4.MEFA
United States
Environmental Protection
Agency
industrial Environmental Research EPA-dOO/7-79~217
Laboratory September 1979
Research Triangle Park NC 27711
Symposium Proceedings:
Environmental Aspects
of Fuel Conversion
Technology, IV
(April 1979,
Hollywood, FL)
Interagency
Energy/Environment
R&D Program Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the
effort funded under the 17-agency Federal Energy/Environment Research and
Development Program. These studies relate to EPA's mission to protect the public
health and welfare from adverse effects of pollutants associated with energy sys-
tems. The goal of the Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology. Investigations include analy-
ses of the transport of energy-related pollutants and their health and ecological
effects; assessments of, and development of, control technologies for energy
systems; and integrated assessments of a wide range of energy-related environ-
mental issues.
EPA REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved
for publication. Approval does not signify that the contents necessarily reflect
the views and policies of the Government, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/7-79-217
September 1979
Symposium Proceedings: Environmental
Aspects of Fuel Conversion Technology,
IV (April 1979, Hollywood, FL)
Franklin A. Ayer and N. Stuart Jones
(Compilers)
Research Triangle Institute
P.O. Box 12194
Research Triangle Park, NC 27709
Contract No. 68-02-3132
Task No. 1
Program Element No. EHE623A
EPA Project Officer: T. Kelly Janes
Industrial Environmental Research Laboratory
Office of Environmental Engineering and Technology
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
PREFACE
These proceedings for the symposium on "Environmental Aspects of
Fuel Conversion Technology" constitute the final report submitted to
the Industrial Environmental Research Laboratory, U.S. Environmen-
tal Protection Agency (IERL-EPA), Research Triangle Park, N.C.
The symposium was held at the Diplomat Hotel, Hollywood, Florida,
April 17-20, 1979.
This symposium acted as a colloquium for discussion of environmentally
related information on coal gasification and liquefaction. The program
included sessions on program approach, environmental assessment, and
control technology development. Process developers, process users,
research scientists, and State and Federal government officials par-
ticipated in this symposium, the fourth to be conducted by IERL-RTP
on the subject since 1974.
Mr. T. Kelly Janes, Chief, Fuel Process Branch, EPA-IERL, Research
Triangle Park, N.C., was the Project Officer. Mr. William J. Rhodes, Pro-
gram Manager, Synthetic Fuels, Fuel Process Branch, EPA-IERL-
RTP, was the General Chairman of the symposium. Dr. N. Dean Smith,
Project Officer, Fuel Process Branch, EPA-IERL-RTP, was the Techni-
cal Chairman.
Mr. Franklin A. Ayer, Manager, Technology and Resource Management
Department, and Mr. N. Stuart Jones, Analyst, Technology and Re-
source Management Department, Center for Technology Applications,
Research Triangle Institute, Research Triangle Park, N.C., were sym-
posium coordinators and compilers of the proceedings.
-------
17 April 1979
Contents
Page
Keynote Address 1
Steven Reznek
Session I: GENERAL APPROACH 5
T. Kelly Janes, Session Chairman
Multimedia Environmental Goals 7
Carrie L. Kingsbury and James B. White
Source Analysis Models for Environmental Assessment 23
Larry R. Waterland and L. B. Anderson
Interagency Research on the Environmental Transport
and Effects of Synfuels-Related Substances 37
W. Gene Tucker and Gerald J. Rausa
Department of Energy Environmental Assessment
Program for Coal Conversion 43
Fred E. Witmer
NK>SH Programs for Evaluation and Control of Industrial
Hygiene Hazards In Coal Conversion 59
James Evans and Barry G. Pallay
EPRI Clean Fuel Program 71
S. B. Alpert and B. M. Louks
Monitoring and Testing Program of Low-Btu GasrHers 85
K. E. Cowser, G. V. McGurl, and R. W. Wood
The Use of Low-Btu Gas for Iron Oxide Pellet Induration—
An Interim Report 97
Robert K. Zahl and J. C. Nigro
18 April 1979
Session II: ENVIRONMENTAL ASSESSMENT: GASIFICATION 109
Charles F. Murray, Session Chairman
Synthetic Fuels Implementation 111
Orcutt P. Drury
Pollutant Evaluations for a Laboratory Semi-Batch
CoalGasrHer 113
John G. Cleland and John Pierce
iii
-------
Contents
Page
Environmental and Engineering Evaluation of the
Kosovo Coal Gasification Plant, Yugoslavia (Phase I) 137
j W * »
Becir Salja, Mira Mitrovic, and Dragan Petkovic
Kosovo Gasification Test Program Results-Part II,
Data Analysis and Interpretation 181
Karl J. Bombaugh and William E. Corbett
Environmental Assessment Report: High-Btu Gasification Technology 203
Masood Ghassemi, K. Crawford, S. Quinlivan, and Donald L. Strehler
Environmental Assessment Report for Wellman-Galusha
Gasification Systems 251
William C. Thomas and Gordon C. Page
Fate of Phenols During the Gasification of Coal 279
John Fillo and Michael J. Massey
Predictions on the Disposition of Select Trace
Constituents in Coal Gasification Processes 303
Gerald L. Anderson, Andrew H. Hill, and Donald K. Fleming
19 April 1979
Session II: ENVIRONMENTAL ASSESSMENT: LIQUEFACTION 333
J. Wayne Morris, Session Chairman
Initial Sampling of the Fort Lewis SRC Pilot Plant 335
David D. Woodbridge
Environmental Assessment of SRC-It—An Update 357
C. Raymond Moxley and David K. Schmalzer
Environmental Assessment Report: Solvent-Refined Coal 383
Kevin J. Shields
Combustion of Liquid Synfuels 405
G. Blair Martin, W. Steven Lanier, G. C. England,
M. P. Heap, and D. W. Pershing
Session III: ENVIRONMENTAL CONTROL 423
Robert P. Hangebrauk, Session Chairman
Control Assay Screening Procedures 425
William F. Longaker, Alfred B. Cherry, and Sohrab M. Hossain
Evaluation of Coal Conversion Wastewater Treatability 457
Philip C. Singer, James C. Lamb, III, Frederic K. Pfaender,
Randall G. Goodman, Randy Jones, and David A. Reckhow
iv
-------
Contents
Control Technologies for Paniculate and Tar
Emissions from Coal Converters 479
Daniel Kennedy, Leonard Breitstein, and C. Chen
A Coal Gasification-Gas Cleaning Pilot Plant:
Operating Experience and Initial Results 499
James K. Ferrell, R. M. Felder, and R. W. Rousseau
Chemical Analysis and Leaching of Coal Conversion
Solid Wastes 521
Robert A. Griffin, R. M. Schuller, S. J. Russell,
and N. F. Shimp
20 April 1979
Hazardous Waste-Definition and Regulation 531
Alan Corson, Mathew Straus, and David Friedman
Factors Considered in Effluent Limitations
Guideline Development 537
John W. Lum
Water Requirements for Synthetic Fuel Plants 539
Harris Gold and David J. Goldstein
Applicability of Petroleum Refinery and Coke
Oven Control Technologies to Coal Conversion 559
Robert A. McAllister
-------
KEYNOTE ADDRESS
Steven Reznek
Office of Energy, Minerals, and Industry,
U.S. Environmental Protection Agency, Washington, D.C.
Welcome to the fourth edition of our sym-
posia on the environmental aspects of fuel con-
version technology. On behalf of the U.S. En-
vironmental Protection Agency (EPA) I wish to
thank you for attending and participating in our
discussions. The outside observer comparing
the agenda for edition 4 to those for editions 1,
2, and 3 might have a sense of deja vu because
we continue to use many of the same topical
headings from session to session. However,
those of you who have persisted in the research
areas that comprise our program of environ-
mental assessment of synthetic fuels technol-
ogies recognize that we are dealing with a net-
work of technical, economic, institutional, and
environmental issues analogous to a kaleido-
scope. Every time something shakes the system
or every time a different vantage point is used,
new features surface.
Upon surveying the progress that has been
made since the first symposium in St. Louis 5
years ago, I was reminded of the words of a song
made popular by Ray McKinley many years ago:
"You've come a long way from St. Louie, baby,
but you still have a long way to go." In keeping
with that theme, I would like to review briefly
where we have been, where we are now, and
then give you my opinion of what our path must
be over the next few years if we are to meet our
obligation to guide the development of synthet-
ic fuels technologies in an environmentally
sound manner.
The first symposium in May 1974 came short-
ly after the oil embargo. The intervening 5
years have presented us with a natural gas
shortage and a coal strike. The price of our fossil
fuels has steadily climbed, pressed by the
steady escalation in the price of our imported
oil. The most recent OPEC oil price increases
continue this dismal trend. Yet we persist in our
high per capita consumption of energy because
our economy is based on it.
The trend in U.S. energy consumption over
the past 30 years is illustrated in Table 1. The
growth in the use of oil and natural gas is re-
flected in massive capital investments in distri-
bution systems and equipment utilizing these
fuels. During this time the amount of coal used
has been comparatively static, undergoing a
slight decline and then recovering to the point
that 1977 use was slightly above 1950 use. How-
ever, as can be seen, coal consumption is ex-
pected to more than double by 1990. The histori-
cal and anticipated distribution in the use of this
coal is illustrated in Table 2.
Included in the industrial/nonenergy cate-
gory for 1990 is an estimated 75 million tons of
coal per year, totaling 5 percent of expected
U.S. production for use in synthetic fuels
production. It is conceivable that this estimate
may be too conservative. In 1977 imports ac-
counted for about 45 percent of the 19.1-million
bbl/d used. The projected figures for 1990 in-
dicate that imports will account for 51 percent
of some 22 million bbl/d consumed in the United
States. In 1978, oil imports dramatically esca-
lated the U.S. trade deficit to over $28 billion.
The economic burden of imports will escalate
still further. Compared to 1940, per capita
(worldwide) energy consumption will double by
1980, increasing demand —and therefore
prices —in the oil market. These economic
forces, plus the impacts of geopolitical events
(witness the oil embargo of 1973, the recent
interruption of oil production in Iran and the
present fuel shortage) will certainly accelerate
the technological development and commer-
cialization of synthetic fuels from coal.
The U.S. Department of Energy (DOE) has
recognized the need to develop a viable synthet-
ic fuels industry and has formulated plans to
spur commercialization. As with most Federal
agencies, however, budgetary limitations have
been imposed, which have pared DOE commer-
cialization support down to a small fraction of
the recommended amounts. At the same time,
the private sector has been reluctant to risk
equity capital investment to attract a sufficient
amount of debt capital needed to finance syn-
thetic fuels plant construction. This is par-
-------
TABLE 1. PATTERN OF U.S. ENERGY CONSUMPTION 1950-1990
Consumption in Quads (1015 Btu) by Year
Source 1950* I960* 1970* 1977*
Hydro, nuclear
Dry, natural gas
Petroleum liquids
Coal
Total, quads (1015 Btu)
Average growth rate (%)
1.43
6.15
13.5
12.9
34.0
2.75
1.6
12.7
20.1
10.J?
44.6
4.17
2.9
22.0
29.5
12.7
67.1
1.78
5.1
19.6
37.1
14.1
75.9
3.
14.6
19.1
48.2
30.3
112.2
05
*Data from U.S. Bureau of Mines cited in: H. R. Linden et al. Perspec-
tives on U.S. and World Energy Problems. Institute of Gas Technology
February 1979.
fData from Exxon Company, U.S.A. Energy Outlook 1978-1990. May 1978.
TABLE 2. DISTRIBUTION OF U.S. COAL DEMAND 1960-1990
Consumption in Quads (1015 Btu) by Year*
Sector I960 1977 1990
Electric utility
I ndustri al/nonenergy
Residential /commercial
Total quads
4.45
4.65
1.1
10.2
10.2
3.7
(L2
14.1
22.6
7.7
—
30.3
*Data from Exxon Company, U.S. Energy Outlook 1978-1990. May 1978.
-------
ticularly true for liquefaction and high-Btu
gasification plants, which typically require well
over billion-dollar investments.
Although this may appear to be a classic con-
frontation between an irresistable force (of ac-
celerating energy demand) and an immovable
object (of investor reluctance), I believe the bar-
rier will quickly crumble when the price of liq-
uid and gaseous fuels from coal becomes com-
petitive with petroleum liquids and natural gas.
Then there will be a rush to build synthetic
fuels plants.
If EPA is to do its job correctly, the Agency
must be prepared with environmental protec-
tion guidelines and regulations before or at
least coincident with the rush. Otherwise, we
face the charge of causing delays and thereby
jeopardizing the nation's energy supplies and,
consequently, the nation's economic well-being;
or we face the charge of underestimating the
environmental impacts, thus jeopardizing the
nation's environmental well-being. I, for one,
don't wish either tag. Let me bring you up to
date in our efforts to ensure that we will be
prepared. Initially, we reviewed what was al-
ready known about the processes for producing
synthetic fuels from coal, the types of pollutants
emanating from these processes, and the tech-
niques that were available for controlling the re-
lease of pollutants. These control methods were
evaluated in terms of their adequacy under the
air and water standards extant circa 1974. One
conclusion drawn was that the amount of ex-
isting data available were inadequate and that
the reliability of this data were questionable.
EPA began to remedy this situation by funding
field studies—environmental assessments, we
called them—on low- and medium-Btu gasifiers,
high-Btu gasifiers, and liquefaction systems. We
also researched various methods to control
pollutants in product and byproduct streams,
waste streams, and fugitive emissions.
As work progressed, it became obvious that
environmental assessment involved more than
pollutant quantification and control methods
evaluation. Removal efficiency for Pollutant X
might be, say, 98 percent; but was that good
enough? The only way to know was to quantify,
or at least estimate the acceptable concentra-
tion level of Pollutant X in the environment for
which the risk of hazard was very small. The
necessity to establish goals against which to
measure the effectiveness of control methods
led to the development of what we call multi-
media environmental goals (MEGs). These were
initially developed for a limited number of
chemical species, but as the scope of EPA's man-
dated responsibilities has grown through the
Toxic Control Substances Act, the Priority Pol-
lutant List, and similar addenda, the list of MEG
species has expanded to keep pace.
A companion development to MEGs were the
source analysis models, or SAMs. SAMs were
based on the use of simplified assumptions
about pollutant transport and transformation in
the environment in order to provide a rapid,
standardized method of estimating pollutant
loadings, which are then compared to MEG
values.
As this framework for effects estimation has
been developed, we have also refined our meas-
urement data base through more comprehen-
sive and more precise measurements. Addition-
al chemical species have been identified and
assayed, sampling and analysis techniques have
been refined, data quality control procedures
have been employed, and biological methods
have been added to our evaluation arsenal.
Over the next 4 days you will share with each
other the details of your individual contribu-
tions since our last meeting. As you listen, I
believe that you will agree with me that, indeed,
we have come a long way from St. Louis. The
next question is: Where do we go from here?
We believe we have laid a solid foundation of
research to scope the environmental problems
associated with developing a synthetic fuels in-
dustry. Now, our thrust is to integrate this in-
formation into a format that will make it direct-
ly useable to guide plant designers, plant oper-
ators, EPA regulatory offices, EPA enforce-
ment offices, and the State counterparts to
these EPA offices.
Presently, we are planning two major vehi-
cles for presenting this information. These are
two report series that we term environmental
guidance documents (EGDs) and environmental
assessment reports (EARs). A given EGD will
summarize for a given technology — for exam-
ple, low-Btu gasification—what is known about
the environmental effects from pollutants that
have been identified for the various systems en-
compassed within the energy technology. It will
survey the techniques available to control these
pollutants and give suggested pollutant dis-
charge limits. In addition, it will project the
-------
future development of effluent, emission, and
solid waste disposal standards.
In projecting future standards, it is obvious
that we will need the active involvement of the
EPA program and enforcement offices. Even
though these offices are heavily committed to a
multitude of other currently more pressing
problems, they have responded quite well to our
overtures for a coordinated approach on syn-
fuels.
The second report series, EARs, will focus on
specific systems, such as the Lurgi system for
low- and medium-Btu gasification. Technical,
economic, and environmental effects informa-
tion will be assembled to present process de-
scriptions; characterization of input materials,
products, and waste streams, performance, and
costs for various control alternatives; analysis
of environmental regulatory requirements; and,
finally, a projection of environmental impacts.
Lest I appear overly optimistic as I con-
template the prospect of producing a series of
"best sellers," I remind myself and you that we
still have a long way to go. Budgetary limita-
tions and shifting priorities, both internal to and
external to EPA, will provide the usual hurdles
to program continuity and intensity. But even if
we manage to avoid these bureaucratic prob-
lems, we still face substantive problems in
collecting, refining, validating, and analyzing
data to provide a solid quantitative basis for our
conclusions and recommendations. I hope these
recommendations will achieve a balance of tech-
nology and economy to provide environmental
protection as the synfuels industry expands.
We look forward to working with you to meet
this challenge. Thank you for your support.
-------
Session I: GENERAL APPROACH
T. Kelly Janes, Chairman
Industrial Environmental Research Laboratory,
U. S. Environmental Protection Agency
Research Triangle Park, North Carolina
-------
MULTIMEDIA ENVIRONMENTAL GOALS
Garrie L. Kingsbury* and James B. White
Research Triangle Institute, Research Triangle Park, North Carolina
Abstract
A key element in the Industrial Environmen-
tal Research Laboratory's (IERL) environmen-
tal assessment methodology is the development
of multimedia environmental goals (MEGs).
MEGs addressed here are desirable control
levels of chemical substances in discharges to
the environment. The purpose of MEGs devel-
opment, the strategy used in deriving MEG
values, and the format established for their
presentation are discussed briefly. Several
aspects of MEGs for polycyclic organic com-
pounds are discussed.
INTRODUCTION
It is widely recognized that coal conversion is
accompanied by the formation of a myriad of
chemical substances and that, in large-scale
operations, these chemicals may be released to
the environment. It is the responsibility of the
U.S. Environmental Protection Agency (EPA)
to ensure that when commercial conversion
processes become operational in the United
States, they will reflect optimum strategies for
environmental control.
Conservative estimates suggest that by the
year 2000, more than 3.3 M bbl/d oil equivalent
will be supplied by synthetic fuels from coal. In
addition, as much as 38 million tons of coal may
be gasified for ammonia synthesis.1 The dimen-
sion, complexity, and severity of potential
environmental impacts of the conversion in-
dustry require that a holistic environmental ap-
proach be adopted if problems are to be pre-
vented. Potential effects on workers and on the
surrounding human populations and ecosystems
will be influenced by the quantity and composi-
tion of environmental discharges. Releases to
air and water, as well as solid wastes disposed
to land, must be considered. In addition to the
acute effects, chronic effects from low-level ex-
posures must be evaluated. Problems resulting
"Speaker.
from process upsets must be anticipated. Clear-
ly, compliance with present environmental
standards will not be sufficient to safeguard
public health and welfare because many conver-
sion-related pollutants are not currently ad-
dressed by Federal guidelines. Consideration
must be given to impacts from all chemical
species that might be discharged.
To provide for comprehensive environmental
assessments, the Industrial Environmental
Research Laboratory at Research Triangle
Park (IERL-RTP) is developing a procedure to
facilitate quantitative evaluation and com-
parison of streams and processes with respect
to their potential environmental impacts. The
methodology prescribes a systematic approach
to interpreting data obtained in environmental
assessment projects. The need for multimedia
environmental goals (MEGs) arises in this con-
text. To fully characterize waste streams for en-
vironmental assessment, pollutant levels must
be related to their environmental effects. The
development of MEGs is a first attempt at a pro-
cedural approach to evaluate and rank a large
number of pollutants for the purpose of en-
vironmental assessment.
DEVELOPMENT OF MEGs
Multimedia environmental goals are defined
as levels of contaminants or degradents (in am-
bient air, water, or land or in emissions or ef-
fluents conveyed to ambient media) that will not
produce negative effects in the surrounding
populations or ecosystems, or that represent
control limits demonstrated to be achievable
through technology. Emphasis thus far in the
MEGs development has centered around speci-
fying three types of goals-levels desirable in
ambient media, levels existing in ambient media
(natural background), and levels believed safe
for exposure of limited duration.
There are several practical considerations in
the development of environmental goals:
• A method is needed to classify pollutants for
comprehensive coverage without having to
-------
TABLE 1. TYPES OF INFORMATION SUPPLIED IN BACKGROUND
INFORMATION SUMMARIES
Information Type
Identifying Information
Properties
Characteristics, Associated Compounds
Occurrence 1n A1r
Occurrence In Water
Occurrence In Land
Other Occurrences
Human Toxidty Data
Animal Toxldty Data
Information Relative to Genotoxlc Potential
(Oncogenldty, Teratogenldty, Mutagenldty)
Aquatic Toxldty
Phytotoxldty
Standards, Criteria, Recommendations, Recognition
Specific Data Supplied
Category; subcategory; Identification number; pre-
ferred chemical name; subspecies; formula; syno-
nyms; description; WIN
Atomic number; periodic group; atomic weight;
molecular weight; melting point; freezing point;
boiling point; density; vapor density; vapor pres-
sure; solubility In water; solubility in liquid;
octanol partition coefficient; p
-------
consider individually millions of compounds.
• A procedure is needed to quantify relative
toxicity of classes of compounds as well as
individual compounds. This procedure would
allow a "most-toxic-first" ordering approach
logically derived from a step-by-step evalua-
tion and linkage between chemical and bio-
assay characterization.
• Existing standards and guidelines should be
incorporated.
• Overall format for presentation of goals
should allow direct comparison of existing
and projected "best techology" controlled
emission levels with emission goals having
environmental significance.
• A uniform approach in applying existing
health and ecological effects data is needed
as well as a means for directly delineating
health/ecological effects data gaps.
The primary objective in compiling MEGs is
to provide an index that will allow quantitative
comparison and evaluation of the hazard as-
sociated with a large number of chemical sub-
stances. The MEGs project began with the com-
pilation of a list of chemical contaminants as-
sociated with fossil fuel processes. The more
than 600 chemicals on the list were organized in-
to categories that effectively grouped chemical-
ly and toxicologically similar substances. (Iden-
tification numbers for specific compounds were
subsequently assigned on the basis of the cate-
gory organization.) In the next step of the MEGs
development, existing Federal regulations and
guidelines applicable to chemical substances
were assembled. Other types and sources of
available information relevant to environmental
goals were also identified. Finally, a suitable
presentation format, the MEG chart, was
adopted, and a 1-page information summary was
prepared to accompany and support the numer-
ical goals for each chemical. Types of informa-
tion provided in the summaries are listed in
Table 1.
Quantitative Comparison of
Pollutants
In order to establish goals, it is necessary to
define equivalent or normalized concentrations
of individual chemical species in multimedia; i.e.,
air, water, or soil (or solid waste). These equiv-
alent levels are needed to serve as reference
values and thus provide a basis for comparing
diverse pollutants. One common denominator
selected for this purpose in the MEGs method-
ology is a conservative threshold (no effect)
level.
The concept of threshold is based on the
premise that there exists for every chemical
substance some definable concentration below
which that substance will not produce a toxic
response. Threshold level is illustrated in
Figure 1. It is defined on a dose vs. response
plot as the intersection of a dose-response curve
with the ordinate. Unfortunately, the human
threshold has been determined (with con-
fidence) for very few pollutants (e.g., lead).
-THRESHOLD x'
x
X'<«-3L_ZERO THRESHOLD
POLLUTANT
RESPONSE
Figure 1. Illustration of chemical pollutant
threshold level.
In the absence of data necessary to obtain
precise threshold levels from dose-response
curves, alternative methods for projecting
"safe" levels for chemical pollutants have been
proposed. Conservative no-effect levels (below
threshold concentrations) may be predicted
from mathematical models. Continuous ex-
posure to these predicted levels may be con-
sidered "safe" by some factor more than or
equal to 1.
Estimated Permissible Concentrations—
Predicted, continuous-exposure, no-effect
levels constitute one set of MEGs called esti-
mated permissible concentrations (EPCs). EPCs
for air, water, and soil are derived through sim-
-------
pie mathematical models that relate available
information (such as threshold limit values
[TLvV ] for workroom air2 or animal toxicity
data) to a conservative continuous-exposure, no-
effects level. Separate EPCs are specified on
the basis of human health effects or ecological
effects.
The objective in formulating a method for de-
riving EPCs was to establish a hierarchy of
models that would permit the calculation of
EPCs for a large number of compounds on the
basis of information available. In the absence of
the preferred types of data (i.e., levels as-
sociated with observed effects on man), models
allowing estimation of EPCs on the basis of
minimal data were needed. Models adopted to
derive EPCs for air and water were selected
from models previously suggested in the litera-
ture. The 1976 report, Estimated Permissible
Concentrations of Pollutants for Continuous
Exposure by Handy and Schindler,3 was used
extensively as a source for these models. In ad-
dition, a simple method was developed to relate
water and soil EPCs. The systems of models
used in specifying EPC values for human health
and for ecology are outlined in Table 2.
Calculated EPCs together with the existing
Federal guidelines applicable to specific chem-
icals comprise the class of MEG values called
ambient level goals. Another class of MEG
values, emission level goals, is discussed below.
Emission Level Goals—
Emission level goals describe maximum con-
centrations of chemicals in waste streams dis-
charged to the environment. Emission goals for
chemicals may be derived from EPCs through
application of a dilution factor. Dilution factors
applied to EPCs represent the ratio of emission
concentration to consequential ambient concen-
trations and must be source-specific. The re-
sulting emission goals reflect acceptable chem-
ical loadings in ambient media.
Alternatively, a second set of MEG values
called minimum acute toxicity effluents
(MATEs) may be used as emission level goals.
These values describe concentrations believed
safe for short-term exposure (acute exposure).
EPCs and MATEs differ in exposure duration.
MATEs are derived through a procedure sim-
ilar to the method for calculating EPCs. EPC
values for a given chemical are always lower
than corresponding MATE values. However, di-
lution factors are not used with MATEs, so
these values represent very conservative emis-
sion goals.
Natural background levels constitute a third
set of MEGs applicable to discharge streams.
Used with a dilution factor, these levels may be
applied as ultraconservative emission level
goals. For some chemicals, levels detected in
ambient media are reported, but toxicity data
are unavailable. This set of goals may serve as
guidelines in the absence of environmentally
significant MEG values.
Zero Threshold Pollutants
The number of compounds classified as sus-
pected carcinogens grows larger every year.
The 1977 Registry of Toxic Effects of Chemical
Substances* lists 2,091 substances as "Sus-
pected Carcinogens"; i.e., reported to cause
tumors in some animal species.
Many researchers in the field of oncology be-
lieve that the concept of threshold levels, as
previously discussed in this paper, does not ap-
ply to carcinogenic, teratogenic, or mutagenic
pollutants. Historically, these compounds have
been referred to as "zero threshold" chemicals
to indicate that a nonzero threshold level could
not be specified. The theoretical dose-response
curves for such compounds intersect the origin
as indicated by the dotted line in Figure 1. The
"zero threshold" approach implies that ex-
posure to a carcinogen at any concentration,
even for a brief interval, can be expected to in-
crease the risk of a tumorigenic effect. This
philosophy precludes specifying EPCs for any
chemical reported to be carcinogenic or terato-
genic. It also precludes setting priorities among
zero threshold pollutants except in terms of
associated risk.
Although EPA has recently adopted an ap-
proach to assess health risks from environmen-
tal carcinogens,5 relatively few such assess-
ments are expected to be completed in the near
future. (Studies on cadmium and POM were
drafted in 1978 in response to the Clean Air Act
Amendments of 1977,87 and analyses for
suspected carcinogens listed in the Consent De-
cree have recently been published in the Feder-
al Register.9) However, limited resources re-
quire that risk assessment analyses be con-
ducted on a priority basis.
Clearly, an alternative to risk assessment is
10
-------
TABLE 2. MODELS USED IN THE DERIVATION OF
ESTIMATED PERMISSIBLE CONCENTRATIONS (EPCs)
BASED ON HEALTH AND/OR ECOLOGICAL EFFECTS
Midi. HI
Air
Watar
UgAI
Land
U«/gl
Air
Watar
Land
(»g/9)
Midal
MmkH
AHS
AHf
AM-
AHS*
AC1
AC2
AT
WHS
WH1-
WH2'
we
WT
LH1
LT
AES
AE
WES
WEI
WE2
WE3
WE4
LE
EKMadat
Lowest imodnt standard or criarion for air band
103
— x TLV or NIOSH recommendation Img/m3)
420
0.107 x LOjo (mg/kg)
0.011 x LOM (mg/kg)
10*
— x TLV or NIOSH racommandition Img/m3)
l?f Adjusted ordiring numbar band on carcinoganic
6 potantiil.
15? + Adiustad ordiring numbir band on taratoganic
1 potential.
Most stringent drinking watar standard or crinrion (ng/s)
«< EKAM,,2.orj(''lA»3l
13.8 x TLV or NIOSH recommendation (rag/m3)
or
0.4 x LOjn (mg/kg)
15 x EPCAC1orj(Mg/m3l
15 x EPCAT Ug/m3)
0.2 x Most stringent EPCyyH U/«)
0.2 x EPCw-r lvg/>)
Lowest ambiant standard or criterion for air band on
•tfactt Uig/m3)
Q.t x Lowest 24-hr concentration hivinj, en ttfext an
vagetation
-------
needed if priorities are to be established among
suspected carcinogens in the near term. Also a
straightforward method for comparing sus-
pected carcinogens with noncarcinogens must
be adopted to evaluate environmental control
strategies. The dictum that "every effort should
be made to reduce environmental contamina-
tion by carcinogens to the lowest possible level"
is not specific enough to be useful in inter-
preting data for comprehensive environmental
assessments. Specifying MEGs for suspected
carcinogens is one alternative to risk assess-
ment that can supply the suspected carcinogen
rankings needed now for decisionmaking by
IERL.
In fact, not all researchers in oncology agree
with the zero threshold concept. Cornfield9 con-
cluded, on the basis of statistical analysis of the
dose-response relationship, that "the existence
of a no-effect or threshold level for the car-
cinogenic compound administered is not pre-
cluded." Dinman has presented evidence to sug-
gest that a finite number of molecules are re-
quired within a cell before a carcinogenic
response can be triggered.10
In reality, there may or may not be a nonzero
"safe" level for carcinogens. At any rate, there
are MEG values for some carcinogens called
(with some misgiving') EPCs and MATEs. To
preface the discussion of EPCs for zero thres-
hold pollutants, some general remarks on car-
cinogens and the nature of the data available
from carcinogen testing are presented.
Information on Carcinogens
Relevant to MEGs—
Epidemiological data provide the most reli-
able indication of carcinogenic risk to human
health, but these data are sparse and difficult to
evaluate. Precise human exposure levels result-
ing in cancer are almost never known. Most of
the available human effects data refer to
chemical mixtures rather than to specific chemi-
cal compounds. For example, occupational can-
cer associated with coal and petroleum products
has long been recognized, but the specific
chemicals responsible are not positively iden-
tified. Mixtures rather than specifics remain in-
dicted. A February 1978 National Cancer Insti-
tute listing names the following among the com-
pounds observed to cause cancer in man: soots,
tars, pitches, asphalts, cutting oils, shale oils,
creosote oils, high-boiling petroleum oils, coke
oven effluents, and various combustion prod-
ucts. No specific polycyclic compounds are in-
cluded in the list.11
As early as 1947, the latent period associated
with occupational exposure to oils, pitch, and
tar products was documented. Figure 2, adapted
from a paper by S. A. Henry appearing in the
British Medical Bulletin, indicates the time
elapsing from onset of employment to manifes-
tation of neoplasia in two groups of workers.12
The activity associated with all these mixtures
is probably attributable to certain polycyclics,
but without better information on the chemical
composition of the substances, the information
remains unusable for assessment.
Presently, the best qualitative evaluations of
carcinogenic risk for chemicals are supplied in
the monographs prepared by the International
Agency for Research on Cancer (IARC). This
agency is part of the World Health Organization
ib
TIME IN TEARS
Figure 2. Latent period associated with expo-
sure to oil and tar substances.
(The graph indicates time elapsing
from onset of employment to
manifestation of a cutaneous
papilloma or epithelioma in 1,335
persons in contact with pitch, tar,
or tar products compared with
1,719 persons in contact with
shale oil or mineral oil.)
12
-------
and has prepared monographs on the evaluation
of carcinogenic risks of chemicals to man for
some 65 substances. The stated objective of the
IARC program is "to achieve and publish a bal-
anced evaluation of data through the delibera-
tions of an international grqup of experts in
chemical carcinogenesis and to put into perspec-
tive the present state of knowledge with the
final aim of evaluating the data in terms of
possible human risks..."18 The evaluations by
IARC reflect biological data, epidemiological
studies and other observations in man, and en-
vironmental data.
Almost all carcinogenic compounds in man
have been demonstrated carcinogenic in one or
more animal species. It is generally accepted
that animal studies provide important informa-
tion to evaluate carcinogenic risk to man. It
must be emphasized that all compounds re-
ported to be carcinogenic are not equally po-
tent. Effective dosages vary widely. For exam-
ple, pyrene produced tumors in mice only after
10 g/kg were administered. At the other ex-
treme, benzo(a)pyrene with n-dodecane is re-
ported to cause skin cancer in mice at a level of
2 Mg/kg.u Carcinogen studies in animals also in-
dicate that latent periods associated with
specific dosages may vary widely between com-
pounds. Latent period is the length of time be-
tween the initial application of carcinogen and
the appearance of the first tumor. In general,
potent carcinogens have shorter latent periods
than weak carcinogens.11 Response to car-
cinogens in experimental animals may be re-
ported as the occurrence or frequency of neo-
plasms compared to control animals.
Results of carcinogenic studies in experimen-
tal animals (without evaluation) are available in
two compendiums. The Registry of Toxic Ef-
ftcts of Chemical Substances* reports species
tested and lowest effect dosages for suspected
carcinogens, although no details of the studies
are given. Another reference, Survey of Com-
pounds Which Have Been Tested for Carcino-
genic Activity (commonly referred to as the
Public Health Series List No. 149)," gives more
complete information. Unfortunately, the list is
not current, the most recent volume coverings
compounds tested in 1972 to 1973.
MEGs for Suspect Carcinogens—
MEGs for individual compounds that are
suspected carcinogens are based on "adjusted
ordering numbers." These numbers, derived
from the available experimental animal data for
each compound, serve as an index to indicate
the potency or hazard associated with a given
chemical. Adjusted ordering numbers are in-
fluenced primarily by the lowest effective
dosage reported and animal species affected.
The numbers have no physical meaning because
they are obtained using an arbitrary weighting
system. They are used in MEGs because they
allow ranking of carcinogens on the basis of
available information. Adjusted ordering num-
bers used in the MEGs methodology are de-
rived from the ordering numbers developed in
the 1976 EPA report, An Ordering of the
NIOSH Suspected Carcinogens List Based On-
ly on Data Contained in the List17
Adjusted ordering numbers for organic sus-
pected carcinogens currently addressed by
MEGs range from 1 to 3 x 106. EPCs for air for
suspected carcinogens are calculated using the
model outlined below. MATEs for suspected
carcinogens are calculated using a similar equa-
tion.
The following assumptions are made in
formulating the model.
Adjusted ordering numbers increase with
carcinogenic potency, indicating genotoxic po-
tential. Goals for a given substance should be in-
versely proportional to the adjusted ordering
number.
An ambient air concentration of 1 ng/m3 may
be considered the lowest concentration of con-
cern. Therefore, the model for zero threshold
pollutants should predict a goal of s 1 ng/m3 for
highly potent carcinogens or teratogens.
K
where K
adjusted ordering number
1/6 to satisfy the <1 ng/m8
assumption for B(a)P.
APPLICATION OF MEGS METHODOLOGY
TO POLYCYCLIC ORGANIC COMPOUNDS
Polycyclic organic compounds are chemicals
containing two or more fused aromatic rings.
Hetero atoms of oxygen, nitrogen, or sulfur may
be present as well as alkyl, hydroxy, or other
ring substituents. Polycyclic organics constitute
a class of compounds of particular interest in
13
-------
EPA's synfuels program. They are known to be
present in conversion processes and have re-
ceived special attention because certain poly-
cyclics are recognized carcinogens. Application
of the current MEGs methodology to polycyclics
as a group has effectively organized and ranked
many of these compounds.
Organization
A total of 124 polycyclic compounds are in-
cluded in the MEGs master list. Six major
MEGs categories contain subcategories devoted
to polycyclic compounds. These groupings were
adopted in order to relate compounds according
to structural similarities that affect chemical
separation and analysis. Subcategories are dis-
tinguished by numbers of rings and by the
presence of heterocyclic 0, N, or S. Nonalter-
nant structures are separated from other poly-
cyclic hydrocarbons because of their unique
resonance characteristics. Descriptions of
MEGs subcategories containing polycyclic com-
pounds and representative structures are pre-
sented in Table 3.
Background Information
Background information summaries address-
ing all polycyclics appearing on the MEGs
master list have been prepared, and MEG
values are specified for 85 of these compounds.
(MEG values presently cannot be provided for
the remaining polycyclics because sufficient in-
formation is not available.) Interesting high-
lights and statistics from the information sum-
maries for polycyclics follow:
• Conflicting rules of nomenclature for
polycyclics have led to confusion. Nomen-
clature endorsed by the International Union
of Pure and Applied Chemistry is used in
the MEGs.
• Molecular weights for polycyclics addressed
range from 128 (naphthalene) to 342 (tri-
benzylene benzene, a nonalternant struc-
ture). Water solubilities are reported to be
quite low for polycyclics, although the pres-
ence of impurities may alter solubilities
substantially.
• Almost all the polycyclics addressed are
associated with coal tar. Many have been
identified in atmospheric particulate sam-
ples. Concentrations of specific compounds
in ambient media are reported in 45 sum-
maries.
• Substantial concentrations in foods are
reported for certain polycyclics (e.g., chry-
sene in vegetables: 395 /*g/kg). Many
polycycb'cs including heterocycles occur
naturally in plants.
• Lipid solubilities, although seldom reported
explicitly, may be deduced from animal test
data when the material is administered in a
lipid-type vehicle. Indications of lipid solu-
bilities for 10 polycyclics are reported in the
summaries.
• Degradation of polycyclics in the atmos-
phere is affected by solar radiation.
• Most polycyclics are planar structures. An
exception is benzo(c) phenanthrene, a four-
ring compound.
• Very limited acute toxicity data are avail-
able for polycyclics. Virtually no toxicolog-
ical data are available for the oxygen heter-
ocycles or for the sulfur heterocycles. No
evidence suggests that these heterocyclic
compounds are carcinogenic.
• Of the 37 polycyclic hydrocarbon groups ad-
dressed by MEGs, 24 are reported to be tu-
morigenic in animals. Alkylation of parent
structures may strongly influence their car-
cinogenicity. (Example: alkyl derivatives of
benzo(c)phenanthrene.)
• Nine nitrogen heterocycles have demon-
strated carcinogenic activity. Many others
in this group have no data available.
• The TLV® recommended by the American
Conference of Governmental Industrial Hy-
gtenists for particulate polycyclic aromatic
hydrocarbons (PAH) is 0.2 mg/m3. This
TLV® recognizes the carcinogenic potential
of PAH collectively. A TLV® of 0.2 mg/m3 is
also recommended for coal tar pitch vola-
tiles. This specification includes naph-
thalene, anthracene, acridine, phenanthrene,
and fluorene, collectively. The purpose of
the TLV*is to minimize concentrations of
higher weight polycyclic hydrocarbons that
are carcinogenic.
EPC Values and Amblant
Concentrations for Air
EPC values and ambient concentrations for
14
-------
TABLE 3. MEGs SUBCATEGORIES FOR POLYCYCLICS WITH REPRESENTATIVE STRUCTURE
SUBCATEGORY
REPRESENTATIVE STRUCTURES
18C. Fused Rim Hydroxy Compoundi •
21A. Two-end Three-Ring Fused Polvcvclic Hydrocarbons
218. Four Ring Futed Polycyclic Hydrocartaont
21C. Five Bini Futed Polycyclic Hydrocarbon;
21D. Compounds with Mora Thin Five Fuad Rings
22A. Two-end Three-Ring Fuad Nonalt«rn«nt Polycyclic Hydrocirbont
22B. Four Ring Futrt Nomlttrnint Polycyclic Hydrocirborn
22C. Fivt Ring Fund Nonalttrnant Polycyclic Hydrocarbons
22D. Nonilurnint Compoundi with More Than Five Fused Rings
23B. Nitrogen Heterocyclet with Fused Six-Membered Rings
23C. Pyrrole and Fused Ring Derivatives of Pyrrole
230. Nitrogen Heterocydes Containing Additional Httiroatoms
24B. Oxygen Hatirocycln with Thru or Mon Fund Ringi
25B. Sullur Hstirocycleiwlth Two or Mon Fund Rings
3-Methylcholinthrene
a.
Mithyllhiophsns
15
Btnzo(c)phintnthrene
Coronene
Benzo(b)fluoranthene
Oibenzo(c,glcarbaiole
'"XT'
Olmethylthlophena
-------
air for selected polycyclic compounds are
presented in Table 4. The bases for the MEG
values for each compound are indicated.
Rankings for polycyclics derived by the
MEGs methodology are basically consistent
with the broad rankings supplied by the Na-
tional Academy of Science (NAS) and the eval-
uations by IARC. In Table 5, carcinogen rank-
ings furnished by MEGs are compared with sug-
gested ratings used by NAS and comments by
IARC. The table includes all polycyclics ad-
dressed by MEGs with adjusted ordering num-
bers greater than 4. All polycyclics with
positive carcinogen codes (as assigned in Refer-
ence 21) are also listed in the table. It should be
noted that the only major inconsistencies in the
rankings of highly potent carcinogens occur for
benzo(a)anthracene and dibenz(a,h)pyrene. The
IARC evaluations for these compounds are of
particular interest.
On the basis of lung cancer mortality in the
United States and in other countries, some in-
vestigators have concluded "that the lung
cancer death rate in men increases by approx-
imately 5 percent for each increment of pollu-
tion as indicated by 1 ng/m3 of B(a)P." Par-
ticipants of the symposium on General Air
Pollution and Human Health with Special Ref-
erence to Long-term Effects (held in Stockholm,
March 1977) have agreed that this estimate; i.e.,
5 percent, should be "regarded as an upper limit
of the possible effect of atmospheric pollu-
tion."18
No United States standards for polycyclics
exist, although 13 polycyclic compounds are
listed in the EPA Consent Decree List. Stand-
ards for polycyclics established by other coun-
tries, however, are of interest. In 1972, the
U.S.S.R. adopted a level of 150 ng/m3 as a
maximum acceptable concentration of benzo(a)
pyrene in workplace air. In 1973, the U.S.S.R.
adopted a standard of 1 ng/m3 for benzo(a)py-
rene in ambient air.1' The Republic of Germany
has adopted a standard of 250 ng/L for car-
cinogenic polycyclic aromatic hydrocarbons in
drinking water. The German standard became
effective January 31,1975.20
The Russian standards for benzo(a)pyrene are
based primarily on work by Janysheva1' involv-
ing intratracheal instillations of benzo(a)pyrene
into the lungs of laboratory rats. The maximum
noncarcinogenic dose to the rat was determined
to be 0.02 mg of benzo(a)pyrene. This noncar-
cinogenic dose in the rat was extrapolated to a
maximum noncarcinogenic dose for humans on
the basis of organ mass (1,000 g for human lung,
1.5 g for rat).
STATUS OF MEG PROGRAM
To date, background information summaries
and MEG charts addressing a total of 640 chem-
icals have been prepared. In November 1977,
216 summaries and charts were published, and
publication of charts and summaries addressing
586 additional compounds is pending. The new
MEG volumes will contain updated summaries
and charts for 195 organics previously ad-
dressed so all organics data are contained in one
reference.
The methodology for generating MEGs has
been applied successfully to yield numerical
goals of at least one type for 572 chemical sub-
stances. (This total does not include all com-
pounds addressed by preliminary MATE
values. Background information summaries
have not yet been compiled for all of the in-
organic substances listed in the preliminary
MATE tabulations). Preparation of MEGs for in-
organics is currently in progress.
A candidate list of compounds to be ad-
dressed by MEGs in the future has been com-
piled. Criteria for inclusion of compounds in the
candidate list are association with fossil fuels
processes or interest to EPA regulatory offices.
A large number of alkylated polycyclics and ni-
trogen heterocycles appear on the candidate
list.
Early this year, a peer review of EPA's en-
vironmental assessment21 methodology raised
several issues that will influence future MEGs
work. One source of concern was that per-
sistence in the environment is not specifically
considered in the current MEGs. An indication
of uncertainty of the range associated with
specific MEG values is also needed. The safety
factors incorporated in the models for deriving
EPCs and MATEs deserve careful review be-
cause compounding safety factors may result in
overly stringent MEG values. A footnote sys-
tem to indicate the basis for each MEG value
should be used in final tabulations in future
reports. The need for a systematic review of the
overall methodology and of specific MEG values
was also pointed out. Finally, the appropriate-
ness of the nomenclature used in the MEGs, par-
16
-------
ticularly regarding the term minimum acute
toxicity effluents (MATE), was questioned.
It should be emphasized that the multimedia
environmental goals are not to be used as regu-
lations. They are designed to specify levels that
may be compared for various pollutants in order
to assess various control technologies. A conser-
vative approach has been applied consistently
to specify MEG values. Models used for MEG
calculations incorporate safety factors to ensure
that the values generated err on the safe rather
than on the high side. Also, where conflicting in-
formation required for MEGs is reported, the
more conservative value is used in the MEGs
calculations.
Projecting optimum control strategies for an
industry slated for future operation is an am-
bitious undertaking. It involves identifying en-
vironmental problems that might arise and as-
sessing their magnitude. Even while environ-
mental effects posed are speculative, priorities
must be established so research efforts may
focus on the problems believed to be most
serious. Clearly, priorities must be established
in a systematic manner. MEGs provide the vital
link between environmental effects and desir-
able control levels needed for comprehensive
environmental assessment. It is imperative that
the methodology for generating MEGs remain
flexible so that the most reliable and most up-to-
date information can be reflected in the values.
The ultimate goal in environmental assessment
is to assure that regulations necessary to pro-
tect the environment can be formulated and
that control technology to meet such require-
ments is available when needed.
ACKNOWLEDGMENTS
Work on the MEGs is funded by the U.S. En-
vironmental Protection Agency, Industrial
Environmental Research Laboratory, Research
Triangle Park, North Carolina, Contract 68-02-
3132. The project officer is T. Kelly Janes.
REFERENCES
I.Mann, Charles E., and James N. Heller.
Coal Profitability: An Investor's Guide.
Coal Week. McGraw-Hill Publications Co.
1979. p. 47-53.
2. TLV's: Threshold Limit Values for Chem-
ical Substances and Physical Agents in the
Workroom Environment with Intended
changes for 1976 and 1977. American Con-
ference of Industrial Hygienists. 1976,1977.
3. Handy, Robert, and Anton Schindler. Esti-
mation of Permissible Concentrations of
Pollutants for Continuous Exposure. U.8.
Environmental Protection Agency. Re-
search Triangle Park, N.C. EPA/600/2-76/
155. p. 43.
4. Fairchild, Edward J., ed. Registry of Toxic
Effects of Chemical Substances, Volumes I
and II. National Institute for Occupational
Safety and Health, U.S. Department of
Health, Education, and Welfare. Cincinnati,
Ohio. DHEW Publication Number (NIOSH)
78-104-B. September 1977.
5. Interim Procedures and Guidelines for
Health Risk and Economic Impact Assess-
ments of Suspected Carcinogens. Federal
Register. 41:2402. May 25,1976.
H>. Carcinogen Assessment Group's Assess-
ment of Carcinogenic Risk from Population
Exposure to Cadmium in the Ambient Air
(draft). Office of Research and Development,
U.S. Environmental Protection Agency.
Washington, D.C. May 1978. 44 p.
7. Carcinogen Assessment Group's Prelimi-
nary Report on POM Exposures (draft). Of-
fice of Research and Development, U.S. En-
vironmental Protection Agency. Washing-
ton, D.C. July 1978.15 p.
8. Environmental Protection Agency Water
Quality Criteria. Federal Register. 44(82):
15926-15981. March 15,1979.
9. Cornfield, Jerome. Carcinogenic Risk As-
sessment. Science. 198:693-699. November
18, 1977.
10. Dinman, Bertram D. "Nonconcept" of "No-
Threshold": Chemicals in the Environment.
Science. 175:495-497. February 4,1972.
11. National Cancer Institute List of Carcino-
genic Chemicals and Mixtures. National
Institutes of Health, U.S. Department of
Health, Education, and Welfare. Bethesda,
Md. February 3,1978.
12. Henry, S. A. Occupational Cutaneous
Cancer Attributable to Certain Chemicals
in Industry. British Medical Bulletin (Lon-
don). 4:389-401.1947.
13. IARC Monographs on the Evaluation of
the Carcinogenic Risk of Chemicals to
Humans, Volumes I-XVIII. International
Agency for Research on Cancer, World
17
-------
TABLE 4. SUMMARY OF AIR EPCs AND AMBIENT AIR CONCENTRATIONS FOR
SELECTED POLYCYCLICS
ID Number
Compound
Ambient Level Most Stringent
Reported (yg/m3) Air EPC (ug/m3)
Basis for EPCs
21A140 Anthracene 0.00035-0.002
21A180 Phenanthrene 0.0004-0.006
21B040 Benz(a)anthracene 0.029
21B060 7,12-Dimethylbenz(a)anthracene
21B080 3-Methylcholanthrene
21B100 Benzo(c)phenanthrene and alkyl
derivatives
21C060 Dibenz(a,c)anthracene
21C080 Dibenz(a,h)anthracene
21C100 Benzo(a)pyrene
21D080 Benzo(ghi)perylene
21D100 Coronene
0.006
-4
21B120 Chrysene _ 2.3 x 10
21B160 Triphenylene 0.0024
21B180 Pyrene 450
21C040 Benzo(g)chrysene
0.009
0.0032-0.032
4 x 10"5 to
4 x 10
-4
-4
-5
21C120 Benzo(e)pyrene 9.0 x 10
21C140 Perylene 0.0001
21C160 Picene 6.5 x 10
21D020 D1benzo(a,h)pyrene
21D040 D1benzo(a,1)pyrene
21D060 Dibenzo(a,l)pyrene
0.003
8.0 x 10"7 to
2.13 x 10"'
22B020 2,3-Benzofluorene 3.05 x 10
22B040 Fluoranthene
-3
133
3.8
0.11
0.0006
0.009
0.5
5.29
556
38
23.5
0.0002
5 x 10
-5
Lowest effective dose for tumor-
igenic response in mice is
3300 mg/kg.
Lowest dose for tumorigenic
response in mice is 71 mg/kg
Lowest effective dose for tumor-
igenic response in mice is
2 mg/kg.
Tumors in 7 species reported.
Lowest effective dose is
21 ug/kg.
Tumors in 8 species reported.
Lowest effective dose is 0.312
mg/kg.
Lowest effective dose for tumor-
igenic response 1n mice is
10 mg/kg.
Lowest effective dose for tumor-
igenic response in mice is
99 mg/kg.
Lowest effective dose reported
for tumorigenic response in
mice is 10 g/kg.
Lowest effective dose reported
for tumorigenic response in
mice is 720 mg/kg.
Lowest effective dose reported
for tumorigenic response in
mice is 440 mg/kg.
Tumors in 5 species are reported.
Lowest effective dose is
0.006 mg/kg.
Tumors in 6 species are reported.
Lowest effective dose is
2 ug/kg.
Tumors in 2 species are reported.
Lowest effective dose 1s
140 mg/kg.
Lowest effective dose for tumor-
igenic response in mice is
111 mg/kg.
Lowest effective dose for tumor-
igenic response in mice 1s
165 mg/kg.
Tumors in 2 species reported.
Lowest effective dose is
2 mg/kg.
Lowest effective dose for tumor-
igenic response in mice 1s
48 mg/kg.
162
LD50 (oral,rat): 2000 mg/kg.
18
-------
TABLE 4 (continued)
ID Number
22C020
22C040
22C060
22C080
22D020
23B120
23B140
23B160
23B180
23B200
23B220
23B240
23B260
23C080
23C120
23C140
23C160
23C180
25B040
Compound
Benzo ( k )f 1 uoranthene
Benzo( j )f 1 uoranthene
1 ,2:5,6-Dibenzof1uorene
Benzo (b)fl uoranthene
Indeno(l,2,3-cd)pyrene
Phenanthridine
Benzo(f)quinoline
Benzo(h)quinoline
Benz(a)acridine
Benz(c)acrid1ne
Di benz ( a, j) acridine
Dibenz(a,h)acridine
Dibenz(c,h)acridine
Carbazole
Benzo(c)carbazole
Dibenzo(a,i )carbazole
Dibenzo(c,g)carbazole
Dibenzo(a,g)carbazole
Benzo(b)th1ophene
Ambient Level Most Stringent Basis for EpCs
Reported (ug/m ) Air EPC (ug/m )
3.9 Lowest effective dose for tumor-
1gen1c response in mice is
72 mg/kg.
0.001 15 Lowest effective dose for tumor-
Igenic response in mice is
288 mg/kg.
31.5 Lowest effective dose for tumor-
Igenic response in mice is
590 mg/kg.
0.0005-0.02 2.1 Lowest effective dose for tumor-
igenlc response 1n mice is
40 mg/kg.
3.9 Lowest effective dose for tumor-
igenic response in mice 1s
72 mg/kg.
162 Based on acridine. LD50 (oral,
rat): 2000 mg/kg.
0.2 x 10"3 162 Based on acridine. LD50 (oral,
rat): 2000 mg/kg.
3 x 10" 162 Based on acridine: LD50 (oral,
rat): 2000 mg/kg.
2 x 10"4
0.0006 25 Lowest effective dose for tumor-
igenic response in mice is
468 mg/kg.
4 x 10 0.59 Lowest effective dose for tumor-
igenic response 1n mice is
11 mg/kg.
8 x 10" 0.5 Lowest effective dose for tumor-
igenic response in mice is
10 mg/kg.
54.5 Lowest effective dose for tumor-
igenic response 1n mice 1s
1 ,020 mg/kg
41 Lowest lethal dose (oral, rat):
500 mg/kg.
45 Lowest effective dose for tumor-
i genie response In mice Is
840 mg/kg.
28 Lowest effective dose for tumor-
i genie response in mice 1s
510 mg/kg.
0.2 Tumors in 4 species reported.
Lowest effective dose 1s
8 mg/kg.
14 Lowest effective dose for tumor-
Igenic response 1n mice 1s
270 mg/kg.
41 Lowest lethal dose (intraperl-
toneal , mouse): 512 mg/kg.
19
-------
TABLE 5. COMPARISON OF CARCINOQENICITY RATINGS FOR POLYCYCLICS
Compound
Beniodlpyrene
Oibenid,h)entfiricene
7,12-Dimethylbenz(e)
•nthrKini
3-Methylcholinthrene
Dibenz(i,i)pyrene
Benzdlenthracine
Dibenzo(c,g)carbazole
Benzofclphenanthrene
(and -CH3 dtrivitivn)
OibenzoMlecridine
Dibeniodjlecridine
Benzofblfluoranthene
DibenzodJIPVrene
PhintnthraiM
Benzolklfluorenthene
lndeno(1,2,3-cd)pynme
Mtthyl chrysenes
Chrysene
Picine
Benzodlpyrena
Dibenzod,h)pyr»ne
Dibenzod,g)carbazole
Benzo(j)fluoranthene
Cholenttirene
Dibenz(a,c)anthracene
Benzolclacridino
Benzodlcirbizole
Oibenz(a,i)carbazole
Dibenz(e,Ji)etridine
Dib«nz(t,g)fluonni
Dibenzlajlanthrecene
Dlbenz(e,clfluorene
Adjured Ratings suajasted
Ordlrlni Numkifi by National Aceeemy
(BnliforEPC-i) otfckwc«B1>«
3,314,500 «•+
754,133 m
272,809 ++++
18,683 +++
1,812 +++
1,562 *
679 ++*
312 «•+
312 ++
284 ++
78 ++
64 +
44
43
43 +
39 Not linid
32 ±
28 Not lilted
23
19 *++
12 ±
11 «•
Not liittd +*
7 +
7 Not lilted
6 ±
6 ±
8 ±
5 +
No viliui givin +
Not listed ±
indfeittom of Carcmogeiilcrty
(ARC Comme«i<13>
Produced tumori in ill 9 inimil tptcin ripornd tittid. Litint ptrlodi ihortir thin for other
polycycllci with powibli ixciption of dlbenz(i,h)inthriceni.
Tumori produced in 6 inimil ipeciet. Both locil md tyitemic circinogenic effecti obterved.
Effective it low dowt. Single doll effective in newborn mice.
Not iddrened
Not iiJdnmd
Rapid ippeenncff of loci) iircomi observed from ubcutaneoui injection in mice end him-
tteri. Skin piinting on mice w« ilto effective but leu ective then benzodlpyreni.
Carcinogenic in mice by leveral router Effective oral doee similar to methyl cholanthrena but
without gastrointestinal tract tumori.
Carcinogenic in rat, mouse, hamster, and possibly dog. Both local and systemic effects observed.
Appears to be stronger respiretory tract carcinogen then banzodtpyrene in hamster.
Not addressed
Skin tumori and lung tumors in mice observed following skin peinting or subcuteneous ad-
min istration. Not tested adequetely by other routes or in other species.
Skin tumors in mice followed topical epplicetion. Subcutaneous edministrations at highest
dosage produced local urcomai and lung tumori. Not teited in other species.
Produced skin tumors in mice following repeated skin painting, but only it levels 10 times
higher than effective benzodlpyrene levels. Not teited by other routes.
Subcuteneoui administration in mice resulted in sircomas in all animils. Not treated by other
routes or in other species.
Not addressed
Not addressed
A complete carcinogen and initiation of skin carcinogenesis in mica, but of lower potency
than benzodlpyrene. Local sircomis followed subcutaneoui injection in mice. Not treated
by other routes as in other species.
Not addressed
Skin tumori in mice followed repeated painting at high concentrations only. High dose, sub-
cutaneoui injections produced low incidence of tumors often long induction time.
Not addressed
Date from e skin painting experiments in mice evoked wnker response than benzodlpynne
or dibenzd.hlenthrecane. Not tested by other routes.
Carcinogenic effects demonstrated following repeated skin painting in mice and injections in
mice and rats. Not teited by other routes or In other species.
Not addressed
A high incidence of skin carcinomas results from repeated skin piinting in mice. Not teited in
other speciet or by other routes.
Not addressed
Not addressed
Skin tumors in mice followed topical application. Bladder tumors in ran followed paraffin wax
pellet implantation. Mot tested by other routes or in other species.
Not addressed
Not addressed
Not eddretsed
Not addressed
Not addressed
Not addressed
•Carcinoganicirv code given by NAS:
- not circinogenic
± uncertein or weekly carcinogenic
+ carcinogenic
++,+++,++++ strongly carcinogenic
Indications of carcinojenicity, refer to the Public Health Service Survey
listed in reference 16 of this paper.
20
-------
Health Organization. Geneva, Switzerland.
October 1978.
14. Cleland, J. G., and G. L. Kingsbury. Multi-
media Environmental Goals for Environ-
mental Assessment, Volumes I and //.In-
dustrial Environmental Research Labo-
ratory, U.S. Environmental Protection
Agency. Research Triangle Park, N.C.
EPA-600/7-77-136a. November 1977.
15. Grice, D. H., T. DaSilva, et al. The Testing
of Chemicals for Carcinogenicity, Mutagen-
icity, and Teratogenicity. Department of
Health and Welfare Canada. September
1973.183 p.
16. Survey of Compounds Which Have Been
Tested for Carcinogenic Activity, 1972-73
Volume. National Cancer Institute, U.S. De-
partment of Health, Education, and Wel-
fare. Bethesda, Md. PHS Publication Num-
ber 149.1973. p. 1.
17. An Ordering of the NIOSH Suspected Car-
cinogens List Based Only on Data Con-
tained in the List Office of Toxic Sub-
stances, U.S. Environmental Protection
Agency. Washington, D.C. EPA-560/1-78-
003. March 1976.132 p.
18. Lucier, George W., and Gary E. R. Hood,
ed. Environmental Health Perspectives.
22:185 p. February 1978.
19. Shabad, L. M. On the So-Called MAC (Max-
imal Allowable Concentrations) for Carcino-
genic Hydrocarbons. Neoplasma. 22(5):
459-468. 1975.
20. Eisenbeiss, F., H. Hein, R. Joester. and G.
Naundorf. The Separation by LC and De-
termination of Polycyclic Aromatic Hydro-
carbons in Water Using an Integrated En-
richment Step. Chromatography Newslet-
ter. 6*(1):8. February 1978.
21. Biologic Effects of Atmospheric Pollutants.
Particulate Polycyclic Organic Matter. Div-
ision of Medical Sciences, National Re-
search Council, National Academy of Sci-
ences. Washington, D.C. 1972.
22. Environmental Assessment Methodology
Workshop. Sponsored by the Environmen-
tal Protection Agency, Office of Energy
Minerals & Industry, Industrial Environ-
mental Research Laboratory, Research Tri-
angle Park; and Industrial Environmental
Research Laboratory, Cincinnati. Airlie
House, Va. January 16-18,1979.
21
-------
SOURCE ANALYSIS MODELS FOR ENVIRONMENTAL ASSESSMENT
L. R. Waterland* and L. B. Anderson
Acurex Corporation, Mountain View, California
Abstract
A series of source analysis models (SAMs)
have been developed to treat the results of emis-
sions assessments of stationary pollutant
sources to evaluate potential environmental im-
pact. These models provide a framework for
making structured comparisons between meas-
ured effluent stream pollutant concentrations
and threshold, allowable ambient concentra-
tions (termed multimedia environmental goals
or MEGs). Model outputs can thus be used to
rank sources or effluent streams from a source
with respect to potential environmental impact,
evaluate alternate emission control strategies,
or set priorities for control technology develop-
ment. Three models have been or are being
developed:
• A rapid-screening model based on direct ef-
fluent stream concentration comparisons,
• A screening model incorporating a dilution
factor approximation to pollutant disper-
sion, and
• A regional site evaluation model based on
detailed treatment of pollutant dispersion.
This paper describes features of each of these,
contrasts elements of their development, and
notes example applications in environmental as-
sessment programs.
INTRODUCTION
Since 1975, the Energy Assessment and Con-
trol Division of the U.S. Environmental Protec-
tion Agency's Industrial Environmental Re-
search Laboratory (EPA-IERL/EACD) has con-
ducted environmental assessment (EA) pro-
grams that focus on identifying and resolving
multimedia environmental risks from energy
systems and fuel processes. The primary pur-
poses of these EAs are to provide the research
data base to support standards development by
EPA program offices and to guide IERL control
technology development programs to ensure
•Speaker.
that appropriate controls are available when
needed. Thus, these programs centralize data,
quantify emissions and risks, evaluate control
options, and recommend R&D priorities.
To coordinate the approach and output of
each of the EAs, IERL/EACD is conducting sev-
eral methodology development tasks that de-
fine standardized procedures to be followed in
obtaining and evaluating process and environ-
mental data. Standardized sampling, chemical
analysis, and bioassy procedures are being
specified; environmental objectives are being
defined; and formats for comparing emission
data and environmental objectives are being
developed. This paper describes results of ef-
forts to date to develop a series of source
analysis models (SAMs) that address the need to
define methods of comparing emission data with
environmental objectives or multimedia envi-
ronmental goals (MEGs).
Figure 1 illustrates the environmental assess-
ment approach. This figure shows the two paral-
lel activities involved in an EA: control tech-
nology evaluation and environmental data ac-
quisition, and environmental objectives devel-
opment. These two activities are brought to-
gether in the task labeled environmental alter-
natives analysis. In this analysis, results from
process or effluent stream emissions assess-
ments are compared to MEGs to form the basis
for defining the outputs of an environmental
assessment, as shown in the figure. The tool
used to perform these comparisons is the source
analysis model (SAM). The SAM, therefore, is
the format used to compare pollutant loadings
to the environment from a pollutant source to
defined MEGs, thereby quantifying the poten-
tial environmental impact of a discharge stream
or pollutant source. Results from these compari-
sons can subsequently be used to define such
EA outputs as:
• Establishing more detailed sampling and
analysis needs,
• Identifying problem pollutants,
• Quantifying discharge stream or source haz-
ard potential,
23
-------
REGULATORY
BACKGROUND
• POTENTIAL POLLUTANTS
AND IMPACTS IN ALL
MEDIA
• DOSE/RESPONSE DATA
ARDSCRITERIA
• TRANSPORT MODE L£
• SUMMARIZE INDUSTRY.
RELATED OCCUPATIONAL
HEALTH/EPIOEMIOLOGI
CAL LITERATURE
ENVIRONMENTAL DATA
ACQUISITION
• EXISTING DATA FOR EACH
PROCESS
• IDENTIFY SAMPLING AND
ANALYTICAL TECHNIQUES
INCLUDING BIOASSAYS
• TEST PROGRAM DEVELOP
MENT
• COMPREHENSIVE WASTE
STREAM CHARACTERI2A
TION (LEVELS I, II. Ill)
• INPUT OUTPUT MATERIALS
CHARACTERIZATION
• CONTROL ASSAYS
ENVIRONMENTAL OBJECTIVES
DEVELOPMENT
• ESTABLISH PERMISSIBLE
MEDIA CONC. FOR CONTROL
DEVELOPMENT GUIDANCE
• DEFINE DECISION CRITERIA
FOR PRIORITIZING SOURCES,
PROBLEMS
• DEFINE EMISSION GOALS
• PRIORITIZE POLLUTANTS
• BIOASSAY CRITERIA
ENVIRONMENTAL ENGINEERING
ENVIRONMENTAL SCIENCES
ENVIRONMEN
TECHNOLOGY
ITAL SCIENCES!
f TRANSFER T~
CONTROL TECHNOLOGY ASSESSMENT
• CONTROL SYSTEM AND DISPOSAL
OPTION INFORMATION AND DE
SIGN PRINCIPLE
• CONTROL PROCESS POLLUTION
AND IMPACTS
• ACCIDENTAL RELEASE. MALFUNC
TION, TRANSIENT OPERATION
STUDIES
APPLICATIONS
• DEFINE BEST CONTROL TECH
NIQUE FOR EACH GOAL
• POLLUTANT CONTROL SYSTEMS
STUDIES
SELECT AND APPLY
ASSESSMENT ALTERNATIVES
ALTERNATIVE SETS OF MULTI
MEDIA ENVIRONMENTAL
GOALS (MEG'S)
• BEST TECHNOLOGY
• EXISTING AMBIENT STANDARDS
• NATURAL BACKGROUND (ELIM-
INATION OF DISCHARGE)
• SIGNIFICANT DETERIORATION
• QUANTIFIED CONTROL RID NEEDS
• QUANTIFIED CONTROL ALTERNATIVES
• DEFINED RESEARCH DATABASE FOR
STANDARDS
ENVIRONMENTAL
ENGINEERING
TECHNOLOGY TRANSFER
MEDIA DEGRADATION AND
HEALTH/ECOLOGICAL
IMPACTS ANALYSIS
• AIR, WATER, AND LAND
QUALITY
Figure 1. Environmental assessment program approach.
-------
• Ranking discharge streams and sources
with respect to potential for adverse envi-
ronmental impact,
• Evaluating alternative control/disposal op-
tions for a given discharge stream, and
• Establishing control technology R&D needs.
In line with EA needs, SAMs are being devel-
oped in three levels of detail in the treatment of
dispersion. The simplest model, SAM/IA, has
been designed for rapid-screening purposes and,
as such, includes no effluent transport or trans-
formation analysis.1 Goal comparisons employ
the minimum acute toxicity effluent (MATE)
MEG. The second model, SAM/I, has been de-
signed for intermediate screening.2 It includes
some simple approximations to effluent stream/
pollutant transport and employs ambient-based
multimedia environmental goals. SAM/11 will be
designed for regional site evaluation and will in-
clude more sophisticated consideration of pollu-
tant transport and transformation, cross media
effects, and population exposure. All levels of
the SAM, however, assume that:
• Only the MEG master list of about 650 spe-
cific chemical compounds3 4 need be con-
sidered in potential toxicity evaluations,
• The set of MEG values defined are appro-
priate indicators of threshold levels for ad-
verse health or ecological effects, and
• Pollutant synergisms and antagonisms can
be neglected.
Figure 2 illustrates the projected application
of each level of SAM in the tiered EA approach.
As the figure shows, the rapid-screening
SAM/IA model finds primary use in evaluating
Level 1 and Level 2 sampling and chemical anal-
ysis results. SAM/IA evaluations of Level 1 data
serve to identify potential problem discharges
and to point out pollutant species requiring
Level 2 analysis. SAM/IA evaluations of subse-
quent Level 2 data close the loop and give
screened pollutants and screened problem dis-
charges.
The intermediate-screening SAM/I model
finds optional use in treating Level 1 results but
is primarily used to evaluate Level 2 data.
Results from SAM/I-Level 2 evaluations con-
firm problem discharges and identify Level 3
monitoring needs. In turn, the regional site
evaluation SAM/II model is designed to treat
Level 3 data and quantify the potential impact
of the problem discharges and pollutants.
The following section discusses features of
each level of analysis model being developed,
contrasts specific elements of each model, and
presents examples of each application. The dis-
cussion, in turn, treats the rapid-screening
model, the intermediate-screening model, and
the projected form of the regional site evalua-
tion model.
RAPID-SCREENING SAM/IA
As noted in the introduction, the SAM/IA
model has been designed as a rapid-screening
tool for evaluating environmental assessment
sampling and chemical analysis data. Thus, the
model approach does not include treatment of
effluent stream dispersion or pollutant chemical
transformation. Instead, potential hazard esti-
mates employ the minimum acute toxicity ef-
fluent (MATE) MEG.3
Two hazard indices are defined in the model:
the potential degree of hazard (PDOH) and the
potential toxic unit discharge rate (PTUDR).
The PDOH is the ratio of the effluent stream
concentration of a pollutant species to that
species' MATE value:
PDOH - Pischai"ge concentration of compound i
1 MATE of compound i
Thus, the PDOH is a measure of the existence of
a potential hazard. Both health and ecological
PDOHs can be defined for gaseous, liquid, and
solid effluent streams consistent with appro-
priately defined MATE values.
The PTUDR is defined as the product of the
PDOH with the effluent stream discharge rate:
PTUDRj - PDOH..stream flow rate.
Thus, the PTUDR is a measure of the magni-
tude of a potential hazard.
The PDOH and PTUDR are calculated for
each pollutant species analyzed in the discharge
stream, or in the case of Level 1 evaluations, for
the most toxic species in an analyzed Level 1
sample fraction. Thus, to obtain a measure of
the toxic potential presented by the total ef-
fluent, individual pollutant PDOH and PTUDR
are summed to give total stream PDOH and
PTUDR. In turn, to estimate the magnitude of a
multieffluent pollutant source, stream total
PTUDRs may be summed to give a source total
PTUDR.
25
-------
Level 1
Sampling &
Analysis
Level 1 Results
\
SAMIA
i r
SAM I
• Recommended Level 2 Analyses
• Potential Problem Discharges
Level 2
Sampling &
Analysis
Level 2 Results
I
SAMIA
• Screened Pollutants
• Screened Problem Discharges
I
SAM I
1
J • Recommended Level 3 Analyses
• Confirmed Problem Discharges
Level 3
Sampling &
Analysis
Level 3
Results
SAM II
•Optional
• Quantified Problem
Pollutants
• Quantified Problem
Discharges
Figure 2. SAM application in the tiered assessment approach.
-------
Tables 1 and 2 illustrate the use of these
SAM/IA concepts to evaluate inorganic analysis
data from a coal-fired utility boiler. Table 1
shows the results of a SAM/IA assessment of
the inorganic component of the boiler's flue gas
stream, including particulate composition. The
table lists PDOH and PTUDR values for the 30
components assayed. Table 2 shows total
stream PDOH and PTUDR values for the four
effluent streams coming from the boiler, for two
sets of operating conditions: baseline or uncon-
trolled for NOX, and controlled for NOX. The
table indicates that the flue gas stream domi-
nates the unit's potential hazard. Further, when
TABLE 1. PDOH AND PTUDR FOR UTILITY BOILER FLUE GAS (INORGANIC): SAM/IA
MEG
Category
32
36
41
42
45
46
47
49
50
51
53
54
55
56
57
62
63
65
69
71
72
74
78
81
82
83
TOTAL
Component
Be
Ba
Tl
CO
CO?
Sn
Pb
NOX
NM|
As
Sb
Bi
SO?
so3
S04
Se
Te
F
Cl
Ti
Zr
V
Mo
Mn
Fe
Co
Cu
Zn
Cd
Hg
Flue Gas
Concentration
(yg/dscm)
9.0
2250
2.6
3.07 x 104
2.72 x 108
6.4
74
1.16 x 106
10.5
95
3.9
44
4.18 x 106
1.45 x 10*
6500
9.9
4.1
84
270
6100
270
260
150
240
4.5 x 10^
66
280
420
1.8
3.1
MATE:
Health
(vig/m3)
2
500
100
4.0 x 10*
9.0 x 106
1.0 x 104
150
9000
150
2
50
410
1.3 x 10*
1000
1000
200
100
2000
400
6000
5000
500
5000
5000
700
50
200
4000
10
50
PDOH:
Health
4.5
4.5
0.026
0.768
30.2
6.4 x 10-4
0.49
129
0.07
47.5
0.078
0.107
322
14.5
6.5
0.050
0.041
0.042
0.675
1.02
0.054
0.52
0.03
0.048
64.3
1.32
1.4
0.105
0.18
0.062
630
PTUDR: a
Health
(Mg/s)
0.312
0.312
0.002
0.053
2.09
4.4 x 10-5
0.034
8.93
0.005
3.29
0.005
0.007
22.3
1.00
0.450
0.003
0.003
0.003
0.047
0.070
0.004
0.036
0.002
0.003
4.45
0.092
0.097
0.007
0.012
0.004
43
Flue gas flow rate of 69.3 kg/s.
27
-------
TABLE 2. PDOH AND PTUDR FOR UTILITY BOILER DISCHARGES UNDER
BASELINE AND NOX CONTROLLED OPERATION: SAM/IA
Stream
Flue gas
Cyclone ash
ESP ash
Bottom ash slurry
TOTAL
Baseline
PDOH
PTUDR
(kg/s)
630 43,300
18.1
23.3
18.0
43,400
18.8
6.1
56.9
Controlled
PDOH
502
15.6
22.7
17.1
for NOX
PTUDR
(kg/s)
34,900
15.8
6.1
52.8
35,000
NOX emissions are controlled (31 percent reduc-
tion), potential flue gas hazard is reduced, reduc-
ing overall total source potential hazard.
A second example serves to illustrate how
the SAM/IA model can be used to identify Level
2 sampling needs based on Level 1 results. This
example also introduces the concept of
"looping," in which results from successive Lev-
el 1 analytical steps are used to rule out the
existence of certain compound categories in a
sample and thereby decrease the calculated
PDOH.
The concentration of total vapor phase organ-
ic compounds (as collected by the organic mod-
ule of the source assessment sampling system
[8AS8] train) in the coal feeder vent discharge
stream was 780 mg/m3 as shown by recent data
from a Level 1 source test of a low-Btu gasifier."
With liquid chromatography separation, the
LC6 fraction accounted for 79 mg/m8. Based on-
ly on this information, the calculated PDOH for
the LC6 fraction would be 460 as shown in Table
3, based on the conservative assumption that
the entire LC6 fraction consisted of the most
toxic species potentially present in the fraction:
2-aminonaphthalene. This LC6 PDOH would
thus be added to those for the other organic and
inorganic categories to obtain the stream total
PDOH. Further Level 2 analyses would be re-
quired for at least the 38 compounds in MEG
categories 5, 8,10,13,19, and 23 listed in Table
3. These compounds have MATE values less
than 79 mg/m , so based only on the information
that compounds eluting in the LC6 fraction are
emitted at 79 mg/m3, these compounds would be
flagged for Level 2 elucidation.
However, when results from the LC fraction
infrared analysis are included with the TCO and
gravimetric data, calculated PDOH decreases.
In the example the IR report noted that the
sample consisted of phenols and cresols (MEG
category 18), carboxylic acids (category 8), and
heterocyclic nitrogen compounds (category 23).
Thus, assuming that amines (category 10), thiols
(category 13), alcohols (category 5), and halo-
phenols (category 19), are present only in negli-
gible quantities, the calculated PDOH for the
LC6 fraction decreases to 360, based on
dibenz(a.h) acridine, as shown in Table 4. The
table also indicates that Level 2 elucidation of
28 species in MEG categories 8,18, and 23 would
now be suggested.
Finally, incorporating results from the low-
resolution mass "spectrometry analysis further
reduces the calculated PDOH and the scope of
needed Level 2. In the example, the LRMS re-
port noted the strong presence of phenols (in-
tensity - 100) and a weaker presence of heter-
ocyclic nitrogen compounds (intensity - 10).
Based on this, we can assume that the maximum
28
-------
TABLE 3. LOW-Btu GASIFIER, COAL FEEDER VENT ORGANIC EXTRACT:
LC6 EVALUATION: TCO/GRAV DATA
MEG
Number
10C220
23B240
23B220
10C080
13A140
23C160
10A020
18B060
13A100
23C020
23D020
IOC 100
23D040
08D280
23C180
ISA 120
23B200
23C040
23C140
ISA 140
18A180
23B020
05B100
23C120
18B020
18A040
23C080
19B020
23B260
18A100
18A160
10D020
18A080
18A060
18B080
23C060
23B040
10B100
Compound
2-Aminonaphthlene
Dibenz (a,h) acridine
Dibenz (a,j) acridine
Ansidines
Perch loromethaneth io 1
Dibenzo (c,g) carbazole
Methyl ami ne
1,4-Dihydroxybenzene
Benzenethiol
Pyrrole
Benzothiazole
1,4-Diaminobenzene
Methyl benzothiazoles
Phthalate esters
Dibenzo (c,g) carbazole
2,2'-Dihydroxydiphenyl
Benz (c) acridine
Indole
Dibenzo (a, i) carbazole
Xylenols
Polyalkyl phenols
Qu i no 1 i ne
1-Phenylethanol
Benzo (a) carbazole
Catechol
Cresols
Carbazole
Chlorinated cresols
Dibenz (c,h) acridine
Phenylphenols
Alkyl cresols
N,N-Dimethly aniline
Ethylphenols
2-Methoxyphenol
1,2,3-Trihydroxybenzene
Methylindoies
2-Methylquinoline
Morpholine
MATE
0.17
0.22
0.25
0.50
0.80
1.0
1.2
2.0
2.1
2.7
4.3
4.5
4.7
5.0
6.0
6.8
11
11
12
13
15
16
18
19
20
22
23
23
23
23
24
25
25
33
36
45
55
70
PDOHa (Entire Assayed
Level is the Compound)
470
360
320
160
99
79
66
40
38
29
18
18
17
16
13
12
7.2
7.2
6.6
6.1
5.3
4.9
4.4
4.2
4.0
3.6
3.4
3.4
3.4
3.4
3.3
3.2
3.2
2.4
2.2
1.8
1.4
1.1
Emission level = 79 mg/nf
29
-------
TABLE 4. LOW-Btu GASIFIER, COAL FEEDER VENT ORGANIC EXTRACT:
LC6 EVALUATION; IR + TCO/GRAV DATA
MEG
Number
23B240
23"B220
23C160
18B060
23C020
23D020
23D040
08D280
23C180
18A120
23B200
23C040
23C140
18A140
18A180
23B020
23C120
18B020
18A040
23C080
23B260
ISA 100
18A160
18A080
18A060
18B080
23C060
23B040
Compound
Dibenz (a,h) acridine
Dibenz (a,g) acridine
Dibenzo (c,g) carbazole
1,4-Dihydroxybenzene
Pyrrole
Benzothiazole
Methyl benzothiazoles
Phthalate esters
Dibenzo (a,g) carbazole
2,2' -Di hydro xydiphenyl
Benz (c) acridine
Indole
Dibenzo (a,i) carbazole
Xylenols
Polyalkyl phenols
Quinoline
Benzo (a) carbazole
Catechol
Cresols
Carbazole
Dibenz (c,h) acridine
Phenylphenols
Alkyl cresols
Ethylphenols
2-Methoxyphenol
1,2,3-Trihydroxybenzene
Methyl indoles
2-Methylquinoline
MATE
(mg/m3)
0.22
0.25
1.0
2.0
2.7
4.3
4.7
5.0
6.0
6.8
11
11
12
13
15
16
19
20
22
23
23
23
24
25
33
36
45
55
PDOHa (Entire assayed
level is the compound)
360
320
79
40
29
18
17
16
13
12
7.2
7.2
6.6
6.1
5.3
4.9
4.2
4.0
3.6
3.4
3.4
3.4
3.3
3.2
2.4
2.2
1.8
1.4
Emission level = 79 mg/nf
concentration of LC6 category 18 compounds
would be 72 mg/m3 and that the maximum con-
centration of category 23 species would be 7.2
mg/m3, with negligible category 8 compounds
present. This information reduces the calcu-
lated PDOH for the LC6 fraction to 36, based on
1,4-dihydroxybenzene, as shown in Table 5. The
table also shows that Level 2 would now be in-
dicated for only 18 compounds in MEG cate-
gories 18 and 23.
INTERMEDIATE-SCREENING SAM/I
The SAM/I model has been designed for in-
termediate screening purposes to evaluate Lev-
el 1 (optionally) and Level 2 data. To address
these objectives, the model includes elementary
treatment of pollutant dispersion or dilution to
ambient levels but does not incorporate am-
bient chemical reaction or transformation. Be-
cause pollutant dispersion is treated, potential
30
-------
TABLE 5. LOW-BTU GASIFIER, COAL FEEDER VENT ORGANIC EXTRACT:
LC6 EVALUATION; LRMS + IR + TCO/GRAV DATA
MEG
Number
18B060
236 240
23B220
18A120
23C160
18A140
18A180
18B020
18A040
18A100
18A160
18A080
23C020
18A060
18B080
23D020
23D040
23C180
Compound
1,4-Dihydroxybenzene
Dibenz (a,h) acridine
Dibenz (a,j) acridine
2,2-Dihydroxydiphenyl
Dibenzo (c,g) carbozole
Xylenols
Polyalkyl phenols
Catechol
Cresols
Phenylphenols
Alkylcresols
Ethylphenols
Pyrrole
2-Methoxyphenol
1,2,3-Trihydroxybenzene
Benzothiazole
Methyl benzothiazoles
Dibenzo(a,g) carbazole
MATE
(mg/m3)
40
0.22
0.25
6.8
1.0
13
15
20
22
23
24
25
2.7
33
36
4.3
4.7
6.0
PDOHa (Entire assayed
level is the compound)
36
33
29
11
7.2
5.5
4.8
3.6
3.3
3.1
3.0
2.9
2.7
2.2
2.0
1.7
1.5
1.2
aEmission level = 72 mg/m 3for category 18 species.
7.2 mg/m for category 23 species.
hazard estimates employ the minimum ambient
level goal (ALGm) MEG.
As in SAM/IA, two hazard indices are defined
in SAM/I: the PDOH and the PTUDR. Here,
though the PDOH is defined as the ratio of the
estimated maximum ambient concentration of a
pollutant species resulting from the effluent
stream to the ALGm for that species:
PDOH, -
Estimated maximum ambient
concentration of compound i
Minimum ambient level goal
for compound i
Again, the PDOH is a measure of the existence
of a potential hazard. Correspondingly, the
PTUDR is defined as the product of the PDOH
with the effluent stream mass discharge rate
and represents a measure of the magnitude of
the potential hazard:
PTUDRj - PDOHj 'Stream mass flow rate.
The PDOH is calculated for each pollutant
species analyzed in the discharge stream, or in
the case of Level 1 evaluations, for all species in
an analyzed Level 1 sample fraction whose po-
tential ambient level concentration exceeds its
corresponding ALGm. The PTUDR is calculated
for each species whose PDOH is greater than
unity. Stream total PDOH and PTUDR and
source total PTUDR are obtained as they were
in SAM/IA, with specific provision for incor-
porating the concept of "looping" described
above.
To obtain the estimated maximum ambient
concentration of a pollutant because of the dis-
charge stream, SAM/I employs approximate
dispersion models to account for the dilution of
a discharge concentration to an ambient concen-
31
-------
tration. Models have been developed for gas-
eous, liquid, and solid discharges into appro-
priate receiving bodies within air, water, and
land media. Figure 3 illustrates the discharge
stream/receiving body combinations treated.
The figure shows that any given gaseous, liquid,
or solid waste stream from a source can be dis-
charged in a number of ways to air-, water-, or
land-receiving media. For example, a liquid or
solid stream can be discharged to a river-, lake-,
or ocean-receiving body. In these cases the final
receptor medium is surface water; thus, the use
of water MEGs is appropriate for potential haz-
ard evaluations. Adverse effects both to human
health and to ecosystems are possible for river-
and lake-receiving bodies, so health and ecologi-
cal evaluations are appropriate. For ocean
dumping, only ecological evaluations are mean-
ingful because direct human health impacts
from ocean dumping are assumed negligible.
Similarly, liquid and solid streams can be dis-
charged to deep well, sump (or waste pond), irri-
gated field, wastepile, plowed field, cavity, or
fill site-receiving bodies in the land medium. For
liquid discharges and solid leachates, the final
receptor medium is groundwater, so water
MEG, health-based evaluations are appropriate.
For leached soil residue the final receptor
medium is the land, so land MEG, ecologically
based evaluations are appropriate.
The underlying physical picture for all the
SAM/I dispersion models is that of a discharge
stream entering an entraining ambient flow.
After mixing takes place, the pollutant stream
dispersion, or dilution factor can be approx-
imated by the ratio of the entraining stream
volumetric flow rate to the discharge stream
flow rate. SAM/I defines a discharge stream di-
lution factor, K, in just such a manner:
K -
Entraining stream volumetric flow rate
Discharge stream volumetric flow rate
Therefore, the estimated maximum ambient
concentration for a pollutant species is the ratio
of the discharge concentration to the dilution
factor.
Dilution factors have been defined for all the
receiving boilers shown in Figure 3. In the dis-
persion models used to calculate these dilution
factors, entraining flow characteristics and cer-
tain discharge stream characteristics have been
internally parameterized based on estimates of
nationwide averages of these characteristics.
Thus, only discharge stream flow rate remains a
model variable. Further, several model dis-
charge stream flow rates have been defined,
spanning discharge flow rate range of interest.
Typical dilution factors have been assigned to
each of these model streams. Therefore, the
SAM/I user need only know the discharge rate
of the stream under evaluation, and receiving
body discharged into, to perform SAM/I calcula-
tions.
As an example, for gaseous effluent streams
discharged into the atmosphere, a Gaussian
plume dispersion model6 was used to predict
maximum ground level pollutant concentra-
tions. Here, the entraining flow is the atmos-
phere. The entraining flow characteristics, at-
mospheric stability, and wind speed are given
values within the model typical of nationwide
average conditions. Further, discharge stream
stack height is internally parameterized by
relating average stack height to average source
flow rate (e.g., large utility power plants,
sources with flue gas discharge rates in the
Mg/s range have stack heights around 200 m,
whereas small commercial or industrial boilers,
with flue gas flow rates in the kg/s range have
about 10 m stacks). Thus, for SAM/I evaluations
a user need only know discharge flow rate to be
able to assign an approximate dilution factor.
The defined SAM/I dilution factor, as a func-
tion of effluent stream discharge rate, for the
various effluent stream/receiving bodies is sum-
marized in Table 6. Details of the models used to
assign these dilution factors are reported else-
where.2
An example of the use of the SAM/I model is
presented in Table 7, where the Level 1 inor-
ganic analysis data for the coal-fired utility
boiler, treated by SAM/IA in Table 1, is eval-
uated through SAM/I. In this example, one
notes that the flue gas flow rate is 69.3 kg/s.
Reference to Table 6 requires a dilution factor
of 1,000. Calculated PDOH and PTUDR values
for the 26 components assayed that have ALG
values as well as stream totals are shown in
Table 7. Further use of SAM/I calculations in
evaluating control technology application and in
identifying Level 2 analysis needs is analogous
to the use of SAM/IA as presented in Tables 2
through 5.
32
-------
Waste Stream
Gas, Liquid, Solid)
Control
Device
O On
D Off
\
\
\
»
Gas
Residual
Liquid
Residual
1
Solid
Residual
nv\»dviiiy
Medium
^. Aim
X
Lx^ |
^^Land-l
.s*
^ J
a) S. water. Surface water, G. water. Ground water
b) A: Air, W: Water, L: Land
c) H: Health, E: Ecological
__ Recehfino ^
Body **
fRtver
JLake
^^ lOcean
[ fDeepWell "
JSump
| — — •"" ^Irrigated Reid
fRrver
I -4 Lake
| [Ocean
\ Surface-j plowed Reid
j LSump
Interior fcavWy 1
\FIII Site
Final
1 Receptor
Medium
Air
S. water
S. water
S. water
^G. water
S. water
S. water
S. water
G. water,
'Land
G. water,
Basis
b)
A
W
W
W
W
W
W
W
r
f Receptor
c)
H,E
H.E
H,E
E
H
H, E
H, E
E
H
E
W 1 H
Land L ', E
Figure 3. SAM/I pollutant discharge overview.
-------
TABLE 6. SUMMARY OF MODEL STREAM DISCHARGE RATES AND DILUTION FACTORS
Discharge Stream Type
Receiving Body
Discharge Rate Q (g/s)
and Dilution Factor K
Gas
Air
0
2.5 x 106
6.5 x 105
1.3 x 105
6.8 x 103
5.4 x 102
K
1 x 102
3 x 102
1 x 103
5 x 103
2 x 104
Liquid/Soluble Solid
River/Lake
0
1 x 105
1 x 104
1 x 103
1 x 102
1 x 101
K
1.6 x 102
1.6 x 103
1.6 x 104
1.6 x 105
1.6 x 106
Ocean
0
3 x 104
(Barge)
K
1 x 103
Discharge Stream Type
Receiving Body
Discharge Rate Q (g/s)
and Dilution Factor K
Liquid
Deep Well
Q
Any
K
1
Liquid/Soluble Solid
Irrigated
Field
Q
Any
K
100
Sump, Waste Pile, Plowed
Field, Cavity, Fill Site
Q
Any
K
10a
10Qt>
Leached Solid
Any Land Body
Q
Any
K
1
aLarge receiving body with base diameter d > 10m.
Large receiving body with base diameter d < 10m.
T-1688
REGIONAL SITE EVALUATION, SAM/11
The SAM/II model will be designed for re-
gional site evaluation purposes for specific
evaluation of Level 3 data. It will be the most
mathematically detailed model in the SAM
series in its treatment of pollutant dispersion
and will include treatment of population ex-
posed to ambient levels to measure the impact
of a potential environmental hazard. Where
possible, the model will factor in pollutant
species chemical transformations.
Individual components of the SAM/II model
are presently being developed. To date, only the
formulation for gaseous stream emissions to the
atmosphere is sufficiently developed to be re-
ported. In this model the hazard index used is
termed the potential impact factor, I. This is
defined to be the sum of the number of people
exposed to ambient pollutant levels, weighted
by the ambient PDOH exposure, wherever the
ambient PDOH exceeds 0.1. Mathematically, the
potential impact factor can be expressed as:
I - £ P.PDOH,.dA; PDOH.stO.l .
Here, i denotes a pollutant species, P is the ex-
posed population (function of A), PDOHj is the
ambient potential degree of hazard as defined in
the SAM/I model (also function of A), and A de-
notes the area of integration, defined as being
that area where PDOH; exceeds 0.1.
The same Gaussian dispersion model used in
SAM/I is employed to estimate PDOHj as a func-
tion of distance for the discharge. However, in
SAM/II source characteristics (stack height, ef-
fluent flow rate) are not parameterized and
model sources are not defined. Instead, these
characteristics are treated as user-supplied in-
puts.
Table 8 is an example of the use of this impact
34
-------
TABLE 7. PDOH AND PTUDR FOR UTILITY BOILER FLUE GAS (INORGANIC): SAM/I
MEG
Category
32
36
41
42
45
46
47
49
50
51
53
54
55
62
65
69
71
72
74
78
81
82
83
TOTAL
Component
Be
Ba
Tl
CO
Sn
Pb
NOv
NHj
As
Sb
B1
SO?
SOs
S04
Se
Te
Ti
V
Mo
Mn
Fe
Co
Cu
Zn
Cd
Hg
Flue Gas
Concentration
(yg/dscm)
9.0
2250
2.6
3.07 x 104
6.4
74
1.16 x 106
10.5
95
3.9
44
4.18 x 106
1.45 x 104
6500
9.9
4.1
6100
260
150
240
4.5 x 104
66
280
420
1.8
3.1
ALGm:
Health
(yg/m3)
0.01
1.0
0.24
1.0 x 104
0.24
0.36
100
43
0.005
1.2
0.7
80
2.4
2.4
0.03
0.24
14
1.2
12
12
107
0.10
0.50
9.5
0.02
0.01
PDOH: a
Health
0.09
2.3
0.011
0.003
0.03
0.20
12
2.4 x lO'4
19
0.003
0.063
52
6.0
2.7
3.3
0.017
0.44
0.22
0.012
0.020
0.42
0.66
0.56
0.044
0.090
0.31
100
PTUDR :b
Health
(Mg/s)
0.16
0.83
1.3
3.6
0.42
0.19
0.23
6.7
n • -i j. • £ j. * innn nnrtli
Dilution factor of 1000; PDOH = IQQQ
Flue gas flow rate of 69.3 kg/s.
factor formulation in ranking the potential envi-
ronmental hazard of stationary combustion
sources.7 The table shows calculated potential
impact factor for flue gas emissions of the cal-
culated 10 potentially most hazardous sources.
Total emissions estimates for the criteria pollut-
ants—NOX, SOX, CO, and hydrocarbon—with
the addition of particulate phase sulfates, trace
elements, and polynuclear aromatic compounds
were used in the calculation, along with esti-
mates of nationwide urban and rural population
densities. The table shows, not surprisingly,
that coal-fired utility and industrial sources
dominate the potential hazard ranking.
35
-------
TABLE 8. POTENTIAL IMPACT FACTOR RANKING FOR STATIONARY CONVENTIONAL
COMBUSTION SOURCES: FLUE GAS EMISSIONS
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
Equipment Type/Fuel
Small Watertube Stoker — Coal
Small Firetube Stoker — Coal
Tangential Utility — Coal
Wall Fired Utility — Coal
Wall Fired Industrial -- Coal
Large Watertube Stoker — Coal
Vertical & Stoker — Coal
Cyclone Utility — Coal
Opposed Utility « Coal
Tangential Utility — Oil
Potential Impact Factor
6.7 x 1011
5.6 x 1011
1.9 x 10H
1.1 x 10}}
7.8 x 1011
7.6 x 1010
5.7 x 1010
4.1 x IQlO
2.1 x 1010
2.7 x 109
SUMMARY
A series of source analysis models for
evaluating tiered environmental assessment
sampling and chemical analysis results in terms
of quantifying the potential environmental im-
pact of a discharge stream or pollutant source is
under development. Elements of the form of
each of these have been presented and illus-
trated through several example applications
demonstrating potential uses in an environmen-
tal assessment.
REFERENCES
1. Schalit, L. M., and K. J. Wolfe. SAM/IA: A
Rapid Screening Method for Environmental
Assessment of Fossil Energy Process Ef-
fluents. EPA-600/7-78-015. February 1978.
2. Anderson, L. B., et al. SAM I: An Interme-
diate Screening Method for Environmental
Assessment of Fossil Energy Process Ef-
fluents. Acurex Corp., Mountain View, Calif.
Acurex Report TE-79-154. December 1978.
3. Cleland, J. G., and G. L. Kingsbury. Multime-
dia Environmental Ooals for Environmental
Assessment, Volume L EPA-600/7-77-136a.
November 1977.
4. Kingsbury, G. L. Master List of Organic
Substances to be Addressed by Multimedia
Environmental Goals. Research Triangle In-
stitute. Research Triangle Park, N.C. Oc-
tober 1978.
5. Page, G. C. Environmental Assessment-
Source Test and Evaluation Report—Chap-
man Low-Btu Gasification. EPA-600/
7-78-202. October 1978.
6. Turner, D. B. Workbook of Atmospheric Dis-
persion Estimates. U.S. Public Health Ser-
vice AP-26.1970.
7. Salvesen, K. G., et al. Emission Characteriza-
tion of Stationary NOX Sources: Volume I
Results. EPA-600/7-78-120a. June 1978.
36
-------
INTERAGENCY RESEARCH ON THE ENVIRONMENTAL TRANSPORT AND
EFFECTS OF SYNFUELS-RELATED SUBSTANCES
W. Gene Tucker*
Industrial Environmental Research Laboratory,
U.S. Environmental Protection Agency, Research Triangle Park, North Carolina
and
Gerald J. Rausa
Energy Effects Division,
U.S. Environmental Protection Agency, Washington, D.C.
Abstract
The Interagency Energy/Environment R&D
Program, initiated in late 1974, comprises over
300 major research and development projects
being conducted by 17 Federal agencies and de-
partments. The program is planned and coordi-
nated by the U.S. Environmental Protection
Agency's (EPA) Office of Research and Devel-
opment. The projects in this program cover a
wide spectrum of investigation—from basic and
applied research on the movement of energy-re-
lated substances through the environment and
their health and ecological effects, to develop-
ment of systems for control of these substances,
to socioeconomic studies of the future impacts
of the U.S. energy development.
This presentation gave an overview of the ob-
jectives and the general status of selected proj-
ects under the interagency program. The proj-
ects described were selected to present a cross-
section of the work being done on the health and
ecological effects, and transport through the
environment, of substances potentially released
by synthetic fuel production and use.
The presentation on this subject was made by
Dr. Tucker. He began by briefly reviewing the
history of the Interagency Energy/Environmen-
tal R&D Program, the Federal agencies in-
volved, the energy processes of concern, and the
historical breakdown of funding for various re-
search areas. Documentation of this informa-
tion is available in References 1 through 14.
Of the 200-plus projects sponsored under this
program that relate to the movement and fate of
substances in the environment and their effect
on human and ecological health, approximately
*Speaker.
50 deal with synthetic fuels processes or sub-
stances that could be released from synfuels
production or use. For purposes of this presenta-
tion, 20 of those projects were briefly described.
These projects were selected to illustrate the
breadth of ongoing research in the biological
and physical sciences and how it complements
the Environmental Assessment and Environ-
mental Control Technology programs that were
the primary topics of this symposium.
The projects were organized into four general
areas:
• Human health effects
• Ecological effects
• Transport and fate
• Measurement and instrumentation
The projects that were discussed are listed in
Tables 1 through 4. Project personnel and refer-
ence documents are listed for those who are in-
terested in obtaining detailed information on
the individual projects.
There is potential for mutual benefit from
greater contact between some of these projects
and the various synfuels environmental assess-
ments being sponsored by the U.S. Environ-
mental Protection Agency (EPA) and other Fed-
eral and private groups. A recommendation was
made that presentations on several transport
and effects projects be included on the program
for the next EPA synfuels symposium.
REFERENCES
1. Interagency Energy/Environmental R&D
Program. U.& Environmental Protection
Agency. EPA-600/7-77-007. March 1977.
2. Who's Who in the Interagency En-
ergy/Environmental R&D Programs IV.
U.S. Environmental Protection Agency.
37
-------
TABLE 1. SELECTED HUMAN HEALTH EFFECTS PROJECTS FROM THE
INTERAGENCY ENERGY/ENVIRONMENTAL R&D PROGRAM
PROJECT
CONTACT
REFERENCE
Repository of Samples
D. L. Coffin
EPA/HERL
Research Triangle Park
North Carolina 27711
(919) 541-2586
FTS 629-2586
Genetic and
Carcinogenic Hazards
Numerous contacts;
see ref. 3, pp. 9-10
3, 5, 6, 7, 8
Extrapolation and
Risk Assessment
D. G. Hoel
NIEHS
Research Triangle Park
North Carolina 27711
(919) 541-3441
FTS 629-3441
2, 3
Industrial Hygiene
A. Thomas
NIOSH
5600 Fishers Lane
Room 8-48
Rockville, MD 20857
(301) 443-3843
3, 7
38
-------
TABLE 2. SELECTED ECOLOGICAL EFFECTS PROJECTS FROM THE
INTERAGENCY ENERGY/ENVIRONMENTAL R&D PROGRAM
PROJECT
CONTACT
REFERENCE
Aquatic Effects of
Synfuel Discharges
K. E. Biesinger
EPA/ERL
6201 Congdon Blvd.
Duluth, MN 55804
(218) 727-6692
FTS 683-9512
H. L. Bergman
University of Wyoming
Laramie, WY 82071
(307) 766-4330
3, 8
Coastal Ecosystems
H. Tait
USFWS
NSTL Station, MS
(601) 688-2091
FTS 494-2091
3, 4, 5
39529
E. D. Schneider
EPA/ERL
South Ferry Road
Narragansett, RI 02882
(401) 789-1071
FTS 838-4843
Vegetative Stabilization
of Spent Shale E. F. Harris
EPA/IERL
5555 Ridge Avenue
Cincinnati, OH 45268
(513) 684-4417
7, 8
Subsidence from In-Situ
Coal Gasification
E. R. Bates
EPA/IERL
5555 Ridge Avenue
Cincinnati, OH 45268
(513) 684-4417
39
-------
TABLE 3. SELECTED TRANSPORT AND FATE PROJECTS FROM THE
INTERAGENCY ENERGY/ENVIRONMENTAL R&D PROGRAM
PROJECT
Dynamics of Refinery/
Petrochemical Wastes
in Marine Waters
Dynamics of Refinery
Wastes in Lake Michigan
CONTACT
H. M. McCammon
DOE/OHER
Washington, D.C.
(301) 353-5547
FTS 233-5547
Same as above.
REFERENCE
3, 5
20545
3, 5
TABLE 4. SELECTED MEASUREMENT AND INSTRUMENTATION PROJECTS
FROM THE INTERAGENCY ENERGY/ENVIRONMENTAL R&D PROGRAM
PROJECT
Secondary Organic Air
Pollutants from Gasifi-
cation Plants
Composition of Synfuel
Wastes
Portable GC for
Organics
Standard Reference
Materials
CONTACT
R. K. Patterson
EPA/ESRL
Research Triangle Park
North Carolina 27711
(919) 541-2254
FTS 629-2254
A. Alford
EPA/ERL
Athens, GA 30605
(404) 546-3525
FTS 250-3525
L. Doemeny
NIOSH
4676 Columbia Parkway
Cincinnati, OH 45226
FTS 684-4266
C. Gravatt
NBS
Washington, D.C. 20234
(301) 921-3775
REFERENCE
14
3, 4
40
-------
EPA-600/9-78-002 (NTIS Number PB 284
375). June 1978.
3. Who's Who V: The Interagency En- 9.
ergy/Environmental R&D Program Direc-
tory and Index. U.S. Environmental Pro-
tection Agency. EPA-600/9-79-017. June 10.
1979.
4. Interagency Energy/Environmental R&D
Program—Status Report III. U.S. Environ-
mental Protection Agency. EPA-600/7- 11.
77-032 (NTIS Number PB 265 443). April
1977.
5. Fiscal Year 1976/Health and Environmen-
tal Effects Research Program Abstracts.
U.S. Environmental Protection Agency. 12.
EPA- 600/7-77-004 (NTIS Number PB 265
381). January 1977.
6. Interagency Program in Energy-Related
Health and Environmental Effects Re-
search: Project Status Report. U.S. Envi- 13.
ronmental Protection Agency. EPA-600/7-
79-009 (NTIS Number PB 290 578). January
1979.
7. EPA Program Status Report: Oil Shale. 14.
U.S. Environmental Protection Agency.
EPA-600/7-78-020. February 1978.
8. EPA Program Status Report: Oil Shale
1979 Update. U.S. Environmental Protec-
tion Agency. EPA-600/7-79-089. March
1979.
Energy/Environment II. U.S. Environmen-
tal Protection Agency. EPA-600/9-77-025.
November 1977.
Energy/Environment III. U.S. En-
vironmental Protection Agency.
EPA-600/9-78-022 (NTIS Number PB 290
558). October 1978.
Gage, Stephen J., et al. Final Report of the
Interagency Working Group on Environ-
mental Control Technology for Energy
Systems. The Council on Environmental
Quality. November 1974.
King, Donald, and Warren R. Muir, et al.
Report of the Interagency Working Group
on Health and Environmental Effects of
Energy Use. The Council on Environmental
Quality. November 1974.
Ray, Dixy Lee. The Nation's Energy Fu-
ture. U.S. Atomic Energy Commission.
WASH-1281, U.S. GPO Stock Number
5210-00363. December 1973.
Identification of Components of Energy-
Related Wastes and Effluents. U.S. Envi-
ronmental Protection Agency. EPA-600/7-
78-004 (NTIS Number PB 280 203). January
1978.
41
-------
DEPARTMENT OF ENERGY ENVIRONMENTAL ASSESSMENT PROGRAM
FOR COAL CONVERSION
F. E. Witmer
Environmental Control Technology Division,
U.S. Department of Energy, Washington, D.C.
Abstract
Implicit in the U.S. Department of Energy's
(DOE) charge to foster the development of com-
mercially viable coal gasification and liquefac-
tion processes is the requirement that this tech-
nology be environmentally acceptable. While
wide-scale use of this technology over the next
several decades is not predicted, synfuel alterna-
tives such as coal conversion will significantly
contribute to domestic energy supply over the
long term.
DOE's environmental assessment activity,
oriented toward evaluating the environmental
impact ultimately associated with a sizable syn-
thetic fuels industry, is conducted to guide here-
and-now RD&D and policy decisions. A series of
time-phased environmental assessments paral-
lel the development and scaleup of the technol-
ogy. Major areas of environmental concern and
uncertainty are singled out in NEPA documen-
tation that accompanies scaleup activity. En-
vironmental concerns that go beyond current
regulatory and siting requirements for energy
technologies are addressed in DOE documents
including environmental development plans
that are prepared for each emerging technology.
Individual, detailed project milestones require
formalized environmental status reports to en-
sure that environmental concerns and issues are
satisfied.
It is the objective of DOE environmental as-
sessment to look beyond the single demonstra-
tion plant facility and to project the potential
impact of a mature coal conversion industry.
Such an assessment is complex and ambitious.
It involves integrating and synthesizing a num-
ber of environmental factors—emission, ef-
fluent, and solid waste source characterizations;
control capabilities; health effect determina-
tions; anticipation of regulatory requirements;
resource demands; social-economic consid-
erations; and cost-benefit analyses.
Because complex technical, economic, and
public issues are at stake, conclusions tend to be
judgmental and, of course, are sensitive to the
scenario under consideration. DOE has a com-
plement of interdepartmental and intradepart-
mental activities to expand the data base, both
in the field and in terms of studies, to improve
the analysis process and the credibility of the
assessments. This presentation will overview
these evolving assessment processes.
INTRODUCTION
Webster defines assessment as the act of de-
termining the importance, size, or value of a
given thing. The U.S. Department of Energy's
(DOE) environmental assessment activity is con-
cerned with:
• Evaluating human and ecological effects of
the environmental intrusions that are asso-
ciated with energy conversion processes,
especially those involving coal conversion;
• Fully characterizing the nature of these pol-
lutant releases; and
• Determining the efficacy and practicability
of control technology that is deployed to mit-
igate and limit such releases.
Obviously, assessment is highly dependent on
the qualities singled out in accounting. In the
area of synfuels conversion, considerable effort
is needed to help define and quantify the con-
trolling environmental qualities. Ideally, the as-
sessment will key to those areas where signifi-
cant environmental impacts are experienced.
In this presentation I hope to:
• Briefly outline some of the obvious difficul-
ties confronting an environmental assess-
ment activity, the primary one being the
lack of hard operating data for synfuels facil-
ities;
• Describe the manner and means by which
DOE is forming a set of interdisciplinary
teams to address a series of process/site-
specific environmental characterizations to
43
-------
DOE ENVIRONMENTAL ASSESSMENT PROGRAM FOR COAL CONVERSION
BACKGROUND
• MOVING TARGETS AND COMPLEXITY OF PERFORMING OVERALL INDUSTRY
ORIENTED ASSESSMENT
• EVOLVING DOE ORGANIZATION
• DEVELOPING ENVIRONMENTAL ASSESSMENT PROGRAM
PROCEDURES TO ENSURE "ENVIRONMENTAL ACCEPTABILITY"
• PLANNING ANO DOCUMENTATION
• COMPLIANCE REGULATIONS - OFFICE OF FOSSIL ENERGY PROGRAMS (FEP)
• FUTURE STANDARDS AND COMMERCIALIZATION IMPACTS - ASSISTANT
SECRETARY FOR ENVIRONMENT (ASEV)
• PROCESS/SITE SPECIFIC CHARACTERIZATION AND ASSESSMENTS
ON.QQINO ACTIVITIES
• BASE PROGRAM HIGHLIGHTS
• CONTROL TECHNOLOGY ASSESSMENTS
SUMMARY
Figure 1. DOE environmental assessment program for coal conversion.
fulfill the aforementioned data need; and
• Indicate the important and significant role
generic environmental (core program) R&D
has had and will continue to have on such an
endeavor (Figure 1).
The intent is to recognize the diverse environ-
mental assessment and acquisition activity on-
going within the department.
BACKGROUND
Public Law 95-91, which on August 5, 1977
created the U.S. Department of Energy (DOE),
charges the Agency with promoting and devel-
oping "environmentally acceptable" energy
technologies. Recent legislation to further guide
EPA regulatory efforts in the area of emissions,
the Clean Air Act Amendments of 1977, speaks
of encouraging the use of best control technol-
ogy that is "economically achievable." DOE's
main thrust is to provide and support economic
energy options, while EPA is dedicated to
preserving and improving the quality of the en-
vironment, cost and resource considerations
having been of critical although secondary im-
port. Administration officials indicate that in-
creased attention is to be given to economic and
resource considerations in the setting of emis-
sions standards. This is as it should be, for ideal-
ly DOE and EPA form a true complement in sus-
taining and improving life quality. Enhanced
energy and environmental goals are insepa-
rable.
One should take a systems approach in arriv-
ing at "preferred energy choices." "Envi-
ronmentally acceptable," "economically achiev-
able," and "preferred choices" are "soft," quali-
tative terms that are in the process of acquiring
quantitative meaning as legislation is imple-
mented and new energy options and concomi-
tant environmental regulations develop. But
this takes time. With synfuels there are pres-
ently a number of moving targets: changing
resource availability, improvements to and new
processes for the technology, promulgation, and
tightening of environmental standards, and
changing economic climate. In addition to
"uncertainty" because of the evolving nature of
the technology and boundary conditions in
which the technology must operate is the uncer-
tainty associated with incomplete knowledge.
This is especially critical to the life sciences, to
allow prediction of long-term human health and
ecological impacts. The situation is complicated
by an overlap of largely subjective social-
economic value judgments and of speculative
future energy resource demands.
In an ideal decisionmaking process, compre-
hensive environmental assessment of "benefits"
vs. "costs" might be made to establish priorities
for energy options. One could envision the selec-
tion of individual "benefit" criteria (energy
44
-------
COST
CONTRIBUTION
PER UNIT
PRODUCT, t
INDEPENDENT
LEVEL OF CONTROL
EXAMPLES OF CONTROL SYNERQISMS
100%
MULTIPURPOSE CONTROL
1. PHYSICAL COAL CLEANING
2. ACTIVATED SLUDOE
3. SCRUBBER
POLLUTANT
X BULPIDES
V ASH
INDEPENDENT CONTROL
CHEMICAL EXTRACTION
PULVERIZATION/FLOATATION
CELLS
X REFRACTORY ACTIVATED CARBON/
ORGANIC* CHEMICAL OXIDATION
V OfORADABLE HOLDUMPOND
OROANICS
X PARTICULATEI
V SULFUR DIOXIDE
ESP/BAQ HOUSE
INCREASED CONTACT AND
REAOINT
Rgure 2. Examples of control synerglsms.
availability, jobs, affluence) and corresponding
"cost" criteria (environmental control costs,
health and safety risks, loss in aesthetics, etc.).
The weak link in such a comparison is the
assignment of weighing factors and, ultimately,
values to "equate" the individual criterion. At
present one is left with "apples and oranges."
Thus, the assessment by necessity becomes sub-
jective and judgmental.
Environmental Controls
In the area of environmental controls, as tech-
nology becomes more efficient, one might com-
pare process control cost and emission level for
a single pollutant. Real-world emissions, ef-
fluents, and solid waste boundary conditions,
which control plant design and operation, are
multivalued. In many instances, the function
and performance of the environmental control
processes are multipollutant-oriented (Figure 2).
Thus, even with characterization of control
hardware, analysis of cost-effectiveness relative
to complying with a set or series of environmen-
tal standards is exceedingly complex and not
wholly practical.
In scrutinizing environmental process control
costs in such a manner one must consider total
energy costs. When processes are compared, it
is erroneous to compare only environmental
control costs. Quite conceivably, an overall proc-
ess systems tradeoff may exist with low costs
45
-------
associated with the coal conversion train bal-
anced by high environmental control costs and
vice versa. Because pollution regulations are
boundary conditions, (i.e., specifications similar,
say, to product purity) that must be met by the
coal conversion process as integral parts to the
overall process, scientists are cautioned against
segregating pollution controls from the rest of
the process.
Occupational and Public Health Effects
As one can appreciate, the health inputs to a
cost-benefit analysis are several orders more
difficult. All of the coal conversion processes in-
volve bioactive materials (both organics and
trace elements) that have not been previously
produced on the scale envisaged for commercial
synfuels operations. Many of the potentially
adverse human effects are low-level and take
years to diagnose and quantify. Human (i.e.,
worker) exposure in pilot-plant operations
represents an exceedingly small "window" in
time and exposure and, because of the develop-
mental nature of such pilot operations, they can-
not be considered truly representative of com-
mercial synfuels activity. Existing sister in-
dustries; e.g., coking and petroleum industries,
are being drawn on to provide quidance. For the
present, one must resort to progressively more
sophisticated biological screening tests and
make the tenuous extrapolation to man. This
should not be construed to say that synfuels is
in a unique position because with increasing
vigilance toward toxic and bioactive materials
(RCRA and TSCA), a number of established in-
dustries are and will be subject to similar
mammal-to-man extrapolations. Emphasis is
placed on the considerable progress being made
in facilitating this animal-man linkage as part of
DOE's base/core research program. In the
human health area, the present state of
knowledge and statistical base are too uncertain
to quantitatively translate worker and ambient-
exposure levels into sickness, disease, and loss
of life, except in insolated cases.
Ecological Effects
A similar situation exists in extrapolating
ecological effects observed in pilot operations to
full-scale facilities. In many instances, the pilot
plant is located in an industrial area that is al-
ready highly contaminated, the contribution of
pilot plant being insignificant relative to the ex-
isting baseline. Again, the question arises of
how representative the operations of a pilot fa-
cility are relative to a full-scale plant. Ecological
effects tend to be regional, site and process, spe-
cific. This erects an additional barrier against
meaningful ecological input to overall technol-
ogy assessment.
Assessment Methodologies
Regional and national environmental impact
assessments suffer from similar uncertainties;
e.g., the accuracy of the dispersion models used
in analysis and assumed pollutant source re-
leases. Perhaps the weakest link, for want of
better input data, is the energy development
scenario and concomitant source terms. Normal-
ly, the synfuels technologies are expected to
comply with assumed standards, with nonregu-
lated pollutants considered in a cursory manner.
The resultant predicted ambient emission levels
are no better than these assumptions. They will,
however, give some index of potential ambient
"hot-spots" and regional problem areas. The
real uncertainty in the evaluation is how this
data translates to human health effects and life
quality. This uncertainty has been the same
problem EPA has had to wrestle with and has
dealt with primarily through the pragmatic ap-
proach of going to standards oriented at "best
available control technology" at the point of
release (e.g., as in the case of the utility in-
dustry). One can probably expect a similar ap-
proach with synfuels.
Thus, the various inputs to comprehensive
environmental assessment of future coal con-
version industry (e.g., control technology,
health, ecological, social, economic, and resource
considerations) are at various stages of develop-
ment, making it difficult to grant creditability
to any overall future environmental assess-
ment.
The approach DOE is taking with this diffi-
cult problem is to integrate the environmental
assessment activity with specific technology
and projects activities, with generic-related en-
vironmental research providing complementary
support. The organization, the methodology,
and ongoing environmental assessment activi-
ties, including general support activities, will be
46
-------
Figure 3. Department of Energy.
-------
ASSISTANT SECRETARY FOR
ENVIRONMENT
OO
OFFICE OF ENVIRONMENTAL
COMPLIANCE AND OVERVIEW
OFFICE OF TECHNOLOGY
IMPACTS
OFFICE OF HEALTH AND
ENVIRONMENTAL RESEARCH
HEALTH EFFECTS I
RESEARCH
DIVISION
ECOLOGICAL I
RESEARCH !
DIVISION I
I
.____—J
J POLLUTANT .
•CHARACTERIZATION '
aSAFETY |
RESEARCH I
DIVISION !
Figure 4. Assistant Secretary for Environment (ASEV).
-------
discussed. These activities provide target en-
vironmental assessments that will ultimately
rule on the "environmental acceptability" of a
candidate technology.
DOE ORGANIZATION AND
ENVIRONMENTAL GOALS
The Department has been in existence for
about 18 mo. It was created primarily to inte-
grate and consolidate the Federal energy pro-
grams then distributed among several agencies
with differing energy mandates and objectives.
While the basic structure has been formalized,
responsibilities are still being refined by assist-
ant secretaryships (Figure 3). Responsibility for
past synfuels environmental activities has main-
ly resided with the Assistant Secretary for En-
vironment (A8EV) and Fossil Energy Programs
(FEP) organizations (Figures 4 and 5). The fact
that environmental concern naturally spreads
from top management to the line-divisions that
are developing the technology has rightly led to
considerable environmental activity outside of
ASEV. What has developed is a logical interface
based on line-divisions having primary responsi-
bility for meeting NEPA and compliance re-
quirements, with ASEV exercising overview re-
sponsibility and implementing a comprehensive
research program oriented to environmental ef-
fects. ASEV has in large part assumed an antici-
patory role in assessing the environmental im-
pact of commercialization activities and future
standards, with ASET providing consultation.
The end product of such assessment activity
provides complete input and support to DOE
policy decisions. It is useful to single out major
organizational components within DOE's syn-
fuels environmental assessment along with key
personnel (Figure 6). The responsibility for site
and process activities (e.g., NEPA requirements
[EIS], securing permits, meeting compliance
standards both for discharges and plant opera-
tion) resides with the FEP project officer, with
assistance from an environmental support
group within FEP. For environmental informa-
tion-gathering and assessment activities beyond
those legally required for plant operation, FEP
looks for support from ASEV.
METHODOLOGY OF ENSURING
ENVIRONMENTAL ACCEPTABILITY
Major DOE programmatic efforts are re-
OFFICE OF FOSSIL
ENERGY PROGRAM!
DIVISION OF SYSTEMS
ENGINEERING
PROCESS ECONOMICS
1 ENVIRONMENT
| MATERIALS
1 1
DIVISION OF FOSSIL ! DIVISION OF FOSSIL |
FUEL EXTRACTION FUEL PROCESSING I
L _ J
1 1
1 1
DIVISION OF FOSSIL DIVISION OF
FUEL UTILIZATION MAONETOHYOROOYNAMICS
1 1
•ARTLESVILLE ETC ONAND FORKS ETC LARAMIE ETC MOROANTOWN ETC FITTSIUROH ETC
Figure 6. Office of Fossil Energy Programs (FEP).
49
-------
DOE
ORGANIZATION KEV PERSON
NATIONAL LABORATORIES VARIABLE
POLLUTANT CHARACTERIZATION * SAFETY RESEARCH DIVISION P. OUHAMEL
HEALTH EFFECTS RESEARCH DIVISION G. STAPLETON
ASEV < ECOLOGICAL RESEARCH DIVISION R.LEWIS
ENVIRONMENTAL CONTROL TECHNOLOGY DIVISION F. WITMER
OPERATIONAL AND ENVIRONMENTAL SAFETY DIVISION D. LILLIAN
.TECHNOLOGY ASSESSMENTS DIVISION B. ALMUALA
fSYSTEMS ENGINEERING DIVISION GASIFICATION B. BARATZ
I LIQUEFACTION J. NAROELLA
| FOSSIL FUEL PROCESSING DIVISION VARIABLE
^ (^FOSSIL ENERGY TECHNOLOGY CENTERS VARIABLE
EPA INDUSTRIAL ENVIRONMENTAL RESEARCH LABORATORY W. RHODES
NIOSH DIVISION OF ENVIRONMENTAL INVESTIGATIONS B. PALLAY
DIVISION OF PHYSICAL SCIENCES AND ENGINEERING J. TALTY
FEP
FUNCTION
IMPLEMENTER
OVERVIEW ft
GUIDANCE
> ADVISORY
Figure 6. Project/site-oriented environmental synfuel assessments—major participants.
quired to have a technology program task defi-
nition for budgeting and scheduling, consistent
with the developmental sequence necessary to
advance and evaluate the technology. An envi-
ronmental development plan (EDP) is prepared
to accompany and supplement the technology
program plan to assure that anticipated en-
vironmental uncertainties are addressed and
become part of the DOE program at each stage
of development, as appropriate (Figure 7). The
synfuel EDPs have been prepared jointly by
ASEV and FEP. Early versions of EDPs (FY
1977 and 1978), because of the difficulty of
achieving a meaningful technology-environ-
mental couple, listed environmental concerns
and requirements and tentative milestones for
addressing the concerns but did not assign
priorities or funding requirements to the
specific tasks. Subsequent update is anticipated
to outline environmental priorities along with
budgetary requirements.
The technology program (e.g., liquefaction,
high-Btu gasification, and low/intermediate-Btu
gasification) is comprised of a series of process-
specific projects with individual milestones and
timelines. As a project evolves from early R&D,
a series of developmental stages are oriented
toward scaleup and ultimate commercialization.
At each major phase of such a system acquisi-
tion train, ASEV, in an overview function, pre-
pares an environmental readiness document
(ERD), which critically reviews the environmen-
TECHNOLOCY
DEVELOPMENT
PLAN
ENVIRONMENTAL I
DEVELOPMENT
PLAN IEDPI
DEVELOPMENT STAGES
1
1
TECHNOLOGY
DEVELOPMENT
ENVIRONMENTAL
COMPONENT
ENGINEERING
DEVELOPMENT
ENVIRONMENTAL
COMPONENT
DEMONSTRATIONS
ENVIRONMENTAL
ASSESSMENT
COMMERCIALIZATION
PRODUCTION
OPERATION
<$>
DECISION
POINTS
A
EDP
UPDATE
Figure 7. Energy systems acquisition program—EDPs.
50
-------
DEVELOPMENT STAGES
<£
DECISIC
POINT
TECHNOLOGY ENGINEERING niu«u«T»i
DEVELOPMENT DEVELOPMENT PtMOHSTfU
ENVIRONMENTAL ENVIRONMENTAL ENVIRONM
COMPONENT COMPONENT ASSESSM
* UPDATE V'M }! V^ /
sN •» /\'l ^ /\'» /
\\r- ' "k Vv.-^^\ !
v- ENVIRONMENTAL v, >
READINESS {HO
DOCUMENT
ERD
COMMERCIALIZATION
mONS PRODUCTION
OPERATION
ENTAL
ENT
/ \^r
IT
<*
t-
ERD
Figure 8. Energy systems acquisition program—ERDs.
tal status of the developmental project and em-
phasizes environmental gaps that may exist and
that require resolution (Figure 8).
The resolution of environmental uncertain-
ties that still may exist when a process is
brought to pilot and/or demoscale is addressed
in a site- and process-specific environmental
characterization and assessment. An example of
such an activity is the Gasifiers-in-Industry En-
vironmental Assessment program currently
conducted by Oak Ridge National Laboratory
(ORNL) for ASEV, which will be reported on
later. A series of field-oriented environmental
plans are currently in preparation for major
FEP projects; e.g., H-Coal, SRC, and high-Btu
gasification. It should be noted that these
process-specific environmental evaluations are
developed jointly with FEP in concert with ad-
vice and counsel from EPA and NIOSH. Suffice
it to say that DOE does not consider "environ-
mental acceptability" and "meeting current
compliance standards" synonymous. Thus, DOE
has initiated a hierarchy of program control
documentation and planning to assure that en-
vironmental requirements are systematically
addressed in the synfuels process development
sequence. What, of course, is needed now is im-
plementation of these planning exercises in con-
cert with pilot, demo, and commercial plant ac-
tivities. Hard data above and beyond that re-
quired for compliance purposes are critically
needed. It is emphasized that normal emission
releases, effluent discharges, solid waste dis-
posal practice, and ecological characterizations
for facility operations can be adequately
handled under existing regulatory licensing and
permit procedures. The direction the detailed
incremental field environmental characteriza-
tions take must respond to the technical direc-
tion and priorities assigned to the individual
processes by FEP technologists and DOE
energy planners.
ONGOING ENVIRONMENTAL ACTIVITIES
While comprehensive, process-specific envi-
ronmental assessment studies constitute a rela-
tively new undertaking, DOE has had a broad-
based environmental characterization and as-
sessment program for the emerging synfuels
processes. Varied generic RD&D activity within
ASEV and FEP has laid a foundation for much
of the process-specific joint activity currently
underway. In recognition of ongoing environ-
mental support activity, select recent accom-
plishments of various groups are highlighted
(Figures 9 and 10), the organizational and func-
tional relation between groups having been
previously identified. Detailed information
relating to ongoing inter- and intra-DOE en-
51
-------
AREA
HEALTH EFFECTS RESEARCH
• DEVELOPED WIDELY USED AMES-TEST BIOLOGICAL (MlTMGENIC) SCREENING
PROTOCOLS
• DETERMINED THAT SYNFUEL BIOACTIVITY DISPROPORTIONATELY RESIDES IN
BASIC FRACTION (E.G. 90% OF BIOACTIVITY IN 10% OF MATERIAL)
• VERIFIED THAT HUMAN CELL IN VITRO (OUT OF BODY) BIO SCREENING
REPRESENTIVE OF IN VIVO (IN BODY) RESULTS
• DEVELOPED SCREENING TESTS (SKIN GRAFT TECHNIQUES) WHICH PERMIT
ANIMAL-HUMAN RESPONSES TO BE COMPARED
POLLUTANT CHARACTERIZATION «. SAFETY
• DEVELOPED PORTABLE SPILL MONITOR FOR AROMATIC COMPOUNDS
Figure 9. Select accomplishments DOE base/generic environmental program.
ECOLOGICAL RESEARCH
• DEVELOPED PREDICTIVE AND MODELING CAPABILITY FOR NUCLEAR FALLOUT
PHENOMENA
• MODELED MULTI-POLLUTANT REGIONAL IMPACT (INCLUDING ACID RAIN)
ENVIRONMENTAL CONTROL TECHNOLOGY
• CORROBORATED DESIGN ADEQUACY OF SYNFUEL PROCESSES ICOALCON, DRY BOTTOM
ASH LURGI ETC.)
• SINGLED OUT AREAS OF SECONDARY CONCERN (WASTEWATER. SOLIDS DISPOSAL)
SYSTEMS ENGINEERING
• DEVELOPED PROCEDURAL GUIDELINES FOR FIELD BASELINE MONITORING
• QUANTIFIED WATER RESOURCE IMPACTS (INCLUDING USE OF SALINE WATERS) FOR
SYNFUEL FACILITIES
Figure 10. Select accomplishments DOE base/generic environmental program (continued).
vironmental efforts are well documented in
publications, annual reports, and symposia pro-
ceedings. While highlights are shown of several
activities, I am not familiar enough with them to
discuss in detail these activities and their
EH&S ramifications.
In recognition of the depth of the individual
program within DOE and the types of generic
activity in which a single organizational entity
is involved, I have taken the liberty of selecting
representative synfuel-related activities within
the Environmental Control Technology Division
(EOT). This seems fitting considering the engi-
neering orientation of this symposium.
ECT assessment studies have attempted to
parallel major scaleup activities within industry
and within FEP. Proposed gasification and liq-
uefaction facility designs have been analyzed
52
-------
GENERIC, PROCESS
ORIENTED ISSUES
SITE SPECIFIC
ISSUES
CURRENT STANDARDS
COMPLIANCE STANDARDS WILL
BE MET.
UNCERTAINTY WITH REGARD TO
SOLIDS DISPOSAL TECHNIQUES
RESULTING FROM RCRA.
• TO BE ADDRESSED IN EIS.
• REVIEW OF DETAILED DESIGN
AND OPERATING PROCEDURES
TO ENSURE COMPLIANCE.
• ON-SITE CHARACTERIZATION
TO VERIFY.
FUTURE STANDARDS
TIGHTER CONTROL OF REFRACTORY
ORGANICS CONTAINED IN AQUEOUS
EFFLUENTS MAY BE EXERCISED (TSCA).
USE OF CONTAMINATED WATERS IN
COOLING TOWERS MAY BE PRE-
CLUDED.
PROCESS SLUDGES. BLOW-DOWNS AND
EVAPORATION POND SLUDGES MAY
REQUIRE TREATMENT (RCRA).
• UNKNOWN AT PRESENT
Figure 11. Perceived adequacy of environmental control technology for gasification.
GENERIC. PROCESS
ORIENTED ISSUES
SITE SPECIFIC
ISSUES
CURRENT STANDARDS
SAME AS GASIFICATION
SAME AS GASIFICATION
FUTURE STANDARDS
SAME AS GASIFICATION PLUS
• TRANSPORTATION AND
HANDLING OF HIGH BOILING
AROMATIC FUELS MAY POSE
SPECIAL PROBLEMS. I.E.
SPILL CONTROL AND CLEAN-
UP ETC.
UNKNOWN AT PRESENT
Figure 12. Perceived adequacy of environmental control technology for liquefaction.
from the standpoint of meeting compliance
standards and evolving NSPS for the reference
technology. Most of these studies come to the
same conclusion; namely, that the proposed en-
vironmental control processes appear adequate,
except for some minor uncertainties that can
only be resolved through in-plant monitoring
and surveillance (Figures 11 and 12). Of course,
the cost of implementing the control options is
the subject of debate.
Uncertainty exists regarding future stand-
ards relating to evolving TSCA and RCRA im-
53
-------
QUENCH WATER FROM
OETC SLAOGINQ-BED
UNIT. PPM
QUENCH/CONDENSATE
WATER FROM ORNL
HYOROPYROLVSIS
UNIT, PPM
TYPICAL OUTPUT FROM
LABORATORY WASTEWATER
TREAT ABILITY PROCESS
TRAINS *. PPM
ANTICIPATED
REGULATIONS
PHENOL
NHj
H,S
CN~
SCN-
PNA
TOC
BOD
COO
4.000- 6.000
8,000-10,000
N. D.
N.D.
N.O.
N. 0.
9,000 -10,000
3,000
30,000
600
70
1,000
6-10
20,000
0.010
6-10
0.1
0.06
0.005
6-60
NIL
NIL
0.3
6.0
0.2
0.1
30
360
AFIELD DEMONSTRATION PLANNED - TREATMENT TRAINS INCLUDED VARIOUS COMBINATIONS OF STEAM STRIPPING SOLVENT STRIPPINQ
BIOLOGICAL DIGESTION, CARBON ABSORPTION, AND CHEMICAL OXIDATION IOZONATION).
N. D. - NOT DETERMINED.
Figure 13. Composition of select coal gasification wastewaters.
GAS
DRYER
n
OREC
OZONE
GENERATOR
45'/h'°J
02 CYLINDER
pH
RECORDER METER
v-o-cK,
REACTION
LOOP
SPARGER'
I
IZOOml
MAGNETIC STmftEfti
GAS WASHING
BOTTLES
FLUOROMETER
[_ frl
1 RECO
RECORDER
MASTERFLEX
PUMP
SAMPLE
POUT
CALIBRATION
LOOP
GAS WASHING
BOTTLES
ROTAMETER
CAN BE
CONNECTED
TO EITHER
LINE
fl
CONTRACTOR-ORNL
Figure 14. Schematic drawing of batch ozonation system.
54
-------
plications with respect to effluents and solid
wastes. Several programs ongoing within EOT
point the way toward control options to meet
more stringent standards in a cost-effective
manner. As the technical and economic feasibil-
ity of these process variations is confirmed,
development and scaleup is picked up by FEP,
at their option.
One area of concern in coal conversion is, of
course, process water contamination. The con-
densate waters from liquefaction and quench
waters from gasification typically contain a high
organic loading. While nearly all the organics
are biodegradable, a trace fraction of ring-struc-
tured compounds (50 to 100 ppb) that resist con-
ventional biological treatment usually remains
(Figure 13). Carbon and char adsorption and
ozonation are being explored as polishing steps
to reduce the level of these trace compounds not
0.8
£
b 0.6
u
O
I
V)
u
0.4
u.
0.2
4
PNA RANGE 50-100 ppb
GAS FLOW RATE
(liters/min)
A 0.27
o 0.38
O 0.61
pH: 7.8
TEMP = 21°C
4 6
TIME (min)
8
10
CONTRACTOR-ORNL
Figure 15. Screening tests— ozonation of
hydrocarbonization wastewater (effect of
gas flow rate).
currently regulated (Figures 14 and 15).
Biological screening tests are being performed
to determine sensitivity thresholds for these
materials along with high-sensitivity gas
chromatographic analysis to determine actual
compounds. An alternative to intensive waste-
water post-treatments is water reuse within the
process proper. Concentration processes such
as freezing and membrane separation are being
investigated to maintain water balance and to
produce a concentrated contaminated aqueous
stream as input to an entrained gasifier (or ther-
mal oxidizer) where the organics are gasified
and salts are collected with the slag (Figure 16).
Wastewater quality requirements for cooling
tower concentration operations are being eval-
uated.
Another area of more immediate concern is
the impact of RCRA on coal conversion wastes
(gasifier slag and water treatment sludge dis-
posal). A screening program is underway to de-
termine if gasifier slags would be classified as
hazardous under candidate EPA protocols.
While preliminary tests indicate slags from en-
trained gasifiers may not be classified as haz-
ardous under the procedure, DOE does not en-
dorse meaningfulness or relationship of the pro-
tocols relative to actual land fill operations. In
my judgment, ecological and field characteriza-
tion studies are in order to verify true en-
vironmental acceptability of waste disposal
practice. Wastewater treatment sludges will be
characterized as quantities of these particular
materials become available.
Improved control technology for hydrocarbon
control in tail gases and within gasifier sulfur
scavenger options is being investigated (Fig-
ures 17 and 18). Controls for auxiliary opera-
tions such as boiler/power plants are being
evaluated under a family of assessments
oriented toward power generation.
SUMMARY
DOE incorporates all required compliance en-
vironmental, health, and safety safeguard moni-
toring and assessment within the project prop-
er as the responsibility of the FEP line-division
and process operator. As an additional precau-
tionary measure, overview responsibility has
been granted DOE's own internal environmen-
tal group, A8EV, to advise and assist the line-
division in these matters. DOE feels, however,
55
-------
BASFCASF ^^^
REVERSE OSMOSIS ^^
PRECONCENTRATION ^^
+gm
^^m
FREEZING CASE M^^
^•B
^^™
DISPOSAL CASES
IMO PMFMTII RFPAVFRVI
REVERSE ^^
OSMOSIS ^^
^
REVERSE
OSMOSIS
CONC.
•
FREEZING
CONC.
_J
REVERSE
__l
FREEZING
CONC.
•
l"
,
PHENOL
RECOVERY
VIA 1
SOLV. EXTR.
1
PHENOL
RECOVERY
VIA
SOLV. EXTR.
PHENOL
VIA
PPT.
AMMONIA 1
RECOVERY r
AMMONIA
RECOVERY
AMM. BICARB.
RECOVERY
AMMONIA
RECOVERY
AMM. BICARB.
RECOVERY
Up
1
\
J
n
— i*>
— «•
L
P^
L— ,
BIOLOGICAL
TREATMENT
<4i
TREATMENT
BIOLOGICAL
TREATMENT
-«—
IMriMF RATION
rat=as3!=a"
i GASIFICATION
INCINERATION
RELATIVE C
]1.0
. i 7
J
0.4'
0.0'
j 7
09
1 1*
03*
•LOW TREATMENT COSTS DEPENDENT ON BVPMOOUCT CREDIT OF
NM4 HCOj ISM/TON ASSUMED).
CONTRACTOR- CONCENTRATION SPECIALISTS. INC.
Figure 16. Candidate gas-Kquor wastewater reuse options.
-------
PROCESS
NO.
DESCRIPTION OF PROCESSES
APPROXIMATE COST
(t/105 Btu SNG>*
TECHNICAL FEASIBILITY
FOR THIS APPLICATION
1
2
3
4
6
6
7
B
9
10
INCINERATION IN A COAL-FIRED BOILER 4
INCINERATION IN A BOILER USING (EPA Reference Ciw) 11
DESULFURIZED MEDIUM Btu GAS
CATALYTIC INCINERATION 5
AQUA CLAUS PROCESS c
HOT CARBONATE SCRUBBING 42
COLD WATER SCRUBBING 66
CUPROUS AMMONIUM SOLUTION ABSORPTION 16
ADSORPTION c
CRYOGENIC SEPARATION 64
POROUS MEMBRANE SEPARATION c
GOOD
GOOD
UNPROVED
UNPROVED
DOUBTFUL
DOUBTFUL
DOUBTFUL''
NO PRACTICAL
ADSORBENT KNOWN
TECHNICALLY
FEASIBLE
DOUBTFUL
'MOST OF THE PROCESSES LISTED HERE HAVE NOT BEEN DEMONSTRATED IN THIS TYPE OF APPLICATION AND CANNOT BE CONSIDERED
AVAILABLE FOR COMMERCIAL USE.
6COST INCLUDES BOTH CAPITAL AND OPERATING CHARGES.
'INSUFFICIENT DATA TO PERMIT COST ESTIMATION IN THIS APPLICATION.
''THIS PROCESS REMOVES ONLY co.
CONTRACTOR-ORNL
Figure 17. Approximate cost of candidate hydrocarbon and CO emissions
control for Lurgi-type SNG plants.8
that "environmental acceptability" of a given
energy technology goes beyond meeting here-
and-now compliance standards. One must deter-
mine the potential environmental impact of a
mature industry, operating under future envi-
ronmental regulations. To support this activity,
DOE has initiated a series of detailed process-
specific field-oriented environmental character-
izations on advancing synfuel technologies. A
multidiscipline systems approach has evolved
deploying specialists in conversion technology,
control processes, health and safety, plant
operations, ecology, and systems analysis. Each
of these specialities is being supported by a
strong, ongoing generic program.
57
-------
C./S MOLE RATIO - 3
E
i
D
<
Z
tu
IU
K
E
K
t
90
80
70
»
so
40
30
20
10
°0
D
O TREATED WITH C»O IN SLURRY AT 30 C, 1000 PSIG
£ TREATED WITH C.O (STEAM CUT OFF DURING COMBUSTION
STAGE)
/\ TREATED WITH C«O IN SLURRY AT 30 C.O PSIG
^ TREATED WITH C«O (STEAM CUT OFF DURING COMBUSTION
STAGE)
^^ MIXED DRY WITH C*(OH)2
MIXED DRY WITH C» (OH), (STEAM CUT OFF DURING
COMBUSTION STAGE)
MIXED DRY WITH C«O
MIXED DRY WITH CjCO,
RAW COAL
s
1
1
10 15 3D
PERCENT STEAM IN FEED GAS
25
30
CONTRACTOR-fiATTELLE
Figure 18. Laboratory screening tests-feasibility of in-gasifier sulfur scavenging
(stream air gasification of treated coal in fluidized bed).
58
-------
NIOSH PROGRAMS FOR EVALUATION AND CONTROL OF INDUSTRIAL
HYGIENE HAZARDS IN COAL CONVERSION*
James Evans
Enviro Control, Inc., Rockville, Maryland
and
Barry Pallay
National Institute for Occupational Safety and Health, Morgantown, West Virginia
Abstract
It is well known that hazardous chemical sub-
stances and physical agents are present in coal
liquefaction and gasification and that the poten-
tial for occupational exposure is high. To make
certain that the workers in this new industry
will be protected, NIOSH first had defined prac-
tical means of protecting the worker and now
has initiated a multidisciplinary, in-depth as-
sessment of occupational health characteristics
and control technology for these conversion
processes through their principal investigator,
Enviro Control, Inc. This paper summarizes the
efforts to date.
The first NIOSH work was directed toward
protecting the worker from apparent problems.
This effort and its results are defined in the
documents, Recommended Health and Safety
Guidelines for Coal Gasification Plants and Cri-
teria for a Recommended Standard: Occupa-
tional Exposures in Coal Gasification Plants.
This is currently being followed by three in-
depth studies: Industrial Hygiene Characteriza-
tion of Coal Gasification Plants, Industrial
Hygiene Characterization of Coal Liquefaction
Plants, and the Assessment of Engineering
Control Technology for Coal Gasification and
Liquefaction. This paper describes the tech-
niques used for sampling and analyzing in liq-
uefaction and gasification plants and for the
ultimate use of the data. The interdependency of
the two characterization projects with the Engi-
neering Control Project will also be discussed.
While hard data are not included in this paper,
sufficient information will be available to show
the direction the three projects are taking.
*Unpresented paper.
INTRODUCTION
The National Institute for Occupational Safe-
ty and Health (NIOSH) has been interested in
studying coal liquefaction and coal gasification
since 1975. The basic objective of NIOSH, and of
the studies to be described, is to protect the
safety and health of American workers.
NIOSH has implemented programs to achieve
this objective through the development of a real
understanding of what the workers are exposed
to, and through promoting better measures that
will avoid occupational exposures.
Hazardous chemical substances and physical
agents are known to be present in coal liquefac-
tion and gasification operations, and the poten-
tial for occupational exposure is high (see
Tables 1 and 2). Because of this potential for oc-
cupational exposures, NIOSH undertook to pre-
pare safety and health criteria documents even
before the results of longer term detailed tech-
nical studies were available. Thus, the criteria
documents represent rough cuts at a standard
for coal conversion processes based on the best
available information at the time. These docu-
ments may be revised as other studies including
those to be described in this paper are com-
pleted.
THE NIOSH STUDY PROGRAM
The first NIOSH study resulted in the docu-
ment, Recommended Health and Safety Guide-
lines for Coal Gasification Pilot Plants.1 This
was followed by Criteria for a Recommended
Standard: Occupational Exposures in Coal Gas-
ification Plants.2 Both were done by Enviro
Control under the direction of NIOSH Project
Officer Mr. Murray Cohen. Currently, JRB,
59
-------
TABLE 1 POTENTIAL OCCUPATIONAL EXPOSURES IN HIGH-Btu
COAL GASIFICATION*
Unit Process
Potential Exposures
Coal handling and preparation
Coal feeding
Gas1f1er operation
Ash removal
Quenching
Shift conversion
Gas cooling
Gas purification
Methanation
Sulfur removal
Gas-I1quor separation
Phenol and ammonia -recovery
Byproduct storage
Coal dust, noise, and fire
Coal dust, noise, gaseous toxicants, and
asphyxia
Coal dust, high-pressure hot raw gas, high-
oressure oxygen, high-pressure steam, fire,
and noise
Heat stress, high-pressure steam, high-
pressure oxygen under Impact conditions,
hot ash, and dust
High-pressure hot raw gas, hot tar, hot
tar oil, hot gas liquor, fire, and noise
High-pressure hot raw gas, high-pressure
hot shifted gas, high-oressure steam, tar,
tar oil (naphtha), hydrogen cyanide, fire,
catalyst dust, and heat stress
High-pressure hot raw gas, hot tar, hot
tar oil, hot gas-liquor, fire, heat stress,
and noise
Sulfur-containing gases, methanol, naphtha,
cryogenic temperatures, high-pressure
steam, and noise
High-pressure Rectisol product gas, high-
pressure methanated gas, steam, nickel
carbonyl, nickel catalyst dust, fire, and
noise
Hydrogen sulflde, other sulfides, and
sulfur oxides
Noise, tar oil, tar, and gas-liquor with
high concentrations of ohenols, ammonia,
hydrogen cyanide, hydrogen sulfide, carbon
dioxide, and trace elements
Phenols, ammonia, acid gases, gas-liquor,
ammonia recovery solvent, and fire
Tar, tar oil, phenols, ammonia, methanol,
phenol-recovery solvent, and fire
National Institute for OccuoationafSafety and health, Criteria for~a
ended Standard.. .Occupational Exposures in Coal Gasification "Tants.
(NlOSH), Publication No. 78-191, September, 1978, po 32-33.
60
-------
TABLE 2. POTENTIAL OCCUPATIONAL EXPOSURES IN COAL LIQUEFACTION*
Unit Process
Potential Exposures
Coal handling and preparation
Coal slurrying
Coal dissolving
Liquid product flashing and
gas recycle
Filtration
Product distillation
Solvent Recovery
Gasification
Shift conversion
Gas cooling
Coal dust, noise, fire, inhala-
tion of combustion products.
Coal dust, noise, middle distil-
late-
High-pressure hydrogen, high-
pressure hot coal slurry, raw gas,
fire.
High-pressure hot gas, acid gas,
light oils, naphthas, middle oils,
hot high-pressure slurry, oil-
liquor, noise, steam, fire.
Precoat dust, light oil, hot
slurry, hot middle oil, hot fil-
trate, hot filter cake, solvent,
dissolved and undissolved coal,
steam, noise, heat, fire.
Hot filtrate, hot naphtha, hot
middle distillate, hot process
solvent, hot solvent refined coal,
vapors from SRC cooling, SRC dust,
steam, noise, fire.
Filter cake, oil-liquor, hot oil,
char dust, combustion gas, inert
gas, steam, noise, fire.
Ammonia, carbon dioxide, carbon
monoxide, hydrogen cyanide, hydro-
gen sulfide, hot raw gas, trace
elements, high-pressure steam,
char and coal dust, noise, fire,
trace elements.
High-pressure hot raw gas, high-
pressure hot shifted gas, high-
pressure steam, hydrogen sulfide,
hydrogen cyanide, fire, catalyst
dust, heat stress.
High-pressure hot raw gas, hot
condensate, fire, heat stress,
noise.
61
-------
TABLE 2 (continued)
Unit Process
Potential Exposures
Gas purification
Methanation
Sulfur removal
Hydrotreating
Oil-liquor separation
phenol and anroonia recovery
Byproduct storage, handling,
cleanup
Sulfur-containing gases, methanol,
oil condensate, cryogenic tempera-
tures, refrigerant gases, high-
pressure steam, noise.
High-pressure Rectisol product
gas, high-pressure methanated
gas, steam, nickel carbonyl,
nickel-catalyst dust, fire, noise.
Hydrogen sulfide, other sulfides,
and sulfur oxides.
Hot naphtha, hot middle distil-
lates, hot synthesis gas, high-
pressure steam, sour water, acid
gas, catalyst dust, fire, noise,
heat.
Coal oils, oil-liquor with high
concentrations of phenols, am-
monia, hydrogen cyanide, hydro-
gen sulfide, carbon dioxide,
trace elements, noise.
Phenols, ammonia, acid gases,
oil-liquor, fire, peroxide com-
pound explosion hazard.
Tar, SRC-I solid product, hydro-
genated oils, phenols, ammonia,
benzene-type light methanol,
phenol recovery solvent, fire.
— —•
* Taken from an interim draft report prepared on NIOSH Contract
No. 210-78-0101.
62
-------
under the direction of NIOSH criteria manager
Mr. Lynne Harris, is preparing a criteria docu-
ment recommending standards for occupational
exposures in coal liquefaction plants. NIOSH
Medical Officer Dr. William McKay is preparing
a medical protocol designed to identify the ap-
propriate means of medical monitoring in pres-
ent and future coal plants. Arthur D. Little, on a
contract directed by Dr. McKay, has prepared
some of the material for the medical protocol.
And Enviro Control, under the direction of
NIOSH Project Officer Mr. William Todd, is
conducting a study entitled Respiratory Protec-
tion in Coal Preparation Plants. Since relatively
little specific information is currently available
regarding occupational exposures and health ef-
fects of coal conversion, the NIOSH criteria doc-
uments on gasification and the liquefaction doc-
ument being prepared make no attempt to de-
velop permissible levels of exposure to toxic
substances specific to coal conversion plants.
Rather, they recommend that applicable exist-
ing Federal occupational exposure limits (or
NIOSH recommendations) be observed. The
documents also recommend that specific safety
procedures, engineering controls, work prac-
tices, workplace monitoring, medical surveil-
lance, personal protective clothing and equip-
ment, sanitation, labeling and posting, and
informing employees of hazards and record-
keeping be considered. The NIOSH documents
include specific information on these recommen-
dations. They also note the need for research ef-
forts to determine and project potential expo-
sures and, in particular, the need for industrial
hygiene and control technology efforts.
NIOSH has several coal studies in progress,
including the following, which are the subject of
this paper:
• A Study of Coal Liquefaction Processes:
Coal Liquefaction and Industrial Hygiene
Characterizations (Contract 78-0101).
NIOSH Project Officer: Mr. Barry Pallay.
• Industrial Hygiene Characterization of Coal
Gasification Plants (Contract 78-0040).
NIOSH Project Officer: Mr. Barry Pallay.
• Control Technology Assessment for Coal
Gasification and Liquefaction Processes
(Contract 78-0084). NIOSH Project Officer:
Mr. James Gideon.
In order to develop a program of this magni-
tude at this particular moment when the coal
conversion program in the United States is not
past the pilot-plant stage, two questions had to
be answered. First, Why bother now? Second
(and perhaps the more serious question), Can
sufficient information be obtained from the pilot
plants to assess potential occupational health
exposures in demonstration or commercial oper-
ations?
In response to the first question, the time for
obtaining this information is now, before dem-
onstration plants or commercial plants have
been built, to enable management and labor to
focus on the development of better work prac-
tices and engineering controls, which will result
in a more healthful workplace environment.
NIOSH, the U.S. Department of Energy (DOE),
the U.S. Environmental Protection Agency
(EPA), industry, and labor all agree that it is
preferable to have the controls built into the
plants in the design and construction stage,
rather than to retrofit them at a later date at
great expense and after workers have been ex-
posed.
This approach is particularly appropriate to
the coal conversion industry, as it now stands.
We are interested in identifying what the
workers may be exposed to, in determining
what the exposure levels may be, and then in
identifying cost-effective controls that will
reduce or eliminate these exposures, promote
productivity, and enhance the feasibility of coal
conversion technology being implemented on a
commercial scale.
As to the second question, those conducting
the study must thoroughly understand the dif-
ferences between a pilot-plant facility and a
demonstration or commercial installation. A
pilot plant is designed to obtain engineering
data to optimize operating conditions or to pro-
vide information on specific process feasibility
and practicality. A commercial plant is designed
for economical operation. Pilot-plant equipment
and operation is not optimized but rather is se-
lected to allow varied test conditions. Often the
equipment does not function adequately at the
conditions found to exist during the testing.
More potential exposure exists in pilot plants, if
only because they are continuously going on-
stream and offstream, either to change pro-
grams, to change layouts, or to repair equip-
ment. In general, the pilot-plant layout is more
compact and does not utilize all elements that
63
-------
would be installed at the commercial or demon-
stration level. Thus, it is important to remem-
ber that pilot plants are not small-scale replicas
of commercial plants. They are built to test cer-
tain defined parts of the process, using available
equipment, and do not represent the complete
commercial-scale process.
If this is understood, NIOSH, as well as other
agencies and industry, should be able to extrap-
olate industrial hygiene characterization infor-
mation and control technology information from
the pilot operation to the demonstration and
possibly to the commercial plant. If successfully
extrapolated, then industry will be able to im-
plement control designs to ensure the safety
and health of workers.
INDUSTRIAL HYGIENE
CHARACTERIZATION
The two industrial hygiene characterization
projects have four primary purposes:
• To determine potential worker exposures
from analysis of process streams, byprod-
ucts, and workplace levels of toxic materials;
the latter by area and personal sampling.
• To identify specific areas within plants
where carcinogens and other toxic chemical
and physical agents are concentrated by par-
ticular unit processes.
• To identify areas where control technology
assessment studies are now required or may
be required in the future.
• To extrapolate the data thus collected in
such a manner that anticipated worker expo-
sures may be approximated at the demon-
stration plant level and at the commercial
plant level.
Studies of coal liquefaction were to be made
on four different types of processes, including
noncatalytic high-pressure hydrogen transfer,
donor solvent process, catalytic hydrogenation,
and pyrolysis. Coal gasification characteriza-
tions were to include a high-Btu operation and a
low-Btu operation. The plants chosen for these
characterization studies are shown in Table 3.
The first to be studied was the solvent-refined
coal (SRC) process at the SRC pilot plant located
in Fort Lewis, Washington. The SRC-I process
includes a high-pressure noncatalytic hydrogen
donor transfer type, while the SRC-II process
includes high-pressure natural catalytic hydro-
gen donor transfer. The second plant chosen
was the Cresap test facility located in Cresap,
West Virginia, and operated by the Liquefied
Coal Development Corporation. This process in-
cludes low-pressure hydrogen donor solvent
transfer and catalytic hydrogenation of the sol-
vent-refined coal. The third plant selected was
the H-Coal pilot plant located in Catlettsburg,
Kentucky, and operated by the Ashland Oil
Company; this process includes direct high-pres-
sure catalytic hydrogenation of coal. The loca-
tion of the fourth process has not been finalized.
For the coal gasification characterization
studies, the Synthane plant located in Bruceton,
Pennsylvania, was to have been the high-Btu
plant, and the Combustion Engineering en-
trained-bed facility located at Windsor, Con-
necticut, was to have been utilized for the low-
Btu characterization study. In addition, the in-
dustrial hygiene characterization data was ex-
pected to be available through the Gasifiers in
Industry program at the University of Minne-
sota at Duluth facility. This facility uses a fixed-
bed, stoic low-Btu gasifier.
To date, walk-through surveys have been
completed at the SRC facility, the Cresap facil-
ity, the Synthane facility, and the Combustion
Engineering facility. However, the survey at
the Synthane plant was not completed when
DOE shut down the facility in December 1978. A
replacement for Synthane will be selected.
Since the H-Coal pilot plant will not be in opera-
tion until 1980, no survey has been scheduled
there.
The walk-through surveys are made to test
sampling and analytical methodology, and to de-
fine the range and level of toxicants in the pilot-
plant workplace. Pilot plants often do not have
predictable operating schedules. Therefore, to
facilitate the program, walk-through surveys
are sometimes carried out under the conditions
at which the pilot plants then happen to be oper-
ating. Sometimes conditions are not at steady
state. Data taken at nonsteady-state operations
are usually adequate for range-finding purposes
but might be misleading if used for other pur-
poses. Therefore, DOE and NIOSH agreed that
these data would not be published and would be
used only for the development of the sampling
plan and methodology for the comprehensive
studies.
Coal conversion facilities contain at least
eight categories of toxic compounds, as shown
in Table 4. The walk-through surveys include
64
-------
TABLE 3. COAL CONVERSION FACILITIES TO BE SURVEYED ON NIOSH
INDUSTRIAL HYGIENE CONTRACTS3
OS
en
Processes
Liquefaction
Solvent Refined Coal
Cresap Test Facility
H-Coal
Gasification
Synthane Fluid Bed
Gasification
Combustion Engineering
Entrained-Bed Gasifier
Developing ,
Company Location
Pittsburg & Midway Ft. Lewis, WA
Coal Mining Co.
Conoco Coal Develop- Cresap, WV
went Co.
Hydrocarbon Research, Catlettsburg,
Inc. KY
Lunmus Co. Bruce ton, PA
Combustion Windsor, CN
Engineering (C-E)
Noninal
Coal Feed
Rate
50 t/d pilot
plant
20 t/d pilot
plant fa-
cility
200-600 t/d
pilot plant
72 t/d
120 t/d
Status
600 t/d pilot
plant being
designed for
Morgantown, WV
Project terminat-
ed June 79
Under construction
Project terminated
Dec. 78
Termination date
uncertain
Contracts 210-78-0101 and 210-78-0400.
Fourth liquefaction plant not yet selected.
'Since Synthane terminated, a replacement plant will be selected.
-------
TABLE 4. TOXIC COMPOUNDS THAT MAY BE PRESENT AT
COAL GASIFICATION PLANTS
Category
Example
1. Polynuclear aroma tics
2. Polynuclear aza-heterocycllc compounds
3. Aromatic aMlnes
4. Nitrosamines
5. Trace elements
6. Participates
7. Gases
8. Other organlcs
Benz(a(anthracene
Benzo(a)pyrene
Benzo(fc)fluoranthene
Benzo(e)pyrene
Chrysene
Dibenzo(a,fc)pyrene
Indeno( 1,2,3-«f Jpyrene
Dibenz(a,fc)anthracene
Dibenzo(a,Z)pyrene
7.12-Dimethylbenz(a)anthracene
Benz(c)acrid1ne
D1benz(a,/i)acridine
Dibenz(aj)acr1d1ne
1-NaphthylaMlne
2-Naphthylamine
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium
Copper
Iron
Magnesium
Manganese
Mercury
Nickel
Phosphorus
Selenium
Strontium
Tellurium
Vanadium
Organic solubles
Particle sizing
Respirable and total particulates
Ammonia
Arsine
Carbon dioxide
Carbon disulfide
Carbon monoxide
Carbonyl sulfide
Aldehydes
Benzene
Cresols
Mercaptans
Methyl thiophene
Toluene
Xylenes
Carbonyls
Cobalt
Hydrogen cyanide
Hydrogen sulfide
Iron
Nickel
Nitric oxide
Hltrogen dioxide
Sulfur dioxide
Thiophene
-------
sampling for polynuclear aromatic hydrocarbon
(PNA) materials, aromatics, trace metals, par-
ticulates including coal dust and benzene sol-
ubles, aromatic amines, various gases, and hy-
drocarbon vapors. Thermal stress and noise
levels are also noted.
The walk-through sampling for PNAs is con-
ducted using closed-face 35-mm cassettes
(shown in Figure 1) consisting of a silver mem-
brane, a stainless steel screen, chromosorb 102
sorbent, and a cellulose support pad. Mass flow
through the cassette is controlled by a critical
orifice calibrated at 9.2 L/min. Sampling is con-
ducted at breathing zone level for 8-hr sampling
periods. To comply with the need to use intrin-
sically safe equipment at the pilot plants, air-
driven pumps are used. These pumps are rela-
tively small, simple, rugged, and inexpensive.
No difficulty has occurred in using this equip-
ment since the pilot plants have plant air piped
to all sections. Personal sampling for PNAs is
performed during the detailed surveys and uses
the train shown in Figure 2 with an MSA Model
S high-flow pump run at 2 L/min for 8 hr.
PVC filters are used to collect samples of
total particulates, and cellulose acetate filters
are used for trace metals. Flow rates are
2 L/min. For respirable dust sampling, the
closed-face cassette is preceded in the sampling
train by a miniature cyclone to remove nonre-
spirable particulates. Flow rates are 1.7 L/min
and MSA Model S high-flow pumps are used.
Charcoal tubes are used to collect samples of
organics such as benzene and toluene; silica gel
tubes are used for aromatic amines, phenols,
and cresols. Low-flow MSA C-200 pumps cali-
brated at 100 mL/min are used for charcoal and
silica gel. MSA and Draeger detector tubes are
used to check for the presence of toxic gases
such as HjjS. S02, CO, C02. N02, HCN, NH8,
CS2, arsine, and mercaptans.
PNA samples were analyzed by the Iowa
State Hygienic Laboratory, located at the
University of Iowa, under subcontract to En-
viro. The analytic methods used are described
in the literature. However, the analysis of the
pilot-plant PNA material was not straightfor-
ward and required developmental work before
the complex mixtures found in plant process
samples could be characterized. A paper de-
scribing the analytical procedures is being de-
veloped for presentation. It is sufficient to say
that Iowa State Hygiene Laboratory has devel-
oped a combination of gas chromatography/
mass spectrometry and high-pressure liquid
chromatography for the analysis. They are able
to back up these analyses by using glass capil-
lary column chromatography as defined by
White et al.»
After reviewing the sample results from the
SRC plant walk-through and current toxicity
studies (primarily Ames tests) being carried out
Casselte (37 mm I.D.)
Critical orifice
and adapter
Cellulose Support
Pad
Chromosorb 102
Silver membrane
Stainless steel screen
Figure 1. High-volume sampling device for PNA.
67
-------
Glass wool
Cassette (37 mm I.D.)
Tygon tubing
I
I 1 //I"
Chromosorb 102
1/4" OD
glass tubing
Cellulose gasket
Row
Cellulose gasket
Silver membrane
Rgure 2. Personal monitoring device for PNA.
at Oak Ridge National Laboratories and at Bat-
telle-North west Laboratory, NIOSH and En-
viro scientists recommended that the compre-
hensive sampling studies should concentrate
and prioritize sampling and analytical efforts.
PNAs were considered highest priority, fol-
lowed in descending order by aromatic amines;
hazardous gases such as GO and H2S; hydrocar-
bons such as benzene, toluene, and xylene; par-
ticulates; and trace metals.
While NIOSH had originally stated that the
presence or absence of nitrosamines should also
be investigated, a low priority was placed on
this analysis, especially in the complex mixture
potentially present at the pilot plants. Research
Triangle Institute's technical staff has since
pointed out that nitrosamines are not present in
their bench-scale reactor and are not expected
to be present in other gasification or liquefac-
tion facilities.
The walk-through studies indicate that, al-
though operating pilot plants have a pervasive
asphalt-like odor, in general the benzene-soluble
content of the atmospheric samples is well be-
low the NIOSH-recommended standards. A first
examination of these data also indicates that
there is a direct relationship between benzene-
soluble material in the atmosphere and house-
keeping, leaking equipment notwithstanding.
CONTROL TECHNOLOGY ASSESSMENT
The control technology assessment (CTA)
program for coal gasification and coal liquefac-
tion has three prime objectives:
• To bring together as much information as
possible on the control technology related to
coal gasification and liquefaction.
• To evaluate this information and publish it
along with recommendations for further re-
search.
• To use this information as one means of pro-
tecting the workers.
To accomplish these ends, we are examining
the following categories of control technology:
• Category I: Elimination by substitution of
unit process or hazardous material.
• Category II: Application of current technol-
ogy to specific equipment designed to con-
tain emissions.
• Category HI: Devices to control hazardous
emissions once they enter the work environ-
ment.
• Category IV: Controls used to isolate the
worker or prevent contact with the agent.
• Category V: Monitoring systems that warn
workers of hazards and initiate corrective
measures.
This study attempts to examine all aspects of
68
-------
the processes that might lead to exposures of
the workers and will examine means of prevent-
ing these exposures. In short, we will attempt to
look at the conditions and chemistry of the proc-
ess and must examine almost every aspect of
equipment design—seals, flanges, packing,
valves, rotating equipment, etc. It may be asked
how this study differs from the documents that
have already been completed, such as the Rec-
ommended Health and Safety Guidelines for
Coal Gasification Pilot Plants and Criteria for a
Recommended Standard: Occupational Expo-
sures in Coal Gasification Plants. The question
is valid and has been raised several times.
For the pilot-plant document, NIOSH at-
tempted to determine where and why work-
place emissions occur in order to alleviate such
emission sources. The emphasis in the criteria
document was on the technology currently
available for commercialization (i.e., Lurgi).
Those studies also made a thorough investiga-
tion into what was currently known about the
toxicology and epidemiology of coal conversion
products. The control technology assessment
program has two central ideas. First, if better
equipment design can reduce emissions, there
will be less worker exposure. Second, if equip-
ment maintenance requirements can be re-
duced, there also will be less worker exposure.
The OTA study focuses on the process and,
more particularly, on the equipment itself—
what the technical problems are, what is being
or can be done about these problems to reduce
the emissions, what is being developed in the
pilot plants, as well as what is commercially
available today. We will also attempt to define
the problems that may lead to worker exposure
and may require further research. The CTA
study is not a traditional industrial hygiene
survey; few samples will be taken. However,
this study and the two characterization studies
previously described were designed to comple-
ment each other, so that a maximum amount of
information could be cross-correlated.
In order to obtain the information required
for the CTA study, NIOSH/Enviro investigators
plan to make approximately 40 site visits, in-
cluding coal liquefaction and coal gasification
pilot plants, which will be visited in conjunction
with the industrial hygiene survey visits. In
gathering information for this study, we will
visit architectural and engineering firms such
as Dravo and Fluor, which have extensive ex-
perience in the design and construction of these
plants; we will talk with the designers of demon-
stration plants such as the Conoco Slagging
Lurgi and the SRC-II operation; we will visit
several plants operating or being constructed
under the Gasifiers in Industry program; and
we will visit the ANG Coal Gasification Com-
pany, which, in 1980, may begin construction of
the first high-Btu coal gasification plant in the
world, providing proper Federal Energy Regu-
latory Commission (FERC) permits are forth-
coming this summer.
We will also visit commercial installations
with analogous processing systems, such as
petroleum refineries and ammonia-manufactur-
ing operations. With the exception of several
low-Btu facilities, we will not be able to visit an
operating commercial gasification facility in this
country; therefore, we hope to visit several op-
erations in Europe that we have not seen be-
fore. We also hope to benefit from several proj-
ects initiated by DOE, including those that look
into the instrumentation needs of demonstra-
tion and commercial facilities and into the
availability of commercially sized equipment.
When making the-site visit, the NIOSH/En-
viro team first gathers as much information as
possible concerning the site. If we are visiting
an operating facility, we generally have the op-
portunity to inspect the facility in detail and, at
the same time, to take a number of samples with
direct-reading instruments, primarily for car-
bon monoxide and organic vapors. (At the three
plant sites visited thus far—Combustion Engi-
neering, SRC, and Cresap—we have been un-
able to find detectable measurements of either
the light organic vapors or carbon monoxide, ex-
pect in a hot well and over an open manhole in a
vessel that contained water saturated with car-
bon monoxide.) We then have the opportunity to
talk with supervisors, engineers, and workers
at the site about various processing, operation,
and mechanical problems. We base our conver-
sations on a pre-prepared site-specific question-
naire, which is generally used to start the con-
versation and to lead us into areas where little
or no information has been reported in the
literature.
To date we have made six such visits: Com-
bustion Engineering Entrained Bed Gasifier in
Windsor, Connecticut;. Solvent-Refined Coal
Pilot Plant in Ft. Lewis, Washington; Synthetic
Fuels Pilot Plant in Cresap, West Virginia;
-------
Dravo Corporation's Synthetic Fuels Division in
Pittsburgh, Pennsylvania; Synthane Pilot Plant
in Bruceton, Pennsylvania; and Synthoil PDU in
Bruceton, Pennsylvania.
As anticipated, many of the coal conversion
operations have similar problems with similar
pieces of equipment; for instance, valves pass-
ing high-pressure three-phase liquids erode
rapidly. Pumps are another area where severe
erosion problems occur. Conventional pump im-
pellers and volutes erode out within days.
Several of the plants are investigated hard-
surface applications on the pump interiors; they
are also looking at a number of different solu-
tions to the omnipresent pump seal problem.
These efforts do not seem to be tightly coor-
dinated, and information is slow in traveling
from one facility to another. In this area alone,
we would hope that our efforts will provide a
significant contribution, where the net result of
our efforts will be an integrated report on all
that we have learned, as well as our assessment
of the best ideas currently available and the
pressing needs for future research.
CONCLUSION
With the integration of the coal gasification
and coal liquefaction industrial hygiene charac-
terization studies and the CTA studies, we hope
to relate detailed analysis of emissions with
process, operating, and mechanical problems. In
other words, we now have the opportunity to
develop an understanding of the real breadth
and depth of the potential occupational health
problem in coal conversion.
The information available from the CTA
studies will be invaluable to the industrial hy-
giene studies, particularly for the extrapolation
of the sampling data from the pilot plant up to
the commercial operation. Industrial hygiene
data from the pilot-plant situation has never
been extrapolated to a commercially sized facil-
ity. The parameters for obtaining this data have
not been established. Thus, if the data are ob-
tained properly and the proper means of extrap-
olation are used, we should be able to provide
sufficient information so plants can be built with
emission levels lower than the current antici-
pated levels.
In summary, it must be recognized that these
three programs are a pioneering effort. Never
before has NIOSH had the opportunity to take
pilot-plant industrial hygiene data and extrap-
olate it for the protection of future workers in
what we see as a future major industry. As this
precommercialization effort moves forward, we
expect that, through the combined efforts of all
of the participating individuals and all of the
programs, we will obtain sufficient information
regarding potential occupational hazards and
their control to not only ensure the health and
safety of workers in the coal conversion indus-
try but also to establish it in a cost-effective
manner.
REFERENCES
1. Recommended Health and Safety Guide-
tines for Coal Gasification Pilot Plants. Na-
tional Institute for Occupational Safety and
Health. Rockville, Md. DHEW (NIOSH) Pub
lication Number 78-120. January 1978.
2. Criteria for a Recommended Standard: Oc-
cupational Exposures in Coal Gasification
Pilot Plants. National Institute for Occupa-
tional Safety and Health. Rockville, Md.
DHEW (NIOSH) Publication Number
78-191. September 1978.
3. White, C. N., M. L. Lee, and D. L. Vassilaros.
A New Retention Index System to Program
Mid-Temperature Capillary/Column Gas
Chromatography of Polycyctic Aromatic
Hydrocarbons. DOE, PETC (submitted for
publication, 1979).
70
-------
EPRI CLEAN FUEL PROGRAM
S. B. Alpert* and B. M. Louks
Electric Power Research Institute,
Palo Alto, California
Abstract
The EPRI program has been underway for
several years. It is aimed at furthering the de-
velopment of advanced systems that can pro-
vide clean synthetic fuels from coal that can be
used to generate power. The EPRI program and
its technical and economic methodology will be
described A key consideration is the benefit to
the environment. The technical and economic
attractiveness of technology and the ability of
new technology to satisfy existing and pro-
jected environmental standards are also con-
sidered.
The environmental assessment of the technol-
ogy with regard to plant siting and fuel utiliza-
tion is best handled as an essential portion of
the R&D contract Unrelated environmental as-
sessment can be counterproductive and waste-
ful, especially in situations where assessments
are made for technologies that ultimately fail to
meet technical and economic goals in pilot plant
test programs.
INTRODUCTION
For 5 yr the Advanced Fossil Power Systems
Department has been directing and managing
the research and development for new ad-
vanced systems that have potential application
to the production of electric power. These
systems need to be cost-competitive and must
satisfy increasingly tight enviromental stand-
ards. Major emphasis has been on flexibility in
using U.S. coals in these R&D projects.
This paper describes the EPRI program in
coal liquids and gaseous fuels and the method-
ology used to assess technology and to imple-
ment the environmental program associated
with the development of advanced systems.
GENERAL OBJECTIVES
Table 1 provides a list of deliverables that
'Speaker.
EPRI expects from an integrated R&D program
aimed at commercial acceptance in the power in-
dustry. In order to receive the attention of the
EPRI staff, each of these items needs to be ad-
dressed and dealt with in the R&D program
that is to be carried out. Table 2 provides a list
of factors that need to be addressed in process
evaluations and that are optimized during pro-
gram development.
COAL LIQUEFACTION
In general, orderly program development
begins with bench-scale equipment to prove the
technological feasibility, moves to operation of
integrated process development units, and cul-
minates in large pilot-plant testing at the 100- to
500-ton/day scale to set the design of commer-
cial plants. Two coal liquefaction pilot plants are
under construction, each of which represents
about 1,000 construction and management per-
sonnel. The cost to the participants in these
first-of-a-kind facilities is $100 million. The
operation of the pilot plants and the associated
support R&D represent a total cost of about
one-quarter of a billion dollars. Such programs
are expensive and highly risky. Until they are
successfully operated for a significant period of
time using the design coal that is to be used in a
commerical plant, there is a chance of technical
failure. Table 3 is an outline of types of syn-
thetic fuels by potential market applications of
interest to utilities.
A partial list of key technical issues that re-
main to be resolved in the R&D program for
producing clean liquid and solid fuels is shown
in Table 4.1 shall not discuss in any detail this
simplified list, but it indicates that a number of
significant technical issues remain to be solved
before coal liquefaction technology reaches a
state of readiness wherein we can confidently
construct commercial plants.
Incentives For Coal Liquid Fuels
Coal liquefaction offers the utility industry an
71
-------
TABLE 1. DELIVERABLES FROM A PROCESS-ORIENTED R&D PROGRAM
• CORRELATIONS AND DATA SUMMARY OF EXPERIENCE
• A DEFINITION OF OPERABLE AND INOPERABLE COMBINATIONS OF
PROCESS VARIABLES
• SUSTAINED DURATION OPERATIONS AT DESIGN CONDITIONS
• ENGINEERING DATA ON DESIGN FEED NEEDED FOR SCALE-UP
t A SERIES OF COMMERCIAL PLANT EVALUATIONS
• AN OPERATING AND MAINTENANCE MANUAL
t A SKILLED TEAM OF SPECIALISTS
TABLE 2. FACTORS REQUIRING OPTIMIZATION IN PROCESSES
PRODUCT VALUE FUEL BALANCE
PRODUCT SLATE AND MARKETS WASTE STEAM CLEANUP
CAPITAL COST, OPERATING COST HYDROGEN, UTILITY GENERATION
THERMAL EFFICIENCY INTEGRATION OF RECYCLE STREAM
STEAM BALANCE QUALITY OF RECYCLE STREAMS
___ 72
-------
TABLE 3. SYNTHETIC FUELS OF INTEREST TO UTILITIES
FUEL TYPE
PROCESS
POTENTIAL MARKETS
METHANOL
• PEAKING COMBUSTION
TURBINE
TURBINE FUELS
HYDROTREATED FRACTIONS
FROM;
• H-COAL
• EXXON
• COMBUSTION TURBINES
t INTERMEDIATE LOAD
COMBINED CYCLE UNITS
DISTILLATE
BOILER
FUELS
HEAVY LIQUID
BOILER FUELS
FRACTIONS FROM;
• H-COAL
t EXXON DONOR
SOLVENT
t SRC-11
FRACTIONS FROM:
• H-COAL
• EXXON
DONOR
SOLVENT
t RETROFIT GAS FIRED
BOILERS
• RETROFIT OIL BOILERS
FOR PEAKING SERVICE
• RETROFIT EXISTING OIL
FIRED BASE LOAD UNITS
SOLID BOILER
FUEL
SOLVENT REFINED
COAL
RETROFIT EXISTING
INTERMEDIATE LOAD PLANT
SPECIFICALLY DESIGNED
SIMPLIFIED BASE LOAD
PLANTS
73
-------
TABLE 4. MAJOR AREAS REQUIRING OPTIMIZATION
AND TECHNICAL RESOLUTION
t PROCESS APPLICATION TO A VARIETY OF COALS
t SOLIDS SEPARATION (SOLID, HEAVY FUEL OIL)
• HYDROGEN PRODUCTION VIA GASIFICATION OF RESIDUES
• FIRED SLURRY HEATER DESIGN
• FEED SLURRY MIXING AND DRYING
§ VACUUM TOWER DESIGNS
§ LET DOWN VALVES, PUMPS
• PUMPS HANDLING SLURRIED COAL, PRODUCTS
option, based on domestic energy resources,
with which to meet its need for liquid fuels. In
1977, generation of electricity consumed 188,000
bbl/d of distillate fuels and 1,469,000 bbl/d of
residual oil (see Table 5). The National Elec-
trical Reliability Council projects, in their
August 1978 report, that this requirement will
grow to 366,000 bbl/d and 1,809,000 bbl/d, re-
spectively, by 1987. In addition, natural gas re-
quirements that can be met by substituting
clean liquid fuels will decline from the 1977 level
of 1,209,000 bbl/d FOE (fuel oil equivalent) to a
still substantial 457,000 bbl/d FOE. This com-
bination calls for 2,632,000 bbl/d of hydrocarbon
fuels in 1987 and perhaps 4,000,000 bbl/d by the
year 2000.
The same report discusses the potential for
additional requirements for liquid fuels because
of a 1- or 2-yr delay in completion of coal and
nuclear plants. If electricity growth averages
5.6 percent per year compounded, an additional
1,041,000 bbl/d could be required if such a delay
occurred. The experience of 1977, when liquid
fuels were utilized to cope with the combination
of a severe winter that curtailed natural gas
supplies used for power generation and a coal
strike, demonstrates that liquid fuels can be
quickly utilized to meet emergency situations.
Today, the planned installation of new oil-
fired steam boilers is essentially nil. Approx-
imately 96,000 MW of capacity will remain in
place in 1987. These units were put into service
primarily in the mid-1960's and have 10 to 30 yr
of useful life remaining. Installed capacity of
liquid-fueled combined-cycle units is expected to
grow from 3,000 to 8,000 MW over this time
period. These units generate electricity more ef-
ficiently than conventional boilers. Combined
cycle capacity is projected to be used more ex-
tensively than in the past. As a result, the an-
ticipated quantity of power generated from
combined-cycle equipment may increase nine-
fold from 4 to 36 billion kWh. Unfortunately, the
future use of petroleum liquids for this kind of
operation has been jeopardized by the recently
legislated Fuel Use Act. This act requires coal
to be used instead of petroleum for new power
stations.
Liquid fuels are attractive to utilities for the
following reasons:
• They are clean and satisfy environmental re-
strictions.
74
-------
TABLE 5. ELECTRIC UTILITY INDUSTRY USE
OF GASEOUS AND LIQUID FUELS
ACTUAL ESTIMATED
1977 1987
OOP's B/D FOE OOP's B/D FOE
DISTILLATE OIL - STEAM 57 70
COMBUSTION TURBINE 116 152
COMBINED CYCLE 15 144
RESIDUAL OIL - STEAM 1,466 1,797
COMBUSTION TURBINE 1 1
COMBINED CYCLE 2 11
CRUDE OIL - STEAM 9 8
SUB TOTAL 1,666 2,183
GAS - STEAM 1,149 425
COMBUSTION TURBINE 23 9
COMBINED CYCLE 37 23
SUB TOTAL 1,209 457
GRAND TOTAL 2,875 2,640
75
-------
• They are easily stored and transported.
• They have properties that can be tailored to
meet user requirements.
• They can be used in new combustion tur-
bines and combined-cycle machines to meet
intermediate and peaking power require-
ments at lower cost than coal-fired plants.
Technology Assessment For Coal-Derived
Processing Routes
EPRI's selection of major investment in the
Exxon EDS project ($30 million) and the H-Coal
project ($12 million) was based on a thorough
review of the processing route alternatives.
Based on the status of technological options, the
capabilities of the management and technical
teams, and the status of the technology, the
EDS and H-Coal routes were selected for major
support.
Actually, EPRI has had a similar investment
in supporting R&D for clean solid fuel via the
SRC-I process. To date the support of the pilot
plant at Wilsonville has represented expendi-
tures of about $17 million on the part of EPRI
over the last 5 yr.
Economic Assessments
There is no evidence to indicate that any
single liquefaction process offers a significant
economic advantage over all others if the de-
sired product slate is fixed. At our current level
of understanding, all leading process candi-
dates, H-Coal, Exxon Donor Solvent, and SRC-
II, appear to produce a specified slate of prod-
ucts at approximately the same cost from a
given coal. The uncertainty in the absolute costs
is larger than the difference between processes
producing similar product slates and quality.
Thus, economic assessments do not aid in selec-
tion of technology choices; selection depends on
factors such as whether the deliverables can be
realized from the project by an experienced or-
ganization.
Combustion Testing Programs
Utilization of coal-derived fuels seems to offer
no more of a challenge than using any new fuel,
such as many low-sulfur fuel oils or low-sulfur
western coal. Test results on SRC-I and SRC-II
fuels in utility tests are shown in Tables 6 and 7.
76
The utility industry requires comprehensive,
large-scale, and long-term tests in utility equip-
ment prior to accepting any new fuel. As an
example, the changeover from eastern coal to
western coal was traumatic for many utilities
because a large number of new maintenance
problems and emission control difficulties were
generated. In line with these requirements,
EPRI has set up a multitiered synthetic fuel
combustion test program to establish accept-
able safe handling procedures. Large-scale util-
ity test programs will require 10,000-40,000
bbl/d of fuel. Sustained test programs, which
will last approximately 6 mo, must await suc-
cessful operation of demonstration or pioneer
commercial plants, which is not scheduled to oc-
cur until after 1985.
Based on EPRI tests performed to date, there
are data that indicate that coal-derived solid and
liquid fuels can be safely handled and complete-
ly combusted in existing utility boilers to avoid
exposure of the public to potentially harmful
aromatic chemical species.
Although obviously not a coal liquefaction
product, shale oil represents another synthetic
fuel option. During the last quarter of 1979, the
U.S. Department of Defense arranged with
Standard Oil of Ohio through the Paraho Devel-
opment Corporation to refine 100,000 barrels of
raw shale oil. EPRI arranged for delivery of
4,500 barrels of the hydrotreated 700° F resi-
due. This product will be used for a utility site
combustion test during 1979. Other test work is
underway using methanol in combustion tur-
bine equipment at a utility site.
Environmental Tests And Issues
Plant Siting Issues—
A major purpose of the operation of the large
coal liquid pilot plants is to obtain information
required to design commercial plants that can
be sited at specific locations. Thus, each of the
major projects has recognized, as an essential
objective, the need to provide necessary design
data for commercial plants. It is not useful to
face the plant-siting issues if the technical
hurdles cause development to be abandoned.
Table 8 shows a list of recent process failures. It
is more efficient to address environmental ques-
tions when the basic process is developed.
Air quality will be monitored at the pilot-
plant sites. Water samples will be handled in
-------
TABLE 6. SRC-I TEST RESULTS
FUEL FUEL ANALYSES EMISSIONS
%S ZN S02 NOX
LB/106 BlU PPM LB/106 BTU PPM
COAL 0,88 l.M 1.01 319 0,47 315
SRC-I 0,71 1,60 0,97 335 0,40 320
TABLE 7. SRC-II TEST RESULTS
FUEL FUEL ANALYSIS MOX EMISSIONS
NORMAL BOILER Low N0y
ZN SETTING BOILER SETTING
PETROLEUM DERIVED
#6 FUEL OIL 0,23 155 100
COAL DERIVED
SRC-II 1,00 270 175
77
-------
facilities either onsite or offsite, and solid
wastes will be analyzed and disposed to moni-
tored landfills. The environmental aspects are
an essential part of these projects and are the
responsibility of the organization carrying out
the project.
It is not advisable to separate the environ-
mental plant cleanup from the rest of the proj-
ect. It is an essential part of the development
and cannot be assigned to a separate uninvolved
organization.
Fuel-Handling iMues—
Handling coal-derived fuels safely at utility
sites with proper protection of plant personnel
and the general public does not seem to repre-
sent formidable problems. Refineries and chem-
ical plants have a long history of dealing with
similar fuels in a satisfactory and acceptable
manner. In this instance, the problem is of a
general nature and a program of research is
likely to be separable from the development of
specific technologies.
Timing of Environmental Work
As indicated, there are still technological
hurdles in the development of clean fuels de-
TABLE 8. PROCESS FAILURES FOR
PRODUCING COAL LIQUIDS
COALCON
CLEAN SYNTHETIC FUELS (CSF)
GULF CCL
SYNTHOIL
SOLVENT REFINED LIGNITE
rived from coal. The environmental program
should be integrated into the R&D program and
not performed by outside contractors having no
understanding of the technical development.
Elaborate measurement of effluents at the
bench scale and at the process development unit
scale of operations is useless. For example, the
elaborate programs on the Synthoil products
and Synthane processes were wasted because
they were terminated for a variety of reasons.
The detailed reports are filed away. Perhaps
the procedures and protocols will be useful, but
if the support had been used to solve a number
of technical problems, we might have had a bet-
ter chance of technical success.
GASIFICATION FOR ELECTRIC POWER
GENERATION
Gasification is a process of converting a solid
fuel, such as coal, into a clean, easy to manage
gaseous product containing substantial quanti-
ties of carbon monoxide and hydrogen. This gas
can be processed further to produce transport-
able and storable fuels such as SNG and metha-
nol, or it can be burned directly in an environ-
mentally acceptable manner for electric power
generation.
There are two fundamentally different ways
in which coal gasification can be used for elec-
tric power generation. The most obvious meth-
od involves a total decoupling of the gasification
process from the power generation facility.
Examples of such systems are:
• Gasification for SNG production;
• Gasification for methanol production; and
• A remotely located gasification plant sup-
plying intermediate-Btu fuel gas over rela-
tively short distances to be burned in con-
ventional oil or gas-fired steam power
plants, combined-cycle equipment, or fuel
cells.
All of these options are technically viable.
However, studies conducted by EPBI and
others have shown that fuels produced in this
manner will be expensive and the overall effi-
ciency of converting coal to electric power will
be poor. Table 9 provides estimates of delivered
fuel costs and coal-to-power efficiencies for
some of the above options. Based on the rela-
tively high fuel costs shown in Table 9, and con-
sidering current economic dispatch constraints,
it is clear that the above options will probably
78
-------
TABLE 9. FUEL COSTS AND EFFICIENCIES FOR DECOUPLED SYSTEMS
SNG METHANOL INTERMEDIATE BTU GAS
FROM COAL FROM COAL FROM COAL
COST OF FUEL DELIVERED
TO THE POWER PLANT SITE,
$/MMBTU(1) $6,00-18.00 $6.00-$8.00 $3,50-$5.00
HEAT RATE, BTU/KWH 16,000 15,500 12,000
EFFICIENCY, PERCENT
21,3 22,0 28.1
(i) MID 1976 DOLLARS; $1,00/MMBTU COAL, ILLINOIS #6 COAL,
-------
only be considered for intermediate and peak
load service in the near future if direct coal fir-
ing continues to exist as an environmentally ac-
ceptable option for baseload power generation.
A possible exception to this conclusion could be
utility systems that have a large fraction of oil-
or gas-fired baseload equipment that can be re-
trofitted.
A second option that exists for applying coal
gasification technology to electric power gen-
eration is the concept of an integrated, dedi-
cated power plant. The gasification plant would
be constructed at the power plant site, closely
coupled to the power-generating equipment.
Studies conducted by EPRI and others have
shown that integrated gasification/combined-
cycle (GCC) power systems have the potential
for extremely efficient operation and for the
production of competitively priced baseload
electric power while providing an environmen-
tally superior power plant.
Environmental Aspects of Gasification/
Combined-Cycle Power Plants
Gasification is a well-known technology for
converting coal into an ultra-clean, low- or inter-
mediate-Btu fuel gas that is capable of meeting
the increasingly more stringent environmental
control requirements dictated by the Glean Air
Act Amendments of 1977.
During gasification, the bulk of the sulfur in
the coal is converted to hydrogen sulfide, which
can be removed from the fuel gas to practically
any extent required by any one of a number of
commercially proven liquid absorption proc-
esses.
Experimental evidence obtained from Texa-
co's Montebello pilot plant demonstrates that
particulate matter carried over from the gasi-
fier into the fuel gas can be removed by water
scrubbing, bringing levels down orders of mag-
nitude lower than either existing or particulate
emission control requirements.
Nitrogen oxide emissions from a GCC power
plant will be mainly a function of combustion
conditions in the turbine combustor. Proper
combustor design coupled with the relatively
low-firing temperatures of currently available
combustion turbines will tend to reduce NOX
emissions to levels below those required by cur-
rent regulations. The contribution from the
gasification plant to reduced NOX emissions will
come from the commerically proven ability to
remove all ammonia from the fuel gas by water
scrubbing, thereby essentially eliminating the
fuel-bound nitrogen.
Solid and liquid effluents from a GCC power
plant will be lower in quantity than from other
coal-based power generation technology. Solid
wastes will be limited essentially to a dry, inert
ash and liquid effluents will consist mainly of
low-volume purge water streams.
Finally, the total makeup water requirement
for a GCC power plant will be approximately 60
percent of that for a conventional coal-fired
steam plant because the bulk of the electric
power will be generated by the combustion tur-
bines, which do not require condensers.
Economics of Gasification/Combined-Cycle
Power Systems
Over the past 4 yr, EPRI has been examining
the costs associated with power production
from gasification/combined-cycle systems. Much
of this work has been conducted by Fluor Engi-
neers and Constructors, Inc. Because of incom-
plete understanding of the effects of combustion
turbine performance on overall system efficien-
cy, much of the initial effort concentrated on
GCC plants employing advanced high-tempera-
ture turbomachinery. Emphasis has recently
been redirected to consideration of the costs of
GCC power plant employing current technology
combustion turbines.
Also, earlier evaluations were aimed at iden-
tifying the gasification technologies offering the
greatest economic incentives for development.
The general conclusion reached from these ini-
tial studies was that the cost of electricity to be
expected from a GCC power system based on a
variety of second-generation gasification tech-
nologies would be somewhat unrelated to the
particular gasifier being used. On this basis, the
choice of a gasification technology to be
employed for the first commercial-scale plants
should be based on state of development and
downstream processing requirements imposed
on the power system rather than on economic
considerations alone. Based on this conclusion
and the status of various gasifier developments,
analytical effort at EPRI has concentrated on
the evaluation of Texaco gasification-based
systems.
Table 10 presents a performance and cost
80
-------
oo
TABLE 10. COST AND PERFORMANCE COMPARISON OF TEXACO-BASED GCC PLANT
WITH A CONVENTIONAL COAL-FIRED POWER PLANT
TEXACO BASED GCC
PLANT WITH 2,000°F
COMBUSTION TURBINE
(1)
COAL TYPE
PLANT LOCATION
EMISSION REGULATIONS
DESIGN CAPACITY, MW
DESIGN HEAT RATE, BTU/KWH
TOTAL INSTALLED PLANT COST, $/KW(S) 774
TOTAL CAPITAL REQUIREMENT, $/KW(3)(4) 903
ILLINOIS #6
ILLINOIS
1976 EPA NSPS
1,000
9,100
CONVENTIONAL COAL
FIRED PLANT WITH
STACK GAS SCRUBBERS
(2)
ILLINOIS BITUMINOUS
WISCONSIN
1976 EPA NSPS
1,000
9,900
743
906(5)
(i) BASED ON EVALUATIONS CONDUCTED BY FLUOR ENGINEERS AND CONSTRUCTORS, INC,
(RP-239) AND GENERAL ELECTRIC COMPANY (RP-986-3),
(2) EPRI REPORT AF-1011, WORK PERFORMED BY BECHTEL NATIONAL, INC. (RP-1080-1).
(3) MlD-1978 DOLLARS,
(4) INCLUDES CONSTRUCTION LOAN INTEREST, INVENTORY CAPITAL, START-UP COSTS,
ROYALTIES, INITIAL CATALYST AND CHEMICAL COSTS AND LAND,
DOES NOT INCLUDE THE COST OF LAND AND EQUIPMENT FOR SOLID WASTE DISPOSAL,
-------
oo
to
TABLE 11. ECONOMIC COMPARISON OF ELECTRICITY COSTS FROM
GASIFICATION/COMBINED-CYCLE PLANTS
(1976 DOLLARS)
PROCESS
CAPITAL, $/KW
COST OF SERVICES,
MILLS/KWH
COAL a $!/MM/BTU
COAL a $2/MM/BTu
COAL FIRED POWER
PLANT WITH F.G.D,
838
40,9
51,2
LURGI*
906
41,2
51,4
BGC
SLAGGER
711
32.8
41,6
COMBUSTION
ENGINEERING
860
39,0
47.6
TEXACO
816
37.2
46,5
'WESTERN COAL
-------
comparison of a Texaco-based GCC plant em-
ploying currently available combustion turbines
with a conventional coal-fired power plant using
nonregenerable limestone scrubbers for S02
emission control. Because these two cost esti-
mates were prepared by different contractors
with designs based on somewhat different coals,
they are not strictly comparable. However, they
do demonstrate that the Texaco-based GCC
plant with available turbomachinery will be
more efficient than and cost-competitive with
conventional coal-fired technology based on
1976 environmental control requirements.
If environmental regulations become more
stringent (as they already have, based on EPA's
proposed regulations stemming from the Clean
Air Amendments of 1977), studies have shown
that the cost differential between GCC systems
and coal-fired plants will increase; i.e., the
economic incentives to build GCC plants will
become greater.
Economic Evaluations
The EPRI economic evaluations have been
extensive in order to direct the selection of gasi-
fication technologies to be supported. Table 11
summarizes costs of power from conceptual
plants based on a variety of coal gasification
processes. As can be seen, the second-genera-
tion processes offer competitive costs for elec-
tric power. In part, such economic analyses are
used to guide the projects in which EPRI has in-
vested R&D funds. We have supported the
three technologies indicated in the last three
columns and have test work underway on these
three coal gasification processing routes.
Table 12 summarizes the steps underway,
with EPRI support, to further develop these
second-generation coal gasification processes.
Environmental Considerations
A large incentive for applying coal gasifi-
cation/combined-cycle gasification technology to
the production of electric power is the ability to
cope with increased requirements for reducing
emissions to the lower levels mandated by regu-
latory agencies. In the opinion of EPRI, the
status of second-generation gasification of coal
is now at a point where 100-MW capacity plant
handling 1,000 tons/day of coal could be de-
signed, constructed, and operated. Major objec-
tives of such a project are the acquisition of data
required for siting even larger plants and the
complete assessment of the environmental ef-
fects of electric generation by such advanced
techniques.
Based on the status of Texaco gasification,
which has been operated at a scale of 150 tons/-
day both in the United States and in Western
Germany, support of a demonstration plant at a
California location is under negotiation. The
assessment of the environmental impact is one
of the major objectives of the demonstration
plant. While typical laboratory environmental
TABLE 12. SECOND-GENERATION COAL GASIFICATION PROGRAMS
PROCESS
PROGRAM
COMBUSTION ENGINEERING
BGC SLAGGER
TEXACO
OPERATION OF S-TON/HR. P.D.U, ON
EASTERN COAL, VARIETY OF COALS,
ENRICHED AIR OPERATION
DYNAMIC TESTS USING U.S. COAL
(300 T/D)
DYNAMIC AND ENVIRONMENTAL DATA
(15 T/D) ^^^___
83
-------
data, as shown in Table 13, provide some infor-
mation, actual demonstration plant operations
are better data sources. Data from other gasifi-
cation technologies, such as conventional fixed
bed, are of little importance or relevance. Here
again, environmental information is an essential
part of the project and cannot be separated
from the necessary R&D.
TABLE 13. TYPICAL WATER SLOWDOWN QUALITY
pH
TOTAL ORGANIC CARBON
TOTAL INORGANIC CARBON
AMMONIA
FORMATE
CHLORIDE
SULFIDE
SULFATE
CALCIUM
MAGNESIUM
SODIUM
THIOCYANATE
TH10SULFATE
FLUORIDE
CYANIDE
ALUMINUM
SILICON
IRON
C6 +VOLATILE ORGANICS
TOLUENE
BENZENE
ALL OTHERS
87
230 ppm
445 ppm
1020 ppm
492 ppm
432 ppm
264 ppm
166 ppm
140 ppm
IOO ppm
80ppm
7O ppm
69 ppm
39 ppm
31 ppm
20 ppm
5.0 ppm
37 ppm
20 ppb
IO ppb
< IO ppb
84
-------
MONITORING AND TESTING PROGRAM OF LOW-Btu GASIFIERS
K. E. Cowser*
Oak Ridge National Laboratory, Oak Ridge, Tennessee
and
G. V. McGurl and R. W. Wood
U.S. Department of Energy, Washington, D.C.
Abstract
Demonstration offow-Btu gasifier technology
includes an extensive environmental and health
study as part of the Gasifiers in Industry Pro-
gram sponsored by the U.S. Department of En-
ergy (DOE). Monitoring and testing plans have
been developed to investigate the gasifiers
located on the campus of the University of
Minnesota-Duluth and in a planned community
development at Pike County, Kentucky.
Four general areas of study are emphasized in
the plans including on-line studies, in-plant
studies, and local area studies to be integrated
through multidisciplinary assessments. This
paper provides a description of the processes
and facilities, of the rationale for the en-
vironmental and health study, and of the prin-
cipal program components. Results are limited
to startup experience at the UMD gasifier.
INTRODUCTION
The Gasifier in Industry Program (Gil) of the
U.S. Department of Energy (DOE) is part of a
broader activity to develop and improve tech-
nologies for converting coal to synthetic gas and
liquid fuels.1 Specifically, this program involves
demonstrating the integration of existing low-
Btu gasification technology in various opera-
tional environments. State-of-art technology is
to be applied in six selected gasifier projects,
one located on the campus of the University of
Minnesota-Duluth (UMD) and another included
in a planned community development at Pike
County, Kentucky (PCK).
Information to be gathered during the dem-
onstration period will consider questions of en-
vironmental acceptability as well as those re-
lated to technical and economic uncertainties.
•Speaker.
DOE requested that the Oak Ridge National
Laboratory (ORNL) develop for their considera-
tion comprehensive, environmental, and health
plans to study the UMD and PCD gasifiers.2 3
The final version of each plan incorporates U.S.
Environmental Protection Agency (EPA) and
National Institute for Occupational Safety and
Health (NIOSH) comments made through an En-
vironmental Working Group, GIL Following is a
description of the monitoring and testing ac-
tivities involved in developing the environmen-
tal and health data base.
Characteristics of Program Plan
A number of environmental and health con-
cerns in coal gasification were identified
previously by DOE and listed in an Environ-
mental Development Plan.4 The issues and in-
formation requirements to satisfy these con-
cerns for coal gasification were subsequently
enhanced by staff of the Assistant Secretary for
Environment, DOE. It was the latter determina-
tion of environmental and health tasks that was
used to guide plan development.
Design of the program plans is based upon
several premises:
• The study period will be limited to 3 yr;
• State-of-the-art capabilities in monitoring
and testing will be applied wherever practi-
cable;
• The first-year program will emphasize scop-
ing and screening activities to delineate the
requirements for more detailed investiga-
tions; and
• Program activities will be conducted with-
out interrupting normal plant operation.
Although efforts will be made to utilize methods
and instruments already available, some devel-
opment in monitoring and testing protocols may
be required to address unexpected problems.
Screening activities during the first year will be
85
-------
followed by detailed investigation of the major
concerns and important constituents, and by in-
itiating monitoring and testing efforts into po-
tential problem areas identified in the screening
studies.
Four general areas of study are emphasized
in the study plan: on-line studies, in-plant
studies, local area studies, and multidisciplinary
assessments that encompass the entire effort.
• On-line studies, or process characterization,
provide guidance for sample testing and in-
formation for control technology evaluation;
• In-plant studies provide information for oc-
cupational health controls and for correla-
tions of potential to actual personnel ex-
posures;
• Local area studies identify pollutant fate
and potential effects and confirm projections
based upon effluent monitoring; and
• Local impact assessments are concerned
with the potential impacts on health and en-
vironment and with the adequacy of envi-
ronmental and engineering controls.
Because the study plans for the two gasifiers
are similar, subsequent descriptive material
focuses on the UMD facility. Discussion of the
PCK plant is limited to description of the proc-
ess.
PROCESS MEASUREMENTS
AND CONTROLS
Process conditions at the two projects will
differ in that different gasifiers, feed stock, and
environmental control systems will be used.
UMD Project
An existing oil-fired heating plant at UMD
has been converted to burn low-Btu gas pro-
Buck«t Elevator
ORNL-DWG. 78-13526
UNIVERSITY OF MINNESOTA
HEATING PLANT DULUTH CAMPUS
STOIC TWO-STAGE GASIFIER
To Atmosphere
Suck
Fmc« (To Binl
Figure 1. Schematic diagram of University of Minnesota-Duluth heating plant.
86
-------
duced by coal gasification.5Tar byproducts from
the gasifier will be collected and used for peak
heating requirements in an existing oil-fired
boiler. The major components of the heating
plant, illustrated in Figure 1, include a coal
handling section, the gasifier, environmental
control devices, and the boiler-steam-off gas sec-
tion.
A Wyoming bituminous coal from the Elkol
Mine, containing 6.6 percent ash and 0.5 percent
sulfur, will be the initial feed stock. Coal will be
received at the Duluth docks as 90 percent 11/4
in x 3/8 in, then screened and trucked to the
heating plant. Several other lignite and bitumi-
nous coals have been proposed for testing.
After tramp iron removal and another screen-
ing for fines removal, coal will be dropped
through purged lock hoppers into the gasifier.
The gasifier is a Foster-Wheeler, Stoic two-
stage design. Gas and tars are produced as the
coal falls through the 250° to 1,100°F devolatili-
zation zone and are removed from the top of the
gasifier. Combustion and gasification of the de-
volatilized coal in an 1,000° to 1,800° F zone, fed
by air and steam, produce bottom gas. Ash is re-
moved beneath the gasifier from a water-filled
pan, which serves to quench the hot ash and seal
against operating pressure (less than 50 in H20).
Top and bottom gases must be cleaned of tars
and particulates before combination into boiler
feed. Because bottom gas at 1,100° F is primari-
ly laden with particulates, a hot cyclone re-
moves the dust for storage or disposal. In con-
trast, top gas (250° F) will contain tars and some
particulates, which will be removed in a hot
electrostatic precipitator and stored in heated
underground tanks for use as boiler feed during
the winter months.
Two modified 25,000-lb/hr steam boilers will
burn low-Btu gas. Tars collected from the un-
derflow of the electrostatic precipitator will be
burned directly in an existing 50,000-lb/hr Com-
bustion Engineering boiler. Gas-fired boiler flue
gases vent to the main heating plant stack,
while tar-fired boiler flue gases vent to a stub
stack. Figure 2 shows the recently completed
addition of the gasifier to the heating plant, on
which shake-down test began October 24,1978.
PCK Project
The gasifier plant now under construction at
the Douglas site in Pike County, Kentucky, will
support a multiuse community composed of resi-
dences, a hospital, a school, municipal buildings,
and industries. As such, it will initially provide
both hot and chilled water, and in the future,
low-Btu producer gas. The project, scheduled
for completion by early 1980, is shown schemat-
ically in Figure 3.
Two standard design, air blown, agitated
fixed-bed Wellman-Galusha gas producers will
be installed in this facility. Each has been
designed to handle 3,000 Ib/hr of Pike County
coal selected to meet air effluent standards for
sulfur emissions. The producer gas will be indi-
vidually piped to two steam-producing boilers.
Each gasifier system is capable of being oper-
ated independently or in parallel. A standby
supply of fuel oil will be used to meet excess de-
mand. Steam from the boilers will be used to
produce hot and/or chilled water by the use of
three steam hot water converters and two
steam absorption chilled water generators.
Kentucky bituminous coal from local Pike
County mines will supply the feedstock for the
plant. This relatively low-sulfur coal (0.8 to 2.0
percent) is in good supply and will be delivered
directly to the site by truck. A 30-day supply of
coal will be stored in a covered and floored
storage area and conveyed by front-end loader
to a coal feed pit outside the plant.
After screening and crushing, the coal will be
conveyed to storage bins, one for each gas pro-
ducer. Coal will be injected into the gasifiers
through coal valves and will travel downwards
through a coal devolatilization stage, a reducing
zone, an oxidation zone, and an ash zone. Air
will be introduced through an annular water
jacket, become saturated with water vapor, and
enter the gasifier just below the slowly revolv-
ing eccentric grate.
Gas will exit the reactor at 1,000° to 1,200° F
into a cyclone where large particles of ash will
be removed. Gas can then be piped to two boil-
ers to produce steam used to heat water or op-
erate a steam absorption refrigeration unit or
sent to the gas cleaning system. Initial plans call
for the reactor gas to be utilized directly in the
boilers, with a gas desulfurization system com-
ing on-line in the future. A secondary cyclone
separator will be utilized to remove the majori-
ty of the remaining particulate matter escaping
from the boilers with the combustion gases.
87
-------
•J.
V.
Figure 2. University of Minnesota Duluth heating plant and coal cjasifier.
-------
PIKE COUNTY PROJECT PIKEVILLE, KY
TWO WELLMAN-GALUSHA GASIFIERS
SINGLE STAGE
ORNL-DWG 78-21484
TO PUMPS
TO ATMOSPHERE
BUCKET ELEVATOR
BITUMINOUS
COAL
CRUSHER
CONVEYOR
Figure 3. Schematic diagram of Pike County coal gasification facility.
Process Sampling and
Characterization - UM D
Numerous sampling points have been desig-
nated to achieve the requirements of process
measurements. Locations of each sampling
point at the UMD project are identified in the
flow schematic of the heating plant (Figure 4).
The details of process sampling and analyses
are described in the UMD product plan.2 In
general, process sampling strategy provides for
characterizing materials introduced into the
process, the intermediate or final product, and
the recycle or waste streams.
The sampling schedule, analytical proce-
dures, and constituents or parameters to be
measured were chosen to allow early measure-
ment of traditionally monitored or suspected
materials, and to maximize the probable detec-
tion of unexpected and hazardous constituents.
Results must be adequate to document process
conditions, to evaluate the efficiency of en-
vironmental control technology, to identify lim-
itations in sample size or analytical methodolo-
gies, to identify possible biological hazards in
potential fugitive emissions, and to establish
priorities for subsequent bioassay.
On-line instrumentation required for process
sampling and monitoring is summarized in
Table 1. Gas chromatographs will monitor the
primary gases (N2, CO, C02, H2), water vapor,
and sulfur compounds (H2S, COS, CS2) at the
electrostatic precipitator, cyclone, and stack ef-
fluents. SOX and NOX will be monitored initially
in the main stack effluent. Grab samples and
samples classified by use of four special sampl-
ing trains will also be used in process and ef-
fluent characterization. Table 2 includes a
general description and the intended applica-
tion of each sampling train. Twenty-three chem-
ical and physical tests will be used initially in
characterizing some 400 process samples col-
lected the first year.
89
-------
ORNL-DWG. 77-1838A
<£>
O
STORAGE
Figure 4. Flow schematic and sampling points for University of Minnesota-Duluth heating plant.
-------
TABLE 1. ON-LINE INSTRUMENTATION FOR CONTINUOUS PROCESS MONITORING
Instrument Monitored streams
Gas chromatograph
Gas chromatograph
Gas chromatograph
Continuous monitor
Continuous monitor
13, 17
18, 20
13, 17
18, 20
13, 17
18, 20
18
18
N2, CO,
N2, 02,
H20*
H2S, COS
ethyl
S02
S0x
NOX
Analysis
C02 , C} , C2 > ^3
C02
, CS2, methyl mercaptan,
mercaptan, thiopene
/ORNL-DWG.\
(78-13533J
OCCUPATIONAL EXPOSURE
AND EFFECTS
Potential exposure of man in the working en-
vironment includes consideration of plant area
controls and effects on man if exposures occur.
Monitoring and testing activities thus involve
the requirements of worker protection and the
potential effects of exposure to primary efflu-
ents and fugitive emissions.
Plant Area Sampling and
Characterization—UMD
The primary objective of an industrial
hygiene program is to recognize, evaluate, and
. control exposures that may be capable of pro-
ducing overt health effects. An industrial
hygiene and medical surveillance program has
been established in cooperation with the Uni-
versity. The University has prime responsibil-
ity for protecting the health of its employees,
and we have participated to complement the
University requirements and to provide infor-
mation for occupational health control assess-
ments.
Two types of monitoring for potential ex-
posures are provided. Area monitoring for CO,
PAH, NH3, NOX, fugitive emissions, heat, noise,
and various chemical stresses indicates possible
exposures and will be accomplished by various
instruments providing real-time monitoring. A
partial listing of area monitors and their func-
tions is provided in Table 3. Personnel monitor-
ing defines the actual exposures. A variety of
standard industrial hygiene techniques employ-
ing filter cassettes and gas badges will be used
to define the time-weighted exposures to gase-
ous and particulate contaminants.
Medical surveillance is necessary to ensure
full protection of all personnel involved in
operating and maintaining the gasifier. Informa-
tion recorded by such surveillance will be cor-
related with results of personnel monitoring
and become part of the assessment activity. The
University provides for complete physical ex-
aminations, with special attention given to skin
abnormalities and sputum cytology tests for em-
ployees at the gasifier.
Occupational Toxicology
The principal focus of occupational toxicology
is the testing of primary effluents and fugitive
emissions for potential effects on man. Informa-
tion will be developed in response to questions
of relative toxicity of byproducts and effluents,
toxicity variation with process conditions, and
toxicity potential of fugitive emissions.
A two-level bioassay program is designed to
test effluents and potential fugitive emissions.
Level one, or cellular bioassays, will be used to
ascertain how the relative toxicity of effluents,
91
-------
TABLE 2. SAMPLING TRAINS FOR UMD GASIFIER
Gas
sampling
train
1
2
3
4
In-stack Heated
particle isokinetic
saapler probes
X
X X
X
X
Heated Knock-out drum. Gas XAD-2
3-cyclone electrostatic Heated cooler/ organic
seriesa>* precipitator3 filter0 condenser sorbent
X XX
X
X XXX
X XXX
Ice-cooled impingers Vacuum pump
Reagent and dry
solutions Empty Orierite test meter Purpose
XX X Measurement of tar loading.
XX X Measurement of particulate
loading and sizes.
X XX Assessment of tar loading;
collection of samples for
organic, aqueous, and trace
element analyses.
X XX Assessment of particulate
loading and size; collec-
tion of samples for organic .
aqueous, and trace element
analysis.
aHeated to 300°f to prevent water condensation.
^Alternative staged particle separators could be used.
-------
TABLES. AREA MONITORS
ORNL-DWG. 78-13539
Control pollutant
Type of instrument and capability
CO
NH3, NO, S02,
naphthalene and
its derivatives
Respirable aerosol
and dust particles
(coal dust, tarry
fumes and ash
particles)
Multipoint, continuously operating sensor station
for CO analysis with visual and audio alarm
Second derivative, UV absorption spectrometer
with multipass gas cell for real-time monitoring
of selected effluents
Piezobalance, portable monitor for measuring
respirable aerosols with mass concentrations
readout each minute; analyses for particulate
polycyclic aromatic hydrocarbons (as benzene
solubles) to determine integrated exposures will
be conducted as part of the conventional indus-
trial hygiene program; attempts will then be made
to correlate the mass concentration and.benzene
soluble fraction-for specific locations in the
gasifier plant. If such correlations are found
to exist then one would have indirect, but near
real-time method for measuring benzene solubles
byproducts, and fractions thereof vary with
process changes, to screen for further testing,
and to correlate with whole animal, somatic ef-
fects. Tests in this category use a variety of
biological systems, including bacteria, yeast,
and mammalian cells, to investigate mutagenic
effects. These shorter term tests will provide
guidance and be complemented by longer term
validating assays using drosophila, cultured
mammalian cells, and whole animal (mouse) sys-
tems. Not all tests will be run on all samples col-
lected at a given point, but priorities will be
established based on the biological activity
detected in the screening assays.
Level two, or mammalian somatic toxicity
tests, complement mutagenic and cytotoxic
testing. These assays use whole animals to
characterize the acute, subacute, and chronic
toxicity of products and effluents. Initially, only
selected samples will be used in the more expen-
sive toxicity tests, with selection based on the
probability of direct or indirect human exposure
and on current information of potential emis-
sions. Additions to the toxicity testing program
are anticipated as the information base on
biological activity develops.
ENVIRONMENTAL FATE AND EFFECTS
Environmental area monitoring includes sam-
ple collection and analyses, operation of con-
tinuous monitors, and application of appropriate
ecological toxicity tests. Information derived
from these activities is used to characterize and
quantify air, water, and solid effluents that may
impact the immediate environs of the plant.
Design of the monitoring program for the
UMD environment considers the ambient envi-
ronmental conditions and the expected oper-
ating characteristics of the gasifier. The follow-
ing information guided development of the mon-
itoring program:
• The Duluth-Superior urban area is in-
dustrialized, and operation of the heating
plant is not expected to modify the ambient
air to a discernable level;
93
-------
• Water use at the gasifier is expected to be
primarily consumptive and not result in any
liquid effluents; and
• The principal solid waste is ash from the
gasifier.
Two instrumented monitoring stations will
primarily monitor criteria air pollutants (CO,
NOX, hydrocarbons, S02, oxidants, and par-
ticulates) with periodic sampling for total
organics and organic speciation. The monitoring
scheme and sampling frequency are listed in
Table 4. If stack monitors indicate sufficient ef-
flux of noncriteria pollutants (e.g., COS, NH3,
HCN), additional measures will be adopted to
monitor for these pollutants.
Water quality measurements will be limited
to samples taken from wells in a sanitary land
fill used for ash disposal. In the event of unusual
plant operating conditions, liquid effluents and
surface streams will be monitored. Gasifier ash
will be leached to investigate this important en-
vironmental parameter, and the water samples
and leachates will be analyzed for a variety of
organic and inorganic constituents. Screening
activities will be used, as appropriate, to test
the toxicity, transport, degradation, and bioac-
cumulation characteristics of either whole ef-
fluent streams, selected chemical fractions, or
specific model compounds.
ASSESSMENTS
Site-specific assessments will be used to en-
sure maximum integration and utilization of in-
formation developed by the program elements
of sample collection, analytical characterization,
biological and environmental testing, and occu-
pational control and medical surveillance.
Analyses of potential impacts include consid-
eration of:
• Human health-related assessments, includ-
ing the industrial worker and the general
public;
• Ecologically related assessments, both ter-
restrial and aquatic systems in the site area;
and
• Operational assessments involving: environ-
mental control equipment, its efficiency and
reliability; and occupational health control
and the engineering systems used to reduce
fugitive emissions.
Information developed through these assess-
ments will be combined with information from
studies of other low-Btu gasifiers and will be
ORNL-DWG. 78-13546
TABLE 4. ON-LINE ENVIRONMENTAL MONITORING
Analyses
Instrumentation
Sampling frequency
Particulates:
Infra-red spectrometer
Chemiluminescence detector
Gas chromatograph
Flame photometric detection
Chemiluminescence detector
Total particulatesa High volume sampler
Continuous
Continuous
Continuous
Continuous
Continuous
24-hr sampling,
collection weekly
Gravimetric analyses carried out by sampling personnel.
94
-------
used to investigate potential impacts of antici-
pated industry growth.
Sample and Data Management
Successful execution of this program requires
that a large number of samples be characterized
by many investigators, and that the data and in-
formation developed be of high quality and
readily accessible in assessment activities. Sev-
eral thousand samples subjected to numerous
analyses and tests and the on-line monitoring
equipment output must be handled the first
year. Both sample and data management are re-
quired.
Initially, all samples other than those
characterized onsite will enter the Sample
Management Center at ORNL. Samples will be
treated as required, forwarded to project
leaders responsible for various discipline-
oriented tasks, and distributed to individual in-
vestigators. The Center will serve as the inter-
face between the UMD sampling staff, the disci-
pline task groups, and the Data Management
Center.
The Data Management Center will provide a
computerized data management system for
storage and retrieval of data and information
and will include: structure for data base devel-
opment; procedures to ensure proper identifica-
tion and recording of data; network to provide
user access to the files; and data analyses
routines.
PROGRAM IMPLEMENTATION AT UMD
All on-line instruments to monitor process
streams, stack effluents, areas within the plant,
and environmental air quality have been install-
ed. Several modifications were made in the
original plan. For example, the process gas
chromatographs were equipped with an addi-
tional cleanup system consisting of electrostatic
precipitators and perma-pure dryers, with the
latter installed to permit optional use of a dry-
ing step during GC operation. Preliminary test
results by Radian Corporation indicate that re-
moval of aerosols from the sample stream
should reduce GC maintenance requirements
without affecting the concentration of the com-
ponents monitored. A data acquisition system
has been provided to monitor process variables,
and computer programs were developed to per-
mit visual display of these variables on a real-
time basis.
Plant operators and environmental monitor-
ing personnel have completed initial tests as
part of the medical surveillance program. The
program under University direction consists of
complete physical examinations and laboratory
studies including routine blood analyses,
pulmonary function tests, audiograms, elec-
trocardiograms, chest X-rays, and color photog-
raphy of the skin. Sputum cytology testing has
also been recommended. All records become
part of the UMD Health Service file.
Our original plan provided only general guid-
ance for in-plant worker protection. Sub-
sequently, in collaboration with University of-
ficials, an industrial hygiene monitoring strat-
egy was developed specifically for the UMD
Gasifier. Program details were identified after
completion of the major components of the
facility and were the results of a site visit by a
team of industrial hygienists and engineers. Ma-
jor consideration was given to monitoring re-
quirements for carbon monoxide, particulate
polycyclic aromatic hydrocarbons (PPAH), heat,
noise, miscellaneous chemical stresses, fugitive
emissions, and personnel sampling. Eleven loca-
tions were identified for continuous CO monitor-
ing and sample collection for PPAH analyses,
portable CO monitors and UV light were speci-
fied for fugitive emission surveys at potential
points of leakage, and CO dosimeters and per-
sonnel air sampling devices were recommended
for worker application. This strategy will be
reviewed after 3 mo of plant operation.
The gasifier has been operated during two
separate periods, primarily to determine oper-
ating characteristics and the need for modifi-
cations in the coupled equipment. Coke was the
fuel used most often although several short-
term runs were made with a coke and coal
blend. Not unexpectedly, a number of leaks oc-
curred at flanges, valves, gaskets, and seals;
their repair required immediate recommenda-
tions by the industrial hygienist for worker pro-
tection.
Many of the detected leaks were of small
volume and did not cause acute exposure to car-
bon monoxide. However, several major emis-
sions took place, which required a change in
gasifier operation and a clearing of the area un-
til levels were reduced to less than threshold
limit values (TLV). Three types of CO monitor-
95
-------
ing were employed: fixed area monitors, hand-
held portable analyzers, and personnel dosime-
ters. Sixty man-days of dosimeter data from the
first shake-down period showed that no 8-hr
time-weighted average (TWA) was over 20 ppm
and that most were less than 10 ppm
(TLV/TWA = 50 ppm).
The value of an industrial hygiene capability
especially during startup operations is well
documented by the above example. This experi-
ence will provide a useful guide to personnel
protection not only when the UMD Gasifier
becomes operational but also for other similar
facilities. Details of this experience will be
discussed in subsequent reports.
CONCLUDING REMARKS
Design of the monitoring and testing program
for UMD involves all of the uncertainties in the
characteristics of a gasifier only recently opera-
tional, and consequently in the nature of the
process streams, byproducts, and effluent
streams. Parameters and tests were chosen ini-
tially to focus on screening methodologies as op-
posed to only selected constituents. Program
changes can be expected after the first full year
of study, with emphasis on investigating the
more significant components.
REFERENCES
1. Gasifiers in Industry: a Program of Coal
Conversion and Utilization. U.8. Energy
Research and Development Administration.
ERHQ-0015. August 1977.
2. Cowser, K. E. (ed.). Proposed Environmental
and Health Program for University of Min-
nesota Gasification Facility. Oak Ridge Na-
tional Laboratory. January 23,1978.
3. Cowser, K. E. (ed.). Proposed Environmental
and Health Program for Pike County Coal
Gasification Facility (draft). Oak Ridge Na-
tional Laboratory. December 18,1978.
4. Environmental Development Plan (EDP):
Coal Gasification Program FY 1977. U.S.
Department of Energy. DOE/EDP-0013.
March 1977.
5. Soderberg, W. E. Coal Gasification Duluth
Campus Heating Plant, Program Opportuni-
ty Notice for the Integration and Evaluation
ofLow-Btu Coal Gasification Technology in
Operational Environments. University of
Minneapolis. July 9,1976.
6. Rutherford, W. T. Coal Gasification for a
Utility Heating/Cooling Plant, Douglas Site,
Pike County, Kentucky, Volume I, Technical
Proposal. Pike County, Kentucky. July 6,
1976.
96
-------
THE USE OF LOW-Btu GAS FOR IRON OXIDE PELLET
INDURATION: AN INTERIM REPORT
R. K. Zahl* and J. C. Nigro
Twin Cities Metallurgy Research Center, Bureau of Mines,
U.S. Department of the Interior, Twin Cities, Minnesota
Abstract
The U.S. Department of the Interior's Bureau
of Mines is conducting a pilot-plant test pro-
gram to evaluate low-Btu gas generated from
the gasification of bituminous, subbituminous,
and lignite coals. The program explores the
technical feasibility of utilizing the gas as an
alternate fuel for high-temperature induration of
iron oxide pellets. A 2-m diameter single-stage,
atmospheric, fixed-bed gasifier has been erected
at the Bureau's Twin Cities (Minn.) Metallurgy
Research Center adjacent to its pilot pelletizing
facility. The gasifier provides a hot, raw, low-
Btu gas for firing iron oxide pellets in an 0.85-m
diameter by 10.7-m long rotary kiln. Pellets were
made from commercial magnetic taconite con-
centrations and were processed at a rate of 430
kg/hr. Gasifier operating data at fuel rates of
900 to 1,200 kg/hr are presented for coke, eastern
Kentucky bituminous coal, Colorado-Wyoming
subbituminous coal, and a North Dakota lignite,
along with corresponding coal and gas analyses,
Btu content, and thermal efficiency. Approxi-
mately 90 to 140 Mg of raw coal was processed
in each of the 5-day around-the-clock tests.
Pellet induration with low-Btu gas is described,
and some preliminary information on pellet
quality is presented.
INTRODUCTION
The U.S. iron ore pelletizing industry pro-
vides about 60 percent of the nation's iron ore
consumption and relies heavily upon natural gas
for pellet induration, consuming an estimated
1.7 km8 annually. During the past few years, the
industry has seen its gas supplies interrupted
while costs have tripled. In the short term, sup-
plies are likely to become even more restricted,
interruptions longer and more frequent, and
costs may be further increased. In the long
term, natural gas may not be available to the in-
•Speaker.
dustry. Presently, fuel oil is the only proven al-
ternative that is similarly restricted, costly, and
unreliable as a future base fuel because of de-
clining supplies.
To face the challenge of dwindling premium
fuel supplies to an industry so basic to the na-
tion's economy, the Federal Bureau of Mines
and others had previously examined direct coal-
burning methods as an alternate energy source.1
Pulverized coal-firing tests at both the pilot
plant and commercial scale have shown that no
single coal type is compatible with the three
commercial pelletizing systems. For the grate-
kiln process, only the premium quality eastern
coals with high ash fusibility temperatures have
shown real promise. Coals having ashes with
low fluid temperatures are required for the
straight-grate process because the prototype
design consists of external "wet bottom" com-
bustion chambers. The shaft-pelletizing fur-
naces have not yet been adapted to direct coal
firing. The major problems are the distribution
of powdered coal to a large number of combus-
tion chambers and the potential for blockage
with coal ash of the inaccessible refractory
passageways. The general conclusions regard-
ing coal firing for iron ore pelletizing are that
although it can be used, coal selection will be
restrictive, premium coals may have to be used,
and with the required plant modifications and
increased refractory costs, it may be no more
economical than some form of coal gasification.
A recent study2, funded by the Bureau and
conducted by the Arthur G. McKee Company,
pointed out that with currently available
technology, production of a hot, raw, low-Btu
gas generated by an atmospheric producer
would offer a viable, economical alternative to
natural gas or oil. An onsite facility would pro-
vide a high overall thermal efficiency and mini-
mize the capital costs of the coal gasification
plant. This system of gas production would give
the pelletizing industry a wider selection of
coals and would be even more cost-effective if
the low-rank western subbituminous and lignite
97
-------
coals could be used. Although the study indi-
cated that the use of low-Btu gas for pellet in-
duration appears technically feasible, the prac-
tical aspects of using this fuel must first be
demonstrated on a pilot-plant scale.
The research program conducted by the Bu-
reau in its pelletizing pilot-plant facility is a
cooperative effort with the U.S. Department of
Energy (DOE) and a consortium of 17 companies
with interests in iron ore, coal, gas, and in-
dustrial engineering. The Bureau's goal is to
determine whether pellet firing with coal gas of
low heating value is technically feasible and
practical, while DOE is interested in gasifier
operations and technology. The U.S. Environ-
mental Protection Agency (EPA) is monitoring
the tests to characterize various gaseous and
liquid streams in the process. Coal gas to be
used in the pelletizing program will be derived
from gasifying bituminous, subbituminous, and
lignite coals.
The data presented in this paper represent
the initial test campaign. It is expected that the
project will be completed by the fall of 1979.
PROCESS AND PLANT DESCRIPTION
The Bureau's pelletizing facility is a fully in-
tegrated pilot plant capable of taking concen-
trate through all pelletizing steps of balling,
drying, preheating, induration, and cooling.
Plant capacity is nominally 500 kg/hr dry feed.
The balling circuit consists of a table feeder for
concentrate, a screw feeder for bentonite, a
belt-type paddle mixer for blending the bento-
nite binder with the concentrate, and a 1.5-m
diameter pelletizing disk to form the nominal
1.2-cm diameter green pellets. The green pellets
are first dried and then preheated, to approxi-
mately 1..270 K on a 0.3- by 3-m long traveling
grate with one updraft drying zone and two
downdraft preheat zones. Then, pellets are in-
durated in an 0.85-m diameter by 10.7-m long
rotary kiln operating at 1,570 K to 1,620 K and
discharged through a shaft-type cooler. The pel-
let cooler supplies preheated air to the kiln. The
low-Btu kiln burner is a scroll-type unit with ad-
justable register vanes for flame shaping. Com-
bustion air supplied to the burner can be pre-
heated to 720 K.
Pellet plant instrumentation and controls are
centrally located in a control room. Tempera-
ture and most pressure and flow data are ob-
tained with a data logger and later processed in
a computer. The low-Btu kiln burner control
system was designed to adapt quickly to
changes in gas composition. The gas flow is con-
trolled by kiln temperature, and the air flow is
controlled by a fully electronic flow ratio control
scheme. Producer gas flow to the kiln is meas-
ured with a "low loss" venturi flow element and
a mass flow computer.
The gasification pilot plant, shown in Figure
1, is adjacent to the pelletizing plant and in-
cludes a 2-m diameter, fixed-bed, atmospheric
producer with a water-cooled agitator arm and
has a nominal capacity of 1.35 Mg/hr of bitumi-
nous coal. Steam is self-generated by passing
the air over water heated by the gasifier cooling
jacket, whereby the air becomes saturated at
some desired temperature. The producer gas
flows through a refractory-lined dry cyclone and
is then transmitted via a 61-cm ID. duct to a
combustion chamber and a 20-cm ID. duct to the
pelletizing kiln. Both ducts are lined with 10 cm
of refractory. The combustion chamber is de-
signed to match the full capacity of the gas pro-
ducer because, at maximum output, the pellet-
izing kiln would use only 10 to 15 percent of the
producer gas. A scroll-type burner is also in-
stalled on the combustion chamber and includes
adjustable register vanes to control flame
shape. Exhaust gases from the combustion
chamber are cleaned with an impingement tray-
type scrubber with pH control. A combination
ignitor-incinerator is installed on the gasifier
cent stack to ignite the gases during flaring or
completely combust the small amount of gases
generated during banking. The gasifier building
was constructed to satisfy an electrical
classification of Class I, Group B, Division n,
and includes a building exhaust fan, open grat-
ing floors, and a hooded vent over the coal feed
bin. Carbon monoxide monitors are present at
three locations with alarm capability between 0
and 100 ppm.
Instrumentation and controls for the gasifier
are centrally located in a control room adjacent
to but isolated from the main operating floor of
the gasifier building. In addition to the normal
complement of instrumentation for such a pro-
ducer, the 9-Mg capacity coal storage bin is sup-
ported on precision load cells, and the producer
gas flow is measured with a "low loss" venturi
98
-------
Cool Vent and flare
<£>
CO
Ash
Green
pellet
feed
Sludge
Pellet
product
Figure 1. Process flow diagram of Bureau of Mines Gasification and Pelletizing
Pilot-Plant, Twin Cities Metallurgy Research Center.
-------
flow element. The air flow is measured with an
Annubar* flow element and controlled by the
producer offtake pressure. The system is de-
signed to operate with a maximum offtake pres-
sure of 1.25 kPa; the ducts, flow elements, etc.,
were designed to contribute no more than 0.5
kPa permanent pressure loss at two-thirds
rated capacity. Temperature, pressure, and
flow data are obtained with a data logger and
later processed by computer. Combustion cham-
ber instrumentation is minimal, and most con-
trols are local and manual. The chamber pres-
sure is automatically controlled and a tempera-
ture profile is recorded. Flame supervision at
the combustion chamber burner includes inter-
locks to shut down and vent the gasifier when
the 61-cm gas safety shutoff valve to the burner
closes. A fully automated gas chromatograph
with a thermal conductivity detector provides
the producer gas analyses. The gas sampling
and conditioning system, which was built in-
house, was designed to obtain tar, oil, and
moisture contents and to deliver a dry, clean
gas to the chromatograph.
DESCRIPTION OF INITIAL
TEST CAMPAIGN
The initial campaign consisted of one 7-day
and three 5-day continuous tests conducted
from November 13 to December 15, 1978. All
tests contained downtime of one to three shifts.
The gasifier was banked between tests.
The test program was based on the assump-
tion that the producer would gasify bituminous,
subbituminous, and lignite coal under reason-
ably stable conditions and would produce gases
typical of atmospheric producers for these fuels.
The producer was originally designed for tar-
free coke and anthracite, and it was expected
that some modifications would have to be made
to the unit for it to operate successfully with
this wide range of fuels. The first few tests
would yield information on the required modifi-
cations.
Radian Corporation, Austin, Texas, repre-
senting EPA, was onsite during the entire test
campaign and obtained gas analyses of the com-
bustion chamber exhaust and producer gas dur-
ing the last three tests. Radian's analyses were
obtained with three gas chromatographs on
•Reference to specific trade names does not imply en-
dorsements by the Bureau of Mines.
each sampling system; producer gas samples
were obtained independently of the Bureau's
samples. There was some overlap in the suite of
constituents analyzed to allow better coordina-
tion of the results between the two systems.
During the test with lignite, a full-scale sampl-
ing program was run to fully characterize the
producer gas, all gasifier effluents, and combus-
tion chamber exhaust. Sampling included iso-
kinetic sampling of the producer gas and com-
bustion chamber exhaust gas streams. Samples
of water discharges were also obtained. Results
of Radian's sampling program are not included
in this report.
The operating philosophy for all tests was to
bring the entire pilot plant on-stream as quickly
as possible and to stabilize each section of the
plant later. After the gasifier was running, the
pelletizing plant and combustion chamber were
brought on-stream with natural gas prior to
switching the producer gas from the flare to the
combustion chamber. After the combustion
chamber and gasifier were on-line, the low-Btu
gas flow was started to the pelletizing kiln, and
pellet making commenced shortly thereafter.
System malfunctions, however, prevented the
startup sequences from proceeding smoothly.
Numerous equipment and system failures oc-
curred during the test campaign, and repairs
had to be made almost continually during the
operation. The abnormally cold weather caused
most problems because the ambient tempera-
ture decreased steadily throughout the cam-
paign. Because the plant was not fully winter-
ized, water lines, air lines, and valves required
an inordinate amount of attention. Consequent-
ly, the stable test periods, especially in the
pellet plant, were too short to attain steady
state. All coals, however, were gasified, and
pellets were indurated with gas from each coal.
Coke was used to start the gasifier and was
run for periods long enough to stabilize the fire
and heat the refractory-lined ducts. The first
coal to be gasified was a high-quality, closely
sized bituminous coal from eastern Kentucky.
The second and third coals were a Colorado-Wy-
oming subbituminous coal and a North Dakota
lignite. All coals were screened at 1.9 cm just
prior to loading into the storage bin via a bucket
elevator. Analysis and sizing of the "as fed" coal
are given in Table 1 and Table 2. The data were
developed from composite analyses of samples
taken four times per shift.
100
-------
TABLE 1. ANALYSIS OF FUELS "AS FED"
Solid fuel tested
Source
Proximate analyses,1 wt-pct
Moisture
Volatile matter
Fixed carbon
Ash
Ultimate analysis,1 wt-pct
Hydrogen
Carbon
Nitrogen
Oxygen
Sulfur
Ash
Heating value, kJ/kg
Ash fusibility,2 °K
Initial deformation
Softening
Fluid
Free Swelling Index
Coke
-
4.8
0.9
87.1
7.2
0.9
74.1
0.5
16.7
0.6
7.2
28,428
1,444
1,505
1,566
0
Bituminous
E.Kentucky
3.3
37.2
54.1
5.4
5.1
79.0
1.6
8.1
0.8
5.4
31,520
1,716
1,744
1,810+
4
Subbiturninous
Colorado-Wyo.
10.8
36.2
46.2
6.8
5.4
63.8
1.3
22.0
0.7
6.8
25,660
1,644
1,672
1,744
0
Lignit-
N.Dakc:a
30.6
2£.3
33.4
7.7
6.3
•U.7
0.6
39.8
C.8
7.7
17,562
1,322
1.35C
1,377
C
lnwet basis" as received.
2ASTM reducing conditions.
RESULTS AND DISCUSSION
Because of the nature of the initial campaign,
the tests yielded results primarily of a "mechan-
ical" nature, although significant process infor-
mation was also collected. The detailed equip-
ment modifications found to be required are not
reported here. Instead, observations are pre-
sented along with a small amount of process
data.
A subjective observation was made that the
gasifier operation was just "settling out" at the
end of a 5-day test. Because all operators agreed
on this point, it has been decided that 10 days
will be the minimum operating period for future
tests. This "settling out" period is a combination
of many factors, not the least of which is learn-
ing the behavior of the specific coal being used.
At 900 kg/hr of coal feed with 7 percent ash
(which is typical of the coals of interest to the
Bureau's test program), the coal residence time
is approximately 4 hr, while the ash residence
time is an additional 24 to 48 hr, depending on
ash bed depth. It can be easily seen, therefore,
that changes to the gasifier operation may not
fully show up for 2 days.
Gasifier operating difficulties were expected
with subbituminous coal and lignite because of
the friable nature of the fuels and the ash char-
acteristics. Low-rank coals generally have a
lower ash fusion temperature, and clinker for-
mation is considered a major problem. Gasifier
air saturation-temperature (i.e., stream con-
sumption) is the major method for controlling
clinker formation. Maintenance of ash bed depth
is affected by clinker formation, grate speed,
and the size of the grate openings. A variable
speed drive was installed on the grate; however,
this modification did not appear sufficient to
compensate for the wide differences in ash be-
101
-------
TABLE 2. COAL SIZING "AS FED"
Solid fuel tested I Coke { Bituminous I Subbltuminous | Lignite
Size, mm
Cumulative wt-pct passing
57.2
50.8
45.3
38.1
32.0
25.4
19.0
16.0
12.7
9.51
6.35
100
100
95.6
75.3
41.2
10.9
3.1
1.9
1.6
1.4
1.1
100
99.9
99.5
97.5
93.7
72.3
32.6
14.1
6.1
3 5
2 5
100
99.9
97.8
85.2
64.9
43.9
28.7
21.1
13.6
8 0
3 8
100
99.4
98 5
94.4
77.0
42 6
15.0
6.7
3.4
2 0
1 2
havior among the three coals. Although serious
clinkering problems were not encountered with
the subbituminous coal, the operation was not
smooth. Auxiliary steam was necessary at one
point because the fire zone became very thin
and the jacket water temperature was not high
enough to saturate the air to the desired level.
The operators were able to stabilize the opera-
tion after 8 hr, and self-generation of steam
became possible again. During the test with
lignite, however, clinkering was a serious prob-
lem, although the gasifier did not appear to
operate as poorly as expected. At the end of the
test, 80 percent of the 12.4-cm wide grate open-
ings were found to be solidly plugged with ash
fines, and very large clinkers were found float-
ing near the top of the ash bed. During all tests,
the air saturation-temperature was unsteady,
and the control scheme proved to be inadequate
for this test program. This contributed to the
generally unstable gasifier operation and the
clinkering problems with lignite.
Numerous other observations were made
during plant operations. It was noted that con-
trol and safety shutoff valves in the producer
gas lines should have oversized actuators for
reliable operation. Coal and char fines entrained
in the producer gas increased dramatically with
the low-rank coals as opposed to bituminous
coal. Quantitative measurements were not ob-
tained; however, a higher dust loading was
noted in the combustion chamber scrubber
water and gasifier cyclone water seal. Gasifier
operating data are given in Table 3, which
shows that the pressure drop across the bed in-
creased for the low-rank fuels. Also, as ex-
pected, the offtake temperature decreased for
low-rank fuels, but the temperatures obtained
were lower than expected.
Operation of the venturi flow elements was
102
-------
TABLE 3. GASIFIER OPERATING DATA SUMMARY
Solid fuel tested
Saturated air, m3 /hr
Saturation temperature, °K.
Coal burn rate, kg/hr
Bed pressure drop, kPa
Offtake temperature, °K....
Gas yield, m3 /hr
Coke
1,388
335
ND
.87
624
ND
Bituminous
2,355
334
657
1.19
692
3,058
Subbituminous
2,740
334
1,179
4.53
559
3,993
Lignite
1,789
332
1,161
2.71
416
3,143
better than expected. The flow element in the
20-cm ID gas duct to the palletizing kiln re-
mained operable until the last test with lignite.
During the lignite test, the low gas tempera-
ture, which caused tar, oil, and moisture conden-
sation, along with the high dust loading, com-
bined to build a 3-mm thick coating of coal fines
on the converging cone of the flow element and
a 1.5-mm thick coating on the venturi throat.
The pressure taps were coated over but were
not internally plugged. The flow element in the
61-cm gas duct became inoperative during the
first test because of plugging of the pressure
taps and would plug up again within minutes
after cleaning. Because both venturies were
well insulated, the difference in performance
was attributed to the orientation of the flow
elements. The small flow element was installed
in a horizontal duct, and its pressure taps were
installed vertically at the top, allowing conden-
sate to flow back to the process. The large flow
element was installed in a vertical duct (flow
downward), and its pressure taps were horizon-
tal, causing any condensate to stay in the taps.
The 61-cm venturi had not yet been removed for
inspection; there may be other problems asso-
ciated with its orientation.
Gasifier process data and material balances
are summarized in Table 4. The data shown are
for short periods during the tests and are based
on average rates for periods that were most
stable. The producer gas flow meter on the gasi-
fier outlet was inoperative, so the flow was
estimated and selected to yield the best total
weight and carbon balance. Carbon, hydrogen,
oxygen, nitrogen, and total weights were bal-
anced independently. For all balances pre-
sented, the output/input ratios were within the
range of the 0.9- to 1.1-range. Also, during the
test with subbituminous coal, the gas sampling
system was not operating well enough to pro-
vide a water vapor content in the producer gas.
Gas moisture content for this test was calcu-
lated with a hydrogen balance; dry gas analyses
for this test were also limited but are con-
sidered acceptable for characterizing the short
test period. Ash balances were exact because
they were based entirely on the known ash feed
rates. Because the carbon in the ash varied sig-
nificantly during the tests, the values used in
the balances were based on residence time esti-
mates and were not modifed to produce better
material balances.
Total thermal efficiency was calculated for
the hot, raw gas at the cyclone outlet and was
approximately 90 percent for all coals. Trans-
mission heat losses were an additional 1.8 per-
cent to the kiln and 2.9 percent to the combus-
tion chamber for bituminous coal with a 1.0-per-
cent total loss in the cyclone. For the sub-
bituminous coal, there was a 2.1-percent trans-
mission heat loss to the combustion chamber
with a 1.2-percent total loss in the cyclone.
Representative gas analyses are presented in
103
-------
TABLE 4. GASIFIER MATERIAL BALANCE AND PROCESS DATA SUMMARY
Solid fuel tested
Inpute, kg/hr
Air
Output, kg/hr
Input ratios, kg/ kg
Total thermal
Gas yield
Heating value, MJ/m3
Tar contribution (wet basis).
Sensible heat (wet basis)....
Bituminous
657
360
2,298
2,961
53.5
186
40.8
6.8
.55
3.51
92
4.68
5.55
5.18
.63
.48
Subbutiminous
1,179
425
2,663
3,717
62.6^
345
88.5
11.8
36
2 26
89
3.43
6.67
5.70
.606-/
.22
Lignite
1 161
239
1,800
2,602
44. 9^'
464
98 0
13 6
21
1 56
89
2 68
6.33
5 10
.63*/
.15
e/
— Estimated.
Table 5. The analyses do not necessarily repre-
sent the same operating periods that were used
for the balances shown in Table 4. The analyses
were corrected to air-free values and were ob-
tained from raw gas analyses having less than 3
percent oxygen with most oxygen contents in
the 1- to 2-percent range. Oxygen in the pro-
ducer gas analysis was assumed to come entire-
ly from leaks in the sampling system. The
sampling system performed very well during
the last test with lignite. Oxygen values of 1
percent or less were consistently obtained dur-
ing the last test after correcting for argon
superimposed on the oxygen peak. Tar and oil in
the gas from lignite were measured but were an
order of magnitude lower than values typically
reported. It was felt that the tar fraction con-
densed into droplets in the gas duct and the
sampling geometry prevented obtaining a rep-
resentative sample. Gas temperature at the
sampling point was above the dew point, so
moisture measurements were considered repre-
sentative.
Results of the pelletizing test with magnetic
taconite concentrates and the three coals are en-
couraging, but they are limited because of mech-
anical difficulties and a generally unstable
gasifier operation. Desired pelletizing tempera-
104
-------
TABLE 5. PRODUCER GAS ANALYSES
Solid fuel tested
Constituent, volume
pet.1
Ha
CO
CH4
C2H4
CaHs
CO,
Na+A
H30
Tar, g/m3
Gas heating value,
kJ/m3
Gross
Net
Tar contribution,
kJ/m3
Gross
Net
Sensible heat at
cyclone exhaust,
kJ/ra3
Coke
Dry
10.7
23.8
.001
.0
.0
9.0
53.6
ND
4,136
3,949
ND
ND
ND
Wet
10.1
22.4
.001
.0
.0
8.5
50.4
6.0
ND
3,875
3,726
ND
ND
376
Bituminous
Dry
18.3
26.7
1.7
.10
.20
6.2
46.8
17.7
6,260
5,812
633
604
ND
Wet
16.7
24.4
1.6
.09
.18
5.7
42.8
8.6
16.2
5,738
5,328
578
551
458
Subbituminous
Dry
16.9
28.5
2.0
.43
.14
6.5
46.9
ND
6,558
6,148
ND
ND
ND
Wet
14.7
24.9
1.7
.37
.12
5.7
40.9
12. #
ND
5,701
5,365
ND
ND
347
Licnitc
Dry
17.5
28.9
1.5
.15
.10
6.2
45.6
ND
6,297
5,924
ND
ND
ND
Wet
14.1
23.3
1.2
0.12
0.08
5.0
36.7
19.4
ND
5,067
4,769
ND
ND
149
1 Corrected air free.
By hydrogen balance.
ture were not sustained as a result of short-test
duration. Pallatizing data for all tests are sum-
marized in Table 6. The pellet strengths listed
were the best obtained during the tests. Al-
though the strengths were 25 to 40 percent too
low for a commercially acceptable product, they
were increasing in all cases. The strengths are
typical of induration at these lower than normal
temperatures.
Pellet chemistry is given in Table 7. The
analyses are also shown normalized to zero
Fe++ by "adding" the required oxygen. The
analyses indicate there may have been some
minor ash pickup by the pellets in the kiln dur-
ing the tests with the western coals. The high
sulfur content in the grate discharge pellets in-
dicates a recirculating load of sulfur between
the grate and kiln during all tests. These data
can be used qualitatively only because the anal-
yses represent short operating periods and only
a few samples. High ferrous iron in the pellets is
indicative of "underburning" and is consistent
with the low strengths attained.
SUMMARY AND CONCLUSIONS
Four 100- to 160-hr around-the-clock tests
were completed during November 13 to Decem-
ber 15,1978, in the Bureau's iron ore pelletizing
and coal gasification pilot plant. Pellets were in-
durated in the 10.7-m long rotary kiln at rates of
400 to 544 kg/hr using hot, raw, low-Btu gases
generated from an eastern Kentucky bitumi-
nous coal, a Colorado-Wyoming subbituminous
coal, and a North Dakota lignite. The minus
50.8-mm plus 19.0 mm sized coals were gasified
at rates of 900 to 1,180 kg/hr in a 2-m diameter
single-stage, fixed-bed atmospheric gas pro-
ducer originally designed for tar-free anthracite
and coke. The ability to gasify these widely dif-
ferent solid fuels in this gasifier, generate good
quality gas, and obtain or approach the
105
-------
TABLE 6. PELLETIZING DATA SUMMARY
Solid fuel tested
Concentrate feed rate
Average grate hood temperature, °K.
Bituminous
456
995
1 625
1.62
Subbituminous
424
955
1,520
1.49
Lignite
544
895
1,515
1.80
TABLE 7. PELLET CHEMISTRY SUMMARY
Constituent, wt pet
Grate feed
Grate dischar«e
Kiln product
Bituminous coal
Fe
Fe84
SiOa+AlaO-.
CaO+MgO
NajCH-KaO
S
66.0 (64.5 )!
21.8 ( 0 )
6.63 ( 6.47 )
.40 ( .39 )
.079 ( .077)
.016 ( .015)
65.7 (64.4 )
14.1 ( 0 )
6.56 ( 6.44 )
.41 ( .41 )
.067 ( .065)
.032 ( .032)
65.7 (64.9 )
7.6 ( 0 )
6.50 ( 6.43 )
.42 ( .42 )
.055 ( .055)
<.005 ( <.005)
Subbituminous coal
Fe
Fe2 +
SiOg+AlaO,
CaCHttgO
NajgO+KaO
S
67.0 (65.0 )
21.6 ( 0 )
5.55 ( 5.39 )
.46 ( .45 )
.05 ( .048)
.011 ( .011)
66.8 (65.2 )
16.8 ( 0 )
5.57 ( 5.44 )
.46 ( .44 )
.053 ( .052)
.022 ( .021)
65.5 (65.4 )
1.8 ( 0 )
5.69 ( 5.68 )
.48 ( .47 )
.043 ( .042)
<.005 ( <.005)
Lignite
Fe
Fe2 +
SiOa+AljjCb
CaO+ttgO
NaaO+KaO
S
66.2 (64.2 )
21.8 ( 0 )
5.79 ( 5.58 )
1.11 ( 1.07 )
.05 ( .048)
.015 ( .015)
65.9 (64.7 )
13.0 ( 0 )
6.13 ( 6.02 )
1.06 ( 1.04 )
.07 ( .07 )
.038 ( .037)
65.3 (64.8 )
5.3 ( 0 )
6.29 ( 6.24 )
1.19 ( 1.18 )
.06 ( .06 )
.004 ( .004)
1 Analyses in parentheses are normalized to zero Fe
independently rounded.
2 +
to show trends. Numbers are
106
-------
necessary palletizing temperatures in the kiln is
considered a major accomplishment of the in-
itial startup campaign. Although pellets were
indurated with the raw, low-Btu coal gas, fre-
quent interruptions in gas flow to the kiln as a
result of erratic gasifier operations prevented
achieving the fully stabilized kiln temperature
profile required to produce pellets of commer-
cially acceptable quality. The palletizing results
obtained from these initial tests indicate that
acceptable quality pellets probably can be made
from magentic taconite concentrates in a rotary
kiln with raw, low-Btu coal gas. However, fur-
ther demonstration tests under more stable
gasifier and palletizing operations are needed
before this approach can be considered for a full-
scale palletizing facility. Future tests must also
be run on hematite concentrates because both
the maximum induration temperature and total
heat requirements are greater than those re-
quirements for magnetite.
Problems in gasifier operation and in trans-
porting and burning the raw producer gas be-
came more prevalent with the decrease in coal
rank. It is now apparent that some modifi-
cations will have to be made to the gasifier to
allow safe and stable operations when high
moisture and friable western subbituminous
coal and lignite are treated. Some of the more
important future modifications will include:
• Changing the cooling water piping and
replacing the air saturation-temperature
controller for closer and more stable control
over steam consumption;
• Increasing the revolving grate spacings to
account for the different ash characteristics
of the low-rank western subbituminous coals
and lignites; and
• Installing traps in the gas ducting to pre-
vent buildup of tar and condensate at low
points in the distribution mains. The neces-
sary modifications to the plant are now un-
derway, and testing is expected to be re-
sumed in the summer of 1979.
REFERENCES
1. Nigro, J. C. Alternative Coal-Firing Methods
for Indurating Iron Ore Pellets. Mining Con-
gress J. 64(6): 19-26. June 1978.
2. Ashworth, R. A., K. C. Byas, and D. G.
Bonamer. Utilization of Low and In-
termediate Btu Gas from Coal for Iron Ore
Pelletizing. Bureau of Mines, U.S Depart-
ment of Interior. Open File Report 36-77.
1977. 283 pp. (Available for consultation at
each of the Bureau of Mines Metallurgy Re-
search Centers; at the Natural Resources
Library, U.S. Department of the Interior,
Washington, D. C.; and from National Tech-
nical; Information Service, Springfield,
Virginia, PB 264-702/AS.)
107
-------
Session II: ENVIRONMENTAL ASSESSMENT:
GASIFICATION
Charles F. Murray, Chairman
TRW
Redondo Beach, California
109
-------
SYNTHETIC FUELS IMPLEMENTATION
Orcutt P. Drury
Office of Domestic Economic Policy and Coordination,
U.S. Department of Commerce, Washington, D.C.
Abstract
The need is urgent to start building large syn-
thetic fuel plants. Supply side action is essential
to meet social expectations that rest on energy.
New energy supplies will help counter interna-
tional inflationary pressures. Synthetic fuels of-
fer us a significant mobilization alternative and,
in sufficient quantities, can provide a cap for
world oil prices.
Of the many implementation alternatives,
three suffice to show the range of opportunity
available: shale oil tax credit, competitive bid-
ding for Federal support, and The Petroleum
Substitutes Requirements Program (PETSUB).
INTRODUCTION
It is a real honor for me to be with you in this
symposium, for your work here is vital. There
are problems to be identified and resolved, but
the need for synthetic fuels is critical. There
may be little we can do to avoid brownouts dur-
ing this summer and successive summers. There
will be other energy shortages, for as a society
we have remained unconvinced about the energy
crisis for too long. Synthetic fuels will have to
fill part of the void, but the void is huge.
The void will be created by nuclear gener-
ating plants that are and will be closed, by other
plants—coal, hydroelectric, and nuclear—that
will not be built. To maintain a per capita zero
rate of energy growth for the growing world
population we need new electric utility plants
faster than they are now being planned or built.
But we need to increase our electric energy
growth rate to shift (indirectly) to coal. At a con-
struction rate we cannot maintain, we are
already over 2 years behind. We will not suc-
ceed in building all the electric plants we need,
but somehow, we need to fill the void—or at
least part of it—with synthetic fuels.
This energy shortage—and it is not only in
electricity—cannot be made up solely by order-
ly conservation. The difference can only be
111
made up by new investment or denial. The prob-
lem with denial is America's social promise: that
denied minorities can expect their share of what
only energy can provide. President Carter said
clearly that our energy crisis has the "moral
equivalent of war." To me, the moral aspect in-
corporates these social expectations. So I em-
phasize this basic energy need, heightened and
sharpened by social expectations for real goods
and services. It is easy for me to assert the
need, but I have two documents here—the Com-
merce Department's forecasting effort—that at-
tempt to prove it.1'2
Beyond a basic energy need and growing
social expectations, a need exists for an
emergency supply capability. Synthetic fuel
plants can give us this additional mobilization,
or emergency capability, in addition to below-
ground storage.
Finally, synthetic fuels produced in sufficient
volume offer a cap on world oil prices. Suc-
cessful work on the supply side of energy may
relieve international inflationary pressures.
Because we are limited to what we can accom-
plish by disciplining demand, we need to devote
renewed effort to the supply side, perhaps the
answer to many of our productivity problems.
How do we go about it? You know the envi-
ronmental problems, and I leave the technical
solutions to you. But, from a management view,
there are many implementation alternatives.
Three, however, will suffice to show the range
of opportunities open to us.
First, the Administration is working on a
draft bill entitled "Shale Oil Tax Credit of 1979."
Because a 50,000-bbl/d plant is expected to cost
$1.2 billion to build, and because shale oil is ex-
pected to sell profitably only at $5 or $6 more
than world oil, considerable incentive is needed.
This bill would provide a $3-credit against tax
liability that would shelter from $5 to $6 per
barrel and provide the needed incentive. This
kind of legislation could be applied to other syn-
thetic fuels.
Second, an earlier Department of Commerce
-------
proposal suggested that prospective producers
bid to produce units of 50,000 barrels of oil —or
the equivalent in Btu content of synthetic gas or
other fuel — with the bid including those aspects
of Federal assistance that the producers wanted.
This would encourage competition but would not
limit the kind of support any firm could seek:
front end grants, per barrel tax credits, guaran-
teed loans, guaranteed prices, etc. This proposal
is not active today, but it is representative of
the range of implementing plans that have been
considered.
Finally, PETSUB, The Petroleum Substitutes
Requirement Program,3 is under active consid-
eration by the Administration. Under that pro-
gram, natural gas distributors, importers, and
other large users of oil and gas would be re-
quired to purchase a specified amount of substi-
tute fuels. The amount would be a small fraction
of their earlier consumption of oil and/or gas.
Producers of substitutes would issue certifi-
cates of quantities delivered to prove customer
compliance.
The certificates would be salable so certain
specialized firms could escape use of the substi-
tutes where other firms had surplus certificates
to sell. All users would bear the program costs
as the synthetic fuel cost differentials were
passed through by refineries and others. Since
synthetic producers would not be limited in
price, market entry would be encouraged. Ac-
tive competition would be expected.
In closing, let me repeat that your job is vital;
technical solutions to identified problems are
needed now. The benefits are three:
• Social expectations can be met,
• Our mobilization capability can be expanded
for emergencies, and
• International inflationary pressure can be
countered.
To reach these benefits, significant im-
plementation alternatives are available to per-
mit your work to be brought to fruition.
But everyone must work quickly— our energy
supply and distribution system are tenuous. We
need to make significant changes swiftly and
carefully.
REFERENCES
1. Forecast of Likely U.S. Energy Supply/De-
mand Balances for 1985 and 2000 and Impli-
cations for U.S. Energy Policy. PB 266 240.
2. Preliminary Forecast of Likely U.S. Energy
Consumption/Production Balances for 1985
and 2000 by States. PB 287 486.
3. Petroleum Substitutes Requirement Pro-
gram. Sobotka & Co., Inc. March 12, 1979.
BIBLIOGRAPHY
Gustaferro, J. F., M. Maher. and R. Wing.
Forecast of Likely U.S. Energy Supply/Demand
Balances for 1985 and 2000 and Implications for
U.S. Energy Policy. National Technical Informa-
tion Service. PB 266 240.
Gustaferro, J. F., C. S. Warlick, A. M. Maher,
and R. Wing. Preliminary Forecast of Likely
U.S. Energy Consumption/Production Balances
for 1985 and 2000 by States. National Technical
Information Service. PB 287 486.
Sobotka & Co., Inc. Petroleum Substitutes Re-
quirement Program. March 12,1979.
112
-------
POLLUTANT EVALUATIONS FOR A LABORATORY SEMI-BATCH
COAL GASIFIER
John Cleland* and John Pierce
Research Triangle Institute, Research Triangle Park, North Carolina
Abstract
Nine U.S. solid fuels have been gasified in the
RTI laboratory unit. Oasifier streams have been
extensively and quantitatively defined in terms
of process and chemical pollutant parameters.
Experimental results have received preliminary
analysis on the basis of:
• Coal-associated influences on pollutant pro-
duction,
• Stream pollutant level comparisons,
• Comparison with similar pollutant and proc-
ess operations data reported in the litera-
ture,
• Correlation of process parameters with
pollutant production, and
• Cross-correlations of pollutant data.
Integrated results from the semibatch gasifier
have evidenced good simulation of fixed-bed,
full-scale, continuous units in terms of product
composition, throughputs, and effects of opera-
tional variables. Mass balances have been im-
proved, and consistent chemical analyses of
potential environmental hazards have allowed
evaluation of production trends. Specific com-
pounds consistently contributing to significant
potential environmental hazards have been
identified. Compounds posing threats to health
(on the bases of both quantity and toxici-
ty/carcinogenicity) appear to be limited to a
reasonable number, allowing routine quanti-
tation. Analysis of this limited number of com-
pounds is being augmented by bioassay re-
search to encompass total materials and
synergistic effects.
INTRODUCTION
The gasification reactor is the primary,
unique source of pollutants in this type of coal
conversion plant. A large amount of information
is already available on such subsidiary proc-
esses as coal storage, water processing, utility
•Speaker.
stack gas contamination, and control and treat-
ment.
In attempting an environmental assessment
applicable to various types of coal gasification
reactions, RTI has been operating a small semi-
batch reactor along with extensive sampling
and analysis of reactor streams. This approach
has proved reasonable for fixed-bed gasifier
simulation, where the complications of wall ef-
fects and slugging, present in fluidized beds, are
minimized. We have concentrated on evaluating
various U.S. solid fuels to determine the pollut-
ant loads that control systems must handle.
This screening is substantially complete with
nine different fuels (Pittsburgh #8, Illinois #6.
western Kentucky #9, Montana Rosebud, Wyo-
ming subbituminous, North Dakota lignite.
North Carolina humus peat, a western Ken-
tucky char, and Bottom Red anthracite) charac-
terized in more than 35 tests.
Typically, tests have included air-blown auto-
thermic operation, making a low-Btu producer
gas. Figure 1 illustrates the reactor and sampl-
ing systems. Five main streams are character-
ized: input coal, gas product, tar, aqueous con-
densate, and ash or char. Data acquisition and
analysis have been augmented by a PDP 11/34
with RSX11M operating system. As shown in
Table 1, pollutant data are typically arrayed.1
All data are presented in integrated form for
each stream, although time-dependent data are
recorded where possible to take advantage of
the information available from the batch reac-
tor. Integrating pollutant output allows approx-
imating steady-state continuous gasifier opera-
tion, while the distinct phases of drying, devola-
tilization, and steam/char reaction can also be
defined.
Important process variables such as reaction
temperature, process gas composition and vol-
ume, water gas shift predictability, combustion
characteristics, and fuel heating values compare
with those of other processes, as seen in Table 2.
The consistency of process variable control and
comparisons with other gasifiers have been pre-
113
-------
REACTOR SYSTEM
AND
SAMPLING
STEAM GENERATION
PRODUCT GAS SAMPLING,
ANALYSIS, METBHNG
REACTANT GAS SUPPLY AND CONTROL
A. 1.4 «m
Unimmit
' ra*MM
io-woo i
n-4 ta*KI tmOnT««
KV-I falMlir friniK
KV-t
4OtO-4»CrM
44-IKW
44-1100
44-1100
tm-tt-oo
*-4
n i
>i.«
Fl-l
n •!
n •
Cn»l* ..... lili|l )
I o-mnr
OfMI
4M nto
PtV-IIM«IVMiiM*«Mk.
nip nun mirm.
•«• FMMCI s LMta<« JWa. no-iaox
LUM UNI RlMHrllwMiri «-KX»-OO4
«-i OM i*»m\*i+t*»r*( taoo. o-oo»
*i-t cot inniii matt m*-m mot, o-ioo*
AI a eo «^w IMM MM- maoon «-«»»
. e-»o»
0-f*
TXKHOCOUTUJ
I
2
3
4
a
s
4
0
T I
T*
c
0
Figure 1. Reactor and sampling systems.
-------
REACTOR SYSTEM
AND
SAMPLING
STEAM GENERATION
PRODUCT GAS SAMPLING,
ANALYSIS, METBHNG
REACTANT GAS SUPPLY AND CONTROL
A. 1.4 «m
Unimmit
' ra*MM
io-woo i
n-4 ta*KI tmOnT««
KV-I falMlir friniK
KV-t
4OtO-4»CrM
44-IKW
44-1100
44-1100
tm-tt-oo
*-4
n i
>i.«
Fl-l
n •!
n •
Cn»l* ..... lili|l )
I o-mnr
OfMI
4M nto
PtV-IIM«IVMiiM*«Mk.
nip nun mirm.
•«• FMMCI s LMta<« JWa. no-iaox
LUM UNI RlMHrllwMiri «-KX»-OO4
«-i OM i*»m\*i+t*»r*( taoo. o-oo»
*i-t cot inniii matt m*-m mot, o-ioo*
AI a eo «^w IMM MM- maoon «-«»»
. e-»o»
0-f*
TXKHOCOUTUJ
I
2
3
4
a
s
4
0
T I
T*
c
0
Figure 1. Reactor and sampling systems.
-------
TABLE 1. TYPICAL DATA ARRAY
CONCENTRATION OF POLLUTANT (MICROGfiAMS/CUBIC METER)
COMPOUND MEG NUMBER MATE
BENZALDEHYBE 07A140 5.9E+04
ACETOPHFNONE 07B120 4 . 1E+04
ACETIC ACID 08A040 2.5E+04
BENZENE 15A020 3.0E+03
TOLUENE
ETHYLBENZENE/C2-BENZENE
STYRENE
BIPHENYL
t'IPHENYLMETHANE
C4H7-BENZENE
C4-BENZFNF.
C5-BEN7ENE
INDAN
INPENE
XYLENES
DIETHYLBENZENE
TRIMETHYLBENZENE
METHYL INUENE
C3-BENZENES
DIMETHYLBIPHENYL
PHENOL
CRESOLS
C.2-PHENOLS
XYLENOLS
NAPHTHALENE
ALPHA-METHVLNAPHTHALENE
BETA-METHYLNAPHTHALENE
ACENAPHTHENE
ANTHRACENE
PHENANTHRENE
PROPENYLPHENANTHRENE
C15H12! 3 RINGS
C16II10! 4 RINGS
PYRENE
FLUORENE
FLUORANTHENE
PYRIDINE
BENZOFURAN
METHYLBENZOFURAN
DIMETHYLBENZOFURAN
HI BENZOFURAN
METHYL THIOPHENE
DIMETHYLTHIOPHF.NE
C2-THIOPHENES
C3-7HIOPHENE
C4-THIOPHENES
BENZ07HIOPHENE
NITROGEN (BY DIFFERENCE)
POLLUTANTS AS PERCENTAGE
15A040
15A060
15A080
15A160
15AN01
15AP03
15AP30
15AP31
15B020
15B040
15B080
15B100
15B180
15BP01
15BP21
15BP22
18A020
18A040
18A080
18A140
21A020
21A041
21A042
21A100
21A140
21A180
21AP03
21
21
21B180
22A020
22B040
23A020
24A040
24A140
24AP01
24B020
25A040
25A060
25AP02
25AP20
25APO4
25B040
3.8E+05
4.4E+05
4.2E+05
1.0F.+03
2.2E+05
7.7E+04
.7.7E+04
7.7E+04
2.3E+05
4.5E+04
4.4E+05
2.3E+05
1.2E+05
4.5E+04
2.2E+05
l.OE+03
1.9E+04
2.2E+04
2.5E+04
1.3E+04
5.0E+04
2.3E+05
2.2E+05
1.6E+04
5.6E+04
1.6E+03
2.4E+04
2.4E+04
9.0E+02
2.3E+05
9.0E+02
9.0E+05
1.5E+04
5.3E+06
5.3E+06
5.3E+OA
5.3E+0&
2.2E+04
2.6E+04
2.6E+04
1.3E+03
2.6E+04
2.3E+04
-
6
7.7E+03
5.7E+03
2.2E+04
2.9E+03
4.4E+02
1.1E+04
4.6E+04
1.5E+04
5.4E+03
8.1E+04
2.2E+04
7.1E+04
6.2E+04
2.9E+04
1.2E+05
3.8E+05
A.2E+04
4.4E+02
4.4E+02
7.4E+01
l.OE+04
1.2E+03
2.0E+03
1 .OE+04
4.4E+04
2.3E+03
6.5E+03
3.3E+04
3.7E+03
4.8E+07
16 20
2.0E+00
2.7E+04
1.1E+OA
7.8E+06
8.2E+03
1.4E+03
4.0E+02
3.8E+03
2.0E+03
1.3E+04
9.1E+04
2. OE+04
2.4E+03
9.6E+02
9. OE+04
2.8E+05
2.0E+03
5.2E+03
4.4E+03
5.0E+03
2.8E+04
l.OE+03
5.0E+03
5.2E+02
5.7E+02
7.7E+00
l.OE+02
7.0E+00
7.0E+01
1.3E+05
4.8E+00
7. OE+04
1.7E+03
2.7E+03
3.4E+08 3.6E+OD
21
8.8E+04
1.9E+03
3.1F+02
4.9E+03
1.9E+03
8.8E+04
2.8E+04
6.9E+04
6.9E+04
1.7E+05
1 .6E+03
6.9E+02
2.1E+04
9.8E+03
2.0E+05
7. BE +04
3.9E:^Ofl
23
8.6E+04
2.7E+03
2.7E+03
5.9E+02
6.3E+04
3.9E+04
9. OE+04
5.9E+05
1.2F+03
7.8E+02
5.5E+04
1.2E+04
1.9E+05
9.4R+04
r..lE+08
25
5.8E+04
6.2E+02
3.1E+0?
1.5E+03
3.3E+04
3.1E+04
1.1E+04
2.2E+04
4. OE+04
6.2E+02
1 .OE+04
9.2E+02
2.3Ff03
l.OE+03
6.4E+OB-
TEST NUMBER
26 31 32
1.2E+04
3.9E+04
6.4E+05
3.2E105
l.VE+05 8.0E+02
2. OE+04 3.0E+01
3.9E+03
4.1E+04 2.0E+02
4.0E+05 l.OE+02
1.8E+05
4.3E+05
2.1E+05
7.8E+05
2.AE+05 3.1E+03
4. OE+04
5.9E+04
8.0E+03
7.AE+03 7.0E+00
2.0E+01
3.1E+03
1.3E+05
1.8E+04
7.SE+04
1.4E+04
3. OE+04
1.6E+04
l.OE+08 5.7E+08
4.9E+04
1.2E+03
4.9EIO2
2.0E+03
2.4E+04
6.4E+04
3.5E+04
6.4E+04
1 . 4E 1 Of.
4.0E+02
7.9E+03
1.3E+03
9.3E+03
1.6E+03
•5.1F. + OQ
33
5.1E+04
2-9E+03
1 .3E+03
3.5E+04
2.5E+05
1.1E+05
5.6E+05
l.5t+05
6.5E+05
1 ,2Ef05
1 . 1E+03
8.4E+03
4.AE+04
1.3EI04
5 . 7E+08
35
5.2E404
1 . '.?F. 103
6.OEI02
7.0E-I03
5. AE;i04
4. OE+04
2.4E+05
1.1E+05
2.8E+05
3.4Ef04
2.0E+02
4. 7F+02
8. OF + 0.1
1.7E+O2
4.2F. + 04
2.9E-I03
1.2EH04
9.3E+03
7. 1E+08
OF TOTAL GAS
STREAM ( BULBS f TENAX , XAD2 f SCRUBBER )
MASS FRACTION
MASS FRACTION
MOLE FRACTION
MOLE FRACTION
U/ CO
U/0 CO
W/ CO
W/0 CO
3.4E-01
4.3F.-0?
1 .8E-01
1 .8E-O2
2.1E-01 2. BE- 01
2.5E-02 9.,SE-03
2.0E-O1 3. OF '••01
1.-5E-02 8.4E-03
2.0E-01
1 .v'E-02
2..?f.-01
1 . IE- 02
1.6E -01
1.9E-02
2.5F.-01
2. JET-02
2.6F-01
2.9E-OJS
4.9F-OJ
3. IE- 03
6.6E-01 1.4E-01
1.3E-02 7.8E-03
4.;
-------
TABLE 2. EXPERIMENTAL TEST PARAMETERS AND COMMERCIAL GASIFIER
OPERATING CONDITIONS
RTI TESTS
Air/Coal 8/g
Stean/Coal 8/g
Carbon Conversion X
Coal Residence Time (Min.
Tar Produced 8/g
Gas Produced SCF/lb
IBIV Btu/SCF
Thruput lb/hr ft2
Coal Type
Pressure psia
Mesh Size
Max Temp *C
Heatup Time to 800°C
(Hin)
Gas Composition
CO
co2
ai4
"2'
H2
V
IHIV Btu/SCF
1. Cilmorc, D.W. and A.J
21
1.1
3.1
97
) 340
.035
48
106
16
111.
06
200
8x16
1015
20
16
18
5.4
30
30
0.4
200
23
2.2
1.2
96
300
.033
56
96
19
111.
96
200
8x16
1050
11
10
18
3.1
13
55
0.8
100
25
1.7
0.50
99.7
180
.018
41
142
30
Montana
Sub.
200
8x16
1060
3
24
9.1
2.4
13
52
0.06
140
. Liberatore, "Pressurized
Presented at Texas Symposium
2. Cavanagh, E.G., et.al
32
1.5
0.37
99.5
110
.011
32
183
44
Wyoming
Sub.
200
8x16
1050
5
29
9.1
5.7
20
36
0.07
210
33
1.5
0.36
98.9
110
.012
35
201
45
Wyoming
Sub.
200
8x16
1040
8
32
4.9
5.7
20
37
0.07
210
35
1.7
0.37
97
110
.029
40
128
46
Wyoming
Sub.
200
8x16
910
23
16
12
3.7
14
54
0.08
130
WELLHAN
MERC
2.3
0.31
98.7
120-540
.022
47
153
107
111.
«6
315
LURGI
3.0
1.5
95
60
N/A
52
195
248
Sub. C
N.M.
300
2"xO 1.75"x0.08"
—
—
21.8
6.9
2.0
17.8
51.5
0.2
150
, Stirred, Fixed-Bed Gasification,"
on Environmental Aspects of Fuel Conversion
, Technology Status REnort
Controls, Radian Corporation,
3. Cleland, J.G., et.al,
1977.
"Pollutants from
—
—
17.4
14.8
5.1
23.3
38.5
N/A
200
Morgan town
GALUSHA
3.5
0.4
99+
120-540
0.06
N/A
168
899
Bitun
ATM
2"xl.25"
1300
—
28.6
3.4
2.7
15.0
50.3
N/A
170
WOODALL
DUCKHAH
2.3
0.25
99
N/A
0.075
N/A
175
70
HVCB
ATM
1.5"x
.25"
1200
—
28.
4.
2.
17.
47.
0.
170
3
5
7
0
2
3
Energy Research Center
Technology. EPA-600/2-76-149,
: Low/Medium Btu Coal Gasification
Synthetic Fuels
Production: Faci
and Related
June 1976
Envrlonmcn tal
tity Construction and Preliminary
Tests," EPA-600/7-78-171.
-------
viously discussed.12 3 Confidence in the approxi-
mation of pollutant production from actual gasi-
fiers is supported by comparison with results of
others reported in the literature. Table 3 shows
some chemical compounds analyzed in the RTI
producer gas stream and also in the streams of
other reactors that have been environmentally
evaluated under U.S. Environmental Protection
Agency (EPA) funding.
Tables 4 and 5 examine RTI tar compositions
compared to other coal tars. It should be noted
that typically less than 20 percent of total tar is
quantitated as specific compounds in our analy-
ses. In the case of tar and other streams, a low-
percent quantitation is partially owed to restric-
ting quantitation to compounds that represent
high-priority hazards. In the case of tar, how-
ever, it is also owed to difficulties in specifically
defining the heavier fraction of the tar. Outside
research7 8 indicates that from 10 to 75 percent
of a coal conversion tar fraction may lie in the
boiling point range above 400° C. Currently ap-
plied RTI techniques on gas chromatography/
mass spectrometry restrict elution tempera-
tures to about 260° C, but some compounds
whose boiling points slightly exceed 400° C and
whose vapor pressures are sufficient have been
detected. Methods in high-performance liquid
chromatography and other analytical tech-
niques are being developed to extend RTI's
range of analysis in this area.
It is also notable that RTI has detected few of
the five-ring and above compounds such as ben-
zo(a)pyrene. This could result from analytical
limitations, although these compounds have
been routinely found and quantitated in other
RTI programs, including coal tar analyses. It is
also possible that low throughputs, slow heating
rates, and high fixed-bed length/diameter ratios
promote secondary reactions that reduce the
TABLE 3. REACTOR GAS STREAM
Found in(4)
RTI Detected MERC
Compounds Cond.
Methylthiophenes
C2-thiophenes
C2-benzenes
Benzofuran
Indan
Indene
Phenol
Cresols
Xylenols
Naphthalene
Biphenyl
Diphenylmethane
Dibenzofuran
Anthracene
Phenanthrene
C3-Benzenes
Acenaphthene
X
X
X
X
X
X
X
X
X
X
X
Found in(4)
MERC
Tar
X
X
X
X
X
X
X
X
X
X
X
Found in(5> Found in(5)
C-W C-W
Vent Gas Cond.
X
X X
X X
X
X
X X
X
X
X
Found in(5>
C-W
Tar
X
X
X
X
X
117
-------
TABLE 4. TAR AROMATICS ing PER GRAM TAR)
Aromatic Group
Naphthalenes
Phenanthrenes
Chrysenes
1-2 Benzanthracenes
3-4 Benzphenanthrenes
Pyrenes
5-ring compounds
% of total tar
"Coal Tar"(6)
1.2E5
1.6E5
5.0E4
3.E4
3.1E4
1.3E4
40
RTI Tar
III. #6
6.7E4
2.3E4
8.0E3
2.2E4
9.0E3
2.7E4
17
RTI Tar*
III. #6
1.7E5
5.9E4
2.1E4
5.6E4
2.3E4
6.9E4
40
* Normalized to 40% to account for nonquantitated compounds.
TABLE 5. TAR COMPOSITION (WEIGHT PERCENT)
(9) (10) (4) Chapman(5)
Compound Coal Tar 1 * Coal Tar 2V RTI Tar* MERCV Wllputte
Naphthalene
Phenanthrene
Ruoranthene
Pyrene
Fluorene
Chrysene
Anthracene
Dtoenzofuran
2-Methyl-
naphthalene
Cresote
Acridene
Phenol
Quinoflne
Xytonoto
Indote
Peryfene
Benz(a)anthracene
Benzo(a)pyrene
10.0
5.0
3.3
2.1
2.0
2.0
1.8
—
1.5
0.9
0.6
0.4
—
0.2
0.2
—
—
—
_
4.8
1.8
1.1
—
0.3
0.4
—
—
—
—
—
—
—
—
0.08
0.6
0.2
5.2
2.3
1.3
0.9
0.7
0.8
0.7
0.7
0.9
1.0
0.2
0.3
0.3
0.3
0.01
0.4
—
—
2.1
—
—
—
0.5
—
—
0.4
—
1.8
—
0.04
—
—
—
—
—
0.4
0,2
—
0.1
—
0.2
0.3
0.6
—
—
—
0.09
0.2
1.9
—
—
0.08
—
0.08
. *6 Coal
118
-------
heavier tar fractions.
The occasional pronounced variations (be-
tween processes in the tables) in tar composi-
tion are primarily caused by the differences per-
taining; to what gasifier stream is defined as
"tar." A large amount of contaminated water
may be mixed in streams sampled onsite at pilot
or full-scale units.4 5 Relative levels of the com-
ponents are consistent. The extrapolation of the
RTI aromatic values in Table 4 is reasonable be-
cause the total aromatics determined by frac-
tionation for this test comprised about 55 per-
cent of the total tar. The linear extrapolation of
each group may be somewhat inconsistent, of
course.
Many of the best analyses for such liquids
have been conducted on the synthesis products
from such processes as Synthoil. The differ-
ences between these reaction processes and
gasification make comparison difficult, how-
ever.
Aqueous condensate compositions are com-
pared in Table 6. The characteristics of a proc-
ess, especially steam-to-carbon ratio, will in-
fluence water concentrations. Because the sepa-
ration of tars and aqueous condensate may not
be well defined, major water contaminants are
shown both with and without tar inclusion. The
major tar contribution is increased xylenols con-
centration.
While benzene, toluene, and xylene can be
considered useful byproducts, their potential
toxic and carcinogenic hazards as fugitive emis-
sions require attention. These substances are
measured in the reaction gas stream through-
out testing. Integrated tests results for a max-
imum production case are presented in Table 7.
While the pilot units compared are fluidized bed
types, results are quite similar, possibly reflec-
ting the constancy of lighter devolatilization
products under varying conditions. Production
of BTX is highly dependent upon coal type, with
eastern or other highly volatile coals producing
the highest levels.
Sulfur balances for the reactor have pre-
sented problems. These have ranged from 30 to
140 percent. Closures are typically best for
high-sulfur coals. Some sulfur is lost during
pressure letdown of the condensate trap
stream, but predicted solubility levels for H£
in water do not explain the sulfur losses in-
dicated. Efforts are being made to characterize
more extensively the early rapid devolatiliza-
tion of sulfur species.
An interesting effect in H2S (and COS) output
has been noted in the batch reactor, especially
for the more reactive coals. Production of the
sulfur species often increases near the end of
testing where the combustion reaction begins to
dominate and steam/char reactivity has nearly
ceased. In this phase, concentration curves very
closely parallel those of increasing carbon diox-
ide. The liberation of additional sulfur as S02
(finally reduced in the upper bed) appears to be
the mechanism for the phenomenon. The behav-
ior is unique to sulfur compounds.
Table 8 compares some RTI sulfur distribu-
tions with data from other processes. Ash sulfur
is notably higher in subbituminous and lignite
coals at oxygen breakthrough (approximate to-
TABLE 6. CONDENSATE COMPOSITIONS (ppm)
Compound
Phenol
Cresols
Xylenols
Trimethylphenol
Lurgi (11)
1200-5650
480-1965
100-450
—
PERC(T2>
HI. #6
3400
2840
1090
110
Synthane(13)
1000-4480
530-3580
140-1170
20-150
RTI Water
400-4100
340-1100
65-230
18
RTI (Max)
Water + tar
4600
1400
670
72
119
-------
tal carbon conversion). H2S, COS, and thiophene
levels are typical for coals tested. Mercaptans
appear to be higher for lignite, as found in
larger scale units.18
Attempts to summarize the large amounts of
pollutant data have included formulation of com-
posite values for total reactor-stream hazard
factors per test. This allows both reactor stream
and coal-type comparisons on an environmental
basis. Each chemical substance quantitation in a
reactor stream is expressed as:
• Concentration in terms of jig/m3 gas, pgIL
tar or water, and ng/g ash;
• Potential hazard level expressed as (stream
concentrationMMATE); and
• Micrograms of pollutant produced per
grams of carbon converted in a reaction test.
Here, the MATE value for each substance is a
minimum acute toxicity effluent19 level to which
pollutants should be controlled in the environ-
ment to prevent detrimental health effects.
MATEs are estimates based on available toxici-
ty data and current environmental regulations
and criteria.20 These values, along with exten-
sive chemical information, have been stored in
the RTI synfuels data processing system.
Table 9 outlines the various approaches to
total stream evaluation for each gasification
test. Pollutants expressed as a mass fraction of
the total stream include those substances rou-
tinely quantitated and presenting hazard poten-
tials. Fuel gas products (e.g., CO) are not in-
cluded here as pollutants. Gross variations in
pollutant mass fraction may reflect limited
stream analyses rather than actual stream com-
position.
The stream hazard factor calculation is de-
rived from the EPA SAM/IA21 (source analysis
model) scheme. This summation calculates rela-
tive stream environmental problems but is a
function of the stream flow volume and there-
fore system operating conditions. For example,
air and steam flow rates and percent conversion
levels vary from test to test, changing the nitro-
gen, unconverted water, and tar levels in which
pollutants are diluted.
Calculation of pollutant loading based on
mass production per mass of coal or carbon con-
version eliminates stream volumes from con-
sideration. Summation of pollutant masses
divided by MATE levels results in the amount
of ambient diluent (air, water, soil) required to
reduce environmental pollutant concentrations
to safe levels (assuming dispersion of the entire
stream into the environment). Percent de-
creases in a minimum required diluent are di-
rectly proportional to the efficiency of control
technologies. The term "minimum" is appb'ed
because pollutant quantitations are limited.
An example of coal environmental compari-
TABLE 7. BENZENE, TOLUENE, XYLENES (RECOVERED FROM GAS STREAM)
PROCESS
Hygas Pilot Plant (III. #6)
BTX
(Liquid Liters/Kg Coal)
0.01-0.02
(14)
Synthane
Synthane PDU
RTI (maximum values)
0.006fl5)
Benzene: 10"4-0.01;
Toluene: 0.001 (max)(16>
Benzene: 0.02
Toluene: 0.006
Xylene: 0.003
BTX: 0.03
120
-------
TABLES. SULFUR BALANCES (PERCENT)
Stream
RTI TESTS
OTHER PROCESSES
Ash
Condensate
Tar
Gas
IN. #6
(21)
1.95
4.06
2.10
91.9
Montana
(25)
15.3
7.5
2.1
75.1
Wyoming
(33)
20.7
—
—
33.9
N.D. Lig.
(43)
9.00
—
—
38.8
Lurgi Hygas
5a 0.9C, 1.0d
— —
1 .4b, 2.9a -
— —
Synthane
1.5e
6.9e
1 .4C, 7.8e
84.08
"WESCO ESTIMATES*17'
bN.D. LIGNITE114'
c,L #6(14)
dMONTANA SUB-BITUMINDUS*14)
ePITTSBURGH #8-NORMALIZED TO 100% CHAR CONVERSION116)
-------
TABLE 9. POLLUTANT STREAM SUMMATIONS
POLLUTANT COMPOSITES
PER TEST
n
FRACTION OF _ , ^
TflTAI QTRPAIM ~" -• - - • •
1 \J IML. O 1 nCMIVI
(% mass) STREAM DENSITY
HAZARD FACTOR £ ^
OF STREAM " 7 MATEj
MINIMUM DILUENT , M
FOR STREAM ~ E
(M3 air; liters water; AMC ~ MATE,
grams sofl-per gram
AC)
RANGE OF VALUES-ALL TESTS
GAS
0.2-
4.0
300-
2200
1.4-
12.0
TAR
0.7-
19.0
1 x 105-
2x107
10-
800
WATER
.008-
1.1
7x10*-
1x106
100-
900
ASH
.04-
2.9
10-
120
1.4-
120
to
C = Polutant Concentration
M = Pollutant Mass
MATE = Minimum Acute Toxicity Value
AMC = Mass of Carbon Converted
n = Total Number of Pollutants
-------
sons on this minimum diluent basis is shown in
Figure 2. Ranges of minimum diluent (or pollut-
ant mass production divided by MATE, on the
basis of carbon converted) are given for the
most relevant experimental gasification tests
for seven coals. The required dilution of pollut-
ant streams in the ambient environment has
been normalized on a "mass required" basis for
air, water, and soil. Rough significance can be
attached to these results by considering two hy-
pothetical cases:
• A plant with 10,000 tons per day carbon con-
version and a gas stream minimum diluent
value of 7,500 vents this gas stream to the
atmosphere. If the gas uniformly diffuses
throughout a hemispherical volume, the ra-
dius of diffusion reaches 1.8 mi before 1
day's gas output is sufficiently diluted.
• If the same plant, with a condensate
minimum diluent value of 1.5 x 108, dumps
its raw condensate stream to surrounding
waters, about 1 trillion gal of diluting water
is required per day.
No surprising hazard variations among the
coals are noted on this generalized basis of
Figure 2. High-sulfur coals naturally present
highest gas stream hazards. Tars and conden-
sate ranges are strong functions of the amount
of phenol, cresols, and xylenols produced. In
such composite evaluations, the contribution of
some constituents may be masked by those with
high concentrations or MATEs. This is evident
in Table 10, where it is obvious that phenols
dominate the hazard level evaluation for con-
densed streams, primarily because of their low
MATE values (based on EPA and NAS/NAE
Water Quality Criteria, and Public Health Serv-
ice Drinking Water Regulations).1'
Compounds with associated relative hazard
values of less than 1 have not been quantitated
at stream concentrations that exceed their
MATE values. Relative values of more than 1
are based on a mass of minimum diluent values
for each compound or element, averaged over
all relevant test cases. Within analytical limita-
tions, the table summarizes the RTI evaluation
of potential environmental hazards on a "per
substance" basis.
Tar fractions have also been routinely sep-
arated through a solvent extraction process.
Mean distribution of the polynuclear aromatic,
nonpolar neutral, polar neutral, organic base,
organic acid, and insoluble fractions are shown
in Figure 3. Average tar levels, as a percent of
coal mass input, are included for each coal. Tar
densities typically range from 1.05 to 1.20 glee.
It should be emphasized that operational con-
ditions, such as maximum temperatures or gas
velocities and pressures, influence pollutant
production and must be considered on a test-by-
test basis. Parametric test conditions are being
investigated for possible reactor design controls
of pollution.
Research into reactor parameter effects on
contaminants has been initiated through statis-
tical correlations of coal screening tests results
(the first phase of research) and continued labor-
atory experimentation.
Simple linear statistical correlations were
begun by first dividing the data into four
groups: test variables (e.g., steam/air ratio, max-
imum temperature), coal characteristics (e.g.,
percent volatiles), tar fractions (organic bases,
PNAs, etc.), and pollutant levels by stream. Two
examples of correlating these groups are shown
in Figures 4 and 5. Coefficients of correlation,
precision of fit, and number of correlated values
are evaluated and best correlations plotted.
Some preliminary results include:
• Phenols in condensates inversely correlate
with percent tar and coal rank (see Figure 4).
• The percent of organic bases in tar increases
with higher steam/air ratio.
• Carbonyl sulfide production is higher when
air/coal is high.
• Coals producing more tar also produce more
sulfur gases and increased PNA and organic
base fractions in tar. The gas stream hazard
factor for these coals is higher and the tar
hazard lower (reduced phenols).
• Organic acids, polar neutrals, and nonpolar
neutrals in tar directly correlate. This group
inversely correlates with the related polynu-
clear aromatic and organic base fractions.
• Extending the time period to reach maxi-
mum temperature reduces the tar hazard
factor.
• For high-sulfate sulfur in a coal, more sulfur
remains in ash following reaction.
• Percent of PNAs in tar correlates poorly
with coal rank and only slightly with tar
mass produced.
• Known correlations are reverified; e.g., gas
heating value and coal rank, gas hydrogen
percent and steam/air ratio, and heating
rate and volatile production (see Figure 5).
123
-------
z
1
UJ
••J
0
z
o
o
h-
LU
•mi
5
o
H
0
LU
i
cc
z
§
fiC
^
o
o
LJJ
rf
oc
z
§
cc
o
106
105
104
103
102
10
1
.1
-
V
-
.'1
V
/
CONDENSATE
1^
-
1 1 1
t £ d
^^H ^^^ ••
E ^
105
104
103
102
I ^^
10
1
-
••• 1
-
V
1 1 1
t 5 d
^ ^
I I 1
zoo
0 Z >
2 £
1 1 1
™
GAS
i i i
zoo
0 Z >
5 £
,
r-
<
/
!
z
<
I'll
TAR
EST
i it t i i J
I I ^ O i | 1
E £ s i <
ASH
I
^^
I I
z o
o z
COAL TYPE
Figure 2. Coal-type comparisons.
124
-------
TABLE 10. POLLUTANT RANKING-POTENTIAL HEALTH EFFECTS
TAR
Xytenob (106)*
Cresote HO5)
Phenol (106)
Trimethylphenol (106)
0-feopropylphenol (10s)
Phenanthrene (103)
Chromium <103)
Banziolne (103)
Fkiorane (102)
Perytone (102)
9-meanthracene (102)
Chrysane (102)
Sulfur (102)
Naphthalene (101)
Anthracene (101)
Arsenic (101)
Fluoranthana (101)
Lead (101)
Quinolna (101)
Pyrened)
2-mehaphthatene (1)
Cadmium «1)
Dlbenzofuran «1)
AcrWne«1)
AnMna(<1)
GAS
Carbon monoxide (104)
Benzene (103)
Hydrogen suffide HO3)
Hydrogen (102)
Carbon dioxide (101)
Thiophene (101)
Xytenols (101)
Ammonia (101)
Methanethiol (101)
EthanethioJ (101)
Methane (101)
Phenols (101)
Creaols (101)
Methyttniophene (1)
Naphthalene (1)
Blphenyim
Hydrogen cyanide (1)
Indene (1)
Toluene (1)
C2-thiophenes(1)
Carbon disuffides (<1)
Carbonyl sulfide (<1)
CONDENSATE
Phenol (107)
Cresols (106)
Xytenols (10s)
Ammonia (104)
Thiocyanate (102)
Cyanide (102)
HCN (102)
Sulfur (1011
Chromium (1)
Lead«1)
Arsenic (< 1)
Cadmium (<1)
Phosphorus «1)
Antimony (<1)
Nitrates (<1)
ASH
Arsenic (103)
Sulfur (102)
Lead (101)
Mercury (101)
Cadmium (1)
Chlorida(<1)
•Rahrtiva hazard, equate 1 where stream hazard is zero.
125
-------
3%
jORAcip;
YYYYYYY
ORBASE
mill
PNEU
NPNEU
2.78%
1.82% 1.78%
0.78%
..............
W.Ky.
III #6 Wyoming Montana
Figure 3. Tar partitions.
126
N.D. Lig.
-------
IV
II.'
1 n. y
3TA11S7ICAL ANALYSIS Srb1t«( J/«iS TUESDAY, APHlt 10,
PLOT OF Lt'Ht5«h'LTT»KCL Ltr.tHDf A = 1 1US, b r d Uflb, f. IC,
©
,i 0,b o.1* t.2 1.5 ».8 ?.l ?.« e.7 J,0 J. 1 S.h
PLTIAHCL
Figure 4. Tar/coal (weight percent) vs. phenol condensate concentration (log ;
-------
('L"T IIF nt.'lS, tTC,
00 Hill1
« 0 'I
/•O'.1
©
0
©
10
it 45
«7
bj b7 M
HlPI
87
Rgura 5. DevotatHzation heating rate (°C/min) vs. benzene concentration in gas stream (/*g/m3/[MATEJ).
-------
While few of the results are definitive, cer-
tain trends are indicated. Multiple regres-
sions may improve some results.
As mentioned, a slow heating rate has been
characteristic of the system. Figure 6 illustrates
a typical temperature history. The entire coal
load for each test is added cold to the preheated
reactor. Autothermic heat addition by air com-
bustion is immediately begun. The period fol-
lowing is critical for pollutant production
because the coal temperature now moves slowly
through the range of pyrolysis temperatures
(300°-700° C) within which 90 percent of the
volatiles may appear.
Heating rates through devolatilization aver-
age about 50° C/min, low compared to those ob-
tained in most pyrolysis studies.22 Reactor
throughputs of steam have been typically low,
although recent tests have emphasized an ap-
proach to optimum reaction rates by increasing
air and steam flows.
Low heating rates should decrease both tar
and total volatile production. This does not ob-
viate the important results obtained during
these test periods. Integrated pollutant outputs
have been demonstrated closely comparable to
those of commercial units, and indeed (for fixed-
bed reactors that are fed by lock hoppers) dis-
TIME FACTORS
ALL TESTS
MAXIMUM
TMEAN
PYROLYSIS HEAT RATES
(° C/MIN)
TIME TO REACH TMAXIMUM
(MIN)
TEST TIMES: >90% AC
(MIN)
RANGE
25-89
24-84
40-350
MEAN
51
51
180
20 40 60 80 100 120 140 160
TIME INTO TEST (MINUTES)
Figure 6. Reactor heating.
129
-------
continuous feed and slow heatup may be a more
realistic approximation.
The batch reactor permits investigation of
the time-dependent evolution of pollutants, as
shown in Figures 7 and 8. The typical rate ex-
pression for devolatilization is first order,
dV/dt = k(Vi - V)
where V — Vj as t —
-------
o
1,000
600
100
99
50 100 150 200 250 300
TIME FROM COAL LOAD (MINUTES)
Figure 8. Pollutant kinetics.
tion; i.e., dV/dt vs. (V, - V)/Vj. This approach
and that above indicate that many compounds
tend to devolve in a second-order manner
through the nonisothermal phase of heatup.
Then the kinetics become first order at temper-
ature stability. This approach agrees to some
extent with the results of Wiser, et al.23 Noniso-
thermal pyrolysis, for the case where tempera-
ture is a linear function of time, has been dis-
cussed previously.24 Commonly, more than 75
percent of the pollutant species will have been
produced before 20 percent of the gasification
test has been completed. H2S and COS often
react relatively slowly, while mercaptans are
expended quickly. Carbon conversion past the
heatup phase is essentially zero order, as ex-
pected for reaction under conditions of constant
steam partial pressure. A summary of the se-
quence of gas stream pollutant production of
five coals is given in Table 11. The percent of
conversion may be taken from the figures.
It has been noted that the major reaction se-
quence during devolatilization for this reactor
closely adheres to that commonly promulgated;
131
-------
TABLE 11. REACTIONS SEQUENCE
#43 (N.D. Lignite)
50%
#41 (W. Ky. #9)
50% 90%
Methanethiol
Benzene
Toluene
H2S
Thiophene
COS
Carbon
Methanethiol
Benzene
Toluene
COS
H2S
Thiophene
Carbon
Methanethiol
Thiophene
Benzene
Toluene
COS
H2S
Carbon
123 (111. #6)
#25 (Montana Sub)
Methanethiol
Thiophene
Benzene
Toluene
COS
Carbon
133 (Hyonlng Sub)
50%
Naphthalene
Phenol
Benzene
Thiophene
Indene
Toluene
Cresols
Xylenols
Biphenyl
H2S
COS
Carbon
90% 50%
Thiophene
Phenol
Benzene
Cresols
Xylenols
Naphthalene
Indene
Toluene
Biphenyl
COS
Carbon
Cresols
Naphthalene
titf,± LVI 1 1
Methyl
Thiophene
Indene
Benzene
Toluene
Thiophene
Xylenols
Methanethiol
Phenol
Biphenyl
H S
COS
Carbon
90% 50%
Methanethiol
Benzene
Thiophene
Methyl
Thiophene
Methyl thiophene
Indene
Naphthalene
Cresols
Xylenols
Phenol
Biphenyl
Toluene
COS
Carbon
Methanethiol
H2S, COS
Benzene
Toluene
Indene
Phenol
Cresols
Naphthalene
Biphenyl
Xylenols
Carbon
90%
Methanethiol
Benzene
Toluene
Indene
Biphenyl
Naphthalene
Xylenols
Cresols
Phenol
COS
H2S
Carbon
i.e., the appearance first of water vapor, fol-
lowed by C02, CO, tars, ethane, methane, and
hydrogen.25
The possibilities of reaction rate control other
than by chemical kinetics have been investi-
gated. It appears from available criteria26 that
internal particle heat transfer is not limiting.
However, external heat transfer (gas diffusion
to particle surface) and mass transfer do seem
to play a role. The common particle sizes util-
ized vary from about 32 to 4 mesh, and there is
evidence of particle size influence on reaction
rate. The primary control of rate at this point is,
as stated, the time required to satisfy the heat
capacity of the entire coal charge to bring the
coal to reaction temperature (basically through
convective heat transfer).
Briefly, some conclusions and recommenda-
tions include:
• The gas stream essentially contains pollut-
132
-------
ants that are well recognized. Major pollut-
ant factors in the gas stream result from car-
bon monoxide, benzene, hydrogen sulfide,
and other sulfur species. Removal of
benzene, xylenols, ammonia, and sulfur spe-
cies should prevent problems. The removal
of pollutants to the ranking level of methane
(see Table 10) could, on a toxic hazard basis,
leave the producer with only the same con-
trol requirements placed on natural gas
pipelines.
While there is little positive evidence of im-
portant levels of heavy hydrocarbons found
in the gas stream, certain trace constituents
may deserve increased attention. This
would require examination beyond acute ef-
fects, where synergistics and persistence in
the environment are considered in terms of
chronic health effects. A definite need exists
for better characterization of aerosols and
gas-stream-suspended particulates. Never-
theless, it can be stated with assurance that
if ambient carbon monoxide, hydrogen sul-
fide, and benzene concentrations are moni-
tored at a coal gasification plant, the pri-
mary fugitive emissions from the product
gas stream will have been accounted for.
Tar presents a more consistent environmen-
tal hazard. While phenol and cresols domi-
nate the hazard picture, elimination of these
reduces the hazard factor to only 103. The
presence of fused aromatic hydrocarbons
such as phenanthrene, chrysene, and 9-meth-
ylanthracene disallows obvious elimination
of the hazard problem. Preliminary bioassay
work has also shown that RTI gasifier tars
are mutagenic.27 More extensive study of
the heavier tar fraction is needed. It is in-
teresting to note that western coal tars pre-
sent as high a hazard as eastern coals be-
cause of the comparable levels of organic
acids (e.g., phenols) produced in gasifying
these coals. More research is required on the
storage, handling, and utilization of the coal
tars. It is hi these areas of plant operation
that tars become an environmental hazard.
It has been noted that the overall pollutant
potential for various coals varies little in
terms of orders of magnitude. While varia-
tions may occasionally be important for con-
trol technology development (e.g., different
sulfur loads on such removal systems as Rec-
tisol, Stretford, and Glaus units), the results
seem to indicate that a uniform approach to
reactor parameter control can be taken for
various coals. As mentioned, more work is
intended in comparing pollutant results ob-
tained thus far on the basis of test conditions
such as average temperatures, combustion
vs. nonautothermal testing, and steam/air
and air/coal ratios. Most variations in the
results are predictable and related to the
major reactor parameters or coal character-
istics such as reaction temperature, volatiles
content, and sulfur content.
• The aqueous condensate is contaminated
primarily by phenols, as is well known. If
phenols and the other important byproduct
in solution—ammonia—are removed, the
condensate hazard factor is reduced to ap-
proximately 10. Further analysis of these
species and cyanides is required to validate
the conclusion.
• The small semibatch reactor works well in
terms of simulating process variables and
outputs and pollutant amounts. Ash, water,
and carbon balances are satisfactory for this
process, but improvements are required on
sulfur balances and the nonair nitrogen bal-
ance. While slow heating rates may reduce
tars and volatiles, the distribution of pol-
lutants throughout the various streams is
quite comparable to those reported in the
literature and may simply point to one ap-
proach for controlling contaminants in full-
scale reactors. The pyrolysis phase is by far
the most important one for pollutant produc-
tion and should be studied more extensively,
including research on nonisothermal kinet-
ics. Most pyrolysis research has been lim-
ited to studying total volatile production
rather than examining individual pollutant
species.
• Important pollutants requiring extensive
examination can easily be limited to a num-
ber that can be quantitated on a per-test
basis. While more than 420 compounds have
been detected and more than 100 routinely
quantitated, there is good evidence that con-
trol of a few priority pollutants beginning at
the top of the pollutant ranking list in Table
10 should ensure environmentally safe coal
gasification. The concentrations or micro-
grams produced per grams of carbon con-
verted for various species from test to test
are most notable for their consistency. Pol-
133
-------
lutants routinely detected have been found
in every type of coal analyzed, from peat to
anthracite. The North Dakota lignite has
shown some peculiar characteristics that
have not yet been fully examined. These
characteristics include the detection of both
unique organic and mineral species in all
streams. It is believed that baseline levels
for pollutants from the U.S. coals tested are
now better defined, allowing more con-
fidence in studies dealing with variations in
these levels.
Many of the results obtained thus far cannot
be presented briefly or generally. Some must be
evaluated on a test-by-test basis. The experi-
mental model is being improved for evaluating
the effects of reactor parameters on pollutant
production or prevention. Such variables as
pressure, coal mesh size, bed depth, tempera-
ture, heating rate, steam/air ratios, and rate of
gas removal from the bed are being considered.
If modeling problems can be overcome, fluidized
bed operation will also be investigated. Fabrica-
tion of a continuous coal feeder and an improved
pretreatment setup for eastern coals is nearly
completed and will be included in further ex-
perimentation. More extensive evaluation of
trace elements as pollutants utilizing neutron
activation analysis is also intended.
ACKNOWLEDGMENTS
The Research Triangle Institute, extends its
greatest appreciation to the Fuel Process
Branch, Industrial Environmental Research
Laboratory, U.S. Environmental Protection
Agency, for supporting this research.
REFERENCES
1. Pottutants from Synthetic Fuels Produc-
tion (annual report). Research Triangle In-
stitute. Research Triangle Park, N.C. EPA
Grant Number R804979-02. February 1979.
2. Pollutants from Synthetic Fuels Produc-
tion (quarterly report for the period May 1,
1978-July 31, 1978). Research Triangle In-
stitute. Research Triangle Park, N.C. EPA
Grant Number R804979-02. August 1978.
3. Cleland, J. G., F. 0. Mixon, and D. G.
Nichols. Small Semi-Batch Oasifier Opera-
tion for Pollutant Determination and Con-
trol Evaluation. (Presented at National
AIGhe Meeting. Philadelphia. June 4-8,
1978.)
4. Pellizzari, E. D. Identification of Com-
ponents of Energy-Related Wastes and Ef-
fluents (final report). Research Triangle In-
stitute. Research Triangle Park, N.C. EPA
Contract Number 68-03-2368.1978.
5. Personal Communication with Dr. Gordon
Page, Radian Corp., Austin, Texas. 1978.
6. Project H-Coal Report on Process Develop-
ment (final report). Hydrocarbon Research,
Inc. PB-234 579, OCR/RDR 26/. 1978.
7. Personal Communication with W. J.
McMichael, Pittsburgh Energy Research
Center, on GLC Simulated Distillation for
Synthane Tars Based on ASTM D288773
Method. 1978.
8. King, B., et al. An Investigation of
Aromatic Fractions from Coal Tar Pro-
duced by an Underground Coal Gasifica-
tion Task. Laramie Energy Research Cen-
ter. CONS-77-0510-7.1977.
9. Shults, W. D. (ed.). Preliminary Results:
Chemical and Biological Examination of
Coal-Derived Materials. ORNL/NSF/EATC-
18. Oak Ridge National Laboratory. Oak
Ridge, Tenn. 1976.
10. Lijinsky, W., G. Singer, and W. Taylor. Car-
cinogenesis by Environmental Pollutants
Related to Energy Production. In: Bio-
medical and Environmental Research for
Support and Evaluation of Coal Conver-
sion Technology. Oak Ridge, Oak Ridge Na-
tional Laboratory, 1975.
11. Jolley, R. L., W. W. Pitt, and J; E. Thomp-
son. Organics in Aqueous Process Streams
of a Coal Conversion Bench-Scale Unit
Using the Hydrocarbonization Process:
HPLCand OC/MS Analysis. (Presented at
the Environmental Technology Annual
Technical Meeting of the Institute of Envi-
ronmental Sciences. Los Angeles. 1977.)
12. Sharkey, A. G., Jr., J. L. Shultz, C. E.
Schmidt, and R. A. Friedel. Mass Spectro-
metric Analysis of Coal-Derived Liquids.
In: Proceedings of the Third Environmen-
tal Protection Conference. Chicago, Sep-
tember 23-26,1975.
13. Forney, A. J., W. P. Haynes, S. J. Gasior, G.
E. Johnson, and J. P. Strakey. Analysis of
Tars, Chars, Oases, and Water in Effluents
from the Synthane Process. U.S. Pitts-
134
-------
burgh Energy Research Center. Pitts-
burgh, Pa. Bureau of Mines Technical
Progress Report 76.1974. 21.
14. Glaser, F., A. Hershaft, and R. Shaw. Emis-
sions from Processes Producing Clean
Fuels. Booz-Allen Applied Research. Beth-
esda, Md. BA 9075015.1974.
15. Kalfadelis, C. D., and E. M. Magee. Evalua-
tion of Pollution Control in Fossil Fuel Con-
version Processes. Gasification. Sec. 1. Syn- 22.
thane Process. EPA-650/2-74-009b. 1974.
16. Forney, A. J., W. P. Haynes, S. J. Gasior, R.
M. Kornosky, C. E. Schmidt, and A. G. 23.
Sharkey. Trace Element and Major Compo-
nent Balances Around the Synthane PDU
Gasifier. Pittsburgh Energy Research
Center. Pittsburgh, Pa. PERC/TPR-75/1. 24.
1975.
17. Beychok, M. R. Process and Environmental
Technology for Producing SNG and Liquid
Fuels. EPA-660/2-75-011. 1975.
18. Personal Communication on Lurgi Gasifica- 25.
tion with Gordon Page and Karl Bom-
baugh, Radian Corp., Austin, Texas. 1979.
19. Cleland, J. G., and G. L. Kingsbury.
Multimedia Environmental Goals for En- 26.
vironmental Assessment Research Tri-
angle Institute. Research Triangle Park,
N.C. EPA Contract 68-02-2612, EPA-600/7-
77-136a. November 1977.
20. Greenwood, D. R., G. L. Kingsbury, and J. 27.
G. Cleland. Draft-A Handbook of Key
Federal Regulations and Criteria for Mul-
timedia Environmental Control Research
Triangle Institute. Research Triangle Park,
N.C. EPA Contract Number 68-02-2612. Jan-
uary 1979.
.Shalit, L. M., and K. J. Wolfe. SAM/IA: A
Rapid Screening Method for Environmen-
tal Assessment of Fossil Energy Process
Effluents. Acurex Corp./Aerotherm Divi-
sion. Mountain View, Calif. EPA Contract
Number 68-02-2160, EPA-600/7-78-015. Feb-
ruary 1978.
Anthony, D. B., and J. B. Howard. Coal
Devolatilization and Hydrogasification.
AIChE Journal 22(4):625-656. July 1976.
Wiser, W. H., G. R. Hill, and N. J. Ker-
tamus. Kinetic Study of the Pyrolysis of a
High-Volatile Bituminous Coal. Ind Eng
Chem Process Design Develop. 6:183.1967.
Pollutants from Synthetic Fuels Produc-
tion. Research Triangle Institute. Research
Triangle Park, N.C. (Research Grant Ap-
plication Submitted to Environmental Pro-
tection Agency.) July 1976.
Howard, H. C. Pyrolytic Reactions of Coal.
In: Chemistry of Coal Utilization, Sup-
plementary Volume, Lowry, H. H. (ed.).
New York, Wiley, 1963. p. 340-394.
Koch, V., H. Juntgen, and W. Peters. Nicht-
Isotherme Reaktionskinetik der Kohlen-
pyrolyse, Teil III: Zum Readtionsablanf bei
hoben Aufheizgeschwindingkeiten, Brenn-
stoff. Chem. 50:369.1969.
Hughes, T. J., et al. Mutagenicity of Coal
Gasification Fractions (Presented at Tenth
Annual Meeting of the Environmental Mu-
tagen Society. New Orleans. 1979.)
135
-------
ENVIRONMENTAL AND ENGINEERING EVALUATION OF THE KOSOVO
COAL GASIFICATION PL ANT-YUGOSLAV! A (PHASE I)
Becir Salja and Mira Mitrovic
Kombinat Kosovo, Obilic-Pristina, Yugoslavia
and
Dragan Petkovic*
Rudarski Institute, Beograd-Zemun, Yugoslavia
INTRODUCTION
At the symposium on Environmental Aspects
of Fuel Conversion Technology, III, held in
September 1977 in Hollywood, Florida, we fully
accounted for our investigations of this project.
In addition, we stated why this research project
was conducted in 8FR Yugoslavia and identified
local institutions engaged in the investigations.
Preliminary data from pilot operations indi-
cate that a multiplicity of pollutants are emitted
by the gasification reactor. Material in effluent
and process streams includes major pollutants
such as sulfur, nitrogen, NH8, particulate tars
and oils, and minor pollutants such as trace
elements and hydrocarbons. The purpose of in-
vestigating these pollutants was:
• to identify emissions and determine their
concentrations in the existing gasification
process;
• To determine the composition and amounts
of pollutants originating to a greater or
lesser extent from all process streams in
various stream effluents or materials; (i.e., in
air, water, and particulates);
• To identify the pollutants whose presence
degrades the environment;
• To assess the capability of existing cleaning
and purification systems; and
• To develop improved equipment and
technology designed to reduce or eliminate
environmental danger accompanying the op-
eration of current technology.
In our investigations, priority was given to
examining pollutants occurring in large
amounts such as sulfur and nitrogen com-
pounds, ammonium, coal, and tar and oil par-
ticulates. The investigations included pollutants
•Speaker.
occurring in lower or trace amounts in the proc-
ess and served to evaluate selected methods of
sampling and sample analysis.
Investigations were carried out in Obilic, near
Pristina, Socialist Autonomous Province Koso-
vo, in a plant for the production of gas under
pressure (clean gas net heating value 3,600
kcal/m*) from dried Kosovo lignite (Lurgi gener-
ators, Dia 3,6 m). Plant capacity is 480 million
m* of clean gas per year.
Prior to sampling and analysis, the following
tasks were completed:
• Detailed description of the lignite gasifica-
tion plant was given, including required
process flow sheets and description of indivi-
dual operating stages of gas and byproducts
production, various waste materials, and
medium effluents;
• Presentation of fuel grade at the inlet and
outlet, byproducts composition, and types of
media;
• Detailed study of plant operating data; i.e.,
of the technological process and location of
measurement points for pollutant sampling;
• Selection of the methodology for sampling
and analysis of solid, liquid, and gaseous
pollutants;
• Selection of the methodology for flow and
particulates measurement; and
• Sampling test operating plan.
The obtained results are given below.
RESULTS
General
Investigations were completed during the
course of three sampling campaigns carried out
in periods with normal operating conditions. At
each point of emission of solid, liquid, or gaseous
137
-------
media in the lignite gasification plant at Kosovo,
preliminary quality investigations were per-
formed. The amounts of the emissions were
measured, estimated, or taken as designed by
the project in order to evaluate the types and
volumes of pollutants. This led to eliminating a
number of measurement points because of their
emissions volume and quality.
ASTM methods were used for sampling, anal-
ysis, preparation of measurement lines, volume
measurements, etc. The content of fixed gases
in gaseous streams was determined by an "Or-
sat" apparatus or gas chromatographic method.
Chemical methods were used for H2S, NHs,
phenol, and HCN content determinations. Gas
chromatography was used for determining the
components occurring in lower amounts.
Generators
For investigations in the generator section,
the following measurement points were se-
lected:
• 2.0 Inlet dried lignite,
• 2.2 Dedusting cyclone discharge into the
atmosphere,
• 3.1 Decompression of coal lock bucket,
• 3.2 Generator vent,
• 3.4 Vent of the collecting tank for tar con-
densate and other contaminated
waters in generator section,
• 3.5 Vent from ash lock expander cyclone,
• 3.6 Coal lock expansion gases,
• 12.2 Slag (dry), and
• 12.3 Water.
Tables 1 through 10 present the data on
amounts and quality of the most important
generator section streams.
Rectisols
For investigations of gas streams in the Rec-
tisol section, the following measurement points
were selected:
• 7.3 Raw gas: feed for Rectisol section (it
contains H2S, cyanides, higher hydro-
carbons, etc.);
• 7.2 Waste gas (C02) (it contains in addition
to C02, methanol, H2S, and higher hy-
drocarbons);
• 7.1 H2 waste gas (it contains H2S, methane,
and other hydrocarbons); and
• 7.4 Clean gas: final product.
Tables 11 through 16 present the data on
amounts and quality of the most important Rec-
tisol section streams.
Tar Separation
For investigations of gas streams in the tar
separation section, the following measurement
points were selected:
13.1 Tar tanks,
13.2 Unclean tar tank,
13.3 Medium oil tank,
13.4 Uncleaned oil tank,
13.5 Gas condenser tank,
13.6 Expansion gases to waste gases flare,
and
• 13.7 Phenolic water tanks.
Tables 17 through 23 present the data on
amounts and quality of the most important tar
separation section gas streams.
Phenosolvan and Expansion
Gases Large Flare
For investigating streams in the phenosolvan
section and expansion gases large flare, the
following measurement points were selected:
14.1 Cyclone vent;
14.2 Tank for gas water, tar, oil, and phen-
olic water separation;
14.3 Unclean oil tank;
14.4 Phenolic water tank;
14.5 Column vent;
14.6 Vent between pos. 25 and 16;
14.7 Column vent;
14.8 Phenosolvan section wastewater tank;
14.9 Crude phenol tank vent;
14.10 Diisopropylether tank;
20.1 Gases to large flare; and
14.11 Wastewaters to biological treatment.
Tables 24 through 28 present the data on
amounts and quality of the most important gas
streams from the phenosolvan section and from
expansion gases large flare.
Storage
For investigating gas streams in the storage
section, the following measurement points were
considered:
• 15.1 Tar tank vent,
• 15.2 Medium oil tank vent,
• 15.3 Gasoline tank vent, and
138
-------
• 15.4 Phenol tank vent.
Tables 29, 30, and 31 present the data on
amounts and quality of the most important stor-
age section gas emission (M.P. 15.3) into the
atmosphere. Data on amounts and quality of
gasoline, medium oil, and tar are included in
Table 32.
SUMMARY AND RECOMMENDATIONS
FOR CONTINUOUS ACTIVITY
Results presented in this report were ob-
tained by testing the lignite gasification plant
according to the Lurgi process. Prior to gasifica-
tion, the run-of-mine Kosovo lignite was
screened, classified, and dried by the "Fleiss-
ner" process. Investigations were completed
through three sampling campaigns (i.e., during
winter, summer, and autumn). Their objective
was to evaluate the effect of the Lurgi process
on environmental pollution. Composition and
volume were determined for all major process
waste streams. The volumes were calculated for
pollutants that are discharged at a high rate in-
to the plant, ambient, and broader surroundings
in the form of waste gases, particulates, and
wastewaters. ASTM methods were mainly used
in the investigations.
Accurate sampling was difficult at some
points (discontinuous, short-lasting discharges
in some cases with high contents of water
vapor, more than 90 percent). Despite modern
equipment mainly provided by the U.S. Envi-
ronmental Protection Agency (EPA) and Radian
Corporation, the capabilities of the instruments,
both for sampling and analysis, made the meas-
urements time-consuming because of the num-
ber of components to be determined in a plant of
this size. It was not possible to maintain con-
stant operating conditions during a campaign (7
to 14 days).
Although the differences in results obtained
chemically and chromatographically were ob-
served and partially explained, it was con-
sidered necessary to repeat the check of various
determination methods in heavily polluted
streams.
Velocity measurements for flow determina-
tions during short explosive discharges proved
insufficient, and some flows were determined
calculatively.
The EPA Method 5 for particulates deter-
mination proved inadequate for determinations
in streams with high-water vapor contents
(more than 80 percent), so the method with "wet
impingers" was used.
During our recordings, the percent of con-
veyed heat in the form of clean gas was 62.88
percent of the heat fed in the form of coal. The
balance of carbon conveyed in the form of clean
gas was 24.98 percent. The major amount of sul-
fur (about 91.1 percent) was combusted with the
waste gases in the large flare. Approximately
4.13 percent of the sulfur remained in the liquid
products.
The results of completed investigations in-
dicate that during the production of gas by the
Lurgi process, the following amounts of pollut-
ants were emitted from 10 tons of dried lignite
and by complete incineration of waste gases
through the flare:
• Sulfur "S" 4.0kg
(about 90 percent as H2S)
Ammonium (NH3) 0.4 kg
Phenols 4.7 kg
Cyanides 0.06 kg
Hydrocarbons (CnHm) 12.6 kg
Hydrogen (H2) 2.6 kg
Carbon monoxide (CO) 4.5 kg
Carbon dioxide (C02) 9,632.7kg
Methane (CH4) 10.0kg
Nitrogen oxide (N02) 5.2 kg
Sulfur dioxide (S02) 180.0kg
Particulates 148.5 kg
Total 10,005.3kg
With regard to the Lurgi flow sheet, par-
ticular care should be paid to generator opera-
tion. Experimental operation on a single genera-
tor with systematic variation of technological
operating parameters could indicate the condi-
tions affecting the variation of pollutants con-
centration and amount (pressure, steam-to-
oxygen ratio, coal bed thickness, throughput
capacity, charge size distribution, pour, etc.).
Such investigations could lead to maximum gas
production and gas quality in line with the
reduction of pollutants volume production to a
minimum. For the Lurgi plant in Kosovo, the
presence of a large number of "minor" vents
significantly increasing pollution is charac-
teristic.
In Combine Kosovo, efforts are made to
charge the optimum size distribution of dried
lignite into the generators in order to achieve
more efficient heat balance and higher produc-
139
-------
tion of raw gas per generator in operation. A
project was developed and is currently under-
way to combust the expansion gases in one of
the boilers of the adjacent power generating
plants. Hitherto, the gases were incinerated by
the large flare, and combustion was unneces-
sary. Completion of this project will increase
the gasification heat balance and reduce envi-
ronmental pollution.
It is also necessary to investigate the effect of
the stockpile of this type of coal on the environ-
ment; i.e., to examine the properties of coal dust
particulates, the properties of groundwaters
around the coal stockpiles, and the dissolution of
various mineral matters and chemical com-
pounds from the coal in atmospheric precipita-
tions.
The lignite drying plant was not investigated.
It is considered important to investigate the ef-
fect of the "Fleissner" process; i.e., coal-drying
process on the living environment. Of impor-
tance here are: composition and volume of
waste gases, wastewaters properties, and prop-
erties and composition of dried coal dust par-
ticulates.
The impact of trace elements in generator
slag on the environment and humans should also
be investigated. It is important to determine
the degree of mineral matters and individual
element oxides dissolved in the water as well as
the increase of their concentration in ground-
water, and to examine the effect of the increase
of concentration of different chemical element
oxides in the water on agricultural products and
other fauna and flora. In any case, the effect of
slag dumps should be investigated, both those
on the surface and underground.
Further investigations concentrate on the
pollutants occurring in small amounts.
140
-------
TABLE 1. VOLUMES OF GENERATOR SECTION EMISSIONS
Measurement
point
A
Measured
l
m o u
from
Z(Te
n t
aesi
t/h)
s
gn
Cal
Est
^(
c u 1 a t ed
i ma ted
10,3 t/
h)
2.0. Inlet dried lignite
2.2. Dedusting cyclone
discharge Into the
atmosphere
3.1. Decompression of coal
lock bucket
3.2. Generator vent
circa 16 t/h 10,3 t/h
5400 mjj/h 3476 mjj/h
circa 9 mjj/h 5,8 mj/h
circa 36 mjj/h 23,2 mj^/h
3.4. Vent of the collecting
tank for tar gaseous
water and other con-
taminated waters in
General Section
39 m^/h 25,1 mjj/h
3.5. Vent from ash lock
expander cyclone
9 m3/h 5,8 mjj/h
3.6. Coal lock expansion
gases
326 mj*/h 209,9 mj/h
12.2 Slag (dry)
12.3 Wastewater
2,6 t/h 1,7 t/h
1,56 n»3/h 1,0 m3/h
141
-------
TABLE 2. PROPERTIES OF DRIED LIGNITE KOSOVO, GRAIN SIZE: - 60 + 6 mm
(COMPOSITE SAMPLES OF LIGNITE) M.P. 2.0
Proximate and ultimate analysis:
Sulfur forms and ash chemical composition:
Sulfur forms
Moisture X
Ash %
Sulphur total . X
S bound X
S combust, X
Coke X
C fix X
Volatile* X
Combustibles X
Heating value
Gross keal/kg
Net kcal/kg
Carbon dioxide
(C02) X
Carbon X
Hydrogen X
Nitrogen X
Oxygen X
Bulk weight, t/m3
Mlcum test
(+6 mm) X
Tar X
20,72
10,33
1.06
0,90
0.16
40,18
29,85
39,10
68,95
4335
4035
1,44
46,30
3.79
1,13
17,57
0,50
78
3.3
24,30
17.74
1,15
1,01
0,14
40.96
23.22
34.74
57,96
3470
3190
3.32
37,80
2,96
1,03
16,03
0,55
76
2,1
Sulfur total, %
Sulfur bound. %
Sulfur combust, %
Sulfur pyritic, %
Sulfur sulphate, %
Sulfur organic, %
Moisture %
Ash chemical
composition
sio2x
Fe2°3
A1203
CaO
MgO
so3
P2°5
T102
Na20
K20
Ratio: add/Base
A1;,0,+S102+T10,
/ t J *• t
6 1,34
0 1,13
6 0.21
1 0,90
5 0,06
0 0,38
2
15,21
6,78
4,74
35,55
11.35
23,30
0,30
0,50
1 ,58
0,46
. ._. ) • 0.367
1,15
1 ,01
0,14
0.78
0,08
0,28
24,30
27,08
7.18
7,27
36,05
5,49
14,55
0,22
0,70
0,91
0,46
0.699
1,52
1,34
0,18
1 ,04
0,11
0,37
.
2°3 '
Ash fusibility:
(Oxldatlve atmosphere)
Initiation of sintering °C 925 1070
Softening temperature °C 1200 1250
Hemisphere temperature °C 1325 1275
Flow temperature °C 1335 1285
142
-------
TABLE 3. PARTICLE SIZE OP DRIED LIGNITE KOSOVO, GRAIN SIZE -60 + 6 mm
(COMPOSITE SAMPLES OF LIGNITE)
Particle size 1n mm
+-•
-
-
-
-
-
-
-
-
-
-
mm
-
60
60
50
40
30
25
20
15
10
6
3
2
1
+ 50
+ 40
+ 30
+ 25
+ 20
+ 15
+ 10
t 5
+ 3
+ 2
+ 1
+ 0
% share 1%
15,79
15,79
7,37
16,84
6,32
7,37
8,42
7,37
2,10
7,37
1,58
1,05
2,64
15,79
31,58
38,95
55,79
62,11
69,48
77,90
85,27
87,37
94,74
96,32
97,37
100,00
% share
3,33
4,45
7,78
24,44
6,67
13,33
23,33
12,22
1.11
1.11
0,56
0,56
1.11
£%
3,33
7,78
15,56
40,00
46,67
60,00
83,33
95,55
96,66
97,77
98,33
98,89
100,00
143
-------
TABLE 4. RESULTS OF THE WASTE GAS ANALYSIS FROM M.P. 2.2.
Gas Stream
M.P. 2.2
Section Generators
Campaign
a)
b)
c)
d)
e)
f)
Gas Composition vol-X
"Orsat" and G.C." "Orsat"
- H2 0,8*
- CnHm 0,0
- 02 20,4
- N2 78.7
- CH4 0,0
- CO 0,0
- co2 0,1
Chem.meth. g/100 mjj (dry)
- H2S n.f
- Phenols
- HCN
G.C. meth. g/100 m^ (dry)
- H2S
-NOX
- COS
- so2
- methyl mercaptan
- ethyl mercaptan
- unknown pp»
Moisture X
Participates g/100 mj] (dry)
(method 5)
Dissolved sol Ids
Tar Components
Total "e"
Flow: m2/h (dry)/Gener. 1n Operation
- designed
- calculated 5400
- measured
G.C. G.C
n.f* n.f*
trace n.f
19.81 20.8
78.99 7B.2
trace n.f
trace n.f
n.f n.f
n.f n.f
0,8-2.3 n.f
0.059-0,149
n.f
n.f
n.f
n.f
n.f.
2,47; 1,02
44 3,89
44 3.89
5768
Hot
»: For a) n.f • not found • <
For b) n.f « < 5 pprav
For c) n.f . <0,1 ppmv
* For other hydrocarbons n.f • < 0,0001 vol-X
0,01 vol.-X; trace • < 0,1 vol.-X
144
-------
TABLE 5. RESULTS OF THE WASTE GAS ANALYSIS FROM M.P. 3.2
Gas Stream
M.P. 3.2
Section Generators
Campaign
a) Gas Composition vol-X
("Orsat" and G.C) "Orsat"
- H2 39,2
- CnHm 0,6*
- o2 o.o
- N2 5,0
- CH4 11,3
- CO 7,7
- C02 36,2
b) Chem.meth. g/100 mjj (dry)
- H2S 1371
- NH3 132
- Phenols 2,27
- HCN
c) G.C. math, g/100 mj| (dry)
- H2S
-NO,
- COS
-so2
- methyl mercaptan
- ethyl mercaptan
- unknown ppm
d) Moisture X
e) Partlculates g/100 mj| (dry)
(method 5)
Dissolved solids
Tar Components
Total "e"
f) Flow: mjj/h (dry)/Gen.1n Operation
- designed
- calculated 36
- measured
G.C. G.C.
44,28 34,1
1,32* 1,18*
1,52 0,70
2.17 2 , 54
9.31 9.38
11,23 9.26
28,49 42.0
945 576
21,9 529
0,574 909
5,77
28,8 98.2
0.44
S4.9
23.4
14,1
44,2; 8,4; 16,2
^ 1,37** 2,37**
259,8 145,3**
913.0 1119,0**
1174,17**1266.67**
Hot*: * Other hydrocarbons
** Wet 1mp1nger
145
-------
TABLE 6. RESULTS OF THE WASTE GAS ANALYSIS FROM M.P. 3.5
Gas Stream
M.P. 3.5.
Section Generators
Campaign
• )
b)
c)
d)
e)
f)
Gas Composition vol-x
(Orsat and G-C) "Orsat"
- H2 0.0
- cnHm 0.0
- o2 o.o
- N2 78.5
- CH4 0.0
- CO 0,0
- C02 21.5
Chem.meth. g/100 mjj (dry)
- H2S 0,0
- NH3 0,57
- Phenols O.OZO
- HCN
G. C. meth. g/100 mj] (dry)
- H2S
- NOX
- COS
- S02
- methyl mercaptan
- ethyl mercaptan
- unknown ppm
Moisture X
Partlculates 9/100 mj| (dry)
(method 5)
Dissolved solids
Tar Components
Total "e"
Flow: mjj/h (dry)/Gen.1n Operation
- designed
- calculated 9
- measured
G.C. G.C.
n.f. n.f
trace trace
0.0 73,43
84,46 0,00
0,27 0.04
n.f. n.f
18,17 26.53
10,0 12,5
22,6 261
4,62 0,217
6,5
n.f
n.f.
n.f.
n.f.
96.6; 90.6
210 130,9** 1,05**
97.3** 348.8**
89.3** 368,7**
210 317,5** 718,55**
H o t »: * Other hydrocarbons; n.f. » < 0,0001 vol-X
** wet Implngers
For a) n.f. • < 0,01 vol.-*; trace • < 0,1 vol.-*
For c) n.f. » < 0,1 ppmv
146
-------
TABLE 7. RESULTS OF THE WASTE GAS ANALYSIS FROM M.P. 3.6
Gas Stream
M.P. 3.6
Section Generators
Campaign 1
•)
b)
c)
<1)
• )
f)
H
Gas Composition vol.-* "Orsat"
("Orsat" and G.C)
- H2 37. 0-37, a
- CnHm 0.9-1.2
- 02 0,3-0,2
- N2 4,8-11,7
- CH4 9.6-2,9
- CO 8,0-12,0
- C02 39.4-34,8
Chen. meth. g/100 wj* (dry)
- H2S 421-363
- NH3 26
- Phenols 0.027
- HCN
G.C. meth. g/100 m* (dry)
- H2S
- NOX
- COS
- so2
- methyl mercaptan
- ethyl mercaptan
- unknown pp*
Moisture X
Partlculates g/100 mjj (dry)
(Method 5)
Dissolved solids
Tar Components
Total "e"
Flow: rajj/h (dry)/ Gen.ln Operation
- designed
- calculated 326
- measured
o t •: * Other hydrocarbons
** w»t Imnlnnert
2
G.C.
23,40
1 ,16*
1,20
7,30
9,95
13,20
36,80
235
5,8
0,465
71,3
trace
trace
23,1
161,9
18b
3
G.C.
31,6
0.95*
0,23
6,07
11 ,0
12.7
37.3
364
n.f 1
180
10.2
120,4; 101.8
76.9
86.1; 1.1
22.3
found
10,4; 10,3
1.2** 7,1**
33,8** 23,3**
81,1** 31,0**
1 16, 1 ww 61 ,*w"
For b) n.f, » < 5 ppnv
For c) trace » < 1 ppmv
147
-------
TABLE 8. HYDROCARBON CONTENT IN SELECTED GENERATORS SECTION
GAS STREAMS
Measurement
points
2.2.
3.2.
3.5.
3.6
Other hydro-
carbons
vol.- %
trace
trace
not found
1,33
1,32
1,18
trace
trace
trace
1,16
9,30
1,02
C2
trace
trace
C3
trace
trace
vol . -
C4
trace
trace
not found
0,84
0,79
0,72
trace
trace
trace
0,63
2,68
0,71
0,38
0,32
0,29
trace
trace
trace
0,17
6,25
0.21
0,11
0,16
0,09
trace
trace
trace
0,19
0,08
0,02
%
C5
trace
trace
.,. Ben-
Cg zene
trace
trace
not found
trace
0,03
0,05
trace
trace
trace
0,12
0,20
0,02
trace
0,02
0,03
trace
trace
trace
0,05
0.09
0.01 0,05
Note: Trace for hydrocarbons » < 0,001 vol.-%
not found (n.f) » < 0,0001 vol.-%
148
-------
TABLE 9. SLAG PROPERTIES BECAUSE OF DRIED LIGNITE GASIFICATION; M.P. 12.2;
(COMPOSITE SLAG SAMPLES ANALYSIS)
Proximate and ultimate analysis:
Moisture %
Ash%
Sulfur total %
Sulfur bound %
Sulfur combust %
Coke X
C fix X
Volatlles mat. X
Combustibles mat. X
Carbon dioxide (C02)X
Carbon X
Hydrogen X
(Nitrogen + Oxygen) X
Slag chemical composition:
36.46
58.06
0,08
0,06
0.02
58.76
0.70
4.78
5.48
4.64
1.35
0,36
3.75
_
91.38
0,12
0,10
0,02
92,47
1,09
7,53
8,62
7,30
2,12
0,57
5.91
30.87
62.27
0,07
0,06
0,01
63.65
1.38
5,48
6.86
5,83
1,72
0,39
4,74
_
90,07
0.10
0.09
o.ot
92,07
2.00
7.93
9.93
8,44
2,49
0,57
6,86
Slag
34,86
57,83
0,12
0,08
0,04
57,94
0,11
7,20
7,31
6,50
2.46
0,40
4,01
spectrochemlcal
.
88,78
0,18
0.12
0,06
88,94
0.16
11,06
11,22
9.98
3,78
0,61
6,77
analysis *ppm"
S102 X
Fe203 X
A1203 X
CaO X
MgO X
50 3 *
P2°5 *
T102 X
Na20 X
KZO x
Ratio Add/Base:
A1203*S102 + T102
Fe2Q3+cao+Hgo+al kalles
Ash fusibility:
(Oxldatlve atmosphere)
Initiation of sintering °C
n »
37.74
7.50
13.31
31.60
6,08
0.29
0.24
0.90
1,15
0.81
1.102
1130
35,77
5.49
13,42
35.80
5,98
0,27
0.24
0.90
0.99
0,73
0.982
1100
30,23
10.38
8.73
41.05
6.44
0.36
0.27
0.80
0,98
0,47
0,670
1130
B
Ba
Be
Mn
Se
Pb
Cr
Ga
N1
Ho
V
Cu
Y
Zn
Co
Sr
Sc
Cd
630
1670
below
2700
2
29
240
37
180
30
137
48
39
56
15
4100
20
1,2
detection
Softening temperature
1180
1205 1280
Hemisphere temperature C
Flow temperature °C
1195
1205
1220
1240
1290
1300
149
-------
TABLE 10. PROPERTIES OF WASTEWATER (SAMPLING POINT 12.3)
Components
pH value
Suspended sol ides mg/1 (105°C)
Total Residue of Evaporation
mg/1 (105°C)
Fixed Remainder of Total Eva-
poration's residue mg/1 (600°C)
Evaporation's residue of dissolved
matter mgA (105° C)
Fixed remainder of Evaporation
residue of dissolved matter,
mg/1 (600°C)
COD (K2Cr207) mg02/l
Permanganate value, mg/1 (KMn04)
BOD5 mg02/l
Volatile Phenols, mg/1
Ammonia free, mg/1
Ammonia fixed, mg/1
Cyanide (CN~), mg/1
Hydrogen sulfide mg/1
Tar+Oil (ether extracts) mg/1
Chloride (Cl~) mg/1
Sulfates, mg/1
Rhodanate (CNS~), mg/1
Thiosulfates ($203), mg/1
Fluorides (F~), mg/1
Nitrites (N02), mg/1
Nitrates (N03), mg/1
Sulfites (SOp, mg/1
10,9
570
760
11,7
559
1432
11,0
460
12,1
204
1330
130
1991
1780
2550 2314
240 1778
2090 2110
90 1275
18
33
28
0,11
trace
1.6
0,01 <
trace
0,0
20,5 2
345
0,025
trace
0,90
0.60
5,5
trace
-
-
-
0.3
1,5
0,01
0,0
5,5
515
0,03
0,65
0,29
4,0
215 1588
49 154
94 139
90
4,25 0,25
trace
2,3 2,2
trace <0,01
trace
0,0
36 36,5
339 668
0,03 0,02
trace
1,0 1,19
0,30 0,82
4,3 5,61
trace
150
-------
TABLE 11. VOLUMES OF RECTISOL SECTION STREAMS
Measurement Amounts
Point Measured Estimated From
Calculated design
1 2(10,3 t/h) 3 (16 t/h)
7~3 10131 mj*/h17.220 mjj/h
Raw gas 10410 mjj/h
(Feed for Rectisol
Section)
7.2 4870 niM/h 1753 mjj/h 2174- mj*/h
N N -5300 N
Waste gas C0« and
other components
7.1. 3490 mj/h 2958 mj*/h 2.546 mj/h
H2S Waste gas
and other components
7.4. 7235 mjj/h 5775 mj*/h 12.500 mjj/h
Glean gas
(Final Product) 5934 mj*/h
151
-------
TABLE 12. RESULTS OF THE RAW GAS ANALYSIS FROM M.P. 7.3
Gas Stream
M.P. 7.3
Section: Rectlsol
Campaign 1 2
a)
b)
c)
<*)
e)
f)
Gas Composition vol.-X "Orsat" 6.C.
(Orsat and G.C. Methods)
- H2 39.8-42.8 38.07-45.2
- CnHm 0.4-0.4 1.60-2,41*
- 02 0,2-0,2 1,63-2,56
- N2 0,9-1,0 1,33-9,48
- CH4 9,9-8,8 11,9 -8,92
- CO 11.6-9,8 9,65-10,07
- C02 37.2-37.0 35,82-21,37
Chen. meth. g/100 mj| (dry)
- H2S 1097-1181
- NH3 130-138
- Phenols 0,352
- HCN 84-85
G.C. meth. g/100 mjj (dry)
- H2S 150-425
- NOX
- COS
- so2
- methyl mercaptan 21-73
- ethyl mercaptan
- unknown ppm
Moisture t
Partlculates g/100 m^ (dry)
(method 5)
Dissolved solids
Tar Components
Total "e"
Flow: mj|/h (dry)/ Gen. 1n operation
- designed 17.220
- calculated 10131; 10410
- measured
3
G.C.
36.1
1,21*
0,55
1,55
12.8
13.5
33,4
673-804
0,25
0.129
7.30
681,5
-
19.8
-
116.6
27,2
-
Note: * Other hydrocarbons
152
-------
TABLE 13. RESULTS OF THE WASTE GAS ANALYSIS FROM M.P. 7.2
Gas Stream
M.P. 7.2
Section: Rectlsol
Campaign 1 2
a) Gas Composition vol.- 1 "Orsat" G.C.
(Orsat and G.C. Methods)
- H2 0,0 0.63-0,20
- CnHm 0,0 0,96-1,45*
- 02 0,3 0,10-0,62
- N2 1,4 2.47-3.74
- CH4 9,4 1.41-1 ,81
- CO 1,2 - -
- C02 87,7 93,98-91,77
b) Chem.meth. g/100 mjj (dry)
- H2S 0,0-21 13,7-00
- NH3 0,0 0,0
- Phenols 0,0-0,027 0,009-0,068
- HCN
c) G.C. meth. g/100 mj| (dry)
- H2S
- NOX
- COS
- so2
- methyl mercaptan
- ethyl mercaptan
- unknown ppm
3
G.C.
0,83
0,49*
0.06
0,32
0.94
94,08
10,0
0.35
1.53
0,67
trace
1,8
0.9
4,0
d) Moisture t
e) Partlculates g/100 ra^ (dry)
(method 5)
Dissolved solids
Tar Components
Total "e"
f) Flow: mjj/h (dry)/Gen.1n operation
- designed 2174-5300
- calculated 4870
- measured
note: * Other hydrocarbons
For c) trace « <
1 ppmv
153
-------
TABLE 14. RESULTS OF THE WASTE GAS ANALYSIS FROM M.P. 7.1
a)
c)
d)
e)
t}
9)
1 2
Gas Composition vol-t "Orsat" G.C.
(Orsat and G.C. Methods)
- H2 1.6 0,02-0,07
- CnHm 0,2 ), 46-1, 54*
- 0. 0,0 0,41-0,23
- Nz 0,2 0,81-0.59
- CH4 8,6 4,44-1,65
- CO 1,8 2,93-1,65
- C02 87,0 87,94-91,45
Chem.meth. g/100 mj] (dry)
- H2S 1519 4138 - 4224
- NH3 0,86 0,0 - 1,7
- Phenols 0.028 0.007-0,011
- HCN
G.C. meth. g/100 m£ (dry)
- H2S 3293
• N0x
- COS trace
- soz
- methyl mercaptan 210
- ethyl mercaptan
- unknown ppm
Moisture t
Partlculates g/100 m^ (dry)
(method 5)
Dissolved sol ids
Tar Components
Total "e"
Flow: mjj/h (dry)/Gen.1n operation
- designed 2546
- calculated
- measured 3490**
Heating value*"*
Gross kcal/mjj 965 840-555
Net kc«l/mjj 870 785-525
3
G.C.
NF
0,77*
0.51
1,39
4,15
2,64
86,94
3541
167
O.Z7
10.1
4083
133,9
786,4
201,5
650
600
* o t * : 'Other hydrocarbons; "at M. P. 20, 1 (1.7.1978); For c)
rwS,",* P&mv; "*Ca)culated without Sulfur Compounds
Combustion (Prof. G. Wagener)
154
-------
TABLE 15. RESULTS OF CLEAN GAS ANALYSIS FROM M.P. 7.4
Gas Stream
M.P. 7.4
Section: Rectisol
Campaign
•)
*>)
c)
d)
e)
Gas Composition vol. -I
(Orsat and 6.C. Meth. ) "Orsat" "Orsat" 6.C.
- H2 66,1 62,4-65.0 64,78-62,09
- CnHm 0,3 0,4-0,4 0,42-0.54*
- 02 0,1 0,1-0.1 1.50-1,76
- N2 1,5 0,8-1,0 2,71-2,46
- CH4 13,5 16.1-14.2 16,25-15.22
- CO 16,5 17.3-16.5 11.06-15.34
- C02 2.0 2,6-2,6 2,65-2.22
Chem.meth. g/100 mjj (dry)
- H2S 0.0 0,0
- NH3 0,0 0,24-0,20
- Phenols - 0,016-0,014
- HCN -
G.C. raeth. g/100 mjj (dry)
- H2S
- NOX
- COS
- so2
- methyl mercaptan
- ethyl mercaptan
- unknown ppm
Moisture X
Particulates g/100 mjj (dry)
(method 5) N
Dissolved solids
Tar Components
"Orsat" G.C.
62.2 63,9
0.3 0,27
0.1 1,23
2.0 5.87
17.6 11.93
17.1 14.33
0,7 n.f.
-
0,20
n.f.
n.f.
0.2
trace
Total "e"
n
9)
Flow: mjj/h (dry)/Gen. In operation
- designed 12.500
- calculated 5775; 5934
- measured 7235
Heating value**
Gross kcal/mjj 3870 4050 3925 3950 3930
Net kcal/uijj 3415 3590 3470 3480 3480
4155 3580
3685 3155
Note: * Other hydrocarbons; n.f • not found.
For a) n.f » 0.01 vol.-*; For c) n.f. « < 0,1 ppmv: trace
** Calculated (Prof. G. Wagener)
< 1 ppm;
155
-------
TABLE 16. HYDROCARBON CONTENT IN SELECTED RECTISOL SECTION GAS STREAMS
Measurement Other
Points hydro-
carbons
7.3
7.2
7.1
7.4
1
2
1
0
1
0
1
1
0
0
0
0
,60
,41
,21
,96
,45
.49
,46
,54
,77
,42
,54
,27
vol-%
C2
1
1
0
0
0
0
0
0
0
0
0
0
.04
,11
,65
,37
,59
,29
,63
.73
,34
,35
.45
,25
C3
0,
0,
o,
o.
0,
o,
o,
o,
o,
0,
o,
o,
35
40
35
27
37
20
32
39
22
07
09
004
C4
0
0
0
0
0
vol. - %
C5
,20
,24
.15
,21
,23
trace
0
0
0
n
n
n
,27
,19
,14
.f
.f
.f
0,01
0,37
0,04
0,09
0,17
trace
0,21
0,11
0,06
n.f
n.f
n.f
+ Ben-
Cg zene
trace
0.
0.
0.
0.
n
0,
0,
0,
n.
n.
n,
29
02
02
09
.f
03
12
01
f
f
02
Note: n.f • not found » < 0,0001 vol.-*
Trace for hydrocarbons » < 0,001 vol.-%
156
-------
TABLE 17. VOLUMES OF TAR SEPARATION SECTION EMISSIONS
Measurement
Point
Amounts
Measured From design Evaluated and
calculated
1 2(16 t/h)
3 (10,3 t/h)
13.1. Tar Tanks
0,5 mjj/h
0,32 mj/h
13.3. Medium Oil Tank
94
60,5 mj/h
13.5. Gas Condenser Tank
9 mj/h
5,8 mjj/h
13.6. Expansion gases to
waste gases flare
30-360 mj*/n 210,3 mj/h
13.7. Phenolic
water tanks
13 mj]/h
8,4 mj/h
157
-------
TABLE 18. RESULTS OF WASTE GAS ANALYSIS FROM M.P. 13.1
Gas Stream
M.P. 13.1
Section: Tar Separation
Campaign
a)
b)
c)
d)
e)
f)
Gas Composition vol. -X
(Orsat and G.C) "Orsat" G.C
- H2 1,1 n.f
- CnHm 0,0-0,6 0.01*
- 02 15,6 19,20
- N2 80,4 78,90
- CH4 0,6 0,08
- CO 0,3 n.f
- C02 2,0 1.11
Chem meth. g/100 m^ (dry)
- H2S 297 52,3-1140
- NH3 2,81 151
- Phenols 0.0185 2,046
- HCN - -
G.C. meth. g/100 mjj (dry)
- H2S 59,2-75,9
' N°x
- COS
- soz
- methyl mercaptan trace - 26
- ethyl mercaptan
- unknown ppm
Moisture t
Participates g/100 mjj/dry/
(method 5)
- dissolved sol Ids
- Tar Components
Total "e"
Flow: m?/h (dry)/ Gener. In Operation
-designed
-calculated 0,5
-measured
G.C
trace
0,01*
19,60
72,12
0,10
n.f
3,09
1920 (1920)
198 (198)
22,06
15,37
273
n.f
130,1
66,7
-
26,8
Ho t t: Other hydrocarbons xj Trace for hydrocarbons •
n.f • not found
< 0.001X
158
-------
TABLE 19. RESULTS OF WASTE GAS ANALYSIS FROM M.P. 13.3
Gas Stream
M.P. 13.3
Section Tar Separation
Campaign
a)
b)
c)
d)
e)
n
Gas Composition vol.-*
(Orsat and G.C)
" H2
- CnHm
- o2
- N2
- CH4
- CO
- C02
Chem. meth. g/100 m]j (dry)
- H2S
- NH3
- Phenols
- KCN
G.C. meth. g/100 mjj (dry)
- H2S
' N0x
- COS
- S02
- methyl mercaptan
- ethyl mercaptan
- uknown ppm
Moisture S
Participates g/100 mjj (dry)
(method 5)
- Dissolved sol Ids
Tar Components
Total "e"
Flow: njj/h (dry)/ Gener.ln
Operation
- designed
- calculated 94
- measured
"Orsat" G.C
20,4 22.48
0,9 2,75*
0,6 0,84
1.1 3.02
9.7 2,74
5.3 3.06
62,0 50,67
5639-3647 6275
3,49 1,3
0,0177 0.114
-
952
135
G.C
n.f
0,96*
0.89
3,36
7.64
n.f
86,36
940
408
45,2
6,31
1882
216.4
126.5
""
11.4
* o t •: * Other hydrocarbons
n.f • not found
159
-------
TABLE 20. RESULTS OF WASTE GAS ANALYSIS FROM M.P. 13.5
Gas Stream
M.P. 13.5
Section Tar Separation
Campaign 1 2
a)
b)
c)
d)
•)
Gas Composition vol.- J "Orsat" G.C
(Orsat and G.C. }
- H2 16,2 13,78
- CnHra 0,8 0,13* - 2,63*
- 02 12.8 15,14
- N2 49,8 58.01
- CH4 4,6 2.10
- CO 3,6 n.f
- C02 12,2 9,06
Chem. meth. g/100 mjj (dry)
- H2S 788 1055
- NH3 3.72 62
- Phenols 0,0177 4.79
- HCN
G.C. meth. g/100 m* (dry)
- H2S
- N°x
- COS
- so2
- methyl mercaptan
- ethyl mercaptan
- unknown ppm
Moisture X
Partlculates g/100 mj] (dry)
(method 5)
Dissolved solids
Tar Components
Total "e"
3
G.C
14,63
0,22*
16.60
60.99
1.19
n.f
6. IS
938.4
n.f.
0.456
20,34
125,8
45.2
19,76
•
1.0
f) Flow: mjj/h (dry)/Gen.1n Operation
- designed
- calculated 9
- measured
H o t t: * Other hydrocarbons
Not found - n.f
160
-------
TABLE 21. RESULTS OF WASTE GAS ANALYSIS FROM M.P. 13.6
Gas Stream
M.P. 13.6
Section Tar Separation
Campaign
«)
b)
c)
d)
Gas Composition vol.- t
(Orsat and G.C. )
- H2
. CnHm
- o2
-N2
-CH,
- CO
-co2
Chen. meth. g/100 mjj (dry)
-H2S
- NH3
- Phenols
- HCM
G.C. meth. g/100 mjj (dry)
- H2S
- N°x
- COS
- S02
- methyl mercaptan
- ethyl mercaptan
- unknown ppm
Moisture I
"Orsat" G.C
25,4 9,87
0.8 2.69*
0,6 1,10
3,5 1.78
9,3 5.91
7.8 6,75
52.6 71.73
1594 2936
4.0 32.0
0.056 4.454
~ ~
G.C
11,1
1,23*
0,47
0,56
6,07
7,17
72,1
2081
1484**
4,157
8,22
1745
n.f
195,2
76.7
-
e) Participates g/100
(method 5)
Dissolved solids
Tar Components
(dry)
Total "e"
f) Flow: mjj/h (dry)/Gener. 1n Operation
- designed 30-360
- calculated
- measured
Note.- * Other hydrocarbons; Trace for hydrocarbons • < 0,001X
Condensate had 1436 g/100 m? phenols 1n Campaign 2
1.1 j * Ik *• A« |J A_ * • * _ "
** with condensate
n.f • not found
161
-------
TABLE 22. RESULTS OF WASTE GAS ANALYSIS FROM M.P. 13.7
Gas Stream
M.P. 13.7
Section Tar Separation
Campa tgn
a)
b)
c)
d)
e)
Gas Composition vol.-i
(Orsat and G.C)
- H2
. Cnum
- °2
- N2
- CH4
- CO
- co2
Chem. meth. g/100 mj| (dry)
- H2S
- NH3
- Phenols
- HCN
G.C. meth. g/100 m* (dry)
- H2S
- COS
- S02
- methyl mercaptan
- ethyl mercaptan
- unknown ppra
Moisture t
Partlculates g/100 m^ (dry)
(method 5)
Dissolved sol Ids
Tar Components
Total "e"
"Orsat" G.C. G.C.
0,0 0,02 Trace
0,0-2,2 0,14* 0,25*
13,0 10.78 12,60
53,4 48,65 52,65
0,4 0,20 0,18
0,0 n.f n.f
33,0 39,32 28,9
1054 2518 981
3.75 618 895
0,021 7.12 0.366
4,64
74.4 274,7
n.f.
trace 131,9
110,4
-
41.0
f) Flow: mjj/h (dry)/Gen.1n Operation
- designed
- calculated 13
- measured
tote: * Other hydrocarbons; trace for hydrocarbons • < O.OOIt;
Condensate In Campaign 2 had 14,36 g/100 mjj phenols;
not found = n.f
162
-------
TABLE 23. HYDROCARBONS CONTENT IN SELECTED TAR SEPARATION SECTION
GAS STREAMS
Measurement Other
Points hydrocarbons C2
vol.- %
13.1
13.3
13.5
13.6
13.7
0
0
0
1
2
0
1
0
0
1
1
2
0
1
0
0
0
,01
,31
,01
,56
,76
,956
,07
,13
,22
.49
,19
,69
,85
,23
,28
,14
,25
0
0
,01
,04
trace
0
0
0
0
0
0
0
0
1
0
0
0
0
0
,39
,42
,62
,40
,09
,07
,91
.72
.05
,53
.4
,11
,01
,02
C3
trace
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
,03
,004
.24
,32
,16
,21
,02
,05
,37
,29
.62
.21
,33
,06
,08
,01
vol . -
C4
trace
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
,24
,005
,39
,66
,006
,27
,02
,03
,21
,11
,72
,09
,27
,02
,05
,05
* +
r r
L5 L6
n.f
trace
trace
0,21
0,20
0,08
0,13
trace
0,04
trace
0,05
0,15
0,02
0.15
0,07
trace
0,09
n.
n.
0,
0,
0,
0,
0,
Ben-
zene
f
f
001
33
16
09
06
trace
0,
03
trace
0,
0,
02
15
trace
0,
0,
n.
0,
08
02
f
08
Note: Trace for hydrocarbons
not found (n.f) * <
< 0,001 Vol.- %
0,0001 vol.- %
163
-------
TABLE 24. VOLUMES OF PHENOSOLVAN SECTION STREAM EMISSIONS AND OF
EXPANSION GASES LARGE FLARE
Measurement Amounts
Points
Measured From design Evaluated and
calculated
1 2 (16 t/h)3 (10,3 t/h)
14.5 Column 1 vent 144 mj|/h 92,7 m3/h
14.9 Grude phenol
tank vent 0,1 m3/h 0,06 m3/h
20.1 Gases to Large
Flare 2990-3320 mjj/h 3448,4 m3/h
14.11 Waste waters 13 m3/h 8,4 m3/h
N o t e: During sampling the Phenosolvan Section was not in
normal production conditions; wastewater had a high
content of phenols. For that reason the quality data
of wastewaters are not given.
164
-------
TABLE 25. RESULTS OF WASTE GASES ANALYSIS FROM M.P. 14.5
Gas Stream
M.P. 14.5
Section Phenosolvan
Campaign 1
a) Gas Composition vol.-X "Orsat"
(Orsat and G.C. methods)
- H2
- CnHra 0,4
- oz
- N2
- CO
- C02 99,0-17,6
b) Chem. meth. g/100 mj| (dry)
- H2S 0,0
- NH3 2611
- Phenols 0.0925
- HCN
c) G.C. meth. g/100 mjj (dry)
- H2S
- NOx
- COS
- soz
- methyl mercaptan
- ethyl mercaptan
- unknown ppm
d) Moisture t
2
G.C.
n.f
0,07*
2,30
1,99
trace
n.f
91.42
534
16.6
614.3
-
trace
trace
3
G.C
n.f
trace*
16,07
59,20
trace
n.f
24,50
6510-7058
16931-43563
15758-15529
29.97
1093
n.f.
n.f.
30,85
8.8
82.1; 88.3
e) Partlculates g/100 mjj (dry)
(method 5)
Dissolved sol Ids
Tar Components
Total "e"
f) Flow: m^/h (dry)/Gen. 1n Operation
- designed
- calculated 144
- measured
< 0,001 vol.-«
• < 0,01 vol.-«
N o t «: * Other hydrocarbons; trace
n.f « not found; For a) n.f • < 0,01 vol.-J For c) n.f.«<
ppmv; Content of H2S in condensate « S48-615 g/100 mjj
Content of NH3 In condensate * 26632 g/100 mjj (Campaign 3)
0.1
165
-------
TABLE 26. RESULTS OF WASTE GASES ANALYSIS FROM M.P. 14.9
Gas Stream
M.P. 14.9
Section Phenosolvan
Campaign
a) Gas Composition vol.-X
(Orsat and G.C. Methods)
- H2
- CnHm
- °2
- CH4
- CO
' C02
b) Chem.meth. g/100 mjj (dry)
- H2S
- NH3
- Phenols
- HCN
c) G.C. meth. g/100 m^ (dry)
- H2S
- NOX
- COS
- S02
- methyl nercaptan
- ethyl mercaptan
- unknown ppm
1 2 3
"Orsat" G.C G.C
0,0 n.f n.f
0,0 trace* trace*
18.6 18.79 20,45
80,8 79,89 76,26
0,6 trace trace
0,0 n.f n.f
0,0 n.f n.f
0.0 456-1070 27,3
0,0 19,8 0.92
0.0174 28.7 8,62
4,07
n.f
n.f
n.f
d) Moisture X
e) Participates g/100 mj] (dry)
(method 5)
Dissolved solids
Tar Components
Total "e"
f) Flow: nifl/h (dry)/Gen. In Operation
- Designed
- Calculated 0,1
- Measured
K o t a:
n.f • not found;
For a) trace •
For c) n.f. •
Other hydrocarbons
0,1 vol.- X
0,1 ppmv
166
-------
TABLE 27. RESULTS OF EXPANSION-WASTE GASES ANALYSIS FROM M.P. 20.1
Gas Stream
M.P. 20.1
Section Expansion
gases Large Flare
Campaign
a) Gas Composition vol.- X
(Orsat and. G.C. Methods) "Orjat" G.C
- H2 - n.f
- CnHm - o,79*
- 02 - 0.37
- H2 - 1,86
- CH4 - 4,83
- CO - n.f
- C02 - 91,55
b) Chen. neth. g/100 njj (dry)
- H2S 0,4 vol.-X 2900
- NH3
- Phenols
- HCN
c) G.C. meth. g/100 njj (dry)
- H2S 167
- COS
-so2
- methyl mercaptan trace
- ethyl mercaptan
- unknown ppm
G.C
trace
2,84*
0,06
0.5
10.41
n.f
88,10
1295-1625
n.f
0,424-0,467
12,5
2747
75,5
317
165
d) Moisture X
e) Participates g/100 mj* (dry)
(method 5)
Dissolved solids
Tar Components
Total "e"
f) Flow: mjj/h (dry)/Gener. 1n Operation
- designed
- calculated 2990-3320
- measured 3490
H o t »: * Other hydrocarbons; Not found • n.f.
for a) trace • < 0,1 vol.- X n.f. » < 0,01 vol.-X
For b) n.f. • < 5 pprav
For c) trace • < 1 ppmv
167
-------
TABLE 28. HYDROCARBONS CONTENT IN SELECTED PHENOSOLVAN SECTION GAS
STREAMS AND IN EXPANSION GASES LARGE FLARE
Measurement
Points
14.5
14.9
20.1
Other
Hydrocarbons
vol.- *
0,07
trace
trace
trace
trace
0,79
0,52
2,84
C2
0,02
trace
trace
trace
trace
0,38
0,19
1.01
C3
0,01
trace
trace
trace
trace
0,27
0,11
1,03
C4
0,04
trace
trace
trace
trace
0,09
0,12
0,59
C5
n.f
trace
n
.f
trace
n
0
0
0
.f
.04
,08
,14
«
n
+ Ben-
6 zene
.f
trace
n
n
n
0
0
0
.f
.f
.f
.01
.02
.07
N o t a: Trace for hydrocarbons « < 0,001 vol.- %
Not found » n.f - < 0,0001 vol.- X
168
-------
TABLE 29. VOLUMES OF THE MOST IMPORTANT STORAGE SECTION GAS EMISSION
INTO THE ATMOSPHERE
Measurement
Point
Amounts
Measured From design
Evaluated and
calculated
2 (16 t/h) 3 (10,3 t/h)
15.3 Gasoline
Tank Vent
0,14 mj5/h 0,09 mjj/h
169
-------
TABLE 30. RESULTS OF THE WASTE GAS ANALYSIS FROM M.P. 15.3
Gas Stream
M.P. 15.3
Section Storage
Campaign
a)
b)
d)
«)
e)
Gas Composition vol.-X
(Orsat and G.C. Methods)
- H2
- CnHm
- °2
- N2
- CH.
- CO
- coz
Chem.meth. g/100 mjj (dry)
- H2S
- NH,
- Phenols
- HCN
G.C. Meth. g/100 mjj (dry)
- H2S
-NOX
- COS
-so2
- methyl mercaptan
- ethyl mercaptan
- unknown ppm
Moisture X
Participates g/100 mj| (dry)
(method 5)
Dissolved solids
Tar Components
Total "e"
."Orsat* G.C G.C
0,0 n.f n.f
0,2 0,223* 0,69*
9,0 4,12 3,89
90,6 95,29 95.32
0,0 trace n.f
0,0 n.f n.f
0,4 n.f n.f
28,56 126-329 237
1,77 0,9 n.f
0.034 0.268 0,0562
129,45
n.f 10,5**
-
n.f.
-
n.f 872.1
n.f 1857
18
f) Flow: mjj/h (dry)/Gen. 1n Operation
- designed
- calculated 0.14
- measured
Not*: * Other hydrocarbons; ** unsure Identification
For a) trace * < 0,1 vol.-X; Not found • n.f • < 0,01 vol-X
For b) not found • n.f. • < 5 ppmv
For c) not found • < 0,1 ppmv
170
-------
TABLE 31. HYDROCARBON CONTENT IN SELECTED STORAGE SECTION GAS EMISSION
INTO THE ATMOSPHERE
Measurement
po1nt
Other
hydrocarbons
vol.- X
vol.- X
Ben-
zene
15.3
0,223
0,69
0,007 0,004 0,030 0,095 0,087
0,009 0,007 0,10 0,39 0,18
171
-------
TABLE 32. DATA ON LIQUID PRODUCTS
Products
Measurement Point
Amounts, calculated
from design
Water, %
Ash, %
Total Sulfur %
Heating value
Gross kcal/kg
Net kcal/kg
Carbon %
Hydrogen X
Phenols X
Other Phenol s X cca
(o,m,p,cresol ; ethyl
phenol; dimethyl phenol ;
trlmethyl phenol)
Pyrldlnes X cca
Spec. gravity, g/cm
Residue after extraction
with toluene and benzene
Gaso-
11ne
15.3
0,65
t/h
0,0
1,45
9395
8925
78,07
8,72
-
0,2
10
0,845
-
Medium
011
15.2
1,55
t/h
0,80
, -
0,95
9880
9395
82,43
8,96
2,1
12,0
•
0,972
-
Tar
15.1
2,2
t/h
1,08
0,92
0,75
8710
8275
72,51
8,06
0,7
3,8
«•
1,059
6,9
Phenol
15.4
0,38
t/h
7790
172
-------
TABLE 33. HEAT BALANCE
Feed
Coal
Steam
Eltctrlc
curr.
Output
Clean gas
Liquid prod.
(gasoline.
•td.oll ,tar,
phtnol )
Wastt gases
flart
Heavy tar
C02 vent
Other vents
Heat consump-
tion for re-
quired power
generation
Heat consump-
tion for re-
quired stead
generation
Conveyed heat
(Steam-raw gas)
conveyed neat
(hot dry raw
gas)
Slag losses
Other not sta-
ted heat losses
and balancing
error
Amount
1 kg
0.8 kg/kg
of coal
0.1S35 KM/ kg
of coal
Amount
0.6062 njj
per kg coal
0.0594 kg/kg
of coal
0,3348 HJJ
per kg coal
0.00625 kg
per kg coal
0.1702 Mjj
per kg coal
0.3615 M*
460.5 kcal
§er kg coal
94,4 xcai
per kg coal
0.7 kg/kg
of coal
1.0636 mj!
per kg coal
0.1625 kg
per kg coal
Heating value
kcal/kg
3.470 Kcal/kg
743 kcal/kg
3.000 kcal/KU
Heating value
kcal/mjj; kcal/kg
3600 kcal/mj}
8042 kcal/kg
766 kcal/mjj
7000 kcal/kg
190 kcal/mj|
19,95 kcal/m^
n
70X
10»
660 kcal/kg
101 kcal/m^
120 kcal/kg
Amount of
heat (kcal/kg
of lignite)
3.470.0
594.4
460.5
4.524.9
Amount of heat
kcal/kg of coal
2.183.3
477.7
256.5
43.8
32.4
7.2
322
69.4
462
107,6
19.5
554.5
4. 5Z4.9
S share
76,68
13.14
• 10.18
100.00
X share
48,22
10.56
5,67
0,97
0,72
0,16
7,11
1,32
10,21
2,38
0.43
12.25
100,00
173
-------
TABLE 34. CARBON BALANCE
Feed
Output:
Clean gas
1 kg of coal Carbon con-
tent "C"
42.72%
0,6062 mj*/kg 0,176 kg/mj*
coal
Amount
kg C/kg
of coal
0,4272
0,1067
% share
100,00
24,98
Waste gases
flare
C02 vent
Other vents
Liquid prod-
ucts (gaso-
line, medium
oil, tar, phe-
nol)
0,3348 mjj/kg
coal N
0,1702 mjj/kg
coal N
0,3615 m»/kg
coal
0,490 kg/mjj 0,1640 38,39
0,512 kg/mjj 0,0871 20,39
0,0326 kg/mjj 0,0118 2,76
0,0594 kg/kg coal 78,15% 0,0464 10,86
Heavy tar
0,00625 kg/kg
coal 72%
0,0045
1,05
Slag-ash
0,1625 kg/kg
coal
2,83% 0,0046 1,08
Losses (waste-
water, etc.)
and balancing
error
0,0021 0,49
0,4272 10l),UO
174
-------
TABLE 35. SULFUR BALANCE
Coal feed
Output;
Slag
waste gases
flare
C02 flare
Clean gas
Liquid pro-
ducts
Heavy tar
Other vents
Total
Balancing error
1 kg
0,1625 kg/kg
coal
0,3348/mjJ/kg
coal
0,1702 mj/kg
coal
0,6062 m^/kg
coal "
0,0594 kg/kg
coal
0,00625 kg/kg
coal
0,3615 mjj/kg
coal N
S content Amount
1<15% (gS/kg of
coal )
11.5
0,1375% 0,223
31,29 10,477
gs/mjj
0,0258 gS/m-j 0,0038
2.14.10-6 1.30.10-6
0,8% 0,475
0,71% 0,044
1,06 g/mj* 0,383
11,61
+ 0.11
% share
100
1,94
91,11
0,033
0,000
4,13
0,38
3,33
100,923
+ 0.923%
175
-------
TABLE 36. EMISSIONS OF OTHER MAJOR POLLUTANTS
(During gasification of 10 t/h of dried lignite)
As determined by measurements, the following is emitted
during gas production according to Lurgi process at a
rate of 10 t of dried Kosovo lignite per hour:
a) From various vents in sections: Generators, Phenosolvan,
Tar Separation and Storage (Measurement Points: 2.2;
3.2; 3.5; 13.1; 13.3; 13.5; 13.7; 14.5; 14.9; 15.3)
Pollutants frlow rate Concentra- Amount
(mj/h) t1on
Sulfur (H2S; COS,
CH3SH, CH3 CH2SH)
As "S"
3615 1,06 3 832
Ammonium (NH3) 3615 0,11 398
Phenols 3615 1,28 4 627
Hydrocyanic acid (HCN) 3615 0,0099 35,8
Hydrocarbons (CnHm) 3615 0,4 1 446
Hydrogen (H) 3615 0,3773 1 364
Carbon monoxide (CO) 3615 1,239 4 478
Carbon dioxide (C02) 3615 86,59 313 032
Methane (CH4) 3615 0,0627 227
Particulates 3615 41,08 148 498
176
-------
TABLE 36. (continued)
b) From C02 Vent; Section Rectisol; Measurement
Point 7.2.
Pollutants
Sulphur (H2S, COS,
CH3SH)
(CH3 CH2SH)
As "S"
Ammonium (NH0)
Phenol s
Hydrocyanic add (HCN)
Hydrocarbons (CnHm)
Hydrogen (H)
Carbon monoxide (CO)
Carbon dioxide (CO,)
Methane (CH.)
Flow rate
(mjj/h)
1702
1702
1702
1702
1702
1702
1702
1702
1702
Concentration
(g/mj5>
0,0258
0,01
0,068
0,0153
6,557
0,719
0,0
1860
5,713
Amount
9
43,9
17,0
115,7
26,0
11169
1 224
0,0
3 165700
9726
177
-------
TABLE 36. (continued)
c) From Plant waste gases Large Flare, Measurement
Point 20.1. (Waste gases from MP 3.6; 13.6; 7.1)
Pollutants
Sulphur (H2S; COS,
CH3SH, CH3 CH2$H)
As "S"
Ammonium (NH7)
Phenols
Hydrocyanic add (HCN)
Hydrocarbons (CnHm)
Hydrogen (H)
Carbon monoxide (CO)
Carbon dioxide (C0?)
Methane (CHA)
Nitrogen oxides (N00)
Sulphur dioxide (S0?)
Flow rate
(»j|/h)
3348
3348
3348
3348
3348
3348
3348
3348 1
3348
3348
3348
Concentra-
tion
(g/mjj)
26,88
0,52
0,04
0.1
13,4
2,534
45,2
635,6 5
3,344
Amount After Inci-
neration, g
89 994
1 741
134
335
44 863
8 485
151 247
475 896 6 154 000
11 196
5 164
179 988
178
-------
TABLE 36. (continued)
d) In the slag (Measurement point 12.2)
Pollutants Flow rate
(t/h)
Sulfur as "S" 1»625
e) In Wastewater (Measurement
Pollutants Flow rate
m3/h
Sulphur "S"
(Sulphites, thiosul-
phates, sulphates,
rhodanldes, hydrogen
sulphide) 1,0
Ammonium 1,0
Phenols 1,0
Concentration
(9/t)
1,33
Point 12.3)
Concentration
g/m3
155
1,9
1,227
Amount
9
2,161
Amount
g
155
1.9
1,227
179
-------
KOSOVO GASIFICATION TEST PROGRAM RESULTS-PART II
DATA ANALYSIS AND INTERPRETATION
Karl J. Bombaugh* and William E. Corbett
Radian Corporation, Austin, Texas
Abstract
This presentation is a progress report on an
EPA-sponsored program to characterize envi-
ronmental problems associated with the gasifi-
cation of lignite in a commercial-scale plant
using Lurgi gasifiers. The data acquisition ac-
tivities associated with this program are being
conducted at a gasification complex in the
Kosovo region of Yugoslavia as an internation-
al, cooperative effort between the United States
and Yugoslavia.
The Kosovo test program is being imple-
mented in two phases. Phase I, now completed,
addressed major and minor pollutant emissions.
Phase II, to begin in the summer of 1979, will
focus on significant trace pollutant emissions,
such as trace elements and hazardous trace or-
ganics.
Because this presentation is based on the data
that was gathered during the first test phase, it
addresses primarily the bulk properties of the
plant's major emission and effluent streams. It
will be presented in two parts. The first part, by
M. Mitrovic, addresses test procedures and re-
sults. The second part, by Radian Corporation,
considers the implications of those results in
relation to control requirements for U.S. gasifi-
cations plants.
INTRODUCTION
The overall objective of the Kosovo test pro-
gram is to characterize the environmental prob-
lems associated with an operating, state-of-the-
art, commercial-scale, Lurgi gasification sys-
tem. Because the Lurgi process has been promi-
nently mentioned in several companies' plans
for pressurized gasification systems in the
United States, the U.8. Environmental Protec-
tion Agency (EPA) is anxious to develop a sound
basis for ensuring the environmental acceptabil-
•Speaker.
ity of those facilities. Through its participation
in the Kosovo test program, the EPA hopes to
gather critical data needed to specify control
priorities and support reasonable performance
standards for future U.S. gasification facilities
based on Lurgi technology.
The Kosovo test program is divided into two
phases of effort. In Phase I, a broad screening
study of the Kosovo Plant and its emission
streams was conducted. Approximately 50 key
process and emission streams were sampled,
and analyses were performed to determine the
concentrations of the major components pres-
ent in those streams. Phase I testing was com-
pleted in November 1979.
In Phase II of the Kosovo test program, a
more select group of process and emission
streams (approximately 30) will be character-
ized in greater detail. A major portion of this
test phase will involve measuring the concen-
trations of trace and minor components in the
plant's "high priority" emission streams. Work
on this test phase is scheduled to be initiated in
early summer of 1979.
The first part of this paper summarizes the
processes and emission streams studied, the
procedures used, and the results obtained dur-
ing the Phase I test period. This writeup in-
cludes an interpretive analysis of the Phase I
test results. The topics to be addressed include:
• Lurgi process environmental problems and
control priorities, and applicability of the
Kosovo data to the U.S. gasification indus-
try;
• Key data gaps, additional questions raised,
and problems unresolved by the Phase I test
results; and
• Phase II test plans.
As a result of the Phase I test program, the
emission streams specifically associated with
the Kosovo gasification facility and generally
associated with Lurgi technology have been
defined. In subsequent sections of this paper,
the data gathered to characterize the air emis-
181
-------
sions, liquid effluents, and solid wastes
generated in the Kosovo plant are discussed,
and plans for future testing are summarized.
In order to provide a consistent basis for the
discussions that follow, the reader's attention is
directed toward Table 1 and Figures 1 through
8. These materials indicate the sources for all of
the plant's major emission streams.
Air Emission Streams
The major sources of air emissions in the
Kosovo plant are summarized in Table 2. As in-
dicated in that table, there are nine major
classes of air emission sources in the plant.
One of the most significant air emission
sources at Kosovo is the plant's Rectisol unit.
Because the Kosovo Rectisol unit is a selective
(Rectisol II) acid-gas removal process, a CC^-rich
stream that normally contains minor amounts of
H2S and other sulfur species is generated along
with an H2S-rich stream that should contain
most of the other acid gases and sulfur species.
The CC>2-rich stream is vented directly to the at-
mosphere at Kosovo. The same approach has
been proposed in several conceptual U.S. plants.
Phase I data do not indicate that this would be a
serious problem, except perhaps during upset
conditions. Components other than C02 that
were found in the C02-rich vent gas include
methane and other light hydrocarbons (which
may present hydrocarbon emission problems in
some areas of the United States because of the
relatively large flow rate of this stream). Minor
amounts of H^, HCN, and mercaptans were
also found in this stream.
The H2S-rich gas stream generated in the
Rectisol unit is a significant waste stream. At
Kosovo, this stream is flared. In the United
States, a treatment process (e.g., Glaus, Stret-
ford) that produces elemental sulfur is the
preferred approach. However, potential prob-
lems with this approach are indicated by the
Phase I data. The 002 content of this stream
may be too great to permit the economical use
of a Glaus system. Also, the presence of mercap-
tans and hydrocarbons in this stream could re-
sult in residual sulfur or hydrocarbon emissions.
As shown in Figure 3, several potential air
emission sources are associated with the
Kosovo plant lignite drying system. This section
of the plant is not addressed in this paper
because it will not be studied until Phase II.
The coal-feeding system at Kosovo is another
significant air emission source. The high-pres-
sure gas stream from the coal lock system is
flared, while the low-pressure vent stream is re-
leased directly to the atmosphere. Venting this
stream would not be an environmentally accept-
able option in the United States.
The generator startup vent gas stream was
not studied in Phase I. Variations in the flow
and composition of this stream will be studied in
Phase II.
In the tar separation section, the condensa-
tion of tars, oils, and phenolic water (at about 25
atm pressure) and the subsequent depressuriza-
tion and release of those liquids into a series of
surge tanks results in generation of:
• A low-pressure flash gas stream that is
routed to the flare; and
• Flash gases from the tar, medium oil, and
phenolic water surge tanks that are vented
directly to the atmosphere.
The vent gas streams leaving the medium oil
and phenolic water surge tanks at Kosovo are
particularly significant because of their high
flow rates and relatively high concentrations of
problem pollutants.
The only remaining waste stream that is par-
ticularly significant at Kosovo is the phenosol-
van unit condensate stripper vent. This stream
results from the steam stripping of process con-
densate upstream of the phenol plant ether ex-
traction section. As anticipated, this stream was
found to contain NH3, acid gases, and a variety
of other volatile components that leave the tar
separation section with the process gas liquor.
All of the streams mentioned above would re-
quire considerable attention in U.S. gasification
plants. With the exception of the two Rectisol
section acid-gas vent streams (which require
special attention), all of these streams should be
collected and either incinerated or recycled.
Other air emission streams that are not as
significant as those mentioned above but that
will require attention in a U.S. gasification facil-
ity are the following:
• Coal bunker and ash lock vent gases: these
streams mainly represent potential sources
of particulate emission; and
• Storage tank vent gases: these sources
should be controlled in a U.S. Lurgi plant,
but their collective impact is considerably
less than the impact caused by the surge
tank vents in the tar separation section.
182
-------
TABLE 1. SIGNIFICANT KOSOVO PLANT PROCESS AND EMISSION STREAMS
Stream
Number
Fleissner
1.0
l.l
1.2
1.3
1.4
Stream Description
Drying - See Figure 3
"Wet" coal from mine
Coal bunker vent
Autoclave vent
Fleissner Condensate
Condensate tank vent
Stream
Type
S
G
G
L
G
Estimated
Flow Rate8
24 MT/hr
7 \
? f
?
?
Conmvnts on Components of Environmental Concern
Detailed characterization desired
Coal dust plus volatile organics and possibly
Inorganics
Detailed characterization desired
Volatile organics/ inorganics
Gasification - See Figure 4
2.0
2.1
2.2
3.1
3.2
3.3
3.4
3.5
3.6
12.1
12.2
12.3
Dried sized coal
Coal bunker area - ambient sample
Coal bunker vent
Coal bucket vent
Low pressure coal lock vent
Start-up vent (to flare)
Liquor tank vent
Ash lock vent
High pressure coal lock vent
(to main flare)
Gaslfler ash (dry)
Gaslfier ash (wet)
Gasification section wastewater
S
G
G
G
G
G
G
G
G
S
S
L
16.0 MT/hr
- I
4000 Nm'/hr I
26 Nm»/hrb)
36 Nm'/hr /
? /
40 Nm'/hr
28 Nm*/hrb
350 Nm'/hr
2.7 MT/hr
>2.7 MT/hr
3 m'/tir
Detailed characterization desired
Mostly air with traces of coal dust and possibly raw
gas components
Coal dust plua raw gas components
Raw gas components
Steam plus ash dust
Coal dust plus raw gas components
Leachable species
Leachable species
Coal and ash dust plus soluble contaminants leached
from ash
Continued - Next page
-------
TABLE 1 (continued)
Stream
Number
Stream Description
Stream
Type
Estimated
Flow Bate*
Comments OD Components of Environmental Concern
oo
Tar Separation - See Figure 5
13.1 Tar tank vent
13.2 Impure tar tank vent
13.3 Medium oil tank vent
13.4 Impure medium oil tank vent
13.5 Condensate tank vent
13.6 Expansion gases (to main flare)
13.7 Phenolic water tank vent
13.8 Heavy tar and dust
13.9 Heavy tar
13.10 Light tar
13.11 Medium oil
13.12 Phenolic water to phenoaolvan
Rectlsol - See Figure 6
7.1 H2S rich gas (to main flare)
7.2
7.3
7.4
7.S
7.6
7.7
7.8
002 vent gas
Rectisol inlet gas
Rectisol outlet gas
Cyanic vater
Product gasoline to storage
Raw gas to COj absorber
Regenerated methanol
G
C
G
C
G
G
G
L/S
L
L
G
G
G
G
L
L
G
L
.4 N»'/hrc Volatile organlca/inorganics
? Volatile organics/inorganics
.25 Nm'/hrc Volatile organlcs/inorganics
? Volatile organlcs/inorganics
? Volatile organica/inorganlcs
26 Hm'/hr Volatile organics/inorganics
13 Hm*/hrc Volatile organica/inorganlcs
.1 HT/hr Volatile organics/inorganics
4 HT/hr Volatile organics/inorganics
Volatile organics/inorganics
.25 HT/hr Volatile organics/inorganics
13 m'/hr Detailed characterization desired
2.500 Hm'/hr Acid gases, sulfur species, hydrocarbons
2,200 Hm'/hr Acid gases, sulfur species, hydrocarbons
17,200 Hm'/hr Acid gases, sulfur species, hydrocarbons
12.000 Hm'/hr Acid gases, sulfur species, hydrocarbons
.8 m'/hr Acid gases, sulfur species, hydrocarbons
.13 HT/hr Volatile components which can escape with storage
tank vent gases
14,500 Hm'/hr Acid gases, sulfur species, hydrocarbons
200 m'/hr Acid gases, sulfur species, hydrocarbons
Continued - Hext page
-------
TABLE 1 (continued)
oo
en
Stream
Number Stream Description
Phenosolvaa - See Figure 7
14.0 Phenosolvan inlet water
U.I Cyclone (Cl) vent
14.2 Phenolic water tank (T2) vent
14.3 Unclean oil tank (T3) vent
14.4 Filtered water tank (T5) vent
14.5 Degasing column (C7) vent
14.6 NH} stripper cooler (E2S) vent
14.7 2nd degasing column (C9) vent
14.8 Slop tank (T10) vent
14.9 Phenol storage tank (T24) vent
14.10 DIPE tank (T22) vent
14.11 Treated waatewater
14.12 NH, absorber (C26) vent
14.13 NH, storage tank (T27) vent
14.14 NH«OH product to storage
14.15 Unclean oil to storage
14.16 Raw phenols to storage
By-Product Storage - See Figure 8
15.1 A/B/Cd Tar tank vent
15.2 A/B/Cd Medium oil tank vent
15.3 A/B/Cd Gasoline tank vent
15.4 A/B/Cd Raw phenol tank vent
15.5 A/B/Cd Unclean oil tank vent
15.6 A/B/Cd NIK OH tank vent
19.1 Cooling cower vent gases
20.1 Waste gases to Hare
(3.6 + 7.1 + 13.6)
Stream
Type
L
G
G
C
G
G
G
G
C
G
G
L
G
G
L
I
L
G/L/S
G/L/S
G/L/S
G/L/S
G/L/S
G/L/S
G
G
Estimated
Flow Rate" Comments on Components of Environmental Concern
13.1 m'/hr Comprehensive characterization desired
2 Nm'/hrt
7 \
7
7
9 Nm'/hr
4 Nm'/hr
.4 Nm'/hr
7
Volatile orgaoics/lnorganlcs; particularly acid
gases, sulfur species, hydrocarbons
.08 Nm'/hr0.
.5 Nm'/hr Ether vapors, other volatile organics
13 m'/hr Comprehensive characterization desired
\ Volatile organics /inorganics, acid gases, NHi
7 }
.2 MT/hr NH, + other volatile species (acid gases, organics)
.03 MT/hr Volatile organics
.09 MT/hr Comprehensive characterization desired
.5 Nm$/hrc
.25 Nm'/hrc
.13 Nm'/hrc
.08 Nm'/hrc
.03 Nm'/hrc
.2 Nm'/hrc
k
Volatile species present ID all by-product streams
7 Volatile species resulting from process leaks Into the
circulating cooling water system
2,900 Nm'/hr Behavior of hazardous species in flare
*Flow data normalized to a one-gaeifler-in-service basis.
Process gas flow only; does not consider the steam which Is present.
'T'ank vent flows assumed equal to the volume displaced by normal process stream flow.
A - vent gas; B - liquid in tank; C - sludge in bottom of tank.
-------
riM
Mtt.r «•*
Ir.ttM.t • rlu. S»
«•»«"«« 1 ». P. ,e»l
y*»«r *| $!•.•• Vi/Vi/viJtL/ Vl/^i/%3'' I »r !».-*» 4 A.
w«r » ,|Mt
M.« *"•
11
SI.
••« 1 »
Ca.l Br»l.«/Sll
(Llfllte) ^ ClMilflc.t
St... 1
0>«nli«4
Coal
Hu to-
il tor
£? Attlttm
QO i
" 1
(•!*>)
— »>Co»
». P. VMt
*"" Co.^cnsit. Ui^* lick M.S
. . tntt "•««""" T ur. »,.„ ,tlA ^ ^
cH*^«« OIH 1 1 on
»" - P.!.. -J r* ..tpt- —t'tor
L I - __. , . Ttn, Huvy Olli 1 PkMltc MiUr *"
r_» H, ^J Mi „, |f|i
P.rlllc.tlo. S»»tfc«l»
c.'
•> C.I (•) . > .
OI,trik.tl.V -»K"."
St.tlOU |ty C<>
it. on •
«•«•««•• .„.. ,„..
to At..
' 1
o ;;» : . ._^ © -»«.......
S.Mrr.tlo. 0!U_ 'sto'rlr; — ^""l
| AtkQMKk | T
•{urn, *••" u.t '»•;•"« Tb««*r
Syitkttlt »«» "»ttr ^.j,
U At..
, t
© ^Ir«t- Mmr— » B.t.r ® , ,., CMlUf ©* ^to/fr!!!"'' *••"•* ••
*«*• A..ltl>*> »• "•"•••t ""r jJiJi. ^^r^r^fX SU" *
11 1 \jr*jr*^ty
llw Ututmm
Slmtfi Sl« Pk.no Is
Mli »
Pk«.ol«
AMItlwi
M.t.r @
""' ., f.,0., ^ ««««
x.l.r I
RADIAN oottt>oM*TioM
• O. M> OOM/Wltl.. IfUVIWM
FIGURE 1
OVERALL PL AMT fiOW SCttE/ȣ
KOS010 iUK(.l C,ASIFICAT/ON PLANT
tm coot nen HO c
mime m.
10-1 21 'f -3
KAU «EA ] ISMOI; Of/
Figure 1. Overall plant flow scheme for Kosovo Lurgi gasification plant.
-------
WATER-
TO OTHER
ISTEAM IN-PLANT
USES
LIGNITE
oo
TO NHa
SYNTHESIS
MEDIUM
BTU GAS
BY-PRODUCT
TARS. OILS.
GASOLINE.
PHENOLS. NH3
PLANT SECTIONS WHICH ARE
BEING STUDIED IN THE
KOSOVO TEST PROGRAM
PLANT SECTIONS NOT BEING STUDIED
PLANT SECTION NOT IN SERVICE
PLANT SECTION IS NOT A DIRECT SOURCE OF
MAJOR PROCESS. WASTE OR BY-PRODUCT
STREAMS REQUIRING CHARACTERIZATION
(EXCLUDES CONSIDERATION OF FUGITIVE EMISSIONS)
PLANT SECTION IS A SIGNIFICANT WASTE STREAM
SOURCE: HOWEVER. ADEQUATE CHARACTERIZATION
DATA ALREADY EXISTS FOR THIS TYPE OF SOURCE
Figure 2. Simplified flow schematic: Kosovo gasification complex.
-------
"WET" RUN OF
MINE COAL t
-••TO BAGHOUSE
AUTOCLAVE
VENT
1.2)
EAM p<3 ^
n. )
1
I
AUTO-
CLAVE
V*-
PRESS.
10 atm.)
1 AUTOCLAVES
P-X
r^^t \
PQl INTERME
1 ,
PRESSURE (
j M •* STEAM TO
AUTO'*' Aur<
-1- nU 1 V
Jl«U/4V t.
COAL
^-v FLASH '
"N. DRUM CONDENSATE
^V TANK
i ^"- \^^
DRIED
COAL
BUNKER
^ ^-*-
~^
DRIEC
1
i
t ,
VENT
^«^^
—/I 4
^T^I * •*
_ COHnPN5ATF
TANK
\^x
COAL C'Vl
TO SIZING OPERATION ""•"
FLEISSNER
CONDENSATE
70-1467-1
Figure 3. Process flow diagram showing sampling points
in Kosovo plant Flelssner drying section.
188
-------
oo
<£>
COAL
WA1LR + OUST
(S.JV-
GASIFIER
70 WASfC GAS
RAW GAS TO
4t t
CKE
fATEt
^
M
•v
\
OCI
r
u
in
/
/
<
'
ft
\
AH
1
(
J
B>-
• ^
AH
11
I
f
MU
s
1
@h
res. —
T
i — -fii
(cm
V"
coNoe
~|
^
/
HSAT£
a
COOLING SeCTIOM
TO TAR SEPARATIO*
»CT AM
POINT
TO
MATtR
Figure 4. Process flow diagram showing sampling points in Kosovo plant generator section.
-------
RELtASED
WATER
ioPHOun\piSS~
FROM RECTISOL
COOLING WATER
HVY JAR ( OUST
TO PUMP
Figure 5. Process flow diagram showing sampling points in Kosovo plant tar separation section.
-------
S G»S
KICH WASTE C4S
G4S
6AS TO
/CINERATOR
TO VENT
-^ CLEAN
HETIIAHI.
*fe
»
-------
PHENOLIC
WATER
STEAM
l-'NL/.E'AN OIL
~ro r,
...
70 TAR
SEPARATION
<£>
iVAS'ff
\*'A7f~r. 7~<3
e/o
j«
^ '
»J
Irl
ff/vr
L^JY
O/PE + PHENOL
i
-<>4-7
CVAPORAltD 0/flE
(I « . D
17
VENT
RAH
STORAGE
I
^-0_^_D
Figure 7. Process flow diagram showing sampling points in Kosovo plant phenosolvan section.
-------
CO
CO
TO POWER
STATION
LOADING
70-1466-1
LEGEND
A - VENT GAS
B - LIQUID IN TANK
C - SLUDGE
Figure 8. Process flow diagram showing sampling points in Kosovo plant byproducts storage area.
-------
TABLE 2. KOSOVO PLANT: MAJOR AIR EMISSION SOURCES
Approximate Flow
Per Ga«ifiar*
(Nm'/hr)
Disposition
Studied in
Phase I
To Be Studied
in Phase II
1. Rectisol Process
HzS Rich Gas (7.1) [2500]
C02 Rich Gas (7.2) 5000
2. Fleissner Lignite Drying Process
Autoclave Vent (1.2) Unknown
Condensate Tank Vent (1.4) Unknown
3. Coal Feeding System (Lock Hopper) Vents
High Pressure (3.o) 400
Low Pressure (3.2) 40
4. Generator Startup Gases (3.3) •
5. Gas Cooling/Tar Separation Section
Flash Gases
Flare
Vent
Vent
Vent
Flare
Vent
Vent
x
x
X
X
6.
7.
3.
9.
High Pressure Expansion Gases (13.6)
Tar/Medium Oil Surge Tank Vents
(13.1-13.4)
Condensate Surge Tank Vents
(3.4; 13.5; 13.7)
Incinerator (20.1)
Phenosolvan Condensate Strippers
(Primarily 14.5)
By-Product Storage Tank Vents (15.1-15.6)
Air/Oxygen-Rich Vents
Ash Lock Vent (3.5)
Coal Bunker Vents - Fleiasner (1.1)
- Gasification (2.2)
[30]
50
40
[2900]
400
1
30
Unknown
4000
Flare
Vent
Vent
Flare
Vent
Vent
Vent
Vent
Vent
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Note: Data points in brackets are derived from plant design data. Other data, points are measured values.
A reasonable basis for establishing priorities
not only for the air emission streams studied in
the Phase I program but also for the individual
components present in those streams is pro-
vided by EPA's source analysis model (SAM).
This analysis tool, which was developed under
EPA contract by Acurex Corporation,1 provides
a convenient format for assessing the potential
environmental problems associated with partic-
ular emission streams. The SAM analysis ap-
proach relies heavily on health-effect related
multiple acute toxicity effluent (MATE) values
that are listed for a variety of organic and in-
organic compounds in Reference 2.
The health-effect related MATE values for
the specific gaseous species measured in the
Phase I test program are listed in Table 3. An
inspection of Table 3 shows that the most toxic
ambient pollutants addressed in the Phase I
test program were benzene and methyl and
ethyl mercaptans. MATE values can be used to
establish priorities for emission stream control
by making the following calculations:
POOH, - Potential degree of hazard for
component i
_ Measured concentration of pollutant i .
" MATE value for pollutant i
194
-------
PTUDR - Potential toxic unit discharge rate
All
components
F x L
where F - stream flow rate.
Because of the nature of these calculations,
PDOHs are useful in establishing priorities for
the components present within specific streams,
and PTUORs are useful in comparing one
stream to another.
Potential degree of hazard calculations for
the high-priority emission streams that were
discussed above are shown in Table 4. The gas
stream analytical data used to support these
calculations are provided in Table 5.
By examining the data presented in Table 4,
the following conclusions can be drawn:
• With respect to the fixed gases analyzed, CO
appears to be the most significant pollutant.
TABLE 3. KOSOVO GASES AND THEIR RESPECTIVE MATE VALUES
(AIR-HEALTH)
Component
(yg/m3)
Fixed Gases
H2
02
CO
C02
NA
NA
3.3 x 10s
4.0 x 10"
9.0 x 10s
Hydrocarbons
C2's
C3's
-------
TABLE 4. CALCULATED POOH AND PTUDR VALUES FOR "HIGH-PRIORITY" AIR EMISSION
STREAMS SAMPLED AT KOSOVO DURING THE PHASE I TEST PROGRAM
3.2
3.6
Lock Hopper Vent Gases
Low
Fixed Cases
Oj
H.
CN,
CO
CO,
Hydrocarbons
c,
c,
c.
C5
c.
•SOMOS
Tolusna
Sulfur Species
M,S
COS
CHiSH
CjHsSH
Other
HHi
HCM
POOH (Stream)*
Stream Flow Rate (Na'/br)
PTUDR (Ita'/hr)
All
Components
* __
POOH (Streasj) - V i
Pressure
:
-
200
2900
91
1.6
0.7
1.9
0.9
3.2
2300
-
71
1.1
590
250
290
5.3
6700
40
2.7ES
CPDOH).
High Pressure
-
-
240
4100
80
1.6
0.4
0.4
0.4
1.1
-
-
110
1.9
910
750
NF
19
6200
400
2.5E6
13.1
Tar
Tank
Vent
-
-
2.2
MF
7.0
TR
TR
TR
TR
TR
4700
1.1
190
NF
1400
690
110
14
7100
2
1.4E4
13.3
Medina
Oil
Tank Vent
:
-
170
NF
190
1.3
0.4
3.7
1.8
11.0
5800
22
1300
<2.5
2200
1300
-
-
11,000
50
5.5E5
13.6
Tar
Separation
Expn. Cases
-
-
130
2300
160
0.9
0.7
5.6
3.6
'll.O
12,000
43
1200
-
2100
810
830
7.5
20.000
26
5.1E5
13.7
Phen. H20
Tank Vent
-
-
4.3
MF
63
TR
TR
TR
TR
TR
19,000
65
190
MF
1500
1200
510
4.2
23.000
40
9.0E5
14.5
Stripper
Vent
-
-
TR
NF
72
TR
TR
TR
TR
HF
HP
NF
760
NF
330
83
2900
130
4300
400
1.7E6
7.1
HjS Vent
-
-
91
810
190
0.7
0.4
1.9
1.8
1.1
-
-
2300
3.5
9300
2100
94
9.1
15.000
2500
3.7E7
7.2
CO] Vent
-
-
20
NF
200
0.7
0.7
TR
TR
NF
-
-
0.5
NF
18
9.7
0.2
1.4
250
5000
1.3E6
7.3
Rectlsol
Inlet Gas
-
-
280
4100
72
1.5
0.9
3.7
0.7
2.1
700
-
470
0.5
1200
280
0.2
6.6
7100
-
l-l
-------
TABLE 5. GAS STREAM ANALYTICAL DATA FOR "HIGH-PRIORITY" PHASE I EMISSION STREAMS
3.2 3.6
Lock Hopper Vent Gases
CoBpound
Fixed Cases (Vol. X)
HI
0,
NI
CH»
CO
CO,
Hydrocarbons (Vol. Z)
c,
C,
C,
C5
C.
Isniana
Toluene
Sulfur Species (pp*)
"»s
COS
CH,SH
CjlUSH
Others (g/100 MB')
KM,
HCM
Low Pressure
34.0
0.7
2.5
9.4
9.3
42.0
0.7
0.3
0.1
0.05
0.03
0.2
-
700
170
270
90
530
5.8
High Pressure
32.0
0.2
6.1
11.0
13.0
37.0
0.7
0.2
0.02
0.02
0.01
-
-
1100
300
420
270
NF
21
13.1
Tar
Tank
Vent
TR
21.0
76.0
0.1
NF
3.2
TR
TR
TR
TR
TR
0.4
0.01
1900
NF
630
250
198
15.3
13.3
Madlim
Oil
Tank Vent
.
NF
0.9
3.4
7.6
NF
86.0
0.6
0.2
0.2
0.1
0.1
0.5
0.2
13.000
<400
1000
480
-
-
13.6
Tar
Separation
Expn. Cases
11.0
0.5
0.6
6.1
7.2
72.0
0.4
0.3
0.3
0.2
0.1
1.0
0.4
12,000
-
950
290
1500
8.2
13.7
Phen. H20
Tank Vent
T»
13.0
53.0
0.2
NF
29.0
TR
TR
TR
TR
TR
1.6
0.6
1900
NF
680
420
920
4.6'
14.5
Stripper
Vent
NF
9.0
58.0
TR
NF
32.0
TR
TR
TR
TR
•F
NT
*
7500
NF
150
30
5300
140
7.1
H2S Vent
-
0.5
1.4
4.2
2.6
86.0
0.3
0.2
0.1
0.1
0.01
-
-
23.000
<560
4300
740
170
10
7.2
COj Vent
0.8
0.1
0.3
0.9
NF
94.0
0.3
0.3
TR
TR
NF
-
-
4.6
O.S
8.5
3.5
0.4
1.5
7.3
Rectlaol
Inlet Gas
36.1
0.6
1.6
13.0
13.0
33.0
0.7
0.4
0.2
0.04
0.02
0.06
-
4700
80
570
100
0.3
7.3
Data froa Canpalgn Three Test; Moveaber 1978
-------
• Negligible CO levels exist in several flash
gas streams, apparently because of the low
solubility of CO in both condensed organic
and aqueous liquids.
• With respect to the light hydrocarbons ana-
lyzed, benzene appears to be by far the most
significant source of environmental concern.
• With respect to the sulfur species analyzed,
mercaptan levels appear to be at least as
great a source of concern as H2S.
• Between the two nitrogen species analyzed,
NH3 appears to be more of a problem than
HCN.
While the data presented in Table 4 show
some interesting trends, several factors limit
the extent to which these data can be used to
draw firm conclusions about the Phase I test re-
sults. Some of these factors are:
• The data presented in Table 5 are single-
point measurements, not necessarily repre-
sentative of either normal plant operation or
the range of operating conditions likely to be
encountered.
• Some degree of judgment is involved in se-
lecting specific levels for most MATE val-
ues.
• Inherent inaccuracies exist in the sam-
pling/analytical procedures used to gather
Phase I test data.
• Other components (those not measured in
Phase I) may have more of an impact on final
PDOH and PTUDR values than any of the
components measured thus far.
Generally, the Kosovo Phase I test data pro-
vide a reasonable definition of the scope and
magnitude of the air emission problems that
will have to be addressed in a U.S. Lurgi plant.
These results also justify continued testing at
Kosovo.
Some of the data needs indicated from Phase
I test results are outlined below. Most of these
needs will be addressed in subsequent phases of
work at Kosovo.
• Levels of other hazardous components such
as trace elements and trace organics en-
countered in key emission streams.
• Rectisol process performance information
(this unit is the source of two key streams):
• Further characterization of the H2S-rich
gas stream is desired to assess its suitabil-
ity for feed to an elemental sulfur recov-
ery unit. More specifically, levels of prob-
lem components such as mercaptans,
COS, hydrocarbons, and C02 (effectively a
diluent) should be monitored as functions
of time.
• Further characterization of the C02-rich
vent gas stream is necessary to confirm
that this stream can be safely vented to
the atmosphere (as proposed in several
U.S. designs). Possible sources of prob-
lems with this approach should be identi-
fied.
• Generator startup gases: Significance rela-
tive to the other key emission streams needs
to be addressed. Although not addressed in
the Phase I program, this effort will be initi-
ated as part of the Phase II program.
• Fates of hazardous gas stream components
in a combustion process: In a U.S. gasifica-
tion facility, most of the emission streams
identified in this paper would be collected
and either recycled back into the process
gas stream (unlikely because of the compres-
sion requirements involved) or incinerated
(for example, in the firebox of onsite stream
generators). If an incineration approach is
used, the fates of hazardous species present
in those streams needs to be assessed. Cur-
rently, no plans for making this type of
measurement are incorporated into the
Phase II test program.
• Fugitive emissions: A program to charac-
terize the fugitive emissions from the
Kosovo plant is now being discussed with
the Yugoslavs.
Liquid Effluents, Liquid Byproducts, and
Solid Wastes
Generally, the liquid and solid wastes pro-
duced in the Kosovo facility did not receive the
same level of attention that the air emissions
did in the Phase I test program. Considerable
useful data concerning these streams was ga-
thered, however.
The major Kosovo plant liquid effluent, liquid
byproduct, and solid waste streams are summa-
rized in Table 6. This table also indicates which
streams are being studied in the Phase I and
Phase II test programs.
The major aqueous waste stream at Kosovo is
the phenosolvan effluent water stream. Accord-
ing to the plant design, this stream was to be
treated in a biological oxidation process, but
currently this system is not in operation.
198
-------
TABLE 6. KOSOVO LURGI GASIFICATION PLANT-MAJOR SOURCES OF LIQUID EFFLUENTS,
LIQUID BYPRODUCTS, AND SOLID WASTES
Aqueous Wastes
Phenosolvan Effluent
Fleissner Condensate
Generator Section Wastewater
Liquid By-Products
Tars, Oils, Gasoline
Phenols
NHi,OH
Solid Wastes
Gasifier Ash
Heavy Tar & Dust
Other Process Residues
Approximate
Flow*
13 MT/hr
Unknown
3 MT/hr
.8 MT/hr
.1 MT/hr
.2 MT/hr
Studied in To Be Studied in
Phase 1 Phase 11
X X
X
X X
X X
X X
X X
(By-Product Stor-
x age Residues)
*
Design values; normalized to a one gasifier in service basis.
-------
Preliminary data obtained from a series of
source screening samples indicate that the
Kosovo plant's phenosolvan unit is effective in
recovering the phenols present in the raw proc-
ess gas liquor. However, as the data in Table 7
indicate, the organic loading in the effluent
water from the phenosolvan unit is still substan-
tial. The indicated phenol concentration is not
sufficient to account for the COD figures that
were obtained for that stream. As a result, it
can be anticipated that the organic characteriza-
tion work to be done in Phase II will shed con-
siderable light upon the nature of the environ-
mental hazards and control needs associated
with that particular stream.
As shown in Figure 3, the generator section
wastewater stream is a composite stream. It
consists primarily of ash quench water. How-
ever, small quantities of coal bunker and ash
lock vent gas scrubber blowdown liquid are also
discharged via this stream. This stream has a
relatively high pH because of the highly alkaline
nature of the Kosovo ash.
Very little characterization data on the
nature of the Kosovo plant liquid byproduct
streams were gathered during the Phase I test
period. What little data were gathered are pre-
sented in Table 8. One of the major points to be
noted here is that the sulfur contents of the liq-
uid byproducts become progressively higher
with a move from "heavies" to "lights." These
data indicate that heavy hydrocarbon byprod-
ucts similar to those generated at Kosovo could
be used to satisfy onsite fuel needs in the
United States without causing serious S02
emission control problems.
As mentioned previously, the bulk of the
work to characterize the liquid and solid wastes
associated with the Kosovo plant will be per-
formed as part of the Phase II program. Some of
the concerns in this area include:
• Trace and minor components present in all
significant liquid and solid waste streams
will be quantified. Of particular concern are
the leachable species present in the solid
waste streams and the soluble components
found in the aqueous wastes. /
• One of the most practical disposal options
for the liquid hydrocarbon byproducts is to
use these materials to satisfy onsite fuel
needs. The fates of hazardous species pres-
ent in those streams in a combustion process
could cause concern, although no specific
plans to study this problem have been made.
TABLE 7. KOSOVO WASTEWATER PROPERTIES (PHASE I DATA)
Phenosolvan
Effluent Water
Generator Section
Wastewater
Units
PH
Susp. Solids
Diss. Solids
COD (K2Cr207)
9.2-9.4
150-190
880-1300
3100-3300
11.4-12.1
180-590
1100-2100
.8-150
mg/H
mg/fc
mg/2,
mg
Phenols
CN~
Cl~
SO**
CNS"
F~
N03~
170-270
.02
16-120
100-110
3
Trace
11-12
.01 Max.
20-70
320-670
.01-.03
.6-1.2
4-6
mg/fc
ing/*
wg/i
mg/S,
ng/i
200
-------
TABLE 8. KOSOVO LURGI GASIFICATION PLANT
LIQUID BYPRODUCT DATA
c
H
N
S
Ash
02
HV(ir)
ng S02
J
Feed Coal
(Dry)
1.1
21.6
510
Heavy Tar
+ Dust
56.0
7.6
0.87
0.33
6.6
28.6
26.5
120
Tar
81.9
8.4
1.3
0.49
0.22
7.8
37.3
130
Medium
Oil
81.2
8.9
1.0
0.71
0.03
8.2
38.3
190
Gasoline
85.7
9.8
0.2
2.2
-
2.1
41.6
530
S02 Emission Limitations
Solid Fuels 86-516 ng/J (0.2-1.2 lb/106 Btu)
Liquid Fuels 344 (0.8 lb/106 Btu)
Assuming 100% conversion of S to SO2
REFERENCES
1. Schalit, L. M., and K. J. Wolfe. SAM/IA: A
Rapid Screening Method for Environmental
Assessment of Fossil Energy Process Ef-
fluents. Acurex Corp./Energy and Environ-
mental Division. Mountain View, Calif. EPA
Contract Number 600/7-78-015 (NTIS Num-
ber PB 277-088). February 1978.
2. Cleland, J. G., and G. L. Kingsbury. Multime-
dia Environmental Goals for Environmental
Assessment, Volumes landII(final report).
Research Triangle Institute. Research Tri-
angle Park, N.C. Report Number EPA-600-7-
77-136a, b, NTIS Number PB 276-919 (Vol-
ume I), PB 276-920 (Volume H). EPA Con-
tract Number 68-02-2612. November 1977.
201
-------
ENVIRONMENTAL ASSESSMENT REPORT:
HIGH-Btu GASIFICATION TECHNOLOGY
M. Ghassemi, K. Crawford, S. Quinlivan, and D. Strehler*
Environmental Engineering Division, TRW, Redondo Beach, California
Abstract
As part of a comprehensive program for the
environmental assessment ofhigh-Btu gasifica-
tion technology, the available data on high-Btu
gasification and associated operations and proc-
esses have been analyzed, and gaps in the ex-
isting data base have been identified This paper
describes the data analysis methodology and
identifies limitations of the available data. The
program was sponsored by the Fuel Process
Branch of the U.S. Environmental Protection
Agency's Industrial Environmental Research
Laboratory (EPA IBRD, Research Triangle
Park, N.C.
BACKGROUND
As part of its 3-yr program sponsored by the
U.S. Environmental Protection Agency (EPA)
for environmental assessment of high-Btu coal
gasification, TRW has recently completed a
three-volume document entitled Environmental
Assessment Data Base for High-Btu Gasifica-
tion (report number EPA-800/7-78-186a, b, and
c). The document represents the summary and
analysis of the existing data base and includes
identification of data gaps.
The preparation of the data base document
drew information from several sources, in-
cluding published and unpublished EPA docu-
ments, open literature, process developers and
EPA/DOE contractors, and authorities in in-
dustry and academic institutions. Gasification
and related processes judged to have the great-
est likelihood of being employed in commercial
SNG facilities are discussed in the data base
document.
DATA BASE METHODOLOGY
To facilitate systematic analysis, the tech-
nologies for high-Btu coal gasification were
•Speaker.
divided into four "operations" (Figure 1). They
include coal preparation, gasification, gas
purification, and gas upgrading. In addition, the
auxiliary processes to be used in commercial
SNG facilities for pollution control were
grouped into air pollution control processes,
water pollution control processes, and solid
waste management processes.
For analysis, the operations and auxiliary
processes were further subdivided into mod-
ules, each module comprised of nearly inter-
changeable processes or processes applicable to
different operating conditions and input re-
quirements.
For each process within a module, a data
sheet was prepared with key information items,
thereby imparting high visibility to engineering
facts and figures, allowing ready comparison
between alternate processes in a given module,
and underlining specific areas where significant
gaps existed in the available data. Represen-
tative data sheets for the dry-ash Lurgi gasifica-
tion process and the Rectisol acid-gas removal
process (single-adsorption mode) are contained
in Appendixes A and B, respectively. Data
sheets were prepared for 11 gasification proc-
esses, 22 gas purification processes, 4 gas
upgrading processes, 18 air pollution control
processes, 17 water pollution control processes,
and 3 solid waste disposal processes.
TECHNICAL DISCUSSION
The 11 gasification processes that were in-
vestigated are presented in Table 1. These proc-
esses use five different types of gasifier de-
signs, as shown in the table. Data contained in
the gasification data sheets are summarized in
Table 2. Typical data include: developmental
status, coal feed and pretreatment, coal feeding
method, gasifier design, gasifier temperature
and pressure, quench and dust removal, ash/char
removal, typical product gas composition, tar/oil
production, and gas yield. As can be seen in
Table 2, there is a wide range of gas composi-
203
-------
> COAL PREPARATION OPERATION
GASIFICATION OPERATION •
GAS PURIFICATION OPERATION
-GAS UPGRADING
OPE RATION
a am t-rrt or COAL AMD PIAMT
MOUCNCC Of racrAMATO* ITVt MAY M
M »«MH AMOVE M t. Qfl> MG
IHEDHDiOMV ASH'
EDWD 1SL*GGM«G>
UIDlZf DIED INTERNAL CHAR GASIFlCATiO**
UIOlZEOtEO EXTERNAL CHAR GASif ICAIiON
MTRAMMDMD SLAGGING
Figure 1. High-Btu gasification operations and process modules.
-------
TABLE 1. GASIFICATION PROCESSES EVALUATED
Lurgi (dry ash)
Lurgl Slagging Gasifier
Hygas (steam-oxygen)
Cogas
C02-Acceptor
Hydrane (Hydrogasification)
Synthane
Self-Agglomerating Ash
Bigas
Koppers-Totzek
Texaco
Fixed bed (dry ash)
Fixed bed (slagging)
Fluidized bed (internal char gasification)
Fluidized bed (external char gasification)
Entrained bed (slagging)
tions and yields from the gasifiers investigated.
The Hydrane gasifier, for instance, produces 57
to 79 percent (volume) methane, while the Lurgi
(dry ash) only produces 8 to 11 percent methane.
Table 3 presents a matrix of the advantages
and disadvantages of the 11 gasifiers. These
characteristics are based on operational charac-
teristics, waste streams, and utility require-
ments.
Gas purification processes are employed to
remove acid gases from the raw product gas to
prevent methanation catalyst poisoning and to
produce a product with a heating value equiva-
lent to that of natural gas. Processes were in-
vestigated that remove H^ and C02 simulta-
neously or selectively. Three types of acid-gas
removal processes were included in the in-
vestigation: hot gas H2S removal, solvent proc-
esses for acid-gas removal, and methanation
guards. The solvent processes are most com-
mon, being extensively employed by petroleum
refineries. Table 4 presents key features of the
solvent processes included in the data base doc-
ument. Listed in the table are the solvents em-
ployed by each process, operating pressure, se-
lectivity, component distribution, solvent
losses, and utility requirements.
The gas upgrading operation generally in-
cludes a shift conversion step, an acid-gas
removal step, and a methanation and drying
step. The data base for both shift conversion
and methanation steps is limited by the lack of
commercial-scale facilities or operating ex-
perience with these processes.
The air pollution control section reviews the
sources and characteristics of gaseous waste
streams associated with:
• The gasification, gas purification, and gas
upgrading operations;
• Water pollution control and solid waste
management; and
• Other auxiliary processes unique to the
operation of commercial high-Btu gasifica-
tion facilities.
Processes that have been used for or that may
apply to the control of gaseous emissions in gas-
ification facilities are reviewed. Alternative
control strategies for integrated facilities are
205
-------
TABLE 2. KEY FEATURES OF HIGH-Btu GASIFICATION PROCESSES
Process
Lurgi (dry ash)
Lurgi (Slagging
Gasifier)
Hygas
(steam-oxygen)
Cogas
C0,,-Acceptor
i
Synthane
Bigas
Hydra ne
Development
Status
Commercial for
fuel and syn-
thesis gas
production
Pilot scale.
demonstration
plant under
design
Pilot scale;
demonstration
plant under
design
Pilot scale)
demonstration
plant under
design
Pilot scale;
no demon-
stration or
commercial
project
planned
Pilot scale
Pilot scale
Bench scale
Coal Feed and
Pre treatment
Limited to non-
caking coals.
Fine coal sizes
must be
briquetted
Limited to non-
caking coals.
Fine coal -sizes
may be utilized
by injection
into center of
gasifler bed
Can use all
domestic coals.
Caking coals
are pretreated
with air and
steam in
fluidi zed bed
at 31i-400°K
Can use all
domestic coals.
Pretreatment
for caking
coals is
accomplished in
first stage
pyrolyzer
Limited to more
reactive coals
(e.g. , lignite
and sub-bitum-
inous coal)
Can use all
domestic coals.
Caking coals
are pretreated
with Oj and
steam within
the gasifier
in a free fall
fluidized bed
zone
Can use all
domestic coals.
Ho pretreat-
ment is
required
Caking coal
permitted with-
out pretreat-
ment.
Coal
Feeding
Method
Pressurized lock-
hopper
Pressurized lock-
hopper
Coal Is slurried
with light
aromatic oil and
charged to gasi-
ficr by high
pressure slurry
pump
Pneumatic feed-
ing with recycle
product g.is
Pressurized lock-
hopper
Pressurized lock-
hopper
Coal Is slurried
with water and
injected into
pressurized drier
before entering
gasifier
Injection nozzle
Gasifier Design
Fixed bed, counter-current
gas/solids flow, tempera-
ture increases downward to
effect pyrolysis and
gasification
Same as dry ash Lurgi
Two stage, fluidized bed
hydrogasification.
Fluidized steam-oxygen
gasification stage pro-
vides heat and gas for
hydrogasificatioh
Coal Is pyrolyzed in four
fluidized stages with
progressively higher
temperatures. Char pro-
duced from pyrolysis of
coal is sent to gasifier.
Crude gas is produced
from the reaction of char
and steam, obtaining heat
indirectly from the com-
bustion of char with air.
Gasifier gas flow counter-
current to coal and char
In tiie gasifier, calcined
dolomite supplies heat for
stean gasification of
coal. Carbonated dolo-
mite Is recalcined In a
regenerator by burning
char with air. Both
vessels fluidized
Steam and oxygen used
to gasify coal in
fluidized bed gasifier
Coal is gasified in an
entrained bed with a
stsan/synthesis gas
mixture. Char is
gasified in an
entrained bed using
Oz and steam to gener- •
ate synthesis gas
Direct hydrogasification
of coal with hydrogen in
a fluidized bed. Hydro-
gen would be produced by
char gasification v/ith
suD>equer,t purification
Gasifier
Temperature
°K(°f)
Max. bed temp.
1255-1644
(1800-2500)
Max. bed temp.
1255-1644
(1800-2500)
Hydrogasifica-
tion
750-1000
(900-1350)
Steam-oxygen
gasification:
1100 (1600)
Pyrolyzers
500-1000
(450-1500)
Gasifier:
1200 (1700)
Casifier:
1090 (1500)
Regenerator:
1230 (1860)
960-1090
(1280-1500)
Upper stage:
1200 (170)
Lower stage:
1755 (2/30)
-6000 (-1500)
Gasifier
Pressure
MPa(psia)
2.1 - 3.2
(300-465)
0.7 - 3
(95 - 415)
6.2 - 7.1
(911-1040)
0.13 (20)
0.20 (29)
1.0 (150)
1.0 (150)
4.2 - 6.8
(600-1000)
8 (117b)
7.0 (1015)
(continued)
206
-------
TABLE 2 (continued)
Process
lurgl (dry ash)
Lurgi (Slagging
Gaslfler)
Hygas
(steam-oxygen)
Cogas
C02-Acceptor
Synthane
81gas
Hydrane
Quench and
Dust Removal
Water spray cooler
to condense tars/
oils and remove
bulk participates
Sa.ne as dry ash
Lurgl
Cyclone followed
by water quench
for oil and parti -
c ul ate removal
Cyclone followed
by venturl scrub-
ber for removal
of char fines and
for recovery of
011
Internal gaslfler
cyclone, external ,
water spray tower
for particulate
removal
Internal gaslfler
cyclone, venturl
scrubber
Cyclone, water
spray tower for
particulate
removal
No Information
Ash/Char
Removal
Lockhopper
water quench.
water slurry
transport
Lockhopper.
followed by
water quench
of slag
Water quench
at gaslfler
pressure.
water slurry
transport
Slag quenched,
transport not
known
Coal ash
leaves regen-
erator with
flue gas and
Is collected
by cyclone
and scrubbing
systems
Lockhopper,
water quench.
steam trans-
port
Slag quenched
followed by
lockhopper
No Information,
char utiliza-
tion has not
been determined
Typical Product Gas
Composition* (vol 5)
CH4
8-11
5-8
13-28
8-15
14
7-13
5-8
57-79
H2
40
28-30
26-37
5-40
56-59
23-35
32-38
21-28
CO
15-20
57-61
8-10
4-19
15
3-12
15-19
1-6
co2
28-31
3-7
28-35
22-29
9-11
37-64*
21-23
1
Tar/011
Production
Yes
Yes
Yes
Yes
No
Yes'
No
?
Gas Yield*
fon3/kg
(scf/lb) of
Dry Feed Coal
0.9-1.7 (16-30)
2.0-2.1 (34-36)
1.0-1.2 (17-20)
Gas: 0.12-.60
(2-12)
Oil: 0.04-0.2 I/kg
(0.005-0.025 gaTV
Ib) coal
1.35 (23)
1.2-1.5 (20-25)
2.0-4.0 (32-68)'
0.6-1.0 (10-17)
•Based upon data for actual operation for the most advanced stage of development
•*N2 free basfs
* Includes CO., used to pressurize the lockhopper
!W1th "free-fall" node of coal Injection; recent pilot plant runs involvir>j "deeo-bed" injection of coals have
Indicated little tar production
207
-------
TABLE 3. ADVANTAGES AND DISADVANTAGES OF HIGH-Btu
GASIFICATION PROCESSES
Process
Lurgi (dry ash)
Lurgi
•(Slagging Gas i fie
Hygas
(steam-oxygen)
Cogas
CO.-Acceptor
Syn thane
Bigas
Hydrane
>.'
"ioTJ
u %.
i- 0
£ 01
E >
s&
Yes
No
•)
No
No
No
No
No
No
.E
.c
1— 4J
C
•o «
as:
o
O) O
oS:
Yes
Yes
Yes
Yes
Yes
No
No
No
V)
s
(J
^
ID
C
3
No
No
Yes
Yes
No
Yes
Yes
Yes
•M
C
N
10
^
0)
r— C
O
C 10
No
No
Yes
Yes
Yes
Yes
Yes
Yes
S-
O.O
Yes
Yes
Yes
Yes
No
Yes*
No
Yes
+J
3
O.
JZ
I
L-
J—
f O)
01 *J
No
Yes
No
No
No
No
Yes
Yes
|
4->
1 *-
O
Moderate
Low
Moderate
Moderate
Low
High
t
Moderate
High
3
on
|
CJl
c
'^J
1_
QJ
S.
O
Moderate
Moderate
High
Low
Moderate
High
High
High
fl!
t— c
0 0
QJ 4J
QJ N
5=
Yes
Yes
Yes
Yes
Yes
No
Yes
No
^
£
|^
<
c
Q
c
V-
Iligh
High
Low
Low
Low
Low
Moderate
Low
n
§-o
u
ce. ••-
csj cr
O. Ol
Yes
Yes
Yes
Yes
No
Yes
Yes
Yes
c
> Ol
X Ol
oz
Yes
Yes
Yes
No
No
Yes
Yes
7
Comments
Commercial operations not for
high Btu gas production at
present. Basis for several
proposed commercial SNG projects
Extensive tests at a modified
dry ash Lurgi plant. Basis
for a DOE-sponsored demonstra-
tion plant.
Pilot plant has demonstrated
operations with several coals.
High carbon utilization has not
been attained to date. Basis
for DOE-sponsored demonstration
program.
Integrated pyrolysis and gasi-
ficationycombustion operations
not demonstrated. Basis for
DOE-sponsored demonstration
program.
Successful demonstration at
pilot plant stage. High cost
of acceptor is a major obsta-
cle to further demonstration
of process.
High pressure lockhopper
feeding not demonstrated.
Pilot plant has limited steady
state operating time.
Ability to control slag flow at
a pilot plant has not been
demonstrated.
Small scale test only. Char
utilization and hydrogen pro-
duction not tested.
•With "free-fall" node of coal injection; recent pilot plant runs involving "deep-bed" injection of coal have indicated little tar production.
-------
TABLE 4. KEY FEATURES OF SOLVENT PROCESSES FOR ACID-GAS REMOVAL
§
Process Nam
PHYSICAL SOLVENTS
Selexol
Purl so)
CHEMICAL SOLVENTS
Anine Solvents
Sulfiban
0£A
AB1P
Alkazld
Carbonate Solvents
Rcnfield
MIXED SOLVENTS
Sulfinol
Ami so I
WDOll.J'JBOCESSK
Claim rco~
vetrocoke
Stretford
Solvent/Reagent
Dimethyl ether of
polyethylene glycol
»- methyl
Z-pyrro! idone
carbonate
phosphate
MonoeHiarolanlne (MCA)
amine
Diethanolamine
01 1 sopropanolaaf ne
Di«ethyl or diethyl
glycine
Potassfun carbonate
and diethanolamlne
and anine bora lei
CyclotctramethylCTe
sulfonc and dlisopro-
panolacrinc
Nethanol and mono- or
diethanol amine
Potassium carbonate
and arsenate/arsenite
Alkaline retavanadate
and anthraqulnone d1-
sulfonic add
Operating Pressure
(add gas partial
press-jrc)
High
High
Low
Low
Low
Low
Moderate
Moderate
Moderate
Moderate
Moderate
HzS/CO? COj/HC
Gnnri
Good Moderate
Good Moderate
Poor Good
Poor Good
Poor Good
Moderate Good
Moderate excellent
Poor Moderate
Poor Moderate
Good Excellent
Good * Excellent
Component Distribution*
r, „. Higher water
COS CS? DSH NH3 HCH Orqantcs »aoor
a,b a,b a.c.d c,d a.c.d a.b.c.d d
a,b a,b a,d a.d a.c.d a.b.c.d d
e e a,b,d a.d e a.d d.g
a.b a.b a.b.d a.d e a.d d.g
a.b a.b a.b.d a.d e a.d d.g
f.g f.g d.g a.d e a.d d.g
f.y r.g f.t] a.d f.fl.d g g
'.9 f.S f.g «i* f.a.d g g
g g g g e g g
Solvent Losses
(Replacement
Requirement!
Low
Low
High
High
Moderate
Lo«
Low
High
Low
Low
Utility t
Regulreaenti
Low
Low
Very high
Very high
Nigh
High
Moderate
Moderate
federate
•Moderate
Koderate
• a) with acid gas stream after simultaneous CO; and HjS removal
b) with CO; stream after separate CO; and HjS removal
c) with HjS stream after separate CO; and H;S renoval
d) with aqueous or organic liquid phase prior to or integral with process
e) degrades solvent
g) remains with treated gas
'Depends on add gas partial pressure, selective vs. non-selective design, and residual sulfur allowed; rating is for moderate to high pressure application
with «10 ppm residual KjS In treated gas.
^Selectivity good, but high COj lowers H?S absorption rate and requires large systems for efficient «?S removal.
-------
discussed. Table 5 shows the air pollution proc-
esses reviewed according to applicability to
high-Btu gasification and the purpose of each
type of control process. Key features of each
process are compared in the data base docu-
ment. Options for the management of sulfur-
bearing waste gases in integrated facilities are
shown in Table 6. It can be seen that a variety of
acid-gas streams are expected to be present in
an integrated facility and that several options
are available for their handling. An integrated
approach to the handling of acid gases, as well
as of other wastes, will be required when envi-
ronmentally acceptable SNG plants are de-
signed.
Several process and air and water pollution
control modules in an integrated facility would
generate aqueous wastes requiring treatment.
Only those aqueous wastes that are specific to
high-Btu gasification and related facilities were
considered. Table 7 lists aqueous waste streams
associated with the different gasification proc-
esses. Each stream —with possible control
methods —is characterized in the data base.
The sources of solid waste in a gasification
plant include: chars and ashes from gasification
and air pollution control, spent catalysts from
shift conversion and methanation, inorganic
solids and sludges from acid-gas removal and air
and water pollution control, tar and oil sludges,
and biosludges from water pollution control. Of
these, only ash, spent catalysts, and inorganic
solids and sludges would be generated in all gas-
ification facilities. The other types of waste may
or may not be generated, depending on specific
processes chosen. Solid waste management op-
tions included in the data base were: resource
recovery, incineration, soil application, and land
burial/landfilling. In comparison with aqueous
and gaseous wastes (for which some composi-
tion and treatability data are available for cer-
tain streams), the composition of solid wastes
and disposal hazards of such wastes are essen-
tially unknown.
DATA GAPS AND LIMITATIONS
A primary goal of the first phase of the en-
TABLE 5. AIR POLLUTION PROCESSES REVIEWED
Sulfur Recovery
Tail Gas Treatment
S02 Control and/or Recovery
Incineration
CO, Hydrocarbon and Odor
Control
Particulate Control
Compression and Recycling
NO Control
n
Claus, Stretford, Giammarco-Vetrocoke
SCOT, Beavon, IFP-1, IFP-2, Sulfreen,
Cleanair
Wellman-Lord, Chiyoda Thoroughbred 101,
Shell copper oxide, lime/limestone
slurry scrubbing, double alkali, and
magnesium oxide scrubbing
Thermal oxidation, catalytic oxidation,
Thermal oxidation, catalytic oxidation,
activated carbon adsorption
Fabric filter, electrostatic precipita-
tion, venturi scrubbing, cyclones
Compression and recycling
Combustion modification and dry and wet
processes
210
-------
TABLE 6. OPTIONS FOR THE MANAGEMENT OF SULFUR-BEARING WASTE GASES
Waste Gas
Control Options*
Comments
Concentrated Acid Gases
1. Claus plant sulfur recovery
2. Claus plant sulfur recovery and
tail gas incineration
3. Claus plant sulfur recovery and tail
gas treatment
4. Same as 1 plus SOg control and/or
recovery
5. Stretford or G-V sulfur recovery
6. Same as 5 plus tail gas treatment
7. Same as 6 plus incineration
8. Incineration
9. Same as 8 plus SOg control and/or
recovery
0. Incineration, treatment for
control and/or recovery in combi-
nation with flue gases from
utility boilers or char combustion
1. Probably unacceptable because of high concentration of total sulfur
in the tail gas; only applicable to streams containing more than 15% HjS.
2. Probably unacceptable because of high levels of SO? in the tail gas; only
applicable to streams containing more than 15% H2$.
3. Tail gas treatment not highly effective when feed gases contain high levels
of CO;; only applicable to streams containing more than 15% H?S.
4. Reasonable option when feed gases contain more than 15% H2S; total sulfur
removal efficiency may be less than option 5.
5. Inapplicable to waste gases containing high levels of HjS; may not be
economical for gases containing high CO? levels; discharge may contain
high COS and HC levels.
6. Same as for Option 5.
7. Same as for Option 5 except for oxidation of CO and HC compounds
8. Unacceptable because of high S02 emissions.
9. Many S02 recovery processes generate sludges requiring disposal;no by-product
sulfur is recovered; regenerable SO^ removal processes must be operated in
conjunction with sulfur recovery units.
10. Same as for Option 9; some economy of scale may be realized if flue gas
desulfurization is required on utility boilers.
Depressurization
and Stripping
Gases
1. Combining with concentrated *cid
gas streams and use of any of the
treatment options listed above
2. Compression and addition to product
gas stream
3. Use as fuel
4. Incineration
5. Same as 4 plus $03 control and/or
recovery
1. See individual options above; may have considerable dilution effect on the
concentrated acid gas streams.
2. Permits material recovery; some energy input required for compression.
3. Stripping gases may have limited fuel value; may have high SOj emissions.
4. High levels of SO? emissions.
5. See comments for Options 9 and 10 for Concentrated Acid Gases.
Pretreatment
Off-Gases
1. Combining with product gas
2. Injection into gasifier
3. Use as fuel
4. Incineration
5. Same as 4 plus $02 control and/or
recovery
1. Product gas dilution and energy requirement for compression; permits
material and energy recovery.
2. Permits material and energy recovery; will require gasifier design modifi-
cation and energy Input for compression.
3. Nay have high SO? emissions.
4. See comment for Option 4, Depressurization and Stripping Gases.
5. See comment for Option 5, Depressurization and Stripping Sases.
Lockhopper Vent
Gases
1. Compression and recycling
2. Incineration
3. Same as 2 plus SOj control and/or
recovery
4. Use as fuel
1. See comment for Option 2, Pretreatment Off-Oases.
2. See comment for Option 4, Depressurization and Stripping Gases
3. See comments for Options 9 and 10, Concentrated Acid Gases.
4. See comment for Option 3, Depressurization and Stripping Gases.
Catalyst Regeneration/
Decomissionlng
Off-Gases
1. Incineration
2. Same as 1 plus $03 control and/or
recovery
1. See comment for Option 4, Depressurization and Stripping Gases.
2. See consents for Options 9 and 10, Concentrated Acid Gases.
Char Combustion,
Incineration and
Treatment Gases
1. Incineration (for transient gases)
2. Sane as 1 plus SO- control and/or
recovery
1. See comment for Option 4, Depressurization and Stripping Gases.
2. See comments for Options 9 and 10, Concentrated Acid Gases.
•Except where gas compression and recycling Is used, all options culminate in discharge of the treated gas to the atmosphere
-------
TABLE 7. AQUEOUS WASTE STREAMS ASSOCIATED WITH DIFFERENT HIGH-Btu
GASIFICATION PROCESSES
Wastewater Category
Parti cul ate scrubber waters
from treatment of:
Pretreater Flue Gas
Lockhopper Vent Gas
Char Combustion Flue Gas
Raw Gas Quench Haters
Cyclone Slurry
Quench Slowdown
Ash Quench Water
Shift Condensate
Me th a nation Condensate
Waste Sorbents & Reagents
Miscellaneous Uastewaters
Gasification Process
0^— S
•C
CO
c
•r-
cn
•i— O>
CD
-------
vironmental assessment was to identify the
data gaps and limitations for study in the next
phase of the program. The limitations and gaps
fall into two categories: data that are nonex-
istent or unavailable, and data that are available
but either incomplete or obtained under condi-
tions significantly different than those antici-
pated in an integrated commercial SNG plant in
the United States.
Examples of the gaps in the first category are
the lack of detailed data on: emissions asso-
ciated with decommissioning spent methanation
catalyst, combined effluent in an SNG plant, and
sludges resulting from the treatment of such ef-
fluent or from the treatment of tar and oily con-
densates. Because no integrated SNG facility
currently exists, this type of information is not
available from actual operation. Even though
environmental characteristics of SNG plant
wastes can be estimated through engineering
studies, to date only a limited number of such
studies have been conducted. In the case of
emissions from catalyst decommissioning, even
though some data might exist, such data are not
publicly available because of proprietary con-
siderations.
Examples of the second category of data gaps
and limitations are the lack of trace element,
organic, toxicological, and ecological character-
istics data for various waste streams in a gasifi-
cation plant, and data on the performance of
various control systems in SNG service. In com-
parison with the limited data available on most
gasification processes, considerable data are
available on the characteristics of aqueous
wastes from the Hygas and dry-ash Lurgi proc-
esses. These data, however, do not cover
organic and trace element constituents, bio-
assay information, waste treatability, and
hazardous characteristics such as biodegrad-
ability, health effects, and potential bioaccumu-
lation and environmental persistence. For the
Stretford process, which has been used in
refinery and byproduct coke applications for
H2S removal from acid gases containing
relatively low levels of C02, limited commercial
experience exists with acid gases containing
high levels of C02 that would be encountered in
an SNG plant. With the exception of a few pollu-
tion control processes (e.g., flaring for hydrocar-
bon and H2§ control, venturi scrubbing for
particulate removal, Phenosolvan for recovery
of phenols from wastewaters, sour water strip-
ping for NHs/HgS removal, and trickling filters
for biological treatment), the various air, water,
and solid waste control processes that would be
potentially employed at commercial facilities
have not been used in coal gasification applica-
tions. Even for the few processes that have
been used for coal gasification, very little data
are available on the characteristics of the
treated streams and on the performance and
costs of these applications.
The first category of data gaps can only be
partially filled (e.g., through engineering
analysis) at the present time because SNG facil-
ities do not exist and the existing pilot plants do
not incorporate all the units or design features
of a large-scale facility. Many gaps in the second
category, however, can be and should be filled
through multimedia environmental sampling
and analysis of the process/discharge streams at
pilot plants and foreign gasification facilities,
through bench-scale studies and engineering
analysis. Even though some of the unit opera-
tions and conditions in the gasification pilot
plants are not scalable to or representative of
commercial facilities, in the absence of such
commercial facilities, sampling at the pilot
plants represents the best and only means of ac-
quiring meaningful data on process and waste
stream characteristics and on the performance
of various processes. Such sampling and analy-
sis programs, coupled with related engineering
studies and bench-scale testing, can provide
valuable and timely input to the evolution of the
SNG industry that would ensure:
• Inclusion of environmental considerations in
selection of processes, equipment, and waste
management options for commercial SNG
plants; and
• Drafting of new source performance stand-
ards for SNG facilities based on sound tech-
nical and engineering data.
Several programs are currently underway or
planned that involve testing/sampling at pilot
plants, bench-scale units, or foreign commercial
facilities.
Major programs that are expected to gener-
ate some of the data needed for high-Btu gas-
ification environmental assessment fall into
three categories: EPA-sponsored programs,
DOE-sponsored programs, and miscellaneous
programs. Limited data are available on the
programs in the miscellaneous category that
are primarily carried out under private funding.
213
-------
Of the EPA programs, the one most directly
related to the high-Btu gasification is the TRW
environmental assessment effort for which the
data base development effort has been the first
step. DOE synthetic fuel pilot and demonstra-
tion programs include sampling and analysis at
various facilities, bench-scale studies for proc-
ess and environmental data acquisition, and
related environmental engineering studies.
Preparation of the data base document repre-
sents completion of the first phase of the TRW
program. The second phase of the program in-
cludes data acquisition through sampling and
analysis of process/waste streams at selected
gasification facilities.
214
-------
APPENDIX A
DRY-ASH LURGI GASIFICATION PROCESS
GENERAL INFORMATION
1. Operating Principles: high-pressure coal gasification in a gravi-
tating bed by injection of steam plus oxygen with countercurrent
gas/solid flow; ash is maintained below the fusion temperature.
2. Development Status: commercially available since 1940.
3. Licensor/Developer: Lurgi Mineralb'technik GMbH.
American Lurgi Corporation
377 Rt. 17 South
Hasbrouch Heights, N.J.
4. Commercial applications: See Table A-l.
PROCESS INFORMATION
1. Commercial Scale: see Figure A-l for flow sheet.
A. Gasifier: see Figures A-2 and A-3.*
(1) Equipment1 2
Construction: vertical, cylindrical steel pressure vessel.
Gasifier dimensions:
2.5 to 3.8 m (8.5 to 12.3 ft) in diameter,
2.1 to 3.0 m (7 to 10 ft) coal bed depth, and
5.8 m (19 ft) approximate overall height of gasifier.
Bed type and gas flow: gravitating bed; continuous counter-
current gas flow; lateral gas outlet near the top of the gasifier.
Heat transfer and cooling mechanism: direct gas/solid heat
transfer; water jacket provides gasifier cooling.
Coal feeding: intermittent; pressurized lock hopper at the
top of the gasifier dumps the coal onto a rotating, water-
cooled coal distributor.
Gasification media introduction: continuous injection of steam
plus oxygen at the bottom of the coal bed through a slotted ash
extraction grate.
Ash removal: rotating, slotted grate at the bottom of the
coal bed; refractory-lined, pressurized lock hopper collects
the ash and dumps it intermittently.
Special features:
Direct quench gas scrubber and cooler that knocks out the
majority of particulates, tars, oils, phenols, and ammonia;
is attached to the gasifier at the gas outlet.
Gasifier water jacket supplies approximately 10 percent
of the required gasification steam.
Rotating coal distributor provides uniform coal bed
depth.
*Figure A-2 shows the evolution of Lurgi gasitiers with corre-
sponding increases in capacity. Figure A-3 presents the commercial
model that is the basis for further discussion.
215
-------
TABLE A-1. LURGI, DRY-ASH, COMMERCIAL INSTALLATIONS1
Plant
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Location
Boh 1 en,
Central Germany
Boh Ten,
Central Germany
Most, CSSR
Zaluzi-Most,
CSSR
Sasolburg,
South Africa
Dors ten,
West Germany
Morwell ,
Australia
Daud Khel,
Pakistan
Sasolburg,
South Africa
Westfield,
Great Britain
Jealgora, India
Westfield,
Great Britain
Coleshill,
Great Britain
Naju, Korea
Sasolburg,
South Africa
Luenen, GFR
Sasolburg,
South Africa
Year
1940
1943
1944
1949
1954
1955
1956
1957
1958
1960
1961
1962
1963
1963
1966
1970
1973
Type of Coal
Lignite
Lignite
Lignite
Lignite
Sub-Bitum. with 30%
ash and more
Caking Sub-Bitum.
with high chlorine
content
Lignite
High Volatile coal
with high sulfur
content
Sub-Bitum. with 30%
ash and more
Weakly Caking Sub-
Bitum.
Different grades
Weakly Caking Sub-
Bitum.
Caking Sub-Bitum.
with high chlorine
content
Graphitic anthracite
with high ash
content
Sub-Bitum. with 30%
ash and more
Sub-Bitum.
Sub-Bitum. with 30%
ash and more
Gasifier
I.D.
8'6"
8' 6"
8' 6"
8' 6"
12'1"
8 '9"
8'9"
8'9"
12T
8' 9"
N/A
8'9"
8 '9"
10 '5*
12'1"
IT 4"
12'4"
Capacity
(MMSCFD)
9.0
10.0
7.5
9.0
150.0
55.0
22.0
5.0
19.0
28.0
0.9
9.0
46.0
75.0
75.0
1400 MM
Btu/hr
190.0
No. of
Gasifiers
5
5
3
3
9
6
6
2
1
3
1
1
5
3
3
5
3
216
-------
LEGEND:
to
1.
2.
3.
4.
COAL
02
STEAM
FEED LOCK
HOPPER GAS
ASH LOCK
HOPPER GAS
FEED LOCK
HOPPER VENT GAS 16. TAR
9. ASH
10. PRODUCT GAS
11. COMBINED LIQUID STREAM
12. SEPARATOR FLASH GAS
13. OIL
14. LIQUOR
15. RECYCLE TAR
RAW GAS
ASH LOCK
HOPPER VENT GAS
17. RECYCLE LIQUOR
Figure A-1. Lurgi gasifier (Based on Westfield Lurgi Installation).
-------
year
to
i-«
00
first generation
1936 -1954
•GAS
coal grade
capacity
MM BTU coal input
hr
lignite
100
second generation
1952-1965
all coal grades
180-250
GAS
non-caking coals
400-500
third generation
ifi from 1969
all coal grades
450-570
Figure A-2. Stages of Lurgi gasifier development.
-------
FEED COAL
RECYCLE TAR
DRIVE
GRATEN
DRIVE
STEAM*
OXYGEN
SCRUBBING
COOLER
GAS
WATER JACKET
Figure A-3. Lurgi pressure gasifier.
219
-------
Tar injection nozzle at the top of the gasifier permits
recycle of byproduct tar (separated external to the
gasification module), which also helps to reduce coal
fines carryover in the product gas (optional features).
Rotating, water-cooled coal bed agitator aids gasifi-
cation of strongly caking coals (optional feature).
(2) Operating Parameters1 2
Gas outlet temperature:
Range: 644 K to 866 K (700° F to 1,100° F).
Normal: 727 K (850° F).
Coal bed temperatures:
1,255 K to 1,644 K (1,800° F to 2,500° F).
Gasifier pressure:
Range: 2.1 to 3.2 MPa (300 to 465 psia).
Normal: 2.1 MPa (300 psia).
Coal residence time in gasifier:
Approximately 1 hr.
(3) Raw Material Requirements1 2
Coal feedstock:
Type: All types; strongly caking coals require agitator-
reduced
throughput and increased steam rate.
Size: 3.2 to 38.1 mm (0.125 to 1.5 in):
Coal is usually fed in two size ranges; coal with up
to 10 percent minus 3.2 mm (0.125 in) can be accepted.
Rate:f 136 to 544 g/sec-m2 (100 to 400 lb/hr-ft2).
Coal pretreatment: crushing and sizing, drying to less than
35 percent moisture; partial oxidation is required for use of
strongly caking coals in gasifiers without agitators.
Steam: 1.11 to 2.59 kg/kg coal.3
Oxygen: 0.26 to 0.62 kg/kg coal.3
Quench water: 3.3 x io"4 m3/kg coal.2
(4) Utility Requirements1
Water: , ,
Boiler: 2.42 x lo"J roVkg coal (580 gal/ton coal).
Cooling: ?
Electricity: 25 kwh/metric ton (23 kwh/ton).
(5) Process Efficiency
Cold gas:3 63 to 60 percent
r-i [Product gas energy output] ,nn
H [Coal energy Tnput] 10°
Overall thermal:1 76 percent.
Total energy [Product + ' HC by +
L output \ gas products steam/I „ ,QO
[Total energy input (coal + electric power)] '
tRate varies with gasifier design and coal type.
220
-------
(6) Expected Turndown Ratio1 = 100/25.
[Full capacity output]
[Minimum sustainable output] '
(7) Gas Production Rate/Yield:3
0.37 to 0.68 m3/sec-m2 (4875 to 900 scf/hr-ft2)
0.93 to 1.70 NmVkg coal (16 to 30 scf/lb-coal).
2. Coal Feed Pretreatment: coal feed is from pressurized lock hop-
pers; no pretreatment is required in third-generation gasifiers.
3. Quench and Dust Removal: crude gas leaves the top of the gasifier
and flows through a scrubber cooler, where it is washed by recircu-
lating quench liquor from the tar-oil separation section. The
gases then pass through a waste heat boiler and a final cooler.
Dust, tars, and condensables are collected from these units.
PROCESS ECONOMICS
Because of advanced development of the Lurgi gasifier, numerous
studies related to costs have been completed.4 5 6 7 However, most of
these studies address themselves only to integrated facilities rather
than to the gasification module. The one exception, in which equipment
lists are presented and detailed cost estimates made, is the Bureau of
Mines Study.4 For a 250-MMSCFD SNG facility costing $737,538,000 in
1974 dollars, 27.1 percent is estimated to be attributable to the gasi-
fication section. Lurgi7 estimates total plant costs of $440,000,000
also in 1974 dollars. No gasification section cost estimates are made.
PROCESS ADVANTAGES
Present gasifiers can accept caking and noncaking coals.
Pressurized operation favors formation of methane In the
gasifier and reduces upgrading costs. The high pressure of
the product gas would also reduce the cost of gas transmission
via pipeline. High pressure may be advantageous for combined-
cycle synthesis gas utilization.
Gasifier has been operated commercially for many years.
Small reactor size may be advantageous for small-scale indus-
trial applications.
PROCESS LIMITATIONS
Caking coals reduce throughput rate and increase steam consump-
tion, which also increases the amount of liquid waste to be
treated.
Maintaining the coal-bed temperature below the ash fusion
temperature limits the maximum process efficiency.
Process condensate and byproducts require additional processing
for environmental acceptability.
Maintaining a low coal-bed temperature results in low steam
conversion in the gasifier.
Limited reactor size may necessitate use of multiple units in
parallel for large Installations.
221
-------
INPUT STREAMS3 8
1. Coal (Stream No. 1): see Table A-2.
2. Oxygen (Stream No. 2)
Coal No. 1 2345
Rate: kg/kg
(Includes 6
percent
Inerts) 0.26 0.48 0.49 0.62
Pressure: HPa
(psia) 3.6(370) 3.5(360) 3.5(360) 3.5(360) --
3. Steam (Stream No. 3)
Coal No. 12345
Rate: kg/kg 1.11 1.97 1.84 2.59
Pressure: (psia) (370) (362) (360) (360) --
Temperature: K (°F)
4. Feed Lock Hopper Gas (Stream No. 4): no data reported.
5. Ash Lock Hopper Gas (Stream No. 5): no data reported.
INTERMEDIATE STREAMS
1. Gaseous Streams
A. Feed Lock Hopper Vent Gas (Stream No. 6): no operational data
reported.
B. Raw Gas (Stream No. 7): no operational data reported.
C. Ash Lock Hopper Vent Gas (Stream No. 8): no operational data
reported.
2. Liquid Streams
A. Combined Liquid Stream (Stream No. 11): no data reported.
B. Recycle Liquid (Stream No. 17): no data reported.
C. Recycle Tar (Stream No. 15)3
Coal No. 12345
Toluene
(wt percent) 32.3 8.6 8.0 3.2
Insoluble
ash (dust) 29.2 10.8 11.1 12.2
Composition (See Tars--Stream No. 16)
DISCHARGE STREAMS
1. Gaseous
Product gas (Stream No. 10): see Table A-3.
Separator flash gas (Stream No. 12): see Table A-4.
2. Liquid Streams
Tars (Stream No. 16): see Tables A-5 and A-9.
Oils (Stream No. 13): see Tables A-6 and A-9.
Liquors (Stream No. 14): see Tables A-7 and A-9.
3. Solids Streams
Ash (Stream No. 9): see Tables A-8 and A-9.
222
-------
TABLE A 2. PROPERTIES OF COAL FEED TO LURGI GASIFICATION (STREAM NO. 1)
Coal No.
Type/Origin
Size: rim (in)
HHV (dry):
Kcal/kg (Btu/lb)
Swelling No.
Caking Index
Compos i ti on :
Moisture: %
Volatile matter: %
Ash: %
C: %
H: %
0: %
S: %
N: %
Trace Elements*(ppm)
Be
Hg
Ca
Sb
Se
Mo
Co
Ni
Pb
As
Cr
1
Montana Rosebud*
Subbituminous A
6.4-31.8
(1/4-1 1/4)
6553
(11,436)
0
0
24.70
29.20
9.73
67.15
4.22
13.02
1.45
1.20
—
—
_.
--
—
—
—
--
--
--
--
2
Illinois #6*
High Volatile
6.4-31.8
(1/4-1 1/4)
7094
(12,770)
3
15
10.23
34.70
9.10
71.47
4.83
9.02
3.13
1.35
1.6
1.1
< .03
0.1
—
7
4
14
10
1
20
3
Illinois #5*
Bituminous
6.4-31.8
(1/4-1 1/4)
7228
(13,010)
2.2-5
15
11.94
35.21
8.13
72.80
4.95
7.99
3.56
1.39
2.0
0.2
< .03
.2
9
7
4
32
28
2
15
4
Pittsburgh
#8
6.4-31.8
(1/4-1 1/4)
7826
(14,087)
7.5
30
4.58
37.37
7.74
77.71
5.28
4.74
2.64
1.42
—
--
—
—
--
—
—
—
—
5
South African1"
Subbituminous
4989
(8,980)
—
—
8.0
—
31.6
52.4
2.6
11.7
0.43
1.2
—
—
--
--
--
--
—
—
—
_ i ^_
:
(continued)
09
-------
TABLE A 2 (continued)
Coal No.
Type /Ori gin
Trace Elements*(cont)
(ppm)
Cu
B
Zn
V
Mn
F
Cl
1
Montana Rosebud*
Subbituminous A
~
—
—
.
—
--
400
2
Illinois #6*
High Volatile
12
132
43
29
20
79
600
3
Illinois #5*
Bituminous
10
307
200
21
22
57
800
4
Pittsburgh
#8
—
—
—
—
—
--
1000
5
South African
Subbituminous
__
__
__
--
--
--
™ ™
From trials of American coals at Westfield
tData from SASOL unit in South Africa^8^.
fData from trials of American coals at WestfielcP .
-------
TABLE A-3. PRODUCTION RATE AND COMPOSITION OF LURGI
PRODUCT GAS-STREAM NO. 103 8
Coal No.
Production Rate:
NnvVkg coal
(C02, N2> and 02
free basis)
Gas Analysis:
H2
02 (includes
N2+Argon)
CO
CH4
co2
C2H6
£ D
H2S
Total Organic
Sulfur
NH.
w
HCN
Naphthalene
St. ClairdeVille
Condensable
1
0.98 m3/kg
41.1%
1.2
15.1
11.2
30.4
0.5%
666g/100Nm3
12-40
0.09
0.27g/100Nm3
0.24
389
2
1.36
39.1
1.2
(N2-0.6)
17.3
9.4
31.2
0.7
1510
23
0.18
2.8
0.68
460
3
1.79
38.8
1.5
(N2-0.7)
17.5
9.2
31.0
0.5
(C2H4-0.3)
1420
30
not
detectable
8.7
1.1
531
4
1.32
39.4
1.6
(N2-0.8)
16.9
9.0
31.5
0.7
(C2H9-0.1)
1010
15
0.18
0.50
1.2
277
5
1.36
40.05
—
20.20
8.84
28.78
0.54-
422
--
. -_
--
—
_*.
225
-------
TABLE A-4. COMPOSITION OF LURGI SEPARATOR FLASH GAS-STREAM NO. 12
(VOLUME PERCENT)3
Coal No.
H2S
NH3
co2
CO
H2
02+Argon
N2
CH4
1
Tar Oil
Sep. Sep.
3.8 8.6
6.3 12.0
64.7 59.3
5.9 4.7
2.9 2.3
3.1 2.5
8.0 6.4
5.3 4.2
2
Tar Oil
Sep. Sep.
5.7 5.5
1.0 1.8
84.9 85.5
1.5 0.8
3.5 3.6 •
0.4 0.6
1.2 1.0
1.8 1.2
3
Tar Oil
Sep. Sep.
6.2 6.8
4.6 2.7
62.9 67.0
4.5 4.2
11.7 13.3
1.3 1.4
5.9 2.3
2.9 2.3
4
Tar Oil
Sep. Sep.
4.4 5.5
2.9 3.5
71.3 73.9
4.7 3.8
12.0 9.6
0.3 0.2
1.0 0.8
3.4 2.7
5
—
—
--
--
—
--
--
—
226
-------
TABLE A-5. PROPERTIES OF LURGI TAR-STREAM NO. 10
Coal No.
Production Rate:
kg/kg coal
Water: wt. %
Toluene
insoluble wt. %
Density: grams/cc
Phenols: (wet) wt. %
Calorific Value
Gross: Kcal/kg
(Btu/lb)
Ultimate Analysis
(dry, dust- free
basis)
C wt. %
H wt. %
N wt. %
S wt. %
Cl wt. %
Ash wt. %
0 (by difference)
wt. %
1
0.02
30.0
22.0
1.025
5.3
8794
(15,830)
83.06
7.. 69
0.65
0.28
0.04
0.05
8.23
2
0.03
26.7
4.5
1.145
2
8829
(15,893)
85.48
6.44
1.18
1.70
N.D.
0.03
5.17
3
0.04
10.4 ,
7.1
1.148
4.7
8837
(15,906)
85.85
6.40
1.19
2.39
N.D.
0.01
4.16
4
0.03
11.9
8.5
1.175
1
8956
(16,120)
88.51
5.93
0.87
1.52
N.D.
0.01
3.16
5
0.02*
—
--
--
--
--
--
0.3
--
--
--
227
-------
TABLE A-6. PROPERTIES OF LURGI OIL-STREAM NO. 133
Coal No.
Production rate kg/kg
Water: wt. %
Dust: wt. %
Density: grams/cc
Phenols: (dry, dust-
free) wt. %
Calorific Value
Kcal/kg (Btu/lb)
Ultimate Analysis:
C: wt. %
H: wt. %
N: wt. %
S: wt. %
Cl: wt. %
Ash: wt. %
Oxygen: (by
difference) wt. %
1
0.02
22.3
0.4
0.937
19.1
(16,960)
81.34
9.17
0.46
0.50
0.04
0.03
8.46
2
0.003
4.3
0.8
1.015
20.1
(16,482)
84.82
7.77
0.70
2.40
N.D.
0.01
4.30
3
0.007
5.4
0.1
1.011
19.2
(16,578)
8.488
7.65
0.49
2.27
N.D.
0.01
4.70
4
0.01
15.4
0.02
0.991
10.0
(17,134)
87.33
7.61
0.45
1.50
N.D.
0.01
3.10
5
0.004
--
—
--
_ _
—
--
—
—
0.25
—
«
--
228
-------
TABLE A-7. PROPERTIES OF LURGI LIQUORS-STREAM NO. 143
Coal No.
Prod. Rate kg/ kg
Tar: ppm
Analysis on
tar free
basis
Tar free basis
PH
S.G. at 60°F
T.D.S.:
ppm
T.D.S.
after
ignition
ppm
Sulfide
H2S, ppm
Total S;
ppm
Fatty acids:
ppm
Ammonia:
Free: ppm
Fixed ppm
Carbonate:
ppm
1
0.93
350 650
Inlet Inlet
tar oil
sep. sep.
9.6 8.3
1.003 1.025
4030 1765
45 35
130 115
150 265
1250 1670
3990 14015
395 525
4070 19460
2
2.11
1130 2150
Inlet Inlet
tar oil
sep. sep.
9.8 8.5
1.003 1.032
2770 1570
110 35
25 440
180 730
490 280
1700 17650
280 210
1280 6550
3
1.77
2150 2200
Inlet Inlet
tar oil
sep. sep.
9.5 8.3
1.002 1.027
3180 1120
85 25
15 490
160 930
400 260
1520 13970
410 330
680 9210
4
2.60
300 1100
Inlet Inlet
tar oil
sep. sep.
9.3 8.2
1.000 1.026
1550 1240
105 120
65 520
155 720
275 610
1600 14000
320 250
1360 10740
5
1.06
5000
(tar & oil)
—
—
__
--
-_
0.03%
10,600
150-200
—
(continued)
-------
TABLE A 7 (continued)
Coal No.
Total phenols:
ppm
Cyanide:
ppm
Thiocyanate:
ppm
Cl : ppm
BOD: ppm
COD: ppm
1
4200 4406
2 4
6 15
45 40
9900 13400
22700 208CO
2
2200 1900
3 11
65 160
135 75
3800 4700
10100 12000
3
2900 3750
7 14
79 158
290 170
6000 6200
9300 10600
4
1400 2150
1 12
70 185
240 210
4100 5400
650 7500
5
3250-4000
6
^^
--
—
—
-------
TABLE A-8. PROPERTIES OF LURGI ASH-STREAM NO. 93 10
Coal No.
Production Rate:
kg/kg
Angle of repose
Bulk Density
Poured:
kg/Nm3 (lb/ft3)
Tapped :
kg/Nm3 (lb/ft3)
Ash Fusion Temp.
Oxidizing:
I.F.: oc
H.P.: oc
F.P.: oc
Reducing:
I F • or
• • I • • w
H P • Or
1 1 • r • • v*
F P • or
r • r • • **\/
Partial analysis
Carbon: wt. %
Si02: wt. %
Al£03: wt. %
fA O fl *> * 1*1 T T«
ICXV4* TV w • /w
CaO: wt. %
MgO: wt. %
Sulfur (as
S03): wt. %
Cl: wt. %
1
0.097
24°
918 (57.4)
1078 (67.4)
1240
1260
1290
1165
1175
1210
6.5
46.8
17.7
11.2
8.3
3.9
1.7
0.01
2
0.090
330
762 (47.6)
894 (55.9)
1350
1365
1390
1090
1150
1225
3.2
49.6
20.5
17.2
2.1
1.0
1.3
0.01
3
0.087
41°
990 (61.9)
1106 (69.1)
1280
1300
1330
1030
1060
1070
2.0
46.1
18.1
19.7
3.9
0.7
0.6
0.01
4
0.077
43°
(42.1)
(48.9)
.
1340
1360
1380
1145
1170
1180
7.6
43.6
20.7
15.0
3.0
0.7
0.8
0.01
5
0.313
„
--
—
«
—
—
—
—
52
28
5
7
1.7
0.2
—
(continued)
231
-------
TABLE A-8 (continued)
Coal No.
Trace Elementst (ppm)
Be
Hg
Cd
Sb
Se
Mo
Co
Ni
Pb
As
Cr
Cu
B
Zn
V
Mn
F
1
--
—
—
—
--
—
--
—
—
--
—
—
—
--
—
--
--
2
14
.04
<0.3
0.2
--
6
.40
456
96
0.1
750
239
622
469
301
200
5
3
20
.016
<0.3
19
--
8
.38
462
200
0.3
592
273
673
1600
181.
305
4.6
4
—
—
—
--
--
—
—
--
—
—
--
—
—
—
—
—
5*
—
—
--
—
—
--
--
--
—
--
—
--
--
--
—
--
*Trace element balance for SASOL is presented in Table A-9
tFrom Reference 10.
232
-------
TABLE A-9. TRACE ELEMENT BALANCE FOR LURGI AT SASOL*
(PERCENT OF ELEMENT IN COAL)8
Element
Be
B
V
Mn
Ni
As
Cd
Sb
Ce
Hg
Pb .
Br
F
Cl
Ash
1
36
72
154
154
36
40
40
72
40
180
3.6
54 1
511
Liquor
1.6
3.5
0.06
0.36
0.64
90
35
36
0.1
32
3.2
32
42 r
46 r
Tar
0.5
0.8
0.005
0.005
0.05
2.5
0.5
3
0.003
4.9
8.2
0.05
0.08
0.24
Oil
0.01
0.002
<0.001
<0.001
0.01
5.2
1.1
0.5
0.001
0.5
0.02
—
0.003
0.008
Total
3
40
72
154
155
134
77
80
72
77
191
36
96
97
*Analysis by spark source mass spectrometer (which can give a semi-
quantitative analysis) for El Paso by SASOL.
t% distribution calculated on analyses as done by Sasol previously.
233
-------
DATA GAPS AND LIMITATIONS
Even though the Lurgi gasifier has the most complete data of any
gasifier because of its advanced development, the available data are
not comprehensive in that not all streams (e.g., lock hopper vent gas)
are addressed, and not all potential pollutants and toxicological and
ecological properties are identified. An environmental data acquisition
effort that would lead to generation of the needed data corresponds to
EPA's phased level approach to multimedia environmental sampling and
analysis.9
RELATED PROGRAMS
Environmental assessments of commercial-scale Lurgi SNG facilities
have been prepared by El Paso Natural Gas for its proposed Burnham
facility and by ANG Coal Gasification Company for its proposed North
Dakota Coal Gasification Project. Documents on process and environmental
considerations for other projects have also been released. Chief among
these is the Wesco SNG facility. The Department of Energy (DOE) recently
conducted tests at the British Coal Board's Lurgi plant at Westfield,
Scotland. The tests involved operating the Lurgi gasifier in the
slagging mode (this is the subject of another gasifier data sheet). EPA
has released a report, Control of Emissions from Lurgi Coal Gasifica-
tion Plants (EPA 450/2-78-012, March 1978), which is to provide informa-
tion to States and regional EPA offices involved in setting standards
for or evaluating impacts from proposed Lurgi gasification facilities.
REFERENCES
1. Handbook of Gasifiers and Gas Treatment Systems. Dravo Corp. ERDA
FE-17772-11. February 1972.
2. The Lurgi Process: The Route to S.N.G. from Coal. (Presented at
the Fourth Synthetic Pipeline Gas Symposium. Chicago. October
1972.)
3. Woodall-Duckham, Ltd. Trials of American Coals in a Lurgi Gasifier
at Westfield, Scotland (final report). Crawley, Sussex, England.
Research and Development Report No. 105, FE-105. November 1974.
4. Preliminary Economic Analysis of Lurgi Plant Producing 250 Million
SCFD Gas from New Mexico Coal. Bureau of Mines. Morgantown, W.Va.
Report No. ERDA-75-57. March 1976.
5. Gallagher, J. T. Political and Economic Justification for Immediate
Realization of a Synfuels Industry, Third Annual International
Conference on Coal Gasification and Liquefaction: What Needs to be
Done Now. Pittsburgh, Pa. August 1976.
6. Kasper, S. Lurgi Gasification Process: Prospects for Commercializa-
tion, Symposium on Coal Gasification and Liquefaction. Pittsburgh,
Pa. August 1974.
7. The Lurgi Pressure Gasification: Applicability. Lurgi Express
Information Brochure Number 01145/6.75. January 1974.
8. Information Provided to the Fuel Process Branch of EPA's Industrial
Environmental Research Laboratory (Research Triangle Park) by South
African Coal, Oil and Gas Corporation, Ltd. November 1974.
234
-------
9. Dorsey, J. A., and Johnson, L. D. Environmental Assessment Sampling
and Analysis: Phased Approach and Techniques for Level 1. EPA-
600/2-77-115. June 1977.
10. Sather, N. F., et al. Potential Trace Element Emissions from the
Gasification of Illinois Coal. Illinois Institute for Environmental
Quality. Number 75-08. February 1975.
235
-------
APPENDIX B
RECTISOL PROCESS
(SINGLE-ABSORPTION MODE)
GENERAL INFORMATION
1. Operating Principles: physical absorption of the sour components
(H2S, C02, COS, mercaptans, etc.) of a gas stream using methanol as
the sorbent. Selective regeneration can provide a rich sulfur-
containing gas stream and a relatively pure C02 stream.
2. Development Status: commercially available.
3. Licensor/Developer: Lurgi MineralSltechnik GmbH
American Lurgi Corporation
377 Rt. 17 South
Hasbrouck Heights, N.J.
4. Commercial Applications
Purification of low/medium-Btu gas produced from coal gasifi-
cation. Gasificatibn plants using the process include
Sasolburg, South Africa; Westfield, Scotland; and Pristina,
Yugoslavia.
Carbon dioxide removal and drying of coal-derived ammonia
synthesis gas. One of the facilities using this process is
located in Kutahya, Turkey.
Carbon dioxide removal from low-temperature fractionation feed
gas. The locations of facilities using the process are not
known.
Carbon dioxide and water removal from a feed gas to LNG plants.
Plant location(s) are unknown.
PROCESS INFORMATION1 2 3 6
1. Flow diagram (see Figure B-l, B-2, and B-3): the Rectisol process
can be used in a variety of modes to achieve different treatment
objectives. Only three operation modes that have been used or
proposed appear most pertinent to coal conversion and are discussed
here. The pertinent features of these operation modes are summa-
rized i£ Table B-l.
2. Equipment: conventional absorbers, stripping columns, distillation
columns, heat exchangers, separators, and regenerators.
Construction: vessels may be fabricated from carbon steel,
dimensions dependent on application.
3. Feed Stream Requirements:"' g£s should be cooled to reduce solvent
losses; high pressures (close to 2.0 MPa or 300 psia) are usual.
Gas temperatures between 253 K and 213 K (-5° F to -75° F) are
usual, depending on conditions.6
"'These conditions are for optimum performance; other input condi-
tions can be handled with increased solvent losses and reduced effi-
ciency.
236
-------
PREWASH COLUMN
L
«_
•••••••
•>—
— ••»•
£
<
u.
I
PREWAS
L
«v
;
•»
h
• NAPHTHA
SEPARATOR
AZEOTROPE COLUMN
I
i
*— 12
•>
ABSORBER
•v-
— ^
REGENERATOR
5
<
5
13
— *i
t
_i
_i
k
O
51
i
1
«—
—
— ^
— *i
REGENERATOR
O
LEGEND:
1. RAW GAS
2. WATER
3. MAKE-UP M»OH
4. PREWASH FLASH GAS
5. REGENERATOR FLASH GAS
6. PRODUCT GAS
7. EXPANSION GAS
8. COMBINED FLASH GAS
9. REGENERATOR OFF-GAS
10. STILL BOTTOMS
11. NAPHTHA
128.13 STRIPPING GAS
t
-•-10
-•» 11
Figure B-1. Rectisol type A3 (removal of CO2 from gas mixtures containing
little or no H2S).
-------
to
CO
00
f
1
OLUMN
o
I
PREWAS
«•
1
i
i
(
c
n
<
j.
r
n
<
LJ
c
>.
«-
M
L—
'
I
C02
ABSORBER
JLFUR
SORBER
(/] CD
1
J
<
(
i
\
u
f.
j
CM
•
t
5
(/)
u.
SULFUF
\
1
t
oc
0
Ul
UJ
o
UJ
oc
0
1
1
-0— -
w
i
c
3
:
i
^
«
k
j
j
>
>
•
^
t
NAPHTHA
SEPARATOR
LEGEND:
1. RAW GAS
2. WATER
3. MAKE-UP MeOH
4. PREWASH FLASH GAS
5. SULFUR F LASH GAS
6. PRODUCT GAS
7. LEAN H2S COMBINED GAS
a CO2 F LASH GAS
-••11
9. REGENERATOR GAS
10. STILL BOTTOMS
11. NAPHTHA
12. 13, 8. 14. STRIPPING GAS
Rgure B-2. Rectisol type B3 (removal of CO2 and H2S with separate recovery).
-------
o
ABSORBER
B-'rVB
I
O
I i
[
18
"P
CO
[ STAGE
1
7 „
3RD STAGE
ABSORBER
14
12-
NAPHTHA
SEPARATOR
i
LEGEND:
1. RAW GAS
2. WATER
3. MeOH
4. NH3 COOLANT
5. NH3 COOLANT
6. PRODUCT GAS
7. INTERMEDIATE GAS
8. REGENERATOR OFF-GAS
9. 5TH STAGE FLASH GAS
10. 6TH STAGE FLASH GAS
11. 1ST. 2ND. 3RD.4TH STAGE FLASH GAS
12. AROMATICS
13. STILL BOTTOMS
14. CONDENSATE
10
Figure B-3. Rectisol type C3 (removal of CO2 and H£S with separate recovery).
-------
TABLE B-1. PROCESS DESCRIPTIONS FOR RECTISOL TYPES A, B. AND C
OPERATING MODES
Process Application/
Type Treatment Objective
Process Description
(Fig. B-1)
Removal of C0£ from
gas mixture con-
taining little or
no sulfur.
to
B
(Fig. B-2)
(Fig. B-3)
Simultaneous removal
of C02 and sulfur
compounds with sep-
arate recovery.
Same as Type B
A methanol stream rich 1n CO? and HzS is used in the prewash column to remove water,
naphtha, ammonia and residual heavy hydrocarbons from the raw gas. The exit
solvent enters the prewash flash column where a flash stream lean in H2S and
rich in CO? is produced (Stream 4). The liquid bottoms from the flash vessel
are routed to a separator where water (Stream 2) is added so that the naphtha
and heavy hydrocarbons can be separated. In the main absorber raw gas contacts
a pure methanol stream from the hot regenerator. A slipstream of saturated
methanol is sent to the prewash column. The remaining methanol is sent to a
flash regenerator where the absorbed gases are removed. Methanol from flash is
sent to the hot regenerator where the final traces of COg and l^S are removed.
Water is removed from the prewash methanol in the methanol/water still with off
gases going to the hot regenerator. Stripping gases (usually nitrogen) may be
used.
Except for the use of a two-stage absorber and two separate flash columns, Type B
Rectisol is very similar to Type A. The raw gas (after leaving the prewash
absorber) is first contacted with a (^-saturated methanol stream. This first
stage absorber removes H2S. In the second stage a pure methanol stream removes
C0£. The methanol for first stage comes from the second stage absorber. The two
methanol streams are flashed separately to create a stream rich in H£S (No. 5)
and a nearly pure (X>2 stream (No. 8). Regeneration is the same as in the Type A.
The primary difference in Type C as compared to Type B is in the regeneration pro-
cess. The first stage acts like the prewash in Type B with second and third stages
similar to first and second in Type B. A multistage flash unit is used to desorb
gases from first and second stage absorption. First stage methanol is first com-
bined with heavy hydrocarbons and water removed from the raw gas and sent to the
separator. The separator works in the same manner as the separators in Types A
and B. The multistage flash reduces the regeneration requirements. The third
stage methanol is handled in a conventional hot regenerator to provide a pure
methanol for final absorption. A split stream regeneration section is also shown
in Figure B-3. Similar gas cooling sections are used in Types A and B but are not
shown on the figures.
-------
4. Operating Parameters
Absorption: 0.3-7.1 MPa (45 to 1066 psia) approximately
303 K (80° F).
Regeneration: see discharge streams, Section 8.0.
5. Process Efficiency and Reliability
C02 better than 97 percent.4
H2S better than 99.9 percent.4
Reliability is considered high with a simple solvent and
construction.
6. Raw Material Requirements
Solvent: CH3OH; purity - ?
Solvent losses can be estimated using equilibrium constants;
however, considerable errors could be involved. No information
available on solvent losses based on actual operating data.
7. Utility Requirements: ?
8. Miscellaneous: ?
PROCESS ADVANTAGES
Lower energy consumption than conventional amine solvent acid-
gas removal processes.2
Can be adapted for removal of all impurities in one pass or
for selective removal.2
Production of a product gas with very low water content.2
Noncorrosive nature of the solvent.3
Unlimited solubility of methanol in water.3
Chemical stability and low freezing point.3
Good for high-pressure applications.
PROCESS DISADVANTAGES
Complex flow scheme.2
Solvent carryover losses may be high.2
Not suited for operation at pressures below 1.1 MPa (165
psia).2
PROCESS ECONOMICS - ?
INPUT STREAMS
1. Gaseous
Stream No. 1: Raw Gas--see Table B-2.
Stream Nos. 12, 13, and 14: Types A and B stripping gas:Ta
When used, the stripping gas is nitrogen from an oxygen plant.
'This corresponds to the range from Type A and B facilities re-
ported in Table B-2 from Reference 3.
241
-------
TABLE B-2. RECTISOL GASEOUS INPUT STREAMS
Constituents/
Parameters
M ^
CO
CH4
co2
N2 + Ar
H2S
COS
cs2
RSH
Thiophene
c2+
MeOH
Temp: °K (°F)
Pressure:
MPa (psia)
Rate: Nm3/hr
(SCFM)
Stream Number Reference
Type A Type A Type B Type B Type c
40.05
20.20
8.84
28.78
1.59
4220 mg/Nm3
10 ppm
--
20 ppm
—
0.54
--
303 (86)
2.5 (380)
381,000
(236,000)
58.4
0.3
0.2
21.9
19.2
—
—
—
--
--
--
--
--
2.4 (356)
153,100
(94,300)
62.31
3.25
0.17
33.25
0.53
0.49
10 ppm
—
—
—
—
—
—
3.2 (480)
142,340
(88,250)
61.59
2.60
0.33
34.55
0.41
0.52
—
—
--
--
--
--•
--
7.1 (1066)
137,000
(84,940)
63.74
4.13
0.13
31.62
0.12
0.26
63 ppm
--
--
--
--
--
303 (86)
0.3 (45)
80,000
(49,600)
*A11 values, unless otherwise noted, are in volume percent.
242
-------
Rate: 231,300 to 693,500 NnrVhr (153,400 to 430,000 SCFM).
Temperature: ?
Pressure: 0.1 to 0.5 MPa (20 to 80 psia).
2. Liquid
Stream No. 2: Water to Separator—quantity ?
Stream No. 3: Methanol Makeup—quantity ?
INTERMEDIATE STREAMS
1. Gaseous
Stream Nos. 4 and 5: Types A and B Flash Gases--?
Stream No. 7: Type C Intermediate Gas--?
DISCHARGE STREAMS
1. Gaseous
Stream No. 6: Product Gas—see Table B-3.
Stream Nos. 7, 8, and 9: Types A and 6 Offgases—see Table
B-4.
Stream Nos. 8, 9, 10, and 11: Type C Offgases-see Table B-4.
2. Liquid
Stream No. 10: Types A and B Still Bottoms--?
Stream No. 11: Types A and B Hydrocarbons and Stream No. 12--
Type C Hydrocarbons—?
Stream No. 13: Type C Still Bottoms:4
Rate: 16 mVhr
pH: 9.7
Phenol: 18 nig/L
Cyanide (as CN): 10.4 mg/L (includes thiocyanate)
Ammonia (as N): 42 mg/L
Sulfides (as S): Trace
Oxygen absorbed: 286
COD: 1,606 mg/L
Conductivity: 1,111 umhos/cm
Stream No. 14: Type C—?
DATA GAPS AND LIMITATIONS
The major limitation in the data is that not all input and dis-
charge streams are characterized, and the characterizations are not
comprehensive in that all potential pollutants and toxicological and
ecological properties are not identified. An example is the total lack
of data on MeOH carryover.
RELATED PROGRAMS - ?
REFERENCES
1. Sinor, J. E. Evaluation of Background Data Relating to New Source
Performance Standards for Lurgi Gasification. EPA-600/7-77-057.
June 1977.
243
-------
TABLE B-3. RECTISOL PRODUCT GAS STREAMS
Constituents/
Parameters
H2
CO
CH4
co2
N2 + Ar
H2S
COS
cs2
RSH
Thiophene
MeOH
Temp: °K (°F)
Pressure:
MPa (psia)
Rate: Nm3/hr
(SCFM)
Stream No.
Type A Type A Type B Type B Type C
57.30
28.40
11.38
0.93
1.77
0.05 mg/Nm3
total sulfur
—
—
—
--
--
288(59)
2.3(345)
263,000
(163,000)
74.8
0.38
0.25
60 ppm
24.57
—
—
—
—
—
--
—
2.2(327)
118,500
(73,500)
94.08
4.86
0.24
10 ppm
0.82
--
--
—
--
—
__
--
3.0(450)
94,040
(34,300)
94.92
3.94
0.47
50 ppm
0.67
1 ppm
—
--
—
—
--
—
6.9(1037)
88,530
(54,890)
93.58
6.06
0.19
—
0.17
—
--
--
—
—
._
295(72)
2.9(440)
54,500
(33,800)
244
-------
TABLE B 4. RECTISOL OFFGAS STREAMS
Stream Number Reference
Constituents/
Parameters
H,
CO
CH4
CO,
H2 + Ar
H2S
COS
C2*
MeOH
cs2
RSH
Thlophene
Temp: °K (°F)
Pressure:
MPa (psla)
Rate: Nm3/hr
(SCFM)
Type A(3) Type B(3) Type B(3) Type B(<) Type C(4)
897 8 9 789 789 11 9 10 8
0.4
0.014
0.017
73.95
25.62*
—
—
-.
—
„
..
—
—
0.1(15) --
45.090 —
(27.956)
0.15 0.79
0.04 0.22
0.05
76.81 98.91 64.6
23.0* 0.05 0.1
2 ppm 2 ppm 35.2
0.1
..
—
—
—
--
..
0.1(15) 0.24(36) 0.24(36]
41.480 14.130 1980
(25.845) (8.760) (1230)
0.76 — —
0.11 -- --
0.06 -- —
90.85 — 68.31
8.22* -- . 1.92
5 ppm -- 29.77
„ ..
„ „
„
.. -_
—
„
..
0.1(16) — 0.2(28)
50.280 - 2390
(31.170) (1480)
0.33 — —
0.14 -- --
0.00 -- --
80.19 — 68.46
19.34* -- -
<5 ppm .— 30.78
8 ppm — 0.76
-. —
— —
„ —
-. —
-. —
295(72) -- 322(121)
0.1(15) — 0.5(73)
30.800 -- 673
(19.100) (417)
21.4 2.6 0.14
18.2 4.8 0.0
11.4 7.2 0.9
46.7 83.4 97.2
1.5 0.8 0.03
3176 ppm 4941 ppm 8824 ppm --
0.003
0.7 1.1 0.7
—
0.0002
0.028
0.0002
273(32) 273(32) 26a(23)
1.3(195) 0.46(70) 0.1(15)
4500 15.000 98.000
(2852) (9.300) (60.760)
to
lU
en
'Includes «2 stripper gas.
-------
2. Kohl, A., and Riesenfeld, F. Gas Purification. Gulf Publishing Co.
Houston, Texas. 1974.
3. Scholz, W. H. Rectisol: A Low-Temperature Scrubbing Process for
Gas Purification. Advances in Cyrogenic Engineering, 15. 1969.
4. Draft: Standards Support and Environmental Impact Statement
Volume 1: Proposed Standards of Performance for Lurgi Coal Gasi-
fication Plants. November 1976.
5. South African Coal, Oil & Gas Corp., Ltd. Information Provided to
the Fuel Process Branch of EPA's Industrial Environmental Research
Laboratory, Research Triangle Park, N.C. November 1974.
6. Maddox, R. N. Gas and Liquid Sweetening. Campbell Petroleum
Series. 1974.
(DUAL-ABSORPTION MODE)
GENERAL INFORMATION
1. Operating Principles: physical absorption of acid gases (C02, H2S,
COS, CS2, etc.) using methanol. When operated in the dual-absorp-
tion mode, C02-saturated methanol is used in the first absorption
step to remove H2S and other sulfur compounds. In the second
absorption step, pure methanol is used for the absorption of C02.
2. Development Status: commercially available.
3. Licensor/Developer: Lurgi Mineralo'technik GmbH
American Lurgi Corporation
377 Rt. 17 South
Hasbrouck Heights, N.J.
4. Commercial Applications: a Rectisol of this type is installed at
Modderfontein, South Africa, for purification of synthetic gas from
coal for manufacture of ammonia.
PROCESS INFORMATION
1. Flow Diagram: see Figure B-4.l
Process Description: C02 and H2S are absorbed in separate
columns with CO shift occurring between operations. In essence,
two separate Rectisol units, each with its own stripper column
(but with common still and regenerator) are employed. C02-
saturated methanol is used to absorb H2S in the first absorber.
Pure methanol from the regenerator is used in the C02 absorber.
2. Equipment: conventional absorbers, stripping columns, distillation
columns, heat exchangers, and knockout drums.
Construction: vessels may be fabricated from carbon steel;
dimensions depend on application.
3. Feed Stream Requirements: ?
4. Operating Parameters1 2 3
Absorption: H2S: 297 K (75° F) 3.0 MPa (440 psia).
C02: 213 K (-75° F) 4.9 MPa (720 psia).
Regeneration: ?
246
-------
to
lU
' LEGEND
1. RAW GAS
3. N. STRIPPER GAS
J. Nj STRIPPER GAS
4 MiOH MAKE-UP
S. H,S SCRUBBER GAS
C. l(frOT TO CO, REMOVAL
7. PRODUCT GAS
8. LEAN H,S FROM NO. 1 STRIPPER
9. LEAN H-S FROM NO. 7 STRIPPER
10. COMBINED LEAN HjS
11 CONCENTRATED HjS
»1 PURE CO,
13. CONDENSATE
14. CO, SATURATED METHANOL
16 PURE METHAIMOL
IS. LEANMETHANOL
Figure B-4. Rectisol—dual-absorption flow diagram (as installed at
Modderfontein. South Africa).
-------
5. Process Efficiency and Reliability: removal of acid gases to a few
micrograms per cubic meter. Reliability is high because of rela-
tively simple operation.
6. Raw Material Requirements
Solvent: methanol
7. Utility Requirements: utility requirements are high because of
large refrigeration requirements. Exact amounts are unknown.
8. Miscellaneous: ?
PROCESS ADVANTAGES
A single solvent (methanol) is used for absorption of both C02
and H2S.
Noncorrosive environments.
H2S streams rich enough to be processed in a Claus unit can be
obtained.
Good selectivity between acid and product gases.
Unlimited solubility of solvent in water.
Solvent is chemically stable and has a low freezing point.
PROCESS LIMITATIONS
Solvent retains heavy hydrocarbons.
Solvent losses during regeneration may be high.
High utility requirements.
PROCESS ECONOMICS - ?
INPUT STREAMS
Stream data are based on the Modderfontein plant.
1. Gaseous
A. Stream No. 1
Composition, wt % Ref. 1 Ref. 2
C02 11.6 13.37
CO 55.02 54.45
H2 31.2 30.00
N2 1.0 0.95
Ar 0.5 0.54
CH4 0.1 0.10
H2S 0.5 0.59
(includes COS)
COS 0.8
MeOH 0 0
Volume Nm3/(scfm) 91,700
(53,370)
Pressure, MPa (psia)
Temperature, K (°F)
B. Stream Nos. 2 and 3: nitrogen from air separation plant, rate
unknown.
248
-------
2. Liquid
A. Stream No. 4: methanol
INTERMEDIATE STREAMS
1. Gaseous
A. Stream No. 5
Composition, wt %
C02
CO
H2
N2
Ar
CH4
H2S
COS
MeOH
Volume, Nm2/hr (scfm)
Pressure, MPa (psia)
Temperature, K (°F)
makeup, rate unknown.
B.
Stream No. 6
Composition, wt %
Pressure, MPa (psia)
Temperature, K (°F)
C. Stream No. 8: ?
D. Stream No. 9: ?
DISCHARGE STREAMS
1. Gaseous
A. Stream No. 7
Composition, wt %
C02
CO
H2
N2
Ar
CH4
H2S
COS
MeOH
Volume, Nm3/hr (scfm)
Pressure, MPa (psia)
Temperature, K (°F)
Ref. 1
12.00
54.60
31.80
1.00
0.50
0.10
93,300
(58,370)
3.0(440)
298(75)
Ref. 1
C02
CO
H2
N2
Ar
CH4
H2S
COS
MeOH
Volume, NmVhr (scfm)
41.30
3.00
54.64
0.70
0.30
0.06
—
--
--
140,000
(87,590)
5.0(735)
308(95)
Ref. 1
4.60
93.50
1.20
0.60
0.10
80,000
(50,110)
4.9(720)
213(-75)
Ref. 2
11.27
56.02
31.06
0.98
0.57
0.10
Ref. 2
41.29
3.00
54.63-
0.64
0.37
0.07
Ref. 2
5.02
93.14
1.12
0.61
0.11
249
-------
B. Stream No. 10: ?
C. Stream No. 12: mostly C02, trace constituents unknown.
D. Stream No. 11
Composition, wt % Ref. 1 Ref. 2
C02 75.00
CO
H2
N2
Ar
H2S 22.00
COS 3.00
MeOH
Volume, NmVhr (scfm) 21,000
(13,140)
Pressure, MPa (psia) 0.1(15)
Temperature, K (°F) 313(105)
2. Liquid
A. Stream No. 13: ?
DATA GAPS AND LIMITATIONS
Limitations in the data for the selective absorption Rectisol
relate primarily to the stream compositions. These limitations include
the following:
Input gas streams: few data on minor component concentrations.
No data on N2 stripper gas rates.
Makeup methanol: no data on amount of makeup methanol required.
Intermediate and product gas streams: limited data on minor
components.
Discharge gas streams: limited data on compositions of off-
gas streams from the strippers and regenerator.
Condensate stream: no data on compositions and rates of
regenerator condensate stream.
Operating parameters: utility requirements, regeneration
parameters, etc., are not reported.
RELATED PROGRAMS
No known programs are presently undertaken to assess the discharges
from this process.
REFERENCES
1. Staege, H. Ammonia Production on the Basis of Coal Gasification.
Chemical Industry Developments. 1973.
2. Schellberg, W. Coal-Based Ammonia Plants. ICI Operating Symposium
Paper 21. 1974.
3. Goeke, E. K. Status of Coal Gasification Technology. FAI Symposium
on Coal as Feedstock for Fertilizer Production. New Delhi. 1974.
250
-------
ENVIRONMENTAL ASSESSMENT REPORT FOR WELLMAN-GALUSHA
GASIFICATION SYSTEMS
William C. Thomas* and Gordon C. Page
Radian Corporation, Austin, Texas
Abstract
Radian Corporation has just entered the
fourth year of a 6yr contract with the U.S. En-
vironmental Protection Agency (EPA) to per-
form a comprehensive environmental assess-
ment of low- and medium-Btu coal gasification
technology. As part of that program, Radian
has conducted a number of source test and eval-
uation programs at operating low-Btu gasifica-
tion facilities in the United States. The results
of those test programs, along with data coir
lected from the open literature, vendors, process
licensors, and other industry contacts, have
been incorporated into an Environmental
Assessment Report (EAR) for Wettman-Oa-
lusha low-Btu gasification systems. This paper
presents the preliminary results and findings of
the Wellman-Oalusha EAR work. Included are:
• An overview of Welbnan-Galusha low-Btu
gasification systems,
• Identification of waste streams and pollut-
ants of major concern,
• The status of environmental protection
alternatives,
• Future data needs and recommendations,
and
• Issues and areas of concern of EPA program
offices.
INTRODUCTION
In March 1976, Radian Corporation entered
into a contract with the U.S. Environmental
Protection Agency (EPA) to perform a compre-
hensive environmental assessment of low- and
medium-Btu coal gasification technology. Orig-
inally a 3-yr effort, the low-Btu program has
recently been extended for an additional 3 yr.
Both the original program and the extension are
being directed by the Fuel Process Branch of
EPA's Industrial Environmental Research Lab-
oratory (IBRD in Research Triangle Park,
•Speaker.
North Carolina.
The initial activity of Radian's low-Btu assess-
ment program involved a comprehensive infor-
mation search aimed at compiling a data base
for low- and medium-Btu gasification technolo-
gy. While a significant amount of data was ob-
tained, data gaps and areas of questionable or
incomplete data were still identified.
In order to obtain the missing data, the sec-
ond major phase of the low-Btu program—data
acquisition—was initiated. This phase involved
conducting source test and evaluation programs
at a number of operating gasification facilities.
To date, data acquisition test efforts have been
conducted at three low-Btu gasification facilities
in the United States and a medium-Btu facility
in Yugoslavia.
The main purpose of this paper is to present a
part of the results of the third phase of the low-
Btu assessment program; Le., the results com-
munication phase. Several documents will be
prepared over the next 3 yr in order to com-
municate the assessment program's results and
findings. One type of document is the environ-
mental assessment report, or EAR. The pur-
pose of an EAR is to provide EPA administra-
tors, program offices, and policy and planning
with a document that represents the EPA Office
of Research and Development's research input
to standard-setting activities for gasification
facilities. Each EAR addresses a unique seg-
ment of gasification technology. An EAR in-
cludes a detailed evaluation of process, waste
stream, and control data collected from field
testing programs; open literature; vendors;
process licensors; and computer modeling activ-
ities. As such, an EAR is a data base for the sub-
ject technology.
In 1978, Radian initiated preparation of the
first of four environmental assessments reports
that will be prepared over the next 3 yr. This
EAR addresses Wellman-Galusha low-Btu gasi-
fication systems. Incorporated into the Well-
man-Galusha EAR are the process, waste
stream, and control data collected by Radian at
251
-------
the three U.S. test sites. The preliminary
results and findings of the Wellman-Galusha
EAR are presented in the following text.
OVERVIEW OF WELLMAN-GALUSHA
LOW-Btu GASIFICATION SYSTEMS
Wellman-Galusha gasifiers are one of the
commercially available gasifiers used to pro-
duce low-Btu (-5.9 x 108 J/Nm8, 150 Btu/scf)
gas from a variety of coal feedstocks. The
Wellman-Galusha gasification systems exam-
ined in this report are described along with
their status, industrial applicability, energy effi-
ciency, costs, and commercial prospects.
System Description
Wellman-Galusha low-Btu gasification sys-
tems have three basic operations: coal pretreat-
ment, coal gasification, and gas purification.
Each operation includes processes with specific
functions, inputs, and outputs. Figure 1 is a
generalized flow diagram showing various com-
binations of operations and process modules for
Wellman-Galusha gasification systems. Table 1
summarizes the input and output streams and
the function associated with each process.
In this study, four coal feedstocks and three
product gas specifications were considered:
• Coal feedstocks
Anthracite (0.6% S; 11.7% ash)
Low-sulfur HVA bituminous (0.7% S; 2.9%
ash)
High-sulfur HVA bituminous (3.9% S; 8.4%
ash)
Lignite (0.9% S; 8.3% ash)
• Product gas specifications
Can meet current NSPS for direct coal com-
bustion
Can meet proposed NSPS for direct coal
combustion
"Very clean" gas
Combinations of these coals and product gas
specifications were selected as the study bases.
Those combinations resulted in three basic gasi-
fication systems being considered. The first
system is typical of what would be required to
produce fuel gas capable of complying with cur-
rent New Source Performance Standards
(NSPS) for direct combustion of low-sulfur coals
(>0.7% S; HHV 30 MJ/kg or 13,000 Btu/lb). This
system has only three process modules: coal
handling and storage, gasification, and particu-
late removal (hot cyclone). This system also rep-
resents currently operating facilities that use
anthracite and low-sulfur HVA bituminous
coals.
A variation of this first system has an addi-
tional process module: raw gas quenching and
cooling. This additional module removes tars
and oils from the raw product gas and reduces
the potential for fouling of equipment used to
transport the low-Btu product gas to its end use.
This system is similar to a facility using Chap-
man (Wilputte) gasifiers to produce a low-Btu
combustion gas for process heaters.
The second Wellman-Galusha gasification
system is used to produce a clean gas from an-
thracite coal. This system contains the following
process modules: coal handling and storage, gas-
ification, gas quenching and cooling, and sulfur
removal. In this system, the product gas is
cooled to 316 K (110° F) before entering the
Stretford sulfur-removal process. The Stretford
sulfur-removal process is effective in removing
H2S, but organic sulfur species (i.e., COS, CS2,
etc.) will essentially remain intact in the product
gas stream.1 H2S removal efficiencies greater
than 99 percent have been achieved with residu-
al outlet H2S concentrations less than 10 ppmv.2
An advantage of the Stretford process is that it
not only removes H2S but also converts the H2S
into elemental sulfur, which can be recovered as
a byproduct.
The third system is used to produce a clean
gas from the following coal feedstock: bitumi-
nous (HVA, or low-sulfur, and HVB, or high-sul-
fur) coal and lignite. In this system, the
quenched and cooled product gas is sent to a
tar/oil removal process followed by a sulfur
removal process. An electrostatic precipitator
(ESP) is used to remove tars and oils that would
cause operating problems with the downstream
sulfur removal process. As in the second sys-
tem, the Stretford process was chosen for re-
moval of sulfur species from the product gas. In
addition, the Monoethanolamine (MEA) process
was examined. The MEA process is capable of
removing both H2S and organic sulfur com-
pounds. However, the sulfur removal efficiency
is dependent upon the pressure of the product
gas. For example, at 0.44 MPa (50 psig) residual
sulfur concentrations of 16 ppmv can be
achieved. At a higher pressure of approximate-
ly 1.5 MPa (200 psig), residual sulfur concentra-
252
-------
•^ MPFUI i ran
T TlT
MB. OILS.
©
©
©
©
•Mtfmm « U»-*UIB cwu (u»iMi«.n am* m nu »n» mm MUD «f * w/« « u.on nwu) TQ nma « no. «• «u 10 caur mm o«ar an ran* COBWTHP or com.
em CMLS re mna * no. on nun aou «n« nmo en ni m COMSTHI OF WL: Mjinarioi OF namtcm conu ro mnn » -oar MS
MOB. as-tmuna. «• iHan am re PMHCC > no. MS «u n avir Him mwa an rat m CMIIMIIOH OF can. MSIFIOTIO OF
amn «• ncraa nn «• m COMMTW OF CML; Msinaniai OF «u OMnwcrre cov 10 FMHCC * -CLEW MS.
MI mi i* ctn.i i
Figure 1. Wellman-Galusha system process modules and multimedia discharges.
-------
to
TABLE 1. OPERATIONS/PROCESS MODULES IN WELLMAN-GALUSHA LOW-Btu
GASIFICATION SYSTEMS
Operation/Process Module
Input Serena
Output Streams
Function
Remarks
Coal Pretreatment
Coal Rand Hug
and Storage
Coal Gasification
Fixed-Bed, Atmospheric
Pressure, l>ry Ash
GasIftar - Uullmaa-
Galuaha
Gas Purification
Presized coal
Preslred coal
Steam
Air
Ash sluice water
Presizcd coal
Coal dust
Coal pile runoff water
Raw product gaa
Coal hopper gaaes
Fugitive gases
Start-up vent gases
Ash
Ash sluice water
Store and transport
coal feedstock
React coal with •
Mixture of steam and
air to produce a raw
low-Btu gas
Coal storage piles would contain a 30 day coal
supply (2-12 Gg, 2000-13.000 short tons of coal
for a plant producing 18-88 Ml, 60-3OO million Btu/
hr of low-Btu gas).
Coals that have been used Include anthracite and
bltuiilnous. Coal size specifications are 7.9 to
14.3 a* for anthracite and 26-51 mm for bituminous.
Larger particle siges can be used for more reactive
coals.
Paniculate Removal -
Hot Cyclone
Gas Quenching
and Cooling
Tar /Oil Removal -
Electrostatic
Preclpltator
Sulfur Humovul -
Stretford
Sulfur Removal -
Monoethanolamine
Process
Haw product gas Product gas
Removed partlculates
Product gas Quenched/cooled
Quenching liquor product gas
Quench liquor
Tars
Oils
Partlculate Batter
Cooled product gas Cooled/detarred
product gas
Tars
Oils
Detarred product gas Clean product gas
Stretford solution Oxidize* vent gas
Air Sorbent blowdown
Sulfur
Detarred product gas Clean product gas
MEA solution NEA blowdown
Acid gases
Sulfur from acid gaa
treatment processes
Tall gases from acid
gas treatment processes
Remove large partlcu-
late matter from the
hot. raw product gas
Remove tars and oils
from the product gaa
and cool the product
gas to approximately
316«K (1WF)
Remove tar and oil
aerosols from the
cooled product gas
Remove HiS from the
detarred product gas
Remove sulfur species
and CO j from toe
detarred product gas
Total partlculate removal efficiencies have been
determined to be between 50-BOZ. Small partlculate
matter will not be removed. Collected partlculates
have characteristics similar to devolatllized coal
particles.
The amount of tars and oils removed Is dependent
upon the coal feedstock. Anthracite coal will pro-
duce essentially no tars, however, bituminous coal
will produce a significant amount of tars.
Emissions from the tar/liquor separator may contain
potentially hazardous compounds. Spent quench
liquor will require treatment before disposal.
ESP's have been used to remove tars and oils pro-
duced by two-stage, fixed-bed, atmospheric gasiflers
and good removal of tars and oils have been demon-
strated by ESP's used In sampling systems.
Vent gases from tar/oil storage tanks may contain
potentially harmful compounds and may need to be
controlled.
Other sulfur species (I.e.. COS, CSj, etc.) will not
be removed from the product gas. If the HCN concen-
tration Is high, then a cyanide guard may be needed.
Blowdown sorbent will require treatment before dis-
posal. If the sulfur IB to be disposed of. tests
need to be performed (I.e., RCRA tests for solid
niques required.
Removal efficiency Increases with Increasing inlet
gas pressure. Acid gases have to be treated to
control sulfur emissions. MEA blowdown will require
treatment before disposal.
-------
TABLE 2. CURRENT WELLMAN-GALUSHA COAL GASIFICATION FACILITIES IN THE
UNITED STATES
Ol
Ol
Number of
Company/Locations Gasifiers
*
Glen-Gery Brick Co.
• York, PA 2
• Reading, PA 2
• Shoemakersvllle, FA 1
• Watsontown, PA 1
National Lime and 2
Stone Co.
• Carey, OH
Can Do, Inc. 2
• Hazel town, PA
Bureau of Mines 1
• Fort Snelling, UN
Pike County 2
• Pikeville, KY
Aluminum Refinery 1
• PA
Coal Feedstock
Anthraclde, low
sulfur fv 0.7Z)
Bituminous low
sulfur (•». 0.7Z)
Anthracite, low
sulfur
KY Bituminous
CO Subbltumlnous
WY or MT Sub-
bituminous NI)
Lignite; (all coals
low sulfur (<1Z S)
Bituminous, low
sulfur (0.6-1.2Z)
Anthracite, low
sulfur (•>- 0.7Z)
Gas Purification
Process
• Hot cyclone
•
• Hot cyclone •
.
•
• Hot cyclone •
•
• Hot cyclone •
• SOj scrubbers
(on combustion
gas)
•
• Hot cyclone
- Gas quench/cooling
• Tar/liquor separation
• Stretford sulfur
removal
• Dehydration
• Electrostatic
precipltators
• Hot cyclone
Remarks
Currently operational
Product gas used to fire brick kilns
Currently operational
Product gas used to fire lime kiln
Lime partially removes sulfur from gas
Product gas to supply industrial park
To be operational In 1979
Product gas to fire pilot iron pelletizing
kiln
Excess product gas will be combusted
Operational in 1978
To be operational in 1979
Product gas to fire boilers and process
heaters
To be operational in 1979
f
• Five gasifiers may be added later
-------
tions of less than 4 ppmv are attainable.3 4 As
mentioned above, the Stretford process also
converts the removed sulfur species into ele-
mental sulfur. Unfortunately, the MEA process
does not have this advantage. Instead, it pro-
duces an acid-gas stream that requires further
treatment, for example, in a Glaus unit.
Status
Wellman-Galusha gasifiers have been com-
mercially available since 1941. Approximately
150 gasifiers have been installed worldwide. In
the United States, eight Wellman-Galusha gasi-
fiers are currently being used to produce a low-
Btu gas from anthracite and low-sulfur bitumi-
nous coals. Table 2 summarizes the locations,
processes, and coal feedstocks for each plant.
Industrial Applicability
Wellman-Galusha gasification systems have
been used to provide a low-Btu fuel gas and a
synthesis gas for ammonia production. A sum-
mary of past applications is given in Table 3.
In the near term, Wellman-Galusha gasifiers
will be used primarily to produce a fuel gas for
onsite use. Potential uses of the product low-Btu
gas include fuel to provide direct heat for such
processes as brick and lime kilns; fuel for small
industrial boilers; and, possibly, synthesis gas
for ammonia production. Production of gas for
offsite use will probably not be significant be-
cause of the cost of transporting atmospheric
pressure, low-Btu gas.
Energy Efficiency
The energy efficiency of Wellman-Galusha
gasification systems will be a significant factor
affecting commercialization potential. However,
this factor becomes less critical when compared
to use of natural gas, which may be either un-
available or too expensive.
The following two kinds of energy efficiencies
are used to describe gasification systems:
(Qg) out
(Qc)in
x 100
and
(QT) out
(QT) in
x 100
TABLE 3. PAST USERS OF GAS PRODUCED BY WELLMAN-GALUSHA GASIFIERS8
• chemical plants
• glass plants
• steel mills
• magnesium manufacturers
• silk mills
• bakeries
• wire mills
• foundries
• potteries
• aluminum and stainless steel
manufacturers
• ordinance plants
• tin plate mills
• lime plants
• brick plants
• zinc smelters
• iron ore processors
• fertilizer plants
Specific uses varied from heat treating (in glass and steel
mills) to synthesis gas (for synthetic fertilizer manufacture)
Materials gasified included charcoal, coke, anthracite and
bituminous coals.
256
-------
where 7?cg is the coal/gas efficiency (in percent),
Tfr is the overall thermal efficiency (in percent),
(Qg)out is the output product gas energy, (Qc)in is
the input coal energy, (QT)out is the total output
energy (product gas + byproducts + steam),
and (Qt)in is the total input energy (coal +
steam + electricity.
Calculated energy efficiencies for the Well-
man-Galusha sysems considered in this report
are shown in Table 4. These calculated efficien-
cies show that the types of processes used,
byproducts produced, and the nature of the coal
feedstock affect the coal/gas and overall
thermal-energy efficiencies of the system.
Capital and Operating Costs
Capital investment requirements and oper-
ating costs were calculated for the following
Wellman-Galusha gasification systems produc-
ing 17.6 MW (60 x 106 Btu/hr) and 87.9 MW
(300 x 108 Btu/hr) of product low-Btu gas:
• System 1 produces a hot raw product gas.
• System 2 produces a desulfurized product
gas (down to 10 ppmv H2S but retaining all
organic sulfur) using a Stretford sulfur-re-
moval process.
• System 3 produces a desulfurized product
gas (-200 ppmv total sulfur) using an ME A
sulfur-removal process operating at 0.21
MPa (30 psia).
• System 4 produces a desulfurized product
gas (-10 ppmv total sulfur) using an ME A
sulfur-removal process operating at 1.6 MPa
(230 psia).
Tables 5 and 6 summarize the capital invest-
ment requirements and operating costs for
Wellman-Galusha gasification systems using
various coal feedstocks. The cost data shown
are for systems without environmental controls.
As shown in Tables 5 and 6, the product gas
costs are dependent upon coal feedstock, prod-
uct gas specifications (tar/sulfur content), and
plant size. Product gas costs for producing a hot
raw gas for onsite use (System 1) range between
$1.90 and $3.60 per GJ ($2.00 and $3.80 per 106
Btu) depending upon the coal feedstock and
plant size. For systems using a Stretford sulfur-
removal process, product gas costs range from
$3.30 to $5.60 per GJ ($3.50 to $5.90 per 106 Btu),
again depending upon the coal feedstock and
plant size. If an MEA sulfur-removal process is
used to remove gaseous sulfur species, product
gas costs would range from $3.80 to $620 per
GJ ($4.00 to $6.50 per 10° Btu) depending upon
the sulfur content of the clean product gas and
the plant size.
For most of the gasification systems, the ma-
jor component of the annualized costs is the coal
feedstock cost. For systems using anthracite
coal, the coal costs represent 36 to 56 percent of
the total costs of the product gas. For systems
using low-sulfur bituminous coal, coal costs are
between 36 and 70 percent of the product gas
costs, and for high-sulfur bituminous coal, coal
costs are 28 to 42 percent.
Commercial Prospects
In the near term, low-Btu gas from fixed-bed
atmospheric-pressure gasifiers like the Well-
man-Galusha will be used primarily as a substi-
tute fuel by industries threatened with natural
gas curtailments. The low-Btu gas will princi-
pally be considered for use as a fuel gas in onsite
furnaces, heaters, kilns, and small boilers. Its
substitution for natural gas will most likely oc-
cur when the costs of retrofitting for use of the
low-Btu gas are small and the low-Btu gas re-
quires minimal purification.
In both new and retrofit applications, coal
gasification is mainly competing with the alter-
native of direct coal combustion. Factors affect-
ing the selection of coal gasification or direct
coal combustion include the suitability of the
coal conversion technology for satisfying the
needs of the specific end use, the cost of the
technology, the cost and difficulty of retrofit-
ting, the cost of environmental controls, and the
cost of the coal.
The increased commercialization of low-Btu
gasification systems like the Wellman-Galusha
depends on the demonstration of the environ-
mental acceptability of the various gasification
systems. Although commercially available con-
trols seem to be adequate, some of the controls
(such as sulfur removal) have not been adequate-
ly demonstrated on coal gasification systems.
The cost of these controls are also uncertain.
Gasification systems featuring Wellman-Ga-
lusha gasifiers are most suitable for relatively
small applications, with fuel demands ranging
from about 8.8 to 88 MW (30 million to 300 mil-
lion Btu/hr). Energy demands greater than
about 88 MW may be better served by gasifica-
tion systems using gasifiers with larger capac-
257
-------
TABLE 4. CALCULATED ENERGY EFFICIENCIES OF VARIOUS WELLMAN-GALUSHA
GASIFICATION SYSTEMS
s
For Systems Producing a Hot
Coal Feed Typical Raw Coal/Gas
Type Gas Temperature Efficiency3, %
Anthracite 700°K (800°F)
Low-S HVA 840°K (1050°F)
Bituminous
High-S HVB 640eK (700°F)
Bituminous
Lignite 420°K (300°F)
91
93
NA8
NA8
Product Gas
Overall Thermal
Efficiency", %
90
92
NA8
NA8
For Systems Producing
Quenched/Clean Gas at
317°K (110°F)
Coal/Gas
Efficiency3, %
83C
68C
61d
62e
78C
Overall Thermal
Efficiency15, %
84C
82C
83C
72d
64/616'11
89 c
Coal/gas efficiency is calculated as: output product gas energy divided by input coal energy.
Overall thermal efficiency is calculated as: total output energy (product gas + by-products + steam)
divided by total input energy (coal +• steam . + electricity).
°These systems produce a cooled, cleaned product gas and feature the Stretford process for sulfur removal.
These systems produce a cooled, cleaned product gas (<200 ppmv total sulfur) by using an amine (MEA)
absorption process to remove sulfur species. In these systems, some of the product low-energy gas is
assumed to be used to meet the energy requirements of the amine process. Alternately, by-product tar may
be used to meet at least part of these energy requirements.
CThese systems produce a "very clean" gas (<10 ppmv total sulfur) by using an amine absorption system (MEA)
to remove sulfur species.
fThe first efficiency is for the 16MW (54.7xl06 Btu/hr) system which uses an electric motor to drive the
gas compressor. The second efficiency is for the 80.1 MW (273.5xl06 Btu/hr) system which uses a steam
turbine to drive the gas compressor.
^lot applicable - Given the coal quality data which were assumed for purposes of conducting this
assessment, these coals cannot be used in systems in which the raw product gas is burned directly.
-------
TABLE 5. CAPITAL INVESTMENT REQUIREMENTS AND ANNUALIZED COSTS OF
UNCONTROLLED WELLMAN-GALUSHA GASIFICATION SYSTEMS
PRODUCING NOMINALLY 17.6 MW (60 X 10* Btu/HR)
OF PRODUCT LOW-Btu GAS (LATE-1977 DOLLARS)8
Coal Feedstock/Type of Product Ca»
Anthracite
to
•S
Capital Investment Requirements , $1,OOO
Design Plant Capacity, MW
Annual Operating Factor
Annuallied Costs, $l,000/yr
Operating and Maintenance Costs
CoalS
Labor/Overhead (8 $15.00/man-hr)
Electricity (S $0.04/kWh)
Steam''
Chemicals
Maintenance (0 61 of direct equipment
costs)
Taxes, Insurance, and GSA Costs (9 4Z
of depreciable Investment)
Capital Related Charges*
TOTAL Annual ized Costs, $10*/yr
Average Gas Costs, $/CJ
Hot Gas"
3,250
19.9
90Z
1,040
131
16
-
149
117
602
2.055
3.64
Cold Gasc
6.110
18.3
90Z
1,040
197
48
(17)
8
276
229
1,116
2,897
5.58
Low Sulfur Bituminous
Hot Gasb
1,730
24.9
90Z
919
66
18
-
74
58
326
1,461
2.07
Cold Gas*
5,200
18.2
90Z
919
131
79
8
233
194
950
2,514
4.87
High Sulfur Bituminous
Stretfordc
5,500
18.2
90Z
702
131
118
63
248
207
1,003
2,472
4.78
HEA (200 ppmv)0
3,890
15.7
90S
702
131
225
55
175
143
715
2.146
4.82
(Cold Gas)
MEA (neg.)'
4,700
15.9
90Z
702
131
643
55
210
171
867
2,779
6.16
"Each system has a basic capnclty of 17.6 MW (60 x 10* Btu/hr) of tar/oil-free product gns at 43.3*C (110'F). The actual total energy supplied to the end-
user though Is as Indicated. Difference!! in the indicated useful energy supplied and the basic capacity of 17.6 MW (60 x 10* Btu/hr) are a result of
1) energy credits taken for the sensible heat and/or tar/oil content of the product gas for the hot gas systems, and 2) use of a portion of the product
gas to supply energy to the stripper reboller In the systems that use the MEA process.
These systems use only a cyclone for product gas purification and deliver a hot product gas to the end user.
cTheso systems use the Stretford process to remove HjS from the cooled product gas. Residual H?S levels are nominal 10 ppmv. Organic sulfur compounds,
such as COS and €87t are not removed by the Stretford process.
This system uses the MEA process operating at 0.21 MPa (30 psla) to remove sulfur species from the cooled product gas. Residual sulfur species amount to
the equivalent of 200 ppmv H7S.
'tills system uses the HF.A process operating at 1.6 MTa (230 psla) to remove sulfur species from the cooled product gaa. Negligible sulfur species are left
In the product gas.
In estimating capital Investment requirements, a spare gasifler/cyclone unit la Included for all systems and cooling liquor pumps are spared 1001.
*Assumed coal properties and delivered costs are: Anthracite: 29.7 Hi/kg (12,800 Btu/lb) and $50/metrlc ton ($45/short ton) .
Low sulfur bituminous: 33.2 MJ/kg (14,300 Btu/lb) and $40/metrlc ton ($36/short ton)
High sulfur bituminous: 29.0 Hi/kg (12,500 Btu/lb) and $28/metrlc ton ($25/ahort ton)
hSteam costs vere assumed to be $0.Oil/kg ($5/10' Ib). Steam credits were taken as $1/GJ ($1.05/10* Btu).
for capital related charges: Utility financing method 1001 equity financing
Uitr-1977 dollars without infl.itlon 15Z after tax return on equity
25-year economic project lifetime 46Z federal Income tax rate
4Z per year straightline depreciation 10Z pretax return on working capital
of depreciable Investment
-------
TABLE 6. CAPITAL INVESTMENT REQUIREMENTS AND ANNUALIZED COSTS OF
UNCONTROLLED WELLMAN-GALUSHA GASIFICATION SYSTEMS
PRODUCING NOMINALLY 87.9 MW (300 X 106 Btu/HR)
OF PRODUCT LOW-Btu GAS (LATE 1977 DOLLARS)8
Coal Feedstock/Type of Product Cas
Capital Investment Requirements^ $1.000
Design Plant Capacity, MU
Annual Operating Factor
Annual lied Costs, $1,000 /yr
Operating and Maintenance Costs
Coal*
Labor/Overhead (9 $15.00/man-hr)
Electricity (9 $0.04/kWh)
Stem."
Chemicals
Maintenance (9 6Z of direct equipment
costs)
Taxes, Insurance, and CSA Coats (9 4Z
of depreciable investment)
Capital Related Charges1
TOTAL Annuallzed Costs, $103/yr
Average Gas Costs, $/GJ
ft"
Hot Gas"
13,300
99.6
90Z
5,198
524
81
-
596
468
2,476
9,343
3.30
thraclte
Cold Gasv
19,700
91.4
90Z
5.198
657
238
(86)
40
871
713
3,640
11,271
4.34
Low Sulfur
Hot Gasb
4,770
99.7
90Z
3,676
263
72
-
189
149
916
5,265
1.86
Bituminous
Cold Gas1-
13,100
91.1
90Z
4,595
394
396
40
563
465
2,436
8,889
3.44
High Sulfur
Stretfordc MEA
14,200
90.8
90Z
3,510
394
590
315
617
512
2,614
8,552
3.32
Bituminous
(200 ppmv)d
11,600
78.5
90Z
3,510
394
1,125
274
499
406
2,165
8,373
3.76
(Cold Gas)
MEA (neg.)E
14,000
79.7
90Z
3,510
394
334
3.39O
274
582
474
2,625
11,583
5.12
Each system, except the one producing a hot product gas from low sulfur bituminous coal, has a basic capacity of 87.9 HW (300 x 10* Btu/hr) of tar/oil-
free product gas at 43.3'C (110'F). The actual total energy supplied to the end-user though is as Indicated. Differences In the Indicated useful energy
supplied and the basic capacity of 87.9 MW (300 x 10* Btu/hr) are a result of 1) energy credits taken for the sensible heat and/or tar/oil content of the
product gas for the hot gas systems, and 2) use of a portion of the product gas to supply energy to the stripper reboller In the systems that use the MEA
process. For the hot gas, low sulfur bituminous system, the tar/oil-free product gas rate Is 70.3 MW (240 x 10* Btu/hr). But, the sensible heat and
tar/oil content of the hot product gas raise the total system capacity to 100 MW (341 x 10* Btu/hr). This capacity was used In the cost analysis because
It Is comparable to the capacity of the other systems examined.
These systems use only a cyclone for product gas purification and deliver a hot product gas to the end user.
°These systems use the Stretford process to remove HjS from the cooled product gas. Residual H7S levels are nominal 10 ppmv. Organic sulfur compounds,
such as COS and r.S2, are not removed by the Stretford process.
This system uses the UFA process operating at 0.21 MT.i (30 psla) to remove sulfur species from the cooled product gas. Residual sulfur species amount to
the equivalent of 200 ppmv IIZS.
CThls system uses the MEA process operating at 1.6 MPa (230 psla) to remove sulfur species from the cooled product gas. Negligible sulfur species are left
In the product gas.
In estimating capital Investment requirements, a spare gaslfler/cyclone unit Is Included for all systems and cooling liquor pumps are spared 100Z.
BAssumed coal properties and delivered costs are: Anthracite: 29.7 HJ/kg (12,800 Btu/lb) and $50/metrlc ton ($45/short ton)
Low sulfur bituminous: 33.2 HJ/kg (14,300 Btu/lb) and $'iO/mctrlc ton ($36/short ton)
High sulfur bituminous: 29.0 HJ/kg (12,500 Btu/lb) nnd $28/metrlc ton ($25/short ton)
''steam costs were assumed to be $0.Oil/kg ($5/10] Ib). Steam credits were taken as $1/OJ ($1.05/10* Btu).
Basis for capital related charges: Utility financing method 100Z equity financing
Lato-1977 dollars without inflation 15Z after tax return on equity
25-year economic project lifetime 46Z federal Income tax rate
4Z per year stralghtllne depreciation 10Z pretax return on working capital
n( depreciable investment
-------
ities (for example, pressurized gasifiers).
Systems featuring two to four gasifiers and
gas purification facilities will require 18 to 24
mo from initial feasibility studies to full-scale
operation.5 McDowell-Wellman can deliver
Wellman-Galusha gasifiers 6 to 8 mo from the
date of order.'
Wellman-Galusha gasification systems will be
most widely applied in the industrial areas of
the Northeast and Midwest. States in those
regions have large reserves of bituminous coal.
WASTE STREAMS AND POLLUTANTS
OF MAJOR CONCERN
Wellman-Galusha low-Btu gasification sys-
tems are sources of gaseous, liquid, and solid
waste streams. Also associated with these sys-
tems are process and byproduct streams that
may contain toxic substances. The multimedia
waste streams and pollutants of major concern
are summarized in Tables 7 through 9. Process
and byproduct streams that may contain po-
tentially toxic compounds are summarized in
Table 10.
Gaseous emissions from Wellman-Galusha
systems contain a significant amount of
pollutants that may have harmful health and
ecological effects. Gaseous pollutants (CO, H2S,
HCN, NH3, and light hydrocarbons) from the
coal feeder and gasifier pokeholes need to be
controlled. Startup vent gases will contain com-
pounds found in the raw product gas (CO, sulfur
species, light hydrocarbons, tars, and oils),
which will require control before venting to the
atmosphere. Vent gases from the byproduct tar
recovery process will contain significant
amounts of potentially harmful pollutants and
will, therefore, need to be controlled. Emissions
from sulfur removal processes are not yet char-
acterized since there are currently no sulfur
recovery processes being used with fixed-bed,
atmospheric pressure, low-Btu gasification
systems.
The amount of liquid effluents from Wellman-
Galusha systems will be limited to blowdown
streams, ash sluice water, and coal pile runoff.
Of these effluents, the blowdown streams will
contain significant quantities of potentially
harmful constituents. Ash sluice water and coal
pile runoff will contain compounds leached from
the ash and coal, which may effect health and
the environment.
Solid waste streams from Wellman-Galusha
systems will consist of ash, collected particu-
lates, sulfur, and blowdown from the MEA sul-
fur-removal process. Ash and sulfur may con-
tain leachable constituents that may be poten-
tially harmful. Collected particulates resemble
devolatilized coal and therefore may be classi-
fied as a solid combustible material. MEA blow-
down sludge contains potentially harmful con-
stituents and needs to be treated before dis-
posal.
The byproduct tar and quench liquor repre-
sent process streams that contain partially
harmful organic and inorganic compounds.
Worker exposure and accidental releases of
these streams should be carefully controlled.
It should be emphasized that the chemical
characteristics and potential biological effects
of the various streams present in a gasification
facility are highly dependent upon the coal feed-
stock and processes used. For example, tars will
not be produced when anthracite coal is gasified;
however, process condensate may contain light
oils.
STATUS OF ENVIRONMENTAL
PROTECTION ALTERNATIVES
The assessment of the status of environmen-
tal protection alternatives involves identifying
and evaluating control alternatives to deter-
mine the most effective control alternatives and
the costs and energy impacts of those alter-
natives. The secondary waste streams from the
most effective control alternatives are also com-
pared to existing and proposed regulations and
to the multimedia environmental goals (MEGs).7
Most Effective Control Alternatives
The criteria used to identify the most effec-
tive control alternatives are applicability to
treating waste streams from low-Btu gasifica-
tion systems, control effectiveness, develop-
ment status, and secondary waste streams.
Costs and energy considerations are not con-
sidered in the selection of the most effective
controls. Table 11 shows the most effective con-
trol alternatives to treat the multimedia waste
streams and potential toxic substances asso-
ciated with Wellman-Galusha gasification sys-
tems.
261
-------
TABLE 7. GASEOUS WASTE STREAMS AND POLLUTANTS OF MAJOR CONCERN FROM
WELLMAN-GALUSHA LOW-Btu GASIFICATION SYSTEMS
Operation
Process
Gaseous Haste Stream
Pollutants of Major Concern
Remarks
Coal Preparation
Coal Storage and
Handling
Coal Gasification
Vellman-Galusha
Gaslfier
Coal dust
Coal feeder vent
gases
Start-up vent gases
to
Fugitive emissions
(pokehole gases)
Participate matter similar in composition to the
coal feedstock.
Gaseous species in the product gas (CO. HzS,
IICH, light hydrocarbons).
NHj,
Raw product gas constituents. Partlculate matter
(coal dust, tar. oil aerosols) and gaseous
species (CO, H2, H2S. COS. HH,. HCN, light hydro-
carbons, etc.). Organlcs of concern include fused
aromatic hydrocarbons, heterocyclic nitrogen,
sulfur and oxygen compounds, carboxylic acids,
amines, sulfonlc acids, sulfoxldea, phenols,
thiols, benzene, and substituted benzene hydro-
carbons. Inorganics of concern include CO,
ethylene. Cr, Hg. U, V. Al. P, As, Cu. Cd. H2S,
CO,, HCN, Li. Tl, Si, Pb. Sb, SO,, CS,, Cl, Tl.
Zr, Fe, Co, Hi, Ag and Zn.
Caseous species in the product gas (CO, HjS, HH>,
11CH, light hydrocarbons).
Bituminous coal gave slightly positive results
for the Ames test. Anthracite coal results were
negative.
High levels of CO were found in the coal hopper
area.
The amount of tars and oila will depend upon the
coal feedstock. Bituminous coals will have a
significant amount of tars where anthracite will
not. Tars from the gasification of bituminous
coals gave positive results on the Ames test.
Emissions of tars and oila will occur when poke-
hole valves are open; however, the major emissions
from the pokeholes will be from gaseous species
in the product gas leaking from the pokehole
valves.
Gas Purification
Gas Quenching and
Cooling (Tar/
Liquor Separation)
Sulfur Removal-
Stretford
Sulfur Removal-
NEA
Separator vent gases
Evaporator and
oxidizer vent gases
Acid gas stream
Organlcs of concern Include fused aromatic hydro-
carbons, amines, heterocyclic nitrogen and sulfur
compounds, ethylene, phenols, methane, and
carboxylic acids. Inorganics of concern include
CO, NHi, NOj, C02. Cr, Ag. V, Cu, P, Li, As, Fe.
Nl, and U.
Volatile compounds In the Stretford liquor (H2O,
CO2, Hi, 02, and possibly Mil)).
CO2, H2S, COS, CS2. mercaptans, and light
hydrocarbons.
These pollutants of concern are associated with
bituminous coals.
This stream has not been sampled because no
Stretford processes are currently used to remove
sulfur species from low-Btu gas.
This stream is sent to a sulfur recovery unit
consisting of a Claus process followed by a Claus
tail gas clean-up process to remove the sulfur
species in the acid gas stream. This stream has
not been sampled since MCA processes have not
been used to remove sulfur species from low-Btu
gas.
-------
TABLE 8. LIQUID WASTE STREAMS AND POLLUTANTS OF MAJOR CONCERN FROM
WELLMAN-GALUSHA LOW-Btu GASIFICATION SYSTEMS
Operation
Process
Liquid Haste Stream
Pollutants of Major Concern
Remarks
Coal Preparation
Coal Handling and
Storage
Coal Gasification
Wellman-Galusha
Gasifler
Cas Purification
Gas Quenching and
Cooling
Sulfur Removal-
Stretford
Coal pile runoff
Ash sluice water
Process eondensate
Solvent blowdown
Contain leachable organlcs and inorganics.
Inorganics of major concern Include P, Tl, V, Cu,
Fe. Ba, Cd. Cr, CM', Li and Hi. Organic concen-
trations of 65 sjg/t have been found; however, It
la not certain whether these were present In the
plant's service water used to sluice the ash
from the gasifler.
May contain organic and Inorganic pollutants
found in the quench liquor (see Table 10).
Thlosulfate and thlocyanate salts.
The composition of this stream will depend upon
the coal feedstock and site-specific conditions
(I.e. pH of leachate).
The amount of sluice water is low and highly
variable. Negative Ames tests were obtained
with low to nondetectable results Indicated for
the cytotoxlcity and rodent acute toxlclty teats.
The amount of process condensate produced will
depend upon the system operation and type of
processes used.
The amount of these salts produced will depend
upon the sulfur and cyanide content of the cooled
product gas entering the Stretford process.
-------
TABLE 9. SOLID WASTE STREAMS AND MAJOR POLLUTANTS OF CONCERN FROM
WELLMAN-GALUSHA LOW-Btu GASIFICATION SYSTEMS
Operation
Process
Solid Haste Str
Pollutants of Major Concern
,rka
Coal Gasification
tlellman-Galuaha
Caslfler
Caslfler aah
Aah laachate
(anthracite coal)
Inorganics of major concern Include Be, P. Fe.
Ca. Al. LI. Ba, Se. Pb. Ca, Cu. Ti, Cd. Sb. V.
Co, U, Mg. Sr. Si. llg. Zr. F. Kb. Aa, Mn, Cr. Hi,
Th, Al, II, Ag, T. Total extractable organics in
the ash ia low ranging f rosi 40-116 |ig/g. Organlcs
of potential concern include phthalate esters,
phenols, nltrophenola, and fused aromatic hydro-
carbons.
Inorganics of concern include P, Zn, Cd and Ag.
Results fro* the ABBS, cytotoxicity, and rodent
acute toxlclty teats for ash produced from gasi-
fying anthracite and bituminous coals were nega-
tive, low or nondetectable. Effects on soil
microcosms were also low.
Results fro* the ASMS, cytoxlcity and rodent acute
toxlclty tests of leachate frost ash produced from
gasifying anthracite coal were negative, low or
nondetectable.
Gas Purification
Particulate Removal-
Hot Cyclone
Collected partlculate
matter
Sulfur Riemoval-
Stretford
Sulfur Removal-
MU
Collected partlculate
natter leachate
(anthracite coal)
Sulfur
MKA sludge
Inorganics of major concern include Hi, Pb, P,
Mn, Fe, Cu, Ba, Sb. Tl, Cr, Ca. Al, V. LI, Mg,
Zr. Co, As, Si, Se, Be. Cd. Ag, Th, Zn, F, Ga,
Hf, Hg, Sr, Tl, Y. Low concentrations (40 to
800 I'g/g) of extractable organics have been
determined, Organlcs of concern include phthalate
eaters, phenols, nitrophenola, amines, cresols.
Inorganics of major concern include Mn, Pb, Li,
Zn, Al. Cd. Co. Cu and Fe.
May contain organics and inorganics Including
thlocyanate and 'thlosulfate salts.
Degradation products Including oxazolldon-2,
l-(2-hydroxyethyl) imldasollmdone-2. dlethanol
urea, dithlocarbamatea, thlocarbamides and other
high molecular weight nonregenerable compounds.
Negative results fro* the Ames test have been
obtained with low to nondetectable results from
cytotoxicity and rodent acute toxlclty teats.
High effects on soil microcosms were found. Col-
lected partlculates resemble devolatllized coal
with carbon contents ranging from 70 to 80Z.
Negative Ames test results were obtained and
nondetectable cytotoxicity teat results.
No data are currently available on the chemical
and biological aspects of the recovered sulfur.
Ho data are currently available on the character-
istics (chemical or biological) of MEA aludge.
-------
TABLE 10. POTENTIAL TOXIC STREAMS AND COMPOUNDS OF MAJOR CONCERN FOR
WELLMAN-GALUSHA LOW-Btu GASIFICATION SYSTEMS
Operation Potential
Process Toxic Stream Compounds at Major Concern Remarks
Caa Purification
Gas Quenching and By-product tar Organic* of major concern Include fused Tar will be produced from bituminous and lignite
§j Cooling and oils aromatic hydrocarbons, bensene, substituted coals. Positive Amea test results have been
Oi benzene hydrocarbons, beterocycllc nitrogen. obtained. Safe handling and controlling tar
sulfur and oxygen compounds, carboxyllc acids, leaks procedures are required.
aliphatic hydrocarbons, phenols and amines.
Inorganics of concern Include Cu, Pb, Sb, Cr,
Cd. Ba, Hg, V. Mg. and As.
Quench liquor Organic* of najor concern Include phenols. Results from aquatic teats Indicated a high
fused aromatic hydrocarbons, beterocycllc potential effect on aquatic species. Health
nitrogen and sulfur compounds, carboxyllc effects tests were low; however, because of the
acids, thlols, glycols, and epozldes. Inorganics chemical characteristics of the quench liquor,
of concern include HHi, cyanides, P, Se, As. F, safe handling and control of leaks are required.
Cl, Ca. Pe and Cd.
-------
TABLE 11. SUMMARY OF MOST EFFECTIVE EMISSION, EFFLUENT, SOLID WASTES,
AND TOXIC SUBSTANCES CONTROL ALTERNATIVES
Waste Stream
Most Effective Control Technology
Air Emissions
• Fugitive dust from coal storage
• Fugitive dust from coal handling
• Coal feeding system vent gas
• Ash removal system vent gas
• Start-up emissions
• Fugitive emissions and pokehole
gases from gasifier
• Fugitive emissions from hot cyclone
• Separator gas
• MEA acid gas
• Stretford oxidlzer vent gas
• Stretford evaporator vent gas
Liquid Effluents
• Water runoff
• Ash sluice water
• Process condensate
• Covered bins
• Asphalt and polymer coatings
• Enclosed equipment, collect gas
and recycle to gasifier inlet
air or treat with baghouse
• Collect gas and recycle to
gasifier inlet air or combine
with product gas
• No control necessary in a
properly designed system
• Incinerator
• Adherence to good operating
and good maintenance procedures
• Same as for gasifier
• Combine with product gas
• Recycle to gasifier
• Stretford
• Claus with tail gas cleanup
• None required with existing
applications. However, via-
bility of this approach needs
to be confirmed in a gasifica-
tion process application.
• Same as for oxidizer vent gas
• Use covered bins for coal
storage
• Contain, collect and reuse for
process needs
• Collect and recycle to ash
sluice system
• Containment and treatment at
hazardous waste facility
(Continued)
266
-------
TABLE 11 (continued)
(Continued)
Waste Strata Most Effective Control Technology
• Stratford blowdown • Containment and treatment at
hazardous waste facility
• Reductive incineration at
high temperature
Solid Wastes
• Ash • Secured landfill
• Cyclone dust • Combustion in incinerator
or coal-fired boiler
• Recovered sulfur • Purify for sale or disposal
• MEA blowdown • Containment and treatment at
hazardous waste facility
Toxic Substances
• Tars and oils • Combustion in boiler or
furnace
*Based only on effectiveness in eliminating or reducing emissions.
267
-------
Cost and Energy Considerations
Costs of the "best available" candidate con-
trol methods (identified in Table 11) are sum-
marized in Table 12. Most of the control alter-
natives have negligible costs when compared to
the costs of the product low-Btu gas. The most
costly control alternatives are those for treat-
ment of the MEA acid-gas stream and process
condensate. The most costly control methods
are also the largest energy consumers. Tars and
oils represent a large energy credit (-25 per-
cent of the product gas energy content), depend-
ing upon the coal feedstock.
One method to reduce the costs and energy
consumption of process condensate treatment is
to reduce the size of the condensate stream.
This can be accomplished by drying the coal
before gasification (the dryer offgas may con-
tain large amounts of coal volatiles). Alterna-
tively, the size of the stream can be reduced by
minimizing the amount of steam fe'd to the gas-
ifier.
Impacts on Air Quality
The potential air quality impacts of gaseous
waste streams from Wellman-Galusha low-Btu
gasification facilities were estimated and com-
pared to the following air standards and
guidelines:
• New Source Performance Standards (NSPS)
for stationary sources,
• National Emissions Standards for Hazard-
ous Air Pollutants (NESHAP),
• National Ambient Air Quality Standards
(NAAQS), and
• State and Federal emission standards.
The air quality impact of specified pollutants
(CO, H2S, COS, NH3, HCN, NOX, SOX, and non-
CH4 hydrocarbons) in gaseous waste streams
from Wellman-Galusha systems using a low- and
high-sulfur bituminous coal was estimated by an
atmospheric dispersion model. The waste
streams considered were coal feeder gases, tar/-
quench liquor separator vent gases, and Claus
tail gas incinerator gases (high-sulfur case only).
Table 13 compares maximum ground-level
concentrations of CO, nonmethane hydrocar-
bons, NOX, and SOX with the NAAQS. With the
exception of nonmethane hydrocarbons, the
predicted pollutant concentrations for both the
low- and high-sulfur coals are below the
NAAQS. Carbon monoxide concentrations
-------
TABLE 12. SUMMARY OF MAJOR COSTS AND ENERGY CONSUMPTION OF
ALTERNATIVE CONTROL METHODS
Uaatt Screaa. Media
«««c* Strata.
Cooerol Method
Control Cosca tn«rc7 C-mauapcloo
(S/GJ)« (J/J)b
Coal Handliag i
Storagt
Fugitive duae
Liquid Efglutnti
• Coal piia runoff
Covered bin* '0.01
Aaphalt and polyatr coatings -rO.Ol
Cncloaed equipnent, collection <0,01
systeM
Covered bin*
Ca.stti.er
Ca» Purification
P articulate Raw
Hoc Cyclone
and Cao Ling
Sulfur U.WT.U-
Stretford
Sulfur Rrnoval
MEA
• Coal feedlflg vent ga*«a
* Aatb renoval vent gaats
• Start-up V*BC gates.
• ru,|ifciv« *mla»l9ne
(pokaKbolt ga«e*}
Liquid Effluents
• Ajh sluice vtiter
Solid Vaatee
• Aah Ucrv-S Bitueftiaoua)
- Stratford
- «*•
- MEA (Stringent)1
• Ajh (Aacaracite)
• A*b (Li(Bite)
• Collected ?ertlculatea
Cueoua laitsioas
• OU.WCA liquor/tar
•eparaur vnnc
Liquid Effluents
• Process Coadeoaata
• sertttord
• MEA (Seriatmt)'
- lignite
• Fro«..ea Coe4eA.Mea
- Hlgh-S lltUaUnoua
• Stratford
• a**
• ME4. (jerlaiaac)1
Gaaaaua laiaaloaa
• Oxldlaar vaae (aa
• evaporator vane gaa
llama !(;iuanc>
* llawdotra aolvaat
Solid 'Jaataa
• Sullur
- Uav-9 lltuaUooua
- SlJh-I alcuaoaoua
- Aaehracita
- Llialca
• Acid iai
• U » produce faa
- 74 « product laa
• Acid caa
- 15 id produce iaa
- 74 W produce j.«
Solid aaac.a
• JXA lloodoua
• iulfur
Ciar lalcc air ar product
• Mono required
• flare or Incinerator
^aat- '0.01
iaa
—
<0.1
• MetlljiBla
—
• »AC
• Good aalnijenence end aoerattae, — —
practlcee
- Collecclaa aad ;auae
• Secured landfill
(Wit. tmelo.)
• Coatuatloa
• Coealae vtch Che produce j
. CoacaloaMac aad creacawec
uaata craanaac facllley
* Evaporation oa-elte
• Hona required
• Xoae raeutred
' Uducelve Uciaeraclon
• iacuiad laadflll
* SCreefard acid faa reaaval
• Claud tflcaauc call |aa
cleaaua
• Concalnavat and creacaanc
ac a hazardeua vaaca faclll
<0.01
0.01-0.02(0.01-0.03)
0.02-4. 06(0. 04-0.08)
0.03-0.07(0.05-0.10)
0.03-4. OKO. 04-0. 10)
0.04-0.10(0.07-0.15)
0.04-0.10(0.07-0.13)
<0.01
l*a <0.01
0.40-0.59
O.M-1.32
1.16-1.69
1.43-2.01
0.06-0.07
0. U-0. 14
0.16-0.19
o.ia-o.20
—
«
3. 002-0. J0«
0.02-0.07
0.002-0.004
O.OOJ-0.020
1.2-1.6
0.6-0.8
0.3-0.6
0.2
sy
• :u*Uilbla
• Xeili(lble
• tteill.lble
• N.|ll(lble
• Hefli|lble
• I.UHlble
• Veilliibla
• SA*
• !(e»llflble
:'^
• !Uh
0.019
0.042
0.055
0.063
—
*A
* Se|ll|xbla
. ?4e|Hclbla
0.007
0.007
0.008
Q.30J
MA"
^ — SaaM aa cfte Screcford juliyr resoval ^aae . e»
3ata not avai.Ubl.1 for caicjiatioo zt aaergy comu.mptiinu.
Coat* art unu.Uiitd eo»cs ?ar GJ of cooled, letarred product gu.
loerp fiontuejpciant *re J of energy required by tht control Mdied per J 9t cooled. J*urred produce ju.
naunpcion will depend upon the «acarials (eok«. coaJL. wood, oil, «tc.) uaad co sure up the gaaifier
tlon 9t the gM auriag the start up tisM period.
?ood lAinteKUac* ud optneing proceduraa should already be daflaad and included la tht -J3l:i aperating costs,
•ISA product* a product gas to Met prapoted XSFS (or coal coajbuatLoa (96 ng/J, 0.2 Ib/Stu).
XZf- (striagtAt) product! a "clean" produce gai caatalaiaf 6 ngVUv1 (10 ppew) of sulfur species.
and che cevpoai-
-*e* *ri aot *»aLlabl» oa eh* entrtr caoauvptioa of craatlai pracua et>od»a»ata ac aa 6tt~+ltm h*iardou* vaats i
ftCillCT-
269
-------
TABLE 13. COMPARISON OF PREDICTED POLLUTANT CONCENTRATIONS TO THE
NAAQS AND STATE OF TEXAS H2S AMBIENT AIR STANDARDS
Low-Sulfur Coal*
Predicted Concentrations (Mg/V)
Pollutant 24-h 3-h 1-h
1
CO 2,300
Non-CHi, Hydrocarbons 650
N0x 20
SOX NA
H2S 10
9.800 13.700
2.800 3.900
70 90
NA NA
50 70
High-Sulfur Coalb
Predicted Concentrations (pg/m9)
24-h 3-h 1-h
2.300
650
20
110
90
9.800
2.800
70
380
390
13.700
3.900
90 •
560
540
NAAQS (Mg/-1)
Primary Secondary
Standards Standards
10.000 (8-h)c
160 (3-h)C
1OO (aam)
365 (24-h)
State of
10,000 (8-h)c
>d 160 (3-h)c'd
100 (aam)
c 1,300 (3-h)c
Texas Regulations
122 ug/«3
NA - Not applicable, SOX emissions are trom the high-sulfur case using an Incinerator to coobust the Claus unit's tail gaaes.
aam - Annual arithmetic Bean.
"For the low-sulfur coal case, a Stretford sulfur removal process Is used.
For the high-sulfur coal, an HEA sulfur removal process is used followed by a Claus process and a Claus tall gas Incinerator.
Concentration not to be exceeded mire than once a year.
d6:00 a.m. to 9:00 a.m.
-------
TABLE 14. LIQUID EFFLUENTS FROM WELLMAN-GALUSHA LOW-Btu GASIFICATION
SYSTEMS EXCEEDING THE MOST STRINGENT EFFLUENT
STANDARDS AND MATE VALUES
Liquid Effluent
Constituents Exceeding
Most Stringent
Effluent Standards
Constituents Exceeding Health
and Ecological MATE Values in the
Multimedia Environmental Goals
Ash Sluice Water
Process Condensate8
(Bituminous Coal)
Stretford Slowdown
Fe, Cr, CN and suspended
solids
NH3, As, Cl~, CN~, B, F~,
Fe, Phenols, P, Se, S0i»=,
BOD, COD, and suspended
solids
Fe
P, Fe, Ti, Ba, La, Li, Cd, Cu, CN~, Ni and V
Phenols, Fused Aromatic Hydrocarbons,
Heterocyclic Nitrogen and Sulfur Compounds,
Carboxylic Acids, Thiols, Glycols, Epoxides,
NHi,, CN~, P, Se, As, F~, Cl~, Ca, Fe and Cd
Vanadate, Fe, EDTA and possibly Thiocyanates
and Thiosulfates
MATE: Minimum Acute Toxiclty Effluent
Process condensate produced from gasifying anthracite coal should not contain the high amounts of
organic constituents found in process condensate from gasifying bituminous or lignite coals.
-------
guidelines for the land disposal of solid wastes
(40 CFR 241). These standards set minimum
levels of performance for any solid-waste land
disposal site. The guidelines apply to the land
disposal of all solid material. Additional stand-
ards have been proposed for hazardous solid
wastes (40 CFR 250).
The solid waste streams from a Wellman-
Galusha gasification facility that will be
regulated under the RCRA are: gasifier ash,
cyclone dust, sulfur cake, and MEA blowdown.
Table 15 shows the characteristics of these solid
waste streams and how the proposed RCRA
regulations may apply. All of the solid waste
streams may be classified as hazardous wastes
under the proposed RCRA regulations.
Product/Byproduct Impacts
The product gas and byproduct tar produced
by Wellman-Galusha facilities may be regulated
by the Toxic Substances Control Act (TSCA) of
1976. However, polychlorinated biphenols
(PCBs) and chlorofluorocarbons are the only
specific substances for which regulations have
been issued.
The product low-Btu gas may contain toxic
substances even after extensive purification.
The byproduct tar does contain substances and
positive Ames test results for mutagenicity
have been obtained.
Radiation and Noise Impacts
Wellman-Galusha low-Btu gasification facili-
ties may have radiation and noise impacts.
Some radioactive species in the coal may be con-
centrated in the entrained particulate matter in
the raw product low-Btu gas and in the ash. Proc-
ess blowers and turboblowers, coal conveyors,
coal bucket elevators, and pumps are sources of
potential noise impacts in Wellman-Galusha fa-
cilities.
DATA NEEDS AND RECOMMENDATIONS
Data needs and recommendations for obtain-
ing those data are divided into the following
categories:
• Gaseous, liquid, and solid waste stream
characterizations and control;
• Process and process streams; and
• Health and environmental impact assess-
ments.
The data needs for the multimedia waste
streams and the process and process streams
associated with Wellman-Galusha gasification
systems are summarized in Tables 16 and 17,
respectively. In general, data associated with
the gasification of high-sulfur bituminous coal
are currently not available. Since existing and
planned commercial Wellman-Galusha gasifica-
tion plants are low-sulfur bituminous and an-
thracite coals, data on high-sulfur coals may
have to be obtained from bench-scale units. Data
are not available on the performance of sulfur
recovery processes and waste streams from
those processes. These data may be obtained if
a Stretford sulfur-removal process is included in
the Pike County gasification facility.
Data needs associated with performing
health and environmental assessments include
data required by existing and proposed regula-
tions, and data required to assess health and en-
vironmental (air, water, and land) impacts of
nonregulated pollutants or streams. The data
needs for existing and proposed environmental
regulations mainly involve pollutant-specific
determinations (i.e., consent decree pollutants,
solid waste leaching tests defined in 40 CFR
251), bioassay tests (i.e., proposed in the RCRA
(40 CFR 2501), and accurate pollution control
costs. Also, long-term monitoring of specified
pollutants is required to assess the effec-
tiveness of a control technique.
Data requirements for assessing the health
and environmental impacts of nonregulated
pollutants and streams will involve pollutant-
specific determinations, long-term monitoring,
and biological testing (including both acute and
chronic tests for health and ecological effects).
The specific methodologies to be used in
performing these impact assessments are still
under development. Therefore, the specific data
needs are not totally defined.
ISSUES AND AREAS OF CONCERN
BY PROGRAM OFFICES
The EPA program offices1 issues and areas of
concern for Wellman-Galusha low-Btu gasifica-
tion technology are briefly discussed here. The
basic issues and areas of concern include:
• Wellman-Galusha gasification technology
At what stage should existing standards ap-
ply to a developing technology?
When will the technology be commercial-
ized?
272
-------
TABLE 15. SOLID WASTES FROM WELLMAN-GALUSHA LOW-Btu GASIFICATION
SYSTEMS THAT WILL BE REGULATED BY THE RCRA
(20 CFR 241)
Solid Waste Stream
Characteristics of the Waste Stream
that may be Classified as Hazardous
to
Gaslf ier Ash
Cyclone Dust
Sulfur Cake
MEA Slowdown
High levels of trace elements are present and may be leached
from the ash.
High levels of trace elements are present. The dust contains
high levels of carbon (70-90Z) and may be classified as
ignitable.
The sulfur will contain various components such as vanadium
salts, thiocyanates , and thiosulf ates .
This stream will contain oxazolidin-2, l-(2-hydroxyethyl)
imidazolindone-2; diethyl urea; dithiocarbamates ; thiocarbamides ;
and other high molecular weight compounds resulting from the
formation of nonregenerable complexes.
-------
TABLE 16. SUMMARY OF WASTE STREAM CHARACTERIZATION AND CONTROL
DATA NEEDS AND PLANNED ACTIVITIES TO OBTAIN
THOSE DATA
y««t« Stream Media
Waste Streem
Additional CheroeterlEallom
Control Technology
Performance Heeded
Planted Actl*UiM to
Obtain Data Naado
Caseous emissions
Coal feeder vent gaa
Start-up veat gaa
Pokahola gaaa*
T.ir/«kli|
o*e.ratlon
Iffactlveneaa of via In t
automatic pokara
Ifrectlveiieaa of re
to tha product gaa
lone ahould bo rooelrad
Iffactlvenoaa of ualat a Claua
and tail ta* claaaup aracaoa
for a«lfur removal
Thla control "111 bo evaluated
by tadlaa and OMt at tlM
Ualwraity of Hlnaaaota (Duluth)
(UH5) roatar Whaelac/Stolc
maalUcation facility
Tha Uellmam-Oaluaha teat facility
at the U.I. tmroau of Minea at
Ft. SMlUni Htnm. has a atart-
ua TOM data that nay bo
available far taatina,
Tha IM> facility vlll uaa
•utoMtle pokara
Tha UMD facility will uaa chli
for their car atoraaa tank.
taut aaoaa will ba charactorltad
by Udtan and OHO.
Tha Pika County Uallnan-Calualia
facility nmy have a Stratford
autfwr ranoval procwa. Thla
atraao) will bo characterlaad by
Radian and OWL. other potential
teet altaa are currently being
pwrauad by Radian.
Mo NIA proeaaona are planned far
removing sulfur from) lov-gtm gaa
at atmmephortc freeeere
LUutd ttflaenta
Aah aluica vater
Prockaa condanaata
Stratford blovdovn
iolta Vaatoa
Oaoifler a
Chemical and biological char-
actarlsatioma for aflluont
gmiaallna atandnrda and con-
parlaon to the HEO'a for hlgh-
•alfur bltuminaua and llgnlta
Cbomtcal amd biological char-
actarlaaciama for affluent
guidellaee and commarlooa to
the HaV'a far high-aulfur
bltumiooua, anthracite amd
llgalta coal*
Chemical and biological char-
accarlaatiomm far efflumfit
guidellaea and commartaoa to
tha MEB'a for high- and lou-
aulfur bltuninoua, aathraclta
amd llgaita comla
Chemical and biological char-
actartiatiara for hlgh-
aulfur bltmniaoun and llgnlta
canla. LaachUg atudln ara
maadad to datarmlma if the aah
la claaalfled aa haxardaua by
Sffactlvofieea of collection
and rauaa of tha aah alwlcm
water
Effaetlvaoaaa of concentrating
proceaa eandonoaco by farced
evaporation
Iffactlvanaa* of reductiva
Incineration
Control requirements will ba
defined by tha KCXA baoad oa
chanical and biological
charactariatica
Aah aluica watar for tha gaal-
ficatlon of llgnltn at tha ft.
Snalllng facility will ba
eharmctarlaod by ladlen
Laboratory taata may ba performed
to evaluate the gaamoua enlaalona
generated by forced evaporation
Reductive Incineration* nay bo
aeod at tha Pika County facility
Laaching teata for Ucnlta aah
ara planned. Other leaching
taata for lovavlfur bltuwinaua
aan any alao am parfonad
Cyclone dual
MCA blovnown
Chemical and biological chnr-
ectarlsatlono of dust collected
tram goalfying high- and low-
eulfur bitunlnone and lignlca
coala ara needed for tha KM
Chemical and biological char-
actarlutlona of sulfur ara
needed far tha nr.RA
Chemical and biological char-
acterisation* ara needed for
the KM
Control requlrtmente vlll ba
defined by the ftCXA baamd on
chemical and biological .
characteristics
Mfactlvanaaa of conbumclng
tha dnet mey ba required
Control rn.|ui renents will ba
defined by the KM baaad on
chanical and biological
characteristics
Control requirements will be
defined by tha KM baaad on
chemical and biological
characteristics
teaching teats for llgnlta ara
planned, other leaching teats
.far low-sulfur bltmminoua coal
nay ba performed
laboratory tests any bn performed
to evaluate diiac conauatlon
characteristic*
Sulfur produced hy the Stratford
procoas will ba charecterlfed If
a Stratford proceaa la uaad at
Pirn* County or If another test
sits con ba obtained.
Ho MIA procoaao* ara currently.
planned to remove sulfur from
etnoeohorlc lov-ltm goo stroams
274
-------
TABLE 17. PROCESS AND PROCESS STREAM DATA NEEDS AND PLANNED
ACTIVITIES TO OBTAIN THOSE DATA
Proceaa
Data Need*
Planned Activities
Ucllman-Calusha Caaifler
Pnrtlculatc Removal -
Hot Cyclone
Ca« Quenching/Cooling
Tar Removal -
Electrostatic Precipitation
Sulfur Removal - Stretford
Knd Use - Combustion
Fate of pollutants (aulfur species, nitrogen apeclea, tara and olla)
for various gaalfler operating condition* and coal feedatocka.
Operating conditions that need to be evaluated Include steam/air
ratio, coal throughput, and bed depth. High-sulfur bituminous coal
haa not been tested alace all commercial facilities uae low-aulfur
bituminous and anthracite coals.
Collection efficiencies of hot cyclones are needed since the
particulatee not removed will affect downstream gaa purification
processes and the raw gas combustion process characteristics and
flue gases.
Fate and distribution of sulfur species, nitrogen spectea, tara,
olla and paniculate matter are needed. The quenched and cooled
gaa characteristics will affect the performance and design of
downstream purification processes.
Tar removal effectiveness needs to be determined since residual
tar/oil aerosols will affect the performance and design of
downstream sulfur removal processes.
Sulfur removal effectiveness needs to be determined. There are
currently no data on the performance of the Stretford process
used to remove H2S from low-Btu gaa.
Combustion gases from burning hot raw gas, quenched gas and
desulfurlzed gas are needed along with tar combustion gases.
Research Triangle Inatltute and North
Carolina State University will be performing
parametric studies on bench-scale gaalflera
using various coal feedstocks.
Partlculate removal efficiency studies for
the hot cyclone at the UHD facility are
planned.
The Pike County facility may hava a gas
quenching/cooling process. The Chapman
facility may be used to evaluate thla process.
The tar/oil removal effectiveness will be
determined at the UMO gasification facility.
Stretford process performance will be
evaluated by EPA and DOE If a Stretford unit
la used at Pike County. Other test sites
are currently being identified.
Combustion gases will be characterized at
the Ft. Snelllng and UMD facility.
-------
• Waste streams from Wellman-Galusha facil-
ities
What are the potentially harmful pollutants
in gaseous, liquid, and solid waste streams,
including potential fugitive emissions?
What are the emission rates of those pollut-
ants?
What potentially harmful pollutants in those
streams are not currently regulated?
What, are the health and ecological effects of
those pollutants and streams?
• Pollution control technology
What technologies have been demonstrated
in controlling gaseous, liquid, and solid
waste streams from Wellman-Galusha facil-
ities?
What are the economics and energy usage
associated with controlling those streams?
Each program office needs representative and
accurate data concerning:
• Chemical, physical, and biological character-
istics of the waste streams to air, water, and
land;
• Technology required to control those waste
streams; and
• Chemical, physical, and biological char-
acteristics of fugitive emissions resulting
from the processing, storage, and transport
of waste streams, products, and byproducts.
The waste stream and fugitive emission data
must be able to stand up to a traditional peer
review and court review before the data are
used for recommending standards. Control tech-
nology data should be obtained on demonstra-
tion-scale control equipment.
The following text contains a summary of the
specific issues and areas of concern by each
EPA program office.
Office of Air Quality Planning
and Standards (OAQPS)
OAQPS prepares a Standards Support and
Environmental Impact Statement (SSEIS) docu-
ment to back up its regulatory activity. These
SSEISs address specific source categories, and
each document contains the following items:
• The process or processes and associated air
emissions;
• Emission control techniques, including proc-
ess modifications and "add-on" control
equipment; ;
• Environmental impacts to air, water, and
276
land resulting from air emissions and the
control of those emissions; and
• Energy and economic impacts associated
with controlling air emissions.
Accurate, well-documented data for emission
measurements are needed to prepare an SSEIS,
along with representative data concerning con-
trol equipment performance.
All methods used to measure emissions must
be documented. Where EPA reference methods
are used, citation of the appropriate part of Ap-
pendix A, 40 CFR 60, is sufficient. Any new
methods or modifications of the standard sam-
pling and/or analysis methods must be clearly
defined and discussed. Confidence intervals on
the data obtained from modified or new meth-
ods are required along with a discussion con-
cerning the representativeness of those data
with respect to long-term emissions.
As part of the SSEIS, OAQPS needs accurate
data concerning control technology perform-
ance, costs, and energy usage. These data also
include water pollution control and solid waste
disposal when there are liquid and solid waste
streams resulting from an air pollution control
device. If possible, data on control technology
should be collected on demonstration-size units.
The following OR&D inputs are desired by
OAQPS:
• Development and evaluation of continuous
monitoring devices for selected pollutants in
air emissions,
• Identification of other potentially harmful
pollutants in air emissions,
• Assessment of a control technology's per-
formance in controlling potentially harmful
pollutants, and
• Evaluation of control technology perform-
ance, costs, and energy usage.
Results from the above inputs need must be dis-
cussed in detail and to be well documented.
Office of Water Planning
and Standards (OWPS)
OWPS has data needs and requirements sim-
ilar to those of OAQPS, except with respect to
effluent streams. Effluent measurements to
determine the presence and concentration of
the 129 priority pollutant species are needed.
These measurements should be performed by
techniques established as adequate for stand-
ards support. Other standard measurements,
-------
such as total suspended solids, biological oxygen
demand, pH, etc., are also needed along with the
identification of other potentially harmful
species in process effluents.
Accurate control technology performance
and economic and energy usage data are inputs
needed by the OWPS. If possible, these data
should be obtained from demonstration-size
processes. Sampling and analysis techniques
and control technology performance data must
be thoroughly discussed and well documented.
Office of Solid Waste (OSW)
08W has issued proposed regulations estab-
lishing the criteria for methods of testing for
and handling and disposal of hazardous wastes.
Their present needs from OR&D for Wellman-
Galusha gasification technology are minimal.
However, the application of the test methods
and identification of hazardous waste streams
from the various processes in Wellman-Galusha
gasification systems will provide necessary data
for the various cognizant enforcement and
monitoring agencies at the local, State, and
Federal levels.
Office of Toxic Substances (OTS)
OTS needs information on toxicity and expo-
sure potential of pollutants in the product and
byproduct streams associated with Wellman-
Galusha facilities to guide its regulatory efforts.
Although OTS will rely on the other program of-
fices (OAQPS, OWPS, and OSW) to regulate
waste streams and residuals, it will probably
serve in an advisory capacity to guide efforts of
these offices relative to toxic substances.
Office of Radiation Planning (ORP)
ORP may consider in FY80 the radiation
hazards posed by the operation of Wellman-
Galusha gasification facilities as well as other
synthetic fuels plants. Radon 222 in air emis-
sions from these plants, or as fugitive emissions
from coal piles and ash piles associated with
plant operation, would be one concern. Another
concern would be the Radium 226 trace impur-
ities in coal pile runoff. A key question for ORP
is: Providing that Radon 222 is found to be a
hazard in conventional combustion technology,
would synthetic fuels plants function as an ef-
fective control technology? From ORP's point of
view, OR&D environmental assessments must
include measurement of Radon 222 and Radium
226.
Gross a and 0 measurements on waste
streams are not adequate to fulfill ORP needs.
Gamma-ray spectrometry followed by U235,
Th282, and K40 elemental analysis are required.
Accurate particle size distribution data from
emission sources are also needed.
Office of Enforcement (OE)
The needs of the Office of Enforcement are
very similar to those of OAQPS and OWPS. If
OE is to advise on the issuance of permits, or in
some cases, issue permits, for Wellman-Galusha
plant operation, it must have a comprehensive
view of Wellman-Galusha low-Btu gasification
technology.
Many of the EPA program offices' general
and specific issues and areas of concern can be
addressed for Wellman-Galusha low-Btu gasifi-
cation systems. However, because of a limited
budget and a limited number of available test
sites having best available control technology,
the data collected on gaseous, liquid, and solid
waste stream characteristics (chemical, physi-
cal, and biological) and technologies to control
those streams must be prioritized. Priorities
will be based upon the program offices' R&D
needs and standards support schedule that are
defined in the Standards Support Plan for Syn-
thetic Fuels, to be published by IERL/RTP of
the Office of Research and Development.
REFERENCES
1. Kohl, A. C., and F. C. Riesenfeld. Gas
Purification (second edition). Houston, Gulf
Publishing Co., 1974.
2. Williams, Dale, and A. F. (Buzz) Zey. Per-
sonal Communication, J. F. Pritchard & Co.
August 11978.
3. Sigmund, Paul. Personnal Communication, 5
and 7. Union Carbide Corp. June 1978.
4. Perry, Charles R. Basic Design and Cost
Data on MEA Treating Units. In: Pro-
ceedings of the 1967 Gas Conditioning Con-
ference University of Oklahoma, April 4-5,
1967, Norman 1967. P. C1-C9.
5. Chem, C. L., and T. L. McCaleb. Coal Proc-
essing: Low Btu Gas as an Industrial Fuel.
277
-------
Chem Eng Progr. 73(6):82-88. (1977).
6. Woodruff, David D. Personal Communica-
tion. McDowell-Wellman Engineering Go.
December 19 1977.
7. Cleland, J. G., and G. L. Kingsbury.
Multimedia Environmental Goals for En-
vironmental Assessment, Volumes I and II.
Research Triangle Institute. Research
Triangle Park, NC. Report No. EPA-600-7-77-
136a, b, EPA Contract No. 68-02-2612. No-
vember 1977.
278
-------
FATE OF PHENOLS DURING THE GASIFICATION OF COAL
John P. Fillo*
Carnegie-Mellon University, Pittsburgh, Pennsylvania
and
Michael J. Massey
Environmental Research & Technology, Inc., Pittsburgh, Pennsylvania
Abstract
An investigation of the engineering relation-
ships governing the production of phenols dur-
ing coal gasification processing is described.
Experiments were conducted on bench-, PDU-,
and pilot plant-scale facilities to assess the im-
pact of initial formation and subsequent decom-
position phenomena on observed levels of phe-
nolic compound production. Experimental ef-
forts included:
• Bench-scale investigation of the decomposi-
tion characteristics of select phenolic com-
pounds in the homogeneous gas phase and
over fixed beds of lignite char and limestone
acceptor solids,
• PDU-scale experimentation on the Pitts-
burgh Energy Technology Center's Syn-
thane PDU gasifier to assess the effects of
changing initial devolatilization conditions
on the formation of phenols from coal, and
• Pilot-scale investigation of coupled forma-
tion/decomposition phenomena via probe
sampling of the spatial chemical composi-
tion within the CO2-acceptor pilot plant gas-
ifier in the vicinity of the fresh coal feed loca-
tion.
Integrating the results of these three separate
experimental studies facilitates an understand-
ing of phenolic compound behavior during coal
gasification. Major behavioral characteristics
identified indicate that:
• Phenols are formed inherently during the
devolatilization stage of coal processing;
• Production of phenols, which, are highly
susceptible to thermal and catalytic decom-
position, is controlled by physical and opera-
tional characteristics of the gasification
process that could enhance thermal and cat-
alytic cracking, and
•Speaker.
• Conditions responsible for enhancing reduc-
tion of.phenolic compound production do not
adversely affect production of the primary
product of coal gasification.
INTRODUCTION
Phenolic compounds comprise a family of aro-
matic hydrocarbons produced during coal gasi-
fication. They report to aqueous and hydro-
carbon condensates (when produced) and are
removed from raw gasifier product gas during
quenching operations. Production of phenolic ef-
fluents during gasification processing is highly
variable and is a strong function of both
gasification process conditions and quench sys-
tem operation. This apparent variability, both
between and within individual processes, pro-
vides the incentive to investigate the relation-
ships that govern production of phenols during
coal gasification.
Given the current status of coal gasification
research in the United States, it is clear that the
environmental acceptability of a process must
be determined based on data from subcommer-
cial facilities. As a result, a substantive engi-
neering basis is needed to properly obtain and
interpret environmental data taken at PDU and
pilot-plant scales of development. In view of the
fact that significant changes in operating condi-
tions can occur in scaling to commercial-size
facilities, data must be obtained that permit
adequate decoupling of process variable interac-
tions so quantitative projections of phenolic
compound production can be made. The ac-
curacy of these projections affects the design
and operation of all modes of processing down-
stream of the primary gasification system.
An experimental strategy is developed that
effectively uses three different experimental
scales of process development. Production is
qualitatively segregated into initial formation
279
-------
and subsequent secondary reaction steps within
the gasification environment. The choice of
experimental efforts reflects the desire to study
these phenomena individually and in a highly
coupled manner. The primary data base needed
to delineate the pathway governing production
of phenols in coal gasification processing is gen-
erated from these experiments.
PHENOLIC COMPOUND PRODUCTION
DURING COAL GASIFICATION
Phenols are of primary interest in coal gasi-
fication processing. Although produced in small
quantities relative to the major product of gasi-
fication, their presence is important in spec-
ifying particular modes and configurations of
downstream raw product gas processing. When
produced in sufficient quantities, they can
represent a valuable byproduct of gasification.
If byproduct recovery is uneconomical or pro-
duction is limited, they represent an effluent
that must be eventually processed. The amount,
type, and physical characteristics of phenols
produced during coal gasification determine*
the physical and chemical nature of the process-
ing scheme required.
Basic Production and Processing
Patterns of Phenols
The general pattern of phenolic compound
production and processing is illustrated in Fig-
ure 1. Phenols produced during gasification exit
the gasifier with the raw product gas and are re-
moved during quenching operations. They re-
port to both aqueous and hydrocarbon conden-
sates, when produced, because of their partial
solubility in water and the aromatic nature of
the compounds. Relative quantities of con-
densates produced are a strong function of gas-
ification process conditions.
Two basic options are available for process-
ing phenolic condensates: byproduct recovery of
crude phenols, and destructive treatment of
phenols. Crude phenols can be recovered from
either aqueous or hydrocarbon condensates.
Solvent extraction and distillation are two
frequently used recovery techniques. In the
event that byproduct recovery is not feasible,
phenols in both aqueous and hydrocarbon con-
densates can be destructively treated. Aqueous
condensates containing high levels of phenolic
material can be processed via biological oxida-
tion in the presence of bacterial organisms.
Reduction of phenols in hydrocarbon conden-
sates can be accomplished by injection of tar
back into the gasifier, which is frequently done
for commercial fixed-bed gasification tech-
nology.
Variable Phenolic Compound
Production Rates
Data summarized in Table 1 indicate substan-
tial variation in measured phenolic compound
production rates, both within and across proc-
essing concepts. Given the widely different
processes represented, this is not surprising.
Large changes in macroscopic operating condi-
tions and physical gasifier geometries must af-
fect the production of phenols. In addition, vari-
abilities within processes should be expected if
significant differences in operating conditions
can be effected.
Phenol Production Variation
Across Processes—
Processes listed in Table 1 are significantly
different, both physically and operationally. Dif-
ferences exist in operating temperatures and
pressure, contacting geometry, and coal type.
These differences are summarized in Table 2
based upon characteristics of operating pilot-
plant facilities.
The type of gas-solid contacting varies consid-
erably, implying substantial differences in
modes of mixing. Pressure and temperature
variations are considerable, with the latter seg-
regated to specify the initial thermal conditions
the coal meets upon entering the gasification
environment. Coal type is indicated more from
an operational viewpoint because neither C02-
Acceptor nor the slagging fixed bed facilities
can operate on bituminous coals.
The most obvious difference in production of
phenols occurs for the CO2-Acceptor process
(see Table 1), where production is fully two to
three orders of magnitude less than for any
other process. Differences in processing condi-
tions are also evident. Most notably, pressure
and initial coal devolatilization temperature are
consistently higher and lower, respectively, for
all other processes. Further cross-process
comparisons are difficult because of con-
siderable observed variability within processes.
280
-------
Coal
*-
G
A
S
I
F
I
E
R
Q
u
E
N
C
H
•
1 ^*
Phenolic
Condensates
«-
Raw Product Gas
Option 1: By-Product
Recovery
Option 2: Destructive
Treatment
Figure 1. Basic production and processing patterns of phenols during coal gasification.
-------
Phenol Production Variation
Within Processes-
Differences in phenol production rates within
individual processes are significant and suggest
that changes in processing conditions can also
significantly affect production. The complexity
of pilot-plant systems, operationally and
physically, precludes straightforward analysis
of this behavior. However, results of experimen-
tation conducted on the Synthane Pilot Devel-
opment Unit (PDU) dramatically illustrate that
production of phenols is substantially reduced
simply by feeding coal deeper into the gasifier.8
Although production of phenols and hydro-
carbon tars decreased substantially as coal was
injected deeper into the gasifier (see Table 3),
the extent of these observed reductions dif-
fered. Alteration of coal feed geometry signifi-
cantly changed process variables such as tem-
perature, vapor residence time, gas-solid con-
tacting, and coal heating rates. Changing from
free-fall (i.e., gravity feed through'solids disen-
gaging zone) to shallow bed-injection (i.e., feed
directly into the fluidized bed) of coal resulted in
major increases in all variables except resi-
dence time. Residence time was the only vari-
able to increase substantially as coal was fed
deeper into the fluidized bed. Coupling of hydro-
carbon production reductions and process vari-
able changes suggested:
• Phenol production was reduced by increas-
ing temperature and residence time, consist-
ent with thermal cracking mechanisms, and
• Hydrocarbon tar production was reduced
primarily by changing initial coal devolatiliz-
ation conditions (i.e., coal heating rates, gas-
solid contacting).
STRATEGIC EVALUATION OF PHENOLIC
COMPOUND PRODUCTION PATTERNS
As a result of the demonstrated variability in
production of phenols, determination of basic
production patterns for phenols during coal gas-
ification was considered desirable. Production
of phenols, as well as of any other effluent, is a
manifestation of two distinct phenomena: initial
formation from coal followed by subsequent sec-
ondary reactions within the gasification envi-
ronment. Proper delineation of these character-
istics required minimizing the inherent com-
plexity of the individual reacting systems. Ex-
TABLE 1. SUMMARY OF RANGES OF PHENOLIC COMPOUND
PRODUCTION FROM COAL GASIFICATION PROCESSES
Process
Phenol Production
Ib/ton MAP coal
(a)
Bigas
C02-Acceptor
Hygas(b)
Slagging Fixed Bed
Synthane PDU
Footnotes:
(c)
< 0.01
1-16
10-30
1-12
a) Data from Reference 1, except where noted.
b) Includes data from Reference 2.
c) Data from References 3-7.
282
-------
Process
TABLE 2. SUMMARY OF COAL GASIFICATION PILOT-PLANT OPERATING CONDITIONS
Temperature(fa), ° (j
BIgas
CC^-Acceptor
Hygas
Slagging Fixed
Bed
Synthane PDU
Subbituminous
NR
(d)
Lignite and
Subbituminous
NR<<0
Coal Tyr>e(a)
NR
Lignite and
Contacting
Geometry
Entrained Flow
Fluid ized Bed
Pressure,
atm
69
11
Devolatilization ^c^
925-1200
815
Gasification
1650
815
Staged 69
Fluidized Beds16'
Fixed Bed < 30
Fluidized Bed <40
425-650
175
400 - 700
870
1650
870
Footnotes;
(a) Coal types include lignite, Subbituminous and bituminous, with NR referring to no restrictions.
(b) Temperatures represent averages, accurate to within at least +10 percent. Wide temperature variations
are noted.
(c) Initial temperature condition to which coal is subjected.
(d) Use of bituminous coal requires an oxidative thermal pretreatment step.
(e) Coal is devolatilized in an upflow entrained-flow riser tube.
-------
TABLE 3. COMPARATIVE STEADY-STATE PRODUCTION RATES FOR PHENOLS AND TAR: FREE-FALL,
SHALLOW, AND DEEP-BED INJECTION OF NORTH DAKOTA LIGNITE^
Coal Injection
Geometry
Phenols
Production,
Ib/ton MAF coal
Percent
Reduction
Cb)
Tar
Production,
Ib/ton MAF coal
Percent
Reduction
Free-Fail
11.9 + 1.3
74.1 + 27-
71
86
Shallow Bed
3.5 + 1.9
10.1 + 5
86
38
Deep Bed
0.5 + 0.6
6.3 + 2.2
Footnotes;
(a) Source: Reference 8.
(b) Reduction achieved by injecting coal deeper into the gasifier.
-------
periments were strategically designed to segre-
gate these phenomena. The inherent advan-
tages of various experimental scales, as il-
lustrated in Figure 2, were used to define an ex-
perimental program consisting of:
• Bench-scale investigation of thermal
decomposition characteristics,
• PDU-scale investigation of formation char-
acteristics, and
• Pilot plant-scale investigation of coupled
formation/decomposition characteristics.
Integrating the results of these investigations
provided the basis necessary to delineate the
patterns of phenolic compound production dur-
ing coal gasification.
Decomposition Characteristics of Phenols
The experimental strategy was designed to
investigate patterns of phenolic compound de-
composition:
• Homogeneous gas phase decomposition of
phenol,
• Homogeneous gas phase decomposition of
ortho-cresol, and
• Heterogeneous decomposition of phenol
over fixed beds of gasifier solids.
Homogeneous gas phase experimentation
keyed on the effects of various combinations of
temperature, residence time, and hydrogen par-
tial pressure. The reactivity of two distinctly
different phenolic compounds was needed to
characterize phenol reactivity relative to cre-
sols and xylenols previously reported.'10 Decom-
position of phenol in the presence of fixed beds
of gasifier solids was necessary to assess poten-
tial catalytic effects in gasification systems. The
bench-scale equipment used to study these
phenomena is illustrated in Figure 3.
Homogeneous Phenol Decomposition—
To facilitate initial work, experiments were
conducted using pure component phenol at at-
mospheric pressure. Although a variety of
phenolic compounds are normally found in coal
gasification aqueous condensate, phenol is
typically the largest single constituent (i.e., 40
to 60 percent of total phenols).11 uu Process con-
ditions were varied to study the effects of
temperature, residence time, and reaction gas
LIGNITE
HOMOGENEOUS
PHENOL
DECOMPOSITION
GAS-SOLID CONTACTING
HEATING/DIFFUSION RATES
*- FORMATION PATTERN
•INHERENT/INHIBITED
• PREDOMINANT
COMPOUND TYPES
*-DECOMPOSITION PATTERN
•TEMP/RESIDENCE
TIME DEPENDENCE
•EFFECTof HYDROGEN
•EFFECTof CATALYTIC
SOLIDS
• RELATIVE REACTIVITYof
HOMOLOGUES
ACCEPTOR
Figure 2. Strategy for mukiscale experimental investigation of phenolic compound
production in coal gasification processing.
285
-------
Distilled
Phenolic
I
•DO—
storage Bellows
pump
I
Inert gas
storage
Reactont
gas
storage
Vaporizer!
0-II5V0
AC
-CD-
\
RReactor
Pumnm -J
Furnace
Steady state
condensate collectJon
system
Rotometer
Constant
differential
flow fie
controller ^
Startup /shutdown
condensate collection
system
i >
A
Inlet gas
sampling
_J
Gas to
vent
Wettest
meter
Valves
Kovor fitting
Ground-gloss joint
Heat tracing
Tygon tubing
Figure 3. Thennal (tecomposftion reactor—basic bench-scale ecjuipment train.
-------
composition on rates of phenol decomposition
and the nature of the decomposition products.
The range of process conditions in these ex-
periments included:
• Reactor temperatures from 300° to 975° C,
with primary emphasis on the range 760° to
900° C;
• Reaction gas residence times of 2 to 4 s; and
• Hydrogen partial pressures of 0.0, 0.2 and
0.5 atm, each at a constant water partial
pressure of approximately 0.5 atm.
Such homogeneous thermal decomposition rate
measurements at atmospheric pressure pro-
vided a distinctly conservative estimate of
potential phenolic compound decomposition
rates in the presence of char at system pres-
sure. Experimentation under well-controlled
conditions also provided quantitative decom-
position kinetics for the most prevalent single
compound in coal gasification aqueous conden-
sates.
Homogeneous Ortho-Cresol
Decomposition—
Ortho-cresol was chosen as a second phenok'c
compound in these studies because cresols are
the largest class of phenolic compounds in coal
gasification wastewaters, with ortho-cresol the
most reactive of the cresols.'10" The range of
experimental conditions included:
• Reactor temperatures from 600° to 900° C,
and
• Reaction gas residence times of 2 to 4 s.
Hydrogen and water partial pressures were
maintained at approximately 0.2 and 0.5 atm, re-
spectively. These studies, in conjunction with
studies of phenol, would define a "decomposi-
tion envelope" for approximately 80 + percent
of the phenolic compounds typically found in
coal gasification aqueous condensates.
Heterogeneous Phenol Decomposition—
Char solids occupy significant portions of coal
gasifiers. In addition, the C02-Acceptor process
required the use of a lime-bearing acceptor. Ex-
periments were conducted to evaluate sepa-
rately the decomposition of phenol in the pres-
ence of fixed beds of North Dakota lignite char
from the Synthane PDU and lime-bearing ac-
ceptor from the C02-Acceptor gasifier. Similar
conditions of residence time, water, and hydro-
gen partial pressures were used at reactor tem-
peratures ranging from 350° to 750° C. Such ex-
periments permitted assessment of the relative
magnitudes of reaction rates in the presence of
potentially catalytic solid surfaces typical in
coal gasification processes.
Formation Characteristics of Phenols—
A series of six gasification trials were con-
ducted on the Synthane PDU (see Figure 4) to
examine the sensitivity of phenolic effluent
production and composition to critical changes
in devolatilization process parameters. Varia-
tions in gas-solid contacting and heating/diffu-
sion rates were effected by altering coal injec-
tion geometry and mean coal particle size, re-
spectively. The relative effects of thermal/cata-
lytic decomposition, identified during bench-
scale experimentation, were minimized by
injecting fresh coal on top of the fluidized bed.
This mode of coal injection provided devolatil-
ization conditions similar to previous shallow
and deep bed-injection trials,8 while essentially
eliminating residence time of devolatilized
species in the hot fluidized bed.
Gas-Solid Contacting—
Gas-solid contacting in the Synthane PDU
was varied by utilizing both free-fall and top
bed-injection of North Dakota lignite coal. Free-
fall injection of coal permitted devolatilization
to occur in a relatively dilute, unmixed environ-
ment. Top bed-injection of coal onto the
gasifier's fluidized bed produced the intense
gas-solid mixing thought to enhance secondary
reactions of devolatilized species with hot char
surfaces. Because residence time of devolatil-
ized species in the hot fluidized bed was mini-
mized, the impact of devolatilization conditions
was effectively isolated.
Coal Particle Size-
Variation of coal particle size influenced rates
of coal heating and diffusion of devolatilized
species from the coal particles. Transient trans-
port by either of these mechanisms (generally
described in terms of the Fourier number) con-
tains the same functional dependencies (i.e.,
proportional to diameter squared). Particle sizes
used in this study produced initial heating and
diffusion rates that differed by more than an
order of magnitude. Initial heating rates (i.e.,
assuming an isothermal coal particle) were
4,000° and 84,000° C/s for 220 (i.e.,~20 x KKf
mesh) and 50 (i.e., 70 percent through 200 mesh)
287
-------
Cool feed
Carbonization zone
"6ft high,IOinches i.d.
Fluidized bed
P6ft high,4 inches i.d.
Char
removal system
~900°C
Product gas and
condensible effluent
Free fall coal injection
point ~5ft above the
fluidized bed
3/4"O.D. dip tube
Top bed-injection on
surface of the fluidized
bed
~ 700 °C
Steam /oxygen
Char
Figure 4. Basic configuration and coal feed locations of the
Synthane POU gasifiar.
288
-------
micron coal particles, respectively. Overall
heating rates (i.e., coal particle reaches 95 per-
cent of reactor temperature) were 200° and
3,000° C/s, respectively. The range and order of
magnitude changes in these rate processes
were sufficient to identify the combined effects
of heating and diffusion rates on phenolic com-
pound formation during coal devolatilization.
Operation of the Synthane PDU during gasifi-
cation of such widely varying coal particle sizes
required maintenance of either constant carbon
conversion or fluidizing conditions (i.e., gas-solid
contacting). Because coal devolatilization is a
rapidly occurring phenomenon, it should not be
affected by levels of carbon conversion typical
in the Synthane PDU (i.e., 50 to 95 percent). As a
result, "constant" fluidization conditions were
maintained throughout these gasification trials
as defined by Damon.15 Operating velocities
necessary to achieve these fluidization condi-
tions were 10.4 and 3.7 cm/s, for mean particle
sizes of 220 and 50 microns, respectively. Selec-
tion of fluidization velocities in these trials was
based solely on the operating constraints of the
Synthane PDU gasifier.
Coupled Formation/Decomposition
Characteristics of Phenols
Negligible quantities of hydrocarbon tars,
oils, and phenols were produced from the C02-
Acceptor pilot-plant gasifier. Coal was injected
at the base of the gasifier's fluidized bed,
operating at 11.5 atm and 815° C. Volatile
species released from the coal had to travel
through over 8 m of a fluidized char bed to exit
with the product gas. In view of the complex-
ities governing production of phenols during
coal gasification, formation and decomposition
were impossible to decouple by measuring ex-
ternal production characteristics.
This experiment was designed to investigate
phenol production characteristics on a large-
scale gasification system. The inherent coupling
of formation and decomposition phenomena
made it an attractive site for strategically
designed experiments where both phenomena
could be observed. This was accomplished by
sampling the spatial chemical composition of the
gasifier in the vicinity of fresh coal feed. Sampl-
ing probes able to withstand the severe gasifier
conditions were designed to quantitatively
remove process gas from within the three-phase
gasifier environment. Complete characteriza-
tion of process gas samples provided data on
both phenols and hydrocarbon tars/oils, as well
as for inorganic effluents (e.g., hydrogen
cyanide and ammonia), and C02-Acceptor
gasifier process dynamics (i.e., steam-carbon
gasification kinetics and fluid mechanical
behavior).1' A schematic of the base of the COg-
Acceptor gasifier and the three longitudinal
probe locations is shown in Figure 5.
Formation of Phenols—
Effective segregation of phenolic compound
formation within the COg-acceptor gasifier re-
quired sampling in the immediate vicinity of the
location of fresh coal feed. As a result, a primary
location for a sampling probe necessarily had to
be opposite this point in the gasifier. Then,
sampling could occur progressively closer to the
fresh coal feed location through an approach
from the opposite side of the gasifier. The
sampling point closest to fresh coal feed was ap-
proximately 25 cm above, offset by 16° (see
Figure 5).
Decomposition of Phenols-
Sampling probes were located at various levels
around the coal feed location to track the fate of
phenols within the gasifier following their
release during coal devolatilization. Since gas
and solid mixing patterns within the gasifier
were not known a priori, two additional probes
were located approximately 59 cm above and 36
cm below the coal feed location. Probe entry in-
to the gasifier was offset by 110° and 225° for
top and bottom probes, respectively. Combined
with the capability to perform a radial traverse,
the environment within the gasifier could be ef-
fectively sampled.
PATHWAYS TO PRODUCTION OF
PHENOLS IN COAL GASIFICATION
PROCESSING
Results of these investigations cover essen-
tially the full range of parameters for each in-
dividual effort. Bench-scale phenolic compound
decomposition studies were previously pre-
sented for the initial phenol work," and for later
ortho-cresol and solids experiments.1'lg " Re-
sults of experimentation on the Synthane PDU
were reported for characterization of all ef-
fluent and product species.11( Process gas and
289
-------
Top Probe.
Middle
Probe
Bottom
Probe
Fresh Coal
Feed
Figure 5. Radial and longitudinal probe orientations in the
C02-acceptor gasifier bed.
290
-------
environmental characteristics within the C02-
acceptor gasifier have also been analyzed and
reported.16 The following discussion evaluates
experimental results that help delineate the for-
mation and decomposition patterns of phenols
during coal gasification.
Delineation of Formation
Characteristics of Phenols
Bulk Formation of Phenols—
Results of experimental investigations con-
ducted on the Synthane PDU point distinctly to
inherent formation of phenols during coal
devolatilization. Data summarized in Table 4 in-
dicate essentially invariant production of total
phenols (i.e., 8 ± 3,7 ± 4 and 6 ± 1 Ib/ton MAF
coal) over the full range of varying coal
devolatilization conditions. In contrast, note
that changes in devolatilization conditions
result in significantly reduced production of
hydrocarbon tars and oils. In the case of the
C02-Acceptor gasifier, probe sampling at the
point nearest fresh coal feed (i.e., ~ 25 cm above
coal feed) identified levels of phenols of 1 to 2
Ib/ton MAF coal. This is in contrast to levels of
phenols fully three orders of magnitude lower
(see Table 1) as measured in the raw product
gas. The release of phenols from coal, based
upon observed characteristics in the Synthane
PDU, must necessarily occur through the lower
end of the coal's thermal processing (i.e., less
than 650° C).
Phenolic Compound Homologues Formed
During Coal Gasification—
The types of phenolic compounds present in
aqueous condensates cannot be determined
simply by characterization of total phenols,
measured colorimetrically. Selected condensate
samples from both the C02-Acceptor and Syn-
thane PDU gasifiers were analyzed via direct
aqueous injection gas chromatography to assess
the types of compounds produced. Results were
significant from an analytical standpoint as well
as for delineating phenolic compound formation
characteristics.
Comparison of Phenolic Compound Produc-
tion Levels from Total Colorimetric and GC
Analyses—Data from experimentation on the
Synthane PDU suggest a negative bias in deter-
mining phenols by the colorimetric technique.
Results, summarized below, include data from
runs CHPFL-284 and 285 (free-fall coal injection)
and CHPFL-287 and 288 (top bed coal injection,
220 and 50 micron particle diameters, respec-
tively);
Coal feed
geometry
Free-fall
Top bed
Ratio
colorimetric/GC
0.61
0.74
± 0.02 (4)
± 0.02 (4)
On the average, only 61 to 74 percent of total
phenolic material measured gas chromatogra-
phically is detected in the colorimetric deter-
mination. This behavior is significant because
the colorimetric technique, an accepted stand-
ard method for determination of phenols in
gasification wastewaters, does not detect be-
tween 26 and 39 percent of the aqueous phenols
present in these aqueous condensates.
Primary Phenol Homologues Formed During
Coal Gasification—The only phenolic com-
pounds detected in these experimental investi-
gations were single aromatic ring phenols (i.e.,
phenol, cresols, and xylenols) present in Syn-
thane PDU aqueous condensate. Phenol and cre-
sols were the only phenols detected in conden-
sates from probe sampling in the C02-Acceptor
gasifier. Unfortunately, formation was not de-
coupled entirely in the COg-Acceptor probe
studies, and hydrocarbon condensates produced
in the Synthane PDU were not analyzed for phe-
nols. However, published data for the Synthane
PDU and the Grand Fork's Slagging Fixed Bed
gasifiers, where both aqueous and hydrocaron
condensates were analyzed for phenols, were
available and are summarized below:
Phenolic compound
production16
Run
Synthane PDU1'
CHPFL-111
CHPFL-118
Ib/ton MAF
coal Single
Single Multi- rings,
ring ring percent
13
13
5
3
71
83
291
-------
TABLE 4. SUMMARY OF STEADY-STATE CONDENSIBLE HYDROCARBON PRODUCTION LEVELS FOR
GASIFICATION OF NORTH DAKOTA LIGNITE IN THE SYNTHANE PDU
Trial No.
Coal Feed
Geometry
Mean Particle
Size, microns
Condensible Hydrocarbon Production, Ib/ton MAP coal
TarOilsPhenols
CHPFL-284
CHPFL-285
Free-Fall
50
13 +_ 3
(6)
(b)
54 ^ 11(C) 8^3
(2)
to
CHPFL-288
CHPFL-289
Top Bed
50
0.4 +_ G.
(6)
49 + 38
CHPFL-286
CHPFL-287
Top Bed
220
0.3 +_ 0.2
(4)
Footnotes:
a) Minimum total phenols as phenol, determined colorimetrically.
b) Values in parentheses are number of data points averaged.
c) Data for CHPFL-285 only.
d) Includes single data points for either CHPFL-286 or 289.
-------
Run
GFETC SFB
RA-2167
RA-314
RA-40*
Phenolic compound
production18
Ib/ton MAP
coal Single
Single Multi- rings,
ring ring percent
27
29
19
4
4
4
88
82
These data show that single aromatic ring
phenols are by far the most predominant
phenolic compound type formed from coal dur-
ing gasification processing. Single aromatic ring
phenols comprise 71 to 83 percent of Synthane
PDU phenolic condensates. The Grand Fork's
Slagging Fixed Bed gasifier produces from 82 to
89 percent single aromatic ring material for the
data shown.
Correspondence of the Pattern of Phenolic
Compound Formation with State-of-the-Art
Coal Chemistry -
Ironically, one must address decomposition of
coal itself to explain the formation of phenols
from coal. In effect, candidate reactions respon-
sible for phenolic compound formation from a
hypothetical chemical structure of the coal
"molecule" are proposed. As a result, the pres-
ent understanding of specific chemical group-
ings and their orientation in the coal "molecule"
makes this prognostication speculative at best.
Based on the behavior identified, certain con-
clusions can be drawn in addressing formation
of phenols from coal.
• Phenolic compound production from
gasification systems that minimize thermal
decomposition (i.e., Lurgi, Slagging Fixed
Bed, Synthane) varies by less than a factor
of 2 to 3, regardless of coal rank;
• Coal oxygen content varies by as much as an
order of magnitude between lignite and bitu-
minous (on a moisture- and ash-free basis);
and
• Phenolic effluents from coal gasification
processes typically contain less than 2 per-
cent of the coal's original oxygen (on a
moisture-and ash-free basis).
What should be noted is the apparent in-
variability in phenolic compound production
compared to the significant variability in coal
oxygen content. Drastic swings in coal oxygen
content do not result in similar variability in
phenolic compound production. The small
amounts of coal oxygen that report as phenolic
effluents only hamper accurately tracking the
fate of coal oxygen during formation of phenols.
Delineation of Decomposition
Characteristics of Phenols
The major conclusion from these experimen-
tal investigations is that production of phenols
during coal gasification is controlled primarily
by decomposition processes. Phenolic com-
pounds are susceptible both to thermal and
catalytic decomposition, although not to the
same extent. The unique atmosphere present in
coal gasification processing (i.e., high relative
H2 and H20 partial pressures) precisely deter-
mines the decomposition pathway for the
formed phenolic compounds. Experimentation
across bench-, PDU-, and pilot plant-scale proc-
esses amply supports this behavior.
Relative Phenolic Compound React!vity-
A significant data base is reported that
defines the decomposition of phenolic com-
pounds, specifically the methyl-phenols. De-
alkylation rates of phenols increase with in-
creasing molecular weight (i.e., additional alkyl
substitution).14 Xylenols as a compound class are
more reactive than cresols.9 Of the three
isomeric cresols, meta-cresol is the least reac-
tive and ortho-cresol is the most reactive.10"
Homogeneous Gas Phase Decomposition of
Phenol and Ortho-Cresol—Experimental
results confirm that substantially quantitative
decomposition of phenol and ortho-cresol can be
achieved by homogeneous gas phase reaction
above 900° and 825° C, respectively. Data on
the relationship of decomposition to reaction
temperature and vapor residence time for
phenol and ortho-cresol are shown in Figures 6
and 7, respectively. As anticipated, experimen-
tal results show that decomposition depends
strongly on reaction temperature and vapor
residence time. Greatest increases in phenol
decomposition occur at temperatures between
750° and 900° C. In contrast, similar increases
occur for ortho-cresol decomposition between
less than 600° and 825° C, suggesting increased
reactivity of ortho-cresol relative to phenol.
293
-------
Observed decbmposition rates are independent
of hydrogen partial pressure for phenol. The ef-
fects of changing hydrogen partial pressure on
ortho-cresol decomposition were not inves-
tigated.
Heterogeneous Decomposition of Phenol—
Experimental results indicate significantly
different behavior for lignite char and lime-
bearing acceptor solids (see Figure 8). Substan-
tially complete decomposition of phenol occurs
for reaction over fixed beds of lignite char from
the Synthane PDU at temperatures as low as
600° C in less than 2 s. Greatest increases in
phenol decomposition occur between 400° and
600° C. In contrast, decomposition of phenol
over fixed beds of lime-bearing acceptor is only
slightly higher than that observed during homo-
geneous gas phase reaction at the same temper-
ature. Only 11 percent decomposition occurs at
a temperature where complete decomposition
occurs in the presence of lignite char. It is ex-
pected that the high surface area of the lignite
char (i.e., - 360 m2/gm), as opposed to that for
the lime-bearing acceptor (i.e., -1 m2/gm),
is responsible for providing the potential for a
catalytically enhanced reaction.
Relative Reactivity of Phenol and Ortho-
Cresol—Based on the demonstrated inde-
pendence of phenol decomposition on hydrogen
partial pressure, first-order kinetics of phenol
decomposition were developed similar to that
for decomposition of unsubstituted aromatic
hydrocarbons.20 The rate-controlling step in the
reaction sequence was the initial thermal de-
composition of the aromatic ring. First-order
rate constants for both homogeneous and heter-
ogeneous decomposition were plotted individu-
ally as a function of reciprocal absolute temper-
ature. Arrhenius parameters were calculated
by a least-squares linear fit of the first-order
rate constants. Ortho-cresol data were similarly
treated, as the reaction appears first order for
large relative molar ratios of hydrogen to ortho-
cresol (i.e., fully 300:1 in these experiments).10
Analysis results are summarized in Table 5.
Arrhenius parameters summarized in Table 5
are used to determine reactivities of ortho-cre-
sol relative to phenol during homogeneous de-
composition and those for phenol, heterogene-
ous relative to homogeneous decomposition. As
shown in Table 6, ortho-cresol is 4 to 15 times
more reactive than phenol under typical gasifi-
cation temperatures. Considering that ortho-
cresol is no more than twice as reactive as meta-
cresol (i.e., at - 700° C),10 phenol is the least
reactive of the phenols. Most notable is the ap-
proximate three order of magnitude rate en-
hancement for decomposition of phenol over
fixed beds of lignite char.
Effect of Reaction Atmosphere on the
Decomposition Pathway of Phenols-
Two distinct characteristics of phenolic com-
pound reaction products were demonstrated
throughout the course of bench-scale experi-
mentation:
• No dehydroxylation products (i.e., benzene
or toluene) were ever detected in more than
trace quantities during either phenol or or-
tho-cresol decomposition experiments, and
• Substantial quantities of heavy hydrocarbon
tars were formed only in the absence of hy-
drogen during these experiments.
The first result was not surprising considering
the excessive amounts of water present in the
reacting system (i.e., fully 1000:1 on a molar
basis) relative to the phenols. Excessive quan-
tities of water essentially act to stabilize the
phenolic hydroxyl to dehydroxylation reac-
tions.21 a The presence of hydrogen in the reac-
ting atmosphere acts to prevent repolymeriza-
tion of free radicals formed during the decom-
position reactions. In this particular instance,
amounts of hydrogen relative to phenols were
fully 300:1 on a molar basis.
Results of this experimental program, com-
bined with previously demonstrated decomposi-
tion characteristics of higher phenols (i.e., cre-
sols and xylenols), define a precise reaction
pathway for decomposition of phenols in coal
gasification:
• Methyl-phenols formed from coal undergo
successive dealkylation to the next lowest
phenolic compound until phenol is produced.
Phenol decomposes via pathways similar to
those for unsubstituted aromatic hydrocar-
bons.20
• Minimal amounts of heavy hydrocarbon tars
are formed.
The unique atmosphere present in coal gasifica-
tion processes is primarily responsible for
determining phenolic compound decomposition
characteristics. Quantities of hydrogen and
steam present in reaction gases are fully hun-
294
-------
100
'5
2 second residence time
m 40 -
o
a.
O
O
u
o
20 -
0
600
700 800 900
AVERAGE REACTOR TEMPERATURE,°G
Figure 6. Measured phenol decomposition as a function of average temperature for
2, 3, and 4 s nominal residence times.
1000
-------
100
80 —
§
u
n
Q)
Cu
w 60
OC
o
M
H
S 40 —.
O
I
w
Q
20 —
2 seconds residence time
4 seconds residence time
600
700 800
AVERAGE REACTOR TEMPERATURE, °C
900
Figure 7. Measured ortho-cresol decomposition as a function of average reactor
temperature for 2 and 4 s nominal residence time.
-------
100
80
2 seconds residence time (char solids)
3 seconds- residence time (char solids)
lime-bearing acceptor solids
homogeneous
c
-------
TABLE 5. SUMMARY OF PHENOL AND ORTHO-CRESOL FIRST-ORDER
DECOMPOSITION KINETIC PARAMETERS
Reaction
Homogeneous
Phenol (25)
(a)
Homogeneous
Ortho-Cresol (9)
Heterogeneous
Phenol (4)
Frequency Activation Energy, Coefficient of
Factor, In A kcal/gmole Determination,
16.5
11.0
19.8
39.1
23.2
31.9
0.96
0.97
0.997
(a) Number of data points.
TABLE 6. COMPARATIVE RATE CONSTANTS FOR PHENOL AND
ORTHO-CRESOL DECOMPOSITION
Phenol
Temperature, C
k /k
hetero homo
Homogeneous
o-cresol phenol
600
700
800
900
1840
1200
840
630
40
15
7
4
dreds of times higher (as high as 1,000:1 for
steam) on a molar basis.
Pathway to the Production of Phenols
During Coal Gasification
Production of phenols during coal gasification is
a complex function of gasifier configuration,
reaction conditions, and probably the chemistry
of the coal processed. The pattern underlying
production of these effluents includes initial
formation followed by subsequent decomposi-
tion within the gasification environment. The
primary pathway explaining production of phe-
nols during coal gasification is illustrated in
Figure 9 and consists of:
• Formation: Phenols are formed inherently
298
from coal, primarily as single aromatic ring
species. The formation of phenols is not ex-
pected to alter significantly as a result of dif-
ferent processing concepts. Similarly, coal
type (at least among those currently util-
ized) should not affect formation character-
istics to a major extent.
Decomposition: Thermal and/or catalytic
cracking phenomena controls production
characteristics of phenols. Sequential de-
composition of phenols occurs by dealkyla-
tion through lower homologues to phenol,
which decomposes to primarily gaseous
species. The rate of decomposition of phe-
nols is significantly enhanced by the pres-
ence of char solids. The rate-limiting step in
the reaction sequence is the final decomposi-
tion of the compound phenol.
-------
SINGLE-RING PHENOLS
MULTI-RING PHENOLS
MULTI-
_ _
PHENOLS
I
I
LOWER I
, PHENOLS _J
•-GASEOUS PROD.
•- HEAVY
HYDROCARBON
TARS
STARVED
H2 STARVED-
LOWER PHENOLS
GASEOUS PRODUCTS
LIGHT HYDROCARBON
OILS
•-XYLENOLS
•-METHANE
CRESOLS
•-METHANE
•-PHENOL
•-LIGHT HYDROCARBON
OILS
GASEOUS PRODUCTS
•-HEAVY HYDROCARBON
TARS
Figure 9. Pattern of phenolic compound production in coal gasification processing.
-------
IMPLICATIONS AND FUTURE WORK
ACKNOWLEDGMENTS
Necessary and quantitative data were gener-
ated during the course of this experimental
investigation, which effectively defines the de-
composition characteristics of phenols under
typical gasification conditions. The studies, com-
bined with previous work in the literature, pro-
vide a stepping stone for launching a variety of
additional experimental investigations. How-
ever, the following issues must be resolved:
• The effects of variations in process gas
steam content must be addressed. The dem-
onstrated role of steam in delineating a spe-
cific decomposition pathway for phenols
should be evaluated at lower steam-to-phe-
nol molar ratios.
• The relative effects of amount and type of
solid surface in enhancing phenol decomposi-
tion rates need to be addressed. The source
of catalytic activity on the char surface in
addition to definition of precise modes of
gas-solid mixing during gasification need to
be defined.
Implications of strategy and quantitative ex-
perimentation performed in this work have sig-
nificant impact upon the assessment and design
of coal conversion technology. From these
studies, it is evident that experimental deter-
mination of effluent production characteristics
at a single experimental scale is inadequate. A
particular scale of development provides either
too little information (i.e., an incomplete char-
acterization) or behavior that is too highly
coupled to measure without resorting to com-
plex sampling techniques. Judicious choice of
experiments across a range of process scales
can provide the information necessary to syn-
thesize quantitative effluent production pat-
terns amenable to process scaleup.
Quantitatively, the demonstrated sensitivity
of hydrocarbon effluent production (including
phenols) to changes in processing conditions
provides an alternative to the conventional
strategy of post-gasification effluent treatment.
Relationships previously developed, along with
those developed in the course of these studies,
can be used to control production of undesirable
hydrocarbon effluents. This strategy can be im-
plemented during process development, on
scaleup to commercial facilities, or for develop-
ment of generically similar novel processing
technologies.
Results presented herein reflect cooperative
efforts between Carnegie-Mellon University
and a range of personnel. The authors wish to
thank the Pittsburgh Energy Technology Cen-
ter for their support and assistance during
bench-scale phenolic compound decomposition
experiments and the Synthane PDU gasifier
trials. The authors also wish to thank Conoco
Coal Development Company and Steams-Roger,
Inc., for assistance in performing the C02-
Acceptor gasifier probe studies.
REFERENCES
l.Nakles, D. V. Significance of Process Vari-
ables on Liquid Effluent Production in Coal
Gasification (Ph.D. thesis). Carnegie-Mellon
University. Pittsburgh, Pa. 1978.
2. Jonardi, R. J., L. J. Anastasia, M. J. Mas-
sey, and R. H. Karst. Hygas Environmental
Characterization: Data Synthesis, Analysis
and Interpretation—Tests 37-64 (interim re-
port FE-2433-25 from the Institute of Gas
Technology). U.S. Department of Energy.
February 1979.
3. Quarterly Technical Progress Report,
Grand Forks Energy Research Center. U.S.
Energy Research and Development Admin-
istration. GFERC/QTR-76/5. November
1976. p. 16-32.
4. Ellman, R. C., B. C. Johnson, H. H. Shobert,
L. E. Paulson, and M. M. Fegley. Current
Status of Studies in Slagging Fixed-Bed
Gasification at the Grand Forks Energy
Research Center. (Presented at the Ninth
Biennial Lignite Symposium. Grand Forks.
May 1977.
5. Fillo, J. P., and M. J. Massey. Analysis of
RA-21 Effluent Data: GFERC Slagging
Fixed Bed Gasifier (interim report
FE-2496-24 from Carnegie-Mellon Universi-
ty). U.S. Department of Energy. April 1978.
6. Johnson, B. C., M. M. Fegley, R. C. Ellman,
and L. E. Paulson. Gasification of North
Dakota Lignite in a Slagging Fixed-Bed
Gasifier. Grand Forks Energy Technology
Center, U.S. Department of Energy. 1978.
7. Paulson, L. E., H. H. Shobert, and R. C.
Ellman. Sampling, Analysis, and Character-
ization of Effluents from the Grand Forks
Energy Research Center's Slagging Fixed-
300
-------
Bed Gasifier. Am Chem Soc Div Fuel Chem
Preprints. 23(2):107.1978.
8. Nakles, D. V., M. J. Massey, A. J. Forney,
and W. P. Haynes. Influence of Synthane
Gasifier Conditions on Effluent and Prod-
uct Gas Production. Pittsburgh Energy Re-
search Center, U.S. Department of Energy.
Pittsburgh, Pa. PERC/RI-75/6. December
1975.
9. Wells, G. L., and R. Long. Thermal Dealkyl-
ation-Hydrocracking of Alkyl Phenols. Ind
Eng Chem Process Des Develop 1(1X73.
1962.
10. Davies, G. A., and R. Long. The Kinetics of
the Thermal Hydrocracking of Cresols. J
Appl Chem. 15:117.1965.
11. Schmidt, C. E., A. G. Sharkey, and R. A.
Friedel. Mass Spectrometric Analysis of
Product Water from Coal Gasification.
Pittsburgh Energy Research Center, U.S.
Department of Energy. Bureau of Mines
Report TPR No. 86.1974.
12. Ho, C. H., B. R. Clark, and M. R. Guerin.
Direct Analysis of Organic Compounds in
Aqueous Byproducts from Fossil Fuel
Conversion Processes: Oil Shale Retorting,
Synthane Coal Gasification and COED Coal
Liquefaction. J Environ Sci Health.
Auami. 1976.
13. White, C. M., and C. E. Schmidt. Analysis of
Volatile Polar Organics in Untreated By-
product Waters from Coal Conversion
Processes. Am Chem Soc Div of Fuel Chem
Preprints. 23(2):134. 1978.
14. Jones, B. W., and M. B. Neuworth. Thermal
Cracking of Alkyl Phenols— Mechanism of
Dealkylation. Ind Eng Chem. 44(11)2872.
1952.
15. Damon, D. A. Aspects of Fine Particle
Fluidization (M.S. thesis). Carnegie-Mellon
University. Pittsburgh, Pa. 1972.
16. Fillo, J. P. An Understanding of Phenolic
Compound Production During Coal Gasifi-
cation Processing (Ph.D. thesis). Carnegie-
Mellon University. Pittsburgh, Pa. 1979.
17. Fillo, J. P., M. J. Massey, J. P. Strakey, D.
V. Nakles, and W. P. Haynes. Decomposi-
tion Characteristics of Phenol Under Syn-
thane Gasifier Conditions. Pittsburgh En-
ergy Research Center, U.S. Department of
Energy. Pittsburgh, Pa. PERC/RI-77/6.
April 1977.
18. Fillo, J. P., and M. J. Massey. Studies of
Phenolic Compound Decomposition Under
Synthane Gasifier Conditions (quarterly
technical progress report, April-June, 1978
from Carnegie-Mellon University). Pitts-
burgh Energy Research Center, U.S. De-
partment of Energy. July 1978.
19. Fillo, J. P., and M. J. Massey. Studies of
Phenolic Compound Decomposition Under
Synthane Gasifier Conditions (quarterly
technical progress report, July-September,
1978 from Carnegie-Mellon University).
Pittsburgh Energy Technology Center,
U.S. Department of Energy. November
1978.
20.Virk, P. S., L. E. Chambers, and H. N.
Woebke. Thermal Hydrogasification of Ar-
omatic Compounds. In: Coal Gasification,
Massey, L. G. (ed.). Washington, D.C. Amer-
ican Chemical Society, 1974.
21. Given, P. H. Reactions of Alkyl Phenols
Over Cracking Catalysts—I. Comparison of
Catalysts and Study of Reaction Condi-
tions. JAppl Chem. 7:172.1957.
22. Saha, N. C., N. G. Basak, and A. Lahiri. Hy-
drogenolysis of Higher Phenols to Lower
Homologues Part I—Factors Affecting Hy-
drogenolysis. J Sci Ind Res. 19Bffl. 1960.
301
-------
PREDICTIONS ON THE DISPOSITION OF SELECT TRACE
CONSTITUENTS IN COAL GASIFICATION PROCESSES
G. L. Anderson, A. H. Hill, and D. K. Fleming*
Institute of Gas Technology, Chicago, Illinois
Abstract
Many factors may affect the formation and
disposition of minor and trace constituents in
coal gasification processes; e.g., the coal feed-
stock, the gasification conditions, and the gas-
processing conditions. Adequate knowledge of
the forms and amounts of these constituents
and the controlling factors that dictate their
final disposition would be desirable for the com-
plete design of a full-scale plant However, the
current data base is weak for some of the more
volatile inorganic trace constituents that may
be gasified with the coal
In this paper, theoretical projections are made
on the amount and final disposition during coal
gasification of volatile trace constituents
formed from arsenic, boron, lead, selenium, and
mercury present in coals. Note that these predic-
tions are theoretical; they are intended to pro-
vide insight into what might occur in coal gasifi-
cation processes, to provide direction for future
experimental work for improving the data base
on these constituents, and to indicate areas
where further investigations would prove useful
for the design of full-scale processes.
The results of these studies indicate that high
recoveries of these trace inorganics are to be ex-
pected, with low discharge to the environment,
in most coal gasification process designs.
INTRODUCTION
Certain minor and trace inorganic constitu-
ents found in coal undergo chemical transforma-
tions during gasification. Some of these reac-
tions produce compounds that are volatile un-
der gasification conditions and leave the gasifier
as part of the raw gasifier product gas.
Present environmental assessment studies
are concerned with the disposition of potentially
toxic substances. However, because of the enor-
mous number of possible substances that may
•Speaker.
be present, emphasis has been on elemental ma-
terial balances around the primary gasifier and
quench system. For some elements, closing the
material balance is difficult because a signifi-
cant fraction of the material may be part of the
quenched product gas. In most cases, the quanti-
ty of these elements in the quenched product
gas is estimated by difference because these
materials are difficult to analyze. Further, the
analytical techniques often used for trace inor-
ganics are not of high accuracy. Sampling is also
difficult in certain cases because of absorption
or reaction of these volatile materials in the
sample containers. However, knowledge of the
compounds present and their approximate con-
centrations would simplify some of these analyt-
ical problems. Then, through further experi-
mental investigations, knowledge of the disposi-
tion of these volatile constituents, which are im-
portant both from an environmental viewpoint
and a processing viewpoint, can be increased.
As an indication of what might occur to cer-
tain elements in coal during gasification, a
theoretical analysis has been performed on the
formation and disposition of compounds contain-
ing arsenic, selenium, boron, lead, and mercury.
These elements were chosen for this study be-
cause earlier work had indicated that they were
the most likely elements to be removed from
the coal during gasification.1 Much of the
theoretical analysis is based on engineering
estimates and should not be taken as hard data.
Rather, this analysis should provide a starting
point for more definitive future investigations.
BEHAVIOR OF THE SELECT
TRACE ELEMENTS UNDER
GASIFICATION CONDITIONS
Processes Analyzed
The major differences between available coal
gasification processes are the operating condi-
303
-------
tions and the amount and distribution of hydro-
carbons produced. Gasification processes oper-
ate under reducing conditions, and the major
constituents produced, other than hydrocar-
bons, are H2, CO, C02, H^, H^, and NH8. In
this study the transformations that the select
elements may undergo were investigated as in-
fluenced by the operating conditions of three
gasification processes. The three processes
selected were the Koppers-Totzek process, the
Lurgi process, and the Hygas® SNG process
with steam-oxygen. Brief descriptions of these
processes follow.
Koppers-Totzek Process-
In this process, pulverized coal is reacted at
low pressure and high temperatures (> 1,800 K)
with steam and oxygen in an entrained bed with
cocurrent gas/solids contacting. A simplified
diagram of the gasifier is shown in Figure 1. The
high gasification temperature assures nearly
complete gasification of the carbon in the feed
coal. Approximately 50 percent of the ash in the
coal flows down the gasifier walls as molten slag
and drains into a slag quench tank. The remain-
der of the ash leaves the gasifier as fine par-
ticles entrained in the exit gas. These particles
are solidified at the gasifier exit by water
sprays and are subsequently separated from the
scrubber water and disposed of with the solid-
ified slag.
The high operating temperature and low op-
erating pressure of this process produce a raw
product gas from the gasifier primarily com-
prised of H2, CO, C02, and steam with minimal
amounts of hydrocarbons and tars. A typical
product gas is shown in Table 1.
Lurgi Process—
The Lurgi process employs a gravitating bed
of coal with continuous countercurrent gas flow,
as shown in Figure 2. Coal is fed intermittently
to the top of the reactor through pressurized
lockhoppers, while oxygen and steam are mixed
and fed into the bottom of the gravitating coal
bed. Gas temperature ranges from 590 K at the
top of the gasifier to 1,260 K at the bottom of
the gasifier. Normal operating pressures are 20
to 32 atm, and coal residence time is approx-
imately an hour. Typical raw product gas from
the gasifier is shown in Table 1. The major dif-
ference in the product gas from that of the
Koppers-Totzek process is that roughly 19 per-
cent of the feed carbon reacts to form methane
and ethane rather than carbon oxides. Approxi-
mately 12 percent of the feed carbon results in
tar, fatty acids, phenols, and BTX production.
Hygas Process—
The Hygas process uses three separate reac-
tion stages for gasifying coal. A diagram of the
gasifier is shown in Figure 3. Coal is fed to the
gasifier as a slurry made with either aromatic
oil or water. The oil or water is vaporized in a
fluidized bed (the slurry dryer) in the top of the
gasifier using heat available in the product gas
from the top reaction stage. The dried coal is
then gravity fed to the first stage, entrained,
and reacted in a low-temperature reactor stage
(LTR) with product gas from the lower sections
of the reactor. The operating temperature for
this section of the gasifier is usually between
920 and 1,060 K. The reacted coal from the first
stage is then disengaged from the gases and
gravity fed to the fluidized-bed second-stage hy-
drogasifier (the high-temperature reactor),
where it is reacted with product gas from the
lowest stage of the gasifier at temperatures be-
tween 1,030 and 1,170 K. Finally, the reacted
coal from the second stage is gravity fed to the
lowest stage of the gasifier, the 80G (steam-
oxygen gasifier), where the remaining carbon in
the coal is reacted with high-pressure steam and
oxygen at temperatures between 1,170 and
1,280 K. The normal operating pressure of the
Hygas process is 69 to 100 atm. A typical raw
product gas from this gasification process is
shown in Table 1. In this process, methane and
ethane account for 27 percent of the feed car-
bon, while 8 percent or less of the feed carbon
produces BTX and phenol with minimal produc-
tion of tars and fatty acids.
Trace Element Chemistry
During Gasification
With the background above, the analysis of
the transformations that arsenic-, selenium-,
boron-, lead-, and mercury-containing com-
pounds might undergo during gasification
follows.
Arsenic Chemistry-
Arsenic concentrations in U.S. coals range
from 0.5 ppm to 93 ppm, with an average of 14
ppm. The major form of arsenic in coal was
304
-------
GAS
FUDW
LOW
PRESSURE
STEAM
co
o
01
BOILER
FEED
WATER
FOUR
HEADED
GASIFIER
COAL
STEAM
OXYGEN
BOILER FEED WATER » ^Z
•-ASH DISCHARGE
A79 020334
Figure 1. Koppers-Totzek gasifier.
-------
TABLE 1. OPERATING CHARACTERISTICS OF THE KOPPERS-TOTZEK. LURGI, AND
HYGAS STEAM-OXYGEN GASIPIERS USING SUBBITUMINOUS COAL
7 1
Pressure , atm
Temperature , K
Koppers-Totzek
2-3
1800-2000
Lurgi
20-32
590 (top)
1260 (bottom)
L
HYGAS
69-100
590 (top) 920-1060 (1st stage)
1030-1170 (2nd stage)
1170-1280 (SOG)
Product Gas, mole fraction
H20
H-
2
CO
co2
N2
CH4
C2'C5
BTX
H2S
COS
NH_
3
HCN
HC1
C^HCOH
O J
Tar
Fatty Acids
g-raol/g-coal
* Product gas does
t Product gas does
0.0801*
0.3039
0.5428
0.0651
0.0052
—
—
-
0.0026
0.0002
0.00003
0.00002
0.0001
—
—
—
0.0778
not include water added
not include oil or water
0.4659
0.2172
0.0800
0.1638
0.0005
0.0591
0.0045
0.0023
0.0026
0.0002
0.0032
0.000002
0.00002
0.0004
0.1660 g/g-mol
0.0167 g/g-mol
0.1165
from water sprays.
used for slurrying coal.
0.4265T
0.1500
0.0843
0.1950
0.0003
0.1248
0.0064
0.0068
0.0023
0.0001
0.0034
0.00006
0.00002
0.0005
—
—
0.0969
-------
FEED COAL
RECYCLE TAR
DRIVE
SCRUBBING
COOLER
GRATE
DRIVE"
STEAM +
OXYGEN
WATER JACKET
79020333
Figure 2. Lurgi pressure gasifier.
307
-------
RAW GAS OUTLET
TO PURIFICATION
AND METHANATION STEPS
INLET FOR SLURRY
OF CRUSHED COAL
AND LIGHT OIL
NITROGEN-PRESSURIZED
OUTER SHELL
FLUIDIZED BED IN
WHICH SLURRY OIL IS
VAPORIZED BY RISING,
HOT GASES AS
COAL DESCENDS
HOT GAS RISING
INTO DRIER
DRIED COAL FEED
FOR FIRST-STAGE
HYDROGASIFICATION
HYDROGASIFICATION
IN COCURRENT FLOW
OF GAS AND SOLIDS
HIGH VELOCITY GAS
FROM SECOND-STAGE
MIXES WITH DRIED COAL
HOT GAS RISING
INTO FIRST-STAGE
CHAR FROM FIRST STAGE
FEEDS INTO SECOND-
STAGE FLUIDIZED BED
RISING GASES CONTACT
DESCENDING CHAR
HYDROGEN-RICH GAS
AND STEAM RISE
TO SECOND-STAGE
FLUIDIZED BED
HYDROGASIFIEO CHAR
FROM SECOND-STAGE
FEEDS INTO STEAM-
OXYGEN GASIFIER
SLURRY
DRIER
RAW GAS
X FIRST-STAGE
/ HYDROGASIFICATION
\ SECOND-STAGE
/ HYDROGASIFICATION
\ STEAM-OXYGEN
/ GASIFIER
ASH
NOTE' THIS SIMPLIFIED SKETCH
IS NOT DRAWN TO SCALE
D-IOS-I9IS
Figure 3. Hygas gasifier with steam/O2 gasification.
308
-------
TABLE 2. FREE ENERGIES OF FORMATION OF ARSENIC-CONTAINING
COMPOUNDS, kcal/mole
Basis: Elements in
Compound
As2(g)
As4(g)
FeAsS*
FeAs2* (loellingate)
FeAs *
Fe2As *
As2S *(high orpiment)
As»S * (low orpiment)
As2S *(low realgar)
AsH3(g)
As203(g)
As20?(s)
As203(l)
FeS
* 9
Barton
Their Standard
600 K
+ 28.93 +
+ 10.0 +
- 30.6
- 14.61
- 9.24
- 9.90
- 33.26
- 33.24
- 23.50
-I- 17.88 +
- 116.8
- 119.0
- 119.5
- 28.7
State
iem{
800 K
21.48
3.0
27.46
16.03
10.92
11.50
21.38
21.32
15.30
19.52
111.5
106.8
110.2
26.2
at Temperature
1000 K
+ 14.23
4.1
- 24.20
- 17.45
- 12.60
- 14.10
- 9.50
- 9.40
- 7.10
+ 21.50
- 106.0
- 94.8
- 101.1
- 23.75
of Interest
1200 K
+ 7.18
- 11.2
- 20.94
- 18.87
- 14.28
- 16.70
- 2.38
- 2.52
- 1.10
+ 23.50
~ 100.6
~ 82.3
" 92.0
- 21.8
deduced by Duck and Himus as arsenopyrite.8
Under gasification conditions when tempera-
tures exceed 820 K, arsenopyrite begins to
decompose into pyrrhotite (FeS) and metallic
arsenic.
FeAsSts) >820K PeSfa) + As(s).
(1)
At temperatures greater than 1,025 K, the de-
composition proceeds rapidly. This decomposi-
tion has been observed in laboratory studies by
many investigators including Zhuchkov,'
Zviadaze et al.,7 and Lukesh.8 This observation
is not totally consistent with thermodynamic
data on the iron-arsenic-sulfur system measured
by Barton* but is within the experimental error
associated with these measurements. Barton's
data, along with other available thermodynamic
data, are given in Table 2.
Once arsenopyrite decomposes into pyrrho-
tite and metallic arsenic, the metallic arsenic
can theoretically vaporize as As4 from the coal.
However, this does not—apparently —occur.
The trace element data from the Hygas process
indicate that arsenic loss from the coal does not
occur until the coal reaches the SOG stage of the
reactor where the temperature ranges from
1,170 to 1,280 K.10 Therefore, either the arseno-
pyrite is embedded in the coal-ash matrix and
volatilization is diffusion controlled, or the ar-
senopyrite is so highly dispersed that formation
of As4 is limited and volatilization occurs by
means of A$2 or As, which have lower vapor
309
-------
pressures.
Assuming elemental arsenic volatilization
does not occur until temperatures in excess of
1,170 K are reached, the volatilized elemental
arsenic must travel through the gasifier. At
these conditions, the stable forms for arsenic
are As4, AsH3 (arsine), and As2. The question
that arises is: How fast will elemental arsenic
react with available hydrogen to form AsH3?
The normal preparation of AsH3 is by reaction
of ALAs or Na3As with water11 or from a mix-
ture of NaBH4 and AsCl3 in water at pH 8 to
pH 10 and 5° to 40° C.12 No data are reported on
the production of AsH3 from arsenic and H2
because this route would provide extremely
small yields of AsH3 at temperatures where the
kinetics would be favorable.
To estimate the rate at which arsine is pro-
duced from elemental arsenic and hydrogen, the
gas phase arsine decomposition kinetics studied
by Kedyarkin and Zorin13 were combined with
the free energy of formation of arsine from As4
and H2 using the law of microscopic reversibili-
ty to derive an expression for the forward reac-
tion of:
d[AsH
dt
0 1
z- - 5.22 x 107
exp (- 54,6lO/RT)[As4]1/4[H2]8« (2)
Concentrations are in atmospheres and time is
in seconds. Based on this equation, equilibrium
control occurs when temperatures are greater
than 900 K.
Because of the rapid cooldown of the product
gases in the lower temperature zones of the Hy-
gas and Lurgi reactors, the arsine-arsenic-
hydrogen reaction is assumed to be frozen at an
equilibrium temperature of 900 K. In the Kop-
pers-Totzek process, the reaction is assumed to
be frozen at 1,800 K because of rapid cooldown
of the gases with the water sprays. However, if
quench of the Koppers-Totzek gases is not in-
stantaneous, small amounts of arsine may be
formed.*
The predicted distribution of elemental ar-
senic and arsine in the raw product gases from
these three processes is shown in Table 3 for
three different arsenic levels. Based upon data
'•'This effect would be more pronounced in Texaco or
Shell gasifiers, which operate in a mode similar to a
Koppers-Totzek gasifier but at higher pressure.
from the Hygas pilot plant, 50 percent of the
arsenic in the feed coal is assumed to be
volatilized in the Hygas and Lurgi processes,
whereas 100 percent is expected to be volatil-
ized in the Koppers-Totzek process because of
the high-temperature, single-stage operating
mode. It is evident that operation at moderate
temperatures, high pressures, and high arsenic
concentrations increases the amount of arsine.
Operation at high temperatures and low pres-
sures favors the formation of elemental arsenic
with negligible amounts of arsine formation.
Verification of these predictions is not yet
possible; the search for the presence of arsine in
operating coal gasification plants is not known
to have been performed. As discussed later, ex-
perimental investigation of the amount of arsine
formation should be of considerable importance.
Selenium Chemistry—
Similar to arsenic, selenium may be initially
present in coals as selenopyrite, FeSeS. Seleni-
um concentrations in U.S. coals range from a
minimum of 0.45 ppm to a maximum of 7.7 ppm,
with an average of about 2.25 ppm. In moderate
temperature gasification processes such as
Hygas and Lurgi, between 30 percent and 70
percent of the selenium is volatilized from the
coal. Based on evidence from the Hygas pilot
plant, most of this gasification occurs in the hot-
test section of the gasifier.10 In the high-
temperature Koppers-Totzek process, no infor-
mation on percent selenium loss is available, but
quantitative volatilization is expected.
The mechanism for volatilization of selenium
from the coal is postulated as either decomposi-
tion of FeSeS to FeS and Se(g) or reaction of H2
with FeSeS to produce H2Se and FeS. This ap-
proach, lacking kinetic data, assumes initial
FeSeS decomposes, liberating Se2(g). Then the
Se2 can react with H2 to form H2Se, which is the
thermodynamically preferred form.
Estimation of the kinetics of forming H2Se
from Se2 and H2 was based on the following
mechanism:
Se2
HSe
H -
H2
HSe + Se.
H2Se
H.
(3)
(4)
Reaction 4 is the reverse mechanism for the ini-
tial step of H2Se decomposition, which has been
determined to decompose according to the fol-
lowing rate expression:14
310
-------
TABLE 3. PREDICTED ARSENIC DISTRIBUTION IN GASIFIER RAW PRODUCT GAS
Process
Equilibrium Temp . , K
H Partial Pressure, atm
Wet Raw Gas Production, g-mol/g-coal
Estimated Arsenic Volatilization, %
At Maximum Arsenic Concentration
in Feed Coal (93 ppm)
AsH3
As4
As2
At Average Arsenic Concentration
in Feed Coal (14 ppm)
AsH3
HYGAS
900
12.05
0.0969
50
4.79 X 10~7
1.48 X 10~6
4,96 X 10*~9
2.79 X 10~?
1.71 X 10~?
Lurgi
900
4.43
0.1165
50
mol r T*3c tion
2.86 X 10~7
1.26 X 10~6
9.00 X 10~9
1.71 X 10~7
1.56 X 10~7
Koppers-Totzek
1800
0.62
0.0778
100
1.64 X 10"14
4.20 X 10~10
7,98 X 10~6
2.47 X 10~15
9.56 X 10~12
As
1.77 X 10
-9
3.20 X 10
-9
1.20 X 10
-6
At Minimum Arsenic Concentration
in Feed Coal (0.5 ppm)
AsH
As
3.40 X 10
3.54 X 10
2.48 X 10
-8
-11
-11
2.78 X 10
1.13 X 10
8.74 X 10
-8
-10
-11
1.76 X 10
4,88 X 10
4.29 X 10
-16
-14
I
-8
-------
-d[H2Se]
dt
. 8.8 x 10
exp(-1500/BT)[H2SeIH].
(5)
Concentrations are in moles per liter and the
units of time are seconds.
Assuming microscopic reversibility, the for-
ward rate of Reaction 4 is given by:
"
dt
, 2.4x10
exp ( - 26,792/RT)[HSeIH2].
(6)
The forward reaction given by Equation 3 can
be estimated using available correlations for ab-
straction reactions.15 The rate of reaction of 863
with H is given approximately by:
_,e2 - 1018 exp (- 8118/BT)[Se2IH]. (7)
dt *
Assuming the concentration of atomic hydrogen
at these conditions is always in equilibrium with
the amount of H2 present, Equation 7 can be
written as:
-d[SeJ 1B
2 - 1.11 x 1016
exp (- 61,612/RT»tSe2IH2]1/2. (8)
dt
At temperature less than 2,000 K, Reaction 4 is
much faster than Reaction 3, indicating that
Reaction 3 is rate controlling. If the expression
is correct, equilibrium control occurs at temper-
atures greater than 900 K. Assuming 900 K
equilibrium control in the Hygas and Lurgi
processes and 1,800 K in the Koppers-Totzek
process, the selenium present in the raw prod-
uct gas is almost exclusively H2Se.
Boron Chemistry-
Boron is present in coal at concentrations be-
tween 2 and 224 ppm, with an average of 67
ppm.-Evidence suggests that most of the boron
is chelated.18 Environmental assessment data
from the Hygas and Lurgi processes indicate
that about 60 percent of this boron is volatilized
during gasification. In the Koppers-Totzek proc-
ess, quantitative volatilization is anticipated.
Chelated boron, when treated with hydrogen
at high temperatures, produces BH8. However,
B(OH)8 is the thermodynamically preferred
form in coal gasification environments based on
data in Table 4.
Hydrolysis of the borane produced should oc-
cur. The hydrolysis of diborane B^ to boric
acid and hydrogen has been used for quantita-
tive analysis of diborane in gas mixtures. The
mechanism suggested for this reaction is given
by:18
(9)
BH3 + H20 = BH8 *
BH20H + H20 = BH2OH *
— BH(OH)2 + H2-
BH(OH)2 + H20 = BH(OH)2 *
— B(OH)2 + H2-
BHgOH +
(10)
(11)
(12)
The rate-controlling step is believed to be the
initial attack of H20 on BH8. The kinetics of this
reaction have been deduced to be:
d[BH,OH] ,
- — ^— - - 1.9 x 106
exp(-6000/BT)(BH8IH20], (13)
where concentrations are given in moles per
TABLE 4. FREE ENERGIES OF FORMATION17 OF BORON-CONTAINING COMPOUNDS,'
COMPOUNDS, kcal/mole
Compound
BH3(g)
B2H6(g)
BH3CO(g)
B(OH)3(g)
B(g)
600 K
28.06
35.50
-16.01
-205.6
111.5
Temperature
900 K
30.33
50.13
-8.206
-118.7
100.7
1200 K
32.95
65.02
0.8710
-171.5
90.0
1800 K
38.64
94.67
21.87
-136.8
68.88
312
-------
liter and time is expressed in seconds.
Therefore, rapid attainment of equilibrium is
assured at all temperatures in a gasification
process with quantitative production of B(OH)3.
Lead Chemistry-
Lead in coal is generally believed to exist in-
itially as PbS, with an average lead concentra-
tion of 39.2 ppm. Environmental assessment
data on lead losses from coal during gasification
indicate conflicting results. Minimal loss of lead
is reported in Lurgi operations and Hygas pilot-
plant operations. However, in Hygas PDU stud-
ies, between 30 percent and 60 percent of the
lead contained in the coal was volatilized. This
disparity is because of the single-stage high-
temperature conditions used in the PDU stud-
ies. At these temperatures, the vapor pressures
of many lead-containing compounds are appre-
ciable, leading to loss from the feed coal. The
free energies of formation for a number of lead
species at 600 K, 1,200 K, and 1,800 K are shown
in Table 5. At 1,200 K in a coal gasification en-
vironment, the vapor pressure of PbS is the
largest of the lead-containing compounds at
8.83 x 10"1 atm. If this vapor pressure were
achieved in the Hygas PDU studies, quantita-
tive loss of the lead from the coal would occur.
However, in an integrated Hygas process and in
the Lurgi process, the product gas from the hot-
ter sections of the gasifier is eventually cooled
down to temperatures of 600 K by the raw feed
coal. At this temperature, the vapor pressures
of lead-containing compounds are significantly
reduced. The anticipated concentration of
gaseous lead compounds in the raw product gas
from the Lurgi and Hygas processes is shown in
Table 6. This loss represents parts-per-trillion
levels of equivalent lead in the feed coal.
The Koppers-Totzek process, which operates
at 1,800 K, is expected to volatilize all the lead
present in the feed coal but, again, these will be
solidified during quench by the water sprays.
The only difference postulated between the
Koppers-Totzek and the Hygas and Lurgi proc-
esses is that the volatilized lead will be Pb and
PbS rather than PbCl2.
Mercury Chemistry
The average concentration of mercury in coal
is 0.2 ppm. The range of concentrations is 0.02
ppm to 1.60 ppm. At the high temperatures em-
ployed in coal gasifiers, quantitative loss from
the coal is expected. Thermodynamic calcula-
tions have been used to estimate the probable
chemical form of mercury in a coal gasification
environment, because the initial form and kinet-
TABLE5. FREE ENERGIES OF FORMATION OF LEAD-CONTAINING COMPOUNDS,17
kcal/mole
Temperature
PbS (g)
PbS (s)
PbC03 (s)
PbCl2 (s)
PbCl2 (g)
PbO (g)
PbO (s)
Pb (g)
600 K
10.00
-22.23
-126.07
-64.86
-48.24
6.634
-37.94
31.10
1200 K
0.29
-12.89
-91.73
-51.18
-50.62
-0.82
-24.06
17.93
1800 K
-9.423
-3.556
-57.40
-37.50
-52.96
-8.266
-10.17
4.768
313
-------
TABLE 6. LEAD CONCENTRATIONS IN COAL GASIFICATION RAW PRODUCT GASES
*LureiHYGAS
-g-mol/g-mol product gas
PbS
PbCl.
2
PbO
Pb
Total
8.13 X 10~14
1.11 X 10~U
2.81 X 10~25
6.94 X 10~16
1.11 X 10~U
2.05 X 10~14
2.46 X 10~U
7.65 X 10"26
1.42 X 10~16
2.46 X 10'11
ics of transforming mercury from one form to
another are not known. Mercury species includ-
ed in the calculations were Hg(g), HgS(g),
HgH(g), HgCtyg) and HgF2(g). The thermody-
namically preferred form in the presence of the
gas is Egig).
Based on this assumption, the quantity of
mercury in the raw gasifier product gas from
the Hygas, Lurgi, and Koppers-Totzek proc-
esses is shown in Table 7. Mercury concentra-
tions assumed in the feed coal were 0.02, 0.2,
and 1.8 ppm.
EFFECT OF GAS
PROCESSING ON TRACE
ELEMENT DISPOSITION
The Purification System
Estimates of the quantities and chemical
forms of the trace elements in the gasifier raw
product gas permit projections on the final
disposition of these compounds in downstream
processes. A typical gas-processing scheme for
a coal gasification plant to produce substitute
natural gas (SNG) is shown in Figure 4.
The first step is a cooling of the raw product
gas to about 300 K using waste heat recovery,
air cooling, and, finally, water cooling. Equi-
librium between gas and liquids is usually
assumed. In this system excess steam; condensi-
ble impurities such as oil and tar; and soluble
impurities such as phenol, ammonia, hydro-
chloric acid, and thiocyanate are removed at
pressure.
The product gases then enter a selective HgS
acid-gas removal section where 99 percent of
the H2S and part of the C02 are removed.
Regeneration of the solvent in this system pro-
duces an acid-gas stream containing about 15
percent HgS, with the balance primarily G02.
After HgS removal, the remainder of the C02
is removed in a second acid-gas removal section.
The product gas, now free of acid gas and oil, is
TABLE 7. MERCURY CONCENTRATIONS IN RAW GASIFIER PRODUCT GASES
procegg
Mercury in
Feed Coal, ppm
HYGAS
g-mol/g-mol raw product gas
Koppers-Totzek
0.02
0.2
1.6
1.03 X 10
1.03 X 10
8.23 X
-9
"8
8.55 X 10
8.55 X 10
-10
-9
6.84 X 10
-8
1.28 X 10
-9
1.28 X 10
1.03 X 10
-8
-7
314
-------
COAL
CO
>-i
Ol
GASIFIER
i
ASH
H
(
o
i
1 OIL
H2S
REMOVAL
DISSOLVED
GASES H2S
OIL
DEPRESSURIZATION
I
WASTEWATER
TREATMENT
t
BY PROD
OIL
i
CLAUS PLANT
UCT 1 1
SULFUR TAIL GAS
1 F
)-C02 REMOVAL
SOLIDS
FROM H2S
REMOVAL
C02(I)
REMOVAL
_^b
COMBINED SHIFT
AND METHANATION
_^
^
C02(II)
REMOVAL
-^-
GLYCOL
COOLER
t ill
C02(I) H20 C02(II) H20
SNG
Figure 4. Typical gas producing configuration for producing SNG.
-------
heated to temperatures greater than 500 K and
enters a catalytic processing section where a
combination of the water gas shift reaction and
methanation reaction occurs producing a gas
containing only CH4, C02, and H20, with
residual amounts of H2 and CO. The water is
removed by cooling and the C02 is removed in a
final acid-gas removal section. Moisture in the
product SNG is then removed by a glycol cooler,
and the SNG is sent to the pipeline for distribu-
tion.
Effects In Quench
Assuming a coal feed with average concentra-
tions of As, Pb, B, Se, and Hg, the chemical form
and estimated quantities of these elements in
the raw gasifier product gas for the Lurgi,
Hygas, and Koppers-Totzek processes are
shown in Table 8. Quenching the gases from
these processes should quantitatively remove
the As4, As2, B(OH)8, PbS, PbCl2, PbO, and Pb
because of vapor pressure or solubility consid-
erations.
The elemental arsenic will exist as suspended
solids, probably condensed on coal fines from
the gasifier, in the excess condensate. The lead
compounds will also be present as suspended
solids because the presence of dissolved H.jS in
the water will suppress the solubility of these
compounds. These materials will, therefore, be
removed from the process with the other sus-
pended solids in the excess condensate, as it is
purified for recycle.
The boric acid will dissolve in the excess con-
densate and report to the sour water treatment
TABLE 8. FORM AND DISTRIBUTION OF SELECT TRACE ELEMENTS
IN RAW GASIFIER PRODUCT GAS
Process
Operating Pressure, atm
20
HYGAS
80
Koppers-Totzek
2
g-mol/g-mol raw product gas
1.71 X 10
1.56 X 10
3.20 X 10
-7
-7
2.79 X 10
-7
1.71 X 10
-7
1,77 X 10
-9
2.47 X 10
-15
9.56 X 10
-12
1.20 X 10
-6
H2Se
B(OH).
1.36 X 10
-7
4.88 X 10
-5
1.13 X 10
-7
2.43 X 10
-5
3.68 X 10
7.92 X 10
-5
PbS
PbCl,
i
PbO
Pb
8.13 X 10
-14
1.11 X 10
-11
2.81 X 10'
-25
6.94 X 10
-16
2.05 X 10
-14
2.46 X 10
-11
7.65 X 10
-26
1.42 X 10
-16
7.62 X 10
1.30 X 10
-7
-12
2.31 X 10
-9
1.67 X 10
-6
Hg
8.55 X 10
-9
1.03 X 10
-8
1.28 X 10
-8
316
-------
section of the plant. The boron content in the
sour water stream is computed to be 34 ppm for
the Hygas process, 63 ppm for the Lurgi proc-
ess, and 590 ppm for the Koppers-Totzek proc-
ess. These values compare favorably with ex-
perimental values obtained in a survey analysis
of potential toxic/inhibitory elements to biolog-
ical oxidation of a Hygas pilot-plant sour
water.19 The boric acid wUl be removed from the
water in the system that removes other soluble
salts.
The amount of arsine in the quenched raw
product gas is expected to be unaffected by the
quench system. Arsine solubility in water is
negligible. The solubility of AsH8 in H20 at
300 K and atmosphere partial pressure is
1.787 x 10"4 g-mol/g-mol H20. Therefore, the
fraction of AsH8 that may dissolve in the con-
densate derived from quenching the raw prod-
uct gases is 1.03 percent for the Hygas process,
0.31 percent for the Lurgi process, and 0.004
percent for the Koppers-Totzek process. The
resultant concentrations of arsine in the excess
condensates will be on the order of 1 ppb, sev-
eral orders of magnitude below current environ-
mental standards, even before water treatment
for recycle.
In the Hygas and Lurgi processes, a signifi-
cant quantity of aromatic oil is also recovered
during quench. The fraction of AsH9 that may
dissolve in this oil is estimated at 4.10 percent
for the Hygas process and 0.31 percent for the
Lurgi process. However, depressurization of
this oil will liberate most of the AsH8. These
liberated gases, because of their quantity, will
be recompressed and returned to the quenched
raw product gas.
Hydrogen selenide removal in the quench sys-
tem is also expected to be negligible. The solu-
bility of H2Se in water is slightly less than that
of HgS. The Henry's constant for H2Se in water
is 963.76 atm at 300 K.20 Therefore, the pre-
dicted concentrations of H2Se in the excess con-
densate are 0.07 ppm for the Hygas process,
0.02 ppm for the Lurgi process, and 0.002 ppm
for the Koppers-Totzek process. These values
are also well below proposed environmental
standards for discharge, even before treatment.
In the Hygas and Lurgi processes, about 7
percent and 0.6 percent of the H2Se will initially
be dissolved in the product oil. However, as
with AsH8, depressurization will flash the H^e.
The H2Se will, therefore, be returned to the
quenched raw product gas.
The quench system will not remove much of
the mercury in the raw product gas. After con-
densation of steam and oil from the product
gases, the partial pressure of Hg in all these
processes is below its vapor pressure of 3.4 x
10~° atm at 300 K. To estimate the solubility of
mercury in the excess condensate, it is assumed
that its solubility was proportional to its partial
pressure with a value of 0.25 mg/L at 3.4 x 10~6
atm, which is the solubility of metallic mercury
in water at 300 K. Based on this assumption, the
amount of Hg removed from the raw product
gas by the excess condensate is 4 percent for
the Hygas process, 1.1 percent for the Lurgi
process, and 0.01 percent for the Koppers-
Totzek process. The resultant concentration of
Hg in this condensate is 0.011 ppmw for the
Hygas process, 0.0023 ppmw for the Lurgi proc-
ess, and 0.0002 ppmw for the Koppers-Totzek
process.
Solubility of mercury in the condensible oil
fractions of the Hygas and Lurgi processes is
not known. Because of the recompression of
flashed gases from this oil, it is assumed to be
negligible.
In summary, the compounds under considera-
tion that remain in the product gas after quench
are AsH8, H^e, and Hg. The projected amounts
remaining are shown in Table 9.
Effects During Sulfur Removal
The next process these gases encounter in
the typical gas processing scheme is the selec-
tive H2& removal system. For high-pressure
processes such as Hygas and Lurgi, a physical-
type solvent would probably be used because of
lower costs. For low-pressure processes such as
Koppers-Totzek, a chemical-type solvent would
be used. However, the Koppers-Totzek system
might also use a physical solvent, if the gas is to
be compressed for delivery, and the chemical
system might, on occasion, be used with the
other gasifiers. The analysis below is based on
the most likely acid-gas removal system to be
used. Also, for completeness, an analysis is per-
formed on the use of a chemical acid-gas re-
moval system with a high-pressure process.
In the physical solvent systems, solubility
data on these species are unavailable; however,
nearly quantitative removal of AsH8 and H^e
is expected. This assumption is based on the
317
-------
TABLE 9. FORM AND DISTRIBUTION OF SELECT TRACE ELEMENTS
IN QUENCHED GASIFIER PRODUCT GAS
Operating Pressure, atm
AsH,
Hg
Process
HYGAS
80
Koppers-Totzek
2
g-mol/g-mol quenched product gas
3.19 X 10
2.50 X 10
-7
4.81 X 10
-7
-7
1.85 X 10
-7
1.58 X 10
-8
1.72 X 10
-8
2.68 X 10
-15
4.00 X 10
-7
1,38 X 10
-8
lower vapor pressures of AsH3 and H2Se rela-
tive to H2S; this indicates AsH3 and
should be more soluble in the solvent than
The removal of mercury in physical solvent
systems is more difficult to predict because
solubility of Hg in solvents is not given by
Raoult's law. However, these physical solvent
systems operate at temperatures where a sig-
nificant part of the Hg may condense. The
Selexol process operates at temperatures
around 280 K. The vapor pressure of Hg at this
temperature is ~6 x 10 ~7 atm with 56 percent
condensation of Hg. Condensation of up to 99
percent of the mercury in the Lurgi-Rectisol
quenched gas stream may occur at operating
temperatures of 230 K.
In the Benfield process, which is a chemical-
type solvent system that might be used with the
Koppers-Totzek process, the normal operating
temperature is 390 K. The removal of arsine
predicted from its solubility in water is negligi-
ble. Removal of mercury is also negligible be-
cause of the high temperature and low partial
pressure of Hg (-2.36 x 10 ~8 atm). However,
quantitative removal of H2Se is expected. The
pKa of H2Se is 4. The pKa of H2S is 7. Therefore,
dissociation of H^e into H+ and HSe~ in a
chemical-type solvent is greater than that of
H2S.
Similarly, if the Benfield process were used in
a Hygas plant, hydrogen selenide would be
nearly quantitatively removed. Arsine and mer-
cury removal will be slightly larger than that
predicted in the Koppers-Totzek process be-
cause of higher partial pressures for these com-
pounds.
The resulting distribution of arsine, hydrogen
selenide, and mercury for these processes is
shown in Table 10. This distribution is based on
the assumption that the H2S selective removal
system is designed to produce an H2S-rich acid-
gas stream containing 15 percent H2S.
The H2S-rich acid-gas stream is then assumed
to go to a Glaus process for production of
elemental sulfur. Although minimal process
problems are anticipated because of the
presence of these trace constituents, contamina-
tion of the byproduct elemental sulfur may oc-
cur. The typical levels of selenium and arsenic in
industrial grade sulfur are less than 2 ppm and
less than 0.25 ppm, respectively. The arsine and
hydrogen selenide in the feed should convert to
arsenious oxide and elemental selenium in the
combustion zone of the Glaus plant. These forms
will precipitate with the elemental sulfur. Based
on the predicted concentrations of these
elements in the H2S-rich acid-gas, the concentra-
tions of arsenic and selenium in the product
sulfur will range from 0 to 280 ppm by weight
and 113 to 348 ppm by weight, respectively. The
presence of elemental sulfur and H2S in the
combustion and catalytic zones of the plant
should convert the mercury to HgS if minimal
H2 is present. The anticipated range of the mer-
cury content of the product sulfur is 0.006 to 20
ppm.
This contamination could render the product
sulfur unfit for many industrial applications.
However, most sulfur is used for sulfuric acid
production for fertilizer. Sulfuric acid manufac-
318
-------
TABLE 10. PROJECTED ARSINE, HYDROGEN SELENIDE, AND MERCURY LEVELS
IN H2S-FREE PRODUCT-GAS STREAM, H2S-RICH
ACID-GAS STREAM, AND PRODUCT SULFUR
Gasification Process Lurgi
Acid-Gas Process Rectisol
AsH- < 0.003
H2Se < 0.002
Hg 0.00016
00
« AsH3 10.38
HjSe 6.47
Hg 0.489
As 162
Se 106
Hg 20.4
*Neg - negligible
HYGAS
Selexol
i 2»
< 0.005
< 0.002
0.0078
HC D-I/>Vt
17.97
8.79
0.387
i roduct
280
144
16.15
Koppers-Totzek
Benfield
: Produc t Gas , ppiuv
Neg
< 0.004
0.014
Neg
19.36
0.00014
Sulfur, ppmw
Neg
318
0.006
HYGAS
Benfield
0.49
< 0.002
0.017
0.37
8.79
0.00014
5.8
144
2.24
-------
turers do have means of accommodating these
contaminents in new, properly designed plants;
an older plant might not be able to use this ma-
terial, reducing its byproduct value. The proper
solution, of course, is to manufacture byproduct
acid, rather than elemental sulfur, at the gasifi-
cation plant. This option not only recovers the
initial sulfur byproduct value but produces a
more valuable byproduct than elemental sulfur.
The predicted concentrations of AsH3 and
H2Se in the H2S-free product-gas streams from
the Lurgi and Hygas processes are conserva-
tively based on only 99 percent removal for
these compounds in the H2S selective removal
step. Likewise, only 99 percent removal of H2Se
is assumed for the Koppers-Totzek process.
More realistically, more than 99.9 percent re-
moval should be expected.
Effects During Initial C02 Removal
These H2S-free product gases then enter the
C02 removal process, which is similar to the
H
-------
co
to
TABLE 11. PROJECTED CONCENTRATIONS OF ARSINE, HYDROGEN SELENIDE,
AND MERCURY IN THE H2S/CO2-FREE PRODUCT-GAS
AND CO2 VENT-GAS STREAMS
Gasification Process
Acid-Gas Process
AsH3
H2Se
Hg
AsH3
H2Se
Hg
* Neg = negligible
Lurgi
Rectisol
*
Neg
Neg
0.00016
<0.009**
<0.007**
0.00016
HYGAS
Selexol
HO 1 f*(\ T?i
~O/ L.U0— r]
2. 2.
Neg
Neg
0.0078
2
<0.014**
<0.005**
0.0078
Koppers-Totzek
Benfield
ree Product-Gas Stream, ppmv -
Neg
Neg
0.015
Neg
<0.001
Neg
HYGAS
Benfield
0.694
Neg
0.021
0.0631
<0.005
0.0090
**Predictions based on conservative 99% removal in the first stage of acid gas
removal. Actual concentrations are expected to be an order of magnitude lower.
-------
TABLE 12. PROJECTED ARSINE, ARSENIC, AND MERCURY CONCENTRATIONS IN
RAW METHANATION PRODUCT-GAS STREAM, QUENCHED METHANATION
PRODUCT-GAS STREAM, AND PRODUCT SNG
CO
Gasification Process Lurgi
Acid-Gas Process Rectisol
AsH3 Neg*
As4 Neg
Hg 0.00032
AsH3 Neg
As4 Neg
Hg 0.00032
AsH3 Neg
As4 Neg
He 0.00032
HYGAS
Selexol
Neg
Neg
0.0120
Neg
Neg
0.0120
„ j _^
Neg
Neg
0.0045
Koppers-Totzek
Benfield
Product GHS , ppuiv
Neg
Neg
0.0234
an Product Gas, ppmv -
Neg
Neg
0.0234
Neg
Neg
0.0045
HYGAS
Benfield
Neg
0.266
0.0322
Neg
Neg
0.0322
Neg
Neg
0.0045
*Neg * negligible
-------
methane and removal of CQ& condensation can
once again occur in the low-temperature acid-
gas removal processes. The concentration of
mercury in this vent-gas stream is predicted to
be about the same as for the previous COg-
removal step.
This product SNG will then be compressed in
the Lurgi and Koppers-Totzek processes to 70
atm.* Then, in all processes, the compressed
gases will be treated in glycol coolers at -274 K
to reduce the dew point of the gases. This
should reduce the amount of mercury in SNG
from the Koppers-Totzek and Hygas processes
to 0.0045 ppm (40 /tg/m3) or less.
Direct combustion of this gas in an industrial
process or use in the home should pose no health
hazards because of the mercury content. A
stack-gas concentration of about 4 /tg/m9 will
result from combustion of this gas compared to
the MATE value of 50 /ig/m8. In the home, the
primary nonvented appliance is the gas range.
The average annual cooking load is 10.2 million
Btu per customer, which results in an annual
mercury discharge into the home of 0.01 g.
Assuming the average home contains 425 m9 of
air with one-half air turnover daily, the average
concentration of mercury in the home would be
0.13 /ig/m8. Typical concentrations of mercury
inside residences is 0.07 fig/m9. The mercury
concentrations are 0.1 to 0.2 pg/m8 3 to 6 mo
after an interior repaint of a house.21 Thus, the
use of SNG from coal gasification plants should
not pose any health effects problems because of
mercury content. Additionally, some attenua-
tion of the mercury levels in the SNG because of
information of HgS in the pipeline is expected,
as well as dilution of the SNG by natural gas.
REVIEW BY ELEMENT
The theoretical analysis performed on the
disposition of arsenic, selenium, boron, lead, and
mercury indicates many areas where further re-
search efforts and environmental assessment
work could be most useful in designing coal
gasification facilities.
•Note that if the Koppers-Totzek gas had been com-
pressed prior to acid-gas removal, a different
H2S/C02-removal process might be economically
preferred, with different disposition of these trace
inorganics.
Arsenic
The projected mass flow rates of arsenic in
the various inlet and outlet streams of the gas-
ification processes are shown in Table 13 and
Figure 5. These projections, of course, depend
on the postulated occurrence of AsH3 in the raw
gasifier product gas. The presence of arsine and
its concentration should be investigated fur-
ther. The solubility of arsine, if present in acid-
gas removal processes, requires study. Finally,
if arsine is present, differing sulfur manage-
ment schemes in coal gasification processes
should be investigated. In the worst case, pro-
jected air emissions of arsenic from a full-scale
coal gasification plant, including boiler, are 6
kg/day compared to 37 kg/day for a coal-fired
power plant delivering the same energy.
Boron
The projected mass flow rate of boron in coal
gasification processes is shown in Table 14.
Boric acid, B(OH)3, is projected to be the major
route for removal from the feed coal. The boric
acid, if produced, will be recovered in the dis-
solved-solids recovery section of wastewater
treatment. No problems are anticipated because
of its presence.
Load
Volatile lead components should only exist in
raw product gases from high-temperature gas-
ification processes such as the Koppers-Totzek
process. This is shown in Table 15. These lead-
containing components, however, will precipi-
tate during quench of the raw gasifier product
gases and be recovered with other suspended
solids in the condensate.
Mercury
The projected mass flow rate of mercury in
the analyzed gasification processes is shown in
Table 16 and Figure 8. As can be seen, mercury
disposition depends upon the gas processing
scheme used. These projections, however, are
based on estimates of solubility and condensa-
tion, and the estimates need to be verified. In
this analysis, the projected emissions of mer-
cury from most of the various gas discharge
streams are below current MEG-MATE values.
Total mercury emissions from the process in-
323
-------
TABLE 13. PROJECTED ARSENIC DISPOSITION (14 ppm IN FEED COAL)
Process
Acid-Gas Removal
Input Stream
Coal Feed
Output Stream
Discharge Ash
Solids from
Wastewater Treatment
Product Sulfur
Sulfur Recovery Tail Gas
C02 Vent Gas (I)
Methanation Quench
C02 Vent Gas (II)
Glycol Cooler Recovery
Product SNG
* Neg - negligible
Lurgi
Rectisol
257.96
128.98
100.06
28.73
Neg*
< 0.275
Neg
Neg
Neg
Neg
HYGAS Koppers-Totzek
Selexol
kg/day
225.37
112.68
79.94
32.41
Neg
< 0.321
Neg
Neg
Neg
Neg
Benfield
257.96
Neg
257.96
Neg
Neg
Neg
Neg
Neg
Neg
Neg
HYGAS
Benfield
225.37
112.68
79.94
0.67
Neg
1.34
31.27
Neg
Neg
Neg
eluding the boiler house are estimated to be be-
tween 1.5 and 22 kg/day for a full-scale facility.
A coal-fired power plant producing the same
amount of energy would emit - 5 kg/day of mer-
cury using the same coal.
Satonium
The projected selenium disposition in typical
coal gasification processes is shown in Table 17
and Figure 7. The projected selenium disposi-
tion is controlled by the fate of HgSe in the gas-
processing section of the plants. If HgSe is pre-
sent as predicted, the major dispositions will be
with the discharge ash and either the product
elemental sulfur or product sulfuric acid from
which it can be removed. Formation of H^e in
gasification processes should be checked, as
well as its solubility in various processing liq-
uids. In these calculations, maximum gas-phase
324
-------
co
COAL ^
KT-258
CASIFIER
HY-225 1
?
ASH
LU-129
KT-NEG.
HY-113
(
H20
1 OIL
t i
H2
REMO
DISSOLVED
GASES
OIL
DEPRESSURIZATION
t
WASTEWATER
TREATMENT
I 1
t
BY PROD
OIL
1
S
VAL
H2S
CLAUS PLANT
UCT 1
SULFUR
LU-29
HY-32
1
TAIL GAS
TO C02 REMOVAL
SOLIDS
LU-100
KT-258
HY-80
FROM H2S
REMOVAL
C02(D
REMOVAL
T
co2(i)
LU-<0.3
HY <0.3
5.
COMBINED
AND
C02(ID
REMOVAL
T
C02(II)
GLYCOL
COOLER
7
H2°
disposition (kg/day) in 250 X 106 SCF/day coal gasification plants.
(14 ppm in coal)
SNG
-------
TABLE 14. PROJECTED BORON DISPOSITION (10.2 ppm IN COAL)
Process
Input Streams
Coal Feed
Output Streams
Discharge Ash
Solids from Wastewater
Treatment
Product Sulfur
Sulfur Recovery
Tail Gas
CO- Vent Gas (I)
Methanation Quench
C02 Vent Gas (II)
Glycol Cooler Recovery
Product SNG
* Neg = negligible
Lurgi
188.3
94.1
94.1
Neg*
Neg
Neg
Neg
Neg
Neg
Neg
HYGAS
/,
kg/day
164.5
82.2
82.2
Neg
Neg
Neg
Neg
Neg
Neg
Neg
Koppers-Totzek
188.3
Neg
188.3
Neg
Neg
Neg
Neg
Neg
Neg
Neg
emissions of selenium in a full-scale process are
estimated at 6.6 kg/day. These emissions are
primarily due to the boiler house. A coal-fired
power plant delivering the same energy is esti-
mated to emit 25 kg/day of selenium into the at-
mosphere.
It is emphasized that the analysis presented
is based primarily on theoretical projections
and engineering assumptions. This analysis
should provide insight into a better understand-
ing of the factors important in determining the
formation and disposition of some of these con-
stituents. Further experimental investigations
are desirable to increase this understanding.
ACKNOWLEDGMENTS
The financial support of the U.S. Environmen-
tal Protection Agency (Contract No. 68-02-2648)
in performing this work is greatfully acknow-
ledged.
326
-------
TABLE 15. PROJECTED LEAD DISPOSITION (35 ppm IN COAL)
Process
Input Streams
Coal Feed
Output Streams
Ash Discharge
Solids from Wastewater
Treatment
Product Sulfur
Sulfur Recovery Tail Gas
C02 Vent Gas (I)
Methanation Quench
C02 Vent Gas (II)
Glycol Cooler Recovery
Product SNG
* Neg - negligible
Lurgi
641.2
641.2
0.005
Neg*
Neg
Neg
Neg
Neg
Neg
Neg
HYGAS
kg /day
560.0
560.0
0.008
Neg
Neg
Neg
Neg
Neg
Neg
Neg
Koppers-To tzek
641.2
Neg
641.2
Neg
Neg
Neg
Neg
Neg
Neg
Neg
327
-------
TABLE 16. PROJECTED MERCURY DISPOSITION (0.2 ppm IN COAL)
Process
Ac id- Gas Removal
Input Streams
Coal Feed
Output Streams
Discharge Ash
Solids from Wastewater
Treatment
Product Sulfur
Sulfur Recovery
Tail Gas
C02 Vent Gas (I)
Methanation Quench
C02 Vent Gas (II)
Glycol Cooler Recovery
Product SNG
* Neg - negligible
Lurgi
Rectisol
3.672
Neg*
Neg
3.635
Neg
0.0092
Neg
0.0173
Neg
0.0097
HYGAS
Selexol
. trn /i
Kg/0
3.213
Neg
Neg
1.854
Neg
0.3397
Neg
0.543
0.202
0.275
HYGAS
Benfield
| ,, ,,
iciy
3.213
Neg
Neg
0.257
Neg
0.514
Neg
0.424
1.743
0.275
Koppers-Totzek
Benfield
3.672
Neg
Neg
0.0007
Neg
0.0015
Neg
0.0015
3.394
0.275
328
-------
COAL
LU-3.67
KT-3.67
HY-3.21
GASIFIER
1
ASH
H20
1
1 OIL
DISSOLVED
GASES
OIL
DEPRESSURIZATION
WASTEWATER
TREATMENT
i f
SOLIDS H.O
t
BY PRODUCT
OIL
H2S
^ REMOVAL
H2S
CLAUS PLANT
1 1
SULFUR TAIL GAS
LU-3.64
KT-0.0007
HY-1.85
TO CO REMOVAL
FROM H2S
REMOVAL
C02(I)
REMOVAL
I
co2(i)
LU-0.0092
KT-0.0015
HY-0.3397
COMBINED
AND
C02(II)
REMOVAL
C02(II)
LU-0.0173
KT-0.0015
HY-0.543
CLYCOL
COOLER
1
-^- SNC
LU-0.0097
KT-0.275
HY-0.275
LU-NEG.
KT-3.394
HY-0.202
Figure 6. Projected mercury disposition (kg/day) in 250 x 106 SCF/day coal gasification plants.
(0.2 ppm in coal)
-------
TABLE 17. PROJECTED SELENIUM DISPOSITION (2.08 ppm IN COAL).
Process
Input Streams
Coal Feed
Output Streams
Discharge Ash
Solids From WaStewater
Treatment
Product Sulfur
Sulfur Recovery
Tail Gas
C02 Vent Gas (I)
Methanation Quench
C02 Vent Gas (II)
Glycol Cooler Recovery
Product SNG
Lurgi
38.3
19.1
Neg*
19.1
Neg
< 0.2
Neg
Neg
Neg
Neg
HYGAS
kg /day
33.5
16.7
Neg
16.7
Neg
< 0.1
Neg
Neg
Neg
Neg
Koppers-Totzek
38.3
Neg
Neg
38.0
Neg
< 0.4
Neg
Neg
Neg
Neg
* Neg = negligible
330
-------
T.iM*,T .„ <
KT-38. 3
f H2°
ASH
LU-19.1
KT-NEG.
HY-16.7 1
in¥I DISSOLVED
01 L CASES
OIL
DEPRESSURIZATION
t
i BY PRODUCT
ATI
WASTEWATER
TREATMENT
* 1
H2S
REMOVAL
i
CLAUS PLANT
1 i
SULFUR TAIL CAS
LU-19.1
KT-38. 0
HY-16.7
TO C02 REMOVAL
SOLIDS
FROM H2S
REMOVAL
C02(D
REMOVAL
*
COMBINED SHIFT
AND METHANATION
*
QUENCH
*
C02(II)
REMOVAL
GLYCOL
COOLER
C02(I)
LU-<0.2
KT-<0.4
HY-<0.1
co2(ii)
SNG
Figure 7. Projected selenium disposition (kg/day) in 250 X 106 SCF/day coal gasification plants.
(2.08 ppm in coal)
-------
REFERENCES
1. Institute of Gas Technology. Initial Envi-
ronmental Test Plan for Source Assess-
ment of Coal Gasification. EPA Contract
Number 68-02-1307, Report No. EPA
600/2-76-259.
2. Bodle, W. W., and Vyas, D. C. Clean Fuels
from Coal Symposium II Papers. Institute
of Gas Technology. June 23-27, 1975. p.
30-31.
3. Woodall Duckman, Ltd. Trials of American
Coals in a Lurgi Gasifier at Westfield
Scotland. U.S. Department of Interior.
Report RI15. November 1974.
4. Institute of Gas Technology. Pipeline Gas
From CoalrHydrogenation. U.S. Depart-
ment of Energy. Report PE-2134. February
1978.
5. Duck, N. W., and Himus, G. W. On Arsenic
in Coal and Its Mode of Occurrence. Fuel
50:267-271.1951.
6. Zhuchkov, I. A., et al. Behavior of
Arsenopyrite During Roasting in an Inert
Gas Atmosphere. Zh Prikl Khim.
40(12):2703-2710.1967.
7. Zviadadze, G. N., and Rtskhiladze, V. G.
Kinetics and Mechanism of the Thermal
Decomposition of Arsenopyrite FeAsS.
Soobschch Akad Mauk Gruz SSR.
52U):125-132.1963.
8. Lukesh, J. 8. Thermal Decomposition of
Arsenopyrite. Am Mineral. 25:539-542.
1940.
9. Barton, P. B., Jr. Thermochemical Study of
the System Fe-As-S. Geochim Cosmochim
Acta. 53(7)^41-951.1969.
10. Institute of Gas Technology. Environmen-
tal Assessment of the Hygas Process.
Chicago, 111. U.S. Department of Energy.
Contract Number EX-76-C-2433, Report
No. FE-2433-16. February 1978.
Fray, D. J. Inst Min Metatt Trans, Sect C.
83:194. 1974.
Zaburdyeav, V. S., et al. Zh Prikl Khim
(Leningrad). 47(7):1459-1463.1974.
11.
12.
13. Kedyarkin, V. M., and Zorin, A. D. Tr Khim
Khim Tekhnol 3):161-164.1965.
14. Dobson, D. C., et al. Photolysis of Hydrogen
Selenide. J Phys Chem. 79(8).
15. Laidler, J. J. Theory of Chemical Reaction
Rates. New York, McGraw-Hill, 1969. p. 36
ff.
16. Inorganic Constituents of Australian Coals,
Part I. Nature and Mode of Occurrence. J
Inst Fuel 57:422-434.1964.
17. JANAF Thermochemical Tables. The Dow
Chemical Co. Midland, Michigan.
18. Weiss, H. G., and Shapiro, L. Mechanism of
the Hydrolysis of Diborane in the Vapor
Phase. / Amer Chem Soc. 75:1211-1224.
1953.
19. Luthy, R. G., and Tallon, J. T. Experimen-
tal Analysis of Biological Oxidation Char-
acteristics of Hygas Coal Gasification
Wastewater. Carnegie-Mellon University.
Pittsburgh, Pa. NTIS Report Fe-2497-27.
1978.
20. Dubeau, C., et al. Solubility of Hydrogen
Selenide Gas in Water. J Chem Eng Data.
J6(l):78-79.1971.
21. Foote, R. S. Mercury Vapor Concentrations
Inside Buildings. Science. 277:513-514.
August 11,1972.
332
-------
Session II: ENVIRONMENTAL ASSESSMENT:
LIQUEFACTION
J. Wayne Morris, Chairman
Hittman Associates, Inc.
Columbia, Maryland
333
-------
INITIAL SAMPLING OF THE FORT LEWIS SRC PILOT PLANT
David D. Woodbridge
Hittman Associates, Inc., Columbia, Maryland
Abstract
During the first phases of the measurement
program at the solvent refined coal (SRQ pilot
plant at Fort Lewis, Washington, emphasis was
primarily directed at determining the quality of
the pollutant streams entering the environment
The first measurements were directed at obtain-
ing information relative to the operation of the
wastewater treatment facility. Because the pilot
plant is not a miniature version of a commercial
facility, it was also necessary to obtain samples
from streams feeding the wastewater treatment
facility. Liquid, gaseous, and solid streams that
could affect the environment were sampled and
analyzed according to the U.S. Environmental
Protection Agency's (EPA) Level 1 and Level 2
procedures.
Coal conversion processes are highly complex
systems consisting of a wide variety of inter-
related components. Level 1 sampling has
shown that certain streams have no significant
environmental impact These data result in
fewer streams requiring sampling for the Level
2 analysis. A detailed evaluation of the data
resulting from the Level 2 sampling and analy-
sis will indicate the streams and/or specific
pollutants that require the attention of a Level 3
methodology.
INTRODUCTION
A number of different processes are under
development for producing synthetic fuels from
coal. One of these technologies is the solvent
refined coal (SRC) system. The process was
originally developed by Spencer Chemical Com-
pany for the U. S. Department of Interior, Office
of Coal Research. Gulf Oil subsequently ac-
quired Spencer Chemical Company and is conti-
nuing development under the Pittsburg and
Midway Coal Mining Company.1 A pilot plant
was constructed at Fort Lewis, Washington,
which has the capability of converting 45 metric
tons of coal per day to the SRC products.
The SRC pilot plant at Fort Lewis, Washing-
ton has the capability to operate either in the
SRC-I or the SRC-II mode. In the SRC-I config-
uration the facility produces a solid product
with a sulfur content of less than 1 percent and
an ash content of 0.2 percent or less.2 When the
facility is in the SRC-II operating configuration,
it produces a liquid product.
Solid, liquid, and gaseous products and by-
products enter the environment as a result of
the liquefaction process. The initial sampling of
the various streams at the SRC pilot plant at
Fort Lewis was designed to obtain preliminary
environmental assessment data, identify the
potential problem areas, and establish priorities
for further considering the streams. The direct
streams from a pilot-plant facility to the envi-
ronment are not the same as those from a com-
mercial operation. To obtain information
related to some of the pollution problems that
may be associated with a commercial operation,
it was necessary to sample the products and all
streams that fed the waste treatment compo-
nents.
Wherever possible, sampling and analysis
procedures were followed in accordance with
the U.S. Environmental Protection Agency's
(EPA) Level 1 and Level 2 procedures.345
Samples were obtained for physical, chemical,
and biological testing. In accordance with the
philosophy of the phased approach, all streams
were surveyed using sampling and analytical
methods that permit priority ranking of the
streams relative to their containment of poten-
tially toxic materials.
Two field sampling and measurement opera-
tions have been performed to obtain samples
from the SRC pilot plant at Fort Lewis. The
first field operation occurred in March 1978 and
the second in February 1979. During both of
these field sample-gathering operations, the
pilot plant at Fort Lewis was in the SRC-II
mode. In March of 1978 the products were
sampled, and samples and data were obtained
from various locations throughout the waste-
water treatment plant. During the February
1979 sampling operations samples and data
335
-------
were obtained for the products, the source coal,
liquid streams, gaseous emissions, and the sur-
rounding atmosphere. Preliminary analysis and
evaluation of these data are presented in this
paper.
SRC PILOT PLANT AND OPERATION
The SRC pilot plant at Fort Lewis, Washing-
ton, was constructed on the military base near
Tacoma, Washington, which began operating in
October of 1974. Figure 1 shows a block diagram
of the SRC system.* In the coal preparation and
handling area raw coal is unloaded, crushed, and
stored in piles. The coal is sized, pulverized, and
mixed with a recycled hydrocarbon solvent. The
resulting coal/solvent slurry is mixed with a
hydrogen-rich gas and preheated. The pre-
heated mixture enters the hydrogenation zone,
which operates at 425° C to 480° C and 6.9 MPa
to 13.8 MPa, with about a 30-min holding time.
The coal is liquefied by reacting with hydrogen.
The liquefied product contains some undis-
solved material, primarily mineral matter and
undissolved coal. The excess hydrogen and
gases produced in the reaction are separated
from the slurry of undissolved solids and coal
solution. The gaseous stream passes through a
cleanup system to remove H2S and C02 and is
then recycled to the reaction zone. Fresh
hydrogen from the hydrogen production area is
added to this recycled gas stream. The slurry of
solids and coal solution is cooled; the solids are
separated from the coal solution, stored, and
used for hydrogen generation. The coal solution
is further separated into a light oil fraction, a
wash solvent fraction, the process solvent, and
the solvent-refined coal. The SRC is solidified
by cooling. The gasification system will gasify
either the residue or a mixture of residue and
coal.7
A diagram of the plant system is shown in
Figure 2. Each area in Figure 2 is numerically
designated as follows:
• 01 Coal preparation and slurry missing,
• 02 Preheating, dissolving, and pressure
letdown,
• 03 Mineral separation: very little use for
SRC-II,
• 04 Fractionation and solvent recovery,
• 05 Gas recovery and recompression,
• 081 Sandvik belt-vacuum bottoms solidifica-
tion,
• 082 Solid product storage area, and
• 091 Wastewater treatment area.
Figure 2 also shows the locations from which
solid, liquid, gaseous, and atmospheric samples
were obtained. These will be discussed later in
the paper. An aerial view of the SRC plant at
Fort Lewis, Washington, is shown in Figure 3.
Operational constraints of a pilot plant limit
the time when samples and data can be ob-
tained. A number of parameters within the pilot
plant can be changed, which may result in
changes in the constituents of various streams
within the system. Operational temperature
and pressure and the rate and amount of hydro-
gen introduced into the process are often
changed. The type of coal is also changed. Even
if gross parameters of the coal are similar, the
presence or absence of trace elements may vary
the constituents of potential pollutant streams.
Sudden shutdown or recycling procedures will
change the equilibrium of the system. A change
in the mode of operation of the plant from the
SRC-I to the SRC-II or order will change the
constituents in the various streams. Consider-
able differences in the temperature and pres-
sure of the process as well as the amount of
hydrogen injected into the system exist be-
tween the two modes of operation. Comparable
data can be obtained only if operational param-
eters are similar and the operation is stabilized.
During both March 1978 and February 1979
sampling operations, the SRC plant at Fort
Lewis was in the SRC-II mode. Because the
plant was operating in the SRC-II mode, data
resulting from the sampling operation may in-
dicate some of the conditions that can be ex-
pected in demonstration or commercial facility.
Typical Pacific Northwest winter conditions
prevailed during the February 1979 field opera-
tion. The temperature ranged from -0.6°C
(31° F) to 5.6° C (42° F) with some rainfall near-
ly every day. Conditions included complete
cloud cover approximately 86 percent of the
time. On the night of February 11, 1979, a
severe windstorm damaged power lines and
shut down the plant for nearly 2 days.
SAMPLING RATIONALE
The phased approach, developed by the Proc-
ess Measurements Branch CPMB) of EPA re-
quires three separate levels of sampling and
analytical effort. The first level. Level 1, utilizes
336
-------
COAL
RECYCLED
SOLVENT
COAL PREPARATION
AND HANDLING
RECYCLED
SLURRY
SRC
(HYDROGENAT IOH)
SOL I OS
SEPARATION
FILfRATE
WASH
SOLVENT
SOLVENT
RECOVERY
KEY
SRC II
RECYCLED
HYDROGEN
GAS CLEANUP
AND BY-PRODUCT
RECOVERY
Figure 1. Solvent refined coal system.1
HYDROCARGO;
AND WATER
SULFUR
MINERAL
RESIDUE
SRC PRODUCT
. LIGHT
•*" LIQUIDS
337
-------
oo
\
09.2 O
O O O
OQ°
BOUNOARV EXCESS STORM
OUTFALL DRAINAGE UNIT
GOING TO WD
SYSTEM
AT JUMfOVER BOX
OR AIMS TO SURGE
RESERVOIR
1
L
LEGEND:
STORM SEWER
WASTE DRAIN
HI-VOL LOCATIONS
LIQUID SAMPLING LOCATIONS
GAS SAMPLING LOCATIONS
Figure 2. SRC plant and sampling locations.
-------
Figure 3. Aerial view of the SRC plant at Fort Lewis, Washington.
339
-------
quantitative sampling and analysis procedures
that yield final analytical results accurate to
within a factor of 3 of the sample.3 Level 1 is de-
signed to:
• Provide preliminary environmental assess-
ment data,
• Identify problem areas, and
• Provide data to order priorities for the var-
ious streams and/or components.
Level 2 sampling and analysis procedures are
designed to:
• Confirm and expand the Level 1 results, and
• Determine exact quantities of organic or in-
organic constituents that could provide a
health or ecological problem.
Level 3 sampling and analysis are directed at
monitoring the problems identified in Level 1 to
provide information for control device design
and development.3
The basic rationales of the Level 1 and Level
2 sampling and analysis procedures were fol-
lowed in planning Phase 1 (March 1978) and
Phase II (February 1979) sample acquisition and
analysis tasks. Phase I sampling was designed
to provide preliminary environmental assess-
ment data on the wastewater treatment facility
of the SRC plant and of the SRC-II products.
The Phase II sampling and analysis were de-
signed to confirm the results obtained in Phase
I on the wastewater treatment facility and the
SRC-II products and to perform complete Level
1 sampling and analysis on:
• All streams flowing into the wastewater
treatment facility,
• All emissions to the atmosphere, and
• The atmosphere surrounding the SRC plant.
Level 2 sampling and analysis were planned for
all liquid streams leading to and through the
wastewater treatment facility.
Figure 4 is a general diagram of streams that
were scheduled to be sampled. Sampling point
locations with respect to plant operations are
shown in Figure 2, where: .
• Numbered locations represent the liquid
sampling locations,
• Lettered locations represent the gaseous
sample locations, and
• (X) represents location of atmospheric sam-
ples.
The following factors limited the extent to
which the Level 1 and Level 2 procedures could
be followed:
• All electrical equipment had to be equipped
with explosion-proof motors and connections
to operate inside the plant area.
• No holes could be made in any pipe of stack.
• Safety regulations prevented obtaining flare
exhaust samples.
• High-volume samplers did not exist in the
area that were equipped with sorbent mod-
ules.
• Certain components of the plant failed to
operate normally.
These limitations mainly restricted the acquisi-
tion of gaseous samples. The source assessment
sampling system (SASS) train was not allowed
to operate on the premises. In most locations it
would have been impossible to operate the
SASS train because of the design of the instru-
ment.
SAMPLING OPERATION
In March of 1978, samples were acquired from
locations throughout the wastewater treatment
system at the SRC pilot plant at Fort Lewis. A
block diagram of the wastewater treatment sys-
tem showing locations of sample acquisition is
shown in Figure 5. All samples were 1 gal
(3.79 L) grab samples taken during the same day
of plant operation. Liquid samples were pre-
served with acid and stored in ice during hold-
ing and shipping to the Hittman Associates lab-
oratory for analysis. Solid samples were placed
in polyethylene bags for shipment.
A more complex field sampling operation was
required for Phase II in February 1979. The
operation was designed to acquire:
• Level 1 and Level 2 liquid chemical samples,
• Samples for bioassay analyses,
• Level 1 gas samples,
• Surrounding atmospheric air samples,
• Coal and product samples, and
• Samples throughout 5 days of plant opera-
tion.
A schedule of the samples acquired from the
Fort Lewis SRC pilot plant is shown in Table 1.
Table 1 also shows location, method, and reason
for sample acquisition. This table does not in-
clude the gaseous or atmospheric samples or on-
site analysis. Liquid samples of the inflow and
effluent of the wastewater treatment facility
were also sent to Gulf South Research Institute.
Liquid samples were collected in 5 gal
340
-------
VENT
SLURRY
B-LJND FLARE EXHAUST
rnicr 1
DUST
COAL V^AL
AND P
r.i
TT ,,. . , _ | /« L.^^I>llk<
SEPARATION, I FUEL PREPARATK
, DRYFP
AGENT 1
DP.AIM
S T *—
OXIDIZER
| TANK
LEACHATE RPWT VACUUM
1 BOTTOM
DRAINS S/
B
W
ON
)L
I
DN
c
h-
\NDV
:LT
ET
FROM '„ SURGE
o:
-^ DUST
PREHEATER
STACK
GAS PRODUCT
STORAGE
1
FLARE
KNOCK
SATER LEACHATE
CK
- V/ASTE
. UATCB __EFFLUEr
Figure 4. Pollutant streams at SRC system. Fort Lewis, Washington.
-------
INLET
WATER
SURGE
RESERVOIR
20^4
CLARIFIER A
FLOTATION
""SKIMMING
CLARIFIER
SETTLED
HYDROCARBONS
303
DISSOLVED
AIR
FLOTATION
UNIT
PRODUCT SOLIDIFICATION
WATER
SAND
FILTER
SAND
FILTER
CHARCOAL
FILTER
207
CLARIFIER
SEDIMENT
NH ADDITION
HYDROCARBON
STEAM
ADDITIOM
206
BIOLOGICAL
UNIT
205
HOLDING
TANK
305
DIGESTED
BACTERIA
FILTER
BACK-
WASH
TANK
208
-» DISCHARGE
NOTE: 200a - liquid samples
300*8 - solid samples
Figure 5. Overall flow schematic of the SRC pilot plant wastewater system.
342
-------
TABLE 1. SAMPLE SHIPMENTS TO INDIVIDUAL LABORATORIES FOR ENVIRONMENTAL
SOURCE TESTS FOR THE SRC SAMPLES COLLECTED FEBRUARY 11
THROUGH FEBRUARY 17 AT FORT LEWIS, WASHINGTON
Sam-
pling
Meth-
od
1st Day (2/11/79) 2nd Day (2/12/79)
*"*
4th Day
(2/16/79)
5th Day
(2/17/79)
HAI TRW COM AM EPA HAI LIT COM UC HAI TRW AM HAI UW HAI TRW AM
Sample
Identification
pnos^or- o
M'5i«* *
L 'fill i
ll
S SH o § ^SHOOS^O Q8 3 S= 8
S. B. Jm 2. £. m
af f
-2 •»
1 CPA
Coal Preparation
Area
2 CSPL
Coal Storage Pile
Leachate
3 DSAD
Dissolve & Separation
Area Drain
4SRAD
Sulfur Recovery Area
Drain
6 RPWT
Recycle Process Water
Tank
7 BB
Boiler Slowdown
8 CW
Cooling Water
9 WWTPI
Wastewater Treat.
Plant Inf.
10 WWTPE
Wastewater Treat,
Plant Eff.
11 VBSD
Vacuum Bottoms
Storage Drain
12 SBCW
Sandvik Belt Cooling
Water
13 SFAD
Solvent Fractionation
Area Drain
Naphtha
Middle Distillate
Heavy Distillate
Recycle Slurry
Pulverized & Dried <
Coal
Vacuum Bottoms '
Raw Coal '
1st QC Sample
2nd QC Sample
• • ••
• ••• ••• ••
1 • ••
^Includes 3 separate bottles for oil and grease, TOC, COD, Phenolics, Alkalinity, Acidity, TDS, TSS and Hardness.
..Also shipped a 1 gallon unextracted RPWT to TRW.
Number five was to be the sample from the flare knockout drum, which was dry.
343
-------
(18.93 L) bottles and then split for the various
analyses. Ten L of the samples were extracted
with methylene chloride for organic analysis.
Extractions from each of the liquid streams
were then shipped by air to the appropriate
analytical laboratory. All liquid samples that
were not extracted or sent for trace metal
analysis were iced to keep below 4° C. Because
atmospheric temperature was generally below
40° F (4.4° C), no difficulty was encountered in
sample preservation. Samples were also pre-
served as shown in Table 2.
Samples of gaseous emissions and atmospher-
ic particulates were obtained from locations
shown in Figure 2. Locations indicated in Figure
2 as positions from which gaseous samples were
obtained are designated in Table 3, relative to
the source of the emissions. A sample of each
source was placed in a 10-L, chemically inert
mylar container and taken immediately to the
laboratory for analysis. With explosion-proof
pumps, 50 to 200 L of gas from each source were
also passed through impingers. The SASS train
was not used because of the stipulated use of
explosion-proof equipment within the plant's
operating area and the denial of the request for
entrance ports to the vent stacks.
Product and solid samples were also acquired
for chemical analysis and bioassay. These
samples, the method of acquisition, and their
disposition are shown in Table 1.
One severe problem was encountered during
the second field trip relative to the operation of
the wastewater treatment facility. A plug de-
veloped in the line between the aeration tank
and the clarifier, which produced a malfunction
of the aeration system. Because the wastewater
treatment facility was operating beyond the de-
signed capacity, the malfunction reduced the fa-
cility's efficiency.
PRELIMINARY SAMPLE ANALYSES
Analysis and preliminary evaluation of data
obtained from the samples gathered from the
various locations through the wastewater treat-
ment facility indicate that the system was per-
forming adequately when all aspects of the SRC
plant and the wastewater treatment facility
were operating normally. Table 4 shows the re-
sults of a spark source spectrometer analysis
for trace elements of the wastewater treatment
facility effluent. As a comparison, the Washing-
ton State limiting concentrations are also
shown. Concentrations of all regulated trace
elements were reduced to levels below those re-
quired by Washington State.
Figure 6 shows the distribution of organic
constituents as a function of location through
the wastewater treatment facility. Table 5
shows the percent of total reduction of the
organics.
Data from the laboratory of the SBC pilot
plant at Fort Lewis indicate that the waste-
water treatment facility obtains removals as
shown in Table 6. These values, however, may
be atypical of results that would be observed in
a commercial system for the following reasons:
TABLE 2. SAMPLE PRESERVATION
Samples for Analysis of:
Oil to Grease
TOG
COD
Phenolics
Method of Preservation
to pH of 2 and coal to 4°C
H.PO to pH of 41.0 g CuSO /I and
coal to 4°C
Trace Elements
Organics
Volatiles
HN03 to pH of 2
CH-Cl,, extraction
Coal to 4°C
344
-------
TABLE 3. GASEOUS SAMPLING LOCATIONS
Location
Designation
A
B
C
D
E
F
G
H
I
Gaseous Source
Slurry blend tank vent
Pre-heater stack gas
N? stripper vent
Oxidizer tank vent
Input to flare
Hot well tank vent
Process solvent accumulator vent
Sandvik belt vent
Process liquor tank vent
TABLE 4. COMPARISON OF TRACE ELEMENT DATA ON TREATED WASTEWATER
WITH REGULATED LIMITATIONS OF WASHINGTON STATE
Element
Sodium (ppm)
Potassium
Nickel
Iron
Zinc
Bromine (ppb)
Selenium
Chromium
Thorium
Rubidium
Antimony
Arsenic
Cone. Observed by
Washington State
(mg/1)
25-50
2-10
0.05-0.10
0.1-0.5
0.5
40-60
0.2-2
0.0-4
0.04
0.8-5.0
0.5-3.0
2-4
Spark Source
Analysis
(mg/1)
5.9
1.4
0.04
0.31
0.4
13
B
1
B
3
B
1
B = Below detection limit.
345
-------
10,000 [-
1
FEED TO
CLARIFIER
FLOW FROM
FLOTTAZUR
BIO-UNIT
EFFLUENT
CAF'.RON
FILTER
EFFLUENT
BACKV/ASH
FILTER
DISCHARGE
Figure 6. Effects of the waste water treatment process on organics.
346
-------
TABLE 5. PERCENT REDUCTION IN ORGANICS BY THE
SRC WASTEWATER TREATMENT FACILITY
Oil and grease
COD
TOC
CR-C1, hydrocarbons
99%
89%
98%
99%
The process water is only about 1 percent of
the total feed to the wastewater treatment
system. The actual COD of the foul process
water has been reported to range from
25,000 to 43,600.'
Phenols have not been recovered from the
wastewater at the SRC-II pilot plant, as
would be the case in a commercial system.
Results of the Level 1 organic analysis in-
dicate that phenols represent about 30 per-
cent of the total organics. Phenols are readi-
ly biodegradable at concentrations from 500
to 1,000 mg/L,'10 u and the high concentra-
tions of phenols relative to the more refrac-
tory classes of organic compounds detected
in the Level 1 analysis explain the high de-
gree of biodegradability.
Although the level of organics was too low fol-
lowing biological treatment to require a Level 1
analysis, results of the infrared analysis indi-
cate that the following classes of hydrocarbons
were still present:
• Aromatics, including substituted benzenes,
naphthalenes, and other polynuclear hydro-
carbons;
• Compound classes with C-0 and C-0
stretches representing aldehydes, acids, and
esters;
TABLE 6. RANGES OF WASTEWATER PARAMETERS AT
THE FORT LEWIS PILOT PLANT8
PH
BOD, mg/1
COD, mg/1
TSS, mg/1
Phenol, mg/1
Extrac table
oil, mg/1
Surge
Reservoir
5.0-9.0
-
1,000-9,600
90-400
30-1,500
10-250
Clarifier
Effluent
6.2-6.8
-
650-5,000
50-300
25-1,100
6-150
Flottazur
Flotation
Unit
6.2-6.8
135-350
500-4,000
30-200
10-1,000
4-30
Bio-Unit
Effluent
6.2-7.4
10-110
20-250
20-300
0.1-1.0
0-4
Plant
Effluent
6.2-7.4
4-23
5-75
0-20
0.0-0.4
0-3
347
-------
500
1*00
300
o
<
ce.
S 200
100
(mg/1)
\
\
\
\
\
\
\
\
\
\
v / DISSOLVED SOLIDS
V
SUSPENDED SOLIDS
I
FEED TO
CLARIFIER
EFFLUENT
FROM
FLOTTAZUR
BIOUNIT
EFFLUENT
CARBON
FILTER
EFFLUENT
DISCHARGE
Figure 7. Distribution of dissolved and suspended
solids hi the wastewater treatment system.
348
-------
TABLE 7. SUMMARY OF THE SAM/IA MODEL FOR THE EQUALIZED FEED TO THE
WASTEWATER TREATMENT SYSTEM AND THE CARBON FILTER EFFLUENT
Effluent Stream Potential Degree of
Hazard
• Health-based MATE
• No. of MATEs exceeded
• Ecological-based MATE
• No. of MATEs exceeded
Feed to
Clarifier
6.49
3/31 (10%)
15,224
7/18 (39%)
Carbon Filter
Effluent
2.23
0.24 (0%)
1,528
5/15 (33%)
Potential Toxic Unit Discharge
Rate (I/sec)
• Health-based MATE
• Ecological-based MATE
32.77
7.67 x 10
11.25
7710.86
• Aliphatic hydrocarbons of alipathic substitu-
tion on ring compounds;
• Compounds with C-N stretch including
amines; and
• Phenols.
The results of the analysis of the wastewater
for suspended and dissolved solids are illus-
trated in Figure 7. While the results indicate 98
percent suspended solids removal in the waste-
water treatment system, the results for dis-
solved solids do not show a consistent trend.
Net reduction in the treatment system was
found to be approximately 14 percent. The sus-
pended solids results agree with available plant
data, which indicate that suspended solids
levels average 15 mg/L in the biounit effluent
and 5.5 mg/L in the carbon filter effluent.
Overflow from the backwash filter was found to
contain 10 mg/L of suspended solids, well within
the Washington State effluent limitations of 50
mg/L.
The SAM/IA model was applied to the feed to
the clarifier and the carbon filter effluent, which
represent the equalized feed to the treatment
system and the treated wastewater, respective-
ly. The results of the SAM/IA application yield
an "effluent stream potential degree of hazard"
based on comparison of the stream components
to the ecologically and health-based MATEs,
and a "potential toxic unit discharge rate" based
upon the flow rate of the stream, thereby allow-
ing the relative hazard of various streams to be
compared on a flow-rate basis. Table 7 summa-
rizes the results of the SAM/IA.
During the second phase of the field sampling
program a mobile laboratory was established in
a covered truck at the SRC plant. This labor-
atory was established to obtain on-the-spot
measurements of pH, conductivity, ammonium,
nitrate, chloride, sulfide, and cyanide. All of
these immediate onsite measurements were
made by ion probes. Table 8 shows the results of
these measurements. A great deal of variability
is evident in the data. Extreme care was taken,
and multiple measurement acquired, in the at-
tempt to obtain readings as accurate as possi-
ble. However, interference often made the de-
gree of accuracy less than desired. Particular at-
tention should be given to the recycle process
water tank because of its high concentrations of
cyanide, chlorides, sulfides, and ammonium.
Analysis of the samples from the SRC pilot
plant for trace elements is being performed by
both spark source spectroscopy and plasma jet
spectroscopy. The plasma jet has the advantage
of excellent quantified results but is limited to
only those elements for which the computer has
been programmed. At the present time, only
metals are spectrographically determined and
quantified for the computer. Table 9 shows the
concentration of each of the metals for each of
the sampling locations.
349
-------
TABLE 8. METALS IN SRC WASTE STREAMS (PLASMA JET SPECTROGRAPHIC ANALYSIS)
Metals
#1
CPAD
#2
CSPL
#3
DSAD
#7
BB
#8
CW
#9
WWTPI
#10
WWTPE
#11
VBSD
#12
SBCWD
#13
SFAD
Dectection
Limit
Aluminum
Barium
Boron
Calcium
Copper
Iron
g Magnesium
Manganese
Phosphorus
Potassium
Silicon
Sodium
Strontium
Titanium
Zinc
L = Below
0.47
0.029
0.043
12.8
0.016
1.45
3.66
0.020
16.8
3.62
27.6
78.3
0.063
0.011
0.77
detect:
99.1
0.066
1.15
245
1.40
1850
30.4
6.75
22.0
3.44
30.1
140
2.63
0.11
8.75
Lon limiJ
0.25
0.080
0.078
4.07
0.029
0.35
0.29
0.014
1.34
0.40
4.82
11.8
0.016
0.007
0.32
L
0.010
0.094
0.37
L
0.52
0.057
0.007
41.1
1.98
18.8
170
0.005
L
L
L
0.014
0.049
30.9
L
2.18
10.4
0.041
6.66
8.70
30.9
16.7
0.17
L
3.11
2.11
0.019
1.02
15.3
0.053
4.90
4.76
0.031
5.16
30.0
17.3
71.7
0.11
0.017
0.38
L
0.028
0.12
17.3
L
0.079
4.26
0.007
L
1.37
12.4
19.1
-
0.014
L
212
0.050
1.21
235
2. .40
2700
156
11.7
254
2.51
45.0
113
1.65
0.080
5.10
L
0.085
0.051
11.9
0.023
0.23
3.76
0.005
L
1.02
11.7
5.90
0.058
L
0.041
0.23
0.28
0.044
3.43
0.039
0.52
0.70
0.014
3.90
0.59
8.12
20.0
0.014
0.008
0.18
0.15
0.001
0.01
0.01
0.015
0.030
0.001
0.003
0.4
0.01
0.08
0.2
0.001
0.006
0.015
-------
TABLE 9. PARTICULATE MEASURED BY HIGH VOLUMES
AT THE SRC PILOT PLANT
Location
Indicator Location
A Ground level 04 area
B E side - 5* north of guard shack
C W side - 300' north of generator
D 200' SW of flare tower
E 75' SW of 091 shack
F Outside fence - 091 area
G Outside fence - south 082 area
H Outside fence - south 01 area
2-12-79
(stormy)
Ifg/m3
116.0
21.2
30.0
18.0
9.1
21.5
23.4
20.2
2-15-79
54.3
3,1.6
47.8
19.8
23.3
26.4
22.7
26.0
Location of the eight high-volume air sam-
plers is shown in Figure 8. Numbers next to the
location numbers are the measured concentra-
tion of particles in micrograms per cubic meters
of air passing through the instrument. Values
are the mean of two measurements. The first
measurements were obtained February 12,
1979. That evening a severe windstorm dam-
aged power lines and shut down the plant. This
storm resulted in the shutdown of the high-
volume samplers after approximately 18 hr of
operation. A 24-hr operation of the high-volume
samplers was obtained February 15 and 16,
1979. Comparison of the data obtained from
these two periods of field measurement, shown
in Table 10, indicates similarities. During the
stormy period, the high-volume sampler near
the center of the SRC pilot plant recorded more
than twice the concentration of particulates. A
definite plume structure toward the northeast
is indicated from the mean data plotted on
Figure 8.
Analysis by liquid chromatography of the
middle and heavy distillates, which are the
products of the SRC-II facility, is shown in
Figures 9 and 10.
CONCLUSION
Evaluation of the data analyzed at this time
indicates rough establishment of priorities
for the wastewater streams associated with an
SRC-II pilot plant operation. The most poten-
tially toxic waste stream is the recycle process
water tank. Establishment of initial priorities
for the wastewater stream is shown in Table 11.
One important fact is demonstrated by pre-
liminary analysis of the samples obtained! from
the SRC pilot plant: that large variations in
chemical concentrations, and perhaps even their
existence, occur during minor shifts in operat-
ing conditions. Thus, a single grab sample re-
veals very little about the chemicals or concen-
trations that can exist in a waste stream from
an SRC plant. In order to evaluate changes in
wastewater constituents, detailed information
is required on operating conditions and changes
in coal type, feed rate, temperature, pressure,
and other physical parameters.
Extrapolation of data to different operating
conditions or to other modes of operation is
meaningless at this time. A complete set of
nearly identical samples must be obtained and
analyzed under the SRC-I mode of operation.
Ordering priorities for the wastewater
stream for the SRC-II mode appears to indicate
that the following should be evaluated under
EPA's Level 3 criteria:
• Recycle process water,
• Sulfur area drain,
• Wastewater treatment plant inflow, and
• Wastewater treatment plant effluent.
351
-------
CO
u
®26
b »«
i i
>— CMTUTS/CLMC0 -/
LEGEND:
STORM SEVER
WASTE DRAIN
Hgure 8. Hgh-vokime locations and 24-hr values.
-------
N FROM VENT AT THE SRC PILOT PLANT
Sample
Site*
1
Coal Process
Area Drain
2
Coal Storage
Area Drain
3
Dissolved
Separator
Area Drain
4
Sulfur
Recovery
Area Drain
6
Recycle
Process
Water Tank
7
Boiler
Slowdown
8
Cooling
Tower Basin
9
Wastewater
Treatment
Plant
Influent
10
Wastewater
Treatment
Plant
Effluent
11
Drain from
Blacktop
Area
12
Sandvick
Belt Cooling
Water
13
Solvent
Fractionation
Area Drain
• • •mri — i
Date
Taken
2/11
2/12
2/15
2/16
2/17
2/11
2/12
2/15
2/16
2/17
2/11
2/12
2/15
2/16
2/17
2/11
2/12
2/15
2/16
2/17
2/11
2/12
2/15
2/16
2/17
2/11
2/12
2/15
2/16
2/17
2/11
2/12
2/15
2/16
2/17
2/11
2/12
2/15
2/16
2/17
2/11
2/12
2/15
2/16
2/17
2/11
2/12
2/15
2/16
2/17
2/11
2/12
2/15
2/16
2/17
2/11
2/12
2/15
2/16
2/17
pH
9.65
11.80
10.30
8.20
2.45
1.85
1.95
1.90
7.80
8.60
7.25
10.70
9.35
9.60
9.30
8.60
8.25
9.05
8.90
9.00
8.85
11.55
11.60
11.40
6.25
6.55
7.10
8.80
8.60
7.80
7.50
7.20
7.10
7.25
7.05
7.90
2.50
2.30
2.50
7.05
7.15
7.50
7.05
7.40
8.60
6.86
7.45
7.25
6.85
NOf
1.93
2.83
2.20
0.55
<0.10
<0.10
<0.10
2.33
0.67
0.54
0.30
1.76
<0.10
<0.10
<0.10
<0.10
21.20
6.51
<0.10
<0.10
<0.10
4.88
0.09
2.29
1.30
0.87
<0.10
1.39
0.03
<0.10
<0.10
0.33
0.27
0.01
<0.01
<0.10
0.50
0.92
0.60
0.47
0.27
0.76
1.49
<0.10
<0.10
<0.10
0.58
Field Analysis _
lmff/1) . tan/l| (mc/I| (mgAI
74.0 <0.1
35.0 0.5
100.0 1.0
6.0 1.0
<1.0 <0.1
<1.0 <0»1
27.0 < 1.0
2.0 <1.0
<1.0 <1.0
< 1.0 0.3
23.0 <0.1
52.0 <1.0
2600.0 <1.0
730.0 8.0
> 100000.0 <0.1
38300.0 1.1
> 100000.0 26.0
> 100000.0 2200.0
> 100000.0 800.0
> 100000.0 16700.0
> 100000.0 817.0
7.2 <0.1
115.0 1.0
2300.0 <1.0
34.0 <0.1
12.0 0.1
33.0 <1.0
55.0 1.9
755.0 2.1
8600.0 0.6
2430.0 <1.0
193.0 <1.0
5.6 <0.1
12.0 <0.1
148.0 0.2
11300.0 <1.0
580.0 <1.0
<1.0 <0.1
<1.0 <0.1
<1.0 <0.1
12.0 <0.1
<1.0 <0.1
<1.0 <0.1
4100.0 <0.1
12.0 <0.1
15.0 <0.1
180.0 2.3
340.0 9.0
275.0 <1.0
988.0 <1.0
0.09
0.80
1.50
0.50
<0.09
<0.10
<0.10
<0.10
0.09
0.10
<0.10
<0.50
0.50
9.30
570.00
1700.00
93.00
1400.00
672.00
5700.00
1300.00
6.13
'
1.30
0.20
0.09
<0.10
0.10
8.83
3.00
0.60
16.00
0.80
0.16
0.30'
<0.10
6.00
0.40
<0.09
<0.10
<0.10
0.11
<0.10
<0.10
1.00
0.20
2.18
8.20
<0.10
2.40
0.70
0.1
0.1
0.2
0.1
0.1
0.1
0.1
0.1
0.4
0.1
0.5
0.1
0.4
5.0
27.0
6.4
14.0
8840.0
300.0
3200.0
1200.0
0.1
1.6
0.5
0.2
0.1
0.4
100.0
6.0
<0.1
10.2
0.5
11.5
15.0
10.5
7.9
11.0
0.3
0.1
<0.1
0.6
<0.1
0.2
4.3
0.2
5.6
6.0
4.0
0.5
Conductivity
-------
80 <
70
z 60
0
i 50
t__
LTI l^O
O
z 30
UJ
o
CC
£ 20
1C
_
-
-
11. 1.1
^ r^ ri
MIDD1E DISTILLATES
63. *i
\—\
23^56?
LC FRACTIONS
Figure 9. Relative distribution of total organic* in seven LC fractions.
90 p
80 -
70
HEAVY DISTILLATES
50 h
0
z
LU
£ 30
a.
20
10
_
8.8 7 7
n n
12.3 9 g _!!L
_JH-
LC FRACTIONS
Figure 10. Relative distribution of total organic* in seven LC fractions.
354
-------
TABLE 11. ESTABUSHING PRIORITIES FOR WASTEWATER STREAMS
FROM SRC PILOT PLANT
Order
Stream
1
2
3
4
5
6
7
8
9
10
11
12
Recycle process water tank
Sulfur recovery area drain
Wastewater treatment plant inflow
Wastewater treatment plant effluent
Coal storage area drain
Drain from general surfaced area
Coal preparation area drain
Dissolver/separator area drain
Solvent fractionation area drain
Boiler blowdown
Sandvik belt water
Cooling water
In addition, the flare knockout drum water was
not sampled because of operational difficulties
and should be evaluated for potential toxic sub-
stances.
ACKNOWLEDGMENTS
Information for this paper was compiled from
work performed by Hittman Associates, Inc.,
under EPA Contract Number 6842-2162. This
work has been guided and supported by EPA's
Fuel Process Branch of the Industrial Environ-
mental Research Laboratory at Research Tri-
angle Park, N.C.
REFERENCES
1. Coal Liquefaction (quarterly report). U.S.
Energy Research and Development Admin-
istration. Washington, D.C. ERDA 7633-4.
October-December 1975.
2. Koralek, C.S., and S. 8. Patel. Environmen-
tal Assessment Data Base for Coal Lique-
faction Technology, Volume I, Systems for
14 Liquefaction Processes. U.S. Environ-
mental Protection Agency. EPA -600/7-78-
184a. September 1978.
3. Lentzen, D. E., D. E. Wagoner, E. D. Estes,
and W. F. Gutkneck. IERL-RTP Procedure
Manual: Level 1 Environmental Assess-
ment Draft, (second edition). U.S. Environ-
mental Protection Agency. EPA-600/7-78-
201. January 1979.
4. Duke, K. M., M. E. Davis, and A. J. Dennis.
IERL-RTP, Procedure Manual: Level 1
Environmental Assessment Biological
Tests for Pilot Studies. U.S. Environmental
Protection Agency. EPA-600/7-77-043.
April 1977.
5. Harris, J. C., and D. L. Levin. EPA/IERL-
RTP Interim Procedures for Level 2 Sam-
pling and Analysis of Organic Materials.
U.S. Environmental Protection Agency.
EPA-600/7-78-016. February 1978.
6. Ferretti, E. J. Coal Liquefaction Gains
Prominence. Coal Mining and Processing.
1312). February 1976.
7. Schmid, B. K. Status of the SRC Project.
Chemical Engineering Progress. 71(4).
April 1,1975.
355
-------
8. Solvent Refined Coal (SRC) Process-
Development of a Process for Producing an 10.
Ashless, Low-Sulfur Fuel from CoaL U.S.
Energy Research and Development Admin-
istration. Report Number 52.1977.
9. Goldstein, D. S., and D. Yoring. Water Con-
servation and Pollution Control in Coal
Conversion Processes. EPA Industrial En- 11.
vironmental Research Laboratory. Re-
search Triangle Park, N.C. EPA-600/7-
77-065. June 1977.
Ghassemi, M., et al. Applicability of Petro-
leum Refinery Control Technologies to Coal
Conversion. EPA Industrial Environ-
mental Research Laboratory. Research
Triangle Park, N.C. EPA-600/7-78-190. Oc-
tober 1978.
Manual on Disposal of Refinery Waste. In:
Volume on Liquid Wastes. American Pe-
troleum Institute, 1969.
356
-------
ENVIRONMENTAL ASSESSMENT OF SRC-II-AN UPDATE
C. Raymond Moxley*
Gulf Mineral Resources Company, Denver, Colorado
and
David K. Schmalzer
Pittsburg & Midway Coal Mining Company, Merriam, Kansas
Abstract
This paper describes the activities that have
been undertaken, as well as future environmen-
tal activities that will occur in the succeeding
phases of the 6,000 T/D SRC-II Coal Liquefac-
tion Demonstration Project This plant will be
built in the Morgantown, West Virginia area
under sponsorship of the U.S. Department of
Energy (DOE).
Currently, the phase is characterized by ef-
forts in two main areas:
1. Collection of baseline data for incorporation
into an environmental impact statement. A
brief description of our data-gathering ef-
fort is given with special attention to:
• Results from the air-monitoring station,
especially regarding the ozone attain-
ment/nonattainment status in the area.
• Existing levels ofPNAs in soils, ambient
suspended particulate matter, ground-
waters, and surface waters/sediments in
the Monongahela River and the various
tributaries that traverse the project site.
• Expected impact of the project on the ex-
isting socioeconomic climate of the re-
gion.
2. Identification of the following anticipated
major environmental concerns of the proj-
ect:
• Current plans for the onsite disposal of
approximately 800 T/D of a potentially
hazardous waste.
• Industrial hygiene and potential health
effects of the plant A medical surveil-
lance program for plant workers and the
status of the toxicology programs for
SRC-II (solvent refined coal)products and
intermediate streams will be addressed.
• Consumptive use of water and its impact
•Speaker.
on the Monongahela River.
• Status of combustion tests on SRC-II
product oil and anticipated environmen-
tal concerns of its use.
INTRODUCTION
Increased use of America's coal supplies is
not only a goal of the National Energy Policy
but also a highly desirable fuel supply alter-
native for the electric power-generating in-
dustry. Unfortunately, conversion from fuel oil
to coal is not only expensive but prohibited in
certain regions of the country because of exist-
ing and proposed environmental regulations.
We are all familiar with the U.S. Environmental
Protection Agency's (EPA) proposed regula-
tions of 85 percent sulfur removal for the utility
industry. The economic consequences of this re-
moval rate are formidable, with estimates rang-
ing around $800 to $l,000/ton of S02 removed.
The objective of the SRC-II project is to use
our coal reserves to provide a liquid fuel that is
competitive with petroleum-derived boiler fuel,
both environmentally (low sulfur and low ash)
and economically. Longer range utilization of
SRC-II products could include heating oils,
gasoline, and feedstocks for chemical produc-
tion.
HISTORY
In July of 1978, the U.S. Department of
Energy (DOE) entered into a contract with the
Pittsburg & Midway Coal Mining Company to
undertake a conceptual design of demonstration
plant. This plant would have a coal feed rate of
6,000 T/CD and produce the equivalent of 20,000
bbl/d. Table 1 traces part of the historical devel-
opment of the process that led to this contract.
357
-------
TABLE 1. HISTORICAL DEVELOPMENT OF THE
SRC-II PROCESS
CO
1965 - Technical feasibility of Solvent Refined Coal is proven by Spencer
Chemicals under sponsorship of the Office of Coal Research.
1972 - The Pittsburg & Midway Co. is contracted by DOE to construct
a 50 T/D pilot plant at Ft. Lewis, Washington.
1974 - Start-up of pilot plant.
1978 - DOE contracts The Pittsburg & Midway Coal Mining Co. to
undertake:
- Conceptual design of a 6,OOO T/CD Demonstration Plant
- Marketability and economic assessments
- An Environmental Analysis of the plant site including the
defining of all air, liquid and solid waste emissions.
-------
PROCESS DESCRIPTION
BASELINE COLLECTION EFFORT
Figure 1 shows the location of the proposed
demonstration plant site, approximately 5 mi
north of Morgantown and bordering on the west
side of the Monongahela River. The location of
the meteorological and air quality monitoring
station (MAQS), which is actually in Pennsyl-
vania, is indicated by a star on the figure.
Figure 2 shows a schematic flow diagram of
the process. The feed, a typical high-sulfur
Pittsburgh seam coal, is mixed with a recycle
slurry produced by the process. Hydrogen
enters the coal-slurry mixture and is pumped
through a preheater to the reactor where the
coal is dissolved and hydrocracked.
The effluent from the reactor enters a series
of vapor-liquid separators. The light process
gases containing hydrogen, H2S, and C02 are
sent through an acid-gas removal system fol-
lowed by a cryogenic unit to separate the
hydrogen that is recycled to the process. The
hydrocarbon gases are refined into a methane,
ethane, propane, and mixed C4 product streams.
The light liquid stream is fractionated into a
naptha product (C5-350° F and End Point) and a
middle distillate (350° to 600° F). The product
slurry is split so that part is recycled to the
front end to be mixed with the feed coal. The
other portion is sent to vacuum distillation
where a heavy distillate is produced and mixed
with the middle distillate from the atmospheric
tower. These two streams (heavy and middle dis-
tillates) comprise the final SRC-II fuel oil prod-
uct.
The vacuum tower bottoms are sent to a high-
pressure slagging gasifier for production of syn-
thesis gas, a mixture of hydrogen and carbon
monoxide. Part of this gas is catalytically
reacted with steam (water gas shift) to repro-
duce hydrogen. The other part of the synthesis
gas is treated to remove acid gases and is used
for plant fuel.
The temperatures in the gasifier are suffi-
cient to liquefy the mineral matter in the feed.
This molten ash is cooled and solidified by a
water quench and, after appropriate dewater-
ing steps, is sent to the ash disposal area located
onsite. Approximately 800 T/CD of slag will be
produced by the plant.
The main objective of any demonstration
plant must be that all aspects of the project are
investigated to determine their feasibility for
commercialization. These aspects include not
only engineering technology, economics, and
marketability of products, but also the environ-
mental acceptability of the technology and its
products. If a demonstration project does not
adequately address all these issues, it has not
accomplished its objectives.
Before one'can characterize the environmen-
tal impacts, it is necessary to undertake exten-
sive environmental baseline monitoring. This,
plus the post-operational monitoring, will facil-
itate scientifically sound conclusions regarding
these impacts and provide a firm foundation on
which to judge the impacts of a commercial
plant.
A brief summary of the baseline collection ef-
fort underway at Morgantown is presented in
Tables 2 through 7. During establishment of a
detailed work plan for this effort, two objectives
were of prime concern:
• That sufficient data be collected on the ex-
isting environment to comply with the re-
quirements of NEPA documents and per-
mits; and
• To characterize aspects of the environment
that the plant might affect so that the post-
operational monitoring would logically con-
tinue from the baseline data collection ef-
fort.
Only at the demonstration phase of any fuel
conversion technology development can real en-
vironmental issues be quantified and judgments
made regarding the environmental acceptabili-
ty of the process.
RESULTS OF BASELINE
DATA COLLECTION
Currently, the data collection effort is ap-
proximately 75 percent complete. It is antici-
pated that all sampling, analyses, and compila-
tion of data will be complete for inclusion in the
draft EIS scheduled for January 1,1980. DOE
plans to issue the final EIS in July 1980.
In this section, no attempt will be made to list
359
-------
GREENE CO.
MONONGALIA CO*
. Meteorological
IT Tower
PENNSYLVANIA
WEST VIRGINIA
Demo
Process
Plant
'Area
^ - - -\
FUTURE SLAG ,,i
DISPOSAL AREA
DEMO
SLAG
DISPOSAL
AREA
Morgantown is
approx. Smiles
south
0 IQOO 2000
GRAPHIC SCALE IN FEET
Product Tankage
and Shipping
Figure 1. Proposed site for SRC-II demonstration plant.
-------
METNANATION
DRIEI
POLVERIZEI
com
PURIFIED HYDROGEN
SLURRY
MIXING TANK
RECIPROCATING
PUMP
VAPOR-LIQUID SEPARATORS
DISSOLVER
SLURRY
PREHEWR
0
0
0
PRODUCT
SLURRY
SHIFT CONVERSION
AMI PURIFICATION
E-UP
OGEN
ACII GAS
REMOVAL
§00
\
0
d
4
1 A
PUNT
F«L
• SULFUR
OXYGEN
PLANT i STEAM i
GASIFIER
LIGHT
LIQUID
FRACTION-
ATOR
CRYOGENIC
SEPARATION
I
ACII GAS REMOVAL
SULFUR
MINERAL RESIDUE SLURRY
I
VACUUM TOWER
^PIPELIME
GAS
+-MPHTIA
OIL
INERT SLAG
Figure 2. SRC-II process.
-------
TABLE 2.
(1) Collect one-year of meteorological data utilizing the 60-meter
meteorological tower located near the Morgantown site.
" (2) Collect one-year of air quality data for SO2, NOX, O3, NMHC, CO, total
suspended particulates (every third day). On a quarterly (seasonal)
basis, analyze collected particulate matter for various trace
metals (approximately 70 elements) and trace polynuclear aromatic
hydrocarbon (four PNA's have been selected that have been historically
used as "indicator" compounds of possible carcinogenic activity.)
-------
TABLE 3.
(1) Make quarterly analyses of surface waters near the site: Monongahela
River (three stations), Robinson Run (two stations) and Crooked Run
(three stations). In addition to the "standard" water quality parameters,
e.g., COD, BOD, TSS, pH, dissolved oxygen, TOC, O&G, TDS, etc., trace
metals and polynuclear aromatics concentrations are also being
determined.
(2) The same analyses, except BOD, are being conducted on groundwater
samples taken from four existing wells near or on the plant site.
(3) A hydrogeology study involving six drill holes will be conducted in the
slag disposal and coal storage and preparation areas to determine:
• Rock identification
- Potential seepage rates and pathway identification
- Depth to nearest aquifer
- Piezometer measurements
- Ability of existing clay to prevent seepage from ash disposal
- Logs of all holes and cores
-------
TABLE 4.
(1) Seasonal studies to qualitatively analyze the existing plant communities.
Included in this work is species inventory, plant community identification,
construction of vegetation maps, and examination for unique, rate, and
proposed special-status plant species.
(2) Seasonal sampling and observation of existing wildlife at the site via
live-trap transects, mist nets, and vehicular and walking transects.
(3) Sample aquatic flora and fauna quantitatively and qualitatively at three
stations on the Monongahela River and two on Crooked Run. Under
investigation will be such groups as fish, macroinvertibrates, zooplankton,
and phytoplankton.
-------
CO
TABLE 5.
(1) Conduct a study to identify the existing relationship between population,
economy, land use, and the demand for public and private utilities,
services, and facilities in Morgantown and neighboring communities. A
£ labor availability study will be conducted, as well as an evaluation of the
adequacy of the present roads and highways.
(2) Conduct a study of the possible presence of cultural resources (historical/
archaeological) within the 2600-acre area.
-------
TABLE 6.
(1) Characterization and mapping of soils present within the project as area
boundaries will be conducted. Trace metals and polynuclear aromatics will
also be analyzed in the soils and in sediments collected from the
Monongahela River (two stations) and at the mouths of Robinson and
co Crooked Runs.
(2) A revegetation plan for the solid waste disposal area will be developed to
satisfy West Virginia solid waste regulations. Stabilization considerations
will include both material and procedural aspects of top soil handling,
seedbed preparation, seeding, application of soil amendments, mulching
and maintenance.
-------
bo
3
TABLE 7.
(1) Conduct two noise surveys (one winter and one summer) to establish
existing noise levels at established measuring points.
(2) Sampling for ichthyoplankton (fish larvae) will be conducted during
spawning times to assess the environmental affects of the raw water
intake structure.
-------
in detail all results of the data collection; rather,
only issues judged to be environmentally signif-
icant will be discussed.
Air
Table 8 summarizes the data collected in the
first 3 mo of the program. Of particular note is
the fact that the mean nonmethane hydrocarbon
(NCMC) value for the 3-mo period is essential-
ly twice the EPA guideline. This is not con-
sidered unusual for heavily wooded areas such
as the Morgantown site. Sulfur dioxide values
average about 7 percent of the National Am-
bient Air Quality Standard (NAAQS), although
few data have been obtained with SE wind di-
rection (i.e., the direction of a major source of
S02 in the area—an electrical power-generating
station).
The maximum 1-hr concentration for ozone
was 0.115 ppm, only slightly below the new
NAAQS of 0.12 ppm. This level was observed in
September (the only month that the instrument
operated properly). Of the 595 hourly observa-
tions, 16 (2.6 percent) exceeded the old 1-hr
NAAQS of 0.08 ppm. We see this as a potential
problem in that the theory of photochemical oxi-
dant formation suggests that the highest levels
will be observed during the hottest months of
the year. Consequently, the exact situation
regarding ozone will not be known until summer
data are collected. This area of Pennsylvania
was previously listed as a nonattainment area.
A new designation, if any, has not been pub-
lished in the Federal Register. All other aspects
of air quality are well within the NAAQS.
Table 9 shows the data that have been ob-
tained on the background levels of polynuclear
aromatics. These analyses are performed at
Gulf Science & Technology on the particulate
matter collected by high-volume air samples.
While no Federal or State standards exist for
these materials, the values are judged to be
quite low. We envision these analyses to be a
very important part of the post-operational
monitoring program.
A composite of the particulate material col-
lected on four high-volume filters was analyzed
for trace element concentrations using mass
spectrographic and atomic absorption tech-
niques. Table 10 shows some of the data re-
sulting from these analyses. Of the elemental
concentrations, silicon, aluminum, copper,
calcium, and potassium represented the major
constituents. However, the concentrations of
these five elements were within the ranges nor-
mally observed in the atmosphere. Of the toxic
elements, only copper was slightly above nor-
mal background levels. Beryllium, chromium,
fluorine, lead, molybdenum, and selenium were
all within normal ambient ranges while arsenic,
cadmium, mercury, nickel, vanadium, and zinc
were below normal measured ranges.
West Virginia has no ambient standards for
trace elements. EPA has only a standard for
lead of 1.5 /ig/m8,3-mo average.
Pennsylvania has ambient standards for the
following trace elements:
Element Allowable Concentration
Lead 5 /*g/m3, 30-day mean
Beryllium 0.01 ^g/m3, 30-day mean
Sulfates 10 uglm3, 30-day mean
Fluoride 5 pg/m3, 24-hr mean
The measured concentrations of lead, beryllium,
and fluoride were within these ambient stand-
ards, while the sulfate concentrations slightly
exceeded the standards.
In terms of expected S02 emission from the
demonstration plant, Table 11 shows the results
of some preliminary dispersion calculations.
The table also shows the corresponding PSD
allowable increments. There are potentially 12
sources of S02 in the plant, but by far the major
source is the incinerator on the sulfur recovery
system. Current plans are to use a combination
of Claus unit and Super-Scot tail gas cleanup
unit.
During normal operations, all nonmethane hy-
drocarbons within the process will be within a
completely enclosed system with vapor recov-
ery systems on all storage tanks and vessels.
The only source of fugitive emissions will be
leakage losses from valves, flanges, etc. If
measured according to the EPA's publication,
AP42, "Guideline for Emission Factors," these
leakage emissions of total hydrocarbons from
the SRC-II Demonstration Plant will be 28 Ib
per 1,000 barrels of liquid hydrocarbon product.
With 16,058 barrels of liquid hydrocarbon prod-
ucts per day and 88.1 MM SCF (million standard
cubic feet) per day of gaseous product, fugitive
hydrocarbon emissions are estimated to be 450
Ib/day, 18.7 Ib/hr, or 2.36 g/s. Because some por-
368
-------
TABLE 8. SUMMARY OF BACKGROUND AIR QUALITY DATA
(PPM)
POLLUTANT
Sulfur Dioxide
Nitrogen Oxides
Carbon Monoxide
Ozone
NMHC
Participate Matter
OBSERVED LEVELS
3-hr. Max. = O.044
24-hr. Max. = 0.012
3-Month Mean = 0.002
24-hr. Max. = 0.03
3-Month Mean = 0.01
1-hr. Max. z 2.35
8-hr. Max. = 1.37
1-hr. Max. = 0.115
3-Month Mean = 0.5
24-hr. Max. = 114 ug/m3
3-Month Mean = 49 ug/m
NAAQS
0.50O (Secondary)
0.140
0.030 Annual
0.05 Annual
35.0
9.0
0.12
0.24 Guideline
o
260 ug/m
o
75 ug/m° Annual
-------
TABLE 9. AMBIENT LEVELS OF POLYNUCLEAR AROMATICS
COMPOUND
CONCENTRATION (ug/m°)
Benz( a) Anthracene
Benz(a)Pyrene
Benz(e)Pyrene
Benz(g, h, DPerylene
Pyrene
0.000458
0.001044
0.000027
(none found)
0.000516
TABLE 10. TRACE ELEMENTS
ELEMENT CONCENTRATION (ug/m°)
Aluminum
Bromine
Calcium
Copper
Iron
Lead
Magnesium
Manganese
Mercury
Phosphorous
Potassium
Silicon
Sodium
Sulfate
Sulfur
Tin
Titanium
Zinc
1.26
0.01
0.46
O.59
0.30
0.09
0.25
0.02
0.0002
0.16
0.39
2.06
0.11
13.97
0.25
0.01
0.05
0.08
tion of the leakage hydrocarbon losses will be
methane, leakage losses of nonmethane hydro-
carbons will be somewhat less than this amount.
The leakage losses for the SRC-II Plant will be
less than this amount. The leakage losses for the
SRC-II Plant will be less than losses from con-
ventional gas plants and oil refineries.
No specific estimation for benzene release to
the atmosphere has been made; however, such
releases are expected to be less than releases in
petroleum refineries. Analytical results on the
SRC-II light oil (naptha) conducted at Gulf
Research & Development and at independent
laboratories have shown that the benzene con-
tent is less than 1.0 percent.
Water
In the area of surface water analyses,
quarterly (seasonal) analyses are being made on
the Monongahela River at three stations and on
the various tributaries that traverse the plant
site (i.e., Robinson Run, Crooked Run, and
Crafts Run). Table 12 summarizes the summer
and fall analyses. Rather than show the com-
370
-------
TABLE 11. MAXIMUM GROUNOLEVEL SULFUR DIOXIDE CONCENTRATION ESTIMATES
SRC-II DEMONSTRATION PLANT
MAXIMUM SO2 CONCENTRATIONS (ug/m3)
EMISSION SOURCE
1
2
3
4
5
6
7
8
9
10
11
12
TOTALS
Federal PSD
Class II Standards
3-HOUR
9.9
2.1
8.4
1.7
0.2
80.3
1.2
0.2
1.0
2.1
0.2
0.6
107.9
512
24-HOUR
3.9
0.8
3.3
0.7
0.1
31.7
0.5
0.1
0.4
0.8
0.1
0.2
42.6
91
ANNUAL
0.1
0.0
0.1
0.0
0.0
0.8
0.0
0.0
0.0
0.0
0.0
0.0
1.0
20
-------
to
TABLE 12. SUMMARY OF SUMMER AND FALL SURFACE WATER ANALYSES
pH - exceeds existing and proposed criteria in Robinson
^
and Crafts Run, August and November.
Arsenic - exceeds existing and proposed criteria in Robinson
Run, August.
Lead - exceeds existing and proposed criteria in Robinson
Run, November.
Manganese - meets or exceeds the proposed criteria in every
sample, August and November.
Iron (total) - exceeds the proposed criteria in many of the
tributary samples, August and November.
Phenols - exceeds proposed criteria in Crooked Run, August.
-------
plete list, the table indicates parameters that
currently exceed proposed water quality cri-
teria. Our sampling program has been modified
to obtain additional data on these particular
parameters and streams.
The polynuclear aromatic analyses of surface
water is shown in Table 13. As in the case with
ambient air levels of PNAs, we view the contin-
uation of these analyses to be an important part
of the post-operational monitoring.
In terms of the effect of the plant on water
quality, current engineering design calls for
zero discharge of liquid effluents. This is ac-
complished by (Table 14):
• Recycling of sour water after cleanup,
• Recycling of boiler and cooling tower
blowdowns after evaporation,
• Recovery of process sewer water via an oil-
water separation,
• Collection and processing of rainwater
runoff resulting from a 10-yr 24 hr storm,
• Tertiary treatment of sanitary sewage, and
• Incineration of all sludges and solids ob-
tained by these operations.
In addition, ammonia and tar acid recovery
units will be an integral part of the plant design.
Consumptive use of water is a problem that
all conversion plants must face. While it is not as
severe a problem in the East as in some West-
ern States, when a project consumes approxi-
mately 4,000 gal/min of water, there are en-
vironmental concerns that must be addressed.
The U.S. Army Corps of Engineers has indi-
cated that the Monongahela River can supply
sufficient water for the project except during
periods of extreme drought (e.g., droughts from
the 1930's and 1950's. However, with the com-
pletion of Stonewall Jackson Dam, now under
construction on the West Fork River, the Corps
has indicated that ample water flow should be
available not only for the demonstration plant
but also for a full-size commercial plant.
Solid Wastes
By far, the largest volume of solid waste
generated in the SRC-II demonstration plant
will be the gasifier bottoms. This material,
which amounts to approximately 800 ton/day, is
expected to be very similar to the bottoms from
a coal-burning facility. Analytical programs
aimed at characterizing this material are under-
way at Oak Ridge National Laboratory, Battelle
(Pacific Northwest Laboratory), and Gulfs Re-
search Laboratory in Harmarville, Pennsylvan-
ia. With no firm analytical data on the hazard-
ous nature (as defined by the Resource Conser-
vation & Recovery Act) of the waste material,
current plans call for managing and disposing of
the initial material according to the most re-
strictive regulations. If the analytical results ob-
tained during the initial phase of demonstration
plant operation show the material to be "non-
hazardous," appropriate changes will be made
to the disposal plans.
Health Effect
The overall development of the SRC process
has included, in addition to the technical devel-
opment of the process, various health programs,
environmental studies, trace element studies,
engineering studies, and product characteriza-
tion and market development studies. The
health programs under the SRC pilot-plant con-
tract include an industrial hygiene monitoring
program, an employee hygiene and education
program, a medical surveillance program, and a
toxicology program. Similar programs are an-
ticipated for the SRC-II demonstration plant.
The principal objectives of the pilot plant
health programs are:
• Protecting the workers from exposure to
materials that could result in adverse health
effects;
• Monitoring the worker environment to
measure the extent and nature of exposure,
both to safegard health and to identify needs
for additional engineering controls or proc-
ess modifications; and
• Assessing the toxic characteristics of the
SRC materials through extensive bioassay
studies.
Process modification and control technology
needs identified in the pilot plant can be incor-
porated in the demonstration plant.
Only limited prior experience is available in
the area of hydroliquefaction of coal. Similar
technology was practiced in Germany during
and prior to World War II with a maximum of 12
plants operated; peak production was about
100,000 bbl/d of distillate products. Little health
information was obtained from these opera-
tions. Union Carbide operated a 300-ton per day
coal liquefaction plant at the Institute of West
Virginia, from 1952 to 1956. Elevated levels of
373
-------
TABLE 13. POLYNUCLEAR AROMATIC CONTENT OF SURFACE WATERS
(PPT)
Benz(a) Benz(a) Benzo(e) Benzo(g,h, i)
Anthracene Pyrene Pyrene Perylene Pyrene
« Robinson Run 0.0 0.0 - O.4 0.0 0.0 2.0
*>.
Crooked Run 0.0 - 39.0 0.0 - 3.0 0.0 0.0 43.0
Crafts Run 0.0 - 34.0 0.0 - 7.0 0.0 0.0 24.0
Monongahela River 0.0 0.0 - 0.5 0.0 0.0 11.0 -28.0
-------
TABLE 14. WASTEWATER TREATMENT
(1) Recycling sour water after cleanup.
(2) Recycling boiler and cooling tower blowdowns after evaporation.
(3) Recovery of process sewer water via an oil-water separator.
(4) Collection and processing of the quantity of rainwater runoff equivalent
to the 10-year 24-hour flood.
(5) Tertiary treatment of sanitary sewage.
(6) Incineration of all sludges and solids obtained by these operations.
-------
skin cancer were observed in the Institute plant
population.1 Many precautionary measures of
the health programs at the SRC pilot plant were
designed based on the Institute experience.
Industrial Hygiene Monitoring Program—
A two-phase industrial hygiene program was
designed and implemented at the SRC pilot
plant. The first phase was an intensive data-
gathering effort. Following the data gathering
and interpretation, an ongoing monitoring pro-
gram was begun to document the continuation
of the low exposure levels observed during the
initial phase and to alert plant personnel to in-
creased exposures from equipment failures or
process modifications. Findings of the monitor-
ing program may indicate the need for addi-
tional engineering controls of process modifica-
tions in the pilot plant or in subsequent plants.
Table 15 summarizes the principal studies
under the industrial hygiene monitoring pro-
gram and reports typical findings during SR-I
and SRC-II operation. The results of these
studies indicate, in general, low worker ex-
posures. Details of these studies have been
reported elsewhere.2'3 Monitoring and charac-
terization development work is underway in
two areas: the development of accurate and
reproducible measurements of particulate
polycyclic aromatic hydrocarbons, as benzene
solubles;4 and quantification of dermal ex-
posure.
Results of the pilot-plant industrial hygiene
monitoring program will be directly useful in
planning and implementing appropriate pro-
grams in the SRC-II demonstration.
Employee Personal
Hygiene and Education-
The pilot-plant employee personal hygiene
and educational program has two major objec-
tives:
• To inform the employee of the known and
potential hazards in the work environment,
particularly those associated with exposure
to coal-derived materials, and to motivate
the employee to use the protective meas-
ures available; and
• To provide the employees with protective
equipment, clothing, facilities, and tech-
niques needed to minimize the potential haz-
ard.
The employee educational program consists
of new employee orientation and continuing
education. The new employee orientation con-
sists of an audiovisual slide presentation des-
cribing the plant, potential exposures in the
plant, and appropriate protection techniques.
After the slide presentation, the new employee
is given the SRC Health Protection Manual and
required to read it. Then the employee is taken
through the locker room change house area and
shown the proper entrances and exits within
the area, proper disposition of soiled clothing,
and proper storage of clean work clothing and
street clothes. Similar training is anticipated for
the SRC-II demonstration plant staff.
Each process area employee is issued rubber
boots and/or leather safety shoes, work uni-
forms, underclothing, socks, work coats, hard
hats, barrier creme, and skin emollient. Employ-
ees working in areas of possible exposure are
required to wear the company-supplied cloth-
ing, shoes, and appropriate safety equipment.
Pilot-plant employees are required to change
into the company-supplied clothing before going
into the work area and to remove that clothing,
shower, and change to their street clothing be-
fore leaving after their shift. To change, facili-
ties are divided into clean areas, dirty areas,
and shower areas to minimize contamination of
individuals and clothing. These procedures have
been discussed in detail elsewhere.2's
Experience gained in the pilot plant, together
with the available results from the toxicology
program and estimates of potential exposures
from the SRC-II demonstration design effort,
will be used to develop appropriate protective
clothing and personal hygiene programs for the
SRC-II demonstration plant.
Medical Surveillance Program—
Each process-explored employee at the pilot
plant is given a detailed preemployment medi-
cal examination and an annual followup exami-
nation. This is supplemented by a quarterly skin
examination by the plant nurse and the referral
of observed skin problems to a dermatologist.
The preemployment and annual examinations
consist of a medical history, a complete physical
examination, a complete blood count, blood
chemistries, urinalysis, chest X-rays, and care-
ful examination of the skin for evidence of le-
sions. Pulmonary function tests are performed
376
-------
TABLE 15. SOLVENT-REFINED COAL PILOT-PLANT MAJOR INDUSTRIAL
HYGIENE STUDIES AND TYPICAL RESULTS
OPERATION
SRC-I
SRC-II
W
Airborne Organic Vapors, ppm
Benzene Vapor, ppm
Total Suspended Particulates, mg/m
Asbestos Fibers, fibers/ml
Hydrogen Sulfide, ppm
Sulfur Dioxide, ppm
Phenolic Vapors, ppm
~
0.7
Trace
<0.04
<0.008
~ 0.6
Trace
<0.04
< 0.008
* No asbestos used in plant during SRC-II operation,
-------
by a plant nurse. A detailed description of the
program has been previously presented.2
Evaluation of the findings of the medical sur-
veillance program has indicated no discernible
changes in the medical profiles of the exposed
employees. The only known occupational health
problems encountered at the SRC pilot plant
are mild transient dermatitis from skin contact
with coal-derived materials. Table 16 summa-
rizes medical observations during the period of
pilot-plant generation.
The most common medical problem has been
eye irritation with 50 to 60 cases, approximately
10 of which involved substantial quantities of
coal-derived solvents contacting the employees'
eyes. In all cases, these eye irritations respond-
ed satisfactorily to first aid treatment con-
sisting of eye irrigation with saline solution.
Followup medical examination by an ophthal-
mologist confirmed the absence of any pro-
longed or permanant eye damage. The strongly
irritating characteristic of the lower boiling
fractions of SRC liquids is attributed to their
phenolic content.
About 25 cases of transient erythema and
multiple cases of mild foliculitis (mechanics'
acne) have been observed. These cases have
responded well to temporary suspension of ex-
posure.
One employee developed a squamous cell can-
cer of the lower lip. The employee had pre-
viously worked 9 yr in a petroleum refinery and
was a cigarette smoker. The Washington State
Board of Industrial Insurance Appeals deter-
mined that the cancer was not related to his
employment in the SRC pilot plant.
The present experience with the SRC pilot-
plant employee population has not revealed any
of the problems experienced at the Institute
plant population, where 60 skin lesions were ex-
cised from a group of 359 coal hydrogenation
workers during a 5-yr period.1 The intensive
employee health programs, functioning at the
SRC pilot plant essentially since startup, seem
to account for the major differences between
the plants. Similar continuing medical surveil-
lance of process-exposed personnel will be im-
plemented at the SRC-II demonstration plant.
Toxicology Program-
In early 1975, a toxicology program on SRC-I
materials was recommended to ERDA. The pro-
posed work included various acute and subacute
tests and chronic skin painting studies. The
scope of work was subsequently expanded to in-
clude chronic inhalation studies and teratogenic
studies. Animal testing at a contract laboratory
began early in 1977. This initial program (sum-
marized in Table 17) was, of course, devoted to
SRC-I products and process materials. It was
terminated in June 1978 because of several prob-
lems in the contract toxicology laboratory.5 A
revised toxicology program that includes stud-
ies of SRC-II materials has been proposed. The
SRC-II portion of the proposed program is sum-
marized in Table 18. A complementary program
has been developed at the Battelle Pacific
Northwest Laboratory.
Summary-Health Effects
Extensive industrial hygiene monitoring,
employee education and hygiene, medical sur-
veillance, and toxicology programs have been
implemented during the development of the
SRC process. The level of effort is probably un-
precedented for a nonnuclear fuel of commodity
process development effort. Results to date
have been generally reassuring; measured
worker exposures have been low; medical pro-
files of plant personnel have remained essential-
ly unchanged; preliminary toxicology work has
not indicated exceptional toxicity problems.
The worker protection, employee education,
industrial hygiene monitoring and medical
surveillance programs employed during the
pilot-plant program will provide the basis for
those activities in the demonstration plant. As
additional information on toxicological proper-
ties, worker health experience, and demonstra-
tion plant worker exposures becomes available,
the health programs will be reviewed and modi-
fied as needed.
ACKNOWLEDGMENTS
Portions of the work discussed in this paper
were conducted by the Pittsburg & Midway
Coal Mining Company and its subcontractors
under Contract Number EX-76-C-01-0496 with
the U.S. Department of Energy, Division of Coal
Conversion and Utilization.
378
-------
TABLE 16. SOLVENT-REFINED COAL PILOT-PLANT
MEDICAL OBSERVATIONS
NO.RELATED TO
DESCRIPTION NO. OF INCIDENTS SRC WORK
Eye Irritation 50-60 Most
Erythema 25 25
Foliculitis (mechanics acne) Multiple Most
Skin Cancer 1 0
-------
TABLE 17. SOLVENT-REFINED COAL PROCESS SUMMARY OF
ORIGINAL TOXICOLOGY PROGRAM
Test ^ — -"""^
^^----^^~ Material
Acute Oral Range Finding
in Rats
Acute Dermal Toxicity in
Rabbits
Guinea Pig Skin Sensitization
Eye Irritation in Rabbits
Acute Inhalation Range
Findings in Rats
Subacute Dermal Study in
Rabbits
Subacute Inhalation Toxicity
in Swiss Mice
Dermal Teratogenicity
in Rats
Dermal Teratogenicity in
Rabbits
Inhalation Teratogenicity in
Rats and Rabbits
Two-year Skin Painting in
Mice
Two-year Inhalation
Carcinogenesis
Process
Solvent
X
X
X
C
C
X
C
X
P
P
X
X
Coal
Slurry
X
X
X
C
C
X
Filter
Feed
X
X
X
C
C
X
P
D
X
Dry
Mineral
Residue
C
C
X
C
P
X
Wet
Mineral
Residue
X
X
X
C
C
X
P
X
Light
Oil
X
X
X
C
C
C
X
P
X
X
Wash
Solvent
X
X
X
C
C
X
C
X
P
P
X
X
Pulver-
ized
SRC
C
C
X
C
X
X
P
P
X
X
Pulver-
ized
Coal
P
X = material to be studied
C = study completed
P = pilot study completed
D r study deleted, impractical to aerosolize filter feed
-------
TABLE 18. REVISED TOXICOLOGY PROGRAM PROPOSAL
SRC-II
Materials
Coal Slurry
Stripper Tower
Bottoms
Product Fuel
Oil
Vaccuum
Bottoms
Task 1
Acute Studies
•— o
n in
1-0
o — i
X
X
X
X
§o
tn
tia
a _>
X
X
X
X
+*
£t
ujV-.
X
X
X
X
Cl
C VI
•r- O> C
3 •#- •#
o a. t/>
X
X
X
X
^
•£S
25
X
Task 2
Dermal Studies
90-Day*
Rabbit
X
X
18-Mo.*
Mouse
X
X
X
Task 3
Reproductive Toxicology
Teratology
Dermal*
Rat/Rabbit
X
X
X
X
Inhal.*
Rat/Rabbit
X
Dermal
Multlgen.
Rat
X
X
Task 4
Inhalation
Inhal.
LC50
X
90-Day*
Rat
X
2-Year
Rat
X
Task 5
Huta-
genlclty
In Vitro
X
X
X
X
X - Material to be studied
* • Pilot study necessary
-------
REFERENCES
1. Secton, R. J., M.D. Archives of Environ-
mental Health. 1:208. 1960.
2. Solvent Refined Coal (SRC) Process: Health
Programs, Research and Development
(Report No. 53, Interim Report No. 24,
Volume III, Part 4.) Industrial Hygiene,
Clinical and Toxicological Programs. FE/496-
T15. January 1978.
3. Solvent Refined Coal (SRC) Process: Health
Programs, Research and Development
(Report No. 53, Interim Report No. 28,
Volume III, Part 4.) Industrial Hygiene,
Clinical, and Toxicological Programs.
FE/496-T19. April 1979.
4. Jackson, J. 0., and J. A. Cupps. Cor-
cenogenesis: Poly nuclear Aromatic Hydro-
carbons (Volume 3). New York, Raven Press,
1978. p. 183-191.
5. Schmalzer, D. K. SRC Pilot Plant: Health
Programs and Observations. (Electric Power
Research Institute Advisory Workshop on
Carcenogenic Effects of Coal Conversion.
Pacific Grove. September 26,1978.)
382
-------
ENVIRONMENTAL ASSESSMENT REPORT: SOLVENT-REFINED COAL
Kevin J. Shields
Hittman Associates, Inc., Columbia, Maryland
Abstract
Environmental assessments reports (EARs)
have been developed by the U.S. Environmental
Protection Agency (EPA) to provide assistance
in meeting commitments to preserve environ-
mental quality. EARs are applicable both to
emerging coal gasification and liquefaction sys-
tems. This paper addresses the environmental
assessment of coal liquefaction via solvent re-
fined coal (SRC).
An overview of the hypothetical SRC system
considered is made. Potential sources of air
emissions, water effluents, and solid waste dis-
charges are identified. Applicable control alter-
natives for the discharges are discussed. Based
on utilization of these controls, a summarized
version of the multimedia environmental goals
(MEGs) and source analysis models (SAMs) ap-
plied to SRC system discharges is presented,
highlighting existing areas of environmental
concern. Research needs for subsequent envi-
ronmental assessments of SRC also are noted.
INTRODUCTION
As part of its goal of maintaining the nation's
environment, the U.S. Environmental Protec-
tion Agency's (EPA) Industrial Environmental
Research Laboratory at the Research Triangle
Park (IERL/RTP), N.C., is directing an effort to
evaluate the environmental aspects of emerging
coal conversion technologies. Hittman Asso-
ciates, Inc. (HAD, a prime contractor to
IERL/RTP, is responsible for environmental
analysis of coal liquefaction systems. Environ-
mental assessment reports (EARs) were devel-
oped to provide best available environmental
assessment data on specified coal conversion
systems in a standardized format, thereby
facilitating utilization by EPA personnel and
other researchers in the field. This paper
discusses a draft EAR prepared by HAI ad-
dressing solvent refined coal (SRC) liquefaction
systems.
SRC systems convert high-sulfur coal and ash
coal into clean-burning gaseous, liquid, and/or
solid fuels by noncatalytic direct hydrogenation.
There are two basic system variations: SRC-I,
which produces a solid, coal-like primary prod-
uct of less than 1.0 percent sulfur and 0.2 per-
cent ash by weight; and SRC-II, which produces
low-sulfur fuel oil (0.2 to 0.5 percent sulfur by
weight) and naphtha as primary products. Both
system variations produce significant quantities
of gaseous hydrocarbons, which are further
processed to yield substitute natural gas (SNG)
and liquefied petroleum gas (LPG) products.
Some constituents formed during coal hydro-
genation may be recovered as byproducts.
ENVIRONMENTAL OVERVIEW OF SRC
SYSTEMS
Major inputs to SRC systems consist of coal,
water, and air. Major products consist of gas-
eous and liquid hydrocarbons. Sulfur, ammonia,
and phenols are recovered from waste streams
as byproducts. The SRC-I and SRC-II systems
are defined to consist of the following system
operations,1 which perform specific functions
essential to solvent refining:
• Coal pretreatment: preparation of the coal
feed to meet system specifications for size
and moisture content.
• Coal liquefaction: reaction of feed coal with
hydrogen, yielding a three-phase mixture of
increased liquid and gaseous hydrocarbon
content.
• Separation: includes all necessary phase
separations. Gas separation and solids/liq-
uids separation processes are employed in
SRC systems.
• Purification and upgrading: a fractionation
process is used to separate components of
the raw liquid products mixture by distilla-
tion, because of differences in boiling points.
A hydrotreating process may be optionally
employed to upgrade the quality of frac-
tionated product liquids.
In addition, SRC systems require the following
auxiliary processes incidental to the functions of
383
-------
the system operations:1 coal receiving and
storage, water supply, water cooling, steam and
power generation, hydrogen generation, oxygen
generation, acid-gas removal, hydrogen/hydro-
carbon recovery, sulfur recovery, ammonia
recovery, phenol recovery, and product/byprod-
uct storage facilities.
Figure 1 is a flow schematic of the SRC-I
(solid product) system that shows how the sys-
tem operations and auxiliary processes trans-
form the major input materials into products
and byproducts.1 Comparison of Figure 1 with
Figure 2, the SRC-II (liquid product) system
flow diagram, identifies the major differences in
the two processing schemes as follows:1
• The SRC-I feed slurry consists of feed coal
mixed with system-derived solvent pro-
duced in the fractionation process. SRC-II
feed slurry consists of feed coal mixed with
product slurry from the gas separation proc-
ess.
• In the SRC-I system, solids/liquids separa-
tion precedes fractionation; in the SRC-II
the sequence of these processes is reversed.
Solids/liquids separation in SRC-I is most
likely to be performed by filtration, produc-
ing the filter cake sent to hydrogen genera-
tion. In SRC-II, solids/liquids separation is
achieved by vacuum distillation, which pro-
duces a bottom residue of high mineral mat-
ter content to be gasified in the hydrogen
generation process.
Waste discharges to air, water, and land
media are identified in Figure 3. Discharges
specific either to the SRC-I or SRC-II system
are noted. Subsequent discussions of discharge
characteristics, applicable control technologies,
and environmental impact assessment are
based on a hypothetical SRC-II commercial-
scale facility, although the preliminary results
may be considered representative of SRC-I.
Waste Streams to Air
As shown in Figure 3, air emissions are
associated with a majority of the processes that
make up the SRC systems. In addition to the air
emissions sources shown, fugitive emissions,
such as vapor leaks from pressurized process
equipment, may occur in the SRC systems.1
Emissions shown in the figure are outlined
below.
• Flue gases: flue gases are produced by com-
bustion units (primarily preheaters) during
liquefaction, fractionation, solids/liquids sep-
aration, hydrotreating, hydrogen genera-
tion, and sulfur recovery. Assuming the
SNG and LPG products are used as fuel in
these units, minimal environmental effects
are anticipated.
• Coal dust: coal handling, processing, and
storage in coal receiving and storage, and
coal preparation result in particulate coal
dust entering the atmosphere. Composition
of the dust is the same as that of the raw
coal.
• Dryer stack gas: to conform to system feed
specifications for moisture content, feed coal
is dried in the coal pretreatment operation.
The stack gas produced by coal drying con-
tains particulate coal and possible volatil-
ized hydrocarbons present in the raw coal.
• Vapors and particulates from cooling: min-
eral residue resulting from solids/liquids
separation (in the SRC-II mode) and SRC
product from fractionation tin the SRC-I
mode) require cooling. Air cooling of these
substances may result in emissions of par-
ticulate solids and hydrocarbon vapors. In-
sufficient data exist to characterize these
emissions and estimate environmental ef-
fects.
• Drift and evaporation: the cooling tower
loses water to the environment as water
vapor. Chemical additives used in water
cooling may also be present in this emission.
• Boiler stack gas: presumably, coal is fired in
the boilers of the steam and power genera-
tion auxiliary process. The resulting stack
gas contains oxide of sulfur and nitrogen and
particulates in the form of fly ash. Utiliza-
tion of SRC system products is one alterna-
tive for reducing these emissions.
• Nitrogen-rich gas: the cryogenic oxygen
generation process separates an oxygen-rich
gas from ambient air for use in the hydrogen
generation process. Other components of
the air (mainly nitrogen) are discharged as
an air emission.
• Carbon dioxide-rich gas: production of hy-
drogen by gasification produces a mixture of
gases. An acid-gas removal unit separates
sulfur gases (primarily hydrogen sulfide)
from the gasifier product gas. This stream is
sent to sulfur recovery. An additional acid-
gas removal stage removes a stream of near-
384
-------
CONOCNStO OILS
RECYCLE NAREUP
SSI!0?1"^ FROHl (U)
COAL ' YOAL ~
f*°". PREPARATION
COAL TO
FEED SLURRY L.OJIEFACTION
2
PRODUCT
SLURRY
INCONOENSIRLE
(Tj)» UASTEUATEK
1 t™©
CAS
SEPARATION
)
RECYCLE SOLVENT
SEPARATED
SLURRY
FILTI
SOLIDS/LIQUIDS
SEPARATION
t
FILTERED
FRACTIONATION
5
i ^ t
R CAKE TO (lO 1 MASH SOLVENT i
.
U
HONCOMCNSIILE
GASES TO @
twRSTEVA
%1
HYDROTREATINC
*
S FROM © — *
NAPHTHA
— TO @
SOLID SJ
TO (1»
00
Ol
DELIVERED
COAL
FILTER CA
FROM (J
OIYCEN 1 .
FRO»(g> .
WATER »
r—
ACIO CAS
« T0®
H'DKOCEH
GENERATION
II
HYDROGEN-RICH
^CAS TO (T) AND
WATER
8
PROCESS
PROCESSES
RECIRCULA
COOLING
WATER
U
r
WATER
9
COOLING WATER
TO PROCESSES
FROH •
WATER ••
STEAM
POWER
10
BOILER SLOWDOWN
FROM
w
Rtoun
AND
GENERATION
STEAK TO
PROCESSES
^ELECTRICITY
' TO PROCESSES
IBM TO ft\
©
PURIFIED
GAS TO rtTl PURIFIED CAS HYDROGEN
FROM
CONCENTRATED
ACID CAS TO
©
S«C TO @
LPC T0@
ACID CASES
(0)
SULFUR
WASTEMATER
PHENOL
RECOVERY
.PHENOLS
'0 ®
LIQUID SRC
SNG
LPC II
FUEL OIL 1 t
LIGHT OH ^ — f
NAPHTHA | . {
SULFUR 1 p
AXHONIA 1
PHENOLS
PRODUCT/BY-PRODUCT
STORAGE
III
PRODUCTS T«
DISTRl'lUTION
. IT-PRODUCTS U
DISTRIBUTION
Figure 1. Flow diagram of SRC-I system.
-------
MM
ftCCVCLE MUttUT
HVWOCCII^ MYMMCtH
no* Tt> TO Qi".
•ONCOMCNSItLE CA&IS TO ^7)
DCll VERIO LOAL
1
WSIHC r»W (T1
Mn< r
Kit CAU>
•«0- 111 MO il 11
C9AI HECEIVIMC
AMI
STMACl
;
CIO CAS TO 11!
|
CEMCKAttOoi
WHFUH
MtQVtftf
iS
t»*t I0 '-' W- iMTCR
>
>«*e*OCEIi-RlfM
— ' * *-- Art
;>jv.iu* %muvAUM
'- .}? FMM &
stiiti
a
12
AMMONIA
UCDvEar
it
U.
TO rMCCSSCS COOLKH. HATE*
MIL»
KOMOOIM
'•0" -iSj
t . *«* fc)
AMMOMIA «ASTE.arE*
TO IB '«>- 1
WATE*
COOLIMh
?
nwifiEO
CAS 18 -t^
1)
PMC«OL
NECOvEAT
i;
1001INC WATEH COAL "~ ""
fcWM- . ^ STCAn AND
TO MKESVCS ffton i fQntll U»EMTiON
u&ita 'O ,
1
OlOKMM TO (^
AC 10 CAS TO ^Ts) F1W* -'j.
^€»OlS L,t .
10 vlj" ru£t OIL .
SULFUR •
AMMONIA - -— •
" STEAM TO
MOUSSE i
ELECTRICITY
MTORQCAMON
ucovf**
u '
i-J
MOOtXT/IT-HOOUCT
18
OMOCIH TO i
-^ ~
SMC TO I81
^
CUT OILS TO •(!•;
bi^TliijTiTM
-SmmsTTw10
Figure 2. Flow diagram of SRC-II system.
-------
COM. OUST
MTER STACK CAS
COAL riLE
ft
.THICKENER
UNOWLOU
COAL CLEAN IM
UFUSE
COAL FILE
•RUNOFF
PMHEATER
FLUE CAS MMON
DIOXIDE
MOCtSS
H«ST(W«Tf
•RCNEATEA FLU* GAS
t
LIQUEFACTION
2
UATE*
SUfPlT
«
\
SLUOU
•ITKOCtH
men CAS
t
OITCEII
CEHEDATim
12
1
AnMNIA
«ECOVE«Y
It
CAS
SEPARATION
)
PROCESS
WASTEUATER
VAPORS AM MPORS AM
^ARTICULATES PREHEATER PMTICULATES PREHEATER
(SRC-I) FLUE CAS (SRC-II) FLUE GAS
rm>CC» SOLIDS/LIQUIDS
WASTtWTER FRACTIONATION SEPARATION
» 5
j
EUESS MSIDUE (SKC-II)
PREHEATER
FLUE CAS
HVOROTREATINO
t
PROCE
WASTE
i
SPENT CATALYST
M FILTER CAKE (SXC-I)
Ml FT AMP
EVArOMTIOH
1
M>TE«
coot ixe CMimc TOWE*
~ ILOUOOUN
s
niLER STACK CAS
1
STEAN AN
B
POUEI CEMRATION
10
1
ASH
ACIB CAS
KEMIVAL PROCESS
13
S«C H
Hr(MOCAR*0«
ANO HTOROCCN
\k
MIOCCSS
VASTEWATER
HYDROCARRON
VAPOI
ST 1
(S«-l) 1
J.
1
— — — PMWULI/K-
PHENOL STOKACE
MCOVEKY PKOCESS >'"•«•'
17
IS
SULFUR
DOST
f
PRODUCT
LIQUID
"EFFLUENT
Figure 3. Source of waste discharges in SRC systems.
-------
ly pure carbon dioxide.
• Low-sulfur effluent gas: sulfur-bearing acid
gases from hydrogen generation and SRC
system operations are treated to convert
sulfur gases to elemental sulfur.
• SBC dust (SRC-I mode) and sulfur dust:
handling and storage of SRC system solid
products and byproduct sulfur result in re-
lease of dust to the environment.
• Hydrocarbon vapors: liquid products of
SRC systems contain volatile hydrocarbon
components. Care must be exercised in han-
dling and storage of these liquids to mini-
mize emissions.
Waste Streams to Water
Sources of wastewater shown in Figure 3 are
briefly discussed below.
• Coal pile runoff: precipitation striking the
raw coal in coal receiving and, storage and
coal preparation infiltrates the coal pile.
During this contact, leaching of both organic
and inorganic constituents of the raw coal
occurs. Runoff water is collected for treat-
ment.
• Thickener underflow: wastewater from the
coal pretreatment operation is routed to a
thickener. Clarified water is recycled to coal
preparation. The underflow stream contains
a high level of suspended solids and coal-
derived organic constituents.
• Cooling tower blowdown: drift and evapora-
tion from the cooling tower result in in-
creased concentrations of dissolved and sus-
pended solids in the process cooling water.
A blowdown or "bleed" stream is withdrawn
to maintain dissolved and suspended solids
concentration within design specifications.
• Process wastewater from hydrogen genera-
tion: wastewater from hydrogen generation
may contain tars, oils, and ammonia. This
stream is directed to the main wastewater
treatment facility.
• Process wastewater from acid-gas removal:
a purge stream is removed from the amine-
based acid-gas removal process to maintain
the concentration of amine and to remove
spent amines that have formed nonregener-
able compounds. This stream is directed to
the main wastewater treatment facility.
• Process wastewater from ammonia recov-
ery process: wastewaters from hydrotreat-
ing, hydrogen generation, and hydrogen/hy-
drocarbon recovery contain significant quan-
tities of ammonia. These wastewaters are
combined and input to the ammonia recov-
ery process. The effluent wastewater exit-
ing ammonia recovery contains hydrogen
sulfide, phenols, hydrocarbons, and traces of
ammonia. This stream is directed to the
main wastewater treatment facility.
• Process wastewater from phenol recovery
process: the gas separation operation re-
moves gaseous constituents of the lique-
faction reactor effluent. Condensation of the
gases yields a phenol-rich aqueous phase,
which is sent to the phenol recovery process.
After phenol recovery the wastewater
stream, containing hydrocarbons, ammonia,
hydrogen sulfide, and traces of phenol, is
combined with other process wastewaters
(from hydrogen generation, acid-gas remov-
al, and ammonia recovery) during waste-
water treatment.
Waste Streams to the Land
Sources of solid wastes in SRC systems are
also shown in Figure 3. Sources and character-
istics of solid wastes are described below.
• Coal-cleaning refuse: refuse is a mixture of
mineral matter (such as slate and tramp
iron), water, and coal. Refuse is recovered
during coal sizing and drying.
• Excess residue (SRC-II mode) or filter cake
(SRC-I): depending on the method of hydro-
gen production employed in SRC systems,
the possibility exists that excess SRC-II
mineral residue of SRC-I filter cake may be
produced. These solids consist of mineral
matter present in the feed coal and high
molecular weight hydrocarbon species.
• Spent catalysts: the hydrotreating opera-
tion uses a catalyst to upgrade coal liquids.
A catalyst also may be employed in the shift
converter of the hydrogen generation proc-
ess. In order to maintain conversion efficien-
cies, catalysts must be withdrawn period-
ically and replaced with fresh ones.
• Ash from steam and power generation: ash
is the oxidized mineral matter present in
coal fed to the boilers.
• Slag or ash from hydrogen generation: gasi-
fication of mineral residue or filter cake to
produce hydrogen converts mineral matter
388
-------
to ash. If a high-temperature gasifier is used,
the ash may fuse and be recovered as a slag.
CONTROL TECHNOLOGY FOR
SRC SYSTEMS
Environmental impact assessment of waste
streams from SRC systems is based on applica-
tion of the control methods described in this sec-
tion. Selection of control practices is primarily
based on work efforts contributing to prepara-
tion of the EAR1 and a previous report by Rogo-
shewski et al.2
Control of Emissions to Air
Suggested control alternatives for controlling
air emissions from SRC systems are given in
Table 1. Final selection of controls for an actual
facility should be based on regional, regulatory,
economic, and site-specific considerations.1 Acci-
dental vapor discharges may occur because of
leaks caused by mechanical failure of equip-
ment. Accidental release control is best
achieved by routing emergency vent gases into
a header that directs them to the flare system.
Development and implementation of preventive
maintenance measures are essential to mini-
mize accidental air emissions because of equip-
ment failure.2
Control of Water Effluents
Table 2 summarizes the preferred control
alternatives for treating water effluents from
SRC systems. In addition to the discharges
shown in the table, accidental leaks may occur,
although they can be minimized by good preven-
tive maintenance procedures. In addition, SRC
facilities should develop a material spills con-
tingency plan including provisions for spills
detection, containment, recovery, and disposal.2
Runoff from coal preparation, receiving, and
storage is combined with thickener underflow
from coal preparation and sent to a tailings
pond. Overflow from the thickener is recycled
to the coal-cleaning process.
Cooling tower blowdown is treated to remove
dissolved solids. Lime softening, ion exchange,
and reverse osmosis are processes used to re-
duce dissolved solids content. Selection of side-
stream treatment should be based on more de-
tailed analysis of regional, economic, regulatory,
and site-specific factors. The treated water is
then discharged to receiving waters.
The remaining process wastewater dis-
charges are combined during treatment in the
main wastewater treatment facility. Two alter-
native wastewater treatment schemes, shown
in Figure 4, are considered applicable to treat-
ment of the water discharges.
Control of Solid Wastes
Preferred control and disposal alternatives
for solid wastes discharged from SRC systems
are summarized in Table 3. Most of the solids
appear suitable for direct landfilling or minefill-
ing without predisposal treatment. Spent cata-
lysts produced may be returned to the manufac-
turer for analysis and subsequent regeneration
or disposal. Should catalyst regeneration be
technically or economically unfeasible, addi-
tional research is recommended to determine if
predisposal treatment of the catalysts is re-
quired. Mineral residue from SRC-II and filter
cake from SRC-I are not well-characterized
materials. If economically feasible, it is recom-
mended that these materials be gasified to
recover available energy. The slag or ash pro-
duced by gasification may be disposed of as
solid waste.
ASSESSMENT OF ENVIRONMENTAL
IMPACTS
This section discusses environmental impacts
associated with SRC waste discharges to air,
water, and land media.1 In addition, environ-
mental aspects of handling and utilization of
SRC products are addressed.
Impacts on Air
Analysis of existing information indicates
that dust emissions from coal receiving and
storage and coal preparation, low-sulfur effluent
gas from sulfur recovery, boiler flue gas from
steam and power generation, and the emission
from the flare system should be regarded as
those emissions to air of greatest environmental
concern. Component pollutants of concern are
summarized in Table 4, based on SAM/IA analy-
sis using health-based minimum acute toxicity
effluent (MATEs) for evaluation of degree of
hazard.1 Trace element data given in these
389
-------
TABLE 1. SUMMARY OF AIR EMISSIONS CONTROL TECHNOLOGY
APPLICABILITY TO SRC SYSTEMS
Operation/Process
Air Emissions Discharged
Preferred Control Technology Applications
Coal preparation
Liquefaction
Gas separation
Fractionation
Solids/liquids separation
Hydrotreating
Coal receiving and storage
Coal dust
Particulate-laden flue
gas from coal dryers
Preheater flue gas
Pressure letdown releases
Pressure letdown releases
Preheater flue gas
Particulate-laden vapors
from product cooling (SRC-I)
Pressure letdown releases
Preheater flue gas
Particulate-laden vapors
from residue cooling (SRC-II)
Pressure letdown releases
Preheater flue gas
Pressure letdown releases
Coal dust
(1) Spray storage piles with water or
polymer.
(2) Cyclones and baghouse filters for
control of dust due to coal sizing.
(1) Cyclones and baghouse filters.
(2) Wet scrubbers such as venturi.
(1) None required (fired by SNG).
(1) Flaring
(1) Flaring
(1) None required (fired by SNG).
(1) Cyclone and baghouse filter.
(2) Wet scrubbers.
(1) Flaring
(1) None required (fired by SNG).
(1) Cyclone and baghouse filter.
(2) Wet scrubbers.
(1) Flaring
(1) None required (fired by SNG).
(1) Flaring
(1) Spray storage piles with water or
polymer.
(Continued)
-------
TABLE 1 (continued)
Operation/Process
Air Emissions Discharged
Preferred Control Technology Applications
Water supply
Water cooling
Steam and power generation
Hydrogen generation
Oxygen generation
Acid gas removal
Sulfur recovery
Hydrogen/hydrocarbon recovery
Ammonia recovery
Phenol recovery
Product/by-product storage
None
Drift and evaporation
Boiler flue gas
Carbon dioxide rich gas
Preheater flue gas
Nitrogen rich gas
Pressure letdown releases
Flue gas
Low-sulfur effluent gas*
Pressure letdown releases
None
None
SRC dust (SRC-I)
Sulfur dust
Hydrocarbon vapors
(1) No controls available - good design
can minimize losses.
(1) Sulfur dioxide scrubbing with aqueous
magnesium oxide solution.
(1) None required.
(1) None required (fired by SNG) .
(1) None required.
(1) Flaring
(1) None required (fired by SNG).
(1) Carbon adsorption.
(2) Direct-flame incineration.
(3) Secondary sulfur recovery.
(1) Flaring.
(1) Spray storage piles with water.
(1) Store in enclosed area.
(1) Spills/leaks prevention.
* A secondary sulfur recovery process may be necessary to meet specified air emission standards.
-------
TABLE 2. SUMMARY OF WATER EFFLUENTS CONTROL TECHNOLOGY
APPLICABILITY TO SRC SYSTEMS
Operation/Process
Water Effluents Discharged
Preferred Control Technology Applications
Coal preparation
Liquefaction
Gas separation
Fractionalion
Solids/liquids separation
Hydrotreating
Coal receiving and storage
Water supply
Water cooiin?
Steam and power generation
Hydrogen generation
Oxygen generation
Acid gas removal
Sulfur recovery
Hydrogen/hydrocarbon recovery
Ammonia recovery
Coal pile runoff
Thickener underflow
None
None
None
None
None
Coal pile runoff
None
Cooling tower blowdown
None
Process wastewater
None
Process wastewater
None
None
Process wastewater
(1) Route to tailings pond.
(1) Route to tailings pond.
(1) Route to tailings pond.
(1) Sidestream treatment (electrodialysis,
ion exchange or reverse osmosis) per-
mits discharge to receiving waters.
(1) Route to wastewater treatment facilitv.*
(1) Route to wastewater treatment facility.*
(Continued)
(1) Route to wastewater treatment facility.*
-------
TABLE 2 (continued)
Operation/Process Water Effluents Discharged Preferred Control Technology Applications
Phenol recovery Process wastewater (1) Route to wastewater treatment facility.*
Product/by-product recovery None
* Two alternatives for the wastewater treatment facility are shown in Figure 4
00
-------
WASTEWATER
FROM AMMONIA
RECOVERY
WASTEWATER
FROM
PHENOL RECOVERY
WASTEWATER
FROM
HYDROGEN
GENERATION
STEAM
STRIPPING
-••HYDROGEN SULFIDE TO SULFUR RECOVERY
-GAMMON IA TO STORAGE
EFFLUENT
WATER
API
SEPARATOR
EFFLUENT
WATER
^EQUALIZATION
EFFLUENT
WATER
DISSOLVED
AIR
FLOTATION
EFFLUENT
WATER
TO ALTERNATIVE I OR II
ALTERNATIVE I
EFFLUENT WATER
BIOLOGICAL TREATMENT-
EXTENDED AERATION
DISCHARGE
TO RECEIVING
WATERS
ALTERNATIVE II
EFFLUENT WATER
I
BIOLOGICAL TREATMENT
AERATED LAGOON
EFFLUENT
WATER
DISCHARGE
TO RECEIVING
WATERS
Figure 4. Two wastewater treatment alternatives
applicable to SRC systems.
394
-------
TABLE 3. SUMMARY OF SOLID WASTES CONTROL TECHNOLOGY
APPLICABILITY TO SRC SYSTEMS
Operation/Process
Solid Wastes Discharged
Preferred Control Technology Applications
co
CO
en
Coal preparation
Liquefaction
Gas separation
Fractionation
Solids/liquid separation
Hydrotreating
Coal receiving and storage
Water supply
Water cooling
Steam and power generation
Hydrogen generation
Oxygen generation
Acid gas removal
Sulfur recovery
Refuse
None
None
None
Excess residue (SRC-II)
or filter cake (SRC-I)
Spent catalyst
None
Sludge
None
Ash
Ash or slag
None
None
None
(1) Landfill
(2) Dumping (Minefill)
(1) Gasification to recovery energy content
followed by disposal (landfill or
minefill)
(1) Return to manufacturer for regeneration
(1) Dewatering followed by landfilling
(1) Landfill
(2) Dumping (Minefill)
(1) Landfill
(2) Dumping (Minefill)
(Continued)
-------
TABLE 3 (continued)
Operation/Process
Solid Wastes Discharged
Preferred Control Technology Applications
Hydrogen/hydrocarbon recovery None
Ammonia recovery None
Phenol recovery . None
Product/by-product storage None
OS
-------
TABLE 4. AIR EMISSIONS OF CONCERN* ASSOCIATED
WITH SRC SYSTEMS BASED ON SAM/IA ANALYSIS
Air Emission
Pollutant
Health-Based MATE
Potential
Degree of Hazard**
Particulate
coal dust***
Aluminum
Arsenic
Chromium
Iron
Lithium
Silicon
5200.
2.0
1.0
1000.
22.0
1.0x10*
2.3x10-^-1.7
4.9x10-3.6
1.5x10-2-11.0
1.3x10-9.9
1.4x10-3-1.1
2.1xlO-3-1.5
Sulfur re-
recovery tail
gas****
Carbon dioxide
9.0x106
87.0
Boiler flue
gas
Flare system
emission
Arsenic
Carbon monoxide
Chromium
Iron
Nitrogen oxides
Sulfur dioxide
Carbon dioxide
Carbon monoxide
2.0
4.0x10*
1.0
1000.
9000.
1.3x10*
9.0x106
4.0x10*
3.0
1.3
7.3
3.7
56
49
20
14
* Based on liquefaction of "average" U.S. coal.
Projected air concentration
** Degree of hazard ^ Health baged ^
*** Ranges due to different types of particulate controls employed.
**** Carbon monoxide and ammonia concentrations exceed ecological-based
MATE but not health-based MATE.
397
-------
discussions are projections based on an
"average" U.S. coal converted to SRC. Parti-
tioning factors based on analyses of SRC waste
materials were used to simulate distribution of
trace elements in streams exiting an SRC facil-
ity. Results of trace elements' degree of hazard
should be viewed as indicative, but not conclu-
sive, of SRC behavior.
Two important conclusions can be drawn
from Table 4. First, all emissions cited are
associated with existing industries (coal mining,
petroleum refining, and steam-electric power
generation). Concern with these emissions is not
directly attributable to operations or auxiliary
processes unique to SRC systems. Second, in
the case of coal dust, application of the more ef-
fective recommended control technology (cy-
clone and baghouse filter) reduces degree of
hazard values below one; i.e., below the health-
based MATE value.
Impacts on Water
Coal pile runoff and effluent water from the
wastewater treatment facility are considered
water effluents of concern.1 Specific pollutants
of concern are shown in Table 5. The character-
istics of coal pile runoff do result from SRC tech-
nology; however, combined wastewater charac-
teristics do result from SRC liquefaction.
TABLE 5. WATER EFFLUENTS OF CONCERN* ASSOCIATED
WITH SRC SYSTEMS BASED ON SAM/IA ANALYSIS
Air Emission
Pollutant
Health-Based MATE,
Potential
Degree of Hazard**
Coal pile
runoff
Aluminum
Calcium
Chromium
Iron
Manganese
Mercury
Nickel
Sulfate
8.0 x 104
2.4 x 105
250.
1500.
250.
10.
250.
1.5 x 1C4
9.1
1.2
8.0
6000.
272.
1.4
4.3
170.
Combined
wastewater
Bismuth
Cresols
C3~phenols
Naphthols
Phenol
Xylenol
6.1 x 103
5.
5.
5.
5.
5.
5.2
188.
18.0
60.0
78.0
76.0
* Inorganics based on "average" U.S. coal.
Organics based on characteristics of SRC bio-unit effluent,
** Degree of hazard -
Projected water concentration (ng/1)
Health-based MATE
398
-------
TABLE 6. SOLID WASTES OF CONCERN* ASSOCIATED
WITH SRC SYSTEMS BASED ON SAM/IA ANALYSIS
Air Emission
Pollutant
Health-Based MATE,
(Mg/g)
Potential
Degree of Hazard**
SRC-II mineral
residue***
API separator
bottoms
Biosludge
Aluminum
Arsenic
Barium
Beryllium
Calcium
Cobalt
Iron
Lead
Manganese
Nickel
Potassium
Selenium
Arsenic
Beryllium
Cadmium
Cobalt
Dysprosium
Lead
Mercury
Nickel
Selenium
Aluminum
Mercury
Vanadium
1.6 x 104
50
1000
6
4.8 x 10A
150
300
50
50
45
6000
10
50
6
10
150
4.6 x 102
50
45
10
1.6 x 10*
5.0 x 101
500
3.7
1.1
1.2
1.2
2.2
2.4
310.
1.4
4.8
2.1
3.0
2.0
2.0
80.0
5.0
250.
350.
364.
530.
51.0
260.
1.1
7.0
1.1
* Based on liquefaction of "average" U.S. coal.
..
**
_ , , , Projected pollutant concentration
Degree of hazard - Health-based MATE
*** Similar characteristics expected for SRC-I filter cake.
399
-------
Impacts on Land
SRC Product Utilization
Solid wastes of environmental concern, based
on SAM/IA analysis with the health-based
MATEs, are shown in Table 6. API separator
bottoms and biosludge from the wastewater
treatment system and SRC mineral residues
contain component pollutant species that ex-
ceed their MATE values. These solids are con-
sidered greater risks to the environment than
either SRC air emissions or water effluents.1
The mineral residue or filter cake produced
during solids/liquids separation in SRC-II and
SRC-I systems respectively contains high mo-
lecular weight organic species. It is recom-
mended that all such material be gasified to
render it safe for land or mine burial. Energy
recovered by gasification of excess residue can
be used onsite or sold as additional SNG prod-
uct.
A review of available analysis data on organ-
ics in SRC-II mineral residue indicated no or-
ganic species present in concentrations exceed-
ing the land-based health MATE value. Organ-
ics associated with SRC mineral residue are
shown in Table 7. Or games presently not as-
signed MATE values are also included in the
table.
With reference to the hazardous nature of
several SRC solid wastes, the following precau-
tions should be considered prior to disposal:1
• That the solids, singly or in mixture, should
be chemically stabilized.
• That the potential physical/chemical reac-
tions of sludges, singly or in mixture, should
be known.
• That the compatibility of the hazardous
waste with appropriate liners, sealants, and
container materials should be established.
• That the life span of the land disposal site
should meet the most stringent State stand-
ards (500 yr for the most hazardous wastes).
Toxic Substances hi Products
Polynuclear aromatic species detected in
analysis of light oil and solid SRC product pro-
duced in the SRC-I mode are shown in Table 8.
The variety of polynuclear species indicated in
the table illustrates the need to exercise care in
handling these materials.
To potential industrial and utility users, the
environmental benefits of using synfuels are of
primary concern. To date only two large-scale
tests have been conducted:3
(1) SRC-I, June 16-24, 1977, Georgia Power
Company's Plant Mitchell, Albany,
Georgia; and
(2) SRC-II, September 10-15,1978, Common-
wealth Edison's 74th Street Generating
Station, New York.
The SRC-I combustion test used 2,700 mg of
SRC-I material from a 3.9-percent sulfur coal.
No particular problems were experienced dur-
ing the 18-day test burn, and the following
levels of emission were achieved:
Constituent
Concentration,
ppm (vol)
Nitrogen oxides 175—300
Particulates 0.015—0.025
Carbon monoxide 50
Uncombusted hydrocarbons 3
Sulfur trioxide 1
The SRC-II combustion test used about 800 m3
of liquid SRC-II product produced at the pilot
plant in Fort Lewis, Washington. Reported lev-
els of emission are shown below:
Constituent
SOX
NOX
Particulates
Current EPA
requirements"1
0.52
0.30
0.04
SRC*
0.43
0.20
0.02
*UnitsareKg/GJ.
Data on sulfur dioxide concentrations were not
reported.
Based on these data, it appears likely that
SRC-II can be utilized in compliance with pro-
posed emissions standards for coal-derived liq-
uids.1
400
-------
TABLE 7. ORGANIC SPECIES PRESENT IN SRC-II MINERAL RESIDUE
MATE Value, jug/g
Organic Constituent Concentration, ug/g Health-based Ecological-based
indane
methylindane
dimethyl indane
tetralin
6-methyltetralin
naphthalene
2-methylnaphthalene
1-methylnaphthalene
dimethylnaphthalene
2-isopolynaphthalene
1-isopolynaphthalene
Cy-naphthalene
cyclohexylbenzene
biphenyl
acenaphthylene
dimethylbiphenyl
dibenzofuran
xanthene
dibenzothiophene
methyldibenzylthlophene
dimethyldibenzylthiophene
thioxanthene
fluorene
9-methylfluorene
1-methylfluorene
anthracene/phenanthrene
methylphenanthrene
1-methylphenanthrene
fluoranthrene
dihydropyrene
pyrene
85
40
25
110
50
1500
740
180
470
2
1
15
1
5
270
61
60
20
70
8
20
5
80
40
50
500
100
50
10
200
10
200
6.8x10;:
6.8x10
6.8xl05
4.0xl05
4.0x10"
5xl05
8xl05
200
200
6.8x10^
6.8X103
8xl05
6.8x105
6.8x10^
3000
1.7x10.
9.1x10*
9.1x10*
2.8xl05
401
-------
TABLE 8. POLYNUCLEAR AROMATICS DETECTED IN SRC-I
LIGHT OIL AND SRC SOLID PRODUCTS
Concentration, ppm (wt) MATE (water-based,
Organic Constituent Light Oil SRC Health-based Ecological-based
0-ethylbenzene 9800 6.5xl06
C3-benzene 3900 3.3xl06
indane 4300 6.8x105
methylindane 180-510 6.8xl05
dimethylindane <5 6.8x10^
tetralin 330 4.0x10;?
dimethyltetralin <5 4.0x10;?
6-methyltetralin 110 4.0x10;?
naphthalene 1630 1 1.5x10;?
2-methylnaphthalene 690 8 6.8x10;?
1-roethylnaphthalene 110 5 6.8x10;?
ditnethylnaphthalene 10-80 3-6 6.8x10
biphenyl 80 2 3000
acenaphthylene 2 8
dimethylbiphenyl 15-21 7-9
dibenzofuran 8 9
xanthene 10 5
dibenzothiophene 3 30
methyldibenzothiophene 4
dimethyldibenzothiophene 5 13
thioxanthene 3
fluorene 15 27
9-methylfluorene 15 11
1-methylfluorene 10 18 c
anthracene/phenanthrene 25 300 1.75x10"
methylphenanthrene 6 50 9.1x10,
1-methlyphenanthrene 6 30 9.1x10
C2~anthracene 6 1 r
fluoranthrene 15 180 2.8x10
dihydropyrene 6 1
pyrene 20 280 .6.9x10
l.OxlO3
l.OxlO3
200
200
200
402
-------
ADDITIONAL DATA REQUIREMENTS
AND RECOMMENDATIONS
Currently, the pilot plants at Fort Lewis,
Washington, and Wilsonville, Alabama, are the
most advanced SRC facilities in existence. Infor-
mation obtained during solvent-refining opera-
tions at Fort Lewis and Wilsonville is being
used to design SRC demonstration plants. In an
analogous manner, data from demonstration
plants will be used to permit successful commer-
cialization of SRC systems.
The draft EAR is based on the best existing
information, namely SRC pilot data, bench-scale
data, and conceptual design studies.1 Just as ad-
ditional operating data are required to commer-
cialize SRC systems, additional environmental
assessment data are necessary to adequately
characterize discharges, estimate environmen-
tal impacts, and evaluate control technology ap-
plicability relevent to SRC systems. Expansion
of the existing environmental assessment data
base for SRC systems should include the follow-
ing areas:
• SRC stream characterization: with the pur-
pose of developing representative physical,
chemical (inorganic and organic), and biologi-
cal (with bioassays) characteristics of SRC
plant streams, in particular before and after
treatment waste streams. While character-
ization of waste streams is essential to en-
vironmental assessment, better character-
ized process streams will permit construc-
tion of an advanced material balance, ideal-
ly permitting one to "track" pollutants
through the SRC system to the environ-
ment.
• Determination of the variability of waste
stream characteristics because of changes in
system operating characteristics: an ex-
panded data base on stream characteristics
may permit such correlations, possibly sug-
gesting ideal operating conditions for mini-
mized environmental effects.
• Performance evaluations and costs of ap-
plicable control technology alternatives.
• Reassessments of environmental impacts
based on the expanded data base.
Because of the relative applicability of SRC
pilot-plant data, the above efforts would be
more beneficial if performed at SRC demonstra-
tion facilities.
Environmental assessment methodologies
such as multimedia environmental goals (MEGs)
and source analysis models (SAMs) have been
developed to provide an organized, consistent
approach for evaluating emerging energy tech-
nologies such as SRC. Technically, there are
many differences between existing SRC pilot
facilities and the demonstration and commercial
plants of the future. Consequently, operating
data on process and waste stream character-
istics from the pilot plant are only an indication
of commercial or demonstration plant behavior.
However, sampling, analysis, and application
of environmental assessment methodologies to
pilot-plant data are essential to permit the fol-
lowing prior to emergence of SRC systems into
the commercial sector:
• Sampling and analysis techniques may be
tried and problem areas identified, thereby
permitting refinement of the techniques.
• Sampling and analysis priorities for the
demonstration/pilot SRC facilities may be
identified based on pilot studies.
• Application of the environmental assess-
ment methodologies to SRC pilot data will
allow additional development and evalua-
tion.
• Each of the above activities will accord SRC
system personnel with the expertise to con-
fidently assess commercial SRC systems at
the time technical progress and economic
conditions permit their emergence.
The following recommendations can be made
regarding future environmental assessments of
SRC systems:
• Efforts to characterize waste streams, proc-
ess streams, products, and byproducts
should be continued at an increased level of
effort. In so doing, numerous benefits are de-
rived including expanding the preliminary
data base on SRC systems, perfecting sam-
pling and analysis procedures, and develop-
ing more sophisticated environmental im-
pact methodologies. Results of these efforts
will be invaluable in establishing research
needs for environmental characterization of
SRC demonstration/commercial facilities.
• Efforts should be undertaken to define suit-
able sites for commercial SRC facilities. Sub-
sequent to definition, applicable sites should
be identified. Information required to per-
form site-specific environmental impact
analyses should be collected for those sites
identified as potentially suitable for SRC
403
-------
facilities, including preconstruction ambient
air and water quality monitoring. Initiating
expanded background monitoring studies in
applicable locations would be useful for en-
vironmental assessment and could hasten
construction of commercial facilities.
Candidate control technologies identified as
applicable to control of wastes from SRC
systems should be tested at SRC pilot and
demonstration facilities to the extent techni-
cally and economically feasible. Sampling
and analysis of discharge streams before
and after treatment would greatly expand
the environmental assessment data base.
Small-scale, slid-mounted control technology
units could be placed on flatbed trucks and
moved to pilot or demonstration facilities for
testing with continuous samples of the
plant's waste stream, thereby providing a
cost-effective means of performance testing
numerous candidate control options.
Continued efforts should be made to pro-
mote cooperation, coordination, and informa-
tion exchange between the various private
and government organizations involved in
development and environmental analysis of
SRC systems. Preparation and presentation
of technical papers at appropriate symposia
and meetings is an excellent way to infor-
mally stimulate interaction of researchers,
leading to more formal interaction during
performance of research. The benefits in-
clude reduced duplication of environmental
assessment efforts, permitting more effi-
cient use of available research funds.
ACKNOWLEDGMENTS
Information for this paper was compiled from
work performed by Hittman Associates, Inc.,
under EPA Contract Number 68-02-2162. This
work has been guided and supported by EPA's
Fuel Process Branch of the Industrial Environ-
mental Research Laboratory at Research Tri-
angle Park, N.C., under the direction of W. J.
Rhodes, Project Officer.
REFERENCES
1. Information obtained and developed by Hitt-
man Associates, Inc., During Preparation of
Draft Document, Environmental Assess-
ment of Solvent Refined Coal (SRC) Systems
for Producing Solid and Liquid Fuel From
Coal. EPA Contract Number 68-02-2162.
2. Rogoshewski, et al. Standards of Practice
Manual for the Solvent Refined Coal Lique-
faction Process. Hittman Associates, Inc. Re-
search Triangle Park, N.C. EPA 600/7-78-
091.1978.
3. Morris, J. W., and K. J. Shields. Environ-
mental Impacts of Coal Liquefaction. (Pre-
pared for National Conference on the Impact
of the National Energy Act on Utilities and
Industries Due to the Conversion of Coal.
December 4-6,1978.)
404
-------
COMBUSTION OF LIQUID SYNFUELS
G. Blair Martin* and W. Steven Lanier
Industrial Environmental Research Laboratory,
U.S. Environmental Protection Agency, Research Triangle Park, North Carolina
and
G. C. England, M. P. Heap, and D. W. Pershing
Energy and Environmental Research Corporation, Santa Ana, California
Abstract
This paper summarizes the available informa-
tion on the state-of-the-art emission control
technology for the use of petroleum-, shale-, and
coalrderived liquid fuels in stationary com-
bustion sources. Because the data on combus-
tion of alternative liquid fuels in practical
systems are limited, the properties of these fuels
are compared to those of petroleum-derived
fuels as a basis for postulating the effectiveness
of combustion process modifications on emis-
sions from alternative fuels. The formation and
control of nitrogen oxides are related to fuel
characteristics, particularly the distribution of
the fuel-bound nitrogen. The effectiveness of
staged combustion techniques is correlated
with a defined measurement of volatile nitro-
gen. The effect of fuel composition on carbon
paniculate formation is also discussed. Finally,
based on promising results for heavy petroleum
fuel oils and coal, it is concluded that burner and
combustion process design modifications have a
high probability of success for alternative fuels.
INTRODUCTION
In the search for energy supplies, the United
States is projected to place heavy reliance on
coal, which is the most abundant fossil fuel
available. Many methods of extracting the ener-
gy from coal are being pursued; however, the ul-
timate decisions on the paths to be followed de-
pend on both economic and environmental con-
siderations. These considerations cover the full
range from resource extraction, through proc-
essing, to end utilization. On the economic side,
it is necessary to include not only capital and
operating costs but also the overall energy effi-
•Speaker.
ciency of the process. On the environmental
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., com-
bustion systems) will be considered. The charac-
teristics of the combustion also influence the
route that will be chosen. For mobile sources
(e.g., automobiles and aircraft), light liquid
hydrocarbon fuels probably will be required for
a significant period in the future. In this in-
stance, the necessity for a specific fuel type may
overcome some of the other potential obstacles
(e.g., economics). For stationary sources, the
fuel used may not be constrained as signifi-
cantly by the requirement of a fuel of specific
characteristics, and the choice of approach may
be wider.
The ways in which coal can be used in an en-
vironmentally acceptable manner depend on the
type of combustion source. The pollutants that
must be controlled include sulfur oxides, nitro-
gen oxides, carbon monoxide, unburned hydro-
carbons, and total particulate. Perhaps the most
options exist for utility generation of electric
power. One option currently being used is the
direct combustion of coal with stack gas clean-
ing for sulfur oxides and particulate, and com-
bustion modifications for control of nitrogen ox-
ides, carbon monoxide, and unburned hydrocar-
bons. Improvement of the existing technology is
being pursued in a number of U.S. Environ-
mental Protection Agency (EPA) projects. A
second option is the conversion of coal into low-
sulfur gaseous, liquid, or solid fuels to be uti-
lized in conventional steam boilers or combined
cycle plants. The use of liquid fuels in power
generation appears to be most applicable to ex-
isting plants already burning petroleum-derived
heavy fuel oils. The large energy losses current-
ly associated with fuel-cleaning processes ap-
405
-------
pear to require use of the advance design com-
bined cycle with integrated gasifier to achieve
energy efficiency comparable to the first option.
Major unknowns in these designs are the crite-
ria for minimizing nitrogen oxides and other
combustion-related pollutants. The third option
is the use of fluidized-bed combustion to mini-
mize sulfur oxides and other pollutants. The
control of sulfur oxides has been a major con-
sideration in the development of all three tech-
nologies;1 however, other pollutants have been
considered less extensively for the latter two
options. For other stationary source applica-
tions, such as residential and commercial heat-
ing, low-sulfur high-Btu fuels will be required,
which may include distillate and/or residual
fuels derived from coal or shale. The purpose of
this paper is to summarize available information
on pollutant formation and control during com-
bustion of petroleum-derived liquid fuels as
related to synthetic liquids where combustion
data are much more limited. The effect of fuel
properties on emission control technologies is
also discussed.
BACKGROUND
A wide range of subject matter relates direct-
ly to combustion of alternate fuels. The topics
include pollutant formation mechanisms, appli-
cable emission-control techniques, fuel charac-
teristics, and end use equipment type. Since
these areas have been treated in detail for alter-
nate fuels previously,2 the background pre-
sented is a brief general summary. The most re-
cent information on combustion and emission
characteristics is summarized.
Pollutant Formation Mechanisms
The mechanisms of formation of nitrogen
oxides (NOX) have been discussed extensively;3 4
however, a brief summary is in order. (Nitric ox-
ide [NO] is the primary form of NOX found in the
flue gas of conventional combustion equipment;
the N02 that is present is believed to be the
product of oxidation of N02 after the combus-
tion process is completed.) The mechanisms for
formation of NO during combustion are as fol-
lows:
• Thermal NO is formed from fixation of at-
mospheric nitrogen by Zeldovitch reactions,
which have a strong temperature depend-
ence.
• Fuel NO is formed through oxidation of
chemically bound nitrogen in the fuel by
reactions with a low-temperature depend-
ence but a strong oxygen availability de-
pendence.
There is also experimental evidence5 to show
that nitrogen species (e.g., NH9 and HCN) can
be synthesized in fuel-rich flames as postulated
by Fenimore8 and subsequently oxidized to NO
as is fuel nitrogen. The other pollutants of con-
cern are SOX, CO, hydrocarbons, POM, carbon
particulate, and metallic particulate. One of the
primary incentives for alternate fuels is sulfur
removal; therefore, SOX levels should be low.
Since proper system designs for stationary
sources can minimize CO and hydrocarbon emis-
sions, no problem is anticipated with alternate
fuels. Carbon particulate emissions for heavy
liquid fuels pose a potential problem that may
be complicated further by the higher carbon-to-
hydrogen ratios of many synthetic liquid fuels.
Metallic particulate is dependent primarily on
the mineral content of the fuel and, therefore,
on the extent of coal ash removal during fuel
processing. Metal form and particle size distri-
bution also may be affected by the combustion
process; however, no detailed information is
presently available. Use of alternate fuels in
combustion systems will require careful design
to minimize these emissions.
Emission-Control Techniques
The basic combustion modification techniques
for NO control can be summarized as follows:
• Diluent addition to reduce flame tempera-
ture is accomplished through the addition of
either water or recycled flue gas to the com-
bustion air.
• Staged combustion is based on operation of
burners at a fuel-rich condition with delayed
secondary air addition to complete heat re-
lease, thereby limiting both peak flame tem-
peratures and primary zone oxygen avail-
ability.
• Burner modifications involve changes in fuel
and air mixing conditions to promote local-
ized fuel-rich conditions and/or combustion
gas recirculation.
• Novel techniques, such as catalytic combus-
406
-------
tion, may allow NO emissions lower than
those achievable for combustion of clean
fuels in conventional systems and may be
particularly applicable to redesign for main-
taining system efficiency.
The first technique controls only thermal NO,
whereas the last three also may control fuel NO.
The emissions of the products of incomplete
combustion (CO, unburned hydrocarbons, and
carbon particulate) are subject to increase as
NO is decreased past a critical point for fixed
system design. However, there is a body of evi-
dence that indicates that these emissions can be
controlled if the system is designed or modified
with both NO and carbonaceous emissions-con-
trol requirements in mind. Since stack measure-
ments are for nitrogen oxides (NOX), that term
will be used in subsequent discussion of control
techniques.
Fuel Characteristics
The properties of alternate fuels have been
summarized previously,2 and only a brief up-
dated discussion is presented below.
Synthetic liquids may be grouped into two
general categories: Those synthesized from the
products of coal gasification, and those derived
directly as liquids by hydrogenation of coal or
by retorting of oil shale. The fuels in the first
category tend to be clean low-boiling fuels such
as alcohols and Fischer-Tropsch liquids. Since
these fuels are also essentially free of both ni-
trogen and sulfur, combustion problems are
minimal. The liquids in the second category may
be compared to crude petroleum oils because
both consist of a wide range of hydrocarbon
compounds with boiling points from 300 to over
900 K. In the crude synthetic liquid fuels, the
bound nitrogen content is generally quite high
(more than 0.5 percent). In addition to this, the
nitrogen is distributed more evenly over the
range of fuel cuts than it is in crude oil. The
most complete information is available on a
2.19-percent nitrogen Paraho shale crude, as
shown in Figure I.7 The nitrogen content is
above 1.2 percent by weight for all fuel fractions
shown here. The sulfur levels are below 1 per-
cent and decrease in the higher boiling frac-
tions. For comparison, a Wilmington, California,
crude8 that contained 0.65 percent nitrogen
yielded only 0.07 percent in the distillate oil
product (corresponding in boiling point to the
33-percent volume fractions of the shale crude).
The nitrogen and sulfur can be removed to low
levels by hydrotreating;9 however, it is an ex-
pensive process. Where other considerations
are paramount (e.g., fuel stability for aircraft
uses), severe hydrotreating may be unavoid-
able. For other applications, it may be possible
to achieve NOX control by combustion modifica-
tion of fuels with minimum hydrotreating to up-
grade the carbon-to-hydrogen ratio for smoke
suppression or to modify the nitrogen com-
pounds to more volatile forms without substan-
tial denitrifica'tion.
In a recent review of commercialization po-
tential of coal-derived liquid fuel processes,10
Whitaker summarized the Electric Power Re-
search Institute's view of coal-derived liquids.
The discussion dealt with three processes: sol-
vent refined coal (SRC-II), H-Coal, and Exxon
Donor Solvent (EDS). While the article indi-
cated that the fuel properties would probably
depend both on the process and on the coal feed-
stock, it did not specify these properties.
DISCUSSION
For all combustion systems (except recipro-
cating engines, which are not a subject of this
paper), the common feature is the use of a
burner for initial fuel and air mixing. Although
the characteristics of specific systems signifi-
cantly differ, the combustion zone conditions
that lead to minimum emissions are quite simi-
lar for two broad classes of fuels (i.e., nitrogen-
free and nitrogen-containing). A large body of
information has been built up on burner designs
for emission control for conventional fuels,
much of which should apply directly to systems
burning alternate fuels.
The primary emission category that is con-
trollable by combustion technology is NOX. Con-
trol of carbonaceous emission (e.g., CO, hydro-
carbon, POM, and carbon particulate) also is af-
fected by combustion technology; however, it
may be treated as a second-order effect, except
for gas turbines operating at low load. This is
not based on establishing priorities for health or
environmental effects of the pollutants but
rather on the approaches necessary to control
all emissions by combustion technology. Many
conventional design approaches are currently
used that offer the potential for low car-
bonaceous emissions by employing conditions
407
-------
2.4
2.2
2.0
1.8
1.6
1.4
BO E
£ I 1.2
CE 5
= o
_
< <
I- I-
O O
1.0
no
0.8
0.6
0.4
0.2
NITROGEN
SULFUR
PARAHO CRUDE SHALE OIL
10 20 30 40
CUM. (MID-VOLUME) DISTILLATION FRACTION (percent)
Figure 1. Total weight of nitrogen and sulfur as a function of
the cumulative midvolume distillation fraction.12
SO
408
-------
that lead to high levels of NOX. Therefore, it is
necessary to approach the problem from the
other direction; that is, to employ the special-
ized design concepts that give low levels of NOX
and optimize that technology to achieve
minimum carbonaceous emissions. In this way,
optimum control of all emissions becomes a
primary design criterion and a goal that can be
achieved during the development of the com-
bustion process for a specific application.
The following discussion identifies key sys-
tem features that relate to emission control for
these two general classes of fuels. Emphasis is
placed on nitrogen-containing fuels.
Nitrogen-Free Fuels
Fuels that do not contain chemically bound
nitrogen produce only thermal and "prompt"
NOX, for which there is a substantial body of
control technology already developed. For sta-
tionary heat and steam generation systems, the
primary techniques are external flue gas recir-
culation and burner designs that maximize in-
ternal recirculation of relatively cool combus-
tion products. The burner techniques can be
coupled with combustion chambers designed to
achieve early heat removal, thereby further
reducing peak temperature and optimizing the
NOX reduction. These techniques are compati-
ble with low carbonaceous emissions and with
low excess air operation for maximum system
thermal efficiency. For gas turbines, a number
of approaches are being explored. Substantial
effort has been devoted to achieving premixed
prevaporized primary combustion zones, which
can be operated at conditions giving lower
flame temperatures (e.g., fuel lean) and,
therefore, lower thermal NOX. These concepts
also produce low carbon particulate levels, but
may produce excessive CO, particularly over
the full operating load range of the engine. This
technique is well suited to the gas turbine that
normally operates at high excess air levels (i.e.,
300 to 400 percent). A major consideration for
this concept is burner stability. These relatively
conventional technologies are well documented
and do not require further discussion.
Nitrogen-Containing Fuels
Nitrogen compounds chemically bound in the
fuel are oxidized to form what is termed fuel
NOX. This is a significant concern for alternate
fuels because virtually all untreated coal- and
shale-derived fuels have large concentrations of
bound nitrogen species. Before processing, the
liquid crudes derived from coal and shale have
more than 0.5 percent nitrogen. Because sulfur
levels are generally below 1 percent, minimum
hydrotreating is desirable to limit efficiency and
economic penalties. For coal-derived solid fuels
(e.g., SRC-I), the nitrogen compounds are not
removed to a significant degree by the fuel con-
version processes. For both conventional solid
and liquid fuels, the nitrogen is bound within
the fuel structure as single or multiple heter-
ocyclic ring compounds, and a similar structure
is believed to exist in the alternate fuels.
In addition to the absolute amount of nitrogen
contained in the fuel, it appears that the degree
of control achievable may also depend on nitro-
gen distribution. The evidence indicates that
the nitrogen in the fuel is converted to simple
gas-phase species (HCN and NH3) before it is ox-
idized to NO or reacts to form N2. The
heterocyclic nitrogen compound in the parent
fuel appears to undergo a sequential pyrolysis
through lighter organic forms to HCN or NH3.
The extent of this pyrolysis depends on temper-
ature, residence time, and ambient conditions
(oxidizing or reducing). Although it has been
shown that a quantitative conversion of pyri-
dine (C5H5N) to HCN can be achieved at 1,373 K
under inert conditions,11 comparable conver-
sions have not been shown for any real fuel at
residence times achievable in practical com-
bustors, even at considerably higher tempera-
tures. The balance of the nitrogen is contained
in the fuel residue that may be char or tar. The
nitrogen evolved into the gas phase is referred
to as "volatile nitrogen." The significance of this
distribution of nitrogen compounds is discussed
in greater detail below.
The basis of fuel NOX control techniques is
the same regardless of the fuel type. A fuel-rich
primary combustion zone, is used to facilitate
the conversion of fuel nitrogen to molecular ni-
trogen (N2). A fraction of the nitrogen is evolved
as XN species (e.g., HCN and NH3), which par-
tially oxidize to form NO. The NO then reacts
with the residual SN to form N2. Because XN
species remaining in the rich mixture undergo
high-efficiency conversion to NO in the lean
secondary stage and because any NO will be re-
tained almost quantitatively, the rich-zone con-
409
-------
ditions must give a minimum value of £XN (i.e.,
HCN + NH3 + NO). The stoichiometry re-
quired to achieve minimum LXN depends on
several factors, including:
• The rate of evolution of nitrogen species
from the fuel;
• The inevitable distribution of stoichiom-
etries from rich to lean, which exists in an
overall fuel-rich zone of a diffusion flame;
• The overall temperature of the reaction
zone; and
• The overall residence time in the reaction
zone.
The interaction of these four factors depends on
the aerodynamic mixing of a turbulent diffusion
flame, as well as the nitrogen distribution for a
given fuel. It is desirable to extract some ener-
gy from the rich products prior to second-stage
air addition to reduce the thermal NOX forma-
tion. In the lean second stage, a significant frac-
tion of the gaseous EXN and a smaller fraction
of any residual nitrogen in the char or tar will
be converted to NOX. Based on evidence for coal
char and petroleum coke, the conversion of this
nonvolatile nitrogen to NO occurs at a low frac-
tion efficiency (i.e., less than 10 percent) for solid
fuels. In fact, the NOX levels from these fuels
are insensitive to burner design changes that
significantly reduce NOX from pulverized coal.
This char NOX may impose a minimum level
below which NOX cannot be reduced for a given
primary zone condition. There are indications
that the nonvolatile nitrogen species from liquid
fuels may undergo higher fractional conversion
to NOX than those in solid fuels.
Emission Performance: Boilers
Primarily because of the small quantities of
synthetic liquid fuels available, the data on their
combustion emissions and performance in prac-
tical systems are limited. Therefore, this discus-
sion will review the available data from experi-
mental systems and compare the performance
of synthetic liquids to that of conventional
petroleum fuels. The data on some of the earlier
work have been summarized previously;12 only
an update is presented below.
Blazowski and Maggitti13 believe that the
alternative fuel characteristics that are most
likely to affect future gas turbine design are the
hydrogen and nitrogen content and the thermal
stability. Carbon-to-hydrogen ratio influences
soot formation, which leads to increased flame
emissivities increasing liner temperatures and
smoke emissions. Fuel-bound nitrogen in jet
fuels contributes to higher NOX emissions
unless the combustor is designed to minimize.
fuel nitrogen conversion to NOX.
Increased NOX emissions have also been
observed when coal- or shale-derived liquids
have been burned in boilers. Muzio14 carried out
tests with SRC-II and found that its combustion
characteristics were similar to those of No. 2
fuel oil except that NOX emissions were higher
(400 ppm compared to 80 ppm) mainly because
the SRC-II contained 1.12 percent nitrogen.
However, emissions can be reduced by blending
with lower nitrogen petroleum-derived fuels or
staged combustion. Similar experiences are re-
ported by Mansour15 when Paraho shale oil was
burned.
The most comprehensive comparison of emis-
sion characteristics to fuel properties has been
carried out by Heap and coworkers.18 A variety
of petroleum-derived residual oils and synthetic
fuels derived from coal and shale have been
burned in two experimental systems; a small
(20 kW thermal) down-fired tunnel17 and a
900-kW cold wall axisymmetric combustor that
simulates the firetube of a package boiler.18 The
down-fired combustor allowed direct determina-
tion of fuel nitrogen conversion by substitution
of argon/oxygen for the combustion air, thereby
eliminating thermal NOX. The package boiler
simulator allows the smaller scale results to be
generalized to practical equipment. The results
of these experiments provide significant insight
into pollutant control for both petroleum-de-
rived and synthetic fuels. The properties of the
petroleum oils have been summarized previous-
ly,17 and properties of the synthetic fuels are
shown in Table I.16 NOX emissions for all fuels
tested in th*e tunnel furnace are summarized in
Figure 2. Since the data are for a system where
very fine oil droplets (about 25 ion) are well
dispersed in the oxidizer at a fuel-lean condition,
it is not surprising that the fuel NOX emissions
(lower curve) are a strong function of fuel
nitrogen content. The NOX levels are high
because the percentage conversion under these
premixed conditions is higher (50 to 75 percent)
than expected in practical systems (25 to 45 per-
cent). The upper curve shows that thermal NOX,
which is determined using air as the oxidizer, is
relatively constant for most fuels. For some
410
-------
TABLE 1. ALTERNATIVE LIQUID FUEL PROPERTIES18
Ultimate Analysis
Carbon, %
Hydrogen, %
Nitrogen, %
Sulfur, %
Conradson Carbon Residue, %
Asphalt ene, %
API Gravity at 60°F
Viscosity SSU at 140°F
Gross Heat of Combustion, Btu/lb
DFM
86.18
13.00
0.24
0.51
4.1
0.036
33.1
36.1
19,430
SRC-II
Blend
89.91
9.27
0.45
0.065
6.18
4.10
10.0
40.6
17,980
Shale
Derived
Residual
86.71
12.76
0.46
0.038
0.19
0.083
29
54.3
19,350
SRC-II
M:H Dist.
85.91
8.74
0.97
0.30
0.51
-
11
-
-
Synthoil
86.30
7.44
1.36
0.80
23.9
16.55
-
10,880
16,480
Paraho
Shale
84.6
11.3
2.08
0.63
2.9
1.33
-
97
18,290
*Paraho Diesel Fuel Marine
-------
2000
1600
oc
Q
O*
g 1200
O
z
800
400
1
-&i
o/ x
V cc'
D 0^
/,
1 1 1 1 1 1 1 1
xx«
X'x
/ x^
•x x*
V X
/ • FUEL NOX
TOTAL NOX/ X
/^bo^X^
xl^
3' QX O PETROLEUM DERIVED
>r D DFM
jAQ A SRC II BLEND
^P A SHALE RESIDUAL
/ ^ SRC II
• SYNTHOIL/BLENDS
• PARAHO SHALE/BLENDS
1 1 1 1 1 1 1
0.4
0.8 1.2
WEIGHT % NITROGEN
1.6
Figure 2. The effect of fuel nitrogen content on total and
fuel NOX (5 percent excess oxygen).16
2.0
412
-------
unexplained reason, the alternative fuels ap-
pear to produce a somewhat higher thermal
NOX level than the petroleum fuels. This points
out the need for NOX control techniques for the
alternative fuels.
To examine the effects of control technolo-
gies, Heap16 also ran the tunnel furnace under
staged conditions. The total NOX data for the
Paraho crude shale oil, shown in Figure 3 at two
primary residence times, indicate that very
high levels of control (90 to 95 percent) can be
achieved at reasonable primary stage stoichiom-
etries (i.e., 70 to 80 percent theoretical air).
Figure 4 compares the results of the shale crude
to a residual liquid from the same crude that has
been extensively hydrotreated, SRC-II and a
blend of 8RC-II with the donor solvent. While
the uncontrolled levels are substantially dif-
ferent, the minimum levels under staged condi-
tions are quite similar. It is interesting to note
that the minimum NOX for the 2.08-percent
nitrogen shale crude is lower than for the
0.97-percent nitrogen SRC-II, a result to be
discussed at greater length. A comparison of
results in the tunnel and in the package boiler
simulator is shown in Figure 5. Although condi-
tions were maintained as consistently as possi-
ble between the systems, the minimum NOX
from the tunnel furnace is significantly lower
than conditions for the simulator using the same
type of ultrasonic atomizer (curve B). This might
be attributable to differences in a number of
primary zone factors including amounts of wall
cooling affecting the rate of nitrogen evolution
in the primary zone; fuel/air mixing rates creat-
ing wider distribution of off-optimum stoichiom-
etries in the boiler simulator; or control of
residence time for secondary air addition be-
cause of recirculation patterns. Comparison of
curves A and B for different nozzles in the
boiler simulator shows that nozzle A, which pro-
duces a coarser spray than B, has lower baseline
emission (primary zone stoichiometric ratio of
1.17) but higher emissions under staged condi-
tions. This points out the importance of optimiz-
ing the combustion system for minimum emis-
sions.
These results and others suggested that
under staged conditions, factors other than total
percent nitrogen affected the minimum attain-
able emissions. A comparison of minimum NOX
under staged conditions for synthetic and petro-
leum-derived fuels is shown in Figure 6. For
many fuels with approximately the same fuel ni-
trogen level (0.4 to 0.6 percent) a significant
spread exists. The minimum is a nitrogen- and
sulfur-doped distillate fuel where all of the
nitrogen is volatile (i.e., has a boiling point of
about 400 K), and the maximum is the SRC-II
blend. It should also be noted that at fuel nitro-
gen levels above 0.6 percent, there is only a
small increase in the minimum achievable NOX.
In an attempt to relate the effects of fuel prop-
erties to emissions, a bench-scale vacuum distil-
lation technique was selected as a relatively
simple and rapid method of quantifying the
amount of "volatile" nitrogen in the fuel. Each
fuel was distilled into as many as five fractions,
and the total mass of oil and nitrogen content of
each fraction was determined. The data for pe-
troleum oils have been presented previously by
Pershing.19 The results for the specific alter-
native fuels tested are compared in Figure 7 to
those for the range of petroleum oil. The shaded
area shows the extremes of individual residual
oils from less than 10 percent of the nitrogen
evolved at 811 K (1,000° F) to over 40 percent at
the same temperature. By comparison, all of the
synthetic fuels show greater than 40 percent
evolved at 700 K (800° F). It is particularly in-
teresting to note that even the "residual" de-
rived from a highly hydrotreated Paraho crude
has bound nitrogen more volatile than in any pe-
troleum-derived residual.
These data were used by Heap" to correlate
the effectiveness of staged combustion vs. vola-
tile nitrogen for various fuels, as shown in Fig-
ure 8 (which uses the same symbols as previous
figures). The ratio of NOX staged to unstaged,
which represents the fraction not controlled by
staged combustion, increases as the nitrogen
volatility decreases. The dotted lines are for the
tunnel furnace at two primary residence times,
where the lower line is the longer residence
time, and the solid line is the package boiler
simulator. This figure indicates that for a given
system the nitrogen volatility has a strong ef-
fect on the degree of NO, control achievable;*
however, the system design also is a significant
factor.
Emission Performance—Turbines
The information available on combustion of
synthetic liquid fuels in gas turbines has been
for baseline combustors without NOX control
413
-------
2000
1600
£ 1200
ct
o
o*
o
800
400
FIRST STAGE RESIDENCE TIME
AT 70 PERCENT THEORETICAL AIR
O 0.83 SEC
A 1.68 SEC
60
80 100
FIRST"STAGE THEORETICAL AIR, %
120
Figure 3. The influence of staged operation in the tunnel furnace on NOX emissions from
the Paraho shale oil (5 percent excess oxygen).16
414
-------
800
A SRC II BLEND
A SHALE RESIDUAL
• SRC II
• PARAHOSHALE
600
tt
0
~ 400
200
I
60
70 . 80
FIRST STAGE THEORETICAL AIR, %
90
Figure 4. Minimum NOX levels achieved with alternative fuels (tunnel furnace primary
zone residence time 0.83 sec).16
415
-------
1800
1400
cc
O
cT 1000
1
i
%
x
O
600
200
O A ^1
OB J
PACKAGE BOILER
SIMULATOR
A TUNNEL FURNACE
0.7
0.8 0.9 1.0
PRIMARY ZONE STOICHIOMETRIC RATIO
1.1
Figure 5. NOX emissions staged-Paraho shale (5 percent overall excess oxygen).10
416
-------
200
oc
Q
O
s
i
z
I
100
O PETROLEUM DERIVED
A SRCII/BLEND
A SHALE RESIDUAL
^ SRC II
• PARAHO SHALE
O PETROLEUM DISTILLATE PLUS
PYRIOINE AND THEOPHENE
0.4
0.8 1.2
WEIGHT % NITROGEN
1.6
2.0
Figure 6. Minimum NOX levels obtained in the tunnel furnace as a function of
fuel-bound nitrogen content.16
-------
50
D
A
DFM
SHALE RESIDUAL
PARAHOSHALE
SRC II BLEND
SYNTHOIL
40
Q
£ 30
_i
O
tit
Z
ui
O
O
GC
Z
s?
20
10
0
400
RANGE FOR
PETROLEUM ~~
DERIVED FUELS
I
600 800 1000
DISTILLATION TEMPERATURE °F
Figure 7. Fuel nitrogen volatility as determined by vacuum distillation.16
1200
418
-------
0.4
0.3
u
O
<
0.2
0.1
O PETROLEUM DERIVED
• PARAHOSHALE
A SRC II BLEND
A SHALE RESIDUAL
PACKAGE
BOILER
SIMULATOR
xv*-
TUNNEL FURNACE
O
.0*^
0.4
0.6 0.8
FRACTION OF NITROGEN IN RESIDUE
1.0
Figure 8. The influence of volatile fuel nitrogen on staging effectiveness (reference 16).
419
-------
technology applied. There are indications that
the wet control techniques designed to control
thermal NOX, such as water injection, will have
little beneficial effect on fuel NOX, and possibly
will be detrimental.
A promising low NOX combustor concept for
gas turbine engines has been reported by
Pierce.80 The program goals were to achieve
NOX emissions below 50 ppmv (at 15 percent 02)
for clean fuels and 100 ppmv (at 15 percent 02)
for fuels with less than 0.5 wt percent bound ni-
trogen. The bench-scale version of the combus-
tor, which burns a premixed, prevaporized, fuel-
rich fuel and air mixture, followed by rapid addi-
tion of secondary air to avoid high peak temper-
atures during fuel burnout, has achieved mini-
mum emission levels of 20 ppmv (at 15 percent
02) for No. 2 fuel oil and 35 ppmv (at 15 percent
02) for No. 2 fuel oil doped with 0.5 wt percent
nitrogen as pyridine.21
Based on the previous discussion on boiler ex-
perience, it may be expected that the nitrogen-
doped oil would provide a good indication of the
control potential for distillate synthetic liquid
fuels, although the NOX level for the higher boil-
ing nitrogen compounds in these fuels may be
somewhat higher. It also appears that the pre-
mixed and prevaporized nature of the primary
zone should provide the maximum opportunity
for minimizing fuel NOX. The bench-scale com-
bustor has been scaled up to the size of a single
can for a practical engine, and preliminary
testing appears to show similar performance.
Practical Implications
Based on the above discussion of the combus-
tion characteristics of synthetic liquid fuels,
some generalizations about system design and
fuel properties are possible.
System Design—
There are some obvious- differences in the
emission performance between the tunnel fur-
nace and the package boiler simulator18 that
cannot be fully explained at this time; however,
speculation is possible if we adopt the volatile
nitrogen hypothesis discussed earlier. It may be
restated as follows:
• That nitrogen species should be evolved as
early as possible in the fuel-rich primary
zone to allow maximum possible reaction of
nitrogen species (XN) to N2; and
• That residual XN not reacted in the first
stage will oxidize to NOX with a relatively
high conversion in the fuel-lean second
stage.
The conditions that appear to favor maximum
N2 formation include high temperature to
evolve nitrogen species as early as possible and
longer residence times. There is also an indica-
tion that an atomizer that yields small droplets
that are well dispersed in the airstream im-
proves the degree of control achieved; some
doubt exists that a completely prevaporized
fuel, premixed with air, is desirable, even if
possible. While it is obvious that prevaporiza-
tion is not possible for petroleum-derived
residual oils, it may be possible for many of the
synthetics (e.g., the shale residual tested). A
directly comparable test is required to deter-
mine if a well-dispersed spray of small droplets
with combustion in a diffusion flame over the
range of stoichiometries is superior to premixed
combustion of a vaporized fuel at a single stoi-
chiometry.
The degree of success achieved in controlling
NOX from any given fuel will depend on integra-
tion of the fuel atomizer, the air mixing device,
and the primary zone thermal environment
(e.g., cooled or refractory). While the available
information is encouraging, additional work is
necessary to optimize emissions for systems
burning heavy liquids that cannot be completely
vaporized.
Fuel Properties-
Compared to petroleum residual oils, the syn-
thetic fuels tested to date appear to have a
larger fraction of the nitrogen bound in low-boil-
ing fuel fractions' and, therefore, to be more
amenable to NOX control technology. The main
problem with this conclusion is that the fuels
are not directly comparable. That is, the petro-
leum residual fuels are the heaviest ends of the
crude that contain most nitrogen of the refrac-
tory compounds, whereas most of the synthetic
fuels should be regarded as crudes. (Note that
the one exception, the Paraho residual, results
from distilling a heavily hydrotreated crude.) In
actual practice it would probably be desirable to
distill the synthetic crude, using the lighter
fractions for jet fuels and distillate oils, thereby
leaving the heavier fractions for boiler fuels.
While it may be argued that such a synthetic
residual would still contain substantially less
420
-------
refractory nitrogen compounds than petroleum
residual (see Figure 7), such a heavy synthetic
must be tested to determine its performance.
The second aspect that remains to be estab-
lished is the need for hydrotreating the various
synthetic fractions. For the lighter jet and
distillate fuel fractions, substantial removal of
nitrogen compounds is apparently required to
enhance storage stability. One approach is to
hydrotreat the full crude prior to distillation;
however, an alternative is to distill the light
fractions and then hydrotreat to remove nitro-
gen to required levels. The primary decision
here would probably be based on an economic
tradeoff of the smaller fraction of the crude bar-
rel available as premium fuel vs. the cost of
heavy hydrotreating the full crude.
If one assumes that an unhydrotreated resid-
ual containing a significant nitrogen content
(e.g., more than 1 percent) is to be used as a
boiler fuel, a second question of extent of hydro-
treating must be addressed. Assuming that the
data shown in Figure 6 prove applicable to prac-
tical systems, the economic impact of hydro-
treating from 2.08 percent nitrogen to about 0.4
percent must be balanced against a 25-percent
reduction of NOX (i.e., 200 vs. 150 ppm NOX, re-
spectively). However, if the true untreated
shale residual produces substantially more NOX
than the crude for a comparable nitrogen con-
tent, yet another tradeoff may be possible. The
extreme case is to deeply hydrotreat the crude
and achieve the relatively low nitrogen residual,
with the attendant potential increase in distil-
late fraction, or to mildly hydrotreat the resid-
ual fraction simply to upgrade the nitrogen into
a more volatile form without substantial denitri-
fication. In either case it might be expected that
the smoke-forming tendencies of the fuels would
be decreased by hydrotreating, which might
provide yet another consideration in the deci-
sion process.
Based on the current state of knowledge, it is
not possible to draw a firm conclusion about fuel
processing. Careful experimental work on fuels-
of specific properties is required.
CONCLUSIONS
The data on combustion of synthetic liquids in
practical combustion equipment are still very
limited. From the available data from experi-
mental apparatus and comparison of emission
characteristics of synthetic liquids to petro-
leum-derived fuels, the following conclusions
can be drawn:
• Under fuel-lean conditions, the nitrogen con-
tent of the fuel is the dominant factor for a
given system. Conversion to NOX is similar
for petroleum and synthetic liquids, with
fractional yield NOX decreasing with in-
creased nitrogen content.
• Under staged combustion conditions, the
volatility of the fuel nitrogen compounds is
an important factor in the degree of control
achievable, and the absolute level of NOX at-
tainable may be a weak function of fuel ni-
trogen content.
• System design is a significant factor in the
degree of control achievable with a given
fuel. The key variables appear to be primary
zone stoichiometry, residence time, and
thermal environment. The methods of atom-
ization and of air/fuel mixing strongly in-
fluence the performance of a practical sys-
tem.
• The experimental results appear to have im-
portant implications for fuel treatment
strategies, particularly denitrification; how-
ever, experiments with a wider range of
fractions from a common crude are required.
METRIC CONVERSION
While it is EPA policy to use metric units, the
nonmetric data used in this paper have been se-
cured from published literature and have not
been altered. Metric conversion can be accom-
plished with the following factors:
°C - 579(°F-32)
J/g - Btu/lb x 2.326
REFERENCES
1. Economic Commission for Europe, Second
Seminar on Desulfurization of Fuels and
Combustion Oases. Washington, D.C. No-
vember 11-20,1975.
2. Martin, G. B. Environmental Considera-
tions in the Use of Alternate Fuels in Sta-
tionary Combustion Processes. In: Sym-
posium Proceedings: Environmental As-
pects of Fuel Conversion Technology.
EPA-650/2-74-118, NTIS PB 23&304/AS.
October 1974. p. 259-276.
3. Brown, R. A., H. B. Mason, and R. J.
Schreiber. Systems Analysis Require-
421
-------
ments for Nitrogen Oxide Control of Sta-
tionary Sources. Aerotherm-Acurex Corp.
EPA-650/2-74-091, NTIS PB 237-367/AS.
September 1974.
4. Martin, G. B. Overview of U.S. Environ-
mental Protection Agency's Activities in
NOX Control for Stationary Sources. In:
Joint t/.S.-Japan Symposium on Counter-
measures for NOr Tokyo, June 28-29,
1974.
5. Yamagishi, K., M. Nozawa, T. Yoshie, T. To-
kumoto, and Y. Kakegawa. A Study ofNOx
Emission Characteristics in Two-Stage
Combustion. (Presented at 15th Sympo-
sium [International] on Combustion. Tokyo.
August 1974.)
6. Fenimore, C.P. Formation of Nitric Oxide
in Premised Hydrocarbon Flames. (Pre-
sented at 13th Symposium [International]
on Combustion. Salt Lake City. August
1970.)
7. Bartik, H., K. Kunchal, D. Switzer,
R. Bowen, and R. Edwards. Final Report:
The Production and Refining of Crude
Shale Oil into Military Fuels. Applied
Systems Corp. Office of Naval Research
Contract N00014-75-C-0055. August 1975.
8. Ball, J. S., and H. T. Rail. Nonhydrocarbon
Components of a California Petroleum. In:
27th Midyear Meeting of the American Pe-
troleum Institute Division of Refinery. San
Francisco, 1962.
9. Dzuna, E. R. Combustion Tests of Shale
Oils. In: Central States Section of the Com-
bustion Institute. Columbus, April 5-6,
1976.
10. Whitaker, R. Scaling up Coal Liquids.
EPRIJournal 3(7W-13. September 1978.
11. Axworthy, A. E., and V. H. Dayan. Chemi-
cal Reactions in the Conversion of Fuel Ni-
trogen to NOX: Fuel Pyrolysis Studies. In:
Proceedings of the Second Stationary
Source Combustion Symposium, VoL IV:
Fundamental Combustion Research.
EPA-600/7-77-073d, NTIS PB 274-029/AS.
July 1977. p. 39-78.
12. Martin, G. B. NOX Considerations in Alter-
nate Fuel Combustion. In: Symposium Pro-
ceedings: Environmental Aspects of Fuel
Conversion Technology. EPA-600/2-76-149,
NTIS PB 257-182/AS. June 1976. p. 373-
394.
13. Blazowski, W. S., and L. Maggitti. Future
Fuels in Gas Turbine Engines. In: Progress
in Astronautics and Aeronautics VoL 62,
Alternative Hydrocarbon Fuels: Combus-
tion and Chemical Kinetics, Bowman, C. T.,
and Birkeland, J. (ed.). American Institute
of Aeronautics and Astronautics, 1978.
14. Muzio, L. J., and J. K. Arand. Small Scale
Evaluation of the Combustion and Emis-
sion Characteristics of SRC Oil (Paper Pre-
sented at the American Chemical Society
Fuel Chemistry Symposium on Combustion
of Coal and Synthetic Fuels. Anaheim.
March 1978)
15. Mansour, M. N., and D. G. Jones. Emission
Characteristics of Paraho Shale Oil as
Tested in a Utility Boiler. EPRI Report
Number AF-709. March 1978.
16. Heap, M. P., G. C. England, and D. W.
Pershing. Emission Characteristics of
Alternative Liquid Fuels. (Presented at the
Institute of Gas Technology Symposium on
New Fuels and Advances in Combustion
Technology. New Orleans. March 26-30,
1979.)
17. Heap, M. P., et al. The Influence of Fuel
Characteristics on Nitrogen Oxide Forma-
tion—Bench Scale Studies. In: Proceedings
of the Third Stationary Source Combustion
Symposium, Vol II. EPA-600/7-79-050b,
NTIS PB 292-540/AS, February 1979. p. 41-
71.
18. England, G. C., et al. The Control of Pollu-
tant Formation in Fuel Oil Flames—The In-
fluence of Oil Properties and Spray Charac-
teristics. In: Proceedings of the Third Sta-
tionary Source Combustion Symposium,
Vol. II. EPA-600/7-79-050b, NTIS PB
292-540/AS, February 1979. p. 3-39.
19. Pershing, D. W., G. C. England, M. P. Heap,
and G. Flament. Control ofNOx from Liq-
uid FuelrFired Industrial Boilers. (Pre-
sented at the 86th National AIChE Meet-
ing. Houston. April 1-5,1979.
20. Pierce,*R. M., C. E. Smith, and B. S. Hinton.
Low NOX Combustor Development for Sta-
tionary Gas Turbine Engines. Proceedings
of the Third Stationary Source Combustion
Symposium, VoL III. EPA-600/7-79-050c,
NTIS PB 292-541/AS. February 1979.
p. 137-173.
21. Pierce, R. M. Advance Combustor System
for Stationary Oat Turbine Engines.
Monthly Progress Report No. 23 on Con-
tract 68-02-2186. December 1977.
422
-------
Session III: ENVIRONMENTAL CONTROL
Robert P. Hangebrauck, Chairman
Industrial Environmental Research Laboratory,
U. S. Environmental Protection Agency
Research Triangle Park, North Carolina
423
-------
CONTROL ASSAY SCREENING PROCEDURES
William F. Longaker,* Alfred B. Cherry, and Sohrab M. Hossain
Catalytic, Inc., Philadelphia, Pennsylvania
Abstract
Control assay (CA) screening procedures are a
significant and important part of the U. S. En-
vironmental Protection Agency's (EPA) overall
data acquisition program for environmental
assessment of fuel conversion systems. This
paper presents a background of the develop-
ment of CA screening procedures as they relate
to the Industrial Environmental Research Labo-
ratory's (IBRD Level 1 sampling protocol The
logic involved in selecting or rejecting specific
unit processes is presented. Screening proce-
dures to be used by a field team for gaseous and
aqueous waste treatment are described. The
development of detailed screening procedures
from the CA methodologies required laboratory
work for confirmation. Test results, conclusions,
and revised CA methodologies are presented in
the paper.
Biological oxidation screening procedures
were the most difficult problem in the develop-
ment ofCA screening procedures; therefore, lab-
oratory data derived from biological tests along
with recommendations for future work are pre-
sented. The feasibility of using a dry bacteria
culture for biological oxidation is discussed.
Laboratory data are presented from specific
gas treatment tests conducted using a modified
Source Assessment Sampling System (SASS)
train. Setup, operation, and required adjust-
ments to the SASS train for proper field opera-
tion are described.
INTRODUCTION
Control assay development (CAD) is the term
applied to a field-testing program for determin-
ing the best potential control techniques based
on Level 1 evaluation of effluent samples before
and after treatment by combinations of labora-
tory procedures that simulate control proc-
esses.
The physical and chemical characteristics and
'Speaker.
health/ecological effects of waste streams must
be determined to establish the potential pollu-
tion problem and the need for control technol-
ogy. The CAD approach for wastewater and for
air emissions provides practical and economical-
ly feasible screening procedures for a number of
treatment technologies without prior know-
ledge of all pollutant parameters. This is possi-
ble when broad criteria such as biological ox-
ygen demand (BOD), chemical oxygen demand
(COD), total organics, etc., can be used as a
measure of the effectiveness of treatment. Spec-
ific pollutants or health/ecological effects will
also be determined after completion of the
screening tests. The methodologies are designed
to produce reliable data indicating the degree of
effectiveness of each control process on a Level
1 basis.
During the formulation of CAD methodol-
ogies, it became apparent that certain proce-
dures should be verified in the laboratory
before being adopted for use in the final proto-
cols.
The objectives of the laboratory study were:
• To determine logistical problems of sample
handling.
• To assess the adequacy of the proposed de-
signs and operation of appropriate test
units.
• To examine the possibility of using a dry
bacterial culture for biological oxidation
studies.
• To evaluate the feasibility of using Source
Assessment Sampling System (SASS) com-
ponents for air testing.
CAD field procedures for coal conversion
wastewater treatment require processing rela-
tively large volumes of water as compared to
standard process development testing proce-
dures for determining treatability of a given
waste. Volumes of 200 L or more have to be
processed to accommodate normal system re-
quirements and to provide 10-L samples for the
Industrial Environmental Research Labora-
tory's (IERL) Level 1 analyses.
CAD air methodologies specify the use of a
425
-------
modified 8AS8. The minimum sample volume
required by IERL Level 1 air analyses for par-
ticulate, organic, and inorganic materials is
1,000 ft3. This volume allows for collection of
sufficient quantities of trace components to
reach detectable levels.
The principal control approaches for solids
(e.g., incineration and fixation) are not easily
conducted in the field. Incineration equipment
becomes impractical to outfit and operate in a
mobile facility. Chemical fixation or encapsula-
tion techniques are proprietary in nature and
cannot be satisfactorily duplicated in the CAD
test program. Samples would have to be for-
warded to a selected process vendor if data are
to be developed. These approaches are not rea-
sonable until a Level 1 analysis establishes the
need for treatment; therefore, no screening pro-
cedures have been recommended for solid waste
evaluation.
WASTEWATER SCREENING
PROCEDURES
Wastewater streams encountered during
CAD testing are expected to contain phenolic
compounds, ammonia, sulfides, and cyanide.
These materials should be present in large
enough quantities to make their recovery eco-
nomical in a full-scale plant; however, pilot-plant
operations may not be able to afford the capital
investment for recovery equipment. It is ex-
pected that CAD testing procedures will be em-
ployed using wastewater streams produced by
pilot plants. Therefore, it becomes necessary to
provide pretreatment of these samples in order
to simulate the characteristics of the waste ef-
fluent that could be expected from a full-scale
plant.
The analytical effort expected of the field
team members is not extensive for any of the
CAD testing. However, during pretreatment
some analyses must be performed to evaluate
the need for pretreatment and the efficiency of
removal when a sample is processed through a
pretreatment step. The individual streams used
to make up the composite sample will be ana-
lyzed with inexpensive test kits, and a decision
will be made by the team leader as to which
streams will be subjected to byproduct removal
treatment before compositing. After byproduct
removal and compositing, the sample will be an-
alyzed to determine the effectiveness of any
pretreatment steps that have been employed.
Phenolic compounds will be removed by liq-
uid-liquid extraction by using isopropyl ether as
the solvent. Ammonia and sulfides will be re-
moved by air stripping at appropriate pH
ranges. A high percent removal rate of hydro-
gen sulfide can be accomplished in a matter of
minutes at pH 7. Ammonia stripping will take
place at a pH of 11, and the wastewater will be
heated to 60° C to reduce the required stripping
time. Complete destruction of the cyanide ion
will be accomplished by the addition of sodium
hypochlorite with agitation.
Figure 1 shows the preliminary test sequence
for wastewater screening. The recommended
screening procedures are not intended to pro-
vide design data for a treatment plant but will
indicate the applicability of a particular treat-
ment process and provide information to be
used as a basis for further studies. The tests
have been limited to those unit processes that
have proven to be most successful in practice
and that have been most universally applied.
Two other processes (wet air oxidation and
evaporation/distillation) were initially consid-
ered for wastewater methodology, but both
were rejected because they are normally used
in special applications and would require more
sophisticated testing procedures than are war-
ranted for CA screening.
To accomplish the proposed objectives of this
portion of CAD, a 200-L synthetic wastewater
sample was processed as it would be by a sampl-
ing team in the field, with the exception of the
byproduct recovery steps and the treatment by
chemical oxidation. Complete Level 1 analytical
procedures were not applied to the treated sam-
ples because of time and cost restrictions. In-
stead, traditional wastewater parameters (COD,
BOD, solids, and metals analyses) were used to
measure the performance of each unit process.
Separate studies were conducted to determine
the effectiveness of using dry bacteria vs. an ac-
climated activated sludge for the biological ox-
idation assessment.
Synthetic Wastewater
Because of the difficulty of obtaining an ac-
tual coal conversion process waste, it was de-
cided to use a synthetically prepared waste for
426
-------
SOURCE A
SOURCE B
BYPRODUCT
REMOVAL
COMPOSITE SAMPLE
I
SOLIDS SEPARATION
!
BIO-OXIDATION
CARBON ADSORPTION
ION EXCHANGE
CHEMICAL OXIDATION
FOR
LEVEL I
ASSAY
I
CARBON I
ADSORPTION !
-*- 3
-*- 4
ION EXCHANGE
6
Figure 1. Preliminary wastewater test sequence.
427
-------
TABLE 1. ORGANIC COMPOSITION OF SYNTHETIC WASTE
Compound
Waste A
Concentration mg/1
Synthetic Waste
Concentration mg/1
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
15.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
Phenol
Resorcinol
Catechol
Acetic Acid
o-Cresol
p-Cresol
3,4 Xylenol
2,3 Xylenol
Pyridine
Benzoic Acid
4-Ethylpyridine
4-Methylcatechol
Acetophenone
2-Indanol
Indene
Indole
5-Methylresorcinol
2-Naphthol
2,3,5 Trimethylphenol
2-Methylquinoline
3,5 Xylenol
3-Ethylphenol
Aniline
Hexanoic Acid
1-Naphthol
Quinoline
Naphthalene
Anthracene
2000
1000
1000
400
400
250
250
250
120
100
100
100
50
50
50
50
50
50
50
40
40
30
20
20
20
10
5
0.2
200
500
500
400
40
25
25
25
120
100
100
50
50
-
50
50
25
50
5
40
4
3
20
20
20
10
5
0.2
428
-------
the laboratory verification studies. The organic
portion of the synthetic wastewater used for
verification purposes was derived from a formu-
lation developed by Dr. Philip Singer from re-
search conducted at the University of North
Carolina, Chapel Hill.1 The concentrations of
organic compounds proposed by Dr. Singer de-
fined a coal gasification wastewater with no
byproduct recovery steps. Because the labora-
tory verification was intended to test CAD
methodologies after byproduct recovery, the in-
itial organic concentrations were modified to
simulate a phenol recovery step. The Phenosol-
van® process was selected as a typical phenol
extraction process. Extraction recoveries ex-
pected from this process were estimated to be
99.5 percent monohydric phenols, 60.0 percent
polyhydric phenols, and 5.0 percent for other
organics.2
The phenolic compounds listed for waste A
(Table 1) were segregated by chemical struc-
ture, and values of 90 percent and 50 percent
removal were used to calculate the concentra-
tions remaining after byproduct recovery of
monohydric and polyhydric phenols, respective-
ly. No concentration adjustments were made for
"other organics."
The inorganic components of the synthetic
mix were selected after actual sample data from
several operating plants were reviewed. Table 2
lists the target inorganic concentrations in the
synthetic mixture.
Solids Separation
Four candidate approaches were considered
for separation of solids by physical means: cen-
trifugation, sand filtration, microstraining, and
cartridge filtration. Although it was felt that all
the above physical separation methods would
be applicable, the first three were discarded
after evaluation of various factors including
degree of solids removal required; the kind of
specialized apparatus needed; the question of
logistics for storing, transporting, and obtaining
new filter media; the ease of operation; and the
reproducibility of results.
Filtration of the composite sample using a
polypropylene cartridge was deemed to be the
most favorable method for solids removal in the
CA screening procedure. A pore size of 75 /*m
was selected as being descriptive of the particle
TABLE 2. INORGANIC COMPONENTS OF
SYNTHETIC WASTE
Component
F
Fe
Pb
Hg
P°4
S
Zn
As
Cd
Cr
Cu
Concentration (me/1)
2.0
0.2
0.04
0.007
2.5
12.0
0.08
0.2
0.02
0.03
0.1
1.0
size discharged from a well-designed primary
settler. A 200-L sample of synthetically pre-
pared waste was passed through the cartridge
filter with no difficulty. The synthetic waste
typically had a fairly low suspended solids level
at the outset, and no problems with filter plugg-
ing were encountered. It was noted, however,
that the waste did exhibit a tendency to precipi-
tate solids from solution upon standing. Several
filtrations were made at various times during
the laboratory study and the 200-L sample could
be passed through the filter in 15 min or less us-
ing a standard laboratory pump. Aeration oc-
curred because of the pumping action, which
caused some foaming in the sample; but this
situation was not considered to be a significant
problem. It is possible that actual wastewater
samples will have a much higher level of solids
than was encountered in the synthetic waste.
Also, during chemical pretreatment for bypro-
duct recovery, conditions could develop condu-
cive to the formation of precipitates, thereby in-
creasing the total amount of suspended solids in
the sample.
The filter cartridges are relatively inexpen-
sive and easy to change when their filtering
capacity has been exhausted. It would be possi-
429
-------
ble to make several filter changes during a run,
if it became necessary, without a significant loss
of time. Cartridge filters are also available in
various pore sizes, and two or more filters of
gradually decreasing size could be used in series
to obtain a higher degree of solids removal, if re-
quired. The synthetic waste had no visible ef-
fect on the integrity of the cartridge or the filter
holder (both polypropylene).
Activated Carbon
Evaluation of the effects of activated carbon
as a unit operation involves selection of a partic-
ular carbon, measurement of adsorptive capaci-
ty using batch isotherms, and development of a
breakthrough curve and regenerability charac-
teristics determined from a continuous-flow
pilot column test. In a detailed concept design
study, a number of different carbons are ex-
amined using a particular wastewater before
the best candidate is selected for the column
tests. Considering the basic purposes for CA
screening procedures and the field time con-
straints imposed, the use of a single, somewhat
broad-based carbon is proposed. This approach
may not produce data using the best suited car-
bon, but the results will be sufficiently indica-
tive of the applicability of carbon as a treatment
step and will still keep the investigations within
practical bounds.
Because it is a relatively simple matter to
perform carbon isotherms on a wastewater sam-
ple in the field to determine the approximate
organic loading and optimum pH conditions for
a specific wastewater, they have been included
as a prescreening procedure. Results of iso-
therm testing provide useful guidelines for the
column test runs in addition to the data they
furnish directly.
Two methods were considered for treating
the composite sample by activated carbon: con-
tinuous feeding through a series of carbon col-
umns, and batch testing. Each batch treatment
of a composite sample represents only one equi-
librium condition. It is anticipated that a micro-
filtration step for removal of suspended carbon
fines would be necessary before subsequent
processing steps could be performed.
Pilot column testing normally requires con-
tinuous sampling throughout the run at several
points in the carbon system to determine wave-
front movement and breakthrough, which are
among the data needed for an actual column
design. Because only a limited number of sam-
ples can be taken during CA screening, it is not
proposed, nor is it necessary, to conduct this
detailed type of design study. Based on the fore-
going considerations, continuous column opera-
tion was selected for use in screening; however,
the number of samples to be collected was lim-
ited to the feed and the composite effluent. The
volume of the feed to the carbon system will be
the amount needed to produce the samples for
analysis after the carbon test as well as from
any subsequent screening procedures, plus the
amount needed to displace "fill water" in the
columns. The feed volume will be contained in a
single vessel, pumped continuously through the
carbon beds, and collected in another vessel at
the effluent end. After an aliquot sample is
withdrawn for subsequent laboratory analysis,
the remaining effluent becomes the influent for
screening steps to follow. To determine general
column operation parameters, several iso-
therms are to be run on a small quantity of the
feed sample prior to the continuous run.
Table 3 summarizes the results of the ac-
tivated carbon verification testing. A Freund-
lich isotherm was developed on the synthetic
waste sample to establish the effectiveness of
carbon treatment and to gain some insight into
the amount of carbon required to produce ac-
ceptable organic removal rates. The standard
COD analysis was used as a measure of organic
removal. The values of X/M (quantity of COD ad-
sorbed per unit weight of carbon) were cal-
culated and plotted vs. concentration of residual
COD in solution. The plot of the data shows a
definite break at carbon dosages of 20 g/L and
higher. The sudden change in slope indicates
that two (or more) classes of organics present
are not uniformly adsorbable (Figure 2).
Carbon column runs were made using the col-
umn design specified by the CAD wastewater
methodology—four 2-in I.D. glass columns con-
nected in series, each charged to the 3-ft level
with activated carbon (7.8 Ib of carbon). The test
sequence for CAD (Figure 1) requires the use of
carbon at two points, before and after bio-oxida-
tion. After filtration the sample was equally div-
ided (84 L per run) for use during the column
tests.
In view of the apparent dual-adsorption re-
430
-------
TABLES. CARBON ISOTHERM RESULTS
Carbon Dose (M) COD Remaining (C) COD Removed (x) X/M
(KM/1 Sample) (*) Qng/1) Qng/1) (mg COD/gm Carbon) (**)
0
1
5
10
20
50
100
(*) Corrected for 100
(**) Equivalent to Ib.
Bun Linear Flow
number Rate (ml/min.)
BOD
1A 190
80
IB 190
58
1 (A4B) 190
2 200
5000 0 0
4653 347 347
3931 1069 214
3657 1343 134
3259 1741 87
1866 3134 63
1000 4000 40
ml sample size used.
COD adsorbed/ 1000 Ib. Carbon.
CARBON COLUMN TEST RESULTS
Loading Rate Influent Concentration Effluent Concentration (+) Z Removal
(gpm/ft2) COD mg/1 BOD me/1 COD rag/1 BOD mg/1 COD
2.3 6864 2200 1714 440 75
2.3 1714 440 334 186 80
2.3 6864 2200 334 186 95 91
2.4 3581 1940 347 197 90 90
(+) Corrected for dilution water in columns.
-------
I03
C* RESIDUAL COO (mg/l)
Figure 2. Carbon adsorption isotherm.
I04
gimes demonstrated by the batch isotherm, it
was decided to collect data during the first test
run in two stages. The 84 L of filtered waste
was pumped through fresh carbon in the col-
umns, and the effluent was retained (Run A).
After the columns were rebedded with new car-
bon, the effluent from Run A was used as the in-
fluent to Run B.
The second portion of synthetic waste was
treated by the bio-oxidation screening proce-
dure and then fed to fresh carbon in the col-
umns. Results of this test are indicated as Run
2. To varying degrees, carbon is effective in
reducing the COD and BOD of the synthetic
waste sample in both applications. By referring
to Table 3, it is seen that the combined Run 1
achieved essentially the same effluent COD and
BOD concentrations and percent removals as
Run 2. It must be recalled, however, that Run 1
was conducted in two stages and that twice the
carbon was bedded. The specified CA screening
procedures are more closely simulated by Run 1
alone. The data show that substantially fewer
(BOD/COD) organics are removed than in Run 2,
432
-------
which follows bio-oxidation.
It can be postulated that lower molecular
weight organics were not retained in the four-
column system but were captured in an equiva-
lent eight-column setup. Apparently, the four-
column system was able to produce a better ef-
fluent quality after one pass by virtue of the
reactions taking place during the bio-oxidation
procedure.
The run time required to process an 84-L sam-
ple through the four-column system at a super-
ficial velocity of 2 to 3 g/min/ft2 is approximately
8 hr. By increasing the column size to 3-in I.D.,
the sample could be processed in slightly less
than 3 hr at an identical superficial velocity. On
the other hand, the amount of carbon available
would be increased more than twice.
One disadvantage of increasing the column
size is that the dilution from the "fill" water ex-
isting in the carbon bed at the beginning of the
run becomes larger in relation to the size of the
sample being passed through the columns. In
any event, the dilution factor has to be con-
sidered when test results are interpreted and
should not substantially affect the evaluation of
activated carbon as a unit process, provided
that a sufficiently large sample is processed.
The synthetic waste demonstrated a tenden-
cy to form some additional solids on standing,
which were removed by the carbon bed. If real
wastes react similarly, it may be necessary to
perform a supplemental cartridge filtration be-
fore feeding the sample to the columns to pre-
vent bed blinding.
The column design was modified slightly be-
cause plugging problems arose using the origi-
nal fitted glass support materials. These were
removed and replaced with 50-mesh screen,
which was satisfactory for all subsequent runs.
Biological Oxidation
The original intent of wastewater treatment
evaluation was to have a field team onsite to
perform all aqueous screening procedures in ap-
proximately 1 week. Standard biological treat-
ability testing using activated sludge normally
requires.2 weeks to 1 mo of continuous opera-
tion for acclimation of the biomass to the specif-
ic waste being studied. After acclimation, an ad-
ditional 3 to 4 weeks of data gathering under
steady-state conditions are required to provide
system performance and design parameters for
that particular wastewater. Control assay
screening procedures are not developed for the
purpose of obtaining design data; therefore, the
continuous sampling after acclimation is not
necessary. However, to properly evaluate a
biological system as a unit process, it is im-
perative that an acclimated seed be used.
The requirement for an acclimated seed on-
site posed several problems. A "wet" seed must
be continuously aerated and provided with
some type of feed substrate during transporta-
tion to a plant and on location. The possibility of
acclimating a sludge from a local municipal
treatment plant was also considered. While a
viable option, such an approach could introduce
unwanted contaminants to the system, depend-
ing on the type of industrial waste normally
treated at the local plant. Biological sludge from
a plant that normally treats coke oven wastes
would be ideal because components of this type
of wastewater are similar to many materials
found in coal conversion wastes. However, the
likelihood of being near this type of treatment
plant would be small and could not be realistical-
ly incorporated into the screening methods. In
essence, it was desirable to determine if there
were any feasible alternatives to using a wet
seed for the screening procedure.
By private communication, one investigator
reports experimentation examining the possi-
bility of quick-freezing activated sludge for sub-
sequent use. While interesting, the work is still
in an early trial stage and the results are too
tentative for inclusion in a screening procedure
at this time. A second alternative is the use of
dry bacterial cultures offered commercially by
several vendors.
Dry bacterial cultures are grown on an inert
material. The organisms are selectively mu-
tated and segregated in accordance with their
ability to biologically degrade specific classes of
compounds. One such culture is purported to
specifically oxidize phenolic compounds, cya-
nides, and other similar contaminants. The
culture is marketed in a dry powder form and,
according to the vendor, the organisms are reac-
tivated when added to warm water and aerated
for 24 hr. The dry bacterial culture route offers
a potential solution for the transportation and
acclimation problems posed by CA methodolo-
gy-
433
-------
It was decided to test a dry bacterial culture
to ascertain whether or not it would serve as a
practical alternative for a wet seed and/or to try
to establish a relationship between system per-
formance using dry bacteria as compared with a
seed acclimated to a waste in the more usual
manner. Tests performed to evaluate biological
screening procedures were divided into two
categories: batch testing and continuous sys-
tems. Additionally, experimental work was con-
ducted to gain better familiarity with the char-
acteristics and application of the dry bacterial
culture; and to explore some side issues that
arose during the test work that were relevant
to the overall bio-oxidation verification pro-
cedures.
The batch tests were performed either in 2-L
glass beakers or in 7-L cylindrical, stainless
steel containers. Vessels used for the continu-
ous systems testing were 7.5-L capacity stain-
less steel tanks fitted with baffle plates at the
outlet to provide a quiescent zone for solids set-
tling. The volume of the aerated portion of these
tanks was about 6 L.
An attempt was made to start a continuous
system using the dry bacterial culture. After
several days of feeding with dilute synthetic
wastewater, there was no apparent biological
growth. It was believed that the bacteria were
present as a dispersed growth and were being
lost in the effluent because there was no meas-
urable solids production in the system and ef-
fluent COD values were consistently higher
than the feed analyses. Millipore filtration of
the effluent samples did not significantly reduce
the effluent COD results.
Data collected during the early exploratory
work with the dry bacterial culture contained
two anomalies:
• Effluent COD concentrations were higher
than influent values.
• The COD concentration in the open feed con-
tainer dropped rapidly on standing.
The latter effect was substanially reduced—but
not totally eliminated—by covering the feed
vessel during the subsequent continuous bio-
testing studies.
The phenomenon of organic (BOD/COD) loss
from the synthetic waste mixture was ad-
dressed several times during verification test-
ing through studies involving aeration of dif-
ferent batches of synthetic waste under varying
test conditions. The data collected during these
runs are presented in Tables 4, 5, and 6.
Air-stripping tests were performed on batch
sampled of the synthetic waste to quantify the
loss of COD material (presumably) by volatiliza-
tion and/or oxidation of the organic compounds
in the waste (Table 4). At the same time, tests
were conducted to determine the amounts of
COD and BOD added to a batch system by the
dry bacterial culture alone (Table 7). A supple-
mental air stripping/oxidation run was con-
ducted near the end of the laboratory test to ex-
amine the effect of volume size on BOD/COD re-
ductions. For convenience, these data are shown
in Table 5.
The bulk of the results support the proposi-
tion that the losses occur primarily through vol-
atilization. However, there is some evidence
that chemical oxidation of the organics could
also be involved. Whatever the actual mecha-
nisms might be, Tables 4 and 6 (Unit 1) show
that the cumulative effect of air stripping/ oxi-
dation is essentially reached after 48 hr of aera-
tion. Table 5 evaluates the effect of volume size
on BOD/COD reduction. A stripping action is de-
finitely indicated by the fact that the (smaller)
units with greater air-to-liquid ratios demon-
strated higher reductions.
The supplier's recommended standard proce-
dure was followed for reactivating the dry bac-
terial culture. First, a measured amount (25 g) of
bacteria/ substrate material was added to 3 L of
distilled water, heated to 38° C (100° F), and
mixed for 2 hr. The batch was then aerated for
24 hr, and aliquots were taken to produce vari-
ous concentrations for analysis. The test results
indicated that the BOD and COD concentration
will increase as a result of adding the dry cul-
ture. Relationships are depicted in Figure 3.
The zero hour time did not include the initial
24-hr aeration period; therefore, the total aera-
tion time from start of reactivation to the end of
the test was actually 96 hr. These tests indi-
cated that the substrate material will provide
the bacteria with an adequate nutrient supply
for at least 72 hr, while also adding organic food
(COD) material to the system. Measurements of
oxygen uptake rates on similar systems con-
firmed the continued high biological activity
over the same time period.
Dry bacterial cultures can also be used as an
additive to an existing biological system. Be-
434
-------
TABLE 4. AIR STRIPPING/OXIDATION TESTS
Aeration Time
>£>
CO
cn
Run #2
Run #3
Run #4
(hrs)
0
1
2
4
24
48
72
COD
(mg/1)
5660
4228
2686
2412
1965
Rem.
%
0
25.3
52.5
57.4
65.3
COD
(mg/1)
5504
2046
1450
1580
Rem.
%
0
62.8
73.7
71.3
COD
(mg/D
4761
2637
2030
1834
Rem.
%
0
44.6
57.4
61.5
BOD
(mg/1)
3306
1408
960
760
Rem.
%
0
57.4
71.0
77.0
COD
(mg/1)
4280
3412
2410
2222
Rem.
%
0
20.0
43.7
48.1
BOD
(mg/1)
2340
1980
1200
Rem.
%
0
15.4
48.7
NOTE: Sample volume used was 15 liters.
TA0LE 5. EFFECT OF VOLUME SIZE ON AIR STRIPPING/OXIDATION
Run #5- 22 eal. Volume Run #6- 7 liter Volume
Parameter
BOD
COD
Infl.
(mg/1)
1080
7560
Aeration Only
Effl. Rem.
(mg/1) (%)
780
5520
27.8
27.0
Dry Bacteria
Effl.
(mg/1)
740
5680
Rem.
(%)
31.5
24.9
Aeration Only
Effl. Rem.
(mg/1) (%)
780
3760
Dry Bacteria
Effl. Rem.
(mg/1) (%)
27.8
50.3
870
3840
18.4
49.2
NOTE: 24 hours aeration period on all units
-------
TABLE 6. BIOLOGICAL OXIDATION BATCH REACTOR RESULTS
Onlt tl-Kir Stripping/Oxidation
Influent Effluent X Removal
BOD («g/D
COD («g/l)
BOD («s/D
COD
BOD
-------
TABLE 7, DRY BACTERIA-COD AND BOD DATA
BOD
Dry Bacteria Concentration
Average
Aeration Time Adjusted Adjusted Adjusted Adjusted Adjusted
(hours) 0.75 gm/1 Value* 1.5 gm/1 Value* 2.25 gm/1 Value* 3.0 gm/1 Value* Value
24
48
72
44
60
86
Average BOD increase: 91
Aeration Time
(hours)
24
48
72
0.75 gm/1
80
102
245
59
80
115
mg/l/gm Dry
Adjusted
Value* 1
107'
136
327
92
112
106
61
75
71
290
274
140
128
121
62
386
268
314
128
89
104
94
91
88
Bacteria added
Dry
.5 gm/1
165
177
280
COD
Bacteria Concentration
Adjusted
Value* 2.25 gm/1
110
120
187
490
500
578
Adjusted
Value*
217
222
256
3.0 gm/1
722
725
895
Adjusted
Value*
240
242
298
Average
Adjusted
Value
169
180
267
Average COD increase: 205 mg/l/gm Dry Bacteria added
*Mathematically adjusted to a Dry Bacteria concentration of one mg/1.
-------
300
250
COD
o:
LJ
U 200
<
CD
tr
Q
fc
150
100
BOD
50
24
48
72
TIME (hrs)
Figura 3. COD and BOD addition by dry bacteria.
438
-------
cause poor results were being obtained from the
continuous system, this operation was discon-
tinued and replaced by two new continuous
units, each containing biomass taken from a
coke oven waste treatment plant. Identical
amounts of the synthetic waste were fed to each
of the units. Additionally, doses of the dry bac-
terial culture were introduced to one of the
units on a daily schedule prescribed by the sup-
plier's instructions. Gradually decreasing
amounts of dry culture were added to this sys-
tem until a "maintenance" dosage level (2 g/6 L)
had been reached. This dosage was continued
for the duration of the testing period. Sludge
from these units was later used for additional
batch tests. Results of the continuous reactor
testing will be discussed later in this report.
Batch Testing
Three sets of batch tests were conducted,
each set consisting of four batch reactors
aerated for 72 hr. Samples from the reactors
were taken every 24 hr and analyzed for COD
and BOD. Air flow to each system was stopped
for 1 hr before sampling to allow for solids settl-
ing. One reactor (Unit 1) in each series contained
wastewater only (no biologically active seed in-
troduced) for the purpose of comparing the ef-
fects of air stripping/ oxidation of the waste to
biological oxidation. The contents of the other
three reactors were prepared as follows:
• Unit 2 —Wastewater plus coke oven sludge
(from continuous Unit A).
• Unit 3—Wastewater plus coke oven sludge
with dry bacteria (from continuous
Unit B).
• Unit 4 —Wastewater plus dry bacteria.
The batch testing (Table 6) revealed no
significant differences in BOD and COD remov-
als between the dry bacteria system (Unit 4) and
the air stripping system (Unit 1).
Both of the systems (Units 2 and 3) using
coke-oven-activated sludge as the bulk of the
seed, performed similarly, with better removals
than the stripping unit and the dry bacteria
unit. In these batch tests, no significant dif-
ference was observed between coke oven sludge
alone (Unit 2) and the system containing sup-
plemental dry bacterial culture (Unit 3).
Average COD and BOD removals were calcu-
lated to compare the effectiveness of the differ-
ent units. After 24 hr, there was little difference
among any of the reactors in either BOD or COD
removal, except for Unit 1, which was some-
what lower. The units containing coke oven
sludge (with and without dry bacteria) began to
show greater removals at 48 hr, and this trend
continued for the 72-hr samples. The reactor
containing dry bacteria alone showed very lit-
tle, if any, superiority over the air stripping/ox-
idation reactor during the first day; and by the
end of the test, the removals were essentially
equivalent. Unit 3 (coke oven sludge plus dry
bacteria) had a slightly higher COD removal
rate than Unit 4 (coke oven sludge only), but the
difference was so small that it cannot be attrib-
uted to the dry bacteria. BOD removals for
these two units were identical.
Two continuous units were set up and oper-
ated for approximately 2 1/2 mo. Both units (A
and B) were seeded with a coke oven sludge; one
unit (Unit B) also received a daily dose of dry
bacteria. The systems were contained in iden-
tical stainless steel reactor tanks each having a
removable baffle to aid in clarification of the ef-
fluent streams. The influent to both systems
was from a common tank, and various concen-
trations of synthetic wastewater were used as
the feed material. Initially, the synthetic waste
was diluted to one-tenth of the original strength
and later changed to one-quarter strength. Dur-
ing the final 3 weeks of testing, both units were
fed full-strength synthetic wastewater.
Figure 4 shows influent and effluent COD
data for both continuous units during the entire
test period. During the early part of the run, the
unit with dry bacteria addition (Unit B) showed
higher effluent values. Vendor instructions on
the use of the dry bacterial culture as a supple-
mental addition were followed in Unit B. The
procedure specified a relatively high initial dose
followed by a dosage rate decreasing to a point
where only a maintenance dose is applied daily.
Presumably, the effluent COD pattern demon-
strated in Unit B reflects the changing dosage
rate of the bacterial culture. (The effect of
culture dose on effluent COD has already been
discussed.) When the dry bacteria addition
reached the maintenance dosage level, COD re-
movals for this system (Unit B) reached a level
equivalent to the coke oven sludge system (Unit
A).
During the final 3 weeks of testing, both units
were fed full-strength waste. The unit with the
439
-------
7000-
6000-
5000
£
i
o
o
1000
500-
IOO
0
INFLUENT-
BOTH UNITS
10
20
30
40
50
DAYS
60
70
80
Figure 4. Continuous biological reactor results.
-------
dry bacteria showed a much greater ability to
cope with the shock loading conditions encoun-
tered when the feed was abruptly changed to
full strength. The companion unit was adversely
affected by the change in feed, although it
gradually recovered over a 3-week period when,
because of time limitations, operation of all
units was discontinued.
Results from verification testing of the bio-
oxidation screening procedure have produced
valuable information relevant to CAD waste-
water methodology. If the synthetic waste mix-
ture used in the experimental work closely
simulates a real-life coal conversion aqueous
waste, then a substantial portion of the organic
removals usually attributed to oxidation by
biological organisms may well be physically
stripped from the bioreactor as an air emission.
Consequently, a simple aeration step in parallel
with the biological treatment step appears war-
ranted to ascertain the extent to which organic
removal through stripping/oxidation is occur-
ring.
Based on results developed with one commer-
cial dry bacterial culture mixture, the use of this
type of dehydrated product as a biological seed
does not meet the needs of the screening proce-
dure. A wet seed approach must be adopted.
Moreover, the wet seed must be acclimated for
about 3 weeks to a waste stream that is general-
ly descriptive of the material that will eventu-
ally be tested by the CA procedure.
Clearly, two choices present themselves. One
is to disregard the biological oxidation step en-
tirely, which is not really reasonable, since this
approach will eliminate consideration of the ef-
fects of a major waste treatment unit process.
The second option is to begin biological acclima-
tion (using a locally available activated sludge
as seed) 3 weeks in advance of the wastewater
screening study. During this time, the CA team
could be generating the air samples for IERL
Level 1 analyses.
At the outset of biotesting verification, it was
presumed that the team would use COD analy-
ses as the prime performance monitoring meth-
od, backed up by an occasional reference BOD.
In view of the experience gained during this
test work, some doubt is now cast upon the val-
idity of using COD for these purposes. Changes
produced by aeration in the oxidation state of
dissolved waste organics may be clouding the
dichromate chemistry and possibly producing
misleading data. It is recommended that the
team should be equipped with a TOC analyzer
for quantification of waste organic content and
for process monitoring purposes.
Ion Exchange
After discussions with an ion exchange resins
manufacturer, it was decided to employ a three-
glass (2-in ID.) column system set up in series.
The first column contained a strong-acid type
resin, the second column was filled with a weak-
acid resin, and the final column contained a
strong-base resin. Prior experience by the man-
ufacturer suggested that this combination of
resins would remove the majority of ions ex-
pected to be present in a typical coal conversion
wastewater. To minimize pumping require-
ments, a single pump was to be used to intro-
duce the sample into the first column and, by
proper positioning of the second and third col-
umns, a continuous gravity flow would be main-
tained.
The ion exchange system was tested to evalu-
ate its ability to process the required aqueous
sample within 1 work day. Excess solids in the
wastewater caused a flow rate problem in the
columns that was solved by filtering the sample
through the 75-/tm cartridge and changing the
resin bed support media. A single pump was
used to introduce the wastewater into the first
column, and gravity flow was employed through
the second and third columns. Constant adjust-
ments to the column height and piping were
necessary to produce a continuous flow through
all of the columns.
CAD methodology specifies the use of ion ex-
change at two points in the test sequence (Fig-
ure lh after bio-oxidation and after bio-oxidation
plus carbon adsorption. Reference analyses of a
few selected metals were made for these runs
and the results are shown in Table 8.
The gravity flow concept was not acceptable
because unequal pressure drops through the
columns, caused primarily by differences in res-
in particle diameters, necessitated constant ad-
justments to the column heights to maintain a
continuous flow. It has been determined that
the sample should be pumped through one col-
umn at a time to eliminate this problem. Fur-
thermore, to reduce the possibility of plugging
441
-------
TABLE 8. RESULTS OF ION EXCHANGE TESTING
Parameter
Iron as Fe, mg/1
Copper as Cu, mg/1
Cadmium as Cd, mg/1
Zinc as Zn, mg/1
Influent
0.7
0.18
0.06
0.36
Run #1
Effluent
1.5
N.D.
0.05
0.22
Run 12
Effluent
0.7
0.034
0.05
0.15
Notes;
Run //I was made on a sample after bio-oxidation plus carbon adsorption.
Run 92 was made on a sample after bio-oxidation only.
N.D. Indicates Not Detectable (less than 0.05 mmg/1).
the resins with solids, a cartridge filter should
be placed in-line before the first resin column.
The analytical data indicate that the ion ex-
change resins did remove metals, although
there was some performance variability from
metal to metal. The principal impact on CAD
methodology is that an overall comparison of
the effluents from both runs shows them to be
reasonably similar; therefore, two ion exchange
runs are not required for CAD purposes. The
ion exchange run after carbon adsorption is the
more appropriate site selection in the test se-
quence.
In view of the increase in column size (from
2-in to 3-in I.D.) suggested for the carbon screen-
ing procedure, it is logical to also change the ion
exchange column size to 3 in. This alteration will
gain some time during the ion exchange test run
and will serve to standardize the column sizes
for both screening procedures.
Chemical Oxidation
Phenolic compounds and numerous other
organic chemicals can be destroyed by reaction
with an oxidizing agent. The choice of an oxidiz-
ing agent rests primarily on its rate of reaction,
selectivity, cost, and ease of handling. Some
commonly used chemical oxidants are:
• Ozone and oxygen,
• Hydrogen peroxide,
• Potassium permanganate, and
• Chlorine and chlorine-containing com-
pounds.
For thermodynamically reversible reactions,
the oxidation reduction potentials can be used
as a quantitative measure of oxidizing power.
However, most reactions involving oxidation of
organic chemicals are irreversible and, there-
fore, the redox potentials are of little use for
predicting expected behavior.
Hydrogen peroxide will be added to the sam-
ple to oxidize any organic components remain-
ing after being processed through the classical
treatment processes. This procedure was not
tested during laboratory verification.
Conclusions and Recommendations
Laboratory verification of the CA screening
procedures revealed several problems with the
original wastewater methodologies. Minor
equipment changes were made to facilitate sam-
ple handling, and a revision of the biological ox-
idation procedure was necessary. Figure 5
shows the steps in the initial CAD treatment se-
quence and includes verification testing results
for those processes examined.
Conclusions and recommendations developed
from the study are:
• Solids separation using an in-line cartridge
filter presented no difficulties, and this ap-
442
-------
COMPOSITE
BOD: 2260
COD: 6666
SS : 382
VSS: 226
pH : 8.0
FILTRATION
BOD: 2200
COD: 6860
117
71
SS :
VSS:
PH :
7.9
BIO-OXIDATION
BOD: 2110
COD: 3571
SS : 362
VSS: 271
pH : 7.6
CARBON-2
BOD:
COD:
SS :
VSS:
pH :
197
347
73
48
7.6
CARBON-1
BOD:
COD:
SS :
VSS:
PH :
186
334
30
30
7.7
ION EXCHANGED!
BOD: 2100
COD: 3490
SS : 85
VSS: 42
Fe: 1.5
Cu: N.D.
Cd: 0.05
Zn: 0.22
ION EXCHANGE-2
BOD:
COD:
SS :
VSS:
pH :
194
340
62
30
7.6
Fe: 0.7
Cu: 0.03
Cd: 0.05
Zn: 0.15
Figure 6. Results for synthetic waste sample.
443
-------
proach will be adopted as originally con-
ceived. If precipitates form in the waste-
water sample, supplemental solids filtra-
tions may be required to prevent blinding of
the carbon and/or ion exchange resin beds.
• The effect of carbon adsorption should re-
main where proposed by the GAD waste-
water methodology; i.e., both before and
after bio-oxidation.
• The carbon column diameter should be
changed from the 2-in I.D. specified to 3 in. A
few minor column design modifications are
also suggested.
• Verification testing data strongly support
the proposition that a substantial portion of
the BOD and COD removals demonstrated
during the bio-oxidation screening proce-
dure can be attributed to air stripping (vola-
tilization). Therefore, the CAD wastewater
methodology should be modified to include
an air-stripping step running in parallel with
the specified bio-oxidation screening proce-
dure.
• Insufficient benefit is derived from the use
of a dry bacterial culture during the bio-
oxidation screening procedure to warrant
its adoption in the testing protocol.
• To be effective, bio-oxidation screening
must use an activated sludge that has been
acclimated to the wastewaters under consid-
eration for a period of 3 weeks prior to the
formal initiation of the CAD wastewater
methodology. While acclimation is under-
way, it is anticipated that the CAD team
would be pursuing the screening procedures
specified by CAD air methodologies.
• Based on experience derived during the
verification testing, the use of COD analyses
as the monitoring method should be re-
placed by TOC to provide a faster and more
accurate analysis of the organic composition
of the samples.
• The gravity flow concept throught the ion
exchange columns is not acceptable as a
CAD screening procedure. The wastewater
sample should be pumped through each col-
umn.
• Evaluation of the effects of ion exchange
should be studied only after carbon adsorp-
tion and not before. The wastewater testing
sequence should be altered accordingly.
• The ion exchange column diameter should
be standardized at 3 in.
Figure 6 shows the final version of the
wastewater screening test sequence.
GASEOUS EMISSIONS SCREENING
Control technology for screening gaseous
samples to determine potential treatment
methods must include unit operations for the
removal of particulates and gases/vapors of con-
cern. Either class of materials may be organic or
inorganic. The types of control technology for
gas treatment include mechanical collection,
electrostatic precipitators, filters, liquid scrub-
bers/ contactors, condensers, solid Sorbents, and
incineration.
Sampling of air streams for Level 1 CAD is
much more difficult than the simple grab proce-
dures specified for liquids. The inability to bring
sufficient sample volume into the CAD mobile
test facility, as is possible with liquid samples,
limits the use of a number of unit operations
and/or desirable strategy that can be applied in
the air methodology. The practicality of per-
forming certain types or large numbers of CAD
tests at the source may be restricted by such
factors as limited working space on a platform,
logistical problems servicing a platform, plant
restrictions on use of nonexplosion-proof equip-
ment, personnel safety, requirement for special-
ized equipment (e.g., SASS train), and the ana-
lytical load generated by a broad test plan.
Based upon the above considerations, the
Level 1 air methodology was developed to be
flexible but more reliant on process information.
This permits the user of CAD to be selective in
choosing a screening system and may allow a
more simplified approach to certain streams.
The various screening sequences available in
Level 1 CAD are presented in Figure 7.
Unit operations considered for the air meth-
odology but not being evaluated in the sequence
are electrostatic precipitation, flaring, and in-
cineration. The following sections indicate the
reasons for their exclusion.
Electrostatic Precipitation
The selection of electrostatic precipitation
technology depends heavily on conductivity and
resistivity properties of the gas stream. Instead
of testing a prototype electrostatic precipitator
unit as a CA screening procedure, measurement
of the following properties is recommended to
444
-------
SOURCE A
SOURCE B
BY-PRODUCT
REMOVAL
I
FOR LEVEL 1
ASSAY
COMPOSITE
SOLIDS REMOVAL
AIR STRIPPING/
OXIDATION
CARBON
ADSORPTION
I 3
BIO
-OXIDATION
~_ _ _ _ r _r
CARBON
ADSORPTION
""* - *^ 5
_mi ^LL «i^ £
ION EXCHANGE
CHEMICAL
OXIDATION
1 8
Figure 6. Final wastewater test sequence.
445
-------
SOURCE
PARTICULATE
REMOVAL
GAS
COOLING
CARBON
ADSORPTION
SCRUBBING
i
CARBON
ADSORPTION
FOR LEVEL
ASSAY
IA, IB, 1C
-3A,3B,3C
Figure 7. Preliminary air testing sequence.
446
-------
supplement existing Level 1 protocols:
• Particle resistivity,
• Particle size—average diameter,
• Specific gravity,
• Bulk density, and
• Particle size distribution curve.
Direct Combustion (Flare)
Flaring is acceptable control technology for a
number of applications, principally in the petro-
leum refining and other industries where upset
conditions involving large volumes of flammable
gases can be economically handled. It is not
recognized or recommended as best available
control technology by regulatory agencies due
primarily to lack of a sufficient data base. A ma-
jor disadvantage is the absence of equipment
and practical techniques to sample the products
of combustion and monitor performance. Meth-
ods and equipment sizes used in pilot-plant test
runs are not practical for CAD and have not
yielded data that can be used for scaleup design
or prediction of performance. The disadvan-
tages of flares are presently too great for the
unit operation to be useful in CAD.
Direct Flame Incineration
Thermal incineration is one of the most effec-
tive means for disposal of hazardous waste
gases and, despite high capital and operating
costs, will likely be specified more frequently in
the future for problem pollutants. A proper
evaluation of the capability of incineration
would involve study of key parameters such as
residence time and temperature. The manipula-
tion of a number of variables is beyond the
scope of Level 1 CAD and, coupled with the gen-
eral difficulty of handling large volumes of sam-
ple, screening tests on incineration become im-
practical and are not recommended. Incinerator
manufacturers, however, have compiled a large
data base on the thermal oxidation of organic
materials, and there is also a high level of con-
fidence that almost any organic material can be
destructed.
The Level 1 air methodology is applicable to
any point source where a Level 1 environmental
assessment might be performed. This is gener-
ally intended to mean those sources that dis-
charge directly to the atmosphere and does not
normally include process lines, internal recycle,
or waste gas lines directed to control devices.
Open vents or stacks that are considered
sources of uncontrolled fugitive emissions are
not recommended for CAD. Examples of these
sources include relief systems, pressure let-
down or control systems, emergency vents,
leaks, spills, etc. They are normally highly
variable in composition, rate, frequency, and
duration, and control technology is often uneco-
nomical or difficult to apply. When the materials
are hazardous, it is common to collect the va-
pors in an exhaust system and direct the com-
bined flow into a central control system such as
a scrubber or flare. Discharges from control
systems are usually of interest to CAD.
Vents, stacks, and other point sources of air
emissions are usually too numerous in the plant
site to permit a CAD assessment of each dis-
charge. A cost-effective program can best be
achieved by performing a reasonably complete
engineering review of the available data before
finalizing sample points. Process and engineer-
ing flow sheets, process and treatment descrip-
tion, and all other information should be studied
prior to a preliminary site visit. During the
visit, information gaps may be filled by discus-
sions with plant personnel and/or inspection of
equipment and devices. If it can be established,
for example, that the emission is a vapor and
contains no particulate matter, the most com-
plex and costly test configuration requiring par-
ticulate sampling modules can be avoided. Fur-
thermore, if the source is a pure, single-compo-
nent organic material (such as breathing and fill-
ing vapors from a storage tank), CAD may not
be needed at all because emissions can be cal-
culated and potential control technology
selected based on the material properties.
IERL Level 1 sampling protocols are em-
ployed in Level 1 CAD air methodology. The
sampling apparatus for a Level 1 assessment
are the grab bulb, for gaseous samples only, and
the SASS, for gaseous streams containing par-
ticulate. The control technologies recommended
for CAD air methodology are particulate remov-
al, gas cooling (condensation), liquid scrubbing,
and carbon adsorption. The equipment for these
operations is constructed and assembled as mod-
ules (Figures 7 and 8). Following is a brief de-
scription of each module and its function in
CAD.
447
-------
SOURCE
PARTICULATE
REMOVAL
CONDENSATION
SCRUBBING
CARBON
ADSORPTION
FOP LEVEL 1
ASSAY
CARBON
ADSORPTION
Rgure 8. Final air tasting saquanca.
448
-------
Paniculate Ramoval
The module is a standard SASS train cy-
clone/filter assembly, contained in a heated
oven. For CA screening purposes, this module
serves only to pretreat the gas when participate
is present.
Gas Cooling
Hot gases must be cooled to at least 55° C
(130° F) before entering an activated carbon
module. In commercial practice, gases are often
cooled to permit use of cheaper materials of con-
struction (e.g., plastics) in downstream ducts
and equipment. In addition to cooling as a pro-
tective measure, condensation of volatile mate-
rial is a valuable control technology. This mod-
ule also will be a standard SA88 train compo-
nent, except that the sorbent cartridge is not
used and will be taken out of line.
Scrubbing
Liquid scrubbing, using an aqueous alkaline
solution, is specified as the primary control
technology in Level 1 CAD screening for remov-
al of pollutants in acid gases. Several media
were investigated and sodium carbonate was se-
lected. COg is a common component in many
gaseous streams and will be absorbed in media
such as sodium hydroxide, requiring a large vol-
ume of solution and causing logistical problems.
The capacity to remove acidic components at ex-
pected concentrations cannot be handled in the
standard SASS impinger assembly. Therefore,
a small counter-current scrubber must be used.
Carbon Adsorption
Activated carbon is being studied for removal
of trace quantities of organic and inorganic ma-
terials. The economics of regeneration usually
preclude carbon being used as the primary tech-
nology for removal of high concentrations of
organics. Regeneration will not be studied in
Level 1 GAD. The module is a column canister
sized to contain a sufficient quantity of ac-
tivated carbon. Calculations show that the
capacity of a standard SASS sorbent module is
not adequate for CAD studies.
The general principles of IERL sampling ap-
ply to CA but may be modified to accommodate
a more flexible approach in air methodology.
This it illustrated in Figure B, which outlines
alternative screening arrangements and as-
sociated sampling requirements. For CA screen-
ing procedures, the standard SASS modules are
used in the following manner:
• The particulate removal module (cyclones
and filter) is used for preconditioning of the
stream prior to entering control devices.
• The gas-cooling module of the SASS train is
used in CAD for evaluating condensation
control technology. Operating this module
according to Level 1 assessment parameters
will serve both as condensation screening
technology and the means to provide a sam-
ple for evaluation of the applicability and ef-
fectiveness of condensation.
• The XAD-2 cartridge and the impinger mod-
ule in the sampling system (see Figure 9) is
designed to collect the residual. A side bene-
fit is the removal of corrosive material that
would cause damage to the vacuum pump,
dry gas meter, and other components down-
stream.
The complete Level 1 analytical protocols
shall be performed on the gas samples pro-
duced. The CAD sample sizes shall meet the re-
quirements for Level 1 analytical protocols.
These are presently:
• GC analysis: 3 L (grab);
• Physical/chemical testing and health effects:
30 m3 (passed through SASS train); and
• Ecology effects: 1,360 L (grab).
Laboratory Verification
In developing the CAD air methodologies,
typical unit operations needed to remove par-
ticulates and gases/vapors from air emissions
were evaluated. For various reasons, some of
these operations had to be excluded from con-
sideration as CA screening procedures. Control
technologies eventually selected for the CAD
methodology included particulate removal, gas
cooling (condensation), carbon adsorption, and
liquid scrubbing.
The SASS, developed for IERL Level 1 sam-
pling, made use of all these mechanisms for
separation and collection of gas stream contami-
nants and therefore initially seemed to be an
ideal system for use in CA screening proce-
449
-------
.
s
STANDARD
SASS TRAIN
CONDENSOR
MODULE
*
PARTICULATE
REMOVAL
SCRUBBER
STANDARD
IMPINGER
MODULE
ORIFICE DRY PUMPS
GAS
METER
Figure 9. Combined screening train.
-------
dures. It was thought that activated carbon
could replace XAD-2 in the same cartridge.
However, subsequent calculations showed that
the capacity of the standard XAD sorbent mod-
ule used in the SASS train would not be ade-
quate for these studies.
Several scrubbing media were investigated
and sodium carbonate was selected as the most
promising. The capacity needed to remove acid-
ic components at expected concentrations was
also calculated, and it was determined that the
standard SASS impinger assembly would not
hold the required volume. The existing conden-
sation module in the SASS train was not ex-
pected to be a problem because sample flow
rates and test duration would be similar to
those encountered in IERL Level 1 sampling.
In order to provide the extra capacity re-
quired for scrubbing, a counter-current, packed-
column scrubber with an 8-L reservoir was
designed. A 4-in I.D. by 5-ft glass column con-
taining 3 ft of Raschig rings as packing was used
during verification testing.
Likewise, a larger canister to contain the ac-
tivated carbon was specified. A 4-in I.D. by 3-ft
glass column containing 10 Ib of activated car-
bon (3-ft bed depth) was used for testing.
Figure 9 shows the configuration of the
modified screening train as assembled to
evaluate scrubbing followed by activated car-
bon. Both control technologies can be evaluated
separately if a process review indicates no need
to study both systems in series.
The solids removal module of the standard
SASS has been incorporated into the train.
However, particulate removal technology will
not be evaluated during screening because data
for evaluating the effects of solids removal
technologies/control devices are obtained by the
standard IERL Level 1 sampling protocols, as
amended by CAD methodologies. When a gas
stream with a high particulate loading is sam-
pled, this module will prevent particle buildup
on the activated carbon. The condenser module
serves two purposes: for cooling of the gas
stream (to a carbon influent temperature of
55° G or less), and as a separate unit process for
removal of low-boiling organics.
The standard SASS train presently requires
two vane-type pumps arranged in parallel in
order to maintain a sample flow rate of 4 ft3/min
through the sample collection portion of the
train.
During a sampling run, particulates gradually
build up on the filter causing an increase in
vacuum at the pumps. If this vacuum becomes
too great, the desired flow rate cannot be main-
tained and the system must be shut down in
order to replace the filter. Incorporating two ad-
ditional modules in the train (scrubber and car-
bon adsorption modules) increases the total
pressure drop across the system.
A SASS train was obtained from the
manufacturer to quantify the effects of the add-
ed components on the system. Testing was ac-
complished by drawing room air through the
SASS train alone, SASS train with carbon in-
line, and the complete system (SASS plus car-
bon canister and' scrubber modules). Vacuum
hoses with an I.D. of 1/4 in were used to connect
the extra modules to the SASS train. Tests
were also performed to determine the pressure
drop across these lines. All vacuum readings
were taken from the gauges supplied with the
pumps, and gas flow rate measurements were
made using the gas meter and timer that are
part of the SASS train control unit. Before the
tests were conducted, a filter was placed in the
filter holder, three of the impingers were each
filled with 750 mL of tap water, and the fourth
impinger was charged with silica gel. XAD-2
resin was placed in the sorbent cartridge
assembly. Results of these tests are presented
in Table 9.
Proper operation of the cyclones is dependent
on the sample gas flow rate through the system,
with 4 ft3/min being the optimum design flow
rate. At this rate, a typical test run collecting
1,000 ft3 of sample has an approximate duration
of 4.5 hr. Depending on particulate loading in
the gas stream, it may become impossible to
maintain a 4-ft3/min flow rate through the modi-
fied SASS train (scrubber and carbon modules
in line); however, the only problem this presents
is an extended sampling period. For the pur-
poses of the screening procedures, it is not ab-
solutely necessary to maintain the 4-ft8/min flow
rate.
The sample flow piping in the standard train
is 1/2-in I.D.; it is recommended that this size
tubing be used for the design of the actual
screening train to eliminate the pressure drop
caused by the smaller diameter tubing. The
modular construction of the entire screening
train makes it a simple matter to add or delete
components or rearrange the sequence of any of
451
-------
TABLE 9. SCREENING TRAIN PRESSURE DROP TESTING
Standard SASS
Flow Rate
(cfm)
4.0
Vacuum
(in. Hg)
8.5
Flow Rate
(cfm)
3.0
Vacuum
(in. Hg)
6.0
Scrubber and Connecting Lines
Connecting Lines (Only)
Scrubber
4.0
4.0
4.0
8.5
6.5
2.0
3.0
3.0
3.0
5.0
4.0
1.0
Carbon Columns and Connecting Lines 4.0
Connecting Lines (Only) 4.0
Carbon Columns 4.0
5.0
4.5
0.5
3.0
3.0
3.0
4.0
3.5
0.5
TOTAL SYSTEM
(Standard SASS with both scrubber
and carbon columns on-line)
4.0
18.5
3.0
3.7
9.0
15.0
the units, depending on prior knowledge of the
gas stream constituents and/or the desired ap-
plication of the train at a particular source.
Preliminary calculations indicated that 8 L of
scrubbing solution (1-Normal sodium carbonate)
would be required to scrub 1,000 ft8 of sample
with an H2S concentration of approximately
2,000 ppmv. Additional calculations indicated
that 5 Ib of activated carbon would be adequate
for removal of organic compounds expected in a
waste gas stream. To verify these calculations,
the special gas blend with the following compo-
sition was utilized:
• Carbon dioxide 70 percent
• Nitrogen 29.55 percent
• Hydrogen sulfide 2,000 ppmv
• Ethylene 2,500 ppmv
Two gas cylinders were required to obtain this
blend, the first containing the N2, H2S, and
C2H4, and the second containing the C02. Flow
rates from both cylinders were monitored by
the use of rotometers and dry gas meters and
were adjusted to obtain the desired final gas
composition (Figure 10).
The gases were first introduced into a mixing
chamber where initial samples were taken to
determine both H^ and total hydrocarbon con-
centrations. From the mixing chamber, the
gases then flowed through the scrubber unit
and the carbon canister. Several test runs were
made on each unit separately, and one run was
conducted to determine H2S and hydrocarbon
removals with both units in series. Total hydro-
carbons were measured by taking a 100-cm8 gas
sample and injecting directly into a gas chroma-
tograph equipped with a flame ionization detec-
tor. Methane was used as the standardization
gas, and, therefore, the results are presented as
total hydrocarbons expressed as methane. Hy-
drogen sulfide levels were measured by draw-
ing a sample of the gas directly through H2S
detector tubes. Results of the testing are
presented in Table 10.
The results of pilot scrubber testing indicate
that 8 L of sodium carbonate scrubbing solution
will not be adequate when a 1,000-ft3 sample is
drawn that has an acid-gas concentration (H2S,
S02, etc.) of 2,000 ppmv or greater. It was
observed during the test period that the scrub-
ber solution became totally ineffective at a pH
of 10.0 or less. It is recommended that the solu-
tion concentration be increased to 2-Normal,
452
-------
©-SAMPLE POINTS
OPTIONAL SCRUBBER BYPASS
ROTOMETERS
MIXING
CHAMBER
SCRUBBER
GAS
CYLINDERS
GAS
METERS
CARBON
CANISTER
EXHAUST
TO
HOOD
Figure 10. Pilot scrubber and carbon adsorber test apparatus.
-------
TABLE 10. RUN #1 -SCRUBBING FOLLOWED BY CARBON ADSORPTION
Tine
(minutes)
0
25
60
90
105
120
150
Gas Volume
(cubic feet)
-
37.2
94.6
143.2
167.1
192.6
240.6
Ii
H2S (pi
2400
2400
2100
2200
2400
2200
2400
* (ppn as methane)
Inlet Concentration
Outlet Concentration
Removal
H.S (ppm) Total Hydrocarbon* H_S (ppm) Total Hydrocarbon* HJ* Total Hydrocarbon
1060
1250-
-
-
-
-
_
5
10
40
100
240
500
1250
1000
1275
-
-
-
-
_
99.8
99.6
98.1
95.4
90.0
77.2
47.9
5.7
-------
01
Ol
«-J
s
3
J
R
f^
^
-
k — *
STANDARD SASS TRAIN
CYCLONES
AND
FILTER
COOLER
AND
XAD-2
IMPMGERS
PUMPS
METER
CONTROLS
SCREENING TRAIN -OPTION No.1
1 I i i
I ii i
•RMrnajLATEL-J CAS i
r" ~ REMOVAL j^ COOLING j^
i II 1
1— J L -1
CARBON
ADSORPTION
SCREENING TRAIN - OPTION No.
i — - -i i -i
i it i
i ii i
• ^PARTICULATEt ^ GAS i
"] REMOVAL | | COOLING |
1 ' ' '
i _J L_ _ J
SCRUBBING
SCREENING TRAIN - OPTION No.
r~ 1 r- - - i
: ; i ;
I ^JPARTICULAT^ . GAS i
1 REMOVAL | | COOLING |
• : : ;
i__ — — j i. _ _ .. _ j
SCRUBBING
2
3
l»
(3
U
(SL)
CARBON
ADSORPTION
XAO-2
GAS
DRYING
AGENT
PUMPS
METER
CONTROLS
XAD-2
IMPINGE RS
PUMPS
METER
CONTROLS
XAD-2
IMPINGERS
PUMPS
METER
CONTROLS
OPTIONAL - DEPENDS ON PROCESS INFORMATION
Figure 11. Screening train options.
-------
and that the total volume available in the reser-
voir be increased to 16 L. As an extra precau-
tion, a pH meter should be used to monitor the
condition of the scrubbing medium. If it is neces-
sary to halt the run for a filter change at any
time during the test, the scrubbing solution
should also be replaced at that time.
Removal of ethylene from the test gas stream
by activated carbon was very poor. It is not
known whether this was due to an inherently
low adsorption capacity for this compound onto
the test carbon, or if the large quantity of car-
bon dioxide present in the stream resulted in
flushing the ethylene through the system. Or-
ganics with higher molecular weights stand a
much better chance of being adsorbed on the
carbon and, for this reason, it is recommended
that the carbon module be retained in the
screening program. It is not practical to sub-
stantially increase the amount of carbon used in
the screening train because the train already
consists of many modules large enough to pre-
sent problems when the sample location is dif-
ficult to reach, and space at the sample point
will be restricted in most cases. The screening
procedure for carbon during Level 1 may be
somewhat limited, but will, nevertheless, be in-
dicative of the potential of the process for
removing organic contaminants and will serve
as a guide for future studies.
In order to obtain meaningful results from
the tests, it is imperative that each source to be
evaluated be sampled according to the Level 1
IERL methods, in addition to the screening
sampling. Ideally, both tests will be run simul-
taneously. If this ia not possible, process data
for each source must be evaluated to determine
the constancy of operation, and judgment must
be used to assess the reliability of comparing
data from two nonsimultaneous test runs.
The 3-L grab samples will be taken as shown
in Figure 11. In addition, an optional sample of
1,360 L will be taken at these sample points for
use in the stress ethylene test. This sample is
listed as optional at this time pending modifica-
tions of the analytical procedure.
Conclusions and Recommendations
A summary of conclusions and recommenda-
tions based on the laboratory work with simu-
lated waste gas is presented below:
• The screening procedures using scrubbing,
carbon adsorption, and condensation should
be adopted.
• Special supplemental scrubber and adsorber
modules will be required to be used in con-
junction with the 8ASS equipment.
• The supplemental modules increase the
pressure drop across the sampling system.
It is recommended that the sample flow rate
be reduced to 3 ft3/min (Level 1 IERL pro-
cedures specify 4 ft3/min for optimum opera-
tion of the particle sizing module).
• A 2.0 normal solution of sodium carbonate
will be used as the scrubbing media. This
solution should be replaced during the test
whenever the pH falls below 10.0 standard
units.
• Figure 11 shows the screening train options
available for air sampling.
REFERENCES
1. Singer, P. C., et al. Assessment of Cool Con-
version Wostewater: Characterization and
Preliminary Biotreatability. EPA 600/7/78-
181. p. 95.
2. Beychok, Milton R. Coal Gasification and the
Phenosolvan Process. (Presented at the
168th National Meeting of the American
Chemical Society, Division of Fuel Chemis-
try. Atlantic City. September 1974.) Volume
19, No. 5. p. 85-93.
456
-------
EVALUATION OF COAL CONVERSION
WASTEWATER TREATABILITY
Philip C. Singer,* James C. Lamb HI, Frederic K. Pfaender,
Randall G. Goodman, Randy Jones, and David A. Reckhow
University of North Carolina, Chapel Hill, North Carolina
Abstract
This paper describes preliminary results
from an experimental program that evaluates
biological treatability of coal conversion
wastewater. The experimental approach in-
cludes preparation of a synthetic wastewater
designed to simulate a practical coal conver-
sion discharge. Design and operation of four
biological reactors and the preliminary results
from the first few months of synthetic waste-
water treatment are described. Data analyzed
include chromatographic analyses of the waste-
water and reactor effluents, as well as cytotox-
icity analyses using Chinese hamster V79 cells.
INTRODUCTION
Most coal conversion technologies incor-
porate or project aerobic biological treatment
as the principal means of removing phenols and
other organic impurities from process waste-
waters. However, the nature and biodegrad-
ability of many of these other organic mate-
rials are not known, and the extent to which
they can be removed by biological treatment
cannot be reliably predicted. Synergisms and
antagonisms resulting from the complex na-
ture of real wastewaters are especially uncer-
tain. Because even well-operated biological
treatment processes typically remove only 85
to 95 percent of the influent BOD and a signifi-
cant portion of the wastewater organics may
not be biodegradable, biological treatment
alone may not provide an environmentally ac-
ceptable discharge. In view of these considera-
tions, a need exists to identify the nature and
characteristics of aqueous discharges from coal
conversion processes, assess their environmen-
tal impact, and develop satisfactory waste-
*Speaker.
water treatment so they may be disposed of in
an environmentally acceptable fashion.
In an earlier report, Singer et al.1 presented
the results of a literature review and survey
showing that the composition of wastewaters
from different coal gasification and liquefac-
tion technologies is relatively uniform, espe-
cially with regard to the phenolic constituents.
Phenol appears to be the major organic constit-
uent, and phenolics as a class constitute 60 to
80 percent of the total organic carbon (TOG) in
the wastewater. Other classes of organics, such
as mono- and polycyclic nitrogen-containing
aromatics, oxygen- and sulfur-containing
heterocyclics, and polynuclear aromatic hydro-
carbons, appear to be present at significant
concentrations. In this paper, the preliminary
results of a study directed at evaluating the
biological treatability of coal conversion waste-
water is presented.
APPROACH
Biotreatability studies require the use of ac-
climatized microbial cultures to insure ac-
curate evaluation of biological treatment sys-
tems and for preliminary assessment of key pa-
rameters in establishing the effectiveness of
such treatment. Meaningful assessment of po-
tential toxicity of wastewater constituents in
biological treatment is impossible unless the
test cultures have been acclimatized to the
wastewater in question.
Ideally, biotreatability studies should be
conducted using the specific wastewater for
which the treatment is being developed. In this
study, however, it is not feasible to use actual
wastewaters from coal conversion operations
because coal conversion processes are still in
the developmental stage and it is unlikely that
a suitable, consistent, and representative
wastewater could be obtained. Accordingly, a
synthetic organic wastewater was formulated
457
-------
to provide a mixture of organic compounds, at
known and reproducible concentrations, to be
used in acclimatizing and maintaining micro-
bial cultures for preliminary biotreatability
studies. The synthetic wastewater is used to
feed several bench-scale pilot reactors. In addi-
tion to generating acclimatized organisms for
biodegradability studies (to be reported else-
where), analysis of effluents from the reactors
provides information on wastewater character-
istics at various levels of biological treatment.
FORMULATION OF SYNTHETIC COAL
CONVERSION WASTEWATER
Several criteria were employed in choosing
specific compounds and their concentrations to
be included in the synthetic wastewater. Be-
cause this waste would be used as a means of
developing an acclimatized culture of microor-
ganisms, most of the compounds selected are
known or thought to be biodegradable. How-
ever, not all of the identified constituents of
coal conversion wastewaters can be used by
microorganisms. Accordingly, some com-
pounds presumed to be slowly degradable or
nondegradable, as deduced from earlier biode-
gradation experiments, ' were included (e.g.,
2-indanol, indene, 2-methylquinoline, and
3,5-xylenol).
When the composition of the synthetic
wastewater was formulated, it was desired
that concentrations of the various components
should be similar to those encountered in real
wastewaters. Accordingly, reference was
made to a summary of the constituents iden-
tified in coal conversion wastewaters1 and the
range, midrange, and median concentrations
were determined for each constituent and for
each class of compounds (e.g., cresols, xylenols,
heterocyclic N-compounds, etc.). From each
class, one or more compounds were chosen
based on biodegradability and reported con-
centration. The specific compounds chosen
were usually the compounds within each class
that were reported at the highest concentra-
tions in the real wastewaters. Often, if a class
contained many components, or if differences
in biodegradability among the components of a
given class were anticipated, more than one
chemical from that class was chosen. The con-
centration selected was the median value re-
ported for that compound in the real waste-
water, or the median of the class if only one
compound from that class was picked. When
the concentration of a specific compound
selected was not known, it was included in the
synthetic wastewater at the median concentra-
tion for its class.
Table 1 presents the composition of the
wastewater formulated in this manner. Twen-
ty-eight organic components are included, as
well as inorganic nutrients and pH-buffers. The
synthetic wastewater represents all major
classes of organics present in real wastewaters
for which data are available, and virtually all
specific organic compounds that have been re-
ported to be present at high concentration. The
total organic carbon (TOO concentration of all
the components is 4,636 mg/L.
DESCRIPTION OF PILOT UNITS
Four 25-L biological reactors were con-
structed for use in the initial phases of the pilot
program. Each reactor consists of a 7Vi in ID
lucite tube, 4 ft long, fitted at the bottom to a
stainless steel cone with a 45° slope (Figure 1).
Each reactor has overflow and sampling con-
nections located at appropriate heights to re-
tain the desired volume of contents in the reac-
tor and to permit withdrawal of samples from
desired elevations. The stainless steel cone is
equipped with connections to permit draining
of the unit and nipples for introducing air and
feed solution at the bottom of the cone.
A compressor, operating through a pressure
regulator, supplies air to each reactor at a rate
adequate to insure thorough mixing and main-
tenance of aerobic conditions in the mixed liq-
uor at all times. The rate of air supply is con-
trolled through the use of rotameters and
needle valves.
The units are fed synthetic wastewater from
a glass storage reservoir mounted on a large
magnetic mixer. The wastewater is fed to each
reactor by a variable-speed peristaltic pump.
The reactors are operated as continuous-flow
activated sludge systems with no recycle of
solids (biomass). Hence, solids residence time
or sludge age equals hydraulic detention time.
The pump feeding Reactor 1 (with a 5-day
hydraulic detention time) is operated con-
tinuously. Pumps supplying feed to the other
three reactors (operated at 10-, 20-, and 20-day
hydraulic detention times, respectively) are ac-
458
-------
TABLE 1. COMPOSITION OF SYNTHETIC COAL CONVERSION WASTEWATER
Compound Concentration, mg/1
1. Phenol 2000
2. Resorcinol 1000
3. Catechol 1000
4. Acetic Acid 400
5. o-Cresol 400
6. p-Cresol 250
7. 3,4-Xylenol 250
8. 2,3-Xylenol 250
9. Pyridine 120
10. Benzoic Acid 100
11. 4-Ethylpyridine 100
12. 4-Methylcatechol 100
13. Acetophenone 50
14. 2-Indanol 50
15. Indene 50
16. Indole 50
17. 5-Methylresorcinol 50
18. 2-Naphthol 50
19. 2,3,5-Trimethylphenol 50
20. 2-Methylquinoline 40
21. 3,5-Xylenol 40
22. 3-Ethylphenol 30
23. Aniline 20
24. Hexanoic Acid 20
25. 1-Naphthol 20
26. Quinoline 10
27. Naphthalene 5
28. Anthracene 0.2
theoretical ZTOC - 4636 mg/1
NH.C1 (1000 mg/1 as N) 3820
MgSO • 7H.O 22.5
CaCl* 27.5
FeNaEDTA 0.34
Phosphate Buffer: KH0PO, 170
435
• 7H.O 668
459
-------
SAMPLING PORT
OVERFLOW
REACTOR STAiJD
EXHAUST
SYSTEM
c
D -*
I^^H
^M
•
tt
1
«o >
^
p
\ .FEED SOLUTION
g VARIABLh bPttU
1 PFRT.qTAI Tf PUMP
T^lf /*^AIRSUPPLY
£.3 *J KOTAnETER
PLEXIGLASS REACTOR
U ri ACC CCCn TIIRP
ULMOC) FLLU 1 UDC.
1
^•STAINLESS STEEL CONE
Figure 1. Schematic of experimental biological reactors.
460
-------
tuated by a clock that operates them for a pre-
determined time once every half hour. (Two re-
actors operate at the same 20-day detention
time to allow one reactor to be isolated for use
as a chemostat to provide seed organisms for
parallel biodegradation investigations; the
other 20-day reactor is used with the 5- and
10-day reactors to provide operating data to
characterize reactor performance as a function
of solids residence time.) Overflow from each
reactor is collected in a glass reservoir and the
amount of wastewater actually fed is deter-
mined daily by measuring the amount of ef-
fluent collected in that container.
Because of the potential hazard of some
chemicals in the wastewater and the need to
eliminate objectional odors in the working
area, an exhaust system was installed to vent
the units continuously to the outside of the
building. The exhaust system consists of a
blower mounted at the outside wall, thereby
maintaining the air ducts under a slight vac-
uum to insure that gases from the reactors
always flow into the exhaust system and not in-
to the room. The feed reservoir is also vented
to the exhaust system to prevent the escape of
gases from that unit into the room.
OPERATION OF PILOT UNITS
The synthetic wastewater is made up in 16 L
batches. Carbon-filtered Chapel Hill, North
Carolina, tap water is used as dilution water to
which the 28 constituents, shown in Table 1,
are added. This is accomplished by adding ap-
propriate quantities from concentrated stock
solutions which are prepared periodically from
reagent grade chemicals and stored under re-
frigeration until use. In order to prepare some
of the concentrated solutions containing com-
pounds of limited aqueous solubility, an
organic solvent was required to maintain sol-
ubility in the stock solutions. Accordingly,
acetone was employed for this purpose. While
this introduced an extra constituent into the
wastewater, it was believed that much of the
acetone would be removed through air strip-
ping during the long detention times in the
reactors. Hence, the TOC concentration of the
raw wastewater is actually somewhat higher
than that shown in Table 1.
The reactors were started up using acti-
vated sludge from one of the Durham, North
Carolina, municipal wastewater treatment
plants. The feed of synthetic wastewater was
increased gradually over a period of several
days to allow time for acclimatization of the
microorganisms to the wastewater. However,
during the first few weeks after startup, all of
the units began to fail as evidenced by in-
creased TOC concentration in the effluents and
decreased solids concentration in the reactors.
Failure occurred first in the 5-day reactor, then
in the 10- and 20-day reactors. The exact reason
for failure is unknown, but several possibilities
have been considered. Operating procedures
during the early stages of the investigation
were uncertain and made it possible for the
concentration of dissolved oxygen in the reac-
tors to drop occasionally to zero. Also, the pH
decreased to low levels (approximately 4.0) and
remained there for extended periods. Further,
there is a possibility that some wastewater
constituents could have exerted a toxic effect
on the microorganisms as concentrations of the
constituents built up in the reactor during the
period following startup. The pattern of fail-
ure, in order of increasing reactor detention
time, is consistent with the latter hypothesis.
Because of the possibility of toxic effects and
a desire to stabilize operations as quickly as
possible, it was decided to reduce the strength
of the synthetic feed during these initial in-
vestigations to one-quarter of that listed in
Table 1. Other investigators have had to resort
to similar dilution procedures in order to treat
coal conversion wastewaters biologically. The
resulting diluted version, with a theoretical
TOC of 1,159 mg/L, is not inconsistent when
compared with biotreatability experiments
being conducted by others. (The concentration
of TOC measured in the feed averaged 1,600
mg/L over the course of the runs because of the
addition of acetone to solubilize the organic
constituents in the feed.) At a later date, the
question of treating the synthetic wastewater
at higher strengths will be addressed. Accord-
ingly, the reactors were started up again using
a synthetic wastewater diluted to one-quarter
of the concentration specified in Table 1.
A significant change in the color of the syn-
thetic feed solution was observed over the
several days during which it is used to feed the
reactors. Attempts have been made to deter-
mine possible changes in wastewater composi-
tion during this time through periodic meas-
461
-------
urements of TOC and chromatographic scans
using high performance liquid chromatography
(HPLC). Chemical changes accompanying the
change in color from clear to brown appear to
be minimal.
Routine sampling of each reactor is per-
formed three times a week. Parameters meas-
ured include temperature, pH, mixed liquor
suspended solids (MLSS), mixed liquor volatile
suspended solids (MLVS8), sludge volume in-
dex (8VI), and TOC. PH is measured potentio-
metrically. MLSS concentrations are deter-
mined using glass fiber filters in a Buchner fun-
nel, followed by drying of the filter in an alum-
inum dish at 103° C for 24 hr. Filtrates from
MLSS analyses are collected for TOC determi-
nations using a Beckman 915 Carbon Analyzer.
SVI is determined by allowing mixed liquor
from the reactors to settle for 30 min in a 1-L
graduated cylinder and calculating the settled
volume occupied by the MLSS.
Other samples are collected as desired for
the measurement of biochemical oxygen de-
mand (BOD), chemical oxygen demand (COD),
and for more detailed analyses such as specific
organic compounds using HPLC and GC/MS,
aquatic bioassays, and assessment of health ef-
fects. BOD and COD analyses are conducted on
samples from which suspended materials have
been removed through glass fiber filtration.
Samples for HPLC and GC/MS analysis and for
aquatic bioassay and health effects assessment
are centrifuged, filtered, and frozen.
PRELIMINARY RESULTS
Figures 2 through 5 show performance char-
acteristics for each reactor over the period
from May to October 1978. The reactors oper-
ated without serious incident from the begin-
ning of May to the middle of June. The opera-
tional data suggested that they had reached
approximate steady-state performance, and in-
tensive data collection for this pattern of oper-
ation was initiated in early June. Five sets of
filtered samples from the reactors were ana-
lyzed for BOD, COD, nitrogen species, and
phosphorus, as shown in Table 2.
It had been planned that the analyses would
be continued at intervals of 2 days over a
period of at least 2 weeks. If the data then in-
dicated that steady-state had been attained, in-
tensive sampling would have been discontin-
ued and the operations modified to another set
of reactor conditions. During the intensive
sampling period in early June, however, the
data for TOC and MLSS indicated clearly that
steady-state operation had not been attained.
Effluent TOC in all of the reactors rose sharp-
ly, beginning about June 9, leading to a deci-
sion to postpone the intensive analysis pro-
gram until a more consistent performance
could be achieved.
The exact cause for the substantial change in
performance that occurred in June is unknown.
However, a short time earlier the time clock
controlling the feed to the reactors malfunc-
tioned, resulting in an overfeed of Reactors 2,
3, and 4.
During July, August, and September, MLSS
and TOC data indicated a reasonably steady
performance, with the possible exception of
Reactor 1 (5-day hydraulic detention time),
which had performed irregularly since startup.
In all units there was a pronounced tendency
for pH to drift downward during this period, al-
though the change in pH did not appear to af-
fect the stability of the MLSS and effluent
TOC. Accordingly, additional samples were
taken during September for detailed chemical
analysis, as shown in Table 2. Because of its er-
ratic performance, Reactor 1 was not sampled
intensively during this period. Reactors 2, 3,
and 4 produced very low effluent BODs, indi-
cating that almost all of the biodegradable ma-
terial had been removed. The COD reductions
were consistent with the reduction in TOC ex-
hibited in Figures 2 through 5. The nitrogen
and phosphorus measurements indicated that
nutrients were sufficient for biological activity
and that microbial growth was not inhibited by
a lack of nutrients. The distribution among the
nitrogen species showed that no nitrification
took place.
Although the performance of the reactors
appeared to be reasonably consistent during
the September sampling period, the pH was
unstable and continued to drift downward, in-
dicating clearly that steady-state operation
had not been attained. During October, the pH
in the reactors reached levels lower than 4.0,
causing concern about reactor stability. This
concern was compounded by sharp rises in ef-
fluent TOC following loss of aeration for
several hours because of compressor failure.
462
-------
/IM
50(1
I'll)
TOTAL ORGANIC CARBON
IN EFFLUENT
10 20
APPIL
10 20
KAY
JUNE
10 20
IUU
10 20
AUGUST
10 20
SEPTEMBER
iO 20
OCTOBER
tHOG
1200
1000
400
'.Illl
41 )i I
an
n
10 20
flARCM APRIL
MIXED LIQUOR SUSPENDED SOLIDS
1C 20
SEPTEMBER
10 20
OCTOBER
I 10 20 I
REACTOR PH
MARCH APRIL
10 20
IWY
10 20
JU1E
10 20
JULY
10 20
AUGUST
10 20
SEPTEHBER
10 20
OCTOBER
Figure 2. Performance characteristics of Reactor 1 with 5-day residence time.
463
-------
350
300
250
§ »
X 150
100
SO
i)
OVERFEED
I!
LOSS OF
«RATIOH
TOTAL ORGANIC CARBON
IN EFFLUENT
MARCH
10 tQ 1 10 <0
APRIL MY
10 20
JUNE
Mil
10 20
AUGUST
SEPTEMBER
OCTOBER
1WO
1200
1000
800
600
400
200
0
MIXED LIQUOR SUSPENDED SOLIDS
MARCH
10 20
APRIL
10 20
NAY
JU i
JUNE
JULY
10 20
AUGUST
10 23
SEPTEMBER
13 20
OCTOBER
REACTOR PH
MARCH
10 20
APRIL
10 20
MAY
13 20
JUNE
10 20
JULY
10 20
AUGUST
10 20
SEPTEMBER
10 20
OCTOBER
Figure 3. Performance characteristics of Reactor 2 with 10-day residence time.
464
-------
V*)
•;oo
?so
/no
I'jfl
100
OVERFEED
II
LOSS OF
AERATION
U
TOTAL ORGANIC CARBON
IN EFFLUENT
10 20 ' 10 20
,1/\RCH APRIL MAY
10 20
JIM
19 20
JULY
10 20
AUGUST
10 20
SEPTEMBER
10 20
OCTOBER
Klllil
sjn
MIXED LIQUOR SUSPENDED SOLIDS
'KID
/'fl'l
11 || II 1 1 1 1 | II
io a 10 20 n 20 10 20 ' ^ :o 10 20 \ :o 20
MAKUl APRIl 1AY JUNE INLY A"GUST SEPTEMBER OCTOBER
REACTOR PH
I 10 20
MARCH APRIL
10 20
HAY
10 20
JUNE
10 20
JULY
13 ZO
AUGUST
10 20
SEPTEMBER
10 20
OCTOBER
Figure 4. Performance characteristics of Reactor 3 with 20-day residence time.
465
-------
50(1
ft
mi
I'M
mo
'X.
OVfRFlEU
II
LOSS OF
AtMIIOd
TOTAL ORGANIC CARBON
IN EFFLUENT
I 10 20
MUCH APRIL
10 20
NAY
10 20
JU:IE
10 20
IDLY
10 20
AUGUST
10 20
SEPTEMBER
10 20
OCTOBER
Ituo
1200
inno
81)0
&no
'HID
MIXED LIQUOR SUSPENDED SOLIDS
WO!
10 20
AMU
10 20
HAY
10 20
IITIE
10 20
JULY
10 20
AUGUST
10 20
SEPTEMBER
10 20
OCTOBER
REACTOR PH
lTll20™l'l™10120~™l™lf"™
MAY JUNE JULY
tARCII
10 20
APRIL
10 20
AUGUST
10 20
SEPTEMBER
10 20
OCTOBER
Figure 5. Performance characteristics of Reactor 4 with 20-day residence time.
466
-------
TABLE 2. SUMMARY OF REACTOR PERFORMANCE:
MAY TO SEPTEMBER, 1978
Date
5/30
6/5
9/12
5/30
6/1
6/3
6/5
6/7
5/30
6/1
6/3
6/5
6/7
9/8
9/12
9/14
9/16
9/18
5/30
6/1
6/3
6/5
6/7
9/8
9/12
9/14
9/16
9/18
5/30
6/1
6/3
6/5
6/7
9/8
9/12
9/14
9/16
9/18
Sample
Raw Waste
H it
ii it
Reactor 1
it
ii
ii
it
, Reactor 2
it
ii
it
ii
it
it
it
H
it
Reactor 3
it
it
ii
it
ii
it
it
it
it
Reactor 4
ii
it
it
it
ti
it
ii
ti
ii
TOG
mg/1
430
399
463
469
521
95
93
98
130
143
90
112
112
116
119
47
64
65
70
70
34
51
47
53
57
57
59
57
99
123
39
53
51
54
56
BOD
mg/1
3520
2880
4140
1115
870
960
1055
1100
179
140
171
245
240
-43
26
25
33
47
30
45
80
52
5
7
8
7
73
18
38
170
183
5
5
4
7
COD
mg/1
5880
5800
5450
1600
1648
1728
1744
2112
400
360
488
532
616
275
315
330
320
340
348
352
400
368
190
180
210
190
292
280
356
496
552
____
200
195
220
240
NO, NO NH.
mg/1 mg/1 mg/1
as N as N as N
<0.03 11.0 243
0.005 3.3 228
228
0.12 6.8 209
0.064 2.0 234
222
0.05 5.6 222
0.07 5.5 217
231
0.07 5.6 225
0.07 3.2 247
249
0.06 4.4 240
TKN Total
mg/1 Phos.
as N mg/1
239 423
243 68
273
370 42
231 106
243
330 35
242 369
246
330 42
244 435
254
290 50
Ortho .
Phos.
mg/1
^«^
46
50
99
38
333
41
400
51
467
-------
Accordingly, in late October thii series of ex-
periment! was terminated.
DISCUSSION OF PRELIMINARY RESULTS
Overall performance of the units from March
through October may be summarized with a
few pertinent observations. All of the reactors
have shown excellent TOG removals from the
feed level of approximately 1,600 mg/L. Fig-
ure 6 summarizes TOG removal data for the
months of July, August, and September before
major excursions in pH were experienced.
With 5-day detention, Reactor 1 was capable of
producing an average effluent TOG of about
200 mg/L, with a range extending from about
80 to 300 mg/L. Reactor 2 (10-day detention)
produced an average effluent BOD of about 80
mg/L, with more consistent performance as
shown by the narrower range of approximately
60 to 120 mg/L. Reactors 3 and 4 (both with
20-day detention) performed in substantially
identical fashion, with effluent TOCs averag-
ing 45 mg/L and a rather narrow operating
range of approximately 40 to 60 mg/L. Table 3
summarizes the average performance of the
reactors for the months of July, August, and
September, taken from the data in Figures 2
through 5 and Table 2.
Kinetic Analysis
In order to design an activated sludge proc-
ess for treatment of coal conversion waste-
water, the parameters describing the kinetics
of microbial growth and substrate utilization
for the given wastewater mutt be determined.
The data collected to date can be used to make
a preliminary determination of these requisite
microbial growth coefficients as follows:
The kinetics of microbial growth can be
described by the equation*
dx/dt -yda/dt-kdx
(1)
where:
x « concentration of microorganisms (bio-
mass) in mg of MLSS per L;
s - substrate concentration, in mg per L, on a
BOD, COD, or TOG basis;
t - time, in days;
y - microbial yield coefficient, in mg of bio-
mass (MLSS) produced per mg of sub-
strate (on a BOD, COD or TOG basis) con-
sumed;
kj - microbial die-away coefficient, in days'1.
Taking finite differences in equation (1) and di-
viding through by xT the mean biomass concen-
tration over the time period At, yields
(Ax/At)/x~- y (As/AtVJT- k*
(2)
For the continuous-flow, completely-mixed reac-
tors used in this investigation, x~is the steady-
state biomass concentration in each reactor, and
At is the detention time of the reactor. Equation
(2) can be rewritten as
(3)
TABLE 3. AVERAGE QUALITY OF EFFLUENT FROM BIOLOGICAL
TREATMENT UNITS (ALL VALUES IN mg/L)
Raw
Waste
BOD 3510
COD 5710
TOG 1600
MLSS
Reactor Detention
5 10
1020 32
1770 310
200 80
700 900
Time
20
7
192
45
950
(Days)
20
5
214
45
900
468
-------
200
CJ
o
I
100
REACTOR NO, 1
LEGEND:
AVERAGE
APPROXIMATE
RANGE
REACTORS 3 &
0
0
10
DETENTION TIME, 9. (DAYS)
15
20
Figure 6. Effect of residence time on reactor performance and stability.
-------
Here, de can be defined as the mean cell
residence time, solids retention time, or sludge
age, and is equal to the steady-state quantity of
biomass in the reactor, divided by the rate of
biomass production. 0C has units of time and for
reactor operation with no recycle of biomass,
the solids residence time is equal to the hy-
draulic retention time. The quantity U in equa-
tion (3) is defined as the process loading factor,
or food to microorganism ratio, and is equal to
the quantity of substrate consumed during the
given reactor detention period divided by the
steady-state biomass concentration (compare
equations (2) and (3)). The process loading fac-
tor can be computed on a BOD, COD, or TOG
basis. If the reciprocal of the sludge age is
plotted against the process loading factor in
accordance with equation (3), a straight line
should result and the microbial kinetic coeffi-
cients y and k
Uc>
V
Qct Days
mg BOD/mg MLSS-day
mg COD/mg MLSS-day
mg TOC/mg MLSS-day
5 10 20
0.71 0.39 0.18
1.13 0.60 0.29
0.40 0.17 0.082
20
0.19
0.31
0.08
470
-------
0.2 -r-
0.15 -•
>,
«
T3
CD
0.05 ..
Y = 0.27 mg MLSS / mg BOD
0.2 0.4 0.6 0.8
Figure 7. Effect of solids residence time on BOD loading.
cd
•o
0.2 _^
0.15 - -
0.1 - •
0.05 - -
Y = 0.18 mg MLSS / mg COD
0.75 1.00 1.25
Figure 8. Effect of solids residence time on COD loading.
471
-------
0.2 ..
0.15 ..
0.1 ..
ffl
•o
©
0.05 ..
Y - 0.52 mg MLSS / mg TOG
0.1
0.2
0.3
.4
Figure 9. Effect of solids residence time on TOC loading.
polar cellular metabolites of these compounds.
More complete analyses are necessary to quan-
tify the removal of the raw synthetic waste-
water constituents as a function of residence
time in the biological reactors and to ascertain
the nature of the components comprising the
residual peaks.
These chromatograms have been compared
with others using 254 nm UV absorbance de-
tection and simultaneous fluorescence detec-
tion at 275 nm excitation and 310 nm emission
wavelengths. (Fluorescence spectrophotom-
etry combined with HPLC is a much more sen-
sitive and selective detection technique than
simple UV absorbance.) From the relative
responses of each peak, these comparisons in-
dicate that very little of the residual organic
material is phenolic. This is important from the
standpoint of reactor performance because a
large portion of the organic carbon in the reac-
tor feed is comprised of phenolic compounds.
HPLC traces of the reactor effluents were
used to obtain approximate concentration val-
ues for several of the major constituents fed to
the reactors. These data are given in Table 6.
The maximum effluent concentrations listed in
Table 6 should be interpreted with great care
because they have been calculated by assum-
ing that a particular chromatographic peak is
caused entirely by the specific compound in
question. It is more likely, however, that each
peak is due to several compounds. Therefore,
472
-------
oo
Figure 10. HPLC chromatograms of raw synthetic feed and reactor effluents.
-------
TABLE 5. IDENTIFICATION OF HPLC CHROMATOGRAPHIC PEAKS FOR RAW FEED
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
Acetic Acid, Benzole Acid, Hexanoic Acid
Solvent
Acetone
Resorcinol
Catechol
Aniline
Phenol
5-Methylresorcinol
4-Methylcatechol
Unidentified
Unidentified
p-Cresol
o-Cresol
2-Indanol
Acetophenone
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
4-Ethylpyridine
Pyridine
Quinoline
3, 4-Xylenol
3, 5-Xylenol
2, 3-Xylenol
Indole
3-Ethylphenol
2-Methylquinoline
1-Naphthol
2-Naphthol
2, 3, 5-Trimethylphenol
Indene
Naphthalene
Anthracene
the actual effluent concentrations are probably
much less that those listed in the table. Recent
studies on the reactor effluents using fluores-
cence/HPLC have indicated that effluent con-
centrations of the compounds listed in Table 6
are probably much less than those reported
there.
Cytoxicity Analysis
A clonal toxicity assay, employing the
Chinese hamster V79 cell line, was used to com-
pare the relative acute toxicities of the ef-
fluents from the biological reactors and the
raw synthetic wastewater. This assay meas-
ures the colony forming ability of cells exposed
to toxicants. The purpose of this test was to
evaluate the effectiveness of biological treat-
ment in alleviating potential human health ef-
fects associated with coal conversion waste-
waters.
Effluent samples were collected from Reac-
tors 2,3, and 4 on September 17,1978, and from
Reactor 1 on October 28, 1978. The samples
were centrifuged and then filtered through a
series of Nuclepore polycarbonate filters con-
sisting of a 1.0-pm prefilter and a 0.2-^m
ultimate filter. The filtrates were collected and
aliquoted in small glass prescription bottles,
which were then frozen and stored at - 80° C.
A sample of the raw synthetic wastewater,
which had been aged for 2 days, was collected,
treated, and stored in a similar manner. Indi-
vidual aliquots of frozen reactor effluents and
474
-------
TABLE 6. REMOVAL OF SELECTED CONSTITUENTS BASED ON HPLC-UV ABSORBANCE ANALYSIS
tffc
-J
01
Compound
Resorcinol
Aniline
Phenol
p-Cresol
o-Cresol
Pyridine
Qu incline
Xylenols
Feed
Concentration
mg/1
250
5
500
62.5
100
30
2.5
135
REACTOR
Maximum
Effluent
Concentration
mg/1
1.24
0.6
4.2
8.0
2.6
0.6
3.5
2
Minimum
% Removal
99.5
87.8
99.2
87.2
97.4
98.2
97.4
REACTOR
Maximum
Effluent
Concentration
mg/1
1.2
0.4'
6.6
5.1
1.2
0.4
1.4
3
Minimum
TL Removal
99.5
92.2
98.7
91.9
98.8
98.8
99.0
REACTOR
Maximum
Effluent
Concentration
mg/1
1.2
0.4
6.7
5.2
1.5
0.5
1.7
4
Minimum
% Removal
99.5
92.8
98.7
91.6
98.6
98.5
98.7
-------
raw feed were thawed immediately prior to
their use and the remainder of that aliquot was
discarded at the end of the day.
A series of dilutions of each wastewater was
made in distilled-deionized water. The addition
of 2 x or 4 x nutrient medium to the dilution
tubes maintained physiological conditions at
final test concentrations ranging from 0.25 to
75 percent of the wastewater sample being
tested. Two hundred cells were seeded per
60 mm of tissue culture dish and allowed to in-
cubate and attach for 3 hr in 3 mL of normal
cell growth medium. Duplicate dishes were
then treated with appropriate dilutions of a
test wastewater. Each pair of dishes received a
single concentration of the test materials.
After an exposure period of 20 hr, growth
medium containing the test materials was
removed. The cells were washed once in a
phosphate buffered saline solution and rein-
cubated in 3 mL of fresh growth medium. Ex-
posed single cells were allowed to grow into
colonies and were then fixed and stained after
7 days. The number of colonies for each ex-
posure condition was calculated as a percent of
the number of colonies in untreated control
plates, and expressed as the relative plating ef-
ficiency.
The results of the clonal toxicity assay are
shown in Figure 11, where concentration-de-
pendent survival curves have been plotted us-
ing the average of the data points from dupli-
cate clonal toxicity experiments. The concen-
trations indicated represent dilutions of the
samples being tested. Concentrations produc-
ing 50 percent lethality (LCsg) are shown in
Table 7, along with the corresponding TOG
concentrations. As indicated in Figure 11 and
Table 7, V79 cytotoxicity decreases with in-
creasing degree of wastewater treatment as
measured by residence time.
It is interesting to note in Table 7 that while
Reactor 1 provided an 87.5-percent reduction
in TOC compared to the raw wastewater, the
LCjjo was reduced only three-fold. This sug-
gests several possible explanations. Most of
the easily degradable TOC may not be very
cytotoxic. On the other hand, it is possible that
a reduction in TOC below certain threshold
levels, which occurs in the reactors with longer
detention times, accounts for the observed
changes in cytotoxicity. The 95-percent TOC
TABLE 7. SUMMARY OF MAMMALIAN CYTOTOXICITY DATA
Sample
Raw Wastewater
No. 1-5 day residence time
No. 2 - 10 day residence time
No. 3-20 day residence time
No. 4 - 20 day residence time
TOC,
mg/1
1600
200
80
45
45
LC
%
150'
1.0
3.0
23.5
80*
80*
*Reactors 3 and 4 did not produce 50% lethality
at the highest concentrations tested (75%).
The LC,Q values shown are extrapolated from
the plots in Figure 11.
476
-------
KEY:
• Synthetic feed
DReactor 1 effluent
+Reactor 2 effluent
oReactor 3 effluent
A Reactor 4 effluent
1 2
10 20 30 40 50 60 70 ' 80
Wastewater Concentration (%)
Figure 11. Results of 20-hr V79 Chinese hamster clonal toxicity assay.
90 95
98
-------
reduction produced by the 10-day reactor cor-
responded to a 23-fold reduction in cytotoxic-
ity, while the 97-percent reduction in TOG pro-
duced by the 20-day reactors corresponded to
an 80-fold reduction in cytotoxicity compared
to the raw wastewater.
CONCLUSIONS
A synthetic coal conversion wastewater,
representative of wastewaters from coal gasifi-
cation and liquefaction processes, has been pre-
pared. The wastewater appears to be biologi-
cally treatable, but some degree of dilution
may be necessary. Biological treatability, as
measured by BOD, COD, and TOC removal, im-
proves with increased solids residence time
(sludge age), but it appears that a minimum
sludge age of 10 days may be necessary to
achieve a reasonable degree of treatment. A
mammalian cytotoxicity assay, used as an in-
dicator of potential human health effects
associated with the wastewater, shows that
cytotoxicity decreases with increasing degrees
of biological wastewater treatment.
Due to continued difficulties with pH varia-
tions, recent changes in the character of the
synthetic wastewater have been made to pro-
vide additional buffer capacity and to eliminate
acetone in preparing the synthetic feed. It does
not seem appropriate to develop more detailed
conclusions at this interim point in the experi-
mental program. Continued operation of the
reactors should lead to more stable perform-
ance in the near future, allowing detailed
analysis of performance and operating param-
eters and more conclusive results.
ACKNOWLEDGMENTS
The authors would like to thank Dr. Dean
Smith and Dr. Thomas Petrie of the Industrial
Environmental Research Laboratory (Re-
search Triangle Park) of EPA for their as-
sistance in the performance of this research,
and to EPA for sponsoring this project.
REFERENCES
1. Singer, P. C., F. K. Pfaender, J. Chinchilli,
A. F. Maciorowski, J. C. Lamb III, and
R. Goodman. Assessment of Coal Conver-
sion Wastewaters: Characterization and
Preliminary Bio treatability. U.S. Environ-
mental Protection Agency. Washington,
D.C. EPA-600/7-78.181. September 1978.
2. Metcalf and Eddy, Inc. Wastewater Engi-
neering, McGraw Hill Book Co., 1972.
3. Luthy, R. G., and J. T. Tallon. Biological
Treatment of Hygas Coal Gasification
Wastewater. U.S. Department of Energy.
Washington, D.C. FE-2496-43. December
1978.
478
-------
CONTROL TECHNOLOGIES FOR PARTICULATE AND
TAR EMISSIONS FROM COAL CONVERTERS
D. M. Kennedy
Dynalectron Corporation, McLean, Virginia
L. Breitstein**
Booz, Allen, and Hamilton, Inc., Bethesda, Maryland
and
C. Ghent
J.R.B. Associates, Inc., McLean, Virginia
Abstract
Raw product gases from coal converters gen-
erally contain particulates and tars that must be
controlled to a level compatible with environ-
mental regulations and process and equipment
requirements. Alternate control technologies for
removing particulates and tars from product
gases were identified and evaluated.
Paniculate and tar emissions in raw product
gases from several types of coal gasifiers were
characterized in terms of their total quantities,
chemical composition, and size distribution. The
emissions data were organized and summarized
according to generic gasifier type, with fixed-,
fluid-, and entrained-bed gasifiers being con-
sidered. The design and operating features of
each identified alternate control technology
were described, with emphasis on characteriz-
ing collection efficiencies as a function of parti-
cle size and other important parameters. These
data were also organized into generic categories
such as cyclones, wet scrubbers, electrostatic
precipitators, fabric filters, and granular bed
filters.
The applicability of each of the identified con-
trol technologies was assessed with respect to
the generic gasifier types and various end uses.
These assessments were based on existing and
proposed environmental regulations and proc-
ess requirements for product gas purity. End
uses considered include combined cycles and
gas-fired boilers. The fate of the particulate and
tar emissions from the various gasifiers was
assessed in terms of their presence in the puri-
*Speaker.
'Formerly of Dynalectron.
fied product gases, liquid effluents, and solid
wastes or sludges. In addition, gaps in the data
base required for these assessments were identi-
fied.
INTRODUCTION
The energy supply problems of the United
States and most of the major industrialized na-
tions are well known and well documented. Cur-
rent projections indicate that the world demand
for petroleum and natural gas will exceed sup-
ply sometime during the 1980's. One obvious ap-
proach to increasing domestic fuel supplies, and,
consequently, to reducing demand for imported
gas and oil, is to utilize the vast coal resources of
the United States to produce synthetic oil and
gas.
In recent years, the electric utility and indus-
trial sectors of the economy together accounted
for about 55 percent of the energy consumption
in the United States. Natural gas and petroleum
supplied about 80 percent of the industrial
energy consumption and 30 percent of the utili-
ty consumption. The use of coal-derived fuels to
replace natural gas and petroleum in these
areas could benefit the United States economic-
ally, in addition to reducing the nation's
dependence on foreign, unreliable sources of
energy. Such coal-derived products might be
employed in a wide variety of end uses, such as
industrial process heat, industrial and utility
boilers, gas turbines, and reducing or synthesis
gas for various industries.
In the case of product gases from coal gasi-
fiers, each particular end use for the gases
would have different environmental regulations
and/or process requirements governing the
479
-------
allowable particulate and tar levels in the prod-
uct gases. Thus, the use of coal-derived product
gases to replace nature gas and oil on a large
scale will require adequate control technology
to remove tars and particulates from the prod-
uct gases to levels compatible with the various
possible end uses. The overall objective of this
study was, therefore, to assess the applicability
of alternate control technologies both commer-
cially available and under development for the
removal of particulates and tars from coal-con-
verter product gases.
The first step in carrying out these control
technology evaluations involved the identifica-
tion and collection of pertinent sources of in-
formation. Computerized literature searches
covering the Chemical Abstracts, Engineering
Index, Pollution Abstracts, U.S. and foreign
patents, government publications, and numer-
ous journals were made to identify sources of
information. These computerized searches were
complemented by thorough library and patent
searches. In addition, other U.8. Environmental
Protection Agency (EPA) contractors, process
developers, and equipment vendors were con-
tacted for relevant data. When identified
sources of information had been reviewed, ap-
propriate data were employed to carry out the
control technology evaluations, as discussed in
the following sections.
This study was performed under EPA Con-
tract Number 68-02-2601 for the Fuels Process
Branch of the Environmental Assessment and
Control Division of the Industrial Environmen-
tal Research Laboratory (IERL) at Research
Triangle Park. The methodology and results
summarized herein are described in detail in
Reference 1.
CHARACTERISTICS OF PARTICULATE
AND TAR EMISSIONS
As an initial step in the evaluation of
technologies for the control of particulates and
tars in gaseous streams originating from coal
gasifiers, emissions and process data were ob-
tained for a wide variety of gasifiers. The avail-
ability of pertinent data was generally found to
be limited. The emissions data were organized
and summarized according to generic gasifier
type, with fixed-, fluid-, and entrained-bed
gasifiers being considered. Because of the
uncertainties in the emissions data for the dif-
ferent types of gasifiers, these results are pre-
sented in terms of best-case, worst-case, and
average-case (or typical) analyses. The worst-
case condition represents the estimated upper
limit of particulate load, with a relatively high
percentage of small particles, which are difficult
to remove. The best-case condition represents
the estimated lower limit of particulate load
with a relatively low percentage of small parti-
cles. All available data on the characteristics of
gasifier emissions were considered in esti-
mating these upper and lower bounds.
Typical operating parameters and raw prod-
uct gas stream characteristics are presented in
Table 1 for several different fixed-bed, fluid-
bed, and entrained-bed gasifiers. Several of the
fixed-bed gasifiers are commercially available,
whereas the Winkler and Hoppers Totzek are
the only commercial fluid-bed and entrained-bed
gasifiers, respectively. It can be seen from
Table 1 that the fixed-bed gasifiers produce
tars, while the entrained-bed gasifiers do not.
Most of the fluid-bed gasifiers produce tara,
while the entrained-bed gasifiers do not. Most of
the fluid-bed gasifiers also do not produce tars.
The particulate and tar loading data are sum-
marized in Table 2. The best-, worst-, and aver-
age-case data for the particulate and tar load-
ings from each generic type of gasifier were
estimated from the detailed data for the indivi-
dual gasifier types in Table 1. It can be seen that
fixed-bed gasifiers produce the smallest particu-
late loadings, while the entrained-bed gasifiers
produce the highest loadings.
Particle size distribution data are presented
in Figures 1, 2, and 3, and are summarized in
Table 2. The particulate collection efficiency of
most control devices is especially sensitive to
particle size. Such data were generally found to
be scarce and incomplete. More complete data
over a broad, specified range of gasifier oper-
ating conditions are needed. In the case of fluid-
and entrained-bed gasifiers, particle size data
were not available below approximately 35 and
20 jim, respectively. Extrapolation of the ex-
isting data for large size particles down to the
small size particle range was, therefore, re-
quired for these two types of gasifiers. Large
particles are removed more easily than parti-
cles below approximately 5 jun; therefore,
future R&D programs should concentrate on
the collection of particle size data down to the
submicron size range.
480
-------
TABLE 1. OPERATING AND RAW PRODUCT GAS STREAM CHARACTERISTICS
Gasi fler
Part Iculjto
Coal Temperature Pressure Loading
Type °C «"a g/nmi
"articulate' Tar loading Tar
Composition (|/run* Compos i t ion
MxeJ Bed
We 11 man anthracite 1(30-920 0.10
Galusha(2,3) bituminous
coke
Lurgl "variety 370-590 2.07-3.21 0.5-6.0
(2,3,
tar oi 1
10-20.
C-82.1
H-7.6
GFERC(2,1|) lignite 85-370 0.60-2.%
I ignite char
bit. cha r
Fluid Bed
Winkler(2,3,li) several 590-730 0.10
coal types
Synthane
al1 types 760
C02 Acceptor lignite 815
(3.*,5) sub-bit.
Hygas(3,'<) all coals 1100
6.90
ash-70'.
C-80?
ash-20''
ash-3fl/.
C-55''.
ash-iiO?
tar-IO
tar oil-25
2.'t-l7
HOIK-
CoGas (3) all types 8?0 O.VI-0."tl
Hydrane (3) all types 5
char-96-*97 None
volat i Ins -
T. xaco(2,3) lignite 200-260 2.10-8.27 "
Combustion all types 870 0.10
CnTineerinq(3)
B ' W (3) jll types 910 0.10-2.10
CoalRx (2) all types 925-950 0.10
Foster non-cakIng unprr stage 2,b\
Wheeler (2,J) <)80-U50
lower staqe
IJ70-l5"lO
Nonr
481
-------
TABLE 2. SUMMARIZED PARTICULATE AND TAR LOADINGS AND
PARTICLE SIZE DISTRIBUTIONS
6
to
Fixed Bed
Best Case
Worst Case
Average
Fluid Bed
Best Case
Worst Case
Average
Entrained Bed
Best Case
Worst Case
Average
Part icul ate
Loading
(g/m3)
0.5
6.0
3.0
1.2
120.0
26.0
30.0
230.0
110.0
Tar
Loading
(g/ra3)
10.0
50.0
18.0
None
None
None
None
None
None
Percent Particles (by weight) Less Than
Specified Diameter (In
1 5
<0.1 0.1
<0.1 4.0
<0.1 2.0
0.1 1.0
0.5 5.0
0.3 3.0
<0.1 0.5
<0.1 0.5
<0 . 1 0.5
10
1
30
15
2
12
7
2
k
3
Micrometers)
50
23
67
45
13
52
33
12
66
39
100
50
76
63
22
78
50
2k
90
57
-------
oo
co
u
o
6
5JC
too
50
10
1
4>^
s
f
j™ '
/O
/o
^
^£^
^
O j
X
-X]
^
x-^'xx^'
x^*»
X
^
'
X
jf
^
^<
^ —
i
' I
/A
/
0
X
>X^
'O
_/
r
/
I/
/ ,.
A'\
,/
_/• '
j/r
r
1
1
r
i
4
^^ !
- ^ ,
^«
>
/
/
/
A
\A
$
/
t
f
*/
y
/
^
i
A/
7
/
5**'
J
-^
o MERC Data ( 1 1 inois No. 6,
B i turn! nous) (7)
A MERC Data (Rosebud, Sub-
SituminousJ (7)
O Lurgi Data
Average of Worst and
Ssst r.rses
c.: ' 13 3C 3C iff 50 SO 70 to SO S5 S3 S3. 9 SS.i
Cumu ative Probability
Figure 1. Particle size distribution for fixed-bed gasifiers.
-------
set
10
i
!
/
, /
/ /
/ /
i /
/i / /
/
/ /
///
f /
/
f
/\
/
/
r
/
f
i
/
/
)
4
J
*
f i
\
'
j
1 1
)
/
/
/
/
: /
y
A
^
/,^
' I /
1
r
/
f
/
/
/
r
/
o
>>c<
; ;
!
'•
i
i
1
/
y
0 f
/
/
^
o
^
/
s
'
'*"*'!
..,
•*'"
f
s^
r
**^"
.,
x-^
s1
o CO
B
1
\
^
*^
2~Acc«
est-Cc
gni f li
r.xtra{.
range
Ave rac
Best C
— •
ptor Data
se based o
id data (5
>olated bey
of data
e of Worst
ases
t
(8)
n
ond
and
j.j ? 10 30 30 iO K (0 73 »0 SO K 33 MJ »S»
Cumulative Probab i i ty
Figure 2. Particle size distribution for fluid-bed gasifiers.
-------
;
-------
Particle size distribution measurements are
usually based on either aerodynamic or optical
properties of the particles. Measurements in the
same gas stream by these two different tech-
niques often yield inconsistent results. Particle
sizes are especially difficult to measure at high-
temperature and high-pressure (HTHP) condi-
tions. The collection of reliable particle size
distribution data for coal-gasifier product gases
will require the development of improved meth-
ods and instrumentation suitable for HTHP con-
ditions.
Additional data are also needed to accurately
estimate particulate and tar loadings from the
various types of gasifiers, particle and tar
compositions, and other pertinent properties
such as particle resistivity. It should be noted
that complete data sets were not available for
any of the gasifier types. For example, the parti-
cle size distribution might be available for a spe-
cific type of gasifier at a given or unspecified set
of conditions, whereas particulate loadings and
compositions might be available for another
type of gasifier within the same generic class,
but at a different set of conditions. There is,
then, a need for R&D programs to provide com-
plete data for all of the above parameters at the
same specified gasifier operating conditions.
ALTERNATE CONTROL TECHNOLOGIES
Alternate control technologies (both commer-
cially available and under development) for re-
moving the particulates and tars from the raw
product gases from coal converters were iden-
tified and evaluated. The performance charac-
teristics of the commercial types of control
devices are generally well known and well docu-
mented. Sufficient data were available for the
following six generic control technologies to
permit performance of detailed applicability
assessments: conventional cyclones, rotary flow
cyclones, venturi (wet) scrubbers, fabric filters,
electrostatic precipitators (ESPs), and granular
bed filters (GBFs). Typical collection efficiencies
for each of these control technologies are pre-
sented as a function of particle size in Figure 4.
The fabric filter and ESP are most efficient for
small particle sizes; the fabric filter, venturi
scrubber, and rotary flow cyclone are most effi-
cient for the relatively large particles. Detailed
descriptions of most of these control devices are
readily available in the literature. Brief sum-
maries of important design and operating fea-
tures are presented below.
Cyclones utilize the centrifugal force created
by a spinning gas stream to separate particu-
lates from the carrier gas. The advantages of
the conventional cyclone are that it is a simple
device—there are no moving parts—and it is a
proven technology. However, cyclones suffer
from the disadvantages of having low removal
efficiencies for particulate sizes less than 5 pm.
Because of their relatively low capital and oper-
ating costs, cyclones are commonly used as pre-
cleaners to remove most of the large particles in
a gas stream upstream of a more expensive con-
trol device (e.g., venturi scrubber or electro-
static precipitator) required to remove the
smaller size particles. Particulate collection effi-
ciency increases with an increase in particulate
diameter, particulate density, inlet velocity,
cyclone body length, ratio of cyclone body diam-
eter to outlet diameter, and the smoothness of
the inner wall. Efficiency decreases as the gas
viscosity, gas density, body diameter, and gas
outlet diameter increase. Because the gas vis-
cosity is proportional to temperature, an in-
crease in temperature results in a decrease in
the collection efficiency.
Rotary flow cyclones are designed to aug-
ment the normal tangential swirl of the inlet gaa
by the addition of a secondary airflow. By doing
so, the possibility of short-circuiting particu-
lates from inlet to outlet is greatly reduced. In
the Aerodyne rotary flow cyclone,14 particulate-
laden gas enters the collection chamber and
passes a stationary vane, which imparts a
rotary motion to the flow. Particulate matter is
thrown toward the outer wall by centrifugal
force and then swept downward to the collec-
tion hopper by the secondary flow. The vendor
data for the Aerodyne Series "S" rotary flow
cyclone are presented in Figure 4. Westing-
house has also tested an Aerodyne Tornado
Cyclone. The grade efficiency data obtained in
these tests show a discrepancy with respect to
the claimed performance by the manufacturer.
This may result from the difficulty of holding
design control specifications when small unit is
tested. Thus, the fractional collection efficiency
data presented in Figure 4 need to be verified.
Although wet scrubbers are available in a
wide variety of designs, all operate on a com-
mon principle of contacting a pollutant-laden
gas with a liquid (usually water) that captures
486
-------
99.99
19-9 ~
i j i I I 1 I 1 I
I I i I I I I'I
0.02
0.) 1
Particle Diameter, Micrometers
Figure 4. Typical collection efficiencies of control devices.
487
-------
the pollutants. Wet scrubbers can be used to
remove both particulates and tars. The objec-
tives of good scrubber design are to provide
good liquid-gas contact, minimize energy con-
sumption and equipment size, and minimize
water requirements. All wet scrubbers produce
a liquid slurry for disposal or further treatment.
Most modern applications attempt to concen-
trate the solids to simplify their ultimate
disposal, and to recirculate as much of the
scrubbing liquid as possible. The collection effi-
ciency of wet scrubbers is strongly dependent
on particle size. In order to achieve high collec-
tion efficiencies with small particles, a high-
energy input is required. For particles above ap-
proximately 10 pm, simple wet scrubber designs
are usually adequate, with a pressure drop of
0.25 kPa being typical. Fine particulates with
diameters of 1 /xm or less require more complex
scrubbers with pressure drops usually well
above 1.25 kPa. In exceptional circumstances,
pressure drops up to 25 kPa have been em-
ployed. Wet scrubbers have been found to be
very effective in removing tars from raw prod-
uct gases. Commercially available gasification
systems generally have employed various types
of wet scrubbers to quench and cool the gases
and knock out the tars, along with a portion of
the particulates.
The venturi scrubber employs a venturi-
shaped constriction and high throat velocities to
atomize the scrubbing liquid. As with wet collec-
tors in general, the collection efficiency in-
creases with higher pressure drops. Different
pressure drops are achieved by designing for
varied gas velocities in the throat. Some venturi
scrubbers are manufactured with adjustable
throats, allowing a range of pressure drops for a
given air volume. The collection efficiency of the
venturi scrubber can generally be considered
highest of the wet collectors.
Electrostatic precipitators (ESPs) for clean-
ing particulates from gases, which have been
used by industry for over 70 years, have also
been found to be an efficient means of detarring
the gases. ESPs operate by using a high-volt-
age, direct current to create gas ions that im-
part an electrical charge to particulates by bom-
bardment. The charged particles are collected
by exposing them to an electric field, which
causes them to migrate and deposit on elec-
trodes of opposite polarity. The electrode clean-
ing system is dependent upon the type of pre-
cipitator. The conventional dry-type precipita-
tor collects particulates on a dry electrode and
removes them periodically by mechanical shak-
ing or rapping. The new wet-type precipitator
collects and removes particulates with a thin,
continuous flowing film of water. The operating
temperatures are generally less than 65° C.
The resistivity of particulates is a critical fac-
tor in the design and operation of a dry precipi-
tator. Particulates with low resistivity (below
50 Q«m) are difficult to collect efficiently be-
cause they tend to loosely adhere to the collec-
tor and are, therefore, easily reentrained in the
gas stream. On the other hand, if the participate
resistivity is too high (above 0.2 G fl>m), the
voltage drop across the deposited particulate
layer becomes so large that the discharge elec-
trode electron emission rate drops, which leads
to a decline in the overall collection perfor-
mance. Hot-side precipitators, which operate at
temperatures up to 540° C, were developed for
certain applications involving high-resistivity
particulates. Research Cottrell, under an EPA
contract, has demonstrated the ability of an
ESP to generate stable corona at temperatures
up to 1,100° C and pressures up to 51 MPa."
While their limited data are encouraging, more
data are required at high temperatures.
One of the oldest and the most widely used
techniques for removing particulates from a gas
stream is the use of fabric filters. The baghouse
design is very commonly used, and is highly ef-
fective even for small particulates. However,
commercially available baghouses are not suit-
able for use at high temperatures. A number of
high-temperature-resistant ceramic fabrics
have become commercially available. Because of
the lack of a suitable high-temperature, inorgan-
ic fiber lubricant needed for the fiber-to-fiber
abrasions, many of these developed ceramic
fabrics are presently unsuitable for filtration
purposes. Still, ceramic fabric filters offer a
potentially promising solution of the problem of
controlling particulates in the high-tempera-
ture, high-pressure environment. The advan-
tages of baghouse filters include high collection
efficiencies, even for submicron particles, rela-
tively low energy use and pressure drop (typi-
cally less than 7.5 kPa), and collection of par-
ticles in dry form, which simplifies ultimate
waste disposal. Disadvantages include large
form, which simplifies ultimate waste disposal.
Disadvantages include large space require-
488
-------
ments, high initial costs, and proven tempera-
tures limited to about 290° C.
A granular bed filter (GBF) employs a station-
ary or moving bed of granules— sand, gravel,
coke, or sintered material—as the filter medi-
um. In order to maintain a steady operating per-
formance, a granular bed filter needs to remove
the collected particulates from the collecting
surface. Several different designs are reported
in the technical literature. In general, they may
be classified as continuously moving, intermit-
tently moving, or fixed-bed filters with respect
to the cleaning methods. GBFs are a promising
technique for high-temperature and high-pres-
sure operation. They have the advantages of be-
ing able to use either inert or sulfur-absorbent
material, and of accommodating high face veloc-
ities while incurring a moderate pressure drop.
The collection mechanism is similar to that of
fiber filters, with impaction predominating and
particulates being collected in the interstices of
the filter. After the initial collection at the filter
surface produces a filter cake, further collection
is accomplished essentially by cake sieving.
Granular bed filters have received increased at-
tention recently, and a number of research pro-
jects are underway to further develop these
systems. The GBF developed by Combustion
Power Company is the most advanced of this
generic class of control devices. This GBF
employs granular filter media between two ver-
tical, louvered screens. To avoid particulate
saturation, the medium is continuously recircu-
lated and cleaned. Commercial devices, re-
stricted to temperatures below 430° C and to
near atmospheric pressures have been available
for a few years.13
In contrast to the six generic classes of con-
trol devices discussed above, several other con-
trol devices are still in the developmental stage;
data are insufficient to permit meaningful eval-
uations of their applicability to coal converters.
Several of these newer, relatively advanced
control devices are discussed below. Additional
collection efficiency data and/or large-scale
testing to determine operational reliability are
required to evaluate these control devices.
Several advanced types of wet scrubbers are
under development to improve the collection of
fine, submicron particles. These newer types of
scrubbers include foam, steam-assisted, and
electrically augmented devices. At present,
their principal disadvantage appears to be high
initial cost compared to other types of wet
scrubbers. In addition, operating and perfor-
mance experience with these devices is limited.
Porous ceramic filters appear to be especially
promising for highly efficient collection of parti-
cles down to the submicron size range at high
temperatures.18 Such devices can take the form
of porous thick-walled filters or thin-walled
(0.2-mm) monolithic honeycomb structures.
While preliminary data at high temperatures
are encouraging, additional testing with larger
scale devices is required for confirmation.
Several novel devices are in the early stages
of development, with only limited preliminary
data available. Such devices include the A.P.T.
dry scrubber,19 molten salt scrubber,20 elec-
trofluidized bed,21 and the Apitron charged
filter.22 The latter appears to have especially
high collection efficiencies down to submicron
size particles, but operation is restricted to the
same temperature range as a conventional bag-
house filter.
APPLICABILITY OF CONTROL DEVICES
Applicability assessments were made for
various combinations of particulate control
devices and gasifier end use pairs. These assess-
ments were made for the three major generic
classes of coal gasifiers discussed previously
(fixed-, fluid-, and entrained-bed).
Each potential end use for the product gases
has different environmental regulations and
process requirements governing the allowable
particulate levels in the product gases. For the
purposes of this study, two particular end uses
were selected for consideration. These end uses
were selected to cover a wide range of particu-
late removal requirements for the control de-
vices under consideration. The use of product
gases as a boiler fuel was selected to represent
those end uses with low to moderate particulate
cleanup requirements. On the other hand, the
use of product gases as a fuel for gas turbines
was selected to represent end uses with rela-
tively restrictive cleanup requirements. The
New Source Performance Standard established
by EPA to limit particulate emissions from coal-
fired steam generators (0.10 lb/106 Btu heat in-
put) was assumed to apply to boilers firing coal-
derived fuel gases. This is equivalent to 0.24
489
-------
g/m3 of particulates for low-Btu fuel gas with an
average heating value of 0.66 MJ/m8 (150
Btu/scf).
Coal-derived product gases can be used as a
fuel for gas turbines employed in combined-
cycle power stations. The tolerance of a gas tur-
bine to particulates is not known with a high
degree of certainty. Stringent specifications for
fuels to be burned in gas turbines have been es-
tablished by various turbine manufacturers.
Results obtained by the U.S. Department of
Energy's High Temperature Turbine Technol-
ogy Program5 suggest a maximum allowable
particulate concentration of 0.0046 g/m8 of ex-
pansion gas, or 0.041 g/m8 of unburned fuel gas,
with no particulates larger than 6 /on in diame-
ter. These results were used as the basis for the
particulate control requirements for the gas tur-
bine end use. It should be noted that there are
presently no environmental regulations govern-
ing the emission of particulates from gas tur-
bines.
Detailed applicability assessments were
made for the six generic classes of control
devices previously discussed. These assess-
ments are based primarily on the capability of a
control device to achieve the required degree of
particulate removal for a specified gasifier end
use pair. In some cases where obvious operating
difficulties would be expected, such potential
problems are also considered in evaluating the
applicability of a control device.
As discussed previously, the removal efficien-
cy of any particulate control technology is a
strong function of particulate size. Thus, a
meaningful applicability assessment of control
technologies requires knowledge of the particu-
late size distribution in the gases to be treated,
along with collection efficiencies of the control
technologies as a function of particle diameter.
The overall collection efficiency of each control
device can then be obtained from the grade effi-
ciency data of the control device and the parti-
cle size distribution data by means of graphical
integration.28 This graphical technique can be il-
lustrated by the following example for deter-
mining the overall collection efficiency of a con-
ventional cyclone operating on an effluent with
the "best-case" particle size distribution of a
fixed-bed gasifier, as shown in Figure 1. The
particle size distribution data in Figure 1 and
the fractional collection efficiency data in
Figure 4 are presented in Table 3. Figure 5 was
then obtained by plotting these tabulated data.
The overall collection efficiency for particulates
in the size range from 0 to 6 /an was determined
by locating the point at which the areas above
and below the curve are equal. An overall collec-
tion efficiency of 86 percent was thereby ob-
tained. It should be noted that the accuracy of
this graphical technique is limited by uncertain-
ties in the particle size and grade efficiency
data, as discussed previously. Errors introduced
by the graphical procedure itself are negligible.
The graphical technique discussed above was
employed for particulates less than 6 jim in
diameter for all control devices. For the particu-
lates greater than 6 /on, a representative value
for removal efficiency could be selected for each
generic control device, with the exception of the
conventional cyclone. This is because of the fact
that the collection efficiency of a conventional
cyclone usually reaches a maximum at a much
larger particle size than 6 /on; for most other
control devices the removal efficiencies are
nearly constant for particles greater than 6 pm.
Thus, the same general graphical method was
used to calculate the overall collection efficien-
cies of a cyclone for each gasifier effluent over
the particulate size ranges below and above 6
/tin.
A compilation of the overall collection effi-
ciencies for each combination of generic control
device and gasifier effluent is presented in
Table 4. With the overall removal efficiency of
each generic control device thus determined,
the applicability assessments were then carried
out on the basis of the estimated particulate
loadings from each gasifier, as presented in
Table 2. The amount of particulates not re-
moved was then calculated. The results for each
generic control device under consideration are
presented in Table 5. The applicability can then
be determined by comparing the amount of par-
ticulates remaining in the product gases to the
maximum allowable concentration of particu-
lates for each end use.
The results of the applicability assessments
are summarized in Table 6. Conclusions drawn
from these results are discussed below sepa-
rately for End Use 1 (combined-cycle fuel gas)
and End Use 2 (conventional boiler fuel gas). As
for End Use 1, the very restrictive requirement
of removing all particles larger than 6 /on has
limited the potential control devices to fabric
filters, a high-efficiency venturi scrubber, and
490
-------
TABLE 3. COLLECTION EFFICIENCY OF HIGH-EFFICIENCY CYCLONE
FOR PARTICULATES FROM FIXED-BED GASIFIER
Particulate Size(Dp)
micrometers
Amount^ Dp,*
% by weight
Cyclone
Efficiency,**
16
-15
}k
13
12
11
10
9
8
7
6
5
k
3
2
1
A
3
2.5
2.0
1.5
1.3
1
0.7
O.k
0.3
0.25
0.11
0.07
0.02
0.01
0.001
>99
99
99
98.5
98
97
96
95
9*
92
90
87
83
77
68
53
* Cumulative size distribution data for fixed-bed gasifier (see Figure 1)
** Collection efficiency of conventional cyclone for particles with diameter
of Dp (see Figure 4)
the Aerodyne rotary flow cyclone. Among these
three control devices, the fabric filter was found
to be the only device capable of achieving the re-
quired product gas purity (0.041 g/m8) for End
Use 1 for all gasifier effluents. However, a
fabric filter should not be employed for gases
containing high levels of liquid or "sticky" par-
ticles. Thus, fixed-bed gasifiers, in particular,
may not be compatible with fabric filters, be-
cause of the quenching operation commonly used
to condense and remove tars and oils. The high-
efficiency venturi scrubber is applicable for End
Use 1 for all gasifier effluents, except for the
worst-case fluid-bed gasifier. However, with a
high-efficiency cyclone upstream as a scalping
device, the venturi scrubber is capable of
achieving this requirement for the worst-case
fluid bed, based on the assumption that the par-
ticulate size distribution for particulates less
than 6 /on remains unchanged after passing
through the cyclone. The Aerodyne rotary flow
cyclone is found to be inapplicable for the aver-
age and worst-case fluid-bed gasifier. It should
be noted that the results presented herein for
the rotary cyclone should be considered ten-
tative until the vendor-supplied data employed
in these assessments are confirmed.
Because the particulate removal requirement
491
-------
c
o
4-»
u
o>
O
o
50
1*0
30
Remova 1
micrometers
Dp=5
Dp*6
\
0 0.05 0.1 0.2
Amount less than stated size, % by weight
Figure 5. Graphical procedure for estimating overall collection efficiency
for paniculate* up to 6 jum in diameter.
492
-------
TABLE 4. OVERALL PARTICULATE REMOVAL EFFICIENCIES OF GENERIC CONTROL
TECHNOLOGIES FOR TYPICAL GASIFIER OUTPUTS
T>
6
<6
>6
<6
>6
*6
>6
<6
>6
<6
>6
<6
>5
^6
>6
^6
?6
? Particulate Removal Vs. Particulate Size
Conventional
Cyclone
86
99
86
93
86
97
82
98.8
79
98.7
76
98.5
SA
98.6
83
98.8
82
98.9
Rotary
Cyclone
98.5
100
9B.8
100
99
100
90
100
91.5
100
93
100
96.5
100
97.3
100
98
100
Venturi
Scrubber
99.93
100
99.9^
100
99.95
100
97
100
97.8
100
98.5
100
99.7
100
99.83
100
99.95
100
Fabric
Filter
99.99
100
99.99
100
99.99
100
99.2
100
99. *»
100
99-6
100
99. 9A
100
99.97
100
99.99
100
E.S.P.
N.A.
N.A.
N.A.
N.A.
N.A.
N.A.
93
99.8
98. k
99.8
98.8
99.3
99
99.8
99.2
99.8
99. ^
99.8
Granular
Bed Filter
9^.6
95
9^.7
95
94.7
95
94. k
95
9*».5
95
9^.5
95
9^.6
95
94.7
95
94.7
95
£
co
-------
TABLE 5. EFFECTIVENESS OF PARTICULATE COLLECTION BY ALTERNATE
CONTROL DEVICES
Gasifier
TY
•c
0)
CD
0)
X
u.
•D
OB
2J
3
iZ
«
DO
•D
0)
C
W
C
UJ
>e
Best
Case
Ave rage
Case
Worst
Case
Best
Case
Average
Case
Worst
Case
Best
Case
Ave rage
Case
Worst
Case
Percent
Particulate
Distribution
Size
(// m\ )
<6 0.3
>6 99.7
<6 5.4
>6 94.6
<6 10.5
>6 89.5
<6 1
>6 99
<6 3
>6 97
<6 5
>6 95
<6 0.5
>6 99.5
<6 0.7
>6 99.3
<6 0.8
>6 99.2
Particulates Remaining in Product Gases
Downstream of Control Device, g/nv*
Conventional Rotary Flow Venturi Fabric
Cyclone Cyclone Scrubber ESP Filter G8F
0.003 0.00002 0 ... 0 0.00003
0.0046 0 0—0 0.0229
0.0023 0.0023 0.0001 --_ 0.00002 0.0085
0.0557 0 0—0 0.142
0.0869 0.0069 0.0002 — 0.00006 0.033
0.1603 0 o—O 0.269
0.0023 0.0012 0.005 0.0002 0.00009 0.00067
0.0137 0 0 0.0002 0 0.025
0.1649 0.0664 0.017 0.013 0.0042 0.0419
0.3275 0 0 0.050 0 1.218
1.44 0.421 0.089 0.071 0.0024 0.31S
1.65 0 0 0.227 0 5-422
0.023 0.0051 0.0005 0.002 0.00009 0.0077
0.419 0 0 0.060 0 1.372
0.131 0.021 0.0014 0.006 0.00023 0.0397
1.312 0 0 0.210 0 5.281
0.339 0.036 0.0009 0.011 0.00019 0.095
2.512 0 0 0.456 0 10.95
To be compared to maximum allowable parttculate loads of:
0.041 g/m3 for combined cycles and 0.24 g/m3 for boiler fuel.
-------
TABLE 6. SUMMARY OF APPLICABILITY ASSESSMENTS
Appl icabi
End Use/Control Device
COMBINED-CYCLE
conventional cyclone
rotary cyclone
venturl scrubber
fabric filter
E.S.P.
granular bed f i Iter
rotary cyclone*
venturi scrubber*
fabric filter*
E.S.P.*
granular bed f i 1 ter*
BOILER FUEL
conventional cyclone
rotary cyclone
venturi scrubber
fabric filter
E.S.P.#
granular bed f i 1 ter
rotary cyclone*
venturi scrubber*
fabric fi Iter*
E.S.P.*
granular bed filter*
lity
of
Fixed Bed
B
X
X
P
X
X
P
X
X
X
P
X
X
X
P
X
W
X
X
P
X
X
P
X
X
P
X
X
P
A
X
X
P
X
X
P
X
X
X
P
X
X
X
P
X
Control
Flui
B
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Devices
d
W
X
X
X
X
X
X
X
X
X
Bed
A
X
X
X
X
X
X
X
X
X
X
X
X
X
X
for Gas
ifier
Entrained
B
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
W
X
X
X
X
X
X
X
X
X
X
X
X
X
Types**
Bed
A
X
X
X
X
X
X
X
X
X
X
X
X
X
X
* A conventional cyclone is assumed to be employed as a scalping device
upstream of the indicated primary control device.
** B - Best Case
W - Worst Case
A - Average Case
P - Designates probable inapplicability due to operating problems, although
particulate removal is adequate.
X - Designates control device is applicable.
j ESP is not applicable to a fixed bed gasifier due to high carbon content and
low resistivity of particles.
495
-------
of End Use 2 is not as restrictive as End Use 1,
the number of control devices applicable to End
Use 2 is increased considerably as compared to
End Use 1. Both the fabric filter and the venturi
scrubber are capable of achieving the require-
ments of End Use 2 for all gasifier effluents. The
Aerodyne rotary flow cyclone was found to be
applicable to all the gasifier effluents except for
the worst-case fluid-bed gasifier. However, with
a conventional cyclone upstream as a scalping
device, it would be applicable to this worst case
as well. A conventional high-efficiency cyclone
by itself would be applicable to the best and
average cases of the fixed-bed gasifier, and the
best case of the fluid-bed gasifier. The CPC
granular bed filter is found to have the same ap-
plicability as the high-efficiency cyclone men-
tioned above. Two cyclones in series are capable
of achieving the same efficiency as an Aerodyne
rotary flow cyclone. A cyclone followed by a
CPC granular bed filter would be applicable to
two more cases than the CPC filter by itself-
the average case of the fluid-bed gasifier and
the best case of the entrained-bed gasifier. It
was found that a dry-type electrostatic precipi-
tator is not applicable to fixed-bed gasifier ef-
fluents because the particles in these effluents
have very high carbon contents (55 to 80 per-
cent) which result in low resistivity of the par-
ticles and inefficient collection. The electrosta-
tic precipitator was found to be applicable to the
best and average cases of the fluid-and
entrained-bed gasifiers for End Use 2. With a
cyclone upstream as a scalping device, the elec-
trostatic precipitator would also be able to
achieve the required removal efficiency for the
worst cases of the fluid- and entrained-bed
gasifiers.
FATE OF POLLUTANTS
In the previous sections, technologies for con-
trolling the particulate and tar levels of the con-
verter product gases have been discussed and
evaluated. Each control device, in turn, gener-
ates solid, liquid, and/or gaseous wastes that
also must be disposed of in an environmentally
acceptable manner. By identifying those
streams in which certain pollutants tend to con-
centrate, proper disposal and control technolo-
gies can be selected to minimize environmental
degradation.
Data on the fate of the particulates and tars
emitted in the product gases, in terms of their
ultimate presence and concentrations in solid,
liquid, and gaseous discharge streams, are pre-
liminary and limited for all gasifier types. The
distribution of these particulates and tars in the
various discharge streams is dictated both by
the removal technology and the physical and
chemical characteristics of the contaminants.
The conclusions summarized below should,
therefore, be considered tentative until confirm-
ed by additional data. The data on which these
conclusions are based are presented in detail in
Reference 1.
In the case of fixed-bed gasifiers, the quench
liquor employed to condense and remove the
tars contains high concentrations of phenolic
compounds. These compounds, together with
ammonia and dissolved acid gases, must be re-
moved from the quench liquor. Mercury tends
to concentrate in the tar, while most other
volatile elements tend to become concentrated
on the particulates. Selenium concentrations in
the quench liquor are very high.
In the case of fluid-bed gasifiers, most of the
available data on the fates of the various con-
taminants were obtained with the Synthane
unit, which also produces tars. Since most other
fluid-bed gasifiers do not produce tars, these
data may not be representative of this generic
type. The available data indicate that many of
the trace elements tend to concentrate in the
particulates and char. Some of the more volatile
elements such as As, Pb, and Hg are also found
in potentially harmful concentrations in the tar.
In the case of entrained-bed gasifiers, organ-
ics tend to concentrate on the particulate mat-
ter rather than the scrubber water. Volatile
elements such as Hg, Se, and As are not absorb-
ed in the scrubber water. Tars are not produced
by entrained-bed gasifiers, so they do not pre-
sent a disposal problem.
Of the six generic control technologies pre-
viously assessed, the venturi scrubber is the on-
ly wet process. The other five processes
generally produce a dry, granular, or powdery
solid waste. In the case of a venturi or other wet
scrubber, the collected fly ash will be wet, com-
plicating disposal of the ash and necessitating
wastewater treatment. Liquid waste streams
from scrubbing or quenching operations must
be treated prior to final disposal or discharge to
surface waters or groundwaters. Present and
proposed regulations for liquid discharges gen-
496
-------
erally require a high degree of water recycle
and reuse within the plant, thereby minimizing
the amounts of liquid to be released from the
plant. The collected ash, whether wet or dry,
must be disposed of in a landfill or in any other
environmentally acceptable manner. Undesir-
able elements can sometimes be leached from
the collected particulate matter. Even if a dry
collection system is used, the solid wastes will
ultimately be exposed to leaching by ground-
water if they are disposed of as landfill or
returned to the mine. Use of liners and entrap-
ment of runoff and drainage water will minimize
the likelihood of ecological degradation.
Additional sampling is required for all gasi-
fier types to identify and determine the concen-
trations of contaminants in quench water, solid
wastes, tars, and scrubber water under better
defined conditions. Laboratory analyses should
include trace metals and identification of the
chemical forms in which they appear, as well as
other inorganic and organic compounds. Studies
to determine the teachability of trace elements
from captured particulates and tars into quench
and scrubber water and into groundwater after
ultimate disposal would be very helpful.
REFERENCES
1. Chen, G., C. Koralek, and L. Breitstein.
Control Technologies for Particulate and
Tar Emissions from Coal Conveners
(draft). Dynalectron Corporation. (Prepared
for U.S. Environmental Protection
Agency.) January 1979.
2. Cavenaugh, E. C., W. E. Corbett, and G. C.
Page. Environmental Assessment Data
Base for Low/Medium-Btu Gasification
Technology, Volume II, Appendices A-F.
EPA-600/7-77-1256. November 1977.
3. Dravo Corporation. Handbook of Oasifiers
and Gas Treatment Systems. FE-1772-11.
February 1976.
4. Becker, D. F., and B. N. Murthy. Feasibility
of Reducing Fuel Gas Clean-up Needs. FE
1236-15. June 1976.
5. Meyer, J. P., and M. S. Edwards. A Survey
of Processes for High Temperature—High
Pressure Qas Purification. ORNL/TM-6178.
November 1978.
6. Sinor, J. E. Evaluation of Background
Data Relating to New Source Performance
Standards for Lurgi Gasification.
EPA-600/7-77-057. June 1977.
7. Moore, A. S., Jr. Cleaning Producer Gas
from MERC Gasifier. U.S. Energy Re-
search and Development Administration.
May 1977.
8. Commercial Plant Conceptual Design and
Cost Estimate—CO2 Acceptor Process
Gasification Pilot Plant Conoco Coal
Development Corp. and Steams-Roger En-
gineering Co. Vol. 10, FE/1734-43. August
1976-December 1977.
9. Parker, R., and S. Calvert. High-
Temperature and High-Pressure Particu-
late Control Requirements. EPA-600/7-77-
071. July 1977.
10. Whiteacre, R. W. Personal Communication
with Koppers-Totzek. August 1978.
11. Shannon, L. J., P. G. Gorman, and M.
Reichel. Particulate Pollutant System
Study, Vol 11: Fine Particulate Emissions.
Midwest Research Institute. (Prepared for
U.S. Environmental Protection Agency.)
NTIS PB-203 521.1971.
12. Shannon, L. J. Control Technology for Fine
Particulate Emissions. Midwest Research
Institute. (Prepared for U.S. Environmen-
tal Protection Agency.) NTIS PB-236 646.
1974.
13. Wade, G. L. Performance and Modeling of
Moving Granular Bed Filters. In: Pro-
ceedings of EPA/DOE Symposium on High
Temperature High Pressure Particulate
Control Acurex Corp. EPA-800/9-78-004.
1977. p. 133-192.
14. Gordon, M. Aerodyne Series "SV" Dust
Collector, Aerodyne Development Corp.
Cleveland, Ohio. Bulletin Number 1275-SV.
1978.
15. Calvert, S., and R. Parker. Effect of
Temperature and Pressure on Particle Col-
lection Mechanisms: Theoretical Review.
A.P.T., Inc. Prepared for U.S. Environmen-
tal Protection Agency.) NTIS PB-264 203
1977.
16. Klett, M. G., W. Szwab, and J. P. Clark. Par-
ticulate Control for Pressurized Fluidized-
Bed Combustion. Gilbert/Commonwealth,
R&D Division. (Prepared for U.S. Energy
Research and Development Administra-
tion.) FE-2220-16. January 1977.
17. Bush, J., P. Feldman, and M. Robinson.
497
-------
Development of a High-Temperature/
High-Pressure Electrostatic Precipitator.
Research Cottrell, Inc. (Prepared for U.S.
Environmental Protection Agency.) EPA-
600/7-77-132.1977.
18. Drehmel, D. C., and D. F. Ciliberti. High
Temperature Fine Particle Control Using
Ceramic Filters. Westinghouse Research
Labs. (Prepared for U.S. Environmental
Protection Agency.) EPA-600/72-77-207.
NTIS PB 274485.1977.
19. Calvert, S., R. G. Patterson, and D. C.
Drehmel. Fine Particle Collection Efficien-
cy in the A.P.T. Dry Scrubber. In: Pro-
ceedings of EPA/DOE Symposium on High
Temperature High Pressure Particulate
Control Acurex Corp. EPA-600/9-78404.
1977. p. 399414.
20. Moore, R. H., G. F. Schiefelbein, G. E.
Stegen, and D. G. Ham. Molten Salt Scrubb-
ing For Removal of Particles and Sulfur
From Producer Gas. In: Proceedings of
EPA/DOE Symposium on High Tempera-
ture High Pressure Particulate Control
Acurex Corp. EPA-600/9-78-004. 1977. p.
430463.
21. Zahedi, K., and J. R. Melcher. Elec-
trofluidized Beds in the Filtration of A Sub-
micron Aerosol. Journal ofAPCA. 26(4)^45-
352.1976.
22. Kirsten, L. Private Communication with
Apitron Divison, American Precision In-
dustries, Inc. September 1978.
23. Peters, J. M. Predicting Efficiency of Fine-
Particle Collectors. In: Calculation and
Shortcut Deskbook. Chemical Engineering.
New York, McGraw Hill, Inc.
-------
A COAL GASIFICATION-GAS CLEANING PILOT PLANT:
OPERATING EXPERIENCE AND INITIAL RESULTS
J. K. Ferrell,* R. M. Felder, and R. W. Rousseau
North Carolina State University, Raleigh, North Carolina
Abstract
An integrated computer-controlled coal gasifi-
cation-gas cleaning pilot plant at North Caro-
lina State University is currently in preliminary
stages of testing. The gasifier is a 6-in diameter
fluidized-bed unit, with a coal feed capacity of
23 kg/hr (50 Ib/hr). The gas cleaning system con-
tains a cyclone, a venturi scrubber, and an ab-
sorber-flash tank-stripper system for acid gas re-
moval This paper describes the plant and asso-
ciated facilities for data acquisition, data log-
ging, and process control; summarizes proce-
dures for chemical analysis of all solid, liquid,
and gas feed and effluent streams; reviews re-
sults of recent runs; and outlines plans for fu-
ture tests.
INTRODUCTION
Many of the factors currently limiting the
large-scale development of coal conversion tech-
nology are environmental in nature. Many proc-
esses exist to gasify coal, some of which are
available commercially,' but the technology of
synthesis gas cleanup is less developed and the
total environmental impact of the implementa-
tion of gasification technology is not yet under-
stood.
Recognizing this problem, the U.S. Environ-
mental Protection Agency (EPA) in 1977 con-
tracted for the design and construction of a
pilot-plant coal gasification-gas cleaning test
facility at North Carolina State University
(NCSU), to be operated by faculty and staff of
the Department of Chemical Engineering. Con-
struction was begun in January 1978, and the
plant was completed and turned over to the Uni-
versity the following summer.
The principal components of the pilot plant
are a continuous fluidized-bed gasifier; a cyclone
separator and a venturi scrubber for removing
participates, condensables, and water-soluble
•Speaker.
species from the raw synthesis gas; and absorp-
tion and stripping towers and a flash tank for
acid-gas removal and solvent regeneration. The
gasifier operates at pressures up to 100 psig
(791 kPa), has a capacity of 50 Ib coal/hr (23
kg/hr), and runs with either steam-air or steam-
02 feed mixtures. The acid-gas removal system
is modular in design, so alternative absorption
processes may be evaluated. Associated with
the plant are facilities for direct digital control
of process systems and on-line data acquisition,
logging, and graphical display. Facilities for
sampling and exhaustive chemical analysis of all
solid, liquid, and gaseous feed and effluent
streams are also available.
The overall objective of the project is to char-
acterize completely the gaseous and condensed-
phase emissions from the gasification-gas clean-
ing process, and to determine how emission
rates of various pollutants and methanation cat-
alyst poisons depend on adjustable process pa-
rameters. Specific tasks to be performed are to:
• Identify and measure the gross and trace
species concentrations in the gasifier prod-
uct, including concentrations of sulfur gases
(H2S, COS), condensable organics (e.g., BTX
and polynuclear aromatic hydrocarbons),
water-soluble species (e.g., ammonia, cya-
nates, cyanides, halides, phenols, sulfates,
sulfides, sulfites, and thiocyanates), and
trace metals (e.g., antimony, arsenic, beryl-
lium, bismuth, cadmium, lead, mercury, se-
lenium, and vanadium).
• Correlate measured emission levels with
coal composition and gasifier operating var-
iables, particularly temperature, pressure,
and solid and gas phase residence time dis-
tributions.
• Perform material balances around the gas-
ifier, the raw gas cleanup system, and the
acid-gas removal system, and determine the
extent to which selected species are re-
moved from the synthesis gas in each of the
components.
• Correlate measured extents of conversion
499
-------
and removal efficiencies for various species
with system-operating variables, including
temperatures, pressures, holdup times, and
solvent circulation rates.
• Evaluate and compare the performance
characteristics of alternative acid gas
removal processes, considering both C02
and H2S removal capabilities and the
degrees to which the processes remove
trace pollutant species from the sour syn-
thesis gas. Evaluate the buildup of con-
taminates in the various acid-gas removal
solvents.
• Use the results obtained in the above
studies to develop models for the gasifica-
tion and the gas cleanup processes. The
models will take as input variables the com-
position and feed rate of the coal, bed depth,
steam and air (or oxygen) feed rates and in-
let temperatures, gasifier pressure, and
operating conditions (temperatures, pres-
sures, solvent flow rates, etc.) for the gas-
cleaning systems, and will predict the coal
conversion and the product gas flow rate
and composition, including trace pollutant
levels. The model will be used as a basis for
perfecting the pilot-plant operating condi-
tions, and for estimating emission levels for
scaled up versions of the processes investi-
gated.
The sections that follow briefly describe the
plant and its operation, provide illustrative re-
sults obtained in test runs, and outline future
test plans.
DESCRIPTION OF PLANT FACILITY
The pilot-plant facility consists of six sub-
systems:
'• Gasifier, coal feed, and char removal system;
• Particulates, condensables, and solubles re-
moval (raw gas-cleaning) system;
• Acid-gas removal system;
• Utilities system;
• Instrumentation and process control sys-
tems;
• Data acquisition and display system.
These subsystems are represented schematical-
ly in Figures 1 through 4.
Gasifier
The gasifier (Figures 1 and 2) is a 6-in (15.2-cm)
I.D. Schedule 40 pipe (316 88) enclosed in sever-
al layers of insulation and contained in a 24-in
(61-cm) I.D. Schedule 80 carbon steel pipe. The
overall height of the unit is roughly 12 ft (3.7 m).
Thermocouples are mounted in the center of the
bed at positions 10,20,30,40,50, and 60 in above
the gasifier feed cones to monitor the bed tem-
perature profile. Differential pressure taps are
set at 15 and 35 in above the feed cones, and the
pressure drop between these taps is used as an
operating parameter. The cones are three 1/2-in
(12.7-mm) diameter tubes arranged triangularly,
with each tube tapering out to 1 in (2.54 cm) for
better flow distribution.
Coal is fed and removed by screw conveyers
from pressurized hoppers at either end of the
vertical reactor. The bed height may be as low
as 3 ft (1 m) and as high as 5.5 ft (1.7 m). The level
of the fluidized bed is monitored with a nuclear
level gauge and kept constant by adjustment of
the char removal screw rotation rate. The coal
feed and removal systems contain nitrogen
purges to prevent back-flushing of any reac-
tants. The insulation section around the gasifier
is also equipped with a nitrogen purge flow for
safety considerations. The gasifier typically
operates at 100 psig (791 kPa) and between
1,600° and 1,800° F (1,150-1,250 K). Steam and
carbon react to form CO and H2; carbon combus-
tion also occurs. Carbon coversions on the order
of 30 to 50 percent have been obtained in prelim-
inary runs.
Particulates, Condensables, and
Solubles Removal (PCS) System
The raw gas produced in the gasifier is fed to
the PCS subsystem (Figure 2). A cyclone sepa-
rator removes most particulates, and a venturi
scrubber quenches the gas stream, removing
water-soluble and condensable compounds at
the same time. The quenched gas stream is fed
through a shell and tube heat exchanger to a
condensate-receiving tank. The heat exchanger
was added after excessive temperature in-
creases in the receiving tank and consequent
losses of volatile condensate components were
observed in initial runs.
Water in the receiving tank can be used on a
once-through basis or recirculated to the ven-
turi scrubber. The gas leaving the tank goes
through a second heat exchanger, to a mist elim-
inator, and then through either a coalescing or a
500
-------
N<
A
S
GASIFIER
PRE-HEATER
STEAM
SUPER-HEATER
onn
H2S
MIXED
GAS
BY-PASS
DRAIN
Figure 1. Utilities system.
-------
N2 PURGE
COAL
FEED
HOPPER
N2 PURGE.
N2 PURGE.
A
G
A
S
I
F
I
E
R
MEL
CYCLONE
V
VENTURI
SCRUBBER
CHAR
RECEIVER
N2 PURGE
CIRCULATION
PUMP
N2
STEAM
PLANT WATER
A
FILTER
>k
ELIMINATOR
HEAT
EXCHANGERS
V
PCS
TANK
AGRS
E><3> DRAIN
Figure 2. Gasifier—PCS system.
-------
ACID
GAS
I SOLVENT
.^NCHILLER
SOUR GAS
COMPRESSOR
HEATER
EXCHANGER
PUMP
Figure 3. Acid-gas removal system.
-------
PROCESS
INSTRUMENTS 1
D/A
CONVERSION
MICRO COMPUTER
VIDEO
DISPLAY
CRT
TERMINAL
OPERATOR'S CONSOLE
Figure 4. Data acquisition system.
DISK
STORAGE
TELEPRINTER
-------
cartridge filter. The pressure drop around the
filter is monitored; if plugging is observed, the
flow is directed to a parallel filter while the first
filter is cleaned or replaced. After leaving the
filter, the sour gas is either burned in a shielded
flare located on the roof or fed to the acid-gas
removal system.
Acid-Gas Removal System (AGRS)
The acid-gas removal system (Figure 3) is
designed to operate in four different modes —
with refrigerated methanol, with hot potassium
carbonate, with monoethanolamine, or with di-
methyl ether of polyethylene glycol. All exper-
ience to date has been with methanol, and only
this mode of operation will be described in the
remainder of the paper.
The AGRS can accept either a sour gas feed
stream from the gasifier or a synthetic gas
(Syngas) feed stream. The feed gas is first
passed through a dehydrator, then compressed
to 500 psig (3.54 mPa), cooled, and fed to an ab-
sorption column. The absorber contains approx-
imately 21.5 ft (6.5 m) of 1/4-in ceramic Intalox
saddles. The 5-in (12.7-cm) diameter column can
accept solvent feed at any of three locations,
which provides flexibility for mass transfer
studies. The sweet gas (whatever remains after
C02, H2S, and other sulfur gases are absorbed)
is then burned in the shielded flare.
The recirculating methanol is refrigerated to
about - 30° F (239 K) before being routed to the
absorber. After passing through the absorber,
the methanol is sent to a flash tank to reduce its
pressure to about 100 psig (791 kPa). It is then
sent to a trim heater before being fed to the
stripping column. The 6-in (15.2-cm) diameter
stripping column containing 22.5 ft (6.9 m) of
1/4-in ceramic Intalox saddles is operated at
about 10 psig (170 kPa), with nitrogen used as
the stripping gas. The column feed temperature
can be regulated by a trim heater. The solvent
is regenerated and sent through a gas chiller (to
further cool the entering sour gas) before being
sent to the refrigeration unit to undergo
another cycle.
Utilities System
Both the gasifier and AGRS are linked to the
utilities subsystem (Figure 1), which provides
'the feed streams to both systems. Nitrogen,
oxygen (or air), and steam are all regulated
through flow control loops to the gasifier, while
a prepared mixture of N2, C02, HgS, or other
gas mixtures can be fed to the AGRS in place of
gasifier make gas. The feed stream to the gasi-
fier is first preheated (N2,02/air) or superheated
(steam). The Syngas feed to the AGRS is mixed
and regulated through a flow control valve on
the sour gas compressor outlet.
Data Acquisition and Process Control
Systems
Plant operation is monitored and regulated
from a control room. Signals from 96 sensors
(temperature, pressure, flow rate, etc.) are sent
to a control panel, where they are processed and
sent to a video display terminal and/or a Honey-
well TDC 2000 process control computer and/or
a microprocessor-based plant data acquisition
system. The TDG 2000 regulates 16 different
control loops in the plant. An alarm panel super-
imposed on a process schematic provides visual
and auditory indications of potentially hazard-
ous conditions.
The data acquisition system has two main ob-
jectives: to provide rapid, easily read informa-
tion to the operator during plant operation, and
to provide a permanent record of run data. A
block diagram of the system hardware is shown
in Figure 4. Each process instrument is wired to
a channel of an L.F.E. Model 6100 Remote Ter-
minal Unit. The LFE 6100, a 96-channel analog-
to-digital converter, digitizes the 1- to 5-V
transmitter signals to 12-bit resolution and
transmits the results to the computer through a
serial communication line. The conversion takes
place every 15 a upon command from the micro-
computer.
The microcomputer used in the acquisition
system is an INTEL 8080A-based system. The
system includes two 8-in floppy disk drives and
32 kbytes of read/write memory. Another 4
kbytes of video display memory are directly ad-
dressable by the processor. The operator com-
municates with the system through a standard
CRT-type terminal, and hard-copy output is
available on a Decwriter II teleprinter.
Once every 15 a the remote terminal unit
transmits signals proportional to all 96 process
variables to the microcomputer. The informa-
tion is translated to engineering units with a
stored calibration function for each channel.
505
-------
Several calculated variables, such as the super-
ficial gas velocity within the fluidized bed, are
also displayed. Running sums are kept to allow
interval averaging of data.
The operator, through the use of commands
entered at the CRT terminal, controls subse-
quent data processing. By entering a "Display"
command, the operator selects one of eight
schematic representations of the plant shown
on the video display. Process information is
superimposed on these displays to provide an
easily readable display of information. The
display information is updated after every 15-s
scan.
Using the "print-on" and print-off* com-
mands, the operator can control hard-copy out-
put during a run. The interval average of se-
lected channels is printed out at the end of a
designated interval. Similarly, the "save" and
"no-save" commands are used to control the
storage of data on disk. Data are written in
blocks including all channels and calculated
variables. This interval is also specified by the
"Interval" command. One 8-in disk holds all in-
formation collected during a 12-hr run.
In the initial runs of the pilot plant it became
apparent that the acquisition system could be
useful for more than simple information collec-
tion. A capability of displaying trends of par-
ticular variables vs. time would greatly facili-
tate plant operation, for example, and the imple-
mentation of a "snapshot" function to record the
sequence of events preceding an alarm-oriented
shutdown of the plant might provide enough in-
formation to prevent a second similar occur-
rence. To provide the computing capability
needed to implement such data management
functions, a Digital Equipment Company PDF
11/34 minicomputer and a color graphics ter-
minal have been ordered to replace the present
microcomputer facility.
ANALYTICAL LABORATORY FACILITIES
Solid, liquid, and gas samples from the pilot
plant are analyzed in four analytical labora-
tories. Compounds and major, minor, and trace
elements that are analyzed are listed in Table 1.
Brief descriptions of the laboratory facilities are
given in the sections that follow.
Main Laboratory
The main laboratory is a general purpose
laboratory, in which ultimate and proximate
analyses of coals and chars are carried out.
Equipment available for these analyses includes
furnaces, ovens, and combustion trains con-
structed and installed following American
Society for Testing and Materials (ASTM)
guidelines.
The main laboratory also houses a water puri-
fication system consisting of a deionizer and a
water still; several macrobalances, semimicro-
balances, and microbalances; glassware; re-
agents; and four instruments for analysis of se-
lected pollutants in the plant wastewater. These
instruments are a Dionex System 10 ion chro-
matograph, an Orion Model 901 selective ion-
analyzer, a Dohrmann Model DC -50 carbon ana-
lyzer, and a Bausch & Lomb-Shimadzu Spec-
tronic 210 UV-Visible spectrophotometer.
Trace Analysis Laboratory
This laboratory is devoted to the analysis of
trace elements by atomic absorption spectro-
photometry. Instruments housed in the labora-
tory include a Perkin-Elmer Model 603 atomic
absorption spectrophotometer with a deuter-
ium arc and various types of flames, a Perkin-
Elmer HGA-2200 graphite furnace, a Perkin-
Elmer mercury analysis system, an LFE Model
LTA-504 low-temperature plasma asher, and a
Barnstead water deionizer.
Coal Research and Analysis Laboratory
This laboratory is equipped for the study of
coal pyrolysis and the analysis of sulfur, nitro-
gen, and free-swelling index in coals and chars.
The instruments housed in the laboratory in-
clude a Fisher Scientific Model 470 sulfur ana-
lyzer, an Antek Model 707 nitrogen analyzer,
and a laminar flow furnace reactor capable of
operation at temperatures up to 1,273 K with
particle residence times as low as 50 ms.
Gas Chromatography Laboratory
The chromatography laboratory is equipped
for analysis of fixed and condensable species in
gas samples, and for analysis of BTX and pheno-
lics in wastewater samples. Instruments in this
laboratory include two Tracer 550 gas chromat-
ographs equipped with flame ionization and
thermal conductivity detectors, a Varian 3700
506
-------
TABLE 1. SUMMARY OF COAL GASIFICATION ANALYTICAL PROGRAM
Sample Type
Analysis
Analyte
coal/char
Proximate
Ultimate
Trace Element
Sieve analysis, density, free swelling
Index
Moisture, ash, volatile matter, fixed carbon
C, H, N. 0, S
As, Be, Cd, Cr, Hg, Ni, Pb, Sb, V
Gas/solvents
Compounds
, CO, C02,
i, COS,
CS,
Trace Elements As, Be, Cd, Cr, Hg, Ni, Pb, Sb, V
Wastewater
Major Elements
Compounds
Trace Elements
C, N, S
Ammonia, total organic carbon, chloride, COD,
cyanate, cyanide, pH, phenolics, residue,
sulfate, sulfide, sulfite, thiocyanate,
benzene, toluene, xylene.
As, Be, Cd, Cr, Hg, Ni, Pb, Sb, V
gas chromatograph equipped with thermal con-
ductivity and dual-flame photometric detectors,
and a Perkin- Elmer Sigma X chromatograph
data station.
PLANT OPERATIONS
Gasifier Operation
The gasifier and PCS system are pressurized
by starting a flow of process nitrogen through
the gas feed preheater. The preheater and gasi-
fier pressure controllers are set at 1,000° F
(811 K) and 100 psig (791 kPa), respectively. Coal
feed is commenced when the reactor tempera-
ture reaches about 450° F(506 K), with the nitro-
gen flow maintained at a level sufficient to
fluidize the bed as it forms. During this time,
steam flow is started through the steam super-
heater, also set at 1,000° F (811 K), and the reac-
tor bypass duct.
When the bed temperature has reached
700° F (644 K) with the bed height between 20
and 30 in (50 and 76 cm), a small flow of oxygen
is started. At this temperature, the bed almost
always ignites. After ignition, the bed tempera-
ture is brought to about 1,450° F (1,061 K) by a
slow increase of oxygen flow, and a small flow of
superheated steam is diverted into the reactor
from the bypass. To achieve the desired steady-
state conditions, nitrogen flow is gradually de-
creased, steam flow is increased, and oxygen
flow is adjusted to maintain the reactor temper-
ature at the desired value. All of the above
changes must be made smoothly; good results
are usually obtained if sudden large changes in
the superficial gas velocity can be avoided.
The steady-state coal feed rate is established
through control of the speed of the coal feed
screw to maintain the desired feed rate and ad-
justment of the removal screw speed to main-
tain the desired bed height as indicated by the
nuclear bed level gauge. During startup, the bed
height can be monitored if signals are observed
from the temperature sensors located in the
bed, the bed differential pressure sensor, and
the nuclear bed level gauge.
507
-------
When the bed is well fluidized from the out-
set, the process described works well, and reac-
tor startup is last and smooth. For a number of
reasons, the bed is often not well fluidized dur-
ing the startup period, and a variety of difficul-
ties occur. The probable causes are hot spots
because of poor mixing in the bed and bed ag-
glomeration, which result in the bed being lifted
to the top of the reactor.
Figure 6 shows a history of a startup of the
gasifier. Shown plotted vs. time are the reactor
bed temperature at 10 in above the gas feed
cones, the pressure drdp measured across 20 in
of the bed, and the calculated superficial gas
velocity in the bed. Noted on the figure is the
time when coal feed and oxygen feed were
started. Apparently, one or more of the upset
conditions noted above occurred after oxygen
feed was started, and a steady operating condi-
tion was obtained only after several hours of er-
ratic behavior.
Once a good steady state has been obtained,
the operation is stable and cannot be easily
upset. We believe that one reason for the dif-
ficulty of operation during startup and ease of
operation during steady state is the difference
in the manner in which the fluidizing gas is dis-
persed. During startup, with no reaction in the
bed, all of the fluidizing gas emerges in jets
from the three feed cones and is not well distrib-
uted. During steady-state operation, the carbon-
steam reaction causes a progressive increase in
the gas flow rate, and the carbon-oxygen reac-
tion increases the gas temperature in a zone
just above the cones. Both of these factors act to
increase the gas turbulence and to improve the
distribution across the bed.
Researchers carry out the startup procedures
by using the TDC 2000 controller, making set
point changes to effect changes in process var-
iables. During startup, the reactor temperature
is controlled manually by adjustment of the coal
and gas feed rates. When the desired steady-
state values of bed level, and coal, steam, and
nitrogen feed rates have been established, the
reactor temperature and oxygen flow control
loops are cascaded so the temperature is con-
trolled by the oxygen flow rate.
The steady-state operation of the gasifier-
PCS system is illustrated by plots of selected
process variable vs. time in Figures 5, 6, and 7.
Figure 5 shows three different steady-state con-
ditions. The first of these, designated Run
GO-5, used a coal feed rate of 60 Ib/hr (22.7
kg/hr), a bed height of 52 in (132 cm), a steam
feed rate of 25 Ib/hr (11.4 kg/hr), and flows of
nitrogen and oxygen adjusted to give a bed tem-
perature of 1,800° F. For the second steady
state, GO-6, the steam feed rate was increased
to 30 Ib/hr (13.6 kg/hr) while the partial pressure
of steam in the feed gas and the reactor temper-
ature was held constant. The gas residence time
in the reactor was thus decreased. For GO-7,
the steam rate was reduced to 20 Ib/hr (9.0
kg/hr), and the velocity and gas residence time
were made the same as those of GO-5.
As noted, when the bed is well fluidized a
good steady state can be achieved, as indicated
by the constancy of the feed flow rates, bed
temperatures, reactor pressure, etc. An
example of a poorly fluidized bed and poor
steady state is shown in Figure 6 for Run
GO-13, and an example of operation with a well-
fluidized bed is shown in Figure 7 for Run
GO-14. During the early part of Run GO-13 the
bed was obviously not well fluidized, as evi-
denced by the erratic behavior of most of the
process variables shown in Figure 6. The data
indicate that while the upper portion of the bed
may have been well fluidized, the region in the
vicinity of the 10-in thermocouple was not. A
zone nearly devoid of solids probably exited at
this point, suggesting that the bed had agglom-
erated and lifted. At approximately 13:30 the
bed temperature was raised to 1,250 K for a
time and then returned to its former set point.
As can be seen, this upset resulted in an im-
proved operation.
Also shown on Figures 6 and 7 are the times
when samples were drawn for analysis at the
sample point locations shown on the plant sche-
matics (Figures 2 and 3). All gas samples are
taken in heated 1-L sample cylinders. Raw gas
samples at 100 psig are drawn from the cyclone
and PCS system exits. Also available are high-
and low-pressure samples of cleaned and cooled
gas drawn from a sampling train at the cyclone
exit (Figure 8). In addition to providing a clean
gas sample, the sampling train allows for a
gravimetric determination of the water content
of the gasifier effluent and provides liquid
samples that may be analyzed for condensable
and soluble species in the effluent. Integrated
liquid samples can also be taken from the receiv-
ing tank following the venturi scrubber. Wher-
ever they are obtained, liquid samples are im-
508
-------
10
0.8
8
u
tt
0.4
QQ 4-
0.2
2-
8:00
BED TEMPERATURE
j I
10:00 12:00 14:00
TIME OF DAY
2000
1400
1200
a
800 ca
400
I I I I
16:00
Figure 5. Startup and steady-state data for runs GO-5, GO-6, GO-7.
-------
en
i-«
o
2 S6.75
«J2
£s.7o
E0.40
3-
S £ Q.35
0.30
REACTOR PRESSURE
0 00 0
MAKE GAS FLOW
11:00
12:00
13:00
14:00
15:00
1150
1100
2-5
«,
m
2.0 >
sr
1.5 2
1.0
16:00
Figure 6. Run GO-13 steady state.
-------
en
1250
1200
0.45
UJ
=>
0.30-
BED TEMPERATURES
REACTOR PRESSURE
MAKE GAS FLOW
o
11:00
12:00 13:00
TIME of DAY
14:00
6-75 =
CO
6.70£
2.5
2.0 ^
CL
«a
I
-------
HEAT
TRACING
01
i-»
to
RAW GAS
SAMPLING
PORT
CYCLONE
COOLING
WATER
SIGHT
GLASS
->TO PCS
REMOVAL
SYSTEM
SPIRAL
EAT EXCHANGER
FLOW
CONTROLLER
COOLING
WATER
ROTAMETER v-»
-M-
HEATER
LOW PRESSURE
SAMPLING PORT
\HIGH PRESSURE
SAMPLING PORT
FILTER & DEMISTER
WATER
SAMPLE PORT
CONDENSATE TRAP
Figure 8. Sample system located at cyclone exit.
-------
mediately subjected to appropriate preserva-
tion steps and are stored to await subsequent
analysis.
AGRS Operation
The absorber is pressurized to 95 psig
(756 kPa) using Syngas nitrogen bypassed
around the sour gas compressor. The flash tank
is pressurized to 30 psig (308 kPa) with process
nitrogen, and the stripper is pressurized to
10 psig (170 kPa) with stripping nitrogen. After
these pressures are achieved, solvent flow is
begun at 1.5 g/min (5.7 L/min) and the solvent
chiller is started and set at - 30° F (239 K).
After the solvent flow is well established, the
absorber is pressurized to 500 psig (3.5 MPa)
with Syngas nitrogen using the sour compres-
sor, and the flash tank is brought to 70 psig (584
kPa) using process nitrogen. These pressures
and flows are maintained during the remainder
of the cool-down period. During this period, the
gas flow rate to the absorber is kept as low as
possible to help increase the cooling rate.
When the absorber and stripper are near
their final temperatures, the solvent and sour
gas flow rates are set at their steady-state val-
ues and the desired flow rates of Syngas and
stripping nitrogen are also set. The composition
of the gas leaving the flash tank is monitored,
and when acid gas is detected in appreciable
quantities, the process nitrogen is turned off
and the flashing gas is used to maintain the de-
sired pressure.
The approach to steady state is monitored
using an on-line carbon dioxide analyzer. In the
near future an on-line analyzer that will monitor
both hydrogen sulfide and total sulfur will also
be used to define the approach to steady state.
Figures 9 and 10 show the transient and
steady-state values of selected process var-
iables for a typical AGRS run. Shown plotted vs.
time are feed gas temperature, two tempera-
tures in the stripper, and three temperatures in
the absorber. The results will be discussed in
more detail in a later section.
After steady-state conditions have been
achieved, samples are taken of the feed gas,
sweet gas from the top of the absorber, flash
tank gas, and acid gas from the stripper. At-
tempts to sample gas at various points in the
columns have been complicated by liquid en-
trainment. Liquid sampling, especially at var-
ious points in the columns, has also proven diffi-
cult. Both problems are currently being worked
on, with different sampling port designs consid-
ered.
ILLUSTRATIVE RESULTS:
GASIFIER OPERATION
Shown on Figure 7 are process conditions for
Run GO-14, carried out February 2, 1979. The
long residence time of solids in the bed (roughly
30 min) probably accounts for the long time re-
quired for the make gas flow rate to reach
steady state.
The raw plant operating and gas analysis
data for Run GO-14A were processed to gen-
erate input for a data logging and material bal-
ance program. (The designation 14A refers to
the period between 12:30 and 13:30, after which
conditions changed in the plant.) The output
from this program is shown in Figures Ha and
lib. The paragraphs that follow summarize the
calculated results and the calculations used to
generate them, more or less in the order in
which the results appear on the computer print-
out.
Reactor Specifications
The reactor pressure and average bed tem-
perature were 103 psig (811 kPa) and 1,792° F
(1,251 K). The reactor diameter is a fixed 6 in
(15.2 cm), and the bed height was controlled at
38 in (97 cm).
The pressure drop in the bed over a fixed
length was measured and used to calculate an
apparent density of the expanded bed. From
this quantity and the known densities of the
solid and gas phases, the bed voidage was deter-
mined to be 0.79 ft3 void/ft8 reactor. The expan-
sion factor is then calculated from this value and
the known settled bed density is 1.95 ft9 ex-
panded bed/ft3 settled bed.
It has so far not been possible to eliminate
leakage from the reactor, particularly around
the feed and char removal screw conveyors. The
magnitude of this leakage is estimated before
each run in both static and dynamic tests, and
the result is incorporated into material balance
calculations. In Run GO-14A, the leakage rate
was estimated to be 0.55 stdft8/min (16 L [STP]/
min), roughly 3 percent of the product gas flow
rate.
513
-------
60H
ABSORBER LOWER SECTION
h290
f-280
9:00
U270i
as.
I
1-260
1-250
240
10:00 11:00
TIME of DAY
12:00
13:00
Figure 9. Run AM-4 startup temperature data.
-------
60-
-290
40-
20-
en
i-*
01
o -
-20-
I
12:00
ABSORBER SOUR GAS FEED
'Sampling point
Interupt to change C02 tank
STRIPPER LOWER SECTION
STRIPPER TOP
ABSORBER BOTTOM
ABSORBER LOWER SECTION
-280
r-270 £
3
I—
<£
as.
UJ
a.
E
f-260 £
-250
-240
13:00 14:00 15:00 16:00
TIME of DAY
Figure 10. Run AM-4 steady-state temperature data.
I
17:00
-------
A******************************************
* NCSLLDEPA.RItlEUT OE..CHtMICAL ENGINEERING 5
* FLUIDIZEO BED COAL GASIFICATION REACTOR *
* *
RUN CO-HA 2/20/79, 12:30-11 IS RUN *IAH£TF« = 6
i:NT s 0
, 10X80 Hf.5>H
12.2 Lb/Ft**3 C1.797 G/CM»*3)
46. 2. Lb/£ T* *3 CO»7ttO C/CM«*3)
51.01 HTCRC.NS
.OOhJ
ULTIMATE ANALYSES(ESTIMATED)
CARBdN
HYDROGEN
OXYGF.N
NITPOGEN
SULFUR
ASH
STEAM/CO
02/CQAL
COMBINED
C
-------
RUN CO*1«A 2/20/791 12I30-H15 RUN »4 ON EXPERIMENTAL PLAN
CONTROL VARIABLES OUTPUT VARlABLtS
TEMPERATURE "= 1792.0 DEGTF'PRESSURE OHOP OVER 20 "IN.s~~'i,ir i*f."H2o~Y~bV27sPSD
N2/02 MOLAR FEED RATIO s l»26 = 1.9 KP*
STEA& PARTIAL PRESSURE = **.16 PSI PCS GAS FLOH RATE = 2.02 LB-MOLE/HK (12.08 SCFM)
PACE TIME I \b\\ M!NCYCLONE GAS >LOW RATE
CE TIME = 5.31 S
GAS SPACE TIME = 5.31 S a 2.01 LR-MHLt/HKCDP.r 6A3IS)
LJii
50
L6
FUEL PROPERTIES CONVERSION
JASJSJ __ _. CARBON.cfiNytRsroN = 33.7x
* 0,047 LB HZ PHOOUCED/LB COAL(MAF)
CH<» s I ix THULAK BASIS)
_. * OtSll LB CK/» PHODUCEO/L0 CfUL(HAF) SOLID MATERIAL BALANCE;
HEATING VALUE OF MAKE GAS = 3S99.6 BTU/LB COAL FED = 233.2 LB
= 211.5 PTu/stF sp^NT CHAM COILECTEQ s i«s,i L^S 6?»2i_LF F
« -flSbS.n Kj/KfcCYCLONE OUST COLLtCTEOs 6,5'Ch -~ 2.8* or"F
CUAL GASIFIED 9 61.6 LB s 35.OX OF '
HEATING VALUE OF SHEET GAS * 6124.1 BTU/LB
= 268,2 BTU/SCr SPENT CHAR REMOVAL RATF = 20^« L8/HR
= 14232!* KJ/KG CHAR RATE FOR HAS5 BALAN~CF"sr23t6~TB7HR
CYCLONE EXIT GAS ANALYSIS PCS EXIT GAS AN»LTSIS ELEMENTAL MATERIAL BALANCES | FLH»4 c.oi1) " >.oo i.oo A."2o TOTAL INPUT as.o 26.33 2,95 3i,7t i^.^u
C02 U.91 16.90 0.340 16.63 16.66 •».»36
N2 22,62 32,JO 0,617 32.37 32.15 ^fi>^^ CHAR - 20.8 17.38 0.03 0.3<- 0,10
«AsTtH«TtR o!o o!o do "o!o o!o
.. 0.0 0.0 0.0 0.0 0,0
TOT*L OUTPUT 82.1 26.79 3.IP 30.94 lb.21
— r DIFFERENCE -3.3X IT?* 5",2X ~^Z^* S673T
Hgura 11b. Run GO-14A data output.
-------
Solid Feed
The feed to the reactor was a devolatilized
Western Kentucky #11 coal, pretreated at
2,000° F (1,100° C) and pulverized and screened
to 10 x 80 mesh. The proximate and ultimate
analyses of the feed coal are shown on the out-
put page, along with ultimate analyses of the
spent char and dust collected in the cyclone. The
term "estimated" above the ultimate analyses
signifies that the given number was obtained in
a previous run under similar conditions; time
and manpower limitations prohibit analysis of
solid samples following every run.
Feed Specifications
Goal was fed at a rate of 31.2 Ib/hr (14.15
kg/hr). Steam was fed at a rate corresponding to
0.95 Ib H20/lb coal (moisture- and ash-free basis),
and oxygen was fed in a ratio of 0.35 Ib 02/lb coal
(MAP). To prevent the feed nozzles from being
burned, nitrogen was fed at a rate correspond-
ing to 1.3 mol N2/mol 02. Therefore, the reactor
could not be considered strictly oxygen fired or
air fired but was much closer to the former.
In the operation of the gasifier, a separate
stream of nitrogen (purge nitrogen) is fed
through the insulating shell and the feed and
char removal screws, eventually combining
with the reactor effluent gas stream. The flow
rate of this stream was 1.9 stdft3/min (4 kg/hr).
The superficial gas velocity in the reactor is
evaluated by assuming a molar gas flow rate
equal to that of the feed gas (steam + 02 + N2),
converting to a volumetric flow rate at the
mean reactor temperature and pressure, and di-
viding by the total reactor cross-sectional area.
The calculated velocity in Run GO-14A was 0.60
ft/s (0.18 m/s).
The minimum fluidization velocity was calcu-
lated from a correlation of Babu et al.,1 after the
feed gas viscosity at the reactor temperature
and pressure was determined using correlations
of Rohsenow and Hartnett.2 The actual super-
ficial velocity was found to be 1.8 times the esti-
mated minimum fluidization velocity.
Control and Output Variables
Several parameters to be used for subse-
quent correlation analysis are summarized in
the output sheet. They include the solid holdup
(14.7 Ib, 6.7 kg), estimated as the apparent solids
density in the bed times the bed volume; the
solid space time (28.3 min, solids holdup/coal
feed rate), and the gas space time (5.31 s, bed
height/superficial gas velocity). Also shown are
the measured pressure drop in the bed, the gas
flow rate (corrected for leakage) measured
following the PCS removal system, and the gas
flow rate at the cyclone outlet, calculated from
the PCS gas flow rate by assuming that the
molar flow rate of dry gas is the same at the two
points.
Product Fuel Properties, Conversion
Variables, and Solid Material Balance
Most of the remaining quantities shown in
Figure 11 are derived from a chromatographic
analysis of the cyclone exit gas. As of the date of
the run shown, reliable measurements of sulfur
gases could not be obtained, so values shown on
the output page referring to sulfur have no sig-
nificance.
The fuel properties of the make gas are first
summarized: these include the molar percent-
ages of carbon monoxide (22 percent), hydrogen
(37 percent), and methane (1 percent), and the
heating values of the make gas and sweet gas.
The make gas is defined as the cyclone effluent
gas with water and purge nitrogen subtracted,
and the sweet gas is the make gas with C02 and
sulfur gases removed.
The carbon conversion in the gasifier is calcu-
lated as the mass flow rate of carbon in the
product gases divided by the feed rate of carbon
in the coal. A 34-percent carbon conversion was
obtained in Run GO-14A. The steam conversion
was 39 percent.
A solid material balance for the total time
period of the run was obtained by weighing the
total amounts of coal feed and spent char and
cyclone dust collected, and determining the coal
gasified by difference. The value of 35 percent
gasified is consistent with the previously cited
34 percent carbon conversion.
The rate at which spent char is removed dur-
ing the steady-state period (21 Ib/hr, 9.5 kg/hr) is
determined from the known rotational speed of
the screw conveyor and the total mass of spent
char collected. Also shown on the output page is
the char removal rate that would close the total
mass balance on the gasifier.
518
-------
Gas Analyses and Elemental Material
Balances
Chromatographic analyses of the gases at the
cyclone and PCS system exits are shown next
on the output page. The measurement of water
in the cyclone gas was subject to considerable
error in this run, and the estimated value of 29.5
percent may be off by as much as 5 percent.
The mass flows in and out of the unit of C, H,
0, N, and total mass are listed, and the percent-
age differences between input and output are
shown. Better closures in the material balances
are anticipated as sampling and analysis proce-
dures become more refined.
ILLUSTRATIVE RESULTS:
AGRS OPERATION
The acid-gas removal system functioned well
mechanically during three initial runs with a
pure nitrogen gas feed. The objects of these
runs were to check the mechanical operation of
the system, to obtain column hydraulic data for
pressure drop and flooding calculations, and to
calibrate and tune all instrumentation and con-
trol loops.
Thus far, only one run (AM-4) has been con-
ducted using a synthetic acid gas (C02 and N2)
feed. The objectives of Run AM-4 were to eval-
uate system performance and on-line sampling
and analysis techniques, observe system per-
formance over an extended period of operation,
provide operating experience for project per-
sonnel, and obtain qualitative information for
the development of an experimental plan. Other
sulfur gases, including H2S, were not used but
will be used in future runs. The results of the
gas analysis for this run appear in Table 2. All
gas compositions are reported on a methanol-
free basis; only trace quantities of methanol
were detected in gas analyses.
A temperature-time plot of several system
parameters for Run AM-4 is shown in Figure 9
(transient period) and Figure 10 (steady-state
period). As can be seen from these plots, approx-
imately 3.5 hr were required for the system to
cool down to its desired value of - 30° F (239 K),
with solvent flow set at 1.5 g/min (5.7 L/min). All
three packed tower sections were used for mass
transfer. Gas was fed to the absorber at approx-
imately 7.5 stdft3/min (212 L [STPl/min). The
feed rate of N2 to the stripping tower was 1.1
stdft8/ min (31 L [STP]/min). Quantitative
measurement of all outlet flows and composi-
tions for mass balance purposes was not possi-
ble at the time of the run. The temperature of
the solvent feed to the stripper was not con-
trolled but was fixed by the absorber bottom
temperature.
TABLE 2. DATA FOR AGRS RUN AM-4
Time
Location
Composition
Hole %
CO,
15:30
16:00
Feed gas
Absorber top
Flash tank
Stripper exit
Feed gas
Absorber top
Flash tank
Stripper exit
76.8
100.0
72.3
35.1
76.9
100.0
73.2
32.5
23.2
-
27.7
64.9
23.1
-
26.8
67.5
519
-------
After the system was started with N2 flow,
C02 was added to the feed gas. A substantial
solution exotherm quickly became apparent in
the absorber. While the top section in the ab-
sorber showed only a slight effect, the temper-
ature in the absorber bottom reservoir rose con-
siderably. The temperature measured in the
lower section of the column packing also re-
flected a milder exotherm than that observed
for the absorber bottom. These observations
suggest that for these solvent and gas flows, a
significant amount of the mass transfer takes
place in a small fraction of the packed tower.
This is further substantiated by the fact that all
the C02 was absorbed in the column. In future
runs, information will be obtained using only
the bottom section of packing for mass transfer
with higher inlet C02 concentrations and lower
solvent rates. Also, additional temperature
measurement capability will be installed in the
lower section.
The temperature profile in the stripper also
varied. The stripper inlet temperature rose as a
result of the absorber bottoms temperature in-
crease, while temperature in the lower section
of the stripper fell as a result of the desorption
endotherm. The thermal effects were not con-
fined to a particular column section, as they
were in the absorber.
FUTURE PLANS
The gasifier will be run with devolatilized
bituminous coal feed through the summer of
1979. The precision of the analyses and the mass
balance closures associated with the gasifier
operation are nearly at a satisfactory level, but
some refinement in procedures is still required.
Once these refinements have been imple-
mented, a designed series of experiments will
be carried out to study the effects of operating
temperature, solid and gas phase residence
times, and feed gas composition on carbon con-
version and sulfur gas and trace pollutant emis-
sion levels.
The acid-gas removal system will be sub-
jected to a series of tests with Syngas feeds—
mixtures of C02, H2S, CO, and H2 in nitrogen
with refrigerated methanol as the solvent. Mass
transfer parameters will be determined and cor-
related with the absorber and stripper tempera-
ture and pressure and the gas and solvent flow
rates. During this period, test runs of the in-
tegrated gasifier-gas cleaning facility will be
performed, with the PCS system effluent gas
serving as the AGRS feed gas. This will even-
tually be the normal mode of operation of the
plant; Syngas runs will only be performed in the
initial stages of the test program for each new
solvent. The development of mathematical
models to correlate the performance of both the
gasifier and AGRS systems will be carried out
in parallel with all experimentation.
Beginning in the fall of 1979, a nondevolatil-
ized lignite or subbitumipous coal will be used
as the feedstock to the gasifier, with refriger-
ated methanol still being used as the AGRS sol-
vent. After several months of integrated plant
operation, a new absorption process will be im-
plemented and tested. A decision concerning
process has not yet been made.
REFERENCES
1. Babu et al. Institute of Gas Technology, Re-
port No. E<49-18)-1930. Chicago, 111. 1976.
2. Rohsenow, Warren, and J. P. Hartnett.
Handbook of Heat Transfer. New York,
McGraw-Hill, 1972.
520
-------
CHEMICAL ANALYSIS AND LEACHING OF COAL
CONVERSION SOLID WASTES
R. A. Griffin*, B. M. Schuller, S. J. Russell, and N. F. Shimp
Illinois State Geological Survey, Urbana, Illinois
Abstract
Five solid wastes from coal conversion proc-
esses were characterized chemically and miner-
alogically. The wastes included three Lurgi
gasification ashes and mineral residues from the
SRC-/ and H-Coal liquefaction processes. Chem-
ical analyses of the solid wastes were performed
for 60 constituents. Mineralogical character-
ization of the solid wastes was carried out using
X-ray diffraction, Mdssbauer spectroscopy,
scanning electron microscopy, and optical
techniques.
Leachates generated from the solid wastes at
eight pH levels and under two different gas at-
mospheres were analyzed for over 40 chemical
constituents. Thermodynamic speciation of in-
organic ions and complexes in solution were
modeled. There were 115 aqueous species con-
sidered in the model, and saturation data were
computed for over 100 minerals.
Results of the mineralogical characterization
and leachate analyses showed a wide range in
constituent concentration and in the minerals
present in the solid wastes. However, thermo-
chemical modeling demonstrated that similar
mineral phases controlled the aqueous solubility
of the major ionic species for all five solid
wastes.
INTRODUCTION
Although the fuels produced by coal gasifica-
tion and liquefaction processes are free of cer-
tain pollution hazards (e.g., sulfur), accessory
elements from the coal may be present in these
fuels or concentrated in the waste streams.
These waste products must be characterized
before environmentally acceptable methods for
their disposal can be developed.
Until recently, primary emphasis had been on
characterizing airborne contaminants from coal
conversion processes. However, several investi-
•Speaker.
gators, including Cavanaugh and Thomas,1
Cavanaugh et al.,2 and Somerville and Elder,8
have recently characterized the waste streams
from low/medium-Btu gasifiers. Filby et al.,4
have characterized the trace elements in the
solid wastes from the SRG-I liquefaction proc-
ess. These waste characterizations are impor-
tant, as demonstrated by the work of Sinor,5
who determined that the flow rate of Ni, As, Cd,
and Pb from a Lurgi gasification plant may be
environmentally significant. Because of the
large quantities of raw materials consumed,
potentially hazardous accessory elements may
be discharged, even though these elements may
be present in the waste in low concentrations.
Because the quantity of solid wastes pro-
duced from coal conversion processes can be
large and variable (Griffin et al.9), the wastes
must be characterized in detail. However, char-
acterization alone is insufficient for evaluating
acceptable waste disposal methods. Therefore,
it is necessary to determine which elements can
be leached from the wastes and under what cir-
cumstances.
The solubility of the accessory elements in
coal conversion ashes and residues has not been
thoroughly investigated. Some gasification
ashes and liquefaction residues are produced
under relatively severe conditions, namely, at
high temperatures and/or pressures. Liquefac-
tion residues are produced under a reducing at-
mosphere. Such conditions can alter the miner-
alogy and subsequent solubility of accessory
elements in the feed coals, thus affecting poten-
tial release of pollutants.
The application of equilibrium solubility
models can provide useful insights into the
chemistry of aqueous systems. Equilibrium
models provide, at a minimum, boundary condi-
tions within which questions may be framed.
For example, a typical environmental problem
solved by equilibrium models is one of predict-
ing the highest concentration of a given consti-
tuent that can be achieved in solution before
precipitation occurs with a given solid phase.
521
-------
Solutions to such problems can be useful in de-
veloping a "worst case" scenario for a given pol-
lutant leaching from a solid waste. Such solu-
tions set the upper boundary for concentrations
of the pollutant that will have to be dealt with
under a given set of conditions.
Applications of solubility models to environ-
mental problems must be interpreted with care.
For example, it is not uncommon to find large
discrepancies in literature values for the solubil-
ity products of some mineral phases. The value
of the solubility product may depend on the ap-
proach to equilibrium, using well-defined
crystals vs. precipitation, and phenomena such
as phase transitions, aging, colloid formation,
and differences in particle size. These factors,
along with slow attainment of equilibrium and
the presence of impure minerals in nature as op-
posed to the pure minerals used to determine
solubility constants, may obscure solubility rela-
tionships and their application to practical en-
vironmental problems.
Important factors controlling the solubility of
mineral phases include pH, redox environment
of the system, oxidation state of the mineral
components, concentration and speciation of in-
dividual inorganic and organic ions and com-
plexes in solution, and ionic strength (total solu-
ble ions). Application of results from solubility
models to real environmental conditions re-
quires considerable caution. Nevertheless,
assuming that the activities are calculated cor-
rectly and that the equilibrium constants are nu-
merically factual, the models should accurately
predict the solubility of an ion under a given set
of conditions for an exhaustive list of solid
phases.
Purpose
The purpose of this study was to investigate
the potential pollution hazards of selected coal
conversion solid wastes. The project is part of
ongoing research by the Illinois State Geologi-
cal Survey into the characterization of coal and
coal residues (Ruch et al.,7 Ruch et al.,8 Ruch et
al.,' Gluskoter,10 and Gluskoter et al.11). The five
wastes chosen for this study included three
Lurgi gasification ashes from runs employing
three different feed coals and two liquefaction
residues —an SRC-I dry mineral residue and an
H-Coal vacuum still bottoms mineral residue. It
is beyond the scope of this study to describe the
three coal conversion process technologies,
which are available elsewhere (e.g., Braunstein
et al,12 and Parker and Dykstra13).
To assess the solubility of the accessory ele-
ments contained in the solid wastes, this study
was developed in four stages:
• Chemical characterization of the solid
wastes,
• Mineralogical characterization of the solid
wastes,
• Determination of the soluble constituents
from the solid wastes, and
• Application of thermochemical equilibrium
modeling to determine the mineral phases
controlling the solubilities of accessory
elements in the solid wastes.
CURRENT STUDIES OF THE
SOLUBILITY OF COAL GASIFICATION
AND LIQUEFACTION SOLID WASTES
Sources of Gasification Ashes
and Liquefaction Residues
During 1973 and 1974, the American Gas As-
sociation and the Office of Coal Research
studied the performance and suitability of vari-
ous American coals for gasification by the Lurgi
process. Four different coals were sent to
Scotland, where they were gasified in the Lurgi
plant at Westfield. Among these four coals were
5,000 tons each of Illinois No. 6 and No. 5 (seam)
coals and a Rosebud (seam) coal from Montana
that was gasified. The unquenched waste ash
was then sent back to the United States, where
it has been used in several studies. The samples
of Illinois No. 5 and No. 6, and Rosebud Lurgi
ash, for which data are reported here, were sup-
plied to us by Peabody Coal Company's Central
Laboratory at Freeburg, Illinois.
The H-Coal liquefaction residue was obtained
from Hydrocarbon Research, Inc., Trenton, New
Jersey. The residue was the vacuum still bot-
toms generated during production of a fuel oil
product using an Illinois No. 6 (seam) coal and
the H-Coal® pilot development unit at the HRI
Trenton Lab May 3, 1976.
The SRC-I liquefaction dry mineral residue
was obtained in September 1976 from the Pitts-
burg and Midway Coal Mining Company solvent-
refined coal pilot plant at Fort Lewis, Washing-
522
-------
ton. A Kentucky No. 9 (seam) coal was being
processed at the time the sample was obtained.
Chemical Characterization
The chemical composition of the five solid
wastes has been characterized for over 60 chem-
ical constituents including major, minor, trace,
and rare earth elements using the methods de-
scribed by Gluskoter et al.11 The results of these
analyses have been reported previously by Grif-
fin et al.,4 Schuller et al.,14 and Griffin et al.15
Nine elements were generally found in concen-
trations greater than 1,000 mg/kg of the solid
wastes: Al, Ca, Fe, K, Mg, Na, S, Si, and Ti. An
additional group of elements was found in con-
centrations generally between 100 mg/kg and
1,000 mg/kg. These included B, Ba, Ce, Cl, Cr, F,
Mn, Sr, V, Zn, and Zr. Another 20 elements
were found in detectable concentrations less
than 100 mg/kg.
The most noteworthy differences in chemical
composition between the wastes were the high
levels of Ca and Mn in the ashes from the Lurgi
process and the high levels of S and Cl in the
two liquefaction residues. The wastes from the
Rosebud and Kentucky No. 9 coals contained
higher levels of P and F than did the wastes
from the Illinois coals. Trace element composi-
tion was highly variable, and no clear pattern
could be distinguished. Concentrations of Zn
varied over the widest range (13 to 1,500 ppm),
presumably in response to the presence or
absence of ZnS in the feed coals. Correlation
between the chemical characterization of the
wastes from this study and other investigations
is quite difficult. The difficulty arises from the
variability within the feed coals employed and
the process parameters used. Changes in tem-
peratures and pressures affect the fate of con-
stituents and the nature of the various coal con-
version process waste streams.
Mineralogical Characterization
Samples of the five solid wastes were analyzed
by X-ray diffraction, Mb'ssbauer spectroscopy,
scanning electron microscopy, and optical tech-
niques. The minerals identified are listed in
Table 1. Comparison of the mineralogy of the
samples from gasification and liquefaction proc-
esses is instructive from the standpoint of the
mineral transformations occurring during coal
conversion. For example, pyrite (FeS2) is the
dominant form of iron in the feed coals but is not
detected in any of the solid wastes. The pyrite
has been converted to hematite and magnetite
during the Lurgi gasification process. In con-
trast, the pyrite has been converted to pyr-
rhotite and troilite during the liquefaction proc-
esses. Another interesting contrast occurs in
the clay minerals present in the feed coals. They
remain unaltered during the liquefaction proc-
esses but are converted to feldspar and mullite
during the Lurgi gasification process.
Aqueous Solubility
To determine the soluble constituents of the
five solid wastes, large-volume, static leaching
tests were used. This involved making 10 per-
cent (weight to volume) slurries of solid waste
with distilled water in large glass carboys. The
slurries were made in a series of four and ad-
justed to pH values over the range 2 to 11. The
pH values of the slurries were monitored and
readjusted to the specified values when neces-
sary. Chemical equilibrium was assumed when
the pH remained' constant. The period for
achieving equilibrium lasted 3 to 6 mo. How-
ever, studies conducted with the Lurgi ashes in-
dicated that they had reached over 90 percent
of their equilibrium concentrations within 1
week. Duplicate sets of the slurries were made;
one set was equilibrated under an argon
(oxygen- and C02-free), atmosphere and the
other under an air atmosphere.
The leachates from the wastes were analyzed
for 43 constituents, and these concentrations
were compared to recommended water quality
levels (EPA1'). The results of the actual leachate
analyses have been reported previously by Grif-
fin et al.' and Griffin et al.15 Table 2 lists consti-
tuents found to exceed the recommended levels
over the pH range studied and under the labora-
tory conditions described above. Although
many constituents exceed the recommended
levels under acid conditions, those that exceed
the recommended levels over the entire pH
range or at their natural pH were felt to repre-
sent the highest potential for pollution. These
constituents are listed under the "Natural pH"
column in Table 2. Also given in Table 2 are the
pH ranges of the leachates used and the pH
values of the two natural pH solutions for each
aerobic (air) and anaerobic (argon) set of slurries
523
-------
TABLE 1. MINERALS IDENTIFIED IN COAL CONVERSION SOLID WASTES
Minerals
Identified
Lurgi
IL 5
Lurgi
IL 6
Lurgi
Rose-
bud
H-Coal
IL 6
SRC
KY 9
Hematite
(Fe203)
Magnetite
Goethite
(FeOOH)
Pyrrhotite
(Fei_xS)
Troilite
(FeS)
Sphalerite
(ZnS)
Quartz
(Si02)
Gypsum- Anhydrite
(CaSOit)
Calcite-Dolomite
(CaCQ3-CaMg(C03)2)
Wollastonite
(CaSi03)
Plagioclase Feldspar
Na(Ca)AlSi3Os
Mullite
(3Al203'2Si02)
Clay Minerals
X
X
X
X
X
X
X
for each waste. The natural pH slurries are slur-
ries that were allowed to equilibrate without
pH adjustment by addition of either nitric acid
or sodium hydroxide. Table 2 indicates that
there is a strong similarity between the soluble
constituents found in a solid waste and the
treatment it undergoes; i.e., the three Lurgi
ashes yielded nearly the same major soluble
constituents for all three feed coals employed.
The same was true for the two liquefaction
wastes. The Illinois No. 6 coal was used in both
the Lurgi and H-Coal processes but resulted in
the derivation of different soluble constituents
from their wastes. The levels of Cd, K, Mn, Na,
Pb, and Sb found in the Lurgi ash leachates
were higher than those found in the leachates
from the H-Coal and SRC residues under the
conditions employed.
In addition to constituents listed under the
"Natural pH" column in Table 2, Al, Be, Cr, Co,
Cu, F, Fe, Mg, Ni, P, V, and Zn were found in the
leachates at concentration levels exceeding the
recommended levels in water under certain pH
conditions, generally when the pH was acidic.
524
-------
TABLE 2. CONSTITUENTS WITH CONCENTRATIONS EXCEEDING RECOMMENDED
WATER QUALITY LEVELS UNDER THE LABORATORY TEST CONDITIONS
Natural pH
Sample Air Argon
Lurgi Ash 7.6 8.9
Illinois #6
Coal
Lurgi Ash 8.3 10.9
Illinois #5
Coal
Constituents
B, Ca, Cd, K,
Mn, NHi,, Pb,
SOi, , Sb
B, Ca, K, Mn,
NHi», Pb, SOi,,
Sb
Adjusted pH
Additional
constituents
Range leached
8.9-2.7 Al, Cr, Co, Cu,
Fe,' Zn
10.9-3.1 Al, Cd, Cr, Co,
Cu, Fe, Ni, Zn
Lurgi Ash 8.5 11.1
Rosebud Coal
B, Ca, cd, F,
K, Mo, NHi», Pb,
SOi», Sb
11.1-3.1 Al, Cr, Co, Cu,
Fe, Mg, Mn, P,
Zn
SRC-I 6.4 7.5
Kentucky #9
Coal
H-Coal 8.8 11.3
Illinois #6
Coal
B,
NHi,
Al,
Ca , Fe , Mn ,
, SOi,
B, Ca, NHi,
10.2-2.9 Al, Be, Cd
Co, K, Ni,
P, V, Zn
11.3-2.3 F, Fe, Mn,
Zn
, Cr,
Pb,
Pb,
Discharges of the constituents listed in Table 2
at the levels found in this study could cause
some environmental degradation and require
wastewater treatment.
Equilibrium Solubility Model
It is difficult to explain the aqueous chemistry
of a complex system such as the leachates from
coal conversion solid wastes. Possible complexa-
tion, ion pair formation, and the effects of
organic components on the formation of organo-
metallic complexes hinder the description of
these systems. On the other hand, it is still of in-
terest to examine these systems in an effort to
account for their soluble components, and we
progress if we prepare diagrams showing the
relations of the known aqueous species to the
mineral solid phases.
The solubility and mineral stability diagrams
were prepared as described by Garrels and
Christ.17 The thermodynamic solubility model
used in this study (WATEQF) considered the
speciation of 115 aqueous inorganic ions and
complexes and computed saturation data for
over 100 minerals. The theory of the model and
its computer implementation have been dis-
cussed previously by Truesdell and Jones,18-19
and by Plummer, Jones, and Truesdell.20
The stability relations of the iron oxides and
sulfides in water are shown in Figure 1 plotted
as a function of Eh and pH. The data from the
leachates of the five wastes and a pyrite stand-
ard, equilibrated under the same conditions as
the solid wastes, are shown plotted on the dia-
gram.
Some explanation of the diagram may aid in
interpreting the data. The upper and lower
limits of water stability are shown on the
diagram and mark the upper and lower bound-
aries of Eh and pH of concern. That is, at Eh and
pH values above the upper boundary shown,
525
-------
Aih Sup«mittnt
Solutloni Ugtnd
PYRRHOTITE F*S
MAGNETITE
Figure 1. Stability relations of iron oxides
and sulfides hi water at 25° C
when the sum of S - 10~3M
and Fe + 2 = 10~6M
water decomposes into oxygen gas and at the
lower boundary decomposes into hydrogen gas.
Thus, Eh and pH values outside this range are
not normally of concern when the aqueous
chemistry of natural systems is interpreted.
The solid lines between solid phases such as
hematite and magnetite mark the boundaries of
mineral stabilities. Data points falling within
these regions indicate that the samples are
within the stability field* of the particular
mineral. Most of the data points shown in
Figure 1 fall within the hematite stability field.
This is reasonable because hematite was iden-
tified by X-ray diffraction as being present in
most of the samples. However, magnetite and
pyrrhotite were also identified as minerals pre-
sent in the solid wastes. These diagrams il-
lustrate that these two minerals are unstable in
these systems and, given sufficient time, will
decompose to other mineral phases.
Data points that fall on or near a boundary
line, such as the pyrite standard plotted in
Figure 1, illustrate a solution in simultaneous
equilibrium with the various solid phases
described by the boundary. The pyrite used in
this study was a technical grade material that
contained both hematite and magnetite as im-
purities. Thus, it is reasonable that the solution
would be in equilibrium with these three miner-
al phases and serves to illustrate that the elec-
trodes used in the measurements were operat-
ing properly.
The boundaries between solid phases and
aqueous species such as between hematite and
the aqueous Fe+2 ion serve as true "solubility"
boundaries; as such, they are a function of the
activity of the ion in solution. Two boundaries
are shown in Figure 1, one for 10 "6M and
another for 10~2M Fe+2aQ. The 10~6M bound-
ary is chosen by convention, on the premise that
if the activity of an ion in equilibrium with a
solid phase is less than 10~6M, the solid will be
immobile in the particular environment. This
convention was developed largely from experi-
ence but seems to correlate well with natural
geologic systems. The 10~2M boundary was
chosen because it corresponds to the upper limit
of Fe*2 concentrations measured in the
leachates from the solid wastes.
The boundary between two aqueous species
such as Fe+2 and Fe+3 ion is drawn where the
concentration of each ion is equal. Thus, the
labeled areas are areas where the particular ion
dominates but where small concentrations of
other ions may also be present.
The 10~6M boundaries of the metastable
minerals maghemite and freshly precipitated
ferric hydroxide are shown as dashed lines on
the diagram. It is certain that these two miner-
als are unstable with respect to hematite,
pyrite, and magnetite, and that, given sufficient
time, will convert to the thermodynamically
stable minerals. However, these minerals are
clearly of more than transitory existence in
natural environments and warrant considera-
tion as mineral phases likely to control iron con-
centrations during the initial leaching of solid
wastes, which may be the environmentally
critical period.
The data plotted in Figure 1 indicate that
amorphous ferric hydroxide is a likely control
on iron concentrations in the leachates at pH
values less than 7. Indeed, computations of ion
526
-------
activity products for the leachates yield good
agreement with the solubility constant for the
amorphous ferric hydroxide in the acid solu-
tions. The iron concentrations tend to drop
below detectable levels in the alkaline solutions.
It is clear from the plot of the data in Figure 1
that the Eh-pH relations of the alkaline leach-
ates are not controlled by equilibria between
minerals shown on the diagram. Figure 2 shows
the aqueous stability relations of the manga-
nese oxide-carbonate system. The data indicate
that the manganese oxides and carbonate are in
equilibrium in the alkaline leachates, while the
data points for the acid leachates fall in the
aqueous Mn+2 ion field. This conclusion is sup-
ported by the computations of the ion activity
products for the manganese minerals. These
computations showed that the alkaline solutions
were generally in equilibrium with the manga-
nese oxides or carbonate on which boundary the
particular data points shown in the diagram fell.
The acid leachates were undersaturated with
respect to the various manganese minerals, as
deduced from Figure 2. Thus, it appears that
manganese oxides control the Eh-pH relations
+1.0-
+0.6-
0-
-0.6-
-1.0-
Aih SupwiMttnt
Solution* L*g*nd
(AIR)
(ARGON)
(AIR)
(ARGON)
(AIR)
(ARGON)
(AIR)
(ARGON)
(AIR)
(ARGON)
I
2-
3-
4-
5
6
7
SUPERSATURATED
•log Ci activity
Figure 2. Stability relations of manganese
oxides in water at 25° C.
Figure 3. Calcium sulfate equilibria of
leachates from five coal conver-
sion solid wastes.
of the alkaline leachates and metastable freshly
precipitated ferric hydroxide in the acid leach-
ates.
The solubility relations of anhydrite and gyp-
sum are shown in Figure 3. Here, the solubility
of gypsum exerts a dominant influence over cal-
cium and sulfate concentrations in the leachates
at all pH levels, with the exception of the H-Coal
leachates. The H-Goal leachates are all undersat-
urated with respect to gypsum, but gypsum still
provides the upper boundary for prediction of
calcium and sulfate concentrations. This is note-
worthy because the H-Coal residue contained
the highest concentrations of sulfur but had the
lowest water-soluble sulfur levels, including all
sulfur species considered. This underscores the
need for information on mineral forms in the
solid waste in addition to chemical analysis of
the waste.
The calcium and magnesium carbonate equi-
libria of alkaline (pH >7.6) leachates from the
five solid wastes in contact with air are shown
in Figure 4. Calcium concentrations of the acid
leachates were controlled by gypsum equilibria,
but it is expected that calcium concentrations in
alkaline solutions in contact with atmospheric
carbon dioxide would be controlled by calcite
solubility. However, the data plotted in Figure 4
indicate that the solutions are supersaturated
527
-------
2-
3-
f.J
r
it
8
7-
10 -I
11
UNDERSATURATED
•log M" Mtivity
Figure 4. Calcium and magnesium car-
bonate equilibria of alkaline (pH
>7.6) leachates from five coal
conversion solid wastes in con-
tact with air.
• H.SK3,aq
A AT*,
SUPERSATURATED
•morphoui SIO,
0 1 i 3 4 6 6 7 8 9 10 11 12 13 14
Figure 5. Silicon dioxide and aluminum
hydroxide solubility equilibria of
leachates from five coal conver-
sion solid wastes.
with respect to calcite while some samples are
either in equilibrium with or undersaturated
with respect to magnesite. Other workers have
also noted higher solubility of calcite in the
presence of Mg. Hassett and Jurinak21 found
that calcites with low levels of Mg showed an in-
crease in solubility. Likewise, Berner22 showed
incorporation of Mg within the calcite to be con-
siderably more soluble than pure calcite. Fur-
ther, the presence of Mg and S04 have been
shown by Aken and Lagerwerff23 to enhance the
solubility of calcite. Thus, it appears that
magnesite equilibria in alkaline leachates can be
used to predict the upper bounds on Mg concen-
trations but that problems with supersaturation
preclude the use of calcite to make similar pre-
dictions of Ca concentrations.
The silicon dioxide and aluminum hydroxide
solubility equilibria are shown in Figure 5. Most
samples were found to fall within the range of Si
solubilities expected from amorphous glass and
quartz. This is consistent with the experimental
design, which employed glass carboys as the
equilibration vessel and in which quartz was
identified as being present in all the solid
wastes. Amorphous Si02 is clearly not the most
stable phase, and silica concentrations, after
long time periods, would be expected to be con-
trolled by alumino-silicate minerals or quartz.
The Al equilibria in the mid-acid and alkaline
(not shown) pH range were dominated by the
amorphous hydroxide. Similar to the Fe and Si
equilibria, a metastable mineral phase was ap-
parently controlling the solubility. It is clear
that these metastable mineral phases must be
considered when the environmental impact is
predicted during the initial leaching of coal con-
version solid wastes.
The aqueous chemistry of some other poten-
tial contaminants was examined. For example,
computation of ion activity products for BaS04
indicated that Ba concentrations in the leach-
ates would never exceed 0.1 ppm, even in very
acid solutions. Fluoride concentrations in the
leachates were predicted to be controlled by
precipitation of fluorite (CaF2) and fluorapatite
528
-------
(Cas(P04)8F). Phosphate levels in the alkaline
leachates would never exceed 1 ppb; this was
predicted from the ion activity product calcu-
lations for fluorapatite and hydroxyapatite
(Ca6(P04)8OH). In the acid leachates, phosphate
levels are predicted to be controlled by pre-
cipitation of insoluble iron and manganese phos-
phates.
The data from this study strongly suggest
that removal of trace metals such as Cd, Co, Or,
Cu, Ni, Pb, and Zn from slurry pond leachates
may be controlled by adsorption on or copre-
cipitation with iron, manganese, and aluminum
oxides and hydroxides. The removal of trace
metals by this mechanism would be operative
for long time periods because the adsorptive
capacity of the solid plase would be continually
replenished by formation of new metal oxides in
the leachates. In any case, the partitioning be-
tween trace metals and solid phases must be
considered when trace metal mobility is
evaluated in these systems. Further, these
studies show that hydroxide, sulfate, and car-
bonate are the major inorganic ligands that
must be considered.
Thus, application of thermochemical solubili-
ty models to the coal solid waste leachates ex-
amined in this study has yielded valuable in-
sight into the potential pollution hazards of
these wastes. It has shown that, while the con-
centrations of chemical constituents in the solid
wastes and leachates varied over a wide range,
similar mineral phases controlled the aqueous
solubility of many major, minor, and trace ionic
species for all five of the solid wastes.
ACKNOWLEDGMENTS
We gratefully acknowledge the U.S. En-
vironmental Protection Agency, Fuel Process
Branch, Research Triangle Park, North Caro-
lina, for partial support of this work under Con-
tract 68-02-2130, Characterization of Coal and
Coal Residues. We also are indebted to the
Peabody Coal Company, Freeburg, Illinois; to
Hydrocarbon Research, Inc., Trenton, New Jer-
sey; and to Pittsburg and Midway Coal Mining
Co., Fort Lewis, Washington, for supplying us
with samples.
The authors are indebted to G. V. Smith, C. C.
Hinkley, and M. Saporoschenko of Southern
Illinois University, Carbondale, for Mossbauer
Spectroscopic Analysis of the samples. The
authors also wish to thank the Analytical
Chemistry Section of the Illinois State Geo-
logical Survey under the direction of Dr. R. R.
Ruch, and to thank Dr. H. D. Glass and T. M.
Johnson for assistance in portions of this
research.
REFERENCES
1. Cavanaugh, E. C., and W. C. Thomas. En-
vironmental Assessment of Low/ Medium
Btu Gasification: Annual Report U.S. En-
vironmental Protection Agency. Publica-
tion Number EPA-600/7-77-142. Washing-
ton, D.C. 1977.
2. Cavanaugh, E. C., W. E. Corbett, and G. C.
Page. Environmental Assessment Data
Base for Low/Medium Btu Gasification
Technology. U.S. Environmental Protec-
tion Agency. Publication Number EPA-
600/7-77-125a. Washington, D.C. 1977.
3. Somerville, M. H., and J. L. Elder. A Com-
parison of Trace Metal Analyses of North
Dakota Lignite Laboratory Ash with Lurgi
Gasifier Ash and Their Use in Environmen-
tal Analysis. In: Environmental Aspects of
Fuel Conversion Technology III, Ayer, F.
A. (ed.). Research Triangle Park, U.S. En-
vironmental Protection Agency, 1978.
4. Filby, R. H., K. R. Shah, and C. A. Sautter.
Trace elements in the Solvent Refined Coal
Process. In: Environmental Aspects of Fuel
Conversion Technology III, Ayer, F. A.
(ed.). Research Triangle Park, U.S. En-
vironmental Protection Agency, 1978.
5. Sinor, J. E. Evaluation of Background
Data Relating to New Source Performance
Standards for Lurgi Gasification. U.S. En-
vironmental Protection Agency. Publica-
tion Number EPA-600/7-77-057. Washing-
ton, D.C. 1977.
6. Griffin, R. A., R. M. Schuller, J. J. Suloway,
S. J. Russell, W. F. Childers, and N. F.
Shimp. Solubility and Toxicity of Potential
Pollutants in Solid Coal Wastes. In: En-
vironmental Aspects of Fuel Conversion
Technology III, Ayer F. A., (ed.). Research
Triangle Park, U.S. Environmental Protec-
tion Agency, 1978.
7. Ruch, R. R., H. J. Gluskoter, and J. E. Ken-
nedy. Mercury Content of Illinois coals. II-
529
-------
linois State Geological Survey. Environ-
mental Geology Note 43.1971.15 p.
8. Ruch, R. R., H. J. Gluskoter, and N. F.
Shimp. Occurrence and Distribution of
Potentially Volatile Trace Elements in
Coal: An Interim Report. Illinois State
Geological Survey Environmental Geology.
Note 61.1973. 43 p.
9. Ruch, R. R., H. J. Gluskoter, and N. F.
Shimp. Occurrence and Distribution of
Potentially Volatile Trace Elements in
Coal: A Final Report Illinois State Geo-
logical Survey Environmental Geology.
Note 72. 1974. 96 p.
10. Gluskoter, H. J. Mineral Matter and Trace
Elements in Coal. In: Trace Elements in
Fuel (Advances in Chemistry Series 141),
Babu (ed.). American Chem Soc. p. 1-22.
11. Gluskoter, H. J., R. R. Ruch, W. G. Miller,
R. A. Cahill, G. B. Dreher, and J. K. Kuhn.
Trace Elements in Coal: Occurrence and
Distribution. Illinois State Geological
Survey. Circular 499.1977.
12. Braunstein, H. M., E. D. Copenhaver, and
H. A. Pfudere. Environmental Health, and
Control Aspects of Coal Conversion: An In-
formation Overview. Oak Ridge National
Laboratory. ORNL/EIS-94 and 95, Volumes
1 and 2.1977.
13. Parker, Leon C., and Dewey L. Dykstra.
Environmental Assessment Data Base for
Coal Liquefaction Technology. U.S.' Envi-
ronmental Protection Agency. EPA-800/7-
78-184a and 184b. September 1978.
14. Schuller, R. M., J. J. Suloway, R. A. Griffin,
S. J. Russell, and W. F. Childers. Identifi-
cation of Potential Pollutants from Coal
Conversion Wastes. In: Proceedings of the
Society of Mining Engineers—American
Institute of Mining Engineers Annual
Meeting, New Orleans, February 18-22,
1979.
15. Griffin, R. A., R. M. Schuller, J. J. Suloway,
N. F. Shimp, and W. F. Childers. U.S. Envi-
ronmental Protection Agency. Final
Report for Contract 6&O2-2190, Task D.
Research Triangle Park, N.C. 1979.
16. 1972 Water Quality Criteria. U.S. Environ-
mental Protection Agency. R3-73-033.
Washington, D.C.
17. Garrells, R. M., and C. L. Christ. Solutions,
Minerals, and Equilibria, New York, Har-
per and Row, 1965.
18. Truesdell, A. H., and B. F. Jones. WATEQ,
a Computer Program for Calculating Chem-
ical Equilibria on Natural Waters. National
Technical Information Service. P.B. 220464.
1973.
19. Truesdell, A. H., and B. F. Jones. WATEQ,
a Computer Program for Calculating Chem-
ical Equilibria on Natural Waters. U.S.
Geol Survey J. Research. 2:233-248.1974.
20. Plummer, N. L., B. F. Jones, and A. H.
Truesdell. WATEQF-A Fortran IV Ver-
sion of WATEQ, a Computer Program for
Calculating Chemical Equilibrium of
Natural Waters. U.S. Geological Survey
Water-Resources Investigations 76-13.
NTIS No. PB-261-027.1976.
21. Hassett, J. J., and J. J. Jurinak. Effect of
Mg+2 ion on the Solubility of Solid Car-
bonates. Soil Sci Soc Am Proc. 55:403406.
1971.
22. Berner, R. A. The Role of Magnesium in the
Crystal Growth of Calcite and Aragonite
From Sea Water. Geochimica et Cosmo-
chimica Acta. 59:489-504.1975.
23. Akin, G. W., and J. V. Lagerwerff. Calcium
Carbonate Equilibria in Solutions Open to
the Air. II. Enchanced solubility of CaC03
in the presence of Mg+2 and SO"2.
Geochimica et Cosmochimica Acta.
29:253-260.1965.
530
-------
HAZARDOUS WASTE-DEFINITION AND REGULATION
Alan S. Corson, Mathew A. Straus, and David Friedman*
Hazardous Waste Management Division,
U.S. Environmental Protection Agency, Washington, D.C.
Abstract
The Resource Conservation and Recovery Act
of 1976 (RCRA), in amending the Solid Waste
Disposal Act, created a regulatory framework in
which to control waste materials disposal The
Act further requires the U.S. Environmental
Protection Agency (EPA) to set and enforce
standards for managing hazardous wastes. This
paper summarizes the development of the defi-
nition of hazardous waste, based on the require-
ments of RCRA and the other standards man-
dated by RCRA. These regulations, proposed in
the Federal Register on December 18,1978, are
briefly reviewed.
INTRODUCTION
The Resource Conservation and Recovery
Act (RCRA), which substantially amends the
Solid Waste Disposal Act, creates a regulatory
framework in which to control the disposal of
wastes defined as hazardous. Subtitle C of
RCRA requires the U.S. Environmental Protec-
tion Agency (EPA), in consultation with State
governments, to develop national standards to
define hazardous wastes; generators and trans-
porters of hazardous waste; performance, de-
sign, and operating requirements for hazardous
waste treatment, storage, and disposal facili-
ties; a permit system for such facilities; and
guidelines describing conditions under which
State governments will be authorized to carry
out the hazardous waste control program.
This "cradle-to-grave" concept is somewhat
unique but necessary to ensure that wastes re-
quiring special management are handled only at
facilities with proper permits. All stages of the
hazardous waste management cycle are con-
trolled, whether the waste is managed "onsite,"
at the point of generation, or transported to an
"offsite" waste management facility.
The national standards mentioned previously
have been proposed for public comment and are
to be finalized no later than December 31,1979.
RCRA provides that these standards will go
into effect 6 mo after final promulgation, or in
early summer of 1980.
The proposed regulatory strategy uses a
pathways approach wherein the path and desti-
nation of any hazardous waste are controlled
without particular attention to the source of the
waste. This approach is basically different from
the approach used to regulate air and water pol-
lution, where specific standards are written for
and tailored to each industrial category. The
pathways approach was chosen because hazard-
ous wastes are mobile and can be disposed of at
locations far from. the generating sources,
whereas industrial, air and water pollution
sources are fixed and relatively easy to identify.
I will briefly review the regulations within
the proposed hazardous waste program and pro-
vide additional detail on the proposed definition
of hazardous waste.
HAZARDOUS WASTE DEFINITION
RCRA requires hazardous waste to be de-
fined by inherent characteristics (e.g., flamma-
bility and corrosiveness) and by listing of partic-
ular hazardous wastes.
HAZARDOUS WASTE GENERATORS
The proposed standards for hazardous waste
generators require recordkeeping, annual re-
ports, proper containing and labeling of hazard-
ous waste shipped offsite for disposal, and a
transport manifest document for each shipment.
Retailers, farmers, and generators of small
amounts of waste Qess than 100 kg/mo) are ex-
cluded from these requirements provided they
dispose of waste in State-approved facilities.
Generators do not need permits.
•Speaker.
581
-------
HAZARDOUS WASTE TRANSPORTERS
Hazardous waste transporters are required
to take the hazardous waste shipments only to
the permitted facility designated by the gener-
ator, to keep appropriate records, and to report
spills enroute. Transporters (as in the case with
generators) do not need permits in the Federal
system, but some States require hazardous
waste transporters to be registered.
HAZARDOUS WASTE FACILITY
STANDARDS AND PERMITS
National standards for hazardous waste
treatment, storage, and disposal facilities not
only establish acceptable levels of performance
that such facilities must achieve but also are the
criteria against which regulatory officials will
measure applications for permits. In setting fa-
cility standards, EPA has relied primarily on
specific design and operating standards (as op-
posed to general ambient or source emission
standards) because they are more easily under-
stood and enforced than other types of stand-
ards.
STATE HAZARDOUS WASTE PROGRAMS
Congress intended that EPA establish na-
tional standards for hazardous waste manage-
ment but that the individual States implement
and enforce this new regulatory program. EPA
has developed a guideline that describes the
elements a State hazardous waste program
must have in order for that State to have
authority to carry out the national program.
Among other things, States must have legisla-
tion and regulations for hazardous waste man-
agement that are no less stringent than in the
Federal analogs and must demonstrate that
they have adequate resources to administer and
enforce the program.
DEVELOPMENT OF THE PROPOSED
DEFINITION
I would like to highlight the development of
the definition of hazardous waste in the Decem-
ber 18,1978 Federal Register. Before a material
can be defined as a hazardous waste, it must
first be established that the material is a solid
waste. RCRA defines "solid waste" as "any gar-
bage, refuse, sludge from a waste treatment
plant, water supply treatment plant, or air pol-
lution control facility and other discarded mate-
rial, including solid, liquid, aemisolid, or con-
tained gaseous material resulting from indus-
trial, commercial, mining, and agricultural oper-
ations and from community activities. The term
does not include solid or dissolved material in
domestic sewage or solid or dissolved materials
in irrigation return flows or industrial dis-
charges that are point sources subject to per-
mits under Section 402 of the Federal Water
Pollution Control Act, as amended, or source,
special nuclear, or byproduct material as de-
fined by the Atomic Energy Act of 1964; as
amended." There are three noteworthy aspects
of a solid waste definition. The term encom-
passes not only solids, but liquids, semisolids,
and contained gases; it explicitly excludes cer-
tain materials; and it includes "other discarded
material." EPA has grappled with the meaning
of "other discarded material" for over a year be-
cause it is one of the more ambiguous yet impor-
tant parts of the definition. For example, are
byproducts of manufacturing processes "dis-
carded materials"? Sometimes they are, and
sometimes they aren't. Are materials sent to
recycling or reprocessing centers "discarded
materials"?
After substantial discussion and comment in-
side and outside the Agency, EPA has judged
this phrase to mean any material that is aban-
doned or committed to final disposition; reused,
if such use constitutes land disposal; and a
waste oil, if it is incinerated or burned as a fuel.
Under this definition, for example, used sol-
vents sent to a reclaiming facility would not be
considered a discarded material and, therefore,
would not be considered a solid or a hazardous
waste. Similarly, materials being transferred
between industrial facilities, perhaps via a
waste exchange, would not be subject to hazard-
ous waste controls. On the other hand, materials
reused in a way involving land application (i.e.,
soil conditioners, fill materials, dust sup-
pressants, etc.) would be considered discarded
materials because reuse of materials in this
manner could result in serious adverse impacts
from uncontrolled release and dispersion of con-
taminants into the environment. Similarly, EPA
has singled out waste oils for special control
532
-------
because they are ubiquitous and there are docu-
mented health and environmental problems as-
sociated with their reuse.
Criteria of Identification and Listing
In defining a hazardous waste as mandated in
Section 3001 of RCRA, EPA is required to de-
velop and promulgate criteria for identifying
the characteristics of hazardous waste and for
listing hazardous waste, and to identify the
characteristic of hazardous waste and list par-
ticular hazardous wastes. As a first step in this
definition process, EPA has developed a set of
criteria in defining the characteristics of a
hazardous waste and for listing these wastes.
These criteria are identified in Section 250.12 of
the proposed rule and are:
• Criteria for Identifying Characteristics of a
Hazardous Waste
• Damage cases: certain wastes are known
to have caused substantial public health
or environmental damage in documented
cases.
• Availability of economical sampling and
analysis procedures for a particular prop-
erty of the waste.
• Criteria for Listing Hazardous Wastes
• The waste is known to meet, or strongly
suspected of meeting, one of the defined
general characteristics.
• The waste meets the statutory definition
of a hazardous waste.
Based on these criteria, EPA has elected to
define the general characteristics of ignitability,
corrosivity, reactivity, and certain aspects of
toxicity to identify hazardous wastes. It should
be noted that EPA also attempted to define
characteristics of infectious and radioactive
waste and other aspects of toxicity such as
genetic change potential and bioaccumulation.
However, in developing this regulation, EPA
encountered difficulty in describing these prop-
erties and elected to deal with potentially infec-
tious, radioactive, and certain toxic wastes by
listing known sources of these wastes or proc-
esses likely to produce them. EPA does intend
to explore the appropriateness of additional
characteristics to further define toxicity and
radioactivity. To this end, it has published an
advanced notice of proposed rulemaking seek-
ing additional data related to these concepts. It
should also be emphasized that neither the char-
acteristics nor the listing is static. Both may be
added to or changed, after opportunity for
public comment, as new information develops.
Hazardous Waste Characteristics
In order to provide specific descriptions of
wastes meeting these characteristics, each char-
acteristic was defined in terms of specific defin-
able properties. The following is a brief descrip-
tion of each characteristic and its properties.
• Ignitability. The objective of the ignitabil-
ity characteristic is to identify wastes that
may present a fire hazard under routine
waste disposal and storage conditions. The
resulting fires at disposal and storage facili-
ties present not only the immediate danger
of heat and smoke but can initiate explo-
sions, generate toxic vapors, and provide a
pathway by which toxic particulates can
spread to the surrounding area. (The term
ignitable was chosen to avoid confusion with
the U.S. Department of Transportation's
(DOT) category of "flammable" in its hazard-
ous materials transportation regulations).
Several methods can be used to identify
ignitable wastes, depending on the physical
state. For liquid wastes, flash point was se-
lected as the property to use because testing
methods are available and are the most re-
producible. The flash point proposed for
identifying ignitable wastes is 140° F
(60° G); this value was selected after consid-
eration of ambient temperatures to which
wastes may be exposed during manage-
ment.
For solid wastes, a prose definition was
selected because test methods are not avail-
able for ignitable solids that simulate the
field conditions to which a waste is subject
during handling and management. For
waste gases, EPA proposes to use the DOT
identification for flammable compressed
gases because the major hazard from ignit-
able gases would arise during transport.
• Corrosivity. A corrosivity characteristic
has been included to identify wastes that
must be segregated from others because of
ability to extract and solubilize toxic con-
taminants (especially heavy metals) that
might otherwise not migrate, and to identify
wastes requiring special containers during
transportation and storage.
538
-------
While heavy metal solubilization is an ex-
tremely complex phenomenon, pH has been
found to be its most important indicator.
The pH limits chosen in these proposed reg-
ulations were based upon skin corrosion
limits and heavy metal solubilization data.
The metal corrosion limits were taken from
DOT hazardous materials regulations be-
cause EPA's concern about container dam-
age is identical to that of DOT's in this case.
• Reactivity. The object of the reactive waste
characteristic is to identify wastes that
under routine management present a hazard
because of instability or extreme reactivity.
Reactivity includes the tendency to auto-
polymerize, to create a vigorous reaction
with air or water, to exhibit shock and ther-
mal instability, to generate toxic gases, and
to explode.
In their proposed regulation, EPA in-
cluded a descriptive definition of a reactive
waste, together with test methods for ther-
mal and shock instability, because of the
problem in developing general test methods
for identifying reactive wastes. While there
are many inputs of energy that may cause a
waste to react or exhibit hazardous proper-
ties, there is no one stress that can cause all
reactive waste to do so. To compound the
problem, reactivity is not just a function of
the composition, temperature, and availabil-
ity of initiating agents. It is also affected by
the mass and geometry of the waste. Thus,
the reactivity of a tested waste sample may
not necessarily correspond to the reactivity
of the waste as a whole.
Because reactive waste is dangerous to
the generator's own operations (as well as
being hazardous for long-term disposal), gen-
erators of reactive waste tend to be aware
that their waste has that characteristic. For
this reason, EPA considers the proposed de-
scriptive definition an adequate identifica-
tion method when it is used in conjunction
with the test methods identifying thermal
and shock instability.
• Toxicity. The toxicity characteristic is in-
tended to identify waste which, if improper-
ly disposed of, may release toxicants in suffi-
cient quantity to pose a substantial hazard
to human health or the environment. The
RCRA definition of hazardous waste re-
quires EPA to judge the hazard posed by a
waste "when improperly treated, stored,
transported, or disposed of, or otherwise
managed." For waste containing toxic con-
stituents, the hazard depends on two fac-
tors: the intrinsic hazard of the constituents
of the waste, and the release of the constit-
uents to the environment under conditions
of improper management.
To assess the intrinsic hazard posed by
the constituents, a series of toxicity indi-
cators were initially considered: acute and
chronic toxicity to humans, animals, and
plants; potential for bioaccumulation in tis-
sue; oncogenicity; mutagenicity; and terato-
genicity.
However, the toxicity definition proposed
December 18, 1978, has been limited as noted
earlier to include only toxicants for which Na-
tional Interim Primary Drinking Water Stand-
ards (NIPDWS) have been developed.
To determine whether toxic constituents in
the waste might migrate in the disposal envi-
ronment, a procedure has been developed to
measure the tendency of the constituents of a
waste to leak or leach out and become available
to the environment under poor management
conditions.
Numerous studies and reports indicate that
damage to ground- and surface water frequently
results from migration of toxic chemicals from a
disposal site. Groundwater contamination is a
particularly important concern because ground-
water provides drinking water to almost one-
half of the population. In addition, once con-
taminated, an aquifer's usefulness as a source of
drinking water may be impaired for years. It
was thus decided that use of a groundwater con-
tamination scenario to "model" improper dis-
posal would be advisable. By selecting a ground-
water contamination scenario, we did not mean
to imply that other vectors are not important.
However, we do feel that except in rare cases,
control levels set using this model will be suffi-
cient to protect against other routes of con-
tamination.
The model is based on wastes creating a prob-
lem through migration of chemicals out of the
disposal site and into a drinking water aquifer. I
want to emphasize that the contamination mod-
el has been developed for definitional purposes
only. It does not address particular disposal
534
-------
methods that might be used by the regulated
community.
The test scheme commonly referred to as the
extraction procedure (EP) has been devised to
meet the limited definition of toxic waste. The
EP coupled with a model scenario of leachate
transport related the concentrations of certain
toxic components found in the extract of the
waste to the EPA NIPDWS. Any waste whose
EP extract contains heavy metals or pesticides
controlled by the NIPDWS in a concentration
greater than 10 times the drinking water stand-
ard is considered a hazardous waste.
A waste that has any of the above character-
istics is a hazardous waste by RCRA definition
whether or not that waste is listed. Conse-
quently, use of characteristics in the hazardous
waste definition implies responsibility on the
part of waste generators to evaluate their
wastes for these characteristics (or to declare
their wastes hazardous) if there is any doubt
about the status of their waste.
Hazardous Waste Listings
The second way a waste can be brought into
the hazardous waste regulatory program is by
including that waste on a list. Actually, EPA
has developed four separate hazardous waste
lists including:
• A list of generic hazardous wastes common
to many different sources (i.e., electroplat-
ing wastes, paint wastes, etc.);
• A list of known sources of infectious wastes,
such as hospital wastes from the labora-
tories;
• A list of industrial processes known to pro-
duce hazardous waste, such as heavy ends or
distillation residues from carbon tetrachlor-
ide fractionation; and
• A list of some 275 substances, which, if dis-
posed of in pure form or as a result of off-
specification production, would be hazard-
ous.
There are approximately 175 specific wastes,
waste sources, and wastes from certain proc-
esses that EPA has identified as hazardous
based on previous studies of industrial wastes,
damage cases, testing of wastes, and State haz-
ardous waste program data.
There may be cases, however, where a partic-
ular facility within a listed source or process
category believes that its waste is nonhazard-
ous because the facility uses different raw mate-
rials than normal, or has made process modifica-
tions or provides onsite treatment prior to dis-
position. In such cases, the individual facility
can petition for exemption from the Subtitle C
control program by submitting appropriate
waste-testing data and requesting a determina-
tion of noncoverage of Subtitle C for the facili-
ties' waste.
Summary
In summary, EPA is required to define haz-
ardous wastes using dual approaches of identi-
fying general characteristics and listing specific
hazardous wastes. As regulation development
evolved, the Agency found it necessary to defer
proposing certain characteristics considered
earlier pending further study. At the same time,
EPA has added to and sharpened the focus of
the hazardous waste list. We believe the net
result of these changes will make it much easier
for waste handlers to determine whether they
are in or out of the Subtitle C regulatory pro-
gram, and at the same time, focus the program
on those wastes of most concern.
535
-------
FACTORS CONSIDERED IN EFFLUENT LIMITATIONS
GUIDELINES DEVELOPMENT
John W. Lum
Effluent Guidelines Division,
U.S. Environmental Protection Agency, Washington, D.C.
Abstract
In the area of coal conversion, studies con-
ducted by the U.S. Environmental Protection
Agency (EPA) have been directed primarily
through its Office of Research and Development
in Research Triangle Park. Information gener-
ated has proven useful in regulation writing.
Most of the past studies were conducted in the
laboratory. Onsite studies to evaluate waste-
water characteristics and wastewater treatment
technology applicability are necessary and en-
couraged. In addition to waste characterization
and technology assessment, cost, industry pro-
file, water quality criteria, nonwater quality-
related impacts, and other factors must be con-
sidered in regulation development.
The Federal Water Pollution Control Act (as
amended in 1977) required the establishment of
technology-based and industry-specific effluent
limitations guidelines for point source dis-
chargers. The Effluent Guidelines Division
(EGD) of the U.S. Environmental Protection
Agency (EPA) has been delegated that responsi-
bility. The Act does not specify the date when
guidelines must be promulgated for coal conver-
sion facilities.
In the area of coal conversion, activities by
the Effluent Guidelines Division have been min-
imal. EGD has been relying on its Office of
Research and Development to gather funda-
mental information. When EGD actively pur-
sues guideline development, it is unlikely that
all conversion processes will be addressed at
the same time. The low-Btu gasification proc-
esses will probably be the first group to be con-
sidered.
Because EGD does not have any active pro-
gram in this area, the only subject I can address
at this time is the type of information that EGD
will consider in its regulation development. The
task of developing information to support tech-
nology-based effluent limitations guidelines can
be divided into five discrete elements:
• Industry profile,
• Wastewater characterization,
• Selection of candidate treatment technol-
ogies,
• Cost analysis of candidate treatment op-
tions, and
• Subcategory review.
Information on the industry's current and
projected distribution is important because the
Agency must consider environmental impact
and national cost implication in writing the
guidelines. The type of data needed to define
the industry includes:
• Description of the process,
• Number and size of existing and projected
facilities using each of the processes,
• Geographical location and the type of coal
used,
• Economics of this industry and its competi-
tive industries,
• Current or anticipated regulations the in-
dustry is or will be subjected to, and
• Stage of process development.
Reports prepared by EPA's Office of Research
and Development have provided much of the
needed information.
Data on wastewater characteristics are nec-
essary to assess the degree of environmental
impact and applicability of treatment tech-
nology. The Agency must define the quality and
quantity of pollutants from aqueous effluents.
The effluents of concern include those which are
process- and nonprocess-related. Normally, EGD
conducts sampling and analytical studies at full-
scale facilities whenever possible. Smaller units
can be sampled if full-scale facilities are not
available or accessible. In the past, EGD has
successfully cooperated with the regulated in-
dustries and anticipates that this industry will
be just as cooperative. The FWPCAA (Section
308) authorizes the Agency to obtain informa-
537
-------
tion necessary for regulation writing.
Analytical data of wastewater characteristics
from bench-scale .operation can be used in pro-
viding "order of magnitude" estimates of the
potential problem and determining the appli-
cability of wastewater treatment technologies.
The Agency is required to assess the discharge
of the 129 toxics substances as well as conven-
tional and nonconventional pollutants. The ana-
lytical method used must be able to quantify the
pollutants at parts-per-billion level. A lower
detection level is required because the water
quality criteria (proposed) for some of the
pollutants are quite low. Studies have been con-
ducted by EPA's Office of Research and Devel-
opment on wastewater characteristics. These
studies quantify the concentration of pollutants
that are present at the 1-mg/L level and above.
Some of the latter studies attempted to quantify
the 129 priority pollutants to lower levels. The
streams analyzed are primarily from bench-
scale process operations. Despite some of the
excellent studies conducted by ORD, an addi-
tional data base will be required prior to regula-
tion writing.
Once the pollutants discharged are defined,
the Agency must evaluate technology available
to reduce the level of discharge. The Agency
can require both end-of-pipe treatment and in-
plant water use modification. The first technol-
ogy option to be considered will be complete
water recirculation and reuse. Other technology
options such as end-of-pipe treatment without
recirculation and best management practice re-
quirements will also be considered. Alternative
technologies must be evaluated during guide-
lines development in terms of cost, energy con-
sumption, water requirement, water quality
criteria, effluent quality, pollutant reduction,
and impact on air and solid media.
Currently, ORD is conducting wastewater
treatment assessment programs on a bench-
scale basis. Once the information from these and
other programs becomes available, applicable
technology options may be determined. These
options should be tested in the pilot-plant scale
whenever possible.
Cost strongly affects the selection of treat-
ment options. An economic impact assessment
will be performed to determine whether the
cost would make the process economically un-
feasible and the extent to which production cost
would be increased.
Effluent limitation guidelines are national
regulations. This does not mean that the ef-
fluent limitations will be the same for all the
coal conversion facilities. The Agency can pro-
mulgate different guidelines for plants with cer-
tain unique features (subcategorization). The
justification for subcategorization can be waste
characteristic, land availability, cost, and treat-
ment technology applicability.
In summary, EGD must consider various fac-
tors in writing regulations. The work that has
been done to date is useful, but more studies are
needed to generate the information necessary
for regulation development. We may be at the
stage (at least for some of the coal conversion
processes) for EPA to interact directly with
DOE and perform studies at the site where the
processes are being developed. Use of informa-
tion that represents real situations would be
beneficial to both the regulator and the
regulated.
538
-------
WATER REQUIREMENTS FOR SYNTHETIC FUELS PLANTS
Harris Gold* and David J. Goldstein
Water Purification Associates, Cambridge, Massachusetts
Abstract
The water requirements for several synthetic
fuel technologies have been estimated at given
locations in the major coal- and oil shale-bearing
regions of the United States, The synthetic fuel
technologies examined include: coal gasification
to convert coal to pipeline gas (Lurgi, Synthane,
Hygas, and Bigas processes); coal liquefaction
to convert coal to low-sulfur fuel oil (Synthoil
process); coal refining to produce deashed low-
sulfur solvent-refined (clean coal) (solvent re-
fined coal process); and oil shale retorting to pro-
duce synthetic crude (Parana Direct, Paraho In-
direct, and TOSCO II processes). A total of 90
plant-site combinations were studied in the Ap-
palachian, Illinois, Powder River, Fort Union,
and Four Corners coal regions and in the Green
River oil shale region. Results are presented for
the total net water consumed by conversion proc-
ess for each coal and oil shale region.
Particular emphasis has been placed on deter-
mining the water consumed for cooling because
cooling is often the largest consumer of water in
a conversion plant. Three cooling options were
considered representing different degrees of wet
evaporative cooling for turbine condensers and
gas compressor interstage coolers. The cost and
availability of water determines the degree to
which wet or evaporative cooling should be used.
Estimates have been made of the cost of trans-
porting water from different sources to the con-
version plant in the Western States to deter-
mine the most suitable coating option. The cri-
terion of water availability was used to deter-
mine the most suitable cooling options in the
Eastern and Central States.
INTRODUCTION
Since the 1973 oil embargo, there has been
considerable debate in the United States con-
cerning the role coal or oil shale will play in solv-
ing the energy problem. There is one important
'Speaker.
issue on which all sides agree, and that is the im-
portance of water in the production of synthetic
fuels. Converting coal or shale to a synthetic
fuel is basically a hydrogenation process in
which water is the source of the hydrogen. The
weight ratio of carbon to hydrogen is higher for
the raw material than for the gaseous or liquid
synthetic fuel. In the conversion, sulfur and
nitrogen are reduced to produce a cleaner fuel,
and ash, oxygen, and nitrogen are reduced to
produce a product with a higher heating value
than the original coal or shale. Water is re-
quired to dissipate to the atmosphere the ther-
mal energy not recovered in the process of con-
verting the coal or shale to the synthetic fuel, to
mine and prepare the raw material, and to dis-
pose of the spent ash and shale and any other
unwanted constituents removed in conversion.
Water is also required for fugitive dust control
within the plant, for sanitary and potable water
usage in the mine and the plant, and for recla-
mation of the disturbed land.
This paper deals specifically with water re-
quirements for integrated mine-plant designs
for manufacturing gaseous, liquid, and solid syn-
thetic fuels from coal and shale and how these
requirements are affected by the local cost and
availability of water. The work is based on a
report recently completed for the U.S. Depart-
ment of Energy (DOE Contract No. EX-76-C-01-
2445) and the U.S. Environmental Protection
Agency (EPA Contract No. 68-03-2207).1 The
range of water requirements for each conver-
sion process—with no distinction made between
coal- and oil shale-bearing region —is summa-
rized in Reference 2. In the present paper, par-
ticular emphasis is placed on determining the
water consumed for cooling because cooling is
often the largest consumer of water in a conver-
sion plant. The results will be summarized by
conversion process and by coal- and oil shale-
bearing region.
PROCESS AND SITE SELECTION
The synthetic fuel technologies examined in-
539
-------
elude: coal gasification to convert coal to pipe-
line gas; coal liquefaction to convert coal to low-
sulfur fuel oil; coal refining to produce a de-
ashed, low-sulfur solvent-refined (clean) coal;
and oil shale retorting to produce synthetic
crude. A number of processes were chosen for
each conversion. Detailed conceptual designs
for integrated mine-plant complexes were made
for each of the representative conversion proc-
esses.1 The processes and products chosen for
comparison are shown in Table 1. Except for the
commercially available Lurgi process, the proc-
esses chosen are representative of those that
have undergone extensive development and that
are sufficiently described in the available litera-
ture so detailed process calculations can be
made. The products chosen are synthetic fuels;
the production of chemicals from coal or shale
(e.g., ammonia or methanol) was not considered.
The specific designs given in Reference 1 are
based on standard-sized plants with the given
product output. A number of processes were
chosen for each conversion. Detailed conceptual
designs for integrated mine-plant complexes
were made for each of the representative con-
version processes.1 The processes and products
chosen for comparison are shown in Table 1. Ex-
cept for the commercially available Lurgi proc-
ess, the processes chosen are representative of
those that have undergone extensive develop-
ment and that are sufficiently described in the
available literature so detailed process calcula-
tions can be made. The products chosen are syn-
thetic fuels; the production of chemicals from
coal or shale (e.g., ammonia or methanol) was
not considered. The specific designs given in
Reference 1 are based on standard-sized plants
with the given product output.
A large number of site and process criteria
combinations were studied to obtain meaningful
assessments on a regional and national level
from detailed local results. For coal conversion,
the process criteria have been defined based on
the quality of the foul condensate recovered
after gasification or liquefaction. Low-
temperature gasifiers (e.g., Lurgi and Synthane)
give a very dirty process condensate (typical
values for bituminous coals: BOD - 10,000 mg/L,
phenol ~ 3,000 mg/L, and ammonia ~ 4,500
mg/L), while high-temperature gasifiers (e.g.,
TABLE 1. PRODUCT FUEL OUTPUT OF STANDARD-SIZED SYNTHETIC FUEL PLANTS
Technology and
Conversion Process
Coal Gasification
Lurgi
Synthane
Hygas
Bigas
Coal Liquefaction
Synthoil
Coal Refining
SRC
Oil Shale
Paraho Direct
Paraho Indirect
TOSCO II
Product
Output
Pipeline Gas 250x10 scf/day
Product
Heating Value
(1011 Btu/day)
2.4
Fuel Oil
50,000 barrels/day 3.1
Solvent Refined 10,000 tons/day
Coal
3.2
Synthetic Crude 50,000 barrels/day 2.9
540
-------
Koppers-Totzek and Bigas), give a relatively
clean condensate (typical values: ammonia ~
4,500 mg/L, BOD and phenol ~ small). The Hy-
gas gasifier, which is an intermediate tempera-
ture gasifier, gives a process condensate of in-
termediate quality. Both the solvent refined
coal (SRC) and Synthoil processes have the
foulest condensates. For oil shale conversion,
the degree of water management depends on
the type of retort used. For direct-heated
retorting processes (e.g., Paraho Direct) most of
the water is recovered. However, for indirect-
heated processes, (e.g., Paraho Indirect and
TOSCO II) the water in the combustion prod-
ucts is generally lost up the furnace stack and
not recovered.
As for site criteria, brackish groundwater
would have to considered an important conjunc-
tive supply to surface waters in the West, while
surface waters are considered primarily in the
East. Eastern and Central States have humid
climates, while climates in the West are arid
and semiarid. Eastern and central coals are both
underground- and surface-mined, while western
coals are primarily surface-mined. In the West,
underground mining followed by surface retort-
ing of oil shale has been investigated extensive-
ly. In-situ retorting was not considered in the
present study because it is still under devel-
opment and cannot yet be considered commer-
cial, although it could drastically reduce the
water consumption.
Site selection was based primarily on the
availability of coal and oil shale, the rank of coal
or oil shale, the type of mining (underground or
surface) and the availability of surface water
and groundwater. The coal mining regions cho-
sen were those where the largest and most easi-
ly mined deposits are located. In the West,
these include the Powder River and Fort Union
regions in Montana, Wyoming, and North
Dakota, and the Four Corners region in New
Mexico. In the Central and Eastern regions, the
Illinois and Appalachian coal basins were
selected. Western coals are principally low-
sulfur subbituminous and lignite, while eastern
and central coals are mainly high-sulfur bitum-
inous. Only high-grade shale from the Green
River Formation was considered. Specific
design examples were restricted to shales with
yields of about 30 to 35 gal per ton (0.13 to 0.15
m3/metric ton), as might be found in Colorado or
Utah. A total of 90 plant-site combinations are
listed in Table 2 for the Eastern and Central
States and in Table 3 for the Western States.
The locations of these sites with respect to the
major energy reserves and the primary water
resources characteristics are shown in Figures
1 and 2. The maps show more sites than the ones
given in the tables. Primary sites correspond to
sites listed in Tables 2 and 3, and secondary
sites were selected to provide a larger study
area for water availability.
WATER REQUIREMENTS
Estimates of water consumption are net; all
effluent streams are assumed to be recycled or
reused within the-mine or plant after necessary
treatment. These streams include the organical-
ly contaminated waters generated in the con-
version process, which are unfit for disposal
without treatment, and the highly saline water
blown down from evaporative cooling systems.
Water is only released to evaporation ponds as
a method of salt disposal. These wastes may
also be disposed of with the coal ash if the prob-
lems of runoff and groundwater contamination
are adequately handled in an economic manner.
The rest of the water consumed leaves the plant
as vapor, as bonded hydrogen after hydrogena-
tion, or as occluded water in the solid residues.
Dirty water is cleaned, but only for reuse and
not for return to a receiving water.
Conversion can never be fully efficient in any
real process. All of the available energy of the
coal or shale cannot be fully recovered in the
synthetic fuel, and the unrecovered thermal
energy must be dissipated to the atmosphere.
Some of the unrecovered heat is lost directly to
the atmosphere; e.g., in hot flue gases and in
coal drying. The remainder of the unrecovered
heat is dissipated either through wet cooling or
dry cooling, depending on economic considera-
tions. In general, the quantity of water evap-
orated in cooling is the prime determinant to
the total quantity of water consumed in a plant.
There are four principal types of cooling loads
in any synthetic fuel plant: process streams, gas
purification, turbine condensers, and gas com-
pressors. As shown,3 the most economical proce-
dure for process streams is to cool them to
about 130° F to 140° F with an air cooler and to
cool below these temperatures by using a wet
system. The acid'gas removal regeneration con-
denser can be economically dry cooled at all
541
-------
TABLE 2. PLANT-SITE COMBINATIONS FOR EASTERN AND CENTRAL STATES
State
Alabama
Illinois
Indiana
Kentucky
Ohio
Pennsylvania
Heat Virginia
County
Jeffereon
Marengo
Bureau
Shelby
St. Clair
White
Bureau
Pulton
St. Clair
Saline
Gibson
Wgo
Sullivan
Harrlck
Floyd
Bar Ian
Nuhlenberg
Pike
Gallla
Tuecarawaa
Jefferson
Armstrong
Somerset
Fayette
Kanavha
Nonongalia
Preston
Hiago
Surface Ground
Alabama ft.
Tomblgbee ft. X
X
Ohio ft.
Ohio ft.
Ohio ft.
Illinois ft.
X
Ohio *.
Ohio ft.
White ft.
White ft.
Ohio ft.
Ohio ft.
Ohio ft.
Ohio ft.
Green ft.
Ohio ft.
Ohio ft.
MuskinguB ft. X
Ohio ft.
Allegheny ft.
Allegheny ft.
Kana«ha ft.
Kanavha ft.
Allegheny ft.
Xanateia «.
KanaieHa ft.
a b
Mining Coal
U
S
U
0
u
u
S
S
S
S
u
a
S
S
0
o
S
S
a
D
S
o
D
0
O
0
U
S
Coal Casifii
•igh TBBp.Gaslfier
flygai Blgaa
X
X
X
X
X
X
X
X
X
X
X
X
X
X
ration
fcov Temp.Gasifier
Uirgi Synthane
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Coal Liquefaction
and Coal meflnlng
Synthoil SMC
X
X
X
X
X
X
X
.
X
X
X
X
X
X
X
Plant-Site Combinations
No. Total State
j
« *
3
1
1
1
1
2
1
1 11
3
1
1
2 7
1
1
1
1 4
1
4
3 a
2
1 3
1
1
1
1
2 C
TOTAL 4S
a O -
b » -
i » - Surface.
I L - Lignite
-------
TABLE 3. COAL AND OIL SHALE CONVERSION PLANT-SITE COMBINATIONS
FOR WESTERN STATES
State
Montana
New
Mexico
North
Dakota
Nyoalng
NllM
Dacker-Dletz
Foater Cre«k
U.S. Steel Chupp Mine
Eaat Hoorhead
Pumpkin Creek
Otter Creek
Colatrip
Coalridge
Gallup
El Paao
Heaco
Scran ton
Bentley
Underwood
Knife River
Center
Slope
Dickinaon
Hilliaton
Belle Ayr
Glllettc-ttyodak
Spotted Horse Strip
Hanna
Antelope Creek Mine
Lake-de-sawt
Kemawrer
Jin Brldoer
Rainbow 18
Water Source
Surface Ground
X
Tongue R.
Yellow* tone R.
Powder R.
Tongue R.
X
Tel Iowa tone R.
Missouri River
X
San Juan R.
San Juan R.
Grand R.
Knife R.
L. Sakakawea
Knife R.
Knife R.
Yellowstone R.
L. Sakakawaa
Missouri R.
Crazy woman Cr.
Crazy Woman Cr.
Powder R.
Medicine Bow
Beaver Cr. x
Tongue R.
HIM Fork
Green R.
Green R.
a °
Mining Coal
S S
S S
S L
S L
S L
S L
S S
S L
S S
S S
S S
S L
S L
S L
S E
S L
S L
S L
S L
S S
S S
S S
S S
S S
S S
S B
S S
a B
Coal Gasification
High Temp.Gaaifier Low Tnsp.Gaaifier
Hygaa Bigaa Lurgi Synthane
X X
X
X
X
X X
1
X X
X X
X
X
X
X
X
X
X
X
X
X
X X
X X
Coal Liquefaction
and Coal Refining
Synthoil SRC
X
X
X
X
X
X
X
X
X
X
X
X
X
Plant-Sit* Combination*
No. Total State
2
1
1
1
1
1
3
1 11
3
2
1 6
1
1
1
1
1
1
1
i a
i
2
1
1
a
i
2
2
1 14
Ol
£>•
O3
Tom*
State
Colorado
Mine
Parachute Creek
Water Source
Surface) Ground
Colorado R.
a e
Mining Shale
0 HC
Direct Retort
Paraho Direct
X
Indirect Retort
Paraho Indirect TOSCO II
X X
Plant-Site Conblnatiom
No. Total state
3 3
O - Underground i S - Surface
B - Bltuminouai L - Lignite i S
HG - High grade shale
Subbittsalnoua
TOTAL
-------
SITE LOCATIONS
RANDOLPH
/y
0
-------
WATER AVAILABILITY
mmmmm inadequate
marginal
adequate
SITE UDCATIONS
H PRIMARY SITES
0 SECONDARY SfTES
-v
KENTUCKY
ILLINOIS COALREGION
Figure 1 (continued)
545
-------
NORTH DAKOT;
COALRIDGE
MISSOURIRIVE
UPPER
MISSOURI
*™*.B«*c R/VER BAS|N
WYOMING
COLORADO
• TRACT C^a
LOW DEVELOPMENT
COST OF WATER
(S/1,000 GALLONS)
SITE LOCATIONS
NEW MEXICO
UPPER COLORADO ^s
RIVER BASIN
O PRIMARY SITES
• SECONDARY SITES
Figure 2. Site locations and the cost of transporting water In Western States.
546
-------
plants when the hot potassium carbonate proc-
ess is used and 90 percent dry-10 percent wet
cooled when a physical solvent process is used.8
The gas purification system chosen by the ori-
ginal designers and assigned to each process is
somewhat arbitrary and has little effect on
cumulative water consumption.
The cooling of steam turbine condensers and
of gas compressor interstage coolers depends
on the cost and availability of water and, there-
fore, on the site. Three cooling options were con-
sidered representing different kinds of wet
evaporative cooling for turbine condensers and
gas-compressor interstage coolers (Table 4). The
cooling option determines whether turbine con-
densers are all wet cooled, whether parallel wet
and dry condensers are used, and whether gas
compressor interstage coolers are all wet cooled
or whether series dry and wet coolers are used.
The decision depends in part on the economics
of cooling.
Figure 3 shows the cost of steam turbine con-
denser cooling in Farmington, New Mexico. It is
clear that there is a cost of water above which it
is economical to use parallel wet/dry condens-
ers. This cost is approximately $0.20/1,000 gal.
The load on the wet cooler is about 10 percent of
the case for all wet cooling. Figure 4 shows the
effect of cost on water consumption for cooling
turbine condensers at two sites in the East and
two sites in the West.
Figure 5 shows the effect of cost on water
consumption for interstage cooling when 1,000
Ib of air is compressed. When the price of water
exceeds about $1.60/1,000 gal, the use of series
dry/wet interstage cooling is the least expen-
0.3
0.25 _
0.2
0.15
0.1
\
\
\
_ v.
I
o.s s;
0
0 0.2 0.4 0.6
WATER CONSUMPTION. GAL/KU-HR
O.b
Figure 3. Cost of steam turbine condenser
cooling in Farmington,
New Mexico.
sive option. The fraction of the cooling load to be
carried by the dry cooler varies significantly
with the cost of water. The effect of the cost of
water is more gradual than was found from the
calculations on turbine condensers. Above a
cost of $1.50/1,000 gal, approximately 50 percent
wet to 50 percent dry cooling should be used.
Where water is plentiful and inexpensive to
transport, high wet cooling should be used. The
cooling loads on both the turbine condensers
TABLE 4. THREE COOLING OPTIONS FOR CONVERSION PLANTS
Cooling
Option
High
Inter-
mediate
Minimum
Practical.
Water Cost and/or Water
($/1000 gals) Availability
<0.20
0.20-1.50
>1.50
Plentiful
Marginally
Available
Scarce
% Turbine
Condenser
Cooling Load
Wet Cooled
100
10
10
% Gas Compressor
Interstage
Cooling Load
Wet Cooled
100
100
50
547
-------
g
u.
O
LU
t/>
00
80
60
40
20
0
1
—
CASPER
X_
•\
— FARMINGTON
^.CHARLESTON
AKRON
10
20
30
40
WATER COST, CENTS/10 GAL
Rgure 4. The effect of water cost on
water consumed for cooling tur-
bine condensers.
100
M
S
8 80
60
AKRON
\CASPER
VA
' — FARMINGTON -V\ \ \
\\ \\
CHARLESTON -— \\ \ \
I
40
20
100
150
200
250
WATER COST. CENTS/10J GAL
Figure 5. The effect of water cost on
water consumed for interstage
cooling when compressing
1,000 Ib air.
and interstage coolers are taken to be all wet
cooled. When water is marginally available or
moderately expensive to transport, interme-
diate cooling should be used. Intermediate cool-
ing assumes that wet cooling handles 10 percent
of the cooling load on the turbine condensers
and all of the load on the interstage coolers.
Where water is scarce and expensive, minimum
practical cooling should be used. Minimum prac-
tical cooling assumes that wet cooling handles
10 percent of the cooling load on the turbine con-
densers and 50 percent of the load on the inter-
stage coolers. The amount of unrecovered heat
dissipated by wet cooling varies from 33 per-
cent for the Synthane process for high wet cool-
ing, to 18 percent for intermediate cooling, to 15
percent for minimum practical cooling. The high
value of 33 percent falls within the range of
Lurgi design data. The El Paso design4 indicates
that 36 percent of the unrecovered heat is dissi-
pated by evaporative cooling, while the Wesco
design5 indicates 26 percent dissipation.
Besides cooling, water consumption esti-
mates include process water requirements,
water required for mining and preparation of
the coal and shale, and for the disposal of ash or
spent shale, which is a function of location
through the amount of material that must be
mined or disposed. Sulfur removal also con-
sumes water: the amount depends not only on
the coal but also on the conversion process.
Water is also essential for other purposes (e.g.,
land reclamation) dependent on climate. Gen-
erally, because any one requirement is not
large, its needs can be met with lower quality
water. Nevertheless, when the requirements
are combined, they are significant and cannot be
neglected in any plant water balance, although
general rules for the amount consumed are not
easily stated. Differences in consumption in this
category for a given coal conversion process,
however, do not vary by more than 15 percent
between regions, except for the Four Corners
region. The difference is greater when this
548
-------
region is compared with others because larger
amounts of water are needed for handling the
high-ash Navajo coal and for dust control and
revegetation.
REGIONAL RESULTS
Table 5 summarizes the total net water con-
sumed for the three different cooling systems
and for all of the conversion technologies and
processes studied. The ranges in the total water
consumed reflect the variation with site. For oil
shale only intermediate cooling was considered.
The water requirements for standard-sized
plants range from 4 to 7 x 10e gal/d for coal gas-
ification and clean coal and from 3 to 8 x
10° gal/d for coal liquefaction; the range of net
water consumed for oil shale conversion is 5 to
8 x 106 gal/d.
To explain the similarities and differences in
net water consumed between the conversion
technologies, it is necessary to examine the
totals on a regional basis (Tables 6 and 7). For a
limited number of process-region-coal rank com-
binations not covered in this study, the results
given in Reference 6 have been used. It should
be noted that a larger percentage of the unre-
covered heat in the Lurgi process is dissipated
by wet cooling in Reference 6 as compared to
the present study, while for the SRC process
the overall conversion efficiency is lower in the
present study than that assumed in Reference 6,
resulting in larger wet cooling loads. However,
the data of Reference 6 present a useful data
base for the present study. Figures 6, 7, and 8
show a breakdown of the average net water con-
sumption by region and by process and for the
three cooling options. Four water use categories
are presented for each coal conversion process
in each region: net process water based on reuse
of all condensate; cooling water; flue gas desul-
furization water, if necessary; and water for
mining, dust control, solids disposal, water
treatment, revegetation, and other uses. For oil
shale it is convenient to break down the water
use categories in a different way to reflect the
large quantities of water required for spent
shale disposal: net process water for retorting
and upgrading; cooling water; water for spent
shale disposal and revegetation; and water for
dust control, mining, and other uses. For the
cases where the net process water is negative
(i.e., net water is produced in the process), the
cooling water requirements can be obtained
from Figures 6, 7, and 8 by adding the absolute
value of the process water to the cooling water
component.
Except for the Hygas process, the net water
consumed for the Four Corners region is higher
than for the other regions because of the larger
amount of water needed for dust control and the
handling of ash for the high-ash Navajo, New
Mexico coal. Water is required for revegetation
in New Mexico because the rainfall is less than
10 in/yr but is not required at any other location.
For the Hygas process, there are many compet-
ing demands that make the above generaliza-
tion invalid.
In the Illinois coal region, the average water
requirements for coal gasification are relatively
independent of the particular conversion proc-
ess, with the variation being no more than 15
percent for the high and intermediate wet-cool-
ing options and no more than 25 percent for the
minimum practical wet-cooling option. More
water is required for coal gasification than for
coal liquefaction which, in turn, requires more
water than coal refining. The water require-
ments range from a low of 9 gal/106 Btu to a high
of 28 gal/106 Btu, greater by more than a factor
of 3. In the Appalachian coal region, water re-
quirements (normalized with respect to the
heating value of the product fuel) for coal gasifi-
cation are greater than the requirements for
coal liquefaction for plants using bituminous
coal. For plants using lignite coal, water re-
quirements for coal gasification are slightly
lower than for coal refining. In the latter case,
this can be attributed to the high moisture con-
tent of the lignite coals and the very large quan-
tities of process water produced in the Lurgi
process. The Lurgi process accepts wet coal,
and the large quantities of dirty condensate pro-
duced are treated for reuse (at a cost) and are
subtracted from the process requirement. It
should also be pointed out that the net water
consumed in the Synthane, Hygas, and Synthoil
processes is virtually identical in both the Il-
linois and Appalachian coal regions for bitumi-
nous coals. However, the net water consumed in
the SRC process is higher for lignite coals than
for bituminous coals because of the lower con-
version efficiency attributed to the larger quan-
tity of energy required for drying the higher
549
-------
TABLE 5. SUMMARY OF NET WATER CONSUMED FOR STANDARD-SIZED
SYNTHETIC FUEL PLANTS
Coal Gasification
Lurgi
Syn thane
Hygas
Bigas
Coal Liquefaction
Synthoil
Coal Refining
SRC
Oil Shale
Paraho D: rect
Paraho Indirect
TOSCO II
Total Water Consumed (10 gpd)
High Wet Intermediate Minimum
Cooling Cooling Practical Cooling
4-7 2-5 2-5
5-6 4 4
5-6 4-5 4-5
6 4 3-4
5-6 3-5 3-4
4-7 3-4 2-4
5
8
8
Total Water Consumed
High Wet Intermediate
Cooling Cooling
18-30 9-22
22-27 16-19
21-26 16-19
25-27 16-18
17-21 11-14
13-21 8-13
(gal/106 Btu)
Minimum
Practical Cooling
7-21
15-17
15-19
14-17
10-14
7-11
18
28
29
wi
§
-------
TABLE 6. REGIONAL SUMMARY OF NET WATER CONSUMED IN 106 gal/d FOR
STANDARD-SIZED SYNTHETIC FUEL PLANTS
CMl CuiflcMlon
CMl Liq\*« fact ion
Co*l rafinlA9
Oil Sh«l«
P*r*ho Olr*ct
Parana Indirect
TOSCO IX
App&lachiu
•itiminou.
123
5. 3-5. 7 3. 9-4. 2 3. 6-3. 9
i toilon
Lljnlu
129
IllinoU Itogion
•Itwinou*
123
5 3-5.5 3.9-4.1 3.6-4.1
Povdar Rlwr/rt. Unio
120
6.0-6.4 4.1-4.4 3.7-4.1
S 9 3.7 3.4
n A*gioa*
Llfnlt*
123
S.7* 3.5* 3.1*
6.3-6.5 4.2-4.3 3.9-4.0
Four OoiiMra
SabbltiMiitmu
123
*.5« 4.1* 3.»*
Gram Uwf
rocmicto*
Oil Sh«l.
J
.
^
S.I
1.2
1.1
1 - High W*t Cooling, 2 - Znter**dl«t« W*t Coolinq. 3 - nlnimm Pr*ctic*l V«t Cooling
•OAtA fro« R«f. 6; only *pplie« to particulAT nurttcr *n4 not r«nq«.
-------
TABLE 7. REGIONAL SUMMARY OF NET WATER CONSUMED NORMALIZED WITH
RESPECT TO THE HEATING VALUE IN THE PRODUCT FUEL IN gal/106 Btu
Coal Gasification
Lurgi
Syn thane
Hygas
Bigas
Coal Liquefaction
Synthoil
Coal Refining
SRC
Oil Shale
Paraho Direct
Paraho Indirect
TOSCO II
Appalachian Region
Bituminous
123
27- 24- IS-
22-24 16-17 15-16
23-26 18-19 17-19
18-21 13-15 12-14
11* 7- 6*
Lignite
123
18 9 7
21 16 15
-
21 12 11
Illinois Region
Bituminous
123
25-28 19-21 17-19
22-23 16-17 15-16
24-25 19-20 18-19
25-27 16-18 15-16
19 13 12
15-17 10-13 9-12
Ponder River/Ft. Union Regions
Subbituminous-Bituminous
123
23-29 15-21 14-20
25-27 17-19 16-17
21-23 16-18 15-17
24 16 14
17 11 10
13-15 8-9 7-8
Lignite
123
22-24 14-15 12-13
24- 15- 13-
21 16 15
26-27 18 16-17
19- 14- 13-
15-21 8-9 7-8
Four Comers
Subbituminous
123
29-30 21-22 20-21
28- 18- 16-
23 18 17
20-22- 14-16- 13-16-
15- 11* 10*
TTrssn River
Formation
Oil Shale
2
_
-
-
18
28
29
8!
1 . (jph H«t Cooling. 2 • Intermediate Met Cooling. 3 - Minimum Practical Wet Cooling
-Data from Kef. 6: only applies to particular nusm>er and not rang*.
-------
10
01
§5
2000
-1000
ILLINOIS REGION
0 UUST CONTROL AND OTHER
Eg FLUE GAS KSULFURIZATION
O COOLING
G3 NET PROCESS
3
APPALACHIAN REGION
BITUHIHOUS COALS
1 - HIGH WET COOLING
2 - IHTERHEU1ATE WET COOLING
3 - MINIMUM PRACTICAL COOLING
APPALACHIAN REGION
LIGNITE COALS
LURGI SYNTHANE HYGAS B1GAS SYNTMOIL SRC SYHTHAHE HYGAS SYKTHOIL LURGI HYGAS SRC
Figure 6. Summary of average net water consumed for standard-sized coal
conversion plants located in the Central and Eastern States.
-------
POWDER RIVER AIIO FORT UUIOt! REGIONS
SUBS ITUI11HOUS COALS
POWDER RIVER AND FORT UNION REGIONS
LIGNITE COALS
ra
FOUR CORNERS
Gl
2
tt DUST CONTROL AND OTHER
E3 FLUE GAS DCSULFUB1ZATION
O COOLING
C3 NET PROCESS
1 - HIGH UET COOLING
2 - INTERH01ATE UET COOLING
3 - MINIMUM PRACTICAL WET COOLING
1
1
1
LURGI STNTNAHE HVGAS 8I6AS STHTMOH SK LURGI HTGAS B1GAS
LURGI HTGAS SrNTHOIl
Rgure 7. Summary of average net water consumed for coal conversion plants
located in the Western States.
-------
I
3
3000
2000
1000
0
GREEN RIVER FORMATION
OIL SHALE
INTERIC01ATE WET COOLING
-
^
'l
8
Si
1
fed
^
/>
\
I
i
1
1
•;•;
51
^n
P
%
1
1
1
1
1
1
::.:|
1
i
X
-
-
-
-
Q
7
6
5
u
4
3
2
1
0
-1
PARAHO PARAHO TOSCO
DIRECT INDIRECT II
30
20
5 10
GREEN RIVER FORMATION
OIL SHALE
P23 DUST CONTROL AND OTHER
fj& SPEUT SHALE DISPOSAL
CD COOLING
E9 RETORTING AHU UPGRADING
PARAMO PARAHO TOSCO
DIRECT INDIRECT II
Figure 8. Summary of net water consumed for oH shale conversion plants located
in the Western States.
moisture lignite coals prior to dissolution. The
alight difference in the results for the Hygas
process is caused by different process water re-
quirements for lignite and bituminous coals.
For each of the three basin-coal combinations
in the West, the net water requirements are
largest for coal gasification, followed in turn by
coal liquefaction, and coal refining (see Figure 7).
The larger requirement for the Four Corners re-
gion is attributed to the high-ash Navajo, New
Mexico coal. In the Powder River and Fort
Union coal regions the average wet water re-
quirements for the Lurgi, Hygas, and Bigas
processes are virtually identical for lignite and
subbituminous coals. The differences in the SRC
water requirements between the lignite and
subbituminous coals are attributed to the large
difference between the moisture content of the
two coals.
, The net water requirements for the Synthoil
and oil shale plants can be compared because
the products are roughly the same. The water
consumed in the Synthoil and Paraho Direct
processes is about equal. However, the water
consumed in the two indirect-heated oil shale
processes is 60 percent higher mainly because
of larger requirements for spent shale disposal
and revegetation.
Differences in water consumption between
the Illinois coal region and the Powder River
and Fort Union regions for subbituminous coals
for a given coal conversion process are relative-
ly small—no more than 15 percent with the ab-
solute difference being no more than 2.5 gal/106
Btu. However, for lignite coals, differences be-
tween the Appalachian coal region and the Pow-
der River and Fort Union regions are much
larger, the maximum being about 6 gal/106 Btu
for the Lurgi process and 4 gal/108 for the SRC
process, with the Lurgi water requirements
555
-------
being smaller in the Appalachian region and the
SRC requirements being smaller in the Powder
River and Fort Union regions.
In a particular coal-bearing region, differ-
ences in the water requirements for the four
coal gasification processes that we have consid-
ered are principally caused by differences in the
process water requirement and the differences
in the estimated overall efficiency resulting in
different cooling water requirements.
WATER AVAILABILITY AND COSTS
Two limiting cases were examined with
respect to water availability in the West: low
water demand and high water demand.1 Low
water demand corresponds to the production of
approximately 1.0 x 10fl bbl/d of synthetic
crude, or its equivalent in other fuels. For high
water demand, 1 x 106 bbl/d of synthetic crude,
or its equivalent in other fuels of 5.8 x 10
Btu/d, were produced in each of the three princi-
pal coal-bearing regions (Fort Union, Powder
River, and Four Corners) and in the principal oil
shale region (Green River Formation), for a total
production of 4 x 106 bbl/d.
Low water demand can be accommodated by
available supplies in most of the hydrologic re-
gions. However, chronic water shortages do
exist, especially in the northern Wyoming area
of the Powder River coal region and the
Tongue-Rosebud drainage area in the Fort
Union coal region. In the Four Corners-San
Juan region in northwestern New Mexico and
the Belle -Fourche- Cheyenne basin in northeast
Wyoming, water demands are excessive. For
high water demand, projected loads cannot be
accommodated by available supplies in most
subregions. Only in the Yellowstone, Upper
Missouri, Lower Green, and Upper Colorado
mainstem basins does it appear that sufficient
supplies are available for the expected loads of
energy production. However, water availability
in the Upper Colorado River Basin may be
limited because the water rights to most of the
free-flowing water in the Basin are already
allocated. These rights would have to be trans-
ferred to support additional energy develop-
ment or water transferred by transbasin di-
version.
Estimates have been made of the coat of
transporting water to the point of use from ma-
jor interstate rivers and riverways. Figure 2
shows the cost of transporting water to all sites
for low water demand. The cost of water deter-
mines the degree to which wet cooling should be
used. If water costs less than $0.25/1,000 gal, a
high degree of wet cooling should be used; if it
costs more than $1.50/1,000 gal, a minimum de-
gree of wet cooling should be used. In between
these extremes, intermediate wet cooling
should be used. Figure 2 shows that except for
plants located near the mainstem of major
rivers or near large reservoirs, intermediate or
minimum practical wet cooling is desirable for
most of the sites in the Western study area.
For large-scale synthetic fuel production, it is
more economical to have a large single pipeline
built to transport water to a large number of
plants than to have a large number of individual
pipelines supplying individual plants. Figure 9
shows the cost of transporting large quantities
of water (for high water demand) to some of the
major coal-producing areas and indicates that
except for large-scale development near the
mainstem of major rivers, intermediate cooling
is desirable for most of the study region.
The criterion of water availability is used to
determine the most suitable cooling option in
the Eastern and Central States. In this region
the adequacy of the water supply was assessed
by comparing a typical plant use with expected
low flows in the stream.1 In the Appalachian
coal region where coal is available, there are
many large rivers contiguous or adjacent to
sites with sufficient and reliable supplies of
water to support one or more large mine-plant
coal conversion complexes. This applies to all
plant sites in the vicinity of the Ohio, Allegheny,
Tennessee, Tombigbee, and Kanawha-New
Rivers. In most of these instances present
water use data and future demand projections
indicate a significant surplus beyond expected
use, even under low flow conditions.
The surface water supplies are less reliable in
the smaller streams, away from the major riv-
ers. Regions generally found to have limited
water supplies for energy development include:
the upper reaches of the Cumberland and Ken-
tucky Rivers in eastern Kentucky; the eastern
Kentucky and adjacent West Virginia coal re-
gions in the Big Sandy River Basin; and north-
ern West Virginia and western Pennsylvania in
the Monongahela River Basin, except those
areas that can be supplied from the Allegheny,
Ohio, or Susquehanna Rivers. Under future con-
556
-------
UPPER
MISSOURI
RIVER BASIN
rv:
KX [m
& \ WYOMING I
UPPER COLORADO
RIVER BASIN
Cost Of Water
($/iooo GALS)
<0-25
:~0 -25- 1-50
V\\\\ > I -50
STE LOCATIONS
m primary sites
• secondary sites
t pipeline
Figure 9. Cost of transporting water to coal regions in the Western States.
557
-------
ditions a minor surplus will exist for the
Tuscarawas River in Ohio. In these water-lim-
ited areas, extreme low flows are practically
zero, and a coal conversion complex could easily
represent a significant portion of the seasonal
low flow. In order for a plant to be sited here an
alternative or supplemental supply must be as-
sured. Figure 1 shows the availability of water
in the Appalachian coal region.
Within the Illinois coal region, the Ohio and
Mississippi Rivers have sufficient and reliable
water supplies to support one or more large
mine-plant coal conversion complexes. The
lower section of the Kaskaskia, Illinois, and
Wabash Rivers in Illinois; the Wabash and
White Rivers in Indiana; and the Green River in
Kentucky also have reliable supplies. Under fu-
ture conditions, deficit supplies are indicated
for the Wabash River in Illinois.1 Figure 1 shows
the availability of water in the Illinois coal re-
gion.
For each process, the average* water con-
sumed is relatively insensitive to the coal-
bearing region, and variations for a given cool-
ing option from site to site within the region are
expected to be small for all of the processes ex-
cept for possibly the SRC process, as discussed
above. However, within a given region, water
availability and cost may vary, and different
cooling options at different sites will produce
large differences in the cooling water consumed
and the plant water requirements.
REFERENCES
1. Gold, H., and Goldstein, D. J. Water-Related
Environmental Effects in Fuel Conversion.
Volume I. Summary and Volume II. Appen-
dices. Office of Research and Development.
U.S. Environmental Protection Agency,
Washington, B.C. EPA 600/7-78-197a, b. Oc-
tober, 1978 (also to be published as a DOE
report, 1979).
2. Gold, H., Nardella, J. A., and Vogel, C. A.
Water-Related Environmental Effects in
Fuel Conversion. (Paper presented at AIChE
71st Annual Meeting. Miami Beach. Novem-
ber, 1978. [also to be published in Chemical
Engineering Progress, 1979P
3. Goldstein, D. J., and Yung, D. Water Conser-
vation and Pollution Control in Coal Conver-
sion Processes. U.S. Environmental Protec-
tion Agency. Research Triangle Park, N.C.
Report No. EPA-600/7-77. June 1977.
4. Gibson, C. R., Hammons, G. A., and Cam-
eron, D. S. Environmental Aspects of El
Paso's Burnham I Coal Gasification Com-
plex. In: Proceedings, Environmental
Aspects of Fuel Conversion Technology.
Research Triangle Park, U.S. Environmen-
tal Protection Agency, October 1974. p.
91-100.
5. Berty, T. E., and Moe, J. M. Environmental
Aspects of the Wesco Coal Gasification
Plant. In: Proceedings, Environmental
Aspects of Fuel Conversion Technology.
Research Triangle Park, U.S. Environmen-
tal Protection Agency, October 1974. p.
101-106.
6. Probstein, R. F., and Gold, H. Water in Syn-
thetic Fuel Production—The Technology
and Alternatives. Cambridge MIT Press,
1978.
558
-------
APPLICABILITY OF PETROLEUM REFINERY AND
COKE OVEN CONTROL TECHNOLOGIES TO COAL CONVERSION
R. A. McAllister
Industrial Environmental Research Laboratory,
U.S. Environmental Protection Agency, Research Triangle Park, North Carolina
Abstract
There are similarities between many of the
process and waste streams of the petroleum
refining and coke oven industries, particularly
the latter, and streams in the coal conversion in-
dustry. The well established environmental and
process control technologies of petroleum refin-
ing and coke oven industries have been re-
viewed. The process/waste streams from several
coal conversion processes were characterized
and streams with refinery and coke oven coun-
terparts were identified. The control technol-
ogies currently used in refining and coke oven
industries for the management of the identified
streams were evaluated for their applicability to
the counterpart coal conversion streams.
For many of the major controls needed (e.g.,
desulfurization, fugitive emissions, and waste-
water treatment), the current industrial practice
seems adequate for the coal conversion indus-
try. Significant composition differences prevail
among the comparable streams, however, and
additional testing and development of pollution
control strategies for the coal conversion indus-
tries are indicated.
Based on current technology, primarily de-
rived from petroleum refining and coke oven in-
dustrial practice, the evolution of the technol-
ogy needed to operate a coal conversion facility
in an environmentally safe manner appears pos-
sible. Cost analyses have not been made here,
but they are expected to be significant
INTRODUCTION
A major effort has been mounted by govern-
ment and private industry to develop the tech-
nology necessary to increase the Country's ca-
pability to become less dependent on foreign
sources of energy. A significant part of this ef-
fort is directed at conversion of coal to gaseous
and liquid energy sources (fuels), and to sources
of industrially useful chemicals. The U.S. Envi-
ronmental Protection Agency has initiated a
comprehensive assessment program to antici-
pate potential environmental problems in the
coal conversion industry, to help evaluate and
develop suitable control measures, and to build
the data bases needed for establishing appro-
priate regulations.
This paper reviews the technologies current-
ly used by two large fossil fuel industries to
identify those environmental control processes
that may be applicable to the coal conversion in-
dustry. The petroleum refining and coke oven
industries both have extensive experience in de-
veloping pollution control strategies and are
continuing to improve their control techniques
to meet even more demanding regulations.
Those refinery streams having compositions (or
components) which have similarities in coal con-
version processes have been examined, and the
applicability of the control technology has been
evaluated. Only the recovery type of coke oven
plant was studied here. This so-called "by-prod-
uct" plant has several process streams with
components and compositions similar to those
expected in the coal conversion industry. The
by-product coke oven industries have developed
several control strategies that appear to be
useful in the coal conversion industry. Both the
petroleum refining and coke oven industries
face similar problems, and both have shared in
the development of control technologies.
Several selected conversion processes have
been scrutinized for the applicability of pe-
troleum refinery or by-product coke oven proc-
ess/effluent control technologies. These proc-
esses are the Koppers-Totzek low/medium-Btu
gasification, the Lurgi (dry ash) high Btu gasifi-
cation, and the COED (Char-Oil-Energy Devel-
opment) liquefaction processes. Some compar-
ison was made to the Synthane gasification
process and the solid-product Solvent-Refined
Coal (SRC-I) process. SRC-I is a specialized
559
-------
process whose primary purpose is deashing of
coal, rather than forming coal-derived liquids or
gases. Selected processes represent a range of
operating conditions, typify processes with
more data available on waste stream character-
istics, and have commercial status impending or
already realized. Figure 1 is a generalized flow
diagram for both liquefaction and gasification of
coal. Normally only one of the paths from coal to
product would be followed depending on
whether the major product was a gas or a liquid.
Two primary references12 were used in de-
veloping the material presented here. Both are
excellent reviews. The first pertains to the ap-
plicability of petroleum refinery control technol-
ogies to coal conversion. The second reviews
coke oven processes and control technologies
and assesses their applicability to the coal con-
version industry.
No attempt has been made to consider all the
refinery or the coke oven industry control tech-
niques. The coal conversion industry suffers
from the disadvantage that urgency, technol-
ogy, and design have outdistanced full-scale
plant experience. As a result, the control tech-
nology must be based on some uncertainty
coupled with the expectation that further devel-
opment and even new techniques will be neces-
sary as more data and experience become avail-
able.
REFINERY CONTROL TECHNOLOGY
A major obstacle to a detailed assessment of
the applicability of refinery control technologies
to coal conversion waste streams stems from
the nonexistence of commercial substitute nat-
ural gas (SNG) and liquefaction facilities in the
United States. Commercial gasification and liq-
uefaction facilities operating in foreign coun-
tries do not generally incorporate those design
and operating features to minimize waste gen-
eration and to control discharges that would be
employed in a similar facility in this Country.
The coals used at the foreign facilities differ
from those available to commercial plants in the
United States; thus, the waste stream charac-
teristics would also differ. The availability of
detailed data from foreign commercial facilities
is not extensive, although some progress is
being made in this^area.
Data from U.S. pilot coal conversion facilities
are normally not completely applicable to com-
mercial-sized plants; nevertheless, they give
certain clues upon which some generalizations
may be made. Relatively much data are avail-
able for many refinery waste streams. Figure 2
is a highly schematic flow diagram for a petro-
leum refinery.
Based on the review of the available data and
from a control technology applicability view-
point, a limited number of refinery and coal con-
version process/waste streams appear to have
certain similar characteristics. These streams
and the basis for their similarities are listed in
Table 1. Despite the noted similarities, there ap-
pears to be significant composition differences
between the analogous streams which would af-
fect applicability and design of a control technol-
ogy. For example, while both the refinery proc-
ess sour gases and the quenched product gas
from coal gasification contain H2S and C02, the
H2S concentration is considerably higher and
the C02 level is significantly lower in most
refinery sour gases. Even when selective H^
removal processes are used, the treatment of
the coal conversion raw product gas results in
production of a concentrated acid gas stream
with C02 levels much higher than those in refin-
ery sour gases. High C02 levels significantly af-
fect the efficiency and cost of operating subse-
quent equipment used to remove sulfur. Unlike
sour waters from refineries which contain high
levels of both sulfides and ammonia, most coal
conversion condensates contain low levels of
sulfide and moderate levels of ammonia. Be-
cause of the differences in the nature of the raw
material and the processing steps employed, the
dissolved and particulate organics found in coal
conversion wastes are different than those in re-
finery wastewaters. The organics in coal con-
version wastes are generally more aromatic; or-
ganics in refineries are largely aliphatic. The
differences in wastewater characteristics also
are reflected in the characteristics of oily
sludges and biosludges resulting from waste-
water treatment. In comparing coal conversion
waste streams with their analogues in refin-
eries, it should be noted that there can be wide
differences between stream compositions from
different coal conversion plants depending on
the coal processed, conversion process used, and
on-site product upgrading methods employed.
The refinery control technologies which may
find application to coal conversion are listed in
Table 2. Some of the control processes, such as
560
-------
COAL
STORAGE, HANDLING,
AND PREPARATION
LIQUEFACTION
PRODUCT
SEPARATION
RAW
GAS
ACID
GAS
HYDRO-
TREATING
T
ACID GAS
TREATMENT
SHI FT AND
METHANATION
LIQUID PRODUCT
LOW ENERGY
PRODUCT GAS
HIGH ENERGY
PRODUCT GAS
Figure 1. Coal conversion processes.
561
-------
CRUDE OIL
PRODUCTS
Rgure 2. Petroleum refinery.
562
1
,
r 1
*
DISTILLATION
j RAW P
TREATING— HYDRO
SOLVENT EXTR
f
RODUCTS •
Y
FREATING, COKING
ACTION, ETC.
1
r
i '
\ i
PROCESSING—CATALYTIC CRACKING
HYDROCRACKING, REFORMING,
ALKYLATION, POLYMERIZATION, ETC.
>
r <
[ i
r i
. > i
STABILIZATION, TREATING, FRACTIONATION,
ABSORPTION, EXTRACTION, ETC.
i
•
^
<
r
i
f i
BLENDING
^
r
»
1
1 r
r
SOUR GASES
SOUR GASES
EFFLUENTS
-------
TABLE 1. SIMILAR REFINERY AND COAL CONVERSION WASTE STREAMS
Refinery Streams
Coal Conversion Counterparts
Major Similarities
Gaseous
Process sour gas
Catalyst regenerator
off-gas
Fugitive emissions
Liquid
Sour waters
Oily waters
Solid
Spent catalysts
Sludges
Quenched product gas, acid gas, and
fuel gas (from liquefaction)
Raw product gas and char combus-
tion flue gas
Fugitive emissions
Raw product gas quench condensate,
waste liquor purge (from lique-
faction), and shift condensate
Raw product gas quench condensate
and waste liquor purge (from
liquefaction)
Spent shift, methanation, hydro-
treating, and Claus plant catalysts
Oil and biosludges
High H~S and ammonia content;
presence of COp
High CO and particulates, NO , and N
A
Hydrocarbons, sulfur compounds, and
ammonia
Ammonia, sulfide, phenols, oils, and
grease/tars
Oil and grease/tar; phenols
Metals (Ni, Co, Mo, etc.), bauxite
Oil and grease/tar, inerts, biomass,
refractory organics
-------
TABLE 2. REFINERY CONTROL TECHNOLOGIES AND THEIR APPLICABILITY
TO COAL CONVERSION
Refinery Control Technology
Applicability to Coal Conversion Waste Streams
Acid Gas Treatment
Diethanolamine (OEA),
Fluor Econamine,
Oiisopropanolamlne (AOIP),
etc.
Physical Solvents
Selexol,
Rectisol,
etc.
Sulfur Recovery
Claus
Stretford
Tail Gas Treatment
IFP-1, Sulfreen
SCOT, Beavon, and Cleanair
Potentially suitable for non-selective re-
moval of H2S and C(L from product gases from
atmospheric/low pressure gasification/lique-
faction. Also suitable for hydrocarbon re-
moval from concentrated acid gases and for
concentrating dilute H2S streams for feeding
to Claus plant. Extensive solvent degrada-
tion may be encountered in coal conversion
applications.
Potentially suitable for selective removal
of HoS and C02 from product gases. Best
suited to hign pressure application. The
resulting concentrated acid gas stream may
contain high levels of hydrocarbons, thus
requiring further treatment prior to sulfur
recovery.
Split-flow mode applicable to coal conver-
sion acid gases containing more than 15%
H2S. Sulfur burning mode applicable to
feeds containing as low as 5% FLS. Removal
of ammonia and hydrocarbons from feed gases
would be required to prevent ammonium bi-
carbonate scaling and carbon deposition on
catalyst, respectively.
Most existing applications are to acid gases
containing low levels (around 1%) of hLS.
High C02 levels necessitate pH adjustment
and result in high blowdown rates. Rela-
tively large unit sizes would be required
with high CCL gases. Process does not
remove non-Hp sulfur compounds.
Suitable for Claus plant tail gas treatment;
cannot achieve very low levels of total sul-
fur in the off-gas which may be required by
emission regulations. Efficiency decreases
with increasing C02 level in the feed.
Sulfur removal efficiencies decrease and CCL
levels in tail gas increase when acid gases
contain high COp levels.
564
-------
TABLE 2 (continued)
Refinery Control Technology
Applicability to Coal Conversion Waste Streams
Tall Gas Treatment (cont.)
CMyoda Thoroughbred 101,
Wellman-Lord, IFP-2, and
Shell CuO
Fugitive Emissions and Odor
Control
Vapor recovery, incineration
source elimination
Sour Water Stripping
Conventional Stripping and
Chevron Wastewater
Treatment Process
Oily Water Treatment
API Separator and Flota-
tion
Biological Wastewater Treatment
Carbon Adsorption and Chemical
Oxidation
Slop 011s and Sludge Treatment
(thickening, centrlfugation,
emulsion breaking, drying beds)
In-Plant Waste Volume and
Strength Reduction
Resource Recovery
Incineration
Land Disposal
Potentially suitable. Requires feed incin-
eration to convert reduced sulfur to S02-
Applicable to analogous sources.
Applicable to coal conversion sour waters.
The design must be modified to allow for the
lower sulfide and often higher ammonia
levels in coal conversion sour waters.
Applicable; units must be designed based
on specific wastewater characteristics.
Generally applicable; biodegradability of
coal conversion waste components not
established.
Should be applicable; design basis must be
established for the specific wastewater.
Generally applicable; design basis must
be established for the specific waste.
Applicable.
Applicable to spent catalysts for material
recovery; sale of tars/oils
Applicable to organic wastes; incinerator
and emission control designs would be feed
specific.
Applicable.
565
-------
sulfur recovery plant tail gas treatment, would
be applicable to waste streams in a coal conver-
sion plant, and their design may be essentially
the same as in refinery applications. However,
as noted in Table 2 under Tail Gas Treatment,
with high C02 levels in the tail gas, as would be
expected in coal conversion applications, the ef-
ficiency is low for some of the processes (IPF-1,
Sulfreen, SCOT, Beavon, and Cleanair). Other
processes such as Stretford, Glaus, and steam
stripping would require extensive modifications
to account for differences in waste composi-
tions. Because of limited data on certain waste
characteristics (e.g., the biodegradability of
some aromatic organic compounds and the set-
tleability of coal conversion solids in waste-
waters) the applicability and efficiencies of proc-
esses such as bio-oxidation, flotation, sludge
dewatering, and emulsion breaking in coal con-
version application cannot be accurately assess-
ed at this time. With the exception of the few
processes that have been tested in"coal conver-
sion applications, such as the Rectisol and the
Fluor Econamine (diglycolamine [DGA]) acid gas
treatment processes and the Stretford tail gas
treatment process, the processes listed in
Table 2 have not been employed in such an ap-
plication. For the processes that have been used
in coal conversion, only limited data are avail-
able on process design and performance. Even
though the processes listed in Table 2 appear
applicable to coal conversion wastes, additional
testing will be required to confirm applicability
and to define criteria for large-scale design and
cost estimation. It should be noted that the suit-
ability of a control process for use in coal con-
version plants cannot be determined separately
from other processes and waste treatment oper-
ations within an integrated coal conversion fa-
cility. The selection of a specific control process
is merely an element in the overall facility
waste management plan, which includes con-
siderations of overall emission/effluent limita-
tions, energy and raw material availability, and
costs.
Some of the components in refinery and coal
conversion wastes are important from the
standpoint of presenting potential occupational
health hazards to plant workers and adverse
health impacts on the general population. Sev-
eral hazardous waste compounds (e.g., H2S, CO,
and mercaptans) are not unique to refinery or
coal conversion wastes and are emitted from a
variety of other industrial and nonindustrial
sources. The hazardous characteristics of many
of these commonplace substances are generally
well documented. The hazardous chemicals
which are unique to coal conversion and refin-
eries fall into three categories: polynuclear
aromatics, heavy metals and organometallic
compounds, and low molecular weight aromatic
substances. Many of the control technologies
used in both refineries and coal conversion
plants should result in partial or total removal
of the hazardous waste components. The fate of
many of the hazardous components in pollution
control processes is not well known, and the re-
quirements for additional controls cannot be
defined at this time.
COKE OVEN CONTROL TECHNOLOGY
Coke is produced by destructive distillation—
also referred to as pyrolysis or carbonization —
of low-sulfur bituminous coal in an oven or re-
tort in the absence of air. The coking tempera-
ture of about 1,100° C is generally higher than
most coal conversion pyrolysis and is conducted
in a reducing atmosphere in the coke oven as
contrasted to a more oxidizing atmosphere in
coal conversion units. The latter are generally
operated at a higher pressure than the coke
oven, which is essentially at atmospheric pres-
sure. Coal used in coke making is usually a blend
of high-volatile coal with a 10 to 50 percent low-
volatile coal. The blend usually does not contain
over 1.5 percent sulfur or 9 percent ash. Ap-
proximately 16 percent of the bituminous coal
mined in the United States is converted to coke,
which is used principally in blast furnaces and
foundries. More thdn 98 percent of the total U.S.
coke is produced from by-product coke oven sys-
tems. The by-product process is oriented
toward the recovery of the gases and chemicals
produced during the coking cycle.
Figure 3 shows a typical by-product coke
oven process. The major steps or process units
involved in the by-product coke plant, in se-
quence, are: coal handling and preparation
equipment, coke ovens, quench stations, pri-
mary cooler, tar separator, tar extractor, am-
monia removal unit, final cooler, light oil scrub-
ber, and sulfur removal unit(s). In addition, some
modern coke plants have chemical refining facil-
ities for recovery of benzene, toluene, and
xylene from the light coal oils.
566
-------
COAL
i
STORAGE, HANDLING,
AND PREPARATION
FUGITIVE
EMISSIONS
i
FLUE GAS
COKE OVEN, QUENCHING, AND COOLING
TAR REMOVAL
I
NH3 REMOVAL
NAPHTHALENE
REMOVAL
i
LIGHT OIL REMOVAL
AND REFINING
I
DESULFURIZATION
COKE OVEN
GAS PRODUCT
PHENOL REMOVAL
40% COKE OVEN GAS FOR
COKE
PRODUCT
RECOVERED
COKE
BREEZE
INDIRECT HEATING OF OVENS
Figure 3. Byproduct coke oven process.
567
-------
The core of the process is the coke ovens,
which are narrow chambers usually about 12 m
long, 5 m high, and tapering in width from about
50 cm at one end to 40 cm at the other. The
ovens hold about 18 Mg of coal each and are
built in batteries of about 100 ovens. Although
coke production from each oven is basically a
batch process, a coke oven plant operates such
that the battery of ovens continuously produces
coke oven gas and byproduct chemicals. In the
by-product coke oven process, coking is accom-
plished at temperatures of 1,090° to 1,150° C
and atmospheric pressure for a period of 16 to
27 hr.
One Mg of the low-sulfur bituminous coal fed
into a by-product coke oven would yield the fol-
lowing:
Coke
Coke breeze
Tar
Anhydrous ammonia
Light oil
Gas, 293m3 (10,350 std ft3)
Water
kg
715
46.5
39.0
2.5
10.0
154.5
32.6,
1,000
Coal gasification processes may be subdi-
vided into low-, intermediate-, and high-temper-
ature operations. These may be further subdi-
vided by operating pressures. The low-tempera-
ture gasification processes tend to show a com-
plete product and by-product slate, including
oils, tars, and phenols. As the gasification
temperature increases, the quantity of oils, tars,
and phenol decreases in preference to lighter
products. The operating pressure also affects
the yields. As the pressure increases, the prod-
uct slate becomes heavier.
Table 3 is a comparative listing of coke oven
and coal conversion process and waste streams.
The gaseous streams listed in Table 3 include
the raw gas from the coke oven and from coal
conversion counterparts. Fugitive emissions are
listed under gaseous streams, but a significant
component in coke oven fugitive emissions re-
sults from airborne coal particles and coke. Fu-
gitive emissions in the coal conversion process
are varied in composition and source. Gas-borne
solid particulates include coal from the coal pile
and coal particles airborne in such handling
processes as crushing, sizing, transporting, and
oven loading. Coke particulates in the unrecov-
ered coke breeze caused by the coke handling
and quenching operations account for a major
share of the gaseous fugitive emissions. Solid
tar particulates are among the fugitive emis-
sions from the tar separator, exhauster, and
electrostatic precipitator. Liquid H2S04 mist
and solid ammonium sulfate particulates may be
generated in the ammonia removal steps. Solid
particulates may also be generated in the under-
firing of the coke ovens by the clean coke oven
gas. Odors are among the fugitive emissions
from the coke ovens, coke handling and quench-
ing operations, tar separators, ammonia re-
moval, naphthalene removal, light oil recovery,
and desulfurization steps. Other specific major
sources of fugitive emissions include coal-charg-
ing hole lids, coke-pushing operations, and door-
seal leaks. Additional sources include pumps,
compressors, valves, and flanges. Most of the
latter group are universal problems in facilities
where chemicals are processed. Diligence in
simple maintenance procedures can often
significantly reduce emissions from many of
these sources.
The coke breeze listed under the coke oven
solid waste streams is the solid coke fines that
are recovered during the quenching operation.
Table 4 presents data comparing gas streams
from a coke oven plant, a refinery, and two
gasification plants. Many similarities are ap-
parent in the components present and in their
compositions. Differences, some of which are im-
portant from a process standpoint, can also be
seen. The hydrocarbon content of the refinery
process sour gas stream is much higher than
that of either the coke oven gas or the coal con-
version gases. There is more hydrogen sulfide
in the refinery stream than in the other
streams. Note the bottom entry in the table, the
ratio of C02 to H2S in the streams. For the
refinery gas, the ratio is much lower than either
the coke oven gas or particularly the coal con-
version streams. High C02/H2S ratios in the lat-
ter make sulfur removal and recovery more dif-
ficult in the coal conversion processes.
A number of processes are being utilized to
remove hydrogen sulfide and recover sulfur
from coke oven gas. These processes are di-
vided into three major categories: liquid absorp-
tion processes (Vacuum Carbonate, Sulfiban
[amine], Firma Carl Still); wet oxidative proc-
esses (Stretford, Takahax, Giammarco Vetrc-
568
-------
TABLE 3. COKE OVEN AND COAL CONVERSION STREAM SIMILARITIES
Coke Oven Streams
Gaseous
Raw gas and acid gas
Fugitive emissions
Ammonia liquor
quench water
Coal pile run-off
Solid
Coke breeze
Oily solids and
biosludges
Tar, naphthalene,
light oil, phenol,
and ammonia
Coal Conversion
Counterparts
Raw gas and acid
gas from gasifica-
tion, and off-gas
from liquefaction
Fugitive emissions
Process wastewater
Coal pile run-off
Coal fines, chars
Oily solids and
biosludges
Tar, naphthalene,
light oil, phenol,
and ammonia
Major Common Pollutants
or Similarities
H2S, NH3, CO, C02, COS,
CS2, and hydrocarbons. See
Table 4 for further details.
Same as above, plus particu-
lates. See Table 6 for some
detail.
NH3, phenols, oils, sul-
fides, and cyanides. See
Table 5 for some details.
Suspended solids and
organic extracts.
Similar by-products.
Oil, grease and tar,
biomass, and refractory
organics.
Similar by-products.
coke); and dry oxidative processes (Iron Oxide
or dry box). Historically, the Iron Oxide process
has been used most extensively. However, the
Vacuum Carbonate process, the Stretford proc-
ess, and more recently, the Sulfiban process
have moved into commercial prominence. The
liquid adsorption processes are called sulfur
removal processes, in that they remove sulfur
compounds, notably H2S, COS, and CS2, from
gaseous streams. When the solvent is regener-
ated, generally a gaseous stream more concen-
trated in H2S results. The oxidative processes
described are sulfur recovery processes in
which elemental sulfur is the product. The
Stretford process does not remove COS or other
organic sulfur compounds from the gas stream.
The Claus sulfur recovery process is also used
but initially had some problems associated with
hydrogen cyanide, iron sulfide, and iron cya-
nide. These problems were resolved by adjust-
ing the Claus unit. A Sulfiban unit removes both
C02 and H2S from the coke oven gas stream
utilizing a nonselective solvent. A Claus unit is
required to convert H2S to sulfur to recover the
sulfur.
The H2S removal or sulfur recovery efficien-
cies achievable for the processes in the coke
oven industry are: Iron-Oxide process, 99 per-
cent (for low gas volumes); Vacuum Carbonate
process, 93 to 98 percent; Sulfiban process, 90 to
-------
TABLE 4. COMPARISON OF GASES
Component/Parameters
H2
CH4
C2H4
C0 to C,
o o
CO
co2
°2
N2
NH3
HCN
H2S
COS
cs2
Light 011
Tar 011
Tar
Phenol
H20
Total
Temp., °C
pressure, MPa
C02/H2S
Raw
Coke Oven
Gas
Vol %
38.22
25.51
2.99
—
6.18
1.33
1.26
0.452
0.70
0.16
0.51
0.018
0.01
0.79
—
0.78
0.04
21.05
100.00
538 •
0.099
2.6
Ref 1 nery
Process
Sour Gases
Vol %
—
8.4
5.2
19
—
4.9
—
—
—
62.5
—
—
—
—
--
«
—
100.00
48
0.10
0.078 '
Gasification
Lurgi
Vol %
22.63
6.75
0.23
—
11.65
16.16
0.18
0.55
0.16
0.203
0.017
~
0.14
o.n
0.10
0.05
41.07
100.00
188
3.10
79.6
Koppers-
Totzek
Vol %
26.37
—
—
—
51.79
8.82
0.69
0.08
0.02
0.41
0.04
—
—
--
—
—
11.78
100.00
1,500
0.105
21.5
570
-------
98 percent; Stretford process, 99.5+ percent
(for H2S only); and Claus (sulfur recovery) proc-
ess, 95 to 96 percent.
Among the acid gas removal processes in the
coke oven industry, the amine and carbonate
solvent processes should have application in
low-pressure gasification processes or in treat-
ing low-pressure off-gases from liquefaction
processes. The two most common sulfur recov-
ery processes in the coke oven industry are the
Claus and Stretford processes. Both of these
processes will have wide application in the coal
conversion industry. Care must be taken to con-
sider the effect of the C02 composition on both
the Claus and Stretford processes when used
for coal conversion applications having high C02
compositions. High C02 affects the stability of
the flame in the Claus reactor and also results in
higher COS concentrations in the tail gas from
the Claus unit. In the presence of NH3, an am-
monium bicarbonate can form that reduces the
performance of the Claus catalyst. C02
neutralizes the Stretford solution and reduces
the absorption rate of the H2S, thus
necessitating a higher solvent circulation rate
and larger units. For coal conversion applica-
tions, such as a gasification process having a
high hydrocarbon and C02 composition in the
acid gas stream, an enrichment step using an
amine process such as ADIP would probably be
effective. The enriched gas would be fed to a
Claus unit for sulfur recovery. Additional treat-
ment of the tail gas from the Claus unit would
be required before discharge to the atmosphere.
Generally, the Stretford process is more
economical when the acid gas stream contains
less than 15 percent HgS, whereas the Claus
process is the choice for levels about 15 percent.
The wastewater characteristics of the differ-
ent processes are compared in Table 5. All of
the major coke oven wastewater treatment
processes should have applications in coal con-
version waste treatment. The process waste-
waters from the by-product coke plants contain
large amounts of phenol, ammonia, sulfide, cya-
nide, and oil and grease. Various control tech-
nologies are being used to remove these pollut-
ants.
Ammonia is being removed and recovered by
steam stripping at alkaline pH, or by Phos-
am-W, a proprietary (U.S. Steel) process that
uses an ammonium phosphate scrubbing solu-
tion and distillation in combination to produce
an anhydrous ammonia product. Sulfide re-
moval from wastewater by steam stripping is
not commonly practiced in the coke oven indus-
try.
Phenols are being removed by solvent extrac-
tion, steam stripping and/or biological oxidation,
and carbon adsorption. Biological treatment has
been successful with coke oven wastewaters in
meeting existing phenol regulatory limitations.
Phenol removal efficiency of about 99.8 to 99.9
percent has been achieved by the activated
sludge system: BOD removal has ranged from
85 to 95 percent. Activated carbon adsorption as
a final polishing treatment has been practiced in
the coke oven industry. Carbon adsorption may
have applicability in coal conversion processes,
especially if char could be used as an activated
carbon.
Many coke oven plants recycle cyanide-con-
taining wastewaters and use them for coke
quenching. There would be no counterpart
operation in coal conversion operations with the
possible exception of the ash quenching. In the
coal conversion industry, levels of HCN are
generally lower than in the coke oven industry.
Some coke oven plants use a by-product light
oil upgrading process which has a potential ap-
plication in the coal conversion industry. This
process, called the Litol process, has been
developed and licensed by the Houdry Division
of Air Products and Chemicals, Inc. It is a
catalytic process by which coke oven light oils
are refined and dealkylated to produce high-
quality, even reagent-grade benzene at essen-
tially stoichiometric yields.
The coke ovens are a major source of air
pollution emissions in the steel industry. Top-
side coke oven workers have a substantially
higher risk of lung cancer than the average
worker, probably from carcinogenic materials
associated with the particulate fraction of the
coke oven emissions. Various schemes to control
these emissions and alleviate potentially ad-
verse health effects are being developed includ-
ing collecting and removing the smoke, particu-
late matter, and gaseous emissions that occur
during the charging, coking cycle, and pushing
and coke-quenching operations. An enclosed
coke-pushing and quenching system is being
developed jointly by the EPA and the National
Steel Corporation. In this system, the coke will
571
-------
TABLE 5. COMPARISON OF PROCESS WASTEWATERS
Constituent
Ammonia
Carbonate
Chloride
Cyanide
Phenol
Sulfide
Thiocyanate
pH
COD, mg/1
Suspended Sol Ids, mg/1
Coke-Oven
Plant
Liquor
mg/1
5,000
—
6,000
50
1,000
1,250
1,000
8.4
10,000
4,000
Synthane
mg/1
8,100
6,000
500
—
2,600
1,400
152
8.6
15,000
600
Lurgi
mg/1
11,200
10,000
—
—
3,500
--
--
8.9
12,500
5,000
Koppers-
Totzek
mg/1
25
1,200
600
0.7
--
—
—
8.9
70
50
SRC- 1
mg/1
5,600
—
--
--
4,500
4,000
--
8.0
15,000
300
remain totally enclosed, from the moment it
leaves the oven until after it is quenched. Emis-
sions evolved during the push and transfer to
the quench station are drawn off and removed
by a high-energy scrubber. Another system,
developed by Koppers Company, is being tested
at the Ford Motor Company to abate coke oven
fugitive emissions. Principal features of this
system are a fume-collecting hood, a fume main,
a venturi scrubber, and a modified quench car
with a synchronization system for coordinating
the quench car's movement with that of the
pusher. The Ah* Pollution Control Association's
April 1979 conference on "Control of Air Emis-
sions from Coke Plants" reflects the industry's
continuing efforts to improve the technology in
this area. These types of fugitive emission con-
trols may have potential applications in the Syn-
fuels industry in analogous situations; e.g., in
ash quenching or SRC solidification unit opera-
tions. A summary of the coke oven control
technology for fugitive emissions is shown in
Table 6. In general, the problem of fugitive
emissions is expected to be much less in a coal
conversion plant than in the coke oven industry.
Analogous operations, after the coal storage,
handling, and preparation steps, would be in
feeding the lockhoppers in the coal conversion
industry and charging the coke ovens. The aspi-
ration systems, the closed charging systems,
and the "smokeless" charging systems used in
the coke oven industry would have applications
to the lockhopper charging operation in coal
conversion industry. Other possible applications
are indicated in Table 6.
Table 7 summarizes the various coke oven
control technologies that may have potential ap-
plications in the coal conversion industries.
572
-------
TABLES. COKE OVEN CONTROL OF FUGITIVE EMISSIONS
Operation/Emission Source
Pollutants
Control Technology
Coal Conversion
Applicability
Charging
Discharging
Quenching
Improved Operating
Procedures and
Maintenance
Coal Participates. SO ,
Hydrocarbons. CO. NO .
& Ammonia
Coke Particulates,
Hydrocarbons. Ammonia,
& CO
Coke Particulates. &
Coke Breeze
Particulates, Hydro-
carbons, CO, Ammonia.
4 NOU
Aspiration Systems
Larry-Mounted Scrubbers
Smokeless Charging System with
Steam Jets
Closed Charging Systems
Bench-Mounted Self-Contained
Hoods
Coke Car - Mounted Hoods
Fixed Duct Hoods
Spray Systems
Coke-Side Enclosures
Internal Baffles
Dry Quenching
Closed Quenching
Mechanical Lid Lifters
Electric Eye Synchronization
Oven and Door Maintenance
Applicable
N.A.
P.A.
Applicable
N.A.
N.A.
P.A.
P.A.
N.A.
N.A.
P.A.
Applicable
N.A.
N.A.
P.A.
N.A. - Not Applicable
P.A. - Possibly Applicable
-------
TABLE 7. COKE OVEN PLANT CONTROL TECHNOLOGIES AND THEIR
APPLICABILITY TO COAL CONVERSION
Coke Oven Plant Control
Technology
Acid Gas Treatment
Amine solvents
Carbonate solvents
(e.g., Vacuum Carbonate
and Benfield)
Sulfur Recovery
Stretford
Claus
Fugitive Emissions Control
Coal handling and loading
Enclosed coke pushing
and quenching system
Fume recovery and
scrubbing
By-product Recovery/Refining
Ammonia from wastewater
(Stripping, Phosam - W)
Ammonia from raw gases
(Scrubbing, Phosam - W)
Phenol from wastewater
(Solvent extraction)
Applicability to Coal Conversion
Systems
Suitable for removal of HLS and C02 from
low pressure raw product and off gases.
Solvent degradation may be encountered.
Can produce high H2S concentration
streams.
Same as above. Processes partially remove
carbonyl sulfide and cyanides. Benfield
process suitable for high pressure
application.
Suitable for low H-S (less than 15%)
containing gases. Organic sulfur not
removed.
units.
High C02 levels require large
Applicable for high HpS (greater than
15%) containing gases. Removal of high
levels of cyanide, ammonia, and hydro-
carbons will be required.
Potentially suitable.
Potentially suitable for ash quenching,
SRC solidification applications.
Applicable to analogous sources.
Suitable for sour waters.
Applicable for low pressure gas
purification.
Suitable for process wastewater
containing 1,000 ing/1 or more phenol
574
-------
TABLE? (continued)
Coke Oven Plant Control
Technology
Applicability to Coal Conversion
Systems
Tar, naphthalene, and light
oil from raw gases
Light oil refining
(e.g., L1tol process
and solvent extraction)
Wastewater Treatment Technology
Biological oxidation;
carbon adsorption;
ammonia, phenol, and
oil removal processes
Suitable, but design must be modified
for different pressures, temperatures,
and compositions.
Suitable for recovery of benzene, toluene,
and xylene (BTX) from coal derived naphthas.
Generally applicable; design basis
must be established for the specific
waste.
Moat of the control technologies listed in Table
7 have been tested in coal conversion applica-
tions; however, most of these applications have
been in process development units or pilot-scale
coal gasification and liquefaction systems. A few
successful uses have been with commercial first-
generation coal gasification processes; e.g., the
Lurgi process. Applicability of the control tech-
nologies does not mean that the control technol-
ogy can be duplicated from the coke oven design
to the coal conversion application. In general,
the composition, flow rate, temperature, and
pressure of the specific coal conversion system
wastes will not be identical to the coke oven
case. These differences, however, must be taken
into consideration during the design of the spe-
cific controls. Design information or scale-up
factors in comparison to coke oven application
should be developed through laboratory or,
pilot-scale testing with actual coal conversion
wastes to determine the system design and to
develop its costs.
The health effects of coke oven emissions
were recently assessed.3 The summary findings
are:
• Exposure to coke oven emissions provides
an elevated risk for cancer and nonmalig-
nant respiratory diseases to coke oven
workers and an increased risk among lightly
exposed workers (nonoven workers in the
coke plant).
• The general population, which includes the
young, the old, and the infirm in the vicinity
of a coke oven plant, should be considered
more susceptible than the workers, especial-
ly for development of chronic bronchitis.
• Lightly exposed workers are exposed to
emissions about 100 times more concen-
trated than the people living in the imme-
diate vicinity of a coke plant. Since the peo-
ple living in the immediate vicinity of a coke
plant experienced an elevated risk for can-
cer and nonmalignant respiratory disease, it
is reasonable to assume that levels as high
as 1 percent of those to which lightly ex-
posed workers are subjected could cause an
increased risk to the general population.
• Coke oven emissions contain an array of
identified carcinogens, irritants, particulate
matter, trace elements, and other chemicals.
The toxic effects observed in both humans
and animals are greater than the effects that
can be attributed to any individual compo-
nent. Thus, "coke oven emissions" as a
whole should be considered the toxic agent.
Since coke oven and coal conversion systems
have many of the same hazardous waste compo-
nents, such as H2S, CO, C02, hydrocarbons, and
575
-------
polynuclear aromatics, there is a potential occu-
pational health hazard to coal conversion plant
workers and the general population in the
vicinity of the plant. Indications are that the
population living within a radius of 15 km from a
coke oven plant would suffer the maximum po-
tential exposure risk. Many of the new control
technologies under development, especially
those for fugitive emissions control, should re-
sult in significant removal of these hazardous
pollutants for the coke oven industry. The ap-
plicability of these control technologies to coal
conversion processes is not altogether clear, ex-
cept, perhaps, in the coal preparation and han-
dling areas.
For the most part, the types of emissions ex-
pected from the coal conversion plant would not
only be less concentrated, but also far less toxic.
The fumes and particulate matter from the coke
ovens themselves, and the subsequent pushing
and quenching operations, account for the major
pollutants which result in the majority of health
hazards encountered in the coke oven industry.
The particulate and fugitive emission problems
in the coal conversion industry are expected to
be several orders of magnitude lower than are
presently found in the coke oven industry. The
coal conversion industry will need to continue to
be vigilant to avoid health problems similar to
those found in the coke oven industry.
CONCLUSIONS
Acid gas and tail gas treatment processes
used in both the petroleum refinery and in the
coke oven plant are adaptable to coal conversion
processes. The efficiency for most of these proc-
esses would decrease by the C02 levels which
are expected to be higher for coal conversion
processes. The current technology for removing
ammonia, phenol, cyanides, hydrocarbons, oil,
and grease from vapor and liquid streams all
seem applicable to coal conversion plants.
Much of the fugitive emission control technol-
ogy, particularly that found in the coke oven in-
dustry, would have applications in the coal con-
version industry. Many new developments are
emerging in this field pertaining to the coal
pyrolysis and quenching operations which
would positively impact on the coal conversion
industry.
Wastewater treatment involving biological
action appears to be useful in coal conversion,
but the biodegradability of coal conversion
waste components has not been established.
Carbon adsorption of organic components from
wastewaters may be necessary for many waste-
water streams, especially those containing
polynuclear aromatic compounds.
Sludge, oily solid waste, and other solid waste
disposal techniques now in use seem currently
applicable and satisfactory for the control tech-
nology needed in the coal conversion.
The fate and the composition of trace organic
compounds (e.g., benzo(a)pyrene and polynu-
clear aromatic compounds) and inorganic com-
ponents (e.g., arsenic, lead, and selenium) are
presently not well known for coal conversion
processes. Whether control strategies will need
to be developed for these components remains
to be seen.
Even though many of the control technologies
appear applicable to coal conversion wastes, ad-
ditional testing will be required to confirm the
applicability for large-scale design and cost
estimation. It is expected that additional
development of control technologies will be
needed for the coal conversion industry.
REFERENCES
1. Ghassemi, M., et al. Applicability of Petro-
leum Refinery Control Technologies to Coal
Conversion. TRW, Inc. Redondo Beach,
Calif. EPA-600/7-78-190 (NTIS PB 288630).
October 1978.
2. Hossain, 8. M., et al. Applicability of Coke
Plant Control Technologies to Coal Conver-
sion (draft). Catalytic, Inc. Philadelphia, Pa.
U.S. Environmental Protection Agency.
Contract Number 68-02-2167. December
1978.
3. Stellman, J. M. An Assessment of the
Health Effects of Coke Oven Emissions (ex-
ternal review draft). U.S. Environmental
Protection Agency, Office of Research and
Development, Office of Health and Ecologi-
cal Effects. Washington, DC. April 1978.
576
-------
TECHNICAL REPORT DATA
(Please read Imttructions on the reverse before completing)
REPORT NO.
2.
. RECIPIENT'S ACCESSION NO.
. TITLE AND SUBTITLE
Symposium Proceedings: Environmental Aspects of Fuel
Conversion Technology, IV (April 1979, Hollywood, PL)
REPORT DATE
PERFORMING ORGANIZATION CODE
. AUTHOR(S)
Franklin A. Ayer and N. Stuart Jones (Compilers)
8. PERFORMING ORGANIZATION REPORT NO.
. PERFORMING ORGANIZATION NAME AND ADDRESS
Research Triangle Institute
P.O. Box 12194
Research Triangle Park, NC 27709
10. PROGRAM ELEMENT NO.
EHE623A
11. CONTRACT/GRANT NO.
68-02-3132, Task 1
2. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final: 9/78 - 9/79
14. SPONSORING AGENCY CODE
EPA/600/13
5. SUPPLEMENTARY NOTES IERL-RTP project officer is T.
2851.
Kelly Janes, Mail Drop 61, 919/541-
6. ABSTRACT
The proceedings document presentations made at the symposium on Environmental
Aspects of Fuel Conversion Technology. The symposium acted as a colloquium for
discussion of environmentally related information on coal gasification and liquefactioi
The program included sessions on program approach, environmental assessment, and
control technology development. Process developers, process users, research scientist;
and state and federal government officials participated in this symposium, the fourth
to be conducted by IERL-RTP on the subject since 1974.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
COSATI Field/Group
Pollution
Fuels
Coal Gasification
Liquefaction
Assessments
Pollution Control
Stationary Sources
Fuel Conversion
Environmental Assessment
13B
2 ID
13H
07D
14B
18. DISTRIBUTION STATEMENT
Release to Public
18. SECURITY CLASS (Thtt Report)
Unclassified
21. NO. OF PAGES
582
30. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (»-7J) 577
" a U.S. GOVERNMENT PRINTING OFFICE: 1979-640.Ol? 39m+REGIONNO_4
------- |