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

-------