EPA-650/2-74-052
STUDY OF POTENTIAL PROBLEMS
AND OPTIMUM OPPORTUNITIES
IN RETROFITTING INDUSTRIAL PROCESSES
TO LOW AND INTERMEDIATE
ENERGY GAS FROM COAL
by
D. Ball, G. Smithson,
R. Engdahl, and A. Putnam
Battelle—Columbus Laboratories
505 King Avenue
Columbus, Ohio 43201
Contract No. 68-02-1323
Task 1
ROAP No. 21ADD-30
Program Element No. 1AB013
EPA Project Officer: William J. Rhodes
Control Systems Laboratory
National Environmental Research Center
Research Triangle Park, North Carolina 27711
Prepared for
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, B.C. 20460
May 1974
-------
This report has been reviewed by the Environmental Protection Agency
and approved for publication. Approval does not signify that the
contents necessarily reflect the views and policies of the Agency,
nor does mention of trade names or commercial products constitute
endorsement or recommendation for use.
11
-------
TABLE OF CONTENTS
Page
SUMMARY vii
CONCLUSIONS X
RECOMMENDED AREAS FOR CONSIDERATION xii
LIST OF ABBREVIATIONS
LIST OF CONVERSION FACTORS xiv
INTRODUCTION 1
OBJECTIVE 2
I. GASIFICATION AND GASIFICATION SYSTEMS 3
1-1. Current Commercial Gasification Processes 3
Lurgi 3
Koppers-Totzek 7
Winkler 9
Wellman-Galusha 11
1-2. Fuel Gas Characteristics 14
1-3. Comparison of Commercial Gasification Processes . . 17
Thermal Efficiency 17
Steam-Raising Capacity 17
Fuel Requirements 20
Unit Capacities 21
Operational 21
II. FUEL GAS CLEANUP 22
III. CONSIDERATIONS IN UTILIZING LOW AND INTERMEDIATE-ENERGY
GAS FROM COAL IN INDUSTRY 26
III-l. Comparison of Characteristics of Low and Inter-
mediate Gas to Natural Gas 26
III-2. Environmental Considerations 30
IV. UTILIZATION OF LOW AND INTERMEDIATE ENERGY GAS IN
INDUSTRY 36
IV-1. Criteria Used for Evaluating Ease of Retrofit
of Industrial Processes 36
TV-2. Classification of Common Industrial Processes
According to Their Ease of Retrofit 38
iii
-------
TABLE OF CONTENTS
(Continued)
Page
IV-3. Utilization of Gas From Coal in Industrial
Boilers 40
IV-4. Retrofit of Gas Turbines to Low- or Intermediate-
Energy Gas From Coal 43
V. ANALYSIS OF POTENTIAL MAJOR INDUSTRIAL USERS OF LOW AND
INTERMEDIATE ENERGY GAS FROM COAL 45
V-l. Chemical and Allied Products Industry (SIC 28). . . 45
V-2. Primary Metals (SIC 33) 48
Basic Iron and Steel 48
Ferrous Foundries (SIC 332) 52
Primary Nonferrous Metals (SIC 333) 54
V-3. Petroleum and Coal Products Industry (SIC 29) ... 60
V-4. Stone Clay and Glass Products (SIC 32) 65
Glass 66
Cement 67
Lime 70
Ceramics 71
VI. UTILIZATION OF LOW OR INTERMEDIATE ENERGY GAS FROM COAL
IN ELECTRIC POWER GENERATION 73
VII. REFERENCES 78
APPENDIX A
INTERCHANGEABILITY OF FUELS WITH RESPECT TO THE COMBUSTION
PROCESS A-l
SUMMARY A-l
CONCLUSIONS A-2
LIST OF SYMBOLS A-3
INTRODUCTION A-4
Presentation of Data A-5
Discussion of the Data A-12
Premix Burners A-12
Delayed Mixing Burners A-17
Nozzle Mix Burners. A-22
References A-25
iv
-------
TABLE OF CONTENTS
(Continued)
APPENDIX B
Page
FUEL GAS CLEANUP PROCESSES B-l
Sulfur Removal B-l
Waste or By-Product Recovery of Sulfur B-ll
Particulates Control ... B-12
Nitrogen Oxides B-13
References • B-14
APPENDIX C
DETERMINATION OF RELATIONSHIPS FOR SULFUR EMISSIONS GIVEN IN
FIGURES B-l THROUGH B-4 C-l
APPENDIX D
CALCULATION OF RELATIVE GAS PRESSURE DROP D-l
LIST OF TABLES
Table 1. Typical Gas Compositions for Commercial Gasifiers . 15
Table 2. Typical Concentrations of Nonhydrocarbon
Impurities in Coal Gas 16
Table 3. Typical Unit Characteristics for Commercial
Gasifiers 18
Table 4. Steam and Oxygen Requirements for Oxygen Blown
Gasifiers 19
Table 5. Comparison of Volumes of Fuel Gas to Natural Gas. . 27
Table 6. Energy Consumption in Chemical and Allied
Products Industry 46
Table 7. Energy Use in Basic Iron and Steel in 1971 51
Table 8. Melting Furnaces in the Ferrous Foundry Industry
in 1965 53
Table 9. Energy Use in the Ferrous Foundry Industry (1964) . 53
Table 10. Energy Requirements of the Primary Nonferrous
Metals Industry (1971) 55
Table 11. Process Energy Requirements for Copper Manufacture
in 1971 56
Table 12. Process Energy Requirements for Lead Manufacture
in 1971 58
-------
LIST OF TABLES
(Continued)
Page
Table 13. Process Energy Requirements in Zinc Manu-
facture in 1971 59
Table 14. Process Energy Requirements in Aluminum
Manufacture in 1971 61
Table 15. Energy Usage in the Petroleum and Coal Products
Industry 61
Table 16. Energy Requirements of Various Petroleum Refining
Processes per Barrel of Feed 65
Table 17. Energy Requirements in the Stone, Clay, and Glass
Products Industry (1971) 66
Table 18. Process Energy Requirements in Glass Manufacture
in 1971 68
Table A-l. Fuel Consumption and Thermal Properties A-6
Table A-2. Fuel Stability Factors A-7
Table B-l. Conventional Processes for Fuel Gas Desulfuri-
zation ..... B-7
LIST OF FIGURES
Figure 1. Lurgi Gasifier , . . . 4
Figure 2. Koppers-Totzek Gasifier 8
Figure 3. Winkler Gasifier 10
Figure 4. Wellman-Galusha Gasifier ... 12
Figure 5. Typical Gasification Scheme With Cold Gas Cleanup . 31
Figure A-l. Flash Back Velocity Gradient Versus Mixture Ratio. A-9
Figure A-2. Critical Heat Supply Rate Per Unit Volume Versus
Mixture Ratio A-10
Figure A-3. Flash Back Velocity Gradient Times High Heating
Value Versus Mixture Ratio A-11
Figure A-4. Premix Burner A-13
Figure A-5. Delayed Mixing Burners A-19
Figure A-6. Nozzle Mixing A-21
Figure B-l. SO* Emission Versus Percent Sulfur in Coal for
Different Gasification Efficiencies B-2
Figure B-2. Sulfur Removal Required to Meet Standard Versus
Percent Sulfur in Coal B-3
Figure B-3. H2S in Fuel Gas Versus Percent Sulfur in Coal for
Different Gas High Heating Values B-4
Figure B-4. SC^ Emissions Versus ^S Concentration in Fuel Gas
for Different Gas High Heating Values B-5
vi
-------
SUMMARY
This study involved an analysis of the state-of-the-art of
commercial coal gasification processes and fuel gas cleanup processes
applicable to coal gasification. The study also included an analysis of
the potential problems in retrofitting existing industrial processes to
the use of low or intermediate energy gas from coal.
It was found that presently there are four gasification processes:
Lurgi, Koppers Totzek, Wellman-Galusha, and Winkler that have been com-
mercially proven for gasifying coal and for which there exists a current
manufacturing capability. It is doubtful, however, that any of these four
processes could be produced in significant numbers for industrial application
at the present time due to limited manufacturing capacity.
Only one of the four processes, the Wellman-Galusha, has been
used commercially in this country for making low energy gas, and presently
there are only a few known cases in which these gasifiers are still used
either for standby service or as a primary source of fuel. The other three
processes are manufactured in Germany, and though they have had little or
no application in this country, they have been used extensively in other
parts of the world for producing town gas (intermediate energy fuel gas)
or synthesis gas for production of ammonia and higher hydrocarbon fuels.
These processes are presently receiving considerable attention in this
country, however, for both the production of synthesis gas to make synthetic
natural gas (SNG) and as a direct source of fuel gas.
It was also found that commercial processes have been developed
for reducing particulate and sulfur levels in the fuel gas well below
those concentrations that would be considered environmentally acceptable
for a coal-based fuel. At the present time, however, no cleanup processes
are available for cleaning the fuel gas at temperatures above 1000 F, and
thus, increasing the overall thermal efficiency of the process through
utilization of the sensible heat in the gas; but a hot gas cleanup
ability is not felt to be critical to make low or intermediate energy
from coal attractive for industrial use. Much of the sensible heat in the
gas can be recovered in a wasteheat boiler while cooling the gas for cleanup,
and the steam generated can be used in the gasification process itself
vii
-------
or for other processes in the industrial plant. Also, distribution of a cold
gas to various processes in a given industry would be easier than distribution
of a hot gas.
An analysis of potential problem areas in retrofitting existing in-
dustrial processes revealed the following:
• From a combustion standpoint, designing or modifying
burners to allow the substitution of low or intermediate
energy gas for oil or natural gas will probably not be a
serious problem, though at the present time, the wide variety
of "off-the-shelf" burners used in industrial processes are
not available for lower energy gases; development of such
burner lines, in many cases, may require 1 to 2 years of
work. It is felt, however, that the development of adequate
burners would be easier for intermediate energy gas of around
300 Btu/scf (2664 kcal/m ) from oxygen-blown gasifiers than
o
for low energy gas of around 150 Btu/scf (1332 kcal/m ) from
air-blown gasifiers.
• The major problems are expected to lie in handling the in-
creased volumes of fuel gas and combustion products necessary
to obtain comparable energy input to that supplied by oil or
natural gas. For low or intermediate energy gas, larger fuel
distribution mains and burner nozzles will be required over
those for natural gas. The higher volume of combustion prod-
ucts will require induced or forced draft fans to operate at
higher pressures than with oil or natural gas. Substitution
of low or intermediate energy gas from coal in existing direct
coal-fired processes should not present serious problems in
this area, since processes designed for coal usually have more
generous gas-flow passages to allow for ash buildup and, also,
less excess air should be necessary in switching to a gaseous
fuel, resulting in less of a difference in the volumes of the
products of combustion.
viii
-------
• The amount of air pollution in the form of S0_ and particu-
lates generated by utilizing low or intermediate energy gas
from coal will depend heavily on the cleanup processes.
With existing commercial cold gas cleanup processes, however,
it should be possible to reduce S0_ emissions well below the
environmental standard for coal-or oil-fired processes. Also,
particulate levels should be reduced to near zero when the
common cleanup combination of particulate removal followed by
wet scrubbing for H-S is used. NO emissions will depend a
great deal on the combustion process, though they should in
general be less than or, at most, equal to those from two-
stage coal combustion.
• Solid and liquid waste from the gasification and gas cleanup
process represent a potential environmental problem area.
The recovery of elemental sulfur represents a potentially
marketable product, though due to the large amount of sulfur
that would be generated there is good reason to believe it would
probably represent a waste to be disposed of. Also, nearly all
the current processes involve wet scrubbing with a variety of
sorbents, resulting in liquid waste streams that would necessitate
cleanup prior to disposal.
Based on their general characteristics, a wide variety of industrial processes
were found to be potentially attractive for retrofit to low or intermediate
energy gas from coal. Criteria were set up for evaluating these process
characteristics and their effect on ease of retrofit. Various processes were
then classed in three groups: those easiest to retrofit and requiring minimal
testing and modification; those in which retrofit is an attractive alternative
but where modification or extensive testing would be necessary; and those pro-
cesses in which retrofit is unattractive. An analysis was made of the four major
energy-consuming industrial sectors (according to the latest U.S. Census of
Manufacturers) to determine the potential for use of low or intermediate energy
gas from coal and to estimate the potential amount of energy substitution
possible.
ix
-------
CONCLUSIONS
The conclusions from this background study into the potential for
utilization of low or intermediate energy gas from coal in industrial processes
are as follows:
• Commercial gasification and gas cleanup systems currently
exist for the gasification of coal to an environmentally
acceptable gaseous fuel for industrial process use.
• Obtaining suitable combustion of the lower energy gases
from coal in processes designed for higher energy fuels
(such as oil and natural gas) should not be a serious
problem, though a full range of "off the shelf" replace-
ment burners presently do not exist for the wide range of
process requirements that exist, and development of such
a line may require from 1 to 2 years.
• The major problems in retrofit will involve the handling
of the increased volumes of fuel gas and products of
combustion resulting when low or intermediate energy gas
from coal is used to supply the same heat input as oil
or natural gas.
• Emissions of sulfur dioxide and particulates should be con-
trollable with existing cleanup processes within the current
environmental requirements for oil- and coal-fired processes
and should in both cases be much less. Emissions of other
potential pollutants must be assessed as systems are utilized.
• Liquid and solid waste generation from gasification could
be a problem in comparison with use of clean fuels such as
oil and natural gas, but this generation of waste may be less
of a problem than that from the alternative of tail gas cleaning
of flue gases following the combustion of dirty fuels (such as
high-sulfur oil or coal).
• There is a wide range of industrial processes representing a
significant portion of energy use in industry that should be
applicable to retrofit to low or intermediate energy gas from
coal, providing that suitable burners and gasification equipment
are available.
x
-------
3. Tnough not all processes require a pressurized fuel gas,
pressurized gasification is attractive in that it allows greater gas
production with smaller equipment and requires the least energy in com-
pression to provide a pressurized gas for distribution or use in gas turbines.
Currently, the Lurgi is the only commercial process capable of pressurized
gasification, and it uses a complex system of lock hoppers requiring
frequent maintenance and downtime for coal feeding and ash removal. The
developing of a less complex pressurized gasification process capable of
gasifying a wide variety of coals would be highly attractive for industrial
use.
4. Another area for consideration in industrial use of coal
gasification as a fuel source is gas cleanup. Almost all the processes
in Table 5 involve liquid scrubbing. The two dry processes, iron sponge
and activated carbon, are mostly applicable where extremely low concen-
trations of sulfur are demanded and where the initial sulfur concentration
is relatively low. With high inlet sulfur concentrations, both processes
require frequent sorbent replacement. For industrial use, a cleanup system
capable of reducing relatively high H^S concentrations in the gas to moderate
concentrations that are environmentally acceptable with a minimum of expense
and waste would be attractive. It is also desirable that the cleanup
system be specific for H?S, as removal of other constituents such as C09
requires unnecessary extra cleanup capacity and complicates sulfur recovery.
For many industrial applications, an unclean gas is not considered especially
valuable. If the steam generated by a waste heat boiler in cooling the hot
raw gas to a temperature suitable for current cold gas cleanup processes can
be utilized, then much of the sensible heat in the gas can be conserved. The
resulting difference in thermal efficiency between the process with a hot clean
gas and a cold clean gas is therefore less than would result if the sensible
heat were not conserved and the greater ease and economy with which the cold
gas can be handled and distributed may make it more attractive in some cases.
5. In addition to product gas cleanup, there are emission and effluent
streams that require characterization in order that environmental assessments
may be made of their control adequacy.
xi
-------
RECOMMENDED AREAS FOR CONSIDERATION
1. Though it: was concluded from the study presented in Appendix A
and from discussions with burner manufacturers that development of adequate
burners for low or intermediate energy gas from coal would probably not
pose serious technical problems, these burners, in general, are not
currently available for the wide range of process requirements. Burners
that used to be used for low energy gas from coal and burners used in some
industries today for low energy waste gases may not in many cases have
the required characteristics of flame stability and safety demanded in
many processes. Prior to any significant application of coal gasification
in industry, a wide selection of burner types and performance will have
to be available from the manufacturer. Development of new burner lines
requires considerable time and may require 1 to 2 years of design and
testing before their becoming available as catalog items.
2. In conjunction with Item 1, there presently are very little
data available on the burning characteristics of various gas compositions
obtained in coal gasification. It would be useful in designing burners
to have accurate data on some of the basic combustion parameters (such as
flame velocity and temperature) for different gas mixtures. Other useful
areas of knowledge that may not evolve from normal burner development
programs are data on emissions of NO and noise« An independent study
determining these various characteristics of typical gas mixtures produced
in coal gasification should prove useful in burner design and in estimating
some of the potential environmental effects of burning low or intermediate
energy gas from coal. It would also provide a sounder basis for optimal
matching of gasification systems and processes, especially relating to
the tradeoffs of using more expensive 300 Btu/scf gas from oxygen-blown
producers versus cheaper 150 Btu/scf gas from air-blown producers.
xii
-------
LIST OF ABBREVIATIONS
scf - Standard cubic foot as measured at 60 F and 30 inches
of mercury (American Gas Association standard temperature
and pressure)
psia - Pounds pressure per square inch absolute
psig - Pounds pressure per square inch gage
Real - Kilocalorie
Kg - Kilogram
Ib - Pound mass
Btu - British Thermal Unit
m - Meter
mm - Millimeter
C - Degrees centigrade
F - Degrees Fahrenheit
N - Newton
ppm - Parts per million by volume
HHV - High heating value of fuel including the latent heat
of vaporization of water formed during combustion
Ng - Gasification efficiency. Energy content of product
gas divided by energy content of coal
-------
LIST OF CONVERSION FACTORS
Btu (@ 60 F) x 0.252 = Kcal
ft x 0.3048 = m
in. x 25.4 = mm
F subtract 32 x 0.555 = C
Ib x 0.453 = Kg
Ib x 0.453 = Kg
scf (@ 60 F & 30 in. Hg) x 0.0284 = m3 (@ 15.5 C & 762 mm Hg)
Btu/scf x 8.88 = Kcal/m3 (@ 15.5 C & 762 mm Hg)
lb/106 Btu x 1.798 = Kg/106 Kcal
ton x 1.104 = metric ton
xiv
-------
INTRODUCTION AND OBJECTIVE
-------
STUDY OF POTENTIAL PROBLEMS AND OPTIONAL
OPPORTUNITIES IN RETROFITTING INDUSTRIAL PROCESSES
TO LOW AND INTERMEDIATE ENERGY GAS FROM COAL
by
D. A. Ball, G. R. Smithson, R. B. Engdahl, and A. A. Putnam
INTRODUCTION
In recent years the United States has become increasingly dependent
on oil and natural gas as a source of clean, convenient energy. According to
the 1972 Census of Manufacturers , natural gas supplied 59 percent of the
energy needs of American industry in 1971 and oil contributed an additional
11 percent. Many segments of industry have become dependent on a source of
clean, convenient gaseous or liquid fuel, and with the growing scarcity and
higher costs of natural gas and oil, low and intermediate energy gas from coal
may be a viable alternative.
The use of low energy gas from coal (or producer gas as it is some-
times called) was once widespread in industry; in fact, it was the first
gaseous fuel used in the iron and steel industry for firing open-hearth fur-
(2)
naces between 1860 and 1920 , when by-product coke oven gas displaced it.
In 1920 there were reportedly over 11,000 gas producers supplying gaseous fuel
for a wide variety of industries from brick plants to bakeries. With the ad-
vent of inexpensive natural gas and oil, however, the incentive for gasifying
coal decreased until today there are but a few bituminous coal gas producers
in this country and these have largely been placed in standby positions.
This study is a general analysis of the potential problem areas in
retrofitting existing industrial processes to the use of low or intermediate
energy gas from coal. Also included is an analysis of the current state of the
art of commercial coal gasification systems and gas cleanup systems applicable
to industrial use.
-------
OBJECTIVE
The objectives of this study are to analyze and provide background
information on the status of current commercial coal gasification and pertinent
gas cleanup technology and to evaluate the potential for utilization of low or
intermediate energy gas from coal in existing industrial processes, specifically
evaluating the expected problems in retrofit.
-------
SECTION I
GASIFICATION AND GASIFICATION SYSTEMS
-------
GASIFICATION AND GASIFICATION SYSTEMS
Low-energy gas from coal, often called producer gas, is made by supply-
ing a sufficient mixture of air and steam to coal for partial combustion result-
ing in a gas rich in CO, H , and possibly some hydrocarbons giving the gas
its characteristic heating value. The air-steam mixture generally yields a gas
with about 50 percent nitrogen content and a heating value of around 100 to 200
o
Btu/scf (888 to 1780 kcal/m ). In most gasification processes oxygen can be
substituted for air, yielding a gas essentially free of nitrogen with a heating
o
value of around 300 Btu/scf (2670 kcal/m ). Oxygen-blown gasifiers are frequent-
ly used in other parts of the world for the production of town gas, which is used
in much the same way that natural gas is used in this country. This gas is also
frequently used for synthesis gas in the production of higher hydrocarbon prod-
ucts, and ammonia.
1-1. Current Commercial Gasification Processes
Presently, there are four gasification systems that can be considered
commercially viable in the sense that they have been proven commercially for the
gasification of coal and are available today through commercial dealers. These
four systems are the
(1) Lurgi
(2) Koppers Totzek
(3) Wellman-Galusha
(4) Winkler.
None of these processes are in widespread use in the United States at the present
time due to the past abundance of cheap natural gas and oil, fuels which today
are becoming increasingly scarce.
The Lurgi gasifier, shown schematically in Figure 1, was first developed
on lignite in 1931 by Lurgi Gesellschaft fur Warme und Chemetechnik mbH, Frankfurt,
(3)
West Germany. Lurgi gasifiers have been widely used in other parts of the
-------
IGNITION
COKE
INLET
COAL BUNKER
AL FEED SYSTEM
SCRUBBING
COOLER
CAS
ORATE.
DRIVE r
STEAM.
OXVGEN
»- DEPRESSURIZING VESSEL
Lurgi (Lurgl Gesellschaft fur Warme und Chemotechnik mbH)
Characteristics :
-Pressurized process (300 to 500 psig) (2070 to 3450 N/m2)
-Fixed bed oj eration with rotating grate and coal-feed distributor,
countercurrent gas flow
-Coal feed and ash removal through lock hoppers
-Typical gasifier outlet temperature 900-1100 F (480 to 590 C)
-Uses double screened non-caking or possibly mildly caking coal
(3-30 mm) (1/8 to 1-1/4 in.)
FIGURE 1. LURGI GASIFIER
(6)
-------
world, primarily utilizing an oxygen-steam gasifying medium for the production
of town gas and synthesis gas for hydrocarbon products; the most notable appli-
cation is the largest coal gasification plant in the world at Sasolburg,
(4)
South Africa. ' Lurgi gasifiers also have used air-steam gasifying mediums to
2
produce a low-Btu gas of around 180 Btu/scf (1600 kcal/m ). Currently, Lurgi
is beginning to operate a demonstration-scale gas-turbine power plant in Lunen,
Germany with a battery of air-blown gasifiers.
The Lurgi gasifier utilizes a fixed-bed pressurized process with the
gasifier operating under approximately 20 atmospheres of pressure. Pressuriza-
tion requires coal feed and ash removal to be handled by lock hoppers at the
top and bottom of the unit. The lock hoppers allow coal and ash to be processed
in and out of the unit intermittently without interrupting the process. The
gasifying medium is introduced under pressure through a rotating grate in the
bottom of the unit and flows countercurrently through the coal bed, forming
the fuel gas which is taken off at the top of the unit. The rotating grate also
allows ash to fall into the bottom lock hopper for removal and serves to prevent
agglomeration of the coal to insure an evenly distributed flow of the gasifying
medium through the bed. At the top of the gasifier is a rotating horizontal
arm that distributes the coal from the feed hopper evenly over the surface of
the bed. The gasifier body itself consists of a water-cooled double-shelled
cylinder with the water flowing in the jacket formed by the annulus of the two
shells. Steam formed by the heat absorbed in the cooling water is capable of
supplying approximately 20 percent of the process steam requirements of the
unit. The overall thermal efficiency of the process including the sensible
heat in the hot fuel gas is 91 percent.
As is typical of fixed-bed processes, the Lurgi process is sensitive
to caking coals. The process was developed with lignite as a fuel and in
nearly all commercial installations has been restricted to lignite or other
weakly caking fuels such as anthracite and sub-bituminous coals. Lurgi claims
to be capable of specifically designing a gasifier to gasify caking coals
typical of the Eastern United States and is currently conducting trials at a
town gas plant in Westfield, Scotland, on a variety of American coals. The
trials are still in progress and, according to the most recent information,
the highest caking coal gasified was an Illinois No. 6, a milding caking coal.
Trials were conducted in 1963 at Dorsten, West Germany, with highly caking
Pittsburgh bed coal on a Lurgi gasifier with a specially designed water-cooled
-------
rotating agitator arm just below the surface of the coal bed. Those tests were
inconclusive, however, as the longest run was only 12 hours. The relatively
short tests were due to the small amount of coal sample available, and the coal
ash had to be increased from 7 percent to 30 percent by the addition of ash from
the noncaking coal normally gasified at the site. To date, Lurgi has not pub-
lished information demonstrating conclusively the ability of its process to
gasify highly caking coal.
In addition to the coal caking properties, operation of the Lurgi
process also is sensitive to coal size. Normally, the coal is crushed and
double screened to sizes from roughly 1/8 inch (3 mm) to 1-1/4 inch (30 mm).
Some small amount of undersize (around 7 percent) can be tolerated, though the
presence of fines tends to disrupt the gasification process.
Due to the large number of moving parts and pressurized operation, the
Lurgi process is mechanically more complex than the other three gasifiers con-
sidered. Seals and packing glands around the lock hoppers and rotating shafts
extending through the gasifier shell require a sophisticated maintenance pro-
gram to insure proper gasifier operation. The Lurgi reactor also has a rela-
tively large number of operating variables. Both the grate and coal distri-
butor arm rotational speed, as well as the coal feed rate, ash removal rate,
and air-steam or oxygen-steam ratio can be varied to adjust to changing demands
on the gasifier. While this may make the gasifier more flexible in some situa-
tions, it also makes it more difficult to operate, requiring highly trained
personnel.
The size of fixed-bed gasifiers is somewhat limited by the ability
to make large rotating parts that will function effectively inside the gasi-
fier. Lurgi presently is capable of making gasifiers 12 feet (3.68 m) in
diameter, but is planning to produce a unit 16 feet (4.88 m) in diameter.
Approximately 7 to 9 billion Btu per day* (1.7 to 2.3 billion kcal per day)
can be produced in the 12-foot unit. In the 16-foot unit, 12 to 16 billion
Btu's per day (3 to 4 billion kcal per day) can be produced. The Lurgi gasi-
fier can operate satisfactorily down to 50 percent of its design load.
2 sz\
* Based on a gasification rate of 300 Ib coal/hr ft (cross section) (1455
kg coal/hr-m^).
-------
Koppers-Totzek
The Koppers-Totzek process shown in Figure 2, utilizes pulverized
coal in suspension with oxygen and steam to produce an intermediate energy gas
3
of around 300 Btu/scf (2660 kcal/m ). The gasifier was first commercially
/Q\
operated in Finland in 1952V ' by H. Koppers Company of Essen, West Germany.
Currently, there are 16 commercial plants that have been built or are under
construction around the world primarily for the production of a synthesis gas
for ammonia manufacture.
The design of the gasifier is relatively simple, and it operates at
nearly atmospheric pressure. Either two or four burners diametrically opposed
on the circumference of the gasifier fire a mixture of oxygen, steam, and
pulverized coal into the center of the gasifier, where the coal is gasified
under slagging conditions at around 3300 F (1850 C). About 50 percent of the
ash is carried away in the gas stream and the remainder drops out as slag into
a water bath at the bottom of the unit. Koppers has conducted tests firing the
gasifier with air instead of oxygen and found that the air had to be preheated
to at least 1800 F (982 C) to maintain slagging conditions. Oxygen-enriched
air can be successfully used in the gasifier with a minimum of 32 percent
oxygen. In subsequent studies, Koppers has concluded that firing with oxygen
instead of air is economically attractive in spite of the necessity of an
oxygen plant, especially for larger units. This is due to the reduced size
necessary for the gasifier, waste heat boiler, and gas cleanup equipment.
The Koppers-Totzek gasifier is the only one of the four discussed
here that is not sensitive to the caking properties of the coal. The coal is
pulverized, with 90 percent passing a 200-mesh screen, fed to the gasifier,
and then gasified in suspension. The coal is normally dried to less than 1
percent moisture to facilitate pulverizing, though up to 8 percent moisture
could be tolerated.
The fuel gas leaves the gasifier at about 2700 F (1480 C) and is
passed through waste heat boilers prior to cleanup. Due to the high temperature
of the gas, it contains significant sensible heat--up to 20 percent of the
(9)
original heating value of the coal. Recovery of some of this heat in a
waste heat boiler makes the gasifier a net exporter of steam, and it is capable
-------
WASTE HEAT
BOILER -
HIGH PRESSURE
STEAM
-GAS OUTLET
COAL
Koppers Totzek (Koppers Company, Pittsburgh, Pennsylvania)
Characteristics:
-Atmospheric pressure process
-Pulverized tangentially fired coal feed
-Typical gasifier outlet temperature 2000-2700 F (100 to 1500 C)
-Can use caking of non-caking coal
-90 percent past a 200 mess sieve
FIGURE 2. KOPPERS TOTZEK GASIFIER
(6)
-------
/g\
of producing roughly 150 pounds of high-pressure steam per million Btu
(270 kg/million kcal) of fuel gas produced. The overall thermal efficiency
of the Koppers process including the sensible heat in the hot fuel gas is
86 percent.
Due to the uncomplicated mechanical design of the gasifier, operation
is relatively simple and maintenance minimal. Commercial installations have
demonstrated the capability of being on line 95 percent of the time. Also, shutdown
can be accomplished instantaneously and the unit can achieve full gas production
within 30 minutes from a standby condition, with the refractory lining heated
by gas or oil torches.
Currently, Koppers makes both two-burner and four-burner units. The
two-burner unit is capable of gasifying up to 400 tons per day (352 metric
tons/day) of coal--equivalent to about 7 to 9 billion Btu per day (1.7 to 2.3
billion kcal/day) of hot gas--and the four-burner unit is capable of gasify-
ing up to 850 tons per day (780 metric tons/day) of coal—equivalent to 14 to
18 billion Btu per day (3.5 to 4.5 billion kcal/day). Koppers clainT ' that
any two opposing burners can be reduced in load 35 percent. In addition, on
the four-burner unit two burners can be completely shut off for a total load
reducing capability of up to 65 percent.
Winkler
The Winkler gasifier, shown in Figure 3, is an atmospheric-pressure
fluidized-bed process first used commercially in Germany in 1926 for the produc-
tion of fuel gas. Since then, 16 gasification plants, totaling 36 gasifiers,
have been designed and built by Bamag Verfahrenstechnik GmbH, a German affiliate
of Davy Power Gas, Inc.
The Winkler process can be fired with either oxygen and steam--making
o
an intermediate energy gas of about 300 Btu/scf (2670 kcal/m )--or with air and
o
steam--making a low energy gas of about 120 Btu/scf (1060 kcal/m ). The gas
leaves the gasifier at 1500 to 1900 F (810 to 1040 C) and, similar to the Koppers
product, it contains considerable sensible heat--up to 10 percent of the heating
value of the coal. If waste heat boilers are used to recover part of this heat,
the Winkler gasifier can be a net exporter of high-pressure steam producing
around 80 pounds per million Btu (144 kg/million kcal) of hot fuel gas.^*-'
Coals with higher reactivity, such as lignite, yield lower exit gas
temperatures and, hence, higher efficiencies than do low reactivity bituminous
-------
10
d
PURGE I INERT CAS LINES
W-LTLJV
FUEL BUNKER
HATER COOLE:
SHAFT
RATCHET OR IV
STEAK
WATER JACKETED
SCREW CONVEYOR
GAS TO DUST
COLLECTOR
WASTE HEAT
STEEL SHELL
REFRACTORY LINING
SCRAPER FOR ASH
REMOVAL
OXYGEN OR*
ENRICHED AIR
RATCHET ORIVE
MATER COOL SHAFT
ASH
RECEIVER
WATER JACKETED
SCREW CONVEYOR
Winkler (Davy-Power Gas, Lakeland, Florida)
Characteristics:
-Atmospheric pressure process
-Fluidized bed operation
-Typical gasifier outlet temperature 1600 F (870 C)
-Can use non-caking or mildly caking coal
-Coal crushed to less than 3/8 inch (9 ram)
FIGURE 3. WINKLER GASIFIER
(6)
-------
11
coals and anthracite. The Winkler unit has the lowest overall thermal efficiency
of the four gasifiers considered--about 75 percent for the hot gas. This is due
primarily to the relatively low carbon conversion efficiency, which results in
high char loss. Also, about 70 percent of the ash in the coal is carried away
with the gas stream. Since the ash is usually in a molten form, a radiant
heat boiler is used at the offtake to solidify it and thus prevent the molten
ash particles from fouling the discharge ports.
The Winkler process is not capable of gasifying caking coal unless
the coal has been pretreated to render it noncaking. Caking coals tend to
agglomerate in the fluidized bed and drop out, resulting in unacceptably high
char loss and low efficiency. The coal must be crushed to a size less than
3/8-inch diameter (9 mm) prior to gasification. Normally, the coal should be
dried to less than 8 percent moisture before crushing, though Winkler claims to
be able to tolerate up to 30 percent moisture as long as the surface of the coal
is not wet.
Presently, the largest available Winkler gasifier is 18 feet (5.5 m)
in diameter. The estimated capacity of the unit is 8 to 10 billion Btu (71 to
89 billion kcal) of hot gas per day.* Winkler reportedly is capable of design-
ing a gasifier for 2,000,000 scf/day (57,000 m /day) that would operate any-
where from 500,000 to 3,000,000 scf/day (14,000 to 85,000 m3/day) (1.5 to
9 billion Btu/day) for a turndown ratio of 83 percent. A plant in Wesselin,
West Germany, however, is only rated at from 300,000 to 650,000 (8,5000 to
3
18,500 m /day) or a turndown ratio of 53 percent.
Wellman-Galusha
The Wellman-Galusha gasifier, Figure 4, is very similar to the Lurgi
unit except that it operates at atmospheric pressure. The gasifier is designed
and produced by McDowell Wellman of Cleveland, Ohio, and at one time it was used
for making fuel gas for a variety of industrial processes in the United States,
including kiln firing in the ceramics industry and furnace firing in the metals
2 2
* Based on 170 Ib coal/hr ft (826 kg coal/hr m ) cross section
-------
12
TYPICAL BUILDING
AND FUEL ELEVATOR .-,
OUTLINE I
WATER JACKET
DISTRIBUTOR'
COMBUSTION
ZONE
FUEL CIN
VALVES CLOSED
LOCK HOPPER
WATER SEAL
AND OUST .
COLLECTOR
GASIFICATION
ZONE
Wellman-Galusha (McDowell Wellman, Cleveland, Ohio)
Characteristics:
-Atmospheric pressure process
-Fixed bed operation with rotating eccentric grate,
countercurrent gas flow
-Typical gasifier outlet temperature 1250 F (675 C)
-Uses double screened non-caking coal (1/2 to 2 inches) (12.5 to 50 no
FIGURE 4. WELLMAN-GALUSHA FUEL GAS GENERATOR^12)
-------
13
and glass industry. There are currently several plants in the United States
that still have Wellman-Galusha gasifiers that either currently are being used
or are kept in standby condition.
Since the Wellman process is a fixed-bed operation, it has restrictions
in fuel properties similar to the Lurgi process. The Wellman process is not cap-
able of gasifying strongly caking coals, but should be capable of gasifying
mildly caking bituminous coals when a rotating agitator arm is used just below
the surface of the coal bed. Also, the coal must be crushed and double screened
to sizes from 1/2 to 2-inch (12.7- to 51-mm) diameter before being charged to
(9)
the gasifier. According to one source , atmospheric units are also more
sensitive to the presence of fines than are pressurized units.
The gasifier vessel is a cylindrical water-cooled shell with a unique
eccentric, stepped rotating grate at the bottom. The gasifying medium is intro-
duced through the grate and flows countercurrently through the coal bed, producing
a fuel gas which is taken off at the top of the unit. Spent ash falls through
the grate into the ash hopper, where it can be removed. Since the unit is not
pressurized, there is no need for the complex pressure-sealed lock-hopper-type
coal-feed and ash-removal system necessitated by the Lurgi concept. Also
requirements on seals and packing glands around rotating shafts extending into
the gasifier should be less stringent than for a pressurized unit, resulting
in lower maintenance.
The largest gasifier McDowell Wellman currently manufactures is 10
feet (3.05 m) in diameter. The estimated capacity of the unit is up to 2-billion
Btu per day (0.5 billion kcal/day) of hot fuel gas.* McDowell Wellman^ ' claims
e
2
2 2
to be able to gasify from 99 Ib coal/hr-ft (482 kg coal/hr m ) based on the
cross-sectional area of the gasifier down to as little as 7.5 Ib coal/hr ft
2
(36.5 kg coal/hr m ) without significantly affecting gas composition. This
results in a turndown ratio in excess of 90 percent.
The Wellman unit can either be operated with air and steam--yielding
2
a low-energy gas of about 170 Btu/scf (1500 kcal/m )--or with oxygen and steam--
o
yielding an intermediate energy gas of about 280 Btu/scf (2480 i kcal/m ).
The gas leaves the gasifier at roughly 1250 F (677 C) and, as such,
is only capable of generating low-pressure steam. No steam is generated in the
2 (12) 2
* Based on 99 Ib coal/hr ft cross section 482 kg coal/hr m .
-------
14
cooling water from the gasifier shell, as the water leaves the shell at around
180 F (82 C). The oxygen or air blast is normally saturated with the vapor
from the 180 F (82 C) cooling water, however, supplying the steam requirements
of the unit. The overall thermal efficiency of the Wellman-Galusha including the
sensible heat in the hot fuel gas is 95 percent.
1-2. Fuel Gas Characteristics
Table 1 gives typical compositions of the fuel gas products for the
four gasifiers covered. For the Lurgi, Winkler, and Wellman-Galusha processes,
values are given for both oxygen-blow and air-blown systems. The Koppers proc-
ess, however, when air blown requires extensive air preheat to maintain slagging
conditions and, according to Koppers, is more economical to fire with oxygen
and steam, reducing the required equipment sizes.
In addition to the basic constituents of the fuel gas shown in Table 1,
there are also a variety of impurities present, the concentrations of which de-
pend largely on the fuel being gasified. Table 2 gives an estimate of the rela-
tive magnitudes of concentrations of some of the nonhydrocarbon contaminants
that might be expected in gas made from coal. Some of these impurities, de-
pending on their concentrations, are unacceptable from an environmental stand-
point, and may be objectionable when the gas is used in some process. In this
case, they must be removed from the gas or reduced to an acceptable concentration
by a subsequent gas cleanup process. These processes are discussed in Section II.
In addition to the nonhydrocarbon contaminants shown in Table 2, the
raw gas also contains some ash and possible condensable tars. The Winkler gas
has the highest ash content, with 70 percent of the ash in the coal being carried
over in the gas stream. The Koppers process also yields a gas with relatively
high ash content, with 50 percent of the ash in the coal being carried over in
the gas stream. The Lurgi and Wellman-Galusha gases have lower ash contents in
the raw gas than the Winkler or the Koppers, but due to their relatively low gas
discharge temperatures produce relatively high concentrations of condensable
tars. Unless the tars are removed, they could present problems if the gas used
in a hot, raw state by condensing and fouling waste heat boiler tubes, burners,
and gas distribution mains. The raw gas from the Winkler and Koppers-Totzek
processes is at a high enough temperature to cause cracking and disintegration
of most of the tar compounds into less objectionable forms.
-------
15
x—s
w
0
P4
K
w
g
£
>
v^
CO
M
a
fe
H
CO
3
t-3
3
O
oi
g
o
U
04
O
fe
CO
2
O
M
H
M
CO
2
§
CO
3
rJ
6
£
e
•
r-l
w
3
PQ
13
4J
C
0)
U
t-l
ID
CM
*"*^ 3d
3 3d
4J
PQ
CM
Z
a*
CM
a
8
CM
O
O
r-4
(U
3
Pn
0
&
1
TJ
Q)
s
bO
rj
•H
>S
M-l
•rl
CO
1*1
(U
•H
t4H
•H
CO
ct3
O
O m
Is* CM
CM r-l
CM CO
CM 00
CO r-l
i-< O
"*
m in
r-i
vo in
CO CM
CO v£>
r-l r-l
CM »tf
CO r-l
r-l r-l
cd cd
0 O
U 0
C C
PQ PQ
C2
cd
0)
4-1 B
co cd
1 (U
G 4J
(U CO
00 I
•H
bO
14
3
O
oo
CM
O
ON
CM
CM
O
in
CO
<£>
in
(•%»
r-4
cd
o
u
CO
3
O
G
•rl
•jj
PQ
B
a
a)
4J
CO
1
c
(U
00
0)
H
4J
O
H
I
CO
}_4
n)
p,
cx
0 0
•tf CM
0
03 CO
1 a)
C 4-1
V CO
00 1
13
^1
-------
16
TABLE 2. TYPICAL CONCENTRATIONS OF NONHYDROCARBON
IMPURITIES IN COAL GAS(14)
Concentration,
Constituent volume percent
H S 0.3 to 3,.0
CS2 0.016
COS 0.009
Thiophene 0.010
Mercaptans 0.003
NH3 1.1
HCN 0.10 to 0.25
Pyridine bases 0.004
NO 0.0001
-------
17
1-3. Comparison of Commercial Gasification Processes
Thermal Efficiency
As a result of the inherent efficiency of the fixed-bed process and
the low gas exit temperature, the Lurgi and the Wellman-Galusha gasifiers have
the highest thermal efficiency of the four processes, of over 90 percent for
the hot, raw gas. The thermal efficiency of the hot gas for the Koppers proc-
ess is around 86 percent, though nearly 20 percent of this is in the sensible
heat of the 2700 F (1480 C) gas. (Table 3 also gives cold gas efficiency at a gas
temperature of 70 F and sensible heat content.) The Winkler process has the
poorest overall thermal efficiency of 76 percent, which is due primarily to the
low carbon conversion efficiency and high char loss of the fluidized-bed process.
This efficiency may be somewhat misleading, however, as the char represents a
usable by-product and may be used as a boiler fuel or recycled to the gasification
process.
Steam-Raising Capacity
In some processes, most notably that of Koppers, a considerable amount
of the energy in the fuel gas is in the form of sensible heat. The common
commercial practice today for all four processes is to recover at least part of
this heat in a waste heat boiler prior to going through a gas cleanup stage.
Table 4 gives estimates of the relative amounts of steam that could be raised
in a waste heat boiler for each of the four processes.
Due to the low temperature of the gases in the Lurgi and Wellman-
Galusha processes, they have the least sensible heat content and thus the lowest
steam raising capacity of the four processes. They also are limited to generat-
ing low pressure steam (due to the low temperature), which in the case of the
Lurgi process does not allow the steam to be used in the 20-atmosphere process.
The Lurgi process, however, does generate steam in the cooling water jacket of
sufficient pressure to be used in the process. This steam generally will supply
about 20 percent of the process steam requirements, and the rest of the steam
must be imported. The Wellman-Galusha process uses much less steam per Btu of
fuel gas than the Lurgi and generally the air or oxygen blast is saturated with
vapor from the 180 F (82 C) cooling-water flow, satisfying the steam requirements
-------
CO
g
!n
3
a
u
OS
1
pfci
en
o
M
H
CO
t— i
c3
u
1
CJ
H
z
5!
U
M
<*)
3
9
-o
U r-l
B C
o
IW *-*
U u
o
X
01 (u
00
b 01
u 4!
a Q
— ' b
B E
3S
CO
* 0)
t* 4
a a
a a
u C
u ^
Q)
I
(
1
u
c
f— <
i
p-4
I
CO
o
in
CJ
t
§
o*>
0
o
CNl
CO
o
r*
CT-
0
O
o
•o
•o
S
tv
eo
b
3
fcj
C
00
1 —
V
irl
00
o
r-4
1
^
u
B
01
U
>H
U4
01
:herma
*^
B
b
5
0
M
•
^•^
01
u
B
01
b
01
U-4
0)
OS
B
O
o
O)
B
•*4
energy
01
£
O
B
O
•*4
4J
U
a
b
B
B
a
01
b
3
4-1
a
b
01
Q
01
£
U
a
•O
01
£
*•*
a
B
B
oo
f-4
01
3
01
4J
E
a
01
01
B
B
01
B
O
u
b
01
B
U
U
18
>H
b sj
a a
9 B
tr oo
B
b o
S.U
b§
o
£ U
b JD
01 3
O. U
r^ b
a cu
u
b
t-l 3
0 0
£
a
•0 b
B 01
a a
0
ai-i
cS
*r4 O
•O "4-4
01 0
B
B B
01 -O
b B
Q. 3
K 0
oi a
01 B
*J O
B
b -0
B B
U -1
-•s
31
S"
.
^^
jg
B
01
vw
B
So
o
i
•o
•o
•
B
01
|
M
O
(-4
00
6
•r4
4J
a
b
01
§•
e
|
w
B
1
r>4
oo
B
4J
a
b
01
§•
•r<
E
B
1
1
1-4
p4
3
-------
19
TABLE 4. STEAM AND OXYGEN REQUIREMENTS FOR
OXYGEN BLOWN GASIFIERS (lb/106 Btu
of cold fuel gas)(a)
Gasifier
Lurgi
Koppers-Totzek
Wellman-Galusha
Winkler
Oxygen
Required
45
80
90
70
Steam
Required
184
30
50
20
Jacket
Steam
37
23
—
--
Export
Steam
67(b)
150
-------
20
(13)
of the unit. The low-pressure steam from a waste heat boiler would then be
available for other use.
Due to their relatively high gas temperatures of around 2000 F (1090 C)
or above, both the Koppers-Totzek and Winkler generators are exporters of high-
pressure steam. The Koppers unit also generates steam in the cooling water flow
which should be equivalent to about 70 percent of the process steam requirements.
The Winkler gasifier is refractory lined and has no cooling water flow, but with
a waste heat boiler should be able to generate approximately 100 pounds of steam
per million Btu (180 kg/10 kcal) of cold fuel gas, with only 20 percent required
for process needs.
Fuel Requirements
The Koppers-Totzek process is the only one of the four gasification
processes discussed that is capable of gasifying strongly caking coal with no
difficulty. Both Lurgi and Wellman-Galusha manufacture gasifiers with rotating,
water-cooled agitator arms operating just below the surface of the coal bed in-
side the gasifier to enhance their ability to handle caking coals. These units,
however, have only demonstrated an ability thus far to handle mildly caking coal
and, then, only with greater difficulty than noncaking coal. In the Winkler
process some pretreatment of caking coal would be necessary to render it non-
caking. Agglomeration of particles in the fluidized-bed process results in low
carbon-conversion efficiency, high char loss, and either air or oxygen break-
through in the fuel bed, resulting in inconsistent gas composition.
All four processes require some coal sizing prior to gasification.
Both the Lurgi and Wellman-Galusha units require the coal to be crushed and
double screened, allowing restricted sizes for gasification. The operation of
the Wellman-Galusha process is more sensitive to the presence of fines than that
of Lurgi and is restricted to sizes from 1/2 to 2-inch diameters. The Lurgi
process accepts sizes from 1/8 to 1-1/4 inch (3- to 30-mm) diameters and will
accept a small percentage of fines. Normally, however, fines must either be
discarded from the process or briquetted prior to gasification. The Winkler
process is restricted to the use of crushed coal below 3/8-inch diameter; also,
more uniform size contributes to more efficient gasification. The Koppers-
Totzek process uses pulverized coal, with at least 90 percent passing a 200-mesh
screen.
-------
21
Unit Capacities
Currently, the Koppers-Totzek gasifier has the highest unit capacity
9 9
of up to 18 x 10 Btu/day (4.5 x 10 kcal/day). The turndown capacity is about
35 to 65 percent, depending on whether it is a two-burner or four-burner unit,
respectively. The current Lurgi and Winkler units are estimated to have about
the same maximum capacity of around 7 to 10 billion Btu/day (1.7 to 2.5 billion
kcal/day). The Lurgi process reportedly has a turndown capacity of about
50 percent and the Winkler unit is estimated to be between 50 and 80 percent
The Wellman-Galusha unit has the smallest unit capacity of approximately 2 billion
Btu/day (0.5 billion kcal/day), with an estimated turndown ability of up to
90 percent.
Operational
The Lurgi unit is the most complex of the four gasifiers, both from
a mechanical and operational standpoint. A sophisticated maintenance program
would be necessary to maintain the various mechanical components and high-
pressure seals to minimize downtime. Town gas installations have claimed an
availability factor as high as 80 percent , though this is not typical of
operation for sustained periods at peak load due to daily and seasonal demand
variations. Start-up and shutdown of the Lurgi require a significant amount of
time and complicated procedures. Start-up of the gasifiers at Westfield, Scotland,
requires approximately 7 hours.
The Wellman-Galusha unit, though similar in design to that of the
Lurgi, is simpler to operate and maintain due to nonpressurization. According
to McDowell Wellman , one man ordinarily handles two gasifiers.
Both the Winkler and Koppers-Totzek units are mechanically uncompli-
cated, with no moving parts inside the gasifier. Both can be started up from
standby position in the relatively short time of about 1/2 hour, and both
companies claim on-stream factors greater than 90 percent.
-------
SECTION II
II. FUEL GAS CLEANUP
-------
22
II. FUEL GAS CIEANUP
The hot raw gas from coal gasification contains many impurities, as
was brought out in Section I and Table 2. Since coal gasification is basically
the partical combustion of coal, with a slight addition of steam, these impuri-
ties are typical of those found in any coal combustion process. Due to the
partial oxidizing conditions of coal gasification, however, these constituents
appear in a reduced form from that in which they would appear in a complete
combustion process.
Two impurities of primary interest from an environmental standpoint
that form gaseous pollutants are sulfur and nitrogen compounds. In the prod-
ucts of gasification, sulfur appears in the reduced state of primarily hydro-
gen sulfide (H_S), with generally very small amounts of carbonyl sulfide (COS),
carbon disulfide (CS_), and trace amounts of mercaptans. Nitrogen compounds
in the fuel are converted primarily to ammonia (NH_), with small amounts of
hydrogen cyanide and trace amounts of nitrogen oxide (NO) and pyridine bases.
Also, the raw gas contains particulates in the form of fly ash. Two
of the processes, the Winkler and Koppers-Totzek, carry at least 50 percent of
the ash over in the gas stream, which would be somewhat typical of a normal coal
combustion process. The Lurgi and the Wellman-Galusha processes, due to the low
gas velocities through the fixed bed of coal, would have less ash carryover,
though these two processes have higher concentrations of condensable tar-forming
compounds than either the Winkler or the Koppers-Totzek processes. If these
tar compounds are not removed from the gas prior to use they can condense in
burners and the distribution mains, causing considerable difficulties.
One advantage of coal gasification over direct combustion of coal is
the relative ease with which harmful pollutants can be removed from the products
of gasification compared with their removal from the flue-gas products of com-
plete combustion. In regard ^to the products from complete combustion, the sul-
fur and nitrogen compounds are in the oxidized forms of SO and NO diluted in
£ X
large volumes of flue gas, effective removal of which has proven extremely
difficult and for which no fully proven commercial processes of removal presently
exist. In the reduced forms of H2S and NR,, these compounds are more easily
removed and commercial processes are available that are capable of extremely
-------
23
high removal efficiencies.* Also, the volume of gas to be cleaned in the
products of gasification for atmospheric pressure units is less than half
that of the eventual volume of flue gas products and can be up to 40 times
less with a pressurized process like the Lurgi. This results in smaller
equipment being needed for gas cleanup and greater ease of removal due to
higher concentrations of impurities.
It appears that the most troublesome area of fuel-gas cleanup
currently is desulfurization. For meeting the current standard for coal-
fired steam generators of 1.2 lb SC>2 per million Btu heat input, (-^) SOme
degree of desulfurization will be necessary whenever coal of over 1 percent
sulfur by weight is gasified (Figure B-l, Appendix B). A gas sulfur removal
of slightly over 90 percent should be sufficient for achieving environmental
standards with coals of relatively high sulfur contents of about 7 percent
(Figure B-2, Appendix B).
The only desulfurization processes that have been commercially
proven operate at temperatures below 300 F. Since 10 to 20 percent of
the original heating value of the coal is in sensible heat in the gas, there
is considerable incentive to clean and use the gas at elevated temperatures.
Four; processes that are currently being developed to desulfurize the gas
at temperatures above 1500 F (815 C) are the Battelle-Northwest molten
salt scrubbing process, the Air Products and Chemical Company's solid ab-
sorbent process, the Bureau of Mines iron oxide process and Consolidation
Coal Company's dolomite process, though it is not certain when any will be
commercially available. Three of these four cleanup processes, the Bureau
of Mines hot iron oxide, the Air Products solid absorbent processes, and
the Consolidation Coal process, will also require development of a hot parti-
culate removal system, and it may also be desirable to remove condensable
tars and ammonia at high temperatures.
The conventional cold cleanup processes listed in Table B-l
(Appendix B) were mostly developed for cleaning synthesis gases or refinery
gases. All but two--the iron sponge and activated carbon processes—use
some form of liquid scrubbing to either chemically or physically absorb
H?S and sometimes C02 and HCN. In most of the processes, the sorbents are
regenerated by stripping off the I^S and other acid gases with air or steam.
The concentration of H2S in the stripped gas is generally high enough
(at least 10 percent) to allow recovery of sulfur in a Glaus conversion unit.
*For more complete descriptions of typical commercial cleanup processes
see Appendix B.
-------
24
All of the cleanup processes mentioned in Table B-l are capable
of achieving quite low concentrations of H^S in the clean fuel gas--on
the order of 10 ppm or below. To meet current environmental standards of
1.2 Ibs S02/million Btu heat input(l&), concentrations on the order of
1000 ppm H2S would be acceptable (Figure B-4, Appendix B). Almost all of
the processes in Table B-l employ expensive sorbents, often at temperatures
below ambient or at high pressures. Some processes require high circulating
liquor rates and large amounts of stripping steam for effective sulfur
removal and regeneration.
Removal of nitrogen compounds (NH-j) may not be critical for
achieving acceptable NO levels. Because gasification followed by final
X
combustion is basically a two-stage combustion process, control of NO can
X
be handled by suitable control of combustion conditions in the final
combustion stage. If a water scrub is used for particulate and tar removal
in a cold gas cleanup process, a substantial amount of ammonia will also
be removed. Since some of this NHo would be converted to NO in the final
•^ X
combustion stage with removal of ammonia resulting, total NO emissions
would thus be reduced.
For cold gas cleanup processes, particulate loadings, below the
environmental standard of 0.1 lb per million Btu^ ' should be easily
attainable by conventional collection devices. However, for the proposed
hot gas cleanup processes, development of a suitable high-temperature
particulate removal process may be necessary.
Some of the liquid scrubbing processes have the desirable characteris-
tic of oxidizing H2S directly to elemental sulfur after absorption, negating
the need for a Claus kiln or other type of sulfur recovery device. These
processes have not been well developed, however, and seem to be declining
in favor. (^) Also, the presence of arsenates in the Thylox and Vetrocoke
processes may make them undesirable, due to the potential for leakage of
a hazardous pollutant.
There are several advantages to a dry process, though neither of
the two commercial processes mentioned appear readily adaptable to current
use. The iron sponge process is most adaptable to gases with low H0S
(17)
concentrations of around 300 ppm , which would already be environmentally
acceptable (see Figure 8). It is also a batch process that would result
in a significant amount of solid waste. The activated carbon process is
-------
25
capable of high sulfur removal, but would also need frequent bed replacement
due to sulfur forming in the pores.
-------
SECTION III
CONSIDERATIONS IN UTILIZING LOW AND INTERMEDIATE-
ENERGY GAS FROM COAL IN INDUSTRY
-------
26
III. CONSIDERATIONS IN UTILIZING LOW AND INTERMEDIATE-
ENERGY GAS FROM COAL IN INDUSTRY
III-l. Comparison of Characteristics of Low and Intermediate-
Energy Gas to Natural Gas
An important consideration in analyzing the ease of retrofitting any
combustion process from one fuel to another is a comparison of the characteris-
tics of the fuels involved. Ideally, a thorough comparison of the characteris-
tics would rely on laboratory and field experiments which are beyond the scope
of this study. By comparing certain parameters calculated from typical gas
compositions for the commercial gasifiers considered in this study, however,
some insight can be gained into various aspects of the relative ease of apply-
ing them in retrofit situations. For purposes of comparison, natural gas was
chosen as the fuel that is to be replaced by low energy gas (150 Btu/scf/1332
3 3
kcal/m ) or intermediate energy gas (300 Btu/scf/2664 kcal/m ) from coal.
Since natural gas is the most widely used fuel in industry and also since many
processes have capability for firing either natural gas or oil, this was felt
to be a valid basis for comparison.
Table 5 gives high heating values for the gases in Table 1 compared
with natural gas, as well as the heat release per cubic foot of stoichiometric
mixture, which is roughly comparable to the heat release per cubic foot of com-
bustion products. The heating value of the fuel is useful in comparing the
relative fuel flow rates and pressure drops necessary in fuel supply mains and
nozzle mix type burners to obtain comparable heat release rates. The heating
value of the stoichiometric mixture is useful in comparing the relative flow
rates and pressure drops of the gases through premix type burners, (where fuel
and air are mixed prior to being burned), and flues, fans, and heat exchangers
downstream of the combustion zone.
Table 5 also gives the ratio of flow rates of the lower energy fuels
to natural gas to obtain the same heat input along with the resulting ratio of
flow rates of the stoichiometric mixtures. These are useful in comparing the
flow rates of combustion products. As can be seen from the data, typical inter-
mediate energy gas from oxygen-blown producers must be delivered to the point
of application at 3 to 4 times the flow rate of natural gas to achieve the same
-------
27
CO
cS
.J
i
H
J2J
o
H
CO
$
. 1
l_ J
s
fa
fa
O
CO
S
2
g
VH/
>
fa
O
a
o
CO
I—I
Prf
<
P-i
^
o
u
m
1
<
H
P
O
^J
0
cu
3
CO
CO
cu
P-l
0)
>
4J
CO
01
OtJ
CO
0)
4J
pa
Q
i
fa
CU
>
•H
4-1
CO
tS
CO
CO
0
r-l
I-i
3
4-1
CO
CO
CO
O
,_(
cu
fa
CO
.3
CO
31
±J
cfl
a
CO
3
r-i
§
fa
^w
o
CO
**^».
3
4-1
PQ
>
ss
o
•H
^4
4J CU
cu i-i
6 3
O 4-1
,J ^A
JS •!-<
y S3
•H
o
4J
CO
r-l
B
fa
U
>r4
4-1 CU
8 S
O 4->
•H X
3s
0
4J
co
r-l
S
-J
CJ
•H
4-1 0)
CU I-l
§ 3
w 4-J
•H X
X< *H
o S3
•r4
O
4J
CO
i-H &^
CU r-l
3 C
fa O
•I-l
1
00
C
•H
^
U-l
••-1
CO
nj
CO
CO
cu
CJ
o
}^
PM
*X) **O C^ *tO CO
CM \O i— 1 CM
co r-» ^ CM CM
VO •d' CM \O CM
oo r^- cy\ oo *o
CM OO O CO 1^-
CM OO CT» 1^- r-l
CO «~l CM CM r-l
CO cfl CO
cu cu cu
4J B 4J 4-1 B
CO CO CO CO CO
i cu i i cu
C 4J C C 4-1
cu co cu cu co
00 1 00 60 1
^"t }_f ^ ^ ^1
X -H ^! X -H
O
vO
••O
vO
o
r-l
CO
CO
O
rH
CO
i-4
3
4-1
CO
-------
28
heat input. The low energy fuels from air-blown producers require anywhere
from 6 to 9 times the flow of natural gas. The differences in the flow rates
of the stoichiotnetric mixtures are less pronounced than those for the fuel,
since the higher heating value fuels require proportionately greater amounts
of air for complete combustion. A 5 to 12 percent increase in the flow rate of
combustion products would be expected in switching to an intermediate-energy gas
from an oxygen-blown producer, and a 30 to 55 percent increase would be expected
in switching to a low energy gas from an air-blown producer. The corresponding
relative increase in pressure drop for the various fuels that would result if
the same fuel and flue gas equipment is used is also shown. For intermediate-
energy fuels about 10 times the pressure drop would be incurred through exist-
ing distribution mains and burners with a corresponding 10 to 30 percent
increase in pressure drop through heat exchangers and other gas passages down-
stream of the combustion zone. For the low energy fuels, however, pressure
drops of 30 to 80 times that for natural gas would be expected in existing
mains and burners, and corresponding pressure drops in passages downstream of
the combustion zone would be from 1-1/2 to 2-1/2 times those for natural gas.
The increased flow rates and pressure drops in fuel supply systems,
burners, heat exchangers, and exhaust flues that would be encountered in retro-
fitting a process from the use of natural gas to low or intermediate energy gas
while maintaining the same process heat input could pose a serious problem.
Supplying the necessary increased fuel supply rates to various processes through-
out an industrial plant will require either pressurized distribution mains,
larger distribution mains, or some combination of the two. Pressurized mains
would complicate the problem of potential leakage of a toxic carbon monoxide-
laden gas into working areas. Increasing the size of distribution systems to
handle the increased flow at lower pressures could create problems for processes
widely dispersed throughout the plant or in areas where space is at a premium.
Only one of the four gasification systems considered commercial in this study--
i.e., the Lurgi process—delivers the fuel gas under pressure (300 to 500 psig,
2070 to 3430/m ). Fuel gas from the other three processes would have to be
compressed either before or after the gas cleanup stage for pressurized
distribution.
-------
29
The increased flows and pressure drops occurring downstream of the
combustion zone, though less than those in fuel supply systems, will potentially
be a more serious problem. Induced draft and forced draft fans would have to
be boosted to higher operating pressures to compensate for higher flow rates.
In some cases it may be possible to reduce the pressure drop through the proc-
ess, such as by removing tubes in boiler heat exchangers, to allow greater
volumes of flow at lower pressure drops without upsetting the heat transfer
characteristics of the process.
If changes in the process cannot be made to compensate for increased
flows and pressure drops, process derating may be necessary. This problem will
probably be most severe for handling the increased volumes of flue gases rather
than the handling of increased volumes of fuel. Analysis of Table 5 reveals
that inability to handle additional flue gas volume could result in a derating
of from 5 to 10 percent for intermediate energy (300 Btu/scf; 2664 Kcal/tn^) gas
or from 25 to 35 percent, for low energy gas (150 Btu/scf; 1332 Kcal/m3).
Alterations necessary in burner designs are uncertain without labora-
tory data on which to base the redesign. Burners that at one time were used
for low energy gas are not readily available today as production items. Also,
the old designs would probably no longer be acceptable in most cases, as
advances in burner and process technology have resulted in burners with generally
wider stability ranges, intermittent instead of continuous piloting, and sophis-
ticated combustion monitoring and control.
As part of this study, an analysis was made comparing combustion
characteristics relating to flame stability of low and intermediate energy gases
typical of the commercial gasifiers considered in this study to the characteris-
tics of natural gas. The complete results of the study are given in Appendix A.
The analysis was based on the critical flashback velocity gradient cal-
culated from typical gas compositions. This parameter has been found very useful
in estimating the inherent flame stability of various fuels in that it is related
to the blowoff limits in an enclosed system, the burning velocity, and the
chemically controlled reaction rate. From the analysis it was concluded that
there should be relatively little difficulty in designing burners with adequate
flame stability for the intermediate energy fuels of around 300 Btu/scf; 2664 kcal/n
from oxygen-blown producers; however, stability problems would be more severe with
the lower energy fuels from air-blown producers.
-------
30
It is generally felt that adequate industrial burners could be
developed for intermediate and most low energy gases. Generally, flow areas in
fuel supply lines and burner parts would have to be increased to handle the in-
creased fuel flows necessary to maintain the same energy input as with natural
gas or oil. Overall burner diameters or tile diameters probably would not be
increased in most cases, however, minimizing the amount of modification neces-
sary in the walls of the furnace.
Larger general-purpose burners should be easiest to retrofit to low
or intermediate energy gas. However, the performance of some specialty types
of burners may be difficult to duplicate. These include high-intensity or high-
velocity burners, burners where particular flame shapes are necessary (such as
flat flame burners), and applications requiring carefully controlled mixing and
combustion rates such as flow-through or pull-through radiant tubes.
III-2. Environmental Considerations
The gasification of coal followed by its use as a fuel in a process
is basically two-stage combustion. If the intermediate product gas or low
energy gas is used directly in a process, then environmental considerations are
basically the same as those for normal two-stage coal combustion. One advan-
tage of gasification, however, is that certain undesirable pollutants can be
removed from the fuel gas more easily than they can from the eventual flue
gases or products of combustion. The two primary gaseous pollutants, S and N,
occur in higher concentrations in the fuel gas than in the combustion products
and also appear as the reduced forms of H_S and NIL, for which commercial re-
moval processes have been developed (see Table B-l). Also, the volume of gas to
be cleaned in the fuel gas stage is about half the eventual volume of the products
of combustion for nonpressurized units and can be much less for pressurized units
like Lurgi gas producers.
Figure 5 shows a schematic diagram of a gasification system with a cold
gas cleanup process for sulfur removal and subsequent sulfur recovery. The
diagram illustrates various points in the system where discharges to the environ-
ment can occur and must be evaluated in any given application.
-------
31
Q. 4-
5) <1>
* -p
c
o
•*— ^—
c o
- c
O 0)
CO en
0
I!
3 O
co
Si
CO
X
0)
o
CD
V)
o
o
o
o
o
c
0)
o o>
< 0
en
to
•H
S
Ed
o
a)
CO
«J
O
0)
CO
-H
S
a"
-H
J
H
O
-a
w
.J
U
o
Q
£
CJ
H
o
M
to
<
O
O
-------
32
If the H»S in the fuel gas is not removed prior to combustion, it will
be converted primarily to SCL. Commercial processes have been developed (see
Table 5) for removing H2S from fuel gas down to concentrations in the range of
10 ppm or less. For most gases, a level of less than 1000 ppm would result in
SO. emissions less than the current environmental standard for steam generators
of 1.2 Ib S02/106 Btu heat input (see Figure B-4). All the processes in Table B-i
remove H-S with some kind of sorbent. For most of the processes, the sorbent
is regenerated by stripping off the H2S and other acid gases. The concentration
of H-S in the stripping gases should be sufficiently high (at least 10 percent)
to allow recovery of elemental sulfur in a Glaus process. A Glaus kiln is cap-
able of removing approximately 95 percent of H~S from the stripping gas; the
remaining H»S in the tail gas represents a potential pollutant and should at a
minimum be burned to S0? with a flare. The tail gas from the sulfur recovery
process could be further scrubbed for additional H.S removal, though it is felt
in general that the SO. emissions from the raw tail gas combined with the
emissions from the scrubbed fuel gas would be less than the environmental stand-
ard of 1.2 Ib SO-/10 Btu heat input (2.16 kg/kcal).
All but two of the processes in Table B-l involve liquid scrubbing with
a variety of absorbents. Though these absorbents may be regenerable, eventually
they deteriorate to a point where a bleed stream must be taken off and fresh
absorbent added to maintain acceptable reactivity. The spent sorbent from the
bleed stream represents a potential pollutant. Some of the processes mentioned
in Table 4 for oxidizing H_S directly to elemental sulfur--Thylox and Vetrocoke
processes involve the use of poisonous arsenates in the absorbing solution which
would complicate the handling of spent sorbent.
Another potential pollutant is the sulfur formed in the sulfur recovery
process. If a significant amount of low energy gas were used in industry and
elemental sulfur was the chief product of the H_S removal process, the market-
ability of the sulfur would be questionable and the sulfur would thus become a
solid waste.
In any sulfur removal process, however, the sulfur compounds must be
disposed of in some form. Proposed lime and limestone scrubbing systems for
removing SO^ from stack gases convert the sulfur to solid waste in the form of
CaS03 and CaSO , which involve a greater amount of solid waste than with elemental
-------
33
sulfur. Some schemes would convert SCL to sulfuric acid, H0SO., as a usable
L- 2. 4
by-product which would suffer from more difficult storage and marketability
problems than elemental sulfur and would be less stable and more difficult to
dispose of.
NO emissions evolve by basically two mechanisms: fixation of air-
X
borne free nitrogen to NO with high temperatures in the combustion zone and
X
oxidation of fuel-bound nitrogen to NO , which occurs at lower temperatures.
X
In the reducing atmosphere of the products of gasification, fuel-bound nitrogen
is converted to NIL. If the ammonia is not removed from the fuel gas, such as
in a water scrubbing operation, it would be carried through to the final com-
bustion process and some of it would be oxidized to NO in the same manner as
would occur in two-stage coal combustion. The formation of NO due to thermal
X
fixation of free nitrogen in the air should be similar to that from firing
natural gas, as this mechanism will depend primarily on the temperature of the
final combustion process. In summary, if NH,. is stripped from the fuel gas
prior to combustion, emissions of NO should be similar to that from natural
gas. If NH., is not stripped from the fuel gas, emissions of NO should be no
•J X
greater than those from two-stage coal combustion. Since NO formation is
X
somewhat dependent on the characteristics of the combustion process and because
little data exist on NO emissions from two-stage combustion of coal, an accurate
X
assessment of NO emissions will depend on actual test data.
x *
Particulates can be removed from the fuel gas by standard particulate
removal processes such as water scrubbing, filtration, or electrostatic preci-
pitation. There should be no problem in obtaining environmentally acceptable
dust loadings in the fuel gas itself. When the fuel gas is used in processes
that are essentially particulate free--that is, when the process itself does not
contribute particulates to the gas stream--there should be little problem attain-
ing environmentally acceptable particulate emissions. In some processes, cement
and lime kilns for example, particulates are entrained in the combustion gases
and must subsequently be removed. In general, low energy gas will result in
slightly higher--possibly 20 to 30 percent higher—volume throughputs of com-
bustion gases than natural gas, which could result in greater entrainment of
particulates and, hence, higher dust loadings in the flue gases of these proc-
esses. In some cases, where particulate emissions are marginally acceptable
and the process uses oil or natural gas, dust loadings in the flue gases may
-------
34
become unacceptable. In those cases that were using coal directly, such as
pulverized coal-fired cement kilns, the removal of ash in the low energy gas
may offset the increased dust entrainment in the kiln itself. In summary, for
those processes that do not themselves contribute to the dust loading of the
combustion gases, particulate emissions should be well within environmental
standards. In those processes originally firing natural gas or oil where
particulates are entrained in the combustion gases, dust loadings in the flue
gases will increase.
Another source of particulates in using low energy gas is in fugitive
dust from coal storage and handling facilities. This potential source of dust
is inescapable whenever coal is used as a fuel. Its control will depend largely
on properly designed and maintained coal-handling equipment.
The particulates removed from the gas stream as well as the ash from
the gasifier itself represent a solid waste. Some industries may have suffi-
cient room and facilities for ash disposal, other industries will require the
ash and other solid waste by-products to be shipped to a suitable disposal site.
If a water scrub is used to cool and clean the gas prior to sulfur
removal, the constituents removed by the water represent an additional source
of pollution. Normally, the water would remove particulates, condensable tars,
NHL, some CO-, and negligible amounts of H-S and other acid gases. The parti-
culates and tars can be removed by conventional filtration devices. Tars can
be separated as usable by-products and recycled to the gasifier, used directly
as fuels in some processes, or sold as feedstock for petrochemical synthesis.
NH_ and other acid gases can be stripped from the water with heat and NH_ can
be recovered as a marketable ammonium salt by scrubbing with strong tUSO, or
H-PO, , NH_ can be recovered as refrigeration grade anhydrous ammonia with a
process like the Phosam Process. "
Another area of environmental concern includes hazards created by the
use of low energy gas in industrial processes in the plant itself. One of
these is the increased toxicity of low or intermediate energy gas over natural
gas due to its content of carbon monoxide. Where small leakages of natural
gas could be tolerated in distribution mains and around burners and other
points of application, these leakages may be intolerable from a health-hazard
standpoint for low energy gas. The fact that many industries will want to
distribute low energy gas under pressure to obtain the necessary increased
-------
35
burner pressures and higher flow rates while keeping distribution mains as
small as possible will only compound the problem.
The higher pressures and flow rates that would be necessary in dis-
tributing and burning low energy gas over natural gas creates another in-plant
hazard in the form of increased noise. The two primary sources of increased
noise are those stemming from higher pressure drops across pressure-reducing
devices and combustion roar resulting from increased volumes of flue gas prod-
ucts and increased pressure drops across burners.
Both problems--increased toxicity of low energy gas over natural gas
and increased noise levels inside factories--require further study to accurately
assess their effects.
-------
SECTION IV
UTILIZATION OF LOW AND INTERMEDIATE ENERGY GAS IN INDUSTRY
-------
36
IV. UTILIZATION OF LOW AND INTERMEDIATE ENERGY GAS IN INDUSTRY
IV-1. Criteria for Evaluating Ease of Retrofit of Industrial Processes
When analyzing the potential applicability of low or intermediate
energy gas as a retrofit fuel in the wide variety of fuel uses that commonly
occur in industrial furnaces and other types of processes, some basic criteria
are necessary for judging the relative applicability or ease of retrofit based
on the general characteristics of the process. The criteria which have evolved
as most significant, based on analyses presented in this study and conversations
with individuals closely associated with industrial fuel use, are the following:
• Processes with a relatively few number of large burners should
be easier to retrofit than processes with a large number of
small burners that require extensive manifolding and control.
• Processes in which specific flame shapes or other flame
characteristics are required, such as flat flame burners,
will be difficult to duplicate in retrofit and may require
field testing and process modification. General-purpose
burners, where specific flame characteristics are not im-
portant, would be easiest to retrofit.
• Processes requiring particular fuel mixing or combustion
rates such as those using radiant tubes will be difficult
to convert and may require extensive testing and process
modification.
• Processes with relatively large combustion chambers and the
ability to easily increase volume throughput of combustion
products by at least 10 percent and possibly up to 30 per-
cent will be relatively easy to retrofit.
• Processes sensitive to small amounts of sulfur in the com-
bustion gas will be more difficult to retrofit and will
place greater demands on the gas cleanup system employed.
• Processes in which the lower flame temperature of some
lower energy gases from coal will cause reduced furnace
temperatures and lower process rates may require regenerators
-------
37
for preheating the fuel and thus will be more difficult to
retrofit.
• Processes that are physically localized in the plant will be
easier to retrofit than those widely dispersed around the
plant. This will depend on the extent to which distribution
mains must be replaced or modified and how much power is
required for distributing the gas.
• Industrial plants or large processes with widely varying
load factors or frequent start-up and shutdown will be
difficult to match to the load-following capability of
the gasifier. Processes that are operated more or less
continuously with only periodic shutdowns for maintenance
will be most amenable to gasification.
• Processes already using lower energy gases (such as blast
furnace top gas, coke oven gas, or low-grade refinery gases
mixed with natural gas) should be little affected by sub-
stitution of low or intermediate energy fuels for natural
gas in the mixture.
• Processes in relatively open areas where a high degree of
ventilation can minimize the toxic effect of gas leakages
and where installation of larger distribution mains is not
an immediate problem should also be easier to retrofit.
Fuel use in industrial processes encompasses an exceedingly wide
variety of specialized applications involving a myriad of furnace and proc-
ess configurations. Analyzing the potential applicability of low or inter-
mediate energy gas from coal to the broad range of potential industrial uses
can, therefore, only be attempted here on a general basis. Even when con-
sidering a particular type of process, furnace designs and other aspects of
the process relevant to its ease of retrofit can vary extensively from appli-
cation to application. Therefore, even though a particular type of process
may be considered relatively attractive from a retrofit standpoint, a parti-
cular application may be highly unattractive.
-------
38
IV-2. Classification of Common Industrial Processes According
to Their Ease of Retrofit
Based on the preceding criteria describing general characteristics
of a process that should directly affect its ease of retrofit, some categori-
zation of the broad range of industrial processes can be made. For purposes
of this study, processes were broken down into three groups indicative of
their relative ease of retrofit or applicability to the use of low and inter-
mediate energy gas.
• Group I; These are processes that are felt to be most
attractive to retrofit of low or intermediate energy gas.
Processes in this group should require minimal modification
and minimum reduction in process rate or product quality.
These processes are also felt to represent relatively large
energy uses in industrial plants. It is felt that immediate
efforts in applying low energy gas to industrial plants can
best be expended on large concentrated energy uses rather
than the wide variety of diverse fuel uses common in nearly
every industry.
The exact modifications necessary for retrofit are difficult
to assess in general and will depend largely on each individ-
ual application. In many cases, modifications will primarily
involve new burners, with little or no modification to the
walls of the furnace, and possibly increasing the volume throughput
of flue gases to account for the greater volume of flue products
for the low and intermediate energy gases over higher grade
fuels. Depending on the particular low energy gas used, the
addition of recuperators for fuel air preheat may be
necessary to attain the necessary furnace temperatures.
This will primarily be a problem with the low energy fuels
(150 Btu/scf) from air-blown producers; the intermediate
energy fuels should not be significantly different from
natural gas in this respect. Those processes considered to
be in Group I are
-------
39
Sintering (primary metals industry)
Palletizing (primary metals industry)
Incinerators (all industries)
Afterburners (all industries)
Various kilns (ceramics industry, lumber industry)
Reheat furnaces (primary metals industry)
Reverberatory furnaces (primary metals industry)
Soaking pits (iron and steel industry)
Air preheating (iron and steel industry)
Open hearth furnaces (iron and steel industry)
Calcining (cement, lime, aluminum industries)
Heat treating where finish is not important (primary metals, ceramic
industries)
Forging furnaces (iron and steel industries)
Direct-fired atmosphere generators (primary metals, ceramics industrii
Smelting operations (nonferrous metals industry).
Group II« This group includes processes that may be attractive
for retrofit to low or intermediate energy gas but where more
extensive considerations are necessary. These considerations
may include more extensive R&D work to develop suitable burners
than for those processes in Group I, or they might also involve
substantially more process modification or derating. Processes
included are
Direct firing on finished products (ceramics, primary metals industr
Indirect fired furnaces with pull-through radiant burners
Glass tanks (glass industry)
Finish annealing operations (primary metals industry)
Processes using flat flame burners (ceramics, primary metals industr
Processes using partial premix burners
Blast furnace injection (iron and steel industry)
Paper and print drying (pulp and paper industry).
Group III. Processes that are felt to be relatively unattractive
for retrofit to low or intermediate energy gas. Generally, these
processes are felt to be sufficiently unattractive for retrofit
that they should either be allowed to continue on the fuels they
are presently using or be replaced with new processes. In general,
-------
they represent relatively small energy requirements compared
with other processes in most industries. These include
Processes using flow-through radiant burners
(ceramic or metallic grid) (all industries)
Automated flame heating systems such as soldering,
brazing, and lamp sealing (glass, pottery, special metals)
Glass and ceramic fiber production (glass industry)
Glass cutting, scarfing, and annealing (glass industry)
Direct-fired or indirect-fired space heaters (all industries).
IV-3. Utilization of Gas From Coal in Industrial Boilers
Industrial boilers represent an important potential for utilization
of low and intermediate energy gas and, because they are a class of process in
themselves and are used in some way in almost every industrial plant, they are
treated separately in this study. In 1968, nearly half the energy input to the
industrial sector was used for raising steam, either for process use or for
(1 8^
on-site generation of electricity. Of this, around 80 percent was in the
form of natural gas and petroleum products.
Industrial boilers fall into two broad categories--fire-tube and
watertube. As the name implies in a fire-tube boiler, the combustion and gas
flow takes place inside the boiler tubing or heat exchanger and steam is raised
by water contacting the outside of the tubing. Because the pressure of the
steam must be contained by the shell of the boiler vessel they are limited to
smaller sizes and lower steam pressures. Sizes are generally given in boiler
horsepower and range from around 10 to 20 horsepower to around 600 to 800
horsepower (1 boiler horsepower equals 34.5 pounds of steam per hour at 212 F).
Maximum steam pressures range up to about 300 psi. These boilers are nearly
always sold in complete packages and as such are quite compact. The vast
majority of fire-tube boilers burn natural gas, oil, or by-product gases and
liquid fuels.
In watertube boilers, steam is raised by passing water through the
tubing or heat exchanger, with combustion and gas flow taking place outside of
the tubing. This type of boiler is suited to larger sizes and higher pressures.
Sizes generally start at around 10,000 Ib steam/hr (4530 kg steam/hr) and range
-------
41
up to large utility boilers of over 1,000,000 Ib steam/hr. For industrial
purposes, boilers seldom exceed 500,000 Ib (226,500 kg steam/hr). Up to about
250,000 Ib steam/hr (113,250 kg steam/hr), watertube boilers can be shop assembled
and shipped as a packaged unit. Above that capacity, boilers must be erected in
/ 1 n\
the field either from scratch or from prefabricated parts. About 85 percent '
of the installed capacity of industrial boilers is in the range below 250,000 Ib
steam/hr (113,250 kg steam/hr), and of all industrial watertube boilers up to
500,000 Ib steam/hr (226,500 kg steam/hr), it is estimated that 78 percent^19^
of the installed capacity is in natural gas or oil. The steam pressures of
industrial watertube boilers vary considerably and may range from several
9
hundred psi to over 2000 psi (1380 N/m ).
Considering the distribution of fire-tube and watertube boilers in
industry, fire-tube boilers are most prominent below about 16,000 Ib steam/hr
(7250 kg steam/hr), accounting for an estimated 78 percent of total sales.
From 16,000 to 100,000 Ib steam/hr (45,300 kg steam/hr), watertube boilers
account for an estimated 79 percent of sales, and above 100,000 Ib steam/hr
essentially all boilers are of watertube construction.
As an industrial process, boilers are considered an attractive retro-
fit application for low or intermediate energy gas. This would include both
fire-tube and watertube boilers whether shop erected or field erected.
In the analysis of the preceding section, industrial boilers would be
classed in Group I, as a process requiring a minimal amount of modification for
retrofit and as sustaining only minimal derating, if any. The characteristics
of boilers that relate to their relative ease of retrofit are similar to those
for industrial processes in general, as outlined in the previous section.
The effect of conversion of boilers to low or intermediate energy gas
will depend a great deal on the development of suitable burners as for other
processes. Also, like other processes, the increased volume of flue gas products
will result in higher pressure drops through heat exchanges and exhaust flues.
This is not felt to be a severe problem, however, and can be compensated for by
removing tubes in the convective heat exchanger or by increasing the load on the
induced draft and forced draft fans.
Package boilers, for which a premium is placed on compactness and
minimal combustion volume, will generally require more careful design consider-
ations and judicious choice of burners than more generously sized field erected
-------
42
units. In these cases, retrofit to intermediate energy gas (300 Btu/scf/
2
2664 kcal/m ) should present little difficulty, but with lower energy fuels--
o
below 150 Btu/scf (1332 kcal/m )--more difficulty will be encountered and some
derating may be necessary. This will depend extensively on burner design,
however.
Though up to this point discussion has been primarily directed at oil-
and gas-fired boilers, there may be impetus to convert existing coal-fired units
to low or intermediate energy gas to avoid the high cost of installing stack gas
cleanup systems to meet environmental regulations. Boilers originally designed
for coal or other solid fuels inherently have more generous combustion volumes
and wider tube spacings in the convective heat exchangers. Both of these
characteristics are attractive in retrofitting to low or intermediate energy
gas, and there should be little difficulty in achieving comparable heat release
to that of the original fuel.
The vast majority of industrial coal-fired boilers under 200,000 Ib
steam/hr (90,600 kg steam/hr), are stoker fired, however. Unless these units
have the capability of also firing gas or oil, there may be no provision for
the installation of burners in the furnace walls. In these cases the furnace
walls would have to be modified to accept low or intermediate Btu gas burners.
Another problem affecting the conversion of boilers designed for
solid fuels such as coal to gaseous fuels is the possible upset in heat trans-
fer balances as a result of the differences in the luminosity of gas flames
from those of solid fuels like coal. In boilers with water wall construction,
the wall tubes receive heat primarily by radiation from the flame. When super-
heaters are used they generally consist of banks of tubes through which the hot
flue products flow, with the primary heat transfer mechanism by convection.
The lower liminosity of the gaseous flame compared with that of coal or other
solid fuels would result in less radiant heat transfer and the higher volume of
flue products with low or intermediate energy gas would result in greater con-
vective heat transfer. This may result in higher superheat temperatures than
the boiler and its associated process load was designed for, hence, de-
superheaters might be required.
-------
43
IV-4. Retrofit of Gas Turbines to Low- or
Intermediate-Energy Gas From Coal
Direct-fired gas turbines are frequently used for generating electri-
city when coupled to an electrical generator or possibly for driving a large
piece of machinery. Gas turbines used in industry are of two basic types, air-
craft and industrial. Aircraft turbines, as the name implies, are very similar
to engines used to power aircraft and tend to be of a very compact design
generally requiring a higher grade of fuel such as natural gas, kerosene, or
No. 2 distillate oil. Industrial-type gas turbines, however, tend to be more
rugged and generously sized for a particular output and can generally handle
lower grades of fuel than aircraft turbines. In retrofitting from high energy
gaseous or liquid fuels such as natural gas and oil to lower energy gases from
coal, aircraft turbines would tend to be affected more severely than industrial-
type turbines, though in general each should experience the same kinds of
problems.
In firing a lower energy gas from coal into an existing gas turbine,
the gas first should be as free of particulates and sulfur as possible to pre-
vent erosion and corrosion of blade surfaces. Generally, the degree of clean-
liness offered by the cleanup systems in Table B-l should provide for sufficiently
clean gas, though no long-term test data are available to verify this. Another
severe requirement on the gas is that it be available at the turbine pressure,
which usually runs somewhere between 100 and 200 psig. Of the gasifiers dis-
cussed, only the Lurgi gasifier supplies gas at a higher pressure; the other
three gasifiers would require compression of the gas after gasification. With
o
300 Btu/scf (2670 kcal/m ) gas from oxygen-blown gasifiers, about three times
the volume of gas would have to be compressed as natural gas, and with air-
blown gasifiers nearly six times the volume is involved. This will require
additional compression capacities for the fuel lines, and the increased fuel
volumes will also require larger gas manifolds inside as well as outside the
turbine.
Simple modification of the combustors may be sufficient to handle the
lower energy gases, especially with 300 Btu/scf gas from oxygen-blown producers
o
in industrial-type turbines. With gases lower than 200 Btu/scf (1776 kcal/m ),
however, new combustors may be necessary, especially in compact aircraft-type
-------
44
turbines; however, no serious problems would be expected until the gas heating
3 *
value drops below 150 Btu/scf (1332 kcal/m ). At this point potential derating
of the unit may be necessary due to a lack of sufficient combustion volume or
inability of the turbine to handle the increased volume of combustion products.
Other adjustments may also be necessary to balance turbine-compressor
loadings for the new fuel. These adjustments will be more severe with low
o
energy (150 Btu/scf, 1332 kcal/m ) gas than with intermediate energy gas due to
lower volume of combustion products. In aircraft-type turbines, balancing of
compressor load with turbine load is somewhat difficult, as part of the turbine
frequently is coupled directly to the compressor. In these cases, a new turbine
rotor may be necessary to obtain proper operating balance. Industrial turbines
in many cases will have inlet guide vanes for adjusting compressor loading.
Overall, gas turbines would be rated as attractive for retrofit to
low or intermediate energy gas from coal providing the required gas pressure and
cleanliness are achieved. No real problem is foreseen in converting to inter-
3
mediate energy gas at 300 Btu/scf (1664 kcal/m ) for either industrial or aircraft-
type turbines. Converting to lower energy gas, especially below 150 Btu/scf
3
(1332 kcal/m ), may pose more serious problems with combustion and balancing of
turbine and compressor performance, in which case derating may be necessary.
*Battelle estimate.
-------
SECTION V
ANALYSIS OF POTENTIAL MAJOR INDUSTRIAL USERS OF
LOW AND INTERMEDIATE ENERGY GAS FROM COAL
-------
45
V. ANALYSIS OF POTENTIAL MAJOR INDUSTRIAL USERS OF
LOW AND INTERMEDIATE ENERGY GAS FROM COAL
The criteria set up in Section IV-1 can be used in evaluating the
relative ease of retrofitting a process to low or intermediate energy gas based
on the general characteristics of the process. The following is a general
analysis of processes in the four major energy consuming industries in the
United States in 1971 according to the 1971 Census of Manufacturers . They
are the chemical and allied product industry (SIC 28), the primary metals
industry (SIC 33), the petroleum and coal products industry (SIC 29), and the
stone-clay-glass products industry (SIC 32).
These four industries account for approximately 63 percent of
total energy use in all industries, excluding utilities, and account for
65 percent of the consumption of oil and natural gas in industry.
V-l. Chemical and Allied Products Industry (SIC 28)
Of all basic industries, the chemical and allied products industry is
the largest single consumer of energy. The industry is broken down into eight
categories shown in Table 6 along with the amounts of energy used in different
forms in 1971.
In general, the chemical industry can be characterized by a large
number of specialized plants producing a wide variety of products. As can be
seen from Table 8, the bulk of the industry energy use (over 70 percent) is
consumed in the industrial chemicals sector (SIC 281). This sector produces
alkalines and chlorine (SIC 2812), industrial gases (SIC 2813), cyclic inter-
mediates and crudes from coal tar (SIC 2815), inorganic pigments (SIC 2816),
and various organic chemicals (SIC 2818) including noncyclic organic chemicals,
solvents, poJyhedric alcohols, synthetic perfumes, and plasticizers. Also
included are various inorganic chemicals (SIC 2819) including inorganic salts
of sodium, potassium, and aluminum, ammonia compounds, fertilizer materials,
and rare earth metal salts.
The next largest energy consuming sector is plastics materials and
synthetics (SIC 282). This sector includes plastics materials and resins
-------
46
TABLE 6. ENERGY CONSUMPTION IN CHEMICALS AND ALLIED
PRODUCTS INDUSTRY^1) (1012 Btu/year) (1971)
SIC
28
281
282
283
284
2851
2861
287
289
Description
Chemical and allied products
Industrial chemicals
Plastics materials and synthetics
Drugs
Soap, cleaners, toilet goods
Paints and allied products
Gum and wood chemicals
Agricultural chemicals
Miscellaneous
Oil
206.3
96.3
59.9
22.0
8.7
1.9
2.5
3.5
11.5
Gas
1472.7
1157.9
158.9
17.5
21.3
6.4
5.4
57.8
48.2
Coal
476.4
268.8
165.1
9.4
4.9
0.3
--
2.6
24.1
Total(a)
2436.8
1719.3
404.8
56.9
48.4
17.3
9.2
71 .9
108.1
(a) Includes other energy sources besides those listed.
-------
47
(SIC 2821), synthetic rubber (SIC 2822), cellulosic man-made fibers (SIC 2823),
and organic noncellulosic fibers (SIC 2824). The other six industry sectors
are relatively small energy consumers and combined account for around 12 percent
of the industry energy consumption.
The different types of energy-consuming processes in the chemical
industry are as varied as the products. Basically, however, chemicals are
generally either made at high temperatures or temperatures below ambient, and
energy is necessary to fire heat exchangers or furnaces for fluid heating or
for cooling with refrigeration.
fossil fuels can be used directly or are commonly used to raise low-
pressure steam, which is then used for a variety of heating and process needs.
Refrigeration can be supplied by either gas or oil-fired absorption machines,
or steam or electric driven compressor units.
The bulk of fossil fuel energy in a chemical plant goes for either
steam raising in a boiler or for firing furnaces. Figures giving the relative
amounts of various fuels used in different processes are not generally available.
The applicability of low and intermediate energy gas to industrial boilers was
discussed in the previous section and that discussion would apply to boilers in
the chemical industry. The use of low or intermediate energy gas in furnaces
used for heating chemical feeds will depend on the design of the furnace, but
in general should follow the criteria listed in Section IV-1. A great many
furnaces will have tube-type heat exchangers fired externally with the chemical
(19)
teed inside the tube in much the same manner as watertube boilers , which
should be readily adaptable to either low or intermediate energy gas. Other
applications with indirect firing, or where special flame shapes or luminosity
are required, will demand more careful consideration.
Duo to the wide variety of chemical plants and processes it is
difficult to assess the potential for utilization of low or intermediate energy
gas to the in.luscry as a whole. Many chemical plants are very small, employing
20 people or u-ss, and other plants are much larger, employing several hundred
or possibly more. Also, many plants tend to run continuously for long periods
of time, putting strenuous requirements on the dependability of a gasification
system.
-------
48
In conclusion, utilization of low or intermediate energy gas in the
chemical industry would be best evaluated on an individual plant basis.
Particular attention should be paid, however, to those cases where by-products
or wastes from the gasification system could be used in the chemical process.
Examples are
Sulfuric acid plants (elemental sulfur, S00)
Fertilizer or nitric acid plants (ammonia)
Cyclic intermediates and crude production (tars) plants.
V-2. Primary Metals (SIC 33)
According to the Census of Manufacturers the primary metals
Industry in i971 ranked second as a consumer of energy among various industries.
rhc primary metals classification is divided into three basic subclassif ications:
• Basic iron and steel (SIC 331)
w Ferrous foundries (SIC 332)
* Nonferrous metals (SIC 333).
Basic Iron and Steel
Basic iron and steel consists of the production of basic steel shapes
including simple structural shapes, steel wire, plate, and sheet. There are
two basic types of mills—primary integrated mills and secondary mills.
Primary integrated mills convert iron ore and coke to pig iron by
means of a blast furnace. The pig iron is then made into steel by primarily
one of two processes in this country: the basic oxygen furnace (EOF) or the
opt i;-h< .(rih furnace. The BOF oxidizes the impurities in molten iron, thus
producing S'.eel by direct injection of oxygen. The self-sustaining exothermic
r.-action n-quires no additional energy source. The open-hearth process, on
i lie other h^id, uses a regenerative reverberatory furnace heated by radiation
i rom a lumi io
-------
phasing out the open hearth in integrated mills and this trend is expected to
continue, as no new open hearths have been installed in many years nor are any
new ones planned.
Secondary mills are similar to integrated mills except they don't
have the capability of making their own iron from iron ore and coke, but
rather receive the cold, raw iron from other plants. Since the iron in a
secondary mill is not received in a molten state the basic oxygen process is
not as attractive as the open hearth. The open hearth, however, is being
challenged in secondary mills by the electric arc process, which is becoming
a common choice whenever additional melting capacity is needed. The electric
arc process uses an electric current to melt the cold iron charge for conver-
sion to steel.
The iron-making process provides two valuable sources of by-product
gaseous fuel.-, for integrated mills. The coking operation converts coal suit-
able for iron loaking to coke by processing in a battery of coke ovens. The
resulting off-gas (coke oven gas) is similar to intermediate energy gas from
3
coal with slightly higher heating value of 550 Btu/scf (4884 kcal/m ). In the
blast furnace coke is burned with iron ore to produce iron, and the resulting
off-gas (blast furnace top gas) is similar to a low energy gas with a
3
lower heating value of around 90 Btu/scf (800 kcal/m ).
Both coke oven gas and blast furnace top gas are used extensively in
most processes in an integrated steel mill, though natural gas and other
i
purchased fuels are necessary as a supplement. In some mills, all gases--coke
oven gas, blast furnace gas, and natural gas--are mixed to form a single gas
of generally about 700 BLu/scf (6216 kcal/m ), which is then used throughout
the plant. Other mills, however, will keep these various gases separate, each
one being intended for use in specific processes. In analyzing the retrofit
of an integrated steel plant to low or intermediate energy gas, this aspect
should be taken into consideration.
After the pig iron has been converted to steel by one of the four
basic s tee 1-uia k Lnf processes—basic oxygen furnace, open hearth, electric arc,
or Bessemer < onwrter (accounts for very small amount of steel production)--the
molten metal is ast into ingots. The ingots are then heated in soaking pits
prior to being foiged, drawn, or extruded into desired shapes. Besides these
-------
50
basic "working" operations, other processes carried out in integrated and
secondary mills include annealing, tempering, and normalizing operations for
providing proper stress levels and grain structures in the metal and various
coating operations such as galvanizing, tin and terneplating and coating with
organic chemicals such as paints, varnishes, enamels, and lacquers.
Table 7 gives estimates of total energy use and the amount of energy
supplied by natural gas to various processes in iron and steel production in
(2i)
1971 . The amount of energy supplied by natural gas can be used to estimate
the potential substitution of gas from coal.
• Sintering and pelletizing operations are for beneficiating
the iron ore prior to charging in the blast furnace. The
fuel gas requirement is primarily for ignition of the ore
prior to the sintering or pelletizing operation.
• Coke ovens are fired primarily by coke oven gas, though
estimates show that some of the energy is supplied by
natural gas.
• Blast furnace injection involves injection of oil or a
reducing gas, such as natural gas, into the blast furnace
to reduce the coke requirement for a given amount of iron
ore. If low or intermediate Btu gas were used it would
have to be at a pressure of about 35 psi, a temperature of
around 2600 F, and would have to be relatively free of C0«
(25)
and water vapor to minimize endothermic effects
• Open-hearth furnaces used low energy gas from coal for fuel
over 50 years ago, though natural gas and oil are the primary
fuels today. Coal or tar is usually injected with the primary
fuel to give luminosity to the flame.
• Air preheaters are used to preheat air prior to firing in the
blast furnace. Blast furnace top gas is most frequently
used since it is readily available. Approximately 10 percent
of this energy is supplied by natural gas, however.
• Scrap steel is usually preheated and charged to any of the
four basic steel-making processes along with the iron. Pre-
heating is usually done in direct gas-fired reverberatory
furnaces and serves to minimize melt times.
-------
51
TABLE 7. ENERGY USE IN BASIC IRON AND STEEL IN 1971
All figures adjusted to 1971 production of steel.
Process
Basic iron and steel (SIC 331)
Sintering
Pelletizing
Coke oven underfiring
Blast furnace injection
Open hearth firing
Air preheating
Scrap preheating
Working and heat treating
Soaking pits
Reheat furnaces
Annealing, tempering, normalizing
Coating and finishing
Galvanizing
Tin and terne plate
Prepainted steel stock
Other
Total
Energy
1350
6.3
3.2
185
49.6
153
183
(b)
550
176
267
107
7.6
4.1
0.07
0.26
0.46
Natural Gas,(a)
1012 Btu/year
691.8
0.63
3.2
1.6
23.4
56.5
18.3
(b)
291
93
141
57
3.8
2.0
0.04
0.13
0.23
Retrofit
Group
—
1
1
1
2
1
1
1
—
1
1
2
--
1
1
2
2
(a) Estimated amount of purchased natural gas used using Reference 30.
(b) Not available.
-------
52
• The working and heat treating operations have already been
discussed to some degree. The cast ingot starts in the
soaking pits where it is heated for primary forging opera-
tions. During the working processes the metal may have to
be reheated in reheat furnaces several times to relieve
stresses and allow further shaping. Finally, annealing,
normalizing, and tempering operations (controlled heating
and cooling) are used to provide the desired mechanical
properties of the metal.
• Many coating and finishing operations are done electro-
lytically and as such require little or no energy other
than electricity. The operations listed either use hot
baths or molten coating material or require drying or
curing of the coating after application.
As of 1969 there were 49 integrated steel mills in the United States
with a total of 178 blast furnaces' '. The largest complex of blast furnaces
is at U.S. Steel in Gary, Indiana, with 12 furnaces. Secondary plants are more
numerous than integrated plants, but are generally smaller and more specialized.
In 1969 there were 100 secondary plants in the United States with a total of
408 melting furnaces. The largest of these, Cameron Iron Works in Houston,
Texas, has 19 melting furnaces.
Ferrous Foundries (SIC 332)
The ferrous foundry industry produces steel castings and iron cast-
ings of gray, malleable, and ductile grades. Ferrous charges are melted in a
variety of fuel-fired furnaces and electric furnaces. Fuel-fired furnaces
include cupolas (coke fired), open hearths (gas and oil fired), air furnaces
(pulverized coal fired), and reverberatory furnaces (gas and oil fired).
Electric furnaces include electric arc and electric induction. Table 8 gives
relative numbers and installed capacities of the various types of melting
furnaces in 1965. Because total steel and iron production has not changed
significantly since 1965, these figures should still be relatively accurate.
-------
53
TABLE 8. MELTING FURNACES IN THE FERROUS
FOUNDRY INDUSTRY IN 1965(25)
Furnace
Cupolas
Open hearths
Air furnaces
Electric arc
Electric induction
Number
2275
75
117
1121
2273
Total Capacity,
tons/charge
22,822
3,506
3,337
3,426
1,709
The two primary uses of energy in ferrous foundries are for melting
and heat treating. Table 9 gives a breakdown of the amounts of various fuels
used for these two operations in 1964. Though these figures are relatively
old, the various proportions should be relatively unchanged. About 80 percent
of the energy use in the ferrous foundry industry is for melting and 10 percent
is used in heat treating operations. The remainder is used for a variety of
operations including mold and core preparation, ladle heating and drying, and
heating makeup air.
TABLE 9. ENERGY USE IN THE FERROUS FOUNDRY INDUSTRY
(1012 Btu/year)(1964)
Process
Gray and ductile iron
Melting
Heat treating
Malleable iron
Melting
Heat treating
Steel
Melting
Heat treating
Gas Oil
8.2
1.9
0.17
1.8
4.0 4.0
5.7
Electricity Coal Coke
1.6 -- 55.0
0.7 7.8 1.6
0.2 1.8
4.2
Total
64.8
1.9
10.2
3.8
12.2
5.7
-------
54
Open-hearth and reverberatory melting furnaces represent the most
significant potential use of low or intermediate energy gas in the ferrous
foundry industry. Both should be relatively easy to convert and would be
classed under Group I in the analysis of Section IV-1.
The majority of heat treating operations in the foundry are direct-
fired applications where surface finish of the casting is not critical. These
applications should also be relatively easy to retrofit to low or intermediate
energy gas and as such would also be placed in Group I. In some heat treating
operations, however, surface finish is important and protective atmospheres
are necessary requiring either very close combustion control or indirect firing
with radiant tubes. In either case, retrofit would probably require greater
consideration and modification than for more easily converted processes, hence
these processes were placed in a Group II classification.
Primary Nonferrous Metals (SIC 333)
The nonferrous metals industry consists chiefly of the manufacture of
aluminum, copper, lead, and zinc. Other metals included in this classification
are
Antimony
Cadmium
Cobalt
Manganese
Molybdenum
Nickel
Tin
Titanium
Tungsten
Various ferroalloys.
Table 10 gives energy requirements of the various segments of the
nonferrous metals industry by standard industrial classification for different
fuels.
-------
55
TABLE 10. ENERGY REQUIREMENTS OF THE PRIMARY NONFERROUS
METALS INDUSTRY (1012 Btu/year) '^ (1971)
SIC
333
3331
3332
3333
3334
3339
Industry Group
Primary
Primary
Primary
Primary
Primary
nonferrous metals
copper
lead
zinc
aluminum
Nonferrous (n.e.c.)
Oil
16.
14.
0.
--
1.
0.
1
0
1
0
9
Natural
Gas
205
45
4
20
124
10
.4
.0
.5
.8
.7
.3
Coal
37
4
1
15
15
0
.8
.7
.8
.4
.7
.1
Total(a)
277
66
13
42
142
11
.2
.8
.3
.3
.9
.9
(a) Includes other energy sources than those shown.
Copper is smelted from sulfides such as chalcocite and chalcopyrite.
The resulting metal is then used for a variety of products, the major ones being
copper wire, brass, and bronze. Table 11 shows the various major fossil fuel
energy-consuming processes in the making of copper along with the estimated
energy usage in 1971. Also shown is the estimated usage of natural gas in these
processes, which serves to indicate the potential substitution of low or inter-
mediate energy gas from coal. Other major fuels used are coal, oil, and
electricity.
The smelting operation is by far the largest single energy consuming
process in copper production. The raw ore in many cases may be roasted or
wintered prior to smelting. Smelting is usually done in large reverberatory
furnaces, producing "blister copper" which is 98 percent pure copper. Both the
roasting and smelting operations use predominantly natural gas as an energy
.-.ource and by the analysis of Section 1V-1 should be relatively easy to convert
to low or intermediate energy gas from coal.
I'libLer copper is then refined in basically one of two ways: fire
refining ii, barrel-shaped or reverberatory furnaces or electrolytic refining
if greater purity is required.
-------
56
TABIE 11. PROCESS ENERGY REQUIREMENTS FOR COPPER
MANUFACTURE IN
Primary copper (SIC 3331)
Roasting concentrates
Primary smelting
Secondary smelting
Fire refining
Melting cathodes
Hot working, wire
Hot working, brass
Melting, brass
Foundry melting
Total
Energy,
1012 Btu/year
66.8
2
24
2
5
6
3
4
3
4
Natural Gas,
percent
75
90
90
90
90
90
80
80
60
95
Ease
of
Retrofit*-13)
--
1
1
1
1
1
2
2
1
1
(a) Figures from reference adjusted to 1971 production of copper as
listed in 1971 Minerals Yearbook.
(b) See Section IV-1.
The fire refining process should be relatively easy to convert to
lower energy gases. The blister copper is melted in the furnace to oxidize
impurities, and then the oxygen is reduced to around 0.1 percent by thrusting
poles of green wood into the bath. The initial steps in electrolytic refining
are essentially the same as the fire refining process. Once the oxygen has
been reduced to 0.1 percent, the carbon is cast into anodes and placed in an
aqueous bath where, with an electric current, pure copper is collected on the
cathode. The melting of the cathodes represents another potential area for
substitution of low or intermediate energy gas, as this process usually takes
place in gas or oil-fired furnaces.
-------
57
Most of the refined copper (60 percent)v ' goes to wire mills for
copper wire. Primary energy uses in wire mills are heating the copper bars to
around 1600 F (870 C) in direct gas-fired furnaces prior to rolling and
annealing of the wire after final drawing in radiant tube furnaces with cracked
ammonia atmosphere. The direct-fired preheating would be considered relatively
easy to convert to lower energy fuels, but the annealing operations would be
considered more difficult. About 1.5 x 10 Btu/ton (0.43 kcal/metric ton) of
copper are required for preheating and about 0.6 x 10 Btu/ton (0.17 kcal/
metric ton) are required for annealing.
Most of the remaining copper goes for the production of brass and
bronze. Areas of energy consumption are in melting brass or bronze prior to
casting and in subsequent annealing operations. Melting operations are done
(21)
in gas or oil-fired reverberatory furnaces or electric furnaces. In 1965
about one-third of the copper used in this manner was melted in reverberatory
furnaces which predominate in older mills. These furnaces should be adaptable
to lower energy fuels. Annealing operations are often carried out in indirect-
fired radiant tube furnaces and as such would be more difficult to convert to
lower energy fuels as was pointed out in Section IV-1.
Lead bullion (97 percent pure) is recovered from galena and other
minerals. The ore is mixed with coke fines and roasted in a Dwight Lloyd
sintering machine, then crushed, mixed with coke, and charged to a blast furnace,
The result is lead bullion which is then refined by oxidizing impurities in a
reverberatory furnace. About 50 percent of secondary or recovered lead is
refined in the same way as lead ores.
Table 12 gives estimated energy requirements for various processes
in lead production. Both sinter ignition and refining should be amenable to
conversion to lower energy gases from coal.
-------
58
TABLE 12. PROCESS ENERGY REQUIREMENTS FOR
LEAD MANUFACTURE IN 197l(2D(a)
Primary lead (SIC 3332)
Sinter ignition
Primary refining
Secondary refining
Total
Energy,
1012 Btu/year
13.3
0.7
3.4
3.8
Natural Gas,
percent
34
40
90
90
Ease of
Retrofit^)
--
1
1
1
(a) Figures from reference 30 adjusted to 1971 production of lead as
listed in 1971 Minerals Yearbook.
(b) See Section IV-1.
Primary zinc ("slab zinc") is made from sphalerite (ZnS), which is
67 percent zinc. The zinc ore is first oxidized by roasting, then mixed with
coke and reduced to zinc with carbon monoxide in sealed, heated retorts. Prior
to charging in the retort, the ore is sintered in a Dwight Lloyd sintering
machine.
There are basically three types of retort in use:
Vertical retort
Belgian retort
Electrolytic retort.
The electrolytic retort requires only electric energy, though there is a sub-
sequent requirement for gas or oil in melting zinc cathodes from the
electrolytic cell.
Both the Belgian and vertical retort method require gas or oil heat.
The Belgian process involves heating zinc oxide sinter with a carbonaceous
reducing agent in cylindrical retorts and condensing the zinc vapor emitted
from one end. For heating purposes, about 60 x 10 Btu/ton zinc is required
in the Belgian retort in natural gas. The vertical retort process is continuous
-------
59
and consists of charging "coked bilcks" of concentrate into a vertical cylinder.
The process requires about 18 x 10 litu/ton zinc (4.98 x 10 kcal/metric ton) . ^
The molten zirv from the retort is then refined in a reverberatory
furnace requiring an additional 3 x 10 Btu/ton zinc (0.83 x 10 kcal/metric
(21)
ton). Table 13 gives process energy requirements for zinc production.
Essentially, all the proot sses shown should be relatively easy to convert to
low or intermediate energy gas from «.oal.
TABLE 13. PROCESS ENERGY REQUIREMENTS IN
ZINC MANUFACTURE IN 197l(21)(a)
Primary zinc (SIC 3333)
Roasting concentrates
Sinter agglomeration
Smelting (retorting)
Melting cathode zinc:
Refining
Total
Energy,
10 J2 Btu/year
42.3
0.9
0.7
27.7
0.5
2 . 8
Natural Gas,
percent
25
90
45
85
90
90
Ease of
Retrofit(b)
--
1
1
1
1
1
(a) Figures from Reference 30 adjusted to 1971 production of zinc as
listed in 1971 Minerals Yearbook.
(b) See Section IV-1.
Aluminum is made from alumina, which in turn is derived from bauxite.
Alumina is separated ouc of bauxite- almost universally by the Bayer process
involving the digestion of bauxite in a pressurized hot caustic soda solution.
In the process, alumina (A^O-j) is converted to soluble AfO?Na and subsequently
is separated from the insoluble impurities--"red muds". Alumina trihydrate is
then precipitated out of solution and converted to anhydrous alumina by calcina-
tion in a rotary kiln at around 2200 F (1450 C). Except for calcining, energy
for the Bayer process is supplied from turbogenerators in the form of
-------
60
2 2
electricity and 110 psi (760 N/m ) exhaust steam generated by 950 psi (6500 N/m )
boilers. The rotary kilns for calcining are usually gas fired and require
approximately 10 x 10 Btu/net ton(2.77 x 10 kcal/metric ton) of alumina.
In the United States the method used for purifying alumina is pri-
marily the Hall Heroult electrolytic process. The primary fossil-fuel energy
requirement for this process is in the baking of carbon anodes. A great deal
of electricity is also necessary (15,000 kw-hr/net ton aluminum)(21) and may
be supplied from steam-driven turbogenerators on-site.
Other energy-consuming processes in aluminum manufacture are melting
in foundries prior to casting and for various heat treating operations. Melt-
ing is usually carried out in various types of reverberatory and crucible-type
furnaces, though electric melting is becoming more popular due to the propensity
for molten aluminum to absorb hydrogen from fossil fuels. Heat treating pro-
cesses include homogenizing, annealing, solution heat treating, and artificial
aging.
Table 14 gives energy consumption in different processes in aluminum
production in 1971, along with the estimated usage of natural gas. Calcining
and melting operation should be relatively easy to convert to lower energy gases
from coal. Heat treating operations, however, may require more consideration,
as surface finish often is important, requiring the use of indirect heating
through radiant tube burners or in circulated air furnaces with close temperature
control (± 5 to 10 F)C'21) (-fc 3 to 6 C).
V-3. Petroleum and Coal Products Industry (SIC 29)
The petroleum and coal products industry is the third largest energy
consumer of the basic industries in the United States. The industry can be
broken down into five major areas where relative fossil fuel energy require-
ments in 1971 are shown in Table 15. Of the five sectors, petroleum refining
(SIC 2911) consumes by far the most energy (95 percent of the industry total) .
The other four sectors are relatively minor consumers, accounting for the
remaining 5 percent of the industry consumption.
-------
61
TABLE 14. PROCESS ENERGY REQUIREMENTS IN ALUMINUM
MANUFACTURE IN
Primary aluminum (SIC 3334)
Calcining
Scrap melting
Foundry melting
Heat treating
Total
Energy,
10 ^ Btu/year
142.9
63.0
7.0
3.8
5.9
Natural Gas,
percent
87
90
80
80
75
Ease of
Retrofit^)
--
1
1
1
2
(a) Figures from Reference 30 adjusted for 1971 production of aluminum as
listed in 1971 Minerals Yearbook.
(b) See Section IV-1.
TABLE 15. ENERGY USAGE IN THE PETROLEUM AND
COAL PRODUCTS INDUSTRY (SIC 29) ")
(1012 Btu/year)
SIC
29
2911
2951
2952
2992
2999
Description
Petroleum and coal products
Petroleum refining
Paving mixtures and blocks
Asphalt felts and coatings
Lubricating oils and greases
Petroleum and coal products (n.e.c.)
Oil
86.3
71.2
9.4
4.4
1.0
0.3
Gas
1363
1332
10.6
12.5
1.6
6.5
Coal
9.3
9.0
—
0.2
0.1
0.1
Total (a)
1510
1440
36.5
22.5
5.1
7.2
(a) Includes sources other than those listed.
The use and disposition of purchased energy in petroleum refining is
difficult to quantify, as refineries vary in design and disposition of the
various by-product fuels that are inherently generated in petroleum processing,
-------
62
The major energy-consuming processes in the refining operation, however, are
distillation, catalytic cracking, thermal cracking, and steam raising in
boilers .
In the distillation process, crude oil is separated into various
components such as naptha, gasoline, kerosene, and lubricating oils, leaving
heavy residual as the remaining part of the crude feed. The two basic methods
of distillation are one-stage pipe stills for long residuums or high per-
centage of bottoms and two-stage pipe stills for short residuums or low per-
centage of bottoms.
Prior to entering the still the crude feed is heated to approximately
650 F to 665 F (344 to 352 C) in a pipe-still furnace. This furnace is the
major area of fuel use in distillation and usually consists of a refractory-
lined steel-jacketed shell with a tube-type heat exchanger through which the
crude oil is pumped. The furnace is generally fired with gas or oil in some-
what the same manner a watertube boiler would be fired.
After heating in the furnace, the crude oil is fed into the pipe
still, which consists of a tall, vertical cylinder containing numerous bell-
cap fractionating trays. The different constituents in the hot crude oil
vapor condense out as the vapor passes through the fractionating trays, which
are held at the different saturation temperatures of the constituent they are
designed to separate. The fractionated liquids are then stripped of steam
absorbed in the distillation process, then cooled and sent to storage.
Cracking operations are employed to gain increased yields of gaso-
line by cracking heavier gas oils (anything from kerosene to heavy distillates)
from the distillation still. The thermal cracking process involves heating the
heavy feed stock to 900 F to 1000 F (480 to 535 C) under pressures of 250 to
750 psig in a process similar to the pipe still used in distillation. The
vapors in the heavier gas-oil feed crack to lighter gasoline fractions and
the remainder comes off as tar. Gasoline yields of up to 65 percent* can be
obtained with a kerosene feed and as low as 20 percent with a heavy distillate
(26)
feed . The main fuel requirements are in the heating of the feed stock in
furnaces similar to those used for pipe stills.
Thermal cracking has given way to more modern catalytic cracking
operations which give higher gasoline yields and octane ratings. In 1959
* Fraction with boiling point lower than gasoline.
-------
63
less than one-third of the cracking in the United States was done in thermal
units, with no new units planned, resulting in a much lower fraction of
thermal cracking being done today.
Catalytic cracking is composed primarily of three processes--cracking,
regeneration, and fractionation. After start-up, a specialized procedure
involving auxiliary burners, the cold feedstock is passed through heat ex-
changers where it picks up heat from tower sidestreams. It is then injected
into a hot catalyst stream and on to a reactor which is held at 900 to 975 F
/O£N
and 10 to 12 psig^ . Once in the reactor the feed oil, in constant contact
with catalyst, is cracked into lighter gasoline fractions. Conversions of up
to 92 percent' ' have been obtained in commercial practice.
(23)
Three basic types of catalyst are usually employed:
• Silica-magnesia, which gives highest yields
• Silica alumina, which gives the highest octane product
• Bentonite clays, which are intermediate in both respects.
Catalysts are continuously regenerated by burning off the carbon deposited
from the oil at temperatures normally maintained between 1050 and 1125 F
9 / n/r \
(562 and 610 C) and 10 psig (69 N/m ). Close temperature control is
necessary to prevent catalyst damage. The flue gas resulting from burning
off the carbon is normally processed through a cyclone separator and then
passed through a waste-heat boiler resulting in significant energy savings
for the refinery as a whole. No fuel is usually required for catalytic
cracking other than the carbon on the catalyst, though some plants have added
gas-fired furnaces for additional capacity.
Naptha reforming is also an important source of higher octane gaso-
line fuels in petroleum refining. As in the case of cracking of heavy frac-
tions, naptha can be reformed thermally or with catalysis. Since 1949, when
catalytic reforming was used for aromatic production, thermal reforming has
(26)
been rapidly replaced . Catalytic reforming of naptha has superseded
catalytic cracking as the principal source of gasoline.
The thermal reforming process is similar to the thermal cracking
process. The naptha feed is pumped through a furnace heat exchanger and
heated to around 1000 F (535 C) at 500 to 1000 psig (3450 to 6900 N/m2).
The naptha is thus cracked and separated into desired fractions in a recti-
fying column.
-------
64
Catalytic reforming (or hydroforming) involves catalytic naptha
reforming in the presence of a large volume of hydrogen. The process basically
involves the same steps as for catalytic cracking. The naptha feed is heated
2
to around 900 F (480 C) at pressures from 250 to 500 psig (1740 to 3480 N/m ),
depending on the particular process, and is reacted with a catalyst causing
the naptha to crack. The cracked naptha is sent to a fractionating column,
where various desired constituents are separated. There are various hydro-
forming processes in use by refineries, though they are basically of two types:
• The platforming process, which uses a fixed-bed catalyst
and in which the naptha feed is reheated and rereacted
with catalyst several times prior to fractionating. An
expensive platinum catalyst is used which needs regener-
ating only about once every 6 months.
• The fluid hydroforming process, which uses a less expen-
sive fluidized catalyst bed that is continuously
regenerated. Only one heating of the naptha feed is
necessary, since the continuously circulating hot
catalyst bed maintains constant temperatures in the
reactor.
There are various other processes in a petroleum refinery that require
energy inputs in various forms. These processes include alkylation and iso-
merization, an important process for making high-grade gasoline and aviation
fuel; polymerization, an adjunct to catalytic cracking for producing high-
octane gasoline components; various methods of refining and dewaxing lubricating
oils; and finishing operations on various fuels such as automotive and aviation
gasoline, kerosene, jet fuel, furnace oil, and diesel fuel. The energy require-
ments for the processes discussed and for some not discussed are given in
Table 16 for various energy forms.
Determining the potential for utilization of low or intermediate
energy gas from coal in the petroleum industry is difficult. Processes and
plant designs are quite varied, making it difficult to determine the disposition
of various waste gases inherently produced in refineries that are used as fuel
-------
65
sources. The three basic demands for energy in refining are for heating feed
streams in tube-type furnace heat exchangers, raising steam in boilers, and
catalyst regeneration.
TABLE 16. ENERGY REQUIREMENTS OF VARIOUS PETROLEUM
REFINING PROCESSES PER BARREL OF FEED<26)
Process
Description
Distillation
Single-stage still
Two-stage still
Thermal cracking
Catalytic cracking
Thermal reforming
Catalytic reforming
Polymerization
Alkylation
Steam,
Ib
8
15
5
110
70
33
100
200
Electricity,
kwh
0.30
0.40
7.0
0.35
1.0
0.72
1.0
4.4
Fuel,
Btu
100,000
140,000
620,000
0
300,000
450,000
0
150,000
Both the tube-type furnace heat exchangers and industrial boilers could
become significant users of low or intermediate energy gas from coal, and neither
should be very difficult to retrofit. Many times in catalyst regeneration, the
carbon deposited from the oil being burned off will furnish the bulk of the fuel
requirement. Those cases in which additional fuel is necessary could represent
a significant potential use of gas from coal, however.
V-4. Stone Clay and Glass Products (SIC 32)
The stone,clay,and glass products industry consists of the manufac-
ture of glass; cement; various structural clay products including brick,
ceramic wall and floor tile and clay refractories; pottery and related products;
and concrete gypsum and plaster products including concrete block and brick,
-------
66
ready-mix concrete, lime, and gypsum products. Table 17 gives a breakdown of
the fossil fuels used in the manufacture of some of the major energy-consuming
products in this industry.
TABLE 17. ENERGY REQUIREMENTS IN THE STONE, CIAY, AND
G1ASS PRODUCTS INDUSTRY (1012 Btu/year)(1971)
SIC
32
3211
3221
3229
3241
3251
326
3273
3274
3275
Industrial
Segment
Stone, clay, and glass products
Flat glass
Glass containers
Pressed and blown glass
Cement
Brick and structural tile
Pottery and related
Ready-mix concrete
Lime
Gypsum products
Oil
117.9
0.4
8.8
2.0
41.9
3.1
1.0
23.2
4.1
5.3
Natural
Gas
726.4
52.0
119.9
56.8
208.1
59.1
16.0
5.5
39.2
30.5
Coal
252.4
3.8
--
0.1
178.8
3.1
0.7
0.1
47.3
1.4
Total (a)
1219.1
55.9
130.3
63.1
429.3
76.7
23.9
80.1
90.0
38.5
(a) Total includes energy sources other than those shown.
Glasi
The glass industry can be divided into three basic sections: flat
glass (SIC 3211), glass containers (SIC 3221), and pressed and blown glass
(SIC 3229). Glass is made by melting a mixture of silica, soda ash, and
limestone, together with cullet (broken glass) in an enclosed refractory-
lined hearth-type furnace. The melting operation is by far the largest energy-
consuming process in glass manufacture, consuming from 75 to 80 percent of the
process energy requirement or 65 to 75 percent of the total energy requirement
(assuming 5 to 10 percent nonprocess energy use) ^ '.
-------
67
Natural gas is the preferred fuel in glass making, accounting for
approximately 90 percent of the energy requirement of the industry (Table 17).
Oil and electricity are generally used as standby fuels, with a trend toward
increasing use of electricity. The fuel is fired into the furnace through
either end or side ports above the surface of the charge in the melting and
refining end of the furnace, which is usually separated from the working end
by a bridge wall. Good combustion control is necessary in glass melting to
maintain the relatively high temperatures and uniform heat release required
by the process for maintaining product quality.
When natural gas is used as a fuel, the air is usually preheated by
regenerators outside the furnace. If low or intermediate energy gas were
used, double-compartmented regenerators may be necessary for preheating both
air and fuel. This could create problems in furnaces designed for natural
gas, as extensive modification may be necessary to gain sufficient space for
the regenerators.
At one time, low energy gas from coal was a common fuel in glass
manufacture. However, since natural gas has become the dominant fuel the
technology of glass melting has advanced considerably. The required melting
area per ton of glass has decreased by nearly a factor of two in the past
(22)
20 yearsv ', largely due to improvements in furnace design and batch formulae,
but the availability of high energy fuels such as natural gas also has been an
important factor in obtaining more rapid batch melting. This could also be an
important factor in switching to lower energy gases.
Other significant areas of energy use in glass manufacture are
annealing, tempering, flame polishing, and various heating operations carried
on throughout the plants. These operations generally make use of many small
burners with a high degree of control and are generally not considered attrac-
tive processes for immediate retrofit to low or intermediate energy gas.
Table 18 gives estimated energy requirements for melting as well as annealing
and other processes in the manufacture of glass.
Cement
The making of Portland cement requires the combination of two basic
types of materials: those primarily containing calcium and those containing
-------
68
silica alumina and iron. The calcium-bearing material is most commonly lime-
stone but may be marble, marl, or waste sludge from lime plants. The three
other constituents are generally derived from argillaceous materials such as
(?
clay, shale, or bauxite
TABLE 18. PROCESS ENERGY REQUIREMENTS IN
GIASS MANUFACTURE IN
Flat glass (SIC 3211)
(a) Melting
(a) Annealing
(a) Other
Total
Energy,
1012 Btu/year
55.9
38.8
3.9
13.2
Ease of
Retrofit^)
2
3
3
Glass containers (SIC 3221)
(a) Melting
(a) Annealing
(a) Other
130.3
107.7
11.1
11.5
2
3
3
Pressed and blown glass (SIC 3229)
(a) Melting
(a) Annealing
(a) Other
63.1
52.8
3.5
6.8
2
3
3
(a) Process energy requirements based on 1971 energy consumption using
proportions of Reference (21) .
(b) See Section IV-1.
The raw materials are blended in the proper proportion and ground to
a fine consist in one of two manners, depending on the type of process. The
wet process is used generally where the raw materials already contain
-------
69
significant moisture, and additional water generally is added to form a fine
slurry during the grinding operation. In the dry process, the raw materials
are blended and dried prior to grinding to a fine powder.
Once ground, the raw materials are fed to long rotary kilns where,
under extreme heat up to 3000 F^ ' (1660 C), they are transformed to round,
marble-sized, glass-hard balls called "clinker". This clinker is actually
Portland cement which undergoes finish grinding to the fine powder normally
associated with the final product. Once mixed with an aggregate, such as
gravel or sand, the cement becomes concrete and sometimes is sold as "ready
mix" concrete.
The kiln represents the single major user of fossil fuel energy in
cement manufacture. The rotary kiln consists of a long (possibly over 500
feet or 150 meters)refractory-lined steel cylinder sloped at a slight angle
of about 1/2 to 3/4 inch per foot length (4.3 to 6.3 cm/m). Gas, oil, or
pulverized coal fuel is fired at the lower end of the kiln and the raw materials
are fed at the top end. The kiln rotates slowly causing the raw materials to
migrate to the lower discharge end, where they are taken off as clinker. The
principal difference between the wet and dry process is the water content of
the raw material feed. The energy requirement for the wet process is about
40 percent higher than the dry process (1.4 million Btu per barrel of clinker
versus 1.0 million Btu per barrel).
The most critical aspect in firing cement kilns is to tailor the
flame pattern to maintain the proper burning area in a given section of the
kiln. Once the combustion has been balanced, there is generally little problem
in maintaining the balance for a given fuel. Changing to lower energy gases
from coal, however, would require rebalancing the kiln and may require some
testing. Also, lower flame temperatures of some low energy gases from coal
could result in a slight reduction in capacity. Rotary kilns, however, should
be fairly easy to adapt to low or intermediate energy gas. Since kilns are
the only major consumer of fossil fuel energy in cement manufacture, the
figures given in Table 17 can be used to estimate the potential for use of low
or intermediate energy gas from coal.
-------
70
Lime
The lime industry is similar to the cement industry in that kilns
represent the single major consumer of energy, with relatively small amounts
of energy used elsewhere. The kiln is used to calcine limestone (CaCCO or
dolomite (CaCO -MgCO ), resulting in so-called "quicklime" (CaO). Since quick-
lime is highly perishable, water is generally added producing hydrated or
slaked lime (Ca(OH)2).
Three types of kilns are currently in use for calcining limestone.
• By far the greatest amount of limestone is calcined in
rotary kilns similar to those used in the cement industry.
The same discussion used for cement kilns applies also
to rotary-type lime kilns relative to the potential for
utilizing low or intermediate energy gas from coal.
* Another common type of kiln in lime production is the
vertical type consisting of a vertical lined steel shaft.
The shaft is stationary and limestone is fed in at the
top through a charging door, where it progresses down
through the heating zone and is taken off at the bottom
as lime. In shaft-type kilns, only the hot combustion
gases pass through the charge and the actual combustion
takes place outside the kiln.
• A fairly recent innovation in the lime industry is the
rotary-hearth kiln. This type of kiln has a rotating
horizontal hearth which is round and may vary in width.
Material is fed in at one point on the circumference of
the unit and calcined lime is removed at another point
as the hearth rotates.
Utilization of low or intermediate energy gas from coal in lime
kilns should not present serious difficulty. National Lime and Stone Company
in Carey, Ohio, and Glenn Gary Manufacturing in Schumakersville, Pennsylvania,
are two lime plants that currently still have gasifiers either in active or
standby service. Potential problems of retrofitting rotary kilns are the same
as those in the cement industry discussed previously. Retrofitting of
-------
71
shaft-type kilns should be relatively straightforward since the combustion
process takes place outside the kiln itself and, hence, the balancing of the
combustion process with the kiln operation would not be as critical as with
a rotary kiln.
A typical large lime plant would be capable of producing around
150,000 tons of lime per year (136,000 metric tons/year) at an average energy
(22")
consumption of around 6 million Btu per tonv ' (1.66 million kcal per metric
ton). Assuming operation for 347 days per year, this results in an average
energy requirement of 2.6 billion Btu per day (0.66 billion kcal/day), which
would require one Lurgi, Winkler, or small Koppers-Totzek unit or two large
Wellman-Galusha gasifiers (see Table 3).
Ceramics
The ceramics industry consists of a variety of products including
brick, clay refractories, ceramic tile, and various pottery products. The
primary uses of energy in this sector are for firing kilns and various baking
and drying operations.
The most important segment of the ceramics industry from the stand-
point of energy use is brick and structural tile (SIC 3251). Energy require-
ments for this segment are given in Table 16. The products in this segment
include common brick, face brick, and glaze brick. After the raw materials
(clay, some crushed minerals, and water) are mixed and set, they are dried in
drying rooms or tunnel driers. Heat can be supplied from a variety of sources
including direct-fired air heaters, exhaust heat from a tunnel kiln, steam
coils, or infrared heaters.
Kilns in this industrial segment generally are of three types: field,
beehive, or tunnel kilns. Field kilns use green brick which form firing arches
by being stacked in a predetermined pattern, with firing by gas or oil into
each arch on a schedule of 5 to 9 days. Beehive kilns are large, circular
downdraft kilns gas or oil fired on cycles of 24 to 72 hours. Tunnel kilns are
large car-type kilns in which the product is fired continuously. Low or inter-
mediate energy gas from coal should be readily adaptable to most drying and
kiln operations in the structural brick and tile industry and in the ceramics
industry in general. In many cases, however, the product may require a fine
-------
72
finish, such as in whiteware and other pottery products, and firing or drying
operations will require careful consideration in switching to a low or inter-
mediate energy gas. This is especially true if the sulfur content in the gas
is significantly higher than in the fuel being replaced. In general, kiln
firing and drying operations on unfinished parts would be considered as
Group I under the analysis of Section IV-1, while firing operations on finished
parts where high surface finish is required would be considered Group 2 or 3.
-------
SECTION VI
UTILIZATION OF LOW OR INTERMEDIATE ENERGY GAS
FROM COAL IN ELECTRIC POWER GENERATION
-------
73
VI. UTILIZATION OF LOW OR INTERMEDIATE ENERGY GAS
FROM COAL IN ELECTRIC POWER GENERATION
Though the thrust of this study is to analyze the potential for
utilization of low or intermediate energy gas from coal in the industrial
sector, a cursory analysis of its applicability to the utility sector is
presented here.
A variety of energy sources are used in the United States for the
generation of electric power including fossil fuels, nuclear fuels, hydro-
electric energy, and geothermal energy. The total energy input to utilities
12 12
in 1971 was estimated at 17,563 x 10 Btu (4426 x 10 kcal). Of this amount,
12
fossil fuels (natural gas, oil, and coal) accounted for 14,239 x 10 Btu
12 (27}
(3588 x 10 kcal) or 81.5 percentv ' of the total. The relative amounts of
(28)
energy supplied by the three major fossil fuels in 1971 were
Fuel Percent Amount
Coal 55 327.9 x 106 short tons
Natural gas 29 3993 x 10^ cubic feet
Oil 16 396.2 x 106 barrels.
Fossil fuel generating plants are presently of two general types,
steam electric or gas turbine. Steam electric plants use large boilers to
generate high-pressure steam which is then used to generate power by expansion
in a steam turbine. Gas turbine plants burn the fossil fuel directly in the
turbine to generate power.
Fossil-fuel-fired steam electric plants account for the major amount
of electrical generation in this country, with 253,819 megawatts of installed
capacity. The total combined fossil fuel and nuclear installed capacity is
/ogN
284,280 megawatts.v These plants can be coal, oil, or gas fired, and in
many cases they have the capability of firing more than one type of fuel. The
boilers themselves are similar in many respects to industrial watertube boilers
(discussed in Section IV-2), though in general they are many times larger, some-
what more complicated, and tend to operate at higher pressures. Relative to the
retrofit of existing fossil-fueled steam electric plants to low or intermediate
energy gas from coal, there are four broad areas of concern.
-------
74
(1) The operating cycle or load factor (actual power generated
divided by the potential power generation based on installed
capacity of electrical generating plants) can be generally
classified into three areas:
a. Base-load plants operate at load factors of around
70 percent and greater and usually consist of the
most efficient, hence largest, units in a power
network. An attempt is made to keep these plants
operating at peak efficiency or near full capacity
as much of the time as possible. These plants tend
to be the newest, largest, most sophisticated, and
most efficient plants in a power network, and they
typically have a steady demand for fuel.
b. Swing-load or mid-range plants operate at load fac-
tors from around 20 percent to 70 percent and many
times consist of the older or less efficient plants
on a power network. As the name implies, the load
on these plants varies greatly to compensate for
broad variations in power demand. The design of
these plants is usually the same as plants used for
base-load operation.
c. Peaking plants operate at load factors of 20 percent
or less and are used to satisfy the short-term peak
demands for power that occur on a daily basis. These
plants in some cases may be the oldest or least effi-
cient plants in a power network. Some new plants have
been built for this purpose--primarily fired with gas
or oil--with the objective of lowest possible investment
cost; the result is units of moderate steam pressures
and temperatures.
The matching of gasifier output to power-plant demand
needs consideration and will probably require control
systems not found on current gasification systems. From
this standpoint it appears that base-load plants would
be most amenable to retrofit, though swing-load plants
-------
75
with slowly changing load patterns many also be
amenable. Peaking plants may operate at full
capacity for short periods of time and at other
times be almost idle or at very low load. It is
important that these plants be able to respond
quickly to changes in load, and from this standpoint
these plants do not seem readily adaptable to current
gasification technology.
(2) Heat transfer in a boiler takes place by primarily two mechanisms.
The walls of the boiler receive heat primarily from flame radia-
tion and the amount of heat they receive depends a great deal on
the luminosity of the flame. Tube banks usually used to super-
heat and reheat the steam are placed downstream from the combus-
tion zone and receive heat primarily by convection as the hot
flue gases flow through them. The balance between these two types
of heat transfer is an important parameter in determining the
outlet temperature of the steam. Controlling steam outlet tem-
peratures within a specified range is important, as unnecessarily
high temperatures require expensive, high-alloy steam lines with
high mechanical properties and low-steam temperatures result in
poor turbine efficiency and the potential for condensation in
the last turbine stages, which promotes rapid blade erosion.
Several methods are used to control steam temperatures out of
the superheater, e.g., spraying water into the steam lines at
various points in the superheater, using flue-gas dampers to
control the amount of superheater tube surface exposed to the
hot flue gas, and flue gas recirculation. When switching from
fuel with a luminous flame (such as oil or coal) to one with
relatively low luminosity (like low energy gas from coal) the
ability of the boiler to control the steam temperature out of
the superheater is an important consideration. In general, the
lower energy gases from coal, due to lower flame luminosities and
higher flow rates of combustion products, result in less radiation
heat transfer and greater convective heat transfer.
-------
76
(3) Increased flows of flue-gas products could also create a prob-
lem in retrofitting existing power plants to low or intermediate
energy gas from coal. As can be seen from Table 5, the volumes
of flue gas (stoichiotnetric mixture) from intermediate energy
gas burned in oxygen-blown units would be on the order of 5 to
12 percent greater than those for natural gas, and for low-
energy gas burned in air-blown units the volumes would be 30 to
55 percent greater. The ability of the induced and forced draft
fans in a unit to handle the increased volume flow of flue gas
will be an important factor in the amount of derating a unit may
have to undergo in retrofit.
Units originally designed for coal inherently have wider tube
spacings and more generous combustion volumes than those designed
for gas or oil to allow for slag or ash buildup. It is antici-
pated that little, if any, derating would result in coal designed
units as a result of increased flue gas volume from intermediate
or low-energy gas from coal. In more compact units designed for
natural gas or oil, however, retrofit to intermediate energy gas
3
(300 Btu/scf) (2670 kcal/m ) may be possible with little derat-
ing, but, in retrofit to low-energy gas (150 Btu/scf) (1335
o
kcal/m ) some derating would probably be necessary.
(4) The last point to be considered in retrofitting existing fossil-
fuel-steam electric plants (or any power plant, for that matter)
to a substitute fuel is dependability. Utilities for many years
have placed paramount importance on the dependability of power
generating systems. Any gasification plant used to supply fuel
to a power plant would have to be capable of very high on-steam
reliability either through the use of dependable systems or with
many standby units. Unlike a stack gas cleanup system (an al-
ternative to gasification for coal-fired plants), the gasifier
cannot easily be bypassed when failure occurs; this characteris-
tic allows the plant to continue in operation while repairs or
adjustments are made. A standby natural gas or oil fuel supply,
however, may be an attractive supplement to any power plant sup-
plied substantially by a coal gasification plant.
-------
77
Retrofitting gas turbines to low or intermediate energy gas from coal
was discussed in Section IV-4. Gas turbines in utilities are used almost exclu-
sively for peaking service. Frequently, they will be run at full load during
the peak demand hours of the day and completely shut off or idled during off
hours, which would result in a strenuous demand on any gasification system.
Frequently, peaking gas turbine plants are at the same site as larger base-load
or swing-load plants and in these cases could use the same gasification facili-
ties as the larger plants, presenting less of a load-following problem for the
gasifier. For isolated gas turbine peaking plants, however, as for steam
electric peakers, retrofit to coal gasification would require sophisticated
controls not yet available in gasification. Gas turbines would also require
pressurized gasification of at least 200 psig to be attractive. Currently,
only one of the four commercial gasification processes, the Lurgi, operates at
such a pressure (20 atmosphere). The Lurgi gasifier is also the most complex
and reportedly requires considerable maintenance and downtime--characteristics
that are not attractive for power generation.
In summary, the most attractive type of power plant to retrofit to low
or intermediate energy gas from coal is a base-load plant originally designed
for pulverized coal with a subcritical natural or forced circulation-type
boiler. This type of unit is actually quite common and would have the additional
attraction of being located close to a source of coal. The least attractive re-
trofit application would be a plant designed for gas or oil, either steam-electric
or gas turbine, performing peaking or low sx\ring-load duty.
-------
SECTION VII
REFERENCES
-------
78
REFERENCES
(1) 1972 Census of Manufacturers. United States Department of Commerce,
Bureau of Census.
(2) The Making .Shaping and Treating of Steel, United States Steel Corpora-
tion, 1964.
(3) Gas Production. Lurgi Gesellschaf ten, 1968.
(4) SASOL Today. The South African Coal Oil and Gas Corporation Limited.
(5) New Fossil-Fueled Power Plant Process Based on Lurgi Pressure Gasifica-
tion of Coal. Lurgi Gessellschaften, 1970.
(6) Indian Government Sponsored Comparative Study of Commercial Coal Gasifi-
cation Processes --Koppers-Totzek. Lurgi. and Winkler. reproduced by
Koppers, 1969.
(7) Trials of American Coals in Lurgi Pressure Gasification Plant at West-
field. Scotland. Elgin Fifth Synthetic Pipeline Gas Symposium, 1973.
(8) The Production of Gas from Coal Through a Conmerciallv Proven Process.
Farnsworth et al., Koppers, 1973.
(9) Technological and Economic Feasibility of Advanced Power Cycles and
Methods of Producing Non-Polluting Fuels for Utility Power Stations.
Robson, et al., United Aircraft Research Laboratories, 1970.
(10) Personal conversation with Koppers, January 21, 1974.
(11) The Winkler Process for the Production of Low-Btu Gas from Coal.
Banchik.
(12) Wellman-Galusha Gas Equipment. McDowell Wellman.
(13) Gasification of Solid Fuels in the Wellman-Galusha Gas Producer.
Hamilton, G. M., McDowell Wellman, American Inst. of Mining Met. and Pet.
Engineers, 1961.
(14) Sulfur Dioxide Removal From Waste Gases. Slack, A. V., Pollution Control
Review,
(15) Oxygen-Steam Producer Blast. Mitchell, Consolidated Mining and Smelting
Company of Canada, 1946.
(16) United States Federal Register. Vol. 36, No. 247, December 23, 1971.
(17) A Survey of R&D Projects Directed Toward the Conversion of Coal to
Gaseous and Liquid Fuels. IGT, Chicago, 1971.
-------
79
(18') Patterns of Energy Usage in the United States. Office of Science and
Technology, January, 1972.
(19) Design Trends and Operating Problems in Combustion Modification of
Industrial Boilers. D. W. Locklin, H. H. Krause, A. A. Putnam, et al.,
Battelle Annual Progress Report to U.S. Environmental Protection Agency,
Grant No. 802402, 1973.
(20) 1971 Minerals Yearbook. Vol. 1. Metals and Minerals. U.S. Bureau of
Mines.
(21) A Study of Process Energy Requirements for U.S. Industries, American Gas
Association, 1969.
(22) Study of Costs of Production and Potential Future Markets for Low Btu
Industrial Fuel Gas and Industrial Hydrogen, Nelson, Layne, Hein,
Battelle Memorial Institute, 1966.
(23) Petroleum Processing. Principles and Applications. Hengstebeck, Standard
Oil Company (Indiana), McGraw-Hill, 1959.
(24) The Making of Portland Cement, Portland Cement Association, Chicago,
Illinois.
(25) Hot Sulfur Removal From Producer Gas, Schultz, Lewis, 3rd International
Conference on Fluidized Bed Combustion, Hueston Woods, 1972.
(26) Petroleum Refining Manual. Noel, Reinhold Publishing, New York, 1959.
(27) United States Energy Through the Year 2000. Dupree, W, G., and West, J. A.,
U.S. Department of the Interior, 1972.
(28) Steam-Electric _Plant Construction Cost and Annual Production Expenses.
Twenty-Fourth Annual Supplement, 1971, Federal Power Commission.
(29) A Cost Analysis of Air Pollution Controls in the Integrated Iron & Steel
Industry, Barnes, T.M., Battelle Columbus Laboratories, NAPCA Contract
PH-22-68-65, 1969,.
-------
APPENDIX A
INTERCHANGEABILITY OF FUELS WITH
RESPECT TO THE COMBUSTION PROCESS
-------
APPENDIX A
INTERCHANGEABILITY OF FUELS WITH
RESPECT TO THE COMBUSTION PROCESS
SUMMARY
This appendix is a study to estimate the substitutability of
selected low-energy gases from Table 1 to natural gas using what minimum
data is available on their burning characterisitcs. The various fuels
were compared on the basis of the critical flashback velocity gradient.
The value of the gradient is obtained by firing a Bunsen-type burner in
the open and reducing the flow rate until flashback occurs. The gradient
is significant in that it is related to the blowoff limit in an enclosed
system and the burning velocity of the flame along with other combustion
properties associated with flame stability.
Values of the critical flashback velocity gradient were compared
on three bases:
(1) The usual case of the gradient alone for stoichiometric
conditions and for the air/fuel mixture that maximizes
the gradient.
(2) The stoichiometric and maximum values of the gradient
times the high heating value of the stoichiometric
mixture.
(3) The stoichiometric and maximum values of the gradient
times the high heating value of the fuel itself.
The Wobbe number was also calculated and compared for various
selected low-energy fuels and natural gas. This parameter is basically
the higher heating of the fuel divided by the square root of the fuel
specific gravity. It is useful in estimating the substitutability of
various fuels in aspiration-type burners and in burners where pressure
sensitive controls are employed.
Finally, the adiabatic flame temperature was calculated and
compared for each of the gases under consideratio.
-------
A-2
CONCLUSIONS
This analysis aas shown that there should be no problem with stabiliz-
ing medium energy fuels as a substitute for natural gas. In some instances,
for low energy fuels, a problem may result. In most cases, the principal
problem is the increased pressure drops through both the burners, if the design
is unchanged, and through the heat exchanger section. Suitable design changes
in the burner can normally be made. In the case of the heat: exchanger, removal
of some tubes of the heat exchanger with same loss in capacity appears the
simplest approach for retrofitting.
Flame radiation might change sufficiently to cause problems both
with control instruments and with heat flux balance. In some instances, it
might be possible to rectify the problem by changing to a flame type of
different luminosity or by preheating.
No consideration has been given to special uses (such as in glass
forming, wherein the product gas composition, the flame shape, and the flame
temperature must be carefully controlled); in these instances, the details
of the installation must: be considered carefully.
-------
A-3
LIST OF SYMBOLS
D Characteristic burner diameter
G_ Critical flashback velocity gradient as calculated by modification of
techniques presented in Reference A-2 to be published in an upcoming
issue of Combustion & Flame Magazine.
G, Blow-off velocity gradient
U Average flow velocity through the burner
Uf, Average flow velocity at flash back = DGf.
U Average flow velocity at blow off = DG
Q Volume flow rate of combustible mixture
Q£ Fuel flow rate
Qa Air flow rate
Hm Heating value per unit volume of combustible mixture
H.p Heating value per unit volume of fuel
Ha Heating value per unit volume of air
N Number of burners
Ap Pressure drop across burner
a Air/fuel ratio relative to stoichiometric air/fuel ratio
p Density of combustible mixture
pf Fuel density
pa Air density
-------
A-4
INTRODUCTION
Low energy gas obtained from various gasification processes has
been suggested as a source of clean fuel for usage in utilities, industrial
boilers, and industrial processes. One of the problems to be considered in
making such a change to low energy gas is its flame stability in a combustion
system. Tnis discussion is concerned with these stability problems.
It should be noted that this direction of change is the reverse
to that made decades ago; as natural gas became available and the use of
various combustible mixtures from coal-gasification processes were phased
out. A similar more recent process occurred in England with the development
of the North Sea gas supplies. Generally speaking, these fuel changes were
accompanied by changes in types of burners. For instance, in the residential
area the quiet, soft diffusion flame burners in heating units were replaced
by the noisier, harder, but more compact premixed flame burners. Unfortunately,
such changes have reinforced a connotation that low energy gas implies large
combustor systems. Yet, the real reason is that the low burning velocity of
natural gas permits the use of premixed burners that lead to more compact
designs in the case of household heating applications. Tnis example clearly
shows that each potential conversion must be analyzed in detail in order
to draw valid conclusions.
Basically, flames may be either of the premixed flame type, wherein
the fuel and air are uniformly mixed before entering the combustion zone,
or of the diffusion flame type, wherein the fuel and air are separated until
they reach the combustion zone. In the latter case, however, the leading
edge of the flame surface is premixed locally; thus, the flame stability is
related to premixed flame characteristics. The flames may be either laminar,
wherein the rate of mixing (of mass, momentum, and energy) is controlled by
the molecular kinetic properties, or turbulent, wherein the rate of mixing is
controlled to a significant degree by the turbulence characteristic of the
-------
A-5
flowing gases. Most industrial burners have turbulent flames; however, on
a local basis, laminar flame characteristics usually control the local phenomena.
Practical burners can combine features of both types of combustion.
For instance, many premixed burners use fuel-rich mixtures; secondary air is
added to the products of the prerr.ix flame to produce a diffusion flame.
Nozzle mix burners (for example, a fuel jet surrounded by multiple air jets
firing into a burner tile) may show either a diffusion flame or premixed
flame, depending on where the flame is stabilized. Thus, it is difficult to
single out one feature of a combustible mixture that can be considered to
characterize the fuel for comparison purposes, even if the burner is not
changed in the process.
If a comparative parameter must be chosen, however, the most easily
available pertinent parameter seems to be the flash-back velocity gradient.
The value of this parameter is obtained by firing a Bunsen-type burner in
the open. The flow rate of the combustible mixture to a laminar flame is
slowly reduced until the flame flashes back. The velocity gradient at flash-
back in laminar flow is independent of duct sizes over a wide ranee of sizes
and ambient atmospheres. Values are available from one source (A-2)* of
information for a wide variety of fuels, and some combination rules have been
developed for those fuel mixtures not listed.
The great significance of the flash-back velocity gradient in
studies of industrial combustors is that it is related closely to several other
significant combustion parameters. For instance, the flash-back velocity
gradient is proportional to the blow-off limit in an enclosed system and
to the burning velocity of the flame (chemically controlled reaction rate
per unit volume), and it is inversely proportional to the ignition delay
time mentioned by many investigators. It also has been suggested that it
is proportional to the peak frequency of the combustion roar spectrum.
Presentation of Data
Table A-l presents the information on the composition and higher
heating valuesCA~1) Of the various fuels considered in this analysis. Table
A-2 presents computed values of the stability limits considered from three
points of view:
'^Superscripts refer to references at end of Appendix A. A list of symbols
used in this appendix is presented on page A-3.
-------
CO
w
w
PH
O
tO 0)
**-* r-i
O 3
•rl ""^ CU
&3 3 3
rS 4J fa
m
ffl*
O
3
o
r-l CM
gffi
,
c
o
•r-l O
4-1 O
•rl
to
o
o
U CM
o
•-< u
CO
3
fa
CN
•3
in o m
VD CO *•&
, >,
03 -r-l X X
O
i-l
•r-l
14-1
•r-l
CO
to
O
< o o
ly>
co
N
4J
O
H
tfl
r-l
•H 01
oo eu
t^ ^^
3 0
r-5 $£
s£) \Q
r-l O
i— I
in
CO i-l
m i
r-l 1
o** t
CM 1
ON
•
CO O
MD
O 0
m
g
to
CO
4-1
to
1
^4
•1-1
<^
cO
"to ^3
3 **~*
r-l CO
tO tfl
o o
1
C r-l
tfl rt
B rl
r5 3
r-l 4-1
01 cfl
3 Z
to
a
o
rQ
J_J
OS
O
O
M
TJ
^>
43
M
CO
1 1
•4-J
o
tJ
c
ffl
oo
re
CO
CJ
C
O 4-1
•rl 0
4-1
tfl CO
•rl 4-J
O C
0 3
tO O
to B
•rl tO
TD
I—I
4J r-l
3 tO
o B
43 03
4J
•rl CO
5 C
•rl
T3 fO
CO 4-1
4J C
cO O
r-l 0
3
O O
i— i CO
cfl r-l
o <;
x— v /— X
CO rO
x_x x_x
*
y—s
r-4
|
-------
A-7
CO
frf
0
H
O
<
£
£
M
r4
M
PQ
IS
CO
rJ
g
P"H
•
CN
1
<
W
a
PP
<
H
"V
^^
cu •
rO O
rQ Z
O
•> O
33 CO
p-{ |
en
X w
MH
4J -^
C 3
CU 4J
•rl PQ
T3
CO ,
4J
1-1
CJ
o
•H 4J
CU C
> OH
•rll
r*5 T3 0
U CO CU
CO 1-1 CO
M O
I
X!
to
CO
r-4
fa
g
3
5
X
J5?
U
•rl
(H
4-1
CU
1
•rl
J3
U
•r-l
O
4-1
CO
d
0
•rl
4-1 S
Ctf 3
CJ -H
•rl T3
U-4 V
•r< S
M
CO
O
rl
CU
•H
M-l
•H
CO
eg
0
SO CN sO r-l O
.* oo r~» CN en * •*
CN
vj so r- m en O
m so oo CM » r^ \o r-l <}• O
r-l SO r-* O 00 CN
r^ i^ o C C 4J 4J
CO CU CU CO CO
I 00 00 I I
rl >•> >i U M
•rl & R -rl i-l
< 8 S <; 1
4-1
•rl
>
CO
l-i
60
U
•rl
t I j
H"4
•rl
0
0)
p~
CO
I— 1
CU
3
<4-J
0)
JS
4-1
4-1
O
4-1
O
o
1-1
O)
>-l
to
3
cr
cc
o>
^
4J
>^
XI
T3
Ol
•0
•r-l
>
•r-l
T3
1— 1
ff
•rl
4-1
ct)
CU
J3
r(
Q)
rC
00
•rl
aa
/-^
CO
N^X
-------
A-8
(1) The usual critical flash-back velocity gradients
at stoichiometric and the maximum flash-back velocity
gradients are presented.
(2) The stoichiometric and maximum values of the product
of the gradient and the corresponding higher heating
value per cubic foot of fresh mixtures are given.
(3) The stoichiometric and maximum values of the products
of the gradients and the corresponding higher heating
value of the fuel are tabulated.
Also included are the Wobbe Numbers (the high heating value divided
by the square root of the specific gravity) which comprise a useful parameter
in evaluating fuel changes in aspirating-type premix burners or burners in
which pressure sensitive controls are used to regulate fuel or oil flows.
The gradient values of Table A-2 are obtained from Figures A-l,
A-2, and A-.3. Figure A-l presents the flash-back velocity gradients as a
function of the ratio of the fuel/air ratio to stoichiometric fuel/air ratio.
These gradients were constructed using a modification of the techniques
presented in Reference A-2 and data from the same source* Figure A-2 pre-
sents the critical value of the heating rate per unit volume, based on the
fresh mixture properties. Figure A-3 represents the curves of Figure A-l
in an alternate form, each curve being multiplied by the corresponding higher
heating value for the fuel. It is noted that the natural gas curve peaks
close to stoichiometric, while the remainder of the fuel-air mixtures peak
on the fuel-rich side.
A consideration of Figures A-l and A-3 shows that: natural gas
behaves mucn like the fuels that have the lower HHV. Other than natural
gas, the produced fuels (principally consisting of H~, CO, and inerts) line
up roughly in order of the amount of inert present. Considering Figure A-l,
natural gas (with no appreciable inarts) has stability limits lower than all
but two of the listed fuels resulting from various coal gasification processes.
In some uses of low and medium energy gas, the gas may be preheated.
Similarly, there are instances wherein the air is preheated. These cases may
be analyzed in a manner similar to that discussed below for the nonpreheated
cases. However, suitable stability curves similar to those in Figures A-l,
A-2, and A-3 must first be constructed.
* This modified calculation scheme is soon to be published in an upcoming
issue of Combustion and Flame.
-------
A-9
5 xlO5
o
c
o>
•6
o
o
o
0)
o
o
.0
o
U_
2x 10
0.8
Koppers Totzek
(Oxygen steam)
Lurgi (Oxygen steam)
Lurgi (Air steam)
Wellman
Galusha-
(Air steam),
Winkler
(Air steam)
0.9 1.0 I.I
( Fuel/Air)/( Fuel/Air) Stoichiometric
1.2
1.3
FIGURE A-l. FLASH BACK VELOCITY GRADIENT VERSUS MIXTURE RATIO
-------
A-10
Koppers Totzek
(Oxygen steam)
Lurgi (Oxygen steam)
Lurgi (Air steam)
Winkler
(Air steam)
0.9 l.O I.I
(Fuel/Air)/( Fuel/Air) Stoichiometric
FIGURE A-2. CRITICAL HEAT SUPPLY RATE PER UNIT VOLUME
(FLASH BACK VELOCITY GRADIENT TIMES HIGH
HEATING VALUE OF THE STOICHIOMETRIC MIX-
TURE) VERSUS MIXTURE RATIO
-------
A-11
10
o
0>
CO
I
fO
CD
i
r
c
«
TJ
O
L.
C9
>N
O
o
0)
o
o
CO
^
u.
10-
2x10
Koppers Totzek
(Oxygen steam)
Lurgi (Oxygen steam)
Wellman Galusha
(Air steam)
Winkler
(Air steam)
0.8
0.9 1.0 I.I 1.2
(Fuel/Air)/(Fuel/Air) Stoichiometr'ic
1.3
FIGURE A-3. FLASH BACK VELOCITY GRADIENT TIMES GAS HIGH
HEATING VALUE (HHV) VERSUS MIXTURE RATIO
-------
A-12
Discussion of the Data
Three general types of burners are considered in the discussion
of the data--pretnix burners, delayed-mixing burners, and nozzle-mixing burners.
While it is not difficult to distinguish premix burners from the other two
types, the distinction between delayed-mixing burners and nozzle-mixing burners
is sometimes rather vague. For the purpose of this discussion, combustion
in a nozzle-mixing burner will be more intense, with at least a significant
fraction of the combustion taking place within the burner tile. Delayed-
mixing flames will generally extend a considerable distance from the burner
and often ba characterized by a low turbulence level and mixing rate.
Significant amounts of furnace gases might be recirculated into the base of
their flames.
Premix Burners*
Premixed flames are reasonably common in industry and are the
easiest to analyze. The premixed fuel and air are supplied through a duct
to the combustion region. For high firing rates with turbulent flow, the
flame will not hold at the end of the duct. Therefore, a variety of flame
holding systems are used. For burners used in open or semiopen applications,
an annulus to which some of the mixture is diverted is commonly used. Figure
A-4 is an example of sych a burner. The flame in the protected annular
region acts as a pilot to maintain or hold the main flame. In closed systems
(such as tunnel burners), steps, recesses, grids, and other obstacles are
used to hold the flame. Tnese form protected recirculation zones, which
hold the flame and from which the flame spreads.
In all these cases, the key factor is a term proportional to the
velocity gradient at flashback.** As a simple example, consider a closed
* "Usually, a burner applied with gas and air which has previously been mixed,
but sometimes a burner within which the gas and air are mixed before
they reach the nozzle (as opposed to nozzle-mixing burners)."
** Often, in the case of flame holding by obstacles, a delay time is invoked;
this is proportional to the reciprocal of the flash-back gradient.
-------
A-13
FIGURE A-4. PREMIX BURNER
-------
A-14
system where the blow-ofjf velocity gradient, G, , is, say four times the
flash-back velocity gradient, G . For turbulent systems, the gradient is
usually given merely in the form of U/D--the average flow velocity, U, divided
by a characteristic diameter, D. Then, U and U = 4 DG . Now, if the
critical velocity gradient is doubled by a change in fuel, then both the
blow-off and flash-back velocities will double. In many burners, the
equivalent of single or multiple steps are used, so that on premature
flashback the characteristic diameter is decreased as the flame moves up-
stream; this decreases the critical flash-back velocity at the same time as
the flow velocity increases, thus stopping the flashback. For such a design,
increasing the critical value of G will increase the range of flow rates
for stable flames, but not necessarily the heating rate as will be shown.
Considering a single burner with a volume flow rate of combustible
mixture, Q, and a heating value per unit volume of mixture, H , the heat
release rate will be QH . When the critical blow-off condition is reached,
m 23
the heat release rate is given by OH ^ UD H ~ D GH . For a single size of
J m m m °
burner, the key term for comparison is GH . The relative values of this term
are plotted on Figure A-2. It is seen that, on the basis of the heat release
rate at blow-off, the lowest Btu fuel is definitely below natural gas
in the excess air region. Three more low energy fuels do not differ too much
from the natural gas, whereas on the basis of velocity gradient consideration
alone (Figure A-l) there is a clear separation.
If the fan power is limited }* the change from, say, natural gas
to low energy gas may be complicated by this power limit. The air power is
3 4
given by QAp, which varies with pQ /D . Assuming a constant heat release
4 3
rate, that is, if QH is constant, the air power varies with p/D H . For
m m
a stoichiometric mixture in the air, p does not vary appreciably in comparison
3
to H . Thus, the relative values of H are of great importance. It is seen
m m
from Table A-l that the values vary from 97 for natural gas to 62 for
gas from a Winkler gasifier. Only gas from a Koppers-Totzek gassifier is
* Similar results are obtained if fan pressure is considered to be
controlled.
-------
A-15
interchai geable with natural gas on this basis. We note that, if D is
increased to compensate for the lower value of H , flashback will be
encouraged. (This is the reason that in shifting to a medium energy gas
from natural gas, there is a tendency to shift to nozzle-mixing or delayed-
mixing burners). If the number of burners or the number of elements in
243
some burner designs) can be changed, then the constant term is p/N D H .
Assuming that burner designs for comparative fuels are to be limited by the
/ / o o c / o
critical velocity gradient, then the constant term is pG /N H . It is
seen that the number of burners must be increased in changing to the low
energy gas from natural gas while at the same time the total area of the
2 -1/3
burners (ND ), must also be increased with Efe(GH )
Although the volume of products is not exactly proportional to the
4 3
volume of fresh mixture, it is close enough that the term p/D H can be
considered also as a measure of flow power loss through the heat exchanger
part of a furnace or boiler. Thus, again, Koppers-Totzek gas is comparable
with natural gas, whereas Winkler gas will require almost four times the
pressure as that of natural gas to move the products of combustion through
the heat exchanger. The higher velocities associated with a change to
Winkler gas would result in higher heat fluxes to the first part of the heat
exchanger and excessive cooling of the products of combustion in the latter
part. Correcting these faults might require rebuilding the heat exchanger,
though removal of the proper number of tubes might prove acceptable.
The Wobbe Number which is the ratio of the higher heating value
of a fuel to the square root of the specific gravity of the fuel is the
common measure of interchangeability in simple combustion units with a fixed
firing rate, where (a) fuel is used to aspirate air (which is quite common
for simple premixed burners), (b) air is used to aspirate fuel, or (c) a
pressure type control is used to control the ratio of the fuel gas and air.
The reasoning that leads to the Wobbe Number is as follows.
Consider a unit in which the fuel is used to aspirate the air.
In this combustion unit of fixed configuration, with, say, a constant pressure
2
drop on the fuel spuds, x> Q is a constant. For the heat release rate
2
to be constant, QfHf is also constant, and it follows that Hf //°f is a
constant. Normalizing the fuel density to specific gravity and taking the
square root results in the Wobbe Number. Therefore, if the Wobbe Number
-------
A-16
changes, the firing rate of this simple type of unit changes (unless
spud size is changed).
But this is not the entire story. In a combustion unit of fixed
configuration, with any of the types of interconnections between fuel and
air mentioned above, the ratio of momentum flux of the fuel to air remains
2 / 2
constant. Thus,/o Q /a Q is a constant. If ut denotes the air/fuel ratio
ILL. ! 3 a
relative to stoichiometric air/fuel ratio and H is the heating value of
a 1/2
air, Q H = Q H /a. By substitution, it follows that (Yifl(.pflf ) /(OH )
3 3 z r L , L ( Q 3.
is a constant. As the heating value of the air that is used in burning any
hydrocarbon fuel does riot vary greatly, a change in Wobbe Number also results
in a change of excess air in the combustor.
As a result of these effects, for residential units the current
practice is to permit only a + 5 percent change in Wobbe Number.
No consideration has been given in the above discussion to the
change in radiant properties of the flame as a result of the change in fuel.
In the absence of particulate matter, the curve of Figure 6-14 of Reference
A-4 can be used for comparison purposes. As an example, radiation from the
product gases from stoLchiometric combustion of the natural gas and Winkler
gas is compared below.
The slightly greater amount of (CCL + H~0) for the natural gas
leads to about 2 percent greater temperature-emissivity product for natural
gas, while the change in H-O/CO ratio from 1.90 to 0.41 leads to about 10
percent greater temperature-emissivity product for the natural gas flame
(actual amount increases with flame thickness). The absolute temperature
ratio of the natural gas to Winkler gas is about 1.14. Thus, even though
one temperature term is already in the temperature-emissivity product, there
is a further 50 percent increase of natural gas radiation compared to Winkler
gas. Thus, the gas temperature itself has the largest effect. Convection
heat transfer effects and any soot radiation effects will reduce the signi-
ficance of these differences, but the differences will still be sufficiently
large so that they must be evaluated.
Another aspect of radiation is that associated with flame detection
and safety considerations. From the above discussion, it is clear that the
performance of any radiation activated controls on a furnace must be con-
sidered, if the fuel is changed.
-------
A-17
Delayed-Mixing Burners*
Turbulent mixing is usually considered as the rate controlling
factor in turbulent diffusion flames of industrial importance. The chemically
limited reaction rate, which is far greater than the gross reaction rate of
the furnace, is not considered to be controlling or even important,
other than through its effect on flame stability. However, the effect of
turbulence itself is not well understood in complex flow systems, and additional
complications arise from the presence of a flame that adds a random set of
volume sources as the gases expand by heat from random pockets of combustion.
Nonturbulent and turbulent diffusion flames have one feature in
common: the flames must be held at some point, line, or area. In a non-
turbulent flame, the adjacent fuel and air interdiffuse over the edge of the
burner. At some distance, a combustible mixture of varying composition is
reached over a region greater than the normal flame thickness. In this region,
at a distance equal to or greater than the quenching distance, a premixed
flame develops and holds (or "seats") the diffusion flame. In fact, the
diffusion flame may be pictured as a stepwise series of premixed flames,
each with hotter but more dilute initial compositions.
In a turbulent flame, a firm seating of the flame often does
not occur. One notes that local cells of the fuel and air are of different
compositions, temperature, and velocities and have different molecular and
thermal dilutions as they approach the reaction zone.** Thus, there are only
local regions where the maximum turbulent flame speed can exceed the velocity
of the oncoming fuel-air mixture. Therefore, the flame-holding points
shift about in space as the local low-velocity regions shift about in the
turbulent stream. Furthermore, all of the leading edges of the flame must
move at close to the maximum premixed flame speed through the turbulent
mixture, stretching and spreading the flame.*** When the flame no longer
* Delayed-mixing burners are those "in which the fuel and air leave the
burner nozzle unmixed and thereafter mix relatively slowly largely
through diffusion. This results in a long luminous flame called a
diffusion flame, luminous flame, or long flame."
** This variation from the average of local time and space concentrations is
known as the unmixedness of the fluid.
*** Otsuka and Niioka (A-5) suggest that, in cases where the flame is being
rapidly stretched as would be the case in a turbulent flame front, the
flame forms in the maximum temperature region rather than the stoichiometric
region often assumed in the literature.
-------
A-18
contacts enough local regions where it can "buck" the oncoming stream and
not be extinguished, it will blow off unless held by some independent energy
source.
It thus appears that the critical stability parameter in an enclosed
turbulent diffusion flame will be related to the maximum flash-back velocity
gradient rather than the velocity gradient specific to the average mixture
ratio.
Figure A-5 shows typical delayed mixing burners that will result
in a long luminous flame. Figure A-5a is a version in which the fuel and
air velocities are similar. Natural gas and low Btu gases are interchangeable
in this burner with change in gas pressure. We note that a pilot flame is
incorporated for ignition and/or piloting of the diffusion flame. The pilot
flames are usually premix or nozzle-mix flames. Therefore, if the stability
of the diffusion flame depends on the pilot flame, then the stability con-
ditions of the pilot flame are of prime importance. However, even with a
pilot flame, the diffusion flame may not be sufficiently held so that a
satisfactory flame results. Thus, the stability characteristic of the
diffusion flame must also be considered. Current practice in design is to
insure satisfactory flame performance without a pilot flame for safety reasons,
It is noted that the protective effect of the short tile of this burner helps
insure satisfactory holding of the frame.
Figure A-5b shows a delayed mixing burner in which the fuel remains
in a high-velocity, coherent jet for a considerable distance surrounded by
a low-velocity air mantle. The flame is piloted through the effect of the
recirculation and mixing annular region surrounding the fuel jet.
In neither of the burners is there any problem of flash back.
Thus, only the possibility of blowing off the flame need be considered in
comparing its performance to various fuels. Considering the fuels in Table
A-l, it is seen from the order numbers for the maximum flash-back velocity
gradient in Table A-2 that natural gas is the most unstable of the tabulated
fuels. For medium energy fuels, the combustion systems are much more stable.
However, this argument does not take into account the necessary change in
fuel flow rate with low energy fuels if the burner remains unchanged.
-------
A-19
PILOT
TIP
(a)
GAS
AIR
(b)
FIGURE A-5. DELAYED MIXING BURNERS
-------
A-20
Figure A-5b nay be considered as just a simple diffusion flame of
the Bunsen burner type., with only fuel in the central jet with a change
in fuel, the maximum diameter of the flame increases as the stoichiometric
air/fuel ratio increases. For turbulent flames* at a constant heat input
rate, the length of flame changes little. For a constant shape of burner
and considering a constant heat input rate and a low velocity of the air
in relation to the fuel jet, the holding point of the flame will be determined
roughly by the product of the higher heating values and the maximum flash-
back velocity gradient. Figure A-3 (and Table A-2) present the values of
this point for the various fuels considered. It is seen that the order of
fuels has changed from those noted in the previous discussion. Koppers-
Totzek gas is now the. most stable fuel, and natural gas is above the midpoint
of the fuels. Winkler gas is again the least stable. When the fuel velocity
is higher than the average air velocity, the movement of the combustible
interface outward with increasing air/fuel value ratio at stoichiometric
also improves the stability of natural gas relative to the remainder of the
fuels.
This is not the entire story, however. For most delayed mixing
burners, such as shown in Figure A-5a, the fuel and air velocities are about
the same to inhibit premature mixing.''"'- Therefore, stability of the flame,
if the flame is held within the tile, is governed by whichever velocity is
controlling—the fuel velocity or the air velocity, or a combination thereof--
in the exact region of holding. Furthermore, if changes in fuel are made
without concomitant changes in burner dimensions, the relative values of
fuel and air velocity will change, and the controlling significance may
change. The flame may find a stable region of attachment around the
annular air jet, rather than the fuel jet. In this case, the flames stabilize
closer to the air jet as the air/fuel volume ratio at stoichiometric decreases,
* While the aspiration rate of the fuel jet cannot be significantly altered,
care is usually taken to eliminate as much swirl and turbulence from the
air flow as possible to keep from increasing the mixing rate unnecessarily.
** These are sometimes called laminar flow burners, but this does not denote
viscous flow (Reynold's numbers are still high). Rather, it denotes
flow without unnecessary high-intensity turbulence.
-------
A-21
Furthermore, and more important, the air velocity does not change much with
fuel at a constant heat input rate. Thus, the curves of Figure A-l should be
considered for stability.
If the flame does not stabilize close to the inlets in either position,
then the slow mixing can result in other diffusion flames starting beyond the
tile in the region where recirculating gases will slow the flow velocity and
dilute the air annulus.
When these burners are used in radiant tubes, it is often desirable
to have the heat flux peak near the burner and hold at that value or fall off
gradually, rather than increase slowly to a peak value some distance down the
tube. To accomplish this, a small amount of air may be bled into the fuel jet
so that the boundary of the fuel jet as it emerges from the fuel tube is a
combustible mixture. This portion of the boundary burns as a premixed flame,
both boosting the heat flux to the wall near the inlet and serving as a pilot
for the downstream diffusion flame. However, because of the diffusion effects,
there is still a composition gradient, and the stability even in this case
should probably be treated as would a diffusion flame stability problem be
treated.
The pressure drops that are involved in supplying the fuel are now
considered briefly. As may be deduced from the discussion of the Wobbe Number,
in connection with premix flames for a pair of fuels in which this number does
not vary too much, the fuels are interchangeable in diffusion-flame applica-
tions as well as premix-flame applications. It is seen from Table A-2 that
the medium energy gases are the closest to natural gas, but are far from being
within the 5 percent variation usually allowed. Furthermore, a massive addi-
tion of propane, about 32 percent by volume, would be required to boost the
values sufficiently to bring them within range. But it is noted that the
energy values of the stoichiometric mixtures are about the same for these fuels
as for natural gas, so that changes only in the burner would be required to
obtain satisfactory operation of the system.
Interestingly, increasing the orifice sizes for the medium energy
gas sufficiently to maintain the same stoichiometry percent results in a de-
crease in gas pressure while maintaining a constant heat release rate. Chang-
ing the orifice size a lesser amount so as to maintain the back pressure on
-------
A-22
the fuel, and maintaining a constant heat release rate results in an increase
in the excess air using fuel aspiration. This, of course, may be handled by
an additional adjustment.
One can conclude, therefore, that in replacing natural gas in a
diffusion flame with medium energy <-nanufactured gas, no stability problems
will be encountered. However, there can be a stability problem with lower
energy fuels if some changes in burners are not made. For diffusion-flame
burners, the conclusions relative to pressure drop through heat exchangers
and the comments on flame radiation made above for premix burners are unchanged,
Nozzle-Mix Burners"
Nozzle-mix burners combine the advantage of the relatively short
flame of the premix burner and the lack of flash-back problems of the diffusion
flame. The short flames are obtained by three different methods. Figure A-6a
shows the use of multiple high-velocity air jets parallel with the fuel jet.
The air jets aspirate the fuel in around them and form short flames because of
the small jet diameter and potential core length.** Figure A-6b shows the use
of nonparallel jets. These may impinge or may be canted to produce a swirl
flame and even a heavy recirculation zone in the center of the jet. If a disk
is added to the end of the fuel jet in Figure A-5b, a high-velocity air flow
and a recirculation zone are formed which lead to an intense mixing. The
burner in Figure A-5b then becomes a nozzle-mix burner. Some of the fuel in
this case may be diverted radially to improve mixing further. In all these
cases, the internally recirculating hot gases plus the hot ceramic tile wall
provides good flame stability.***
If the flame is held as a diffusion flame in a nozzle mix burner,
then the flame might either be held around the central fuel jet or the
* "A burner in which fuel and air are not mixed until fuel just as they leave the
burner port, after which mixing is usually very rapid. The flame cannot
flash back in this type of burner."' •*
** On occasion, the roles of the fuel and air jets are reversed.
-** Care must be taken to prevent aspiration of cold furnace gases back into
the tile at the flame base.
-------
A-23
GAS
(a)
GAS
(b)
FIGURE A-6. NOZZLE-MIXING BURNERS
-------
A-2 4
peripheral air jets. The argument here is exactly the same as for the delayed
mixing burners. The main difference is that, when the flame is not attached
close to the inlet of either the fuel or air, rapid mixing may take place
before a stable region for the flame to seat is encountered. In this case,
the action of the flame is much like a premix burner.
Therefore, it is concluded that in changing from natural gas to lower
energy fuel in a nozzle mixing burner, the position of the flame base may change
from around the fuel jet to around the air jets. Therefore, an unqualified
comparison of stability cannot be made. As a result, it is not clear whether
a flame might satisfactorily contain itself within a nozzle-mixing burner tile
with a specific change in fuel.
Again the overall conclusions are similar to those for premix and
delayed mixing burners.
-------
A-25
References
(A-l) Combustion of Gases. Gas Engineers Handbook, The Industrial Press,
1965, Table 2-59.
(A-2) Grumer, J., Harris, M. E., and Rowe, V. R., Fundamental Flashback.
Blowoff. and Yellow Tip Limits of Fuel Gas-Air Mixtures, U.S. Bureau
of Mines, RI5225, 1956.
(A-3) North American Combustion Handbook. The North American Manufacturing
Company, 1952.
(A-4) Hottel, H. D., and Sarofim, H. F., Radiative Transfer. McGraw-Hill
Book Company, 1967, Figures 6-14.
(A-5) Otsuka, Y., and Niioka, T., The One-Dimensional Diffusion Flame in
a Two-Dimensional Counter Flow Burner. Combustion and Flame, 21 pp,
163-176, 1973.
-------
APPENDIX B
FUEL GAS CLEANUP PROCESSES
-------
APPENDIX B. FUEL GAS CLEANUP PROCESSES
Sulfur Removal
Figure 5 shows the expected S0? emissions per million Btu heat input
versus percent sulfur in the coal for different gasification efficiencies. A
coal-heating value of 13000 Btu/lb (7200 kcal/kg) on a moist, ash-free basis was
assumed in this case.
For a typical coal combustion process, nearly all the sulfur in the
coal is converted to gaseous products in the form of S0_ in the flue gas. An
/gi\ ^
analysis of Koppers data v ' reveals that nearly all the sulfur in the coal is
converted primarily to H S. For purposes of this study, it was assumed that at
least 90 percent of the sulfur in the coal is converted to tLS.
From Figure B-l it can be seen that emissions of S02 with no sulfur
removal would be expected to exceed the current new source standard of 1.2 Ib
S0? per million Btu heat input of fuel gas for coals with more than 1 per-
cent sulfur by weight.
Figure B-2 shows the required removal efficiencies of a gas-cleaning
device to achieve the Federal standard for a combustion process. For high
sulfur coals (up to 7 percent), removal efficiencies of around 90 percent would
be required, and for coals above 3 percent sulfur, removal efficiencies above
75 percent would be required to achieve the standard.
Figure B-3 shows the expected concentrations of I^S in fuel gases of
different heating values for an assumed gasification efficiency of 70 percent
with coal at 13000 Btu/lb (7200 kcal/kg). Figure B-4 shows how the concentration
of H_S in the fuel gas relates to emission of SO- in pounds per million Btu heat
input. It can be seen from Figure B-4 that a concentration of about 700 ppm for
3 3
100 Btu/scf (888 kcal/m ) gas to 2000 ppm for 300 Btu/scf (2660 kcal/m ) gas
would be the allowable concentration of H-S in the fuel gas to meet the standard
^ f
for steam generators of 1.2 Ib S02/million Btu heat input (2.1 kg/10 kcal).
-------
B-2
20
18
16
14
CM 10
O l2
10
8
0
Coal at 13,000 Btu/lb
Fed std = 1.2 lb/!06Btu
I
"0 12345
Sulfur in Coal, percent
FIGURE B-l. S02 EMISSION VERSUS PERCENT
SULFUR IN COAL FOR DIFFERENT
GASIFICATION EFFICIENCIES
Note: See Appendix C for Determination of Relationships for Sulfur
Emissions.
-------
B-3
0)
o
0)
Q.
o
TJ
C
CO
o3
0)
T3
0)
5
cr
a:
"5
o
E
or
i_
^
H—
CO
100
90
80
70
60
50
40
30
20
10
0,
Gasification
Efficiency
60°/<
0
2345
Sulfur in Coal, percent
FIGURE B-2. SULFUR REMOVAL REQUIRED TO MEET
STANDARD VERSUS PERCENT SULFUR
IN COAL
Not
See Appendix C for Determination of Relationships for Sulfur
Emiss ions.
-------
B-4
20
15
E
Q.
Q.
to
O
O
CD
10
CO
oo
X
Coal HHV, 13,000 Btu/lb
Gasification efficiency, 70%
2345
Sulfur in Coal, percent
FIGURE B-3. H2S IN FUEL GAS VERSUS PERCENT
SULFUR IN COAL FOR DIFFERENT
GAS HIGH HEATING VALUES (HHV)
Note: See Appendix C for Determination of Relationships for Sulfur
Etniss ions.
-------
i-5
20
•3
£
10
o
15
10
0
Coal HHV, 13,000 Btu/lb
Gasification efficiency,
70%
0
Federal std = 1.2 Ib/IO Btu
5000 10,000
H2S in Fuel Gas, ppm
15,000
FIGURE B-4. S02 EMISSION VERSUS H2S CONCENTRATION
IN FUEL GAS FOR DIFFERENT GAS HIGH
HEATING VALUES (HHV)
Note: See Appendix C for Determination of Relationships for Sulfur
Emissions.
-------
B-6
Commercial processes currently available for removing sulfur from fuel
gas are listed in Table B-l. All these processes except two, the iron sponge proc-
ess and activated carbon process, utilize some form of liquid scrubbing for re-
moving sulfur. The highest temperature at which any of them are capable of
operating is 250 F (122 C).
Normally, when one of these systems is employed for sulfur removal,
the gas from the gasifier is first processed through a waste heat boiler to
utilize some of the sensible heat in the gas to raise steam. The estimated
steam raising capabilities of the four gasifiers discussed here are given in
Table 4. Some of the sensible heat in the gas can also be used to preheat the
air or oxygen feed to the gasifier to further improve the overall thermal
efficiency of the unit.
The minimum temperature the gas can be cooled to and still be usable
for steam raising or gas preheating is around 300 to 500 F (150 to 260 C), de-
pending on the desired steam pressure or preheat temperature. The gas is then
cooled still further by scrubbing with water to the temperature required by the
cold sulfur removal process. The water also removes particulates, ammonia, and
condensable tars if present. Some H_S and CO. may also be removed in the water
scrubber though, due to the low solubility of H S in water (l/3000th that of
SO.)(B~3* the amount of H-S removed should be negligible.
After water scrubbing, the gas is ready for the sulfur removal process.
The water used for scrubbing is generally filtered to remove particulates and
tar and then heated to strip off ammonia along with some other acid gases prior
to going through a cooling tower and being returned to the scrubber. The ammonia
in the stripped gas (possible 15 percent by volume) can be recovered as a
marketable product by existing processes.^ '
Amine Scrubbing. The first group of processes in Table B-l remove sul-
fur with chemical absorption in an organic amine compound. These processes
operate at ambient temperatures or slightly above. The applicability of these
processes to cleaning of the gas from coal gasification is doubtful, as the sor-
bents are expensive and tend to be rendered nonregenerable by other impurities
in the gas such as heavier hydrocarbons and carbonyl sulfide.
-------
z
Q
M
H
C"3
M
f¥*
EC
P;j
J^
t3
w
§
w
^
w
B
o
en
M
en
W
O
Pd
P-i
,
O
M
E-i
5
u
r-i
«
a
M
o
§
V
w
B
ht
0
CO
g
0)
•o
i
o.
g
o
41
rH
f%
09
hi
41
B
41
60
41
B
O
B
i
o
(b
*
T!
B
1 '
V O
o
B X
4) 4J
A -H
Irf ft
o
CO
rV
O
O
rH
£
"•
^.
V
CM
S
w
03
af
g
rH
0
09
J3
1
!S
W
>
72
NH
09
B
41
O
If
u
o
B
4J
9
J5
e
(Q
4J
§
a
&
M
41
41
O
6 *
r/j
B U
•O O
M W
0
co
(x
o
0
£
r-.
O
o
V
u
^
CO
B
1
O
B
01
4J
a
a
Q
60
B
a
S
B
O
•pH
09
O
W
hi
o
u
j:
jj
•H
09
§
XI
O
.
>
^H
B
X
4)
u
B
-Q
hi
0
U
r>
0
o
rH
(J,
^
•*
V
CM
O
U
n
CO
CM
a
41
O
1
^^4
O
U
60
•rl
a
a
•rl
g
B
O
U
M
60
B
a
o
B
O
•H
09
O
hi
0
u
f*
4J
•H
«
§
•H
•8
a
.
41
•H
09
E
41
a
X
41
U
C
41
JS
hi
O
CO
B
41
•rl
J
"*
O
1
rH
CM
O
O
4)
09
»
CO
,H
4J
•O
M ° "
10 (j
o "3
s § M
C 09 *
O M 41
-I co B
U U.H
3 ° i
O
CO
3
i
X
rH
•O
a
41
14
09
C
o
rO
1,4
09
U
O
U
•=
§
i
.
>
•rl
09
E
X -0
41 41
J3
U hi
S 0
4) 09
U <0
O
CO
U
a
4)
•rl
J
"^
0)
>
«r4
4J
u
CM 4)
g -3
a
CO CO
CM CN
&
hi
•a o
XT)
-§u
hi 0
U Tl
« -a
u
. B
3 41
P. a.
o
41 iH
B f
r
o
B
•rl
•M
9
CO
X
41
O
U
41
ON
u
CO
pf1
o
z
o
a>
•o
41
41
•a
E
O
U
D
tr
41
hi
u
o
3
cr
60
E
•rl •
iJ CO
a O
•3°
u o
hi U
•rl
U 41
41 0)
U B
10 41
10
J= B
60M
O
in
CM
E
a
X
n
u
I*
o
3
tr
60
E
«
41
hi
•rl
3
cr
4)
mi
09
41
n
09
41
N
0L 41
U
B-7
g
M
60
•O
a
hi
a
4)
2 c
BE
<0 o
m a
X 41
rH hi
a s
u »
u
(».
o
CM
a.
CO -rl <9
SZ b
" « TJ
n a
J-' O S
10 rH
hi u a
M 3 0)
00 Tl
3-1 §
60
M U
4i a
r
o
in
CM
1
g
u
X
41
41
CO
o
u
o o
CM 41 CM 41
O "4! O 41
U
CO
O
-N
en
8
CO
O
u
CM
*
•V
hi
|
<0
41
CO
o
3
4)
a
o
co
O
°CM
o
u
09
U
•e
rH
4)
2
u
£
U
§
tr
8
B
o
-------
3
•r-l
4J
O
U
PQ
PQ
<
H
o
4J
c <
Lrect Oxidatlo
Q
Cft
01
JD
r~t
10 X
3 B
3 1 !
ir. hi
Mj 33
t-4 O W-l
2 7) -i
* kl 3
,. 00 00
P "
zL *
ft On
*W • B)
" T) U
S -C P co
* M 0 ki
O 3 H
• kl O CO U
25! c j: o
0)0. 41 3
3 r-4 CT
in 01 co H-I
P C7\ p -H _l
O 1 34J
•wo a. a j;
_ °» 01 oo
•Si p 4J ,*
2P 3 0 X
3 • • 14-1 CX
o x oi 7i
C O u 3 CO •
4) J3 01 co u
"* s
k, r- 5 "-1 °
5 c ^ C • co r-l
u oi » hi 0) c
<*•!
•H -H B (N
• CO
4) 4) 2
*O *O kl
T4 *H O -i.
x x
O O B n c*>
5 o O
C c ••< co O
0 0 13 < CM
kl kl O fr} CO
M W CO Ui .8
O
i b
u u o
3 O •*>
M M kt U
-I *O u CX
3 X 3 W 41 p
0 O 3 41 > O
J3 — 1 " i CO
-< O <0 O -O XI
ro C 01 u ki <
u O « X hi O
CU XoXco >4-i i-l
CUT* OCO Q 4J CO
= u «j i^ § oi o
QJ4J U S X tO kl -H
-1 4) 01 -C f* «J CO
d S U. Hcj CO >
e-
•
,
u
c
41
•H
JJ
B
CO
g ^ .
r-l 01 4)
01 > ki
•0 O 3
S co
•O 41 CO
01 u CU
i-l H
o o o.
O CO
U r-l i-l
CO CO
41 *•!
.O CO 4J
r-4 kl
4-1 >r4 CO
CO 0 O.
B x-e
? 00
CO <0 -r<
co 01 J=
o a
41
^
. a
u u .C
c c
OI 01 4J r-l r-l
J3 ja co o o
hi kl 3 CO CO
O O B -r4 T<
CO <0 kl kl
X 3 3
41 oi u a. o.
> > -r*
•r4. i-l kl CO CO
co m 3 co co
c c ex
01 o B 01 41
CX O. r-l Q g
X X co CO
w w co to
4J JJ U
s s s
•rJ >rj -^
S -° •" -°
"III
° § g ?
_ ^H *— 1 i-^
° 11) 41 4)
pa M M
r-l CM ^ st
r-l CU CO
0 C
ki - . O
kl P -O
X 4) ki
(X J3 »
1 4J U
01 * «
1 V
1-4 r-4 r-4 G
O X > «>
C j: J=r-< r-l
CO 4J 4-1 O X
_C 01 41 O O.
4J B B X o
4) 1 -rlr-4 M
£ 2 O 60 £
c
o
ir4
U
r-l 0
O r-l r-4 hi
• O O 0
-1 CO X hi M
4J -r4 01 O T)
O kl r-l • 3 <
01 3 4> •"*
as a. to («4 >
O
•
jj
01 c
CO 4)
41 S
-H §
S
C r- 1
o ex
T< 41
«-* kl
CO
kl 4J
U C
C 41
41 3
o cr
C 41
O hi
u *w
u C 00
41 O C
r-4 -H -r<
C 4J kl
•H CO v4
kl 3
p 4) cT
O O. 4)
<« O ki
W JC C
CMC! O
K " -2
CO p
W CO CO
41 U
>
O • c
B B o
4) CX
kl & CO
xo g
t-( O o
01 <*"( U-4
4^
G. C %
i—4 tD 3
CX J= »W
S" -3
o w
•u u
s s
*r4 «H
1 1
0 J
CO CO
sT4 sT1
"2
§
O
c
o
0>
•a
1-4
M
O
C
o
\4
•V4
•e a
« M
*J C Q
2 -S 2
P •> ja
•On hi
dfrl 3
0
X)
kl
Iron sponge
Activated C<
B-8
•
01
•r-l
4-J
CO
4-1
C
0)
co
CU
P
&
0)
rJ
<4-l
0
4-1
cn
-r-l
1— 1
-------
B-9
A Ika I i Sc r ubb ing . The next group of processes listed in Table B-l in-
volve chemical absorption in alkali or inorganic absorbents. The Seaboard
process developed by Koppers is obsolete, as there is no means for regenerating
the absorbent. The vacuum carbonate process is similar to the Seaboard process
but allows for sorbent regeneration under 25 in. Hg vacuum with steam. The
sorbent, Na9CO , is not very soluble in water and requires large circulating
£» O
liquor rates. The next process (originally developed by the Bureau of
Mines) uses hot I^CO-j as the sorbent. This process has the highest operating
temperature (250 F, 120 C) of any of the cold gas cleanup systems. The Benfield
and Catacarb processes are very similar to the process developed by the Bureau
of Mines, except that they employ proprietary catalysts to increase the rate of
absorption and stripping, thus decreasing the required liquor flow rates. The
2
hot carbonate processes operate best at about 300 psig (2060 N/m -hr) pressure.
The sorbent is regenerated with steam, and, due to the high liquor flow required,
3
a relatively large amount of steam is necessary (up to 1 Ib of steam per 6 ft
of acid gas or 0.453Kg/0.17 m3).
The Shell process uses hot K-PO instead of K_CO_ . K_PO is more soluble
in water than K_CO_, thus reducing the required liquor flow rate. A large amount
of steam is still required for regenerating the sorbent, however.
A potentially attractive process is scrubbing with an ammoniacal liquor,
since a certain amount of ammonia is already available from the water scrubbing
process. Ammoniacal scrubbing is currently planned for use on the air-blown
gasifiers at the Lurgi gas-turbine power plant demonstration unit in Lunen,
/•p_£\
Germany. v
Direct Oxidation. Direct oxidation processes have the desirable
characteristics of converting H~S from the gas stream directly to elemental
sulfur and being selective for H«S. Absorption of C0_ along with H_S requires
unnecessary liquor flow and larger equipment. The Wet Iron Box process, of which
the Manchester, Gluud, and Ferrox are modern applications, react H_S with iron
oxide (Fe20,j or Fe(OH)g) in a sodium carbonate slurry to oxidize l^S directly to
sulfur. Reportedly, sulfur of 90-95 percent purity can be attained. (B-5) These
processes have not been well developed, however, and immediate commercial availability
is doubtful.
The Thylox and Giammarco-Vetrocoke processes use sodium, ammonium, or
potassium arsenate solutions to absorb and oxidize IS to elemental sulfur. The
-------
B-10
sorbent is expensive, however, and the presence of arsenate makes it potentially
hazardous as a pollutant if leakage should occur.
The Stretford process uses a mixture of sodium carbonate and sodium
vanadate as the absorbing oxidizing solution. The process is capable of opera-
ting over a wide range of pressures but requires high solution-flow rates with
an expensive absorbent, and the resulting sulfur content may be too high to
render the fuel usable.
Physical Absorption. Physical absorption processes were developed
primarily to attain an e>xtremely low concentration of sulfur in synthesis gases
to prevent contamination of catalysts used in synthesizing methane and other
higher hydrocarbon products from the gasification of various fuels including
coal. They all are designed to operate best at temperatures below ambient,
though some can operate at ambient.
The Lurgi-Rectisol process uses very cold (-80 F, -62 C) methanol as
the absorbent. This process is currently used to clean synthesis gas at the
/r>_ -i \
giant Lurgi gasification plant in Sasolburg, South Africa.1 The absorption
takes place at temperatures from 0 to -80 F (-28 to -62 C), resulting in a
high degree of gas purity (total sulfur less than 3 ppm).(B-4)
The Purisol, Selexol, and Fluor processes are similar to the Lurgi-
Rectisol except that they use different absorbents and the absorption takes
place at near ambient temperature and elevated pressure. Attainable gas purity
with all three processes is around 4 ppm total sulfur.
In all four physical absorption processes the sorbent is expensive
and the gas must be cooled to at least ambient temperature. The gas purity
obtained in these processes is more in keeping with the strict requirements of
catalytic synthesis operations rather than the more lenient: requirements of
environmental acceptability.
Dry Adsorption Processes. The iron sponge process utilizes hydrated
iron oxide impregnated on wood shavings to selectively remove H0S. It is a
batch process, however, and requires frequent replacement of the sponge.
Activated carbon can also be used to obtain very low concentrations
of H^S in the product gas. Sulfur tends to form in the pores of the carbon,
however, resulting in need for frequent replacement of the bed.
-------
B-ll
Hot Cleanup Processes. Currently, three processes are under develop-
ment for sulfur removal from flue gases at elevated temperatures, though none
are presently commercially available. These are the Battelle-Northwest molten
salt scrubbing process, the Air products and Chemical Company's solid absorbent
process, and the Bureau of Mines iron oxide process.
The Battelle process utilizes a mixture of molten alkali metal carbon-
ates containing calcium carbonate in solution as the working fluid for a venturi
scrubber, which is designed to remove particulates and sulfur compounds at tem-
peratures from 100 to 1650 F (592 to 895 C). The molten salt is filtered to
remove particulates and regenerated with a steam-C02 mixture.
The Air Products process utilizes an absorber comprised of fully
calcined dolomite (MgO-CaO), with the l^S in the fuel gas reacting with CaO to
form CaS over a temperature range from 1100 to 1650 F (592 to 895 C). Regenera-
tion is by reaction with steam. This process utilizes the approach initially
suggested by A. M. Squires.' *^'
The Bureau of Mines iron oxide process utilizes pellets of fly ash
impregnated with iron oxide in a packed bed to remove t^S at temperatures from
1000 to 1500 F (535 to 815 C). This process has been tested on producer gas at
the Bureau's Morgantown gasification site in a pilot-scale demonstration unit
and achieved 95 to 97 percent removal of H2S.(B"8>9) Regeneration is by air
blown through the bed at around 1000 F, releasing the sulfur in the form of SCv.
The Consolidation Coal Company dolomite process(B-ll) reacts HoS in
the fuel gas with half calcined dolomite (MgO-CaCO^) in a fluidized bed at
Ib52 F and 15 atmospheres pressure. The spent sorbent (MgO-CaS) is regenerated
with CC>2 and recalcined at 1300 F. t^S is liberated in the regeneration step.
Waste or By-Product Recovery of Sulfur
In most of the liquid scrubbing processes in Table B-l the sorbent is
regenerated by stripping off H^S and other absorbed gases. The H-S at a suf-
ficiently high concentration of 10 percent or more can be converted to elemental
sulfur in a Glaus conversion unit.
In the Glaus split-stream process^ ' , one-third of the gas stream
is split and combined with air to completely oxidize H2S to SO and hydrocarbons
to CO . This stream is then combined with the remaining two-thirds of the gas
stream over a bauxite catalyst, where H^S and SO. combine to form elemental
sulfur. Typically, 90 to 96 percent of the sulfur is removed from the feed gas.
-------
B-12
Further sulfur removal may be necessary on the tail gas of the Glaus unit,
depending on the amount of sulfur originally removed from the fuel gas.
Some of the liquid scrubbing processes in Table B-l convert IL^S directly
to elemental sulfur. These processes have declined in favor, however, and are
not well developed. Also, in some cases the sulfur is of too low a purity to
be efficiently recovered for further use.
Recovery of sulfur or waste disposal from the proposed hot cleanup
processes is not well defined due to lack of development. In the Bureau
of Mines iron oxide process (the most highly developed of the three currently),
the sorbent is regenerated with air, stripping the sulfur off in the form of
S0?. The disposition of SO is more difficult than H~S, since production of
elemental sulfur requires a reducing agent. S00 can be converted to sulfuric
t~
acid, H_SO,, by conventional processes, though these are more complicated than
the Glaus process; also, the sulfuric acid (H«SO.) is more difficult to handle
and dispose of than sulfur.
Particulates Control
All four of the gasification processes discussed earlier will require
some form of particulate control. The Winkler reactor has the highest reported
ash carryover of the four gasifiers, with 70 percent of the coal ash being
carried over in the gas. The Koppers process carries over about 50 percent of
the ash into the gas, wich the nature of the ash similar to that from a pulverized
coal fired boiler. The Wellman-Galusha and Lurgi processes would have lower
dust loadings in the fuel gas than either that of the Winkler or Koppers, but
would also contain condensable tars that should be removed to prevent fouling
of burners and other fuel-handling processes.
Only one of the proposed hot sulfur removal processes, the Battelle-
Northwest molten salt process, proposes to remove particulates along with sul-
fur at elevated temperatures* Both the Air Products dolomite process and the
Bureau of Mines iron oxide process also require some form of particulate re-
moval at elevated temperatures. Work has been done on developing high-
temperature precipitators operating at temperatures above 1500 F (815 C)
though these systems have not yet seen significant commercial application.
Filtration through beds of sized granular solids also looks promising though
this concept must be demonstrated commercially.
-------
B-13
The presence of condensable tars in the gas could also be troublesome.
If tarry compounds are allowed to condense on collecting surfaces or filter
media they could seriously affect performance of the unit. None of the hot
sulfur removal processes are intended to also remove tars. If these tars are
not removed prior to gas distribution and use, care must be taken to maintain
high temperatures and generously sized gas passages to prevent plugging and
fouling of gas mains and burners.
If the gas is to be cleaned and used cold--below 300 F (149 C)--the
gas temperature must be lowered from the gasifier exit temperature to between
300 F (149 C) and 500 F (260 C) by either preheating the air or oxygen feed,
raising process steam in a waste heat boiler, or some combination of both.
The gas then can be processed through conventional particulate removal devices.
The most common method is water scrubbing, which also serves to remove conden-
sable tars and ammonia while lowering the gas temperature still further for the
subsequent sulfur removal process.
Nitrogen Oxides
Coal gasification is basically the first stage of a two-stage combus-
tion process. As a result of the reducing atmosphere in the gasification zone,
nearly all of the nitrogen in the coal is converted to ammonia (NH_), with
small amounts going to HCN and other trace compounds (Table 2). Only a very
small amount of NO would be formed in the gasification step.
2C
If water scrubbing is employed in a cold gas cleanup system, a large
amount of the ammonia would be removed. If cleanup processes are used that
don't remove ammonia (hot gas cleanup systems, for example), then the ammonia
would be carried through to the final combustion process, in which case the
eventual NO emissions from oxidation of ammonia during combustion would be
x
expected to be similar to those of normal two-stage coal combustion.
-------
B-14
References
(B-l) The Production o£ Gas from Coal Through a Commercially Proven Process,
Farnsworth et al., Koppers, 1973.
(B-2) United Stateis Federal Register, Vol. 36, No. 247, December 23, 1971.
(B-3) Chemical Engineer's Handbook, Fifth Edition, Perry and Chilton, McGraw-
Hill, 1973.
(B-4) A Survey of R&D Projects Directed Toward the Conversion of Coal to
Gaseous and Liquid FueJLs, IGT, Chicago, 1971.
(B-5) Technological and Economic Feasibility of Advanced Power Cycles and
Methods of Producing Non-Polluting Fuels for Utility Power Stations,
Robson, et al., United Aircraft Research Laboratories, 1970.
(B-6) New Fossil-Fueled Powe£ Plant Process Based on Lurgi Pressure Gasifica-
tion of Coal, Lurgi Gessellschaften, 1970.
(B-7) SASOL Today, The South African Coal Oil and Gas Corporation Limited.
(B-8) Hydrogen Sulfide Removal From Hot Producer Gas With Sintered Absorbents,
Schultz, Berber, APCA Journal, February, 1970.
(B-9) Hot Sulfur Removal From Producer Gas, Schultz, Lewis, 3rd International
Conference on Fluidized Bed Combustion, Hues ton Woods, 1972.
(B-10) Cyclic Use of Calcined Dolomite to Desulfurize Fuels Undergoing
Gasification, A.M. Squires, Fuel Gasification. Advances in Chemistry
Series No. 69, American Chemical Society, Washington, D.C., 1967.
(B-ll) Production of Clean Fuel Gas From Bituminous Coal, Curran, et al. ,
Consolidation Coal Company, EPA Report 650/2-73-049, 1973.
(B-12) Air Pollution Control, Edited by Werner Straus, 19/1.
-------
APPENDIX C
DETERMINATION OF RELATIONSHIPS FOR
SULFUR EMISSIONS GIVEN IK
FIGURES B-l THROUGH B-4
-------
APPENDIX C
DETERMINATION OF RELATIONSHIPS FOR
SULFUR EMISSIONS GIVEN IN
FIGURES B-l THROUGH B-4
Relationships between Ibs S0~/106 Btu versus the percent sulfur
in coal. (Figure B-l)
(1) Ibs S/lb coal = % S/100
(2) Ibs S/Btu (coal) = Ibs S/lb coal x r—-
HHV
C
(3) Ibs S/106 Btu (coal) = Ibs S/Btu (coal) x 106
(4) Ibs S/106 Btu (gas) = Ibs S/106 Btu (coal) x rf-
Mg
Assuming 90 percent sulfur carryover from coal to gas, equation
(4) becomes:
(5) Ibs S/106 Btu (gas) = 0.90 Ibs S/106 Btu (coal) x ~
or by substitution:
%S 106 0.90
_
106 Btu (gas) 100 X HHVC X Ng
Because 1 mole S yields 1 mole S02 and
Molecular weight S = 32
Molecular weight S02 = 64.
The weight of SC>2 = 2 x weight of sulfur, therefore
(7) Ibs S02 = %S 106 0.90 ?
10° Btu (gas) 100 X HHVC X Ng
or
Ibs S02 m 1.8 x 104 %S
Btu (gas) HHVC Ng
-------
C-2
Relations ip between percent sulfur removal required to meet
standard of 1.2 Ibs SO™/ 10 Btu versus percent sulfur in coal. (Figure B-2)
.n. ,n . Ibs SOW106 Btu (gas) - 1.2 ...
(9) Percent S00 removal = - — — g. ..'?' n ''ft""' - x 100
v ' 2 Lbs 802/10° Btu (gas)
or
_ 1..2 _
(10) Percent S02 removal = 1 - I-8 x 1()4 %s/HHV Ng
Relationship between tbS content of fuel gas ""ppm) and percent sulfur
in coal, (Figure B-3)
moles S Ibs S 1 _ moles H2S
106 Btu (gas) = 106 Btu (gas) x 32 io6 Btu (gas)
moles H2S _ iTioles H?S
10ft Btu (gas)" X HHVg " 10& scf (gas)
At 32 F and 29.92 in. Hg, 1 mole of H~S occupies 359 ft3. Therefore,
at 607» and 30 in. Hg (standard gas condition for this report), 1 mole
of IUS occupies
,sq v 520 29.92 __ 3
359 x 492 x 30.0 - 378 ft
,10v moles H2S -_Q ftj H?S „
(13) rrr - : — r^ — r x 3/8 = TTTT - T~^ - r = ppm H0b
' 10& scf (gas) 10" scf (gas) PF 2
By combining equations (11), (12), and (13), and using equation (6)
the following expression can be developed.
°/c 1 f)6 n nn -i
(14) ppm H2S (in fuel gas) = ^ x ^- x -g- x ^ x HHVg x 378
or
/ICN „ c 10.6 x 104 (%S) (HHVg)
(15) ppm H2S = }
L-
-------
C-3
Relationsip between lb S02/10 Btu and ppm H2S in fuel gas,
(Figure B-4)
By equation (13):
fif.\ ppm H9S _ moles H2S _ moles SO?
(Lb) 3?0 - 1()6 scf (gas) ~ 106 scf (gas)
moles S02 64* _ Ibs S02
scf (gas) X HHVg 10b Btu
*64 is molecular weight of S02, therefore
Ibs S02 _ Q 1 64
106 Btu ~ ppm H2b X 378 X
or
S02 0.169 ppm H2S
=
106 Btu HHVg
List of Symbols for Appendix C
7oS - Percent sulfur content of coal on moisture ash-free basis.
HHVC - High heating value of coal on a moisture ash-free basis.
HHVg - High heating value of fuel gas.
Ng - Gasification efficiency. Heat content of product gas divided
by heat content of coal.
scf - Standard cubic foot (@ 60 F and 30 in. Hg).
-------
APPENDIX D
CALCULATION OF RELATIVE GAS PRESSURE DROP
-------
APPENDIX D
CALCULATION OF RELATIVE GAS PRESSURE DROP
This appendix describes the calculation procedure used in cal-
culating the relative pressure drops in distribution mains, valves, burners,
and flow passages downstream of the combustion zone when typical low-energy
gases from coal are substituted for natural gas.
The values given in Table 5 are relative pressure drops expressed
as multiples of the pressure drop incurred with natural gas in a given
distribution and combustion system. Values are given for two cases.
• Based on the fuel gas only (Column 5, Table 6). These
values apply to increases in pressure drop that might
be expected in distribution mains, valves, and premix
type burners when the low-energy gases are substituted
for natural gas. The values are based on the relative
increases in flow rates of fuel necessary to achieve
equivalent heat release to natural gas (Column 3, Table 6).
• Based on the stoichionetric mixture (Column 6, Table 6).
These values apply to increases in pressure drop that
might be expected in nozzle mix type burners and flow
passages downstream of the combustion zone. The values
are based on the relative volume flows of the stoichio-
metric mixtures (Column 4, Table 6) which are roughly
comparable to the relative volumes of combustion products
that would result when sufficient fuel and air are mixed
to achieve equivalent heat release to natural gas.
The following is a derivation of the expression used to calculate
the relative pressure drops given in Columns 5 and 6 of Table 6 including
the assumptions involved.
The pressure loss in a piping system, ignoring changes in elevation
and entrance and exit losses, can be expressed with the following:
fE (.1)
where Ap - pressure loss
f - friction factor
L - pipe length
D - pipe hydraulic diameter
KC - contraction loss coefficient
Ke - expansion loss coefficinet
K.f - fitting loss coefficient
P - gas density
V - average flow velocity at the system inlet.
*Expression is a modified form of equation 5-77 on page 115 of Reference B-l.
-------
D-2
Equation (Bl) is an adaptation of the Bernoulli equation and expresses the
pressure loss as a number of "velocity heads" [__ \-
\2 /
The first term in brackets in equation (Bl) accounts for the
pressure loss in straight pipe section due to friction between the walls
and the flowing fluid. The friction factor (f) is dependent to varying
degrees on the pipe Reynolds number and the surface finish or relative
roughness of the pipe material. Figure B-1 is a plot of the relationship
between the friction factor and the Reynolds number time Vf for various
types of pipe. In turbulent flow, which was assumed for these calculations,
f varies only slightly with Reynolds number for a given pipe material.
Therefore, because these calculations involve the comparison of relative
pressure drops of different gases in the same piping system the term
(4f L/'D) was assumed to be constant.
The terms KC and Ke are contraction and expansion loss coefficients
respectively, and account for pressure losses in sections of changing cross-
sectional area in tVie flow passage. They are a function of the ratio of
the upstream to downstream cross-section area and for a given pipe system
can be assumed constant over a wide velocity range.
K£ is a fitting loss coefficient accounting for pressure loss
in elbows, valves, orifices, and other flow obstructions in the flow stream.
Kr varies for different types of obstructions but for a given pipe system
can be considered constant.
Following from the above discussion where the bracketed terms
in equation Bl are considered as constant for a given flow system over a
wide range of flow velocities equation Bl can be replaced by a simpler form.
AP - ^- (B2)
2K
Here, K is considered as an overall pressure loss coefficient for the entire
flow system.
For these calculations the pressure loss of various low energy
gases is to be compared to natural gas when using the same piping or flow
system to achieve equivalent heat release. Equation B2 can be expressed as
-------
D-3
9
o
a
Ed
>
CO
oi
Z Efc
o ^
M OS
H
P S
^V
O
H
E~ Ed
i—i 2J
fe S
-------
D-4
^PLg = PLg V3Lg
APng - Png ^ng
Since K is only a function of the particular piping of flow system which
is assumed to be the same in both cases it can be dropped yielding:
uPLg = PLg
APng Png
(B3)
where Lg refers to the low energy gas
ng refers to natural gas.
The velocity V is related to the flow rate Q of the gas by
V = £-
where A is the characteristic cross-section area.
Again, assuming the same piping system hence (A) is the same for
both the low energy gas and natural gas, the term
APng
Png
.(Trig P Lg j
or
APng
PLg
i"QLs'a
!Qng
The term (t^1) in equation B4 is the relative flow rate of low-
Qng
energy gas to natural gas either based on the fuel alone or the stoichiometric
mixture and appears in Columns 3 and 4, respectively, of Table 6. The resulting
APLe
relative pressure drop (•._ °)
APng'
is shown in Columns 5 and 6 of Table 6.
Reference B-1. Unit Operations of Chemical Engineering, W.L. McCabe, J.C. SmitV
McGraw Hill, 1967.
-------
TECHNICAL REPORT DATA
(i'L'aSc read lu^l/iu tionii on the n iv/>c hcfw coniplfliiit;/
1. HLPORT NO. 2.
EPA-650/2-74-052
4. TITLE AND SUBTITLE
Study of Potential Problems and Optimum Opportunities
in Retrofitting Industrial Processes to Low and
Intermediate Energy Gas from Coal
7. AUTHOR(S)
D. Ball, G.Smithson, R. Engdahl, and A. Putnam
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Battelle — Columbus Laboratories
505 King Avenue
Columbus , Ohio 43201
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
NERC-RTP, Control Systems Laboratory
Research Triangle Park, NC 27711
3. RECIPIENT'S ACCLiiSION-NO.
4
5. REPORT DATE '
May 1974 i
6. PERFORMING ORGANIZATION CODL
8. PERFORMING ORGANIZATION RhPOFH NO
10. PROGRAM ELEMENT NO. f
1AB013; ROAP 21ADD-30
11. CONTRACT/GRANT NO.
68-02-1323 (Task 1)
13. TYPE OF REPORT AND PERIOD COVERED
Final; Through 5/21/74
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
The report describes a study to compile background information, including
environmental considerations, on the potential for retrofitting existing industrial
processes to the use of low- and intermediate-energy gas from coal. Potential
problems in retrofitting processes are analyzed. Processes where retrofit is
most attractive are identified, along with estimates of their annual energy
requirements. Also, current commercially available gasification systems and
representative gas cleanup systems are described and available data summarized.
17. KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
Air Pollution
Coal Gasification
Gas Burners
Furnaces
Processing
Gases
Desulfurization
18. DISTRIBUTION STATEMENT
Unlimited
b.lDENTIFIERS/OPEN ENDED TERMS
Air Pollution Control
Stationary Sources
Industrial Processes
Low-Energy Gas
Inter mediate -Energy Gas
Gas Cleaning
Particulates
19. SECURITY CLASS (This Report/
Unclassified
20. SECURITY CLASS (This page)
Unclassified
c. COSATI field/Croup
13B
13H
ISA
07D
07A
21. NO. OF PAGES
139
22. PRICE
EPA Form 2220-1 (9-73)
------- |