effort will
provide sufficient information in the areas of physical
characteristics, organic and inorganic chemical species
characterization, and biochemical assaying such that
stream control priorities and an initial estimate of
process/control system regions of overlap can be
established.
In this paper, the various sampling procedures for
environmental assessments will be quickly reviewed for
level 1 and level 2 requirements. In addition, a newly
designed flue gas sampling system will be presented.
The Lurgi Coal Gasification Complex proposed by
the EI Paso Natural Gas Company for New Mexico was
chosen as a somewhat typical coal conversion process for
275
-------
Component
Gases
Organic
Liquids and
.501 ids
I\)
......
0)
Trace Elements
Coal, Ash,
and
Particulate
~la ter
Table 1. Some gasification product and byproduct species of interest
Cl ass
Inorganic
Acid Gases
Organic
Sulfurous
Hydrocarbons
Polynuclear
Aromatic Compounds
Nitrogen Compounds
Phenols
Sulfur Compounds
Nonvolatile
Volatile
Ultimate
Proximate
Sulfur Forms
Ash (Particulate)
Dissolved Gases
Organics
Trace Elements
Ions
Gross Characteristics
Item or Property of Intepest
HZ' CO, 02' N2' Ar, NH3' H20
HZS, C02' SOx' NOx' HF, HCl, HCN
CH4' C2H6' CZH4' C3H8' C3H6' C4HI0' C4HS
COS, CS2' CH3SH, C2H5SH
C5 - C12 hydrocarbons, benzene, toluene, xylene, indenes, naphthalenes
pyrenes, fluoranthenes. phenanthrenes, fluorenes, acenaphthenes, benzopyrenes, chrysenes
coronene
pyridine, picolines, 1utidines, quinoline, isoquinoline. quinaldine, indole, carbazole,
acridine
phenol, cresols, xyleno1s, naphthols
mercaptans (thioalcoho1s), thiophenol, thiocresol, benzothiophene
Ba, Be, Ca, Cr, Cu, Mn, Mo, N'j, Sr, V, Zn
As, B, Cd, F, Hg, Pb, Sb, Se, Sn
C, H, N, S, Cl, °
moisture, heat content, ash, volatiles, fixed carbon
total, pyritic, sulfate, sulfide
Si02' A1Z03' Fe203' TiOZ' PZ05' CuO, CaO, MgO, NaZO, KZO, F, 504 ~
i See items 1 i s ted above
S", 504""
N03-' F-, C1-, Br-, CN-, P04A, C03"', HC03-' SCN-, H+
TOC, suspended solids, oil and grease, specific conductance
BOD, COD,
-------
purposes of planning and analyzing environmental assess-
ment strategies and requirements. The various sampling
locations are shown superimposed on a model of the EI
Paso Plant in figure 1. A generalized flow scheme for
sampling purposes has been simplified for qu ick refer-
ence use into the single block diagram in figure 2 which
shows input and output streams only (refs. 1,2.4,5).
I n the coal gasification complex, with the exception
of the stacks and possibly the solids and slurry streams,
there are generally little or no differences between level 1
and level 2 sampling locations. This is due to the fact
that the predominant number of streams are homogen-
eous throughout the length of their respective transfer
systems and since most samples are of the grab or com-
posite type on both levels 1 and 2, the area of sample
withdrawal need not change. However, sampling tech-
niques will be different for levels 1 and 2 (ref. 3).
The basic sampling strategy (figure 3) has been orga-
nized around the six general types of samples found in
coal gasification or other complex process technologies
rather than around the analytical procedures that will be
required on the collected samples. The six sample types
are delineated below:
Solids and solid slurries. These are the coal input,
bottom ash, and any aqueous stream containing
more than 5 to 10 percent solids.
Liquid streams
Aqueous. Water streams containing less than 5
to 10 percent insoluble solids.
Nonaqueous. Homogeneous streams other than
water streams (usually organic).
Gaseous streams (containing no significant
particulate). These samples include process streams,
process vents, and ambient air samples.
Flue gas containing particulate
Flue gas without particulate. These streams are
essentially the same as the gas streams except special
procedures are used to sample for acid gases and
PAH compounds.
Fugitive Dust Sampling. This includes local source
sampling and plant perimeter sampling employing
high volume sampler techniques.
An examination of error analysis techniques make it
apparent that if one particular error source has a much
larger error than the other sources, the system error is
very close to the largest single error. For this reason, it is
logical to select sampling and analysis procedures for
both levels that yield results of comparable accuracy. In
the phased approach, the selection of sampling points
and analysis techniques relies on the concept that level 1
sampling is oriented towards obtaining survey and/or
semiquantitative results only, whereas level 2 sampling
and analysis is intended to acqu ire more accurately the
data necessary to perform an environmental assessment.
Consequently, a level 1 sample may be taken from any
easily accessible port within the flow scheme of a given
unit. For example, in obtaining a level 1 stack sample,
the probe may be inserted in any convenient location
along the duct ,leading to the stack and a psuedo-
isokinetic sample may be taken in order to obtain quali-
tative data. On level 2, however, where quantitative data
are required, isokinetic samples must be withdrawn from
specific locations away from ducting bends and other
obstructions in order to ensure a more quantitative sam-
ple. Stream parameters such as flow rates, temperature,
pressure, and other physical characteristics will be
obtained on both levels within the accuracy require-
ments of a given level of sampling. Application of these
concepts to sampling a coal gasifier is shown in figure 3.
I n the phased and direct sampling efforts for the
gasifier, many effluent parameters must be determined
onsite due to their unstable nature and it is assume.d
that most other large-scale plants will have similar
samples. Mobile laboratories in the form of vans or trail-
ers are recommended for source assessment efforts such
as stack sampling, ambient air sampling and water quali-
ty measurements. In general, commercially available vans
have been outfitted with a variety of equ ipment neces-
sary for assessing environmentally significant factors for
the source of interest.
The sampling procedures to be employed for solid
materials are quite varied and must be tailored to the
stream being sampled. Sites vary from large piles of
material, contained solid piles such as in hoppers, silos or
vehicles and flowing systems as in pneumatically driven
flowing streams, process streams, or conveyor systems.
The most important factor in the choice of the sampling
device is to obtain a truly representative sample.
The extremely wide variety of situations requiring
sampling makes it difficult to specify exactly what
method should be used or what equipment manufactur-
ers should be specified. Level 1 sampling program should
employ simple procedures that will provide sufficiently
representative sample in a short time period. This sample
can often be a simple grab sample, taken at one time. If
the site to be sampled is a pile of material, equal por-
tions at random locations can be removed, combined,
coned, and quartered. If the sample is to be taken from a
conveyor or elevator, the device could be stopped and a
sample removed over the complete width and depth of
the stream. This is essential so that all size particles be
sampled as size segregation does occur in such a moving
system. If the device is a moving sample in an enclosed
duct or pipe, then a cutter to separate a portion of the
stream can be employed.
I n level 2 and control device development sampling
277
-------
!\oJ
....,
co
Figure 1.
Model of EI P
( aso Lur .
total area - gl coal gasifica .
1.5 sq m') tlon com I
. I. P ex
-------
COAL
WATER
N
....,
CO
EMISSIONS TO AIR
STACK FOR
POWER PLANT
AND BOILERS
ACID GAS
OR
SULFUR PLANT
EFflUENT
INCINERA TlON
STACK
NITROGEN
GAS
PLANT
FLARE
SYS TEM
COOLING
TOWER
LOSSES
VENTS
WATER
EV APORA TED
FROM
PROCESS UNITS
INPUT AND
EFFLUENT STREAMS
FROM
PLANT
FOR MANUFACTURE OF
GAS FROM COAL
I..
ASH TO COAL MINE WATER TO
OR SLURRY POND EVAPORATION POND
OR TREATMENT
-----I
y
REFUSE
Figure 2. Gasification plant effluent flow diagram.
SALEABLE PRODUCTS
PRODUCT GAS
@
(BY PRODUCT
SULFUR)'
@
(BY PRODUCT TAR)'
BY PRODUCT OILS 0
BY PRODUCT
NAPHTHA 0
AMMONIA SOLUTION
(201'0 AMMONIA) @
BY PRODUCT
PHENOLS @
, NOT ALL UNIT£
-------
!
~COAI & e(-'lT'~M
A~II 1:'ICI.ULJIr-iG
1-.' H SllJ!-'!\, STr~L\,\\S
I"
Lf'/h I
C, '.\B
SAI.\PlE
~ SITES
!\)
CD
o
lE'/~l II
COM~O\ITE
SN-iFLE
lEVfL I LEVEe II
TAP lAP
SAMPlf SAMP lE
8 SITES . SITES
LEVEll LEVEL I' LEVEL I LEVeL II LEVEL I LEVel II LEVEL I LEVel II
TAP COMPOSITE I~IDICATOR ACID GAS SASS MODIFIED HIGH VOL HIGH VOL
SAMPLE TAP SAMfLE lUBES IMPINGERS TRAIN SASS SAMPLER SAMPLER
TRAIN 5" h
A SITES 4 SITES 1~ SITlS 8 SITES 5 SITES 3 SITES CYCLONE CYCLONE
11 SITES 11 SITES
LEVEL I LEVeL II LEVEL I LEVEl II
ORGANIC ORGANIC ORGANIC ORGANIC
SOR3EW SOR3ENI
TRAP TRAP SOR3Et~T SORBENT
I SITE 1 SITE TRAP TRAP
(PRODUCT (PRODUCT 5 SITES 3 SITES
GAS) GAS)
LEVEl I LEVEL II LEVEL I
GRAB
SAMfLl S INTEGRATW GAS GRAB
AMBIENT & A/,'.BlENT
PROCfS»
15 SITES 16 SITES 5 SITES
LEVEL II LEVEL I
GRAB
SN~PLES REACTIVE
(PROCESS) GAS IMPS.
4 SITES 5 SITES
8 SITES
Figure 3. Coal gasifier - sampling site survey chart.
-------
programs, the procedures should provide an accurate in-
tegrated sample. Samples of piles or hopper contained
material should be obtained using a pipe borer, slotted
tube (thief) or auger samplers that can reach from the
top to the bottom of the material. The choice of the
device will largely depend on the physical condition such
as moisture content, particle size and degree of agglome-
ration of the material. The locations of the sample sites
within the material should follow a specific pattern so
that each area of the material can be sampled. Moving
streams should be sampled in one of two ways. One is to
use a riffler or whistle pipe which removes a portion of
the stream continuously. The other method is remove a
portion of the complete stream at specific time intervals.
Numerous devices of this type have been designed that
involve cutters, scrapers, swinging arc samplers, etc. The
most important consideration is that the entire cross-
section of the stream by sampled and that the sample be
removed at the stream velocity (refs. 1,6).
Procedures for sampling aqueous streams are well
established and many suitable procedures are available.
Some of these procedures are based on chemical analysis
of specific components for analysis and biological effect
evaluation and should be chosen for the specific applica-
tion (ref. 3). The organic content of an aqueous stream
can vary widely in composition and nature. Low molecu-
lar weight polar compounds may be water soluble. High
molecular weight material may be insoluble and thus
result in a two-phase nolThomogeneous system. The
non-homogeneous streams tend to be stratified and thus
make the sampling job more difficult. Composite sampl-
ing with a multi-point probe system is the most conven-
ient method to obtain a truly representative sample. If
the stream is homogeneous, then single point probes on
the sampler are sufficient (refs..1 ,6).
At the present time, there is a number of suitable
single point and mUltipoint sampling devices on the
market. Sampling can be for a single or very short time
period, essentially a grab sample, a long term composite
sample or individual samples obtained by a sequential
sampler. The choice largely depends on the reason for
the sample and the available time. For level 1 studies, a
one time sample should suffice, while for level 2 sampl-
ing, it is recommended that the composite type or the
sequential sampler be employed. The latter device yields
separate samples which can be analyzed individually or
composited as desired.
Nonaqueous samples will be primarily organic in
nature with little or no inorganic constitutents. In the EI
Paso Natural Gas plant, a portion of the raw coal input
into the gasifier (4 to 8 percent) is converted into a
variety of liquid organic byproducts, ranging from low
boiling naptha to extremely high molecular weight tars
and tar solids. Many of the sampling procedures for
aqueous streams are applicable for other liquid streams
as well. This sampling has been practiced for a consider-
able time by the petrochemical industry, and has been
adequately specified by ASTM and API Standards D 270
and 2546, respectively (ref. 7). I f automatic samplers are
available for the streams in question, they should be
used in accordance with the plant sampling schedule as a
first preference.
When selecting a method for sampling aqueous
liquid streams, it is important to keep in mind the nature
of the stream and the purpose for which it will be
analyzed. For example, samples of low viscosity and
boiling point such as naphtha, light oil, and ammonia do
not stratify or precipitate solids. Thus, it is very likely
that a single grab sample will be sufficient. On the other
hand, tar and heavy oils are viscous materials that can
stratify and precipitate solids when cooling. I n these
cases, care must be exercised to take a hot sample and
use a more elaborate method that avoids sampling only
one part of a stratified stream. In addition, it should be
remembered that pure streams or streams with a narrow
boiling point range are much easier to sample than a
stream ~at contains naphtha through tar. - .-
Fugitive dust sampling will definitely be a consider-
ation for coal conversion processes. The sampling and
analysis of dust in areas utilizing large quantities of coal
and ash establishes an overall ambient particulate con-
centration profile. The primary consideration in bound-
ary sampling involves positioning the high volume air
samplers in such a way that samples representative of the
overall atmosphere are obtained. One way of doing this
is to place four samplers equidistantly at 90° angles to
each other in order to account for fluctuations in wind
intensity and direction. Samples taken over a one week
period will provide additional information in establishing
an accurate data profile (refs. 3,8,9).
There are several types of gas sampling requirements
in coal gasification systems. These include:
Process streams, effluents, and vents
Ambient air samples
Flue gas
Gasifier output.
Gas sampling methods utilize either a grab sampling ap-
paratus or an integrated air sampling apparatus. One grab
sampling and integrated air sampling system undergoing
development utilizes an in-line silica gel absorption tube
for the purpose of trapping reactive constituents (acid
gases and particulates) which might otherwise alter the
sample composition by chemical reaction and/or sorp-
tion.
281
-------
Ambient air samples are taken to assess the extent
of the problem by determining the overall plant contri-
bution of toxicants to the atmosphere. The primary ob-
jective in obtaining this type of sample is to ensure that
the atmosphere sampled consists of a representative seg-
ment of the process environment. This is best accomp-
lished by establishing a profile derived from the simulta-
neous operation of four integrated sampling units placed
equidistantly from one another in the same manner as
for fugitive dusts.
Obtaining a representative sample of the effluent
from process power plant and incinerator stacks is a
complex task, since all process parameters vary with
both time and location within the stack. Consideration
must be given to a variety of factors, such as flow rate,
sampling time, changes in loading (attributable to cyclic
process changes). and suitable impingement techniques
needed to entrain the various components of interest
(table 1) (refs. 9,10,11). In the Lurgi system, the power
plant and acid gas incinerator stacks (streams 7 and 13
of figure 2) will also need to be sampled.
Recently, a new train for sampling flue gases for
environmental assessments has been designed and tested.
The project was an integral aspect of an overall EPA
program to acquire a variety of industrial flue gas partic-
ulate samples in sufficient quantity to permit evaluation
of comprehensive biological and chemical testing meth-
odology for environmental assessments. I n the project,
the utility of a prototype series cyclone sampling train
was carefully evaluated. The train was fabricated and
calibrated by TRW based on a Southern Research Insti-
tute cyclone design and utilizing an Aerotherm Corpor-
ation high volume stack sampler (all designed under EPA
funding). The train (figure 4) was composed of three
cyclones designed to provide large quantities of particu-
late matter, size classified in the ranges of (a) >1 OJ..!, (b)
3J..! to 10J..!, and (c) 1J..! to 3J..!. A final filter to provide a
submicron size cut and a commercial high volume stack
sampler pump and control system (Aerotherm HVSS)
completed the system (ref. 11).
The primary objectives of this program were to ac-
quire the particulate emission samples from eight to ten
industrial sources utilizing the newly designed series
cyclone train and to disburse sample aliquots to partici-
pating agencies and contractors for the following tests:
Cytotoxicity
Mutagenicity
Spark Source Mass Spectrometry
High Resolution Mass Spectrometry
Gas Chromatography/Mass Spectrometry and/or
Others as directed by EPA.
Mitre Corporation functioned as project coordinator for
EPA and the results are presented in a separate paper.
The probes, transform, cyclones and interconnect
lines were fabricated exclusively from 316 CRES stain-
less steel. This alloy was extensively tested and found to
be nontoxic and non-inhibitory in biological compatibil-
ity testing on TRW's Viking Lander Biology Instrument
program to test for life on Mars conducted for NASA.
The probe, transform and 10J..! cyclone had thermo-
couples placed as shown in figure 4 and were heat
wrapped and insulated.
The 3J..! and 1J..! cyclones and Aerotherm filter hous-
ing were installed in the standard Aerotherm HVSS
oven. The filter material chosen by EPA for use was a
Teflon needle felt material pretested and found accept-
able for cytotoxicological purposes. The standard glass
fiber filter material stipulated for the EPA Method 5
train was found to cause an interference in the biological
testing.
One of the major concerns of the project was that
sample integrity must be maintained, i.e., cross-contam-
ination between samples and spurious external contam-
ination from "dirty" brushes, spatulas, bottles, the envir-
onment, etc., must be prevented. It was essential that a
valid correlation be possible between cytotoxicity or
mutagenicity and the trace chemical composition of the
sample and possibly the general type of process. Toward
this goal, all equipment that would contact the samples
(nozzle, probes, cyclones, interconnect tubing, fittings,
filter housings, bottles, brushes, spatulas, petri dishes,
etc.) were cleaned, inspected and packaged per TRW
procedures demonstrated under the VLBI Program to
preclude potentially toxic contaminants from the samp-
les.
Flue gas sampling procedures for organic compon-
ents are currently in the research and development stage.
Organic emissions occur in the condensed, readily con-
densible and gaseous forms. Of these, the gaseous forms
present a serious challenge to the design of efficient
sampling and concentration approaches. In general, the
condensed organic species can be removed by filtration
of an integrated sample over a period of time sufficient
to satisfy the detection limit of the analytical procedure.
Similarly, readily condensed species can be sampled em-
ploying trapping procedures at ice or dry-ice tempera-
tures with care taken to avoid loss of material by micro-
fog (submicron aerosol) formation.
The detection and quantitative measurement of
trace organic vapor species generally requires a concen-
tration step to attain the required detection limits. The
most frequently employed concentration techniques are
solvent scrubbing, condensation (cryogenic trapping),
adsorption on activated carbon, chromatographic equi-
Ii br ation, chemical reactions, and chromatographic
co I u m n t rapping. Potentially, the most attractive
282
-------
STACK
II.)
CD
W
OVEN OR
THERIM.'. BLANKET
1---1
I - - I
I --~-
I I
T~NSFOi :
I I
I I
~ ~
L__J
10 JJ CYCLONE
AEROTHERM OVEN
1---
IMPINGERS
FILTER I
I 1~ CYC;LONE I
I 3 ~CYCLONE I
L___~__J
ICE BATH
FINE ADJ.
BYPASS VALVE
VAC.
LINE
DRY TEST
ORIFICE t:,. P METER
MAG NEHELIC GAGE
COARSE
ADJ. VALVE
Figure 4. Schematic of series cyclone stack sampling train.
-------
method for collecting and concentrating organic substan-
ces from ambient air or stationary emission sources em-
ploys the adsorption and/or partitioning properties of
materials normally used in gas chromatographic analysis
to retain organic substances selectively wh ile removing
the major diluent gases, such as air, nitrogen and water
vapor. The porous polymer materials are used in particu-
late laden streams as a supplemental collection media
following a filter or cyclone collection device. Standard
Method 5 systems have been used for collection of the
particulate in the probe and filters and the volatile
organics in the impinger liquid. A more efficient proce-
dure is to employ the porous polymer in place of or in
addition to the impinger collection.
The possibility of employing the inorganic multi-
cyclone approach for both organic and inorganic sampl-
ing is currently being explored. Aerotherm Corporation
is currently repackaging the cyclone train described
earlier to reduce the size and weight of the cyclone
system, to provide the capability of sampling high tem-
perature systems and to incorporate organic vapor traps.
The system, with the enlarged adsorber, should be cap-
able of collecting both organic and inorganic samples at
the same time. At 5 scfm, sampling times should be
about one hour to obtain sufficient material for both
level 1 chemical analysis and biological testing. The
probe, cyclone and filter collected material would be
analyzed for both organic and inorganic constituents, as
would the adsorber, and the condensed trap. This system
will be tested in 1976 on the Combustion Power
Company CPU-400 test facility during an ERDA funded
test of the coal-fired combined cycle potential of the
system.
REFERENCES
1. J. W. Hamersma and S. L. Reynolds, "Tentative Pro-
cedures for Sampling and Analysis of Coal Gasifica-
tion Processes," EPA Contract 68-02-1412, Task
No.3. (Prepared for the Office of Research and
Development, Environmental Protection Agency,
Research Triangle Park, N. C., March 1975).
2. J. W. Hamersma and S. L. Reynolds, "Review of
Process Measurements for Coal Gasification Pro-
cesses," EPA Contract 68-02-1412, Task No.3.
(prepared for the Office of Research and Develop-
ment, Environmental Protection Agency, Research
Triangle Park, N. C., April 1975).
3. J. W. Hamersma and S. L. Reynolds, "Environmen-
tal Assessment of Sampling and Analytical Strate-
gies and Test Sampling and Analytical Costs," EPA
Contract No. 68-02-1412, Task No.9. (Prepared for
the I ndustrial and Environmental Research Labora-
tory, Environmental Protection Agency, Research
Triangle Park, N. C., Draft Submitted October
1975).
4. H. Shaw and E. M. Magee, "Evaluation of Pollution
Control in Fossil Fuel Conversion Processes," Gasi-
fication; Section 1, Lurgi Process, EPA-
650/2-74-009-C, July 1974.
5. "EI Paso Natural Gas Company Burnham Coal Gasi-
fication Complex-Plant Description and Cost Esti-
mate," Application of EI Paso Natural Gas Co. Be-
fore U.S. Federal Power Commission, Docket No.
CP 73-131, August 16, 1972 (Rev. September 20,
1972).
6. "Approved Procedures for Process Measurements -
Trace Inorganic Materials," TRW Systems Group,
EPA Contract No. 68-02-1393, February 1975.
7. "Standard Method of Sampling Petroleum and
Petroleum Products," ASTM Committee D-2, 1971
Annual Book of ASTM Standards, Part 18,
D270-65,47-71.
8. "Planning the Sampling of the Atmosphere," ASTM
Committee D-22, 1971 Annual Book of ASTM
Standards, Part 23, D1357-57, 291-297.
9. D. L. Brenchley C. D. Turley, and R. G. Yaime,
"Industrial Source Sampling," Ann Arbor Science,
Ann Arbor, Michigan, 1973.
10. "Air Sampling I nstruments," American Conference
of Governmental Industrial Hygienists, 4th Edition,
1972.
11. T. Chamberlain, D. Jones, J. Trost, and A. Grant,
"Fabrication and Calibration of a Series Cyclone
Sampling Train," EPA Contract No. 68-02-1412,
Task No.7. (prepared for the Office of Research
and D eve I opment, Environmental Protection
Agency, Research Triangle Park, N. C., April 1975).
284
-------
ANALYTICAL TECHNIQUES FOR SAMPLE CHARACTERIZATION
IN ENVIRONMENTAL ASSESSMENT PROGRAMS
C. H. Lochmuller*
Abstract
A tiered approach to chemical analysis in an envi-
ronmental assessment program can be both the most
cost effective and information effective strategy. On the
other hand, a tiered approach requires the development
of a more general scheme for analysis than would a po-
tentially more costly direct, single-phase program.
In a tiered approach, the first level of analytical
effort is a standardized inorganic/organic scheme applied
to all samples collected at the site to be assessed. The
second tier analytical effort, however, is optimized for a
given sample type and requires a great deal more profes-
sional judgment and flexibility in its application. This
paper discusses methods of choice, criteria for selection,
analysis costs and report contents for an analytical effort
which could be applied to a tiered environmental assess-
ment program.
INTRODUCTION
The modern analytical chemist has at his disposal a
veritable arsenal of sophisticated techniques applicable
to the task of sample characterization. Given unlimited
resources, each of these techniques could be applied in
turn to every sample taken from a process feedstock,
waste or product stream. Where resources are finite, it is
the analyst's first obligation to devise an analysis strat-
egy which is both cost and information effective. Two
extremes are immediately apparent: (1) direct analysis
of each sample for all its components, or (2) direct anal-
ysis for a limited list of known hazardous materials. The
former approach would be highly information effective
but very costly; the latter, certainly inexpensive but does
not provide information about unsuspected substances
which might be hazardous but are not included in the
list and for which there are no analyses.
An intermediate strategy is a tiered or phased ap-
proach in which analysis occurs at three levels of investi-
gation. The Level I scheme is a general one which
includes physical characterization (e.g., particle size and
morphology), inorganic and organic chemical analysis
and bioassay for cytotoxicity and mutagenicity. It was
*The author is on leave to the U.S. Environmental Protec-
tion Agency, Research Triangle Park, N.C., from Duke Uni-
versity where he is an Associate Professor of Chemistry.
designed to give maximum cost and information effec-
tiveness and have a broad range of appl icabil ity. The
information from this phase is used to prioritize the
various streams in rank order of their potential environ-
mental hazard and to plan an optimized sampling and
analysis strategy for those streams where a hazard exists.
Execution of this latter strategy is Level II of the tiered
approach. In some cases, a further refinement may be
necessary after Level II, e.g., isolation and identification
of a particular chemical species from within a given class.
Level III seems to follow certain key materials over the
extremes of process variation by providing continuous
analysis methods. This paper attempts to show the
elements of the overall strategy of a phased analytical
approach with special emphasis on the component parts
of the generally applicable first phase, or Level I,
scheme. Included are some actual results obtained both
in-house at the I ndustrial Environmental Research
Laboratory, Research Triangle Park (I E R L/RTP) and
externally in the laboratory of a Process Measurements
Branch contractor.
THREE-TIERED ANALYSIS SCHEME
In general, the methods to be described will be
applied in laboratories that are remote to the process
sampling site. As has been previously discussed (viz. Stat-
nick). certain volatile or unstable substances are best
analyzed for onsite by techniques such as gas chromatog-
raphy. The majority of the analyses are done in remote
laboratories because of the need for skilled technical
personnel, special operation conditions and primarily
because of the higher cost-effectiveness of this approach.
A flow chart of the Level I analytical scheme (fig. 1)
shows the various components of the overall scheme and
their interrelationships. The basic divisions are: (1)
physical, (2) chemical and, (3) bioassay. Chemical analy-
sis is subdivided into separate schemes for inorganic and
organic components. The bioassay procedure includes in
vitro test procedures, e.g., cytotoxicity or mutagenicity.
PHYSICAL CHARACTERIZATION OF SOLlDS-
LEVEL I
Level I physical characterization is inexpensive and,
in most cases, nondestructive. I n flue and stack gas sam-
ples, particle size distribution information is provided by
285
-------
Solids
I\,)
CIO
C)
Morphology
- Inorganic .Spark
Mass
Extraction
Field If Necessar~ Separation
Samples - - Organic into 8 Fract
WT% Determin
Recombined f
''-' (if nccessar
provide suf
Illi:lterial)
t
in vitro
- Bioassay Cytotoxi ci ty
Mutaqcnicity
Cost:
$15
Source
Spectroscopy
IR of
each fraction
Organic
Cl ass-
Interpretation
j
---...
ions
ation
ractions
y to
ficient
and-
Figure 1. Flow chart of level 1: environmental assessment analysis scheme.
Cost:
$500
._.Cost:
$800
-------
the "cut-point" (ds 0) values of the series cyclones of the
sampling train (i.e., 10, 3, lum respectively). For other
solid samples, optical microscopy for gross morphology
is one suitable technique and, in most cases, will detect
distinctive substances such as asbestos. A photomicro-
graphic record of the sample is both worthwhile and low
cost. More sophisticated techniques such as scanning
electron microscopy are considered too expensive for
routine inclusion.
CHEMICAL ANALYSIS -
LEVELl
The intent of the chemical analysis scheme at this
phase of assessment is to provide qualitative and quanti-
tative information relevant to the element composition
of process samples and the identity of the major classes
of organic chemical compounds present.
INORGANIC ANALYSIS
Inorganic analysis at Level I utilizes the Spark-
Source Mass Spectrometry (SSMS) technique to achieve
qualitative and semiquantitative elemental analyses on
all solids, particulates, filterable solids from liquid
streams and the residues of evaporated liquid samples.
This technique was selected from the many others avail-
able (e.g., atomic absorption, emission and x-ray fluores-
cence spectrometry and neutron activation analysis)
because of its general multi-element capability, accept-
able detection limits, speed and cost. Precision and
accuracy of a factor of 2 and a detection limit of 1 ppm
for elements analyzed are specified. Spark source mass
spectrometric analysis for some 80 elements costs about
$300 commercially including sample preparation. A
comparison of SSMS and several other techniques can be
found in table 1.
ORGANIC ANALYSIS
The objective of Level I organic analysis (table 2) is
to achieve quantitative estimate of the predominant
classes of organic compounds present in samples taken
from process streams. Organic compounds are often
described in terms of classes. This terminology has a very
specific connotation. Classes of organic compounds
include, for example, alcohols, ketones, aldehydes,
caboxylic acids, amines, etc. While it is indeed possible
to fractionate complex mixtures into classes, this is a
difficult and costly procedure. At Level I, the strategy is
to isolate well-defined fractions by conventional liquid
chromatography. Under controlled conditions, the con-
tent of a chromatographic fraction (in terms of class
types) is fairly predictable. An example of the relation
between organic class and chromatographic fraction is
illustrated in figure 2.
Organic extracts are resolved into eight fractions by
liquid/solid chromatography on silica gel, utilizing a pre-
scribed series of eight solvent mixtures as eluants. The
fractions are evaporated to constant weight by methods
that minimize evaporative loss of the constituents of
interest. The weight of each fraction is determined to
0.1 percent or 10 g. As an aid to identification of the
constituents of each fraction, an infrared adsorption
spectrum is obtained and interpreted. I nterpretation at
this stage is in terms of classes which may be present.
For the infrared analysis, a grating instrument covering
the range 4,000-600 cm -1 is required. The total cost of
a Level I organic analysis is estimated at $450, not in-
cluding the infrared interpretation.
BIOLOGICAL ANALYSIS
The biological tests included in the Level I analysis
scheme are intended to provide indications of potential
biohazards independent of chemical analysis. This
environmental assessment tool is an important aspect
since chemical information, as mentioned earlier, is an
ambiguous predictor of biological activity. Levell chem-
ical analysis is a search for danger signals; bioassay pro-
vides additional indicators.
Like the chemical analysis, Level I bioanalysis must
remain relatively simple and inexpensive. For this rea-
son, only in vitro or "test tube" methods may be
seriously considered part of the first level. The selected
methods discussed in this section provide an estimation
of acute toxicity and certain types of mutagenic activity.
I n addition to being of direct interest, mutagenic behav-
ior is an effective screening indicator for carcinogenic
activity. Although not all mutagens are carcinogenic,
nearly all carcinogens cause mutagenesis.
An estimate of the acute toxicity could be deter-
mined by means of the rabbit alveola macrophage
procedure. Although the system presents opportunities
for considerably more sophisticated and subtle studies,
only cell mortality at a single pollutant concentration,
compared to controls, would be utilized as part of Level
I.
Particulate samples are added-in weighed quanti-
ties, to the culture medium and incubated before adding
the test cells. After a second incubation, the number of
dead and living cells are determined by dye exclusion
techniques. Water-immiscible liquid samples are treated
identically except for the initial incubation period.
Water-immiscible samples may be solubilized by treat-
ment with dimethyl sulfoxide. Organics trapped by
287
-------
Table 1. Comparison of analytical techniques
Technique
DC Arc/AC Spark (OES)
ICPOES
I\)
co
co
X-Ray Fl uores cen ce (XRF)
Atomic Absorption
Spectroscopy (AAS)
Neutron Activation (NAA)
Spark Source Mass
Spec (SSMS)
Isotope Dilution (SSMS)
Multielement Capability
Unable to run Hg, Se, Sb
or F
Unable to run F, additional
elements up to 40 depends
on instrument
Unable to run Li, Be or B
None
Be not done effectively
F, Pb requi re sped a 1
handling
All elements can be
determined
Cannot do Hg, As, Be,
Mn, F or Na
Accurary
+30-50% semi quant.
~10% quant.
-+40%
Qualitative to +10% depend-
ing on element and matrix
+10-<1% depending on con-
centration and standards
+30% semi quant.
2-3% quant.
100-300% 'Iqui ck and di rtyll
+30% semi quant.
<5%
Sens i ti vi ty
Generally acceptable
varies with matrix
Acceptable
Vari ab 1 e depen ds on
element and matrix
Acceptable, varies
with element
Acceptable, varies
with matri x
Acceptable
Acceptab 1 e
-------
Table 2. Level 1 organic analysis
Compound
Compound Class Fraction Contained
hydrocarbon 1
aromatic 1 ,2
ester* ~,i
ketone 4
phenol 4,~
alcohol 5
urea (diamide, 6
aromatic
alkyl aryl sulfonic acid 2,8
amine (6)
sulfoxide 7
carboxylic acid n.f
Identified as
n-octadecane
hydrocarbon
anthracene
anthracene
ester*
stearone
ketone
phenol
l-octanol
phenol
l-octanol
5-diphenyl urea
aromatic amide
I\J
co
co
p-toluenesulfonic acid
alkyl benzene sulfonate
(acid or salt)
octadecylamine
amine
dimethylsulfoxide
DMSO
suberic acid
n. i
*Suberic acid octylester formed in sample work-up from the acid, octanol and the presence of
an aryl sulfonic acid
Note:
underline indicates fraction containing major portion of material.
-------
1
2
3
4
5
6
7
8
ALIPHATICJ AROMATIC HYDROCARBONS
ESTERS
KETONES
f'.)
co
o
ALCOHOLSJ PHENOLS
AMI DES
ALKYLJ AROMATIC SULFONATES
SULFOXIDES
Figure 2. Liquid chromotographic fractions vs. class types.
-------
porous polymer adsorbents are extracted with a
compatible solvent (pentane in the case of Tenax.GC,
ENKA, Holland), reduced in volume, and then taken up
in dimethyl sulfoxide for addition to the test system.
The recommended test for mutagenesis is carried
out using a B. Ames.type procedure with selected strains
of micro.organisms. At Level I, only one concentration
is tested and only one solvent system is employed.
Dimethyl su Ifoxide has been selected as the solvent most
likely to give "worst case" results, a desirable goal at
Leve I I.
Solid samples are dissolved as completely as possible
in dimethyl sulfoxide, filtered, and added to the test
system. Liquid samples are added to dimethyl sulfoxide
and then to the test system.
Skilled technicians under close supervision of a pro.
fessional cytologist and bacterial geneticist are required
for the cytoxicity and mutagenicity testing, respectively.
Laboratories must be well equipped and carefully
managed to prevent chemical or biological contami.
nation of samples and to avoid health hazards to the test
personnel. I nterpretation of results must be accom-
plished by experienced personnel. Therefore, this work
would only be conducted in laboratories approved by
EPA health effects programs. Estimated combined cost
is $800.00 for cytotoxicity.mutagenicity testing.
CHEMICAL COMPONENTS OF
LEVEL IIN PRACTICE
Of the chemical analysis procedures used in Levell,
SSMS is the most straight-forward. It is tempting to list
many of the analytical results we have achieved with this
technique, but such lists would have little meaning with.
out background information related to each source.
More to the point of this paper are the data from some
feedstock samples run "in-house" using both SSMS and
NAA. As can be seen from table 1, the two methods are
essentially equivalent within the errors involved for the
elements analyzed. SSMS yielded more overall informa.
tion without special effort.
The results of a Level I organic analysis on a syn.
thetic sample prepared in our labs and run "blind" by
our contractor is shown in table 3. As can be seen, the
combination of liquid chromatography- infrared spec-
troscopy yielded a sign ificant amount of specific as well
as class information. Utilizing conventional infrared
techniques, the limit of detection varied from 20.200 /1g
depending on the relative infrared sensitivity. Utilizing
more advanced techniques such as Fourier Transform
Infrared Spectroscopy, the limit of detection could be
reduced 100.fold and more speciation information could
be obtained at very little increased cost.
LEVE III - Analysis Strategy
level II analyses are carried out on streams that
have been prioritized in Level I Environmental Assess-
ment programs. It is not possible to present a chemical
analysis strategy for Level II in the detail possible in
Level I. I n Level II, every problem is stream specific and
will require individual planning. The techniques used in.
clude the entire "arsenal" referred to previously, but
could be a more quantitative application of Level I
methods, modifications of these or new techniques.
What can be said is that Level II methods will either be
classification or speciation techniques. The former iden.
tifies a particular compound type-e.g., oxide, phenolic,
amine, etc., while the latter identifies a particular com.
pound itself-e.g., PbO, m.bromophenol, aniline, etc. The
latter will rarely be used and only in those cases where
specific control devices must be implemented. The
speciation techniques are often more expensive to exe.
cute because of the need to separate the material of
interest from its matrix.
In Level II, bioassay is expanded to include "dose-
response" measurements and tests for carcinogenicity.
This further delineates the potential human hazard and
the nature of the problem.
LEV E L III - Analysis Strategy
In Level III, the limits of process variation and its
effect on certain "indicator" species, e.g., total sulfur for
sulfur compounds, nondispersive I R for nitro com.
pounds, are studied. In this phase, the long-term average,
process-related environmental hazard is assessed. It is
anticipated that at regular intervals, Level I or II
methods will be used as a check on the significance of
the continuous process.monitor results.
SUMMARY
A phased approach to chemical, physical and biolog-
ical analysis is proposed, which is generally applicable.
The approach promises to be cost and information effec.
tive as a component of an overall environmental
sampling and analysis plan. The advantages over alterna.
tive, direct approaches would seem to be the possiblity
of comprehensive survey of all process streams within
the limits of finite resources for process and control
technology development.
291
-------
Table 3. Comparison of NAA and SSMA analysis of coal sample
COAL
~g/gram of Sample
Neutron Activation* Spark Source* Neutron Activation Spark Source
Li «.2) Se 1 1
Be «.2) Br 10 2
B 50 Rb 20
F 5 Sr 30
Na 2000 3000 Y 3
- Mg 800 Zr 4
Al 6000 Nb 1
Si 9000 Mo 3
P 5 Ru .5
f',) S 900 Pd .3
co Cl 1100 1500 Ag .3
f',)
K 2000 5000 Sb 0.5 .3
Ca 4000 4000 Cs 0.8 1
Sc 3 10 Ba 30
Ti 500 Eu. .2
V 20 La 4 2
Cr 50 30 Ce 3
Mn 160 160 Pr 1
Fe 11 000 6000 Nd 3
Co 4 5 Gd 1
Ni 25 Pb 1
Cu 10 Th <1
Zn 60 4 U <1
Ga 5 Sm 0.8
Ge 8 Hg 0.05
As 2 20 Cd 4
*Mean values, carrying a variability of +2, -1/2 of mean value
-------
EVALUATION OF PARTICULATE CHARACTERIZATION TECHNIQUES*
H. Mahar, Ph.D.
N. Zimmerman, Ph.D.t
Abstract
The Environmental Protection Agency's measure-
ment program for environmental assessment is address-
ing the need for a rapid, effective, and inexpensive
means for evaluating the potential environmental haz-
ards associated with industrial particulate emissions.
Chemical analyses and cellular biological assays were
performed on size~/assified particulate material col-
lected at nine industrial sites. The exercise was formu-
lated to determine whether the chemical analyses or the
bioassays, alone or in combination, were sufficient to
assess the hazards associated with particulate emissions.
This program has identified the need for size~/assi-
fied particulate matter to be utilized in the various
chemical or biological tests. Elemental analysis and par-
tial organic characterization of the particulate samples
have been performed. A cellular bioassay, utilizing rabbit
alveolar macrophages (RAM), has been used to provide a
rank ordering of particulate samples in terms of their
observed cytotoxic activity. A screening technique utiliz-
ing several bacterial indicator strains is being used to
evaluate the mutagenic potential of the particulate
samples. Attempts to correlate observed biological
activity with chemical analyses are ongoing.
Several proposed screening methods which may be
used to characterize the potential hazards associated
with particulate material released from various sources
are currently being field tested. Both biological activity
and chemical composition of the various samples are
being determined. Chemical analysis, when used alone,
cannot provide sufficient data for complete evaluation
of pollutant emissions due to the effects of synergism,
antagonism, and differing degrees of biological avail-
ability. The presence or absence of given toxic compo-
nents does not preclude nor indicate the relationship of
the effluent to a suspected toxic effect. Bioassay tech-
*This work was carried out under Contract No. 68-02-1859
with the Environmental Protection Agency.
tThe authors are members of the technical staff,
Environmental Planning and Engineering Department, MITRE
Corp., McLean, Virginia.
niques, on the other hand, can be used effectively to
assess potential toxic effects, including interaction ef-
fects.
Whereas cellular bioassay takes into account the
biological activity of the sample, including synergistic
and antagonistic effects, it in no way pinpoints the spe-
cific compounds responsible. The most cost-effective
screening method would appear to involve the use of
cellular bioassay to determine which effluent samples are
biologically active, together with the use of chemical
fractionation and analysis to ascertain which agents are
responsible for the observed activity.
Size-classified particulate material collected at nine
industrial sites has been subjected to chemical analyses
and biological assays to determine whether the proce-
dures, alone or in combination, provide a rapid, effec-
tive, and inexpensive means to evaluate potential hazards
associated with particulate emissions. The exercise was
formulated to provide actual field testing data to verify
the measurement program for environmental assessment
being initiated by the Process Measurements Branch,
Industrial Environmental Research Laboratory (lERL/
RTP), Environmental Protection Agency. The measure-
ment program has been designed to provide a cost-effec-
tive means to determine the magnitude of the potential
environmental problems associated with emerging energy
technologies during the various development stages of
that technology (refs. 1,2). The measurement program
provides for multi phase testing and characterization,
with increased effort directed towards those process or
effluent streams identified as potential hazards.
I norganic elemental analysis and partial organic
characterization, with emphasis on polycyclic hydro-
carbons of known carcinogenic potential, have been
instituted. The rabbit alveolar macrophage cytotoxicity
bioassay and mutagenicity screening. tests using several
bacterial strains (Salmonella typhimurium) are utilized
to predict the acute toxicity and mutagenic behavior of
the particulate matter.
This research program will evaluate these biological
and chemical tests, performed alone and in combination,
for their ability to assess the potential hazards associated
with particulate emissions. An ancillary objective will be
to consider the need for and performance of the sequen-
tial cyclone sampling train used to collect the particulate
sample. At this time, however, the exercise is not com-
plete and only preliminary conclusions can be drawn.
293
-------
SAMPLE COLLECTION
A series cyclone sampling train was developed to
provide a capability to collect large quantities of size-
classified particulate matter from a variety of industrial
sources so that chemical and biological characterizations
of the sample could be achieved. The cyclone train
design called for collection of samples according to
particle aerodynamic diameter in ranges of >1011, 3-1011,
1-311, and <111 (by filter). The sampling train was
intended to provide representative material from the
different industrial sources.
TRW Systems Group* fabricated the sampling train
according to drawings provided by Southern Research
Institutet (SRI). Since particulate samples were to be
collected for cytotoxicity testing, construction materials
were selected for their nontoxic qualities. Assessment of
toxicity was based upon information generated by TRW
on the NASA-sponsored Viking Biology Lander Instru-.
ment Program. The field sampling configuration of the
series cyclone train is presented in figure 1. The sampling
train, as prepared by TRW Systems Group, should not
be considered optimal since the intent was to provide a
fail-safe sampling system capable of collecting a variety
of particulate samples under a variety of operating con-
ditions. In its report, TRW Systems Group urges that the
train be redesigned and recalibrated to minimize weight,
to reduce the number of components, and to identify
more precisely the minimum critical flow that gives
acceptable size classifications (ref. 3).
Industrial sites sampled were selected to provide
particulate material possessing a variety of physical and
chemical characteristics, as well as an anticipated range
of cellular toxicity. In order to evaluate the cyclone
train, a wide range of sampling conditions under which
the train would operate was chosen. Table 1 provides the
sampling logistics at nine industrial sites. A tenth site
was sampled, but due to unexpected changes in source
emission levels, the sample mass collected was too small
to permit further analysis. Note that the actual sampling
locations at the industrial sites were not consistent with
respect to control devices. The particulate material
collected should not be construed to represent the actual
emissions of that particular site. All sampling, handling,
shipment, and storage procedures were intended to
provide minimal contamination in subsequent chemical
and biological characterizations. An intensive effort was
expended during field operations to ensure sample
integrity.
*One Space Park, Redondo Beach, California 90278.
t200 Ninth Avenue, South Birmingham, Alabama 35205.
CHEMICAL ANALYSIS
Three types of chemical analyses have been per-
formed on the 3-1011 and 1-311 sized particulate samples.
Assays were performed for 75 different elements using
spark source mass spectrometry (SSMS). The SSMS tech-
nique provided a lower detection limit of 0.1 parts per
million by weight (ppmw) for each element. No repeti-
tive analyses of individual particulate samples were
performed. Polycyclic organic constituents of the partic-
ulate samples were determined using gas chromatograph-
ic mass spectrometry (GC-MSJ and high resolution mass
spectrometry (HRMS). The identification of the poly-
cyclic hydrocarbons focused on those of known carcino-
genic potential or of structurally similar compounds.
Using GC-MS, the detection limit for individual hydro-
carbon species was slightly less than 10 ng. No repetitive
analyses of individual particulate samples were per-
formed.
BIOLOGICAL ASSAYS
In this study, two in vitro biological assays have
been utilized to determine the acute toxicity and the
mutagenic potential of the 3-1011, 1-311, and <111 particu-
late samples from the various industrial sources. The
mutagenic bioassay utilizes several bacterial indicator
strains (histidine deficient Salmonella typhimurium:
T A-1535, T A-1537, and T A-1538), with reversion to
prototrophy indicative of mutation (ref. 4,5). Bacterial
exposure to the particulate samples, using dimethyl-
sulfoxide as the solvent vehicle, occurs on plates by the
agar overlay method. All tests are run with and without
a mouse liver activation system.
The rabbit alveolar macrophage (RAM) has been
used to determine the potential acute cytotoxicity of
size-fractionated particulate matter collected from a
variety of industrial sources. The alveolar macrophage
exists as a pulmonary free cell and provides an early line
of defense against inhaled foreign bodies. Because of its
phagocytic activity, it is particularly useful in the toxico-
logic evaluation of airborne particulate material.
The RAM testing procedure is a modification of
that developed by Waters et al. (refs. 6,7)- An initial
cytotoxicity screening of a particulate sample at a final
concentration of 1,000I1g/ml of culture medium was per-
formed in triplicate. The particulate sample was incu-
bated for 20 hours in the culture medium to allow for
dissolution of any soluble component(s) of the particu-
late matter. Macrophage cells were then added to achieve
a final concentration of approximately 5 X lOs cells/ml.
The cultures were then incubated for 20 hours (@ 37° C
in a humidified 5 percent CO2 atmosphere). Cell number
was estimated by hemocytometer count and cell viabili-
ty was estimated by light microscopy on the basis of
try pan blue dye exclusion.
294
-------
PITOT
SAMPLING
PROBE
I\,)
CD
U'I
D
OVEN OR
THERMAL BLANKET
I--
I
I
I
I
I
I
TRANSFORM I I
I I
I I
~ ~
L__J
10 ILCYCLONE
D ; DISASSEMBLY POINT FOR
CLEANING
AEROTHERM OVEN
r--
I
I
I I
I I
I 31L CYCLONE I
L______J
AEROTHERM
GAS METER
T.C.
~~ ~~
DRY TEST
ORIFICE AP I METER
MAGNEHELIC GAGE
3 PUMPS
Figure 1. Series cyclone train, field sampling configuration.
VAC
LINE'
VACUUM
\ GAGE
COARSE
ADJ. VALVE
FINE ADJ.
BYPASS VALVE
AI R TIGHT
V AC. PUMP
-------
Table 1. Logistics of sample collection
SOURCE OPEN HEARTH COKE OVEN BASIC OXYGEN IRON OIL FIRED CERAMICS COPPER ALUMINUM WASTE WATER
FURNACE HEATER FURNACE SINTERING POWER PLANT PLANT SHELTER REFINERY TREATMENT
(ROTATING PLANT, SLUDGE
KILN) INCINERATOR
SAMPll NG ELECTROSTATIC BASE OF DOWNSTREAM OF INLET TO WET SCRUBBER BETIIEEN OUTLET OF INLET TO DUCT BETI/EEN
LOCATION PRECIPITATOR STACK ESP. IlAG HOUSE INLET PRIMARY AND ROASTER BAG HOUSB FURNACE AND
(4 DUCT DIAM- DOWNSTREAM OF SECONDARY REVERBERATOR, IIATER QUENCH
ETERS DOWN- INDUCTION FAN, CYCLONE INLET TO IlAG
STREAM FROM ESP, UPSTREAM OF SYSTEM HOUSE
4 DUCT D IAM- STACK
ETERS UPSTREAM
N OF STACK)
c.o
en
yo . 350' - 425' 400' 150' - 225' 400' 170' 510' 250' 210' 1100'
STACK
T' CYCLONE' 350' 3BO' 270' 350' 195' 400' 275' 300' 410'
(10")
T' OVEN' 350' - 400' 380' - 400' 270' 390' 250' 400' 300' 300' 380'
flOW RATE 4.8 SCFM 3.5 SCFM 4.6 SCFM 5 SCFM 4.5 SCFM 3.8 SCFM 4.7 SCFM 5 SCFM 4.7 SCFM
THROUGH
SAHPll NG
TRAIN
TOTAL 5 HOURS 5 HOURS 5 HOURS 2 HOURS 2 HOURS 1. 25 HOURS 1 HOUR 2 HOURS 5 HOURS
SAMPl/ NG (CONTINUOUS) (CONTINUOUS) (CONTINUOUS) (INTER- (CONTINUOUS) (INTER- (CONTINUOUS) (INTER- (CONTINUOUS)
TIME HITTENT) HITTENT) HITTENT)
. OEGREES FAHRENHEIT
-------
Samples found in the initial screening to produce
net cell death of greater than 15 percent, as compared
with controls, were retested in a preliminary concentra-
tion-response t~st using particulate concentrations of
1,000/1g/ml, 300/1g/ml, arid 100/1g/ml of culture medi-
um. The pH values of the cultures were monitored
throughout, and if shifts below 6.8 or above 7.6 oc-
curred, the sample was tested under both unadjusted and
adjusted conditions.
In addition, an attempt was made to ascertain
whether the toxicity of a given particulate sample was
due to the particles themselves and/or soluble compo-
nent(s) released into the medium. The particulate matter
was incubated in the culture medium for 20 hours and
then removed from the medium via centrifugation. The
supernatant and centrifuged particles (resuspended in
fresh medium) were then tested for cytotoxicity. Figure
2 indicates the testing sequence used in this study.
RESULTS
Analytical results presented here have been chosen
for illustrative purposes. * Particulate samples collected
in the largest cyclone (>10/1) were not analyzed because
it was felt they represented a minor air pollution hazard
since they would, in all probability, settle out of the
atmosphere in a short period of time. Those particulate
samples collected on filters «1/1) were subjected only to
partial organic characterization by GC-MS and to acute
cytotoxicity screening by the RAM procedure.
Selected results of spark source mass spectrometry
analyses are provided in table 2. For brevity, the concen-
trations of only 20 of the 75 elements assayed are in-
cluded. Selection of those 20 elements was based upon
Threshold Limit Values (TL V's), concentration, and
inherent toxicities. The partial organic characterization
of the samples, accomplished by gas chromatographic
mass spectrometry, is provided in table 3. Only those
organic species present in detectable amounts are listed.
Note that three particulate samples were not analyzed
(i.e., coke oven heater, 3-10/1 sample; oil-fired power
plant, 1-3/1 and 3-10/1 samples). Regarding table 3, it is
noted that if collection cyclones were operating at tem-
peratures above 3500 F, much of the organic fraction
would have been lost as vapor (see table 1).
Selected results of the rabbit alveolar macrophage
(RAM) cytotoxicity tests are reported in figures 3 and 4
to indicate the utility of the testing procedure. The
graphic displays provide mean values 1 one standard
deviation. Figure 3 indicates the relative ranking of
"Complete tabulations are available from the authors.
cytotoxic particulate samples collected from nine indus-
trial sources. Cytotoxicity is reported as percent viability
(percent of viable cells remaining after exposure period)
at a particulate concentration in the test medium of
1,000/1g/ml. The data represent an average observed
cytotoxicity from both the initial screening test and the
preliminary concentration-response test (@ 1 ,000/1g/ml)
including pH adjusted and unadjusted conditions (i.e.,
there were six determinations for each sample).
Figure 4 provides preliminary concentration-
response relationships between different fractions of
particulate samples collected from a sludge incinerator.
The log of the particle concentration in the test medium
is plotted against the viability of the RAM cells, reported
as percent of controls (i.e., percent survival) for the
particulate plus supernatant (P+S) fraction, the particu-
late (P) fraction, and the supernatant (S) fraction. In
figure 4, the least-squares regression lines for the data are
provided along with mean values (1 one standard devia-
tion) of the observed viability, reported as percent of
control. Note that the 1-3/1 sample appears more cyto-
toxic than the 3-10/1 sample.
The bioassay screening procedure to determine the
mutagenic potential of the various particulate samples
has not been completed. No conclusions can be drawn at
this time.
SUMMARY AND CONCLUSIONS
The utility of several testing methods to assess the
potential hazards associated with industrial particulate
emissions has been established. A rigorous statistical
evaluation is underway to identify any relationships that
exist between observed biological activity and resultant
chemical characterizations. Several preliminary conclu-
sions can be drawn from the completed work:
a. The cyclone train has proven itself a useful tool
for the gathering of size-classified particulate
material present in a variety of industrial efflu-
ents. It has also been shown that the samples
collected in 1-5 hours of sampling train opera-
tion are of sufficient mass to be utilized for
biological and chemical analyses. Furthermore,
the need for size-classified particulate material
has been established, since the bioassays and
chemical analysis have indicated that particu-
late fractions from the same industrial source
do not necessarily possess similar characteris-
tics.
b. Initial studies indicate the potential value of the
RAM cyctotoxicity test as a preliminary screen
to assess the toxicity of particulate material.
The procedure can be used to provide a rank
297
-------
PARTICULATE SAMPLE
INITIAL SCREENING TEST
(at 1000Ilg/ml)
viability after 20 hrs exposure
to particles plus pre-incuba-
tion supernatant (in triplicate)
"PARTICLES PLUS SUPERNATANT"
(if viability of initial
screen <85% of controls)
PRELIMINARY CONCENTRATION-RESPONSE TEST
'"
(0
co
CONCENTRATION:
10001lg/ml
CONCENTRATION:
3001lg/ml
CONCENTRATION:
100\lg/ml
VIABILITY AFTER 20 HR
EXPOSURE TO PARTICLES
PLUS PRE-INCUBATION
SUPERNATANT
(in triplicate)
"PARTICLES PLUS SUPERNATANT"
VIABILITY AFTER 20 HR
EXPOSURE TO PARTICLES
WITHOUT PRE-INCUBATION
SUPERNATANT
(in triplicate)
"PARTICLES"
VIABILITY AFTER 20 HR EXPOSURE
TO PRE-INCUBATION SUPERNATANT
(in triplicate)
"SUPERNATANT"
Figure 2. Rabbit alveolar macrophage cytotoxicity screening test procedure
-------
Table 2. Partial elemental analysis of particulate samples as determined
by spark source mass spectrometry
N
co
co
~ OPEN HEARTH COKE OVEN BAS I C OXY GEN IRON OIL FIRED CERAMICS COPPER ALUMINUM SLUDGE
FURNACE HEATER FURNACE SINTERING POWER PLANT PLANT SMELTER REFINERY INCINERATOR
SlZ ,
ElliMENT 1-3~ 3-10~ 1-3~ 3-10~ 1-3~ 3-1D~ 1-3~ 3-10~ 1-3~ 3-10~ 1-3~ 3-10~ 1-3~ 3-10~ 1-3~ 3-1O)J 1-3~ 3-1O)J
Bi 21 11 48 2.9 5 100 5.7 1.1 140 79 2% 2% 270 59 1500 1900
Pb 1900 900 940 210 170 5000 2100 450 60 33 35% 4% 380 78 1300 2200
11\ I':> 0.53 7.5 0.97 0.78 35 11 0.78 10 5.3 3900 HOO 3 0.19 2.5 5.3
Te U.4 0.15 0.85 0.08 0.15 2 0.58 0.12 2.7 1.3 3600 1300 26 4.3 0.29 0.34
Sb 70 27 3.7 9.7 17 3.5 2.5 7.8 10 5.3 8% 3% 130 66 25 30
Sn 610 340 66 40 140 6.1 4.4 14 36 2. S 4% 4600 100 33 U 5000
Cd 44 5.4 9.4 12 9.7 22 14 0.97 13 6.6 5000 1400 16 1.8 630 790
Ag 65 41 4.2 43 440 35 6.1 4.4 0.57 0.64 1100 2200 2 1.1 600 500
Mo 79 35 17 10 38 17 H 814 23 12 500 2500 69 51 110 33
So 2.3 2.7 9.8 0.4 0.51 310 89 63 42 20 5pOO 3400 22 14 73 52
As 480 190 3.9 52 67 120 35 9.3 890 110 21% 6% 4100 700 100 43
Zn 2% 2% 1100 4600 5000 HOO 760 900 360 220 1.7% 0.7% 160 56 5000 5000
Cu 2900 3600 170 800 HOO 2200 1000 300 210 93 5% 20% 2000 4600 3% 3%
Ni 470 490 51 46 93 64 26 5.2 88 36 1300 760 2800 3800 4500 3000
Co 111 120 2.6 57 130 59 86 220 3.4 18 180 460 43 27 850 1000
Mn 4400 5000 250 5000 1% 321 640 220 1600 1800 310 880 20 6.3 850 1000
Cr 860 880 94 170 220 52 76 1000 150 120 190 79 90 56 4000 2300
V 330 290 38 23 40 33 48 27. 540 210 49 98 5000 1500 94 55
Ti 15 33 8 9.6 0.44 370 290 210 5000 5000 560 1100 130 140 5000 3800
Be 0.34 0.24 0.25 0.15 0.26 0.39 0.57 8.1 2.3 1.2 5.9 0.59 2.9 1.8 2.4 14
-VALUES IIj. PPH WEIGHT. EXCEPT AS NOTED
-------
Table 3. Carcinogenic and structurally similar polycyclic organic constituents
in particulate samples as determined by gas chromatographic mass spectrometrya
(,.)
o
o
SOURCE OPEN HEARTH COKE OVEN BASIC OXYGEN IRON OIL FIRED CERAMICS COPPER ALUMINUM SLUDGE
FURNAC E HEATER FURNACE SINTERING POWER PLANT PLANT SMELTER REFINERY INCINERATOR
PARTICLE SIZE 1-3" 3-10" 1-3" 3-10" 1-3" 3-10" 1-3" 3-10" 1-3" 3-10" 1-3" 3-10" 1-3" 3-10" 1-3" 3-10" 1-3" 3-10).
lUENTIFIED COMPONENT
Anlhr acene/Phenanthrene Z.7 0.4 3.9 b 24.8 6.9 289.5 b b 22 0.5 24.7 172.6
Methyl Bnthracenes 27.4 9 51.3
Fluoranthene 1.3 0.2 7.4 1.4 31. 2 6.2 34.1 137.2
Pyrene 0.4 2.9 1 16.1 4.1 42.9 155.8
Methyl Pyrene/Fluoranthene 18.0 3.9 32.5
Chry&ene! Benz (a)anchracene 0.4 1.9 93.6 287.3
Methyl chrysenes 25 79.7
Benzo f luoranthenes 1.6 0.5 1.6 1.8 5.8 248 786.3
Benz (a)pyrene 0.9 0.3 4.3 5.8 46.5 405.6
Benz(e)pyrene 3 2 1.8 9.0 13.8 807.2 1014
3-Hethy lcholanthrene 59.5 23.5
lndeno (1.2.3 .-<.:d)pyrene 1.1 lOG 3.6 341. 4 249.2
Henzo (ghi) pery lene 0.4 0.5 0.8 426 786.3
IHbenzo (a. h) anthracene 4.3 4.5 1387 1213.3
Dibenzo (c .g)carbazole 84.7 77.6
Uibenz (a1 and ah)pyrenes 1.5 352 216.1
Co ronene 87.3 50.2
dreported as parts per million by weight
boOt analyzed
-------
A I OIL FIRED POWER PLANT (1-3~)
A I ALUMINUM PLANT (1-3~)
A I COKE OVEN HEATER (3-10~)
A I COPPER SMELTER (3-1O~)
A I COPPER SMELTER (1-3~)
A I SLUDGE INCINERATOR (1-3~)
A I IRON SINTERING (1-3~)
w A I OPEN HEARTH (3-10~)
--..J
£:L. A I ALUMINUM PLANT (3-1O~)
:E:
c::( I OPEN HEARTH (1-3~)
C/) A
A I COKE OVEN HEATER (1-3~)
A I CERAMICS PLANT (1~3~)
(.oJ ~ IRON SINTERING (3-10~)
0
-
A I SLUDGE INCINERATOR (3-1O~)
~ BASIC OXYGEN FURNACE (3-10~)
~ CERAMICS PLANT (3-10~)
~ BASIC OXYGEN FURNACE (1-3~)
. " ' I ' I
10 20 30 40 50 60 70 80 90 100
PERCENT VIABILITY (lOOO~G/ML)
Figure 3. Viability of rabbit alveolar macrophages after 20 hours/particles
plus supernatant (data reported as percent viability :t 1 standard deviation;
data include screening and preliminary concentration-response trials at the
1,000 JiG/M L particle concentration).
-------
SLUDGE INCINERATOR (3-10">
100 ::::'-~0 S ~F~'''=
--.... . ~F P+S
P ~._---
80
....
ffi
u
~ 60
>-
....
::;
;;:;
""
:;: 40
:::j
~ KEY
0' S
20 GJ= P
(;). P+S
/ , " ,
/ 100 300 500 1000
CONCENTRATION (.G PARTICLES IML MEDIUM)
SLUDGE INCINERATOR (1-3,)
100
--_n" <;1
~
S <;1--m-
-;: 80
z
'"
~
..
~ 60
::;
~
: 40
~
....-- ~
...._-':~--
~
~r-
p+~l:t.
P -l~:::::
KEY
20
0'S
GJ" P
(;)" P+S
I
/
,
100
,
300
CONCENTRATION <.G PARTICLES IML MEDIUM)
500
1000
Figure 4. Viability of rabbit alveolar macrophages exposed
to various particulate fractions
302
-------
ordering of samples from industrial sources in
terms of their acute cytotoxic behavior. Addi-
tional functional and biochemical measure-
ments can be made to enhance the resolution of
the bioassay system.
It should be clearly understood that this has been a
preliminary report, the purpose of which was to assess
selected testing methods of characterizing the potential
hazards associated with particulate emissions. Refine-
ment of these methods, including assessment of sensi-
tivity, specificity, and external validity, is' of current
concern. A comprehensive report on these activities will
be available in the first quarter of 1976.
REFERENCES
1.
R. M. Statnick and L. D. Johnson, "Measurement
Programs for Environmental Assessment," presented
at the EPA Symposium, Environmental Aspects of
Fuel Conversion Technology, II, Hollywood,
Florida, Dec. 15-18, 1975.
C. H. Lochmuller, "Analytical Techniques for
Sample Characterization," presented at the EPA
2.
303
3.
Symposium, Environmental Aspects of Fuel Conver-
sion Technology, II, Hollywood, Florida, Dec.
15-18, 1975.
TRW Systems Group, "Fabrication and Calibration
of a Series Cyclone Sampling Train" (Interim
Report), April, 1975.
F. Gletten, V. Weekes, and D. J. Brusick, "In Vitro
Activation of Chemical Mutagens. I. Development
of an In Vitro Mutagenicity Assay Using Liver
Enzymes for Activation of Dimethylnitrosamine,"
Mutation Res., 28: 113, 1973.
B. W. Ames, W. E. Durston, E. Yamaski, and F. D.
Lee, "Carcinogens are Mutagens: A Simple Test
System Combining Liver Homogenates for Activa-
tion and Bacteria for Detection," Proc. Nat. Acad.
Sci. USA, 70:2281, 1973.
M. D. Waters, D. E. Gardner, C. Aranyi, and D. L.
Coffin, "Metal Toxicity for Rabbit Alveolar Macro-
phages In Vitro," Environ. Res., 9:32-47,1975.
M. D. Waters, D. E. Gardner, and D. L. Coffin,
"Cytotoxic Effects of Vanadium on Rabbit Alveolar
Macrophages In Vitro," Tox. Appl. Pharm.,
28(2):253-263,1974.
4.
5.
6.
7.
-------
18 December 1975
Session V:
ENVIRONMENTAL PROBLEM DEFINITION, PART B
Robert P. Hangebrauck
Session Chairman
305
-------
WATER REQUIREMENTS FOR AN
INTEGRATED SNG PLANT AND MINE OPERATION
D. J. Goldstein and R. F. Probstein*
Abstract
Part of the environmental assessment of large plants
to make synthetic natural gas from coal is the determi-
nation of the water consumed. This is particularly
important in the West where coal is available but water is
scarce. The water consumed includes not only water for
the conversion process but also water evaporated for
cooling and consumed in mining, land reclamation, and
solids disposal. Published information varies by more
than fourfold. In this paper, details of the procedures for
determining water requirements are given. The determi-
nation of the cost of not evaporating water for cooling
but of using air cooling and condensing is also described.
It is shown that water requirements are dependent on
process design, mine location, and climate and that
generalized assessments which are not site specific and
design specific are of limited value. It is also shown that
the published water requirements for integrated SNG
plants and mine operations in the West may be high and
that the actual requirements could, depending on the
location, be half the lowest estimate to date.
INTRODUCTION
An important part of the environmental assessment
of large plants to make synthetic natural gas from coal is
the determination of the quantity of water consumed.
This is particularly important in the West where coal is
available but water is scarce. As examples of the many
studies which have required this information see refs.
1-3. These studies conclude that, by about 1980, water
may be the resource which limits the development of
coal conversion technology.
For a plant making 250 x 106 set/day of SNG,
Davis and Wood of the United States Geological Survey
(USGS) have summarized the water requirements to be
between 8.8 x 106 and 38 x 106 gal/day (ref. 4). The
lower end of this range of water consumption comes
from one of two designs for Lurgi process plants to
make SNG from Navajo coal in the Four Corners area
*The authors are partners in Water Purification Associates,
238 Main St., Cambridge. Mass. 02142. R. F. Probstein is also a
Professor of Mechanical Engineering at the Massachusetts I nsti-
tute of Technology.
near Farmington, N.M. (refs. 5-8). From refs. 5-8 we
have prepared table 1, in which it should be noted that
the figures for the EI Paso plant have been scaled from
288 to 250 million set/day. It is important to remember
that table 1 and all the numerical answers given in this
paper are illustrative examples. Even for detailed designs,
consumed water estimates between different process de-
signs at the same location cannot be expected to agree to
within more than 20% at this stage of development. The
water requirements shown in table 1 have been divided
into four categories under which detailed discussions will
be given in this paper; namely (i) process, (ii) cooling,
(iii) mining and reclamation, (iv) evaporation, solids dis-
posal and other uses.
The upper end of the range of water consumption
comes from the table in ref. 9 which is also reproduced
in refs. 4 and 10. With a change of units, the table of ref.
9 is reproduced in table 2. The plant size used in table 2
(250 x 109 Btu/day) is within 6% of the plant size used
in table 1 (250 x 106 scf/day). Tables 1 and 2 are in
disagreement. One important reason for this is the assess-
ment of the water required for cooling.
Hittman Associates in a recent report (ref. 11) find
that 55-66 x 106 gal/day of water are evaporated for
cooling in a coal refinery consuming 1.8 x 1012 Btu/day
of coal. For a plant consuming 3.5 x 1011 Btu/day of
coal (as do the SNG plants used as examples in this
paper). this comes to 10.7-12.8 x 106 gal/day. Assuming
that about 75% of the makeup water is evaporated and
25% is blown down, the makeup rates would then be
approximately 14-17 x 106 gal/day. These values are
roughly in agreement with the figures of table 2 for a
cooling water makeup rate of 3% of the circulating
water, which in turn is consistent with a 25% blowdown
rate if a cooling water range of 25° F is assumed.
In (ref. 12) cooling water makeup requirements of
20-35 x 106 gal/day are suggested, based on a waste heat
calculation. The plant is taken to be 60"10 efficient so
that 160 x 109 Btu/day are lost by evaporating water.
This requires about 16.7 x 106 gallons/day evaporated,
assuming 1,150 Btu is needed to evaporate a pound of
water. The makeup requirements refer to situations
where approximately 3% to 5% of the circulating water
is evaporated. Here again, we have a number in approx-
imate agreement with table 2 and in disagreement with
table 1.
307
-------
Table 1. Approximate water requirements to make 250 x 106 set/day of SNG
by the Lurgi process in New Mexico (refs. 5-8)
Quantities in 106 gal/day:
EI Paso
Wesco
Input
Output
Net
Consumption
Input
Output
Net
Consumption
Process, including
fuel gas generation 5.3 4.4 0.9 5.7 4.9 0.8
Cooling 4.1 0.4 3.7 3.2 0.7 2.5
Mining, reclamation 1.9 1.9
Evaporation,
solids disposal,
other uses 2.3 2.1
TOTAL 8.8 7.3
Table 2. Water requirements for 250 x 109 8tu/day SNG plant (ref. 9)
Quantities in 106 gal/day:
Bituminous & Subbituminous Lign ite
Water Cooled
Cooling makeup
(% circulating water) 3 5 7 3 5 7
Process water 2.5 2.5 2.5 2.5 2.5 2.5
Boiler makeup 0.6 0.6 0.6 0.5 0.5 0.5
Cooling makeup 17.4. 29.0 40.7 14.5 24.2 33.9
TOTAL 20.5 32.1 43.8 17.5 27.2 36.9
Partially Air Cooled
Process water +
boiler makeup
(from above) 3.1 3.1 3.1 3.0 3.0 3.0
Cooling makeup
(by subtraction
from total) 7.1 13.0 18.8 5.8 10.6 15.5
TOTAL (ref. 9) 10.2 16.1 21.9 8.8 13.6 18.5
308
-------
So far we have been discussing cooling water. Water
consumed in (a) mining and reclamation and (b) evapo-
ration, solids disposal, and other uses is stated to be
another large need. For example, in table 1 these catego-
ries are seen to represent about 50% of the total water
requirement for both designs. However, they are some-
times not included in assessments of water consumption
and when included are often not detailed or explained.
It is one of the theses of this paper that the water re-
quirements for these categories are very much site de-
pendent. Moreover, the published requirements cited for
them may be high and that in fact the actual water need
may be much more modest.
Water for the process is the smallest quantity but is
notable for the very high quality influent need (which is
boiler feed water) and very dirty effluent process con-
densate (ref. 13).
The present paper is a progress report of our on-
going study and gives the detailed procedures necessary
for determining the quantities of water consumed. From
these details it will become apparent why such widely
differing quantities have been recorded and, for specific
processes at specific locations, it will be possible from
the procedures given to more precisely define the water
quantities required for the process, for cooling, for
mining, for disposal, and for other uses.
PROCESS WATER
Figure 1 shows a simplified block diagram of the
important sections of an SNG plant employing oxygen,
along with the principal process influent and effluent
streams. The stream numbers in figure 1 are the same as
those in table 3, in which illustrative examples of process
influent and effluent water quantities are given together
with the references from which they were taken or de-
rived. Each example has been verified by a complete
hydrogen balance. We emphasize again that these are
specific examples and that repeatability to within 20% is
all that can be expected between different process de-
signs. I n considering the Lurgi process, note that tables 1
and 3 differ because fuel gas generation is not included
in table 3 for any of the processes.
The influent streams are mostly of boiler feed qual-
ity. Of the effluent streams, the gasifier off-gas conden-
sate stream 5 may be very dirty. The stream can be as
low as 95% water with the rest mostly ammonia, carbon
dioxide, phenol, and other organic material. The conden-
sate after the shift converter, stream 6, is dirty but not
as dirty as stream 5. Stream 7 depends on the gas purifi-
cation process. Water made by the methanation reaction,
stream 8, is high quality water. The quantity of meth-
anation water is dependent on the fraction of the prod-
uct methane made in the gasifier as is shown in figure 2
which shows the four processes listed in table 3 along
with several other processes.
COOLING WATER USES
As was shown in the Introduction, the published
literature contains a fourfold range in the quantity of
water required for cooling. I n table 4 is given a typical
set of thermal data for an SNG plant. This is not for any
particular plant but is given so that numerical examples
can be provided for this discussion. The assumption that
all of the unrecovered heat ends up evaporating water
leads to the upper limit of cooling water requirement.
To determine whether this is necessary or desirable, we
must investigate the points of cooling individually.
I n table 5 is shown a partial summary of the relative
uses of circulating cooling water for two Lurgi process
plants. When water is limited, it is clear that the choice
of where to use wet cooling is a matter of engineering
discretion. The major sections of an SNG plant using
oxygen are shown in figure 1, and our study of numer-
ous detailed plant designs has shown in which of these
sections heat is not recovered and in which sections it is
either all or mostly recovered (see table 6). I n what
follows, we will first discuss each section to determine
the quantity and level of heat that must be disposed of.
Once this has been set forth, we will then point out the
considerations necessary to evaluate whether evapora-
tion of water for cooling is necessary or desirable. The
actual choice will usually be specific to a process a de-
sign and a climate. '
COOLING POINTS AND LOADS
Gas Purification
The energy consumed and the cooling required in
gas purification is dependent on the process used. As an
example of a chemical absorption process we have con-
sidered the hot potassium carbonate process. In our
example, 13 x 106 scf/hr of CO2 are to be absorbed. If
the liquid absorbs 3.5 scf/gal with an absorber at 1000
psi, the circulation pumps require 4.5 x 104 hp. About
half of this power can be supplied from a liquid expan-
sion recovery turbine and about 0.15 x 109 Btu/hr in
steam is consumed by the pump drive, or about 3.3% of
the unrecovered heat in the example of table 4. The
major energy consumption is in the stripper. The heating
load, which is approximately the cooling load, is about
100,000 Btu/1000 scf CO2 (approximately 10 scf
CO2/lb steam) (ref. 19).* This load is quite low and
* During the conference, ref. 28 was brought to the atten-
tion of the authors, from which a figure of 80,000 Btu/1,OOO scf
C02 would appear to be more accurate for the high pressures
used.
309
-------
STEAM
BOILER
coal
DRYING
ELECTRICAL
GENERATION
r
I
~
~ EXYGEN
r PLANT
I
t
process
points
COAL
FEEDING
wash water or
steam
4
GAS
PURIFICATION
METHANATION
y
condensate
@
..
I condensa te
I CD
L -1 SULFUR I
PLANT
stearn or
water stearn or
stearn G) water
cp I 3
SCRUBBING & SHIFT
GASIFIER QUENCHING CONVERTER
.
condensate condensate
@ @
COOLING &
DRYING
SNG
Figure 1. Major sections of an SNG plant
310
-------
Table 3. Illustrative examples of process water streams for the production
9f 250 x 106 sef /day of SNG
Quantities in 106 gal/day:
I nfluent Stream Effluent Stream
Process 2 3 4 Total 5 6 7 8 Total
Lurgi
(ref. 5) 4.5 0.2 4.7 2.9 0.2 0.8 3.9
Bigas
(ref. 14) 1.2 3.6 0.3 5.1 2.5 0.4 0.7 3.6
Synthane
(refs. 15, 16) 5.0 0.9 5.9 3.3 1.1 0.4 4.8
Hygas/Oxygen
(ref. 17) 2.1 3.0 5.1 2.4 0.8 0.2 3.4
applies to a system with a high pressure absorber, such as
the one considered; therefore, in our example, 1.3 x 109
Btu/hr are consumed as low pressure steam or about
29% of the unrecovered heat in the example of table 4.
This hea! is removed in the stripping column condenser.
Off-Gas Quenching and Scrubbing
In some plants, the gas leaving the gasifier must be
rapidly quenched and scrubbed free of oils and tars. This
is done, for example, in the Synthane process. Because
the tar is sticky when hot, waste heat recovery is not
possible. We fi nd that the quench removes about 1.1 x
109 Btu/hr (24% of the unrecovered heat in our ex-
ample) of which about two-thirds is for condensing
steam and one-third is for cooling ,the gas.
Boiler Steam Generation
Coal, in addition to the quantity consumed in the
gasifier, is consumed in the power plant. (In the Syn-
thane process, all the coal is passed through the gasifier
and char is burnt in the power plant.) Power is needed to
generate electricity, to drive the gas compressors for
oxygen production and the big pumps in the gas purifi-
cation stage. I n the simplest case the gas compressors,
which are a big consumer of energy, are driven by steam
turbines and the steam is condensed as it leaves the tur-
bine. This is an important case for the study of cooling
and it is discussed in detail below. However, we must
emphasize that this is not the only possibility.
If the coal is of so Iowa sulfur content that stack
gas scrubbing is not required, production of steam by
burning coal is probably the best route. It, however,
stack gas scrubbing is required, gasification of the coal to
produce a low-Btu sulfur-free gas should also be con-
sidered. If a low-Btu gas is made, gas turbines will prob-
ably also be involved. Even if steam turbines are the only
motors used, they are not necessarily condensing tur-
bines.
In our example the extra coal or char burnt for
power is 3.2 x 109 Btu/hr. In the particular case where
this is all burnt to produce steam in a boiler, about 12"10
of the energy goes up the stack and requires no cooling.
This amounts to 0.4 x 109 Btu/hr, or 9% of the unre-
covered heat in our example.
Oxygen Production
The important details of an oxygen plant are shown
in figure 3. Air is compressed to about 90 psia and cool-
ed to about 90° F. In this condition it enters the separa-
tion plant which operates with only a small energy and
cooling consumption to give oxygen at about 20 psia
and about 95° F. The oxygen must then be compressed
to gasifier pressure which we have taken to be 1,015
psia. For a discussion on energy and cooling, oxygen
production can be treated as a series of compressors.
--. . - -- -~----
When a gas is compressed, it heats up and must be
cooled before it can be compressed further-that is, cool-
ers are needed between the stages of a compressor. Also,
as we are here assuming condensing steam turbine drives
for the compressors, condensers are required so that the
exhaust steam can be condensed and returned to the
boiler. The condensers and the interstage coolers were
considered separately for both the air compressor and
the oxygen compressor. With the air compressor han-
dling 3.46 x 104 Ib/min and the oxygen compressor han-
dling 8~06 x 103 Ib/min, the cooling loads are 0.2 x 109
Btu/hr in the air compressor interstage coolers, 0.1 x
311
-------
1.0
>. 0.9
n:1 MOLTEN
'd SALT
.........
.-I
n:1 0.8
b>
u:>
0
.-I BIGAS
0.7 .
z
0
H
1:-1
u 0.6
::>
Q
g SYNTHANE
il<
0.5 .
r:.::
>:4
1:-1
.::x:
3: 0.4
Z
w 0
... H
ro.J 1:-1
.::x: .
z 0.3 HYGAS
.::x:
::r: ELECTROTHERMAL
1:-1
~ 0.2
.
0.1
0
0.4 0.5 0.6 0.7 0.8 0.9 1.0
CH4 PRODUCED IN GASIFIER
CH4 IN PRODUCT PIPELINE GAS
Figure 2. Quantity of methanation water as a function of product methane produced
in gasifier for SNG plants producing 250 x 106 scf/day.
-------
Table 4. Typical thermal balance of an SNG
plant producing 250 x 106 scf/day
Heating Value
Material 10. Btu/hr % of Feed
Feed coal 15.0 100
Product gas (9.8) (65)
Byproducts (0.7) (5)
Unrecovered heat. Net 4.5 30
.The anrecovered heat if used to evaporate water in II
cooling tower, will evaporate about 11.8 x 10. gal/day
in summer (assuming 1100 Btu/lb water evaporated) or
about 10.3 x 10' gal/day in winter (assuming 1250 Btu/
Ib water evaporated). Makeup water quantities will be
higher than the evaporated water.
Table 5. Principal uses of cooling water in
Lurgi process plants
Percent of Total Circulated Water
Section (ref. 5)
Refrigeration 35
Gas Liquor Separation 17
Gas Cooling 11
Sulfur Recovery 9
Phenol Extraction 8
Air Separation 8
Gas Purification
TOTAL 90%
(ref. 18)
9
11
61
10
91%
Table 6. Disposition of unrecovered heat in SNG plants
Sections Where Heat Is Unrecovered
Gas Purification
Scrubbing and Quenching of Gasifier Off-Gas
Steam Boilers
Oxygen Production and Compression
Electrical Generation and Use.
Coal Drying
Coal Feeding
Product Cooling and Drying
Sections With Little Unrecovered Heatt
Gasifier
Methanator
Shift Converter
Sulfur Plant
.Unrecovered electrical energy.
tHeat recovery is practiced in most.of these sections.
109 Btu/hr in the oxygen compressor interstage coolers
and 0.6 x 109 Btu/hr in the condensers. The total is 0.9
x 109 Btu/hr which is 20% of the unrecovered heat in
oUJ example.
Electric Power Generation
If steam turbines are used for generating electrical
power, then for estimating purposes a heat rate of
10,000 Btu/kwh may be used-that is, an efficiency of
about 34%. The amount of electricity used depends very
much on the designer. For an example we may consider
18,000 kw to be a middle-to-!ow requirement. To gen-
erate this quantity of electricity, 0.18 x 109 Btu/hr are
consumed of which about one-third or 0.06 x 109
Btu/hr (1.3% of the example's unrecovered heat) is con-
verted to electricity and consumed without cooling and
about two-thirds or 0.12 x 109 Btu/hr (2.6% of the
unrecovered heat) is lost in the turbine condensers.
Coal Drying
The coal fed to many gasifiers must be dry. The
energy consumed to dry coal depends on the moisture in
the coal. An extreme case is exemplified by a Montana
lignite of 300fo moisture and 8,700 Btu/lb HHV. To feed
15 x 109 Btu/hr to the gasifier, about 0.52 x 106 Ib/hr
of water must be evaporated consuming about 0.5 x 109
Btu/hr (11% of the unrecovered heat in our example).
Cooling water is not required because if it is used, nearly
313
-------
DIrO C>t<) TRIM WATER
COOLER
90 psia
90°F
-----
oxygen stages of
SEPARATION 20 psia compression
9SoF
PLANT not shown
air
to gasifier
1015 psia
Figure 3. An oxygen plant shown with air coolers.
the same amount of quite clean water would be recover-
ed. We are presently looking into ways of recovering this
water using air cooling. For the case considered, were all
of the evaporated water recovered, this would amount to
1.3 x 106 gal/day.
Coal Feeding
We have not completed our study of the energy
consumed in feeding coal to 1000 psig. If lock hoppers
are used, the energy is consumed by the lock hopper gas
compressors and the need for cooling water is similar to
the discussion on oxygen preparation and electrical gen-
eration. If a slurry in toluene is used, the cooling load is
that required to condense the toluene in the gasifier off-
gas. If a slurry in water is used, it is probable that this
water should not be evaporated in the gasifier but should
be evaporated before the coal enters the gasifier at 1000
psig-for example, by the hot off-gas from the Bigas gasi-
fier (ref. 20).
Product Cooling and Drying
The SNG product must be cooled and dried. It is
necessary to cool to 900 F and not to 1400 F because
the size of a glycol dehydrator to accept gas saturated at
1000 psig and 900 F must be multiplied by four if it is
to accept gas. saturated at 1000 psig and 1400 F (ref.
41). For purposes of illustration, we have assumed that
the gas stream is 250 x 106 scf/day of methane and that
it is cooled to 3500 F by heat exchange with incoming
gas or by waste heat recovery. The total load to cool this
gas down to 900 F is about 0.14 x 109 Btu/hr or about
3% of the unrecovered heat in our example.
Summary of Cooling Points and Loads
In tabie 7, we have summarized the heat dissipation
rates required at the cooling points in our illustrative
SNG plant. This unrecovered heat load is given in both
Btu/hr and as a percent of the total load. The load per-
centages differ slightly from the percentages cited
throughout the text because the total calculated load of
4.7 x 109 Btu/hr is slightly larger than the value of 4.5 x
109 Btu/hr given in the "typical thermal balance" of
table 4.
For the particular example given, it is evident from
table 7 that there are three principal points of cooling
load; gas purification, off-gas scrubbing and quenching,
and oxygen production. Another important point to be
observed in this regard is that in the example chosen, it
is not necessary to supply cooling for 20% of the total
"unrecovered heat." Of the remaining 80% of the heat
to be dissipated, the three cooling points cited account
for about 93% of the cooling which must be supplied.
In the following section, we will attempt to estab.
lish the economic and design criteria for determining
whether wet, dry, or wet/dry cooling is the desired
approach. As previously noted, the choice will usually be
site, process, and design specific. We shall only attempt
314
-------
Table 7. Summary of examples of cooling points and loads.
Point
Gas Purification (hot potassium carbonate)
Regenerator condenser
Pump drive
Off-gas scrubbing & quenching
Boiler steam generation
Oxygen production
Air compressor intercoolers
Oxygen compressor intercoolers
Turbine condensers
Electric power generation
Turbine condensers
Electricity
Coal drying (lignite)
Product cooling & drying
TOTAL
Load
Cooling
Needed
(10. Btu/hr)
(%)
1.3
0.15
28.0
3.2
Yes
Yes
1.1
23.5
Yes
0.4
8.5
No
0.2 ~
0.1
0.6
19.2
Yes
0.12
0.06
2.6 Yes
1.3 No
10.7 No
3.0 Yes
100
0.5
0.14
4.67
to outline those calculations and concepts which are
important to arriving at a decision.
COOLING METHOD CHOICE
We have investigated the major heat loads shown in
table 7 which require cooling, to determine under what
conditions dry cooling, wet cooling, or a combination
should be used. The formulae and data used in the
calculations are given in the appendix. I n each case we
have determined the cost of water which makes the total
annual operating cost equal for wet and dry cooling. The
procedure can best be exemplified by detailing the case
of gas purification, which is the simplest.
Gas Purification
The condenser in the stripping column of a hot
potassium carbonate system must remove 1.3 x 109
Btu/hr and can be dry or wet. The vapors were assumed
to enter at 2300 F and to be cooled to 1400 F. Using the
information from the appendix, the areas of the con-
densers were found to be 2.5 x 105 ft2 dry or 1.3 x 105
ft2 wet and the circulation water rate to be 105 gp-m.
The cost estimate is given in table 8. In this example, the
choice is between supplying and paying for capital or
supplying and paying for water. It should be noted that
the cost of water in table 8 and all other tables is for
water evaporated and includes the cost of supply, as well
as the cost of treating the circulating cooling water and
disposing of the blowdown. In the example shown here,
dry cooling should probably be used.
Off-Gas Quenching and Scrubbing
The basic system studied for off-gas quenching is
shown in figure 4. Circulating quench water is heated
from TV F to 2700 F in the scrubber and is cooled back
to TO F in the cooler, with 1.1 x 109 Btu/hr transferred
to the cooling medium. The lower the temperature T,
the less the energy needed to circulate the water. We
have investigated dry cooling with T = 1300 F, wet cool-
ing with T = 900 F and with T = 1300 F, and dry cooling
followed by wet cooling with T = 900 F. In this case, the
energy needed to circulate the cooling water more than
offsets the larger capital cost of a dry cooler and even
with no charge for water, dry cooling is cheaper.
Oxygen Production
By far the most complex optimization of a cooling
system is that around the compressors. As part of our
ongoing project, we will be doing this optimization at
selected sites in the western United States. When, how-
ever, one is exploring possibilities with the aim of deter-
mining cooling water requirements to an accuracy of
315
-------
Table 8. Example of a cost estimate for a dry and wet cooled condenser
for a hot potassium carbonate gas purification system
Dry Wet
Capital Cost (106 $1
Heat transfer area 4.5 0.66
Cooling tower 0.70
TOTAL (106 $) 4.5 1.36
Installed Power (103 hp)
TOTAL (103hp) 4.4 4.2
Operating Cost (8000hr/yr) (105 $/yr)
Capital @ 15%/yr 6.7 2.0
Energy @ 2.24d/hp-hr 7.8 7.5
Water @ 46d/1 000 gal evaporated 5.0
TOTAL (105$/yr) 14.5 14.5
gas
2700p
summer time design conditions are dry interstage cooling
to 1400 F and wet interstage cooling to 900 F. The
maximum exit temperature from a compressor stage is
3000 F.
- Because the horsepower of a compressor is directly
proportional to the absolute temperature of the entering
gas, dry interstage cooling will involve more horsepower
and a bip' - capital investment. This is clearly shown in
tables 9 ~ 10. In w."ter, when the air and cooling
water are colder, the power consumed to compress the
gas by a controlled pressure ratio will decrease. When
using dry interstage coolers, we have the choice of alter-
ing the pitch of the fans to save the fan energy (this is
usually done automatically), or of obtaining a lower
interstage temperatu re and saving compressor energy. In
this example the compressor energy that will be saved is
more than the fan energy, so we assumed that the fans
draw full power all year.
When using wet interstage coolers, we will circulate
the water at full rate all year to avoid fouling the heat
exchangers. However, in the winter, we may operate at
one of two extreme conditions or somewhere inbetween.
We may run the cooling tower at full capacity. There
will be only a small change in the rate of water evapora-
ted, but there will be a large decrease in hot water and
cold water temperatures -and therefore in the gas temp-
erature entering the compressor stages. This will result in
a saving in compressor horsepower.
gas
TOp
C
R
U
B
B
E
R
oil & condensate
Figure 4. System for off-gas quenching.
about 25%, many of the calculations can be approxi-
mated and simplified. We will briefly outline some of the
possibilities indicating which have proved to be impor-
tant. The compressors are to compress 3.46 x 104
Ib/min of air from 14.7 psi a to 90 psia and 8.06 x 103
Ib/min of oxygen from 20 psia to 1,015 psia. The
316
-------
Table 9. Example of air compressor interstage cooler calculations for
dry and wet cooling
Number of Stages
Summer Design Conditions
Compressor hp
Total dry cooling area (ft2)
Total wet cooling area (ft2)
Circulated cooling water (gpm)
Capital Cost
Compressors @ $135/hp
Dry cooling area @ $18/ft2
Wet cooling area'~ $5.'1 ttr
Cooling tower @ $7/gpm
TOTAL (106 $)
Auxiliary Power
Air cooler fans @ 0.015 hp/ft'
Cooling tower fans @ 0.012 hp/gpm
Circulation pumps @ 0.03 hp!gpm
TOTAL (10' hp)
Operating Cost (8000 hr/vr)
Capital cost@ 15%/vr
Annual avg. compressor power @ 2.24d/hp-hr
Auxiliary power @ 2.24d/hp-hr
Water @ $9.4/1000 gal evaporated
TOTAL (106 $/vrl
Dry Wet
3 3
86,170 81,260
1.62 x 10'
3.80 X 10' 2.43 X 105
1.82 x 10' 1 .52 x 10'
(106 $)
11.63 10.97
2.92
0.19 1.24
0.01 0.11
14.75 12.32
(10' hpi
2.43
0.02 0.18
0.05 0.46
-
2.50 0.64
( 1 06 $/Vr)
2.2 1.8
15.1 14.3
0.4 0.1
0.2 1.7
17.9 17.9
Alternatively we can hold the hot water and cold
water temperatures, and so hold the compressor power
constant, and bypass the cooling tower-saving water
and fan energy. Using, as an example, a climate approxi-
mating that of New Mexico, we find that it is best to run
the cooling towers at full capacity all year, to save com-
pressor energy and not to save water and cooling tower
fan energy.
The results of the interstage cooler sample calcula-
tions are summarized in tables 9 and 10. The cost of
cooling water is offset by the cost of capital (for the
cooling system and for the compressors) plus the cost of
energy to drive the compressors. An annual average com-
pressor power has been used in the approximate cost
comparisons and because this is so important, the break-
even water cost will be dependent on climate. In our
examples, wet cooling should probably be used if water
is available.
in considering condensers for the steam turbine
drives, an additional complication becomes important;
the "heat rate," which is the rate of heat consumption
per horsepower output, varies with the condenser tem-
perature. We assumed 7,700 Btu/hp-hr below 1000 F,
7,800 Btu/hp-hr at 1200 F and 8,070 Btu/hp-hr at 1400
F. Because the heat rate is not Ii nearly dependent on
temperature we have, when using a dry condenser, run
the fans all year-round and saved compressor energy;
but, when using a wet condenser, we have bypassed the
tower in winter and saved water and fan energy. If the
tower is bypassed for the condensers and not bypassed
for the interstage coolers, two separate cooling tower
basins and circulating systems must be installed and
maintained.
The approximate cost calculation is shown in table
11. The added compressor energy is due to the change in
heat rate and is dependent on the climate; so, therefore,
317
-------
Table 10. Example of oxygen compressor interstage cooler calculations
for dry and wet cooling
Dry Wet
Number of Stages 7 5
Summer Design Conditions
Compressor hp 39,710 38,455
Total cooling area (ft') 3.89 x 10' 4.99 X 10'
Circulated cooling water (gpm) 7.26 x 103
Capital Cost (10. $)
Compressors@ $135!hp 5.36 5.19
Cooling area (see Appendix) 0.70 0.26
Cooling tower @ $7!gpm 0.05
TOTAL (106 $) 6.06 5.50
Auxiliary Power hp
Air cooler fans @ 0.015 hp!ft2 584
Cooling tower fans @ 0.012 hp!gpm 87
Circulation pumps @'Q.03 hp/gpm 218
-
TOT AL 1111'1 584 305
Operating Cost (8000 hr!yd (10. $!yd
Capital @ 15%!yr 0.91 0.83
Annual avg. compressor power @ 2.24i!hp-hr 6.99 6.83
Auxiliary power @ 2.24d!hp.hr 0.10 0.05
Water@ $3.3/1000 gal evaporated 0.29
TOTAL (10. $!yr) 8.00 8.00
is the breakeven water cost. Wet cooling should probably
be used if water is available.
Throughout our calculations so far we have com-
pared wet coolers to dry coolers. I n fact, dry inter-
coolers followed by wet trim coolers are accepted prac-
tice. The optimization of these designs is quite complex
as may be seen in ref. 21 where it is shown that an
optimum pressure drop and heat transfer coefficient
exist for the air part of the interstage cooling. In north-
ern climates, a dry condenser can condense at temper-
atures below 100° F for many months in winter obviat-
ing the need for cooling water. Larinoff and Forster (ref.
22) have discussed the various ways of applying dry/wet
cooling to turbine condensers (e.g., dual surface dry/wet
condensers ~ith a wet cooling tower or a wet condenser
with a dry/wet tower in series or parallel arrangement).
All of these possibilities represent a potential saving in
water at the cost of a higher investment in the cooling
system. We will be studying these possibilities.
Product Cooling and Drying
! n figure 5 is shown the cost of cooling a high pres-
sure hydrocarbon gas from 350° F to an exit tempera-
ture of TO F. The cost is expressed per unit of heat load
and, provided the heat transfer coefficients do not
change, it does not matter if the heat load includes a
condensing load. When cooling SNG, the condensibles
are such a small fraction of the stream that figure 5
applies. We have, therefore, chosen to cool the product
SNG by dry cooling to 140° F and then by wet cooling
to 90° F; 86% of the load is taken by the dry cooler.
Summary of Wet or Dry Cooling Choices and Evapo-
rated Water
Our suggestions for wet or dry cooling are shown in
table 12 which is an extension of table 7. Wet cooling
has been assumed for the condensers of all drive turbines
because the calculation is similar to that for the air and
oxygen compressors. The load on wet cooling is about
1.20 x 109 Btu/hr. About 0.95 x 109 Btu/hr is the load
318
-------
Table 11. Approximate cost comparison for dry and wet compressor turbine condensers
Units in 10' $/yr:
Dry Wet
Operating Cost
Capital @ 15%/yr 3.3 1.0
Fans & pump energy @ 2.24d/hp-hr 4.4 2.5
Added compresssor energy @ 3d/104 Btu 4.9
Water at $2.62/1000 gal evaporated 9.1
TOTAL (10' $/yr) 12.6 12.6
Air Compressor Drive
Oxygen Compressor Drive
Dry
Wet
1.6
0.5
2.1
1.2
2.3
6.0
4.3
6.0
on turbine condensers and wet cooling of a process
stream following dry cooling. If, as we have suggested,
this type of cooling load requires a constant cold water
temperature all year round with the tower bypassed in
winter, the rate of evaporation of water will vary with
the climate from site to site. We have made preliminary
calculations using the climates of Farmington, New
Mexico; Casper, Wyoming, and Bismarck, North Dakota.
The annual average rate of heat removal is about 1,100
Btu/lb water evaporated in New Mexico and 1,300
Btu/lb in the more northern locations. This applies to
that portion of the cooling load which permits tower
bypassing. For estimating purposes then, about 3.1 x
106 gal/day of water would be evaporated for cooling in
the Four Corners area of New Mexico, while about 2.6 x
106 gal/day would be evaporated in the lower Northern
Great Plains region of Wyoming. As we have noted, these
water requirements are very reasonable estimates and do
not necessarily represent minimum values.
WATER CONSUMED FOR OTHER
THAN PROCESS AND COOLING
The consumptive water requirements shown in table
1 for (a) mining and reclamation and (b) evaporation,
solids disposal, and other uses, together constitute about
50% of the total usage for both Lurgi SNG designs in
New Mexico. From these designs (refs. 5-8, 18, 23) we
list in table 13 the principal categories of consumed
water. The mining and reclamation group is essentially
independent of the process, while the other group is
somewhat dependent on the process. Both groups, how-
ever, depend strongly on the geographical location and
we may expect different answers for a New Mexico plant
when compared with one in the Great Plains.
I n the following sections, we will discuss separately
each principal category of water consumption and
attempt to provide a general methodology for calculat-
ing the amount of water consumed. When water is limit-
ed, its use or loss prevention becomes a matter of eco-
nomic tradeoff. The water quantities we will present will
be characterized as reasonable bounds, with no effort
made to provide as detailed cost estimating as was done
for cooling. Thus, for estimating road dust control re-
quirements, we shall not treat all roads as paved, but
neither shall we consider them all unpaved. Similarly,
when considering evaporation suppression in ponds, we
shall not consider unusual pond depths nor expensive
chemical retardent procedures. However, where accept-
able procedures exist for conserving water as in sulfur
byproduct packaging, these will be noted. On the other
hand, where detailed capital cost tradeoffs must be made
against water costs, as in boiler tube cleaning of coal-
fired boilers, this will also be pointed out. Detailed cost
tradeoffs are now being carried out to determine
whether water use is necessary or desirable for several
special and relatively small categories. These detailed cal-
culations will be reported on at a later date as part of
our ongoing program.
MINING AND RECLAMATION
WATER REQUIREMENTS
In this section, we will discuss the principal con-
sumptive water requirements for mining and land rec-
lamation associated with an SNG plant producing 250 x
106 scf/day. By way of example, we will consider a
319
-------
.3 I
DRY COOLING I
--- WET COOLING I
.25
I
I
.2 I
I
==' 1
+I
I!j
\.D /1
0 .15
r-i
"
-u)- WATER, 50<::/1000 GAL ~ I
--
, -
+I
U)
0 /
()
.1 /
.-"'"
---
WATER, NO COST
.05
300
200
100
85
. 0
EXlt Temperature, F
Figure 5. The cost of cooling a high pressure gas stream.
320
-------
Table 12. Summary of the choice of wet or dry cooling and approximate amount of
evaporated water for an SNG plant producing 250 x 106 set/day in New Mexico and Wyoming
B rea keven
evaporated Wet Evaporated Water
Load water cost or N.M. WYO.
Point (109 Btu/hr) ($/10' gal) Dry (10' gal/day) (10' gal/day)
Gas purification (hot potassium
carbonate) Regenerator condenser 1.3 0.46 Dry
Pump drive 0.15 Wet 0.39 0.33
Off-gas scrubbing & quenching 1.1 zero Dry
Boiler steam generation 0.4 no cooling
Oxygen production
Air compressor intercoolers 0.2 9.4 Wet 0.53 0:45
Oxygen compressor intercoolers 0.1 3.3 Wet 0.26 0.22
Turbine condensers 0.6 2.6 Wet 1.57 1.33
Electric power generation
Turbine condensers 0.12 Wet 0.31 0.27
Electricity 0.02 no cooling
Coal drying (lignite) 0.5 no cooling
Product cooling & drying 0.14 zero 85% Dry
14% Wet 0.05 0.04
TOTALS 4.67 3.11 2.64
Table 13. Principal categories of consumed
water in SNG plants for other than process
and cooling
Mining and Reclamation
Road Dust Control
Coal Handling
Coal Washing
Revegetation
Evaporation, Solids Disposal
and Other Uses
Pond Evaporation
Ash Disposal
Sludge Disposal
Other Uses
plant in which steam is produced in coal-fired boilers.
Our emphasis is principally on the difference in water
needs between different sites. The absolute values of the
requirements given are to some extent a matter of en-
gineering discretion and dependent on a wide variety of
conditions. Nevertheless they can be considered to be
representative. The locations considered for our ex-
amples are the Four Corners area of New Mexico and
Campbell County, Wyoming, in the lower Powder River
Basin near Gillette, Wyoming. We emphasize that only
the principal consumptive requirements are discussed so
that, for example, domestic and industrial uses are not
considered here, although they will be in the next sec-
tion, because most of the water is reused.
Road Dust Contra/
Truck traffic on unpaved haul roads will generally
give rise to fugitive dust. The amount of fugitive dust
will depend principally on the miles of unpaved road,
the traffic density, and the dust raising capacity of the
vehicles, which increases with vehicle speed and size. It
will, of course, also depend on the soil type and the
weather in that more dust will be raised during dry
periods or when there is wind.
To a large extent, the length of the haul roads will
depend on the area's productivity, as measured by the
amount of coal or more exactly, heating value recover-
able, per unit area of stripped land. As stripping pro-
ceeds, the further out from the gasification plant or the
more haul spokes required to obtain the needed coal
input, the larger will be the number of miles of haul
321
-------
roads. It follows that the haul road lengths are site de-
pendent. Needless to say, the soil type and weather con-
ditions are also site dependent. I n table 14 are shown
typical strip-mined coal yields in tons/acre for the Four
Corners area of New Mexico and for the Campbell
County area in Wyoming, as estimated from the data of
refs. 23, 24. Also shown are the corresponding lengths of
45-foot-wide unpaved haul roads, as cited for the pro-
posed UII-Wesco 9.6 x 106 ton/year mining operation
(ref. 23), and as estimated for the Arco 10 x 106
ton/year Black Thunder Mine (ref. 24). There is no im-
plicit assumption here that the level of truck traffic is
the same to yield the same heating value output per
year.
Table 14. Typical coal yields and unpaved haul
road lengths for a 10 x 106 ton/year strip
mining operation in Four Corners, New
Mexico and Campbell County, Wyoming
Four Corners,
New Mexico
Campbell County,
Wyoming
Land yield of 8,500
Btu/lb coal !tons/acre)
T otallength of 450ft
wide unpaved haul
roads (miles)
37,000
125,000
8
2
One way to essentially eliminate this source of dust
is to pave the roads. This involves a cost and in our
ongoing program we shall undertake to compare this
cost with the costs for alternative methods of holding
down road dust. The principal alternative is to wet down
the haul roads, with or without the aid of chemical
evaporation retardents and soil stabilizing agents. Lack-
ing precise information on road dust suppression, we
may assume that the roads can be kept in a wetted con-
dition through an annual deposition rate of water on the
roads equal to the net annual evaporation rate. Any rain-
fall is taken to be an additional safety factor and is not
subtracted from the amount of water to be laid down,
because how much of it is actually absorbed in the road-
way and how much runs off is variable. I n table 15, we
have shown both the annual precipitation and evapo-
ration rates for the Four Corners area of New Mexico
and the Campbell County area of Wyoming. Knowing
the evaporation rate, we have evaluated the water re-
quirement for road dust control using the relation
lay down rate = evaporation rate x road length
x road width.
For Four Corners, we find a requirement of 147 gpm
and for Campell County, a requirement of 32 gpm. It is
of interest to compare these values, respectively, with
the estimated annual average of 270 gpm for the UII-
Wesco operation (ref. 27) and the estimated 36 gpm
requirement for the Arco Black Thunder Mine (ref. 25).
Table 15. Annual precipitation and evaporation
rates for Four Corners, New Mexico and Camp-
bell County, Wyoming (refs. 23, 25, 26)
Four Corners,
New Mexico
Campbell County,
Wyoming
Precipitation
(ll1j:hes/year)
Pond evaporation rate,
90% pan rate
(inches/year)
8
13
65
56
We would er ilasize that the requirements cited
can be reduced through the use of chemicals which pene-
trate the road soil and either stabilize it or retard evapo-
ration. In this regard we would also note that waste
blowdown salts might also be put on the road as a means
of retarding the evaporation rate. In any event, the
example values should be considered maximum ones for
the given road areas and that economies in this particular
water need can undoubtedly be made.
Coal Handling
In all coal preparation plants dust is generated in the
stages of unloading, conveying, crushing, general screen-
ing, and storage. The two main categories of dust collec-
tion systems employed are wet collectors and dry collec-
tors. For the operations described, dry collectors are
most commonly used. Although the amount of dust
generated will depend on the coal size required for the
process, this obviously can have no direct effect on
water needs, so long as dry collectors are employed. In
the Wesco Lurgi design, a fabric filter dry dust collection
system is planned, so that water will not be required for
dust control (ref. 23). Some small amount of water may
be needed to wet down storage piles, to prevent fines
from becoming airborne during dry periods of high
wind, although covered storage areas could minimize this
problem. To the extent that the weather depends on the
site there can be some variability in this water usage. In
322
-------
any case, the amounts involved would be small enough
so that for estimating purposes we need not allocate
water for this category of dust control.
A water requirement may be imposed by the fact
that different processes require different size coal. For
example, a Lurgi gasifier uses relatively large size coal
which requires only primary crushing and secondary
crushing and screening to yield a 1 W' x 0 product. On
the other hand, the Synthane and Bigas processes use a
200-mesh pulverized coal. Depending upon the process,
some water may be needed for preparing the coal for
pulverizing but no general requirements can be stated
and, in any case, they would not appear to be large.
Coal Washing
When steam is produced by coal-fired boilers, as in
the Wesco Lurgi plant (refs. 8, 23), it may be desirable
or necessary to upgrade the run-of-mine coal by reducing
the ash and sulfur levels. In the Wesco design, washing
pilot plant tests by Utah International Inc., using wet
concentrating tables, have led to upgraded properties
that are shown in table 16, where they are compared
with the properties of the run-of-mine Navajo coal. Also
shown for comparison is a representative composite of
run-of-mine Wyoming subbituminous coal from beds
near Glenrock and Gillette in the lower Powder River
Basin. Washing the Navajo coal is seen to decrease the
ash by about 52% and the sulfur about 24%, as measured
per unit of heating value. Stack gas cleanup for S02
removal would still be required to meet the emission
standards of 1.2 Ib S02/106 Btu.
A wet concentrating table is an easy-to-operate, in-
expensive and reliable coal washer. However, it is a rela-
tively large consumer of water. The throughput of water
for these devices can be taken to be about 12 gpm per
ton per hour of feed coal (ref. 30). Although the make-
up water rate is to some extent discretionary, we may
use the overall figure for coal preparation plants in New
Mexico of about 12%. This value is somewhat lower than
the national average of 18.5%, which however includes
such water surplus areas as Washington and Alaska, but
is higher than in the large water-consuming humid States
of Ohio and Pennsylvania where the comparable values
are only 8.3% and 10.2%, respectively. It is also higher
than for the semiarid mining regions of Montana where
it runs at 9.1%. Figures for the other Northern Great
Plains regions of interest are not available (ref. 30).
The amount of upgraded coal required for the steam
plant is about 3,760 tons/day (157 tph) which may be
compared with 21,860 tons/day of run-of-mine coal
needed for the gasifier (ref. 23). With about a 30"10 solids
concentration in the tailings from the concentrating
tables, we may estimate the sized raw coal fed into the
tables at about 175 tph. The consumptive water use is
then 175 tph x 'j 2 gpm/tph x 0.12 or about 250 gpm.
This may be compared with the average annual value of
450 gpm estimated by Utah International (ref. 27).
From table 16, comparison of the ash and sulfur
content of the washed Navajo coal with the run-of-mine
Wyoming coal shows clearly that if the same processes
were contemplated in the lower Northern Great Plains,
the coal would not have to be washed and there would
be no associated water requirement.
Revegetation
As part of any reclamation of strip-mined land in
arid and semiarid regions there exists a potential re-
quirement for supplemental irrigation water associated
with the establishment of soil stabilizing plant cover on
mine spoils. A report by the National Academy of
Sciences (ref. 26) on the rehabilitation of Western coal
lands concludes that areas with greater than 10 inches of
mean precipitation annually can be reclaimed without
supplemental irrigation, but that with less than 10 inches
Table 16. Properties of Navajo coal before and
after washing (ref. 23) and of a composite
Wyoming subbituminouscoal (ref. 29)
Ash Sulfur Higher Heating Value
Coal (wt%) (wt%) (Btu/lb)
Navajo, run-of-mine 26.0 0.96 8,310
Navajo, washed 14.1 0.87 9,870
Wyoming, run-of-mine 8.1 0.62 9,043
323
-------
supplemental irrigation is required. This conclusion may
be somewhat oversimplified in that factors other than
precipitation must undoubtedly enter, including the
evaporation rate, soil properties, seasonal precipitation
periods, land exposure, etc. (ref. 26).
Numerous studies have recently become available on
the techniques for reclaiming strip-mined land (refs.
31,32). Among the most recent efforts is an attempt to
revegetate mined land in the Four Corners region, where
the rainfall is less than 10 inches (see table 15), without
supplemental irrigation by using water harvesting tech-
niques. In any case, however, we can conclude that par-
tially reshaped mine spoils can be successfully revegeta-
ted in this region with supplemental irrigation of about
10 inches during the first growing season, with no
further requirement during subsequent growing seasons
(ref. 33).
For the Wesco example of a Lurgi plant in Four
Corners, as previously noted, the coal required would be
about 9.6 x 106 tons/year which, based on the yield
shown in table 14, would indicate a mined land area
requiring reclamation of about 260 acres annually. The
product of the required depth of irrigation water per
year and the area of land mined annually leads to a
supplemental irrigation requirement for our example of
135 gpm. This requirement is less than the 335 gpm
estimated by Utah International (ref. 27) where account
was taken of seedlings not taking hold the first season
and where higher initial requirements were assumed (ref.
27).
Data on reclamation in Wyoming (refs. 34, 35) in-
dicate that in the Powder River Basin, where the annual
precipitation is 12 to 14 inches or somewhat higher,
supplemental irrigation is in fact not needed to revege-
tate mine spoils. Of course, proper reclamation tech-
niques must be followed including returning top soil,
slope reduction, grass drilling, timely seeding, fertilizer
addition, mulching, etc. (ref. 34). On the basis of the
published information, it does appear that water alloca-
tion for revegetation is not required in such areas as the
Powder River Basin.
Summary of Mining and Reclamation Needs
The importance of site in determining water require-
ments is perhaps most clearly illustrated from our results
for the mining and reclamation examples. In table 17 we
have summarized the principal consumptive uses for the
mining and reclamation needs associated with an SNG
plant producing 250 x 106 set/day in Four Corners, New
Mexico and in Campbell County, Wyoming. Two things
stand out from these results; the large difference in
water needs between the two sites and the fact that even
the larger of the two values represents but a small frac-
tion of what had previously been considered to be the
total requirements for the plant and mine (see table 1).
CONSUMED WATER IN EVAPORATION,
SOLIDS DISPOSAL, OTHER USES
In this section we estimate the amount of water
consumed in evaporation, ash quenching, sludge dis-
posal, and other side uses associated with the operation
of the gasifier and complex, and with product removal.
The consumptions in these categories are somewhat de-
pendent on the specific gasification process employed
Table 17. Summary of principal consumptive
water requirements for mining and reclamation
examples for an SNG plant producing 250 x 106
scf/day
(gpm)
Four Corners,
New Mexico
(10. gal/day)
(gpm)
Road dust control" 147 0.21
Coal handling
Coal wash i ng 250 0.36
Revegetation 135 0.20
TOTAL 532 0.77
Campbell County,
Wyoming
(10. gal/day)
32
0.05
32
0.05
324
-------
and on the plant design, as well as on the site. Whenever
possible, we will try to provide a general methodology
for determining the amounts of consumed water. Often,
however, we will fall back on specific process character-
istics associated with proposed Lurgi designs, since most
of the published information available for comparison is
for these designs.
Pond Evaporation
Any estimate of the amount of water lost through
evaporation from settling ponds, reservoirs, or clarifiers
at the plant site will depend principally on three factors;
(1) the geographical location as it affects precipitation
and evaporation rates, (2) the pond areas, and (3)
whether evaporation suppressants are used. The fi rst
factor cannot be controlled once the plant location has
been fixed. To a large extent the pond area is governed
by the site, since it affects the quality of the supply
feedwater. For plain sedimentation, the dirtier or more
turbid the feed, the longer is the settling time and hence
the larger is the basin volume, which generally implies a
larger basin area. The extent to which use is made of
chemical agents, such as lime, to enhance flocculation
and speed up the sedimentation process is a matter of
economic tradeoff. The use of evaporation suppression
techniques is also a matter of economic tradeoff against
the price or need for water.
The principal uses for pond ing are; (1) settling
basins or clarifiers for the supply feed stream, (2) a
storage and surge reservoir for the clarified supply, and
(3) settling ponds for clarification of water to be re-
cycled, such as water from tailing ponds, etc.
If the volume flow rate (throughput) of water into
any pond (basin or reservoir) is known, the water surface
area is given by the relation
pond area = throughput x settling
time/water depth.
A conservative settling time for a basin to clarify a
turbid surface stream, where lime is used as a precipi-
tant, may be taken to be 24 hours. It is also reasonable
to assume a 1 week storage in the reservoir to accom-
modate for any interruption of the supply. Peak usage
and surge periods can be accounted for by designing for
a throughput which is 50% greater than the mean annual
average. For our example of an SNG plant in Four Cor-
ners, New Mexico, we estimate the total mean withdraw-
al from the feed stream to be 3,800 gpm, while for the
Campbell County, Wyoming, example, we take 2,600
gpm. Any small changes resulting from a difference in
these estimated totals can be accounted for once the
plant water requirements are determined in detail. A
water depth of 3 ft for the settling basin is characteristic,
while 21 ft for the reservoir would be reasonable. These
data along with the derived surface areas are shown in
table 18.
To estimate the mean loss of water from these hold-
ing areas, reference is made to table 15 where it is shown
that the annual difference between the precipitation and
evaporation is 57 inches in New Mexico and 43 inches in
Wyoming. This corresponds to a net evaporation rate of
about 0.1 gal/ft2/day in New Mexico and about three-
fourths that value in Wyoming. In table 19 are shown
the water loss rates for the surface areas of table 18
based on the net evaporation minus precipitation figures.
We would expect the total values to be maximum ones
since they assume a quite dirty feed stream and rather
conservative design criteria.
Table 18. Surface area requirements for settling
basins and reservoirs for SNG plant examples
in Four Corners, New Mexico and Campbell
County, Wyoming
Throughput Settling Water Su rface A rea
N.M. WYO. Time Depth N.M. WYO.
(gpm) (gpm) (days) (ft) (103 tel (103 te)
5,700 3,900 3 365 250
5,700 3,900 7 21 365 250
Settling Basin
Reservoir
325
-------
Table 19. Water loss rates from evaporation for settling
basins and reservoirs in SNG plant examples in Four
Corners, New Mexico and Campbell County, Wyoming
Settling Basin
Reservoir
TOTAL
Four Corners,
New Mexico
(gpm)
36
36
Campbell Cou nty,
Wyoming
(gpm)
18
18
36
72
The use of greater depths and other design changes
would reduce the surface area and hence the loss rate.
Furthermore, the application of monomolecular films on
the water surface could also greatly suppress the evapo-
ration, at probably a quite reasonable cost because of
the relatively large surface areas involved (ref. 36).
We may compare the total evaporation derived for
our Four Corners plant with the Fluor-Wesco design by
scaling our base throughput from 3,800 gpm to 5,100
gpm. In that case we would arrive at a water loss rate of
about 100 gpm compared with 420 gpm for the Wesco
plant (ref. 8). Of course, there are undoubtedly a
number of design assumptions which account for the
difference.
Additional pond evaporation losses are not evalu-
ated since they have, in part, already been taken into
account implicitly in the assumed makeup rate for the
coal preparation calculations. Other ponding recycle
losses are not expected to be large and are not evaluated
further here since they are quite specific to the treat-
ment system.
Ash Disposal
The quantity of ash or slag leaving a gasifier that
must be disposed of is a function of the ash content of
the coal fed to the gasifier. The ash content of the coal
in turn depends on the location of the mine from which
the coal was taken (see table 16). On the other hand, the
process defines the temperature at which the ash or slag
leaves the gasifier. This temperature is important for
water consumption since the ash or slag must be cooled
down before disposal, mainly for reasons of safety.
Normally, the cooling is done by quenching with water
down to a temperature somewhat below the boiling
point, generally around 200° F.
For most coals, ash will begin to deform around
2,000 to 2,500° F and fuse at temperatures about 100°
F higher (ref. 37). The temperature in the gasification
o
section of a Lurgi gasifier is between 1,400 and 1,600
F. However, the temperature is moderated at the bottom
of the reactor by the steam which is introduced there.
As a result, the 2sh probably leaves at a temperature of
from 800 to 1,400° Fe On the basis of reasonable
assumptions we have deduced from the data on ash dis-
posal given in ref. 23 that a temperature of about 1,000°
F appears to have been selected for the design value. In
slagging gasifiers, such as Bigas, the slag will be quenched
at temperatures closer to 2,700° F. For purposes of our
example we shall assume that in disposal the ash is
quenched from 1,000° F down to 200° F.
In our example, as outlined in the section on coal
washing, 21,860 tons/day of coal is fed to the gasifier.
From table 16, the weight percent of ash in the Navajo
coal is 26% and that in the Wyoming coal is 8.1 %. The
product of the coal rate and percent ash gives 5,685
tons/day of ash to be disposed of in our New Mexico
example and 1,770 tons/day in the Wyoming example.
The rate at which water is evaporated in cooling down
the hot ash can be determined from the relation for the
rate of heat removal, which in consistent units is given
by
heat removed/time = specific heat x temperature drop
x ash quenchedltime.
uividing the heat removal rate by the heat of vaporiza-
tion (1,000 Btu/lb) determi nes the evaporation rate.
Based on the analysis of bituminous ash (ref. 37), we
have estimated the ash specific heat to be about 0.2
Btu/lbf F. Using the estimated temperature drop of
800° F the amount of water evaporated is 152 gpm for
the New Mexico example and 47 gpm for the Wyoming
example. The figure given for the Four Corners case is
essentially the same as cited in refs. 5 and 23, as it
should be, since the ash loadings are the same and since
we derived our design temperature drop from the data in
these references.
326
-------
The weight of water remaining in the quenched ash
is taken to be 30% of the ash weight (ref. 23) or about
284 gpm for Four Corners and 89 gpm for Campbell
County. The total water requirement for quenching the
gasifier ash is thus 438 gpm and 136 gpm for New
Mexico and Wyoming, respectively.
In addition to the gasifier ash, if as in our example,
coal-fired boilers are used to generate steam, boiler ash
must also be disposed of. In the boilers, about three-
quarters of the ash will be removed by wet scrubbing of
the stack gases, while the remainder will be removed as
bottom ash. We can approximately estimate the amount
of consumed water by taking the water requirement per
unit weight of ash to be disposed of to be the same as
that for the gasifier. In the section on coal washing it
was pointed out that 3,760 tons/day of washed Navajo
coal was required for the boilers so that on an equivalent
heating value basis, 4,104 tons/day of the composite
Wyoming coal of table 16 would be needed. The respec-
tive ash disposal rates are 530 tons/day for Navajo coal
and 332 tons/day for Wyoming coal, with the corre-
sponding rates of consumed water equal to 40 gpm and
25 gpm.
Sludge Disposal
The largest amount of consumed water in the dis-
posal of sludge will be that water remaining in the sludge
from the settling basin that is used to remove the
suspended solids from the intake feed. The feed water is
assumed to enter the settling basin with a sediment con-
centration of 5,000 mg/1. In addition, since this is good
quality water, it is assumed that the sediment sludge will
be dewatered to 50% weight solids by centrifuging (ref.
23). Using the estimate of a total mean withdrawal from
the feed stream of 3,800 gpm for the New Mexico ex-
ample and 2,600 gpm for the Wyoming example, it fol-
lows that the quantities of water remaining in the sludge
will be .005 times the withdrawal or 19 and 13 gpm,
respectively.
Other Uses
Included among some other consumptive uses of
water cited are domestic and utility potable water for
plant and mine use, water for sulfur pelletizing, and
steam for cleaning the economizer tubes in coal-fired
boilers.
An integrated mine and plant complex of the type
described would employ about 1,000 people. We may
estimate the potable water usage for domestic supplies,
fire water, equipment, and plant washing at about 80
gpd/person. The total requirement is therefore about 56
gpm. However, about 90% of this water will be recover-
ed for reuse so that the consumptive use is quite small
amounting to only about 6 gpm.
in ref. 8, a relatively large quantity of water is indi-
cated for boiler tube cleaning-that is, cleaning of thl.
economizer tubes in the stacks of the coal-fired boilers.
When steam cleaning is employed, the requirement for
this purpose is usually about 2% of the steam generated.
The amount of steam generated can be estimated by
noting that 3,760 tons/day of 9,780 Btu/lb coal is burnt
producing steam at about an 88% efficiency. This is
approximately 2.7 x 106 Ib/hr giving a consumptive
water requirement of 108 gpm for boiler tube cleaning,
which may be compared with the Wesco design estimate
of 200 gpm (ref. 8). However, in our design example we
shall not consider this particular use of water since com-
pressed air can also be lIsed for economizer tube clean-
ing. The tradeoff of capital cost for the compressors
against the cost of water will be examined in the ongoing
study.
In ref. 8, 250 gpm of water is shown as consumed in
pelletizing sulfur. Here again, we shall not consider this
use because of the water-short nature of the areas
examined and since the sulfur can be shipped either as a
liquid or in cast blocks. These costs will also be examin-
ed in detail.
Summary of Evaporation, Disposal, and Other Needs
Once more, the importance of site is seen to stand
out in determining the water consumed in evaporation,
ash, and solids disposal. In table 20 we have summarized
these quantities and although the total difference be-
tween the two locations is not as dramatic as for mining
and reclamation, it is nevertheless quite significant.
Again, even the larger of the two totals is seen to be
significantly lower than had previously been suggested.
SUMMARY AND CONCLUSIONS
Taking the process net consumption of water equal
to 0.8 x 106 gal/day, consistent with the Lurgi design
(ref. 8), we can then combine this value with the results
of the present estimates to present a summary picture of
the total consumed water in our illustrative integrated
Western mine operations and SNG plants producing 250
x 106 set/day. This summary is given in table 21. It must
be emphasized strongly that these results are for the
illustrative conditions chosen for the examples of this
paper. They should not be taken as representative of all
SNG plants nor should they be used to make generalized
assessments without further interpretation. In this re-
gard, it is worthwhile mentioning again that even for
detailed designs, the consumed water estimates between
different process designs at the same location cannot be
expected to agree to within better than 20%.
327
-------
Table 20. Summary of amounts of water consumed in evaporation,
solids disposal and other use examples for an SNG plant producing
250 x 106 sef /day
Four Corners, Campbell Cou nty,
New Mexico Wyoming
(gpm) (106 gal/day) (gpm) (106 gal/day)
Pond Evaporation 72 0.10 36 0.05
Gasifier Ash Disposal 436 0.62 136 0.19
Boiler Ash Disposal 40 0.06 25 0.04
Sludge Disposal 19 0.03 13 0.02
Other Uses 6 0.01 6 0.01
TOTAL 573 0.82 216 0.31
Table 21. Summary of amounts of water consumed for specific
examples of integrated western mine operations and SNG plants
producing 250 x 106 scf /day
Four Corners, Campbell County,
New Mexico Wyoming
(106 gal/day) (106 gal/day)
Process, including fuel
gas generation 0.80 0.80
Cooling 3.11 2.64
Mining, reclamation 0.77 0.05
Evaporation, solids disposal,
other uses 0.82 0.31
TOTAL 5.50 3.80
A principal conclusion which can be drawn from the
results is that generalized estimates of water consump-
tion for Western mine-plant complexes to make synthe-
tic natural gas from coal are of limited value, if they are
not both site specific and design specific. Another con-
clusion is that the published water requirements for
integrated SNG plants and mine operations in the West
may be high and that the actual requirements' could,
depending on the location, be half the lowest estimate to
date.
ACKNOWLEDGMENTS
The present paper summarizes some results from a
general study on all aspects of water for synthetic fuel
production, including quantity, quality, and treatment.
This study is supported by the National Science Founda-
tion under RANN Grant No. SIA74-19080-A01 and the
Environmental Protection Agency under Contract No.
68-03-2207. We wish to acknowledge the assistance of
Joseph Shen in the preparation of this paper. We also
wish to acknowledge the assistance under subcontract of
Stone & Webster Engineering Corp. in various aspects of
the cooling study reported here. This work was per-
formed by W. D. Comley, C. Jones, and W. Paniciocco.
REFERENCES
1. Federal Energy Administration, Water Resources
Task Force, "Project Independence Blueprint-Water
for Energy," Report prepared by A. D. Little, Inc.,
Sept. 5, 1974.
328
-------
2. U. S. Department of the Interior, Water for Energy
Management Team, "Report on Water for Energy in
the Upper Colorado River Basin," July 1974.
3. U. S. Department of the Interior, Water for Energy
Management Team, "Report on Water for Energy in
the Northern Great Plains Area with Emphasis on
the Yellowstone River Basin," July 1975.
4. G. H. Davis and L. A. Wood, "Water Demands for
Expanding Energy Development:' U. S. Geological
Survey Circular 703, 1974.
5. EI Paso Natural Gas Company, "Second Supplement
to Application for Certificate of Public Convenience
and Necessity," Federal Power Commission Docket
CP73-131,1973.
6. C. R. Gibson, G. A. Hammons, and D. S. Cameron,
"Environmental Aspects of EI Paso's Burnham I
Coal Gasification Complex:' in Proceedings, En-
vironmental Aspects of Fuel Conversion Technology
(May 1974, St. Louis, Missouri). pp. 91-100, U. S.
Environmental Protection Agency.
EPA-650/2-74-118, Oct. 1974.
7. J. D. Strasser, "General Facilities Offsite, and
Utilities for Coal Gasification Plants," IGT
Symposium on Clean Fuels from Coal, pp. 639-652,
Institute of Gas Technology, Chicago, III., Sept.
10-14,1973.
8. T. E. Berty and J. M. Moe, "Environmental Aspects
of the Wesco Coal Gasification Plant," in Proceed-
ings, Environmental Aspects of Fuel Conversion
Technology (May 1974, St. Louis, Missouri). pp.
101-106, U. So Environmental Protection Agency,
EPA-650/2-74-118, Oct. 1974.
9. Federal Power Commission, National Gas Survey,
Vol. II. Task Force Reports, Section XII, p. 503,
1973.
10. E. Schmetz, H. Podall, and G. A. Mills, "Water Re-
quirements for a Synthetic Fuels Industry Based on
Coal:' in Proceedings, Workshop on Research Needs
Related to Water for Energy, Water Resource Center
Publ ication UILU-WRC-74-0093, University of
Illinois at Urbana-Champaign, Nov. 1974.
11. Hittman Associates, "Environmental Effects, I m-
pacts and Issues Related to Large Scale Coal Re-
fining Complexes, Vol. 2," Energy Research and
Development Administration R&D Report No. 101
- Interim Report No.2, (NTIS No. FE-1508-T2).
pp. 75-76, May 1975.
12. University of Oklahoma, Science and Public Policy
Program, "Energy Alternatives: A Comparative
Analysis," pp. 1-113, Univ. of Oklahoma, Norman,
Oklahoma, May 1975.
13. A. J. Forney et aI., "Analyses of Tars, Chars, Gases,
and Water Found in Effluents from the Synthane
Process," in Proceedings, Environmental Aspects of
Fuel Conversion Technology (May 1974, St. Louis,
Missouri), pp. 107-113, U. S. Environmental Pro-
tection Agency, EPA-650/2-74-118, Oct. 1974.
14. Air Products and Chemicals Inc., "Engineering
Study and Technical Evaluation of the Bituminous
Coal Research, Inc. Two-Stage Super Pressure Gasi-
fication Process:' Office of Coal Research R&D Re-
port No. 60, 1971.
15. J. P. Strakey Jr., A. J. Forney, and W. P. Haynes,
"Effluent Treatment and Its Cost for the Synthane
Coal-to-SNG Process," Am. Chem. Soc., Div. of
Fuel Chemistry Reprint Vol. 19, No.5, p. 94,1974.
(Paper presented at ACS 168th Natl. Meet., Atlantic
City, N. J., Sept. 1974).
16. U. S. Department of the Interior, "An Economic
Eva I uation of SY NTH AN E Gasification of
Pittsburgh Seam Coal at 1000 psia Followed by
Shift Conversion, Purification, Single-Stage Tube
Wall Methanation and Pollution Control:' Report
No. 74-31, Bureau of Mines, Morgantown, W. Va.,
1974 (RESTRICTED).
17. C. L. Tsaros, S. J. Knabel, and L. A. Sheridan,
"Process Design and Cost Estimate for Production
of 265 Million SCF/Day of Pipeline Gas by the
Hydrogasification of Bituminous Coal," Office of
Coal Research R&D Report No. 22 - Interim Re-
port No.1, 1965.
18. Transwestern Pipeline Co., Transwestern Coal Gasi-
fication Co., Pacific Coal Gasification Co., Western
Gasification Co., "Amended Application for Cer-
tificate of Public Convenience and Necessity,"
Federal Power Commission, Docket No. CP73-211,
1973.
19. J. H. Crow and J. R. Dungan, "Hot Potash Process
Purifies Gas at Lone Star's Fashing Plant," Oil &
Gas J., 97-99, Nov. 24,1958; also H. E. Benson and
J. H. Field, "New Data for Hot Carbonate Process:'
Petroleum Refiner, 39(4):127-132, April 1960; also
literature from Benfield Corp., Pittsburgh, Pa.
20. Air Products and Chemicals, Inc., "Feasibility Study
of a Coal Slurry Feeding System for High Pressure
Gasifiers:' Office of Coal Research R&D Report
No. 68 (Final). December 1971.
21. G. F. Cavaliere and R. A. Gyepes, "Cost Evaluation
of I ntercooler Systems for Air Compressors:'
Hydrocarbon Process., 52(1):107-110, Oct. 1973.
22. M. W. Larinoff and L. L. Forster, "Dry and Wet-
Peaking Tower Cooling Systems for Power Plant
Applications," Paper presented at ASME Winter
Annual Meeting, Houston, Texas, December 1975.
23. Batelle Columbus Laboratories, "Detailed Environ-
mental Analysis Concerning a Proposed Gasification
329
-------
~Idnt for Transwestern Coal Gasification Co., Pacific
Coal Gasification Co., Western Gasification Co. and
The Expansion of a Strip Mine Operation Near
Burnham, New Mexico Owned and Operated by
Utah International Inc.," Federal Power Commis-
sion, Feb. 1, 1973.
24. Atlantic Richfield Co., "Preliminary Environmental
Impact Assessment for the Proposed Black Thunder
Coal Mine, Campbell County, Wyoming," also
"Revised Mining and Reclamation Plan for the
Proposed Black Thunder Coal Mine, Campbell
County, Wyoming," Denver, Colorado, 1974.
25. Atlantic Richfield Co., "Black Thunder Mine, 10
Million Ton Per Year Water Supply," (Personal
communication, Hugh W. Evans), Denver, Colorado,
March 6,1975.
26. National Academy of Sciences, Rehabilitation
Potential of Western Coal Lands, p. 168, Ballinger
Publishing, Cambridge, Mass., 1974.
27. Utah I nternational Inc., "u II Water Requirements
for Fluor Wesco Lurgi Plant," (Personal communica-
tion, L. E. McCloskey), San Francisco, Cal., Oct. 3,
1975.
28. D. H. McCrea and J. H. Field, "The Purification of
Coal Dervied Gases: Applicability and Economics of
Benfield Processes," Paper No. 29b, AIChE 78th
Natl. Meeting, Salt Lake City, Utah, August 1974.
29. Northern Great Plains Resources Program, Mineral
Resources Work Group, Discussion Draft, Feb.
1974.
30. J. W. Leonard and D. R. Mitchell (editors). Coal
Preparation, pp. 17-5, 17-6, Am. Inst. of Mining,
Metallurgical, and Petroleum Engineers, New York,
1968.
31. Proceedings, Second Research and Applied Tech-
nology Symposium on Mined-Land Reclamation,
Oct. 22-24, 1974, National Coal Association,
Washington, D.C., 1974.
32. Proceedings, Third Symposium on Surface Mining
and Reclamation, Vol. I, Oct. 21-23, 1975, National
Coal Association, Washington, D.C., 1975.
33. F. E. Aldon, "Techniques for Establishing Native
Plants on Coal Mine Spoils in New Mexico," in Pro-
ceedings, Third Symposium on Surface Mining and
Reclamation, Vol. I, pp. 21-28, National Coal
Association, Washington D.C., 1975.
34. R. L. Tresler, "Strip Mine Reclamation in
Wyoming," in Proceedings, Second Research and
Applied Technology Symposium on Mined-Land
Reclamation, pp. 22-28, National Coal Association,
Washington, D.C., 1974.
35. T. Yamamoto, "Coal Mine Spoils as a Growing
Medium: Amax Belle Ayr South Mine, Gillette,
Wyoming," in Proceedings, Third Symposium on
Surface Mining and Reclamation, Vol. I, pp. 49-61,
National Coal Association, Washington, D.C., 1975.
36. Office of Water Resources Research, "Evaporation
Suppression, A Bibliography," WRSIC Report No.
73-216, (NTIS No. PB-231-200). U. S. Dept. of the
Interior, Washington, D.C., Dec. 1973.
37. T. A. Hendrickson (ed.), Synthetic Fuels Data
Handbook, Cameron Engineers, Inc., Denver,
Colorado, 1975.
38. R. N. Maddox, Gas and Liquid Sweetening, p. 57,
John M. Campbell and Co., 1974.
39. R. F. Neerken, "Compressor Selection for the
Chemical Process Industries," Chem. Engineering,
82(2):78-94, Jan. 20,1975.
40. Elliott Division Carrier Corp., "Elliot Multistage
Compressors," Bulletin, p. 25,1973.
41. F C. Riesenfeld and A. L. Kohl, Gas Purification,
2nd ed., p. 485, Gulf Publishing Co., 1974.
330
-------
APPENDIX
REPRESENTATIVE FORMULAE AND DATA USED FOR CALCULATING COST OF WET
AND DRY COOLING
Heat Transfer Coefficients (U)
1.
U, Btu/(hr)(ftz)(o F)*
Condensing steam from turbine drives
Condensing steam in the presence of nonconden-
sible gases in gas purification regenerator (ref.
38)
Cooling methane at 1.000 psig
Cooling water for off-gas scrubbing
Air and oxygen compressor interstage coolers
Dry
130
Wet
400
75
70
120
110
100
170
psig
10
50
100
300
~500
10
20
30
40
50
12
20
40
50
70
* Unfinned area for dry coolers.
2.
Heat Transfer Area
(a) Dry cooling
The following empirical
temperature rise was used
[T1 + Tz
~t = 0.005 U 2
formu la for the air
~tl
Here, t1 is the inlet air temperatu re and T 1, T z are the
process stream inlet and outlet temperatures. The other
formulae for calculating the heat transfer area are stand-
ard.
(b) Wet cooling
Standard formulae were used with assumed
cooling water temperatures of 80° F cold and 105° F
hot.
331
-------
3. Compressor Formulae
These may be found, for example, in refs. 39 and
40. The polytropic efficiency was taken to be 77 per-
cent.
4.
Energy for Cooling
Dry cooler fans
Horsepower = 0.015 X Area (U < 50)
= 0.0175 X Area (50 ~ U < 100)
= 0.02 X Area (U ~ 100)
Cooling tower fans
Horsepower = 0.012 X gpm circulated
Cooling water circulation pumps
Horsepower = 0.03 X gpm circulated
5. Costs
Area Cost Pressure (p, psig)
Dry cooling $ 18/ft2
Wet cooling $5.1/ft2 p<300
$5.6/ft2 30~ p<450
$6.1/ft2 45~ p<600
$8.9/ft2 ~600
Other
Cooling tower $7/gpm circulated
Compressors $135/hp
Electrical energy 2.24i/hp-hr (3i/kWh)
Steam $3/106 Btu
332
-------
SULFUR EMISSION CONTROLS FOR A COAL GASIFICATION PLANT
Milton R. Beychok*
Abstract
Through a summary of the WESCO coal gasification
project, an example of environmentally sound fuel con-
version technology is presented. It is proposed that first
generation gasification plants will be predominantly of
the Lurgi type.
The following recommendations are made concern-
ing sulfur control regulations specific to coal gasification
processes:
The regulations should not be in terms of
concentration.
Carbonyl sulfide emissions should not be ex-
cused. The regulations should include all sulfur
species.
The burning of coal, oil, and gas in the ae ~iIiary
steam boilers and superheaters is already cov-
ered by existing EPA regulations. The gasifica-
tion plant regulations should not include those
auxiliaries.
An acceptable and equitable regulation migh t
be:
THE TOTAL EMISSION OF ALL SUL-
FUR SPECIES IN ALL VENTS AND
STACKS FROM A COAL GASIFICA TlON
PLANT (EXCLUDING ANY STEAM
GENERATION FACILITIES) SHOULD
NOT EXCEED 1.5-3.0 PERCENT OF THE
SULFUR CONTAINED IN THE GASI-
FIER FEEDSTOCK COAL.
I very much appreciate this opportunity to present
an industry viewpoint on the control of sulfur emissions
from coal gasification plants. Later in this presentation, I
will discuss the detailed sulfur control features of an
actual design for a full-scale commercial coal gasification
project. But first I would like to talk for a few minutes
on how industry and the EPA might improve their work-
ing relationship. There would be far less court litigation
between industry and the EPA if industry could effec-
tively participate in developing emission control regula-
tions-and if industry viewpoints were truly given serious
consideration. I fully realize that some industry view-
points are nonproductive and perhaps overly opposed to
meaningful regulations. But the large majority of indus-
try is willing to accept emission regulations based upon
*The author is a consulting engineer in Irvine, California.
realistic data and commercially proven technology. The
EPA should solicit industry viewpoints very early in
their process of developing regulations, and provide
industry a real opportunity to participate in their deci-
sionmaking. It is not enough merely to invite formal
written comments within a legalistic time limit of 30 or
60 days after publication of a background document or
proposed promulgations-especially when those docu-
ments or proposed promulgations represent months or
years of work and many irreversible decision steps by
the EPA and their contractors. What industry wants is
early participation, face-to-face dialogues, and a mean-
ingful opportunity to help formulate proposed regula-
tions through all the stages of decisionmaking. I submit
that a working partnership between industry and the
EPA would drastically reduce the court litigation involv-
ing emission rej!ulations.
In considering the regulation of emissions from an
emerging industry such as coal gasification, we must first
determine what process technology that emerging indus-
try will utilize. There are at least ten full-scale commer-
cial coal gasification projects currently proposed and
seriously under design or study in the United States.
They include the following:
Location Owner
New Mexico WESCO (Western Gasification Co.)
New Mexico EPNG (EI Paso Natural Gas)
North Dakota Natural Gas Pipeline Co.
North Dakota ANG (American Natural Gas)
Wyoming Panhandle Eastern.
Wyoming Cities Service
Illinois Texas Eastern Transmission
Kentucky Texas Gas Transmission
Montana Burlington Northern
Utah (Confidential)
All but perhaps one or two of these plan to use Lurgi
coal gasification technology. The WESCO and EPNG
projects in New Mexico and the ANG project in North
Dakota have filed for regulatory approval with the
Federal Power Commission (FPC)-and some of the
other projects plan to file with the FPC fairly soon. The
WESCO project has already obtained a very favorable
FPC ruling and fully expects final approvals. WESCO has
also obtained permit approval from the New Mexico En-
vironmental I mprovement Agency and the New Mexico
Surface Mine Commission, and the final WESCO EIS will
soon be issued by the Department of I nterior. The plant
is fully designed and major equipment specifications are
ready for purchase. By all standards, the WESCO project
333
-------
is the most advanced of all the coal gasification projects.
All that is needed for the WESCO project to start con-
struction tomorrow is government assistance with fi-
nancing for coal gasification plants producing high-Btu
pipeline gas. There can be little doubt that the first gen-
eration of coal gasification plants will use Lurgi gasifica-
tion technology.
Another consideration in developing regulations for
an emerging industry is how to obtain a realistic basis for
such regulations. When developing standards for an exist-
ing industry, the EPA first examines existing plants to
determine the emission abatement levels attained in
"exemplary" plants. Since we have no exemplary plants
in the emerging coal gasification industry, we can only
examine "exemplary" designs. This brings me to my first
major point:
The EPA should recognize that the first-
generation coal gasification plants will
most probably use Lurgi technology. The
EPA should develop regulations specific to
Lurgi technology and should base those
regulations on the "exemplary" designs of
WESCO, EPNG, and ANG.
In my opinion, the WESCO, EPNG, and ANG exemplary
designs are much more meaningful than any "paper"
designs that could be or have been developed by any
EPA contractor. Having been with the WESCO project
for almost 5 years, I know that millions of dollars have
been spent to bring that design and project to its current
status. An EPA contractor limited to $100,000 or
$200,000 of funding could never hope to equal the
design expertise expended on the WESCO project.
It is also my opinion that the EPA should not dilute
their efforts by attempting at this time to develop regu-
lations that would include the many second-generation
gasification processes under development in the United
States by ERDA, the American Gas Association, Bitumi-
nous Coal Research, the Institute of Gas Technology,
and others. When and if the second-generation demon-
stration plants funded by ERDA have been proven
successful, the EPA will have access to their operating
data and could then develop emission regulations for the
subsequent commercial exploitation of the second-
generation processes. This brings me to my next major
point:
The EPA should develop regulations speci-
fic to second generation coal gasification
processes after the ERDA demonstration
plants have shown which of those processes
will truly be commercially viable.
This recommendation is perfectly consistent with the
existing EPA policy of developing regulations specific to
process categories within a given industry. For example,
the EPA's effluent guidelines for the petroleum refining
industry are divided into five sets of guidelines for five
categories of refinery process arrangement. As another
example, the NSPS for lead smelters (40 CFR 60, Sub-
part L) specifies different emission standards for rever-
beratory furnaces than for pot furnaces; thus it would be
quite consistent for the EPA to:
First develop coal gasification emission regu-
lations for the first generation Lurgi technology
based on the available exemplary designs.
Then develop regulations for the second-
generation coal gasification plants based on
operating data from the ERDA-funded demon-
stration plants.
In my opinion, the Booz-Allen report, contracted by the
EPA to develop a data base for coal gasification sulfur
emission regulations, encompassed too broad a scope by
including first &rJd second generation plants. As noted
earlier, a $100,000-200,000 report could not hope to do
an adequate study of Lurgi technology alone--and it cer-
tainly couldn't do an adequate study of the many first
and second generation processes as the Booz-Allen re-
port attempted to do.
WESCO DESIGN AND RATIONALE
Now, let us examine the sulfur emission control de-
sign for the WESCO project and the rationale for the
design. But first I want to emphasize that the WESCO
design is based on meeting the gasification emission
standDrds developed by the State of New Mexico, which
are the only such standards yet promulgated in the
United States. The auxiliary coal-fired steam boiler in
the WESCO design is also based on meeting the New
Mexico regulations for coal combustion which are far
more stringent than the EPA's standards for coal com-
bustion. If I may digress for a moment and refer back to
my opening remarks about letting industry participate in
developing regulations, the New Mexico coal gasification
regulations are an outstanding example of such coopera-
tion. In late 1971, WESCO was asked by New Mexico to
assist them in developing regulations and WESCO re-
sponded affirmatively. The New Mexico EIA and
WESCO worked together over a period of many months
to hammer out a set of draft regulations. The draft was
then reviewed in depth at a public meeting of the New
Mexico EIA's Technical Advisory Committee which in-
cluded representatives of local environmentalist organi-
zations. A redrafted version was then reviewed at a pub-
lic hearing before the New Mexico Environmental Im-
provement Board, and a final version was subsequently
adopted into law in September 1973. The final version
was not completely in agreement with WESCO's original
334
-------
suggestions, but it was accepted and implemented in
WESCO's plant pesign. When WESCO submitted their
design along with a permit application, the New Mexico
EIA granted a permit and approval to construct. As of
today, there has been no litigation and no appeal either
by industry or by the strong local environmentalist orga-
nizations. All in all, it was a story of excellent coopera-
tion between industry and the State EIA.
Now, let me explain the WESCO sulfur emission
control design (see figure 1). For those of you who may
not be familiar with coal gasification to produce high-
Btu pipeline gas, the key processing steps are those
within the bold outlines on the flow sheet (figure'):
Gasifiers - pressure vessels where coal, steam, and
oxygen are reacted to yield a crude gas containing
methane, hydrogen, carbon oxides, excess steam, and
various byproducts and impurities. About 40% of the
plant's ultimate methane (SNG) product is produced
here. Subsequent reaction steps produce the other 60%.
Shift Conversion - excess carbon monoxide in the crude
gas is catalytically converted ("shifted") to carbon
dioxide to provide the 3-to-' hydrogen-to-carbon mon-
oxide ratio required for additional methane synthesis in
the subsequent methanation step.
Rectiso/ Gas Purification - the Rectisol absorption proc-
ess, using refrigerated methanol, selectively removes
hydrogen sulfide and carbon dioxide from the crude gas.
Precooling at the Rectisol process entry recovers by-
product naphtha.
Methanation - hydrogen and carbon monoxide (in a
3-to-' ratio) are catalytically combined to form methane
and byproduct water. About 60% of the end product
methane (SNG) is produced here, and a final Rectisol
scrub removes any residual carbon dioxide.
The gasifier feedstock coal contains 226.4 tons/day
of sulfur. About 5% (11.3 tons/day) of that remains in
the byproduct tars, oils, and naphtha and the remaining
95% (215.1 tons/day) is gasified. About 2.19 tons/day
of the gasified sulfur are in the form of carbonyl sulfide
(CaS) and the remaining 212.91 tons/day are in the
form of hydrogen sulfide.' Thus about 94% of the coal
sulfur appears as gasified H2 S and about 1 % appears as
cas. Since the emission control of cas is somewhat
more difficult than the control of H2 S, it is important to
define the amount of cas produced during gasification
when comparing different designs.
*AII sulfur quantities on figure 1 (including CaS) are ex-
pressed as tons/day of elemental sulfur. Also, the "all sulfur
species" quantities include the cas quantities-they are not
cumulative.
335
As shown in figure 1, some of the gasified sulfur
distributes among seven minor gas streams, but most of
it (202.9 tons/day) remains with the crude product gas
entering the Rectisol unit. As noted earlier, the Rectisol
unit removes sulfur and carbon dioxide from the crude
gas. Most of the sulfur (141.6 tons/day) is removed as a.
rich-H2S gas stream containing roughly 21 vol % H2S,
Another lean-H2S gas stream contains roughly 0.9 vol %
H2S (61.1 tons/day of su Ifur). The lean-H2~ stream. is
much larger than the rich-H2S stream because It contalils
about 8,800 tons/day of C02 while the rich-H2S stream
has only about 700 tons/day of C02'
The small rich-H2S stream would be ideally suited
for routing through a Claus unit which converts H2S
into byproduct sulfur-but it contains more hydrocar-
bons (38 tons/day) than a Claus unit can tolerate. The
large lean-H2S stream (0.9 vol % H2S) is much too lean
for a Claus unit and yet contains more sulfur (61.1
tons/day) than point source regulations would permit
releasing to the air-as well as 94 tons/day of hydrocar-
bons which, if released to the atmosphere, would create
serious difficulty in meeting the ambient air quality
standards (AAOS) for hydrocarbons. I ncineration of the
lean-H2S stream would eliminate the hydrocarbon prob-
lem but would still release the sulfur to the atmosphere
as sulfur dioxide. In order to solve the problems of
point-source sulfur emissions and ambient air quality
hydrocarbon standards, the WESCO design does the
following:
The small rich-H2S stream is further treated in
a second gas purification unit for the selective
se paration of hydrocarbons (shown as an
"H2S-concentrator" in figure 1). The resultant
enriched-H2S stream contains about 70 vol %
H2S and is ideally su ited for routing through
the Claus unit along with one of the minor gas
streams (from naphtha separation within the
Rectisol unit). The Claus unit will convert 95%
of this H2S to byproduct sulfur.
The large lean-H2S stream and the offgas from
the phenol extraction unit are routed through
Stretford units for essentially complete conver-
sion of their H2S to byproduct sulfur (98.6%
conversion). However, the lean-H2S stream con-
tains about 0.9 tons/day of sulfur as cas which
is not converted in a Stretford unit.
The Claus unit tail gas, the Stretford unit tail
gas, the hydrocarbons from the H2S-concen-
trator, and other minor gas streams are all in-
cinerated in the steam boiler fireboxes. This
converts their 12.3 tons/day of residual sulfur
species to S02 and prevents any hydrocarbon
emissions. The resultant S02 is then removed at
-------
(+.2) [0.') (2)
(137.+0) SUL.FUR
BY PROOLfC. T
(~5~ I!£WVEKf)
';~
'" °
"d
~'--'
BOILER
I FIREBOXES
"' ~'"
~~! ASH
(1.1)
(.....0) (..5.05) SULFuR
[O.~] BY PROOUCT
(98... 7. RECOVERY
~..../'. ~~A
.0", :::~T (0.21)
i~~ ~~~ [0.19]
(71'} (0.20)
~ 224 TONS/D OIL
11.250 Bru/LB HHV
O.O~ WT% S
(1.23)
NAPTHA 5EP. GAS
(3.23)
(:3.01)
CAT. REGEN. GAS,"
GAS-LiQ. vENT;
GAS-LiQ. FLASH
SULFUR REMOVED
(~:;; REMOVAL)
STARTUP VENT,
COAL LOCK VENTS
(0.08)"
<11.'1)
(.U
(.U
O'J
(226.4)
GASIFIER COAL
GASIFIERS,
SHIFT CONVERSION,
RECTI SOL PRECOOL
~
u. ~
° ..-
~£
w
..
..
0;
f--
U1
(0.21)
-----
(0.20)
SUPERHEATER
FIREBOX
------
Z+.820 TONS/D
8.32S 8TU/L8 HHV
O.~12 WT,. S
SNG PRODUCT
< 5,600)
TARS, OILS ('1.2)
NAPTHA (2.1)
NH. PRODUCT
PHENOL PRODUCT
REUSE WATER
TOTAL GASIFICATION PLANT SULFUR EMISSIONS
= 0.08 + 1.2l\ + 0.:2.1 - 1.52 TONS/DAY
= 0.7% OF GASIFIER COAL SULFUR
ASH
(
[
<
) ALL SULFUR SPECIES, TONS/ D
) CARBONYL SULFIDE, TONS/D
) HYDROCARBONS, TONS/D
.. DAILY EMISSIONS ON ANNUALIZED BAsiS
(a) 21 VOL'j(, H,S
(b) 0.1 VOL 7. H,S
(c) APPROX. 70 VOL % H,S
TOTAL IIOILER PLANT SULFUR EMISSIONS
= 3.23 + 0.20 = 3.4-3 TONS/DAY
= 10 7. OF BOILER COAL AND FUEL OIL SULFUR
SULFUR DISPOSITION
ALL SULFUR QUANTITIES ARE EXPRES5/00 AS TONS/DAY
OF ELEMENTAL SULFUR. THE '~LL SULFUR SPECIES. QUANTITIES
INCLUDE THE (.05 QUANTlT/E5.
Figure 1. Sulfur disposition
-------
90% efficiency in the boiler stack gas scrubbers.
The final Rectisol offgas (containing about
8,000 tons/day of C02' 0.21 tons/day of sulfur
and 77 tons/day of hydrocarbons) is incinerat-
ed in the steam superheater firebox. This pre-
vents any hydrocarbon emissions and the small
amount of resultant S02 does not require re-
moval.
The net result (as shown in figure 1) is that the total
gasification plant sulfur emissions amount to only 0.7%
of the sulfur in the total gasifier coal feedstock. This
excellent control is achieved by incinerating the tail
gases from the primary sulfur recovery units (Claus and
Stretford) which converts any residual sulfur species into
S02 that is then removed in the boiler stack gas scrub-
bers. Thus, even the COS which is not converted in the
Stretford units (or from any other source) is removed
after incineration to S02.
Table 1 (emissions vs. regulations) illustrates how
the gasification plant sulfur emissions will meet the New
Mexico regulations. It also illustrates how the emissions
from fuel combustion in the steam boilers and super-
heaters will better the New Mexico regulations which are
2-3 times more stringent than the Federal regulations.
This design exemplifies a number of points which
are germane to any proposed EPA regulations:
1.
The problem of point-source sulfur emission control
is interrelated with the prevention of hydrocarbon
emissions in amounts that would violate the ambi-
ent air quality standards. They cannot be solved in-
dependently of each other.
Carbonyl sulfide emissions can be prevented by in-
cineration followed by S02 removal, and inciner-
ation is quite probably required in order to control
hydrocarbon emissions.
2.
Table 1. Emissions v.s: regulations
Gasifier Coal Heating Value
= (24,820) (2000) (8,325)
'" 413,253 x 10' Btu/day
Gasification Plant Sulfur Emissions
= 0.08 + 1.23 + 0.21
= 1.52 tons/day
= (1.52) (2000)/413,253
= 0.0074 pounds/1 O' BtU
----------
Boiler Plant Coal Fines Heating Value
= (3,8701 (2000) (10,040)
= 77,710 x 10' Btu/day
(New Mexico regulation = 0.0080)
--~------_._---
Boiler Plant S02 Emissions (from coal-firing)
= (3.23) (2)
'" 6.46 tons/day
= (6.46) (2000)/77,710
'" 0.166 pounds/10' Btu
(New Mexico regulation = 0.340)
(Federal regulation = 1.20)
-----------------------~---------
Superheater Fuel Oil Heating Value
= (224) (2000) (17,250)
'" 7,728 X 10' Btu/day
Sl.Iperheater S02 Emissions (from oil-firing)
'" (0.20) (2)
'" 0.40 tons/day
= (0.40) (2000)/7,728
0.104 pounds/10' Btu
(New Mexico regulation = 0.340)
(Federal regulation = 0.80)
337
-------
3. The number and diversity of gas streams (13 are
shown in figure 1) cannot be simplified into a single
regulation expressed as point-source concentration
in ppm. Also, the common use of stack gas S02
scrubbers for the boiler plant flue gases and the
gasification residual tail gases complicates the prob-
lem. If the regulation is expressed as concentration
in ppm, it will cause a great many problems of inter-
pretation regarding common stacks, common scrub-
bers, and how to "average" the various exit gas
streams. It might also encourage plant designers to
take advantage of C02 or N2 dilution or excess
combustion air in any number of ways.
Table 2 summarizes my recommendations regarding
coal gasification sulfur control regulations, and I urge
you to give them serious consideration:
The regulations should not be in terms of con-
centration.
Carbonyl sulfide emissions should not be
excused. The regulations should include all sul-
fur species.
The burning of coal, oil, and gas in the auxiliary
steam boilers and superheaters is already cov-
ered by existing EPA regulations. The gasifica-
tion plant regulations should not include those
au xi I i aries.
An acceptable and equitable regulation might
be:
The total emission of all sulfur species in
all vents and stacks from a coal gasification
plant (excluding any steam generation
facilities) should not exceed 1.5-3.0 per-
cent of the sulfur contained in the gasifier
feedstock coal.
The 1.5-3.0% should be subcategorized for regional
differences in coal sulfur contents. Thus, low-sulfur coals
might be limited to 1.5% emission and high-sulfur coals
to 3.0% emission. This would provide an equalizing
incentive to gasify high-sulfur Eastern coals. Otherwise,
our large energy reserves of Eastern coal might not be
utilized. In any case, even a 3% emission of sulfur in
gasification is only one-tenth of the sulfur emitted by an
equivalent coal-burning power plant.
CLOSING
In closing, I want to emphasize two caveats concern-
ing my presentation. First, I do not and cannot speak for
Table 2. Recommendations
1.
Regulations should not be in terms of concentrations /i.e.,
ppm by volume).
2.
Regulations for gasification should not include auxiliary
steam boilers and superheaters. Burning of coal, oil,and gas
in steam plants is covered by existing EPA regulations.
3.
The gasification plant regulations should include all sulfur
species from all vents and stacks /including annualized in-
termittent operations).
4.
Carbonyl sulfide emissions should not be "excused."
5.
Recommended wording for high-Btu gasification regula-
tion:
"THE TOTAL EMISSION OF ALL SULFUR SPECIES IN
ALL VENTS AND STACKS FROM A COAL GASIFI-
CATION PLANT (EXCLUDING ANY STEAM GEr -R-
ATION FACILITIES) SHOULD NOT EXCEED 1.5-3.0
PERCENT OF THE SULFUR CONTAINED IN THE
GASIFIER FEEDSTOCK COAL"
The 1.5-3.0% should be subcategorized for regional differ-
ences in coal sulfur contents. Thus, low-sulfur coals might
be limited to 1.5% emission and high-sulfur coals to 3%
emission.
338
-------
the entire energy industry. All of my remarks and
opinions are based on my involvement with the specific
WESCO design to meet the New Mexico regulations.
Secondly, I should explain my recommended 1.5-3.0 %
emission limit in contrast to WESCO's design limit of
0.7% as shown in figure 1.. As an engineer who has spent
a lifetime in the process design of industrial plants, I can
assure you that the difference between 0.7% and 1.5% is
a realistic hedge against the difference between design
expectations and actual plant performance.
Finally, I want to thank you for giving me the
opportunity to make this presentation. And I want to
assure you that most of us in industry share with you a
real concern for the integrity of our environment. I am
sure that we can both accomplish much more by
listening to each other than by shouting at each other.
339
-------
LOW ENERGY GAS RETROFIT TO INDUSTRY
D. A. Ball*
Abstract
Low- or intermediate-energy gas from coal (with a
heating value of 150 to 400 Btulscf) is an alternate fuel
for industries faced with reduced availability of natural
gas and an uncertain fuel oil supply. Such a gas can be
produced and cleaned of such contaminants as
particulates, sulfur, and nitrogen compounds using
commercially available technology.
This paper describes conversion of two hypothetical
industrial plants (a petroleum refinery, and a secondary
steel mill) to onsite production of low- and
intermediate-energy gas from coal as a source of process
energy. Specific gasifier and gas cleanup equipment
requirements are discussed for the hypothetical plants,
considering such factors as peak plant load, assumed coal
source, and process and piping requirements. The
possibility of utilizing existing plant piping and gas
burners is discussed. General material and energy flows
are presented. The potential environmental impact of
conversion to gas from coal is also discussed in terms of
emissions from both the gasification process and the
combustion process.
INTRODUCTION
I n recent years, American industry has been faced
with an acute shortage of natural gas as a source of
energy. As a result, many industries have become
increasingly reliant on oil and propane, substitutes that
are easiest to utilize minimum equipment modifications.
However, increased demand for these alternate fuels-
along with limited supply and foreign politics-have
resulted in their cost increasing dramatically and the
availability being uncertain.
Table 1 shows the dependence of industry on natu-
ral gas and petroleum. In 1973 industry consumed
approximately 15.9 Quads (1015 Btu) of natural gas and
petroleum to satisfy 70 percent of its total energy
requirement of 22.9 Quads.
Projections of the future supply of natural gas and
oil vary. Figure 1 shows a recent projection of domes-
tically produced natural gas. This projection shows that,
with stimulation techniques, supplies of natural gas may
improve somewhat in the near future. However, many of
these techniques are unproven and their use would result
*The author is with Battelle's Columbus Laboratories,
Columbus. Ohio.
in higher costs of recovering gas. Expensive LNG imports
will also aid supply; but, over the next 40 years, con-
sensus is that natural gas will all but disappear as a
significant energy source in the United States. Either
through allocation or increased price this fuel will
become less attractive to industry at a much earlier date.
Oil and propane have been in somewhat greater
supply to industry than natural gas, though at higher
prices. Much of this supply, however, has been satisfied
by foreign imports. In the future, the U. S. government
will undoubtedly place increased pressure on reducing
these imports to improve its balance of payments posi-
tion. However, the projection of domestic production of
oil, shown in figure 2, would predict a steadily decreas-
ing supply of this fuel. Therefore, oil will probably be
more expensive and less available to industry in the
future, especially as it is used more heavily as a substi-
tute for natural gas.
One proposed alternate source of energy for indus-
try has been low- or intermediate-energy gas produced
from coal. Commercial processes for producing this gas
were used widely in American industry over 50 years ago
and, although only a few commercial applications
remain in this country today, many are still being used
in various foreign countries.
Low-energy gas with a heating value of 120 to 200
Btu per standard cubic foot (using air and steam for
gasification) and intermediate-energy gas with a heating
value of 280 to about 400 Btu per standard cubic foot
(using oxygen and steam for gasification) could con-
ceivably be generated onsite and substituted for natural
gas and oil in industrial fuel using processes. Such a sub-
stitution, however, involves many considerations such as
environmental effects, gas distribution, process and
burner modification, and economics.
The objective of this study was to evaluate these
considerations and their implications for two selected
Table 1. Energy use in industry (ref. 1)
Quads Percent
Natural Gas 10.4 46
Petroleum 5.5 24
Coal 4.6 20
Electric 2.4 10
341
-------
30
20
z 24
o
- 21
I-
g 18
81
&: 12
-1 9
-------
industries. Selection of the two industries was based on a
preliminary analysis of their relative attractiveness and
potential for being retrofitted to low- or intermediate-
energy gas. Model plants were selected for analysis in
each of the two industries. A suitable gasification-gas
cleaning system was then selected for each of the two
model plants. Material and energy balances were calcu-
lated for each gasification plant based on their'generat-
ing sufficient low- or intermediate-energy gas to satisfy
the energy requirements of the model plants. Analysis
was then made of the potential environmental impact of
such a retrofit and the modifications necessary in the
plant fuel distribution and fuel using processes. This
paper will deal primarily with the potential environ-
mental effects of such a retrofit.
INDUSTRY SELECTION
The major criterion used in selecting candidate
industries for study was existing use of natural gas and
oil by 4-digit SIC code. Table 2 shows the 10 top con-
sumers of natural gas, oil, and coal in the United States
in 1971 along with their relative ranking in use of these
three fuels and amounts of the fuels consumed.
These industries were then evaluated according to
other criteria such as amenability of processes to conver-
sion to low- or intermediate-energy gas, location of the
industry relative to supplies of coal, and dependence on
a supply of gaseous fuel or inabilitv to convert to direct
use of coal. Finally, the basic iron and steel industry
(SIC 3312) and the petroleum refining industry (SIC
2911) were selected for study.
Basic Iron and Steel Industry
The basic iron and steel industry was subdivided
into two parts: integrated mills and secondary mills.
Integrated mi lis are generally large mi lis that produce
steel from raw iron ore by reducing the ore to iron in a
blast furnace using coke. This is followed by conversion
of the iron to steel, primarily by the Basic Oxygen
Process. Because these processes produce significant
quantities of waste fuels-particularly the blast furnace
and coking operation, integrated steel mills are not as
dependent on oil and natural gas as other types of steel
Table 2. Energy use by the 10 major industrial
consumers of gas, oil, and coal by 4-digit SIC
codes for 1971 (ref. 2)
. Ranking Energy use 1012 Btu /yeara
SIC Description Gas Oil Coal Gas Oil Coal Totalb T otalC
3312 Blast furnaces and steel 2 1 3 653 167 133 953 1254
2818 Industrial organic chemicals 3 10 4 605 26 132 763 908
(not elsewhere classified)
3241 Hydraulic cement 5 5 1 208 41 180 429 429
2621 Paper mills 6 2 2 201 161 149 511 526
2631 Paperboard mills 7 3 5 181 157 82 420 436
2911 Petroleum refining 1 4 1332 68 9 1409 1440
2819 Industrial inorganic chemicals 4 375 23 35 433 453
(not elsewhere classfied)
3334 Primary aluminum 8 125 1 16 142 143
3221 Glass containers 9 120 9 129 130
3352 Aluminum rolling and drawing 10 61 2 3 66 72
2821 Plastics and resins 7 10 54 29 37 120 134
2611 Pulpmills 6 38 38 4 80 81
2815 Cyclic intermediate and crudes 8 13 28 72 113 126
2824 Organic fibers (noncellulosic) 9 8 42 26 60 128 132
2812 Alkalies and chlorine 6 47 4 80 131 139
2823 Manmade fibers (cellulosic) 7 12 2 64 78 80
3313 Electrometallurgical products 9 2 1 51 54 56
aGas at 1038 Btu/scf; Oil at 6.0 x 10. Btu/42 gallon barrel; Coat at 26.2 x 10. Btu/short ton.
bTotal of purchased fossil (gas, oil, coal).
CTotal of all purchased fuels.
343
-------
mills or other industries. Secondary mills, on the other
hand, have no blast furnaces or coking facilities and
produce steel from iron and steel scrap, primarily in
electric-arc furnaces. These mills are somewhat smaller
but more numerous than integrated mills and are heavily
dependent on natural gas and petroleum products for
fuels. A recent survey (ref. 3) revealed that a sample of
integrated steel mills were only 18.4 percent dependent
on natural gas and only 26.3 percent dependent on
natural gas and petroleum products combined; second-
ary steel mills were 65.8 percent dependent on natural
gas and 75.3 percent dependent on natural gas and
petroleum products combined.
It was estimated that 22 percent of the natural gas
and petroleum products used by the basic iron and steel
industry (SIC 3312) were consumed in secondary mills.
Also, it was estimated that 70 percent of secondary mill
capacity is located in the 20 major coal producing and
coal bearing states.
Fuel use in secondary steel mills is similar to that in
integrated mills and consists of heating large pieces of
metal in a variety of furnaces for forming and heat treat-
ing operations. Other significant fuel uses include boilers
and space heating. I n preliminary analysis, these proc-
esses were considered relatively amenable to conversion
to low- or intermediate-energy gas.
Petroleum Refining Industry
The refinery industry (SIC 2911) is by far the top
industrial consumer of natural gas for fuel, consuming
about twice as much as the next major consumer, basic
iron and steel (SIC 3312). In 1973 natural gas accounted
for about 36.1 percent of energy consumed in refinery
operations and oil accounted for an additional 11.4 per-
cent (ref. 4). An additional 35.1 percent of refinery
energy consumption was satisfied by off-gases from
refinery processes. This so-called refinery gas-consisting
primarily of methane, ethane, butane, propane, and
hydrogen, is a high-grade gas mixture and could poten-
tially be classed as a premium fuel and sold.
Fuels consumed in refinery operations are primarily
for feedstock heating in process heaters and steam
raising in boilers. Although refineries vary considerably
in type and size, the fuel-consuming processes (process
heaters and boilers) are relatively uniform in characteris-
tics and on preliminary analysis were considered rela-
tively amendable to conversion to low- and intermedi-
ate-energy gas. The amount of fuel consumed in a
refinery is largely a function of (1) refinery size, which
can vary from less than 5000 barrels of crude per day to
over 400,000 barrels per day; and (2) refinery complexi-
ty, a function of the number and type of different
processing operations.
GASIFIER CLEANUP SYSTEM CONSIDERATIONS
A variety of commercially available gasification
systems were considered as possible candidates for each
of the two model industries selected. Table 3 is a list of
gasification processes which are commercially available.
Several characteristics were taken into account in select-
ing gasification systems for each of the model plants.
These considerations were:
(1) Unit size and turndown ratio of the gasifier.
Unit size determines the number of units required for
satisfying the energy demand of the industrial plant.
Unit size of turndown ratio is combined to determine
the flexibility of the system in meeting the varying load
demands of the plant. For each model plant, a gasifier
unit size was selected such that three to four gasifiers
could satisfy the energy demand of the industrial plant.
This number of units would supply adequate flexibility
and yet not be so numerous as to cause undue plant
complexity and high cost.
(2) The ability of the gasification system to gasify
the coals that are available at the industrial plant was
also considered an important aspect of the design. Non-
agitator fixed-bed units, shown in table 3, are generally
restricted to coals with free-swelling index less than 3.
Agitator-type fixed-bed units are generally capable of
handling slightly higher caking coals, up to a free-
swelling index of about 7. Entrained slagging type
processes are generally considered capable of handling
any type of coal, although coals with a high grindability
index and low moisture content are considered more
attractive du"e to the necessity for pulverization. The one
fluid-bed unit shown in table 3 would be restricted to
mildly caking coals.
(3) Another important consideration is the by-
products produced in the gasification process. I n the
low-temperature processes, i.e., fixed-bed processes,
these byproducts are generally tars, oils, phenols, and
ammonia. Some steam could be produced by the waste
heat from the gasifier, but this would not be considered
a substantial product. Conversely, the high-temperature
entrained slagging processes produce significant amounts
of steam, but-due to their high temperature, produce
little chemicals such as oils, tars, phenols, or ammonia. If
a particular industry has a need or use for either the
steam or chemical byproducts produced in the gasifica-
tion process, then one type of process or the other may
be considered more attractive.
A wide variety of gas desulfurization systems were
also considered in the selection. These included physical
sorbent systems which usually operate at pressures of
about 300 psig, chemical absorption systems which
usually operate at about atmospheric pressure, and
344
-------
Table 3. Commercial gasifiers considered for model industry plants
Unit Output
Gasifier Bed Type 10' Btu/day Potential Byproducts Limitations
Lurgi fixed/agitator 8-12 tar, oil, phenols, NH3 needs sized low-caking
coal
Kopper9-Totzek entrained/slagging 8-16.5 steam oxygen-blown only,
Weliman..Galusha pulverized coal
fixed)'agitator 1.5-2.5 tar, oil, phenols, NH3 needs sized low-caking
coal
Applied Technology fixed/2-stage 0.2-2.3 tar, oil, phenols, NH3 needs sized coal,
FSI <3
Winkler fluidized 1.3-14 steam needs crushed low-caking
coal
Riley Morgan fixed/agitator 1.5-3.0 tar, oil, phenols, NH3 needs sized low-caking
coal
Woodall Duckham fixed /2 -stage 1.0 tar, oil, phenols, NH3 needs sized coal,
FSI <3
Babcock & Wilcox entrained/slagging 15 steam primarily O2 -blown,
pulverized coal
M.W. Kellogg fixed/agitator 2.5-3.5 tar, oil, phenols, NH3 needs sized low-caking
coal
Wilputte fixed/agitator 0.6 tar, oil, phenols, NH3 needs sized low-caking
coal
physical-chemical absorption systems which involve a
combination of physical and chemical absorbents. Physi.
cal sorbent processes have a high absorptivity for sulfur
in the gas, but also remove hydrocarbons. The removal
of hydrocarbons from the fuel gas is undesirable. They
also tend to remove carbon dioxide as well as sulfur
compounds, thus complicating sulfur recovery. Chemical
absorption processes are selective for hydrogen sulfide
over carbon dioxide and have very low absorptivities for
hydrocarbons. These processes, however, are not capable
of removing sulfur to as Iowa concentration as physical
sorbent processes. Physical-chemical absorption proc-
esses combine some of the positive features of both the
physical and chemical type processes, but also involve
some compromises. I n the chemical absorption proc-
esses, sulfur can either be recovered from a hydrogen-
sulfide rich gas stream resulting from the stripping of the
sorbent solution or from direct oxidation of hydrogen
sulfide to sulfur in the sorbent solution itself. After
reviewing these systems, it was decided to select a clean-
up system which is commonly recommended by the
particular gasifier vendor of the selected gasification
system, as this would be the most likely choice should a
plant actually be constructed.
A final sulfur level in the fuel gas of 300 ppm was
selected somewhat arbitrarily for both plants on the
basis of several considerations.
(1) The limit would allow both plants to operate
within the most stringent state S02 emissions standards
for coal-fired processes.
(2) This low limit was judged adequate to limit
corrosion in piping and possible su Ifidation of products
although more research is necessary to accurately define
these limits.
(3) It represents a lower practical limit for atmos-
pheric cleanup systems such as those studied. 80th types
of systems studied can achieve lower concentrations, but
at higher cost and with the possibility of more complex
sulfur recovery.
MODEL STEEL PLANT DESCRIPTION
The secondary steel plant selected as a model has a
steel production capacity of 1,110,000 tons per year.
The mill has production facilities for producing rolled
bars, cut and coiled rod, and merchant products such as
angles and channels. Some finished products such as wire
and nails are also made in the model plant. The energy
demand of this plant ranges from a minimum of 15 X
109 Btu/day to a maximum of 23 X 109 Btu/day with
an average energy demand of 20 X 109 Btu/day. Mini-
mum demand usually occurs on weekends and during
down time for plant maintenance. Table 4 summarizes
the characteristics of the steel plant model.
345
-------
t..)
,I:>.
m
storage
N2
Coal
pulverizer
Oxygen
Saturated
. steam
/ ----- Wo,t. h.ot
recovery
To
pulverizer
Air
Clean gas
-
MDA sulfur
removal
Filter
Sulfur
Sludge
Make up
Figure 3. Koppers Totzek/MDEA steel mill gasification plant.
-------
Table 4. Steel plant statistics
ton/year
Molten steel capacity
Rolled bars
Cut and coiled rod
Angles, channels
1 ,110,000
170,000
250,000
350,000
Energy demand
Maximum
Average
Minimum
109 Btu/day
23 (peak production)
20 (normal production)
15 (weekends, downtime)
A Koppers-Totzek gasification systen. Nas selecteu
for the steel plant model. This plant would consist of
four two-headed gasifier units and would gasify about
1900 ton/day of coal producing approximately 22.3 X
109 Btu/day of fuel gas with a high-heating value of 286
Btu per standard cubic foot. An MDEA (methyl-
diethanolamine) cleanup system was selected for use in
desulfurizing the fuel gas. This process, which operates
at atmospheric pressure is commonly recommended by
Koppers in such applications. Figure 3 shows an overall
flow sheet of the gasifier and gas cleanup system.
The Koppers-Totzek process is restricted primarily
to oxygen-steam gasification producing intermediate-
energy gas of about 300 Btu per standard cubic foot.
The intermediate-energy gas was considered desirable for
the steel mill application after reviewing the combustion
requirements of the various furnaces in the plant. Con-
version of these processes to a gas with a lower heating
value would require extensive conversion and modifica-
tion.
The coal selected for the steel mill model was a
lignitic coal with a moisture content of about 27 percent
and a su Ifur content of 1.7 percent. This coal would be
pulverized to 70 percent through a 200-mesh screen and
dried during the pulverization step to less than 4 percent
surface moisture before feeding to the gasifier.
Gas produced in the gasification step would be
passed through a waste-heat boiler for steam recovery
and then through a two-stage venturi scrubber for
removal of particulates and any tars that may be formed
in the process. Because the Koppers-Totzek gasifier
operates at such a high temperature (about 3300° F),
only trace amounts of tars, phenols, oils, and a relatively
small amount of ammonia are present in the fuel gas.
From the venturi scrubber the gas would be processed
through a cooler and then into the MDEA sulfur-removal
system for removal of sulfur compounds.
The final fuel gas would contain about 300 ppm of
sulfur compounds and the overall efficiency of the gasifi-
cation process would be about 70 percent. Statistics on
the gas plant design are summarized in table 5.
Figure 4 shows a representative plot plan of the
steel mill with the gasifier and coal-storage piles in-
cluded. The steel mill proper covers about 230 acres of
land with approximately 70 acres of land in building
area. Though the buildings in figure 4 are shown linked
together, in reality they would be interspersed through-
out the available area and relatively little free land would
be available in the steel mill proper. The gasification and
gas-cleaning plant is estimated to require about 15 acres
of land and storage for a 1-month supply of coal would
require additional 2-1/2 acres.
MODEL REFINERY DEtiRIPTION
The model plant developed for representing the
refinery industry would process about 25,000 barrel/day
of crude oil. (See table 6.) Products produced at the
refinery wou Id consist primarily of residual oil, gasoline,
and distillate oil. In the summer the refinery would con-
vert a large percentage of residual oil to asphalt and in
winter this residual oil would be sold as heating oil. The
refinery processes a low-sulfur crude oil with less than 1
percent sulfur.
Total energy demand of the refinery is 7.2 X 1 O~
Btu/day in the winter and 5.79 X 109 Btu/day in the
summer. This demand could be satisfied by a mixture of
refinery waste gas and low-energy gas from coal. The
refinery gas contains a significant amount of propane,
butane, and hydrogen-equivalent to about 0.67 X 109
Btu/day. For this study, these constituents are assumed
to be 98 percent removed and sold. The remaining
refinery waste gas consists primarily of methane, ethane,
and some nitrogen and has a heating value of about
1,062 Btu per standard cubic foot. This gas would sup-
ply about 2.46 X 109 Btu/day of energy, and the re-
mainder would have to be satisfied by a low-energy gas
from coal.
Table 5. Gasification plant design for
steel mill model
Gasifier - Koppers-Totzek
Desulfurization - MDEA {methyl diethanolamine}
Number of units - 4
Gas production rate - 22.3 x 109 Btu/day
Gas heat value - 286 Btu/scf
Coal consumption - 1900 ton/day
Efficiency - 70 percent
347
-------
Gasification.
B Cleaning
.15 acres
Coal
storage
2.5 acres
Steel Mill Proper:
Land area 230 acres
Building area 70 acres
1-- - - ---,-- --- - -T-~-- --..
I I Bor joist I Rolls I
I I ,...------1
: Wore house ~ - - - - - --I Long spon I
I I Mill building r:S~ -.- - -~
I I I hipping I
L_____~__--__L___-~
r------,
Ie b . I
1'0 rleoting I
'---T--1
I I I
I I I
I Nails, I I
I wire I I
I . ~I
I mesh 1"0 U I
I ., "
I I ::~ I
~__l~~
r--.,
r------, , I
: Moto r room I Ie'
I ..10
r----,----r----rJ-------,-- .c; I
~ Serop yord :Electrie I Sooking IBlooming: IBIliel : ~:::d
L- --- - _~~~e- .LP!!.s_--l.:n~ _J. ~l!.I__- ~~ ~l
Figure 4. Steel mill plot plan.
The coal selected for use in this system was an
Eastern bituminous-type coal with 6 percent moisture, 8
percent ash, and sulfur content of 3.9 percent. The free-
swelling index of this coal is about 5, dictating the use of
an agitator-type fixed-bed gasifier.
Table 6. Refinery plant statistics
Crude oil capacity 25,000 barrel/day
Products (barrel/day) Summer Winter
Residual oil 1600 4750
Gasoline 12,700 11,525
Distillate oil 5625 4600
Total Energy Demand 5.79 7.20
(109 Btu/day)
Refinery waste gas 2.46 2.46
Low-energy gas d\!mand 3.33 4.74
Due to the low overall energy demand of the
refinery, an air-blown fixed-bed, Wellman Galusha gasifi-
cation system was selected for study. The gasification
plant would supply about 4.74 X 109 Btu/day in the
winter and 3.33 X 109 Btu/day in the summer. The
mixture of refinery waste gas at 1,062 Btu per standard
cubic foot and low-energy gas from the Wellman Galusha
at 168 Btu per standard cubic foot would have a heating
value of about 264 Btu per standard cubic foot in the
.winter and 234 Btu per standard cubic foot in the
summer.
Figure 5 shows the flow sheet for the Wellman-
Galusha gasification plant.
Due to the small size of the plant, it was not felt
practical to install coal preparation facilities; therefore,
crushed, sized coal would be purchased from the mine
and stored at the gasification plant. The gas plant would
consist of three 10.ft diameter Wellman-Galusha units
capable of producing a total of 4.77 X 109 Btu/dayof
fuel gas With a heating value of 168 Btu per standard
348
-------
Coo i storage
w
.j:>o
co
Stretford
absorber
Air
Cooling wa ter
Ash "
Makeup waler
Ammonia
stripping
Cooling pond
Figure 5. Flow sheet for the Wellman-Galusha gasification plant.
- Clean gos
~SUJfur
~
Phenol
removal
-------
w
(TI
o
~
Coal storage
I acre
Gasifi ca-
tion B
Cleaning
1.0 acre
Coo Ii ng pond
! .acre
River bonk
~
--
Processing. 4 acres
I - Ir " I
I G0501lne lilight g05 II Cotol ylic I
~~.::uct~~ J ~~~~~J ~~:..k~n~_J
,- - - - - - - - -.,
I Proce55 I
LS~pJ:~~ -- -_J
,------- ----,
: Crude Oil :
Lo~s~ ~a~on___J
Ref inery storage. 23 acres
-----
~
---
Figure 6. Refinery plot plan.
~
--
-------
cubic foot. The coal consumed would be about 252
ton/day and the overall thermal efficiency of the plant
would be 76.i percent. The raw gas from the gasifier is
processed directly through a scrubber for the removal of
tars, oil, phenols, and ammonia, and then through a
cooler section where additional ammonia, tars, and other
condensible constituents are removed. The gas is then
fed into a Stretford-type desulfurization system which
oxidizes sulfur compounds to elemental sulfur in solu-
tion, eliminating the need for a Claus plant. The final gas
product would contain 300 ppm of sulfur, or less, and
would be mixed with refinery gas and distributed to the
various processes in the refinery. Table 7 summarizes the
pertinent characteristics of the refinery gasification
plant.
Table 7. Gasification plant design for
refinery model
Gasifier - Wellman-Galusha (3 unitsl
Desulfurization - Stretford
Gas production rate - 4.77 x 10' Btu/day
Gas heat value - 168 Btu/scf
Coal consumption - 252 tonlday
Efficiency - 76.7 percent
Figure 6 shows an overall plot plan of the refinery
with gasification and coal storage facilities and also the
required cooling pond. The processing facilities of the
refinery itself occupy about 4 acres of ground, and
storage capacity requires an additional 23 acres of
ground. The gasification plant for the refinery is esti-
mated to require about 1 acre of ground with an addi-
tional 1 acre required for a cooling pond. Storage facili-
ties for 1-month supply of coal would require an addi-
tional 1 acre.
EMISSIONS FROM THE GASIFICATION
PROCESSES
Model Steel Plant
Table 8 summarizes the major emissions from the
gasification process for the model steel plant. The major
points of emissions in this process are the coal storage,
coal pulverizing and preparation facilities, the oxygen
plant, the filter which separates water from slag and
clarifier sludge, cooling tower, and the Claus sulfur
recovery process.
Emissions from the coal storage pile will involve
fugitive dust picked up by the wind and leachate result-
ing from rain water filtering through the coal pile. The
coal pile should be packed tight to limit dust loss and
conveyors should be hooded with the hood exhaust
processed through a baghouse or electrostatic
precipitator. Leachate from the coal pile would resemble
acid mine drainage in many respects-containing acids,
organics, and soluble metals. This water should be col-
lected and ponded for biological reduction of pollutants
before being discharged to a water course. These prob-
lems can be minimized by coating the coal pile with a
plastic material and drawing from it only during periods
of emergency. The coal would normally then be taken
directly from unit train or barge by covered conveyors.
Emissions from the coal pulverizing preparation step
consists of pollutants in the gas used in drying the coal.
A portion of the final product gas is combusted to heat
air which is then supplied to the pulverizer for drying
purposes. This stream is then vented from the pulverizer.
The stream consists primarily of carbon dioxide, nitro-
gen, some water vapor, and oxygen. The stream would
also contain particulates and possibly some small
amounts of sulfur dioxide oxidized from the coal.
The vent stream from the coal preparation step
should be processed through a baghouse or electrostatic
precipitator or some other efficient particulate removal
device for controlling particulate emissions. Emissions of
other constituents would be relatively minor at this
stage.
To limit dust loss, the entire coal pulverization
facility should be located in a building with positive
ventilation control. The exhaust from the building
would then be processed through a particulate control
device.
The discharge from the oxygen plant would involve
primarily nitrogen which is not considered a harmful
emission and would require no control.
The sludge from the filtration process which
separates water from the slag and clarifier sludge is rela-
tively inert and could be filled or ponded. As a result of
the use of a filter, this sludge would have a low water
content. The sludge would typically contain compounds
common in coal ash and should require little or no treat-
ment.
A significant discharge to the atmosphere would be
the cooling tower plume. The cooling tower water would
contain dissolved constituents from the scrubber circuit
that overflows from the clarifier. These constituents
would be present to some extent in the drift loss or
plume from the cooling tower. Although many of these
compounds may be present only in infinitesimal
amounts when combined with the water in the plume,
351
-------
Table 8. Discharges from steel mill model gasification plant
Area of
Source Impact Amou nt Composition Percent
Coal storage - fugitive dust Air, dependent on coal dust
and leachate from rain Water wind and rain acids
conditions organics
soluble metals
Coal pulverizing Air 492,160 Ib/hr CO. 2
(or 112, 144 scfml N. 68
H.O 13
0. 17
Oxygen plant Air 72, 704 scfm N. 100
Filter Water, 21,616 Ib/hr C 17.8
Solid Waste Ash 72.2
H.O 10.0
Cooling tower plume Air 17,500 Ib/hr H.O 100
Claus Solid Waste 1,704lb/hr S 100
(byproduct)
Claus tail gas Air 675 scfm H.S 2.7
CO. 97.3
'they may create a corrosion or health menace in the area
around the plant. A solution to this problem is to use
dry cooling towers or a cooling pond-either of which
would involve much greater cooling area.
About 1704 Ib/hr of elemental sulfur would be
produced from the Claus plant in this DroCP.!:S. This
sulfur would be of marketable quality and could be
stored and shipped. The Claus process, however, only
removes about 95 percent of the sulfur compounds of
the inlet stream. The resulting tail gas or vent stream
from the Claus process, therefore, would contain hydro-
gen sulfide and CO2. With the system shown this tail, gas
could be blended with the product gas from the gasifier
and com busted without exceeding even the strictest
state limitations on sulfur dioxide emissions.
Model Refinery Plant
The major emissions from the refinery model gasifi-
cation plant are shown in table 9. Sources of emissions
are coal storage, the gasifier itself, scrubber effluent, and
emissions from the Stretford desulfurization process.
Effluents from coal storage would involve similar
considerations to those discussed for the steel plant
model. Because crushed, sized coal would be purchased
from the mine, however, dust loss for the refinery would
be less than for the steel plant due to the lower per-
centage of fines or small particles.
About 1768 Ib/hr of dry ash would be emitted from
the gasifier in the form of bottom ash and could either
be trucked or slu iced to a pond or landfill site. This ash
should present no serious disposal problems.
The effluent from the scrubber system contains
significant amounts of tars, ammonia, and phenols,
which would have to be treated prior to disposal. In
some cases these products may be able to be used in the
industrial plants or marketed. For instance, in the case
of the refinery, the recovered tars could possibly be used
in the making of asphalt to supplement residual oil
supplied. However, this would have to be evaluated as to
the effect of these tars on the asphalt production process
of the plant. Tars would be recovered by decantation
and would result in a composition of about 91 percent
tar, and 9 percent water. Ammonia and other
compounds such as trace amounts of hydrogen sulfide
which may be dissolved in the scrubber water, could be
steam stripped and recovered for sale. Phenols could also
be recovered for use through the so-called Phenolsolvan
process, or they could be biologically reduced to sludge
and separated from the water for disposal.
There is no immediate use for phen01s in the
refinery so biological reduction would probably be
employed. The economics of this versus recovery of the
phenols in a potentially more expensive process would
have to be evaluated furthr-r.
352
-------
Table 9. Discharges from refinery model gasification plant
Area of
Sou rce Impact Amount Composition
Coal storage - fugitive dust Solid Waste depends on wind coal dust
and leachate from rain and rain conditions acids
organics
soluble metals
Scru boor effluents Water
Tar separation 1153 Ib/hr tar
H20
NHa stripping 1095Ib/hr NHa
H20
Phenols 120 Ib/hr phenols
Stretford Solid Waste 777 Ib/hr sulfur
(byproduct)
Percent
91
9
20
80
100
100
The Stretford desulfurization process for this
particular design would produce about 770 Ib/hr of
elemental sulfur which would be stored and sold.
EMISSIONS FROM COMBUSTION PROCESSES
Emissions from combustion processes fall generally
into four categories: (1) emissions of sulfur dioxide, (2)
oxides of nitrogen, (3) particulates, and (4) trace constit-
uents such as polycyclic organic matter or heavy metals.
In the gasification process,sulfur compounds in the
coal are converted to sulfur compounds in the gas on
almost a quantitative basis. The major su Ifur-bearing
constituent is hydrogen sulfide with minor amounts of
carbon-sulfide (COS), carbon disulfide (CS2), and
mercaptans. If these compounds are not removed prior
to combustion of the fuel gas, they are oxidized quanti-
tatively to sulfur dioxide in the eventual flue gas from
the combustion process. These expected emissions are
shown in figure 7 as a function of coal sulfur and heat
content. As can be seen from figure 7, a coal-sulfur con-
tent of from 0.5 to 0.8 percent would be required for
most coals before compliance with the Federal standard
of 1.2 pounds of S02 per million Btu heat input could
be met. Coals for both the steel plant model and refinery
plant model were assumed to contain over 1 percent
sulfur, thus requiring some form of desulfurization.
Figure 8 shows the expected emissions of sulfur dioxide
from combustion processes as a function of sulfur con-
tent in the fuel gas for different fuel-gas heating values.
To meet the Federal standard for new sources of 1.2
pounds of S02 per million Btu heat input based on the
10.0
8.0
"
Cjj
..
Q 6.0
....
:2
.;
C
Q
-.;
i 4.0
1>.1
..
o
VI
2.0
2e~!..!!!!~~~~8~
1.0
2.0 3.0
Percent Sulfur in Cool
4.0
~.O
Figure 7. 502 emissions versus sulfur in coal.
gas heating value, a sulfur level in the fuel gas of ~nly
about 1000 ppm would be required even for low-energy
gas. As the heating value of the gas rises, the allowable
sulfur content also rises. Many states, however, have
tighter standards for S02 emissions than the Federal
standard, and it appears that the trend is for tighter
standards to be promulgated. The New Mexico standard
(.34 pounds of 802 per million Btu) would require a
maximum of about 300 ppm sulfur in the gas to be
353
-------
N
o
-------
8 10 12 14 16
MGSS Fraction 01 Nitrogen in Fuel Gas
Figure 9. Fuel nitrogen converted to NOx (ref. 5).
Table 11. Relative NOx production of various
fuels at 10 percent excess air
Fuel
Relative NOx
Production
Natu ral Gas
Kopper&-Totzek (steel mill)
Wellmim-Galusha (refinery)
1.00
0.93
0.74
Data on oxidation of fuel-bound nitrogen to NOx'
however, shows that the fuel-bound nitrogen is not
quantitatively converted to NOx' instead part of it is
converted to elemental nitrogen which would not be
considered a pollutant. It appears that the conversion
rate of fuel-bound nitrogen to NOx in the combustion
products decreases as the concentration of fuel-bound
nitrogen increases. It also appears that the conversion
rate is relatively independent of the form in which the
fuel-bound nitrogen exists in the gas. Figure 9 shows the
results of a compilation of data from various sources on
conversion of nitrogen in the fuel to NOx in the combus-
tion products. This curve shows that for low concentra-
tions of nitrogen in the fuel gas, conversion of fuel-
bound nitrogen to NOx is essentially 100 percent and
the conversion decreases to a minimum of about 15
percent at high concentrations of nitrogen.
Data on fuel-bound nitrogen content or ammonia
content in fuel gas from coal is lacking. Table 12 gives
some values of NH3 concentrations obtained from the
literature for four gasification processes. These processes
1.0
"-
o
Z
a
"0 0.8
~
c:
a
U
c: 0.6
o
:<
0.0
o
2
4
6
include entrained slagging and fixed-bed operation using
both air and oxygen as the gasifying medium. Table 12
also shows the expected NOx emissions as a function of
gas-heat value using the curve of Dykema-Hall from
figure 9 for nitrogen conversion. The values shown in
table 12 do not include thermally produced NOx but, in
some cases, they still exceed the Federal standard of 0.7
Ib NO/106 Btu for solid fuel-fired systems and in all
cases exceed the Federal standard of 0.2 Ib NO/106 Btu
for gas-fired systems. As stated above, these values are
for the raw gas from the gasifier and in cleanup systems
where water scrubbing is used (such as the.two model
plants), nearly all the ammonia in the fuel gas can be
removed prior to combustion.
Particulate content in the final clean product gas
from both of the gasification plant models in negligible.
Combustion of this gas, therefore, would be expected to
result in negligibie particulate emissions to the atmos-
phere and no particulate control would be required. In
both model plant cases, the low-energy gas would be
replacing the firing of some heavy oil which would result
in an overall decrease in particulate emissions from these
two industries.
Emissions of trace organic constituents such as poly-
cyclic organic matter (POM) are a function of the
number of long chain hydrocarbons or ring-type hydro-
carbons in the fuel itself and of the combustion condi-
tions. Coal and oil both contain significant numbers of
these compounds. However. the product gas from gasi-
fication should contain few, if any. long chain hydro-
carbon components or ring-type hydrocarbon compo-
nents. Combustion conditions for firing the fuel gas
would be similar to those for firing natural gas; thus,
18
20
22
24
355
-------
Table 12. Ammonia in raw gas
Process
NH3 Volume
Percent
NOx
Ib/10' Btua
Koppers-=l"otzek (°2) (ref. 6)
2-Stage Entrained (aid (ref. 7)
Lurgi (pressurized °2) (ref. 8)
Fixed Bed (atmospheric air) (ref. 7)
0.17
0.38
0.70
0.25
0.44
1.46
0.63
1.16
aUsing relationship of figure 9.
emissions of these types of materials would be expected
to be similar to that of natural gas. They would be signif-
icantly less than if the coal were fired directly or if oil
were used directly as the fuel.
Other trace constituents, such as trace metals that
may be vaporized in the combustion process, are also
potential pollutants. The more volatile metals (mercury,
etc.,), would be vaporized in the gasification process but
subsequently condensed in the coal-gas cleaning process.
Table 13 shows representative concentrations of trace
elements, in tar and condensate from the synthane
reactor; thus emissions of trace elements would pri-
marily be confined to the liquid and solid residue from
the gasifier rather than to the combustion products.
Table 14 summarizes emissions from the two indus-
trial plant models. These emissions would be compared
with those from the industrial plant prior to conversion
to gasification. The most notable increase in emissions
from the combustion process is S02 em issions. I n both
cases, these emissions would be below the new-source
Federal standard for S02 emissions for coal-fired
sources. I n the case of the steel plant model, it is as-
sumed that the tail gas from the Claus plant is combined
with the clean gas from the gasification plant. This ap-
proximately doubles the sulfur emissions from this
particular process.
Other Aspects of Converting Industry to Gas From Coal
Most industrial furnaces can be converted to at least
an intermediate-energy gas, without extensive modifica-
tion. Many furnaces could also successfully be converted
to low-energy gas,although the lower flame temperatures
and larger flue gas volumes impose certain limitations on
the type of furnace that could be successfully converted
without extensive modification. In some cases, existing
burners could probably be successfully modified for
intermediate energy gas, especially if the gas heating
value could be increased to about 400 Btu per standard
cubic foot by removal of CO2. However, in many cases,
new burners would be required. For low-energy gas, new
burners would almost always be required.
Most industrial natural gas distribution systems are
designed for up to 125 to 150 pounds per square inch
gage. However, natural gas is seldom distributed at this
pressure and in many cases is distributed at 50 pounds
per square inch gage or less. The distribution pressure is
then normally reduced at the burner to pressures on the
order of a few ounces per square inch gage. Under these
conditions it may be possible to substitute at least an
intermediate-energy gas in an existing natural gas dis- I
tribution system although pressurization of the gas
would be required.
Table 13. Trace elements in tars and condensate
from Synthane - 40 atm PD U gasifier (ref. 9)
ConstitUent
Concentration (ppm) jI.g/g
Tar" Condensate
Arsenic (As)
Bismuth (Bil
BOFon (B)
Br~mine (Br)
Cadmium (Cd)
Fluorine (F)
Galium (Ga)
Germanium (Ge)
Lead (Pb)
Mercury (Hg)
Nickel (Nil
Selenium (Se)
Zinc
0.71-1.6
0.1 7-0.30
12-17
0.01-0.08
0.15
0.97-12
0.02-0.17
0.02-0.08
0.16-0.48
0.21-1.2
0.95-1 .6
0.08-0.24
0.48-3.3
'0.001
43-82
0.001
32-39
0.001
0.003
0.027-0.030
0.001-0.Q18
0.14-0.18
0.007-0.13
356
-------
Table 14. Effect on emissions from combustion sources
Particulate Emissions
Reduced - Particulate content of gas is negligible.
Gas would replace some current oil- and coal-firing
Increased -
Steel Mill - 0.383 Ib SO, /1 O. Btu of gas
0.267 Ib SOi /1 O. Btu og coal
Refinery - 0.200 Ib S0, /1 O. Btu of gas
0.232 Ib S0, /10. Btu of coal
So,
NOx Emissions
Reduced - Fuel gas contains negligible nitrogen
(NH3, HCN) and thermal NOx would be less due
to lower nitrogen content of flue gas.
CONCLUSIONS
Retrofitting industrial plants to low- or intermedi-
ate-energy gas presents many environmental considera-
tions, but there appears to be little reason why such a
retrofit cannot be done in an environmentally acceptable
manner.
Sulfur and nitrogen compounds in the fuel gas (pri-
marily H2S and NH3) represent the primary sources of
atmospheric emissions from combustion processes.
These constituents are also potentially corrosive, espe-
cially if any water vapor is present. Although the corrosi-
vity of these constituents has not been accurately de-
fined, it appears that their concentration in the fuel gas
should be reduced below that which would be consid-
ered environmentally acceptable to minimize corrosion
in the intricate and extensive gas distribution systems
found in most industrial plants.
The major environmental hazard involves the gasifi-
cation plant itself. The potential atmospheric emissions
in the fuel gas become potential liquid and solid ef-
fluents. A variety of processes will be required for treat-
ing the various liquid streams used in cleaning and cool-
ing the fuel gas. I n many industries, water treatment
systems are already in use similar to those that would be
required for a gasification plant. In the refinery industry,
for example, water treatment processes are commonly
used for removing oils, phenols, ammonia, and H2S;
however, in almost all cases these treatment systems
would have to be enlarged or additional processes added
if a gasification plant were installed. The expertise and
technology for treating these various streams in an envi-
ronmentally acceptable manner are available.
The major factor in determining whether an in-
dustry would convert to low- or intermediate-energy gas
is economics. Even a small gasification plant would in-
volve many processes, most of which involve either
cleaning the gas or treating various effluent streams as-
sociated with gas cleaning. These various cleaning sys-
tems also constitute the major cost in a coal gasification
plant. Modifications required in furnaces are primarily
determined by the heating value of the gas with higher
heating value gases requiring fewer modifications. The
requirement for complex gasification plants and attrac-
tiveness of a high grade fuel gas (such as intermediate-
energy gas from an oxygen-blown gasifier) tend to favor
large-scale industrial applications. Thus, gasification is
most attractive for large, individual, industrial plants or
groups of smaller plants in an industrial park.
REFERENCES
1.
"A National Plan for Energy Research, Develop-
ment, and Demonstration: Creating Energy Choices
for the Future, Volume I: The Plan," Energy Re-
search and Development Administration, ERDA~,
June 1975.
"1972 Census of Manufacturers:' U.S. Department
of interior.
"Special Survey on Gross and Net Consumption of
Fuels and Energy in the Iron and Steel Industry,"
American Iron and Steel Institute, May 1974.
Oil and Gas Joumal, April 7, 1975.
Dykema, O. W., and Hall, R. A., "Analysis of Gas,
Oil, and Coal Fired Utility Boiler Test Data:' U.S.
EPA Symposium on Stationary Source Combustion,
September 1975.
Farnsworth et aI., "Clean Environment With the
Koppers-Totzek Process," EPA Meeting, Environ-
mental Aspects of Fuel Conversion Technology,
May 1974.
2.
3.
4.
5.
6.
357
-------
7. Robson, F. l., Giramonti, A. J., "The Environ-
mental Impact of Coal Based Advanced Power Gen-
erating Systems," Symposium Proceedings, Environ-
mental Aspects of Fuel Conversion Technology, pp.
237-257, October 1974, EPA 650/2-74-118.
8. Shaw, H., and Magee, E. M., "Evaluation of Pollu-
358
9.
tion Control in Fossil Fuel Conversion Processes,
Gasification, Section 1, Lurgi Process," July 1974,
EPA 650/2-74-069C.
Forney, A. J., et aI., "Trace Element and Major
Component Balances Around the Synthane PDU
Gasifier," August 1975, PERCITPR-75/1.
-------
COMBINED-CYCLE POWER SYSTEMS
F. L. Robson, W. A. Blecher, and A. J. Giramonti*
Abstract
The performance, cost, and emissions of integrated
fuel processing/combined-cycle power systems are iden-
tified and compared to those of a coal-fired steam sta-
tion.
Of the many types of advanced power systems being
proposed as partial solutions to this country's long range
energy problem, the COmbined Gas And Steam
(COGAS) system is the only one currently in the
commercial stage, with several thousand Mw already in-
stalled in utility systems. The systems now installed and
those more advanced versions being proposed require a
very clean fuel in order to assure reasonable gas turbine
performance and lifetime. While these clean fuels can be
produced from coal, the processes are inefficient, typi-
cally losing 20-30 percent of the original coal heating
valve. By integrating the fuel processing system with the
COGAS power system, much of the inefficiencies which
appear as process heat can be recovered, resulting in
overall plants having efficiencies, costs, and emissions
which are more attractive than present-day conventional
coal-fired steam stations.
INTRODUCTION
The United States has been blessed with an abun-
dance of coal, much of it in close proximity to large
urban areas. This coal was used widely by the electric
utilities to fuel their boilers. However, the advent of
inexpensive imported oil displaced coal from many
urban locations on the eastern seaboard, wh ile many
southern and western power stations switched to even
cheaper natural gas. The use of coal has been further
reduced by stringent sulfur emission regulations in many
of the coal-producing States. By mid-1973, less than 55
percent of the power in the United States was generated
from coal.
Research on methods of converting coal to a more
environmentally acceptable fuel has been underway for a
number of years. Additional impetus was given to these
efforts by the OPEC-imposed oil embargo after the
October War of 1973. Most of the coal conversion
methods under investigation involve changing the coal to
a new form, usually liquid or gas. This requires a con-
siderable amount of energy, thus, it is typical that only
*The authors are with the United Technologies Research
Center, East Hartford, Connecticut.
75 percent or less of the coal heating value is available as
chemical energy in the resulting fuel. In order to be
utilized economically in the utility industry, some
method of more efficient electrical generation is re-
qu ired to overcome all or at least a major portion of this
fuel cycle inefficiency.
One such system, identified some years ago (e.g.,
ref. 1,2), is the COmbined Gas And Steam or COGAS
power system. More recently, the COGAS system has
again been identified as the most attractive advanced
power system (see figure 1 taken from ref. 3).
COMBINED-CYCLE POWER SYSTEM
There are a number of cycle configurations possible
with COGAS or combined-cycle power systems. The
steam boiler may be fired with additional fuel or it may
be pressurized (supercharged) and serve as the combus-
tor for the gas turbine. For applications using advanced,
high-temperature gas turbines, the simple waste-heat
recovery cycle shown schematically in figure 2 is the
most attractive. Currently, several thousand Mw's of
combined-cycle systems of this configuration are in-
stalled or on order for use with clean liquid fuels.
The major reasons for the current interest in
COGAS systems are the high performance and low
capital cost. These systems typically cost only 50-60 per-
cent of a conventional steam power plant and have in-
stallation times of only 24-32 months. The performance
on a clean liquid fuel such as home heating oil or jet fuel
is currently around 41 percent with potential increases
to 50 percent or more as gas turbine technology
advances (figu re 3).
COAL GASIFICATION
The gasification of coal has been carried out for
commercial purposes for quite a long time. One of the
first instances in this country was the use of gasified coal
to produce illuminating gas for the Baltimore Gas light
Company in 1817. Prior to the widespread availability of
inexpensive natural gas, much of this country's industry
used atmospheric pressure gasifiers to provide fuel for
metallurgical use, glassmaking, bakeries, etc.
For use in power generation with gas turbines, it is
necessary to provide the fuel at pressures of 10 atmos-
pheres or above. The only commercially available gasifier
operating at pressure with coal is the fixed-bed type such
as that offered by Lurgi. This system has several draw-
backs, not the least of which is its high cost. There are
359
-------
35
30
25
a: 20
I
$
~
---
U) 15
....J
....J
~
....J
UJ 10
~
u.
+ 9
~ 8
~ 7
0
6
5
4
3
4 5 6 7 8 9 10 15 20 30 40
CAPITAL MILLS/ KWHR
Figure 1. Preliminary ECAS results.
several more advanced pressurized gasifiers in various
stages of development. These have the potential for
more efficient and less costly production of fuel gas than
does the fixed-bed gasifier. As part of a program on
advanced COGAS systems carried out by UTRC under
EPA Contract 68-02-1099, evaluations were made of
several types of gasifiers including the Bureau of Mines
fixed-bed gasifier and the Bituminous Coal Research
(BCR) two-stage entrained flow gasifier. Because of the
potential attractiveness, the BCR will be emphasized
here.
A schematic of this gasifier with pertinent flows is
shown in figure 4. A comparison of the off gases from
this gasifier and the Bureau of Mines gasifier is given in
table 1. As can be seen in table 1, both fuel gases contain
quantities of HzS, cas, and NH3 which must be re-
moved before combustion if environmental regulations
are to be met.
GAS CLEANUP SYSTEMS
The fuel gas coming from the gasifier must be clean-
ed not only to meet the environmental regulations, but
also to meet restrictions set by the gas turbine. The
latter are often more stringent than those for the envi-
ronment as can be seen in table 2. Here it is seen that
particulates must also be removed.
Today, there is a large number of commercially
available cleanup systems as has been pointed out by
prior speakers (ref. 4). These systems operate at tem-
peratures of 420 K or less, well below gasifier outlet
temperatures. A simple schematic of a typical low-
temperature cleanup system is shown in figure 5. Several
of these systems have nearly equivalent performance.
Consequently, a representative physical absorption
system, the Selexol process of Allied Chemical Corpor-
ation, was selected for further consideration.
360
-------
AIR
POWER
TURBINE
COMPRESSOR
TURBINE
ELECTR IC
GENERATOR
100MW
FUEL
STEAM
BOILER
---
ELECTRIC
GENERATOR
50MW
r
~
TO STACK
L -cc)-
CONDENSER
PUMP
Figure 2. Waste-heat COGAS system.
>
I
I
56
DISTILLATE FUEL OIL
CERAMIC VANES AND
CONVENTIONAL
COOLED VANES
COMPRESSOR
PRESSURE
RATIO
*-
I
UJ
U
z
«
2:
c::
o
LL
c::
UJ
CL
Z
o
I-
«
I-
C/J
C/J
«
t9
o
U
52
40
48
- -
,"""'.......- --- 24
". "..., - -:::- - - - 20
,,,./,,.""'" ,,-
7 ...., - - - -- -- - 16
// //" ...,,- --
~// " ----------12
~ ---
--- 8
1480 -1590 -1700 1810
TURBINE INLET TEMPERATURE, K
44
40
250
550
350
450
650
750
NET POWER PER UNIT AIRFLOW - kW/kg/sec
Figure 3. COGAS station performance.
361
-------
COAL
28.4 MJ/kg
(12,200BTU/lb)
TRANSPORT GAS
422° K (300° F)
3.83 MN/m2
(555 psia)
0.410 m3/kg
(6.56 scf / Ib COAL)
AIR
7000K (800°F)
3.55 MN/m2
(515 psia )
0.193 m3/kg
sef / Ib COAL)
SLAG HOPPERS
5.499 MJ/m3
(147.6 BTU/scf)
MOL. WT ~ 24.240
FUEL GAS
HHV ~ 5.287 MJ/m3
(141.9 BTU/scf)
MOL. WT. ~ 24.519
4.83m3/kg
(77.3 scf/lb COAL)
GAS
GASIFIER STAGE II
1,2550K (1,8000F)
3.14 MN/m2
(455 psia ,)
CHAR HOPPER
STEAM
7830K (9500K)
6.31 MN/m2
(915psia)
(O.57kg/kg COAL)
QUENCH WATER
SLAG (0.087 kg/kg COAL)
Figure 4. BCR entrained flow gasifier.
The use of the low-temperature cleanup system,
while meeting all turbine and environmental restrictions,
results in some degradation of performance because of
the cooling of the fuel gas. By cleaning the gas at higher
temperatures, at or near gasifier outlet for example,
much of the sensible heat can be utilized in the cycle at
maximum efficiency. Unfortunately, there are presently
no high-temperature sulfur removal systems in more
than simple bench-scale or rudimentary demonstration
stages. One of the potentially more attractive methods
of high-temperature cleanup involves the use of half-
calcined dolomite as a sulfur remover. Considerable
work on this process has been done by the Consolidated
Coal Company, a part of the Continental Oil Corpor-
ation, and the CONOCO Process has been selected for
presentation here.
It should be noted that in addition to sulfur re-
moval, the low-temperature systems will also remove
nearly all the soluble nitrogen compounds such as
ammonia, cyanide, pyridine, etc. This necessitates an
362
-------
Table 1. Fuel gas characteristics
Gasifier Type
Fixed-Bed*
Temperature, K (F)
Pressure, MN/m2 (psia)
811
3.45
( 1000 )
( 500)
Volume %
47.61
20.55
5.88
13.83
2.76
0.60
0.10
0.25
8.42
100.00
Tar, kg/kg Coal 0.11
HHV, MJ/m3 (8tu scf) [Tar Free] 5.212 ( 139)
N2
CO
C02
H2
CH4
H2S
CDS
NH3
H20
Sensible Heat/Coal HHV, %
12.2
*
No tar recycle.
363
Two-Stage
Entrained-Flow
1255
3.45
( 1800 )
( 500)
47.70
16.74
8.84
11.98
3.14
0.46
0.10
0.38
10.46
100.00
0.0
4.687 (125)
24.6
-------
Table 2. Fuel gas cleanup requirements
Low-Btu Gas
Typical Current Spec
Sulfur
0.05 Mol % or Less
Than Amount to Form
0.6 ppm Alkali Metal
Sulfate
<1.0 Mol % or Less Than
Amount to Form 5 ppm
Alkali Metal Sulfate
Particulates
4 ppm Weight or
0.0012 gr/ft3 > 2 ~
3
30 ppm or 0.01 grift
Metals
Vanadium
< 0.003 ppm Weight
See Sulfur Spec
< 0.02 ppm Weight
< 0.6 ppm
Nitrogen
500 ppm as NH3
TREATED
GAS
RAW
GAS
FEED
STEAM
RICH
SOLVENT
LEAN SOLVENT
Figure 5. Typical low-temperature acid gas removal unit.
364
-------
ammonia removal system with additional utility loads.
The high-temperature systems do not remove ammonia,
or at least have not been documented to be capable of
reviewing ammonia or other nitrogen compounds. Thus,
an appreciable amount of fuel-bound nitrogen would be
contained in the power system fuel.
INTEGRATED POWER SYSTEMS
To ensure that the highest performance results from
coupling the coal processing and power systems, a very
careful integration of components must be carried out.
These are gasifier requirements for high-pressure air and
steam which can be supplied from the power system;
similarly, some of the cooling of the fuel gas prior to
low-temperature cleanup could supply heat for use in
the power system. Somewhat simplified schematics of
CLEAN GAS TO BURNER
BOOST COMPRESSOR
GAS TURBINE
GAS
TUABINE
EXHAUST
COAL 8-
GAS
SCA GASIFIER
STEAM CYCLE
COOLING
TOWER
PROCESS
COOLING
WATER
the integrated systems based upon the SCR-type gasifier
with low- and high-temperature cleanup are shown in
figures 6 and 7.
The performance of the integrated systems is very
much a function of gasifier type and turbine operating
characteristics. For systems based upon fixed-bed gasi-
fiers, a turbine inlet temperature of 1477 K (2200 F)
was selected. Such an integrated system could be avail-
able in the late-1970 decade. Performance for this first-
generation system with both low- and high-temperature
cleanup is shown in figure 8 where it is compared to the
performance of the distillate-fired COGAS system. The
performance penalty associated with the low-tempera-
ture system is due to both thermodynamic penalty
associated with cooling the gas and also to the relatively
high utility loads of the cleanup system. The high-
temperature cleanup process used in this system is based
CHAR
TOWATER
SCRUBBER
f;
CLEAN FUEL
STACK GAS
SULFUR
SULFUR RECOVERY
Figure 6. Integrated COGAS/BCR/selexol system.
365
-------
CLEAN GAS TO BURNER
INLET
AIR
BOOST COMPRESSOR
COAL & GAS
WATER
COOLING
TOWER
PROCESS COOLING WATER
H.P.STM.
OOLOMITE
ABSORBER
OOLOM ITE
REGEN.
OOLOMITE
MAKE-UP
SPENT
OOLOMITE
H2S
OOLOMITE
SLUOGE
C02 & STE AM
TRANSPORT
GAS
Figure 7. Integrated COGAS/BCR/CONOCO system.
upon an iron-oxide process operating at gasifier outlet
temperature of 810 K (1000 F). Because of the need to
convert the S02 resulting from the iron-oxide process to
elemental sulfur (the preferred form), about 8 percent of
the fuel gas must be used resulting in a significant per-
formance penalty.
With the BCR-based systems, a 1700 K (2600 F)
turbine temperature was assumed. This temperature is
the target goal of an ERDA program to be initiated in
early 1976 to develop second-generation gas turbines for
utility use. Performance of second-generation integrated
systems with both low- and high-temperature cleanup is
shown in figure 9. The efficiency of the integrated
system with the low-temperature cleanup is comparable
to conventional steam stations with flue gas desulfuri-
zation (FGD); the systems with high-temperature clean-
up are appreciably better.
SYSTEM ECONOMICS
Because of the lower capital costs of the power
system, the cost of the gasification and cleanup systems
are partially offset when compared to a conventional
steam station with FGD. A comparison of the costs of
power for the various systems is shown in figure 10.
Here it is seen that the first-generation systems (fixed-
bed gasifier, 1477 K [2200 F] turbine) are marginally
competitive with the steam station. However, second-
generation systems (entrained-flow gasifier, 1700 K
[2600 F] turbine) are more attractive with the potential
366
-------
TURBINE INLET TEMPERATURE = 2200F
CONVENTIONAL COOLING
45
2~
PR = 32/ DIS;;L8LATE FUEL OIL
>
:I:
:I:
40 ~
?fi
>-
U
Z
w
U
LL
LL
W
35 ~
16
~
3~~ I' 'U MINES/IRON OXIDE
~ TB BU MINES/SELEXOL
I _1
30 I-
50
100
150
200
250
SPECIFIC POWER - KW/LB/SEC
Figure 8. Performance estimate for first generation systems.
TURBINE INLET TEMPERATURE = 2600F
CERAMIC VANES, CONVENTIONAL BLADES
32 24
50 PR = 40
> DISTILLATE OIL
:I:
:I: 45
?fi
>- 8
U
Z
w
U
LL
LL 40 BCR/CONOCO
w
8
BCR/SELEXOL
35 B
150 200 250 300 350
SPECIFIC POWER - KW/LB/SEC
Figure 9. Performance estimate for second generation systems.
367
-------
1..j!:1 ,
BU MINES SELEXOL
(FIRST GENERATION)
CI:
I
S
~
---
(/)
...J
...J
~
I
f- 30
(/)
0
U
Z
0
f-
«
CI:
w
Z
w
19
CI:
w
S
0
a..
20
a
I
a
l CONVENTIONAL
f COAL WITH
STACK GAS
CLEANUP
/
BU MINES/IRON OXIDE
(SECOND GENERATION)
BCR/SELEXOL
(FIRST GENERATION)
0.50
1.00
CQAL COST $/GJ
2.00
0.50
1.00
$/MILLION BTU
1.50
I
2.00
1.50
Figure 10. Busbar power cost.
of generating electricity at costs some 15-20 percent
lower than the steam station.
ENVIRONMENTAL ASSESSMENT
The foregoing has served to identify the perform-
ance and economic potential of integrated power
systems. An equally important aspect is the environ-
mental intrusions such a system might have.
The primary guideline used in establishing the goals
for the air emissions was that the overall plant would
have to meet the regulations for large coal-fired power
stations; i.e.,
80z 0.52 kg/GJ (1.2Ib/million Btu)
NOx 0.30 kg/GJ (0.7 Ib/million Btu)
Particulates 0.04 kg/GJ
(0.1 Ib/million Btu).
Both low- and high-temperature cleanup processes could
easily achieve the SOz goal with the low-temperature
processes being able ,to go to 100 ppm of Hz S in the fuel
gas quite readily. The high-temperature processes varied
between 100 ppm and 500 ppm of HzS~ again easily
within the 'limits. Emissions from the stack of the Claus
plant associated with each cleanup process were appre-
ciably higher than from the power system stacks, but the
sum was still within the limits (see table 3 for the various
368
-------
Table 3. Estimated environmental intrusion of integrated power systems
(All values in kg/GJ coal unless otherwise noted)*
Water Therma111 S02 NOx Particulate Solids SulfurY AmmoniaY
BOM/Se1exo1 3/ ~~~~:~ . 1 38&1 .153 6.28 1.27 0.59
94.04/ ~0048/
1477 K COGAS 12.~ .038 .075
TOTAL 106.0 0.005 .176 0.228 .004 6.28 1.27 0.59
BOM/Iron Oxide 12~Q!I ~~~~~~ 0.334&1 .137/ ~o327 6.28 1.08
1477 K COGAS .145 1.41-
Total 12.0 0.018 .489 1. 54 .03 6.28 1.08
BCR/Se1exo1 3/ ~~~~~Y .209 .129 3.05 1.18 0.41
224.04/ ~004W
1700 K COGAS 10.~ .034 .005
Total 234.0 0.016 .243 .134 .004 3.05 1.18 0.41
t.) BCR/CONOCO 3/ ~~~~:Y .014 3.05 1.17
C) 32.04/ 1~34V ---97
co
1700 K COGAS 9.~ .224 .02-
Total 41. 0 0.002 .238 1. 34 .02 3.05 1. 17
*
1 kg/GJ = 0.430 1b/mi11ion Btu.
lIGJ's rejected per GJ's coal.
YRecovered with no credit for subsequent sale.
1ISee Table 4 for water characteristic~
1/Inc1udes boiler makeup and cooling tower losse~
YMechanica1 draft cooling tower losses.
&lInc1udes emissions from Claus plant. A 90%+ reduction could be obtained with
ZlSum of thermal and fuel-related NOx.
~Assumed to be equal to methane-fired systems.
~Sum of carryover and combustion products.
tail-gas cleanup.
-------
emissions). Additiollal treatment of these stacks could
further reduce the S02, but at a cost and performance
penalty.
The emissions of NOx' however, were not as amen-
able to reduction. The low-temperature processes re'-
moved essentially all the fuel-bound nitrogen, thus only
the atmospheric NOx (that resulting from the combina-
tion of atmospheric nitrogen with oxygen during com-
bustion) was formed. Because of the low flame temper-
atures associated with the low-calorific value fuel gas,
these emissions are well below those for other fuels. Un-
fortunately, the fuel-bound nitrogen compounds pass
throuah the high-temperature cleanup processes and, at
the concentrations involved, are 90 percent or more con-
verted to NOx' Thus, these emissions from the integrat-
ed system are far above the prescribed value as can be
seen in table 3.
The particulate story is somewhat analogous to that
for NOx' Aqueous scrubbing using available scrubber
technology in the low-temperature systems removed
essentially all the particulates. There are, however, no
developed high-temperature particulate removal devices
in commercial service capable of removing small « 2 J1)
particles at the required efficiency. However, it can be
postulated that cyclones, followed by granular beds or
fabric filters made of high-temperature material, would
be capable of 99+ percent removal of particulates to the
2J1 level and 90 percent or more in the <2J1 to submi-
cron levels (e.g., see ref. 5). Thus, all the integrated
systems considered were assumed able to meet the regu-
lation for coal-fired stations.
In addition to the airborne emissions, there are a
number of sources of water-borne and solid wastes from
the coal processing system. Perhaps the worst offender
would be the fixed-bed gasifier and low-temperature
cleanup since, in that instance, a significant amount of
coal tars would be present. The chemical characteristics
of the process-.£ondensates from such a system are given
in table 4. According to reference 4, all of the water
emissions from the integrated system can be treated so
there will be no release of water-borne pollutants and
the system could be operated in compliance with the
effluent controls currently mandated for 1983 with
respect to the Best Available Technology Economically
Achievable.
SUMMARY
The integration of coal gasification, fuel gas cleanup
and combined-cycle power systems could be realized by
the end of this decade using technology currently avail-
able or in late stages of development. While this first-
generation system would be only marginally attractive in
Table 4. Characteristics of process
condensate from Bureau of Mines/
selexol process
(All values in ppm except pH)
Component
Total Ammonia
Total Sulfur
Phenol
Thiocyanate
Cyanide
Fatty Acid
Chloride
Carbonate
COD
80D5
Sulfi des
Heavy Metals
pH
200-400
500
100
1-10
500
250
2500
10-100
1 0- 20
9
an economic sense, its operation would provide the
necessary experience which would lead to higher perfor-
mance systems.
The second-generation integrated systems which
could appear in the mid-1980's would offer higher effi-
ciencies and, possibly, lower capital costs than conven-
tional steam stations with FGD. This would result in
lower generating costs, perhaps by 20 percent. With con-
tinued research into methods of removing or combusting
fuel-bound nitrogen compounds, these systems would
also have considerably less environmental intrusion than
the steam stations.
REFERENCES
1.
P. F. H. Rudolph, "New Fossil-Fueled Power Plant
Process Based on Lurgi Pressure Gasification of
Coal," Proceedings of the Division of Fuel Chemis-
try, American Chemical Society, May 1970.
F. L. Robson and A. J. Giramonti, "An Advanced-
Cycle Power System Burning Gasified and Desul-
furized Coal," Proceedings of the First Seminar on
the Desulphurization of Fuels and Combustion
Gases, Geneva, Switzerland, November 1970.
L. Shure, "Energy Conversion Alternatives Study
(ECAS), Comparison of Task I Results by NASA-
Lewis Research Center," May 27,1975.
2.
3.
370
-------
4.
C. Colton, "Low- and Intermediate Btu Fuel Gas
Cleanup," Environmental Aspects of Fuel Conver-
sion Technology II, Hollywood, Florida, December
1975.
371
5.
A. K. Rao et ai., "Particulate Removal from Gas
Streams at High Temperature/High Pressure,"
EPA-600/2-75-020, August 1975.
-------
NO CONSIDERATIONS IN ALTERNATE FUEL COMBUSTION
x
G. Blair Martin*
INTRODUCTION
I n the search for energy supplies, the United States
is projected to place heavy reliance on coal which is the
most abundant fossil fuel available. There are many
methods of extracting the energy from coal being pur-
sued; however, the ultimate decisions on the path(s) to
be followed depend on both economic and environmen-
tal considerations. These considerations cover the full
range from resource extraction through processing to
end utilization. On the economic side it is necessary to
include not only capital and operating costs but also the
overall energy efficiency of the process. On the environ-
mental side, there are potential impacts in every step and
the overall effect on air, water, and land quality must be
assessed. For the purposes of this paper, only the end
use processes (i.e., combustion systems) will be consid-
ered.
The ways in which coal can be used in an environ-
mentally acceptable manner are dependent on the type
of combustion source. The pollutant which must be con-
trolled includes sulfur oxides, nitrogen oxides, carbon
monoxide, unburned hydrocarbons, and total particu-
late. Perhaps the most options exist for utility genera-
tion of electric power. One option currently being used
is the direct combustion of coal with stack gas cleaning
for sulfur oxides and particulate and combustion modifi-
cations for control of nitrogen oxides, carbon monoxide,
and unburned hydrocarbons. Improvement of the exist-
ing technology is being pursued by a variety of EPA
programs. A second option is the conversion of coal into
low sulfur fuels to be utilized in conventional steam
boilers or combined cycle plants. The large energy losses
currently associated with fuel cleaning processes appear
to require use of the advanced design combined cycle
with integrated gasifier to achieve energy efficiency com-
parable to the first option. Major unknowns in these
designs are the criteria for minimizing nitrogen oxides
and other combustion related pollutants. The third op-
tion is the use of fluidized bed combustion to minimize
sulfur oxides and other pollutants. The area of sulfur
oxides control was the subject of a recent conference
(ref. 1) and many excellent papers were prepared. One
particularly pertinent summary on fuel conversion proc-
*G. Blair Martin is with the Combustion Research Branch,
Industrial Environmental Research Laboratory, U.S. Environ-
mental Protection Agency, Research Triangle Park, North Caro-
lina.
esses by Spaite (ref. 2) contained the following conclu-
sion:
"The competitiveness of low and intermediate
Btu gas systems for power plant application appears to
depend on a high level of success in development of
improved gas turbines, which can be integrated into
advanced high efficiency power cycles which com-
pensate for losses suffered in conversion of coal to
clean fuels. . . ."
For other applications, such as residential and com-
mercial heating, low-sulfur, high-Btu fuels will be re-
quired. Candidates under consideration include synthetic
natural gas (SNG), synthetic oils, and alcohol fuels.
This paper will concentrate on combustion and
emission characteristics of alternate fuels, an area where
information is still very limited. The topics to be covered
include a summary of the most recent combustion data
for both boiler and gas turbine systems and a presenta-
tion of some concepts that may have potential for con-
trolling the formation of NOx and other pollutants in
the combustion process.
BACKGROUND
A wide range of subject matter relates directly to
combustion of alternate fuels. The topics include pollut-
ant formation mechanisms, applicable emission control
techniques, fuel characteristics, and end use equipment
type. Since these areas have been treated in detail for
alternate fuels previously (ref. 3), the background pre-
sented below is a brief general summary. The most
recent information on combustion and emission charact-
eristics is summarized.
Pollutant Formation Mechanisms
The mechanisms of formation of NOx have been
extensively discussed (refs. 4,5,6,7); however, a brief
summary is in order. (Note: Subsequent discussion will
be limited to nitric oxide (NO) which is the primary
form of nitrogen oxides found in the flue gas of conven-
tional combustion equipment.) The mechanisms for
formation of NO are as follows:
1. Thermal NO is formed from fixation of atmos-
pheric nitrogen by Zeldovitch reactions which have a
strong temperature dependence.
2. Fuel NO is formed by oxidation of chemically
bound nitrogen in the fuel by reactions with a low tem-
perature dependence but a strong oxygen availability
dependence.
373
-------
There is also recent experimental evidence (ref. 8)
to show that nitrogen specie (e.g., NH3 and HCN) can be
synthesized in fuel rich flames as postulated by Feni-
more (ref. 9) and subsequently oxidized to NO as is fuel
nitrogen.
Other pollutants such as carbon monoxide, unburn-
ed hydrocarbons, and carbon particulates are normally
low in well designed conventional systems.
Emission Control Techniques
The basic combustion modification techniques for
NOx control can be summarized as follows:
1. Diluent addition to reduce flame temperature is
accomplished through the use of either water or recycled
flue gas added to the combustion air;
2. Staged combustion is based on operation of
burners at a fuel rich condition with delayed secondary
air addition to complete heat release, thereby limiting
both peak flame temperatures and primary zone oxygen
avai lability;
3. Burner modifications involve changes in fuel
and air mixing conditions to promote localized fuel rich
conditions and/or combustion gas recirculation;
4. Novel techniques, such as catalytic combustion,
may allow NO emissions lower than those achievable for
combustion of clean fuels in conventional systems and
may be particularly applicable to redesign for maintain-
ing system efficiency.
The first technique controls only thermal NO, while
the last three may also control fuel NO.
The emissions of the products of incomplete com-
bustion (carbon monoxide, unburned hydrocarbons, and
carbon particulate) are subject to being increased as NOx
is decreased past a critical point for fixed system design;
however, there is a body of evidence that indicates that
these emissions can be controlled if the system is de-
signed or modified with NOx control requirements in
mind.
Fuel Characteristics
The fuels of concern are clean liquid and gaseous
fuels which can be derived from coal and for which some
combustion and emission information is available. The
fuels to be considered are discussed below and properties
are shown in tables 1 and 2.
Table 1. Comparison of various gaseous fuels
Higher Gas Composition, Mole %
Heating Value
Type of Fuel KJ/M3 (Dry) CO H2 CH4 C2H6 C02 N2 H20
Texarkana Natural Gas a 36,083 0.0 0.0 96.0 3.2
0.0 0.8 0.0
Cleveland Natural Gas a 42,200 0.0 0.0 80.5
18.2 0.0 1.3 0.0
Synthetic Natural Gas
Koppers-Totzekb 35,896 0.08 2.5 94.2 0.0 0.4 2.8 0.0
Synthetic Natural Gas
Lurgib 36,586 0.0 0.8 96.8 0.0 1.2 1.1 0.0
Medium-Btu Gas 11,120 50.4 33.1 0.0 0.0 5.6 0.0 9.6
Koppers-Totzekb
Mediu~-Btu Gas 11,269
Lurgi 9.2 20.1 4.7 0.5 14.7 0.0 50.2
Low-Btu Gas-Lurgib 6,717 13.9 19.6 5.5 0.0 13.3 37.6 10.1
Low-Btu Gas-Winkler b 4,403 19.0 11.7
0.5 0.0 6.2 51.1 11.5
Low-Btu Gas-l!ygas b 8,806 13.5 16.6 8.4
0.6 12.7 28.9 18.2
Low Gas-V-Gas b 5,597 17 11. 6 4.1 45.4
8.8 12.0
aHandbook of Chemical Engineering, John H. Perry--Editor, 3rd edition.
bit II Institute of Gas Technology, September 10-14, 1973
Clean Fuels from Coal-Symposium Papers,
374
-------
Table 2. Comparison of various liquid fuels
lIigher Ultimate Analysis, Weight %
Heating Value
Fuel KJ/KG C H N S 0 Ash
Distillate Oil a 44,892 86.9 13.1 0.023 0.096 0.0 < 0.002
Residual Oila 44,147 86.8 12.5 0.22 0.89 0.0 0.03
COED Crudeb 83.0 8.4 1.1 0.35 7.15 0.0
COED b
Hydrotreated 88.0 11. 6 0.01 0.01 0.38 0.0
Synthoilb 39,077 89.6 7.6 0.9 0.31 1.6 (1. 3)
H-Coa1 Crudeb 9.5 0.68 0.19
H-Coal Low Sulfur 8.4 1.05 0.43
Toscoa1 Oilb 37,681 80.7 9.1 0.7 0.2 9.4 0.2
Shale Oil 43,496 84.6 11. 7 1.77 0.76 0.01
Alcohol Fuel 22,278 37.5 12.5 0.0 0.0 50 0.0
aCRB in-house fuels
b"C1ean Fuels from Coal-Symposium Papers," Institute of Gas Technology, Sept. 10-14, 1973.
High-Btu Fuels. These fuels. are the type which
would be potentially economical. 'to transport for use at
points remote from the coal conversion sites. Fuels of
this type include Synthetic Natural Gas (SNG), alcohol
fuels, synthetic coal liquids, and shale oil, for which a
number of processes for production are being investi-
gated. These SNG, alcohol fuels, and light synthetic oils
can be considered as supplements or replacements for
natural gas and light fuel oils and may be used predom.
inantly in combustion equipment currently burning
these fuels (e.g., residential, commercial, and small
industrial boilers). SNG is produced to be similar to
natural gas. The alcohol fuels will be predominantly
methanol possibly with up to 10 percent higher alcohol.
The heavy synthetic oils may be used as boiler fuels or
may be distilled to produce light fuels suitable for tur-
bines or residential furnaces.
Medium-Btu Gas. This is the product of oxygen
blown gasifier systems and is currently projected pri-
marily as a synthesis gas for production of SNG and
methanol. Its use as a boiler fuel may depend more on
economics than technical considerations.
Low-Btu Gas. Although identified by the single
term 10w.Btu gas (LBG). the actual nature of the fuel is
as varied as the airblown gasification systems that can be
used to produce it. It is projected primarily as a fuel
utilized in utility boilers or by industrial parks, either of
which could be located close to the gasification site, as
long distance transportation is not economical. The
actual utilization projected for the lBG depends on the
time frame. In the near term it may be used in conven-
tional utility boilers and gas turbines either as retrofit to
existing systems or integrated new designs. However, in
the long run, combined cycles based on supercharged
boilers or high temperature gas turbines as the primary
combustion device may predominate. To increase the
efficiency of advanced design gasifier systems, there is a
strong interest in high temperature H2S removal proc-
esses to provide a clean fuel gas at 810 to 1090K. Again,
both economics (capital cost and energy efficiency) and
technology may come into play in the final decision.
Trace Constituents. The fate of the sulfur is fairly
well considered in most studies of medium- and low-Btu
gases, and removal of H2 S and cas is well documented.
On the other hand, the fate of the nitrogen content of
the coal has received less attention. Farnsworth (ref. 10)
shows the final concentration of bound nitrogen specie
in the medium-Btu gas after low-temperature cleanup
from the Koppers-Totzek process to be less than 1 ppm.
On the other hand, there is considerable evidence of the
375
-------
presence of large amounts of NH3 in the product gases
(ref. 11). Robson estimates up to 600 ppm of NH3 in a
low-Btu gas with a low-temperature Hz S cleanup and
3,800 ppm from high-temperature processes under devel-
opment (ref. 12). It is not clear that these nitrogen
compounds could be present in the high-Btu fuels; how-
ever, it is quite possible that they be present in the low-
and medium-Btu gaseous fuels. Finally, many coal con-
version systems producing gaseous fuels also form signifi-
cant amounts of heavy organic tars, and there is some
indication that some of this material may be present in
the fuel gas after cleanup. This tar may also contain
bound nitrogen compounds (ref. 13).
COMBUSTION AND EMISSION
CHARACTERISTICS OF FUELS
Each of the candidate fuels has its own distinct
combustion characteristics, pollutant emission potential,
and operating constraints. Since these fuels have only
recently been considered for widespread application to
stationary combustion sources, there is a scarcity of
experimental combustion and emission information and
these characteristics have not been documented for any
practical system operating on these fuels. The most
recent available data are summarized below and related
information for each fuel is discussed.
Synthetic Natural Gas. The composition of Syn-
th~tic Natural Gas (SNG) is normally shown as closely
approximating that of natural gas (i.e., principally
methane) with all trace contaminants (e.g., HzS, NH3)
removed. If this is the case, combustion and emission
properties are well, known and thermal NO control tech-
nology for utility boilers is well established.
Alcohol Fuels. The alcohol fuels under considera-
tion are predominantly methanol or mixtures of metha-
nol and higher alcohol. Limited results on combustion of
"Methyl Fuel" in an unspecified system as reported by
Duhl (ref. 14) show NO levels to be about a factor of 4
less than those for methane. Duhl has subsequently
reported on methanol combustion in a 50 MW utility
boiler and shows only minor differences in emissions of
methanol (about 60 ppm NO) and natural gas (about 75
ppm) at comparable conditions (ref. 15). The author has
presented a summary of inhouse and contract research
on methanol combustion (ref. 16) and the results can be
summarized as follows:
1. For any given condition alcohol fuels produce
lower emissions of NOx than distillate oil, methane, or
propane.
2. The NO emissions of alcohol fuels increase as
the percentage of higher alcohols increases.
3. The low NO emissions for alcohol fuels appear
to be a function of the presence and level of oxygen in
the fuel molecule, which can be viewed as a diluent car-
ried in the fuel. The operative mechanism appears to be
related to thermal effects of the fuel, latent heat of
vaporization, and/or decreased flame temperature; how-
ever, chemical effects cannot be totally ruled out.
4. The magnitude of the difference between
methanol and any conventional fuel is a characteristic of
system design; however, the difference appears to be
smaller for cold wall systems.
5. The emissions of CO, UHC, and particulate
from alcohol fuels were generally the same as, or less
than, those for the conventional clean fuels tested (pro-
pane and distillate oil). Aldehydes do not appear to be a
significant problem based on limited data.
6. From a technical standpoint, methanol appears
to be a satisfactory fuel for stationary combustion sys-
tems; however, the final commercial feasibility will prob-
ably depend more on cost and availability than on the
ability to burn the fuel.
Synthetic Oil Products. The data on combustion of
synthetic oils derived from coal and shale containing
high levels of chemically bound nitrogen (>1 percent)
are very limited; however, some data on combustion of
lower nitrogen « 0.5 percent) synthetic oils have been
reported by Haebig for a small atmospheric pressure
burner (ref. 17) and by Hardin for a gas turbine
combustor (ref. 18).
Haebig used a small firebrick chamber, fired by a
modified flame retention residential oil burner with an
air atomizing nozzle. The heat input rate was a nominal
one ml per second of oil. Emissions from three
coal-derived fuels were measured and compared to
conventional distillate oil. The coal-derived oils consisted
of one distilled product with 0.31 percent nitrogen and
two full range filtered crudes with about 0.5 percent
nitrogen. The trends are similar to those previously
reported for doped distillate by Martin (ref. 19) and for
petroleum derived residual oils by Turner (ref. 20). One
particularly significant point is that the distillate from
the full range coal liquids contained 0.31 percent
nitrogen, apparently indicating that the nitrogen
compounds are considerably lower boiling fractions than
those found in conventional crudes where Haebig found
only 0.01 percent N in a comparable range distillate. At
low excess air levels significantly higher fractional
conversion of the nitrogen to NO was observed for the
0.3 percent distillate than for the 0.5 percent N full
range crude. It was also observed that when the
petroleum-based distillate oil was doped to 0.5 percent
N with pyridine, the fractional conversion was more
similar to the coal distillate than the full range coal oil.
Both of these observations may relate to the volatilit'( of
376
-------
the nitrogen compounds and the ability to control the
formation of fuel NO from synthetic oils through
cOlT)bustion control as discussed by the author
previously (ref. 3).
Hardin reported the results of combustion of two
COED oils (a light fraction and a blend of 20 percent
light and 80 percent heavy fraction), of Sea Coal and of
JP-4 and 5 fuels in a T63 gas turbine combustor can. The
fuel nitrogen contents for the synthetic oils ranged from
0.15 percent for the 20/80 blend to 0.4 percent for the
Sea Coal and total NOx emissions were reported to
increase with increased fuel nitrogen. It was also con-
cluded that combustor modification for fuel NO control
did not appear promising; however, no specific attempts
at modification are discussed. The conclusion is even less
_~urprising if the combustor operated at the very fuel lean
70
conditions stated (i.e., F/A of 0.0117 to 0.022) and with
a lean primary combustion zone. The fuel analysis shows
0.19 percent in the light product and 0.13 percent in the
heavy product, indicating that for the distilled COED oil
the nitrogen also concentrates in the lighter fractions.
Low-Btu Gas. The combustion data on these fuels
are also quite limited; however, some additional data
have also become available for both boiler and gas tur-
bine type combustors.
The data from the EPA contract with the IFRF
have been presented previously (ref. 3); however, to
place some subsequent data in context it is appropriate
to show these data again. Figure 1 indicates that the
simulated low-Btu gas at 298 K and combusted with 105
percent of theoretical air at 573 K resulted in an NOx
emission of less than 4 ppm.
50
o
UJ
a:
::>
en
<{
UJ
::iE 40
en
~
E
c-
c-
UJ'
o
X 30
o
S::
a:
I-
z
67 PERCENT NATURAL GAS - 33 PERCENT LOW Btu GAS
20
10
33 PERCENT NATURAL GAS - 67 PERCENT LOW Btu GAS
----------------
o
o
2
100% LOW Btu GAS BELOW DETECTION LIMIT
10
4
6
8
RELATIVE SWIRL INDEX, Rs
Figure 1. Nitrogen oxide emissions as functions of combustion
air swirl and fuel gas composition.
377
-------
Lisauskas (ref. 13) used a small refractory furnace
to burn the product gas from a pilot scale fixed-bed
gasifier. The burner was a scaled down version of a bluff
body stabilized, register type boiler burner. The fuel gas,
from 400 to 1,900 ppm NH3 was dependent upon the
gasifier operating conditions, and was delivered to the
burner at temperatures between 305 and 316 K with the
organic tar removed in most cases. In one case the NH3
was scrubbed to the range 11-43 ppm in the fuel gas
before combustion to establish the thermal NO baseline.
A surprising result is that the baseline level with low
temperature fuel and no air preheat is stated to be
90-140 ppm of thermal NO, which is a marked contrast
to the IFRF experience. Based on this thermal NO value
and the subtraction technique for determination of the
fuel NO, the conversion to NO decreases from 65
percent at 240 ppm NH3 in the fuel gas to 23 percent at
1,900 ppm NH3.. Again the trends are generally similar
to those observed for oils (refs. 19 and 20). Although
some points in the report by Lisauskas raise questions in
the author's mind, this is a particularly valuable
contribution of data on combustion of a real
coal~derived low-Btu gas containing NH3.
Pillsbury (ref. 21) has presented results from firing
simulated low-Btu gases with heating values from 4,625
to 10,000 KJ/m3 in subscale and full-size gas turbine can
combustors of similar design. The gases were low tem-
perature, bound nitrogen free blends of pure gases. The
results show 5 to 25 ppm NO over the combustor exit
temperature range of 920 to 1,160 K as compared to 30
to 60 ppm for natural gas over the same range of exit
temperatures. The CO was found to be greater than for
natural gas in the subscale, but considerably reduced in
the full size combustor. These results are similar to those
previously reported by Klapatch (ref. 22), although it is
unclear in both cases on what basis the NOx measure-
ments are reported (i.e., ppm as measured, corrected to
some O2 level, etc.).
DISCUSSION
Based on the foregoing summary of available com-
bustion data on alternate fuels, it is obvious that more
intensive experience with the actual fuels is still re-
quired. A number of systematic studies need to be con-
ducted with specific end use systems in mind. The
author's opinion is that efficient, environmentally
acceptable use of these fuels may necessitate abandoning
or substantially modifying current conceptions of the
combustion systems. With this in mind, it is possible to
propose some concepts that can potentially form the
basis for combustor design experiments and to postulate
their success in NOx control based on existing knowl-
edge and tools. The concepts for combustioh of low-Btu
gas are discussed in general for boilers and gas turbines
and then in detail for gas turbines. The significant prop-
erties of synthetic oils are discussed in general.
Liquid Fuels
The same general concepts discussed above should
be applicable to combustion of high nitrogen synthetic
liquid fuels in boilers and turbines. As previously dis-
cussed, the key feature for liquid fuels may be the way
in which the fuel nitrogen compounds are bound in the
fuel. For conventional petroleum derived heavy fuels,
the degree of NOx control with staged combustion has
!1enerally been limited to about 50 percent by the onset
of unacceptable smoke formation in the experiments
performed to date. It should be noted that some recent
results have shown promise for better control by opti-
mizing the burner. On the other hand, experiments with
doped distillate oils, where the nitrogen is 100 percent
volatile, have resulted in an 80 to 90 percent reduction
(see figu re 2).
The limited data available on crude and distillate
fractions of several synthetic oils derived from coal sug-
gest that the nitrogen compounds may be more evenly
distributed over the boiling range of the oil (refs.
17,18,31). This can be seen by reference to table 3.
I n the case of a high nitrogen California petro-
leum crude, the distillate fraction had only about one-
tenth the nitrogen concentration of the parent crude.
For the three synthetic crudes listed the distillate fuels
have 0.6 to 1.45 times the nitrogen content of the
parent crude. Necessary information that is not given for
any of the fuels is the fraction of the crude that will
appear as a light distillate oil. This would allow a better
assessment of the "volatility" of the nitrogen com-
pounds in the crude. In any case the comparison indi-
cates that the synthetic oils have significantly more of
the nitrogen distributed into the lighter ends and, there-
fore, may be more amenable to NOx control by staged
combustion.
Low-Btu Gas
It now appears that low-Btu gases will be used pre-
dominantly as a low-sulfur fuel supply for utility power
generation. The processes used to convert the fuel gas to
electric energy will be either conventional steam power
plants or combined cycle systems. These combined cycle
systems can be either a supercharged boiler with a recov-
ery gas turbine or a fired gas turbine with waste heat
recovery boilers. The latter system appears to be favored
for advanced gas turbine capable of accepting higher tur-
bine inlet temperature (i.e., >1,590 to 1,640 K). The
378
-------
E
0.
0.
OVERALL STOICHIOMETRIC'RATIO = 1.4 "
1000 ~
,,/
I-TOTALNO
"
/
"
/
"
/
900
800
z'
~Ci] 700
<>:(j;
0:<>:
""CD
a:i~ 600
~C
00'
u,..; 500
~11
X~
0- 400
So!
0:
""
2 300
rI"" THERMAL NO
0_0-0-
~7 0--0----
0-
200
100
o
0.6
0.9 1.0 1.1 1.2
PRIMARY STOICHIOMETRIC RATIO
1.3
1.4
Figure 2. Reduction of fuel NO by staged
combustion.
overall system efficiency (i.e., Btu of coal input to Btu
equivalent of electricity output) is a function of the
product of the thermal efficiency of the gasification
system and the thermal to electrical efficiency of the
power cycle. The thermal efficiency of gasification sys-
tems using low temperature H2 S cleanup systems is in
the range of 60 to 70 percent. Combined with a 50
percent efficiency of an advanced generation turbine
(i.e., turbine inlet temperatures of 1,813 K) the overall
system efficiency is 30 to 35 percent which is less than
or equal to a modern coal-fired boiler with S02 removal.
However, Robson (ref. 23) estimates that this can be
raised to 85 percent by using high-temperature H2 S
cleanup processes, which yield a 42.5 percent system
efficiency with a high-temperature turbine. This points
very clearly to the need for consideration of the
combustion and emission problems associated with a
high temperature fuel gas which may also contain NH3.
The possible combination to be considered is a
complex matrix of fuels (i.e., high and low temperatures,
with or without bound nitrogen) and combustion sys-
tems (i.e., gas turbines or conventional and supercharged
boilers). For the purposes of this discussion, the general
types of control strategies for both boilers and turbines
Table 3. Comparison of fuel nitrogen
content of crude and distillate
fractions of synthetic and petro-
leum crudes
Fuel
Wilmington,
Ca 1 iforn i a,
oil (ref. 32)
Coal derived
( ref. 1 7)
COED coal
derived
(ref. 18)
Shale derived
(ref. 31)
Nitrogen content %
Crude Distillate
0.65
0.07
0.5
0.31
0.13
0.19
2.05
1.60
are presented below, then one particular set of cases is
chosen for more detailed analysis.
Boiler Concepts. For boiler appl ications a key com-
mon feature, which gives a great flexibility in designing
for emission control, is the ability to remove heat from
the combustion gases in a controlled manner and to use
that heat in an efficient manner for steam generation.
For instance, with a high-temperature fuel gas, the sen-
sible heat could be removed prior to combustion there-
by significantly reducing thermal NO. The syst~ms do
require operation with overall excess air as close to
stoichiometric as possible to maximize thermal effi-
ciency and this limits the control options somewhat.
Two schemes for application of combustion modifica-
tion to boilers (either supercharged or conventional) are
shown in figure 3 and are discussed below:
1. Flue gas recirculation as represented by figure
3A is based on removal of relatively cool combustion
gases from the flue for mixing with the incoming com-
bustion air. This reduces the peak flame temperature and
controls thermal NOx' The technique does not affect the
total mass of combustion gases to the stack and, if the
heat transfer is properly designed, also does not affect
stack temperature. As a result the only efficiency impact
is the FGR fan power which may be counterbalanced to
a degree by an ability to operate at lower excess air.
2. Staged combustion as represented by figure 3B
is based on a rich primary combustion zone to limit
oxygen availability followed by heat removal and sec-
ondary air addition to complete combustion at a lower
379
-------
Q
AIR !l.
LEAN COMBUSTION ZONE
WITH INERT DILUTION
FLUE ~
GAS Ly'
[)
FUEL
-
FLUE GAS RECIRCULATION
.,./
"
A) FLUE GAS RECIRCULATION
AIR
RICH
PRIMARY
,/'
..........
FUEL
Q
B) STAGED COMBUSTION
--
_.
--
--
'0
MIXING
ZONE
LEAN
SECONDARY
(>
[)
FLUE
GAS
Figure 3. Conceptual representation of NOx control
techniques for boilers.
peak temperature. The most important feature of this
approach is the control of fuel NOx that can be
achieved. Data from a small scale system fired with a
distillate oil doped with 1 percent pyridine are shown in
figure 2. For this volatile nitrogen compound, the fuel
NOx can be reduced from about 750 ppm for a single
staged combustion at a stoichiometric ratio (SR = actual
air to theoretical air for complete combustion) of 1.2 to
150 ppm with 0.7 SR in the primary and 1.2 overall.
Even for this nonoptimized system with no interstage
cooling, the potential for substantial control of fuel NOx
is evident. The presence of interstage cooling should
further reduce total NOx by suppressing thermal NOx
formation in the second stage. These two techniques,
combined with optimized design of the burner for the
particular technique should provide the basis for control
of both thermal and fuel NOx in boilers.
Gas Turbines. The gas turbine has features which
are considerably different from boilers and, therefore,
requires somewhat different approaches to emission con-
trol. A main constraint on the system is the inlet temper-
ature of the hot gas from the combustor to the power
extraction turbine. This requires significant quantities of
dilution air to be added to combustion gases from the
lean primary zone for control of the temperature at the
turbine inlet. However, since the efficiency of the tur-
bine increases as the turbine inlet temperature increases,
a significant emphasis is being placed on developing
power turbines from the current stationary system
capability of 1,370 K toward a future goal in the 1,813
to 1,920 K range. There is no heat extraction as such;
however, the use of secondary and dilution air to cool
the primary zone is a feature that may be used for NOx
control in advanced turbine design. The overall design
380
-------
concepts can be represented in general form in figure 4
and are discussed in more detail below:
1. A schematic of the features of a conventional
gas turbine combustor is shown in figure 4A; however,
no attempt is made to represent the complex geometrics
and air admission patterns of actual combustors. The
important features are the near stoichiometric lean pri-
mary air and fuel introduced into the primary zone, fol-
lowed by secondary air to complete combustion and
dilution air for temperature control. The features of this
system combined with operation at high pressure lead to
high combustion intensities and attendant high forma-
tion of thermal NOx' The level of conversion of fuel
nitrogen to NOx is also high for this type of combustor
(refs. 18,24). The redesign of this conventiollal com-
bustor for NOx control is being pursued by many organ-
izations.
2. An alternative concept based on the use of
catalytic combustion is shown schematically in figure
4B. In this concept the total combustion air could be
premixed with the fuel, forming a mixture outside the
normal flammability limits, then passed into a catalyst
bed to promote fuel oxidation. For this stoichiometry
the adiabatic flame temperature is relatively low and the
rate of thermal NO formation is postulated to be slow.
Due to the premixed and oxygen rich nature of the
system, high levels of conversion of fuel nitrogen to NOx
can be anticipated. Engelhard Industries' work on cata-
lytically supported thermal (CATATHERMAL@)
combustion (ref. 25) indicates that both catalytic sur-
face reactions and homogeneous surface reactions are
important in achieving complete combustion and low
emissions. Catalyst exit temperatures up to 1700 K with
NOx emissions of less than 7 ppm are reported for a
variety of clean fuels including a simulated coal gas.
Carrubba (ref. 26) has recently reported 70 to 90
percent conversion of fuel nitrogen to NOx based on
propane doped with NH3 and a diesel oil doped with
pyridine. As previously postulated, this limits the
one-stage lean catalytic combustor to use on fuels
without bound nitrogen.
3. A concept which should offer the potential for
control of fuel NO is shown in figure 4C. Although this
system shown is all catalytic for purposes of discussion,
other systems using diffusion flame combustors or
hybrid conventional-catalytic systems can also be envi-
sioned. The system as shown points out one of the diffi-
culties that must be overcome for practical application,
that is, the large volume of see<;>ndary air that must be
mixed with the primary zone combustion products. The
mixing must be rapid to prevent local temperature
excursions and the attendant formation of thermal NOx'
The optimum method of secondary mixing to eliminate
potential emission and operability problems should be a
major emphasis in any development of this concept. The
concept also does not show any primary zone heat
removal by the secondary air, although this may be a
necessary and/or desirable design feature for practical
systems.
Concept Evaluation. Although the concepts dis-
cussed above appear to have strong potential for
control of both thermal and fuel NO, data on actual
LBG are not available to validate the concepts at the
higher temperatures projected for the advanced systems.
In view of this, an alternate method of evaluating the
emission control potential of the concepts was needed.
After some consideration, it was decided that some
pseudo-quantitative estimates of emissions could be
obtained by using existing mathematical models. The
two types of such models are: 1) the fluid dynamic
models which include aerodynamic and chemical phe-
nomena; and 2} the stiff chemical kinetics codes with
simple stirred reactor and streamtube configurations.
Since the geometry of the combustor concepts is com-
pletely undefined at this time, it seems unnecessary to
do complex aerodynamic calculations until some initial
determination of the desired flow properties can be
made. For this reason the second approach was chosen.
With the emphasis on use of existing technology, the
method chosen was the Modular Kinetics Application
Program (MKAP) developed by Ultrasystems, Inc., based
on the numerical analysis of Tyson and Kliegel (ref. 27).
MKAP allows the solution of a large number of simul-
taneous kinetic rate equations in streamtube (plug flow)
and perfectly stirred reactor modes. Complex flow situa-
tions can be partially simulated by combining the two
reactor modes in various ways. The combination chosen
for these calculations was a one-millisecond stirred reac-
tor to achieve ignition, followed by one or two plug flow
sections. This configuration was picked because it can be
argued that the results of the calculation bear a direct
resemblance to one combustor concept, catalytic com-
bustion. The factors to be considered are as follows: 1)
in a catalyst monolith each cell may be viewed simplis-
tically as a stream tube; 2) if a significant amount of the
reaction occurs in the gas phase as proposed by Pfefferle,
the thermal NOx formation chemistry may be reason-
ably described by homogeneous kinetics; and 3) for a
minimum heat loss from the catalyst, at steady state a
nearly uniform temperature may exist over the flow
path. This also assumes that the catalyst surface does not
playa significant role in the formation or destruction of
either fuel or thermal NO. In addition, for this initial
evaluation, the calculations were not constrained by one
381
-------
AIR
[)
[)
DILUTION [)
~Q
[)
[)
[)
[)
TTI
A) CONVENTIONAL PRACTICE
..
I .
t, [) [)
[) PREMIXED [)
FUEL AND
J AIR TTI
[) [)
. I
AIR+-
FUEL=
B) SINGLE STAGE CATALYTIC
AIR
SECONDARY
QQ
r)
r)
r)
TTI
C) TWO STAGE CATAl YTIC
Figure 4. Conceptual representation of NOx control
techniques for gas turbines.
332
-------
limit imposed by current catalysts, that is, temperature
of operation. As the concepts are further developed, this
limit must be faced and/or other approaches tried.
The calculations performed represent a first attempt
at defining the limit of control of NO)l imposed by
chemical kinetics in idealized flow situations and, there-
fore, should be viewed merely as an indication of poten-
tial rather than as a quantitative representation of the
achievable emission levels. It should also be emphasized
that the practical combustor configurations that might
resul,t from these combustor concepts may bear very
little' resemblance to current practice; however, this in
itself should not constitute a barrier to their utilization
if practical constraints can be accommodated. It is par-
ticularly important to consider that stationary gas tur-
bine combustors are not necessarily constrained by the
same factors as aircraft engines (e.g., size and weight). In
addition, if the control concepts require, it may be pos-
sible to use residence times considerably longer than the
10 mill iseconds that approximate current practice for
stationary gas turbines.
Having chosen the basic calculation approach, the
next step was in choice of the chemical kinetic rates to
be used. Again using existing information, the rate data
were chosen from the survey performed by Engleman
(ref. 28). I n an attempt to lend credence to the kinetic
rates, the results of the calculation were compared to the
jet stirred reactor data of Bartok et al. (ref. 29), for
methane with and without added nitrogen species (i.e.,
NH3 and NO). With minor adjustment of several rate
constants, the same set of kinetics was shown to fit the
experimental results of both thermal and fuel NO over
the range from 60 to 150 percent of theoretical air (Le.,
from fuel rich to fuel lean). It should also be emphasized
that in spite of the good agreement between calculated
Table 4. Composition of Lurgi gas
Constituent
Mole percent
H2
CO
C02
CH4
H20
N2
19.6
13.9
13.3
5.5
10. 1
37.6
and measured NO values, there are no reported values
for HCN from the stirred reactor under fuel rich condi-
tions. The oxidation of this HCN to fuel NO in a lean
second stage can be quite significant, as will be shown.
The kinetic rates used to predict HCN formation may be
questioned, although they fall well within the bounds of
current estimates. It should also be noted that the HCN
is formed only by CH fragments of the fuel, and does
not include synthesis from the initial fuel nitrogen
species. The details of the rate selections and compari-
sons will be documented separately.
An LBG derived from an air blown Lurgi system
was chosen for the first series of calculations because of
projected near-term availability and, more importantly,
the high methane content. The composition of the gas is
shown in table 4. For the purposes of this analysis it is
assumed that the H2S is removed from the fuel gas at
high temperature (Le., 1,088 K) by some unspecified
process. Following Robson (ref. 12) it was assumed that
the gas could contain 4,000 ppm NH3 following the high
temperature cleanup and this value was used in all calcu-
lations of fuel NO formation.
Finally, it is necessary to define the combustor inlet
and outlet conditions to be studied. The case chosen was
a combustor for an advanced gas turbine capable of
accepting turbine inlet temperature of 1,813 K with a
fuel inlet temperature of 1,088 K and 810 K air at 10
atmospheres. These conditions coupled with 4,000 ppm
of NH3 in the fuel gas and no heat loss from the entire
combustor were chosen to represent the most difficult
probable case for NOx control. For these conditions, the
overall fuel lean equivalence ratio ~ (defined as actual
fuel to air ratio divided by stoichiometric fuel to air)
yielding an adiabatic flame temperature of 1,813 K is
0.45. The results of the calculations for three cases are
described below.
1. The first case considered was the single stage
lean combustion case for the fuel gas without NH3. The
calculated thermal NO as a function of plug flow resi-
dence time is shown in figure 5. It should be noted that
the presence of the one millisecond stirred reactor for
ignition results in about 23 ppm of NO at the start of
the plug flow for the 10 atmosphere case and 7 ppm at
one atmosphere. The thermal NO then increases linearly
with residence time for both cases; however, the high
pressure case increases more rapidly. This points out two
factors that should be carefully considered in future
studies. First, it cannot be assumed that the thermal NO
kinetics at 1,813 K are too slow to be significant, and
second, great care must be taken in using low pressure
results to infer thermal NO at high pressure for this
temperature level. It is also important to point out that
for a residence time that might be experienced in an
383
-------
actual combustion (i.e., 100 ms) the thermal NO level is
about 87 ppm corrected to stoichiometric. This appears
to be a very favorable result considering the severity of
the conditions chosen.
2. The sec,?nd case was identical to the first
except that the f'uel gas contained 4,000 ppm of NH3. It
is not particularly surprising that 86 percent of the NH3
was converted to fuel NO (Le., about 890 ppm NO as
calculated). The calculation indicates that the conversion
occurred in the ignition stirred reactor and the first few
milliseconds of the plug flow. This is considerably more
rapid than the formation of the thermal NO in the same
system. It should be noted that the calculation also com-
pares very well with the 70 to 90 percent conversion of
TAIR = 810 K
TFUEL = 1088K
-------
fuel nitrogen to NO in a lean catalytic converter previ-
ously discussed (ref. 26). This result also reconfirms the
postulate that fuel NO requires a different control
approach than thermal NO and leads to the next calcula-
tion.
3. The final case was the two stage concept with
the rich primary zone to control fuel NO fol-
lowed by secondary air addition to achieve fuel burnout.
This concept has a much larger number of variables that
must be explored to determine the optimum condition,
including: primary zone equivalence ratio; primary zone
residence time; and the rate of secondary air mixing with
the partially burned primary zone gases. For fuel NO
control the primary zone is expected to be the dominant
feature of the concept. The goal is to maximize the con-
version of the fuel nitrogen specie to molecular nitrogen
which will have a negligible effect on the NO formation
in the second stage and, thereby, minimize the residual
nitrogen species (e.g., NO, HCN, NH3) that result in fuel
NO in the exhaust of the second stage. For the chosen
input conditions the two key factors in this are the
equivalence ratio and residence time of the primary
zone. (It should be noted that the temperature of the
primary zone and the fuel composition may also have an
effect; however, for this study they were not consid-
ered.) The results of the calculation are summarized in
figure 6. It should be noted that the curve connecting 4>
= 0.45 and 4> = 1.33 is hypothetical but that the shape is
based on experimental observations that fuel nitrogen
conversion is relatively independent on the fuel lean side
until the equ ivalence ratio approaches unity. The term
residual nitrogen specie is the sum of all major nitrogen
compounds (i.e., NO, HCN, and NH3) in the rich pri-
mary that can result in NOx formation upon secondary
air addition. The total residual nitrogen species decrease
rapidly as the equivalence ratio becomes rich; however, a
minimum shown by these calculations occurs at 4> = 1.33
for all residence times. As the equivalence ratio increases
toward 2, the residual nitrogen species increase for all
residence times; however, the increase is progressively
larger for shorter residence times. The reason for this can
be seen by reference to figures 7 and 8. Figure 7 for 4> of
1.33 shows that the NH3 has disappeared very early in
the plug flow, while NO and HCN from the exit of the
stirred reactor decrease steadily toward equilibrium
levels as residence time increases. Figure 8 for 4> of 2.0
shows a high NH3 concentration and a moderate NO
concentration of the exit of the stirred reactor, both of
which decrease significantly with time; however, the
concentration of HCN shows a significant increase with
residence time. [It should be noted that subsequent cal-
culations have shown that removal of the methane con-
tent of the fuel essentially eliminates the HCN forma-
tion. If this is borne out by subsequent experimental
studies, it may have a significant influence on the
ultimate level of fuel NOx control achievable by staged
combustion for fuel gases of varying CH4 content.] The
hot partially oxidized fuel gas containing residual nitro-
gen species can then be used as the input to the second
stage operating fuel lean overall to complete combustion
of the fuel. It may be expected that the residual nitrogen
species will go predominantly to NOx in this second
stage if very rapid mixing with the secondary occurs.
Since the credibility of the overall results depends
heavily on the validity of the calculation of the effluent
nitrogen species from the rich first stage, it is necessary
to seek some experimental data for comparison. As pre-
viously mentioned, most data on fuel nitrogen conver-
sion as a function of equivalence ratio in simple systems
have measured only NO and not the other residual com-
pounds on the fuel rich side. I n addition none of the
measurements have been made at elevated pressure,
although the results of the calculations on the fuel lean
side indicates that pressure has a very small effect on the
conversion. For this reason two comparisons were
chosen: 1) the author's data on staged combustion of a
doped distillate oil previously presented in figure 2; and
2) recently reported data ofAxworthy et al. (ref. 30),
performed on a low pressure methane flat flame doped
with NH3 and HCN. The details of the systems and
conditions are presented in table 5. For the oil-fired
experimental furnace it may be argued that the residual
nitrogen species from the rich primary are essentially
completely converted to NOx in the second stage and
that measured NOx constitutes a valid measurement of
the residual nitrogen species from the primary. In the
case of the Axworthy results only NO is reported;
however, recent results in the same system indicate that
at 25 ms NO is essentially the only species of importance
even under rich conditions. I n this low-pressure flat
flame, argon was substituted for nitrogen in the inlet
mixture, thereby eliminating thermal NO. The compari-
son of these results is shown in figure 9. First, note that
the premixed flat flame data correspond quite well with
the calculation at 4> of 0.45 but diverge substantially on
the fuel rich side. This could be a function of at least
four differences between the systems: 1) the fact that
the flat flame burner is operated with pure methane
which may significantly influence the HCN in early parts
of the flame; 2) the flat flame has a much shorter resi-
dence time and the NO concentration may not have re-
laxed to the same extent as the longer residence time
calculation; 3) the flat flame is nonadiabatic, which may
affect specie concentration as related to 2 above; and 4)
on the fuel rich side the initial stirred reactor of the
calculation may playa dominant role. Second, note that
385
-------
w-
I
NH3° = 4000 PPM IN FUEL
TFUEL = 1088K
TAIR = 810 K
PRESSURE= 10 ATM
RESIDUAL NITROGEN SPECIES = L (NO, NH3 AND HCN)
1.0
0- - ----
\
0.8 \
\
0 \
M
J:
:2 \
....
0 \
c:
0
.;; \
u
CIj 0.6
...
.... I'
CI)' ~
w I,f
u
w I "
Q.,
CI) I \
:2 ", \
w " , \
(.::J .
0 " , \
a: 0.4
~ ',\ \
:2
...I '\ \ \ t = 50 ms
«
::J '\ \ \
Q \ \ \ \
CI)
w \ \ \ .
a:
0.2 \ \ \
\ \ \ .
\\~ ~~~
\\
\\.
\ ----. t = 500 ms
.
0.5 1.0 1.5 2.0
EQUIVALENCE RATIO, rf>
Figure 6. Residual nitrogen species as a function of equivalence
ratio and residence time.
386
-------
10,000
ORDINATE VALUES FROM STIRRED REACTOR IGNITION ZONE T = 1 m~1!C
[NH310 = 2009 ppm (4000 pprn IN FUEL)
Tf=2190K
~UMVACUE
~ 1000
a.
Z'
o
t-
«
a:
t-
Z
UJ
U
Z
o
U
100
10-
NH3
1
1
10
[NO!
~[HCN[
100
1000
TIME (msec)
Figure 7. Nitrogen species concentration as a function of residence time at
equivalence ratio of 1.33.
on the lean side, the staged system with the diffusion
flame primary shows a much lower level of conversion
than the estimated trend of the calculation, which is
consistent with trends in experimental results on conver-
sion of fuel nitrogen in premixed and diffusion flame
systems. However, on the fuel rich side, these results
compare remarkably well with the calculations. A most
interesting feature is that the author noted a tendency
for the NOx to level off sharply or even turn up as the
primary zone equivalence ratio of the staged system was
pushed below 1.43. Perhaps the most important aspect
of these comparisons may be in the guidance that they
may provide for future combustor design exercises that
will be undertaken. In any event, in spite of wide system
differences, the comparisons lend sufficient credence to
the calculated residual nitrogen values, to use them as a
basis for the next step in evaluation of a staged system.
Working from the rich primary calculation, several
cases were calculated for secondary air injection to
achieve the overall lean equivalence ratio of 0.45 and a
final temperature of 1,813 K. It was anticipated that the
manner in which the secondary air was added would
have a significant effect on the thermal NO produced in
the second stage burnout zone. If the air is mixed grad-
ually the burnout will occur over a distribution of tem-
peratures from the inlet value, through the adiabatic
temperature at rp equal unity, to the final value. If the air
could be mixed instantaneously with the primary zone
product, the temperature would increase to a final value
at which little thermal NO is formed.
For the first attempt, a primary rp of 2.0 and a 500
msec residence time primary zone was chosen. The
secondary was calculated on the basis of instantaneous
mixing and a gradual mixing with air mass distribution
over 55 msec. These represent two rather extreme cases
for purposes of comparison. It must be realized that
instantaneous mixing is not possible and that 2 to 3
milliseconds is the shortest known time achieved in
experimental attempts to date. It is expected that the
thermal NO excursion for these times will be closer to
the instantaneous result than to the 55 ms one. This is
another parameter subject to further evaluation and
optimization. The results are shown in figure 10. The
gradual mixing case produces the large expected temper-
ature excursion with the attendant high rate of forma-
tion of thermal NO as well as oxidation of the residual
387
-------
f'
10,000
ORDINATE VALUES FROM STIRRED REACTOR IGNITION ZONE T = 1 msec
(NH3)0 = 2406 ppm ( - 4000 ppm IN FUEL)
Tf = 1900K
E
8: 1000
z'
o
f--
«
a:
f--
Z
w
~ 100
o
u
HCN
NO
10
1
1
10
100
1000
TIME (msec)
Figure 8. Nitrogen species concentration as a function of residence time at equivalence ratio of 2.0.
Table 5. Comparison of systems
2
Pressure
atm. Fuel
0.1 Methane
1.0 Oi sti 11 ate
Dil
10.0 Low-Btu gas
Estimated
primary ZDne
Fuel Primary Overall residence
NitrDgen nitrDgen equivalence equivalence time
cDmpDund cDncentratiDnc ratio- range ratio. millisecDnds
NH3 Dr HCN 1 540-5000b 0.5 to' 1.5 Same 25
Pyridine 1550 0.715 to' 1.43 Co.nstant at 250
0.715 Dr 0.835
Systema
3
NH3
2009
0.45 to' 2.0
Same
200
a 1.
2.
3.
LDW pressure flat flame (ref. 30) - ArgDn substituted fDr nitrDgen in Dxidizer
Experimental furnace with rich primary and lean secDndary (ref. 19)
CalculatiDns fDr this paper
CDnstant 2500 ppm in fuel and air fDr ~ = 1.5 to' 0.5
b
C
ShDwn as ppmNO fDr 100% cDnversiDn cDrrected to. ~ = 1.0
388
-------
1.0
0.8
o
M
J:
2
....
o
c:::
o
'..
(,)
C\I
~. 0.6
en
w
U
W
0..
en
2
w
e"
o
a:
!:: 0.4
2
...J
«
::>
c
en
w
a:
0.2
o
. -----
. '\
\
LEGEND
. THIS STUDY t = 200 ms
. AXWORTHY - HCN
(REF.30)
o AXWORTHY - NH3
(REF. 30)
. MARTIN - PYRIDINE
(REF. 19)
0.5
\
\
\
A \
\ \
A
\\
~
A
~ .
\6" .~
" . - A c::::::::::'
A
\
1.0
FUEL LEAN
1.5
FUEL RICH
EQUIVALENCE RATIO,
-------
1100 1100
2500 1000 1000
900 900
800 800
E T T
~ 700 E 700
~ 0.
SECONDARY AIR 0. SECONDARY AIR
w' :2
a:: 0 MIXING TIME.::::. 55 msec :2 INSTANTANIOUSL Y MIXED
::J i= 600 Q 600
I- <:( I-
<:( 1 500 a:: IjJ OVERALL = 0.45 <:( IjJ OVERALL = 0.45
a:: I- a::
w :2 I-
c.. w 500 :2 500
~ CJ w HCN AND NH3-0 IN < 0.1 msec
w :2 CJ
I- 0 :2
CJ 0 INITIAL HCN = 119 ppm
400 CJ 400 INITIAL NO = 17 ppm
PROMPT NO = 39 ppm
300 300 INCLUDES
DILUTION
EFFECT
200 200
NO
500 100 100
0 0
0 50 100 0 50 100
TIME AFTER INITIATION OF SECONDARY AIR INJECTION (MSEC)
Figure 10. Comparison of instantaneous and gradual mixing of secondary air
with primary zone products at ~ = 2.0.
nitrogen species from the primary. The instantaneous
mixing case produces very rapid conversion of the resid-
ual nitrogen spe<:ie as well as a small thermal NO spike
identified as "prompt NO:' followed by a gradual rise of
thermal NO with time. In the latter case a final value of
200 ppm NO is approached, which corresponds to
approximately 450 ppm NO corrected to stoichiometric.
This level of NO is about the same as that of the current
state-of-the-art for direct coal combustion; however,
there appears to be the potential to achieve even low
NOx with further design optimization.
A second case was chosen bas~d on a primary equiv-
alence ratio of 1.33, and 500 msec at which the residual
nitrogen species were a minimum, followed by instan-
taneous mixing of secondary air to achieve an overall
equivalence ratio of 0.45. The results of this calculation
are shown in figure 11. The final NO value in the ex-
haust is 67 ppm at IjJ of 0.45 or about 150 ppm cor-
rected to IjJ of unity. This is a factor of 3 improvement
over the first case (1jJ = 2.0) with instantaneous mixing.
Summary. The results of the calculations indicate
that advanced combustor designs have a good potential
390
-------
T AIR = 810K
I)
SECONDARY
AIR TO ACHIEVE
<:> OVERALL OF 0.45
AIR
PRIMARY
<:> = 1-.33
T1 = 1088K
FUEL
4000 ppm NH3
PRIMARY ZONE
RESIDENCE
TIME = 500 msee
T = 2190K
~
o
NOm = 43 ppm
HCNm = 8 ppm
[)
[)
TTI = 1813K
if> = 0.45
NO = 67 ppm
MIXING I
ZONE
t-
SECONDARY
ZONE
RESIDENCE
TIME = 100 msee
INSTANTANEOUS
NOTE:
NOm AND HCNm INCLUDE
SECONDARY DILUTION
EFFECT
Figure 11. Optimization of the initial design concept.
for controlling total NOx from a high temperature,
nitrogen containing low-Btu gas. It should be empha-
sized that the results shown here are a very limited set
and in no way can be interpreted as representing an
optimized application of the technology. However, they
are sufficiently encouraging to pursue both the calcula-
tion and experimental verfication of the concepts. Some
limits of the calculations should be reemphasized as fol-
lows:
1. The flow schemes were simplified based on the
catalytic combustion scheme represented as a plug flow
reactor with only homogeneous reactions influencing
NOx formation. It must be pointed out that the adi-
abatic temperature for both of the rich primary cases
exceeded the postulated maximum allowable bed tem-
perature of 1813 K (i.e., for I/J of 1.33, the adiabatic
temperature is 2190 K). At these elevated temperatures
the catalyst and monolith degradation is a severe prob-
lem; however, if only a hot non catalytic surface is need-
ed at elevated temperatures, some of the high tempera-
ture. -""onol~ths- may serve.
2. Related to the above is the fact that all calcula-
tions were done under adiabatic conditions, whereas, in
practical turbines there is some heat removal from the
primary zone by secondary and dillution air. The non-
adiabatic cases need to be considered for at least two
reasons. First, the temperature of the primary zone may
affect some of the fuel nitrogen species reactions, espe-
ciallythe HCN formation. Second, proper use of pri-
mary zone cooling could significantly relax the require-
ment for instantaneous secondary air addition to mini-
mize mixing zone thermal NOx'
3. Some of the chemical kinetic rates have not
been validated by experimental data, especially for HCN
formation.
4. The calculations have been run for only one
fuel gas composition. It is quite important to establish
the effect of different compositions, especially methane
content, on the results.
5. Only a single set of input conditions (i.e., fuel
temperature, air temperature, NH3 content, etc.) has
been run.
6. The initial stirred reactor has only been run at
one residence time (i.e., one millisecond); however, it
appears to playa major role in the fate of fuel nitrogen
species. The effect of residence time needs to be evalu-
ated.
7. Alternate concepts, such as a diffusion flame
preburner followed by a rich plug flow relaxation zone,
may offer even greater potential for NOx reduction.
It is obvious that extensive work needs to be done
to validate these concepts for advanced gas turbine de-
signs.
In addition, similar concepts should be applicable to
steam generators with equal or greater success. The most
closely related steam generator would be the high pres-
sure (supercharged) boiler; however, with the exception
of size, the general approaches should be applicable to
391
-------
~r
all boilers. As previously mentioned, the boiler applica-
tion may be somewhat easier to accomplish as significa~t
quantities of heat can be removed from the primary
zone. This is particularly essential when the boiler must
operate at low overall excess air for optimum efficiency.
Otherwise, even with instantaneous secondary air mix-
ing, the second stage temperature would be high and
significant thermal NO formation could occur.
CONCLUSIONS
Based on the information presented in this paper,
the following conclusions were drawn:
1. The data on emission levels and control tech-
niques for synthetic fuels combustion are still very
limited; however, some experimental information is now
available. The same statement is true for the fuel prop-
erties. Additional data are required on both aspects of
the problem.
2. In the absence of hard data, an existing compu-
tational method was used to estimate the potential of
LBG combustor design concepts for NOx control. The
results of those calculations show the potential f,or over
90 percent control of fuel NO for a high temperature,
ammonia containing low-Btu gas is theoretically pos-
sible. While these estimates cannot be taken as being
quantitative, it is believed that they provide a good
indication for the potential of the concepts. The achiev-
able results are obviously limited by practical considera-
tions including size, cost, mixing, and material con-
straints and duty cycle requirements.
3. A more extensive study is required to examine
the effect of other variables and design concepts on NOx
emissions and, thereby, "optimize" the techniques for
relatively simple flow systems. This needs to be followed
by an experimental evaluation to verify the techniques.
The author plans to pursue this line of R&D.
4. Perhaps the most important point is that the
comQustor design concepts here represent a significant
departure from current gas turbine combustors; how-
ever, it may be necessary to abandon conventional
approaches in order to control NOx emissions for the
severe conditions imposed by the advanced power gener-
ation systems.
5. While control of NOx emissions for high-nitro-
gen synthetic liquid fuels was not specifically consid-
ered, it appears that similar concepts should apply. The
major uncertainties are in the area of fuel properties as
related to smoke formation and/or particulate resulting
from solid material remaining in the fuel following refin-
ing. If combustion control of fuel NOx can be used, it
will reduce the necessity for hydrotreating of the syn-
thetic fuels to remove nitrogen compounds.
ACKNOWLEDGMENT
The author wishes to acknowledge the thought and
effort devoted by Drs. T. J. Tyson and M. P. Heap of
Ultrasystems, I nc., to performing the kinetic calculations
and their judgment employed in modifying some of the
conditions run in the optimized mode. The author takes
sole responsibility for the interpretations placed on
those calculations.
REFERENCES
1. Economic Commission of Europe Second Seminar
on Desulfurization of Fuels and Combustion Gases,
Washington, D.C., November 11-20, 1975.
2. P W. Spaite, "Liquefaction and Gasification of
Solid Fuel Prospects for Production of Synthetic
Boiler Fuels from Coal," introductory report pre-
sented at Economic Commission of Europe, Second
Seminar on Desulfurization of Fuels and Combus-
tion Gases, Washington, D.C., November 11-20,
1975.
3. G. B. Martin, "Environmental Considerations in the
Use of Alternate Fuels in Stationary Combustion
Processes," symposium proceedings; Environmental
Aspects of Fuel Conversion Technology, EPA-
650/2-74-118, NTIS No. PB 238-304/AS, October
1974, pp. 259-276.
4. W. Bartok et aI., "Systems Study of Nitrogen
Oxides Control Methods for Stationary Sources,
Vol. II," Esso Research and Engineering, Linden,
New Jersey, Report G R-.2 NOS-69, prepared for
Division of Process Control Engineering, National
Air Pollution Control Administration, EPA No.
APTD 1286, NTIS, No. PB 192-789, November
1969.
5. R. A. Brown et aI., "System Analysis Requirements
for Nitrogen Oxide Control of Stationary Sources,"
Aerotherm-Acurex Corporation, EPA-650/2-74-091,
NTIS No. PB 237-367/AS, September 1974.
6. G. B. Martin, "Overview of U.S. Environmental
Protection Agency's Activities in NOx Control for
Stationary Sources," presented at the Joint U.S.-
Japan Symposium on Countermeasures for NOx'
Tokyo, Japan, June 28-29, 1974.
7. D. G. Lachapelle, J. S. Bowen, and R. D. Stern,
"Overview of Environmental Protection Agency's
NOx Control Technology for Stationary Combus-
tion Sources," presented at the 67th Annual AIChE
Meeting, Washington, D.C., December 1974.
8. K. Yamagishi et aI., "A Study of NOx Emissions
Characteristics in Two Stage Combustion," pre-
sented at the 15th Symposium (International) on
Combustion, Tokyo, Japan, August 1974.
392
-------
9. C. P. Fenimore, "Formation of Nitric Oxide in Pre-
mixed Hydrocarbon Flames," 13th Symposium
(I nternational) on Combustion, Salt Lake City,
Utah, August 1970.
10. J. F. Farnsworth et aI., "Clean Environment with
Koppers-Totzek Process," symposium proceedings;
Environmental Aspects of Fuel Conversion Tech-
nology, EPA-650/2-74-118, NTIS No. PB
238-304 AS, October 1974, pp. 115-130.
11. A. J. Forney et aI., "Analyses of Tars, Chars, Gases
and Water Found in Effluents from the $ynthane
Process," symposium proceedings; Environmental
Aspects of Fuel Conversion Technology, EPA-
650/2-74-118, NTIS No. PB 238-304/AS, October
1974, pp. 107-113.
12. F. L. Robson and A. J. Giramonti, "The Environ-
mental I mpact of Coal-Based Advanced Power
Systems," symposium proceedings; Environmental
Aspects of Fuel Conversion Technology, EPA-
650/2-74-118, NTIS No. PB 238-304/AS, October
1974, pp. 237-257.
13. A. Lisauskas and S. A. Johnson, "NOx Formation
During the Combustion of Low Btu Gas from
Coal," presented at the AIChE 80th National Meet-
ing, Boston, Massachusetts, September 7-10, 1975.
14. R. W. Duhl and T. O. Wentworth, "Methyl Fuel
from Remote Gas Sources," presented at 11th
Annual Meeting of the Southern California Section
AIChE, Los Angeles, California, April 1974.
15. R. W. Duhl and E. Allegrini, "Methyl Fuel - A Boiler
Alternate" presented at the Symposium on Impact
of Methanol on Urban Air Pollution, 80th National
AIChE Meeting, Boston, Massachusetts, September
7-10,1975.
16. G. B. Martin and M. P. Heap, "Evaluation of NOx
Emission Characteristics of Alcohol Fuels for Use in
Stationary Combustion Systems," presented at the
Symposium on Impact of Methanol on Urban Air
Pollution, 80th National AIChE Meeting, Boston,
Massachusetts, September 7-10, 1975.
17. J. E. Haebig, B. E. Davis, and E. R. Dzuna, "Prelimi-
nary Small Scale Combustion Tests of Coal Liq-
uids," presented at ACS Division of Fuel Chemistry,
Philadelphia, Pennsylvania, April 1975.
18. M. C. Hardin, "Evaluation of Three Coal Derived
Liquid Fuels in A Standard T63 Combustor,"
Detroit Diesel Allison Report RN 74-28, November
14,1974.
19. G. B. Martin and E. E. Berkau, "An Investigation of
the Conversion of Various Fuel Nitrogen Com-
pounds to Nitrogen Oxides in Oil Combustion,"
AIChE/Symposium Series Air Pollution and Its Con-
trol, Volume 68, 1972.
20. D. W. Turner, R. L. Andrews, and C. W. Siegmund,
"Influence of Combustion Modification and Fuel
Nitrogen Content on Nitrogen Oxides Emissions
from Fuel Oil Combustion," presented at 64th
Annual AIChE Meeting, San Francisco, California,
November 1971.
21. P. W. Pillsbury, E. N. G. Cleary, P. P. Sigh, and R.
M. Chamberlain, "Emission Results from Coal Gas
Burning in Gas Turbine Combustors," paper No.
75-GT-44, presented at ASME Gas Turbine Confer-
ence, Houston, Texas, March 2-6, 1975.
22. R. D. Klapatch and G. E. Vitti, "Gas Turbine Com-
bustor Test Results and Combined Cycle Systems,"
Combustion Vol. 45, No. 10, April 1974, pp. 35-38.
23. R. L. Robson and A. J. Giramonti, "The Environ-
mental Impact of Coal-Based Advanced Power
Systems," symposium processings; Environmental
Aspects of Fuel Conversion Technology, EPA-
650/2-74-118, NTIS No. PB 238-304/AS, October
1974, pp. 237-257.
24. C. Wilkes and R. H. Johnson, "Effects of Fuel
Nitrogen on NO" Emissions from Gas Turbines,"
pres~nted at the EPR I NOx Control Technology
Seminar, San Francisco, California, February 5-6,
1976.
25. N. C. Pfefferle, R. V. Carrubba, R. M. Heck, and G.
W. Roberts, "Catathermal Combustion: A New
Process for Low-Emissions Fuel Conversion," pre-
sented at the ASME Winter Meeting, Houston,
Texas, November 30 to December 4, 1975.
26. R. V. Carrubba, M. Chang, W. C. Pfefferle, and L.
M. Polinski, "Catalytically Supported Thermal Com-
bustion for Emission Control," presented at the
EPRI NOx Control Technology Seminar, San Fran-
cisco, California, February 5-6, 1976.
27. T. J. Tyson and J. R. Kliegel, "An Implicit Integra-
tion Procedure for Chemical Kinetics," Paper No.
68-180, AIAA 6th Aerospace Sciences Meeting,
1968.
28. V. S. Engleman, "Survey and Evaluation of Kinetic
Data on Reactions in Methane/Air Combustion,"
Exxon Research and Engineering Company, EPA-
600/2-76-003, NTIS No. PB 248-139/ AS, January
1976.
29. W. Bartok et aI., "Laboratory Studies and Mathe-
matical Modeling of NOx Formation in Combustion
Processes," Exxon Research and Engi.neering Com-
pany Final Report No. GRU-3GNOS-71, EPA No.
APTD 1168, NTiS No. PB 211-480, 1972.
30. A. E. Axworthy, G. R. Schneider, M. D. Schuman,
and V. H. Dayan, "Chemistry of Fuel Nitrogen Con-
version to Nitrogen Oxides in Combustion,"
Rocketdyne Division of Rockwell International,
393
-------
7)"7
?f
EPA 600/2-76-039, NTIS No. PB250-373/AS,
February 1976.
31. E. R. Ozuna, "Combustion Tests of Shale Oils," pre-
sented at the meeting of the Central States Section
of the Combustion Institute, Columbus, Ohio, April
5-6, 1976.
32. J. S. Ball and H. T. Rail, paper presented at 27th
Midyear Meeting of the American Petroleum In-
stitute ~ivision of Refinery, San Francisco, Calif-
ornia, 1962.
394
-------
ENVIRONMENTAL IMPACT AND R&D
NEEDS IN COAL CONVERSION*
E. M. Magee, R. R. Bertrand, and C. E. Jahnigt
Abstract
The environmental impact of coal conversion plants
is very large. This impact manifests itself in coal and
water requirements and solid, liquid, and gaseous emis-
sions. These resource requirements and effluents are dis-
cussed for eight gasification and three liquefaction
plants. Research and development needs, aimed at
minimizing this impact, are also discussed. Special atten-
tion is given to areas where sufficient information is
tacking for a reliable estimate of the magnitude of the
environmental impact.
INTRODUCTION
The impact of coal conversion on the environment
can be very large. The magnitude of this impact has been
indicated for gasification in references 1 and 2, but the
total extent of the impact has not been evaluated be-
cause of lack of information concerning many of the
factors involved.
Under contract with the Environmental Protection
Agency, Exxon Research and Engineering Company has
prepared preliminary designs for eleven coal conversion
processes, and one chemical coal cleaning process. These
processes are listed in table 1. The designs are based on
the nonconfidential information available at the time
and emphasize the areas of environmental concern.
It is necessary to consider each process when a de-
tailed environmental assessment is to be carried out, but
there are certain generalizations that can be made con.
cerning environmental impact of conversion and R&D
needs. A paper from this laboratory was presented
earlier that dealt with control technology R&D needs
(ref. 15). The present paper will emphasize environ-
mental impact and areas where data assessment is need-
ed.
ENVIRONMENTAL IMPACT
The impact of coal conversion manifests itself in
resource requirements and in gaseous, liquid and solid
*This work was carried out under contract No. 68-02-0629
with the U.S. Environmental Protection Agency.
tAuthors are with the Government Research Laboratories
of Exxon Research and Engineering Company, Linden, New
Jersey.
Table 1. Preliminary process
designs prepared*
o Low- and Medium-Btu Gasification
- Koppers-Totzek (3)
- U-Gas (4)
- Winkler (5)
o High-Btu Gasification
- BI-GAS (6)
- HYGAS (7)
- C02 Acceptor (8)
- Lurgi (9)
- Synthane (10)
o Liquefaction
- COED (11)
- H-Coal (12)
- SRC (13)
o Chemical Coal Cleaning
- Meyers (14)
*Numbers in parentheses refer to
references.
effluents. Resource requirements are principally coal and
water; there are numerous gaseous, liquid, and solid
streams that enter the environment. These process
demands on the environment are discussed in this
section.
The plants discussed vary in size and no attempt is
made to relate the various quantities to products since
the purpose is to illustrate the overall environmental re-
quirements. Besides, the various bases differ consider-
ably from process to process. The ranges of the heating
value of the products are given in table 2. In this and
other tables, approximate values are given for a 100,000
bbl/day petroleum refinery and a 1,000 MWe power
plant for comparative purposes.
395
-------
~..
r
Table 2. Heating value of products
Processes
Product Higher Heating Value,
106 Btu/hr
Lo~- and medium-Btu gasification
High~Btu gasification
Liquefaction
100,000 bbl/day refinery (example)
1,000 MWe power plant (example)
3,600-10,400
9,800-12,800
6,900-24,400
27,000
3,400
Raw Water Requirements
Coal conversion plants are large users of raw water,
needed for three principal purposes. First, the conver-
sion of coal to liquid or gaseous fuel requires a source of
hydrogen since coal is deficient in hydrogen compared
to gaseous or liquid hydrocarbons. The only source of
this hydrogen is water and this water is chemically con-
verted so that it represents a net water loss. A second
water requirement is for quenching and washing various
process streams. This water is not destroyed but be-
comes contaminated with various organic and inorganic
chemicals. The third use for water is for cooling process
streams. A large fraction of this water is evaporated to
provide the cooling effect and this causes a buildup of
inorganic salts in the cooling water so that a portion
must be purged. Evaporation losses. drift losses and
blowdown must be replaced with fresh water.
The ranges of raw water requirements of the proc-
esses studied are given in table 3. Also shown for com-
parison are the requirements for a medium-size petro-
leum refinery and a large power plant utilizing a recycle
cooling system. The largest gasification and liquefaction
plants require almost twice as much water as the refinery
and about as much as the power plant. These quantities
of water are very significant and would place a large
environmental burden on many areas where coal is avail-
able. Some reduction in water usage is possible with air
cooling, but water requirements are still large as some of
the process designs include a large fraction of air cooling.
Table 3. Raw water requirements
Processes
Raw Water, gal/min
Low- and medium-Btu gasification
High~Btu gasification
Liquefaction
100,000 bbl/day oil refinery (example)
1,000 MWe power plant (example)
2,400-3,200
2,800-7,100
3,600-7,600
4,000
8,000
396
-------
Table 4. Energy losses
Processes
Energy Losses,
109 Btu/Day
50- 116
Low- and medium-Btu gasification
High-Btu gasification
Liquefaction
122-154
118- 193
100,000 bbl/day oil refinery (example)
1,000 MWe power plant (example)
65
130
Energy Losses
The next impact to be discussed is net energy con-
sumption of the processes studied. This net energy con-
sumption, or energy "loss," has environmental implica-
tions for several reasons. First, it has a direct effect in
heat rejected to the surroundings-that is, to air and
water. This direct effect may not, however, be as
important as the resulting side effects that include the
consumption of a proportionate quantity of resources
(coal and raw water). the increase in the size of certain
plant equipment, and the increased quantities of chemi-
cal effluents. Table 4 lists the ranges of the energy losses
for the processes studied together with comparative ex-
amples for an oil refinery and power plant. The losses
for the coal conversion processes are larger than for the
oil refinery and the average is about equivalent to that of
the power plant. The magnitude of the losses can be
better appreciated from the fact that the lost heat is
sufficient to provide heating for approximately 100,000
homes.
An approximation of the units responsible for the
energy losses is shown in table 5. The assignment of
responsibility for the losses shown in table 5 are some-
what subjective-for example, the loss of sensible and
latent heat of the main gas stream is assigned to the shift
and cooling area. This should, unlike losses due to com-
pression in oxygen production, be prorated over the
whole plant. However, no realistic method of such pro-
ration is obvious. The sensible heat losses in many
minor, miscellaneous streams are assigned rather arbi-
trarily, but l:1ave little effect on the numbers in the table.
The first column of numbers shows losses in the
utility and fuel gas sections of the plant. These sections
are not themselves part of the gasification scheme; there-
fore, the losses due to these sections were prorated over
those units using fuel gas, electricity, and steam. This
proration has a large effect on the gasification section
losses because the gasification section is a major user of
steam.
The losses due to methanation and pipeline com-
pression, as shown in the table, are distorted due to the
fact it was expedient to use extraction turbines in these
sections. If the latent heat losses from steam used in
steam drives is prorated over all the steam drives, the
percentage losses in methanation and pipeline compres-
sion become a more realistic 11.9 percent and 6.9 per-
cent respectively. The other numbers in the table would
be reduced slightly to make up to 100 percent.
The major revelation from table 5 is that there is no
single section of the plant where improvements would
have a pronounced effect on the overall thermal efficien-
cy. However, since the total heat losses amount to over
6,000 million Btu/hr, there is room for improvement.
Gaseous Effluents
The major gaseous effluent streams from the proc-
esses studied are shown in table 6. These streams are
almost as large as the corresponding streams from a large
power plant using recycled cooling water. The gaseous
effluents from the latter are also shown in the table.
The CO2 and flue gas streams can contain combus-
tibles, sulfur compounds and oxides of nitrogen. The
effluent cooling tower air can contain volatile organic or
inorganic compounds that may be present in cooling
water from leaks or other sources. The waste nitrogen
streams from the oxygen plants are usually clean.
The size of the streams are so large that extreme
care must be taken to ensure that they are reasonably
clean. Some techniques for controlling pollutants in
397
-------
!iY
(
Table 5. Thermal losses by unit in Lurgi gasification(1)
Plant Section
Percent of Total Energy Loss
Pipeline compression
Before Proration After Proration
of Utility and of Uti 1 ity and
Fuel Gas Losses Fuel Gas Losses
0.4 2.2
13.4 22.6
5.7 22.8
14.3(2) 14.5(2)
15.1 18.7
6.7 7.7(4)
1.1 1.7(4)
1.3 2.4
6.4 7.4
17.5(3)
18.1
Coal preparation
Oxygen production
Gasification and quench
Shift and cooling
Purification
Methanation
Sulfur recovery
Gas liquor treating
Utilities
Fuel gas production
(1)
(2)
(3)
(4)
Based on information from reference 16.
Major losses due to cooling.
Includes miscellaneous areas totaling 0.4%.
Extraction turbines used; if total losses in condensing
steam to steam drives is distributed evenly, th~se
numbers become 11.9% for methanation and 6.9% for compression
with equivalent reductions in all other areas.
these streams have been discussed earlier in this sympo-
sium (ref. 15).
There are other sources of gaseous effluents in coal
conversion plants that, though not as large as those
shown in the table must, nevertheless, be considered.
Such areas as coal storage, evaporation ponds, solids
quench, and others must be considered to prevent air
pollutants from escaping.
Liquid Effluents
Table 7 lists the ranges of liquid effluents from coal
conversion plants. from an example oil refinery and
from a large power plant. The numbers listed in the table
represent total dirty water except for storm drainage and
miscellaneous purges. All of this water may not be dis-
charged as techniques exist for cleaning the water for
reuse. The two major sources of dirty water are process
398
-------
Table 6. Major gaseous effluent streams
Processes
Gaseous effluent streams, ton/day
C02 + Flue Gas C.T. Air Nitrogen
Low- and medium- 4,900-19,200 600,000-2,221,000 0-38,000
Btu gasification
High-Btu gasifi- 45,000-94,000 546,000-3,264,000 0-21,000
cation
Liquefaction 39,000-66,000 1,250,000-2,655,000 6,000-16,000
1,000 MWe power 110,000 3,500,000 0
plant (examrrl e)
water streams and cooling tower blowdown. The former
usually contains organics such as oil, phenols, etc., and
inorganics such as ammonia, sulfur compounds, and
others. The cooling tower blowdown contains a high
concentration of dissolved solids, but may contain
organic materials due to leaks.
Also included in table 7, for comparison, are typical
dirty water streams for an oil refinery and a large power
plant using recycle cool ing. As can be seen, the water
from coal conversion plants is about equivalent to that
from a petroleum refinery and is much greater than that
from a power plant. The treatment of these quantities of
water for discharge or reuse can be very expensive.
Solid Effluents
Table 8 shows the ranges of solids from coal con-
version plants. Also listed is an example quantity of solid
waste from a large power plant. Solids from petroleum
refineries are relatively modest.
The major solids effluent streams consist of ash
from combustion operations and from the gasification or
liquefaction steps. The quantities to be disposed of are
larger than those from a power plant or from oil re-
fineries. The disposal of these materials may present
difficulties.
Impact Summary
The conclusion that can be drawn from this infor-
Table 7. Liquid effluents
Processes
Total Dirty Water,*
gal/min
Low- and medium-Btu gasification
High-Btu gasification
Liquefaction
100,000 bbl/day oil refinery (example)
1,000 MWe power plant (example)
600-2,000
1,300-3,500
1,100-2,400
2,000
1,500
*Does not include rain runoff and miscellaneous purges.
399
-------
(~;.:
'if
Table 8. Solid effluents
Processes
Solids*, ton/day
Low~ and ~edium-Btu gasification
High~Btu gasification
Liquefaction
1,000 MWe power plant (example)
1,000-1,300
1,000-4,900
700-3,000
1 ,000
*Does not include coal cleaning, flue gas scrubbin~ and
miscellaneous sludges.
mation is that the impact of a coal conversion plant on
the environment is as great as that of a medium-size
petroleum refinery or a large coal burning power plant.
R&D NEEDS
The potential, large environmental impact of coal
conversion plants requires that effective means be avail-
able to control the pollution from exiting streams and to
minimize resource requirements. Control technology
R&D needs were discussed earlier in this symposium
(ref. 15), and the present secti on of th is paper wi II
emphasize data acquisition needs.
Trace Elements
The concentrations of trace elements in coal are
fairly well known (refs. 17,18), but far less is known of
the disposition of these potentially hazardous chemicals
in coal conversion plants. If 20,000 tons/day of coal is
used as feed to the plant and this coal contains 1 ppm of
a trace element, then 40 pounds/day of that element
must appear in streams from the plant.
It is obvious that all materials entering the plant
must also leave via the effluent or product streams.
Many of the trace elements volatilize to a small or large
extent during processing, and many of the volatile com-
ponents can be highly toxic. This is especially true for
me,cury, selenium, arsenic, molybdenum, lead,
cadmiurT), beryllium, and fluorine. One study has been
made or1 the trace element content of coal solids after
various 'stages of treatment (ref. 19). A very recent
attempt has bee>n. made to make a material balance on
trace elements in coal gasification (ref. 20). The recovery
was variable, ranging from 17 to over 100 percent. Some
information is available on the trace element content of
liquid products from the SRC process (ref. 13). In
general, though, comparatively little information is avail-
able as to the eventual fate of the trace elements enter-
ing a coal conversion plant and much more data is need-
ed before the potential problem can be evaluated.
R&D Needs in Air Conservation
It will be necessary, of course, to know what ele-
ments and compounds are present, and in what quanti-
ties, in all gaseous effluent streams. The very magnitude
of some of the effluent streams, as shown in table 6,
makes the presence of very small traces of pollutants a
matter of concern. Other streams, though smaller, may
contain larger concentrations of potential pollutants
and, therefore, must be controlled.
Table 9 lists some sources of pollutants that could
Table 9. R&D needs in air conservation
o
Determination of quantities of
volatiles from cooling towers
Quantification of organic losses to
air from evaporation ponds and biox
units
o
o
Analysis of combustibles removed
with ac i d gas
Control technology where problems
exist
o
400
-
-------
very well be overlooked in an investigation of pollution
in coal conversion plants.
Cooling water-especially that used against high
pressure streams, can pick up trace quantities of volatiles
through leaks. These volatiles then enter the atmosphere
from cooling towers. Individual petroleum refinery
cooling towers in the Los Angeles area were reported to
be emitting 3-1,500 pounds of hydrocarbons per day
(ref. 21). This source of atmospheric pollution from coal
conversion must be evaluated and corrective control pro-
cedures taken where necessary.
Evaporation ponds and biological oxidation (biox)
units can be sources of atmospheric pollution. Several
papers have pointed out the increase in microbial con-
centrations in the air near sewage treatment plants (refs.
22,24). Thibodeaux and Parker (ref. 25) have developed
equations for calculating the concentration of volatiles
in the air near aerated basins. Techniques for handling
dirty water, such as evaporation ponds and biox units,
can-unless care is taken-transfer a water pollution
problem into an air pollution problem. If this happens,
more R&D is necessary to produce technology to allow
water treatment without air contam ination.
Another area deserving attention is the removal of
combusti bles along with acid gas from coal conversion
streams. This possibility is recognized for absorption
systems and means to control the effluent of the com-
busti bles have been indicated in one case (ref. 16). Other
systems such as hot potassium carbonate also absorb
gases such as hydrogen, carbon monoxide,and methane
(ref. 26). A better definition of potential problems that
may result in coal conversion plants is needed.
When these problems in the area of air conservation
are better defined, appropriate R&D is needed to pro-
duce technology to solve the problems.
R&D Needs in Water Conservation
Table 10 lists some specific areas for data acqui-
sition needed in the area of water pollution control in
coal conversion plants. One item needed is a better water
analysis than is presently available where materials are
lumped as oxygen demand. Since many compounds-
such as certain aromatics-are not readily biodegradable,
their presence in water effluents does not show up as
biological oxygen demand. Analysis for specific chem-
icals in effluent water is needed together with a deter-
mination of their fate in conventional wastewater treat-
ment facil ities.
Another area for R&D is an effort to determine
quantities of miscellaneous purge streams and their
compositions. For example, published information on
the solution purge from Stretford units used in desul-
furizing coke oven gas indicates a large variation in the
quantities of purge (ref. 27). A better definition of such
Table 10. R&D needs in water conservation
o Analysis for specific constituents
in dirty process water
o Documentation of quantities and
qualities of miscellaneous purge
streams (e.g., from acid gas removal)
oAnalysis of leachate from coal
storage, ash, other solids
o Control technology where problems
exist
streams in coal conversion plants will indicate what
further R&D is necessary to minimize the environmental
impact of such streams.
An examination of the leachabil ity of inorganic
materials from coal, ash, and other solids from coal con-
version plants is needed. The ash from liquefaction units
and gasifiers will probably differ from the ash from coal
combustion. The knowledge as to whether water runoff
or seepage from such solids is innocuous would be of
value.
Control technology necessary to remove any envi-
ronmental danger from water pollution must await
further problem definition in the area.
Data Acquisition in Coal Conversion Plants
An analytical test plan has been formulated to facil-
itate data acquisition in coal conversion plants. Using an
example coal gasification plant and a coa I liquefaction
plant, a detailed list of streams where analyses might be
required was made (ref. 28). Table 11 shows a summary
of the streams.
Twenty streams in gasification and 23 in liquefac-
tion are the minimum that would have'to be sampled
and analyzed to make a pollutant material balance. If a
balance could not be obtained due to errors or other
causes, it might be necessary to sample and analyze as
many as 29 other streams in gasification and 38 in
liquefaction in order to determine the cause of the im-
balance. The example gasification 'plant has 72 streams
and the example liquefaction plant has 75 streams. For
various reasons, it is conceivable that all these streams
would have to be sampled and analyzed.
The magnitude of the problem is increased by the
number of analyses that may be required for each
stream. The Analytical Test Plan gives the met'lod~. -
401
-------
~/
/
Table 11. Stream analysis requirements for typical
conversion plants
Gas ifi cati on
Li quefacti on
Minimum streams to be analyzed
for pollutant material balance
Additional stream analyses to
determine errors
Total streams identified where
analyses may be required
20
23
28-29
37-38
72
75
those analyses shown in table 12. Of course, not all the
analysis would have to be carried out for each sample,
but the magnitude of the effort required is truly large.
CONCLUSIONS
The potential impact of coal conversion plants on
the environment is large. The magnitude of the impact is
comparable to that of a medium-size petroleum refinery
Table 12. Analyses to be performed
Material
Number of
Analyses
Metals
17
Gases
16
Polynuclear aromatics
14
Other organic materials
9
Inorganic ions
7
Coal, ash and char
11
Water quality indicators
12
Particulates
or a large coal-burning power plant. When it is realized
that the impact of coal conversion is an additional
impact-that is, it does not replace refining or ultimate
fuel utilization-the drive to minimize the effects of coal
conversion becomes even more significant.
The large size of the potential environmental impli-
cations of coal conversion requires a comparable amount
of new information and technology needs. The R&D
effort needed to assure that coal conversion can be
carried out in an environmentally sound way is indeed
also very large.
REFERENCES
1. W. J. Rhodes, "Environmental Impact of Coal
Gasification," in "Air: I. Pollution Control and
Clean Energy," C. Rai and L. A. Spielman, Eds.,
AIChE Symposium Series 147, Vol. 71, 1975, pp.
154-159.
2.. C. E. Jahnig and R. R. Bertrand, "Environmental
Aspects of Coal Gasification," presented at the
AIChE Meeting, Boston, Mass., Sept. 1975.
3. E. M. Magee, C. E. Jahnig, and H. Shaw, "Evalua-
tion of Pollution Control in Fossil Fuel Conversion
Processes, Gasification; Section 1: Koppers-Totzek
Process," EPA-650/2-74-009a, January, 1974. NTIS
PB-231 675.
4. C. E. Jahnig, "ibid Section 7: U-Gas Process,"
EPA-650/2-74-009i, August 1975.
5. C. E. Jahnig, "ibid Section 8: Winkler Process,"
EPA-650/2-74-009j, September 1975.
6. C. E. Jahnig, "ibid Section 5: BI-GAS Process,"
EPA-650/2-74-009g, May 1975. NTIS PB-243 694.
7. C. E. Jahnig, "ibid Section 6: HYGAS Process,"
EPA-650/2-74-009h, August 1975.
402
-------
8. C. E. Jahnig and E. M. Magee, "ibid Gasification;
Section 4: CO2 Acceptor Process,"
EPA-650/2-74-009d, December 1974. NTIS PB-241
141.
9. H. Shaw and E. M. Magee, "ibid Section 3: Lurgi
Process," EPA-650/2-74-009c, July 1974. NTIS
PB-237694.
10. C. D. Kalfadelis and E. M. Magee, "ibid Section 2:
Synthane Process," EPA-650/2-74-009b, June 1974.
NTIS PB-237 113.
11. C. D. Kalfadelis and E. M. Magee, "Evaluation of
Pollution Control in Fossil Fuel Conversion Proc-
esses, Liquefaction; Section 1: COED Process,"
EPA-650/2-74-Q0ge, January 1975. NTIS PB-240
372.
12. C. E. Jahnig, "ibid Section 3: H-Coal Process,"
EPA-650/2-74-009m, October 1975.
13. C. E. Jahnig, "ibid Section 2: SRC Process,"
EPA-650/2-74-009f, March 1975. NTIS PB-241
792.
14. E. M. Magee, "Evaluation of Pollution Control in
Fossil Fuel Conversion Processes, Coal Treatment;
Section 1: Meyers Process," EPA-650/2-74-009k,
September 1975.
15. C. E. Jahnig, R. R. Bertrand and E. M. Magee, "En-
vironmental I mpact and R&D Needs in Coal Conver-
sion," this symposium.
16. Anon, "EI Paso Natural Gas Company Burnham
Coal Gasification Complex-Plant Description and
Cost Estimate," prepared by Stearns-Roger, Inc.,
August 16, 1972 (Rev. 9/20/72). Application of EI
Paso Natural Gas Company before U. S. Power
Commissions, Docket No. CP 73-131, November 15,
1972 (Rev. Oct. 1973).
17. E. M. Magee, H. J. Hall and G. M. Varga, Jr.,
"Potential Pollutants in Fossil Fuels,"
EPA-R2-73-249, June 1975. NTIS PB-225 039.
18. H. J. Hall, G. M. Varga, Jr. and E. M. Magee,
"Symposium Proceedings: Environmental Aspects
of Fuel Conversion Technology," St. Louis,
Missouri, May 1974, EPA-650/2-74-118, pp. 35-47.
19. A. Attari, "The Fate of Trace Constituents of Coal
During Gasification," EPA-650/2-73-004, August
1973.
20. A. J. Forney et aI., "Trace Element and Major Com-
ponent Balances Around the Synthane PDU Gasifi-
er," Pittsburgh Energy Research Center, Report No.
PERC/TPR-75/1, August 1975.
21. U.S. Department of Health, Education, and Welfare,
"Atmospheric Emissions from Petroleum Refiner-
ies," PHS NO. 763,1960. NTIS PB-198 096.
22. P. J. Napolitano and D. R. Rowe, Water and Sewage
Works, 113, Dec. 1966, p. 480.
23. A. P. Adams and J. C. Spendlowe, Science, 169,
Sept. 1970, p. 1218.
24 L. J. Thibodeaux and N. J. Carter, "Recent Ad-
vances in Air Pollution Control," AIChE Sympo-
sium Series, Vol. 70, No. 137, AIChE, New York,
1974.
25 L J. Thibodeaux and D. G. Parker, "Desorption
Limits of Selected Industrial Gases and Liquids
from Aerated Basins," Paper No. 30D, AIChE Meet-
ing, Tulsa, March 1974.
26. J. H. Field et aI., "Pilot Plant Studies of the. Hot-
Carbonate Process for Removing Carbon Dioxide
and Hydrogen Sulfide," Bureau of Mines Bulletin
597, U.S. Gov't Printing Off., 1962.
27. A. J. Moyes and J. S. Wilkinson, The Chemical En-
gineer, Feb. 1974, p. 84.
28. C. D. Kalfadelis et aI., "Evaluation of Pollution Con-
trol in Fossil Fuel Conversion Processes, Analytical
Test Plan," EPA-650/2-74-009-1, October 1975.
403
-------
rr
(
TECHNICAL REPORT DATA
{Please rcad J"s1.l11c/iol1s 011 the rel'crse before eOIl1/J/c/il1g}
1. REPORT NO. /2. 3. RECIPIENT'S ACCESSIO~NO.
EPA-600/2-76-149
4. TITLE AND SUBTITLE S . 5. REPORT DATE
ymposlUm Proceedings: Environ- June 1976
mental Aspects of Fuel Conversion Technology, n
(December 1975, Hollywood, Florida) 6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S) B. PERFORMING ORGANIZATION REPORT NO.
Franklin A. Ayer (Compiler)
9. PERFORMING ORGANIZATION NAME AND ADDRESS 10. PROGRAM ELEMENT NO.
Research Triangle Institute EHB52 9
P.O. Box 12194 11. CONTRACT/GRANT NO.
Research Triangle Park, North Carolina 27709 68-02-1325, Task 57
12. SPONSORING AGENCY NAME AND ADDRESS 13. TYPE OF REPORT AND PERIOD COVERED
EPA, Office of Research and Development Final' 2-4/76
14. SPONSORING AGENCY CODE
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711 EPA-ORD
15. SUPPLEMENTARY NOTES IERL-RTP Task Officer for this report is William J. Rhodes,
Mail Drop 61, Ext 2851.
16. ABSTRACT Th t EPA d . th . tit
e repor covers TS secon symposIUm on e enVlronmen a aspec s
of fuel conversion technology. The symposium was conducted at the Diplomat Hotel,
Hollywood, Florida, December 15-18, 1975. Its main objective was to review and
discuss environmentally related information in the field of fuel conversion technology.
Specific topics were environmental problem definition, process technology, control
technology, and process measurements.
17. KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS b.IDENTIFIERS/OPEN ENDED TERMS C. COSA TI Field/Group
Air Pollution Process Variables Air Pollution Control l3B 13H, OV
Fue Is Industrial Processes Stationary Sources 21D
Conversion Control Problem Definition
Environmental Measurement Process Technology 14B
Engineering Control Technology 05E
Environment s Process Measurements
18. DISTRIBUTION STATEMENT 19. SECURITY CLASS (This Report) 21. NO. OF PAGES
Unclass Hied 410
Unlimited 20. SECURITY CLASS (This page) 22. PRICE
Unclassified
EPA Form 2220-1 (9-73)
'U.S. COVEfu~..IEN", PRINTING OFFICE: 1976-641-317/5515 Region No.4
404
~--
-------
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Technical Information Staff
Cincinnati, Ohio 45268
OFFICIAL BUSINESS
PENALTY FOR PRIVATE USE. S3OO
AN EQUAL OPPORTUNITY EMPLOYER
POSTAGE AND FEES PAID
US ENVIRONMENTAL PROTECTION AGENCY
EPA-335
Special Fourth-Class Rate
Book
D
O
V
O
MS P LEVIN
U5EPA REGION III
6TH AND WALNUT STS
CURTIS BLDG 6TH AND WALNUT
PHILADELPHIA P*
CINN-3
STS
19106
If your address is incorrect, please change on the above label,
tear off- and return to the above address.
If you do not desire to continue receiving this technical report
series. CHECK HERE ["").- tear off label, and return it to the
above address.
------- |