United States Office of Air Quality EPA-450/3-85-009
Environmental Protection Planning and Standards July 1984
Agency Research Triangle Park NC 27711
__
•oEPA Industrial Boiler
SO2 Technology
Update Report
-------
EPA-450/3-85-009
Industrial Boiler SO2
Technology Update Report
Prepared by:
Radian Corporation
Under Contract No. 68-02-3816
Prepared for:
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Radiation
Office of Air Quality Planning and Standards
Emission Standards and Engineering Division
Research Triangle Park, NC 27711
July 1984
-------
DISCLAIMER
This report has been reviewed by the Office of Air Quality Planning and Standards, U.S. Environmental
Protection Agency, and approved for publication as received from the Radian Corporation. Approval does
not signify that the contents necessarily reflect the views and policies of the U.S. Environmental Protection
Agency, nor does mention of trade names or commercial products constitute endorsement or recommenda-
tion for use. Copies of this report are available from the National Technical Information Services, 5285 Port
Royal Road, Springfield, Virginia 221 61.
-------
TABLE OF CONTENTS
CONTENTS i
LIST OF TABLES "" fv
LIST OF FIGURES V1-
Chapter 1 - INTRODUCTION 1-1
Chapter 2 - POST COMBUSTION CONTROL APPROACHES 2-1
2.1 WET SCRUBBING PROCESSES 2-1
2.1.1 Sodium 2-1
2.1.1.1 Process description 2-2
2.1.1.2 Factors affecting performance 2-5
2.1.1.3 Applicability to industrial boiler 2-12
2.1.1.4 Development Status 2-29
2.1.1.5 Reliability 2-29
2.1.1.6 Emissions data 2-32
2.1.2 Dual Alkali 2-36
2.1.2.1 Process description 2-38
2.1.2.2 Factors affecting performance 2-39
2.1.2.3 Applicability to industrial boilers 2-41
2.1.2.4 Development status 2-43
2.1.2.5 Reliability 2-44
2.1.2.6 Emissions data 2-46
2.1.3 Limestone Wet Scrubbing 2-48
2.1.3.1 Process description 2-50
2.1.3.2 Factors affecting performance 2-50
2.1.3.3 Applicability to industrial boilers 2-55
2.1.3.4 Development status 2-57
2.1.3.5 Reliability 2-61
2.1.3.6 Emissions data 2-62
2.1.4 Lime Wet Scrubbing 2-62
2.1.4.1 Process description 2-63
2.1.4.2 Factors affecting performance 2-63
2.1.4.3 Applicability to industrial boilers 2-65
2.1.4.4 Development status 2-65
2.1.4.5 Reliability 2-67
2.1.4.6 Emissions data 2-69
111
-------
TABLE OF CONTENTS (Continued)
2.2 DRY PROCESSES 2-70
2.2.1 Spray Drying 2-70
2.2.1.1 Process description 2-70
2.2.1.2 Factors affecting performance 2-73
2.2.1.3 Applicability to industrial boilers 2-75
2.2.1.4 Development status 2-76
2.2.1.5 Reliability 2-78
2.2.1.6 Emissions data 2-79
2.2.2 Dry Alkali Injection '. 2-81
2.2.2.1 Process description 2-81
2.2.2.2 Factors affecting performance 2-83
2.2.2.3 Applicability to industrial boilers 2-84
2.2.2.4 Development status 2-84
2.2.2.5 Reliability 2-85
2.2.3 Electron-Beam Irradiation 2-85
2.2.3.1 Process description 2-85
2.2.3.2 Status of development 2-87
2.3 REFERENCES 2-89
Chapter 3 - COMBUSTION MODIFICATION CONTROL APPROACHES.. 3-1
3.1 FLUIDIZED BED COMBUSTION 3-1
3.1.1 Process description 3_2
3.1.2 Factors affecting performance 3-6
3.1.3 Applicability to industrial boilers 3-9
3.1.4 Development status 3-12
3.1.5 Emission test data 3-13
• 3.2 LIMB 3_16
3.2.1 Process description 3-16
3.2.2 Factors affecting performance 3-19
3.2.3 Applicability to industrial boilers 3-19
3.2.4 Development status 3-21
3.2.5 Emissions data 3-22
3.3 COAL/LIMESTONE PELLETS 3-23
IV
-------
TABLE OF CONTENTS (Continued)
3.4 REFERENCES 3-24
Chapter 4 - PRECOMBUSTION CONTROL APPROACHES 4-1
4.1 PHYSICAL COAL CLEANING 4-1
4.1.1 Process description 4-1
4.1.2 Factors affecting performance 4-2
4.1.3 Applicability to industrial boilers 4-3
4.1.4 Development status 4-3
4.1.5 Performance 4-6
4.2 COAL GASIFICATION 4-6
4.2.1 Process description 4-7
4.2.2 Factors affecting performance 4-7
4.2.3 Applicability to industrial boilers 4-9
4.2.4 Development status 4-10
4.2.5 Reliability 4-10
4.2.6 Performance 4-10
4.3 COAL-LIQUID MIXTURES 4-14
4.3.1 Process description 4-15
4.3.2 Factors affecting performance 4-16
4.3.3 Applicability to industrial boilers 4-16
4.3.4 Development status 4-17
4.3.5 Reliability 4-20
4.3.6 Emissions data 4-20
4.4 COAL LIQUEFACTION 4-21
4.4.1 Process description 4-21
4.4.2 Factors affecting performance 4-25
4.4.3 Applicability to industrial boilers 4-27
4.4.4 Development status 4-29
4.4.5 Reliability 4-31
4.4.6 Emissions data 4-31
4.5 REFERENCES 4-37
-------
LIST OF TABLES
Table
Page
2.1-1 PREDOMINANT SODIUM SCRUBBING ABSORBER TYPES WITH THEIR
THEIR TYPICAL S09 REMOVAL EFFICIENCIES AND OPERATING
PARAMETERS f 2-10
2.1-2 TABLE OF SODIUM SCRUBBING SYSTEMS 2-13
2.1-3 POPULATION OF SODIUM SCRUBBERS BY APPLICATION 2-22
2.1-4 TOTAL S0? TREATED BY APPLICATION FOR CURRENT SODIUM
SCRUBBER SAMPLE 2-23
2.1-5 POPULATION OF SCRUBBERS ON UNITS FIRING OIL, COAL
AND OTHER FUELS 2-24
2.1-6 POPULATION OF WASTE DISPOSAL METHODS OF SODIUM SCRUBBERS 2-25
2.1-7 RECENT RELIABILITY DATA FOR SODIUM SCRUBBERS 2-31
2.1-8 EMISSIONS DATA USING EPA TESTING METHODS 2-33
2.1-8a AVERAGE RESULTS FROM SODIUM SCRUBBING SYSTEMS 2-34
2.1-85 S02 REMOVAL EFFICIENCIES BY ABSORBER TYPE 2-35
2.1-9 APPLICABILITY OF DUAL ALKALI SYSTEMS INSTALLED ON
INDUSTRIAL BOILERS 2-42
2.1-10 RELIABILITIES FOR DUAL ALKALI SYSTEMS 2-45
2.1-11 EMISSIONS DATA FOR DUAL ALKALI SYSTEMS USING EPA
TESTING METHODS 2-47
2.1-12 SUMMARY OF LIMESTONE SYSTEMS OPERATING ON U. S.
INDUSTRIAL BOILERS AS OF OCTOBER 1983 2-56
2.1-13 SUMMARY OF WET LIME FGD SYSTEMS INSTALLED ON U. S.
INDUSTRIAL BOILERS AS OF OCTOBER 1983 2-66
2.1-14 THIOSORBIC LIME APPLICATIONS TO UTILITY BOILER
FGD SYSTEMS 2-68
2.2-1 SUMMARY OF INDUSTRIAL BOILER SPRAY DRYING SYSTEMS 2-77
-------
2.2-2 SUMMARY OF EMISSION DATA FOR FOUR INDUSTRIAL LIME
SPRAY DRYING FGD SYSTEMS 2-80
3.1-1 PARTIAL SUMMARY OF COAL-FIRED INDUSTRIAL AFBC
BOILERS IN THE U. S 3-11
3.1-2 SUMMARY OF SO, EMISSIONS DATA FOR VARIOUS AFBC
CONFIGURATIONS 3-14
4.1-1 PREPARATION AND THERMAL DRYING OF BITUMINOUS COAL AND
LIGNITE BY STATE - 1978 (THOUSAND) SHORT TONS 4-5
4.2-1 CURRENT APPLICATIONS OF LOW AND MEDIUM BTU
GASIFICATION TECHNOLOGY 4-11
4.3-1 TEST EXPERIENCE WITH COM-FUELED PACKAGE WATERTUBE
BOILERS 4-18
4.3-2 INSTALLED AND ANNOUNCED DOMESTIC COM PLANTS 4-19
4.4-1 PROPERTIES OF SRC-II FUEL OILS AND COMPARABLE
PETROLEUM PRODUCTS 4-30
4.4-2 EMISSION RESULTS FOR SRC TEST BURN AT
PLANT MITCHELL 4-33
4.4-3 EMISSION RESULTS FOR DOE/PETC TESTS ON SRC 4-36
vn
-------
LIST OF FIGURES
Figure
Page
2.1-1 Process Flow Diagram for a Sodium Scrubbing System
Using Soda Ash Slurry Storage 2-3
2.1-2 Equilibrium partial pressure of SCL Over Aqueous
Sodium Sulfite Solutions 7 2-6
2.1-3 Limestone Process Flow Diagram 2-51
2.2-1 Typical Spray/Dryer Particulate Collection
Flow Diagram 2-71
2.2-2 Dry Alkali Injection Flow Diagram 2-82
2.2-3 E-beam/Ammonia Process Flow Diagram 2-86
3.1-1 Schematics of Traditional Dense-Bed FBC Power-Generation
Systems 3.3
3.1-2 Schematics of Two-Stage and Circulating-bed AFBC
Power-Generation Systems 3-7
3.2-1 Multistage Combustion in a Distributed Mixing Burner
(Top Half of Burner Only Depicted) 3-18
3.2.2 Fuel Rich Fireball Burner Design for Tangentially-Fired
Boiler 3_20
4.1-1 Kitt Mine Coal Preparation Plant - Hourly Incremental
Data for Sulfur Dioxide Emission Parameter 4-4
4.2-1 Low/Medium-Btu Gasification Process Options for
Supplying an Industrial Boiler Fuel Gas 4-8
4.4-1 Flow Diagram for SRC-I Demonstration- Plant 4-23
4.4-2 Flow Diagram for SRC-II Demonstration Plant 4-26
-------
1.0 INTRODUCTION
This document was prepared to provide the public and industry with
additional background information on the industrial boiler source category
in support of potential new source performance standards for sulfur dioxide
(SOp) emissions. The document is to be used as a supplement to the
Background Information Document for Industrial Boilers prepared for the
U. S. Environmental Protection Agency (EPA), Office of Air Quality Planning
and Standards by Radian Corporation in March 1982 and the series of
Individual Technology Assessment Reports (ITARs) for industrial boiler
applications prepared under the direction of EPA's Industrial Environmental
Research Laboratory at Research Triangle Park, N.C. The overall objective
of this report is to update the information and data contained in the
above-referenced reports as it relates to SOp emission control technologies.
To minimize duplication of material, this document assumes that the
reader is familiar with the earlier reports and makes liberal reference to
those reports. In the case of some S0? control technologies, the principles
of operation and factors affecting performance have changed to such a great
extent that a substantial re-write of the technology description was in
order. Where this is not the case, only supplemental information is
presented.
The S02 control technologies selected for examination and updating
are those which are either in current use by industrial boiler operators or
under active investigation in research and development programs. The
technologies are generally categorized as post combustion control approaches
(Section 2.0), combustion modification control approaches (Section 3.0), and
fuel pretreatment control approaches (Section 4.0). Each technology is
discussed and evaluated from the standpoint of process principles, factors
affecting performance, applicability to industrial boilers, development
status, operability and reliability, emissions data, and process economics.
1-1
-------
The process economics sections of this report contain information
describing the impacts of process design and operating parameters on system
costs. Direct comparisons of capital and annual costs for the technologies
judged to be most applicable to industrial boilers for S0? emissions control
are contained in the S00 Model Boiler Cost Report.
1-2
-------
2.0 POST COMBUSTION CONTROL APPROACHES
Post combustion techniques for controlling SOp emissions from
industrial boilers are discussed in this section. These techniques remove
S02 from flue gases produced from fuel combustion. Post combustion
techniques have been divided into wet and dry processes according to the
final form of the recovered SO^ product.
2.1 WET SCRUBBING PROCESSES
Wet flue gas desulfurization (FGD) processes use an alkaline solution
or slurry to absorb S02 from boiler flue gas. The absorbed S02 exits the
system either as a liquid waste stream or as a semi-solid waste sludge. The
wet FGD processes discussed here are:
o Sodium
o Dual alkali
o Limestone
o Lime.
Each of these technologies is currently being used commercially to remove
SOp from industrial boiler flue gases.
2.1.1 Sodium
Sodium scrubbing comprises approximately 98 percent of all industrial
wet FGD installations and is treating roughly 80 percent of all the S0~
treated by wet FGD scrubbers. If oil field generators are excluded from the
industrial boiler population, then sodium scrubbers represent about 80
percent of the total industrial boiler wet FGD system population. The
predominance of sodium scrubbers is primarily because of their ease of
operation and their reliability, which is reported to be about 98 percent on
average (see Section 2.1.1.5).
S02 removal efficiencies for sodium scrubbers have been consistently
high. For tests conducted on 45 scrubbers using EPA testing methods, the
2-1
-------
average S02 removal efficiency was 96.2 percent with a standard deviation of
2.9 percent.
Sodium scrubbing is also the most economical of the wet FGD systems for
most industrial boiler applications. For high sulfur coals, the total
annualized costs (TAG) for sodium scrubbing systems become greater than
TAC's for dual alkali systems in the 150 - 250 x 106 Btu/hr range. For low
sulfur coals, this range is much beyond 400 x 106 Btu/hr, so that for all
practical purposes sodium scrubbing is less expensive than dual alkali
scrubbing for low sulfur industrial boiler applications. The predominant
factors affecting total annualized costs are reagent and liquid treatment
costs, which together comprise 65-85 percent of the total operating costs.
Sodium scrubbing capital costs including wastewater treatment represent
between 35 and 50 percent of dual alkali capital costs.
The March 1982 Background Information Document (BID) for Industrial
Boilers assumed that sodium scrubbing use would be significantly limited by
wastewater regulations. However, in many areas of the country, the
wastewater stream is already being permitted by the local water authorities.
Because of the assumption that the water regulations would be strict, it was
predicted that treatment and disposal of the wastewater would be
prohibitively expensive. It was therefore assumed that this technology
would be applied only to those few plants that had either an inexpensive
reagent source or a readily available disposal mechanism, or both. However,
only about 20 percent of the plants currently using sodium scrubbers can be
grouped into this category, indicating that wastewater treatment and
disposal is not prohibitively expensive. For a further discussion of the
wastewater issue, the reader is referred to Section 2.1.1.3.
2.1.1.1 Process Description. The following discussion includes
additions and updates to the March 1982 BID and repeats only that
information which is considered essential for the the discussions in the
other sodium scrubbing sub-sections. A simplified sodium scrubbing process
flow diagram is presented in Figure 2.1-1 to replace the one presented in
the BID. Sulfur-dioxide is absorbed from boiler flue gases into an aqueous
2-2
-------
To Stack
Absorber
Flue Gas C
r
ro
i
CO
Make-up water
Saturated NagC03
Solution
Slurry
Storage
Tank
Feed
Water
\ /
V
/ \
\J
Recirculation
Tank
Recycle Stream
Aqueous Waste To
Treatment And Disposal
Figure 2.1-1.
Process flow diagram for a sodium scrubbing
system using soda ash slurry storage.
-------
solution in the scrubber. The scrubber effluent flows (usually by gravity)
to a recirculation tank where it is mixed with make-up reagent and water.
If the reagent is in the form of NaOH, it is typically added as a 50 weight
percent solution. If the reagent is Na0CO,, it is usually added as a
12
saturated solution. ' The aqueous reactions that take place in the
scrubber and, to a greater extent, in the recirculation tank are:
2NaOH + S02 >* Na2S03 + H20 (2.1-1)
or
Na2C03 + S02 >> Na2S03 + C02 (2.1-2)
and
Na2S03 + S02 + H20 ^ 2NaHS03 (2.1-3)
S02 + H20 ^ H2S03 (2.1-4)3
Na2S03 + iO? ^ Na?SO, (2.1-5)
C. J L, •.-mui_^™n«™_^^^. £ ^
The residence time in the tank is typically three to four minutes. The
aqueous solution leaving the recirculation tank contains primarily NaOH,
Na2C03, Na2S03, NaHS03, H2S03, and Na2S04. Most of this stream is recycled
to the scrubber, while a small fraction is bled for treatment and disposal.4
This wastewater stream may be treated on-site by oxidation to reduce
chemical oxygen demand (COD) and to reduce the potential for SO-
re-emissions. This stream may also be allowed to settle in order to filter
out fly ash and other insoluble compounds. Disposal of the wastewater
stream is handled in one of several ways: evaporation ponding, deep-well
injection, or discharge to a sewer, river, or ocean.
The system's operation is monitored by the specific gravity and pH of
the recirculation tank. In some systems, the specific gravity is controlled
by the addition of make-up water. It determines both the buffering capacity
of the scrubbing solution and the flow rate of the blowdown stream. The
higher the specific gravity, the greater will be the buffering capacity of
the solution and the lower will be the blowdown flow rate. The pH is
controlled by the addition of sodium reagent. If, for example, the process
experiences a transient increase in S02 loading, then the pH in the
2-4
-------
recirculation tank will drop. This, in turn, signals the addition of
make-up reagent to re-establish the pH to normal. Make-up water and
blowdown flow rates will then both be increased to maintain the set-point
value of the specific gravity.
2.1.1.2. Factors Affecting Performance. The major operating variables
affecting scrubber performance are the pH and total sulfite concentration
(TSC) of the scrubbing solution. The pH primarily affects SCL removal
efficiency while TSC affects this as well as reagent consumption (for those
systems not using an oxidation system for wastewater treatment) and
transient performance. Other variables affecting scrubber performance are
absorber type and L/G ratio. Each of these factors will be discussed in
this section.
£H
The pH of the scrubbing liquor is determined primarily by the ratio of
Na2S03 and NaHSOj. Since HS03" is a weak acid (with a pKa of 7.45 at 50°C),
the greater the Na2S03/NaHS03 ratio is the higher the pH of the scrubbing
liquor will be. According to Figure 2.1-2, raising the pH will lower the
equilibrium S07 partial pressure of the scrubbing liquor which will in turn
67
increase the driving force for SCL absorption. ' This means that if all
other design and operating parameters are held constant, increasing the pH
of the scrubbing solution will increase the SCL removal efficiency of the
scrubbing system.
The pH of the scrubbing solution is controlled simply by adding reagent
to the recirculation tank (see Figure 2.1-1). Typically, the pH of the
scrubbing solution is maintained around 7.0, which means that the
o
NagSO-j/NaHSO-j ratio is approximately 1:2. At this pH, the equilibrium
partial pressure is less than 20 ppmv for most sodium scrubbing solutions.
Since inlet concentrations of S02 range from 1,000 to 3,000 ppmv, the
theoretical SO^ removal efficiency is greater than 95 percent. Due to the
reactiveness of dissolved SOp in aqueous sulfite solutions and the mass
transfer capabilities of most absorber designs, these equilibrium values are
2-5
-------
Log(S02)
(ppmv)
ro
cr>
4.0-1
3.0-
2.Q_
1.0-
0 -
-1.0-
-2.Q.
3.0
4.0
5.0
6.0
PH
0.1 Molar
01 Molar
001 Molar
7.0
Figure 2.1-2. Equilibrium Partial Pressure of S02 Over Aqueous
Sodium Sulfite Solutions?
-------
approximated in practice. Commercially operating systems have consistently
reported S02 removal efficiencies greater than 95 percent (see
Section 2.1.1.6).
Total Sulfite Concentrations (TSC)
The total sulfite concentration (TSC) is defined as the sum of all
_2
SCU and its corresponding cations. For the sodium scrubbing liquor, this
?
includes primarily Na2SO, and NaHSCL. It should be noted that SO, will
also be dissolved in the scrubbing solution typically in a ratio of 1:3
Q
relative to the sulfite species. Sodium sulfite and sulfate are the
primary dissolved species and together comprise the total dissolved solids
(TDS) of the scrubbing solution. Sulfate is a very stable species and has
little effect on scrubber performance except when it becomes so concentrated
that it can promote precipitation of the sulfite species and significantly
reduce SCL removal efficiency.
Figure 2.1-2 shows that as the TSC increases, the equilibrium SCL back
pressure will also increase. For example, using the range observed for
commercially operating systems, 0.01M to 1.7M, the S02 partial pressure will
vary from 0.11 ppmv to 19 ppmv within this range at a pH of 7.0 and at 50°C.
Assuming that all other operating and design parameters remain constant, the
SOp removal efficiency will theoretically decrease as TSC increases.
However, when compared to inlet SOp partial pressures of 1,000 to
3,000 ppmv, this 170-fold change in equilibrium exit partial pressure does
not significantly affect the overall S02 removal efficiency. This fact has
been substantiated by commercially operating systems which have shown no
trend in S02 removal efficiency as a function of TSC.
Although increasing TSC may reduce S02 removal efficiency by a small
degree, it can significantly improve transient performance by stabilizing
pH. Since HS03" is a weak acid, NaHS03 and Na,,SO, serve as a buffer in
the scrubbing solution. The higher their concentrations are, the greater
the buffering capacity of the scrubbing liquor will be. Scrubbers operated
in the concentrated mode (conventionally defined as TDS levels exceeding
five weight percent) will typically have an inlet pH of 7.0 - 7.5 and an
2-7
-------
8 12
outlet pH of 6.5 - 7.0. ' On the other hand, scrubbers operated in the
dilute mode (conventionally defined as IDS levels less than or equal to 5
weight percent) have an inlet pH of 9-10 and an outlet of pH of 4-5.12
Buffering is important because it increases reliability and improves
transient performance. At pH's above 8.0, the likelihood of calcium scaling
is high. Within most sodium scrubbing loops there is some background Ca+2
(e.g. from make-up water or ash leachate) which will combine with available
sulfite and sulfate ions. At pH's above 8.0, CaSO-, and CaSO, will
precipitate out of solution and cause scaling. Sometimes this scaling will
lead to plugging, especially in the recirculation lines and spray nozzles.13
As a result, the S02 removal efficiency can be impaired; in the extreme
cases, the unit will have to be shut down and cleaned. At low pH's,
substantial corrosion of the scrubber, tank, and pipe internals can occur,
especially if the scrubbing solution has a high chloride ion concentration.
This corrosion will increase the maintenance costs of the scrubbing unit and
decrease the scrubber's reliability.
Buffering serves another function in that it helps to prevent large pH
fluctuations from occurring, even when inlet S02 concentrations vary
dramatically, as is typical of industrial boiler operation. This insures
relatively constant outlet S02 concentrations or, in other words, good
transient S02 removal performance.
Absorber Design
The design of an absorber determines i'ts mass transfer characteristics
and thus S02 removal capabilities. Absorber designs can be grouped into
three categories: open vessels, vessels with internals, and combinations of
the two. Open vessel absorbers, such as venturi scrubbers, spray towers,
and liquid jet eductors, rely on a combination of high gas- and liquid-side
pressure drops to provide adequate mass transfer. Vessels that have
internals, such as packed beds and tray towers, rely primarily on solid
internal surface area for absorption. Combination absorbers, such as disc
and donut contactors and spray baffles, use a combination of the
characteristics employed by both open vessels and vessels with internals.
2-8
-------
Table 2.1-1 summarizes SCL removal efficiencies for these three categories
of absorbers, as reported by plants, vendors, and governmental agencies for
approximately 290 scrubbers. Included also in Table 2.1-1 are the typical
values for the absorbers' gas- and liquid-side pressure drops as well as
their typical liquid-to-gas ratios (L/G's). A high gas-side pressure drop
assures adequate mixing and a high liquid-side pressure drop assures not
only an adequate liquid-flow rate but sufficient atomization as well. The
L/G will be discussed further under its own sub-heading, and the
applicability and reliability of each absorber type will be discussed in
their respective sections.
Table 2.1-1 provides vendor, plant and government data for
approximately 290 sodium scrubbers. The average S02 removal efficiency for
the seven scrubber types in the table was 93.9 percent. The standard
deviation was 3.9 percent. (Actual emissions as determined by EPA testing
methods alone are provided in Section 2.1.1.6). As shown by the table, the
open vessel category has reported the second highest SOp removal efficiency
of the three absorber categories. For the approximately 115 scrubbers
within this category, the average S02 removal efficiency was 93.3 percent.
In general, open vessels have high liquid-side pressure drops to atomize the
scrubbing solution. Atomization produces small droplets with high surface
area/volume ratios. The gas-side pressure drop varies a great deal within
this category. Spray towers, for example, have very low gas-side pressure
drops to provide a low velocity gas. This low velocity provides high
residence times and prevents re-entrainment of the liquor droplets. It
should be noted that these low velocities may produce laminar flow in small
diameter towers. Therefore, SO,-, removal efficiencies may be low for small
17
spray tower systems. Venturi scrubbers, on the other hand, have
relatively high gas-side pressure drops to cause turbulent mixing. In these
scrubbers, S02 removal efficiency improves as gas velocity increases.
The average SO^ removal efficiency for approximately 50 tray towers was
90.6 percent. Tray towers, too, operate with some atomization, and
therefore the liquid side pressure drop is moderate. Gas-side pressure
drops are relatively high because of the trays. Towers with two trays are
2-9
-------
TABLE 2.1-1 PREDOMINANT SODIUM SCRUBBING ABSORBER TYPES WITH THEIR TYPICAL SO,, REMOVAL EFFICIENCIES AND OPERATING PARAMETERS
ro
i
o
>^1 SO, REMOVAL EFFICIENCY (%) TYPICAL OPERATING PAfiAMFTFR<;d
Absorber Absorber bultur Range
Category Type (Mt. %) Actual* Guaranteed6
Open Venturl scrubbers 1.0-2.5 91.6f(5.3)9 NAh
Vessels
Spray towers 1.1-1.7 88.9(4.9) 90
Liquid jet eductors 1.0-2.4 94.2(3.3) 91.9
Vessels Tray absorbers
with (2 stage) 0.7-5.5 75-90 (NA) NA
Internals (3 stage) 00.6(3.3)
Packed bed 1.7 73.0(DNA)' HA
Combination Spray baffle 0.6-6.0 96.6(0.8) 95.0
Disc and Oonut 1.0 95.0(DNA) 95.0
contactor
93.9J(3.9)
Gas-side AP Liquid-side AP L/G Applicability*
Theoretical (in. H20) (psig) (gal/1000 acf)
90-95 8-20 25 10
95-99 1.5-2.5 75 30-50
90-99 0.5-1.0 40 50-120
«« 8-12 30 20
95-97
95-99 2 nominal 1-10
95-99 5 30 30
95-99 5 NA 10-20
7 n
3.4
29
17
11
32
0.7
Actual values are those reported by vendors, plants and governmental agencies.5
Guaranteed values represent those guaranteed by vendors.5
Theoretical values are those reported in literature.'3"'5
Typical operating parameters are those compiled from vendors, plants,
Applicability represents the percent that each absorber type comprises
population. The sample population Is located in Table 2.1-2 and lists
and literature in References 13-28.
of the total 292 scrubbers that specified absorber types
356 scrubbers In all.
in the sample
9These values denote standard deviations as determined in Reference 3.
h.
NA = Not Available
J4w»/ °°™ "Ot aPP]y- ,There was »nl> one source <>f information and therefore calculation of the standard deviation was meaningless
Average SO^ removal efficiencies for all scrubbers except for packed beds. meaningless.
-------
not expected to achieve SCL removal efficiencies as high as the other
29
absorber types. Only one source reported using a packed bed and this
30
obtained 73 percent removal. However, as shown by the theoretical values,
it is expected to have a much higher S02 value. Since the data were taken
in 1973 when the emission limits were less-than-stringent, it is believed
that this figure does not represent the performance of packed beds under
typical operating conditions.
Those absorbers combining the features of the previous two categories
have demonstrated the highest S0? removal efficiencies. For the
approximately 90 spray baffles within the sample, an average of 96.6 percent
S09 removal efficiency has been achieved with a standard deviation of less
L c
than one percent. For the two disc and donut contactors, the efficiency
has averaged 95 percent. The spray baffle absorber combines a unique
spiral mixing technique along with a moderate liquid-side pressure drop for
atomization and a moderate gas-side pressure drop. In addition, it
provides solid surface area on the baffles themselves for mass transfer.
The disc and donut contactor provides solid surface area as well. Mixing,
•31
too, is achieved with a baffle-type arrangement.
Liquid- to- ga_sr ratio
The liquid-to-gas ratio (L/G), measured in gal/103 acf, also affects
S02 removal efficiencies. In general, as the L/G is increased the S0?
removal capabilities will increase up to the flooding point of the
scrubber. This depends on the type of absorber and the degree to which
the gas- and liquid-side pressure drops are increased to accommodate the
increased liquid flow. Once the desired S02 removal efficiency is specified
for a particular absorber, the L/G is set as well. Table 2.1-1 provides
typical L/G's for the seven predominant absorber types for S0? removal
efficiencies at or above 90 percent. Although there are exceptions, in
general, vessels with internals require the lowest L/G, followed by
combination type absorbers, and then by open vessels. In addition, those
absorbers using greater degrees of atomization will, in general, require
higher L/G's than those that don't.
2-11
-------
It should be noted that L/G's are lower for sodium-based systems than
for calcium-based systems. This is primarily because sodium is much more
soluble in water than calcium and thus requires less water for dissolution.
As shown by Table 2.1-1, except for liquid jet eductors, L/G's for
sodium-based systems range from 1-50 gal/103 acf. For calcium-based systems
(exclusive of Thiosorbic lime and Thiosorbic limestone), they range from 60
to 120 gal/103 acf.
2.1.1.3 Applicability to industrial boilers. The March 1982 BID indicated
that sodium scrubbing was applicable only to very few types of plants.
These plants had either an inexpensive reagent source or a readily available
disposal technique, or both. Installation at other industrial boiler sites
was assumed to be limited because of the predicted zero discharge
requirement and because of the prohibitive costs of treating the waste
stream (see Sections 4.2.1 and 7.2.1 of the BID). However, about 98 percent
of the estimated 680 wet FGD systems installed on industrial boilers are
sodium scrubbing systems (see Table 2.1-2).5 They represent at least
80 percent of all industrial boiler wet FGD systems that treat flue gas from
boiler size equivalents (BSE)* greater than 100 x 106 Btu/hr. Even after
eliminating what the BID considered as "special" applications (such as oil
field steam generators, paper mills, soda ash, and textile plants) sodium
scrubbers represent at least 70 percent of the total wet FGD systems
currently operating. These data suggest that the assumptions set forth in
the BID and ITAR should be investigated and revised where necessary. A
review of some of these assumptions is presented below.
Zero Discharge Requirement
The previous analysis assumed that zero wastewater discharge
requirements would restrict the use of sodium FGD systems. In some areas of
the country this will be the case. For example, in California and in the
*
Boiler size equivalent is used instead of boiler size because in many cases
more than one boiler is ducted to a common scrubber.
2-12
-------
TABLE 2.1-2. TABLE OF SODIUM SCRUBBING SYSTEMS
ro
i
Company/
Location
General Motors
St. Louis, MO
Dayton, OH
Tonawanda, NY
Pontiac, MI
St. Regis Paper
Cantonment, FL
Texaco
San Ardo, CA
Santa Monica, CA
American Thread
Marion, NC
Mobil Oil
San Ardo, CA
Kern Co. , CA
Buttonwillow, CA
McKitterick, CA
Start-up
Date
1972
1974
1975
1976
1973
1973
1979
1982
1980
1974
1974
1981
1981
1981
1981
1982
1982
1982
198?
1982
1979
1979
1980
Vendor
A.D. Little
Entoleter
FMC
GM
Neptune/Airpol
Ceilcote
Ducon
Andersen 2000
Thermotics
U.W. Sly
Manufacturing
In-house
MA
NA
NA
NA
NA
HA
NA
NA
Healrr
Technology
Heater
Technology
Heater
Technology
Absorber
Type3
TA
VG
VS
TA
NA
PB
ST
SB
NA
TA
TA
NA
NA
NA
NA
NA
NA
NA
NA
NA
LJE
LJE
LJE
Fuel
Typeb
C
C
C
C
B.O
0
0
0
0
C
0
0
0
0
0
0
0
0
0
0
0
0
0
%S 1
3.2
0.7-2.0
1.2
0.8
<1.0
1.7
1.7
3.2
3.5
<1.0
2-2.5
1.66
1.61
1.39
1.65
1.56
1.58
1.47
1.0
1.0
1.1
1.1
1.1
Boiler Size
Equivalent
[10° Btu/hr)
313
82
84
188
250
375
70
200
30
75
50
NA
108
22
50
50
22
50
50
22
50
22
23
27.5
64
50
50
Number of %SO, Removal*1
FGD Units Actual Guaranteed
2
1
4
1
1
1
25
2
5
3
17
1
2
20
8
1
1
1
1
1
1
1
1
1
7
7
2
90
86
90
NA
80-90
73
95
97
98
70-90
90
95.8*
94.4*
96.8*
90.7*
87.5*
89.7*
93.9
91*
89*
85
95
96
NA
NA
NA
NA
NA
NA
NA
95
NA
None
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
90
96
Waste Disposal Method
Oxidation/neutralization/
sewer
Clarify/adjust PH/sewer
Aeration/treatment
Dewatered/landfill
plant
Clar i f ication/aeration
NA
NA
NA
NA
pond
pond
NA
NA
NA
NA
NA
NA
NA
NA
NA
pond
NA
NA
-------
TABLE 2.1-2. TABLE OF SODIUM SCRUBBING SYSTEMS (Continued)
ro
i
Company/ Start-up
Location Date
Taft, CA
Bakersfield, CA
Bel ridge. CA
Georgia Pacific
Crosett, Ak
Great Southern Paper
Cedar Springs, GA
ITT Rayonier
Fernandlna Beach,
FL
Mead Paperboard
Stevenson, AL
Husky Oil
San Ardo, CA
Texasgulf
Granger, WY
Nekoosa Paper
Ashdovm, AK
FMC
Green River, MY
Alyeska Pipeline
Valdez, AK
Getty Oil
Bakersfield, CA
Santa Maria, CA
McKittrick, CA
1980
1980
1980
1981
1982
1975
1975
1975
1975
1976
1976
1976
1976
1977
1977
1979
1977
1980
Vendor
Heater
Technology
Heater
Technology
Heater
Technology
Neptune/
Airpol
Heptune/
Airpol
Neptune/
Airpol
Neptune/
A i rpo 1
Heater
Technology
Swemco
Neptune/Airpol
FMC
FMC
FMC
In-house
In-house
In-house
Heater
Technology
Absorber
Type
LJE
LJE
LJE
LJE
LJE
VS
VS
VS
TA
LJE
TA
VS
DO
NA
TA
TA
NA
NA
LJE
Fuel
Typeb
0
0
0
0
0
B.C.O
B.C.O
B.O
0
0
C
C
C
0
0
0
0
0
tt
1.2
1.2
1.0
1.0
1.2
1.5-2
1-2
2-2.5
1.5-3
1.4
0.75
1-1.5
1.0
.0.1
1.1
1.1
4.0
1.1
Boiler Size
Equivalent
(10b Btu/hr)
25
25
50
50
50
25
650
1650
400
340
100
25
340
700
400
800
NA
300
450
500
NA
75
Number of «SO,, Removal*1
FGD Units Actual Guaranteed
2
8
3
10
3
2
1
2
1
1
1
1
2
2
1
1
1
1
5
4
1
1
96
95
95
80
85-90
80-85
95
85
90
90
95
96
90
96
96
94
96
95
90
90
NA
NA
NA
NA
80
NA
NA
95
NA
NA
NA
NA
NA
OR
y J
Waste Disposal Method
NA
NA
NA
Sewer
Pond
Wastewater treatment/pond
To digester In pulping
process
NA
Deep well injection
pond
To digester in pulping
process
Salt pond
Oxidation/dilution
Deep well injection
Deep well Injection
Pond
NA
Ml
lltt
pond
pond
-------
TABLE 2.1-2. TABLE OF SODIUM SCRUBBING SYSTEMS (Continued)
ro
Company/ Start-up
Location Date
Fellows, CA
Taft, CA
Kern Co. , CA
Union Oil
Guadalupe, CA
Kern Co. , CA
Taft. CA
McKittrick, CA
Bel ridge Oil
McKittrick, CA
Elf Aquttaine
Kern Co., CA
Kerr-McGee
Trona, CA
Chevron
Bakersfield, CA
Maricopa, CA
Kern Co. , CA
1980
1982
1982
1978
I960
1978
1979
1980
1981
1978
1979
1978
1983
1978
1978
1979
1980
1979
1980-
1982
1982
1982
1982
Absorber
Vendor Type
Heater
Technology
Andersen 2000
NA
Andersen 2000
Heater
Technology
Heater
Technology
Andersen 2000
Heater
Technology
Heater-
Technology
Andersen 2000
Heater
Technology
CE Mdtco
Andersen 2000
Andersen 2000
CEA
Koch
Koch
Neptune/Airpol
Heater
Technology
Andersen 2000
NA
NA
NA
LJE
SB
NA
SB
LJE
LJE
SB
LJE
LJE
SB
LJE
LJE
ST
SB
SB
TA
TA
TA
NA
LJE
SB
NA
NA
NA
Fuel
Type1 *S
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
l.l
1.1
1.0
2.8
2.4
2.4
1.1
1.1
%
1.1
0.7-1.2
1.5
1.1
1.1
1.1
1.1
0.7-5.5
1.1
1.1
NA
1.1
1.3
0.8
1.04
1.2
Boiler Size
Equivalent
(10e Btu/hr)
75
37.5
375
50
25
50
50
25
50
NA
50
50
50
50
50
750
130
300
50
50
50
437.5
462.7
NA
Number of
FGD Units
1
2
2
2
2
1
5
1
7
4
1
2
1
1
1
2
3
2
11
1
7
1
1
1
%so2
Actual
95
96
98*
98
90
96
96
95
96
95
95
90
90
96
96
98+
90
90
95
96
97
96*'
96*f
97*'
Removal
Guaranteed
90
95
NA
95
80
95
95
90
95
NA
80
NA
NA
95
95
NA
NA
NA
NA
95
95
NA
NA
NA
Waste Disposal Method
NA
Hauling to
NA
Hauling to
NA
NA
Hauling to
NA
NA
NA
secure site
secure site
secure site
NA
Haste water treatment/pond
NA
Hauling to
Hauling to
Salt pond
Pond/waste
Pond/waste
NA
NA
NA
NA
NA
NA
secure site
secure site
treatment
treatment
-------
TABLE 2.1-?. TABLE OF SODIUM SCRUBBING SYSTEMS (Continued)
Company/
Location
Start-up
Date
Vendor
Absorber
Type3
Fuel
• " ' !• ' ~-Tm-
Type
%s
Boiler Size
Equivalent
(10° Btu/hr)
Number of JS00 Removal d
FGD Units Actual1- Guaranteed Waste Disposal Method
General American Oil
Taft, CA
Gulf Oil
Lost Hills, CA
Kern Co., CA
Sun Production Co.
Fellows, CA
Newhall, CA
Olldale, CA
Phillip Morris
Chesterfield, VA
Tenneco Oil
Bakersfleld, CA
Green River, HV
Shell Oil Co.
Coalinga, CA
Bakersfleld, CA
1978
198?
1978
1982
1979
1979
1979
1979
1979
J982-
1983
1982
1980
1981
1981
1983
Andersen 2000
Andersen 2000
Andersen 2000
Andersen 2000
Andersen 2000
U Jl
NA
Andersen 2000
CE Natco
CE Natco
CE Natco
Flakt
Andersen 2000
Andersen 2000
NA
Flakt
Oucon
Ducon
Npptune/Airpol
NA
UA
nM
HA
MA
ni\
SB
SB
SB
SB
SB
NA
SB
ST
ST
ST
ST
SB
SB
NA
NA
NA
NA
NA
NA
NA
NA
NA
0
0
0
0
0
0
0
0
0
0
C
0
0
0
C
0
0
0
0
0
0
0
1.2
1.2
1.2
0.6
1.0
1.2
1.3
1.4
1.3
1.2
NA
1.0
1.0-1.6
1.04
1.5
0.6
0.6
1.1
0.80
0.80
0.79
0.79
50
75
50
55.2
25
30
25-50
50
NA
50
237
50
50-150
261.8
300
NA
NA
50
100
180
200
100
100
52
50
1
1
2
1
1
I"
1
2
2
1
5«
1
2
2
1
2
2
98
98
97
95*
99.5*
93.5*
97
85
85
85
NA
99
96-98
99.4*
93
90
90
96.4
99.2*
98.6*
98.8*
99.8*
95
96
95
NA
NA
NA
95
NA
NA
NA
90
95
95
NA
NA
NA
NA
NA
NA
NA
NA
NA
Hauling to
Hauling to
Deep well
NA
NA
NA
Hauling to
Pond
NA
Pond
Aeration
Hauling to
Hauling to
NA
Pond
NA
NA
NA
NA
NA
NA
NA
secure site
secure site
Injection
secure site
secure site
secure site
Kernrldge Oil
McKlttrlck, CA
1980
Andersen 2000
SB
1.2
50
95
95
NA
-------
TABLE 2.1-2. TABLE OF SODIUM SCRUBBING SYSTEMS (Continued)
Company/ Start-up
Location Date
Kern Co. , CA
Santa Fe Energy
Fellows, CA
Coalinga, CA
Bakersfleld. CA
Grace Petroleum
Pismo Beach, CA
Sun Oil
Kern Co. , CA
Miles Labs
Clifton, NJ
Struthers Oil & Gas
Red Lodge, MT
St. Joe Paper
Port St. Joe, FL
Ami noil
Hunt) tig ton Beach,
Occidental Petroleum
Taft, CA
Exxon
Kern Co. , CA
Petro Lewis
Kern Co. , CA
1982
1980
1981
1982
1982
1982
1980-
1981
1981
1981
1982
1982
CA
1982
1982
1982
1983
Vendor
CE Natco
Due on
CE Natco
CE Natco
CE Natco
CE Natco
Heater
Technology
Heater
Technology
Heater
Technology
NA
NA
Thermotics
Andersen 2000
Andersen 2000
Andersen 2000
Neptune/Alrpol
Andersen 2000
Andersen 2000
Andersen 2000
NA
CE Natco
Thermotics
Absorber
Type
VS
NA
NA
VS
NA
NA
LJE
LJE
LJE
NA
NA
NA
SB
SB
SB
NA
SB
SB
SB
NA
NA
NA
Fuel
Type
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
B,0
0
0
0
0
0
0
*S
1.0
1.1
0.85
1.10
1.10
1.0
1.2
0.8
0.8
1.52
1.52
1.2
1.1
2.8
2.2
NA
2.0
1.1
1.2
0.8
1.15
1.02
Boiler Size
Equivalent
(10° Btu/hr)
62.5
125
62.5
125
62.5
NA
50
25
50
50
50
50
NA
25-100
42.8
17.8
NA
50
50
25
62.5
62.5
62.5
Number of
FGD Units
9
1
1
1
1
1
6
1
5
10
1
1
4
41
1
1
1
13
2
1 I
1
1
1
*S00 Removal*1
Actual Guaranteed
96*
96*
97*
99*
96*
94*
96
96
96
97.5*
96.2
98
96
96
98
NA
96
97
97
95*{
98*T
NA
NA
NA
NA
NA
NA
95
95
95
NA
NA
NA
95
95
95
NA
95
95
95
NA
NA
NA
i
Waste Disposal Method
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
Hauling to secure site
Sewer
NA
NA
NA
NA
NA
NA
NA
NA
-------
TABLE 2.1-2. TABLE OF SODIUM SCRUBBING SYSTEMS (Continued)
r , «-.. Boiler Size
Company/ Start-up Absorber Fuel. Equivalent
Location Date Vendor Type3 Type" XS (10G Btu/hr)
USS Tenneco Chem 1983 Andersen 2000 SB OW NA 100
Pasadena, TX
Tosco Petroleum 1983 Andersen 2000 SB Oil ?«;
Kern Co., CA
Marathon Oil 1983 Andersen 2000 SB 012 50
Kern Co. , CA
Garwood Paper 1983 Andersen 2000 SB 042 50
Garwood, NJ
American PetroHna 1983 Andersen 2000 SB 06 240
Big Spring, TX "U
Bradford Dyeing 1983 Andersen 2000 SB 04 214
Assoc.
Westerly, RI
Optimum Energy 1983 Andersen 2000 SB 0 1.2 300
Kern Co. , CA
Number of ISO Removal*1
FGD Units Actual Guaranteed Waste Disposal Method
2 NA 95 NA
1 95 NA NA
1 97 95 Hauling to secure site
1 98 95 Oxidation/sewage
1 NA 96 Oil field injection
1 NA 95 NA
2 NA 96 Hauling to secure site
VA = vane gage. ..... ------ '
C = coal; 0 - oil; B - bark; PC = petroleum coke; OW = organic waste.
Boiler size equivalent represents the heat load applied to each scrubber.
spray baffle; VS = venturl scrubber; ST = spray tower; PB = packed bed;
eNA = not available.
Calculated assuming HHV = 18,500 Btu/lb.
j|lt is assumed that a 50, 100 and 150 MM Btu/hr boiler are cached served by 3 FGD units.
^lt is assumed that there are 2 25 MM Btu/hr units and 3 50 MM Btu/hr units each served by one scrubber.
It is assumed that there are 1 25 MM Btu/hr unit, 2 50 MM Btu/hr units, and 1 100 MM Btu/hr unit each served by one scrubber.
-------
Southwest where the surface and ground water supplies have a high background
salinity and where crops are especially sensitive to high IDS water, sodium
blowdown streams can not, in most cases, be discharged into the local water
32
systems either directly or via a local sewage plant. Similarly, it is
doubtful that the blowdown could be discharged in North Dakota's water
33
supplies since they already have high background sulfate concentrations.
In these cases, the zero discharge requirement would probably have to be
met, not because of national regulations but because of local restrictions.
Implicit in the assumption that the sodium scrubber wastewater would
not be allowed to be discharged to surface waters and sewers are the
assumptions that the aqueous waste from sodium scrubbers significantly
impacts the quality of the receiving water body and that the zero discharge
laws which were being considered in the late 1970's would become a reality
by 1985. However, two of the waste stream characteristics which were
previously thought to be deleterious already comply with current water
regulations or can be made to do so with minimal treatment. The typical pH
levels of sodium waste streams comply with the water quality standards set
forth in the 1976 Quality Criteria for Water and thus do not require
35
neutralization. The blowdown stream's chemical oxygen demand (COD), due
to the oxygen scavenger NapSCU, can be reduced adequately by aeration.
The only potential secondary pollution concerns associated with the
sodium scrubbing wastewater stream is its total dissolved solids (TDS)
concentration and its trace metals concentration. In most industrial
plants, the scrubber wastewater is diluted by combining it with other
wastewater streams produced in the plant. The total plant effluent is not
treated for TDS and is generally not treated specifically for trace metals.
However, it is very likely that many of the trace metals will precipitate
out as metal hydroxides and be removed along with the other suspended solids
in the plant effluent. Nevertheless, in most cases the receiving water
body, whether it is a sewer or river, has the capacity to dilute the
scrubber wastewater that is discharged with the plant effluent to such a
degree that water quality impacts are negligible. If the impacts had been
significant, then the local water authorities would not have issued permits
2-19
-------
for the various sodium scrubbing streams that are currently being
discharged. Information collected in this update show that 14 plants using
sodium scrubbers are known to discharge their wastewater either to a sewer
or directly to a river. One sodium scrubbing vendor claims to know of 30
such plants that discharge to a sewer.
That the zero effluent discharge has not become a reality is
substantiated by these numerous plants that are discharging to surface
waters. Currently, there are no plans by the Effluent Guidelines Division
to regulate the sodium scrubbing wastewater itself. Furthermore, a
similar stream, the scrubber sludge stream, is not specifically regulated
under the effluent standards for the Steam Electric Point Source Category.38
One further secondary pollution impact that might limit the use of
sodium scrubbers but was not discussed in the BID or ITAR is the potential
for S02 re-emission due to the back pressures illustrated in Figure
2.1-2. ' The concern over S02 re-emissions is that the overall S02 removal
efficiency of sodium scrubbers can be substantially reduced. An extensive
study of this issue has shown that this potential can, for all practical
purposes, be eliminated by converting all of the sulfite to sulfate in a
well-operated oxidation tank. Conversion to the very stable sulfate will
also minimize reduction of sulfur to hydrogen sulfide gas in sewer
39
systems.
Presentation and Analysis of Applicability Data
Table 2.1-2 presents a list of approximately 360 sodium scrubbers,
representing approximately half of the estimated 670 sodium scrubbers in use
today. This figure of 670 was derived from one vendor's approximation of
•3 C
its market share and the number of units it had sold. Specifically, this
vendor estimates that its 233 systems represent 30-40 percent of the sodium
scrubbers applied to industrial boilers. Although this claim alone should
not be used in determining the total number of sodium scrubbers, it is
substantiated by the following information:
2-20
-------
Only 92 of this vendor's units are accounted for in the sample
population. This represents less than 30 percent of those c
scrubbers within our sample for which a vendor's name was given.
If the remaining 141 of this vendor's scrubbers are added to the
number of sodium scrubbers in the sample population, then
approximately 500 would be accounted for.
Several major vendors either did not respond or were not
contacted. Their recent installations, therefore, are not
accounted for in the sample. The 500 figure mentioned above
should therefore be conservative.
One official from Kern County estimates that there are about 1000
steam generators and virtually all have sodium scrubber. It
should be noted, though, that a single scrubber will, in many
cases, treat the flue gas from more than one steam generator.
Nevertheless, it suggests that there are many sodium scrubbers in
the field today that are not accounted for by the sample in
Table 2.1-2.
Other pertinent information presented in Table 2.1-2 include actual and
guaranteed SO,, removal efficiency, start-up date, boiler size equivalent,
fuel type and sulfur content, and wastewater disposal technique. Those
scrubbers that were found to have been shut-down since the last survey were
deleted from the list. Within this revised sample, California scrubbers
comprise roughly 90 percent of the total sodium scrubber population, with
the remaining 10 percent fairly evenly distributed throughout 15 other
states. Sixty-one percent of the California scrubbers are located in Kern
County.
Tables 2.1-3, 2.1-4, 2.1-5, and 2.1-6 were derived from Table 2.1-2 and
provide further analyses of the information contained within it.
Table 2.1-3 presents the population of sodium scrubbers by application. It
also categorizes these applications into three boiler size equivalents: 0 -
100 x 106 Btu/hr, 100 - 250 x 106 Btu/hr, and greater than 250 x 106 Btu/hr.
The most pertinent conclusions to be derived from this table are listed
below.
Sodium scrubbers installed to treat flue gas from oil field
2-21
-------
TABLE 2.1-3 POPULATION OF SODIUM SCRUBBERS BY APPLICATION3)b
Application
Paper mill
Textile mill
Oil generators
field
Refinery
Soda Ash
Other Industrial
'Total
Mean Boiler Size
Equivalent
Boiler Si
0 - 100 100
Before After Before
1980 1980 1980
1 1
2
99 161 5
2
522
104 166 10
50.78 46.20 153.10
ze Equivalent (MMBTU/hr)
- 250
After
1980
1
15
1
17
137.29
250+
Before After
1980 1980
8
12 7
2
6 2
3
29 11
547.28 337.45
Total Number
Within Sample
Before After
1980 1980
9 1
2 1
116 183
0 5
6 2
10 2
143 194
Percent of Sodium
Scrubber
Before
1980
6.3
1.4
81.1
0
4.2
7.0
100
Sample
After
1980
0.5
0.5
94.3
2.6
1.0
1.0
100
Reference 5.
The sample itself represents 96 percent of the total wet FGD systems that have been installed up to
October 1983. If the estimated 670 sodium scrubbers are used, then sodium scrubbing represents about
98 percent of the total number of wet FGD systems.
cMean Boiler Size Equivalent (Q > 100 MMBTU/hr): Before 1980 446.21
After 1980 215.93
Overall 349.97
-------
ro
to
TABLE 2.1-4 TOTAL SO-
SODIUM SCRUBBER
TREATED BY APPLICATION FOR,CURRENT
SAMPLE0 'c (1000 Ib S02/yr)J
Boiler Size Equivalent (MMBTU/hr)
0 - 100
Before After
Application 1980 1980
Paper mill 1,300 1,200
Textile mill
Oil field generators 51,500 56,000
Refinery 900
Soda Ash
Other Industrial 3,400 900
> 100
Before After
1980 1980
85,960
1,900 4,500
38,500 21,900
8,200
41,700 8,000
23,700
Percent of S0? Treated
by Sodium Scrubbers
Before
1980
35.7
0.8
36.8
0
17.0
11.0
After
1980
1.2
4.8
76.4
8.9
7.9
0.9
Total 56,200 59,000 191,760 42.600 100 inn
-------
TABLE 2.1-5. POPULATION OF SCRUBBERS ON UNITS FIRING OIL, COAL AND OTHER FUELS5
Number of Units Number of Units
Firing Coal Firing Oil
^ate 0-100 100-250 250+ 0-100 100-250 250+
Alabama i
Arkansas 2
California 262 20 21
Florida
Georgia
Michigan 1 i
Missouri 2
Montana 1
North Carolina 2
New Jersey 2
New York 4
Ohio 1
Rhode Island 1
Texas i
Virginia 1
Wyoming 6
lota I 5 4 11 265 23 21
Coal Oil Other Fuel
Vercent 6 92 2
Average Sulfur Content 1.35 1.52 1.71
Standard Deviation 0.67 0.727 0.485
Number of Units Average
Firing Other Fuels Sulfur Content
0-100 100-250 250+ mean
.25
1 1.42
1.49
3 1.83
2 1.50
0.80
3.2
2.2
1.00
3.50
1.20
1.35
4.0
6.0
_
1.08
8 -or
std. dev.
—
0.29
0.63
0.72
0.00
0.00
0.00
_
0.00
0.99
0.00
*.
_
-
0.34
0^3
bBoiler Size Equivalent (MM Btu/hr)
c# reporting fuel type = 356
d# reporting sulfur content = 323
# reporting fuel type and boiler si?p
•\-\~i
-------
TABLE 2.1-6 POPULATION OF WASTE DISPOSAL METHODS OF SODIUM SCRUBBERS5
ro
i
ro
en
Sewer then
State RWB
Alabama
Alaska
Arkansas
Cal ifornia
Florida
Michigan . 1
Missouri 1
North Carolina
New Jersey 2
New York
Ohio 1
Rhode Island
Texas
Wyoming
Total 5
Percent of 10
of total
Total excluding 5
CA & WY
Percent of 31
Disposal Method
Discharge Deep-well
to a RWB Ponding injection
2
21 9
2
2
1
1
1 1
2 1
9 23 11
18 47 23
9 0 1
56 06
Treatment Method
Used in Oxidation Removal
Plant Processes for COD of TSS
1
1
2 2
2 2
1 1
1
2
1
1
1
1 Total 9 8
2
1
6
Dilution
1
1
total excluding
CA & WY
RWB = Receiving Water Body
TSS = Total Suspended Solids
-------
generators comprise about 89 percent of all industrial boiler
sodium scrubbers and about 94 percent of those installed after
1980. In addition, sodium scrubbers on oil field generators
comprise about 91 percent of all wet FGD scrubbers installed after
1980.
The overall mean BSE for sodium scrubbers is 109 x 10 Btu/hr.
For the units greater than OF equal to 100 x 10 Btu/hr, the
overall mean BSE is 350 x 10 Btu/hr and since 1980 has been
220 x 10° Btu/hr.
Of all the sodium scrubbers in operation today, approximately
74 percent treat BSE's less than 100 x 10 Btu/hr.
The number of installations since 1980 for all applications except
for oil field generators and refineries have decreased relative to
those before 1980.
Although Table 2.1-3 presents useful data concerning the number of
sodium scrubbers, it gives no indication of the amount or percentage of S0?
treated. Table 2.1-4 presents the total S02 treated in each industrial
application. This sample itself treats about 70 percent of the S02 treated
by all wet FGD scrubbers. If all of the estimated 670 sodium scrubbers are
considered, then sodium scrubbers treat approximately 82 percent of the S0?
treated by all FGD scrubbers. Other conclusions to be derived from this
table are listed below.
- Approximately 360,000 tons per year of SO- are currently being
treated by sodium scrubbers.
- Although oil field scrubbers represent 89 percent of all sodium
scrubbers, they treat only 49 percent of the S02 treated by
all industrial boileg sodium scrubbers. For BSE's greater than
or equal to 100 x 10 Btu/hr, they treat only 28 percent of the
S02 treated by all sodium scrubbers above that size.
- Sodium scrubbers in paper mills and soda ash plants treated
over 50 percent of the total SO- treated by pre-1980
installations. Post-1980 installations treat less than 10
percent. It is speculated that most if not all pre-1980
installations were retrofits.
Table 2.1-5 presents those scrubbers within our sample for which fuel
type and sulfur content information were provided. It breaks down the
2-26
-------
analysis into state and BSE as well. The following are pertinent
conclusions:
- The average fuel sulfur content of this sample is 1.51 weight
percent with a standard deviation of 0.73. If California is
excluded, the average sulfur content is 1.78 with a standard
deviation of 1.23.
- Ninety-two percent of the scrubbers service boilers firing oil.
When California scrubbers are ignored, only 20 percent of the
remaining scrubbers service boilers firing oil and 60 percent
service boilers firing coal.
- The sulfur content of the coal is about 10 percent less than
that of the oil.
- There appears to be a growing demand for sodium scrubbers in
refineries to treat the flue gas from process boilers firing
high sulfur oil (6-8 weight percent). This trend is confirmed
by a prominent sodium scrubbing vendor.
Table 2.1-6 presents the population of waste disposal methods by
category and by state. Thirty-nine of the 72 plants in the sample provided
this information. The important conclusions are summarized below.
- About 50 percent of the plants reporting wastewater disposal
procedure use evaporation ponds. However, all of these plants are
located in California and Wyoming.
- Approximately 10 percent of the plants dispose of their waste
in a sewer. If California and Wyoming are ignored, the sewerage
option represents 31 percent.
- Approximately 20 percent of the plants use aeration to treat the
scrubber wastewater prior to discharge.
Table 2.1-1 presents the distribution of scrubber types for those 290
scrubbing units within our sample that reported scrubber type. The three
different categories (open vessels, vessels with internals, and combination
vessels) comprise approximately an equal fraction each of the total sample
population. Specifically, the open vessels comprise 39 percent; the vessels
with internals comprise 28 percent, and the combination vessels comprise 33
2-27
-------
percent. Spray baffles represent the highest percentage of the population
of scrubbers; liquid jet eductors and tray towers are next. Packed beds
follow these; however, the packed beds' statistic is probably a distortion
of the absorber's overall representation. Unlike most of the other absorber
types whose data were derived from many plants, the packed bed data came
from one source. Moreover, this source made its report in 1973. Since
packed beds were the most popular absorption device among the first
generation scrubbers (for both sodium and calcium reagents) it is not
surprising to see such a large number at one plant at that time. However,
demand for these units has diminished substantially in recent years
primarily because of reliability problems. Therefore, it is doubtful that
they will be applied widely in the future.
Reasons for the Current Popularity of Sodium Scrubbers
The reasons for the prevalence of sodium scrubbing systems as compared
to other industrial FGD systems are listed below:
- Ease of operation
- High reliability
- Relatively low initial capital costs
- Relatively low total annualized costs (for low SCL loadings)
The primary reason sodium scrubbers are popular is that they are
relatively easy to operate, requiring little operator attention. Although
reliability is a function of ease of operation, it is nevertheless a
separate reason for sodium scrubbing popularity. Some industrial boiler
plants must shut their whole process down if the scrubber malfunctions. In
these cases, the penalty associated with process downtime may be high. In
other cases, loss of the scrubbing system will require the boiler owner to
burn more expensive, low-sulfur fuels. Process economics (capital and
annualized costs) also substantially affect system applicability and should
thus be mentioned here as well.
2-28
-------
Since oil field generators use such a large percentage of the sodium
scrubbers, a separate applicability explanation is warranted for them.
There are four primary reasons for their predominance, some of which are
unique to their application. First, since steam generators are remotely
located, they must be able to operate without operator attention. Second,
because of the strict air regulation in their areas, they must be
exceptionally reliable because the generator must be shut down if the
scrubber malfunctions. Third, generators are moved frequently from one oil
well to the next, and sodium scrubbers are the most portable wet FGD
systems. And fourth, since SCL loadings from oil field generators are
generally low, sodium scrubbers are usually the least expensive of the FGD
technologies.
2.1.1.4 Development Status. Sodium scrubbers are well demonstrated.
Approximately 400 units have been installed since 1980,5 resulting in a
great deal of process refinement which has translated into reductions in
cost. For example, capital costs are now approximately 40 percent of what
they were five years ago. As would be expected with any maturing
technology, this cost reduction can be attributed both to the increased
economies of high production volumes and to increased standardization. In
addition, process control has become more sophisticated for sodium systems,
resulting in an increase in system reliability as well as reductions in
labor and maintenance requirements.
2.1.1.5 Reliability. Reliability, operability, and availability are
typically used interchangeably throughout industry without any rigorous
definition of each. Thus, the term "reliability" is subject to different
interpretations, and in normal usage is understood to mean that a system is
either "free from failure" or it is "able to function when needed." To
avoid confusion, the EPA has standardized these terms by defining them
quantitatively. These definitions are as follows:
Availability: Hours the FGD system was available (whether operated or
not) divided by the hours in the period, expressed as a
2-29
-------
percentage.
Operability: Hours the FGD system was operated divided by boiler
operating hours in the period, expressed as a
percentage.
Reliability: Hours the FGD system was operated divided by the hours
the FGD system was called upon to operate, expressed
as a percentage.
When requesting "reliability" data from plants, reliability was the
index requested and was presented to the plants as the definition above.
The numbers presented in Table 2.1-7 are the values provided by various
plants and vendors in response to this definition. It should be noted that
no actual operating data or logs were obtained for these systems, nor were
time periods of data collection specified by the plants and vendors. These
data are therefore not to be taken as rigorous measures of sodium scrubbing
performance. Nevertheless, because all are consistently high, they show
that sodium scrubbing is exceptionally reliable. Earlier information from
the EPA Industrial Boiler FGD Survey: First Quarter 1979 showed reliability
figures similar to these recently acquired data. Fifteen boiler operators
reported reliability and/or operability indices of between 89 and 100
percent with an average of 97.8 percent. Of the 15 responses gathered in
that survey, 9 reported a 100 percent reliability and only two reported
4?
reliabilities less than 95 percent.
These high reliabilities are due primarily to the simplicity of both
the chemistry and design of the process. The sodium species in the
recirculation stream remain in solution at the TDS concentrations and
temperature ranges typically found in the operation of sodium scrubbing
43
systems. Solution scrubbing minimizes the erosion of pumps and pipes, as
well as the scaling of mist eliminators all of which contribute to a sub-
stantial fraction of the downtime in calcium-based systems. Calcium,
leached from coal ash and sometimes present in the make-up water itself, is
the predominant precipitable species. However, its concentration generally
2-30
-------
TABLE 2.1-7 RECENT RELIABILITY DATA FOR SODIUM SCRUBBERS
Plant or Vendor Number of Scrubber Units Reliability (%)
A41 15 99 - 99.5
B18 2 99+
C25 1 100
D3'20 233 98+
E21 5 98
2-31
-------
is too low to cause scaling problems, even at relatively high pH's.
Operating the system in the concentrated mode reduces the risk of calcium
precipitation by reducing system pH. Even more importantly, the
concentrated solution provides a buffer which is effective in preventing the
pH excursions that can result from widely fluctuating inlet SCL flow rates.
In addition, a vast majority of the sodium systems operate with open
vessels, or combination vessel scrubbers. Those scrubber systems that have
vessels with internals are expected to have lower reliabilities. Compared
to vessels with internals, open vessels increase reliability by reducing
both the horizontal solid surface area available for scale formation and the
residence time of the scrubbing liquid on a solid surface. These factors in
turn minimize isolated pH excursions that can cause scaling and corrosion.
2.1.1.6 Emissions Data. Table 2.1-8 presents SO,, removal efficiencies and
outlet S02 emissions for 45 scrubbers at 18 different sites.44 S02 removal
efficiencies for these scrubbers ranged from 89.3 to 99.4 percent, while
outlet S02 emissions ranged from 0.007 lb/106 Btu to 0.23 lb/106 Btu. The
average sulfur content in the fuels fired at all sites was 1.30 weight
percent with a standard deviation of 0.64 weight percent. For the boilers
at the sites firing oil only, the average sulfur content was 1.17 weight
percent with a standard deviation of 0.28 weight percent. Other relevant
data in this table include the year in which the tests were performed,
method of testing, absorber type, and pH of the scrubbing liquid. All but
one test were performed after 1980, and all but one scrubber, which is at
Site #1, operated on a boiler that fired oil. The scrubber at Site #1
treated the flue gas from a boiler firing coal that had a sulfur content of
3.64 weight percent.
Table 2.1-8a presents the results of the average S02 removal
efficiencies and average S02 outlet emissions for the 45 scrubbers listed in
Table 2.1-8. The average S02 removal efficiency was 96.2 percent with a
standard deviation of 2.9 percent. However, if the data for the two tray
absorber site (Site #10) are deleted, the average S02 removal efficiency per
scrubber becomes 96.7 percent with a standard deviation of 2.3 percent. The
2-32
-------
TABLE 2.1-8 EMISSIONS DATA FROM SELECTED SODIUM SCRUBBING FGD SYSTEMS44
to
CO
Source
(Company)
1
2
2
2
2
3
4
5
6
7
7
8
9
10
11
11
12
13
bC= Coal:
Site
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
0 = Oil.
HI i „ .
Year
Number of of
Scrubbers Test
1
?
1
1
1
2
2
1
4
3
4
10
2
3
2
1
3
2
1979 -1980
1983
1983
1983
1983
1981
1983
1983
1983
1981-1982
1981-1982
1982
1982
1982-1983
1982
1982
1982
1982-1983
Test
Method
CEMd
EPA 8
EPA 8
EPA B
EPA 8
EPA 8
CEM
CEM
CEM
CEM
CEM
EPA 8
EPA 8
EPA 8
EPA 8
EPA 8
EPA 8
EPA 8
SO, Removal
Efficiency
(*)
92. 2f
96.6
99. Of
98.1
98.1
99.4e>f
96.9
99.4
99.1
89. 3f
95.2
96. 2f
98. Of
98.1
96. 5f
95. Of
96.2
90. Of
Outlet
Emissions Fuel
(lb/S02/10° Btu) Type3
0.20°
0.055
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
02
,03
,03
,026e
053
007
008
23
099
047
022
022
039
038
058
095
C
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Fuel's
Sul fur
Content
(wt. X)
3.64
1.50
1.14
1.46
1.46
1.1
1.51
1.04
0.79
1.60
1.43
1.02
1.0
1.0
1.1
0.6
1.01
1.0
Absorber
Type6
TA
LJE
TA
TA
SB
ST
NA
NA
NA
TAC
TA
VS
VS
SB
VS
SB
NA
LJE
Scrubber
Inlet pH
8.1
NA"
NA
NA
NA
7.0
NA
NA
NA
7.45
7.96
NA
NA
NA
NA
NA
NA
NA
CIA = iray Absorber: VS = Venturi Scrubber: SB = Spray Baffle: ST = Spray Tower; LJE =
This tray absorber was known to contain only two trays. Two try absorbers are known to
rf than those of three tray absorbers.
Liquid Jeft Eductor; NA = Not Available
lower SO, removal efficiencies
At s1te
fS02 outlet emissions were determined in lb/10 Btu for one scrubber only.
S02 removal efficiency was determined from measured inlet and outlet SO, emissions.
NA means not available. *-
-------
44
TABLE 2.1-8a. AVERAGE RESULTS FROM SODIUM SCRUBBING SYSTEMS
SO,, Removal Efficiencies ± Standard Deviations, in Percent
1) Average efficiency for all sites. 96.5 ± 2.9
2) Average efficiency for the nine sites that
measured inlet/outlet SO^ emissions. 95.5 ± 3.6
3) Average efficiency for all sites excluding the
site with the two tray scrubber (site 10). 96.9 ± 2.3
4) Average efficiency for all scrubbers. 96.2 ± 2.9
5) Average efficiency for all scrubbers excluding
the two tray scrubbers (Site 10). 96.7 ± 2.3
S00 Outlet Emissions ± Standard Deviations, in Ib S00/106 Btu
c. ———— .—-—•——— ^
1) Average S02 outlet emissions for all sites, 0.060 ± 0.062
2) Average S0? outlet emissions for oil-fired boiler 0.052 ± 0.053
sites £-
2-34
-------
TABLE 2.1-8b. S02 REMOVAL EFFICIENCIES BY ABSORBER
Type
of
Absorber
Venturi scrubber
Tray Absorber
Spray Baffle
Liquid Jet Eductor
Spray Tower
Total
Number
of
Scrubbers
11
7
3
3
2
26
Number
of
Sites
2
4
3
2
12
Range of S0? Removal
Efficiency (%)
96.3 ± 1.2
96.3 ± 1.8
97.5 ± 2.3
94.1 ± 4.4
99.4 ± 0.6
96.4 ± 2.2
aThe data from the two-tray absorbers at Site #10 in Table 2.1-8 were
excluded from the data set.
2-35
-------
reason for deleting the data for the two-tray absorber is that most tray
absorbers have three trays, and three-tray absorbers are known to have
higher S02 removal efficiencies than two-tray absorbers. The average outlet
------ j.oe ib scyio6
ie oil-fired b
outlet S0? emissions was 0.052 Ib S0,/106 Btu with a standard deviation of
£ C C-
S02 emissions per site for all sites was 0.06 Ib S02/106 Btu with a standard
deviation of 0.062 Ib S02/106 Btu. For the oil-fired boilers, the average
0.053 Ib S02/106 Btu
Table 2.1-8b shows that absorber type (ignoring the two tray absorbers)
has only a slight, if any, effect on S02 removal efficiency. The average
S02 removal efficiency for the 26 scrubbers identified in this table was
96.4 percent with a standard deviation of 2.2 percent.
As discussed in Section 2.1.1.2, S02 removal efficiency is a strong
function of pH. Due to the lack of pH data in Table 2.1-8, this contention
can be neither supported nor refuted. In addition, S02 removal efficiency
does not appear to be a function of fuel sulfur content.
All sodium scrubbing test results reported in Table 2.1-8 except those
for Site #1 were from short-term compliance tests. At Site #1, the scrubber
treating flue gas from a coal-fired boiler averaged 96.2 percent S02 removal
efficiency, which is consistent with the scrubbers treating flue gas from
oil-fired boilers. The data from Site #1 were collected from 30 days of
continuous emission monitoring (CEM). At sites #7 through #11, short-term
CEM compliance tests were performed using ultraviolet photometry. These
tests were classified by the EPA as an alternative method to measure S02.
The EPA Method 8 was the test method used at sites #2 through #6 and sites
#12 through #18. Both inlet and outlet S02 emissions were measured at nine
sites, while at the other nine sites, only outlet S02 emissions were
measured. The inlet SO^ emissions at the latter nine sites were calculated
from the sulfur content in the oil and AP-42 correlations.
2.1.2 Dual Alkali
The dual (or double) alkali process is the second most prevalent wet
FGD technology being applied to industrial boilers today. Since 1974, 13
regeneration systems servicing 27 sodium scrubbers have been installed on
2-36
-------
industrial boilers. Five regeneration systems have been installed since
1980 to service nine scrubbers. It should be noted, however, that four of
the dual alkali systems are known to be currently inoperative.
Of all the industrial wet FGD systems in operation today, approximately
one percent are dual alkali systems. Moreover, the use of dual alkali
systems has not increased in recent years. The reasons for the lack of
current interest relative to the interest in sodium systems are discussed in
Section 2.1.2.3. Although dual alkali units comprise a small fraction of
the number of wet FGD systems, they account for almost 16 percent of the
total S02 currently being treated by wet FGD systems. This is because the
average boiler size equivalent (BSE) and fuel sulfur content for operating
dual alkali systems are both much higher than those for sodium scrubbing
systems.
Dual alkali systems are characterized by slightly lower S02 removal
efficiencies and slightly lower reliabilities than those achieved by sodium
scrubbing systems. Based on data from EPA approved tests, the average S0~
removal efficiency for dual alkali units has been about 90 percent. The
reported reliabilities of dual alkali FGD systems range from 80 to 99
percent.
The lime dual alkali (IDA) process is a relatively mature technology.
One vendor, however, is currently developing a way to reduce the capital
cost for surge capabilities. Surge capabilities are necessary to
accommodate the widely fluctuating loads which are characteristic of
industrial boiler operation. Furthermore, this vendor predicts that in the
near future, dual alkali regeneration plants will be constructed in areas of
high sodium scrubber density such as Kern County and New Jersey. These
plants will treat the blowdown from sodium scrubbers already operating in
these areas and regenerate the sodium reagent for reuse. In addition,
methods for substituting limestone for lime in the regeneration section are
being investigated. Currently, the cost for raw limestone is about one
tenth of the cost for lime. This cost differential provides a considerable
incentive for reagent substitution if the technical and economic feasibility
of this change can be demonstrated.
2-37
-------
2.1.2.1 Process Description. The dual alkali process and chemistry are
described in detail in Section 4.2.2.1.1 of the BID. However, there are
additions and corrections that deserve mention and are listed below
according to the process area.
Absorption
The absorption process is almost identical to that of sodium
scrubbing systems, except for elevated concentrations of trace
substances (notably Cl" and Ca+2) which are the result of closed
1 3 28
loop operation. ''
Regeneration
The washwater is not returned to the scrubber directly but rather
via the thickener.31'45
All dual alkali processes currently operating in the U.S. use
lime as the regenerating alkali. One vendor, however, has
recently announced plans to offer a limestone dual alkali
process and will begin a pilot operation in 1984.17 The relative
merits of this process are discussed in Section 2.1.2.4.
Solids Separation
The regeneration reactor effluent, which contains a 1.0 to 1.5
percent suspension of calcium sulfite and sulfate solids as well
as soluble sodium sulfite and sulfate, is sent to the solids
separation section where the solids are concentrated via a
thickener and vacuum filter to approximately 50 percent solids.
In most systems, the filter cake is washed with make-up water to
reduce the soluble sodium salts in the adherent liquor prior to
disposal.31'45
The filter cake can be disposed of directly, stabilized with fly
ash, and/or fixated with lime. Fly ash reduces the filter cake's
moisture content, thereby improving its stability. By adding lime
to the fly ash mixture, a long-term cement-forming process known
as pozzolanic action begins. Pozzolanic action is similar to
cement curing in that chemical bonds between the lime and the
alumina and silica-containing components of fly ash are formed.
2-38
-------
The mixture strength increases over a period of several
45
months.
2.1.2.2 Factors Affecting Performance
Since the dual alkali system uses a sodium scrubber, many of the
factors affecting sodium scrubbing performance as described in Section
2,1.1.2 are also applicable to the scrubbing section of the dual alkali
process. However, there are some important differences. Dual alkali
systems generally operate in a very concentrated mode. Whereas the IDS
concentrations for sodium scrubbers are typically 5-10 weight percent, those
for dual alkali systems are 15-20 percent.3'46'47'48 The benefits of high
TDS concentrations have already been discussed in Section 2.1.1.2. Those
reasons that pertain especially to the more concentrated dual alkali systems
are discussed in more depth as follows:
At high TDS concentrations, oxidation is minimized by maintaining
a relatively high, steady state sodium sulfate concentration
46 49 50
throughout the system. ' ' This inhibits further sulfate
formation. Sulfate ions are much more difficult to precipitate
out of solution than sulfite ions. The sulfate ion will instead
leave the process with sodium. Therefore, high TDS levels in the
dual alkali system reduce sodium consumption by reducing sulfate
formation.
Since dual alkali systems are operated essentially in a closed
loop mode, it is important that liquid recirculation rates be
minimized to minimize the size of vessels and tanks as well as
pumping and filtering requirements. These variables are
minimized as the TDS concentration increases.
The higher TDS concentration provides a stronger buffer solution.
The control of pH within the narrow range of 6.2-6.8 is essential
in dual alkali systems because of corrosion at low pH's and
scaling at high pH's. Because of the closed operating mode,
chloride accumulates to very high concentrations thus increasing
3 31
corrosion potential at pH's below 6.0. In addition, since
2-39
-------
some calcium ions from the regeneration section are carried over
to the scrubbing section, there is a great potential for rapid and
substantial plugging via calcium scaling above a pH of 7.
This same buffering capacity leads to stable outlet S02
concentrations, even with fluctuations in load and inlet SCL
concentration. However, with high concentrations of NaHSO., the
S02 absorption reaction, Na2$03 + S02 + H20 •»• ZNaHSCL, is
equilibrium constrained.
The other factors affecting scrubber performance such as liquid-to-gas
ratio and absorber design are the same as those discussed in the sodium
scrubbing section. These will not be repeated here; instead, the reader is
referred to Section 2.1.1.2.
The predominant factors affecting regeneration are the sulfur and
chloride contents of the fuel. The combustion of low sulfur coal results in
a higher ratio of oxygen to sulfur dioxide in the flue gas than does the
combustion of high sulfur coal. A higher relative oxygen content promotes
the oxidation of a higher percentage of sodium sulfite to sodium sulfate.
This can cause two major problems: (1) a lower liquid phase alkalinity due
to higher S04~ and lower S03~ levels in solution and thus a lower SCL
removal capability, and (2) potentially higher sodium losses due to the
requirement for a Na2S04 purge stream.
Dual alkali systems operate in a relatively closed-loop mode, except
for the water evaporated in the scrubber and that which is occluded in the
waste sludge. Because of this, substances that exist only in trace amounts
in sodium scrubbing systems can build up to high steady state concentrations
in the dual alkali process. Chloride, which is volatilized from the coal in
the boiler and absorbed in the scrubber, is the most corrosive of the
substances contained in coal. In some cases, chloride can reach
"3
concentrations of up to 40,000 ppm in the scrubbing liquor.
High chloride concentrations lead to high sodium losses and contribute
to stress corrosion. Sodium losses increase because sodium, a positive ion,
will pair with chloride, a negative ion, to insure charge conservation in
the scrubbing liquor. This effectively ties up the sodium that would
2-40
-------
otherwise be associated with an alkaline ion such as SCL". Since chlorides
are removed in soluble form (either as a liquid purge or as occluded liquor
in the solid waste), a corresponding amount of sodium will be lost along
with any chloride purged from the system. High chloride concentrations at
pH's below 6.2 can substantially and rapidly corrode all sections of a dual
alkali system.
One proposed solution to the chloride problem is to use a prescrubber
to remove chlorides before the flue gas enters the dual alkali system.
However, the use of a prescrubber with a separate liquor loop will cause
water balance problems in the system. Since all the evaporation loss would
occur in the prescrubber, the only water loss from the double alkali system
would be the water included with the sludge. The make-up water rate for
this small water loss would not be sufficient to meet requirements for
normal cake washing (more than one displacement wash), demister washing,
pump seals, and lime slaking. Another source agrees that prescrubbers are
not a viable solution to the chloride problem. This source contends that
the installation of Inconel or Hastelloy G alloys is the only feasible
deterrent to corrosion.
2.1.2.3 Applicability to industrial boilers
Table 2.1-9 presents a summary of double alkali scrubbing systems, both
currently operative and inoperative, that have been installed on domestic
industrial boilers. It presents such pertinent information as operating
status, boiler size equivalent, number of scrubbers per unit, and fuel type
and sulfur content. Thirteen dual alkali systems have been installed since
the early 1970's. All of these plants except one have only one dual alkali
system; the exception has two complete dual alkali systems. The 13
regeneration sections service a total of 27 scrubbers and an average of 520
x 10 Btu/hr BSE (the average BSE for the scrubbers is 230 x 106 Btu/hr).
The average fuel sulfur content is 2.81 weight percent. The dual alkali
systems that are currently operating represent about one percent of the
total industrial wet FGD systems operating today and they treat about
16 percent of the total SO,, treated by industrial boiler wet FGD systems.5
2-41
-------
TABLE 2.1-9. APPLICABILITY OF DUAL ALKALI SYSTEMS INSTALLED ON INDUSTRIAL BOILERS
,
IN3
Boiler Size
Company/Location
Caterg;jl^r Tractor
East Peoria, IL
Joliet, IL
Mapleton, IL
Morton, IL
Mossville, IL
Firestone53'54
(now Occidental)
Potts town, PA
General Motors55
Parma, OH
Santa Fe Energy Corp •
Bakersfield, CA
ARCO Polymers31
Monaca, PA
GrissoguAir Force
Base58
Peru, IN
St. Regis Paper59
Sartell, MN
Mississippi Army ,n
Ammunitions Plant
Bay St. Louis, Miss
Total
Average Per Plant
These values represent
THaCti u A 1 nac v£jr»*r»ac art t-
Start-up Operating Equivalent No. of No of . Fuel Absorber
Vendor Date Status (106 Btu/hr)a FGD Units Scrubbers6 Type8 K ?JpeS
FMC
Zurn
FMC
Zurn
Zurn
FMC
GM
57 FMC
FMC
Neptune/
Airpol
Neptune/
Airpol
Zurn
1978
1974
1974
1978
1975
1975
1974
1979
1980
1981
1983
1983
the BSE for the
Operative
Operative
Operative
Operative
NA
Shut-down
(Company change)
Operative
Operative
Operative
Operative
Inoperative
(Plugging)
Inoperative
(plugging)
940 1 4 C 3.2
300 1 2 C 3.2
1060 1 2 C 3.2
170 1 2 C 3.2
630 1 4 C 3.2
36 1 1C 2.5-3.6
570 1 2 C 2.5
310 1 10 1.5
1360 1 3 C 3.0
140 1 2 C 3.0-3.5
590 1 2 C 1.4
150 2 2 C 3.0
6300 13 27
520 1.1 2.3 2.81
VS
TA
VS
TA
VS
VS
TA
DD
DD
VS
PB
PB
overall scrubbing system, not for the Individual scrubbers.
These values represent the number of scrubbers serviced by the regeneration section(s).
CC = Coal; 0 = Oil.
VS = Venturi Scrubber: TA = Tray Absorber: DD = Disc and Oonut Contactor: PB = Packed Bed
-------
Nine of the thirteen regeneration systems are known to be operating at
this time; four are known not to be operating (see Table 2.1-9). Of these
six, three are inoperative due to plugging, while one is inoperative due to
54 59 60
a company takeover. ' ' Two of these systems are operating as sodium
scrubbing systems, disposing of their wastewater directly to a local
sewerage system. The other unit is removing only particulate matter; the
boiler it serves is currently burning a compliance coal to meet state SO,
59
regulations.
Since 1980 five systems have been installed, and of these five, only
two are operative. The lack of current demand relative to the demand for
sodium systems may be explained to a large extent by the following:
Reliability factors for industrial dual alkali systems are
slightly lower than those for sodium scrubbing systems (see
Section 2.1.2.5).
Dual alkali systems are more complex than sodium
scrubbing systems and therefore require more operator attention.
Dual alkali systems are less economical than sodium scrubbing
systems, especially for low SOp loadings.
2.1.2.4 Development status
Although lime dual alkali is a mature technology, it is undergoing
several developments. One source is predicting that, in the near future,
centralized dual alkali regeneration plants will be installed in areas of
high sodium scrubber density. It claims that there is enough economic
incentive now for private contractors to construct plants to regenerate the
spent sodium salts of the scrubber blowdo'wn. After the sodium has been
exchanged for calcium, it will be sold back to the scrubber operators for
reuse in their systems. In most cases, the cost of calcium sludge waste
disposal will be lower than the cost of disposing the sodium blowdown
stream. Currently, there are bids out to build two regeneration systems:
one to service sodium scrubbers operating on 63 steam generators and the
other to service sodium scrubbers operating on 20 steam generators.
One vendor is evaluating methods of substituting limestone for lime in
the regeneration step. Currently, the cost for raw limestone is about
2-43
-------
one-sixth the cost of lime. This cost differential provides a considerable
incentive for reagent substitution if the technical and economic feasibility
of this change can be demonstrated. A limestone dual alkali (LSDA) process
was tested in a pilot scale system at Gulf Power Company's Plant Scholz in
1981; another extensive test is planned by EPRI at Northern Indiana Public
Service Co. (NIPSCO) for early 1984.17'47
At the Scholz plant, the average SCL removal efficiency was
95.8 percent. However, this S02 removal efficiency might not be typical of
LSDA systems since it appears that the unit was operated in the dilute mode.
The pH of the effluent from the scrubber was consistently below 6.0, which
is not the pH that would have been expected if the system had been operated
at the design TDS concentration of 20 weight percent. In other words, the
operating TDS level appears to have been much lower than the design TDS
level.
Limestone utilizations were high, over 97 percent. However, the waste
sludge solids content (at 35 to 45 percent) was well below the design value
of 55 percent. In addition, the soda ash consumption of 0.29 moles of
Na2C03/mole of S02 removed far exceeded the design value of 0.04. These two
problems might have been the result of the mechanical performance of the
equipment, which, recomissioned after three years of inactivity, was poor.47
Despite the initial poor performance in these two areas, one vendor claims
that sodium consumption should be between about 0.02 to 0.05 moles Na/mole
S0? removed and that the solids content should be in the range of 50 to
17
70 percent for LSDA systems.
2.1.2.5 Reliability
Reliability data for industrial boiler and utility boiler dual alkali
systems are presented in Table 2.1-10. Included with the table are the
capacities of the systems, the period over which the data were collected,
and the type of index reported by the plant. Most of the plants kept
availability indices; Louisville Gas & Electric reported a reliability
value; Caterpillar reported an operability value; and neither Santa Fe
Energy nor Occidental Petroleum specified its index of reliability. Behrens
2-44
-------
TABLE 2.1-10. RELIABILITIES FOR DUAL ALKALI SYSTEMS
i
-p>
en
FGD Systems Capacity
(106 Btu/hr)
Sante Fe Energy Co.56
ARCO31
Firestone Tire & Rubber
(now Occidental Petroleum)
Caterpillar
Louisville Gas & Electric3'62
Southern Indiana Gas & Electric Co.3'63
Central Illinois Public Service Co.3'64
310
1,360
40
940
2,700
2,600
5,700
Period
of Data Index
Collection Reported Reliability(%)
48
12
NAb
NAb
12
13
9
NAb
Availability
NAb
Operability
Reliability
Availability
Availability
99
97.6
80
90
94.1
96.7
96
Utility systems.
5NA = Not Available.
-------
reports an average availability for the three utility systems presented in
Table 2.1-10 of 96.5 and an average operability of 79.7.65
Since the data are for various indices and from both utility and
industrial systems, a statistical analysis is not justified. Nevertheless,
the consistently high values indicate that dual alkali systems are highly
reliable.
2.1.2.6 Emission data
The emissions data for industrial boiler dual alkali systems are pre-
sented in Table 2.1-11 in terms of percent S02 removal efficiency and outlet
S02 emissions (Ib S02/106 Btu). Data for all six scrubbers were obtained
using EPA methods. The General Motors test lasted more than one month; the
Grissom Air Force Base and ARCO data were from compliance tests; and the
Santa Fe Energy data were from a recent in-house study.31'56'57'66'67'68
The average S02 removal efficiency was reported to be 91.0 percent with a
standard deviation of 2.4 percent. The average SCL emissions in the
C C-
scrubber outlet was 0.38 Ib S02/10 Btu with a standard deviation of 0.20 Ib
S02/10 Btu. Actual outlet emissions ranged from 0.091 to 0.65 Ib S02/106
Btu, depending on both the fuel sulfur content and the actual S02 removal
efficiency.
Two other tests not listed in the table deserve mention. They are the
long-term testing at Gulf Power Company's Plant Scholz pilot operation and
the recent year long test at Louisville Gas and Electric's Cane Run 6
system. Average S0? removal efficiencies for these two tests were 95.5 and
en C9
92.0 percent, respectively. '
Theoretically, any wet FGD system can achieve very high (99+ percent)
S02 removal efficiencies. This applies even to systems that use no alkaline
reagent and scrub with water only - however, only if very high L/G's are
used. Likewise, dual alkali systems can theoretically achieve very high S02
removal efficiencies. However, under normal operating conditions, they have
shown that they can achieve only around 90 percent. This is because most
are operated with high TDS concentrations (15 - 20 percent), which as was
discussed in Section 2.1.1.2, increase the equilibrium partial pressure of
2-46
-------
ro
i
TABLE 2.1-11. EMISSIONS DATA FOR DUAL ALKALI SYSTEMS USING EPA TESTING METHODS
Company/Location
ARCO Polymers31
Monaca, PA
General Motors67
Parma, OH
rn
Grissom Air
Force Base
Peru, IN
Santa Fe Energy '
Bakersfield, CA
Fuel
Type
Coal
Scrubber I Coal
Scrubber II Coal
System I Coal
System II Coal
Oil
Sulfur Content
of Fuel (wt.%)
2.5 - 2.8
2.5
2.5
3.0 - 3.5
3.0 - 3.5
1.5
Outlet Emissions
(Ibs S02/10° Btu)
0.65
0.30
0.32
0.56
0.38
0.091
S09 Removal
Efficiency (%)
88
92.2
91.6
88 1
94 2
91 7
Average 2.61 0.38
91.0
-------
S02. When dual alkali tests have been operated in the dilute mode, they
have shown removal efficiencies similar to those achieved by sodium
scrubbing systems. For example, testing at Plant Scholz of both the IDA and
LSDA processes provided SCL removal efficiencies of 95.5 and 95.8 percent,
50
respectively. Although not stated specifically, the absorbers effluent pH
indicated that both systems were operated in the dilute mode. Therefore,
the relatively low efficiencies reported by commercial-scale industrial
systems, which typically operate in the concentrated mode, appear to be
consistent with theory.
2.1.3 Limestone Wet Scrubbing
Limestone wet scrubbing has been applied at only one industrial boiler
site and its industrial sector demand appears limited within the near future
for the following reasons:
Compared to clear liquor (sodium or dual alkali) scrubbers,
limestone systems require considerably more operator
attention and skill, due to the potential for scaling.
Scaling is a result of the relative insolubility of limestone
(calcium carbonate) in water, being 1/14,000 as soluble as
sodium carbonate. Since industrial applications are less
likely to have sophistica-ted instrumentation, a pool of
skilled operators and technicians, or spare scrubber modules
than their utility counterparts, scaling and lower scrubber
reliabilities are likely.
The high initial capital costs of limestone scrubbing favor
sodium-based scrubbers for the smaller boiler sizes
encountered in industrial applications.
Due to the potential for scaling in limestone systems, it may
be more difficult to achieve high sustained SOp removal
efficiencies in limestone scrubbers compared
to more soluble lime and sodium-based scrubbers.
In general, unbuffered calcium-based absorption (limestone/lime)
achieves lower sustained SOp removal efficiencies than sodium-based
absorption on comparable applications. As an illustration, the two
limestone demonstrations at the Springfield Utilities Southwest Station and
2-48
-------
Rickenbacker Air National Guard Base (RANGE) both measured unbuffered,
limestone system removal of SCL on high sulfur coal application of 50 to
70 percent. As shown by sections 2.1.1.6 and 2.1.2.6, typical removal
efficiencies on high sulfur coal controlled by sodium and dual alkali
scrubbers have ranged from 85 to 99 percent. Mass transfer additives such
as adipic acid and dibasic acid can significantly increase S0? removal
efficiencies for limestone systems. S02 removal efficiencies of 90 to
96 percent have been achieved at several utility sites, while a 30-day
average of 94.3 percent SO^ removal was achieved at one industrial site.
Because of the limited number of limestone industrial applications, and
the unique design features of that one application, the performance of
limestone scrubbing in industry can only be estimated based on utility
experience. Behrens reports that limestone FGD availability in the utility
industry is the lowest of all absorbents, averaging 73.5 percent. This
does not compare favorably with lime at 84.1 percent and dual alkali at 96.2
percent. While other studies have shown how limestone FGD reliabilities
can be significantly improved through improved instrumentation, maintenance
and operating practices, and spare modules, the perception remains that
where minimal operator attention and expertise is applied, the most reliable
FGD systems, and the overwhelming choice of industry, are, and will continue
to be, sodium-based systems.
One noteworthy statistic that has not been discussed previously is the
effect of S02 loading on FGD reliability. Since the Behrens data above
include a disproportiate number of medium and high sulfur coal applications,
limestone FGD reliabilities on low sulfur coal applications tend to be much
higher than those values previously cited. For example, one limestone FGD
system tested by the EPA in 1979, which achieved better than 95 percent S02
removal for a 30-day period on a 0.55 percent sulfur coal, has achieved,
along with a sister unit, essentially 100 percent reliabilities in recent
years. Therefore, on low sulfur coal applications in industrial boilers,
limestone FGD may be a reasonable alternative.
2-49
-------
2.1.3.1. Process description. The process chemistry, equipment, and
operations for this system are described in the March 1982 BID. However, a
more detailed process flow diagram is included in Figure 2.1-3 to replace
the simplified diagram in the BID.69
2.1.3.2 Factors affecting performance. Limestone wet FGD systems are
confronted with two major chemical-related problems affecting their
performance. These are the relative insolubility of the reagent and the
susceptibility of the systems to scaling or plugging. These two problems
are the major considerations in the design and operation of limestone
systems. They affect the following variables:
- Reagent requirements
- Liquid-to-gas ratio (L/G)
- Usage of soluble species and additives
- Slurry pH
- Reaction tank residence time
- Scrubber design
- Reaction tank configuration
The S02 removal performance of sodium or dual alkali FGD systems is
limited only by gas-liquid mass transfer in the scrubbing step 'since all of
the alkalinity required for reaction with S02 is available in soluble form.
Calcium-based systems, on the other hand, rely on solids dissolution to
provide most of the alkalinity required for SCL absorption. Since
liquid-solid mass transfer tends to be significantly slower than gas-liquid
mass transfer, lime and limestone systems must be operated differently than
the other two wet FGD system, as described below.
Reagent requirements Unlike sodium scrubbers which operate at an
Na2C03/S02 stoichiometric equivalent ratio of less than one, limestone
systems operate at higher ratios because of limestone's relatively slow
dissolution rate. In other words, reagent utilization for limestone systems
is generally much lower than that for sodium-based systems. Although it is
generally agreed that SO,, removal is a function of the amount of excess
:-5o
-------
STACK
ADSORBER
LIMESTONE
ftTORAQE TANK
LIMESTONE
8LURKY
( WATER
8LUOGE TO
DISPOSAL
FIGURE 2.1-3. LIMESTONE PROCESS FLOW DIAGRAM
-------
limestone in solution, there is not a consensus concerning the levels
required. For instance, data from TVA's Shawnee Power Station in Paducah,
Kentucky show that S02 removal increases rapidly with increasing limestone
stoichiometry up to approximately 1.4 moles CaC03/mole S02 absorbed. Data
collected by Radian Corporation, however, indicate that little improvement
in SCL removal is realized by increasing the limestone stoichiometry above
1.1. . The differences between these results might be attributable to the
different sources of limestone and the specific hold tank configurations for
each system.
L/6 Ratio - Higher S02 removal efficiencies are achieved, and the
chances of scaling are reduced, at higher L/G ratios up to the point where
flooding and poor gas distribution occur. Typical L/G's for all
calcium-based systems range from 9 to 15 &/m3 (60 to 120 gal/1,000 ft3) with
the higher number being more typical for high sulfur coals.17'71 For
comparison, typical L/G's for sodium scrubbing systems range from 0.7 to 7.8
A/m (5 to 50 gal/1,000 ft3). As a result of the greater liquid flow,
pumping requirements and thus electricity costs will be several times
greater for limestone systems than for sodium-based systems.
Effects of Soluble Species - The concentration of dissolved ions other
than Ca in the scrubbing slurry directly affects the liquid phase
alkalinity and hence the system's ability to remove sulfur species from
boiler flue gas. The important ions are Na+, Mg++, and Cl". These soluble
ions can enter the system as Na-,0 or MgO in the ash, MgC03 from Thiosorbic
limestone, or HC1 in the flue gas. They can also be added to the system
with an additive. Magnesium and sodium assist in S0? scrubbing by
maintaining additional alkaline species in solution. This improvement in
S02 scrubbing due to increased magnesium and sodium concentrations in
scrubbing liquors is well documented. *3'74'75 On the other hand, high
chloride levels are generally thought to be detrimental to S02 removal since
chloride ions tie up the alkaline species and result in excessive alkalinity
losses. Some organic acids, such as adipic acid and dibasic acid, have been
2-52
-------
used to enhance limestone dissolution and S02 removal. Their effect is much
the same as an increase in alkalinity due to Mg+2 and Na+ addition.70
Slurry pH - The operating pH selected for calcium-based systems
involves a tradeoff between reagent utilization and SCL removal efficiency.
The more acidic the slurry is the greater the reagent utilization will be;
whereas the more alkaline the slurry is the greater the SO- efficiency will
be. For all calcium-based systems, operating at too high a pH can cause
scaling. For limestone systems, the optimum slurry pH is between 5 and 6.76
Reaction Tank Residence Time - Residence time in the reaction or hold
tank is determined by the size of the reaction tank and the liquid flow
rate. It is an especially important parameter in limestone systems because
of limestone's relatively slow dissolution rate. Larger reaction tank
residence times lead to greater limestone dissolution and hence higher
limestone utilization. However, the actual residence times used will be a
tradeoff between the costs associated with the reaction tank, pumping
requirements, and reagent costs. Besides reducing limestone utilization,
large hold tanks can reduce operating costs by producing large, easy to
dewater crystals.
Another important factor affecting limestone utilization is particle
size. The smaller the limestone particle is the greater the surface/volume
ratio and thus the greater the limestone dissolution rate will be. However,
too small a size can actually render the particle ineffective because water
effectively "blinds" the particle by shielding it from other water
molecules. As a result, the limestone does not dissolve as well.49
Scrubber Design - Because packed bed scrubbers are very efficient gas
absorption devices, they were used in the original lime and limestone wet
FGD systems. However, due to problems with plugging in packed beds, the
trend, over the last several years has been away from packed beds to open
spray towers, which are more reliable and easier to maintain.77
2-53
-------
Gas maldistribution can be a major problem in limestone spray
absorbers, particularly in large units. Portions of the scrubber can become
liquid phase alkalinity-limited due to gas maldistribution, even though the
total alkalinity entering the scrubber is sufficient for good SCL removal.
Some scrubber designs, therefore, incorporate straightening vanes and/or
open packing to promote good gas distribution.76
Reaction Tank Configuration - Reaction tank configuration also has been
shown to have an effect on both limestone utilization and SCL removal. One
source indicates that plug flow reaction tank designs can yield significant
improvements in limestone utilization and SCL removal. (A plug flow design
is one that allows the reacting stream to flow through the reactor such that
there is no backmixing. A plug flow situation can be approximated by a
number of mix tanks in series.) For a constant limestone addition rate, the
S02 removal efficiency at TVA's Shawnee Station increased from 70 to 79
percent by changing the reaction tank from a single stirred tank to three
identically-sized tanks in series. This plug flow effect apparently drives
the limestone dissolution reaction further toward completion and makes more
liquid phase alkalinity available for reaction with absorbed S02.70
Solid Waste Disposal - An important operational area which is
associated with all calcium-based flue gas desulfurization (FGD) systems is
the dewatering and disposal of the solid phase reaction products.
Conventional limestone systems produce a sludge composed primarily of
calcium sulfite which, because of its crystalline properties, may req'uire
special handling. This sludge is thixotropic: that is, it reliquefies upon
application of stress and does not dewater well. Consequently, ponding is
the normal method of ultimate disposal. However, further problems can arise
with ponding due to process liquor infiltration of ground water. Plastic or
clay liners are usually required to prevent this type of contamination.
Currently, FGD wastes are classified as non-hazardous according to RCRA
regulations, pending the outcome of an EPA study examining these wastes. A
recent innovation, forced oxidation, is being employed at many new systems
2-54
-------
to oxidize calcium sulfite to calcium sulfate, thereby improving the
JO
dewatering and handling properties of the sludge. Many commercial
processes also fixate the sludge by adding dry lime or limestone.
2.1.3.3 Applicability to industrial boilers. Currently the only
limestone wet FGD system operating on an industrial boiler is located at the
Rickenbacker Air National Guard Base (RANGE) (see Table 2.1-12). Since the
boiler is used only to provide space heating, it is used only six months out
80
of the year. No other calcium-based wet FGD system for an industrial
boiler application has been reported to have used or to be using limestone.
Furthermore, the major suppliers of utility limestone FGD systems predict
that few, if any, new industrial boiler limestone systems will be installed
in the near future.81'82'83
It should be noted, on the other hand, that the limestone wet FGD
process is a proven technology in the utility industry. As of November
1982, limestone systems represented over 50 percent of the 190 utility FGD
systems that had been installed and about 56 percent of the total scrubbing
capacity (both wet and dry). This discrepancy in limestone wet scrubbing
usage between industry and electric utilities is primarily due to the
perception that limestone systems are less reliable than sodium-based
systems. Also, despite the recent developments in the use of mass transfer
additives, forced oxidation, and spray tower designs, limestone systems
still have higher capital and annualized costs than sodium-based systems,
especially for small BSE's. The capital costs are much higher for limestone
systems because there are more equipment items, and the similar equipment
items (such as the scrubber) are larger to compensate for the relative
insolubility of limestone. Also, it is much more expensive to maintain high
reliabilities for limestone systems than it is for sodium scrubbing systems.
For example, studies have shown that more sophisticated instrumentation,
greater maintenance costs and operator attention, as well as spare absorbers
are required for a limestone wet FGD system to have consistently high
reliabilities. Despite these additional costs, limestone FGD systems are
economically attractive for utility boilers because the cost differences
2-55
-------
TABLE 2.1-12
SUMMARY OF LIMESTONE SYSTEMS OPERATING ON U.S.
INDUSTRIAL BOILERS AS OCTOBER 1983
Process Vendor
Lime Research
Company/Location
Rickenbacker Air
National Guard Base
7Q
Columbus, OH
Start-up Number of Size Fuel
Uate \-b[} Units (actm) Type Sulfur(%)
3/76 1 55,000 Coal 3.6
ro
i
en
cr>
-------
between soda ash and limestone exceed the additional capital and maintenance
costs. However, even for large industrial applications where limestone
systems appear to enjoy an annual cost advantage, boiler owners seem to be
reluctant to apply the technology because of the large capital investment
required.
2.1.3.4 Development status - Limestone wet scrubbing technology is
well established and widely applied in the utility industry. Currently,
there are several efforts focused on ways to improve performance, cost
effectiveness, reliability, and waste disposal for limestone systems. The
recent innovations worth mentioning are: the use of mass transfer additives,
forced oxidation, and advanced absorber designs.
Mass Transfer Additives - Both inorganic and organic additives have
been used to improve the S02 removal efficiency and reagent utilization of
limestone systems. These additives are used because of the limestone's low
solubility in water. This low solubility results in a low liquid phase
alkalinity, making it necessary to contact the acidic flue gas with large
slurry volumes. The liquid to gas ratio (L/G) for limestone systems is
typically 60 to 120 gallons per thousand cubic feet of gas, depending on
the S02 concentration and the desired removal level. Recirculating this
large amount of slurry consumes a large portion of the system's electrical
power. Maintenance on these large pumps is often difficult and time
consuming due to their size.
Several additives are being used commercially which increase the liquid
phase alkalinity of limestone systems. Magnesium oxide is currently the
most widely used additive. However, since chlorides effectively tie up the
magnesium, the application of MgO will be limited to open loop systems. In
closed loop systems, chlorides (originally present in the feed coal) are
concentrated in the recirculating slurry to high levels. Prescrubbers for
chloride control can mitigate this problem, but create other operating
problems such as wastewater disposal, water balance impacts, and chloride
stress corrosion. For these reasons, prescrubbers are not commonly
2-57
-------
employed for chloride control. Adipic acid and waste dicarboxylic acids
obtained during the production of adipic acid are also being used in utility
installations. Since these acids do not react with chlorides, their ability
to enhance S02 removal is not affected by the high chloride concentrations
that are sometimes encountered in closed loop operations.
Adipic acid is a commercially available dicarboxylic organic acid in
powder form, used primarily as a raw material in the nylon-manufacturing
industry and with some applications as a food additive. The capability of
carboxylic acids to improve SCL removal and limestone utilization has been
known for over 10 years. Most of the initial research in this area was
performed by Dr. G. T. Rochelle of TVA. More recently, EPA has sponsored
adipic acid testing at its RTF laboratory facility, the Shawnee Prototype
unit, a full-scale utility in Springfield, MO, and the Rickenbacker Air
National Guard Base (RANGB) near Columbus, Ohio.71 Dibasic acid (DBA) is a
by-product of adipic acid and costs about half as much. It is a mixture of
adipic, succinic, and glutaric acids. It has been shown to have the same
effects as adipic acid and is currently preferred because of its lower
cost.71
DBA effectively buffers the pH in limestone absorbers and improves the
SO^ removal efficiency. This buffering action limits the drop in pH at the
gas/liquid interface during absorption. The resulting higher concentration
of SO,, at the interface accelerates the liquid-phase mass transfer. Thus,
SOg absorption becomes less dependent on the limestone dissolution rate to
provide the necessary alkalinity. This makes it possible to achieve a
higher S02 removal efficiency at a lower L/G and limestone stoichiometry.
The optimum concentration range of DBA for effective S02 removal is at 700
to 1500 ppm with a pH greater than 5.2 at the scrubber inlet.86 However,
there are difficulties with increased degradation of DBA when pH's greater
than 5 at the scrubber inlet are used. These impacts, though unfavorable,
pC
are not seen to be a serious threat to the process. Preliminary economic
evaluations have shown that DBA can reduce both the capital investment and
the operating costs of limestone systems while simultaneously improving the
performance, even where the actual addition rate of DBA is three to five
2-58
-------
times the theoretical requirement due to the degradation of the acid.87 One
study shows a decrease of about 10 percent in total levelized costs for a
500 MW system firing a high sulfur coal even at high degradation levels.88
Studies indicate a substantially greater limestone utilization when
either adipic acid or DBA is used. The Shawnee test showed that at pH
levels lower than 5.2, the limestone utilization is usually greater than 85
percent for an adipic acid-enhanced limestone system, as compared to 65 to
70 percent utilization at the higher pH needed in unbuffered limestone
systems to achieve an equivalent SO,, removal. Thus, an adipic acid- or
DBA-enhanced system consumes less limestone, generates less waste sludge,
and reduces cost. In addition, high limestone utilizations contribute to
QQ
more reliable scrubber operation.
Forced Oxidation - Forced oxidation is a process modification in which
air is sparged into a reaction tank - usually the recycle tank - to oxidize
calcium sulfite ions to calcium sulfate ions. This improves system
operation by preventing scaling and by making the FGD sludge easier to
handle and dispose.
During the operation of first generation lime and limestone FGD
systems, calcium sulfate scaling on system internals was often a serious
problem. Oxygen from the flue gas reacted with sulfite ions and formed
sulfate ions. This "natural" oxidation is generally between 10 and 30
percent of the total S02 removed. It has been found that for oxidation of
less than 15 percent, calcium sulfate scaling does not occur. This is due
to a coprecipitation mechanism in which sulfate ions are interspersed
throughout the crystal lattice replacing sulfite ions. This coprecipitation
mechanism can keep the scrubbing solution subsaturated with respect to
calcium sulfate. At oxidation levels above 15 percent, the coprecipitation
mechanism is not capable of removing all of the sulfate ions from the
solution. Since very few calcium sulfate, or gypsum (CaS04 - 2H20), seed
crystals are present, crystal growth (scaling) on the system internals may
71
occur to reduce the relative saturation.
2-59
-------
One possible solution is to control the process such that the oxidation
fraction is less than 15 percent. However, since oxidation is a function
not only of flue gas oxygen levels but also of S02 concentrations, contactor
design, and other liquid and gas phase parameters, it is difficult to
control this simply by maintaining low excess air levels. An alternative to
operating with low oxidation fractions is, paradoxically, to operate at the
other end of the spectrum -- with high oxidation fractions. This provides
sufficient gypsum seed crystals in solution to prevent crystal growth on
scrubber and pipe surfaces. Additionally, sludge handling characteristics
with forced oxidation are greatly improved over unoxidized solids due to the
high concentration of gypsum. Gypsum is a structurally stable solid which
can be stacked to heights of over 100 feet for temporary and permanent
disposal.
In most forced oxidized systems, air is sparged into the reaction tank
at a stoichiometry of 2 to 4 moles of oxygen per mole of S02 removed. Other
methods of oxidation have been examined by TVA at the Shawnee pilot
scrubbers. At least one commercial vendor uses a double loop system to
produce gypsum. The second liquor loop, at a higher pH, is used to remove
the bulk of the S02 via oxidation. Excess liquor is passed to the front
where the pH drops and most of the oxidation takes place. This design also
promotes good limestone utilization.
Absorber Design - As discussed previously, the major trend in absorber
design has been away from packed contactors and towards open spray towers.
This is due to the relatively high reliability and easy maintenance of open
vessels. The classical spray tower, which uses small high-energy nozzles
and relatively low gas velocities, is not practical for limestone FGD
systems. However, spray towers with high L/G's developed for these
applications have been shown to exhibit exceptionally outstanding
performance. Very high S02 removal efficiencies (95 percent) have been
achieved with reliabilities approaching 100 percent.77 There is one major
problem with spray tower operation, however, and that is the mist
eliminator. In this part of the system, gas flow is restricted and the
2-60
-------
potential for plugging and scaling is increased. To reduce the severity of
this problem, various washing schemes have been developed.77
Two new absorber designs have recently been evaluated which offer
alternatives to the more conventional counter current design: the cocurrent
absorber, such as that recently evaluated by EPRI and TVA, and the jet
bubbling reactor used in the Chiyoda CT-121 process.71
2.1.3.5 Reliability. Reliability data for industrial limestone FGD
systems are scarce since only one system is currently operating. Scrubber
performance at the RANGB facility has generally been quite good except for
the early stages of operation during which several start up problems caused
significant amounts of downtime. From November 1976 through December 1978,
the RANGB system demonstrated that an industrial boiler limestone FGD system
can operate with high reliability. During this period, it operated about 95
percent of the the time, excluding the downtime caused by a severe
90
blizzard. It should be noted that part of its high reliability might have
been attributable to the fact that it operates only 6 months out of the year
and thus would provide more time for maintenance and repair than is typical
for other systems. Also, the unique design of RANGB's FGD system gives it
a more steady operation with a constant liquid-to-gas ratio. However, this
steady operation is achieved at the expense of higher electricity, solid
waste disposal, and reagent costs.
In a recent study performed for the EPA, 24 utility-size limestone
systems were evaluated. These systems had an average availability of only
73.5 percent and an average operability of 73.8 percent. The primary
components of failure and the percentage of system outages resulting from
these failures were: dampers (28 percent), duct systems (19 percent), fans
(17 percent), absorber towers (16 percent), and mist eliminators (9
percent). While this study affirms that reliabilities are in general
relatively low, one EPA test conducted in 1979 showed that reliabilities for
systems in which low sulfur coal is used can achieve exceptionally high
reliabilities. For example, one limestone FGD system which achieved better
than 95 percent reliability for a 30-day period on a 0.55 percent sulfur
2-61
-------
coal has achieved, along with its sister unit, essentially 100 percent
reliabilities over the last several years.91'92
Although many improvements in design and operation have been identified
to improve limestone systems, the conclusion reached by industrial clients
is that additional effort has to be expended in order to make limestone
systems acceptable where reliable steam supply is paramount. The
implication is that the extra risk/effort associated with limestone FGD is
offset by economic advantages on utility applications, but this economic
advantage is not present for industrial size applications. It is therefore
not surprising that the vast majority of industrial FGD systems are
sodium-based and not calcium-based.
2.1.3.6 Emissions data. Emissions data for limestone and
limestone-adipic acid systems for the industrial unit at RANGB are reported
in Section 4.2.5 of the BID. These data are from a 30-day test on a boiler
firing 3.5 percent sulfur coal. To summarize this section, 50-70 percent
S02 removal efficiency was achieved with limestone alone, and 94.3 percent
was achieved-when adipic acid was used. Actual long term emission data at
Springfield Utilities in Missouri confirm RANGB's test results.93 Without a
mass transfer additive the S02 removal efficiencies at that facility were
50-70 percent. With a mass transfer additive, target S0? removal
efficiencies of 80, 90 and 95 percent were achieved over periods ranging
from 7 to 30 days. One 31-day test at the TVA Shawnee Plant with adipic
acid gave an average of 96.1 percent S02 removal efficiency.93
2.1.4 Lime Wet Scrubbing
Although lime is about 100 times more soluble in water than limestone,
it still presents the same chemical problems that are inherent to all
calcium-based systems. Like limestone systems, there is only one lime
system currently operating in the industrial boiler market, and this system
has unique cost advantages. The plant at which this scrubber is located
uses the lime slurry blowdown to neutralize and precipitate metal ions out
of wastewater streams generated by other processes within the plant.
2-62
-------
According to the vendors of utility wet lime systems, it is unlikely that
very many, if any, new systems will be installed on industrial boilers over
pi Op QO
the next five years. ' ' This lack of demand in the industrial sector
is primarily because lime systems are affected by the same problems that
affect limestone systems as discussed in Section 2.1.3.
The owner of the only currently operating industrial lime system,
Pfizer, Inc., reports achievements of greater than 90 percent SO- removal
efficiency and 95 percent reliability. S02 removal efficiencies and
reliabilities for utility lime systems have typically been much lower for
comparable high-sulfur coal applications.
2.1.4.1 Process description. The process, chemistry, equipment, and
operations for this system are described in the March 1982 BID and will
therefore not be repeated here.
2.1.4.2. Factors affecting performance. Like limestone systems, lime
wet FGD systems are confronted with two major chemical-related problems
affecting their performance. These are the relative insolubility of the
reagent and the susceptibility of these systems to scaling or plugging.
These two problems are the major determinants in the design and operation of
lime and limestone systems.
Because they have similar chemistry, lime and limestone systems operate
similarly. For example, the L/G ratios are about the same (although L/G's
for lime systems are slightly less), and new lime scrubbers tend to be open
vessels rather than packed or tray towers.
However, since lime is more soluble than limestone, several of the
factors discussed in Section 2.1.3.2 are not accurate of or relevant to wet
lime scrubbing. For example, reaction tank residence times are typically
much shorter in lime-based systems compared to those in limestone processes.
Thus lime systems have smaller hold tanks resulting in lower capital and
operating costs. Unlike limestone systems, which show a much higher S0?
removal performance with simulated plug flow reactors, lime systems show
little improvement with these as compared to batch reactors. The pH of the
2-63
-------
slurry at the scrubber inlet is generally around 8 to 9 for lime systems as
86
opposed to 5-6 for limestone systems. The solids from the lime based
systems are not thixotropic and are much easier to dewater. Therefore,
forced oxidation does little to improve the handling characteristics of the
waste solids. Waste solids from lime systems are also more stable after
disposing to a landfill, but still may require fixation depending on
disposal site requirements. The use of excess reagent is not required
with lime systems because reagent utilization is typically above
95 percent. Also, tests have shown that mass transfer additives such as
dibasic and adipic acid, which have significantly improved the performance
of limestone systems, have little if any effect on utilization or SCL
removal for lime systems. However, magnesium oxide enhanced or Thiosorbic
lime (a particularly reactive lime) has proved to be effective as will be
discussed below.
Thiosorbic lime - Thiosorbic lime is a unique type of lime with a high
magnesium concentration (typically around 4 to 8 weight percent MgO).
Currently, a mine in Maysville, Kentucky is the only natural source of
97 98
Thiosorbic lime in the United States. ' Dravo Lime Company owns the
Maysville facility and has patents on all Thiosorbic lime systems, both
natural and synthetic. Since MgO is about 600 times more soluble in water
than corresponding calcium compounds, the amount of available alkalinity in
99
the scrubbing solution is increased with its use. Thus, for the same
system configuration, S02 removal efficiency will increase (over 90% S02
removal has been achieved). On the other hand, use of Thiosorbic lime will
reduce the required L/G ratio (and associated pumping costs) for a constant
S02 efficiency target. In addition, reliabilities have been substantially
higher due to a reduction in scaling. This is partially because MgO
enhances buffering in a pH range of 5.8 - 6.5, well below the pH at which
the onset of scaling occurs. Although waste disposal problems might be
anticipated with a more soluble reagent, the supplier states that 45-50%
97 98
solid sludges are routinely attained. '
2-64
-------
2.1.4.3 Applicability to industrial boilers. Lime processes have
found only limited application on industrial boilers and, like limestone
systems, are more applicable to larger utility size boilers for the same
reasons described in Section 2.1.3.3. The March 1982 Background Information
Document had reported three industrial wet lime systems in operation (Table
2.1-13). However, it was found out that Carborundum Abrasives scrubs only
for particulate matter and has always burned a compliance coal to meet S0?
regulations. The Armco Steel plant took its lime scrubber out of
operation around 1979-1980. The boiler on which it is installed currently
uses process waste gas when it is called upon to operate.101 The system at
Pfizer, which was developed by Pfizer and the National Lime Association, is
still operating. It should be noted that at least part of the reason for
installing a lime FGD system at Pfizer was to take advantage of the chemical
properties of the scrubber blowdown. This slurry stream, which has a solids
content of about 4 weight percent, is pumped to the plant's industrial
wastewater pretreatment unit. There, it is used to neutralize the
wastewater and precipitate metal ions generated by other processes within
the plant. The resulting sludge is concentrated in a vacuum filter and
hauled to a non-hazardous waste landfill.
In contrast to the limited application of wet lime FGD systems in the
industrial boiler sector, lime systems are the second most prevalent type of
FGD system for utility boilers. As of November 1982, 35 lime systems were
installed on utility boilers, representing approximately 18 percent of the
total number of utility FGD systems.84
2.1.4.4 Development Status. The lime wet scrubbing technology is well
established in the utility industry. Currently, several efforts are focused
on ways to improve reliability. For example, as with utility limestone
systems, lime systems are beginning to use spray towers to increase
reliability while maintaining, if not improving, SO- removal efficiency (see
Section 2.1.4.2). Of the six utility lime systems scheduled to begin
operation in the 1982-1984 period, the majority will have spray towers.
2-65
-------
TABLE 2.1-13. SUMMARY OF WET LIMF FGD SYSTEMS INSTALLED ON U. S
INDUSTRIAL BOILERS AS OF OCTOBER 1983
Process
Lime
Lime
Lime
Vendor
In-house design
Koch Engineering
Carborundum
Company/Location
Pfizer, Inc. „.
East St. Louis, IL q
Armco Steel f--, lm
Middletown, QHSJ'JUI
Ca rborundum/ Abras i ves
Buffalo, NYJjfl°°
Start-up
Date
1978
1975
1980
Number Of
FGD Units
1
1
1
Size
(acfm)
100,000
140,000
40,000
Fuel
Type Sulfur (%}
Coal 3.6
Coal 1.0 - 1.2
Coal 2.2
Current Status
Operational
Shut down
Uses a
Compliance Coal
o>
CT)
-------
However, unlike limestone systems, these lime systems will be using neither
mass transfer additives nor forced oxidation. This is because reagent
insolubility and thixotropic sludges are not as great a concern for lime
systems as they are for limestone systems.
There are currently 13 utility FGD systems using Thiosorbic lime, all
located in the Ohio River Valley. These plants along with other pertinent
go
information are listed in Table 2.1-14. As the table shows, availability
figures for these systems are about 95 percent on average. A number of
studies have shown 10-15 percent lower capital investment costs and lower
operating costs using Thiosorbic lime over comparable limestone scrubber
98
systems for electric utility plants.
The Pfizer plant does not currently use Thiosorbic lime. However,
Thiosorbic lime is being used in the industrial fluidized bed combustion
units at the Ashland Petroleum plant in Catlettsburg Kentucky.
2.1.4.5 Reliability. Reliability of lime FGD systems for industrial
boiler applications is difficult to assess because of little available data.
The one industrial application reports an availability of 95 percent.
According to the plant, the only major problem encountered with this system
has been plugging of the scrubber's inlet and outlet gas ducts.
Although there is a paucity of reliability information for industrial
wet lime systems, a substantial data base exists for utility installations.
In a study performed for the EPA using this data base, 23 lime systems were
evaluated. The lime systems were reported as having an average availability
of 84.1 percent and an average operability of 75.4 percent. As with the
limestone systems, the primary components of failure were dampers, duct
systems, fans, absorber towers, and mist eliminators; yet significant
improvements can also be expected with improved design, operating and
maintenance. Nevertheless, wet lime FGD systems are not expected to be able
to compete with sodium-based FGD systems in the industrial market for the
same reasons discussed in Section 2.1.3.5 for limestone systems.
2-67
-------
TABLE 2.1-14. THIOSORBIC LIHE APPLICATIONS TO UTILITY BOILER FGO SYSTEMS 98
CTv
GO
Plant/Location
BIG RIVERS
Green 1
Green 2
(at Sebree, Ky.)
CINCINNATI GAS ft ELECTRIC
East Bend 2
(at Rabbit Hash, Ky.)
COLUMBUS & SOUTHERN OHIO
Conesville 5
Conesville 5
(at Conesville. Oh.)
MONONGAHELA POWER
Pleasants 1
Pleasants 2
(at Willow Island. W. Va.)
PENN POWER
Bruce Mansfield 1
Bruce Mansfield 2
Bruce Mansfield 3
(at Shippingport, Pa.)
DUQUESNE LIGHT
•Phillips
(at South Heights, Pa.)
*E1rama
(at W. Elizabeth. Pa.)
WEST PENN POWER
Mitchell
(at Courtney. Pa.)
Turrentlu »/-hla.
-------
2.1.4.6. Emissions data. Industrial lime systems have achieved over
90 percent S02 removal efficiency. In a 30-day test at the RANGB FGD system
using lime as a reagent and operating on a 3.5 percent sulfur coal, the S02
removal efficiency was 91.5 percent. The actual emissions were 0.4
Ib S02/10 . Pfizer has reported an S02 removal efficiency for its system
to be above 90 percent for 3.5 percent sulfur coal; however, the actual S02
removal efficiency was not given, nor was the method used for determining
it.
2-69
-------
2.2 DRY PROCESSES
Dry processes that have potential applicability to industrial boilers
include spray drying of a lime or sodium reagent, dry injection of a sodium
reagent, electron-beam irradiation of flue gas containing ammonia or lime
and combustion of a palletized or pulverized coal and limestone mixture.
Each of these processes results in a dry product for waste disposal. The
use of the coal/limestone fuel mixture is discussed in Section 4.2.
2.2.1 Spray Drying
Spray drying FGD technology has developed rapidly over the past several
years and is an applicable S02 control method for all industrial boilers.
The technology is offered by more than 10 system vendors and 21 industrial
spray drying units have been sold. Seven of these units are currently
operational.
Spray drying involves contacting the flue gas with an atomized lime
slurry or a solution of sodium carbonate. The hot flue gas dries the
droplets to form a dry waste product while the absorbent reacts with S02 in
the flue gas. The dry waste solids, consisting of sulfite and sulfate
salts, unreacted absorbent and fly ash, are collected in a baghouse or ESP
for disposal.
2.2.1.1 Process description. A schematic diagram of the spray drying FGD
process is shown in Figure 2.2-1. Flue gas containing fly ash and SOp
enters the spray dryer and is contacted with a finely atomized alkaline
solution or slurry. During the approximately 10-second residence time in
the dryer, the flue gas is adiabatically humidified as the water in the
slurry or solution is evaporated. Simultaneously, flue gas S02 reacts with
the alkaline species to form solid sulfite and sulfate salts. The solids
formed are dried to generally less than 1 percent free moisture. The flue
gas, which has been humidified to within 11 to 28°C (20 to 50°F) of its
2-70
-------
Clean Gas to
Atmosphere
Hot or Warm Gas Bypass
Flue Gas
Combustion Air
Partial Recycle of Solids
U
-Water
Sorbent
Slurry
Tank
}
Product Solids ft
Fly Ash Disposal
Sorbent Storage
Figure 2.2-1. Typical Spray Dryer/Particulate Collection Flow Diagram
-------
adiabatic saturation temperature, passes through the dryer and into a high
efficiency particulate matter control device. In some system designs a
portion of the solids drops out of the dryer, but the bulk of the desulfuri
zation products are collected with fly ash in a baghouse or an ESP. The
most common reagent is lime, although sodium-based reagents are also used.
Atomization designs vary with regard to the use of rotary disk or two-fluid
nozzle atomizers, wheel speed in rotary atomizers, external or internal
mixing in nozzle atomizers and the number of atomizers per dryer.
The reaction between the alkaline material and flue gas S0? continues
as the gas passes through the ductwork and the baghouse or ESP. Reaction
mechanisms and mathematical models have been postulated for the lime spray
dryer process. ' The overall chemical reactions for lime and sodium
carbonate are shown below.
S02 + CaO + 1/2 H20 >» CaS03 1/2 H20 (2.2-1)
or
S02 + Na2C03 >* Na2S03 + C02 (2.2-2)
In addition to these primary reactions, sulfate salts are produced by the
following reactions:
S02 + CaO + 1/2 02 + 2H20 >* CaS04 2H20 (2.2-3)
or
Na2S03 + 1/2 02 >- Na2S04 (2.2-4)
S03 + Na2C03 >* Na2S04 + C02 (2.2-5)
Auxiliary equipment associated with the spray drying process includes a
reagent preparation system. Sodium carbonate reagent is prepared as a
concentrated solution in a stirred tank. In lime systems, pebble lime is
generally slaked in ball mill, paste or detention slakers, although ball
mill and paste slakers are more common.
2-72
-------
Reagent utilization can often be improved by recycle of the waste
solids, particularly in lime systems, where the unreacted reagent in the
waste solids can be reused. The recycle solids are either slurried
separately and added to the reagent feed just upstream of the spray dryer or
they are added directly to the fresh reagent holding tank.103'104
Additional advantages to the use of solids recycle include (1) a more easily
dried atomizer slurry because of a higher initial weight percent solids and
(2) reduced scaling potential compared to once-through lime systems with
feed slurries containing less than 10 percent solids.105'106 Disadvantages
to solids recycle include the added capital costs and operating complexity
associated with the solids recycle equipment. Solids recycle may not be
used on some systems, depending on the amount of unreacted reagent in the
waste solids and vendor or operator preference.
2.2.1.2 Factors affecting performance. The performance of a spray
dryer FGD system depends on several factors, the two most important being
the flue gas approach to saturation temperature at the dryer outlet and the
amount of reagent added per unit of inlet S02 (reagent ratio). Unlike wet
scrubbing systems, the amount of water that can be added to the flue gas
(the L/G ratio) is set by heat balance considerations for a given inlet flue
gas temperature and approach to saturation. Typical L/G ratios range from
0.03 to 0.04 £/m3 (0.2 to 0.3 gal/1000 ft3). The amount of reagent added
(reagent ratio) is varied by raising or lowering the concentration of a
solution (sodium system) or weight percent solids of a slurry (lime system)
containing this set amount of water. While holding other parameters such as
temperature constant, S02 removal increases with increasing reagent ratio.
However, as the reagent ratio is increased to raise the level of S0?
removal, two limiting factors are approached:
- Reagent utilization decreases, raising reagent and disposal costs.
- An upper limit is reached for the solubility of the reagent in the
solution, or for the weight percent of solids in the slurry (due to
pumpability considerations).
2-73
-------
There are at least two methods of circumventing these limitations. One
method is to utilize solids recycle, using the solids that have dropped out
in the spray dryer or collected in the particulate emission control device.
Recycle has the advantage of increasing reagent utilization, and it can also
increase the opportunity for utilization of any alkalinity in the fly
The second method of avoiding the above limitations on S0? removal is
to operate the spray dryer at a lower outlet temperature; that is, a closer
approach to saturation. Operating the spray dryer at a closer approach to
saturation has the effect of increasing both the residence time of the
liquid droplets and the residual moisture level in the dried solids. As the
approach to saturation is narrowed, S02 removal and reagent utilization
increase dramatically.
The approach to saturation at the spray dryer outlet is set by either
the requirement for a margin of safety to avoid condensation in downstream
equipment or restrictions on stack temperature. The design approach to
saturation for spray drying systems generally ranges from 10 to 28°C (18 to
50°F). Operation at a relatively close approach to saturation, 10 to 14°C •
(18 to 25°F), is common for applications where S02 removal requirements
approach 85 to 90 percent. However, operation at a close approach to
saturation may also be used to decrease reagent use in lower efficiency
applications.
Some spray dryer system designs, particularly on large utility boilers,
allow for warm or hot gas bypass around the spray dryer to reheat the dryer
outlet gas (see Figure 2.2-1). Warm gas (from downstream of the boiler air
heater) can be used at no energy penalty, while the use of hot gas (upstream
of the air heater) has an energy penalty associated with the decrease in
energy available for air preheat.
Another factor that may affect the performance of spray drying systems
is the inlet flue gas temperature. For inlet flue gas temperatures
significantly below approximately 121°C (250°F), S02 removal may be limited
by the amount of water and reagent that can be added to the flue gas. The
2-74
-------
limiting inlet temperature for a particular system depends on the fuel
sulfur content, desired SO^ removal and reagent quality.
Spray dryer system performance can also be affected by the choice of
the particulate collection device. Baghouses have been chosen over ESP's in
most commercial spray drying applications. Baghouses have an advantage over
ESP's in that unreacted alkalinity in the solids and fly ash collected on
the filter bag surface can react with the remaining SCL in the flue gas as
the gas passes through the baghouse. Pilot studies have shown that SCL
removal across the baghouse may account for 15 to 20 percent of the overall
SO? removal, depending on reagent ratio, approach temperature and baqhouse
107
pressure drop. Data from recent tests on a full-scale (110 MWe) utility
system show baghouse S02 removals ranging from 9 to 10 percent of overall
removal during low sulfur (1.2 percent) coal testing and
13 to 15 percent during high sulfur (3.5 percent) tests. Overall SOp
removal during these tests was 90 percent and the system operated at a
10°C (18°F) approach to saturation.108
The factors that are important in making the choice between ESP's and
fabric filters include:
- Use of solids recycle (increased dust loading increases the size
and cost of an ESP).
- Fly ash resistivity (high ash resistivity often requires larger,
more expensive ESP's).
- Pressure drop considerations (an ESP will result in lower pressure
drop costs than a fabric filter).
Baghouse designs for spray dryer applications vary primarily with
regard to bag fabric, cleaning frequency and cleaning mode. Thirteen of the
21 industrial spray drying units sold will use pulse-jet baghouses; the
others will use reverse-air baghouses.
2.2.1.3 Applicability to industrial boilers. Spray drying FGD is an
applicable S02 control method for all industrial boilers. Early development
work on spray drying systems demonstrated applicability to boilers firing
2-75
-------
low to medium sulfur fuels (less than 3 percent sulfur).109 Recent test
results reported for two industrial spray drying systems and a small utility
system (100 MWe) show that the S02 control method is applicable to high
sulfur (3 to 4 percent sulfur) fuels as well.
For spray drying systems using sodium carbonate as the reagent,
disposal of the waste product may entail additional requirements. The waste
consists of highly soluble sodium salts, such as Na2$03 and Na-SCL. Land
disposal of the waste solids may require clay- and/or plastic-lined
landfills in areas where the potential exists for groundwater contamination.
All industrial boiler spray drying systems sold so far will be or are
currently using lime as the reagent.
2.2.1.4 Development status. Spray drying technology for removing S0?
from boiler flue gas has developed rapidly over the past several years. The
technology is commercially offered by more than 10 system vendors and
21 spray drying FGD units have been sold for industrial boiler applications.
Seven of these units are currently operational and four other units are
expected to be in start-up by the end of 1983.
The commercial systems sold for industrial boiler applications are
summarized in Table 2.2-1. These systems are being applied to boilers
burning coals with a fairly wide range of sulfur contents (0.6 to
3.5 percent sulfur). The systems have S02 removal guarantees ranging from
70 to 90 percent and at least five of the systems have outlet S00 emission
fi ^
guarantees for a maximum of 520 ng/J (1.2 lb/10 Btu) or lower.
In addition to the systems for industrial boilers, 17 utility spray
drying systems have been sold. The applications range in size from 44 to
860 MWe and total about 6,800 MWe in FGD system capacity. The utility
systems are being applied to low sulfur (less than 2 percent) coal-fired
units and S02 removal guarantees from the vendors are as high as 90 percent.
Six of the utility systems are operational and one system is in the initial
startup stages.110"113
2-76
-------
TABLE 2.2-1. SUMMARY OF INDUSTRIAL BOILER SPRAY DRYING SYSTEMS
113
System Purchaser/Location
Argonne National Laboratory
Argonne, IL
Strathmore Paper Company
Uoronoco, MA
Celanese Fiber Company
Cumberland, MD
Container Corporation
Philadelphia, PA
University of Minnesota:
Units 1 & 2
Minneapolis, MN
Austell Box Board Co.
Austell , Georgia
General Motors Buick
Division
Flint, MI
Fairchild Air Force Base:
Units 1, 2, & 3
Spokane, MA
Puget Sound
Naval Shipyard:
Units 1, 2, S 3
Bremerton, WA
Maelstrom AFB:
Units 1, 2 & 3
Grent Falls, MT
Griffis AFB
Units 1,2,3 « 4
Rome, NY
Vendor3
Niro/Joy
Mikropul/Koch
Engineering
Rockwell Int./
Wheel abrator-Frye
Ecolaire, Inc.
Flakt, Inc.
Wheel abrator-Frye
Niro/Joy
Niro/Joy
G. E. Environmental
Services
Niro/Joy
Ecolaire, Inc.
Size, Mg/hr
(Ib/hr) Steam
77
(170,000)
39
( 85,000)
50
(110,000)
77
(170,000)
40 Mile6
114
(250,000)
204
(450,000)
50
(110,000)
64
(140,000)
each
41
(90,000)
hot water each
41
(90,000)
each
Coal Data
Type Sulfur
Illinois 3.5%
bituminous
Eastern 2.3 to 3%
bituminous
Eastern 2% maximum
subbituminous
Eastern U
subbitu-
minous
Subbitu- 0.6 to 0.7*
mi nous each
Bituminous 1.0 to 2.5%
Indiana 1 to 3%
bituminous
Western 1%
subbituminous
NA 1.6% maximum
Western 1.0*
subbituminous
Eastern 3.0%
bituminous
SO, Removal
Removal Outlet
78.7%
75%
70% for 1% S
coal; 86% for
2% S coal
Design removal
of 90%
70% f
Varies with
sulfur content
70 to 90%
85%
84%
85%
85%
Guarantee*"
(lb/106 Btu)
1.2
1.2
70 Ib/hr SO,,
outlet c
NAe
NA
1.2
1.2
NA
NA
NA
0.71
Startup Date/Status*1
Operational. Turned over
purchaser.
Operational. Turned over
purchaser.
Operational. Turned over
purchaser.
Operational. Turned over
purchaser.
to
to
to
to
One unit operational; second
startup in September 1983.
Operational. Not turned
to purchaser.
Operational. Not turned
to purchaser.
Initial startup stages.
Late 1987.
Spring 1985.
Late 1984
over
over
Niro/Joy = Niro Atomizer Inc./Joy Western Precipitation Division.
Electrical output, part of cogeneration system.
cWhere guarantee information not available, design values are reported.
dAs of October 1983.
eNA = not available.
At reagent ratio of 1.0.
-------
2.2.1.5 Reliability. Reliability of industrial spray drying systems
is difficult to assess because only four systems have been operational for a
long period of time. These are operated by Strathmore Paper, Celanese
Fibers, Argonne National Lab and Container Corporation. The data available
indicate that lime spray drying FGD systems applied to industrial units are
reliable when operating at S02 removal efficiencies in the 60 to 75 percent
range on both bituminous and subbituminous coals of 3 weight percent sulfur
or less.
Availability of the spray dryer system at the Strathmore Paper Company
has been quite high except during the early stages of operation. Availa-
bility is defined as the percentage of hours that the FGD system is
available for operation (whether used or not) divided by the hours in the
period. Initial startup problems in late 1979 resulted in significant
amounts of downtime. These problems were a result of poor initial spray
dryer design combined with an actual gas flow that was 25 to 35 percent
higher than the design flow. However, following system design modifications
in March 1980, the system operated for nearly 1.5 years with approximately
80 percent overall system availability while achieving 70 percent S0?
removal on a bituminous coal of 3.0 weight percent sulfur. Strathmore
subsequently switched to a low sulfur coal, 1 percent, and lowered the S0?
removal to 60 percent. The system has operated in this mode for 1.5 years
and experienced 94 percent availability during this period.1 The system
normally operates 24 hours per day throughout the year. Sudden and wide
variations in boiler load are common at the plant because of changes in
process steam demand. These load changes are reported to have little effect
on the spray drying system and downstream baghouse.
The spray dryer system at the Celanese Fibers Company also showed
relatively low availability (65 percent) during the early stages of
operation. Initial operating problems were related to variable coal
quality, slurry feed pump wear, ineffective grit removal and atomizer slurry
maldistribution. Solution of these problems required minor modifications in
system design and operation. Following these modifications, system
availability averaged between 90 and 95 percent for the period from October
2-78
-------
1980 to mid-1982.113'116 During this period the FGD system averaged
70 percent SO- removal on a subbituminous coal with an average 1.0 weight
percent sulfur content. Solids recycle was not employed at this site but a
fabric filter was used for particulate matter control.
Startup of the Argonne National Laboratory system began in November
1981 and the system was fully operational in February 1982. Problems
encountered during the startup involved auxiliary equipment such as slurry
pumps, agitators and blowers. During the past year, the system has
operated with approximately 80 to 85 percent availability, while achieving
about 80 percent SO^ removal on a 3.5 weight percent sulfur coal, excluding
two major down periods. If these two major down periods are included the
availability drops to 56 percent. The first down period resulted from
delamination of 40 percent of the filter bags after nine months of
operation. These felted fiberglass bags were replaced with woven fiberglass
bags. The other major down period was also the result of a baghouse
failure. This facility is the only one of the four to operate with solids
recycle.
The spray drying system at Container Corporation of America has
operated for 2.5 years achieving 75 percent SO- removal with 0.6 percent
sulfur coal. The availability has steadily increased since start-up and the
overall availability during this period has been 80 percent. The primary
operating problem has involved failure of the atomizer. The plant keeps a
spare atomizer on site and can change atomizers very quickly thereby
114
minimizing downtime.
2.2.1.6 Emission Data. Recently available emissions test data for
four industrial spray drying systems are shown in Table 2.2-2. As shown in
Table 2.2-2, outlet S02 emission rates of less than 366 ng/J (0.85 lb/106
Btu) were achieved with all four systems. Results of short-term tests show
S02 removal efficiencies above 90 percent were achieved at locations A, B,
and D for coals ranging from 0.6 percent to 3.8 percent sulfur. No
long-term continuous monitoring data for these systems are currently
available. Comparison of the data presented for locations A and B shows a
2-79
-------
TABU 7.2-2. SUMMARY OF EMISSION DATA FOR FOUR INDUSTRIAL LIME SPRAY DRYING FGD SYSTEMS
EPA Method 6 Test Results
Location
A
A
A
B
C
D
No. of
Runs
3
3
3
6
6
3
Inlet S00 (nq/J)a Outlet SO,, (nq/J)a
Average
2,877
2,550
2,630
2,316
1.4306
516e
Range Average
NAC 585
NA 258
NA 116
NA 176
323 - 452 366
12.5 - 17.2 14.3
% SO, Removal
Average
79.7
89.9
95.6
92.4
74f
97. 2f
Boiler
Load
35%
70%
82%
75%
100%
NA
Reagent
Rat1ob
0.8
1.5
2.0
1.9
NA
NA
Coal Sulfur
Content
3.0%
3.0%
3.0%
3.8%d
1.5 - 2.5%
0.6%
Approach
Temperature
°C (°F)
13
13
13
14
19
(23)
(23)
(23)
(25)
(35)
NA
Solids Recycle
Rate (kg Solids/
kg Lime Feed)
2:1
2:1
2:1
None
None
None
no
Moles of calcium per mole of inlet S0?.
CNA = not available.
Coal/oil mixture with 94.2% coal hent input.
Estimated from coal properties.
Estimated from coal properties and measured outlet emission rate.
-------
considerably lower reagent requirement to achieve 90 percent SCL removal on
the system with solids recycle. This is expected since the use of solids
recycle improves reagent utilization.
2.2.2 Dry Alkali Injection
In the dry injection process, a dry alkaline material is injected into
the flue gas just ahead of a particulate control device. The alkaline
material reacts with the S02 in the flue gas and the solids and fly ash are
collected for disposal.
Dry injection technology has been developed through pilot and
laboratory scale studies but is not yet commercially applied to industrial
boilers. Application of the technology is planned, however, for a 500 MWe
utility boiler.
2.2.2.1 Process description. A generalized flow diagram of the dry
alkali injection process is shown in Figure 2.2-2. Dry injection schemes
generally involve pnuematic injection of a dry, powdery sodium-based reagent
into the flue gas with subsequent particulate collection in a baghouse. The
point of alkali injection has been varied from the boiler furnace all the
way to the inlet of the baghouse. Although other alkaline reagents, such as
lime, limestone and magnesium dioxide, have been tested, only certain sodium
compounds have shown the capability for high S02 removal from the flue gas.3
Both baghouse and ESP collection devices have been tested with dry injection
processes. However, the effect of the reaction between unspent reagent on
the filter bag surface and S02 remaining in the flue gas seems overwhelm-
ingly to favor the bag collector.
Nahcolite and trona ores, which contain naturally occurring sodium
compounds appear to be the most promising reagents for dry injection in
iiq ion
terms of reactivity and cost. 5l Nahcolite, which contains 70 to
90 percent sodium bicarbonate (NaHC03) has been shown to be more reactive
with S02 in flue gas than trona ore (Na2C03 NaHC03 2H20).121'122
2-81
-------
The principle reaction product from nahcolite and trona injection is
sodium sulfate (Na-SOj, according to the following overall
123 1?4
reactions:1"'1^
2NaHC03 + S02 + i02 ^ Na2$04 + 2C02 + H20 (2.2.3)
2(Na2C03 NaHC03 2H20) + 3S02 + 3/2
(2.2.4)
Prior to reaction with SCL, it appears that both nahcolite and trona
I pC
must undergo a decomposition step as shown below.
2NaHC03 ^ Na2C03 + H20 + C02 (2.2.5)
2(Na2C03 NaHC03 2H20) > 3Na2C03 + C02 + 5H20 (2.2.6)
The decomposition reaction increases the porosity and reactive surface
area of the reagent particles. The SCL reaction proceeds as follows:
(2.2.7)
2.2.2.2 Factors Affecting Performance. In addition to reagent type,
major factors affecting SO,, removal by dry injection include the amount of
reagent added (stoichiometric ratio), the temperature at the point of
injection and the size of the reagent particles.
As expected, the removal of SCL by dry injection increases with
increasing "normalized stoichiometric ratio" (equivalent moles of Na?0 per
mole of inlet SO,,) because additional reagent is available to react with the
SCL. However, higher stoichiometric ratios also result in lower reagent
126
utilization.
Nahcolite and trona undergo a decomposition prior to reaction with SCL;
the temperature at the point of reagent injection affects the rate of this
decomposition. In general, injection at higher temperature increases the
2-82
-------
i
oo
co
Reagent
Storage
Air Preheater
—tXh
X
Baghouse Compartments
Reagent
Holding
Bin
Injection
Fan
Figure 2.2-2. Dry Alkali Injection Flow Diagram
-------
decomposition rate and increases the initial rate of reaction with
127 128
S02- ' The evolution of H,,0 and C(X, during decomposition increases the
pore volume of the particles, creating more surface area for chemical
reaction and a lower resistance for S02 diffusion.128 As the reaction of
S02 and Na2C03 proceeds, it appears that the pores begin to plug; the
reaction then becomes limited by diffusion of SCL into the particle.129
Injection of the reagent at too low a temperature will reduce the
initial rate of S02 reaction and may limit the overall S0? removal
achievable with the dry injection system. For nahcolite, it appears that
S0? removal may drop off dramatically below an injection temperature of
125
approximately 135°C (275°f). 3 The minimum injection temperature for trona
is currently unknown but is estimated to be below 93°C (200°F).21 Injection
of sodium compounds at too high a temperature (about 343°C, 650°F) reduces
1 OC
their reactivity due to particle sintering.
Another factor affecting S02 removal and reagent utilization in dry
injection systems is particle size. In general, pilot and laboratory scale
studies have shown that higher SCL removals are obtained with smaller
particles. The majority of these studies were conducted with particles
1 OQ
ranging in size from 30 to 200 microns.
2.2.2.3 Applicability to industrial boilers. Dry alkali injection is
an applicable S02 control method for industrial boilers firing fuels with
low to moderate sulfur contents (up to 2 percent sulfur). The applicability
of dry injection to boilers firing higher sulfur fuels is difficult to
assess because limited data are currently available.
As with sodium-based spray drying systems, the high solubility and
leaching potential of the sodium waste solids may require special disposal
handling techniques. Land disposal of the solids in clay- and/or
plastic-lined landfills may be called for in areas with potential for
groundwater contamination.
2.2.2.4 Development Status. Dry alkali injection technology has not
yet been commercially applied to either industrial or utility boilers.
However, the first planned commercial application of trona injection has
2-84
-------
been announced for a 500 MWe utility installation scheduled for startup in
1990, Numerous pilot and laboratory scale studies have been conducted on
124
the technology. Demonstration scale tests were recently executed on a
small utility boiler (22 MWe) firing a low-sulfur western coal (0.44 percent
122 127
sulfur). ' Four to eight hour tests on this system showed that SOp
removals of 70 and 90 percent can be achieved with nahcolite at stoichio-
metric ratios of approximately 0.8 and 1.1 respectively. For trona ore,
this same system showed SO,, removals of 70 and 90 percent at stoichiometric
ratios of 1.3 and 2.4 respectively. During the testing, the normal baghouse
inlet temperature ranged from 143 to 149°C (290 to 300°F).122'127
The application of trona dry injection had been previously constrained
by questions regarding S02 removal limitations and cost. However, the
recent demonstration-scale studies have shown that SOp removal efficiencies
of 70 to 80 percent can be achieved with trona on low sulfur coals at
reasonable stoichiometric ratios. Trona ore is currently mined in large
quantities for conversion to sodium carbonate.
The application of nahcolite dry injection has been constrained by
uncertainties regarding reagent cost and availability. Nahcolite is
currently not mined in the United States, but at least one firm has
announced intentions to develop a nahcolite mining operation and several
other companies are investigating the possibility of supplying nahcolite
IOC 1O9
through solution mining techniques. ' However, a market commitment for
a minimum of 909,000 Mg/yr (1,000,000 ton/yr) of nahcolite may be necessary
to off-set the large capital investment associated with opening a commercial
mine. ' This production level corresponds to the nahcolite demand of
5000 MWe of utility generating capacity burning 1 percent sulfur coal with
i 90
70 percent SO- removal.
2.2.2.5 Reliability. Since dry alkali injection has not yet been
commercially applied, no data are available on the reliability or
operability of these systems. However, due to their inherent mechanical and
chemical simplicity, dry injection systems are expected to be at least as
reliable and operable as wet scrubbing systems and spray drying systems.
2-85
-------
2.2.3 Electron-beam Irradiation
Electron-beam (E-beam) irradiation processes are still in the very
early stages of development. These processes involve the irradiation of
flue gas containing a reactant, such as ammonia or lime. The process
removes both SC^ and NOX from the flue gas and produces a dry waste product
that must be subsequently removed in a particulate collector.
2.2.3.1 Process description. A schematic diagram of the
E-beam/ammonia process is shown in Figure 2.2-3. In this process, incoming
flue gas is cooled and humidified in a water quench tower, resulting in a
gas moisture content of about 10 percent. Ammonia is injected into the
cooled gas and the gas is passed through an E-beam reactor. In the reactor,
oxygen and water are ionized to form the radicals [HO], [0] and [H02] by the
application of electrons at a dose of 1 to 3 Mrads (1 Mrad is equivalent to
10 joules/g of flue gas). These radicals react with S09 and NO to form
C- X
sulfuric acid (^SO^) and nitric acid (HMO-,). The acids are neutralized by
ammonia and water in the flue gas to form solid ammonium sulfate ((NH.LSOJ
and ammonium sulfate nitrate ((NH4)2$04 2 NH4N03). The reaction time for
formation of the sulfate and nitrate salts is less than one second. Product
solids are collected in a hopper below the E-beam reactor or in a downstream
particulate collector.
In another version of the E-beam process, the water quench tower is
replaced with a lime-based spray dryer (see Section 2.2.1). Reactions in
the E-beam reactor occur in the same manner as above except that the
products formed are calcium salts (CaSO,, Ca(NO,)0 and CaSO,) instead of
135 J
ammonium salts.
Factors impacting S09 and NO removal by electron-beam irradiation
C- A
include gas moisture content, gas temperature, oxygen content, reagent ratio
and electron dosage. In addition, efficient penetration of the gas stream
by the beam requires a unique discharge pattern and other special design
considerations.
2-86
-------
Ammonia
ro
i
oo
Flue
Gas
f i I i f
•Quench Water
E-gun
E-beam
Reactor
Particu-
late
Collector
ID
Booster
Fan
-^-Product Solids
Drain
Figure 2.2-3. E-beam/ammonia process flow diagram.
-------
2.2.3.2 Status of Development. The electron-beam process is in an
early developmental state. The process has not yet been applied to a real
coal-fired flue gas. However, pilot studies on both the lime and ammonia
based E-beam process configurations are currently underway. The DOE has
signed cost sharing agreements with both Research-Cottrell and the joint
1 "?fi
venture EBARA/Avco-Everett. These pilot systems will treat flue gas from
coal-fired boilers.
Research-Cottrell will evaluate the E-beam/lime slurry process with a
10,000-acfm pilot plant currently being installed at TVA.137 NO and S0?
A £
removal optimization tests will be conducted at electron irradiation rates
between 0.5 and 1.5 Mrad. During the scheduled 2-year program, Research-
Cottrell will also conduct nitrate fixation tests and electron-gun cost
138
reduction studies. Research-Cottrell performed bench-scale studies on
the E-beam process under DOE Funding in 1979 and recently developed a
mathematical model for the E-beam/lime slurry process.
EBARA/Avco-Everett will conduct 10,000 to 20,000 acfm pilot studies on
the E-beam/ammonia injection process. A host site for this study is still
being negotiated. Current plans are to conduct the testing on flue gas
from a high-sulfur eastern coal. Following optimization and reliability
testing, EBARA and Avco plan to investigate the potential use of waste
138
products from the process as fertilizer.
The EBARA Manufacturing Company in conjunction with Japan Atomic Energy
•3
Research Institute (JAERI) has operated a 1000 Nm /hr pilot plant treating
flue gas from an oil-fired boiler. In 1976, EBARA tested a 3000 Nm3/hr
pilot plant on the off-gas from an iron ore sintering furnace at Nippon
Steel.140
2-88
-------
2.3 REFERENCES
1. Struthers - Anderson Pollution Control Systems. Points to Consider in
Evaluation of SCL Emission Control Systems for Steam Generators in the
Oil Fields. Atlanta, Ga. November 1980. pp. 4-5, 7, 12-13, 14-15.
2. Durkin, T. H., Southern Indiana Gas and Electric Company, et al.
Operating Experience with a Concentrated Alkali Process. (Presented at
the American Power Conference, 1980).
3. Memo from Berry, R. S., Radian Corporation, to Charlie Sedman and list
of Addresses. November 11, 1983. Notes from September 8 meeting with
Jack Brady of Andersen 2000 and from subsequent phone conversations
with him.
4. Memo from Shareef, G. S., and Berry, R. S., Radian Corporation, to the
Industrial Boiler S02 Docket. December 30, 1983. Development of
Material Balance for the Sodium Scrubbing FGD Algorithm.
5. Memo from Berry, R. S., and Maddox, J. A., to the Industrial Boiler S0?
Docket. February 15, 1984. Sodium Scrubber Data Sample and Analysis.
6. Oestreich, D. K. Equilibrium Partial Pressure of Sulfur Dioxide in
Alkaline Scrubbing Processes. (Prepared for the U.S. Environmental
Protection Agency, Office of Research and Development.) Washington,
DC. Publication No. EPA600/2-76-279. October 1976. p. 13.
7. Technical Note from Berry, R. S., Radian Corporation, to the Industrial
Boiler SOp Docket. May 31, 1984. SO- Re-emissions from the Sodium
Scrubbing Wastewater Stream in Aerobic Environments. Chapter 3.
8. Technical Note from Berry, R. S., Radian Corporation, to the Industrial
Boiler S02 Docket. January 25, 1984. Update of the Sodium Scrubber
Wastewater Issue, p. 6.
9. Brady, J. D., Andersen 2000. Particulate and S02 Removal with Wet
Scrubbers. Chemical Engineering Progress, pp. 73-77. June 1982.
10. Letter from Mayrsohn, H., Kern County Air Pollution Control District,
to Berry, R. S., Radian Corporation. August 9, 1983. Sodium Scrubber
Source Data.
11. Reference 1, p. 7.
12. Reference 1, pp. 4-5.
13. Reference 1, pp. 12-13.
2-89
-------
14. Letter from Clum, D. N., Heater Technology, to Berry R. S., Radian
Corporation. August 5, 1983. Heater Technology Response to Radian's
Letter.
15. Cornell, C. G., and Dahlstrom, D. A., Environtech. Sulfur Dioxide
Removal in a Double Alkali Plant. Chemical Engineering Progress.
pp. 47-53. December 1973.
16. Brady, J. D., Struthers - Andersen Pollution Control Systems Emission
Control for Oil-Fired Steam Generators. Atlanta, GA. November 14,
1979. pp. 19, 30.
17. Memo from Berry, R. S., Radian Corporation, to Meeting Attendees and
Durkee, K. R., EPA. July 1, 1983. Revised notes from June 7, 1983 FMC
meeting on the Dual Alkali Technology.
18. Letter from Seipp, D. D., FMC-Wyoming, to Berry, R. S., Radian
Corporation. August 11, 1983. Response to sodium scrubbing letter.
19. Letter from Cameron, H., General Motors, to Berry, R. S., Radian
Corporation. October 20, 1983. Response to sodium scrubbing letter.
20. Letter from Brady, J. D., Andersen 2000, to Berry, R. S., Radian
Corporation. July 28, 1983. Response to sodium scrubbing letter.
21. Letter from Anderson, D. F., Grace Petroleum to Berry, R. S., Radian
Corporation. October 10, 1983. Response to sodium scrubbing letter.
22. Letter from Souza, M., Bradford Dyeing Association, to Berry, R. S.,
Radian Corporation. November 10, 1983. Response to sodium scrubbing
letter.
23. Letter from Vossler, D. A., Union Oil, to Berry, R. S., Radian
Corporation. September 7, 1983. Response to sodium scrubbing letter.
24. Letter from Segnar, C. N., Chevron USA, Inc., to Berry, R. S., Radian
Corporation. November 1, 1983. Response to sodium scrubbing letter.
25. Letter from Maxwell, G., Cranston Print Works, to Berry, R. S., Radian
Corporation. September 1983. Response to sodium scrubbing letter.
26. Letter fro;r Borenstein, M., Neptune/Airpcl, to Berry, R. S., Radian
Corporation. September 14, 1983. Response to sodium scrubbing letter.
27. Letter from Borenstein, M., Neptune/Airpol, to Berry, R. S., Radian
Corporation. June 29, 1983. Response per previous phone conversation
with him.
2-90
-------
28. Letter from Maddox, J. A., Radian Corporation, to Camponeschi, B., FMC.
May 23, 1984. Confirmation of telephone conversation between
B. Camponeschi and R. S. Berry (Radian) on June 20, 1983.
29. Letter from Morgester, J. J., California Air Resources Board to Berry,
R. S., Radian Corporation. August 9, 1983. Source test data.
30. Memo from Read, B. S., and Jennings, M. S., Radian Corporation, to
Larry Jones, EPA. October 8, 1982. FGD Survey, Work Request B.
31. Francis, D. V., ARCO Chemical Company, and R. V. Biolchini, FMC
Corporation, Double Alkali Flue Gas Desulfurization Retrofit on an
Industrial degeneration Facility. Proceedings: Symposium on Flue Gas
Desulfurization. Volume 1. (Prepared for the U.S. Environmental
Protection Agency, Industrial Environmental Research Laboratory and
Electric Power Research Institute.) March 1983. pp. 183-200.
32. Letter from Anton, E. C., California Water Resources Control Board to
Maddox, J. A., Radian Corporation. June 7, 1984. Confirmation and
corrections to telephone call report dated August 26, 1983 between E.
C. Anton and R. S. Berry (Radian Corporation).
33. National Academy of Sciences. Water Quality Criteria 1972. (Prepared
for the U.S. Environmental Protection Agency.) Washington, DC, 1972.
p. 89.
34. Reference 8, Chapter 3.
35. U.S. Environmental Protection Agency. Quality Criteria for Water.
Washington, DC. Publication No. EPA-440/9-76-023. 1976. p. 337.
36. Memorandum from Berry, R. S., Radian Corporation, to Industrial Boiler
S02 Docket. July 20, 1984. Potential Water Quality Impacts of Trace
Metals in the Wastewater from Sodium Scrubbers Installed on Coal-Fired
Industrial Boilers.
37. Letter from Hereth, M., Radian/Washington, to Berry, R. S., Radian/RTP.
October 1983. Projected Promulgation Dates for Wastewater Standards to
Point Sources.
38. U.S. Environmental Protection Agency. Development Document for
Effluent Limitations Guidelines and Standards for the Steam Electric
Point Source Category. Publication No. EPA No. 440/1-80/029-b.
September 1980.
39. Technical Note from Martinez, J. A., Radian Corporation, to the
Industrial Boiler S02 Docket. April 2, 1984. Emissions from Sodium
Scrubbing Wastewater Streams in Sewer Systems. Chapter 1.
2-91
-------
40. Letter from Calderwood, C., Kern County APCD, to Maddox, J. A., Radian
Corporation. June 4, 1984. Confirmation of information collected in
July 25, 1983 telephone conversation between Calderwood and R, S.
Berry, Radian Corporation.
41. Letter from Maddox, J. A., Radian Corporation, to Vossler, D., Union
Oil Corporation. May 23, 1984. Confirmation of information collected
in the September 25, 1983 telephone conversation between Vossler and R.
S. Berry (Radian).
42. U.S. Environmental Protection Agency. Background Information Document
for Industrial Boilers. EPA Contract No. 68-02-3058. March 25, 1982.
pp. 4-69.
43. Weast, R. C. CRC Handbook of Chemistry and Physics. 57th Edition.
CRC Press, Cleveland, OH. 1973. pp. B-158 - B-163, B-128.
44. Memo from Martinez, J. A., et al., Radian Corporation, to the
Industrial Boiler SO- Docket. March 26, 1984. Compilation of S0?
Emission Data Obtained by EPA Approved Methods from Sodium Scrubbing
Systems.
45. Sachtschale, J. R., Santa Fe Energy Company, and J. F. Dydo, FMC
Corporation, Operation and Performance of a Double-Alkali Scrubber.
Journal of Petroleum Engineering, pp. 2630-2632, 2633-2635. November
1982.
46. Chang, J. C. S. and Dempsey, J. H., Acurex Corporation, and Norm
Kaplan, U. S. EPA. Pilot Testing of Limestone Regeneration in Dual
Alkali Processes. Proceedings: Symposium on Flue Gas Desulfurization.
Volume 1. (Prepared for the U.S. Environmental Protection Agency and
Electric Power Research Institute.) March 1983. pp. 201-222.
47. Valencia, J. A., Lunt, R. R., and Ramans, G. J., Industrial
Environmental Research Laboratory. Project Summary. Evaluation of the
Limestone Dual Alkali Prototype System at Plant Scholz: System Design
and Program Plan. (Prepared for the U.S. Environmental Protection
Agency.) Research Triangle Park, NC. Publication No.
EPA-600/S7-81-141a. September 1981.
48. Durkin, T. H., et al. Operating Experience with a Concentrated Double
Alkali Process. (Received as literature from FMC Corporation, May
1983.)
49. Boward, W. L., et al., FMC Corporation. FMC Limestone Double Alkali
Process. (Presented at the meeting of the American Institute of
Chemical Engineers in Cleveland, OH.) August 30, 1983.
2-92
-------
50. LaMantia, C. R.s et al, Arthur D. Little. Final Report: Dual Alkali
Test and Evaluation Program. Volume III. Prototype Test Program -
Plant Scholz. (Prepared for the Industrial Environmental Research
Laboratories.) pp. 1-4 - 1-8.
51. Dickerman, J. R., and Johnson, K. L. (Radian Corporation.) Technology
Assessment Report for Industrial Boiler Applications: Flue Gas
Desulfurization. (Prepared for the U.S. Environmental Protection
Agency.) Research Triangle Park, NC. Publication No.
EPA-600/7-79-178i. pp. 2-34 - 2-43, 2-88 - 2-91. November 1979.
52. Letter from Compton, B., Caterpillar Tractor Company, to Martinez, J.
A., Radian Corporation. January 5, 1984. Response to Radian questions
concerning Caterpillar's dual alkali systems.
53. Tuttle, J., et al., PEDCo Environmental. EPA Industrial Boiler FGD
Survey: First Quarter 1979. (Prepared for the U.S. Environmental
Protection Agency, Washington, DC.) Publication No. EPA-600/7-79-067b.
April 1979.
54. Letter from Maddox, J. A., Radian Corporation, to Baldwin, D.
Occidental Chemical Company. May 23, 1984. Confirmation of
information obtained in the October 19, 1983 telephone conversation
between J. A. Martinez (Radian) and Baldwin.
55. Reference 51, pp. 2-84.
56. Letter from Sachtschale, J. R., Santa Fe Energy Company, to Maddox, J.
A., May 31, 1984. Confirmation and corrections to October 18, 1983
telephone conversation between Sachtschale and J. A. Martinez (Radian).
57. Reference 45, pp. 2633-2635.
58. Letter from Maddox, J. A., Radian Corporation, to Friesenhahn, Grissom
Air Force Base. May 23, 1984. Confirmation of information collected
in the July 5, 1983 telephone conversation between Friesenhahn and
R. S. Berry (Radian).
59. Letter from Turner, J. J., St. Regis Paper Company, to Berry, R. S.,
Radian Corporation. November 1, 1983. Dual alkali information.
60. Letter from Maddox, J. A., Radian Corporation, to Weber, J., Army Corps
of Engineers. May 23, 1984. Confirmation of information obtained in
the July 5, October 11, and October 12, 1984 telephone conversations
between Radian employees and J. Weber.
61. Valencia, J. A., Lunt, R. R., and Ramans, G. J., Industrial
Environmental Research Laboratory. Project Summary - Evaluation of the
Limestone Dual Alkali Prototype Systems at Plant Scholz: System Design
and Program Plan. Publication No. EPA-600/S7-81-141a. September 1981.
2-93
-------
62. Behrens, G. P., Radian Corporation. The Current Status of Commercial
Flue Gas Desulfurization Systems. (Prepared for the U.S. Environmental
Protection Agency, ORD.) Washington, DC. EPA Contract No. 68-02-3171.
pp. 205-208. November 1982.
63.
Grant, R. J. and Simpson, 0. L., CIPS. Full-scale DAFGD Experience at
Central Illinois Public Service Company's Newton Station. No date.
64. Van Meter, J. A., and Legatski, J. K., SIGECO, and Durkin, T. H., FMC
Corporation. Operating Experience with the FMC Double Alkali Process.
(Presented at the EPA Symposium on Flue Gas Desulfurization.) October
29, 1980.
65. Reference 62, pp. 12-15.
66. Letter from Reiners, M., ARCO Chemical, to Maddox, J. A., Radian
Corporation. June 6, 1984. Dual Alkali Scrubbers for Industrial
Boiler NSPS.
67. U.S. Environmental Protection Agency. Background Information Document
for Industrial Boilers, Appendices A-E. (Prepared for OAQPS by Radian
Corporation.) March 25, 1982. pp. C-168 - C-186.
68. Memo from Hancock, D. F., Indiana Emissions Sampling Section, to
Profit, F. P., Indiana State Board of Health. January 7, 1982.
Grissom Air Force Base Emissions Data.
69. Reference 62, p. 41.
70. Corbett, W. E., Hargrove, 0. W., Merrill, R. S. (Radian Corporation.)
A Summary of the Effects of Important Chemical Variables Upon the
Performance of Lime/Limestone Wet Scrubbing Systems. (Prepared for
EPRI.) Palo Alto, CA. December 1977. pp. 2-11 - 2-17.
71. Reference 62, pp. 51-54.
72. Wen, C. Y. and Fan, L. S. Absorption of Sulfur Dioxide in Spray Column
and Turbulent Contacting Absorbers. Final Report. Publication No.
EPA-600/2-75-023. Morgantown, WV. West Virginia University. August
1975.
73. Knight, R., Gorden and Steve L. Pernick, "Duquesne Light Company,
Elrama and Phillips Power Stations Lime Scrubbing Facilities", "in
Proceedings, Symposium on Flue Gas Desulfurization, New Orleans, March
1976, Volume 1, Research Triangle Park, NC, 1976, pp. 205 ff.
74. Borgwardt, R. H., Limestone Scrubbing of S02 at EPA/RTP Pilot Plant,
Progress Report No. 16. Research Triangle Park, NC. Undated.
2-94
-------
75. Cronkright, Walter A. and William J. Leddy, "Improving Mass Transfer
Characteristics of Limestone Slurries by Use of Magnesium Sulfate",
Environmental Science & Technology. 10(6), 1976. pp. 659-672.
76. Reference 51, pp. 2-34 - 2-43.
77. Saleem, A. (Chemico Air Pollution Control Corp.) Spray Tower: The
Workhorse of Flue Gas Desulfurization. Power Magazine. 124 (10):
73-77. October 1980.
78. Reference 62, p. 42.
79. Clarke, et al. Evaluation of the Adipic Acid Enhanced Limestone Flue
Gas Desulfurization Process on an Industrial Boiler. EPA Contract
68-02-3173. Project Summary. Research Triangle Park, NC. July 1981.
80. Letter from Haines, R. P., RANGB, to Maddox, J. A., Radian Corporation.
June 1, 1984. Verification of telephone call information - September
19, 1983.
81. Letter from Maddox, J. A., Radian Corporation, to Saleem, A., General
Electric Company. May 23, 1984. Confirmation of information collected
in June 9, 1983 telephone conversation between Saleem and R. S. Berry
(Radian) about wet FGD developments.
82. Letter from Maddox, J. A., Radian Corporation, to Diomigo, A.,
Thyssen-CEA. May 23, 1984. Confirmation of information collected in
June 9, 1983 telephone conversation between Saleem and R. S. Berry
(Radian) about wet FGD developments.
83. Letter from Byrne, R. E., Research Cottrell, to Berry, R. S., Radian
Corporation. August 16, 1983. Wet and dry FGD developments.
84. Reference 62, p. 5-15.
85. U.S. Environmental Protection Agency. The Effect of Flue Gas
Desulfurization Availability on Electric Utilities. Volume II.
Technical Report. Publication No. EPA-600/7-78-031b. March 1978.
86. Reference 51, pp. 2-81 - 2-91.
87. Doctor, R. D. Utility Flue Gas Desulfurization: Innovations and
System Availability. (Prepared for U.S. Department of Energy.)
March 1982.
88. Reference 62, p. 65.
2-95
-------
89. Wange, S. C. and Burbank, D. A. (Bechtel National, Inc.) Adipic
Acid-Enhanced Lime and Limestone Testing at the EPA Alkali Scrubbing
Test Facility. Volume 1. (Prepared for the U.S. Environmental
Protection Agency). Washington, DC. p. 274. July 1981.
90. Clarke, P. A., et al_. (PEDCo.) The Adipic Acid-Enhanced Flue Gas
Desulfurization Process for Industrial Boilers. Volume I. Field Test
Results. (Prepared for U.S. Environmental Protection Agency.)
Research Triangle Park, NC. November 1982. Chapter 3.
91. Kelly, W. E. et al. U.S. Environmental Protection Agency. Air
Pollution Emission Test. Third Interim Report: Continuous Sulfur
Dioxide Monitoring at Steam Generators. Volume I. Summary of Results.
EMB Report No. 77-SPP-23C. p. 1. March 1979.
92. Melia, M. T., et al. PEDCo Environmental, Inc. Utility FGD Survey
July 1982 - March 1983. Volume II. Design and Performance Data for
Operational FGD Systems, Part I. EPRI Contract No. RP982-32/EPA
Contract No. 68-02-3173. pp. 442-454.
93. Reference 62, p. 189.
94.
Hargrove, 0. W., et al. Full Scale Utility FGD System Adipic Acid
Demonstration Program. Publication No. EPA-600/7-83-035. October
1983.
95. Letter from Carlton, J. C., Pfizer Corporation, to Berry, R. S., Radian
Corporation. September 26, 1983. Response to lime scrubbing
questions.
96. Reference 62, pp. 78-79.
97. Letter from Stowe, D. H., Dravo Lime Corporation, to Maddox, J. A.,
Radian Corporation. June 8, 1984. Status of the Thiosorbic process.
98. Henzel, D. S. and D. H. Stowe, Dravo Lime Corporation. A Proven
Reagent for High Sulfur Coal Flue Gas Desulfurization. (Prepared for
the 7th Symposium on Flue Gas Desulfurization. Hollywood, FL.
May 17-20, 1982.)
99. Reference 43, p. B-128.
100. Letter from Maddox, J. A., Radian Corporation, to Anderson, Carborundum
Abrasives Company. June 5, 1984. Verification of information
collected in July 12, 1983 and May 30, 1984 telephone conversations.
2-96
-------
101. Letter from Barker, J. E., Armco Steel, to Maddox, J. A., Radian
Corporation. June 8, 1984. Verification of information collected in
the July 12, 1983 telephone conversation between Barker and R. S. Berry
(Radian).
102. Downs, W., W. J. Sanders and C. E. Miller. Control of S0? Emissions by
Dry Scrubbing. (Presented at the American Power Conference. Chicago,
Illinois. April 21-23, 1980.)
103. Getler, J. L., H. L. Shelton, and D. A. Furlong. Modeling the Spray
Dryer Absorption Process for S02 Removal. Journal of the Air Pollution
Control Association. £9(12): 1270-1274. December 1979.
104. Kelly, M. E. and M. A. Palazzolo. (Radian Corporation.) Status of Dry
S02 Control Systems: Fall 1982. (Prepared for U.S. Environmental
Protection Agency, Research Triangle Park, N.C.) EPA Publication
No. 600/7-83-041. August 1983. pp. 2-28, 12-13, 54-58, 71.
105. Palazzolo, M. A. and M. A. Baviello (Radian Corporation). Status of
Dry S02 Control Systems: Fall, 1983. pp. 7-38, 40-42, 78-80.
106. Reference 105, pp. 88-104.
107. Stevens, N. J., et al. Dry S02 Scrubbing Test Program. Draft report.
(Prepared for U.S. Environmental Protection Agency, Research Triangle
Park, N.C.). EPA Contract No. 68-02-3190. May 1981. p. 48.
108. Blythe, G. M. Field Evaluation of a Utility Dry Scrubbing System.
(Presented at the 1983 EPA/EPRI Symposium on Flue Gas Desulfurization.
New Orleans, Louisiana. November 1-4, 1983.)
109. Reference 42, pp. 4-85 - 4-90.
110. Reference 105, pp. 78-80.
111. Reference 105, pp. 32-38.
112. Reference 105, pp. 40-42.
113. Reference 105, pp. 7-23.
114. Memo from Jones, G. D., Radian Corporation, to Sedman, C., EPA/ISB.
July 30, 1984. Evaluation of Reliability and Availability of
Industrial Spray Drying Systems.
115. Mudgett, J. S., R. S. Sadowski, W. W. West, and M. Mutsakis. Dry Flue
Gas Desulfurization System in an Industrial Plant. In: Proceedings of
the American Power Conference. Volume 44. Chicago, 1982.
pp. 1003-1012.
2-97
-------
116. Crowe, R. B. First-year Operational Experience Celanese Fibers Company
Coal-Fired Boiler Using a Dry Flue Gas Cleaning System. (Presented at
the 74th Annual Meeting of the Air Pollution Control Association.
Philadelphia, Pennsylvania. June 21-26, 1981. Paper #81-35.2.)
117. Farber, Paul S. Start-up and Performance of a High Sulfur Dry Scrubber
System. (Presented at the 75th Annual Meeting of the Air Pollution
Control Association. New Orleans, Louisiana. June 20-25, 1982. Paper
#82-40.5).
118. Burnett, T. A. and K. D. Anderson. Technical Review of Dry FGD Systems
and Economic Evaluation of Spray Dryer FGD Systems. (Prepared for
U.S. Environmental Protection Agency and Tennessee Valley Authority,
Research Triangle Park, N.C.) EPA Publication No. 600/7-81-014, TVA
Publication No. EDT-127, NTIS No. PB 81-206476. February 1981.
119. Apple, C. and M. E. Kelly. Mechanisms of Dry S0? Control Processes.
(Prepared for U.S. Environmental Protection Agency, Research Triangle
Park, N.C.) Publication No. EPA-600/7-82-026. April 1982. pp. 61-63,
80-100.
120. Yen, J. T., R. J. Demski and J. I. Joubert. Control of S02 Emissions
by Dry Sorbent Injection. In: Flue Gas Desulfurization, ACS Symposium
Series 188. Washington, DC. American Chemical Society. 1982. p. 350.
121. Reference 119, pp. 80-87.
122. Muzio, L. J., et al. Demonstration of S0? Removal on a Coal-Fired
Boiler by Injection of Dry Sodium Compounds. In Proceedings:
Symposium on Flue Gas Desulfurization - Volume Z. Palo Alto, Electric
Power Research Institute. March 1983. pp. 628-649.
123. Reference 104, pp. 54-58.
124. Davis, W. T. and T. C. Keener. Chemical Kinetics Studies on Dry
Sorbents - Literature Review. (Prepared for U.S. Department of Energy,
Grand Forks, North Dakota.) Publication No. DOE/FC/10184-2. August
10, 1980. pp. 42-56.
125. Naulty, D. J. Economics of Dry FGD Sorbent Injection. (Presented at
the 76th Annual Meeting of the Air Pollution Control Association.
Atlanta, Georgia. June 19-24, 1983. Paper #83-38.6.) 23 p.
126. Reference 119, pp. 80-100.
2-98
-------
127. Muzio, L. J., et al. Dry S0~ - Particulate Removal for Coal-Fired
Boilers. Volume 1. Demonstration of S02 Removal on a 22-MW Coal-Fired
Utility Boiler by Dry Injection of NahcoTite. (Prepared for Electric
Power Research Institute, Palo Alto, CA.) EPRI CS-2894. March 1983.
128. Reference 119, pp. 61-63.
129. U.S. Environmental Protection Agency. Evaluation of Dry Sorbents and
Fabric Filtration for FGD. Research Triangle Park, N.C. Publication
No. EPA-600/7-79-005. January 1979. p. 60-62.
130. Reference 105, pp. 29-30.
131. Reference 42, pp. 4-160.
132. Solution Mining of Nahcolite (natural sodium bicarbonate) may begin by
1986. Chemical Engineering. 90(12}:2Q. June 13, 1983.
133. Shah, N. D. (Multi Mineral Corporation.) Dry Scrubbing of S0?.
Chemical Engineering Progress. _78_(6):73-77. June 1982.
134. Reference 129, p. 2.
135. Reference 104, pp. 12 and 13.
136. Trexler, E. C. DOE's Electron Beam Irradiation Developmental Program.
In Proceedings: Symposium on Dry Flue Gas Desulfurization. Palo Alto,
Electric Power Research Institute. March 1983. p. 359.
137. Reference 105, pp. 30-31.
138. Reference 104, p. 71.
139. Menegozzi, L. and P. L. Feldman. Removal of NO , S02 by Electron Beam
Irradiation; A Phenomenological Model. (Presented at the 74th Annual
Meeting of the Air Pollution Control Association. Philadelphia,
Pennsylvania. June 21-26, 1981.) 19 p.
140. Reference 42, pp. 4-160.
2-99
-------
3.0 COMBUSTION MODIFICATION CONTROL APPROACHES
Approaches for reducing SO., and NO emissions from coal-fired
^ X
industrial boilers through combustion modification are assessed in this
section. Control methods assessed include fluidized bed combustion
combustion (FBC), limestone injection in multi-stage burners (LIMB) and
combustion of coal/limestone pellets. Information concerning the
reliability and economics of these technologies is generally unavailable.
3.1 FLUIDIZED BED COMBUSTION
Fluidized bed combustion (FBC) is being investigated as an alternative
to conventional combustion techniques for industrial coal-fired boiler
applications (e.g., stoker-fired, pulverized-coal, etc). Fluidized bed
boilers (FBC) offer potential advantages in both boiler design and emissions
control. The fluidized bed promotes higher heat transfer rates which
results in reduced heat transfer surface requirements. The fluidized bed
also operates at a lower temperature which produces lower NO emissions.
A
Addition of limestone to the bed allows sulfur to be captured in-situ which
eliminates the need for an FGD system to control SOp emissions. Also, the
ability of an FBC unit to burn a wide variety of fuels provides fuel
flexibility to users. The primary motivation for development of FBC
technology in Europe and Asia has been fuel flexibility; the technology is
being developed in the U.S. to comply with environmental regulations and for
retrofit applications.
During the past decade, a number of development programs have been
sponsored by both governmental and private organizations to quantify the
advantages of FBC technology and to evaluate its feasibility in commercial
applications. Issues of concern with respect to FBC commercialization
included performance, cost, reliability, and environmental impact. The
earlier development programs did not demonstrate a clear-cut advantage for
FBC compared to conventional boilers. However, more recently,
industrial-sized FBC have become available commercially in the United
States; over 80 FBC installations are operating or scheduled for start-up
prior to 1985. As a part of the recent commercialization activity, several
3-1
-------
new design concepts have been introduced. These new design concepts have
produced configurations which include: 1) the use of recycle for traditional
dense-bed F8C systems, 2) circulating beds, and 3} staged-beds.
The first set of SO^ emissions data for industrial FBC units burning
coal were collected at the Georgetown University (GU) facility. This unit
was operated to meet an emissions limit of 0.78 Ib SCL/106 Btu, cor-
responding to a median of 85 percent removal on a 3 percent sulfur coal.
Results from the GU unit showed a high degree of emissions variability due
to design problems and operating procedures; the performance results are
probably a conservative estimate of what a well-designed, well-operated FBC
dense-bed system can achieve. Emissions data have also been collected on
the Tennessee Valley Authority's (TVA's) 20 MWg dense bed FBC pilot plant
designed for utility applications. This system demonstrated 87 percent S0?
removal with no solids recycle and 98 percent removal with solids recycle.
These results are not directly translatable to an industrial FBC system,
however, due to the large freeboard height associated with the TVA plant;
greater freeboard height facilitates S02 emissions reduction. Emissions
from advanced bed design show mixed results: 82-83 percent SO- removal from
two-stage beds and 90-96 percent removal from circulating bed facilities.
3.1.1. Process Description
Simplified schematic diagrams of several traditional dense-bed FBC
system designs are presented in Figure 3.1-1. While the figure illustrates
configurations generating electrical power, these same systems can produce
steam for industrial applications. An atmospheric fluidized bed combustion
(AFBC) boiler equipped with a separate carbon-burnup cell is presented in
Figure 3.1-la. This design concept has been abandoned in favor of the
recycle configuration presented in Figure 3.1-lb where elutriated particles
from the bed are collected by a cyclone and recirculated back to the bed.
In addition to the cyclones, downstream fabric filters or ESPs are necessary
to further reduce flue gas particulate emissions. A pressurized fluidized
bed combustion (PFBC) system operating in the combined-cycle mode is
presented in Figure 3.1-lc. Since it appears that AFBC boilers will
dominate the industrial FBC market in the near future only AFBC designs are
considered in the following sections.
3-2
-------
OJ
CO
Primary
Combustor
Additive
Coal
Primary
Cyclone
Final Dust Stack
Collector
Secondary
Cyclone
Air
Ash
T
Air
Steam
Turbine
Condenser
Water
Boiler
Feed Water
Disposal
Carbon Burnup CelI
(a) Water-Cooler! AFBC Combustor with Separate Carbon-Burnup Cell
Figure 3.1-1. Schematics of Traditional Dense-Bed FBC Power - Generation Systems2
-------
Primary
Combustor
Primary
Cyclone
Air Ash
Final Dust
Collector
Stack
Secondary
Cyclone
Disposal
(b) Water-Cooled AFBC Combustor with Recycle from Primary Cyclone
Figure 3.1-1 (con't.)
-------
Air
oo
en
Compressor
Participate
Removal
Pressurized
Combustor
Additive-
Coal-
Gas Turbine
Stack
Steam Turbine
Water
ondenser
Boiler
Feed Water
Ash Disposal
Heat
Recovery
(c) PFBC Water-Cooled Combustor/Combined-Cycle Plant
Figure 3.1.1 (con't.)
-------
Two newer AFBC design configurations are presented in Figure 3.1-2.
The two-stage system shown in Figure 3.1-2a is a traditional dense-bed FBC
system where coal is fired with a substoichiometric amount of air in the
lower stage and additional air is added in the upper stage. This approach
decreases the amount of N0x formed in the first stage but allows acceptable
combustion efficiency to be achieved in the second stage.
A circulating fluidized bed (CFBC) is illustrated in Figure 3.1-2b.
The CFBC utilizes smaller limestone particles and high combustion-air
velocities to carry all of the solid particles out of the combustion reactor
in a dilute phase. The particles are then collected and returned to the
combustor. The required heat transfer can be accomplished either in the
dilute gas phase section as pictured in Figure 3.1-2b or externally by heat
exchange with the collected hot particles prior to reinjection into the
combustor. Potential CFBC advantages include lower NO emissions, improved
/\
limestone utilization, increased combustion efficiency, a simpler
coal/limestone feed system, and improved load-following capability.2
3.1.2. Factors Affecting Performance
The following major factors that affect sulfur capture in the AFBC
boiler were identified and discussed in the March, 1982 Industrial Boiler
New Source Performance Standard Background Information Document (BID).3
These factors include:
- calcium-to-sulfur molar feed ratio (Ca/S);
- limestone sorbent particle size;
- gas phase residence time (related to bed depth and superficial gas
velocity);
- solid phase residence time (related to bed depth, feed mechanism, and
solids recycle rate); and
- bed temperature.
These factors can be varied to obtain the optimum sulfur capture.
However, it should be emphasized that these factors also affect other
important performance variables including boiler operation (e.g., combustion
efficiency, boiler efficiency, etc.) and control of other flue gas emissions
3-6
-------
Stack
Final Dust
Collector
Additives
»
Secondary Air
Additive
Coal
r
Primary Ash Disposal
Air
(a) Water-Cooled AFBC Unit with Two-Staged Combustor
Combustor
Additive
CoTT
Ash Disposal
(b) Water-Cooled AFBC Unit with Circulating Bed
Stack
Figure 3.1-2. Schematics of Two-Stage and Circulating-Bed AFBC
Power-Generation Systems^
3-7
-------
(e.g., N0x and participates) and solid waste characteristics. Therefore, a
number of important design compromises must be made between boiler
performance and environmental impact.
Recent designs have been more sophisticated in response to needs for
optimizing the tradeoffs resulting from coupling combustion and in-situ
emissions control. The effects of the newer configurations on SCL emissions
control and associated tradeoffs with other performance variables are
itemized below:
- Recycle of elutriated fines from traditional dense beds improves
combustion efficiency and limestone utilization and reduces SCL
and NO emissions.
A
- Staged beds overcome the design tradeoffs associated with a one-bed
unit by allowing combustion and emissions control to be optimized
more independently.
- Circulating beds allow the units to be operated at different
conditions than traditional dense-beds (e.g., limestone size,
superficial velocity, residence time, mixing) and allow
performance to be optimized under more favorable conditions
(e.g., improved limestone utilization, SCL control, NO control,
C- /\
and combustion efficiency).
Another important point that should be discussed based on recent test
data is the effect of coal characteristics on S0? emissions. In addition to
the sulfur content, the form of the sulfur and the alkalinity and quantity
^of the ash in the fuel will affect SOp emissions. Tests conducted by DOE's
Grand Forks Energy Technology Center (GFETC) and METC on low-rank fuels
indicate that some lignites and low-sulfur subbituminous western coals
contain a significant quantity of calcium and sodium alkalinity in their
45
ash. ' The relatively large quantity of alkaline ash and low sulfur
content combine to provide significant sulfur capture. In tests conducted
with a Beulah, North Dakota lignite, the inherent alkali-to-sulfur ratio
ranged between 0.5 and 1.2 for low-sodium and high sodium lignite ashes
respectively. To achieve 90 percent sulfur capture, the low-sodium Beulah
lignite required only enough limestone to be added to produce an external
3-8
-------
alkali-to-sulfur ratio of 0.75. Furthermore, the high-sodium lignite
contained sufficient inherent alkalinity in the ash to achieve 90 percent
sulfur capture without the addition of any limestone. Recent tests with
Texas lignites indicate that ash and sulfur characteristics other than
alkali-to-sulfur ratio also affect sulfur capture efficiency (i.e., silica-
to-sodium ratio).
Although sodium in the fuel contributes to improved sulfur capture, it
also increases the agglomerating tendency of the fuel. High sodium levels
in lignite lower the melting point of the ash and cause the particles in the
bed to stick together. Agglomeration can cause a number of operating
problems including loss of fluidization, loss of bed temperature uniformity,
plugging of recycle lines, reduced combustion efficiency, and decreased heat
transfer rate.
3.1.3. Applicability to Industrial Boilers. Only a handful of
vendors offered industrial AFBC boilers in the U.S. in 1979 on a commercial
basis. Today, approximately 40 manufacturers offer AFBC boilers capable of
producing from 10,000 to 600,000 Ib/hr of steam at conditions comparable to
conventional boilers. Many are offering guaranteed systems for a wide
variety of applications.
The price and availability of premium fuels as well as long-term
environmental concerns have made AFBC a viable option compared to
stoker-fired and pulverized-coal fired units. S09 and NO emissions control
^ A
achieved within the combustion chamber can eliminate the need for scrubbers,
lower sulfur coal purchases, or elaborate combustion modifications. The
fuel flexibility provided by FBC technology allows a wide range of solid
fuels with varying ash and moisture contents to be successfully burned
within a single boiler. In the U.S., AFBC boilers are generally cost-
competitive with conventional industrial boilers equipped with scrubbers.^
The number of AFBC boilers operating throughout the world has increased
dramatically in recent years. In China alone, over 2,000 AFBC boilers
combust low grade fuels containing up to 70 percent ash. The AFBC units are
3-9
-------
used because of their ability to combust the low grade fuels. In general,
limestone is not added for SO,, removal in China. These boilers are
generally small and are frequently located in remote areas.6
Outside of China, over 130 industrial-sized AFBC boilers are operating
or planned for operation in the near future. These boilers are designed to
represent a wide range of requirements such as size, fuel type, and steam
conditions for a variety of industrial applications. Their sizes range from
10,000 to 600,000 Ib/hr of steam. Over 22 different types of fuels are
planned for use including low rank fossil fuels (lignite and peat) and waste
from process industry, agricultural, and municipal sources. Steam pressures
in excess of 2,500 psi are generated. These installations also represent
all of the major types of design configurations including the more recently
introduced staged and circulating bed designs.1
Of the over 130 units outside China, 80 units are located in the United
States. Excluding AFBC boilers that are test, demonstration, or uncompleted
units, only eight AFBC boilers in the United States burn coal. Information
describing five of the coal-fired AFBC units is summarized in Table 3.1-1.
This type of information is not currently available for the remaining three
units. Comparisons of the five units presented in Table 3.1-1 indicate the
variability in design and operating conditions for these initial'commercial
installations.
Despite the availability of commercial units and the increasing number
of installations, some potential users of AFBC boilers remain skeptical of
the overall technical and economic advantage of this relatively new approach
for steam and power generation. To reduce the reluctance of potential
users, the technology must continue to be improved and optimized to address
continuing issues associated with unit subsystems. Then, the technology
must be adequately demonstrated in various industrial applications to prove
its flexibility in meeting specific process requirements.
3-10
-------
TABLE 3.1-1. PARIIAL SUMMARY OF COAL-FIRED INDUSTRIAL AFBC BOILERS IN THE U. S.
Construction
Bed Configuration
Plant A
Field
Circulating
Plant B
Field
Circulating
Plant C
Field
Conventional
Bubbling Bed
Plant D9
Meld
Conventional
Bubbling Bed
Plant E
Package
Conventional
Bubbling Bed
Features
Solids Recycle
Staged Combustion Air
Limestone for SO, Removal
Recycle Ratio
Primary/Stoichiometric Air Ratio
Ca/S Ratio
Percent SOj Removal
Fuel
Type
Heating Value (HHV), 3tu/lb
Sulfur Content, ',
Alternate Fuels
Boiler Efficiency, *
Yes"
Yes
Yes
NAh
0.6
3.5
90
Coal
7,937
0.5b
Petroleum Coke
72
Yes
Yes
Yes
Not Determined
Confidential
3 or 4
Not Determined
Coal
10,000
0.6
Cokec
Not Determined
Yes
No
Yes
Not Determined
NA
2
Not Determined
Coal
No
No
Nod
NA
NA
NA
NA
Coal
Not Available
1.0
None
Not Determined Not Available
Not Available
0.8/1.5f
e
Yes1
No
No
NA
NA
NA
NA
Coal
12,085
3
None
83.5
Availability, S
CEMk Equipment
85J
Not Determined
Not Determined Not Available
Not Available
30.
N0x
CO
CO,
Particulates
Recurring Problems
Status
Yes
Yes
Yes
Yes
Yes
None
Operational ,
Dec, 1981
Compliance testing
Completed July 1983
Yes
Yes
No
No
Yes
NA
Operational
July, 1983
Yes
Yes
Yes
No
Yes
NA
Operational
August, 1983
No
No
No
No
No
NA
Ooerational
August, 1981
Currently ooerating
with cost-cutting
measures.
No
No
No
No
No
Water Tube 4
Wall Erosion
Operational
AprM, 1980
"roblems with
erosion of water
tubes and walls.
Additional solids recycle, beyond that provided by the circulating bed, is available but not being used.
Average total fuel stream contains approximately 2 percent sulfur. Petroleum coke contains approximately 7 percent sulfur and
has a higher heating value of 14,943 8tu/1b.
After the unit comes on-line, oil-impregnated diatamaeous earth will be tested for use as a fuel.
Limestone used only for bed material due to liberal emission requirements and as a cost-cutting measure.
The decision to use or not to use alternate fuels has not been made.
Two different coals with differeit sulfur contents will be used depending an ecanomics.
^Information gathered from manufacturer at suggestion of operator.
hNot applicable.
Solids recycle incorporated originally, but presently inoperable due to mechanical problems.
JDoes not include down time resulting from electrical power outrages.
Continuous emission monitoring.
3-11
-------
3.1.4. Development Status. Research and development (R&D) began in
England in the 1960's to develop FBC technology as an improved method for
burning coal. In the United States, a significant R&D effort was conducted
during the 1970's. Much of the work was sponsored by the U.S. Department of
Energy (DOE). Recently, the DOE's overall mission has shifted from large,
demonstration projects to bench-scale, high-risk, advanced concept
research. Since DOE regards conventional AFBC technology as commer-
cialized, further commercial development of AFBC technology has become the
responsibility of the private sector. DOE has halted its participation in
large-scale demonstration programs at Georgetown University, Great Lakes
Naval Station, Shamokin Area Industrial Corporation, and United Shoe
Manufacturing Corporation. As an example of DOE's shift of emphasis toward
more advanced technology, the goal of DOE Morgantown Energy Technology
Center's (METC) advanced AFBC projects is to achieve 90 percent sulfur
capture on high sulfur coal with a calcium-to-sulfur (Ca/S) molar ratio of
o
1.5 or less.
The Electric Power Research Institute (EPRI) is sponsoring programs
aimed at developing FBC technology for utility applications. EPRI is
sponsoring testing at the Babcock and Wilcox (B&W) 6' x 6' unit at Alliance,
Ohio. EPRI is supporting a test program initiated in 1982 at a 20 MW AFBC
pilot plant operated by the Tennessee Valley Authority (TVA) at their
Shawnee Generating Station. While these programs are directed toward
utility applications, many of the technical issues addressed are directly
applicable to industrial boiler facilities.
Currently, 80 industrial FBC installations are operating or scheduled
for start-up in the United States prior to 1985. Only nine of these
installations are designed for coal combustion. Most of the other units
will use wood, oil, natural gas, process gas, or process wastes as a fuel
source.
3-12
-------
3.1.5. Emission Test Data
In the past, SO,, emissions data have been collected primarily from
small scale test equipment operating over a wide range of conditions. More
recently, S02 emissions data have been collected at large commercial scale
operations. A summary of available S02 removal data for the various AFBC
configurations is presented in Table 3.1-2.
One of the first sources of continuous emissions data from a commercial
scale facility was the dense-bed AFBC unit located at Georgetown University
(GU). At GU, the unit was operated to meet the District of Columbia
emission limit of 0.78 Ib S02/106 Btu. The median S02 removal efficiency
has been about 85 percent with 3 percent sulfur coals at Ca/S ratios between
4 to 6. Actual S02 emissions varied over a broad range due principally to
the coal sulfur variability. During periods when the coal was sampled on an
hourly basis, S02 removal efficiency ranged from 80 to 90 percent at Ca/S
ratios of 4 to 7. However, significant design and operating problems have
been encountered at GU which have resulted in higher Ca/S ratios than
originally anticipated. The Ca/S ratios observed at GU are probably higher
than would be required if the system design and operation were optimized.
The effect of solids recycle on dense bed performance is dramatically
illustrated at TVA's 20 MWe pilot unit where removal at a Ca/S of 3.0
increased from 87 percent with no recycle to 98 percent with a recycle ratio
of 1.5. EPRI's target for sulfur capture is to achieve 90 percent removal
at a Ca/S of 2.0. This target is predicted based on an evaluation of test
performance results.
The TVA 20 MWg pilot plant design provides greater sulfur capture
efficiency than the older unit at Georgetown University. It should be
noted, however, that the outstanding S0? removal performance of the
TVA. 20 MWg pilot plant operating with solids recycle may be aided by the
higher freeboard of this unit. Freeboard height at the TVA unit is over
20 feet compared to about 10 feet for a typical industrial fluidized bed
boiler. The higher freeboard allows more time for S02 capture by entrained
sorbent, effectively increasing the in-bed gas residence time.
3-13
-------
TARLE 3.1-2. SUMMARY OF S02 EMISSIONS DATA FOR VARIOUS AFBC CONFIGURATIONS
Capacity
AFBC Configuration lb steam/hr
Conventional Bubbling Bed
Georgetown University 100,000
TVA 20 MW(e)
- no recycle . 150,000
- Recycle ratio = 1.5
Staged Bed
Wormser - United Shoe 2,500L
Manufacturing Corp
Wormser -- Iowa Beef 70,000
Processors
Circulating Bed
Lurgi ND
Battelle MS-FBC 55,000
ND = flaf-A nnt a u ;i i 1 JfTTnT ~
Type of
S0? Emissions
Coal Type Combustion Ca/S Monitoring/Length of SO
(Percent Sulfur) Temperature Ratio Monitoring Period Removal
Eastern bituminous
(1.5 - 2.0%S)
Eastern bituminous
(3.7%S)
Eastern bituminous
(1.5*5)
Midwestern bituminous
(4.2%S)
Eastern bituminous
(37.S)
Various (2%S)f
1550°F 3-6 CEM/23 days 75 - 95%
1550°F 3.0 CEM/ND 87%
3-0 98%
1800°F
-------
The capability of staged combustion is also illustrated in Table 3.1-2.
Wormer's staged combustion concept achieved approximately 80 percent SO^
removal at a Ca/S molar ratio of 3.0 at two different installations, one
11 12
firing high sulfur coal and the other firing low sulfur coal. '
The performance of two circulating bed design concepts is also
summarized in Table 3.1-2. The Lurgi circulating bed data demonstrates a
significant improvement in limestone utilization and removal efficiencies
over the other design configurations. Lurgi's staged circulating bed
achieved 90 percent sulfur capture at a Ca/S ratio of 1.5 while operated at
full capacity with excess air levels of 15 to 20 percent. Battelle's
Multi-Solids FBC unit obtained 95 percent removal at a Ca/S ratio of 4.5.
The data presented in Table 3.1-2 serve to compare the trends provided
by the newer configurations with respect to S0? emissions. More data are
necessary to provide a direct comparison of optimum S0? control for
configurations at comparable design and operating conditions.
3-15
-------
3.2 LIMB
LIMB is a developing technology which is capable of achieving
simultaneous reductions in sulfur oxides (SO ) and nitrogen oxides (NO )
A A
emissions from pulverized coal boilers. The term "LIMB" is short for
Limestone Injection Multistaged Burners. This technology is based on the
use of low NO combustion techniques in combination with dry limestone
A
injection into the furnace for simultaneous SO control.
A
The goal of ongoing LIMB R&D is to develop a technology which can
substantially reduce SO and NO emissions for a capital investment of
X A
$30-40/kW -- about 1/5 of the cost of conventional flue gas desulfurization
(FGD) systems. For new coal-fired boiler applications the goal is to
achieve S02 removals of 70 percent with simultaneous NO emission levels of
86-130 ng/J (0.2-0.3 Ib/million Btu).
Since LIMB techniques are only now beginning to be evaluated in
commercial-scale combustion equipment, it will be several years before any
meaningful data on the long term costs, benefits and/or problems associated
with this technology are known. Initial test results in small scale
equipment have been promising, however.
3.2.1 Process Description
In the LIMB process, dry, finely ground limestone is injected into the
furnace either through the burners, or through separate injection ports
installed in the furnace wall.
If limestone is used as the alkaline reagent in a LIMB system, the
following are among the key reactions which will occur:
Calcination -
CaC03 + CaO + C02 (3.2.2-1)
Reaction with reduced sulfur species in fuel-rich zones -
CaO + H2S •»• CaS + H20 (3.2.2-2)
3-16
-------
Sulfation -
CaO + S02 + 102 -> CaS04 (3.2.2-3)
Product solids, along with any unreacted limestone are entrained in the flue
gas and collected along with fly ash in a downstream particulate control
device such as an electrostatic precipitator or fabric filter.
In order for LIMB to be effective in controlling SCL emissions, the
alkaline reagent must be injected under conditions which are favorable for
sulfur capture via Equations 3.2.2-2 and 3.2.2-3. This requires the
integration of limestone injection control for SCL with the use of low NO
C— A
combustion techniques. The two techniques presently under development for
use in LIMB technology are the distributed mixing burner (DMB) for
wall-fired pulverized coal boilers and the fuel rich fireball for
tangentially-fired boilers.
The distributed mixing burner is shown conceptually in Figure 3.2-1.
Coal and primary combustion air are injected through a central fuel
injector. The coal begins to burn in a very fuel-rich zone. Secondary air
admitted through two concentric throats gradually mixes -with the primary
reactants. The final mixture in the burner zone is still fuel-rich, having
about 70 percent of the air required for complete combustion. These fuel-
rich conditions minimize the formation of fuel NO by promoting maximum
A
conversion of the chemically-bound nitrogen in the coal to molecular
nitrogen. The balance of the air necessary for complete combustion is
admitted through tertiary ports spaced around the burner periphery. This
delayed combustion approach also reduces peak flame temperatures which
minimizes NO formed by thermal fixation of the nitrogen present in '
A
combustion air.
Tangentially-fired boilers require a different approach to achieve the
same results. In this case, the coal and primary combustion air are
introduced in a jet which penetrates most of the width of the furnace. The
jet is directed along the tangent of an imaginary circle in the center of
the furnace. Secondary air is introduced in the same vertical plane at
3-17
-------
00
I
CO
Tertiary Air
Outer
Secondary Air
Inner
Secondary Air
Burner
Center!1ne
Very Fuel Rich
Zone (Average
Stoichiometry 40%)
Progressive Air Addition Zone
(Overall Stoichiometry 70%)
Final Air Addition Zone for Burnout
(Overall Stoichiometry 120X)
Figure 3.2-1.
Multistage combustion in a distributed mixing burner (top
half of burner only depicted).3
Figure redrawn from figure presented in reference 15.
-------
elevations both above and below the fuel jet. The balance of the combustion
air is introduced in the same horizontal plane as the fuel jets but directed
at an angle closer to the furnace wall. By mounting one such assembly at
each corner of the furnace, a fuel-rich fireball is formed in the center of
the furnace. This design generates the same type of delayed mixing as the
DMB and likewise reduced NO formation. In most boilers, multiple burner
A
elevations are used to provide the necessary energy input. A plan view of
the fuel rich fireball approach is shown in Figure 3.2-2.
3.2.2 Factors Affecting Performance
The variables which appear to have the greatest effect on the SCL
capture rate are temperature, residence time and limestone stoichiometry
(Ca/S ratio). Temperature effects are important because of their impacts
upon both the thermodynamics and kinetics of the calcination and sulfur
capture reactions. Temperatures substantially below about 800°C (1500°F)
will cause the reactions shown in Equations 3.2.2-2 and 3.2.2-3 to proceed
at rates which are too slow to be of commercial significance. Very high
temperatures on the other hand (well above 1000°C) can deactivate the
sorbent and lower the driving forces for sulfur capture. Because of these
effects, the LIMB process achieves its best results when the sorbent is
injected and the coal firing is controlled so that the residence time of the
particles at the optimum temperatures for reaction is maximized. Some of
the same conditions which favor efficient sulfur capture are also favorable
from the standpoint of minimizings NO formation.
3.2.3 Applicability to Industrial Boilers
The current emphasis of LIMB technology is on utility application.
The major factors influencing the compatibility of LIMB with new boilers
appear to be the coal properties and the design of the boiler furnace,
convection section, and ash removal system. Depending on these factors,
potential problems arising from LIMB applications include increased
3-19
-------
Coal and
primary air
Secondary A1r Injected
Above and Below Flame
Figure 3.2-2. Fuel rich fireball burner design for tangentially
fired boiler.
Figure redrawn from figure presented in reference 15.
3-20
-------
furnace slagging, plugging of tight convection section passes, overloading
or plugging of ash removal systems, and incomplete coal combustion. These
problems must be dealt with through alterations in boiler operating
procedures or system design modifications. In addition, LIMB technology
will also increase boiler thermal losses by 1 to 2 percent, and will require
higher efficiency downstream particulate controls due to the increase in
uncontrolled particulate matter emissions. It is expected that similar
problems will have to be dealt with in applying LIMB to industrial boilers.
3.2.4 Development Status
A primary driving force behind LIMB technology development at present
is the need for low cost, NO and S09 control systems for retrofit
A C. ---..-
applications to pulverized coal boilers.
The largest scale test effort to date has been carried out by Dr. Klaus
Hein of Rhienisch-Westfailsches Electrizitatswerk (RWE) in the Federal
Republic of Germany. His work involved the firing of brown coal in
tangentially-fired pulverized coal boilers. Brown coal is a low rank coal
similar to a low quality lignite. These boilers operate at relatively low
combustion zone temperatures due to the high moisture content of the coal
(up to 60 percent) and the flue gas recirculation used for drying. These
conditions are thought to be favorable for sulfur capture by the sorbent as
well as the generation of relatively low NO emissions. This system was
tested on a 60 MW boiler where SOp reductions of over 60 percent were
achieved on a low sulfur coal. Tests on a 300 MW boiler are planned in
1982-83.
Other tests include those conducted by Steinmuller, a major German
boiler manufacturer. Steinmuller ran bench-scale experiments using natural
gas doped with sulfur compounds. They also conducted 2 MW, pilot-scale
tests with sulfur-doped natural gas and pulverized coal. The pilot-scale
burner is a staged burner, the design of which is based on earlier EPA work
on the distributed mixing burner. Using a proprietary calcium-based
sorbent, Steinmuller has achieved up to 70 percent SO control at a
A
calcium-to-sulfur stoichiometry of two-to-one.
3-21
-------
The development of LIMB will continue to be affected by ongoing low NO
A
combustion technique development efforts. First generation low-NO burners
X
developed by various boiler manufacturers are already being installed on
utility-scale coal-fired boilers. Further, EPA's low NO -program has
A
produced very encouraging pilot-scale test results with more advanced burner
designs such as the distributed mixing burner for wall-fired units and the
fuel rich fireball for tangentially-fired units. Evaluations of a wall
fired burner on an industrial boiler and of a tangential burner on a utility
boiler are currently in progress. Results now show that NO emission levels
x f
from these advanced burners can be maintained at levels of 0.3 lb/10 Btu.
3.2.5 Enri$s_jon_s_pata_
Recent LIMB performance data were discussed in previous sections. No
long term commercial scale performance data are available at present. LIMB
testing on a 700 MW utility boiler in West Germany is planned for late 1983
1C
and laboratory- and pilot scale testing by EPA is continuing.
3-22
-------
3.3 COAL/LIMESTONE PELLETS
Coal/limestone pellet technology is an SCL removal technique currently
being developed by the EPA. In this process, coal/limestone pellets are
fired as ordinary fuel in stoker boilers; the SCL formed during combustion
reacts with the limestone present in the fuel pellets to form calcium
sulfate and calcium sulfite salts.
No significant developments have occurred for this technology since the
preparation of the March 1982 Industrial Boilers Background Information
Document (BID). A 14-day continuous test burn of the pellets had been
scheduled for a 60,000 Ib stream/hr chaingrate stoker boiler. However, the
14-day test has been cancelled or delayed indefinitely due to problems with
the pellets as manufactured by Banner Industries. Adequate drying of the
pellets on a large-scale production basis was not possible with Banner's
existing process equipment.
Work on the development of coal/limestone pellet technology has only
recently been resumed and no new results are yet available. Current
research efforts are being directed toward the development of a
coal/limestone briquette production process that will produce pellets with
mechanical strength and durability characteristics superior to those
produced using auger extrusion.
3-23
-------
3.4 REFERENCES
1. Makansi, J. and B. Schwieger. "FTudized-Bed Boilers." Power
.126(8): 126. August 1982.
2. Hubble, B.R. Fluidized-Bed Combustion: A Review of Environmental
Aspects. Argonne National Laboratory Report No. ANL/ECT-12.
January 1982.
3. U.S. Environmental Protection Agency. Fossil Fuel Fired Industrial
Boilers — Background Information Document. EPA Report No. EPA-450/
3-82-006. March 1982.
4. Golirsch, G.M., S.A. Benson, D.R. Hajicek, and J.L. Cooper. Sulfur
Control and Bed Material Agglomeration Experience in Low-Rank Coal
AFBC Testing. Volume II. Proceedings of the Seventh International
Conference on Fluidized-Bed Combustion. Grand Forks Energy Technology
Center/Combustion Power Company, Inc. October 1982. pp. 1107-1120.
5. Golirsch, G., et al. Atmospheric Fluidized Bed Combustion Testing
of North Dakota Lignite. Volume III. Proceedings of the Sixth
International Conference on Fluidized Bed Combustion. April 1980.
pp. 850-862.
6. Schwieger, B. "Fluidized-Bed Boilers Keep Chinese Industry Running on
Marginal Fuels." Power. _127_(3):59-61. March 1983.
7. Mares, J.W. Keynote Address. Proceedings of the Seventh International
Conference on Fluidized-Bed Combustion. Volume I. U.S. Department of
Energy. October 1982. pp. 1-4.
8. Aul, E.F. Notes from meeting between Radian Corporation and U.S.
Department of Energy/Morgantown Energy Technology Center, Morgantown,
West Virginia. June 14, 1983.
9. Young, C.W., et al. Continuous Emission Monitoring at the Georgetown
University Fluidized Bed Boiler. Publication No. EPA-600/S7-81-078.
March 1983.
10. Castleman, J.M., et al. Campaign I Report: Technical Summary of
TVA/EPRI 20-MW AFBC Pilot Plant Test Program. Volume I. TVA/Energy
Demonstrations and Technology Division. May 1983.
11. Fraser, R.G. Operation and Testing of the Wormser Grate Fluidized
Bed Combustor at the USM Corporation at Beverly, Massachusetts.
Publication No. DOE/ET/15460-193. May 1981.
3-24
-------
12. Sadowski, R.S., et al. ( Wormser Engineering, Inc.) Operating
Experience with a Coal-Fired Two Stage FBC in an Industrial Plant.
(Presented at the 76th Annual Meeting of the Air Pollution Control
Association. June 1983.)
13. Lund, T. (Lurgi Corporation.) Lurgi Circulating Fluid Bed Boiler:
Its Design and Operation. Volume I. Proceedings of the Seventh
International Conference on Fluidized Bed Combustion. October, 1982.
pp. 38-46.
14. Jones, 0. and E.C. Seber. (Conoco, Inc.) Initial Operating Experience
at Conoco1s South Texas Multi-Solids FBC Steam Generator. Volume I.
Proceedings of the Seventh International Conference on Fluidized Bed
Combustion. October 1982. pp. 381-389.
15. Brna, T. G. and G. B. Martin. New Development: Dry Flue Gas
Desulfurization and Combined SO and NO Removal. (Presented at the
Third E.C.E. Conference on Desuifurizatton of Fuels and Combustion
Gases. Salzburg, Austria. May 18-22, 1981.) 36 pp.
16. Kelly, M. E. and S. A. Shareef (Radian Corporation). Third Survey of
Dry SO- Control Systems. (Prepared for U.S. Environmental Protection
Agency.) Research Triangle Park, North Carolina, Publication
No. EPA600/7-81-097. June 1981.
17. Letter from J. A. Maddox (Radian), to Jack Wasser (IERL), May 23, 1984.
Status of Coal/Limestone Pellet Technology.
3-25
-------
4.0 PRECOMBUSTION CONTROL APPROACHES
Precombustion control techniques for reducing PM, NO , and S09
A £-
emissions from industrial boilers include physical or chemical coal cleaning
and the production of clean synthetic gaseous or liquid fuels from coal.
Recent developments impacting the applicability of these technologies to
industrial boilers are summarized in this section.
The use of coal-liquid mixture (CLM) as industrial boiler fuels is also
discussed in this section. While CLM use is not a control technique per se,
this treatment reflects the recent interest shown in CLM's for liquid
fuel-fired boiler retrofit applications.
4.1 PHYSICAL COAL CLEANING
Physical coal cleaning (PCC) or coal washing is a cost competitive
method of reducing the sulfur and ash contents of coals containing
significant quantities of pyrite sulfur and/or ash. Several recent economic
studies have indicated that it may be possible to reduce coal sulfur levels
with PCC at no net cost to the fuel purchaser. This finding result from the
credits which can be taken for reduced coal transportation costs, reduced
ash and scrubber sludge disposal costs, reduced FGD system reagent
requirements, reduced boiler maintenance costs, and increased boiler
efficiency and operability which result from the use of a higher grade coal.
These costs can more than offset the costs of the coal cleaning plant.
4.1.1 Process Description
In a modern PCC plant, coal is typically subjected to size reduction
and screening, separation of coal-rich and impurity-rich fractions,
dewatering, and drying. Commercial PCC methods achieve a separation of the
coal from its impurities by relying on differences in the specific gravity
(gravity separation) or the surface properties of the coal and its mineral
matter (froth flotation).
The overall process design philosophy in most modern PCC plants is to
treat precise fractions of the crushed coal feed with specific unit
4-1
-------
operations which best meet the overall cleaning plant objectives. A
characteristic of this design philosophy is that multiple product streams
evolve, each with its own set of physical and chemical properties. These
separate product streams may be blended prior to shipment to produce a
composite coal precisely meeting the consumer's specifications. Within the
context of supplying small industrial boilers, many opportunities exists for
premium (low-ash, low-sulfur) size fractions to be segregated from the final
blending operation and targeted for specific end users.
4.1.2 Factors Affecting Performance
The primary factor which determines the amount of sulfur reduction
which is achievable by physical cleaning is the distribution of the sulfur
forms in the coal. There are three general forms of sulfur in coal;
pyritic, sulfate, and organic. Pyritic sulfur generally exists as
individual particles (0.1 micron to 25 cm in diameter) distributed uniformly
through the coal matrix. Pyrite is a dense mineral (4.5 g/cc) compared with
bituminous coal (1.3 g/cc) and is not water-soluble; the best means of
removing pyrite sulfur from coal is by specific gravity separation (dense
media washing).
Sulfate sulfur is usually present in very small amounts (O.T percent by
weight or less) in coal. This form of sulfur, is usually water soluble and
can be removed by washing the coal.
Organic sulfur is usually chemical bonded to the organic carbon of the
coal and cannot be removed unless the chemical bonds are broken. The amount
of organic sulfur present thus defines the lowest limit to which a coal can
be cleaned with respect to sulfur removal by physical methods.
Other factors affecting the performance of PCC technology include: the
size to which the coal is crushed, the unit processor employed, the
densities of the separating media and the percent recovery of cleaned coal
on a mass or energy input basis. Higher removal percentages can be achieved
only at the cost of lower mass or energy recovery rates (higher percent
rejected material).
4-2
-------
4.1.3 Applicability to Industrial Boilers
The firing of physically cleaned coal in industrial pulverized coal-
fired boilers offers several advantages over the use of raw coal. Because
physical cleaning partially removes pyrite, ash, and other impurities, both
S02 and particulate emissions are reduced. Physical cleaning also results
in the production of fuel with much more uniform properties than the raw
coal (see Figure 4.1-1). This results in greatly improved combustor
performance characteristics. As compared to raw coal, physically cleaned
coal is easier to handle and feed, and burns more efficiently and uniformly
with less chance for clinkering. This reduces boiler maintenance and ash
disposal problems. Physical cleaning of coal should also improve the
overall performance of stoker-fired boilers provided the resultant coal size
is acceptable for stoker firing.
4.1.4 Development Status
As shown in Table 4.1-1, over 224 million tons of bituminous coal and
lignite were cleaned by mechanical means in 1978, the last year for which
cleaning plant statistics were developed. This represents about one third
of the total US. production of bituminous coal and lignite for that year.
The majority of the cleaning plants currently in operation are designed for
ash removal rather than sulfur removal, although many take out 20-30 percent
of the sulfur in the raw coal. The capabilities of individual plants vary
widely from less than 200 to more than 25,000 metric tons per day.6
Most of the PCC plants which are currently in service operate with
fairly low capacity factors. This characteristic is due to a combination of
two effects:
the fluctuating (e.g. seasonal) nature of coal demands and
the maintenance requirements associated with any solids handling
operation.
4-3
-------
12'
10—
t—
CO
ID
O
CM
O
1/1
GO
CO
4 —
00
CO
A ^^^
DAY BASIS
ROM COAL
\^^
CLEAN COM.
0 20
FIGURE 4.1-1
I ' I ^ I " I
40 60 80 100
PRODUCTION TIME (HRS)
120
140
Kitt Mine Coal Preparation Plant - Hourly Incremental Data
for SrKur Dioxide Emission Parameter
(Source: Reference 4 }.
-------
TABLE 4.1-1.
PREPARATION AND THERMAL DRYING OF BITUMINOUS COAL AND LIGNITE BY
STATE - 1978 (Thousand Short Tons)
-
Alabama
Alaska
Arlsona
Arkansas
Colorado
Georgia
Illinois
Indiana
Iowa
Kansas
Kentucky:
Eastern
Western
i Total
en
Maryland
Missouri
Montana
New Mexico
North Dakota
Ohio
Oklahoma
Pennsylvania
Tennessee
Texas
Utah
Virginia
Washington
Neat Virginia
Wyoming
Number of
Cleaning Plants
28
1
-
2
4
1
37
14
1
2
50
14
64
1
3
-
1
-
18
5
66
2
1
6
24
2
135
1
Mechanically
Cleaned
8,584
59
-
109
2,584
-
38.691
15,767
-
652
26.380
14,170
40.550
38
1.023
-
665
-
16,550
457
35,546
1.568
1,417
2,641
8.953
4,708
44.186
34
Crushed or
Screened
2.974
525
9.054
202
8.764
89
9,554
4,868
361
512
55,043
20.650
75.693
1.164
2,493
17.535
11,381
9,245
13,378
4,765
31,578
5.743
18,332
5.869
14.986
-
27.884
55,405
No
Processing
8,996
147
-
208
2,466
24
355
3,547
89
62
14,810
4.636
19,446
1,797
2,150
9,065
586
4,783
11,309
847
14,353
2,721
271
630
8.007
-
13,244
2.889
Total a/
Production
20,553
731
9,054
519
13,814
113
48,600
24,182
450
1,226
96.233
39,456
135.689
2,998
5,665
26,600
12,632
14.028
41,237
6,070
81,477
10,032
20.O20
9,141
31.946
4,708
85.314
58.328
Number of
Thermal Drying
Units
1
-
-
-
1
-
6
-
-
-
7
1
8
1
-
-
-
1
6
2
15
1
9
1
17
-
49
-
Tons
Thermally Dried
414
-
_
_
881
-
3.852
-
-
-
2.094
373
2,467
37
-
-
-
75
603
175
2,926
100
1,417
150
2.293
-
7,892
-
Total United States •/
419
224.780
332,353
107.994
665.127
•/ Data may not add to totals shown due to Independent rounding.
(Source: Reference 5).
118
23.282
-------
4.1.5 Performance
PCC will typically remove about 50 percent of the pyritic sulfur
present in coal, although the actual removal will depend on the washability
of the coal (the ratio of pyritic to organic sulfur), the size to which the
coal is crushed, the unit processes employed, and the densities of the
separating media. An analysis of levels and forms of sulfur found in
typical U.S. coals indicates that the high sulfur bituminous coals mined in
the northern appalachian and midwestern states typically contain up to 70
percent pyritic sulfur. As much as 70 percent of this sulfur can be removed
in cleaning processes which achieve a 90 percent recovery of the energy
content of the input coal. Coals from the southern appalachian and western
coal producing states more typically contain about 30 to 40 percent pyrite
sulfur. When these coals are cleaned by physical methods, total sulfur
reductions of about 20 to 30 percent (calculated on a lb/106 Btu basis) are
typically achieved.
4.2 COAL GASIFICATION
A number of commercially available coal gasification/gas purification
technologies have been proven to be capable of substantially reducing the
emissions of S0?, PM and NO that result from the direct combustion of coal.
£• A
At the present time, however, there is limited interest in the construction
of new gasification facilities to produce fuel gases for new industrial
boilers. The primary reason for this lack of interest is the current low
cost (relative to coal-derived gases) and high availability of natural gas
from conventional sources.
The key to the SO,, control capability of a coal gasification system is
the performance of the acid gas removal (AGR) unit of the gas purification
section of the plant. Industrial boiler fuel supply systems requiring an
AGR unit will generally not be cost competitive with conventional natural
gas, oil, or coal-fired boilers equipped with post-combustion controls.
This will limit most new gasifier applications to systems requiring a non-
interruptable gaseous fuel supply which do not have stringent product gas
sulfur specifications.
4-6
-------
4.2.1. Process Description
As shown in Figure 4.2-1, a complete gasification-based fuel production
system (including pollution control) consists of three steps: coal pretreat-
ment, coal gasification and gas purification. Coal pretreatment is
necessary to supply a feedstock with the proper physical and chemical
characteristics to the gasifier. In the gasification step, pretreated coal
is reacted with a steam/air or steam/oxygen mixture to produce a gas with a
heating value of approximately 5.6 MJ/Nm (150 Btu/scf) in the air-blown
(low Btu) case or 13 MJ/Nm (350 Btu/scf) in the oxygen-blown (medium-Btu)
case. In the gas purification step, particulate matter (including condensed
heavy hydrocarbons), sulfur and nitrogen species may be removed from the raw
product gas. The extent of gas purification required is determined by the
desired end use of the product gas and/or the applicable emission standards
for the end use combustion equipment.
4.2.2 Factors Affecting Performance
The most critical parts of a coal gasification system from the
standpoint of final fuel gas'specifications (which in turn determine the
ultimate emissions from any downstream process) are the gas scrubbing and
acid gas removal (AGR) operations. Removal of coal dust, ash, and tar
aerosols entrained in the raw product gas leaving the gasifier can be
accomplished with cyclones, or ESPs, or with water, oil or solvent
scrubbers. In the gas quenching and cooling section, tars and oils can be
condensed and particulates and other impurities, such as ammonia, sulfides
and cyanides can be scrubbed from the raw product gas.
Acid gases such as H2$, HCN, COS, C$2, mercaptans, and S02 are only
partially removed from a raw fuel gas in a simple gas quenching and cooling
section. For this reason, either low sulfur coal or an AGR system must be
used to significantly reduce the level of sulfur emissions in a boiler flue
gas stream. Commercially available AGR techniques include physical and
chemical solvent (absorption) processes, direct conversion, catalytic
conversion processes and fixed-bed adsorption processes. The specific gas
4-7
-------
Pretreatment
Gasification
Gas (and by-nroduct) Purification
Product Utilization
steam or water-
Coal
-P.
i
co
Coal Handling
and
Pretreatment
Coal Gasification
May Include:
Crushing
Sizing
Pulverizing
Transport
Storage
Drying
Partial Oxidation
Air (for
low Btu)
Air
•Ash
02 (for
medium Btu)
Crude by-products
(acid oases, r*<-
gas liquor) I
Hot Gas Cleanup
(Bulk Particulate
Removal)
COg to incineration
or tail gas treatment"
May Include:
Cyclone
ESP
L
Gas Purification
(tar, oil, sulfur,
nitrogen removal)
May Include:
Quenching/Scrubbing
Acid Gas Removal
Wastewater to POTII or
further treatment
sulfur-
NH ~
By Product
Recovery
Hot Particulate-Free Fuel
(To on-site combustor or
further cleanup)
Gas
I
• Needs for additional
I cleanup dictated by
combustor fuel specs.
or emission limits.
Cooled Desulfurized
Fuel Gas to Combustor
or Pipeline
Tars/oils to Fuel,
Upgrading or Sale
Figure 4.2-1.
Low/medium-Btu gasification process options
for supplying an industrial boiler fuel gas.
-------
cleanup process used will generally depend on the raw fuel gas pressure and
composition as well as the desired levels of contaminant removal.
Essentially complete removal of the particulate matter and reduced
nitrogen species present in the quenched raw product gas stream will be
achieved in most commercial AGR system. The level of reduced sulfur species
removal which is achieved will be dictated by the S02 emission limits of the
combustor. In synthesis gas applications, product gas specifications for
residual sulfur species concentrations are typically 1 ppm or less and these
levels have been achieved in commercial systems. Since combustion
applications do not usually require these stringent removal levels, so
• commercial AGR units for fuel gas production units can be designed for
almost any desired level of removal of the reduced sulfur present in the
quenched raw gas.
4.2.3 Applicability to Industrial Boilers
Low- and medium-Btu gasification systems are applicable to any
industrial boiler that can accept a gaseous fuel. Since low-Btu gas
combustion requires higher fuel flows and generates'higher flue gas volumes
than natural gas on an equivalent energy input basis, new boilers will have
to be equipped with slightly larger fuel and flue gas handling systems to
burn low-Btu gas. Otherwise, there are no technical obstacles to the use of
low- or medium-Btu gas as a boiler fuel.
In most cases, the use of low- or medium-Btu gas as an industrial
boiler fuel will be more costly than a direct-coal-fired unit equipped with
post combustion controls. Because of these economic considerations, most
future gasification system applications will not involve a dedicated fuel
gas production facility for a new industrial boiler. Future applications,
like current ones, are more likely to involve direct process heating where
(1) a clean gaseous fuel is required and (2) a non-interruptable supply of
natural gas is not available or cannot be guaranteed.
4-9
-------
4.2.4 Development status
There are only a limited number of coal gasifiers operating in the
United States on a commercial basis at the present time (see Table 4.2-1).
Most of these units are used to produce fuel gas for process heaters or
furnaces. Two of the units were designed to produce fuel gas for an
industrial boiler. These are the UMD/Foster Wheeler/Stoic and the Can-Do
facilities. It should be noted that both of these units were constructed
with significant support funds provided by the U.S. DOE.
Only two of the gasification systems listed in Table 4.2-1 are equipped
with gas cleanup systems that include acid gas removal units. Both of these
systems utilize Stretford AGR/sulfur recovery technology. The All is
Chalmers demonstration unit, which is designed to supply approximately 400 x
10 Btu/hr of low Btu fuel gas to a utility boiler, is now undergoing
startup. The Caterpillar Tractor AGR unit operated at reduced loads (less
than 40 percent of design) during the 1979-1982 time frame but this unit has
since been shut down indefinitely.
4.2.5 Reliability
Because of the limited commercial operating history of coal
gasification/gas purification systems in this country, there are no detailed
statistics on the frequency and severity of operating problems with these
units. The Caterpillar Tractor Stretford system apparently experienced no
significant operating problems during its two-plus years of operation, but
this system never operated above 40 percent of its design load.8 The
numerous gasifiers which are currently operating around the country have a
long history of reliable operation. However, this experience is not
necessarily applicable to units equipped with extensive gas cleanup
facilities.
4.2.6 Emissions Data
No certified test data for the Caterpillar Tractor AGR system are
available. This unit was not operated under any regulatory constraints and
there were no requirements for routinely reporting any fuel gas quality or
4-10
-------
TABLE 4.2-1. CURRENT APPLICATIONS OF LOW AND MEDIUM BTU GASIFICATION TECHNOLOGY
Facility
Gasifier and Coal Type
Extent of Product Gas
Clean-up and Fuel End Use
Commercial Units-Domestic
Holston Army Ammunition Plant
Kingsport, TN
Glen-Gery Brick Company;
Nine sites in Eastern, PA
National Lime and Stone, Co.
Cary, OH
Caterpillar Tractor
York, PA
University of Minnesota
Duluth, MN
Allis Chalmers
East Alton, IL
Can Do, Inc.
Hazel ton, PA
Howmet Aluminum
Lancaster, PA
Elgin-Butler Brick Co.
Austin, TX
Ten, air-blown Chapman gasifiers;
bituminous coal
Twelve, air-blown Well man Galusha
gasifiers, anthracite
Two, air-blown Wellman Galusha
gasifiers; bituminous coal
Two, air-blown, Wellman incandes-
cant gasifiers; bituminous coal
One, air-blown, Foster-Wheeler/
Stoic gasifier
One, air-blown, Kilngas gasifier
Two, air-blown Wellman Galusha
gasifiers; anthracite
One, air-blown Wellman Galusha
gasifier; anthracite
One air-blown SSF gasifier;
lignite
Hot cyclone; quenching; scrubbing; gas
used as fuel in process furnace; tar
burner in boiler.
Hot cyclone; hot gas used as brick
kiln fuel
Hot cyclone; hot gas used as lime kiln
fuel
Quenching; scrubbing; ESP; AGR
(Stretford); gas used as process
heater fuel.
Hot cyclone; ESP; quenching;
scrubbing; gas used as boiler fuel;
tars incinerated
Hot cyclone; quenching; scrubbing;
AGR (Stretford); gas used as utility
boiler fuel
Hot cyclone; cooling; gas used as
fuel for industrial park
Hot cyclone; gas used as process
heater fuel
Hot cyclone; gas used as brick
kiln fuel
-------
CN>
TABLE 4.2-1. CURRENT APPLICATIONS OF LOW AND MEDIUM BTU GASIFICATION TECHNOLOGY
(Continued)
Extent of Product Gas
Facility Gasifier and Coal Type Clean-up and Fuel End Use
Commercial Units - Foreign
Numerous foreign facilities Lurgi and Koppers-Totzek are the As needed to meet process require-
in Europe, Asia and Africa most widely used commercial systems; ments; synthesis gas and industrial
low rank coals generally used as fuel gas are the most common
feedstocks. applications.
Developmental Units - Domestic and Foreign
Numerous systems are under development both in the U.S. and abroad which offer the potential for improve-
ments in operating efficiency, reliability, fuel flexibility, environmental control effectiveness or
cost effectiveness relative to competitive technologies. These technologies have not reached a stage
of development which would be characterized as commercially demonstrated however:
BGC/Lurgi and GFETC slagging gasifiers, Westinghouse, U-gas, Pressurized Wellman Galusha (METC)
Exxon (Catalytic), Bigas, GEGAS, Shell (Koppers) and Texaco.
-------
combustor flue gas emissions data. The operators of this system claim,
however, that they had no problems in meeting the design outlet fuel gas
sulfur specification of less than 10 ppm total reduced sulfur.
4-13
-------
4.3 COAL-LIQUID MIXTURES
A coal-liquid mixture (CLM) is any blend of coal, liquid fuels (e.g.
fuel oil, methanol), water and additives (dispersants) that allows coal to be
handled as a liquid rather than as a solid fuel. The objective in using CLMs
is to substitute a less expensive, readily available solid fuel for a more
expensive, premium liquid fuel. With coal-water slurries (CMS), total
substitution of coal for oil is achieved whereas only partial substitution is
achieved with coal-oil (COM) or coal-oil-water (COW) mixtures. Because of
the economic advantages of a complete substitution for oil, recent interest
in the use of CWS has been increasing at the expense of COM and COW use.
The main applications of CLMs are expected to be in retrofits of
existing oil-fired boilers. In new applications, a conventional coal-fired
unit will generally be more cost effective. Incentives for converting
existing oil-fired boilers to CLM firing are provided by the lower cost of
coal on an equivalent energy input basis and concerns over the future
availability and price of premium liquid fuels. Another consideration is the
compatibility of CLM technology with deep coal cleaning methods (as discussed
in Section-4.1). Since both of these technologies require finely ground
coal, coal cleaning techniques which will improve the quality of the final
CLM blend can be easily and cost effectively integrated into a CLM
preparation plant.
In most applications, there will be no direct environmental benefits
associated with the use of CLMs. Uncontrolled PM and S02 emissions from
CLM-fired boilers are similar in character and present in quantities that are
predictable from the properties of the parent fuels. Uncontrolled emissions
of N0x with CWS-firing will be reduced due to the effect of water in lowering
the flame temperature. This benefit is not realized, however, when staged
combustion techniques are used for NO control. Any environmental benefits
A
derived from CLM-firing are associated more with the use of a cleaned coal or
an S02 adsorbent as a fuel additive than the use of a CLM directly.
4-14
-------
4.3.1 Process Description
Preparation of coal-oil mixtures (COM), coal-water slurries (CWS), or
coal-oil-water (COW) mixtures involves several steps. Coal pulverizing and
blending of the mixtures may be done either on-site or off-site depending on
the sizes of the units involved and a number of other site-specific factors.
In most small boilers, the CLM fuel would be prepared off-site in a large
centralized preparation plant and transported to the end-user in order to
realize the most favorable cost savings. A typical, large COM plant pro-
ducing 10,000 bpd of a 50/50 (wt./wt.) COM mixture would supply the fuel
input needs of approximately 750 MWt (2500 x 106 Btu/hr) of industrial boiler
capacity.
All coal-liquid mixtures are prepared by grinding the feed coal to a
very fine mesh size (usually to at least 70 percent through 200 mesh) prior
to preparing the final CLM blend. A finer grind provides better fuel
stability (less tendency for the coal particles to settle), better combustion
characteristics and reduced erosion problems in the fuel handling/ feeding
system. One utility boiler application in Florida, for example, has tested a
coal-oil mixture (COM) with a coal feed ground to 98 percent through 325
mesh. The disadvantages of a smaller grind size include: higher grinding
costs, a higher fuel viscosity (which will impact the design of the fuel
pumping, agitation and atomization systems) and potentially the generation of
finer fly ash particulates in the combustion flue gas.
The choice of a final blend mixture is dictated by a complex set of
site-specific constraints and economic trade offs. Generally the maximum
fuel savings is realized by maximizing the coal and minimizing the liquid
fuel content of the final blend mixture. With CWSs in particular, a high
coal content is required in order to maintain an acceptable furnace
efficiency. CWSs containing up to about 75 percent (wt.) coal have been
tested to date. COM blends are not limited by these same thermal
efficiency constraints and so mixtures containing as little as 10 percent
(wt.) coal have been tested.11 The maximum practical coal content of a COM
4-15
-------
is limited by coal handling pumping, erosion and viscosity concerns to about
50 percent (wt.) coal.
4.3.2 Factors Affecting Performance
The most important factors which affect the performance of a CLM-fired
boiler from an emissions point of view are the characteristics of the fuels
fired in the boiler and the capabilities of the control devices applied to
the unit. Generally, about 80 percent of the ash and 90-plus percent of the
sulfur present in a fuel will leave a boiler as flue gas particulate and S02
emissions respectively. Since most CLMs will have a higher ash content
and may have a higher sulfur content than the fuel oils they replace,
additional control equipment for both PM and S02 may be needed. Careful fuel
selection and blending or the use of fuel cleaning technologies upstream of
the CLM blending step could minimize or eliminate the need for additional S02
control equipment. However, most boilers converted from oil to CLM's will
need additional PM controls.
CLMs contain about 1 percent additives and stabilizers, which are often
alkaline compounds. Although these additives may reduce S02 slightly by
reacting with SOp to form sulfate and sulfate salts, their use increases flue
gas PM loadings and may contribute to increased furnace slagging, fouling and
refractory degradation problems. For these reasons, CLM producers are
examining the use of alternative additives such as ammonium-based compounds.
4.3.3 Applicability to industrial boilers
Almost any liquid fuel-fired boiler can be converted to burn CLMs. The
types of modifications that may be needed to accomplish this conversion
include: the addition of an agitator and possibly a heater to the liquid
fuel storage tank, additional liquid fuel pumping capacity (larger pumps and
possibly larger fuel supply lines), modified burners with special erosion
resistant tips, additional steam for soot blowing and fuel atomization,
modifications to the furnace bottom and ash handling system to accommodate
higher ash flows, and new or upgraded flue gas treatment equipment to
maintain compliance with applicable environmental regulations. Also, some
4-16
-------
derating of the boiler may be necessary in order to provide enough residence
time for the slower coal combustion reactions to occur. Because of these
considerations, the units which are best suited for a conversion to
CLM-firing are those which were originally designed for coal firing (with "V"
step bottoms, low plan area heat release rates, low furnace liberation rates
and adequate equipment for handling increased flue gas and ash loadings).
Typical conversion costs for units which are reasonable candidates for
CLM-firing were estimated to be in the range of $100-150/kW in one recently
published study, while costs for equivalent new coal-fired units were
determined to be about $500/kW.
4.3.4 Development Status
Most of the CLM development, testing and demonstration work which was
done in the late 1970's was focused on coal-oil mixtures rather than coal-
water slurries. By 1981, COMs containing up to 50 percent (wt.) coal had
been tested in a 400 MW utility boiler during a 1-year demonstration
13
program. According to another report, 21 units representing nearly
5000 MWg of electric utility generating capacity have- been converted to
CLM-capable units and another 10,000 MW of conversions are planned. No
equivalent statistics on industrial boiler conversions were found although
COMs have been tested in several industrial package boilers ranging in size
up to about 35 MWt (120 x 10 Btu/hr). Table 4.3-1 summarizes the recent
test experience with COMs in package watertube boilers. Current locations
and sizes of domestic COM preparation plants are shown in Table 4.3-2.
Based upon the above facts, COM preparation, handling, and combustion techno-
logy is considered to be comrnercially proven.
Coal-water slurries did not receive nearly as much attention as COMs
during the last 1970's. This was due primarily to concerns about the
feasibility, costs and impacts (e.g., derating) of converting existing
oil-fired boilers with limited ash handling and pollution control capabili-
ties to coal-only firing. However, recent design studies indicate that unit
deratings of only 3.5-5.5 percent are obtained when coal-water slurries
containing 65-75 percent (wt.) coal are fired in a furnace with an adequate
4-17
-------
TABLE 4.3-1. TEST EXPERIENCE WITH COM-FUELED PACKAGE WATERTUBE BOILERS
oo
/TkAimjuv
LOW AMI
GJI
Safclnav, HI.
PETC
Pittsburgh, FA
Island Creel.
Coal/Hooker
Chemical
White Springs. Fl
Uumko Products
Champagne, IL
EICON
Vlcksburg, MS
BOILER
MFC
C. E. Ulckes
120.000 Ib/hr
(preheater)
Nebraska
24,000 Ib/hr
B4W
120,000 Ib/hr
C. E. Wickes
40,000 Ib/hr
Superior
125.000 Ib/hr
TlfPE
A
D
«
A
U
FUEL
COM
35/50Z coa
COM
40X coal
COM
50Z coal
(COM Energy
COM
35X coal
(ERGON)
SOZ coal
(CoaliquiUa
COM
35Z coal
MODIFICATION
• Forney Verloop
burner
• Fuel storage
and handling
e Coen burner
e Fuel atorage
and handling
e Soot blower
in convection
paas
e Baghouae
e ID fan
e Modified burners
e Modified fuel
handling system
e Additional soot
blower
e Economizer
e Baghouae
• ID fan
e Howe Baker
turner
e Fuel handling
sy steal
• Modified nozsle
in Howe Baker
burner
• Fuel storage
and handling
OPERATING
EXPERIENCE
Phase I - 250 hrs. ,
35Z coal; Phase II.
494 hi-a., 50Z coal;
75Z naxlmusi load
500 hrs. over
approximately two
•onths; considera-
ble ash accumulation
in furnace
Short term tests
performed; evaluated
effect of particle sice
on burner erosion
Short term tests up to
full load satisfactory
.
Used to provide process
steam for fluid energy
•ill
STATUS
Completed 1977
COM tests completed
1981; currently
conducting CW tests
Long tern tests planned
Completed in 1978.
Initial operation
completed In 1981;
fuel development
continuing
Source: Reference
15
-------
TABLE 4.3-2. INSTALLED AND ANNOUNCED DOMESTIC COM PLANTS'
UNIT
INSTALLED
Nepsco
Coal Liquid
Florida Power & Light
Ergon
ANNOUNCED
Ashland Oil
COMCO
Island Creek Coal
Coal Liquid
Mt. Airy
Belcher
Wyatt
Amcom
Arco
a
LOCATION
Salem Harbor, MA
Shelbyville, KY
Sanford, FL
Vicksburg, MS
Southpoint, Oh
Bartow, FL
Jacksonville, FL
Jacksonville, FL
Dravosburg, PA
Mobile, AL
New Haven, CT
Chester, PA
West Virginia
CUMULATIVE
ACTUAL
1979
1980
FORECAST
1981
1982
STABILIZATION
Chem.
Cottell
Chem.
Ul trafine
Chem.
Ul trafine
Chem.
Cottell
Cottell/Chem
Chem./Mech.
Chem. /Fine
N/A
Chem.
COM CAPACITY -
3
18
25
49
BPD
2,000
1,500
10,000
5,000
1,200
3,000
3,000
6,000
1,500
5,000
3,000
3,000
5,000
BPD
,500
,500
,700
,200
CONSTR.
1978
1978
1979
1980
1980
1980
1981
1981
1981
1981
1981
1981
1981
OPER.
1979
1979
1980
1980
1981
1981
1981
1982
1982
1982
1982
1982
1982
Source: Reference 16
-------
combustion volume. Because of the feed coal size reduction requirements of
CWS, deep coal cleaning technology can be easily integrated into a CWS
preparation plant. This potential to generate a low ash, reduced sulfur,
liquid fuel through the combined use of coal cleaning and CWS technology is
one driving force behind much of the current CWS research, development and
commercialization activities.
CWS technology would best be characterized as near commercial at the
present time. The current U.S. CWS preparation plant capacity of only 40,000
tons/year will limit the number and scale of near-term CWS demonstration
1 ft
projects. However, several recent tests have demonstrated the feasibility
of CWS firing in commercial-scale equipment. A 75 percent (wt.) CWS has been
successfully test-fired in a 12 MVIt (40 x 106 Btu/hr) industrial boiler.19
Another report indicates that a 70 percent (wt.) CWS has been successfully
fired in a 23 MWt (80 x 106 Btu/hr) test burner.20 EPRI and DOE are jointly
sponsoring a CWS combustion test in a 65 MW. (225 x 106 Btu/hr) industrial
pi t
boiler in September, 1983.
To date, there has been almost no commercial interest in coal-alcohol
mixtures due to the high costs of fuel grade alcohols relative to those of
petroleum-based fuel oils. Coal alcohol mixtures containing up to 40 percent
(wt.) methanol have been burned successfully in a 1.5 MW. (5 x 106 Btu/hr)
pp t
industrial boiler.
4.3.5 Reliability
No data defining the operating histories of commercial-scale systems
firing CWS, COM, or COW mixtures have been published.
4.3.6 Emissions Data
Measurements of the emissions from a 65 MW, (225 x 106 Btu/hr) CWS-fired
industrial boiler in Memphis, Tennessee were conducted in September 1983.
Results from these tests will be available in early 1984.
4-20
-------
4.4 COAL LIQUEFACTION
Technical developments among coal liquefaction processes in the past
five years have occurred primarily at the pilot plant scale as no large
demonstration scale or commercial scale facilities have been constructed.
The major technical advances that have occurred are the addition of
two-stage liquefaction (TSL) to the SRC-I process and the use of solvent
deashing for the SRC-I and H-Coal processes.
No firm commitments have been made at this time for the construction of
a commercial-size coal liquefaction plant that could supply fuels to the
industrial boiler market, although a number of proposed plants are in the
advanced planning stages. Given the long construction and start-up lead
times for plants of this type, no significant volumes of coal-derived liquid
fuels will be available to industrial boiler owners in the next five years
and probably not in the next ten years.
Emissions data from test burns with coal-derived liquids indicate that
(1) S02 emissions depend on the sulfur content and heating value of the coal
liquid (which can be adjusted by varying the liquefaction process operating
conditions); (2) NOX emissions are higher than comparable petroleum-derived
fuels owing to the higher nitrogen content of coal liquids; and (3) uncon-
trolled PM emissions are comparable to petroleum-derived fuels but will
probably require control by fabric filter rather than ESP due to low ash
resistivity.
4.4.1 Process Description
As described in the Synthetic Fuels ITAR, coal liquefaction processes
can be divided into two general categories: direct and indirect.23 The
indirect processes, also known as catalytic synthesis, gasify coal to
generate a synthesis gas which is subsequently converted over a catalyst to
a wide variety of fuels. Since the catalytic synthesis process starts with
carbon monoxide and hydrogen, lower molecular weight products are favored
such as LPG, gasoline, and diesel oil. Economic considerations dictate
against the production of fuel oils that would be of interest to industrial
boiler owners. Moreover, no commercial indirect liquefaction plants are
4-21
-------
operating or under construction in the U.S. today. Tennessee Eastman will
use the process principles to produce acetic anhydride from coal at their
Kingsport, TN plant, scheduled to come on-line in Fall, 1983; large indirect
liquefaction plants are operated in South Africa to produce primarily motor
fuels. The indirect process will not be considered further as a source of
industrial boiler fuels.
Direct liquefaction processes fall into one of three categories:
carbonization, extraction, and hydrogenation. Very little development work
has occurred in the first two categories over the last five years and no
commercial plants are under serious consideration. Hydrogenation processes
do show some promise of eventually contributing to boiler fuel supplies.
Of the hydrogenation processes, the four which have reached the most
advanced state of development are the SRC-I, SRC-II, H-Coal, and Exxon Donor
Solvent (EDS) processes. The process descriptions provided in the
Synthetic Fuels ITAR are generally accurate with the following exceptions:24
" Use of TSL in the SRC-I Process - In an effort to increase the yield of
clean"premium fuels and the efficiency of hydrogen utilization, a
second stage of hydrogen processing has been added to the SRC-I
process. In the first stage, raw coal is converted into solvent
refined coal (SRC), distillates, and fuel gas. In the second stage,
expanded-bed catalytic hydrogenation is used to produce high quality
liquids and solids from a portion of the first-stage SRC.25 For the
6000 tons per day demonstration plant proposed for Newman, KY (see
Figure 4.4-1), one-third of the first-stage SRC will be solidified as
solid, another third will be feedstock for a delayed coker/calciner to
produce anode coke, and the final third will be treated in the
?fi
second-stage hydrocracker.
Use of Critical Solvent Deashing in the SRC-I Process - A second major
technical change to the SRC-I process is the use of the Kerr-McGee
Critical Solvent Deashing (CSD) process for solid-liquid separation in
place of filters. This process uses a deashing solvent to extract
4-22
-------
i
ro
oo
Coal
Coal
Preparation
Air
Separation
02
SRC
Vacuum
Distillation
KH Critical
Solvent Deashlng
Gaslfler
Hydrogen
Separation
Solidification
Hethan*tion
Syngas
Processing
Acid
Gas
Gas Processing
Product
Fractlonatlon
Coker/
Calciner
Expanded-Bed
Hydrocracker
Claus
Beavon
Gas
aw Naphtha
1st. Fuel
Oil
Two-Stage
-^•-liquefaction
SRC
-Sulfur
Aggregate
Figure 4.4-1. Flow Digram for SRC-I Demonstration Plant
-------
soluble coal liquids and reject the mineral matter and unconverted coal
near the critical point of the deashing solvent. Recovery of approxi-
mately 90% of the SRC has been demonstrated by the CSO process at the
Wilsonville pilot plant.
Use of Solvent Deashing in the H-Coal Process - The H-Coal process flow
diagram in Figure 2.3-2 of the Synthetic Fuels ITAR is significantly
out of date. The H-Coal process can be operated in two different
modes: the syncrude and the fuel oil modes. In the syncrude mode,
high yields of distillate liquids are achieved. Hydroclones are used
to reduce the solids content of the reactor effluent slurry. The
low-solids stream is recycled as slurry oil for feed coal; the
high-solids stream is fractionated to produce an all-distillate product
and a residuum stream which can be fed to a partial oxidation (i.e.,
gasification) unit to produce hydrogen or used as in-plant fuel. In
the fuel-oil mode, a heavier product slate is generated by operating
the reactor at less severe conditions. Heavy fuel oil will be
recovered using a solvent deashing technique such as the Kerr-McGee CSD
process described above or the Lummus anti-solvent deashing process.
The latter process uses a promoter liquid which causes precipitation of
heavy coal liquids on ash particles. Separation occurs as these
pQ
particles agglomerate and settle in a settler.
Use of Partial Oxidation in the EDS Process - The flow diagram for the
EDS process in Figure 2.3-2 of the Synthetic Fuels ITAR shows that
hydrogen is produced by steam reforming of the light hydrocarbon gases
from vacuum distillation. Fuel gas and liquid products are generated
by feeding the vacuum bottoms stream to a Flexicoking unit. An
alternative arrangement was investigated in a design study for a
commercial size EDS plant: the bottoms stream from the vacuum column
is split, with about one-half going to the Flexicoking unit and the
remainder converted to hydrogen in a partial oxidation (i.e.,
gasification) unit. Study results indicate that the alternative
4-24
-------
arrangement leads to a significant improvement in yield and plant
OQ
thermal efficiency and a slight reduction in capital investment.
- Use of Atmospheric and Vacuum Distillation for the SRC-II Process -
Figure 4.5-6 of the March 1982 BID indicates that liquid product from
the letdown and flash system of the SRC-II process is directed to a
vacuum column for solids removal followed by an atmospheric column for
separation of recycle solvent and liquid products. This is no longer
an accurate representation of the process as currently configured. The
process flow diagram for the proposed SRC-II demonstration plant (see
Figure 4.4-2) shows that reactor effluent flows through a series of
vapor-liquid separations where it is ultimately separated into process
gas, light hydrocarbon liquid, and product slurry. The product slurry
is split into two streams, the first being recycled to the process for
slurrying with feed coal and the second directed to a vacuum tower. In
the vacuum tower, a lighter distillate stream is removed overhead and
sent to fractionation; a heavier distillate product is removed as a
side stream, and the residue is sent to a gasification unit for
hydrogen production. The vacuum tower overhead, together with the
light hydrocarbon liquid from vapor-liquid separation, are sent to an
atmospheric fractionation tower to produce naphtha and a middle
distillate stream. Atmospheric tower bottoms are returned to the
30
vacuum tower.
4.4.2 Factors Affecting Performance
From the standpoint of New Source Performance Standards for industrial
boilers, the most important performance criteria for coal liquefaction
processes are the reduction of sulfur, nitrogen, and ash contents from
parent coals to product liquids and the combustion characteristics of
product liquids. In the liquefaction process, sulfur and nitrogen in parent
coal react with hydrogen to form hydrogen sulfide (H,,S) and ammonia (NH3),
respectively. Ash in the parent coal is removed via distillation and
solids-liquid separation techniques (e.g., hydroclones, filters, ad solvent
4-25
-------
Purified fly
-------
deashing processes). The Synfuels ITAR provides a thorough discussion of
the impacts of key process parameters on liquefaction product
characteristics. The principal impacts have been summarized in Section
4.5.5.2 of the BID.
One other parameter which has been found to be of importance to the
combustion properties of SRC is the solids-liquid separation scheme.
Combustion Engineering, in a study funded by EPRI, examined the combustion
characteristics of SRC produced by pressure filteration deashing (PFD),
anti-solvent deashing (ASD), and critical solvent deashing (CSD) under
combustion conditions similar to those achievable in boilers originally
designed for coal firing. The major conclusions drawn from the study
include :
From an overall combustion efficiency standpoint, both the CSD and PFD
SRC are relatively reactive solid fuels comparable in reactivity to
subbituminous coal. The ASD SRC is relatively unreactive in
comparison.
Compared to PFD and ASD, CSD SRC has the potential for producing a low
carbon (<10£) fly ash under low NO , staged combustion conditions if
A
flame temperature can be maintained sufficiently high during both fuel-
rich and fuel-lean'stages, thereby making the CSD fly ash amenable to
collection in electrostatic precipitators.
The SRC's, due to their relatively high fuel nitrogen contents, have a
high NOX formation potential under conventional firing conditions.
However, staging the combustion air can result in lower NO emissions
A
without jeopardizing their combustion efficiencies.
4.4.3 Applicability to Industrial Boilers
Commercial coal liquefaction facilities, if built, will produce fuels
in much larger quantities than are required by any one industrial boiler.
Therefore, the liquefaction plant would be considered an off-site supplier
of fuel.
4-27
-------
The solid fuel from the SRC-I process cannot be used in conventional
stoker boilers but can be used in pulverized coal-fired boilers with minor
modifications. Solid SRC cannot be used as speader stoker feed due to its
low melting point (approximately 155°C); the SRC would melt on the grate
before being combusted and fall into the plenum for removal with the ash.
Satisfactory combustion of solid SRC has been demonstrated for
pulverized coal-fired boilers with only minor modifications. Depending on
the site, these modifications may include the use of water-cooled burners,
addition of moisture prior to pulverization, or minor adjustments to
pulverizers. Combustion tests by DOE/PETC, discussed below, have shown that
SRC may be fired as a pulverized solid, a molten liquid, or as a slurry with
recycle process solvent.
The data presented in Table 2.3-6 of the Synthetic Fuels ITAR for solid
SRC and parent coals support the following comparisons:
SRC ash contents are significantly reduced from parent coal levels to
around 0.3 percent;
SRC heating values are about 25% greater than parent eastern coals and
50% greater than parent western coals;
The SRC sulfur content for eastern coals can be reduced to 0.7 to
1.0 percent under normal reactor conditions. Even lower sulfur content
SRC can be produced by increasing the severity of reactor operating
conditions. For western coals, SRC sulfur contents as low as
0.1 percent can be achieved under normal conditions. On a Ib SO-/106
Btu basis, these sulfur figures correspond to over 80% reduction for
eastern coals and almost 90% for western coals.
SRC fuels have slightly higher nitrogen and hydrogen contents than
parent coals and significantly lower oxygen contents.
Coal-derived liquids from the SRC-II, H-Coal, and EDS processes can be
substituted for petroleum-based fuels in oil-fired industrial boilers with
only minor modifications for coal liquid handling and storage. Studies by
Gulf on SRC-11 fuel oil showed satisfactory performance with respect to
viscosity, flash point, pour point, and stability. However, many elastomers
commonly used in fuel handling systems were destroyed by simple swelling
tests; viton and nylon 616 being the exceptions.
4-28
-------
The coal liquefaction processes cited above produce a variety of fuel
oil products with characteristics ranging from those of No. 2 fuel oil to
those of No. 6 fuel oil. The data summarized in Table 2.3-5 of the
Synthetic Fuels ITAR and Table 4.4-1 support the following general
observations: '
Coal liquid sulfur contents will range from 0.2 to 0.4 percent under
normal conditions compared to 0.04 to 0.5 percent for petroleum
products. As with solid SRC, the sulfur (and nitrogen) content of a
given coal liquid product can be reduced by adjusting the operating
conditions of the reactor and/or hydroprocessing operations;
The nitrogen content of SRC-II fuel oils is significantly higher (at
0.9 to 1.2 percent) than petroleum products (less than 0.3 percent);
nitrogen contents for H-Coal and EDS distillate are comparable to
petroleum products.
Heating values for coal liquids are slightly below those for petroleum
products and tend to increase with increasing process severity;
The oxygen contents of coal liquids (at 1-3 percent) are significantly
above those of petroleum products (at 0.01-0.4 percent);
Coal liquids are more aromatic in nature than petroleum products, which
is consistent with their lower hydrogen contents.
4.4.4 Development Status
At this time, no large demonstration-size or commercial-size coal
liquefaction plants are operating or under construction. As identified in
the Synthetic Fuels ITAR, the SRC-I process has been investigated at the
45 TPD pilot plant in Ft. Lewis, WA and the 6 TPD pilot plant in
Wilsonville, AL; the SRC-II process was developed at the Ft. Lewis pilot
plant; the EDS process has been under development at a 227 TPD pilot plant
in Baytown, TX; and the H-Coal process has been demonstrated at a 546 TPD
pilot plant in Catlettsburg, KY.36 At the present time, operations at all
of these pilot plants have been terminated with the exception of the
Wilsonville plant.
4-29
-------
TABLE 4.4-1 PROPERTIES OF SRC-II FUEL OILS AND COMPARABLE PETROLEUM PRODUCTS37
OJ
o
ANALYSIS (DRY): % BY WT.
CARBON
HYDROGEN
NITROGEN
SULFUR
OXYGEN
SATURATES: % BY VOL.
AROMATICS: % BY VOL.
DENSITY
VISCOSITY: SUS @ 100°F
FLASH POINT: °F
POUR POINT: °F
NICKEL: ppm
VANADIUM: ppm
SODIUM: ppm
SRC-II
MIDDLE DISTILLATE
(350°-550°F)
86.0
9.1
0.9
<0.2
3.6
35
62
0.974
36.3
>160
<-45
<0.1
<0.1
-
NO. 2 FUEL
OIL
87.0
12.8
<0.2
0.04-0.48
<0.09
>65
<32
<0.876
32.6-37.9
>130
<+5
<0.1
<0.1
<0.5
SRC-II
HEAVY DISTILLATE
89.1
7.5
1.2
0.37
1.4
-
-
1.072
231
-
<+45
<0.3
<0.1
2-11
NO. 5 FUEL
OIL
88.3
10.7
<0.3
0.07-1.9
<0.4
-
-
0.940
124-900
-
<+Rf)
46
180
2-20
-------
Commercial design studies have been completed for all four processes.
Various levels of detailed design have been completed for demonstration
plants (nominally 6000 TPD of coal feed) for the SRC-I and SRC-II processes;
no firm commitments are in place to construct and operate these plants due
to the withdrawal of support by the U.S.DOE and the lack of necessary
support from the private sector.
In view of the long lead times associated with the design,
construction, and start-up of plants of this size, it seems certain that
significant quantities of coal-derived liquid and solid fuels will not be
available to industrial boiler operators for the next five years, and
probably will not be available for the next ten years.
4.4.5 Reliability
To date, no commercial coal liquefaction plants have been built and
only limited combustion tests have been performed on the coal-derived
liquids. As a result, information regarding maintenance requirements and
the impact these coal-derived fuels would have on an industrial boiler are
not available. However, the impacts and maintenance requirements for coal
liquids-fired boilers should be similar to those of oil-fired boilers.
4.4.6 Emissions Data
The results of three major combustion tests performed with coal-derived
solid and liquid fuels are discussed in this section. These results are
primarily concerned with sulfur dioxide, nitrogen oxides, and particulate
matter emissions.
Plant Mitchell Tests on SRC-I - An 18-day test burn on solid SRC was
conducted in the 22.5 MWe Unit 1 boiler of Georgia Power Company's Plant
Mitchell near Albany, Georgia on June 14, 1977. Boiler modifications which
were made to accommodate the burning of SRC included:
Use of specially developed water-cooled dual register burners, and
Use of ambient primary air, reduced ball spring pressure, and variable
speed feeder motors in the pulverizers.
4-31
-------
The SRC fuel was produced at the Ft. Lewis pilot plant from
approximately 3.9 percent sulfur coal. As fired, the SRC had a heating
value of 15,274 BTU/lb, sulfur content of 0.71 percent, nitrogen content of
1.60 percent, and ash content of 0.57 percent. Boiler efficiency while
firing SRC was equivalent to that of coal-firing at full load and averaged
near 86%. Emissions results for S0? and NO are summarized in Table 4.4-2.
£ A
Two ESP's were operated in series after the boiler but the design for the
first precipitator was considered to be obsolete. The respective average
particulate emissions into the first precipitator, after the first
precipitator, and after the second precipitator were 1.0, 0.9, and
0.04 lb/106 BTU.
The Plant Mitchell tests demonstrated that SRC could be successfully
fired in a pulverized coal boiler and meet EPA emission requirements in
force in 1977. In addition, the SRC tests demonstrated overall low ash
loading and a non-abrasive ash which are expected to mitigate problems with
tube cutting and boiler deslagging and generally reduce maintenance on ash
hand!ing equipment.
Consolidated Edison Tests on SRC-II Fuel Oil - In September/October,
1978, a combustion demonstration test using SRC-II fuel oil was conducted on
a 450,000 Ib steam/hr utility boiler located at the 74th Street Generating
00 -3Q
Station of the Consolidated Edison Company of New York. ' The SRC fuel
oil was produced at the Ft. Lewis pilot plant from a variety of parent
coals. The heating value, sulfur content, nitrogen content, and ash content
of the fuel were 17,081 BTU/lb, 0.22, 1.02, and 0.02 percent, respectively.
The objectives of the test were to demonstrate combustion of SRC-II fuel
oil, to characterize NO emissions, and to investigate the potential to
A
reduce NO levels through combustion modifications. Major results of the
)\
test program can be summarized as follows:
No major operational problems were encountered due to combustion of
SRC-II fuel oil and performance on SRC-II fuel oil met all applicable
emission regulations;
4-32
-------
I
GO
OJ
TABLE 4.4-2 EMISSION RESULTS FOR SRC TEST BURN AT PLANT MITCHELL35
CONDITIONS
Low Load
Medium Load
Full Load
°2
(%)
11.0
7.3
5.6
so2
(ppm)
222
255
335
S62
(lb/10D BTU)
1.09
1.00
0.97
N0x
(lb/106 BTU)
0.43
0.45
0.40
-------
Boiler thermal efficiency levels with SRC-II fuel oil were comparable
to those with No. 6 fuel oil;
Nitrogen oxide emissions for SRC-II fuel oil at full load were
0.35 lb/106 BTU at baseline conditions and 0.23 lb/106 BTU at low NO
X
(staged combustion) conditions. NO emissions for SRC-II fuel oil were
A
nominally 70% greater than those for No. 6 fuel oil at both conditions.
This result was expected in view of SRC-II fuel oil's higher nitrogen
content.
Use of staged combustion reduced NO emissions on the order of 35% for
A
both SRC-II and No. 6 fuel oils.
Particulate emissions for SRC-II fuel oil were below 0.03 lb/105 Btu
under all test conditions and typically 40-60 percent lower than
equivalent emissions for No. 6 fuel oil.
It should be noted that the test boiler in this program was ideally
suited to take maximum advantage of the staging concept (i.e., a well-mixed
flame in the fuel rich zone, and adequate space for soot burn-out in the
40
fuel lean zone). The NO emissions level which can be anticipated with
A
other types of boilers with more intense flames is not certain.
DOE/PETC Tests on SRC - Tests were conducted with solid SRC using a
3450 Ib steam/hr firetube boiler, designed to burn No. 6 fuel oil, at the
U.S. Department of Energy's Pittsburgh Energy Technology Center (PETC).
The tests were designed to demonstrate the feasibility of using this fuel in
more compact oil and gas-fired units with higher heat release rates than the
coal-fired utility boiler of the Plant Mitchell test. The fuel was produced
at the SRC pilot plant in Wilsonville, AL from high-sulfur Kentucky coal.
The solid SRC had a heating value of 15,927 Btu/lb, sulfur content of
0.8 percent, nitrogen content of 2.0 percent, and ash content of
0.3 percent. The SRC was fed to the boiler in three different physical
forms: a slurry of 70 percent SRC-I process solvent and 30 percent
pulverized SRC; a molten liquid at approximately 600°F; and a solid,
pulverized to 90 percent minus 325 mesh. The major results of the program
are summarized below:
4-34
-------
Carbon conversion and boiler efficiencies for slurry and molten forms
were equivalent to those for No. 6 fuel oil (at 99.7 percent and
82 percent, respectively).
For pulverized SRC, boiler efficiency was the same but carbon
conversion efficiency was slightly reduced (98.6 to 99.6 percent);
pulverized SRC was burned at 50 percent load due to burner limitations.
Emissions results are summarized in Table 4.4-3. The data suggest that
SCL, NO , and particulate emissions are proportional to the sulfur,
£ A
nitrogen, and ash contents of the respective fuels.
Results indicate that SRC, including the solid form, can be burned in
larger oil-designed boilers of watertube design without significant
derating.
4-35
-------
-pi
I
OJ
cr>
TABLE 4.4-3 EMISSION RESULTS FOR DOE/PETC TESTS ON SRC 39
EMISSIONS
(lb/10° BTU)
so2
N0x
Particulate Matter
NO. 6a
FUEL OIL
0.628-0.671
0.223-0.265
0.139
SRC/SOLVENTa
SLURRY
0.537-0.693
0.668-0.850
0.122-0.214
MOLTEN3
SRC
0.953-1.085
0.669-0.772
0.184-0.849
PULVERIZED13
SRC
1.130-1.194
0.770-1.134
0.13-0.70
a Full Load
D Half Load
-------
4.5 REFERENCES
1. Buder, M.K., et al. (Bechtel National, Inc.), Impact of Coal Cleaning
on the Cost of New Coal-fired Power Generation. EPRI CS-1622.
March 1981. p. 4-9. 129.
2. Buroff, J., et al. (Versar, Inc) Technology Assessment Report for
Industrial Boiler Applications: Coal Cleaning and Low Sulfur Coal.
Publication No. EPA-600/7-79-178c. December 1979. p. 124.
3. Reference 1, p. 127.
4. Jung, A., et al., (Versar, Inc.) Determination of the Attenuation of
Sulfur Variability by Coal Preparation. Draft Report. EPA Contract
No. 68-02-2199. U.S. Environmental Protection Agency, Industrial
Environmental Research Laboratory. Research Triangle Park, N. C.
March 1981.
5. Bituminous Coal and Lignite Production and Mine Operations - 1978,
U.S. DOE/Energy Information Administration, Energy Data Report,
Publicatin No. DOE/EIA-0118(78), June 16, 1980. p. 55.
6. 1982 Keystone Coal Industry Manual. New York, McGraw-Hill Inc., 1982.
7. Cavallaro, J. A., et al. (U. S. Bureau of Mines). Sulfur Reduction
Potential of U. S. Coals: A Revised Report of Investigations,
BOM RIV 8118, Publication No. EPA/600/2-76-091. April 1976. pp. 296,
312, 314, 321, 322.
8. Letter from J. A. Maddox (Radian) to Stan Curtis (Caterpillar),
May 23, 1984.
9. Proceedings of the Fifth International Symposium on Coal Slurry
Combustion and Technology. Sponsored by the U. S. DOE (Pittsburgh
Energy Technology Center). Tampa, Fl. April 25-27, 1983. p. 314,
502-521, 580-587.
10. Reference 9, p. 502.
11. Proceedings of the Third International Symposium on Coal-Oil Mixture
Combustion. Sponsored by the U.S. DOE (Pittsburgh Energy Technology
Center). Orlando, Fl. April 1-3, 1981. p. 114, 171-183.
12. Reference 11, pp. 171-183.
13. Proceedings of the Ninth Annual International Conference on Coal
Gasification, Liquefaction and Conversion to Electricity. Sponsored by
the University of Pittsburgh and the U. S. DOE (Pittsburgh Energy
Technology Center). Pittsburg, PA. August 3-5, 1982. p. 210.
4-37
-------
14. Reference 9, p. 580-587.
15. Foo, 0. K., et al. An Assessment of Package Boilers for Industrial
Coal-Liquid Mixture Applications. (Presented at the Fourth
International Symposium on Coal Slurry Combustion and Technology.
Orlando, Florida.) May 10-12, 1982.
16. Hawkins, G. T. (Coal Liquid, Inc.). Industrial Utilization of Coal-Oil
Mixtures, (Presented at Coal Technology '81 Conference. Houston,
Texas. November 17-19, 1981.) pp. 215-224.
17. Synfuels, March 11, 1983, p. 4.
18. Synfuels, March 18, 1983, p. 4.
19. Reference 9, pp. 502-521.
20. Chemical Engineering, April 18, 1983, p. 27, 29.
21. Chemical Engineering, June 27, 1983, p. 15.
22. Pan, Y. S. et al. Coal-Methanol Mixture Combustion Tests with Coals of
Different Ranks. (Presented at the Fifth International Symposium on
Coal Slurry Combustion and Technology. Tampa, FL. April 25-27, 1983.)
23. Thomas, W. C. (Radian Corporation.) Technology Assessment Report for
Industrial Boiler Applications: Synthetic Fuels. (Prepared for
U.S. Environmental Protection Agency.) Research Triangle Park,
North Carolina. Publication No. EPA-600/7-79-178d. November 1979.
pp. 2-38 - 2-50, 2-54.
24. Reference 23, pp. 2-38-2-50.
25. Smith, M. R., D. A. Hubbard, and C. C. Yang (Kellogg-Rust Synfuels,
Inc.). Logic Technology and Effect of Coal Liquefaction Conditions on
Final Up-Graded Product State. Proceedings of the Ninth Annual
Conference on Coal Gasification, Liquefaction and Conversion to Energy.
Pittsburgh, PA. August 3-5, 1982. p. 440, 438.
26. Tao, J. C. and J. P. Jones (International Coal Refining Co.). SRC-I
Technology and Its Economics. Energy Progress. 2_(l}:2. March 1983.
27. Reference 25, p. 438.
28. Eccles, R. M. and G. R. DeVaux (Hydrocarbon Research Inc.). Current
Status of H-Coal Commercialization. Chemical Engineering Progress.
May 1981. pp. 80-85.
4-38
-------
29. Epperly, W. R. and D. T. Wade (Exxon Research and Engineering Co.).
Donor Solvent Coal Liquefaction. Chemical Engineering Progress.
77_(5) :73-79. May 1981.
30. Free!, J., D. M. Jackson and B. K. Schmid (SRC International, Inc.).
The SRC-II Demonstration Project. Chemical Engineering Progress.
77_(5):86-91. May 1981.
31. Reference 23, p. 2-54.
32. Goetz, G. J., T. C. Lao, A. K. Mehta, and N. Y. Nsakala (Combustion
Engineering, Inc.). Effect of Liquefaction Processing Conditions on
Combustion Characteristics of Solvent-Refined Coal. Interim Report.
(Prepared for Electric Power Research Institute. Palo Alto,
California.) Publication Number AP-2328. March 1982. pp. S-l - S-7.
33. Free!, J. and D. M. Jackson (SRC International, Inc.) SRC-II Product
Applications-Recent Developments. Proceedings of the Eight Annual
International Conference on Coal Gasification, Liquefaction and
Conversion to Electricity. Pittsburgh, PA. August 4-6, 1981.
pp. 231-233, 241, 242-243, 232.
34. Reference 23, p. 2-54.
35. Reference 33, p. 241.
36. Reference 33, p. 2-42 - 2-43.
37. McRanie, R. D. (Southern Company Services, Inc.). Full-Scale Utility
Boiler Test With Solvent Refined Coal (SRC). Interim Report. Work
Performed Under U.S. Department of Energy Contract No. EX-76-C-01-2222.
April 1978. pp. 1-15.
38. Piper, B. F., S. Hersh, and W. Nazimowitz (KVB Incorporated).
Combustion Demonstration of SRC-II Fuel Oil in a Tangentially Fired
Boiler. Final Report. (Prepared for Electric Power Research
Institute. Palo Alto, California. May 1979.) pp. 7-1 - 7-22.
39. Hersh, S., et. al. (KVB Incorporated). Emissions Characterization of
SRC-II Fuel Oil Fired in a Utility Boiler. Proceedings of the
Governor's Conference on Expanding Use of Coal in New York State:
Problems and Issues. 1981. pp. 323-329.
40. Reference 33, p. 232.
4-30
-------
41. Pan, Y. S., et. al., (U.S. Department of Energy). Final Test Report on
the Combustion of Solvent Refined Coal in a 100-HP Firetube Boiler.
(Prepared by Pittsburgh Energy Technology Center, Pittsburgh, PA.)'
Publication No. DOE/PETC/TR-82/5 (DE 82007670). pp. 17-22.
4-40
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing}
1. REPORT NO. 2.
EPA-450/3-85-009
4. TITLE AND SUBTITLE
Industrial Boiler S02 Technology Update Report
7, AUTHOR(S)
Ed Aul , Suzanne Margerum, Robert Berry, et al .
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Radian Corporation
3200 E. Chapel Hill Road/Nelson Highway
Research Triangle Park, North Carolina 27709
12. SPONSORING AGENCY NAME AND ADDRESS
Office of Air Quality Planning and Standards
U. S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
3. RECIPIENT'S ACCESSION NO.
5. REPORT DATE
6. PERFORMING OJ^/WMIJ QBfcpN CODE
8. PERFORMING ORGANIZATION REPORT NC
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-02-3816
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
EPA/200/04
15. SUPPLEMENTARY NOTES
Project Officer - Dale Pahl , OAQPS/ESED/MD-13
16. ABSTRACT
Tins document is a resource document for the development of Federal standards
of performance for control of sulfur dioxide (SOg) emissions for new industrial
boilers. Various precombustion, combustion modification, and post combustion
control technologies are identified with respect to each technology's applicability
to industrial boilers, development status, and factors affecting performance
Emissions data for each technology are also presented. Post-combustion technologies
examined include wet flue gas desulfurization (FGD) systems (sodium, dual alkali
lime, limestone) and dry processes (spray drying FGD, dry alkali injection
electron-beam irradiation). Combustion modification approaches examined include
fluidized bed combustion, limestone injection multistage burners, and coal/1imestone
pellets. Precombustion approaches include physical coal cleaning, coal gasification
coal-liquid mixtures, and coal liquefaction.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Held/Group
S02 Emissions
Coal Air Pollution
Industrial Boilers
Pollution Control Technology
Fuel Standards
Coal Cleaning
Flue Gas Desulfurization
Coal
Air Pollution Control
18. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
172
20.
22. PRICE
EPA Form 2220-1 (Rev. 4—77) PREVIOUS EDITION is OBSOLETE
------- |