United States Industrial Environmental Research EPA-600/7-79-1S8a
Environmental Protection Laboratory December 1979
Agency Research Triangle Park NC 27711
Chemically Active Fluid
Bed for SOX Control:
Volume I. Process
Evaluation Studies
Interagency
Energy/Environment
R&D Program Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the
effort funded under the 17-agency Federal Energy/Environment Research and
Development Program. These studies relate to EPA's mission to protect the public
health and welfare from adverse effects of pollutants associated with energy sys-
tems. The goal of the Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology. Investigations include analy-
ses of the transport of energy-related pollutants and their health and ecological
effects; assessments of, and development of, control technologies for energy
systems; and integrated assessments of a wide range of energy-related environ-
mental issues.
EPA REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved
for publication. Approval does not signify that the contents necessarily reflect
the views and policies of the Government, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/7-79-158a
December 1979
Chemically Active Fluid
Bed for SOX Control:
Volume I. Process
Evaluation Studies
by
D.L Keairns, W.G. Vaux, N.H. Ulerich,
E.J. Vidt, and R.A. Newby
Westinghouse Research and Development Center
1310 Beulah Road
Pittsburgh, Pennsylvania 15235
Contract No. 68-02-2142
Program Element No. EHB536
EPA Project Officer: Samuel L. Rakes
Industrial Environmental Research Laboratory
Office of Environmental Engineering and Technology
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
PREFACE
The Westinghouse Research and Development Center Is carrying out a
program under contract to the United States Environmental Protection
Agency (EPA) to provide experimental and engineering support for the
development of the Chemically Active Fluid-Bed (CAFB) process. The pro-
cess was originally conceived at the Esso Research Centre, Abingdon, UK
(ERCA), as a fluidized-bed gasification process to convert heavy fuel
oils to a clean, medium heating-value fuel gas for firing in a conven-
tional boiler. Westinghouse, under contract to EPA, completed an ini-
tial evaluation of the process in 1971.1 Conceptual designs and cost
estimates were prepared for new and retrofit utility boiler applications
using heavy fuel oil. Westinghouse continued the process evaluation
from 1971 to 1973 and formulated an atmospheric pollution control demon-
stration plant program for retrofit of a utility boiler utilizing a
high-sulfur, high-metal-content fuel oil (for example, vacuum bottoms).2
The CAFB process represented an attractive option for use of these low-
grade fuels for which pollution control using hydrodesulfurization or
stack-gas cleaning was uneconomical. Application of a pressurized CAFB
concept with combined-cycle power plants was also assessed.2 Experi-
mental support work was initiated between 1971 and 1973 to investigate
two areas of concern - sorbent selection and spent sorbent processing -
to achieve an acceptable material for disposal or utilization. The
preliminary design and cost estimate for a 50 MWe demonstration plant at
the New England Electric System (NEES) Manchester Street Station in
Providence, RI were completed in 1975.-* Commercial plant costs were
projected and development requirements identified. Experimental support
of the sulfur removal system continued in order to provide a basis for
the detailed plant design. A number of design and operating parameters
from the preliminary design study that required further development were
identified. This three-volume report presents results of process
iii
-------
analyses, experimental studies, and application evaluations carried out
from 1976 to June 1979. This volume contains an assessment of the
market potential, sulfur control studies on limestone selection and
attrition, alternative sulfur sorbents, particulate control, and pro-
cess assessment. Our conclusions are based on available CAFB experi-
mental data and on Westinghouse analyses of CAFB plant designs and
performance projections. Results and analyses of exploratory tests with
lignite in a CAFB pilot plant (^1 MW ) currently being completed by Esso
Research Centre and of the lignite tests scheduled at Central Power and
Light Co.'slO MW plant are not included in this report. Our conclusions
are subject to the results and analyses of these experimental programs.
Volumes II and III of this report and prior reports issued under
this contract include:
Chemically Active Fluid Bed for SOX Control: Volume 2.
Spent Sorbent Processing for Disposal/Utilization, EPA-
600/7-79-158b, December 1979
Chemically Active Fluid Bed for SOV Control: Volume 3.
A. *
Sorbent Disposal, EPA-600/7-79-158c, July 1979
Solids Transport between Adjacent CAFB Fluidized Beds,
EPA-600/7-79-021, January 1979
Sorbent Selection for the CAFB Residual Oil Gasification
Demonstration Plant, EPA-600/ 7 -77-029, NTIS PB 266 827,
March 1977.
iv
-------
ABSTRACT
Selected process evaluation studies are reported in support of the
development of an atmospheric-pressure fluidized-bed gasification
process referred to as the Chemically Active Fluid Bed (CAFB) process.
The basic concept was designed for liquid fuels and utilizes a regenerative
limestone sulfur sorbent and produces a low- to intermediate-Btu fuel
gas. Limestone sorbent selection, sorbent attrition, alternative metal
oxide sorbents, particulate control, residual fuel feedstock availability,
and an updated process assessment are investigated. Limestone sorbent
selection results are presented for the EPA-sponsored CAFB demonstration
plant. Sorbent attrition and economics provide the primary criteria, as
most limestones are not limited by sulfur removal. Trace element, regen-
eration, and disposal characteristics should be considered. An attrition
tendency procedure was developed and utilized to measure the attrition
tendency of the Brownwood limestone sorbent selected for the demonstration
plant. Alternative metal oxide sulfur sorbents are reviewed that could
reduce the environmental impact of solids disposal and may improve process
economics. Three sorbents are identified for further study. Particulate
control requirements are identified for coal and residual fuels. The
availability of residual fuels for the process are reviewed, as are the
environmental impact of the process and operational considerations.
Application of the process will depend on the availability of suitable
feedstocks. Process modification for solid fuel application could
permit utilization of the process since availability of high-sulfur
residual oils will be increasingly limited to refinery applications.
-------
TABLE OF CONTENTS
Page
1. INTRODUCTION 1
2. SUMMARY AND CONCLUSIONS 2
Market 2
Sulfur Removal 3
Particulate Control 4
3. RECOMMENDATIONS 5
4. MARKET 6
Fuels 6
Utility Applications for the CAFB Process 11
Industrial Applications for the CAFB Process 16
5. SULFUR REMOVAL 28
Calcium-Based Sorbents 28
Alternative Metal Oxide Sorbents 38
6. ATTRITION OF FLUIDIZED-BED GASIFICATION SORBENTS 47
Conclusions 47
Apparatus and Procedure 48
Results 52
Discussion 72
7. PARTICULATE CONTROL 73
Control Requirements 73
Control Options 74
Ass-essment 76
8. ASSESSMENT 79
Process Economics 79
Potential 81
Environmental Impact 84
9. REFERENCES 87
APPENDIX
A. Attrition in the Bubbling Zone of a Fluidized Bed 91
vii
-------
LIST OF TABLES
Page
1. Gross Energy Purchased Compared with Shipments and Value 20
Added
2. Petroleum Refining Census Data: Useful Energy by Major 22
Source
3. Lime Industry Distribution of Energy Utilization by Source 24
4. Blast Furnaces and Steel Mills Census Data: Energy 25
Utilization 1958-1971, by Major Energy Source
5. Aluminum, Estimated U. S. Industry Sources and Uses of 26
Energy, 1971
6. Analysis of Brownwood Limestone 33
7. Summary of Sulfide Oxidation Tests 37
8. Desulfurizer Basis 39
9. Regenerator Basis 40
10. Sorbent Screening Results 45
11. Possible Sources of Particle Attrition in a Fluidized-Bed 49
System
12. Mean Values and Standard Deviations of the Size 52
Frequency Distributions for Limestone Sorbents
13. Before and After Size Distributions and Composition Data 53
for Test Sorbents
14. Specific Surface Increase Data and Analysis of Variance 69
of Differences
15. Increases in Sorbent Fractions <495 ym and Analysis of 70
Variance of Differences
16. Frequency Increase Data for Sieves Smaller than 701 um 71
and Two-Way Analysis of Variance
viii
-------
LIST OF TABLES (continued)
Page
17. CAFB Particulate Control Requirements 77
18. Cost Comparison 82
ix
-------
LIST OF FIGURES
Page
1. Sulfidation Rate of Brownwood Limestone 34
2. Influence of Sorbent Residence Time on the Sulfidation of 34
Brownwood Limestone
3. Sorbent Attrition Test System 51
4. Mean Values and Standard Direction of the Size Frequency 51
Distributions for Limestone Sorbents
5. Particle Size Frequency Curves for Attrition Screening 64
Treatment of Sorbents
6. Oil-Fueled CAFB Particulate Control Requirements 75
7. Coal-Fueled CAFB Particulate Control Requirements 75
-------
NOMENCLATURE
CAFB - Chemically active fluidized bed
DOE - Department of Energy
EPA - Environmental Protection Agency
ERCA - Esso Research Centre, Abingdon, UK
FW - Foster Wheeler Corporation
MeO - Metal oxide
NEES - New England Electric System
PER - Pope, Evans and Robbins
TGA - Thermogravimetric analysis (analyzer)
xi
-------
ACKNOWLEDGMENT
The achievements of this program are the result of the contribu-
tions of many individuals. The commitment, support, guidance, and
patience of S. L. Rakes, the EPA project officer is gratefully acknowl-
edged. P. P. Turner and R. P. Hangebrauck, Industrial Environmental
Research Laboratory, EPA, are acknowledged for their continuing contri-
butions and support for this work since its inception.
Numerous individuals at Westinghouse contributed. We gratefully
acknowledge the thoughtful and perceptive reviews and contributions of
Dr. D. H. Archer, Manager Chemical Engineering Research; and the work of
A. W. Fellers, C. A. Hill, and R. E. Brinza, who implemented the experi-
mental test programs.
xii
-------
1. INTRODUCTION
The CAFB (Chemically Active Fluidized Bed) gasification process, in
which limestone or dolomite removes the sulfur from fuel gas during the
gasification process, was developed to permit the utilization of high-
sulfur residual fuel oil or refinery bottoms in conventional boilers by
producing a low-sulfur fuel gas. Coal is also being investigated as a
fuel. The process can be operated as a once-through, limestone sorbent
system, a sorbent regeneration/sulfur recovery system, or a sorbent-
regeneration system without sulfur recovery by capturing the sulfur-rich
gas from the regenerator with the spent stone. The spent stone from
each system alternative can be processed to minimize the environmental
impact of the waste stone for disposal or to provide material for poten-
tial market utilization.3*4
Under contract to the U. S. Environmental Protection Agency (EPA),
Westlnghouse has carried out system analyses and laboratory support work
on sulfur removal, solids transport, processing of spent sorbent for
disposal or utilization, and the environmental impact of processed and
unprocessed residue disposal.3*4 Esso Research Centre, Abingdon,
UK (ERCA) has carried out pilot-scale tests to investigate sulfur
removal. ^ At San Benito, Texas, a 10 MW demonstration plant has been
retrofitted by Foster Wheeler Energy Corporation and Central Power and
Light Co. and is being tested.*>
Work was performed to assess the potential market applications, to
develop a basis for calcium-based sulfur sorbent selection, to determine
the potential for alternative sulfur sorbents, to evaluate particulate
control requirements, to identify and compare spent sorbent processing
options, and to determine the environmental impact of the disposal of
spent calcium-based sulfur sorbents. The results of this work has been
reported and provides the basis for the engineering evaluation.
-------
2. SUMMARY AND CONCLUSIONS
MARKET
Oil and gas represent about 30 percent of today's electric genera-
ting capacity, and the Fuel Use Act of 1978 precludes the use of oil and
natural gas for existing plants after 1990. Many technology options
exist for using coal, low-grade fuels, or alternative energy supply-and-
deraand technologies; and many international, institutional, and legisla-
tive actions will affect the choices to be made. This market assessment
was carried out by evaluating the present CAFB concept within the pre-
sent energy environment.
The CAFB process was conceived and designed for liquid fuels. The
CAFB demonstration plant** is scheduled to carry out tests with residua
and lignite fuels. The performance of the present CAFB configuration
using solid fuel will provide perspective on the ability of the present
concept to achieve required efficiency and environmental objectives.
Our evaluation indicates that the CAFB configuration will require modi-
fication to achieve these objectives with solid fuels. After modification
of the present CAFB configuration to process solid fuels, further evalua-
tion would be required to include comparisons with other fluidized-bed
processes designed for the gasification of solid fuels. The market
assessment considers both liquid and solid fuels. Demonstration plant
tests with lignite will provide further information that will result in
clarification of the present conclusions:
High-sulfur vacuum bottoms containing high-metal organic
complexes and produced from vacuum distillation of atmos-
pheric residual oil remain the most attractive fuel for
the CAFB process.
The availability of high-sulfur residual oil for CAFB pro-
cessing is decreasing.
-------
Solid feedstocks such as lignite, tire scrap, or wood
refuse are attractive fuel options; efficient utilization
is projected to require design modifications in the pre-
sent configuration, which was conceived for liquid fuels;
demonstration plant tests are planned to determine per-
formance with lignite.
The potential utility capacity exceeds 100,000 MW; imple-
mentation will depend on the availability of suitable
feedstocks; unless the concept is modified for solid fuel
application that availability will be low.
The generation of steam within a petroleum refinery is the
most promising industrial application for the CAFB pro-
cess; only the availability of residual fuels will limit
the market.
SULFUR REMOVAL
Sorbent selection will be determined by the attrition
characteristics of the sorbent and its cost.
A test apparatus and procedure was developed and demon-
strated to compare sorbent attrition tendency; the mechan-
isms include calcination, thermal shock, grid jets,
bubbling bed, and freeboard phenomena.
Brownwood limestone was compared with three reference
sorhents to assess its attrition tendency.
Brownwood limestone, selected for the demonstration plant,
is acceptable for regenerative operation; preliminary
tests indicate air oxidation of the spent sorbent from
once-through operation would not result in an environment-
ally acceptable material for direct disposal.
Zinc oxide (ZnO) is the most attractive alternative sor-
bent for in-situ desulfurization in a CAFB gasifier based on
thermodynamics.
Iron oxide (FeO) could be used for external desulfurization.
-------
PARTICULATE CONTROL
CAFB liquid fuel gasification is expected to require con-
ventional cyclones before and after the boiler to meet
emission standards.
Lignite gasification is expected to require conventional
cyclones before the boiler and electrostatic precipitators
or filters for final control.
-------
3. RECOMMENDATIONS
Fluidized-bed gasification technology can be utilized to process re-
sidual oil, coal, or other low-grade fuels to produce a fuel gas for in-
dustrial or utility applications. This report has reviewed the capability
of a specific fluidized-bed concept, the CAFB process. The following areas
are recommended for further study and investigation in order to determine
the capability of the CAFB process.
Review the market potential following the scheduled demonstration
plant tests using lignite. Available data support application of
the technology to low-grade petroleum liquids. Our evaluation
indicates, however, that the present CAFB configuration will re-
quire modification if it is to utilize solid fuels. This would
limit the market potential, and it is thus important to understand
and evaluate the scheduled lignite test results.
Continue development work on the high-temperature sulfur removal
system. This work has broad application (the CAFB process, designed
for liquid fuels alternative gasification processes using "hot"
gas cleaning, and fluidized-bed combustion processes, e.g.,
selected regenerative FBC and PFBC concepts) and merits further
development. Specifically:
Continue development work on the processing of spent limestone
sorbent from the regenerative process for disposal and utili-
zation. Options recommended include dry sulfation, direct
disposal, and briquetting.
Investigate methods for air oxidation of once-through
calcium-based sulfur sorbents to achieve a material
acceptable for disposal or utilization.
Continue development of alternative sulfur sorbents applic-
able to CAFB and other gasification processes - ZnO and FeO.
-------
4. MARKET
It was clear when EPA began CAFB development work several years ago
that a process to produce a clean, hot, fuel gas from high-sulfur
residua (atmospheric or vacuum) had substantial potential. At that
time
Sulfur dioxide (802) emission regulations were in effect
and stricter regulations were planned, so combustion of high-
sulfur residua required some provision for emission control.
Residua hydrodesulfurization was (and is) expensive.
Natural gas supplies were shrinking rapidly due largely to
federal price controls, and many gas-fired boilers were
going to be either shut down or switched to oil firing.
Thus, a clear path to CAFB commercialization was the use of a non-
compliance fuel, high-sulfur residua, to feed a CAFB retrofit onto
either a residua-fired boiler capable of gas firing or a gas-fired
boiler. That path became even clearer in 1975 when Foster Wheeler
developed interest in a Central Power & Light 10 MW CAFB retrofit on a
San Benito, Texas, utility boiler,7 and the Texas Railroad Commission
issued a directive," since rescinded, to Texas utilities in December of
1975 to schedule large reductions in natural gas consumption for power
generation. This section will examine two aspects of what has happened
since 1975 to that apparent market for CAFB: fuels and applications.
FUELS
Clean, low-sulfur fuels were never proposed for CAFB processing
because they were suitable for direct combustion, i.e., they were in
compliance with EPA requirements. High-sulfur oils that contained
-------
nitrogen compounds and usually heavy-metal organic complexes, however,
were proposed for CAFB processing. As listed in Table 5 of Volume I of
our March 1975 report,^ those fuels were either high-sulfur residual oil
produced from atmospheric distillation of sour crude or high-sulfur
vacuum bottoms produced from vacuum distillation of atmospheric residual
oil. We concluded in the 1975 feedstock assessment that atmospheric
residual oil would not be imported for a CAFB application because of a
national need to minimize oil imports. That position has been strength-
ened considerably since 1975. Another conclusion in the 1975 assess-
ment, that high-sulfur crude oil was not likely to be a CAFB feedstock,
has also been strengthened since 1975. The recent "brink of disaster"
situation regarding gasoline and home heating oil production in the
United States due to shrinking domestic crude oil supplies and the ceil-
ing imposed on oil importation absolutely dictates that all crude oil
and atmospheric residual oil be processed by domestic refiners for dis-
tillate fuel production. Thus, the only CAFB feedstock possible from
those mentioned in the 1975 assessment is high-sulfur vacuum bottoms.
Since 1975, other feedstock possibilities have developed. Foster
Wheeler has proposed coal, particularly lignite, as a feedstock for the
San Benito and other CAFB units.6 Also, the burgeoning synfuels program
in the U.S. raises the possibility of coal- or oil-shale-derived liquid
feedstocks for CAFB. Finally, wastes and refuse from pulp and paper,
petrochemical and plastics plants, and scrap rubber have been proposed.
Recently, the price decontrol of heavy crude oils (16ฐ API gravity or
lower) also raised the possibility of an increased quantity of vacuum
bottoms derived from refining these heavy crudes.
In order to assess the use of the possible CAFB feedstocks just
mentioned, two principal considerations need to be addressed, availabil-
ity and suitability.
Availability of Fossil Fuels
Recent discussions with oil refinery architect/engineering firms
and with synfuel project sponsors have provided insight into the
-------
probability of utilization of shale oil, coal-derived oil, or heavy
crude refining residua for a CAFB unit. Also, recent pronouncements
from the legislative and executive branches of the federal government
indicated the direction of future political action.
A continuing transportation fuel (gasoline, jet fuel, and diesel
fuel) squeeze is projected for the domestic market. We must, for
reasons of national security and a sound economy, decrease our depend-
ence on imported oil on the one hand, and, on the other, increasingly
obtain these distillate fuels (plus home heating oil and domestic com-1
bustion turbine peaking fuel) from declining domestic sources. The
efforts to do so have a common drawback. Distillate fuels contain more
hydrogen than the raw materials (heavy crude oil, oil shale, or coal),
from which they must be derived. That hydrogen deficiency, plus
the increased need for hydrogen to desulfurize or denitrogenate the sour
crudes and syncrudes that will make up a growing percentage of our dom-
estic supply, places a substantial new demand on domestic hydrogen sup-
ply. The hydrogen, wherever possible, will be supplied by residua from
the refining operation through partial oxidation, either in a Shell/
Texaco-type partial oxidation system or in a fluid coking/Flexicoking-
type system. The ash-containing residual materials from direct lique-
faction of coal will also be consumed in hydrogen production. Present
plans of the developers of H-Coal, SRC-11, Exxon Donor Solvent, and
COGAS processes indicate that all ash-containing material will probably
be used for hydrogen generation in oxygen-blown gasification units.
Tar sands operations will use the residual as coker feedstocks and use
the low-sulfur coke as boiler fuel for raising steam to be used in the
tar/sand separation. Oil-shale-derived oils will require denitrifica-
tion from hydrogen produced via partial oxidation of residua. Heavy oil
refining will, in turn, require more fuel and more hydrogen for upgrad-
ing the reformer and cat cracker feed streams and will consume the high
yields of residua to produce that fuel and hydrogen. CAFB is not
presently designed for such hydrogen production.
-------
Politically, restrictions on imports of oil plus the increasing
demand by third world nations^ for distillate fuels, which is shrinking
the supply of residua from the Caribbean area, and the pressure by the
Department of Energy (DOE) for utilities to convert boilers from oil to
coal and to natural gas all indicate shrinking availability of residua
for CAFB processing for utility applications.
Alternative Fuels
While the CAFB was proposed for gasification of high-sulfur
residua, other nonfossil fuels are available.
-.
The following abstracts and titles have been reviewed to identify
alternative fuels:
Engineering Index (manual search) Abstracts 1974-1976
Appl. Sci. & Tech Index (manual search) Titles 1974-1976
Engineering Index (Lockheed computer search) Abstracts
1970-1976
Chemical Abstracts (Lockheed computer search) Abstracts
1970-1976
EPA Solid Waste Int. Retrieval Service (computer) Abstracts
1970-1976
This search yielded about 400 references and abstracts describing
fluidized-bed gasification, pyrolysis, and incineration.
Waste fuels that have been identified for gasification, pyrolysis,
or incineration are listed below.
The extensiveness of this list suggests that practically any burn-
able waste may potentially be incinerated or gasified in a fluidized-bed
process such as CAFB. Some wastes such as sewage sludge are very wet,
while others such as waste plastic are very dry.
-------
Alternative Fuels
Bark
Groundwood mill sludge
Kraftmill sludge
Paint sludge
Rubber waste
PVC
Municipal refuse
Vinyl chloride monomer
Ore sludges
Mixed plastics (PE,PP,PS)
Polyethylene
Sewage sludge
Plastics from municipal
refuse
Manure
Tires
Sawdust, peanut shells, rice
husks
Coal washery rejects
Refinery waste
Agricultural & forestry waste
Distillery slops, packing-
house waste
It seems that the fuels best suited to the CAFB are those that are
rich in sulfur and that will exploit the sulfur-capturing potential of
the process. The sulfurous fuels include rubber tires (0.95 to 1.1 wt %
sulfur), rubber scrap (about 2% sulfur), sulfide-containing wood-
digestion liquors, and coal washery rejects. There are, in addition,
vast stores of obsolete chemical munitions stored in the USA containing
over 300 million pounds of toxic fill (TRW, VII, p 255, XIV, p 153).
Some of these are sulfur compounds, such as sulfur mustard (ClCt^C^^S,
for which the recommended disposal method is incineration. Such obso-
lete munitions may well be suited to destruction in the CAFB along with
recovery of useful energy and control of sulfur emissions.
Suitability
The CAFB process as presently configured was not designed to use
the solid feedstocks proposed as possible fuels, such as lignite, tire
scrap, or wood refuse. Tests planned at the San Benito, Texas demon-
stration plant will provide perspective on performance with lignite.6
10
-------
The liquid residua feedstocks are suitable (but as discussed above will
be increasingly unavailable). Their suitability stems from the gasifi-
cation mechanism proposed by ERCA:^
In the shallow fluidised bed of the gasifier there is a rapid
circulation of lime between top and bottom. Indications are
that coke is laid down on the lime in the upper portion of
the fluid bed by oil cracking and coking reactions and that
this coke burns off in the lower portion where oxygen is sup-
plied by the air distributor.
Oil as a feedstock permits this coke laydown on the circulating .
limestone; solids feedstocks such as lignite do not. Instead, lignite
is present at a high concentration in the gasifier bed to maintain an
adequate gasification rate and is fed at that concentration from the
gasifier to the regenerator. The concentration of unreacted fuel fed to
the regenerator, therefore, is much higher for lignite than for oil gas-
ification. As a result, proper regeneration of sulfided stone may not
be achieved because of the high heat release from the reaction of the
air in the regenerator with unreacted lignite. This difficulty has not
yet been overcome by the CAFB unit at Abingdon or by the plant at San
Benito, and until it is, we cannot say that solid feedstocks are suit-
able for CAFB processing.
These constraints on the availability and suitability of fuels for
CAFB indicate, at best, a very restricted market based on special local
conditions that may make a residuum available for a sufficient time to
justify a CAFB investment. Generally, such an availability of a suit-
able feedstock for the present configuration cannot be expected in the
forseeable future in the U.S. Modifications to the concept to permit
utilization of solid fuels would extend the feedstock availability.
UTILITY APPLICATIONS FOR THE CAFB PROCESS
Development of the CAFB has been devoted almost entirely to the
atmospheric pressure operation of the process. At atmospheric pressure
11
-------
the CAFB is suitable for providing low-heating-value gas to a conven-
tional utility boiler. The feasibility of operating the CAFB at pres-
sures suitable for combined-cycle power generation has previously been
evaluated.^ These two utility applications of CAFB are influenced by
several market factors.
Atmospheric Pressure CAFB
As mentioned briefly earlier, the use of a CAFB retrofit to a gas-
/oil-fired boiler has been its most likely application. The present
U.S. policy of conversion away from oil to coal or to the newly plenti-
ful deregulated natural gas or even to nuclear* has an obvious effect on
a process designed to use oil, even high-sulfur, high-heavy-metals-
content vacuum bottoms oil.
Potential utility sites for a CAFB retrofit are abundant. A survey
of the gas-fired utility boilers (nearly 85 percent of which can also fire
oil) installed in the 48 contiguous states can be summarized as
follows:
Year Commissioned Total MW
1978 516
1977 1,191
1976 2,686
1975 2,878
1974 5,014
1973 5,286
1972 4,531
1971 5,356
1970 4,280
1969 2,650
1968 3,462
1967 6,007
*(President Carter in his announcement of a new NRC Chairman on
Dec. 7, 1979).
12
-------
Year Commissioned Total MW
1966 4,479
1965 2,771
1964 3,465
1963 3,325
1962 2,042
1961 3,104
1960 2,742
1959 3,216
1958 3,201
Pre-1958 33,360
TOTAL 105,562
The preponderance of this capacity is in the "sunbelt":southern
California, Nevada, Arizona, New Mexico, Texas, Louisiana, Mississippi,
and Florida; and essentially all of the capacity commissioned since 1958
has gas/oil capability. The gas-only units are mostly small, pre-1950
installations, many of which are retired or are on peaking service. The
utility groups, such as the Florida Operating Group, Middle South Util-
ities, Texas Utilities Companies, the SCEC Power Pool, and the
California-Nevada Area Group, account for a large majority of the capac-
ity listed. These utility groups are in the region where population
growth is pressuring the utilities to continue to expand generation
capacity. Admittedly, there are no new, gas-fired units scheduled to
come on line in the future due to the severe gas shortage in the mid-
1970s, but the point is that all of the gas-fired utility capability
installed since 1958 must be under pressure to continue power genera-
tion. The question is, will this need be assisted by CAFB retrofit?
The answer to that question is a matter of fuel supply, timing, and
economics. The fuel supply was discussed in the preceding section.
Presently, the supply of fuel is dominated by a sharply increased
13
-------
natural gas supply. The "gas bubble" is certainly a transient phenome-
non. But how transient? The American Gas Association, in its publica-
tion, A.G.A. Monthly, has been indicating at least five years and
probably ten years as the forseeable duration of the bubble. Lower-48
gas production will be nearly 5 trillion ft^/yr higher in 1990 due to
deregulation than was projected under Federal Price Control.H Also,
agreements with Mexico for supply from their large and expanding gas
fields, plus future supplies of gas from Alaska, to take the pressure
off the southwestern U.S. supply of gas, indicate that the natural-gas-
fired utility boilers could remain natural-gas-fired for about another
decade. After 1990, CAFB may have available to it some of the 70+ GW
of, by then, 20-to-30-year-old units for retrofit. Our opinion, as pre-
sented in the preceding section, is that the general likelihood then of
suitable feedstocks being available for CAFB after 1990 is low, possibly
a special situation here or there, but no more than that.
The industrial boiler situation is similar. The AGA recognizes
that eventually natural gas will be too scarce a domestic resource to
permit utility boilers to burn it; thus, the projected availability of
utility boilers to CAFB retrofit in the 1990s. The industrial market,
however, is one that the gas industry is presently actively promoting
for installation of new capacity, all in the political guise of reducing
oil imports and accompanied by a campaign to "enable states to classify
dual-fuel customers to firm category (i.e., permit firm gas supply com-
mitments to industry) recognizing environmental benefits of gas
use."12 The industrial oil offsets now available total over
700,000 bbl/day of imported oil, with a projection by AGA to nearly
1.2 million bbl/day in 1980! We see little possibility that industry,
interested in low first cost, efficiency, and minimum environmental
intrusion, will be persuaded to use a CAFB system in the forseeable
future.
Previous assessment of the utility boiler population has shown that
the coastal regions (Federal Power Commission regions I, III, V, and
VIII) represent the areas of greatest interest to the CAFB with about
14
-------
700 gas- and oil-fired boilers existing up to 1000 MWe in capacity. *
FPC region I (the Northeast) is probably not applicable because of a
trend to convert these boilers to the use of coal. Almost all of the
boilers smaller than 50 MWe ( 300 in number) are 20 to 40 years old and
represent a market of limited applicability. The number of applicable
boilers, between 50 and 400 MWe in capacity, number about 300. They are
also located in the regions of potentially greatest low-grade residual
oil availability, if we assume a special local condition as discussed
earlier.
The ability to retrofit these boilers with the CAFB is an important
concern. A 50 MWe CAFB demonstration plant design has shown that space
in close proximity to the boiler may be very limited, requiring either
very long, hot, low-heating-value fuel gas piping, or removal of equip-
ment to provide space for the CAFB process.3 Burners, hot air ductwork,
windboxes, water walls, and I. D. fans may have to be modified. Two
studies of boiler retrofit with cold, low-heating-value fuel gas gen-
erated by coal gasification indicate that a fuel gas having the charac-
teristics of the CAFB fuel when fired in an extensively modified boiler
could achieve the maximum rating of the boiler, but the modification
could cost typically $!4/kW for a boiler whose original design fuel is
gas^'1^ (based on cold, low-heating-value fuel gas). The steam gener-
ator efficiency at maximum rating would also be reduced, and, if the
original design fuel for the boiler were gas, the unit would no longer
be capable of gas-firing. This is an area requiring further
definition.
Pressurized Low-Grade Residual Oil Gasification
The gasification of low-grade residua at elevated pressures in a
process similar to the atmospheric pressure CAFB can be used to supply
low-heating-value gas to a highly efficient and economically attractive
combined-cycle power plant.15
15
-------
The pressurized CAFB operation should be carried out with dolomite
as the preferred sorbent and can be operated with either sorbent regen-
eration by the steam/C02 reaction
CaS + C02 + H20 ^^ CaC03 + H2S ,
or with once-through sorbent utilization. The air regeneration scheme
used with the atmospheric pressure CAFB process does not appear economi-
cally feasible at pressures suitable for combined-cycle operation.
The critical market factors influencing the pressurized oil gasifi-
cation process are the availability of low-grade residual oils and the
competing economics of alternative power generation techniques. The
pressurized oil gasification process appears to be economically attrac-
tive when compared with alternative technology.^ Low-grade residua
oils will be no more available for the pressurized than for the atmos-
pheric CAFB process (see FUELS section).
INDUSTRIAL APPLICATIONS FOR THE CAFB PROCESS
The development of the CAFB process has been directed toward the
generation of steam using low-grade petroleum residua specifically for
electric utility application. The potential for utilizing the CAFB low-
heating-value fuel gas for industrial purposes - steam generation for
process steam, process heating, or power generation; process direct or
indirect heating; or process gas supply - has been assessed for the pur-
pose of identifying alternative applications that should be developed.
The goal of applying the CAFB process industrially would be to
reduce the industrial consumption of clean fuels, such as natural gas
and distillate fuel oils, or to permit the utilization of low-grade pet-
roleum residua that might already be consumed industrially in an envir-
onmentally acceptable manner. The feasibility of achieving this goal
has been evaluated by a survey of U.S. industries - their energy con-
sumption and process characteristics. A similar study has been carried
out by Battelle for the industrial application of low- and intermediate-
heating-value gas generated from coal.l"
16
-------
Limitations of the CAFB Process as an Industrial Low-Heating-Value Gas
generator
Characteristics of the CAFB process and industrial low-heating-
value gas applications limit the applicability of the CAFB process:
CAFB has been extensively tested only for low-grade petro-
leum residues. This limits fuel availability and
restricts application to industries located in regions
where these fuels may be available. In general, it would
be inconsistent with national policy to substitute
petroleum derivatives where coal is already in use.
The CAFB low-heating-value gas is hot (~870ฐC) and would
be difficult to cool because of its high tar content.
Cooling would also reduce the system efficiency by 10 to
15 percent. Many industries supply a large number of pro-
cess fuel needs by means of extensive gas or oil distribu-
tion systems. Cooling the CAFB gas would be required
and/or replacement of the distribution system by an expen-
sive, high-temperature gas distribution system. Also,
because of space limitations, in many retrofit cases the
CAFB gasifier may not be placed in close proximity to a
single large user of the low-heating-value gas, again
requiring cooling or expensive high-temperature piping.
The CAFB is an atmospheric-pressure, operating-gas producer
that cannot fill pressurized process gas requirements in
its present state of development. Also, many existing
fuel gas distribution systems are designed on the basis of
natural gas delivering at ~345 kPa and could not carry a
corresponding energy rate of low-heating-value gas even if
it were cooled.
The purity of the CAFB low-heating-value gas (containing
particulate, tars, etc.) would not satisfy the constraints
17
-------
of many process gas users. The flame temperature require-
ments of some industrial processes could not be supplied
by the CAFB fuel gas.
Many new-plant industrial applications such as steam gen-
eration or process heating could be satisfied by alterna-
tive techniques such as fluidized-bed combustion or
conventional direct combustion of low-grade fuels with
fuel gas cleaning. These techniques would probably be
economically superior to the CAFB process.
Many industrial boilers and process heaters may be incapa-
ble of retrofit to low-heating-value gas due to space con-
straints or they may suffer because of economics or
performance. ^*
On the basis of these generalizations, we conclude that the CAFB
could be applied economically only to industrial situations consisting
of the retrofit of large existing steam generators (process steam, pro-
cess heating steam, or power generation steam). Industries with large
steam requirements that are presently supplied by clean fuels and are
located in regions with potential low-grade petroleum fuel availability
could be considered.
Industries Surveyed
Energy consumption, pollution characteristics, and process needs of
the large U.S. industrial energy users were surveyed.2>3,17 -phe six
general industrial categories - food; paper; chemicals; petroleum and
coal; stone, clay, and glass; and primary metals - represented 77 per-
cent of the purchased energy consumed in manufacturing in 1967.
Table 1 indicates the 1967 energy consumption and energy intensity
(1000 Btu energy consumed/$(1967) value added) of the major U.S. indus-
tries. The largest energy-consuming category is primary metals, which
is led by blast furnaces and steel mills. The second highest consumer
is the chemical industry. The most energy intensive major category is
18
-------
petroleum and coal products at 284.38 (1,000 Btu/$(1967). Among the
individual industries the lime industry is the most energy intensive.
In addition to consuming purchased energy, several industries consume
significant captive energy (raw materials that are converted to products
and subsequently used to provide energy). The major consumers of cap-
tive energy are the petroleum industry and blast furnaces and steel
mills.
Assessment
Because of process requirements and the availability of residual
fuel oils the food and paper industries do not represent areas of poten-
tial CAFB application. The remaining four general industrial cate-
gories, however, do provide applications that satisfy some of the CAFB
criteria.
The chemical industries could utilize CAFB to supply the steam
requirements for a large chemical complex. Two factors will limit the
applicability to the chemical industries: the availability of residual
fuel oil and the feasibility of retrofiting existing steam generators
with the CAFB process. Only chemical plants located in regions of high
potential residual oil availability could be considered. This would
probably limit interest to U.S. coastal regions (PADs I, III, and V)1^
for the chemical industries or most other industrial application. Small
chemical plants with small steam utilization rates or large chemical
plants with numerous small steam generators distributed within the com-
plex are probably not of interest. The distribution of steam generator
sizes in U.S. chemical plants is unknown.
The criteria for the CAFB process may be most clearly satisfied in
the petroleum refining industry. Table 2 summarizes the energy con-
sumption by fuel source for petroleum refining. Large amounts of
natural gas are consumed by refineries, while large amounts of captive
energy are available in the form of residual oil, petroleum coke, and
refinery (still) gas. Large steam generators present in refineries
should be capable of CAFB retrofit, but, again, space may be limiting.
19
-------
Table 1
GROSS ENERGY PURCHASED COMPARED WITH SHIPMENTS AND VALUE ADDED,
HIGH-ENERGY-USING MANUFACTURING INDUSTRIES, 1967
NJ
o
20 Food and kindred products
2026 Fluid Bilk
2037 Frozen fruits and vegetables
2042 Prepared feeds
2611 Pulp nllls
2653 Corrugated and solid fiber boxes
28 Chemicals and allied products
2813 Industrial gases
2815 Cyclic Intermediates and crudes
2824 Organic fibers, noncellulosic
3221 Glass containers
3251 Brick and structural clay tile
33 FT 1 aery seta! Industries
3351 Copper rolling and drawing
3352 UuBlmiB rolling and drawing
ross energy
{trillion
BTUs)
CD
1.097.7
81.5
54.3
36.4
59.7
98.0
36.3
3,257.1
112.3
149.8
971.3
160.9
107.4
135.5
74.1
101.6
3,339.9
1.810.6
118.8
43.7
96.0
hlpBents
111 Ion
1967 S)
(2)
83,972
7,826
3,468
2,082
4,797
730
2,960
42,148
589
1,597
4,248
3,974
2,033
1,352
886
362
46,731
19.621
2,636
2,391
2.959
to shipments
(D(2)
(1,000 BTUs/
1967 $)
(3)
13.07
10.41
15.65
17.50
12.44
134.27
12.28
77.28
190. 1
93. 9
228. 4
46.04
52.83
IOO.21
83.62
280.72
71.47
92.28
45.03
18.28
32.45
46.53
alue added
(llllon
1967 S)
<4)
26,620
2,351
1,413
764
1.227
334
1,130
23,550
401
730
2,295
1,535
507
1,252
842
659
812
251
19.978
8,910
193
1,543
791
812
704
939
607
to value added
(1)<4>
(1,000 BTUs/
1967 $)
(5)
41.24
34.67
38.42
47.66
48.66
293.68
32.12
138.31
280.12
205.35
423.15
97.85
85.80
160.89
112.46
404.62
167.18
203.21
678.05
76.99
70.1
720.0
62.0
102. Z
96.6
Sources: Energy: Ready-mixed concrete (SIC 3273)-Table 22.1. Ml other Industries-like "reported energy' In Table \t\ (with cor ret-
-------
petroleum and coal products at 284.38 (1,000 Btu/$(1967). Among the
individual Industries the lime industry is the most energy intensive.
In addition to consuming purchased energy, several industries consume
significant captive energy (raw materials that are converted to products
and subsequently used to provide energy). The major consumers of cap-
tive energy are the petroleum industry and blast furnaces and steel
mills.
Assessment
Because of process requirements and the availability of residual
fuel oils the food and paper industries do not represent areas of poten-
tial CAFB application. The remaining four general industrial cate-
gories, however, do provide applications that satisfy some of the CAFB
criteria.
The chemical industries could utilize CAFB to supply the steam
requirements for a large chemical complex. Two factors will limit the
applicability to the chemical industries: the availability of residual
fuel oil and the feasibility of retrofiting existing steam generators
with the CAFB process. Only chemical plants located in regions of high
potential residual oil availability could be considered. This would
probably limit interest to U.S. coastal regions (PADs I, III, and V)18
for the chemical industries or most other industrial application. Small
chemical plants with small steam utilization rates or large chemical
plants with numerous small steam generators distributed within the com-
plex are probably not of interest. The distribution of steam generator
sizes in U.S. chemical plants is unknown.
The criteria for the CAFB process may be most clearly satisfied
in the petroleum refining industry. Table 2 summarizes the energy con-
sumption by fuel source for petroleum refining. Large amounts of
natural gas are consumed by refineries,while large amounts of captive
energy are available in the form of residual oil, petroleum coke, and
refinery (still) gas. Large steam generators present in refineries
should be capable of CAFB retrofit, but, again, space may be limiting.
21
-------
Table 2
PETROLEUM REFINING CENSUS DATA: USEFUL ENERGY BY
MAJOR SOURCE, 1958 TO 196717
Energy
Sources
1958
1962
1967
In Physical Units
Purchased
Coal, million short tons 1.069 .789 .777
Petroleum, million bbls 1.933 7.334 7.263
Gas, billion ft3 783.694 942.488 1,100.756
Other fuels, $ million* 7.300 11.200 20.600
Electric energy, billion kWh 9.115 12.147 17.474
Captive Consumption
Residual oil, million bbls 43.147 34.582 41.638
Other fuels, incl.
petroleum coke million bbls. . . . 17.415 40.827 42.055
Refinery (still) gas, billion ft3. . . 676.970 776.351 714.568
In Trillions of Btus
Total Energy 2,093 2,283 2,508
Coal
Captive Consumption
Other fuels, incl.
Refinery (still) gas
.... 904
.... 28
. . . . 11
. . . . 811
.... 23
.... 31
1,033
.... 271
.... 82
.... 680
1,115
20
43
975
35
41
1,189
217
192
780
1,336
20
42
1,139
76
60
1,172
262
198
711
*Includes gasoline, LPG, wood and purchased steam, and fuels not
specified by kind.
Source: Bureau of the Census, Census of Manufactures, Fuels and
Electric Energy Consumed, 1967.
22
-------
Stone, clay, and glass product industries do not appear to be prom-
ising applications for CAFB. The highest energy user in this classifi-
cation, hydraulic cement, consumes most of its energy in firing rotating
kilns. These kilns are capable of direct coal firing, and retrofit by
CAFB fuel gas may not be feasible. The most energy-intensive industry
in this industrial category, the lime industry, is a large user of coal,
coke and breeze, and natural gas but uses very little fuel oil as shown
in Table 3. Most of the energy consumption in the glass industry
requires high flame temperatures which would limit the applicability of
the CAFB fuel gas.
The primary metal industries fail to satisfy the criteria for CAFB
application. Blast furnaces and steel mills, which are the higher
energy consumers in this category, are also large consumers of captive
energy in the form of coke, breeze, blast furnace and coke oven gas
(Table 4). Residual oil would have very limited applicability in this
industry. The secondary steel industry, which does not have available
to it the captive energy of the integrated mills, could use a low-
heating-value gas for furnace operations of heat-treating and forming
where presently natural gas is used.17 We expect this application to be
unsuitable for CAFB because of extensive gas distribution and furnace
modification problems.
The primary aluminum industry is a large consumer of electrical
energy, some natural gas, and almost no residual oil (Table 5). Because
of its high electrical energy consumption the primary aluminum industry
is located in regions of cheap hydroelectric energy where residual oil
is generally unavailable.
Conclusions
The single most promising alternative application for the CAFB pro-
cess is the generation of steam within a petroleum refinery. No other
alternative applications have been identified, although others may exist
in special circumstances. Also, alternative applications may become
23
-------
Table 3
LIME INDUSTRY DISTRIBUTION OF ENERGY UTILIZATION, BY SOURCE
(trillion Btu, and percentages)*'
1947 19
(78.4) (63
Fuel Oil (total) 1.9 2
(3.3) (4
Gas 6.0 14
(10.5) (25
Other Fuels & Fuels n. s.k . 1.2 0
(2.1) (0
54 1958 1962 1 1967
.2 29.4 39.6 42 0
.0) (56.0) (64.5) (51.3)
.4 1.7 1.8 2 3
.3) (3.2) (2.9) (2.9)
.5 15.9 15.7 31 1
.9) (30.3) (25.6) (38.0)
.5 2.2 0.5 1.2
.9) (4.2) (0.8) (1.0)
Electric Energy
(gross energy
consumed basis)
Electric Energy
(useful energy)
Total
(gross energy
consumed basis)
Total
(useful energy)
3.0 3.3 3.3 3.7 5.3
(5.7) (5.9) (6.3) (6.2) (6.9)
0.7 0.9 1.0 1.2
56.9
(100)
54.5
55.9
(100)
53.5
52.5
(100)
50.2
61.4
(100)
58.9
1.7
81.9
(100)
78.3
Note: Figures in parentheses represent percentage distribution of total
which is based on gross energy consumption.
24
-------
Table 4
BLAST FURNACES AND STEEL MILLS CENSUS DATA: ENERGY UTILIZATION
1958-1971, (Trillion Btu) BY MAJOR ENERGY SOURCE17
Energy Source 1958 196
Useful Energy 2,423 2,7
Purchased 1,143 1,4
Coal 174 1
Coke 293 3
Petroleum 211 2
Gas 374 6
Other fuels* 30
Fuels, n.s.k.** 5 n.
Captive consumption 1,280 1,2
Coke & breeze 942 9
Blast furnace & coke 338 3
oven gas***
Gross Energy 2,548 2,9
Purchased 1,268 1,6
Captive consumption 1,280 1,2
2 1967 1971
68 3,223 n.a.
71 1,566 1,472
78 148 131
64 286 266
00 179 161
16 748 655
36 44 93
a. 43 8
77 119 158
97 1,657 n.a.
93 1,236 n.a.
04 421 n.a.
30 3,467 n.a.
33 1,810 1,802
97 1,657 n.a.
Estimates based on Bureau of the Census data.
*Includes gasoline, LPG, wood and purchased steam.
**Fuels not' specified by kind.
***Blast furnace gas is a coke by-product and included in the coke
energy.
n.a. - not available.
25
-------
Table 5
ALUMINUM, ESTIMATED U.S. INDUSTRY SOURCES AND USES OF ENERGY, 1971
(data are (1012) Btu except as marked)17
Distillate oil
Refining
18 3
0 8
75 1
A
94.2
9 7%
Baking
*
*
5.8
*
6.0
0.6%
Smelting
665 9
665.9
68.3%
Fabrication
94 0
f, 9
3.2
1.6
88.3
6.0
200.0
20.5%
Vehicles
0. 1
*
2.6
4.6
7.4
0.8%
Other
0. 1
1.9
2.0
0.2%
Total en
778. 2
6.9
3.4
2.4
171.2
8.7
4.6
975.5
ergy by type
Percentage
79.8%
0. 7
0.3
0.2
17.5
0.9
0.5
100%
N3
Note: Distribution of data along the columns is based on proportions derived from Table 31-1. Data exclude fuel usage in production of
alumina not sold to aluminum with the difference being taken from gas row. This single deduction is made on the assumption that the
difference lies primarily in calcining alumina.
*Denotes less than 0.05C1012) Btu.
-------
apparent when it is demonstrated that alternative fuels such as munici-
pal wastes, industrial wastes, or coal can be utilized by the CAFB.
Constraints on the availability and suitability of fuels for CAFB
indicate, at best, a very restricted market based on special local con-
ditions that may make a residuum available for a time sufficient to
justify a CAFB investment. Generally, such availability of a suitable
feedstock cannot be expected in the U.S. in the forseeable future.
27
-------
5. SULFUR REMOVAL
CALCIUM-BASED SORBENTS
The desulfurizing action of the CAFB process is usually represented
by the chemical reaction:
CaO + H2S > CaS + H20 .
The apparent simplicity of the process conceals the complex mechanism of
interaction between the fuel sulfur and the calcium-based sorbent as oil
is converted into a low-sulfur fuel gas. Thermodynamically, the equi-
librium for the reaction lies far to the right and predicts >95 percent
sulfur capture. Kinetic effects, process conditions, and the physical
and chemical state of the calcium sorbent, however, are dominant in
determining the extent of sulfur capture. Sorbent stone type, particle
size, the previous thermal and chemical history of the sorbent, and its
mechanical strength all influence the desulfurizing effectiveness of the
process and its operability. ERCA, for example, found one sorbent -
Conklin limestone - to be impossible for use in a fluidized bed because
of the high rate of attrition and elutriation of the stone as it was fed
to the gasifier.19 The large variety of potential calcium-based sor-
bents (e.g., limestone, dolomite, impure limestone, marble, aragonite,
marl) make it necessary to develop sorbent specifications from the
available data and to devise screening methods by which the suitability
of a particular candidate material can be assessed.
The relevant data come essentially from three sources:
The operating experience of ERCA on the continuous pilot
plant and on the batch gasifiers at Abingdon^
The laboratory tests and data assimilation carried out by
Westinghouse for the CAFB process evaluation^
28
-------
The laboratory tests and fluidized-bed work on other
sulfur-removal systems using calcium-based sorbents car-
ried out by Westinghouse and other contributors (Exxon,
Argonne National Laboratories [ANL], Consolidated Coal
[CONOCO], Foster-Wheeler [FW], Pope, Evans and Rabbins
[PER], and Combustion Power) on programs for EPA and DOE.
The suitability of a particular sorbent can be defined in an ideal-
ized manner. If a sorbent has a sufficiently high reaction rate with
the liberated fuel sulfur under the process conditions, it should effec-
tively desulfurize the fuel gas. Westinghouse has measured the reaction
rate of several stones with hydrogen sulfide (I^S) in a fuel gas using a
thermogravimetric apparatus (TGA); in all cases the reaction rate has
been sufficiently fast to capture 90 percent of the fuel sulfur, accord-
ing to the predictions of a model of fluidized-bed desulfurization
applied to CAFB operating conditions.3 None of the stones tested showed
a marked difference in reaction rate below 30 percent utilization of the
calcium fraction in the stone. These results lead to three general
conclusions:
The sulfur removal capability of different stones should
be similar at low calcium utilization (<30 percent).
The sulfur removal should improve as the bed height is
increased.
The sulfur removal should be high (>90 percent) at
calcium-to-sulfur (Ca/S) feed rates as low as 3/1.
The operating experience of ERCA5 can be compared with predictions
from laboratory studies. In batch studies, at a Ca/S makeup rate of 1.5
to 1.6, the sulfur removal efficiency for three stones (BCR 1691,
Denbighshire, and BCR 1350) was 75, 76, and 76 percent. At Ca/S makeup
rates of 2.83 and 2.71, the sulfur retention for BCR 1359 and Pfizer
calcite (Adams, MA) was 89 percent. These results apparently show that
sulfur removal is independent of the type of stone used; but ERCA, in
evaluating the effect of variation in the run conditions, concluded
29
-------
that Denbighshire was the superior sorbent, the remaining three being
equally active. Later tests, however, showed that Limestone 1359 was
marginally better than Denbighshire. It seems probable that there is
indeed, no difference in the inherent sulfur removal ability of the sor-
bents tested and that slight changes in operating conditions are
responsible for the differences noted.
Other tests by ERCA have shown that sulfur removal is improved
very little by increasing the bed depth; further, the Ca/S mole ratio has
always been much higher than 3/1 when high sulfur removal efficiencies
were achieved. Later test runs did show, however, that a deeper bed
gave greater sulfur removal. Recent evidence suggests that the sulfur
is not entirely released as H2S and that organic sulfur compounds in
tars escape from the bed of lime. Hydrogen sulflde introduced into the
bed is efficiently fixed by the lime, a result that agrees with
fluidized-bed studies of lime sulfidation at Westinghouse.
Although testing of a candidate sorbent will give information on
its reactivity with t^S, high reactivity does not ensure successful
operability of the process with a particular stone. Other factors, such
as stone attrition, fines recirculation, and air injection, may be dom-
inant in controlling the desulfurizing action.
Development of Sorbent Selection Criteria
Westinghouse has carried out sorbent selection studies for a CAFB
demonstration plant site in Providence, Rhode Island.-*
This evaluation of the relevant data has led to the definition of
stone selection criteria based on:
Acceptor properties of the stone for sulfur removal
Attrition resistance of the stone
Trace element emission characteristics
Regeneration characteristics
Suitability of spent sorbent for final processing for
disposal
30
-------
Economic availability of the stone.
A change in the demonstration plant site to San Benito, Texas mandated
that limestones available in the Texas area be assessed for their suita-
bility in the CAFB process.
Limestone from Texas and Mexico were evaluated as candidate sulfur
sorbents for the CAFB gasification demonstration plant at San Benito.
The procedure followed was to identify candidate stones using available
literature and expertise on the limestone industry in the area surround-
ing San Benito. Attrition was measured by elutriation losses suffered
by samples of these stones in a small fluidized-bed unit under calcina-
tion conditions. The samples were evaluated for their reactivity to l^S
in a fuel gas mixture at 870ฐC. Trace element analyses of the minerals
were carried out. This test procedure left unclear the distinctions
between most of the stones tested, and it was recommended that the cost
of the sorbent determine the choice.
A separate topical report*" was issued describing the results of
the sorbent selection study for the CAFB demonstration plant.
Brownwood Limestone Tests
The sorbent selected by FW for the La Palma demonstration plant,
Brownwood limestone, was evaluated. The received size distribution was
determined and the sulfur removal performance of the composite distribu-
tion obtained from TG sulfidation tests on nine size fractions of the
sorbent. The possible deactivation of the sorbent during prolonged
exposure to high temperatures and the feasibility of oxidizing sulfided
Brownwood limestone in air for disposal in the sulfated form were
examined.
Experimental Procedure
The reaction rate of limestone with l^S was determined in a modi-
fied Du Pont thermogravimetric reactor. A 20-mg sample of double-
screened limestone was suspended from the balance arm in a platinum mesh
31
-------
basket. A sheathed, chromel-alumel thermocouple, located about 1 cm
above the sample, measured the nominal sample temperature. The sample
was heated to temperature at a programed rate of 10ฐC/min. After com-
plete calcination, 0.5 percent H2S in a fuel gas mixture (2.5% CH^,
10% CO, 25% H2, 16.4% C02, N2) flowing at 600 ml/min (STP) was intro-
duced. The fraction of calcium sulfided was monitored, with time, by
the weight change of the sample.
Experimental Results
Brownwood Limestone Sulfidation. The analyses made on Brownwood
limestone are summarized in Table 6.
Sulfidation tests on Brownwood limestone were carried out on
nine particle size fractions. The particle size distribution received
consisted of fairly large particles, more than 50 percent of them larger
than 3000 ym. The coarse particles,however, were extremely reactive.
Data from the tests were compounded (by a weight-averaging basis) to
determine the Sulfidation rate as a function of sorbent utilization for
the material, as received. Figure 1 shows the rate of reaction obtained
for the 2380 to 3360 ym (6 to 8 mesh) size fraction and the composite
rate curve for the as-received material.
A simple model developed for desulfurization in fluidized beds2l>22
can be used to estimate Ca/S molar feed requirements for desulfurization
in once-through processes, using rate constants derived from TG data.
The model projections have agreed very well with fluidized-bed combustor
22
pilot plant data. The use of the model for desulfurization projec-
tions for the CAFB process however, requires that the sulfur be in the
form of H2S. The ability of the limestone to absorb organic sulfur is
unknown.
For 85 percent desulfurization in a 0.9 m bed fluidized at 1.4 m/s,
the reaction rate required is shown in Figure 1. (The method of making
the projections is detailed elsewhere.22) At this reaction rate with
32
-------
Table 6
ANALYSIS OF BROWNWOOD LIMESTONE
Chemical Analysis
Compound % Wei girt
Ca as CaO 53.3
Mg as MgO 0.53
C02 43.6
Al as A1203 0.94
Fe as F6203 0.97
Si as Si02 1.84
Na as Na20 0.026
K as K20 0.17
Cl 0.0032
Total 101.4
Ignition 42.4
Wt. Loss, %
Particle Size Distribution
U.S. Mesh Sieve % Weight
+5 31.3
5-6 20.6
6-8 42.2
8-10 3.9
10-12 1.2
12-14 0.3
-14 0.5
Grain Size
20-100 pm
33
-------
ft
I
-w
M
s
S>
O
O
H
H
U
O
-1
-2
RUN 385
BROWNWOOD TEXAS LIMESTONE ; 2380-3360 ym
870 C ; 600 ML/MIN.
0.5Z H2S ; FUEL GAS BALANCE
CALCINED NONTSOTHERMALLY UP TO
870 C IK JOOS FUEL GAS
85Z SO Removal
2
1-4 in/sec
C.9 m depth
DISTRIBUTION AS RECEIVED
.1
.4 .5 .6 .7
FRACTION SULFIDED
.8
.9
00
o
Figure 1 - Sulfidation Rate of Brownwood Limestone
Brownwood Limestone, 6/8 mesh
870ฐC, 0.5% H.S in fuel gas
1
-------
the composite rate data, 30 percent sorbent utilization is obtained.
The estimated Ca/S molar feed requirement for 85 percent desulfurization
is, therefore, 3/1 in a once-through process.
Effect of Sorbent Residence Time. In order to determine the effect
of time at temperature on the reactivity of Brownwood stone, a large
particle size fraction, 2380 to 3360 pm (6 to 8 mesh), of the sorbent
was calcined and held at 1000ฐC for three hours in fuel gas before sul-
fidation. The rate curve for sulfidation, shown in Figure 2, is nearly
identical to the rate curve of Brownwood limestone that was not exposed
to the three-hour treatment at 1000ฐC. No loss of sorbent reactivity
due to high temperature exposure is indicated.
Oxidation of Sulfided Brownwood Limestone. The oxidation of sul-
fided Brownwood limestone in air was tested on the TG apparatus as an
alternative possibility for disposal of spent gasifier bed material.
Since previous tests have generally shown that only a small fraction of
sulfided limestone can be oxidized in air before an impenetrable sulfate
shell forms, a method of activating the sorbent before it picks up sul-
fur was tested. The activation method used was to precalcine the sor-
bent at conditions that have been proved to produce a calcine with
wide-mouthed pores.^3 The larger pores formed should be better able to
accommodate the large sulfate ion formed when the sulfide is air oxi-
dized to sulfate.
Air oxidation was tested on three particle size fractions of
Brownwood limestone sulfided to levels of 20 to 70 percent. The sot
bents were initially calcined at three conditions: 870ฐC in fuel gas
(calcination simulating CAFB process), 900ฐC in 60 percent C02, and
850ฐC in 30 percent C02 (conditions under which calcines formed have
wide-mouth pores). The results are summarized in Table 7. The calcium
sulfide (CaS) fraction of 3000 urn particles that can be air oxidized at
800ฐC tripled as a result of sorbent pretreatment. The extent of
35
-------
oxidation also increased with decreased sorbent particle size and
decreased sulfidation of the sorbent. The maximum extent of oxidation
that occurred, however, was about 70 percent.
Conclusions
Selection of sorbents based on their sulfidation rates is impos-
sible because all sorbents are very active toward f^S absorption, and
organic sulfur removal by sorbents is not understood. Current sorbent
screening techniques, therefore, are based on evaluating the attrition
resistance of the stones and their economical availability. Trace ele-
ment, regeneration, and disposal characteristics should also be consid-
ered when the information is available.
Brownwood limestone appears to be an acceptable sulfur sorbent.
Limited data suggest that it is not deactivated by high-temperature
(1000ฐC) exposure. Air oxidation of the sulfide to sulfate at 800ฐC
does not appear to be an acceptable method of sorbent disposal. Methods
for improving the possible extent of sulfide oxidation, however, were
identified.
Recommendations
To develop generalized sorbent selection criteria for the CAFB pro-
cess the following areas should be investigated:
The reactivity of sorbents after exposure to the regenera-
tion- process.
The desulfurizing mode that is not described by I^S or S02
absorption. Because pilot plant data indicate that a
fraction of the sulfur escapes the sorbent bed, possibly
as organically based sulfur, it may be impossible to
achieve 90 percent sulfur oxide (SOX) removal by bed
height, superficial velocity, and sorbent activity
adjustments.
36
-------
u>
Particle Size,
Mm
2380-3360
420-500
44-74
Table 7
SUMMARY OF SULFIDE OXIDATION TESTS
% CaS Formed
CaS Oxidized after 20 Minutes
No Activation,
870ฐC/Fuel Gas Calcination
Activation by Calcination
900ฐC/60% C02 Calcination 850ฐC/30% C02 Calcination
40
71
40
66
28
40
50
20
15
11
34
13
51
44
42
56
62
66
71
63
-------
ALTERNATIVE METAL OXIDE SORBENTS
The CAFB process has been developed exclusively on the basis of
using natural, calcium-based sorbents. Alternative metal oxide (MeO)
sorbents may exist that could improve the regenerative performance of
the CAFB process or reduce attrition losses and improve the process eco-
nomics and environmental impact.
In order to investigate the potential of alternative sorbents, a
three-phase screening assessment is being conducted. The first phase
consists of an evaluation of sorbent thermodynaraic equilibrium desulfur-
ization and regeneration performance. The second phase considers mate-
rial and energy balance limitations characteristic of the sorbents. .. The
third phase surveys the cost and availability of the alternative sor-
bents and support materials. Reported here are the results of the phase
I activities.
Criteria and Basis for Selecting Alternative Sorbents
A range of probable operating conditions for the CAFB process must
be developed in order to provide a basis for assessing alternative Me
sorbents. Table 8 lists the conditions applied to the desulfur-
izer-gasifier. Both the atmospheric-pressure and the pressurized CAFB
concepts are evaluated. Both in-situ (as in the present calcium-based
CAFB concept) and external desulfurization is considered in the thermo-
dynamic assessment. The range of temperatures explored is based on a
lower limit that may result in excessive tar formation or in limiting
gasification reaction rate and an upper temperature limit that may
result in sorbent sintering, deactivation, or sorbent melting. Because
the gasifer can be operated over a range of air/fuel ratios from about
15 percent of stoichioinetric upward to very high levels, and various
methods of temperature control may be used (stack-gas recycle, steam or
water injection, heat transfer surface, etc.), the fuel gas compositions
may cover a very broad range. The composition presented in Table 8 is
based on ERCA experimental results^ that have been significantly
broadened to provide a reasonable range for studying its impact on the
38
-------
Table 8
DESULFURIZER BASIS
Fuel Residual fuel oil
Pressures 100 kPa (1 atm) and 1500 kPa
(15 atra)
TemperaLure Range 500ฐC (900ฐF) to 1200ฐC (2200ฐF)
or sorbent melting point
Low-Heating-Value Gas 1-5 mole % H20, 5-10% H2, 5-12%
Composition CO, 3-8% Cfy, 1-5% C2H4, 5-15%
C02> remainder N2.
Required Thermodynamic H2S 100 ppm
Control Level
alternative sorbents. The thennodynamic level of H2S control selected
in Table 8 is based on providing sufficient kinetic driving force to
satisfy the existing emission standards. Again, this number is somewhat
arbitrary since the dilution of the fuel gas may vary widely, depending
upon the operating conditions used. These assumptions are considered to
be sufficiently accurate for the therraodynamic screening of alternative
sorbents.
The following reactions are considered in the gasifier:
^5- MeO + C02
MeO + H2S ^=^ MeS + H20
reductants
MeO - >Me
Sorbent carbonate stability and MeO stability may be important sorbent
limitations.
Table 9 summarizes the basis for the sorbent regenerator.
Two pressure levels are selected that correspond to the atmospheric-
pressure and the pressurized CAFB concepts. A temperature range of
39
-------
Table 9
REGENERATOR BASIS
Pressures 100 kPa (1 atm) and 150 kPa
(15 atm)
Temperature Range 400ฐC to 1400ฐC or sorbent
melting point
Required Thermodynamic 10 mole 7,
S02 or H2 Level
100ฐC is applied in order to consider all potential sorbents. The sor-
bent melting point must not be exceeded. Thermodynamic levels of HoS or
S02 are selected to permit the application of relatively economical
sulfur recovery technology. Three regeneration schemes that result in
H2S or S02 products will be considered:
MeS + 3/2 02 ^=^= MeO + S02
MeS + H20 ^=^ MeO + H2S
MeS + C02 + H20 ^F^ MeC03 + H2S .
Competing reactions may also occur in the regenerator:
C + 02 =S=^ C02
C + H20 ;^=s CO + H2
MeS + Z02 ^^ MeS04
MeO + C02 ^^ MeC03 .
Carbon deposited on the sorbent during gasification will be present at
levels dependent upon the gasifier operating conditions. The influence
of this deposited carbon is neglected for this screening study, but the
effects of sulfate formation and sorbent carbonation are considered.
Alternative Sorbent Considered
Simple metal oxides (MeO^) have been screened thermodynamically.
All the MeOA in the periodic table, ranging from lithiumoxide (Li20)
to uranium oxide (UO-j) , were initially considered, but sufficient
40
-------
therraodynamic data could be found only for the following systems :^
sodium oxide (Na20) , magnesium oxide (MgO), silicon dioxide (SiOฃ),
calcium oxide (CaO), manganese oxide (MnO) , iron oxide (FeO), ferric
oxide (Fe203), ^6304), cobalt oxide (CoO), cupric oxide (CuO), cuprite
(Cu20), zinc oxide (ZnO), molybdenum dioxide (MoC^) , molybdenum trioxide
(Mo03), wolfram dioxide (W02) , litharge (PbO) .
Complex metal oxide forms of some of the above simple oxides such
as NaALC^, ^2X103, CaA^O^, and CaV20$ could be evaluated on the basis
of thermodynamic data for sulfate formation,^" but this has not been
attempted.
The sorbent CaO is included in the evaluation because it may be
superior to natural sorbent limestone when it is in the form of active
CaO carried on an inert support such as alumina. It also provides a
comparison between the well-known limestone potential and the alterna-
tive sorbent potential.
Desulfurization Performance
Four areas critical to the desulfurization performance of the
alternative sorbents were considered:
Metal oxide stability
Metal carbonate stability
Sorbent melting points
Desulfurization potential.
Metal Oxide Stability
The reduction of the MeO sorbent to the base metal in the reducing
atmosphere generated by the gasifier could lead to several problems:
loss of desulfurization potential, the generation of low-melting point
components, and so forth. The equilibrium for the reaction
MeO + H2 =5=^ Me + H20
was examined for a fuel gas having a ratio of XH20/XH2 = ฐ* 1 to 1*ฐ ^see
Table 8). Other reducing components were ignored (carbon monoxide [CO],
41
-------
methane [CH^] , etc.) for this feasibility screening. If the equilibrium
value of XH 0/XH2o for a given sorbent is greater than the actual fuel
gas value, then we assumed that the base metal would be stable.
Of the 16 metal oxides considered five were found to be clearly
unstable: CoO, CuO, CuC>2, MoC>3, and PbO. The metal oxides FeO, Fe203,
Fe3ฐ4> and MฐC>3 are uncertain ( thermodynamically) , with FeO being the
most stable of the iron oxide forms. The remaining sorbents are clearly
stable oxides. The uncertain sorbents are considered in further screen-
ing because they may well be kinetically stable oxides.
Metal Carbonate Stability
Limited data are available for the equilibrium
C02 .
Specific data could be found for only Na20, CaO, and MnO, indicating
that sodium carbonate (Na2C03) would be the stable sorbent form under
all desulfurizer conditions, calcium carbonate (CaCX^) would be stable
at atmospheric pressure for temperatures lower than 700ฐC and at
1500 kPa (15 atm) pressure for temperatures lower than 900ฐC. Manganese
carbonate (MnC03) would be unstable under any desulfurizer conditions.
The remaining eight sorbent materials, MgO, Si02, FeO,
> ZnOt Mo02, and W02 are believed to be unstable as carbonates, but
specific data could not be found.
Desulf urization Potential
The reaction equilibrium
MeO + H2S z~^MeS + H20
42
-------
for the remaining 11 sorbents was compared to the acceptance criteria in
Table 8 (100 ppm I^S). With water contents in the fuel gas of 1 to
5 percent, the ratio Xn2s/xH2ฐ at ecluilibrium must be less than 10~2 to
2 x 10-3.
All of the sorbents were found to meet this criterion except MgO,
N32C03, Fe203, Fe30^, and Si02ซ Manganese oxide would satisfy this
requirement at temperatures ranging from 400 to 800ฐC, W02 at from 400
to 500ฐC, Mo02 at from 400 to 500ฐC, and FeO at from 400 to 650ฐC. The
other sorbents, CaO and ZnO, satisfy the constraint over the entire
temperature range considered, 400 to 1300ฐC.
Depending upon the desulfurizer operating conditions (air/fuel
ratio, temperature control method, etc.) and the kinetics of the desul-
furization reaction, the I^S constraint applied could be relaxed to a
level as high as 1000 ppm. With this relaxation Fe203 would probably
become an acceptable sorbent and would broaden the temperature ranges
for MnO, W02, Mo02> and FeO.
Sorbent Melting Points
The melting points of the remaining six sorbents, CaO, MnO, ZnO,
W02ป Feฐป an<* MoOz* in tne oxideป sulfide, and sulfate forms were com-
pared to the applicable operating temperature ranges of the desulfurizer
and regenerator. None of the sorbents appeared to be limited by melting
except Mo02, which melts in its oxide form at about 800ฐC.
Regeneration Performance
The equilibrium S02 generation from the reaction system (sulfide
oxidation)
MeS + 3 MeS04 ^=S: 4 MeO + 4 S02
was determined for CaO/CaC03, Fe203, and ZnO. The most stable oxide
form of iron under oxidizing conditions is Fe203 rather than FeO. No
data could be found for MnO, W02, or
43
-------
In order for the sorbent to be acceptable for the sulfide oxidation
scheme, the S02 level generated should exceed 10 mole % (Table 9). The
sorbents Fe203 and Znฐ wil1 satisfy this constraint over the entire
temperature range considered (400-1 300ฐC) and at atmospheric or pressur-
ized operation and will provide huge reaction-driving forces. The sor-
bent CaO will satisfy the contraint at atmospheric pressure for tempera-
tures above 975ฐC and at 1500 kPa (15 atm) pressure at temperatures
above 1150ฐC.
Alternatively, the sorbents Fe203, MnO, ZnO, W02, and Me02 may be
regenerated from their sulfide form to their oxide form by reaction" with
steam: MeS + H20 ^=^ MeO + H2S. The H2S criterion calls for levels
greater than 10 mole % or, if we assume a pure steam reactant, the equi-
librium ratio X^-S/Xj^O 0.111. Only Fe203 can meet this constraint
for steam regeneration at temperatures above 500ฐC, with a maximum H2S
level of 30 mole %.
Calcium carbonate can be regenerated by the steam/C02 reaction
CaS + H20 + C02 ^^ CaC03 + H2S .
The constraint of 10 percent H2S can be met at neither atmospheric nor
pressurized operation. A 3 mole % H2S level can be reached at 1500 kPa
(15 atm) pressure and temperatures below 700ฐC.
Conclusions
Six sorbents remain after thermodynamic screening: CaO/CaC03, MnO,
ZnO, W02, FeO, and Mo02ซ Table 10 summarizes the results of the
screening.
The results indicate that only the CaO/CaC03 and the ZnO alterna-
tive sorbents could be used for in-situ desulf urization in a CAFB-type
gasifier. The FeO-based sorbent could be used for external desulfuriza-
tion at either 100 kPa (1 atm) or at 1500 kPa (15 atm) pressure. Exten-
sive work has already been performed on FeO sorbents, indicating great
potential on the basis of their highly regenerative nature. 29-31
44
-------
Table 10
SORBENT SCREENING RESULTS
Ul
aAchieves 100 ppm SC>2 ^n fuel gas.
Generates at least 10 mole % SC>2-
C0ener-ices at least 10 mole Z HฃS.
"The carbonate Is unstable at sufficiently high temperatures.
eNo data available.
fFor Fe20j.
^Generates only 3-5% !l2S.
Sorbent
CaO
CaC03
InO -
7.n3
W02
FeO
Mo02
Regeneration
Maximum Deaulfurlzer
Temperature ฐCa
100 kPa (1 atra) Press.
1300
700
800
1300
500
650
500
1500 kPa (15-atm) Press.
1300
900
800
1300
500
650
500
Sulftde Oxidation
Temperature Range, ฐCฐ
100 kPa (1 atra) Press.
>975
Nod
e
<1300
e
<1300f
e
1500 kPa (15 atm) Press.
>H50
Hod
e
<1300
e
500f
No
1500 kPa (L5 at*) Press.
No
<700S
No
No
No
>500f
No
-------
The sorbents MnO, W02, and Mo02 cannot be therraodynamically evalu-
ated because of lack of data. On the basis of the limited data avail-
able, however, their potential does not appear great, and further
assessment will be terminated.
Evaluation of the alternative sorbents CaO/CaC03, FeO (Fe203), and
ZnO will continue with an assessment of material and energy balance con-
straints, availability, and cost feasibility.
46
-------
6. ATTRITION OF FLUIDIZED-BED GASIFICATION SORBENTS
Natural materials vary in their resistance to attrition. To select
sorbents one must screen them by some laboratory procedure. The objec-
tive of this study was to develop a reproducible procedure for measuring
the attrition resistance of granular sorbents applicable to the CAFB.
Sorbent added to the CAFB bed first experiences thermal shock, then
calcination. Jets at the grid and bubbling above the grid tumble the
sorbent particles. The sorbent screening process we have developed
includes all of these processes that attrite particles by thermal, chem-
ical, and mechanical means.
The test apparatus developed is a 9.5-cm-id cell with a three-hole
grid. Test temperatures are maintained by a furnace surrounding the
cell. Our test procedure was to determine the gas flow required to form
8-cm-high jets in a bed of a particular sorbent. Sorbent was added to
an empty bed at 900ฐC and fluidized for 1 hr at 815ฐC at the predeter-
mined gas flow rate. Solids were sieved for particle size distribution
before and after the attrition treatment.
Replicate testings of Grove, Greer, Brownwood, and Pfizer sorbents
showed good repeatability between replicate tests and decisive differ-
ences in attrition tendency among different sorbents.
The apparatus and procedure developed here are not presented as a
universal method but rather as a prototype. This study demonstrates
that sorbents can be ranked decisively with regard to attrition
tendency.
CONCLUSIONS
An apparatus and a procedure have been demonstrated for
measuring the attrition tendency of granular sorbents.
47
-------
The procedure includes the attrition mechanisms present in
the grid region, the bubbling bed region, splashing in the
freeboard, thermal shock, and calcination.
The procedure for sorbent screening tested in this study
discriminates decisively between the attrition tendencies
of different sorbents.
The apparatus and procedure described here are not pro-
posed as a standard. This method serves, rather, as a
prototype and demonstrates that a standard screening
method can be developed.
Brownwood limestone, while not superior in attrition resistance,
is within acceptable attrition resistance limits as compared with
other sorbents in these tests.
APPARATUS AND PROCEDURE
The limestone sorbents used in the CAFB are subject to attrition
caused by gradual erosion or by sudden shattering. Our work has shown
that several mechanisms contribute to attrition in fluidized-bed
gasification.
The conditions of attrition-tendency testing must approximate con-
ditions in the CAFB. In any event, the testing should be at fluidized-
bed temperature and include the principal mechanisms of attrition.
Identified causes of attrition and factors affecting attrition included
in this test procedure are listed in Table 11.
The test procedure involves charging cold (room-temperature) stone
to a hot (900ฐC) reactor and fluidizing it in such a manner that there
are zones of jet action and free bubbling. A high freeboard allows
ejection of particles and uncushioned falling back to the bed surface.
Apparatus
The apparatus developed for measuring attrition tendency is a
cylindrical pipe 9.5 cm in diameter. The grid has three perforations so
spaced that jets will be equidistant from the wall and each other.
Pressure taps just above and below the grid allow the pressure drop
48
-------
Table 11
POSSIBLE SOURCES OF PARTICLE ATTRITION
IN A FLUIDIZED-BED SYSTEM
Attrition Source
Application to the Test Method
1. GRID JETS. Particles are accelerated
to high velocity and smash into the
roof of the jet. Particles tend to
shatter rather than abrade.
2. BUBBLING ABOVE THE GRID JET REGIME.
Bubbles cause rubbing and tumbling of
particles, tendining to abrade fine
chips from larger particles.
3. THERMAL SHOCK. Sudden heating of room-
temperature or colder particles to
above 800ฐC causes severe stress and
particle failure.
4. SPLASHING IN THE FREEBOARD. Bursting
bubbles throw particles into the free-
board; falling particles collide and
attrite.
5. TRANSFER LINES
6. CYCLONES
7. ROTARY VALVES
8. CHEMICAL REACTION. Changes in crystal
lattice structure cause interfacial
stress leading to fracture.
9. FLUIDIZED-BED SHAPE. A large value of
the bed height to bed-diameter ratio
encourages slugging and alters the
extent of attrition in the bubbling
zone and freeboard.
10. BED DEPTH. Bed depth contributes an
attrition force comparable to hydro-
static pressure. Local attrition rate
'is proportional to bed depth; average
attrition rate varies with the square
of bed depth.
1. The apparatus comprises three jets, each
8 cm high.
2. There is a 10-cm space above the top of
the jets in which there ia a vigorous
bubbling.
3. The test procedure includes pouring sor-
bent at 25ฐC into an attrition test cell
preheated to 900ฐC, then maintaining a
temperature of 815ฐC.
4. The test cell geometry is not designed
to lessen attrition from splashing.
5, 6, 7. Particle attrition occurs in
pneumatic transport and mechanical
valves. While these may be part of a
fluidlzed-bed system, they do not com-
prise a fluldized bed proper and are not
Included in the test equipment.
8. Not Included. Sulfation is not too dif-
ficult to achieve. Sulfidation causes
formation of metal eutectlcs and results
in severe fouling and corrosion of test
cell parts.
9. The height/diameter ratio is kept at
less than 2 and there is no slugging.
10. Bed depth is constant at 18 cm among
teats.
49
-------
across the grid to be measured. The entire
test cell is contained in a furnace. The sys-
tem is pictured in Figure 3. The high free-
board and moderate gas velocity controlled par-
ticle carry-over; although a filter was
installed in the exhaust line it captured only
negligible amounts of fines.
Procedure
r/R =
2+JT
Our first objective was to establish gas flow conditions that would
form jets 8 cm high in a bed of calcined sorbent. We filled the pipe
with calcined sorbent 8 cm deep, heated the bed to 815ฐC and gradually
increased the gas flow rate until the jets broke the surface of the bed
of solids. We recorded this gas flow rate and AP across the grid for
the following tests.
Our next objective was to measure the repeatability of extent of
attrition in a bed of sorbent 18 cm deep with 8-cm-high gas jets. We
heated the empty unit to 900ฐC and quickly poured in uncalcined sorbent
at room temperature through a feed pipe to a depth of 18 cm, then capped
the feed pipe. Thermal shock effects were evident: the sorbent
crackled and jumped as C02 was liberated and the particles were heated
swiftly. After capping the feed pipe we set the gas flow for jets 8 cm
high, maintained a bed temperature of 815ฐC, and let the gas flow for
one hour.
At the end of the test we turned off the furnace power and main-
tained a trickle flow of nitrogen through the bed to prevent intrusion
of C02 or humid room air.
After the system had cooled we weighed and sieved the bed solids
and assayed the solids for C02 content. It is worth noting the
replicability of sieve analysis. Figure 4 shows the means and
standard deviations for three replicate sievings of uniformity split
50
-------
O.g. 1700(40
Port for
Adding
Solids
ฃ^ป Exhaust
Pressure G*O9ป
Flow meter
_}Rซgul*tor
I-
1
.001
.0001
100
Figure 3. Sorbent Attrition Test System
Figure 4. Mean Values and Standard Deviation
of the Size Frequency Distributions
for Limestone Sorbents
-------
masses of Brownwood, Grove, and Greer sorbents. The figure Is to be
used in conjunction with Figure 5 in the results section to determine
the significance in differences between sieve analyses. The data in
Figure 4 are listed in Table 12.
RESULTS
Table 13 lists the before and after size distributions and
composition data for three replicates each of three sorbents. The tests
with Grove, Greer, and Brownwood limestones were run randomly to
minimize time trends. Pfizer dolomite was run in three sequential tests
as an afterthought. Particle size distribution data are graphed in
Figure 4. Particle frequencies are shown on logarithmic ordinates to
emphasize differences in small frequencies and on arithmetic ordinates
to accentuate differences in the large frequencies near the mode.
The results have been interpreted in three different ways.
Table 12
MEAN VALUES AND STANDARD DEVIATIONS OF THE SIZE FREQUENCY
DISTRIBUTIONS FOR LIMESTONE SORBENTS
Sieve Size
Mesh
8
12
16
24
32
42
60
115
250
325
Pan
Tabul.
Mm
2380
1810
1180
835
570
420
294
175
87
51
36
ited vai
Brownwood
Mean
0.1235
1.1056
0.4812
0.3695
0.2094
0.0024
0.0003
0.0001
0.0001
0.0002
0.0002
les are t
Std. Dev.
0.0118
0.0144
0.0027
0.0095
0.0057
0.0004
0.0001
0.0001
0.0000
0.0005
:he differet
Grove
Mean
0.0568
0.7986
0.5861
0.5970
0.4048
0.0079
0.0010
0.0006
0.0018
0.0009
itial free
Std. Dev.
0.0146
0.0948
0.0402
0.0649
0.0548
0.0001
0.0002
0.0001
0.0001
0.0003
Greer
Mean
0.0148
1.3723
0.6657
0.0662
0.0177
0.0013
0.0013
0.0016
0.0019
0.0015
0.0004
Std. Dev.
0.0065
0.0199
0.0227
0.0055
0.0019
0.0001
0.0002
0.0004
0.0003
0.0001
0.0001
[uencies (fraction at solids
mass within the size range d to d + Ad) * In [d + Ad) * d] .
52
-------
Table 13
BEFORE AND AFTER SIZE DISTRIBUTIONS AND
COMPOSITION DATA FOR TEST SORBENTS
53
-------
BROWNWOOD 1
ART FELLERS
7311A3219A1
16
15
2000
,1
OPERATOR NAME ;
ROTOMETER SERIAL NUMBER
ROTOMETER SETTING
ROTOMETER PRESSURE
STONE WEIGHT BEFORE (GRAMS)
STONE WEIGHT AFTER (GRAMS) 1494
STONE DEPTH (cm) 18
ATTRITION CELL USED 10cm
ATTRITION CELL PRESURE ATM.
GAS COMPOSITION N2
COMMENT : PRE-HEAT CELL 900C LOAD 2000gm STONE INTO CELL RE-HEAT
CELL TO 315C FLOW N2 FOR ONE HOUR DECREPITATION A
43.57% CO,
21.32% OX:
3 HOLE JET
************ BEFORE *************
SUM Fi/Dpi= 7.7501
I/SUM = 0.1290
SPECIFIC SURFACE= 46.5009
MESH MICRONS MEAN D2 MASS
In GRAMS
Dl
0
12
16
24
32
42
60
115
250
325
PAN
(3366)
2380
1397
991
701
495
351
246
124
61
43
(30)
2830
1810
1180
335
570
420
294
175
37
51
36
.346
.533
.346
.346
.346
.346
.355
.685
.709
.349
.346
0
0
0
0
0
0
0
0
0
0
0
.042
.589
.166
.127
.072
.000
.000
.000
.000
.000
.000
***FRACTION***
DIFFER- CUMULA-
ENTIAL TIVE
0.0428
0.5891
0.1664
0.1279
0.0725
0.0008
0.0001
0.0001
0.0002
0.0001
0.0001
1.0001
0.9573
0.3682
0.2018
0.0739
0.0014
0.0006
0.0005
0.0004
0.0002
0.0001
0.0000
0.
3.
1.
1.
1.
0.
0.
0.
0.
0.
0.
f/D
1798
2547
4105
5312
2715
0190
0034
0074
0253
0196
0278
. f/ln
0.
1.
0.
0.
0.
0.
0.
0.
0.
0.
0.
1237
1052
4810
3695
2095
"**-ป ^f
0023
0003
0002
0003
0003
0003
************* AFTER **************
SUM Fi/Dpi=
I/SUM
SPECIFIC
MESH
8
12
16
24
32
42
60
115
250
325
PAN
SI
SURFACE-
MICRONS
(3366)
2380
1397
991
701
495
351
246
124
61
43
(30)
MEAN
2380
1810
1180
835
570
420
294
175
37
51
36
13.
0.
80.
D2
In
Dl
.346
.533
.346
.346
.346
.346
.355
.685
.709
.349
.346
3974
0746
3847
MASS
GRAMS
2
45
15
12
7
0
0
0
2
0
0
.790
.820
.300
.550
.220
.370
.290
.930
.100
.770
.180
***FRACTION***
DIFFER-
ENTIAL
0.0315
0.5177
0.1729
0.1418
0.0816
0.0042
0.0033
0.0105
0.0237
0.0087
0.0020
CUMULA-
TIVE
0.9980
0.9664
0.4487
0.2758
0.1340
0.0524
0.0482
0.0450
0.0345
0.0107
0.0020
0.0000
0.
2.
1.
1.
1.
0.
0.
0.
2.
1.
0.
f/D
1325
8604
4651
6983
4313
0995
1115
6005
7274
7060
5650
f
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
/In
0911
^* -* *. j^
9714
* t ^ ซ^
4997
4098
2358
0121
** *. f- j.
0092
0153
0335
0249
*+ *> ~r y
0059
54
-------
BROWNWOOD 2
OPERATOR NAME ;
ROTOMETER SERIAL NUMBER
ROTOMETER SETTING
ROTOMETER PRESSURE
STONE WEIGHT BEFORE (GRAMS)
STONE WEIGHT AFTER (GRAMS)
STONE DEPTH (cm)
ATTRITION CELL USED
ATTRITION CELL PRESURE
GAS COMPOSITION
ART FELLERS
7311A3219A1
16
15
2000
1515.4
18
10cm 3 HOLE JET
ATM.
N2
43.57% OX
17.75% OX
roMMENT : PRE-HEAT CELL 900C LOAD 2000gm STONE INTO CELL
RE-HEAT CELL TO 815C FLOW N2 FOR ONE HOUR ^ECREPITATIOIT
************ BEFORE *************
SUM Fi/Dpi- 7.7501
I/SUM - 0.1290
SPECIFIC SURFACE- 46.5009
MESH MICRONS
b
12
16
24
32
*dfc42
W60
^115
250
325
PAH
(3366)
2380
1397
991
701
495
351
246
124
61
43
(30)
MEAN D2
In
Dl
2830
1810
1180
835
570
420
294
175
87
51
36
.346
.533
.346
.346
.346
.346
.355
.685
.709
.349
.346
MASS
GRAMS
0.042
0.589
0.166
0.127
0.072
0.000
0.000
0.000
0.000
0.000
0.000
***FRACTION***
DIFFER- CUMULA-
ENTIAL TIVE
1.0001
0.0428 0.9573
0.5891 0.3682
0.1664 0.2018
0.1279 0.0739
0.0725 0.0014
0.0008 0.0006
0.0001 0.0005
0.0001 0.0004
0.0002 0.0002
0.0001 0.0001
f/D
0.1798
3.2547
,4105
,5312
,2715
,0190
,0034
,0074
,0253
0196
1,
I.
1.
0,
0,
0,
0,
0,
0.0001 0.0000 0.0278
************* AFTER **************
SUM Fi/Dpi- 1^'5917
I/SUM - 0.0735
SPECIFIC SURFACE- 81.5504
MESH MICRONS
(3366)
8 2380
12 1397
991
701
495
351
246
124
61
43
(30)
MEAN D2
In
Dl
16
24
32
42
60
115
250
325
PAN
2380
1810
1180
835
570
420
294
175
87
51
36
.346
.533
.346
.346
.346
.346
.355
.685
.709
.349
.346
MASS
GRAMS
3.830
43.900
14.160
12.630
7.790
0.420
0.290
0.900
1.920
0.770
0.270
***FRACTION***
DIFFER- CUMULA-
ENTIAL TIVE
0.9969
0.0439 0.9530
0.5037 0.4492
0.1625 0.2867
0.1449 0.1418
0.0894 0.0524
0.0048 0.0476
0.0033 0.0443
0.0103 0.0340
0.0220 0.0119
0.0088 0.0031
0.0031 0.0000
f/D
f/lt
0.1237
1.1052
0.4810
0.3695
0.2095
0.0023
0.0003
0.0002
0.0003
0.0003
0.0003
f/ln
0.1847
2.7830
1.3769
1.7356
1.5682
0.1147
0.1132
0.5901
2.5323
1.7324
0.8606
0.1270
0.9451
0.4696
0.4189
0.2583
0.0139
0.0094
0.0151
0.0311
0.0253
0.0090
55
-------
BROWNWOOD 3
OPERATOR NAME ;
ROTOMETER SERIAL DUMBER
ROTOMETER SETTING
ROTOMETER PRESSURE
STONE WEIGHT BEFORE (GRAMS)
STONE WEIGHT AFTER (GRAMS)
STONE DEPTH (era)
ATTRITION CELL USED
ATTRITION CELL PRESURE
GAS COMPOSITION
COMMENT : PRE-HEAT CELL
ART FELLERS
7311A3219A1
16
15
7000.0
3
1658
13
10cm
ATM.
N2
9QOC LOAD
43.57%
28.32%
CD
3 HOLE JET
RE-HEAT CELL TO S15C FLOW H2 FOR ONE HOUR; FILTER
FILTER SINTERED METAL 3u 646.Igm BEFORE AND AFTER
20r>0gra STONE INTO CELL
ON EXHAUST
DECREPITATION
************
SUM Fi/Dpi=
BEFORE
1 / SUM
SPECIFIC
SURFACE=
MESH MICRONS MEAN
3
12
16
24
32
42
60
115
250
325
PAN
(3366)
2330
1397
991
701
495
351
246
124
61
43
(30)
************* AFTER
SUM Fi/Dpi-
I/SUM
SPECIFIC
SURFACE'
MESH MICRONS MEAN
8
12
16
24
32
42
60
115
250
325
PAN
(3366)
2380
1397
991
701
495
351
246
124
61
43
(30)
*************
7.7405
0.1291
46.4434
MEAN D2 MASS
In GRAMS
Dl
2830
1310
1180
835
570
420
294
175
87
51
36
.346
.533
.346
.346
.346
.346
.355
.635
.709
.349
.346
n.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
042
590
166
127
072
000
000
000
000
000
000
***FRACTION***
DIFFER- CUMULA-
ENTIAL TIVE
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0421
5907
1664
1273
0722
0008
0001
0001
0002
0001
0001
1.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0001
9530
3673
2009
0736
0014
0006
0005
0004
0002
0001
0000
0.
3.
1.
1.
1.
0.
0.
0.
0.
0.
0.
f/D
1769
2633
4103
5247
2668
0190
0034
0057
0230
0196
0278
0
1
0
0
0
0
0
0
0
0
0
f/ln
.1217
. 1082
. 4810
. 3680
.2087
.0023
.0003
. 0001
.0003
.0003
.0003
**************
MEAN
2380
1810
1180
835
570
420
294
175
87
51
36
13.
0.
83.
D2
In
Dl
346
533
346
346
346
346
355
685
709
349
346
9600
0716
7601
MASS
GRAMS
***FRACTION***
DIFFER-
ENTIAL
5.
58.
16.
13.
8.
0.
0.
1.
2.
1.
0.
540
560
090
610
030
460
330
220
770
090
370
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0511
5400
1484
1255
0741
0042
0030
0113
0255
0101
0034
f/n
CUMULA-
f/ln
TIVE
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
9966
9455
4055
2571
1316
0575
0533
0503
0390
0135
0034
0000
0.
2.
1.
1.
1.
0.
0.
0.
2.
1.
0.
2147
9835
2574
5031
2991
1010
1035
6429
9361
9709
9478
0
1
0
0
0
0
0
0
0
0
0
. 1477
.0132
. 4288
.3627
.2140
.0123
.0086
.0164
.0360
. 0288
.OOQ9
56
-------
GREER 1
OPERATOR NAME ;
ROTOMETER SERIAL NUMBER
ROTOMETER SETTING
ROTOMETER PRESSURE
STONE WEIGHT BEFORE (GRAMS)
STONE WEIGHT AFTER (GRAMS)
STONE DEPTH (era)
ATTRITION CELL USED
ATTRITION CELL PRESURE
GAS COMPOSITION
ART FELLERS
7311A3219A1
13
20
2000.0
1412.9
18
10 cm
ATM.
37.34% OC
10.52% 00
3 HOLE JET
N2
COMMENT : PRE-HEAT CELL 900C LOAD 2000gm STONE INTO CELL
RE-HEAT TO 815C FLOW N2 FOR ONE HOUR DECREPITATION
WEIGHT OF FILTER ON EXHAUST DID NOT CHANGE
BEFORE *************
SUM Fi/Dpi- 6.7872
I/SUM = 0.1473
SPECIFIC SURFACE- 40.7235
MESH
8
12
16
24
32
42
60
115
250
325
PAN
MICRONS
(3366)
2380
1397
991
701
495
351
246
124
61
43
(30)
MEAN
2830
1810
1180
835
570
420
294
175
87
51
36
D2
In
Dl
.346
.533
.346
.346
.346
.346
.355
.685
.709
.349
.346
MASS
GRAMS
0.005
0.731
0.230
0.022
0.006
0.000
0.000
0.001
0.001
0.000
0.000
***FRACTION***
DIFFER- CUMULA-
ENTIAL TIVE
1.0001
0.0051 0.9950
0.7316 0.2634
0.2303 0.0331
0.0229 0.0102
0.0061 0.0041
0.0005 0.0036
0.0005 0.0032
0.0011 0.0021
0.0014 0.0007
0.0005 0.0001
0.0001 0.0000
f/D
0.0214
4.0420
1.9520
0.2743
0.1070
0.0110
0.0156
0.0629
0.1609
0.1039
0.0361
************* AFTER **************
SUM Fl/Dpi- 7.4552
I/SUM - 0.1341
SPECIFIC SURFACE- 44.7317
MESH
8
12
16
24
32
42
60
115
250
325
PAN
MICRONS
(3366)
2380
1397
991
701
495
351
246
124
61
43
(30)
MEAN
2380
1810
1180
835
570
420
294
175
87
51
36
D2
In
Dl
.346
.533
.346
.346
.346
.346
.355
.685
.709
.349
.346
MASS
GRAMS
0.870
49.240
18.430
2.210
0.710
0.170
0.130
0.130
0.190
0.070
0.050
f/D
f/In
0.0147
1.3726
0.6657
0.0662
0.0176
0.0013
0.0013
0.0016
0.0020
0.0015
0.0004
f/ln
***FRACTION***
DIFFER- CUMULA-
ENTIAL TIVE
0.9993
0.0120 0.9873
0.6811 0.3062
0.2549 0.0513
0.0306 0.0207
0.0098 0.0109
0.0024 0.0086
0.0018 0.0068
0.0025 0.0043
0.0026 0.0017
0.0010 0.0007
0.0007 0.0000 0.1921 0.0020
0.0506
3.7627
2.1603
0.3661
0.1723
0.0560
0.0612
0.1423
0.3021
0.1898
0.0348
1.2778
0.7367
0.0883
0.0284
0.0068
0.0051
0.0036
0.0037
0.0028
57
-------
GREEK 2
OPERATOR NAME ;
ROTOMETER SERIAL NUMBER
ROTOMETER SETTING
ROTOMETER PRESSURE
STONE WEIGHT BEFORE (GRAMS)
STONE WEIGHT AFTER (GRAMS)
STONE DEPTH (cm)
ATTRITION CELL USED
ATTRITION CELL PRSSURE
GAS COMPOSITION
ART FELLERS
7311A3219A1
13
20
2000.0
1430.6
18
10cm
ATM.
37.34% 00
15.10% OO
3 HOLE JET
N2
COMMENT : PRE-HEAT CELL 900C LOAD 2000cra STONE INTO CELL RE-HI*
CELL TO 815C FLOW N2 FOR ONE HOUR DECREPITATION FILTER ON
EXHAUST LINE NO WEIGHT CHANGE
************ BEFORE *************
SUM Ft/Dpi- 6.7872
I/SUM - 0.1473
SPECIFIC SURFACE- 40.7235
MESH
8
12
16
24
32
42
60
115
250
325
PAN
MICRONS
(3366)
2380
1397
991
701
495
351
246
124
61
43
(30)
MEAN
2830
1810
1180
835
570
420
294
175
87
51
36
D2
In
Dl
.346
.533
.346
.346
.346
.346
.355
.685
.709
.349
.346
MASS
GRAMS
0.005
0.731
0.230
0.022
0.006
0.000
0.000
0.001
0.001
0.000
0.000
***FRACT
DIFFER-
ENTIAL
0.0051
0.7316
0.2303
0.0229
0.0061
0.0005
0.0005
0.0011
0.0014
0.0005
0.0001
ION***
CUMULA-
TIVE
1.0001
0.9950
0.2634
0.0331
0.0102
0.0041
0.0036
0.0032
0.0021
0.0007
0.0001
0.0000
f/D
0.0214
4.0420
1.9520
0.2743
0.1070
0.0110
0.0156
0.0629
0.1609
0.1039
0.0361
f/ln
0.0147
1.3726
0.6657
0.0662
0.0176
0.0013
0.0013
0.0016
0.0020
0. 0015
0.0004
************* AFTER
SUM Fi/Dpl-
I/SUM
SPECIFIC SURFACE-
**************
6.9750
0.1433
41.8504
MESH
3
12
16
24
32
42
60
115
250
325
PAN
MICRONS MEAN
(3366)
2380
1397
991
701
495
351
246
124
61
43
(30)
2380
1810
1180
835
570
420
294
175
87
51
36
D2
In
Dl
.346
.533
.346
.346
.346
.346
.355
.685
.709
.349
.346
MASS
GRATIS
0.
69.
21.
1.
0.
0.
0.
0.
n.
0.
0.
550
030
600
800
360
060
070
150
150
060
070
***FRACTION***
DIFFER- CUMULA-
ENTIAL TIVE
0.0059
0.7346
0.2299
0.0192
0.0038
0.0006
0.0007
0.0016
0.0016
0.0006
0.0007
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
9993
9934
2538
0289
0098
0060
0053
0046
0030
0014
0007
0000
f/D
0.0246
4.0585
1.9480
0.2294
0.0672
0.0152
0.0253
0.0ฐ12
0.1835
0.1252
0.2069
f/ln
0.
**
0.
o
'
o.
o.
0.
^-
0.
o.
o
*-
0.
01 fto
^ A o 7
fifiAl
u o < j
0 5 *5 A
0111
A A 1
001 A
v ' A O
OO91
'* \J ฃ A
OO91
V. 1 ^ J
flm o
"'J A o
002?
58
-------
GREEK 3
OPERATOR NAME ;
ROTOMETER SERIAL NUMBER
ROTOMETER SETTING
ROTOMETER PRESSURE
STONE WEIGHT BEFORE (GRAMS)
STONE WEIGHT AFTER (GRAMS)
STONE DEPTH (cm)
ATTRITION CELL USED
ATTRITION CELL PRESURE
GAS COMPOSITION
ART FELLERS
7311A3219A1
20
13
2000.0
1672.1
18
10cm 3 HLOE JET
ATM.
N2
37.34% CD
24.25% CCt
COMMENT : PRE-HEAT CELL 850C LOAD 2000gm STONE INTO CELL RE-HEAT
CELL 750C FLOW N2 FOR ONE HOUR DECREPITATION
FILTER ON EXHAUST LINE NO WEIGHT CHANGE
************ BEFORE *************
SUM Fi/Dpi- 6.7838
I/SUM - 0.1474
SPECIFIC SURFACE- 40.7030
MESH MICRONS MEAN D2 MASS
In GRAMS
Dl
8
12
16
24
32
42
60
115
250
325
PAN
(3366)
2380
1397
991
701
495
351
246
124
61
43
(30)
2830
1810
1180
835
570
420
294
175
87
51
36
.346
.533
.346
.346
.346
.346
.355
.685
.709
.349
.346
0.005
0.731
0.230
0.022
0.006
0.000
0.000
0.001
0.001
0.000
0.000
***FRACTION***
DIFFER- CUMULA-
ENTIAL TIVE
0.0051
0.7316
0.2303
0.0229
0.0061
0.0005
0.0005
0.0011
0.0014
0.0005
0.0001
1.0001
0.9950
0.2634
0.0331
0.0102
0.0041
0.0036
0.0032
0.0021
0.0007
0.0001
0.0000
f/D.
0.0180
4.0420
1.9520
0.2743
0.1070
0.0110
0.0156
0.0629
0.1609
0.1039
0.0361
f/ln
0.0147
1.3726
0.6657
0.0662
0.0176
0.0013
0.0013
0.0016
0.0020
0.0015
0.0004
************* AFTER **************
SUM Fl/Dpi- 6.7484
I/SUM - 0.1481
SPECIFIC SURFACE- 40.4906
MESH
8
12
16
24
32
42
60
115
250
325
PAN
MICRONS MEAN
(3366)
2380
1397
991
701
495
351
246
124
61
43
(30)
2830
1810
1180
835
570
420
294
175
87
51
36
D2
ln~
Dl
.346
.533
.346
.346
.346
.346
.355
.685
.709
.349
.346
MASS ***FRACTION***
GRAMS DIFFER- CUMULA-
ENTIAL TIVE
0.
36.
10.
1.
0.
0.
0.
0.
0.
0.
0.
730
740
560
080
300
040
040
050
070
030
010
0.0147
0.7398
0.2126
0.0217
0.0060
0.0008
0.0008
0.0010
0.0014
0.0006
0.0002
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
9998
9851
2453
0326
0109
0048
0040
0032
0022
0008
0002
0000
0.
4.
1.
0.
0.
0.
0.
0.
0.
0.
0.
f/D
0519
0875
8021
2605
1060
0192
0274
0575
1620
1185
0559
f/ln
0.0425
1.3881
0.6146
0.0629
0.0175
0.0023
0.0023
0.0015
0.0020
0.0017
0.0006
59
-------
GROVE 1
OPERATOR NAME ;
IIOTOMETER SERIAL NUMBER
ROTOMETER SETTING
ROTOMETER PRESSURE
STONE WEIGHT BEFORE (GRAMS)
STONE WEIGHT AFTER (GRAMS)
STONE DEPTH (cm)
ATTRITION CELL USED
ATTRITION CELL PRESURE
GAS COMPOSITION
COMMENT : PRE-HEAT CELL
RE-HEAT CELL 815C FLOW N2 FOR
ART FELLERS
7311A3219A1
12
11
2000.0
5
43.97% CO,
28.07% CCv
3 HOLE JET
1524,
18
10cm
ATM.
N2
900C LOAD 2000gn STONE INTO
ONE HOUR ONLY MADE 765C
CELL
FILTER INSTALLED ON EXHAUST LINE NO WEIGHT CHANGE RECORDED
TO DECREPIATI01!
************ BEFORE
SUM Fi/Dpi=
I/SUM
SPECIFIC SURFACED
MESH MICRONS
8
12
16
24
32
42
60
115
250
325
(3366)
2330
1397
991
701
495
351
246
124
61
43
*************
SUM Fi/Dpi=ป
AFTER
I/SUM
SPECIFIC
SURFACE-
MESH MICRONS MEAN
12
16
24
32
42
60
115
250
325
PAN
(3366)
2330
1397
991
701
495
351
246
124
61
43
(30)
A************
9.3947
0.1064
56.3685
MEAN D2 MASS
In GRAMS
Dl
2830
1810
1180
835
570
420
294
175
87
51
.346
.533
.346
.346
.346
.346
.355
.685
.709
.349
0
0
0
0
0
0
0
0
0
6
************
MEAN
2380
1810
1180
835
570
420
294
175
87
51
36
3.
0.
53.
D2
In
Dl
.346
.533
.346
.346
.346
.346
.355
.685
.709
.349
.346
9431
1118
6587
.019
.425
.202
.206
.140
.002
.000
.000
.001
.000
**
MASS
GRAMS
o
43
20
13
11
0
0
0
0
0
0
.370
.3<)0
.630
.010
.070
.320
.090
.OHO
.050
.010
.010
***FRACTION***
DIFFER- CUMULA-
ENTIAL TIVE
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0197
4257
2028
2066
1401
0027
0004
0004
0013
0003
1.
n.
0.
0.
n.
0.
n.
0.
0.
0.
0.
ooon
9803
5546
3513
1452
0051
0024
0020
0016
0003
0000
0.
2.
1.
">..
I.
0.
0.
0.
0.
0.
***FRACTION***
DIFFER-
ENT
0.
0.
0.
0.
0.
0.
o.
0.
0.
0.
0.
IAL
0245
4543
2141
1364
1146
0033
0009
0009
0005
0001
0001
cu
MULA-
f/D
0828
3521
7187
4744
4580
0643
0122
0206
1529
0588
f/D
f/ln
0.
0.
0.
o.
0.
0.
o.
o.
0.
o.
f
0560
7087
5862
5971
4040
0078
0010
onos
0019
OQO?
/In
TIVE
0.
0.
o.
o.
0.
0.
0.
0.
0.
0.
n.
0.
9909
9754
5210
3069
1205
0050
0026
0017
0007
0002
0001
0000
0.
2.
1.
2.
2.
0.
0.
0.
0.
0.
0.
1031
5102
8142
2328
0105
0789
0317
0532
0595
0203
0288
0.
0.
0.
0.
0.
0.
0.
n.
n.
0.
0.
070ฐ
v* * w ^
3524
6187
538S
3312
*+ -J .t t
0096
002^
OO14
0007
0003
*-* T v J
0003
60
-------
GROVE 2
ART FELLERS
/311A3219A1
12
11
2000.0
15.13.8
13
10cm
ATM.
43.97% CO
25.84% CO
3 HOLE JET
OPERATOR NAME ;
ROTOMETER SERIAL NUMBER
ROTOMETER SETTING
ROTOMETER PRESSURE
STONE WEIGHT BEFORE (GRAMS)
STONE WEIGHT AFTER (GRAMS)
STONE DEPTH (cm)
ATTRITION CELL USED
ATTRITION CELL PRESTIRE
GAS COMPOSITION N2
COMMENT : PRE-HEAT CELL 900C LOAD 2000gm STONE INTO CELL RE-HEAT
CELL TO 815C FLOW N 2 FOR ONE HOUR NO DECREPITATION
FILTER ON EXHAUST LINE NO CHANGE
************ BEFORE *************
SUM Fi/Dpi- 9.3316
I/SUM - 0.1065
SPECIFIC SURFACED 56.2896
MESH
8
12
16
24
32
42
60
115
250
325
MICRONS MEAN
(3366)
2380
1397
991
701
495
351
246
124
61
43
2830
1810
1180
835
570
420
294
175
87
51
T)2
In
Dl
.346
.533
.346
.346
.346
.346
.355
.685
.709
.349
MASS ***FRACTION***
GRAMS DIFFER- CTJMULA-
ENTIAL TIVE
0.019
0.425
0.202
0.206
0.140
0.002
0.000
0.000
0.001
0.000
0.0197
0.4257
n.2028
0.2066
0.1401
0.0027
0.0004
0.0004
0.0013
0.0003
1.0000
0.9803
0.5546
0.3518
0.1452
0.0051
0.00.14
0.0020
0.0016
0.0003
0.0000
f/D
0.0696
2.3521
1.7187
2.4744
2.4580
0.0643
0.0122
0.0206
0.1529
0.0588
f /In
0.056ฐ
0.7987
0.5862
0.5971
0.4049
0.0078
0.0010
0.0005
0.0019
0.0009
************* AFTER
SUM Fi/Dpi=
I/SUM
SPECIFIC SURFACE-
**************
8.7928
0.1137
52.7570
MESH MICRONS MEAN
D2
Dl
MASS
GRAMS
f/D
8
12
16
24
32
42
60
115
250
325
PAN
(3366)
2830
1397
991
701
495
351
246
124
61
43
(30)
2830
1810
1180
835
570
420
294
175
37
51
36
.346
.533
.346
.346
.346
.346
.355
.685
.709
.349
.346
2.810
44.070
19.570
17.120
9.810
0.320
0.110
0.100
0.050
0.020
0.010
f/ln
***FRACTION***
DIFFER- CUMULA-
ENTIAL TIVE
0.9999
0.0299 0.9700
0.4685 0.5015
0.2080 0.2935
0.1820 0.1115
0.1050 0.0065
0.0034 0.0031
0.0012 0.0019
0.0011 O.OOOQ
0.0005 0.0003
0.0002 0.0001
0.0001 0.0000 0.0295 0.0003
0.1056
2.5883
1.7630
2.1795
1.8426
0.0810
0.0398
0.0607
0.0611
0.0417
0.0863
0.8790
0.6013
0.5260
0.3035
0.0098
0.0033
0.0016
0.0007
0.0006
61
-------
GROVE 3
OPERATOR NAME ;
ROTOMETER SERIAL NUMBER
ROTOMETER SETTING
ROTOMETER PRESSURE
STONE WEIGHT BEFORE (GRAMS)
STONE WEIGHT AFTER (GRAMS)
STONE DEPTH (era)
ATTRITION CELL USED
ATTRITION CELL PRSSURE
GAS COMPOSITION
ART FELLERS
73.11A3219A1
12
11
2000.0
1618.4
18
10cm 3 HOLE JET
ATM.
N2
43.97% CO,
30.28% CO-
COMMENT : PRE-HEAT CELL 900C LOAD 2000gra STONE INTO CELL RE-HEAT
CELL TO 815C FLOW N2 FOR ONE HOUR :*0 DECREPITATION FILTER ON EXH
BEFORE *************
SUM Fi/Dpl= 9.3947
I/SUM = 0.1064
SPECIFIC SURFACE= 56.3685
MESH
8
12
16
24
32
42
60
115
250
325
MICRONS MEAN
(3366)
2380
1397
991
701
495
351
246
124
61
43
2830
1810
1180
835
570
420
294
175
87
51
D2
In
Dl
.346
.533
.346
.346
.346
.346
.355
.685
.709
.349
MASS
GRAMS
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
019
425
202
206
140
002
000
000
001
000
***FRACTION***
DIFFER- CUMULA-
ENTIAL TIVE
0.0197
0.4257
0.2028
0.2066
0.1401
0.0027
0.0004
0.0004
0.0013
0.0003
1.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0000
9803
5546
3518
1452
0051
00?. 4
0020
0016
0003
0000
0,
2.
1.
2,
2,
0.
0.
0,
0.
0.
f/D
,0823
,3521
,7187
,4744
,4580
.0643
,0122
,0206
.1529
,0588
f/ln
0.0569
0.7987
0.5862
0.5971
0.4049
0.0078
0.0010
0.0005
0.0019
0.0009
************* AFTER
SUM Fi/Dpi-
I/SUM
SPECIFIC SURFACE-
**************
8.9284
0.1120
53.5704
MESH
8
12
16
24
32
42
60
115
250
325
PAN
MICRONS
(3366)
2330
1397
991
701
495
351
246
124
61
43
(30)
MEAN
2380
1810
1180
835
570
420
294
175
87
51
36
D2
In
Dl
.346
.533
.346
.346
.346
.346
.355
.685
.709
.349
.346
MASS
GRAMS
1.510
22.4.50
19.8.30
9.340
5.860
0.170
0.040
0.050
0.0.30
0.010
0.010
***FRACTION*** f/D f/ln
DIFFER- CUMULA-
ENTIAL TIVE
0.9998
0.0255 0.9744 0.1070
0.3785 0.5959 2.0913
0.3343 0.2615 2.8334
0.1575 0.1040 1.8860
0.0988 0.0052 1.7334
0.0029 0.0024 0.0682
0.0007 0.0017 0.0229
0.0008 0.0008 0.0482
0.0005 0.0003 0.0581
0.0002 0/0002 0.0331
0.0002 0.0000 0.0468 0.0005
0.0736
0.7102
9663
4551
2856
0083
0019
0012
0007
0005
62
-------
PFIZER 1
ATTRITION WITH 3-HOLE JET IN 10-CM B*O
UNCALCINED -18+32 PFIZER SOR3ENT
OPERATOR NAME:
ROTOMETER SETTING:
ROTOMETER PRESSURE 5
STONE WEIGHT BEFORE {GRAMS)
STONE WEIGHT AFTER (GRAMS):
STONE DEPTH (CM):
ATTRITION CALL USED:
ATTRITION CALL PRESSURE:
GAS COMPOSITION:
COMMENT: REPESENTATIVE
ART FELLERS
20*
13.
2000*0
777.2
1 j.O
CM 3HO
ATM
N2
SAMPLE BEFORE
(HOLE JET
AND AFTER
PREHEAT CELL 860C POUR 200UG STONE IN
REHEAT CELL 7&5C THEN FLOW N2 IHR
DECREPJATION
BEFORE
MESH MICRONS MEAN LN(02/Dl)
*ปปFRACTIONซปซ*
MASS DIFFER- CUMULA
GRAMS ENTIAL TIVE
- F/D
SURFACE MEAN EQUIVALENT PARTICLE SIZE, CM ป *1028
SPECIFIC SURFACE, CMปซ(-1ป 58.3661
F/LN
8
12
16
2ซ4
32
42
60
1 IS
250
325
PAN
(3337)
2360
1397
991
701
H95
351
2M6
12<4
61
13
30
2806
1815
1 176
833
589
ซU6
293
I7ซ4
86
51
35
3H6
.521
0*43
.3*46
.3*48
.3*4*4
.355
.685
.709
.350
.360
.000
1 1.920
27.M30
2*4.380
3.870
.020
.000
.000
.000
.000
.000
0000
1763
.ซ40S6
3605
.0572
.0003
.0000
.0000
.0000
.0000
0000
1 .0000
1 .0000
.8237
-------
Figure 5
PARTICLE SIZE FREQUENCY CURVES FOR ATTRITION
SCREENING TREATMENT OF SORBENTS
64
-------
10'
10'
C] BF.F-ORE
CD PFTER.REPLiCP.TE 1
& PFTER.REPL1CPTE 2
-f aFTE.R.RF.FLiCPTE 3
10'
2 5 ID1 2 5 10 2
FflRTiCLF D1PMETFR, MICROMETERS
Particle-Size Distribution of
Greer Limestone. Attrition
Testing in 10-cm Bed with 3-Hole
Grid. Ordinate is Logarithmic
Scale
3 BEF 8RF.
O PfTF.R.RFFL iCPTE 1
A PFTER.RE^LiCPTE 2
-t- flFTEK.REPLiCPTF 3
-*-
5 \a! 2 5 10' 2
PflRTlCLE DIPMETF.R. MICROMETERS
Particle-Size Distribution of
Greer Limestone. Attrition
Testing in 10-cm Bed with 3-Hole
Grid. Ordinate is Linear Scale
!0
a er.f ORE
O PFTER.REPLiCPTE 1
A PFTER .Rf.PL iCPTE 2
+ SFTER .REPLICPTE 3
10' 2 & 10' 2 5 101 2 5
Op. PflRTlCLE D1PMETER. MICROMETERS
Particle-Size Distribution of
Greer Limestone. Attrition
Testing in 10-cm Bed with 3-Hole
Grid. Ordinate is Probability
Scale
10'
65
-------
o I
' I
CD BEF BRE
O PFTER .REFL iCPTE !
A PFTF.R .REFL iCPTE 2
+ RFTER.REFL iCPTf. ")
2 "5 1 CT ~2 5 " 1 0' 2 ' 5
PRRT1CLE DIPMETER. MICROMETERS
a BEFORE
O PETER .REPl1CRTE 1
A PFTER.REFLICPTE 2
+ PFTER.REFLiCRTE 3
10' 2 ~ 5 ]0! 2 5 10 2
FRRTiCLE DIPHETER. 1ICROMETF.RS
!0
Particle-Size Distribution of
Grove Limestone. Attrition
Testing in 10-cm Bed with 3-Hole
Grid. Ordinate is Logarithmic
Scale
Particle-Size Distribution of
Grove Limestone. Attrition
Testing in 10-cm Bed with 3-Hole
Grid. Ordinate is Linear Scale
a BEFORE
O PFTER.REPLICRTE 1
A PFTER.REPLICRTE 2
+ RFTF.R .REFLICRTE 3
10' 2 5 102 2 5 10' 2 5 "u
Dp. PRRT1CLE OlflMETER, MICROMETERS
Particle-Size Distribution of
Grove Limestone. Attrition
Testing in 10-cm Bed with 3-Hole
Grid. Ordinate is Probability
Scale
66
-------
rj B( r RF
O PFUR.RFrL ICflTf.
f, P r r f. R . R F c L i C ? T F-
+ PFTt R ,RFrL iCSTF
10' 2 5 10' _
f. DiPMF.TF_R. MICROMETERS
Particle-Size Distribution of
งrownwood Limestone. Attrition
Testing in 10-cm Bed with 3-Hole
Grid. Ordinate is Logarithmic
Scale
10'
E BF.f iRt
CD PFTF.K .RFPLiCPTE 1
i PFTER.RF.PLiCflTE Z
+ HFTCR.RFFLiCSIF 3
') 10' 2 rj 10 2 S
PRRTiCtE DIP1F.TF.R. niCROMFTF.RS
Particle-Size Distribution of
Brownwood Limestone. Attrition
Testing in 10-cm Bed with 3-Hole
Grid. Ordinate is Linear Scale
in'
BF.f CRF.
O PFTER.REPUCPTF !
& PF1ER .RF.FLiCPTF. 2
+ 3FTER.RF.FL PTF 3
10'
2 5 iff 2 5
Dp. PPRT1CLE DIP^FTER.
Particle-Size Distribution of
Brownwood Limestone. Attrition
Testing in 10-cm Bed with 3-Hole
Grid. Ordinate is Probability
Scale
67
-------
10
2 5 10'' 2 5 10! 2 5
PRRT1CLF DIP1F.TF.R. MICROMETERS
!04
Attrition with 3-Hole Jet in 10-cm
B Uncalcined -18 +32 Pfizer Sorbent.
Ordinate is Logarithmic Scale
5 10' 2 5 10' 2 5
PPRT1CIE DiPMF.TER. MICROME1F.RS
10
Attrition with 3-Hole Jet in 10-cm
B Uncalcined -18 +32 Pfizer Sorbent.
Ordinate is Linear Scale
68
-------
First, the increases in specific surface, listed in Table 14, are
compared by completely randomized analysis of variance. Testing at the
5-percent level shows significant differences among all three means and
between the Greer and Grove means.
Second, in interpreting these data, the increase in mass fraction
of particles smaller than 495 ym are listed and analyzed in Table 15.
Again, testing at the 5-percent level shows significant differences
among all three means and between the Greer and Grove means.
The third approach is comparing effects of both sieve size and sor-
bent on frequency. We cannot compare all sieves because the frequen-
cies are not independent but choose those sieves smaller than 701 ym.
Table 14
SPECIFIC SURFACE INCREASE DATA AND ANALYSIS
OF VARIANCE OF DIFFERENCES
Repli-
cate
1
2
3
Sorbent
1.
Before
46.50
46.50
46.44
Brownwood
After
80.38
81.55
83.76
Analysis of Variance
Source
Column Means
Within Cols.
To
Indivi
F =
Diff.
33.88
35.05
37.32
2.
Before
40.72
40.72
40.70
on Differences
Sum Sq
2620.6
17.8
tal 2638.4
dual DF Test on Greer & Grove
4.01 + 1.13 - 0.21 + 2.71 + 3
3(1
Greer
After
44.73
41.85
40.49
DF
2
6
Diff.
4.01
1.13
0.21
3
Before
56.37
56.37
56.37
Mean Sq
Grove
After
53.66
52.76
53.57
Diff.
-2.71
-3.53
-2.79
F-Ratio
1310.3 442
2.96
8
Effects:
.53 + 2.79 _ ,
2 + (-1)2)2.96
0.97 with 1 & 8
.6
df.
69
-------
Table 15
INCREASES IN SORBENT FRACTIONS <495 pm AND ANALYSIS
OF VARIANCE OF DIFFERENCES
Repli-
cate
1
2
3
Sorbent
1 . Brownwood
Before
0.0014
0.0014
0.0014
After
0.0524
0.0524
0.0575
Analysis of Variance
Source
Column Means
Within Columns
Total
Diff.
0.051
0.051
0.056
2.
Before
0.0041
0.0041
0.0041
on Diferences:
Sum Squares
0.0000384
0.00519
0.00519
Individual DF Test on Greer
,., _ 0.0068 + 0.0019 + 0.0007
3d2
+ (-1)'
Greer
After
0.0109
0.0060
0.0068
DF
2
6
8
Diff.
0.0068
0.0019
0.0007
3
Before
0.0051
0.0051
0.0051
Mean Square
0.00258
0.00000640
and Grove Effects:
- 0.0008 - 0.0014 - 0.0001
2 ) (0.00000640)
Grove
After
0.0059
0.0065
0.0052
Diff.
0.0008
0.0014
0.0001
F-Ratio
402
56259
184
with 1 & 8 df
Table 16 lists frequency-increase data of a two-way analysis of
variance. The column (sorbent) means give rise to an F-ratio of 3.56
which signals a significant difference (tabulated F = 3.2 with 2 and
42 df) at the 5-percent level.
Inspection of any of three different aspects of these data show
that there are indeed significant differences in degree of attrition
among the sorbents tested.
70
-------
Table 16
FREQUENCY INCREASE DATA FOR SIEVES SMALLER THAN 701 ym AND
TWO-WAY (randomized block) ANALYSIS OF VARIANCE
Sieve
Size
Range, wn
495-701
351-495
246-351
124-246
61-124
43-61
<43
Values of Frequency Increase
Sorbent
1 . Brownwood
0.0213
0.0488
0.0053
0.0098
0.0116
0.0100
0.0089
0.0091
0.0083
0.0151
0.0149
0.0163
0.0332
0.0308
0.0357
0.0246
0.0250
0.0285
0.0056
0.0087
0.0096
2. Greer
0.0001
0.0065
0.0108
0.0010
0.0005
0.0055
0.0010
0.0008
0.0038
0.0001
0.0005
0.0020
0.0000
0.0003
0.0017
0.0002
0.0003
0.0013
0.0002
0.0018
0.0016
3 . Grove
0.0737
0.1014
0.1193
0.0018
0.0020
0.0005 ,
0.0016
0.0023
0.0009
0.0009
0.0011
0.0017
0.0008
0.0008
0.0012
0.0006
0.0003
0.0004
0.0003
0.0003
0.0005
71
-------
Table 16 evidences some consistent negative frequency changes with
the Greer and Grove sorbents. These values are small and probably
attributable to random chance as loss by attrition rather than to gain
by attrition. The Grove sorbent curve (Figure 5) consistently shows a
net loss of particle mass in the 400 to 800 urn size range. This size
appears to attrite preferentially, perhaps because of grain structure.
DISCUSSION
Our purpose here has been to develop a screening test that embodies
the attrition mechanisms active in the CAFB.
The procedure tested involves the principal attrition sources.
There is probably a "reasonable" range of effects for each attrition
cause. We have attempted to duplicate this, but the balance is imper-
fect. For example, the sources of attrition in a given system may be
almost entirely grid jet effects; in the test described here, thermal
shock is a prime contributor to attrition.
It is premature to specify a standard piece of equipment for attri-
tion testing. It is practicable, however, to specify a procedure, des-
cribe the apparatus, and recommend a "good enough" reference or standard
sorbent against which others may be compared.
For attrition testing of candidate sorbents for CAFB we suggest use
of an attrition test cell congruent with that described here and the
same test procedure. Either Grove 1359 limestone or Greer limestone is
suggested as an adequate reference because they are comparable in
attrition resistance, and both are adequate in attrition and sorption
performance.
72
-------
7. PARTICULATE CONTROL
The control of particulate emissions from the CAFB is a critical
area since the process operability and environmental acceptability
depend on the control success. Particulate control requirements and
control options are discussed on the basis of parametric projections.
CONTROL REQUIREMENTS
The CAFB particulate control requirements have been considered par-
ametrically for two cases: gasification of liquid fuels (residual fuel
oils and bitumen) and gasification of lignite. The gasifier is the
major source of particulate emissions, with the regenerator, the spent
sorbent processing system, the sulfur recovery system, and the sorbent
(and lignite) handling systems of secondary importance.
The control of particulate from the gasifier must meet three gen-
eral requirements: minimization of coarse sorbent particle losses; pro-
tection of the fuel gas piping, the burner, and the boiler from erosion
and deposits; and environmental particulate emission standards of 4.3
x 10~5 kg/GJ (0.1 lb/106 Btu). In general, particulate control will be
required before and after the boiler in order to meet these require-
ments.
The following two figures represent the particulate control effi-
ciency requirements before and after the boiler as a function of the
sorbent elutriation rate for commercial CAFB installations. The sor-
bent elutriation rate is expressed as a fraction of the fresh sorbent
feed rate.
73
-------
Figure 6 considers two fuels, residual fuel oil and bitumen with
2.6 wt % sulfur and 3.75 wt % sulfur, respectively. Two calcium-to-
sulfur ratios are shown - 0.5 and 1.0 - based on the expected range of
gasifier performance from pilot plant experience.
Figure 7 considers lignite having a 15 wt % ash content and 3.6 wt
% sulfur content. The fraction of the lignite ash elutriated from the
gasifier is a parameter in the figure having three values - 0.25, 0.5,
and 1.0.
The control requirements before the boiler are difficult to quan-
tify, but we expect that if control before the boiler is utilized the
control device should be of as high an efficiency as can be tolerated in
terms of the fuel gas pressure drop and system operability, since the
capital investment will probably not be too sensitive to collection
efficiency for a given control technique. A very high elutriation rate
(>100% of the fresh sorbent feed rate) probably would result from a high
carry-over of coarse material which should be captured to the greatest
extent possible. A small elutriation rate (<10% of the fresh sorbent
feed rate) probably would result from attrition and carry-over of fine
material which should be removed from the system (for spent sorbent
processing) rather than recycled to the gasifier or regenerator.
The lignite ash should be removed from the system without recycle
to the greatest extent possible in order to avoid ash agglomeration
problems and high ash carry-over rates. Ash separation from coarse
sorbent particles may be possible, depending on the nature of the ash
(size, density, shape). Multiple stages of particulate control may be
required.
CONTROL OPTIONS
The control technology available for the CAFB is as follows:
cyclones, granular-bed filters, conventional filter systems, scrubbing
systems, electrostatic precipitators. Any of these could be used after
the boiler, while only the cyclones or granular-bed filters will be con-
sidered for the hot low-heating-value gas cleaning before the boiler.
-------
Curve 686569-A
1.0
0.1
~ 0.01
r~
0.001
0.01
With
(41
Residual Fuel Oil With 2.6 Wt % Sulfur
and 19,000 Btu/lb (44,194k.!/Kg) Heating
Value
Bitumen
3.75 Wt* Sulfur and 17,900
635 kJ/Kg) Heating Value
Overall Particle Collection Efficiency
of Particulate Control Equipment
Before Boiler (1) and Following
Boiler (2)
Ca/S=0.5
Ca/S = 1.0
0.1 LO 10.0
Sorbent Elutriation Rate as Fraction of Fresh Sorbent Feed Rate
Figure 6. Oil-Fueled CAFB Particulate Control Requirements
Curve 68656B-*
0.1
f
0.01
'5
S"
>
J. 0.001
Fraction of
Ash Elutriated
Lignite With 3.6 Wt% Sulfur. 15 Wt % Ash and
11,000 Btu/lb (25,586 kj/Kg) Heating Value
Ca/S = 1.0
- Ca/S =0.5
0.1 LO 10.0
Sorbent Elutriation Rate as Fraction of Fresh Sorbent Feed Rate
Figure 7. Coal-Fueled CAFB Particulate Control Requirements
75
-------
Because the nature of the low-heating-value gas is to form deposits
and cause plugging and because of the possible need for sorbent recycle
from the collection device, we recommend that conventional cyclones be
used before the boiler. High efficiency from these cyclones can not be
expected because of deposit formation. We have selected an upper limit
of 90% overall efficiency for sorbent material as a basis. As in Fig-
ure 6 for the gasification of residual fuel oil or bitumen and if we
assume a cyclone collection efficiency of 90 percent, no particulate
control would be required after the boiler if less than 15 percent of
the fresh sorbent rate were elutriated from the gasifier in the case of
bitumen and 25 percent in the case of residual fuel oil.
As in Figure 7 for the gasification of lignite, we expect that
either multiple cyclones will be required before the boiler or some form
of particulate control will be required after the boiler under all reas-
onable conditions of elutriation. Depending on the nature of the lig-
nite ash, either a conventional cyclone or an electrostatic precipitator
would be recommended after the boiler.
ASSESSMENT
Table 17 summarizes recommendations and limitations of the recom-
mendations. On the basis of pilot plant elutriation results, we expect
that liquid fuel gasification will require particulate control by con-
ventional cyclones before and after the boiler. We project lignite gas-
ification to require a conventional cyclone before the boiler and an
electrostatic precipitator for final control. These conclusions will be
valid over a broad range of performance.
Additional alternatives that could be applied to reduce particle
elutriation from the gasifier are: reduced fluidization velocity, shal-
low bed operation, increased freeboard height or baffles in the free-
board, improved distributor plate design, limited recycle of fines from
the particulate control system to the gasifier, improved sorbent feeding
76
-------
Table 17
CAFB PARTICULATE CONTROL REQUIREMENTS
Cases
No Control
Required3
Cyclone before
Boilerb
Series Cyclones
before Boiler or
Residual
Fuel Oil Bitv
<2.5 <1.
>2.5 >1.
<25 <1!
Lignite,
% Ash Elutriated
Linen 2 5
5 Always
required
5 Not
> sufficient
>25 >15 >0
<250 <150 <65
100
Always
required
Not
sufficient
Not
sufficient
before and after
Boilerc
Cyclone before
Boiler and ESP
after Boiler^
>250
>150
>65
>0
aAssumes gas lines, burners, boiler unaffected by erosion, deposits.
"Maximum overall cyclone efficiency 90%; dependent upon size distribu-
tion, deposit formation, pressure drop limitations.
cAssumes maximum efficiency of 2 cyclones in series of 99%.
"Assumes maximum cyclone efficiency of 90% and electrostatic precip-
itator (ESP) efficiency of 99%. Very sensitive to lignite ash
characteristics.
77
-------
method to avoid fast heating and calcination, sorbent selection based on
attrition resistance, sorbent preparation by sizing or prehardening, and
lignite sizing and washing.
There are no development requirements for cyclone or electrostatic
precipitator particulate control except to demonstrate long-term reli-
ability and performance of the cyclone with hot, low-heating-value gas
particulate control. The alternatives listed to reduce particle elutri-
ation would require design evaluation and/or development work before any
of them could be implemented.
Data gaps exist in the areas of sorbent attrition and elutriation
behavior, lignite ash characteristics and elutriation behavior; commer-
cial cyclone performance in the CAFB low-heating-value gas environment;
and erosion and deposit effects in the fuel gas line, burner, and
boiler. The availability of such data would permit improved projections
of particulate control requirements but would probably not change the
general conclusions developed.
78
-------
8. ASSESSMENT
PROCESS ECONOMICS
There are many options to consider in assessing CAFB Process
Economics:
New vs retrofit
Once-through vs regenerable
CAFB vs stack scrubbing
CAFB vs hydrodesulfurization
CAFB vs gasification
CAFB vs Flexicoking(ฃM)
CAFB vs coal liquefaction
Some of these choices can be resolved relatively easily. It is
unlikely, for example, that a new CAFB-fueled boiler could be justified.
New boilers on feedstocks appropriate for CAFB should probably utilize
fluidized-bed combustion (FBC). Also, the federal policy of coal utili-
zation for new boilers dictates against CAFB, and, in general, the feed-
stock problems discussed in Section 4 of this report indicate little
likelihood of fueling new boilers with residue from oil refining or from
synfuel production. Further, the 1975 CAFB^ assessment indicates that
both once-through and regenerative stack-gas cleaning processes are
lower in investment and operating cost than a CAFB unit for 50, 200, and
500 MW power boilers, so these stack-scrubbing options are likely to be
used to meet the requirements placed on new units by the EPA New Source
Performance Standards (NSPS).32 The economic assessment prepared by GCA
Corporation for EPA in 1979 reaches the same conclusion in its compari-
son of 250 MW regenerable systems.^3 The GCA report indicates $260/kW
investment and $5.90/bbl of fuel oil feed operating cost for CAFB vs
$92/kW and $2.63/bbl for regenerable MgO flue gas desulfurization.
Thus, new regenerable CAFB is too costly.
79
-------
The GCA report also provides a basis for an ecoraonic assessment of
once-through vs regenerable options. The once-through CAFB system
costs, if we assume dry sulfation of the sulfided limestone, are $188/kW
and $4.38/bbl, which is less than 75 percent of the regenerable CAFB
system cost. This once-through CAFB system, however, is still substan-
tially more costly than the regenerable flue gas desulfurization system,
so a once-through system is also too costly for new CAFB installations.
It does appear that once-through CAFB has significantly better economics
than regenerable CAFB and should be the process to use in any comparison
with options such as hydrodesulfurization or Flexicoking. (SM) CAFB does
provide the potential to use a high heavy-metal-content residuum that
cannot be burned directly in conventional boilers because of boiler tube
corrosion/deposition problems. In such a case the flue gas desulfuriza-
tion option would not apply, and some fuel processing system would need
to be used.
The GCA report also addresses Flexicoking^^ ancj hydrodesulfuriza-
tion (LC-Fining)34 economics relative to once-through CAFB.
Once-through CAFB Flexicoking LC Fining
Investment, $/kW 188 107 95
Operation, $/bbl 4.38 3.45 3.84
In its analysis of these results, GCA concludes that "in order to
operate the CAFB on a competitive basis . . . high sulfur, high
metals crudes must be $2 to $3 per barrel cheaper" and "at present this
per barrel differential requirement is roughly twice the market situa-
tion. " As discussed in Section 4, the tight crude oil supply today, the
projection that this situation will be the norm for the 1980s, the
increasing need for hydrogen-^ in processing heavy crudes, the develop-
ing needs for transportation fuels in Third World Nations, and a reduc-
tion in U.S. oil imports as a matter of national security all indicate
that the $2 to 3/bbl price differential is unlikely.
80
-------
To supplement the GCA economic assessment we have prepared an
assessment of CAFB relative to some "synfuels" options such as oil and
coal gasification and coal liquefaction. The results of this assessment
reinforce the conclusion that refiners will utilize vacuum bottoms for
hydrogen production before using coal, so that every oil fraction will
be used before they resort to coal. In order to minimize confusion (the
GCA assessment was for a 200 MW unit in 1980; the Westinghouse assess-
ment was for a 200 MW unit in 1977) we have normalized our cost data to
the $5.90/bbl operating cost of GCA's regenerable CAFB and Resox Sys-
tem. Table 18 indicates relative costs to operate each of the systems.
It is obvious that:
Vacuum bottoms will be used as a hydrogen source in pref-
erence to coal (02 gasification of resid vs coal).
Vacuum bottoms will be used as a fuel gas source in pref-
erence to coal (air gasification of resid vs coal).
LC-Fining of vacuum bottoms for both desulfurizing and
deraetallizing34 is a potentially attractive route to
hydroprocessing for transportation fuel production.
Airblown gasification of resid is competitive with regen-
erable CAFB, so a Texaco partial oxidation system with
preheated air feed, which has been successfully operated
in a 1 MW pilot plant at Montebello, California, can pro-
vide clean fuel gas to a refinery distribution system,
which CAFB cannot do.
Our conclusion from the economic assessment summarized above is
that no definable market exists for CAFB.
POTENTIAL
The development program for CAFB was funded by EPA to investigate
the possibility of a boiler pollution control system functioning to
81
-------
Table 18
COST COMPARISON
$/bbl EQUIVALENT
Regenerable CAFB
Regenerable FGD
Nonregenerable FGD
LC-Finlng
Coal Liquefaction
02/Coal Gasification
Air/Coal Gasification
02/Resid Gasification
Air/Resid Gasification
GCA
5.90
2.63a
2.40
3.84
-
-
-
-
-
1 Westinghouse
5.90
4.70b
3.02
3.68
15.60
17.70
9.67
17.70
5.56
aMag-Ox
bWellman-Lord
clean the fuel and not the products of combustion.* The initial esti-
mates for the CAFB retrofit to a 600 MW power boiler indicated "as much
as 50% less" capital cost "than an add-on wet scrubbing system" for a
residual-oil-fired installation.
In 1975, after a thorough assessment of the data from CAFB tests at
the Esso-UK 750 kW unit in Abingdon, England, and after a 50 MW retrofit
design for a CAFB system had been prepared by Stone and Webster, Inc.,
for New England Electric Systems,-* it was apparent to us that "CAFB is
25 to 50 percent greater" in capital cost than "limestone scrubbing
stack gas cleaning costs." Also, we concluded in 1975 that a larger
hydrodesulfurization unit could treat vacuum bottoms with low metals
content competitively with CAFB. Thus, the feedstock appropriate for
CAFB was narrowed to high-heavy-metals-content residua because such
material could not be fired in a stack-gas-scrubber-equipped boiler
82
-------
because of corrosion/deposition problems and could not be hydrodesul-
furized because of catalyst deactivation by the heavy metals in the
residua.
Now, in 1979, CAFB continues to be evaluated as more expensive than
stack-gas scrubbing. Several relatively new hydrodesulfurization pro-
cesses, however, are reported to be capable of processing high heavy
metal residuum. One of these, LC-Fining, was concluded by GCA to have
lower processing costs for residuum cleanup than CAFB in their report to
EPA.* It, thus, appears that processing any residual oil for pollution
control can best be done by other than CAFB means. Also, the availabil-
ity of residual oils for CAFB feedstock is dissappearing since the oil
industry has both the ability and the urgent need to process all of
their by-product distillates into transportation fuels to
Reduce imports of oil
Provide hydrogen for hydroprocessing
Provide fuel gases for process equipment
Meet increasing worldwide demand for distillate fuels
Assure adequate domestic U. S. supplies of gasoline,
diesel fuel, jet fuel, and home heating oil.
The potential application of CAFB to solid fuels (lignite, tire
shreds, etc.) has yet to be satisfactorily demonstrated. In any case,
air-blown gasification with core gas desulfurization is indicated to be
competitive with CAFB and can supply clean, basically distributed fuel
gas (within utility site or refinery site battery limits) to a large
single boiler or to a multiplicity of units such as process heaters,
steam reformers, and hydrotreaters. We conclude that the only possible
market for CAFB may be a special situation where a suitable feedstock,
not directly combustible, and an existing gas-fired boiler of moderate
size, exist in the same size.
83
-------
ENVIRONMENTAL IMPACT
The major environmental concerns associated with the CAFB process
are SOX, nitrogen oxide (NOX), particulate control, solid waste dis-
posal, and resource utilization.
The ability to control SOX emissions has been partially demon-
strated by the ERCA CAFB pilot unit operated in the regenerative mode,
using liquid fuels and limestone sorbents. Integrated regenerative
operation with sulfur recovery has not been performed. Only limited SOX
emission data using solid fuels have been collected on the pilot unitป.
While the ability of the CAFB demonstration plant to control the SOX
emissions is uncertain (e.g., the sulfur capture efficiency of the RESOX
process and the ability to operate the demonstration plant regenera-
tively using solid fuels), it is likely that acceptable levels of SOX
control can be achieved with the CAFB process by selecting appropriate
design and operating conditions. Once-through sorbent operation -
rather than regenerative - may be required with some fuels.
Nitrogen oxide emissions from the CAFB process should be acceptable
if the proper burner design is selected. Previous CAFB pilot unit
operation has indicated low NOX emission levels.
In the CAFB process particulate emissions are more a problem of
process operability (i.e. , deposition and erosion) than they are of
environmental protection. Existing technology can reduce the particu-
late from a CAFB retrofited boiler to acceptable levels. If the gas
passes through high-temperature cyclones before entering the boiler,
particulate emissions will probably be only partially reduced and, in
order to satisfy environmental standards, an electrostatic precipitator
or baghouse will still be required after the gas exits from the boiler.
Westinghouse has investigated the environmental impact of the dis-
posal of unprocessed and processed CAFB solid waste extensively. On the
basis of laboratory testing results, we judged that the unprocessed CAFB
84
-------
spent sorbent would be environmentally unacceptable for direct land dis-
posal. Available test data, however, show that environmental accept-
ability can be achieved by further processing.
The major environmental concerns for direct disposal are heat
release, sulfide, pH, calcium, sulfate (804), and total dissolved solids
(IDS). The major environmental concerns about disposal after processing
are pH, calcium, 804, and TDS.
Results suggest that the disposal of processed CAFB solid waste may
cause environmental effects comparable to (due to its chemical proper-.
ties) or perhaps less negative than (due to its physical properties) the
disposal of the residue from the currently commercialized FGD process.
Several processing techniques for CAFB solid residues have been
identified, including both high-temperature and low-temperature
options.
Although on the basis of its leachate quality the high-temperature
processed compact appears to be environmentally superior to the other
alternatives, the energy requirements would have to be evaluated in
relation to the benefits. On the basis of environmental impact, dry
sulfation would be the recommended process, followed by dead-burning and
low-temperature fly ash blending.
As a subsystem, dry sulfation is the most expensive option, either
as a percentage of plant cost or relatively, but its ultimate cost
advantage results from elimination of a sulfur recovery plant. Back-up
options are direct disposal, which is attractive if a consumer is able
to utilize the material, and briquetting. The direct disposal option,
with utilization of the material in building block, for example, is an
option for the CAFB demonstation plant in San Benito, Texas.
The CAFB process provides some potential environmental benefits in
the area of resource utilization. Some low-quality fuels suitable for
consumption in the CAFB process are not easily utilized by conventional
85
-------
technology. Such fuels represent an energy resource that should be
utilized in the most effective manner, which in some cases may be gasi-
fication in the CAFB process.
86
-------
9. REFERENCES
1. Archer, D. H., D. L. Keairns, J. R. Hamm, R. A. Newby, W.-C. Yang,
L. M. Handraan, and L. Elikan, Evaluation of the Fluidized Bed Com-
bustion Process, Vols. I, II, and III. Report to EPA, Westinghouse
Research and Development Center, Pittsburgh, PA, November 1971, GAP
Contract 70-9, NTIS PB 211-494, 212-916, and 213-152.
2. Keairns, D. L., D. H. Archer, R. A. Newby, E. P. O'Neill,
E. J. Vidt, Evaluation of the Fluidized-Bed Combustion Process,
Vol. IV, Fluidized-Bed Oil Gasification/Desulfurization. Report to
EPA, Westinghouse Research and Development Center, Pittsburgh, PA,
December 1973, EPA-650/2-73-048d, NTIS PB 233-101.
3. Keairns, D. L., R. A. Newby, E. J. Vidt, E. P. O'Neill, C. H.
Peterson, C. C. Sun, C. D. Buscaglia, and D. H. Archer, Fluidized
Bed Combustion Process Evaluation - Residual Oil Gasification/
Desulfurization Demonstration at Atmospheric Pressure. Report to
EPA, Westinghouse Research and Development Center, Pittsburgh, PA,
March 1975, EPA-650/2-75-027 a&b, NTIS PB 241-834 and PB 241-835.
4. Keairns, D. L., C. H. Peterson, and C. C. Sun, Disposition of Spent
Calcium-Based Sorbents Used for Sulfur Removal in Fossil Fuel Gasi-
fication, presented at the Solid Waste Management Session, 69th
Annual Meeting, AIChE, November 28 - December 2, 1976, Westinghouse
Scientific Paper 76-9E3-FBGAS-P1.
5. Craig, J. W. T., et al. , Chemically Active Fluid Bed Process for
Sulfur Removal during Gasification of Heavy Fuel Oil - Second
Phase. Report to EPA, Esso Research Centre, Abingdon, UK, Novem-
ber 1974, EPA-650/2-74-109, NTIS PB 240-632/AS.
87
-------
6. Chemically Active Fluid Bed Process (CAFB). Monthly report to EPA,
Foster Wheeler Energy Corporation, Livingston, NJ, May 29 -
June 25, 1978, Contract 68-02-2106.
7. Chemically Active Fluid Bed Process (CAFB). Monthly report to EPA,
Foster-Wheeler Energy Corporation, Livingston, NJ, June 17 -
July 20, 1975, Contract 68-02-2106.
8. Chemical Week, 125(21): November 21, 1979; p. 48.
9. Saxton, A. L., et al., Assessment of the PEAB Process, Report to
EPA, Exxon Engineering Company, Linden, NJ, June 1977.
10. Craig, J. W. T., et al., Study of Chemically Active Fluid Bed Pro-
cess for Sulphur Removal during Gasification of Heavy Fuel Oil.
Interim report to EPA, Esso Petroleum Company, Abingdon, UK,
July 1972 to April 1973, Contract 68-02-0300.
11. AGA Monthly, 60(9): September 1978; p. 2.
12. AGA Monthly, 61(8): August 1979; p. 26.
13. Assessment of the Capability of Firing Clean Low-Btu Gases in
Existing Coal, Oil, and Gas-Fired Steam Generator. Final report to
EPRI, Combustion Engineering, Inc., Windsor, CN, December 1975,
EPRI 265-1.
14. Low-Btu Gas Study. Final report to EPRI, Babcock & Wilcox,
Alliance, OH, January 1976, EPRI 265-2.
15. Winning More From Heavy Oils, Chemical Engineering, Dec. 5, 1977;
p. 118.
16. Ball, D. A., et al. , Environmental Aspects of Retrofitting Two
Industries to Low- and Intermediate-Energy Gas from Coal. Report
to EPA, Battelle-Columbus Laboratories, Columbus, OH, April 1976,
EPA 600/2-76402.
88
-------
17. Energy Consumption In Manufacturing, Conference Board, Cambridge,
MA: Ballinger Publishing Co.; 1974.
18. Monk, J. A., et al. , Residuum and Residual Fuel Oil Supply and
Demand in the United States - 1973-1985. Report to EPA, Arthur D.
Little, Inc., Cambridge, MA, June 1976, EPA 600/2-76-166.
19. Johnes, G., Esso Research Centre, Abingdon, UK, private
conraunication.
20. O'Neill, E. P., D. L. Keairns, andM. A. Alvin, Sorbent Selection
for the CAFB Residual Oil Gasification Demonstration Plant. Report
to EPA, Westinghouse Research and Development Center, Pittsburgh,
PA, March 1977, EPA-600/7-77-029, NTIS PB 266-827.
21. Keairns, D. L., D. H. Archer, J. R. Hamm, S. A. Jansson,
B. W. Lancaster, E. P. O'Neill, C. H. Peterson, C. C. Sun,
E. F. Sverdrup, E. J. Vidt, andW.-C. Yang, Fluidized Bed Combus-
tion Process Evaluation, Phase II-Pressurized Fluidized Bed Coal
Combustion Development. Report to EPA, Westinghouse Research
Laboratories, Pittsburgh, PA, September 1975, EPA-650/2-75-027c,
NTIS PB 246-116.
22. Newby, R. A., N. H. Ulerich, E. P. O'Neill, P. F. Ciliberti, and
D. L. Keairns, Effect of S02 Emission Requirements on Fluidized-Bed
Combustion Systems: Preliminary Technical/Economic Assessment.
Report to EPA, Westinghouse Research and Development Center,
Pittsburgh, PA, August 1978, EPA-600/7-78-163.
23. Ulerich, N. H., E. P. O'Neill, and D. L. Keairns, The Influence of
Limestone Calcination on the Utilization of the Sulfur-Sorbent in
Atmospheric Pressure Fluid-Bed Combustors. Report to EPRI, West-
inghouse Research Laboratories, Pittsburgh, PA, August 1977, Pro-
ject 720-1, EPRI FP-426.
24. JANAF Thermocheraical Tables.
89
-------
25. National Bureau of Standards.
26. National Bureau of Standards.
27. Richardson, F. D., J. H. E. Jeffes, J. Iron Steel Inst., 171(167):
1952.
28. Lowell, P. S., andT. B. Parsons, Identification of Regenerable
Metal Oxide S02 Sorbents for Fluidized-Bed Coal Combustion. Report
to EPA, Radian Corporation, Austin, TX, July 1975, EPA 600/2-
75-065.
29. Reeve, L., Desulphurization of Coke-oven gas at Appleby-
Frodingham, Journal of the Inst. of Fuel, 319, July 1975.
30. Schultz, F. G., and P. S. Lewis, Hot Sulfur Removal from Producer
Gas, 3rd International Conference on Fluidized Bed Combustion,
Hueston Wood, OH, October 1972.
31. Abel, W. T., F. G. Schultz, and P. F. Langdon, Removal of Hydrogen
Sulfide for Hot Producer Gas by Solid Absorbents, U.S. Bureau of
Mines Report 7947, 1974.
32. Environmental Protection Agency, 40 CFR Part 60 (FRL 1240-7),
June 11, 1979.
33. Preliminary Environmental Assessment of the Lignite-Fired CAFB,
EPA-000/7-79-048, Feb. 1979, Pgs B-8 & B-9.
34. Van Driesen, R. P., J. Gaspers, A. R. Campbell, and G. Lunin,
LC-Fining Upgrades Heavy Crudes, Hydrocarbon Processing, 58(5):
May 1979; p. 107.
35. Cornelisse, C. L. E., H. J. Madsack, and Z. Supp, Gasify Residuum
for Plant Utilities, Hydrocarbon Processings, 58(7): July 1979;
p. 126.
90
-------
APPENDIX A
ATTRITION IN THE BUBBLING ZONE OF A FLUIDIZED BED
All of the studies relating to attrition in the bubbling zone are
described in Appendix A.
ATTRITION IN THE GRID ZONE OF A FLUIDIZED BED
Attrition Mechanisms
The frequently considered source of attrition in a fluidized bed is
the obvious grinding and shattering collisions of particles. There are
several causes of particle wear, which include the following:
Abrasion
In this process defects, edges, and corners are knocked from par-
ticles by low-energy collisions. Abrasion can occur during passage of a
gas bubble through the bed of solids.
High-Energy Collisions
Particles may be accelerated to high velocity; for example, when
entrained in a jet at the distribution plate, the high-velocity particle
can strike another particle or vessel wall and shatter into relatively
large fragements.
Blinichev, Strel'tsov, and Lebedeva^ have distinguished two zones
in a fluidized bed - the lower, which they call the "nozzle" effect
zone, in which gas jets accelerate large particles to energies suffici-
ent for shattering; and the upper zone, characterized by intensive mix-
ing and low-energy impacts that grind particle surfaces.
91
-------
Thermal Shock
When cold particles are added suddenly to a bed of red-hot solids
there is severe thermal stress on the cold particles. One expects spal-
ling of the particle surface and perhaps shattering of the entire parti-
cle into large fragments.^
Chemical Stress
Sorbent particles calcine, then react with sulfur dioxide (SC>2);
calcium oxide (CaO) forms calcium sulfite (3803), with subsequent
changes in the lattice structure. This change in particle structure at
its surface hardens particles in some cases, or in other cases causes
internal stresses leading to spalling or weakened particle
surfaces.A2~A4
Internal Gas Pressure
When cold limestone or dolomite makeup sorbent is added to a hot
flviidlzed bed, the resulting calcination generates carbon dioxide (C02)
within the particle. Esso Research Centre in Abingdon, UK (ERCA) found
that a slower calcination rate of fresh limestone results in lower pro-
duction of fines.^5 Similarly, water within particle cracks will flash
when heated to bed temperatures. While CC>2 pressures are moderate
(100.0 kPa equilibrium at 900ฐC), steam pressures are high and can
explode particles.
Transfer Lines and Cyclones
These are not a part of the fluidization process but are generally
included in a fluidized-bed system. Sorbent breakage rate is related to
the circulation rate of the solids and is controlled by equipment design
effects on solids impact.
Kutyavina and Baskakov explain, "With fluidization, particles are
ground by abrasion and splitting. .. Abrasion is evidently predominant
even for brittle and insufficiently strong materials."^
92
-------
Similarly, Wei describes two mechanisms of particle attrition:
'grinding' or the abrasive removal of a layer of
crystallites and matrix from the skin, and
'shattering' or the deep disintegration of the
matrix material.
The former mechanism leaves behind a large particle somewhat
reduced in size and a pile of very fine particles; the latter mechanism
leaves an assortment of fragments from the very small to the very large.
The former is controlled by the hardness of the crystallites and the
abrasion resistance of the matrix; the latter is controlled by the
impact elasticity of the matrix and imperfections in the structure.^
Doheim, Ghaneya, and Rassoul^ observed with fluidized iron ores in- a
nonreacting system that the primary mechanism of attrition is by abra-
sion, not breakage. Blinichev and others^- report that the wear of hard
fluidized particles is by abrasion; soft materials split, then abrade.
Forsythe and Hertwig^", Kutyavina and Baskakov,A6 Zenz^^ make the same
observation.
In this report we have limited discussion to only the first source
of attrition, grinding caused by rising gas bubbles in a fluidized bed.
In most fluidized beds several attrition mechanisms will act. In this
study we eliminated the grid (distribution-plate) jets by using a por-
ous, sintered-metal grid and avoided temperature and chemical effects by
operating at room temperature. Energy collisions also occur above the
bed where particles splash into the freeboard as bubbles break.
EXPERIMENTAL STUDIES
Table Al presents an index of experiments carried out.
93
-------
Table Al
INDEX OF EXPERIMENTS
Experiment
Title
Jet Observations in
a Semicircular Bed
Dependence of Jet
Length on Orifice
Dimensions and Par-
ticle Diameter
Measurement of Par-
ticle Material
Strength in a
Circulating Bed
Measurement of Par-
ticle Material
Strength in a Jet
without Circulation
Measurement of
Variation in the
Composition of Cal-
cined Limestone
Measurement of
Attrition Attri-
buted to Grid Jets
Testing for Attri-
tion Tendency of
Fluidized-Bed Gasi-
fication Sorbents
Purpose
Observe the character of
solids flow into a jet
Determine jet lengths to avoid
unnecessary depth of cover
solids and thus minimize
unwanted bubbling-bed
attrition
Provide relative measures of
stone attrition resistance in
Jet attrition
Measure the hardness of par-
ticle material by extent of
attrition where particles
hit a target only once and
do not recirculate
Determine if the composition
of limestone varies between
small particles
Investigate and describe
attrition In the vicinity of a
grid
Develop an apparatus and pro-
cedure for screening sorbents
on the basis of attrition ten-
dency
Apparatus
7- and 20-cm
semicircular bed
7-cm circular
fluldized bed
7-cm circular
fluidized bed
Target impact Ion
device
Wet chemical
assay
7-cm circular
fluidized bed
10-cm circular,
high-temperature
fluidlzed-bed
system
Reported In
Section
of This
Report
Appendix
Monthly
Report
for
1/78, 2/78
2/78
3/78, 4/78
Not
reported
Not
reported
-------
First Experiment: Jet Observations in a Semicircular Bed
Rationale and Purpose
An unknown in grid jet attrition is the character of particle
entrainraent into a jet. The purpose of these experiments was to observe
the flow of particles toward and into a jet.
Apparatus
In these experiments we used a semicircular transparent cell, 7-cm
in inside diameter. The apparatus shown in Figure Al has one semicirr
cular orifice.
Procedure
The apparatus was used for observing the circulation of particles
in a jet. We filmed motion and trajectories in 7- and 20-cm-id semicir-
cular beds. About one percent of the bed particles was colored red to
clarify the motion of individual particles.
Dwa. 7692A08
Bed of
Granular
Solids
Semicircular
Orifice
Figure Al. Semicircular Jet Model
95
-------
Results
The product of this study was close-up films of particle motion in
the vicinity of a jet. Repeated viewing of the films revealed that:
There is no small-scale turbulence in the bed of particles
except very close to the jet- Bed particles follow
smooth, parallel trajectories.
Particles follow roughly elliptical trajectories starting
at the top of the jet. Particles migrate to the jet and
are entrained and delivered by the jet to its top, where
they begin another circulation.
We observed an envelope such as Merry^11 described enclos-
ing the jet circulation region (Figure A2).
Owg.
Boundary of Dilute Phase Jet
-V
/
Dividing Streamline
In
N
\
\
\
,
X ,
/ \ /
Region of Intense
Particle Movement
.Adjacent to Jet
Boundary
Figure A.2. Graphic Representation of the Interaction of Particle
and Fluid Flow Fields in the Vicinity of the Jet
96
-------
Discussion
Inspection of these films verified the complexity of particle cir-
culation and attrition In the region of a jet. None of the rates Is
described for particle entralnment and circulation. Study and descrip-
tion of particle circulation rates, entralnment rates, and local attri-
tion rates will be needed for a basic understanding of grid Jet attri-
tion. For this study we decided to concentrate on a statistical
approach to defining grid jet attrition rather than to investigate the
interconnected complex of mechanisms.
Conclusions
Solids are entrained Into a jet over its entire length.
The density of solids in a jet increases with height above the
orifice.
The mass flow of entralnment Into a jet is constant over its
length. The mass flux (mass/area/time) decreases with height
approximately as
-------
Apparatus
The 7-cm-id attrition test cell used in this experiment is pictured
in Figures A3 and AA. It accepts interchangeable grids with a single
orifice.
Procedure
We filled the cell to a measured depth and increased the gas flow
rate very slowly until the jet broke through the bed surface.
Results
Figures A5 and A6 show the results of these jet length measure-
ments. All of these curves have positive slopes, affirming that a -
greater grid, AP, increases gas flow and causes a longer jet. Sim-
ilarly, increasing the orifice diameter causes a longer jet. We mea-
sured jet length in beds of two particle sizes (dp). For the particle
sizes tested, 500 to 710 urn and 1000 to 1410 um, the jet length is about
inversely proportional to particle diameter.
Conclusions
We can conclude from Figures A5 and A6 that:
Jet length is about proportional to grid AP.
Jet length is about inversely proportional to particle
diameter for a given material and grid AP.
Jet length increases with increasing orifice diameter.
We have drawn these conclusions from limited experiments. They
apply only to the conditions encountered in this apparatus; the results
are not generally applicable.
98
-------
Figure A3. Overall Photo of Attrition Test Cell
99
RM-74891
-------
Kc
C/
ta meter
/
/
Pressure
Taps
Plenum
Filter
Plexiglas Fluidized Bed
6. 99 cm ID x 91.44 cm High
^Sintered- Met* Wrtrtfcrter
PWe
ฉ Pressure Gauge
Figure A4
SIK 31827
Rotameter
j Valve
' Regulator
1 House Air
Flow Diagram for Room-Temperature
Fluidized Bed
Third Experiment: Measurement of Particle Material Strength in a Jet
with Circulation
Purpose
In the process of grid jet attrition, particles entrain into the
high-speed jet, accelerate, and smash into the roof of the jet. We
hypothesize that soft materials (such as chalk) will attrite readily and
hard materials (such as diamonds) will attrite slowly; in other words,
attrition rate varies inversely with particle strength. The purpose of
this experiment was to assign a measure of particle hardness to several
materials ranging from very soft co very hard.
Apparatus
The test apparatus was the 7-cm-id test cell (Figure A3) fitted
with the orifice-and-target device shown in Figures A7 and A8. In this
ment the orifice diameter was 0.257 cm.
100
-------
Curve 694597-A
12
I I I I I
I I T I I I I
?oa Bed of 500-710 urn particles
ปซ*Bed of 1000-1410 urn particles
5 10 15 20 25 30 35 40 45 50 55 60 65 70
AP across the Orifice, psi
Figure A5. Jet Lengths in Beds of Tymochtee Dolomite as
Affected by Grid AP, Orifice Diameter, and
Stone Particle Diameter, 18ฐC
:urve 6Mi96-5
"1 1 1 1 1 1 1 1 1 1 1 1 1 1 T
15
12
I
Bedof 500-710 urn Particles
Bed of 1000-1410 urn Particles
j i i i i I I I i
i i i i i
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
AP across the Orifice, psi
Figure A6. Jet Lengths in Bed of Tymochtee Dolomite as
Affected by Grid AP, Orifice Diameter, and
Stone Particle Diameter, 18ฐC
101
-------
o
K)
Figure A7. Jet-and-Target Device for Measuring Particle Strength
-------
Threaded 4.40
Stainless Steel
i i f
0.476 - cm Holes. 2.54 - cm Bolt Circle
n0.635 cm
T
6.985cm
-0.9525cm
Brass
\
0.1575 cm orifice
Threaded 4.40
Owg. 169*1896
6 - 0.635 - cm Holes
12.7-cm Bolt Circle
Figure A8. Target Assembly for 7-cra Attrition Cold Test Cell
-------
Procedure
The procedure In these tests was to charge the test cell with 1000
to 1410 urn ground stone to a depth of 9 cm, turn on the gas flow to a
plenum pressure of about 100 kPa gauge, and allow the jet to operate for
5.0 minutes. The detailed test procedure was:
Presieve all stone to 1000 to 1410 ym.
Place filter in 100ฐC oven for 1 hour, cool in desiccator for
1/2 hour, and weigh.
Assemble 7-cm bed with the orifice and the 5.1-cm-diameter
target 7.0 cm above the orifice.
Place presieved stone in 100ฐC oven for 1 hour, cool in desic-
cator for 1/2 hour. Weigh.
Fill bed to a total depth of 9 cm with stone and weigh.
Install filter on bed exhaust.
Set rotameter to 65 percent. Run for 5.0 minutes. Record
plenum pressure and rotameter pressure.
After test remove filter, dry in 100ฐC oven for 1 hour, cool in
desiccator for 1/2 hour, and weigh.
Remove stone from bed at 100ฐC for 1 hour, cool in desiccator
for 1/2 hour, weigh.
Perform sieve analysis on recovered solids.
We removed the solids from the system and carefully measured the size
distribution of the product.
Results
The motion of solids in the jet attrition apparatus was evident
when we reviewed particle motion through the clear plastic cell. Par-
ticles could be seen circulating downwards as shown in the inset
figure.
104
-------
Dwq. 7692A07
Target
Results of these tests are listed In Table A2 and in Figures A9 and
A10. Figure A10, which shows the differential size distribution on an
arithmetic frequency ordinate, clearly shows the relative amounts of
attrition products. The softer stones, tuff and marble, attrited more
severely after 5.0 minutes of jet action; the effect on the harder
aplite and diabase was much less noticeable. For all four minerals the
mode (peak) of attrition products we conjecture is at about 50 pm, with
marble showing an additional mode at 300 pm. We conjecture these modes
to be related to grain or crystallite sizes. Table A2 summarizes sieve
analysis data and lists the specific surfaces of powders. Specific sur-
face data are based on the entire charge of powder, with an assumed par-
ticle diameter of 10 pro for lost powder. We analyzed filter fines from
the tuff attrition by Coulter Countervjy. The logarithmic median diam-
eter was <(> - 7.1 (7.5 pm), with a deviation measure of $ = 1.24. The
Hazen effective sand size of filter fines was 1.95 pm (fine silt), with
a uniformity coefficient of 4.77.
During the tests, the pressure drop averaged 212-101 ป 111 kPa
across the 0.257-cm diameter orifice.
Discussion
These results suggest that we can differentiate between stone
types; we can identify the easily attritable and hard-to-attrite mate-
rials. The specific surface, a, and mean diameter, dav, are related
105
-------
Table 2
SUMMARY OF SIEVE ANALYSES AND POWDER STATISTICS AFTER 5.9 MINUTES OF JET
ATTRITION WITH A 0.256-cm JET
Stone Type
Specific Gravity g/cm3
Qualitative Hardness
Specific Surface, I06cm-l
Mean Diameter, cm*
% Wt Loss of Initial Size
Sieve Size, D2-Di.um
1410 - 2000
1000 - 1410
710 - 1000
500 -710
355-500
250-355
125-250
63-125
43-63
30-43
Filter + Losses
Before
All
( Same as after)
( Same as after)
50.4
0.119
0
Mean Diam.
1680
1190
840
595
420
300
180
90
52
36
10
Fraction
0
1.00
0
0
0
0
0
0
0
0
0
Frequency**
0
2.89
0
0
0
0
0
0
0
0
0
After Fluidization
Tuff
1.53
Soft
571
0. 0105
43.8
Grams
1.20
143. 70
40.68
5.16
3.04
2.20
3.60
6.60
19.72
19.16
12.54
Frequency**
0.0135
1.6123
0.4564
0.0579
0.0341
0.0247
0.0202
0.0370
0.2213
0.2150
0.0221
White Marble
2.18
Soft
164
0.037
36.3
Grams
14.40
251.70
45.54
14.60
15.96
20.17
28.81
13.83
7.23
3.34
1.92
Frequency**
0.0997
1.7424
0. 3153
0.1011
0.1105
0.1396
0.09%
0.0478
0.0501
0.0231
0.0021
Aplite
2.40
Hard
96.5
0.062
18.1
Grams
3.30
345.20
44.12
8.97
6.60
4.75
5.00
2.98
2.24
0.76
1.68
Frequency"
0.0224
2.3442
0.2996
0.0609
0.0448
0.0323
0. 0170
0.0101
0.0152
0.0052
0.0018
Diabase
2.88
Hard
80.0
0.075
15.0
Grams
7.10
428.30
46.00
7.58
5.50
4.56
5.12
3.36
3.77
0.36
0.75
Frequency**
0.0400
2.4158
0.2595
0.0428
0.0310
0.0257
0.0144
0.0095
0.0213
0.0020
0.0007
(Mi/Dpi)
-1
Frequency^ Fraction/In ( D-/D. '
-------
Curve 6961 77-B
1.0
o Diabase 3-22-78
Q Aplite 3-22-78
a Tuff 3-23-78
ฐ White Marble 3-23-78
0.1
>^
o
0.01
0.001
Size Range of ฐ
Starting Material
10
100
1000
10000
Particle Diameter,
Figure A9. Size Frequency Distribution Curves for Attrition Hardness
Testing of Several Minerals. Logarithmic Ordinate-
f - fraction on sieve * ^n(dpmax/dpmin)
107
-------
Curve 716191-B
2.5
2.4
2.3
2.2
2.1
2.0
1.9
1.8
1.7
1.6
1.5
1.4
>s
ฃ 1.3
O>
I 1.2
u.
1.1
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0
o Diabase 3-22-78
a Aplite 3-22-78
* Tuff 3-23-78
o White Marble 3-23-78
Diabase
Aplite
White
Marble
100
Particle Diameter, urn
Figure A10. Size Frequency Distribution Curves for Attrition
Hardness Testing of Several Minerals.
Logarithmic Ordinate. f * (fraction on sieve)
108
-------
by d
sv
6 * a. There is no formula relating the mean diameter of
solids after jet attrition and the percent of solids reduced smaller
than the starting sieve size. When compared on a graph (Figure All).
however, the diameter dsv and percent loss of coarses* correlate well;
they are related by the regression line
loss of coarses a 50.5 - 480.3 d
svป
0.980
for 0.0105 < dsv < 0.0750 cm. This correlation means that we can deter-
mine the easily measured percent loss of coarses and obtain a precise
measure of either the mean particle diameter or the specific surface of
the powder.
Curve 716190-A
CD
O
I/I
3
40
30
20
10
Tuff
Marble
Aplite
Diabase
.02 .04 .06 .08
Mean Particle Dlam, cm
Figure All - Relation between Loss of Coarses
and Final Particle Diameter
*Coarses (as opposed to fines) are solids in the largest sieve class,
1000 to 1410 urn.
109
-------
Stone strength values (attrition resistance) are calculated from
this summary of Table A2.
Before Treatment
After Treatment
Stone
All stones
Tuff
White marble
Aplite
Diabase
Specific Surface (cm"1 x 106)
Increase over "before"
50.4
571
164
96
80
0
521
114
46
30
We calculated the change in specific surface for each stone after five
minutes of attrition in a single jet. Increases in specific surface for
the four types of stone are
Tuff
521
Marble
114
Aplite
46
Diabase
30
Relating these values to that of tuff = 100, the softest mineral gives
100
21.9
8.8
5.8
These values'are relative measures of ease of breakage. Their inverses
measure stone strength.
100
457
1132
1740
110
-------
We chose tuff and marble as the two solids for study in the grid-jet
attrition tests. Their relative hardnesses taken as
- Tuff = 100
P
- Marble - 457
P
are sufficiently different, and both are in the same hardness range as
calcined limestone sorbent.
Conclusions
Single jet attrition testing of different minerals gives
increases in specific surface (or loss of coarse particles
or decrease in mean particle diameter) in the ranking
expected.
The attrition fragments for the four minerals tested have
modes at about 50 vim. White marble shows a second mode at
about 300 inn. These modes are conjectured to be related
to the diameters of grains or crystallites comprising the
minerals.
The mean particle diameter (inverse of specific surface)
correlates linearly at 98 percent with the loss of coarse
particles in the orifice-and-target apparatus.
Tuff and marble are two minerals well suited to jet attri-
tion, studies. Their measures of hardness or attrition
resistance are sufficiently different, and both have hard-
nesses comparable to those of calcined limestone or
dolomite.
Fourth Experiment; Measurement of Particle Material Strength in a Jet
without Circulation
Purpose
The apparatus and procedure described in the third experiment
measures the relative strength of circulating particles. The rate of
attrition depends on both the particle strength and the rate of
111
-------
circulation. This procedure provides a measure of particle strength as
related to the jet where particles do, indeed, circulate but does not
give a measure of particle strength alone.
The purpose of this experiment was to measure the relative strength
(attrition resistance) of minerals when attrited by a single impact
against a target.
Apparatus
Figure A12 shows the apparatus used in this experiment. Its pur-
pose is to accelerate particles to near-sonic velocity and shoot them
against a target, thus providing a measure of particle strength.
0.635-cm Tube
Air
Solids
f"i~ 0.953-cmTube
110.2 cm I
Target
Dwo. 6435A14
7.0-cm Plexiglas Exhaust
I Tube f
Filter ""
" JJ1
Figure A12 - Apparatus for Measurement of Particle Strength
in a Nonrecirculating System
Procedure
The test procedure in this experiment was to
1. Presieve all stone - 12 + 16 U. S. mesh.
2. Place filter in 100ฐC oven for one hour; place in desiccator
for one-half hour.
3. Assemble 7-cra bed with target and stone injection system.
4. Place presieved stone in 100ฐC oven for one hour; place in des-
iccator for one-half hour. Weigh.
5. Place 100 g of stone in funnel.
112
-------
Curve 695730-B
1 -
- 0.1
o>
I
0.01
I I I I I
O White Marble 4-4-78
o Diabase 4-4-78
D Aplite 4-4-78
ฃ> Tuff 4-5-78
0.001
J I
100 1000
Particle Diameter, pro
Figure A13 - Size Frequency Distribution for Impingement of
a Stream of Nonrecirculating Particles
against a Target
113
-------
6. Install filter on bed.
7. Set rotameter to Z = 75.
8. After test remove filter and stone. Place In 100ฐC oven for
one hour; place in a desiccator for one-half hour. Weigh.
9. Sieve and analyze attrition products.
Results
Results of testing for particle attrition by impingement on a plate
are listed in Table A3 and graphed in Figure A13. As in the third
experiment, the tangibly softer stones, tuff and marble, attrited more
than did the harder aplite and diabase. Again, there is a mode near
50 Mm (30 Mm in this experiment) and a second mode for marble (more, of a
shoulder in this experiment) near 300 ym. Table A3 summarizes frequency
data and lists the specific surface and surface-volume diameter dsv for
solids after treatment. The mass balances in this experiment were very
close, all within 100 ฑ 0.9 percent. Detailed sieve analysis data are
listed in Table A4.
Discussion
As in the third experiment, we compared the weight % loss of
coarses with the mean particle diameter of the attrited solids. Fig-
ure 14 shows that the produce diameter dsv and percent loss of coarses
correlate closely as
% loss of coarses - 17.9 - 93.3 dsv, r = 0.971
for 0.041 < dsv < 0.109 cm.
Figures All and A14 are not directly comparable as they involve differ-
ent mechanisms; the fluidization mechanism is dependent on time, the
single-impact-jet mechanism is independent of time. Again, however,
mean particle diameter dsv and weight percent loss of coarses are
closely related.
114
-------
Curve 716883-A
co
o>
CO
o
O
M
O
CO
co
3
14
1C
.02 .04 .06 .08 .10
Mean Particle Diam., cm.
Figure A14. Relation between Loss of Coarses and Final Particle
Diameter for High-Speed Impingement
Table A3
SUMMARY OF SIEVE ANALYSES AND POWDER STATISTICS AFTER SINGLE
IMPACT OF PARTICLES AGAINST A TARGET
Stone Type
Specific Gravity, g/cm3
Qualitative Hardness
Specific Surface, cm~l
Mean Diameter, cm*
X Wt Loss of Initial Size
Sieve Size,' Urn
1440-2000
1000-1410
710-1000
500-710
355-500
250-355
125-250
63-125
43-63
30-43
Mean Dia. , Urn
1680
1190
840
595
420
300
180
90
52
36
10
Before
All
Same as after
Same as after
50.4
0.119
0
0
2.8902
0
0
0
0
0
0
0
0
0
After Treatment
Tuff
1.53
Soft
146
0.041
14.9
Size
0.0145
2.4451
0.2896
0.0191
0.0095
0.0058
0.0032
0.0072
0.0162
0.0405
0.0046
White
Marble
2.18
Soft
113
0.053
11.9
Aplite
2.40
Hard
55.1
0.109
7.5
Diabase
2.88
Hard
55.7
0.108
8.2
requency
0.0318
2.5145
0.1789
0.364
0.0275
0.0246
0.0169
0.0066
0.0020
0.0020
0.0039
0.0173
2.6561
0.1832
0.0116
0.0072
0.0049
0.0025
0.0012
0.0009
0.0014
0.0000
0.0491
2.6040
0.2000
0.0127
0.0075
0.0049
0.0026
0.0016
0.0012
0.0020
0.0000
*6/Specific Surface
**Fractlon of posder mass within the size range d ฑ AD - F (d) in
&d
d - 6d
115
-------
Table 4
DETAILED SIEVE ANALYSIS DATA FOR JET ATTRITION WITHOUT CIRCULATION
WT MAR 4-4-78
BEFORE *************
TOTAL f/D- 8.403361344538
1 /TOTAL - .119
SPECIFIC SURFACE- 504201.6806723
T MESH MASS FRACT. f/d
16 100.0000 1.0000 8.4034
************* AFTER **************
TOTAL f/Dซ 18.82833620627
I/TOTAL - 5.31114374E-02
SPECIFIC SURFACE- 1129700.172376
f/(lnD2/Dl)
2.8902
T MESH
12
16
24
32
- 42
60
115
250
325
PAN
F+L
MASS
1.1000
87.0000
6.2200
1.2600
0.9500
0.8500
1.1700
0.4600
0.0700
0.0700
0.8500
FRACT.
0.0110
0.8700
0.0622
0.0126
0.0095
0.0085
0.0117
0.0046
0.0007
0.0007
0.0085
CUM FRACT.
1.0000
0.9890
0.1190
0.0568
0.0442
0.0347
0.0262
0.0145
0.0099
0.0092
0.0085
f/d
0.0655
7.3109
0.7405
0.2118
0.2262
0.2833
0.6500
0.5111
0.1346
0.1944
8.5000
f/(lnD2/Dl)
0.0318
2.5145
0.1798
0.0364
0.0275
0.0246
0.0169
0.0066
0.0020
0.0020
0.0039
APLITE 4-4-78
************ BEFORE *************
TOTAL f/D- 8.403361344538
I/TOTAL - .119
SPECIFIC SURFACE- 504201.6806723
T MESH MASS FRACT. f/d
16 100.0000 1.0000 8.4034
************* AFTER **************
TOTAL f/D- 9.176497162962
I/TOTAL - .1089740433895
SPECIFIC SURFACE- 550589.8297776
f/(lnD2/Dl)
2.8902
T MESH
12
16
24
. 32
ป2
60
115
250
325
PAH
F+L
MASS
0.6000
91.9000
6.3400
0.4000
0.2500
0.1700
0.1700
0.0800
0.0300
0.0500
0.0100
FRACT.
0.0060
0.9190
0.0634
0.0040
0.0025
0.0017
0.0017
0.0008
0.0003
0.0005
0.0001
CUM FRACT.
1.0000
0.9940
0.0750
0.0116
0.0076
0.0051
0.0034
0.0017
0.0009
0.0006
0.0001
f/d f/(lnD2/Dl)
0.0357 0.0173
7.7227
0.7548
0.0672
0.0595
0.0567
0.0944
0.0889
0.0577
0.1389
0.1000
2.6561
0.1832
0.0116
0.0072
0.0049
0.0025
0.0012
0.0009
0.0014
0.0000
116
-------
Table 4 (Continued)
TUFF 4-5-78
************ BEFORE *************
TOTAL f/D- 8.403361344538
I/TOTAL ป .119
SPECIFIC SURFACE- 504201.6806723
T MESH MASS FRACT.
16 100.0000 1.0000
f/d
8.4034
f/(lnD2/Dl)
2.8902
************* AFTER **************
TOTAL f/D- 24.33161495366
I/TOTAL - 4.10987927E-02
SPECIFIC SURFACE- 1459896.89722
MESH
12
16
24
12
2
oO
115
250
325
PAN
F+L
MASS
0.5000
84.6000
10.0200
0.6600
0.3300
0.2000
0.2200
0.5000
0.5600
1.4000
1.0100
FRACT.
0.0050
0.8460
0.1002
0.0066
0.0033
0.0020
0.0022
0.0050
0.0056
0.0140
0.0101
CUM FRACT.
1.0000
0.9950
0.1490
0.0488
0.0422
0.0389
0.0369
0.0347
0.0297
0.0241
0.0101
f/d
0.0298
7.1092
1.1929
0.1109
0.0786
0.0667
0.1222
0.5556
1.0769
3.8889
10.1000
f/(lnfl2/Dl)
0.0145
2.4451
Q.2896
0.0191
0.0095
0.0058
0.0032
0.0072
0.0162
0.0405
0.0046
DIABASE 4-4-78
************ BEFORE *************
TOTAL f/D- 8.403361344538
I/TOTAL - .119
SPECIFIC SURFACE- 504201.6806723
T MESH MASS FRACT. f/d
16 100.0000 1.0000 8.4034
************* AFTFR **************
TOTAL f/D- 9.282539323418
I/TOTAL - .1077291423347
SPECIFIC SURFACE- 556952.3594051
f/(lnD2/Dl)
2.8902
MESH
12
16
24
32
42
60
115
250
325
PAN
F+L
MASS
1.7000
90.1000
6.9200
0.4400
0.2600
0.1700
0.1800
0.1100
0.0400
0.0700
0.0100
FRACT.
0.0170
0.9010
0.0692
0.0044
0.0026
0.0017
0.0018
0.0011
0.0004
0.0007
0.0001
CUM FRACT.
1.0000
0.9830
0.0820
0.0128
0.0084
0.0058
0.0041
0.0023
0.0012
0.0008
0.0001
f/d
0.1012
7.5714
0.8238
0.0739
0.0619
0.0567
0.1000
0.1222
0.0769
0.1944
0.1000
117
-------
We calculated the change in specific surface for each stone after
impact in the jet. Increases in specific surface for the four types of
stone are:
Stone
Before
After
Increase
Tuff
50.4
146.0
95.6
Marble
50.4
113.0
62.6
Aplite
50.4
55.1
4.7
Diabase
50.4
55.7
5.3
Relating these values to that of tuff = 100, the softest mineral gives:
100
65.5
4.9
5.5
These values are relative measures of ease of breakage. Their inverses
(scaled up to tuff = 100) measure stone strength:
100
152.7
2034
1804
These values do differ from those measured in the circulating bed
(Experiment 3):
100
457
1132
1740
but the correlation between measures of hardness is fairly good
(Figure A15).
118
-------
Curve 716189-A
01 O
ฃ 2000
.22 b
1000
CO
CO
0
1-to-l
Correlation
0
1000
2000
Figure A.15. Measure of Particle Hardness by Single Impact
from a Jet
We are confronted with two procedures for estimating material
strength on impact. One procedure (preceding experiment) combines both
particle strength and rate of circulation; the other (this experiment)
eliminates rate of circulation and is affected primarily by particle
strength. We believe that the procedure involving jet action in a bed
of material (preceding experiment) is more pertinent. The rate of cir-
culation is an integral part of the grid jet attrition mechanism. Since
it cannot be.measured separately by any simple test, the circulation
rate is best included with the measurement of material strength.
Conclusions
Particles injected into a horizontal jet and shot against
a target plate attrited appreciably. Between 8 and
15 percent of the coarse solids were fragmented by a sin-
gle impact.
119
-------
As in the jet attrition with circulation of a bed of par-
ticles the mode of fragment sizes was near 50 pm (30-
40 pm in the experiment), with a second mode for white
marble near 300 pm. Tuff fragments show a well-defined
mode at 40 pm.
The weight loss of coarse particles from a single high-
speed impact is well correlated with the mean particle
size of all attrited particles.
Both methods of measuring particle hardness or attrition
resistance (third and fourth experiments) give comparable
results.
We chose the method of measuring particle hardness in a
circulating bed for the following experiments because the
procedure is more like the action in a fluidized bed.
Fifth Experiment; Measurement of Variation in the Composition of
Calcined Limestone
Purpose
In some TGA or wet chemical measurements a single particle of sor-
bent is tested. One assumes that the single particle epitomizes all
particles.
After calcining limestone, however, we have noticed that some par-
ticles remain gray and others become white, as expected in dead-burning
limestone. We separated and analyzed gray and white particles of cal-
cined limestone to see if they differed in composition.
Procedure
We gathered several particles of each color about 1 mm in diameter
and assayed them for chemical composition. Particles had been fluidized
at 815ฐC for 100 hours in nitrogen.
120
-------
Results
Results of the chemical assay are
Chemical Species
Carbon Dioxide
Calcium
Magnesium
%
White Particles
1.47
64.96
0.39
Gray Particles
1.51
4.75
0.00
Discussion and Conclusion
These results suggest that impurities are present in large aggre-
gates, evidenced by both appearance (grayness or whiteness) and chemical
composition. They further suggest caution in gathering small sorbent
samples and the possible need to mill and split sorbents when a small
(
-------
Dwg. 7692A09
Roof of
Jet
Path of
Particles
Entrained
into the
Jet
Orifice
Plenum
Characteristics of the system comprising this model are
(Figure A17):
Particles are entrained into the jet over most of its
length at a flux F(Z) where Z is height measured aboveve
the grid. YangAl^ has measured the rate of particle
entrainment into a jet for two lengths of jet exposure.
His apparatus allows masking the top of a jet by directing
it into a variable-height draft tube. Yang's results show
that the rate of solids entrainment into a jet (mass/unit
time) increases with the length of jet exposed above the
orifice.
Particles accelerate within the jet. The rate of acceler-
ation vanes with radial position in the jet (Figure A16).
The jet velocity decreases with Z because of the jet
expansion (Figure A17). The velocity field within the jet
is not the same as for a jet in a nozzle or a free jet in
air. The entrained particles extract momentum from the
gas, and the particles are entrained at some unknown
rate.
122
-------
Particles strike the roof of the jet and shatter. The
distribution of fragment sizes depends upon particle
strength a, particle velocity Up, and particle diameter
dp. Fragment size distributions B (x, y) (Figure A18)
have been studied for slow crushing or static crushing but
have not been found reported for high-velocity impact.
Particles circulate from the upper part of the jet to the
base of the jet. Merry has shown that the particles may
recirculate from the upper jet to the jet base and be
reground in the jet (Figure A2).A11
Analysis
The foregoing description outlines an exceedingly complex momentary
balance and circulation pattern. The system does not appear amenable to
rigorous model analysis but seems best suited to dimensional analysis
and experimental measurement of coefficients. The quantities involved
in jet attrition are
Symbol Description Dimensions
R attrition rate per jet g/s
U0 jet velocity at orifice cm/s
pg particle density
po gas density at orifice
dp particle diameter cm
d0 ' orifice diameter cm
g gravity acceleration
gc Newton' s law conversion factor g cm/m/dyne. s^
as particle strength dyne/cm^
As increase in specific surface cm /g
This list of variables, if each were tested, would present a formidable
experimental program. "There exists a method between formal mathemati-
cal development and a completely empirical study. It is based on the
fact that, if a theoretical equation does exist among the variables
123
-------
Curve 716192-B
T) A particle migrates
through the bed
4) and smashes into the
top of the jet, shattering
into fragments
3) It accelerates in the
high-speed gas,
2) and is swept into the jet
V////////////
Figure A16. Movement of Particles through a Jet
Gas Velocity Particle Mass
Figure A17. Conjectured Distribu-
tions of Gas Velocity
and Particle Mass in
a Fluidized-Bed Jet
u
Figure A18.
Gas Velocity
Possible Particle
Velocity Profile
Variation of Particle
Velocity and Jet Aver-
age Gas Velocity with
Height above the
Grid Z.
124
-------
Dwg. 7692A10
Low-Velocity
Particle
High-Velocity
Particle
Figure A19. Fragmentation Size Distribution B(x,y). Probably is
affected by particle velocity before impact.
affecting a physical process, that equation must be dimensionally homo-
geneous. Because of this requirement it is possible to group many fac-
tors into a smaller number of dimensionless groups of variables."^*
The independent groups that can be formed from variables in the
preceding list include
Dimensionless
Group Name
Attrition number
Definition Interpretation
R Rate of fines formation
.2 Rate of air mass delivered
Ng0 Bond number
Npr Froude number
ND Diameter ratio
Np Density ratio
(OB/PS)
u2
g
PS
Gravitational force
Attrition resistance
Inertial force
Gravitational force
Particle diameter
Orifice diameter
Gas density
Solid density
125
-------
Our approach to describing jet attrition* was to measure the aver-
age rate of attrition in the grid region and relate it to dimensionless
groups serving as independent variables. The actual attrition measure-
ments were preceeded by a series of experiments for measuring sorbent
hardness or attrition tendency (third and fourth experiments) in which
R = g of fines forraed/s/jet
P0 = gas density at orifice, ps - solids density, g/cm^
d0 = orifice diameter, dp = particle diameter, cm
Uo = gas velocity at the orifice, cm/s
g = gravity acceleration, cm/s^
gc = Newton's law conversion factor g . cm/dyne s^
AO, Al, A2, A3, A4 = regression coefficients
og = solid strength, dyne/cm^
Apparatus
The apparatus we used in these tests was a a 7-cm-id vertical
plastic tube filled with a single-hole grid. The grid was interchange-
able for varying the orifice diameter. Figures A7 and A21 show the
apparatus.
Procedure
In each test we filled the tube to depth of about 10 cm with a bed
of sieved solids, set the gas flow, and let the bed jet for 5.0 minutes.
During the test we recorded plenum pressure, rotameter float, and pres-
sure readings; after the test we sieved the bed solids. Orifice gas
density and temperature were calculated from a compressible-flow func-
tion table for isentropic gas expansion.
*Some researchers^-^ question the existence of grid jets. We can as
well consider this effect to be attrition caused by the high-velocity
gas streaming from the grid orifices.
126
-------
Figure A20. Close-up Photo of 7-cm-id Attrition Cell
127
RM-80594
-------
oo
Table A5
VALUES OF INDEPENDENT VARIABLES AND RESPONSES IN THE GRID JET ATTRITION TESTS
0ปq. 1697B88
Test
J-l
J-2
J-3
J-4
J-5
J-6
J-7
J-8
J-9
J-10
J-ll
J-12
Date
1978
V2
V2
5/2
V3
4/28
5/3
V4
V3
5/4
5/4
5/5
5/1
Stone8
Tuff
Tuff
Marble
Tuff
Tuff
Tuff
Marble
Marble
Marble
Tuff
Marble
Marble
Gas
C02
C02
C02
Air
Air
Air
C02
Air
C02
C02
Air
Air
Orifice,
cm
0.157
0.157
0.112
0.112
0.157
0.157
0.157
0.112
0.112
0.157
0.112
0.112
"p
cm
0.0595 e
0.1190 '
0.1190
0.1190
0.0595
0.0595
0.0595
0.1190
0.0595
0.1190
0.1190
0.0595
0.1190
O/PS
dyne -cm
9
100
100
457
100
100
100
457
457
457
100
457
457
457
AP
psi
10
5
10
10
10
5
5
4
10
5
10
4
4
Pi
psia
24.7
19.7
24.7
24.7
24.7
19.7
19.7
18.7
24.7
19.7
24.7
18.7
18.7
Vl
0.595
0.746
0.595
0.595
0.595
0.746
0.746
0.786
0.595
0.746
0.595
0.786
0.786
Po/Pl
0.670
0.798
0.670
0.690
0.690
0.811
0.798
0.841
0.670
0.798
0.690
0.841
0.831
VT|
0.888
0.934
0.888
0.862
0.862
0.919
0.934
0.933
0.888
0.934
0.862
0.933
0.946
NBO"
0.242
0.242
0.027
0.123
0.242
0.242
0.053
0.027
0.027
0.242
0.027
0.027
0.027
Uoฐ
18452
14440
18179
23716
24322
23656
14440
19178
18179
14440
23716
18897
12956
Npr"
2.21
1.36
3.01
5.12
3.84
3.64
1.36
2.39
3.01
1.90
5.12
3.25
1.53
Nd
y\>
0.379
0.758
1.06
1.06
0.379
0.379
0.379
1.06
0.531
0.758
1.06
0.531
1.06
Po/Pp
0.0144
0.0137
0.0101
0.0098
0.0098
0.0092
0.0096
0.0063
0.0101
0.0137
0.0069
0.0063
0.0095
Increase in.
Sp.Surf
As,cm~l
71.9
32.1
1.8
31.0
139.9
21.9
17.8
6.8
9.4
71.0
6.8
0.7
8.6
FinesFrac.
AF
0.0755
0.0322
0.0054
0.046
0.1223
0.0334
0.054
0.0134
0.0439
0.0584
0.0336
0.004
0.0329
Mean Diam
Ad.iim
135
159
34
155
310
54
77
422
43
269
123
3
150
a Stone Densities: Tuff 1.53 g/cm3; Marble 2.18 g/cm3
b NBO =(9/9c' do /(Vps>
c UQ = Q0/A0=Q'(To/p0)N|p * PRM'/TRT'MR-H n/41
=Q =Q'(To/14.7)
14.7 PR 29/293- 293MR-H it/4) (ฃ = 0.006 ;
M = Molecular Weight, denotes reference conditions, o denotes conditions at orifice,
I denotes conditions in plenum, R denotes conditions in rotameter
24 - 32 Mesh
12 - 16 Mesh
-------
Results
Experimental results are listed in Table A5. Table A6 lists the
associated regression expressions of the form
Attrition = A0 + AXN0 + A2NFr + A3Nd + A4Np
and their error estimates, Se. (Se * halfrange) is the standard error
estimate of the residual error. It is calculated as the normalized
standard deviation of the differences between observed and predicted
responses. Se is an estimate of the standard deviation of the response
at any setting of the independent variable. The coefficient of deter-
mination, R^, is one measure of the quality of the model. It is the
fraction of variation in the response which is accounted for by the
model. The positive square root of R^ is the correlation between the
observed and predicted responses.
The normalized coefficients on different terms (Table A6) are com-
parable. For example the expression estimating Asxdฐ has coefficients
of 2.3, 0.3, 0.4, and 0.1 on the dimensionless variables. We can infer
that the Bond number is predominant in estimating Asxdฐ.
P
Discussion
Inspection of the eight models investigated (Table A6) shows that
the best estimates of grid jet attrition rate are described by the
responses (dependent variables) involving specific surface. Note that
2
in most models the Bond number (g/gc) d /(os/ps) is prominent. The best
fit (highest coefficient of determination, lowest relative standard
error) in Table A6 is given by the last expression for 10^ Asd /Uot).
o
This expression, however, includes time as a linear variable (As a t,
and prior experience suggests that As is a power function of t, As a tm,
0
-------
Table A6 Dซg. 1697687
REGRESSION ANALYSIS RESULTS FOR GRID JET ATTRITION TESTING
Response
Increase in
Specific Surface
Increase in
Fines Fraction
Decrease in
Mean Diameter
Attrition
Number I
Sp. Surf. Incr.
x Orifice Diam.
Sp. Surf. Incr.
x Particle Diam.
Attrition
Number II
Regression Model (T)
AS = 37. 3 + 27. 3 NB() f 12.1 N,/- 9.1 NJ +3.3N*
Af =0.051 + 0.002 NBQ* + 0.019NF* - 0.012N* + 0.024 N*
Ad = 140.8 + 127.8 NgJ - 95. 1 Np* + 92. 5 N* - 142. 8 N*
N.t = 0.007 -0.001 NBo* + 0.004 Np* - 0.002 NJ + 0.003 Np*
tn \i = - 5. 318 +0.082 NBo +0.338 NFr -0.245 NJ + 0.309Np @
(ASxcy*ฎ = -0.489+0.4 N^ + 0.90 Npr -0.113NJ +0.012Np
( ASxdp) ' ฎ = -0.279 + 0. 54 N^, + 0.069 Npr + 0. 109 NJ + 0. 143 N*
? * C^t
AC rfn \3r ซ *
6a^ "".- r- _n A BO i n 19 w n IAA M i n n1; w ^ n ^vt w
1(J uot ~ +0.10 Ng0 U. 106NFr + U.U5Nd + 0. Vfi Np
Se-H Range/2)
0.483
0.486
0.554
0.600
4.20
0.463
0.51
0.43
2
R . Coeff. of Determination
0.58
0.42
0.46
0.37
0.21
0.59
0.63
0.70
u>
o
N*=(N -Midrange Value of Response) -J-Halfrange of Response NB* =(NBo - 0.1345)-HO. 0175 NJ=(Nd-0.72)-:-0.34
NFr ~(NFr -3.24)-*-1.88 NJ =(Np'-0.01035)-5-0.
N' =.UnN -Midrange Value of I nl response! -5- Half range of tn [response I
00405
- 20.275
^-20.125
-------
2
The Bond number d (g/gc) * (os/ps) predominates and implies, first, that
larger grid holes will increase attrition and, second, that harder mate-
rial (larger os) will attrite more slowly. The Froude number accounts
for increased attrition with increasing gas velocity through the orifi-
ces. The density ratio infers that increasing gas density will increase
attrition. This increase is expected because the denser gas provides a
greater drag on particles and accelerates them faster in the jet. Simi-
larly, the positive diameter ratio suggests that larger particles will
attrite faster. This increase in speed, too, is expected where parti-
cles are accelerated to their terminal velocities in the jet: a larger
particle exerts a greater kinetic-energy (o particle mass)/required-
surf ace-energy (a particle surface) than does a smaller particle. For
much larger particles that do not accelerate to the jet velocity before
impact, however, the increase in attrition with particle size may not
apply.
Knowing the increasing in specific surface, As, caused by jet
action is of limited use in predicting fluidized-bed attrition. The
increase in fines content, AF, is the needed practical variable.
Regression analysis of the data in Table A5 gives us a relation between
AF and As:
As - 1908 AF1'399; r = 0.92 .
SCREENING TESTS: ATTRITION TENDENCY OF BROWNWOOD LIMESTONE
This work is described in the main text.
Conclusions
Attrition can be severe in grid jets.
Attrition in the vicinity of the grid occurs through
entrainment of particles in a gas jet issuing from the
grid, their acceleration in the jet, and their being
thrown at high velocity against the roof of the jet.
131
-------
This is a complex mechanism, and the separate processes
comprising the mechanism are unknown. The overall mechan-
ism is not easily amenable to modeling and analysis.
The overall mechanism is amenable to dimensional analysis
in which seven independent variables can be combined into
four independent dimensionless groups-
Experimental attrtrition results correlate well with
linear combination of four dimensionless groups (the Bond
number, Froude number, diameter ratio, and density
ratio).
Seventh Experiment: Testing for Attrition Tendency of Fluidized-Bed
Gasification Sorbents
The full text of this experiment is in the main text-
Summary
Fluidized beds are well suited to gasification of coal. The bed
solids, chemically-active limestone or dolomite, capture sulfur pollu-
tants as soon as they are released from the coal. The continued agita-
tion of particles in a bed, however, causes attrition to fines and a
subsequent loss of solids.
Natural materials vary in their resistance to attrition. To select
sorbents one must screen them by some laboratory procedure. The
purpose of this study is to develop a reproducible procedure for measur-
ing the attrition resistance of granular sorbents.
Coal gasifiers encounter temperatures of about 800 to 900ฐC. Sor-
bent added to an operating bed first experiences thermal shock, then
calcination. Jets at the grid and bubbling above the grid tumble the
sorbent particles. The sorbent screening process we have developed
includes all of these processes to attrite particles by thermal, chemi-
cal, and mechanical means.
132
-------
The test apparatus is a 9.5-cm-id cell with a three-hole grid.
Test temperatures are maintained by a furnace surrounding the cell. Our
test procedure was to determine the gas flow required to form 8-cm-high
jets in a bed of a particular sorbent. Sorbent was added to an empty
bed at 900ฐC and fluidized for one hour at 815ฐC at the predetermined
gas flow rate. Solids were sieved for particle size distribution before
and after the attrition treatment.
Replicate testings of Grove, Greer, Brownwood, and Pfizer sorbents
showed good repeatability between replicate tests and decisive differen-
ces in attrition tendency among different sorbents.
The apparatus and procedure developed in this study are not .presen-
ted as a universal method but rather as a prototype. This study dem-
onstrates that sorbents can be ranked decisively with regard to attri-
tion tendency.
Conclusions
We have demonstrated an apparatus and a procedure for mea-
suring the attrition tendency of granular sorbents.
The procedure includes the attrition mechanisms present
in the grid region, the bubbling bed region, splashing in
the freeboard, thermal shock, and calcination.
The procedure for sorbent screening tested in this study
discriminates decisively between the attrition tendencies
of different sorbents.
We do not propose the apparatus and procedure described
here as a standard. This method serves as a prototype and
demonstrates that a standard screening method can be
developed.
133
-------
REFERENCES
Al. Blinichev, V. N., Strel'tsov, V. V., and E. S. Lebedeva, An Inves-
tigation into the Size Reduction of Granular Materials during Their
Processing in Fluidized Beds, Intl. Chem. Energy., 8(4): 615-18;
October 1968.
A2. Jonke, A. A., et al. , Annual Report on a Development Program on
Pressurized Fluidized-Bed Combustion, Argonne National Labora-
tories, Argonne, IL, July 1976, ANL/ES-CEN-1016.
A3. Paige, J. I., J. W. Town, J. H. Russell, and H. J. Kelly, Sorption
of SC>2 and Regeneration of Alkalized Alumina in Fluidized-Bed Reac-
tors. Report of Investigations 7414, U.S. Bureau of Mines,
August 1970.
A4. Foster Wheeler Energy Corporation, Chemically Active Fluidized Bed
Process Monthly Technical Narrative No. 20, January 24-
February 20, 1977, prepared March 14, 1977.
A5. Craig, J. W. T., et al., Chemically Active Fluidized Bed Process
for Sulfur Removal during Gasification of Heavy Fuel Oil, Second
Phase. Report to EPA, Esso Research Centre, Abingdon, UK,
November 14, 1973, EPA-650/2-73-039.
A6. Kutyavina, T. A., and A. P. Baskakov, Grindings of Fine Granular
Material with Fluidization, Chemistry and Technology of Fuel Oils,
8(3): 210-13; March-April 1972.
A7. Wei, J., L. Wooyoung, and F. J. Krambeck, Catalyst Attrition and
Deactivation in Fluid Catalytic Cracking Systems, Chem. Eng. Sci. ,
32(10): 1211-18; 1977.
A8. Doheim M. A., A. A. Ghaneya, and S. A. Rassoul, The Attrition
Behavior of Iron Ores in Fluidized Bed Reactors, La Chimica e
L'Industria, 58(12): 836-40; December 1976.
134
-------
A9. Forsythe, W. L. Jr, and W. R. Herturg, Attrition Characteristics of
Fluid Cracking Catalysts, I&EC,41(6): 1200-06; June 1949.
A10. Zenz, F. A., Final Attrition in Fluid Beds, Hydrocarbon Proc.
50(2): 103-105; February 1971.
All. Merry, J. M. D., The Flow Fields of a Fluid and Particles around a
Vertical Jet in a Fluidized Bed, Westinghouse Research Labora-
tories, Pittsburgh, PA, Research Memo 75-8X1-PDULA-M1, 2-28-75.
A12. Experimental and Engineering Support of the CAFB Demonstrating
Monthly progress report to EPA, Westinghouse Research and Develop-
ment Center, Pittsburgh, PA, January 1-31, 1978, Contract 78-
02&2142.
A13. Sucin, G. C., and M. Patrascu, Particle Circulation in a Sprited
Bed, Pruder Technology 19, 1979, 109-114.
A14. Yang, W. C., and D. L. Keairns, Design of Recirculating Fluidized
Beds for Commerical Applications, AIChE, Symposium Series, No. 167,
74: 212-228; 1968.
A15. McCabe, W. L., and J. C. Smith, Unit Operation of Chemical Engi-
neering, New York: McGraw-Hill Book Company, Inc.; 1956, 21-22.
A16. Rowe, P. N., H. J. MacGillivray, and D. J. Cheesman, Gas Discharge
from an Orifice into a Gas Fluidized Bed, presented at 71st Annual
AIChE Meeting, Miami Beach, FL, November 1978.
135
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO
EPA-600/7-79-158a
2.
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Chemically Active Fluid Bed for SOx Control;
Volume I. Process Evaluation Studies
5. REPORT DATE
December 1979
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
D.L.Keairns, W.G.Vaux, N.H.Ulerich, E.J.Vidt,
and R. A.Newby
8. PERFORMING ORGANIZATION REPORT NO,
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Westinghouse Research and Development Center
1310 Beulah Road
Pittsburgh, Pennsylvania 15235
10. PROGRAM ELEMENT NO.
EHB536
11. CONTRACT/GRANT NO.
68-02-2142
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final; 7/75-10/79
14. SPONSORING AGENCY CODE
EPA/600/13
is. SUPPLEMENTARY NOTES jERL-RTP project officer is Samuel L. Rakes, Mail Drop 61, 919/
541-2825.
is. ABSTRACT Tne report describes selected process evaluation studies supporting the
development of an atmospheric-pressure, fluidized-bed, chemically active gasifi-
cation process, using a regenerative limestone sulfur sorbent to produce low- to
intermediate-Btu fuel gas. Limestone sorbent selection and attrition, alternative
medtal oxide sorbents, particulate control, fuel supply, and an updated process ass-
essment are investigated. Limestone sorbent selection results are presented for
the EPA-sponsored CAFB demonstration plant. Sorbent attrition and economics are
the main criteria as most limestone are not limited by sulfur removal. Trace ele-
ment, regeneration, and disposal characteristics should be considered. Feasibility
tests of air oxidation for disposal of gasifier solids for once-through operation show
up to 70% conversion of the CaS. Methods for improving performance are identified.
A procedure was developed to measure the attrition tendency of the sorbent selected.
Brownwood limestone has intermediate attrition resistance showing 5.4% mass loss
by attrition for this, test, compared with three reference stones ranging from 0. 5 to
9.1%. Sixteen alternative metal oxide sulfur sorbents that could reduce the environ-
mental impact of solids disposal and may improve process economics were screen-
ed. CaO/CaO3, ZnO, an:- FeO are sorbents identified for further study.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Pollution
Sulfur Oxides
Fluidized Bed Processing
Coal Gasification
Calcium Carbonates
Dust
Aerosols
Pollution Control
Stationary Sources
Chemically Active Fluid
Bed
Particulate
Metal Oxides
13B
07B
13H,07A
11G
07D
13. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (This Report)
Unclassified
20. SECURITY CLASS (This page f
Unclassified
21. NO. OF PAGES
MS
22. PRICE
EPA Form 2220-1 (9-73)
136
------- |